Frontispiece: Potential energy field of morphine surface using a proton probe
at its van der Waals
Opiates George
R. Lenz
Suzanne M. Evans
Health Care Research and Development British Oxygen Corporation Group Murray Hill, New Providence. New Jersey
Department of Medicinal Chemistry and Pharmacognosy University of Illinois at Chicago Chicago, Illinois
D. Eric Walters
A. J. Hopfinger
Molecular Design Group NutraSweet Research amI Development Skokie, Illinois
Department of Medicinal Chemistry and Pharmacognosy University of Illinois at Chicago Chicago. Illinois
With a chapter by Donna L. Hammond Section of Central Nervous System Diseases G. D. Searle & Company Skokie, Illinois
1986
~
ACADEMIC
PRESS, INC.
Harcourt Brace Jovanovich, Publishers Orlando Boston
San Diego New York Austin London Sydney Tokyo Toronto
Contents
@ [986 BY ACADEMIC PRESS, INC. ~PYR[GHT ALL RIGHTS RESERVED. NO PART OF THIS PUBL[CATION MAY BE REPRODUCED OR TRANSMI1TED [N ANY FORM OR BY ANY MEANS, ELECTRON[C OR MECHAN[CAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY [NFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISS[ON IN WRITING FROM THE PUBLISHER.
ACADEMIC Orlando,
PRESS, INC.
Oval
Road,
London
LIBRARY OF CONCRESS
NWI
ix
1 Morphine
f10rida 32887
United Kingdom Edition published by ACADEMIC PRESS INC. (LONDON) 24-28
Preface
I. Introduction II. The Biosynthesis References
LTD.
under
CATALOCING-iN-PUBLICATION
DATA
:2
Biological
Effects
Donna
I. II. III. IV. V. VI. VII.
.
[ncludes index. ]. 'LNarcolics.-Models. 2. Narcotics-Structureactivity relationships. 3. Chemistry, Pharmaceutical. I. Lenz, George R. IDNLM: I. Narcotics. QV 89 06131 RS431.N370/i5 1986 615'.7822 85.1997:i [SBN 0-12-443830.X (alk. paper)
3
Introduction Multiplicity of Opiate Receptors Analgesia Respiratory Depression Gastrointestinal Motility Dependence Liability Summary References
Synthesis
and
of Morphine,
IN THE UNITED STATES OF AMERICA
86871!81!9
of Opioids
L. Hammond)
title:
Opiates.
PRINTED
I 7 24
and Metabolism of Morphine
7DX
(by Main entry
and Its Analogs
29 31 35 37 38 39 40 41
Structure-Activity Codeine,
and
Related
Relationships Alkaloids
I. Syntheses of Morphine, Codeine, and Related Alkaloids II. The Structure-Activity Relationships of Morphine and Related Compounds III. Diels-Alder Adducts of Thebaine IV. The Chemical Anatomy of Morphine and Its Derivatives References
987654321 \I
45 55 101 155 157
Contents
vi
4
Physical Chemistry,
Molecular
QSAR Analysis of the Morphine, and Benzomorphan Analgesics
Modeling,
166 174 185
9
188
10
II. III. IV. V. VI.
Analgesics
I. Methadone and Related Compounds II. Other Open-Chain Compounds References
Naturally Occurring Morphinans Conversion of Morphine and Its Analogs to Morphinans The Total Synthesis of Morphinans Structure-Activity Relationships of the Morphinans The Chemical Anatomy of the Morphinans References
189 190 193 206 242 243
400 435 445
Piperidine I. II. III. IV. V. VI.
Chemistry
and Molecular
Modeling
Analgesics
I. Physical Chemistry Studies of Open-Chain Analgesics II. Molecular Modeling of Open-Chain Analgesics References
6 The Benzomorphans I. Introduction II. Benzomorphan Syntheses III. Structure-Activity Relationships in the Benzomorphan Analgesics IV. The Chemical Anatomy of the Benzomorphans References
Physical
of Open-Chain
11
8
Open-Chain
388 398
The Morphinans I. Introduction
7
vii
III. Molecular Modeling and Quantitative Structure-Activity Relationship (QSAR) Studies References
and
Morphinan,
I. Physicochemical Studies II. Molecular Modeling and QSAR Studies References
5
Contents
Enkephalins I. II. III. IV. V. VI. VII. VIII.
250 252 259 310 311
448 456 457
Introduction Opioid Peptide Precursors Peptide Synthesis Enkephalin Selectivities for the J1.and 0 Opiate Receptors Minimum Enkephalin Chain Length Necessary for Analgesia Structure-Activity Relationships in the Enkephalins Clinically Investigated Enkephalin Analgesics The Chemical Anatomy of the Enkephalins References
459 463 471 473 48] 482 500 502 503
Analgesics 318 319 331 334 352 362 367
Introduction Meperidine Family Bemidone Family Prodine Family Alkyl Family Anilino Family References
Physical Chemistry, and QSAR Analysis
Molecular Modeling, of the Arylpiperidine 377 385
Physical Chemistry of the Enkephalins I. II. III. IV. V.
index
Analgesics I. Physicochemical Studies II. Stereostructure, Conformation, and Biological Activity
12
and Molecular
Introduction Solid-State Conformations Solution Conformations Molecular Modeling Studies QSAR Studies References
Modeling 513 514 516 532 537 537 543
Preface
Research in the pharmaceutical sciences is becoming increasingly interdisciplinary. The days of organic and medicinal chemists and pharmacologists being the only members of the preclinical research team are gone. The trend in research is to study chemical and biological events on the molecular level as well as to work in the more traditional domain of animal pharmacology. This has resulted in the addition of new members to the research team. Experts in molecular spectroscopy and physical chemistry are aiding in the interpretation of structure-activity data. In many cases the biology is divided up among animal pharmacologists, molecular pharmacologists, biochemists, and molecular biologists. The newest member of the research team, the "drug designer," uses the computer to establish predictive criteria relating the physicochemical properties of molecules to their observed biological endpoints. Unfortunately, preclinical research monographs in the pharmaceutical sciences are not usually coauthored by all members of today's preclinical research team. Generally, the result is a discussion deep in some topics and shallow in others. Further, the integration and continuity of the component topics are often fragmented and incomplete. Often the reader is forced to scan through a set of reference books to assemble a comprehensive overview. We believe that a unique aspect to this book on the opiates is that experts in each of the major components-synthetic chemistry, medicinal chemistry, pharmacology, physical chemistry, and drug design-have teamed together to generate a complete text. Thus the discussion of the opiates is uniformly complete and integrated across all subdisciplines. The structure of the chapters reflects the inclusion of detailed reviews of each of the various subdisciplines. Discussions of organic syntheses and reporting of structure-activity relationships predominate throughout the book, reflecting our greater knowledge of certain topics than of others. The book is organized around simplification of the rigid molecular framework inherent in morphine. Chapter I describes the history, biosynthesis, and metabolism of the naturally occurring morphine. A special ix
x
Preface
Preface
effort has been made in Chapter 3 to put together a complete, but also concise, summary of the enormous amount of work done on morphine, codeine, and related alkaloids. Cleaving the dihydrofuran ring in morphine yields the analgesic morphinans (Chapter 5). Continuing simplification by scission of the C-ring in the morphinans forms the benzomorphans (Chapter 6), an area of continuing research interest. Further bond breaking leads to the arylpiperidines (Chapter 7), where analgesics as potent as the thebaine Diels-Alder adducts have been observed. The seemingly ultimate simplification results in the open-chain analgesics (Chapter 9), where an aromatic ring is joined to an amino by a flexible chain. Chapter 11 describes the endogenous ligands for the opiate receptors, the enkephalins, and the biosynthesis and SAR investigations into this fascinating class of peptide analgesics. It is interesting that out of the thousands of enkephalin analogs prepared, only three have made it to clinical investigation and these seem to have undesirable clinical profiles. In Chapter 2, Donna Hammond contributes an invited discussion on the biological effects of opioids. The style of this chapter is such that it is easily read by a synthetic chemist who has a minimal background in opioid biology. The discussions (Chapters 4, 8, 10, and 12) of the physical chemistry, molecular modeling, and QSAR investigations of the various classes of opioids following the medicinal chemistry chapters are unique entries to a treatise in this field. The book leaves several major questions unanswered. Such is to be expected from a text on a dynamic research field. It is the nature of the beast. Nevertheless, we collectively wish that we could provide more insight into, for example, the active conformation and shape of the enkephalins, the common three-dimensional pharmacophore among opiates, and the physicochemical properties governing opiate receptor specificity. Nevertheless, we feel that the most current information on these and other pressing questions is provided to the reader. Lastly, many people besides ourselves are responsible for this book becoming a reality. Professor H. A. Scheraga of Cornell University first suggested to one of us, AJH, the need for a book of this type. At that time all four of us were members of Research and Development at G. D. Searle & Co. of Skokie, Illinois. Interestingly, as this book goes to press, none of us is now a member of Searle R & D. While senior R & D management at Searle did not go out of the way to encourage us on this project, they also did not discourage us and allowed us to use company clerical services to generate a working manuscript. A number of people at Searle helped in the preparation of the manuscript. However, Ms. Sue Christain was of key importance in generating both text and structures. She is our silent, fifth author. Other
xi
Searle personnel who unselfishly gave us their time and skills are Ms. Grace Koek, Ms. Dolores Weiman, and Ms. Linda Tepper. This has been a long and arduous project for all of us. However, it is an undertaking that will provide a common bond to sustain our friendship for years to corne. J'
GEORGE SUZANNE
R. LENZ
M. EVANS
D. ERIC WALTERS A. J. HOPFINGER
1. Morphine
and Its Analogs
J. Introduction A. History. B. Occurrence . C. Production and Use . . .. . II. The Biosynthesis and Metaholism of Morphine
A. MorphineBiosynthesis . . . . . . . ..
.
...
B. Morphine Alkaloid Biotransformation in Animal Species C. Biotransformation in Papaver D. Morphine Disposition
References
I.
. . . . .
Species
I I 2 5 7 7 13 19 20 24
Introduction
A. History The development of the first effective analgesic drug, opium, was almost certainly accidental and occurred in prehistory. When the unripe seed capsule of the opium poppy, Papaver somniferum L., is incised, a viscous milky fluid is exuded. As this exudate is exposed to air, it dries and darkens to a hard, slightly sticky mass known as opium. The potent biological effects of opium were recognized in ancient times, and for many millenia this substance has been used by the practicing physician. Its sedative and euphoric properties have also caused opium to have a long folkloric history. The ancient Egyptians knew of its properties, and it has been variously smoked or eaten and ingested as its alcoholic tincture, laudanum. It has found use as a poison (1), but its major use has been for the relief of pain. Morphine, the major active ingredient of opium, is used today as an analgesic in controlling severe pain despite the development of more potent and efficacius opiates. The intense development of analgesics on the basis of the morphine framework might not have occurred as readily were it not for the numerous other biological activities inherent in the morphine molecule. Deleterious side effects include respiratory depression, constipation, and marked sedation. Morphine also acts as a euphoriant while at the same time causing addiction. These last effects, coupled with the tolerance that develops to it, make morphine a readily abusable substance (2). The separation of the analgesic effect from the others has occupied medicinal chemists for many decades.
2
1
Morphine
and Its Analogs
Although opium contains a variety of alkaloids, its major. constituent and the most potent analgesic is morphine (1) (3). MorphIne has the distinction of being the first nitrogenous base to be isolated from a living source and is one of the most intensively investigated and intriguing materials in the history of chemistry (4). The isolation of an opium constituent in crystalline form was first achieved in 1803 by Derosne, an apothecary living in Paris (5). He ~iluted a syrupy extract of opium with water and precipitated the salt of opIUm wIth potassIUm carbonate. SeguIn, in 1804, presented a paper to the Institute of France entitled "Sur I' opium," in which he described the isolation of morphine from opi~m (6). Morphine was isolated as a crystalline substance In 1806 by Fnednch Wilhelm Sertiirner, an apothecary, in the city of Paderborn in the Kingdom of Hanover (7). It is Sertiirner who is usually credited with the. discovery and isolation of morphine. Johann Bartholomaus Trommsdorff, editor of Trommsdorff's Journal of Pharmacy, where Sertiirner's work appeared, was moved to remark (8): Dee
Versuch
def
das chemische iiber
das Opium
uDd
es
werden
ist
herm sind,
vielmehr
moehte,
enthalten
viel dank so darf Man
zu
wunschen,
urn manche
setzeD. Vorzuglich wiederholt
Prof.
Publikum
wiinschte
ooeh
manche
schuldig die Asten das
sehr
interreste
ist. So vielfach noch
dieser
obwaltenende
leh, das die Versuche
kleineswegs Gegenstand Dunkelheiten mit
etwas
aber
Ansichten,
wofOr
nun 3uch
die Arbeiten
als geschloBen ooeh
weiter
in ein helleres
groBen Mengen
ihm
ansehen, untersucht Licht
zu
mochtern
werden.
As a result of this entreaty, suppliers in various obscure corners of the Near East and Far East hastened to comply with this dictum, and have supplied increasing amounts of morphine and its derivatives ever since. The scientists involved in the study of morphine read like a Who's Who of nineteenth-century chemistry: Liebig, Knorr, Wieland, Pschorr, Gadamer, and many others. Despite the plethora of experimental <;>bservations, the correct structure for morphine (1) was not postulated until 1925 by Sir Robert Robertson (9). The absolute configuration was not determined until the mid-1950s (10), shortly before the total synthesis of morphine was reported, over 150 years after its isolation (lla). B. Occurrence The opium derived from Papaver somniferum contains at least 50 alkaloids, with the major constituent being morphine. The alkaloids contained and the percentage of occurrence of the major alkaloids are presented in Table I-I. The opium alkaloids are derived biogenetically from I-benzylisoquinolines, which, in turn, are derived ultimately from phenylalanine. The more complicated alkaloids, with regard to their therapeutic uses, can be classified structurally either as (a) those containing a reduced benzylisoquinoline group in the form of a hydrophen-
Introduction
)
Table I-I Alkaloids
Found in Opium
a
6-Acetonyldihydrosanguinarine {3-Allocryptopine Berberine
Narceine (0.2%) Narceine imide Narcotine
Canadine
Narcotoline
Codamine Codeine (0.5%) Codeine N-oxides Codeinone
Normorphine
Coptisine
Orientaline
Neopine Nornarceine Norsanguinarine
Coreximine
13-0xocryptopine
Corytuberine
Oxydimorphine
Cryptopine
Oxysanguinarine
Dihydroprotopine
Pacodine
Dihydrosanguinarine
Palaudine
Glaucine Gnoscopine Hydrocotarnine 10- H ydroxycodeine 16-Hydroxythebaine (+ )-Isoboldine (- )-Isocorypalmine
Papaveraldine Papaveramine Papaverine
From Santavy
C
Porphyroxine
Protopine Pseudomorphine
Lanthopine Laudanidine Laudanine Laudanosine Magnoflorine 6-Methylcodeine N-Methyl-14-0-desmethylepiporphyroxine Morphine (10-20%) Morphine N-oxides a
(1%)
Papaverrubine
(z)-
Reticuline
Salutaridine Salutaridinol-I Sanguinarine ( )-Scoulerine Stepholidine Tetrahydropapaverine Thebaine
(0.3%)
Thebaine
N-oxides
(11b,llc).
anthrene nucleus or (b) those containing an intact, albeit modified, benzylisoquinoline nucleus. The most important hydrophenanthrenebased alkaloids are morphine (I), codeine (2), and thebaine (3). Benzylisoquinoline-based alkaloids are exemplified by papaverine (4) and dlnarcotine (5). Despite their common precursor, it is perhaps not surprising that the hydrophenanthrene and the other benzylisoquinoline alkaloids have radically different biological profiles. The phthalideisoquinoline alkaloid dlnarcotine (5) is used primarily as an antitussive, while the closely related bicuculline is a potent antagonist of the central nervous system (CNS) neurotransmitter y-aminobutyric acid. Papaverine (4) is a smooth muscle relaxant with little CNS activity. Codeine (2), a centrally acting analgesic,
4
1
Morphine
and Its Analogs
I
Introduction
5
although weaker than morphine, is characterized by its oral activity and is used extensively as an antitussive. Thebaine (3), on the other hand, is very toxic and produces strychnine-like convulsions. Thebaine is in demand, however, as an intermediate for the preparation of the highly potent 14-hydroxymorphinan derivatives and as a point of entry for the synthesis of compounds derived from Diels-Alder additions to its cyclic diene system. In contrast to its low content in the opium poppy, it is the main alkaloid in another species of poppy, Papaver bracteatum, which does not contain morphine (12). Although thebaine is readily' convertible to codeine, it has to be subsequently de methylated to produce morphine. One means of reducing illegal opium traffic would be to curtail the cultivation of the opium poppy, P. somniferum. This would restrict the availability of alkaloid raw material that is readily convertible to heroin (diacetylmorphine) while allowing the cultivation of P. bracteatum in order to extract thebaine and convert it to pharmaceutically acceptable products. See Gordon's review (2) for an account of the economics of manufacture of and illicit dealings in opium.
Morphine
C. Production and Use
~O
OR 2 Codeine
4 Papaverine
o
3 Thebaine
5 dJ-Narcotine OCR)
Scheme }./. Derivation of the reduced and nonreduced benzylisoquinolines common precursor.
from a
Although opium's most famous alkaloid is morphine, the most widely utilized drug for the relief of mild to moderate pain and as a cough suppressant is codeine (2). Codeine accounts for about 90% of U.S. consumption of opium derivatives (13). The majority of opium available for export in the global market comes from India and, to a lesser extent, Turkey. The Soviet Union, which has a significant opium-producing capacity, consumes most of its production. The global and U.S. consumption of codeine from 1970 to 1974 and the projections from 1975 on are shown in Table 1-2. As the. table shows, in contrast to relatively stable global consumption, the United States had an average increase of over 10% per year in the reported period. In the preceding decade, consumption was relatively stable. A crisis occurred in 1972-1973 that threatened to produce a shortage of opium in the United States. This crisis was due to a number of factors: (a) partial crop failures in India substantially reducing the raw opium supplies, (b) a total ban on opium poppy growth, starting in 1973, by the Turkish government under U.S. government pressure, and (c) the Soviet Union becoming a net importer of opium for the first time. Because of the threatened shortage at the consumer level, the U.S. government was forced to release portions of its strategic stockpile reserves of opium to domestic producers in order to meet domestic requirements. This crisis forced both U.S. and international assessment of the required amounts and suppliers of codeine. It also stimulated interest in the
j 6
1 Morphine and Its Analogs Table 1-2 Codeine Consumption
in the United
Year
(metric
Worldwide
1970 1971 1972 1973 1974 1975 1978 1980 {I
h
States and Worldwide"
tons)
United
158 150 156 163 155 166" 177" 186"
States (metric
23 26.5 28.5 33 34.9 37.5" 45" 50.1 h
From Schwartz (/3). Projections.
Table 1-3 U.S. Drug Enforcement Administration Production Quotas for 1985 Basic Class (Schedule II) Alphaprodinc Codeine (for sale) Codeine (for conversion) Dextropropoxyphene Dihydrocodeinc Diphcnoxylate Fentanyl ydrocodone " Hydromorphone Levorphanol Meperidine Methadone Mix~d Alkaloids of Opium Morphine (for sale) Morphine (for conversion) Opium (tinctures, extracts)" Oxycodone (for sale) Oxycodone (for conversion) Oxymorphone Sufenlanil Thebaine Grams of anhydrous base. b" Grams of powdered opium.
Analgesic
Proposed 1985 Quota 37.300" 54.051,000 3,534.000 75,795,000 1,341,000 550,000 3,500 1,459.(kJO 164,lk~) 21,750 7, 999
,I~k)
tons)
II
The Biosynthesis and Metabolism
of Morphine
7
practical production of opium derivatives by total synthesis. The Turkish ban on opium poppy crops was rescinded, and alternative sources of codeine were investigaled. The most important result was the intensive investigation of thebaine (3), derived from P. bracteatum. The dried latex from this poppy contains up 10 55% thebaine (14), which is readily convertible into codeine while avoiding morphine as an intermediate. This species of poppy has the potential to be a commercial crop in the United States. However, commercial production of P. bracteatum has not been allowed by the federal government, in part due to political considerations concerning the anticipated loss of foreign exchange to the less developed (LDC) producing countries, India and Turkey. Additionally, there is substantial non medicinal economic usage of Papaver species in the LDC areas (15). The poppy nevertheless retains a potential domestic replacement for foreign sources. Each year the U.S. Drug Enforcement Administration sets aggregate production quolas for schedule I and II controlled substances. This information is published at the end of the preceding year in the Federal Register. The aggregate production quotas for 1985 in grams of anhydrous base for the various analgesics are presented in Table 1-3,
II. A.
The Biosynthesis and Metabolism of Morphine Morphine Biosynthesis
The biosynthetic sequence for morphine (1), the major alkaloid of the opium poppy, Papaver somniferum, has been validated through the radiotracer work of various groups, following a significant structural suggestion made by Gulland and Robinson in 1925 (9). The el~boration of the later stages of biogenesis considered the structural resemblance of the morphine skeleton to an unsubstituted 1-benzylisoquinoline (6) which can be ob. tained by breaking bonds A and B. Conversely, a benzylisoquinoline can
1,383,IXkJ 22,300 1,142,00) 58,084.000 2,068,000 1,966,000 6.41JO 5,00) 500 6.890,IJOO
CH3 I
CH3 I
N
N
~
, I
0...
1 Morphine
'- --
6
B
OH / 7 Laudanine
8
I
Morphine
and Its Analogs
serve as precursor to the aporphine skeleton by another type of ring closure. In fact, Robinson suggested that laudanine (7) might be a biosynthetic precursor to morphine and its relatives (16). Thus, early work defined how norlaudanosoline (16), the first 1benzylisoquinoline recognized along the pathway, was elaborated into morphine via the key intermediates salutaridine (20) and thebaine (22). Recent work on the early stages of morphine alkaloid biosynthesis has focused on the formation of the benzylisoquinoline system itself in attempts to identify the species forming the two halves of the molecule, whose origin is the naturally occurring amino acid tyrosine. The biosynthesis of morphine (llb,c,17a-j) proceeds along the pathway shown in Schemes 1-2 and 1-3. Two molecules of the amino acid tyrosine (II) (18) form the basic I-benzyltetrahydroisoquinoline skeleton, with dopamine (10) serving to elaborate one half (19) of this skeleton, ring A with the ethylamine side chain. A L-dopa decarboxylase, isolated in Papaver orientale (20) latex, can effect the necessary decarboxylation of dopa to givedopamine. Dopa, the transamination product of 13, however, is incorporated into only one C-6-C-2 unit of the key intermediate (14) and the eventual product reticuline (18), whereas tyrosine, which occurs naturally in P. somniferum, via 4-hydroxyphenylpyruvic acid (12) and 3,4-dihydroxyphenylpyruvic acid (B), is incorporated into both C-6-C-2 units, the phenethylamine and benzylic portions (21). However, the C-l oftyrosine is specifically the source ofthe carboxyl group in the key intermediate, norlaudanosoline-l-carboxylic acid (14) (22). Decarboxylation of 14 gives the dihydroisoquinoline (15), a known precursor of morphine (23) and the immediate precursor of norlaudanosoline (16) (24) in the Papaver species. Confirmation of this pathway has been demonstrated not only in P. somniferum plants and P. orientale seedlings and latex (25), but also in cell-free systems of P. somniferum stems, seed capsules, and other plant parts, wherein the intermediates 14, IS, and 16 have been formed from dopamine and 3,4-dihydroxyphenylpyruvic acid (26). Dihydroxylation of both aromatic halves is necessary before joining to give 14, as shown by the fact that the intermediate (25) is a poor precursor for morphine (25).
II
The Biosynthesis
and Metabolism
9
of Morphine
H DOPA
i
o~
CO2" HO
HO
14
11 Tyrosine
/HO 13 CO2"
I
HO 12
HO HO
<---
H
<---
HO
16 (~)-Norlaudanosoline
17 (+)-Norlaudanosoline -dimethyl ether
15
'\
18
OH
(~) -Reticuline Scheme ].2.
HO 25
Norlaudanosoline biosynthesis from tyrosine
precursors.
10
18
1
(!-Reticuline)
1
20 (+)-Salutaridine
Morphine
21
i
and Its Analogs
Salutaridinol
j
II
The Biosynthesis and Metabolism of Morphine
II
o
I
18 1
19
23
Neopinone
/
22
26 Diradical
Reticuline
Thebaine 21
22 Scheme
24
2 Codeine
1
Morphine
Codeinone Scheme 1-3.
20 Salutaridine
(ertha-para)
Salutaridinol
Thebaine
-
II
Morphine biosynthesis from norlaudanosoline.
Conversion of norlaudanosoline to reticuline (27) involves both N- and a-methylation. Feeding incorporation studies have shown that 0methylation, with no selectivity for position, precedes N-methylation, giving (:t)-norlaudanosoline dimethyl ether (17) from (:t)norlaudanosoline (16). N-Methylation as the terminal step results in (:t )-reticuline (18). In all cases, the methyl group is invariably derived from the S-methyl of methionine and plays an important role in the further elaboration of the opiate alkaloids. Both isomers of reticuline (28), (+ )-reticuline and (- )-reticuline, have been found to be incorporated into the morphine alkaloids, with a loss of tritium from C-I of the (+ i-isomer, supporting the conclusion that a rapid equilibration (reversible oxidation-reduction) occurs through the 1,2-
}.4.
I
Conversion of reticuline to thebaine.
dihydroreticuline intermediate (19) (29). This racemization is enzymic and substrate specific, with the evidence for 19 being confirmed by observation. (:t )-Reticuline has indeed been isolated from opIUm, provldmg evidence that the racemate is a naturally occurnng alkalOId (30). In fact, (+ )-reticuline exists in excess in mature poppies ~f P. somniferu.m, however the ( - i-isomer is drawn from the interconvertmg set. ConversIOn of (- )-reticuline to the dienone salutaridine (20) (31), a~ shown m Scheme 1-4, occurs directly by an oxidative coupling (32) [the dlradlcal (26) bemg formed from oxidation of the two phenolic hydroxyl groups] as postulated by Barton and Cohen for phenols (33). One additional ri?gclosure needed to give thebaine (22), the first hydrophenanthrene alkalOId m the pathway, is then accomplished by loss of water from the mtermedIate alcohol salutaridinoll (21) (34), whose hydroxyl group is on the same side of the ring system as the C-15-C-16 bridge. The other isomer wIth opposite stereochemistry at C-7, salutaridinollJ, does not serve as a precursor for thebaine, thus showing the spatial arrangement of the phenohc hydroxy and allylic alcohol necessary for ring closure. Biotransformation of thebaine to codeine (2) (35), then, occurs through two intermediate ketones, neopinone (23) and codeinone (24) (36), by initial conversion of the enol ether, not by reduction of thebaine to codeine
12
Morphine
II
and Its Analogs
The Biosynthesis
and Metabolism
t3
of Morphine
II
Thus, in the later irreversible sequence of steps, two O-demethylations (40) are crucial to the biotransformation of thebaine to morphine. Also crucial to the biosynthetic pathway are the specific methylating steps (41). In the sequence from reticuline to thebaine, the methylation of particular hydroxyl groups is critical for directing cyclization, by providing a means of protecting and thereby inactivating phenolic groups in a specific manner. N-Methylation is helpful for the oxidative coupling (42) step; the location of the methoxyls on the rings, however, determines the sites for coupling. Complete methylation, as in tetrahydropapaverine (28), prevents this essential coupling. Differences in methylating systems may thus
oxygenase [0]
22
Thebaine
CH30 1
1
28
27 Oripavine
23
Neopinone 1
Further Scheme 1-5.
Tetrahydropapaverine
account for the fact that various plant species show differing patterns of alkaloid biosynthesis. Evidence for this is found in the Papaver species P. bracteatum, where a lack of de methylating and hydrogenating capabilities results in thebaine being the final product of opiate alkaloid biosynthesis, with no conversion of thebaine to codeine and morphine (39).
biosynthesis
Conversion of thebaine to morphine alkaloids.
methyl ether. Precursor feeding studies in P. somniferum using labeled thebame have shown that this enol ether cleavage involves the loss of the 6-0-.methyl group with retention of the 6-oxygen, as shown in Scheme 1-5, possIbly thro~gh a mechanism involving an oxygenase (37), as is known in other aromatIc ether cleavages. In species such as P. bracteatum and P. orientale, whi~h lack enzymes for this C-6 demethylation, biosynthesis stops a~ thebame and branches to oripavine (27) by demethylation of the phenohc ether at C-3 (38). Migration of the double bond into conjugation produces codeinone (24) from neopmone (23). Codeinone is then reduced to codeine (2), with both the addItIon and removal of hydrogen at C-7 in the later two steps being nonspecIfic. O-DemethylatJOn then occurs to give morphine (1) (39).
B.
Morphine Alkaloid Biotransformation
in Animal Species
The biotransformation (43) of morphine, codeine, and heroin has been studied in vivo in humans as well as in various animal species, namely, the rat, guinea pig, rabbit, cat, and dog. Morphine and codeine typically undergo alkylation, dealkylation, and oxidation reactions (heroin, in addition, undergoes hydrolysis), followed by conjugation prior to elimination. In vitro studies by incubation with isolated enzyme preparations (rat liver and brain, guinea pig liver) have further substantiated the formation of these metabolites. Their importance lies in the biological consequences of analgesia, with biotransformation being the key determinant of drug onset, duration of activity, and potency. The major metabolic pathways for morphine (I), codeine (2), and heroin (29), although relatively species dependent for quantitation, are shown in
14
1
Morphine
and Its Analogs
~"o
0"
~
HO'C
HO
H
H
,fl.
0"
I.,
R''''H 34 R':CH.35
R'O
0" R':H
''0 38
Ft'''CH. 39 A'O
0'R'''H 36 R'''CH,37
Scheme 1~6. Major metabolic pathways of morphine alkaloids. Key' (a) h d I'
N-dea~kYlatlOn. re d uctton.
(c) conjugation,
(d) O-dealkylation,
b « »
( I ) O:i~~tr~~' , g
(e) Q-alkylation'
II The Biosynthesis and Metabolism of Morphine
'
Scheme 1-6 (44). Thus, morphine forms metabolites primarily by: I. conjugation: glucuronidation (43b) at C-3 and C-6 2. conJugatIOn: ethereal sulfate formation at C-3 3. N-demethylationjN-oxidation 4. O-methylation and O-demethylation Conjugation, the ~ost common form of detoxification, encompasses bo th the glucuromdatlOn and sulfate formation reactions. Glucuronidation usually. serves as the major biotransformation pathway for compounds contammg a phenolic or alcoholic hydroxyl group and produces a watersoluble compound that is then excreted easily in bile or urine. The sum of
OH
1\
morphine conjugates generally accounts for 55-65% of the administered dose. The major metabolite in humans, rats, and dogs (45), formed by the conjugation of morphine with glucuronic acid mainly in the liver, and to a lesser extent in the intestine, kidney, and placenta, is morphine-3glucuronide (31) (46), shown to be zwitterionic (pKa values = 3.2, 8.1) (47). Morphine-6-glucuronide (32) and the 3,6-glucuronide form in very small amounts in humans (totaling only 1% of the 3-glucuronide) (46a); however, they have been easily identified in other mammalian species. Although the 3-glucuronide is much less active than morphine, the 6-glucuronide produces a 37-fold increase in analgesic activity and has a prolonged duration compared to morphine sulfate (48). In fact, most manipulations of functionality at the C-6 position of morphine have generated compounds displaying greater analgesic potency than morphine itself. In the reverse reaction, the glucuronide conjugate at position-3 hydrolyzes more readily than that at position-6. Morphine-3-sulfate (MES) (33), the major conjugate and urinary metabolite in cats and chickens due to a deficiency of glucuronyl transferase, is the second major metabolite in humans, being formed in a ratio to morphine-3-glucuronide of I: 4 and accounting for 5-10% of an administered dose (49). This biotransformation again occurs in the liver, the source of the sulfate group thought to be the active intermediate adenosine-3' -phosphate-5-pyrophosphate sulfate, transferred to the phenolic hydroxy by a mechanism analogous to the formation of sulfated metabolites of hydroxylated steroids (50). Other reported metabolites occur in such small quantities in most species that they generally do not contribute appreciably to the pharmacological activities observed for morphine, even if they do possess intrinsic agonist opiate-like analgesic properties. An exception to this generalization occurs for metabolism at the key target sites of brain regions having opiate receptors (99). The enzyme N-demethylase (51), which catalyzes the oxidation of the methyl group to an alcohol and then to an aldehyde, has as its end product normorphine (34) (52), which accounts for 3-5% of an administered dose (53). Reconversion to morphine has been demonstrated both in vitro and in vivo in rats (54). The methyl donor is believed to be S-adenosyl-L-methionine, as in O-methylation reactions. Studies have shown that norcodeine (35), in addition to normorphine (55), and being formed either from normorphine directly or from codeine itself, is formed in humans (56) and rats (57a). The mechanism (58) again is N-dealkylation by microsomal oxidation in hepatic endoplasmic reticulum (P-450 mixed function oxidase) and brain tissue, but its maximal rate does not correlate simply with the lipophilicity of the substrate (59). These N-dealkylases in the liver and brain are different in that the brain enzyme shows high
16
1
Morphine
and Its Analogs
II
The Biosynthesis
and Metabolism
17
of Morphine
Ii
stereoselectivity and displays specificity for agonists. Here, metabolism is essential for the expression of morphine's opiate activity since greater biological efficacy is attained with increased N-demethylation of morphine. Pure narcotic antagonists are not metabolized to the norcompounds in the brain, although they are effectively so transformed in the liver (99). Thus, contrary to early reports, humans do form norcompounds as metabolites, even though in only small quantities. Both normorphine and norcodeine have been shown to be equiactive in analgesic receptor. binding affinity compared to their parents (guinea pig ileum) (60), although these are exceptions to the generally observed phenomenon of reduced agonist activity on N-dealkylation (61). In various other morphine derivatives, the N-methyl group has been shown to be important for agonism in vitro, the effect being more striking in vivo (mice). N-Oxide formation of morphine (MNO) (36) and codeine (CNO) (37), catalyzed by amine oxidase (51), is an alternative oxidation biotransformation in guinea pigs and humans. It gives products of undetermined stereochemistry that arise from oxidation of the tertiary amine, which is catalyzed by amine oxidase in the liver. MNO has weak analgesic activity, bemg only one-tenth as potent as morphine. Although it has been shown to be a normal metabolite of morphine, once it is formed (in amounts less than 2% of free morphine), its only metabolic pathway in vivo has been demonstrated in rats to be reduction back to morphine, a reaction that may not involve the liver microsomal system. Thus, this metabolic process is reversible, producing as a metabolite of MNO morphine itself, which is actually thought to be responsible for the observed activity of MNO (62). The ~xidation product hydromorphone [dihydromorphinone (38)], a thIrd oXIdatIOn product, has been reported as a metabolite in rats, guinea pigS, ~abbits, monkeys, and morphine-dependent humans (46a,53,55). ThIs bIOtransformatIOn accounts for 4% of a dose given to rats and 6-7% of a dose given to monkeys (52d). Hydromorphone itself is a potent narcotic analgesic (both orally and parenterally, with a potency greater than that of morphine) whose metabolic profile (Scheme 1-7) includes C-6 reduced a- and/or I3-hydroxy metabolites, with the later formation being favored in all species studied, including humans (63). However, a- and f3-dihydromorphine (42, 43) are only approximately one-fifth to two-thirds as p~tent as hydromorphone, their contribution to the pharmacological actIVIty of the morphine parent thus being negligible (potency: hydromorphone > 6a-OH > 6f3-0H metabolite). Morphine has been shown to undergo O-methylation (64) to codeine in humans, as well as in rats (53) and dogs, via S-adenosyl-L-methionine methyl group transfer. Codeine and its metabolites are thought to account
Rl
R2
R3
42
8
H
08
43
8
08
8
44
C83
8
08
45
CH3
08
H
Compound
Scheme ]-7.
C-6 metabolites of hydromorphone
and hydrocodone.
for up to 10% of an administered dose (65). O-dealkylation regenerates morphine. Hydrocodone (39), formed by C-6 oxidation and reductIOn of the 7,8-double bond in codeine (humans, guinea pigs, dogs), is a more potent analgesic then codeine itself. Its greater bioavailability and ~ore extensive O-demethylation to hydromorphone, combmed with ItS mablllty to form C-6 conjugates, are thought to be responsible for its greater potency. Changes in the oxidation state of the functional group at C-6 of hydrocodone, however, produce metabolites with only a minimal contribu~ion to the observed analgesic activity. As with morphine, the greatest activity hes in the 6-keto compound, followed by the 6-hydroxy epimers (Scheme 1-7), with the 6a- (44) having only one-seventh the potency of the 613- (45) (57). All three oxidized metabolites nonetheless have a potency greater than that of codeine itself. The major metabolic pathways for the opiate analgesics account for 75-85% of an administered dose. The remaining percentage (51) mcludes extremely small amounts of conjugation, oxidation, and reduction products. These are a- and f3-dihydromorphine (51,66) (guinea pig) synthesized by reduction of the ketone in hydromorphone or the 7,8-double bond in morphine, a- and y-isomorphines (guinea pig, rat, dog, rabbit, humans), hydroxylated morphines, and all of their metabolites. These products,
18
19
II The Biosynthesis and Metabolism of Morphine
1 Morphine and Its Analogs II
however, do not contribute significantly to the observed pharmacological profile of the parent compound or the major metabolites (67). Heroin [Scheme 1-6 (29)], 3,6-diacetylmorphine, in addition to the above metabolic pathways available upon morphine formation, in humans and other animal species studied, undergoes biotransformation by hydrolysis (68). The first, rapid deacetylation yields 6-0-monoacetylmorphine (6-MAM) (30), whose observed pharmacological activity is equipotent to morphine (69). The identity of the enzymes responsible for this hydrolysis has not been established. In plasma, on the basis of in vitro experiments, serum cholinesterase is believed to be responsible; however, human serum cannot hydrolyze 6-MAM to morphine (70). This second, slower deacetylation releases morphine, the major urinary metabolite, which accounts for 50-60% of an administered dose. Nearly all tissues can hydrolyze 6-MAM to morphine (71), but studies have shown that liver tissue exhibits the greatest ability and brain tissue the least (72), furthering the belief that 6-MAM is an important mediator of the pharmacological response to heroin. Heroin thus acts as a pro-drug, providing concentrations of active metabolites 6-MAM and morphine to the systemic circulation, as determined by route of administration (100). The secondary amine metabolites, namely, norheroin (41) and 6-acetylnormorphine (40), are 20 times less potent (mice) than their N-methyl derivatives, and both resemble morphine rather than heroin in their onset, peak, and duration of activity (73). The narcotic antagonists naloxone (46) and naltrexone (48), although displaying species differences in metabolic disposition, produce 6 ,,_ hydroxyl and/or 6j3-hydroxyl metabolites as a result of the reduction of the 6-keto group (Scheme 1-8) (74). In humans and rabbits, for naloxone, the conjugation product naloxone-3-glucuronide (47) is the primary metabolite, although in the chicken the major metabolite is N-allyl-14-hydroxyl7,8-dihydronormorphine-3-glucuronide (49), the 6,,-hydroxyl reduction product. Naltrexone in humans yields a metabolite with the 6j3-hydroxyl configuration, N-cyclopropylmeth yl-14- hydroxyl-7 ,8-dih ydronorisomorphine (52), while in the chicken the metabolite is again the 6,,-epimer (51). Little of the 6j3-hydroxyl naloxone metabolite (SO) is produced in humans. The antagonist (naloxone, naltrexone) potency profiles, as well as agonist (hydromorphone, hydrocodone) profiles, have been shown to be sensitive to functional group changes at the C-6 position. Both antagonists display a separation of potencies in the 6,,- and 6j3-0H metabolites when compared to the ketone parents. The relative order for analgesic antagonism is naloxone/naltrexone > 6,,> 6j3-0H metabolites. In both antagonist series, the hydroxy metabolites are of lesser potency and slower onset of action than the ketones, so that their actual contribution to the antagonist activity of the parent compounds is doubtful (74).
46
(R=B)
47
(R=
48
~ 2H
OB
BO
OH
R
Compound
Rl
R2
49
allyl
OB
B
50
allyl
U
ou
51
cyclopropylmethylene
OB
B
52
cyclopropylmethylene
B
OB
Scheme ].8.
C.
Biotransformation
Metabolites
of naloxone
and naltrexone.
in Papaver Species
Metabolites of the Papaver species h~ve been shown to be. both alkaloidal and non alkaloidal (75). The major degradatlve pathway IS the initial rapid demethylation to normorphine (34), a. step shown to be irreversible occurring solely in stem latex, and makmg both the actIve synthesis a~d degradation rates for morphine very high (76)..This rap~d turnover has led to the belief that morphine plays an actIve role m metabolism, acting perhaps as a specific methylating agent. Subsequent degradation of normorphine yields large nonalkalOldal compounds (77) that account for 80% of morphine metabohtes (78).
20
Morphine
and Its Analogs
II
The Biosynthesis
and Metabolism
of Morphine
V3 Peripheral tissue (low blood flow)
V2 Peripheral tissue
53
54
VI Morphine
Central compartment
;,
Fig. 1-1.
I
I I
55 Scheme 1-9.
Polar metabolites
of Papaver
56
(R=H)
57
(R=CH3)
species:
the N-oxides.
Other polar metabolites (Scheme I-9) in P. somniferum have been character~zed as the two ISOl~eric N-oxides of morphine (55, 56) and the one N-oxIde of code me m WhIChthe N+-O- bond is axial (57) (79). Both oXIdes of thebame (53, 54) have been isolated as natural metabolites in P
bracteatum. D.
Morphine Disposition
1. Eli'!'ination from Plasma Previous confusion about the disposition of morphme resulted mamly from the different analytical methodologies and protocols used for measurement (80). Recently, however, the Issue has been clanfied by the use not only of radioimmunoassay (RIA) (81) but also of the newly developed, more precise, and more sensitive gas-lIqUId .chromatography (GLC) technique for detection (82). The pharmacokmehc profile of morphine following intravenous administration
2t
.
Three-compartment
model
of morphine
Metabolized and excreted drug
disposition
in humans.
has been represented (83) by a three-compartment model (Fig. I-I) (83a) that coordinates the kinetics of the drug in various tissues and the kinetics of the pharmacological effects produced. The central compartment (V,) is the recipient of the drug and the other two compartments (V2, V3) are perfused peripheral tissue; V3, however, comprises tissues with low blood flow (84). Morphine is eliminated from the body primarily by hepatic biotransformation via the central compartment. The major mechanism, then, for removal of morphine from plasma is the formation of water-soluble glucuronides in the liver; the major elimination route is glomerular filtration and/or tubular secretion in the kidney. Clearance of morphine depends partly on the pH of tubular urine. Elimination via bile into the duodenum and feces is small (generally less than 3%), although studies vary widely in quantitizing this pathway. Free morphine may be reabsorbed, however, by hydrolysis of the conjugates (about 20% of the dose in rats) via j3-glucuronidase activity of both the intestinal flora and mucosa (85). This hydrolysis is a prerequisite for enterohepatic recycling. Although urine and feces remain the main routes of excretion, saliva has also been reported. Clearance of morphine from plasma depends on both the distribution and elimination processes. Distribution has been shown to be widespread. Saturable uptake, however, has been shown only by the kidney and brain, although the amount of drug accumulated in these tissues is small compared to that found in skeletal muscle. Subsequent elimination is rapid, the rate-limiting step for this process being hepatic blood flow. Hepatic extraction and biotransformation are thought to be essentially complete (83b); therefore, liver clearance is perfusion limited.
22
1
Morphine
and Its Analogs
Elimination consists of three distinct phases within a 6-hour period, after which time 66% of the dose has been excreted in the urine (83). The first (1T)and second (a) phases represent the distribution to various tissues and organs. Studies in which RIA and solvent extraction methods have been used have demonstrated that typically, within 10 minutes, 97% of a dose is cleared from plasma and distributed. Concentrations of conjugated drug exceed those offree drug by this time. By 1-1.5 hr, metabolites account for more than 90% of the total morphine in plasma. The third ({3) terminal elimination phase consistently ranges from 1.5 to 2.5 hours, independent of dose, although the elimination half-life for the glucuronide is longer (86). Terminal elimination is actually biphasic: from 6 to 48 hours, the terminal elimination half-life has been shown to be 10 hours in rats (86,87). Hepatic morphine clearance parallels hepatic blood flow, which is responsible for regulating reuptake from peripheral tissues and delivery to the liver. The rate of elimination, k8, is greater than the rate of uptake from peripheral tissues with low blood flow, kA, in the compartment model. Although 80% (parenteral administration) of a dose is excreted within 8 hours and 90% after 24 hours, small amounts are still being excreted 72 hours later. This long half-life represents the fraction of morphine coming out of the tissues and being reabsorbed from enterohepatic circulation (88). The long-term appearance of morphine in plasma may be due to slow leaching from high-affinity binding sites like brain (86). Morphine undergoes a high first-pass effect (88,89), which explains its poor oral efficacy, since most of the drug never appears in the plasma but is extracted/metabolized in the intestinal walls and in the liver before concentrating in the central plasma compartment (90). The significance of metabolism in the gastrointestinal tract seems to depend on lipophilicity, which affects the efficiency of gut UDP-glucuronyl-transferase activity, as evidenced by the observed small first-pass intestinal metabolism of the hydrophilic dihydromorphine (91). Across various species, free morphine in urine accounts for 2-12% of an administered dose in a 24-hour period and bound morphine for 20-42% when parenterally administered. Urinary excretion of morphine and its conjugates accounts for 70-83% of a dose; 11-14% is accounted for by biliary secretion (92). 2. Dependence on the Route of Administration Morphine metabolism is largely dependent on the route of administration, and therefore also on the rates of absorption and distribution (93). While plasma levels are determined by the route of administration, plasma half-lives are not. In
II
The Biosynthesis
and Metabolism
of Morphine
2J
humans, various routes have been employed, the major four being intravenous (preferred), intramuscular, and subcutaneous injection and oral administration (94). Studies on the analgesia produced by each route and the accompanying kinetics have focused on the plasma levels of drugs (94) and on renal clearance (95). However, no correlation between plasma levels and intensity of analgesia has been demonstrated. Morphine distribution has been shown in animal species such as the rat, dog, and monkey in tissues such as kidney, lung, liver, spleen, and muscle. A major portion of morphine is in body muscle; only small amounts cross the blood-brain barrier. Studies aimed at quantitating the tissue distribution of drug have been limited mainly to rat brain. Although studies have indicated a direct connection between plasma and brain, the equilibration is slow due to the multicompartmentalization of morphine (96). There is no specific localization of morphine in the brain (97). Decline of analgesia occurs with a decrease in morphine concentrations in plasma and cerebrospinal fluid (CSF), wherein uptake is slower. Elimination from kidney is prolonged compared to plasma pharmacokinetics. RIA studies (94) in humans have shown that the most rapid absorption of morphine occurs by the intravenous (iv) injection route (87b,98). Drug is delivered directly into the plasma, producing high free morphine plasma levels within 15 min, but plateauing over the next 12 hr. The observed maximum plasma level corresponds well to the reported time of maximum pharmacological effect, that is, minutes after injection. Low concentrations of the drug persist in the plasma for up to 48 hours; this is thought to be due to an albumin-binding phenomenon. Second to the preferred iv route is the intramuscular (im) injection route (98). Absorbed both by passive diffusion and by active fluid transport, giving a systemic availability of 100%, the drug shows a maximum biological effect in addition to a maximum serum concentration 60-90 minutes after administration. Equally efficacious is the subcutaneous (sc) injection route, since the absorption and distribution rates seem to be similar to those of the im route. The sc route gives a concentration equivalent to that of the iv route at a 15-min interval. From 15 min to 3 hours, plasma levels are higher with either the im or the sc route than with the iv route. However, at 6-9 hours after administration, all parenteral routes show equal values, but with an equilibrium plateau after 1-1.5 hours. Oral administration is the least preferred route. It results in less than 20% of the maximum plasma level of free morphine, which is produced 15 min to 1 hour after an equivalent iv dose. In all studies, early oral free morphine plasma levels were lower than those from parenteral routes, but equivalent concentrations for all routes were reached after about 9 hours. Neither plasma nor urine concentrations of free morphine were high; in
........ 24
1
Morphine
and Its Analogs
contrast, the concentration of conjugated morphine, produced within the mucosal cells of the small intesiine and the liver, was 16 times the free morphine concentration in plasma after I hour. Factors contributing to these observed effects of poor oral efficacy are thought to be not only the low pH of gastric fluid and the high pK. of morphine, causing extensive ionization, lipid insolubility, and poor absorption, but also (in fact, mainly) the high hepatic extraction ratio determined by blood flow and causing extensive first-pass effects (90). Thus, only about 30% of an oral dose reaches the systemic circulation (98). Any absorption most likely occurs in the duodenum, where the h(gh pH provides a favorable environment and a large surface area exists. In summary, the initial free morphine plasma levels are highest with the iv route of administration, but the distribution, metabolism, and excretion are also more rapid. After a longer time interval (15 min to 3 h), plasma levels are highest with the two other parenteral routes. Oral . administration results in insignificant plasma morphine levels; however, the concentration of morphine conjugates is high. Plasma levels of conjugated morphine are initially highest again with the iv route. All routes result in similar values after about I hr. From 2-3 to 9 hours, however, concentrations following oral administration are highest. Cumulative 48-hr excretion of morphine conjugates is similar via all routes, accounting for 60-80% of the administered dose, although at a I-hour interval the excretion of conjugates is highest with the iv route (95).
11a. M. Gates and G. Tschudi, D. Elad and D. Ginsburg. lib. F. S6ntary, Alkaloids (N.
of Poisons,"
2nd ed.,
16. R. Robinson and S. Sugasawa, J. Chem. Soc. p. 3163 (1931). -*,17a. K. W. Bentley, "The Chemistry of the Morphine Alkaloids." (Clarendon), New York, 1954. 17b. D. Ginsberg,
He/v.
Chim.
Acta 38, 1847 (1955);
Alkaloids."
Wiley
(Interscience),
Univ.
New York,
Press
1962.
R. B. Herbert, in "The Alkaloids" (1. E. Saxon, ed.), Specialist Periodical Reports, Vol. 10, The Chemical Society, London, 1980; ibid. Vol. 9, 1979; ibid. Vol. 6, 1976. 17d. K. K. Kamo, W. Kimoto, A. Hsu, P. G. Mahlberg, and D. D. Bills, Phytochemistry 21, 219 (1982). 17e. G. W. Kirby, Science 155, 170 (1967).
,,"17f. R. H. F. Manske, Alkaloids (N. Y.) 6, 423 (1960); ibid. 12,425 (1970). >K17g. T. Robinson,
York,1968.
"The
Biochemistry
.
of Alkaloids."
Springer-Verlag,
Berlin
and New
17h. L D. Spenser, Lloydia 29, 71 (1966).
~
17L G. Stork. Alkaloids (N. Y.) 6, 219 (1960).
;J 17j. K. W. Bentlev. ed.. ,"The Chemistry of Natural Products. Vol. I. The Alkaloids." Wiley (Interscience), London, 1957. 18a. A. R. Battersby, R. Binks, and B. J. T. Harper, J, Chern. Soc. p. 3534 (1962). 18b. A. R. Battersby, R. Binks, and D. J. LeCount, Proc. Chern. Soc. p. 287 (1960). 18c. A. R. Battersby and B. J. T. Harper, Chem. and Ind. (London) p. 365 (1958). ISd. A. R. Battersby and B. J. T. Harper, Tetrahedron Leu., p. 21 (19~). 18e. E. Leete. Chem. and Ind. (London) p. 977 (1958). 19. E. Leete and J. B. Murrill, Tetrahedron Lell. p. 147 (1964). 20. M. F. Roberts and M. D. Antoun, Phytochemistry 17, 1083 (1978). 21. D. S. Bhakuni, A. N. Singh, S. Tewari, and R. S. Kapil, J. Chern. Soc., Perkin Trans. 1 p. 1662 (1977); S. Tewari, D. S. Bhakuni, and R. S. Kapril, J. Chem. Soc. Chem. Commun. p. 554 (1975). 22. M. L. Wilson and C. J. Coscia, J. Am. Chem. Soc. 97,431 (1975). 23. A. R. Battersby, R. Sinks, R. J. Francis, D. J. McCaldin, and H. Ramuz, J. Chem. Soc. p. 3600 (1964).
pp. 466ff. Lippincott,
K. W. Bentley
"The Opium
Oxford
;i'17c.
24. E. Leete, J. Am. Chem. Soc. 81, 3948 (1959). 25. 26. 27. 28.
10. M. Mackay and D. C. Hodgkin, J. Chem. Soc. p. 3261 (1955); J. K.lvod., P. and O. Jeger,
ibid. 78, 1380 (1956);
15. J. A. Duke, Econ. Bot. 27, 390 (1973).
2. M. Gordon, Annu. Rep. Med. Chern. 9,38 (1974). * 3. A. K. Reynolds and L. O. Randall, "Morphine and Allied Drugs," p. 3. Univ. of -it Toronto Press, Toronto, 1957. .; 4. D. Ginsburg, "The Opium Alkaloids." Wiley (Interscience), New York, 1962. 5. J. F. Derosne, Ann. Chim. 45, 257 (1803). 6. M. A. Seguin. Ann. Chim. 92, 225 (t814). 7. F. W. Sertiirner, Trommsdorffs J. Pharm. 14,47 (1806); see also: Gilbert's Ann. Phys. 55,56 (1817); ibid. 57, 183 (1817); ibid. 59, 50 (1818). 8. J. B. Trommsdorff, Trommsdorff's J. Pharm. 14,93 (1806). 9. J. M. Gulland and R. Robinson, Mem. Proc. Manchester Lit. Phil. Soc. 69,79 (1925). Buchschacher,
J. Am. Chem. Soc. 74, 1109 (1952); ibid. 76, 312 (1954). Y.) 12,333 (1970).
lie. F. 56nt"!,, Alkaloids (N. Y.) 17, 385 (1979). 12. J. W. Fairbairn, Pharm. J. (London) 216, 29 (1976); D. von Neubacher and K. Monthes, P/anta Med. 11,387 (1963); N. Sharghi and I. Lalezari, Nature (London)213, 1244 (1967). 13. M. A. Schwartz, "Prescription Drugs in Short Supply: Case Histories," p. 9. Dekker, New York, 1980. 14. J. W. Fairbairn and K. Helliwell, J. Pharm. Pharmacol. 29,65 (1977).
References 1. T. G. Wormley, "The Micro-Chemistry Philadelphia, Pennsylvania 1885.
25
References
and
A. R. Battersby, R. C. F. Jones, A. I. Scott, S.-L. Lee, and T. A. R. Battersby and R. Binks, A. R. Battersby, G. W. Evans, Tetrahedron Lell. p. 1275 (1965); Leu. p. 1271 (1965).
and R. Kazlauskas, Tetrahedron LeU. p. 1873 (1975). Hirata, Heterocycles 11, 159 (1978). Proc. Chem. Soc., London p. 360 (1960). R. O. Martin, M. E. Warren, Jr., and H. Rapoport, E. Brochmann-Hanssen and B. Nielsen, Tetrahedron
29. A. R. Battersby, D. M. Foulkes. and R. Binks. J. Chem. Soc. p. 3323 (1965); P. R. Borkowski, J. S. Horn, and H. Rapoport, J. Am. Chem. Soc. 100, 176 (1978).
H. M. E. Cardwell, J. Chem. Soc. p. 3252 (1955).
I
26
1 Morphine
and Its Analogs
30. E. Brochmann.Hanssen and T. Furuya, J. Pharm. Sci. 53, 575 (1964). 31a. A. R. Battersby, D. M. Foulkes, M. Hirst, G. V. Parry, and J. Staunton, J. Chern. Soc. C p. 210 (1968). 3Ib. M. A. Schwartz and I. S. Mami, J. Am. Chern. Soc. 97, 1239 (1975). 31c. K. L. Stuart, Chern. Rev. 71, 47 (1971). 32a. D. H. R. Barton, Pu," Appl. Chem. 9, 35 (1964). 32b. A. R. Battersby, "Oxidative Coupling of Phenols" (W. I. Taylor and A. R. Battersby, eds.), p. 119. Arnold, London, 1968. 32c. D. H. R. Barton, and T. Cohen, "Festschrift Arthur Stoll," p. 117.Birkhauser, Basel, 1957. 32d. T. Kametani and K. Fukumoto, J. Heterocycl. Chern. 8, 341 (1971). 33. D. H. R. Barton, Proc. Chern. Soc., London, p. 293 (1963); D. H. R. Barton, A. M. Deftonn, and O. E. Edwards, J. Chern. Soc., p. 530 (1956). 34. D. H. R. Barton, D. S. Bhakuni, R. James, and G. W. Kirby, J. Chern., Soc. C. p. 128 (1967); D. H. R. Barton, G. W. Kirby, W. Steglich, G, M. Thomas, A. R. Battersby, T. A. Dobson, and H. Ramuz, J. Chem. Soc. p. 2423 (1965). 35. E. Brochmann.Hanssen, B. Nielsen, G. Aadahl, J. Pharm. Sci. 56, 1207(1967); H. I. Parker, G. Blaschke, and H. Rapoport, J. Am. Chem. Soc. 94, 1276 (1972). 36a. A. R. Battersby, E. Brochmann-Hanssen, and J. A. Martin, J. Chem. Soc., Chem. Comm. p. 483 (1967). 36b. A. R. Battersby, J. A. Martin, and E. Brochmann-Hanssen,J. Chem. Soc. C, p. 1785 (1967). 36c. G. Blaschke, H. I. Parker, and H. Rapoport, J. Am. Chem. Soc. 89, 1540 (1967). 36d. H. Rapoport, F. R. Stermitz, and D. R. Baker, J. Am. Chem. Soc. 82,2765 (1960). 37. J. S. Horn, A. G. Paul, and H. Rapoport, J. Am. Chem. Soc. 100, 1895 (1978); E. Brochmann.Hanssen and Y. Okamoto, L/oydia 43, 731 (1980). 38. E. Brochmann-Hanssen and S. W. Wunderly, J. Pharm. Sci. 67, 103 (1978). 39. G. W. Kirby, S. R. Massey, and P. Steinreich, J. Chem. Soc., Perkin Trans. 1 p. 1642 (1972). 40. F. R. Stermitz and H. Rapoport, Nature (London) 189,310 (1961); F. R. Stermitz and H. Rapoport, J. Am. Chem. Soc. 83, 4045 (1961). 41. R. O. Martin, M. E. Warren and H. Rapoport, Biochemistry 8, 2355 (1967). 42. B. Franck, G. Blaschke, and G. Schlingloff, Angew. Chem. 75, 957 (1963). 43. H. L. Holland, Alkaloid> (N. Y.) 18,323 (1981). 43b. E. L. Way and T. K. Adler, Bull. WHO 25, 227 (1961). 44. M. R. Johnson and G. M. Milne, "Burger's Medicinal Chemistry" (M. E. Wolfe, ed.), Part III, Chapter 52. Wiley, New York, 1981. 45. A. L. Misra, S. Y. Yeh, and L. A. Woods, Biochem. Pharmacol. 19, 1536 (1970). 46. S. Y. Yeh, C. W. Gorodetzky, and H. A. Krebs, J. Pharm. Sci. 66, 1288(1977). 46h. L. A. Woods, J. Pharmacol. Exp. Ther. 112, 158 (1954). 47. J. M. Fujimoto, E. Way, and E. Leong, J. Pharmacol. Exp. Ther. 121, 340 (1957); J. M. Fujimoto, E. Way, and E. Leong, J. Am. Pharm. Assoc. 47, 273 (1958). 48. K. Shimomura, O. Kamata, S. Ueki, S. Ida, K. Oguri, H. Hoshimura, and T. Tsukamoto, Tohoku J. Exp. Med. lOS, 45 (1971). 49. J. M. Fujimoto and V. B. Haarstad, J. Pharmacol. Exp. Ther. 165,45 (1969); F. W. Oberst and E. G. Gross, J. Pharmacol. Exp. Ther. 80, 188 (1944). 50. R. H. DeMeio and C. Lewycka, Endocrinology (Baltimore) 56, 489 (1955). 51. S. Y. Yeh, H. A. Krebs, and C. W. Gorodetzky, J. Pharm. Sci. 68, 133 (1979). 52a. U. Boern"er, R. L. Roe, and C. E. Becker, J. Pharm. Pharmacol. 26, 393 (1974). 52b. C. H. March and H. W. Elliot, Proc. Soc. Exp. Bioi. 86, 494 (1964).
References
27
52c. A. L. Misra, S. J. Mule, and L. A. Woods, Nature (London) 190,82 (1961). 52d. S. Y. Yeh, R. L. McOuinn, and C. W. Gorodetzky, Drug Metab. Dispos. 5, 355 (1977). 53. A. Klutch, Drug Metab. Dispos. 2, 23 (1974). 54. D. H. Clouet, Biochem. Pharmacol. 12, 967 (1963). 55. S. Y. Yeh, J. Pharmacol. Exp. Ther. 192, 201 (1975). 56. J. W. Miller and H. H. Anderson, J. Pharmacol. Exp. Ther. 112, 191 (1954). 57a. E. J. Cone, W. D. Darwin, and C. W. Gorodetzky, J. Pharm. Pharmacol. 3], 314 ~ (1979). 57b. E. J. Cone and W. D. Darwin, Biomed. Mass. Spectrom. 5,291 (1978). 58. J. R. Gillette, Ann. N. Y. Acad. Sci. 179, 43 (1971). 59. P. H. Duquette, R. R. Erickson, and J. L. Holtzman, J. Med. Chem. 26, 1343(1983). 60. H. W. Kosterlitz, J. A. H. Lord, and A. J. Watt, in "Agonist and Antagonist Actions of Narcotic Analgesic Drugs" (H. W. Kosterlitz, H. O. J. Collier, and J. E. Villarreal, eds.), pp. 45-61. Macmillan, London, 1973. 61. H. W. Kosterlitz and A. A. Waterfield, J. Pharm. Pharmacol 28, 325 (1976). 62. R. L. H. Heimans, M. R. Fennessy, and G. A. Gaff, J. Pharm. Pharmacol. 23, 831 (1971). 63. N. B. Eddy, J. Pharmaco/. Exp. Ther. 56, 421 (1936). 64. U. Boerner and S. Abbott, Experentia 29, 180 (1973). 65. J. W. A. Findlay, E. C. Jones, and R. M. Welch, Drug Metab. Dispos. 7,310 (1979). 66. E. J. Cone, B. A. Phelps, and C. W. Gorodetzky, J. Pharm. Sci. 66, 1709 (1977). 67. A. L. Misra, N. L. Vadlamani, R. B. Potani, and S. J. Mule, Biochem. Pharmacol. 22,2129 (1973). 68. S. Y. Yeh, R. L. McQuinn and C. W. Gorodetzky, J. Pharm. Sri. 66, 201 (1977). 69, E. R. Garrett and T. Gurkan, J. Pharm. Sri. 69, 1116(1980). 70. O. Lockridge, N. Mottershaw-Jackson, H. W. Eckerson, and B. N. La Du, J. Pharma. col. Exp. Ther. 2]5, 1(1980). 71. G. L. Cohn, J. A. Cramer, and H. D. Kleber, Proc. Exp. Bioi. Med. 144,351 (1973). +,72. E. L. Way, J. M. Young, and J. W. Kemp, Bull. Narc. 17,25 (1965). 73. K. C. Rice and A. E. Jacobson, J. Med. Chem. 18, 1033 (1975). 74. J. M. Fujimoto, S. Roerig, R. I. H. Wang, N. Chatterjie, and C. E. Inturrisi, Proc. Soc. Exp. Bioi. Med. 148, 433 (1975). 75. J. W. Fairbairn, S. S. Handa, E. Gurkan, and J. D. Phillipson, Phytochemistry 17, 261 (1978). 76. R. J. Miller, C. Jolles, and H. Rapoport, Phytochemistry 12, 597 (1973). 77. J. W. Fairbairn and S. EI-Masry, Phytochemistry 7, 181 (1968). 78. J. W. Fairbairn, F. Hakim, and Y. E. Kheir, Phytochemistry 13, 1133 (1974). 79. J. D. Phillipson, S. S. Handa, and S. EI-Dabbas, Phytochemistry 15, 1297 (1976), 80. E. R. Garrett and T. Gurkan, J. Pharm. Sci. 67, 1512 (1978). 81. D. H. Carlin, J. Pharmaco/. Exp. Ther. 200, 224 (1977). 82. D. R. Stanski, L. Paalzow, and P. O. Edlund, J. Pharm. Sci. 71, 314 (1982). 83a. C. C. Hug, Jr., M. R. Murphy, E. P. Rigel, and W. A. Olson, Anesthesiology 54, 38 (1981). 83b. M. R. Murphy and C. C. Hug, Jr., Anesthesiology 54, 187 (1981). 84. B. E. Dahlstrom, L. K. Paalzow, G. Segre, and A. J. Agren, J. Pharmacokinet. Biopharm. 6, 41 (1978). 85. C. T. Walsh and R. R. Levine, J. Pharmacol. Exp. Ther. 195, 303 (1975). 86. B. A. Berkowitz, K. V. Cerreta, and S. Spector, J. Pharmacol. Exp. Ther. 191, 527 (1974).
28
Morphine
and Its Analogs
87a. S. Spector, J. Pharmacal. Exp. Ther. 178,253 (1971). 87b. S. Speclor and E. S. Vesell, Science 174, 421 (1971). 88. B. E. Dahlstrom and L. K. Paalzow, J. Pharmacokinet. Biopharm. 6, 505 (1978). 89. K. Iwamoto and C. D. Klaassem, J. Pharmacal. Exp. Ther. 200,236 (1977). 90. B. Dahlstrom, J. Johansson, and L. Paalzow, Acta Pharmacal. Toxicol. 39,46 (1976). 91. M. J. Rance and J. S. Shillingrord, Biochem. Pharmacal. 25,735 (1976). 92. E. R. Garrett and A. Jackson, J. Pharm. Sri. 68, 753 (1979). 93. U. Boerner, S. Abbott, and R. L. Roe, Drug Metab. Rev. 4, 39 (1975). 94. S. F. Brunk and M. Delle. Clin. Pharmacal. Ther. ]6, 51 (1974). 95. B. A. Berkowitz, Clin. Pharmacokinet. I, 219 (1976). 96. B. E. Dahlstrom and L. K. Paalzow, J. Pharmacokinet. Biopharm 3, 293 (1975). 97. P. Bullock, S. Spanner, and G. B. Ansell, Biochem. Soc. Trans. 5, 335 (1977). 98. D. R. Stanski, D. J. Greenblatt, and E. Lowenstein, Clin. Pharmacol. Ther. 24,52 (1978). 99. E. F. Hahn, Medicinal Research Rev. 5,255 (1985). 100. C. E. Inturissi, M. B. Max, K. M. Foley, M. S. Schultz, S-U. Shin, and R. W. Houde, New Engl. J. Med. 310, 1213(1984).
2. Biological Effects of Opioids by Donna
Hammond'
I. Introduction II. III. IV. V. VI. VII.
..
.
29 31 35 37 38 39 40 41
..
Multiplicity of Opiate Receptors Analgesia. . . Respiratory Depression. Gastrointestinal Motility Dependence Liability. Summary .. .. References . . . .
I. Introduction In the mid 1970s, research in the field of opioid pharmacology was galvanized by several significant discoveries. In 1973, Pert and Snyder (1) and Terenius (2), as well as Simon et al. (3), demonstrated that opiates bound with high affinity to homogenates of brain in a saturable and stereospecific manner. Furthermore, the affinity of the opiates for this binding site closely paralleled their in vivo (4) and in vitro (5) potencies. Using biochemical techniques, these studies demonstrated the existence of the "opiate receptor" previously hypothesized on the basis of pharmacological evidence. The high degree of stereoselectivity of opioids for this receptor first observed with dextrorphan and levorphanol was later confirmed with (+)- and (- )-morphine (6) and (+)- and (- )-naloxone (7). Although the existence of receptors for substances endogenous to the body (e.g., the neurotransmitter acetylcholine) had long been recognized, the existence of receptors for substances that were not endogenous to the body, such as morphine, was a novel concept. Identification of an opiate receptor therefore fueled the search for an endogenous morphine-like substance in the brain. In 1975, Hughes (8) identified a substance in extracts of brain that inhibited contractions of guinea pig ileum and mouse vas deferens in a naloxone-reversible manner. Using opiate receptorbinding techniques, Pasternak et al. (9) and Terenius and Wahlstrom (10) ,
Department of Biological Research, G. D. Searle & Co., Research and Development
Division,
Skokie,
Illinois
60077.
29
30
2
Biological
Effects
of Opioids
Table 2-1 Biological
Effects of Opiates
Excitation
(feline,
Sedation (human, Miosis
(rodent,
equine) primate,
rodent)
human)
Mydriasis (feline, primate) Euphoria (in the presence of discomfort) Dysphoria (in the absence Nausea and emesis Analgesia Antitussive Respiratory depression
Multiplicity of Opiate Receptors
31
be provided in one chapter. Consequently, this chapter focuses on the interactions of opioids with the different opiate receptors and briefly reviews those opioid effects of greatest clinical relevance: analgesia, inhibition of gastrointestinal transit, and respiratory depression and dependence (physical and psychic). For a more comprehensive review of opioid pharmacology, the reader is referred to several excellent monographs and texts (24-26).
of discomfort)
Inhibition of gastrointestinal transit Urinary
II
retention
Tolerance Physical dependence
similarly demonstrated a substance in brain extracts that displaced the binding of radiolabeled opiates. These substances were subsequently identified as the pentapeptides, [Metjenkephalin and [Leu]enkephalin (II ,12). An additional morphine-like peptide of larger molecular weight, i3-endorphin, was isolated from pituitary extracts by Goldstein and colleagues (13,14). Currently, three separate, individually gene-derived families of endogenous opioid peptides are recognized: the enkephalins, the endorphins, and the dynorphins (15-18) (d. Chapter 11). The recent isolation of morphine itself from bovine brain (19) and frog skin (20) is particularly interesting in light of the past 10 years of research effort to identify and isolate endogenous morphine-like substances. However, the origin of this morphine remains to be identified. As early as the 196Os, Portoghese (21) and Martin (22) suggested that opioid analgesics interacted with multiple receptors or with multiple modes to a single receptor. However, these concepts were not vigorously pursued again until 1976 when Martin et al. (23) examined the pharmacological profile of morphine and its cogeners in the chronic spinal dog and identified three different syndromes. They attributed these syndromes to interactions of the prototypic opioid agonists with three distinct receptors. Thus, on the basis of pharmacological evidence, Martin hypothesized the existence of not one but multiple opiate receptors. This subsequently sparked an equally attractive hypothesis, still being examined today, that the different biological effects of opioids are mediated by different opiate receptors. Because the biological actions of opioids are numerous, diverse and complex (Table 2-1), an exhaustive review of opioid pharmacology cannot
II.
Multiplicity of Opiate Receptors
Two approaches have been utilized in studies of the opiate receptor. In the first, the pharmacological profile of a series of prototypic opioid agonists and antagonists was carefully characterized using behavioral, physiological, or pharmacological measures. This approach was used in the early studies of Martin et al. (23) and Lord et al. (27). Using the chronic spinal dog. Martin et al. (23) evaluated the effects of a series of opioids on a large number of physiological functions including pulse rate. respiration rate, temperature, pupil diameter, behavioral state. physical dependence, and nociceptive reflexes. The effects of the opioids could be classified according to three syndromes (see Table 2-2 of ref. 23) suggesting the existence of three distinct opiate receptors. These receptors were termed J.L, K, and (J". The J.Lreceptor, for which the prototypic agonist was morphine, was associated with bradycardia, miosis, respiratory deceleration, indifference, and analgesia. In contrast, the (J" receptor was associated with tachycardia, mydriasis, respiratory acceleration, and delirium; its prototypic agonist was SKF 10,047 (N-allylnormetazocine). The K receptor was associated with miosis, analgesia, sedation, and little change in pulse or respiratory rate; its prototypic agonist was ketocyclazocine. The studies by Lord et al. (27) utilized a similar pharmacological approach in investigating the effects of a series of opioids and opioid peptides on electrically induced contractions of smooth muscle. The authors hypothesized that the relative potencies of a series of opioids should be closely correlated among a variety of in vitro assay systems if the opiate receptor population of these systems was identical. The heterogeneity of opiate receptors in different tissues was made immediately apparent by this study. Thus, the opioid pentapeptides [Met]- and [Leu]enkephalin were more potent inhibitors of the electrically induced contractions of mouse vas deferens than of guinea pig ileum. Conversely, morphine was more potent in the guinea pig ileum than in the mouse vas deferens preparation. Furthermore, the concentration of naloxone required to inhibit the actions of the pentapeptides in the mouse vas deferens
..,.... 32
2
Biological
Effects
of Opioids
was ten times that required to inhibit the actions of morphine in the same tissue.! Together, these observations suggested the existence of two types of opiate receptors present in differing amounts. The first, to which morphine bound with high affinity, was termed the I" receptor; it was the predominant opiate receptor in guinea pig ileum. The second, to which the pentapeptides bound with high affinity, was termed the 0 receptor; it was the predominant receptor species in the mouse vas deferens. Using a similar approach to compare the potencies of morphine and benzomorphans in the guinea pig ileum and mouse vas deferens preparations, Hutchinson et al. (28) concluded that certain novel benzomorphans bound to a receptor distinct from that bound by morphine. This receptor was later suggested to correspond to the K receptor postulated by Martin et al. (23). The rabbit vas deferens has been demonstrated to contain a preponderance of K receptors (29). The opiate receptor has also been characterized biochemically using a number of different approaches (30). If the opioid was available in radiolabeled form, the kinetics of its saturable, stereospecific binding were examined using analyses of saturation curves to determine the equilibrium dissociation constant (Kd) and the number of binding sites (Bmax). In an alternative approach suitable for use with ligands that were not radiolabeled, the affinity of the unlabeled ligand for a receptor was estimated from its ability to displace the binding of radiolabeled ligands of known characteristics. Finally, very elaborate "selective protection" studies were conducted in which a specific receptor among a mixed receptor population was pre incubated with unlabeled ligand. This receptor was subsequently "protected" from alkylation (irreversible inactivation) with drugs such as N-ethylmaleimide and phenoxybenzamine. Such studies were used to demonstrate the existence of different opiate receptors in a mixed receptor population and to yield a functionally homogeneous receptor population for study. On the basis of the biochemical and pharmacological studies discussed above, four subtypes of the opiate receptor have been proposed: 1", 0, K, and (]"(30-33). In addition, prototypic agonists for each receptor subtype have been proposed. Table 2-2 lists the receptor subtypes and those I
with different
Receptors
receptors
through
particular
receptor
concentration the same dissociation obtained agonist
binding properties
the use of antagonists.
may be expressed
of antagonist
pharmacological constant
that
may
with
an antagonist
interacts
with
in terms
requires
effects
measured
also be expressed against
different
The
of an antagonist
of its equilibrium
a doubling
of the
against
dissociation
agonist
of the antagonist.
as its negative
logarithm,
in different
in the
tissue
preparations.
an agonist
constant
concentration
in the absence
an agonist
receptors
can be differentiated in a mixed population efficacy
or pA2.
preparations
Table 2.2 Prototypic
Agonists
for the Opiate
Receptors
Receptor
Mu(,,)
Agonist
Morphine
Dihydromorphine
Delta (5)
DAGO DADLE DTLET DPen2 ,LPen~ -enkephalin
Tyr'D-Ala-Gly-MePhe-NH(CH2)20H Tyr-D.Ala-Gly-Phe-D-Leu Tyr-D- Thr-Gly-Phe-LeuThr
(CH,hC-s
I
Kappa (K)
Ethylketocyclazocine
HO Dynorphin'_9
Tyr-Gly.Gly-Phe-Leu-Arg-Arg-IIe-Arg
~NJ;JNMe V-50,488
of
Cl
at a
Ke, that
Cl
to achieve pAz
will differ
I
Tyr-NHCHCO-Gly-Phe-NHCHC02H
The equilibrium The
s.-C(CH,h
values if the
Sigma «T)
SKF 10,047 (N-allylnormetazocine)
HO
34
2
Biological
Effects
of Opioids
opioids commonly considered to be prototypic ligands. Although some opioids exhibit as much as 100 times greater affinity for one receptor than for another, it is important to realize that no opioid is truly specific for any one opiate receptor subtype.2 Rather, opioids are discussed in terms of their selectivity and relative affinity for one receptor as opposed to another. Morphine is the prototypic ligand for the ILreceptor. Other compounds that have greater affinity for the IL receptor include dihydromorphine, phenazocine, etorphine, and levorphanol (30,33). Certain en kephalin analogs also bind preferentially to the ILreceptor. Although the I) receptor was initially viewed as the "peptide" receptor, it quickly became apparent that modification of the C-terminus of the pentapeptides decreased their affinity for the I) receptor relative to the IL receptor (34-36). FK 33-824 [Tyr-DAla-Gly-MePhe-Met-(O)-ol] (37,38) and OAGO [Tyr-DAla-GlyMePhe-Gly-ol] (39) are two opioid pep tides that bind to the ILsite with 20 and 100 times, respectively, greater affinity than to the I) receptor. Morphiceptin [Tyr-Pro-Phe-Pro-NH2] and certain of its analogs also bind to the IL receptor with much greater affinity than to the I) receptor (40). The current prototypic agonist for the I) receptor is oAla2-DLeu'en kephalin (OAOLE). However, it is only three times more selective for the I)site than for the ILreceptor (41). Thus, it is likely to be supplanted as a prototypic ligand by several of the more recently synthesized and substantially more I)-selective enkephalin analogs. These include OSLET [Tyr-DSer-Gly-Phe-Leu-Thr] (36,42) and OTLET [Tyr-DThr-Gly-PheLeu-Thr] (41), which are 8 and 23 times more selective for the I) receptor, respectively. Oimeric pentapeptide [Tyr-oAla-Gly-Phe-LeuNH],. (CH2)n (OPEn) (43) and tetrapeptide [Tyr-oAla-Gly-PheNH]2' (CH2)n (OTEn) (44) enkephalins in which the monomers are linked by methylene chains of varying length have also been determined to bind to the I) receptor with greater affinity than to the IL receptor. In the OPEn series, OPE, has the greatest affinity for the I)receptor, while OTE12 has the greatest affinity for the I) receptor in the OTEn series. Conformationally constrained cyclic 2 Opioid compounds and selectivity between (i.e., that concentration
are frequently compared in terms of potency, affinity for a receptor, receptors. Estimates of potency and affinity are based on IC50 values of drug that produces a 50% reduction of effect or displacement of a
radiolabeled ligand). The ratio of ICso values determined for the different receptors provides a measure of the relative potency of the compound at the different receptors but does not address the selectivity of the compound quite useful in in vitro tissue bath
for a particular receptor. preparations. Selectivity
This approach is. however, can be addressed in the
radiolabeled ligand displacement studies by conversion of the ICso values to Ki values. The ratio of Ki values provides an estimate of relative selectivity for a receptor. It should be noted that estimates of potency, affinity, and selectivity differ among experimental preparations and between studies.
III
Analgesia
35
analogs of enkephalin have also been prepared using penicillamine (13,13dimethylcysteine) residues (45). Two of these rigid cyclic analogs, [DPen', DCys5]-enkephalinamide and [DPen',LCyss]-enkephalinamide, are 20 times more potent at the I) than at the IL receptor. The corresponding carboxylic acid terminal compounds, [DPen2,DCys5]_ and [Open2,LCys5]_ enkephalin, similarly exhibit high affinity for the I)receptor (46). However, the most pronounced I)selectivity of this series has been observed with the bis-penicillamine analogs [DPen',LPen'J- and [open2,Dpen']-enkephalin (47). These analogs are 371 and 175 times more potent, respectively, at the I) receptor than at the IL receptor. Ethylketocyclazocine, ketocyclazocine, and bremazocine bind to the K receptor with high affinity. However, these ligands are not very selective for the K receptor, since they also bind relatively well to the IL receptor (48). Recently, U-50488 (trans-3,4-dichloro-N-methyl-N-(2-(Ipyrrolidinyl)cyclohexyl]benzeneacetamide) and U-69593 (5a,7a,8f3-( -)_ N-[7-(I-pyrrolidinyl)-I-oxaspiro[ 4,5]dec-8-yl]benzeneacetamide) have been demonstrated to bind more selectively to the K receptor (49,50). Among the endogenous opioid peptides, the dynorphins exhibit high affinity for the K receptor (51). The most selective of these is dynorphin'_9' which is 10 times more selective for the K receptor than for the IL receptor (52). SKF 10,047 is the prototypic opioid ligand for the a receptor. The reader is referred to the reviews by Zukin and Zukin (31,32) for an in-depth' review of this receptor and its relationship to phencyclidine. In summary, four subtypes of the opiate receptor have been identified on the basis of their different pharmacological profiles. In addition, biochemical and auto radiographic studies have demonstrated that these receptors are distributed differentially throughout the central nervous system (CNS) and the periphery (54). These observations have led to the suggestion that the different opiate receptors may mediate the different pharmacological effects of the opioids. Thus, since 1976, investigators have attempted, not always successfully, to assign pharmacological significance to each of the opiate receptor subtypes. The following sections briefly review these attempts.
III.
Analgesia
A wide variety of behavioral tests are available to evaluate the potential efficacy of an opioid as an analgesic. These tests differ in the type, intensity, and duration of the noxious stimulus; in the endpoint response that is measured; and in their sensitivity to inhibition by opioids. These tests include the writhing (55-57), formalin (58), tail flick (59), hot plate
36
2
Biological
Effects
of Opioids
(6(}), and flinch-jump tests (61) routinely performed in the rodent, the skin twitch response performed using the dog (62), and the discrete trial shock titration procedure used in the primate (63,64). Table 1 of the review by Martin (24) includes a comparison of the efficacy of different opioids in these various tests. In addition to the analgesia produced upon systemic administration of opioids, an animal's response to noxious stimuli can be attenuated or inhibited by injection of microgram amounts of opioids into the cerebral ventricles, into certain nuclei of the brain stem, and into the spinal cord subarachnoid space (65). These findings suggest that systemically administered opioids produce analgesia by acting at sites within the CNS. This conclusion is supported by studies demonstrating that the analgesia produced by systemically administered opioids can be antagonized by microinjection of opiate antagonists in the cerebral ventricles, spinal cord subarachnoid space, or brain stem nuclei (66-70). Finally, it has been proposed that the analgesia produced by systemically administered opioids is the product of a complex interaction of the many sites in the CNS at which the opioids act (71-73). Thus, microinjection of a subeffective dose of opioids at both a supraspinal and a spinal site produces a profound analgesia that exceeds that anticipated on the basis of an additive effect at both sites. The analgesic activity of opioids was initially attributed to activation of the J.' receptor. This hypothesis was based in part on the differential distribution of the opiate receptor subtypes in the brain and the finding that regions involved in the processing of nociceptive information were enriched in J.' sites (74-76). Although the analgesic activity of K agonists such as ethylketocyclazocine has been recognized since 1976 (23,77), this effect could not be dissociated from their additional affinity for the J.' receptor. A selective K agonist, U-50488, has been shown to produce analgesia (49). This observation, and the analgesic activity of K agonists that have a J.' antagonist rather than agonist activity (e.g., nalorphine), suggest that activation of K receptors also produces analgesia. In addition, K receptors have been visualized in the spinal cord, a CNS site involved in the processing of nociceptive information (78). The analgesic activity of S agonists, particularly after intracerebroventricular or intrathecal administration, has been demonstrated in several studies (79-82), as has the existence of S receptors in regions of the CNS involved in the processing of nociceptive information (74-76). Thus, it appears that the analgesic activity of opioids cannot be attributed to activation of one particular subtype of the opiate receptor. Indeed, it is increasingly apparent that experimental conditions (species, test, agent, etc.) can significantly affect the conclusion~ drawn by studies
IV
Respiratory Depression
37
that attempt to assign functional significance to a particular opiate receptor. The choice of species can be be an important factor. For example, only 10% of the opiate receptor binding sites in rat brain can be characterized as K-like, in contrast to estimates as high as 40% in guinea pig brain (48,83, but see 50). Also, although K-like sites have been demonstrated in mouse and guinea pig spinal cord, they appear to be absent in rat spinal cord (50). The choice of analgesiometric test can also influence the results of such studies. Several investigators have demonstrated that analgesiometric tests utilizing thermal stimuli, such as the tail flick and hot plate tests, are sensitive indicators of the anal esic actions of J.' agonists but are insensitive to the actIOns of K agonists 49,84 86). In contrast, analgesiometric tests utilizIng chemIcal stimuli, such as the ~hing test, are sensitive to the analgesic actIons of both J.' and K agonists. Finally, the relative selectivity (or lack thereof) of the oplOld used as a prototypic ligand is also an important factor. Therefore, the results of studies that attempt to attribute the in vivo effects of an opioid to an interaction with a specific receptor on the basis of its in vitro profile must be reviewed with great care.
IV,
I
I I I
Respiratory Depression
Respiration may be monitored using a number of different techniques including measurement of respiratory rate, tidal volume, minute volume (the product of respiratory rate and tidal volume), the arterial tension of C02 and O2 (also known as pC02 and p02), integrated phrenic nerve activity, and responsivity to C02' Using these techniques, investigators have shown that opioids alter respiratory rate, rhythmicity, pattern, and minute volume (24,87). Morphine has been demonstrated to affect the frequency and tidal volume control mechanisms of the respiratory center independently (88) and to depress the peripheral hypoxic drive to respiration (89). Opioids also decrease responsivity to C02 (24). For example, morphine, pentazocine, and nalbuphine shift the C02 stimulus-respiratory response curve in humans to the right (24,89,90). Indeed, almost the entire decrease in total respiration may be attributed to a failure of the respiratory center to respond fully to C02 (88). Opioid peptides such as [Met]enkephalin, fJ-endorphin, and DAla2-DLeu5-enkephalin also depress respiration (91-94). Several studies have tried to determine whether the respiratory depressant and analgesic effects of opioids are mediated by the same opiate receptor. The results of these studies indicate that the respiratory depressant and analgesic effects of opioids are mediated by different opiate
38
2
Biological Effects
of Opioids
VI
receptors. Thus, McGilliard and Takemori (95) and Pazos and Florez (96) reported that the apparent pAz value for antagonism of the analgesic effect of an opioid differed from the apparent pAz value for antagonism of its respiratory depressant effect. If these effects had been mediated by the same opiate receptor, the pAz values would have been similar. In addition, pretreatment with selective, irreversible I" antagonists such as {3funaltrexamine (97) or naloxonazine (98) has been shown to antagonize the analgesic effect, but not the respiratory depressant effect, of opioids. Taken together, the results of these studies indicate that the respiratory depressant and analgesic effects of opioids are mediated by different opiate receptors. It should be noted that the respiratory depressant effects of opioids cannot be attributed to activation of one particular opiate receptor subtype and that I" as well as Ii opiate receptors are involved in the depression of respiration (24,92,94,96).
VI.
)9
Liability
Dependence Liability
Drug dependence has been defined by the World Health Organization Expert Committee on Drug Dependence as a state, psychic and sometimes
also physical,
from the interaction between a and other responses that to take the drug on a continuous or periodic basis in order
living organism and a drug, characterized always include a compulsion to experience (107)
V. Gastrointestinal Motility The decrease in gastrointestinal motility produced by opioids is a result of an increase in the tone of portions of the stomach, small intestine, and large intestine and a decrease in the number of propulsive contractions of the small and large intestines (99,100). Both peripheral and central sites of action mediate the reduction in gastrointestinal motility produced by opioids (101-103). Thus, gastrointestinal transit of a charcoal or radiolabeled chromium "meal" is inhibited following intracerebroventricular (101-103) or intrathecal (104) administration of opioids. Furthermore, the decrease in gastrointestinal transit produced by systemically administered opioids is attenuated by intracerebroventricular administration of opioid antagonists (101,105,106). The results of these studies suggest that a portion of the gastrointestinal effects of opioids is mediated by sites within the CNS. A peripheral site of action is indicated by the finding that opioids that do not cross the blood-brain barrier, such as loperamide, also decrease gastrointestinal motility (102). Ward and Takemori (106) have suggested that the centrally mediated effect of opioids on gastrointestinal motility is mediated by receptors, whereas both and Kopiate receptors I" I" mediate the peripherally mediated effects of opioids on gastrointestinal motility. Porreca and Burks (104) have similarly concluded that the centrally mediated actions of opioids on gastrointestinal transit are mediated by 1", but not K, opiate receptors. These authors have also concluded that the gastrointestinal effects of opioids exerted at the level of the spinal cord involve Ii as well as 1", but not K, opiate receptors (104).
Dependence
I
its psychic effects,
resulting
by behavioral
and sometimes
to avoid the discomfort
of its absence
This definition recognizes that there are two components to drug dependence, psychic and physical, and that psychic dependence can occur in the absence of physical dependence. The definitions of psychic and physical dependence developed by Eddy et al. (108) further clarify the difference between these two components of drug dependence. Thus, psychic dependence is defined as "a feeling of satisfaction and a psychic drive that require periodic or continuous administration of the drug to produce pleasure or to avoid discomfort." Physical dependence is defined as "an adaptive state that manifests itself by intense physical disturbances when the administration of the drug is suspended. . . . These disturbances, i.e. the withdrawal or abstinence syndromes, are made up of specific arrays of symptoms and signs of psychic and physical nature that are characteristic for each drug type." The different abstinence syndromes and signs that characterize withdrawal from prototypic drugs such as alcohol, amphetamine, morphine, and cannabis reflect the different peripheral and neuronal substrates on which these compounds act to produce their effects. Several animal models of physical drug dependence have been developed including the single-dose suppression test, the precipitated withdrawal test, and the primary dependence test. The single-dose suppression test is used to determine the ability of an opioid agonist to suppress the abstinence syndrome exhibited by opioid-dependent animals in the process of withdrawal. In contrast, the precipitated withdrawal test is used with opioid antagOllist analgesics or mixed agonist-antagonist analgesics to determine their ability to precipitate an abstinence syndrome in opioiddependent animals. The primary dependence capacity of the opioid is usually evaluated last. In this test, the opioid is administered frequently over a period of 30-45 days and is then abruptly withdrawn, at which time the animals are monitored for signs of abstinence (109,110). Animal models of psychic dependence have also been developed. The reinforcing properties of an opioid can be evaluated by studies of its ability to support or initiate self-administration. The rhesus monkey IS an excellent species for these studies because it will self-admimster the
..... 40
2
Biological
Effects
of Opioids
References
majority of the drugs self-administered by humans for nonmedical purposes (109). The discriminative stimulus properties of an opioid can be evaluated using a drug discrimination paradigm in which an animal is trained to press one lever (drug lever) after administration of a prototypic opioid and another lever (saline lever) after administration of a vehicle. If a test compound is substituted for the prototypic opioid and the stimulus properties of the test compound generalize to those of the prototypic opioid, the animal will indicate this fact by responding on the drug lever. Importantly, the classificatio.n of opioids according to their discriminative stimulus properties, as indicated by the results of drug discrimination studies in rodents and primates is strikingly similar to their classification according to their subjective effect as reported by humans (109). The use of animal models to evaluate the dependence liability of opioids is discussed in greater detail in the monograph authored by the Committee on Problems in Drug Dependence in conjunction with the National Institute of Drug Abuse (110) and in the review by Woolverton and Schuster (109). A detailed discourse on the psychic and physical dependence liabilities of various opioids is not within the scope of this section. This information can be obtained from several reviews (24,109) and from the annual reports of the Committee on Problems in Drug Dependence. A review of in vivo studies of the dependence liability of opioids indicates that p.- and K-selective opioids have different dependence profiles. Thus, compounds with preferential affinity for the K receptor did not completely generalize to the discriminative stimulus properties of p. agonists (109,111-113; see also Table 7 of ref. 24) and were not self-administered by rhesus monkey (114,115), unlike p. agonists such as morphine. Furthermore, K agonists did not suppress the abstinence syndrome in morphine-dependent animals (77,114). Finally, the abstinence syndrome exhibited by animals' made dependent on K agonists was different from that exhibited by animals made dependent on p. agonists (24,49,77,114). The dependence profile of selective 5 agonists has not been thoroughly examined at this time.
VII.
41
have attempted to equate each receptor subtype with a particular biological effect of the opioids, such as analgesia, respiratory depression, gastrointestinal motility, and dependence. A brief review of the data to date indicates that it has not been possible to assign a physiological significance to one particular opiate receptor subtype. Rather, the different subtypes appear to mediate, to differing extents, many of the pharmacological effects of opioids.
References
Summary I
Research on opioid pharmacology was galvanized in the 1970s by the demonstration of the existence of an opiate receptor and the subsequent isolation and characterization of its endogenous ligands. At this point, at least four subtypes of the opiate receptor have been well characterized (p., K, and IT) and several others have been postulated (A, 0). Many studies
I
1. C. B. Pert and S. H. Snyder, Science 179, 1011 (1973). 2. L. Terenius, Acta Pharmacol. Toxicol. 32, 317 (1973). 3. E. J. Simon, J. M. Hiller, and I. Edelman, Proc. Natl. Acad. Sci. U.S.A. 70, 1947 (1973). 4. R. S. Wilson, M. E. Rogers, C. B. Pert, and S. H. Snyder, J. Med. Chern. 18, 240 (1975). 5. I. Creese and S. H. Snyder, J. Pharmacal. Exp. Tha. 194,205 (1975). 6. Y. F. Jacquet, W. A. Kleet K. C. Rice, I. Iijima, and J. Minamikawa, Science 198, 842 . (1977). 7. I. Iijima, J.I. Minamikawa, A. E. Jacobsen, A. Brossi, K. C. Rice, and W. A. Klee, J. Med. Chern. 2t, 398 (1978). 8. J. Hughes, Brain Res. 88, 295 (1975). 9. G. W. Pasternak, R. Goodman, and S. H. Snyder, Ufe Sci. 16, 1765 (1975). to. L. Terenius and A. Wahlstrom, Acta Physiol. Scand. 94, 74 (1975). 11. J. Hughes, T. W. Smith, H. W. Kosterlitz, L. A. Fothergill, B. A. Morgan, and H. R. Morris, Nature (London) 258, 577 (1975). 12. R. Simantov and S. H. Snyder, Proc. Natl. Acad. Sci. U.S.A. 73,2515 (1976). 13. H. Teschemacher, K. E. Opheim, B. M. Cox, and A. Goldstein, Ufe Sci. 16, 1771 (1975). 14. B. M. Cox, H. Teschemacher, K. E. Opheim, and A. Goldstein, Ufe Sci. 16, 1777 (1975). 15. J. Hughes, Br. Med. Bull. 39, 17 (1983). 16. B. M. Cox. Life Sci. 31, 1645 (1982). 17. F. E. Bloom, Annu. Rev. Pharmacol. Toxicol. 23, 151 (1983). 18. E. Weber, C. J. Evans, and J. D. Barchas, Trentb NeuroSci. 6, 333 (1983). 19. A. Goldstein, R. W. Barrett, I. F. James, L. I. Lowney, C. J. Weitz, L. L. Knipmeyer, and H. Rapoport, Proc. Natl. Acad. Sci. U.S.A. 82, 55203 (1985). 20. K. Oka, J. D. Kantrowitz, and S. Spector, Proc. Natl. Acad. Sci. U.S.A. 82, 1852 (1985). 21. P. S. Portnghese. J. Med. Chern. 8, 606 (1965). 22. W. R. Martin, Pharmacal. Rev. 19, 463 (1967). 23. W. R. Martin, C. G. Eades, J. A. Thompson, R. E. Huppler, and P. E. Gilbert, J. Pharmacol. Exp. Ther. 197, 517 (1976). 24. W. R. Martin, Pharmacal. Rev. 35, 283 (1984). '*' 25. O. W. Lever, K.-J. Chang, and J. D. McDermed, Annu. Rep. Med. Chem. 19, 1 (1984).
42
2 Biological Effects of Opioids
26. D. H. Clauet, "Narcotic Drugs: Biochemical Pharmacology." Plenum, New York, 1971. 27. J. A. H. Lord, A. A. Waterfield, J. Hughes, and H. W. Kosterlitz, Nature (London)
r
267, 495 (1977).
28. M. Hutchinson, H. W. Kosterlitz, F. M. Leslie, A. A. Waterfield, and L. Terenius, Br. J. Pharmacol. 55, 541 (1975). 29. T. Oka, K. Negishi, M. Suda, T. Matsumiya, T. Ioazu, and M. Ueki, Eur. J. Pharmacal.
30. 31. 32. 33. 34. 35.
73, 235 (1981).
S. J. Paterson, L. E. Robson, and H. W. Koster/itz, 8r, Med. Bull. 39, 31 (1983). R. S. Zukin and S. R. Zukin, Ufe Sci. 29, 2681 (1981). R. S. Zukin and S. R. Zukin, Trends NeuroSci. 7, 160 (1984). P. L. Wood, Neuropharmacology 21, 487 (1982). Y. Audigier, H. Mazarguil, R. Gout, and J. Cros, Eur. J. Pharmacol. 63, 35 (1980). H. W. Kosterlitz, J. A. H. Lord, S. j. Paterson, and A. A. Waterfield, Br. J. PhormacoI.
68, 333 (1980).
68. 69. 70. 71. 72. 73. 74.
20, 484 (1981).
37. K.-J, Chang, B. R. Cooper, E. Hazum, and P. Cuatrecasas, Mol. Pharmacol. 16,91 (1979). 38. D. Roemer, H. H. Buescher, R. C. Hill, l. Pless, W. Bauer, F. Cardinaux, A. Closse, D. Hauser, and R. Huguenin, Nature (London) 268, 547 (1977). 39. B. K. Handa, A. C. lane, J. A. H. Lord, B. A. Morgan, M. J. Rance, and C. F. C. Smith, Eur. J. Pharmacol. 70, 531 (1981). 40. K.-J. Chang, E. T. Wei, A. Killian, and l,-K. Chang,J. Pharmacol. Exp. Ther. 227,403 (1983). 41. J.-M. Zajac, G. Gacel, F, Petit, P. Dudey, P. Rossignol, and B. P. Roques, Biochem. Biophys. Res. Commun, Ill, 390 (1983). 42. G. Gacel, M,-C. Fournie-Zaluski, and B. P. Roques, FEBS Len. 18,245 (1980). 43, D. Rodbard, T, Costa, Y. Shimohigashi, and S. Krumins, J. Rec. Res. 3, 21 (1983). 44. Y. Shimohigashi, T. Costa, H.-c. Chen, and D. Rodbard, Nature (London) 297, 333 (1982). 45. H. I. Mosberg, R. Hurst, V. J. Hruby, J. J. Galligan, T. F. Burks, K. Gee, and H. I. Yamamura, Biochem. Biophys. Res. Commun. 106, 506 (1982). 46. H. I. Mosberg, R, Hurst, V. J. Hruby, J. J. Galligan, T. F. Burks, K. Gee, and H. I. Yamamura, Life Sci. 32, 2565 (1983). 47. H. I. Mosberg, R. Hurst, V. J. Hruby, K. Gee, K. Akiyama, H. I. Yamamura, J. J. Galligan, and T. F. Burks, Proc. Natl. Acad, Sci. U.S.A. 80, 5871 (1983). 48. H. W. Kosterlitz, S. J. Paterson, and L. E, Robson, Br, J. Pharmacol. 73, 939 (1981). 49. P. F. Yon Voigtlander, R. A. Lahti, andJ. H. Ludens,J. Pharmacol. Exp. Tha. 224, 7 (1983). 50. R. A. lahti, M. M. Mickelson, J. M. McCall, and P. F. Von Voigtlander, Eur. J. Pharmacol.
75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89.
109, 281 (1985).
90.
51. J. P. Huidobro-Toro, K. Yoshimura, N. M. Lee, H. H. Loh, and E, L. Way, Eur. J. Pharmacol.
91. 92. 93. 94. I
,
4J
J. J. Loux, S. Smith, and H. Salem, Arzneim.-Forsch.lDrug Res. 28, 1644 (1978). J. Pearl, M. D. Aceto, and L. S. Harris, J. Pharmacol. Exp. Ther, 160,217 (1968). D. Dubuisson and S. G. Dennis, Pain 4, 161 (1977). F. E. D'Amour and D, L. Smith, J, Pharmacol. Exp. Ther. 72, 74 (1941). G. Woolfe and A. D. MacDonald, J. Pharmacol. Exp. Ther. 80, 300 (1944). W, O. Evans, Psychopharmacologia 2, 318 (1961). W. R. Martin, C. G. Eades, H, F, Fraser, and A, Wikler, J. Pharmacol. Exp. Ther, 144, 8 (1964). L. Dykstra, J. Pharmacol. Exp. Ther. 211, 722 (1979). T. L. Yaksh and T. A. Rudy, Eur. J. Pharmacol. 46, 83 (1977). T. L.'Yaksh and T. A. Rudy, Pain 4, 299 (1978). K. Tsou, Ac/a Physiol. Sin. 26, 332 (1963). J. Vigouret, H. Teschemacher, K. Albus, and A. Herz, Neuropharmacology 12, 111 (1973). T. L. Yaksh and T. A. Rudy, J. Pharmacol. Exp. Ther. 202, 411 (1977). Y. F, Jacquet and A. Lajtha, Science 185, 1055 (1974). A, H. Dickenson, J.-L. Oliveras, and J.-M. Besson, Brain Res. 170, 95 (1979). J, C. Yeung and T. A. Rudy, J. Pharmacol. Exp. Ther. 215, 626 (1980). J. C. Yeung and T. A. Rudy, J. Pharmacol. Exp. Ther. 215, 633 (1980). J. D. Levine, S. R. Lane, N. C. Gordon, and H. L. Fields, Brain Res. 236, 85 (1982). R. R. Goodman, S. H. Snyder, M. J. Kuhar, and W, S. Young, Proc. Nail. Acad. Sci, U.S.A. 77, 6239 (1980). A. S. Moskowitz and'R. R. Goodman, J. Neurosci. 4, 1331 (1984). T. Duka, M. Schubert, M. Wuster, R. Stoiber, and A. Herz, Neurosci, Len. 21, 119 (1981). P. E. Gilbert and W. R. Martin, J. Pharmacol. Exp. Ther. 198, 66 (1976). P. Slater and S. Patel, EUr. J. Pharmacol. 92, 159 (1983). A. S. Tung and T. L. Yaksh, Brain Res. 247, 75 (1982), T. L. Yaksh, J. Pharmacol. Exp. Ther. 226, 303 (1983). H. I. Mosberg, R. Hurst, V. J, Hruby, J. J. Galligan, T. F. Burks, K. Gee, and H. I. Yamamura, Biochem. Biophys. Res. Commun. 106, 506 (1982). G. S. F. Ling and G. W. Pasternak, Brain Res. 271, 152 (1983). R. Maura, Neurosci. Lett. 30, 303 (1982). M. B. Tyers, Br. J. Pharmacol. 69, 503 (1980). S. J. Ward and A. E. Takemori, J. Pharmacol. Exp. Ther. 224, 525 (1983), C. Schmauss and T. L. Yaksh, J. Pharmacol. Exp. Ther. 228, 1 (1984). R. A. Mueller, D. B. A. Lundberg, G. R. Breese, J. Hedner, T. Hedner, and J. Jooason, Pharmacol. Rev. 34,255 (1982). H. L. Borison, Pharmacol. Ther. 3, 227 (1977). J. V. Weil, R. E. McCullough, J. S. Kline, and I. E. Sodal, N. Engl. J. Med. 292, 1103 (1975). J. C. Seed, S. L. Wallenstein, R. W. Houde, and J, W, Bellville, Arch. Inl. Pharmacodyn.
72, 265 (1981).
52. C. Chavkin, I. F. James, and A. Goldstein, Science 215, 413 (1982). 53. A. D. Corbett, S. J. Paterson, A. T. McKnight, J. Magnan, and H. W. Kosterlitz, Nature (London) 299, 79 (1982). 54. S. F. Atweh and M. J. Kuhar, Br. Med. Bull. 39,47 (1983). 55. I. Taber, Narcotic Antagonists, M. C. Braude, L. S. Harris, E, L. May, J. P. Smith, and J. E. Villarreal, Adv. Biochem. Psychopharmacol. 8, 191 (1974).
56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67,
36. M.-C. Fournie-Zaluski, G. Gacel, B. Maigret, S. Premilat, and B. P. Roques, Mol. Pharmacol.
References
116, 193 (1958).
J. Florez, A. Mediavilla, and A. Pazos, Brain Res, 199, 197 (1980). J. W. Holaday, Pep/ides 3, 1023 (1982). A. H. Hassen, G. Feuerstein, A, Pfeiffer, and A. I. Faden, Regul. Pep. 4,299 (1982). A. Pfeiffer, G. Feuerstein, 1. J, Kopin, and A. I. Faden, J. Pharmacol. Exp. Ther.225, 735 (1983).
95. K. L. McGilliard and A. E. Takemori, J, Pharmacol. Exp. Ther. 207, 494 (1978). 96. A. Pazos and J. Florez, Eur, J. Pharmacol. 87, 309 (1983).
44
2
Biological Effects of Opioids
3.
97. S.J. Ward and A. E. Takemori. EUT.J. Pharmacal. 87, 1 (1983). 98. G. S. F. Ling, K. Spiegel, S. L. Nishimura, and G. W. Pasternak, EUT.J. Pharmacal . 86, 487 (1983). 99. J. H. Jaffe and W. R. Martin, in "The Pharmacological Basis of Therapeutics" (A. G. Goodman, L. S. Goodman, and A. Gilman, eds.), pp. 494-534. Macmillan New York , 1980. ' 100. M. Weinstock, Narc. Drugs: Biochem. Pharmacal. 394-407 (1971). 101. D. Parolaro, M. Sala, and E. Gorl, Eur. J. Pharmacal. 46, 329 (1977). 102. R. Schulz, M. Wuster, and A. Hen, Naunyn-Schmiedeberg'sArch. Pharmacal. 308 255 (1979). ' 103. J. J. Stewart, N. W. Weisbrodt, and T. F. Burks, J. Pharmacal. Exp. Ther. 205 547 (1978). ' 104. F. Porreca and T. F. Burks, J. Pharmacal, Exp. Ther. 227, 22 (1983). 105. D. E. Burleigh, J. J. Galligan, and T. F. Burks, Eur. J. Pharmaco/. 75,283 (1981). 106. S. J. Ward and A. E. Takemori, J. Pharmaco/. Exp. Ther. 224,359 (1983). 107. WHO Expert Committee on Drug Dependence, WHO Tech. Rep. Ser. 407,5 (1969). )08. N. B. Eddy, H. Halbach, H. Isbell, and M. H. Seever., Bull WHO 32, 723 (1965). 109. W. L. Woolverton and C. R. Schuster, Pharamcol. Rev. 35, 33 1983)..of ,110. Committee on Problems in Drug Dependence, Res. Monogr. er. 52, 15 1984).¥ 111. G. T. Shearman and A. Herz, J. Pharmaco/. Exp. Ther. 2, 1982 . 112. D. W. Hein, S. Herling, A. M. Young, R. J. Valentino, and J. H. Woods, Fed. Proc., Fed. Am. Soc. Exp. BioI. 39, 995 (1980). 113. S. Herling and J. H. Woods, Ufe Sci. 28, 1571(1981). 114. J. H. Woods, C. L. Fly, and H. H. Swain, in "Characteristics and Functions of O?ioids:' (1. Van Ree and L. Terenius, eds.), pp. 403-411. Elsevier/North-Holland BIOmedical Press, Amsterdam, 1978. 115. J. H. Woods, C. B. Smith, F. Medzihradsky, and H. H. Swain, in "Mechanisms of Pain and Analgesic Compounds" (R. F. Beers, Jr., and E. G. Bassett, cds.), pp. 429-445. Raven, New York, 1979.
Synthesis and Structure-Activity Relationships Morphine, Codeine, and Related Alkaloids I. Syntheses of Morphine, Codeine, and Related Alkaloids . . . . II. The Structure-Activity Relationships of Morphine and Related Compounds. . . . . . . .. ...... A. Alteration of Existing Functional Groups and Structures on Morphine . . . . . . . . . . . . . . . . . B. Insertion of Substituents in Nonfunctionalized areas . III. Diels-Alder AdductsofThebaine . . . .. .... A. Ketone, Sulfone, Nitroso. Ester. and Nitrile Adducts B. Functionalization at C-19: Alcohols . . C. Opiate Receptor Probes. . . . . . . . . IV. The Chemical Anatomy of Morphine and Its Derivatives A. The Chemical Anatomy of Morphine . . . . B. The Chemical Anatomy of Diels-Alder Adducts References
I.
of
45 55 56 79 101 103 126 153 155 155 155 157
Synthesis of Morphine, Codeine. and Related Alkaloids
Morphine and codeine have been the subject of numerous successful and unsuccessful synthetic studies since the correct formulation of the gross structure of morphine and codeine by Gulland and Robinson (1). The relatively complex, unsymmetrical structures of morphine and codeine were first constructed by the heroic synthetic efforts of Gates and Tschudi (2). A second, contemporary synthesis, which is not described here, also involved the construction of a suitable hydrophenanthrene and subsequent elaboration of the bridging amine ring (3). For a delightful discussion of these two syntheses by the author of the latter, see ref. (4). Subsequent successful syntheses usually produced dihydrothebainone, which had previously been converted to codeine and then to morphine by Gates and Tschudi, or other intermediates previously converted to these alkaloids. Until recently, the syntheses of these complex alkaloids have been primarily academic exercises aimed at developing new synthetic methodology, not commercial attractiveness. The shortfall in supply in the early 1970s stimulated new efforts to design a commercially viable synthesis of morphine and codeine. The first synthesis of morphine by Gates is illustrated in Scheme 3-1 (2). It started with the readily available dye intermediate, 2,6-dihydroxy-
t
45
46
3
Synthesisof Morphine,
Codeine,
and
Alkaloids
Related
I
Synthesis
of Morphine,
Codeine,
and Related
Alkaloids
47
NO
OH
OH
a,b }
c,d,e
HO
p,q
>
63%
r.s.t 24%
HO
48%
C6H5C02 1 0
HO OCH3
OH
+ 7-hydroxy isomer
OCR3 f
JQJ~OH
q,b,C,d,e>
)
80%
C6H5C02 c,r,s,u,t ) 6%
OCH3 OCH3 g 97%
h,i ) 83%
0 0
0 v 34%
0 j ) 66% CH30 CH30
CH30
k 50%
3
naphthalene (1) and involved the preparation of dienophilic ortho-quinone (2), which eventually comprised the A- and B-rings of morphine. The dienophile (2) was obtained in about 20% overall yield starting from 1. A Diels-Alder reaction of 2 with butadiene generated the hydrophenanthrene (3). The Diels-Alder adduct (3) contained a cyan om ethylene group in the proper array for eventual transformation into the nitrogencontaining ring of morphine. Furthermore, as a result of the Diels-Alder reaction, a strategically placed double bond was introduced. This double bond served as the key for the introduction of the allylic hydroxyl group in ring C. Reduction of 3 led to saturation of the enolic double bond,
CH30
)
Morphine
Scheme 3.1. (cont.) Reagents: p, potassium hydroxide, diethylene glycol, 225°C; q, potassium tert-butoxide, benzophenone; T, bromine; s, 2,4-dinitrophenylhydrazine; t, acetone, HCI; u, pyridine, beat; v, pyridine hydrochloride, 220°C.
CH30
1,m,n 79%
)
)
Codeine 2
4
0
5
)
m ) 65%
o 28%
>
Scheme 3-1. The first synthesis of morphine by Gates.1952. Reagents: a, benzoyl chloride, pyridine; b, nitrous acid; c, PdlC, H2; d, ferric chloride; e, sulfur dioxide; f, dimethyl sulfate, potassium carbonate; g, hydroxide; h, ethyl cyanoacetate, triethylamine; i, potassium ferri cyanide; j, butadiene; k, copper chromite; I, Wolff-Kishner; m,lithium H2' aluminum hydride; n, Eschweiler-Clarke; 0, dilute sulfuric acid.
r
.,. 48
3
Synthesis
of Morphine,
Codeine,
and Related
Alkaloids
fonnation of an imino-lactone, and an unprecedented free radical rearrangement to form the lactam (4). This lactam contained four of the five rings found in codeine and morphine. Unfortunately, at this point, as a result of the reduction, the incorrect stereochemistry was generated at C-14 (morphine numbering). Subsequent steps adjusted the oxidation level of the molecule that ultimately led to the enone (5). Introduction of the enone labilized the proton at C-14, leading to the thermodynamically preferred cis-ring fusion. Closure of the furan ring and reduction of the ketone furnished codeine, which was demethylated in molten pyridine hydrochloride to yield morphine, some 150 years after its isolation from the opium poppy. The overall yield for the synthesis shown in Scheme 3-1 was 3.5 x 10-3%. Subsequent efforts at the total synthesis of the opium alkaloid have utilized either a biomimetic route in which a phenolic 1benzyltetrahydroisoquinoline is oxidized to a morphinandienone that is convertible to thebaine and thence to morphine, or an electrophilic reaction of a suitably functionalized, partially hydrogenated isoquinoline to form the morphinan carbon-nitrogen skeleton. Although not all known syntheses of morphinan will be discussed, selected examples of the above approaches are illustrated. With the biomimetric approach, the control of product distribution ratios and the sensitivity of the morphinan-dienone to subsequent transformations have been frequent and major problems. An example of this approach is illustrated in Scheme 3-2 (5). In this synthesis, a benzyltetrahydro-isoquinoline, N-carbethoxynorreticuline (6), is oxidized with thallium(lII) to give the salutaridine analog (7), a morphinandienone. Hydride reduction then gives the epimeric salutaridinols (8), which are subsequently dehydrated in unspecified yield to form racemic thebaine. It has since been demonstrated that a variety of oxidizing agents and nitrogenprotecting groups are compatible with morphinandienone fonnation in reasonable yields (6). However, as illustrated in Scheme 3-3, attempts at salutaridine (10) formation using a Pschorr reaction on the aminobenzylisoquinoline (9) gave vanishingly small yields (7). The entire concept of a biomimetic approach had previously been investigated using optically active material derived from opium (8). In the alternative approach to morphine alkaloid synthesis, the major method for construction of an azacarbocyclic ring system in morphine depends on the Grewe cyclization (9). The transformation of 12 into 13 in Scheme 3-4 is an example of this type of electrophilic subsfitution. In this approach, dihydrothebainone (15) has been the usual objective. However, predominant cyclization to morphinans with the wrong oxygenation pattern, lengthy synthetic routes requiring symmetrical I-benzyl substituents,
Synthesis of Morphine, Codeine. and Related Alkaloids
a 23%
49
> o 7
6
OH
8 Scheme
3.2.
isoquinoline. (c) hydrochloric
A biomimetic Reagents:
Thebaine synthesis
(a) 1.0equiv.
of thebaine
thallium
proceeding
trifluoroacetate;
from
a I-benzyltetrahydro-
(b) lithium
aluminum
hydride;
acid.
9 Scheme 3-3.
10 Salutaridine Reagents: (a) sodium nitrite, N-sulfuric acid; (b) 70"C.
and the failure of the reaction with compounds substituted to generate the correct oxygenation pattern have hindered synthetic efforts in this area. Methods have now been developed to circumvent many of these problems (10).
50
3
,
Synthesis
a,b 87%
of Morphine,
Codeine,
and Related
Alkaloids
) -
HO
OH 12
11
C6H5 I
N-N OH
O~
II
N_N
T
I
Synthesis
of Morphine,
would
be converted
dihydrocodeinone e ~ 75%
>
and Related
51
Alkaloids
C-2. The extra phenolic group is removed by selective use of the Musliner-Gates reaction (12) to generate optically active dihydrothebainone (15), which is then converted to codeine and morphine. A major problem with this approach is the difficult synthesis of the appropriately substituted phenylacetic acid precursor to compound (11). The most straightforward approach to the synthesis of opium alkaloids achieved so far is shown in Scheme 3-5. This synthesis illustrates the successful execution of a modified Grewe-type synthesis of optically active dihydrothebainone (15), codeine, and morphine in an overall yield of 15-20% from meta-methoxyphenylethylamine (16) (13). Racemic tetrahydroisoquinoline (17) was readily resolved with tartaric acid into its optical antipodes. The undesired enantiomer could be recycled readily (14). After Birch reduction, the amino group was protected as its formamide derivative (18). Bromination ensured that 19 would cyclize correctly to form 20, which
d 45%
Codeine,
readily
to
(- )-dihydrothebainone (15), (-)-
(21), and (- )-nordihydrocodeinone
(22). The versatility CH30
o
CH3O""~NH2
'+ CH30
0 J9l
CO H 2
.:,
)
lQj)H d'
OH
14
13
16
~'4%
b,c 15
(-) -Cihydrothebainone
'--7 86'
OH Scheme 3-4. Grewe-type syntheses of morphine alkaloids using symmetrically substituted 1.benzylisoquinolines. Reagents; a, hydrogen, Pt/C, fonnaldehyde; b, lithium/ammonia; c, hydrochloric acid; d, 5-chloro-l-phenyltetrazole, potassium carbonate; e, hydrogen, Pd/C.
When a symmetrically substituted I-benzylisoquinoline is used to enter the morphine alkaloid series, it is necessary ultimately to remove the extra substituent. An example of such a process is illustrated in Scheme 3.4 (11). The symmetrical, protected phenolic hydroxyl groups in the 3' ,5' -positions of the I-benzyl substituent in the optically active 11 ensure the location of the hydroxyl group in the correct position in 13. After 11 is subjected to the Birch reduction, the resultant (12) cyclizes to the morphinan derivative (13), which has the required phenolic hydroxyl at C-4 and an extra one at
o
---4
~
OH 60%
OH '6'
18
19
20
Scheme 3-5. A potentially commercial total synthesis of morphine and codeine. Reagents: a, 200°C, neat, argon; b, phosphorus oxychloride, acetonitrile; c, pH 4-5, NaCNBH3; d, Lijammonia; e, phenyl formate; C, methane sulfonic acid, ethylene glycol; g, N. bromoacetamide;
h, aqueous
formic acid; i, triftuoromethane
sulfonic
acid.
52
j
3
Synthesis of Morphine.
Codeine, and Related Alkaloids
Synthesis
of Morphine,
Codeine,
and Related
53
Alkaloids
~
)
92%
100%
80%
15
6
c,d,e ~
~4B% N
23
I CH3
Dihydrothebainone
24
I CH3
~::
OCH3
@:
n,g,o,h,p 67%
>
=
f
OCH3
9
)
22 Nordihydrocodeinone
21 Dihydrocodeinone
)
95%
60%
H 26
H 25
Codeine
OCH3
q
)
@:
Morphine
90%
h Reagents: j, hydrochloric acid; k, bromine/acetic acid; I, sodium hydroxide; m, hydrogen, Pd/C in the presence of acqueous formaldehyde; n, trimethylorthoformate, acid; 0. potassiumtert-butoxide, DMSO; p, lithium aluminum hydride; q, boron Scheme
3.5.
(cont.)
95%
>
i BO%
>
-
CHO
tribromide.
of this approach allows the preparation of both enantiomers in the nor series of codeine and morphine, as well as a wide variety of agonists and antagonists, by functionalization of the secondary amine in 22. Improved procedures allow the ready preparation of codeine from either IS or 21 (15), and a rapid, high-yielding de methylation then gives morphine (16,17). The overall yields of natural (or unnatural) morphine, codeine, and thebaine are about 25% from meta-methoxyphenylethylamine with only 6-8 isolated intermediates and the reactions have been run on a large scale (16). An alternative general approach to morphine-based analgesics that proceeds from 4-arylpiperidines has been described (18,19). The arylpiperidines themselves function as analgesics. They are subsequently converted into octahydroisoquinolines possessing a phenyl-substituted bridgehead position (20). The synthesis is illustrated in Scheme 3-6. The
~OCH3
H2C
CH30
j,k,l
) )CH3O
)
B1%
2B
Morphine
Scheme 3-6. The 4a-phenyloctahydroisoquinoline route to the opiu~ alkaloids. Re~g~nts: a, 2,3-dimethoxyphenyllithium; b, p-toluene sul~onic acid, tolue~e, 11.0 ~; c, n-butYII.lthIU~~ d, H2C=C(CH2Br)CH2CH2Br;e, sodium iodide; .f, perchlonc. aCid m met~anol, .g, azomethane; h, dimethyl sulfoxide; i, boron triftuonde: -urc.; J, mes~1 chlonde, tnethy 1amine; k, lithium triethylborohydride;
I, osmium tetroxide, sodium penodate.
54
3
Synthesis
of Morphine,
Codeine,
and Related
OH
Alkaloids
OR2
29
30
R1
R2
H
31
R1
CH3,
R2
32
R1
R2
.
-
CH3CO
H
II
The Structure-Activity
Relationships
of Morphine
and Related
Compounds
55
arylpiperidine (23) is deprotonated to form the enamine anion, which condenses with an allylic dibromide to generate the octahydroisoquinoline (24). Enamine protonation under kinetic control yields the cis ring-fused iminium salt (25), which, after conversion to the aziridinium cation (26), is oxidized and then cyclized to the morphinan derivative (27). This synthetic sequence generates the incorrect trans-stereochemistry at the BC-ring junction in 27. After conversion of the exo-methylene group into a strategically placed ketone in 28, classical transformations as shown in Scheme 3-1 allow epimerization to the thermodynamically more stable cis ring fusion and then conversion to morphine and codeine alkaloids. This sequence is interesting because it appears that only compounds containing the exocyclic methylene group like 26 undergo cyclization. Compounds containing oxygenated functions fail (21).
R1
II, 40
R1
B,
41
R1
R2 = H
42
R1
OCH3,
43
R1
OCH2CH3,
44
R1
OCH2C6BS'
R2
OH
4S
R1
OC(CH3)3'
R2
OH
48
R1
02CCH3'
49
R
H
OH
R2
R2 - OH R2 = OB
R2
.
OH
46
The Structure-Activity Relationships of Morphine and Related Compounds
The use of opium for a variety of medical disorders and analgesia can be traced to the beginning of prehistory. However, attempts to treat pain with discrete chemicals began about 200 years ago with the isolation of morphine from opium. Although the addiction liability and the toxicity of morphine were recognized early on, it was the invention of the hypodermic syringe by Wood in 1853 and the subsequent abuse of parenteral morphine that illustrated the social problems of this drug. These and related events initiated the search for a safe, nonaddicting opioid. Even before the correct structure of morphine was known, chemical investigations were begun to alter its pharmacological effects. One of the earliest derivatives was the diacetyl compound heroin, the heroic drug, which was initially introduced as an antidote to morphine addiction (22). Over time, clinical experience invalidated this claim. In subsequent years, a variety of other morphine derivatives were introduced, but none were demonstrably superior to morphine. In 1929, the first systematic studies of the structure-activity relationships (SAR) among derivatives of the opium alkaloids were begun under the direction of the Committee on Drug Addiction of the National Research Council (23). This early work formed the data base for the eventual semi-systematic investigation of the opioid SAR. The SAR of morphine and related derivatives have been reviewed extensively over the years (24-32).
56
3
Synthesis of Morphine. Codeine. and Related Alkaloids
II
The Structure-Activity
Relationships
\7
of Morphine and Related Compounds
A. Alteration of Existing Functional Groups and Structures on Morphine In examining the structure of morphine, besides its optical mirror image, substituents that can be modified without directly affecting the basic nucleus are the phenolic function at C-3, the alcohol at C-6, the double bond at C-7, and the methyl group on the basic amine. The stereochemistry of these substituents and the stereochemistry at the BC-ring junction have also been investigated. Although these groups and stereochemical relationships have not been varied systematically to determine SAR correlations, a suffiGient data base currently exists to create one. 1. Enantiomeric (+ )-Morphine and Its Analogs To define the enantiomeric requirements of morphine's ability to produce analgesia and to interact with its receptors, the (+ )-enantiomer of morphine was synthesized (33). The absolute configuration of the morphinan skeleton of naturally occurring (- )-sinomenine (29) is enantiomeric to natural (-)morphine. The synthetic pathway proceeding from (29) also allows the preparation of (+ )-codeine (31) and (+ )-heroin (32), as well as (+)morphine (30) (33,34). Since sinomenine (29) is a rare alkaloid and is attainable only with difficulty, a subsequent total synthesis has been developed (13). The unnatural enantiomers of the alkaloids have shown no analgesic activity in standard screening tests for centrally acting analgesics, and (+ )-morphine (30) has minimal opiate receptor affinity (35). 2. Trans-Morphine Morphine possesses the cis-decalin type of junction between rings Band C, while the aromatic ring and ether linkage force the Coring into the boat conformation. In the simplified opiate analgesics such as the benzomorphans, conversion of the cis-ring fusion to the trans-decal in type of ring system furnishes superior analgesics when compared to their respective cis-fused isomers. As a result, the synthesis of trans-morphine (33) was undertaken to prepare a more potent morphinebased analgesic. However, it can be readily seen that the introduction of the BC-trans ring fusion would severely distort the shape of the morphine ring system. The morphine alkaloid isoneopine (34), which contains unsaturation at position 8, serves as the starting material for the preparation of the desired trans-morphine (Scheme 3-7). The 6-,B-tosylate (35) slowly undergoes hydroboration, which after oxidative workup gives the 8-a-alcohol (36) and, more importantly, generates the desired BC-trans ring fusion (36). Solvolysis of the tosylate (36) in the presence of lithium carbonate yields the 6-a-alcohol (37), which is the naturally occurring configuration. Because the 6-a-alcohol is hindered, it is possible to tosylate the 8-a-alcohol selectively, giving 38, which, after elimination, introduces
a
b,c )
)
OB
CH30 IBoneopine 34
35
d
a )
> CH30
37
36
f
e
)
> CH30
CH30 38
HO
0""
--
OH trans-codeine 39
OH trans-morphine 33
Scheme 3.7. The first synthesis of trans-morphine. Reagents: at tosyl chloride. pyridine; b, diborane; c, hydrogen peroxide; d, lithium carbonate in reftuxing dimethylformamide; e, refluxing 2,4,6-collidine; f, lithium diphenylphosphide.
the requisite double bond at position 7, forming trans-codeine (39). The standard demethylation procedure employing pyridine hydrochlonde gives only traces of the desired trans-morphine. The use of the diphenylphosphide anion allows the isolation of trans-morphine (33), but in a yield of only 16%. The overall yield of trans-codeine (39) from isoneopine (34) ISa respectable 10% (37). A substantial improvement was made when It was
58
3
Synthesis
of Morphine,
Codeine,
and Related
Alkaloids
Table 3-]
and Related
of Morphine
Compounds
59
EDso"
trans-Morphine
R
Compound
,.
Relationships
Table 3-2
Morphine
Codeine Trans-Codeine Morphine Trans-Morphine
II The Structure-Activity
42 39 33
mpk, sc 7.5 17.5 1.2 11.7
CH, (47) H CH,CH,
A 0.5 0.6 0.3
R CH) H CH,CH,
A 0.5 0.3 0.3
.. mpk, sc in mouse hot plate assay.
Hot plate activity.
discovered that bis-tosylate of diol (34), prepared from the tosylate of isoneopine, undergoes solvolysis and elimination with potassium acetate in refluxing dimethylformamide to yield trans-codeine directly in three steps from isoneopine, with an overall yield of 30% (38). In distinct contrast to simpler systems, conversion of the naturally occurring cis-ring fusion in morphine to the trans-ring fusion provided compounds with disappointing analgesic activity (Table 3-1). For instance, trans-codeine was about half as potent as codeine in the hot plate assay. More significantly, the analgesic activity of trans-morphine was only 1/lOth that of morphine itself. The rather severe distortion of the C-ring in trans-morphine was thus not consistent with enhanced analgesic activity (38,39). However, in the morphinans (q.v.), the trans stereochemistry generally yields more potent analgesics than the cis. 3. The Phenolic Function at C-3 The phenolic group at C-3 in morphine has been assumed to be important to its biological activity and its ability to bind to the opiate receptor. This concept has been tested by the ~eductive elimination of the phenolic group and also the 6-hydroxyl group 10 morphme (40). The 3-deoxy compound (40) is about one-third as potent an analgesic as morphine while possessing only 1/30th of its receptorbinding affinity. Surprisingly, the 3,6-dideoxy derivative (41) is equipotent as an analgesic and retains one-third of the receptor-binding affinity of morphine. The results indicate that the phenolic hydroxyl group generally has a greater effect on receptor binding than on analgesia and that the 3-phenolic group is not essential for analgesic activity.
4. O-Alkyl and Acyl Substitution at C-3 An intact A-ring is, in general, necessary for analgesic activity. Masking of the phenolic hydroxyl by etherification or esterification generally causes a decrease in morphine analgesic activity, with heroin being a notable exception. Methylation yields codeine (42), which is a weaker analgesic than morphine but is widely used for the relief of mild to moderate pain and as an antitussive. Codeine is orally active, and its analgesic activity is attributed to its metabolic conversion to morphine (41). In terms of potency, codeine is about 1/lOth as active as morphine when administered subcutaneously in the mouse tail flick test. The ethyl ether (43) (Dionin) is somewhat more potent than codeine, while the benzyl ether (44) is intermediate between codeine and morphine (25). The derivative (45), containing a bulky tertiary butyl ether, is inactive in the tail flick assay (42). Alkylation of morphine with N-(2-chloroethyl) morpholine generates pholcodeine (46), which is valuable as an antitussive and is superior to codeine as a centrally acting sedative. Although masking of the phenolic group results in a reduction of its morphine-like effects, the opposite is true of the alcoholic function at C-6; therefore, the diacetyl derivative (47) (heroin, diamorphine) has at least twice the potency of morphine. Heroin (47) and its lower acylhomologs have similar analgesic potencies in mice and high physical dependence liabilities in monkeys (Table 3-2) (43). The activities of 6-monoacylated derivatives demonstrate little difference from those of the diesters; all compounds are two- to fourfold more potent than morphine. In contrast, 3-acetylmorphine (48) and morphine are equivalent (44). These observations indicate that rapid deacylation of heroin occurs at the phenolIc
3
60
Synthesis
of Morphine,
Codeine,
and Related
a
II
The Structure-Activity
b
----7 RO
Alkaloids
of Morphine
and Related
CH30 52
RO
c
d,e
----7
R R
CH3 CH2OCH3
55
58
b
1
RO 52 54
R
CH2OCH3
RO
~ 51 R 53 R 56 R
Scheme 3-8.
Scheme 3.9. diazomethane,
57
Synthesis of codeine, morphine, and heroin epoxides.
Reagents: a, methoxymethyl chloride; b, silver carbonate; c, hydrogen peroxide, sodium hydroxide; d, sodium borohydride; e, 1 N hydrochloric acid; C, acetic anhydride, pyridine.
position in the plasma and that the resultant species that penentrates the CNS (45).
6t
----;> CH30
-----0>
Compounds
a
----7 RO
Relationships
6-acetylmorphine
is the
5. The Importance of the 7,S-Double Bond The catalytic reduction of both morphine and codeine with hydrogen to yield dihydromorphine (49) and dihydrocodeine (50) is a straightforward reaction (46). The resultant dihydro compounds possess equivalent or slightly increased analgesic potency compared to their unsaturated parents. For instance, dihydromorphine (49) is equipotent to morphine in mice and has approximately 2.5 times the potency in cats. From these observations, it is readily apparent that unsaturation at C-7 is not essential for analgesic activity (47). Codeine j3-7,8-epoxide (51) is a minor metabolite of codeine (4S) and is readily prepared by base-catalyzed addition-elimination of hydrogen peroxide to codeinone (52) (Scheme 3-8) (49). Several attempts to synthesize
Synthesis of 713, 813-cycJopropylcodeine. Pd(II); c, sodium borohydride.
~ Reagents:
a, (CH3hSOCH2
-; b,
the equivalent morphine j3-7,8-epoxide (53) failed due to A-ring degradation (50). Successful synthesis of the desired epoxide started with protection of the phenolic hydroxyl of morphine as its methoxymethyl ether and subsequent oxidation of the 6-hydroxyl to the ketone (54). Base-catalyzed epoxidation with hydrogen peroxide then gave the epoxyenone (55). Reduction with sodium borohydride gave the 6a-hydroxy-7j3-8j3epoxide (56), which, after removal of the protecting group, gave morphine j3-epoxide (53). Acetylation of 53 gave heroin epoxide (57) (51). All of the epoxides demonstrated dose-dependent analgesic activity. Not surprisingly, heroin epoxide (57) was the most potent, and all three epoxides (51, 53, 57) were approximately twice as potent as their unsaturated parents. The similarity of ole fins and cyclopropane derivatives with regard to their chemical properties (52) and extrapolation to their biological effects led to the replacement of the 7,8-double bond with a cyclopropane ring. The addition of dimethyloxosulfonium methylide to conjugated enones to yield cyclopropylketones is a general reaction of wide applicability (53). However, when this reagent was applied to codeinone (52) (Scheme 3-9), only methylene transfer to the ketone to form the exocyclic oxirane (58)
62
3
Synthesis
of Morphine,
Codeine,
and Related
Alkaloids
was observed (54). The oxirane (58) was equipotent to codeine in the mouse hot plate test. Subsequent attempts to add various carbenes or dihalocarbenes to codeinone (52) were unsuccessful. The desired transformation was ultimately achieved by the palladium.catalyzed addition of diazomethane to 52 to form the 7/3,S/3-cyclopropylcodeinone (59), which could be reduced to 7/3,S/3-cyclopropylcodeine (60) with sodium borohydride (55). The cyclopropylcodeine (60) and codeinone (59) derivatives were tested for analgesic activity in both the peripheral mouse writhing assay and the central rat tail flick assay and were compared with codeine. Peripherally, cyclopropylcodeine (60) is approximately five times more active than codeine, while centrally it is inactive in analgesic doses of codeine. On the other hand, cyclopropylcodeinone (59) is approximately 20-fold more active than codeine both peripherally and centrally. The reduction of the 7,S-double bond or replacement with epoxy or cyclopropyl isosteres thus appears, for the most part, to be consistent with the retention of analgesic activity. Minimally, these substitutions retain the original activity and can increase it slightly to several fold. 6. The 6-Hydroxyl Group: Its Epimers, Isosteres, and Replacements A significant amount of work has been done on position 6 of morphine in addition to epimerization of the alcohol, acylation, and etherification. This includes oxidation to the ketone, replacement by halogen, and complete removal of the oxygen. The majority of these studies are included in the original work by Eddy and Mosettig (23). All of these compounds have analgesic activity, and they usually have greater analgesic potency and dependence liability than morphine. Additionally, many of these early analogs also possess a reduced 7,S-double bond. The analogs are presented in Table 3-3. The synthesis of the majority of these compounds is usually straightforward. Because of its Qigh potency, significant amounts of the 2:meth lene dihydro compound (61) have been required, but because the initial synt eSls was not capable of producing the amounts necessary for detailed pharmacological investigations, a new synthesis was developed I(.56j1 Protection of the phenolic hydroxyl as its methoxymethyl ether in dmYdromorphinone (62) allowed the condensation with the Wittig reagent, methylenetriphenylphosphorane, to generate the 6-methylene group with an overall yield of 'il'2P after deblocking. Detailed investigation has indicated that 61 has a more rapid onset of action than morphine, and tolerance to the analgesic and sedative effects develop more slowly and to a lesser degree than with morphine. Arterial pressure effects and intestinal motility are also lessened, but both 61 and morphine have about the same respiratory depressant activity.
II
The Structure-Activity
Relationships
and Related
of Morphine
Compounds
63
Table 3-3 The
Effect
of C-6 Substitution
on Analgesic
Potency
a
7 6{1
HO Double Bond
C-6 Substituent
7.8
a-OH /J-OH a-OCH) /J-OCH, =0
6.7
H
CH) CH:zCH) n-C4H<)
C6HS Saturated
a Adapted,
H a-OH /J-OH a-OCH) /J-OCH, =0 ~CH2 fJ-CH, in part,
Analgesic Potency
Species
1 (morphine)
Mouse Mouse Mouse Cat Mice
0.4 6 0.5 0.3
8. 15.
Rat
4 5 2
Rat Rat Rat Rat
8 2.5 0.7 4 JO' 4 60' 50'
Cat Cat Cat Cat Cat Cat Cat Cat
from Table 2.2 of ref. 9.
Because the hydroxy group and the double bond in the Coring constitute an allylic alcohol, the rearrangements in this system have only relatively recently been untangled. The reactions of nucleophiles with codeine tosylate, as well as pseudocodeine tosylate, have been determined kinetically (57). Codeine tosylate (63), when treated with lithium chloride in dimethylformamide at 40°C, undergoes an Sn2 reaction to give the a-chlorocodide (64), which rearranges by an Sn1' mechanism to form /3-chlorocodide (65) when heated at 120°C in the same solvent. The structures of these products have been rigorously defined using nuclear magnetic resonance (NMR) techniques, thus eliminating much of the confusion in the earlier literature (58). Pseudocodeine tosylate (66) reacts
3
64
Synthesis
of Morphine,
Codeine,
and Related
Alkaloids
a,b ':>53%
62
61
a
64
II
The Structure-Activity
Relationships
of Morphine
and Related
Compounds
opiate receptors in living animals. This occupancy is reversed by (-)_ naloxone but not (+ )-naloxone (266). The pharmacologically rare azido substituent has been introduced into the 6-position of dihydromorphine (49) by an inversion process involving conversion of the 6a-alcohol to the tosylate and displacement with azide to yield 6-deoxy-6j3-azidomorphine (67) (59). This compound is a remarkably potent analgesic, being 150-300 times as potent as morphine in the rat and 50 times as potent as morphine in humans. It is also less toxic than morphine, with a low physical dependence liability in rats and monkeys, although withdrawal symptoms follow its use in humans (60). Clinical studies indicated that azidomorphine (67) functions as a typical opiate-like drug in humans. It causes pupil constriction, subjective effects, and morphine-like euphoria, yet suppresses the morphine abstinence syndrome (61). Other azido derivatives studied were 6-deoxy-6j3-azidocodeine (68), which is 13 times as powerful as morphine in the rat hot plate assay (62), and 6-deoxy-6j3-azido-14j3-hydroxymorphine (69), which has properties similar to those of 67 in humans and is also an effective antitussive (63).
67
69
68 65 Scheme 3-10. Reagents: (top): (a) methylenetriphenylphosphorance, (a) LiCI, DMF, 40°C, (b) DMF, 120"C.
66 (b) acid; (bottom):
with weak nucleophiles by an S.l' mechanism [e.g., chloride can give a-chlorocodide (64)] and with more powerful nucleophiles by an S.l mechanism accompanied by retention of configuration (Scheme 3-10). Removal of the methoxyl methyl generates the a.chloromorphide and j3-chloromorphide analogs of 64 and 65, respectively. The a-chloromorphide is about 15 times more potent than morphine but is considerably more toxic, while the j3-chloromorphide is equipotent (58). An exciting extension of this work is the displacement of the 6atrillate with 18F to provide ligands in both the agonist and antagonist series suitable for positron emission tomography (PET) (265). An [18F]analog of naltrexone (ef. Section II,A,7) has been used to visualize
65
70
71
(X
N-NH2)
72
(X
0)
73
(X
Nt2)
Dihydromorphinone (70) is readily available from morphine by reduc. tion and oxidation, or alternatively by a novel acid-catalyzed rearrangement of the allylic alcohol to the enol ether and subsequent tautomeri-
66
3
Synthesis
of Morphine.
