Engineering of Halogenases

Hannah Minges

Engineering of Halogenases towards Synthetic Applications Increasing the Thermostability and Investigations on a Marine Brominase Bmp5

BestMasters

Springer awards „BestMasters“ to the best master’s theses which have been completed at renowned Universities in Germany, Austria, and Switzerland. The studies received highest marks and were recommended for publication by ­supervisors. They address current issues from various fields of research in natural sciences, psychology, technology, and economics. The series addresses practitioners as well as scientists and, in particular, offers guidance for early stage researchers.

More information about this series at http://www.springer.com/series/13198

Hannah Minges

Engineering of Halogenases towards Synthetic Applications Increasing the Thermostability and Investigations on a Marine Brominase Bmp5

Hannah Minges Bielefeld, Germany This master thesis was kindly supervised by Prof. Norbert Sewald from Bielefeld University, Organic and Bioorganic Chemistry.

BestMasters ISBN 978-3-658-18409-4 ISBN 978-3-658-18410-0  (eBook) DOI 10.1007/978-3-658-18410-0 Library of Congress Control Number: 2017941551 Springer Spektrum © Springer Fachmedien Wiesbaden GmbH 2017 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Printed on acid-free paper This Springer Spektrum imprint is published by Springer Nature The registered company is Springer Fachmedien Wiesbaden GmbH The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany

Table of contents List of figures ........................................................................................................................ 7 List of schemes...................................................................................................................... 9 List of abbreviations .......................................................................................................... 11 1

Introduction ............................................................................................................... .13 1.1 Value of biocatalysis in chemical synthesis ................................................ .13 1.2 Natural halogenated compounds ................................................................... 13 1.3 Halogenating enzymes .................................................................................. 15 1.4 The FAD-dependent halogenase family ....................................................... 15 1.5 Biocatalytic application of tryptophan halogenases ..................................... 20 1.6 Directed enzyme evolution ........................................................................... 21 1.7 Halogenated marine metabolites ................................................................... 23

2 3

1.8 FAD-dependent halogenase Bmp5 ............................................................... 23 Motivation ................................................................................................................... 25 Material and Methods ................................................................................................ 27 3.1 Analytics ....................................................................................................... 27 3.2 Laboratory Equipment .................................................................................. 30 3.3 Chemicals and solvents ................................................................................. 31 3.4 Molecular biological methods ...................................................................... 39 3.5 Biochemical methods .................................................................................... 44 3.6 Directed evolution......................................................................................... 48

4

3.7 Biotransformations........................................................................................ 51 Directed evolution of Thal: Results .......................................................................... 59 4.1 Directed evolution of tryptophan 6-halogenase Thal ................................... 59 4.2 Investigating the thermostability of Thal-E2 and -E2R ................................ 73 4.3 Bromination ability of Thal variants and comparison to the wild type................................................................................................................ 75

5

4.4 Preparative synthesis of L-6-bromotryptophan using Thal-PrnFADH combiCLEAs ....................................................................................... 78 Directed evolution of Thal: Discussion ..................................................................... 81

6

5.1 Directed evolution of tryptophan 6-halogenase Thal ................................... 81 Establishment of a marine brominase: Results ....................................................... 89 6.1 Establishment of the marine brominase Bmp5 ............................................. 89 6.2 Identification of Bmp5 .................................................................................. 93 6.3 Enzymatic bromination of 4-HBA after purification of His6-Bmp5 ............ 94 6.4 Generation of GST-tagged Bmp5 ................................................................. 96

5

6.5 Establishment of cofactor regeneration for Bmp5 in E. coli lysate .............. 99 6.6 Investigation of Bmp5's substrate scope ..................................................... 105

7

6.7 Preparative synthesis of 2,4-dibromophenol using Bmp5-ADH combiCLEAs .............................................................................................. 107 Establishment of a marine brominase .................................................................... 109 7.1 Establishment of the brominase Bmp5 for heterologous expression .......... 109 7.2 Expression of GST-Bmp5 fusion protein ................................................... 110 7.3 His6-Bmp5 revealed sufficient enzymatic activity after purification ......... 111 7.4 Establishment of cofactor regeneration system for Bmp5-catalyzed reactions ...................................................................................................... 112 7.5 Evidence of Bmp5 expression by Western blot analysis and peptide mass fingerprint .......................................................................................... 113

8 9

6

7.6 Detected reaction compounds confirmed proposed Bmp5-catalyzed reaction mechanism .................................................................................... 114 Outlook ...................................................................................................................... 121 Summary ................................................................................................................... 123 Literature .................................................................................................................. 127

List of figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Figure 23 Figure 24

Figure 25 Figure 26 Figure 27

Selection of halogenated secondary metabolites. ......................................... 14 Crystal structure of the tryptophan 7-halogenase PrnA. ............................... 17 Crystal structure of PrnA's active pocket. ..................................................... 18 General principle and strategy for directed enzyme evolution. .................... 22 EpPCR of 1.6 kb thal using pET28a-thal as template. ................................. 60 SDS-PAGE analysis of Thal WT expression in E. coli BL21 (DE3) pGro7 pET28a-thal. ........................................................................... 62 Investigation of Thal WT's thermostability. . ................................................ 64 Evaluation of the high-throughput screening. ............................................... 67 Re-screening of Thal mutants. ...................................................................... 68 SDS-PAGE analysis of Thal-E2 expression in E. coli BL21 (DE3) pGro7 pET28a-thal-E2. .. ................................................................... 70 Site-directed mutagenesis of Thal-E2. .......................................................... 71 SDS-PAGE analysis of Thal-E2R expression in E. coli BL21 (DE3) pGro7 pET28a-thal-E2R. ................................................................... 72 Comparison of Thal WT's, E2's and E2R's remaining bromination activity in dependence of heat-shock temperature. ................... 74 Comparison of Thal WT's and E2R's catalytic activity in crude lysate. ............................................................................................................ 76 Comparison of the bromination activity between Thal WT and E2R in crude lysate. . ..................................................................................... 77 Comparison of bromination activity between Thal WT- and Thal-E2R. ...................................................................................................... 78 1 H-NMR spectrum of L-6-bromotryptophan. ................................................ 79 Excerpt of Thal-E2R's proposed structure. ................................................... 87 Separation of bmp5 from residual pMK-RQ vector by agarose gel electrophoresis. ....................................................................................... 90 Identification of bmp5 containing colonies via colony-PCR. . ...................... 90 SDS-PAGE analysis of Bmp5 expression in E. coli BL21 (DE3) pGro7 pET28a-bmp5. . ....................................................................... 91 SDS-PAGE analysis of Bmp5 expression in E. coli BL21 (DE3) pET28a-bmp5. ..................................................................................... 92 Detection of His6-Bmp5 by Western blot analysis. ...................................... 93 Approach for Bmp5-catalyzed bromination of 4hydroxybenzoic acid to 2,4-dibromophenol with cofactor regeneration. .................................................................................................. 95 Separation of the 1 kb insert from desired 6 kb pETM30 vector by restriction digest. ........................................................................... 96 Identification of GST-bmp5 containing colonies via colonyPCR. .............................................................................................................. 97 SDS-PAGE analysis of GST-Bmp5 expression in E. coli BL21 (DE3) pETM30-bmp5. ........................................................................ 98 7

Figure 28 Figure 29

Figure 30 Figure 31 Figure 32 Figure 33 Figure 34 Figure 35

8

Progress of Bmp5-catalyzed conversion of 4-hydroxybenzoic acid. . ............................................................................................................ 101 GC-MS analysis of Bmp5-catalyzed conversion of 4hydroxybenzoic acid to 2,4-dibromophenol via an intermediate. ................................................................................................ 102 GC-MS analysis of Bmp5-catalyzed conversion of 4hydroxybenzoic acid to 2,4-dibromophenol. .. ............................................ 103 GS-MS identification of side products formed during Bmp5catalyzed bromination. . ............................................................................... 104 GC-MS results of phenol bromination catalyzed by Bmp5. . ...................... 106 Bmp5-ADH combiCLEAs-catalyzed conversion of 4hydroxybenzoic acid. .................................................................................. 107 Structural elucidation of 2,4-dibromophenol via 1H-NMR spectroscopy. ............................................................................................... 108 Analysis of Bmp5's substrate scope.. ........................................................... 117

List of schemes Scheme 1 Scheme 2 Scheme 3 Scheme 4 Scheme 5 Scheme 6 Scheme 7 Scheme 8 Scheme 9 Scheme 10

Scheme 11

Chlorination of L-tryptophan by the FAD-dependent halogenase PrnA. . ......................................................................................... 16 Reduction of FAD by NADH to FADH2 and NAD+. ................................... 16 Halogenation cycle of PrnA. . ........................................................................ 19 Biosynthetic pathway of thienodolin. ........................................................... 20 Proposed two-step reaction mechanism of the Bmp5-catalyzed conversion of 4-hydroxybenzoic acid to 2,4-dibromophenol. . ..................... 24 Strategy for directed evolution of Thal towards a thermostable variant. .......................................................................................................... 66 Bmp5 regenerates FADH2 by oxidizing NADPH to NADP+. ...................... 94 Bmp5-catalyzed conversion of 4-hydroxybenzoic acid via an intermediate to the formation of 2,4-dibromophenol. .................................... 95 Bmp5-catalyzed conversion of 4-hydroxybenzoic acid to 2,4dibromophenol employing GDH for regeneration of NADPH. .................... 99 Approach for Bmp5-catalyzed bromination of 4hydroxybenzoic acid to 2,4-dibromophenol in presence of continuous cofactor regeneration. . .............................................................. 100 Proposed reaction mechanism for Bmp5-catalyzed conversion of 4-hydroxybenzoic acid. ........................................................................... 115

9

List of abbreviations ADH amp APS bp BSA 5-Br-Trp 5-Cl-Trp 6-Br-Trp 6-Cl-Trp 7-Br-Trp 7-Cl-Trp cam DNA epPCR E. coli EI ESI EtOH FAD / FADH2 FW GST HCCA HEPES HPLC HRP IMAC iPrOH IPTG kan kDa LK-ADH LC-MS MeCN MeOH MS NAD+ / NADH NADP+ / NADPH NMR OD p.a. PBS PAGE

alcohol dehydrogenase ampicillin ammonium persulfate base pair bovine serum albumin L-5-bromotryptophan L-5-chlorotryptophan L-6-bromotryptophan L-6-chlorotryptophan L-7-bromotryptophan L-7-chlorotryptophan chloramphenicol deoxyribonucleic acid error-prone PCR Escherichia coli electron impact ionization electrospray ionization ethanol flavin adenine dinucleotide forward Glutathione S-transferase α-Cyano-4-hydroxycinnammic acid 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid high performance liquid chromatography horseradish peroxidase immobilized metal ion affinity chromatography iso-propanol isopropyl-β-D-thiogalactopyranoside kanamycin kilodalton ADH from Lactobacillus kefir liquid chromatography-mass spectrometry acetonitrile methanol mass spectrometry nicotinamide adenine dinucleotide nicotinamide adenine dinucleotide phosphate nuclear magnetic resonance optical density pro analysi phosphate-buffered saline polyacrylamide gel electrophoresis 11

PDB PCR ppm P. fluorescens Ponceaus S REV ROESY RP rpm RR-ADH RT SDS SPhos water soluble T7P T7T taq TEMED Tris TFA Tween WT

12

protein data bank polymerase chain reaction parts per million Pseudomonas fluorescens 3-Hydroxy-4-(2-sulfo-4-[4-sulfophenylazo] phenylazo]-2,7naphthalenedisulfonic acid sodium salt reverse rotating-frame overhauser effect spectroscopy reversed phase revolutions per minute alcohol dehydrogenase from Rhodococcus spp. room temperature sodium dodecyl sulfate sodium 2′-dicyclohexylphosphino-2,6-dimethoxy-1,1′biphenyl-3-sulfonate hydrate T7 promoter T7 terminator Thermus aquaticus N,N,N′,N′-tetramethylethylenediamine 2-Amino-2-hydroxymethyl-propane-1,3-diol trifluoroacetic acid polyethylene glycol sorbitan monolaurate wildtype

1

Introduction

1.1

Value of biocatalysis in chemical synthesis

In living systems the majority of biological reactions is being catalyzed by enzymes. The special feature of biocatalysts lies in their ability to increase reaction rates to a million fold without being consumed or permanently altered by the reaction. An interesting and useful aspect to employ enzymes in organic chemistry is their chirality and unique stereostructure, resulting in a high substrate specificity as well as regio- and stereoselectivity.[1,2] Biocatalyzed reactions in comparison to the correlates catalyzed by inorganic, organic or organometallic compounds are superior because of their safety and economic aspects due to milder reaction conditions, e.g. regarding temperature, pressure, the use of less toxic reagents or organic solvents. In addition, enzymes provide new opportunities for synthetic transformations that were otherwise impossible or inefficient to perform. For example, if desired substrate positions are not accessible by conventional methods, biocatalysts might address these sites with remarkable specificity. The resulting chemically well-defined products are of great value for medicine, agriculture or pharmacy and may allow the production of chemicals from renewable resources. Thanks to this more energy saving, environmentally friendly and resource efficient approach, the use of biocatalysts in chemical transformations paves the way to sustainable and green chemistry.[3] 1.2

Natural halogenated compounds

Halogenated compounds are an important and reactive compound class in nature. Prior to 2013 more than 4700 naturally occurring halogenated natural products have been discovered.[4] Biohalogenation occurs on a diversity of organic scaffolds, where the halogen atoms can be incorporated on aliphatic carbons, olefins as well as aromatic and heterocyclic rings.[5]

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_1

13

Figure 1

Selection of halogenated secondary metabolites as representatives for chlorinated, brominated and iodinated natural products.

Chlorination and bromination are more abundant motifs compared to iodination and fluorination due to the higher concentration of chloride and bromide ions in the biosphere. Naturally chlorinated metabolites, such as the antibiotics chloramphenicol (1) and chlortetracycline (2) as well as the antitumor agent rebeccamycin (3), carry chlorine substituents in their aromatic moieties and are especially of therapeutic interest (Figure [5–7] 1). Brominated metabolites such as pentabromopseudiline (4) are in particular synthesized in marine organisms. Due to bromide's natural abundance in seawater compared to its minor terrestrial presence bromination can be seen as characteristic feature for marine natural products.[8] Iodinated natural products occur rarely in nature. One example is the human thyroid hormone L-thyroxine (5).[9] Even though fluoride is highly abundant in the earth's crust, only few fluorinated metabolites have been identified so far. In contrast to the oxidative mechanism of the C-Cl, C-I and C-Br bond formation, the C-F bond formation follows a nucleophilic attack of the fluoride ion. The difficulty of desolvating the electronegative fluoride ion in aqueous media, which is an essential feature to act as a naked nucleophile, as well as the toxicity of accumulating fluorinated metabolites in the producer organism are considered as reasons for the limited amount of known fluorinated metabolites.[5] Only a few native fluorinating enzymes have been characterized in detail so far, such as the fluorinase from Streptomyces cattleya. It is a nucleophilic halogenase and catalyzes the reaction between

14

a fluoride ion and S-adenosyl-L-methionine (SAM) to generate 5’-fluorodeoxy adenosine (FDA) and L-methionine.[10,11] 1.2.1

Synthetic interest in halogenated compounds

The synthetic interest in halogenated compounds lies in their ability to act as important building blocks that can be employed in diverse ways for organic synthesis. They can serve as starting material for further modifications, e.g. by cross-coupling reactions, nucleophilic- or electrophilic aromatic substitution.[12] Electrophilic aromatic substitution belongs to the most important reactions in organic chemistry, leading to valuable precursor molecules that can be further modified to pharmaceuticals[13], agrochemicals[14] or other industrial products. Regarding halogenation of arenes, this reaction proceeds mainly by electrophilic aromatic substitution. Even though this approach requires harsh reaction conditions, the introduction of the halogen atom at electronically and mechanistically unfavoured positions is very challenging. Consequently, a crude mixture of different regioisomers might be obtained, resulting in a low overall yield of the desired product, unwanted byproducts and toxic waste.[15] 1.3

Halogenating enzymes

In contrast to the conventional approach, the enzymatic halogenation offers a milder and safer way for the synthesis of halogenated compounds. It proceeds at almost neutral pH and 25 °C in aqueous media, avoids toxic halogenation agents and does not require any activating or protection groups due to the enzyme's high regio- and stereoselectivity.[16] Halogenating enzymes can be grouped into four classes, each requiring another redox cofactor for catalytic activity. These halogenases are either heme-dependent, non-hemedependent, vanadium-dependent or flavin-dependent. The heme- and vanadiumdependent enzymes belong to the haloperoxidases and require H2O2 as oxidation agent. O2-dependent halogenases are subdivided into flavin-dependent and non-heme, Fe2+dependent halogenases that require dioxygen as reducible co-substrate for their enzymatic cycle.[5] 1.4

The FAD-dependent halogenase family

In 2000 van Pée and coworkers purified and partially characterized the first member of the FAD-dependent halogenase family, the tryptophan 7-halogenase PrnA from Pseudomonas fluorescens.[17] PrnA catalyzes the first step in the synthesis of the antibiotic pyrrolnitrin (6) by regioselective chlorination of L-tryptophan (7) at the C7 position of the indole scaffold. To perform the chlorination or bromination of L-tryptophan the halogenase requires flavin, a halide salt and O2 as stoichiometric components (Scheme 1).[18]

15

Scheme 1

Chlorination of L-tryptophan (7) by the FAD-dependent halogenase PrnA. PrnA catalyzes the chlorination of 7 at the C7-position of the indole moiety at pH 7.4 and 25 °C. The required FADH2 is regenerated by the flavin reductase PrnF.

Van Pée et al. also identified the absolute necessity of a flavin reductase for the activity of PrnA.[17] PrnF regenerates the cofactor FADH2 (11) by oxidizing NADH (12) to NAD+ (13) (Scheme 2).

Scheme 2

The reduction of FAD (9) by NADH (12) to FADH2 (11) and NAD+ (13) is catalyzed by the flavin reductase PrnF.

Apart from PrnA there are other FAD-dependent tryptophan halogenases, such as RebH from the rebeccamycin (3) biosynthesis in Lechevalieria aerocolonigenes. RebH that is highly homologous to PrnA catalyzes the incorporation of a halogen substituent at the C7 position of L-tryptophan.[19] In contrast, PyrH from Streptomyces rugosporus is a Trp 5-halogenase that is involved in the pyrroindomycin synthesis.[20] Thal from Streptomyces albogriseolus catalyzes the first step in the thienodolin (16) biosynthesis and is able to address the electronically unfavoured C6 position of L-tryptophan.[21] Two further FAD-dependent halogenases, KtzQ and KtzR from Kutznerian spp. were identified to chlorinate L-tryptophan in a tandem reaction sequence during the biosynthesis of kutzneride. Firstly, L-tryptophan is chlorinated at the C7-position by KtzQ and afterwards the second chlorine atom is introduced by KtzR at the C6-position of [22] L-7-chlorotryptophan (8).

16

1.4.1

Structure of FAD-dependent tryptophan halogenases

The crystal structure of PrnA was determined by Dong et al. in 2005 to 1.95 Ǻ resolution.[23,24] Up to now this structural information and the identification of the crystal structures of RebH[25] and PryH[26] helped to propose a reaction mechanism for FADdependent halogenases.

Figure 2

Crystal structure of the tryptophan 7-halogenase PrnA shown as monomer (PDB:2AQJ). The ribbon model shows α-helices in red, β-sheets in cyan and turns in green.[23]

PrnA is a 61 kDa homodimer consisting of two pyramidal shaped monomers (Figure [24] One monomer consists of two distinct binding sites for the substrates tryptophan 2). and FAD. The cofactor is bound in a solvent-exposed groove next to the parallel β-sheets which dominate the FAD-binding module. The chloride-binding site is located next to the isoalloxazine ring of FAD. The substrate tryptophan is bound in a distinct enzyme module, separated by a 10 Ǻ tunnel from the FAD-binding site. It is positioned next to H101, F103 and W455 and its indole nitrogen forms a hydrogen bond to E346. Tryptophan's amino group as well as its carboxylate group are hydrogen bonded, and the amino group forms a salt-bridge with E450. All these interactions lead to a rather rigid and fixed position of tryptophan, probably resulting in the high regioselectivity of the halogenation reaction (Figure 3).[23]

17

Figure 3

Crystal structure of PrnA's active pocket (PDB: 2AQJ). The protein scaffold is shown as electrostatic surface (negative: red, positive: blue). Important amino acid residues are highlighted as stick model. The substrates L-tryptophan, FADH2 and the chloride ion are depicted in green (N: blue, O: red). Hydrogen bonds are shown as dashed lines. The amino acid residues K79 and E346 are essential for the halogenase's catalytic activity. The ɛ-amino group of K79 is proposed to react with HOCl to form a stable chloroamine.[27] According to studies by Flecks et al., the chloroamine is supposed to re-compose into HOCl that finally performs the electrophilic aromatic substitution of [18,23,24] L-tryptophan.

Two highly conserved regions can be found in all FAD-dependent halogenases. The G / x / G / x / x / G motif is located near the NH2-terminus and is involved in the binding of the flavin cofactor. The second motif is located in the center of the enzyme and contains two tryptophan residues (W / x / W / x /IP). This region is responsible for the prevention of the binding of organic substrates to the isoalloxazine ring of FADH2, so that the halogenase does not act as a monooxygenase that is similar in mechanism.[23,28] 1.4.2

Proposed reaction mechanism for tryptophan 7-halogenase PrnA

As the mechanism of the enzymatic halogenation is still under discussion, different suggestions exist about the exact halogenation process. Focusing on the chlorination of L-tryptophan catalyzed by PrnA, the reaction is proposed to proceed as follows: the flavin reductase PrnF provides the required reduced flavin (11) by oxidizing NADH (12) (Scheme 2). The reduced FADH2 (11) binds to the halogenase’s active pocket and reacts with molecular oxygen to generate a flavin hydroperoxid intermediate FAD(C4a)-OOH (14). This intermediate reacts with a bound chloride ion to form hypochlorous acid (HOCl). 18

The active pocket of PrnA contains two amino acid residues, K79 and E346 that are essential for the catalytic activity of the tryptophan halogenase. The importance of K79 for the halogenation reaction was demonstrated by the mutation of K79 against an alanine residue that resulted in a complete loss of the halogenation activity. In the active pocket HOCl is guided by K79 from the FAD-binding site to the tryptophan-binding site. In PrnA as well as in RebH, K79 is located directly between the flavin and tryptophanbinding pockets.[23] The exact chlorination mechanism of L-tryptophan is still under controversial discussion, as different intermediates are proposed to act as halogenating species. As proposed by Yeh et al. the rapid reaction of HOCl with the ɛ-amino group of K79 leads to the formation of a long-lived chloroamine (15) as evidenced by radiolabeling using 36Cl. Nevertheless, this chloroamine could only be detected in absence of L-tryptophan, however incorporation of 36Cl into the substrate was observed.[27] In contrast, Flecks et al. proposed that the chloroamine's electrophilicity is too low to enable halogenation of Ltryptophan and therefore the free hypochlorous acid is supposed to be reformed. Further interactions of HOCl with E346 are responsible for the correct orientation of the halide species and probably increase the electrophilicity that finally reacts in an electrophilic aromatic substitution to yield L-7-chlorotryptophan (8) (Scheme 3).[18,23,27,29].

Scheme 3

Halogenation cycle of PrnA. In the first step (a) FADH2 (11) is oxidized by O2 to form the flavin hydroperoxide FAD(C4a)-OOH (14) which reacts with a chloride ion to HOCl (b). The hypochlorous acid reacts with the ɛ-amino group of K79 to form a chloroamine as proposed by Yeh et al [27] (c-d). This chloroamine is assumed to recompose, according to Flecks et al[18], to HOCl that performs the electrophilic aromatic substitution (f). After deprotonation of the Wheland intermediate (16) L-7-chlorotryptophan (8) is released from the active site (g).

19

1.4.3

Tryptophan 6-halogenase Thal

The tryptophan 6-halogenase Thal was discovered in Streptomyces albogriseolus and is involved in the biosynthesis of the plant growth-regulating compound thienodolin (18), where it catalyzes the initial step in its biosynthesis. As 18 is a tryptophan analogue, it is proposed that amino acid 7 serves as a biosynthetic precursor (Scheme 4).[21]

Scheme 4

Biosynthetic pathway of thienodolin (18). The tryptophan halogenase Thal catalyzes the initial reaction step, the chlorination of L-tryptophan (7) at the C6 position of the indole ring. Further biosynthetic reactions lead to the formation of 18. [21]

The comparison of Thal's amino acid sequence with the Trp 7-halogenase PrnA revealed a sequence identity of 57 %. Regarding the active site of Thal and PrnA, this region only differs by three amino acid residues. Even though Thal's crystal structure has not been elucidated yet, it can be proposed that Thal has structural similarity to other FADdependent halogenases due to its high degree of homology to RebH and PrnA. Thal should follow an analogous mechanism of halogenation, even though the orientation of the substrate in the active pocket is assumed to be slightly different, resulting in the distinct preferences for regioselective halogenation. This suggestion is supported by investigations of Zhu et al. who proposed a similar halogenation mechanism for the Trp 5-halogenase PyrH that is also structurally related to RebH.[26,30] 1.5

Biocatalytic application of tryptophan halogenases

1.5.1

Immobilization strategies

Regarding the preparative application of enzymatic halogenation, Frese et al. performed the regioselective bromination of L-tryptophan with RebH on a gram scale. In this approach RebH was co-immobilized with the flavin reductase PrnF from Pseudomonas fluorescens and an alcohol dehydrogenase from Rhodococcus spp. for cofactor regeneration as cross-linked enzyme aggregates (CLEAs). The combiCLEAs could be removed by filtration and after one desalting step the purified brominated amino acid was obtained in high yield and purity.[31] 1.5.2

Examination and modification of the substrate scope

Concerning the substrate scope of FAD-dependent halogenases, O'Connor et al. demonstrated that RebH is able to halogenate the unnatural substrate tryptamine, a direct precursor to many alkaloid natural products, at the C7 position. In addition, a RebH 20

mutant (Y455W) was generated by site-directed mutagenesis which preferentially installed chlorine into tryptamine rather than into its natural substrate tryptophan.[32] Two years later in 2013 Payne et al. focused in more detail on the substrate scope of RebH and demonstrated the chlorination and bromination of unnatural substrates such as several indoles and naphthalenes. As shown by Frese et al. in 2014, RebH is able to override directing electronic effects and halogenates several tryptophan derivatives at electronically unfavoured positions.[33] Even though PrnA and RebH are structurally very similar, PrnA has a narrower substrate scope and halogenates unnatural substrates only at electronically activated positions.[34] A reason for RebH's and PrnA's different selectivity for unnatural substrates might lie in their ability to position these substrates in their active pockets. A slight difference in the amino acid sequence of their active pockets contributes to this observation. In RebH a water-mediated hydrogen bond can be formed between N467 and the incoming substrate which might help to position it for halogenation at disfavoured positions. This hydrogen bond cannot be formed in PrnA, because this position is occupied with L456.[35] 1.5.3

Modifying tryptophan halogenases by site-directed mutagenesis

During the past years several approaches have been made to overcome the limitations of tryptophan halogenases, such as low activity and a narrow substrate scope by creating mutated variants of the enzyme. As shown by the results of O'Connor et al. and Lang et al., site-directed mutagenesis of tryptophan halogenases allows to engineer the enzyme's substrate scope, what opens new ways for the synthesis of halogenated compounds.[32,36] Identification of potential amino acids that are involved in the binding of the substrate requires a lot of knowledge about the enzyme, its structure and mechanism. As effective mutations not necessarily occur in the active pocket, an enormous working effort is needed to exchange all potential amino acids by site-directed mutagenesis. In addition, some beneficial mutations might be abolished by another mutation, so that it is nearly impossible to find the desired mutant with its specific characteristics by this rational approach. In order to counteract these problems, directed evolution, meaning the random introduction of mutations and the creation and subsequent selection of libraries with thousands of mutants, is a powerful tool to select those mutants with the desired features as discussed below. 1.6

Directed enzyme evolution

1.6.1

Principle of directed evolution

In order to generate an effective biocatalyst for industrial processes, genetic modifications such as protein engineering evolved as powerful tool for fine tuning and optimization of enzymes for chemical synthesis. Specific properties of interest can be modified leading to an increase in activity on non-native substrates, enhanced enantio- and product-selectivity and the overall stability of the enzyme. 21

An example for the successful application of evolved biocatalysts in industrial processes is the synthesis of sitagliptin, an antidiabetic compound. Savile et al. were able to replace a rhodium-based catalyst by a modified transaminase for asymmetric amination. By this, the efficiency of the manufacturing process could be improved significantly. After multiple steps of directed evolution an enzyme variant was obtained that showed 13 % increase in overall yield and 53 % increase in productivity (kg/L per day). Apart from the reduced amount of waste and manufacturing costs, toxic byproducts could be eliminated as well, what is an important feature in medical chemistry.[37] Even engineering of promiscuous enzyme functions to catalyze non-natural reactions can be enabled by means of directed evolution as demonstrated by Arnold et al. for numerous reactions like cyclopropanation or aziridination using engineered monooxygenases.[38] The process of directed evolution as a mimic of Darwinian evolution consists of three distinct parts (Figure 4). The first part is the generation of a random gene library, e.g. by error-prone PCR, DNA shuffling, iterative saturation mutagenesis or by using specialized mutator strains.[39] In the following step, the modified genes are transformed and expressed in a suitable expression host organism such as E. coli. The desired mutants are selected in a final high-throughput screening for specific properties. Preferentially, screening and selection are carried out using rapid and facile readouts based on absorption or fluorescence, whereas techniques based on chromatography mostly suffer from a lower throughput. The starting point for the creation of new biocatalysts is usually an existing enzyme as template. After the first round of directed evolution, enzymes that show the desired characteristics are used as parent for further rounds of directed evolution, leading to a stepwise optimization of the enzyme.[40,41]

Figure 4

22

General principle and strategy for directed enzyme evolution. One round of directed evolution consists of (a) the generation of a mutant library, (b) bacterial transformation and (c) a screening method selecting the mutants with the desired properties.

1.6.2

Directed evolution of RebH

Directed evolution of the FAD-dependent halogenase RebH was firstly performed by Poor et al. in 2014. After three rounds of mutagenesis a thermostable RebH variant (3LSR) was identified. The eight randomly introduced mutations resulted in an increased melting temperature of 18 °C compared to the wild type. In addition to the increase in conversion of tryptophan, improved conversion was also detected for unnatural substrates. Comparison of the kinetic parameter of the mutant and RebH WT revealed similar kcat/Km at 40 °C, however a significantly lower kcat/Km was identified for the mutant at 21 °C. Therefore, despite increased temperature tolerance and higher lifetime its kinetic parameter could not be improved. [42,43] 1.7

Halogenated marine metabolites

The reported number of over 2200 marine organobromine compounds demonstrates their high occurrence in the marine ecosystem. Marine organisms are considered as the main source for these compounds and in 2014 Agarwal et al. focused on the biosynthesis of polybrominated aromatic compounds.[8] Polybrominated diphenyl ethers (PBDEs) are especially characteristic for the marine environment, as they are produced by several plants, bacteria, fungi, algae, invertebrates and mammals and accumulate in the marine food chain.[8] As PBDEs consequently enter the human food chain as well, these compounds are potentially harmful because they are toxic and inhibit some mammalian signaling pathways and enzymatic reactions.[44] Agarwal et al. discovered a biosynthetic gene locus that is responsible for the synthesis of PBDE in Pseudoalteromonas spp. As the discovery of this gene locus led to the identification of the brominated marine pyrroles and phenols, the gene locus was abbreviated as bmp. It consists of the genes bmp 1-8, whereas each gene locus encodes for a different enzyme.[8] 1.8

FAD-dependent halogenase Bmp5

As the bmp5 gene locus was identified to encode for a FAD-dependent halogenase, this gene was moved into focus in context of this thesis. Apart from the well-known FADdependent tryptophan halogenases, Bmp5 does not require the addition of a flavin reductase in vitro, but is able to generate FADH2 in situ by oxidizing NADPH.[8] Regarding the substrate scope of Bmp5, 4-hydroxybenzoic acid (4-HBA, 19) is considered as a potential substrate, because it serves as precursor for the bromophenol moiety of several PBDEs such as pentabromopseudiline (4). The Bmp5-catalyzed conversion of 19 is proposed as a two-step mechanism as shown in the following scheme.

23

Scheme 5

Proposed two-step reaction mechanism of the Bmp5-catalyzed conversion of 19 to 23 via a stable intermediate (21). After electrophilic addition of Br+ in ortho position to the hydroxyl group of 19 and proton abstraction by a base (B), 21 is formed. The product 23 is formed after simultaneous decarboxylation and bromination of 21 in para position to the hydroxyl group.[8]

Due to the directing effects of the electron-donating hydroxyl group and the electronwithdrawing effect of the carboxyl group, the first step of the bmp5-catalyzed reaction is an electrophilic bromonium addition in ortho position to the hydroxyl group. The stable intermediate 3-bromo-4-hydroxybenzoic acid (21) is formed after proton abstraction by a base. In the following step 21 is brominated in para position to the phenol hydroxyl group and is subsequently decarboxylated, leading to the formation of 23. A second bromination of the intermediate 21 in ortho position to the hydroxyl group is proposed to be sterically hindered by its localization in the active pocket, therefore leading to decarboxylative bromination in para position. Bmp5's ability of halogenation without additional flavin reductase and the unexpected incorporation of bromide in presence of chloride in vivo differ considerably from the known FAD-dependent halogenases. In addition, Bmp5 is supposed to incorporate iodide. These findings may be explained by the fact that Bmp5 is more homologous to flavindependent oxygenases rather than to flavin-dependent halogenases.[8] Therefore, Bmp5 opens the way to new chemoenzymatic approaches for the synthesis of brominated compounds.

24

2

Motivation

FAD-dependent halogenases catalyze the regioselective halogenation of organic compounds under benign reaction conditions at room temperature and nearly pH 7 thereby merely requiring a halide salt and molecular oxygen. Their convenient handling is an enormous advantage to classical halogenation reagents that usually require hazardous reaction conditions. Biohalogenation is of particular interest for organic synthesis as it enables the preparative scale synthesis of reactive halogenated compounds with remarkable regioselectivity. Even though it was demonstrated that FAD-dependent tryptophan halogenases such as RebH are able to halogenate unnatural substrates and override directing electronic effects[33], the enzyme's limited substrate scope as well as its low activity and instability are factors that impede further applications in industrial processes. Therefore, the desired objective for this thesis is the modification and optimization of the tryptophan 6-halogenase Thal towards an improved thermostability, in order to obtain a catalyst with increased lifetime and activity at elevated temperatures. As previous results regarding an improved thermostability of RebH were achieved by several rounds of directed evolution[42], random mutagenesis of Thal in combination with a robust high-throughput fluorescence screening is considered as promising approach. Apart from the well-established tryptophan halogenases a marine brominase, namely Bmp5 from Pseudoalteromonas luteoviolacea, additionally turns into focus of this thesis. As Bmp5 is reported to catalyze the incorporation of bromide as well as iodide, it is of special interest for synthetic purposes, due to the fact that tryptophan halogenases are unable to accept iodide as substrate. After optimization of enzyme expression as well as examinations on purification strategies, Bmp5-catalyzed halogenations are planned to be performed on analytical as well as preparative scale. Here, the enzyme's applicability with immobilization methodologies such as combiCLEAs could be an interesting approach. Additionally, substrate screening revealing unnatural accepted compounds would be an important feature towards Bmp5's application for synthetic purposes. Finally, the establishment of a robust cofactor regeneration system, enabling continuous supply of necessary cofactors is another goal.

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_2

25

3

Material and Methods

3.1

Analytics

3.1.1

High performance liquid chromatography (HPLC)

HPLC is a versatile technique for the separation, purification and characterization of different compounds, e. g. macromolecules, such as peptides and proteins, but also amino acids, metabolites etc. can be examined. The principle of separation is based on the analyte's different adsorption to the stationary phase and its interaction with the mobile phase. In normal phase HPLC, a polar stationary phase generally consisting of silica media and a non-polar mobile phase are used, whereas RP-HPLC employs a non-polar stationary phase and elution is performed with polar solvents. Peptides and proteins are usually separated efficiently by RP-HPLC. The stationary phase of RP-HPLC consists of hydrophobic alkyl chains, such as C18 and therefore separates molecules according to their hydrophobic properties by elution with a gradient of a solvent mixture, such as H2O / MeCN. Consequently, molecules adsorbing with higher affinity to the hydrophobic surface have a longer retention time than polar molecules which elute earlier.[45] 3.1.1.1

Analytical RP-HPLC

An Accela-chromatography system from Thermo Scientific is used for analytical RPHPLC separations. Autosampler: Pump: Detector: Column:

Accela autosampler Accela 600 pump Accela PDA detector Hypersil Gold C18, 3 μm, 150 × 2.1 mm, Thermo Scientific

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_3

27

Composition of mobile phase Component H2O MeCN TFA

Eluent A / (v/v/v) 95 5 0.1

Eluent B / (v/v/v) 5 95 0.1

Elution method: Method A: linear gradient over 5 min 0 ՜ 100 % eluent B; flow rate: 0.7 mLήmin-1; 22 °C; detection: λ = 220, 254 and 280 nm Time / min 0 5.0 6.0 6.5 8.0 3.1.1.2

Eluent A / % 100 0 0 100 100

Eluent B / % 0 100 100 0 0

Preparative RP-HPLC

Preparative RP-HPLC separations are performed with a LaChrom HPLC from MerckHitachi. The composition of the elution solvents is identical to analytical RP-HPLC. Interface: Pump: Detector: Column:

D-7000 L-7150 L-7420 Hypersil Gold C18, 8 μm, 250 × 21.2 mm, Thermo-Scientific

Elution method Time / min 0 5 105

3.1.2

Eluent A / % 100 100 0

Eluent B / % 0 0 100

Electrospray ionization-mass spectrometry (ESI-MS)

ESI-MS measurements were performed with a Bruker Daltonics, Esquire 3000 ion trap mass spectrometer. Argon serves as cooling gas, whereas nitrogen is provided as nebulizer and dry gas, generated by NGM-11 (Bruker). The samples are injected using a syringe pump. Data acquisition is performed with the EsquireNT 5.2 software and further analysis is made with the DataAnalysis 3.4 software.