Codeine,
and Related
Alkaloids
zation to the 6-ketone (64). With the wide variety of phosphorus and sulfur ylids now available, it is surprising that more use of them has not been made to introduce substituted methylene groups at C-6. A carbethoxymethylene group introduced at C-6 in the antagonist morphine series was disappointing in its biological properties (65). However, these a, 13unsaturated compounds are interesting because they possess the potential for irreversible receptor binding by 1,4-conjugate addition to the acrylate ester group. Another approach used to create irreversibly binding receptor ligands is to prepare the hydrazone of oxymorphone. This simple derivative (71), readily prepared by the reaction of excess oxymorphone with hydrazine, when given in vivo produced a significant inhibition of receptor binding for over 24 hours despite extensive washing of the brain membrane homogenates. The parent oxymorphone (72) had no effect after this time period. Significantly, 24 hours after in vivo administration of 71, analgesia was still demonstrable in mice using the tail flick assay, while oxymorphone (72) analgesia had ceased. The mechanism for the long-lasting effects is unknown, since no evidence of covalent bonding was demonstrated (66a). The hydrazone exists as a mixture of syn and anti isomers (66b). Subsequently, it was found that the hydrazones comprising 71 readily and rapidly disproportionate to the azine (73) in weak acid (67). The azine, oxymorphazine (73), is 20-40 times more potent than the corresponding hydrazone while retaining the same long-lasting effects. Although covalent binding to the opiate receptor may be possible for the azines, the potential for the noncovalently bound drug to dissociate very slowly has been noted. Very slow dissociation would appear to be irreversible binding in a practical sense. Since the azine (73) is a bifunctional molecule, it could bind to two receptor sites at once, greatly enhancing the affinity and thereby decreasing the rate of dissociation. Further studies on the occupation of two receptor sites by two morphinebased molecules covalently linked by a spacer molecule have used the 6-amino group. The original method for reductive amination of a 6-ketone yields a 1:2 mixture of the 613- and 6a-amino epimers (68). Subsequent investigation has yielded stereospecific synthesis of both epimers (Scheme 3-11). Preparation
of the iminium
salt (74, R
= benzyl)
forces the Coring
into a boat conformation due to steric repulsion between the vicinal ether oxygen and the syn-benzyl group. With the Coring in the boat conformation, hydride transfer occurs exclusively to the more accessible a-face, thereby furnishing the 6f3-isomer (75). Catalytic debenzylation then yields the stereochemically pure 6f3-oxymorphamine (76). The 6a-amine (78) is prepared by stereospecific catalytic reduction of the imine (77) prepared from the 6-ketone (72) and benzylamine. The imine retains the energeti-
II
The Structure-Activity
Relationships
of Morphine
and Related
Compounds
67
R
~ RN
a ~
74
(R
-
d,e 72 oxymorphone
R
--;>
benzyl>
75
(R.
benzyl>
76
(R.
HJ
77
78
Scheme 3.) J. Synthesis of the epimeric 6-amino-oxymorphones. Reagents: a. dibenzy). amine; b. sodium cyanoborohydride; c, PdjC, hydrogen;d. benzyl amine; e, sodium boro. hydride.
cally more favorable chair conformation and therefore results in hydrogenation on the f3-face to yield the 6a-amine (69). Solution NMR studies indicate that in the 6a-oxymorphamine, the Coring exists in the familiar boat configuration, while in the 6f3-oxymorphamine, the Coring exists in the chair conformation (70). Biological activity has been reported for the 6-amino epimers in the narcotic antagonist series (68). Bivalent ligands were synthesized where two molecules of (76) were linked by variablelength spacer groups consisting of succinyl-bis-oligoglycine (79). The greatest receptor affinity potency difference occurred when the spacer contained two glycine residues (79, n = 2). Bridging of neighboring opiate receptors therefore occurs when the spacer group possesses a linear length of 18
A.
On the presumption
that the marked
enhancement
of receptor
affinity is a consequence of bridging, it is likely that the linear spacer length is substantially greater than the interreceptor distance, because sufficient translational mobility of the oxymorphamine must be required for docking
3
68
Synthesis of Morphine,
(n
79
Codeine,
II
and Related Alkaloids
0-4)
~
to the initial receptor recognition site. Moreover, the optimal spacer length for effective receptor interaction with both pharmacophores should be dependent on the relative orientation of these neighboring receptors. Both factors indicate that the average distance between occupied receptors is less than 18 A (71). The increase in receptor affinity potency for bivalent ligands with neighboring receptors is a function of entropic factors, because of the restriction 'of the univalently bound bivalent ligand within the interactive volume of the neighboring vacant receptor (Scheme 3-12). Therefore,
(L 99 I I..
99 II
00
~)·
00
99
99 ,,
00
00
II
receptor
singlvoccupied receptor
A
B
unoccupied
00, ~) . ' .
(
L) 00 , I
I I
"
00
00 independently singly occupied receptor.
i
1
() 00
99
II II
II
00
00 bridged occupied receptors
"
l
c l:::
spaced bivalent ligard
Scheme
3-12.
Bridging of spaced bivalent opiate ligands (according to ref. 71).
,
I
The Structure-Activity
Relationships of Morphine and Related Compounds
69
proceeding from state B to state C has to be favored over univalent binding of a second bivalent ligand, provided the spacer length allows bridging. In the extreme case, a bivalent ligand with an excessively long spacer (e.g., a polymer) would be expected to possess a potency that approximates a statistical factor of 2 over the monovalent analog due to the much larger accessible volume of the residual free pharmacophore. In the extreme case, therefore, the infinitely spaced pharmacophores function independently. The previously described qualitative picture can be understood in terms of the Gibbs free energy equation (toGo = toHO - T toSo). The univalent binding of a second bivalent ligand occurs with a greater negative entropy change (toSo) than the bridging of a univalently bound divalent ligand, which has a smaller containment volume. Thus, the free energy change (toGo) in the latter case should be more negative relative to the former one. This assumes that the enthalpy change (toHo) accompanying the univalent binding of two bivalent ligands does not differ substantially from that involved in the bridging of a single bivalent ligand to two independent neighboring receptors. With this reasonable assumption, the more favorable entropy change is responsible for potency enhancement when bridging can occur (71). A successful approach in using highly selective affinity labels to investigate opioid receptors utilizes the principle of recognition site-directed covalent association, which was pioneered by Baker (72). In theory, high selectivity is achieved because two recognition processes are required for covalent bonding: (a) the primary recognition process, which is reflected by the affinity of the ligand for the receptor, and (b) the second recognition step, which involves the alignment of the electrophilic portion of the reversibly bound ligand with a proximal receptor nucleophile. It is the secondary recognition step that is important for irreversible labeling inasmuch as covalent bonding will not occur if the electrophilic center of the ligand and the receptor nucleophile are not in the appropriate relationship to one another. An attractive aspect of this approach, which requires two recognition steps that lead to covalent bonding, is that, in theory, extremely high selectivity can be achieved provided that the electrophile exhibits selectivity in its choice of nucleophiles (73). In one of the first investigations of this theory, oxymorphone (72) was converted into its nitrogen mustard derivative, j3-chlorooxymorphamine (80) (74). In the guinea pig ileum, 80 was equipotent to morphine. However, unlike morphine, this agonist effect could not be reversed by membrane washing or naloxone treatment. lntracerebroventricular (icv) administration to mice resulted in a fourfold increase in the duration of analgesia compared to oxymorphone (75). Other in vivo results were equivocal, but more clear-cut results were obtained with the narcotic antagonist derivative, j3-chlornaltrexamine (81).
70
3
HO
Synthesis
N (CH2CH2Cl>
80
(R
81
(R
83
(R
84
(R
CH3)
of Morphine,
2
Codeine,
and Related
Alkaloids
~~C1CH2CH2
82
CH2-<1
Chlornaltrexamine (81) produced ultralong narcotic antagonism (2:3 days) to morphine analgesia in mice after a single icv administration. In contrast, reversible narcotic antagonists, such as naltrexone, prevent morphine-induced analgesia for less than 2 hours. In agreement with the long duration of action of 81, a single icv dose protected against morphineinduced physical dependence for 72 hours. Corresponding in vitro studies on the effect of 81 on the guinea pig ileum also demonstrated a blockade of the response to morphine that is not reversible either by washing or by subsequent morphine treatment (74). These in vitro and in vivo data suggest that chlornaltrexamine exerts its sustained antagonistic effect by alkylating opioid receptor nucleophiles. Since this compound is highly reactive toward nucleophilic reagents, it presumably reacts with receptor nucleophiles via its aziridinium ion (82). An affinity label that contains a highly reactive electrophile, such as 81, would not be expected to be able to discriminate between opioid receptor subtypes in the second recognition step because it would possess the
T
II The Structure-Activity
Relationships
of Morphine
and Related
Compounds
71
potential to react with a range of weak 10 strong receptor nucleophiles anywhere within its accessible steric volume. Indeed, chlornaltrexamine blocks the effect of both morphine and ethylketazocine, !J.and K agonists, respectively, on the guinea pig ileum. Irreversible antagonism of enkephalin activity in the mouse vas deferens indicates that 8 receptors are blocked as well (76). This observation led to the development of more selective irreversible ligands by modifying the reactivity of ligand electrophilicity. Among the several different electrophilic groups attached to the C-6 amine, the fumarate ester appears to be the best in terms of opiate receptor subtype selectivity. The naltrexone derivative has been named J3-funaltrexamine (83) and the oxymorphone analog J3-fuoxymorphamine (84) (77). The pharmacological properties of both, using the guinea pig ileum, are significantly different from those of their nitrogen mustard counterparts, 80 and 81. Both possess agonistic properties that can be terminated by either naloxone or membrane washing. Comparison of receptor affinities indicates that 83 is about five times more potent than 84, which is equipotent with morphine. The reversible agonistic effect appears to be mediated by different receptors. The agonistic activity of 83 resembles that of the mixed agonist-antagonists, while 84 appears to be a pure agonist. A remarkable feature of the action of 83 is that after repeated washings of the guinea pig ileum to remove noncovalently bonded 83, the response of the muscle strip to morphine is completely inhibited. It appears that the covalently mediated, irreversible antagonism of 83 in the preparation
(presumably
K
receptors) is completely resistant to alkylation by 83,
despite the fact that it mediates its agonistic effect through these receptors. Apparently, there are no sufficiently reactive nucleophiles, such as cysteine sulfhydryl groups, in the accessible steric volume of the 83 electrophilic center to add irreversibly in a Michael fashion. The in vivo results are consistent with irreversible binding. Subcutaneous injection of 83 is able to antagonize the effects of morphine for 4 days, as measured by the tail flick assay. The importance of stereochemistry at C-6 in opiate ligands has been investigated with a series of electrophiles in both the 6" and 613 configurations (78). Comparison was based on the efficiency of irreversible !J.receptor blockade (Table 3-4). The data illustrate the importance of the second recognition step and of covalent bonding, since only the 613 isomers irreversibly blocked !J.receptors and neither of the isomeric series affected the K receptors. The ligands with the 6" stereochemistry appeared to interact with the same receptors as the 613series, however, since the" isomer of 83 could protect against the irreversible effects of 83. Since both the" and 13isomers listed in Table 3-4 are potent reversible
3 Synthesis of Morphine, Codeine, and Related Alkaloids
72
II
The Structure-Activity
Relationships
of Morphine and Related Compounds
73
Table 3-4 Irreversible Effect of Affinity Labels on Morphine and Nalorphine
IC50 Ratio (treated/control) R I
NHCOC HCOC
C
CC02CH3
CC02CH3
NHCOCH=CH2 NHCOCH2I N=C=S N(CH2CH2Clh
C-6 Configuration a {3(83) a {3 a {3 a {3 a {3 a {3(81)
Morphine
Nalorphine
1.2 0.9 6.0 1.3 0.95 0.5 1.5 J. 0 1.2 0.8 1.9 1.1 1.0 1.2 2.4 1.2 1.3 0.8 6.9 1.5 Irreversible agonist 28.5 28.9
agonists in the guinea pig ileum, a deficient secondary recognition step rather than a poor primary association must be responsible for the inability of the 6a isomers to alkyl ate opiate receptors. 7. Substitution at Nitrogen The replacement of the N-methyl group in morphine brings about quantitative ,and, much more significantly, qualitative changes in SAR, yielding compounds that are potent agonists, antagonists, and mixed agonist-antagonists. Although substitution at nitrogen in the opiates is hardly new, the importance of the work early in this century was not recognized until the morphine antagonist properties of N-allylnormorphine (nalorphine) were discovered in the early 1940s (79). This led to a more systematic study of the effects of substitution at the amino group of morphine. The simplest derivative is normorphine (85), which is the starting material for most N-substituted derivatives. A wide variety of procedures has been used to effect this demethylation, including the classic von Braun reaction, utilizing cyanogen bromide (80). Subsequent improvements have
b
;> 90-93% overall
85 Scheme 3-13. The preparation of normorphine. Reagents: (a) phenyl chloroformate, potassium bicarbonate; (b) 95% hydrazine, allyl alcohol, nitrogen stream. \
involved the use o~ benzyl, ethyl, and methyl chloroformate (81). These were eventually replaced by phenyl chloroformate, since the intermediate carbamate formed with this reagent has proved easier to hydrolyze (82). Ethyl azodicarboxylate has been used to demethylate various 6-ester derivatives of morphine and codeine in reasonable yield (83). Normorphine has been prepared from its 2,2,2-trichloroethyl carbamate in a 75% yield (84). The current method of choice utilizes an improved phenyl chloroform ate method (85) that avoids the contamination of the normorphine with dihydromorphine, formed by diimide reduction during the hydrolysis of the carbamate with hydrazine (Scheme 3-13). Norheroin (86) is prepared directly from normorphine using acid-catalyzed acetylation with acetic anhydride (86), while 6-acetyl normorphine (87) is prepared from 3,N-bis(tert-butoxycarbonyl) normorphine, which is synthesized from normorphine and tert-butylazidoformate (86). Biologically, the interest in normorphine came from the hypothesis that the analgesia produced by morphine may be mediated by metabolic demethylation (264). The analgesic effectiveness of normorphine relative to that of morphine varies considerably by species and route of administration. By icv administration in mice, normorphine is equipotent with morphine in accordance with its in vitro receptor affinity (87). However, it is only about 0.10-0.15 times as active as morphine by subcutaneous (sc) or
"f 3 Synthesis of Morphine, Codeine, and Related Alkaloids
74
...
86
(norheroin, Rl
87
(R1
= R2 eOCH3) = H, R2 = COCH3)
88
89
intraperitoneal (ip) administration
(88). In
dogs, normorphine is equipo-
tent by intravenous (iv) administration (89), and, in humans, 0.25 times as active by sc administration (90). Normorphine maintains the addiction profile in addicts, and cessation of treatment after chronic administration results in withdrawal symptoms that are similar to but somewhat milder than those of morphine (91). In addicts, single doses of 85 cause less sedation, temperature depression, respiratory depression, and pupil constriction than equivalent doses of morphine (91). Norheroin (86) and 6-acetylnormorphine (87) have been tested only subcutaneously for centrally mediated analgesia in mice using the hot plate assay (86). In this assay, 86 and 87 are very much alike and possess approximately 0.05 times the potency of their N-methyl derivatives. The secondary amines in the nor-series appear to be too polar to allow facile transport into the CNS even when their hydrophilic hydroxyl groups are esterified, indicating that transport phenomena rather than a lack of intrinsic activity is responsible for their low analgesic potency. Other simple conversions of the amino group are oxidation to the N-oxide and quaternization. The N-oxide (88), a metabolite, is essentially inactive as an analgesic (92). Quaternary salts of opiates have been investigated because of the interest in developing opiate agonists and antagonists that act peripherally and are excluded from the CNS. NMethylmorphine (89), originally synthesized in 1868 (93), is active in the
II
The Structure-Activity
Relationships of Morphine and Related Compounds
75
acetic acid-induced writhing test, indicating peripheral analgesic activity, but is inactive in the hot plate assay, indicating central activity (94). However, the quaternary salts, when given systemically, have curare-like activity, causing neuromuscular paralysis, while on icv administration, 89 is analgesic and equipotent to morphine (95a). Obviously, the completely ionized quarternary salt is excluded from the CNS, but metabolic Ndemethylation can occur, producing morphine that readily penetrates the CNS. Morphinomimetic effects can then occur if a sufficient time course is used for testing. The individual diastereomers of ~-alkyl morphine have analgesic activity by both icv and sc administration. The Ncyclopropylmethylmorphine diastereomer, p.ossessing ~n a~ial ~-~ethyl group, has moderate mixed agonist-antagomst propertIes with sIgmficant CNS penetration (95b). Replacement of the methyl group of morphine with other organic residues became significant following the description of the antagonistic properties of N-allylnormorphine (nalorphine) (96-98), although the first narcotic antagonist, N-allylnorcodeine, was prepared in 1915 (99,100). Nalorphine (90) is a morphine antagonist, which although lacking analgesic properties in ~nimals, is an effective analgesic in humans,. being comp~rable to morphine (101). It was initially used as an antIdote for opIate overdose; it wa~;also used in combination with morphine in an attempt to attain analgesia ~ithout respiratory depression. However, it has respiratory depressant acti~ity and also produces intense dysphori~ and psychotomimetic effects. In addition, nalorphine produces a physIcal dependence different from that observed with morphine (102). These side effects made nalorphine clinically unacceptable as an analgesic. . The antagonist properties of nalorphine have prompted the synthesIs and biological examination of other varients of N-substitution in morphine (103-105). Extending the nitrogen substitution by only one methyl.ene group to the N-ethyl derivative (Table 3-5) drastically reduces morph~n~mimetic activity and reveals antagonistic properties. The 3-carbon cham IS optimal for morphine antagonism, with the n-propyl derivative being ~s potent an antagonist as nalorphine (90). Interestingly, the acetylemc propargyl derivative has very little activity in either direction. The .Nisopropyl group restores weak analgesic activity and potentiates morphme effects in the antagonism test. Lengthening the alkyl chain to n-butyl, n-pentyl, and n-hexyl restores analgesic properties and eliminates antagonism, while branching these side chains gives inactive compounds. The N- hen lethyl roup gives a compound with substantially enhanced ~nalgesic act~ty.. ISresult has fores a 0:ved the e~tenslve wo~k that has since been done on N-phenylalkyl synthetIc analgesIcs. SaturatIOn of t?e aromatic ring reduces the analgesic potency by a factor of 20, WhIle conversion to the phenacyl derivative destroys analgesic potency.
3 Synthesis of Morphine, Codeine, and Related Alkaloids
76 table 3-5
II
The Structure-Activity
Relationships of Morphine and Related Compounds
77
Table 3-6
The Effect of N-Substitution
on Morphine Analgesia
Oxymorphone-Based
Antagonists
HO
R CH3 (morphine) CH2CH3 CH2CH=CH2(90) CH2C==CH CH2CH2CH3 CH(CH3h (CH2hCH3 CH2CH(CH3h (CH2)4CH3 (CH2)sCH3 CH2C6Hs CH2CH2-c-C6H II CH2CH2.C~s CH2COC6Hs (CH2hCN
(nalorphine)
"b Tail flick test in rats Compound potentiates c
Relative Analgesic Potency"
Relative Morphine Antagonist Potency
1.0 <1.0 <0.1 <0.1
o <0.1 1.0 <0.005 1.0
o <0.1 <0.1
o 0.7 0.7 <0.1 0.3 6.1. <0.1 0.3<
Potentiatesb <0.1 <0.1 Potentiates
o Potentiates o o
o
Antagonistic Potency"
R 91 92 93
96
98
CH2CH=CH2 (naloxone) CH2CH=C(CH3h (nalmexone) /CH2 (naltrexone) CH2-CH"" CH2 /CH, ."" CH2 CH2-CH"" / CH2
I
CH~D
99 a
b
I 0 S-c(mfiguration
R-configuration
Relative to nalorphine Relative
to morphine
(90)
= 1.0.
=
Analgesic Potencyb
7-10 0.5
0 0.3
17
0
5
0.2
25
0.2
0
1.0.
<0.01
the action of morphine.
Refcrence 106.
The activity trends observed in the normorphine series has led to the introduction of various N-substituents in other morphine alkaloids that have shown enhanced analgesic potencies compared to morphine, with the reasonable expectation that more potent antagonists would be obtained (107). Oxymorphone (72) is a narcotic analgesic approximately 10 times as potent as morphine. Direct alkylation of noroxymorphone with allyl bromide yields naloxone (91) (108). Naloxone is a potent narcotic antagonist, being approximately 7-10 times as active as nalorphine (90) (Table 3-6), while demonstrating no agonist activity either in the mouse hot plate or writhing tests (109) or in humans (110). Furthermore, naloxone completely antagonizes the analgesic activity of mixed agonist-antagonists such as nalorphine (/11). The antagonistic without agonistic effects of naloxone have been confirmed in humans. Naloxone antagonizes the
respiratory depression produced by oxymorphone but does not produce respiratory or circulatory effects of its own (112). Further studies in humans confirm the lack of dysphoric agonist effects such as the psychotomimetic reaction and sedation (113). Naloxone has an extremely short duration of action, whiCh makes it very suitable for the treatment of acute narcoticism but not for blocking the euphoric effects of opiates in addicts. The ( + )-enantiomer of naloxone has been synthesized and has only 10-3 to 10-4 times the activity of (- )-naloxone (267). Introduction of a dimethylallyl group into noroxymorphone gives nalmexone (92), which, through the simple addition of the two vinyl methyl groups, yields a mixed agonist-antagonist (114). Nalmexone is approximately one-half as potent as nalorphine (90) as an antagonist and about one-third as active as morphine as an agonist. As part of the continuing search for potent, long-acting antagonists for long-term opiate receptor blockade in addicts, the cyclopropylmethyl group has been introduced into
78
3 Synthesis of Morphine. Codeine, and Related Alkaloids
the noroxymorphone molecule, yielding naltrexone (93) (115,116). Naltrexone has been approved for use in the United States for addicts who have been withdrawn from opiate addiction and are in the process of rehabilitation. Naltrexone is several times more potent than naloxone and has a much longer duration of action. The cyclopropyl group in 93 protects against hepatic degradation during the first pass through the liver (115). Naltrexone is 17 times more potent than nalorphine in producing abstinence symptoms in opiate-dependent patients (117). Animal studies indicated that 93 has an opiate antagonist potency 40 times that of nalorphine (90) and 2-3 times that of naloxone (91) (118). The potency difference is not large enough to explain the 20- to 30-fold difference seen in the comparison of naltrexone to naloxone with respect to duration of action and human dosage (119). The difference, therefore, is due to different metabolic pathways. Naloxone is inactivated in humans primarily by hepatic conjugation to form glucuronides that do not readily penetrate the blood-brain barrier (120). Naltrexone, although also conjugated, is reduced primarily to 6,8-naltrexol (94), an active metabolite, which does cross the blood-brain barrier (121). Chemically, a wide variety of reducing agents give the 6a-naltrexol epimer (95). The 6,8 alcohol can be formed by reduction with formamidine sulfinic acid (Scheme 3-14) (122,123). Biologically, 6,8-naltrexol (94) has opiate antagonist activity that is 2-8% that of its parent, naltrexone (121,124), without demonstrable analgesic activity (123). On the other hand, 6a-naltrexol (95) behaves as a mixed agonist-antagonist (123). Expansion of the cyclopropyl ring of naltrexone gives the cyclobutylmethyl analog (96). The compound is reported to be a mixed agonistantagonist with approximately five times the potency of nalorphine and equivalent potency to morphine (Table 3-6) (125). Reduction of 96 yields the 6a-a1cohol, nalbuphine (97). The reported syntheses of nalbuphine are presented in Scheme 3-15 and are illustrative of the various methods for producing N-substituted normorphine derivatives (125). As an analgesic, nalbuphine has four to five times the potency of morphine in the mouse writhing test but is very weak in the hot plate. In humans, 97 has been found to be slightly less potent than morphine, but with a marginally longer duration of action (126). As previously indicated, replacement of the N-methyl group in morphine alkaloids generates a spectrum of activities from pure agonism through pure antagonism. An example of the subtleties inherent in this region of the molecule is demonstrated by the tetrahydrofurylmethylene diastereomers 98 and 99. The R-diastereomer (98) is a mixed agonistantagonist possessing 25 times the potency of morphine in the mouse writhing assay and having the ability to suppress morphine abstinence in
II
The Structure-Activity
Relationships
of Morphine and Related Compounds
79
93
a
95 Scheme 3-14. Reagents: acid.
(a) lithium tri-sec-butylborohydride;
b
94 (b) formamidine
sulfinic
pa~~. The S-diastereomer (99), on the other hand, is a pure antagonist with 0.2 times the activity of nalorphine and no agonist properties (Table 3-6) (127). B.
Insertion of Substituents in Nonfunctionalized
Areas
While manipulation of functional groups in the morphine molecule was undertaken initially to determine general SAR, the introduction of substituents into nonfunctionalized portions was done secondarily to look at specific effects on a given opiate derivative's biological activity. On the whole, these substituent introductions have resulted only in quantitative differences in analgesic activity when compared to the parent molecule. 1. Substituellts ill the AromaticA-Rillg An intact A-ring is, in general, essential for analgesic activity in morphine-based molecules. Substitution in this ring results in reduced analgesic activity. The I-chloro and I-bromo codeine derivatives (100), as well as the I-acetyl compound (101), have
3
80
Synthesis of Morphine, Codeine, and Related Alkaloids
II
The Structure-Activity
Relations~ips of Morphine and Related Compounds
81
a ~CH30
OH
CH30
HO (R
noroxymorphone
~100 101 (R 102
F)
(R
H, dihydrocodeinone)
103
(R
CH 2Cl )
104
(R
CH3)
OH
H
~b
2.
---7 OH
HO 96
=
21
d
c
HO
(R
= Cl, Br) = CH3CO)
97 nalbuphine
Scheme 3-15. Synthesis of nalbuphine. Reagents: a, cyclobutane carbonyl chloride; b, sodium borohydride; c, cyclobutylmcthylene bromide; d, lithium aluminum hydride.
significantly diminished analgesic potency relative to codeine (128). However, the l-fluoro analog (102) is as potent as codeine, indicating that steric bulk factors rather than electronic effects are responsible for the decrease in analgesic potency (129). A I-methyl group has been introduced into dihydrocodeinone (21) by chloromethylation to the l-chloromethyl derivative (103) and zinc-acid reduction to the methyl group (104). The 6-ketone is subsequently reduced to give I-methyl-dihydrocodeine (105) (130).
.
Introduction
of Substituents
at Position 5: The Dihydrofuran
CoRing
JunctIOn Metopon (112), 5-methyldihydromorphinone, was considered to be one o~ the best morphine drugs developed (131). It is more potent t~an morphme on oral and sc administration, produces fewer side effects lIke nausea and vomiting, and is less sedating. Physical dependence develops less rapidly and is less severe than with morphine (132). Its use, however, has been restricted by its lengthy, difficult, and expensive synthesis. The synthesis (Scheme 3-16) of 112 is interesting in that it involves a direct ~isplacement at an allylic carbon by a grignard reagent. The synthesIs starts from the readily available thebaine (106) (133), which is h~drogenated to the enol ether of dihydrocodeinone (107). Reaction of 107 w~th methyl ?rignard results in displacement of the allylic oxygen at C-5 with concomitant cleavage of the dihydrofuran ring. The reason for this unus.ual reaction is twofold: (a) coordination of the allylic oxygen with the LewIs ~cid ma.gnesium iodide and (b) the fact that the leaving group is a pheno~lde amon. Subsequent hydrolysis of the enol ether gives the ~orphl?an ketone (108). To then rebuild the dihydrofuran ring, 108 is dlbrommated to 109 and treated with base to yield the morphine derivative
3
82
Synthesis of Morphine, Codeine, and Related Alkaloids
II
The Structure-Activity
Relationships
of Morphine and Related Compounds
83
Table 3-7
a
b
The Influence of Substituents at C-5 in Dihydromorphinone
~
106
107
R H CH3 (metopon) CH2CH3 CH2(CH3h n-CSHII
c ~ OH
.
CH3
108
Analgesic Potency.,b
b
109
Measured
1 2 1 0.1 . 0.5
(112)
in cats; ref. 135.
Dihydromorphine
= 0.6
dihydromorphinone.
Despite the lengthy synthetic pathways needed to introduce a simple alkyl group at C-5, a number of derivatives have been prepared. Lengthening the chain at C-5 results in a steady decrease in analgesic potency. A n-pentyl group in this position possesses 25% of the analgesic activity of metopon (112), while an isopropyl group has only 5% of the activity of the methyl derivative (Table 3-7) (136). Further manipulation of metopon by reduction to the 6a-alcohol (113) reduces the analgesic potency to 1/50th
a,e ~
110 112
(R
= H, metopon)
Scheme 3-16. Synthesis of 5-methyldihydromorphinone (Metopon) from thebaine. Reagents: a, catalytic hydrogenation; b. methyl grignard; c. bromination; d, base; e, demethylation.
\110). The residual bromine is removed by catalytic hydrogenation, forming 5-methyldihydrocodeinone (111). Demethylation at C-3 then yields metopon (112) (132,134,135). It was reported that treatment of thebaine (106) with butyl lithium at low temperature results in stereospecific deprotonation at C-5. Alkylation with methylfluorosulfonate gives 5f3-methylthebaine, which is then converted into metopon (112), as well as various 5-methylcodeine derivatives (135).
113
that of metopon (112) (137). A later study varied the nitrogen substitution by introducing the strongly agonistic phenethyl group and the antagonist, producing dimethylallyl and cyclopropylmethyl groups. Both the 6-ketone and the 6a-alcohol of these derivatives have been prepared (Table 3-8). Surprisingly, not only do the agonist-enhancing groups produce active
3
84
Synthesis of Morphine, Codeine, and Related Alkaloids
II
The Structure-Activity
Relationships
of Morphine and Related Compounds
Table 3-8 The Effect of C-6 and Nitrogen Substitution on 5-Methyl Substituted Morphine Derivatives
a
115
=0 =0 =0 =0 ...OH ...OH
.
Analgesic Potency..b
Antagonism
CH3 (metopon) H (CHzhC6H, CHzCH=C(CH3h (CHzhC6Hs
1.0 0.2 8 0.25 0.5
No
CHz-<1
0.02
No
No No No
116
117 (7-methoxycodeine)
Hot plate test, sc, in mice. b Morphine = 0.4 metopon.
analgesics, but the antagonist-producing groups give compounds retaining the analgesic activity and do not demonstrate any morphine antagonism (138). :::0
3. The Influence of Substituents at Position 7 on Analgesia As previously indicated, much of the focus on the introduction of substituents into nonfunctionalized areas has been a classical medicinal chemical exercise in potency enhancement. While this is partially true for the C-7 region of the morphine molecule, a great impetus for substitutions has been the tremendous potency enhancements found in the Diels-Alder adducts of thebaine and the related oripavine derivatives (qv). It has been observed empirically that in the Diels-Alder adducts, an alkyl substituent at C-7 is the most important factor in enhanced potency (139). A methoxyl group has been introduced at C-7 using the I-bromo derivative (115) of the naturally occurring morphinan alkaloid, sinomeninone (114) (Scheme 3-17). The 7-enol ether of 115 is brominated and then treated with base to close the oxide bridge to the sinomenine derivative (116). Hydride reduction then reduces the 6-ketone to the 6a-alcohol and reductively removes the I-bromine to give 7-methoxycodeine (117), which in the presence of acid regenerates sinomeninone (114). The analgesic potency of 117 is about one-third that of codeine when administered
CH30
OH
114 (sinomeninone) Scheme 3-17.. Synthesis of 7-methoxycodeine from sinomeninone. Reagents: a, methanol, hydr?gen chlonde; b,. bromine, acetic acid, then sodium hydroxide; c, lithium aluminum hydnde; d, hydrochlonc acid.
subcutaneously in the hot plate test and is equivalent to codeine when given orally. Since 117 is unstable to acid, the product sinomeninone (114) was tested for analgesic activity. While it was active parenterally, it was not active orally, indicating that the oral analgesic activity of 117 is not due to its conversion to 114 (140). The introduction of 7-alkyl groups.has been accomplished in a variety of ways, none of which is a straightforward alkylation process. In contrast to the reaction of thebaine (106) with methyl grignard, ultimately forming
85
3
86
Synthesis of Morphine, Codeine, and Related Alkaloids
II
The Structure-Activity
Relationships
of Morphine and Related Compounds
87
a
---7 21
122
106 (thebaine)
e,f,g ~
c CH3 ~ or d CH30
118
OH R = H,
119
CH3
R = H ( 12 3), OH ( 12 4) , alkyl, phenyl, acetyl
125
Scheme 3-19. Preparation of 7a-substituted morphinones. Reagents: (a) dimethyl formamide dimethylacetal; (b) hydrogenation under various conditions; (c) organo-lithium reagents.
n Scheme 3-18. agents:
a, lithium
Synthesis of 7-methyldihydromorphinone dimethylcuprate;
b, acid; c, catalytic
= 2,3
derivatives from thebaine.
hydrogenation
(for R
=
Re-
H); d, excess
lithium dimethylcuprate (for R = CH3); e, von Braun reaction; f, N-alkylation with cycloalkylmethyl bromides; g, bromine, acetic acid, then sodium hydroxide; h, boron tribromide.
metopon (112), the reaction with lithium dimethy1cuprate gives a 7methylmorphinan (118) (Scheme 3-18) (141). Hydrogenation of the enone in 118 gives the 7O'-methyl morphinan; reaction with lithium dimethy1cuprate gives the 7a-813-dimethylmorphinan (119). Because of the instability of 7-methylcodeine to acid, the N-methyl group is removed and the resultant nor-compound is alkylated with cycloalkylmethylene halides prior to ring closure to the dihydrofuran (120) (135). Removal of the codeinone 3-methoxy methyl group with boron tribromide then yields the
7-O'-methyl compounds (121) (142,143a). An alternative approach (Scheme 3-19) is the reaction of dihydrocodeinone (21) with dimethylformamide acetal to give the vinylogous amide (122), which can be hydrogenated to the 7O'-methyl (123) or the 7a-hydroxymethylene (124). Other 7-substituents could be introduced by the reaction of 122 with organolithium reagents to give 125 followed by hydrogenation. These reactions have also been applied in the 1413-hydroxy series (Scheme 3-19) (142). The effects of these substitutions on analgesic potency are presented in Table 3-9. These data indicate that the introduction of a methyl group into the 7a position in the N-methyl compounds increases the analgesic potency relative to codeine but is equivalent to that of the unsubstituted codeinone. Increasing the size of the group in the 7a position to larger than a methyl causes a drop in potency, as does the incorporation of functionalized side chains. In the N-cycloalkylmethylene compounds, methyl substitution at the 70' position increases the analgesic potency over that of morphine but reduces antagonist activity. Further introduction of
3 Synthesis of Morphine, Codeine, and Related Alkaloids
88
II
The Structure-Activity
Relationships
of Morphine and Related Compounds
89
Table 3-9 The Effects of 7-Substitution
on Analgesia b,c
--=JI. d 21
126
Analgesia
Other
Writhinga
14-0H 14-0H
3.8d 2.2 0.5 I <0.4 1.0 2.0' I 2.7 0.1 I 0.6
Tail Flickb
e
Antagonism'
? CH3 CH3 CH2 CH3 CH3 CH3 H H H H H H a
CH3 CH2CH3 (CH2).CH3 CH2C(CH3h CH2C6Hs CH20H CH3 CH3 CH3 CH3 CH3 CH3
H H H H H H H CH3 H CH3 H H
CH3 CH) CH3 CH) CH3 CH) CH2-c-yHs CH2-c-C3Hs CH2-c-C.H7 CH2-c-C4H7 CH2-c-yHs CH2-c-C.H7
CH30 127 R 0.61 1.7 <0.1 0.4 28 0.2
Mouse, sc.
bRat, sc. , Rat tail flick, sc.
.
to codeine
1 Relative
to nalorphine
d Relative
Relative to morphine
= 1.
= 1. =
1 (Table 3-5).
an 8f3 methyl group almost eliminates analgesic activity without producing remarkable effects on antinociceptive activity. Introduction of a 14f3 hydroxyl group with a 7a methyl yields an antagonist about twice as potent as naltrexone (93) (Table 3-6). The conclusion that was drawn from these data is that introduction of a 7a methyl group into the N-cycloalkylmethylene opiates does not result in agents that have potent mixed agonist-antagonist properties (142). Geminal dimethyl groups have been introduced at C-7 through a novel crossed aldol-Cannizzaro reaction on dihydrocodeinone (21), which results in reduction of the ketone to the 6f3-a1cohol and introduction of the 7,7-bis-hydroxymethylene group (126) (Scheme 3-20) (143b,144). Selec-
128 H, CH3, CH2CH3
Scheme 3-20. Synthesis of 7,7-dimethyl-dihydromorphone and codeine derivatives. Reagents: (a) formaldehyde, calcium hydroxide; (b) p-toluene sulfonyl chloride, pyridine; (c) lithium triethylborohydride; (d) dimethyl sulfoxide, trifluoroacetic anhydride; (e) sodium borohydride.
tive tosylation of the C-7 substituent followed by hydride reduction gives the 7, 7-dimethyl-6f3-hydroxyl compound that is oxidized to the ketone (127) and reduction to the 7,7 -dimethyl-dihydrocodeine (128). Additional substitution can be introduced at C-8, and the N-substituent can be converted to cyclopropylmethylene and the 3-methoxyl converted to the free phenolic hydroxyl, as shown in Scheme 3-18 (144). The biological results of these substitutions are shown in Table 3-10. In the dihydrocodeinone series, a geminal dimethyl group increases the analgesic potency over that of codeine but retains approximately equivalent potency to dihydrocodeinone in both the writhing and tail flick assays. In the dihydromorphinone series, with a free 3-phenolic hydroxyl, the 7,7dimethyl analogs possess enhanced analgesic potency over both morphine and dihydromorphinone (see Table 3-3). With an N-cyclopropylmethylene and 7,7-dimethyl groups, the compounds are pure antagonists devoid of agonist activity (144). In general, the introduction of a 7,7-dimethyl group can enhance the analgesic activity of morphine-based opiates, but this effect depends on the specific groups at C-8 and on the amine nitrogen. The effects are not great compared to those that can be achieved by other substitutions.
3
90
Synthesis of Morphine, Codeine, and Related Alkaloids
II
The Structure-Activity
Relations~ips of Morphine and Related Compounds
91
Table 3-10 7,7-Dialkyl Codeinone and Morphinone
Derivatives a,b ---7
H£CH-CH=CH-C6HS
Analgesia
R1
R2
CH3 CH3 CH) CH) H H H H H
H CH) CHzCH) H H CH2CH) H CH) CH2CH)
R)
Other
CH) CH3 CH) CH) CH) CH) CH2-c-C)H~ CH2-c-C)H~ CHz-c-C)H~
6a-OH
Writhing"
2.6.1 4.1 2.3 0.8 11 ' 3.5 I I I
Tail Flick b
c
Antagonist
"'CH20H
C
~
d
OH
7.7 18 5.4
CH30
Scheme 3-21. Reagents: (a) acetone, acid; (b) Swern oxidation; (c) cinnamyl Wittig; (d) catalytic hydrogenation.
19.5 31
Table 3-11 2.1' 6.5 1.0
7a-Hydroxymethylene
Morphine and Codeine Derivatives
" Mouse, sc. bRat, sc. e Tail flick rat. sc. .I Relative to codeine = I.
e
f
Relative Relative
to morphine to nalorphine
= 1.
=]
OH (Table
3-5).
A series of 7-acyl derivatives of dihydrocodeinone (21) and dihydromorphi none were prepared by acylation of the morpholine enamine of 21 and subsequent boron tribromide de methylation (145). Of the ,B-diketones thus prepared, the 7-hexanoyl derivative (129) is the most interesting, possessing one-fourth the analgesic potency of dihydromorphinone (70) (see Table 3-3), wth very weak antagonist properties (145). The most interesting results in this series occur when a lipophilic 7,B substituent is introduced. The required stereochemical configuration was prepared as illustrated in Scheme 3-21. Because of the method of synthesis, all the derivatives prepared contained a 6,B-hydroxyl instead of the naturally occurring 6a, and a 7a hydroxymethylene group (146). The biological results are presented in Table 3-11. Introduction of a 7,B alkyl group has little effect in the dihydrocodeine series. However, introduction of a phenethyl group increases the analgesic potency to 370 times that of codeine, while a phenylbutyl has over 700 times codeine's potency.
R1
R2
CH) CH) CH) CH) CH) CH) H CH3 H
H CH2CH) C6H~ CH2C6Hs (CHzhC6H~ (CH2hC6H~ (CHzhC6Hs (CHzhC6Hs CHZC6HS
U
R) CH) CH) CH) CH) CH) CH) CH) CH2-c-C3Hs CH2-c-~Hs
Analgesic Potency" 3.Sb O.S 370 26 735 4.4 700" SOb 4.5"
Mouse writhing assay.
b Relative Relative "
to codeine to morphine
= I. = 1 (codeine/morphine
= 5).
Curiously, the phenylpropyl group confers less than 50% of the activity of the phenyl butyl. Above four methylene groups, the analgesic potency rapidly falls off. Similar results have been obtained in the dihydromorphine series. When the N-methyl group is replaced by a cyc1opropylmethylene,
3
92
Synthesis of Morphine,
Codeine, and Related Alkaloids II
The
potent analgesic activity is retained in the codeine analogs, but the morphine analogs are only marginally better than morphine itself. Narcotic antagonism has not been demonstrated for these compounds (146). Through a complex series of steps, 130, which contains the requisite 6a-hydroxyl and a 7a methyl group in place of the previously prepared hydroxymethylene, was synthesized. This compound (130) has about 20 times the potency of codeine but is about one-half as active as its hydroxymethylene analog. Oxidation to the 6-ketone formed the codeinone (131) and subsequently the morphinone (132) derivatives. Although both of these contain the N-cyclopropylmethylene group characteristic of the antagonists, they are equipotent to dihydrocodeinone (21) and dihydromorphinone (70) respectively (146). The syntheses of the much
Structure-Activity
Relationship~
of Morphine
and Related
Compounds
93
52
+
CH30 Scheme 3.22.
Synthesis
of
8-ethyldihydrocodeinones.
Reagent:
(a) lithium diethyl su-
prate.
more interesting compounds (133), containing the N-methyl group in place of the N-cyclopropylmethylene in 131 and 132, have been reported, with no indication of biological activity (147).
129
(CH2)4C6HS /
"CH3
RO
130 131
(R
132
(R
4. C-B Substituents Impetus for the investigation of the effects of substituents at position 8 has stemmed from the high analgesic potencies of the Diels-Alder adducts of thebaine (Section III). In an attempt to explain the potent analgesic activity of this series of compounds, the existence of a lipophilic site on the opiate receptor surface was hypothesized (148). This putative site would interact with the alkyl portion of the tertiary alcoholcontaining side chain of the C-ring. Modeling indicated that this lipophilic receptor site was proximate to both C-7 and C-8 of morphine. Substituents are readily introduced at C-8 by l,4-conjugate addition of lithium dialkyl cuprates to codeinone (52), giving, as the major product, 8{3-alkyl-dihydrocodeinones (149). The 8a epimers are also formed in low yield (150). The synthesis of the 8-ethyl epimers is illustrated in Scheme 3-22. A series of 8-acyl compounds has been prepared by l,4-conjugate addition of acyl anion equivalents. The 8-acyl compounds are of interest because of their ease of conversion to the tertiary alcohols. Biologically, a wide range of 8{3substituted compounds with unsaturated, branched, or large straight chain alkyl groups, together with acyl and tertiary alcohols, have been disappointingly less active as analgesics, both centrally and peripherally, than dihydrocodeinone (149). The 8{3-methyl
-3
94
Synthesis of Morphine, Codeine, and Related Alkaloids
II
The Structure-Activity
Relationships
of Morphine
and Related Compounds
95
a
(CO)3cr
~
134
(R
135
(R
= CH3)
136 (codorphone)
42 (codeine)
= CH2CH3)
138
b,c,d,e
HO
137
(134) and ethyl (135) dihydromorphinone derivatives are equipotent with the unsubstituted parent. Because ot this activity, N-cyclopropylmethyl derivatives have been prepared in both the morphinone and codeinone series. The most interesting compound to come out of the C-8 substituted opiates is codorphone (136), which has a mixed agonist-antagonist profile and has undergone clinical evaluation (149). Surprisingly, introduction of either a heteroatom or a halogenated or otherwise functionalized side chain effectively eliminates analgesic activity in the 8f3 substituted dihydrocodeinone series (151). The 14f3-hydroxy-8-substituted compounds have also been synthesized, but analgesia results have not been reported (152). 5. Natural and Unnatural Substituents at C-10 Position 10 on morphine has been relatively little investigated. However, since C-lD is benzylic on an electron-rich ring, it should be readily oxidatized through radical reactions. The 1O-keto compound (137) has been isolated from morphine solution, but no indication of its biological activity is available (153). In a rare application of organometallic chemistry to the opium alkaloids (Scheme 3-23), the chromium tricarbonyl complex of codeine (138) has been prepared and its configuration determined by X-ray crystallography. Alkylation of a protected version of this complex with methyl iodide gives the lOf3-methyl derivative (139) after removal of the
139
(R
CH3)
140
(R
H)
Scheme 3-23. Synthesis of lO-methyl morphine derivatives. Reagents: a, chromium hexacarbonyl; b, tert-butyl dimethylsilyl chloride; c, sodium hexamethyldisilazine, iodide; d, pyridine; e, tetrabutyl ammonium fluoride.
methyl
chromium and deprotection. The morphine derivative 140 was prepared from 139 using standard methods (154). Biologically, the 1O-methyl derivative in the codeine series (139) is equipotent to codeine, while the lD-methyl-morphine analog (140) is much less potent than morphine (154).
6. Thebaine-Derived Substituents at C-14 One of the earliest substituents introduced into the morphine nucleus was a 14f3-hydroxyl group that was ultimately derived from the then rare alkaloid, thebaine. Thebain (106) occurs as a minor alkaloid in the opium poppy but is the main alkaloid (up to 52% in the dried latex) in another poppy species, Papaver bracteatum (155). The presence of the 3-0-methyl ether in thebaine makes this alkaloid an attractive starting material for the preparation of codeine and codeine-derived antitussives and analgesics, since the syntheses do not go through morphine or other readily abusable substances. The 14hydroxyl group confers a substantial increase in analgesic potency over the parent unsubstituted compound. Chemically, a 14f3-hydroxylgroup is introduced by oxidation of thebaine with either hydrogen peroxide or peracids in organic acids (156,157).
3
96
Synthesis of Morphine, Codeine, and Related Alkaloids
a ~
106 (thebaine)
141
1
142 (l4-jJ
-hydroxycodeinone)
II
The Structure-Activity
Relationship,s of Morphine and Related Compounds
97
unlikely. A more likely mechanism is the epoxidation of the 8 (14)-double bond to form the intermediate (143), which undergoes acid-catalyzed ring opening to give 144. The epoxide ring-opened intermediate (144) can proceed directly to 142 or, alternatively, can undergo water or organic acid addition to form 145. The 8,8,14,8-diol enol ether (145) has been isolated in good yield from the meta-chloroperbenzoic acid oxidation of thebaine in a mixture of acetic and trichloroacetic acids (158). Hydrolysis of the enol ether in 145 gives the known dihydroxydihydrocodeinone derivative (146) (159). Both 145 and 146 are readily convertible into 14,8hydroxycodeinone (142) (158). Further transformations of 142 include hydrogenation to 14,8-hydroxydihydrocodeinone (147). The classical method for converting 147 into oxymorphone (72) is reaction of 147 with hot hydrobromic acid (160). Additional evidence for initial 8,14-addition comes from nitration studies of thebaine. Initially, reaction of thebaine with tetranitromethane was reported to give 14,8-nitromorphinone (148) (161). Subsequently, when thebaine was oxidized with dinitrogen tetroxide, 8-nitrothebaine (149) was isolated, which is consistent only with initial formation of an 8,14-dinitro derivative (Scheme 3-25) (162).
1
HO
147 (oxycodone)
146 Scheme3-24. Synthesis of 14,6-hydroxycompounds from thebaine. Reagents: (a) hydrogen peroxide, organic acid; (b) meta-chloroperbenzoic acid, organic acid.
Initially, the oxidation was postulated to proceed through l,4-addition of the elements of hydrogen peroxide to thebaine to produce the 6,14-diol (141) (Scheme 3-24), which undergoes expulsion of methanol to form 14,8-hydroxycodeinone (142) (156). While l,4-addition is possible, subsequent investigations and ancillary evidence make this direct mechanism
OH
150 151
(no 6,7,8)
Biologically, the effects of 14,8-hydroxyl substitution are species dependent. While oxycodone (147) and oxymorphone (72) have only 1.4 and 11 times the potency of morphine in mice, respectively, oxymorphone has 6.7 times the analgesic potency of morphine in humans, while oxycodone has a potency of 0.7 times (25,29,163,164). The 14,8-hydroxyl derivatives of morphine (150) and its dihydroanalog (151) are equipotent to morphine, at least in mice (165). The 14,8-hydroxyl group in these compounds is strongly hydrogen bonded to the amino nitrogen, which may provide a reason for the potency-enhancing influence of this substitution (166). A very interesting observation on potency SAR has been made for the esters of 14,8-hydroxycodeinone (142); instead of hydrogen-bonding effects,
3 Synthesis of Morphine, Codeine, and Related Alkaloids
98
n
The Structure-Activity
Relationship,s of Morphine and Related Compounds
99
Table 3-12 SAR of the Esters of 14-H ydroxydihydrocodeinone
a ~CH30
OCH3
CH30
OCH3
106
/
R
1
148 Scheme 3-25. Reaction of thebaine with dinitrogen tetroxide or tetranitromethane.
tetroxide.
149 Reagent: (a) dinitrogen
potency may be due to partition coefficient effects (Table 3-12) (167). These derived esters are all analgesics, some with very high potencies. The analgesic potency rises with increasing size, peaking when R is n-hexyl and then diminishing. The benzoyl esters are inactive but phenacyl, cinnamoyl'l and related esters are up to 200 time$ as potent as morphine or 600 times as potent as codeine (167,168). Transport factors are probably the most important component of the SAR relationships here, but conformational changes due to non-hydrogen-bonded interactions between the 14,8 ester and the amino group may also be important. Although the effects of a 14,8-hydroxy group have been known for a long time, only recently have other substituents been introduced at this position. The nitro derivative (148) can be reduced with sodium borohydride to the 14,8-amino group, which is trapped with acetic anhydride to give the acetamido derivative, which can be demethylated with boron tribromide to give the 14,8-acetamido-morphine derivative. Other substituents can be introduced by the reaction of thebaine with thiocyanogen,
H (142) COCH3 COC:zHs COC3H7-n COCH1CH=CHz COC.H9-n COCSHII-n COC6HIJ-n CO~Hwn CO~H1Tn COC11HI9-n COCHZC6HS CO(CHz)zC6HS COCH=CH-C6Hs COCH=CH-CH3
Analgesic Potency.,b 0.3 4 20 30 30 40 50 60 5 I 0.03 50 liS 175:!: 75 31
. Tail clip paradigm, sc, in mice. b Relative to morphine (= 1).
(SCN)z, or positive halogen. Demethylation then gives the 14,8-substituted morphine and morphinone derivative (Table 3-13) (169). The analgesic potency of these derivatives has been determined using the guinea pig ileum with normorphine as the standard, although normorphine analgesia is reported to be strongly species dependent (d. Section II,A,7). In this paradigm, all the 14,8-substituted codeine derivatives are extremely weak. However, codeine itself is extremely weak in this assay (169,170). The analgesic potencies of all the morphine derivatives are less than that of morphine (169). The influence of 14,8 substituents has also been investigated in the naltrexone series (171). A phenylamino group has been introduced by a 1,4-cycIo-addition of thebaine with substituted and unsubstituted nitrosobenzene derivatives (Scheme 3-26) (172). Rearrangement of the initial hetero Diels-Alder
3
100
Synthesis of Morphine, Codeine, and Related Alkaloids
Table 3-13
=0 ...0zCCH) ... OH =0 =0 ...OH
.
NOz NHCOCH) SH CI Br ...OH
Guinea pig ileum. b Relative to normorphine
Diels-Alder
-
-
156
155
= 1.
a
154
? CH30
-
Potencya,b
0.006 0.6 0.4 0.5 0.4 0.8
(85)
101
Adducts of Thebaine
adduct yields".Jhehydroxyamiiie (152), which can be converted to the 1413 phenylamine derivative (153) with concomitant reduction of the 7(8)double bond. Unfortunately, none of the substituted analogs are as potent as morphine when tested in the mouse writhing assay (172). Another hetero-Diels-Alder reaction is used to prepare 14J3-amino morphine and morphinone derivatives. Reaction of thebaine with 1,1chloronitrosocyclohexane yields, after zinc reduction, 14-aminocodeinone
Thc Analgesic Activity of 1413-Substitutcd Morphine and Morphinone Derivatives
Analgesic
III
(154) (173). Conversion into the 14J3-bromoacetamide derivatives 155 and 156 is readily accomplished using standard methodology (174). It will be recalled that reduction of the 14J3-nitro derivative also results in the reduction of the 7(8)-double bond (169). The bromo acetamide derivatives are of interest as irre~ersible ligands (d. Section II,A,6), and both possess
OCH3 106 (thebaine)
high affinity
for the opiate
receptor
(ICso
= 15
and 10 nm, respectively;
morphine = 4 nm) and demonstrate irreversible binding when incubated with brain membranes for at least 30 minutes (174).
III.
Diels- Alder Adducts of Thebaine
Thebaine (106), an alkaloid present in 0.2-0.8% in opium and a major constituent (90% of total alkaloid content) in Papaver bracteatum (which is morphine free), possesses little utility commercially or medically for two reasons: (a) its lack of the depressant and analgesic properties common to 153 152 Scheme 3-26. Reaction of thebaine with nitrosobenzenes. Reagents: (a) nitrosobenzene' (b) hydrochloric
acid; (c) hydrogen.
Pd/C.
'
IOZ
3
Synthesis of Morphine, Codeine, and Related Alkaloids
other morphine alkaloids and (b) its expression of extreme toxicity and CNS stimulation. The value of and interest in thebaine in recent years have thus centered on the pharmacological exploitation of its transformation products, most notably the 6,14-endoetheno and 6,14-endoethano derivatives, those C-ring bridged adducts of general structural type 1581 derived
106
R =
157
R = H
R
CH3
Thebaine Oripavine
Rl
H, CH3 H, alkyl,
R2
alkyl, substituted
halogen
or cyclo X y
C2H2, C2H4 yl = H, alkyl, keto or carboxyl derivative amine, alcohol
Z
H, alkyl, alkenyl, carboxyl derivative
158
III
160 R'
= R" = H,
alkyl, aryl
from condensation of thebaine on the exposed face of the diene system with a variety of dienophiles. Substituents at C-7, especially the ketonic adducts of type 159 and the secondary and tertiary alcohols of type 160, I R, X, Y, and Z's with number superscripts are uscd throughout for substituents directly attached to the morphine skeletal atoms; R's with prime superscripts arc used throughout for substituents at sites removed from the above.
103
Adducts of Thebaine
--
-
.
have been st~died extensively, along with further structural modifications of the aromatic oxygen and piperidine nitrogen functionalities, for example, C-3 demethylation to the phenol and N-cyclopropylmethylation of northebaine bases. Thus, the initial disclosure by Bentley and his colleagues (175), the major contributors in this area, concerning ready access to a series of Diels-Alder adducts of thebaine and related alkaloids has led to exhaustive chemical efforts using various dienophiles and subsequent transformations in the search for new structural types displaying a separation of analgesic activity from undesired toxicity and addictive liability. Thebaine has therefore provided the raw material for new analgesic agents; its ring-C bridged derivatives, in their complexity, have been structural probes for the opiate receptor. The chemistry of the Diels-Alder adducts of thebaine (106), oripavine (157), and derivatives, yielding the six-membered ring 6,14-endoetheno tetrahydrothebaines and oripavines and numerous derivatives, has been thoroughly reviewed (176,177). The highlights of this work include the C-7 substituent classes of ketones, esters, alcohols, a,f3-unsaturated ketones, amin'es, nitriles, alkanes, and others, many of which not only have intrinsic value as potentially potent analgesics but also have value as precursors of othe; structural types that are available through chemical transformation: modification of the C-7 substituent, 3,6-0-demethylation, reduction of the 6,14-etheno bridge, removal of the C-3 oxygen functionality, substitution in the aromatic nucleus, nitrogen demethylation and alkylation, baseand/ or acid-catalyzed rearrangement, dehydrogenation, nuclear substitution, and ozonolysis. Efforts in this area are continuing; a few of the classical examples and a sample of recent work are highlighted below. A.
159
Diels-Alder
Ketone, Sulfone, Nitroso, Ester, and Nitrile Adducts
1. Synthesis Using Ethylene Dienophiles Thebaine readily undergoes Diels-Alder additions of unhindered dienophiles (176,177), various substituted ethylenes (178,179) (Scheme 3-27), and nitroso compounds (180) such as nitroso-carbonyls (181-184) and -arenes (185-187). Due to the e'lectron-rich Coring, a number of Dlels-Alder reactions with other dienophiles and molecular rearrangements of the resulting adducts have been investigated (188). In each case, the dienophile approaches thebaine from the face containing the nitrogen bridge, which is the least hindered side of the molecule. The reaction is under electronic control; unsymmetrical dienophiles such as alkyl and aryl vinyl ketones (189), acrylic esters, or acrylonitriles (178,189) add to give only C-7 substituted 6,14endoethenotetrahydrothebaines (Table 3-14), with classical examples being thevinone (166), nepenthone (168), and ethyl thevinoate (170),
3
104
Synthesis of Morphine,
Codeine,
and Related Alkaloids
III
Diets-Alder
105
Adduets of 1l1ebaine
Table 3-14 Diels-Alder
Products Formed from Thebaine and Ethylene Dienophiles
/
I
-CH3 2
Thebaine 106
/
161
161 Dienophile Substituents R'
106 162
70',80'
163
7 IJ , 81J
166° 167 168' 169 170d 171 172 173 174 175 176 177 178
.
H H H H H Cl Sr OAe H CH3 Cl H H
R2 COCH3 COCH2CH3 COC6HS C02CH3 C02CH2CH3 C02CH3 C02CH2CH3 C02CH2CH3 CN CN CN S02CH3 S02CHCH2
b Reference 189. Thevinone. d Ethyl thevinoate. 'Reference
1647/1,8(¥ Scheme 3-27.
Diets-Alder
165 7u,81J adduct products of thebaine with ethylene dienophiles.
Product (in %)
<5
6 100' 100' 70' 50 lOO' 75'
25 lOO!
lOOt
C Nepenthone. 178. ! Reference
190.
formed from methyl and phenyl vinyl ketones and ethyl acrylate, respectively. There is no evidence for the production of C-8 substituted compounds. The stereochemistry of the product is almost entirely 70' for monosubstituted ethylenes; the acrylonitrile product (174) is an exception, having considerable amount of 7{3 formation (189). Most 70: sulfonyl compounds (190) have been epimerized by treatment with hot ethanoJic sodium hydroxide, whereas other thebaine Diels-Alder adducts that have an electron-withdrawing 70'-substituent and a 7{3-hydrogen suffer basecatalyzed rearrangement. Most of the ketone, ester, and nitrile Diels-Alder adducts of thebaine and their derivatives formed from monosubstituted or 1, I-disubstituted ethylenes have had disappointing biological activity as either analgesic
106
3
Symhesis of Morphine,
Codeine,
and Related Alkaloids
Table 3-16 Analgesic Activity of 7a-~u)fQne_Diels-Alder
Table 3-15 Analgesic Activity of7a-Ketone Adducts
Diels-Alder
Adducts
/ Analgesia,a Tail Pressureb
70., R' Group H CH3 CH(CH3h C6HS (CH2hCH3
b
177 179 180 181
0.6< 1.2 9.0 0.7 0.03
a
0.07d 0.12 0.13 0.17
bRats, ip.
< ED50 of morphine = -1. 7 mg/kg. d Relative to morphine = 1.0; codeine
a Reference 192. bRats, sc. < Relative to morphine = 1.0.
agonists or antagonists. A number of adducts with varying structures, however, have displayed notable pharmacological profiles, as discussed below, with trends being outlined by Lewis (191). The 7a-ketonic adducts of type 161, in which R = hydrogen, methyl, isobutyl, phenyl, and octyl, have displayed analgesic potencies relative to morphine (Table 3-15) in various protocols, namely, rat tail pressure, mouse hot plate, and mouse phenyl benzoquinone writhing. Maximum activity resides in the isobutyl analog, which has nine times the potency of morphine (192). Interestingly, thevinone (166) is equipotent with morphine. By contrast, ethyl thevinoate (170) shows no analgesic activity up to doses of 200 mg/kg. The sulfonyl compound 177 and certain tertiary amines, derived from vinyl sulfone adduct 178 by treatment with secondary amines, have shown analgesic activity in rats. Approximately 5-15 times less potent than morphine, compounds 177, 179, 180, and 181 have a potency similar to that I of codeine (190) (Table 3-16). The adducts derived from addition of aromatic nitroso compounds are valuable intermediates to previously inaccessible analgesic 14arylaminodihydrocodeinones (184), due to facile acid hydrolysis of the precursor 1,2-oxazines (182) to 14-(N-hydroxyarylamino)codeinones (183) followed by catalytic reduction (193) (Scheme 3-28). Codeinones and dihydrocodeinones that have 14-alkyl or 14-alkenyl substituents are
23< 14 13 10
CH3 CH2CH3 2-Morpholinoethyl 2-Piperidinoethyl Reference 190.
Potency
ED so (mg/kg)
70., R' Group Analgesia," Tail Pressure Molar Potency Ratio
potency
= 0.2.
O=N-@X X
II - CH
3 .C1.H
>
182
106
1.
(
b
184 Scheme 3-28. Synthesis or 14-arylaminocodeinones using nitroso adducts. Reagents: (a) acid; (b) catalytic hydrogenation.
183 arene Diets-Alder
108
3
Synthesis of Morphine,
Codeine,
and Related Alkaloids
III -Dieis-Alder
Adducts of Thebaine
109
Table 3-17 Analgesic
Activity
of 14-(Arylamino)eodeinones
a,b
185,186
187 188
Analgesia a Writhing
R Group
185 186 187
Testb
EDso (mg/kg)
CH3 CPM' CH3
Straub Tail Testb ADso (mg/kg) c,d
-1O.5d
-5 >20
a b Reference 193. Mice, se. e EDso of morphine = 0.26 mg/kg. d , EDso of naloxone = 0.09 mg/kg. CPM = cyclopropylmethyI.
isolated as rearrangement products from the 7a-ketones by first treating these Diels-Alder adducts with the appropriate Grignard reagent and then refluxing in formic acid (ct. Section III,B,3). Of the 14-(arylamino)codeinones, only two p-fluorophenyl derivatives (185 and 186) have displayed any activity (Table 3-17). Most N-methyl compounds are only one-third to one-tenth as potent as morphine and display no antagonist activities, such as 185. The N-cyclopropyl derivatives (186) have only slight antagonist activities. A 14-hydroxylamino intermediate (187) has displayed an unusual mixture of agonism in the writhing assay but not in the tail flick test, where instead antagonism and excitation occur (193). Most 7a-substituted 6,14-endoetheno and 6,14-endoethano tetrahydrothebaines, including ketones, esters, and alkyls, can be selectively demethylated. For example, ethyl thevinoate (170) is readily converted to the corresponding oripavine (188) (Scheme 3-29) by heating an intermediate C-19 ketal, formed by mild treatment with a trialkyl orthoformate ester, with potassium hydroxide in diethylene glycol (194). Using the ketal derivative of esters and ketones avoids transformation to the usual
170
189
c,d
190
ylation syntheses. Reagents: (a) trialkyl orthofor. . mate; (b) potassium hydroxide, diethylene glycol; (c) hydrogen bromide, ace lie aCId ; (d ) Seherne 3 29. -
ethanol,
C-3
hydrochloric
and
acid.
C-6
O-demeth
3
lIO
Synthesis of Morphine,
Codeine. and Related Alkaloids
base-catalyzed rearrangement products. Treatment of 170 at room temperature with hydrogen bromide in acetic acid yields, after reesterification, the 6-0-demethyl analog (189). Prolonged treatment yields the bis-demethyl compound 190 (1~5). By analogous procedures, N-substituted demethyl and bis-demethyl analogs have been prepared, some of which have shown slight but not significant antagonist activities (194). 6,14-Endoethenotetrahydrothebaine (196), the C-? unsubstituted pa,rent, can be prepared by Huang MinIon reduction of the 7-oxo bridged 'thebaine derivative (191) (196). 6,14-Endoetheno ketone (191) itself is
,
III
Diels-Alder
Table 3.18
\
Analgesic Activity Derivatives
191 193 192 194 195
191
a e
,
170 198 199 200
R1
R'
CH3 CH3 H H
CH3 H H H
CH2CH3 CH2CH3 CH2CH3 (CH2)3CH3
inactive as an analgesic at 100 mg/kg, but its 6,14-endoethano ketone (192) is half as potent as morphine. The two N-methyl substituted 7-oxotetrahydrooripavine derivatives 193 and 194 are about two and four times more potent than morphine, respectively (Table 3-18). The Ncyclopropylmethyl thebaine analog (195) is a mild antagonist, having one-half the activity of pentazocine in the phenylbenzoquinone writhing
of 7-0xotetrahydrothebaine
R
R1
Analgesia," Tail Pressure Testb
CzH2 CzH2 CzH4 CzH4 CzH2
CH3 H CH3 H CH3
CH3 CH3 CH3 CH3 CPM'
Inactive 1.5e 0.5 4.0 0.18,1
Reference 196. Relative
and Oriparine
X
bRats, ip.
to morphine
d Antagonism
R
III
Adducts of Thebaine
relative
= 1.0.
to nalorphine
= 1.0.
CPM = cyciopropylmethyl.
assay. The thebaine base (196) is equipotent with morphine, while conversion to the oripavine base (197) results in an increase in analgesic potency of 30-40 times (196). Reduction of the double bond in either compound results in little change in analgesic potency compared to the unsaturated bases. The increase due to O-demethylation at C-3 occurs as a general phenomenon, increasing analgesic potency approximately 10-50 times, as evidenced also by the eightfold increase in rat tail pressure test activity (sc) for ethyl ester (199) compared to morphine (195). Demethylation of C-6 also increases analgesic potency. Whereas the 6-methoxy ethyl ester (170) is inactive, the 6-hydroxy analog (198) is equipotent with codeine in the rat tail pressure test. lncreasing the C-7 ester alkyl chain length in the 3,6-dihydroxy compounds increases potency, with the butyl ester (200) being 40 times more potent than morphine (195). In addition to thebaine, other structurally related alkaloids (197-199) react with methyl vinyl ketone, giving only a single isomer, the 7a-acyl product. 6-Demethoxythebaine (201) and ,B-dihydrothebaine (202) form analogous Diels-Alder addition products. ,B-Dihydrothebaine does not, however, give the conventional Diels-Alder adduct products when reacted with acetylenic dienophiles (200).
3
112
Synthesis of Morphine,
Codeine,
and Related Alkaloids
III
Diels-Alder
113
Adducts of Thebaine
----'> a
I I
\ : 7 R
201
2. Base-Catalyzed Rearrangements
Diels-Alder adducts containing a C-7 ketone, such as thevinone (166), suffer base-catalyzed rearrangement (189,201) (Scheme 3-30, R = COCH3), initiated by enolization of the ketone toward C-7. Both tex and 7{3Diels-Alder isomers yield the same rearrangement products, since, under reversible conditions, removal of the C-7 proton to form the enolate makes C-7 no longer asymmetric. C-7 carbanion attack on the 4,5-oxygen bridge, with the formation of a C-7-C-5 carbon-carbon bond, produces a phenoxide anion at C-4 (204), which is either stabilized by alkylation (206) or attacks C-6 (205). Although the three isomeric bases (203, 204, 205) are in equilibrium in methanolic potassium hydroxide, base 205 is favored due to relief of strain arising from the 4,5-oxygen bridge and is therefore the rearranged base obtained in excellent yield. Methoxide attack on 204 at C-6, C-17, or C-19, which is the carbonyl of the C-7 substituent, is also possible in the methanol solvent. The two primary base-catalyzed rearrangement products of thevinone (207, 208) can be hydrolyzed with acid under mild conditions to yield an ex,{3~unsaturated diketone, 18-acetyl-5,14-endoethanothebainone (209). Dlels-Alder adducts containing a C-7 ester or nitrile (Scheme 3-30, R = C02CH2CH3, CN) are also rearranged by base, but the products do not usually proceed beyond bases 210 or 211 due to the decreased electron-accepting power of the C-7 ester or nitrile group compared to that of the C-7 ketone. Such groups are ineffective in producing the equivalent of 205 (201,202). Further hydrolysis by acids yields the ex,{3-unsaturated ketones 212 and 213, which are analogous to structure 209. The Diels-Alder ketone nepenthone (168) readily forms the novel . Isonepenthone (214) (201,203) when refluxed in 5% methanolic sodium hydroxide. This formation was quite surprising when first observed, but the rearrangement product is stable due to relief of steric strain and minimization of nonbonded interactions. Mild acid hydrolysis of isonepenthone initially yields the C-18 {3-epimer, which equilibrates to pseudonepenthone (215). An unusual base-catalyzed rearrangement (Scheme 3-31) is that of the dihydrothebainequinone (216) (204), which gives the phenolic diketone
205
203
202 166
R = COCH3
170
R = C02Et
174
R = CN
168
R = COC6HS
)
207
R = COCH3
21.
R'
COC,',
~
R 206 R
1 =
H,
CH3
1 R = H, CH3 R = COCH3
208
R = COCH3
209
210
R = C02Et
212
211
R = CN
213
R = C02Et R = CN
215
R = COC611S
Scheme 3-30. Base-catalyzed rearrangement products of ketones. esters, and nitriles. Reagents: (a) potassium hydroxide; (b) alkyl halide; (c) acid hydrolysis.
(217) (205). This product is sensitive to air oxidation and thus produces the isolatable dihydroflavothebaone enol methyl ether (218), the reaction being driven by aromatization of the cyc1ohexanedione ring. Hydrolysis yields a new alkaloid base (219), the 6,14-bridge saturated analog of the basecatalyzed rearrangement product of thebainehydroquinone (206), which then autoxidizes in alkaline solution to flavothebaone (220). Much interesting chemistry has evolved from investigation of the base-catalyzed rearrangement products of Diels-Alder ketones, ester and nitriles. Unfortunately, the few compounds that have been tested for analgesic activity have been inactive. The dihydrothebainequinone adduct (216) has only one-tenth to one-fifth the potency of morphine in the rat tail
pressure test by sc administration (204).
114
3
Synthesis
of
Morphine,
Codeine,
and
Related
Alkaloids
III
Diels-Alder
b -..
216
115
Adducts of Thebaine
a ~
217
218
H
221 166
~
Scheme
e
220 (flavothebaone) 3-31. Novel base-catalyzed rearrangement
219
R
R1
R
222
H
CH3
223
CH3
CH3
224
H
CH2CH3
225
CH3
CH2CH3
R'
3-32.
Synthesis of 7-alkyl derivatives. Reagent:
(a) Huang
Minion.
- CH 3
Scheme of dihydrothebainequinone. Reagents: a, potassium hydroxide; b, airoxidationor potassium hydroxide; c, methyl sulfate;
d, hydrolysis;
e, alkali.
3. Alkyl, Carboxyl, and Amino Derivatives Many derivatives of 6,14endoethenotetrahydrothebaine have been obtained from the 7a-ketone, -ester, and -nitrile adducts, which have repeatedly proved to be favorable starting materials. Huang Minion reduction of the 7-acyl products has provided 7a-alkyl oripavines, whereby demethylation of the phenolic ether has also occurred (207). Thus, the Diels-Alder adduct aldehyde (221) and methyl ketone (166) yield 7a-methyloripavine (222) and 7a-ethyloripavine (224) respectively (Scheme 3-32), both of which can be methylated to the corresponding thebaine analogs 223 and 225, all of which can be converted to N-substituted 7a-alkyl derivatives via either the nor-compounds or the N-acyl intermediates. A 7-methylene product (227) has been obtained by Hofmann methylation of the 7dimethylaminomethylene derivative (226), obtained from the ethyl acrylate adduct (221) via the 7-dimethylamino amide intermediate. The 7-ethylidene compound (228) has resulted from the p-toluenesulfonate of the C-19 secondary alcohol of 166.
R 226
CH2N(CH3)2
227
=
CH2
228
=
CH-CH3
The 7a-methyl and -ethyl oripavines 222 and 224 are equipotent with and three times less potent than morphine, respectively, in the rat tail pressure test by sc administration, whereas the methylene and ethylidene thebaines 227 and 228 have little or no activity (196). A tetrahedral carbon at C-7 therefore seems to be associated with higher activity than a trigonal center. The ethyl acrylate adducts of thebaine have been used as starting materials for the incorporation of peptide segments derived from the endogenous opiate ligand leucine-enkephalin (229), in essence resulting in
116
3
Synthesis of Morphine,
Codeine, and Related Alkaloids III
the union of the rigid morphine ring system of the "Bentley adducts" with the lipophilic portion of the enkephalins (208). The rationale rests on the observations of common core elements of leucine-enkephalin compared to morphine skeletons: (a) the tyrosine moiety, in which the tyrosine residue in the natural ligand, in contrast to morphine, is flexible enough to adopt various conformations, and (b) the "tail" of the enkephalins, corresponding to the C-19R alcohol portion of known very potent analgesics, such as etorphine. Thus, the major isomer 7a-ethyl thevinoate (170) has been hydrolyzed with base to the thevinoic acid (230) and then converted to the acid chloride (231) by reacting with oxalyl chloride before coupling with L-Ieucine, L-phenylalanyl-L-leucine, and glycyJ-phenylalanyl-L-leucine ethyl esters to yield thebaines 232, 233, and 234, respectively (Scheme 3-33), which can all be demethylated to give tthe corresponding 3,6dihydroxy derivatives using hydrogen bromide in glacial acetic acid. For the L-phenylalanyl-L-Ieucine analogs, the C-terminal ester of the peptide has been subsequently reduced to the primary alcohol using sodium borohydride. The leucine-enkephalin dihydroxy analog (235) is a potent
a,b
--
230
R'
OH
231 R'
C1
J
-
COR'
1 170
232 233 234 235
R'
OEt
R
R'
CH3
Leu-OEt
CH3 CH3 H
Gly-Phe-Leu-OEt
Phe-Leu-OEt
-
C02R'
Leu-OEt
Scheme 3-33. Synthesis of peptide morphine derivatives. ide; (b) oxalyl chloride.
Reagents: (a) sodium hydrox-
Diels-Alder
Adducts of Thebaine
117
GlY-Gly-Phe-Leu
HO 229 Etorphine
Leucine-Enkephalin (Tyr-Gly-Gly-Phe-Leu)
analgesic in the mouse hot plate assay (sc), displaces etorphine in the opiate receptor binding assay, and has an overall pharmacological profile similar to that of morphine. Opiate receptor probes have been synthesized by alkylation of a 7a-amino oripavine adduct (236) (Scheme 3-34), this Diels-Alder adduct being conveniently obtained from the 7a-acetyl phenol of thevinone (166) via a 7a-acetamide intermediate obtained by the Schmidt reaction (209). Thus, 236 has been converted to the isothio-, cyanato-, bromoacetamido-, and methylfumaramido-derivatives; each alkylating agent has then been
assayed for
po
and 8 receptor selectivity (210). Only 7a-methylfumaramido
6,14-endoethenooripavine (FAO, 237) has proved to be a highly selective alkylator of 8 receptors with no cross-reactivity for po receptors. An alternative synthesis of 7a-amino compounds has involved Curtius degradation of the 7a-ethyl ester (170) via the hydrazide and azide to give the benzyl urethane (238), which yields 236 on hydrolysis (209). The 7a-ethyl ester (170), by way of the acid chloride (231), on treatment with primary or secondary amines yields 7a-amides (239) (211) (Scheme 3-35), including the 7a-acetamido base (240), although some of these are also available by direct Diels-Alder addition of acrylamide dienophiIes to thebaine (212). Reduction with lithium aluminum hydride yields 7-a-aminomethyl compounds (241); reduction of the primary amide (240) yields the methylamine (242) only when a large excess of reducing agent is used (211). The 7,B-methylamino, 7a-methyl compound, howev~r, can also be obtained directly from the 7,B-cyano, 7a-methyl adduct by reduction with lithium aluminum hydride (213). Derivatives with a nitrogen substituent other than methyl and the oripavine bases have also been 'prepared. The 7a-amido and 7a-aminomethyl derivatives in each Nsubstituted series all have analgesic activity less than that of the corresponding secondary alcohols. None of the intermediates has any significant analgesic activity. The effect of C-3 O-demethylation generally accounts for a 5- to lO-fold increase in analgesic activity, as has been observed in other series of the morphine-codeine group.
118
3
Synthesis of Morphine,
Codeine, and Related Alkaloids
III
Diels-Alder
Adducts of Thebaine
119
HNR'R" ~
231
, "
CONR
'R"
166 239
240
236
R'
R"
H
/
IINHR'
OCH3
237
t R'
R' = R"
= COCH=CHC02CH3
H, alkyl, aryl
(FAO)
241 242 R' Scheme 3-35. num anhydride.
= R" = H Synthesis of7a-amido
and methylamino
derivatives.
Reagent:
238
i
c,d H (COR) COR (H)
170 Scheme 3-34. Synthesis of an FAa opiate receptor probe. Reagents: (a) sodium azide, perchloric acid (Schmidt); (b) hydrolysis; (c) hydrazine; (d) sodium nitrite, acid; benzyl alcohol.
'~,~
"'COR (H) " 'H (COR)
4. 7,8-Disubstituted Adducts In addition to monosubstituted and 1,1disubstituted (178) ethylenes, cis 1,2-disubstituted (179) ethylenes add readily to thebaine. Trans 1,2-disubstituted ethylenes, both symmetrical and unsymmetrical, containing bulky groups (179), have been used
243
R
CH)
70',80'
successfully
244
R
CH)
7{3,
245
R
phenyl
70', 80'
246
R
phenyl
7{3,8a
as dienophiles.
For cis and trans diacetylethylene
COCH3, Scheme 3-27) and dibenzoylethylene (Rt =
R2
(R1
=
R2
=
= COC6Hs), the
exclusive isolated product bases have been characterized spectroscopically as the 7a, 8a (243, 245) and 7{3, 8a (244, 246) diester compounds. The
80'
lithium alumi-
120
3
Synthesis of Morphine,
Codeine. and Related Alkaloids
III
Diels-Alder
247
256 CH3
~
RO
'
.
CH3
COR' ~255
o
121
Adducts of Thebaine
I
R'
R"
R'
C6H5
R"
OCH3
C6H5
~257
R'
R"
258
R'
C6H5
R"
OCH3
C6H5
,CH3
-
CH3
.~ ~"'''CH3
OCH 3
250
R = CH3
251
R = H
COR' 254
c,d 1
d' 252 Scheme 3-36. 7.8-Disubstituted adducts. Reagents: (a) hydrochloric or phosphoric acid, ethanol; (b) lithium aluminum hydride; (c) p-toluene sulfonyl chloride; (d) potassium t-butoxide, butanol; (e) hydroxide; (f) acid hydrolysis.
R'
260
R'
C6H5,
R"
OCH3
R"
= C6H5
COR'
Scheme 3-37. Base-catalyzed rearrangement
more stable stereoisomeric products result from the preference of bulky substituents for the a-configuration at C-S due to the proximity of the nitrogen-containing ring. No isomeric 7a,S{3 and 7{3,S{3adduct bases are formed. Entrance to 7,S-disubstituted adducts (214) can be achieved either directly by reaction with trans or cis 1,2-disubstituted dienophiles (179) (Scheme 3-27) or by conversion of the maleic anhydride adduct (247) to the diester (248) (204), which can then be reduced with lithium aluminum hydride to the diol (249) (Scheme 3-36). This diol serves as a useful starting material for the generation of additional 7,S-disubstituted adducts, for
259
hydroxide;
of 7,8-disubstituted
adducts. Reagents: (a)
(b) acid hydrolysis.
example, by conversion to the ditosylate, followed either b7 hydrogen~lysis with lithium aluminum hydride to give the 7a, Sa dimethyl codlde (250) or by reaction with potassium (-butoxide in (-butyl alcohol to yield the diene (252). Diester (248) is unstable toward base and is rearranged in a manner similar to that of the C-7 monoester (170), giving the vicinal trans diester (253) which can be hydrolyzed with dilute acid to the phenolic trans disubstituted a,{3-unsaturated ketone (254). When the 7-substituent is a phenyl ketone (Scheme 3-37), both cis and trans orientations of the C-7, C-S substituents in the compound pairs 255,
122
3 Synthesis of Morphine,
Codeine,
and Related Alkaloids
257 and 256, 258 yield the same base-catalyzed rearrangement hemiketal dicarbonyl intermediate due to C-7 epimerization. On acid hydrolysis, the diketones in each case afford a phenolic trans disubstituted a,{3unsaturated ketone: 255, 257 yielding 259 and 256, 258 yielding 260. The dialkyl compounds show improvement in analgesic activity compared to the monoalkyls. The 7a,8a-dimethyl base (250) is 18 times more potent than morphine; its oripavine analog (251) has 200 times the potency of morphine (213) in the rat tail pressure test (sc). But the tetahydrofuran base (261) has only 0.9 times morphine's analgesic potency. Conversion to
III
Diels-Alder
]23
Adducts of Thebaine
a Thebaine '"
b
R
Rl
261
CH3
CH3
262
H
CH3
263
CH3
Allyl
264
CH3
CPM
266
267 R' = Me, Et 268 R' = Me, Et Scheme3-38. Adducts formed from acetylenic dienophiles. Reagents: (a) dimethyl acetylencdicarboxylate; (b) methyl or ethyl propriolate.
265
the oripavine (262) results in 12 times morphine's potency; its N-allyl and N-cycIopropylmethyl codides (3-methoxy) (263, 264) are weak morphine antagonists. Interestingly, the N-cycIopropylmethyl phenylpyrrolidino compound (265) is a weak analgesic. In general, any unsaturated substitution at C-8a results in a decrease in or loss of analgesic activity, however, a simple alkyl group increases analgesic potency.
Reaction of thebaine with acetylenic dienophiles (215, 216) (Scheme 3-38) such as dimethyl acetylenedicarboxylate yields an unsaturated 7,8disubstituted analog, compound 266. Reaction with methyl or ethyl propiolate yields the 7-monosubstituted products 267. Since both products are thermally unstable, however, reaction usually gives rise to rearrangement benzazocine isomers. Work with propiolic esters has demonstrated that reaction of thebaine with acetylenic dienophiles is solvent dependent (200), with new types of adducts (268) being isolated in polar solvents (216, 217); these abnormal products result from C(9)-N bond scission. Additional work has been done on the irradiation products of 266 (218) and its 6,14-endoethano derivatives.
1Z4
3
Synthesis of Morphine.
Codeine. and Related Alkaloids
III
Diets-Alder
Adducts of Thebaine
1Z5
5. H-813 can be assigned on the basis of spin decoupling experiments since H-813 is coupled to H-7Ci and H-8Ci, whose chemical shift is known. 6. Very small long-range coupling exists between H-813 and H-17, the stereochemical environment being similar to that between H-713 and C-18.
Fig. 3-/.
5. Stereochemical Assignmellls Stereochemical assignment of the C-7 and C-8 epimers of thebaine and oripavine bases formed by Diels-Alder addition of electrophilic olefins and subsequent transformation has relied heavily on proton NMR, including homonuclear field-sweep spindecoupling and nuclear Overhauser effect (NO E) techniques. In addition to elucidation of the absolute configuration of the asymmetric centers generated by these additions, NMR and supportive infrared (IR) and ultraviolet (UV) spectroscopy have elucidated the alkaloid base skeleton and stereochemistry of base-catalyzed rearrangement products. In summary, the NMR findings indicate the following for virtually all Diels-Alder adducts studied (219) (Fig. 3-1): 1. The chemical shift for H-513 is diagnostic for C-7 configuration assignment, the range for the average chemical shift being, with few exceptions, 4.57:t 0.05 ppm for a 7Ci substituent and 5.07:t 0.12 ppm for a 713substituent. This downfield shift in 713compared to 7Ci compounds also holds for all analogs produced by modifications of the initially formed adducts. 2. H-513 and H-713 are coupled to H-18 in the 7Ci-compounds, wherein the three protons and the intervening carbon atoms approximate a plane in which the connecting bonds resemble a W. 3. The C-8 geminal protons show unique and characteristic variations in chemical shift, influenced by the C-17-C-18 double bond as well as by the nature and stereochemical orientation of the C-7 substituent. 4. H-813 and H-8Ci show a large difference in chemical shift based on the difference in the shielding effect of the tertiary nitrogen atom (electron pair) held in proximity to H-813 by the rigid ring system. H-813 appears downfield from H-8Ci, whose average chemical shift is, with few exceptions, for either a 7Ci or a 713substituent, 1.35 :t 0.15 ppm.
This NMR documentation, including the experimentally observed spin systems, coupling constants, and anisotropic tertiary nitrogen effect on the C-8 protons, confirms that the Diels-Alder adducts of thebaine are not C-8 but C-7 derivatives, having an "endo" and not an "exo" disposition of the 6,14-etheno bridge and being "inside" the tetrahydrothebaine ring system. The C-5 proton in the reduced C-19 tertiary carbinol derivatives of the ethyl acrylate product (170) has proved characteristic when compared to the C-5 proton in the methyl vinyl ketone adduct (166) (189): 4.60 ppm (7Ci-) and 5.17 (713-) for the monosubstituted ethyl esters versus 4.55 ppm (7Ci-) and 4.98 (713) for the methyl ketones. Chemical shifts of the methyl and methylene protons of the ethoxy .group, in addition to the C-5 proton, have characterized ethyl ester adducts 'resulting from 1,1-disubstituted dienophiles (178). In the adducts studied (171,172, 173) and compared to the 7,7-bisethoxycarbonyl derivative, the three ethoxy signals absorb at lower fields when the ethoxycarbonyl group is in the l3-configuration. For example, the ethyl 2-acetoxyacrylate dienophile produces a 7Ci-carboxy ethyl ester product (173) that has NMR signals at 4.17 ppm (-CH2), 1.21 (-CH3), and 5.45 (5-H), whereas the 713-epimer has 4.32-, 1.32-, and 4.71-ppm shifts, respectively. Both the H-513resonance and the NOE between the 7Ci-methyl and 8Ci-H and 813-H have characterized the 7Ci-methyl-713-cyano-thebaine adduct (175) (178). In 7,8-disubstituted compounds (244, 246), the coupling constant (6 H~) for H-713 and H-8o: corresponds to a transoid vicinal arrangement that IS distinguishable from the cisoid (11 Hz) (179). In addition, H-5 has long-range coupling with H-8, showing H-5 to be 13. The low field resonance of H-8 is diagnostic for 13stereochemistry, based on the effect of the the nitrogen lone pair. Increasing or decreasing the electron density at nitrogen, that is, preparation of N-oxides or addition of acid, with concomitant measurement of the absorption of the H-8 resonance, allows confirmation of the 813-H assignment. For rearrangement products such as 253, the olefinic bridge signals for H-8 (doublet, 5.83 ppm; J8.7 = 9 Hz) and H-7 (doublet of doublets, 5.08 ppm; J7.8 = 9, J7.5 = 2 Hz) characterize and differentiate the ether from a phenolic product formed from ether ring opening (214).
126
3
Synthesis of Morphine,
Codeine,
and Related Alkaloids
The sulfonyl adducts 177 and 178 have been assigned as 7a on the basis of the difference in the resonance of the CoSH proton when compared to the generated 7{3-epimers (190). The 7{3-sulfonyl group also produces an effect on CoSH that is quantitatively similar to that observed with other 7{3-electron attracting groups. In cases where structural assignment of stereochemistry has not been possible by proton NMR spectroscopy, derivatization has proved useful. For the l,l-disubstituted ethylene, 2-chloroacrylonitrile, the epimeric ad ducts (176) had to be converted to spiroaziridines 269 and 270 by reduction with lithium aluminum hydride (178). Knowing that cyclization
III
Diets-Alder
127
Adducts of Thebaine
166
y
273
CH3
_
CH 0 3
I 269
R'
H
270
R'
H
271
R'
P-C1C6H4CO
272
R'
P-C1C6H4CO
is stereospecific and proceeds with inversion of configuration allowed (a) interpretation of proton NMR data for these spiroaziridines and the two p-chlorobenzoyl derivates 271 and 272, and (b) assignment of stereochemistry at C-7 in the original thebaine adducts, the most abundant being the 7{3-cyano-7a-chloro isomer and producing 269. The aziridines have only slight agonist activity in the rat tail pressure test; their p-chlorobenzoyl derivatives (271, 272), however, are three and seven times more potent than morphine, respectively, as analgesics in the rat tail pressure test (sc) (178). B.
Functionalization
at C-19: Alcohols
Although the adduct of thebaine and methyl vinyl ketone, thevinone (166), is rearranged in base via the C-7 carbanion under reversible enolization, the initial enolization of the ketone takes kinetic control, so that valuable products of reaction at the terminal carbon are possible. One useful sequence (Scheme 3-39) involves treatment of 166 with trimethyl orthoformate and perchloric acid in methanol to yield dimethyl ketal (273), followed by thermal elimination of methanol to yield an enol ether, both of
CHO
f: ~ ~
274
=
tJ
275
R
276
where X
=
where X
NR';
R'
=
0
= H, alkyl, aryl
R277
R =
278
O~R,
(Jl
NH2 R
Scheme 3.39. Reactions of thevinone at the terminal carbon atom. Reagents: orthoformate, perchloric acid; (b) heat; (c) phosgene, dimethylformamide.
=
alkyl (a) trimethyl
which form the methoxy enal (274) by reaction with the Vilsmeier reagent formed from phosgene and dimethylformamide. This enol ether of {3ketoaldehyde condenses readily with a variety of active methylene compounds and reacts with alkyllithiums to form 7a-substituted heterocyclic compounds (220), for example, the bases (275-277) and the a,{3unsaturated ketones (278), which can be reduced to saturated ketones. The dimethyl ketal (273) is the key for the synthesis of 3-hydroxy and N-substituted 7a-ketone and 7a-ester substituted bases. All of these compounds prove to be difficult to make by other conventional routes.
128
3
Synthesis of Morphine,
Codeine,
and Related
Alkaloids
R' : ~IIC ~R" \ OH OCH)
a or b or
c
(A) R'
R'
R"
H
CH)
166
CH)
279
169
OCH)
280
CH3
H
170
OCH2CH)
281
CH3
CH3
III
Diels-Alder
Adducts of Thebaine
129
configuration of the product. When no competitive processes are operative, the grignard reaction is highly stereoselective, giving a high yield of one pure diastereoisomer of the tertiary alcohol, whereas reaction with lithium alkyls shows less stereoselectivity and produces more side products. Reactions of ketones with grignard reagents R"MgX, however, can be complex, leading to the following types of products (221).
1. Tertiary alcohol (A), the normal and the major grignard reaction product
2. Tertiary alcohol diastereoisomeric
with (A), also a normal but a minor reaction product 3. Secondary alcohol ((A), where R" = H), the major grignard reduction product formed when R" contains a l3-hydrogen (borohydride reduction product)
4. Secondary
alcohol
diastereoisomeric
with
(A),
R" = H, a minor
reduction product (Meerwein-Pondorff reduction product) 5. C-4 phenol, from base-catalyzed rearrangement of the ketone followed by grignard, reaction at the carbonyl group
282 R' = R" = CH3 Scheme ]-40. Synthesis of C-19 alcohols of 6, 14-endocthenotetrahydrothebaine. Reagents: (a) grignard reagent or lithium alkyl; (b) sodium borohydride; (c) aluminum isopropoxide.
1. Thevinols and Orvinols To further study modifications of the 7-keto group involving removal of electron density at this position, secondary and tertiary alcohols have been readily prepared from 70'.aldehydes and 7O'.-ketones by reaction with aluminum I alkoxides, sodium borohydride, lithium alkyls, and grignard reagents. For example, reduction of thevinone (166) with aluminium isopropoxide yields the secondary alcohol 279 almost exclusively, although switching to sodium borohydride produces a 50:50 mixture of the two C-19 diastereoisomeric alcohols 279 and 280 (Scheme 3-40). Treatment of thevinone (166) with methyl grignard provides a small amount of opened compound (282) along with the dimethyl carbinol (281) (221), which is also obtainable from 7O'.-esters (169) or ethyl thevinoate (170). Generally, the C-19 alcohols can be easily prepared by reaction of grignard reagents or lithium alkyls on the monosubstituted C-7 aldehydes, ketones, or esters (175), the stereochemistry at C-7 in the starting material being the determinant of the C-19
The grignard reduction process is also. highly stereoselective, usually producing almost exclusively one isomer, although a minor amount of the diastereomer can be isolated. In some cases, this reduction process seriously competes with the normal grignard reaction, such as in the 6,14-endoethano bases, where 40% of the product is the reduced compound. The most extensively studied series of alcohols is that derived from thevinone (166), which is as potent an analgesic as morphine. Both grignard reaction with and grignard reduction of this ketone lead to products of the same stereochemical series (A), compounds 280 and 281. However, the Meerwein-Pondorff reduction produces products that have opposite configuration at C-19, compound 279, resulting from free rotation about the C-7-C-19 single bond and preferential hydrogen transfer to the "top side" of the carbonyl group in the aluminium coordinated complex of the transition state. Reaction of thevinone with n-propyl or isobutylmagnesium halides also results in diastereomer (280). The series of secondary and tertiary alcohols of the N-methyl thebaine series illustrates the higher analgesic activities found in the tertiary derivatives compared to the secondary bases 279, 280, 286, and 295 in series A (alkyl derivatives) and 298 and 300 in series B (aryl derivatives) (Table 3-19). The effect of increasing alkyl chain length in a homologous series is shown most dramatically in series B, where a one-methylene unit increase from phenyl in 299 to benzyl in 301 results in a 2000-fold increase
3
130
Synthesis of Morphine,
Codeine, and Related Alkaloids
Table 3-19
III
Diels-Alder
Adducts of Thebaine
131
Table 3-20
Analgesic Activity of C-19 Alcohols 6,14- Endoethanotetrahyd rothebaine
of (Thevinols)
Analgesic
Activity
of Thevinols
with Similar Substitution
at C-19
CH3 Analgesia:
279 280 281 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302d 303
Analgesia," Tail Pressure Testb
R'
R"
H CH3 CH3 CH3 CH3 CH3 H CH3 CH3 CH3 CH3 CH3 CH3 CH3 CH3 H CH3 C6HS H CH3 H CH3 CH3 CH3
CH3 (Series A) H CH3 CH2CH3 (CH2hCH3 CH2(CH3h CHiCH3h (CH2hCH3 CH2CH2(CH3h CCCH3h (CH2)4CH3 (CH2hCCH3h (CH2hCH3 (CHzhCH3 eyclopentyl cyclohexyl cyclohexy 1
CH3 (Series B) C6HS C6Hs CH2C6Hs CH2C6Hs (CH2hCt;HS (CH2hC6Hs
1< 0.09 2.7 20 96 10 5.3 24 2.5 0.1 IS 30 2 0.03 1.0 9.0 59 0.09 0.01 0.07 7.6 150 500 2.1
a
Reference 221. bRats, SC. < Relative to morphine
= 1.0.
d
Phenethyl thebaine (PET).
in potency; an additional unit increase (302) leads to a further 3-fold increase and then to a sharp 250-fold decrease (303). For alkyl substituents of series A, a peaking effect in potency occurs at a 3- to 5-carbon chain (284, 287, 290). Branching effects in R" vary, usually, however, decreasing in potency relative to the straight chain counterparts (285, 288, 289). In the thebaine N-methyl series, compound 302 is the most potent analgesic.
R' 304 305 306 307 308 309 310 311 312 u
R"
H CH2CH3 (CHzhCH3 C6Hs (CHzhCH3 C6HS CH2CH3 C6HS C6Hs Reference 221.
< Relative
CH2CH3 CH2CH3 (CHzhCH3 (CHzhCH3 (CHzhCH3
Cyclohexyl C6HS C6HS CH2C6Hs
bRats,
to morphine
Tail Pressure
Test
b
O.S< 2.5 3.1 0.2 3.0 0.04
o o
sc.
= 1.0.
Only compounds with a hydrogen or methyl group as one substituent of the alcohol display significant activity. The preferred substituent for R' is methyl. For a given R" increasing the size of R' beyond methyl decreases the analgesic potency (Table 3-20). The alcohol 283 is 30 times more potent than its diastereomer, the R-configuration at C-19 usually being the more active of a diastereomeric pair. In these tetrahydrothebaine alcohols, further modifications have been made in an attempt to increase potency. Esterification of the C-19 hydroxyl group in the secondary alcohols 279 and 280 has not affected the analgesic potency significantly; esterification of the C-19 tertiary alcohols has generally been unsuccessful, as well as a-alkylation (221). Demethylation at C-3 in codeine derivatives to yield morphine derivatives has been well documented to result in increased analgesic potency. Demethylation at C-6 in other opiate alkaloid derivatives results in a significant decrease in potency. Therefore, C-3 hydroxyl and C-6 methoxyl are the preferred substituents for maximizing activity. The 6,14endoetheno and 6,14-endoethano C-19 alcohols have been selectively a-de methylated at C-3 to oripavine derivatives by heating with potassium hydroxide in refluxing diethylene glycol, since reaction with acidic reagents
132
3
Synthesis of Morphine, Codeine, and Related Alkaloids
Table 3-21 Analgesic
Activity
of C-3 Substituted
Thevinols
- CH3
II'C_R' \ OH OCH3
R 281 313 314 283 315 284 316J 317 287 318 291 319 320
.
e
CH3 H CaCH3 CH3 H CH3 H CaCH3 CH3 H CH3 H CaCH3
R'
Analgesia," Tail Pressure Testb
CH3 CH3 CH3 CH2CH3 CH2CH3 (CH2hCH3 (CH2hCH3 (CH2hCH3 (CH2hCH3 (CH2hCH3 (CH2hCCH3h (CH2hCCH3h (CH2hCCH3h
2.7e 63 55 20 330 96 3200 8700 24 5200 30 9200 1300
Reference 222. bRats, sc. Relative to morphine = I. d Etorphine.
leads to acid-catalyzed rearrangement products (222). The 4,5-oxygen bridge, a phenolic ether, is unaffected under these conditions. Esterificaton of the C-3 phenolic hydroxyl group of several 6,14-ethenotetrahydrooripavines has been easily achieved by conventional methodology. A substantially more potent series of analgesics has resulted from the demethylation at C-3; esterification only marginally improves the potency over that of the phenols. The alcohols of the grignard reaction product with thevinone (166) illustrate the importance of the C-3 hydroxyl and C-3 acetate in substantially increasing analgesic potency several hundredfold (Table 3-21). The effect is very pronounced when the two alkyl groups on C-19 are methyl and n-propyl (compound 284, the most active analog in series A of Table 3-19). The oripavine (316) is more than 300 times as active as the thebaine base, while the acetate (317) is about 900 times as active as 284, over two times as active as 316. Usually, however, acetylation, while still retaining more
III
Diels-Alder
133
Adducts of Thebaine
activity than morphine, causes reduction in activity compared to the oripavine bases, as in compound 320. One unique derivative of etorphine (316) that has a benzoylthio group replacing the phenolic hydroxyl at C-3 has been used to test the effect of oxygen-sulfur interconversion on analgesic activity (223). S-etorphine (321), synthesized via a Newman-Kwart rearrangement of a thiocarbamate intermediate, has two (oral) and three (sc) times the potency of morphine in mice. This sulfur derivative, however, has very low JJ-opioid receptor affinity, as measured by the inhibition of tritiated naloxone. Etorphine itself has a ~ receptor affinity 20 times that of morphine, whereas S-etorphine has only 1/40th the affinity of morphine, slightly more than 1/800th that of etorphine.
-
fH3 ,,'c
(CHZ)
ZCH3
\H OCH3 321 S-etorphine
Etorphine (Immobilon) (316) [and its C-3 acetate (317), which has almost 9000 times the analgesic potency of morphine] has been extensively studied both in vitro and in vivo (224) and has found widespread use as an analgesic agent to immobilize large wild game animals due to its considerable margin of safety. The diastereomer of etorphine shows a potency dropoff of 50-fold, so that, as usual, the C-19R alcohol is the more potent. Etorphine is 1000-80,000 times more potent than morphine when administered sc, depending on the test protocol and the animal species used (255a). It is absorbed sublingually in dogs; sublingually in humans, a 0.5- to 1.5-~/kg dose is equivalent to 5-10 mg morphine given im or by iv injection (225b). The undesirable side effects are similar to those of morphine; however etorphine depresses the CNS to a greater extent than morphine does. Interestingly, studies on tritiated etorphine cerebral receptor binding in vivo in rats (224a) have demonstrated an extremely low fractional receptor occupancy at analgesic doses (2% at the EDso, tail flick assay). In rats, this analgesia is rapidly reversed, the in vivo dissociation half-life being about 50 seconds. Results support a receptor binding-effect model based on low fractional receptor occupancy at analgesic doses of a pure agonist such as etorphine.
.
134
3
Synthesis of Morphine,
Codeine, and Related Alkaloids
III
Diels-Alder
Adducts of Thebaine
135
Table 3-22 Analgesic
Activity
of 6, 14-Endoetheno-
6.14-Endoethanotetrahydrothebaine
versus C-19 Alcohols
-
a :>
CH3
"y
II I IC
R
I
"- OH
X 281 322 287 323 324 291 325 326 299 327 301 328 a
CzHz ~H4 ~Hz ~H4 ~H4 ~Hz ~H4 ~H4 ~Hz ~H4 ~Hz ~H4
R
R'
Analgesia: Tail Pressure Testb
CH3 CH3 CH3 CH3 H CH3 CH3 H CH3 CHJ CH3 CH3
CH3 CH3 (CHzhCH3 (CHzhCH3 (CHzhCHJ (CHzh(CHJ)z (CHzh(CH3h (CHzh(CH3h C6Hs C6Hs CH2C6Hs CH2C6Hs
2.7e 2.9 24 240 12,000 30 150 II ,000 0.07 0 150 110
References 221,222. bRats, sc.
e Relative to morphine = I.
The N-methyl 3-deoxy compounds of various 6,14-endoetheno tertiary alcohols have been prepared; generally, these are more potent analgesics than the thebaine analogs but less potent than the oripavine derivatives (226). Catalytic reduction of the 6,14-endoetheno tertiary alcohols to the 6,14-endoethanotetrahydrothebaines has been accomplished only at elevated temperature and pressure with Raney nickel (227), although these products can also be formed in reverse sequence by (a) reduction of the 6,14-endoetheno ketone to the ethano ketone and (b) subsequent transformation of the ketone to C-19 alcohol. Grignard reaction of the 6,14-ethano ketone follows the same general patlerns as the 6,14-etheno analogs with the same stereospecificity (221). Reduction of the 6, 14-etheno bridge usually increases the activity of codides relative to that of their unsaturated analogs, such as the lO-fold increase seen in 323 compared to 287 and the S-fold increase seen in 325 compared to unsaturated analog 291 (Table 3-22). While this holds for alkyl substituents at C-19, it does not
R
H,
X
C2H2,
CH3
329 C2H4
R' Y
= R" C-" \OH
(
331 R1
=
c or d,e
I'Y
330
alkyl, alkenyl, alkynyl
Scheme 3-41. Synthesis of nor-thebaine and oripavine derivatives. Re~gents:. (a) cyanogen bromide, heat; (b) potassium hydroxide, diethylene glycol; (c) appropnate halIde; (d) appropriate acyl halide; (e) lithium aluminum hydride.
hold well for aryl substituents in the codide series, such as compounds 3~7 and 328. Reduction in the morphides yields tremendous increases m analgesic activity. The 6,14-endoethano phenols 324 and 326 have 12,000 and 11,000 times the potency of morphine, respectively (222). 2. Nor-thebaine and -oripavine Bases Many C-19 tertiary alcohols with nitrogen substituents other than methyl have been made, many. of these tertiary amines being analogs of nalorphine. The secondary ammo oripavine and thebaine bases (330) are intermediates, being produced from the N-methyl compounds via the N-cyano compound (329) or from a .N, N I-methylenebisnor adduct, produced by reaction with cyanogen bromide and ethyl or methyl azodicarboxylate, respectively (Sche~e 3-41) ~~22). The N-cyano compounds are then hydrolyzed under alkaline conditIOns, conditions also effecting O-dealkylation at C-3, to the nor compounds
136
3
Synthesis of Morphine,
Codeine,
and Related
Alkaloids
(330), which are then alkylated, alkenylated, or alkynylated to Nsubstituted bases (331) by reflux with the desired halide or treatment with the acyl halide followed by reduction of the amide with lithium aluminum hydride. In all cases, pure C-19 isomers of phenolic bases can be obtained if pure starting alcohols are used. Heating the symmetrical or unsymmetrical bisadduct intermediate with alkyl or acyl halides in the same manner also produces stereochemically pure products; however, only reactive alkyl halides can be used and only methoxy (not phenolic) bases can be obtained. N-substituted compounds in both the 6,14-etheno and 6,14ethano series are accessible by these routes. Replacement of N-methyl by hydrogen and other groups has been done in most series of alcohols, in addition to 7a-ketones and -esters, resulting in varying agonist and antagonist effects (222). No particular group has conferred morphine antagonist properties on all bases; however, groups such as n-propyl, allyl, and cycIopropylmethyl have given rise to the highest proportion of narcotic antagonists. For example, the NcycIopropylmethyl thebaine analogs of the 7 a-methyl, 7 a-ethyl, and 7a-propyl ketones, as well as those of the 7a-methyl and 7a-ethyl esters, show narcotic antagonistic activity. The trends (191,228) shown by these N-substituted bases (Tables 3-23, 3-24) are as follows: 1. Generally, oripavines have greater morphine antagonism than the corresponding thebaines, so that there is less intrinsic agonist activity. For example, in the N-allyl series, a weakly analgesic thebaine, compound 334, gives a strong antagonism on demethylation at C-3 to the oripavine base (335). 2. The size of the R' group and the C-19 configuration both influence the intrinsic activity in the N-substituted 6,14-endoetheno bases. 3. In the N-methyl Diels-Alder adducts of thebaine (Table 3-19), increasing the size of the R" group when R' is methyl tends to increase analgesic activity. In the C-19R series of allyl and CPM northebaine derivatives (Table 3-23), potency also increases with increasing chain length when one C-19 substituent is methyl. Maximum analgesic potency peaks at the R' = n-propyl, butyl range (compare 336 to 340, 334 to 338). 4. In the N-cycIopropylmethyl nororipavines, the change from methyl (333) to ethyl (337) increases antagonism, but a further increase in alkyl chain length to the propyl or isopentyl carbinol (341) confers powerful agonism and no antagonism (Table 3-24). In the N-allyl nororipavines, this change from antagonism to agonism also occurs between methyl (335) and propyl (339). Both the N-CPM and N-allyl ethyl carbinols 337 and 344 are powerful antagonists in the rat tail
III
Diels-Alder
137
Adducts of Thebaine
Table 3-23 Analgesic
Activity
of N-Substituted
Thevinol Derivatives N_R1
"CH 3 II'C_R, , OH OCH3 Analgesia,"
332 333 334 335 336 337 338 339f 340 341
R'
R1
R
CH3 CH3 CH3 CH3 CHzCH3 CHzCH3 n-Propyl n-Propyl !sopentyl !sopentyl
CPM' CPM Allyl Allyl CPM CPM Allyl Allyl CPM CPM
CH3 H CH3 H CH3 H CH3 H CH3 H
Agonism
Tail Pressureb Antagonism OAJ 35
0.02' 2.2 0.2 70 0.75 20 3.8 48
Reference 226. bRats, sc. , CPM cyclopropylmethyl. a
= , d Relative to nalorphine Relative to morphine = 1.0. = 1.0. f Alletorphine.
pressure test, but the analogous n-propyl carbinols 342 and 345, derivatives of etorphine, are agonists. 5. In the C-19S cycIopropylmethyl series, when R' is methyl, the amount of agonist character increases with increasing size of t?e othe~ al~yl substituent, with the change from antagonism to agoms~ begmnmg with propyl or butyl. Here the transition is les~ dramatic: however, since the potency decreases with increasing ch~m length, m contrast to the diastereomeric series, where potency mcreases. . 6. While N-cycIopropylmethyl noretorphin.e (~42) has 1000 11~es morphine's analgesic potency, the opposite ~Iastereom~r ( 19 ) IS a morphine antagonist equivalent to nalorphme (rat tall ~res~ure). However, 342 has 3!lOths the antagonist activity of nalorphme 10 the tail flick assay.
;
Alletorphine (339) is 50-100 times more potent an analgesi~ th:n morphine in rats (sc, rat tail pressure test), showmg much lower resplrato y
138
3
Synthesis of Morphine,
Codeine,
and Related Alkaloids
Table 3-24 Analgesic
Activity
of N-CPM
and N-Allyl
I
III
Diels-Alder
139
Adducts of Thebaine
The 3-deoxy analogs, such as base 346 of N-substituted 6,14endoethenooripavine alcohols that have antagonist properties, such as 337, show weaker antagonism than the 3-hydroxy counterparts and, in most cases, show some weak analgesic activity. The 3-deoxy N-allyl and N-cycIopropylmethyl analogs (347, 348) of analgesic oripavine alcohols (339, 341) show decreased analgesic activity (226).
Orvinols
~H) -
'IIC
F
~R'
I
OH CH)
Analgesia, a Tail Pressureb
333 337 342 343 335 344 345
R'
R
CH3 CH2CH3 (CH2hCH3 (CH2hCH] CH3 CH2CH] (CH2hCH3
CPMc CPM CPM CPM Allyl Allyl Allyl
EDsn (mg/kg)
ADso (mg/kg) O.Ol3d 0.02
0.002' 0.026 0.21 4.2 0.033
Q
Reference 228.
R'
R
346
Et
CPM
347
n-propyl
Allyl
348
isopentyl
CPM
bRats, sc or ip.
C CPM = cyc1opropylmethyl. , EDso of morphine = -1.7
d Antagonism. mg/kg.
depression than either morphine or etorphine and giving a low acute tox~city (229). T~is se~aration of effects, however, is not found as clearly in a different species, mice (230). The antagonist component is not strong enough to prevent some dependence in monkeys. In humans, alletorphine has s~own ~ome success as an analgesic in postoperative abdominal surgery (eqUieffectlve at 1/20th the dose of morphine) and cancer patients (228,231). Two well-studied 6,14-endoethenotetrahydrooripavine N-cycIopropylmethyl analogs are alcohols with a methyl and an isoamyl substituent as R' (232). Codides and morphides having a large group at the C-19 position (e.g., n-butyl, isoamyl, cycIohexyl, phenethyl; see Table 3-19) demonstrate i?creasing analgesic potency, the bases being 50-500 times as potent as morphl~e. The m~thyl analog (333) antagonizes morphine's analgesic and depressive effects 10 the mouse, rat (analgesia antagonism is 35 times that of nalorphi.ne), and dog, without any evidence of agonism. The isoamyl analog .va~les f.rom 250 to 1000 times the potency of morphine as an anal¥e.slc I~ mice, rats, guinea pigs, and dogs by intraperitoneal (ip) adm1OIstratIOn but shows species variation in effects differing from those of morphine.
In the 6, 14-endoethano N-cycIopropylmethyl nororipavines, the primary alcohol (349) and the diastereomeric secondary methyl alcohols (350, 351) are powerful antagonists with low intrinsic activity (28,228). Diprenorphine (352) is not only a morphine antagonist in rats (tail pressure test) but has over 300 times the potency of nalorphine, with no antinociceptive activity in mice (tail flick test) (Table 3-25). It is marketed as Revivon, which is used for reversal in game animals of the immobilization induced by etorphine (316) (Immobilon). Diprenorphine in rats displays in vivo binding kinetics very different from those of pure agonists such as etorphine (224a). Although the cereberal receptor affinities of diprenorphine and etorphine are in the 0.1-0.2 nM range, diprenorphine has a long in vivo receptor halt-life for dissociation (approximately 18 minutes), being retained at several cereberal binding sites for many hours. In contrast, the half-life of etorphine is about 50 seconds. Antagonists are thus theorized to be more effective than agonists in displacing tritiated opiate receptor ligands in vivo. Increasing the size of the alcohol alkyl substituent in the CPM oripavines gives mixed agonist-antagonist activity, compounds such as 353 (n-propyl) and 354 (n-butyl) becoming powerful agonists instead of antagonists in the rat tail pressure test. The partial agonist buprenorphine (355) (Temgesic),
140
3
Synthesis of Morphine,
Codeine, and Related Alkaloids
III
Diels-Alder
Adducts of Thebaine
141
Table 3-25 Analgesic
Activity
of 6,14-EndOethana
N-CPM
Orvinols
R' IIC~~RII
\
OH
OCH3 356 Analgesia".b Rat Tail
R' 349 350 351 352' 353 354 3551
H H CH3 CH3 CH3 CH3 CH3
Pressure
Mouse
Tail
Flick
R" H CH3 H CH3 (CH2hCH3 (CHzhCH3 C(CH3h
0.025 0.011 0.02 0.008
>IOOd 0.008 0.041 0.024
O.OOY 5.6 1.7 0.22
b mg/kg, sc or ip. r Diprenorphine. 100 mg/kg. EDso morphine = 1.8 mg/kg; EDS
ADso morphine> 100 mg/kg; ADS
331
"d Reference 228. C
f
Buprenorphinc.
with a t-butyl group, has analgesic potency 25-75 times that of morphine and morphine antagonist potency 4 times that of nalorphine (233). It has a rapid onset, produces a weaker physical dependence than morphine in rodents (233), has a longer duration of action than morphine in the rat tail pressure test, and has known metabolic degradation pathways (234), making buprenorphine useful in certain physical conditions in humans (235) in whom it is also effective by sublingual administration. 3, Acid-Catalyzed Rearrangements to Unsaturated Ketones The alcohols of 6,14-endoetheno and 6,14-endoethano tetrahydrothebaines and oripavines are all unstable to acids and suffer dehydration and complex rearrangement, depending on the nature of the alcohol group, the 6,14bridge, and the reaction conditions. In summary, all C-19 alcohols (331) give phenolic ketones (356) as stable end products of acid-catalyzed rearrangement, the yields of the intermediate codeinones being very poor. These phenolic ketones are the major end products of the 6,14-endoethano series and 6,14-endoetheno phenyl derivatives. The 6,14-etheno alcohols of straight chain alkyls
R
= H, CH3
RI
other = H, CN, alkyl, H, alkyl, aryl R' = R" = X = C2H2, C2H4
R" 357 Scheme 3-42. Acid-catalyzed rearrangement products of C-19 alcohols. Reagent: (a) acid (e,g., formic, hydrochloric).
rearrange complet~ly to produce primarily cycIohexenodihydrocodeinones of type 357 (Scheme 3-42). The 6,14-ethano alcohols having alkyl substituents with chain branching adjacent to the alcoholic center rearrange in acid in a manner different from that of their straight chain counterparts (236). Brief heating of these alcohols in formic acid initially produces the olefinic side chain product 358 (Scheme 3-43) rather than a C-7 olefin, as confirmed by characteristic olefinic absorbances in the proton NMR spectrum (237). Diastereomeric alcohols at C-19 produce the same olefinic product, which is itself unstable in acidic media. The 6,14-endoetheno olefins of 358 rearrange with prolonged heating to 14-alkenyl codeinones (359), in accordance with their chemical properties and absorption spectra,
142
3
Synthesis of Morphine,
Codeine, and Related Alkaloids
331
358
\
359
X
360
X = C2H4
C2H2
Ie
368
362
R1
R =
Rt'
H; X = C2H2
R'
= CH3;
X = C2H4
Scheme 3-43. Acid-catalyzed rearrangement to 14-alkenykodeinones and a,,B-unsaturated ketones. Reagents: (a) formic acid; (b) hydrochloric acid, room temperature; (c) hydrochloric acid, heat; (d) sodium hydroxide. 2-ethoxyethanol; (e) hydrochloric acid; (f) zinc, acetic acid.
by protonation of the C-19 double bond followed by ring fission. The degree of substitution of the double bond determines the ease of this later rearrangement to the codeinones, with either refluxing formic acid or dilute mineral acid being effective reagents. Again, diastereomeric alcohols produce the same codeinone product. 14-Alkenyldihydrocodeinones
HI
Diels-Aldcr
Adducts of Thebaine
143
(360) are usually not easily produced from 6,14-endoethano alcohols due to further rearrangements in acidic media, but they have been obtained by catalytic or chemical (with sodium borohydride in pyridine) reduction of the 6, 14-endoetheno-14-alkenylcodeinones, for example, as in the case where the R"H2C group is phenyl (238). If 14-alkenylcodeinone bases (359) [or the precursor 6,14-endoetheno alcohols (331), the C-6 hydroxy alcohols (361), and the side chain olefins (358)] are heated with hydrochloric acid, further rearrangement occurs to C-4 phenolic O',,B-unsaturated ketone derivatives of structural type 356, isomeric with structure 359 and characterized by both proton NMR and UV spectroscopy (Scheme 3-43). The carbonyl (or its equivalent) at C-6 seems to be mandatory for this rearrangement, since it may activate the ether bridge. The 6-hydroxy base (361) is readily obtained with hydrochloric acid treatment at low temperature, but this can be subsequently converted to the same C-4 phenolic base (195,237). In cases where R' and R"H2C are different hydrocarbon chains, both cis and trans geometric isomers of 356 are obtained. These O',,B-unsaturated ketones can be reduced to the saturated ketones (362) with zinc and acetic acid. If the secondary and tertiary alcohols of 331 are first converted to the tosylates (207), and then either refluxed in xylene or treated with potassium t-butoxide in boiling [-butanol, C-7 olefins (363), or terminal olefins (358), both with an intact 4,5-oxo bridge result (Scheme 3-44). This method for forming 7-alkylidenes has been used for both 6,14-endoetheno and 6,14-endoethano tosylates when the two substituents on the C-19 alcohol are hydrogens and methyl groups (239). Treatment of 6,14endoetheno olefins (363) with perchloric acid leads directly to 0',13unsaturated ketones (356). The 1,I-dimethyl olefin (364, R" = R' = CH3) and the un substituted olefin, the 7-methylene base (364, RII = R' = H), have been prepared by this route. The C-4 phenolic O',,B-unsaturated ketones can also be obtained from the grignard reduction product of ketones (e.g., 282) by treatment with cold hydrochloric acid, these being milder conditions than those for the dehydration and hydrolysis of alcohols (331) to l4-alkenylcodeinones (Scheme 3-45) (201). Both the C-4 phenol and methyl esters of 356 have thus been prepared using this route. Reaction of base-catalyzed rearrangement products such as 205 also provides a route to C-4 phenolic O',,B-unsaturated ketone products. Reaction with methyl lithium to give an intermediate C-19 alcohol followed by hydrolysis with cold dilute hydrochloric acid to a C-4 phenolic ketone (365) and then with hot hydrochloric acid yields the olefinic ketone 356 (Scheme 3-45). The product is identical to that formed by acidcatalyzed rearrangement of l4-alkenylcodeine derivative 359 or of hydrolysis of the alcohol grignard reduction product 282.
144
3
Synthesis of Morphine, Codeine, and Related Alkaloids
a or
b
)
+
III
Diels-Alder
Adducts of Thebaine
358
: ,R I
Tsa~
y
C' \R"
282 R2
331 R'
145
Rl
=
R
=
R'
CH];
R"
=
H
363
R" = H, CH3
356 b,a
R = R2 = H, CH3 Rl H, CN, alkyl,
othel
R' = R" - H, alkyl,
ary
y 205
356 364
365
R = Rl = R' - CH3 R"
R'
H or CH]
R" R' Scheme 3-44. a,,8-Unsaturated ketones from 7-alkylidenes. Reagents: (a) refluxing xylene; (b) potassium t-butoxide, butanol; (c) perchloric acid, heat.
In the rearrangement of most I4-alkenylcodeinones (359), the 5,14bridged C-3 phenols and methyl ethers of type 356 actually become the minor products of rearrangement and a competitive process produces the major component as a nonphenolic, nonconjugated ketone. This stable end product can also be obtained directly from the 6,I4-endoetheno C-I9 alcohols of type 331 when rearrangement is pushed to completion and is the result of either stereoisomeric alcohol, due to an intermediate C-? carbonium ion formation. Thorough proton NMR studies and chemical transformation studies have determined the product to be cyclohexeno[1' ,2':8,I4]dihydrocodeinones 357a and 357b formed by protonation of the
Scheme 3-45. a,,8-Unsaturated ketones from C-19 alcohols. Reagents: (a) hydrochloric acid, room temperature; (b) alkyl lithium; (c) hydrochloric acid, heat.
enone system of 366 [the proto tropic equilibrium product of 14alkenylcodeinone (359)] and collapse of the carbonium ion intermediate (367) (the product of addition of the C-8 carbonium ion to the side chain) in one of two ways, depending on the nature of the substituents in the unsaturated ring (Scheme 3-46) (240,241). The structure of the rearranged product has been subsequently confirmed by X-ray studies (242). In cases where R' and R" are alkyl groups, the ~5' isomer (357b) is usually favored over the ~4' (357a). This rearrangement process is not operable in 6,14-endoethano analogs due to the need for the enone system in the intermediates. 6,I4-Endoetheno bases that cannot isomerize to olefins (366), such as when R"H2C is replaced by phenyl, give rise only to the 5,14-bridged phenols of type 356 (237,240).
146
3
Synthesis of Morphine,
Codeine, and Related Alkaloids
a
359
366
367
r
III
Diels-Alder
Adducts of Thebaine
147
The biological activity of the two primary types of products resulting from acid-catalyzed rearrangement of C-19 alcohols, namely, bases 356 and 357, has generally been disappointing. Although this transformation would be expected to generate an analgesic, it produces opiate derivatives that have no analgesic effects but that do have central antidepressant (243) as well as other central effects (244). The dehydration and rearrangement of the C-19 alcohols to the 14-alkenylcodeinones (359) results in a substantial decrease in analgesic potency; however, some phenolic a,{3unsaturated ketones (356) do have moderate potency. When R' is methyl, I R" is hydrogen, and Rand R are methyl, the base 368 (Scheme 3-43) has 2 times the potency of morphine and its C-4 methyl ether has 15 times its potency (237). The cyclohexenodihydrocodeinones and morphinones have decreased analgesic or antagonist activity compared to the C-19 alcohol precursors. The cyclohexenodihydromorphinones are interesting, though, in that they represent the only N-methyl series of compounds in which morphine antagonism has been demonstrated. The 6,5' morphinone isomer (369) (R = H) has l/lOth the potency of nalorphine; the codeinone (R = CH3) is weaker (191). 14-AlkenyIcodeinones (359) and their dihydro derivatives (360) can be reduced at the C-6 carbonyl group with sodium borohydride to give the 6-hydroxy analogs (237). Reduction of the codeinones gives exclusively 6a-hydroxy products, whereas dihydrocodeinones give mixtures of the isomeric 6a and 6{3alcohols (dihydrocodeines and dihydroisocodeines), the ratio being solvent dependent (238). Reduction of the carbonyl group in the cyclohexenodihydrocodeinone and dihydromorphinone derivatives of 357 produces the 6a-hydroxy products (240).
4. Modifications and Derivatization In an effort to increase the analgesic potency of the C-19 alcohols of the 6, 14-endoethenotetrahydrothebaines
357b
357a 369
R1
= R" = R' = CH3;
R = H, CH3 Scheme ].46. Synthesis of cyc1ohexenodihydrocodeinones (a) formic or hydrochloric acid.
and morphinones.
compounds Reagent:
Base~ of type 356, ~62, and 357 with substituents other than methyl on t~e tertiary nItrogen (I.e., H, CN, other alkyls) are prepared in moderate Yield by rearrangement of the N-substituted alcohols of type 331. The 3-hydroxy analogs of each can be obtained by demethylation of the 3:methoxy rearrangement products using hydrobromic acid or by coinbilled rearrangement and demethylation of the C-19 alcohol thebaine adducts with the same reagent (237,240).
and oripavines, modification of the aromatic ring has
been studied. Addition of substituents to the aromatic A ring of dimethyl alcohol (281) causes either a loss or a reduction of activity in the resulting compared
to their unsubstituted
analogs. The phenol and
methyl ethers of the chlorinated derivative 370 are less active than the unsubstituted precursor; the phenolic amines, derivatives 371, also have greatly reduced analgesic activity (245). Modification of the piperidine ring by substitution at the carbon adjacent to the nitrogen (C-16) has also been achieved. The 15,16-unsaturated 6,14-endoethenotetrahydrothebaine and oripavine precursors result from dehydration with mercury (II) acetate (246). N-substituted analogs can also be dehydrogenated to enamines if they do not contain a reactive olefin; the 6,14-endoetheno bridge always remains inert. The enamines can be reduced back to the parent bases with sodium borohydride, providing a
]
148
Synthesis of Morphine,
Codeine, and Related Alkaloids
convenient method for selective C-15 labeling with tritium or C-15, C-16 radiolabeling (246,247). Treatment of the enamines with acid yields iminium salts (i.e. perchlorates), which, in additi~n to the 15,16unsaturated compounds, then serve as intermediates for C-l5, C-l6 substituted derivatives (248). The 16-alkyl and 16-aryl derivatives of analgesics in the 6,14-endoethenotetrahydrothebaine series have been prepared by reaction of these iminium perchlorates with grignard reagents or lithium alkyls (249). Other derivatives of both C-15 and C-16 6,14saturated and unsaturated bridge compounds have been made by these methods. The C-15-C-16 modification has not been successful in producing analgesics. 15,16-Unsaturated carbinols, such as 372 and 373, are less
"CH]
y
"'C - CH] 'OH
OCH] 370
R
371
R
H,
=
CH];
H; X
=
Y
H;
H; Y
=
X
=
Cl
CH2Y' where
R1
R2
H
.:CH3 IPC --R' \'H CH3
372 373
n-butyl Cyclohexyl
R' R'
R2
S;H]
R2
'IIC_R' bH CH3
=
374
R'
375
R'
CH)
CH2C6H5 Cyclohexyl
III
Diels-Alder
Adducts of Thebaine
149
potent analgesics than the parent saturated bases in the rat tail pressure test. Compound 373 has only about 1/4000th the analgesic potency of its parent, although it has potent antitussive activity (246). Unsaturated analogs with substituents other than hydrogen at C-l6 also show little analgesic activity. The 6, 14-endoethano and 6,14-endoethenotetrahydrothebaines bearing C-15 and/or C-l6 substituents have all shown little activity as analgesics, although substitution at C-16 with an alkyl group has led to potent antitussives. 16a-Phenyl, benzyl, phenethyl, and alkyl other than methyl in the etheno series completely eliminate analgesic activity in C-19 alkyl and aryl carbinols. A 16a-methyl substituent retains marginal activity in some of the more potent carbinols, for example the benzyl carbinol (374), which is 750 times less active than the unsubstituted base but is only 4-5 times less active than morphine (249). The 16-methyl cyclohexyl carbinol (375) has no analgesic activity but has 12 times the potency of codeine as an antitussive (250). Increasing the chain length at C-16 to ethyl decreases antitussive activity by a factor of 3. Continued increase in the alkyl chain shows further reduction, with the 16-phenyl analog being inactive. Various analogs in the 7,B-carbinol series have been prepared to check analgesic potency in comparison to the 7a-equivalents (241). For example, starting with 7-methyl-epi-nepenthone derivatives 376 and 377, the tertiary alcohols (378-380) result from treatment with the appropriate grignard reagent where the chiral products have the S-configuration at C-l9 (Scheme 3-47). The C-19R secondary alcohols 381 and 382 result from treatment with sodium borohydride. The analgesic potencies of these C-19 carbinols vary, showing a strong dependence on the C-19 configuration. The secondary alcohols 381 and 382 have 40 and 6 times the potency of morphine in the rat tail pressure test, respectively, but the tertiary alcohol 378 has only one-third morphine's potency. The N-CPM methyl alcohols, 380 and 382, have no antagonist activity. A series of related C-7 tertiary alcohols where the hydrocarbon chain is directly attached to C-7 have been prepared by appropriate grignard reaction with the 7-oxotetrahydrothebaine (191) (Scheme 3-48) (251). Attack from the a-face is preferred on steric grounds, approach of reagent from the ,B-face being hindered by the C-5 and C-15 ,B-hydrogens. The isolated products have an H-5,B proton signal at ~4.5 ppm, attributable to the 1,3 deshielding by the 7,B-hydroxyl group, thus allowing assignment of their structures. In those compounds where the hydroxyl is directly attached to C-7 instead of C-19, the effects on analgesic potency observed in a homologous series directly parallel those in the C-19 carbinols, although relative comparative potency is significantly diminished. Derivatives of 383 are less
3
150
Synthesisof Morphine, Codeine, and Related Alkaloids
III
Diels-Alder
Adducts of Thebaine
151
a
CH3 R
R'
378
CH]
379
CH]
380
CPM
191
383
R'
CH3,
384
R'
C6HS'
n
0-4
384c R'
C6HS'
n
2
C6HS'
n
3
384d R' Scheme
376
R
=
CH3
377
R
=
CPM
3-48.
Synthesis
of
7a-substituted-7j3-alcohols.
Reagent:
(a)
grignard
n = 0-2
reagent.
:CH3}6HS
'-C'-H \OH OCH3
Scheme 3-47.
Synthesis
of 7j3-carbinols. Reagents:
381
R
382
R
= =
(A)
(B)
R configuration
S configuration
Fig. 3-2. CH] CPM
(a) grignard reagent; (b) sodium
borohydride.
potent than morphine; the alcohols of 384 show a lO-fold increase in analgesic potency from phenyl to benzyl or phenethyl. The phenethyl and phenpropyl analogs (384c, 384d) are the most active, being about three times as potent as morphine. 5. Stereochemical Assignments Proton NMR spectroscopy, including the special techniques mentioned earlier for obs.erving spin systems and coupling constants, has definitively charactenzed th~ 7a.- and 7f3carbinols of Diels-Alder adducts of thebaine and their aCid-catalyzed rearrangement products. The epimeric C-7a compounds have NMR spectra analogous to those described earlier for C-? ketones and esters (Section III,A,5, Fig. 3-1).
NMR and IR spectroscopies have also provided a basis for assigning the configuration at the C-19 asymmetric center (219). The physical methodology results are consistent with the experimental ones, showing that the synthesis of C-19 alcohols is stereoselective. For the two epimers where the substituents are phenyl and methyl (A and B, Fig. 3-2), the tertiary hydroxyl resonances at 5.26-5.8 ppm (CDCI]) indicate intramolecular hydrogen bonding, also confirmed by the IR spectra, thereby placing the tertiary hydroxyl near the C-6 methoxyl group. Both diastereomers show restricted rotation about the C-19-C-? bond. In A the olefinic protons show marked upfield shifts, indicating the proximity of phenyl to the endo bridge. This results in a shielding effect (this effect being stronger for the closer C-1S than for the C-1? proton) and characterizes the R epimer. By comparison, then, B has the S configuration at C-19. The H-5f3 chemical shift is consistent with earlier work on C-? ketones, indicating
3
152
Synthesis of Morphine,
Codeine, and Related Alkaloids
(B)
(A)
Fig. 3-3.
that the compounds are 70: epimers, the configuration at C-7 remaining unchanged through synthesis. The proton NMR spectra for the acid-catalyzed rearrangement products dihydrocyclohexenocodeinones show the C-5 proton characteristic of the cyclic ether bridge at -4.5 ppm. No olefinic endo bridge proton signals are apparent. The isomers are easily distinguished when R' is phenyl and R" is hydrogen in compounds 357a and 357b (A and B, Fig. 3-3). Even though both contain the styrenoid system and have styrenoid ultraviolet absorption, spin-decoupling techniques indicate that one of the methylene protons of A adjacent to the double bond at position 3' has a down field shift characteristic of being allylic and of being located near the unshared electron pair on nitrogen (240). When both R' and R" groups are alkyl, the lack of an olefinic proton signal characterizes the 6.5' isomer (B). Further NMR experiments on the carbinols has confirmed that the difference in H-80: and H-8,8 chemical shifts is partly due to effects of the ring nitrogen, although both C-8 protons have their chemical shifts affected (upfield shifts) by the C-17-C-18 double bond. H-9,8 and N-methyl protons consistently appear downfield. The 6,14-endoethano carbinols have provided evidence for the double bond deshielding of H-9,8, since the saturated analogs consistently have H-9,8 0.5 ppm upfield from its average chemical shift at 3.13 ::t 0.04 ppm (219). Various competing effects, however, have made a complete analysis of the geminal80: and 8,8 protons in some systems impossible, even though the effects produced by the nearby tertiary nitrogen and C-17-C-18 double bond are consistent (219). The entire molecular structure of one Diels-Alder adduct, etorphine, has been elucidated through X-ray crystallographic analysis of the hydrobromide salt (252). This comprehensive study of the subtle details of geometry has supplemented the NMR spectroscopic studies.
C. Opiate Receptor Probes Several Diels-Alder adducts of thebaine and oripavine have been used as opiate receptor probes (253) in the hopeof defining a three-dimensional
III
Diels-Alder
Adduets of Thebaine
153
receptor model. The extremely potent analgesic activities of the C-7 substituted alcohols of the 6,14-endoetheno and 6,14-endoethano DielsAlder adducts of thebaine, influenced by the size, nature, and configuration of ~he. alco.holic portion, suggested to Bentley and Hardy (222) that Beckett s sImplIfied model (254) of the morphine receptor (three points: anionic, planar, and cavity) might be extended (255) to include additional significant points of attachment by these bases. This fourth binding site for the group at C-7 would have configurational selectivity for the C-19 alcohols. The first sugg.estion by Bentley and Lewis that the alcoholic hydroxyl group ~as reqUIred as a. specific binding site was disproved by the high analgesIc potency (1000 tImes that of morphine) found in the cyclohexanoderivative (385). Subsequently, the presence of a lipophilic site on the receptor that could mediate analgesic effects by interaction with the alkyl and phenyl groups on C-19 (256) and that could discriminate between diasteromers was proposed. In support of this idea, the phenethyl carbinol (302) (PET, analgesic potency 500 times that of morphine; Table 3-19) supposedly owed its increased activity to the presence of the additional phenyl g~ou~ (257). In addition, destruction of the aromatic A ring by ozono~ysls dId not. produce a complete loss of activity in the resulting lactomc esters, whIch cannot adopt a topological arrangement isosteric with the mo~phine aromatic nucleus (257) and yet have an analgesic potency eqUIvalent to that of morphine. Looking at a series of 6demethoxy thevinols and orvinols, Rapoport et ai. (258) demonstrated that th~ hydrogen bond between the C-6 methoxyl and the alcoholic hydroxyl (FIg. 3-2), as supported by quantum mechanical calculations, NMR spectro~c?py, and X-~ay. crystallography, was not essential for analgesic actIvIty. The specIficity for the C-19R absolute configuration, however, sug~ested to them a secondary binding site that had a high discriminatory aptitude in the lipophilic region of the receptor. On the basis of this modified model for the opiate receptor, Michne et ai. (259,260) investigated a series of 2,6-methano-3-benzazocine-ll-propanols (386), resembling various agonists and agonist-antagonists, that should have been capable of the proposed four-point receptor interaction, since t?e compounds have an appropriately substituted nucleus and an asymmetncally substituted alcohol approximating that of the Diels-Alder adduct alcoho.ls. In these compounds, the classical structure-activity profile for na.rcotlc agonist and antagonist activities, as seen in bridged thebaine and ?npavine derivatives, has not been upheld. Interestingly, though, the two Isomers of 387 having an isoamyl alcohol substituent at CAll are three and five times as potent as nalorphine as antagonists (261). In Diels-Alder adducts of a simpler bicyclodecane skeleton (388, 389), Crabb and
3 Synthesis of Morphine, Codeine, and Related Alkaloids
154
IV
The Chemica! Anatomy
of Morphine
and Its Derivatives
..CH..
HO phenyl. thienyl +---~ ,/ furyl "-
CO. CHOH
X ::~H. D
I N
+ ___ __ __+ un.;:. ,,, C H OrCOR'" -- -0
r-' \' CH3
HO
387
386 t-Bu, n-Bu, Me, Et, n-Pr, Me, CPM, R3 = H, Me; Me, Et i-Am; R2 = R4 CH3, R3
Rl
2 = R =
H, C 0Me, C02Et,
A.
CN
The Chemical Anatomy of Morphine and Its Derivatives The Chemical Anatomy of Morphine
Natural morphine is a levorotatory molecule containing five rings. On the basis of the SAR results presented in Section II,A, the following generalizations can be made (see Fig. 3-4): 1. Enantiomeric dextrorotatory morphine is devoid of analgesic activity. 2. A trans-ring junction between rings Band C drastically reduces analgesic potency.
is adapted from
3. A phenolic hydroxyl at C-3 is important for analgesia, but not necessary. The 3,6-deoxy-derivative is equipotent to morphine but may represent metabolic hydroxylation. 4. Other substituents on the A-ring decrease or eliminate activity. 5. Substituents at C-lO in the B-ring maintain or decrease morphinomimetic activity. 6. Some of the chemical features of the Coring of morphine are relatively noncrucial: removal of the double bond [7(8)] or the alcohol at C-6. Numerous other chemical modifications are also compatible with enhanced analgesic activity: a C-5 methyl group, oxygenation or alkylation at C-6, phenylalkyl substitution at 7f3, and short alkyl substitution at 8f3. '\ 7. Hydroxyl substitution at the BC-ring junction (C-14), as well as its acyl derivatives, strongly enhances analgesia. Other substituents eliminate activity. 8. The methyl group on the amine is not optimal. Replacement by aralalkyl or functionalized aralalkyl groups increases morphinomimetic activity severalfold. Replacement by N-lower alkyl (propyl, allyl, cyclopropylmethylene) produces morphine antagonists. 9. Addition of a sixth ring by Diels-Alder additions to thebaine can increase either morphine agonist or antagonist activity by at least four orders of magnitude.
H
Wilkinson (262) demonstrated for most compounds studied a loss of analgesic potency accompanied by toxic side effects. This led them to question earlier hypotheses and to believe that the aromatic ring of the bridged thebaine compounds is actually a structural requirement.
IV.
H. OH, O,CR. = 0, alkyl, OCH, at C-6
Fig. 3-4. Potency-enhancing substituents on morphine. (This representation that reproduced in ref. 30.)
389
388
H,OHorO,CRatC-14 ;:. 7 (8). ~- CH, Or CH. CH, at C- 8 6 (71 or {J- (CH.)n C.H.
, CH, at C-5
385 R1 R2 R4 R1
0",',
J
R1
155
B.
The Chemical Anatomy of Diels-Alder Adducts
Diels-Alder adducts of thebaine and their analogs, having greater complexity and rigidity than morphine, as well as a different shape, have
3
156
Synthesis of Morphine, alkyl. substituted
and Relattd
Alkaloids
References
alkyl. alkenyl
H or COR. R = alkyl Potency-enhancing
-+ n-
substituents
on the Diels-Alder
157
6. The piperidino nitrogen should be tertiary, the secondary amines being less active as agonists. Increasing the size of the nitrogen substituent beyond methyl steadily decreases analgesic activity. Some N-allyl and N-cyclopropylmethyl thebaine and oripavine derivatives of tertiary alcohols are potent antagonists. 7. The piperidine and ether rings should remain intact. Substitution on the piperidine ring near the basic nitrogen creates potent antitussives.
or IJ. CH(OH)R'; tr- or fJ-C(OH)R'R" where R'=CHJ and OCHJ R"=(CH2)nCHJ or (CH2)nAr R configuration preferred
7
Fig. 3-5.
Codeine,
References
adducts.
been used in attempts to test receptor criteria for analgesic activity and the separation of desired from undesired biological effects in an effort to create a superior morphine. These Coring-bridged compounds have exhibited a wide range of agonist -antagonist profiles (228,229); however, the only two modifications creating dramatic increases in biological activity are (a) demethylation at C-3 to the oripavine base and (b) substitution at C-7 with secondary and tertiary alcohols. SAR in the carbinol series is complex, but the physiological activity of each base is dependent on the nature of the substituents on the C-3 oxygen atom, the nature and size of the alcoholic group, and the substituent on the nitrogen atom, with the following trends being observed (191,263) for these structures (Fig. 3-5): 1. The substitution on the carbinol group at C-7 is important: C-19 tertiary alcohols have a higher analgesic potency than secondary ones; however, in cases where both C-7a and C-7{3 epimers have been evaluated, only slight differences in potency have been observed in the pairs. 2. Highest activities have been found when a moderate disparity in size between the two groups on the C-19 alcohol, R" and R', exists. One substituent should be small, that is, H or CH3. Maximal analgesic activity is then found when the second substituent is a 3- to 5hydrocarbon chain; further lengthening results in a steady decrease in activity. 3. The diastereomers of unsymmetrical tertiary alcohols can show markedly different analgesic properties, the 19R configuration being the more potent morphine agonist. 4. The comparable 6, 14-endoetheno and 6,14-endoethano analogs show only marginal differences in analgesic agonist or antagonist potency. 5. The oripavine derivatives (C-3 hydroxyl) or their C-3 acetylated analogs are more potent analgesics than the thebaine bases (C-3 methoxy).
fl.t.of1e.,; f'
1. J. M. Gulland and R. Robinson, Mern. Proc. Manchester Lit. Phil. Soc. 69, 79 (1925). 4'<2. M. Gates and G. Tschudi, J. Arn. Chern.Soc. 78, 1380(1956); M. Gates and G. Tschudi, J. Arn. Chern. Soc. 74, 1109 (1952). ¥ 3. D. Elad and D. Ginsberg, J. Arn. Chern. Soc. 76,312 (1954); D. Elad and D. Ginsberg, J. Chern. Soc. p. 3052 (1954).
*
4. D. Ginsberg,
"The
Opium
Alkaloids,"
p. 55. Wiley (Interscience),
New York,
1962.
rJ;.J..f.1"tt,! 5. M. A. Schwartz and I. S. Mami, J. Arn. Chern. Soc. 97, 1239 (1975). Mor'f;";'1( 6. C. Szantay, G. Blask6, M. Barczai-Beke, P. Pechy, and G. Dornyei, Tetrahedron LeU. 21, 3509 (1980). 7. T. Kametani, M. Ihara, K. Fukumoto, and H. Yagi, J. Chern. Soc. C p. 2030 (1969). 8. D. H. R. Barton, G. W. Kirby, W. Steglich, and G. M. Thomas, Proc. Chern. Soc., London p. 203 (1963); D. H. R. Barton, D. S. Bhakuni, R. James, and G. W. Kirby,]. Chern. Soc. C p. 128 (1967). 'i" (2) A. Grewe and A. Mondon, Chern. Ber. 81,279 (1948); R. Grewe, Naturwissenschaften 33, 333 (1946).
10. K. C. Rice, Proceedings of the Committee on Problems of Drug Dependence, NIDA * Res. Monogr. No. 41, p. 99 (1981). 11. T. S. Lie, L. Maat, and H. C. Beyerman, Reci. Trav. Chirn. Pays-Bas 98,419 (1979); H. C. Beyerman, J. van Berkel, T. S. Lie, L. Maat, J. C. M. WesselS;-H. H. Bosman, E. Buurman, E. J. M. Bijsterveld, and H. J. M. Sinnige, Reci. Trav. Chirn. Pays-Bas 97, 127 (1978); H. C. Beyerman, T. S. Lie, L. Maat, H. H. Bosman, E. Buurman, E. J. M. Bijsterveld, and H. J. M. Sinnige, Reel. Trav. Chirn. Pays-Bas 95, 24 (1976). 12. W. J. Musliner and J. W. Gates, J. Arn. Chern. Soc. 88,4271 (1966); ibid. Org. Synth. 51, 82 (~971. ~g;::5'r. -/,,~(S>''''+1o,)\C~IAJIc./.; , @
~
'f.-@
K. C. RIce, J. Or Chern. 45 3135 (1980). K. C. Rice and A. Brossi, J. Org. ern. 45, 592 (1980). 7'
'y.,-~J~-a.o-.3d:~ P I5- ~
'*~
Heterocycles 6, 1157 (1977).
K. C. Rice, J. Med. Chern. 20, 164(1977); K. C. Rice, in "The Chemistry and Biology
_
J'
of Isoquinoline Alkaloids" (1. D. Phillipson, cd.), p. 191. Springer Verlag. Berlin and New York. otL.11J A. Brossi, Proc. Asian Syrnp. Med. Plants Spices, 4th 1,261 (1980); Chern Abstr. 95,98084 (1981). . D. A. Evans, C. H. Mitch, and R. C. Thomas, J. Am. Chern. Soc. 102, 5955 (1980). 19. W. H. Moos, R. D. Gless, and H. Rapoport, J. Org. Chern. 47, 1831 (1982). 20. D. A. Evans and C. H. Mitch, Tetrahedron LeU. 23, 285 (1982). 21. W. H. Moos, R. D. Gless, and H. Rapoport, J. Org. Chern. Soc. 48, 227 (1983).
~~
A
11. Iv 11.\ $ 70/11
~
at. A't')' I p .~""
J
158
3
References
Synthesis of Morphine, Codeine, and Related Alkaloids
52. 53. 54. 55. .,.~
A. Meijere, Angew. Chem. Int. Ed. Engl. 18, 809 (1979). E. J. Corey and M. Chaykovsky, J. Am. Chem. Soc. 87, 1353 (1965). L. J. Sargent and A. E. Jacobsen, J. Med. Chem. IS, 843 (1972). M. P. Kotick, J. Med. Chem. 24, 722 (1981). M. A. Abdel.Rahman, H. W. Elliott, R. Binks, W. Kung, and H. Rapoport, J. Med. Chem. 9, 1 (1965).~G.d'''-H..'fI.." .t1"y,t M.'"r...~ (,O/< f.t.-.ey , ,., 1.. 57. S. Makleit, T. Mile, and R. Bognar, Mag. Kem. Foly. 81,564 (1975); Chem. Abstr. 84, 90360 (1976). 58. H. J. C. Yeh, R. S. Wilson, W. A. Klee, and A. E. Jacobson, J. Pharm. Sci. 65,902 (1976). 59. R. Bognar and S. Makleit, Acta Chim. Acad. Sci. Hung. 59, 373 (1969). 60. J. Knoll, Pharmacol. Res. Commun. 5, 175 (1973); R. C. Hill, D. Roemer, and H. Buescher, J. Pharmacol. Exp. Ther. 201, 580 (1977). 61. S. C. Clarke, D. R. Jasinski, J. S. Pevnick, and J. D. Griffith, Cli/I. Pharmacol. Ther. 19, 295 (1976). 62. J. Knoll, D. Furst, and K. Keleman, J. Pharm. Pharmacal. 25, 929 (1973). 63. J. Knoll, S. Furst, and S. Makleit, J. Pharm. Pharmacal. 27, 99 (1975). 64. H. Rapoport, R. Nauman, E. R. Bissell, and R. M. Bouner, J. Org. Chem. 15, 1103 (1956). 65. E. F. Hahn, J. Fishman, and R. D. Heilman, J. Med. Chem. 18,259 (1975). 66a. G. W. Pasternak and E. F. Hahn, J. Med. Chem. 23, 674 (1980). 66b. V. M. Kolb, Life Sci. 33 (Suppl. 1), 419 (1983). 67. E. F. Hahn, M. Carroll:Buatti. and G. W. Pasternak, J. Neurosci. 2, 572 (1982). 68. J. B. Jiang, R. N. Hansen, P. S. Portoghese, and A. E. Takemori, J. Med. Chem. 20, 1100 (1977). 69. L. M. Sayre and P. S. Portoghese, J. Org. Chern. 45, 3366 (1980). 70. R. C. Crouch, A. V. Bhatia, and O. W. Lever, Jr., Tetrahedron LeU. 24,4801 (1983). 71. P. S. Portoghese, G. Ronsisvalle, D. L. Larson, C. B. Yim, L. M. Syre, and A. E. Takemori, Life Sci. 31, 1283 (1982). 72. B. R. Baker, "Design of Active Site Directed Irreversible Enzyme Inhibitors." Wiley, New York. 1967. 73. P. S. Portoghese and A. E. Takemori, Spec. Publ.-R. Soc. Chem. 42,181 (1982); A. E. Takemori and P. S. Portoghese, Ann. Rev. Pharmacol. Toxicol. 25, 193 (1985). 74. T. P. Caruso, A. E. Takemori, D. L. Larsen, and P. S. Portoghese, Science 204,316 (1979); P. S. Portoghese, D. L. Larsen, J. B. Jiang, A. E. Takemori, and T. P. Caruso, J. Med. Chem. 21, 598 (1978); L. M. Sayre, A. E. Takemori, and P. S. Portoghese, J. Med. Chem. 26, 503 (1983). 75. T. P. Caruso, D. L. Larsen, P. S. Portoghese, and E. A. Takemori, J. Pharmacol. Exp. Ther. 213, 539 (1980).
22. H. Dreser, DISch. Med. Wochenschr. 24, 185 (1898); C. R. Wright, J. Chern. Soc. 27, 1031 (1874). .
~
23. L. F. Small, N. B. Eddy, E. Mosettig, and C. K. Himmelsbach, "Studies in Drug 24.
25. 26. 27. 28. 29.
30. 31. 32. 33. 34.
35.
36. 37. 38. 39. 40.
Addiction," Pub. Health Rep. Suppl. No. 138. U.S. Gov!. Printing Orfiee, Washington, D. c., 1938."':;>6- O~ ~"'Ilf.gi.ti.... H. L. Holmes, "Structure Activity Relationships for Some Conjugated Heteroenoid Compounds, Catechol Monoethers and Morphine Alkaloids," Vols. I and 2. Defense Research Establishment Suffield, Ralston, Alberta, Canada, 1975. A. H. Beckett and A. F. Casy, Prog. Med. Chem. 2, 43 (1962). A. H. Beckett and A. F. Casy, Prog. Med. Chem. 4, 171 (1965). A. F. Casy, Prog. Med. Chem. 7, 229 (1970). A. F. Casy, Prog. Drug Res. 22, 149 (1978). M. R. Johnson and G. M. Milne, in "Medicinal Chemistry," 4th cd. (A. Burger, ed.), p. 699. Wiley, New York, 1980. P. A. Janssen and C. A. M. Van der Eycken, Drugs Affect. Celli. Nerv. Syst. 2, 25 (1968). D. Lednicer and L. Mitseher, "The Organic Chemistry of Drug Synthesis," p. 286. Wiley, New York, 1977. D. Lednicer and L. Mitseher, "The Organic Chemistry of Drug Synthesis," p. 314. Wiley, New York, 1980. K. Goto, "Sinomenine, an Optical Antipode of Morphine Alkaloids." Kitasato Institute, Tokyo, 1964. I. lijima, J. Minamikawa, A. E. Jacobson, A. Brossi, and K. C. Rice, J. Org. Chem. 43, 1462 (1978); I. lijima, K. C. Rice, and J. V. Silverton, Heterocycles 6, 1157 (1977). K. C. Rice, I. Iijima, W. A. Klee, and Y. F. Jacquet, Proc. Commi. Drug Addict. p. 204 (1977); Y. F. Jacquet, W. A. Klee, K. C. Rice, I. Iijima, and J. Minamikawa, Science 198, 842 (1977). M. Takeda, H. Inoue, and H. Kugita, Tetrahedron 25, 1839 (1969). H. Kugita, M. Takeda, and H. Inoue, Tetrahedron 25, 1851 (1969). H. Kugita, M. Takeda, and H. Inoue, J. Med. Chem. 13, 973 (1970). H. Inoue, M. Takeda, and H. Kugita, Chem. Pharm. Bull 21, 2004 (1973). J. Reden, M. F. Reich, K. C. Rice, A. E. Jacobson, A. Brossi, R. A. Streaty, and W. A. Klee, J. Med. Chem. 22,256 (1979); F.-L. Hsu, K. C. Rice, and A. Brossi, Helv. Chim. ACla 63, 2042 (1980).
41. E. L. Way and T. K. Alder, Pharmacal. Rev. 12, 283 (1960); T. K. Alder, J. Pharmacol. Exp. Ther. 140, 155 (1963); T. Johannesson Toxicol. 20, 165 (1963).
and J. Schon, Acta Pharmacol.
42. E. Mohacsi, W. Leimgruber, and H. Baruth, J. Med. Chem. 25, 1264 (1982); P. H. '''''':;i'""'- Duquette, R. R. Erickson, and J. L. Holtzman, J. Med. Chem. 26, 1343 (1983).
,
~
3. E. L. May and A. E. Jacobson, J. Pharm. Sci. 66, 285 (1979). 44. L. H. Welsh, J. Org. Chem. 19, 1409 (1954). '45. E. L. Way, J. W. Kemp, J. M. Young, and D. R. Grasetti, J. Pharmacol. Exp. Ther. ~~ hf'\. 129, 144 (1960). ,.,"'.... 46. H. Wieland and E. Koralek, Justus Liebigs Ann. Chem. 433, 269 (1923). " 47. N. B. Eddy and L. E. Lee, J. Pharmacol. Exp. Ther. 125, 116 (1959). 48. K. Uba, N. Miyata, K. Watanabe, and M. Hirobe, Chem. Pharm. Bull. 28,356 (1980). 49. K. Uba, N. Miyata, K. Watanabe, and M. Hirobe, Chem. Pharm. Bull. 27,2257 (1979). S. Y. Yeh, H. A. Krebs, and C. W. Gorodetzky, J. Pharm. Sci. 68, 133 (1979). ,~~. 50. 51. N. Miyata, K. Uba, K. Watanabe, and M. Hirobe, Chem. Pharm. Bull. 28,3722 (1980); P. J. Thorpe, Drugs Flit. 8, 793 (1983).
76. P. E. Gilbert and W. R. Martin, J. Pharmacol. Exp. Ther. 198, 66 (1976); A. E. Takemori, D. L. Larsen, and P. S. Portoghese, Eur. J. Pharmacol. 70, 445 (1981). 77. P. S. Portoghese, D. L. Larsen, L. M. Sayre, D. S. Fries, and A. E. Takemori, J. Med. Chem. 23, 233 (1980). 78. L. M. Sayre, D. L. Larsen, D. S. Fries, A. E. Takemori, and P. S. Portoghese,J. Med. Chem. 26, 1229 (1983); L. M. Sayee, D. L. Larson, A. E. Takemori, and P. S. Portoghese, J. Med. Chem. 27, 1325 (1984). 79. L. A. Woods, Pharmacal. Rev. 8, 175 (1956). yO. H. A. Hageman, Org. Reacl. 7, 198 (1953). us\. J. D. Hobson and J. G. McCluskey, J. Chem. Soc. p. 2015 (1967); G. A. Brine, K. G.
~
[
159
N
1 J4"'<
Boldt, C. K. Hart, and F. I. Carroll, Org. Prep. Proced. Int. 8, 103 (1976). '1[~!-.. N' 0 ,.:..,
11-1.
N--ok 160
J-h'1I..~,f.
rv , ",.l..tct/f'"or;J.M,~,.,v;, 3
Synthesis of Morphine, Codeine, and Related Alkaloids
82. K. C. Rice, J. Org. Chem. 40, 1850 (1975); M. M. Abdel-Monem and P. S. Portoghese, J. Med. Chem. IS, 208 (1972). 83. A. Pohland and H. R. Sullivan, Jr., U. S. Patent 3,342,824 (Sept. 19, 1967). 84. T. A. Montzka, J. D. Matiskella, and R. A. Partyka, Tetrahedron Lefl. p. 1325 (1974). 85. K. C. Rice and E. L. May, J. Heterocycl. Chem. 14,665 (1977); R. A. Olofson, J. T. Martz, J.-P. Senet, M. Piteau, and T. Malfront, J. Org. Chem. 49, 2081 (1984). 6. K. C. Rice and A. E. Jacobson, J. Med. Chem. 18, 1033 (1975). 87. M. F. Lockett and M. M. Davis, J. Pharm. Pharmacol. 10, 80 (1958). 88. J. W. Millar and H. H. Anderson, J. Pharmacal. Exp. Ther. 112, 191 (1954). 89. P. D. Orahovats and E. G. Lehman, J. Pharmacol. Exp. Ther. 122, 58A (1958). 90. L. Lasagna and T. J. DeKornfeld, J. Pharmacal. Exp. Ther. 124, 260 (1958). 91. H. F. Fraser, A. Wikler, G. D. Van Horn, A. J. Eisenman, and H. Isbell, J. Pharmacol. Exp. Ther. 122, 359 (1958). 92. M. R. Fennessy, Br. J. Pharmacal. 34, 337 (1968). 93. A. Crum-Brown and T. R. Fraser, Tram. R. Soc. Edinburgh 25, 1 (1868). 94. T. W. Smith, P. Buchan, D. N. Parsons, and S. Wilkinson, Life Sci. 31, 1205 (1982). 95a. R. S. Foster, D. J. Jenden, and P. Lomax, J. Pharmacal. Exp. Ther. 157, 185 (1967). 95b. M. A. Iorio, A. Disciullo, A. Mazzeo-Farina, and V. Frigeni, Eur. J. Med. Chem. 19, 11 (1984). 96. E. R. Hart, J. Pharmacal. Exp. Ther. 72, 19 (1941); E. R. Hart and E. M. McCawley, ibid. 82, 339 (1944). 97. J. Weijlard and A. E. Erickson, J. Arn. Chem. Soc. 64, 869 (1942). 98. K. Unna, J. Pharmacal. Exp. Ther. 79, 27 (1943). 99. J. Pohl, Z. Exp. Pathal. Ther. 17, 370 (1915). 100. J. von Braun, Ber. Dtsell. Chem. Ges. 49, 977 (1916). 101. L. Lasagna and H. K. Beecher, J. Pharrnacol. Exp. Ther. 112,356 (1954); A. S. Keats and J. Telford, ibid. 117, 190 (1956). 102. W. R. Martin and C. W. Gorodetzky, J. Pharmacol. Exp. Ther. 150, 437 (1965). 103. R. L. Clark, A. A. Pessolano, J. Weijlard, and K. Pfister 111;J. Am. Chern. Sac. 75, 4963 (1953). 104. A. F. Green, G. K. Ruffell, and E. Walton, J. Pharm. Pharmacal. 6, 390 (1954). 105. C. A. Winter, P. D. Orahovats, and E. G. Lehman, Arch. 1m. Pharmacadyn. 110, 186 (1957). 106. A. E. Jacobson, K. C. Rice, J. Reden, L. Lupinacci, A. Brossi, R. A. Streaty, and W. A. Klee, J. Med. Chem. 22, 328 (1979). 107. S. Archer and W. F. Michne, Prog. Drug Res. 20, 45 (1976). 108. M. J. Lewenstein and J. Fishman, U. S. Patent 3,254,088 (1966). 109. H. Blumberg, H. B. Dayton, M. George and D. N. Rappaport, Fed. Proc., Fed. Am. Soc. Exp. Bioi. 20, 311 (1961). 110. A. S. Keats and J. Telford, personal communication; quoted in L. S. Harris, Narc. Drugs: Biochem. Pharmacal. 89 (1971); Chem. Abstr. 79,26977 (1973). Ill. H. Blumberg, H. B. Dayton, and P. S. Wolff, Proc. Soc. Exp. Bioi. Med. 123, 755 (1966). 112. F. F. Foldes, J. N. Lunn, J. Moore, and 1. M. Brown, Am. J. Med. Sci. 245,23 (1963). 113. D. R. Jasinski, W. R. Martin, and C. A. Haertzen, J. Pharmacal. Exp. Ther. 157, 420 (1967). 114. M. J. Lewenstein and J. Fishman, U. S. Patent 3,320,262 (1967); Chem. Abstr. 67, 90989 (1967). 115. H. Blumberg and H. B. Dayton, in "Narcotic Antagonists," p. 43 (M. C. Braude, L. S. Harris, E. L. May, J. P. Smith, and J. E. V. Hareal, eds.), Raven, New York, 1974.
161
References
116. M. S. Gold, C. A. Dackis, A. L. C. Pottash, H. S. Sternbach, W. J. Annitto, D. Martin, and M. P. Dackis, Med. Res. Rev. 2,211 (1982). 117. W. R. Martin and D. R. Jasinski, C/in. Pharmacal. Ther. 14, 142 (1973). 118. H. Blumberg, H. B. Dayton, and P. S. Wolff, Taxicol. Appl. Pharmacal. 10, 406 (1967). 119. W. R. Martin, D. R. Jasinski, and P. A. Mansky, Arch. Gen. Psychiatry 28, 784 (1973).
120. J. M. Fujimoto, Proc. Soc. Exp. Bioi. Med. 133, 317 (1970). 121. E. J. Cone, Tetrahedrolz Lett. p.2607 (1973); N. Chatterjie, J. M. Fujimoto, C. E. Inturrisi, S. Rocrig, R. I. H. Wang, D. V. Bowen, F. H. Field, and D. D. Clarke, Drug Metab. Dispos. 2,401 (1974); J. M. Fujimoto, S. Roerig, R. I. H. Wang, N. Chatterjie, and C. E. Inturrisi, Prac. Soc. Exp. Bioi. Med. 148, 443 (1975). 122. V. Nakagawa and K. Minami, Tetrahedron Lefl. p. 343 (1972). 123. N. Chatterjie, C. E. Inturrisi, H. B. Dayton, and H. Blumberg, J. Med. Chern. 18,490 (1975). 124. K. Vereby and S. J. Mule, Am. J. Drug Alcohol Ab,ue 2, 357 (1975). 125. H. Blumberg, I. J. Pachter, and Z. Matossian, U. S. Patent 3,332,950 (1967); Chem. Abstr. 67, 100301 (1967); H. Blumberg and H. B. Dayton, Adv. Bioelzem. Psychopharmacal. 8, 33 (1973). 126. W. T. Beaver, Mod. Med. (Minneapolis) 44, 21 (1976); W. T. Beaver and G. Feise, J. Pharmacol. Exp. Ther. 204, 487 (1978). 127. H. Merz, K. Stockhaus, and H. Wick, J. Med. Chem. 20, 844 (1977); M. D. Aceto, L. S. Harris, W. L. Dewey, and R. L. Balster, Committee on Problems of Drug Dependence, Natl. Acad. Sci. (1976). 128. O. J. Braendon, N. B. Eddy, and H. Halbach, Bu//. WHO 13, 937 (1955). 129. R. J. J. C. Lousberg and U. Weiss, Experientia 30, 1440 (1974). 130. C. Olieman, L. Maat, and H. C. Beyerman, Recl. Trav. Chim. Pays-Bas 95, 189 (1976). 131. G. Stork and L. Bauer, J. Am. Chem. Soc. 75, 4373 (1953). 132. L. F. Small, H. M. Fitch, and W. E. Smith, J. Am. Chem. Soc. 58, 1457 (1936); N. B. Eddy, H. Halbach, and O. J. Braendon, Bu//. WHO 17, 569 (1957). 133. J. W. Fairbairn, Pharm. J. (London) 216, 29 (1976); D. von Neubacher and K. Monthes, Planta Med. 11,387 (1963); N. Sharghi and I. Lalezari, Nature (London) 213, 1244 (1967).
134. L. F. Small, S. G. Turnbull, and H. M. Fitch, J. Org. Chern. 3, 204 (1939). 135. D. D. Weller and H. Rapoport, J. Med. Chem. 19, 1171 (1976). 136. L. F. Small, N. B. Eddy, E. Mosettig, and C. K. Himmelsbach, Stud. Drug Addict., Suppl., p. 138, Washington, D. C. (1938). 137. N. B. Eddy, H. Halbach, and O. J. Braenden, Bu//. WHO 14, 353 (1956). 138. L. J. Sargent and E. L. May, J. Med. Chem. 13, 1061 (1970). 139. J. W. Lewis, K. W. Bentley, and A. Cowan, Annu. Rev. Pharmacal. 11,241 (1971); J. W. Lewis, Adv. Biachem. Psychopharmacol. 8, 123 (1974). 140. I. Iijima, J. Minamikawa, K. C. Rice, and A. E. Jacobson, J. Med. Chern. 21, 1320 (1978). 141. D. L. Leland, J. P. Polazzi, and M. P. Kotick, J. Org. Chem. 45, 4026 (1980). 142. M. P. Kotick, D. L. Leland, J. O. Polazzi, J. F. Howes, and A. R. Bousquet, J. Med. Chem. 24, 1445 (1981).
143a. S. Archer and P. Osei-Gyimah, J. Heterocyc/. Chern. 16, 389 (1979). 143b. C. Mannich and K. Schulte, Arch. Pharm. (Weinheim, Ger.) 276, 593 (1938). 144. D. L. Leland and M. P. Kotick, J. Med. Chem. 24, 717 (1981). 145. J. Quick, P. Herlihy, R. K. Razdan, and J. F. Howes, J. Med. Chem. 25,983 (1982). 146. M. P. Kotick, D. L. Leland, J. O. Polazzi, J. F. Howes, and A. Bousquet, J. Med. Chem.
26, 1050 (1983).
162
3
Synthesis of Morphine, Codeine, and Related Alkaloids
147. D. L. Leland and M. P. Kotick, J. Org. Chern. 48, 1813 (1983). 148. J. W. Lewis, K. W. Bentley, and A. Cowan, Annu. Rev. Pharrnacol. 11,241 (1971); G. D. Smith and J. F. Griffen, Science 199, 1214 (1978). 149. M. P. Kotick, D. L. Leland, J. O. Polazzi, and R. N. Schut, J. Med. Chern. 23, 166 (1980). 150. F. Ghoz1and, P. Maroni, I. Viloria, and J. Cros, Eur. J. Med. Chem.-Chirn. Ther. 18,22 (1983). 151. M. P. Kotick and J. O. Polazzi, J. Heterocycl. Chern. 18, 1029 (1981). 152. D. L. Leland, J. O. Polazzi, and M. P. Kotick, J. Org. Chern. 46, 4012 (1981). 153. B. Proska, Z. Voticky, L. Molnar, J. Putek, and M. Stefek, Chem. Zvesti. 32, 710 (1978). 154. H. B. Arzeno, D. H. R. Barton, S. G. Davies, X. Lusinchi, B. Meunier, and C. Pascard, Nouv J. Chirn. 4, 369 (1980). 155. J. W. Fairborn, Pharm. J. (London) 216, 29 (1976); D. von Neubacher and K. Monthes, Planta Med. 11, 387 (1963); N. Sharghi and I. Lalezari, Nature (London) 213, 1244 (1967); J. W. Fairbairn and K. Helliwell, J. Pharm. Pharmacal. 29, 65 (1977). 156. M. Freund and E. Speyer, J. Prakt. Chem. 94, 135 (1916). 157. F. M. Hauser, T. K. Chen, and F. I. Carroll, J. Med. Chem. 17, 117 (1974). 158. I. Iijima, K. C. Rice, and A. Brossi, Helv. Chim. Acta 60, 2135 (1977). +;.,.. .JJ2.. U. Weiss, J. Org. Chem. 22, 1505 (1957).
~161.
U. Weiss, J. Am. Chem. Soc. 77,5891 (1955)'-7
~ )'",/1.1
o;.;y~7'~
~.
0.,.......
R. M. Allen and G. W. Kirby, J. Chern. Soc., Chern. Corn/lultl. p. 1346 (1970). 162. S. Archer and P. Osci-Gyimah, J. Heterocycl. Chem. 16, 389 (1979). 163. H. Blumberg, S. Carson, and E. Stein, Fed. Proc., Fed. Arn. Soc. Exp. Bioi. 13,451 (1954). 164. N. B. Eddy and L. E. Lee, Jr., J. Pharmacal. Exp. Ther. 125, 116 (1959). 165. U. Weiss and S. J. Daum, J. Med. Chern. 8, 113 (1965); M. G. Lester, V. Petrow, and O. Stephanson, Tetrahedron 21, 771 (1965); D. J. Barron, P. L. Hall, and D. K. Vallance, J. Pharm. Pharrnacol. 18, 239 (1966). 166. T. B. Zalucky and G. Hite, J. Med. Chem. 3, 615 (1961). 167. W. R. Buckett, M. E. Farquharson, and C. G. Haining, J. Pharm. Pharrnacol. 16, 174 (1964); W. R. Buckett, ibid. 16, Suppl. 68T (1964). 168. W. R. Buckett, J. Pharm. Pharrnacol. 17, 760 (1965). 169. P. Osei-Gyimah and S. Archer, J. Med. Chem. 23, 162 (1980). 170. H. W. Kosterlitz and A. A. Waterfield, A1I1l!1.Rev. Pharrnacol. 15, 32 (1975). 171. P. Osei-Gayimah, S. Archer, M. G. C. Gillan, and H. W. Kosterlitz,J. Med. Chern. 24, 212 (1981). 172. L. S. Schwab, J. Med. Chem. 23, 698 (1980). 173. P. Horsewood and G. W. Kirby, J. Chern. Res. Miniprint 401 (1980); J. Chern. Res. Synop. p. 4880 (1980). 174. S. Archer, A. Seyed-Mozaffari, P. Osei-Gyimah, J. M. Bidlack, and L. G. Abood, J. Med. Chern. 26, 1775 (1983). 175. K. W. Bentley and D. G. Hardy, Proc. Chem. Soc. p. 220 (1963). 176a. K. W. Bentley, "The Chemistry of the Alkaloids" (S. W. Pelletier, ed.), p. 117. Van Nostrand Reinhold, Ncw York, 1970. 176b. K. W. Bentley, in "The Alkaloids" (J. E. Saxon, cd.), Specialist Periodical Reports, Vol. 1, pp. 127-135. The Chemical Society, London, 1971; see also Vol. 11. pp. 102-103, 1981. 176c. H. O. Bernhard and V. A. Snieckus, in "The Alkaloids" (1. E. Saxon, cd.), Specialist Periodical Reports, Vol. 2, pp. 138-142. The Chemical Society, London, 1972; see also Vol. 5, pp. 144-145, 1975.
References
163
176d. M. Shamma, in "The Alkaloids" (J. E. Saxon, ed.), Specialist Periodical Reports, Vol. 3, pp. 161-163. The Chemical Society, London, 1973; see also Vol. 13, pp. 149-153, 1983. 177. K. W. Bentley, Alkaloids (N. Y.) 13,75-124 (1971). 178. J. W. Lewis, M. J. Readhead, I. A. Selby, A. C. B. Smith, and C. A. Young,J. Chern. Soc. C p. 1158 (1971). 179. R. Rubenstein, F. Haviv, and D. Ginsberg, Tetrahedron 30, 1201 (1974). 180. P. Horsewood and G. W. Kirby, J. Chem. Soc., Chem. Commun. p. 1139 (1971). 181. G. W. Kirby and J. M. Sweeny, J. Chem. Soc., Chem. Cornrnun. p. 704 (1973). 182. T. L. Gilchrist, M. E. Peek, and C. W. Rees, J. Chem. Soc., Chem. Commun. p. 913 (1975). 183. T. L. Gilchrist, C. J. Harris, F. D. King, M. E. Peek, and C. W. Rees, J. Chem. Soc., Perkin Trans. 1 p. 2161 (1976). 184. G. W. Kirby and J. G. Sweeny, J. Chern. Soc., Perkin TraIlS. I p. 3250 (1981). 185. K. W. Bentley, P. Horsewood, G. W. Kirby and S. Singh, J. Chern. Soc., Chern. Cornrnun. p. 1411 (1969). 186. K. W. Bentley, G. W. Kirby, A. P. Price, and S. Singh,J. Chem. Soc., Perkin TrailS. I p. 302 (1972). 187. G. W. Kirby, K. W. Bentley, P. Horsewood, and S. Singh, J. Chern. Soc., Perkin TraIlS. I p. 3064 (1979). 188. R. Giger, R. Rubenstein, and D. Ginsberg, Tetrahedron 29, 2387 (1973). 189. K. W. Bentley and D. G. Hardy, J. Am. Chern. Soc. 89, 3267 (1967). 190. K. W. Bentley, J. W. Lewis, and A. C. B. Smith, J. Chell!. Soc., Perkin Trans. 1 p. 870 (1972). 191. J. W. Lewis, K. W. Bentley, and A. Cowan, Annu. Rev. Pharrnacol. 11, 241 (1971). 192. R. E. Lister. J. Pharrn. Pharmacol. 16, 364 (1964). 193. L. S. Schwab, J. Med. Chern. 23, 698 (1980). 194. J. J. Brown, R. A. Hardy, Jr., and C. T. Nora Roth, U. S. Patent 3,562,279 (1971); Chern. Abstr. 75, 20225d (1971). 195. K. W. Bentley, J. D. Bower, and J. W. Lewis, J. Chern. Soc. C. p. 2569 (1969). 196. J. L. Lewis, M. J. Readhead, and A. C. B. Smith, J. Med. Chern. 16,9(1973). 197. P. R. Crabbendam, L. Maat, and H. C. Beyerman, J. R. Netll. Chern. Soc. 100, 293 (1981). 198. H. van Koningsveld, L. Maat. and T. S. Lie, Acta Crystallogr. 40, 1082 (1984). 199. J. J. Brown, R. A. Hardy, Jr., and C. T. Nora, U. S. Patent 3,318,885 (1967); Chem. Abstr. 67, 82295v (1967). 200. K. Hayakawa, I. Fujii, and K. Kanematsu, J. Org. Chern. 48, 166 (1983). 201. K. W. Bentley, D. G. Hardy, H. P. Crocker, D. T. Haddlesey. and P. A. Mayor, J. Am. Chern. Soc. 89, 3312 (1967). 202. J. W. Lewis, M. J. Readhead, and A. C. B. Smith, 1. Chern. Soc., Perkin Trails 1 p. 878 (1972). 203. K. W. Bentley and J. C. Ball, J. Org. Chern. 23, 1725 (1958). 204. K. W. Bentley and A. F. Thomas, J. CJIe/ll. Sac. p. 1863 (1956). 205. K. W. Bentley and B. Meek, J. Chern. Soc. C. p. 2233 (1969). 206. Z. J. Barneis, J. D. Carr, R. J. Warnet, and D. M. S. Wheeler, Tetrahedron 24, 5053 (1968). 207. K. W. Bentley. D. G. Hardy. J. W. Lewis, M. J. Readhead, and W. I. Rushworth, J. Chern. Soc. C. p. 826 (1969). 208. H. C. Beyerman, T. S. Lie, L. Maat, and M. Noordam-Wcissdorf, J. R. Netlz. Chern. Soc. 101, 455 (1982).
164
3
Synthesis of Morphine,
Codeine, and Related Alkaloids
References
209. K. W. Bentley, D. G. Hardy, and A. C. B. Smith, J. C/lem. Soc. C p. 2235 (1969). 210. K. C. Rice, A. E. Jacobson, T. R. Burke, Jr., B. S. Bajwa, R. A. Streaty, and W. A.
240. K. W. Bentley, D. G. Hardy, C. F. Howell, W. Fulmor, J. E. Lancaster, J. J. Brown, G. O. Morton, and R. A. Hardy, Jr., J. Am. Chem. Soc. 89, 3303 (1967). 241. J. W. Lewis and M. J. Readhead, J. Chem. Soc. C p. 2296 (1971). 242. A. A. Freer, G. A. Sim, 1. G. Guest, A. C. B. Smith, and S. Turner, J. Chem. Soc., Perkin Trans. 2 p. 401 (1979). 243. A. Cowan and B.A. Whittle, Br. J. Pharmacal. 44, 353P (1972); N.S. Doggett, H. Reno, and P. S. J. Spencer, ibid. 50, 440P (1974). 244. A. Cowan, Fed. Proc., Fed. Am. Soc. Exp. Bioi. 40, 1497 (1981). 245. K. W. Bentley, J. D. Bower, and A. C. B. Smith, J. Chem. Soc. C p. 2241 (1969). 246. D. 1. Haddlesey, J. W. Lewis, P. A. Mayor, and G. R. Young, J. Chem. Soc., Perkin TraIlS. I. p. 872 (1972). 247. J. W. Lewis, M. J. Rance, and G. R. Young, J. Med. Chem. 17,465 (1974). 248. D. 1. Haddlesey, J. W. Lewis, and P. A. Mayor, J. Che,iJ. Soc., Perkin Trans. I p. 875 (1972). ' 249. J. W. Lewis, P. A. Mayor, and D. 1. Haddlesey, J. Med. Chem. 16, 12 (1973). 250. A. L. A. Boura, D. 1. Hadd[esey, E. J. R. Harry, J. W. Lewis, and P. A. Mayor, J. Pharm. Pharmacal. 20,961 (1968).
Klee, Science 220, 314 (1983). 211. K. W. Bentley, J. D. Bower, J. W. Lewis, M. J. Readhead, A. C. B. Smith, and G. R. Young, J. Chem. Soc. C. p. 2237 (1969). 212. O. Hromatka, M. Knollmuller, and G. Sengstschmid, Mona/sh. Chem. 99, 1662 (1968). 213. J. W. Lewis and M. J. Readhead, J. Chem. Soc. C p. 2298 (1971). 214. J. L. Lewis and W. 1. Rushworth, J. Chem. Soc. C p. 560 (1970). 215. H. Rapoport and P. Sheldrick, J. Am. Chem. Soc. 85, 1636 (1963). 216. K. Hayakawa, S. Motohiro, 1. Fujii, and K. Kanematsu, J. Am. Chem. Soc. 103,4605 (1981). 217. A. Singh, S. Archer, K. Hoogsteen, and J. Hirshfield, J. Org. Chem. 47,752 (1982); A. Singh, J. Hirshfield, K. Hoogsteen, and S. Archer, ibid. 48, 173 (1983). 218. R. Rubenstein, R. Giger, and D. Ginsberg, Tetrahedron 29, 2383 (1973). 219. W. Fulmor, J. E. Lancaster, G. O. Morton, J. L. Brown, C. F. Howell, C. T. Nora, and R. A. Hardy, Jr., J. Am. Chem. Soc. 89, 3322 (1967). 220. J. J. Brown and R. A. Hardy, Jr., U. S. Patent 3,474,102 (1969); Chem. Abstr. 72, 21816x (1970); J. J. Brown and R. A. Hardy, U. S. Patent 3,474,103 (1969); Chem. Abstr. 72, 43967z (1970).
251. 252. 253. 254. 255. 256. 257. 258.
J. W. Lewis and M. J. Readhead, J. Med. Chem. 16, 84 (1973). J. H. van den Hende and N. R. Nelson, J. Am. Chem. Soc. 89, 2901 (1967). D. H. Thorpe, Anesth. Analg. (Cleveland) 63, 143 (1984). A. H. Beckett and A. F. Casey, J. Pharm. Pharmacal. 6, 986 (1954). P. S. Portoghese, J. Med. Chem. 8, 609 (1965). H. F. Fraser and L. S. Harris, Annu. Rev. Pharmacal. 7, 277 (1967). K. W. Bentley, D. G. Hardy, and P. A. Mayor, J. Chem. Soc. C p. 2385 (1969). C. W. Hutchins, G. K. Cooper, S. Piirro, and H. Rapoport, J. Med. Chern. 24,773 (1981). 259. W. F. Michne, R. L. Salsbury, and S. J. Michalec, J. Med. Chem. 20, 682 (1977). 260. W. F. Michne, J. Org. Chem. 41, 894 (1976). 261. W. F. Michne, J. Med. Chem. 21, 1322 (1978). 262. T. A. Crabb and J. R. Wilkinson, J. Chem. Soc., Perkin Trans. I p. 644 (1976). 263. K. W. Bentley, Endeavor 23,97 (1964); K. W. Bentley and J. W. Lewis, in "Agonist and Antagonist Actions of Narcotic Analgesic Drugs" (H. W. Kosterlitz, H. O. J. Collier, and J. E. Villarreal, cds.), p. 7. Univ. Park Press, Baltimore, Maryland, 1973. 264. E. F. Hahn, Medicinal Res. Rev. 5, 255 (1985). 265. T. R. Burke, Jr., K. C. Rice, and C. B. Pert, Heterocycles 23, 99 (1985); M. A. Channing, W. C. Eckelman, J. M. Bennett, T. C. Burke, Jr., and K. C. Rice, 1m. J. Appl. Radial. Isol. 36, 429 (1985). 266. C. B. Peat, J. A. Danks, M. A. Channing, W. C. Eckelman, S. M. ,Izrson, J. M. Bennett, T. R. Burke, Jr., and K. C. Rice, FEBS LeU. 177, 281 (1984). 267. 1. lijima, J. Minamikawa, A. E. Jacobson, A. Brossi, K. C. Rice, and W. A. Klee, J. Med. Chem. 21, 398 (1978). 268. M. Erez, A. E. Takemori, and P. S. Portoghese, J. Med. Chern. 25,847 (1982); P. S. Portoghese, and A. E. Takemori, Life Sci. 36, 801 (1985).
221. K. W. Bentley, D. G. Hardy, and B. Meek, J. Am. Chem. Soc. 89, 3273 (1967). 222. K. W. Bentley and D. G. Hardy, J. Am. Chem. Soc. 89, 3281 (1967). 223. M. Hori, T. Kataoka, H. Shimizu, E. Imai, T. Iwamura, M. Nozaki, M. Niwa, and H. Fujimura, Chem. Pharm. Bu//. 32, 1268 (1984). 224a. D. C. G. Perry, J. S. Rosenbaum, M. Kurowski, and W. Sadee, Mol. Pharmacal. 21, 272 (1982). 224b. A. M. Harthoorn, Fed. Proc., Fed. Am. Soc. Exp. Bioi. 26, 1251 (1967). 225a. G. F. Blane, A. L. A. Boura, A. E. Fitzgerald, and R. E. Lister, Br. J. Pharmacal. Chemother.
[65
30, 11 (1967).
225b. H. E. Dobbs, G. F. Blane, and A. L. A. Boura, Eur. J. Pharmacal. 7, 328 (1969). 226. J. W. Lewis and M. J. Readhead, J. Med. Chem. 16, 525 (1970). 227. K. W. Bentley, U. S. Patent 3,433,791 (1969); Chem. Abstr. 70, 78218s (1969). 228. J. W. Lewis, Adv. Biochern. Psychopharmacol. 8, 123 (1973). 229. G. F. Blane, A. L. A. Boura, E. C. Leach, W. D. Gray, and A. C. Osterberg, J. Pharm. Pharmacal. 20, 796 (1968); G. F. Blane and D. S. Robbie, in "Agonist and Antagonist Actions of Narcotic Analgesic Drugs" (H. W. Kosterlitz, H. O. J. Collier, and J. E. Villarreal, cds.), pp. 120-127. Univ. Park Press, Baltimore, Maryland, 1973. 230. J. D. Bousfield and J. M. H. Rees, J. Pharrn. Pharmacal. 21, 630 (1969). 231. J. T. Macbeath and B. A. Moore, Br. J. Anaesrh. 48, 97 (1976). 232. K. W. Bentley, A. L. A. Boura, A. E. Fitzgerald, D. G. Hardy, A. McCoubrey, M. L. Aikman, and R. E. Lister, Nalure (London) 206, 102 (1965). 233. J. Dum, J. Bliisig, and A. Herz, Eur. J. Pharmacal. 70,293 (1981); A. Cowan, J. W. Lewis, and 1. R. Macfarlane, Br. J. Pharmacal. 60, 537 (1977). 234. M. J. Rance and J. S. Shillingford, Biochem. Pharmacal. 25, 735 (1976). 235. J. H. Mendelson, J. Ellingboe, N. K. Mello, and J. Kuehnle, J. Phannacol. Exp. Ther. 220, 252 (1982). 236. K. W. Bentley, D. G. Hardy, B. Meek. J. B. Taylor, J. J. Brown, and G. O. Morton, J. Chem. Soc. C p. 2229 (1969). 237. K. W. Bentley, D. G. Hardy, and B. Meek, J. Am. Chem. Soc. 89, 3293 (1967). 238. J. W. Lewis and M. J. Readhead, Telrahedron 24, 1829 (1968). 239. J. W. Lewis and M. J. Readhead, J. Chem. Soc., Perkin Trans. I p. 881 (1972).
l
I
4.
A.
166 167 170 172 173 174 174 183 185
I. Physicochemical Studies Physicochemical measurements have played a major role in assessing the structure, configuration, conformation, and physical behavior of opiate analgesics. X-Ray crystallography and nuclear magnetic resonance (NMR) spectroscopy have been used extensively. Other methods, such as infrared (IR), optical rotatory dispersion (ORO), circular dichroism (CD), and mass spectra, as well as partition coefficient and dissociation constant (pKa) measurements, have also found some application. X-ray crystallography has been of particular value in studying semirigid opiates such as the morphine series, where the solid-state structure may be expected to resemble closely the biologically active conformation. For more flexible compounds, such as the arylpiperidines and methadone analogs, NMR has been used to assess solution conformation under a variety of conditions of solvent, temperature, and pH. IR spectroscopy has been employed in studying intramolecular hydrogen bonding in the flexible opiates. ORO and CD are potentially useful in the study of chiral compounds, but most reports to date have simply interpreted spectra in terms of known structural features. Measurements of partition coefficients and pKa values have been carried out with the hope of relating these properties to in vitro and in vivo activity as well as examining possible mechanisms of action. The following discussion focuses on the application of physicochemical methods to the study of compounds in the morphine, morphinan, and benzomorphan series. 166
167
Studies
X-Ray Crystallography
Mackay's and Hodgkin's crystallographic study of morphine (1) in 1955 played a key role in elucidating the structure of this substance (I). Since then, numerous compounds in the morphine series have been studied by
Physical Chemistry, Molecular Modeling, and QSAR Analysis of the Morphine, Morphinan, and Benzomorphan Analgesics I. Physicochemical Studies. . A. X-Ray Crystallography B. Proton NMR . . . . . C. Carbon-13 NMR . . . D. Dissociation Constants and Panition Coefficients II. Molecular Modeling and OSAR Studies A. Molecular Modeling. B. OSAR Studies References . . . . . .
Physicochemical
morphine
codeine
this method. Kartha et al. (2) determined the absolute configuration of the morphine skeleton in their 1962 study of codeine (2). Table 4-1 lists the morphine derivatives for which X-ray crystalJographic studies have been reported. All of these compounds have in common a T shape in which rings A and B constitute the vertical part of the T, and ring C and the piperidine ring represent the horizontal portion. The piperidine ring is in a chair conformation, with the N-substituent equatorial for alJ of these compounds. N-Methylnalorphine (3) has a quaternized piperidine Table4-1 X-Ray Crystallographic Studies of Morphine-Type Compounds CoRing Conformation
Compound (I) Morphine Codeine (2) N-Methylnalorphine Nalorphine (4) Oxymorphone (5) (6) Naloxone
(3)
7 3-Hydroxylevallorphan Azidomorphine (9) 14- Hydroxyazidomorphine Dextromethorphan (10) Nalbuphine (11) 19- Propylthevinol (12) 6-Acetyl-l-iodocodeine 13
(8)
Boat Boat Boat Boat Chair Chair Chair Chair Chair Chair Chair Boat Boat Boat Planar
-
References 1,4-6 2 3 7 8,9 10,11 12 13 14 15 16 /7 /8 /9 20
168
4
Benzomorphan
Analgesics
HO
-tN
-
CH2-CH=CH2
I CHJ
1
!!
N-methy1-na1orphine
I
Physicochemical Studies
169
derivative (7)], a 6,B-substituent [3-hydroxylevallorphan (8) and azidomorphine (9)], or no 6-substituent [dextromethorphan (10)] exist in a Coring chair conformation, with 6,B-substituents equatorial (8-16). Nalbuphine (11) has a saturated Coring, but its 6-substituent has a stereochemistry; its Coring has a boat conformation with the 6-hydroxy group in a bowsprit position (17). 19-Propylthevinol (12) has its Coring held rigidly in a boat conformation by the 6,14-ethano bridge (18). The 8-bromo-6,7unsaturated compound 13 has a nearly planar Coring, with the exception of carbon atom 14 at the B-C ring junction (20).
nalorphine
nitrogen; the active (opiate antagonist) isomer has an axial methyl group and an equatorial allyl substituent (3). The isomer having an equatorial methyl and an axial allyl group is inactive, suggesting that the N-allyl group must be in the equatorial position in order for the compound to be active as an antagonist. The Coring conformation of these compounds is dependent on the presence or absence of a 7,8-double bond and on the stereochemistry of the 6-substituent. Compounds such as morphine (1), codeine (2), and nalorphine (4) exhibit a Coring boat conformation (1-7,19); the 6-a-substituent is in a bowsprit position. Coring saturated compounds having a 6-keto group [oxymorphone (5), naloxone (6), and the 8, 14-buteno
2
10
HO
N -CHzO CHJ
:
HO -q
2 oxymorphone
... CH2-CH2-CHJ
CHJ
2 11
naloxone
12
nalbuphine 19-propylthevinol
1
8
11
170
B.
4
Benzomorphan
Analgesics
171
I Physicochemical Studies
Proton NMR
Proton NMR has been used to study the Coring conformations of numerous compounds in the morphine series (21-24). The NMR results are consistent with and extend the X-ray crystallographic findings (see above). In general, the following conclusions may be drawn from H-H coupling constants, chemical shifts, and shielding of 3- and 6-acetoxy groups: 1. When the Coring is saturated, it exists in a chair form. 2. When there is a 7,8-double bond [e.g., morphine (1)], the Coring exists in a boat or half-boat conformation. 3. When there is an 8,14-double bond [e.g., neopine (14)], the Coring exists in a half-chair conformation. 4. When the Coring is a 6,7,8,14-diene [e.g., thebaine (15)], the C-ringis essentially planar.
CN
11
16 B/c
t~ns-morphine
HO
In the B-C trans-fused morphine system (16), the saturated Coring is forced into a boat or skew-boat conformation (24). Proton NMR was used extensively by Fulmor et al. to determine the stereochemistry and configuration of several 6,14-endo-ethenotetrahydrothebaine compounds (25). In the case of compound 17, the stereochemistry of the cyano group was determined by examining long-range couplings of the olefinic protons. In compound 18, NMR and IR spectra showed that the C-19 hydroxyl group is hydrogen-bonded to the 6-methoxy group in both stereoisomers; stereochemistry at C-19 could be assigned on the basis of the observation that in one isomer the 19-phenyl group can shield the olefinic proton resonances by 0.5-0.6 ppm relative to the other isomer.
N -CH~
12 6a-na1trexo1.
R1
=
OH. R2
6p-na1trexo1.
R1 =
H. R2
na1trexone.
R1R2
H OH
= 0
consistent with other NMR and X-ray results; for the 60' compound, a boat conformation might be anticipated on the basis of the X-ray study of
nalbuphine (17). Brine
14 neopine
1.2 thebaine
Brine et al. reported that both 6a- and 6{3-naltrexol (19) and their mono- and diacetates exist in a Coring chair conformation (26). For the 6{3 compound, a chair conformation with an equatorial hydroxyl group is
et al.
noted changes in the H-H coupling constants
and 13C NMR signals, which they attributed to an intramolecular hydrogen bond between the axial 6-hydroxy and the 3-acetoxy group. This hydrogen bond can be formed only when ring C is in a chair conformation. For the spiro-oxirane derivative 20, Jacobson et al. used the coupling constants and the nuclear Overhauser enhancement between olefinic and oxirane protons to assign the configuration at C-6 of the compound, as shown (27). In addition, they showed that the Coring exists in a boat conformation with the oxirane oxygen in a flagpole position. Proton NMR coupling constants and chemical shifts were also used by Jacobson and co-workers to determine the positions and configurations of the halogenated morphine derivatives 21 and 22 (28).
172
4
Benzomorphan
Analgesics
173
I Physicochemical Studies
in two different environments. Subsequently, the X-ray crystal structure of morphine sulfate was determined by Wongweichintana et al. (6); they found no evidence of disorder at C-15. The cause of the split signal for C-1S under these conditions is still not resolved. D. HO 20
21 c:r-ch1oromorphide
22
Dissociation Constants and Partition Coefficients
Dissociation constants (pKa) and partition coefficients are physicochemical properties related to the ability of drugs to pass through lipid membranes. For example, Kutter et al. showed that intravenously, etorphine (23) is 3800 times as potent as an analgesic as dihydromorphine (24); intraventricularly, it is only 40 times as potent (34). This result is attributed to the higher lipid solubility of etorphine, which facilitates its diffusion across the blood-brain barrier. Since only the neutral (un-ionized) form of the molecule is likely to diffuse readily through lipid membranes, the pKa may have considerable influence on the ability of opiates to reach central receptor sites. Kaufman et al. determined the pKa values for a set of 15
/1-ch1oromorphide
Using 600 MHz proton NMR, Glasel demonstrated nitrogen inversion in morphine (29). At low pH, a small amount of N-axial methyl was detected with slow interconversion to the N-equatorial form. The axial form was estimated to be about 1 kcaljmole less favorable than the equatorial form. The interconversion rate at neutral pH was found to be about equal to the NMR time scale, many orders of magnitude slower than the on-rate to the receptor. Thus, the N-axial conformation should be considered as a possible receptor-active conformation.
CHJO CHJ
- CI: -CH2-CH2-CHJ OH
C.
Carbon-13 NMR
Carbon-13 NMR has not been used to any appreciable extent in determining conformation or configuration in the morphine series. Most of the. carbon NMR literature for these compounds reports spectra and assIgns the resonances (30-32). Carroll et al. interpreted their carbon-13 NMR results to mean that 6,14-endo-ethano compounds such as 19propylthevinol (12) have A-, B-, and D-ring conformations similar to those of morphine (32). Hexem et al. reported the high-resolution carbon-13 NMR spectrum of crystalline morphine sulfate (33). The spectrum exhibits resolved resonances for each of the 17 carbon atoms. Each of the carbon atoms bound to nit.rogen exist.s. as an asymmetrical doublet, due to I4N_I3C dipolar couplIng. In addItIOn, the signal for C-lS exists as a nearly symmetrical doublet. These authors suggested that C-1S (part of the D-ring) may exist
n etorphine
214 dihydromorphine
opiate agonists and antagonists from several structural classes (35). In addition, they measured octanol-water distribution coefficients for these compounds at two temperatures (20° and 37°C) and at several pH values. Distribution coefficients increased with temperature, the magnitude of the increases ranging from 21 to 200%. These authors noted that distribution coefficients may be extremely sensitive to pH for amines that have pKa values near physiological pH. They point out that physiological pH, which is considered to be 7.4, may in fact range from 7.1 to 7.7, resulting in a 300-400% difference in distribution coefficient for many opiate compounds. Since the normal pH of the fetus is 7.2, opiates tend to accumulate on the fetal side of the placental barrier.
174
II.
4
Molecular
Benzomorphan
Analgesics
The div~rsi~y of chemical structures that are active narcotic analgesics has made It dIfficult to construct a general structure-activity relationship (SAR). Most opiate agonist and antagonist SAR data are based on morphine and structurally related rigid and semirigid analogs. Thus, it is not surprising that most molecular modeling and quantitative structureactivity relationship (QSAR) studies of opiates have involved the morphine family. A.
II
Molecular Modeling and QSAR Studies
175
Modeling and QSAR Studies
ojf Fig. 4-1. The morphine structural pharmacophore.
Of ~
Molecular Modeling
Morphine and structurally related cyclic analgesics, as well as opiates that have only a single ring in common with morphine, share a common overlap of the phenyl ring and the amine nitrogen. Beckett and Casy (36) have proposed these two common substructures as a basis for constructing a common pharmacophore. The pharmacophore can be described as a 4-phenylpiperidine in the chair conformation with the phenyl ring in the axial position (Fig. 4-1). This model has two limitations. First, potent azabicycloalkanes (25) are restricted to the equatorial form (37), and unrestricted 4-phenylpiperidines (see Chapter 8) are generally most stable in the equatorial state. Second, compounds like etonitazine (26) and fentanyl (27) are 100-10,000 times as potent as morphine but do not exhibit an isomorphic chemical connectivity relationship to morphine.
Fig. 4-2.
I
Axial and equatorial
ri~ .~._.~
~..... ~:'
geometries
of the phenyl ring.
Fries and Portoghese (38) suggested that the analgesic receptor allows narcotic agents to interact with the phenyl ring, either axial or equatorial. Andrews and Lloyd (39) pointed out that such a proposal is not geometrically inconsistent with a common binding site (Fig. 4-2). Belleau and co-workers (40) have hypothesized that the opiate receptor is flexible and can be specifically and divergently perturbed into a shape negating the appearance of morphine-like effects. Specific conformations can be structurally induced that are translated into analgesia while disallowing the pharmacological side effects of narcotics. The authors initially focused attention on the fact that naloxone (6) and naltrexone (19), both uniquely "clean" effectors, carry a unique polar substituent vicinal (at C-14) to the nitrogen atom. This, in turn, suggested that the introduction of such a substituent at an equivalent position into potent but "unclean" morphinans might allow the synthesis of clean analgesics. The corresponding synthetic strategy was pursued (41), leading to the discovery of butorphanol (28), an analgesic with reduced narcotic side effects (42). HO
. 26
28
Further investigations by Belleau and colleagues of the opiate receptor have led to the clastic binding proposal (43). This proposal consists of two
176
4
Benzomorphan
Analgesics
II
p.arts: (a) a p.roductiv: effector-:rece~tor interaction requires a stereospecIfic lone pair (or N -H) onentatIon by the tertiary nitrogen at the receptor site (44); (b) the binding event is accompanied by an oxidative one-electron transfer from the lone pair of the nitrogen to the receptor (45) or by a proton t.ransfe~ fro~ .the corresponding salt to the receptor (46). The hypothesIs of bIOactIvity dependence on the directionality of the nitrogen lone pair has been questioned by Shiotani et al. (47). Kolb (48), on the other hand, favorably evaluated the electron transfer proposal and extended the concept to include the formation of intermediate radicalanion-radical-cation pairs (49). Tollenaere and Moereels (50) performed PCILO calculations to estimate t?e proton affinity (dEpA) of the basic nitrogen atom of morphine, ~orphInan, benzomorphan, phenylpiperidine, and \fentanyl-type analgeSICS.Except for the phenylpiperidine, the dEpA values of these compounds are ~340 kcaI/mole. These authors suggested that dEpA is a quantitative measure of the nature of chemical moieties in the vicinity of a basic nitrogen. They also presented examples that indicate an additive character of dEpA. Snyder et al. (51) have investigated oxidative one-electron transfer mechanisms in a series of pyrrolidine- and piperidine-containing polycycles that are similar to morphine. They used the MM2 force field (52) with full geometry optimization to generate molecular structures. The MM2 calculations were followed by a set of approximate ab initio PRDDO (53) calculations using the molecular geometries determined by the MM2 ca!culations. It may be noted that questions of computational accuracy anse when a molecular mechanics method is used to determine geometries where .orientations of lone pair orbitals are a primary concern. Further, these workers did not optimize molecular geometry in the PRDDO calculations s? that these workers had difficulty in achieving convergence for PRDDO In some of the electronic configuration calculations. Ease of ionization at the nitrogen atom of these compounds was evaluat~d by com~aring relative nitrogen lone pair energies and by computIng energy difference between the neutral amine and radical-cation pai~s. ~bsence of a correlation between the ease of nitrogen lone pair IOnIZatIOnand analgesic activity for a set of compounds was interpreted to mean that oxidation clastic binding is not operative at the f-t opiate receptor. The interaction of the quaternized amine group with an anionic receptor . site has also been considered using intermolecular modeling. Model receptor studies implicate a sulfate or phosphate moiety as a plausible anionic receptor site (54). Loew et al. (55) have used a "supermolecule" approach employing the PCILO method (56) and fixed valence geometry
Molecular Modeling and QSAR Studies
177
m olecular mechanics force field calculations (57) to determine the intersome N-substituted mO lecular energy of complex formation between . . benzomorphans and model anionic receptor sItes. A mmomum met h y I phosphate (AMP) and ammonium methyl sulfate (AMS) .were sel~~te~ as mo del rece p tor sites. The calculated complex formatIon stablhzatton .. d ffi d en ergies with both anionic sites tend to follow the observe a mtles an . potencies, as can be seen in Table 4-2. The three most potent antagonIsts are calculated to form the three most stable complexes with AMS and are among the four most stable complexes with AMP. The two nearly pure gonists have the smallest calculated interaction energies. Although there ~ no measured binding constant for the 2-meth~lthiofuran compound, the weak antagonism does not correlate well wIth the calculated strong complex energy. . Andrews and Lloyd (39) suggest that the search for a common analgesIc pharmacophore is actually hindered by the selection of ~orphine ~s. a conformational standard. They reason that the conformatIOn of a ngld structure like that of morphine may be far from optimal with respect to maximizing analgesic potency. A more rational, if more diffi~ult, approach, from their point of view, would be to use newer, m?re flexible, and more potent analogs to deduce the common stereochemIcal components of analgesic activity. . Andrews and Lloyd (39) pointed out that the presence of an aromatIc ring and a nitrogen atom in most central nervous system (CNS) drugs has long been recognized. However, these workers. note~ that molecu.lar superposition studies of the crystal structur~s ?f elg~t dlvers~ CNS-actIng drugs [chlorpromazine (antipsychotic), Imlpra~mne (anttdep~essant~, amphetamine (stimulant), LSD (hallucinogen),. dlphenylhy~antoIn (antIconvulsant), diazepam (anxiolytic), phenobarbItal (hypnotIc), and .morphine (analgesic)] indicate that the rin~s .and nitrog.ens,. respectIvely, occupy equivalent spatial locations. This IS Illustrated In FIg. 4-3. These authors speculated about the possibility of a common CNS pharm~cophore, of which the original analgesic model of Beckett and Casy (36) IS a specialized subset. .' Loew and co-workers (58,59) carried out molecular orbItal calculatlO~s, including PCILO calculations (56), on a series of mix.ed agonist-anta?omst N-substituted rigid opiates. Two distinct conformatIOns correspondmg to two different induced receptor site conformations are postulated. The two conformations are shown in Fig. 4-4. Using this hypothesis, Loew et al. (59) suggested the synthesis of a series of morphine analogs predicted to have a range of agonist-antagonist potency ratios. Some of these .c~mpounds were synthesized and found to be active in preliminary prechOlcal tests (60).
L
Table 4-2 Antagonist Potency with Model Anionic
_J co
and Binding Affinity Receptor Sites
of Benzomorphans
N-R
ANT"
Kc x 109 c
Methylcyclopropane 2-Methylfuran 3-Methylfuran Propene Ethylbenzene 2-Methylthiofuran 2,3-Dimethylfuran Propane
3.0 0.66 O.66b 0.22 0.03 0.024 0.0
0.8 2.5 3.0 10.0
with Varying
N-Substituents
- AEpCILOd
and Their Energies
d
of Interaction
- AEEMP
-ApCILOd
-AEEMpd
5.2 5.5 5.2 3.8 4.9' 4.1 5.2 -.f
8.0 2.1 2.8 1.0 1.8' 1.4 2.1 -.f
3.2 4.2 4.0 2.9 3.8' 3.8 3.6 -.f
8.2 6.8 7.2 6.2 7.3' 9.1 4.9 3.1
a
Antagonism relative to nalorphine = 1 in guinea pig ileum. h The 3-methylfurfuryl and 2-methylfurfuryl benzomorphans are equally potent antagonists in mice and monkeys, No guinea pig ileum data exist for the 3-methyl analog. C
Stereospecific binding constants Ke. = units of energy in kilocalories per mole.
d -AE
'Calculated for be nzene. fNot calculated.
,.
.., 0
0'::1 .." 3 ~.~. , ~g . r'I>.... .., .
~...., 0 ::s
g
~~:!1
(")
~o;-~ (JQ
CO
Q
....
0
"0
f]. g,
..., ::sn> <= ~"0
.., 0 0' ::s 0'
..,
::s 3n> '" n> .., .., (JQ
'" "0
~.
'<
-J -D
c;: OJ ---~ (JQ ~3 .., 5' 0c:
3' ~"0 c: .., 3 a::> ;" 0' ~::> .., B (JQ '" C;; .'< .., "0 .., 5' 0 ::> ~::; OJ <= 0" c...., -. ::> "0 (JQ ::s-. ::> 0;- .. ~::s'" 0; :§: (JQ a 0' ::I
~! ;" ::> n> .., <> (JQ 0 :;' ':'-
o' ::s a
:J:
0<;' '"
(")
to
c.. :I:
:I:
<'
'" (') Z V>
0,
(JQ 5'
e:
(JQ <=
180
4
Benzomorphan
Analgesics II
The atomic charge densities and bond polarities were computed for both the protonated and free base forms of morphine as part of these studies (58). These physicochemical properties are reported in Table 4-3 for general reference. The calculated electrostatic potential patterns (61) of morphine derivatives were also generated for purposes of comparision and opiate receptor mapping. The electrostatic potential pattern of morphine in the plane of the benzene ring is shown in Fig. 4-5. A hydroxyl group was used as the field probe to generate this figure. The atom numbering code is the same as in Table 4-3. Figure 4-5 indicates that the entire molecule is surrounded by a highly repulsive potential, which is typical of morphine-like molecules, and represents an "excluded region" analogous to a van der Waals radius. The highly positive potential attenuates at varying radial rates to zero for different parts of the molecule. There are regions of negative potential of
Molecular Modeling and OSAR Studies Table 4-3 (com.) C. Piperidine ridge atomic charges (a.u,) Base Atom CIS CI6 HIs HIs HI6 HI6
o Ring atom
Net Atomic Charges and Bond Polarities in Morphine Protonated and Base Forms.
CI
Base Atom
Protonated
PCILO
INDO
PCILO
INDO
NI3 C. CI6 CM H... HI6 HI6 HM HM HM HN
-0,13 +0,11 +0,09 +0.06 -0,03 -0.02 -0.02 -0,01 0,0 0.0
-0.22 +0.14 +0.16 +0.16
+0,07 +0,10 +0.09 +0,05 +0,01 +0,03 +0,05 +0,05 +0.06 +0,05 +0.16
+0,06 +0.12 +0,14 +0.12 +0.00 +0.01 +0,02 +0.02 +0.04 +0,04 +0,15
IA
+0.05
+0,72
+0.72
,,;,,0,04
-0,03 -0,05 -0,03 -0,03 -0.04
+0,02
B, N bond polarities (number of electrons in the bonding atomic orbitals of each atom)
Base N-CII N-C" N-CM N-H
1.06-0.94 1.06-0,94 1.05-0,94
Protonated
PCILO
INDO
PCILO
INDO
0,0 +0,09 0,0 0.0 -0.02 -0.02
+0,04 +0,16 -0,02 -0.02 -0.03 -0,03
-0,02 +0.09 +0.02 +0,04 +0,03 +0,05
+0,03 +0.14 0,00 +0.02 +0,02 +0,02
D. Benzene and ring C atomic charges (a.u.)
Table 4-3
A, Nitrogen group atomic charges (a,u,)
181
Base PCILO
Protonated INDO
PCILO
INDO
C3 C. CI2 CII
-0.03 ~-0,14 +0.16 +0,07 -0,07 +0.06
+0.01 -0,03 +0,18 +0,13 -0,02 +0,02
-0,03 -0.13 +0,17 +0.07 -0.09 +0.06
+0,03 -0,02 +0,19 +0.14 -0,03 +0,00
C6 C7 CR C1. CI3 Cs
+0.13 -0,03 0.0 +0,02 +0.03 +0,10
+0.20 -0,03 0,0 +0.03 +0.03 +0,17
+0,13 -0,01 0,0 +0,01 +0.02 +0.11
+0,20 -0.0 0.0 +0,04 +0,02 +0.18
E, 0 bond polarities (as in Part B) Base arC. OF-CS OrC3 06-C6
.
1.14-0,90 1,14-0.88 1.12-0,90 1.13-0.89
Protonated 1.12-0.90 1.12-0,89 1.10-0.92 1.1 1-0.91
From ref. 58.
Protonated 1.29-0.73 1.26-0.75 1.22-0.78 1.25-0,84
various extents and depths within the overall repulsive (+) potential region generated by the molecule. It is these negative potential regions, together with the large positive potential around the nitrogen cationic head, that appear to constitute the major receptor-sensitive features of the rigid opiates.
182
4
Benzomorphan
Analgesics II
Molecular
Modeling
and QSAR
Studies
183
:RI
o
?
C'~R2
"\
H
30
40 100
C,J
~
100) '0 -40
Fig. 4-5. benzene
Calculated electrostatic potential pattern of morphine in the plane of the
ring.
Kaufman et al. (62-64) have used molecular potential contour maps, generated using a pseudo ab initio quantum chemical calculation (65,66), to differentiate between narcotic agonists and antagonists. In particular, eseroline (29) has been compared to morphine to explain opiate analgesic activity (67). The potential fields about the pharmacophorically significant nitrogen atoms of eseroline and morphine are nearly identical. However, there are significant differences between the fields in the phenolic region as well as other groups in these two molecules. Kaufman et al. have also employed quantum chemical calculations to postulate the respective molecular requisites for interaction with 1-', 0, K, 0', and other opiate receptors (67). Conformational analyses of a series of oripavine derivatives (30) using the PClLO method (56) were carried out by Loew and Berkowitz (68) to identify a structural basis for differences in agonist potency between
OH
diastereoisomers of carbinol substituents at C-7. Low-energy conformers of the carbinol substituents are found with and without hydrogen bonding' to the C-6 methoxyl group. The relative energies of these conformers depend on the R, and Rz groups as well as the diastereoisomerism of the alcohol. The authors suggested that the interaction of specific conformations of C-19 carbinols and a lipophilic receptor site is critical to agonist potency. However, no SAR applicable for molecular design can be extracted from this study. Verlinde et al. (69) explored the role of the extra oxygen, common to most benzomorphans with opioid K properties, on the net charges, bond polarities, and proton affinities using the PClLO method (56). Net atomic charges and bond polarities in the nitrogen region are very similar to those computed by Loew and Berkowitz (59) for some morphine-like opiate narcotics. Bond polarities and proton affinities were found to vary from compound to compound. However, no correlation with any measure of analgesic activity could be identified. This may be due in part to the fact that the calculations were carried out directly on crystal structures without doing any geometry optimization. A quantum mechanical model for the interaction of narcotic analgesics with receptors was constructed by Gomez-Jeria and Peradejordi (70). The dissociation constant was computed as a function of the net charge and delocalizability of the drug. The calculated dissociation constant fit the observed constant at the 98% confidence level. B.
QSAR Studies
Lien et al. (71) formulated a set of QSAR relationships for narcotic analgesic agents using classical linear free energy descriptors. Mager (72) has applied pattern recognition methods to discriminate among morphinomimetic opioids. Fries and Bertelli (73) have carried out a Hansch analysis for a series of l-phenyl-3-aminotetralins and found the I-phenyl substituent
184
4
Benzomorphan
Analgesics
References
is critical to the drug-receptor interaction. However, the conformation of the phenyl ring was not found to be decisive in specifying opiate activity. Simon et al. (74) have attempted to map the analgesic receptor using the steric mapping procedure [the minimum steric difference (MSD) approach] devised specifically for steric fit effects. A relationship was established between the analgesic activity after intravenous and intraventricular administration, respectively, and the hydrophobicity of the molecule. The receptor model obtained by the MSD method agrees with Casy's model of a goblet-shaped receptor site (37). However, the MSD model also predicts a side "lid" interacting with the extra ethylenic moiety of the etorphine compound. A QSAR was derived from the following MSD-based analysis:
A,
~
MSD descriptors and biological activity measures are derived in this study. Also, the relationship between lipophilicity and MSD is not discussed, In general, the QSAR studies on the morphine family are not as useful in rationalizing experimental observations or serving as a base in designing new compounds as the molecular modeling efforts.
References 1. M. Mackay and D. C. Hodgkin, J. Chern. Soc. p. 3261 (1955). 2. G. Kartha, F. R. Ahmed, and W. H. Barnes, Acta Crysta//ogr. IS, 326 (1962). 3. R. J. Kobylecki, A. C. Lane, C. F. C. Smith, L. P. G. Wakelin, W. B. T. Cruse,E. Egert, and O. Kennard, I. Med. Chern. 25, 1278 (1980). 4. L. Gylbert, Acta Crysrallogr., Sect. B B29, 1630 (1973). 5. E. Bye, Acta Chern. Scand., Sect. B B3O, 549 (1976). 6. C. Wongweichintana, E. M. Holt, and N. Purdie, Acta Crysrallogr.,Secr. C C40, 1486 (1984). 7. Y. G. Gelders, C. J. De Ranter, and C. Van Rooijen-Reiss, Crysr. Struct. Commun. 8, 995 (1979). 8. R. J. Sime, M. Dobler, and R. L. Sime, Acra Crystallogr., Sect. B B32, 2937 (1976). 9. S. D. Darling, V. M. Kolb, G. S. Mandel, and N. S. Mandel, J. Pharrn. Sci. 71, 763 (1982). 10. I. L. Karle, Acta Crystallogr., Sect. B B3O, 1682 (1974). 11. R. L. Sime, R. Forehand, and R. J. Sime, Acta Crystallogr. Sect. B B31, 2326 (1975). 12. A. A. Freer, G. A. Sim, I. G. Guest, A. C. B. Smith, and S. Turner,I. Chern. Soc., Perkin T'ans. 2 p. 401 (1979). 13. J. F. Blount, E. Mohacai, F. M. Vane, and G. J. Mannering, J. Med. Chern. 16,352 (1973). 14. K. Sasvari, K. Simon, R. Bognar, and S. Makleit, Acta Crystallagr., Sect. B B3O, 634 (1974). 15. A. Kalman, Z. Jgnath, K. Simon, R. Bognar, and S. Makleit, Acta Crystallogr., Sect. B
3.08- 0.55MSD
n~1I
R~0.93
AI is the measured intraventricular analgesic activity and MSD is the measure of steric fit (75). The numerical values for AI and MSD are given in Table 4-4. There is considerable difficulty in understanding how both the Table 4-4 Stereochemical
Description
and Correlations,
Substance
1 2
Elorphine; Elph( -) Fenlanyl: Fent( -)
3
Hydromorphnne; (Mph) (-), des ~7'= 011 Morphine: Mph( -) Dihydromorphine; Mph(-), des ~, 2(2-Furyl)elhyllevorphanol; Lev( -), R]: CH2CHrfuryi Normorphine; Mph( - ). R]=H Levorphanol; Lev( - ) R,=CH3 Ketobemidonc; Ketb( -)
Reduced
A, +0.22 -0.82
Seriesa
j(x'j= 1) 1-30 4-6,9,13-18,20-21,
MSD 5 8
29-37, 43-46
4 5 6
7 8 9 10
Methadone; Med(R)
Pethidine; Peth(-). R]=Me, R2= R3 =H
II
a
-1.09
1-20,30
8
-1.31
- 1.38
1-20,30 1-20. 30
8 8
-1.38
1-14,16-19,30-36
8
-1.71
1-20
-2.18
1-14,16-19,30
-2.31
1-6,9. 11-14, 16-20 29,30,47 1-6,9,11-14, 16, 17 20, 29, 30, 47, 53, 71-76
-2.78
-3.09
1-6,9,11-14,16-18, 20.29,30,47
185
16. 17. 18. 19.
9 10
20. 21. 22.
9 II
23. 24. 25.
B32, 2667 (1976). L. Gylbert and D. Carlstrom, Acta Crystallogr., Sect. B B33, 2833 (1977). R. J. Sime, M. Dobler, and R. L. Sime, Acta Crystallogr., Sect. B 832, 809 (1976). J. H. van der Hende and N. R. Nelson, I. Am. Chern. Soc. 89, 2901 (1967). A. A. Liebman, D. H. Malarek, J. F. Blount, N. R. Nelson, and C. M. Delaney, I. Org. Chern. 43, 737 (1978). H. van Koningsveld, T. S. Lie, and L. Maat, Acta Crystal/ogr., Sect. C C40, 313 (1984). R. Rull, Bull. Soc. Chirn. h p. 586 (1963). S. Okuda, S. Yamaguchi, Y. Kawazoe, and K. Tsuda, Chern. Pharrn. Bull. 12, 104 (1964). W. Geiger and H. Wollweber, EUr. J. Med. Chern. 17, 207 (1982). H. Kugita, M. Takeda, and H. Inoue, Terrahedron 25, 1851 (1969). W. Fulmor, J. E. Lancaster, G. O. Morton, J. J. Brown, C. F. Howell, C. T. Nora, and
R. A. Hardy, Jr., J. Arn. Chern. Soc. 89, 3322 (1967). 26. G. A. Brine, D. Prakash, C. K. Hart, D. J. Katchmar, C. G. Moreland, and F. 1. Carroll, J. 0'8. Chern. 41, 3445 (1976). 27. A. E. Jacobson, H. J. C. Yeh, and L. J. Sargent, Org. Magn. Reson. 4, 875 (1972).
10
From ref. 74.
.
186
4
Benzomorphan
Analgesics
28. H. J. C. Yeh, R. S. Wilson, W. A. Klee, and A. E. Jacobson. J. Pharm. Sci. 65,902 (1976). 29. J. A. Glasel, Biochem. Biophys. Res. Commun. 102, 703 (1981). 30. Y. Terui, K. Tori, S. Maeda. and Y. K. Sawa, Tetrahedron Let!. p. 2853 (1975). 31. C. J. Kelley, R. C. Harruff, and M. Carmack, J. Org. Chern. 41, 449 (1976). 32. F. I. Carroll, C. G. Moreland, G. A. Brine, and J. A. Kepler, J. Org. Chern. 41,996 (1976). 33. J. G. Hexem, M. H. Frey, and S. J. Opella, J. Am. Chern. Soc. 105, 5717 (1983). 34. E. Kutter, A. Herz, H..J. Teschemacher, and R. Hess, J. Med. Chern. 13,801 (1970). 35. J. J. Kaufman, N. M. Serna, and W. S. Koski. J. Med. Chern. 18, 647 (1975). 36. A. H. Beckett and A. F. Casy, J. Pharm. Pharmacal. 6, 986 (1954). 37. A. F. Casy, Prog. Drug Res. 22, 149 (1978). 38. D. S. Fries and P. S. Portoghcse, J. Med. Chern. 19, 1155 (1976). 39. P. R. Andrews and E. J. Lloyd, Med. Res. Rev. 2, 355 (1982). 40. B. Belleau, "Chemical Regulation of Biological Mechanisms," p. 201. Academic Press, New York, 1981. 41. I. Monkovic, H. Wong, A. W. Piccio, Y. G. Perron. I. J. Pachter, and B. Belleau, Can. J. Chern. 53, 3094 (1975). 42. F. S. Caruso, A. W. Piccio, H. Madissoo, R. D. Smyth, and I. J. Pachter, in "Pharmacological and Biochemical Properties of Drug Substances" (M. E. Goldberg, ed.), Vol. 2, p. 19. American Pharmaceutical Association-Academy of Pharmaceutical Sciences, Washington, D.C., 1979. 43. "Clastic" comes from the word "Klatos," which is Greek for "broken." In biology, "clastic" is used to describe a division into parts. 44. B. Belleau, T. Conway, F. R. Ahmed, and A. D. Hardy, J. Med. Chern. 17,907 (1974). 45. B. Belleau and P. Morgan, J. Med. Chern. 17, 908 (1974). 46. B. Belleau, U. Gulini, B. Gour-Salin, R. Camicioli, S. Lemaire, and F. Jolicoeur, "Proceedings of the Second Camerino Symposium on Recent Advances in Receptor Chemistry," pp. 1-14 Elsevier/North.Holland Biomedical Press, Amsterdam, 1984. 47. S. Shiotani, T. Kametani, Y. Jitaka, and A. Itai, J. Med. Chern. 21, 153 (1978); S. Shiotani. J. Med. Chern. 21, 1105 (1978). 48. V. M. Kolb, J. Pharrn. Sci. 73, 715 (1984). 49. R. B. Silverman, S. J. Hoffman, and W. B. Catus III, J. Am. Chern. Soc. 102, 7126 (1980); T. Shono, T. Toda, and N. Oshino,~. Am. Chern. Soc. 104,2639 (1982). 50. J. P. Tollenaere and H. Moereels, Eur. J. Med. Chern. 15, 337 (1980). 51. J. P. Snyder, T. A. Halgren, and V. M. Kolb, J. Med. Chern., in press. 52. N. L. Allinger, J. Am. Chern. Sac. 99, 8127 (1977). 53. T. A. Halgren and W. N. Lipscomb, J. Chern. Phys. 58, 1569 (1973). 54. H. H. Loh, T. M. Cho, Y. C. Wu, and E. L. Way, Life Sci. 14, 2233 (1974). 55. G. Loew, S. Burt, P. Nomura, and R. Macelroy, in "Computer Assisted Drug Design" (E. C. Olsen and R. E. Christoffersen, eds.), p.243. American Chemical Society, Washington, D.C., 1979. 56. J. L. Coubeils, P. Courriere, and B. Pullman, C. R. Acad. Sci. Paris 272,1813 (1971). 57. A. J. Hopfinger, "Conformational Properties of Macromolecules." Academic Press, New York, 1973. 58. G. H. Loew, D. Berkowitz, H. Weinstein, and S. Srebrenik, in "Molecular and Quantum Pharmacology" (E. D. Bergmann and B. Pullman, eds.), p.355. Reidel, Dordrecht-Holland, 1974. 59. G. H. Loew and D. S. Berkowitz, J. Med. Chern. 18, 656 (1975). 60. J. J, DeGraw, J. A. Lawson, J. L. Crase, H. L. Johnson, M. Ellis, E. T. Uyeno, G. H. Loew, and D. S. Berkowitz, J. Med. Chern. 21, 415 (1978).
References
187
61. P. Politzer and D. G. Truhlar, eds., "Chemical Applications of Atomic and Molecular Electrostatic Potentials." Plenum, New York, 1981. 62. J. J. Kaufman, Int. J. Quantum Chern. 16, 221 (1979). 63. J. J. Kaufman, NIDA Res. Monogr. No. 22, p. 250 (1978). 64. J. J. Kaufman, P. C. Hariharan, F. L. Tobin, and C. Petrongolo, in "Chemical Applications of Atomic and Molecular Electrostatic Potentials" (P. Politzer and D. G. Truhlar, eds.). Plenum, New York, 1981. 65. H. E. Popkie, W. S. Koski, and J. J. Kaufman, J. Arn. Chern. Soc. 98, 1342 (1976). 66. H. E. Popkie and J. J. Kaufman, Int. J. Quantum Chern. 10, 569 (1976). 67. J. J. Kaufman, in "Advances in Endogenous and Exogenous Opioids, Proceedings of 12th International Narcotics Research Conference" (T. Hiroshi and E. J. Simon, eds.), p. 417. Kodansha, Tokyo, 1981. 68. G. H. Locw and D. S. Berkowitz, J. Med. Chern. 22, 603 (1979). 69. C. L. Verlinde, N. M. Blaton, C. J. De Ranter, and O. M. Peeters, J. Med. Chern., in press. 70. J. S. Gomez-Jeria and F. Peradejordi, Bal. Soc. Chilo Quim. 27, 145 (1982). 71. E. J. Lien, G. L. Tong, D. B. Srulevitch, and C. Dias, NIDA Res. Monog. No. 22, p. 186 (1978). 72. P. P. Mager, Act. Nerv. Super. 23, 136 (1981). 73. D. S. Fries and D. J. Bertelli, J. Med. Chern. 25, 216 (1982). 74. Z. Simon, N. Dragomir, and M. G. Plauchithiu, Rev. Rourn. Biochirn. 18,139 (1981). 75. Z. Simon, I. Badilescu, an? T. Racovitan, J. Theor. BioI. 66,485 (1977).
II
5. The Morphinans I. Introduction
. . . . . . . . .
188 189 190 193 193
II. NaturallyOccurringMorphinans. . . . . . . . . . . III. Conversion of Morphine and Its Analogs to Morphinans . IV. The Total Synthesis of Morphinans . A. GreweCyclization . . . . . . . . . . . . B. The Total Synthesis of 14-Hydroxymorphinans
199
C. OxidativeCouplingto Morphinandienones.
201
D. Electrochemical Oxidations . . . . . . . . . V. Structure-Activity Relationships of the Morphinans . . . A. Alteration of Existing Functional Groups and Structures. B.
The
Effect
of Substituents
in Nonfunctionalized
Positions
. . . . . .
204 206 207
on Morphinan
Analgesia. C. Ring Additions, Contractions, Enlargements, and Movements in the. Morphinans . . .. D. Movement of the Nitrogen within the Molecular Framework and Heteroatom Insertion. .. .. VI. The Chemical Anatomy of the Morphinans References
220 230 236 242 243
1. Introduction For many years, investigations on the modification or separation of morphine's biological properties centered on the modification of existing natural products. A complementary approach was the simplification of the basic molecule itself. One of the simpler conversions, on paper at least, was cleavage of the dihydrofuran ring to provide the tetracyclic nucleus, ultimately termed the morphinans. Although several other simpler derivatives of the rigid opiates had been prepared prior to the total synthesis of the first morphinan, the analgesic activity of these tetracyclic derivatives led to intensive investigation of these compounds, particularly in the mixed agonist-antagonist and antagonist series. This chapter discusses the "
188
Naturally
Occurring
189
Morphinans
occurrence of these types of molecules in nature, partial and total syntheses of the tetracyclic structure, and structure-activity relationships. Because of the close similarity of the morphinans and morphine, structureactivity relationships for analgesic activity are presented in a format and approach similar to that used for morphine. For the morphinans, those similar positions substituted in morphine are considered as a group, while the nonfunctionalized positions are a second group.
II.
Naturally Occurring Morphinans
The first isolation of a morphinan from a natural source occurred almost 20 years after the description of the synthetic material. The naturally occurring compounds contain a dienone grouping or a reduced version of it. The first alkaloid of this group identified was salutaridine (1), and it was isolated from Croton sa/utaris (1). Salutaridine was probably identical to floripavine, reported by Russian workers in 1935 (2). . . Interestingly, 1 is an intermediate in the bIOsynthesIs of morphme from tyrosine and was ultimately isolated from opium. The morphinandienone alkaloids have been isolated from a wide variety of plant genera (Table 5-1). The diversity of structure in the morphinan alkaloids parallels, in many respects, the structure-activity relationships of the .mo~hine-bas~d analgesics. Since salutaridine is an intermediate in morphme bIOsynthesIs, it has the same absolute configuration as morphine. However, its enantiomer, sinoacutine (2), has been isolated from Sinomenium acutum (17). This occurrence was taken advantage of when sinomenine (3) was converted into (+ )-morphine, the unnatural enantiomer (18). The presence of a 14-hydroxyl group in morphine derivatives has been underscored in Table S.I Plant Sources of Morphinan Alkaloids Plant Cassytha Cocculus Colchicum Corydalis Croton Fumaria Glaucium Meconopsis
Reference
3 4 5 3 3 6 7 8
Plant Nemuaron Ocotea Papaver Rhiziocarpa Sinomenium Stephania Thalictrum Triclisia
Reference
9 10 11 12 13 14 15 16
190
5 The Morphinans
1
3
( salutaridine)
(sinornenine)
2
4
(sinoacutineJ
III Conversion of
Morphine
191
and Its Analogs to,Morphinans.
this process attractive. The formation of morphinans from morphine and codeine, as well as a host of other alkaloid-derived derivatives, has been observed during the investigations of the structure and rearrangements of these opium alkaloids. The early work in this area, which is often confusing, has been reviewed and summarized (21). While an exhaustive review of cleavage methods will not be attempted, the more commonly used or interesting methods will be indicated. The majority of cleavage methods proceed from a 6-ketone. For instance, the generally used reagent for rupture of the dihydrofuran bridge is zinc and ammonium chloride (22,23). Dihydrocodeinone (7) is rapidly cleaved to dihydrothebainone (8) by this method. A 1413-hydroxyl group, which is important for analgesia enhancement, usefully survives these conditions (24). Reduction of 9 gives 10 in good yield. Other previously used methods included tin (25) and zinc or sodium analgams (26). Interesting approaches are the decomposition of hydrocodeinone hydrazone under Huang-Minion conditions to yield the morphinan olefin (II).
(tridictophylline)
Zn NH4Cl
8
7 5
(nudarine)
6
(dihydronudarine)
Chapter 3. This substituent is usually incorporated using thebaine intermedIates. !n the morphinan series~ a naturally occurring 14i3-hydroxy denvatlve, tnd~ctophylline (4), with the correct enantiomeric configuratIOn, has been Isolated (16). Besides the structural features illustrated in ~tructur~s 1 t~ 4, a substantial variety of morphinan alkaloids have been Isolated In whIch the ketone has been reduced or a double bond saturated for example, nudarine (5) and dihydronudarine (6) (19,20). '
III.
Zn
10
9
Conversion of Morphine and Its Analogs to Morphinans
While many morphinans have been used in the synthesis of morphine and related alkaloids (cf. Chapter 3), the ready availability of large amounts of morphIne, codeIne, and thebaine have made the reversal of
11
192
5 The Morphinans
12 (thebaine)
IV
193
The Total Synthesis of Morphinans
series of addition eliminations to yield 14 as the only isolated product. Methylation of the acidic hydroxyl at C-6 produces salutaridine (I) in a fair overall yield (28). A shorter, more efficient alternative is oxidation of thebaine (12) with air in the presence of sodium bisulfite, whereby 14 is formed directly in excellent yield (29). Cleavage of the dihydrofuran ring in thebaine (12) with retention of the dienol ether allows access to morphinan analogs of the extremely potent thebaine Diels-Alder adducts. Birch reduction of thebaine gives only the enol ether (IS) (30), which cannot be converted to the conjugated diene
13
a.j ~thebaine
(12)~ or c
, 15
au
16 Reagents: f
(a)
Na/NH3,
(b)
LAH,
(c)
KIN"]
(16) (31). Various hydride reductions of 12 to 16 have been described (32), but ultimately the reactions were found to be too carpricious and nonreproducible (30). A partial solution to the problem of the preparation of 16 is the reduction of 12 with potassium in liquid ammonia, whereby equimolar amounts of IS and 16 are formed and separated by crystalliza, tion (33).
b.
1
IV.
The Total Synthesis of Morphinans
A. Grewe Cyclization 14 Scheme 5-1. Reagents: (a) Claisens alkali; (b) diazomethane.
An alternative method for preparing 8 from 7 is heating with ethyl mercaptan and hydrochloric acid. The reaction mechanism proceeds through a phenoxide displacement by mercaptide to form the 5-ethylthiomorphinan, which can be isolated. The thioether is then displaced by additional mercaptan to form 8 (27). A novel way to form the morphinandienone alkaloid salutaridine (I) proceeds through 14-bromocodeinone (13), which is available from thebaine (12) in one step (Scheme 5-1). Heating 13 with alkali results in a
Morphine can be considered both a phenanthrene and an isoquinoline derivative, and many of the early attempts at synthesis focused on the preparation of hydrophenanthrene derivatives. A summary of the model studies has appeared (34). Cyclization of the acid (17) leads to two unexpected structures 18 and 19, the latter of which has striking similarity to the heterocyclic ring system of morphine (35). The structure of 19 and Robinson's hypothesis on the biogenetic formation of morphine (36) led Grewe to publish his speculations on a biomimetic synthesis of morphinelike alkaloids from suitably reduced isoquinolines (37). In using the biomimetic approach, the reduced isoquinoline (20), which carries the
194
5 The Morphinans
IV
The Total
Synthesis
195
of Morphinaos
H3P04
+ 23
17
18
19 1 d
c fl,. OH
e
0<'7"1,
loy;.t,
(
C1
~j.."l>,+;
~y1Y"j''''' lea,,), . OV& P 4.el.(~( ~e 0,
Cp.~c~J..1.
f
20
9
~J.;
< "'r""iI;
25
20 j
th\,
el'\~"""/~
n
<: C",,",+-0.1'7 &. cia f-hrl.,
,
Hy..l,,'~ 20
22
f;;:;,fr.J5:
CC'l'I~,;<:
R~j,"{+7
(",.4.117.
21
requisite phenethyl-cyclohexene grouping necessary for phenanthrene formation, was chosen. Using the acidic conditions that were successful in the alicyclic syntheses, cyclization of 20 should yield the desired tetracyclic morphinan ring system (21). The alternative cyclization to form the hydroaporphine (22) could also occur. Since this approach occurs through a cyclodehydration, it is biomimetic in a formal sense only, since morphine is formed biosynthetically through one-electron transfer oxidations (ef. Chapter I). The realization of this hypothesis is illustrated in Scheme 5-2. Modified Knoevenagel condensation of a-carbethoxycyclohexanone yields the un-
h
21 Scheme 5-2. The Grewe morphinan synthesk Reagents: (a) ethyl cyanoacetate; (~) saponification and decarboxylation; (c) ammonia; (d) phosphorus oxychloride; (e), cat~lytlc hydrogenation; (f) methyl iodide; (g) benzyl magnesium chloride; (h) phosphonc aCid.
196
5 The Morphinans
IV The Total Synthesis of Morphinans
correct regiochemistry (48). However, the I-methyl group in 31 is not removable. This problem has been solved by the use of the bromo (30) (R1 = CHO, R2 = Br) or hydroxy derivative (30) (R1 = CHO, R2 = OH) where the blocking group is ultimately removed by hydrogenation (cf. Schemes 3-4 and 3-5) (45,49). After the pioneering discoveries of Grewe, and as interest increased in the morphinans, the need for a commercial synthesis of this ring system became apparent. In addition, a common intermediate was necessary. This intermediate had to avoid the demethylation step that was inherent in the original synthesis to allow access to a wide variety of derivatives. The approach chosen was to use a cyclohexenylethylamine in place of the tetrahydroisoquinoline and is outlined in Scheme 5-4 (50). The crucial step in the synthesis is the extension of the Bischler-Napieralski reaction to appropriately placed, isolated double bonds. Previously, this cyclodehydration had been used only with aromatic rings. The requisite cyclohexenylethylamine (33) is readily prepared from cyclohexanone (32), and the amide (34) is obtained by heating 33 with a phenylacetic acid. After selective reduction and demethylation of the phenolic ether, the secondary amine (35) can be readily alkylated with a wide variety of substituents, thus meeting the second condition for a general intermediate. Cyclization to the morphinan then occurs according to standard Grewe conditions (50-53). To obtain the pure enantiomers, resolution can be effected either at the isoquinoline (36) (R = CH3) (54) or the morphinan (37) (R = CH3) stage (55). However, on a commercial scale, the isoquinoline (35) is readily resolved using tartaric acid (56). The morphinans produced by the Grewe reaction under the conditions listed above lead to rings BC cis-fused. When the BC ring junction is trans-fused, the resultant tetracycle is termed an isomorphinan. Access to the isomorphinan system has been more difficult than access to the morphinan one. It was first prepared, in a lengthy sequence, during Gate's synthesis of morphine (ct. Scheme 3-1) (57). Subsequently, it was synthesized, again in a difficult sequence, from thebaine (12) (58). In the benzomorphan series, use of aluminum chloride for the catalyst gives significant amounts of the trans isomer (59). Use of this reagent with 38 gives substantial amounts of the isomorphinan (39), which can be separated from the morphinan by differential rates of quarternization with methyl iodide (60). While the Grewe method for the synthesis of morphinans generally gives good yields, by-products, which can occasionally be troublesome, are also formed. One by-product consistently formed in 10-15% yield in the cyclization of 38 is the aporphine derivative (40) (61). Compound 40 probably arises through double bond migration in 38 and subsequent
+
HO 27
(R1""CH3,Ri"H)
30
(R1 =CHO, R2"'CH3)
28
(3%)
31
(85%)
29
t97
(37%)
Scheme 5-3.
saturated ester (24), which is converted to the heterocycle, 5,6,7,8tetrahydroisoquinoline (25). After formation of the methyl imminium salt of 25, grignard addition to the activated double bond yields the I-benzylhydroisoquinoline (26). This can be reduced by catalytic hydrogenation to the penultimate product (20). Today, this enamine reduction is usually done by complex hydride reduction. Treatment of the isoquinoline (20) with phosphoric acid then yields the parent compound of the morphinan series (21) (38,39). A great deal of synthetic effort is required to prepare 25, a crucial intermediate in the synthesis. Although early work indicated that it was not possible to prepare 25 by catalytic hydrogenation of isoquinoline, reduction over platinum in acidic methanol yields 25 in very high yield (40). This cyclization to morphinans has become a name reaction in organic synthesis: the Grewe cyclization (41). Synthetic improvements have included the use of the formamide instead of the N-methylamino in the isoquinoline (42). The use of this protecting group allows the cyclization reaction to proceed at a rate 400 times faster than that of the N-methyl. Cyclization catalysts other than phosphoric acid have included phosphoric and sulfuric acids (42), hydrochloric acid (43), hydrobromic acid (44). trifluoromethane sulfonic acid, and trifluoromethane sulfonic acid containing ammonium hydrogen bifluoride (45). The use of various protecting groups or substituents, as well as the problems inherent in unsymmetrical substitution on the benzyl radical, are illustrated in Scheme 5-3. Cyclization of 27 with either phosphoric or hydrochloric acid yields a mixture of two products, 28 and 29, in a ratio of 1:10 (46,47). The minor isomer (28) possesses the correct regiochemistry for conversion to codeine. With the introduction of a methyl group and the use of a formyl group on the nitrogen, cyclization of 30 with sulfuric acid proceeds more rapidly and in much higher yield to produce 31 with the I
~l
198
IV
5 The Morphinans
6
199
The Total Synthesis of Morphinans
AiCl)
a +
32
39
38
c5
b
~
c
33 HO
40 d
OCNH
H
e,f
~~-CH3
OCH3 34
H3
41
OH
43
cyclization and ether hydrolysis. The Grewe cyclodehydration of 41 has failed to yield any morphinan but is converted into the cyclic ether (42), which slowly rearranges to the aporphine (43). The stereochemistry of 43 and the structures of 42 and 43 have been elucidated by X-ray analysis (62). The other usual accompanying by-product of morphinan synthesis is the isomorphinan formed in 3-5% yield (63).
9
H
35
42
36
B.
The Total Synthesis of 14.Hydroxymorphinans
The presence of a 14,B-hydroxyl group in the morphine series enhances the analgesic potency of these molecules compared to that of the unsubstituted parent. The introduction of this group into the morphinan series represents a synthetic challenge, since the Grewe reaction does not lend itself to the introduction of this functional group. The ultimately successful method is illustrated in Scheme 5-5 (64). The readily available a-tetralone (44) is spiro-annulated (45) and then converted to the hydroxyamine (46).
h
--7
HO
37 S.c~eme 5-4. A commercial synthesis of the morphinan ring system. Reagents: (a) ~yndlDe; (b) H2. Rane~ Co; (c) p.methoxyphenylacetic acid, heat; (d) phosphorus oxychlonde; (e) H2. Raney NI; (f) demethylation; (g) alkyl halide; (h) Grewe cyclization.
l
200
5 The Morphinans
a
IV
The Total Synthesis of Morphinans
b.c
---+
'>
20t
a HO
45
46
d
9H
f b
47
,
-co-O
48
f,g,h
e 50
(H-H)
51
(R;COCF3)
j,c
53 Scheme 5.5. Synthesis of 14-hydroxymorphinans. Reagents: (a) sodium hydroxide. 1,4dibromobutane; (b) acetonitrile, base; (c) lithium aluminum hydride; (d) hydrochloric aqueous; (e) bromine; (f) sodium bicarbonate; (g) 13S"C in DMF; (h) trifluoroacetic anhydride; (i) m-chloroperbenzoic acid; (j) sodium borohydride. .
Treatment with acid causes rearrangement and dehydration to the hydrophenanthrene derivative (47). The fourth bridging ring is introduced by stereospecific bromination from the a face of 47, whereby the intermediate bromonium ion is captured by the amine. The resulting product (48) has the ring system characteristic of the hasubanane alkaloids, naturally occurring opium alkaloids. The free base of 48 forms the aziridine (49), which is isolable at low temperature. Heating (49) in the presence of weak
CH]O Scheme 5-6. Reagents: (a) m-chloroperhenzoic acid; (b) 64% sulfuric acid in 2-butanone; (c) LAH; (e) phosphoric acid.
base causes elimination to morphinan (SO) with 8(14)-unsaturation. After acylation of SO, the amide (51) is selectively epoxidized on the {3face (52), which is reductively hydrolyzed to the 14{3-hydroxymorphinan (53) (64,65). This general reaction has served for the synthesis of many 14{3-hydroxy compounds, as well as another approach to isomorphinans via 47 (66a). The Grewe cyclization has been successfully applied in the preparation of 14-hydroxymorphinans (Scheme 5-6). Epoxidation of the amide in Scheme 5-6 yields a mixture of diasteromers that is converted into the trans diol in 75% yield. After reduction of the amide to the amine and complexation with borane, the diol is smoothly cyclized with phosphoric acid to the 14-hydroxymorphinan in 65-70% yield (66b). C.
Oxidative Coupling to Morphinandienones
The oxidative coupling of l-benzylisoquinolines dienone ring system originated in experiments
to form the morphinandesigned to mimic the
202
5
The Morphinans
IV
The Total
Synthesis
biosynthesis of the morphine
Cu
54
55
CH3
l::,
CH30 CH30
CH3
OR 57 (R=CH3)
(1%)
56
(R=CH3)
61
(R=CI/2C6H5)
60
(R-CH2C6H5)
62
(R=H) (2%)
+
RO 58
(1%)
59 63
(R=CH3) (1.4%) (R=CH2C6H5) (8.4%)
I ! I
I
I I i
OCH3 65
(25%) II
64 I
203
of Morphinans
alkaloids (67).
Two approaches
have been
taken: (a) the use of the Pschorr reaction and (b) one-electron transfer oxidizing agents. The Pschorr reaction (68) has, on the whole and with an occasional exception, given very low yields. This reaction consists ofthe decomposition of a diazotized benzylisoquinoline (58) and has been used primarily to synthesize aporphines (55), although dienones can also be formed (69). The use of one-electron oxidizing agents, on the other hand, started with very low yields of morphinandienones, but as reagents have improved, this became a useful way of synthesizing these complex molecules. The scope of the Pschorr method of producing the morphinan ring system can be illustrated with compounds 56, 60, and 64 (70). Decomposition of the diazonium salt (56), which is readily prepared from the nitro derivative via the amine, yields approximately 1% each of the deamination product laudanosine (57), the aporphine glaucine (58), and the morphinan O-methylflavinantine (59) (71). The cydization products 58 and 59 result from the reaction of the electrophilic species produced in the diazonium decomposition with the starred positions in 56. When the trivial substitution of a benzyl for a methbxyl methyl group is made in 56 to produce 60, the decomposition furnishes small amounts of the two deamination products 61 and 62, but the morphinandienone (63) yield increases sixfold (72,73). When the position that leads to the aporphine type of structure (55) is blocked, the formation of the morphinandienone is enhanced, for example, 64 to 65 (74). In general, the Pschorr approach to morphinandienones remains of academic interest only. One-electron oxidizing agents have been extensively studied, mainly in an effort to mimic biosynthetic pathways. However, some of the reagents that have been developed provide the morphinan ring system in synthetically useful yields. Initially, Barton studied the oxidation of reticuline (66) (R = CH3) to salutaridine (1) in order to provide chemical evidence for the biochemical conversion of reticuline to salutaridine and eventually to morphine (67). A wide range of oxidants was investigated, but 1 was obtained in a maximum yield of 0.015%. Since then, extensive investigations of the oxidation of 66 by a wide variety of inorganic oxidizing agents have shown that the major morphinandienone product (67) is isomeric with I. Compound 67, pallidine, or its derivatives can now be prepared in good yield, especially considering the degree of complexity and reactivity of the molecule formed (Table 5-2). The structure of the molecule being oxidized may exert a profound effect on the yield of the resultant morphinan. While the conversion of 66 to 67 with VOCl3 is very low, the same reagent with the isoquinoline (68) yields the morphinan (69) in 34% yield (81).
104
5 The Morphinans Table 5-2 Formation
of Morphinandicnoncs
Using One-Electron
Oxidizing Agenrs
o>
[0)
rS U :I: ~ Z
66 Oxidant
R
Yield (%)
MnOz silica gel K,Fe(CN),
CH) CH3 CH, COCF, COCF, CO,CH,CH, CHO or CO,CH,CH, CHO or CO,CH,CH,
AgzCO.\-Celite
VOCI, VOF, TI(O,CCH,h TI(O,CCH,h Pb(O,CCH,), trichloroacetic
4
75
0.9 0.5 0.3 8 11 23 18-40
76 77 77 78 79 79 80
o~ c
.g 'i5 c o
u oM c:
i:: U II
acid
C,H,I(O,CCH,h trichloroacetic
Reference
14-31
£u
80
acid
~
c: o
. '"
VOC13
r2
34%
££:£ UUU:I:
rS UN ~ Z
:£
£u
U<
i:: U
ii':£:£ uuuu
68
D.
U
u£
£U
£:£:££ uuuu
:£:£ uu
:£
:£ u
£U
U<
£:£i:£ uuuu
:£i uu
i<
69
:£ u
Electrochemical Oxidations
The important role played by phenolic coupling in morphine alkaloid biosynthesis is described In Chapter 1. As previously discussed
'\
££ uu
:£<
o.... c:
:I: o > '1.0 :I::I: uu
.
£U
U
££££ uuuu
:££ uu
206
5
The Morphinans
V
(Section IV,C), attempts to carry out these phenolic couplings in the laboratory have been only partially successful due to the susceptibility of the products to further oxidation. As an alternative approach to the use of inorganic oxidants, controlled potential electrochemical oxidations have been investigated. This electrochemical method ultimately led to the synthesis of the morphinan ring system in virtually quantitative yield. Initial attempts to oxidize reticuline (66) (R = CH3) were unsuccessful. However, protection of amine nitrogen as a carbamate (66) (R = C02CH2CH3) and anodic oxidation resulted in the isolation of the morphinandienone (67) in 15.5% yield (82). The major breakthrough occurred with the discovery that it was not necessary and, indeed, was undesirable to have free phenolic groups in the benzylisoquinoline. For example, the tetramethoxybenzylisoquinoline laudanosine (57) could be electrochemically oxidized to the morphinandienone (59) in yields of up to 93% (Table 5-3), compared to 15% for the free phenol (83). The conditions for the electrolysis appear to be critical, with the best being anodic oxidation at 1.0-1.2 volts in acetonitrile and fluoroboric acid. The regiochemistry of the cyclization is the same as that for the inorganic oxidant, with oxygenation at the 2,3-positions of the morphinan molecule rather than at the 3,4-position characteristic of the majority of the morphine and morphinan-based alkaloids. A variety of attempts to force the 3,4-oxygenation pattern has been uniformly unsuccessful.
Structure-Activity
Relationships
207
of the Morphinans
Table 5-4 Comparative
Analgesic Effects of Morphinan Enantiomers
Enantiomeric Composition
R
(=) (-) (+) (=) (-) (+) Q
A.
H (racemorphan) H (levorphan) H (de'drorphan) CH) (racemethorphan) CH) (levomethorphan) CH] (dextromethorphan)
Analgesic PotencyII' 2-2.5 5 0.2 0.5
Morphine = 1, SC, in rats .
Alteration or Existing Functional Groups and Structures
I. Enantiomers One of the initial syntheses of the morphinan ring system was the preparation of the 3-phenol (37) (R = CH3) (90.). This compound, racemorphan, is a potent analgesIc, bemg about two times as active as morphine (Table 5-4) (91). In contrast to morphme, which exerts its analgesic action orally only in relatively high doses,. racemorphan IS active in humans both orally and parenterally at eqUivalent doses of 1-2 mg. Resolution of the racemic compound into its enantiomers demonstrated that the analgesic activity resided exclusively in the levorotatory isomer levorphan (levorphanol). The dextrorotatory isomer (dextrorphan) was inactive as an analgesic. The side effects of levorphanol and racemorphan are similar to those of morphine (92). However, dextrorphan, although inactive as an analgesic, p~ssesses cough-suppressant properties and is free from the side effects mherent 10 morphme and levorphanol. As a general rule in the morphinan series, the de~trorota~ory isomers have antitussive properties. In common with the relative acllvltles of morphine and codeine, racemethorphan (Table 5-4) has about 10% of the activity of racemorphan (91) and is more potent as an antitussive agent. Optical resolution again has demonstrated that the analgesIc effects reside 10 the levo-isomer, levomethorphan. The potent antItussive effect resides. 10 the dextro-isomer, dextromethorphan, which is free from the opIate side
V. Structure-Activity Relationships of the Morphinans When research into the analgesic properties of the morphinans became important, extensive structure-activity relationship correlations in the morphine alkaloid series were already available. Since the morphinans were similar to morphine, lacking only the ether bridge, the assumption was made that substitutions that were effective in the morphine series could be transferred directly to the morphinans. As a result, a great deal of the research interest in the morphinans has been in the substitution of the bridge nitrogen. As with morphine, the objective was to prepare either mixed agonist-antagonists or pure antagonists that were analgesics without the undesirable side effects of morphine. A major difference between the two series is that the morphinans are usually made by total synthesis, readily furnishing both enantiomers by resolution techniques and facilitating biological differentiation. This has led to the discovery of potent antitussive agents that lack the usual opiate side effects.
i '\
L
208
5
The Morphinans
Table 5.5 Analgesic
Structure-Activity
Relationships
209
of the Morphinans
Table 5-6 Activity
of Isomorphinans
Enantiomeric Composition
R
(-) (+) (-) (+)
CH, CH, CH, CH)
(-)'
CH,-<]
The Effect of Substituents at C-3 on Morphinan Analgesia
Other Substitution
Analgesic Potency"
t:,' t:,'
Enantiomeric Composition
Inactive at 32 mpk
(-) (-)
-7-8 0.25
(=) (=)
Inactive"
(=) (=)
8-10
Q
Relative to morphine = 1; tail flick test in rats. b Antagonist potency relative to nalorphine = 7. " Agonist activity is less than 1/300th that of cyc10rphan in the PSQ
writhing
V
test.
effects demonstrable in levomethorphan (93-95). Dextromethorphan is wIdely used as a cough suppressant. Similar separations of activities have been observed with a variety of N-substitutions (96). 2. Ring BC trans-Morphinans (Isomorphinans) A substantial amount of effort has gone into the synthesis of both trans-codeine and trans:morphine where the usual cis-decalin stereochemistry at the BC ring JunctlO~ has been Inverted to the trans-decalin stereochemistry (97). The analgesIc potencIes of both trans-codeine and trans-morphine have both been disappointing, bei~g 0.5 that of codeine and 0.1 that of morphine, respeclIvely. The work In the pentacyclic series has been based on data ~btained originally in the morphinan series. Although the Grewe cyclizalIon can be .modlfied to p~oduce substantial amounts of isomorphinans (60), the onglnal preparatIon of the isomorphinan ring system was an outgrowth of the Gates total synthesis of morphine (57,58). The racemic mIxture was resolved using tartaric acid to furnish the pure enantiomers. As c?n be seen In Table 5-5, levo-enantiomers retain analgesic activity. The Isomorphlnan analog of levorphan (Table 5-4) is twice as potent as levorphan and 10 tImes as potent an analgesic as morphine (98). Interestingly, the dextroisomorphinan containing unsaturation at C-6 apparently retains some Inherent analgesic potency. The isomorphinan analog of cyclorphan (R = cyclopropylmethyl), a potent mixed agonist-antagonist,
(=) (=) (-) a Relative
R
Analgesic Potency.
References
H (2t) HO (37) CH,CO, CH,O CH,CH,O CH,CH,CH,O CH,=CH,CH,O
0.5 5 Strong Strong Strong Weak-inactive Weak-inactive
52,101 92 34,90 34,52,90 54,102a 34,102a 34,50
C6:f{sCH2O
Weak-inactive 0.5
34,50 102b
C6HsO to morphine = 1.
is essentially without any analgesic potency (60,99). 11appears that in the agonist isomorphinan series, analgesic potency is increased over that of the equivalent morphinan. In the mixed agonist-antagonist series, the effect is still unclear. 3. The Aromatic Ring and the Phenolic Function at C-3 The presence of a phenolic 3-hydroxyl group and its ethers and esters in the morphinans has been studied for its effect on analgesia as well as for ItS cough suppression. On the whole, the structure-activity relationships follow the observations made for the morphine alkaloids. For instance, the 3hydroxyl group can be readily removed using the Musliner-Gates ~eagent (/00) to furnish the 3-deoxy compound (21), which has approxImately VlOth the potency of levorphan (37) (Table 5-6) (52,lOI). Acetylation of the 3-hydroxyl group yields a compound with equivalent analgesic effects but with a shortened duration of action (90). Formation of the methyl ether results in a reduction of analgesic activity (52,90), while longer chain ethers rapidly cause the loss of analgesia. The allyl and propyl ethers are essentially inactive (50). The phenoxy ether is 0.5 times as potent an analgesic as morphine, with a longer duration of action and decreased physical dependence liability (102b,c). The phenoxy ether may functIOn as a prodrug for levorphanol (37) (lOld).
210
5
V
The Morphinans
Structure-Activity
Relationships
Table 5-7
The antitussive activity of the dextro-morphinan isomers approximately parallels the an~lgeslc potencies in the 3-ethers, with the methyl (dextromethorphan) beIng the most potent. Antitussive activity decreases as chain length mcreases (103,104). [nterestingly, a 3-methyl substituent, which is stencally more bulky than either a hydroxyl or a methox y l grou p d oes no t .. ' d ecrease t h e antItussIve activity (105). [n the ~ripavine Diels-Alder adducts, it is not necessary to have an ~romatlc rmg present for analgesic activity to occur. This effect has been m.vewgated t.n the morphinan series by Birch reduction. Reduction of 37 wIth lithIUm 10 ammOnIa gIves the nonaromatic morphinan 3-ketone (71) after hydrolysIs of the enol ether (70). Similarly, the isomorphinan (72) gIves the BC-trans 3-ketomorphinan (74)' (106). Neither of these compound~ has demonstrated any analgesic activity in the rat tail flick assay, m~'catmg ~he necessity of an aromatic ring in the morphinan and ~somorphman senes.
Analgesic Potency of 6-Ketomorphinans with Single A-Ring Substitutions
1 2
Substitution
U
(BC-cis)
70
72
(BC-trans)
73
Analgesic
Hot plate activity,
ActivityU
2.6 0.5 0.8 Inactive 0.3 0.05 0.1 0.7 1.0 3.2 morphine
= 1.
Iy great analgesic activity (107). This observation indicates that the usual structure-activity relationship inherent in the morphine series does not always hold. To evaluate the importance of a carbonyl group in the 6-position, the 4-hydroxy-6-ketone (75), readily available from either morphine (108) or by total synthesis (109), was deoxygenated to 76 using the Musliner-Gates reagent (100). The 6-ketomorphinan (76) possesses remarkably potent analgesic activity, being three times as potent as morphine in the hot plate assay (101). For comparison purposes, the un substituted compound 21, without the 6-ketone, has about one-half the activity of morphine. Hydroxylation of the 3-position is the most important A-ring substitution in the morphine alkaloids as well as in the morphinans. The 3hydroxy-6-ketomorphinan (77) was originally synthesized by a rather difficult route from dihydrothebainone (8) via the 4-phenyl ether (110,111). A lengthy total synthesis using a modification of the Gates morphine synthesis yielded racemic 77 (112). Subsequently, use of the Grewe morphinan synthesis allowed 77 to be obtained from readily available starting materials in six steps (113). The methyl ether (78) is readily obtained either directly from 77 or as an intermediate in the synthesis. Neither 77 nor 78 is as potent as morphine (Table 5-7), but surprisingly, the methyl ether is significantly better as an analgesic.
71 74
>
75
Pattern
None (76) 3.0H (77) 3.0CH, (78) 1.OH (80) l-OCH, (82) 2-0H (83) 2-0CH, (84) 4-0H (75) 4.02C:CH, (86) 4.0CH, (85)
>
37
211
of the Morphinans
76
4. The 6-P~sition in the Morphinans The 6-ketone substituted morphmans are an mterestmg group of molecules because of their unexpected-
l
212
5 The Morphinans
a
V
Structure-Activity
~b CH3~CIl3
r
15
/,,\
c
'\ /
\\
d,e .>
CH)d
'0 C6HS
OCH1
CH_O
a
7
/
bC"Hc
\\ 0
,
I
f,c
of the Morphinans
a
~)
~~~CH)~O'/~CH) 12
Relationships
213
(OM
b
)
81
f#o0
N-CHO rI\
C"H~CH.,O.
In~
HO c
\
0
d,e
~\\ 0
H~3
-. f \\ o 80 82 Scheme 5-8. Synthesis of I-hydroxy-6-ketomorphinan. Reagents: (a) Fremy's salt, reduction; (b) benzylbromide; (c) 5-chloro-l-phenyl[lH]tetrazolc, H2/Pd; (d) potassium hydroxide; (e) HCHO, hydride, phenyltrimethylammonium mcthoxide.
79 Scheme 5-7. Synthesis of (- )-3-hydroxy-6-ketoisomorphinan from thebaine. Reagents: (a) Na/NH3; (b) C6HSBr, pyridine, Cu; (c) 5% hydrochloric acid; (d) H2,Pd/C; (e) deketalization, (f) potassium hydroxide, triethylene glycol.
The isomorphinan derivative (79) of 77 has also been prepared in a lengthy sequence starting from thebaine (12). The synthetic sequence leading to the isomorphinan is outlined in Scheme 5-7, which gives an idea of the effort that went into the synthesis of some of these molecules (114). The isomorphinan (79) has about 12 times the potency of morphine in the tail flick assay, which makes it about equipotent with the equivalent morphinan (77).
The synthesis of the I-hydroxy-6-ketomorphinan (80) is outlined in Scheme 5-8. The synthesis proceeds from the protected morphinan (81). Oxidation with Fremy's salt, potassium nitrosodisulfonate, gives the p-quinone, which is reduced to the hydroquinone. The I-hydroxyl group is selectively benzylated and the 4-hydroxyl group is removed using the Musliner-Gates reagent. Deformylation followed by N- and a-methylation forms I-hydroxy-80 and I-methoxy-6-ketomorphinan 82 (101). This reaction pathway is necessary because the Grewe cycIization gives only abnormal products (62). In the hot plate assay, the free phenol (80) is inactive, while its methyl ether (82) has about one-third of the activity of morphine. The 2-hydroxy-6-ketomorphinan (83) and its O-methyl ether (84) have been prepared from 3-methoxyphenylethyl amine and 3methoxyphenylacetic acid using the standard Grewe cycIization. The analgesic potency is disappointing, being only 5 and 10% of morphine for 83 apd 84, respectively (115). The 4-hydroxy-75 and 4-methoxy-6-keto-
214
5 '['he Morphinans
morphinan 85 are the most surprising monosubstituted 6-keto compounds to come out of this series. The requisite 4-hydroxy compound (75) is readily prepared either by total synthesis (109,116) or from morphine (108,117,118). Methylation or acetylation by standard methods furnishes the O-methyl ether (85) and the acetyl ester (86). The phenol (75) and its acetyl derivative (86) are approximately equipotent to morphine. However, the 4-methoxy compound (85) has over three times the activity of morphine (118,119).
V
Structure-Activity
Analgesic Potency of ~,4-Disubstituted 6-Ketomorphinans
92 93 94 89 91 8 90
HO B7
(R1=H,
R2=OH)
(R1=OH,
R2=H)
95
The requirement of a 6-ketone for strong analgesia in this s~ries is demonstrable not only by the difference in analgesic potency with the 6-deoxy compound (21) but also with the epimeric 6-hydroxyl compounds 87 and 88. The 613-aIcohol (87) has 7% of the analgesic activity of the 6-ketone (75), while the 6a-aIcohol (88) has 8% (119). Opiate receptor affinity for 75 is equivalent to that of morphine, while the methyl ether (85) has about one-third the affinity of morphine (119). Th~ 3,4-disubstituted 6-ketomorphinans, although more complex than the singly substituted analogs, were prepared much earlier by reductive conversion of codeine to dihydrothebainone (8). The derivatives of 8 were extensively investigated in Japan after 8 was found to retain 40% of the analgesic potency of morphine (Table 5-8) (120). a-Methylation of 8 ~iv~s the dimethoxy compound 89, which at 2.6 times the potency of morphIne IS the most interesting compound in this series (119,120). By simple chain extension of 89 to the 4-ethyl compound (90), half of the analgesic potency is lost, while bridging with a methylenedioxy group (91) effectively eliminates analgesia (117). On the other hand, the diphenol (92) has one-third the potency of isomer 93. The increased lipophilicity at C-4 increases the analgesic activity lO-fold on going from the 4-phenol (92) to the 4-0-ethyl ether (94). A substantial number of disubstituted 6-ketomorphinans containing a 1413-hydroxyl group have been prepared and are discussed in a subsequent section on 1413-substitutions.
215
of the Morphinans
Table 5.8
Compound
BB
Relationships
a
Rt
Rz
Analgesic Potency"
OH OH OH OCH3
OH OCH3 OCzI-I5 OCH3
0.16 0.5 1.5 2.6 0.14 0.4 1.5
OCHzO OH OCZH5
OCH3 OCH3
Hot plate test, sc;.morphine
= I.
5. Substitutions at Nitrogen The substitution of other organic residues for the methyl group in morphinans follows broadly the structure-activity relationships observed in the morphine alkaloid series. The effect ?f various alkyl, arylalkyl, and some functionalized derivatives is presented In Table 5-9. The nor-compound, which contains a secondary amine, is very similar to normorphine in terms of centrally mediated analgesia (121). The substantially increased polarity in these types of compounds apparently prevents efficient crossing of the blood-brain barrier. As the c.hain !e.ngt.h is increased from the methyl group in levorphanol (37), analgesIc activity IS rapidly lost, being nonexistent with n-propyl (122,123) and rapidly rising again as the chain length is further increased to n-pentyl ~~4,96). Bra?ching of the alkyl side chain does not bring about any defimtlve change In analgesic potency. The absence of analgesic activity with an. n-propyl substitution is also found in morphine, since the compound functIons as an opiate antagonist (122,123). The n-propyl derivative is a potent antagonist of morphine analgesia in the rat but a weak antagonist of respiratory depression in the rabbit (122). The cyanoalkyl substit~ent has been referred to as paradoxical, increasing potency severalfold m some classes of opiates and slightly or not at all in others. In the morphinans, :he cyanomethylene decreases analgesia somewhat. The cyanoethylene denvative, however, has lO-fold the analgesic potency of levorphanol and 47-fold that of morphine (124).
216
5
The Morphinans
V
Table 5-9
H CH3 (levorphanol) ~Hs II-C3H7 n-C4H9 n-CsHII CH2CH2CN CH2CN CH2C6Hs (CH2hC6Hs (CH1hC6Hs (CH2)4C6HS CH2COC6Hs (CH2hCOC6Hs
Relationships
of the Morphinans
217
Table 5-10
The Effect of N-Substitution
R
Structure-Activity
on Morphinan Analgesia
Relative Analgesic Potency.
The Effect of N-Alkenyl, Alkynyl, and Cycloalkylmethylene Substitution on Morphinan Analgesia and Opiate Antagonism
Morphine Antagonism R
O.03b 5 0.2
o
n-C3H7 -CH2CH=CH2 (levallorphan) -CH1C==CH
Yes
2 6 47< 0.6 Inactive 3 0.02 0.13 6.5 0.02
-CH1CH=C(CH3h -CHz--
Relative Analgesic Potency.
o 1.4
Relative Morphine Antagonistic Potencyb
Active 1.3
-1 Inactive
-0.1 1.3 11.4
2
40
o
(cyclorphan)
-CHz-<>
a Mouse writhing assay, morphine b Tail flick antagonism, nalorphine
"b Morphine = 1. Tail flick test. C Hot plate test.
= 1. = I.
--" r"'r{ (.
4;,,,;..J >-
~l",t.J'7).-.tth;
counteract the respiratory depression caused by narcotic overdose (127). However, by itself, in common with its morphine analog nalorphine, levallorphan causes respiratory depression (127). Although a potent antagonist of morphine analgesia (128), it also has a strong analgesic effect, although it is not useful clinically because of its nalorphine-like psychotomimetic side effects (46). The propargyl compound is as effective as nalorphine in antagonizing both morphine and phenyl piperidine opiates (123) and has been used as an analgesic in patients with postoperative pain (123,129). In the morphine series, a dimethylallyl substituent yields a mixed agonist-antagonist, nalmexone. In the morphinan series, the dimethylallyl compound is a potent analgesic causing mild respiratory depression but no narcotic antagonism. Its addictive liability is equivalent to that of morphine (123,130). In the cycloalkylmethylene substitution, the cyclopropylmethylene analog (cyclorphan) (60,99) is a potent antagonist in animals (131) and a surprisingly strong analgesic in animals as well as in humans (132). However, cyclorphan produces a high incidence of nalorphine-like psychotomimetic effects, which precludes its use as an analgesic. The related
The introduction of other oxygen or nitrogen functions into the side chain leads to a substantial reduction in or elimination of analgesic properties. The use of arylalkyl substituents also parallels the observations made in the morphine series; a phenethyl group is the best substituent, having three times the activity of morphine (34,121). Substituents on the aromatic ring can further increase the analgesia. The phenacyl derivative in the morphinan series is substantially more potent than morphine in humans (125) and has substituted for morphine in addiction studies (126). The same compound in the morphine series is essentially inactive. The use of /3,y-unsaturated or cycloalkylmethylene substituents in place of the nitrogen methyl group furnishes the same type of narcotic antagonist analgesic found in the morphine alkaloids. While the n-propyl compound is an antagonist, it has been little studied. The N-allyl compound, levallorphan (Lorfan), on the other hand, is used clinically. Levallorphan is a potent narcotic antagonist (Table 5-10) that is used to
L
, 5 The Morphinans
218
V
Structure-Activity
Relationships
of the Morphinans
219
U
f
Table 5-12
Table 5.11 Mixed Agonist-Antagonist 6-Ketomorphinans
N-Substitution
The Effect or N-Heterocyclic Analgesia
in the
Alkyl Substitution on Morphinan
HO
RJ
Rz
Relative Analgesic Potency"
H CH) CH) CH)
CPM< CPM CBMd Allyl
1.25 0.1 0.7 0.3
Relative Analgesic Potency
R
Relative Opiate Antagonistic potencyb
Morphine Antagonism
50"
a
Inactive 0.1 Inactive Not dose responsive
100
Mouse writhing assay, morphine = I.
b Morphine tail flick antagonism, < Cyclopropylmethylene. d Cyclobutylmethylene.
naloxone
0.25<
= I.
cyclobutylmethylene morphinan, on the other hand, appears to be a pure agonist without any antagonist properties (133). The effect of nitrogen substituents known to convert agonists to mixed agonist-antagonists or pure antagonists has been studied in the 4-hydroxyand 4-methoxy-6-ketomorphinan series (l18,134). The phenolic N-cyclopropylmethyl derivative (Table 5-11) has potent analgesic activity without narcotic antagonist properties (118). Its methyl ether has lflOththe activity of both morphine and naloxone (134). The cyclobutylmethylene derivative is similar to the 3-hydroxy compound (Table 5-10) in that it is a pure agonist (107,134). The use of heterocyclic alkyl substitutions has led to extremely strong enhancement of analgesic activity over that observed with the N-methyl group (Table 5-12). The use of thienylethyl and furylethyl substituents gives morphinans 50 to 100 times more potent than morphine (124). Similar, though less potent, enhancements are observed with other heterocycles. The furylmethylene compounds present a striking contrast. Both unsubstituted compounds are weak pure antagonists. However, the addition of a methyl to the furan ring converts both positional isomers into pure agonists with three to five times the potency of morphine (134). An interesting extension of both the phenylethyl and heterocyclic side chains is
{' t!'I'~
5
n1
0.4
3
U
Hot plate test, morphine = 1.
I>Writhing C
test.
Tail clip antagonism, nalorphine = 1.
compound 95 (p. 214), which combines a maleimide substitution on a phenylethyl side chain (135). Compound 95 is one of a series prepared in an effort to obtain a receptor agonist, containing a Michael acceptor, that bonds covalently to the opiate receptor. Although 95 possesses onefifth the activity of morphine, it is also an antagonist. However, 95 does not bond to the receptor in a covalent manner (135).
L
220
5 The Morphinans
V
The effect of variation of N-substitution in the presence of a 1413hydroxyl group is discussed in Section V,B,3. B.
Structure-Activity
Relationships
221
of the Morphinans
Table 5-13 The Effect of A-Ring Substituent
Positions on Morphinan Analgesia
The Effect of Substituents in Nonfunctionalized Positions on Morphinan Analgesia
Since a substantial amount of the structure-activity relationship of morphine was known before the Grewe synthesis made a wide variety of morphinans available, the substituents introduced into the various positions of the morphinan ring have reflected the morphine structure-activity relationship. As a result, less synthetic effort has been expended but more judiciously targeted molecules have been made.
Rl
RI
1. Aromatic A -Ring Substituents and Their Influence on Morphinan Biological Activity An initial indication of the influence of A-ring substituents on the analgesic activity of 6-keto-morphinans was described in Section V,A,3. The broad trend indicated there, where the 3-phenol was the most potent, followed by the 4-, 2-, and I-phenols, has been confirmed in a study where substituents were introduced in both the agonist and antagonist morphinan series (136). In addition, the biological activities have been correlated with opiate receptor affinity (Table 5-13). As expected, the 3-hydroxymorphinan in each N-substituted series is by far the most potent with respect to receptor binding as well as pharmacological activity. Displacement of the phenolic hydroxyl group to the 2- or 4-position decreases opiate receptor binding affinity by 30- to 100-fold. However, the retained affinity is quite significant. It is interesting that the deoxy analog of cycIorphan demonstrates significant binding affinity, albeit 500 times weaker than that of cycIorphan. Alkylating the hydroxyl group by methyl ether formation reduces receptor affinity at least 100-fold compared to that of the parent phenol (42,137). It is apparent that the correlation between binding affinity and pharmacological, analgesic, and opiate antagonistic potency is excellent. The only exceptions are the methyl ethers, which in all cases are considerably more potent analgesics than would be expected from their binding affinities. This is probably due to a partial metabolic de methylation in vivo, which is similar to that of codeine (138). Earlier work indicated that a I-methyl or a 2-methyl group in racemorphan yields analgesics with 0.5 and 1.7 times the activity of the unsubstituted parent, respectively (37, R = H) C!!...)' LI~W3~) 2. The Introduction of Various Substituents at Positions 5 to 10 Metopon, 5-methyldihydromorphinone, is considered to be one of the best morphine-based analgesics developed in terms of potency without side effects (139). An intermediate in the synthesis of metopon (97), which is
3-0H (Ievorphanol) 2-0H 4-0H 3-0CH3 (levomethorphan) 2-0CH3 3-0H (levallorphan) 2-0H 3-0H (cyc1orphan) 2-0H 4-0H H 3-0CH3 2-0CH3
R2
Relative Analgesic Potency
Relative Antagonist Potency
Opiate Receptor Affinity (nm) 4
CH3 CH3 CH3 CH3
0.02 0.2 0.5
400 60 2,000
CH3 Allyl
<0.001 0.34
> 10,000 4
Allyl CPM"
Inactive 2.5
0.02 1.5
CPM CPM CPM CPM CPM
0.1 0.08 0.2 0.8 rnactive
0.04 0.02 <0.05
004 30 35 200 300 10,000
cyc1opropylmethylene " CPM =
readily prepared by the reaction of dihydrothebaine (96) with methyl grignard, has the requisite 513-methyl group (140). O-Methylation gives the 513-methyl-6-ketomorphinan derivative (98) (Scheme 5-9), which has 1.5 times the activity of morphine (141). The analgesic activity is comparable to that of metopon, but the pharmacological profile of 98 has not been described. In contrast to the reaction of dihydrothebaine with grignard reagents, the addition of lithium dimethylcuprate to thebaine results in the stereospecific introduction of a 713-methyl group together with epoxide ring opening (142). Controlled hydrolysis then leads to either the morphinan or isomorphinan series (Scheme 5-10) (142). Reduction of the enone double
L
222
V
5 The Morphinans
Structure-Activity
Relationships
223
of the Morphinans
Table 5-14 The Effect ofa 7-Methyl lsomorphinan Analgesia
a
96
b
Group
on 6-Ketomorphinan
and
97
).
98 Scheme 5-9. Synthesis of a 5-methylmorphinan. Reagents: trimethylammonium chloride.
a
(a) MeMgI;
(b) phenyl-
>
H)
12
/
le.
CH) CH) Scheme 5-10. . dlmethylcuprate;
bond is stereospecific in both series, forming the 7a-methylmorphinan and the 713-methylisomorphinan. In order to prepare the 3-oxygenated compounds, the 4-hydroxyl group is reductively removed. The nitrogen substitution is varied by preparing the nor-compound. The morphinan and isomorphinans containing either a 3-hydroxyl or a methoxyl substituent in both the agonist and antagonist series are potent analgesic agents (Table 5-14) (143). In general, the introduction of a 7-methyl group does not substantially alter the analgesic potency of the parent unsubstituted
H) CH)O
Reaction of thebaine (b) aqueous acetic acid;
with methylcuprate. (c) 5% hydrochloric
Reagents: acid.
(a)
lithium
L
Table 5-15 224
5 Tli'e Morphinans
morphinan, nor does it significantly that of morphine. The reason for the is that 3,4-disubstituted morphinans analgesic potency of morphine in morphine antagonism (144).
7-Carbethoxy-6-ketomorphinans
and -isomorphinans
increase the analgesic strength over interest in a single substituent at C-3 99 and 100 have only 0.2 times the the writhing assay and very weak Isomorphinan
Rz
Relative Analgesic Potency"
Relative Opiate Antagonist Potencyb
Morphinan H H H CH) CH) CH)
CH) CPM' CBMd CH) CPM CBM
1.0 0.2 0.7 0.2 <0.1' <0.1'
Not tested 0.3 <0.1' Not tested <0.3' <0.3'
Isomorphinan H H CH) CH)
CH) CPM CH) CPM
1.3 0.2 <0.1' <0.1'
Not tested 0.3 Not tested <0.1'
RI
99 100
(R=H) (R=CH3)
In a simple model of the thebaine Diels-Alder adducts, a diastereomeric 7-carbethoxy group has been introduced into various N-substituted 6-ketomorphinans and isomorphinans. In contrast to the tremendous potency enhancements with the Diels-Alder adducts, the results with the morphinans have been disappointing (Table 5-15). The analgesic potencies vary from weak to approximately that of morphine, while the antagonist potencies are all substantially less than that of nalorphine (145). In an extension of this work, a series of compounds has been prepared in which a 6-double bond in the isomorphinan residue generates a molecular volume similar to that of the Diels-Alder adducts (146). Insertion of a substituted 7-alkyl group would prepare a simpler analog of the Diels-Alder adducts. Of the compounds prepared (Table 5-16), the target compound 101, containing a secondary pentyl alcohol, is 130 times as potent as morphine (146). In contrast, however, the BC-cis morphinan analog is inactive. The 8-alkylmorphinans are readily prepared from morphinan 102, which is easily prepared from thebaine. Hydrolysis of the enol ether yields the morphinan enone (103) (Scheme 5-11), which undergoes organocuprate addition to stereospecifically introduce the 8f3-alkyl substituent (104) (147). An extensive series of these compounds has been prepared with a 3-oxygen group and varying nitrogen functionality. Table 5-17 shows a representative variety of the compounds and the associated biology. On the whole, analgesia and opiate antagonism decrease with increasing chain length at C-8. Significantly, none of the 8-alkyl derivatives is as active as its unsubstituted parent (147). However, the biology is sufficiently encouraging to cause investigators to study the effect of 6-ketone modification (148). The most interesting compound is 106, which has potent mixed agonist-
U Mouse writhing test, morphine = I. b Rat tail ftick antagonism, nalorphine = 1. e Cyc1opropylmethylene. d Cyc1obutylmethylene. , Calculated from the highest dose tested.
Table 5.16 Analgesic Potencies of 7-Substituted 6,7-Didehydroisomorphinans
l
R,
Rz
CH) CH) CH) H CH)
CH) CH) CH) CH) CPMb
Relative Analgesic Potency"
R) CHO CH(OH)(CHzhCH) CO(CHzhCH) CO(CHzhCH) CH(OH)(CHzhCH3
I. "b Mouse writhing test, morphine = CPM = cyc1opropylmethylene.
(101)
0.2 130 1 17 4
'
(
Ir~1f
Ilf6
J
V a '>
102
103
b
>
Scheme 5-11.
Structure-Activity
Relationships
of the Morphinans
227
antagonist properties. Compound 106 is four times as potent as morphine in the mouse writhing assay and equivalent to nalorphine as an antagonist. In contrast to the 6-ketone, 106 containing a methylene group in place of the 6-ketone does not substitute for morphine in drug-dependent monkeys (148). A series of 813-alkyl-7a-methylmorphinan 6-ketones has been synthesized by organocuprate addition to the 7-methyl analogs of 103. In general, the 7,8-disubstitution does not offer an advantage over the unsubstituted parents either in the agonist or the antagonist series (143). Early work on the morphinans, stemming mainly from the Gates total synthesis, allowed access to the lO-hydroxy (107) and lO-keto (108) morphinans. These compounds do not possess any analgesic activity (34).
104 Reagents: (a) 25% hydrochloric acid, 100°C; (b) lithium dimethylcuprate.
Table 5-17 8-Alkyl Substituents on Morphinan 6-Ketones
106
R1
Rl H H H CH3 CH3 CH3 H H H CH3 CH3 CH3 a
R2 CH3 CH3 CH3 CH3 CH3 CH3 CBMe CBM CBM CBM CBM CBM
R3 CH3 CH2CH3 (CH2)3CH3 CH3 CH2CH3 (CH2)3CH3 CH3 CH2CH3 (CH2)3CH3 CH3 CH2CH3 (CH2)3CH3
Mouse writhing test, morphine
Relative Analgesic Potency"
Relative Opiate Antagonist Potencyh
0.8 0.9 Inactive 1.6 0.65 0.1 7 4 2 1.5 0.15 0.1
Not tested Not tested Not tested Not tested Not tested NO! tested 0.9 3 0.3 0.2 Inactive Inactive
= 1. b Rat tail flick antagonism, nalorphine = 1. e CBM = cyclobutylmethylene.