28

3.1.3

Matrix-Assisted Laser Desorption Ionization -Time of Flight Mass Spectrometry (MALDI-ToF MS)

To identify proteins by a characteristic peptide mass fingerprint using MALDI-ToF MS they are enzymatically digested into smaller peptide fragments with specific proteases. The protease trypsin cleaves at distinct positions of the protein C-terminal to the basic amino acids lysine and arginine leading to a peptide mixture.[46] These peptide segments can be detected and analyzed by mass spectrometry, e.g. MALDI-ToF MS that generates pseudo molecular ions [M+H]+ of the corresponding peptide segments. These masses are matched to the calculated masses of an in silico digestion enabling identification of the protein. The mass spectra are recorded with a Bruker UltrafleXtreme employing a MTP AnchorChip 384 TF target (Bruker). Therefore, 0.5 μL analyte are co-crystallized with 0.5 μL HCCA matrix (5 mg mL-1) on the target. The calibration is performed with a protein standard, thus 1 μL of standard is mixed with 50 μL HCCA matrix and 1 μL of the mixture spotted on the target. Table 1

Protein calibration standard for Maldi-ToF MS calibration

Component [M+H]+ Bradykinin (1-7) Angiotensin II Angiotensin I Substance P Bombesin Renin-substrate ACTH (adrenocorticotropic clip(1-17) ACTH-clip(18-39) Somatostatin(28)

3.1.4

Average m/z / Da 757.39916 1046.54180 1296.68480 1347.73540 1619.82230 1758.93261 hormone) 2093.08620 2465.19830 3147.47100

Gas chromatography-mass spectrometry (GC-MS)

GC-MS is a coupling of gas chromatography with EI-MS and represents a powerful analytical tool for the separation, identification and quantification of compounds in complex mixtures. A prerequisite is that the analyte is sufficiently volatile for evaporation without previous thermal degradation. The individual components are separated due to their different boiling points by using a temperature-controlled column. Components with a lower boiling point have a shorter retention time on the capillary column than less volatile components and get eluted by a carrier gas, e.g. He or N2. Afterwards, the compounds are guided to the mass spectrometer and ionized by EI. During ionization molecular ions are formed that can further degrade into fragment ions to enable compound identification.

29

The GC-MS analysis is performed with a TraceGCUltra gas chromatograph (Thermo scientific) and an ITQ900 ion trap mass spectrometer equipped with an AS2000 auto sampler (ThermoFinnigan, Dreieich, Germany). 1 μl of sample volume is splitlessly injected at 300 °C injector temperature. The gas chromatograph contains a 30 m × 0.25 mm VF-5 ms column coated with 0.25 μm of 5% diphenyl- and 95% dimethylsiloxane (Varian, Agilent Technologies, Deutschland GmbH, Darmstadt, Germany). The temperature of the interface is set to 250 °C, and the ion source to 220 °C. Helium is applied as carrier gas with a constant flow of 0.3 ml min−1. The oven temperature is heated to 80 °C for 3 min and is afterwards increased in steps of 5 °C/min to 325 °C. The mass spectra are recorded at 20 scans s−1 using a scanning range of m/z 50–750. 3.1.5

Nuclear magnetic resonance (NMR)

NMR spectra are recorded with a DRX-500 and AV500 (Bruker) and all measurements are conducted in deuterated solvents at 25 °C. The chemical shift ߜ is referenced to residual non-deuterated solvent signal (DMSO-d6: 1H: 2.49 ppm; 13C: 39.5 ppm). The data are edited with the TopSpin Software, version 2.1 (Bruker) and with Mestrenova, version 8.1, Mestrelab research. Resonance frequencies: 1H: 500 MHz; 13C: 126 MHz 3.2

Laboratory Equipment

Autoclave Vertical floor-standing autoclave 2540 ELV Camera system LAS-3000 Luminescent Image Analyzer French Press French® Pressure Cell Press Electrophoresis equipment Mini-Protean 3 Electrophoresis System Incubator Certomat IS incubator shaker Innova 4000 incubator shaker Heraeus B6 incubator Microplate reader Infinite M200 Millipore system Ultrapure water, Q-Gard® 3-Pack Nano-Drop ND-1000 Spectrophotometer Photometer UV-1280 pH-meter and -electrode MP 220; Inlab 420 30

Systec GmbH Fujifilm Thermo Electron Bio-RAD Sartorius Stedim Biotech GmbH New Brunswick Scientific (Eppendorf) Thermo-Scientific Tecan Merck-Millipore Peqlab Shimadzu Mettler Toledo

Shaker Mini-Rocker MR-1 Powering device Power-Supply EPS 601 Power Pac 200 Platform shakers Microplate shaker Titramax1000 Syringe pump Type 200 Thermocycler Primus 25 advanced Primus 96 advance (96-well-format) Thermomixer Thermomixer compact UV lamp Transilluminator T4 Vacuum concentrators Speedvac RVC 2-18 Alpha 2-4 LCS Vortexer Vortex-Genie 2 Balance XS-105 analysis scale BP 410S Centrifuge MiniSpin tabletop centrifuge Centrifuge 5810R Centrifuge rotor Swing-out rotor A-04-62 Fixed angle rotor F-34-06-38 Fixed angle rotor F-45-30-11

3.3

Biosan Amersham Biosciences Bio-RAD Heidolph KD Scientific Peqlab Peqlab Eppendorf Biometra-Whatman M. Christ Scientific Industries Mettler-Toledo Sartorius Eppendorf Eppendorf Eppendorf Eppendorf Eppendorf

Chemicals and solvents

All employed chemicals are obtained in highest quality (p.a.) from commercial suppliers Sigma-Aldrich, VWR, Carl Roth as well as Acros Organics. Organic solvents are distilled or used in p.a. quality. Water is purified with a Milli-Q water purification system and heat-sterilized for microbiological and protein-chemical work. 3.3.1

Special chemicals

acrylamide 4K solution (40%) ampicillin L-(+)-arabinose 3-bromo-4-hydroxybenzoic acid

AppliChem AppliChem Roth Sigma-Aldrich 31

chloramphenicol 2,4-dibromophenol FAD 4-hydroxybenzoic acid kanamycin sulfate NAD+ NADP+ NADPH dNTPs sodium tetrachloropalladate(II) SPhos water soluble 2,4,6-tribromophenol tryptone yeast extract 3.3.2

Purification kits

QIAprep Spin Miniprep Kit QIAquick PCR Purification QIAquick Gel Extraction Kit

3.3.3

Qiagen Qiagen Qiagen

Enzymes

Enzymes DNA polymerase GoTaq G2, 5 U/μL Pfu-DNA polymerase, 3 U/μL Phusion-DNA polymerase, 2 U/μL DNA ligase T4-DNA ligase, 3 U/μL Flavin reductase PrnF from P. fluorescens, 398 U/mL Alcohol dehydrogenase RR-ADH from Rhodococcus spp., 165 U/mL LK-ADH from Lactobacillus kefir, 26 U/mL Protease Trypsin, 0.1 mg/mL activation: 15 min, 30 °C Glucose dehydrogenase GDH from Bacillus megaterium, 100 U/mL

32

AppliChem Acros Organics AppliChem Sigma-Aldrich AppliChem Roth Roth Roth Promega Acros Organics Sigma-Aldrich Acros Organics Fluka Analytical AppliChem

Supplier Promega New England Biolabs New England Biolabs Promega provided by Christian Schnepel provided by Christian Schnepel Self production Promega

Provided by OCI, Bielefeld University

3.3.4

Restriction enzymes

Restriction enzyme NdeI, 20 U/μL BamHI, 20 U/μL NcoI, 20 U/μL NotI, 20 U/μL DpnI, 20 U/μL

3.3.5

Recognition sequence 5’-CA|TATG-3’ 3’-GTAT|AC-5’ 5'-G|GATCC-3' 3'-CCTAG|G-5' 5'-C|CATGG-3' 3'-GGTAC|C-5' 5'-GC|GGCCGC-3' 3'-CGCCGG|CG-5' 5'-GAMe|TC-3' 3'-CT|AMe G-5'

Supplier New England Biolabs New England Biolabs New England Biolabs New England Biolabs New England Biolabs

Plasmids

Name

Description

pET21a

Cloning and expression C-His6 vector with multiple cloning site Plasmid for expression of Not exLK-ADH pressed

amp

Cloning and expression vector with multiple cloning site Plasmid for expression of Thal, thal was inserted between NdeI/BamHI sites Plasmid for expression of mutant Thal-E2, thal-E2 was inserted between NdeI/BamHI sites Plasmid for expression of mutant Thal-E2R, thal-E2R was inserted between NdeI/BamHI sites

N-His6 C-His6

kan

N-His6

kan

≈7000 Selfproduction

N-His6

kan

≈7000 Selfproduction

N-His6

kan

≈7000 Selfproduction

pET21aLK-ADH

pET28a

pET28athal

pET28athal-E2

pET28athal-E2R

Tag

Resistance

amp

Size / bp 5443

Supplier Novagen

≈ 6200 Provided by OC I, Bielefeld University 5369 Novagen

33

pET28abmp5

Plasmid for the expression of Bmp5, bmp5 was inserted between NdeI/BamHI sites Cloning and expression vector with multiple cloning site

pETM30

pETM30bmp5

pGro7

3.3.6

N-His6

kan

≈7000 Selfproduction

N-His6 N-GST

kan

Plasmid for expression of N-His6 Bmp5, N-GST bmp5 was inserted between NcoI/NotI sites Plasmid for the none expression of chaperones GroEL, GroES, Induction via araB with L-arabinose

kan

6346 Provided by BC IV, Bielefeld University ≈8000 Selfproduction

5400 TakaraClontech

Bacterial strains

Strain DH5α BL21 (DE3) BL21 CodonPlus (DE3)-RIL 3.3.7

cam

Genotype F– endA1 recA1 ΔlacU 169 Ψ80 dlacZ ΔM15 F– ompT hsdSB (rB–mB–) gal dcm (DE3) [lacI lacUV5-T7 gene 1 ind1 sam7 nin5]) E. coli B F– ompT hsdS(rB– mB–) dcm+ Tetr gal λ(DE3) endA Hte [argU ileY leuW Camr]

Resistance none none cam

Supplier Novagen Novagen Novagen

Primers

The employed primers are synthesized by Eurofins Genomics and obtained as salt-free lyophilisate. The underlined sequence represents restriction sites. Sequences encoding a gene of interest are printed in bold type. Name

Nucleotide sequence and description

pET28aNdeI-FW

5’-GGCCTGGTGCCGCGCGGCAGCCATATG-3' Forward primer for epPCR with pET28a as template

pET28aBamHI-StopREV ThalE2K510-FW

5'-GTCGACGGAGCTCGAATTCGGATCCTTA-3' Reverse primer for epPCR with pET28a as template

34

5'-GTTTGCAGATGTTAAACGTAAAGGTGATACCCTGGTTG-3' Forward primer to reverse silent mutation of Thal-E2 at amino acid 510

corresponding to the WT sequence ThalE2K510-REV

5'-CAACCAGGGTATCACCTTTACGTTTAACATCTGCAAAC-3' Reverse primer to reverse silent mutation of Thal-E2 at amino acid 510 corresponding to the WT sequence

NcoI-bmp5FW

5'-GCCGCGCCATGGGAAACAAAACCATTGCCGTTATTGG-3' Forward primer for subcloning of bmp5 from pET28a into pETM30 with NcoI cleavage site and Gly insertion at position 2.

NotI-bmp5REV

5'-ATCGCCGCGGCCGCTTATTTGCTAATTTCACGCAGG-3' Reverse primer for subcloning of bmp5 from pET28a into pETM30 with NotI restriction site

T7P

5’-TAATACGACTCACTATAGGG-3’ T7 promotor primer for sequencing of pET vector derivatives

T7T

5’-CTAGTTATTGCTCAGCGG-3’ T7-terminator primer for sequencing of pET vector derivatives

3.3.8

Culture media

All listed media are heat-sterilized prior use. Medium Luria-Bertani medium (LB)

Composition 10 g L-1 tryptone 10 g L-1 NaCl 2.5 g L-1 yeast extract pH 7.0

LB agar

10 g L-1 tryptone 10 g L-1 NaCl 2.5 g L-1 yeast extract 10 g L-1 agar pH 7.0

SOC-medium

2 % (w/v) tryptone 0.5 % (w/v) yeast extract 10 mM NaCl 2.5 mM KCl 10 mM MgCl2 10 mM MgSO4 20 mM D-glucose pH 7.0

35

3.3.9

Antibiotics

The antibiotics are employed in a thousand-fold dilution of the listed stock solution. Stock solution Ampicillin Chloramphenicol in EtOH Kanamycin

3.3.10

Buffer

Enzyme buffer CutSmart buffer 5x GoTaq buffer 10x Pfu buffer 5x Phusion HF or GC buffer Size standards 1 kb DNA ladder 100 bp DNA ladder Color Plus prestained protein marker (11-245 kDa) 3.3.11

Supplier New England Biolabs Promega Promega New England Biolabs New England Biolabs New England Biolabs New England Biolabs

Gel electrophoresis

Agarose gel electrophoresis 10x DNA loading buffer

Ethidium bromide stock solution 50x TAE buffer

SDS-PAGE 10x PAGE buffer

Coomassie staining solution

Destaining solution 3x SDS loading dye + β-mercaptoethanol

36

Concentration / mg∙mL-1 100 50 60

Composition 70 % (w/v) sucrose 100 mM EDTA 0.1 % bromophenol blue 1 mg∙mL-1 ethidium bromide 2 M Tris 2 M acetic acid 50 mM EDTA 192 mM glycine 250 mM Tris 1 % (w/v) sodium dodecyl sulfate 0.02 % Coomassie Brilliant Blue 5 % (w/v) aluminium sulfate-18-hydrate 10 % (v/v) ethanol 2 % (v/v) phosphoric acid 10 % (v/v) ethanol 2 % (v/v) phosphoric acid 187.5 mM Tris-HCl pH 6.8 6 % SDS 30 %glycerin

Stacking gel buffer Separation gel buffer Stacking gel 4 %

Separation gel 12 %

3.3.12

Western blot analysis

Buffer 10x PBS

PBS-T 10x semi dry blot buffer

1x semi dry blot buffer blocking buffer

3.3.13

15 % β-mercaptoethanol 0.003 % bromophenol blue 0.5 M Tris pH 6.8 1.5 M Tris pH 8.8 2.59 mL MPW 1.26 mL Tris 0.5 M pH 6.8 50 μL10 % SDS 0.5 mL Aa/Bis (40 %) (386 g L-1 acrylamide/13 g L-1 bisacrylamide) bisacrylamide 5 μL TEMED 0.3 mL 1.5 % APS 3.6 mL MPW 2.5 mL Tris 1.5 M pH 8.8 100 μL 10 % SDS 3 mL Aa/Bis (40 %) (386 g L-1 acrylamide/13 g L-1 bisacrylamide) 5 μLTEMED 0.8 mL 1.5 % APS

Composition 1.5 M NaCl 0.2 M Na2HPO4 50 mM NaH2PO4 pH 7.4 1x PBS 0.1 % Tween 50 mM Tris base 40 mM glycine 10 % SDS pH 9.2 10 % (v/v) 10x semi dry blot buffer 25 % (v/v) MeOH 2 % (w/v) milk powder in PBS-T

Tryptic digestion

Solution Washing solution

Composition 30 % (v/v) MeCN 0.1 M NH4HCO3 37

Trypsin buffer Trypsin solution per reaction Peptide extraction solution

3.3.14

10 mM NH4HCO3 1 μL activated trypsin 14 μL trypsin buffer 50 % (v/v) MeCN 0.1 % (v/v) TFA

Protein expression and purification

Buffer IPTG-stock solution PMSF-stock solution

Composition 0.01 M IPTG 50 mM phenylmethylsulfonyl fluoride in iso-PrOH

HisTALON affinity chromatography Equilibration buffer

Washing buffer

Elution buffer

Glutathione S-Transferase (GST) affinity chromatography Equilibration buffer

Washing buffer

Elution buffer

3.3.15

50 mM Na2HPO4 300 mM NaCl 300 mM imidazole pH 7.4

50 mM Na2HPO4 300 mM NaCl pH 8.0 50 mM Na2HPO4 300 mM NaCl pH 8.0 50 mM Na2HPO4 300 mM NaCl 10 mM reduced glutathione pH 8.0

Enzymatic halogenation

Component L-tryptophan-stock solution FAD-stock solution NAD+-stock solution 38

50 mM Na2HPO4 300 mM NaCl pH 7.4 50 mM Na2HPO4 300 mM NaCl 10 mM imidazole pH 7.4

Composition 25 mM L-tryptophan 1 mM FAD 50 mM NAD+

NADP+-stock solution 10x bromination buffer

10x iodination buffer

50 mM NADP+ 100 mM Na2HPO4 300 mM NaBr pH 7.4 (adjusted with H3PO4) 100 mM Na2HPO4 300 mM NaI pH 7.4 (adjusted with H3PO4)

3.4

Molecular biological methods

3.4.1

Cultivation of E. coli

E. coli cells are cultivated in Luria-Bertani medium, containing the appropriate antibiotics, at 37 °C. Liquid cultures are grown in a shaking incubator at 150 rpm, whereas cells streaked on LB agar are stored in an incubator. 3.4.2

Preparation of E. coli glycerol stocks

Glycerol stocks are prepared for long-term storage of plasmids in a desired bacterial strain which can be directly used for inoculation of main cultures. Therefore, a single colony is picked from a freshly streaked agar plate and cultivated overnight in liquid medium containing the appropriate antibiotics. 1.5 mL of the overnight culture are centrifuged (5000 × g, 5 min) and 1 mL of supernatant is discarded. The cell pellet is resuspended in the remaining supernatant and 500 μL glycerin (86 %) are added. The bacterial solution is mixed carefully, flash-frozen in liquid nitrogen and stored at -80 °C. 3.4.3

Determination of bacterial titer

The growth of bacteria is monitored photometrically by measuring the optical density of the culture medium. The absorbance of the sample is measured at a wavelength of 600 nm, whereas the medium without bacteria serves as reference. 3.4.4

Isolation of plasmid DNA

In order to isolate plasmids of E. coli DH5α or E. coli BL21 (DE3), 10 mL of overnight culture are centrifuged (15 min, 10000 × g, 4 °C) and the plasmid isolation is performed according to the QIAprep Spin Miniprep Kit protocol from Qiagen. The cells are initially lysed by alkaline lysis and genomic DNA is precipitated due to subsequent neutralization, whereas the circular plasmid DNA remains in solution. In the following step the plasmid DNA adsorbs on a silica membrane in presence of a chaotropic salt via anion exchange chromatography. After removal of impurities by several washing steps the plasmid DNA is eluted. 3.4.5