107
(R=OH)
lOB
(R= =0)
3. The Importance of a Hydroxyl Group at C-14 In the morphinebased analgesics, the presence of a 1413-hydroxyl group usually increases the analgesic potency and decreases the opiate side effects. Morphinan analogs of all the important morphine derivatives have been prepared by total synthesis. In the agonist series, the 3-hydroxy derivative has five times the activity of morphine (Table 5-18) (149). Shifting the phenolic hydroxyl to position 4 retains the activity but, as in common with the morphinans, the methyl ether is still potent (150). The effects of alkyl chain length on nitrogen parallel the observations made in the morphine series. As a result, a series of alkyl and allyl substituents show mixed agonist-antagonist activities. The most important of these is the cyclobutylmethylene derivative butorphanol (64), which has both significant analgesic actions and antagonistic activity comparable to that of naloxone (151,152). Butorphanol has a low potential for physical dependence in both humans and animals (151,153). Like many of the mixed agonist-antagonists, butorphanol produces some degree of dysphoria and psychotomimetic effects (154). In contrast to other alkyl substituents, a cyclopropylmethylene group results in a relatively clean antagonist, oxilorphan (64,155). The presence of a 14-hydroxyl group increases the potency in antagonists and decreases the intensity of the unpleasant disorienting side effects. However, the
228
5 The Morphinans
V
Structure-Activity
Relationships
of the Morphinans
229
Table 5-18 The Analgesic and Antagonistic and Isomorphinans
Activities of 14-Hydroxy Substituted
Morphinans
a
47
)
109
Rl
1
Morphinan
Isornorphinan
Relative Analgesic Potency.
Ethyl Propyl Allyl 3,3-Dimethylallyl Propargyl
Isomorphinan 3-0H 3-0H 3-0H 3-0I-!
CH3 CPM CBM Allyl
h e
0.02 5 Ie Ie 0.04
c
<0.01 <0.01 110
Inactive 4.5
Scheme 5-12. Reagents: (a) trifluoroacetic sodium t-pentoxide in hot benzene.
of oxymorphone-induced
Hot plate assay, morphine = 1. eydopropylmethyl. ," CPM = CBM = cydobutylmethyl.
anhydride;
(b) m-chloroperbenzoic
acid; (c)
10
CBM'
0.8 0.01 0.01 5 0.01
0.03 0.08 0.2 <0.01 0.1
0.1 0.05 0.07 <0.01
<0.01 0.02 0.01 0.02
Since these 14-hydroxymorphinans are prepared by a total synthesis that has significant flexibility, it is possible to prepare the 14a-hydroxyisomorphinans readily (Scheme 5-12) (158). The key to the synthetic access to the 14a-hydroxy compounds is the stereospecific epoxidation of the hydrophenanthrene (109), which is readily accessible from the common intermediate (47) (cf. Scheme 5-5). Ring closure then forms the isomorphinan. None of the isomorphinan derivatives has significant analgesic or antagonistic activity in comparison with the epimeric morphinan (149). The introduction of a 6-ketone into the 14-hydroxy morphinan series proceeds by reduction of 14-hydroxydihydrothebainone (159). A large series of 3,4-disubstituted-6-ketomorphinans has been prepared (160). The monosubstituted compounds are obtained by suitable reductive deoxygenation from the disubstituted compounds (107,159). The presence of the 6-ketone tends to increase the analgesic potency over that of the unsubstituted parent (Table 5-19). For instance, the A-ring unsubstituted compound is twice as potent as morphine (107). In contrast to oxilorphan, the 6-keto derivative and its O-methyl ether have potent anti nociceptive activity (161-163). Shifting the methoxyl group to the 4-position also increases the agonist activity. Surprisingly, a series of 4-methoxy-6-ketomorphinans containing allyl and cyclopropylmethyl nitrogen substituents lack either agonist or antagonist properties (150). The 3,4-disubstituted
Mouse writhing test, morphine = I. Antagonism
-
b
----
Morphinan 3-0H 3-0H 4-0H 4-0CH3 3-0H (oxilorphan) 3-0H (butorphanol) 3-0H 3-0H 3-0H 3-0H 3-0H
.
Relative Opiate Antagonist potencyb
Straub
tail,
naloxone
=
I.
presence of an oxygen bridge in morphine and its analogs promotes metabolic deactivation at the 3-position. In human studies, oxilorphan has been' found to be very long-acting and to have only one-eighth the side effect activity of the benzomorphan cyclazocine (133,156). However, disturbing psychotomimetic effects have been reported with a high subcutaneous dose of oxilorphan (157). I
L
230
V
5 The Morphinans
Structure-Activity
Relationships
231
of the Morphinans
Table 5-19 Analgesic
Potencies
of 14-Hydroxy-6-ketomorphinans
", COCH)
111
112
Relative Opiate R[
H OCH] OH OCH] H H H H OCH] OCH] OCH] OCH]
.
R]
R2
H H H H OH OCH] OCH] OCH] OH OCH] OCH] OCH]
CH] CH] CPM< CPM CH] CH] AJlyl CPM CH] CH] CH] (6a-OH) CH] (6,B-OH)
Hot plate test, morphine b Naloxone = 1.
<
Relative Analgesic Potency.
Antagonistic
2 3 0.2 2.7 1.5 5.4 Inactive Inactive 0.9~
24
@Sf)
Potencyb
}Z::;
r~F1
)
[..f
J0}-}V3
) d..'D
0.4 0.7
= I.
CPM = cyclopropylmethylene.
c?mpounds are ~otent analgesics, and a large number have been synthesIzed (!20): The Importance of the 6-ketone is evident when compared to the eplmenc 6-hydroxyl compounds (Table 5-19). The alcohols possess only 2-3% of the analgesic potency of the 6-ketone (120). C.
Ring Additions, Contractions, Enlargements, Movements in the Morphinans
and
The Diels-~Ider adducts of thebaine have led to extremely potent . morphme agoms~s and an.tago~ists (Chapter 3, Section III). Surprisingly, however, there IS very lIttle mformation on the Diels-Alder addition reactions of ,B-dihydrothebaine (16) and dienophiles. The reaction of 16 ~ith methyl vinyl ketone yields the adduct (111) (164). Compound 16 and ItS 3-acetate react with acetylenic dienophiles, via a zwitterionic intermediate, t? for~ hydr~phena?threnes. rather than Diels-Alder ad ducts (165). DespIte thIs paucIty of mformatlOn, an efficient synthesis of an analog
'''C-(CH) 2 2 C 6 H5 JH
~IJ
I SO 0.01 0.03
~H)
)-c'1
(?)
CH)O
OCH) 11)
120
(R=CPM)
121
(R=CBM)
(112) of 16, in which the nitrogen has been translocated from C-l7 to C-16, has been reported (166). The analog (112) has subsequently been converted into adduct 113 by Diels-Alder addition followed by grignard addition (166). There have been no reports of biological testing for any of these adducts. A 5,6-benzo-annulated analog of racemorphan has been prepared, as illustrated in Scheme 5-13. Condensation of the naphthylethylamine (114) with a glycidic ester (115) furnishes the Grewe intermediate (116), which cannot be cyclized. After N-methylation, 117 is converted to the morphinan (118) (167). The cyclopropylmethylene derivative is further converted to the 5,6-cyclohexano compound (119). These analogs, when compared to the mixed agonist-antagonist pentazocine, show that the N-methyl compound (118) has two times pentazocine's analgesic activity, while 119 has 0.2 times its analgesic activity but equivalent antagonistic activity (167). A series of 6,7-substituted l4-hydroxymorphinans has been synthesized according to the method shown in Scheme 5-5. Of the CE-trans cyclohexano derivatives, neither the cyclopropylmethylene (120) nor the cyclobutylmethylene (121) compound possesses analgesic activity, while 120 is equivalent to naloxone as an antagonist. Compound 121 is inactive as an antagonist (168).
232
5 The Morphinans
V
Structure-Activity
Relationships
of the Morphinans
233
OH
+ a,b
+
a,b
115
cJ> ~ '1;-
114
/ie'it @ £$ c;J..We; ferCt..rk.
115
c
) Mil I
)
iPB
J
4),3+
,
()Cj5/
-r.et-.J1.I, 1/.:<\ {I'H 118 116
(R=H)
117
(R=CH3)
d,e
HO
122 Scheme 5-14. Reagents: (a) aqueous reaction; (c) hydrogen bromjde.
hydrochloric
acid, heat;
(b) Eschweiler-Clark
)
119 Scheme 5-13. Reagents: (a) aqueous hydrochloric acid, heat; (b) Eschweilcr-Clark r~action; (c). 85% phosph~ric acid, 150°C; (d) conversion to N-cyclopropylmethylene; (e) Birch reductIOn and selective catalytic hydrogenation.
Contraction of the C-ring has been accomplished by using cyclopentenylethylamine in place of the corresponding cyclohexenylethylamine in the Grewe synthesis, as shown in Scheme 5-14 (169). The resulting C-normorphinan (122) as the racemic mixture lacks analgesic activity. The optically active compound has been synthesized in a long sequence from 7-oxodihydrothebainone, but no biological activity has been reported (170). The 6-methyl-C-nor (123) and exo-6-methylene-C-nor (124) compounds have also been synthesized. Compound 123 is reported to be 19 tLmes as active as morphine in the hot plate test (170). --The D-normorphinans can be synthesized by the appropriate modifica. tIon of the synthesis outlined in Scheme 5-5. The D-normorphinans 125 and 126 are both devoid of analgesic activity and do not interact with the
~1"~C/EIl M>"""'Y \r«;#:J:;q)! 124
(X =
CH3)
(X==CH2)
125
(X
H)
126
(X
OH)
opiate receptor- (171). Because of this lack of activity, interesting postulates have been made about the mode of interaction of the nitrogen at the receptor level (172). An X-ray study of 126 has demonstrated that the nitrogen lone pair is oriented toward the aromatic A-ring, while in the morphinans and benzomorphans the orientation is toward the aliphatic Coring. Nitrogen looe pair orientation has therefore been proposed as the key factor in receptor binding. A mechanism has been proposed whereby the nitrogen lone pair interacts with an electrophilic site on the receptor, followed by stereospecific electron transfer from the ligand. As a result, the ligand itself is oxidized. This type of binding has been termed clastic (172) and has been criticized (173,174). In the benzomorphans, the biological results from rigid analogs in which the nitrogen lone pair is directed either toward or away from the A-ring does not support the theory of clastic binding (175). Nevertheless, investigations continue in this area (176).
234
5
a
The Morphinans
V
Structure-Activity
Relationships
of the Morphinans
235
)
128 127
c
(R
=
132
(R
= OH)
f,g,h
b
a i,q,;
OCHJ)
form the 3-methoxy (129) and 3-hydroxy (130) C-homomorphinans. The equivalent isomorphinans 131 and 132 ar~ similarly pr~pared (179). The homomorphinans 129 and 130 have 0.35 times and 1.5 times the hot plate activity of morphine, respectively, while the homoisomorphinans 131 and 132 had about one-third the potency of the stereoisomeric homomorphinans (180). .. . There have been a few examples of moving one of the termlnt of the piperidine D-ring in morphinans. One of the first .is the movement of the carbon terminus to C-14, which also causes the D-nng to contract. The new ring system has been termed metamorphinan and is readily prepared fr~m thebaine (12), as indicated in Scheme 5-16 (181). Thebaine readily
)
e
131
Thebaine
(12)
)
RO
133
129 (R=CHJ) 1JO (R=H) Scheme 5-15. Reagents: (a) dimethylaminoethylchloride; (b) LiCH2C02Et; (c) p-toluene sulfonic acid; (d) H2,Pt; (e) Ba(OHh, then polyphosphoric acid; (f) ethyl chloroformate; (g) hydrochloric acid; (h) formaldehyde; (i) LAH: (j) H2-PdjC.
The C-homo morphinan (127) has been synthesized from cycJoheptenylethylamine using the procedures shown in Scheme 5-14 (169,177,178). The compound is as active as morphine but is more toxic. A C-homomorphinan in which an additional methylene group has been inserted between the nitrogen and the B-ring has been synthesized as outlined in Scheme 5-15 (179). Starting from a phenylcycJohexanone (128), the tetracycJic ring system is constructed in a nonconvergent linear fashion to
c,d,e
)
HO
134 135 Scheme 5-16. Reagents: (a) SnCI2, hydrochloric acid; (b) H2-Pd/C, 18% hyd~ochloric acid; (c) 2,4-dinitrophenyl chloride; (d) Birch reduction: (e) pyridine hydrochlonde.
L
, 5 The Morphinans
236
V
Structure-Activity
Relationships
rearranges to metathebainone (133); the latter is reduced to 134, which has the BC-rings in a trans relationship. Removal of the phenolic group at C-4 and O-demethylation yields metamorphinan (135), which actually belongs to the isomorphinan series. Despite these rather radical changes, 135 retains 10% of the analgesic potency of morphine in the mouse writhing test (181). Movement of the nitrogen terminus to C-8 leads to the morphinan analog (136), which has been synthesized by a method similar to that shown in Scheme 5-15. Compound 136, however, possesses only weak analgesic activity (182). Simultaneous ring expansion of the D-ring
237
of the Morphinans
CH2---@oCH3
00
d,e)
f
139
H 136
Scheme 5-17. Reagents: (a) p-methoxyphenylmagnesium chloride; (c) H2-Pd/C; (d) CH3I; (e) HrPt; (f) 48% hydrogen bromide reflux.
137
(b) dehydration;
16
morphinan (138) (184). The syntheses of t~ese molec~les have all been similar, using either a phenyl or benzyl substItuted, partIally hydr~genated quinoline or isoquinoline and subjecting it to the Grewe cycl~tlOn. For instance, the 6-azades-N-morphinan (139) has been prepared as Illustrate~ in Scheme 5-17 (185). In this synthesis, a substituted benzyl group IS i~troduced at C-5 in the isoquinoline, which, after reduction to the Grewe intermediate, cyc1izes readily to 139 (185). The 6-aza derivative does not possess analgesic activity (186). .. The 7-azades-N-morphinan (140) has been prepared by additIOn of the. grignard in Scheme 5-17 to the 8-position of an isoquin~line followed by Grewe cyclization (187). The distances between the ammo and hydrox~l functional groups and the quaternary carbon in 140 are the same as I.n racemorphan (37). However, the analog possesses only weak. analgesIc properties (188).. The 8-aza-analog has bee~ prepared ~s shown. I~ Scheme 5-18. A phenyl substituent is introduced mto the apical position of an 8-carboxyperhydroquinoline to yield 141. Ring closure of 141 forms the B-ring of the morphinan (142), which is converted into the 8-azades-Nmorphinan (143) (189). Further investigation has demonstrated that 143 has the same stereochemistry as morphine (BC-cis, CD-trans) (190). As a result, the 3-hydroxy compound (144) has been prep.ared (191). Compound 144 has about 10% of the activity of. morp~me (191). The. 3hydroxy-15-aza- and l~-azades-N-morphi~ans, ~n.keepmg with the earlier observations, do not possess any analgesIc actIVIty (192-195). The most
HO
138 (Des-N-morphinan) 140 and movement of the N-terminus to C-14 forms the derivative (137), which does not have any analgesic action (183). Compound 130, which is the corresponding morphinan derivative of 137, is more potent than morphine (180). The variety of chemical manipulations leading to the addition of rings, ring enlargements, and contractions, as well as the shifting of the D-ring termini, have not led to a substantially improved analgesic or biological profile over that of morphinan itself. On the contrary, most changes have strongly diminished or eliminated the analgesic activity. D.
Movement of the Nitrogen within the Molecular Framework and Heteroatom Insertion
The effect of moving the position of the amino group in morphinans has been extensively studied by Japanese scientists. They have used a convention whereby the compounds are named as aza-derivatives of des-N-
L
238
5 The Morphinans
V
Structure-Activity
Relationships
ZJ9
of the Morphinans
R
a,b >
d,e
c
---+
f
141
R
R 142
143
(R=H)
144 (R=OH) Scheme 5-/8. Reagents: (a) acrylonitrile; (b) ethyl oxalate, NaH; (c) base; (d) H2-PdfC; (e) H2-Pt; (f) Eschweiler-Clark reaction; (g) hydrochloric acid, then polyphosphoric acid; (h) Huang-MinIon.
potent analgesic activity is found in the 9-azades-N-morphinan derivative (145), which is approximately equal to that of morphine but is much more toxic (196). A compound (146) has been prepared that is not only a 16-azades-N-morphinan but has the N-terminus of the O-ring moved from C-9 to C-14 (197). Like most of these derivatives, 146 does not possess analgesic activity (197). The azades-N-morphinans, with the exception of 145, have been disappointing on the basis of their biological activities. However, the biological activity observed in 145 has served as the basis for making aza analogs.
147
(R=CH3)
148
(R=CPM)
Scheme 5-/9. Reagents: (a) ethyl bromoacetate; (b) NaOH; chloride, NaH; (e) lithium aluminum hydride; (f) formaldehyde.
(c) hydrazine; (d) benzyl
A nitrogen heteroatom has been introduced into the morphinan derivative (145) with the expectation that an excellent analgesic activity would be observed without addiction liability (198). Among the several methods for synthesizing the desired compound, 147,. is that. shown in Scheme 5-19 (199). Although the biological results obtained wIth 147 have no.t been reported, the N-cyc1opropylmethy! derivative (~48) has been descnbed as "a fantastic analgesic" (200). The levo-enantlOmer of 148 has approximately five times the analgesic activity of pentazocine in the writhing assay but only 10% of the antagonistic properties of levalorphan (200,201).
240
5 The Morphinans
V
Structure-Activity
Relationships
241
of the Morphinans
a,h,c
HO
150
HO 145
151
146
ld,f
152 149
153
156
A different type of aza-derivative has been obtained by the Beckmann rearrangement of dihydrothebainone (8) to form the azalactam (149), which, however, again does not have any analgesic activity (202). The introduction of an oxygen atom into the C-ring of the morphinan molecule has resulted in some potent analgesic agents, in contrast to the aza series. The synthetic scheme is illustrated in Scheme 5-20. The benzomorphan derivative (ISO) is converted to either the 0'-152 or {3-151 alcohol, depending on which reducing reagent is used. Hydroboration followed by oxidation generates a primary alcohol from the allylic group that is ring closed to pyran in the morphinan (154) or isomorphinan (153) series. The reaction sequences employed lend themselves readily to the introduction of a substituent at C-14 in the morphinan products by allowing the use of various organometallic reagents. The N-substitution has also ~een investigated, with various antagonist and agonist groups being mtroduced. The biological results indicate that the oxamorphinans are all more potent analgesics than their oxaisomorphinan counterparts. Additionally, the introduction of the oxa-atom provides a compound equivalent to the parent compound plus the addition of a 14-hydroxyl group. Cyclopropylmethylene substitution yields mixed agonist-antagonist activities that are increased by 14-alkyl substitution (203-206). The 6-oxa derivative (155), proxorphan, of cyclorphan is 7-10 times as potent as morphine in three rodent models of analgesia and has 0.05 times the antagonistic potency of naloxone (207-209). The compound is more potent and has longer lasting antitussive properties than codeine (210). A compound (156) that contains a thiophene ring in place of the
HO
154 Scheme 5-20. Reagents: (a) sodium borohydride; (b) borane; (c) hydrogen peroxide; (d) methane sulfonyl chloride; (e) diisobutylaJuminum hydride; (f) boron tribromide.
155
aromatic A-ring has been described synthetically, but without any biological results (211). In summary, the introduction of oxygen and nitrogen heteroatoms into the morphinan ring system has been significantly more successful in either retaining or increasing the analgesic properties of the molecules than the movement of position 17 of the nitrogen.
242
VI.
References
5 The Morphinans
The Chemical Anatomy
of the Morphinans 6.
The tet:acyclic structure shown in Fig. 5-1 represents the absolute configuratIOn of the analgesically active isomers of the morphinan series All kno~n ac~ive is~mer~ are levorotatory, the dextrorotatory enantio~ mers bemg either mactlve or much less active. The dextrorotator y compounds, ~owever, usually possess antitussive properties. The morphmans have the same absolute configuration as morphine at carb~n at.oms 9, 13, a~d 14. .On the basis of the structure-activity relatlO?shlps developed m SectIOn Y, trends for analgesic potency in morphmans are the following:
7. 8.
9. 10.
1. The B~-trans ring junction found in the isomorphinans can increase analgesIa severalfold over the BC-cis junction. 2. The pr.esence of a free or acetylated hydroxyl at C-3 is necessary; alkylatIon decreases the analgesic potency severalfold. 3. A 6-ket.one can .strongly affect analgesic activity. The 3-deoxy-6ket?ne. IS t~ree tImes as potent as morphine. While the 3-methoxy denv~tlve ]S about equivalent to morphine, the 4-methoxy and 3,4-dlmethoxy compounds have three times the activity of the equivalent morphines. 4. Nitrogen substituents of the type CH2-X-aryl in which X is methylene, CC?, or CHOH and the aryl residue is thienyl, furyl, or phenyl a.re optImal. The thienyl compound is 50 times and the furyl IS 1O? tI.mes. as P?tent as morphine. The simpler cyanoethylene substItutIOn IS eqUIvalent to furyl. 5. Derivatives containing small aliphatic nitrogen substituents from 3 to 5 carbon atoms, such as allyl or cyclopropylmethylene, are potent
11.
I! \Lx~
furyl
-.'
1. D. H. R. Barton, G. W. Kirby, W. Steglich, G. M. Thomas, A. R. Battersby, T. A. Dobson, and H. Ramuz, J. Chem. Soc. p.2423 (1965). 2. R. A. Konovalova, S. Yu. Yunusov, and A. P. Orechov., Ber. Dtsch. Chem. Ges. 68, 2277 (1935). 3. K. L. Stuart, Chem. Rev. 71, 47 (1971). _ 4. D. J. Slatkin, N. J. Doorenbos, J. E. Knapp, and P. L. Schiff, Jr., J. Pharm. Sci. 61, 1825 (1972). 5. A. F. Cameron and C. Hannaway, J. Chem. Soc., Perkin Trails. I p. 1002 (1973). 6. M. Shamma, P. Chinnasamy, S. F. Hussain, and F. Khan, Phytochemistry 15, 1802 (1976). 7. A. Shafiee, A. Ghanbarpour, I. Lalezari, and S. Lajevardi, J. Nat. Prod. 42,174 (1979). 8. S. R. Hemmingway, J. D. Phillipson, and R. Veepoorte, J. Nat. Prod. 44, 67 (1981). 9. I. R. C. Bick, H. M. Lowe, N. W. Preston, and J. J. Wright, Aust. J. Chem. 26,455 (1973). 10. M. K. Yusupov, Kh. A. Aslanov, and D. T. Ngo, Khim. Prir. Soedin. 11,271 (1975); Chem. Abstr. 84,44473 (1976). 11. A. R. Battersby and T. H. Brown, Chem. Commun. p. 170 (1966). 12. D. Dwuma-Badu. J. S. K. Ayim, S. F. Withers, N. O. Agyemang, A. M. Ateya, M. M. EI.Azizi, J. E. Knapp, D. J. Slatkin, and P. L. Schiff, Jr.. J. Nat. Prod. 43, 123 (1980). 13. J.-S. Hsu, S.-Y. Lo, and J.-H. Chu, Sci. Sill. (Ellgl. Ed.) 13,2016 (1964). 14. I. I. Fadeeva, A. D. Kuzovkov, and T. N. ll'inskaya, Tr. Vses. Nauch-1ssled. [fISt. Lek. Aromat. Rast. 15, 334 (1969). 15. M. Shamma and S. S. Salgar. Phytochemistry 12, 1505 (1973).
/"
CHz
Nitrogen insertion <- - - - - --
at C-IG
_
~ \to Hydroxy or methoxyl groups at C-3, C-4 Fig. 5-1. Potency-enhancing
_
~ CH2CHzCN
cis or trons .yBC-ring junction ._::.-:: _>H or OH at C-14 - - - - -:;.Oxygen insertion at CoB C
N'
, , , , ,,
_
-
,, ~
:;.CH(OH)(CH2bCHJ
mixed agonist-antagonists, strongly paralleling the observations made in the morphine series. Substituents at carbon atoms 5, 7, and 8 of the Coring generally decrease or eliminate analgesia. An exception is the 7a-hydroxypentenyl, which is 130 times as potent as morphine. Hydroxyl substitution at the BC ring junction (C-14) enhances analgesia over the unsubstituted parent. Ring contractions and expansions usually decrease activity. Exceptions are the 6-methyl-C-normorphinans, which have 19 times the activity of morphine, and the C-homo, which is equivalent to morphine. Movement of the termini of the D-ring eliminates activity. Moving the position of the nitrogen from 17 in the D-ring eliminates or strongly decreases analgesia. Introduction of an additional heteroatom into the molecular framework, such as a nitrogen at C-9 or an oxygen at C-8, can strongly enhance analgesia.
References
CHz, co. CHOH Phenyl. thlenyl. ~ - - -/i"\\
243
at C-7
Ketone at C-G
substituents on morphinans.
l
II ~:
II
Ii
244
5' The Morphinans
References
16. A. L. Spiff, V. Zabel, W. H. Watson, M. A. Zemaitis, A..M. Ateya, D. J. Slatkin, J. E. Knapp, and P. F. Schiff, Jr., l. Nat. Prod. 44, 160 (1981). 17. Y. Sasaki and U. Veda, l. Pharm. Soc. lpn. 78,44 (1958). 18. K. Goto, "Sinomenine, an Optical Antipode of Morphine Alkaloids." Kitasato Institute, Tokyo, 1964. 19. H. G. Boit and H. Fientje, Naturwlssenschaften 46 514 (1959); ibid. 47, 180 (1960). 20. S. Pfeifer and D. Thomas, Pharmazie 27, 48 (1972). 21. H. L. Holmes, Alka/oids (N. Y.) 2, 2 (1952); H. L. Holmes and G. Stork, ibid. 2, 161 (1952); G. Stork, ibid. 6,219 (1960); K. W. Bentley, ibid. 13, 1 (1971); F. Santavy, ibid. 17, 385 (1979). 22. C. Olieman, L. Maat, and H. C. Beyerman, Reel. Trav. Chim. Pay-Bas. 95, 189 (1976). 23. C. Schopf and T. Pfeifer, l. Liebigs Ann. Chem. 483, 157 (1930). 24. H. SChmidhammer, A. E. Jacobson, L. Atwell, and A. Brossi, He/v. Chim. Acta 64, 2540 (1981). 25. C. Mannich and H. Lowenheim, Arch. Pharm. 258, 295 (1920). 26. C. SchOpf and H. Hirsch, l. Liebigs Ann. Chern. 489,224 (1931). 27. T. D. Perrine and L. F. Small, l. Org. Chern. 17, 1540 (1972). 28. D. E. Rearick and M. Gates, Tetrahedron Lett. p. 507 (1970). 29. L. F. Bje1danes and H. Rapoport, l. Org. Chern. 37, 1453 (1972). 30. K. W. Bentley, R. Robinson, and A E. Wain, l. Chern. Soc. p. 958 (1952). 31. A. J. Birch and M. Fitton, Aust. J. Chern. 22,971 (1969). 32. H. Schmid and P. Karrer, He/v. Chim. Acta 33, 863 (1950). 33. R. K. Razdan, D. E. Portlock, H. C. Dalzell, and C. Malmberg, l. Org. Chem. 43, 3604 (1978). 34. J. Hellerbach, O. Schnider, H. Besendorf, and P. Bellmont in "Organic Chemistry. Vol. 8. Synthetic Analgesics" (D. H. R. Barton and W. Doering, cds.), p. 1. Pergamon, New York, 1966. 35. R. Grewe, Ber. Dtsclz. Chern. Ges. 76, 1072 (1943). 36. R. Robinson and S. Sugasawa, l. Chern. Soc. p. 3163, p. 3173 (1931); Z. Kitasato and R. Robinson, ibid. p. 785 (1932); R. Robinson and S. Sugasawa, ibid. p. 789 (1932); ibid. p. 280 (1933). 37. R. Grewe, Naturwissemchaftell 33, 333 (1946). 38. R. Grewe and A. Mondon, Chem. Ber. 81B, 279 (1948). 39. R. Grewe, Angew. Chern. 59, 194 (1947). 40. J. Z. Ginos, J. Org. Chern. 40, 1191 (1975). 41. D. C. Palmer and M. J. Strauss, Chem. Rev. 77, 1 (1977). 42. W. Leimgruber and E. Mohacsi, V. S. Patent 3,634,429; cited in W. Leimgruber, E. Mohacsi, H. Baruth, and L. O. Roudall, Adv. Biochern. Pyschopharm. 8 45 (1974). ' 43. R. Grewe and A. Mondon, Chern. Ber. 81, 279 (1948). 44. E. M. Fry and E. L. May, l. Org. Chem. 26, 2592 (1961). 45. K. C. Rice, l. Org. Chem. 45, 3135 (1980). 46. R. Grewe and W. Friedrichson, Chem. Ber. 100, 1550 (1967). 47. G. C. Morrison, R. O. Waite, and J. Shavel, Jr., Tetrahedron Lett. p. 4055 (1967). 48. H. C. Beyerman, E. Buurman, and L. Maat, l. Chem. Soc., Chern. Comrnun. p. 918 (1972). 49. H. C. Beyerman, T. S. Lie, L. Mam, H. H. Bossman, E. Buurman, E. J. M. Bijoter. veld, and H. J. M. Sinnige, Reel. Trav. Chim. Pay-Bas. 95,24 (1976).
245
50. O. Schnider and J. Hellerbach, He/v. Chim. Acta 33, 1437 (1950). 51. R. Grewe, R. Hamann, G. Jacobsen, E. Nolte, and K. Riecke, l. Liebigs Ann. Chern. 581, 85 (1953).
52. 53. 54. 55. 56. 57. 58 59: 60. 61. 62.
H. Henecka, J. Liebigs Ann. Chem. 583, 110 (1953). M. Sasamoto, Chem. Pharm. Bull. 8, 980 (1960). O. Schnider, A. Brossi and K. Vogler, He/v. Chim. Acta 37, 710 (1954). O. Schnider and A. Griissner, He/v. Chim. Acta 34, 2211 (1951). J. Hellerbach, A. Griissner, and O. Schnider, He/v. Chim. Acta 39, 429 (1956). M. Gates, R. B. Woodward, W. F. Newhall, and R. Kiinzli, J. Arn. Chem. Soc. 72, 1141 (1950). . M. Gates and T. A. Montzka, J. Med. Chem. 7, 127 (1964). J. H. Ager, S. E. Fullerton, E. M. Fry, and E. L. May, J. Org. Chem. 28,2470 (1963). M: Gates, Adv. Biochem. Psychopharmcol. 8, 51 (1974). A. Griissner, J. Hellerbach, A. Brossi, and O. Schnider, Hell'. Chim. Acta 39, 1371 (1956). . H. Schmidhammer, A. Brossi, J. L. Flippel)-Anderson, and R. Gilardi, Heh'. Oum. " Acta 66, 2437 (1983).
63. Y. K. Sawa, K. Kawaski, and S. Mayeda, (hem. Pharm. Bull, 8, 960 (1960). 64. I. Monkovic, T. T. Conway, H. Wong, Y. G. Perron, 1. J. Pachter, and B. Belleau, l. Am. Chern. Soc. 95, 7910 (1973). 65. I. MonkoviC,H. Wong, A. W. Pircio, Y. G. Perron, I. J. Pachter, and B. Belleau, Can. J. Chern. 53, 3094 (1975). 66a. 1. Monkovic and H. Wong, Can. l. Chem. 54, 883 (1976). 66b. I. Monkovic, C. Bachand, and H. Wong, J. Am. Chem. Soc. 100, 4610 (1978). 67. D. H. R. Barton, D. S. Bhakuni, R. James, and G. W. Kirby, l. Chern. Soc. C p. 128 (1967). 68. G. Daidone, S. Plescia, and J. Fabra, l. Heterocycl. Chem. 17, 1409 (1980). 69. D. L. Hey, J. A. Leonard, C. W. Rees, and A. R. Todd, J. Chem. Soc. C p. 1513 (1967). 70. T. Kametani and K. Fukumoto, 1. Heterocycl. Chem. 8, 341 (1971). 71. T. Kametani, K. Fukumoto, F. Satoh, and H. Yagi, l. Chem. Soc. C p. 520 (1969). 72. T. Kametani, T. Sugahara, H. Yagi, and K. Fukumoto, J. Chern.Soc. C p. 1063(1969). 73. B. Gregson-Allcott and J. M. Osbond, Tetrahedron Lett. p. 1771 (1969). 74. A. R. Battersby, A. K. Bhatnagar, P. Hackett, C. W. Thornber, and J. Staunton, Chern. Cornmun. p. 1214 (1968). 75. B. Franck, J. Lubs, and G. Dunkelmann, Angew. Chem. Int. Ed. Eng/. 6,969 (1967). 76. T. Kametani, K. Fukumoto, A. Kozuka, H. Yagi, and M. Koizumi, l. Chem. Soc. C p.2034 (1969). 77. T. Kametani, A. Kozuka, and K. Fukumoto, J. Chern. Soc. C p. 1021 (1971). 78. S. M. Kupehan, O. P. Dhingra, and C.-K. Kim, J. Org. Chern. 41, 4049 (1976). 79. M. A. Schwartz and J. S. Mami, l. Am. Chem. Soc. 97, 1239 (1975). 80. C. Szantay, G. Blask6, M. Barczai-Beke, P. Pechy, and G. Dornyei, Tetrah~dron Lett. 21,3509 (1980); J. D. White, G. Carzvatti, T. B. Kline, E. Edstrom, K. C. Rice, and A, Brossi, Tetrahedron 39, 2393 (1983). 81. B. Franck and V. Teetz. Angew. Chem. lilt. Ed. £lIg/. 10, 411 (1971). 82. J. M. Bobbitt, I. Noguchi, R. S. Ware, K. N. Chiong, and S. 1. Huang, l. Org. Chem. 40, 2924 (1975).
83. L. L. Miller, F. R. Stermitz, and J. R. Falck, l. Am. Chern. Soc. 93, 5941 (1971). 84. L. L. Miller, F. R. Stewart, J. P. Gillespie, V. Ramachandran, Y. H. So, and F. R. Stermitz, l. Org. Chem. 43, 1590 (1978).
L
246
5 The Morphinans
85. T. Kametani, K. Shishido, and S. Takano, J. Heterocyl. Chem. 12, 305 (1975). 86. L. Christensen and L. L. Miller, J. Org. Chem. 46, 4876 (1981). 87. L. L. Miller, F. R. Stermitz, and J. R. Falck, J. Am. Chem. Soc. 95, 2651 (1973). 88. E. Kotani and S. Tobinaga, Tetrahedron Leu. p. 4759 (1973). 89. S. M. Kupchan, V. Kameswaran, J. T. Lynn, D. K. Williams, and A. J. Liepa, J. Am. Chem. Soc. 97, 5622 (1975). 90. O. Schnider and A. Grussner, Helv. Chim. Acta 32, 821 (1949). 91. W. M. Benson, P. L. Stefko, and L. O. Randall, J. Pharmacol. Exp. Ther. 109, 189 (1953). 92. H. Isbell and H. F. Fraser, J. Pharmacol. Exp. Ther. 106, 397 (1952). 93. P. L. Stefko and W. M. Benson, J. Pharmacol. Exp. Ther. 103, 363 (1951). 94. P. F. Stefko and W. M. Benson, J. Pharmacol. Exp. Ther. 108, 217 (1953). 95. B. Pellmont and H. Biichtold, Schweiz. Med. Woc/zenschr. 84, 1368 (1954). 96. N. B. Eddy, H. Besendorf, and B. Pellmont, Bull. Narc. 10, 23 (1958). 97. H. Kugita, M. Takedo, and H. Inoue, J. Med. Chem. 13, 973 (1970); H. Inoue, M. Takeda, and H. Kugita, Chem. Pharm. Bull. 21, 2004 (1973). 98. M. Gates and W. G. Webb, J. Am. Chem. Soc. 80, 1186 (1958). 99. M. Gates and T. A. Montzka, J. Med. Chem. 7, 127 (1964). 100. W. J. Musliner and J. W. Gates, J. Am. Chem. Soc. 88,4271 (1966); ibid., Org. Synth. 51, 82 (1971). 101. H. Schmidhammer, A. E. Jacobson, and A. Brossi, Heterocycles 17, 391 (1982). 102a. O. Schnider, A. Griissner, and J. Hellerbach, unpublished; quoted in ref. 34. 102b. E. Mohacsi, T. Hayes, and J. Sepinwall, J. Med. Chem. 27, 1219 (1984). 102c. Anonymous, Drugs Fur. 6, 483 (1981); ibid. 7, 600 (1982). 102d. L.-J. Leinweber, A. J. Szuna, A. C. Loh, J. Sepinwall, K. Carter, and M. A. Schwartz, Biochem. Pharmacal. 31, 553 (1982). 103. H. Biichtold, unpublished; quoted in ref. 34. 104. M. Murakami, N. Inukai, and N. Nagano, Chem. Phaml. Bull. 20, 1699 (1972). 105. M. Murakami, S. Kawahara, N. Inukai, N. Nagano, H. Iwamoto, and H. Ida, Chem. Pharm. Bull. 20, 1706 (1972). 106. W. H. Pirkle and M. Gates, J. Org. Clrem. 30, 1769 (1965). 107. H. Schmidhammer, A. E. Jacobson, and A. Brossi, Med. Res. Rev. 3, 1 (1983). 108. F.-L. Hsu, A. E. Jacobson, K. C. Rice, and A. Brossi, Heterocycles 13, 259 (1979). 109. A. Brossi, F.-L. Hsu, and K. C. Rice, J. Org. Chem. 47, 5214 (1982). 110. Y. K. Sawa, N. Tsuji, and S. Maeda, Tetrahedron 15, 154 (1961). 111. C. Olieman, L. Maat, and H. C. Beyerman, Recl. Trav. Chim. Pays-Bas. 99, 169 (1980). 112. R. Maeda and E. Ohsugi, Chem. Pharm. Bull. 16, 897 (1968). 113. C. Olieman, P. Nagelhout, A. D. deGroot, L. Maat, and H. C. Beyerman, Recl. Trav. Chim. Pays-Bas. 99, 353 (1980). 114. Y. K. Sawa, M. Horiuchi, and K. Tanaka, Tetrahedron 21,1133 (1965); Y. K. Sawa and S. Maeda, Tetrahedron 20, 2247 (1964); Y. K. Sawa, N. Tsuji, and S. Maeda, Telrahedron 20, 2255 (1964). 115. H. Schmidhammer and A. Brossi, Can. J. Clrem. 60, 3055 (1982). 116. F.-L. Hsu, K. C. Rice, and A. Brossi, Helv. Clrim. Acta 63, 2042 (1980). 117. M. D. Rozwadowski, F.-L. Hsu, A. E. Jacobson, K. C. Rice, and A. Brossi, Can. J. Chem. 58, 1855 (1980). 118. A. Manmade, H. C. Dalzell, J. F. Howes, and R. K. Razdan, J. Med. Chern. 24 1437 (1981). '
References
247
119. A. E. Jacobsen, F.-L. Hsu. M. D. Rozwadowska, H. Schmid hammer, L. Atwell, A. Brossi, and F. Medzihradsky, He/v. Chim. Acta 64, 1298 (1981). 120. I. Seki, H. Tagaki. and S. Kobayashi, J. Pharm. Soc. 84, 280 (1964). 121. P. A. Janssen and C. A. M. Van der Eycken, Drugs Affect. Cent. Nerv. Sysi. 2, 25 (1968). 122. H. I. Chernov, J. W. Miller, and G. J. Mannering, Fed. Proc., Fed. Am. Soc. Exp. Bioi. 18, 376 (1959). 123. J. Telford, C. N. Pappadopoulos, and A. S. Keats, J. Pharmacol. Exp. Ther. 133,106 (1961). 124. A. E. Jacobsen, K. C. Rice, J. Reden, L. Lupinacci, A. Brossi, R. A. Streaty, and W. A. Klee, J. Med. Chern. 22, 328 (1979). 125. T. J. DeKornfeld, Cllrr. Res. Anesth. 39, 430 (1960). 126. H. F. Fraser and H. Isbell, Bull. Narc. 12, 15 (1960). 127. J. E. Eckenhoff and S. R. Oech, Clin. Pharmacol. Ther. 1, 483 (1960); J. E. Eckenhoff, G. L. Hoffman, and L. W. Funderburg, Am. J. Obslet. Gynecol. 65, 1269 (1953). 128. C. W. White, Jr., R. Megirian, and P. S. Marcus, Proc. Soc. Exp. Bioi. N. Y. 92,512 (1956). 129. S. Archer and L. S. Harris, Prog. Drug. Res. 8, 261 (1965). 130. A. S. Keats; quoted in A. H. Beckett and A. F. Casy, ?rog. Med. Chern. 2,43 (1962). 131. L. S. Harris, A. K. Pierson, J. R. Dembinski, and W. L. Dewey, Arch. 1m. Pharmacodyn. 165, 112 (1967). 132. L. Lasagna, J. W. Pearson: and T. DeKornfeld; quoted in L. S. Harris, Narc. Drugs: Biochem. Pharmacol. p. 89 (1971). 133. I. J. Pachter, Adv. Biochem. Psychopharmacol. 8, 57 (1974). 134. H. Merz, A. Langbein, K. Stockhaus, G. Walther, and H. Wick, Adv. Biochem. Psychopharmacol. 8, 91 (1974). 135. P. S. Portoghese, R. N. Hanson, V. G. Telang, J. L. Winger, and A. E. Takemori, J. Med. Chem. 20, 1020 (1977). 136. H. Schmidhammer, A. E. Jacobson, L. Atwell, and A. Brossi, Heterocycles 16, 1859 (1981). 137. L. D. Simon, F. R. Simon, E. Mohasci, L. Berger, and E. J. Simon, Ufe Sci. 28,2769 (1981). 138. T. K. Alder and F. H. Shaw, J. Pharmacol. Exp. Ther. 103, 337 (1951). 139. N. B. Eddy, H. Halbach, and O. J. Braendon, Bull. WHO 17,569 (1957). 140. L. Small, H. M. Fitch and W. E. Smith, J. Am. Chem. Soc. 58, 1457 (1936). 141. A. Brossi, L. Atwell, A. E. 1acobson, M. D. Rozwadowska, H. Schmidhammer,1. L. Fippen-Anderson, and R. Gilardi, Helv. Grim. Acta 65, 2394 (1982). 142. D. L. Leland, J. O. Polazzi, and M. P. Kotick, J. Org. Chem. 45,4026 (1980). 143. D. L. Leland and M. P. Kotick, J. Med. Chem. 23, 1427 (1980). 144. M. P. Kotick, D. L. Leland, 1. O. Polazzi.1. F. Howes, and A. R. Bousquet, J. Med. Chem. 24, 1445 (1981). 145. P. Herlihy, H. C. Dalzell, 1. F. Howes, and R. K. Razdan, J. Med. Chem. 25, 986 (1982). 146. 1. Quick, P. Herlihy, and 1. F. Howes, J. Med. Chem. 27, 632 (1984). 147. 1. O. Polazzi, R. N. Schut, M. P. Kotick, J. F. Howes, P. F. Osgood, R. K. Razdan, and J. E. Villarreal, J. Med. Chem. 23, 174 (1980). 148. J. O. Polazzi, M. P. Kotick, 1. F. Howes, and A. R. Bousquet, J. Med. Chem. 24, 1516 (1981). 149. I. Monkovic, H. Wong, A. W. Pircio, Y. G. Perron, I. J. Pachter, and B. Belleau, Can. J. Chern. 53, 3094 (1975).
248
5 The' Morphinans
150. H. Schmid hammer. A. E. Jacobson, L. Atwell and A. Brossi, He/v. Grim. Acta 64, 2540 (1981). 151. A. W. Pircio, J. A. Gylys, R. L. Cavanagh, J. P. Buyniski, and M. E. Bierwagen, Arch. Int. Pharmacodyn. 220, 231 (1976). 152. J. Castaner and D. M. Paton, Drugs Fur. 2,231 (1977); ibid., Drugs Fut. 2, 330 (197~); B. R. Belleau, Chem. Can. 32, 1 (1980). 153. A. Delpizzo, Om. Ther. Res. 20, 221 (1976). 154. R. W. Houde, S. L. Wallerstcin, and A. Rogers, Adv. Pain Res. Ther. 1,647 (1976). 155. J. Castaner and D. M. Paton, Drugs FUt. 2, 746 (1977). 156. A. W. Pircio and J. A. Gylys, l. Pharmacal. Exp. Ther. 193, 23 (1975). 157. J. G. Nutt and D. R. Jasinski, Clin. Pharmacal. Ther. IS, 361 (1974). 158. B. Belleau, H. Wong, 1. Monkovic, and 1. G. Pe~ron, l. Chem. Soc., Chem. Commun. p. 603 (1974); 1. Monkovic, H. Wong, B. Belleau, 1. J. Pachter, and Y. G. Perron, Can. l, Chern. 53, 2515 (1975). 159. Y. K. Sawa and H. Tada, Tetrahedron 24, 6185 (1968). 160. 1. Seki, l. Pharm. Soc. lpn. 83,389 (1963); 1. Seki, J. Pharm. Soc. lpn. 84,615 (1964). 161. Y. Sawa, S. Maeda, and N. Tsuji, U. S. Patent 3,085,091 (April 9, 1963). 162. K. Hirose, A. Matsushita, Y. Kojima, M. Eigyo, J. Jyoyama, T. Shiomi, Y. Tsukinoki, H. Hatakeyama, K. Matsubara, and K. Kawasaki, Arch. Int. Pharmacodyn. 241, 79 (1979). 163. Y. Sawa, R. Maeda, and T. Kada, U. S. Patent 3,738,989 (June 12, 1973). 164. K. W. Bentley and J. B. Taylor, unpublished; quoted in K. W. Bentley, Alkaloids (N. Y.) 13, 75, 120 (1971). 165. K. Hayakawa, 1. Fujii, and K. Kanematsu, l. Org. Chem. 48, 166 (1983). 166. K. Wiesner, J. G. McCluskey, J. K. Chang, and V. Smula, Can. l. Chem. 49, 1092 (1971). 167. J. L. Douglas and J. Meunier, Can. l. Chem. 53, 3681 (1975). 168. M. Menard, P. Rivest, B. Belleau, J.-P. Daris, and Y. G. Perron, Can. l. Chem. 54,429 (1976). 169. S. Sugaswa and S. Saito, Chem. Pharm. Bull. 4, 237 (1956). 170. Y. K. Sawa, N. Tsuji, K. Okabe, and T. Miyamoto. Tetrahedron 21, 1121 (1965). 171. B. Belleau, T. Conway, F. R. Ahmed, and A. D. Hardy, l. Med. Chern. 17,907 (1974); T. T. Conway, T. W. Doyle, Y. G. Perron, J. Chapnis, and B. Belleau, Can. l. Chern. 53, 245 (1975). 172. B. Bclleau and P Morgen, l. Med. Chern. 17, 908 (1974). 173. A. F. Casey, Prog. Drug Res. 22, 149 (1978). 174. S. Shiotani, T. Kometani, Y. litaka, and hai, l. Med. Chern. 21, 153 (1978); T. Kometani and S. Shiotani, l. Med. Chern. 21, 1105 (1978). 175. M. Hori, T. Kataoka, H. Shimuzu, E. Imai, and Y. Suzuki, Heterocycles 20, 1979 (1983). 176. V. M. Kolb and S. Scheiner,l. Pharrn. Sci. 73,389,719 (1984); V. M. Kolb,l. Pharrn. Sci. 73, 715 (1984).
177. S. Saito, Chern. Pharm. Bull. 4, 438 (1956). 178. H. Henecka and W. Wirth, Med. Chem. 5, 321 (1956). 179. S. Shiotani, l. Org. Chem. 40, 2033 (]975). 180. S. Shiotani and T. Kometani, l. Med. Chem. 20, 976 (1977). 181. M, Gates and D. A. Klein, l. Med. Chem. 10, 380 (1967). 182. E. L. May and J. G. Murphy, l. Org. Chem. 19,615 (1954); ibid. 19,6]8 (1954); E. L. May, l. Org. Chem. 23, 947 (1958). 183. H. Kugita, Chem. Pharrn. Bull. 4, 29 (1956).
References
249
184. N. Sugimoto, l. Pharm. Soc. lpn. 75, 183 (1955). 185. N. Sugimoto, S. Ohshiro, H. Kugita, and S. Saito, Chem. Pharrn. BrIll. 5,62 (1957); N. Sugimoto and S. Ohshiro, Chern. Pharrn. Bull. 5, 316 (1957). 186. N. Sugimoto, Japanese Patent 3799 (1957). 187. N. Sugimoto and H. Kugita, Chem. Pharrn. Bull. S, 67 (1957); ibid. 6, 429 (1958). 188. N. Sugimoto, Japanese Patent 30213 (1957). 189. N. Sugimoto and S. Ohshiro, Tetrahedron 8, 296 (1960). 190. S. Ohshiro, Tetrahedron 8, 304 (1960). 191. S. Ohshiro, Tetrahedron 10, 175 (1960). 192. N. Sugimoto and S. Ohshiro, Chem. Pharrn. Bull. 4, 353 (186). 193. E. Ochiai and K. Harasawa, Chern. Pharrn. Bull. 3, 369 (1955). 194. K. Harasawa, l. Pharm. Soc. lpn. 77, 168, 172, 794 (1957). 195. N. Sugimoto and S. Ohshiro, Chern. Pharrn. Bull. 4, 357 (1956). 196. N. Sugimoto and H. Kugita, Chern. Pharrn. Bull. 3, 11 (1955); ibid. 5, 378 (1957). 197. H. Kugita, Chern. Pharrn. Bull. 4, 189 (1956). 198. T. Kametani, K. Kigasawa, M. Huragi, and N. Wagatsuma, Chem. Pharm. Bull. 16,296 (1968). 199. T. Kametani, K. Kigasawa, K. Wakisaka, and N. Wagatsuma, Chern. Pharrn. Bull. 17, 1096 (1969). 200. T. Kametani, K. Kigasawa, M. Hiiragi, N. Wagatsuma, U. Kusama, and T. Uryu, Heterocycles 4, 41 (1976). 201. T. Kametani, K. Kigasawa, M. Hiiragi, N. Wagutsuma, K. Wakisaka, F. Satoh, and S. Saito, l. Med. Chern. 13, 1064 (1970). 202. I. Seki. Chern. Pharrn. Bull. 18, 1269 (1970). 203. 1. Monkovic, Can. l. Chern. 53, 1189 (1975). 204. Y. Lambert, J.-P. Daris, and 1. Monkovic, Can. l. Chern. 55, 2523 (1977). 205. M. Saucier, J.-P. Daris, Y. Lambert, 1. Monkovic, and A. W, Pircio,J. Med. Chern. 20, 676 (1977). 206. Y. Lambert, J.-P. Daris, 1. Monkovic, and A. W. Pircio, l. Med. Chem. 21,423 (1978). 207. T. A. Montaka, J. D. Matiskella, and R. A. Partyka, U. S. Patent 4,154,932 (May, 1979). 208. Anonymous, Drugs Fur. 6, 632 (1981). 209. A. W. Pircio and J. P. Buyniski, PharmacologiJt 22, Abstr. 802 (1980). 210. J. C. Reiffenstein, R. L. Cavanagh, and J. A. Gylys, PharmacologiJt 22, Abstr. 803 (1980). 211. F. Sauter, P. Stanetty, E. Hetzl, and F. Fuhrmann, l. Heterocycl. Chem. 20, 1477 (1983).
Introduction
251
6. The Benzomorphans 1. Introduction . . . . . . . . . . . . . .. ..."." II. Benzomorphan Syntheses. " .. "".......... III. Structure-Activity Relationships in the Benzolllorphan Analgesics A. The Effect on Analgesia of Alkyl G~oups in Rings Band C . . B. Nitrogen Substitution . ". . . . . . . . . . . . . . . . . C. The Introduction of Oxygen Containing Functions in the B- and C-Rings of Benzomorphans . . . . . . . . . . . . D. A-Ring Substitutions "and Replacements . . . . . . . . . . . . . E. BC-Ring Enlargements and Contractions . ." . . . . . . . . F. More and Less Complex Benzomorphan Analogs . . ." . ." . ." . . . G. Nitrogen Movements within the Benzomorphan Nucleus Plus Nor and Homo Derivatives . . . . . . . . . . . IV. The Chemical Anatomy of the Benzomorphans
References . . . .'.
I.
. . .
"
. . . . . . .
250 252 259 259 273 289 295 297 304 306 310 311
Introduction
It is tempting to say that the benzomorphans originated in a rigorous scientific study of the simplification of the morphine molecular framework to uncover the pharmacophore. However, this type of molecule is a relative latecomer to the family of opiate-derived analgesics. For example, the open chain analgesic methadone, which was developed in Germany in the early 1940s, preceded the benzomorphans. The present simplification of the morphine skeleton may be visualized as an excision of furan ring oxygen in morphine followed by truncation of the resultant morphinan at the BC-ring junction (Fig. 6-1). The benzomorphan ring system was first synthesized by Barltrop in 1947 (1). The name of the ring system is derived from the trivial name morphan, originally suggested to Barltrop by Sir Robert Robinson to denote the 2-aza[3.3.1] ring system (2,3). Barltrop's original benzmorphan designation was subsequently changed, however, by the editors of The Journal of Organic Chemistry to the current benzomorphan. The Korean conflict served as a stimulus for the rapid development of this type of analgesic. The National Institutes of Health (NIH) in the United States was charged with finding adequate, but not necessarily improved, substitutes for morphine and codeine. The discovery of May and Eddy at NIH that the 6,1I-dimethyl derivative of benzomorphan met
250
morphine
benzomorphan
morphinan Fig. 6-1.
A conceptual
genesis of the benzomorphans.
the above criterion stimulated intensive investigation of these molecules (4). There are two common numbering systems for the benzomorphans: (a) the original and still commonly used 6,7-benzomorphan numberin~ and. (b) the Chemical Abstracts numbering system based on the benzazocme nng. In this chapter, the Chemical Abstracts numbering system together with the trivial name benzomorphan will be used. The Chemical Abstracts name for benzomorphan, which,escaped unscathed in the Temh Collective Index, is 1,2 3 4 ,5 ,6-hexah ydro-2 ,6-methano-3-benzazocine. '.'
~ 4'
3'
0
2'
~-CH3
4
6
9 5
I'
Benzomorphan Numbering
Chemical Abstracts Benzazocine Numbering
Because of the importance of the alkyl groups at 6 and 11 to the analgesic properties of the benzomorphans, it. i~ not fe.asibl~ to use the position substituent approach to structure-actIvIty relatIonsh~ps that .has been utilized for both the morphine alkaloids and the morphmans. FIrst, the important syntheses of this class of compoun.ds will be indic~ted, followed by the effects on analgesia of alkyl groups m the B- and C-nngs, substituents
on the aromatic
A-ring,
nitrogen
substitution,
and ox~ge.n
insertion. The introduction of oxygen into the alicyclic rings can mImIc 14-hydroxylation in morphine, as well as impart interesting opiate receptor selectivities. Ring-expanded and-contracted benzomorphans have also led to interesting analgesics in this series of opiates.
151
II.
6 The Benzomorphans
Benzomorphan
II
Benzomorphan
253
Syntheses
Syntheses e,d,e~
In 1947, Barltrop prepared the first benzomorphan using the tetralone (1) (Scheme 6-1) as the starting material (1,2). The quaternary benzomorphan (2) was prepared as a model for the A-, B-, and D-rings of morphine. Although this synthetic approach was straightforward, the overall yield was poor. Following the initial work, which did not proceed beyond the qu~ternary salt (2), May reinvestigated the synthesis using the dimethylaml?o ~nalog (5). Bo.th the alkylation of the tetralone and the subsequent cychzahon occurred m low yields, and this approach was abandoned in favor of a route starting with hydratroponitrile (Scheme 6-2). Although the second route is much longer, it provides the benzomorphan (3) in 5% overall yield. Starting from hydratroponitrile (4), the overall yield of the tetralone (6) is 60%. The remaining reaction sequence leading to 3 is equal~y. effectiv~. However, this approach is limited to benzomorphans containing subshtuents at position 6. Efforts to prepare 6,11-disubstituted comp~unds failed when the methyl ketone corresponding to the aldehyde (5) faded to undergo the Knoevenagel condensation (6). The tetralone route, however, has remained a useful route to 6-substituted benzomorphans. The most widely used approach to benzomorphan synthesis is based on t~e acid-c.atalyzed c~cJization of appropriately substituted tetrahydropyridmes. ThIs method IS analogous to Grewe's method for the synthesis of
5
4
CZ'
f
?
~(C"2)2NMe2 CH3 6
j, k,l '
?
3 Scheme 6-2. A more efficient synthesis of the benzomorphans. Reagents: (a) sodamide, Cl(CHzhNMe2; (b) lithium aluminum hydride; (c) methyl cyanoacetate, ammonium acetate in acetic acid; (d) H2/Pt; (e) hydrochloric acid; (f) polyphosphoric acid; (g) hydrobromic acid; (h) bromine; (i) ammonia (aqueous); (j) pyrolysis; (k) hydrochloric acid in ethanol; (I) Huang-Minion.
1
b,e
".
2 .Scheme 6-1. Original synlh~sis of the benzomorphan ring syslem. Reagents: (a) sodamlde, CI(CHzhNEtz; (b) bromine; (c) sodium bicarbonate.
morphinans from substituted tetrahydroisoquinolines (d. Chapter 5, Section IV,A) (7). This route is especially useful, since it allows the preparation o( 6,1l-disubstituted benzomorphans that are more effective analgesics and are inaccessible by the tetralone route. The first examples of benzomorphans prepared by the Grewe reaction are the 6,1l-dimethyl derivatives 9 and 11 (Scheme 6-3) (6). Addition of either benzyl grignard or p-methoxybenzyl grignard to 3,4-dimethylpyridine methiodide (7), followed by reduction of the resultant sensitive dihydropyridine, gives the tetrahydropyridines 8 and 10. The grignard has added to the more hindered position of the pyridinium salt. Subsequent cycJization of 8 with phosphoric acid gives the 6,1l-dimethylbenzomorphan (9) in 20% overall yield. Analogously, the tetrahydropyridine (10) undergoes cycJization and
254
6 The Benzomorphans
-
II
Benzomorphan
255
Syntheses
a,b
-
R
CH3
7
(R = HI (R = OH) cyclization of benzyltetrahydropyridines. (c) 85% phosphoric acid.
8
(R
HI
10
(R
OCH3)
9
--b
12
11
Scheme 6-3. Grewe nard; (b) borohydride;
a
Reagents:
(a) benzyl
grig-
ether cleavage to furnish the 8-hydroxy-6,1l-dimethylbenzomorphan in 14% yield from 7. The Grewe synthesis of benzomorphans is very general, and literally hundreds of compounds have been prepared by this method. In many cases, the greatest synthetic challenge has been the preparation of the appropriately substituted pyridine. The Grewe benzom01'phan products 9 and II in Scheme 6-3 are drawn with the alkyl groups cis with respect to the B-ring. This cis stereochemistry has been termed a by May and Eddy (8). Analogous to the Grewe morphinan synthesis, small amounts of the trans, with respect to the B-ring, or {3 isomer can also be isolated in many cases. The initial assignment of stereochemistry was based on analogy with the proven morphinan BC-ring stereochemistry (7). The isomerism in the benzomorphans has been extensively studied because of the important pharmacological differences between the a and {3series. Initially, differentiation was based on the rate of methiodide formation where by the less hindered a isomer quaternized 5-10 times faster than the more hindered {3 isomer (9). Subsequently, proton nuclear magnetic resonance NMR studies demonstrated that the II-methyl group in the cis-isomer resonates about 25 Hz upfield from the lla-methyl (9). These suppositions were then supp.orted by X-ray crystallographic studies (10).
14 (~I 13 (0:) Scheme 6-4. Reagents: (a) methyl grignard. then pyrolysis; (b) thionyl chloride, pyridine; (c) H2/Pt, HCI04; (d) H2/Pt.
I
L
Because the {3isomers usually occur as a minor product in the Gre~e synthesis, methods have been developed to make them more readily available. The ll-ketobenzomorphans are readily available using the tetralone route (Scheme 6-1). Grignard conversion to the tertiary alcohol and subsequent dehydration yields the ll-methylene derivative (12) (Scheme 6-4). Catalytic hydrogenation under neutral cond.itions gives th~ a isomer (13) through addition across the {3face. ProtonattOn of the ammo group in 12, however, causes the addition of hydrogen to occur from the a face of the molecule, resulting in formation of the ll{3-methyl group (14) in 70% yield (11). An alternative method involves the construction of a
256
6 The Beiizomorphans
Me
Benzomorphan
Syntheses
257
Me a,b
cQrHe ~N
N+ I
I Me
Me
~e +
HeD
+ Me
C H 6 5
15
C6HS
Me
+
~~e
NC
~'I
16
Me
He\\ NC
I Me
N I Me
C6HS
C6HS
17
18
Me
Me
~~ Me
19
Scheme
~,,'Me
~,,'He 17
II
;:.,~ NC f
Me
C6HS
~
6-6.
of handling. For construction of the ,B-benzomorphan (21), the cyanide is reductively removed and the resultant tetrahydropyridine is cyclized with aluminum chloride in carbon disulfide. The cycIization catalyst and solvent combination had previously been used to enhance the amount of ,Bisomer formed in Grewe cyclizations (13). A method for generating the tetrahydropyridine from a noncyclic precursor is illustrated in Scheme 6-6. Conversion of the butenyl amine to its phenylacetamide allows cyclization to the dihydropyridine using the Bischler-Napieralski reaction. Subsequent borohydride reduction then yields the requisite Grewe tetrahydropyridine (14). Novel methods for the construction of the benzomorphan framework have included the use of homolytic chloramine cyclizations (Scheme 6-7) (15). The chloramine (22) is oxidized to the protonated aminyl radical, which cyclizes in 92% yield to a 4: 1 mixture of cis and trans piperidines (23). The cis compound readily undergoes a second intramolecular ring formation to yield the benzomorphan in 75% yield (16). A different approach uses the construction of the aromatic ring onto the existing aza-bicyclic framework (Scheme 6-8). By using standard methodology, the aza-bicyclic (24) is converted to 25, which contains a heterocyclic pyrone ring in place of the usual benzene ring in benzomorphans. The pyrone ring in 2S is sufficiently reactive to undergo an inverse electron demand Diels-Alder reaction with the electron-rich dienophile, dimethoxyethylene. The regiochemistry of the cycloaddition is such that a methoxyl
I C6HS
20
21 Scheme 6-5. Reagents: (a) benzyl grignard; (b) perchloric acid; (c) sodium cyanide; (d) hydrochloric acid. ether; (e) hydrochloric acid; (f) sodium borohydride; (g) aluminum chloride, carbon disulfide.
tetrahydropyridine containing substituents with defined stereochemistry (Scheme 6-5) (12). Addition of benzyl grignard to 1,3,4-trimethylpyridinium yields a mixture of iminium dienes 15 and 16, which can be separated as their cyanide adducts 17 and 18. Reaction of the hydrochloride salt of 17 with aqueous acid eliminates cyanide and results in the rearranged iminium diene (19), with a trans relationship between the benzyl and methyl groups. The diene (19) is trapped with cyanide for ease
L
258
III
6 The Benzomorphans
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
259
group is inserted in the appropriate 8-position of the resultant benzomorphan (26) after decarboxylation and aromatization. The superfluous ester group at C-9 is readily removed via 27 in 97% yield to give the 8methoxybenzomorphan (17).
III.
22
--?
A.
23
Reagents:
(a) TiCI). aqueous acetic acid; (b) aluminum chloride.
a
--+
24
25
b c,d
The Effect on Analgesia of Alkyl Groups in Rings B and C
The majority of the benzomorphan analgesics synthesized have substituents at positions 6 and 11, the BC-ring junction and the Coring, respectively. These two positions correspond to the vestiges of the Coring in morphine, which includes the critical 14-position of morphine. Although it is generally felt that substitution at these two positions is necessary, it is instructive to see how this supposition was developed. This section is restricted to the N-methyl substituted benzomorphans. Other examples of N-alkyl substitutions are found throughout this chapter. Despite the diverse array of synthetic approaches to benzomorphans reported in Section II, the parent unsubstituted benzomorphan (32) has been prepared by another method (Scheme 6-9). This reaction sequence starts from the pyridinium salt (28), which is readily available from 4-phenylpyridine, and has as its critical step the intramolecular FriedelCrafts acylation of the 4-phenylpiperidine carboxylic acid (30). The resultant I-keto compound (31) is easily converted to the parent benzomorphan (32) (18). The corresponding 8-hydroxy (33), 8-acetoxy (34), and 8-methoxy (35) derivatives, which are equivalent to the 3-position derivatives in morphine, heroin, and codeine, respectively, have been synthesized using standard Grewe methodology. In the mouse hot plate assay, the parent compound (32) has 10% of the activity of morphine, while the 8-acetoxy derivative (34) has 16%. The methoxy substituted compound (35) is lethal at analgesic doses. The phenol (33) has 25% of the activity of morphine. This activity is of particular interest, since 33 lacks the quaternary carbon atom, equivalent to C-14 in morphine, once thought essential to morphinomimetic activity. Perhaps the most surprising and differentiating observation made in this series, as opposed to that of the more complex morphine and morphinan, is that the racemate of 35 appears to have opiate antagonist properties, causing abstinence symptoms in morphinedependent monkeys at equianalgesic does (19). In the monoalkyl substituted benzomorphans, the 6-position has been most intensively studied, followed by the 11a- and l1,B-substituents. The
b
Scheme 6-7.
Structure-Activity Relationships in the Benzomorphan Analgesics
>
26 27
R = C02Et R = C02H R = H Scheme 6-8. Reagents: (a) dimethyl methoxymethylenemalonate, base; (b) 1,1. dlmethoxyethylene; (c) sodium hydroxide; (d) copper powder, quinoline, 22'C.
l
260
6 The eenzomorphans
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
261
Table 6-1
a,b ~
The Effect of Increasing Alkyl Substituent Monosubstituted Benzomorphans
28
Length
at C-6 in
29 HO
-
Compound
R
Analgesic ActivityU
Reference
33 36 37 38 39 40 41
H CH) CzHs n-C3H7 n-C4H9 n-CsH 11 n-C6H13
0.25 0.14 0.65 0.76 0.48 0.50 0.11
/8 20 2/ 22 23 24 23
f
30
31 U
32
(R
33
(R
34
(R
to morphine
= I, hot plate assay.
to the standard Grewe synthesis (27). A versatile synthesis that allows the preparation of five-, six-, and seven-membered C-rings is illustrated in Scheme 6-10. The readily accessible tetralone ester (42) is converted to its amide (43). Subsequent bromination allows ri~g closure to the keto benzomorphan lactam (44). Removal of the oxygen functions yields 6-phenylbenzomorphan (45) (26). The 6-phenyl compounds are not very active analgesics (Table 6-2). The unsubstituted (45) and the 8-hydroxy (47) derivatives are equipotent, having about 20% of the activity of morphine. The presence of a para-chloro substituent on the 6-phenyl ring retains the analgesic properties, but a para-hydroxy group eliminates it (28). The 1113-methyl benzomorphans have been prepared using a tetralone route similar to that shown in Scheme 6-10 (29). The same tetralone approach to the lla-methyl series yields only a naphthalene derivative. A more circuitous route using intramolecular mercury(II)-induced cyclization of an amino group to the double bond in a dihydronaphthalene yields the lla-methyl analog (30). A subsequent synthesis allows stereospecific synthesis of either epimer from a common intermediate (31). The unsubstituted 1113-methyl (50) and its 8-hydroxy derivative (51) are comparable in analgesic potency to codeine and morphine, respectively (32). Neither compound supports morphine dependence in rhesus monkeys; (51) precipitates withdrawal symptoms when substituted for morphine. This opiate
= R) = OH) = 02CCH3)
35 (R = OCR3) Scheme 6-9. Reagents; (a) KCN; (b) methanol, hydrochloric acid; (c) methyl iodide; Hz/Pt; (e) hydrochloric acid; (f) polyphosphoric acid; (g) Wolff-Kishner reaction.
Relative
(d)
6-methyl A-ring unsubstituted compound was prepared early in the development of this type of molecule and has minimal hot plate activity (5). The effect of increasing alkyl substitution at position 6 in the 8-hydroxybenzomorphans is illustrated in Table 6-1. The 6-methyl derivative (36) is slightly less analgesic than the unsubstituted compound (33), but analgesic potency increases with chain length, with the peak at n-propyl, and then slowly diminishes. The total effect of chain length at this position on analgesia is not very strong. A series of 6-phenyl benzomorphans has been prepared as part of a larger investigation that included the more interesting 6-phenyl-ll-methyl compounds. The presence of the aryl ring allows a variety of synthetic approaches (25,26) to these biologically interesting molecules in addition
i
262
6 The 'Benzomorphans
o II
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
263
Table 6-2
-
a
Analgesic Activity of 6-Phenylbenzomorphans
~C02Et 6 5
42
43 0
II
-
~"' 6~H 5
~~-CII]
c
-d
C6H5
CON II(CII']
Compound
R1
R2
Potency.
45 46 47 48 49
H H OH OH OH
H CI H Cl OH
0.18 Activeb 0.18 Activeb
44 0 II
a
~-CII]
~-CH
o
C6HS]
~C6H5
45 .Scheme 6-1~. Reagents: (a) NH2CH3; (b) bromine; (c) sodium methoxide; Klshner reachon; (e) lithium aluminum hydride.
~_CII
CH3 R
(d) Wolff-
"'CH]]
o HO
50
(R
H)
51
(R
OH)
52
antagonistic activity is similar to that observed with the unsubstituted compound (33). The lla-isomer (52) has approximately 0.25 times the
Relative to morphine = 1, mouse writhing assay. b Mouse tail flick assay.
analgesic potency of the {3isomer (51), but again, it is not morphine-like in addicted monkeys. Compound 52 also possesses antagonist properties and causes long-lasting abstinence syndrome in monkeys (30). During the conceptual development of the benzomorphan molecule, morphine was dissected by cleavage of the Coring, leaving the quaternary carbon and the tertiary carbon at C-14 intact. The retention of these features, using substitution by methyl groups, was originally felt to contribute to analgesic activity by providing steric bulk similar to that of the excised alicyclic Coring (2). A series of A-ring unsubstituted 6,11dialkyl benzomorphans having both the a and {3 conformations is presented in Table 6-3. Even in the absence of the phenolic 8-hydroxyl, substantial analgesic activity is retained. For instance, both the 6,11a and -{3diethyl substituted compounds 53 and 54 are approximately one-half as potent as morphine (33). In this particular unsubstituted series, there is little difference in potency has been observed between the 11a and {3 stereoisomers, the maximum being a factor of 3 with 6, II-dimethyl substitution. The situation is remarkably different in the 8-hydroxy series (Table 6-4), where up to an 80-fold difference in analgesic potency between the
Table 6-3
III
Analgesic Activity of 6, II-Disubstituted
'-, 1
R
{3 R1
9 2] 53 54 55 56
CH] CH] C2Hs C2Hs n-C]H7 n-C]H7
Relationships
in the Benzomorphan
R2
Analgesic Potency"
a-CH] {3-CH] a-C2Hs {3,C2Hs a-CH] {3-CH]
0.08 0.24 0.42 0.50 0.10 0.25
a 3 1.2 2.5
References 8 13,22 33 33 11.34 lJ,34 ".
" Relative to morphine = I. mouse hot plate assay.
Table 6-4 8-Hydroxybenzomorphans
o
Pharmacological Properties of 6- and 6,1l-Alkyl-8-hydroxybenzomorphan
-, 1 R
~:"3
o
HO
]]
R' CH] CH3 CH] CH3 CH3 CH] C2Hs C2Hs C2Hs C2Hs n-C]H7 n-C]H7 n-C]H7 n-C]H7
R2
Enantiomers
~N-," IIIR23 ;;;1
R
HO {3
57 58 59 60 6] 62 63 64 65 66 67 68 69
265
Table 6.5
6, II-Dialkyl Substituted
Compound
Analgesics
stereoisomers is observed. This significantly increased analgesic activity is a general rule with 8-hydroxy substitution (39). On the basis of the limited number of substituents contained in Table 6-4, the optimum substitution at C-6 appears to be ethyl, while at C-lI a J3-methyl group is best (2,37). The resultant benzomorphan (63) is 30 times more potent than morphine. Compound 11, metazocine, is considered the parent of the entire series of dialkyl benzomorphans. The compounds contained in Table 6-4 all show moderate to strong analgesia in the mouse and low or no physical dependence liability in rhesus monkeys, a clear separation of morphine effects. The monoalkyl and dialkyl benzomorphans listed in Tables 6-1 and 6-4 are all racemates, since they have been prepared by total synthesis from optically inactive precursors. Separation of the enantiomers has been achieved using classical resolution techniques, and a comparison of the biological activities is given in Table 6-5. As expected, the levo enantiomers are twice as potent as the corresponding racemates and, like the
@f$:"J o Compound
Structure-Activity
Benzomorphans .without A-Ring Substitution
Analgesic Potency"
a-CH1 {3-CH3 a-~Hs {3-C2Hs a-n-C3H7 {3-n-C]H7 a-CH] {3-CH] a-C2Hs {3-C2Hs a-CH3 {3-CH] a-n-C]H7 {3-n-C]H7
" Relative to morphine = I, mouse hot plate assay.
0.7 4.8 1.5 4.5 0.75 1.7 0.43 30.0 0.50 7.5 0.58 10.0 0.03 2.4
a 16 3 2.3 70 15 17 80
References
Compound
R1
R2
Enantiomer
Analgesic Potency.
22 22,35 22.36 22,36 37 37 22,38 22.38 38 38 37 37 22 22
II
CH] CH] CH] CH] ~Hs ~Hs n-C]H7 n-C]H7 CH] CH] ~Hs C2Hs
CH] CH] IJ-C]H7 n-C]H7 ~Hs ~Hs CH] CH] H H H H
Leva Dextro Leva Dextro Leva Dextro Levo Dextra Levo Dextra Leva Dextra
2.0 Inactive 1.1 Inactive 1.0 0.16 1.5 0.1 0.67 0.05 2.0 0.06
to morphine dependence
= I, mouse liability.
60 64 66 36 37
. b C
Relative Physical
Relative to nalorphine
= I.
hot plate assay.
PDLb None None None None None Intermediate None High None Low None Low
Morphine AntagonismC 0.02-0.03 None Yes None 0.1 None 0.2 None 0.02 None 0.02-0.05 None
266
racemates, do not sustain morphine dependence in rhesus monkeys (this is characteristic of no physical dependence liability). They actually demonstrate a nalorphine-like opiate antagonism in causing the morphine abstinence syndrome in addicted monkeys. The most effective is the levo isomer of the 6-propyl-11a-methyl derivative (66), which is one-fifth as potent as nalorphine as an antagonist but is a more potent analgesic than morphine. Just as surprising is the capacity of the majority of the dextro enantiomers to substitute for morphine (40). The 6-methyl-1l13-propyl racemate (61) has also been resolved into its enantiomers. As expected, the levo enantiomer of 61 is considerably more potent than morphine and exacerbates the morphine withdrawal syndrome in addicted rhesus monkeys, indicating that it too possesses opiate antagonist properties. The dextro enantiomer of 61 lacks opiate activity (41). The 6-phenyl-ll-alkyl benzomorphans have been prepared as an outgrowth of the I1-desalkyl derivatives (d. Table 6-2). In contrast to these simpler derivatives, the II-alkyl derivatives arise from a completely different synthetic approach, as illustrated in Scheme 6-11 (42). The tertiary alcohol (72) is readily prepared by grignard addition to 1,3dimethylpiperidone. Acid-catalyzed dehydration yields an equimolar mixture of the two isomeric tetrahydropyridines 73 and 74. After extended acid
III
6 The Benzomorphans
treatment,
74 becomes
the predominant
isomer
(73:74
=
Structure-Activity
C'o
Relationships
CH)
in the Benzomorphan
(1 /'
~~I
CH)
267
Analgesics
+
I
CH)
CH) 7)
72
-
74
c
75
15:85).
After quaternization of 74 with p-methoxybenzylbromide, Stevens rearrangement affords the necessary Grewe intermediate (75), which cyclizes exclusively to the 1113-methyl derivative (71) (42). The exclusive formation of the 1113-isomer is unusual and has been rationalized in terms of differential stabilities of the phenyl-stabilized carbonium ions in the Grewe cyclization (2,42). The l3-conformation in 71 was ultimately confirmed by X-ray structural analysis (43). The biological activity of a series of 6-phenyl-l1-alkyl benzomorphans, including the enantiomers of 71, is presented in Table 6-6. Of these compounds, the most extensively studied is 71 and its enantiomers (44). Levo-7I is not only a potent analgesic but also an effective opiate antagonist, precipitating withdrawal symptoms in addicted monkeys (45). The antagonistic potency has been compared to that of nalorphine. Morphine antagonism could even be demonstrated in the mouse tail flick analgesia assay. The dextro enantiomer of 71 is not only an analgesic but also possesses a high level of physical dependence liability and causes complete suppression of morphine withdrawal symptoms in monkeys at 5 mpk (45). Clinically, metazocine [( - )-ll] and G PA 1657 [( - )-71] have been studied. In humans, metazocine [(-)-ll] is an excellent pain reliever, but with respect to abuse potential it is similar to morphine rather than
HO
@
71 Scheme 6-11. Reagents: (a) hydrochloric acid, heat, 15.minutes; (b) hydrochlor.ic ac.id. 48-hour reflux; (c) anisyl bromide; (d) KOH. toluene, 108 C; (e) 48% hydrobromIc aCid.
nalorphine. The correspondence between animal and human tests has been at best qualitative and thus disappointing (46). In humans, GPA 1657 [(-)-71] is a potent analgesic, both orally and parenterally. Orally, G.p~ 1657 is 20 times as potent as pentazocine (see below), and parenterally It IS twice as potent as morphine. In spite of its antagonistic properties: GP A 1657 causes most of the usual morphinomimetic side effects: respiratory depression, dizziness, nausea, and euphoria. Interestingly, in a single uncontrolled study, tolerance was not obs,erved after 90 ~ay.s.of t.reatment. There was also no evidence of physical dependence hablltty In a large patient study, which is consistent with its antagonist properties (45).
268
6 The Benzomorphans
111 Structure-Activity
Relationships
Table 6-6 Analgesic
269
Analgesics
Table 6-7 Activity
of 6-Aryl-II,B-alkylbenzomorphans
Compound
Compound
R1
R2
R3
Analgesic Potency"
Reference
70 (:t )-71 ( - )-71 ( + )-71 76 77 78
H OH OH OH OH OH OH
H H H H CI F H
CH3 CH3 CH3 CH3 CH3 CH3 ~H5
A 2.4 6.7 0.4 A A A
28 42 42 42 28 28 28
"Relative
in the Benzomorphan
81 82 83 84 85
R
Analgesic Potency"
CH3 ~H5 n-C3H7 n-C4H9 C(CH3h
0.07 0.16 0.21 0.22 Inactive
a
Relativetometazocine(lI) choline writhing assay.
= l,mouseacetyl-
to morphine = I, mouse hot plate assay,
A hybrid of the benzomorphan system with portions of the extremely potent thebaine Diels-Alder adducts is exemplified by structures 79 and 80. The unique tertiary carbinol function found at position 7 in the Diels-Alder adducts has been similarly positioned at llJ3 in the benzomorphans. Despite their lack of a phenolic group at position 8, these analogs, 79 and 80, possess 40% of the analgesic activity of morphine; this is somewhat disappointing in view of the 400-fold increase seen in the Diels-Alder adducts (47). A series of analogs of 80 has been reported that are 6-methyl-8-hydroxy derivatives in which the length of the aliphatic tail beyond the tertiary alcohol has beeen varied (48), Despite the phenolic hydroxyl, all of these compounds are less active as analgesics than 80 (Table 6-7), and they do not demonstrate any opiate antagonism. Other disubstituted benzomorphans are the 8-methoxy-I, I-dimethyl derivative (86), which is about twice as potent as codeine (49), and the 5,6-dialkyl compounds. Apart from the 6,lla- and -J3-derivatives, the 5,6-dialkylbenzomorphans are obtained in vanishingly small yield in the Grewe cyc1ization (50). The 5,6-dimethyl derivative (87) possesses two-tenths the activity of metazocine (11), while the 5-methyl-6-ethyl-88 has only 0.05 times the analgesic activity (50).
79 80
(R (R
= =
COCH ) 3
86
C(CH3)20H)
87
(R
=
CH3)
88
(R
=
C2HS)
Many potential combinations for three substituents in the BC-rings of benzomorphan exist, but relatively few have been studied. Both a methyl group and a phenyl ring have been introduced into the I-position of metazocine via the I-ketone. Neither of these derivatives, 89 or 90, has significant analgesi,c potency (51).
HO 89
(R
90
(R
= CH3) = C6HS)
270
6 The Benzomorphans
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
271
Table 6-8 Analgesic
Potency
of 6,11,11- Trisubstituted
Benzomorphans
4 steps,
Compound 91 92 93 94 95 96 97 91 Scheme
6-12.
Reagents:
(a) methyl
grignard;
(b) hydrobromic
acid.
A much more interesting group of trisubstituted benzomorphans is the 6,11,11-trialkyl compounds, which, with varying nitrogen substitution, has led to interesting and specific opiate receptor ligands. In the N-methyl derivatives, the pharmacological profile is similar to that of the 6,11disubstituted derivatives, but the trisubstituted compounds are more potent and much longer-acting (52). The construction of trimethyl derivative 91 is briefly outlined in Scheme 6-12. The major synthetic challenge is the construction of a 2-benzyl-1,3,3,4-tetralkylpiperidine-4-ol that undergoes a Grewe-type cycJization on acid-catalyzed carbonium ion formation at C-4 (52). The compounds prepared in this way are presented in Table 6-8. The majority are 11,11-dimethyl derivatives, the most active being the 6-ethyl (92) and 6-phenyl (94) derivatives, which have over 60 times the analgesic potency of pentazocine (53). Acetylation of the 8-hydroxyl increases this activity fourfold. With the exception of the 11,11tetramethylene (95), which has 12% of the antagonist activity of nalorphine, none of these compounds possess opiate antagonist activity. While t~ere is a paucity of opiate receptor affinity data, the 6-ethyl-11,1ldImethyl (92) has significantly higher receptor affinity than morphine (54). The X-ray crystal structure of gemazocine (91) has been determined (55). In summary, the advantage of the 1l,11-disubstitition over the ll-mono-
CH) CH) CH) CH) (CH2). CH3 CH3
CH3 ~H5 n-C3H7 C6H5 ~H5 ~H5 C2H5
H H H H H COCH3 CH3
Analgesic Potency.
Opiate Receptor Affinityb
12.5 62.5 42 62.5 16' 250 16
0.07
a Relative to pentazocine (150) = I, rat tail flick assay. b Relative to morphine 1, IC50 = 3.4 nm. = , Antagonist activity = 0.12 relative to nalorphine = 1.
substituted benzomorphans is its greater potency and much greater duration of action (52). The incorporation of the tertiary hydroxyl side chain in the oripavinebased analgesics, originally described for the 6-methyl benzomorphans, has been extended to the 6,lla-dimethyl derivatives (48,56). The polarization of activity in these compounds, either agonist or antagonist, is illustrated in Table 6-9. The antagonistic potency of N-methyl (100) is five times that of nalorphine, making it among the most potent 3-methyl antagonists known (56). The replacement of the tertiary alcohol by a secondary ketone in these derivatives has been undertaken to prepare analgesics with better agonist-antagonist properties (57). Noteworthy in these benzomorphan ketones (Table 6-10) is the increase in analgesic potency as the linear aliphatic chain of the ketone is extended from 101 to 104. The activity peaks at 104, which has 35 times the potency of morphine in the writhing assay and 100 times that in the tail flick assay. Analgesic activity then falls off rapidly as the chain length is further extended. The smaller straight chain analogs, in contrast to the tertiary alcohols (98-100) do not possess opiate antagonist properties. However, the pentyl derivative (105) retains some analgesic properties and is equivalent to naloxone
Table 6.9
III
Benzomorphan
R
Analgesic Potency.
Antagonist Activityb
0.2 <0.02 <0.02
0.04 0.30 5.0
CH3 C(CH3h (CHzhCH(CH3h
"b Relative to morphine = 1. Relative to nalorphine = 1.
Table 6-10 Ketone Derivatives of the Oripavines
Analgesic
Compound 101 102 103 104 105 106 107 108 109 110 111 112
R CH3 CzHs C3H7 C4H9 CsHlI C6HI3 i-C3H7 i-C4H9 i-CsHll i-C6H13 (CHzhC6Hs (CHzh-c-CsH9
Q
Relative to morphine b Relative
in the Benzomorphan
Analgesics
273
as an antagonist. When the ketone side chain is branched, the analgesic effects in the writhing assay are reduced, and in the tail flick assay they are eliminated for all the derivatives except 108, which is still a very potent analgesic. The derivative 112 appears to be a potent antagonist with no agonist effects, since the reported analgesic activity of 112 is non-enantio selective (58). The receptor affinities of 112 are interesting. Initially, 112 was postulated as a selective K antagonist; however, subsequently, it was demonstrated that 112 has a receptor selectivity similar to naloxone at the ,.", 8, and K receptors (59). In addition, compared again to naloxone, its duration of action is substantially longer (59). The introduction of three alkyl groups at positions 6 and 11 has thus led to potent analgesics and antagonists with either relatively simple functional groups like gemazocine (91) or with the more complex oripavine portions just discussed. B.
Benzomorphan
Relationships
Alcohol Derivatives of the Oripavines
Compound 98 99 100
Structure-Activity
to naloxone
= 1. = 1.
Potency"
Writhing
Tail Flick
Antagonist Activityb
2.4 3.3 26 35 2.5 0.05 5.5 32 8.1 0.1 0.07 0.02
1.6 3.9 39 100 Inactive Inactive 0.2 58 Inactive Inactive Inactive Inactive
Inactive Inactive Inactive Inactive 1.0 0.2 Inactive Inactive 0.03 Inactive 0.5 0.7
Nitrogen Substitution
The effect of substituents in the Coring on induction of opiate antagonist activity in the N-substituted benzomorphans has been noted several times. The effect of substitution at position 3, the amino group, is therefore more complex than in either the morphine or morphinan series, because not only does the amino substitution have to be considered for mixed agonistantagonist activity, but the pattern of substitution in the other portions of the benzomorphan ring system must be considered as well. This section on nitrogen substitution consists of three subsections: N-alkyl substitution, unsaturated and cycloalkylalkyl substitutions, and furan and related substituents. Of the three, the second has been, by far, the most extensively investigated. 1. N-Alkyl Substitution at Nitrogen The investigation of the effect of N-alkyl chain length on analgesic activity in the 6,11a-benzomorphans was one of the earliest structure-activity studies undertaken in the benzomorphans (Table 6-11). (60). The 3-ethyl (114), 3-n-propyl (115), and 3-n-butyl (116) derivatives are devoid of analgesic activity, but the 3-n-pentyl homolog is as active as morphine. This activity parallels almost exactly that seen in the morphine series. Initially, the n-propyl derivative (115) was not investigated for antagonist activity. However, when the potent antagonist activity of n-propylnormorphine became known, npropylbenzomorphan (115) was reinvestigated; it proved to be a strong analgesic antagonist, even surpassing the N-allyl compound (144) (61). The influence of epimeric II-methyl groups has been studied in the 3propylbenzomorphans. For agonist, the ll-trans compounds generally have greater agonist activity than the cis compounds. The difference is less
274
6 The "Benzomorphans Table 6-11
Structure-Activity
"b Relative Relative
in the Benzomorphan
275
Analgesics
R
Analgesic Potency"
Thc Effect of II-Alkyl Substitution and N-Alkyl Length on Benzomorphan Analgesic Activity
Antagonist Activityb
R1
R2
Analgesic Potency"
~Hs ~Hs C2Hs C2Hs ~Hs C2Hs C2Hs ~Hs ~Hs /I-C3H7 fI-C3H7 fI-C3H7 ll-C3H7
H CH3 ~Hs /I-C3H7 /I-C4H9 rr-CSHII fI-C6HI3 rr-~H1s rr-CgH t7 CH3 ~Hs /I-C3H7 n-C6H13
Inactive Inactive Inactive 1.2 Inactive 0.5 Inactive 3.7 0.02 1.1 Inactive 0.13 Inactive 0.04 0.03 Inactive Inactive 0.02 0.44~ ",o..w-1 :. d Inactive Active Active 0.1
Compound 113 114 115 116 117 118
Relationships
Table 6-12
The Effect of N-AlkyI Length on Normetazocine Biological Activity
Compound
III
CH3 ~Hs fI-C3H7 rr-C.Hg /I-CsHIl /I-C3H7 II,B to morphine to nalorphine
0.7 Inactive Inactive Inactive 1.0 Inactive
119 120 121 122 123 124 125 126 127 ~128
7.0
2.4
= I. = I.
pronounced for antagonists. Indeed, the 11,B-methyl derivative (118) has only 35% of the opiate antagonist activity of its I la-methyl epimer (62). A similar study of alkyl chain length effects on nitrogen has been described for benzomorphans with 6-methyl and both II-ethyl and 11propyl substituents. The lla-ethyl series parallels the lla-methyl compounds, with the ethyl (121), propyl (122), and butyl (123) derivatives having weak or nonexistent analgesic properties but being rather potent antagonists (Table 6-12). Analgesic activity is restored with the n-pentyl derivative (124) and then slowly decreases as chain length increases (63). With the 1Ia-n-propyl substitution, the results are different, an example of the complexity inherent in the benzomorphans. The effect of Nsubstitution with 6-methyl-lla-n-propyl substitution does not follow the structure-activity relationship established in the preceding benzomorphans, the morphinans, or the morphine derivatives. Analgesic potency drops substantially as expected in the ethyl (129) and is eliminated in the 3-n-propyl (130), but is not restored with n-pentyl or n-hexyl (131) substitution; all the derivatives from ethyl (129) through n-hexyl (131) have narcotic antagonist properties. Higher homologs are inactive. As the length of the alkyl chain increases, so does the duration of antagonistic activity. The ethyl derivative (129) has a moderate duration of activity,
o/WI
(129 130 131 Relative b Relative U
C
d
to morphine to nalorpine
Antagonist Activityb
Opiate Receptor Affinity
= I, mouse writhing assay.
= I.
Nanomolar. Poor dose response.
while the n-propyl (130) has a very long duration. Other higher homologs are less potent than nalorphine but are faster-acting and of longer duration. The n-hexyl derivative (131) is somewhat different; its onset of activity is slower, but it possesses a very long duration of activity (64). Halogenated alkyl substituents on nitrogen have been synthesized for a variety of reasons. The first attempt was aimed at irreversible receptor labeling using 2-bromoethyl derivatives of normetazocine. The premise was based on the formation of covalent bonds with anionic receptor binding sites. Although the bromoethyl derivative was a more effective analgesic than the corresponding 2-hydroxyethyl derivatives, the significance of this effect could not be assessed, since the bromo derivatives had a sustained depressant action not observed in the hydroxyethyl compounds (65). A series of 2-fluoroethyl substituted normetazocines has been
276
6 The Benzomorphans
III
Structure-Activity
Relationships
Table 6-13
Table 6-14
2-Fluoroethyl Substituted Normetazocines
3-Aralalkyl Benzomorphans
in the Benzomorphan
Analgesics
277
N-(CH2)nR1
"'R3
H Compound 114 132 133 134
R H] HzF HFz F)
Relative Analgesic Potencya 0.03 0.13 0.12 Inactive. toxic
Affinityb
Compound
II
RI
R2
R)
Analgesic potencya
Reference
180 50 1Ooo 1700
137 138 139 140 141 142 143
2 2 2 2 3 2 2
C6HS C6Hs C6HS COC6HS C6HS C6H4-p-NH2 C6H4-p-OCH)
CH] CzHs 1I-~H7 CH) CH) CH) CH)
CH] CzHs CH] CH) CH) CH) CH)
4.8 0.6 1.3 0.5 0.1 11 3.8
69 38 70 7/ 72 72 72
a Relative to morphine = I, mouse writhing assay. b Receptor affinities in nanomolar.
prepared to test the effect of amine basicity on analgesic potency (66). The monofluoroethyl derrivative (132) is a more potent agonist than the almost inactive N-ethyl derivative (Table 6-13), and analgesia falls off as the degree of fluorination increased. The most active member of this series (132) has exceptional toxicity due to metabolic N-dealkylation and subsequent conversion of the fluoroethyl group to fluoroacetate, which is a suicide substrate for the Krebs tricarboxylic acid cycle (66). As with fluorinated side chains, an interesting hybrid between benzomorphans and the major tranquilizer butyrophenones has been synthesized. The resultant compound (135) is a potent analgesic (with 3.5 times the potency of pentazocine) with a significant amount of central nervous system (CNS) depressant activity (67). An interesting substituent is the w-cyanoalkyl group, which leads to compounds whose analgesic potency has a very strong dependence on the alkyl chain length (68). The most potent is the cyanoethylene derivative (136), which is 30 times as potent as morphine and does not substitute for
135
136
a
Relative to morphine
= 1.
morphine in drug-dependent rhesus monkeys. It is worth comparing 136 with the N-propyl derivative liS, which is a very potent antagonist without analgesic properties. A series of aralalkyl substitutions on normetazocine yields some potent analgesics, one of which is clinically useful (Table 6-14). Introduction of a phenethyl side chain yields a potent analgesic, phenazocine (137), which is five times as potent as morphine. Variation of the alkyl length at positions 6 and 11 (e.g., 138 and 139 decreases this activity, as does increasing the chain length (141) or introducing a keto-group (140). Substitution on the aromatic group, 142 and 143, retains analgesia. Phenazocine (137) is a clinically useful compound, although no longer marketed (73). It is an effective analgesic parenterally for most types of severe pain and orally for chronic pain. Tolerance develops more slowly to phenazocine than to morphine, and it appears to have less physical dependence liability (73). 2. Unsaturated Alkyl and (Cycloalkyl) alkyl Substituents on Nitrogen It is well known from the work on the morphine and morphinan series that the change from an N-methyl group in compounds that are opiate agonists to allyl, cyclopropylmethyl, or similar groups usually results in analogs that possess opiate antagonist activity. In the benzomorphans, antagonist activity in N-methyl substituted derivatives has been observed for the first time. It was therefore to be expected ttiat the substitution of the N -methyl group in metazocine by the usual antagonist functions
.
6 The Benzomorphans
278 Table 6-15 Unsaturated
Alkyl 3-Substituted
Compound 144 145 146 147 148 149 150 a
b
Analgesic Potency"
R CH2CH- CH2 (SKF-1O,047) CH20=CH CH2CH=CCI2 CH2CH=CHCI CH2CCI=CH2 CH2CCH3=CH2 CH2CH=C(CH3h (pentazocine)
Relative
to morphine
Relative
to nalorphine
Benzomorphans
Antagonist Activityh 1.9
o o o o o <0.1
0.1 0.02 5.6 0.02 1.0 0.02
= 1. = I.
would lead to interesting biological effects. This expectation has been realized as the generic opiate receptor has been subdivided into several types, primarily on the basis of results obtained with various benzomorphans. A series of substituted and unsubstituted N-allyl and propargyl derivatives of normetazocine is presented in Table 6-15. As can be readily seen, the allyl substitution (144) generates a potent antagonist without analgesic properties (74). Opiate antagonism is significantly reduced in the propargyl analog 145 as well as in the halogenated allyl groups 146 and 148. However, the 3-chloroallyl (147) and 2-methylallyl (149) substitutions retain potent antagonist properties (74). Probably the most important clinically useful compound in this series is the 3,3-dimethylallyl derivative (150), pentazocine (74). Pentazocine (150) is a weak analgesic antagonist, being only 1/50th as active as nalorphine in reversing the analgesic effect of morphine and phenazocine (137) (75). In humans, however, it is an effective analgesic for a wide variety of painful stimuli. Parenterally, 150 has one-third the activity of morphine, while orally it approximates codeine's potency. Pentazocine has a relatively rapid onset of action with a duration of 3-4 hours (76). Pentazocine has a low physical dependence
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
279
liability (77) and is therefore not covered by the Harrison Narcotic Act. However, psychotomimetic effects have been noted (78). Historically, the first subdivision of the opiate receptor into J..L,K, and a types was made o~ the ba.sis of animal behavior studies (79). The litany of subtypes has continued with D(80), A (81), and € and further division of J..L into J..L-land J..L-2(82). In particular, K and a opiates were shown to produce effects through mechanisms that must involve receptors distinct from the classic~l morp?ine J..Lreceptor. The K receptor ligands fail to suppress morphine abstmence and do not cause abstinence in morphine-dependent monkeys. They are also involved in diuresis and feeding behavior. The a receptor ligands mediate mania and psychomimetic effects (83). This receptor is e~t~er identical to or very strongly coupled allosterically with the phencyclidine receptor (84). The prototypic ligand for the a opiate receptor is N-allylnormetazocine, SKF-10047 (144) (79). Compound 144 causes mydriasis, tachypnea, tachycardia, hallucinations, and mania, effects considered to be mediated by the a opiate receptor. In rats and mice, 144 does not demonstrate analgesia in either the hot plate or tail flick assays. However, in the mouse writhing assay, N-allylnormetazocine is approximately one-half as potent as morphine and the analgesia is reversed by naloxone (85). As expected, the analgesic activity resides in the levoenantiomer (86). The affinities of 144 for various receptor subtypes have been determined (87). Compound 144 has high affinity for both the J..L (0.4 nm) and D (0.6 nm) receptors, where it acts as an antagonist and an agonist, respectively (87). The affinity of 144 for the a receptor (7 nm) is significantly less than that of the J..Lor D opiate receptor (85,88). This is a rather common OCCurrence where benzomorphans have relatively high affinities for several types of opiate receptors. Substitution on nitrogen with cycloalkyl groups, primarily with a methylene spacer between the nitrogen and the cycloalkyl function, has produced a highly interesting series of compounds. The 3cyclopropylmethylene derivative (151), cyclazocine, is a potent opiate antagonist that also possesses central muscle relaxant and tranquilizing properties in animal tests (Table 6-16) (74). Cyclazocine (151) has agonist properties in animals and is an effective analgesic in humans, being 40 times as potent as morphine (89). Although it produces addiction, the withdrawal syndrome in addicts is not as severe as that experienced with morphine (90). However, while cyclazocine is a very effective anaglesic in humans, it produces too high a level of psychotomimetic effects to be clinically useful (91). The cyclobutylmethyl derivative (152) is similar to cyclazocine (151) but has weaker muscle relaxant properties and is equivalentrto pentazocine (150) in rodent analgesic tests. In addition, 152 is an opiate antagonist of intermediate potency, being about one-fifth as
280
6 The Benzomorphans Table 6-16 Analgesic Activity of 3-Cycloalkyl Substitution Benzomorphans
~ o
in
N-R "'CH3
_ ~CH 3
HO
Compound
R
151
CHz-c-C3H5 (cyclazocine) CHz-c-C.H7 CH2-c-C5H9 CHZ-C-C6HII (CH2h-c-C3Hs c-C5H9
152 153 154 155 156
Analgesic PotencyO
Antagonist Activityb
0.07
4.3
0.11 0 0 0 0
0.22 0.29 0.006 0.88 0.54
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
281
pentazocine and cyclazocine are compared to the parents, there is virtually no difference in their biological properties (93). This is in contrast to the more striking differences seen in the metazocine analogs (d. Table 6-4). The O-methyl ether of cyclazocine (159) is more potent than pentazocine in rodent analgesic tests and, as expected, less potent than cyclazocine as an antagonist (95). A limited clinical trial with 159 was stopped after the usual side effects of cyclazocine appeared in an unusual time course (96). Psychotomimetic side effects did not appear until 4-12 hours after intramuscular administration and lasted for 2 to 31 hours. The reactions varied from emotional withdrawal, apathy, dysphoria, and communication difficulties to delusions, disorientation, spatial disorientation, and hallucinations (96).
157
-0
0
0.03
159
158
-Cl'lc(J
0
0.48
The alkyl substituents at positions 6 and 11 have been varied to determine the structure-activity relationships for opiate antagonism properties (Table 6-17). Significant enhancement of opiate antagonist properties over those of the parent 6,1l-dimethyl derivatives is seen with a propyl group at C-6 or an ethyl group at C-lI (62). A variety of 6-phenyl-ll j3-methyl benzomorphans with 3-antagonist substituents has also been shown to have antagonist properties with slight to no analgesic properties (28,45). Stereospecific opiate receptor binding has been observed with the benzomorphans using a brain homogenate (Table 6-18) (97). The pharmacological relevance of this affinity is supported inter alia by its ability to predict analgesic and antagonist potencies, as measured by other standard in vitro assays (98). As demonstrated in Table 6-18, all the pharmacologically active benzomorphans bind to the non differentiated brain opiate receptor in the morphine range. Besides the usual 6,1I-disubstitution patterns found in benzomorphans with nitrogen antagonist substituents, a limited number of examples with substitution at C-5 have been reported (Table 6-19). When disubstitution is restricted to positions 5 and 6, comparison of the SKF-10047 analogs 165 and 166 with pentazocine (150) and its 5,6-positional isomer (167) shows approximate equivalence in opiate antagonism (62). Inclusion of an
I, mouse tail flick assay. °b Rclative to morphine = Relative to nalorphine = I.
active as nalorphine (Table 6-16). The cyclobutylmethyl derivative (152) has also been investigated in humans, but it produces such intense psychotomimetic effects that its analgesic properties cannot be accurately assessed (92). The cyclopentylmethyl derivative (153) is still a potent antagonist, but, it has no agonist activity and lessened tranquilizing properties. Increasing the ring size to six carbons (154) causes a marked diminution of antagonist properties and elimination of the other properties. Increasing or decreasing the number of methylene groups between the nitrogen and the cycloalkyl groups generally reduces all the biological properties when compared to the cycloalkymethyl compounds. All of the compounds hitherto discussed are racemates. In cases where optical resolution has been carried out, the major portion of the antagonistic activity has been found in the levo-enantiomer. With pentazocine (150), the levo-enantiomer is 20 times as potent as the dextro, while the ratio for cyclazocine (151) is 500 (93,94). In contrast, when the ll-stereoisomers of
181
6 The Benzomorphans
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
183
Table 6-17 The Effect of 6- and ll-Substituents
on Opiate
Antagonism
Properties
Table 6-19 The Effect of Alkyl Substitution
at Position
5 on Antagonist
Properties
IItCH)
\;:d/\2
HO
HO Antagonist
Compound
R'
R'
R'
Potency"
144 160 161 150 162 151
CH,CH~CH, CH,CH=CH, CH,CH=CH, CH,CH=C(CII,), CII,CH=C(CH,), CH,-
CH, CzlIs n-C3H7 CH, CzHs CH,
CH, CH1 CH, CH, CII, CH)
2.4 2.3 10.2 0.03 0.01 4.3
163
CH,-
C,H,
CII,
2.3
164
CH,-
CH)
CzHs
22.4
a
Relative to nalorphine
Opiate Receptor Affinities for Benzomorphans N _R1 IIJR3
(-)-11 (+)-11 (-)-62 (-)-144 (-)-137 (-)-151 (ot)-150
R,
R,
R,
Potency"
165 166 150 167 151
CII,CH=CH, CH,CII~CH, CH,CH~C(CII,), CH2CH~C(CII,), CH,-
CzHs CzHs CH, CH, CH,
H CH, II CII, H
3.5 3.0 0.11 0.04 4.3
168
CH,-
CH,
CH,
<0.001
Relative to morphine
= 1.
= 1.
Table 6-18
Compound
Antagonist
Compound
R' CII, CII, CII, CH,-CII=CH, C,H,(CII,) CHz-c-CjHs CH,-CH~C(CH,), Morphine
R'
R'
ICsu(nm)
CH, CH, C,II, CH, CH, CH, CH,
CH, CH, CII, CH, CH, CH, CH,
30 >1000 14 2 0.6 0.9 15 3
additional methyl group at C-5, on the other hand, completely eliminates antagonist activity in the cyclazocine analog (168) (99). A series of 6,ll,ll-trisubstituted benzomorphans containing a variety of 3-substituents has led to compounds with surprising biological properties (52). These compounds, spanning a range of mixed agonist-antagonist and antagonist substituents, not only possess strong opiate antagonistic properties compared to nalorphine but are also active as agonists in the tail flick assay, where most benzomorphans show up weakly, if at all (Table 6-20) (53). One of the more interesting compounds to come out of this trisubstituted series is bremazocine (177), which contains the unusual hydroxycyclopropyl methyl group on nitrogen (100). Bremazocine (177) appears to be a
potent and long-acting
K
receptor agonist (100) that also exhibits some I-'
receptor antagonist properties (101). Compound 177 is a potent, centrally acting analgesic with a long duration of action, being three to four times as potent as morphine in the tail flick and hot plate assays. However, in the monkey shock titration test, 179 is 180-fold more active than morphine. Bremazocine is free of physical and psychological dependence liability and does not produce respiratory depression (100). On the basis of its
284
6
.
The Benzomorphans
Table 6-20 Agonist
and Antagonist
Properties
of
6-Substituted-l1,11-dimethylbenzomorphans
HO
RI
Compound
169 170 171 172 173 174 175 176 a
b
CH,CH~CH, CH,CH=CH, CH,C=CH CH,CH~C(CH3), CH,CH=C(CH3), CH;!.c-C3H5 CHrc-C3Hs CH2-c-C4H1
Relative
to pentazocine
Relative
to nalorphine
R' CH3
C,H, C,H, CH3 C,H, CH, C,H, C,H,
Analgesic Potency"
Analgesic Activityb
2.0 <2 <0.5 2.0 16 4.0 <2.0 16
15.5 7.8 3.9 <2.0 <2.0 15.5 31 7.8
= 1, rat tail flick assay. = 1.
(bremazocine)
177
178IR=CH3) 179
pharmacological
prototypic
K
profile,
bremazocine
(R = CICH3)3)
shows a promise
of becoming
a
receptor ligand (100).
A series of 6,11,11-trisubstituted compounds has been reported that . m~orporates the tertiary alcohol side chain at position 7 of the oripavine Dlels-:Alder adduct buprenorphine (48,56). The benzomorphan buprenorphme analogs 178 and 179, while more potent as opiate antagonists than buprenorphine, possess only a small fraction of the analgesic properties of the oripavines (48,56).
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
285
3. Heterocyclic Substituents on Nitrogen The interest in heterocyclic substitution on nitrogen in the benzomorphans originated with observations in the morphine and morphinan analgesic studies, where substitution of the methyl group by phenethyl and subsequently thienylethyl led to significant enhancement of analgesic potency. This area of study evolved when it was realized that some heterocycles bear a resemblance to the opiate antagonisl allylic group. For instance, a furylmethyl group's 11'electrons and oxygen lone pair are delocalized over the furan ring to form an aromatic system. However, the furan ring does exhibit residual olefinic properties and would be expected to resemble an allyl group (102). The thienylethyl substitution at position 3 forms a potent analgesic (180) when compared to either morphine or metazocine (72). Replacement with a furylelhyl (183) results in an analgesic 25 times as potent as morphine and without antagonist properties (Table 6-21) (103). Both 180 and 183 function as typical morphine-like JL-agonists because the heterocyclic olefinic contribution is an additional methylene group removed from the allylic posilion. In contrast to the thienylethyl derivative (180), the thienylmethyl compound (181) is not an analgesic and possesses some opiate antagonist properties (104). Replacement of the sulfur with oxygen yields the a-furylmethyl compound 182, which does have slight analgesic activity but is as potent as nalorphine as an antagonist (105). The corresponding f3-furylmethyl derivative 184 retains the antagonistic activity while losing the analgesic effects. More interesting are the methyl-substituted furylmethyl derivatives 185 and 186 (104,106,107). The derivative 186 is a mixed agonist-antagonist analgesic, while its positional isomer (185) is a potent agonist without antagonist properties. However, 185 does not exert the typical morphine-like side effects, including physical dependence liability. The furylmethyl analogs, particularly 185, display a selectivity toward K opiate receptors (108). The original premise that an appropriate heterocyclic methyl group can produce either mixed agonist-antagonist or antagonist properties seems to be established. A detailed structure-activity investigation has been conducted in the furylmethyl series (182, 184-186). In the 6-substituted compounds, increasing the 6-substituent from hydrogen through n-propyl results in small qualitative changes in either analgesic or antagonistic properties or both (104). Similarly, the difference in activity between the 11a- and 1113epimers of compounds 182 and 184-186 is small. The most striking effect is the sixfold decrease in analgesic activity on going from the a to the 13 conformation in both 185 and 186 (104). Variouscombinationsof methyl and ethyl substitution patterns at posilions 6 and 11 have resulted in little change in comparison to the respective parents (104).
286
6
The BenlOmorphans
Table 6-21
HI
Structure-Activity
Relationships
in the Benzomorphan
287
Analgesics
Table 6-22
Analgesic Activity Benzomorphans
of 3-HeterocycJic
Alkyl
Substituted
Analgesic
Activity
of 3.Suhstituted
Cyclic Ethers of Normctazocine
HO
Compound
Analgesic POlencya
R
Analgesia
Antagonist
Activityh
Antagonist
Compound
180
9.1
-(CH'h-Z)
None 187
181
182
183
-CH,-l:J
0.03
CH,
0.1 188 (MR-2034)
-CH,J:)
1.0 189
-(ClI ) -r-J
186
25
a
None
Inactive
0.12
~(5)
21
Activityb
0.17
31 r<./,+
CH,~
~(R)
1.0
CH;..Q
~""t..\.-qg(>r>\':s.t
Inactive
0.t8
O.ll
(5)
Inactive
9.2
0.53
(R)
O.t7
0.36
Inactive
(5)
0.34
1.1
Inactive
(R)
0.03
{J.{)6
Inactive
(5)
0.15
O.tO
Inactive
CH)
0.8
191
(CH,)_' 0 -H
192
(CH');'{~)
0 0.3
D
None
0.05
0
a
Relative to morphine = 1, mouse writhing test. b Relative to nalorphine = 1.
A variety of analogs containing a tetrahydrofuran ring in place of the fman ring have been prepared (Table 6-22) (/05). The compounds are agonists or mixed agonist-antagonists. The most interesting compound is 188, MR-2034, which is much more potent than morphine. The importance of tbe configuration of the newly introduced asymmetric center can be indicated by comparison with its relatively inactive diastereomer (187). Increasing the length of the alkyl cbain or introducing a methyl group at
193
194
a
Relative b Relative
(CIt')'H
D0
(CH,)""p to morphine = 1. 10 nalorphine = I.
-
Inactive
CH.,
190
-CIID CH1
(R)
'H
0
-CH'D
CHD -CH,
Writhinga
H
CH'''!:) 0
0 185
;:)
Platea
Hot
K-~.9.H'".~'''
" 184
None
S
R
,
I
288
6 The Benzomorphans
the new asymmetric center decreases the analgesic activity. The ll(:! isomer of 187 has 42 times the analgesic potency of 187, equivalent to 5 times that of morphine. The 11(:!isomer in the (S)-series of 188 is one-half as active as the lla isomer in the writhing assay (105). MR-2034 (188) has been studied extensively (109); despite its strong analgesic potency, it does not elicit typical morphine-like symptoms. It does not cause physical dependence liability in monkeys. On the basis of its pharmacological profile and its affinity for the K receptor, 188 has been labeled a K agonist. A thorough investigation of its neurochemical properties, however, indicates that MR-2034 binds extremely well to other subclasses of the opiate receptor, including ",-I, ",-2, U',and 6 (110). It has been called a universal opiate (100). The realization that the compounds listed in Table 6-22 are cyclic ethers has led to the further simplification of the side chain and a dramatic increase in potency. The ether 195, which possesses the two methylene groups between the ether oxygen and the basic amino-nitrogen, as well as the asymmetric center inherent in the tetrahydrofuran derivatives 187 and 188, is 124 times as potent as morphine in the writhing assay and 4 times as active as 188 (1/1). In contrast, the diastereomer of 195,196, is inactive. Further structure-activity investigations (Table 6-23) have reTable 6-23
3-Ether Substituted Normetazocine
Derivatives
III Structure-Activity
Relationships
in the Benzomorphan
Analgesics
289
vealed that the new asymmetric center is not important. The simple ether 198 is even more potent as an analgesic than 195. The requirements for potent agonist activity in this series are strict. Increasing or decreasing the alkyl chain on either side of the oxygen in 198 severely decreases potency, as does the addition of a second methyl group, 197 to 195 (/11,112). These open chain ethers share the same K agonist features as the cyclic tetrahydrofuran elhers. C,
The Inlroduction of Oxygen Containing Functions in the B- and C- Rings of Benzomorphans
1. Hydroxyl and Ketone Groups at C-I The C-I position in benzomorphans corresponds to C-IO in the morphine alkaloids, and of the few examples known from the morphine series, this substitution effectively eliminates analgesic activity. In the benzomorphans, I-ketones are present as part of the tetralone method of synthesis but are usually removed during conversion to the targeted benzomorphan. The preparation of the epimeric I-hydroxy compounds 204 and 205 was actually undertaken to study frozen conformations of noradrenergic ligands (113). However, the conversions illustrated in Scheme 6-13 are general for the benzomorphans. The benzylic position (C-I) in benzomorphans is readily oxidized to the ketone (203), which can be reduced to the la-alcohol (204) or, through a hydroxyl inversion sequence, to the 1(:!-alcohol (205). The 6-methyl analogs of 203 and 204 have been prepared but, as would have been expecled, have minimal analgesic activity (/14). The la-substituted-l(:!-
110
Compound
R
Configuration
Analgesic PotencyQ 206
(R.
CIl3)
207
(R = C6115)
CH, 195 1% 197 t98 t99 200 201 202 a
Relative
I
CH,CHOCH,
(5) (R)
CH,-C(CH,hOCH, (CH,hOCH, (CH,hOC,H, (CH,hOC,H, (CH,),OCH, (CH,),OCH, to morphine
= 1, mouse writhing test.
124 Inactive 3.1 157 7.1 0.7 1.6 0.5
alcohols 206 and 207 are formed by grignard addition to l-ketometazocine O-melhyl ether. In the mouse writhing assay, tertiary alcohols 206 and 207 have 2 and 10% of the activity of morphine, respectively (51). From the limited number of examples available, it seems that oxidation at C-l in the metazocine series does nothing to improve analgesic potency. An example of the preceding is shown in Table 6-24, where the J-ketometazocine derivative (208) has only about 10% of the activity of pentazocine (115). The other entries in Table 6-24 reflect the structure-
290
6
Relationship~
in the Benzomorphan
Analgesics
29t
l_Ketobenzomorphans
b
203
Structure-Activity
Table 6-24
32
III
The Ber1Zomorphans
204
d,e,f
HO
1
Compound
a
205 Scheme 6-13. Reagents: (a) Cr03, sulfuric acid; (b) sodium borohydride; (c) several steps; (d) CN8r; (e) 6% hydrochloric acid; (f) tosyl chloride, pyridine; (g) aqueous base; (h) lithium aluminum hydride.
activity relationships developed in the morphine, morphinan, and prior benzomorphan analgesics. Of particular note is 214, ketocyclazocine, which is an agonist of similar potency to cyclazocine but possessing antagonist properties similar to those of pentazocine. Ethylketocyclazocine (216) is a potent agonist devoid of antagonistic properties. Interestingly, the IIj3 epimer (217) of ethylketocyclazocine (216) is merely as active as pentazocine as an agonist and is twice as effective as an antagonist. Ketocyclazocine (214) and ethylketocyclazocine (216) possess a unique in vivo pharmacology that distinguishes these drugs from classical /J.agonists. Inter alia, they do not substitute for morphine and do not precipitate withdrawal in addicted rhesus monkeys. This led Martin to subdivide the opiate receptor and to designate ketocyclazocine as the prototype K ligand
b
R'
R'
R'
Analgesic Potency"
Antagonist Activityb
208 209 2tO 211 212 213 214
CH, (CH,),C,H, n-C3H, n-C)H7 CH,CH~CH, CH,CH~CH, CHrc-C)Hs (ketocyclazocine)
CH, CH, CH, C,H, CH, C,H, CH,
CH, CH, CH, CH, CH, CH, CH,
0.10 0.52 0.70 2.2 2.1 1.0 14
0.13 Inactive 3 t.45 4.0 1.3 2.0
215 216
CH2-c-C3Hs
CH,
CHz-c-C3HS
C,H,
C2Hs CH,
Inactive
217 218 219 220
(ethylketocyclazocine) CHrc-C)Hs CHrc.C3H5 CHrc.C3H5 CHrc.C4H7
0.52 49
C2H5 n-C3H7 C,H, CH,
fJ-CH, CH, C,H, CH,
1.0 1.5 15 0.56
'
2.0 18 Inactive
Relative to pentazocine (150) = 1, mouse writhing test.
Relative
to pentazocine
= 1.
(79). In general, K agonists lack both /J.and Ii receptor agonism. A general /J. antagonism is unlikely (116). Both 214 and 216 bind strongly to opiate receptors in homogenized rat brains, with ICso values of 18 and 9 nm,
respectively (97). However, besides having a high affinity for
K
receptors,
214 and 216 have moderate to high affinity for /J. and Ii sites (117). A neurochemical profile indicates that ethylketocycJazocine 216 is a /J.-2and Ii antagonist (118), which is a direct receptor antagonism and not a physiological one (116). Reduction of the I-ketone with palladium and hydrogen in acetic acid yields the Ij3-aJcohol. The 3-cycJopropylmethyl derivative (221) has 0.2 times the potency of pentazocine (150) as an agonist and 0.4 its potency as an antagonist. The 3-allyl derivative (222) is more potent, being 0.5 and 4 times as active as pentazocine as an agonist and antagonist, respectively. Both 221 and 222 are more potent as agonists and antagonists than the I-ketone (208) (119).
292
6
The Be'hzomorphans
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
293
Table 6-25 Analgesic
2. A Hydroxyl
Group
=
221
(R
222
(R = CH2CH=CH2)
HO
A hydroxyl group at C-1I in the benzomorphans is equivalent to a hydroxyl group at C-14 in morphine, a substitution known to enhance analgesic potency and decrease side effects. The generation of a hydroxyl group is acccomplished via the tetralone 223, which itself is readily available from ,B-tetralone (ef. Scheme 6-1). Addition of methyl grignard to quaternary amine (223) yields predominantly the tertiary alcohol (226) in the a-methyl series, resulting from equatorial attack of grignard (Scheme 6-14). Addition to the free base (224), on the other hand, yields the ,B-methyl series (225) resulting from axial attack (120,121). Catalytic
224
Scheme 6-14.
R'
R'
Analgesic Activity"
II 227 228 226 225 229 230 23t' 232'
CH, CH, CH, CH, CH, C,H, C,H, CH, CH,
CH, H OH CH, OH CH, OH H CH,
H OH H OH CH, OH CH, OH OH
0.7 0.02 Inactive 0.t7 0.20 0.t8 0.71 1.9 0.16
reduction of the ketones gives similar results, forming the epimeric secondary alcohols. The ketone at C-ll is hindered, and grignard reagents larger than methyl add either sluggishly or not at all (122). Biologically, the introduction of a hydroxyl group has been disappointing. Only the lIa-alcohol (230) is as active as metazocine (11) (120-122). A surprising result is that the secondary alcohol (231) in the phenazocine series is more active than the tertiary 6,1l-disubstituted compound (232) (Table 6-25) (120,121,123). The conclusion reached is that introduction of the equivalent of a morphine 14-hydroxyl group in the benzomorphans does not yield the potentiation of analgesic activity in the benzomorphan series. An extension of the above study had included the results of a systematic examination of the influence of epimeric ll-hydroxyl groups on benzomorphans substituted at C-3 with traditional mixed agonist-antagonist and antagonist groups. The observation has been made that hydroxylation generally decreases analgesic activity. When the hydroxyl has the a orientation, it has a slight effect on antagonist activity. When the hydroxyl group has the ,B orientation, it enhances the antagonist properties (124). A series of 6-allyl and 6-n-propyl-ll,B-hydroxybenzomorphans has been synthesized using the tetralone route (125). Both the allyl and
a,c,b>
HO 223
R'
Relative to morphine = 1, mouse hot plate assay. b 3-Phenethyl instead of 3-methyl.
22S
3
Compound
a
HO
=0
of I1-Hydroxybenzomorphans
CH2-C-C3HS)
at C-l1
+ /CH 3 N 'CH
Activity
226
Reagents: (a) methyl grignard; (b) O-demethylation; (c) pyrolysis.
l
294 6
Th:
Benzomorphans III
Table 6-26 Analgesic
Activity
of 6-AJlyl and
Structure-Activity
Relationships
in the
Benzomorphan
ProPyl-l1J3-hydroxybenzomorphans
Analgesics
295
N-CH2-<] "lOCH)
245 Compound
233 234
:
R' CII, CH2-c-C3Hs
R' CII,CII=CII,
R'
Antagonist Potencyh
II
6.03
II H H H CII, CII, CII,
<0.02
0.63 0.17 0.71 0.20 0.03 0.001 0.22
2.0 0.54 1.6 0.40 <0.02 8.2 1.36
0.001
0.75
235 236 237 238 239 240
CHrc-C4H7 CHrc-C3Hs CHrc-C4H7 CII, CHrc-C3Hs CHrc-C4H7
CII,CII=CII, CII,CII=CII, n-CJH7 n-C3H7 CH,CII=CII, CH,CII=CII, CH,ClI=CII,
241
CII,-Q
CII,CII~CH,
CII,
242 243 244
CII,-C=CII CHrc-C3Hs
CII,CII=CII, n-C)H7 n-C)H7
CH) CII, CII,
CHrc-C4H7
Analgesic Activity"
<0.001 <0.001 0.28
(moxazocine)
HEr
HO
247
is cleaved to yield the 6;S-diol (247). A series of esters of 247 have been prepared, but only the 6-acetate possesses significant analgesic activity (130).
0.70 13.2 0.84
Relat~ve to butorphanol == I, mouse writhing test. RelatIve to butorphanol ==I, butorphanol ==10 times that of morphine,1 times that of naloxone (ct. Table 5-18).
D.
A-Ring Substitutions and Replacements
The S-position in the benzomorphans corresponds to the 3-position in morphine; for this reason, most of the A-ring substitutions have occurred at this position. The relative analgesic activities for a series of metazocine derivatives is shown in Table 6-27. It is readily apparent that these analgesic relationships approximate those in the morphine series, but an oxygen atom at C-S is not necessary for opiate activity (132). The S-acylthio compound (250) is prepared by pyrolysis of the thiocarbamate derived from the S-hydroxy compound-the Newman-Kwart rearrangement (133). The synthesis of a series of derivatives of 250 bearing various N-substituents has been reported (132). These compounds collectively are reported to be strong analgesics compared to morphine, with lessened side effects or physical dependence liability. The metazocine derivative 250 is approximately as potent as metazocine but has only one-seventh of its opiate receptor affinity (134). A series of benzomorphans containing /luoro and chloro substitution at C-S are readily prepared by Grewe cychzatlOn (135). All of these compounds are less potent and more toxic than either the completely unsubstituted or S-hydroxy analogs (135). A nitro group can be readily introduced at C-S by nitration, and this can be reduced to the amine (6). In order to demonstrate definitively that nitration occurred
n-propyl g~oups, together with the lla-methyl group, can conformationally approxImate the C-ring. in the morphinans. The biological results pres.ented In Table 6-26 IndIcate that while opiate antagonist activity is retaIned when .co~pared to the morphinan butorphanol, the agonist properties are slgmficantly decreased (126). An lla-~ethoxyl group is present in moxazocine (245) (127). The ether (245) IS a mIxed agomst-antagonist that is equipotent to cyclazocine (151) as an analgesIc but has only about one-fourth its muscle relaxant and ~NS-depr~ssant properties (128). Moxazocine is effective clinically, causIng a I.ow IncIdence of nausea. Psychotomimetic effects are observed at 6-12 times the analgesic dose (129). 3. A Hydroxyl Group at C-6 A hydroxyl group can be readily . Introduced at C-6 In benzomorphans by cyclization of the benzylpiperidone (246) wIth re/luxIng hydrobromic acid. During cyclization the methyl ether
j
296
6 The Senzomorphans
III
Structure-Activity
Relationship!>
in the
Benzomorphan
297
Analgesics
Table 6-27 Analgesic Activity Benzomorphans
of 8-Substituted
a,b
251
R
d
Compound
9 11 248 249 250
R H OH 02CCH) OCH, SCOC,H,
Analgesic
Potency
a
0.08 0.7 l.Sb 0.21
Relative b C
Reference
to morphine
252
0.61<
Q
= 1, mouse hot plate assay.
Scheme 6-15. Reagents: (a) sodium-ammonia; (b) hydrochloric (d) acetic anhydride, acetic acid, hydrochloric acid.
add; (c) hydroxylamine;
131.
Mouse writhing test.
at C-B, cyclazocine was subjected to the Birch reduction to give the enone (251). After oximation, treatment under Semmler-Wolff conditions reforms the aromatic ring (252), which now contains an acetamide group at C-B (Scheme 6-15) (136). Hydrolysis of the amide to form the B-amino compound yields an analgesic with properties between those of pentazocine and cyclazocine. The replacement of the aromatic ring with a variety of heterocyclic rings has been achieved, but in many cases only the synthesis has been reported. For example, the pyridoazocine (253) has been prepared by a lengthy synthesis, but no biological evaluation has been reported (137). A thiophene ring (254) can replace an aromatic ring in the benzomorphans when a thienylmethyl grignard reagent is used in place of benzyl in the Grewe synthesis (138). An extensive series incorporating the thiophene ring has been prepared, but the compounds possess only weak analgesic activity relative to their toxicity and no morphine antagonism (138). A positional thiophene isomer (255) has been reported but not biologically evaluated (139). A thiazole ring analog of metazocine (256) and its UI'!-isomer (257) bas been reported (140). The Uer-methyl (256) is approximately equivalent to morphine as an analgesic, while the UI'! stereoisomer (257) has only 0.25 times the potency of morphine (140). Other more distantly related benzomorphans containing a thiazole ring have also been reported (141).
N - CH)
253
254
255
CH3
E.
256
(R1
CH3, R2
= H)
257
(R1
H, R2
CH3)
BC-Ring Enlargements and Contractions
Various skeletal modifications of the benzomorphan alicyclic rings have been reported in studies attempting to determine the ring effects on pharmacological activity. These include both B- and C-nng contractIons to
298
6
The Benzomorphans
III
Structure-Activity
Relationships
in the Benzomorphan
299
Analgesics
CH2C02CH3
a,b,c >
o
II
o
NH2"HCl
258
259
a,b
).
262
Scheme 6-17.
263
Reagents: (a) boron trifluoride, IlO.C; (b) LAH.
oxygenated derivative (265) (145). A simpler route starts from 7methoxytetralone and yields the 8-hydroxy co.~pound (266) (147). The phenol
260
261
Scheme 6.16. Reagents: (a) hydroxylamine; (e) B2H6; (f) Eschweiler-Clark reaction. form
norben20morphans
homoben20morphans, these syntheses morphan
have
and
ring
(b) H2/Pt; (c) hydrochloric
expansions
as well as both used
series, and some
methodology
to form
nor homo worked
acid;
B-
and
combinations.
been
found
actIvIty of morphme
(147).
A
(d) base;
Coring
Many
out in the parent
of the derivatives have
(266) has about 2% of the analgesic
facile route to C-norben20morphans has been descnbed;. however, It requires the presence of two aromatic methoxyl groups at posllIons 8 and 9 (148).
of
ben20-
to be potent
R
analgesics.
1. only
B-Norbenzomorphans been
cursively
The synthesisof B-norben20morphans
investigated.
The
first synthesis
danone aceticacid ester(258) (Scheme 6-]6). nished the amino ester(259), which cycli2ed
(260). Hydride reduction
of the
started
from
has an
in-
Reductive amination furto the tricyclicderivative
lactam carbonyl group and subsequent
N-methylation yielded the C-norben20morphan (26]) (142). The analgesic potency of 26] is equivalent to that of codeine. The 8-desmethoxy
derivative of 261 retains one-third of the analgesic potency of 261 (143). An alternative synthesis of this nor-ring system (Scheme 6-] 7) employs the substituted cychzalIon of
piperidone 262, which undergoes a Lewis acid-catalyzed to the hydroxy norben20morphan (263) as a separable mixture
methyl epimers. No biological activity has been reported for 263 (144).
2. C-Norbenzomorphans The C-norben20morphan (264) was prepared in a manner analogous to that of the B-nor compound (261) (145,146). The 8-methoxyl group was introduced by nitration and ulti-
mately by decomposition of the dia20nium salt to yield the requisite
264
(R = HI
265
(R = OCH3) (R = OH)
266
267
268
3 B- and C-Homobenzomorphans The B-homoben20morphans have bee~ little investigated. The ketone (267) has been reported without ~ny biological data (149). Another B-homo variant has be~n prepared usmg
the Grewe reaction on a ]-phenylethyltetrahydropyndme
mstead of the
usual I-benzyl derivative. The reported B-homo derivative (~68) possesses approximately 25% oftheactivity of penta20cine (150). ThIS IS especIally 'nteresting, since the phenolic hydroxyl group IS m an unusual posllIon. C-Homoben20morphans are readily produced uSI~g a chloropropylamine in place of the chloroethylamine in Barltrop s ongmal tetralone approach to ben20morphan (d. Scheme 6-1). The keto-C-hom~be~zomorphan has served as a precursor to a variety of C-homodenvalIves (Scheme 6-]8). (151). The chemistry of this ketohomoben20morphan is interesting because it does not follow that of the analogous
300
6
=0
The Btnzomorphans
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
301
Analgesic Activity of C-Homobenzomorphans
-->
-CH3
Compound
CH30
Analgesic
270 273 274 275
269
ActivityQ
0.13 1.2 0.41 1.1
1 Q
plate
Relative
to morphine = 1, hot
essay.
N-CH3
O-Demethylation yields the free phenols 273 and 274. Wolff-Kishner deoxygenation and O-demethylation furnish the monomethyl phenol (275). An alternative, more practical synthesis of the starting keto-C-homobenzomorphan has been developed (152). The tertiary alcohol (169), equivalent to the 14f3-hydroxyl group in morphine, severely decreases the analgesic potency of the molecule. The f3-methyl derivative (273) is more potent as an analgesic than its a-epimer (274), parallel to observations in the parent benzomorphans. The toxicity of the mono methyl C-homobenzomorphan (275) is so great that the analgesic activity reported in Table 6-28 is equivocal (151). The unsubstituted parent (276) of the C-homobenzomorphan has been synthesized without reported biological activity (153). A series of Chomobenzomorphans containing various substituents on the amine group
'CH3 CH30
270
275
271
(R
= CH3)
273
(R
272
(R
=
=
CH3)
274
(R
=
HI
HI Scheme
6-18.
ketobenzomorphan (151). Addition of methyllithium results in a-addition to yield 269. In the benzomorphans, grignard or alkyllithium addition forms the llf3-alkyl-lla-hydroxy function. The alcohol (269) is extremely resIstant. to dehydratl?n to the exo-methylene derivative (270) under a wIde vanety of condItIons. The desired derivative (270), however, is also formed usmg a methylene Wittig reagent. Catalytic hydrogenation of 270 gIves the f3-methyl (271) and a-methyl (272) compounds in 25 and 7% yield, respectively. The predominance of the f3-epimer is noteworthy, since the ll-methylene benzomorphan stereoselectively provides the a isomer. Hydrogenation of 270 under acidic conditions produces 271 in 85% yield.
276 has been evaluated pharmacologically (Table 6-29) (154). The N-allyl and N-dimethylallyl derivatives (277-279) have either no or weak analgesic activity and weak antagonist activity. On the other hand, the homocyclazocine analog (281) appears to be a pure antagonist with a duration of action equivalent to that of nalorphine. The alternative C-homobenzomorphan has been synthesized using a modified a-tetralone approach where the Coring is formed in an intramolecular Mannich reaction (Scheme 6-19) (155). The analgesic activities
302
6
The S;nzomorphans
III
Table 6.29
277 278 279 280 281 Relative Relative
in the Benzomorphan
Analgesics
30J
C. Homobenzomorphans
C-II Methyl Configuration
Compound
b
Relationships
Table 6-30
C-Homobenzomorphans
Q
Structure-Activity
Trans Trans Cis Trans Cis to pentazocine to nalorphine
Analgesic Activit yO
R C!I,CH~CH, CH,CH~C(CH,), C!I,CH~C(CH,), CHrc-C3Hs CHrc-C]Hs
o o 0.2 o 1.0
Antagonist
Potency
b
0.24 0.02 Very weakly active 1.1 3.7
Compound 283 284 28S 286 282 287 288
= I, mouse writhing test.
= 1. a
-CH3
~-
a
b
CH)O CH)
(CH2) 2NHCH)
CH)O
CH)
/CH)
I
c -->
Scheme 6-19. HI.
Reagents: (a) formaldehyde,
282 hydrochloric acid; (b) LAH: (c) phosphorus,
R' H OH OH OH OH 'OH OH
R' H H H H CH] CH] CH.~
R'
Analgesic PotencyQ
H H a-CH] fJ-CH, H a-CH3 fJ-CH,
0.08 0.41 0.44 0.55 0.63 0.85 1.1
Relative to morphine = 1, mouse tail cuff assay.
of the various ring-substituted derivatives are listed in Table 6-30. In the presence of the phenolic hydroxyl, the effect of varying alkyl substitution is not great, the most potent compound being 288, which is equivalent to morphine. Compounds 282 and 287-288 have been resolved and possess the expected properties (156). The levo-enantiomer of compound 282, eptazocine, has about one-half of the activity of morphine in a variety of analgesic tests, this activity being antagonized by naloxone. Eptazocine also antagonizes morphine's analgesic effect (157). This C-homobenzomorphan appears to be a mixed agonist-antagonist opiate (157). Marketing approval has been requested for eptazocine as an analgesic in Japan under the trade name Sedapain (158). 4. B-Nor-C-homo- and B-Homo-C-norbenzomorphans The B-norC-homobenzomorphans are of relatively recent vintage and have been prepared in the expectation of producing a strong analgesic having neither physical dependence liability nor psychotomimetic effects (159). The initial synthesis of this ring system employed a series of reactions previously unused in the benzomorphans (Scheme 6-20). A Diels-Alder reaction between benzyne, generated in situ, and cyclopentadiene forms the bicyclic system (289), which undergoes a [2 + 2J cycloaddition with isocyanate to
304
6
289
The Benzomorphans
III
Structure-Activity
Relationship!>
in the
Benzomorphan
Analgesics
305
290
HO 293
d
-->
292
e
--> If)
Scheme 6-20. Reagents: (a) cyclopentiaiene, (e) Na,SO,; (d) hv; (e) LAH; (f) alkyl halide.
291 magnesium; (b) chlorosulfonylisocyanate;
form the j3-lactam (290). Photochemical reverse [2 + 2] ring opening generates the functionalized B-nor-C-homobenzomorphan, which can be reduced to the parent B-nor-C-homobenzomorphan (291). A variety of substituents have been introduced on nitrogen, but the most active is the N-methyl (291, R = CH3), which has 20% of the activity of morphine without substituting for or antagonizing it (159). An alternative large-scale synthesis of the N-methyl-B-nor-C-homobenzomorphan (292) has been reported (160). To introduce the requisite phenolic hydroxyl, a more prosaic approach using a methoxylindanone has been used in an alternative B-nor-C-homobenzomorphan (161). Surprisingly, the phenolic derivative (293) is inactive as an analgesic. Derivatives with a variety of other agonist and antagonist nitrogen substituents are very weak analgesics when compared to morphine (161). The B-homo-C-norbenzomorphans have been even less thoroughly investigated. While successful synthetic approaches have been described, no biological data have been made available (162-/64). F.
More and Less Complex Benzomorphan
Analogs
Models of the opiate receptor at the molecular level explain agonist and antagonist effects by the interaction between receptor sites and the functional sites of the analgesics (165,/66). The direction of the lone pair on nitrogen is considered to be important in determining agonist and antagonistic properties, but the models are diametrically opposed in terms of direction. One model requires the lone pair to be equatorial (165), while the other emphasizes its axial direction (166). A series of rigid analogs
294
295
.
HO
296 (n=l) 297 (n=2) (294-297) of pentazocine have been synthesized (/67). Thes~ derivativ~s all possess rigid stereostmctures: the. piperidine rings ~re In the chair conformation with the nitrogen lone paIr aXIal. The bIOlogIcal properties of these molecules are, however, quite different (168). The analgesIc potencies of the seven-membered analogs 295 and 297 are less than those of pentazocine, while the six-membered ring analogs 294 and 296 are almost inactive. The latter pair have antagomst properties, whIle the former potentiate morphine analgesia. All have strong affimty for the opl~te receptor. The conclusion reached is that lone pal.r dlfectlOn. IS not cmctal for discrimination between agonist and antagomst properties. Benzazocines and benzazepines, which result when the ll-carbon atom in benzomorphan is removed, have been investigated as analgesIcs (169).
306
6
The Benzomorphans
III
Structure-Activity
Relationships
in the Benzomorphan
307
Analgesics
Table 6-31 Table 6-32 Benzazocine
Analgesics
Analgesic
Activity
and Benzazocine
of Benzazepine
Derivatives
HO
Compound 298 299 300 301 302 a
Analgesic R'
R'
CH, CH, CH, C6HS C6HS
CH, CH, CH, H H
R' CH, CH2-c-C3Hs CH,CH~C(CH,), CHJ CH2-c-C3Hs
Potency"
Compound
n
R'
R'
1.4 0.74 0.31 0.06 Inactive
303 304 305 306 307 308 309 310
1 1 1 2 2 2 2 2
CH, CH, CH, CH, CH, CH, CH, CH,
CH, CH3 CH3 CH, CH, H H H
Relative to morphine = 1, mouse writhing t~st. !>.,
The ring system is prepared by cyclization of N-dimethylallyl or cinnamyl dimethoxyphenylethylamine (170). Several of these flexible molecules show surprisingly potent analgesia (Table 6-31). Of the dimethyl derivatives, the N-methyl compound (298) is more potent than morphine; however, the mono derivatives have little, if any, analgesic activity (170). The requisite phenolic hydroxyl in the benzazocine and benzazepine ring systems has been prepared. The benzazepines initially were synthesized by the Bischler-Napieralski reaction (17l). The benzazepine analogs of cyclazocine (303), SKF- 10047 (304), and pentazocine (305) are all more potent analgesics than morphine (Table 6-32) (172). Antagonist properties have not been reported. On the other hand, the dimethyl substituted analogs in the benzazocine series (306-307) are relatively weak analgesics (173,174). Surprisingly, the monomethyl analog (308) is twice as potent as morphine (175,176). However, in this case, the cyclazocine (309) and SKF-l0047 (310) analogs are potent opiate antagonists with no apparent analgesic activities (169). Both 309 and 310 induce a phenomenom called ------the quasi-morphine withdrawal syndrome (177). These drugs, )r09 and 310...] produce the effects of morphine withdrawal; however, this occurs in nonaddicted, normal rhesus monkeys (177). G.
a
Analgesic Potency"
R' CH2-c-C)Hs CH,CH~CH, CH,CH=C(CH,), CH,CH=C(CH,), (CH,),C,H, CH)
1.6 1.0 1.7 0.25 0.16 1.0
CHz-c-C3HS
-' -'
CH,CH~CfI,
Relative to morphine:o::1, mouse writhing assay.
b Opiate
antagonists.
benzomorphans. On the whole, the biological results of this extensive investigation have been disappointing. One of the earhest IDvestIgatlOns listed the syntheses of five different posItIOnal Isomers (311-315) (178).
313
312
311
Nitrogen Movements within the Benzomorphan Nucleus Plus Nor and Homo Derivatives
A great deal of effort and ingenuity has gone into the synthesis of nitrogen positional isomers as well as the nor and homo analogs of 314
315
308
6
The Benzomorphans
III
Structure-Activity
Relationships
in the Benzomorphan
Analgesics
309
Table 6-33 Analgesic Activity of the Nitrogen Isomer of Benzomorphan
4-Positional
320
(R = CH3,
HO
CH1CH=CHZ'
CHZ-C-C3HS'
321
CH2CH=C(CH3)2)
HO Compound
316 317 318 319 " b
Relative 8-Desoxy
R
Analgesic
H a-CHJ fJ-CH, H' to codeine
Potency"
1.1 0.33 0.28 <0.18 = 1, mouse
HO
tail clip test.
compound.
The analgesic properties of derivatives of some of these positional isomers were subsequently reported. Derivatives of 315 containing an 8-methoxy group (benzomorphan numbering) and either 6-methyl (179) or 6,11dimethyl substituents (180) have been prepared. Compounds of this type have not been found to be useful analgesics. The positional isomer (311) in which the amine has been moved to position 4 has been more extensively investigated (181,182). As shown in Table 6-33, movement of the nitrogen from position 3 to position 4 seriously diminishes the analgesic activity, particularly since the analgesic activity is compared to that of codeine rather than morphine (183). Surprisingly, the dimethyl compounds 317 and 318 are almost equivalent in analgesic potency to the 8-desoxy monomethyl compound (319). Antagonist activity resulting from this substitution has also been disappointing (181). A further shift of the nitrogen to position 5, compound 312 with an 8-hydroxy group, yields an analgesic with one-half of the potency of codeine (147). A series of nor compounds in which the nitrogen has been moved to position 4 (320) possess little antinociceptive activity in the hot plate assay. Those wIth N-antagonist substitution also have little, if any, antagonist properties, but all are very toxic (184). The C-nor derivative 321 is about one-quarter as active as codeine in the tail clip assay (147). The C-homo
324
325
326
derivatives 322 and 323 have, at most, 10% of the activity of morphine (147,185,186). A variety of compounds more distantly related to the benzomorphans have utilized both nitrogen and bridge head position shifts. For instance, the bridged benzocyclooctane 324 has been reported (187). Compound 325 has about one-half the activity of penlazocine in the mouse writhing assay. The nor compounds with N-phenylethyl and dimethylallyl related to 325 are substantially less active (188). Variation of the nitrogen substitution in 326 does not yield a compound with more than 10% of the activity of cyclazocine as either an agonist or an antagonist (189).
310
6 The Benzomorphans
References
HO
HO 322
323
,
,
The addition benzomorphans)
IV.
of another nitrogen to the benzomorphan ring (azahas a deleterious effect on analgesia (190).
The Chemical Anatomy of the Benzomorphans
The tricyclic structure shown in Fig. 6-2 represents the absolute configuration of analgesically active enantiomers of the synthetic molecules. All known active compounds are levorotatory. The dextrorotatory enantiomers are either less active or inactive as analgesics. The benzomorphans have the same absolute configuration as morphine at atoms 2 and 6. On the basis of the structure-activity relationships developed in Section III, analgesic activity in the benzomorphans shows the following trends with functionalization: I. An aromatic A-ring with a phenolic hydroxyl at C-S displays analgesic activity, although the hydroxyl group is not necessary for analgesic effects. Methylation of the phenol reduces activity severalfold. 2. A lower alkyl substituent at C-6, preferably ethyl or propyl, enhances analgesic activity. CHz.CO
~,
Ln:"
Ketone at Co,
.'
, 'If
phenyl.
thienyl.
311
3. A lower alkyl group at C- II, usually j3-methyl or j3-ethyl, displays analgesic activity, the a-compounds being less active as a rule. The presence of a ketone y to the ring at C-l I can increase morphinomimetic activity up to IOO-fold. 4. Alkyl groups at C-6 and disubstitution at C-l I can increase analgesic activity several hundredfold. 5. The N-methyl-substituted benzomorphans, with or without alkyl substituents at C-6 and C-lI, can possess opiate antagonist activities. This activity resides in the levo isomer. In some cases, the dextro isomer can substitute for morphine. 6. Varying the nitrogen alkyl group length yields agonists at methyl, antagonists at propyl, and agonists again at pentyl. A methoxyethyl group increases analgesia to 157 times that of morphine. 7. Replacing the N-methyl with phenyl ethyl or heterocyclic ethyl can increase the analgesia by up to 25-fold over morphine. S. The use of unsaturated or cycloalkylmethyl substituents on nitrogen yields either mixed agonist-antagonists or antagonists. The N-allyl (SKF-IO047) serves as the prototypic ligand for the "opiate receptor that governs mania and other psychotomimetic effects. The dimethylallyl substitution gives pentazocine, which is a clinically useful mixed agonist-antagonist analgesic. 9. Introduction of a ketone at C- I with various N-substitutions forms mixed agonist-antagonists. Ketocyclazocine is the prototypic ligand for the opiate K receptor. The K agonists do not cause the physical dependence liability observed in the J1.agonists such as morphine. 10. Introduction of a hydroxyl group at C- II, analogous to the potencyenhancing C-14 group in morphine, is not consistent with strong analgesia. II. Ring contractions and enlargements usually decrease activity. Exceptions are some C-homo derivatives. 12. Movement of the nitrogen is not advantageous.
furyl
, , N'CH, ..-.. " '~
References
orfCHzhO-alkyl
, <
"~
,
,
:.
Mono or di.alkyl
lower
alkyl groups
substitution.
or fCH1hCOR
at
at C.G
Hydroxy group at C-B Fig. 6-2.
Analgesia potency-enhancing
substituents
on benzomorphans.
C-' 1
1. 2. 3. 4.
3. A. BarItrop, J. Chern. Soc. p. 399 (1947). D. C. Palmer and M. J. Strauss, Chern. Rev. 77, 1 (1977). J. Bosch and J. Bonjoch, Heterocycles 14, 505 (1980). E. L. May, J. Med. Chern. 23,225 (1980); N. B. Eddy and E. L. May, Science 181,407 (1973). 5. E. L. May and J. G. Murphy. J. Org. Chern. 20,257 (1957).
JI2
6 The Benzomorphans
References
6. E. L. May and E. M. Fry, J. Org. Chern. 22, 1366 (1957). 7. R. Grewe and A. Mondon, Chem. Ber. 81,279 (1948); R. Grewe, A. Mondon, and E. Nolte, Justus Liebigs Ann. Chern. 564, 161 (1949). 8. N. B. Eddy and E. L. May, in "Organic Chemistry. Vol. 8. Synthetic Analgesics" (D. H. R. Barton and W. Doering, eds.), p. 113. Pergamon, New York, 1966. 9. S. E. Fullerton, E. L. May, and E. D. Becker, J. Org. Chern. 27, 2144 (1962). to. W. Fedeli, G. Giacomello, S. Cerrini, and A. Vaciago, J. Chern. Soc. B p. 1190 (1970); W. Fedeli, G. Giacomelli, S. Cerrini, and A. Vaciago, Chern. Commun. p. 608 (1966); J. L. Karle, R. D. Gilardi, A. V. Fratini, and J. Karle, Acta Crystal/ogr., Sect. B 2S 1469
35. E. L. May and J. H. Ager, J. Org. Chern. 24, 1432 (1959). 36. S. E. Fullerton, J. H. Ager, and E. L. May, J. Org. Chern. 27, 2554 (1962). 37. K. C. Rice, A. E. Jacobson, and E. L. May, J. Med. Chern. 18, 854 (1975). 38. J. H. Ager and E. L. May, J. Org. Chern. 27,245 (1962). 39. J. H. Ager, S. E. Fullerton, and E. L. May, J. Med. Chern. 6,322 (1963); E. L. May and M. Takeda, J. Med. Chern. 13, 805 (1970). 40. J. H. Ager, A. E. Jacobson, and E. L. May, J. Med. Chern. 12,288 (1969). 41. K. C. Rice and A. E. Jacobson, J. Med. Chern. 19, 430 (1976). 42. N. Yokohama, F. B. Block, and F. H. Clarke, J. Med. Chern. 13, 488 (1970). 43. F. H. Clarke, H. Jaggi, and R. A. Lowell, J. Med. Chern. 21, 600 (1978). 44. N. Yokoyama, P.I. Almaula, F. R. Block, F. R. Granat, N. Gottfried, R. T. Hill, E. H. McMahon, W. F. Munch, H. Rachlin, J. K. Saelens, M. G. Siegel, H. C. Thomaselli, and F. H. Clarke, J. Med. Chern. 22, 537 (1979). 45. F. H. Clarke, R. T. Hill, J. K. Saelens, and N. Yokoyama, Adv. Biochern. Psychopharrn. 8, 81 (1974). 46. H. F. Fraser, D. E. Rosenberg, and H. Isbell, Committee on Drug Addiction and Narcotics, NRC-NAS, Minutes of 25th Meeting, Addendum 2, p. 1 (1963); reported in ref. 8, p. 153. 47. W. F. Michne, J. Org. Chern. 41, 894 (1976). 48. W. F. Michne, R. L. Salsbury, and S. J. Michalec, J. Med. Chern. 20, 682 (1977). 49. Y. Sawa, T. Kato, T. Matsuda, M. Mori, and H. Fugimara, Chern. Pharm. Bul/. 23, 1932 (1975). . 50. N. F. Albertson, W. F. Michne, and B. F. Tullar, J. Med. Chem. 21, 471 (1978). 51. A. Ziering, N. Malatestinic, T. Williams, and A. Brossi, J. Med. Chern. 13,9(1970). 52. P. A. J. Janssen, Adv. Biochern. Psychopharrn. 8, 109 (1974). 53. A. M. Akkerman and P. A. J. Janssen, U. S. Patent 3,764,606 (Oct. 1973); A. M. Akkerman and P. A. J. Janssen, U. S. Patent 3,883,536(May, 1975);A. M. Akkerman and P. A. J. Janssen, Gee. Offen. 2,027,077 (Dec. 1970); Chern. Abstr. 74, 125486 (1971). 54. M. W. Lobbezoo, W. Soudijn, and I. van Wijngaarden, Eur. J. Med. Chern. IS, 357 (1980). 55. Y. G. Gelders, C. J. DeRanter, and H. Schenk, Acta Crystal/ogr., Sect. B 835, 699 (1979). 56. W. F. Michne, J. Med. Chern. 21, 1322 (1978).
(1969).
11. S. Saito and E. L. May,). Org. Chern. 27, 1087 (1962); C. F. Chignelland E. L. May,1. Med. Chern. 8, 385 (1965). 12. E. M. Fry, J. Org. Chern. 28, 1869 (1963). 13. J. H. Ager, S. E. Fullerton, E. M. Fry, and E. L. May, J. Org. Chern. 28,2470 (1963); B. C. Joshi, E. L. May, H. M. Fales, J. W. Daly, and A. E. Jacobson, J. Med. Chern. 8, 559 (1965). 14. T. Kametani, K. Kigasawa, M. Hiiragi, T. Hayasaka, K. Wakisaka, F. Sateh, T. Aoyama, and H. Ishimaru, J. Heterocyc/. Chern. 8, 769 (1971); T. Kametani, K. Kigasawa, M. Hiiragi, F. Sateh, H. Sugi, and T. Uryn, J. Heterocyc/. Chern. 9, 1065 (1972); T. Kamatani, T. Honda, S..P. Huang, and K. Fukumoto, Can. J. Chern. 53, 3820 (1975). 15. L. Stella, Angew. Chern. 1m. Ed. Engl. 22, 337 (1983). 16. L. Stella, B. Raynier, anp J. M. Surzur, Tetrahedron Lett. p. 2721 (1977); L. Stella, B. Raynier, and J. M. Surzur, Tetrahedron 37, 2843 (1981). 17. D. L. Boger, M. Patel, and M. D. Mullican, Tetrahedron Lett. 23.4559 (1982). 18. K. Kanematsu, M. Takeda, A. E. Jacobson, and E. L. May, J. Med. Chern. 12,405 (1969). 19. E. L. May and M. Takeda, J. Med. Chern. 13, 805 (1970). 20. J. G. Murphy, J.H. Ager, and E. L. May,J. Org. Chern. 25, 1386 (1960); E. M. Fry and E. L. May, J. Org. Chern. 26, 2592 (1961). 21. S. Saito and E. L. May, J. Org. Chern. 27, 948 (1962). 22. J. H. Ager, S. E. Fullerton, and E. L. May, J. Med. Chern. 6, 322 (1963). 23. B. C. Joshi, C. F. Chignell, and E. L. May, J. Med. Chern. 8, 694 (1965). 24. A. G. Deneau and M. H. Seevers, Addenda to the Minutes of the Committee on Drug Addiction and Narcotics, N. A. S.-N. R. c., Washington, D. c., 1958-1965; quoted in P. A. Janssen and C. A. M. Van der Eycken, Drugs Affect. Cent. Nerv. Syst. 2, 39 (1968); Chern. Abstr. 69, 94663. 25. F. H. Clarke and F. B. Block, Neth. Patent Appl. 6,606,057 (November, 1966); Chern. Abstr. 67, 21854e (1967); U. S. Patent 3,449,332 (June, 1969). 26. G. N. Walker and D. Alkalay, J. Org. Chern. 31, 1905 (1966). 27. F. B. Block and F. H. Clarke, J. Med. Chern. 12, 945 (1960). 28. N. Yokoyama, P. I. Almaula, F. B. Block, F. R. Granat, N. Gottfried, R. T. Hill, E. H. McMahon, W. F. Munch, H. Rachlin, J. K. Saelens, M. G. Siegel, H. C. Tomaselli, and F. H. Clarke, J. Med. Chern. 22, 537 (1979). 29. T. Oh-ishi, A. E. Jacobsen, R. S. Wilson, H. J. C. Yeh, and E. L. May,J. Org. Chern. 39, 1347 (1974). 30. H. Inoue and E. L. May, J. Med. Chern. 19, 259 (1976). 31. R. D. Gless and H. Rapoport, J. Org. Chern. 44, 1324 (1979). 32. H. Inoue, T. Oh-ishi and E. L. May, J. Med. Chern. 18, 707 (1975). 33. A. E. Jacobson and E. L. May, J. Med. Chern. 7,409 (1964). 34. B. C. Joshi and E. L. May, J. Med. Chern. 8, 696 (1965).
313
I
~
57. W. F. Michne, T. R. Lewis, S. J. Michalec, A. K. Pierson, and F. J. Rosenberg, J. Med. Chern. 22, 1158 (1979). 58. W. F. Michne, T. R. Lewis, S. J. Michalec, A. K. Pierson, M. G. C. Gillan, S. J. Paterson, L. E. Robson, and H. W. Kosterlitz. Dev. Neurosri. (Amsterdam) 4, 197 (1978). 59. S. J. Ward, A. K. Pierson, and W. F. Michne, Life Sci. 33, (Suppl. 1) 303 (1983). 60. J. H. Ager and E. L. May, J. Org. Chern. 25, 984 (1960). 61. S. Archer, N. F. Albertson, L. S. Harris, A. K. Pierson, and J. G. Bird,J. Med. Chem. 7, 123 (1964). 62. N. F. Albertson, Adv. Biochern. Psychopharrn. 8, 63 (1974). 63. 1. M. Uwaydah, E. L. May, and L. S. Harris, J. Med. Chern. 20, 1374 (1977). 64. K. C. Rice, W. A Klee, M. D. Aceto, H. H. Swain, and A. Brossi, J. Pharm. Sri. 66, 193 (1977). 65. M. May, L. Czoncha, D. R. Garrison, and D. J. Triggle, J. Pharrn. Sri. 59, 884 (1968). 66. W. G. Reifeneath, E. B. Roche, W. A. AI-Turk, and H. L.Johnson, J. Med. Chem. 23, 985 (1980).
JI4
6 The Benzomorphans
References
67. H. Yamamoto, C. Sait~, S. Inaba, T. Inukai, K. Kobayashi, T. Fukumaru, Y. Koga, T. Honma, and Y. Asaml, Arzneim.-Forsch. 25, 795 (1975). 68. A. E. Jacobson, K. C. Rice, J. Reden, L. Lupinacci, A. Srassi, R. A. Streaty, and W. A. Klee, J. Med. Chern. 22, 323 (1979). 69. E. L. May and N. B. Eddy, J. O'g. Chern. 24, 1435 (1959). 70. J. H. Agee, unpublished; quoted in ref. 8, p. 135. 71. E. M. Fry and E. L. May, J. O,g. Chern. 24, 116 (1959). 72. M. Gordon, J. J. Lafferty, D. H. Tedeschi, B. M. Sutton, N. B. Eddy, and E. L. May, J. Med. Chern. 5, 633 (1962). 73. H. F. Fraser and H Isbell, Bull. Narc. 12, 15 (1960). 74. S. Archer, N. F. Albertson, L. S. Harris, A. K. Pierson, and J. G. Bird,i. Med. Chern. 7, 123 (1964). 75. L. S. Harris and A. K. Pierson, J. Pharmacol. Exp. Ther. 143, 141 (1964). 76. A. S. Keats and J. Telford, J. Pharmacol.Exp. Ther. 143, 157 (1964); R. Paddock, E. G. Beer, J. W. Bellville, B. J. CiJiberti, and W. H. Forrest, Jr., Clin. Pharmacol. Tha. 10, 355 (1969). 77. H. F. Fraser and D. E. Rosenberg, J. Pharmacol. Exp. Ther. 143, 149 (1964). 78. W. T. Beaver, S. L. Wallenstein, R. W. Houde, and A. Rogers, Clin. Pharmacol. Ther. 7,740 (1966); R. C. Hamilton, J. W. Dundee, R. S. J. Clarke, W. B. Loan, and J. D. Morrison, Br. J. Anaesth. 39, 647 (1967). 79. W. R. Martin, C. G. Eades, J. A. Thompson, R. E. Huppler, and P. E. Gilbert, J. Pharmacol. Exp. Ther. 197, 517 (1976). 80. J. A. H. Lord, A. A. Waterfield, J. Hughes, and H. W. Kosterlitz, Nature (London) 267, 495 (1977). 81. J. Grevel and W. Sadee Science (Washington, D. C.) 221, 1198 (1983). 82. G. S. F. Ling, J. M. MacLeod, S. Lee, S. H. Lockhart, and G. W. Pasternak, Science (Washing/on, D. C.) 226, 462 (1984). 83. R. S. Zukin and R. S. Zukin, Trends Neurosci. 7, 160 (1984). 84. R. S. Zukin and R. S. Zukin, Mol. Pharmacol. 20, 246 (1981). 85. G. W. Pasternak, M. Carroll-Buatti, and K. Spiegel, J. Pharmacol. Exp. Ther. 219, 192 (1981). 86. M. D. Aceto and E. L. May, Eur. J. Pharmacol. 91, 267 (1983). 87. K..J. Chang, E. Hazum, and P. Cuatrecasas, Proc. Natl. Acad. Sci. U.S.A. 78,4141 (1981). 88. D. M. Patton, Drugs Fut. 8, 439 (1983). 89. L. S. Harris, A. K. Pierson, J. R. Dembinski, and W. L. Dewey, Arch. Int. Pharmacodyn. Tha. 165, 112 (1967). 90. W. R. Martin, C. W. Gorodetsky, and T. K. McClane, Clin. Pharmacol. Ther. 7,455 (1966). 91. L. Lasagna, T. J. DeKornfeld, and J. W. Pearson, J. Pharmacol. Exp. Ther. 144, 12 (1964). 92. A. S. Keats and J. Telford, unpublished; quoted in L. S. Harris, Narc. Drugs: Biochem. Pha,rnacol. p. 89 (1971); Chern. Absl'. 79, 26977 (1973). 93. S. Archer, L. S. Harris, N. F. Albertson, B. F. Tullar, and A. K. Pierson, Adv. Chem. Sa. 45, 162 (1964). 94. B. F. Tullar, L. S. Harris, R. L. Perry, A. K.Picrson, A. E. Soria, W. F. Wetterau, and N. F. Albertson, J. Med. Chem. 10, 303 (1967). 95. D. M. Patton, Dmgs FU/. 3, 238 (1978). 96. W. T. Beaver and G. A. Feise, J. Clin. Pharmacol. 17, 480 (1977). 97. C. B. Pert, S. H. Snyder, and E. L. May, J. Pharmacol. Exp. Ther. 196,316 (1976); S. Snyder and C. B. Pert, Ann. Intern. Med. 81,534 (1974).
315
98. I. Creese and S. H. Snyder, J. Pharmacol. Exp. Ther. 194, 205 (1975). 99. R. Parfitt and S. Walters, J. Med. Chem. 14, 565 (1971). 100. D. Roemer, H. Buscher, R. C. Hill, R. Maurer, T. J. Petcher, H. B. A. Welle, H. C. C. K. Bakel, and A. M. Akkerman, Life Sei. 27, 971 (1980). 101. P. F. Von Voigtlander and R. A. Lewis, Fed. Proc., Fed. Am. Soc. Exp. Bioi. 41, 1314 (1982); P. F. Von Voigtlander, R. A. Lahti, and J. H. Ludens, J. Pharmacol. Exp. Ther. 224, 7 (1983); J. D. Leander, J. Pharmacol. Exp. Ther. 227, 35 (1983). 102. J. D. Leander and D. M. Zimmerman, Drug Dev. Res. 4, 421 (1984). 103. M. Gordon and J. L. Lafferty, U. S. Patent 2,924,603; Chem. Abstr. 54, P1855 (1960). 104. H. Merz, A. Langbein, K. Stockhaus, G. Walther, and H. Wick, Adv. Biochem. Psychopharmacol. 8, 91 (1974). 105. H. Merz, K. Stockhaus, and H. Wick, J. Med. Chem. 18, 996 (1975). 106. W. Warner and M. Puig, Drugs Fur. S, 137 (1980). 107. W. Warner and M. Puig, Drugs Fut. 5, 139 (1980). 108. C. B. Smith, Dev. Neurosci. (Amsterdam) 4, 197 (1978). 109. W. Warner and M. Puig, Drugs Fut. 5, 303 (1980). 110. N. Johnson and G. W. Pasternak, Life Sci. 33, 985 (1983). 111. H, Merz and K. Stockhaus, J. Med. Chem. 22, 1475 (1979). 112. H. Merz and K. Stockhaus, Dev. Neurosci. (Amsterdam) 4, 227 (1978). 113. J. Fauley and J. B. LaP;dus, J. Med. Chern. 16, 181 (1973). 114. E. L. May and J. G. Murphy, J. O,g. Chern. 20, 257 (1955). 115. N. F. Albertson, U. S',Patent 4,161,598 (July, 1979); Chern. Abstr. 91,175227. 116. P. L. Wood, Dmg Dev. Res. 4, 429 (1984). 117. A. Pfeiffer and A. Herz, Biochem. Biophys. Res. Commun. 101,38 (1981); J. Magnan, S. J. Peterson, A. Tavani, and H. W. Kosterlitz, Naunyn Schmiedeberg's Arch. Pharmacol. 319,197 (1982); P. L. Wood, S. E. Lane, and R. L. Hudgin, Neuropharmacology 20, 1215 (1982). 118. P. L. Wood, J. W. Richard, and M. Thakur, Life Sci. 31, 2313 (1982). 119. W. F. Michne and N. F. Albertson, J. Med. Chem. IS, 1278 (1972). 120. E. L. May, H. Kug;ta, and J. H. Ager, J. O,g. Chern. 26, 1621 (1961). 121. E. L. May and H. Kug;ta, J. O,g. Chern. 26, 188 (1961). 122. S. Saito and E. L. May, J. O'g. Chern. 26, 4536 (1961). 123. H. Kugita and E. L. May, J. O,g. Chern. 26, 1954 (1961). 124. N. F. Albertson, J. Med. Chern. 18, 619 (1975). 125. M. Saucier, J.-P. Daris, Y. Lambert, I. Monkovic, and A. W. Pircio,J. Med. Chem. 20, 676 (1977). 126. 1. Monkovic, Can. J. Chern. 53, 1189 (1975). 127. J. Castaner, Drugs FUl. 3,388 (1978). 128. T. A. Montzka and J. D. Matiskella, U. S. Patent 3,956,336. 129. R. J. Novcck and E. S. Caruso, Curro Ther. Res. 22,469 (1977). 130. M. Takeda and E. L. May, J. Med. Chern. 13, 1223 (1970). 131. S. E. Fullerton, E. L. May, and E. D. Becker, J. Org. Chem. 27, 2144 (1962). 132. A. Goldstein and A. Naidu, Biochem. Pharmacol. 27, 1033 (1978). 133. M. Hori, M. Ban, E. Imai, N. Iwata, Y. Baba, H. Fujimura, M. Nozaki, and M. Niwa, Heterocycles 20, 2359 (1983).
134. H. Fujimura, M. Nozaki, M. Niwa, M. Hori, E. Imai, and M. Ban, Dev.
Neurosci.
(Amsladarn) 4, 471 (1978). 135. A. E. Jacobson and E. L. May, J. Med. Chern. 8, 563 (1965). 136. M. P. Wentland, N. F. Albertson, and A. K. Pierson, J. Med. Chem. 23, 71 (1980). 137. J. Adachi, K. Nomura, and K. Mitsuhashi, Chem. Pharm. Bull. 24, 85 (1976). 138. T. A. Montzka and J. D. Matiskella, J. Heterocycl. Chern. II, 853 (1974).
1
316
6 The Benzomorphans
139. J. Bosch, R. Granados, and F. Lopez, J. Heterocyc/. Chern. 12, 651 (1975). 140. K. Katsuura, M. Ohta, and K. Mitsuhashi, Chern. Pharm. Bull. 30, 4378 (1982). 141. K. Katsuura and K. Mitsuhash!, fh~m Pharm. Bull. 31, 2094 (1983); K. Katsuura, K. Yamaguchi, S. Sakai, and K. Mitsuhashi, Chern. Pharm. Bull. 31, 1518 (1983). 142. A. E. Jacobson and M. Mokotoff, J. Med. Chern. 13, 7 (1970). 143. M. Mokotoff and A. E. Jacobson, J. Heterocyc/. Chern. 7, 773 (1970). 144. J. Bosch, M. Rubiralta, M. Moral, and J. Bolos, J. Chern. Soc., Perkin Trans. 1 p. 1459 (1984). 145. K. Mitsuhashi, S. Shiotani, R. Dh-uchit and K. Shiraki, Chern. Pharm. Bull. 17, 434 (1969). 146. G. N. Walker and D. Alkalay, J. Org. Chern. 36, 491 (1971). 147. T. Kometani and S. Shiotani, 1. Med. Chern. 2], 1105 (1978). 148. H. Finch, Tetrahedron Lett. 23, 4393 (1982). 149. G. R. Proctor and F. J. Smith, J. Chern. Soc., Perkin Trans. I p. 1754 (1981). 150. F. J. Smith and G. R. Proctor, l. Chern. Soc., Perkin Trans. I, p.2141 (1980). 151. M. Takeda and H. Kugita, J. Med. Chern. 13,630 (1970). 0). 152. M. Takeda, H. Inoue, M. Konda, S. Saito, and H. Kugita, J. Org. Chem. 37, 2679 (1972). 153. S. Shiotani, T. Kometani, and K. Mitsuhashi, Chern. Pharrn. Bull. 20, 277 (1972); S. Shiotani and T. Kometani, Chern. Pharrn. Bull. 21, 1053 (1973). 154. S. Nurimoto, S. Suzuki, G. Hayashi, and M. Takeda, lpn. l. Pharmacal. 24, 461 (1974). 155. S. Shiotani, T. Kometani, K. Mitsuhashi, T. Nozawa, A. Kurobe, and O. Futsukaichi, J. Med. Chern. 19, 803 (1976). 156. S. Shiotani, T. Kowetani, T. Nozawa, A. Kurobe, and O. Futsukaichi, l. Med. Chern. 22, 1558 (1979). 157. T. Nabeshima, S. Yamada, K. Yamoguchi, K. Matsumo, T. Kameyama, S. Sakakibara, and S. Matsumoto, Folia Pharrnacol. (lpn.) 81, 411 (1983). 158. Annnymous, Drugs FUt. 5, 255 (1980); Updates: ibid., 6, 323 (1981); ibid., 7, 352 (1982); ibid., 8, 453 (1983).
159. P. H. Mazzocchi and A. M. Harrison, l. Med. Chern. 21, 238 (1978). 160. P. H. Mazzocchi, E. W. Kordoski, and R. Rosenthal, l. Heterocycl. Chem. 19, 941 (1982). 161. P. H. Mazzocchi and B. C. Stahly, J. Med. Chern. 24,457 (1981). 162. G. R. Proctor and F. J. Smith, l. Chern. Res. Synop. p. 286 (1980); l. Chern. Res. Miniprinl p. 3544 (1980). 163. R. Achini, /lelv. Chirn. Acla 64, 2203 (1981). 164. S. J. Miller, G. R. Proctor, and D. I. C. Scapes, l. Chern. Soc., Perkin Trans. I p. 2927 (1982). 165. A. P. Feinberg, I. G. Creese, and S. H. Snyder, Proc. Notl. Acad. Sci. U.S.A. 73,4215 (1976). 166. V. M. Kolh, J. Pharrn. Sci. 67, 999 (1978). 167. M. Hori, T. Kataoka, H. Shimizu, E. Imai, Y. Suzuki, and N. Kawamura, Heterocycles 20, 1979 (1983). 168. M. Hori, T. Kataoka, H. Shimizu, E. Imai, Y. Suzuki, N. Kawamura, H. Fujimura, M. Nozaki, and M. Niwa, Chern. Pharrn. Bull. 31,2520 (1983). 169. Anonymous, Drugs Fut. 5,92, 94 (1980). 170. Y. Sawa, T. Kato, T. Masuda, M. Hori, and H. Fujimura, J. Pharrn. Soc. 95, 251 (1975). 171. M. Hori, H. Fujimura, T. Masuda, and Y. Sawa, l. Pharrn. Soc. 95, 131 (1975).
References
317
172. Y. Sawa, T. Kato, T. Masuda, M. Hori, and H. Fujimura, Chern Pharrn. Bull. 23, 1917 (1975). 173. Y. Sawa, T. Kato, A. Morimoto, T. Masuda, M. Hod, and H. Fujimura, J. Pharrn. Soc. 95, 261 (1975). 174. Y. Sawa, T. Kawakami, T. Hattori, T. Masuda, M. Hod, and H. Fujimura, Chern. Pharrn. Bull. 23, 2211 (1975). 175. H. H. Ong and E. L. May, J. Org. Chern. 38, 924 (1973). 176. M. E. Rogers, H H. Ong, E. L. May, and W. A. Klee, J. Med. Chern. 18,1036 (1975). (f7y R. J. Valentino, C. B. Smith, and J. H. Woods, Fed. Proc., Fed. Arn. Soc. Exp,. Bioi.
-
40, 1502 (1981). -'> Q_4'; - ~'1'p].,'>' "'I r<-.(K--&YC~< (I", ...,., "'h178. K. Mitsuhashi, S. Shiotani, R. Oh-Uchi, and K. Shiraki, Chern. Pharrn. Bull. 17, 434 ~I (1969). 179. H. Kugita and T. Oine, Chern. Pharrn. Bull. 11, 253 (1963). 180. M. A. Iorio and A. F. Casy, Gaz. Chim. ftat. 104, 1243 (1974). 181. W. K. Chang, L. A. Walter, and R. I. Taber, l. Med. Chern. 14, IOU (1971). 182. T. Kometani, S. Shiotani, and K. Mitsuhashi, Chern. Pharrn. Bull. 24, 541 (1976). 183. S. Shiotani, T. Kometani, and K. Mitsuhashi, J. Med. Chern. 20, 310 (1977). 184. P. H. Mazzncchi and B. C. Stahly, J. Med. Chern. 22, 455 (1979). 185. S. Shiotani, T. Kometani, and K. Mitsuhashi, l. Med. Chern. 18, 1266 (1975). 186. S. Shiotani, T. Kometani, K. Mitsuhashi, T. Nozawa, A. Kurobe, and O. Futsukaichi, J. Med. Chern. 19, 803 (1976). 187. K. Watanabe and T. Wakabayashi, l. Org. Chern. 45, 357 (1980).
T. Ma'suda, M. Hori, and H. Fujimura, Chern. Pharrn. Bull. 23, 1932 (1975). 189. W. F. Michne, J. Med. Chern. 19, 1159 (1976). 190. T. Kametani, K. Kigasawa, M. Hiiragi, and K. Makisaka, Hetrocycles 2, 349 (1974). 188. Y. Sawa, T. Kato,
II
7. Piperidine
Meperidine
Family
JI9
R
~,
Analgesics
e .g.,
I. Introduction . . . H. Meperidine Family. A. Synthesis . . . . . . . . . . B. Structure-Activity Relationship!> C. Clinical Utility ... . . . . . . . . . . D. Ring-Expanded and -Contracted Analogs . III. Bemidone Family A. Synthesis . .. B. Structure-Activity IV. Prodine Family .
...... Relationships
. . . . . . . . . . B. Structure-Activity Relationships V. Alkyl Family . . . . . . . A. 4-Alkyl-4-Arylpiperidines . B. 3-Alkyl-3-Arylpiperidines . . . . . . . . C. Ring-Expanded and -Contracted Analogs. . . D. Conformationally Rigid and Bridged Analogs. VI. Anilino Family . . . . . . . . . . . . . . . . A. Synthesis . . . . . . B. Structure-Activity Relati~nsh'ip~: G~n~r~ti~n 'of'C~~~u~d's Utility . . .. . ........... C. Rigid Analogs and Conformational Exploration References
A. Synthesis
~ilh Cli'ni~ai
318 319 319 321 328 328 331 331 332 334 335 337 352 352 354 356 359 362 362 363 366 367
I. Introduction One of the m~st intensely studied morphine-like synthetic analgesics, . mtroduced by Elsleb and Schaumann in 1939 (1), is ethyl I-methyl-4phe~ylpiperidine-4-carboxylate, known by 20 synonyms (2) although officially named meperldme and commonly called pethidine in the United States. Structure-activity relationships in the synthetic piperidine analgeSICShave been. explored at numerous positions of the molecule, leading to a most extensive data base that is still growing. Structural types can be grouped into five major classes: (a) 4-carbalkoxy, the meperidine family; (b) 4-ketoxy, the bel~ldone family; (c) 4-acyloxy, the prodine family; (d) 4-alkyl; and (e) 4-an.lhno, the fentanyl family, the later class containing the most potent synthetic analogs. The substituted piperidines are distinguishable from morphme by theIr structural simplicity and increased stereochemical flexibility, and have therefore spawned intense interest as recepJI8
oxygen -C02R,
heteroarornatic,
functionality: -OCOR, -COR, -01
/[J)... o
Ar
/ N
f
H, Me especially
Cn H2n
n
1.2.3
11 R
3-Me
Fig. 7-1.
. Me, alkyl, phenylalkyl aralkyl Structural modification of 4-arylpiperidine
analgesics.
tor probes, allowing the development of hypotheses about the structural requirements of the analgesic receptor. The first four family structural types have been extensively studied and reviewed (3-10a) for synthesis, pharmacology, and addiction liability. Major beneficial modifications of the general formula of the 4arylpiperidines are shown in Fig. 7-1, where the C-4 substituent is the determinant of the classification, R3 being primarily an oxygen-containing function: carbalkoxy, acyloxy, alkyl ketone, alkyl ether (or related alkyl). R1 includes methyl and other related alkyl and phenylalkyl groups. R2 is hydrogen and methyl. The aromatic group is unsubstituted, or contains a variety of functional groups, or is replaced by heteroaromatic groups. And n is 2, 1, or 3 for the classical six-membered piperidine ring and the contracted 5- or expanded 7-azacycloalkyl ring analogs.
II.
Meperidine Family
A. Synthesis The first synthesis of meperidine (4) was reported by Eisleb (11), with later modifications described by others (3,8,12). Through the years, the
320
7
Piperidine Analgesics
II Meperidine Family
321
R b (or
f)
y R 3
~) N ,
oE-
x-c n H 2n -Y-Ar
x-c n H2n -Ar
}
Cn H 2n -Y-Ar
R
~)
~~
N I CH2CH2-Ar
I CH2CH2-Het Scheme
3
4
R = R
= H,
R
1
Z = CH3,
R
=
CzHs
1 3 Z 5 R = R = R = H, R = CZHS of 4-carbalkoxy-4-arylpiperidines. Reagents:
Scheme 7-1. Synthesis R 'N(CII,CH,X,h; (b) NaNH" YCH,CH, Y;(c) R 'NH,; (d) 11,50" R'OH;(e) or hydrolysis; (0 NaNH2. R 'N-CH2CH2 Y
(a)
NaNH2.
bydrogenolysis
I CH(R')CH,Y,
original synthetic schemes (Scheme 7-1) have remained the principal routes to the meperidine family of compounds. The key starting material is the phenylacetonitrile (I), which is converted to the 4-cyano-4phenylpiperidine (2) either directly or via the dialkylated phenylacetonitrile derivative (3), where Y is a substituent replaced by a halogen or
7-2.
Synthesis
of normeperidine
analogs.
arylsulfonyloxy group prior to ring closure. Acid hydrolysis followed by esterification yields the final alkyl carboxylate. U R 1 is a benzyl or benzenesulfonyl group, removal by hydrogenolysis or hydrolysis results in the l-H (nor) compound (5), which can then be reacted with the appropriate functional groups, according to the conventional methods for alkylating amines, to give several classes of derivatives (Scheme 7-2). The" and f3 forms of N-substituted 3-methyl meperidines are prepared from the respective nor derivatives, obtained by separation of the racemic 4-cyano-4-phenylpiperidine intermediate. B.
Structure-Activity
Relationships
Exploration of substituents and their positions in the meperidine, or pethidine, series has focused primarily on the C-4 position, in various substitutions on the aromatic ring and in alteration of the oxygen
322
7
Piperidine Analgesics
functionality, and at the N-l position, in numerous replacements for the methyl group. The effect of replacing the carboxy and amino groups with heteroatoms, of changing both the size of and the substituents on the heteroatom-containing ring, and of introducing conformational rigidity have also been studied.
II
""'OW'" Meperidine 7 I CH3
1. Meperidine Modifications In considering structural modification of the C-4 aryl and C-4 carboxylic acid ester, optimum functionality for analgesic activity is found in meperidine, ethyl I-methyl-4phenylpiperidine-4-carboxylate, itself. General findings are as follows (4,13,14): 1. Substitution within the 4-phenyl group generally reduces activity, for example, o-OH, p-OH, p-NHz, the exception being particular ortho and meta substituents, for example, m-OH (bemidone family), which enhance analgesic activity. 2. Moving the phenyl group and/or carbethoxy groups to an alternate position reduces activity (15-17), for example, 3-phenyl (6, isopethidine), which is half as active as meperidine.
6 3. Changes to C-4 alkyl esters larger or smaller than ethyl reduce activity, for example, isopropyl or methyl, although an adamantyl ester is advantageous for increasing both potency and duration of . action (18). 4. Substitution of the C-4 ester with ketones or ami des or hydrolysis to the carboxylic acid reduce activity, the exception being a C-4 ketone wIth an m-OH phenyl substituent (ketobemidone family), although a tetrahydropapaverine amide (BG-9) has been extensively evaluated m the mouse hot plate assay, rat striatum, and guinea pig ileum; in the last, It shows both agonist and antagonist properties (19). Isosteric replacement of carbon by sulfur has led to a series of 4alkylsulfone meperidine analogs (7) (20), a number of which (R = ethyl, propyl, Isobutyl) are as effective as meperidine in mice (21,22). However, smgle Isosteric replacement of the ring nitrogen in meperidine by sulfur (8, R
~ ethyl)
leads to a loss of analgesic
activity (23). Addition
of an ethylene
bndge across the 2,6-positions (9), which then fixes the conformation
of
J2J
Meperidine Family
8
M
Z
N
C
N + (C) S
S C
R1 R R
R, R
1
= C6HS' 9
COZCZHS
R, R1 = C6H5' COZCZHS 10
the 4-substituents (24), has been reported to give a (3(exo) ethyl ester that is slightly more active (about 1.5 times) than meperidine in mice (25). The "ester (en do) is several times less active than the (3 and is more toxic in rats (26). The biological activity evoked has therefore been partially attributed to a conformation in which the aromatic and piperidine rings approach coplanarity (27), due to a flattened piperidine ring (26). In addition to the phenyl tropanes, the alkyl endo/exo-phenyl azabicyclo[2.2.1]heptane carboxylates (10), which have similar potencies in mice (28,29), have been used to investigate conformational properties of the 4-phenylpiperidine family. These series of rigid meperidine analogs have indicated that analgesic activity is not extremely sensitive to the conformation of the C-4-phenyl group. However, the biological response is dependent on the position of the aryl substituent, as evidenced by the narcotic antagonist, with a lack of analgesic properties, generated in a series of endo-carbomethoxy aryltropanes, wherein the aryl has been moved to the C-3-equivalent position (30). Meperidine (Demerol, Dolantin) is clinically useful in select situations for moderate pain management, such as in smooth muscle spasm, having a potency between that of morphine and codeine. Thus, meperidine is 1/IOth to 2/lOths as potent as morphine. This rank order of relative potency is generally consistent in various animal test protocols (31). In mice and rats, meperidine is one-fifth to one-fourth as potent as morphine (32); in humans, 50-100 mg meperidine is equivalent to 10 mg morphine (33). Meperidine's toxicity is low, and its duration of action is shorter than
324
7
Piperidine
Analgesics
that of morphine. Tolerance develops slowly, and addictive liability is lower than for morphme (9). However, morphine-like side effects such as respiratory depression, nausea, and vomiting are observable. Full clinical and pharmacokinetic profiles of pethidine have been reported (34-37). Cunously, mependme has a low affimty for opiate receptors, only 0.2% that of morphine (38,39), yet a hot plate analgesia about 10% that of morphine (32,40,41). Apparently, meperidine penetrates the brain mOre readily and reaches 600-fold higher brain levels than morphine (42). However, the concentration of each drug necessary to achieve halfreceptor occupancy (rat brain homogenate) corresponds to the brain concentration at half-maximal analgesic response. 2. N-1 Substitutions Normeperidine itself (5) is much less active than meperidine (43). Studies of norpethidine analogs therefore have been extensive, this N-] structural modification contributing the largest number of analogs to the total data base. Most analogs are synthetically available from norpethidine by alkylation with the appropriate alkylj aralkyl halide or by reaction with a suitably substituted aldehyde, aldehyde/ketone, epoxide, or olefin (Scheme 7-2). General findings are as follows (14) (Table 7-1): I. Several phenylalkyl or heteroalkyl substituents increase the potency over that of pethidine, for instance, phenylethyl (II, pheneridine), phenyl propyl (12), p-nitrophenylethyl (13), and p-aminophenylethyl (14, anileridine, Leritine). Anileridine is actually 3 times as potent as pethidine in the mouse hot plate assay, 10-12 times more potent in other animal assays (for example, rat radiant heat), and 2-3 times more potent in humans (34). ]n clinical use, equianalgesic doses of anileridine and morphine are 25 and 10 mg, respectively (51,53-55). Whereas in the unsubstituted phenyl alkyl series peak activity occurs at the 3-carbon chain length, in the p-amino and p-nitro phenylalkyl series peak activity occurs at the 2-carbon chain length. Chain branching gives inactive or weakly active compounds. Compounds with other substituents on the phenylethyl aromatic ring display potencies equivalent to that of meperidine or between those of meperidine and anileridine. N-acylanileridines have shown 4-40 times the activity of morphine in guinea pig ileum, but these activities do not correlate with their analgesic potencies (56). Interestingly, however, a fumaranilate derivative has shown a long-lasting antagonism to morphine analgesia, which suggests a high affinity for analgesic receptors (57). The demonstrated antagonism of morphine analgesia is unique, since no antagonism has been found for the fumaranilate in dependent mice or guinea pig ileum. The compound's ability to alkylate
II
Meperidine Family
325
Table 7.] Analgesic Activities of 4-Carbethoxy.4-arylpiperidines aC6Hs R-N C02C2HS Analgesic Activity" R 4<
5 11< 12 13 J4< t5 16' 17 18 19
CII, 11 C,II,(CII,h C,II,(CII,), p-NO,C,II,(CII,h p-NII,C,II,(CII,h C,II,NII(CII,h C,II,NII(CII,), C,II,CII~CIICII, (t) p-NII,C,II,CII~CIICII,
.
CII,(CII,),
(t)
Mice
Rats
Refercncesb
1.0 0.3 2.5 23 N.A/ 3 64 9 32 N.A. 7
1.0 0.5 2-3 13-20d 6 II 101 31 29-40d 12 7
II 43 44-49 48-50 48 20,48,50-52 58,59 58,59 48,49 48,50,61 32,41,62-64
f\
20
2.5
3-7
65-67
10 18 8
3_Sd 10 5-8
68-70 68-70 70-72
77
t8-28d
70,71,73,74
N.A.
24-33d
70,73,74
5 275 286
75-77 49,78-81 49,81,82
O,--,N(CH2)2 C,1I,O(CII,h C,II,O( CII,). C,II,CII,O(CII,),
21 22 23
~CH2O(CH2)2
24 25
~(CH2)4 26< 27< 28<
IIO(CII,hO(CII,h CoII,CO(CII,h CoII,CII(OIl)(CII,h
"References lOa,10b,I3. t; Clinical utility. d From
N.A. 74_106d 99-150d b Pharmacology several
laboratories.
and synthesis. ~Not available.
analgesic receptors selectively has been explored through its antagonistic property, 2, Two isomeric anilinoalkyl derivatives (IS, 16) have higher activities than anileridine (59), Piminodine (16, Alvodine), whose clinical potency is equivalent to that of morphine, has been marketed for postoperative pain; 7,5 mg is equivalent to 10 mg morphine (60).
326
7
Piperidine
Analgesics
Substitution of an amide linkage [C6H,NHCO(CnH2)n] for NH does not improve activity. 3. Two unsaturated I-cinnamyl derivatives (17, 18) are extremely active. Alkyl groups up to nine carbons, the C-1O analog being inactive, confer slightly greater activity than meperidine, with n-hexyl (19) being optimum. This homologous series shows a good correlation between affinity, in the presence of sodium, for opiate receptor binding sites and analgesic activity in the mouse hot plate assay (64). 4. An oxygen or sulfur functionality at a 2-4 carbon distance from the basic nitrogen center results in good analgesic activity. A 2morpholinoethyl (20, morpheridine) or 2-thiomorpholinoethyl substituent, and several oxyalkyl derivatives, those with 2-ethoxyethyl (21), 2-ethoxybutyl (22), and 2-benzyloxyethyl (23, benzethidine) substituents, have potent activities relative to meperidine. Morpheridine's analgesic activity in both rats and dogs is intermediate between those of morphine and pethidine (65); however, measurement of potency at 60 minutes shows morpheridine to be about 1.2 times as potent as morphine (70). The thialkyl analogs are much less active than the corresponding oxygen ethers. The phenoxy derivative is equipotent with benzethidine, but substituted phenoxy analogs are less potent than the unsubstituted parent. A quantum leap in activity occnrs when substituting a 2-tetrahydrofurfuryl group (24, furethidine) and replacing the ether linkage (25). Both tetrahydrofurfuryl derivatives have approximately 25-35 times the potency of meperidine. These results have indicated that the maximum analgesic activity can be found in an N-substituent of six or seven atoms or of chain length 7-9
A. Simplified
ethers,
however,
also have exploitable
biological properties, with the 2-(2-hydroxyethoxy)ethyl analog (26, etoxeridine) being commercially available in Belgium. 5. A series of aralkyl carbonyl- or hydroxyl-containing norpethidines, prepared from the reaction of norpethidine, formaldehyde, and an acetophenone 'or from condensation of an appropriate haloalkyl aryl ketone and norpethidine, exhibit the highest activity of all 4-phenyl-4-carbalkoxy derivatives. The 2-propiophenone derivative (27), with a 2-carbon alkyl chain, gives the maximum response at 60-200 times that of pethidine in mice and rats. Lengthening or shortening the alkyl chain or placing substituents on the aromatic ring generally decreases the potency. Ethyl is the optimum C-4 ester group. The secondary alcohol (28, phenoperidine, Operidine) obtained from reduction of the 2-propiophenone is slightly more active than the ketone, with a similar activity profile, and has found
II
Meperidine Family
327
clinical application in neuroleptanalgesia, the administration of a potent analgesic with a tranquilizer (83). The R-( + )-enantiomer is four times less active than the (- )-isomer but seven times more potent than morphine in mice (82). The acetate is less active than the ketone. Acetophenone and butyrophenone derivatives, ketones, alcohols, esters, and ethers, exhibit reduced activity compared to the propiophenones. The amino analogs of the reduced acetophenones are inactive (84). Since the number of chemical modifications in the series of N-substituted norpethidines is so large, a semiquantitative estimate of the influence of the above changes on analgesic potency in rats and mice has been made (49). Interestingly, replacement of the N-methyl group with allyl (29) or cyclopropylmethyl (CPM) (30) does not generate an antagonist, as is the
29
R = CH2CH=CH2
30
R
= CH2-<1
case in several morphine-based series (85-87). Only narcotic analgesia has been observed in these series. In fact, N-allylnormeperidine is as potent as meperidine, itself, having about 1/lOth the activity of morphine in rats (43). Methoxycarbonyl compounds with an N-I allyl substituent have also shown good analgesic activity, without any antagonism (88). 3. Ring Methylations One piperidine ring modification has been studied in some detail: addition of a 3-methyl group to meperidine and normeperidine derivatives. Several 3/3-methyl meperidines with various I-alkyl substituents have been synthesized (89), but their biological activities have not been thoroughly explored. This cis (/3) 3-methyl-4phenyl isomer of pethidine has been reported to be 8.5 times as active as the trans (ex) in mice (90) and 12 times as active in guinea pig ileum longitudinal muscle contraction (91). Alkyl 3/3-methyl carboxylates of the parent I-cinnamylnormeperidine (17) and the parent 1-(2-benzoylethyl)normeperidine (27) have been studied (92,93).
328
C.
7
Piperid1ne Analgesics
329
II Meperidine Family
Clinical Utility a
Compounds accepted for medical use include meperidine and nine N-substituted normeperidine relatives. Clinical potency parallels their animal potency (Table 7-1). The analgesic application and typical human dose administered, along with references for the date of introduction to the U.S. clinical market, have been summarized for each drug (5).
C6HS
,
(CH2)2
Ring-Expanded
(CH2)n
I
I
N(CH3)2
X 33
32
D.
X
CN
and -Contracted Analogs
Studies directed at the contracted 5-membered or expanded 7- and 8-membered ring analogs of meperidine have been neither extensive nor interesting (13). The pyrrolidine analog of meperidine (31) as well as C6HS
31 pyrrolidine derivatives are inactive (94,95). The 7- and 8-ring analogs are less active than the corresponding 6-membered piperidine analogs. 1. Synthesis Due to the lower activity of the azacyclopentanes, heptanes, and octanes compared to their piperidine counterparts, very few analogs have been synthesized. The principal syntheses (96-99) allowing variation for the introduction of ring alkyl, usually methyl, and N-I substituents are analogous to the sequence for the piperidine series and involve similar key intermediates. For example, in the ring-expanded N-methyl derivatives, 4-dimethylamino-2-phenylbutyronitrile (32) is alkylated with a 1,3-dihalopropane, followed by ring closure of 33 to yield a 4-cyano-4-phenyl methyleneimine (34) (Scheme 7-3). The nitrile is then hydrolyzed and esterified, giving ethoheptazine (35) and homologous esters. Using a dihalobutane in a similar process, the azacyclooctanes (36) are obtained (100). Treatment of the starting nitrile (32) with trans-2-butenal (crotonaldehyde) yields a dimethylaminohexanal that is converted via the alcohol and chloride to the 1,5-dimethylazacycloheptane nitrile (37) and ester (38) (Scheme 7-4). Using methacrylaldehyde or 3-buten-2-one (methyl vinyl ketone) in place of croton aldehyde generates the 6- and 7-methyl derivatives (39, 40) (99). The 2- and 3-methyl derivatives (41, 42) are synthesized
35
n=3,
36
n =4
R=C2HS
Scheme 7.3. Synthsesis of ring-expanded meperidine analogs. Reagents: X(CH,)"X, (b) beal, -200"C; (e) acid hydrolysis; (d) H,SO" ROH.
(a) NaNH2,
by using /3-dimethylaminoisopropyl chloride and separating the resulting 4-cyano-4-phenyl methyleneimines before completmg the Independent syntheses (101,102). Using substituted phenylacetonttnles m the above schemes, derivatives with substituents on the aromatIc nng can be synthesized (97). 2. Structure-Activity Relationships The prototype of the hexamethylene imine family, ethoheptazine (35, Zactane), IS less acllve than
330
7
a
Piperidioc
Analgesics
III Bemidone Family
JJI
b
--7
---+ CH3
c'"r:/
32 CH3 39 C6a
CH3
40
d <
(CH2)n
N-
~CH 3
'::0'" CH3
I
CH3 37
n=2
I
CH3 41
le'f
42
C02R
Ill.
CH3 (CH2)
n
Recognition of the structural similarity between the phenyljpiperidyl rings of meperidine and the AID rings of morphine has led to exploration of the effects of a meta-hydroxyphenyl substituent equivalent, to C-3-0H in morphine (111). The C-4 carboxy ester and ketone substituents with a meta-hydroxyphenyl constitute the bemidone and ketobemidone families. In rigid 4-phenyl axial morphinomimetics like morphine, a meta-OH is known to enhance analgesic activity. In the bemidones, this substitution has complex biological and conformational implications (112-114) and has produced both strong agonists and antagonists, depending on the N-I substituent.
N../ I CH3 38
n=2 Scheme NaNH"
7-4.
Synthesis
t-CH,CH=CHCHO;
of methylated
azocycloalkane
(b) NaBH,;
(e) SOCI,;
meperidine (d) heat;
analogs. (e) H,SO,;
Reagents: (f)
(a)
ROH.
meperidine in mice and about one-third as active in rats (IOa,103,104). It demonstrates no addiction liability (105) and is less toxic than codeine (98). Clinical interest (106-108) rests on oral efficacy against moderate pain in doses of 50-100 mg. It is commonly administered in combination with aspirin (Zactirin) (106,109). The 2- and 3-methyl derivatives of ethoheptazine are slightly more potent in the mouse (10a), with the 3-methyl (42) appearing to be promising in animals due to an analgesic potency greater than that of codeine or meperidine and few side effects (110). In humans, however, the methylated azacycloheptane analogs appear to be as active as ethoheptazine itself. Only ethoheptazine has undergone extensive investigation, including clinical study.
Bemidone Family
A,
.
Synthesis
The synthesis of the alkyl meta-hydroxyphenyl carboxylates follows that of Scheme 7-1, using the meta-methoxy substituted phenylacetonitrile and treating the resultant methyl ethers with refluxing hydrobromic acid to give the phenols. N-I substitution is accomplished by the usual methods from the norbemidones. From the 4-phenyl-4-cyano intermediate, reaction with the magnesium grignard generates the C.4 ketones. In the ketobemidone series, the N-methyl and nor-compounds are obtained by the ethylchloroformate method (115).
332
B.
7 Piperidin~ Analgesics
Structure-Activity
III
Bemidone Family
333
Relationships Table 7-2
1. Substitutions Resulting in Antagonism General findings concerning antagonism in the bemidone family are as follows (116):
Analgesic
Activities
of 4-ketoxy-4-arylpiperidines R2
1. The free meta-hydroxy group on phenyl promotes antagonist activity. 2. A C-4 ester or ketone, in conjunction, promotes antagonist activity, wIth the methoxycarbonyl being the optimal functional group. 3. N-substituents that are cyclic or acyclic but allyl-like promote antagOnism. Bemidone (43) itself is 1.5 times as potent as meperidine in mice (4,5), so that introduction of a meta-OH into the meperidine structure actually increases analgesic activity. The N-I allyl, dimethyl allyl, and CPM derivatives of bemidone fail to show much, if any, antagonism. The implication that the lack of antagonism in meperidine congeners is simply due to the lack of a meta-hydroxyl group on the C-4 phenyl ring is proved incorrect by these structure-activity relationships. Other structural features seem important for nalorphine-like activity, for instance, shortening of the ester chain. In the carboxymethyl series, the N-I trans and cis 3-chloroallyl, dimethyl. ally, and cyclopropylmethyl analogs show various degrees of antagonism and agonism without displaying many morphine-type side effects. The trans 3-chloroallyl (44) is the most active antagonist of the bemidone series. Several cyclic allyl-like substituents (for example, 45) are antagonists in mice and/or monkeys. The simple allyl and several furfuryl derivatives (46) have no antagonistic properties and show typical morphine-like side effects. Interestingly, N-I allyl and CPM derivatives of ketobemidone also behave like meperidine, having a morphine-like profile of action and failing to demonstrate any antagonist activity in morphinedependent monkeys.
R
HO 44 45
43
46
1 CH3' R = OCZHS
R = CH3, R 1 = CZHS 2. Substitutions in the Ketobemidone Cymidon), a 4-m-hydroxyphenyl-4-propionyl 47
Structure derivative
Ketobemidone of meperidine,
(47, is a
R
R'
R'
Analgesic Activity in Mice"
CH, C,H,(CH,j, c.H,NH(CH,), C,H,NH(CH,), C,H,CH=CHCH, CH,(CH,), - N- f<' \ CH,(CH,), C,H,OCH,CH(OH)CH,
C2H~ C,H, C3H7 C,H, ~C,H,
OH OH H H OH OH OH H
1.0 6 t 9 6 0.9 Qb 1.6 1.4
Meperidine
47' 48 49 50 51 @ 53 54
"References
JOa,b,J3.
C,H, C,H, C)H7
b Pharmacology and synthesis.
References b
8,Il7 Il8,Il9 Il8,Il9 131,120) 31,120 126
(' Clinical utility.
strong agonist, being 10-12 times as active as meperidine and showing no narcotic antagonism (8,34). Without the meta-hydroxyl, the compound is only half as potent as meperidine. Ketobemidone is used primarily in Europe in a dose of 5-10 mg in humans, since its potency and duration are at least equivalent to those of morphine. Studies on norketobemidone equivalents of normeperidines show (5) (Table 7-2): 1. The phenylpropyl substituent (48) decreases analgesic potency to the meperidine level but confers no antagonism or physical dependence capacity. 2. Two anilinoalkyl derivatives (49, 50), as the 4-propionyl and 4butyryl analogs, have greater activity than meperidine, but this substitution does not have the dramatic effect that is seen in meperidine congeners. 3. The N-cinnamylnorketobemidone (51) is equivalent to meperidine in analgesic potency. An ethyl group replacement for methyl almost abolishes activity, while larger alkyl groups show different effects of agonism and antagonism. Only a n-penty substituent (52) is stronger than
ketobemidone,
by almost
ree lImes;
-
exyl (53) and
334 7
Piperidine
AnalgesiC'S
IV
N-heptyl chains are nalorphine-like (with 8% and 4% the potency of nalorphine, respectively) antagonists with a long durations of action. In vivo potencies of the N-alkylnorketobemidones correlate well with mouse brain homogenate receptor affinities and antagonist activities in guinea pig ileum (/2l). Both agonist and antagonist potencies in various animal models correlate well for this homologous series (/22). A 2-cyanoethyl group abolishes anti nociceptive activity (/23), unlike the positive effects seen in morphinans and benzomorphans. 4. An oxygen functionality at a fixed distance from the ring nitrogen, as in 2-hydroxy-3-phenoxypropyl, 2-morpholinoethyl, and tetrahydro_ furfuryl compounds (/24,/25), does not dramatically increase activity. The 2-hydroxy-3-phenoxypropyl derivative (54) is slightly more active than meperidine and is equipotent with codeine as a cough suppressant. 5. Spiro analogs of ketobemidone, (55), obtained by reduction with sodium borohydride followed by treatment with hydrobromic/ acetic acids, as well as an isochromanspiropiperidine lacking any alkyl appendages, are devoid of analgesic activity in mice (/27,/28). They have very low affinities for opiate receptors in rat brain, but are interesting to study as receptor probes, investigating with the rigid system the steric bulk and conformational requirements of the receptor.
Prodine Family
JJ5
crease in activity relative to meperidine, regardless of the nature of the N-1 substituent (49). Investigation of analogs with 3-alkyl groups, ~Iong with all possible mono- and dialkylated analogs, has led to clarification of the stereochemical and conformational complexIty wlthm the senes and has allowed correlation of molecular geometry with analgesic potency among stereoisomers (/3l). A.
Synthesis
The synthetic route (/32-134) to the prodine family (Scheme 7-5), with or without a C-3 substituent, involves acylation, with an acid anhydride or chloride of a 4-aryl-4-piperidinol intermediate (58), obtained by addition of a lithi~m aryl derivative (preferred to an aryl grignard) or ~quivale~t to a 4-piperidone (57). Diastereomeric alcohols are produced If the plperidone has an asymmetric center, for example, at C-3. These can be
separated into the
Ct
(R3-aryl trans) and /3 (R3-aryl cis) forms before R
0 6R3 R3 N 11 R
a ----?
57 58 HO b 1 R .~
55
IV.
Prodine Family
The observation that reversal of the meperidine's ethoxycarbonyl group to propionoxy (56), described by Jensen and co-workers in 1943 (/29), leads to a 5-10 times more active analgesic (4,/30), has led to exploitation of these esters of 4-aryl-4-piperidinols, called the reversed esters of pethidine. This single modification generally accounts for a 20-fold in-
J.
OCOR2
"
R3
N 11 R
56
R = R3 = H, R1 = CH3, R2 = CZHS
Scheme 7-5. Synthesis of 4-acyloxy-4-arylpiperidines. Reagents: (a) LiAr; (b) (R2COhO, pyridine or R2eOCI.
336
7
A)
R
Piperimne
Analgesics
R
B) b ---c>
B.
60
61
58
Scheme 7.6. Synthesis of piperidino! intermediates. Reagents: (a) CHlO, R INH2; (b) R1NH2. CH2=CHC02C2Hs; (c) Na, xylene; (d) acid hydrolysis, heat; (e) NaH, R'X;
(f) dilute
Structure-Activity
Relationships
1. C-4 Aryl and Ester Modifications Considering structural modification and/or replacement of the C-4 aryl and reversed ester groups, the general findings are as follows (5,6,14,143,144);
--
57
337
The 4-acyloxy series of meperidine analogs has been primarily explored by modification of the reversed ester group, substitution on/replacement of the aromatic ring, and substitution in the heteroatom-containing ring, each aspect being considered alone or in conjunction with exploration of N-I substitution. Beyond this, the intermediate piperdinols and their alkyl ethers, along with ring-expanded, ring-contracted, and rigid analogs, have been studied.
R
e,f
Prodine Family
followed by hydrolysis and decarboxylation of the 3-carbalkoxy-4_ piperidone (Scheme 7-6, scheme B) (136-139). The iminodiester is synthesized from an amine and appropriately substituted acrylates, for example, ethyl methacrylate, followed by ethyl acrylate for the 3-methyl-4piperidone product. Using methyl crotonate in place of the methacrylate yields the 2-methyl-4-piperidone (140). This synthetic route is not efficient, however, for C-3 alkyl substituents larger than methyl (141). An alternative, more practical sequence for C-3 substituents involves alkylation of a 3-carbalkoxy intermediate (61), followed by dilute acid hydrolysis. Modification of the N-I substituent has been achieved either by choosing R' in the amine, or by substitution reaction on the I-H piperidinol intermediates (58, R' = H), or by an exchange reaction between the methiodide salt of the N-methyl piperidone and a primary base R'NH2 (142).
59
~
IV
HCI.
acylation if desired. When the lithium alkyl has bulky ortho substituents, for example, 2,6-dimethyl, only one isomer, the trans, is produced. The piperidinol or piperidine itself can be obtained by several different methods. The piperidinol synthesis of Schmidle and Mansfield (Scheme 7-6, scheme A), involving a 4-phenyl-4-methyl intermediate (59) synthesIzed from ~-methylstyrene, formaldehyde, and a primary amine, and g~vmg 30% YIeld, has produced unsubstituted piperidine rings (135). Using hIgher a-substituted styrenes yields 3-alkyl piperidinols. For piperidines with or without a C-3 methyl substituent, the 4-piperidone (57) can be obtained by Dieckmann cyclization of an iminodiester (60, R3 = H, CH3),
1. Isosteric replacement of the phenyl with 2-furyl, 2-pyridyl, and 2-thienyl (steric factor changes) decreases analgesic activity (145149), although two 4-ethoxy derivatives (62, 63) have, respectively, 3.6 and 1.6 times the potency of meperidine in mice and 1-3 times the potency of morphine. /"'1
o
62
# - form
338
7
Piperidine Analgesics
Table 7-3
IV Prodine Family
339
Table 7-4
4-Acyloxy-4-arylpiperidines
Analgesic Activities of 4-Hydroxy- and 4-Ethoxy.4-arylpipcridines R RI_N
C6AS
Q
R2
R'
Analgesic Activity"
References
OC2Hs OC2Hs OH OH 011 OH OH
O.6h 4 0.35 50" 6 150" 20
152 152 /34 156 156 157 157
R' Analgesic Activity R'
R
R'
Mice".b
56 64 65
CH, CH, CHJ
C2Hs CH, n.C]H7
H H H
66 67 68 69 70
C,H,(CH,1, C,H,(CH,1, C,H,(CH,1, C,II,(CH,1, C,H,(CH,1,
C2Hs C2Hs C2Hs C2Hs C2Hs
H o-CH] m-CH3 p.CH, o,p-CH.1
3.5 3.4 4.7 <0.3 0.5
66 71 72 73 74 75
C,H,(CH,1, C,H,(CH,1, C,H,(CH,1, C,H,(CH,), C,H,(CH,1, C,H,(CH,1,
C2Hs CH, CH, CH, CH, CH,
H H o-CH3 m-CH3 p-CH, o-DCH3
3.5 6.3 1.2 1.2 0.5 N.A.
Micet.".d
2.6 0.09 0.44
Morphine
Meperidine
= 1.
= 1.
C,H,(CH,1, C,II,CO(CH,1, C,H,(CH,1, C,H,N(COC,H,)(CH,1, C,H,N(COCH,)(CH,1, C,H,N(COC,H,)CH(CH,)CH, C,H,N(COC,H,)CII,CHCH,
134,152,156,157. "b References Mice, meperidine 1. Rats, d Mice, morphine 1. =
=
17 5.7 3.0 0.5 N.A." 3.0
"b References /34,143. d
76 77 78 79 80 81 82
t."Reference 152. .. Not available.
2. Replacement ~f phenyl with other unsaturated groups, groups that are nonaromatIc but have a 7r electron cloud, for example, olefinic, acetylemc, or. ~aphthylenic, decreases the analgesic profile, although hot plate actIvIty may be retained (147,150,151). 3. Changmg the reversed ester from propionoxy to acetoxy or butyryloxy (56, 64, 65. m Table 7-3) generally reduces activity, the exception perhaps bemg the N-! phenethyl derivatives (66, 71 in Table 7-3) (49,134,152). 4. Substitution within the 4-phenyl ring (66-70, 71-75 in Table 7-3) generally red~ces activity, although ortho-methyl and methoxy subslItuents retam substantial activity; para substitution usually causes more severe losses of activity than ortho substitution (134,152).
t."
meperidine
= 1.
5. Tertiary alcohol ('78 at 0.35 times the analgesic activity of morphine)
and ether (76, 77 in Table 7-4) intermediates are usually devoid of significant activity, even with potency-enhancing N-I or C-3-alkyl substituents (134,153-155); exceptions are certain N-acylated linear and branched chain N-!-phenethylamine piperidinols (79-82 in Table 7-4), two N-(o-chlorophenylethyl) 4-piperidinols with a C-4 N-disubstituted 2-propionamide group (4-propionoxy counterparts essentially inactive) (158), and a rigid analog, a 3-azabicyclo[3.3.!]nonane methyl ether (159). 2. N-l Substitutions Many N-! substitutions in the 4-acyloxy family parallel those of the meperidine family and produce similar trends in analgesic activity. General findings are as follows (Table 7-5): I. Several phenylalkyl substituents increase activity relative to meperidine, for example, phenylethyl (66), phenylpropyl (83), and phenylallyl (84). Peak activity occurs at a saturated chain length of three carbons (45,49). Two 2-oxazolidino-5-ethyl derivatives have increased activity, being equipotent with morphine in rats (161). A 2-indanyl substituent, representing a conformationally locked phenylethyl equivalent, results in retention of activity equivalent to that of meperidine (162) and of activity compared to loss 66. 2. The anilinoethyl derivative (85), analogous to the meperidine compound (15, Table 7-1), is extremely potent (1000 times the potency of
340
7
Piperidine
Analgesics
IV
Pradine Family
341
Tahle 7-5
4. As with meperidine analogs (28, Table 7-1), reduced derivatives of aralkyl carbonyl substituents exhibit the highest activity of all 4-phenyl-4-acyloxy compounds. While propiophenone (88) is about 1300 times as active as meperidine, the N-(3-phenyl-3-propanol) (89) is 3200 times as active. The 3-propionoxy derivative (91) is almost as potent as the alcohol, while the N-(3-phenyl-3-acetoxy) (90) shows a decrease in activity relative to the ketone.
Analgesic Activities of 4-Propionoxy-4-arylpiperidines DC6H5 R-N OCOC2H5 Analgesic Activity" R Meperidine
56 66 83 84 85 86 87 88 89 90 91
CH) C,Il,(CH,j, C,II,(CH,h C,H,CH~CHCH3 (t) C,H,NH(CH,j, C,H,N(COC,H,)(CH,j, F(CH,h C,H,CO(CH,j, C,H,CHOH(CH,j, C,H,C(OCOCH,)(CH,j,< C,H,C(OCOC,H,)(CH,j,
Mice
Rats
1 5-10 17-25< 162 261 N.A.d N.A. N.A. N.A. N.A. 120 1000-1500
1 26 110 572 1100 1301 3 2200 1346 3219 622 3040
Refercncesb
3. C-3 Ring Substitutions Most of the investigations of the piperidine ring have centered on substitution, especially with alkyl and aryl groups, at the C-3 position and determination of the conformation and stereochemistry within the resulting isomers (134,136,151,171,172). The relative analgesic activity of a- and ~-stereoisomers, along with differences in the antipodal forms of each, has been thoroughly studied (132). The racemic (d,l) forms of the diastereomeric a (trans) and ~ (cis) 3-methyl-4propionoxy-4-phenylpiperidines differ in potency by a factor of 5 relative to morphine in rats (173). Whereas a-methyl (92, alphaprodine), the most abundant synthetic isomer, is equipotent with the 3-H unsubstituted piperidine (3-desmethylprodine), ~-methyl (93, betaprodine) has at least three times the potency of morphine.
11 49,/29,/30 49,152 49 49,160 163 156 /65,166 163 163 49 49,/69,/70
a
References JOa,b,13. b Pharmacology and synthesis. d ~ Not available. As 3-acetoxy.
C
From several laboratories. C6XOC2H5 CH3
l )
meperidine in rats), although the acylated propionanilide derivative (86) has lost this remarkable potency. The anilinoalkyl compounds also lose activity when the nitrogen becomes tertiary, for example, by methyl and ethyl substitution (164). The sulfur isostere of 85 loses almost all activity (3 times that of meperidine), although the sulfoxide retains more (43 times that of meperidine) (163). 3. Several substituted alkyl derivatives, such as 7-fluoroheptyl (87) in the 4-propionoxy series and 6-chlorohexyl-, 6-nitrohexyl-, 8fluorooctyl-4-acetoxy piperidines, show enhanced activity relative to meperidine, usually 250-1000 times that of morphine. The Nisopropyl-4-propionoxy analog is several times more potent than morphine in rats (147,167), although isosteric replacement with dimethylamino to give a hydrazine derivative abolishes all activity. This is particularly interesting, since the 2-morpholinoethyl analog is active (168). In 4-acetoxy derivatives, the dimethylamino has only one-third the potency of the isopropyl isostere (134),which makes it inactive.
N I CH3
92
u
93
II
94 56
The influence of C-3 substituents as the a- and ~-isomers, however, does not always follow the trend of 3-methyl (174). However, the ~- (or cis) isomer is the more potent isomer of several 3-methyl 4-
I I
1
342
7
PipcJ.:idine Analgesics
Table 7-6 Analgesic
Activities
of Prodine
CH) -N
a-prodine
OCOC2HS CH)
93
.8-prodine
S6
3-dcsmethj'lprodinc
Isomer
C-3
C-4
(Oo)
RS S S RS S R
RS S R RS S R
(+) (-) (Oo) (+) (-) (-)
Analgesic Potency"
7.0" 13.3 0.6 34.3" 48.0 4.6 9.2
,.
Reference J76. h Mice, meperidine = l. " fJ/a -5.
carbalkoxypiperidines. The 4-acetoxy and 4-butyryloxy {3-3-methylprodines decrease in activity relative to the 4-propionoxy equivalents by factors of 4 and 3, respectively (24). !soprodines (94), isomers with the phenyl and reversed-ester groups moved from the normal C-4 position in the prodine structure (56) to the C-3 position, are inactive (137,175). Interestingly, the importance of the C-4 geometry is greater than that of the C-3, as demonstrated
343
in conjunction with N-I substitu-
C6H5
Configuration
92
Prodine Family
General findings for C-3 substitution, tion, are as follows (Table 7-7):
Isomers
q
IV
by studies of enantiomorphs
of
a- and
{3-prodine
(176). The two dextro (+ )-prodines, with different C-3 but identical C-4 configurations, have greater potencies than those isomers with identical C-3 and different C-4 configurations (Table 7.6). The most potent resolved form of betaprodine is the (+ )-35,45, at eight times morphine, and that of alphaprodine is the (+ )-3R,45, at three times morphine. Physical methodology has demonstrated a high skew-boat population in the more potent analgesics, which means that the best biological response is attributed to conformations in which the aromatic and piperidine rings approach coplanarity (24). Alphaprodine (92, Nisentil) has been used clinically (now withdrawn) in select situations. Its potency is greater than that of meperidine although less than that of morphine, in humans and animals, and its duration is very short (6,173,177). A 40- to 60-mg dose is equivalent to 100 mg meper. idine (5). Unfortunately, addiction liability is high (10a).
I. Certain alkyl substituents produce a high increase in activity over 3-desmethylprodine, for example, methyl, ethyl, and N-propyl. For 3-methyl, the {3(cis) geometrical isomer (93) demonstrates the better analgesic activity (24,178), as seen in 3-methyl congeners of meperidine, although for larger 3-alkyl groups, the a (trans) isomer is more potent (174). The a-ethyl (95, a-meprodine) is eight times as active as the {3(96, {3-meprodine) in mice, but biological activity can vary depending on the testing protocol, both isomers being equiactive in comparison to morphine (136). Interestingly, allyl represents the limit in size for C-3 alkyl substitution, the 3-crotyl derivative falling off in activity. While the {3-allyl (98) is only 1110th as active as morphine, the a-3.allyl compound (97, a-allylprodine) is 13 times as active as morphine and 10 times as active as alphaprodine in mice (136,137). It has 60 times morphine's potency in rats but behaves as a typical narcotic analgesic (172). The analgesic al{3 isomer potency ratio for the 3-ethyl, 3-propyl, and 3-allyl propionate pairs has been measured to be 8.8, 7.4, and 130, respectively (141,172). In the guinea pig ileum also, the a-allylprodine isomer has been found to have superior agonist potency (91). Stereochemical studies of the allylprodines (171), therefore, have shown that the stereochemical structure-activity pattern for 3-alkyl reversed esters holds: the a (trans) isomer is more potent than the {3(cis), with the exception of 3-methyl. This suggests a different mode of receptor interaction for the a-alkyl isomers than for {3-prodine. The carbon-carbon unsaturation in the allyl group has been thought to enhance the enantiomeric potency of (d,l)-a-allylprodine. In comparison to the low activity of the n-propyl analog, 97 shows a 24-fold increase in potency; comparison to morphine shows a IS.fold increase. Receptor affinity studies of 3-alkylated derivatives in rat brain homogenates have confirmed diastereoisomeric al {3 potency differences, with results correlating well with analgesic potencies (179). Higher receptor affinities are found in a-alkylated esters rather than {3 (higher alkyls and allyl), with the reverse situation existing for methylated piperidines. Studies on the activities of antipodal forms of racemic a- and {3-3-alkyl-4acyloxy-4-arylpiperidines have determined receptor discrimination based on the enantiotopic edge of the molecule (6,180; see also Chapter 8). In each case, the 45 configuration is more potent than the 4R. An "Ogston" effect has been postulated to be operative in the ligand-receptor interaction, the receptor having the ability to dis-
IV
~r!
~~ ........
~~~R:~~ .........................
~~~~~~~ .....................--
.
;;0" c <: ~c '60 0: '0
.
.~ :~ u <:
... .,u
.c
"-
:c
;;~~"
....
<:
'"
Prodine Family
HI
criminate configurational and conformational features in enantiomeric ligands (i.e., enantiotopic groups). 2. 3-Phenyl substitution in both a- and /3-prodine, and also on their 4-acetoxy analogs, destroys activity. Benzyl is also disadvantageous (173). 3. N-I alkyl substitution shows a reduction in activity, the isobutyl-3-H norprodine analog showing intermediate activity between meperidine and alphaprodine. N-I allyl substitution does not produce narcotic antagonist activity. Quite the contrary; 99 shows over a fourfold increase in analgesia compared to meperidine, with no nalorphinelike properties but a high physical dependence capacity in monkeys. A 3,3-dimethylallyl substitution, unlike its potency-enhancing effect in the bemidone series, causes a 75% decrease in potency relative to meperidine (181). Interestingly, a I,3-diallyl compound (100) seems to be an extremely active analgesic in mice. 4. N-phenylethylnorprodines (101,102) show increased activity relative to the prodines, with a relative potency similar to that of the prodines, the /3-isomer (101) being the more active, with 22 times the potency of morphine (24,134). 5. Substitution on the phenyl ring in the prodine series usually decreases activity, as in the 3-desmethyl series (Table 7-3), one exception being an artha-methyl in the 4-acetoxy analog (103), which is more active than the 3-methyl-4-artha-tolyl-4-propionoxy equivalent or the 3desmethyl-4-artha-tolyl-4-acetoxy derivative (72, Table 7-3). This compound has been tested clinically for its effect on postoperative pain at a dose of 3-4 mg (5), since it is not only three to four times as potent as morphine but shows separation of morphine-like activities (182). It is not clinically useful, however, since it produces severe respiratory depression. Replacement of phenyl by p-tolyl, o-tolyl, and m-tolyl gives progressively less active compounds (151). Hydroxyl substitution at the meta position in the prodines, the allylprodines, and the phenylethyl 3a-methyl compound (102) results in compounds that are inactive as agonists and antagonists in rats or mice (183,184). The phenolic derivative of 3-desmethylprodine is also inactive, in contrast to 56, which has about 10 times the potency of meperidine. The implication here is lack of a morphinomimetic receptor interaction in these synthetic analgesics, since hydroxyl substitution normally enhances analgesic activity in the rigid opiates. Activity differences between the phenolic arylpiperidines and morphine-type compounds have been attributed to divergent ligandbinding modes, whereby the phenolic piperidines mimic en kephalin rather than morphine binding (184).
346
7
Piperidiqe
Analgesics
IV
Prodine Family
347
Table 7-8 Analgesic
Activities
of Antipodes
of a-Pradine
CH)
Analogs
R
0
+
HC=C-CH=CH2
CH)
C6HS
I CH3-T-C=C-CH=CH2
OCOC2HS JR,4S
3S,4R 3RAS
Isomer
--->
CH)
C6HS OCOCiis
R
)_-
OH
Analgesic
Substitution
Antipode (-)35.4R (+)3R.4S (-)35.4R (+)3R.45 (-)3S.4R (+)3R.45 (- )3S.4R (+ )3R.45
H (desmclhyl) CHJ
a
pro-4R pro-4S
CH:H~
a
pro-4R
C.1H7
a
CH,~CHCH,
a
pro-4S pro-4R pro-4S pro-4R pro-4S
II Mice, subcutaneously. hot plate assay.
AClivity".h
1.0 0.06 1.4 0.04 0.94 0.03 0.85 0.11 28.3
Potency Ratio 45/4R-25 c <--
4S/4W28 4S/4R-25 45/ 4R -260
b References /71,187-189.
promedol Scheme 7-7. Synthesis of 2.S-dimethyl-4-piperidone. Reagents: (a) 50% H2S04;
(b)ltgSO,-H,o; (c) R'NH,. 6. N-I anilinoalkyl substitution in the prodines does not have the dramatic effect seen in the unsubstituted reversed ester of meperidine. N-methylanilinoethyl (104) increases activity relative to the prodines, but the potency of the N-ethylanilinoethyl drops by twothirds, intermediate between that of 104 and meperidine. 7. As with meperidine analogs and 3-desmethylprodine, the propiophenone derivative (105) is extremely active, almost 900 times as active as meperidine. 8. N-I alkoxy groups, such as methoxy (106) and ethoxy, decrease the activity of betaprodine to the level of meperidine, with other alkoxy groups being inactive.
diastereomer (6,14). Studies over the past 20 years have focused on the enantiotopic edges of the molecule (131): . . .(1) the relative configuration of the substituents in the pipe~idine ring, (2) the conformational equilibrium of each isomer, preferably In the protonated state as solute in water, and (3) separatIo~ of chlral dtastereoisomers into antipodal forms followed by estabhshment of the absolute configuration and the analgesic potency of each member of enantiomorphic pairs. The syntheses of the 2,3-, 2,S-, and 3,S-dimethyl ring analogs have used, with various modifications, 4-piperidone syntheses that Involve reactIOn of a substituted divinylketone and a primary amine, as shown for promodol (Scheme 7-7), and that result in a//3 mixtures. SynthesIs of the 2,6dimethyl ketone follows either from Manmch reaction of dimethyl acetonedicarboxylate and acetaldehyde or from catalytIc reductIOn and oxidation of the unsaturated 1,2,6-trimethyl-4(IH)pyridone and gives rise to different stereoisomer ratios. Synthetic methods for obtaining these methylated piperidines have been reviewed and summarized (131 and references cited therein). Interest in the 2-methyl, 3-methyl, 2,3-dimethyl, 2,S-dimethyl, 2,6dimethyl, and 3,S-dimethyI4-propionoxy arylpiperidines has led to specu-
4. Methylated Ring Analogs Resolution of the racemic a- and /3prodine family of compounds has shed light on the relationship of stereostructure to analgesic activity. Separation of a-3-methyl, 3-ethyl, 3-propyl, and 3-allyl isomers has shown that the more potent activity resides in the (+ )-3R,4S antipode, where the 3-substituent is equatorial on the pro-4S edge (Table 7-8). Other mono- and dimethylated derivatives, [2-, 2,3-, 2,6-, 2,S- (promedols), 3,S- (isopromedols)], display complex activities depending on the stereochemical relationships in the antipodes comprising each racemic
1
348
7
Piperidine
IV
Analgesics
Peadine
Family
349
Table 7-10 Analgesic
Activities
of Ring Methylated
Analgesic
Isomers of 4-Propionoxy-4-arylpiperidines
Activities
of Promedol
Isomers
C6"S
=2"S R R
Isomer
3-CH,
a
f3
2-CH,
a f3
2,3-CH,
a f3 y 8 Trans Cis Cis y(107)
2,6-CH)K
3,5-CH,
Meso Meso
Configuration
3e' pro-4R pro-4S 3. pro-4R pro-4S 2. 2e pro-4R pro-4S 2a,3e 2e,3e 2a,3a 2e,3a 2a,6e 2e,6e 2a,6a 3ea,5ea pro-4S,pro-4R pro-4S,pro-4R 3e, Se 3a,5a
a Mice, subcutaneously hot plate assay. t> a = axial, e = equatorial. C d Relative Relative to meperidine meperidine f
g
=
= 1.
1.
Antipode d,1 (-)3S,4R (+)3R,4S d,1 (+ )3S,4S (-)3R,4R d,1 d,1 (- )2S,4R (+)2R,4S d,1 d,1 d,/ d,1 d,1 d,1 dJ d,1 (+ )3S,5S (-)3R,5R dJ d,1
Analgesic Activity" 7.7c 0.6 14.4 40.9 52.4 4.0 O.65d.~ 0.65d....[ 0.75" 0.07 0.53" 0.03 3.0 Inactive 2.3Y 0.25 Inactive 2.29" 5.86 1.15 Inactive Inactive
to 3-desmethyl = 1. "-3.5
CH3
.D References
176,189 176,189 176,189 176,189 176,189 176,189 140,191 131,175,191
175 175 192 192 131,192 6,192 193 193 193 195 195 195 6 6
relative to
0.75 relative to 3-desmethyl, ref. 99. As the 4-acetoxy derivatives.
lalion on Ihe pharmacological significance of conformation relalive to a ~orphine-like or non-morphine-like binding mode (190). Biological activitIes of both racemic and. anlipodal forms of mono- and dimethylated 4-proplOnoxy-4-arylplpendmes have been measured relative to eilher the 3-desmethyl derivative or meperidine (3-desmethyljmeperidine, approxImately 10/1). The results from various laboratories are presented in Tables 7-9 and 7-10 (prodines included for reference), with Ihe general
Isomer
a'
f3 y(108)
S a
Mice.
Configuration
2a.5ac pro-4R, pro-4S pro-4S. pro-4R 2a.5e 2e,5e pro-4S, pro-4R pro-4R, pro-4S 2e, Sa subcutaneously.
Antipode
Analgesic
d,1 (+ )2R,4S,5S (- )2S,4R.5R d.l d,1 (+ )2S,4S.5R (- )2R,4R,55 d.l
26"
Activity"
12' 20 Inactive
8.1 2.9
1.3 1.3 0.t3 2.6
References
188.191 188.191 188.191 191,196 191,197 19I.I97 19I.I97 131
hot plate assay.
"a/I3/8 potency ratio= 9/3/1 (8/2.5/1 as acetates). C
d
a ::: axial; e = equatorial. Relative to meperidine = l.
~
Relative
to morphine
=
1.
Irends describing the relalionship of slereoslructure to analgesic activity discussed in Chapter 8. To resummarize Ihe potency of a- and (3-prodine, Ihe (+ )-anlipode with Ihe 4S configuration is Ihe more polenl of each resolved pair, wilh (+ )-3S,4S (3-prodine being more than 50 limes as potent as meperidine. The potency ratio of racemic (d,/) l3/a-prodine has been consistently measured as about 5.5/1, regardless of Ihe exact increases relative to meperidine (191). The racemic a- and (3-2-melhylpiperidines are equiactive, less active Ihan 3-desmethylprodine, but almost four limes as aclive as meperidine. The 4-aceloxy a or (3 race mates are much less aclive (140). The (3 (Ievo)-2S,4R antipode, equipotenl with the racemate, is the more active antipode by a faclor of 10 (175), which is twice as aClive as morphine. Opiale-binding sludies in guinea pig brain homogenales have shown a parallel potency difference, with ICso (3(- )/( +) values equal 10 9/1 (175). In Ihe racemic 2,3-dimethyl-4-propionoxy piperidines, Ihe diaxial (y) isomer is Ihree times as active as the reversed ester of meperidine, which is interesling, since this is 100 times the potency of the diequalorial ((3) isomer. The y slereoisomer has been reported 10 have 10-12 times Ihe
350
7
Piperidine
Analgesics
activity of morphine (6). A significant divergence is also found in the activities of the a and S racemates. The a isomer has three-fourths morphine's activity; the S isomer is inactive. The {3-2,3-dimethyl isomer has very low activity (practically none) due to unfavorable placement of the 2- or 3-methyl groups on either edge of the molecule (192). 2,6-Dimethylated analogs are mildly active. The cis-2,6-dimethyl-4acetoxy analog, which has a disadvantageous pro-45 equatorial methyl, as in (dextro) {3-2-methyl, has one-fourth the activity of meperidine. The only active racemate is the trans isomer, which is superior to the 3-desmethyl parent. The 2,S- and 3,S-dimethyl series of compounds (promedols and isopromedols) are highly active, with some variation in potency among the diastereomers, which have been studied in conjunction with differences in activity (131,190,194). Only one 3,S-dimethylated 4-propionoxy derivative, the y-isopromedol (107, Isopromedol), is more active than meperidine, by a factor of 2. The (+ )-35 ,S5 antipode is the more potent, equipotent with morphine, with the potency ratio of dextro to levo antipodes being S: 1. Of the four 2,S-dimethyl isomers, the most potent is the S, with a potency equivalent to {3-prodine and a conformation with the two methyl groups cis to the 4-equatorial phenyl. The importance of this conformation has been studied via rigid 2-azabicyclo[2.2.2]octane analogs, considered to be locked-boat conformers of S-promedol, since the 2,S-dimethyl groups are fixed in a cis-diaxial relationship (198). Racemic {3shows activity that is equivalent to that of the unsubstituted 3-desmethylprodine, approximately 10 times that of meperidine. The a-2,S-dimethyl isomer (a-promedol) has 12 times the potency of morphine, with the (+ )-2R ,45 ,S5 antipode being 20 times as potent as both the 3-desmethyl and morphine (almost twice the activity of the racemate). Curiously, the levo antipode is inactive. aPromedol is 10 times more potent than y-promedol, whose activity resides mainly in the (+ )-25,45 ,SR configurational isomer. The potency ratios for the a/{3/yracemates are 9/3/1, which remains constant in the 4-acetoxy derivatives as well (199). Interestingly, in the resolved promedols (a and
IV
CI
Prodine Family
35\
y), the more potent antipode has the 45 configuration, as does resolved {3-prodine. The racemic mixture (108, trimeperidine, Promedol) has demonstrated increased activity relative to pethidine both in humans and in animals. Clinically useful at a 10- to 20-mg dose subcutaneously, superior to pethidine in both potency and duration of action (197), it is used in the U.S.S.R. for smooth muscle spasms, spastic conditions, and obstetrics (14,34). Isopromedol, the stereoisomer of Promedol, however, has been reported to be a superior analgesic for smooth muscle pain. 5. Ring-Expanded and -Contracted Analogs The azacycloheptane (200,201) and azacyclooctane (100,202) analogs of the prodines have been synthesized, usually via the usual Dieckmann cyclization route through appropriate modification of the iminodiester intermediate. The pyrrolidine derivative and its analogs have also been made by Dieckmann cyclization of an iminodicarboxylate diester to the 3-pyrrolidone, followed by addition of an aryl lithium and acylation (203). An alternate synthesis involves a N-carbethoxy-3-pyrrolidone intermediate, obtained by reaction of an a ,{3-unsaturated ester and a N -carbethoxy'a-amino acid, which is converted to the 3-aryl-3-pyrrolidinol then reduced to the N-methyl analog and acylated (204,205). Modification of the sequences to yield the norpyrrolidine analog of prodine allows introduction of N-l substituents. Some ring-expanded analogs have been interesting pharmacologically, since they demonstrate clinically useful levels of analgesia with beneficial separation of other opiate activities (9,206). The 3a-methyl azacycloheptane analog (109, proheptazine) shows 10 times the potency of meperidine, which is the equivalent of the increase observed for the prodines compared to their 3-desmethyl counterparts (13). The 4-propionoxy, along with the 4-acetoxy analog, however, produces high physical dependence in monkeys (10a.b).
C H 2 XOC2HS CH3 CH3""l..)
109 N I CH3
107
(d,l)
lOB
(d,l)
In the pyrrolidine series of prodines. results generally parallel those of the related meperidine and prodine families (reviewed in 5,6). Racemic 2-methyl-3-propionoxy-3-arylpyrrolidine (110, prodilidine) is less potent than either meperidine or codeine, but its physical dependence capacity is
352
7
Piperidine
Analgesics
low (10a). The (+ )-enantiomer has 1.4 times the activity of the racemate by intraperitoneal injection, the (- )-isomer 0.6 times (207), although this varies on oral administration. Studies on modification of prodilidine have shown that the 2-methyl is the optimal alkyl substitution, the 3-propionoxy is the best ester group, and N-I substitution does not dramatically improve activity (208). In the last class, all compounds have activities less than that of prodilidine, with the most beneficial N-I substitution giving a paraaminophenylethyl derivative potency merely equivalent to that of prodilidene. Norprodilidine also maintains the activity of the parent, but it is much more toxic. Prodilidine (110) is the only compound that has been studied clinically (209), with several studies giving differing reports of its effectiveness. Nevertheless, 50- to lOO-mg doses have been used for ambulatory patients with musculoskeletal disorders. Orally, it has analgesic properties similar to those of codeine and greater than those of aspirin in potency, but it lacks aspirin's antipyretic and anti-inflammatory effects. It does not exhibit cardiovascular, respiratory depressive, antitussive, or constipating effects at normal doses.
V Alkyl Family
353
Table 7-11 Analgesic
Antagonist
Activities
of 4-Alkyl-4-arylpiperidines
011
Antagonist
Activity ADso..,b R lit 112 113 114 115 116 117 118
CH, CH, CH2CH~CH2 CPM< C6HsCH2CHz C6HsCOCHzCHz C6HsCOCH2CHz C6HsCOCH2CHz Nalorphine
Naloxone
C,
Form
Rats
Mice
{3 a {3 {3 {3 {3 {3 (3
d,l d,l d,l d,l d,l d,1 (+) (-)
0.24 33 0.47 0.72 0.11 0.056 0.023 0.05 0.40 0.022
1.0 39 0.72 0.72 0.t4 0.049 0.025 0.t4 0.45 0.079
" Milligrams per kilogram, subcutaneously. b Reference 211. C CPM = cyclopropylmethyl.
110
V. Alkyl Family A.
4-Alkyl-4-Arylpiperidines
Exploitation of the biological effects in the C-4-alkyl substituted 4arylpiperidines has provided a unique opportunity within the synthetic analgesics to study a combination of agonist and antagonist effects. Narcotic antagonist behavior has been discovered in derivatives wherein the C-4 position is substituted with alkyl and meta-hydroxyphenyl groups and the ring nitrogen with a methyl group, not exclusively with allyl or cyclopropylmethyl, as found in rigid opiates. Various modifications of the parent, the potent 1,3,4-trialkyl-4-phenylpiperidine prototype, have included substitution on the aryl ring, methylation at and N-I substitution
(210). Significant findings include the importance of a C-3-methyl group and its isomeric and antipodal forms in contributing to antagonist properties and potency differences (211). The 1,3,4-trialkyl-4-phenylpiperidines (Table 7-11) are pure narcotic antagonists (211 ,212), as judged by their response in the mouse writhing assay, the electrically stimulated guinea pig ileum, and tritiated naloxone binding in bovine brain homogenates, both in the presence and absence of sodium chloride. The nitrogen substitution that increases agonist potency in the meperidine family substantially increases antagonist potency. The N-methyl compound (111) has antagonist activity close to that of nalorphine both in rats and in mice; the allyl (113) and cyclopropylmethyl (114) do not have enhanced activity. However, phenylethyl (lIS) or propiophenone (1l6) substitutions show a lOO-fold increase in naloxone competitive 'binding and up to a 20-fold increase in potency compared to Ill, with 116 being equivalent to naloxone. The f3 (trans 3,4-dimethyl) isomer is over 50 times more active than the cis (a, 112). Resolution of the f3propiophenone shows the (+ )-isomer (117) to be two to six times as potent as the (- )-isomer (lIS).
354
7
Piperidine
Analgesics
Comparisons of activities of 1,4-dialkyl, 1,2,4,S-tetraalkyl, and 1,3,4,6tetraalkyl to those of the 1,3,4-trialkyl derivatives indicate that the antagonist activity is the result of a loss of intrinsic activity, not of receptor affinity, which has implications for conformational requirements, studied also in cis and trans phenylpyrindines (119) and 2-methylphenylmorphans (120) (213,214).
V Alkyl Family
355
Table 7-12 Analgesic
Activities
R
HO
120
C6H;<;COCH2
agonists, has equal affinity for both IL and I) receptors, whereas the (- )-isomer is a partial opiate agonist (213). Picenadol has a desirable clinical profile, with its good therapeutic index and low physical dependence liability, and is used clinically in obstetrics (215,216). 3-Alkyl-3-Arylpiperidines
Moving the C-4 substituents to the C-3 position (217) has allowed studies on 3-alkyl-3-(3-hydroxyphenyl)piperidines. both as ring-unsubstituted analogs and as C-2 or C-4 alkylated derivatives. These, plus 3-methoxyphenyl substituted equivalents, have shown divergent activities, depending on the N-l substituent and the C-3 alkyl group. The conventional synthesis involves hydrogenation of a cyano ester or ketone intermediate to give the cyclized 3-arylpiperidine, although other synthetic routes to 3arylpiperidines have been reported and reviewed (218,219). In C-3-methyl compounds, N-l substitution with large alkyl chains results in more potent analgesic activity, whereas in C-3-propyl derivatives, N-l preference is a small alkyl group (220). A 1,3-Dimethyl parent (122, Table 7-12) has only one-half the analgesic activity of codeine and
R'
Analgesic Activity EDso ".'
H a-CH3 JJ-CH, H H
50.6 Inactive
-66 -60 12.1 12.5 24.3
Meperidine
121
Changing a-methyl to a-C-4-n-propyl (C-4-n-propyljC-3-methyl cis) results in a mixed opioid agonist-antagonist (picenadol, 121), whose optical isomers show discrete separation of narcotic agonist and antagonist activities. The racemate is twice as active as meperidine and one-fourth as potent as morphine (subcutaneously or orally). The (+ )-isomer has a full opiate profile and is l/lOth as potent as morphine but, unlike other full
B.
CH, CH, CH, CH2~CHCH2
122 t23 124 t25 126
H
119
of 3-Alkyl-arylpiperidines
Codeine ,.
Milligrams
per kilogram,
/> Reference
mice,
subcutaneously.
220.
one-third that of meperidine, but has lower toxicity in mice. Methylation at C-4, either a or {3 (123,124), as well as eliminating the hydroxy group, decreases the analgesic response; N-allyl substitution (125) does not improve the potency. The N-l acetophenone derivative (126), however, has activity equal to that of meperidine and has a shorter duration of action than either morphine or codeine. Interestingly, its analgesic response is antagonized by nalorphine. In 2,3-dialkyl-3-arylpiperidines, N-phenylethyl (127) and Nacetophenone (128) derivatives are approximately half as potent as mo~phine in mice, with low toxicity (221). The 2-desmethyl analog of 128 tS
HO
127
R
128
R = CHZCOC6HS
129
R = CHZCH=CHZ
:
CHZCH2C6HS
130
356
7
Piperidine
equipotent with meperidine in mice. Both a (2,3-dimethyl trans) and {3 (2,3-dimethyl cis) phenylethyl compounds are active, 0.7 and 0.33 times as active as meperidine, respectively (6). Significantly, however, in this series, the N-allyl derivative (129) shows analgesia antagonized by nalorphine and compound 128, but is devoid of analgesic activity. The {3isomers of N-allyl and N-CPM 2,3-dimethyl-3-(3-methoxyphenyl)piperidines are more potent, with two to four times the potency of nalorphine (6). A report has confirmed the relative stereochemical assignments of the more potent a antipode, a(- )-I-allyl-2,3-dimethyl-3-(3-hydroxyphenyl)piperidine (222). Substitution of 3-methyl with 3-n-propyl or 3-benzyl in compounds 128 and 129 eliminates activity. Receptor affinity studies indicate that the active compounds in this series are pure agonists, the conformational adaptability of the 3-arylpiperidines being related to the rigid opiates (194,223). C.
V
Analgesics
Ring-Expanded and -Contracted Analogs Ring-expanded and -contracted versions of the 3-alkyl-3-arylpiperidines have yielded useful clinical prospects. The 3-ethyl-3-(3-hydroxyphenyl) azacycloheptane analog (130, Meptazinol), whose synthesis, isomer resolution, and clinical profile have been amply reviewed (224), has exploitable agonist-antagonist activities (225,226). The two optical isomers show a potency equivalent to that of the racemate, which is equipotent with pentazocine in rats and has a similar antagonist profile. Meptazinol has a low incidence of side effects, however, and has been used for postabdominal surgery and for patients with acute renal cholic. 11also has the benefits of rapid oral absorption and conjugation followed by excretion. Contraction of the azacylo ring to pyrrolidine (227) with 3-propyl-3-(3hydroxyphenyl) substitution produces an analgesic (131, profadol) 2.5 to 4 times as potent as codeine or meperidine in rats, which is a considerable enhancement over prodilidine (110), the 3-phenyl-3-propionoxy pyrrolidine. The side effects observed in various animal species, however, have been similar to those of prodilidine. The levo-isomer is twice as potent and toxic as the dextro-isomer, but it is the racemate that has been evaluated in humans. Profadol is the prototype of the 3-phenylpyrrolidine series in which an alkyl group replaces the previously exploited reversed esters related to the prodine family. Structural modification of the clinically useful analgesic have focused on the C-3 alkyl, hydroxyaryl, and nitrogen requirements for optimal antinociceptive activity, with the same modifications that produce antagonists in the {3-prodines producing antagonists in the 3arylpyrrolidines. General findings (relative to codeine) are as follows (228,229) (Table 7-13):
357
Alkyl Family
Table'.13 Analgesic
Activities
of 3-Arylpyrrolidines
I I
I
! ~i
R
R'
1 ~1
,
il
i,,
-f
'j
~;j'
110 131 132 133 134 135 136 137 138 139 140 141 142 "b
';1
t ,!
J
1 (
I
CH, CH, CH, CH, CH, p-NH,C,H,(CH,h CH, CH, CH, CH, (CH,hCH, (CH,),CH, (CH,),CH,
OCOC,H, (CH,hCH, CH(CH,h CH,CH(CH,h CH,C(CH,h (CH,hCH, CO,C,H, CH,CH~CH, (CH,hCH, CH,CH~CH, (CH,hCH, (CH,hCH, CH,CH(CH,h
Analgesic Potencyll'.b 0.2 2.5 3.0 2.7 3.8 5.8 0.4 1.2 1.3 0.9 Inactive 1.5 2.5
Rats, intraperitoneally. Relative
to codeine
= 1.
" References 227,228.
"
I
CH, H H H H H H H CH, CH, H H H
R'
f
I. Propyl and 3-hydroxyphenyl groups at C-3, as in profadol (131) itself, give the optimum typical antinociceptive response; limited chain branching, for example, isopropyl (132), isobutyl (133), 2,2dimethylpropyl (134), retains about the same level of potency; combination with N-methyl or beneficial N-l substitution, for example, para-aminophenylethyl (135), gives slight enhancement of analgesia accompanied by decreased lethality [note: methoxyphenyl with C-3propyl (138) loses activity relative to 132]. 2. Oxygenated (136) or unsaturated (137,139) functions at C-3 usually decrease activity in 3-hydroxy and 3-methoxyphenyls. 3. Substituted N-l phenylethyl substituents substantially increase analgesic potency; N-propyl substitution (140) results in a loss of activity; an increase in linear chain length to C-5 (141,142) again results in analgesia, but with a potency less than or equivalent to that of the N-methyl, a trend analogous to that observed in normorphine N-substitutions.
358
7
Piperidine
Analgesics
Two modifications studied in detail have been N-cycloalkyl and Nunsaturated alkyl substitutions (230,231), since morphine agonists are converted to partial antagonists by these changes. N-cyclopropylmethyl produces the best combination of analgesic and antagonist activities, with manipulation of the length and branching of the 3-alkyl group modulating the potencies (maximum analgesia at n-propyl, isopropyl). The most potent racemate, 3-isobutyl-N-cyclopropylmethyl (143), has three times pentazocine's potency as an antagonist and exhibits potent antinociceptive activity against abdominal constriction in mice. The optical enantiomers are active in both morphine agonist and antagonist assays, in contrast to
V
Alkyl Family
HO
147
D
146
D.
HO
R1 = CHZCH(CH3)Z
143
R
CHZ-C-C3HS'
144
R
CHZCH = CHZ' R
145
R
CHZCH = C(CH3)
1
z'
= CHZCH(CH3) R
1
Z
= CHZCH(CH3)Z
profadol, in which the (- )-isomer has greater agonist and the (+ )-isomer greater antagonist activity. Structure-activity relationships in the Nallylpyrrolidines compared to similarly substituted N-cyclopropylmethyl ,analogs show a lower antimorphine/antinociceptive ratio for allyl compounds. While 143 and the 3-isobutyl-N-allyl (144) have about 20 times the potency of pentazocine in a morphine agonism paradigm (abdominal constriction), 144 has over 10 times pentazocine's potency as an antagonist. Substitution in the allyl group reduces the morphine antagonism activity; for example, 3-isobutyl-N-dimethylallyl (145) suffers more than a lO-fold loss of analgesic response and a total loss of antagonism compared to the simple allyl. Interestingly, substitution of the 3-methylene group with N-methyl in several 3-alkyl-3-aryl (meta-hydroxy or meta-methoxy) pyrrolidines to give a set of pyrazolidines results in inactivity, the best of the series, a 4-n-propyl derivative (146), having only half the potency of codeine (232). Ring contraction to an azetidine analog of profadol (147) produces an analgesic potency over twice that of codeine (233).
359
Conformationally
Rigid and Bridged Analogs
Several series of compounds related to the alkyl arylpiperidines have been synthesized for use as narcotic receptor structural probes. From these, both agonists and antagonists have emerged, with some modifications producing mixed agonist-antagonist opiates. Each new analgesic series has its own synthetic scheme and unique set of structure-activity relationships, with several trends paralleling those of the meperidine and prodine families themselves. Although several attempts are being made to relate pharmacological findings to molecular conformations and binding modes, no conclusions have been universally accepted. The goal of the future is, therefore, to characterize narcotic receptor binding modes and explain the structure-activity relationships of the synthetic arylpiperidine analgesics with information gained with these conformationally fixed probes. A summary of the structural types used for optimizing biological activity and for receptor probing is as follows: 1. J-Aryl-3-azabicyclo[3.1.0jhexanes, structure (A), cyclopropyl pyrrolidines (rigid analogs of profadol): Examination of substitution within the aryl ring and of N-substitution has shown that the most potent member of this series has para-methylphenyl and N-H substituents (234,235). Bicifadine (148), as a racemic mixture, with an analgesic (A)
(B)
N I R 148
R
H, X
149 R 150 R =
x = 3' -OCH3 X = 3' -OH
360
7
Piperidine
Analgesics
V
profile in several animal models, has undergone clinical trials for postoperative pain, but with unsatisfactory results (236). Physical methodology and X-ray crystallography have determined that the (+ )-enantiomer, in which all activity resides, has the lR,5S absolute configuration. 2. I-Aryl-3-azabicyclo[3.2.0jheptanes, structure (B), homologs to (A): Examination of ketal intermediates has uncovered two dimethyl ketals (149,150), which have morphine-like analgesic profiles in mice and rats. The phenol (150) has tritiated naloxone-binding displacement ability equivalent to that of morphine (237). A molecular mechanics analysis of the 3-methoxy derivative has attempted to rationalize this opiate-receptor binding (237). 3. Spiro[tetralin-I,3'-pyrrolidinesj, structure (C), conformationally restricted rotamers of profadol: Examination of substitution on the aryl ring and N-substitution gave inactive analgesics, even with hydroxylated phenyls (238). Two unsubstituted phenyl analogs (151,152) display activity equivalent to codeine.
(E)
(F) N-R
-R
R1
HO
X
153 R 154 R
(CHZ)ZC6HS'
X
(CHZ)ZC6HS'
X
3'-OH 4'-OH
155
R
156
R
R1
-
CH
3 CH3, R 1 - H
6. I-Aryl-azabicyclo[3.2.ljoctanes, structure (F), both a bridged profadol and a five-membered analog of the hydroxyphenylmethylmor-
phans: Examination of N-substitution,
ring substitution, and substitution on the aryl ring has given mixed agonist-antagonists, derivatives displaying analgesic activities in the range of codeine to morphine yet also displaying nalorphine-like antagonism (245). In the interesting enantiomeric pairs of 155 and 156, activity resides in the (+ )-isomer. The absolute configuration of (+ )-155 has been determined as lR,5S,7R, which is analogous to the absolute configuration of the more potent hydroxyphenylmorphan (+ )-enantiomer (246,247). This stereoselectivity again, as in the prodine series, suggests that the opiate receptor discriminates enantiotopic edges of the ligand. 7. Oxide-bridged (3-hydroxyphenyl)methylmorphans, (157,158), conformationally restricted phenyl rings: In contrast to the (hydroxyphenyl)methylmorphans with a freely rotating phenyl ring, these restricted analogs have attempted to correlate torsional angles between the phenyl and piperidine rings, as determined by X-ray analysis, with opiate-binding activity (248,249,296). Testing has shown that 157,158 are devoid of agonist and antagonist activity, although 158 has appreciable opiate receptor binding affinity.
x
- Rl _ R21 = R3 _ H
120 R - CH3, Rl = R2 - H, X _ 3' -DH 4. 5-(3-Hydroxyphenyl)-2-methylmorphans, structure (0), bridged 4alkyl-4-arylpiperidine analogs: Examination of the racemate and the enantiomers of the parent (120) has disclosed different opioid profiles: the racemate equipotent with morphine; the (+ )-isomer three times more potent than morphine, with high physical dependence; the (- )-isomer equipotent with morphine but with weak morphine antagonist properties in morphine-dependent monkeys (239). Substitution on the nitrogen, on the phenyl oxygen, and on the piperidine ring (240-242) has indicated that while agonist potency can be increased two to three times in the (+ )-isomers, antagonistic potency in the (- )-isomers is not affected, even with cyclopropylmethyl or allyl groups. 5. 5-Aryl-2-azabicyclo[3.2.ljoctanes, structure (E), conformationally rigid analogs of hydroxyphenylmethylmorphan (0): Examination of phenyl-hydroxy, -alkoxy, -acetoxy, and N-alkyl. -cycloalkyl, -aralkyl groups has indicated that an arylethyl side chain and 3R
361
hydroxyphenyl or 4-hydroxphenyl produce the best analgesic activity (243,244). 3'-Hydroxy N-phenylethyl derivative (153) has an opiate agonist-antagonist profile (equipotent with morphine, half as potent as nalorphine), while 4'-hydroxy (154) is twice as potent as morphine and has weak binding affinity to labeled opiate receptors.
(D)
151
Alkyl Family
152 R = CH3, R - R2 = R3 - H
157
l
158
362
VI.
7
Piperidine
VI
Analgesics
363
Anilino Family
Anilino Family
The 4-anilinopiperidine family has yielded the most potent and distinct class of synthetic analgesics (6,144), related to both the arylpiperidines and open chain anilides (Chapter 9), as evidenced by their structureactivity relationships. The data base of compounds produced since the early 1960s is enormous, a substantial subset of which is clinically effective. The general formula showing structural modifications in this family is given in Fig. 7-2, where each positional modification has been evaluated alone or in combination with others. A.
a
159
Synthesis
b
Literature procedures (250-257) for the synthesis of I-substituted 4-(N-arylalkanoamido)piperidines and relatives (Scheme 7-8) involve reduction of a Schiff base (159), obtained by condensation of an aniline derivative with a l-substituted-4-piperidone, followed by acylation of the 4-anilino intermediate (160). N-substituted derivatives are alternatively obtained through a N-I-carbethoxy intermediate or the l-H piperidine, easily formed by catalytic hydrogenation of the l-benzyl-4-anilide. When these sequences have been unsuccessful, for example, when the piperidine ring is disubstituted or the aryl ring is peculiarly substituted, particular modifications have been described. Various routes to the prototypic compound, fentanyl (161), have also been reported (258-260). OH,
OR,
alkyl
~
1
X
NH4 (
Ac)' I 0010>;"-'
C H
R1 = (CHZ) ZC6HS' Z H, R3 = CZHS X= R = Scheme 7-8. Synthesis of 4-anilino and 4-anilidopiperidines. ZnCI,; (b) NaSH,; (e) (R'CO),O or R'COCI.
2n ~
tI
aralkyl,
heteroalkyl
B.
H,
Fig. 7-J.
alkyl,
Structural modification of 4-anilinopiperidine
Structure-Activity Clinical Utility
Reagents: (a) p-TSA or
Relationships: Generation of Compounds with
Interest in the family prototype (161, fentanyl), l-phenylethyl-4(N-phenylpropionamido )piperidine, has stemmed from its extremely potent analgesic activity relative to other piperidine analgesics (Table 7-14). Fentanyl has almost 500 times the potency of meperidine (in mice, subcutaneously) and 220 times that of morphine (in mice, intraperitoneally) in the hot plate assay, and 500 times that of morphine in the rat tail withdrawal protocol (intravenously) (261-263). The structure-activity relationships of fentanyl analogs and 3-methyl derivatives demonstrate the
1,2,3
N 11 R
H, alkyl
160
161
H, alkyl, or oxygen functionality: COZR, CHZOR, COR, DeOR
/n
A"""';/e
O~
)
x
c
A.y J
H, COZR, COR5 where R5 = R, NR'R"
halogen,
"'" BH~
analgesics.
l
364
7
Piperidine
Analgesics
VI
Anilino Family
365
Table 7-14 Potencies
of Piperidine
Piperidine
Analgesics
Analog
Meperidine (4) Bemidone (43) Alphaprodine (92) Betaprodine (93) Ketobemidone (47) Fentanyl (161) a
b
Relative
potency",b
1 1.5 7 35 10 470
Mice. subcutaneously. References
34, 176,261.
following: 1. The significance of N-phenylethyl and N-propionyl groups in oplimizing analgesic activity (261): alkyl, other aralkyl, and alkyl amino groups on N-l are weaker or inactive; ethoxycarbonyl or hydrogen on the aniline nitrogen diminish activity. 2. The necessity of an unsubstituted phenyl ring to maximize receptor binding (264,265): phenolic hydroxyl and methoxyl derivatives (equivalents of the A-ring of morphine opiates) and rigid analogs are inferior, as measured by stereospecific binding with tritiated fentanyl in rat brain homogenates. 3. The relationship of stereochemistry and configuration to potency (261,262,266-268): only small alkyl groups (i.e., methyl), exclusively at C-3, cis to the 4-anilido substituent and as the (+ )-enantlOmer, enhance activity [cis-3-methyl racemate 6 times, as the (-)enantiomer 0.2 times, but as the (+ )-enantiomer 19 times the potency of fentanyl, with (+ )-cis up to 6684 times the potency of morphine (rat tail withdrawal assay, intravenously)]. 4. The geometry requirements in terms of anilino nitrogen to piperidine nitrogen distance (267): isofentanyl (162) suffers a 300-fold loss in analgesic activity.
162
5. The good correlation (a) between the in vitro affinity for the opiate receptor and the in vivo analgesic potency and (b) between duration of fixation to the receptor sites and duration of action (269). Fenta~Yl (Sublimaze, Leptanal) has been extensively studied since its introduction into clinical practice in the 1960s (270), and its pharmacokinetic profile has been determined using radioimmunoassay techniques on surgical patients (271,272). Due to a potency 50-100 times that of morphine, a '~apid onset of action, and a short duration, fentanyl's clinical utility in surgical anesthesia is widespread, especially in combination with the major tranquilizer droperidol (Innovan, Droleptan, Thalamontal) used in a procedure called neuroleptanalgesia. A 0.2-mg dose of fentanyl is equianalgesic with 10 mg morphine, given intramuscularly. Unfortunately, fentanyl causes the expected morphine-like side effects. Studies with fJ.and S opiate receptors have shown that fentanylisothiocyanate (FIT) is a highly selective alkylator of S receptors in brain membranes (273), and that 3-methylfentanylisothiocyanate (super-FIT), an even more potent S opiate receptor selective affinity ligand, can be used to purify the S receptor subunit (297). (+ )-Cis-3-methylfentanyl (163), whose analgesic action is also more rapid and has a shorter duration than morphine, is of interest clinically due to its lower toxicity (274).
XCOC2HS)-@ CH3
l.)N I
(CH2) 2C6HS 163
Exploration (275-277) of oxygenated substituents replacing hydrogen at C-4, and their C-3-,methyl counterparts, has led to the discovery of novel clinical agents of the fentanyl type with additional beneficial analgesic properties, such as a shorter duration of action and a higher margin of safety. A small polar group at C-4 of the anilidopiperidines, that is, C-4-carboalkoxy, as in the meperidine family, C-4-alkoxymethyl or C-4-oxoalkyl, as in ketobemidone, significantly enhances the analgesic potency relative to fentanyl. Structure-activity profiles found in fentanyl itself also apply to these derivatives, for example, potency-enhancing substituents on the piperidine nitrogen (Table 7-15). Sufentanil (170) (278-280) and alfentanil (174) (281-283) are the most promising candidates for accepted clinical use in the United States, although other analogs, such as 169 (284) and carfentanil (164) (285), are being studied in the clinic (286).
7
366
References
Piperidine Analgesics
Table 7-15 Analgesic
Activities
of 4-Anilidopiperidines
x
o R'
R'
X
C02C11) C02C113 C02CI13 C02CH3 C02CH3 CH,OCH, CH,OCH, COCH, COCH, COC,H, H
C,H, CzHs c-C)Hst C,H, CzHs C,H, C,H, C,H, c-C3Hs C,H, C,H,
H H H H 4-F H H H H H H
R - 164 165 166 ~167 168 ~169 -170 171 172 173 161
(I
C,H,(CH2){ C,H,S(CH,)," C,H,(CH,Jz C,H,CH,CH(CH,) C,H,(CH,Jz C,H,(CH,Jz" C4H)S(CHzhC
C,H,(CH,Jz C,H,(CH,Jz C,H,S(CH,Jz C.H,(CH,h' Morphine Meperidine
Potency Ratioa,b
7682 7159 6176
k"'" ~~~-:;1 r~ ';h(
4038 3987 4921 3795 4436 292 1 0.53
criteria that determine binding to Ihe opiate receplor (287,288)_ These have focused on the structure-activity relationships of rigid fentanyl-type analogs and determination of the absolute configuration of the more potent resolved 3ialkyl enantiomers. Ring expansion to perhydroazepines or ring contractiop to pyrrolidines has shown that ring size does not significantly alter the level of analgesic potency found for a beneficial combination of substituents in the 4-(propananilido )piperidines (289). Whereas the observed analgesic activity does not differ significantly in ring homologs, it is severely \diminished or abolished in analogs with conformational restraints (29q-294). Isomeric N-substituted 3-(propanilido)nortropanes, stereochemi~ally a set of semirigid analogs, have shown that 3{3 isomers are more potent than 3" (295). The potent (+ )-cis 3-methylfentanyl and its isothiocya,nate (super-FIT), both related to the more active (+ )-cis{3-prodine antipode (35,45), have been assigned an absolute configuration of 3R,45 by recent X-ray analyses (298)_
-f"ih.)
Reference 275.
b Tail withdrawal, "d Clinical utility. 2-Thienylethyl. Cyclopropyl. f" Fentanyl.
"r..i )
367
rats, intravenously.
References 1. O. Eisleb and O. Schaumann, Dtsch. Med. Wochschr. 65, 967 (1939). 2. A. Goldstein, L. Aronow, and S. M. Kalman, "Principles of Drug Action: The Basis of Pharmacology," p. 812. Wiley, New York, 1974. crJ F. Bergel and A. L. Morrison, Q. Rev. (London) 2, 349 (1948). /4. 0.3. Braenden. N. B. Eddy, and H. Halbach, Bull. WHO 13, 937 (1955). 5. R. A. Hardy, Jr. and M. G. Howell, in "Analgetics" (G. de Stevens, ed.), pp. 179-222. Academic Press, New York, 1965. 6. A. F. Ca,y. Progr. Drug Res. 22, 168-227 (1978). 7. A. F. Casy, in "Narcotic Drugs Biochemical Pharmacology" (D. H. Clouet, ed.), pp. 1-16. Plenum, New York, 1971. \ O. 3. Braenden and P. O. Wolff. Bull WHO 10, 1003 (1954). 'i 9. N. B. Eddy, H. Halbach, and O. 3. Braenden, Bull. WHO 14, 353 (t956). lOa. L. B. Mellett and L. A. Woods, Progr. Drug Res. 5, 190-203 (1963). lOb. G. A. Deneau and M. H. Seevers, Bull. Drug Addict. Narc., No. 23, Addendum 1, Natl.Acad. Sci.,Natl. Res. Council, Div. Med. Sci.,Washington, D. C. (1961);ibid.,No. 24, Addendum 2. (1962); ibid., No. 25. Addendum 3 (1963). <:; i11 O. Eisleb, Chern. Ber. 74, 1433 (1941). P~S,-~
~
f~!r< f'70+k.
174
C.
Rigid Analogs and Conformational
Exploration
Attempts have been made to discover the relationships between the anilinopiperidines and the arylpiperidines or other analgesic agonistsantagonists with respect to conformational (structural) and stereochemical
LaZ> P. A. 13. 14. 15. 16. 17.
J. Janssen, U. S. Patent 3,004,977 (October 17, 1961); Chern. Abstr. 56, 12861 (1962); R. Kwok and P. Prane. U. S. Patent 3,627,771 (December 14, 1971); Chern. Abstr. 76,99527 (1972). A. H. Beckett and A. F. Casy, Bull. Narc. 9, 37 (1957). A. H. Beckett and A. F. Casy, in "Progress in Medicinal Chemistry" (G. P. Ellis and G. B. West, eds.), Vol. 2, pp. 61-71. Butterworths, London, 1962. A. D. MacDonald, G. Woolfe, F. Bergel, A. L. Morrison, and H. Rinderknecht, Br. J. Pharmacal. 4, 4 (1946). G. Lambrecht and E. Mutschler, Arch. Pharm. (Weinheim, Ger.) 308, 676 (1975). R. L. Jacoby, K. A. Nieforth, and R. E. Willette, J. Med. Chern. 17,453 (1974).
J68
7
Piperidine Analgesics
18. A. N. Voldeng, C. A. Bradley, R. D. Kee, E. L. King, and F. L. Melder, J. Pharm. Sci. 57, 1053 (1968). 19. Drugs Fulure 6, 529 (1981). 20. Merck & Co., Br. Patent 815,808 (July 1, 1959); Chern. Abst,. 54, 1559 (1960). 21. J. Buchi, M. Prost, H. Eichenberger, and R. Lieberherr, Helv. Chim. Acta 35, 1527 (1952). 22. S. Archer, U. S. Patent 2,777,850 (January 15, 1957); Chern. Abst,. 51, 12986 (1957). 23. C. M. Hoffman and M. J. Weiss, J. Org. Chern. 25, 653 (1960). 24. A. F. Casy, J. Med. Chern. 11, 188 (1968). 25. M. R. Bell and S. Archer, J. Arn. Chern. Soc. 82, 4638 (1960). 26. S. J. Daum, C. M. Martini, R. K. Kullnig, and R. L. Clarke, J. Med. Chern. 18,496 (1975). 27. A. F. Casy and J. E. Coates, Org. Magn. Reson. 6, 441 (1974). 28. P. S. Portoghese, J. Pharm Set. 55, 865 (1966). 29. p, S. Portoghese, A. A. Mikhail, and H. J. Kupferberg,J. Med. Chern. 11,219 (1968). 30. R. L. Clarke, A. J. Gambino, A. K. Pierson, and S. J. Daum,J. Med. Chern. 21, 1235 (1978). ill> "Problems of Drug Dependence, 1981" (L. S. Harris, ed.), NIDA Research Monograph 41, Comm. on Prob. Drug Dep., Natl, Inst. Drug Abuse, Rockville, Maryland, 1982. 32. N. B. Eddy and D. Leimbach, J. Pharmacol. Exp. Ther. 107,385 (1953). 33. L. Lasagna and H. K. Beecher, J. Pharmacol. Exp. Ther. 112, 306 (1954). 34. N. B. Eddy, H. Halbach, and O. J. Braenden, Bull WHO 17, 569 (1957). 35. D. J. Edwards, C. K. Svensson, J. P. Visco, and D. Lalka, CIin. Pharmacokinel. 7,421 (1982). 36. C. E. Inturrisi and J. G. Umans, CIin. Anaeslhesiol. I, 123 (1983). 37. B. E. Dahlstrom, L. K. Paalzow, C. Lindberg, and C. Bogentoft, Drug Metab. Dispos. 7, 108 (1979). 38. C. B. Pert and S. H. Snyder, Proc. Nail. Acad. Sci. U.S.A. 70, 2243 (1973). 39. C. B. Pert and S. H. Snyder, Mol. Pharmacol. 10, 868 (1974). 40. M. M. Abdel-Monem and P. S. Portoghese, J. Pharrn. Pharmacol. 23, 875 (1971). 41. S. T. Christian, C. W. Gorodetzsky, and D. V. Lewis, Biochern. Pharrnacol. 20, 1167 (1971). 42. D. L. Larson and P. S. Portoghese, J. Med. Chern. 19, 16 (1976). 43. R. H. Thorp and E. Walton, J. Chern. Soc. p. 559 (1948). 44. T. D. Perrine and N. B. Eddy, J. Org. Chern. 21, 125 (1956). 45. L. Grumbach, Fed. Proc., Fed. Arn. Soc. Exp. Bioi. 18, 61 (1958). 46. K. Pfister, III and A. A. Pessolano, U. S. Patent 2,294,753 (August 2, 1960); Chern. Abs". 55, 12427 (1961). 47. B. Elpern, U. S. Patent 3,109,004 (October 29, 1963); Chern. Absl'. 60,6829 (1964). 48. B. Elpern, L. N. Gardner, and L. Grumbach, J. Arn. Chern. Soc. 79, 1951 (1957). 49. P. A. Janssen and N. B. Eddy, J. Med. Pharm. Chern. 2, 31 (1960). 50. N. V. Smirnova, A. P. Arendaruk, D. D. Smolin, and A. P. Skoldinov, Med. Promst. SSSR 12, 31 (1958). 51. J. Weijlard, P. D. Orahovats, A. P. Sullivan, Jr., G. Purdue, F. K. Heath, and K. Pfister. J. Arn. Chern. Soc. 78,2342 (1956). 52. Merck & Co., Br. Patent 815,926 (July 1, 1959); Chern. Abs". 54, 1559 (1960). 53. P. D. Orahorvats, E. G. Lehman, and E. W. Chapin, J. Pharrnacol. 119,26 (1957). 54. A. S. Keats, J. Telford, and Y. Kurosu, Aneslhesiology 18, 690 (1957). 55. N. B. Eddy, L. E. Lee, and L. S. Harris, Bull. Na". 11,3 (1959).
References
369
56. P. S. Portoghese,
V. G. Telang, A. E. Takemori,
144 (1971).
57. A. E. Takemori, 1051 (1974).
A. Ward, P. S. Portoghese,
\ and G. Hayashi, \
J. Med. Chern. 14,
and V. G. Telang, J. Med. Chern. 17,
I
58. B. Elpern, P. Carabateas, A. E. Soria, L. N. Gardner, and L. Grumbach, J. Arn. Chern. Soc. 81, 3784 (1959). 59. L. A. W~ds, G. A. Deneau, D. R. Bennett, E. F. Domino, and M. H. Seevers, Toxico/. Appl. Pharrnacol. 3,358 (1961). 60. T. J. De Kornfeld and L. Lasagna, J. Chronic Dis. 12, 252 (1960). 61. F. A. Cutler, Jr. and J. F. Fisher, U. S. Patent 2,962,501 (November 29,19(0); Chern. Abs". 55, 10475 (1961). 62. B. Elpern, P. M. Carabateas, A. E. Soria, and L. Grumbach, J. Org. Chern. 24, 2045 (1960).
,
63. 64. 65. 66. 67. 68. 69.
I
B. Elperp, U. S. Patent 2,901,487 (August 25, 1959); Chern. AbSlr. 54, 1555 (1960). C. B. Pert, S. H. Snyder, and P. S. Portoghese, J. Med. Chern. 19, 1248 (1976). A. F. Green and N. B. Ward, Br. J. Pharmacol. 11,32 (1956). R. J. A~derson, P. M. Frearson, and E. S. Stern, J. Chern. Soc. p.4088 (1956). R. A. Millar and R. P. Stephenson, Br. J. Pharmacol. 11, 27 (1956). P. M. Frearson and E. S. Stern, J. Chern. Soc. p. 3062 (1958). P. M. Frearson and E. S. Stern, Br. Patent 797,448(July 2,1958); Chern.Abstr. 54,583 (1960). . 70. A. M. J. N. Blair and R. P. Stephenson, Br. J. Pharmacol. IS, 247 (1960). , .71. R. E. Lister Br. J. Pharmacol. IS, 254 (1960). 72. E. S. Stern and R. L. Watt, U. S. Patent 2,960,507(November 15, 1960);Chern.Abslr. 55, 18779 (1961).
73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88.
P. M. Frearson, D. G. Hardy, and E. S. Stern, J. Chern. Soc. p. 2103 (1960). D. G. Hardy, R. E. Lister, and E. S. Stern, J. Med. Chern. 8, 847 (1965). H. G. Morren and H. Strubbe, Ind. Chirn. BeIge, 22, 795 (1957). E. Merlevede and S. Levis, Arch. Inl. Pharrnacodyn. lIS, 213 (1958). H. G. Morren, Belg. Patent 558,883 (July 31, 1957); Chern Abs". 54, 1555 (1960). P. A. J. Janssen, A. H. M. Jageneau, P. J. A. Demoen, C. van de Westeringh, A. H. M. Raeymaekers, M. S. J. Wouters, S. Sanczuk, B. K. F. Hermans, and J. L. M. Loomans, J. Med. Pharrn Chern. I, 105 (1959). P. A. J. Janssen, A. H. M. Jageneau, P. J. A. Demoen, C. van de Westeringh, J. H. M. de Canniere, A. H. M. Raeymaekers, M. S. J. Wouters, S. Sanczuk, and B. K. F. Hermans, J. Med. Pharrn. Chern. 1, 309 (1959). P. A. J. Janssen, A. H. M. Jageneau, P. J. A. Demoen, C. van de Westeringh, J. H. M. de Canniere, A. H. M. Raeymaekers, M. S. J. Wouters, S. Sanczuk, and B. K. F. Hermans, J. Med. Pharrn. Chern. 2, 271 (1960). A. Pohland, U. S. Patent 2,951,080 (August 30, 1960); Chern. Absl', 55,4540 (1961). R. H. Mazur, J. O'g. Chern. 26, 962 (1961). E. Nilsson and P. Janssen, Acla Anaslhesiol. Scand. 5, 73 (1961). J. O. Jilek, J. Pomykacek, and M. Protiva, Collecl. Czech. Chern. Cornrnun.32, 1285 (1967). P. J. Costa and D. D. Bonnycastle, J. Pharmacol. Exp. Ther. 113, 310 (1955). C. A. Winter, P. D. Orahovats, and E. G. Lehman, Arch. Int. Pharrnacodyn. 110, 186 (1957). S. Archer and L. S. Harris, Prog. Drug Res. 8, 261 (1965). K. Zeile, H. Merz, and K. Freter, Ger. Offen. 1,288,090 (January 30, 1969); Chern. Absl'. 70, 9642 (1969).
370
7
Piperidine Analgesics
89. Research N. V., Be. Patent 941, 748 (November 13, 1963). 90. A. F. Casy, L. G. Cbat!en, and K. K. Khullar, J. Chern. Soc. C p. 2491 (1969). 91. H. W. Kosterlitz, J. A. H. Lord, and A. J. Watt, in "Agonist and Antagonist Actions of Narcotic Analgesic Drugs" H. W. Kosterlitz, H. O. J. Collier, and J. E. Villarreal cds.), pp. 45-61. Univ. Park Press, Baltimore, Maryland 1973. ' 92. P. A. J. Janssen, U. S. Patent 3,012,030 (December 5,1961); Chern. Abstr. 56, 12861 (1962). 93. P. A. 1. Janssen, U. S. Patent 3,004,977 (October 17, 1961); Chern. Abstr. 56, 12861 (1962). 94. F. Bergel, N. C. Hindley, A. L. Morrison, and H. J. Rinderknecht, J. Chern. Soc. p. 269 (1944). 95. G. F. Woods, T. L. Heying, L. H. Schwartzman, L. H. Grenell, S. M. Gasser, W. F. Rowe, and N. C. Bolgiano, J. Org. Chern. 19, 1290 (1954). 96. F. F. Blicke and E.-P. Tsao, J. Am. Chern. Soc. 75, 3999 (1953). 97. J. Diamond and W. F. Bruce, U. S. Patent 2,666,050 (January 12, 1954); Chem. Abstr. 49, 4031 (1955). 98. J. Diamond, W. F. Bruce, and F. T. Tyson, J. Org. Chern. 22, 399 (1957). 99. J. Diamond, M. Dymicky, and W. F. Bruce, Br. Patent 843, 924 (August 10,1960); Chern. Abstr. 55,2705 (1961). 100. J. Diamond, W. F. Bruce, C. Gochman, and F. T. Tyson, J. Org. Chern. 25,65 (1960). 101. J. Diamond and W. F. Bruce, U. S. Patent 2,740,777 (April 3, 1956); Chern. Abstr. 50, 15598 (1956). 102. J. Diamond and W. F. Bruce, Br. Patent 781,002 (August 14, 1957); Chern. Abstr. 52, 9230 (1958). 103. J. Seifter, D. K. Eckfeld, I. A. Letchack, E. M. Gore, and J. M. Glassman, Fed. Proc., Fed. Am. Soc. Exp. Bioi. 13, 403 (1954). 104. J. M. Glassman and J. Seifter, J. Pharrn. Exp. Ther. 115, 21 (1955). 105. H. F. Fraser, Fed. Proc., Fed. Am. Soc. Exp. BioI. IS, 423 (1956). 106. A. J. Grossman. M Golbey, W. C. Gittinger, and R. C. Batterman, J. Am. Geriatr. Soc. 4, 187 (1956). 107. L. J. Cass and W. S. Frederik, Curro Ther. Res. 3, 97 (1961). 108. M. Golbey, W. C. Gittinger, and R. C. Batterman, Fed. Proc., Fed. Am. Soc. Exp. Bioi. 14, 344 (1955). 109. L. J. Casso W. S. Frederik, and A. F. Bartholomay, JAMA 166, 1829 (1958). 110. J. Malis, M. A. Ries, and R. J. Bradley, Fed. Proc., Fed. Arn. Soc. Exp. Bioi. 22, (2, Part 1), 248 (1963). 111. E. L. May, in "Agonist and Antagonist Actions of Narcotic Analgesic Drugs" (K. W. Kosterlitz, H. O. Collier, and J. E. Villarreal, eds.), pp. 17-22. Univ. Park Press, Baltimore, Maryland, 1973. .~ rt "I ~'""""112. J. P. Tollenaere, Trends Pharrnacol. Sci. 4, 367 (1983). ..\oj 113. M. Froimowitz, J. Med. Chern. 25, 1127 (1982). '\i M. R. Bell and S. Archer, J. Am. Chern. Soc. 82, 151 (1960). 115. M. M. Abdel-Monem and P. S. Portoghese, J. Med. Chem. IS, 206 (1972). . A. Langbein, H. Merz, L. Stockhaus, and H. Wick, Adv. Biochem. Psychopharmacol. 8, 157-165 (1974). >I"~'~'
·
~
+f
;
7. A. W. D. Avison and A. L. Morrison, J. Chern. Soc. p. 1469 (1950). 118. B. Elpern, P. M. Carabateas, and L. Grumbach, J. Org. Chern. 26, 4728 (1961). '>s 19 B. Elpern, U. S. Patent 3,096,335 (July 2, 1963); Chern. Abstr. 59, 13955 (1963). T. Oh-ishi and E. L. May, J. Med. Chern. 16, 1376 (1973). 121. R. S. Wilson, M. E. Rogers, C. B. Pert, and S. H. Snyder, J. Med. Chern. 18, 240 (1975). .
References 122. H. W. Kosterlitz, (1975).
371 F. M. Leslie, and A. A. Waterfield, J. Pharm. Pharmacol. 27.73
\
123. I. M. Uwaydah, M. K. Waddle, and M. E. Rogers; J. Med. Chern. 22, 889 (1979). 124. E. S. Stern, Br. Patent 841,120 (July 13, 1960); Chern. Abstr. 55, 3620 (1961). 125. C. H. Boehringer Soho, Belg. Patent 614,688 (September 5, 1962); Chern. Abstr. 58, 11334 (1963).
'r<'.. 126. K. Th0'1'ae, Br. Patent 882,891 (November 22,1961); Chern. Abslr. 56, 11575(1962);
~r f
~
K. Thomae, Fr. Patent M831 (October 23, 1961); Chern. Abstr. 58, 9032 (1963). W. M. E. Rogers, D. S. Wilkinson, J. R. Thweatt, and S. P. Halenda, J. Med. Chern. 23. 688 (1980).
M. A. Ior!o, F. Gatta, and E. Menichini, Farmaco. Ed. Sci. 32, 212 (1977). 129. K. A. Jensen, F. Lindquist, E. Rekling, and C. G. WOltbrandt, Dan. Tidsk. Farm. 17. 173 (1943): 130. A. Ziering,\ L. Berger, $. D. Heineman, and J. Lee, J. Org. Chem. 12.894 (1947). ffi. A. F. Casy; Med. Res. Rev. 2, 167 (1982). 132. A. Ziering and J. Lee, J. Org. Chern. 12, 911 (1947). 133. S. M. McElvain and M. D. Barnett, J. Arn. Chern. Soc. 78. 3140 (1956). 134. A. H. Beckett, A. F. Casy, and G. Kirk, J. Med. Pharrn. Chern. I, 37 (1959). 135. C. J. Schmidle and R. C. Mansfield, J. Arn. Chem. Soc. 78. 1702 (1956). 136. A. Ziering, A. Motchane, and J. Lee, J. Org. Chern. 22, 1521 (1957). 137. W. M. Benson, D. J. Cunningham, D. L. Hane, and S. van Winkle, Arch. Int. Pharmacodyn. 109. 171 (1957). 138. S. M. McElvain and K. Rorig, J. Am. Chern. Soc. 70, 1820 (1948). 139. D. R. Howton, J. Drg. Chern. 10, 277 (1945). 140. A. F. Casy and K. M. J. McErlane, J. Chern. Soc., Perkin Trans. 1 p.726 (1972). 141. K. H. Bell and P. S. Portoghese, J. Med. Chern. 16, 203 (1973). 142. M. M. A. Hassan and A. F. Casy, Tetrahedron 26, 4517 (1970). 143. A. H. Beckett and A. F. Casy, in "Progress in Medicinal Chemistry" (G. P. Ellis and G. B. West, eds.), Vol. 4, pp. 209-218. Butterworths, London, 1965. 144. A. H. Beckett and A. F. Casy, in "Progress in Medicinal Chemistry" (G. P. Ellis and G. B. West, eds.), Vol. 7, pp. 229-284. Butterworths, London, 1970. 145. A. H. Beckett, A. F. Casy, and P. M. Phillips, J. Med. Pharm. Chern. 2.245 (1960). 146. A. F. Casy, A. H. Beckett, G. H. Hall, and D. K. Vallance, J. Med. Pharm. Chern. 4, 535 (1961). 147. A. H. Beckett and J. V. Greenhill, J. Med. Pharrn. Chern. 4.423 (1961). 148. A. F. Casy and A. H. Beckett, J. Pharrn. Pharmacol. 13. 161-T (1961). 149. A. H. Beckett, U. S. Pateot 3,074,952 (January 22,1963); Chern. Abslr. 59,580 (1963).