Determination of DNA concentration

The DNA concentration is measured spectrometrically by the use of a Nanodrop ®spectrometer from Peqlab. 1-2 μL of the undiluted sample are required and the DNA 39

concentration is measured at a wavelength of 260 nm. The ratio of the absorbance at 260 nm and 280 nm indicates the purity of the DNA and should ideally have a value between 1.8 and 2.0. 3.4.6

Polymerase chain reaction

The polymerase chain reaction (PCR) is an in vitro amplification method of a specific DNA sequence. As a first step the DNA double strand is denatured into single strands at 94-98 °C. In the following step the temperature is decreased, so that the primers can anneal to their complementary sequences on the target DNA. The temperature for primer annealing depends on the length and GC content of the applied primers and is often performed between 45 and 70 °C. Subsequent to this step the DNA polymerase binds to the primer and extends the DNA strand by adding nucleotides and therefore synthesizing a new DNA molecule complementary to the target sequence. The optimal working temperature of the thermostable DNA polymerase depends on the enzyme employed, but is often chosen around 72 °C. Generally, high-fidelity polymerases are employed equipped with a proofreading activity to avoid nucleotide misincorporation that might lead to undesired mutations. After the duplication of the DNA strand the first PCR cycle is completed. The process of denaturation, primer annealing and synthesis of new DNA is repeated for 25 to 30 times, so that the amount of specific DNA, neighboured by two primers, is duplicated exponentially after the second cycle of amplification. Pipetting scheme for a standard PCR with Phusion-DNA polymerase: Component 5x Phusion HF or GC buffer dNTPs, 10 mM FW Primer, 10 μM RV Primer, 10 μM Template DNA Phusion-DNA Polymerase (2 U/μL) Nuclease-free water

Final concentration 0.2 mM 0.5 μM 0.5 μM 100 ng 0.02 U/μL Total volume

Volume / μL 10 1 2.5 2.5 variable 0.5 Fill up 50 μL

Cycle conditions for a routine PCR with Phusion-DNA polymerase: Cycle Step Initial Denaturation Denaturation Annealing Extension Final Extension Storage

40

Temperature / °C 98 98 55 72 72 4

Time 2 min 20 s 20 s 30 s / kb 5 min

Cycles 1 30 30 30 1 ∞

3.4.6.1

Colony-PCR

Colony-PCR is a method to determine the existence of insert DNA in a plasmid construct by direct screening of an E. coli colony grown on LB agar. Therefore, several clones are picked and transferred separately into 30 μl H2O. The cells are lysed to release DNA by incubation at 95 °C for 20 min. Afterwards the colony-PCR is performed according to standard PCR procedure using 1 μL of lysate containing the DNA template. As the 3´→ 5´ exonuclease activity of the Phusion DNA polymerase is not required for the analytical purpose of a colony-PCR, GoTaq-DNA polymerase without proof reading activity is employed. The presence or absence of an insert is proven by separating the PCR samples via gel electrophoresis. Positive transformants carrying the desired insert result in a specific DNA band with appropriate size. To prove success of PCR a template specific for the employed primers is prepared in parallel as positive control. 3.4.6.2

Site-directed mutagenesis PCR

Site-directed mutagenesis PCR serves to introduce or eliminate specific mutations in a template DNA. Therefore, primers are employed that preferentially contain the desired mutation in the middle of their sequence, flanked by complementary regions. After incorporating the mutation in to the daughter strands by linear amplification of the complete plasmid due to self-complementary primers, the methylated E. coli template DNA is digested using DpnI. This restriction enzyme recognizes only methylated nucleotides and does not digest the mutated PCR product. Pipetting scheme for site-directed mutagenesis PCR with Pfu-DNA polymerase: Final concentration

Volume / μL 5 0.2 μM 1 0.2 μM 1 1-2 ng/μL 1 0.2 mM 1 0.06 U/μL 1 40 Total volume 50 μL Cycle conditions for site-directed mutagenesis PCR with Pfu-DNA polymerase: 10x Pfu reaction buffer FW primer, 10 μM RV primer, 10 μM Template-DNA, 50-100 ng/μL dNTPs, 10 mM Pfu-DNA polymerase 3 U/μL H2O

Cycle Step Initial Denaturation Denaturation Annealing Extension Final Extension Hold

Temperature / °C 95 95 55 68 68 4

Time 2 min 30 s 1 min 1 min / kb 5 min

Cycles 1 15 15 15 1 ∞

41

3.4.6.3

PCR purification

PCR purification is performed according to the QIAquick Purification Kit. The DNA binds to the silica membrane of the column at pH below 7.5 and high salt conditions. After impurities such as primers, nucleotides and enzymes are removed by washing the DNA is eluted with a low-salt buffer. 3.4.7

Agarose gel electrophoresis

Agarose gel electrophoresis is a well-established technique to separate DNA in an electric field according to its size. Due to DNA's negatively charged phosphate backbone these molecules migrate towards the anode, whereas smaller fragments move faster than bigger ones. In order to determine the size of the separated DNA fragments, a DNA marker runs in parallel as reference that separates into fragments of known size. To prepare an agarose gel, 1 % (w/v) of agarose is dissolved in TAE buffer by heating in a microwave oven. The warm solution is poured into a gel chamber equipped with an appropriate comb to form sample wells. The hardened gel is covered with TAE buffer and is ready for sample loading. The DNA samples are mixed with 6x loading dye, applied to the sample wells and separated for 45-60 min using a voltage of 150 V. In order to visualize the DNA fragments, the gel is stained in an ethidium bromide bath for 10 min and destained in water to remove the unspecific background. The DNA bands are detected by fluorescence of the intercalating dye using an ultraviolet transilluminator. Finally, the gel is photographed for documentation. 3.4.8

DNA extraction from agarose gel

The extraction of DNA bands from an agarose gel is performed according to the QIAquick Gel Extraction Kit from Qiagen. Therefore, the relevant DNA band is cut out of the gel, resolubilized in a suitable buffer and the mixture applied to the provided silica column. The DNA adsorbs to the membrane at high-salt conditions and washing steps are performed to remove impurities. The DNA is finally eluted in a low-salt buffer. 3.4.9

Restriction digest of plasmid DNA

Restriction enzymes recognize and cleave DNA within specific sequences and create either sticky or blunt ends, as they hydrolyze the DNA's phosphodiester bonds. In order to generate a recombinant vector, insert and vector are digested in parallel employing the same restriction enzymes. This methodology facilitates to subclone genes of interest into an expression vector for heterologous protein expression. The restriction digest is performed by mixing 1 μg DNA with 20 U of restriction enzyme as well as the appropriate enzyme buffer in 50 μL total volume. After incubation at 37 °C for 2 h additional 20 U of restriction enzyme are added and incubated for another hour. To receive the digested DNA fragments, the reaction mixture is purified by gel extraction or Qiagen PCR purification Kit.

42

3.4.10

Ligation of DNA fragments

DNA ligation aims to generate a recombinant vector by linking insert-DNA with an analogously digested vector. Commonly both components are ligated by a T4-DNA ligase. This enzyme catalyzes the formation of the phosphodiester bond between the free 3’-hydroxy and 5’-phospho termini and can be used for ligation of sticky as well as blunt ends. The ligation is performed in an appropriate T4-DNA ligase buffer with a molar excess of insert in a 3:1 ratio. The ligation reaction mixture is set up according to the representative scheme below and incubated at 4 °C in a water bath overnight. pET28a NdeI/BamHI (5.3 kb) Thal insert (NdeI/BamHI) (1.6 kb) 10x T4-ligase buffer (contains ATP) T4-DNA ligase (3 U/μL) H2O Final volume 3.4.11

100 ng 90 ng 2 μL 1 μL (0.15 U/μL) Fill up 20 μL

Preparation of chemocompetent E. coli cells

Chemocompetent E. coli cells are prepared using the CaCl2 method. Therefore, 150 mL LB medium is inoculated with 10 mL overnight culture of an appropriate E. coli strain. The bacteria are incubated at 37 °C in a shaker until an OD600= 0.5-0.6 is reached. Afterwards the culture is chilled on ice for 10 min and the bacteria are harvested by centrifugation (3220 × g, 4 °C, 5 min). The supernatant is discarded and the pellet resuspended in 20 mL ice-cold sterile 100 mM MgCl2 solution. After incubation on ice for 30 min the culture is centrifuged again (3220 × g, 4 °C, 5 min). The supernatant is discarded and the pellet resuspended in 10 mL ice-cold sterile 100 mM CaCl2 solution and 2 mL glycerin. The cell suspension is aliquoted to 100 μL per tube, flash-frozen in liquid nitrogen and stored at -80 °C. 3.4.12

Heat-shock transformation

The heat-shock transformation method is used to transform chemocompetent E. coli cells with DNA. Therefore, competent E. coli cells (50 μL), stored at -80 °C, are thawed on ice and 1 μL of DNA is added. The suspension is mixed gently and chilled on ice for 30 min to allow the DNA to settle down. Afterwards, the cells are heat-shocked for 2 min at 42 °C leading to the formation of small pores in the cell membrane so that DNA is able to enter the bacterial cell. Subsequently, the cells are placed on ice and 1 mL SOCmedium is added. For regeneration and development of the antibiotic resistance cells are cultivated at 37 °C for 1 h and finally spread onto preheated agar plates containing the appropriate antibiotics as selection markers. 3.4.13

DNA sequencing

The DNA sequence of isolated DNA is checked by Sanger sequencing, using T7P and T7T as primers. The sequencing is performed by GATC Biotech Company, Konstanz / Köln. 43

3.5

Biochemical methods

3.5.1

Protein expression in E. coli

3.5.1.1

Expression of Thal and its mutants

For the overexpression of Thal and its mutants, the corresponding plasmid is transformed into E. coli BL21 (DE3) pGro7. Afterwards, a single colony is picked from the agar plate and cultivated overnight in liquid medium containing the appropriate antibiotics kanamycin (60 μgήmL-1) and chloramphenicol (50 μgήmL-1). To grow the main culture 3 L LB medium supplemented with the mentioned antibiotics are inoculated with 20 mLήL-1 of an overnight culture. The expression culture is incubated at 37° C until an OD600 of 0.6 is reached. Afterwards the temperature is decreased to 25 °C for 30 min, allowing the bacteria to adapt to the lower temperature. Co-expression of Thal as well as the chaperones GroEL / ES is induced by addition of IPTG (0.1 mM) and L-arabinose (2 gήL-1). For protein synthesis the expression culture is incubated at 25° C for 20 h. The overexpression of Thal mutants is additionally performed in the E. coli strain BL21 CodonPlus(DE3)-RIL that contains extra copies of the argU, ileY, and leuW tRNA genes. The encoded tRNAs recognize the arginine codons AGA and AGG, the isoleucine codon AUA, and the leucine codon CUA. Protein overexpression is performed as described above, expect that no induction with L-arabinose is required due to lack of pGro7. 3.5.1.2

Expression of Bmp5

The overexpression of Bmp5 in E. coli BL21 (DE3) pGro7 is performed according to the expression protocol of Thal as described above. The overexpression of Bmp5 in E. coli BL21 (DE3) is conducted similarly, except that no chloramphenicol as selection marker and L-arabinose for induction are required due to the lack of the pGro7 plasmid. 3.5.1.3

Expression of LK-ADH

The overexpression of LK-ADH in E. coli BL21 (DE3) pET21-LKADH is performed in the same way as for Thal and Bmp5, except that the LK-ADH expression is induced with 0.5 mM IPTG. 3.5.1.4

Cell harvesting

After protein overexpression in E. coli the cells are harvested by centrifugation (30 min, 3220 × g, 4 °C). The remaining pellet is resuspended in 40 mL Na2HPO4 (0.1 M, pH 7.4) and centrifuged again (30 min, 10000 × g, 4 °C). The supernatant is discarded and the pellet is flash frozen in liquid nitrogen and stored at -20 °C. 3.5.1.5

Cell disruption by French Press

The French Press is a mechanical method for cell disruption, especially suitable for large scale cell breakdown. By forcing the cell suspension through a small slit with high pressure, shear forces are applied which lead together with a sudden pressure difference to the burst of the cell wall. The E. coli cell lysate is prepared by resuspending the 44

bacteria of 1.5 L expression culture in 30 mL equilibration buffer (50 mM Na2HPO4 pH 7.4, 300 mM NaCl) as well as 50 μM PMSF for protease inhibition. The sample is lysed three times by French Press under a pressure of 70 bar. A spatula tip of DNaseI is added, the cell debris removed by centrifugation (10000 × g, 30 min, 4 °C) and the resulting supernatant sterile filtered through a 0.45 μM Whatman filter using a water-jet vacuum pump. 3.5.1.6

Determination of LK-ADH activity

The ADH activity of the diluted E. coli BL21 (DE3) pET21-LKADH crude lysate (10 μL, 1:100 dilution) is measured in a fivefold determination in a final volume of 1 mL, containing 250 μM NADP+, 10 mM Na2HPO4 pH 7.4 and 20% (v/v) iso-propanol. As reference serves 1 mL of 10 mM Na2HPO4, pH 7.4 containing 20% (v/v) iso-propanol. The conversion of NADP+ to NADPH is monitored photometrically by detecting the increase of absorbance at 340 nm due to the formation of NADPH. The conversion rate is determined by regression of the linear range within 15 s after reaction start. The ADH activity is calculated with the given equation: ୼୉ሺଷସ଴୬୫ሻ

Units mL-1= ୲ήሺ୚

ౝ ή୊ሻȀሺகή୚౦ ήୢሻ

Vg F ε Vp d

total volume (1 mL); dilution factor (100) extinction coefficient (6.3 mL μmol-1 cm-1) enzyme volume light pathway (1 cm)

3.5.2

Protein purification via immobilized metal ion affinity chromatography

The purification of N- or C-terminal His6-tagged proteins is performed by immobilized metal ion affinity chromatography (IMAC) using a HisTALON column made up of immobilized Co2+ ions from Takara clonetech. The column material consists of a tetradentate chelator that is loaded with Co2+ ions having a high affinity and specificity for His6-tagged fusion proteins. Proteins or contaminations lacking this tag do not interact with the matrix and are removed during washing steps. The elution of the His6-tagged protein is performed in presence of a high imidazole concentration that displaces the bound protein from the matrix. The column, containing a bed volume of 1.5 mL resin, is equilibrated at room temperature stepwise with 10 mL MilliQ water and equilibration buffer at a flow rate of 1 mLήmin-1. All following steps of enzyme purification are performed at 4 °C. The sterile filtered cell lysate is loaded to the column with a flow rate of 0.5 mLήmin-1. Proteins that bind non-specifically to the Co2+ matrix are removed during two washing steps (10 mL, 50 mM NaH2PO4 pH 7.4, 300 mM NaCl, flow rate: 1 mLήmin-1) whereas the second washing buffer contains additional 10 mM imidazole. All fractions are collected for further analysis by SDS-PAGE. 45

The His6-fusion protein is eluted by addition of elution buffer (300 mM NaCl, 300 mM imidazole, 50 mM Na2HPO4 pH 7.4 (flow rate: 1 mLήmin-1), where 10 to 15 fractions of 0.5 mL each are collected. The protein concentration is roughly estimated with a rapid Bradford test and afterwards quantitatively determined by Nano-Drop UV spectroscopy. 3.5.3

Protein purification via Glutathione S-Transferase (GST) affinity chromatography

The purification of N- or C-terminal GST-tagged proteins is based on the affinity of the GST-fusion protein to its substrate glutathione which is linked via a sulfur atom to the sepharose matrix. This interaction is reversible and the tagged protein can be eluted by addition of reduced glutathione. The affinity of GST for glutathione is much higher than for immobilized substrate, thereby it enables to remove the GST-tagged protein from the matrix. The purified enzyme is dialyzed against 50 mM Na2HPO4, 300 mM NaCl pH 8.0 (5 L reservoir) overnight to remove remaining glutathione. The employed resin (Protino® Glutathione Agarose 4B) is purchased from Macherey-Nagel, Düren, Germany.[47] The workflow for the purification of GST-tagged proteins is performed analogously to the described purification of His6-tagged proteins, except from the employed matrix and buffer solutions (see 3.3.14). 3.5.4

Determination of protein concentration

The Bradford protein assay is a technique to determine the total protein concentration of a sample compared to a protein standard. Interactions of the Bradford’s dye Brilliant Blue with basic amino acid residues of the protein result in a shift of its absorption maximum and a color change from red to blue in proportion to the amount of protein present in the sample. The absorption can be measured in a 96-well plate and the concentration is calculated with help of a BSA calibration series. In order to roughly estimate the present amount of protein e.g. of the elution fractions after protein purification, a rapid Bradford test is performed. Therefore, 200 μL of Bradford reagent are mixed with 2 μL of protein sample and the intensity of the color change represents the amount of protein per fraction. For quantitative determinations a Nano-Drop spectrometer is employed by absorption measurement at 280 nm. 3.5.5

SDS polyacrylamide gel electrophoresis

SDS-PAGE (sodium dodecyl sulfate polyacrylamide gel electrophoresis) directs to separate protein samples according to their molecular weight in an electric field. The proteins to be analyzed obtain a negative charge by the addition of the anionic detergent SDS overcompensating the inherent protein charge. The tertiary structure is destructed by the reducing agent β-mercaptoethanol that causes cleavage of inter- and intramolecular disulfide bridges. As response to the electric field the unfolded, negatively charged proteins migrate to the positively charged anode, whereas smaller proteins migrate faster through the gel’s pore structure than larger ones. The SDS gel consists of two different types of gels: the large-pore stacking gel on top and the small-pore separating gel underneath. Due to their different pH-value and pore size, the proteins are initially 46

concentrated in the stacking gel at pH 6.8 and afterwards separated according to their size in the separating gel at pH 8.8. The samples are mixed with 3x SDS loading dye, heated at 95 °C for 10 min and loaded to the sample wells. A voltage of 100 V is applied for approximately 15 min and is increased to 180 V when the dye front reaches the separating gel. The electrophoresis is stopped as soon as the dye front reaches the bottom of the gel. Afterwards the gel is washed with Millipore water, protein bands are detected by Coomassie staining solution and destained using destaining solution overnight. 3.5.6

Tryptic digest of proteins

The identification of proteins can be accomplished by tryptic digestion in combination with peptide mass fingerprint analysis. Therefore, Eppendorf tubes are washed with 0.5 mL 60 % MeCN and 0.1 % TFA and dried overnight. This step helps to remove contaminating plasticizers that impede detection of the peptides via mass spectrometry. The Coomassie stained protein band is cut out of the polyacrylamide gel into small pieces and is transferred to an Eppendorf tube. To extract the Coomassie dye the gel slice is washed several times with 200 μL of washing solution by incubation in a shaking incubator for 10 min and subsequent discarding of the supernatant. This procedure is repeated until the gel slices become colorless and residual solvent is removed in vacuum to complete dryness. Afterwards, gel pieces containing the protein are digested by rehydration in 15 μL trypsin buffer containing 1 μL of protease and are incubated at 37 °C for 30 min. The evaporated solvent is refilled to 15 μL and finally the mixture is incubated at 37° C overnight. The gel slices are dried in vacuum for one hour. After addition of 15 μL peptide extraction solution (50% MeCN/0.1% TFA) the samples are analyzed by MALDI-ToF-MS on a Bruker UltrafleXtreme to give the peptide mass fingerprint. 3.5.7

Western blot analysis

Western blot is an important immunobiological technique to separate and subsequently identify a specific protein from complex mixtures such as a cell lysate. For Western blot analysis proteins are initially separated by SDS-PAGE and then transferred to a nitrocellulose membrane making them accessible for antibody detection. 3.5.7.1

Protein transfer

For protein separation of the sample SDS-PAGE is performed as previously described. The proteins bands are transferred from the polyacrylamide gel to the nitrocellulose membrane by semi-dry transfer, sandwiching the gel and membrane between two stacks of Whatman paper, allowing direct contact with the plate electrodes. Therefore, three layers of Whatman paper, trimmed to SDS-gel size and soaked in semi-dry blotting buffer, are placed on the anode. The nitrocellulose membrane is moistened with semi-dry blotting buffer, placed on top of the stack and carefully covered with the polyacrylamide gel. Afterwards, three additional layers of soaked Whatman paper are placed on the gel, the cathode is positioned on the stack and the protein transfer performed at 1 mA per cm2 47

for one hour. The transfer efficiency is checked by covering the nitrocellulose membrane with Ponceau S dye (Sigma-Aldrich) and shaking it for 5 min. 3.5.7.2

Immunodetection of proteins

In order to avoid non-specific binding of the primary antibody, the membrane is blocked with 10 mL blocking buffer containing milk powder for one hour at RT. Subsequently, the membrane is incubated with the primary antibody overnight at 4 °C. Non-bound antibody is removed by washing the membrane three times with 10 mL PBS-T for 10 minutes. The membrane is incubated for 40 minutes with the secondary HRPconjugated antibody and afterwards washed twice for 10 minutes with PBS-T as well as once with PBS. For imaging and signal development the instructions of the SuperSignal® West Pico Chemiluminescent Substrate kit (Thermo Scientific) are followed. After excess reagent is removed from the membrane it is placed into a plastic wrap and ready for signal development. Antibody (ab) Primary ab Secondary ab

Hexa-His α-mouse-HRP

Species mouse goat

3.6

Directed evolution

3.6.1

Directed evolution of Thal

3.6.1.1

EpPCR for generation of mutant libraries

Dilution 1:3000 1:10000

Supplier GE Healthcare Jackson ImmunoResearch

EpPCR is an important tool for directed evolution in order to introduce random mutations into a gene of interest. The performance of the epPCR is based on recommendations provided by the Genemorph II Random Mutagenesis kit from Agilent technologies. The low fidelity of the GoTaq-DNA polymerase results in an imprecise amplification and the introduction of nucleotide mismatches. The error rate can be further increased by the addition of Mn2+, the increase of Mg2+ or unbalanced dNTP concentrations. Mn2+ ions are assumed to replace the essential cofactor Mg2+ in the active site of the polymerase, thereby favoring misincorporation. The desired number of mutations is controlled by the initial amount of template DNA and the number of amplification cycles. A higher mutation frequency is achieved by lowering the DNA template concentration, because the probability rises that primers bind to incorrect duplicates instead of template DNA. Therefore, more errors can accumulate during every amplification cycle. Consequently, the mutation rate is decreased by increasing the amount of template DNA or lowering the amount of PCR cycles.

48

Pipetting scheme for epPCR with GoTaq-DNA polymerase Final concentration 5x-GoTaq- reaction buffer FW primer, 10 μM RV primer, 10 μM Template-DNA dNTPs, 10 mM GoTaq-DNA polymerase 5 U/μL MgCl2(1 mM) H2O

0.2 μM 0.2 μM 2-10 ng/uL 0.2 mM 0.05 U/μL 150-200 μM Final volume

Volume / μL 10 1 1 Variable 1 0.5 7.5-10 Fill up 50 μL

Cycle conditions for epPCR with GoTaq-DNA polymerase Cycle Step Initial Denaturation Denaturation Annealing Extension Final Extension Storage 3.6.1.2

Temperature / °C 95 95 55 72 72 4

Time 2 min 30 s 45 s 1 min / kb 5 min

Cycles 1 25 25 25 1 ∞

Subcloning and transformation in expression host organism

A mutant library of thal is created by epPCR as described above and the general success of the PCR is checked by agarose gel electrophoresis. The purified PCR samples are digested with the restriction enzymes NdeI and BamHI, in order to create the required sticky ends for subcloning into an expression vector. The methylated E. coli template DNA is removed by DpnI digestion. After ligation of the digested Thal insert with an analogously restricted pET28a vector, the ligation mixture is transformed via heat-shock into E. coli BL21 (DE3) pGro7. Plasmid DNA of representative transformants is sequenced to determine the mutation frequency. 3.6.1.3

Preparation of E. coli BL21 (DE3) pGro7 pET28a-thal master plates

Mutated E. coli BL21 (DE3) pGro7 pET28a-thal colonies are picked from the agar plate and transferred to a 96-well plate containing 200 μL LB medium supplemented with the appropriate antibiotics kanamycin (60 μg∙mL.1) and chloramphenicol (50 μg∙mL.1) per well. The comparison between the mutant and WT activity is achieved by providing Thal WT as reference in each row within the 96-well plate. The culture is incubated at 37 °C overnight in a shaking incubator and 50 μL of 86% glycerin solution are added to each well. Finally, the master plate is frozen at -80 °C for storage and subsequent inoculation.

49

3.6.1.4

Inoculation of the main culture and Thal expression

The master plate serves for inoculation of the preculture containing 200 μL LB medium and the appropriate antibiotics mentioned above. The inoculation is performed with a purpose-built stamp by dipping the stamp into the slightly thawed master plate for 5 s and transferring the clones onto the new plate. This preculture is incubated at 37 °C overnight in a shaking incubator (450 rpm). The main culture is prepared in deep-well plates, containing 1.5 mL of LB medium per well. It is inoculated with 50 μL of preculture and incubated at 37 °C for approximately 4.5 h in a shaking incubator (1000 rpm). The temperature is lowered to 25 °C and after further 30 min the protein expression is induced by simultaneous addition of IPTG (0.1 mM) and L-arabinose (2 mg mL-1). Expression cultures are incubated for additional 20 h at 25 °C at 1000 rpm. 3.6.1.5

Cell harvesting and lysis

The cells are harvested by centrifugation (30 min, 3200 × g, 4 °C), the supernatant is discarded and the cells are frozen for at least one hour at -20 °C. For cell lysis 150 μL lysozyme (2 mg mL-1) dissolved in bromination buffer (10 mM K2HPO4 pH 7.4, 30 mM NaBr) is added to each well and incubated for 45 min at 25 °C. Afterwards, DNaseI is added (10 μL per well, 1 mg mL-1) and the cells are incubated for additional 20 min at 25 °C to hydrolyze released nucleic acids. In order to improve the efficiency of cell disruption, the cells are frozen again for one hour at -20 °C. After thawing them at 30 °C, cell debris is removed by centrifugation for 10 min at 3220 × g, 4 °C. 3.6.1.6

Activity screening for a thermostable Thal mutant

In order to screen for a thermostable Thal mutant, 60 μL of lysate are transferred into a new 96-well plate and heat-shocked at 49.5 °C for 20 min in a thermocycler. 50 μL of thermally treated lysate is added to a 96-well plate, containing 5 mM L-tryptophan, 10 μM FAD, 1 mM NAD+, 5 % (v/v) iso-propanol, 100 μg∙mL-1 ampicillin and 50 μM PMSF in each well. Afterwards 2.5 U∙mL-1 PrnF, 1.0 U mL-1 RR-ADH, 10 mM Na2HPO4 (pH 7.4), and 30 mM NaBr are added. The plate is sealed with an air-permeable film and incubated at 25 °C for 20 h at 450 rpm. Precipitate is removed by centrifugation and the remaining supernatant is used for the following screening by fluorogenic Suzuki reaction. 3.6.1.7

High-throughput fluorescence screening by Suzuki-Miyaura cross-coupling readout

The fluorescence readout is based on a Suzuki cross-coupling reaction. For this reaction 3-aminophenylboronic acid was identified as suitable reaction partner for L-6-bromotryptophan, resulting in a coupling product with a bathochromic shift in its absorption as well as fluorescence emission. This fluorophor is excited at its absorption maximum of 300 nm and the fluorescence emission is specifically detected at 420 nm. To perform the screening assay 10 μL of lysate samples containing the brominated amino acid are transferred to a 96-well PCR plate. Coupling buffer is added in a final concentration of 5 mM 3-aminophenylboronic acid and 7.5 mM K3PO4 per 100 μL reaction. After heat activation (40 °C, 10 min) of the catalyst solution, a final 50

concentration of 0.25 mM Na2Cl4Pd and 0.75 mM SPhos water soluble per 100 μL reaction are added to each well. To carry out fluorogenic Suzuki coupling, the plates are sealed with an adhesive PCR film and incubated for 2 h at 95 °C in a thermocycler. Afterwards, the reaction is quenched by addition of 2 % TFA and the fluorescence emission acquired using a microplate reader. 3.6.1.8

Re-screening of positive mutants

Mutants that show an increased fluorescence emission are re-screened as fivefold in the same manner as described above to validate data from previous high-throughput screenings. 3.6.1.9

Determination of melting point / stability of Thal and its mutants

1.5 L of E. coli culture containing overexpressed Thal WT or mutant is lysed by French press and the lysate is purified via HisTALON affinity chromatography as described above. The purified Thal is dialyzed against 10 mM Na2HPO4 pH 7.4 and 30 mM NaBr (5 L reservoir) overnight. 50 μL of dialyzed Thal is incubated as triplicate at different temperatures ranging from 40-56.3 °C for 20 min in a thermal cycler. Afterwards, 40 μL of thermally treated Thal is added to a 96-well plate, containing 5 mM L-tryptophan, 0.01 mM FAD, 1 mM NAD+, 5 % (v/v) iso-propanol, 100 μg∙mL–1 ampicillin and 50 μM PMSF in order to examine its bromination activity. Afterwards 2.5 U∙mL–1 PrnF, 1 U∙mL– 1 RR-ADH, 10 mM Na2HPO4 pH 7.4 and 30 mM NaBr are added to a final volume of 150 μL. The plate is sealed with an air-permeable film and the bromination reaction performed overnight at 25 °C in a shaking incubator (450 rpm). The final conversion of L-tryptophan is monitored by RP-HPLC and high-throughput fluorescence screening using Suzuki-Miyaura cross-coupling as described above. 3.7

Biotransformations

3.7.1

Tryptophan 6-halogenase Thal

3.7.1.1

Enzymatic bromination of L-tryptophan with purified Thal mutants

The bromination of L-tryptophan is performed with purified Thal-E2 and Thal-E2R from 1.5 L of E. coli culture containing the overexpressed halogenase. The purification is performed via HisTALON affinity chromatography as described before. The purified enzyme is dialyzed against 10 mM Na2HPO4 pH 7.4 and 30 mM NaBr over night in a 5 L reservoir in order to remove the contaminating chloride ions. 200 μL dialyzed enzyme solution is added to 800 μL bromination buffer with a final concentration of 5-10 mM tryptophan, 10 mM Na2HPO4 pH 7.4, 30 mM NaBr, 1 U mL-1 RR-ADH, 2.5 U mL-1 PrnF, 0.01 mM FAD, 1 mM NAD, 5 % (v/v) iso-propanol, 50 μM PMSF and 100 μg mL-1 ampicillin per 1 mL reaction. The mixture is incubated at 25 °C in a shaking incubator and monitored by RP-HPLC. At different time points aliquots of 30 μL are taken from the mixture and quenched in 30 μL MeOH. The mixture is centrifuged (12000 × g, 10 min) and 30 μL of supernatant is analyzed via RP-HPLC.

51

3.7.1.2

Enzymatic bromination of L-tryptophan in E. coli crude lysate

To perform the enzymatic bromination of L-tryptophan in E. coli crude lysate containing Thal WT or its mutant Thal-E2R 1.5 L of E. coli BL21 (DE3) pGro7 culture containing overexpressed Thal WT or Thal-E2R are resuspended in 30 mL bromination buffer containing 10 mM Na2HPO4 pH 7.4, 30 mM NaBr and 50 μM PMSF. The cells are lysed three times by French press and cell debris is removed by centrifugation (10000 × g, 30 min, 4 °C). The supernatant is divided into one 30 mL bromination reaction mixture and five mixtures with a final volume of 1 mL each. To prepare the biotransformation on 30 mL scale, 20 mL of lysate are added to 10 mM Ltryptophan, 1 mM NAD+, 0.01 mM FAD, 1 U mL-1 RR-ADH, 2.5 U mL-1 PrnF, 30 mM NaBr, 50 μM PMSF, 5 % (v/v) iso-propanol and 100 μg mL–1 ampicillin in a final volume of 30 mL. The reaction mixture is incubated at 25 °C in a shaking incubator (150 rpm) and the reaction progress is monitored by RP-HPLC as described before. 1 mL scale reactions are composed of 0.5 mL lysate and added to bromination buffer to reach a final concentration of 1 mM NAD+, 0.01 mM FAD, 1 U mL-1 RR-ADH, 2.5 U mL-1 PrnF, 30 mM NaBr, 50 μM PMSF, 5 % (v/v) iso-propanol, 100 μg mL–1 ampicillin and varying concentrations of L- tryptophan as mentioned below. Sample 1 2 3 4 5 L-tryptophan concentration 5 mM 5 mM 10 mM 10 mM 5 mM Incubation temperature 25 °C 37 °C 25 °C 37 °C 25 °C Heat-shock of lysate prior bromination no no no no yes (49.1 °C, 20 min) The reaction mixtures are incubated at the listed temperatures in a shaking incubator (150 rpm) and the reaction process is monitored by RP-HPLC. 3.7.1.3

Preparative enzymatic synthesis of L-6-bromotryptophan using Thal-PrnF-ADH combiCLEAS

Cross-linked enzyme aggregates (CLEAs) are an easily prepared method for enzymeimmobilization. CLEAs are more stable in comparison to soluble enzymes in their buffer solution, no further carrier material is required and at the end of reaction the solid biocatalyst can be easily removed by filtration. E. coli BL21 (DE3) pGro7 cells from 1.5 L culture containing the overexpressed halogenase Thal (WT or mutant) are resuspended in 30 mL 100 mM Na2HPO4 pH 7.4 and lysed three times by French press. After centrifugation (10000 × g, 30 min, 4 °C). 2.5 UήmL-1 PrnF and 2 UήmL-1RR-ADH are added to the cleared lysate and thoroughly mixed. Proteins are precipitated by addition of 16.2 g finely ground ammonium sulfate (95 % saturation) followed by incubation in a tube rotor for 1 h at 4 °C. The cross-linking reagent glutaraldehyde is added in a final concentration of 1.0 % (v/v) and the mixture is incubated for additional 2 h at 4 °C and centrifuged to sediment CLEAs (10000 × g, 30 min, 4 °C). The supernatant is discarded and the resulting combiCLEAS are washed

52

three times with 40 mL 100 mM Na2HPO4 pH 7.4 each and stored overnight at 4 °C to inactivate tryptophanase. For subsequent bromination the combiCLEAS are carefully resuspended in reaction buffer containing 1.0 mM L-tryptophan, 0.1 mM NAD+, 1 μM FAD, 5 % (v/v) isopropanol, 30 mM NaBr and 15 mM Na2HPO4. The pH value is adjusted with H3PO4 to pH 7.4 and H2O is added to a final volume of 1 L. The mixture is shaken at 150 rpm at 25 °C for several days and the reaction progress is monitored by HPLC. Afterwards, the combiCLEAs are removed by filtration. 3.7.1.4

Workup procedure for Thal-PrnF-ADH combiCLEAS

For subsequent workup the reaction mixture is concentrated under reduced pressure to a final volume of 50 mL, thereby evaporating remaining iso-propanol and water. For desalting the solution is purified with a RP-C18 silica gel column (flow rate: 9 mL min-1). Therefore, the column is equilibrated and salts removed with 300 mL H2O / 0.1%TFA and the elution of bound reaction components performed with 300 mL MeOH / 0.1% TFA. MeOH is removed under reduced pressure and the remaining product lyophilized to complete dryness. The residue is dissolved in 10 mL H2O/MeCN (1:1), separated via preparative HPLC and fractions containing the desired product are collected and lyophilized. Compound identity is checked by NMR spectroscopy, RP-HPLC and ESI-MS 6-bromotryptophan

(24) chemical formula: C11H11BrN2O2 Yield: 87.2 mg; 31 % RP-HPLC: tr = 2.6 min 13

C-NMR (126 MHz, DMSO-d6): δ [ppm] = 171.13 (CO2H), 137.50 (C7a), 126.49 (C3a), 126.40 (C2), 121.80 (C5), 120.43 (C4), 114.39 (C7), 114.28 (C6), 107.53 (C3), 52.95 (Cα), 26.28 (Cß).

1

H-NMR (500 MHz, DMSO-d6): δ [ppm] = 11.24 (s, 1H, indole-NH), 7.57 (d, J =1.6 Hz, 1H, C7H), 7.52 (d, 3J = 8.5 Hz, 1H, C4H), 7.26 (d, 3J = 2.2 Hz, 1H, C2H), 7.14 (dd, 3J = 8.5 Hz, 4J=1.7 Hz, 1H, CH5), 4.14 (t, 3J = 6.1 Hz, 1H, CαH), 3.25 (dd, 2J = 15.9 Hz, 3J= 6.1 Hz,1H, CβH2), 3.22 (dd, 2J = 15.7 Hz, 3J= 6.1 Hz, 1H, CβH2).