150. 151. 152. 153. 154.
N. J. Harper and S. E. Fullerton, J. Med. Pharrn. Chem. 4. 297 (1961). A. H. Beckett, A. F. Casy, G. Kirk, and J. Walker, J. Pharm. (London) 9. 939 (1957). A. F. Casy and N. A. Armstrong, J. Med. Chern. 8.57 (1965). R. Abbas, R. E. Willette, and J. M. Edwards, J. Pharm. Sci. 66, 1583 (1977). R. A. Stokbroekx, J. Vandenberk, A. H. M. T. Van Heertum, G. M. L. W. van Laar, M. J. M. Van der Aa, W. F. M. Van Bever, and P. A. J. Janssen, J. Med. Chern. 16, 782 (1973).
155. M. A. Davis, F. Herr, R. A. Thomas, and M.-P. Charest, J. Med. Chern. 10, 627 (1967). 156. P. M. Carabateas, W. F. Wetterau, and L. Grumbach, J. Med. Chern. 6.355 (1963). 157. O. E. Fancher, S. Hayao, W. G. Strycker, and L. F. Sancilio, J. Med. Chern. 7.721 (1964). . 158. B. S. Huegi, A. M. Ebnother, E. Rissi, F. Gadient, D. Hauser, D. Roemer, R. C. Hill, H. H. Buescher, and T. J. Petcher, J. Med. Chern. 26, 42 (1983).
J7Z
7
References
Piperidine Analgesics
373
199. K. M. J. McErlane and A. F. Casy, J. Chem. Soc., Perkin Traft!. 1 p. 339 (1972).
159. D. Zhou, S. Fang, B. Ge, Q. Lit Z. Li, Q. Dai, R. Wu, Y. Wu, Z. Huang, X. Chen, Yaoxue Xuebao 17, 503 160. B. Elpern, W. Wetterau, (1958). 161. W. J. Welstead. Jr., G. W. H. Funderburk, and
\
200. J. Diamond, W. F. Bruce, and F. T. Tyson,J. Med.\ Chem. 7, 57 (1964).
(1982); Chern. Abstr. 97, 182185 (1982). P. Carabateas. and L. Grumbach, J. Am. Chern. Soc. 80,4916
.
C. Helsley. C. R. Taylor, L. B. Turnbull, J. P. Da Vanzo R. S. Alphin, J. Med. Chern. 16, 1129 (1973). 162. E. Solomons and J. Sam, J. Med. Chern. 16, 1330 (1973). 163. P. M. Carabateas and L. Grumbach, J. Med. Pharm. Chern. 5, 913 (1962). 164. N. Itoh, Chern. Pharrn. Bull. (Tokyo) 10, 55 (1962). 165. K. Thomae, Br. Patent 940,534 (October 30, 1963); Chern. Abstr. 60, 2912 (1964). 166. C. H. Boehringer Soho, Fr. Patent 1914M (August 12, 1963); Chern. Abstr. 60, 5467 (1964). 167. R. H. K. Foster and A. J. Carman, J. Pharmacol. 91, 195 (1947). 168. Wellcome Foundation Ltd., Br. Patent 919,124 (February 20, 1963); Chem Abstr. 59, 2829 (1963). 169. P. A. J. Janssen, Belg. Patent 578,395 (May 30,1959); Chern. Abstr. 54,4623 (1960). 170. A. Pohland, Br. Patent 824,607 (December 2, 1959); Chern. Abstr. 54, 7742 (1960). 171. K. H. Bell and P. S. Portoghese, J. Med. Chem. 17,129 (1974). 172. M. A. Iorio, G. Damia, andA. F. Casy,J. Med. Chem. 16,592 (1973); M. A. Iorio and A. F. Casy, J. Pharm. Pharmacol. 26, 553 (1974). 173. L. O. Randall and G. Lehmann, J. Pharmacol. 93, 314 (1948). 174. M. A. Inrin, A. F. Casy, and E. L. May, Eur. J. Med. Chern. 10, 178 (1975). 175. D. S. Fries, R. P. Dodge, H. Hope, and P. S. Portoghese, J. Med. Chem. 25,9(1982). 176. P. S. Portoghese and D. L. Larson, J. Pharm. Sci. 57,711 (1968). 177. E. H. Bachrach, A. D. Godholm, and A. M. Betcher, Surgery 37, 440 (1955). 178. M. A. Iorio and A. F. Casy, J. Chern. Soc. C p. 135 (1970). 179. M. A. Iorio and W. A. Klee, J. Med. Chern. 20, 309 (1977). 180. A. H. Beckett, Fortschr. Arzneimittforsch. I, 455 (1959). 181. A. F. Casy, A. B. Simmonds, and D. Staniforth, J. Pharm. Pharmacol. 20,768 (1968). 182. A. S. Keats, J. Telford, and Y. Kurosu, J. Pharmacol. 130,218 (1960). 183. A. F. Casy and F. O. Ogungbamila, J. Pharm. Pharmacol. 35, Suppl. 76P (1983). 184. P. S. Portoghese, 8. D. Alreja, and D. L. Larson, J. Med. Chem. 24, 782 (1981). 185. Tanabe Seiyaku Co. Ltd., lpn. Patent 14741/63 (August 13, 1963); Chem. Abstr. 60, 2907 (1964). 186. R. T. Major and F. Dursch, J. Org. Chem. 26, 1867 (1961). 187. K. H. Bell and P. S. Portoghese, J. Med. Chem. 16,589 (1973). 188. D. S. Fries and P. S. Portoghese, J. Med. Chem. 19, 1155 (1976). 189. D. L. Larson and P. S. Portoghese, J. Med. Chem. 16, 195 (1973). 190. M. Froimowitl and P. Kollman, J. Compo Chem. 5,507 (1984). 191. A. F. Casy and K. McErlane, J. Pharm. Pharmacol. 23, 68 (1971). 192. A. F. Casyand F. D. Ogungbamila,l. Chem. Soc., Perkin Trans. I p. 749 (1981). 193. A. F. Casy, J. E. Coates, and C. Rostron, l. Pharm. Pharmacal. 28, 106 (1976). 194. "Principles of Medicinal Chemistry" (W. O. Foye, ed.), 2nd ed., pp. 274-291. Lea and Febiger, Philadelphia, Pennsylvania, 1976. 195. P. S. Portoghese, A. S. D. Gomas, D. L. Larson, and E. Shefter, J. Med. Chem. 16, 199 (1973). 196. A. F. Casy and K. M. McEriane, J. Chem. Soc., Perkin Trans. I p. 334 (1972). 197. D. Fries and P. S. Portoghese, l. Med. Chem. 17,990(1974). 198. C. R. Clark, J. M. Holbrook, and R. F. Borne, J. Med. Chern. 19, 852 (1976).
201. l. Diamond and W. F. Bruce, U. S. Patent 2,775,589 (December 25, 1956); Chem. Abslr. 51, 441 (1957). 202. J. Diamond and W. F. Bruce, U. S. Patent 2,740,780 (April 3, 1956); Chem. Abstr. 50 15600 (1956). ' 203. J. F. Cavall~, J. Davoli, M. l. Dean, e. S. Franklin, and D. M. Temple, J. Med. Pharm. Chem. 4, 1 (1961). 204. Y.-H. Wu, W. A. Gould, W. G. Lobeck, Jr., H. R. Roth, and R. F. Feldkamp, J. Med. Pharrn! Chern. 5, 752 (1962). 205. Y..H. Wu, w. G. Lobeck, Jr., and R. F. Feldkamp, J. Med. Pharm. Chem. 5 762 (1962). \ ' 206. M. A. Iorio and M. Miraglia, Chim. Ther. 1,85 (1973). 207. e. V. Winder, J. Wax, B. Serrano, L. Scotti, S. P. Stackhouse, and R. H. Wheelock,J. Pha,rnacol. Exp.\Ther. 133,117 (1961). 208. J. F. Cavalla, R. ~.Selway,J. Wax, L. Scotti,ande. V. Winder,l. Med. Pharm. Chem. 5, 441 (1962).
209. R. C. Batterman,
G. H. Mouratoff, andJ. E. Kaufman, Am. J. Med. Sci. 247,62 (1964); (1961). 210. D. M. Zimmerman, U. S. Patent 4,081,450 (March 28,1978); Chem. Abstr. 89, 109113 (1978). 211. D. M. Zimmerman, R. Nickander, J. S. Horng, and D. T. Wong, Nature (London) 275,
L. J. CassandW. S. Frederik,Curro Ther. Res. 3,97
332 (1978).
212. l. D. Leander, Pharmacol. Biochem. Behav. 15,65 (1981). 213. D. M. Zimmerman,S. E.Smits,M. D. Haynes,B. E. Cantrell,andR.Nickander,Abstr. Pap. Arn. Chern. Sac, /84 Merr., MEDl, p. 23 (1982). 214. D. M. Zimmerman, S. E. Smits, M. D. Haynes, B. E. Cantrell, M. Reamer, and R. Nickander, NIDA Res. Monogr. Ser. 41, 112 (1982). 215. Drugs Future 9, 204 (1984).
216. M. D. Hynes, S. E. Smits, B. E. Cantrell, R. Nickander, and D. M. Zimmerman, NIDA Res. Monogr. Ser 41, 119 (1982). 217. 218. 219. 220.
S. N. Urimoto and H. Hayashi, lpn. J. Pharmacol. 23, 742 (1973). H. Kugita and T. Oine, Chem. Pharm. Bull. 11,253 (1963). R. Achini, Helv. Chim. Acta 64, 2203 (1981). H. Kugita, H. Inoue, T. Dine, G. Hayashi, and S. Nurimoto, l. Med. Chern. 7,298 (1964). 221. H. Kugita, T. Dine, H. Inoue, and G: Hayashi, l. Med. Chern. 8,313 (1965). 222. F. R. Ahmed,Acta Crystallogr., Sect. B 40,1021(1984). 223. M. A. Iorio, A. F. Casy,E. L. May,and W. A. Klee, Congr. Hung. Pharmacol. Soc. hoc. 2, 119(1976). 224. Drugs Future I, 68 (1976); 2,146 (1977); 3,160(1978); 4,154 (1979); 5,103 (1980); 6,112 (1981); 7,129,908 (1982). 225. A. Hedges, P. Turner, and l. Wadsworth, Br. J. Anesthesiol. 52,295 (1980). 226. A. G. Wade and P. J. Ward, l. Int. Med. Res. 10, 104 (1982); R. K. J. Price and A. N. Latham, ibid. 10,219 (1982); e. E. Parker and A. F. Langrick, ibid. 10,408 (1982). 227. C. V. Winder,M. Welford,l.Max,andD. W. Kaump,l. Pharmacol. Exp. Ther.154, 161 (1966). 228. I. M. Lockhart,N. E. Webb,M. Wright,e. V. Winder,andP. Varner,l. Med. Chem.15, 930 (1972).
374
7
Piperidine Analgesics
229. I. M. Lockhart,N. E. Webb,M. Wright, C. V. Winder, and M. A. Hare,J. Med. Chern. 15,687 (1972). 230. R. E. Bowman, H. O. J. Collier, P. J. Hattersley, I. M. Lockhart, D. J. Peters C. Schneider, N. E. Webb, and M. Wright, J. Med. Chern. 16,1177 (1973). ' 231. R. E. Bowman, H. O. J. Collier, I. M. Lockhart, C. Schneider, N. E. Webb, and M. Wright,J. Med. Chern. 16, 1181 (1973). 232. M. J. Kamet and H. S. t. Tan,i. Pharm. Sci. 61, 188, 1936 (1972). 233. D. C. Bishop,J. F. Cavalla, I. M. Lockhart, M. Wright, C. V. Winder, A. Wong, andM. Stephens.J. Med. Chern. 11,1077 (1968). 234. J. W. Epstein, H. J. Brabander, W. J. Fanshawe. C. M. Hofmann, T. C. McKenzie, S. R. Safir, A. C. Osterberg, D. B. Cosulich, and F. M. Lovell, J. Med. Chern. 24,481 (1981). 235. A. U. Epstein, A. C. Osterberg, and B. A. Regan, NIDA Res. Mongr. Ser. 41, 93 (1982). 236. E. J. B. Porter, M. Rolfe, H. J. McQuay, R. A. Moore, and R. E. S. Bullingham, Curro Ther. Res. 30, 156 (1981). 237. T. C. McKenzie, J. W. Epstein, W. J. Fanshawe, J. S. Dixon, A. C. Osterberg, L. P. Wennogle, B. A. Regan, M. S. Abel, and L. R. Meyerson, J. Med. Chern. 27, 628 (1984). 238. P. A. Crooks and R. Szyndler, J. Med. Chern. 23, 679 (1980). 239. H. Awaya, E. L. May, M. D. Aceto, H. Merz, M. E. Rogers, and L. S. Harris,J. Med. Chern. 27, 536 (1984). 240. D. M. Zimmerman, U. S. Patent 4,278,797 (July 14, 1981); Chem. Abstr. 94,83962 (1981). 241. H. H. Ong, T. Oh-ishi, and E. L. May, J. Med. Chern. 17, 133 (1974). 242. M. D. Aceto, L. S. Harris, and E. L. May, NIDA Res. Monogr. Ser. 41, 338 (1982). 243. H. H. Ong, V. B. Anderson, and J. C. Wilker, J. Med. Chern. 21, 758 (1978). 244. H. H. Ong, V. B. Anderson, J. C. Wilker, T. C. Spaulding, and L. R. Meyerson, J. Med. Chern. 23, 726 (1980). 245. M. Takeda, H. Inoue, K. Noguchi, Y. Honma, M. Kawamori, G. Tsukamoto, Y. Yamawaki, S. Saito, K. Aoe, T. Date, S. Nurimoto, and G. Hayashi, J. Med. Chern. 20, 221 (1977). 246. T. G. Cochran, J. Med. Chern. 17, 987 (1974). 247. G. Hite, P. Salve, J. B. Anderson, M. Rapposch, M. Mangion, and M. Froimowitz, Acta Crystallogr., Sect. B. 40, 850 (1984). 248. T. R. Burke, Jr., A. E. Jacobson, D. C. Rice, and J. V. Silverton, J. Org. Chern. 49. 1051 (1984).
249. T. R. Burke, Jr., A. E. Jacobson, K. C. Rive, and J. V. Silverton, J. Org. Chern. 49, 2508 (1984). 250. P. A. J. Janssen, U. S. Patent 3,141,823 (July21, 1964); Chern. Abstr. 62, 14634 (1965). 251. C. Janssen, Fr. Patent 1,517,671 (March 22, 1968); Chern. Abstr. 70, 115015 (1969). 252. J. W. Cole and R. Hallas, U. S. Patent 3,686,187 (August 22,1972); Chern. Abstr. 77, 139823 (1972). 253. P. A. Janssen and G. H. P. Van Daele, Ger, Offen. 2,610,228 (September 3D, 1976); Chern. Abstr. 86, 29645 (1977). 254. P. A. J. Janssen and G. H. P. Van Daele, U. S. Patent 4,179,569 (December 18,1979); Chern. Abstr. 92, 128743 (1980). 255. R. S. Vartanyan, V. O. Martirosyan, E. V. Vlasenko, L. K. Durgaryan, and A. S. Azlivyan, Khirn.-Farrn. Zh. 15, 43 (1981); Chern. Abstr. 95, 24753 (1981). 256. R. Hallas and C. W. Cole, U. S. Patent 3,691,176 (September 12, 1972); Chern. Abstr. 77, 151984 (1972).
References
37\
57' R.-S. Zong, D.-X. Yin, and R.-Y. Ji, Yao Hsueh H!>ueh Pao 14, 362 (1979). 258. S.-H. Zee, C.-L. Lai, Y.-M. Wu, and G.-S. Chen, K'o Hsueh Fa Chan Yueh K'an 9, 387 (1981).
\s vl-o-<
259. S.-H. Zee and W.-K. Wang, J. Chin. Chem. Soc. (Taipei) 27, 147 (1980). 260. A. Jonczyk, M. Jawdosiuk, M. Makosza, and J. Czyzewski, Przem. Chern. 57, 131 (1978). 261. A. F. Casy, M. M. A. Hassan, A. B. Simmonds, and D. Staniforth, J. Pharm. Pharrnacol. 21, 434 (1969).
~
262. W. F. M;. Van Bever, C. J. E. Niemegeers, and P. A. J. Janssen, J. Med. Chem. 17, 1047 (1974). 263. W. Jin, H. Xu, Y. Zhu, S. Fang, X. Xia, Z. Huang, B. Ge, and Z. Chi, Sci. Sin. (Engl. Ed.) 24, 5 (1981). c;64. -./ M. W. LobbezoQ, W. Soudijn, and I. van Wijndaarden,J. Med. Chern. 24,777 (1981); M. w. Lobbezoo, W. Soudijn, and I. van Wijngaarden, Eur. J. Med. Chern. 15,357
\
(1980). 265. M. W. Lobbezoo, W. Soudijn, and I. van Wijnaarden, 4, 357 (1980).
EUr. J. Med. Chern. Chirn. Ther.
'
266. A. F. Casy and F. O. Ogungbamila, Eur. J. Med. Chern. Chirn. Ther. 18, 56 (1983). 267. T. N. Riley, D. B. Hale, and M. C. Wilson, J. Pharm. Sci. 62, 983 (1973). 268. P. A. Janssen, W. F. M. Stokbroekx, and A. Raymond, U. S. Patent 3,907,813 and 3,907,811 (September 23, 1975); Chern. Abstr. 84, 4821 (1976). 269. J. E. Leysen and P. M. Laduron, Arch. Int. Pharrnacodyn. 232, 243 (1978). 270. J. S. Finch and T. J. Kornfeld, J. Clin. Pharmacol. 7,46 (1967). 271. R. Schleimer, E. Benjamini, J. Eisele, and G. Henderson, Clin. Pharmacal. Ther. 23, 188 (1978). 272. C. C. Hug. Jr., "The Pharmacokinetics of Fentanyl." Janssen Pharmaceutica, Inc., 1981. 273. K. C. Rice, A. E. Jacobson, T. R. Burke, Jr., B. S. Bajwa, R. A. Streaty, and W. A. Klee, Science 220, 314 (1983); T. R. Burke, Jr., B. S. Bajwa, A. E. Jacobson, K. C. Rice, R. A. Streaty, and W. A. Klee, J. Med. Chem. 27. 1570 (1984). 274. Drugs Future 5, 197 (1980). 275. P. G. H. Van Daele, M. F. L. De Bruyn, J. M. Boey, S. Sanczuk, J. T. M. Agten, and P. A. J. Janssen, Arzneim.-Forsch. 26, 1521 (1976). 276. C. De Ranter, O. Peeters, and Y. Gelders, Arch. InI. PhYJiol. Biochim. 87, 1031 (1979).
277. W. F. M. Van Bever, C. J. E. Niemegeers, K. H. L. Schellekens, and P. A. Janssen, Fortschr.
Arzneirnittforsch
26, 1548 (1976).
278. Drugs Future 2, 334 (1977); 3, 410 (1978); 8, 472 (1983). ,J!-~~" 279. J. E. Leysen, W. Gommeren and C. J. D. Niemegeers, Eur. J. Pharmacol. 87.209 (1983). 280. C. J. E. Niemegeers, K. H. L. Schellekens, W. F. M. Van Bever, and P. A. J. Janssen, Fortschr. Arzneirnittforsch 26, 1551 (1976). ,\ 281. Drugs Future 6, 335 (1981); 7, 418 (1982). ~ ~\h~ C-3-1,M.. 282. D. R. Stanski and C. H. Jug, Jr., Anesthesiology 57, 435 (1982). 283. L. E. Mather, Clin. Pharmacokinet. 5, 422 (1983). (-40,,>,
1
1J
284.
Drugs
Future
3,800
(1978).
- Z",,..oIl>l,q
~"~.!,. 285. Drugs Future 5, 410 (1980). - o.,.f~-J.'n; I 286. J. D. Borel, Contemp. AnesIh. PracI. 7, 1 (1983). 287. A. P. Feinberg, I. Creese, and S. H. Snyder, Proc. Natl. Acad. Sci. U.S.A. 73,4215 (1976). 288. D. Lednicer and P. F. Von Voigtlander, J. Med. Chern. 22, 1157 (1979). 289. Z. G. Finney and T. N. Riley, J. Med. Chern. 23, 895 (1980).
376
7
Piperiqine Analgesics
290. W. Klein, W. Back, and E. Mutschler, Arch. Pharm. 308, 910 (1975). 291. R. F. Borne, 5.-J. Law, J. C. Kapcghian, and L. W. Masten, J. Pharm. Sci. 69, 1104 (1980). 292. J. G. Berger, F. Davidson, and G. E. Langford, J. Med. Chern. 20, 600 (1977). 293. B. E. Maryanoff, D. F. McComsey, R. J. Taylor, Jr., and J. F. Gardocki, J. Med. Chern.
24, 79 (1981).
294. R. F. Borne, E. K. Fifer, and I. W. Waters, J. Med. Chern. 27, 1271 (1984). 295. T. N. Riley and J. R. Bagley, J. Med. Chern. 22, 1167 (1979). 296. T. R. Burke, Jr., A. E. Jacobson, K. C. Rice, B. A. Weissman, and J. V. Silverton, NIDA Res. Monogr. Ser. 49, 109 (1984). 297. W. F. Simonds, T. R. Burke, Jr., K. C. Rice, A. E. Jacobson, and W. A. Klee, Proe. Nal/. Acad. Sci. U.S.A. 82, 4974 (1985). 298. T. R. Burke, Jr., A. E. Jacobson, K. C. Rice, and J. V. Silverton, J. Med. Chern. (in press, 1986).
8. Physical Chemistry, Molecular Modeling, and QSAR Analysis of the Arylpiperidine Analgesics l. Physico~hemical Studies
.
377 378 380 383 385
A. X-Ray Crystallography B. Protdn NMR . . C. Carbdn-13 NMR . . . . . . . . . . . . . . . . II. Stereostructure, Conformation. and Biological Activity ....... III. Molecular Modeling and Ouantitative Structure-Activity Relationship (QSAR) Studies ... A. Molecular'Modeling B. OSAR Studies. . References
1.
Physicochemical
388 388 394 398
Studies .
Embedded in Ihe struclure of morphine is a piperidine ring in a chair conformation, with an axial phenyl ring bonded at the 4-position (Fig. 8-1). This was recognized early on and was postulated to be a key pharmacophore. Since the late 1940s, numerous 4-phenylpiperidine-based analgesics have been synthesized and lested (for reviews, see references 1 and 2). Intereslingly, however, most of the active compounds in this series are not expected to exist in a phenyl-axial conformation to any significant extent. Conformation and stereostructure-activity relationships in this series are of further interest because the molecule is prochiral; substitution on the piperidine ring renders the 4-position optically active, and stereoseleclivily is observed in the biological activities of these compounds. The two sides of the piperidine ring may be labeled pro-4R and pro-4S, as shown in Fig. 8-2. Numerous 3-alkyl derivalives and all possible mono- and dimethyl derivatives have been synthesized and studied. X-Ray crystallography and nuclear magnetic resonance (NMR) speclroscopy have played key roles in
Fig. 8-1.
Morphine
structure,
HO'" highlighting
the embedded
377
4-phenylpiperidine
fragment.
378
8
Physical Chemistry, Molecular Modeling, QSAR of Arylpiperidine Analgesics
o
I
Fig. 8-2. (after
Meperidine Portoghese,
structure, reference
showing the pro-4R and pro-4S edges of the piperidine 2).
elucidating the relationships between molecular configurations, conformations, and biological activities in this class of aryl piperidine analgesics. (Note: Some of the early literature dealing with 1,2,5-trimethylpiperidine derivatives has discrepancies in usage of the terms a, /3, y,o. Reference 1 clarifies these ambiguities and should be consulted when reading the literature on these compounds.) The next three sections describe the application of crystallography, proton NMR, and carbon-13 NMR to the characterization of arylpipcridine analgesics. The following section focuses on stereostructure, conformation, and analgesic activity, in which the results of all of these methods are integrated and interpreted in terms of the observed biological activities. A.
caDEt 1
meperidine
2
c2 -prod
ine
X-Ray Crystallography
Crystallographic studies have contributed substantially to an understanding of the steric requirements for receptor binding of arylpiperidines. Table 8-1 lists the compounds whose structures have been determined
'[(0 o
X-Ray Crystallographic
Studies of Arylpiperidine
4
Conformation
Meperidine (I) a-Peadine (2) {J-Prodine (3) a-Allylprodine (4) {J-Attylprodine (5) {3-1,2-Dirnethyl-4-phenyl-4-propionyloxypipcridine a-l,2,3- Trimethyl-4-phenyl-4-piperidinol (7) {3-1,2,3- Trimethyl-4-phenyl-4-piperidinol (8) y-l ,2,3- Trimcthyl-4-phenyl-4-piperidinol (9)
(6)
a-I ,2,5- Trimethyl-4-phenyl-4-piperidinol (1 o)a y-l,2,5- Trimethyl-4-phenyl-4-piperidinol (11) 1,2,6- Trimcthyl-4-phenyl-4-acetoxypiperidine (12) y-I,3 ,5- Trimethyl-4-phenyl-4-propionoxypiperidine 9 incorrectly
labels the a-isomer
prodine
(13)
as the ,B-isomcr.
Chair Chair Chair Chair Chair Chair Chair Chair Chair
phenyl-equatorial phenyl-equatorial phenyl-equatorial phenyl-equatorial phenyl-equatorial phenyl-equatorial phenyl-equatorial phenyl-equatorial phenyl-equatorial
Chair phenyl-axial Chair phenyl-equatorial Chair phenyl-equatorial Chair phenyl-equatorial See reference
J for corrected
Reference 3 4 5 6 6 7 8 8 8 9 9 11 11
OR
/f(0
o 5
6
{3 -allyl
R
= H, COEt
prodine
OR
structures
in this series.
7
\
a -allyl
Analgesics
Compound
Reference
-"')(0 o
3 {3 -prodine
Table 8-1
a
379
Studies
crystallographically. The conformation of the piperidine ring in the solid state is also reported in Table 8-1. In some cases, the structure studied has been the parent alcohol rather than the biologically active ester. Casy has presented reasons for assuming that the alcohol and ester will possess similar solid-state conformations in most cases (12). The solid-state conformation of the piperidine ring is a chair for all compounds. Only one of these compounds (10) has an axial phenyl substituent. Among the three isomeric 2,5-dimethyl-piperidines, this isomer is the most potent analgesic.
~N-CHJ
~'O_4!1 COOEt
ring
Physicochemical
I
1
R:::
H.
Ac
8
R
'"
H, Ac
.380
8
Physical Chemistry, Molecular Modeling, QSAR of Arylpiperidine -Analgesics
Table 8.2 Proton NMR Studies of Arylpiperidine Analgesics Compound a-Prodine-alcohol
State
Solvent
(2)
Base HCI
COCl, COCl,
HCI
0,0
(3)
Base HCI HCI HCI HCI HCI HCI HCI HCI Base
COCl, COCI, 0,0 COCl, COCl, 0,0 COCI, COCl, COCl,
HCI
OMSO
Base HCI Base HCI
COCl, OMSO COCI, COCI,
Base HCI Base
" p-Prodine-alcohol 9
R
= H, Ac
10
It" a-Prodine (2) /3-Prodine
(3)
\
a-Aliylprodine (4) '0" /3-Aliylprodine (5) '0" a-2-Methyl-alcohol (IS)
OH 11
a-2-Methylpropionyl ester (IS)
11-2-Methylpropionyl ester (6) a-I,2,3-Trimethyl-4phenyl-4-piperidinol
13
B. Proton NMR
COCl, COCl, COCl,
/7 /7 /8
Base
COCl,
Chair phenyl-equatorial
/8
Base
COCl,
Chair phenyl-equatorial
18
Base
CCI,
Twist-boat (16)
/8
HCI
COCl,
Chair phenyl-equatorial
/8
HCI
COCl,
Chair phenyl-equatorial
18
HCI
COCl,
Chair phenyl-equatorial
/8
HCI
0,0
Chair phenyl-equatorial
/8
0,0
Since proton-proton coupling constants for vicinal protons are dependent on the torsional angles between the protons, proton NMR is particularly well suited to studying the conformations of six-membered ring compounds (for example, see reference 13). Thus, it is not surprising that many papers have appeared in which the solution conformation(s) of arylpiperidine analgesics have been examined by this method. Proton NMR studies of these compounds are summarized in Table 8-2. It is important to note that the solvent, the presence or absence of the ester group, and the state of the basic nitrogen atom all influence the results. For instance, the 2,3-dimethylpiperidine isomer, 9, was found to exist in a phenyl-equatorial chair conformation with both methyl groups axial in the solid state (8) and, under most conditions, in solution (18). However, the parent alcohol in dilute solution in CCl. exists in the boat form 16, due to the intramolecular hydrogen bond shown. It is tempting to assume that the
14 14 14 /4 14 14 /2 /2 /2 /2 /5,16 /5,/6 /7 /7
17 17 17 17
(7)
/3-1,2,3- Trimethyl-4-
phenyl-4-piperidinol
References
Chair phenyl-equatorial Chair phenyl-equatorial Chair phenyl-equatorial Chair phenyl-equatorial Chair phenyl-equatorial Chair phenyl-axial" 100% chair phenyl-equatorial Chair phenyl-equatorial 75-80% chair phenyl-equatorial Twist-boat (14)" Chair phenyl-equatorial Chair phenyl-equatorial Twist-boat (15) and/or chair, phenyl-equatorial Twist-boat (15) and/or chair, phenyl-equatorial; epimeric conjugate acids Chair phenyl-equatorial Chair phenyl-equatorial Chair, phenyl-axial Chairs phenyl-axial + phenylequatorial 2 : I, epimeric conjugate acids Chair phenyl-equatorial Chair phenyl-equatorial Chair phenyl-equatorial
12 /3-2-Methyl-alcohol (6)
Conformation(s)
(8)
')'-1,2,3- Trimethyl-4-
phenyl-4-piperidinol (9) ')'-1,2,3-Trimethyl-4phenyl-4-piperidinol (9) a-I,2,3- Trimethyl-4phenyl-4-propionyloxypiperidine (7) /3-1,2,3-Trimethyl-4phenyl-4-propionyloxypiperidine (8) 1"1,2.3- Trimethyl-4phenyl-4-propionyloxypiperidine (9) ')'-I,2,3-Trimethyl-4phenyl-4-propionyloxypiperidine (9)
(continued)
\
1
382
8
Physical Chemistry, Molecular Modeling, QSAR of Arylpiperidine
Analgesics
I
Physicochemical
Studies
383
Table 8-2 (Cont.) Compound a-l.2,5- Trimelhyl-4phenylpiperidinol (j-l,2.5-Trimcthyl-4phcnylpipcridinol y-t ,2,5- Trimethyl-4phenylpiperidinol S.l,2.5-Trimethyl-4phcnylpipcridinol 4-Acctoxy-l,2.6trimethyl-4-phenylpiperidine (19) 4-Aceloxy-I,2.6trimethyl-4-phenylpiperidine (20) 4-Acctoxy-I,2,6trimethyl-4-phenylpiperidine (21) Tropane analog (22) a
State
Solvent
Base
COo.,
Chair phenyl-axial
/9
Base
CDCll
Chair phenyl-equatorial
/9"
Base
CDC/_!
Chair phenyl-equatorial
/9
Base
CDCI,
Chair phenyl-equatorial
/9"
HCI
CDCl)
Chair phenyl-equatorial
20
HCI.
COo.,
Chair phenyl-equatorial
20
Conformation(s)
References
(10) (17) 17
18
(II)
CI
(18)
OAe
OAe HCI
CDCI)
Chair phenyl-axial
20 19
HCI
Casy et at. initially interpreted
CDCl)
20
2/
Chair phenyl-axial
the proton NMR spectra of these compounds
in terms of
phenyl.axial-chair and twist-hoat conformations (12). Later, they considered these compounds to be in phenyl-equatorial chair conformations on the basis of carbon.13 NMR spectra (15). h Reference 19 has the lahels {3 and () interchanged; see reference 1 for corrected structural assignments.
21
14
.B-prodine, conformation
22
conformation in aqueous solution is most relevant to the biological activity, but the degree of solvation and the hydrophobicity of the receptor binding site are not yet known. In addition, drug receptor interactions could affect the receptor-bound conformation. Thus, it is necessary to take into consideration all of the observed conformer states in assessing structureactivity relationships.
twist-boat
C. Carbon-I3 NMR
15 R
H, COEt
16
Jones and co-workers interpreted the carbon-13 NMR spectra of the 1,2and 1,3-dimethylpiperidine compounds (IS, 6, 2, and 3) in terms of their conformations as previously derived from proton NMR and X-ray crystallography (22). From these compounds and a few simpler model compounds, they derived a set of additivity parameters for the piperidine ring
384
8
Physical Chemistry, Molecular Modeling, QSAR of Arylpiperidine
Analgesics
II
II.
carbon resonances. They then used these parameters to interpret the carbon-13 NMR spectra of several 1,2,5-trimethylpiperidine compounds (10, II, and 17) in terms of their configurations and conformations (23). (As described previously, reference I should be consulted for corrected usage of the terms a, (3, y, and 8 in this series.) Jones et al. (22) concluded that the a-isomer is in a chair conformation with the phenyl axial, as depicted in 10. They interpreted the spectrum of the {3-isomer (17) in terms of a twist-boat conformation, as shown in 23. The y-isomer was projected to have a chair conformation with the phenyl group equatorial, as shown forIl. OR
R
""
H,
COEt
/
OCOEt 24
OAe 25
The same approach was used by Casy and co-workers in determining the solution conformations of the three isomeric 1,2,3-trimethyl derivatives 7, 8, and 9 (/8). All three of these compounds (as the ester hydrochlorides in COCI3) were concluded to exist in phenyl-equatorial chair conformations, even though this forces both methyl groups to occupy axial positions in yisomer 9.
L
Conformation,
Stereostructure, Biological ,
/
23
Stereostructure,
and Biological Activity
Conformation,
385
and
Activity
Meperidine (I) and the 3-alkyl compounds 2-5 all exist in the solid state and in solution in a phenyl-equatorial chair conformation. Placing a 3-alkyl (methyl, ethyl, propyl) equatorial substituent on the pro-4S edge of the piperidine ring has essentially no effect on analgesic activity, while a 3-alkyl equatorial substituent on the pro-4R edge of the ring results in a substantial decrease in activity. 3-Alkyl axial substituents produce a slight decrease in activity on the pro-4R edge of the ring and an increase in activity on the pro-4S edge of the ring. The 3-allyl substituent behaves differently from simple alkyl (including n-propyl) substituents; the allyl derivatives are substantially more active than the unsubstituted parent compound, suggesting a specific nonsteric interaction with the receptor (6,24). The a-2-alkyl compound 15 is conformationally mobile in solution (/7). The parent alcohol (both the hydrochloride and the base) appears to exist as a mixture of the twist-boat form shown in structure 15 and the phenyl-equatorial chair; the hydrochloride also appears to be a mixture of the epimeric conjugate acids in which either the N-methyl group or the proton may be equatorial. The twist-boat form can be stabilized by intramolecular hydrogen bonding. Esters of 15 in COCl, solution are in the phenyl-axial chair conformation (base) or in a 2:1 mixture of the phenylaxial chair and phenyl-equatorial chair forms. The racemic mixture has approximately the same analgesic activity as the parent compound with no 2-substituent. In view of the conformational flexibility of this compound, it is difficult to draw conclusions about its specific receptor requirements. The {3-2-methyl compound 6 is in a phenyl-equatorial chair conformation with the 2-methyl group in an equatorial position both in the solid state (7) and in solution. (/7). When the methyl group is on the pro-4R edge, the compound is as active as the desmethyl parent; the 4S isomer shows decreased activity, suggesting steric interference with receptor binding. The esters of all four of the 2,3-dimethyl compounds (7, 8, 9 and 24) appear to favor the phenyl-equatorial chair conformation, on the basis of X-ray crystallographic, proton NMR, and carbon-13 NMR studies (8,/8). This finding is surprising only in the case of y-isomer 8, in which both methyl groups are in axial positions. The parent alcohol of this isomer in CCl4 solution can adopt a twist-boat conformation, stabilized by an intramolecular hydrogen bond (16), but the esters show no such tendency. The {3-isomer, 8, has both methyl groups in equatorial positions. Based on the findings that a pro-4R-equatorial 3-methyl group or a pro-4Sequatorial 2-methyl group decreases analgesic activity, it could be expected that neither of the {3-isomers would show good analgesic activity; this is indeed the case. Similarly, 8-isomer 24 must have either an axial 3-methyl
386
8
Physical Chemistry, Molecular Modeling, OSAR of ArylpiperidiDc Analgesics
unfavorably placed on the pro-4R edge or an equatorial 2-methyl unfavorably placed on the pro-4S edge; the racemic mixture is inactive. The racemic y-2,3-dimethyl compound 9 is slightly more active than the desmethyl parent compound. As described above, this compound should have the 2- and 3-methyl groups axial. Since an axial 3-methyl group improves activity on the pro-4S edge and diminishes activity on the pro-4R edge (see above), it follows that the axial 2-methyl group on the pro-4S edge is sterically acceptable. This interpretation is further supported by the finding that the tropane analog 2S (comparable to having 2,6-diaxial substituents) retains activity (25). The racemic a-isomer has activity near that of the desmethyl compound. Most of the activity is expected to reside in the compound having the methyl groups on the pro-4S edge, since neither the 2-axial nor the 3-equatorial methyl should interfere with binding. On the pro-4R edge, the 3-equatorial methyl group was previously shown to decrease activity (see above). All four of the 2,5-dimethyl compounds (the promedols, 10, 11, 17, and 18) have been studied by NMR spectroscopy (19,23). In addition, two of the structures have been determined crystallographically (9). (See reference 1 regarding discrepanci~ in the nomenclature of these compounds in the literature.) The a-isomer, 10, is one of the few compounds in which an axial phenyl group is apparently favored, since it puts both of the methyl groups in equatorial rather than axial positions. For the a-isomer, analgesic activity resides in the enantiomer with the equatorial 2-methyl group on the pro-4S edge and the 5-methyl group in the pro-4R edge; the other enantiomer is inactive. The remaining isomers, 11, 17, and 18, all preferentially adopt the phenyl-equatorial chair conformation. The racemic f3-isomer, 17, is as active as the desmethyl parent. On the basis of the discussion above, it is expected that the activity originates in the isomer having the axial 2-methyl on the pro-4R edge and the equatorial 5-methyl on the pro-4S edge. The y-isomer 11 has been resolved into its enantiomers. When the equatorial 2-methyl is on the pro-4R edge and the equatorial 5-methyl is on the pro-4S edge, the analgesic activity is equivalent to that of the desmethyl parent. In this case, both methyl groups are in sterically acceptable positions, as described above. Conversely, if the equatorial 2-methyl is on the pro-4S edge and the equatorial5-methyl is on the pro-4R edge, both substituents are in unfavorable positions, and this compound is substantially less active. Finally, the racemic o..isomer 18 shows good analgesic activity. Again, the analgesic activity is expected to come from the enantiomer having the equatorial 2-methyl group on the pro-4R edge and the axial 5-methyl group on the pro-4S edge. The other enantiomer of this compound will have both methyl substituents in unfavorable positions, as described above.
II
Stcrcostructure,
Conformation,
and Biological Activity
387
There are three possible isomers (19, 20, and 21) in the 2,6-dimethyl series. Each of these has been studied by proton NMR (20), and one has been examined crystallographically (10). The first two compounds are in phenyl-equatorial chair conformations. Racemic 19 is more active than the desmethyl compound. When the equatorial methyl group is on the pro-4S edge, it is expected to interfere with biological activity, so the activity should reside primarily in the compound having an equatorial methyl group on the pro-4R e.dge and an axial methyl group on the pro-4S edge. Compound 20 has an equatorial methyl group on the pro-4S edge and was found to be inactive. Like compound 20, isomer 21 has two equatorial methyl groups. In 21, the phenyl group is axial. This compound is also inactive as an analgesic. Among the three isomeric 3,5-dimethyl compounds, only one shows any activity: isomer 13 (11). A crystallographic study showed that this compound is in the usual phenyl-equatorial chair conformation. The less active isomer has an axial methyl unfavorably placed on the pro-4R edge; the more active isomer has an equatorial methyl unfavorably placed on the pro-4R, but it has an axial methyl group on the pro-4S edge, which was shown to increase analgesic activity (see above). The results for the phenyl-equatorial compounds are summarized in Fig. 8-3. On the pro-4R edge of the piperidine ring, substitution on the 2-position is acceptable, while 3-substituents decrease activity. On the pro-4S edge, substitution is acceptable on the 3-position and on the 2-axial position; 2-equatorial substitution leads to decreased activity. The data are less extensive for the phenyl-axial compounds. Figure 8-4 summarizes the results determined to date. On the pro-4S edge of the ring, a 2-equatorial substituent decreases activity, while 2-axial and 3-equatorial substituents have no effect. Both axial and equatorial substituents are acceptable on the 2-position of the pro-4R edge, but pro-4R 3-equatorial substitution decreases activity. X-ray crystallographic studies have revealed another structural feature that correlates with analgesic activity (2,Il). The more active enantiomers
.
Fig. 8.3. Summary of structure-activity relationships for methyl-substituted phenylequatorial-chair piperidines. 0, no effect on activity; @, decreased activity; ([Y, increased activity.
388
8
Physical Chemistry, Molecular Modeling, QSAR of Arylpiperidip.e Analgesics
III
Molecular Modeling and QSAR Studies
JS9
a
Fig. 8.4. Summary of structure-activity chair piperidines. Key as in Fig. 8-3.
I
RIal
relationships for methyl-substituted
phenyl-axial-
I~ ,. c (bJ
R Fig. 6-5. Rotation of the phenyl ring relative to the piperidine ring in ~a) mo~e active and (b) less active isomers of arylpiperidine analgesics in the phenyl-equatonal chaiT configuration. See Reference 2 for details.
are those in which the phenyl ring is in the conformation shown in Fig. 8-5a , and the less active enantiomers are as shown in Fig. 8-5b. This conformational feature is "controlled by the presence or absence of substituents in the 3- and 5-positions of the piperidine ring.
III. A.
Molecular Modeling and Quantitative StructureActivity Relationship (QSAR) Studies Molecular Modeling
Froimowitz (26) has carried out a conformational study of various phenylpiperidine analgesics (the prodines, ketobemidone, meperidine, and 1,3,4-trimethyl-4-phenylpiperidines) using Allinger's molecular mechanics program MM2 (27). Phenyl equatorial conformations were found to be preferred for the prodines, ketobemidone, and meperidine. The calculated equatorial and axial phenyl conformations for the prodines are shown in Fig. 8-6. For ketobemidone and meperidine, however, phenyl axial conformations were computed to be only 0.7 and 0.6 kcal/mole, respectively, higher in energy. It was suggested that phenyl axial conformers may be responsible for the potency-enhancing effect of a phenyl meta-hydroxy group in these two compounds. In contrast, phenyl axial conformers were computed to be relatively unfavorable for the prodines, being 1.9, 2.8, and 3.4 kcal/mole higher in energy for 3-demethyl, a-, and j3-prodine, respectively. Froimowitz related
Fig. 8-6. Lowest-energy phenyl equatorial and phenyl axial conformers for the prodines. Relative steric energies are (a) 10.5 and 12.4 kcal/mole for 3-demethylprodine, (b) 12.9 and 15.7 kcal/mole for a-prodine, and (c) 13.2 and 16.6 kcal/mole for l3-prodine.
the relative concentrations of an analgesic conformation to the potencies of these three compounds (see Table 8-3). A phenyl axial conformer was calculated to be preferred by 0.7 kcal/mole for the 3-demethyl compound of 1,3,4-trimethyl-4phenylpiperidine, with phenyl equatorial conformers preferred by 1.3 and Table 8-3 Correlation of the Relative Potencies of 3-Demethylprodine, a-Prodine, with the Relative Concentration of Their Analgesic Conformationa
3-Demethylprodine (3R,4S)-a-Prodine (3S,4S).fJ.Prodine
and I3-Prodine
ED.'\(I (mgjkg)
Relative Potencyh after Brain Penetration
Energy Difference, kcal/mole
Relative Concentration of Analgesic Conformation
1.00 1.45 5.2t
1.00 1.24 1.77
0.0 0.5 3.7
1.00 1.40 2.00
a Differences in brain penetration have been adjusted for. Energy differences are between the two conformers that have identical energies when the piperidine ring does not contain a substituent in the 3-position. The analgesic conformation is assumed to be the one that is favored by a substitution on the pro-4S edge of the piperidine ring. h Reference
28.
390
8
Physical
Chemistry,
Molecular
Modeling,
QSAR
of ArylpipcriQine
Analgesics III
3.3 kcal/mole for the a and {3compounds. Phenyl axial conformers were unexpectedly found to be especially destabilized by a 3-methyl group in the (3 configuration due to the steric crowding of the three piperidine substituents. Comparisons were also made between the computed structures and those observed by X-ray crystallography (4,5). More recently, Froimowitz and Kollman (29) carried out additional conformational energy calculations on various prodine derivatives using both MM2 (molecular mechanics) and the semiempirical quantum mechanical (PClLO) methods. Compounds studied include 3-demethylprodine, a-prodine, {3-prodine, the a-2-methyl derivative, a-promedol, the y-2,3dimethyl derivative, and y-isopromedol. All of the compounds are predicted to activate the opiate receptor in a phenyl equatorial conformation. Optimum activity is postulated to result from a specific orientation of the phenyl and propionoxyl groups. The a-promedol analog is calculated to be most stable in a phenyl equatorial conformation which is in disagreement with experimental data. Two mirror-image phenyl equatorial conformers are preferred for 3demethylprodine. The more active prodine antipodes consistently prefer the conformer in which the phenyl orientation is opposite to (i.e., the mirror image of) that of mOrphine and the morphine-like ( + )_ phenylmorphan. The authors suggest that this observation may be the molecular basis for the non-morphine-like behavior that arises with the introduction of a phenyl meta-hydroxyl into some prodine analogs. The findings of Froimowitz are qualitatively similar to those of Loew and Jester (30), who also carried out PClLO calculations on meperidine and the prodines. The energy differences between phenyl equatorial and phenyl axial conformers were found to increase in the order meperidine, 3-demethylprodine, a-prodine, and {3-prodine. However, the energy differences are three to nine times greater according to the PClLO calculations compared to the molecular mechanics results. This, however, may be a consequenc~ of Loew and Jester's failure to carry out a complete structure optimization. Loew and Jester (30) also computed the charge distributions for the compounds they studied and superimposed the minimum energy conformers on an energy-optimized structure of morphine. Figure 8-7 illustrates, for example, the minimum-energy conformer of meperidine superimposed on morphine (dotted structure). The charge densities are also shown in Fig. 8-7. A correlation between the calculated energy difference between equatorial and axial energies and analgesic potency was observed for four compounds, as reported in Table 8-4.
Molecular
Modeling
and OSAR
Studies
391 +0.16 H
+0.02
-0.05 .0.02
+0.08
.0.07
.0.04 /,.
+0.08
+0.02
-:0.02
/ .0.02 {
"\
0
:
;.
~
0.3...]"
/'
..."
0
:
-0.11:
I
.0.17
-
,
:
J
" "
-0.02 Fig. 8.7. on that
Minimum-energy
of morphine
conformer
(pharmacophore
of meperidine
with a piperidine
ring superimposed
I) and nct atomic charges.
Table 8-4 Relationship between Potency and Relative Axial Energies in Meperidine and Prodines Drug dE (eq-ax)" Potency (EDso)
('" )-IHrodine 21.0 0.32
(:t)-a-Prodinc
DesmethyJprodine
Meperidine
8.6 1.7
6.6 1.3
5.3 13.t
"ilE in kcalj molc.
In a set of analogs of the prodine analgesics, the energy of the highest occupied molecular orbital of the aryl moiety correlates with the I?g ED50 obtained from the mouse hot plate assay by subcutaneous admInIstratIOn (31). The orbital energies were not actually computed, but taken from a data table. Thus, there is a concern regarding the reliability of these measures in the actual compounds of interest. Also, only six compounds have been used to establish the correlation. Nevertheless, this observation suggests a possible charge transfer interaction between the ~ryl groups of the analgesics and their receptors, wIth the aryl groups actIng as charge donors. This model can and should be tested. Isoelectrostatic contour spheres for morphine, meperidine, and aprodine were constructed by Breon et al. (32) usin? quantum mechanics. Minor configurational changes were made In meperIdIne and a-prodIne to approximate the spatial configuration of morphine, the most rigid analog
Table 8-5
0
Hot Plate Analgesic Potency and Physicochemical
Parameters of Substituted Benzoic ACId Esters of l-Methyl-4-piperidinol'
~o-GN-CH,
.ft})Rn
LogO/c)' No.
1 2' 3 4' 5 6' 7' 8 9 HI' 11 12' 13 14 15 16 17 18 19' 20 21 22 23 24' 25 26 27' 28
R,
h
3,4-(OCH3h 4-0C3H3 3-0CH3 4-CN 3,4.(OCH,O) 4-0-n-C~H9 2,4,6-(CHJh 2,3-(OCH3h 3,5-(OCH3h 2-CF) 2-CH) 2-NO] 2,4,6-(OCH)h H 4-0CH3 2-0CH) 2-F 3-0CH),4-CH) 3-CN 4-F 2,5-(CH)h 3,4,5-(OCHJh 3-F,4-CHJ 2.CH3C~H9 2,6-(CH]h 3-F 2.CI 3.0H
~m
0.08 0.38 -0.02 -0.57 -0.05 1.55 1.29 0.08 0.08 0.88 0.56 -0.28 0.06' 0 -0.02 -0.02 0.14 0.54 -0.57 0.]4 1.07k -0.6 0.7 2.01 1.07 0.14 0.71 -0.67
0.04 0 -0.02 0 -0.025 0 0 -0.02 0.08 0 0 0 0 0 0 0 0 -0.02 -0.57 0.0 0.56 -0,4 0.14 0 0 0.14 0 -0.67
~o
0 0 0 0 0 0 0.86 -0.02 0 0.88 0.56 -0.28 0.04 0 0 -0.02 0.14 0 0 0 0.56 0 0 2.01 1.07 0 0.71 0
L" 0 0 0 0 0 0 1.8B 1.92 0 1.24 0.94 1.38 3.84 0 0 1.92 0.59 0 0 0 0.94 0 0 1.57 1.88 0 1.46 0
81,m. 0.35 0 0.35 0 0.2' 0 0 0.35 0.7 0 0 0 0 0 0 0 0 0.35 06 0 0.52 0.7 0.35 0 0 0.35 0 0.35
FI.m -0.55 0 -0.55 0 -0.5V 0 0 -0.55 -1.1 0 0 0 0 0 0 0 0 -0.55 -0.51 0 -1.24 -1.1 -0,46 0 0 -0,46 0 -0.55
Bl,pQ 0.35 0.35 0 0.6 0.2 0.35 0.52 0 0 0 0 0 0.35 0 0.35 0 0 0.52 0 0.35 0 0.35 0.52 0 0 0 0 0
EJ,p
HBo
HBm
-0.55 -0.62~ 0 -0.51 -0.55
0 0 0 0 0 0 0 1.128 0 1.078 0 1.918
1.128 0 1.128 0 1.128 0 0 1.128 1.128 0 0 0 0 0 0 0 0 1.128 1.898 0 0 1.128 0 0 0 0 0 1.0
-0.94' -1.24 0 0 0 0 0 -0.55 0 -0.55 0 0 -1.24 0 -0.46 0 -0.55 -1.24 0 0 0 0 0
0'
0 0 1.128 0 0 0 0 0 0 0 0 0 0 0 0
HB, 1.128 1.248 0 1.898 0 1.248 0 0 0 0 0 0 1.128 0 1.128 0 0 0 0 0 0 1.128 0 0 0 0 0 0
Obsd.
.
Calcd.c
1.91 1.72 1.67 1.65 1.61 1.58 1.55 1.55 1.53 1.49 1,48 1,48 1.47 1.43 1.43 1.43 f,41 1.4 1.39 1.38 1.36 1.35 1.32 1.28 1.27 1.26 1.25 1.2J
1.84 1.61 1.63 1.62 1.65 1.56 1.11 1.49 1.43 1.35 1.37 1.34 1.33 1.44 1.62 1.30 1.40 1.44 1,48 1.37 1.29 1.36 1.07 1.32 1.30 1.26 1.33 1.24
0.07 0.11 0.04 0.03 -0.04 0.02 -0.44 0.06 0.10 0.14 0.1I 0.14 0.14 -0.01 -0.19 0.13 0.01 -0.04 -0.09 0.01 0.07 -O.o.t 0.25 -0.04 -0.03 0.00 -0.08 -0.03
1.21 1.17 1.17 1.14 1.13 1.11 1.08 1.07 0.85 0.63 1.52' 1.24" 1.18" 1.15" 1.15" 0.86" inact
1.31 1.25 1.10 1.31 1.01 1.29 1.08 1.25 0.85 1.15 1.12 1.26 1.23 1.29 1.17 0.97 1.27 1.35
-0.10 -0.08 0.07 -0.07 0.12 -0.18 0.00 -0.18 0.00 -0.52 DAD -0.02 -0.05 -0.14 -0.02 -0.11
~''';{''toot'
29' 30 31 324.11 33 34' 35 36 37' 38' 39"'.11 40 41 42' 43 44' 45' 46
2-Br 4-CH) 2,4,5-(CH)) 4-NO] 4-C(CH3h 4-CI 3,4-CIJ 3,5-(CHJ)3 4-CJ19 2,6-(OCH)h 2-Cr.H9 2-0C6H9 2-0C)H9 2.CzH9 3,4-(CHJ)3 2-C3H~Ct.H) 2-CN 3-CH2
47' 48
11 2,3,5-4)
. fi
0,86 0.56 1.5 -0.28 1.98 0.71 1.25 1.07 1.96 0.08 1.96 2.08 0.38 1.02 0.99 2.66 -0.57 0.56
0 0 0.56 0 0 0 0.71 0.107 0 0 0 0 0 0 0.56 0 0 0.56
1.12 3
0 2.00
0.86 0 0.56 0 0 0 0 0 0 0.08 1.96 2.08 0.38 1.02 0 2.66 -0.57 0
1.77 0 0.94 0 0 0 0 0 0 3.84 4.22 2A5 2.86 2.05 0 6.27 2.17 0
0 0 0.52 0 0 0 0.8 1.04 0 0 0 0 0 0 0.52 0 0 0.52
0 1.12
0 2.17
0 2.3
"I 0 0 -1.24 0 0 0 -0.97 -2,48 0 0 0 0 0 0 -1.24 0 0 -1.24
0 0.52 0.52 0.7 1.59 0.8 0.8 0 2.11/ 0 0 0 0 0 0.52 0 0 0
0 -1.24 -1.24 -2.52 -2.78 -0.97 -0.97 0 -3.82/ 0 0 0 0 0 -1.24 0 0 0
0 2.B
1.15 0
1.4 0
0 0 0 0 0 0 0 0 0 Of 0 01.248 0 0 0 1.898 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0
0 0
0 0
0 0
0 0 0 0 0 0 0
V
0 0
/
!918
1.23 0.00
From reference 35. The listed values are the actual values minus the value for
that by thetheunsubstitutcd compound (14)E.can have zero all J.theMed. parameters. b C: EDm (mmoljkg). Testcd subcutaneously as water-soluble "H" HCIsosalts hot plate method; d. A. Jacobson arM value E. L. for May, Chern. 8,563 < Dcrived from eq 12 of reference 35. 4 These derivatives were synthesized following the Craig's plot analysis. ~ -0.62: E..OCHI + (E..c)~ - E..CH): -0.55 + (-1.31 + 1.24). , Estimated from BI.ocH)' 1 -0.94: E..ocH) + (Ec2H) - E..cH3) :: -0.55 + (-1.63 + 1.24). ~ Omitted in deriving eq II of reference 35.
,
0.06:
1I').S-(OCH))3
+ 1I'OCH) :: 0.08
- 0.02.
(1%5).
I Assuming that a hydrogen bond cannot be formed because of unfavorablc conformation forced by the di-ortho-substituents. . The value of 1I').S-(CH)>J was used. I Because of the preferred perpendicular conformation of the biphenyl, maximum dimensions were used for the steric effect, i.e., E, (L) and B, werc used instead of E. (5) and B,. m Assunring the bulky phenyl group prevents "OCr.H9" to be a hydrogen-bond acceptor. Activity was observed in 4 to 5 out of 10 mice tcsted. "
00
o
~ ~
394
8
Physical
Chemistry,
Molecular
Modeling.
QSAR
of Arylpiperidine
oI
Analgesics
among these three compounds. Common areas of reactivity, potential energy minima, and charge densities were identified. However, this work only reinforces findings from earlier studies and does not provide new insights. B.
z"
§ o,
QSAR Studies
Benzoic acid esters of l-methyl-4-piperidinol possess analgesic activity, as revealed by the hot plate assay (33,34). The more potent members of this structural class are in the range of morphine and codeine but, in general, display little morphine-like physical dependence liability in monkeys. Classical QSAR analyses were carried out on the set of compounds reported in Table 8-5 (35). Among the substituent parameters included in the study, Lonho (length of ortho substituents) and BI (minimal width of substituents) or E, (Taft steric constant) at the meta and para positions yield inverse correlations with analgesic potency. Lipophilicity, especially in the meta position, and the ability of the meta position to also be a hydrogen bond acceptor are found to enhance analgesic potency. Cheng et al. (35) developed individual QSARs for ortho, meta, para, ortho andlor meta, meta, andlor para sets of substituents, and for the entire data base. In SOme cases (for example, the meta-substituted derivatives), the number of terms in the QSAR is large compared to the number in the data set. This leads to concern regarding the statistical significance and reliability of the QSAR and corresponding conclusions. The QSAR derived for all the active compounds in the data base is: log (I/C)
=
(0.14
;j:
(0.51 n
=
44; r
.t.
o ..'" «=
~< ~o ~ . 0< I~
0.03) E,.pam + o
(0.40 ;j: 0.10) HBme'a (0.72 ;j: 0.18) BI.me" + (0.25 ;j: 0.07) HBpam.;"d (0.07 ;j: 0.02) Lonho + ;j:
:;; ~ o,
0.17) Pime" + 1.44
~= 0' :::~
N< <;;0
o . I~< 0< ;:)0 ~ . 0< I~
(I)
= 0.77; s = 0.166; f = 9.14
where C is the EDso (millimoles per kilogram) for average hot plate activity. HB is the indicator variable for hydrogen-bonding effects, and Pi is the measure of lipophilicity. The number of compounds is represented by n, r is the correlation coefficient, s is the standard deviation of fit, andf is the measure of statistical significance. Equation (I) is not a particularly good fit to the biological data. The 2,4,6-trimethyl, 2,6-dimethoxy, and 2-phenyl derivatives were poorly predicted, with rather large residuals. The authors offer no reasons why
N N N N o
c z
N
0: ~ o,
~;!; ~;;; ~~~~~~~0 0
<
,
,
N
N N
" .1f ~U
< ;s
0
'E
0
.;:
~00
N 0
,
,
u z ;s
c
" ~8.eo
OJ 00 ~~~~..
Q~'-'
0_
::EN
.
~~:> + ~...J-
::E °~~~:>
~" 1f
,
0
,
0
i'1
N N
N N
N N
'"
0
,
,
,
0
00. 00 ,
g:'f
:8'f
o!::-
o!::-
o!::-
o!::-
°:>
0"
i
::E
00.
0.~~,
'"
0
...J
o!::-
~.
.8 '0 00 , E ~o. --:~~~~" o II . '" 1= o!::-5 -6 '0~01) ct.:: .. C
~~~". ~c
>. C C
>. -S 0
::;
M
-. 0
0
0
'6
.5
..~0
E .c0 ~U
.
E
C
. 0
3. - ~0, _0~<
"'-
°0
::E
.
:r
6<
cict
~~' -
';S
o!::-
N<~o
,,-, §~~~o. ° ~..!. .z
.
oO
:;;
~° :r
;:;
° ::E
'0 u > ...J "
a' :r-
';S
1;:< _0
~< ::10
-< ::10
,,<
6.<
6..{
6";'
6<
';S
';S
';S
1~
';S
"
a_
::EN d!:r
'"
NO
N< go
~<
6.{
0<-
I~
';s
N
S;;~~
3 0
Q
"'-
"
~~M
I;:-~o. ~z 0_ ,
N< ;;;0
.. c
o..j,
0;;"= 0.0
Compound
tii
o i:I: "
~~~~~No.
.-" 0"
';S
:;a:E~ ~~~~.':f-Q. 1: o ci~
~o.
-
N 8
~° i:>
§
g f f
"":z °,
,
0"
0
E go~ 0-'~"S :
~'0;;. 0
M
0
~e:g §
~~~~"
]
~N ~'"
0.1 ,c"a.
-
c:.>-;::
:2~~ 0
.0 0
:z
.'" 0 0 z
:c U
:c U
C
.~
~:c
D-U
N
:c
I :c ~:c D-U
U o
LJ
:c U
I
N ~:c U
I
I :c
Cn
IN
IN
IN
0
L ~"
:c" u-u
I :c
6 =
N
:c
U
Ie}
§
:c
D-U
a
~c;~M 1::
E
~~M f:'
E E U '0
e
~g;;; «1).-;:: ~~"' ~~"0 o -
D.
DN
E.
1
1.7570
0.2705
0.3761
0.1311
1.5703
0.3769
0.5404
0.1292
3
1.1433
0.2697
0.5905
0.t231
4
0.9970
0.2216
0.9464
0.1396
5
0.8881
0.6291
0.7623
0.1385
7
0.7029
0.7631
0.9633
0.1469
8
0.1523
1.3222
0.6842
0.1645
9
0.1343
10 11 12
-0.0380 -0.1967 -1.2308
1.4197 1.5472
0.8579 1.2577 1.5653
0.1616 0.1760
13
-1.3783
1.5874
0.6903
0.1421
reference
1.1732
36.
serial number of the compound refers to Table 8.6.
For the meaning of symbols see text.
~~II 1 ~";ij::; c-
Li.
Log(1/ED,,)
2
From b The
"0,,
I :c
6
0
~-S
g..B B e :;;13
:c U
:c U
I
.z
IN
IN
chemical
Compounds""b
~.§.
]~'O g I;;
~8_
" 0
quantum
~Activityand Quantum Chemical Indices of Some
~0.. ;;.
~< ::10 ';S
N
and the related
Table 8.7
~f . ~0
o
using
of piperidyl nitrogen atom arising from MOs locating (DN); (c) the energy level of the unoccupied 1T orbital, LUMO or the orbital nearest the LUMO and locates on of the I-ethyl group (Ew). Because the variation of the
.
~~1=
"0
indices calculated
indices
are shown in Table 8-7. The results of several regression analyses are
" " 0. 0 .It '€
0" ';S
~:;::E
",A
mechanical
0'0 u _0 o
~<
6,5
electron density dominantly on it which may be the ~the {3 substituent
~(;j II II
~.. "0
'0 ... ~0;
o
~o ~.
,-
°, -
6..{
using quantum
The activity of the compounds
, -5 ;;.5
ce:-
o!::-
was performed
~evaluatethe effect of amide oxygen and nitrogen on the activity.
~-e"E ~EO"
.D ~.5
°
0
~1>
~"
might be made
~substituent in the I-position of the piperidyl ring has only a small effect on the phenylpropanamide linked to the 4-position, it was not possible to
.5tii c.":: '1:;1
". 5i
::E
E
~~;;:'";:
.
o!::-
0
'" "0 c
N.~, :£'f
" ~. " ~~~~~"M ~--:~
.\
"
"c
"0
o!::-
~. ,
or what changes or additions
chemical indices correlate with the activity: (a) the 1T-electron density on the {3substituent of the I-ethyl group arising from the molecular orbital closest to the HOMO and locating on this substituent (Dw); (b) the
..E":': 0
397
(36). The compounds, EDso values, and some of the quantum mechanical indices are reported in Table 8-6. It was found that the following quantum
< "ii ~~~"E
." ,
are outlyers
Studies
the CNDO/2 molecular orbital method as potential activity correlates
~0 ~~:E ~~~"E
~:!-g, . ~~~~B u o~
o!::-
~~~~:':" . M'" ~--:~
analogs
.~ 0_t::
~-.
~~. ~,
0
~C ".0:= o"" 0
~..
and OSAR
definitive in showing what factors should improve or decrease analgesic potency, that is, the QSAR can be tested. A QSAR analysis of 13 compounds involving 3-methylfentanyl and its
""' ~.-
..
Modeling
~~~~~~to equation 1 in order to improve the fit. On the other hand, equation 1 is
~~° u";
0,
Molecular
~~~these compounds
" "if ..
~~§
:g
~§
o!::-
0'" ,
,
.- II ~f OJ -1!
0
...J
.
III
,
0
0'"
,
'"
': o!::-
~;;;
~~~~:!:
00
00
,
0
0
~. N.~~.
+ :> ...J-
'0 U >
0
:Ii
;!. ~.
0_ ::E-
...J
,
,
0,
0,
0
0
~° '"
ii:
;;;
N
0
~;!;
~;;;
j
398
8
Physical Chemistry, Molecular Modeling, QSAR of Arylpiperidine
Analgesics
References
Table 8.8
11. P. S. Portoghese, Z. S. D. Gomas, D. L. Larson, and E. Shefter,J. Med. Chern. 16,199 (1973). 12. A. F. Casy, J. Med. Chern. II, 188 (1968). 13. E. L. Eliel, N. L. Allinger, S. J. Angyal, and G. A. Morrison, "Conformational Analysis." Am. Chern. Soc., Washington, D. C, 1981. 14. A. F. Casy, Tetrahedron 22, 2711 (1966). 15. M. A. Iorio, G. Damia, and A. F. Casy, J. Med. Chern. J6, 592 (1973). 16. K. H. Bell and P. S. Portoghese, J. Med. Chern. 16, 203 (1973). 17. A. F. Casy and K. M. J. McErlane, J. Chern. Soc., Perkin 1 p. 726 (1972). 18. A. F. Casy, F. O. Ogungbamila, and C. Rostron, J. Chern. Soc., Perkin 1 p. 749 (1982). 19. A. F. Casy and K. M. J. McErJane, J. Chern. Soc., Perkin I p. 334 (1972). 20. A. F. Casy, J. E. Coates, and C. Rostron, 1. Pharrn. Pharrnaco/. 28, 106 (1976). 21. A. F. Casy and J. E. Coates, Org. Magn. Reson. 6,441 (1974). 22. A. J. Jones, A. F. Casy, and K. M. J. McErlane, Can. J. Chern. 51, 1782 (1973). 23. A. J. Jones, C. P. Beeman, A. F. Casy, and K. M. J. McErlane, Can. J. Chern. 5t, 1790 (1973). 24. K. H. Bell and P. S. Portoghese, J. Med. Chern. ]6, 589 (1973). 25. A. F. Casy, Pmg. O'"g Res. 22, 149 (1978). 26. M. Froimowitz, J. Med. Chern. 25, 1127 (1982). 27. N. L. Allinger and Y. H. Yuh, Quantum Chern. Progr. Exch. 13, 395 (1980). 28. M. M. Abdel-Monem, D. L. Larson, H. J. Kupterberg, and P. S. Portoghese, J. Med. Chern. IS, 494 (1972). 29. M. Froimowitz and P. Kollman, J. Cornput. Chern. 5, 507 (1984). 30. G. H. Loew and J. R. Jester, J. Med. Chern. 18, 1051 (1975). 31. K. S. A. Razzak and K. A. Hamid, J. Pharrn. Sci. 69, 796 (1980). 32. T. L. Breon, H. Peterson, Jr., and A. N. Paruta, 1. Pharm. Sci. 67, 73 (1978). 33. J. A. Waters, 1. Med. Chern. 20, 1496 (1977). 34. J. A. Waters, J. Med. Chern. 2], 628 (1978). 35. c.- Y. Cheng, E. Brochmann-Hanssen, and J. A. Waters, 1. Med. Chern. 25, 145 (1982). 36. C Changying and L. Lemin, Int. J. Quanturn Chern. 23, 1597 (1983).
Correlation between Activity and Quantum Chemical Indices of Some Compounds (Bivariate Linear Regression) Significance Test Regression Equation Y=a+ beX, + b2X2
Log(I/ED,,)
= 2.1817
- 1.6027D. Log(I/ED,,) - 0.95680. Log(I/ED,,) - 0.5637DN Log(I/ED,,,) -1.0827DN"
= = = -
0.47000N 2.4332 1.2477DN' 6.1810 36.2678£. 5.9953 30.4102£.
Correlation Coefficient
Standard Deviation
F
0.9361
0.4342
0.9761
Critical Value
T,
T,
24.86
F~.)~ = 9.55
5.161
0.948
0.2346
60.59
F~X)J
= 10.9
4.223
3.844
0.7242
0.8515
3.86
Fg_t~ = 3.26
0.430
1.321
0.9690
0.2669
46.10
Fi.~ = 10.9
2.600
3.521
["
a From reference 36. b Compound 13 has been rejected.
shown in Table 8-8. Table 8-8 shows that for the compounds investigated, three quantum chemical indices listed correlate with the activity to a certain extent, among which the correlation between Dn and the activity is most significant. The authors suggest, on the basis of the QSAR analysis, that the {3substituent functions riot only as a hydrophobic group, but also as an electron acceptor to form an acceptor-donor complex with the receptor through a charge transfer process. (
References 1. A. F. Casy, Med. Res. Rev. 2, 167 (1982). 2. P. S. Portoghese, Ace. Chern. Res. 11,21 (1978). 3. H. van Koningsvcld, Reel. Trav. Chim. Pays-Bas 89, 375 (1970); J. V. Tillack, R. C. Seccombe, C. H. L. Kennard, and P. W. T. Oh, Reel. Trav. Chirn. Pays-Bas 93, 164 (1974). 4. F. R. Ahmed, W. H. Barnes, and G. Kartha, Chem. Ind. (London) p. 485 (1959); G. Kartha, F. R. Ahmed, and W. H. Barnes, Acta Crystal/ogr. 13, 525 (1960). 5. F. R. Ahmed, W. H. Barnes, and L. A. Masironi, Chem. Ind. (London) p. 97 (1962); F. R. Ahmed, W. H. Barnes, and L. D. Masironi, Acta Crystal/ogr. 16,237 (1963); F. R. Ahmed and W. H. Barnes, Acta Crystal/ogr. 16, 1249 (1963). 6. P. S. Portoghese and E. Shetter, J. Med. Chern. 19, 55 (1976). 7. D. S. Fries, R. P. Dodge, H. Hope, and P. S. Portoghese, J. Med. Chern. 25, 9 (1982). 8. M. Cygler and F. R. Ahmed, Acta Crystal/ogr., Sect. B B40, 436 (1984). 9. W. H. De Camp and F. R. Ahmed, Chern. Cornrnun. p. 1102 (1971). 10. K. Hayakawa and M. N. G. James, Can. J. Chern. 51, 1535 (1973).
399
/
Methadone
9.
Compounds
401
-~
NaNH2
~CN
Open-Chain
Analgesics
Compounds . . A. Modification of the Alkylamine Chain B. Modification of the Ketone Chain . . . C. Modification of the Diphcnyl Fragment. 11. Other Open-Chain Compounds ..... A. Thiamoutene and Related Compounds. R. Benzimidazok-Based Compounds. . C. Cyclohcxylamine-Bascd Compounds. D. Bezy!amine-Based Compounds . . . . E. Miscellaneous Open-Chain Compounds
400 406 41\ 425 135 436 436 438 441 444 445
References
Methadone and Related Compounds
Methadone (I), the first of the open-chain analgesics, was discovered at I.G. Farbenindustrie at Hoechst-am-Main in Germany during World War II in the course of work on spasmolytic compounds (1,2). r'
'"
I
0
+
YNMe'i CI
EtMgBr
\~
I. Methadone and Related
1.
and Related
'\
~
I
0
'" 2
Its narcotic-type analgesic activity was unexpected, since it lacked any obvious resemblance to previously known compounds. Despite the fact that there was a morphine shortage at the time of its discovery, methadone was not used as an analgesic until after the war. Its analgesic potency is about twice that of morphine, and it was apparently tested at excessive doses, leading to adverse side effects (2). Much of the early literature uses the name amidon or amidone. This name was soon abandoned in favor of methadone in order to avoid confusion with other products (2). Methadone exhibits all of the usual narcotic-type activities: analgesia, sedation, respiratory depression, constipation, physical dependence, antagonism by nalorphine and related compounds. It is distinguished by its good oral activity and long duration of action. Despite its general similarity 400
I"
+ , Scheme
9-1.
Synthesis
2
of methadone
and isomethadone
(reference
6).
to morphine,
a number of differences have been observed. Physical dependence and withdrawal symptoms are reported to be slower in onset and less intense with methadone than with other narcotic analgesics (3). Methadone exhibits less depressant activity than morphine (2) and produces less euphoria (4). It also has a local anesthetic effect (5) and is reported to be weakly antihistaminic (2). Interestingly, all efforts to produce open-chain narcotic antagonists have been unsuccessful. Methadone is commonly used orally in outpatient maintenance therapy for narcotic-dependent individuals (3) because it has relatively good oral activity and a long duration of action. Its lower level of euphoria and sedation, slow development of tolerance, and slow development of withdrawal symptoms make it particularly suitable for such use. The original synthesis of methadone (6), shown in Scheme 9-1, led to a mixture of four diastereomers, dl-I (methadone) and dl-2 (isomethadone), in approximately equal amounts. Schultz et al. (7) explained the apparent rearrangement on the basis of the postulated aziridinium intermediate 3, which could be opened by attack at either of the ring carbon atoms. The synthesis reported by Bockmuhl and Ehrhart (1), Scheme 9-2, also led to a mixture of methadone and isomethadone.
3
Easton and co-workers (8) reported a synthesis of methadone in which isomethadone was not produced. As shown in Scheme 9-3, diphenylacetonitrile was condensed with propylene oxide to form a furanone-imine. This was opened with phosphorus tribromide to form the alkyl bromide, which
402
9
g;' '"
CN
+
h
B,.
Open-Chain-
Analgesics
g;
8,..
'"
+
I~""
and Related Compounds
Methadone
CN
+
I~""
Be
CI'"'y0E' DE'
1)
403
1)
NaNH2
~ 2) H+,H20
CHO
MeMg I
~ 21 SOC/2
c
EtMgSr ~
Scheme
9-4.
Methadone
synthesis
of Morrison
and Rinderknecht
(reference
9).
+ 2 Scheme
9-2.
Synthesis
of methadone
and isomethadone
(reference
1).
,y
I
CN
f} I"" ~
+
f} '"
1
CN
+
"" I~
-
~
EtMgBr
o
2 Scheme
1 Scheme
9-3.
Methadone
synthesis
of
Easton
and
co-workers
(reference
8).
was subsequently reacted with dimethylamine to form the usual nitrile precursor. Yields in the final step were reported to be less than 10% for both the bromide and the chloride. Another ~ethadone synthesis in which isomethadone contamination was avoided was reported by Morrison and Rinderknecht (9). In this approach (Scheme 9-4), diphenylacetonitrile was condensed with chloroacetaldehyde diethyl acetal. The deprotected aldehyde was converted to the methyl carbinol with methyl Grignard, and then to the chloride. As in the Easton synthesis, the reaction of the chloride with dimethylamine produced a very low yield.
9-5.
Isomethadone
synthesis
(reference
/0).
Isomethadone, which is somewhat less active as an analgesic than methadone, was also synthesized by methods that preclude contamination with methadone. In one case (10), the postulated aziridinium intermediate was avoided by formylating the amine,as shown in Scheme 9-5. The formyl protecting group was subsequently reduced to a methyl group, and the nitrile was converted to the ketone in the usual way with ethyl Grignard. Alternatively, these authors took advantage of the fact that the secondary tosylate is more reactive than the primary chloride (Scheme 9-6) in selectively preparing the primary chloride. In contrast to the methadone syntheses described above, this compound reacted with dimethyl amine in 58% yield. Both methadone and isomethadone contain an asymmetric center. As might be expected, the stereoisomers have different levels of analgesic activity. Thorp and co-workers (11) resolved the nitrile precursor and prepared (+)- and (- )-methadone. Larsen el al. (12) subsequently resolved (+)- and (- )-methadone using d-tartaric acid. Thorp (13) found the
levorotatory
(- )-isomer to be about twice as active as morphine
by
404
9
I
;} '" ~
Open-ChaIn
Analgesics
Methadone
and Related
CN
+
'-("CI
o-a'.
NHMez
OTs
Compounds
__
I
li A I H..
I HO~C
---
405
NH-CHO
"-
HO'-"""
NHCH,
+
I",
diphenyl_
-
SOCIZ ~
CI~
EtMgBr
,CHO N I CH,
ace
ton
i t r i Ie )
HCOOH
---4
EtMgBr 2 Scheme
9.6.
Isomethadone
synthesis
(reference
10).
subcutaneous administration in rats, and he reported the dextrorotatory (+ )-isomer to be inactive. Other workers (14) found analgesic activity in the (+ )-isomer, although the (- )-isomer was reported to be 7-50 times as potent. At least two laboratories (13,14) reported studies in which the racemic mixture was more toxic than either isomer alone. For instance, Scott ef al. determined LD50 values of 21,29, and 31 mg/kg in mice for the (:t)-, (-)-, and (+)-isomers, respectively. Furthermore, they observed that the (- )-isomer produced death in mice over a relatively long period of time (up to 8 hours), while the (+ )-isomer resulted in death very rapidly or not at all, suggesting that toxicity may Occur by different mechanisms for the two isomers. This may account for the apparent synergistic toxicity observed for the (:t) mixture. Differences in activity between the (+)- and (- )-isomers could, in principle, be due to differences in dish-ibution, metabolism, and/or excretion between the two isomers. However, Sung and Way (15) showed that the two isomers have very similar tissue distribution. Furthermore, they showed that the two isomers have almost identical rates of metabolism and pH optima for metabolism in liver slices. Sullivan and co-workers also found little or no difference between the two isomers when comparing metabolic patterns and excretion. The differential analgesic activity between the isomers was ascribed to differences in receptor binding for the stereo isomers (16). This suggestion was substantiated when receptor binding assays became available. The more active (- )-isomer of methadone exhibited receptor affinities of 4 x 10-9 to 2 x 10-8 M, while the (+ )-isomer had values of 1 x 10-7 to 3 X 10-7 M (17-19). Isomethadone also exhibits different levels of analgesia in the two stereoisomers. The levorotatory (- )-isomer was shown to be the more active form, being about 40 times as active as the (+ )-isomer on subcutaneous administration in mice (20,21).
Scheme
9.7.
Synthesis
of (-
)-methadone
(-1-, from D-( - )-alanine (reference 23).
The absolute stereochemistry of the more active (- )-methadone was shown to be R in 1955 by Beckett and Casy (22). They converted both the nitrile precursor of (- )-methadone and D-(- )-alanine to the dextrorotatory amine 4. Beckett and Harper (23) later carried out the synthesis of (- )-methadone using D-(- )-alanine as a starting material, as shown in Scheme 9-7. The analgetically more potent (- )-isomethadone was shown by Beckett ef al. (24) to have the S absolute configuration by relating it to (- )-a-methyl-i3-alanine.
4
At least eight metabolites of methadone have been detected in humans. The major metabolic pathway is apparently N-demethylation, although the N-demethylated compound is not isolated as such. It is reportedly an unstable compound that rapidly cyclizes (25). Accordingly, the first metabolite identified was the cyclic compound 5 (26). Sullivan and co-workers later identified several other cyclic N-demethyl and N,N-bisdemethyl compounds, to which they assigned the structures 6 (27), 7 (28), 8 (28), and 9 (25). In addition, they isolated the carboxylic acid 10 (28), the phenyl ring-hydroxylated compound 11 (25), and the a1cohol12 (25). Of these, only compound 12 exhibited analgesic activity of its own. Sullivan and Due (25) indicated that the methadone N-oxide that Beckett ef al. isolated from urine (29) is a storage artifact rather than a metabolite.
406
9
Open-Chain
Analgesics
Methadone
and Related
Compounds
407
Table 9.1
'"
Structure-Activity Relationships Methadone Analogs with Variation of the Amino Groupo
~5
. OH
c
OH
R
.
7
Activityb
-NMe2 (methadone) .Pyrrolidyl
9
5-10 5-10 7 10
.Piperidyl -Morpholinyl
HO
10
Q
Data from reference 1. b Relative to meperidine
11
12
The structure-activity relationships derived by modification of the methadone structure will be considered in four sections. The first will deal with changes, in the alkylamine chain, the second with changes in the ketone fragment, and the third with the diaryl portion of the molecule. The fourth will cover phosphorus and sulfur analogs of methadone. A.
in
Modification of the Alkylamine Chain
/
Removal of the amine nitrogen completely abolishes analgesic activity (2). Quatemization of the amine decreases activity substantially. The quatemized methadone analog 13 and the isomethadone analog 14 reportedly showed about 1I80th of the activity of methadone ';:hen administered subcutaneously to mice.
= 1.
Dupre el al. (30) carried out a more extensive exploration of the N-substituents in the nor~ethadone series (Table 9-2). Among the N,Ndiaikyl compounds, the dimethyl compound (normethadone, IS) had the highest activity. As the alkyl groups increased in size, analgesic activity decreased rapidly; the di-n-propyl and di-n-butyl compounds showed little or no activity, and the benzyl and dibenzyl derivatives were completely inactive. In compounds having cyclic amines, the pyrrolidine, piperidine, and morpholine derivatives all showed higher analgesic activity than the dimethyl compound, with the best activity in the morpholine derivative. Methyl substituents on the piperidine or morpho line ring decreased activity substantially, except in the case of 4-methyJpiperidine. Compounds 16-19, having 7- or 8-membered ring amine substituents, showed little or no analgesic activity (31).
15
17
13
Numerous variations of the dimethylamino group have been tried. Bockmuhl and Ehrhart (1) found the highest activity in methadone analogs having pyrrolidine, piperidine, or morpholine substitutents (Table 9-1).
19
19
408
9
Open-Chain
Analgesics
Methadone
and Related
Compounds
409
Table 9-2 Table 9-3 Structure-Activity Relationships in Normcthadone Analogs with Variation of the Amine Substituents a
Relative Analgesic Activities Methadone Isomers, Isomethadone Isomers, and Normethadone"
Compound
R -NMe,(15) -NEI2 -N(nPr), -N(nSu), -N(Me)ben,yl -N(benzyl), Pyrrolidine Piperidine Morpholine 2-Methylpiperidine 3-Methylpiperidine 4-Methylpiperidine 2,6-Dirnethylpiperidinc 3-MethylmorphoJine 3,5-Dimethylmorpholine
J 0.7-1 0-0.3 0-0.3
1.6 0.8 26 2.5 1.2 50 16.6
a
Data from reference 32. bED50 in milligrams per kilogram, subcutaneous administra-
o o 4 2-3 7 o 0-0.3 3 0-0.3 1.5 1.5
" Data from reference 30. b Relative to meperidine = 1.
EDs~/
(ot)-Methadone (I) (- )-Methadone (+ )-Methadone (ot)-Isomethadone (2) (- }-Isomethadone (+ )-Isomethadone Normethadone (IS)
Activityb
of
tion in mice. Under these conditions, morphine had an ED50 of
3.t. mg(kg.
/
Changes in the alkyl chain connecting the amine to the~ diaryl ketone have also been explored. The analgesic activities of (-)- and (+)_ methadone (1) and of (-)- and (+ )-isomet.hadone (2) were described previously. In both pairs, the levorotatory-( - i-isomer was substantially more active than the dextrorotatory isomer (13,/4,20,21). In the case of methadone, the more active isomer has the R absolute configuration (22), while the more active isomethadone isomer has the S configuration (24). Normethadone (15) was found to be substantially less active than either (:t)-methadone or (:t)-isomethadone (32), as shown in Table 9-3. Bockmuhl and Ehrhart (1) report that replacing the a-methyl group in methadone with an a-ethyl substituent (20) produced an inactive compound. Lengthening the alkyl chain by one or two atoms, as in compounds 21-24, also produced compounds with little or no analgesic activity (1,30,32).
zo
21
N"\ /
22
0
23
Henkel et al. (33) investigated the stereoisomers of 5-methylmethadone (25). Attempts to utilize the Easton synthesis (Scheme 9-3, reference 8) with cis- and trans-2-butene-oxide were unsuccessful in this system, leading
25
410
9
Open-Chain
Analgesics
I
o..",<.fh>,ta...,:~..
o
R~..
9H
/ \..,H
NHMe2
R~CH)
(.is'
+
Or- +"'hS-~-",<+e/le-o""
~-f-R
1 J BULi
CH'-f-H NMe,
2) TsCI
---7 \
.
dlphenyl_ acetonitri le/
,
.0,CCN
c
JC
CH)-f-H ~ C-CONH,
NMe',
" CH,-C-H
diphenyl_
, I,
CH-C-H
acetonitrile 1 base
NMe, n-bu.tyl "
n i tf
~,C-Co,R ,
i te
H-f-CH)
2'
CH2N2
1
,, N"',
CH-C-H
0,C-CO.R , ,, CH-C-H
Etli
CHT~-H NMe,
1
l..:t I-threo
Scheme 9-8.
Synthesis of erYlhro- and threo-5-methylmethadone
(reference 33).
either to an olefin or a despropionyl compound. The synthesis employed by these workers, shown in Scheme 9-8, introduced the troublesome dimethylamino group at an early stage by opening cis- or trans-2-butene-oxide with dimethylamine. The resulting alcohol was then converted to the tosylate. In the case of the threo isomer, the tosylate readily formed an aziridinium intermediate that could be reacted in situ with diphenylacetic ester anion to form the ester shown, which was readily transformed to the desired ketone with ethyllithium. The erythro tosylate proved to be much less reactive, necessitating a modification of the synthesis. In this case, the tosylate was reacted with diphenylacetonitrile anion to form the nitrile, which was converted to the ketone in four steps. The threo racemate contains the
and Related Compounds
411
5S,6R stereoisomer, which combines the absolute configurations of the most active isomers of both methadone and isomethadone. Interestingly, this racemate showed no agonist or antagonist activity, indicating that the chiral centers do not behave independently. In contrast, the erythro racemate was more than five times as potent as methadone on subcutaneous administration. B.
CH,-cr-H
Methadone
Modification of the Ketone Chain
The ketone functionality of methadone is relatively unreactive, failing to form a semicarbazone or to reduce with aluminum isopropoxide, sodium amalgam, or Raney nickel/hydrogen (34). Isomethadone is even more resistant to catalytic hydrogenation (35). Nevertheless, one of the first modifications made to the methadone structure was reduction of the ketone to an alcohol. I. Methadols and Acylmethadols Reduction of the ketone in methadone can, in principle, give rise to two pairs of diastereomers. Hydrogenation of racemic methadone over platinum oxide catalyst gave rise to a single dl pair (34). Lithium aluminum hydride reduction of methadone produced the same dl pair (35), which was named a-methadol (36). Pohland and co-workers (37) showed that hydrogenation of (+)_ methadone produced (- )-a-methadol, and (- )-methadone produced (+)a-methadol. May and Mosettig (38) later showed that the {3-isomers were the major product in the sodium/isopropanol reduction of methadone; lesser amounts of the a-isomer were formed but could be removed by fractional crystallization. In the case of the {3-methadols, the sign of rotation is the same as that of the parent ketone (38). Reduction to the carbinol substantially decreased the analgesic activity in both the a and {3 pairs, but it was found that analgesic potency could be restored by acetylating the alcohols (20,38). Table 9-4 lists analgesic activity data for racemic and resolved methadone, a- and {3-methadols, and a- and {3acetylmethadols (20). It can be seen that the {3-methadols are generally more active than the a-methadols and that the acetyl derivatives are more active than the parent ketones. It is interesting to note that the most active methadols (a-I and (3-d) are derived from the least active ketone (dmethadone)
.
Isomethadone (2) fails to react with hydrogen-platinum oxide (35,36); however, it can be reduced with lithium aluminum hydride to produce a single dl pair, which is referred to as a-isomethadol (39). As was found for methadone, the (3-isomer was the major product of the sodium-isopropanol reduction of the ketone (39). In the isomethadone series, (+)isomethadone is reduced to ( + )-a-isomethadol and to (- )-{3-isomethadol.
412
9
Open-Chain
Analgesics
Table 9-5 Analgesic Activities of Isomethadone, Isomethadol, and Acetylisomethadol
Table 9-4 Analgesic Activities and Acetylmethadol
of Methadone, Methadol, Stereoisomers'"
Compound
ED",
(::!:)-Methadone
b
1.6
9.2
0.8 25.7
8.0 89.3
a-( )-Methadol a-( '" + )-MethadoI a-( - )-MethadoI
18.9 24.7 3.5
10.9 61.8 3.8
/3-( )-MethadoI '" /3-(- )-Methadol /3-( + )-Methadol
7.3 7.6 63.7
67.3 36.7 70.0
1.2 0.3 1.8
4.0 1.6 1.1
/3-( - )-AcetylmethadoI /3-( + )-AcetylmethadoI
0.8 0.4 4.1
2.6 2.0 5.1
Morphine
2.3
3.7
a-(::!:
)-Acetylmethadol
(3-(:!: )-Acetylmethadol
"Data from reference 200
b ED5Q in milligrams per kilogram cutaneous administration in mice. C
EDso'in milligrams per kilogram administration in mice.
(:!: )-Isomethadone
2.5 1.2 49.8 66.8 60.7 91.7
(- )-Isomethadone
(- )-Methadone (+ )-Methadone
a-( + )-Acetylmethadol a-( - )-Acetylmethadol
ED50l>
Compound
EDsoc
~
C
( + ).Isomethadone a-(:!: )-IsomethadoJ
a-( + Hsomethadol a-( - ).Isomethadol 13-(
)-Isomethadol
12.3
/3-(-'" )-Isome'hadoI /3-(+ )-Isome'hadoI
58.7 6.2 4.8 2.7 62.7 17.4 10.9 70.6
a-(:!: )-Acetylisomethadol
a-( + )-Acetylisomethadol a-( - )-Acetylisomethadol (3-(::!: )-Acetylisomethadol /3-(- )-Acetylisomethadol 13-(+ )-Acetylisomethadol " Data from reference b
EDso
i~ milligrams
administration
Stereoisomers'"
132 30.3 93.9 40.7 11.2 10.4 104 55.4 35.0 164
for subcutaneous
in mice.
C
EDso in milligrams per kilogram for oral adminis. tration in mice. d Inactive at nonlethal doses.
Table 9.6 Analgesic
Activities
of Some
Secondary
Amine
and Tertiary
Amine
However, acetylation reverses the sign of rotation for both the a- and i3-isomers. Leimbach and Eddy (20) measured analgesic activity subcutaneously and orally in mice for racemic and resolved isomethadone, aand i3-isomethadols, and a- and i3-acetylisomethadols, as shown in Table 9-5. i3-Isomethadol was found to be less active than isomethadone, and a-isomethadol was almost completely inactive. Acetylation of the isomethadols restored much of the anagesic activity, and on oral administration some of the acetates had better activity than the parent ketone, but none of these compounds was as active as the corresponding methadone dedvatives. The methadols and acetylmethadols are among the few compounds in which secondary amines show good analgesic activity (40a). In fact, for the examples shown in Table 9-6, the secondary amines are more active than the corresponding tertiary amines. a-Acetylmethadol (26) has been used
23.0 24.4 d d d
20.
per kilogram
for sub. for oral
EDsoc
R H H Ac Ac
Derivatives
R' Me H Me(U) H(27)
of a-Methadol"
ED5Qb 16.9 0.98 1.09 0.48
" Data from reference 40b. b EDso in milligrams per kilogram on subcutaneous administration in mice.
414
9
Open-Chain
Analgesics
Methadone
Table 9-'
Compound (+ )-Methadooe (- )-Methadone a.( - ).Methadol a.( + ).Methadol a-( - )-N-Normethadol a-( + )-N-Normcthadol a-( - )-N,N-Dinormethadol a-( + )-Acetylmethadol a-( - )-Acetylmethadol a-( + )-N-Noracetylmethadol a-( - )-N-Noracetylmethadol
- )-N,N-Dinoracetylmethadol a Data from reference
Ester Analogs of Methadonea
and
IC", eM) Lh 1'\ 1.3 X 10-7 130
I
5.0 x 1O~9 5.0 5.1 x 10-' S 10 2.0 X 10-7 ~O 8.0 X 10-10 O.g 6.3 x 10-' ~30 1.5 x 10-' ISO 4.3 X 10-9 "(..3 3.5 x 10-' 35 3.6 x 10-' :n. 1.0 x 10-' 1.0 1.6 x to-9 I,b
c R .Me .Et -iPf .Et .Et a
Activity (meperidine
R' -NMe2 -NMe2 -NMe2 -morpholinyl -piperidyl
= 1)
Activity (ED""mgjkg) 9.9 18 61
<0.1 0.2
Data from references
J and 38.
19.
clinically as an analgesic (40a) and in maintenance therapy for narcoticdependent persons (41). Its properties are generally similar to those of methadone, with better oral activity and a longer duration of action. The major route of metabolism for this compound is N-demethylation to the N-nor and N,N-dinor compounds 27 and 28 (42). Nickander et al. (43) have showed that in humans 27 and 28 reach plasma levels at which these metabolites may be responsible for part of the analgesic activity and much of the long duration of action. The dinor compound 28 has an EDso about equal to that of the parent compound (43).
26
415
Table 9-8
Receptor Binding of Stereoisomers of Methadone, a-Methadol, a-Acetylmcthadol, Some N-Demethylated Compoundsa
a-(
and Related Compounds
27
2. Esters Replacement of the ketone side chain by an ester is readily accomplished from the nitrile precursor, either by sulfuric acid hydrolysis and esterification (1) or by heating with the alcohol, sulfuric acid, and ammonium chloride in a scaled tube (30). Table 9-8 lists analgesic activities for a number of ester analogs of methadone. As the ester was varied from methyl to ethyl to isopropyl, activity decreased in this series. Eddy et al. (38) found that the ethyl ester analogs of methadone (29) and isomethadone (30) have essentially the same analgesic activity. Compound 29 had an EDso of 18 mg/kg and 30 had an EDso of 19 mg/kg. A further summary of structure-activity relationships in the esters is found in Table 9.9 for compounds in the normethadone series. Here, too, the methyl ester was most active, with activity decreasing as the ester group became larger.
28
Horng and co-workers (19) measured receptor binding for the stereoisomers of methadone, a-methadol, a-acetylmethadol, and some of the N-demethylated compounds. Their results are shown in Table 9-7. The receptor affinities measured are in good accord with the measured analgesic activities (cf. Table 9-4). The N-demethylated compounds that showed good analgesic activity (Table 9-6) also have relatively high receptor affinities.
29
30
3. Ketones Bockmuhl and Ehrhart (1) explored variations of the ketone side chain, as shown in Table 9-10. The highest activity was found in the ethyl ketones. The aldehyde, methyl ketone, n-propyl ketone, and isopropyl ketone showed substantially less analgesic activity. The allyl and isobutyl ketones again showed reasonable analgesia, but phenyl and benzyl ketones were very weak analgesics.
416
9
Table 9-9
Ketone
o -Me -Et .iPe -nBu -phenyl -benzyl
2 0.5 I <0.2 <0.2 0.5
Subcutaneously,
relative to idine = 1.
meper-
and Related Table
Modifications"
Compounds
417
9-11
Amide
Analogs
of Normethadone"
c R
R'
Activityb
-piperidyl -NMcz -morpholinyl -NMez -morpholinyl -NMez
3 o 0.5-0.75 0.5 5 3 1 0.25-0.5 0.25 2-3 2 <0.2 <0.5 o
R
Activityb
" Data from reference 1. b
Methadone
Table 9-10
Ester Analogs of Normethadone"
R
Open-Chain'Analgesics
-H -H -Me -Me -Et -Et -nPc -nPc -iPe -allyl -iBu -phenyl -phenyl -benzyl a
b
-morpholioyl
-NMcz -piperidyl -piperidyl -piperidyl -morpholioyl
-NMcz -morpholinyl
Data from references Subcutaneously,
relative
J and 2. to meper-
idine = 1.
4. Amides Only a few simple ami des were reported by Bockmuhl and Ehrhart (1), and these showed little or no analgesic activity. For instance, the primary amide 31 was completely inactive, and compound 32 was much less active than meperidine. Janssen and Jageneau (44) have reported an extensive study of amide analogs of normethadone, methadone, and isomethadone; their results are summarized in Tables 9-11, 9-12, and 9-13. Amide analogs of normethadone are shown in Table 9-11. Most of these compounds had very weak analgesic activity. Primary and secondary amides were completely inactive, and the best activily was seen when the dimethylamide or pyrrolidine amide was employed. None of these compounds approached the activity of methadone. Table 9-12 lists several amide analogs of methadone. Unlike the ketones, this series exhibited no significant increase on addition of the methyl substituent exto the amine nitrogen. The most active analgesic in this series was an N,N-dimethylamide, which was still much less active than methadone.
ED", (mgfkg)
R'
-NH, -NHMe -NMez -NMe2
-morpholinyl
-pyrrolidyl
-piperidyl
-piperidyl -piperidyl -piperidyl
-pyrrolidyl
-morpholinyl
-piperidyl -piperidyl
-piperidyl
Inactive >100 >100 22.3 70.0 13.6 >100 73.0 >100 95.0 5.2
-morpholinyl
-morpholinyl
-piperidyl
-morpholinyl
-morpholinyl
Methadone
I I
I
" Data
from
references
1 and 44.
Table 9-]2 Amide Analogs of Methadone"
, R -NHMe -NHEt -NMe2 -pyrrolidyl -pyrrolidyl -pyrrolidyl -NEt2 Methadone
R'
ED", (mgfkg)
-morpholinyl -morpholinyl -morpholinyl -NMe2 -pyrrolidyl -piperidyl -piperidyl
" Data from reference
44.
53.9 62.0 24.5 >100 > 50 > 50 > 25 5.2
4]8 9
Open-Chain
Analgesics
I
Methadone
and Related
Compounds 419
Table 9.13 Amide
Analogs
of lsomethadoneu
(oic1x.
base ) C I-CH2-CH2-NR2
Scheme 9.9. Synthesis of phosphorus X = -phenyl, -ethyl, -ethoxy.
c R -NIIMe -NIIMe -NilE' -NIIEt -NIIEt -NHiPr -NHnBu -NlltBu -NH-benzyl
-NMe2 -NMe2 -NMe;! -NIIMeEt -NEt;! -NEt2 -pyrrolidyl ~-pyrrolidyl
R' -NMc;! -morpholinyl
-NMe;! -piperidyl -morpholinyl -morpholinyl -morpholinyl -morpholinyl -morpholinyl
-NMe2 -piperidyl -morpholinyl -morpholinyl -piperidyl -morpholinyl
-NMe2 -pyrrolidyl
-pyrrolidyl
-piperidyl
-pyrrolidyl
-piperidyl
-pyrrolidyl
-morphoJinyl
(+ )-isomcr
-pyrrolidyl
-morpholinyl
(+ )-isomer
-pyrrolidyl
-morpholinyl
(- )-isomcr
-piperidyl
-morpholinyl
-morpholinyl
-morpholinyl
Methadone
U
Data from reference
31
ED", (mg/kg) ]46 44.1 145 54.0 25.9 82.0 >100 >100 >100 21.0 11.4 1.38 26.0 >50 32.1 16.3 20.9 13.2 7.80 1.25 0.645 >150 59.0 70.0 5.2
44.
32 33
Janssen and Jageneau found Ihe highest analgesic activities in amide analogs of isomethadone, as shown in Table 9-13. Among Ihe monoalkyl amide analogs, the elhyl amide showed the best analgesic potency (26 mg/kg); smaller or larger alkyl groups decreased activity. In Ihe dialkyl
I
analogs
of methadone
(reference
46). n = 1, 2;
amide series, the order of potency observed was dimethyl> methylethyl > diethyl. The pyrrolidine amide series exhibited the best analgesic activities, with one compound showing greater analgesic activity than melhadone. The dextrorolatory isomer of this compound is called dextromoramide, 33. This compound had an ED50 of 0.645 mg/kg, in sharp contrast to the levorotatory isomer, which had an ED50 > 150 mg/kg. Changing the pyrrolidine amide tOlhe piperidine amide decreased the analgesic potency by a factor of about 40. In clinical sludies (45), dextromoramide was found 10 be aboul twice as potenl as morphine. Occurrence of side effects such as respiratory depression, sedation, and nausea were about equal for dextromoramide and morphine. Dextromoramide-induced analgesia had a shorter duration of action than morphine-induced analgesia. 5. Phosphorus and Sulfur Analogs Shelver and co-workers (46) prepared a series of monoaryl and diaryl phosphorus analogs of melhadone. These compounds were prepared as shown in Scheme 9-9, by reacting Ihe anion of the arylphosphorus starting material with the appropriate chloroalkylamine. The yield in the alkylation reaction was found 10 be highly dependent on the solvent and conditions. The analgesic activities measured for the monoaryl series are shown in Table 9-14. None of these compounds approached Ihe level of activity shown by morphine. The diphenylphosphine oxides were found to be substantially more active than the diethyl phosphonates and the single dielhylphosphine oxide tested. Surprisingly, Ihe most active compound found in this series was 34, a diethyl amine. This compound was still only about one-fifth as active as morphine. In the only instance where a methyl substituent was tried on the alkylamine chain, Ihe isomethadone analog 35 was about four limes as aclive as the desmelhyl compound.
;, Q " O-p-o u0u '" N----., Ib ..
Q
O~p-o
dYNMe. ~Ib 3.
36
The diaryl phosphorus analogs prepared by Ihese workers are listed in Table 9-15. In this series, no dielhylphosphine oxides and only one
Table 9-14
Methadone
Monoarylphosphorus
Analogs
0I A
0
' CH
-phenyl -phenyl -phenyl -phenyl -phenyl -phenyl -OEt -OEt -OEt -OEt -Et Morphine
grl
,R
",P'
:::: -..::::
R
I R'
R
-CHrCH2-NMe2 -CH2-CHrNEt2 -CHrCHrpyrrolidyl -CH2-CHrpiperidyl -CH2-CHrmorpholinyl -CH(CII,J-CH,-NMe, -CHrCHrNMe2 -CHrCHrNEt2 -CHrCHrpiperidyl -CHrCH2-morpholinyl -CHrCHrpiperidyl
CI
thio ~
re.
RX
l
gr :::: ~
SH
+
(35)
Analogs
I I r
I
-CH2-CHrCH2-NMe2 -CH,-CH,-p;peridyl
"b Data from reference 46. EDso. milligrams per kilogram.
(36)
10.2 15.2 13.3 11.2 17.2 14.6 22.1 11.1 1.3
subcutaneously
SR
'"
I
SOa.R
Scheme
of Methadone"
-CHrCH2-NMe2 -CHrCHrNEt2 -CH2-CHrPyrrolidyl -CHrClI2-piperidyl -CH2-CHrmorpholinyl -CH(CH,)-CH,-NMe,
gr
b.se
subcutaneously
R'
"'l
I""
I",
D
85.1 6.2 47.9 24.7 10.1 19.7 248.7 144.2 105.0 104.4 191 1.3
I
Diaryl Phosphorus
in mice.
421
R'
Table 9-15
-phenyl -phenyl -phenyl -phenyl -phenyl -phenyl -phenyl -OEt Morphine
Compounds
I",
Data from reference 46. b" ED50' milligrams per kilogram, in mice.
R
and Related
of Methadone"
I I
9-10.
Synthesis
of sulfone
analogs
of methadone
(reference
47).
diethylphosphonate were prepared. The latter compound, 36, was the most active diaryl derivative tested, but was still only one-eighth as potent as morphine. Among the diphenylphosphine oxides, variations in the alkylamine side chain seemed to make very little difference. Even the insertion of an extra methylene group in the alkylamine chain caused only a small decrease in analgesic activity. Klenk et al. (47) prepared a series of methadone analogs in which the ketone side chain was replaced by an alkylsulfone or arylsulfone. Scheme 9-10 outlines the general synthetic approach to these compounds. Benzohydryl chloride was first converted to the mercaptan by treatment with thiourea. The mercaptan was treated with the appropriate alkyl or aryl halide and then oxidized to the sulfone with hydrogen peroxide. Finally, treatment with base and a chloroalkylamine produced the target compounds. The structure-activity relationships observed by these workers are summarized in Table 9-16. In general, the alkyl substituent on the sulfone had to be methyl or ethyl in order to obtain good analgesic activity; propyl and p-toluyl sulfones were much less active. The most active compounds were those having a methyl substituent a to the amine nitrogen (as in methadone), although this point was not explored extensively. Tullar and co-workers (48) later resolved the stereoisomers of one of the sui fones, compound 37, by making the d-bitartrate salts. The levorotatory form was found to be about 20 times more active as an analgesic than the dextrorotatory form.
I
r r
37
6. Imines Cheney et al. (49) prepared a series of imine and acylimine derivatives of methadone-type ketones. These imines were readily prepared from the nitrile precursor and were found to be quite stable,
412
9
Open-Chain'
Analgesics
Methadone
Table 9-16 Analgesic
and Related
Table 9-17 Activities
of Sulfone
Analogs
of Methadone" Imine and Acyliminc
:
gx '"
R
a
h
R' ['I
Activityb
-CHrCHrpiperidyl -CH(CH,)-CH,-NMe, -CHz-CHrNMc2 -CHz-CHz-piperidyl -CH2-CHz-NEtz
-CH,-CH(CH,)-NMe, -CH,-CH(CII,)-piperidyl -CH,-CII(CH,)-NMe, -CH,-CII(CH,)-NMe, -CHrCHz-piperidyl
++ ++ ++ ++ ++ +++ +++ +
o o
Data from reference 47. equal to methadone; +++ = approximately equal to meperidine; + + = approximately = approximately
of Methadone"
I
R'
+
Analogs
SO,R
"-
-Me -Me -Et -Et -Et -Et -Et -nPr -p-loluyl -p-loluyl
423
Compounds
equal
to aminopyrine.
R -H
-acetyl -propionyl -prepianyl
-CHrCllr morpholinyl -CII,-CH(CH,)-NMe, -CHrCHrpiperidyl -CHz-CHz-pyrrolidyl
-acetyl -acetyl Methadone
Data from reference 49. '" b Subcutaneous analgesic dose
39
per kilogram)
pigs.
Table 9.18 Imine and Acetylimine
R -H -H -acetyl -acetyl Morphine
Derivatives
of Methadone'"
R' -CH(CH,)-CII,-NMe, (38) -CH(CII,)-ClI,-morpholinyl -CII(CII,)-CH,-NMe, (40) -CH,-CH(CH,)-NMe, (4!)
"b Data from reference 38. ED5o, milligrams per kilogram.
3'
(milligrams
Dosch
15 12.5 15 12.5 12.5 30 t2.5 25 50 75 t2.5
-CHrCHz- morpholinyl -CHrCHz-piperidyl -CII(CII,)-C!!,-NMe, -CII,-CH(CH,)-NMe, -CII,.CH(CH,)-NMe, -CII(CH,)-CH,-NMe,
-H -H -H -acctyl
in guinea
presumably due to steric hindrance. Acylimines were prepared by treating the imines with acetyl or propionyl chloride. Many of these compounds, listed in Table 9-17, showed analgesic activity equivalent to that of methadone; none was found to be more potent. Overall, the acyl derivatives showed somewhat less analgesic potency than the parent imines. The propionyl derivatives were only slightly less active than the acetyl analogs. Eddy and co-workers (38) also examined the analgesic activities of a series of imine derivatives (Table 9-18). Compound 38, the imine of isomethadone, showed about one-fifth the analgesic potency of morphine. The most active imine dierivative examined by these workers was 39, which was equal to or slightly better than morphine in potency. The acetylimine derivatives of isomethadone and methadone (40 and 41, respectively) were both much less active than morphine in this study.
Analgesic
R'
(39)
subcutaneously
15 1-3 60 40 3.t
in mice.
424
9
40
Open-Chain
Analgesics
Methadone
41
and Related
Compounds
425
48
49
50
7. Other Modifications of the Ketone Replacement of the ketone side chain by the methyl or ethyl ether (50) produces the active isomethadone analogs 42 and 43. These were reported to be two and five times as potent, respectively, as meperidine. These compounds also exhibited local anesthetic activity.
42
54
43
56
55
Variable results were seen in acyloxy analogs. The acetoxy- and propionyloxynormethadone analogs 44 and 45 were very weak as analgesics (2). The acetoxyisomethadone analog 46 exhibited moderate activity (32). The (+ )-isomer of the propionyloxyisomethadone analog 47 was quite active, while the (- )-isomer of this compound showed no analgesic activity (51). 58
57
59
o
53
52
o
60
reduction products, the alcohols 54 and 55 (2). Esterification of the alcohols (compounds 56-58) did not significantly alter the pharmacological results (2). Weak analgesic effects were observed for the olefin 59 and the chloride 60 (32,34). 46
47
. Most other modifications of the ketone have produced very weak or mactIve compounds. Complete removal of the ketone side chain of methadone (compound 48) resulted in total loss of analgesic activity (34). The mtnle precursors of methadone and isomethadone (compounds 49 and S0,. r~spectively) were ~lso devoid of analgesic effects (32). No analgesic actIvIty was observed m the carboxylic acid derivatives 51-53 or their
C.
Modification of the Diphenyl Fragment
Most modifications of the phenyl groups have resulted in substantial loss of analgesic activity. However, replacement of the diphenyl-carbon fragment with N-arylpropionamide derivatives has produced a number of highly active compounds. These results will be examined in the following two sections.
426
9 Open-Chai~
Analgesics
Methadone
and Related
Table 9-19 Phenyl
o
Group Modifications
in Normethadone
427
Compounds
benzyl
Analogs"
MgCI )
O'YNM<,
:.;(
propionic anhydr I de
X
R
-phenyl -phenyl -phenyl -phenyl -phenyl -p-Cl-phenyl a
b
Data Activity
x
R'
-allyl -benzyl -cinnamyl
-piperidyl -piperidyl -NElz
-m-OMe-phenyl -m-OH-phenyl -p-Cl-phenyl
-morpholinyl
from reference
-NMcz
o o o <0.2 <0.2
61 Scheme
o
= 1.0.
Table 9-20 Phenyl Group
Modifications
Ester
of Normethadone"
Analogs
R
R'
-phenyl -c-hexyl
-c-hexyl -c-hexyl -ftuorenyl-phenyl -ethyl -ethyl -ethyl -phenyl -benzyl Q
b
Data
9.11.
Synthesis
of
propoxyphene
and
related
compounds
(reference
52).
active compounds, while the para-chI oro compound was inactive. Replacement of a phenyl ring with allyl, benzyl, and cinnamyl also produced inactive compounds. Among the esters (Table 9-20), saturation of one or both phenyl rings produced weak analgesics. Tying the two phenyl rings together into a fluorenyl derivative destroyed analgesic activity, as did replacement of one or both phenyl rings with ethyl groups. Replacing a phenyl ring with a benzyl group also resulted in an inactive compound. When the ketone side chain was replaced by a propionyloxy group and a benzyl group was substituted for one of the phenyls, a number of compounds with moderate analgesic activity were discovered (52). Included in this series was propoxyphene, compound 61, which has found use in the treatment of mild to moderate pain in humans (53). These compounds were prepared as shown in Scheme 9-11 (52). The appropriate benzophenone was treated with benzyl magnesium chloride to form mainly one diastereomer (called a) of the carbinol. Acylation with propionic or acetic anhydride produced analgesics with about 1/lOth the potency of methadone. The {3diastereomer was without analgesic activity. Pohland and Sullivan resolved the a carbinol with camphorsulfonic acid and prepared the a-( +)- and a-( - )-isomers of propoxyphene (51). The analgesic activity was found to reside only in the (+ )-isomer. Pohland et al.
1.
to meperidine
relative
-morpholinyl
ActivitY'
in
Activityb <0.2 <0.2 0 0 0 0
from reference J. relative to meperi-
Activity,
dine = 1.0
1. Modification of the Phenyl Groups In their extensive paper de. scnbmg structure-activity relationships in methadone analogs. Bockmuhl and Ehrhart (1) described several analogs in which one or both phenyl groups were modIfied. As may be seen in Tables 9-19 and 9-20, few of these analogs retained any analgesic activity. ]n the ketone series (Table 9-19), meta-hydroxy and meta-methoxy substitution produced weakly
61
(54) later succeeded in resolving the amino-ketone precursor to the carbinol using dibenzoyl tartaric acid, providing an alternative stereoselective synthesis. The absolute configuration of a-( + )-propoxyphene was established by Sullivan et al. (55) via chemical transformation to products of known
.
428
9
Open.Chafn
Analgesics
Methadone
and Related
Compounds
429
o
TO
li
an i line
A IH..
propionic
propionic
anhydride
)
anhydride
63
Scheme 62 Scheme
9-12.
Synthesis
of diampromid
(reference
57).
configuration. They showed that the analgetically active diastereomer had 2S,3R absolute configuration. Shortly thereafter, Casy and Myers (56) confirmed the 3R absolute configuration and pointed out that this is the same configuration
as in the active
(- )-isomethadone.
Racemic propoxyphene was included among the analgesics for which Pert and Snyder (17) measured receptor binding. Consistent with its relatively low level of analgesic activity, this racemate showed a binding constant of 1 x 10-6 M.
9-13.
Synthesis of phenampromid (reference 57).
(e.g., compound 64) produced by treating the intermediate diamines with alkyl chloroformates (58); however, the analgesic activities seen in the carbanilate series were quite low. Structure-activity relationships for a number of propionanilides related to diampromid are outlined in Table 9-21 (59). Only one substitution was tried on the N-phenyl ring; a meta-methoxy substituent reduced analgesic Table 9-21 Propionyl AnilidesQ
2. N-Aryl-Propionamides In 1959, Wright and co-workers (57) reported a series of N-phenylpropionamide derivatives having good analgesic activity. The most interesting compounds in this series were the methadone analog diampromid, 62, and the isomethadone analog phenampromid, 63. R
62
63
Diampromid was reported to have analgesic activity between that of morphine and meperidine (57). Phenampromid was somewhat less active, being about equal to meperidine in rats (57) and equal to codeine in mice (58). Diampromid was prepared from 2-bromopropionanilide, as shown in Scheme 9-12 (57). The resulting amide was reduced with lithium aluminum hydride to the diamine, then acylated with propionic anhydride. In the synthesis of phenampromid (Scheme 9-13), the 2-bromopropionyl amide of piperidine was first reacted with aniline (57). Reduction with lithium aluminum hydride and acylation were again employed to produce the desired product. These workers also explored a series of carbanilates
.H .H .H .H .H .H
-H -H -OCH,
-H
n
R'
ADso.
1 1 1 1 1 2 2 2 2 3
.H -m-Me -p-Me .p.Cl -p.F -H -m-Me -p.NH2 -H
8 2 3 6 6 4 14 16 15 17 4 11 3
Compound 65 Meperidine
Morphine a
b
Data
from reference 59.
ADso = subcutaneous per kilogram) that elevates heat response animals.
dose (in milligrams the rat tail radiant
time by 100% in 50% of the
430
9
Open-Chain
I
Analgesics
Methadone
and Related
431
Compounds
Table 9.22 Analgesic
Activities
of Stereoisomers
of PhenampromidQ
Compound
ADsOh
(00)-63 (-)-63 (+)-63 Morphine
13 9 36 3
Q
Data from reference
jre50lved
[
tar
tar
with i c: aC
58.
i d)
phenylacetaldehyde
h
AD~ = subcutaneous dose (in milligrams per kilogram) that elevates the rat tail radiant heat response time by 100% in 50% of the animals.
)
I-t-)-diarnpromid
activity by a factor of 4 relative to the unsubstituted compound. In the benzylamine series (n = 1), substitution on the benzyl aromatic ring enhanced the potency somewhat. When the amine substituent was changed from benzyl to phenethyl (n = 2), however, ring substitution decreased analgesic activity significantly. Analgesic activity was also reduced when the amine substituent was lengthened to phenylpropyl (n = 3), but compound 65, with an N-cinnamyl substituent, showed good activity.
Scheme
9-14.
Synthesis
of enantiomers
and (+)-diampromid
Table 9.23
Influence of Alkyl Chain Length on Stereosclectivity in Enantiomeric Propionanilides
n
Wright and co-workers (58) resolved the stereoisomers of phenampromid, 63, using malic acid. As may be seen in Table 9-22, the (- )-isomer was about four times as active as the (+ )-isomer. Portoghese (60) showed that (-)-63 has the same absolute configuration (R) as the more potent stereoisomer of isomethadone (20), indicating that these compounds have the same stereochemical requirements. The stereoisomers of diampromid, 62, and the N-benzyl analog 66 were prepared (61) as outlined in Scheme 9-14. Resolution was carried out on an N-benzyl diamine, with tartaric acid as the resolving agent. The (-)_ diamine was acylated with propionic anhydride to form (- )-66. The benzyl group was removed by catalytic hydrogenation, and the phenethyl side chain was added in the last step to yield (+ )-diampromid. Analgetic potencies measured for these compounds are shown in Table 9-23. Note that the more potent isomers, (- )-66 and (+ )-62, are derived from the same precursor and therefore have the same absolute configuration. Portoghese and Larson (63) related the more active isomers to L-(+)_
of (-)-66
a
Compound
(00)-66 R-(+)-66 5-(-)-66 (00)-62 5-(+)-62 R-(-)-62 (00)-67 5-(+)-67 R-(-)-67
2
3
" Data
from
references
8 Inactive 4.3 3.7 3.6 11.7 12.5 8.9 11.9 61 and
62. b AD50 = subcutaneousdose (in milligrams per kilogram) that elevates the rat tail radiant heat response time by 100% in 50% of the animals.
(reference 61).
432
9
Open-Chain
alanine, indicating that they have the S absolute configuration. This is in surpnsmg contrast to the finding that the more active isomer of methadone has the R absolute configuration (64). Portoghese and Riley (62) extended ~he~r study of stereochemical effects to the phenylpropyl analog 67. As mdlcated m Table 9-23, stereoselectivity in the analgesic activity falls off as the alkyl chain length increases. For compound 67, the difference in activity between the (+)- and (- )-isomers is not statistically significant. H,!tmann et al. (65) published an extensive structure-activity study on proplram, .68, and related compounds in which the phenyl ring of the proplOnamhdes was. replaced with a pyridyl ring. These compounds showed mixed agomst and antagonist properties. The best analgetic acllVlltes were seen for the 2-pyridyl compounds; the 3-pyridyl isomer 69 was somewhat less active, and the 4-pyridyl compound 70 was completely macllve as an analgesic. Methyl substituents on the 2-pyridyl ring decre~sed or abo~ished analgesic activity, but the 4-phenyl compound 71 was 15 times as active as 68. The latter compound was not pursued because it also showed relatively high toxicity.
0,) aNiO 6'
0,)
aNio 70
I
Analgesics
0,)
orNiO 69
0,)
NNiO 71
Variation of the acyl group (Table 9-24) indicates that the propionyl substituent IS clearly superior to other choices. The formamide and cyclohexylcarboxamide are inactive, and the butyryl and benzoyl derivalives show only moderate analgesic activity. Anal~esic activity i~ highly sensitive to substitution on the alkyl chain connectmg the two mtrogen atoms (Table 9-25). A single methyl substituent on the carbon adjacent to the amide nitrogen (i.e., compound 68) appears to be nearly optimum; compounds having an ethyl group or two
Methadone
and Related
433
Compounds
Table 9-24
Table 9-25
Variation of the Acyl Group in the Propiram Series"
Variations of the Alkylpiperidine Propiram Series"
Group in the
[
R .H -Me
-E' (68) -nPr -c-hexyl -phenyl
EDsoh Inactive 100 (in mice)
9.3 20 Inactive 35
68
R,
R,
R.,
EDsob
-H -Me -Et -Me
-H .H
-H
.H
Inactive
-H -H -H -Me -Me
-H -H -H -H -H
9.3 >20 50 5-6 Inactive
-Me
.Me
Inactive
.Me
-H -H
a
Data from reference 65. b ED50. milligrams per kilogram. subcutaneously in rats (except as indicated).
R,
-H -Me -H -H -H
a Data from reference 65. b ED50, milligrams per kilogram, neously
subcuta-
in rats.
methyl groups in this position show much less analgesia. When there are no methyl substituents present or when there are methyl groups on the carbon adjacent to the piperidine, analgesic activity is absent. The exception is the mixture of four diastereomers having one methyl group on each carbon; both racemic pairs were as active as propiram. Nearly 50 variations of the dialkylamino group were reported; some representative examples are listed in Table 9-26. Most of the noncyclic dialkyl amines showed only moderate or weak analgesic activity. A notable exception was compound 72, which was tested as a mixture of two diastereomeric racemates. Among the simple cyclic amines, the piperidyl derivative was most active. Compounds having 7- or 8-membered rings also showed good activity, but those with 5- and 9-membered rings were much less potent. Substituted piperidines showed a range of activities, from the weakly active 2,2-dimethyl derivative to the relatively potent 3, 3-dimethyl derivative. Here again, phenyl substitution produced an anomalous effect; while a 4-methyl substituent decreased activity by a factor of 3, the 4-phenyl derivative was more than 100 times as potent as the unsubstituted parent compound. Numerous bicyclic amines were also tested; several of these are also included in Table 9-26. The absolute configurations of the (+)- and (- )-isomers of propiram, 68, and the analogs 73, 74, and 75 were reported by Wollweber (66). In all
434
9
Open-Chain
Analgesics
II
435
Other Open-Chain Compounds
Table 9.26 Table 9-27 Variations Series"
of the Dialkylamino
Group in the Propiram Analgesic Activities of Stereoisomers of Propiram Related Compounds"
Compound [
Dialkyl
amines
-NMe2 -NEtl -N(nPr), -N(Me)-benzyl -N(Me)-phene'hyl -N(Me)-CH(CH,)-CH,-phenyl (mixture Cyclic
160 18 >20 >20 19.5 1
of 4 diastereomers)
amines
a
-pyrrolidyl -piperidyl
42 9.3 12.0 11.7 42.6 5.89 4.96 26 0.07 34.7 0.98
(68)
-azacycloheptyl -azacyclooctyl -azacyclononyl -2-Mc-piperidyl -3-Me-piperidyl -4-Me-piperidyl -4-phenylpiperidyl -2,2-Mc2-piperidyl -3,3-Mel"piperidyl Bicyclic
(72)
EDsOb
(=)-68 5(+)-68 R(-)-68 (=)-73 5(+)-73 R(-)-73 (=)-74 5(+)-74 R(-)-74 (=)-75 5(+)-75 R(-)-75
NR,
b
11 13 18 0.98 1.6 1.2 2.34 5.07 1.87 100 37 61
o~
I
~N'V'N~ N
I
(mouse) (mouse) (mouse) (rat) (rat) (rat) (rat) (rat) (rat) (mouse) (rat) (rat)
Data from reference 66. EDso,
milligrams
administered
~
and
per kilogram,
subcutaneously.
~
I
72
73
amines
-3-azabicyclo[3.1.0]hexanyl -3-azabicyclo[3.2.1]octanyl -2-azabicyclo[
2.2.2 ]octan
yI
-l,2,3,4-tetrahydroisoquinolyl "b Data from reference 65. EDSIh milligrams per kilogram, rats.
1.58 1.09 11.0 Inactive 75
subcutaneously
74
in
cases, Ihe difference in analgesic activity between the (+)- and (- )-isomers was small. Table 9-27 shows that, for propiram and 75, the S-( + )-isomer is more active, while the R-( - )-isomer is more active for compound 74. The S configuration in this series gives the same arrangement of functional groups as the R configuration in isomethadone (the order of precedence changes in going from the diphenylmethyl substituent of isomethadone to the
acylaminopyridine of propiram), so the more. active stereoisomer of propiram corresponds to the more active stereOIsomer of Isomethadone (20).
II.
Other Open-Chain Compounds
There are several groups of open-chain analgesic compounds that cannot be considered as analogs of the methadone series. These wtll be dtscussed in the following sections.
436
9
Open-Chain
Analgesics
II
Other Open-Chain
Compounds
437
Table 9.28 th i eny
I
Li )
Scheme 9-15.
A.
Synthesis of compounds
Thiambutene
Analgesic Activities of Dithienylbutenylamines
in the thiambutene
7. series (reference
67).
Q
}rNR'
[
and Related Compounds NR,
Adamson (67) described a series of dithienyl-alkenylamines and alkanolamines having analgesic, spasmolytic, anti histaminic, and local anesthetic activity. Included in this series was thiambutene, 76, which has analgesic potency comparable to that of morphine. As shown in Scheme 9-15, thiambutene was prepared by treating an aminoester with two equivalents of thienyllithium to form the tertiary carbinol shown. Yields in this reaction were satisfactory only in the case of tertiary amines. The carbinols in this series were without analgesic activity, although they did possess antispasmodic and local anesthetic effects. Treatment with acid produced the olefin.
(;0)(+)(-)-
.NMe, -NMe, -NMe, -NMeEt -NEl2 (;0)- -pyrrolidyl (+ )- -pyrrolidyl (-)-pyrrolidyl -piperidyl -2-Me-piperidyl -3-Me-piperidyl -4-Me-piperidyl -azacycloheptyl -azacyclooctyl a
76
1.07 1.7 0.27 1.7 1.0 0.7 1.5t 0.64 1.1 0.5-0.6 0.6-0.9 0.6-0.9 1.3-1.8 0.05-0.10
Data from references 68 and 69.
b Activity upon subcutaneous injection in rats. Morphine = 1.
77
The analgesic activities of these compounds (68,69) are listed in Table 9-28. Among the cyclic amine derivatives prepared, the azacycloheptyl analog was most potent. The piperidine and pyrrolidine derivatives also showed good activity, but methyl-substituted piperidines were somewhat less active, and the aza cyclooctyl derivative was quite weak. Beckett and co-workers (69) resolved the optical isomers of thiambutene and found that the dextrorotatory isomer was about six times as potent as the levorotatory isomer. The more active stereoisomer was shown (22) to have the same absolute configuration as the more potent (- )-isomer of methadone. Saturation of the olefinic double bond was tried on several compounds in this series (70). These compounds were generally one-fourth to one-fifth as . active as the parent olefins.
Activityb
of these compounds was accomplished in two steps, as illustrated in Scheme 9-16. The most active compound in their series, 77, was reported to be about 1000 times as potent as morphine on subcutaneous administration in rats. Some of the structure-activity relationships derived for this series are
O,N
O I
.&
.
NH,
+ NHz
o,NOh B. Benzimidazole-Based Compounds Hunger el al. (71,72) reported the synthesis of a number of extremely potent analgesics built upon the benzimidazole ring system. The synthesis
CI-CH2-CH2-NEt~ 77 Scheme 9-16.
Preparation
~Q
of analgesics in the benzimidazole
NEtt OE.t series (reference
71).
438
9
Open-Chain
A"nalgesics
II
Other
Open-Chain
439
Compounds
Table 9-29 Analgesic Compounds
Activities
of Some Benzimidazole-Based
a
9-17.
Scheme
Preparation
of compounds
related
to tilidine
(reference
74).
[
Activity R .H -NOz -N02 -NOz -N02 Q
b
-4-CI -4-CI -3,4-(OMe), .4-0Me -4-0Et
Data from reference Activity
Morphine e
Subcutaneousb
R'
Activity
0.1 1 10 30 WOO
Oral" 0.5 5 3 15 1250
72.
on subcutaneous
injection
= 1. on oral administration
in rats.
summarized in Table 9-30. Most of these compounds have only weak or moderate analgesic activity, but tilidine (78) showed good activity. In a clinical study, (75) a 50-mg dose of tilidine was found to produce analgesia equivalent to that of a 100-mg dose of meperidine. Dubinsky and coworkers (76) found that in rats, analgesia was correlated with levels of a tilidine metabolite in the brain and suggested that the metabolite might contribute to the analgesic activity observed.
Table 9-30 Structure-Activity
in mice.
Morphine = I.
listed in Table 9-29. Substituents on the benzimidazole and phenyl rings were shown to have a profound effect on the level of analgesic activity. Comparing the last two compounds in the table, simply changing a para-methoxy group to a para-ethoxy increased analgesic activity by almost two orders of magnitude in rats and mice. Eddy (73) reported EDso values for these two compounds of 0.015 and 0.001 mglkg.
R
H 'R'
60
Studies
cJ6R'
.
R 78
C.
Cyclohexylamine-Based
Compounds
Several structurally diverse groups of analgesics have in common the aminocyclohexane or aminocyclohexene fragment. Satzinger (74) reported a series of cyclohexene derivatives prepared by the Diels-Alder reaction shown in Scheme 9-17. The reaction produced both cis and trans isomers. In some cases, the olefin was subsequently saturated by hydrogenation. Qualitative measures of analgesic activities for these compounds are
.COOiPr .COOMe .CH,OH .CH,OAc Q
Data from reference
74.
CompoundsQ
er6R'
.
C
R'
Isomer
Activity
A B C D B B B A B B A B B A B B D D
Strong Moderate Weak Weak Moderate Weak Moderate Moderate Weak Weak Moderate Weak Moderate Moderate Weak Moderate Inactive Weak
.NEtz .pipcridyl .NHMe -pyrrolidyl
.COCH, .CN
and Related
cJ6R'
.COOEt
.COOEt .COOEt .COOEt .COOEt 78
for Tilidine
D
440
9
Open-Chain
f\nalgesics
II
Other Open-Chain
Compounds
441
6 R-benzoyl
chloride
[' Scheme 9-/8.
Preparation
of cydohexylmethyl
benzamides
(reference
77).
Harper el al. (77) reported the preparation of 79 and several related benzamidomethylcyclohexylamine derivatives by the method shown in Scheme 9-18. While 79 has a potency close to that of morphine, most of the compounds reported in this series are substantially lower in analgesic activity (77,78). Table 9-31 shows the analgesic activities reported for compounds of this type.
D,
Benzylamine-Based
Compounds
Two compounds (82 and 83) with moderate analgesic activity were described by Wilson and Pircio (79). These compounds were prepared by treating the cyclobutane derivative 84 with either phenyl Grignard or cyclohexyl Grignard reagent. These two compounds were reported to be about equal to codeine on intraperitoneal administration.
NMe'z 82
83
Table 9-31 Analgesic
Activity
Workers at Dai,!ippon in Japan (80,81) prepared an extensive series of diphenyleihylpiperazine compounds such as 85. Selected results from this wor-kale shown' in Table 9-32. Among the piperazine substituents examined, medium-sized (6,7,8) cycloalkyl rings gave the best results. Good activity was also seen for the para-methoxybenzyl group (compound 86) at this position; the para-nitrobenzyl derivative was inactive. Substituents on the two phenyl rings decreased analgesic potency in most cases. The
of Some
~
Cyclohexylmethylbenzamides"
R
-H .2-CI -3.CI .4-CI 3,4.CI, 3,4-CI, 3,4.CI, 4.F -H Morphine Codeine
R'
-NMe2 -NMe2 -NMe2 -NMe2 -NMe2 -piperidyl -N' -Me-piperazinyl
-NMe2 -piperidyl
15.5 60 9.5 5.0 2.5 >!OO >100 5.0 >100 2.0 t7.0
from references 77 and 78. EDso, in milligrams per kilogram. on subcutaneous injection in mice. b"
Data
8.
442
9
Open-Chain
Analgesics
II
Other
Open-Chain
443
Compounds
Table 9-32
8r
Diphenylethylpiperazine
AnalgesicsQ
p-bromophenyl
MgBr )
N) '-N ,
['
R.
R,
R, -H
-II
R, -H
R,
EDsob
-H -cyclopentyl
-cycIohexyl -cycloheptyl -cycIooctyl -cyclododecyl
-phenyl -iBu
-benzyl -4-Me-benzyl -4-NOz-henzyl
-H -H
-H -COMe -COE! -COE' Codeine
-2-Me -3-Me -4-Me -H
-H -2-0Me -4-0Me -2-CI
-4-0Me-henzyl (i:) (+) (-) -4-0Me-benzyl
-4-0Me-benzyl
(;0) (+) (-)
-4-Me -H
-3-Cl -4-Cl -2-Me -3-Me -4-Me -4-NO, -4-Me -H
-4-0Me-benzyl -4-0Mc-benzyl
-H
-2-CI
-4-0Me-benzyl
a Data from references b EDso, in milligrams
>160 36.9 3.09 5.45 3.45 >160 >160 > 70 >160 >160 >160 15.2 40.1 13.5 120 26.4 31.7 32.5 37.4 79.0 58.4 >160 48.1 39.7 65.2 28.0 14.1 8.11 98.6 80-160 73.1 >160 28.1
81 and 82. per kilogram, on subcutaneous
injection
in mice.
Substituents not listed in the table are assumed to be the same as for the preceding compound in the table.
) 21
phenethy
+
Q
80
Br
I "'gBr
H0>aNM~J
()' 0
81
Br Scheme 9-19.
Synthesis of compounds
80 and 81 (reference
stereo isomers of 86 were resolved, and Ihe dextrorotatory
79).
enantiomer was found to be about Ihree limes as potent as the levorotatory enanliomer. Lednicer and Von Voigtlander (82) reported that compound 80 has an ED,o in mice of 0.1 /lg/kg, about 10,000 times as potent as morphine. Receptor binding studies by Ihese aulhors showed 80 to have an IC,o of 8 x 10-10 M (compared to 2.4 x 10-8 M for morphine). In comparison, the stereoisomer 81 was much less active (EDso of 7-8 mg/kg). The synthesis of 80 and 81 is shown in Scheme 9-19. The cyano group of the aminonilrile was displaced by the para-dibromobenzene Grignard reagent to introduce the bromophenyl group. Yardley and co-workers (83) reported the preparalion of ciramadol, 87. They resolved the compound using tartrate and found the ( + )-isomer to be inactive. The (- )-isomer was shown to have mixed agonist and antagonist activity, with analgetic activily about two times Ihal of morphine on intraperitoneal, intramuscular, or oral administration. In a clinical study (84), ciramadol was shown to be somewhal more potenl as an analgesic than pentazocine.
HO~ I", 87
88
444
9
Open.Chain
Analgesics
References
445
Table 9-33 Stereoisomers
References
of Viminol"
1. M. Bockmuhl and G. Ehrhart, Liebigs Ann. Chern. 561,52 (1948).
2. K. K. Chen. Ann. N. Y. Acad. Sci. 51, 83 (1948).
[ Configuration Isomer
R, R, S, S, Meso
A
B
C
ED",
R R S S R
R R S S S
+
>20 0.9 >20 >20
Morphine
+ '"
"
>20
5.0
"b Data from reference 85. EDso. in milligrams per kilogram, on intraperitoneal administration in rats.
E.
Miscellaneous Open-Chain Compounds
Vi~inol, 88, is a pyrrole derivative reported (85) to have good analgesic actIVIty and physIcal dependence liability comparable to that of pentazocIne.. The compound exhibits both agonist and antagonist activities. VlmInol possesses three asymmetric centers. Della Bella and co-workers (85) separated most of the possible stereoisomers, as shown in Table 9-33. The analgesic potency was found to reside only in the levorotatory enantiomer of the isomer in which both sec-butyl groups have the R absolute configuration. Antagonist activity was found to reside in the levorotatory enantjomer in which both sec-butyl groups have the S absolute configuration. Carrano. et al. (86,87) reported the analgesic activity of N-butyroyl-N'. cInnamylplperazIne, 89. ThIs compound was described as having low physIcal dependence liability in rodents. It appeared to be particularly effective when administered orally.
3. A. F. Casy. Prog. Dmg Res. 22, 149 (1978). 4. H. Isbell, A. Wikler, N. B. Eddy, J. L. Wilson, and C. F. Moran, lAMA, J. Am. Med. Assoc. 135, 888 (1947). 5. F. G. Everett, Anesthesiology 9, 115 (1948). 6. E. C. Kleiderer, J. B. Rice, and V. Conquest, Report 981, Office of the Publication Board, Dept. of Commerce, Washington, D. C. (1945). 7. E. M. Schultz, C. M. Robb, and J. M. Sprague, J. Am. Chem. Soc. 69,2454 (1947). 8. N. R. Easton, J. H. Gardner, and J. R. Stevens, J. Am. Chem. Soc. 69,2941 (1947). 9. A. L. Morrison and H. Rinderknecht, J. Chem. Soc. p. 1478 (1950). 10. M. Sletzinger, E. M. Chamberlin, and M. Tishler, J. Am. Chem. Soc. 74,5619 (1952). 11. R. H. Thorp, E. Walton, and P. Ofner, Nature (London) 160, 605 (1947). 12. A. A. Larsen, B. F. Tullar, B. Elpern, and J. S. Buck, J. Am. Chem. Soc. 70,4194 (1948). 13. R. H. Thorp, Br. J. Pharrnacol. 4, 98 (1949). 14. C. C. Scott, E. B. Robbins, and K. K. Chen, J. Pharmacal. Exp. Ther. 93,282 (1948). 15. c.-Y. Sung and E. L. Way, J. Pharmacal. Exp. Ther. 109,244 (1953). 16. A. H. Beckett and A. F. Casy, J. Pharm. Pharmacal. 6, 986 (1954). 17. C. B. Pert and S. H. Snyder, Science 179, 1011 (1973). 18. D. T. Wong and J. S. Horng, Life Sci. 13, 1543(1976). 19. J. S. Horng, S. E. Smits, and D. T. Wong, Res. Commun. Chem. PathoJ.Pharmacal. 14, 621 (1976). ~ R,UjlTo.. iSl">..4-", A J;p.;.i/lu ,F'II1': i--J...,,{ol, 20. D. G. Leimbach and N. B. Eddy. J. Pharmacal. Exp. Ther. 110, 135 (1954). 21. E. J. Jenney and C. C. Pfeiffer, Fed. Proc., Fed. Am. Soc. Exp. BioI. 7,231 (1948). 22. A. H. Beckett and A. F. Casy, J. Chern. Soc. p. 900 (1955). 23. A. H. Beckett and N. J. Harper, J. Chem. Soc. p. 858 (1957). 24. A. H. Beckett, G. Kirk, and R. Thomas, J. Chem. Soc. p. 1386 (1962). l25. H. R. Sullivan and S. L. Due, J. Med. Chern. 16, 909 (1973). 26. A. H. Beckett, J. Taylor, A. F. Casy, and M. M. A. Hassan, J. Pharm. Pharmacal. 20, 754 (1968).
27. 28. 29. 30.