4

53

3.7.2

Biotransformation with Bmp5 on analytical scale

3.7.2.1

Enzymatic bromination of 4-HBA with purified His6-Bmp5

Purified His6-Bmp5 from 1.5 L E. coli BL21 (DE3) or E. coli BL21 (DE3) pGro7 pET28a-bmp5 culture is used for the enzymatic bromination of 4-HBA. 100 μL Bmp5 (c≈1 mg mL-1) is added to 2 mM 4-HBA dissolved in iso-propanol, 10 mM Na2HPO4 pH 7.4, 30 mM NaBr, 0.1 mM FAD, 5 mM NADPH in a final volume of 200 μL and incubated at 25 °C at 150 rpm. The reaction is monitored by RP-HPLC as previously described. 3.7.2.2

Enzymatic bromination of 4-HBA with purified GST-Bmp5

GST-Bmp5 purified via GST affinity chromatography from 1.5 L E. coli BL21 (DE3) culture serves as catalyst for the bromination of 4-HBA. 100 μL GST-Bmp5 (c=0.71 mg mL-1) is added to 2 mM 4-HBA, 10 mM Na2HPO4 pH 7.4, 30 mM NaBr, 0.1 mM FAD, 5 mM NADPH in a final volume of 200 μL and incubated at 25 °C. The reaction is monitored by RP-HPLC as described above. 3.7.2.3

Enzymatic bromination of 4-HBA in E. coli lysate using glucose dehydrogenase

In another approach a cofactor regeneration system employing glucose dehydrogenase is tested with His6-Bmp5. 1.5 L expression culture of E. coli BL21 (DE3) pET28a-bmp5 are resuspended in 20 mL bromination buffer containing 15 mM Na2HPO4 pH 7.4, 45 mM NaBr and 50 μM PMSF. The cells are lysed by French press for three times and the cell debris is removed by centrifugation (10000 × g, 30 min, 4 °C). The supernatant is used for enzymatic halogenation and therefore 2 mM 4-HBA, 2 mM NADP+, 0.01 mM FAD, 10 U mL-1 GDH, 40 mM glucose and 100 μg mL–1 ampicillin are added in a final volume of 30 mL. The reaction mixture is incubated at 25 °C in a shaking incubator (150 rpm) and the reaction progress monitored by RP-HPLC. 3.7.2.4

Enzymatic bromination of 4-HBA in E. coli lysate in presence of NaCl

For the enzymatic bromination of 4-HBA in presence of NaCl, an E. coli BL21 (DE3) pET28a-bmp5 lysate is prepared as described above. The cell lysate is divided into five separate bromination mixtures, each containing different concentrations of NaCl. 4 mL of lysate are added to 2 mM 4-HBA, 2 mM NADP+, 0.01 mM FAD, 2 U mL-1 LK-ADH, 5 % (v/v) iso-propanol and 100 μg mL–1 ampicillin to obtain an overall volume of 6 mL. The NaCl concentrations range from 0 mM; 30 mM, 50 mM, 100 mM to 300 mM. The reaction progress is monitored by RP-HPLC in the way described above. 3.7.2.5

Substrate screening of Bmp5 in E. coli lysate

For the substrate screening of Bmp5 in E. coli lysate, 1.5 L of E. coli BL21 (DE3) pET28a-bmp5 culture are resuspended in 30 mL bromination buffer (10 mM Na2HPO4, 30 mM NaBr pH 7.4) and 50 μM PMSF. The cell lysis is performed as described above. The cell lysate is divided into seven separate bromination mixtures each containing a different substrate (4-HBA, phenol, benzoic acid, aniline, indole, tryosine and phenylalanine). 4 mL of lysate are added to 2 mM substrate, 100 μg mL–1 ampicillin, 54

2 mM NADP+, 0.01 mM FAD, 2 U mL-1 LK-ADH and 5 % (v/v) iso-propanol in a final volume of 6 mL. The conversion of the potential substrate is checked by RP-HPLC. 3.7.2.6

Enzymatic iodination of 4-HBA in E. coli crude lysate

1.5 L of E. coli BL21 (DE3) pET28a-bmp5 culture are resuspended in 30 mL iodination buffer (10 mM Na2HPO4, 30 mM NaI, pH 7.4), 50 μM PMSF and lysed as described above. 5 mM 4-HBA, 2 mM NADP+, 0.01 mM FAD, 2 U mL-1 LK-ADH, 5 % (v/v) isopropanol and 100 μg mL–1 ampicillin are added to the lysate and the reaction mixture is incubated at 25 °C in a shaking incubator (150 rpm). The reaction progress is monitored by RP-HPLC as described above. 3.7.2.7

GC-MS sample preparation

In order to identify the compounds of Bmp5-catalyzed reactions, aliquots of 1 mL are taken at different time points and prepared for GC-MS analysis. The reaction is stopped by heat precipitation (10 min, 95 °C) and the samples are centrifuged to remove precipitated proteins (10000 × g, 10 min). The supernatant is transferred to a new Eppendorf tube and the pH value is adjusted to pH 1-2. The organic compounds are extracted three times with 250 μL ethyl acetate each and the organic solvent is evaporated at 60 °C to complete dryness. The residue is re-dissolved in 100 μL ethyl acetate for recording GC-MS spectra. 3.7.3

Biotransformation with Bmp5 on preparative scale

3.7.3.1

Enzymatic bromination of 4-HBA in E. coli crude lysate

The enzymatic bromination of 4-HBA catalyzed by Bmp5 in E. coli crude lysate is performed in two different approaches regarding cofactor regeneration. In the first approach the cofactor NADPH is regenerated by LK-ADH, which reduces NADP+ by oxidizing iso-propanol. This approach is performed in a reaction mixture with a final volume of 30 mL. Therefore, 1.5 L expression culture of E. coli BL21 (DE3) pET28a-bmp5 containing overexpressed Bmp5 are resuspended in 20 mL bromination buffer containing 15 mM Na2HPO4 pH 7.4, 45 mM NaBr and 75 μM PMSF. The cell lysis is performed by French press for three times and the cell debris is removed by centrifugation (10000 × g, 30 min, 4 °C). The supernatant is employed for enzymatic halogenation and therefore 2 to 5 mM 4-HBA, 2 mM NADP+, 0.01 mM FAD, 2 U mL-1 LK-ADH, 5 % (v/v) iso-propanol and 100 μg mL–1 ampicillin are added into an Erlenmeyer flask and the reaction mixture is incubated at 25 °C in a shaking incubator (150 rpm). To track the reaction progress by HPLC, aliquots of 80 μL are taken from the mixture at different time points and the reaction is stopped by addition of an equal volume of MeOH. After stopping the reaction workup of the reaction mixture is performed as described below. The second reaction mixture is performed on a 1 mL scale and does not contain LK-ADH but an excess of NADPH. The cell lysate is prepared in the same way as described above and to 0.5 mL of lysate 2 mM 4-HBA, 10 mM Na2HPO4 pH 7.4, 30 mM NaBr, 5 mM 55

NADPH and 0.01 mM FAD are added stepwise. The reaction mixture is incubated at 25 °C in a shaking incubator and the reaction progress is monitored by RP-HPLC. 3.7.3.2

Preparative synthesis of 2,4-dibromophenol using Bmp5-ADH combiCLEAS

The preparation of combiCLEAS from 1.5 L E. coli BL21 (DE3) culture containing overexpressed Bmp5 is performed analogously to the Thal combiCLEAs preparation (3.7.1.3) described before with the following exceptions. For continuous cofactor regeneration, LK-ADH (2 U mL-1) is employed, while no flavin reductase is required. After protein precipitation glutaraldehyde is added in a final concentration of 0.7 % (v/v). Bmp5-ADH combiCLEAs are immediately employed for bromination without further storage. For subsequent bromination the combiCLEAS are resuspended in reaction buffer containing 1.0 mM 4-HBA, 0.5 mM NADP+, 1 μM FAD, 5 % (v/v) iso-propanol, 30 mM NaBr and 15 mM Na2HPO4. The pH value is adjusted with H3PO4 to pH 7.4 and H2O is added to a final volume of 0.5 L. The mixture is shaken at 150 rpm at 25 °C for several days and the reaction progress is monitored by HPLC. Afterwards, the combiCLEAs are removed by filtration. The iodination of 4-HBA with combiCLEAS is performed as described above, but using 30 mM NaI except NaBr. 3.7.3.3

Processing of reaction product after enzymatic bromination of 4-HBA with combiCLEAs

After removal of combiCLEAS by filtration, the pH value of the reaction solution is adjusted to pH 1-2 and reaction components are extracted five times with CH2Cl2 (150 mL each). The solvent is removed under reduced pressure, the residue dissolved in approximately 5 mL CH2Cl2 and purified by column chromatography using silica gel as stationary phase and CH2Cl2 as eluent. The elution fractions are examined by thin-layer chromatography for the presence of the desired compound. Product-containing fractions are collected and the solvent is evaporated to dryness under reduced pressure. During storage at 4 °C colorless, needle-shaped crystals of 2,4-dibromophenol are obtained. Compound identity is checked by NMR spectroscopy, RP-HPLC, GC-MS and ESI-MS.

56

2,4-dibromophenol

(23) chemical formula: C6H4Br2O yield: 50.8 mg; 40 % ESI-MS: m/z = [M(79Br/79Br)-H]- obs. 248.7 calc. 248.86; [M(79Br/81Br)-H]- obs. 250.7 calc. 250.8; [M(81Br/81Br)-H]- obs. 252.7 calc. 252.8 GC-MS: tr = 14.3 min / m/z = [M+∙(79Br/79Br)] obs. 251.8 calc. 249.9; [M+∙(79Br/81Br)] obs. 253.3 calc. 251.9; [M+∙(81Br/81Br)] obs. 255.2 calc. 253.9 RP-HPLC: method A: tr= 3.8 min Rf (CH2Cl)= 0.78 1

H-NMR (500 MHz, DMSO-d6): δ [ppm] = 10.57 (s, 1H, OH), 7.66 (d, 4J = 2.3 Hz, 1H, C3H), 7.35 (dd, 3J = 8.7 Hz, 4J =2.4 Hz, C5H), 6.91 (d, 3J = 8.7 Hz, 1H, C6H).

13

C-NMR (126 MHz, DMSO-d6): δ [ppm] = 154.2 (COH), 134.9 (C3), 131.9 (C5), 118.4 (C6), 110.9 (C4), 110.8 (C2).

57

4

Directed evolution of Thal: Results

4.1

Directed evolution of tryptophan 6-halogenase Thal

4.1.1

Optimization of epPCR for random mutagenesis of thal

For random mutagenesis of thal adjustment of the mutation rate during epPCR is an essential precondition to obtain high-quality mutant libraries suitable for the directed enzyme evolution. A mutation frequency of 2-7 nucleotide mismatches per gene is recommended for the initial mutation approach.[48] Regarding the 1.6 kb thal gene a mutation rate of 2 mutations/kb was desired and therefore epPCR optimized with regard to this target. Initially, different template as well as MnCl2 concentrations were tested in the first attempt of epPCR. The employed template concentrations corresponded to the amount of DNA to be amplified and thereby referred to the length of thal insert serving as template in PCR. The tested DNA amounts ranged from 23-115 ng, whereas the mutation rate was expected to increase with decreasing amount of template DNA. As recommended 100 ng template DNA should result in a low mutation rate of 0-4.5 mutations/kb, in contrast a lower template concentration is expected to lead to an increase of mutations per gene[48]. In parallel, the error rate of the employed GoTaq polymerase was increased by the addition of MnCl2 which is known to replace the cofactor Mg2+ in the active site of the polymerase, thereby favoring nucleotide misincorporation. The number of PCR cycles was adjusted to 25 cycles in comparison to a routine PCR without mutagenesis where generally up to 30 cycles are performed. The lowering of PCR cycles achieved fewer target duplications and therefore lowered the probability that an already mutated DNA strand served as template in a new amplification cycle. In order to enable nucleotide misincorporations all over the gene, the employed primers merely flanked the encoding region of thal with complementarity to the surrounding pET28a vector backbone. This strategy was meant to avoid that the terminal regions of the gene were excluded from mutagenesis, as it was unknown which amino acid residues of Thal were crucial for thermal stability.

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_4

59

The specific amplification of thal was checked by agarose gel electrophoresis and all four epPCR mixtures showed the desired thal DNA amplificate with 1.6 kb in length. In an additional epPCR reaction 300 ng template DNA were employed along with 100 μM MnCl2.

Figure 5

EpPCR of 1.6 kb thal using pET28a-thal as template and employing the primers pET28a-NdeI-FW and pET28a-BamHI-Stop-REV. For variation of the mutation rate the reaction mixtures contained either 115 or 23 ng template DNA and 150 or 200 μM MnCl2. 10 μL of DNA sample were separated on 1 % agarose gel and stained via ethidium bromide.

All epPCR products were purified and enzymatically digested with the restriction enzymes NdeI and BamHI for subsequent ligation with the analogously digested pET28a expression vector and transformed via heat-shock into E. coli BL21 (DE3) pGro7. In order to avoid the transformation of methylated template plasmid, that lacked any mutations, but would transform more efficiently than unmethylated DNA, the PCR products were additionally digested by DpnI. This restriction enzyme only recognized methylated DNA for cleavage and thereby removed the template plasmid. The efficiency of the DpnI digestion was checked with a further epPCR mixture as control. This examination was of special importance to ensure that undesired wild type background did not rise by incomplete removal of template. This positive control was prepared in the same manner as all described epPCR samples except that it did not contain DNA polymerase, so that only template plasmid was present and no amplification occurred. In the following this epPCR product was digested with NdeI, BamHI and DpnI, ligated and transformed into E. coli BL21 (DE3) pGro7. As expected no E. coli colonies were identified on the agar plate, proving the successful digest of methylated template plasmid DNA.

60

After successful transformation into E. coli BL21 (DE3) pGro7, positive clones were identified by colony-PCR, pET28a-thal plasmids isolated and the individual mutation rate of two clones checked for each epPCR preparation by Sanger sequencing. Table 2

Identified mutation rate of Thal mutants after epPCR performed under variation of template as well as MnCl2 concentration.

Initial template amount 23 ng 23 ng 115 ng 115 ng 300 ng

MnCl2 concentration 150 μM 200 μM 150 μM 200 μM 100 μM

Mutations/ 500 pb 1-2 4 none 2-5 1

The sequencing results revealed a mutation rate of 1-4 mutations per 500 bp for an initial template amount of 23 ng. Higher concentrations of MnCl2 led to an expected increase of the mutation rate. As 2-3 mutations per thal gene were the aspired goal, this low amount of template would probably lead to much higher mutation rates and was therefore not appropriate for random mutagenesis of thal. In contrast, an initial template amount of 115 ng resulted in a mutation rate varying between 0 and 5 mutations per 500 bp. Again the reaction mixture containing higher MnCl2 concentration showed an increased mutation rate but as the amount of 5 mutations per 500 bp was too high, and the mutation rate varied too strong, this epPCR composition was not used for further mutagenesis of thal. In order to lower the mutation rate to 1 mutation per 500 bp, the initial template amount was increased to 300 ng DNA and the MnCl2 concentration decreased to 100 μM. One mismatch per 500 bp was observed, this result approximates an average of three mutations in the 1.6 kb thal gene. Therefore, this epPCR conditions were chosen for further random mutagenesis of thal. As the amount of sequenced Thal mutants was rather low and the identified mutation rate not precise enough to make a reliable statement about the impact of the employed epPCR composition, these results were not absolutely representative. Still they gave a rough estimation about the required amount of template DNA and MnCl2 concentrations, in order to obtain a suitable mutation rate. 4.1.2

Characterizing the thermostability of Thal WT

The directed evolution of the tryptophan 6-halogenase Thal focused on the development of a thermostable variant, striving for an increased stability, lifetime and catalytic activity of the biocatalyst. To follow this aim initially the thermostability of Thal WT was investigated by heat treatment of the enzyme at different temperatures. Subsequent measurement of the remaining bromination activity, employing L-tryptophan as substrate, revealed the impact of the thermal treatment on Thal's thermal resistance. This enabled

61

the identification of the maximum temperature threshold for Thal WT to be still sufficiently active prior to denaturation. For the investigation on the Thal WT thermostability, the pET28a-thal plasmid, encoding for N-terminal His6-tagged Thal, was transformed into E. coli BL21 (DE3) pGro7. The E. coli BL21 (DE3) pGro7 strain was chosen due to the coexpression of the chaperones GroEL and GroES to enhance the proper folding of overexpressed His6-Thal during protein biosynthesis. Expression conditions of Thal were derived from the wellestablished procedure for the tryptophan 7-halogenase RebH that could be previously transferred to other tryptophan halogenases.[33] 4.1.2.1

Protein purification via HisTALON affinity chromatography

The fusion protein His6-Thal was lysed by French Press and subsequently purified via HisTALON affinity chromatography from E. coli lysate. The resulting fractions were separated by SDS-PAGE to ensure successful purification as well as sufficient purity of Thal.

Figure 6

SDS-PAGE analysis of Thal WT expression in E. coli BL21 (DE3) pGro7 pET28a-thal. After cell disruption via French Press the lysate was purified by HisTALON affinity chromatography. The flow-through, washing fraction 1, washing fraction 2 and elution fractions were analyzed by 12 % polyacrylamide gel (4 μL of FT and WF; 2 μL for elution fractions). All elution fractions showed an intensive protein band at 60 kDa, corresponding to His6-tagged Thal's molecular weight.

During application of the cell lysate, His6-tagged Thal interacted and bound tightly to the matrix, whereas other proteins or contaminations lacking a His6-tag did not bind and were collected in the flow-through. Therefore, this fraction consisted of numerous cellular E. coli proteins differing in molecular weight. A dominant protein band at 60 kDa was identified, corresponding to the molecular mass of the chaperone GroEL as well as nonbound Thal. The analysis of washing fractions revealed the same intense protein band at 60 kDa, corresponding to the mentioned proteins. As the second washing buffer contained a low concentration of imidazole (10 mM), un-specifically bound proteins were eluted, whereas a minor loss of Thal was inevitable. The final elution of Thal was performed in 62

presence of 300 mM imidazole. An intense protein band at 60 kDa could be identified for all elution fractions as well as several protein bands of lower mass that resembled proteolytically degraded Thal, still containing the His6-tag. The Thal containing elution fractions were combined and the overall protein concentration determined to 11.0 mg mL-1. In order to employ Thal for subsequent bromination of L-tryptophan, contaminating chloride ions were removed by dialysis against 10 mM Na2HPO4, 30 mM NaBr pH 7.4 over night. During dialysis a loss of protein was inevitable and the concentration of Thal dropped to 5.2 mg mL-1. 4.1.2.2

Determination of thermostability

Thal's thermostability was examined by incubating purified enzyme as triplicate at different temperatures ranging from 40.0 to 56.3 °C for 20 minutes. This temperature range was assumed to be critical for halogenase stability. The remaining catalytic efficiency of Thal WT was tested by following bromination of L-tryptophan in a 96-well plate in presence of continuous cofactor regeneration. Subsequent to the heat-shock the substrate, cofactors as well as auxiliary enzymes were added to the halogenase, because these reaction components should not be exposed to elevated temperatures. The final conversion of L-tryptophan to L-6-bromotryptophan was analyzed by RP-HPLC measurement and additionally by fluorogenic Suzuki-Miyaura cross-coupling to evaluate the impact of thermal treatment on Thal's enzymatic activity (Figure 7).

63

(a)

(b)

Figure 7

Investigation of Thal WT's thermostability by determination of the total conversion of to L-6-bromotryptophan in dependence of the heat-shock temperature of purified enzyme (mean value, error bars indicate the standard deviation). Thal's activity was monitored by RP-HPLC measurements (a) as well as fluorogenic highthroughput screening (b). Both methods revealed a rapidly decreasing enzyme activity beyond 45.5 °C and nearly no overall activity at 49.1 °C or higher temperatures. L-tryptophan

A triplicate of Thal without prior heat-shock served as control that showed an average conversion of 99 % to prove functional setup of the reaction system. Regarding the Thalcatalyzed bromination of L-tryptophan monitored by RP-HPLC the conversion remained 64

constant above 98 % up to a heat-shock temperature of 45.5 °C. A further temperature increase of 1.8 °C led to a dramatic decrease of enzyme activity, finally dropping to less than 3 % above 49.1 °C (Figure 7a). In addition, the comparable curve progression was also observed for the fluorescence emission of the Suzuki-Miyaura cross-coupling product L-6-(3’-aminophenyl)tryptophan (26) in dependence of Thal's heat-shock temperature (Figure 7b). As the fluorescence signal of the coupling product was proportional to the concentration of 6-bromotryptophan, the fluorescence intensity at 420 nm referred to the halogenation activity of Thal. Further detail is given in chapter 4.1.3. The strikingly bigger error bars for the fluorescence readout compared to the RP-HPLC measurements resulted from more manual pipetting steps as well as evaporation during incubation for Suzuki coupling. The RP-HPLC measurements as well as fluorescence read-out led to consistent results, indicating a thermal stability of Thal WT up to 45.5 °C and a negligibly small catalytic activity above 49.1 °C. Thus, Thal gets inactivated due to denaturation around 49 °C. In order to screen for thermostable Thal mutants, this temperature was chosen as screening criterion to select those variants that are still active above 49.1 °C. 4.1.3

Directed evolution of Thal with screening towards increased thermostability

For random mutagenesis of Thal WT epPCR was chosen as suitable technique and as previously described the reaction conditions were optimized to achieve the desired mutation rate of 2-3 mutations/kb. Based on this strategy randomly mutated thal libraries were generated, the gene variants subcloned into a pET28a expression vector and transformed into E. coli BL21 (DE3) pGro7. Master plates were prepared as described in chapter 3.6.1.3 and served for the inoculation of main cultures and following Thal expression in 96-well plates. Moreover, master plates were required as backup to isolate plasmid DNA of potentially interesting mutants. After cell harvest and lysis by addition of lysozyme and a freeze-thawing step, the Thal-containing lysate was heat-shocked at 49.1 °C, a temperature previously identified to be high enough to prevent any catalytic activity of Thal WT and therefore enabled selection of more thermostable mutants. The variants were employed for subsequent biohalogenation of L-tryptophan and the mutants' halogenation activity was quantified based on the high-throughput fluorescence screening by derivatization using Suzuki-Miyaura cross-coupling (Scheme 6). Regarding directed evolution, one of the most important as well as limiting factors is a reliable and rapid high-throughput screening assay to identify the desired mutants. In this context, Schnepel et al. developed an assay based on Suzuki-Miyaura coupling appropriate for high-throughput screening to quantify the halogenation activity of tryptophan 5-, 6- and 7-halogenases by fluorescence readout (manuscript in preparation).

65

Scheme 6

4.1.3.1

Strategy for directed evolution of Thal towards a thermostable variant.

Evaluation of high-throughput screening for thermostable Thal mutants

According to the previously described procedure, the following figure shows the results of high-throughput screening data targeting the identification of a thermostable Thal mutant. A total of 240 mutants was screened and the assignment of the examined colonies performed by plotting the fluorescence emission against its corresponding position in the 96-well plate (Figure 8).

66

(a)

(b)

(c)

Figure 8

Corresponding to the previously described procedure, the evaluation of the highthroughput screening was performed by plotting the fluorescence emission at 420 nm against the clone's position in the 96-well plate, striving for identification of a thermostable Thal variant. The data show the screening results of 240 analyzed clones, that led to the identification of two putative transformants (B6, F6, increased conversion by factor 1.5, figure b) and one clone (E2, figure c) that exhibited an increased fluorescence emission by factor 2.7 compared to the mean value.

The majority of measured values scattered evenly around a mean value, even though the absolute values of fluorescence emission differed among the presented data. In comparison to figure (a), that did not show a significantly higher fluorescence of one particular mutant, figure (b) revealed two (mutant B6 and F6) and figure (c) one clone (E2) with remarkably higher fluorescence compared to the mean value. The mutants B6 and F6 showed an increase of conversion by factor 1.5 whereas mutant E2 revealed a considerable 2.7-fold increase. These promising results indicated an increased bromination activity and thereby pointed towards an improved thermostability for the identified mutants that were subsequently re-screened to pursue these findings.

67

4.1.3.2

Re-screening of positive Thal mutants

The Thal mutants B6, F6 and E2 were identified by high-throughput fluorescence screening to exhibit a remarkably high catalytic activity despite thermal treatment at 49.1 °C. In order to validate the obtained data, these clones as well as Thal WT were rescreened as fivefold replicate. To evaluate the impact of the thermal treatment on the mutant's activity, the re-screening was additionally performed without heat-shock prior bromination reaction serving as reference.

Figure 9

Re-screening of Thal mutants that revealed an increased conversion by highthroughput fluorescence readout. The mutants E2, B6, F6 as well as the WT were rescreened as fivefold replicate and the average fluorescence emission corresponding to the concentration of L-6-bromotryptophan is plotted (error bars indicate the standard deviation). In addition to the re-screening strategy including thermal treatment at 49.1 °C (20 min), an additional examination of mutant activity was performed without thermal treatment (control). Clone E2 showed in both approaches a significantly elevated fluorescence emission in comparison to the other clones and the WT examined in this experiment.

The results of the re-screening approach supported the assumption of an improved thermostability of the Thal mutant E2. This mutant showed a significantly higher fluorescence emission in both measurements. Subsequent to treatment at 49.1 °C it showed a 1.7-fold higher conversion than the WT enzyme. Without thermal treatment the overall conversion of Thal-E2-catalyzed reactions was elevated by factor 2.7 confirming the observations obtained in high-throughput screening. In contrast, the mutants B6 and F6 did not show an increased fluorescence emission compared to the wild type in both reaction approaches. Therefore, these mutants were not further investigated.

68

4.1.3.3

Sequencing results for Thal mutant E2

The DNA sequence of Thal mutant E2 was checked by Sanger sequencing to identify possible mutations of the putative thermostable variant. The following table summarizes the identified mutations in the gene thal-E2. Table 3

Identified mutations (nucleotide and amino acid exchange) of Thal-E2 in comparison to the WT after DNA sequence analysis.

Thal WT Thal E2

1st mutation Codon AGC Ser-359 GGC Gly-359

2nd mutation Codon AAA Lys-374 AGA Arg-374

3rd mutation Codon AAA Lys-510 AAG Lys-510

The sequencing of Thal-E2 led to the identification of three mutations, whereas two of them resulted in an amino acid exchange from serine to glycine at position 359 and Lys374 to Arg-374. The third mutation revealed a silent mutation, not leading to an amino acid exchange. 4.1.4

Further characterization of Thal mutant E2

As previously described the Thal mutant E2 was initially identified by high-throughput screening to exhibit a remarkably increased catalytic activity in comparison to the WT even after incubation at 49.1 °C. Sequencing of thal-E2 revealed two alternations of amino acid residues. For further characterization of this mutant and examinations on its catalytic performance as well as thermostability, Thal-E2 was overexpressed in E. coli BL21 (DE3) pGro7 on 1.5 L scale and purified for following experiments. 4.1.4.1

Expression and purification of Thal E2

For expression of Thal-E2 in E. coli BL21 (DE3) pGro7 the corresponding master plate containing E. coli BL21 (DE3) pGro7-pET28a-thal-E2 served for inoculation of main cultures and the subsequent overexpression was performed in analogy to the WT. After cell harvest and lysis His6-Thal-E2 was purified via HisTALON affinity chromatography and success of purification was checked by SDS-PAGE. In addition to the previously described purification of Thal WT the cleared lysate prior purification, as well as the insoluble cell debris after solubilization in 8 M urea were analyzed by SDS-PAGE along with the flow-through, washing and elution fractions.

69

Figure 10

SDS-PAGE analysis of Thal-E2 expression in E. coli BL21 (DE3) pGro7 pET28a-thal-E2. The lysate was purified via HisTALON affinity chromatography and the flow-through, washing fraction 1, washing fraction 2 and elution fractions were collected and separated in a 12 % polyacrylamide gel. In addition, the cleared lysate as well as the cell pellet collected after cell lysis were analyzed. (10 μL for elution fractions, 5 μL for remaining samples). All identified protein bands were rather blurred and the elution fractions revealed a very faint band around 60 kDa, corresponding to the molecular weight of His6-tagged Thal-E2.

Separation of the cleared lysate resulted in several blurred protein bands, as it consisted of various differently sized E. coli proteins. A slightly more intense band could be identified at 60 kDa, corresponding to the molecular mass of GroEL and Thal-E2. In analogy, analysis of the flow-through revealed several bands from E. coli proteome. The dominant protein band at 60 kDa probably matching to the mentioned proteins was identified as well. Washing fraction 1 showed the same intense protein band at 60 kDa, whereas this band was barely visible in the second washing fraction. The elution of His6-Thal-E2 was performed in presence of imidazole whereas only a faint protein band in the range of 60 kDa was identified for all elution fractions pointing to a low overall concentration of the desired His6-tagged protein. A few equally weak protein bands of lower mass, probably corresponding to proteolytically degraded Thal-E2 were detected as well. To exclude the possibility of formation of aggregated Thal-E2 in inclusion bodies that might precipitate within the insoluble protein fraction, the components of the cell debris were solubilized in 8 M urea and subsequently analyzed by SDS-PAGE. A faint and uneven band was identified at 60 kDa, corresponding to the chaperone's and Thal-E2's molecular mass. As the band was very weak in comparison to cleared lysate, Thal-E2 was not expected to accumulate significantly in inclusion bodies. After combination of the ThalE2 containing elution fractions the overall protein concentration was determined to 70

2.5 mg mL-1. This protein concentration was much lower than expected, as it resembled only one quarter of purified Thal WT. 4.1.5

Site-directed mutagenesis of Thal-E2

The previously described purification of Thal-E2 resulted in an unsatisfying low total protein concentration, what could prove problematic with regard to biocatalytic purposes employing purified enzyme. Therefore, the sequencing results of Thal-E2 were analyzed in more detail, especially focusing on the silent mutation at amino acid position 510. As not each codon is used with equal frequency in E. coli, it was revealed that the mutation of the AAA codon against AAG in the Thal-E2 sequence resulted in a 68 % less frequently used codon for lysine.[49] This rarely employed codon was initially considered as possible explanation for the decreased yield of Thal-E2 in comparison to the WT after purification. In order to restore the original WT sequence regarding this triplet, a sitedirected reverse mutation PCR of the mutated codon was performed. 4.1.5.1

Reverse mutation of Thal-E2

The reverse mutation of the AAG codon at amino acid position 510 in Thal-E2 according to the wild type sequence AAA was conducted with pET28a-thal-E2 plasmid serving as template. Site-directed mutagenesis PCR was performed as described in 3.4.6.2 and the success of PCR checked by agarose gel electrophoresis (Figure 11).

Figure 11

Site-directed mutagenesis of Thal-E2 to reverse the silent mutation at amino acid position 510. The WT sequence was obtained within this codon by using ThalE2K510-FW and ThalE2-K510-REV primers. Successful amplification of plasmid DNA during mutagenic PCR was confirmed by separation in 1 % agarose gel.

The observed DNA band with a length of 7.0 kb corresponded to pET28a-thal-E2. To complete the mutagenesis the PCR product was purified, the template plasmid digested

71

by DpnI and the linear PCR product pET28a-thal-E2R transformed into E. coli DH5α to receive circular plasmid pET28a-thal-E2R. 4.1.5.2

Sequencing results of Thal mutant E2R

Sequencing results of Thal-E2R confirmed the successful reverse mutation of codon 1530, encoding lysine-510 according to the WT sequence. Table 4

Identified mutations of Thal-E2R in comparison to the WT after reverse mutation PCR of Thal-E2.

Thal WT Thal-E2R

1st mutation Codon AGC Ser-359 GGC Gly-359

2nd mutation Codon AAA Lys-374 AGA Arg-374

4.1.6

Further characterization of Thal-E2R

4.1.6.1

Expression and purification of Thal-E2R

Reverse mutation Codon AAA Lys-510 AAA Lys-510

Thal-E2R was overexpressed in E. coli BL21 (DE3) pGro7 and purified via HisTALON affinity chromatography in accordance to previously established procedure for Thal-E2. The protein content of the sample before induction of protein overexpression, the cleared lysate of E. coli BL21 (DE3) pGro7 pET28a-thal-E2R as well as flow-through, washing fractions and elution fractions of affinity chromatography were analyzed in a polyacrylamide gel.

Figure 12

72

SDS-PAGE analysis of Thal-E2R expression in E. coli BL21 (DE3) pGro7 pET28a-thal-E2R. A sample prior induction as well as the cleared lysate were separated in a 12 % polyacrylamide gel along with the collected flow-through, washing fraction 1, washing fraction 2 and elution fractions after HisTALON affinity chromatography (10 μL for elution fractions, 5 μL for remaining samples). All elution fractions revealed an intense band around 60 kDa, corresponding to the molecular weight of His6-tagged Thal-E2R.

Several protein bands of major and minor size occurred whereas a band around 60 kDa was not detected prior induction according to the expectations. The cleared lysate revealed Thal-E2R's and GroEL's corresponding band at 60 kDa and broadly distributed protein bands of different size. The flow-through and washing fractions contained a decreasing amount of Thal-E2R and GroEL as visible by less intense protein bands at 60 kDa. However, all elution fractions showed a considerable protein band at 60 kDa. These observations indicated an increase in protein expression of Thal-E2R compared to Thal-E2 due to more intense 60 kDa bands obtained in the elution fractions. Despite this observation, the total protein concentration of Thal-E2R was determined to merely 2.5 mg mL-1, thereby showing no significant improvement in protein expression. 4.1.7

Expression of Thal-E2R in E. coli BL21-CodonPlus(DE3)-RIL

As the reverse mutation of the Lys-510 codon to the WT sequence did not reveal the desired increased protein concentration subsequent to purification, the focus was again laid on the sequencing results of Thal-E2R, especially concentrating on the mutation of Arg-374. It turned out that the mutation of the Lys codon AAA into AGA encoding Arg resulted in a triplet that is merely 4 % employed in E. coli relative to all remaining codons encoding arginine.[49] As this is worth to be considered as reason for the decreased concentration of Thal-E2R in comparison to the WT, pET28a-thal-E2R was transformed via heat-shock into E. coli BL21 CodonPlus(DE3)-RIL. This strain contains extra copies of rare tRNAs for the arginine codons AGA and AGG, the isoleucine codon AUA, and the leucine codon CUA. However, subsequent expression and purification of Thal-E2R via HisTALON affinity chromatography still revealed low protein concentrations ranging around 1 mg mL-1. As the purification of Thal-E2 and Thal-E2R proved challenging and could not be improved at this stage of research, the examination of catalytic properties regarding their thermostability was performed with these low protein concentrations. As no improvement of expression level was observed in the E. coli strain BL21 CodonPlus(DE3)-RIL, further investigations were performed with Thal-E2R overexpressed in E. coli BL21 (DE3) pGro7. 4.2

Investigating the thermostability of Thal-E2 and -E2R

After purification of Thal-E2 and Thal-E2R their thermostability was analyzed in comparison to the WT. To examine remaining bromination activity subsequent to thermal treatment the dialyzed enzyme solutions were incubated for 20 min at different temperatures ranging from 41.9 to 58.1 °C and subsequently employed for the bromination of 5 mM L-tryptophan in a 96-well plate at 25 °C. The final conversion of L-tryptophan was monitored by RP-HPLC and high-throughput fluorescence screening. The resulting data were normalized to the employed protein concentration, thereby allowing the direct comparison between the residual activity of Thal WT and its mutants (Figure 13).

73

Figure 13

Comparison of Thal WT's, E2's and E2R's remaining bromination activity in dependence of heat-shock temperature by plotting the relative conversion of L-tryptophan determined by RP-HPLC versus the temperature gradient. Thal-E2R showed the highest catalytic activity nearly throughout the entire temperature range, thereby pointing to an increased thermostability especially in contrast to Thal WT that revealed by far the lowest catalytic activity. As expected, thermostability of E2 and E2R was similar, whereas slight differences can be neglected.

An increase in heat-shock temperature resulted for all Thal variants in a decrease of residual conversion at 25 °C, whereas the individual curve progressions revealed that the enzymes' catalytic activity was affected differently. Thal-E2 and Thal-E2R showed a slight, approximately linear decrease of activity, whereas the WT's conversion remained constant at lower temperatures but then decreased abruptly above 46 °C. In general, activity profile of Thal-E2 proceeded similarly to E2R and varied merely around 5%. The Thal mutant E2R showed the highest relative conversion of 31 % even in the examined temperature range between 41.9 and 54 °C. Nevertheless, residual conversion catalyzed by these Thal mutants remained considerably above the wild type. Notably, at temperatures above 56.0 °C bromination activity of Thal-E2R and Thal-E2 dropped below 5 %. In contrast, Thal WT's relative conversion remained constant at 20 % up to a heat-shock temperature of 45.5 °C, but then decreased dramatically with further temperature increase dropping to complete inactivity above 49.0 °C. Summing up, Thal-E2R revealed the highest enzymatic activity despite previous thermal treatment indicating an improved thermostability especially to the WT. These findings were further confirmed by fluorescence screening assay.

74

4.3

Bromination ability of Thal variants and comparison to the wild type

4.3.1

Bromination attempts with purified Thal mutants

As previously described the purification of the Thal mutants E2 and E2R via HisTALON affinity chromatography proved complicated due to very low protein concentrations, ranging around 2 mg mL-1 originating from 1.5 L E. coli culture. As the enzyme solutions were additionally dialyzed in order to remove chloride ions, protein concentrations dropped to values around 0.2-0.5 mg mL-1, further impeding enzymatic approaches using purified enzymes. Nevertheless, the conversion of 5 mM and 10 mM L-tryptophan was attempted with the purified Thal mutants in 0.5-1 mL reaction volume as described in 3.7.1.1. However, no significant conversion of substrate was identified either for Thal-E2 or Thal-E2R. Consequently, further bromination reactions were performed in E. coli BL21 (DE3) pGro7 lysate containing either pET28a-thal-E2 or pET28a-thal-E2R to circumvent protein purification that led to a serious loss of halogenase. 4.3.2

Bromination of L-tryptophan in E. coli crude lysate

The improved thermostability and enzymatic activity of Thal-E2R was further characterized by bromination reactions of L-tryptophan in crude lysate, with variations in substrate concentration, incubation temperature and by heat-shocking the lysate prior bromination reaction. These different approaches investigated the potential of Thal-E2R for chemoenzymatic applications in comparison to the wild type. Therefore, several reactions were performed on 1 mL scale, whereas an additional bromination reaction of 10 mM L-tryptophan was implemented in a 30 mL reaction volume. All experiments were performed with crude lysate containing either Thal-E2R or Thal WT. The performance of experiments and the results are summarized in the following tables. Table 5

Composition of 1 mL reaction samples regarding substrate concentration, incubation temperature and thermal treatment prior bromination reaction.

Sample (S) L-tryptophan concentration Incubation temperature Heat-shock of lysate prior bromination

1 5 mM 25 °C no

2 5 mM 37 °C no

3 10 mM 25 °C no

4 10 mM 37 °C no

5 5 mM 25 °C yes

75

Table 6

S 1 2 3 4 5

Figure 14

Results of Thal WT- and E2R-catalyzed bromination in crude lysate under differing reaction conditions as mentioned in Table 5.

Reaction time 44 h 52 h 44 h 44 h 44 h

E2R 100 % 53 % 47 % 29 % 97 %

Conversion of Thal variant WT 33 % 53 % 19 % 29 % 49 %

Comparison of Thal WT's and E2R's catalytic activity in crude lysate for five different reaction samples (S1-S5). Thal-E2R showed a significantly increased conversion of Ltryptophan at an incubation temperature of 25 °C (S1, S3, S5), whereas the heat-shock did not have a negative impact on the enzyme activity.