A. Pohland, H. E. Baal, and H. R. Sullivan, J. Med. Chern. 14, 194 (1971). H. R. Sullivan, S. L. Due, and R. E. McMahon, J. Am. Chem. Soc. 94, 4050 (1972). A. H. Beckett, D. P. Vaughan, and E. E. Essian,J. Pharm. Pharmacal. 24,244 (1972). D. J. Dupre, J. Elks, B A. Hems, K. N. Speyer, and R. M. Evans, J. Chem. Soc.p. 500 (1949). e~yl-k.. -b f-h...O 31. F. F. Blicke and E.-P. Tsao, J. Arn. Chern. Soc. 76, 2203 (1954). 5-""J.~yl ~tH..","" 32. G. Satzinger,
.33. 34. 35. 36.
Liebigs
Ann.
Chem.
728, 64 (1969).
J
J. G. Henkel, E. P. Berg, and P. S. Portoghese, J. Med. Chem. 19, 1308 (1976). E. L. May and E. Mosettig, J. Org. Chern. 13, 459 (1948). E. L. May and E. Mosettig, J. Org. Chern. 13, 663 (1948). M. E. Speeter, W. M. Byrd, L. C. Cheney, and S. B. Binkley, J. Am. Chern. Soc. 71,57 (1949). 37. A. Pohland, F. J. Marshall, and T. P. Carney, J. Am. Chern. Soc. 71, 460 (1949). 38. N. B. Eddy, E. L. May, and E. Mosettig,J. Org. Chern.17, 321 (1952). 39. E. L. May and N. B. Eddy, J. Org. Chern. 17, 1210 (1952). 40a. N. A. David, H. J. Semler, and P. R. Burgner, JAMA, J. Am. Med. Assoc. 161,599 (1956).
446
9
Open-Chain
A~algesics
40b. N. B. Eddy, J. Arn. Pha'rn. Assoc., Sci. Ed. 39, 245 (1950). 41. A. Zaks, M. Fink, and A. M. Freedman, lAMA, J. Am. Med. Assoc. 220,811 (1972). 42. R. E. Billings, R. Booher, S. Smits, A. Pohland, and R. E. McMahon, J. Med. Chern. 16, 305 (1973). 43. R. Nickander, R. Booher, and H. Miles, Ufe Sci. 14, 2011 (1974). 44. P. A. J. Janssen and A. H. Jageneau, J. Pharm. Pharmacal. 9, 381 (1957). 45. A. S. Keats, J. Telford, and Y. Kurosu, J. Pharmacol. Exp. Ther. 130,212 (1960). 46. W. H. Shelver, M. Schreibman, N. S. Tanner, and V. Subba Rao,i. Med. Chern. 17,120 (1974). 47. M. M. Klenk, C. M. Suter, and S. Archer, J. Am. Chern. Soc. 70, 3846 (1948). 48. B. F. Tullar, W. Wetterau, and S. Archer, J. Am. Chern. Soc. 70, 3959 (1948). 49. L. C. Cheney, R. R. Smith, and S. B. Binkley, J. Arn. Chern. Soc. 71, 53 (1949). 50. Ger.Patent 1,167,357 (1964); Chern. Abs". 61, 1801. 51. A.Pohland and H. R. Sullivan, J. Am. Chem. Soc. 77, 3400 (1955). 52. A. Pohland and H. R. Sullivan, J. Am. Chem. Soc. 75, 4458 (1953). 53. C. M. Gruber, J. Lab. Clin. Med. 44, 805 (1954). 54. A. Pohland, L. R. Peters, and H. R. Sullivan, J. Org. Chem. 28, 2483 (1963). 55. H. R. Sullivan, J. R. Beck, and A. Pohland, J. Org. Chem. 28, 2381 (1963). 56. A. F. Casy and J. L. Myers, J. Pharm. Pharmacol. 16, 455 (1964). 57. W. B. Wright, Jr., H. J. Brabander, and R. A. Hardy, Jr., J. Am. Chem. Soc. 81, 1518 (1959). 58. W. B. Wright, Jr., H. J. Brabander, and R. A. Hardy, Jr., J. O,g. Chern. 26,476 (1961). 59. W. B. Wright, Jr., H. J. Brabander, and R. A. Hardy,Jr.,J. O,g. Chern. 26,485 (1961). 60. P. S. Portoghese, J. Med. Chem. 8, 147 (1965). 61. W. B. Wright, Jr. and R. A. Hardy, Jr., J. Med. Chem. 6, 128 (1963). 62. P. S. Portoghese and T. N. Riley, J. Pharm. Sci. 54, 1831 (1965). 63. P. S. Portoghese and D. L. Larson, J. Pharm. Sci. 53,302 (1964). 64. A. H. Beckett, P,ag. D,ug Res. I, 455 (1959). 65. R. Hiltmann, F. Hoffmeister, E. Niemers, U. Schlichting, and H. Wollweber, Arzneim. Fonch. 24, 584 (1974). 66. H. Wollwcbcr, Eu,. J. Med. Chern. 17, 125 (1982). 67. D. W. Adamson, J. Chern. Soc. p. 885 (1950). 68. A. F. Green, Br. J. Pharmacol. 8, 2 (1953). 69. A. H. Beckett, A. F. Casy, N. J. Harper, and P. M. Phillips, J. Pharm. Pharmacol. 8, 860 (1956). 70. D. W. Adamson, W. M. Duffin, and A. F. Green, Nature (London) 167, 153 (1951). 71. A. Hunger, J. Kebrle, A. Rossi, and K. Hoffmann, Experientia 13,400 (1957). 72. F. Gross and H. Turrian, Experientia 13, 401 (1957). 73. N. B. Eddy, Chern. Ind. (London) p. 1462 (1959). 74. G. Satzinger, Liebigs Ann. Chem. 728,64 (1969). 75. A. L. Mauro and M. Shapiro, Curro Ther. Res. 16, 725 (1974). 76. B. Dubinsky, M. C. Drew, M. D. Melgar, J. K. Karpowicz, and F. J. DiCarlo, Biochem. Pha'rnacol. 24, 277 (1975). 77. N. J. Harper, G. B. A. Veitch, and D. G. Wibberley, J. Med. Chem. 17,1188 (1974). 78. R. T. Brittain, D. N. Kellet, M. L. Neat, and R. Stables, Br. J. Pharmacal. 49, 158P (1973). 79. A. Wilson and A. W. Pircio, Nature (London) 206, 1151 (1965). 80. K. Natsuka, H. Nakamura, H. Uno, and S. Umemoto, J. Med. Chem. 18, 1240 (1975). 81. N. Shimokawa, H. Nakamura, K. Shimikawa, H. Minami, and H. Nishimura, J. Med. Chern. 22, 58 (1979).
References
D. Lednicer and P. F. Von Voigtlander, J. Med. Chem. 22, 1157 (1979). J. P. Yardley, H. F1etcher Ill, and P. B. Russell, Experientia 34,1124 (1978). F. Camu, Eur. J. C/in. Pharmacal. 19, 259 (1981). D. Della Bella, V. Ferrari, V. Frigeni, and P. Lualdi, Nature (London) New BioI. 241, 282 (1973). 86. R. A. Carrano, K. K. Kimura, R. C. Landers, and D. H. McCurdy, Arch. Int. Pharma. codyn. 213, 28 (1975). 87. R. A. Carrano, K. K. Kimura, and D. H. McCurdy, Arch. Int. Pharmacodyn. 213, 41 (1975).
82. 83. 84. 85.
[
447
Physical Chemistry Studies of Open-Chain
Analgesics
449
10. Physical Chemistry and Molecular Modeling of Open-Chain Analgesics I. Physical Chemistry Studies of Open-Chain Analgesics A. Methadone and homcthadone B. Methadols and Isomethadols C. Dextromoramide D.
Propoxyphcnc
. . .
44R 44R 451 453 453 454 456 457
. . . .
E. Other Open-Chain Analgesics ..... II. Molecular Modeling of Open-Chain Analgesics References
I. Physical Chemistry Studies of Open-Chain Analgesics
Methadone
and Isnmethadone
Beckett (1) proposed that methadone, 1, and isomethadone, 2, are stabilized in a cyclic conformation by intramolecular interactions between the ketone and amino groups. They considered two types of interactions, shown in Fig. 10-1. The hydrogen bond postulated in Fig. IO-Ia should stabilize the ionized form, raising the pK" while the lone pair carbonylcarbon interaction of Fig. IO-Ib should stabilize the un-ionized form and
;:p~ ~~ ~ Fig. 10.1. Possible intramolecular ence 1).
~
431 interactions
448
Proposed solution conformation
of methadone
(reference
I).
lower the pK,. The pK, values measured for methadone and isomethadone were 8.25 and 8.07, respectively, some 0.1-0.3 pK, units lower than the analogous compounds lacking the ketone. Beckett proposed the conformation shown in Fig. 10-2 as the receptor active form of methadone.
Proton nuclear magnetic resonance (NMR) has been a part of several investigations on methadone conformations. Methadone and isomethadone conformations in chloroform were investigated by Smith (2). He found that the two N-methyl groups exhibited nonequivalence, supporting a single preferred conformation. In addition, the C-methyl substituent of methadone was shielded by about 0.5 ppm, consistent with a carbonyl-NH hydrogen bonded conformation that places the methyl group above one of the aromatic rings. The shielding effect was not seen in analogs having cyano-, hydroxy-, or proton substituents in place of the propionyl group (3). Haller and Schneider (4) examined normethadone, 3, and a deuterated derivative. They calculated proton NMR spectra for the folded (gauche) and extended (antiperiplanar) conformations of the dimethylaminoethyl side chain. The experimentally determined spectrum closely resembled the spectrum calculated for the extended conformation.
3
(b)
in methadone
Fig. 10-2.
2
The open-chain analgesics exhibit all the activities, side effects, and physical dependence properties of the cyclic opiates, despite the lack of obvious structural similarity to the traditional opiates. Numerous physical studies have been carried out on these conformationally flexible compounds in order to try to find similarities to morphine and arylpiperidine analgesics. A.
c
and isomethadone
(refer-
In another study of methadone and isomethadone, Henkel et al. (5) found methadone to be more conformationally mobile than isomethadone, since methadone exhibited a solvent-induced inversion in the circular dichroism spectrum. These authors found that methadone (but not
.., 4\0
10 Physical Chemistry and Molecdar
Modeling of Open-Chain
Analgesics
Table 10-1 Vicinal Proton Coupling erythro-5-Methylmethadooea
Compound
Constants
for threo~ and
CDCI..
CD..OD
D,O
6.7 7.2
7.0 7.6
nd' nd
rhreo.4 erythro-4 threo-4. DC! erylhro-4 . DCt
nd
"t
8.3
6.6
"t
6.0
From reference 6. b" Not determined.
isomethadone) underwent rapid proton exchange of the protons a to the carbonyl group, presumably via an intramolecular amino group participation. They interpreted this to mean that methadone may adopt a folded conformation, while isomethadone is held relatively rigidly in an extended conformation.
I
The erythro and threo isomers of 5-methylmethadone, 4, were prepared and their conformational properties examined by Henkel et al. (6). In this series, the threo pair of racemates was without analgesic activity, while the erythro pair was found to be 5.4 times as active as methadone. Table 10-1 shows the vicinal proton coupling constants for these compounds and their salts in three solvent systems. For the most part, the values observed are consistent with conformationally flexible molecules, but the very low value seen for the threo salt was proposed to be due to a strongly hydrogen bond-stabilized folded conformation, as shown in Fig. 10-3. Measurement of pK, values for these compounds gave further support to this suggestion: the threo isomer had a substantially higher pKa than the erythro isomer ,. ,.
l
Fig. 10-3. Folded conformation encc 6).
\/ Me C~H O-trN'=i
inactive threo-4 (refer-
Analgesics
451
(8.71 and 8.16, respectively), indicative of intramolecular hydrogen bonding in the threo compound. Subsequent X-ray crystallographic studies of these compounds (7,8) showed the erythro compound in a folded conformation and the threo in an extended conformation. The more potent enantiomer of erythro-4 was found to have the 55,65 absolute configuration. This finding is consistent with the result in isomethadone and related compounds; the 5-position configuration is apparently more important than the 6-position configuration. Crystallographic studies have generally confirmed the conformational flexibility of the open-chain analgesics; no consistent pattern has been observed in the structures determined to date. Hanson and Ahmed (9) found methadone hydrobromide to be in an extended conformation. The crystal structure of the methadone base, however, showed a folded conformation (10,11). In this case, there is a close contact between the amine nitrogen atom and the carbonyl carbon (the N-C distance was found to be 2.9 A, about 0.3 A shorter than the sum of the van der Waals radii of the two atoms). This arrangement is essentially the same as that proposed by Beckett (1), shown in Fig. IO-Ia. This attraction induces several distortions in the molecule. Several bond angles are distorted from tetrahedral geometry (109') to 114', and the carbonyl carbon is almost 0.1
4
Physical Chemistry Studies of Open-Chain
A out
of the plane formed
by the atoms attached
to it.
In other crystal structure determinations, Shefter (12) found that isomethadone hydrochloride exists in an extended conformation. SimilarIy, an extended conformation was observed for the hydrochloride salt of normethadone, 3 (13). Kaufman and co-workers (14) measured pKa values and distribution coefficients for methadone and a-acetylmethadol. They found that the distribution coefficients are highly dependent on the temperature at which the measurement is made. Since such measurements are routinely made at 20 or 25'C and in vivo distribution takes place at 37'C, these authors suggest that the temperature dependence of the results may have a substantial impact on the observed distribution behavior. B.
MethadoIs and IsomethadoIs Having two asymmetric centers, the methadols have presented a more complex and often contradictory conformational picture. Casy and Hassan (15) carried out an infrared spectroscopic study of a-methadol, {3methadol, and normethadol (compounds 5 and 6) in nonpolar solvents. They observed evidence of strong intramolecular hydrogen bonds in both the free bases and salts of these compounds, and proposed the folded conformations shown in Fig. 10-4.
4\2
10
Physical
Chemistry
and Molecular
Modeling
of Open-Chain
Analgesics
~~M' ~ H (a)
(b)
Proposed hydrogen-bonded conformations the salt form of the methadols (reference 15). Fig.
/0-4.
for (a) the free base form and (b)
6
5
Portoghese and Williams (16) also observed strong intramolecular hydrogen bonds in a- and ,B-methadols. They found pKa values of 8.15 and 7.85, respectively, for these two compounds and ascribed the difference to the greater stability of the NH-O hydrogen bond in the a-isomer. In addition, it was found that the hydroxyl protons of the methadol free bases were at unusually low field in the proton NMR spectra (a = 8.6 ppm; ,B= 7.9 ppm). These proton resonances had peak widths of 30 and 23 Hz, respectively, consistent with a stronger hydrogen bond in the a-isomer. In a later paper (17), these authors carried out further studies on the methadols. Here they observed less temperature dependence of the OH resonance in the ,B-isomer than in the a-isomer in the proton NMR spectra; they also reported some concentration dependence of the OH absorption in the infrared spectrum of the a- but not the ,B-isomer. These results are consistent with stronger hydrogen bonding in the ,B-isomer. In their study of the isomethadols (7), Portoghese and Williams (18) again found evidence in the infrared spectra for strong intramolecular hydrogen bonding for the free bases in nonpolar solvents. In this series, the a-isomer appeared to be more strongly hydrogen bonded. The proton NMR results supported this conclusion; the temperature effect on the hydroxy proton was greater in the ,B-isomer than in a. In aqueous solution, neither appeared to have strong intramolecular association; the pKa values measured were 7.77 for the a-isomer and 7.76 for the
I
Physical
Chemistry
Studies
of Open-Chain
Analgesics
453
,B-isomer, not much different than those of model compounds incapable of such hydrogen bonding. The absolute configurations of the isomethadols were determined to be 35,5R for (+ )-a-isomethadol and 3R ,5R for (- )-,B-isomethadol. These authors point out that the analgetically active isomers of methadol and isomethadol all have a 35 absolute configuration. Casy and Hassan (19) used optical rotatory dispersion studies to determine the absolute configurations of the (+)- and (- )-normethadols. They found that the more analgetically active isomer of normethadol has the 5 configuration at C-3, just as was found previously for the methadol and isomethadol series. Shefter (12) reported X-ray crystallographic structure determinations of a-methadol and a-methadylacetate. The latter compound exhibited an extended conformation, as has been seen in several other open-chain analgesic structures. However, the alcohol showed a torsion angle between the quaternary carbon and the nitrogen atom of 116', midway between an extended and a folded conformation. In a comparison of six open-chain analgesic structures, Shefter concluded that there was no clear pattern relating conformation to analgesia in these compounds.
. C,
Dextromoramide
Dextromoramide is the dextrorotatory (and analgetically more potent) stereoisomer of compound 8 (20). Crabbe et at. (21) inferred from optical rotatory dispersion spectra that dextromoramide has the same absolute configuration as L-(+ )-5-alanine and, therefore, the same absolute configuration as the more active stereoisomer of isomethadone. .. Bye (22,23) carried out X-ray crystallographic structure determmatlOns of dextromoramide, both as the bitartrate salt and as the free base. In both cases, the aminoethyl side chain was found to have an extended ~onformation. Bye compared the X-ray crystal structures of 10 open-cham analgesics. He reached the same conclusion as did Shefter: there IS no clear conformational preference in the open-chain compounds that can be related to analgesic activity.
D. Propoxyphene 7
Three crystal structure determinations have been carried out on propoxyphene derivatives. Bye (24) used anomalous dispersion effects for chlor-
4\4
to
Physical
Chemistry
and Molecular
Modeling
of Open-Chain
Analgesics
ine to establish the absolute configuration of (+ )-propoxyphene (9) hydrochloride as 2S,3R. This result was consistent with the assignment made by Sullivan et al. (9) on the basis of chemical evidence. In this structure the aminoethyl side chain is in an extended conformation. The extended conformation was also found (25) in the crystal structure of the propoxyphene free base. N-Norpropoxyphene, 10, is a major metabolite of propoxyphene. This compound, too, exhibited an extended aminoethyl conformation in the solid state, as determined by Bye (26).
NMe H
NMe,
.
Chemistry
Studies
of Open-Chain
4\\
Analgesics
Table 9-2 Analgesic Propiram
Activity and Conformational and Related Conpounds"
Preference
R
for
EjZ
-CH(CH,)-CH,-piperidyl (II) -CII,-CH(CH,)-piperidyl -CH(CH,)-CH(CH,)-piperidyl (Ihreo) -CH(CH,)-CH(CH,)-piperidyl (erylhro) -C(CH,h-CH,-piperidyl -CH,-C(CH,h-piperidyl -CHrCHrpiperidyl -CII(CH,)-CH,-3,3-Me,-piperidyl -CH(CH,)-CH,-3-azabicyclo[3.2.0]hepryl
E Z E E E Z Z E E
9.7 Inactive 6.4 5.0 50 Inactive Inactive 0.98 2.34
10 28. " From reference b ED50. subcutaneous,
o~
ON10 11 E.
Physical
Other Open-Chain Analgesics
The dissociation constants (pK, values) were determined by Wollweber (27) for propiram, 11, and several of its analogs. Propiram had a measured pKa of 8.92, and the analogs had values ranging from 8.9 to 9.2. No apparent relationship could be seen between pKa and analgesic activity in this series. Geiger and Wollweber (28) used ultraviolet and NMR spectroscopy to study the conformations of propiram and related compounds listed in Table 10-2. The pyridine ring was found to be perpendicular to the carbonyl group, existing in two possible conformations, as shown in Fig. 10-5. In the E conformation the methylene group of the ketone is over the pyridine ring, while in the Z conformation the carbonyl oxygen occupies this position. The E conformation was identified by a single ultraviolet
in rats.
(UV) maximum at about 260 nm and a methylene chemical shift of 1.7-1.9 ppm in the proton NMR spectrum. The Z conformation sh~wed two UV maxima, one at 267 nm and another at 225-229 nm, and m the proton NMR spectrum the methylene adjacent to the ketone had a ~h~mical shift of 2.1-2.2 ppm. As can be seen in Table 10-2, compounds eXlstmg ~n the E conformation showed good analgesic activity, while compounds eXlStmg m the Z conformation were found to be inactive. The predominance of the E conformation also correlates with the presence of methyl substituent(s) on the carbon adjacent to the amide nitrogen. Beckett and co-workers (29) measured the pKa values of thiambutene, 12, and several related compounds. The presence of sulfur lowered the measured pKa to 7.5-9.0, compared to 9.3 for the diphenyl analog. These workers postulate an intra molecularly associated conformation, as shown in Fig. 10-6. Such an arrangement would stabilize the neutral form and, therefore, lower the pK,. In the I ,2-diphenylethylamine series, compound 13 has been reported to be 0.3-0.5 times as potent as morphine in producing analgesia (30). An NMR analysis of the proton couplings indicated that the phenykclipsed conformation exists to an extent of about 22%. OptIcal rotatory dispersIOn
.'-
~~ Fig. 10-5.
Conformation
of propiram
and related
compounds
(reference
28).
Fig. 10-6.
Proposed conformation of thiambutene (reference 29).
456
10 Physical Chemistry and Molccalar
Modeling of Open-Chain Analgesics
References
studies were used to propose that the analgetically active (- )-isomer has the R absolute configuration.
~NM" 12
II.
c
13
Molecular Modeling of Open Chain Analgesics
Quantum chemical calculations have been carried out on methadone, I, to test the hypothesis that it structurally mimics the fused-ring structure of morphine (31). Conformational energies were computed using the (PClLO) method (32) for the nonprotonated and protonated forms. The minimum energy conformation of the protonated form of methadone exhibits intramolecular hydrogen bonding. However, neither the non protonated nor the protonated minimum energy conformations exhibits much similarity to morphine. This can be seen in Fig. 10-7, which contains the
457
minimum-energy conformer of protonated methadone with its nitrogen atom superimposed on that of morphine (dotted outline). A conformational study of methadone and some of its analogs was carried out using molecular mechanics energy calculations (33). The results of the calculations are consistent with experimental findings in the following respects: 1. Isomethadone was found to be less flexible than methadone due to the proximity of the S-methyl group to the phenyl rings. 2. Methadone has a greater propensity to form an intramolecular hydrogen bond than does isomethadone. 3. The SS ,6R-S-methylmethadone isomer (4) can adopt unusual eclipsed conformations due to the methyl groups on C-S and C-6. The N-methyl groups of methadone and isomethadone have markedly different conformational preferences. These conformational differences are postulated to result in each molecule interacting differently with the opiate receptor, which, in turn, may be responsible for the different modes of interaction observed for compounds in this class.
References 1. 2. 3. 4. 5. 6. 7.
'. Fig. 10.7. Minimum-energy (reference 31).
conformer of methadone superimposed
on that of morphine
8. 9. 10. tl. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
A. H. Beckett, J. Pharm. Pharmacol. 8, 848 (1956). L. L. Smith, J. Pha.rn. Sci. 55, t01 (1966). A. F. Casy. J. Chern. Soc. B p. 1157 (1966). R. Haller and H. J. Schneider, Arch. Pharm. (Weinheim, Ger.) 306, 450 (1973). J. G. Henkel, K. H. Bell, and P. S. Portoghese, J. Med. Chern. 17, 124 (1974). J. G. Henkel, E. P. Berg, and P. S. Portoghese, J. Med. Chern. 19, 1308 (1976). P. S. Portoghese, J. H. Poupaert, D. L. Larson, W. C. Groutas, G. D. Meitzner, D. C. Swenson, G. D. Smith, and W. L. Duax, J. Med. Chern. 25, 684 (1982). W. L. Duax, G. D. Smith, 1. F. Griffin, and P. S. Portoghese, Science 220, 417 (1983). A. W. Hanson and F. R. Ahmed, Acta Crystallogr. 11,724 (1958). H. B. Burgi, J. D. Dunitl, and E. Shefter, Nature (London) New BioI. 244, 186 (1973). E. Bye, Acta Chern. Scand., Sa. B 828, 5 (1974). E. Shelter. J. Med. Chern. 17, t037 (1974). E. Bye, Acta Chem. Scand., Ser. B 830, 323 (1976). J. J. Kaufman, N. M. Semo, and W. S. Koski, J. Med. Chem. 18,647 (1975). A. F. Casy and M. M. A. Hassan, Can. J. Chem. 47, 1587 (1969). P. S. Portoghese and D. A. Williams, J. Pharm. Sci. 55, 990 (1966). P. S. portoghese and D. A. Williams, J. Med. Chem. 12, 839 (1969). P. S. Portoghese and D. A. Williams, J. Med. Chem. 13, 626 (1970). A. F. Casy and M. M. A. Hassan, J. Med. Chern. 11, 601 (1968). P. A. J. Janssen and A. H. Jageneau, J. Pharrn. Pharmacol. 9, 381 (1957). P. Crabbe, P. Demoen, and P. Janssen, Bull. Soc. Chim. Fr. p. 2855 (1965). E. Bye, Acta Chem. Scand., Ser. B 829, 22 (1974).
458
10 Physical Chemistry and Molecular Modeling of Open-Chain
A~algesics
23. E. Bye, Acta Chern. Scand., Ser. B 830, 95 (1976). 23a. H. R. Sullivan, J. R. Beck, and A. Pohland, J. Org. Chern. 28,2381 (1963). 24. E. Bye, Acta Chern. Scand. 27, 3403 (1973). 25. E. Bye, Acta Chern. Scand., Ser. B 829, 556 (1975). 26. E. Bye, Acta Chern. Scand., Ser. B 831, 157 (1977). 27. H. Wollweber, Eur. J. Med. Chern. 17, 125 (1982). 28. W. Geiger and H. Wollweber, Eur. J. Med. Chern. 17,207 (1982). 29. A. H. Beckett, A. F. Casy, N. J. Harper, and P. M. Phillips, J. Pharrn. Pharrnacal. 8, 860 (1956). 30. T. Sasaki, K. Kanematsu, Y. Tsuzuki, and K. Tanaka, J. Med. Chern. 9, 847 (1966). 31. G. H. Loew, D. S. Berkowitz, and R. C. Newth, J. Med. Chern. 19, 863 (1976). 32. S. Diner, J. P. Malrieu, F. Jordan, and M. Gilliert. Theor. Chirn. Acta 15, 100(1969). 33. M. Froimowitz, J. Med. Chern. 25, 689 (1982).
11. Enkephalins I. Introduction..
..
II. Opioid Peptide Precursors. A. Proopiomelanocortin B. Proenkephalin . C. Prodynorphin . 111.Peptide Synthesis ...... A. Amino
Acid Protecting
. .
Groups
B. Methods of Amide Bond Synthesis . .. .. ... IV. Enkephalin Selectivit~es for the 1-1.and S Opiate Receptors V. Minimum Enkephalin Chain Length Necessary for Analgesia. VI. Structure-Activity Relationships in the Enkephalins A. The Enkephalin N-Terminus.. . B. Tyrosinel Structure-Activity Relationships. C. Glycine2 Structure-Activity Relationships . D. GlycineJ Structure-Activity Relationships . . . E. Phenylalanine4 Structure-Activity Relationships . . . F. Methionine5jLeucinc5'Structure-Activity Relationships G.
The
Enkephalin
C-Terminus
.
.
.
.
.
H. Enkephalin-Based Opioid Antagonists. . VII. Clinically Investigated Enkephalin Analgesics VIII. The Chemical Anatomy of the Enkephalins References .
I.
459 463 463 465 468 471 471 471 473 481 482 482 484 487 490 491 496 499 499 500 502 503
Introduction
While opium has been known for millennia and morphine for almost two centuries, the discovery and characterization of the endogenous opioid peptides have occurred only within the last decade. The discovery of these endogenous opioid peplides was based on two main approaches that occurred after it was discovered that electrical stimulation of the periaqueductal gray region in rat brains produced naloxone-reversible analgesia (1,2). In the first, the ability of various brain extracts to mimic morphine's effect on mouse vas deferens and guinea pig ilium was investigated (3). After demonstration of naloxone reversibility, purification of the active factor commenced (3). A second approach, utilizing biochemical techniques, resulted in the identification of specific opiatebinding sites in brain tissue (4). Subsequently, three independent research 4\9
460
11
I
Enkephalins
groups reported their discovery of stereospecific opiate-binding sites within the central nervous system, the opiate receptor (5-7). Taken together, the findings of naloxone-reversible analgesia after electrical stimulation and the presence of opiate receptors in the brain allowed the conclusion that an endogenous opiate ligand must exist. In 1975, several research groups demonstrated that pituitary and brain extracts contained compounds with opiate-like activity (3,8-10). Further characterization by Hughes and Kosterlitz resulted in the successful identification of two pentapeptides: methionine-enkephalin (1) and leucine-enkephalin (2) (11). The isolation of these two pentapeptides, which differ only in their C-terminal amino acids, touched off an explosion in peptide-related research that has continued to this day and has resulted in the identification of various classes of opioid peptides. Morley originally divided the opioid peptides into five categories (12,13). Now, however, the opioid peptides can be divided into endogenous and nonendogenous groups that can be further subdivided (Fig. 11-1): A. Endogenous peptides 1. The two pentapeptides: [Met]enkephalin (1) and [Leu]enkephalin (2). 2. Peptides that arise, or are postulated to arise, from biosynthetic enkephalin precursors. This type includes peptides arising from adrenal proenkephalin: [Met]enkephalin-Arg6-Phe7 (3) (14), peptide E (4) (15), [Met]enkephalin-Arg6-Gly7-Leu8 (5) (16), dynorphin (6) (17), ex-and ~-neoendorphin (7) (18), and PH-8P (8) (19). 3. ~-Endorphin (9) (20,21) and the related ex- (10), y- (11), and 8endorphin (12) (22-26). The terms endorphin and opioid peptide were used synonymously in early publications, and confusion resulted. Endorphins refer to those peptides which arise from ~-lipotropin (27). B. Exogenous peptides 4. Pronase-resistant peptides. ~-Casomorphin-5 (13) and -7 (14) are derived from amino acids 60-66 of the milk protein ~-casein (28). 5. Dermorphin (15) and derived peptides. This potent antinociceptive heptapeptide was isolated from the skin of South American frogs of the Phyllomedusa species (29-31). 6. Various other peptides whose opiate-like properties do not arise from a direct receptor interaction. For example, the dipeptide kyotorphin (16) appears to act by releasing en kephalin (32).
Introduction
461 Tyr-Gly.Gly.Phe-Met I ([Met)enkephalin)
Tyr-Gly-Gly-Phe-Leu 2 ([Leu]enkephalin)
Tyr-Gly-Gly-Phe-Met-Arg-Phe 3 ([MetJenkephalin-Arg6.Phe7) I
5
Tyr-Gly-Gly-
"
Tyr-Gly-Gly-Phe-Met-Arg6 -Gly? _Leu") 5 ([Met]enkephalin-Arg6-Gly7-Leu8) W
Phe-Met-Arg-Arg-
~
V al-GIY-Arg-Pro-Glu-
Trp- Trp-Mct
-Asp- Tyr-
20 25 Gln-Lys-Arg- Tyr-Gly-Gly- Phe-Le~ 4 (Peptide E) 1 5 W 15 17 Tyr-Gly-Oly-Phe-Leu-Arg-Arg-Ile-Arg-Pro-Lys-Leu-Lys-Trp-Asp-Asn-Gln 6 (Dynorphin) 1
5
9
Tyr-Gly-Gly-PheLeu-Arg-L ys- Tyr- Pro 7 (I3-Neoendorphin) I 5 Tyr-Gly-Gly.Phe-Leu-Arg-Arg8 I
(PH.8P)
5
Tyr-Gly-Gly.Phe-Met
10 -Thr-Ser-Glu-
15
L ys-Ser.Gln.
20 Phe- Lys-Asn-A
8 lie
Thr. Pro- Leu. Val. Thr-Leu-
25 la-lie-
9 10 11 12
lie. Lys. Asn.
30
31
Ala- Tyr-Lys-Lys-Gly-Glu
(J3-Endorphin) (a-Endorphin, AA 1-16) (y-Endorphin. AA 1-17) (S.Endorphin. AA 1-27)
1
5
7
Tyr-Pro- Phe- Pro-Gly- Pro-lie tJ ({:f-Casomorphin-5, AA 1-5) 14 ({:f-Casomorphin-7) 1 5 7 Tyr- D-Ala-Phe-GlyTyr- Pro-Ser.NH2 IS (Dermorphin)
Tyr-Arg 16 (Kyotorphin) 1
5
10
Tyr-Gly-Gly.Phe-Leu-Arg-Lys17 (a-Neoendorphin) 1 Tyr-Gly-Gly-
5 Phe- Leu-Arg.Arg-Gln18 Fig. II-I.
Tyr- Pro- Lys
10 13 Phe-Lys- Val- V al- Thr
(Rimorphin) Opioid
peptides
462
11
Enkephalins
The various types of endogenous opioid peptides available from the three types of biogenetic precursors are proopiomelanocortin (POMC), proenkephalin, and prodynorphin; these will be discussed in the section on en kephalin biosynthesis. Except for the following brief description of the biological effects of the exogenous opioids, they will not be discussed further unless portions of the molecules were incorporated into the enkephalins. The casein-derived opioid l3-casomorphin-7 (14) has only 4% of the activity of [Met]enkephalin (I) in isolated guinea pig ileum (GPI) and 3% of the activity of morphine by intracerebroventricular (icv) administration (33). The dermorphins possess a unique D-alanine substitution, hitherto found only in bacteria. Despite their recent discovery, an extensive structure-activity profile has emerged. Dermorphin (15) itself is 40 times as potent as [Met]enkephalin (I) in the isolated GPI and 1000 times as potent as morphine in the rat hot plate test by icv administration (34). The structure-activity results on a large number of derivatives have indicated the following (35,36): 1. The N-terminal tetrapeptide is the minimum sequence required for opioid activity. 2. The three amino acid residues of the N-terminal are most important for activity. 3. Substitution of Gly4 retains activity. 4. O-Methylation of Tyr1 retains activity (37), while O-sulfation destroys it (38). 5. Substitution of Tyr' and/or modification of the C-terminal are readily tolerated. The various classes of endogenous opioids have differing selectivities for
the J.I.,13,and Kreceptors. Derived pep tides and many synthetic analogs, some containing unnatural amino acids, have furnished ligands with very high degrees of receptor-subtype selectivity, so that the biological properties of these subtypes can be more readily investigated. The minimum length necessary for anti nociceptive activity in the enkephalins has been investigated, as well as the effects of this truncation on receptor, particularly J.I. and 13,selectivity. ' The structure-activity relationships among the individual amino acids in en kephalin have been resolved. However, the effect of multiple changes in enkephalins does not lend itself to the type of structure-activity reporting employed for the rigid opiates. As a result, except for single amino acid changes, the discussion of synthetic enkephalins with multiple changes will be restricted to those that are biologically and especially clinically interesting.
II
II.
Opioid
Peptide
Precursors
463
Opioid Peptide Precursors
The biosynthesis of the opioid peptides illustrates what appears to be a trend in neurobiology. Like many other central nervous system (CNS) peptides, the opioid peptides are not synthesized individually. Instead, a single gene codes for a large inactive polypeptide that contains within its structure the sequences of several small active molecules that are subsequently split from the precursor by processing enzymes. Often the individual active peptides perform different functions. Such an arrangement confers a great deal of flexibility on the CNS. It may help to coordinate the separate actions that combine to produce complex behavior. Moreover, if the sites where processing occurs vary from tissue to tissue, it will be possible to generate different combinations of peptides from a single gene product. These possibilities have proved true for at least one of three opioid polypeptide precursors. Three separate genes code for the opioid peptide precursors. The first of the large precursors identified was proopiomelanocortin, which contains a wide variety of other biologically active peptide sequences. Subsequently, both proenkephalin and prodynorphin were identified. The biosynthesis of opioids has been extensively reviewed (39-45). A.
Proopiomelanocortin
Soon after the discovery of the two enkephalin peptides, the [Met]enkephalin sequence was found to be present as amino acid residues 61-65 of a large 11,000-dalton polypeptide, l3-lipotropin (I3-LPH) (46), and a 3000dalton fragment of I3-LPH, l3-endorphin (21). These larger peptides were present in appreciable amounts in mammalian pituitary glands (47,48) and were derived from a still larger polypeptide of about 31,000 daltons, proopiomelanocortin (or proopiocortin) (49-53). The gene coding for proopiomelanocortin (POMC) was rapidly cloned and sequenced, and the amino acid sequence is shown in Fig. II-2. (54). The complete POMC molecule contains the amino acid sequences of seven distinct biologically active peptides: l3-endorphin (residues 235265), l3-lipotropin (residues 173-265), adrenocorticotropic hormone (ACTH) (residues 132-170), a-melanocyte-stimulating hormone (a-MSH) (residues 132-144), I3-MSH (residues 215-232), y-MSH (residues 77-88), and corticotropin-like intermediate lobe peptides (CLIP) (residues 149170). The individual peptides in POMC are bounded on both pairs of basic amino acids, combinations of lysine and/or arginine, whose occurrence at strategic sites of cleavage is typical of prohormones (55).
II
CI
Opioid
Peptide
Precursors
465
POMC is made in the anterior and intermediate lobes of the pituitary gland, as well as in several other areas. Processing of POMC is different in the anterior lobe of the pituitary of rodents than in the intermediary lobe. In the anterior lobe, the initial product is cleaved to produce ACTH and ,B-lipotropin. The intermediate lobe also produces ACTH and ,B-LPH, but these are immediately cleaved to produce a-MSH and CLIP, and ,13endorphin and y-LPH, respectively (43). The derived ,B-endorphin has reasonably high affinity for the /L (2.05-
nm), jj (2.36-nm), and
K
(67-nm) opiate receptors (56). By icv administra-
tion, ,B-endorphin has ISO times the potency of morphine as an analgesic (57). Apparently, the entire ,B-endorphin molecule is necessary for analgesia, since deletion of only three amino acids from the C-terminal results in 94% loss of the analgesic potency of ,B-endorphin (58). The [Metjenkephalin moiety in ,B-endorphin is critical for opiate activity, since removal of it, or even of the tyrosine I , eliminates analgesia (58,59). Several other classical amino acid substitutions in the ,B-endorphin amino acid sequence have been reported, but without very significant positive changes in analgesic potency (59). The presence of a free amino group on the N-terminus is critical to activity. N-Acetylated derivatives of ,B-endorphin are without activity (60). Amino acid extensions from the N-terminus also abolish acitivity, indicating that ,B-lipotropin does not have any direct activity at opiate receptors (61). Initially, the presence of the [Met]enkephalin sequence in POMC indicated that [Met]enkephalin was derived from POMC via processing through ,B-lipotropin and ,B-endorphin. However, a series of studies demonstrated that it was doubtful that ,B-endorphin served as a precursor of en kephalin in the brain. First of all, brain levels of ,B-endorphin are only approximately 5-10% those of the enkephalins (62). Also, the regional distributions of ,B-endorphin and the enkephalins differ substantially (62,63). It was necessary to look for additional enkephalin precursors that were structured to be processed to [Leu ]enkephalin, since POMC does not contain this sequence. Furthermore, although the ,B-endorphin sequence is preceded by a Lys-Arg enzyme cleavage sequence, the [Met]enkephalin sequence is not ended by a recognized signal for proteolytic processing. Thus, the ,B-endorphin sequence is programmed to be processed out essentially intact from POMC and does not serve as a precursor to [Met]enkephalin. As a result, it was necessary to determine the precursors and biosynthetic pathways leading to the enkephalins. B.
Proenkephalin Since the question of the nature of the precursor(s) to [Met] and [Leu ]enkephalin had not been resolved with the characterization of
466
11
Enkephalins
POMC, the pituitary gland and the brain were examined for opioid precursors using modern biochemical techniques (42,64). The peptide precursors were present, however, in such low concentrations that a substantial isolation effort would have been necessary. Fortunately, at this same time, bovine adrenal medulla was found to contain high concentrations of opioid peptides (65). Among the many peptides isolated from this source were enkephalins I and 2, C.terminal extended derivatives 3 and 5, and peptide E (4), as well as a large number of longer opioid-containing fragments. Peptide 3, containing C-terminus Arg6_Phe7, was considered to be the terminating carboxyl end peptide sequence in the parent polypeptide, since this dipeptide is not a processing signal for cleavage (66). Simultaneously with the work on peptide isolation, recombinant DNA techniques were being applied to characterize the enkephalin precursor. The complete structure of the enkephalin precursor, proenkephalin, was simultaneously announced by two independent research groups (Fig. 11-3) (54c,67,68). Human proenkephalin has also been sequenced (69). Tbe degree of homology between the two is very high, particularly at the protein level. Each contain six [Met ]enkephalin sequences and one [Leu]enkephalin sequence, as well as peptide E (4) and the C-terminal extended enkephalins 3 and 5. These peptides, except for peptide E, are normal constituents of both adrenal chromaffin cells and brain tissue (70). No peptides with other types of biological activity have yet been identified in the proenkephalin molecule, as was found in POMC. The receptor affinities for the various opiate receptor subtypes have been determined. Initial results for [Met]- and [Leu]enkephalins, using GPI (p.) and mouse vas deferens (MVD) (~), indicated that both were ~ selective, with [Leu]enkephalin being more ~ selective (71). A more sophisticated analysis using receptor-subtype selective ligands yielded the
affinities for the p., ~, and
K
receptors, as well as their relative affinity
(Table 11-1). The earlier results were confirmed with [Leu]- and [Met]enkephalin being ~ selective, without any K affinity (71,72). Extending the C-terminal by two (3) or three (5) amino acid residues yields enkephalins that are effectively equipotent at both the p. and ~ receptors and possess some K receptor affinity (73,74). In this respect, they are very similar to the 31 amino acid peptide j3-endorphin (9) in their receptorsubtype profiles (56). The analgesic activity of [Met]- and [Leu]enkephalin, by icv administration jn a variety of antjnociceptjvetests, is either nonexistent or very weak and transitory (75). Similarly weak and transitory analgesia has been observed for 3 (76). A time study of the opioid effects on mouse tail flick analgesia confirmed the lack of activity of enkephalins I and 2, while the C-terminal extended enkephalins 3 and 5 have 10 and 5% of the analgesic
o~
r::i:<
"
.0.. c ;.,
1. ~
1.'" 3
>, [3
" "> [3 " >, " -,'
[3 >, [3
- <:~;, 8
;;:
" [" 3 q" "> [" 3 ""6 [3
...J
~
i='
468
11
Enkephalins
Table 11-1 Opiate
o~
Receptor
Subtype
Peptide [Met]enkephalin [Leu)enkephalin Arg'-Phe' (3) [Met]enkephalinArg6_Gly7_Leu8
for Proenkephalin-Derived
Receptor
Affinity" (nm)
" (I) (2)
(Met]enkephalin-
fJ-Endorphin
Affinities
9.5 19 27
S 0.9 1.2 29
K
4440 8210 108
Opioids
Relative
=' .., ~6;:~~..2~68";j NO ~o..5 J,N~ ;;... 1......9 c ~>., >. _ ..J >., :E ....J ~~':"~>.1C>6-7 '" ~Y5~b~~ >. ~Eb-;JC>6.< ::s ~'" C1 9 6 -=
"
S
K
Reference
0.09 0.06 0.47
0.91 0.94 0.42
0 0 0.11
71 72 73
6.6
4.8
79
0.41
0.56
0.03
74
2.1
2.4
67
0.52
0.46
0.02
56
a Receptor subtype ligands: J.L,(D-Ala2,MePhe4,Gly(ol)5]enkephalin (DAGO); S, ioAla2,D-Leu5]enkephalin (DADLE); K, bremazocine (see Chapter 6, compound 177).
(:I
e
-<
~b 9
~0;- -7
j
"
;::~>.-7<~< ~6 -= ";j~=uo':"6
>'~..J>"2V:9u v J, ~=: .:.. -7 -7=cXB=''1>. ~;::~b-::~~6 v u
y o!:!
0
..J
b
>.
;;..
,
v
,
0,. .:.. v
..c
' ~c':6o,.~~i3 co':2 <,-J; '7 1
potency of morphine. However, the activity peaked rapidly and was gone within minutes (77), Peptide E (4) was four times as potent as morphine, and this activity was present for at least 30 minutes (77), The lack of analgesic activity of enkephalins is due to their rapid enzymatic degradation in the brain (78). Prodynorphin
Determination of the structure of POMC and pro en kephalin did not account for two other [Leu]enkephalin-containing peptides that were known to be present in porcine pituitary glands and other tissues: a-neoendorphin (17) ({3-neoendorphin-LyslO) (18) and dynorphin (6) (59,79), Thus, at least one additional gene product had to be involved to explain their occurrence. Research into the origins of these enkephalins led to the discovery of a third [Leu]enkephalin-containing peptide, a tridecapeptide, rimorphin or dynorphin B (18) (80), At the same time, a 32-amino acid peptide, containing dynorphin at the amino end and the tridecapeptide at the carboxyl end, was isolated from pig pituitary (81), The common genetic origin of all of these [Leu]enkephalin-containing peptides was convincingly demonstrated by cloning experiments (82), Sequencing of the cloned cDNA from porcine hypothalamus showed a precursor that contains the complete sequence of dynorphin and aneoendorphin but contains no [Met]enkephalin sequences (Fig, 11-4), The precursor polypeptide was originally termed proenkephalin B (82), although pronorphin has also been used (80), The most commonly used and least confusing name is, however, prodynorphin (81), Analysis of the cDNA structure indicates that prodynorphin consists of 256 amino acids,
<
~...,.= 6
b-O:::;~~~_<
I~2~~;;6.:..
0,.
b
>.
>. ..J
-7
~c;
='
0
::I
:;;
i;
;:I
>.9=~j.:2~ 6&-~~~<~
0~N , U N
j
6
12Y-1~1~ ....!.
>. '"
...
, ;:I
~. c::
Q., . "-
Oil
0 0':
.:..
c
.;,
:E >. <
6..
=.2~=~ ~0
:, ::, ~6 i3 ...J
6
:( -<
~i!J:i:b ~on <: <:, ~, 0, , ~~~!22
0 t:'
(3~, ... I\)
, 0': c ......
3...,
::s
V2
Y ;:I
f1
Y ..... 1 0 ;:I
1
;:I
,
....J ~v
;:I
' ~(3 ,
... IU
~~6~~ Y
~>. -;; >7 .cCo
0: "
::;:
.;"
~0: c 6. :a Co
u
::;: -!
::s
.c
~'" >. "i >.
o~S-.
CoO ~-'"
on " u~c.c0"
6
>.
0. >. 6 >.
"
>.
" ..J
u
<~, Sc 6 '">. 0
.c 0.
~N
~u
-'< c
" "~u ~< 6 .::!.
u
:2 >.
o~~i~io':"i6.5oOG f"'):;: I.) o~~_o~'"
;, U'
>. ..J
" ..J
~6 2
u
0u"
0" ~>. Ntz .5
<
~.c6~t;<2(,/j,
......I.)
,
.;"
c1>.-7~b1 i3..56..2~O~ 6..9c91~-:: of-. , -<":I
-'< 0 .J; 0c~u
6
.:..
:E
~'" ;;u
6
~6
~i; J._<
6~~~~~> .3~5:~E1= ::, ..J ,
::, j
. ;:I
<6." 0.u, <6. ..J,
<~~
.c>.
<
.c
'
~-7
0
-;;
6"
~-< -<
.~ .c
>. > .;Co ~..J 6.
~(5
-7
....J ~.
~i
~b -<
i3
-
C.
u
o~
Affinity
(5)
(9)
-
~0
~N1
and I3-Endorphin
1 N,
-;; -;; ~>.
'u_ u
~~8.~
~0 '" " cu~0.;"
< < " ~~if", ~c
-! 'u'" >. .c 0 ..J
~~f, 0."
C
.;"
" < ..J " 6. i>.'" i'">. <:, <<.;" " ~~...J >.
6 .;"
'8
.c
~f-
-!
j~~~== 6 ~~-;; :t >
~.i: ~~i: <: <: <: 0. f- 6"
>.
6
~,,; >. o<,..J
ti::~ ..J
11
470
Enkephalins
including a putative signal sequence of 20 amino acids. Prodynorphin contains three repeating [Leu ]enkephalin sequences, two of which represent the N-termini of the neoendorphin (residues 175-183) and dynorphin (residues 209-225) sequences that are flanked by Lys-Arg couplets. The third enkephalin sequence, leumorphin (residues 228-256), is connected by paired arginine residues to a 22 amino acid C-terminal sequence (see Fig. 11-4). There is remarkable homology between the two enkephalin precursors, proenkephalin and prodynorphin, within the N-terminal region. Both possess shared amino acid residues at equivalent positions, and both possess six cysteine residues at almost identical positions following the hydrophobic signal sequence. Disulfide bond formation and protein folding probably play an important role in the processing of these polypeptides. While processing of POMC and proenkephalin produce peptides that are analgesic, similar processing of prodynorphin does not produce any known analgesic peptides (78). For instance, a-neoendorphin (17), dynorphin (6), dynorphin 1-8, and rimorphin (18) are all inactive in the mouse tail flick assay after icv administration (78,83). The opiate receptor-subtype affinities for prodynorphin-derived peptides are presented in Table 11-2 (84,85). In the prodynorphin series, the outstanding observation is the high affinity for the K binding receptor subtype for the five endogenous peptides. That dynorphin (6) was a selective endogenous ligand for the K receptor was established by bioassay as well as binding studies (86). Both dynorphin (6) and rimorphin (18) have very high affinity for the K site, whereas that of a-neoendorphin (17) is not quite as high. Dynorphin 1-8 and /3-neoendorphin (7) have the
lowest
K
affinity of the five peptides. At present, it is not possible to assess
the physiological significance of these peptides, since, in spite of their high {j K selectivity, they still have significant affinities for the J.I.and receptor sites (85). Table 11-2 Opiate Receptor Subtype Affinities for Prodynorphin-Derived Opioids Receptor Affinity (om) Peptide
"
a-Neoendorphin
(17)
(3-Neoendorphin
(7)
Dynorphin (6) Dynorphin 1-8 Rimorphin (18)
1.24 6.9 0.73 3.8 0.68
5
K
0.57 2.1 2.4 5.0 2.9
0.20 1.2 0.12 1.3 0.12
Relative
" 0.10 0.10 0.13 0.22 0.14
Affinity
5
K
0.23 0.33 0.04 0.16 0.03
0.67 0.57 0.83 0.62 0.83
Reference 84 73 84 84 73
III
Peptide Synthesis
III. A.
t
471
Peptide Synthesis Amino Acid Protecting Groups
A peptide is the formal result of the condensation of amino acids with the elimination of water. If the reaction is to give a single product, there has to be only one free amino group and one free carboxyl group, necessltatlOg protectIOn of the functIonal groups that are not involved in amide bond formation. The situation is complicated by the presence of reactive functional groups in the amino acid side chains that also require protectIve blocklOg groups. Thus, two types of protective groups are needed: one type that intermittently protects the a-amino and carboxyl groups and liberates them again selectively, and a second type that blocks the other functional groups throughout the synthesis. For a protecting group to be useful, it must be (a) easily introduced, in high yield, into amino acids and peptides, (b) completely stable during peptide bond formation, and (c) quantitatively removable without deleterious effects on the peptide. Although many protective groups have been described, only a few have found general and extensive use. Since most peptides are reasonably stable in acidic media, combinations of protective groups havlOg graded acid labilities are preferred. Only a few other methods of cleavage are used, notably reductive ones. An in-depth appraisal of protecting groups for amines (87), carboxyl (88), sulfhydryl (89), and hydroxyl (90) residues in peptide synthesis has appeared. Dual functional protecting groups (91) and differential protection and selective deprotection (92,93) have been reviewed, as well as peptide synthesis with minimal protection of the side chain functions (94). The preceding subjects have also been reviewed in Hoube'n-Weyl (95). B.
Methods of Amide Bond Synthesis
1. Solution Methods The high demand for efficiency in peptide amide bond formation has restricted the number of generally useful methods. To be useful, a good method must provide (a) a very low to nonexistent rate of racemization (96), (b) minimal side product formation, (c) facile workup, (d) high to quantitative yields, and (e) reasonable rates of reaction. The most important requirement is the absence of racemization because the biological activity of the peptides is invariably dependent on their configurational integrity, and removal of unwanted isomers can be extremely difficult (97). No known method meets all these requirements, and so far only four methods have been found to be sufficiently effective for general synthetic use. The methods are (a) the azide method, (b) mixed anhydrides, (c) dicyclohexylcarbodiimide condensations, and (d) active esters.
472
11
Enkephalins
IV
Enkephalin
Selectivities
for the IL and 8 Opiate
Receptors
473
In the azide method (98), an activated carboxyl component is prepared in two stages: the formation of N-protected amino acid or peptide hydrazides followed by conversion to the aZIde pnor to couphng wIth a carboxyl protected amino acid. Azide couphngs were ongmally consIdered to be free of racemization, but substantial racemizatIOncan occur particularly with larger peptide fragments (96,~9.1. The second method, mixed carbonic carboxylic anhydrides (100), utthtzes the reactIOn of an N-protected amino acid with a chlorocarbonate ester in the presence of a weak tertiary base. The mixed anhydride thus formed IS treated m situ with the amine component. The method is valued for ItS rapId reactIOns and facile workup. In the dicyclohexylcarbodiimide method, th~ carboxylic acid is activated by addition across one of the carbodnmlde double bonds and in situ reaction with an amine component (101). ThIs method is used extensively in solid phase synthesis. In the active ester approach, an ester of the N-protected amino. acid is formed with a hydroxy lie function that also serves as a facIle leavmg group upon reaction with an amine. This method IS dlstmgUlshed by the large vanety of esters that have been found suitable (102). The potential for the occurrence of side reactions in the various methods of peptide synthesis has been reviewed (103). . . All of the preceding reactions are used in classical solutIOn peptIde synthesis using two approaches (104). The first IS mcremental cham elongation wherein one amino acid at a time is added startmg from the carboxyl-terminal end. The second is a convergent fragment approach where-in sequential fragments of the peptide are formed and ulttmately assembled using conventional techniques. Fragment condensatIon allows greater flexibility in the choice of protecting groups. The homogeneIty of intermediate fragments can also be ascertamed. The drawbacks of fragment condensation are the potential for low solubility of the larger fragments, slow coupling rates, and increased potential for racemization and intramolecular reactions. 2 The Solid Phase Method A major problem with peptide synthesis is that it is very labor intensive, primarily in the workup and purification of the intermediates. A partial solution to this problem IS the sohd phase method, which is based on the incremental addition of amino acids to a growing chain, which is utilized extensively in solution techniques. However, in this case, the carboxyl-terminal end of an ammo aCId ISattached to an insoluble support resin. Because the growing peptide is covalently bound to a polymeric support, the product is readily separated from the by-products. Each coupling step of the polymer-supported peptIde cham can be driven to completion through the use of excess soluble reagents, and
mechanical losses are diminished by retaining the peptide-polymer beads in a single reaction vessel. The most attractive features of the solid phase approach are as follows: (a) there is high-speed attachment of several amino acids per day to the peptide chain, (b) insolubility problems that can trouble solution syntheses of larger peptides are avoided, (c) the procedure is easy and convenient to perform, and (d) mechanization and automation substantially reduce the labor commitment. The major drawback to this method is the extensive purification of the final product that is necessary due to the slightly less than quantitative yields in each coupling step. The strength of the solid phase method lies in its operational efficiency compared to product quality, which is the strength of conventional solution methods. The solid phase method has been critically reviewed by Merrifield (105).
IV.
Enkephalin Receptors
Selectivities
for the JL and i5 Opiate
At approximately the same time that basic structure of the enkephalins was determined, the existence of various subtypes of opiate receptor was postulated by Martin on the basis of his extensive reseearch on the chronic spinal dog (106). These were designated J.l., K, and u. Subsequently, a fourth subtype, 8, was proposed, which had a relatively high affinity for the enkephalins (cf. Chapter 2) (107). The occurrence of multiple receptor subtypes offers the potential for design of specific compounds targeted for each, with, it is hoped, a specific biological profile resulting. The design of receptor-subtype specific ligands has resulted in some highly selective, but not completely specific, agents. The subtypes important for analgesia are
the J.l.,8, and K receptors. With extended-length enkephalins, dynorphins, and endorphins,
K receptor occupancy becomes important. However, with
the pentapeptide and smaller enkephalins, only J.l. and 8 receptor selectivities usually have to be considered. The following section will therefore concentrate on the effects of peptide chain length and C- and N-terminal substitions on J.l.and 8 opiate receptor-subtype selectivities, followed by a discussion of the various subtype receptor ligands, including cyclic peptides, that have been developed. In the following tables, where potencies are referred to GPI and MVD, GPI represents the potency of the compound in the electrically stimulated guinea pig ileum, a tissue prepara-
tion considered to be rich in
J.l.
receptors. MVD refers to potency in the
mouse vas deferens, a tissue rich in 8 receptors. In both cases, [MetJenkephalin is defined as having a potency of I.
474
11 Enkephalins
IV
Enkephalin
Table 11-3
Table 11-4
The Effect of Enkephalin Chain Length and C-Terminal Substitution on 1J.and 5
The
Receptor
Influence
Selectivities
of Phe4
f(Jr the JL and 5 Opiate
on JL: /) Opiate
Receptor
Receptors
475
Selectivity
Selectivities
Peptide
Peptide
GP[
MVD
GPI/MVD Ratio
References
[LeuJEnk [MctJEnk [D-Ala2,Leu5]Enk [D-Ala2,Met5]Enk
0.36 1 5.5 6.3 I.14 5.6 4.2 5.0 3.3
1.2 1 9.U 5.0 7.2 3.0 2.9 1.6 2.5
0.3 1 0.6 1.2 0.16 1.9 1.4 3.[ 1.3
108 109 I/O III 1/2 1/2.1/3 1/2.1/4 1/1 1/5
1.2 1.2 0.11 25 1.3
0.056 0.054 0.002 1 0.04
2[ 22 55 25 32
1/6 1/6 1/7 1/6 71
0 0.008 7.7 1.34
0 U.0U6 U.U13 U.U3
[3 59 48
1/6 1/6 116 109
19 Tyr-D-Ala-Gly-Phe
[D-Ala2,Lcu5]Enk-OMe [D.Ala2,Met5]Enk-OMe [1)-Ala2,Leu5]Enk-NH2
[D_Ala1,Mct5]Enk. NH,(DAME) Tyr-D-Ala- Phc-Mct -NH2 Tyr-n-Ala- Phc-Lcu-NH2 Tyr-D-Ala-Phe-NH2 Tyr'D-Ala-GlyNH(CH,j,C,H, Tyr-Phe-Met-NHz Tyr'D-Ala-NH( CHzhC(,lls Tyr-n-Ala-NII( CI12hC6Hs Morphine
Table 11-3 demonstrates the effect of both C-terminal substitution and amino acid chain leng[h in IJ. and 8 receptor selectivities. In Table 11-3 a GPI/MVD ratio of less than I indicates 8 selectivity and a ratio of greater than I, IJ. selectivity. All responses are related to [Met]enkephalin. [Leu ]enkephalin is more 8 specific than [Met ]enkephalin. Replacement of Gly2 with D-Ala2 in bo[h enkephalins resulIs essentially in re[en[ion of the same rela[ionships. However, comparison of the resulIs from two different laboratories for [D-Ala2,Met5]enk demonstrates both the uncertainties and dangers in comparing data from different investigators. Since D-Ala2 is a very common substitution, the comparisons in Table 11-3 are based on enkephalins with this substitu[ion. However, incorporation of lipophilic D-amino acids in the second position increases IJ. recep[or selectivity, apparently [hrough an auxiliary binding site (71 ,110). Conversion of the C-[erminal carboxylic acid of [he enkephalins to ei[her ester or amide results in an increase in IJ. specificity, especially for the amides. Deletion of Gly3 to form the tetrapep[ide amides analogous [0 the dermorphins and casomorphins strongly increases IJ. specifici[y. In the [ripeptides, where the C-[erminal is subs[i[u[ed by a phenylpropylamide, IJ.
20 Tyr-o-Ala-Gly-NH (CH,hC,H, 21 Tyr-o-Ala-G[y-NH-CH(CH,)CH,CI[(CH,h 22 Tyr-o-Mct -Gly-NH-CH (CII,)CH,CH( CI I,h '" Relative
to [Met]enk
GP["
MVD"
GPI/MVD Ratio
2.2 4.4 U.32 0.62
0.26 0.18
8.5 24 -32 -62
= 1.
selectivi[y again predominates. In this case, however, the phenylpropylamide may be substi[uting for [he phenylalanine side chain. Deletion of both Gly2 and Gly3 elimina[es any ac[ivity in the tripep[ide Tyr-Phe-MetNHb indicating that [here is a critical distance between the [yrosine and phenylalanine aromatic rings. The dipeptides, con[aining Tyr-D-Ala and with a phenylpropylamide or a phenylethylamide C-terminus, retain surprisingly po[ent activity thaI again is IJ. selec[ive. A dramatic effect on IJ.:{jselectivity involves the influence of Phe4 (Table 11-4). Early work by Kosterlitz (1/8) indica[ed a substan[ial IJ.selectivity when the C-terminal in [Leu]enk is decarboxyla[ed. Independently, others investigated the effect of changing Phe4 to nonaromatic lipophilic groups. The [etrapep[ide (19) is IJ. selective, as is its decarboxylated analog (20). However, a decisive shift towards IJ.receptor specificity is obtained when the phenethyl group (20) is changed to [he strictly aliphatic hydrophobic side chain in 21. This change between 20 and 21 leads to a large decrease in {j receptor recognition. The IJ.-selectivity is fur[her enhanced by [he substitution of D-Me[2 for D;Ala2, perhaps reflecting a better binding a[ the lipophilic auxiliary site (1/9). Since dihydromorphine is a pro[otypic rigid opiate ligand for the IJ. recep[or, it was not surprising that various physical chemical investigations probed the similarities between the rigid opiates and the p.-selective peptides (120). The resulIs of a series of s[udies (7/ ,/21) employing, among others, pep tides 21 and 22, using both NMR investigations and conformational energy minimization techniques, have led to several conclusions: I. Due [0 [heir small size and hydrophobic content, these peptides are able to fir [he conformational space occupied by the morphine alkaloids. 2. The peptides exist in a highly folded conforma[ion, suggested by the poten[ morphinomimetic activi[y of conforma[ionally constrained analogs incorporating a-aminoisobutyric acid (Aib2) (/22).
416
II
IV Enkephalin Selectivities for
Enkephalins
the Jl and 8 Opiate Receptors
411
laboratory investigated l3-casomorphins ranging from four to seven amino acids in length and reported only a fourfold selectivity of 23 for the J1. receptor (129). The reduction of the C-terminal carboxylic acid to the corresponding alcohol, an effect designed to change a hydrophilic group into an essentially lipophilic one, has led to very J1.-selective ligands. For instance, the tetrapeptide (24), syndyphalin or SD-25 (130), possesses a hydrophobic o-amino acid at position 2 and backbone methylation between Gly3 and Phe-oI4, as well as having the usual Phe4 present as phenylalaninol (131). The changes result in a peptide that has parenteral analgesic activity more
3. The folded conformation is within the approximately 5 kcal/mole of energy available to the lowest-energy peptide conformation in solution at room temperature. 4. There is external orientation of the lateral side chain of any o-amino acid at position 2, but not for their L-enantiomers. This allows an additional bonding interaction at a lipophilic binding site on the receptor. 5. There is a good, energetically feasible overlap between the tyrosine and the phenethyl chain of morphine. This obviously also occurs for peptides specific for the 0 receptor. 6. There is similar spatial orientation of the C-6-a-hydroxyl group of morphine and the critically important amide carbonyl group of Gly3 (119,123).
Try-Pro- Phe-Pro-NH2
23 (Morphiceptin) Tyr-D-Ala-Gly-McPhe-Gly-ol 25 (DAGO. glyol)
Circular dichroism studies also confirmed the existence of folded conformations for the J1.,but not the 0, selective peptides (124). In summary, specificity for J1.opiate receptor binding sites is obtained by:
Tyr-o-Met( O)-Gly-MePhe-ol 24 (5yndyphalin. 50-25) Tyr-D-Ala-Gly-MePhe-Met(O)_ol 26 (FK33-824)
potent than that of morphine (132). Syndyphalin (24) is significantly more selective than morphiceptin (23) and displaces tritiated dihydromorphine (ICso, 0.05 nm) much more effectively than DADLE (33) (ICso, 200 nm) (133). The current prototypic opioid ligand for the J1.receptor is the penta peptide 25 (DAGO or glyol), where again the terminal amino acid's carboxyl group, in this case Glys, has been reduced to the alcohol (134). Glyol (25), with a selectivity of lOa for the J1.over the 0 receptor, is thus a better J1.ligand than the rigid opiate normorphine. The effects leading to highly J1.-selective ligands are finely balanced. For instance, retaining the first four amino acids in DAGO (25) and replacing the ethanolamine C-terminus (glycine alcohol) with methionine sulfoxide alcohol (Met(O)-ol) results in the pentapeptide 26 (FK-33-824), a very potent morphine-like analgesic peptide (135). This peptide has in its structure all of the critical components necessary to interact with both J1.and 0 receptors. It exhibits high potency in both the GPI and MVD assays and inhibits equally well both rigid opiate and opioid binding in brain homogenates (1I8). Among the approaches to the preparation of J1.'selective ligands, based on the concept of a folded peptide conformation, is the formation of cyclic enkephalins. Cyclization in en kephalin analogs is particularly powerful way to reduce drastically the conformational degrees of freedom. A successful approach has used the formation of cyclic lactams between basic o-amino acids in position 2 and C-terminal carboxylic acid (120)_ Since in these structures Tyrl is exocyclic, the necessary flexibility in this region is maintained. The initial compound (27) containing a 14-membered ring lactam incorporates 0-2,4-diaminobutyric acid as the bridging ligand. Compared to [Leu]enk (2), 27 is 17 times more potent in the GPI assay and 7 times less potent in the MVD test. Confirming results were obtained J1.
1. Decreasing the number of amino acids in the enkephalin chain. 2. Increasing the lipophilicity at the C-terminal by esterification, amide formation, and especially decarboxylation. 3. Replacement of the aromatic Phe4 residue with a lipophilic aliphatic side chain. 4. Introduction of a hydrophobic o-amino acid as the second en kephalin amino acid. 5. The use of hydrophobic amino acid as the fifth residue to occupy an additional receptor-binding site. This summary, although somewhat oversimplified, explains much of the observed biology. Additionally, the J1.selectivity derives not from an increase in J1.receptor affinity but from a drastic reduction in affinity for the
o receptor. An exception to the above observations is the tetrapeptide morphiceptin (23) (125), which is a derivative of the exogenous opioid l3-casomorphin (126). Morphiceptin (23) is reported to have an affinity for the J1.receptor that is 1000 times greater than that for 0 receptors, hence its name (126). Although the presence of proline as the second and fourth amino acids does involve significant conformational restrictions, replacement of Pro2 by o_Pro2 effectively eliminates biological activity. Morphiceptin has analgesic effects when administered icv (I27), and limited structureactivity relationships have been developed (128). Disparate results from another laboratory (129), however, need to be resolved before morphiceptin can be considered as a J1. receptor opioid ligand. The second
1
II
478
Enkephalins
IV
Table 11-5 Lactam
Formation
H
TYr-~~~-:nN (CH,I,
0
'-N H Compound 27 29 30 31 28'
" 2 1 3 4
H
MVD"
GP!iMVD
10.5 17.5 5.1 51.2 10.6
0.16 U.14 0.02 O.OS
5.8 3.1 9.9 2tiA
=
hydrophilic group at this position could be expected to enhance 8 selectivity. Additionally, since decreasing chain length favors IL selectivity, increasing it slightly may enhance 8 affinity. Empirically, it was found that replacement of Gly2 with the hydrophilic Ser2 and introduction of Thr6 in [LeuJenk furnished a peptide (34) (DSLET) that is 620 times more potent in the MVD assay than in the GPI assay and is about 10 times more specific for 8 receptors than DADLE (33) (71.143). The replacement of Ser' in 34 by Thr2 yields the even more specific deltakephalin (DTLET) (35). Deltakephalin (35) has a 3000-fold separation of activities between the GPI
Ratio
S
1.
= Tyr-D-Nva-Gly-Phc-Lcu-NHz
(Nva
479
(142). Since lipophilic D-amino acids at position 2 enhance IL selectivity, a
GPI"
Relative to (Leu]enk (2) " (28) f>
o
N
Selectivities for the Jl. and 5 Opiate Receptors
disulfides were nonselective toward the IL or 8 receptors (141). However, use of penicillamine (13.I3-dimethylcysteine), which couples the conformational constraints of aminoisobutyric acid and disulfide ring formation, generates highly 8-specific peptides. The peptide usually considered the prototypic 8 receptor ligand is DADLE (33), an analog of [Leu]enk where Gly2 has been replaced by D-Ala2 and Leus is present as its D-Leu' enantiomer (liB). However, DADLE (33) possesses only a threefold preference for the 8 receptor
in Pentaenkephalins
o
Enkephalin
S
I
= norvaline)
I
Tyr-D-Cys-Gly- Phc-n-Cys- N H 2 32
in receptor binding assays (136). Comparison with the corresponding open-chain analog 28 demonstrated that the IL-receptor selectivity of 27 is a direct result of conformational restriction, confirming the previously discussed physicochemical conclusions derived from flexible peptides (137). Some variation in conformational restriction was achieved by varying the number of methylene spacing groups in the basic amino acid2 (13B). All of these analogs showed wreceptor selectivity (Table 11-5) with selectivity increasing with lactam ring size, 13-membered through 16membered. When compared to [Leu]enk (2), the most active analog 31 is 51 times more potent in the GPI assay and 12 times less potent in the MVD assay. The results of further structure-activity relationship studies indicated that these cyclic analogs possess the same configurational requirements in amino acids I, 2, 4, and 5 as the linear enkephalin (139). A second class of cyclic wselective ligands was obtained by substitution of cysteine at positions 2 and 5 and subsequent ring formation through disulfide bond formation. The initial example (32) was a surprisingly strong analgesic by iv administration and possessed substantial respiratory depression properties. It was 100 times more potent than morphine in competing for its receptor (140). A subsequent investigation employing [D-Cys2,D-CyS']- and [D-Cys2, Cys']-enkephalins revealed that these cyclic
Tyr-D-A 33
!a-Gly- Phe-])- Leu (DADLE)
Tyr- Y -Gly-Phe-Leu-Thr
34 35
(Y
~
Scr. DSLET)
(Y = Thr, deltakcphalin,
DTLET)
and MVD assays (144). These results indicate the importance of either the hydrophilicity or lipophilicity of the second amino acid in enkephalin. The importance of Phe4 for 8 selectivity was again underscored by the large d~crease in MVD potency when it was replaced by hexahydrophenylalamne (145). Another striking result is the high degree of 8 specificity encountered when the fifth amino acid is replaced by its D-enantiomer. However, this effect is reversed by the addition of threonine as the sixth amino acid (146). These differences probably reflect the ability of Phe4 to fit into its 8-specific binding site (71). On the basis of both the IL and 8 ligand investigations (71), 8 opiate receptor specificity is engendered by: 1. The presence of an aromatic phenylalanine at position 4.
i
ring at the equivalent
position
of
11 Enkephalins
48(1
2. The presence of a hydrophilic C-terminus that allows the Phe4 aromatic ring to be fitted into its 8-binding site. That is, the amino acids following Phe4 playa key role in determining /j specificity. 3. The presence of a hydrophilic side chain on the second amino acid, such as Thr or Ser, strongly enhances /j specificity. 4. An extended, rather than folded, energetically available conformation of the peptide. While the necessity of extended peptide conformations of enkephalin for /j receptor selectivity has enjoyed considerable support, the emergence of an entire class of cyclic 8-specific peptides indicates that the extended form concept may have to be modified. Although cyclic [o-Cys', 0-, or L-Cys5] enkephalins (e.g., 32) were only slightly JJ.selective (141), replacement of cysteine with penicillamine (Pen, 13,I3-dimethylcysteine) ultimately led to the most 8-specific pep tides currently known. Although the initial experiments were performed on the Pen-containing enkephalin amides (147), the free acid was found to be more selective, in line with the preceding generalizations. Replacement of O-Cys2 in 32 with o-Pen2 yields a highly 8-seleetive analog (36) (Table 11-6) (148), and even higher selectivities were obtained when penicillamine was introduced into positions 2 and 5. The eyclic [o-Pen2,o-Pen'J analog (38), having over a 3000-fold selectivity, is the most selective /j ligand known (149). This /j receptor selectivity probably results from additional conformational rigidity in these cyclic analogs imparted from the gem-dimethyl groups of Pen. An NMR study of both the cyclic cystine and penicillamine analogs indicated similar overall peptide conformations. However, differences in conformation and flexibility were observed at the C-terminal, which may account for the receptor affinity differences in these analogs (150).
b-Selective Ligands Pcntacnkephalins
Chain Length Necessary for Analgesia
Tyr-D-Ala-Gly-Phe-NH
Tyr-X-Gty.Phe-Y.OH GPI
MVD
D-Pen ()-Pen D-Pen D-Pen D-CyS D-CyS
D-CyS Cys D-Pen Pen D-Pen Pen
1350 213 6930 2720 67 40
6.3 0.3 2.2 2.5 0.t3 0.75
I
Tyr-D-Ala-Gly-Phe-NH 42
V.
Minimum Enkephalin Chain Length Necessary for Analgesia
While [MetJenkephalin and [Leu]enkephalin do not demonstrate analgesic activity, due to their rapid inactivation by peptidases, the substitution of Gly2 by o-Ala2 in the enkephalinamides 43 and 44 (Table 11-7) restores analgesic activity to these molecules (145,153). The question then was what was the minimum chain length necessary for analgesia. A series of deletion peptides has been prepared to investigate this point, and
Table 11-' Chain Length Necessary for Analgesia Analgesia"
s I
Y
I
(CH,)"
Peptide
X
481
Another approach, which has also been used in the rigid opiates, is the formation of dimeric enkephalins linked by spacer groups of varying length. Among the dimers of a tetrapeptide, the one with a 12-methylene bridge unit (42) has optimum selectivity and affinity for /j receptors (151,152). The /j selectivity of 43 is approximately four times that of DSLET (34). It has been suggested that the selectivity observed with these dimers is the result of cross-linking of receptors by the two linked individual en kephalin terminals (151).
Based on Cyclic Penicillamine-Containing
s I
36 37 38 39 40 41
Minimum Enkephalin
Minimum Enkephalin
Table 11-6
Compound
V
GPljMVD 2t5 666 3164 t088 515 53
ratio
43 44 4S 46 47 48
Tyr-o-Ala.Gly-Phe.Met.NH, Tyr-D-Ala-Gly-Phe-Leu-NH2 Tyr-o-Ala-Phe-Met-NH2 Tyr-D-Ala-Phc-Leu-NH2 Tyr.o.Mct(O)-Gty-MPA' Tyr-o-Ala-PPAc
49 Tyr-o-Met(O).PPN a Relative to morphine to C
= 1 (AD50 MPA = N-methylphenylethylamide. PP A
= 3-phenylpropylamide.
Tail Flick
Writhing
icy icy
0.5 1.4
0.6 0.5 <0.02 0.05
= 80
nmole/kg).
Mode of Admin.
0.6
sc icy sc sc
References 145,154 145,153 154 154 155,156 116,157 158 158
11 Enkephalins
482
some of the results are presented in Table 11.7. Tetrapeptides 45 and 46 also retain analgesic properties (154). However, while both the tripeptide 47 and the dipeptides 48 and 49 possess potent analgesia by icv administra. tion, they are also active as analgesics by parenteral administration (155-158). However, in pep tides 47-49, the arylalkylamtde groups at the C-terminal appear to be necessary for analgesIc actIvIty. It may be that the aromatic ring in the amide function substitutes for the aromatic ring in phenylalanine in [Met]. and [Leu]enkephalins. In summary, the analgesic properties of the enkephalins parallel the activities fou.nd to be necessary for activity in the GPI and MVD assays, and the mInImum cham length necessary for enkephalin.based analgesia is a dipeptide.
VI.
Structure-Activity
Relationships in the Enkephalins
The development of structure-activity relationships in the enkephalin area began shortly after the disclosure of the structures of the endogenous opioids. Literally thousands of analogs have been prepared, and reviews of this structure-activity relationship have been published (12,159-160). In the majority of,cases, biological activity was primarily expressed in terms of the GPI and MVD bioassays. While these assays indicate the potential for analgesic potency, the effects of transport, metabolism, and tissue distribution are not addressed. Keeping these facts in mind, the structureactivity relationship of the enkephalins will be indicated primarily on the basis of MVD and GPI assays. The approach used will focus on each amino acid in the pentapeptide. C. and N.terminal effects will also be indicated. The structure-activity relationship results will be primarily restricted to the pentapeptides, although tetrapeptides will be included. Additionally, only single replacements will be considered, although common replacements, such as D.Ala' and D.Met', will be considered as equivalent to Gly' in structure-activity relationship discussions. Opioid analgesics related to the casomorphin and dermorphin series will not be discussed; neither will the extensive investigations conducted on the isosteric amide bond replace. ments in the enkephalins (161). Effects of the enkephalin analogs on receptor selectivity have already been discussed in Section IV and will not be differentiated in any discussion of the analgesic effect, although the GPI is rich in I-' receptors and the MVD in S receptors.
A. The Enkephalin N-Terminus The amino group of tyrosine has been modified in a variety of ways: by acylation, alkylation, deamination, and extension by addition of other
VI
Structure-Activity
Relationships
in the Enkephalins
483
Table 11-8 N-Terminal Substitutions
in EnkcphaJins X.Tyr.Gly.Gly.Phe.Y
Compound ] 50 51 52 53 54 55 56 59 60 61 62 63
X H CH) CH., n-C3H7 n-C:,;HH (CH,j, CH,CO (CH,j,CCO,(BOC) Tyr Phe Arg Lys Gly
Y
GP]
MVD
Met Leu Met Met Met Leu Leu Leu Met Met Met Met Met
1.0 0.76 1.0 0.045 0.67 0.08 <0.001 0.002 0.15 0.18 0.17 0.20 0.30
1.0 0.75 0.2 0.(108 0.01 0.02
0.28 0.50 0.22 0.25 0.18
References
112.113.163 111.164 168 168 112.113 113 113 J08 J08 J08 108 J08
amino acids. Removal of the amino group and replacement of Tyr' with 4.hydroxyphenylpropionic acid results in the elimination of biological activity in both [Met]. and [Leu)enkephalin (113,162). N.Methylation of the Tyr' in both [Leu]enkand [Met)enk results in little change in potency in the GPI assay. While the [Leu]enk derivative (50) is equipotent in both the GPI and MVD assays, the corresponding [Metlenk derivative (51) is fourfold less active in the MVD assay than is the GPI (Table 11.8). The esters and amides of [Leu ]enk also show about a 4. to JO.fold reduction in activity in MVD when compared to the GPI results (112,113,163,165). A series of n.alkyl substituted [Met]enk derivatives from ethyl through octyl shows substantial decreases in potency and receptor affinity (168). The series peaks at n.pentyl (53), which has two. thirds the GPI activity of [Met ]enk, but vanishingly small MVD activity. Interestingly N ,N. dimethylation of the amino group (54) in [Leu]enk causes a marked decrease in potency in both assays (112). Substitution by classical rigid opiate antagonist substituents will be considered in Section VI,H. Many derivatives of enkephalins having the N.terminus blocked by acyl groups or carbonates are intermediates in the total synthesis of the parent unblocked molecule. These blocked derivatives 55 and 56 of [Leu]enk are essentially without biological activity, indicating the need of a functioning N.terminal amino group, either mono or unsubstituted, for opioid activity. It has been proposed that the opiate receptor has a region that prefers a cationic species in the area of tyrosine and that this region is restricted to 3.3 A
484
11
Enkephalins
Table 11-9 Guanidine
Substitutions
Tyrosine
Compound
Tyr-D-Met-Gly-Phe-NII, II,N-C-Tyr-D-Met-Gly-Phe-NII,
II
18 83
MVD"
Ratio
0.16 0.80
112 104
Nil "[Met]enk
~ I.
(166). As a result, a series of strongly basic guanidino analogs of enkephalins, exemplified by 58, containing D-Met2 in the tetrapeptide amide, have been prepared with the expectation that the more basic guanidino residue would result in an increase in potency (167). In general, this substitution substantially increases potency in both the GPI and MVD assays, with no general change in selectivity. The guanidino peptide (58) is twice as potent as the amino peptide (57) and 4.4 times as potent as morphine by iv administration in the rat tail flick assay (Table] ]-9) (167). Initial reports on the addition of amino acids to the N-terminal (169), indicated that ArgO, residue 60 of {3-lipotropin, maintained 90% of the GPI activity of (MetJenk. In contrast to this earlier work, two other groups observed that even ArgO substantially lowered GPI potency (108,170). A variety of lipo- and hydrophilic amino acids (59-63), possessing basic and acidic groups, all lower both GPI and MVD potencies without a noticeable change in selectivity (Table] ]-8) (108). Increasing the chain length of the amino acids at the N-terminal causes the GPI potency to decrease rapidly as the chain is extended (169). A series of analogs of [Met ]enk containing (3-alanineo, GABAo, and 5-aminopentanoic acido all lack activity in both the GPI and MVD assays, as well as having minimal opiate receptor affinity (171). On the whole, most substitutions of the N-terminus decrease potency. N-Methylation can maintain, or slightly enhance in some cases, the biological activity. There appears to be very little opportunity for manipulation of this portion of the en kephalin molecule. 8,
Structure-Activity
Relationships
in the Enkephalins
485
Table 11-10
N-Terminal
57 58
VI
Tyrosine' Structure-Activity
Relationships
Of the five amino acid residues in the enkephalins, configurational requirements at tyrosine are the most of the Tyrt residue in both [LeuJ- and [MetJenkephalin activity (113,159,172). Likewise, replacement of Tyrt
the structural and stringent. Deletion abolishes biological by its enantiomer,
Substitutions
X-Gly-Gly-Phe-Y
] 64 65 66 67 68 69 70 7] 72 73 74 75
Y
GPI
MVD
Met Met Met Met Met Met Met Leu Leu Met Met Met Leu
1.0 0.002 0.17 <0.01 <0.01 <0.01 <0.01 0.002
1.0 0.0003 0.004
X
Compound
Tyr Phe Tyr(Me)" p-NII,Phe p-CIPhe p-IPhe 3,4-(OIl),Pheb
(DOPA)
p-OH-phenylglycine
p-homoTyr Gly Ala D-Ala His
" O-Methyltyrosine. b 3,4-Dihydroxyphenylalanine. c 2-Amino-4-(4-hydroxyphenyl)-butyric
0.04
Inactive Inactive Inactive <0.0002
References
162 164,174 159 159 159 159 JJ3 JJ2 159 159 159 JJ3
acid.
D-Tyr', in [Metjenkephalin results in a biologically inactive compound with negligible receptor affinity (112,173). Replacement of the Tyr parahydroxy group by a p-amino (66), chloro (67), or iodo (68), or substitution of Tyr' by DOPA (69) results in a loss of agonist activity (Table 11-10) (159). The stringent structural constraints on Tyrt are demonstrated by the homo (71) and nor (70) tyrosine derivatives, which possess, at most, 4% of the GPI/MVD activity of their parent (112,113). The tyrosine has also been replaced by other amino acids and, as shown in Table 11-10, none of these has demonstrable activity (159). Several rigid analogs that constrain tyrosine in a fixed conformation have been prepared. One of the most interesting approaches has been to use both enantiomers of the benzomorphan metazocine in place of tyrosine in [Met]enkephalinamide (175). Both diastereomers of derivatives 76 and 77 were prepared to test the hypothesis that the tyramine moiety in rigid opiates and opioids serves an identical functional role at the receptor surface. It was found that both 76 and 77 were either inactive or feebly active in the GP] and MVD assays. This may not have been too surprising given the known lack of activity in N ,N-disubstituted tyrosine enkephalins (163). The conclusion reached was that the rigid opiates and opioids interact with the same receptor, but with different modes of binding (175). Another approach has been to constrain the tyramine moiety in an indane,
488
II
Enkephalins
VI
Structure-Activity
Relationships
Table 11.11
CH,
Glycine! Substitutions
II
Tyr-X-Gly-Phe-Y Compound I 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 Ib3 Q
X Gly Ala o*Ala D-Ala fJ-A[a Aib Aib Aib f:-Ahx D-Leu o-Lys D-Met D-Met D-Met D-Phc Pro D-Pro D-Ser v-Ser o-Trp v-Thr D-Val
Y Met Leu Leu Met Leu Leu Leu-NHz Met-NH2 Leu Leu Leu Leu Leu-NH2 Met-NHz Leu Leu Met Met D-Leu Leu Met Met
GP[
MVD
1.0 0.026" 1.8-5.5 1.14-6.3 0.002 0.13
1.0 0.009 8.3-9.4 5.0-7.2
Inactive 1.34" 0.18" 8.0, 63.
Inactive 0.020 3.0a 0.44,3.7 0.54 0.0003
0.01 [4.3" 1.4 O.06Q 0.56 0.42
29.5a [5 0.05
References
112 110,112,113,114 11/ ,1/2,113,/72 113 113 178 /78 /79 112,113 /80 113,/65 114,/45,165 /45 /59 112 /6/ /65 7/ 112,113 /59 /59
Methyl ester.
Among the many approaches to conformation rigidity, the use of dehydroamino acids has been recently introduced mainly because of the Sp2 character of the a carbon. The synthesis of these molecules has been reviewed (181), Although the Z-isomers are readily available, the synthesis of the less accessible E-isomers has been described (182), The use of tl-Ala2 (104) in place of Gly2 has been studied using opiate receptor affinities with DADLE (33) and dihydromorphine as a ligand. In these biochemical assays, 104 is about equipotent to DADLE in displacing dihydromorphine and two-thirds as active as DADLE in displacing itself (183). Consequently, 104 is about one-half as selective as DADLE for the 5 receptor. A more classical approach is the alkylation of the peptide chain backbone, in this case, replacement of Gly2 with sarcosine (Sar), Nmethylglycine, The Sar2-analog (105), however, possesses only 1% or less of the biological activity of [Met]enk (I) (112), An interesting approach is the N-substitution of Gly' with acetic acid or {3-hydroxyethyl (107) residues
-
in the Enkephalins
489
T yr-CO- NII-C-CO-G Iy-Phe- Leu 104 (i>-Ala'-[Leu]enk)
Tyr-Sar-Gly-Phe-Leu 105
Tyr-( R)Gly-Gly- Phe-Leu
Tyr-(
106 (R ~ CH,CO,H) 107 (R ~ CH,CH,OIl)
Gly-Gly-
Phe. Leu
h
108 Tyr.CO-NH-CH,-CS-Gly-Phe109
Leu
(184). While compound 106 has virtually no aClivity, the alcohol derivative (107) has 0.43 limes Ihe aClivily of [MetJenk (I) in the MVD assay and only 0.06 times in Ihe GPI assay. This hydroxy ethyl side chain may be mimicking the known O-selective effects of Ser and Thr2 (143,144). A dimeric form (108) of [LeuJenk, sharing a common Tyr', has also been prepared that also shows MVD selectivity but has only weak polency (184). The thiocarbonyl group has replaced the carbonyl group in Gly2 as another form of conformational constraint. The thioamide analog (109) has 2.6 and 9, I times the activity of [LeuJenk (2) in the GPI and MVD tests, respectively (185), Various approaches to rigid analogs have been described that incorporate rigid amino acids in place of Gly2 or ring formation between Tyrt and Gly2, Methylene bridging of the a-carbon of Gly2 and the nitrogen of Gly2 results in a series of bridged analogs (1l0-1l2), Of these, III has 0.1 times (ICH2Int Tyr-NH~N-CH2CO-Phe-Met-NH2
o 110
(n::
III
(n
2)
112
(n::::
c 3)
4)
r-\ Tyr-N
N-CH2CO-Phe-Leu
L../ 114
113
490
II
Enkephalins
the potency of [Met]enk (1), the others less than 0.01 times the potency (/86,187). The thiazolidinone derivative (113), an analog of a cyclic methionine, has only 3% of the activity of (Metjenk (1) (/86). The amide nitrogens of Gly2 and Gly3 have also been bridged by an ethylene group (114), resulting in a biologically inactive molecule (/88). Similar results were observed with [Metjenkephalin derivatives (189). Formation of an imidazolidinone ring between the Tyrl and Gly2 nitrogen (115) results in a drastic reduction of biological activity (190). Similarly, the diketopiperazine derivative (116) is inactive due to, among other things, the blockade of the Tyrl amino group (191). On the whole, the experience with the use of rigid analogs at Gly2 has been disappointing. D.
Glycine3 Structure-Activity
VI
Structure-Activity
Relationships
in the Enkcphalins
491
0.3% of [MetJenk (1) (112). Replacement of the Gly3 amide bond in [LeuJenk (2) with a thioamide, similarly to that illustrated in 109, retains GP! equivalency to [LeuJenk (2), but MVD activity is reduced by 85% (/85). Another rigid analog (125), where the Gly2 and Gly3 amide nitrogens are linked by an ethylene bridge, is also inactive (189).
, 1\
Tyr-D-Ala-N
N-Phe-Leu-NH2
o
125
Relationships
In terms of the overall enkephalin molecule, Gly3 is a critical section that allows very little manipulation. Deletion of Gly3 (or, equivalently, GlyZ) drastically reduces GPI activity to less than 0.2% of that of [Met]enkephalin (193). Introduction of other natural amino acids also strongly decreases biological activity (Table 11-12). Surprisingly, the incorporation of D-Ala3 119 and 120, so successful at Gly2, also reduces both GPI and MVD activity to 2% or less. The replacement of Gly' by Ser3 and dehydroalanine has been investigated biochemically (/92). With both D- and L-amino acids so ineffective, it was not surprising that the l3-alanine derivative (121) was also inactive, as was Aib' (179). Backbone methylation, replacement of Gly3 with Sar3 in [LeuJenk (2) in common with the preceeding observations, also reduces MVD to only
In summary, structure-activity relationship studies for Gly3 have indicated that all modifications introduced thus far, with the exception of a thioamide group. reduce, usually drastically, the biological activity. E.
Phenylalanine4 Structure-Activity
Relationships
The Phe4 position of the enkephalins, like the others, has been extensively investigated, not only from the structure-activity relationship angle, but also for the determination of receptor type selectivities. Also, in contrast to Tyrl, a significant amount of effort has gone into studying the effects of substituents on the aromatic ring of phenylalanine. However, similarly to Tyrl, replacement of Phe4 by D-Phe4 (129) effectively abolishes biological activity (Table 11-13). Replacement of Phe4 with a variety of other amino acids (Table 11-13), with the ~xception of Trp, eliminates both MVD and GPI activity.
Table 11-12 Gly] Substitutions
Tahle 11-13 Phe4 Substitutions
Tyr-Gly-X-Phc- Y Compound I 117 118 119 120 121 122 123 124
X Gly Ala Ala D-Ala I)-Ala I3-Ala Pro D-Pro SCT
Y
GPI
MVD
Met Leu Met Leu Met Leu Leu Leu Leu
1.0
1.0 0.025
<0.01 O.OIR O.t14 <0.001
0.0031 O.()()1 <0.001 0.(XJ5
Tyr-Gly-Gly-X- Y
References
112 /59 1/2.1/3 1/2./73,/74 1/3 JJ2 Il2 192
Compound
X
Y
GPI
MVD
I 126 127 128 129 130 131 132
Phe Gly Ala D-Ala D-Phe Tyt Tyr Trp
Met Met Met Met Met Leu Met Met
1.0 Inactive Inactive <0.01 <0.01
1.0
<0.001 0.02 0.001 0.27
References
159 150 /59 159.173 1/2 /62 194
492
11 Enkephalins Table
VI
Structure-Activity
11-14
j,Z.Phe4
Table
Substitutions
Relationships
in the Enkephalins
493
11-15
Phe4 Aromatic Nitro Substitions Z
Tyr- X -Gly-j,
Tyr-D-Ala-Gly-(p-NO,)Phe-X
-Phe-Leu
Opiate Binding
Compound
x
DHM (nm)
133 134 Morphine
Gly D-Ala
21 6.4
DADLE (nm) 3.8 1.9
136 137 138
> 50 mpk 13 0.5
a
22.7 174 35 Relative
to [Leu]enkephalin
(2)
MVD"
Reference
10.5 51
198 199 199
= 1.
Table 11-16 Phe4 Aromatic Subst;tuents Tyr-D-Ala-Gly-Phe-(Me
~C6H5
Compound 138 (Metkephamid) 139 140 141 142 143 144 145 146 147 148 149 150 151
CO-Leu
135
Substitution patterns on the aromatic ring of Phe4 have been investigated. The initial results were based on the p-nitro derivatives (Table 11-15). Introducing a para-nitro group on Phe4 (136-138) and varying the fifth reSIdue furnIShed analogs with substantially increased GPI and MVD potencies relative to the unsubstituted [Leu)enkephalin (2) (/98,199). However, smce 2 IS substantially less active in the GPI assay, the Met analog (138) is 8 selective, while D-Leu (137) does not discriminate.
GP!"
An extensive structure-activity relationship investigation has been conducted in meta- and para-substituents on the analgesic peptide, metkephamid (138). The results reported for MVD potency and analgesia in Table 11-16 are all related to this peptide. In the meta-substitution series, an electronegative halogen (139, 140) appears to be the best substitution. Other larger electronegative groups such as trifluoromethyl (141) are less efficient, as are electron-releasing groups (142, 143) (200). Analgesic potency by sc administration parallels the MVD potencies but is not as sensitive. Para-substitutions show a generally similar pattern to the meta, with electron-withdrawing groups yielding the most potent peptides: Fluorine (144) and nitro (ISO) have similar MVD potencies, with (144)
However, the amino acids investigated do not have the large lipophilic side cham necessary to replace the aromatic ring of phenylalanine in J.<-selective ligands (Section IV). Exchange of Phe4 for a'-Phe provides analogs 133 and 134, (Table 11-14), which possess high affinity for the opiate receptor; also 133 has hIgher selectIvIty for the 8 receptor (/83). More importantly, the dehydro analogs possess in vivo analgesic properties, although they are weak (/95). However, the [a'-Phe4,Met5]enkephalin amide derivative has 0.2 times the activity of morphine by intravenous (iv) administration (/95,196). The preparation of E-dehydrophenylalanine has been reported, but has not yet been incorporated into peptides (/82). Another conformationally restricted Phe4 analog is the cyclopropyl derivative (135) (/97). However, both enantiomers are several orders of magnitude less active than the parent in both the GPI and MVD assays (/97).
Tyr-D-Ala-NH
x
Compound Analgesia Mouse Tail Flick (iv)
:]I
1..
Phe4 Aromatic Substitution H m-CI m-Br m-CF) m-CH) m-OCH) p-F p-CI D-p-CI p-Br
p-! p-CF, p-N02 poOH
)Met-NH,
MVD
Analgesia Mouse Jump (sc)
Reference
1.0 17.2 24.9 6.9 4.5 4.2 15.8 5.5 0.07 5.5 0.9 8.1 19.1 0.03
0.36 mpk 0.11 0.36 0.33 0.33 0.15 0.02 0.67 >30 1.6 >30 0.43 0.16 >10
200 200 200 200 200 200 201 201 201 201 201 201 201 201
11 Enkcphalins
494
VI
Structure-Activity
Relationships
Table 11-18
Table 11-17
Hexahydrophenylalamine
N-Alkylatcd
Derivatives
Tyr-D-Met-Gly-Phc(
LPhe4]-Tetrapeptide
Amides R
611 )-X Tyr-D-Ala-Gly.L
Compound
152 153 "Relative
495
in the Enkephalins
x Met-NH2 Leu-NH2 to [MetJenkephalin
Phe-NH2
GP[" Analgesia 3.0 2.3 (I)
0.31 0.10
= 1.
being approximately eight times more analgesic (201). The D-p-chloro derivative (146) is substantially less active than the L-p-chloro derivative (145). The electron-releasing and acidic p-hydroxy group, replacement of Phe4 with Tyr4, is even less active than the D-p-chloro (146) in the MVD assay and has very weak analgesic properties (201). Similar results have been reported for the tetrapeptides, lacking the fifth amino acid, related to (138) (202). The aromatic ring of Phe4 is not necessary for biological activity. Reduction to the hexahydro derivative can furnish analgesic enkephalins, which were origina)ly thought to lack physical dependency (203), although this was later disproved (204). Incorporation of Phe(6H)4 in the [DMet2]enkephalamides 152 and 153 (Table 11-17) yields peptides that are slightly more potent in the GPI test and substantially less active in the MVD assay (145). The effect on GPI and MVD activity on replacement of Phe4 by large alkyl groups has been discussed in Sections IV and V. Backbone alkylation of Phe4 has been used extensively, primarily with the introduction of methyl and ethyl residues. A thorough investigation has been reported in a tetrapeptide where the alkyl group has been varied systematically among both alkyl chains and those groups known to impart antagonist and mixed agonist-antagonist properties in classical rigid opiates (Table 11-18) (205). For MVD activity, potency peaks at N-ethyl (156) for the alkyl substituents, being almost 90 times more potent than the unsubstituted parent (154). Substantial MVD activity is also shown by the opiate antagonist, mixed agonist-antagonist substituents (159-161). Analgesic efficacy is also achieved by alkylation of the Phe4 amide nitrogen, with the N-ethyl analog (156) being some 280 times more potent than the un substituted parent (154) and 150 times more potent than its N-melhyl congener (155). Analgesic potency diminishes with the larger hpophilic substituents (158), as well as with the shorter, more hydrophilic denvalIves (163, 164) (202,205). Whether the derivatives 159-161 possess opiate antagonist properties was not disclosed.
Compound 154 155 156 157 158 159 160 161 162 163 164 165 "Relative
R
MVD"
Mouse Hot Plate (st:)
0.11 0.17 9.5 1.8 0.07 2.4 12.3 3.1 4.7
O.ss mpk 0.47 0.003 0.15 11.7 0.01 0.012 0.4 0.05 1.4 0.33 0.013
II CH3 C2H5 fI-C.~H7 fI-C"HI1 CII,CII~CII, CHrc-C3H5 CII,C![~C(CII,), (CII,),SCII, CII,CO,CH, CII,CH,OIl CH2CII2F to (Met)enkephalin
0.10 0.40 0.52
(I)
=I
(reference
225).
The thioamide grouping has been introduced at Phe4 to generate conformationally restrained enkephalin analogs. The [Leu]enkephalin derivative containing the thioamide in place of the amide group at Phe4 is, however, equipotent to [Leu]enkephalin in both the MVD and GPI assays, as well as in IL and 8 opiate receptor affinities (185). A rigid analog was prepared by joining the Phe4 amide nitrogen to the Leus amide nitrogen by a 2-carbon bridge (189). The diastereomers of the resultant (166) were
166 separated, and neither had appreciable binding for the opiate receptor and only weak analgesic properties. In summary, the Phe4 position has considerable latitude in terms of what substitutions can be accommodated without drastic decreases in biological activity. The phenylalanine aromatic ring can be substituted or reduced,
496
11
VI
Enkephalins
the amide nitrogen can be alkylated, and Phe' can be replaced with other aromatic amino acids such as Trp4. F.
Methionine'/Leucine'
Structure-Activity
Structure-Activity
Relationships
in the Enkephalins
497
of the activity of their corresponding parents, respectively (Il3,207). Replacement of Met'/Leu' with hydrophilic groups, such as hydroxyalkyl amides (25) (DAGO) or aminoalkylamides 175 and 176, leads to a significant enhancement of GPI effects and a decrease in MVD activity (134). These compounds, 175 and 176, are potent analgesics by iv administration. A similar but much stronger effect is observed with 177, which contains a highly lipophilic group. In 177, the GPI potency is 92 times that of [Met]enkephalin (I), while it is only 0.5 times as potent in the MVD assay (208) Reduction of both I and 2 to their corresponding alcohols 178 and 179, retains both receptor affinity and GPI activity (Il2,13i). The corresponding [D-Ala2,Met'-01] derivative (180) and its corresponding sulfoxide both possess strong opiate receptor affinity and are potent analgesics by parenteral administration (135). The reductIOn of the terminal amino acid to the corresponding alcohol has served to prepare biologically potent enkephalin derivaties, for example, (24-26).
Relationships
Structural and conformational changes at the Met'/Leu' position are well tolerated and have led, almost invariably, to active compounds. The tetrapeptide, Try-Gly-Gly-Phe, resulting from removal of the terminal amino acid, retains activity in both the GPI and MVD assays (Il2,Il3,159). Similar but more potent biological activity was observed with the D-Ala2 tetrapeptides reported in Table 11-18. Replacement of Leu' by its optical antipode, D-Leu' (167), results in a decrease in GPI potency, while the MVD activity is unchanged (Table 11-19). A similar decrease is observed with DADLE (33), but the MVD is increased threefold. In [Met]enkephalin (I), replacement with D-Met' (168) significantly decreases GPI activity and virtually eliminates MVD. With D-Ala2, the use of D-Met' (169) yields a peptide qualitatively similar to [DLeu']enkephalin (167), where activity is reduced and MVD unchanged. Oxidation of the sulfide of methionine in I to the sulfoxide results in a greater decrease in GPI than MVD potency, but the compound (170) still retains significant. activity in both assays. Sulfone formation (171) decreases GPI activity IOO-fold. The D-Ala2 sulfoxide (172) has been reported to be 30-50 times more potent as an analgesic than the corresponding methionine derivative (206). Decarboxylation of [Met]- and [Leu]enkephalin I and 2 yields the tetrapeptide amides 173 and 174, which retain 7% (MVD) and 0.6% (GPI)
Tyr.Gly.Gly.Phe.R 173
(R
174
(R
175 176
= NH(CH,),SCH,)
= NH(CH,),CH(CH,),)
Tyr-o-Ala-Gly-McPhc-R (R = NII(CH,),N(CH,),) (R = NH(CII,),NO(CH,),)
Tyr-D.Ala.G Iy. Phe-N.( CH,) ,GI( CH ,),
I CH2Ct,lls 177 Tyr-Gly-Gly.Phe-R 178 179
(R (R
= Met'DI) = Leu-ol)
Tyr-D-Ala-Gly-Phc-Mct-ol 180
Table 11-19 Introduction of D-Lcu5fMet5 and the Effects of Me,s Oxidation
Tyr-Gly-Gly-
Tyr.X-Gly-Phe-Y Compound
2 167 84 33 I 168 85 169 170 171 172
X Gly Gly D-Ala o.Ala Gly Gly D-Ala D-Ala Gly Gly D-Ala
Y Leu f)-Leu Leu D-Leu Me' o-Met Met o-Mct Met(O) Met(O,) Me'(O)-NH,
GPt
MVD
References
0.36 0.15 5.5 3.3 1.0 0.1t 5.6 1.8 0.2 0.01
1.2 1.4 9.0 27 1.0 0.002 5.0 5.9 0.67
/12 /12,159 /18 /18 /18 /59./73 /18 /18 /64 /59 206
CI
Phe-NH( 189
CH, J,CO,H
Tyr-Gly-Gly-Phe-R 190 (NH-CH(CH,CH(CII,),)PO,H,) 191 (NH.CH(CH,CH,SCII,)pO,H,)
Replacement of Met'/Leu' with a variety of other naturally occurring amino acids leads to a significant reduction in biological activity, with the expected exceptions of isoleucine (184) and norleucine (185) (Table 11-20). The use of 6-aminohexanoic acid in position 5 (189) was unsuccessful, but the [D-Met2] analog of 189 was very potent, equivalent to 0.75 times
.6.J
408
II Table
Enkephalins
VI
11-20
Met~/Lcu'"
I 181 182 183 184 185 186 187 188
X
GPI
Met Ala D-Ala Gly
MVD
1.0 0.029 0.065 0.02
lie Nle Phe Pro Val
References
1.0
o.lm 0.03 0.78
0.5 0.11 IWll 0.04
0.IX15
1/8 1/2.1/3 1/3 162 159 2U9 /59 159 1/2
the potency of morphine parenterally (179). The synthesis of a wide variety of amino acids in which the carboxyl group has been replaced by a phosphonic acid and the incorporation of these acids as the fifth residue in enkephalin have been reported (2/0). The Leus/MetS analogs 190 and 191 possess analgesic properties (2/1). The carboxyl group of DADLE (33) has been converted into a chloromethyl ketone (192) in an attempt to prepare an irreversible receptor ligand (212). This ligand, DALECK (192), binds irreversibly to the high-affinity opiate-binding site and induces a long-lasting, dose-dependent analgesia (212). Leu' has also been converted to the dehydroamino acid and incorporated into the analogs 193 and 194 in order to retard enzymatic hydrolysis of the amide bond.
G.
I, ),SCt t dCSNI t(CH,)
195
1\
Tyr-Gly-Gly-Phe-N
H
R 196 (R 197 (R
C- Terminus
The N-substituents necessary for inducing opiate antagonist properties into the rigid opiates have been well characterized in the preceding chapters. The introduction of these substituents, cyclopropylmethyl or allyl, to the N-terminus of enkephalins has led, in the monosubstituted cases, either to agonists with sharply reduced biological activity or, at best, to partial agonists (111,208,216,217). On the other hand, the N,N-diallyl compound (198), and its more stable isosteric analog (199), where the Gly3_Phe4 bond has been replaced by CHz-S-, are pure antagonists with I>
Phe-.l". Leu
t.CH((G
The Enkephalin
H. EnkephaIin-Based Opioid Antagonists
194
Tyr-D-Ala-Gly-Phe-Nt
499
[o-Ala2,Met5]enkephalin showed GPI potency peaking at ethyl and icv hot plate analgesia peaking at n-propyl (2.1 times that of morphine) (215). The increase in chain length to six amino acids by addition of the Thr" (35) or Serb (34) to yield highly I>-specific peptides has also been mentioned (71.143,144),
Phe-.l z. Leu 193
Tyr-o-Ala-Gly-
in the Enkcphalins
The effect of C-terminal amidation on biological activity has been illustrated in Table 11-3. The N-ethylamide of [o-Ala2,Pro5]enkephalin displays similar GPIIMVD properties and is approximately equivalent in potency to the parent amide (165). A series of N-alkyl amides of
CII,CH(CHd, I Tyr-o-Ala.Gly-Phc-l'tl-Ctl.COCI t,Cl 192 Tyr-Gly-Gly-
Relationships
Both peptides possess affinity for both the J.Land I>opiate receptors and are approximately equivalent to [Leu]enkephalin (2). As analgesics, 194 has 2% of the iv activity of morphine, while 193 is inactive (183,213). The N-backbone methylation of [Leu]enkephalin (2) and its o-Ala2 analog (84) does not result in any significant change in either in vitro potency or selectivity (159,169). However, the increased resistance to hydrolytic enzymes by this N-methyl substituent can substantially increase analgesic potency in enkephalin-based analgesics. The thioamide grouping has also been used to provide enzyme resistance and conformational restraints. For instance, the thioamide (195) possesses substantial analgesic potency by icv administration and is reported to have a better therapeutic index than [Met]enkephalin (I) (214). Rigid analogs, in which either Met' or Leu5 amides are constrained in a six-membered ring, have been synthesized (189). The rigid analog (196) of [Leu]enkephalin (2) is 4.5 times as potent iv as morphine in the tail flick test. The methionine analog (197), while possessing greater affinity for the opiate receptor, approximates morphine as an analgesic (189).
Substitutions Tyr-Gly-Gly-Phc-X
Compound
Structure-Activity
NH 0
opiate receptor I
= CII,CH(CII,J,I = (CH,I,SCH,)
phine I
~
selectivity
(218,219).
ta 3D-fold selectivity in antagonizing in MVD
preparations.
Antagonists
198 and 199 possess a IO(2) over normorinvestigations primarily
[Leu ]enkephalin
Structure-activity
\00
11 N.N-( Allyl
h Tyr-Gly-Gly-
En kephalins
VII
Phc- Leu
Clinically Investigated
Enkephalin
Tyr-D-Ala-Gly-Phe202 (melkephamid.
198 N.N-(Allylh Tyr-Gly-NH(CH,hS-CH(CH,C,H,)CO-Leu 199 N. N-(Allylh
Tyr- Aib-Aib-
26
Phe- Leu
Clinically
Investigated
(FK 33-824).
Tyr-D-Mct-Gly-Phe203
MeMet-NHz LY-127.623) 0 )-01
DAMME) Pro-NHz
([D-Met2,ProS]enkephalinamide)
CH ,) (CH,hC,H,
By iv administration, it has somewhat less analgesic potency than by sc administration. The duration of metkephamid-induced analgesia was intermediate between that of meperidine and morphine. A physical dependence study indicated that 202 produced only slightly more dependence in rats than saline_ Metkephamid has minimal effects on respiration, with no effects being noted below 64 mpk (225,226). Clinically, metkephamid was administered to postoperative patients and had an onset of action of 0.5 hour and a duration of about 4 hours. Side effects were noted in 80% of the patients. Besides the usual opiate side effects, actions unique to metkephamid were a feeling of heaviness in the limbs, emotional dissociation, a burning sensation at the injection site, dry mouth, and nasal congestion. It was suggested that some of these effects might be due to 8 receptor activity (226,227). It was stated that the side effects were not overly distressing (227). FK 33-8214 (26) was reported in 1977 by the group at Sandoz (135). A number of in vivo studies have demonstrated that it is a potent analgesic. On icv administration, FK 33-824 is 1000 times more potent than morphine in the mouse tail flick assay_ The peptide was twice as active as morphine on sc administration and orally had 0.2 and 0.3 times the activity of morphine in the tail flick and hot plate assays, respectively. The &nalgesia was dose dependent and naloxone reversible. Tolerance and physical dependence have been produced in monkeys, and withdrawal symptoms were observed after rapid drug withdrawal (135). While respiratory depression is apparent with 26, its therapeutic index, related to analgesia, appears to be large (228). In humans, FK 33-824 has been tested for potential therapeutic effectiveness in schizophrenia (229). The peptide also releases prolactin (230) and inhibits luteinizing hormone secretion (231)_ When administered, FK 33824 demonstrated many of the same symptoms as metkephamid (202): Heaviness of the limbs and muscles, often coupled with a feeling of oppression on the chest or a tightness in the throat (232,233), facial flushing, and dry mouth were observed (232). Other commonly observed morphine-like effects were noticeably absent. As an analgesic in humans with postoperative pain, its efficacy was less than that of morphine, and the
involving the second and fifth enkephalin residues led, in most cases, to congeners with lower affinity and selectivity (220). What came out of this investigation was that structure-activity relationship data developed for the agonist enkephalins were not transferable to an antagonist series_ A highly selective 8 receptor antagonist has been developed by replacing the two glycine residues in 198 with two Aib residues (200) (221). The selective 8 antagonist (200) is equipotent with naloxone at that receptor, but is devoid of activity at the /J.and K receptors below a concentration of 5 /J.m. A different approach to antagonists is based on the secondary ami des of a tetrapeptide_ The prototypic peptide is 201, which is a pure antagonist in the MVD with a iO-fold selectivity for the /J.receptor. An initial structureactivity relationship study did not result in improved affinity or a dramatic change in receptor selectivity. However, both the Phe4 methyl group and the secondary amide are critical for antagonist properties; for example, the structurally similar analog (177) is a pure agonist (208).
VII.
\01
Tyr-o-Ala-Gly-MePhe-Mct(
200 Tyr-D-Ala-Gly-MePhe-N( 201
Analgesics
Enkephalin
Analgesics
Despite the literally thousands of peptides synthesized and tested as analgesics, only three have been investigated clinically: metkephamid, LY-127,623 (222); FK33-824 (DAMME) (223); and [D-Met2,Pro']. en kephalin amide (224). All of the peptides contain a selection of modifications inserted to retard enzymatic hydrolysis: D-amino acids, backbone methylation, or reduction of a carboxyl group to the alcohol. Metkephamid (202) was reported in 198] by the group at Lilly (225). The peptide (202) is slightly more 8 selective (1.7 times) than [Met]enkephalin in the MVD assay. By icv administration, metkephamid was at least 100 times more potent than morphine. By sc administration, it was more potent than morphjne, meperidine, pentazocine, or codeine in the writhing and mouse hot plate jump assays. In the mouse hot plate paw lick and rat tail flick assays, 202 has about one-third the potency of morphine.
1
II
102
Enkephalins
analgesia was unpredictable and insufficient, and with a shorter duration than morphine in at least half of the patients (234,235), The third clinically investigated enkephalin derivative is [D-Met2,Pro'Jenkephalinamide (203), which was reported in 1977 by a Hungarian group (236,237), As an analgesic, by icv administration, 203 was 70 times more potent than morphine in the tail flick and hot plate tests, By iv injection, it was about three times more potent tban morphine. The effects were naloxone reversible. Typical morphine-like side effects were noted in the rodents, Tolerance to the analgesia developed more rapidly in rats than with morphine (238). When administered to healthy human volunteers, the usual autonomic side effects were observed (heaviness in the limbs, dry mouth, etc.). Emotional detachment was again noted. As with 26, prolactin was released. An elevation in the pain threshold was reported (239),
VIII.
The Chemical Anatomy of the Enkephalins
The enkephalins I and 2 and their derivatives represent a special case among the opiate-based analgesics in that they are peptides. As such, they are subject to the action of proteolytic enzymes in addition to the usual metabolic, transport, tissue distribution, and CNS penetration problems inherent in the rigid opiates. This and the observation that the majority of the published biological data on enkephalins are based on the GPI and MVD assays make structure-activity relationship extrapolations in this area more uncertain than usual. Tbat the use of enkephalins for analgesia is a viable approach has been demonstrated by the clinical evaluation of several peptides. With the above reservations in mind, biological activity in the enkephalins shows the following trends: 1. The minimum chain length for analgesia is the first two amino acids with a lipophilic amide side chain. 2. A functioning amino group, eitber unsubstituted or monosubstituted, on tyrosine is necessary, as is the phenolic hydroxyl. Most otber changes result in drastic reductions in activity. 3. Substitution of D-amino acids for Gly2 usually increases activity, whereas L-amjno acid substitution, decreases it. 4. Little change is tolerated at Gly3 5. Phe4 tolerates a great deal of change. The aromatic ring can be reduced or substituted for by electron-withdrawing groups. A de bydrophenylalanine derivative possesses analgesic properties. Phe4 can be substituted for by Trp but not by most other amino acids,
References
103
6. The fifth position also tolerates a great deal of manipulation, with most changes resulting in active compounds. Oxidation of Met' to its sulfoxide, but not its sulfone, results in enhanced analgesic properties. Both Leu5/Met5 can be de carboxylated as well as replaced by other natural and unnatural lipophilic amino acids. Reduction of the carboxyl group to the alcohol results in either retained or enhanced analgesic potency. Formation of amides usually results in enhanced stability. 7. Amide backbone substitution at Phe4 and Met'/Leu' can produce potent analgesics, but introduction at either Gly2 or Gly3 is contraindicated. The functional group dispositions that influence J.Land 8 opiate receptor specificities are: For J.Lreceptor specificity: 1. 2. 3. 4. 5.
Decreasing the number of amino acids in the peptide. Introduction of a lipophilic D-amino acid at Gly2, Replacement of Phe4 by a large lipophilic chain. Hydrophobic amino acids at position 5. A folded conformation.
For /5 receptor specificity: 1. 2. 3. 4.
The presence of the aromatic ring of Phe4 A hydrophilic C-terminus A hydrophilic D-amino acid at Gly2. An extended conformation. However, the /5specificity observed with cyclic penta-en kephalin derivatives may qualify this requirement.
References 1. D. V. Reynolds, Science (Washington, D. C.) 164, 444 (1969). 2. D. J. Mayer, T. L. Wolfe, H. Akil. J. C. Carder, and J. C. Liebeskind. Science (Washington, D. C.) 174, 1351 (1971); H. Akil, D. J. Mayer, and J. C. Licbeskind, Science (Washington, D. C.) 191, 961 (1976). 3. J. T. Hughes, Brain Res. 88, 295 (1975). 4. A. Goldstein, L. T. Lowney, and B. K. Pal, Proc. Natl. Acad. Sci. U.S.A. 68, 1742 (1971). 5. E. J. Simon. J. M. Hiller, and I. Edelman, Proc. Nail. Acad. Sci. U.S.A. 70. 1947 (1973). 6. C. B. Pert and S. H. Snyder, Science (Washington, D. C.) 179. 1011 (1973). 7. L. Terenius and A. Waldstrom, Life Sci. 16. 1759 (1975). 8. L. Terenius and A. Waldstrom, Acta Physio. Scand. 94. 74 (1975). 9. G. W. Pasternak, B. Goodman, and S. H. Snyder, Life Sci. 16. 1765 (1975).
11 Enkcphalins
104
to. B. M. Cox. K. E. Opheim, H. Tcschemachcr, and A. Goldstein, Life Sci. 16, 1777 (1975). 11. J. T. Hughes, I. W. Smith, H. W. Kosterlitz, LA. Fothergill, B. A. Morgan, and H. R. Morris, Nature (London) 258, 597 (1975). 12. J. S. Morley, Br. Med. Bull. 39, 5 (1983). 13. J. W. Thompson, Br. Med. J. 288, 259 (1984). 14. J. Rossier, Y. Audigier, N. Ling, J. Cross, and S. Udcnfriend, Nature (London) 288, 88 (1980). 15. D. L. Kilpatrick, T. Taniguchi, B. N. Jones, A. S. Stern, J. E. Shively, J. Hullihan, S. Kimura, $. Stein. and S. Udenfricnd, Proc. NUll. Acad. Sci. U.S.A. 78.3265 (1981). 16. U. Gubler, P. Seeburg, B. J. Hoffman, L. P. Gage, and S. Udenfriend, Nature (London) 295, 206 (1982); M. Nada, Y. Furutani, H. Takahashi, M. Toyosata. T. Hirose, S. Inayama, S. Nakanishi, and S. Numa, Nature (London) 295, 202 (1982). 17. A. Goldstein, W. Fischli, L. I. Lowney, M. Hunkapiller, and L. Hood, Proc. Nat!. Acad. Sci. U.S.A. 78, 7219 (1981). 18. K. Kangawa, N. Minamino, N. Chino, S. Sakakibara, and H. Matsuo, Biochem. Biophy~'. Res. Commun. 99, 871 (1981); N. Minamino, K. Kangawa, N. Chino, S. Sakakibara, and H. Matsuo, Biochem. Biophys. Res. Commun, 99, 864 (1981). 19. N. Minamino, K. Mizumo, Y. Hayashi, K. Kangawa, H. Mat!;uo, N. Chino, and S. Sakakibara, in "Peptide Chemistry 1980" (K. Osawa, ed.), p. 145. Protein Research Foundation, Osaka, 1981. 20. A. F. Bradbury, D. G. Smyth, C. R. Snell, N. J. M. Birdsail, and E. C. Hulme, Nature (London) 260, 793 (1976). 21. C. H. Li and ,D. Chung, Proc. Natl. Acad. Sci. U.S.A. 73, 1145 (1976). 22. N. Lin, R. Burgus, and R. Guillemin, Proc. Natl. Acad. Sci. U.S.A. 73,3942 (1976). 23. D. G. Smyth, Br. Med. Bull. 39, 25 (1983). 24. R. H. GuiJIemin, Science (Washington, D. C.) 202, 390 (1978). 25. L. Terenius, Annu. Rev. Pharmacol. Toxieol. 18, 189 (1978). 26. F. Bloom, Annu. Rev. Pharmacol. Toxicol. 23, 151 (1983). 27. J. S. Morley, in "Perspectives in Peptide Chemistry" (A. Eberle, R. Geiger, and T. Wieland, eds.), p. 329. Karger, Basel, 1981. 28. A. Henschen, F. Lottspeich, V. Brantl, and H. Teschemacher, Hoppe-Seyler'.\' Z. Physiol. Chern. 360, 1217 (1979). 29. R. de Castiglione, F. Farao, G. Pcrseo, and S. Piani, Int. J. Pept. Protein Res. 17,263 . (1981). 30. P. C. Montecucchi, R. de Castiglione, S. Piani, L. Gozzini, and V. Erspamer, Int. J. Pept. Protein Res. 17,275 (1981). 31. P. C. Montecucchi, R. de Castiglione, and V. Erspamcr, Int. J. Pept. Protein Res. 17, 316 (1981). 32. H. Takagi, H. Shiomi, H. Ueda, and H. Amano, Nature (London) 282, 410 (1979). 33. V. Brantl, H. Teschemacher, J. Blaesig, A. Henschen, and F. Lottspcich, Life Sci. 28, 1903
(1981).
34. V. Erspamer and P. Melchiorri, Trends Pharm. Sci. 1,391 (1980). 35. S. Salvadori, M. Marastoni, R. Tomatis, and G. Sarto, Farmaco, Ed. Sci. 38, 153 (1983). 36. S. Salvadori, G. P. Sarto, and R. Tomatis, Eur. J. Med. Chem. 18, 489 (1983). 37. K. Darlak, Z. Grzonka, P. Janicki, A. Czlonkowski, and S. W. Gumulka, J. Med. Chern. 26, 1445 (1983). 38. R. de Castiglione and G. Perseo, Int. J. Pept. Protein Res. 21, 471 (1983). 39. J. L. Marx, Science (Washington, D. C.) 220, 395 (1983). 40. J. Hughes, Br. Med. Bull. 39, 17 (1983).
References 41. 42. 43. 44.
101
R. V. Lewis and A. S. Stern, AmlU. Rev. Pharmacol. Toxicol. 23, 353 (1983). S. Udenfriend and D. L. Kilpatrick, Arch. Biochem. Biophys. 221, 309 (1983). O. Civelli, J. Douglass, and E. Herbert, Peptides (N. Y.) 6, 69 (1984). H. Imura, Y. Nakai, N. Nakao, S. Oki, I. Tanaka, H. Jingami, T. Yoshimasa, T. Tsukada, Y. Ikeda, M. Suda, and M. Sakamoto, J. Endocrinol. Invest. 6, 139 (1983). 45. A. Z. R6nai, in "Opioid Peptides" (J. I. Szekely and A. Z. R6nai, eds, Vol. 3, p. 107. CRC Press, Boca Raton, Florida, 1983. 46. C. H. Li and D. Chung, Nature (London) 260, 622 (1976). 47. M. Chretien, S. Benjannet, N. Dragon, N. G. Seidah, and M. Lis, Biochem. Biophys Res. Commun, 72, 427 (1976). 48. R. Guillemin, N. Ling, R. Burgus, F. E. Bloom, and D. Segal, Psychoneuroendocrinology (Oxford) 2, 59 (1977). 49. R. E. Mains, B. A. Eipper and N. Ling, Proc. Natl. Acad. Sci. U.S.A. 74,3014 (1977). 50. J. L. Roberts and E. Herbert, Proc. Natl. Acad. Sci. U.S.A. 74, 4826 (1977). 51. J. L. Roberts and E. Herbert, Proc. Natl. Acad. Sci. U.S.A. 74,5300 (1977). 52. M. Rubinstein, S. Stein, and S. Udenfriend, Proc. Natl. Acad. Sci. U.S.A. 75, 669 (1978). 53. M. Chretien, P. Crine, F. Gassard, N. Lariviere, S. Benjamet, S. Gianoulakis, and N. G. Seidah, in "Biosynthesis, Modification and Processing of Cellular and Viral Polyproteins" (G. Koch and D. Richter, eds.), p. 87. Academic Press, New York, 1980. 54. S. Nakanishi, A. Inoue, T. Kita, S. Numa, A. C. Y. Chang, S. N. Cohen, J. Nunberg, and R. Schimke, Proc. Natl. Acad. Sci. U.S.A. 75,6021 (1978); S. Nakanishi, A.lnoue. T. Kita, M. Nakamura, A. C. Y. Chang, S. N. Cohen, and S. Numa, Nature (London) 278,423 (1979); S. Numa, Peptides (N. Y) 6, 1 (1984). 55. D. F. Steiner, D. $. Quinn, S. J. Chan, J. Marsh, and H. S. Tager, Ann. N. Y. Acad. Sci. 343, I (1980). 56. H. W. Kosterlitz, J. Magnan, and S. J. Paterson, Br. J. Pharmacol. 75, Proc. Suppl., 121P (1982). 57. D. Yamashiro, P. Nicolas, and C. H. Li, I"t. J. Pept. Protein Res. 21, 219 (1983). 58. C. H. Li, in "Hormonal Proteins and Peptides" (c. H. Li, ed.), Vol. 10, p. 1. Academic Press, New York, 1981. 59. D. Yamashiro and C. H. Li, Peptides (N. Y.) 6, 191 (1984). 60. J. F. Deakin, J. O. Dostrovky, and D. G. Smyth, Biochem. J. J89, 501 (1980). 61. A. Goldstein, Science (Washington, D. C.) 176, 1081 (1976). 62. J. Rossier, T. M. Vargo, S. Minick, N. Ling, F. E. Bloom, and R. Guillemin, Proc. Natl. Acad. Sci. U.S.A. 74, 5162 (1977); R. Simantov, S. R. Childers, and S. Snyder, Proc. Natl. Acad. Sci. U.S.A. 74, 648 (1977). 63. D. T. Krieger, A. Liotta, and M. J. Brownstein, Proc. Natl. A cad. Sci. U.S.A. 74,648 (1977); F. Bloom, E. Battenberg, J. Rossicr, N. Ling, and R. Guillcmin, Proc. Natl. Acad. Sci. U.S.A. 75, 1591 (1978). 64. S. Udenfriend and D. L. Kilpatrick, Peptides (N. Y.) 6, 25 (1984). 65. M. Schultzberg, T. Hokfelt, J. Lundberg, L. Terenius, L.-G. Elfirm, and R. Elde, Acta Physiol. Scand. 103, 475 (1978); M. Schultzberg, J. M. Lundberg, T. Hokfelt, L. Tercnius, J. Brandt, R. P. Elde, and M. Goldstein, Neuroscience 3, 1169 (1978). 66. A. S. Stern, B. N. Jones, J. E. Shively, S. Stein, and S. Udenfriend, Proc. Natl. Acad. Sci. U.S.A. 78, 1962 (1981). 67. U. Gubler, P. H. Seeburg, L. P. Gage, and S. Udcnfriend, Nature (London) 295, 206 (1982). 68. M. Noda, Y. Furutani, H. Takahashi, M. Toyosato, T. Hirose, S. Inayama, S. Nakanishi, and S. Numa, Nature (London) 295, 202 (1982).
106
11 Enkephalins
69. M. Comb, P. H. Seeburg, J. Adelman, L. Eiden, and E. Herbert, Nature (London) 295, 663 (1982). 70. J. Rossier, Y. Audiger, N. Ling, J. Cros, and S. Udenfriend. Nature (London) 288, 88 (1980); A. S. Stern, R. V. Lewis, S. Kimura, J. Rossier, S. Stcin, and S. Udenfriend, Arch. Biochem. Biophys. 20S, 606 (1980); D. L. Kilpatrick, B. N. Jones, K. Kojima, and S. Udenfriend, Biochem. Biophys. Res. Commun. 103,698 (1981); M. R. Boarder, A. J. Lockfeld and J. D. Barchas, J. Neurochem. 39, 149 (1982). 71. M.-C. Fournie-Zaluski, G. Gacel, B. Maigret, S. Premilat, and B. P. Roques, Mol. Pharmacal. 20, 484 (1981). 72. E. Weber. F. S. Esch. P. Bohlen, S. J. Paterson, A. D. Corbett, A. T. McKnight, H. W. Kosterlitz, J. D. Barchas, and C. J. Evans, Proc. Nall. Acad. Sci. U.S.A. 80, 7362 (1983). 73. A. D. Corbett, S. J. Paterson, A. T. McKnight, J. Magnan, and H. W. Kosterlitz, Nalure (London) 229, 79 (1982); Y. Audiger, H. Mazarguil, J. Rossier, and J. Cros, Eur. J. Pharmacal. 68,237 (1980); M. Wuester, P. Rubini, and R. Schulz, Life Sci. 29, 1219 (1981). 74. J. Magnan, S. J. Paterson, and H. W. Kosterlitz, Life Sci. 31, 1359 (1982). 75. J. I. Szekely, in "Opioid Peptides" (J. 1. Szekely and A. Z. R6nai, eds.), Vol. 2, p. 109. CRC Press, Boca Raton, Florida, 1983. 76. C. K. Inturrisi, T. Umans. D. Wolff. A. S. Stern, P. V. Lewis, S. Stein, and S. Udenfriend, Proc. Nar!. Acad. Sci. U.S.A. 77,5514 (1980). 77. V. Hollt, F. Cankat Tulunay, S. K. Woo, H. H. Loh, and A. lIerz, Eur. J. Pharmacol. 85, 355 (1982). 78. A. Z. R6nai, in "Opioid Peptides" (1. I. Szekely and A. Z. R6nai, eds.), Vol. 3, p. 107. CRC Press, Boca Raton, Florida, 1983. 79. A. Goldstein and V. E. Ghazarossian, Proc. Natl. Acad. Sci. U.S.A. 77,6207 (1980). 80. D. L. Kilpatrick, A. Wahlstrom, H.-W. Lahm, R. Blaeher, and S. Udenfriend, Proc. Natl.Acad. Sci. U.S.A. 79, 6480 (1982); D. L. Kilpatrick, A. Wahlstrom, H.-W. Lahm, R. Blaeher, E. Ezra, G. Fleminger, and S. Udefriend, Life Sci. 31, 1849 (1982). 81. W. Fisehli, A. Goldstein, M. W. Hunkapillar, and L. E. Hood, Proc. Nar!. Acad. Sci. U.S.A. 79,5435 (1982); W. Fischli, A. Goldstein, M. Hunkapillar, and L. Hood, Life Sci. 31, 1769 (1982). 82. H. Kakidani, Y. Furutani, H. Takahashi, M. Noda, Y. Morimoto, T. Hirose, M. Asai, S. Inayama, S. Nakanishi, and S. Numa. Nature (Londo,,) 298, 245 (1982). 83. A. Goldstein, Peptides (N. Y.) 6, 95 (1984). 84. A. D. Corbett, H. W. Kosterlitl, A. T. McKnight, and S. J. Paterson, unpublished, quoted in ref. 85. 85. S. J. Paterson, L. E. Robson, and H. W. Kosterlitz, Peplide.~ (N. Y.) 6, 147 (1984). 86. C. Chavkin, I. F. James, and A. Goldstein, Scie"ce (Wa~'hinglon, D. C.) 215, 413 (1982). 87. R. Geiger and W. Konig, Peptides (N. Y.) 3, I (1981). 88. R. W. Roeske. Peptides (N. Y.) 3, 102 (1981). 89. R. G. Hiskey. Peptides (N. Y.) 3, 137 (1981). 90. J. M. Stewart, Peptides (N. Y.) 3, 170 (1981). 91. B. J. Johnson, Peptide.~ (N. Y.) 3, 304 (1981). 92. J.-L. Fauehere and R. Schwyzer, Peptides (N. Y.) 3, 203, (1981). 93. H. Yajima and N. Fujii, Peplides (N. Y.) 5, 65 (1983). 94. J. K. Inman, Peptide.. (N. Y.) 3, 254 (1981). 95. E. Wunsch, M. Deffner, K.-II. Deimer, E. Jaeger. P. Stelzel, P. Thamm, and G. Wendlberger, "Methoden der organischen Chemic" (Houben-Weyl), (E. Mueller el al., eds.), Vo!. 15, Part 2. Thieme, Stuttgart. 1974.
References
507
96. D. S. Kemp. Peplides (N. Y.) I, 317 (1979); N. L. Benoiton, Peplides (N. Y.) 5, 217 (1983). 97. E. Gross and J. Meienhofer, Peplides (N. Y.) I, 1 (1979). 98. J. Meienhofer, Peplides (N. Y.) I, 198 (1979). 99. P. Sieber, B. Riniker, M. Brugger, B. Kamber, and W. Rittel. Helv. Chim. Acta 53, 2135 (1970). 100. J. Meienhofer, Peplides (N. Y.) I, 264 (1979). 101. D. H. Rich and J. Singh. Peptides (N. Y.) I, 242 (1979). 102. M. Bodanszky, Peptides (N. Y.) 1, 106 (1979). 103. M. Bodanszky and J. Martinez, Peplides (N. Y.) 5, 111 (1983). 104. M. Mutter and E. Bayer, Peplides (N. Y.) 2, 286 (1979). 105. G. Barany and R. B. Merrifield, Peptide~' (N. Y.) 2, I (1979). 106. W. R. Martin, C. G. Eades, J. A. Thompson, R. E. Huppler, and P. E. Gilbert. J. Pharmacal. Exp. Ther. 197, 517 (1976). 107. J. A. H. Lord, A. A. Waterfield. J. Hughes, and H. W. Kosterlitz, Nature (London) 267,495 (1977). 108. G. Gacel. M.-C. Fournier-Zaluski, E. Fellion, B. P. Roques, B. Sennault, J.-M. Lecomte. B. Malfroy, J.-P. Swerts, and J.-c. Schwartz, Life Sci. 24, 725 (1979). 109. M. C. Summers and R. J. Hayes. 1. BioI. Chem. 256,4951 (1981). 110. H. W. Kosterlitz, J. A. H. Lord. S. J. Peterson. and A. A. Waterfield, Br. J. Pharmacal. 68, 333 (1980). Ill. B. A. Morgan, J. D. Bower. K. P. Guest, B. K. Handa, G. Metcalf, and C. F. C. Smith, in "Peptides: Proceedings of the 5th American Peptide Symposium" (M. Goodman and J. Mcienhofer. eds.). p. 111. Wiley. New York, 1977. 112. C. R. Beddell, R. B. Clark, G. W. Hardy, L. A Lowe. F. D. Ubatuba. J. R. Vane, S. Wilkinson. K.-J. Chang, P. Cuatrecasas, and R. J. Miller, Proc. R. Soc. London. Ser. B 198, 249 (1977). 113. A. S. Dutta, J. J. Gormley. C. F. Hayward, J. S. Morley. J. S. Shaw, G. J. Stacey, and M.T. Turnbull, ACla Pharm. Sllec., Suppl. 14, 14 (1977). 114. A. S. Dutta, J. J. Gormley, C. F. Hayward, J. S. Morley, J. S. Shaw, G. J. Stacey, and M. 1'. Turnbull. in "Peptides 1978. Proceedings of the 15th European Peptide Symposium" (Z. Siemion and G. Kupryszewski, eds.), p. 537. Wydawnictwa Uniwersytetu Wroc1awskiego, Wroc1aw, Poland, 1978. 115. Y. Audiger, H. Mazarguil, R. Gout, and J. Cros, Eur. J. Pharmacol. 63, 35 (1980). 116. R. J. Vavrek, R.-L Cui. and J. M. Stewart. Life Sci. 31, 2249 (1982). 117. V. Brantl, A. Pfeiffer. A. Herz, A. Henschen, and F. Lottspeich, Peptides3, 793 (1982). ll8. H. W. Kosterlitz, J. A. H. Lord. S. J. Paterson, andA. A. Waterfield. Br. J. Pharmacol. 68, 333 (1980). 119. B. P. Roques, G. Gacel, M.-C. Fournie-Zaluski. B. Scnault, andJ.-M. Lecomte, Eur. J. Pharmacol. 63,35 (1980). 120. P. W. Schiller, Peptides (N. Y.) 6, 220 (1984). 121. B. ~faigret, S. Premilat, M.-C. Fournie-Zaluski, and B. P. Roques, Biochem. Biophys. Res. Commun. 99, 267 (1981). 122. R. Nagaraj and P. Balaram, FEnS Leu. 96, 273 (1978). 123. F. A. Gorin, T. M. Balasubramanian, C. D. Barry, and G. R. Marshall. 1. Supramol. Slrucl. 9, 27 (1978); F. H. Clark, H. Jaggi, and R. A. Lovell. J. Med. Chem. 21,600 (1978); S. R. Childers, T. Crecse, A. M. Snowman. and S. H. Snyder. Eur. J. Pharmacol. 55, 11 (1979). 124. J. S06s, I. Berzetei, S. Bajusz, and A. Z. R6nai, Life Sci. 27, 129 (1980). 125. J. 1'. Pento, Drugs Future 7, 487 (1982).
11 Enkcphalins
108
126. K..J. Chang, A. Killian, E. Hazum, P. Cuatrecasas, and J.-K. Chang, Science (Washington, D. C.) 212, 75 (1981). 127. K.-J. Chang, P. Cuatrecasas, E. T. Wei, and J..K. Chang, Life Sci. 30, 1547 (1982). 128. K.-J. Chang, E. T. Wei, A. Killian, andJ.-K. Chang, I. Pharmacol. Exp. Ther. 227,403 (1983). 129. V. Brantl, H. Teschcmacher, J. BHisig, A. Henschen, and F. Lottspeich, Ufe Sci. 28, 1903 (1981). 130. D. M. Paton, Drugs Future 8, 875 (1983). 131. Y. Kiso, M. Yamaguchi, T. Akita, H. Moritoki, M. Takci, and H. Nakamura, Naturwissenschaften 68, 210 (1981). 132. H. Nakamura, Y. KisQ, S. Motoyoshi, N. Yoshida, K. Ishii, Y. Yokoyama, T. Kadokawa, and M. Shimizu, Eur. J. Pharmacal. 85. 133 (1982). 133. R. Quirion. Y. Kiso, and C. B. Pert, FEBS Leu. 141,203 (1982). 134. B. K. Handa, A. C. Lane, J. A. H. Lord, B. A. Morgan, M, J. Rance, and C. F. C. Smith, Eur. J. Pharmacol. 70,531 (1981); M. G. C. Gillian and H. W. Kosterlitz, Br. J. Pharrnacol. 77, 461 (1982). 135. D. Roemer, H. H. Buescher, R. C. Hill, J. Pless, W. Bauer, F. Cardinaux, A. Closse, D. Hauser, and R. Huguenin, Nature (London) 268, 547 (1977). 136. J. DiMaio and P. W. Schiller, Proc. Natl. Acad. Sci. U.S.A. 77, 7162 (1980). 137. P. W. Schiller and J. DiMaio, Nature (London) 297, 74 (1982). 138. J. DiMaio, T. M..D. Nguyen, C. Lemieux, and P. W. Schiller,J. Med. Chem. 25, 1432 (1982). 139. P. W. Schiller and J. DiMaio, in "Peptides: Proceedings of the 8th American Peptide Symposium" (y. J. Hruby and D. H. Rich, eds.), p. 269. Pierce Chemical Co., Rockford, Illinois, 1983. 140. D. Sarantakis, U.S. Patent 4,148,786, April 10, 1979. 141. P. W. Schiller, B. Eggimann, J. DiMaio, C. Lemieux, and T. M.-D. Nguyen, Biochem. Biophys. Res. Commun. 101,337 (1981); P. W. Schiller, B. Eggimann and T. M..D. Nguyen, Life Sci. 31, 1777 (1982). 142. J.-M. Zajac, G. Gacel, F. Pettit, P. Dodey, P. Rossignol, and B. P. Roques, Biochem. Biophys. Res. CommUfI. 111, 390 (1983). 143. G. Gaccl, M..C. Fournie-Zaluski and B. P. Rogues, FEBS Len. 118, 245 (1980). 144. J-M. Zajac, G. Gacel, F. Petit, P. Dodey, P. Rossignol, and B. P. Rogues, Biochem. Hiophys. Res. Commun. 111,390 (1983). 145. Y. Audigier, H. Mazarguil, R. Gout, and J. Cros, Eur. J. Pharmacol. 63, 35 (1980). 146. C. R. Bedell, R. B. Clark, R. L. Follenfont, L. A. Lowe, F. 8. Ubatuba, S. Wilkinson, and R. J. Miller, Pharmacochem. Libr. 2, 177 (1977). 147. H. I. Mosberg, R. Hurst, V. J. Hruby, J. J. Galligan, T. F. Burks, K. Gee, and H. F. Yamamura, Biochem. Biophys. Res. Commun. 1M, 506 (1982).
R. Hurst, V. J. Hruby, J. J. Galligan, T. F. Burks, K. Gee, and H. F. Yamamura, Ufe Sci. 32. 2565 (1983). 149. H. 1. Mosberg, R. Hurst, V. J. Hruby, K. Gee, H. I. Yamamura, J. J. Galligan, and 148. H. I. Mosberg,
T. F. Burks, Proc. Nat!. Acad. Sci. U.S.A.
150. H. I. Mosberg and P. W. Schiller, 151. Y. Shimohigashi, (1982).
80, 5871 (1983).
Int. J. Pept. Protein Res. 23,462
(1984). T. Costa, H.-C. Chen, and D. Rodbard, Nature (London)
297, 332
152. A. J. Lipkowski, A. M. Konecka. and I. Sroczynska, Peptides 3, 697 (1982). 153. Y. Audiger, H. Mazarguil, R. Gout, and J. Cros, Eur. 1. Med. Chem. 15,173 (1980). 154. R. E. Chipkin, D. H. Morris, M. G. English, J. D. Rosamond, C. H. Stammer, E. J. York, and J. M. Stewart, Ufe Sci. 28, 1517 (1981).
References
509
155. Y. Kiso, T. Miyazaki, T. Akita, and H. Nakamura, EUr. J. Pharmacol. 71, 347 (1981); Y. Kiso, T. Miyazaki, M. Satomi, M. Inai, T. Akita, H. Moritoki, M. Takei, and H. Nakamura, in "Peptide Chemistry 1981" (T. Shioiri, ed.), p. 65. Protein Research Foundation, Osaka, 1982. 156. Y. Kiso, T. Miyazaki, T. Akita, H. Moritoki, M. Takei, and H. Nakamura, FEBS LeU. 136, 101 (1981). 157. R. J. Vavrek, L.-H. Hsi, E. J. York, M. E. Hall, and J. M. Stewart, Peptides 2, 303 (1981). 158. Y. Kiso, T. Miyazaki, M. Inai, K. Kitagawa, T. Akita. H. Moritoki, and H. Nakamura, in "Peptide Chemistry 1982" (S. Sakakibara, ed.), p. 245. Protein Research Foundation, Osaka, 1983. 159. J. S. Morley, Annu. Rep. Pharmacol. Toxicol. 20, 81 (1980). 160. P. E. Hansen and B. A. Morgan, Peptides (N. Y.) 6, 269 (1984). 161. A. F. Spatola, in "Chemistry and Biochemistry of Amino Acids, Peptides and Proteins" (8. Weinstein, ed.), Vol. 7, p. 267. Dekker, New York, 1983. 162. B. A. Morgan, C. F. C. Smith, A. A. Waterfield, J. Hudges, and H. W. Kosterlitz, J. Pharm. Pharmacol. 28, 660 (1976). 163. R. J. Miller, K.-J. Chang, P. Cuatrecasas, S. Wilkinson, L. Lowe, C. Beddell, and R. Follenfant, in "Centrally Acting Peptides" (1. Hughes, cd.), p. 195. Macmillan, London, 1978. 164. P. W. Schiller, C. F. Yam, and M. Lis, Biochemistry 16, 183 (1977). 165. J. S. Shaw and M. J. Turnbull, Eur. J. Pharmacol. 49, 313 (1978). 166. P. C. Belanger, J. Scheigetz, R. N. Young, S. E. Charleson, R. L. Hudgin, and E. L. Engelhardt, J. Med. Chern. 24, 1297 (1981). 167. S. Bajusz, A. Z. R6nai, J. I. Szekely, E. Miglecz, and I. Berzetei, FEBS Leu. 110,85 (1980). 168. M. C. Summers and R. J. Hayes, J. BioI. Chem. 256,4951 (1981). 169. N. F. Ling, S. Minick, and R. Guillemin, Biochem. Biophys. Res. Commun. 83, 565 (1978); P. Y. Law, E. T. Wei, L. F. Tseng, H. H. Loh, and E. I. Way, UfeSci. 20,251 (1977). 170. A. Day, M. Lujan, W. Dewey, L. Harris, J. Radding, and R. Freer, Res. Commun. Chem. Pathol. Pharmacol. 14, 597 (1976). 171. G. Gacel, M.-C. Fournie-Zaluski, E. Feillon, 8. P. Rogues, B. Senault, J. M. Lecomte,
B. Malfroy,
J. P.
Swerts,
and J. C. Schwartz,
Ufe
Sci. 24, 725 (1979).
172. J. M. Hambrook, B. A. Morgan, M. J. Rance, and C. F. C. Smith, Nature (London) 262, 782 (1976). 173. D. H. Coy, A. J. Kastin, A. V. SchaUy, O. Morin, N. G. Caron, F. Labrie, J. M. Walker, R. Fertel, G. G. Bernson, and C. A. Sandman, Biochem. Biophys. Res. Commun. 73, 632 (1976). 174. N. Ling and R. Guillemin, Proc. Natl. Acad. Sci. U.S.A. 73, 3308 (1976). 175. K. Ramakrishnan and P. S. Portoghese, J. Med. Chem. 25, 1423 (1982). 176. T. Deeks, P. A. Crooks, and R. D. Waigh, J. Pharm. Sci. 73,457 (1984). 177. T. Deeks, P. A. Crooks, and R. D. Waigh, J. Med. Chern. 26,762 (1983). 178. T. S. Sudha and P. Balaram, Int. J. Pept. Protein Res. 21,381 (1983). 179. R. Paruszewski, R. Matusiak, E. Gmitrzuk, W. Gomulka, and P. Janicki, in "Peptides 1982, Proceedings 17th European Peptide Symposium" (K. Blaha and P. Malon, eds.), p. 471. Walter de Gruyter, Berlin, 1983. 180. L. Terenius, A. Wahlstrom, G. Lindberg, S. Karlsson, and U. Ragnarsson, Biochem. Biophys. Res. Commun. 7], 175 (1976). 181. U. Schmidt, J. Hausler, E. Ohler, and H. Poisel, Fortschr. Chem. Org. Naturst. 37,251 (1979).
110
II
Enkephalins
182. R. H. Mazur and D. R. Pilipauskas, in "Pcptides 1982, Proceedings of the 17th European Peptide Symposium" (K Blaha and P. Malon, cds.), p. 319. Walter de Gruyter, Berlin, 1983; R. H. Mazur and D. R. Pilipauskas, in "Peptides, Proceedin.gs of the 7th American Peptide Symposium" (D. H. Rich and E. Gross, cds.), p. 81. Pierce Chemical Co., Rockford, Illinois, 1981. 183. Y. Shimohigashi and C. H. Stammer, in "Peptides, Proceedings ~f the 7th ~merican Peptide Symposium" (D. H. Rich and E. Gross, cds.), p. 645. PIerce Chemical Co., Rockford, Illinois. 1981. 184. R. Tomatis, S. Salvadori, and G. P. Sarto, Farmaco. Ed. Sci. 35, 980 (1980). 185. G. Lajoie, F. Lepine, S. Lemaire, F. Jolicoeur, C. Aubc, A. Turcotte, and B. Belleau, Int. J. Pept. Protein Res. 24, 316 (1984). 186. D. F. Veber and R. M. Frcidingcr, U. S. Patent 4,254,107, March 3, 1981. 187. R. M. Freidinger, in "Peptides, Proceedings of the 7th American Peptide Sympos~u~" (D. H. Rich and E. Gross, eds.), p. 673. Pierce Chemical Co., Rockford, IllinOis, 1981. 188. R. Tomatis, S. Salvadori, and G. P. Sarto, EUr. J. Med. Chern. 16, 229 (1981~. 189. M. W. Moon. R. A. Lahti, P. F. Yon Voigtlander, and J. Samanen, in "Peptides, Proceedings of the 7th American Peptide Symposium" (D. H. Rich and E. Gross, eds.), p. 641. Pierce Chemical Co., Rockford, Illinois, 1981. . 190. M. C. Summers and R. J. Hayes, FEBS Left. 113,99 (1980); M. J. GIdley, L. D. Hall, J. K. M. Sanders, and M. C. Summers, Biochemistry 20, 3880 (1981). 191. R. Tomatis, S. Salvadori and M. Guarneri, Farrnaco., Ed. Sci. 34,698 (1979). 192. Y. Shimohigashi and C. H. Stammer, Int. J. Pept. Protein Res. 20, 199 (.1982). 193. J. M. G. Anton, F. Reig, G. Valencia, and J. G. Dominguez, in "peptides 1978, Proceedings of the 15th European Peptide Symposium" (Z. Siemion and G. Kupryszewski, cds.), p. 549. Wydawnictwa Uniwersytetu Wroc1awskiego, Wroc1aw, Poland, 1979. 194. P. W. Schiller, C. F. Yam, and J. Prosmanne, J. Med. Chern. 21, 1110 (1979). 195. Y. Shimohigashi, C. H. Stammer, T. Costa, and P. F. Von Voigtlander, Int. J. Pept. Protein Res. 22,489 (1983). 196. Y. Shimohigashi, M. L. English, and C. H. Stammer, Biochem. Biophys. Res. Cornmun. 104, 503 (1982). 197. H. Kimura, C. H. Stammer, Y. Shimohigashi, C. Ren-Lin, and J. Stewart, Biochern. Biophys. Res. Commun. 115, 112 (1983). .. 198. J. V. Castell, A. N. Eberle, V. M. Kriwaczck, A. Tun-KYI, P. Schiller, K. Quang Do, P. Thauei, and R. Schwyzer, He/v. Chirn. Acta 62, 525 (1979). 199. J.-L. Fauchere, S. Pfenniger, K. Quang Do, C. Lemieux, and P. W. Schiller, Helv. Chim. Acta. 66, 1053 (1983). 200. P. D. Gcscllchen and R. T. Shuman, U. S. Patent 4,322,339, March 30, 1982. . 201. P. D. Gcscllchen, R. C. A. Frederickson, S. Tafur, and D. Smiley, in "Pepttdes. Proceedings of the 7th American Peptide Symposium" (D. H. Rich and E. Gross, eds.), p. 621. Pierce Chemical Co., Rockford, Illinois, 1981. 202. R. T. Shuman and P. D. Gesellchen, U. S. Patent 4,322,340, March 30, 1982. 203. B. Filippi, P. Giusti, L. Gma, G. Borin. F. Ricchelli, and F. Marchiori, Int. J. Pept. Protein Res. 14, 34 (1979). 204. J. S. Morley and E. T. Wei, Int. J. Pept. Protein Res. 16, 254 (1980). . 205. R. T. Shuman, P. D. Gesellchen, E. L. Smithwick, and R. C. A. Fredenckson, in "Peptides, Procceedings of the 7th American Peptide Symposium" (D. H. Rich and E. Gross, cds.), p. 617. Pierce Chemical Co., Rockford, Illinois, 1981. 206. J. A. Kiritsy-Roy, S. K. Chan, and E. T. Iwamoto, Ufe Sci. 32, 889 (1983).
References
511
207. R. C. A. Frederickson, R. Nickander, E. L. Smithwick, R. Shuman, and F. H. Norris, in "Opiates and Endogeous Opioid Peptides" (H. W. Kosterlitz, ed.), p. 239. NorthHolland Pub!., Amsterdam, 1976. 208. J. D. Bower, B. K. Handa, A. C. Lane, B. A. Morgan, M. J. Rance, C. F. C. Smith, and A. N. A. Wilson, in "Peptides, Proceedings of the 7th American Peptide Symposium" (D. H. Rich and E. Gross, eds.), p. 607. Pierce Chemical Co., Rockford, Illinois, 1981. 209. S. Bajusz, A. Z. R6nai, J. 1. Szekely, Z. Dunai-Kovacs, I. Berzctei, and L. Graf, Acta Biochem. Biophys. Acad. Sci. Hung. II, 305 (1976). 210. L. Kupczyk-Subotkowska and P. Mastalerz, Int. J. Pept. Protein Res. 21,485 (1983). 211. P. Mastalerz, L. Kupczyk.Subotkowska, Z. S. Herman, and G. Laskawiec, Naturwissenschaften 69, 46 (1982). 212. M. Szucs, S. Benyhe, A. Borsodi, M. Wollemann, G. Jancs6, J. Szecsi, and K. Medzihradszky, Ufe Sci. 32, 2777 (1983); M. Szucs, Drugs Future 9, 416 (1984). 213. Y. Shimohigashi and C. H. Stammer, J. Chern. Soc., Perkin Trans. I p. 803 (1983). 214. K. Rolka, M. Kruszynski, and G. Kupryszewski, Acta Pharm. Suec. 21, 173 (1984). 215. R. B. Mathur, B. J. Dhotre, R. Raghubir, G. K. Patnaik, and B. N. Dhawan, LifeSci. 25, 2023 (1979). 216. C. B. Pert, D. L. Bowie, A. Pert, J. L. Morell, and E. Gross, Nature (London) 269, 73 (1977). 217. E. F. Hahn, J. Fishman, Y. Shiwaku, F. F. Foldes, H. Nagashima, and D. Duncalf, Res. Cornmun. Chern. Pathol. Pharmacal. 18, 1 (1977). 218. J. S. Shaw, L. Miller, M. J. Turnbull, J. J. Gormley, and J. S. Morley, Ufe Sci. 31, 1259 (1982). 219. J. J. Gormley, J. S. Morley, T. Priestly, J. S. Shaw, M. J. Turnbull, and H. Wheeler, Life Sci. 31, 1263 (1982). 220. P. Bclton, R. Cotton. M. B. Giles, J. J. Gormley, L. Miller, J. S. Shaw, D. Timms, and A. Wilkinson, Ufe Sci. 33, Supp!. 1, 443 (1983). 221. R. Cotton, M. G. Giles, L. Miller, J. S. Shaw, and D. Timms, Eur. J. Pharmacol. 97, 331 (1984). 222. W. Werner and M. Puig, Drugs Future, 5, 81 (1980); update: 7, 129 (1982). 223. M. Puig, Drugs Future 3,511 (1978); updates: 4, 529 (1979); 5, 368 (1980); 6, 443 (1981); 7, 523 (1982); 8, 648 (1983); 9, 548 (1984). 224. D. M. Paton. Deugs FUMe, 4, 711 (1979); updates: 6, 648 (1981); 8, 897 (1983). 225. R. C. A. Frederickson, E. L. Smithwick, R. Shumam, and K. G. Bemis, Science 211, 603 (1981). 226. R. C. A. Frederickson, E. L Smithwick, and D. P. Henry, Int. Brain Res. Organ. Manage. See. 7, 227 (1980). 227. J. F. Calimlim, W. M. Wardell, K. Sriwatanakul, L. Lasagna, and C. Cox, Lancet p. 1374 (1982). 228. A. Pazos and J. Fl6rez, Eur. J. Pharmacal. 99, 15 (1984). 229. N. Nedapil and E. Ruther, Pharmakopyschiatr. Neuro-Psychopharmakol. 12, 277 (1979). 230. J. Brownell, E. del Pozo, and P. Donatsch, Acta Endocrinol. 94, 304 (1980). 231. Y. Kato, S. Hirota, H. Katakami, N. Matsushita, A. Shimatsu, and H. Imura, Proc. Soc. Exp. Bioi. Med. 169, 95 (1982). 232. B. von Graffenried, E. del Pozo, J. Roubicek, E. Krebs, W. Poldinger, P. Burmeister, and L. Kerp, Nature (London) 272, 729 (1978). 233. G. Stacher, P. Bauer, H. Steinringer, G. Schmierer, B. Langer, and S. Winklehner, Gastroenterology. 83, 1057 (1982).
11
5\2 234.
H. B. Andersen, (1982).
235.
G. Stacher, (\979).
236.
J. 1. Szekely, A. Z. R6nai, Z. Dunai-Kovacs, E. Miglecz, I. Berzetri, S. Bajusz, and L. Graf, Eur. J. Pharmacal. 43,293 (1977). S. Bajusz, Z. R6nai. J. I. Szekely. L. Graf. Z. Dunaj-Kovacs, and I. Bcrzetei. FEBS Letl. 76, 91 (1977).
237.
B. C. Jorgensen.
P. Bauer,
and A. Engquist,
Enkephalins
H. Steinringer,
238.
E. Miglccz, (1979).
1. I. Szekely.
239.
J. Foldes, K. Torok, J. I. Szekely. Suppl. I, 769 (1983).
E. Schreiber,
and Z. Dunai-Kovacs, J. Borvcndeg,
Acta Anaesthesia/.
Scand.
and G. Schmierer,
Psychopharmacology
26,69
12.
Pain 7, 759
(Berlin)
I. Karczag. and J. Tolna,
62, 29
Life Sci. 33,
Physical Chemistry and Molecular Modeling of the Enkephalins I. Introduction . . . . . . II. Solid-State Conformations A. X-Ray Crystallography.. . ...... B. Laser Raman, Infrared, and Solid-State NMR Spectroscopy III. Solution Conformations . . . . A. Studies in Organic Solvents B. Studies in Aqueous Solution IV. Molecular Modeling Studies. V. QSAR Studies References
5\3 5\4 514 515 516 516 525 532 537 537
1. Introduction The enkephalins have numerous degrees of conformational freedom, suggesting Ihat they may exist in many conformalional states. Extensive effort has gone into the study of the solid-state, solution, and free-space conformations of these compounds, with the expeclation Ihat a low-energy conformation can be identified that is closely related to the "active" conformation. Such a conformation would be useful in relating the structural features of enkephalins to those of other opiate-type compounds and could be utilized in the design of new analgesics. Since il is not yet feasible to determine the conformalions of these compounds while they are bound to their receptors, conformalional information has been taken from crystallographic sludies, spectroscopic investigalions, and conformational energy calculations. Each of these melhods has inherent limitations, and in all cases one may question the relevance to the receptor-site situalion. Nevertheless, such studies provide some indication of the range of conformations possible, and they often inndicate types of intramolecular and intermolecular interactions that should be considered. Two excellent reviews have been published on conformation-activity relationships in the enkephalin series (1,2). Conformations of peptides may be described at two levels: the conformation of the peplide backbone and the rotamers of the side chains. Unless otherwise specified, the following sludies presume the existence of trans-amide bonds. The conventions used to describe side chain rotamers are illuslrated in Fig. 12-1. 513
\14
12 Physical Chemistry and Molecular Modeling of the Enkephalins
H H-,CxH -HN~COR t I (g+t J to describe
{19-1 Fig. 12-1.
II.
Conventions
used
III 19-£1") side chain rotamers.
Solid-State Conformations
Several X-ray crystallographic studies of enkephalins and some of their analogs have appeared. In addition, laser Raman and solid-state nuclear magnetic resonance (NMR) spectroscopy have been employed In examIning the cooformational behavior of these compounds. A,
X-Ray Crystallography
Smith and Griffin (3) reported the first crystallographic study of [Leu]enkephalin. They found a Tyr-Gly-Gly-Phe j3-bend conformation (type I'J with hydrogen bonds linking the Tyr-C=O to Phe-NH and
Tyr-NH3 + to Phe-C=O
(Fig. 12-2). These workers reported some dIS-
order in the tyrosine side chain position, with two apparent conformers present. The tyrosine Ca-C~ bond is close to rotamer I of Fig. 12-1, which is similar to the tyramine conformation observed in the crystal structure of morphine (4). However, neither tyrosine conformation observed in [Leu]enkephalin showed a C~-Cy rotation close to that of morphine. The phenylalanine side chain conformation corresponded to rotamer I, as dId the leucine side chain. It was pointed out that the Gly-Gly conformatIOn
+
II
I
Solid-State Conformations \1\
observed in the crystal structure is sterically not available to L-amino acids, although a D-amino acid in position 2 could be accommodated. Subsequent to this work, Blundell et al. (5) reported a different space group for this crystal structure in which the tyrosine side chains were not disordered but exhibited four separate conformations in the unit cell. These four molecules have nearly identical backbone conformations but show two different tyrosine side chain orientations. The remainder of the structure is essentially as described by Smith and Griffin. Karle and co-workers (6,7) examined another crystallographic form of [Leu]enkephalin. The unit cell of this form contains four molecules of [Leu]enkephalin and 40-50 molecules of solvent (water and dimethylformamide). All four molecules have essentially extended backbone conformations. Some differences were found in the conformations of the tyrosine and leucine side chains. For the tyrosine side chain, two molecules were in conformation II and two were in conformation II!. All four phenylalanine side chains exhibited conformation I. The leucine side chain was found in rotamer I in three cases and rotamer II in the other. Japanese workers (8,9) determined the structures of [Metjenkephalin, [4' -Br-Phe4j-[Met]enkephalin, and [4' -Br-Phe4j-[Leu]enkephalin. They observed the latter compound in essentially the same conformation as described by Smith and Griffin (3) for [Leu]enkephalin, with the Tyr-GlyGly-Phe j3-bend and two different tyrosine conformations. The [Met]enkephalin structures could not be solved at high resolution, but lowresolution Patterson maps gave evidence of essentially extended peptide backbones in both compounds. These workers subsequently determined the structure of the tert-butyloxycarbonyl derivative of [4'-Br-Phe4j_ [Met]enkephalin (10), a compound that retains some analgesic activity. In this case, the backbone was found in an extended conformation. The tyrosine side chain exhibited conformation I, the phenylalanine side chain conformation II, and the methionine conformation I. The tetrapeptide fragments Tyr-Gly-Gly-Phe and Gly-Gly-Phe-Leu have been examined crystallographically (11.12). Tyr-Gly-Gly-Phe adopted a j3-bend structure similar to that found by Smith and Griffin, but Gly-Gly-Phe-Leu showed no intramolecular hydrogen bonds.
H' N H-,
~ 6
B.
Laser Raman, Infrared, and Solid-State NMR Spectroscopy
l.u
Fig. 12-2.
Tyr-Gly-Phe
,B-bend conformation.
Han and co-workers (/3) obtained the laser Raman spectra of [Leu]enkephalin, four different l3C-labeled [Leu]enkephalins, mid several model compounds. These workers interpreted the spectroscopic results in terms of the type I' j3-bend reported by Smith and Griffin (3), as
516
12
Physical
Chemistry
and
Molccular
Modeling
III
of the Enkephalins
shown in Fig. 12-2. The same group used infrared spectroscopy to show that [Leu]enkephalin exists as the zwitterion in the solid state (14). Tyrosine labeled with deuterium at positions 3 and 5 of the aromatic ring was used by Rice et at. (15) to prepare labeled [Leujenkephalin. The deuterium-labeled enkephalin was studied with deuterium NMR in the solid state in order to derive information about the tyrosine ring motions. The tyrosine ring was found to execute 1800flips around the CwC, bond at a rate of 50,000/sec.
III.
517
I
"" ""
H
~ II
51
0
'H 00,cyNi 1=1
Fig. 12-3.
0
H
H I,
:,'0
Solution Conformations
OH
NJ~y
H,~
Numerous and sometimes conflicting reports have appeared in the literature regarding the solution conformation(s) of enkephalins. Discrepancies may arise from differences in the methods used, as well as from variables such as sample concentratjon, solvent, pH, temperature, ionization state, and buffers. The majority of studies employed NMR spectroscopy, but ultraviolet, circular dichroism, infrared, and fluorescence methods have also been used. The following discussion will be divided into two parts: first, studies carried out in organic solvents or aqueous-organic mjxed solvents, and second, studies carried out on enkephalins in aqueous solution. In each of these sections, backbone conformation will be examined first, followed by side chain conformations. A.
Solution Conformations
0
b _H
Postulated
0' 0
~ /,
HO dimeric
~ACO,0 H' i.
° H , ~N1f'~
f
R
form
NH,"
H-..;;::
\
0 of enkephalins.
Mct-NH resonance, they proposed that the peptide exists in a type I l3-bend conformation, as shown in Fig. 12-4. Such a conformation should be stabilized by the Met-NH to Gll-c=o hydrogen bond as well as the interaction of the charged end groups. The proposed folded conformation was further supported by the measured 13C spin-lattice relaxation times, which indicated that the internal a-carbons have about the same mobility as the a-carbons of the terminal groups. Similar results were observed when the solvent was a I: I mixture of dimethylsulfoxide and water. Similarly, Jones et at. (20) found the Met NH proton to have a very low temperature dependence in dimethylsulfoxide solution. These workers also proposed a type I l3-bend conformation, as shown in Fig. 12-4. In contrast to the above findings, Bleich and co-workers (21) found no evidence for hydrogen bonding from the temperature dependence of the NH protons of [Met]enkephalin in dimethylsulfoxide solution. All amide protons exhibited temperature coefficients well above the value expected for hydrogen-bonded species. This discrepancy was resolved by Jones et at. (22), who demonstrated that the findings of Bleich and co-workers correspond to the behavior of
Studies in Organic Solvents
Khaled and co-workers (16) showed that the spectra of enkephalins are concentration dependent in many organic solvents (dimethylsulfoxide, trifluoroethanol, methanol) using ultraviolet, circular dichroism, proton NMR, and carbon-13 NMR spectroscopy. At low concentration (1 mM), enkephalins are expected to exist in monomeric form, while dimers and higher associated forms may exist at higher concentrations (100 mM). Khaled et at. interpreted their results in terms of an extended anti parallel l3-pleated sheet structure at high concentration (Fig. 12-3). The concentration dependence is a particularly important consideration in NMR studies because NMR tends to be a less sensitive method, and high sample concentrations are often used to overcome the limitation in signal sensitivity. The first NMR studies of an enkephalin reported were those of Roques and co-workers (17-19). They carried out variable-temperature studies on (Met]enkephalin in dimethylsulfoxide. On the basis of coupling constants, conformational energy maps, and the low temperature dependence of the
OH
~ ~"YNl H'~~QCH ~
~~
''yN~ R
Fig. 12-4.
1
Gly-Gly-Phe-X
~I '"
0 l3-bend conformation.
518
12
Physical
Chemistry and Molecular Modeling of the Enkephalins
III Solution Conformations
the cationic form in dimethylsulfoxide solution. while the ,,-bend conformation is characteristic of the zwitterionic (or neutral) form of [Met jenkephalin. Han et al. (/4) addressed the question of the ionization state of [Leu]enkephalin in dimethylsulfoxide solution. On the basis of infrared studies. they concluded that there is a mixture of zwitterionic and neutral species in dimethylsulfoxide. From this, they inferred that, in dimethylsulfoxide, the carboxylate and amino groups have similar pK, values. The conformation of [Leu]enkephalin in dimethylsulfoxide solution initially appeared to he less well defined. Garbay-Jaureguiberry er al. (23) considered both the type I and type II' ,,-bend conformations to be consistent with the spectral data. In either case, a folded conformation appears to be highly favored under these conditions. Fournie-Zaluski er al. (11) investigated the conformations of the tetrapeptide fragments Tyr-Gly-Gly-Phe, Gly-Gly-Phe-Leu, and Gly-GlyPhe-Met in dimethylsulfoxide solution using proton NMR spectroscopy. Temperature dependence of the amide protons in each case suggested that the carboxyl-terminal NH was involved in an intramolecular hydrogen bond. These results and analysis of the backbone proton coupling constants indicated that all three compounds exist in a type I ,,-bend conformation. [Leu ]enkephalins labeled with carbon- 13 in the carbonyl carbons were prepared by Stimson er al (24). Using carbon and proton NMR, they found a low temperature dependence for the Leu'-NH and the Gly2-CO, in support of a hydrogen-bonded ,,-bend structure. In constrast to the work of Khaled er al. (/6, see above), these workers found no evidence of a concentration dependence. The values of the carbon-13 spin-lattice relaxation times of all of the carbonyl carbons were essentially identical, lending further credence to the existence of a folded structure. With the carbon-13 labels present, these workers were able to measure carbon-proton coupling constants, from which they ruled out type I', type II. and type II' ,,-bend conformations; they interpreted their data in terms of a type I ,,-bend. Han et al. (13) used laser Raman spectroscopy to examine the conformation of [Leu]enkephalin and carbon-I3 labeled derivatives in dimethylsulfoxide solution. Analysis of the amide bands indicated the presence of a "-bend conformation, but it was not possible to determine from these experiments the type of ,,-bend(s) present. A group of French workers (25,26) prepared [Leu]enkephalin enriched in nitrogen- 15. Using nitrogen-IS NMR in dimethylsulfoxide solution, they observed that nitrogen spin-lattice relaxation times were consistent with a folded structure. On the basis of nitrogen-proton coupling constants, they
519
ruled out a type I ,,-bend conformation and proposed that a type II' ,,-bend is favored in solution. These conflicting results may be due to the fact that the measured proton-proton, carbon-proton, and nitrogen-proton coupling constants were interpreted as though they arise from a single conformation, whereas there may be multiple conformations in solution that interconvert rapidly. If the latter is occurring, th'\, coupling constants observed will be a Boltzmann weighted average over all of the conformations present. In fact, several authors have argued for the existence of multiple conformations in dimethylsulfoxide solution. [Leujenkephalins having stereospecific deuterium labels on the glycine residues were prepared by Fischman er al. (27). Using proton-proton coupling constants. they found no evidence for a single preferred conformation in solution and suggested that conformational averaging is probably taking place. Higashijima and co-workers (28-30) have carried out extensive proton and carbon NMR studies of [Met]enkephalin and [Met]enkephalinamide in dimethylsulfoxide solution as a function of temperature and concentration. Like Khaled et al. (/6), these workers ohserved significant concentration dependence of the NMR spectra. Concentration-dependent shifts were largest for the Tyr1 and Met' resonances. In variable-temperature studies, these workers confirmed the low temperature dependence of the Met'-NH proton. but they pointed out that the temperature dependence is nonlinear. Furthermore, several of the a-protons also exhibit temperature dependence. These anomalous temperature dependencies were attributed to aromatic ring current effects occurring in dimeric forms (e.g., Fig. 12-3). In addition to dimeric forms, these workers found evidence for a folded conformation in dimethylsulfoxide. When Gd(III) was used as a relaxation probe, it was bound to the C-terminal carboxylate group; the distancedependent relaxation was greater for the Tyr1 residue than for the Gly2, Gly3, and Phe4 residues. Further, when ammonium perchlorate was added to disrupt the intramolecular head-to-tail attraction the Gd(III) relaxation followed the order Met' > Phe4 > Gly3 > Gly2 > Tyr1. They concluded that there exists in solution a group of conformations that interconvert rapidly. A similar conclusion was drawn by Oi Bello and co-workers (31) in their proton NMR study of' enkephalin analogs containing y-aminobutyric acid (GABA). Analogs incorporating GABA into position 2 or position 5 showed evidence of both folded and extended conformers in dimethylsulfoxide solution. More recently, Renugopalakrishnan er al. (32) published studies on [Leu]enkephalin and [Metjenkephalin in dimethylsulfoxide solution
1
\10
12
Physical
Chemistry
and Molecular
Modeling
of the Enkcphalins
III Solution Conformations
OH
9'
the N-terminal X- Tyr segment is sufficiently flexible to permit interaction with the receptor. The enkephalin anal~g ,!,yr-o-Ala-Gly-Phe-Met was examined by Niccolai et ~I. (36) using proton NMR in dimethylsulfoxide solution. Proton spin-lattice relaxation rates indicated that this compound has a relatively rigid backbone. Nuclear Overhauser enhancement measurements showed that the methyl groups of o-Ala and Met are in close proximity in solution, supporting ~he existence of a f3-bend conformation. Kessler and Holzemann (37,38) conducted proton NMR studies of cyclo-[Leu]enkephalin and cyclo-[Met]enkephalin in dimethylsulfoxide solution. On the basis of the temperature dependence of amide protons, proton-proton coupling constants, and studies on model compounds, they proposed a y, y-conformation (Fig. 12-6). Furthermore, they suggested that acyclic enkephalins might adopt a similar conformation in solution. Unfortunately, these workers were unable to determine whether these cyclized compounds have biological activity due to their poor water solubility. Roques and co-workers (39) obtained evidence for a folded conformation in the analog Tyr-o-Met-Gly-Phe-Pro In dimethylsulfoxide solutIOn. This compound ha~ no amide NH on the fifth amino acid residue to hydrogen bond. A strong head-to-tail interaction was suggested by the observations that titration of the amino group strongly affected the Pro' resonances and that titration of the carboxylate strongly affected the Tyr' resonances. A number of JL-selective enkephalin analogs were studied by FournieZaluski et al. (40). These compounds generally had a o-amino acid in position 2, and the fourth and fifth residues were replaced by hydrophobic amide groups. Proton NMR spectra in dimethylsulfoxide solution indicated
"" H
@_H,N N, COp ~N,H
CH,S) 0'
Fig. 12-5. Folded (reference 34).
conformation
~"O,'.. /::0 H'N
~-i
of [MetJenkephalin
o proposed
by Zctta
and Cabassi
using infrared and laser Raman methods. Their interpretation differed from that of Han et al. (I3, see above). The amide 1 band that Han et al. attributed to a {Hurn is proposed by Renugopalakrishnan and co-workers to represent a f3-sheet structure (i.e., extended conformation, associated as shown in Fig. 12-3). One may conclude that infrared and Raman spectra are not sufficiently definitive (or not sufficiently well understood at present) to permit detailed conformational conclusions to be drawn. Anteunis and co-workers (33) carried out a proton NMR study of [Met]enkephalin in dimethylsulfoxide-water (2:1). They found that titration of the amino group affected the resonances of the methionine residue. Titration of the carboxylate group affected the resonances of the tyrosine residue. These findings suggest that there exists a folded conformation stabilized by head-to-tail interactions at intermediate pH. Zetta and Cabassi (34) also examined [Met]enkephalin in dimethylsulfoxide and dimethylsulfoxide-water mixtures by proton NMR. When the dimethylsulfoxide concentration was 60 mol% or greater, the Met5-NH proton was found to be intramolecularly hydrogen bonded and the Gly3-NH to be shielded from solvent. These workers proposed the conformation shown in Fig. 12-5 to account for the observed spectral properties. The hexapeptides Tyr- Tyr-Gly-Gly-Phe-Met, Phe- Tyr-Gly-Gly-PheMet, Lys-Tyr-Gly-Gly-Phe-Met, and Gly-Tyr-Gly-Gly-Phe-Met all retain significant opiate receptor binding and analgesic activity that appears to be due to the intact peptide (35). These compounds all show proton NMR evidence (backbone proton-proton coupling constants and Met-NH temperature dependence), which supports a f3-bend conformation for the Gly-Gly-Phe-Met segment in dimethylsulfoxide solution. This suggests that the relatively rigid f3-bend portion can still bind to the receptor, while
\11
HO
,
1
Fig. /2-6. 'Y,y-Conformation (references 37, 38).
of cyclo-enkephalins
proposed
by Kessler
and Holzemann
522
12
Physical
Chemistry
and Molecular
Modeling
of the Enkephalins
III
a highly folded conformation for the I"-selective compounds. These authors considered the folded form to be similar in overall shape to morphine. Gairin ef al. (41) carried out proton NMR studies in dimethylsulfoxide solution on Tyr-o-Ala-Gly-Phe-Nva, Tyr-o-Ala-Gly-Phe-Met, and their C-terminal
ami des
(Nva
=
norvaline).
The
amides.
which
Solution Conformations
523
leucine side chains appeared to have significant internal rotational freedom. For the analog Tyr-o-Met-Gly-Phe-Pro in dimethylsulfoxide, Roques and co-workers (39) determined side chain rotamer populations from proton NMR coupling constants. They found all three rotamers populated for the tyrosine side chain, while rotamer I predominated to the extent of about 60% for the phenylalanine slOe chain. Again, carbon-13 spin-lattice relaxation times indicated limited rotation for the tyrosine side chain. Using nitrogen-15 NMR and nitrogen-15 labeled [Leu]enkephalins, Garbay-Jaureguiberry and co-workers (26) studied the rotameric states of the phenylalanine and leucine side chains. On the basis of protonnitrogen, proton-carbon, and proton-proton coupling constants, these investigators found the following ratios of rotamers 1111I11I:phenylalanine, 65/10/25; leucine, 95/015. Stimson ef al. (24) carried out a similar study of [Leu ]enkephalin in dimethylsulfoxide using proton and carbon NMR. They did not distinguish between conformations I and", but they found that these two conformers predominated over rotamer III for all three side chains in [Leu]enkephalin. Rotamers I and" accounted for 80% of the rotamer population of the tyrosine residue, 90% of the phenylalanine rotamers, and 100% of the leucine rotamer population. In contrast, Jones ef al. (20) found substantial populations of all three rotamers for the methionine side chain in [Met]enkephalin. Using protonproton coupling constants, these workers found the ratio of 1111I11I to be about 43/40117. Niccolai ef al. (36) used proton NMR to probe the side chain conformations of Tyr-o-Ala-Gly-Phe-Met in dimethylsulfoxide solution. They reported that the o-alanine, phenylalanine, and methionine side chains have substantial internal motion, while the tyrosine side chain is not freely rotating. Kobayashi and co-workers (46-49) have carried out extensive proton NMR studies on [Met]enkephalins having deuterium label(s) on the Cland {3carbons of the tyrosine and phenylalanine residues. Using these labeled compounds, they carried out an analysis of the tyrosine and phenylalanine side chain rotamers as a function of temperature and of solvent polarity. Fig. 12-7 shows the effect of temperature on the side chain rotamer populations in dimethylsulfoxide solution. As expected, increasing temperature increases the populations of less stable conformers at the expense of more stable ones in the case of phenylalanine. However, the most stable tyrosine rotamer became more populated at higher temperature. These workers attributedthis anomalous behavior to a change in the equilibrium between folded and extended conformations with change in temperature.
showed
selectivity, gave evidence of having different solution conformations andI" more conformational rigidity than their parent compounds. Measurement of energy transfer between donor and acceptor fluorophores permits an estimate of the intramolecular distance between the two groups. In the case of the enkephalins, such experiments provide an estimate of the average distance between the two aromatic rings. Such fluorescence measurements have been carried out in aqueous solution, for the most part, but Reig ef al. (42) found a phenyl-phenyl distance of 9.4 A for ethanol solutions of Tyr-o-Pro-Gly-Phe-Leu-OEt and Tyr-o-Pro-GlyPhe-Met-OEt. Garcia-Anton and co-workers (43) obtained a distance of 10.8 A for Tyr-o-Met-Gly-Trp-Pro in ethanol. Such distances require at least some folding of the peptide backbone. [Met ]enkephalin in trifluoroethanol solution was examined using ultraviolet and circular dichroism spectra (44). Lack of temperature dependence was interpreted to mean that there is no ordered structure under these conditions. In addition, the pH dependence of the spectra was examined; this study indicated that there is no evidence for involvement of the tyrosine hydroxyl group in intramolecular hydrogen bonding. Numerous studies have been made of side chain conformations of the enkephalins in organic solvents. On the basis of spin-lattice relaxation time measurements, Bleich ef al. (21,45) concluded that the tyrosine side chain of [Met]enkephalin is conformationally restricted, whereas the phenylalanine and methionine side chains undergo rapid motion in dimethylsulfoxide solution. Combrisson and co-workers (19) came to a similar conclusion for the same compound in a I: I mixed dimethylsulfoxide-water solvent system. In their study of [Leu]enkephalin in dimethylsulfoxide, GarbayJaureguiberry ef al. (23) concluded from proton-proton coupling constants that all three rotamers of the tyrosine side chain coexist in solution, with rotamer I predominating (see Section I for the conventions used to describe side chain rotamers). Similarly, they found rotamer I to predominate for the phenylalanine side chain, with lesser amounts of rotamers and III. However, they concluded that only rotamer [ was significantly " populated for the leucine side chain. On the basis of carbon-13 spin-lattice relaxation times, they determined that the tyrosine aromatic ring does not undergo reorientation around the Cy-C, axis; the phenylalanine and
1
\24
Physical Chemistry and Molecular
12
of the Enkephalins
Modeling
III
B.
ill
n
40
60
80
100 Temperature
Fig. 12-7. of
Effect
[Metjenkephalin
of temperature
on the rotamer
in dimethylsulfoxide
solution
20
40
60
populations
80
100
of Tyrl and Phe4 side chains 47,
48).
Fig. 12-8 shows the effect of a change in solvent polarity (reported as the logarithm of the dielectric constant) on the rotamer populations in these compounds. Among the phenylalanine rotamers, conformation 111constitutes about 15% of the total throughout the range of polarities studied. The major conformer (I) increases in importance as the solvent polarity increases, with a concomitant decrease in the population of conformer II. For the tyrosine rotamers, conformation 111 again makes up a fairly constant percentage of the total population (about 20% in this case). Unlike phenylalanine, however, the rotamer I population decreases as solvent polarity increases, until rotamers I and II are present in about equal proportions in aqueous solution. Overall, the evidence regarding the conformation of enkephalins in organic solvents is probably best interpreted in terms of multiple con-
1.0
,.
~0.8
.3
~0 ~.
II ~I
0.4
'"" 0.2
'r~
'n
ill
5
10
15
20
'ill~~ o
side
,Effect of a solvent di:le~tric .constant on the chams of fMet)enkephahn In dimethylsulfoxide
0
.
1
2
log €
log € PhF~g. ,12-8.
e
~
0.8 05
rota mer
solution
525
populations
(references
of Tyrl
Studies in Aqueous Solution
Numerous studies (e.g., 16,21,30,50;51) using NMR, ultraviolet, circular dichroism, and fluorescence spectroscopy have indicated apparent conformational differences on going from organic to aqueous solvent conditions. One may argue that the aqueous environment is more relevant to the physiological situation, although the extent of solvation of the molecule at its receptor site is open to question. Many of the same techniques that were described in the previous section for enkephalins in organic solvents have also been applied to aqueous solutions. In addition, interactions with metal ions and lipid systems have been investigated. An important consideration when studying conformational behavior in aqueous solution is the ionization state of the molecule. As would be expected, the neutral form of the enkephalins was shown (using infrared and laser Raman spectroscopy) to exist as the zwitterion in aqueous solution (14). This result was in contrast to the finding (see above) of both uncharged and zwitterionic forms in dimethylsulfoxide solution. Directly related to the question of ionization state is the dissociation behavior of the acidic and basic groups in the molecule. A number of pKa determinations have been reported for [Metjenkephalin. For the Nterminal amino group, values of 7.75 (28), 7.2 (52), and 7.7 :t 0.3 (53) have been reported. Carboxylate pKa values reported include 3.50 (28), 3.55 (52), 3.2:t 0.3 (53), and 2.8 (54). The tyrosine hydroxyl ionization was reported to occur at 10.2 (50), 10.4 (52), 1O.2:t 0.2 (44), and 1O.5:t 0.3 (53). Fischman et al. (27) carried out proton NMR studies on [Leu]enkephalins having stereospecific deuterium labels on the glycine" carbons. On the basis of proton-proton coupling constants, they argued that there are probably many conformations rapidly interconverting in aqueous solution and that the observed coupling constants are the result of conformational averaging. Spirtes and co-workers (44) drew a similar conclusion regarding [Met]enkephalin in aqueous solution. The observation that the circular dichroism spectrum was not affected by temperature suggested the absence of an ordered structure. Using laser Raman spectroscopy, Han et al. (13) also concluded that multiple conformations exist in aqueous solution. Studying [Met]enkephalin in aqueous solution by proton and carbon NMR, Higashijima and co-workers (28) found that titration of the
(oC)
(references
Conformations
formations, including one or more folded forms. The results give some indication of the conformational possibilities but do not permit identification of an "active conformation."
I
20
Solution
and
46, 48).
i
526
12
Physical
Chemistry
and Molecular
t\.1odeling of the Enkcphalins
N-terminal amino group did not affect resonances of the methionine residue and that titration of the methionine carboxylate did not affect tyrosine resonances. These results indicate that the two end groups are not in close proximity in solution and that monomeric extended forms predominate under these conditions. Furthermore, none of the amide NH resonances gave any evidence of intramolecular hydrogen bonding or shielding from solvent. In a natural abundance nitrogen-IS NMR study of [Metjenkephalin, Higuchi et al. (54) also carried out a titration of the methionine carboxylate group. This titration had no effect on the tyrosine nitrogen chemical shift, again suggesting that the end groups arc not close together in aqueous solution. Miyazawa and Higashijima (30) provided further evidence for an extended conformation in aqueous solution. Spin-lattice relaxation rate enhancement by Gd(lII) was found to be greatest for methionine protons [where the Gd(III) was bound to the carboxylate group] and to decrease continously for protons on Gly3, Gly', and Tyrl Since relaxation enhancement is dependent on distance through space, it would appear that extended forms predominate. Zetta and Cabassi (34) examined the proton NMR spectra of [MetJenkephalin in water, dimethylsulfoxide, and mixed solvents. They found extended conformations predominant in aqueous solution and in waterdimethylsulfoxide mixtures containing 40 mol% or more of water. Infrared and laser Raman spectra of [Leulenkephalin and [Met lenkephalin in aqueous solution were examined by Renugopalakrishnan and co-workers (32). On the basis of comparisons with spectra of model compounds, they suggested that [Met]enkephalin exists in a ,B-sheet (extended) conformation and that [Leu]enkephalin exhibits both ,B-sheet and type II ,B-bend conformations. They suggest that such conformational differences may be related to receptor subtype selectivity. In contrast to these indications that enkephalins adopt mainly extended conformations in aqueous solution, other workers have interpreted their results in terms of folded conformations. Jones et al. (20), on the basis of proton-proton coupling constants, proposed that [Met]enkephalin has qualitatively the same folded conformation in water as in dimethylsulfoxide. Tancrede et al. (55,56) measured spin-lattice relaxation times for [Met]enkephalins having the carbon-13 label in Gly2 or Gly3 They found that Gly3 motion was more restricted than that of Glyl, the opposite of what would be expected for an extended, unordered conformation. Using paramagnetic probes [Gd(lII) bound to the carboxylate group and Cr(CN). -3 bound to the amino group], Levine et al. (53) measured average distances between the probe and protons on [Met]enkephalin.
III
Solution
Conformations
m
They compared these measured average values with calculated values for a fully extended conformation and found the measured distances to be up to 0.8 A shorter. This observation indicated to them that folded conformations contribute to the overall conformational population in aqueous solution, even though no head-to-tail or intramolecular hydrogen bond interactions could be demonstrated directly. Zetta et al. (57) were also unable to find direct evidence for intramolecular hydrogen bonds or head-to-tail interactions, but they found that the Gly-Phe-Met segment had less conformational freedom than the Tyr-Gly segment, again consistent with the existence of folded conformations. Fluorescence spectroscopy has been applied extensively to aqueous solutions of enkephalins and enkephalin analogs. This approach permits an assessment of the average distance separating two aromatic groups in a molecule. For [MetJenkephalin and [LeuJenkephalin, Kupryszewska et al. (58) observed both sensitized fluorescence of the acceptor and quenching of donor fluorescence. From these measurements, they calculated average phenyl-phenyl distances ranging from 7.5 to 8.7 A, depending on the method used. Schiller and co-workers (59-63) have carried out fluorescence measurements on several [Trp4]enkephalins, which retain substantial analgesic activity. Intramolecular Tyr- Trp distances in these compounds ranged from 8.8 to 10.7 A, with a mean value of 9.3:t 0.2 A. This result is consistent with either a single folded conformation in solution or an averaged value for many solution conformations, including significant contributions from folded conformers. In either case, the results are consistent with one or more folded conformations, that is, the peptides are not fully extended all the time in aqueous solution. Since the 9.3-A average distance is maintained even at pH 1.5, the head-to-tail ionic attraction between the charged end groups is not required to stabilize the folding. Other workers have found similar results using fluorescence spectra. Garcia-Anton
el al. (43) found a Tyr- Trp distance
of9.0
A for
Tyr-D-Met-
Gly-Trp-Pro-NH2 in water. Using [4'-NH,-Phe4]enkephalin, Siemion and Szewczuk (64) observed a phenyl-phenyl distance of 10 A. Guyon-Gruaz (65) and co-workers carried out similar measurements on enkephalin derivatives having C-terminal dansyl substituents. They obtained tyrosinedansyl distances in water and in trifluoroethanol solution of 12.4-14.9 A, which again requires at least some contribution from folded conformers. Schiller (60) reported, on the basis of fluorescence spectra, that the tyrosine hydroxyl group is not involved in hydrogen bonding, since the phenol fluorescence is not quenched in [Trp4]enkephalins. On the other hand, Filippi et al. (66) compared fluorescence yields of [Leu]enkephalin
528
12
Physical
Chemistry
and
Molecular
Modeling
of the Enkcphalins
and its Ser'-analog to those of the model compound tyrosinamide and concluded that the decreased fluorescence (10-15% lower) strongly supported the presence of intramolecular hydrogen bonding involving the hydroxyl group. Soos et al. (67) considered the aqueous solution conformation of enkephalins to consist of a mixture of random and J3-bend conformers. They determined the circular dichroism spectra of [Met]enkephalin and of the fragments Tyr-Gly-Gly and Gly-Gly-Phe, then subtracted the spectra of the latter two fragments (which presumably have random conformations), and considered that the remaining CD spectrum was due to J3-bend conformation. Repeating the CD measurements for a series of five en kephalin analogs, these workers found a correlation (correlation coefficient
=
0.94)
between
the "J3-bend
content"
and the measured
IDso
values in guinea pig ileum preparations. Interestingly, there was not significant correlation with activity in mouse vas deferens. These authors suggested, on this basis, that the J3-bend conformation is required for activity in the guinea pig ileum receptor population, but not for the mouse vas deferens receptor population. Overall, the evidence amassed to date appears to be most consistent with the interpretation that the enkephalin backbone is highly flexible in aqueous solution. Both extended and folded forms probably exist in equilibrium. Side chain conformers have also been studied extensively in aqueous solution. Khaled et al. (68), using carbon-13 NMR, observed that the tyrosine y-atom showed unusual titration behavior at high pH. They attributed this result to a conformational change in the tyrosine side chain at higher pH. On the basis of spin-lattice relaxation times, proton exchange rates, and temperature and pH effects in the NMR spectra of deuteriumlabeled enkephalins, Bleich and co-workers (21,69) determined that the tyrosine side chain motion is quite rigidly fixed with respect to the peptide main chain. On the other hand, they found that the phenylalanine and methionine side chains undergo rapid motion. Kobayashi and co-workers (46-49), using deuterium-labeled [Met]enkephalin, determlOed the proportions of rotamers I, II, and III for the tyrosine and phenylalanine side chains of [Met]enkephalin in aqueous solutIOn. FIg. 12-9 illustrates the existence of all three rotamers of tyrosine and phenylalanine over a wide range of pH values. Circular dichroism and NMR spectroscopy have been applied to the study of the solution conformation(s) of larger peptide analgesics such as the endorphins, dynorphin, and lipotoropin (see Chapter II for structures). UslOg carbon-I3 NMR spectra, Tancrede et al. (55,56) found a-endorphin to exist in a flexible random coil structure in aqueous solution. Addition of
!II
Solution
Conformations
529 1.0
10 A
.
B
0.8
0.8
0.6
06
~0
.~
.,
~0
~.
----
I
IT O.
OA
~I
0
'" 0.2 o
......
I -, 5
_aAm 10
pH
0
0
0.2
,..-. 0
.8
II
. .... ....m 5
10 pH
Fig. 12.9. Effect of pH on the rotamer populations of Tyr1 and Phe4 side chains of [Met}enkephalin in aqueous solution (reference 48).
lipids had no effect on spin-lattice relaxation times, chemical shifts, or line widths, with the exception of phosphatidylserine at slightly acidic or neutral pH. In the presence of phosphatidylserine, spin-lattice relaxation times were found to be highly dependent on solution pH. Hollosi and co-workers (70) carried out circular dichroism studies of J3-endorphin in water and in trifluoroethanol. In trifluoroethanol, which is a helix-promoting solvent, these workers found a helical structure, probably near the C-terminal region. Comparisons with model compounds suggested that the helical structure was stabilized by interaction with the N-terminal amino group. In contrast, J3-endorphin showed little tendency to form a helical structure in aqueous solution. Levine et al. (53) found the proton NMR spectrum of J3-endorphin in water to differ significantly from that predicted for the peptide in a random coil conformation. The existence of some conformational restraintswas supported by relaxation measurements using Gd(III) ions. In comparison to J3-endorphin, the Met5-sulfoxide derivative showed less opiate activity and less tendency to form a-helix in trifluoroethanol (7/ ,72). A J3-endorphin analog was prepared by Taylor et al. (73) in which residues 1-19 were the same as in J3-endorphin, while the remainder of the residues were selected on the basis of their ability to promote helix formation. The CD spectrum does, in fact, show helical forms, and the synthesized compound 'was shown to have receptor affinities two to three times that of the parent compound. Studying J3-endorphin in trifluoroethanol, Hammonds et al. (74) found that a considerable a-helix structure is present for J3-endorphin, while these authors could find no helical structure for [Met]enkephalin. Bewley and Li (75) examined J3-endorphin fragments of varying size in aqueous
130
12
Physical
Chemistry
and Molecular
Modeling
of the Enkephalins
solution using ultraviolet spectroscopy. They determined that at least nine amino acid residues had to be present in order to demonstrate ordered structure. In /3-endorphin, they found evidence for a tertiary structure around the Tyr1 and Phe4 side chains. The a-amino group, Lys'", and the fragment Thr6-SerlO appeared to be involved in stabilizing a folded form. In a series of circular dichroism studies of /3-endorphin and /3-lipotropin in aqueous solution (76-79), there was little evidence of any secondary structure in water. This conclusion was supported by viscosity measurements and sedimentation coefficients,whichwere consistent with a random coil conformation. Helical structure appeared to be present in methanol solution or when lipids or sodium dodecyl sulfate were added to the solution. The secondary structure induced by addition of lipid was disrupted on addition of calcium ion. Lipotropin was also found to acquire some helical character when the solvent was trifluoroethanol (80). Maroun and Mattice (81) carried out circular dichroism investigations on dynorphin. They found no ordered structure for this compound in aqueous solution. Addition of phospholipid had little or no effect, but addition of sodium dodecyl sulfate induced the formation of a-helical regions. Schiller (63) prepared a synthetic dynorphin fragment having tryptophan in position 4. Unlike the enkephalins (see above), fluorescence spectra showed that the Tyr1_Trp4 average distance was at least 15 A (compared to about 10 A for the enkephalin analogs). It has been proposed that phospholipids may be intimately connected to the opiate receptor, and it has been shown that some lipids can bind opiate analgesics stereospecifically (82,83). For this reason, several studies have focused on the effects of phospholipids on en kephalin conformation. Behnam and Deber (84) examined the tyrosine and phenylalanine /3proton resonances of [Met]enkephalin in the presence of Iysophosphatidyl glycerol micelles. They found that the tyrosine resonances broaden by about 5.5 Hz. The two phenylalanine resonances shifted 30 and 4 Hz, respectively. There was no significant change in the a-/3 proton-proton coupling constants, however, indicating no change in the side chain rotamer populations. In a carbon NMR study of [Met]enkephalin in the presence of phosphatidylserine, Jarrell et al. (85) found evidence for enkephalin-lipid interactions. The spin-lattice relaxation times were altered, indicating a change in the degree of molecular motion in solution. Salt and morphine were shown to interfere with the interaction. The pH optimum for binding was in the slightly acidic range. The binding was shown to involve the N-terminal amino group but not the tyrosine hydroxyl C-terminal carboxyl group. Binding constants for this process were very low.
III Solution Conformations
131 l
r
HO-C6H4
Fig. 12.10.
.1
Proposed
structure
0
o'l-Ir:~l-I FN CH '\,N.: H Ii
0 JL-NH
;.N
" 'i:
~+.' --0 ' l'-~'c
(-~
of the copper
""\
1
:t
C'H,
NH
0
R (II)-cnkephalin
complex
(reference
90).
Bleich and co-workers (86) measured the effee{ of [Met ]enkephalin on the spin-lattice relaxation times of the lipid resonances. They monitored the proton resonances of the choline methyl groups, the alkyl chains, and the w-methyl groups in vesicles of /3,y-dimyristoyl-l-a-Iecithin. They found that enkephalin alters the relaxation times for these resonances, indicating some interaction. However, the interaction was found to be nonspecific, since it was also shown to occur with tetraglycine. There has been considerable interest in en kephalin interaction with metal ions, beginning with the circular dichroism study of Poupaert et al. (52). Addition of sodium or potassium ions caused changes in the carbonyl region of the spectrum, leaving the aromatic region unchanged. The interaction with sodium was weak, but was considered to be potentially significant. Divalent cations produce much more stable complexes, as demonstrated in a number of studies. Hollosi and co-workers (87) found that calcium and manganese form 1:1 complexes with several enkephalin analogs. Using proton NMR, Haran et al. (88) found that zinc ions affect chemical shifts and coupling constants of [Met]enkephalin, implying that a conformational change was induced. It was shown (89) that divalent metal ions have a greater effect on enkephalins than on enkephalinamides. This is consistent with the expected interaction of the cation with the carboxylate group. Addition of divalent copper was shown (90) to affect selectively the following carbon NMR resonances: Met' carboxylate and C-a; Tyr1 carbonyl and C-a; Gly2 C-a; Gly3 C-a. The structure shown in Fig. 12-10 was proposed for the complex. Sharrock et al. (91) demonstrated that, although the copper complex is a relatively strong one, it has very little effect on the biological activity and the in vitro receptor binding. Aided by proton,. carbon, and aluminum NMR, Mazarguil and co-workers (92) studied the binding of aluminum (III) ions to [Leu Jenkephalin. They found evidence for two binding sites, the first involving the tyrosine carbonyl and the C-terminal carboxylate, and the second involving the N-terminal amino group. Aluminum binding altered the temperature dependence of the amide protons in dimethylsulfoxide solution, suggesting that the /3-bend conformation was disrupted.
\32
12
Physical
Chemistry
and Molecular
Modeling
of the EnkephaJins IV
IV.
Molecular
Molecular
Modeling
Studies
IJJ
Modeling Studies
A consequence of the high conformational flexibility in the enkephalin famIly has been to focus computation-based research on conformational behavior and not thermodynamic features, as can be derived from quantltalIve structure-activity relationship (QSAR) approaches. The theoretical conformational energy calculations on enkephalins reported below should be reviewed in light of the recent syntheses and at least partial expenmental conformational characterization of conformationally restncted ~nalogs (37,38,93-95), some of which displayed significant analgesIc activIty In one or more screens (94-105). In particular, the backbone-restricted analog, Tyr-cyclo[N-.-o-Lys_Gly_ Phe-Leu] (FIg. 12-1 I) has been studIed by proton NMR in dimethylsulfox_
,?e. Nonequlvalence of the Gly3 protons suggests some rigidity of the nng str~cture. A l3-bend backbone "conformation at Phe4 and Leus, which ISstablhz,ed by Intramolecular hydrogen bonding involving the N-. proton of o-Lys and the carbonyl group of Gly3, has been suggested from an an~lysis of the coupling constants and the temperature dependence of the ar'."de protons: The proposed conformation is shown in Fig. 12-11. While this conformatIOn has .conslderably more rigidity than open-chain analogs, srgnrficant conformatIOnal freedom still exists regarding the relative onentatlOns of the functional groups on the side chains. The cyclic analog Tyr-cYclo[N-l>-o-O~n-GIY-Phe_Leuj has been shown to retain a single conformatIOn In dImethylsulfoxide solution (107). This conformation is
characterized by two transannular hydrogen bonds, one involving Orn'NH and Orn'-CO and the other involving Leus-NH and Gly3-CO. Molecular mechanics-based conformational energy calculations on yet a third cyclic analog in this class, Tyr-cyclo[N-y-o-Dbu-Gly-Phe-Leu], identified a low-energy conformer resembling the type II' l3-bend model proposed by Clarke and co-workers (108). This conformer lacks a hydrogen bond involving the Leu5-NH and Dbu-CO)groups that should be expected by analogy to the other two cyclic compounds. Overall, the backbone conformations of Tyr-cyclo[N-.-o-Lys-Gly-PheLeu] and Tyr-cyclo[N-l>-o-Orn-Gly-Phe-Leu] can be taken as necessary but not sufficient constraints to define the bioactive conformation. Side chain flexibility and the possibility of conformational changes on binding could negate the structural information realized from these restricted analogs. Taken literally, the difference in backbone conformations of Tyr-cyclo[N.-o-Lys-Gly-Phe-Leu] and Tyr-cyclo[N-l>-o-Orn-Gly-Phe-Leu] suggest that the geometry of the opioid receptors can accommodate at least two different en kephalin backbone conformations. The high conformational flexibility of enkephalin has prevented a complete conformational analysis in terms of a systematic search involving all torsional rotations. Research has focused on conformational strategies to overcome the uncertainty associated with high conformational flexibility. Three types of general strategies, not necessarily independent of one another, have emerged: I. Identifying and exploring the major determinants of conformational specificity. 2. Comparing common conformations available to active analogs but unaccessible to inactive congeners. 3. Matching low-energy conformations of enkephalins to rigid opiates like morphine and its derivatives.
Fig. 12-11.
Conformation
Holzemann (reference /06).
of Tyr-cyclo-N-t:.D-Lys-Gly_Phc_Leu
proposed
by Kessler and
With few exceptions, all conformational energy calculations have been performed using molecular mechanics (109). Many calculations have been based on application of the Empirical Conformational Energy of Peptides and Proteins (ECEPP) program (110). The effect of solvent has not been included in any of the conformational analyses. This is an important point to keep in mind with regard to the relative stability of folded conformations compared to extended structures. In free space, folded conformations can gain added stabilization energy from both dispersion and electrostatic interactions involving distant atoms with respect to primary structure. This is clearly not possible in extended conformations, where favorable solutesolvent interactions have the potential to overcome favorable folded intramolecular effects. The conformational analyses of the enkephalins are
114
12
Physical
Chemistry
and
Molecular
Modeling
of the, Enkcphalins IV
further complicated by the charge state assigned to a molecule. Without knowing the charge state at the receptor, the assignment of charges remains an intrinsic assumption in any such study. Isogai el al. (Ill) carried out a conformational analysis of [Met]enkephalin and concluded that the lowest-energy conformation involves a folded conformation characterized by a type II' bend in the Gly3-Phe. position. This conformation is stabilized by an intramolecular hydrogen bond between the carbonyl group of Gly3 and the hydroxyl group on the tyrosyl side chain. Tyrosine fluorescence studies in water and butanol (60) and pK, determination of the tyrosine hydroxyl in water (44) suggest that this calculated conformation is not highly populated in these solvents even though the structure is consistent with some NMR data (17). The calculated low-energy conformer having a type II' ,l3-bend allows an L-alanine but not a o-alanine to be substituted for Glyz. It has been found that both L-Ala2 and o-Alaz analogs retain high binding potencies in both J.L and Ii receptor assays. On this basis, Isogai el al. came to the conclusion that their lowest -energy conformer is not the bioactive form. Momany (112) followed up the study of Isogai elzal. with extensive conformational analyses of [Metjenkephahn, [L-Ala ,Met ]enkephahn, and [o-Alaz,Met'jenkephalin using the ECEPP program. The enkephahn analogs were assigned different charged end groups to assess the role of charge state on conformation. The lowest-energy conformer f~und for the highly active [o-Alaz,Met']enkephalin differs from the (sogal conformation in that the hydroxyl group on tyrosine is not mvolved m an mtramolecular hydrogen bond. Unfortunately, this conformation is probably not the active conformation. It is incompatible with substitution of a-methylphenylalanine in position 4 and a-aminoisobutyric acid in position 2, both of which result in active analogs (1). The minimum-energy conformations of each residue were used to generate 400 starting conformations in a conformational analysis of the zwitterionic form of [Met]enkephalin (113). The Boltzmann probablhtles were computed for the resulting conformational states. Fifteen of the 400 conformers account for 97% of the total probability of state occupancy. There is considerable variability among the types
Molecular
Modeling
Studie~
535
I 'I I
! ~
The zwitterionic form of the peptide gives rise to predominantly folded conformations that are stabilized by terminal group electrostatic interactions. Maigret and Premilat (116) followed up on their initial Monte Carlo calculations in an attempt to group conformers into common classes using cluster analysis. Five common classes of enkephalin conformers were found for both ionic states. Maigret and Premilat concluded that the solution conformational behavior of en kephalin can be represented by the identified common classes of conformers. The Monte Carlo strategy was also employed by Demonte el al. (117,118) in a conformational analysis of some [Trp.]enkephalin congeners. The singlet-singlet energy transfer efficiencies between Tyrl and Trp., which are distance dependent, were determined in aqueous solution by fluorescence spectroscopy for the zwitterionic form of [Trp.,Met']en kephalin , [o-AlaZ, Trp. ,Met']enkephalin, [AlaZ,Trp. ,Met5jenkephalin, [Me-Trp.,Met5]enkephalin, and [Trp.,Me-Leu'jenkephalin. The corresponding average transfer efficiencies calculated from the Tyr'- Trp. intramolecular distances based on the Monte Carlo-generated ensemble of conformers are in good agreement with the experimental values. There is also reasonable agreement between experimental and computed average NMR proton-proton coupling constants for the backbone protons. The Monte Carlo-based conformational analyses all point toward an equilibrium distribution of conformer states in solution. The tetrapeptide analogs Tyr-Gly-Gly-Phe, Tyr-o-Ala-Gly-Phe, and Tyr-L-Ala-Gly-Phe were the subjects of Monte Carlo conformational analyses (I I). An equilibrium distribution of solution conformers is postulated for Tyr-Gly-Gly-Phe and a highly populated, tightly folded solution conformer for Tyr-o-Ala-Gly-Phe on the basis of agreement between calculated and observed backbone proton-proton coupling constants. The final strategy to be discussed here regarding the search for a bioactive conformation is that of molecular comparison to rigid opiates. On the premise that the Tyrl and Phe. side chains of enkephalin correspond as pharmacophores to the tyramine unit and the phenethyl group, respectively, of PEa, some workers were led to determine lowenergy enkephalin conformers that possess various amounts of structural overlap with PEa (119,120). Both molecular mechanics (119) and quantum mechanics (I20) methods were used to compute the conformatIOnal energies. The enkephalin conformation that maximizes overlap with PEa is high in energy, according to both studies. . An alternate strategy was next adopted (120) in which a constramed conformational search was carried out"to overlap tyramine with Tyrl and to
\36
12 Physical Chemistry and Molecular Modeling of the Enkcphalins
References
537
conformation exhibits the T shape of morphine but is about 8 kcal/mole higher in energy than the calculated global minimum-energy conformer. Such a difference in energy is at the borderline regarding whether or not the receptor interaction will overcome the intramolecular destabilization. PEO permit substitution of D-Ala, but not L-Ala, for Glyo. Thc 52 minimumenergy conformers identified by Isogai et al. (Ill) were used as starting points in this restricted conformational analysis. One of the resulting conformers met the conformational and overlap constraints and is only 3.5 kcal/mole above the global minimum-conformer state. This conformer, superimposed on PEO, is shown in Fig.12-12. This proposed bioactive conformation for [Metjenkephalin is characterized by a type II' j3-bend structure such that considerable overlap is realized in the tyramine section, the side chain of Phe4 coincides with the phenethyl group, and the carbonyl group of Met' shares a near common position with the methoxy group of PEO. Maigret and co-workers (121) also adopted a constrained energy minimization approach in the analysis of some tetrapeptide analogs of Tyr-DAla-Gly-Phe. Four alternative conformations derived from both experimental and theoretical data were uscd in the energy minimizations. The Gly3 carbonyl group was fixed on the Coring hydroxyl of morphine, and the tyrosine residue was restricted to the tyramine locus. Inferential arguments led these workers to select one of the four conformations as the three-dimensional pharmacophore for the peptide at the I'- receptor. This
.
V. QSAR Studies No published QSAR study of the enkephalins has been found by these authors. More generally, the use of classic linear free energy analyses for the derivation of structure-activity relationships in peptide chemistry is remarkably scarce. The focal point in the analysis of peptides has been and continues to be conformational analysis. One reason why little QSAR work has been done on pep tides is the difficulty in assigning the appropriate state of ionization for thc biologically relevant conditions. Partition coefficient, a, and other linear free-energy descriptors are quite sensitive to the ionic state of the peptide. In addition, linear free-energy descriptors, especially the partition coefficient, have been useful in characterizing drug action where transport and/or bioaccumulation are rate-limiting but continuous variables in dictating activity. Direct drug-receptor interactions are of secondary importance in these situations. In the casc of peptides, transport is often an all-or-none issue, that is, transport is a noncontinuous factor in bioactivity, and the specificity of the drug-receptor interaction is the critical event. Finally, linear free-energy descriptors employ the implicit assumption that the individual components are additive and that interactions between terms can bc ignored. For molecules the size of pentapeptides, where intramolecular interactions are often observed, such an assumption may no longer be appropriate in all cases.
References
'ovll Fig. 12-12. Low-energy conformer of {Met]enkephalin superimposed on PEO (reference /20).
1. F. A. Gorin, T. M. Balasubramanian, C. D. Barry, and G. R. Marshall, J. Supramof. St'uct. 9, 27 (t978). 2. P. W. Schiller, in "The Peptides" Vol. 6, p. 219. Academic Press, New York, 1984. 3. G. D. Smith and J. F. Griffin, Science 199, 1214 (1978). 4. M. Mackay and D. C. Hodgkin, J. Chern. Soc. p. 3261 (1955). S. T. L. Blundell, L. Hearn, 1. J. Tickle, R. A. Palmer, B. A. Morgan, G. D. Smith, and J. F. Griffin, Science 205, 220 (1979). 6. I. L. Karle, J. Karle, D. Mastropaolo, A. Camerman, and N. Camerman, in "Peptides: Structure and Function. Proceedings of the 8th American Peptide Symposium" (V. J. Hruby and D. H. Rich, cds.), p. 291. Pierce Chemical Co. Rockford, Illinois, 1983.
138
12 Physical Chemistry and Molecular Modeling of the Enkephalins
7. A. Camerman, D. Mastropaolo, I. Karle, J. Karle, and N. Camerman, Nature (London) 306, 447 (1983). 8. T. Ishida, M. Kcnmotsu, Y. MinD, M. Inoue. T. Fujiwara, K. Tomita, T. Kimura. and S. Sakakibara, in "Peptide Chemistry 1983" (E. Munekata, cd.), Protein Research Foundation, Osaka, 1984. 9. T. Ishida, M. Kenmotsu, Y. MinD, M. Inoue, T. Fujiwara, K. Tomita, T. Kimura, and S. Sakakibara, Biochem. J. 218, 677 (1984). 10. M. Doi, T. Ishida, M. Inoue, T. Fujiwara. K. Tomita, and T. Kimura, FEBS Let!. 170, 229 (1984). II. M..C. Fournie.Zaluski, T. Prange, C. Pascard, and B. P. Rogues, Biochem. Biophys. Res. Commun. 79, 1199 (1977). 12. T. Prange and C. Pascard, Acta Crysta/logr., Sect. R 835, 1812 (1979). 13. S. L. Han, E. R. Stimson, F. R. Maxfield, and H. A. Scheraga, Int. J. Pept. Protein Res. 16, 173 (1980). 14. S. L. Han, E. R. Stimson, F. R. Maxfield, S. J. Leach, and H. A. Scheraga. 1m. J. Pept. Protein Res. 16, 183 (1980). 15. D. M. Rice, R. J. Wittebort, R. G. Griffin, E. Meirovitch, E. R. Stimson. Y. C. Meinwald, J. H. Freed, and H. A. Scheraga, J. Am. Chern. Soc. 103, 7707 (1981). 16. M. A. Khaled, M. M. Long, W. D. Thompson, R. J. Bradley, G. B. Brown. and D. W. Urry, Bioehem. Biophys. Res. Cornmun. 76, 224 (1977). 17. B. P. Rogues, C. Garbay-Jaureguiberry. R. Oberlin. M. Anteunis. and A. K. Lala, Nature (London) 262, 778 (1976). 18. C. Garbay-laureguiberry, B. P. Rogues. R. Oberlin, M. Anteunis, and A. K. Lala, Bioehem. Biophys. Res. Commun. 71, 558 (1976). 19. S. Combrisson, B. P. Rogues, and R. Oberlin, Tetrahedron Left. p. 3455 (1976). 20. C. R. Jones, W. A. Gibbons, and V. Garsky, Nature (London) 262, 779 (1976). 21. H. E. Bleich, J. D. Cutnell, A. R. Day, R. J. Freer, J. A. Glasel, and J. F. McKelvy, Proc. Natl. Acad. Sci. U.S.A. 73, 2589 (1976). 22. C. R. Jones, V. Garsky, and W. A. Gibbons, Bioehem. Biophys. Res. Commun. 76, 619 (1977). 23. C. Garbay-laureguiberry, B. P. Rogues, R. Oberlin, M. Antcunis, S. Combrisson, and J. Y. Lallemand, FEBS LeU. 76, 93 (1977). 24. E. R. Stimson, Y. C. Meinwald, and II. A. Scheraga. Biochemistry 18, 1661 (1979). 25. D. Marion, C.' Garbay-Jaureguiberry, and B. P. Roques. Bioehem. Biophys. Res. Commun. 101, 711 (1981). 26. C. Garbay-Jaureguiberry, D. Marion, E. Fellion, and B. P. Rogues, 1m. J. Pepl. Protein Res. 20, 443 (1982). 27. A. J. Fischman: M. W. Riemen. and D. Om'burn, FEBS Left. 94, 236 (1978). 28. T. Higashijima, J. Kobayashi, U. Nagai, and T. Miyazawa, EUr. J. Biochem. 97,43 (1979). 29. T. Higashijima. T. Miyazawa, M. Fujino. and M. Wakimasu, in "Peptide Chemistry 1980" p. 101. Protein Research Foundation. Osaka, 1981. 30. T. Miyazawa and T. Higashijima, iliopolymers 20, 1949 (1981). 31. C. Di Bello, S. Andini, L. Ferrara, R. Napolitano, and L. Paolillo. 1m. J. Pept. Protein Res. 20, 455 (1982). 32. V. Renugopalakrishnan, R. S. Rapaka, T. W. Collette. L. A. Carreira, and R. S. Bhatnagar, Bioehem. Biophys. Res. Commun. 126, 1029 (1985). 33. M. Anteunis, A. K. Lala. C. Garbay-Jaureguiberry, and B. P. Rogues. Biochemistry 16, 1462 (1977). 34. L. Zetta and F. Cabassi, EUr. J. Bioehem. 122, 215 (1982).
References
.
139
35. G. Gacel, M.-C. Fournie-Zaluski, E. Fellion, B. P. Rogues, B. Senault, J.-M. Lecomte, B. Malfroy, J.-P. Swerts, and l.-C. Schwartz, Life Sci. 24, 725 (1979). 36. N. Niccolai, V. Garsky, and W. A. Gibbons. J. Am. Chem. Soc. 102, 1517 (1980). 37. H. Kessler and G. Holzemann, Angew. Chem. Int. Ed. Engl. 20, 124 (1981); Angew. Chon. 93, 90 (1981). 38. H. Kessler and G. Holzemann. Liebigs Ann. Chem. p.2028 (1981). 39. B. P. Rogues, C. Garbay-Jaureguiberry, S. Bajusz, and B. Maigret, EUr. J. Bioehem. 113, 105 (1980). 40. M.-C. Fournie-Zaluski, G. Gacel, B. Maigret, S. Premil1, and B. P. Rogues. Mol. Pharmacal. 20, 484 (1981). 41. J. E. Gairin, R. Haran, H. Mazarguil, Y. Audigier, and J. Cros, FEBS Lett. 128,51 (1981). 42. F. Reig, J. M. Garcia Anton, G. Valencia, J. J. Garcia Dominguez, A. Englert, J..P. Demonte. and R. Guillard, in "Peptides 1980: Proceedings of the 16th European Peptide Symposium" (K. Brunfeldt, cd.), p. 397. Scriptor, Copenhagen, 1981. 43. J.-M. Garcia Anton, F. Reig, G. Valencia, J. J. Garcia Dominguez, R. Guillard, l.-P. Demonte, and A. Englert, in "Peptides 1982: Proceedings of the 17th European Peptide Symposium" (K. Blaha and P. Mahon, cds.), p. 763. Walter de Gruyter, Berlin, 1983. 44. M. A. Spirtes, R. W. Schwartz, W. L. Mattice, and D. H. Coy, Riochem. Biophys. Res. Commun. 81, 602 (1978). 45. H. E. Bleich. A. R. Day, R. J. Freer, and J. A. Glasel, Bioehem. Biophys. Res. Commun. 74, 592 (1977). 46. J. Kobayashi and U. Nagai, in "Peptide Chemistry 1977" (T. Shiba, ed..), p. 91. Protein Research Foundation, Osaka. 1978. 47. J. Kobayashi, U. Nagai, T. Higashijima, and T. Miyazawa. Biochim. Biophys. Aeta 577, 195 (1979). 48. J. Kobayashi, U. Nagai. T. Higashijima, and T. Miyazawa, in "Peptide Chemistry 1978" (N. Izumiya. cd.), p. 125. Protein Research Foundation, Osaka. 1979. 49. J. Kobayashi, T. Higashijima, U. Nagai, and T. Miyazawa, Bioehim. Riophys. Acta 621, 190 (1980). 50. M. Atreyi, M. V. R. Rao, K. Covindan, M. R. Rajeswari, and K. B. Mathur, Adv. B;osci. 38, 55 (1982). 51. T. S. Sudha and P. Balaram, FERS Lett. 134, 32 (1981). 52. J. H. Poupaert, P. S. Portoghese, and V. Garsky. J. Med. Chem. 19, 1354 (1976). 53. B. A. Levine, D. L. Rabenstein, D. Smyth, and R. J. P. Williams. Bioehim. Biophys. Acta 579, 279 (1979). 54. N. Higuchi, Y. Kyogoku, and H. Yajima, Chem. Lell. p. 171 (1980). 55. P. Tancrede, R. Deslauriers. W. H. McGregor. E. Ralston, D. Sarantakis. R. L. Somorjai, and r. c. P. Smith. Biochemistry 17, 2905 (1978). 56. R. Deslauriers, P. Tancrede, W. H. McGregor, D. Sarantakis. R. L. Somorjai. and I. C. P. Smith, Jerusalem Symp. Quantum Chem. Rioehem. II, 285 (1978). 57. L. Zetta, F. Cabassi. R. Tomatis, and M. Guarneri, Eur. 1. ilioehem. 95,367 (1979). 58. M. Kupryszewska, I. Gryczynski, and A. Kawski, Photochem. Photobiol. 36, 499 (1982). 59. P. W. Schiller, C. F. Yam. and M. Lis. Biochemistry 16, 1831 (1977). 60. P. W. Schiller, Rioehem. Biophy!>'. Res. Commun. 79,493 (1977). 61. P. W. Schiller. C. F. Yam. and J. Prosmanne, 1. Med. Chern. 21, 1110 (1978). 62. P. W. Schiller and J. St.-Hilaire, J. Med. Chem. 23, 290 (1980). 63. P. W. Schiller, Int. J. Pept. Protein Res. 21, 307 (1983). 64. A. Jankowski, I. Z. Siemion, and Z. Szewczuk, Acta Rioehim. Pol. 28, 11 (1981).
540
12 Physical Chemistry and Molecular Modeling of the Enkephalins
65. A. Guyon-Gruaz, l.-P. Demonte, M.-C. Fournie-Zaluski. A. Englert, and B. P. Roques, Biochemistry 20, 6677 (1981). 66. B. Filippi, P. Giusti, L. Cima. G. Borin, F. Ricchelli, and F. Marchiori, Int. J. Pept. Protein Res. 14, 34 (1979). 67. J. Soos, I. Berzetei, S. Bajusz, and A. Z. Ronai, Ufe Sci, 27, 129 (1980). 68. M. A. Khaled, D. W. Urry, and R. J. Bradley, J. Chern. Soc., Perkin Trans. 2 p. 1693 (1979). 69. H. E. Bleich, A. R. Day, R. J. Freer, and J. A. Glasc1, Biochem. Biophys, Res. Commun. 87, 1146 (1979). 70. M. Hollosi, M. Kajtar, and L. GraC, FEBS Left. 74, 185 (1977). 71. L. Graf, M. Hollosi, A. Patthy, I. Berzetei, and A. Ronai, Neuropepr. (Edinburgh) I, 47 (1980). 72. L. Graf, M. Hollosi, I. Barna, l. Hermann, J. Borvendeg, and N. Ling, Biochem. Biophys. Res. Commun. 95, 1623 (1980). 73. J. W. Taylor, D. G. Osterman, R. J. Miller, and E. T. Kaiser, J. Am. Chern. Soc. 103, 6965 (1981). 74. R. G. Hammonds, Jr., A. S. Hammonds, N. Ling, and D. Puett, J. Bioi. Chern. 257, 2990 (1982). 75. T. A. Bewley and C. H. Li, Biochemistry 22, 2671 (1983). 76. J. T. Yang, T. A. Bewley, G. C. Chen, and C. H. Li, Proc. NaIf. Acad. Sci. U.S.A. 74, 3235 (1977). 77. c.-S. C. Wu, N. M. Lee, H. H. Loh, J. T. Yang, and C. H. Li, Proc. Natl. Acad. Sci. U.S.A. 76, 3656 (1979). 78. N. M. Lee, c.-S. C. Wu, J. T. Yang, H. H. Loh. and C. H. Li, Adv. Biochem. Psychopharmacol. 22, 285 (1980). 79. c.-S. C. Wu, N. M. Lee, N. Ling, J. K. Chang, H. H. Loh. and J. T. Yang. Mol. Pharmacal. 19, 302 (1981). 80. P. M. Bayley, C. R. Snell, and D. G. Smyth, Biochem. Soc. Trans. 5,683 (1977). 81. R. Maroun and W. L. Mattice, Biochem. Biophys. Res. Commun. 103,442 (1981). 82. L. G. Abood and F. Takeda. Eur. J. Pharmacol. 39, 71 (1976). 83. L. G. Abood and W. Hoss, Eur. J. Pharmacol. 32, 66 (1975). 84. B. A. Behnam and C. M. Deber, in "Peptides: Structure and Function. Proceedings of the 8th American Peptide Symposium" (V. J. Hruby and D. H. Rich. eds.), p. 445. Pierce Chemical,Co., Rockford, Illinois, 1983. 85. H. C. Jarrell, R. Deslauriers, W. H. McGregor, and I. C. P. Smith, Biochemistry 19, 385 (1980). 86. H. E. Bleich, J. D. Cutnell, A. R. Day, R. J. Freer, and J. A. GlaseL Biochem. Biophys. Res. Commun. 71, 168 (1976). 87. J. Hollosi, Z. Dobolyi, and S. Bajusz, FEBS Leu. 110, 136 (1980). 88. R. Haran, J.-E. Gairin, and H. Mazarguil, C. R. Acad. Sel. Paris 291D, 685 (1980). 89. R. Haran, P. Sharrock, J.-E. Gairin, and H. Mazarguil, Int. J. Pept. Protein Res. 20,380 (1982). 90. J.-E. Gairin, H. Mazarguil, P. Sharrock, and R. Haran, Inorg. Chem. 21, 1846 (1982). 91. P. Sharrock, R. Day, S. lemaire, S. St.-Pierre, H. Mazarguil, J.-E. Gairin, and R. Haran, Inorg. Chim. Acta 66, 191 (1982). 92. H. Mazarguil, R. Haran, and J.-P. laussac, Biochim. Biophys. Acta 717, 465 (1982). 93. D. Hudson, R. Sharpe, and M. Szelke, Int. J. Pept. Protein Res. 15, 122 (1980). 94. T. Deeks, P. A. Crooks, and R. D. Waigh, 1. Med. Chem. 26, 762 (1983). 95. F. A. Gorin, T. M. Balasubramanian, T. J. Acero, J. Schweitzer, and G. R. Marshall, J. Med. Chern. 23, 1113 (1980).
References
541
96. J. DiMaio and P. W. Schiller, Proc. Natl. Sel. U.S.A. 77, 7162 (1980). 97. J. DiMaio, T. M.-D. Nguyen, C. lemieux, and P. W. Schiller, J. !\led. Chem. 25, 1432 (1982). 98. P. W. Schiller, B. Eggimans, J. DiMaio, C. lemieux, and T. M.-D. Nguyen, Biochem. Biophys. Res. Commun. 101, 337 (1981). 99. H. I. Mosberg, R. Hurst; V. J. Hruby, J. J. Galligan, T. J. Burks, K. Gee, and H. I. Yamamura, Biochem. Biophys. Res. Commun. 106, 506 (1982). 100. H. I. Mosberg, R. Hurst, V. J. Hruby, J. J. Galligan, T. J. Burks, K. Gee, and H. I. Yamamura, Life Sci. 32, 2565 (1983). 101. I. Z. Siemion, Z. Szewczuk, Z. S. Herman, and Z. Stachura, Mol. Cell. Biochem. 34, 23 (1981). 102. M. L. English and C. H. Stammer, Biochem. Biophys. Res. Commun. 85,780 (1978). 103. R. E. Chipkin, J. M. Stewart, and C. H. Stammer, Biochem. Biophy.~. Res. Commun. 87, 890 (1979). 104. Y. Shimohigashi and C. H. Stammer, Int. J. Pept. Protein Res. 19, 54 (1982). 105. Y. Shimohigashi and C. H. Stammer, Int. J. Pept. Protein Res. 20, 199 (1982). 106. H. Kessler, G. Holzemann, and R. Geiger, in "Peptides: Structure and Function. Proceedings of the 8th American Peptide Symposium" (V. J. Hruby and D. H. Rich, eds.), p. 295. Pierce Chemical Co.. Rockford, Illinois, 1983. 107. H. Kessler, G. Holzemann, and C. Zechel, In!. J. Pept. Protein Res. (1985). (In press). 108. F. II. Clarke, II. Jaggi, and R. A. Lovell, 1. Med. Chern. 21, 600 (1978). 109. U. Burkert and N. L. Allinger, ACS Monogr. No. 177, (1982). 110. M. J. Browman, L. M. Carruthers, K. L. Kashuba, F. A. Momany, M. S. Pottle, S. P. Rosen, and S. M. Rumsey, Quantum Chem. Program Exchange II, 286 (1975). 111. Y. Isogai, G. Nemethy, and H. A. Scheraga, Proc. Nat/. Acad. Sci. U.S.A. 74,414 (1977). 112. F. A. Momany, Biochem Biophys. Re!1'. Commun. 75, 1098 (1977). 113. J. L. De Coen, C. Humblet, and M. H. J. Koch, FEBS LeU. 73,38 (1977). 114. S. Premilat and B. Maigret, Biochem. Biophys. Res. Commun. 91,534 (1979). 115. S. Premilat and B. Maigret, J. Phys. Chem. 84, 293 (1980). 116. B. Maigret and S. Premilat, Biochem. Biophys. Res. Commun. 104, 971 (1982). 117. J. P. Demonte, R. Guillard, and A. Englert, Arch. Int. Physiol. Biochem. 87, 1038 (1980). 118. J. P. Demonte, R. Guillard, and A. Englert, Int. J. Pept. Protein Res. 79, 1199 (1977). 119. A. F. Bradbury, D. G. Smyth, and C. R. Snell, Nature (London) 260, 165 (1976). 120. G. H. Loew and S. K. Burt, Proc. Natl. Acad. Sel. U.S.A. 75, 7 (1978). 121. B. Maigret, S. Premilat, M.-C. Fournie-Zaluski, and B. P. Roques, Biochem. Biophys. Res. Commun. 99, 267 (1981).
Index
A 6-AcetyJ-t.iodocodeine, x.ray crystallography, t67 Acetylisomethadol,411-414 stereoisomers,412-413 Acetylmethadol,411-414 dissociation constant, 451 metabolism, 414 receptor binding, 414 stereoisomers, 412 x-ray crystallography. 453 6-Acetylnormorphine, metabolite, 14, 18 ACTH, contained in proopiomelanocortin, 463 4-Acyloxy-4-arylpiperidines, 334-352; see also Prodine, 3.Desmethylprodine, specific compound methylation 2,3-dimethyl. isomers, SAR, 346-350 2,6-dimethyl, isomers, SAR. 346-350 3,5-dimethyl, see lsopromedol 2,S-dimethyl, see Promedol C-2 methyl, isomers. SAR. 348-349 C-3 methyl. see Prodine SAR, 337-352 C-4 acytoxy, 337-338 C-3 alkylation, 343-346 C-4 aryl, 338 dimethylation, see methylation C-2 methylation, 346-351; .fee aim methylation C-3 methylation; see also Prodine; see specific compound N.I analogs, 338-341; see al.w Norprodine synthesis, 335-337 D-Ala!-
1)- I .eu~-enkcphalin
543
effect, respiration, 37 relative affinity, for /) receptor, 34 [Ala2,Met~]enkephalin, molecular modeling, 534 (o-Ala1,Mct~lenkephalin. molecular modeling, 534 (o-Ala1,Met~)enkephalin amide. .\'('(' DAME (Ala1,Trp4,Met~lenkephalin. Monte Carlo simulation. 535 ID-Ala!,Trp4,Met~]enkephalin, Monte Carlo simulation, 535 Alfentanil, 365 3-Alkyl-3-arylpiperidines, 354-356; see also 4-Alkyl-4-arylpiperidines, Profadol antagonists, SAR, 358 rigid analogs, 359-361 ring contraction, SAR, 356-369 ring expansion, 356 4-Alkyl-4-arylpiperidines, 352-354; see also 3-Alkyl.3-arylpiperidines antagonists, SAR, 353 rigid analogs, 354, 359-361 AJletorphine, 137-138 N.Allylnormetazocine, 31, 35, 278-279 effect, in chronic spinal dog, 31 opiate receptor selectivity, 279 relative affinity, for u receptor, 35 N.Allylnormorphine, see Nalorphine N.AlIylnoroxymorphone, .fee Naloxone Alphaprodine, 341 absolute configuration, 342 analgesic activity, 341-342 analogs, 345; see also Norprodine antipodes, SAR, 342, 346 clinical use, 342 a.AlIylprodine, 343-344
544 conformation, 385 nuclear magnetic resonance, 380-382 stereochemistry, 385 x-ray crystallography, 378 j1-Allylprodine conformation, 385 nuclear magnetic resonance, 380-382 stereochemistry. 385 x-ray crystallography, 378 Aluminum(lIl), interaction with [Leu]enkephalin.531 Alvodine, see Piminodine Amidon(e), 400 6-Amino-6-deoxymorphine. 66 (4' -Amino-Phe"]enkephaJin, fluorescence spectroscopy, 527 Amphetamine. 177 Analgesia choice of test, 36-37 5 receptor, and involvement, 36 K receptor, and involvement, 36 methods of measurement, 35-36 J.Lreceptor, and involvement, 36 opioid receptor sensitivity, 37 opioid site of action, 36 AniJeridine, 324-325 4-Anilinopiperidines, 362-367; see specific compound, Fentanyl; see alw 4Anilidopiperidines, Fentanyl, specific compound rigid analogs, 366-367 SAR, 363-367 synthesis, 362-363 Arylpiperidines conformation, 378-394 molecular modeling, 388-394 nuclear magnetic resonance, 380-384 physicochemical studies, 377-384 QSAR, 394-398 x-ray crystallography, 378-380 Azabicycloalkanes, 174 Aza-des-N-morphinan SAR, 236-238 synthesis, 236-238 6-Azido-6-deoxycodeine, 65 6-Azido-6-deoxymorphine, 65 Azidomorphine, x-ray crystallography, 167, 169
Index
Index
B Bemidone; see also Ketobemidone analgesic activity, 332 analogs, SAR, 332 analogs, synthesis, 331 Benzazepines, 306 Benzazocines, 305-306 Benzethidine, 325-326 Benzimidazoles, 436-438 Benzomorphans, 250-311 hetrocyclic A-ring substitutions, 296 history, 250 nitrogen positional isomers, 306-310 nomenclature, 251 numbering system, 251 opiate receptor affinity, 271, 281-282 rigid analogs, 305-306 SAR 6-alkyl substitution, 259-261 A-ring substitution, 259-261, 295-296 BC-ring enlargements and contractions, 297-304 6,ll-dialkyl enantiomers, 265 6,ll-dialkyl substitution, 263-266, 268 dialkyI substitution, 268-269 6-hydroxy substitution, 294 II-hydroxy substitution, 292-294 II-methyl substitution, 261-262 N-ether substitution, 288 N-heterocyclicalkyl substitution, 285288 N-substitution, 273-289 alkyl groups, 273-276 cycJoalkyl, 279-285 cyanoethyl group, 276 unsaturated, 277-285 I-oxygen substitution, 289-292 parent nucleus, 259-261 6-phenyl-ll-alkyl substitution, 266-268 6-phenyl substitution, 260, 263, 266268 summary, 310-311 trialkyl substitution, 269, 284 6,1 1,1 I-trialkyl substitution, 270-271 stereochemistry, 254 synthesis, 252-259 lip-alkyl derivatives, 255-256 Bischler-Napieralski reaction, 197,257
I I
545
Diels-Alder reaction, 257 Grewe cyclization, 252-254 homolytic cyclization, 257 hydratroponitrile route, 252 II-hydroxy substitution, 292 II-methyl derivatives, 261 I-oxygenated derivatives, 289-290 from 4-phenyl-pyridine-2-carboxylic acid, 259 stereochemistry of substituents, 254 tetralone route, 252, 255, 261 6,1 1,1 I-trialkyl derivatives, 270 Benzylisoquinolines electrochemical oxidation, 204-206 Pschorr reaction, 48 reaction with one electron oxidants, 203 TI(III) oxidation, 48 Bischler-Napieralski reaction, 197,257 synthesis of benzazepines, 306 [4'-Br-Phe~J-[Leu]enkephalin, x-ray crystallography, 515 [4'-Br-Phe4]-[MetJenkephalin, X-rdy crystallography, 515 Bremazocine, 283 relative affinity, for K receptor, 35 Buprenorphine, 139-140 Butorphanol, 227 molecular modeling, 175 N-Butyroyl-N'-cinnamylpiperazine, 444
c Calcium, interaction with enkephalins, 531 4-Carbalkoxy-4-arylpiperidines, 319-331; see specific compound; see al.w Meperidine, Normeperidine, Bemidone structure-activity relationships, 321-327 synthesis, 319-321 p-Casein, 460 Casomorphine, SAR, 462 p-Casomorphin, 460 derivatives, 476-477 structures, 461 p-Chlornaltrexamine, 69-70 Chlorocodide, 63 ,8-Chlorooxymorphamine,69 Chlorpromazine, 177 Ciramadol, 443 Circular dichroism, 166
1
dynorphin, 528-530 endorphins, 528-530 enkephalins, 516-522, 525, 528, 531 Gly-Gly-Phe, 528 lipotropin, 528-530 Tyf-Gly-Gly, 528 Clastic opiate receptor binding, 233, 304305 CLIP, contained in proopiomelanocortin, 463 Cluster analysis, of enkephalin conformers, 535 CND, see Codeine N-oxide Codeine biology, 59 biosynthesis, 10-12 biotransformation in animals, 13-19 N-oxides, 14, 16, 20 O-demethylation, 14, 17 in Papaver, 20 de methylation, 52 7,8-epoxide,60 7-methoxy,84 morphine precursor, 10-12 N-oxide, metabolite, 14, 16, 20 opium alkaloid, 2 SAR A-ring substitution, 79, 80 C-7 substitution, 84-93 C-8 substitution, 93 synthesis, 45-48, 51-52 x-ray crystallography, 167-168 (+ )-Codeine, 56 synthesis, 51 trans-Codiene, 57-58 Codeinone cuprate addition, 93 morphine precursor, 10-12 Codorphone, 94 Conformationally restricted analogs, of enkephalins, 532 Copper, interaction with enkephalins, 531 Cyclazocine, 279 enantiomers, 280 methyl ether, 281 II-stereoisomers, 281 Cyclohexenylethyl amine, 197 Cyclohexylamines, 438-441
546
Index
cyclo-[Leu]enkephalin, nuclear magnetic resonance, 521 cyclo-[MetJenkephalin. nuclear magnetic resonance, 521 7,S-Cyc1opropylcodeinone. 62 Cyclorphan, 217-218, 221 analog, 208 Cymidon, see Ketobemidone
D DADLE; .H'(' D-Ala1-D-Leu~-cnkephalin DAGO; H'(' Tyr-o-Ala-Gly-MePhe-Gly-ol DALECK, 498 DAME, opiate receptor affinities. 474 DAMME, see FK33-824 Demccol, see Meperidine 3-Deoxymorphine, 58 Dependence. drug, definition, 39 K receptor, and involvement, 40 J.Lreceptor, and involvement, 40 physical animal model, 39 definition, 39 psychic animal model, 39-40 correlation of animal model and clinical experience, 40 definition, 39 Demorphin. 460 SAR, 462 structure, 461 3-Desmethylnorprodine. analogs SAR, 338-341 analgesic activity, 338, 340 3-Desmethylprodine, analogs SAR, 337-339 N-I substitution, see 3-Desmethylnorprodine synthesis, 335-337 Dextromethorphan, 207 x-ray crystallography, 167, 169 Dextromoramide, 419, 453 Dextrorphan, 207 3,6-Diacetylmorphine, metabolism, 14, 18 Diamorphine, see Heroin Diampromid, 428, 430-432
Index 547
Diazepam, 177 lrans- 3 ,4-dichloro-N-methyl-N-(2-( I-pyrrolidinyl)-cyclohex ylJ-benzeneacetamide and analgesia, 36 relative affinity, for K receptor, 35 3,6-Dideoxymorphine,58 Diels-Alder reaction, synthesis of benzomorphans, 257 Dihydrocodeine, 7,7-dimethyl, 89 Dihydrocodeinone, 51, 60 chloromethylation, 80 7,7-dimethyl,89 enol ether, 81 reaction with DMF acetal, 87 reductive cleavage to morphinans, 191 Dihydromorphine, 60 QSAR, 184-185 relative affinity, for 1J.receptor, 34 Dihydromorphinone, 63, 65 metabolite, 14, 16 metabolite, hydroxy, 16-17 Dihydrothebaine, Grignard reaction, 221222 Dihydrothebainequinone, 114 base-catalyzed rearrangement, 112-114 Dihydrothebainone, 51 4-hydroxyl removal, 211 SAR,214-215 synthesis, 48, 192 synthesis from dihydrocodeinone, 191 13-Dimethyl-4-phenyl-4-propionyloxypiperi_ dine, x-ray crystallography, 378 Dimethylpiperidines conformation, 385-388 nuclear magnetic resonance, 380-384 stereochemistry, 385-388 ,B,y-Dimyristoyl-Iecithin, see Phospholipid Dionin, 59 Diphenylethylpiperazines, 441-442 Diphenylhydantoin, 177 Diprenorphine, 139-140 Dissociation constant, 166, 173 of enkephalins, 525, 534 Dolantin, see Meperidine DPE~, .\'i'l' ITyr-o-Afa-Gly-PhcLeuNIIJ,-iCII,h DSLET, .\'i'l' Tyr-o-Scr-Gly-Phe-Lcu-Thr DTEt:, ,\'l'l' ITyr-o-Ala-GlyPhcNHh'(cH~)t:
DTLET, ,\'i'l' Tyr-o-Thr-Gly-Phc-Leu_Thr Dynorphin biology, 470 circular dichroism, 528-530 fluorescence spectroscopy, 530 nuclear magnetic resonance, 528-529 opiate receptor affinities, 470 structure, 461 Dynorphinl_9, relative affinity, for K receptor, 35 E Electrochemical oxidation of I-benzylisoquinolines, 204-206 Empirical conformational energy of peptides and proteins (ECEPP), 533-534 6,14-Endoethenotetrahydrothebaine, 110; see also Thebaine, Diels-Alder adducts C-19 alcohols, see Thevinols, Orvinols Endorphins, 460 circular dichroism, 528-530 contained in proopiomelanocorlin, 463465 nuclear magnetic resonance, 528-529 structures, 461 ultraviolet spectroscopy, 529-530 f3-Endorphin biology, 465 effect, respiration, 37 opiate receptor affinities, 465, 468 SAR, 465 Enkephalins, 459-512, 513-537 biology, 466 clinicaJIy investigated, 500-502 comparison with rigid opiates, 475 conformation, 513-537 cyclic derivatives, 477 deuterium labeled, 516, 519, 523, 525, 528 dimeric, 481 dipeptide derivatives, 481 dissociation constants, 525 fluorescence spectroscopy, 527, 534-535 history, 459 infrared spectroscopy, 516 laser Raman spectroscopy, 515-516, 518 metal ion interactions, 531
1
molecular modeling, 532-537 nuclear magnetic resonance, 516-531 occurrence in adrenal glands, 466 opiate receptor affinities, 466 opiate receptof affinities summary, 503 opiate recept6r selectivities, 473-481 5-opiate receptor selectivity SAR, 479 p.-opiate receptor selectivity SAR, 475 opioid antagonists, 499 opioid antagonist isosteres, 499 penicillamine-containing, 480 QSAR, 537 SAR,481-499 C-terminus, 499 Leu~ replacement by dehydro-Leu, 498 Met~/Leu~, 4%-499 Met~/Leu~ backbone alkylation, 499 Met~/Leu~ replacements, 497 Met5/Leus rigid analogs, 499 minimum chain length for analgesia, 481 N-terminus, 482 N-terminus amino acid additions, 484 Phe~ backbone alkylation, 494 Phe~ replacement by dehydro-Phe, 492 Phe~ reduction to hexahydro, 494 Phe~ rigid analogs, 495 Phe4 substituents, 492 Phe4 substituted, 493 Summary, 502 tyrosine N-substitutions, 483 tyrosine replacement, 483-485 x-ray crystallography, 514-515 Enkephalinamides, metal ion interactions, 531 Eptazocine, 303 EseroJine, 182 Ethoheptazine, 329-330 Ethylketocyclazocine, 290 and analgesia, 36 opiate receptor affinity, 291 relative affinity, for K receptor II-stereoisomer, 290 Ethyl l-methyl-4-phenylpiperidine_4_car_ boxylate, see Meperidine Ethylthevinoate, 105 Etonitazine, 174
548
Index
Index 549
Etorphine analgesic activity. 132 benzoylthio, 133 clinical use, 133 QSAR, 184-185 receptor binding, 133 relative affinity, for jJ. receptor, x-ray crystallography, 153 S-Etorphine, 133 Etoxeridine. 325-326
34
F FAO, 117-118 Fentanyl absolute configuration, 367 analgesic activity, 363-364 analogs, 365-366 clinical use, 365 isothiocyanates, 365, 367 molecular modeling, 174 QSAR, 184-185,397-398 Flavothebaone. 114 Fluorescence spectroscopy (4'.amino-Phe(]enkephalin, 527 dynorphin, 530 [Leu]enkephalin, 527 [Mel]enkephalin, 527, 534 [Ser']enkephalin, 528 [Trp']enkephalins, 527-528, 535 Tyr-Met-Gly-Trp-Pro.522 Tyr-Met-Gly-Trp-Pro-NHh 527 Tyr-Pro-Gly-Phe-Leu, 522 Tyr-Pro-Gly-Phe-Met, 522 FK-33-824, 34, 477 clinical investigations. 500 Fremy's salt oxidations, 213 ,B-Funaltrexamine, 71 antagonism, respiratory depression, by opioids, 38 2-(2-Furyl)ethyl levorphanol, QSAR, 184185 G Gastrointestinal motility I( receptor, and involvement, J.Lreceptor, and involvement, opioid site of action, 38
38 38
Gemazocine, 270 Gly-Gly-Phe, circular dichroism, 528 Gly-Gly-Phe-Leu nuclear magnetic resonance, 518 x-ray crystallography, 515 Gly-Gly-Phe-Met, nuclear magnetic resonance, 518 Glyol, see DAGO Gly-Tyr-Gly-Gly-Phe-Met, nuclear magnetic resonance, 520 GPA-1657,266-267 Grewe reaction by-products, 197-199 catalysts, 196 effects of N-substituents, 196 regiochemistry, 1% stereochemistry of, 197 synthesis of benzomorphans, 252-254 synthesis of 14-hydroxymorphinans, 201 synthesis of morphinans, 193-199
H (+)-Heroin,56 Heroin, see also 3,6-Diacetylmorphine 7,8-epoxide,61 history, 55 SAR, 59 Homobenzomorphans, 299-303 Hydrocodeinone hydrazone, oxidative cleavage to morphinans, 191 Hydrocodone, 14 hydroxy metabolites, 17 Hydromorphone, !iee also Dihydromorphinone QSAR, 184-185 Hydroxyazidomorphine, x-ray crystallography, 167 14-Hydroxycodeinone, see oxycodone 3-Hydroxylevallorphan, x-ray crystallography, 167, 169 14-Hydroxyisomorphinan SAR, 228 synthesis, 229 14-Hydroxymorphinans, synthesis, 199201 (3-Hydroxyphenyl)methylmorphans, 360361
Imipramine, 177 Infrared spectroscopy, 166 [Leu]enkephalin, 516, 519-520, 525526 [Met]enkephalio, 519-520, 525-526 Isofentanyl, 364 IsomethadoJ, 411-414 infrared spectroscopy, 452 nuclear magnetic resonance, 452 synthesis, 411 Isomethadone acyloxy analogs, 424 amide analogs, 416, 418 conformation, 448-451 dissociation constant, 448-449 ether analogs, 424 imine derivatives, 422 molecular modeling, 457 nuclear magnetic resonance, 449-450 stereoisomers, 401, 404-405, 408-409 synthesis, 401, 403 x-ray crystallography, 451 Isomorphinan SAR, 208-209 C-6 substitution, 212 C-7 substitution, 221-225 C-14 hydroxyl substitution, 228-229 synthesis, 197 Isoneopine, 56 Isonepenthone, 112-113 Isopethidine, 322 Isoprodine, 342 Isopromedol, stereoisomers SAR, 348, 350 synthesis, 347
K Ketobemidone analgesic activity, 333 analogs, SAR, 332-334 clinical use, 333 molecular modeling, 388 QSAR, 184-185 synthesis, 331 Ketocyclazocine, 290 effect, in chronic spinal dog, 31
opiate receptor affinity, 291 relative affinity, for I( receptor, 35 6-Keto-morphinans, 213-215 reduction to alcohols, 214 SAR C-7 substitution, 221-225 14-hydroxy substitution, 229-230 N-substitution, 218 4-Ketoxy-4-arylpiperidines, 331-334; see also Ketobemidone SAR, 332-334 synthesis, 33 t ~ Kyotorphin, 460 structure, 461 L Laser Raman spectroscopy [Leu]enkepha1in, 515, 518, 525-526 [Mellenkephalin, 519-520, 525-526 Laudanine, morphine precursor, 7-8 Laudanosine, oxidation, 206 Leptanal, see Fentanyl Leritine, see Anileridine [Leu]enkephalin aluminum interaction, 531 carbon-13 labeled, 518 conformation, 513-537 deuterium labeled, 516, 519, 525 effect, in guinea pig ileum, 31 effect, in mouse vas deferens, 31 fluorescence spectroscopy, 527-528 infrared spectroscopy, 516, 519-520, 525-526 ionization state, 518, 525 laser Raman spectroscopy, 515, 518520, 525-526 Monte Carlo simulation, 534-535 nitrogen-15 labeled, 518-519, 523 nuclear magnetic resonance, 516, 518519,522-523,525,528,531 occurrence in prodynorphin, 470 occurrence in proenkephalin, 466 opiate receptor affinities, 468 opiate receptor selectivities, 474 structure, 461 x-ray crystallography, 514-515 Levallorphan, 216-217, 221 Levomethorphan, 207, 211
550
Index
Levorphanol QSAR, 184-185 relative affinity, for 1J.receptor, 34 Lipotropin circular dichroism, 528-530 nuclear magnetic resonance, 528-529 p- LipotrOf)in, 463 Lonan, see Levallorphan LY-t27,623, see Metkephamid Lysergic acid diethylamide (LSD), 177 Lysophosphatidyl glycerol. see Phospholipid Lys- Tyr-Gly-Gly-Phe-Met, nuclear magnetic resonance, 520
M 6-MAM, see 6-0-monoacetylmorphine Manganese, interaction with enkephalins, 531 Meperidine analgesic activity, 325 analogs, see also Bemidone C-4 acyloxy, see 4-Acyloxy-4-arylpjperidines, Prodine C-4 alkyl. see 4-Alkyl-4-arylpiperidines C-4 anilino. see 4-Anilinopiperidines. Fentanyl azacycloheptane, 328-330 C-4 hydroxy, see 4-Phenyl-4-piperidinols C-4 ketoxy, see 4-Ketoxy-4-arylpiperidines, Ketobemidone N-I substitution, see Normeperidine C-4 aryl, SAR, 322-325 C-4 carbalkoxy, SAR, 322-325 clinical use, 323-324, 328 conformation, 385, 388-394 molecular modeling, 388-394 synthesis, 319-321 x-ray crystallography, 378 p-Meprodine, 343-344 Meptazinol, 356 MES, see Morphine-3-sulfate Metal ions, interactions with enkephalins, 531 Metamorphinan, 235 Metazocine, 265
enantiomers, 265 racemate, 264 [Met]enkephalin carbon-13 labeled, 526 circular dichroism, 528 conformation, 513-537 deuterium labeled, 523, 528 dissociation constants, 525 effect in guinea pig ileum, 31 in mouse vas deferens, 31 respiration, 37 fluorescence spectroscopy, 527 infrared spectroscopy, 519-520, 525-526 ionization state, 525 laser Raman spectroscopy, 519-520, 525-526 nuclear magnetic resonance, 516-517, 519-520,523-528,530-531 OCCUITencein proenkephalin, 466 OCCUITencein proopiomelanocortin, 465 opiate receptor affinities, 468 opiate receptor selectivities, 474 structure, 461 x-ray crystallography, 515 [Met]enkephalinamide, nuclear magnetic resonance, 519 [Met]enkephalin-Arg6-Gly'-Leu8, 460-461 [MetJenkephalin-Arg6_Phe', 460-461 Melanocyte stimulating hormone, contained in proopiomelanocortin, 463 Methadol,411-414 conformation, 451-453 dissociation constant, 452 infrared spectroscopy, 451-452 nuclear magnetic resonance, 452 receptor binding, 414 stereoisomers, 411-412 synthesis, 411 x-ray crystallography, 453 Methadone activity, 400 conformation, 448-451 discovery, 400 dissociation constant, 448-449, 451 metabolism, 404 metabolite, 405 molecular modeling, 456-457 nuclear magnetic resonance, 449-450 QSAR,184-185
Index receptor binding, 404, 414 stereoisomers, 401, 403-405, 408-409 structure-activity relationships, 406-435 alkylamine modifications, 406-411 amide analogs, 416-419 diphenyl modifications, 425-435 ester analogs, 415-416 imine derivatives, 421-423 ketone modifications, 411-425 phosphorus analogs, 419-421 sulfur analogs, 419, 421-422 toxicity, 404 x-ray crystallography, 451 Methadone-N-oxide, 405 5-Methyldihydrocodeinone,82 5-Methyldihydromorphinone, ~'ee Metopon 3-Metbylfentanyl, QSAR, 397-398 3-Methylfentanylisothiocyanate, 365, 367 5-Methylmethadone conformation, 450 dissociation constant, 450-451 molecular modeling, 457 nuclear magnetic resonance, 450 stereoisomers, 409-411 synthesis, 410 x-ray crystallography, 451 N-Methylmorphine, 74 N-Methylnalorphine, x-ray crystallography, 167 Methylpiperidines conformation, 385 nuclear magnetic resonance, 381 stereochemistry, 385 x-ray crystallography, 378 Metkephamid clinical investigations, 500 SAR, 493 Metopon, 81-83 6-aJcohol, 83 [Me-Trp4,Met5Jenkephalin, Monte Carlo simulation, 535 MNO, see Morphine N-oxide Molecular mechanics, see Molecular modeling Molecular modeling [Ala2,Met5]enkephalin, 534 lo-Ala~,Mct~lcnkephalin, 534 azabicycloaJkanes, 174 benzomorphans, 176-179 butorphanol, 175
551 dimethylpiperidines, 390 enkephalins, 532-537 fentanyl, 176 isomethadone, 457 y-isopromedol, 390 ketobemidone, 388 meperidine, 388-394 methadone, 456-457 5-methylmethadone, 457 morphinans, 175-176 morphine, 174-183 nalorphine, 177-179 naloxone, 175 naltrexone. 175 oripavine derivatives, 182-183 4-phenylpiperidines, 174,.-t76, 388-394 prodines. 388-394 a-promedol, 390 Tyr-cyclo[N-y-Dbu-Gly-Phe_Leu], 533 6-0-Monoacetylmorphine, 14, 18 Monte Carlo simulation [Ala2,Trp4,Met5]enkephalin, 535 [D-Ala2,Trp4,Met~]enkephalin, 535 (Leu]enkephalin, 534-535 [Me-Trp4,Met~]enkephalin. 535 [Trp4]enkephalins, 535 [Trp"Me-Leu~]enkephalin, 535 [Trp4,Met~Jenkephalin, 535 Tyr-Ala-Gly-Phe, 535 Tyr-Gly-Gly-Phe, 535 Morpheridine, 325-326 Morphiceptin, 476 relative affinity, for 1J. receptor, 34 Morphinans molecular modeling, 175-176 naturally occurring, 189 opiate receptor affinities, 220-221 SAR, 206-241 annulated derivatives, 231 A-ring reduction, 210 A-ring substitution, 220, 240 9-Aza derivatives, 239 C-7,8 disubstitution, 227 C-IO-homo derivatives, 234 C-14 hydroxyl substitution, 227-230 C-nor derivatives, 232 C-IO oxygen substitution, 227 C-3 substitution, 209-210 C-6 substitution, 210-215, 225, 229-230
218, 221-
552 C.7 substitution, 221-225 C-8 substitution, 224-226 Diels-Alder adducts, 230-231 D-nor derivatives, 232 enantiomers, 207-208 isomorphinans, 208-209 5-methyl substitution, 220-221 N-heterocyclic alkyl substitution, 218219 nitrogen substitution, 215-220, 223230 oxa derivatives, 240-241 synthesis annulated derivatives, 231 9-Aza derivatives, 239 from, I-benzyIisoquinolines, 201-204 C/D nor and homo derivatives. 232235 by electrochemical oxidation, 204-206 Grewe cyclization, 193-199 14-hydroxy derivatives, 199-201 isomorphinan by-products, 197 from, morphine anaJogs, 190-193 by, one electron oxidants, 203 oxa-derivatives, 240-241 by, PschoIT reaction, 201 [otal, 193-206 Morphinandienones. synthesis, 201-207 Morphine 3-acetyl, 59 6-acetyl, 59 biosynthesis, 7-13 from tyrosine, 8-9 biotransformation, 13-20 in animals, 13-19 in Papaver, 19-20 conjugation, 13-15 glucuronide formation, 14-15 sulfate formation, 14-15 N-demethylation, 14-16 N-oxidation, 14, 16 O-methylation, 14, 16-17 route of administration, 22-24 bond polarities, 180-183 charge densities, 180-183 chromium complex, 94 disposition, 20-24 three-compartment model, 21 effect in chronic spinal dog, 31
Index
Index 553
in guinea pig ileum, 31 in mouse vas deferens, 31 respimtion. C02 stimulus-response curve, man, 37 electrostatic potentials, 180-183 elimination from plasma, 20-22 in urine, 21-22 history, 2 isolated from bovine brain, 30 from frog skin, 30 to-methyl. 95 nuclear magnetic resonance, 170, 172 QSAR, 184-185 relative affinity, for J.Lreceptor, 34 SAR,55-101 A-ring substitution, 79 C-3 substitution. 58-60 C-5 substitution, 81-84 C-6 substitution, 62-72 C-7 substitution, 84-93 C.8 substitution, 93 C-IO substitution, 94 C-14 substitution. 95-101 3,6-diesters, 59 3-ethers, 59 history, 55 N-substitution, 72-79 summary, 155 structure, 45 synthesis, 45-55 aryJpiperidine route. 52-53 biomimetic, 48-49 from codeine, 52 Gates, 45-48 Grewe, 48 Rice, 51-52 x-my crystallography, 167-168, 173 (+)-Morphine,56 synthesis, 51 trans-Morphine, 56-68 SAR, 58 synthesis, 57 Morphine 7,8-epoxide, 60-61 Morphine-3-glucuronide, 14-15 . Morphine N-oxide, metabolite, 14, 16, 20, 74 Morphine.3-sulfate, 14-15 Morphinone, 63
Moxazocine, 294 MR-2034, 286-288 opiate receptor affinities, MTD, see Steric mapping
288
N Nalbuphine, 78 effect, respiration, C02 stimulus-re_ sponse curve, man. 37 x-ray crystallography, 167, 169 Nalmexone, 77 Nalorphine. 76 molecular modeling, 177-179 x-ray crystalJography, 167-168 Naloxonazine, antagonism, respiratory depression, by opioids, 38 Naloxone, 18-19.76-77 hydroxy metabolites, 18-19 molecular modeling, 175 x-ray crystallography, 167-168 (+)-Naloxone,77 Naltrexol. 78 nuclear magnetic resonance, 170-171 Naltrexone, 18-19,78 hydroxy metabolites, 18-19 molecular modeling, 175 Neoendorphins, 460 biology, 470 opiate receptor affinities, 470 structures, 461 Neopine, nuclear magnetic resonance, 170 Neopinone, morphine precursor, 10-12 Nepenthone, lOS, 112-113 Nisentil, see Alphaprodine Norbenzomorphans, 298 Norcodeine, N-allyl, 75 Norheroin. 73-74 metabolite, 14, 18 Nor-homobenzomorphans, 303 Norlaudanosoline I-carboxylic acid, 8-9 dimethyl ether, 9-10 morphine precursor, 8-10 Normeperidine analgesic activity, 324-325 analogs clinical use, 325. 328
SAR, 324-327 synthesis, 321 synthesis. 321 N ormethadol infrared spectroscopy, 451 opticaJ rotatory dispersion, 453 stereoisomers, 453 Normethadone, 408-409 acyloxy analogs, 424 amide analogs, 416-417 nuclear magnetic resonance, 449 x.ray crystallography, 451 Normorphine 6-acetyl, 73 N-alkyl substitution, 75-76 biology, 73 metabolite, 14-15 QSAR, 184-185 synthesis, 72 Norprodine. analogs, SAR, 345-346 Nuclear magnetic resonance, 166 a-allylprodine, 380-382 jJ-allylprodine, 380-382 arylpiperidines. 378-380 cyclo-fLeu]enkephalin, 521 cyclo-[MetJenkephalin, 521 dimethylpiperidines, 380-384 dynorphin, 528-529 endorphins. 528-529 enkephalins, 516-531 Gly-Gly-Phe-Leu,518 Gly-Gly-Phe-Met,518 Gly-Tyr-Gly-Gly-Phe-Me[, 520 isomethadone, 449-450 [Leu]enkephalin, 516, 518-519, 522-523, 525,528,531 lipotropin. 528-529 Lys-Tyr-Gly-Gly-Phe-Met, 520 [Met]enkephalin, 516-517, 519-520, 523-528, 530-531 fMet]enkephalinamide, 519 methadone, 449-450 5-methylmethadone, 450 morphine, 170, 172 naltrexol, 170-171 neopine, 170 normethadone, 449 4-phenylpiperidines, 380-384 Phe-Tyr-Gly-GIY-Phe-Met, 520 a-prodine, 380-382
554
Index
{3-prodine, 380-382 a.prodine alcohol, 380-382 ,B-prodine alcohol, 380-382 thebaine, 170
Peffect of activation in chronic spinal dog, 31 in guinea pig ileum, 32 prototypic ligand, 31, 33 multiplicity, 31-35 biochemical characterization, 32-34 pharmacologic characterization, 31-32 probes, 67-72 selective protection, 32
384
[3-t,2,3-trimethyl-4-phenyl-4-piperidinol, 380-382, 384 y-I ,2,3-trimethyl-4-phenyl-4-piperidinol. 380-382, 384 a-I,2,5-trimethyl-4-phenyl-4-piperidinol, 380-382, 384 {3-t,2,5-trimethyl-4-phenyl-4-piperidinol, 380-382, 384
"
y-l,2,5-trimethyJ-4-phenyl-4-piperidinol. 380-382, 384 5-1 ,2,5-trimethyl-4-phenyl-4-piperidinol, 380-382 a-I,2,3-trimethyl-4-phenyl-4-propiony_ loxypiperidine, 380-382 {3-1,2,3-trimethyl-4-phenyI-4-propiony_ loxypiperidine, 380-382 y-I,2,3-trimethyl-4-phenyl-4-propiony_ loxypiperidine, 380-382 trimethylpiperidines, 380-384 Tyr-o-Ala-Gly-Phe-Met, 521-523 Tyr-D-Ala-Gly-Phe-Nva, 522 Tyr-cyc/o[N-e-Lys-Gly-Phe_LeuJ, Tyr-cyc/o[N-S-Orn-Gly-Phe_Leu), 533 Tyr-Gly-Gly-Phe, 518 Tyr-Met-Gly-Phe-Prn, 521, 523 Tyr-Tyr-GIY-Gly-Phe_Met, 520
555
in rabbit vas deferens, 32 prototypic ligand, 31, 33
1,2,6-trimethyl-4-phenyl-4-acetoxypiperi_ dine, 380-382 (t- J ,2.3-trimethyl-4-phenyl-4-piperidinol, 380-382,
Index
532 532-
o Operidine, see Phenoperidine Opiate analgesics, production quotas, 6 Opiate receptor ~ in rat vas deferens, 32 prototypic ligand, 33 initial discovery, 29 K distribution in central nervous system, 36-37 effect of activation, in chronic spinal dog, 31
elTect of activation. in chronic spinal dog, 31 prototypic ligand, 31, 33 Dpioid, biological effects, summary, 30 Opioid peptides, 459-470 biosynthesis, 463-470 from adrenal glands, 466 history, 459-470 types, 460 Opium, 1-5 alkaloids, 2 economics, 5 history, I production, 5 Optical rotatory dispersion, 166 Oripavine biosynthesis, 12 derivatives of, PCILO calculations, 182183 Orvinols; see also specific compound acid-catalyzed rearrangement, 140-147 acetylation at C-3, 132-133 analgesic activity, 131-132, 134 N-substitution, SAR, 137-138 3-deoxy, 139 6,14-endoethano, 139-140 Oxilorphan, 227-228 Oxycodone, %-97; see also Oxymorphone, azine biology, 97 conversion to oxymorphone, 97 ester SAR, 97-99 Oxymorphone, 66, 69, 76 azine, 66 biology, 97 hydrazone, 66 x-ray crystallography, 167-168
P pA2, definition,
32
Pallidine, biomimetic synthesis, 203 Partition coefficient, 166, 173 Pattern recognition, 183 PCILD calculations benzomorphans, 176-181, 183 dimethylpiperidines, 390 fentanyl, 176 y-isopromedol, 390 morphinans, 176 morphine, 176 nalorphine, 177-179 oripavine derivatives, 182-183 4-phenylpiperidines, 176,390-394 a-prodine, 390-394 /J-prodine, 390-394 a-promedol, 390 [D-Pen2,o-Cys~J-enkephalin, relative affinity, for () receptor, 35 ID-Pen2 ,D-Cysl]-cnkephalinamide, rchllive affinity, for () receplor, 35 1[)-Pen2,L-Cys~]-enkephalin. relative affinity. for S receptor, 35 ID-Pen2,L-Cysl]-enkephalinamide, affinity. for S receptor, 35 ID-Pen2.o-Penl]-enkcphalin. ity, for S receplor.
relative
relative
affin-
relative
affin-
35
[D-Pen2,L-Pen~J-enkepharin. ity. for S receptor, 35
Pentazocine, 278-279 effect, respiration, C02 stimulus-response curve, man, 37 enantiomers, 280 ll-stereoisomers, 280 Peptide E, 461 Peptide synthesis, 471-473 solid phase methods, 472 solution methods, 471 PET, 130, 153 Pethidine, QSAR, PH-8P,460 structure,
see Meperidine 184-185 461
Phenampromid, 428-430 Phenazocine, 277 relative affinity, for J.Lreceptor, Phenethyl thebaine, Phenobarbital, 177
130, 153
34
Phenoperidine, 325-326 I-Phenyl-3-aminotetralins, QSAR, 183184 4-Phenyl-4-piperidinols synthesis, 336-337 SAR,339 6-Phenylbenzomorphans SAR 261, 263, 266-268 N-substitution, 277, 281 synthesis, 260-262, 266-267 4-Phenylpiperidines conformation, 378-394 molecular modeling, 388-394 nuclear magnetic resonance, 380-384 PCILO calculations, 176 physicochemical
studies,
377-384
QSAR, 392-398 x-ray crystallography, 378-380 4-Phenylpiperidine-2-carboxylic acids, use in benzomorphan synthesis, 260 Phe-Tyr-Gly-Gly-Phe-Met, nuclear magnetic resonance, 520 Phosphatidylserine, .\'ee Phospholipid Phospholipid interaction with dynorphin. 530 interaction with {Met]enkephalin, 530531 Piminodine, 325 Piperdones, use in benzomorphan synthesis,270 POMC, see Proopiomelanocortin Potassium, interaction with enkephalins, 531 Prodine; see also 3-Desmethylprodine, specific compound analgesic activity, 341-342, 346 analogs C-3 alkylation; see also a-Prodine, {3Prodine C-3 alkylation, isomers, SAR, 341346 configuration, 341-342, 346 N-I substitution, see Norprodine ring-contraction, SAR, 351-352; Jee also Prodilidene ring.expanded, SAR. 351 a.Prodine conformation, 385, 388-394 molecular modeling, 388-394
/
556 nuclear magnetic resonance, 380382 stereochemistry, 385 x-ray crystallography, 378 ,B-Prodine, 341 absolute configuration, 342 analgesic activity, 341-342 analogs, 345; see also Norprodine antipodes. SAR, 342 conformation, 385. 388-394 molecular modeling, 388-394 nuclear magnetic resonance, 380-382 stereochemistry, 385 x-ray crystallography. 378 a-Prodine alcohol, nuclear magnetic resonance, 380-382 J3-Prodine alcohol, nuclear magnetic resonance, 380-382 Prodynorphin, 468-470 structure, 469 Profado!, 356-357 Proenkephalin. 465-468 structure. 467 Proenkephalin 8, see Prodynorphin Promedol, see also 1,2,5-Trimelhyl_4_ phenyl-4-piperidinoJ stereoisomers, SAR, 349-351 synthesis, 347 Proopiocortin, see Proopiomelanocortin Proopiomelanocortin, 463 biosynthesis, 465 processing, 465 structure, 464 Propiram,432-435,454_455 Propoxyphene, 427-428, 453-454 N-Propylnormorphine, antagonist properties, 75 J9-Propylthevinol nuclear magnetic resonance, 172 x-ray crystallography, 167, 169 Proton affinity, 176, 183 Proxorphan, 240 Pyridinium salts Grignard addition, 253, 256 use in benzomorphan synthesis, 253, 256, 259 5a,7a,8{3.( - )-N-[7-(I-PyrrolidinYI)-I_oxas_ piro[4,5]dec.8-yIJbenzeneacetamide, relative affinity, for K receptor, 35
Index
Index 557
Q Quantitative structure-activity relationships (QSAR), 183-185 dihydromorphine, 184-185 enkephalins, 537 etorphine, 184-185 fentanyl, 184-185 2.(2-furyl)ethyJ levorphanol, 184-185 hydromorphone, 184-185 ketobemidone, 184-185 levorphanol, 184-185 methadone, 184-185 3-methylfentanyl analogs, 397-398 morphine, 184-185 normorphine, 184-185 l-phenyl-3.aminotetralins, 183-184 4-phenylpiperidines, 392-398 Quasi.morphine withdrawal syndrome, 306 Quaterary morphine salts, 74-75
s Salutaridine biomimetic synthesis, 203 morphine precursor, 8, 10-11 occurrence, 189 opium alkaloid, 2 synthesis, 48, 192 Salutaridinoll, morphine precursor, 10-11 SD-25, see SyndyphaJin [Ser2,LeujJenkephalin, fluorescence spectroscopy, 527-528 Sinomenine, conversion to (+)-morphine, 189 Sinomeninone, 85 SKF-10047, see N-Allylnormetazocine Sodium, interaction with enkephalins, 531 Steric mapping, 184-185 Sublimase, see Fentanyl Sufentanil, 365 Syndyphalin, 477
R Racemethorphan, 207 Receptor binding acetylmethadol,414 N,N.dinoracetylmethadoJ, 414 N,N-dinormethadol,414 methadol,414 methadone, 414 N-noracetylmethadol,414 N-normethadol, 414 propoxyphene, 428 Respiration, method of measurement, 37 Respiratory depression by opioids, 37-38 {j receptor, and involvement, 38 J.Lreceptor, and involvement, 38 Reticuline morphine precursor, 9-11 oxidation, 206 Reversed esters of meperidine, 334-352; see al.\'O4.Acyloxy-4-arylpiperidines, Prodine, specific compound RevivonR, see Diprenorphine Rimorphin, 468 bioJogy, 470 opiate receptor affinities, 470 structure, 461
T TengesicR, see Buprenorphine Tetrahydropyridines, use in benzomorphan synthesis, 253, 256 Thebaine, 81 air oxidation, 193 Birch reduction, 193,212 bromination, 192 6-demethoxy-, /11-112 Diels-Alder adducts, 101-153; see also specific compound 7-alkyl, analgesia, 115 7-alkyl, synthesis, 114-115 7-amino, 117-119 14-arylamino, analgesia, 108 base-catalyzed rearrangement, 112113,121-122 C-7 derivatives, 126-127 C-7 substitution, 103-107 C-14 substitution, 106-108 chemical anatomy, 156-157 O-demethylation, 108-111 7,8.disubstitution, synthesis, analgesic activity, 122 7,8-disubstitution, synthesis, from ethyJenes, 118-120
7,8-disubstitution, synthesis, from maleic anhydride, 120-121 general structure, 102 7-keto, analgesia, 106 7-keto, conversion to alcohols, 128129; see also Thevinols 7-keto, narcotic antagonism, 136 7-methylfumaramido, receptor probe, 117-118 opiate receptor probes, 153-154 7-oxo, analgesia, 110-111 7-oxo, conversion to 7{3-alcohol, 149150 peptide derivatives, 115-117 stereochemistry, assignment, 103-105, 118-120 stereochemistry, physical methodol. ogy, 124-126 7-sulfone, analgesia, 107 synthesis, acetylene dienophiles, 123 synthesis, aromatic nitroso dienophiles, 106-108 synthesis, ethylene dienophiles, 103107, 118-120 fJ-dihydro-, 111-112 from Papaver brae/ea/urn, 5, 7 hydride reduction, 193 morphine precursor, 8, 10-12 nuclear magnetic resonance, 170 opium alkaloid, 2 organocuprate addition, 221-222 oxidation by hydrogen peroxide, 95 by peracids, 95 reaction with cuprates, 86 reaction with N204, 98 rearrangement to metathebainone, 235236 synthesis, 48 Thevinoic acid, 116 Thevinols A-ring derivatization, 147-148 acid.catalyzed rearrangement, 140-147 C-19 configuration, 128-129 7{3-diastereoisomers, 149-150 15,16-modification, 147-148 16-alkyl, 149 N-substitution synthesis, 135-136
558 SAR, 136-137 O-demethylation, 131-132 at C-3, see Oevinols reduction to 6,14.endoethano, 134 stereochemistry physical methodology, 150-153 x-ray crystallography, 153 synthesis from thevinone, 128-129 Thevinone, 105, 112-113 Thiambutene, 436-437, 455 Tilidine, 439 Trimeperidine, see Promedol .. 1,2.6- Trimethyl-4-phenyl-4-acetoxyplpendine conformation, 387 nuclear magnetic resonance, 380-382 stereochemistry, 387 x-ray crystallography, 378 1,2,5- Trimethyl-4-phenylpiperidines, nomenclature, 378 a-I ,2,3- Trimethyl-4-phenyl-4-piperidinol conformation, 385-386 nuclear magnetic resonance, 380-382, 384 stereochemistry, 385-386 x-ray crystallography. 378 . .. {3-1,2,3- Trimethyl-4-phenyl-4-plpendlnol confonnation, 385-386 nuclear magnetic resonance, 380-382, 384 stereochemistry, 385-386 x-ray crystallography, 378 y-J 23-Trimethyl-4-phenyl-4-piperidinol c~~fonnation, 385-386 nuclear magnetic resonance, 380-382, 384 stereochemistry, 385-386 x-ray crystallography, 378 a-I,2,5- Trimethyl-4-phenyl-4-piperidinol conformation, 386 nuclear magnetic resonance, 380-382, 384 stereochemistry, 386 x-ray crystallography. 378 {3-I,2,5- Trimethyl-4-phenyl-4-piperidinol conformation, 386 nuclear magnetic resonance, 380-382, 384 stereochemistry, 386 y- J ,2,5- Trimethyl-4-phenyl-4-piperidinol
Index
Index 559
conformation, 386 nuclear magnetic resonance, 380-382. 384 stereochemistry, 386 x-ray crystallography. 378 5-1 ,2,5- Trimethyl-4-phenyl-4-piperidinol conformation. 386 nuclear magnetic resonance, 380-382 stereochemistry, 386 a-I,2 ,3-Trimethyl-4-phenyl-4-propionyloxy_ piperidine. nuclear magnetic resonance, 380-382 {3-1,2,3- Trimethyl-4-phenyl-4-propionyloxy_ piperidine, nuclear magnetic resonance, 380-382 y-I ,2,3- Trimethyl-4-phenyl-4-propionyloxypiperidine, nuclear magnetic resonance, 380-382 ')'-1,3,5- Trimethyl-4-phenyl-4-propionyloxypiperidine conformation, 387 stereochemistry. 387 x-ray crystallography, 378 Trimethyl-4-phenylpiperidines, nuclear magnetic resonance, 380-384 [Trp4]enkephalins fluorescence spectroscopy, 527, 535 Monte Carlo simulation, 535 [Trp4,Me-Leu")enkephalin, Monte Carlo simulation, 535 [Trp\Met")enkephalin, Monte Carlo simulation, 535 Tyr-Ala-Gly-Phe, Monte Carlo simulation, 535 Tyr-cyclo[N-y-Dbu-Gly-Phe-Leu], 533 Tyr-cyclo[N-S-Om-Gly-Phe-Leu], 532-533 Tyr-cyclo[N-B-Lys-Gly-Phe-LeuJ. 532-533 Tyr-o-Ala-Gly-MePhc-Gly-ol, relative affinity, for 1J-receptor. 34; ,W>('al.m DAGO. 477 Tyr-D-Ala-Gly-MePhe-Me(O).ol. .\'('(' FK33-824 Tyr-D-Ala-Gly-MePhe-Met-(O).l)I. relative affinity. for 1J-receptor. 34 Tyr-D-Ala.Gly-Phe, Monte Carlo simulation, 535 Tyr-D-Ala-Gly.Phe-D-Leu, see DADLE [Tyr-D-Ala-Gly-Phe-LeuNHh'(CHh, relative affinity, for 5 receptor. 34
Tyr-n-AIa-Gly.Phe-(Me)"-lct_NH!. mctkephamid. 493 Tyr-o-Ala-Gly-Phe_Met. nuclear resonance. 5~ 1-523
s('(' magnetic
[Tyr-D-Ala-UIY-PhcNHh'(CH:.),~, affinily. for 0 receptor. 34 Tyr-n-Ala-Gly-PhC-Nva. resonance. 522 Tyr-D-Cys-Gly-Phe_Cys. Tyr-D-Cys-Gly-Phe_n_Cys.
X-ray
,\'('(' Syn480 480
Tyr-n-Ser-Gly-Phe-Leu_Thr. ity. for 0 receptor.
relative
affin-
34
Tyr-D-Thr-Gly-Phe-Leu_Thr, relalive affinity. for 8 receptor, 34: .I't'(' (/1.\'0 DTLET Tyr.Gly-Gly. circular dichroism. 528
see [LeuJenkepha_
Tyr.Gly-Gly.Phe-Met, see {MetJenkephalin Tyr.Met.Gly-Phe-Pro, nuclear magnetic resonance, 521, 523 Tyr-Met-Gly-Trp_Pro, fluorescence spectroscopy, 522, 527 Tyr-Pro-Gly-Phe-Leu, troscopy, 522
fluorescence
spec.
Tyr-Pro-GJy-Phe.Met, troscopy, 522
fluorescence
spec-
Tyr-Tyr.Gly-Gly-Phe_Met, netic resonance, 520
nuclear
crystallography, 166 &-acetyl-I-iodocodeine. J67 acetylmethadol, 453 a.allylprodine. 378 f3.allylprodine. 378 arylpiperidines, 378-380 azidomorphine, 167, 169
{4'Br.Phe4J-(LeuJenkephalin, [4'Br-Phe4J-(Met]enkephalin, codeine, /67-168 dextromethorphan. dextromoramide,
515 515
167, 169 453
f3-dimethyl-4-phenyl.4-propionyloxypi. peridine, 378
Tyr-Gly-Gly-Phe Monte Carlo simulation, 535 nuclear magnetic resonance, 518 x-ray crystallography, 515 Tyr-GIY-Gly-Phe-Leu, lin, enkephalins
Viminol. 444
x
47X. 4XO 478. 4XO clinical
Tyr-D-Mct(O)-Gly-MePhe_ol. dyphalin l'yr-n-IJen-Gly_Phc_Pcn. Tyr-n-Pen-Gly-Phe_D_Pen,
magnetic
529-530 516, 522
v
relative
nuclear
Tyr-n-Met-Gly-Phe-Pro_NH~. investigations. 500
f3.endorphin, enkephalins,
mag-
u V-50,488, see trans-3,4-dichloro.N.methyl. N.{2.( l.pyrroJidinyl)-cyclohex yl]benze neacetamide U-69.593, see 5a,7a,8f3-( -)-[N-(7-(I-pyrro_ lidinyl)-I.oxaspiro(4.5)dec_8_yl)ben_ zeneacetamideJ Ultraviolet spectroscopy
Gly-Gly-Phe-Leu, 515 14-hydroxyazidomorphine, 167 3-hydroxylevallorphan. 167, 169 isomethadone. 451 [LeuJenkephalin, 514-515 meperidine, 378 [MetJenkephalin, 515 methadol. 453 methadone, 451 5-methylmethadone, 451 N-methyJnalorphine, 167 morphine, 167-168, 173 nalbuphine, 167, 169 nalorphine, 167-]68 naloxone, 167-168 normethadone, 451 N.norpropoxyphene, 454 oxymorphone, 167-168 4-phenyJpiperidines. 378-380 a-prodine, 378 p.prodine, 378 propoxyphene, 453-454 19-propylthevinol, 167, 169 I ,2,&-trimethyl-4-phenyl_4_acetox ypiperidine, 378 a-I ,2,3-trimethyl-4-phenyl-4-piperidinol, 378
Index
56()
{3-I,2,3-trimethyl-4-phenyl-4-piperidinol, 378 '1-1 ,2,3-trimethyl-4-phenyl-4-piperidinol.
378 a-l.2,S-trimethyl-4-phenyl-4-piperidinol, 378
y-l.3,S-trimethyl-4-phenyl-4-propionoxypiperidine. 378 Tyr-Gly-Gly-Phe, 515
z
'1-1.2,5-trimethyl-4-phenyl-4-piperidinol.
378
Zactane, see Ethoheptazine