Regarding sample 1 that contained 5 mM L-tryptophan with incubation at 25 °C, ThalE2R revealed 100 % conversion of the substrate after 44 h, whereas the WT reaction reached only 33 % (Figure 15). This result indicated a significantly improved catalytic activity of Thal-E2R at RT for the mentioned substrate concentration. By increasing the incubation temperature to 37 °C (sample 2), both enzymes equally converted 53 % of substrate after 52 h, pointing to a considerably retarded enzymatic activity at elevated temperatures. An increase of tryptophan concentration and subsequent incubation at 25 °C (sample 3) resulted in a strong difference of conversion, whereby the mutant showed a conversion of 47 % and the WT a conversion of merely 19 %. However, the 30 mL bromination mixture also containing 10 mM tryptophan revealed the same conversion of both the mutant and WT by reaching 95 % after 52 h probably due to a higher enzyme concentration because the volume ratio of employed lysate was higher in the 30 mL reaction. Simultaneous increase of substrate concentration and temperature led to the lowest Thal-E2R-catalyzed conversion of 29 % (sample 4). Thermal treatment of 76

the Thal containing lysate prior bromination reaction resulted in a doubled enzymatic activity for the mutant in comparison to the WT by reaching 97 % conversion after 44 h (sample 5, Figure 15). Notably, in contrast to WT thermal treatment at 49.1 °C had no negative impact on E2R and retained its activity. These finding underline the previously detected improved thermostability of Thal-E2R and emphasize its applicability for preparative biohalogenation. Sample 1: 5 mM L-Trp, 25 °C incubation

Sample 5: Heat-shock, 5 mM L-Trp, 25 °C incubation

Figure 15

Comparison of the bromination activity between Thal WT and E2R in crude lysate. The relative conversion (HPLC) was plotted versus reaction time. Selected reaction samples showed a remarkable difference in Thal-E2R- and Thal WT-catalyzed conversion of L-tryptophan.

77

To conclude, the highest catalytic activity of Thal-E2R was obtained in a small reaction volume, containing 5 mM L-tryptophan under incubation at 25 °C, whereby the heatshock did not lead to a negative impact on the halogenase. 4.4

Preparative synthesis of L-6-bromotryptophan using Thal-PrnF-ADH combiCLEAs

The preparative synthesis of L-6-bromotryptophan (24) was performed as combiCLEAS, thereby immobilizing overexpressed Thal-E2R from 1.5 L E. coli lysate in combination with the required auxiliary enzymes PrnF and RR-ADH for continuous cofactor regeneration. It should be examined whether the increased stability of mutant Thal-E2R has a beneficial impact on the synthesis of 24 in terms of activity and final conversion using the immobilized halogenase. The mentioned enzymes were precipitated by addition of ammonium sulfate, cross-linked with glutaraldehyde and employed for bromination of 1 mM L-tryptophan on 1 L scale. The reaction progress was monitored by RP-HPLC (Figure 16). Subsequently, the solid biocatalyst was removed by filtration and 24 purified by desalting and RP-HPLC.

Figure 16

Comparison of bromination activity between Thal WT- and Thal-E2R- PrnF-ADH combiCLEAs for the preparative synthesis of L-6-bromotryptophan (24). The relative conversion as determined by HPLC is plotted versus the time progress. The WT revealed full conversion after a reaction time of 12 d, whereas the mutant Thal-E2R reached a maximal conversion of 90 % after 20 d.

While the reaction proceeded the Thal-E2R-catalyzed bromination turnover increased steadily until reaching 88 % after 17 d, whereas further incubation did not result in a significant increase of product. Finally, after a reaction time of 20 d, the combiCLEAs revealed 90 % conversion of L-tryptophan. In parallel, the same preparative synthesis of L-6-bromotryptophan (24) was performed with Thal WT combiCLEAs. Beginning from the first day of reaction, this approach revealed a steep increase in conversion, leading unexpectedly to complete turnover after 12 d. 78

As no total conversion was achieved for Thal-E2R-catalyzed bromination, the reaction product was purified by desalting and RP-HPLC and obtained in a yield of 31 %, 87.2 mg. The identity of L-6-bromotryptophan (24) was confirmed by combined interpretation of one- and two-dimensional NMR spectra to confirm retained regioselectivity of Thal-E2R. Figure 17 shows an excerpt of the 1H-NMR spectrum, focusing on aromatic protons.

Figure 17

Excerpt of 1H-NMR (500 MHz, DMSO-d6) spectrum confirming product identity of L6-bromotryptophan (24) after combined interpretation with 1H-1H-ROESY and 1H-1HCOSY NMR spectra. Signal intensity is plotted versus the chemical shift in ppm referenced to solvent signal.

The C7-proton adjacent to the bromo substituent resonated as a doublet (7.57 ppm) due to 4 J coupling to C5 proton. This coupling and an additional 3J coupling (8.5 Hz) to C4H resulted in a doublet of doublets for C5 proton at 7.14 ppm. Correspondingly, C4H (7.52 ppm) could be assigned as a doublet with equal coupling constant as observed for adjacent C5H. The indole proton revealed a singlet at 11.24 ppm. Additionally, C2H signal at 7.26 ppm led to a doublet with weak coupling to the indole proton. Cα-proton (4.14 ppm) resulted in a triplet signal and accordingly two doublet of doublets appeared for the diastereotopic Cβ-protons (3.25 / 3.22 ppm). The clear distinction between 6-bromotryptophan and 5-bromotryptophan to confirm the mutant's regioselectivity that would reveal an almost identical coupling pattern was performed by 1H-1H-ROESY-NMR spectroscopy. The doublet of 4J coupling at 7.56 ppm could be assigned unambiguously to C7H due to dipolar coupling with the indole proton. In contrast, for bromination at C5 a doublet signal at C7 position would be expected with strong 3J coupling to adjacent C6. Accordingly, concerning 5-bromotryptophan indoleNH would give a cross signal to this doublet. Within the present spectra this expectation was not observed. In addition, further regioisomers were excluded, because coupling patterns of indole moiety were not in agreement.

79

5

Directed evolution of Thal: Discussion

5.1

Directed evolution of tryptophan 6-halogenase Thal

5.1.1

Successful optimization of epPCR

This project focused on the improvement of the thermostability of the tryptophan 6halogenase Thal and aimed for the development of a more thermo resistant variant by means of directed evolution. Therefore, the first and essential step was the generation of mutant libraries by introducing random mutations into the gene thal. Commonly employed techniques are e.g. epPCR, iterative saturation mutagenesis, DNA shuffling or the usage of specialized mutator strains.[39] As epPCR is a technique to introduce mutations all over the gene and simply requires the substitution of MgCl2 to MnCl2, thereby increasing error rate of the DNA polymerase, it was chosen as suitable tool for the generation of mutant libraries of thal. However, to obtain a high-quality mutant library, epPCR needed to be optimized with regards to the desired mutation rate. The adjustment of a suitable mutation frequency is the basis for a reliable mutant library and a factor that should not be neglected in generating high-quality mutants. In this project a mutation frequency of 2 mutations per kb was desired, as a lower mutation rate could have resulted in several colonies bearing no mutation at all due to statistical distribution. Even though a mutation frequency of 5-7 nucleotide mismatches per kb significantly reduces the wild type background and increases the possibility and amount of identified mutations, it becomes more difficult to identify those beneficial mutations that support a desired characteristic, as their positive impact might be compensated by another one with unfavorable properties. Although this effect might appear with only two mutations in one gene, it was proposed that identification of positive mutants is more successful when screening libraries with low mutant frequencies. Several approaches of epPCR with variations in MnCl2 concentration and amount of template DNA led to the desired mutation rate of 2 mutations per kb by employing 300 ng template DNA and 100 μM MnCl2. Increasing amounts of MnCl2 and reduction of template led to the expected higher mutation rate. As only two colonies of each epPCR approach were analyzed for their nucleotide mismatches, the obtained results varied

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_5

81

considerably and were therefore not absolutely representative. More reliable results would be obtained by sequencing larger amounts of samples and additional variation of the epPCR composition regarding template concentration, Mg2+, Mn2+ concentration and number of amplification cycles. To overwhelm possible codon bias and enhance library diversity, unbalanced mixtures of dNTPs could also be tested in further approaches. As these optimization steps are associated with high working effort and costs, as plasmid DNA needs to be isolated from each clone and its sequence has to be checked by Sanger sequencing, the performed strategy and obtained results were satisfying to reveal the optimal reaction conditions leading to the desired mutation rate. 5.1.2

Successful expression and purification of Thal WT for further characterization of its thermostability

In order to develop a thermostable Thal variant that should exhibit improved lifetime and catalytic activity, the properties of the WT catalyst were investigated in more detail with emphasis on its thermostability. To enable an efficient selection process towards a thermostable Thal mutant, the WT’s temperature threshold was determined at which the catalytic activity decreased to zero due to denaturation. To pursue this question, Thal WT was initially overexpressed in E. coli and purified via HisTALON affinity chromatography. The strain E. coli BL21 (DE3) pGro7 was chosen as suitable expression host for simultaneous coexpression of chaperones GroEL and GroES to assist Thal's correct native folding and increase its recovery as soluble and active protein.[50] The results confirmed that Thal WT was successfully overexpressed, purified and obtained in a sufficient concentration. Thal's activity profile was determined by incubating the purified enzyme at different temperatures and subsequent bromination of L-tryptophan conducted at 25 °C, indicating Thal's maximal temperature to be still sufficiently active. The experiment revealed a substantially decreased enzyme activity above 45.5 °C and showed no activity at all above 49.1 °C. Therefore, this threshold was chosen as selection factor for the identification of significantly more temperature-stable Thal mutants by directed evolution. In comparison to the temperature stability of tryptophan 7-halogenase RebH, previously examined by Poor et al., the enzyme showed significant denaturation around 42 °C. This was in good agreement to the results obtained for Thal, as both enzymes are related to each other.[42] 5.1.3

Identification of a thermostable Thal mutant by directed evolution

The first approach of directed evolution revealed three potential Thal mutants showing a 1.5-2.7-fold higher final conversion as indicated by the fluorescence emission than residual colonies, pointing towards an improved thermostability. Rescreening of theses outliers supported this assumption only for the Thal-E2 mutant, whereas the others did not show improved conversions in comparison to the WT. Theses colonies can be rated as false positive results probably caused by pipetting inaccuracies or evaporation effects in the 96-well plate. In contrast, Thal-E2 confirmed the assumption of an increased 82

thermostability by showing a 1.7-fold higher conversion subsequent to thermal treatment compared to the WT. Without thermal treatment this mutant even revealed a 2.7-fold higher bromination activity. These results corresponded to the expectations that increased temperatures might lead to partial denaturation, whereas Thal-E2 probably exhibited an increased catalyst lifetime at 25 °C. Further characterization of Thal-E2 by DNA sequencing revealed three mutations in the 1.6 kb gene thal, thereby confirming the aspired goal of 2 mutations per kb. The identified mutations resulted in a substitution of Ser-359 to Gly-359, Lys-374 to Arg-374 and a silent mutation of Lys-510. 5.1.4

Expression and purification of Thal-E2 led to low protein concentrations

For further examinations of Thal-E2's catalytic properties and comparison to the WT, Thal-E2 was overexpressed in E. coli and purified via HisTALON affinity chromatography. As indicated by SDS-PAGE, showing very faint bands in the elution fractions, the obtained protein concentration was 75 % less than for the WT. In search for a plausible explanation the focus was placed initially on the silent mutation of Thal-E2 at Lys-510. It was recognized that the mutation of the AAA codon against AAG resulted in a 68 % less frequently used triplet.[49] Although the genetic code is degenerated, meaning that more than one triplet may correspond to a specific amino acid, not all synonymous codons are used with equal frequency due to the rarity of certain complementary tRNAs. This biased usage may be explained by translational selection among synonymous base triplets or selection against preferred structures in DNA.[51] As the AAG codon encoding lysine has rare occurrence in E. coli, the silent mutation at Lys-510 was mutated reversely into the wild type sequence at this position, thereby striving for an increase of Thal-E2's protein expression. 5.1.5

Reverse mutation of Thal-E2 did not increase protein expression level

Contrary to the expectations, the reverse mutation of Thal-E2 to Thal-E2R, did not reveal a strong increase in protein concentration. Even though after successful expression of Thal-E2R in E. coli BL21 (DE3) pGro7 and subsequent purification SDS-PAGE analysis showed intensive bands corresponding to Thal-E2R's molecular weight, the resulting protein concentration could not be enhanced compared to Thal-E2. Probably the observed protein bands resembled the chaperone GroEL that had a similar weight as Thal-E2R. Therefore, codon exchange concerning Lys-510 did not lead to a notable increase of isolated protein subsequent to purification. 5.1.6

Usage of E. coli BL21 CodonPlus (DE3)-RIL did not increase Thal-E2R concentration

Another factor to be considered that might has contributed to the low expression level of the Thal mutants was the mutation at Arg-374, resulting in a codon that merely represented 4 % of the used codons for arginine.[49] It was therefore suggested, that the strain E. coli BL21 CodonPlus (DE3)-RIL would improve protein expression, as it 83

contained extra copies of genes that encode tRNAs that most frequently limit translation of heterologous proteins in E. coli. Besides codons for isoleucine and leucine, the encoded tRNA argU recognized the arginine codons AGA and AGG.[52] Consequently, it was assumed that the expression of Thal-E2R in this E. coli strain would result in significantly improved expression levels. However, this effect was not observed, as the protein concentrations remained as low as in previous purifications without additional tRNAs. Probably the additional level of tRNAs corresponding to the rare arginine codon AGA was still not sufficient to enhance protein expression. Another attempt to circumvent the rarely employed AGA codon for arginine would be site-directed mutagenesis to a more frequently used triplet such as CGU.[49] As Thal-E2R's further mutation Gly-359 resulted in an abundantly used codon for glycine, this mutation was not considered to have a negative effect on protein level. 5.1.7

Thal mutants E2 and E2R retained high enzymatic activity despite thermal treatment

The thermal treatment of purified Thal-WT, -E2 and -E2R at different temperatures and subsequent utilization for bromination of L-tryptophan supported the proposed increased thermostability of the found mutants, as they revealed significantly elevated catalytic activities in comparison to the WT. Thal-E2 and E2R steadily denaturated, but showed an increased catalytic activity over a longer temperature range, whereas the WT's activity decreased directly above 45.5 °C and became inactive above 49 °C. Noteworthy, E2 and E2R activity dropped to approximately zero only above 58 °C corresponding to a considerable elevation of enzyme stability of 9 °C. As both mutants, E2 and E2R, obtained the identical mutations on amino acid level, comparable results regarding their bromination activity were observed as expected. ThalE2R's slightly higher catalytic activity may be traced back to the fact that the obtained low protein concentrations lied within an area that could not be determined with sufficient accuracy. 5.1.8

Further examinations of Thal-E2R were performed in E. coli lysate

Due to the small amount of purified protein in areas of 0.2-0.5 mg mL-1, it was more applicable to conduct further examinations of Thal-E2R in lysate. In chemoenzymatic purposes this strategy is commonly established to perform biocatalytic reaction approaches and further loss of enzyme could be avoided. On the other hand, the direct comparison between reactions performed in lysate is more difficult as Thal's protein concentration cannot be determined in the lysate due to the presence of plenty other E. coli proteins. As protein concentration dramatically decreases during dialysis, this step could be evaded by performing the cell lysis and protein purification with buffers that do not contain NaCl but NaBr instead.

84

5.1.9

Thal-E2R revealed significantly improved catalytic activity and no negative impact of thermal treatment

Thal-E2R's catalytic activity was further investigated in lysate by choosing different substrate concentrations, incubation temperatures and reaction volumes. Moreover, the impact of thermal treatment was examined in detail. The mutant showed full conversion of 5 mM L-tryptophan, but could not completely convert the twice amount of substrate with the same efficiency. Besides a too low enzyme concentration product inhibition may be considered as one reason to impede the bromination in presence of high substrate loading. Notably, the mutant confirmed a significantly improved catalytic activity not being effected by thermal treatment at 49.1 °C in comparison to the WT: Considering the assumed lower protein concentration of the mutant its catalytic activity is proposed to be significantly higher than observed in this experiment. However, the elevated enzymatic activity of the WT after heat-shock compared to the same reaction approach without thermal treatment was striking. Yet, a reason that might contribute to the observed increasing conversion rate could be the heat inactivation of degrading enzymes, e.g. proteases, thus stabilizing Thal and preventing its degradation. An increase of reaction temperature to 37 °C resulted for both, Thal WT and Thal-E2R in considerably decreased enzymatic activities, possibly caused by degradation of the insufficient heat-resistant auxiliary enzymes as well as cofactors required for bromination. To sum up, Thal-E2R exhibited improved properties for chemoenzymatic halogenation. Probably an increased lifetime contributed to an elevated total turnover of tryptophan especially at 25 °C. 5.1.10

Thal-E2R-PrnF-ADH combiCLEAS allowed preparative synthesis of L-6-bromotryptophan

As previous results emphasized Thal-E2R's improved stability on an analytical scale, preparative scale synthesis of L-6-bromotryptophan 26 was performed as Thal-E2R-PrnFADH combiCLEAs that revealed 90 % conversion of L-tryptophan after a reaction time of 20 d. The steady increase of product during this long reaction time pointed at Thal-E2R's prolonged overall stability, as such a reaction time was even for an immobilized enzyme remarkably high. Increasing degradation of auxiliary enzymes and cofactors probably prevented the mixture from reaching full conversion, especially because E. coli lysate also contained several other enzymes that might participate in degradation of reaction components. Unexpectedly, the desired improved catalytic activity of Thal-E2R in comparison to the WT could not be observed on preparative scale, as the WT already reached full conversion after 12 d. A probable reason for this faster conversion was the significantly higher protein concentration for Thal WT, resulting in a higher amount of immobilized catalyst. In order to quantify the amount of overexpressed halogenase, a quantitative Western Blot could be performed. Even though the obtained contradictory results with combiCLEAs were not expected, Thal-E2R's catalytic lifetime was sufficiently high to lead to an acceptable overall conversion even with high substrate loading of 204 mg L-1 L-tryptophan. This approach 85

confirmed the applicability of combiCLEAs for Thal-E2R obtaining satisfying amounts of product 26. Still further optimization steps, e.g. regarding expression levels, could improve the total conversion and a subsequently facilitated workup procedure to pave the way towards an improved tryptophan 6-halogenase variant for preparative purposes. 5.1.11

Identified mutations influenced Thal-E2R's thermostability

Investigations on Thal-E2R's amino acid sequence revealed the substitution of Ser-359 against Gly-359 and the mutation of Lys-374 against Arg-374. Both substituted amino acids are proposed to contribute to the halogenase's thermostability. The thermal denaturation of proteins is associated with drastic changes of inter- and intramolecular interactions concerning the disruption and new formation of ionic-, polar-, hydrophobic- or hydrogen bonds leading to disorders in secondary structure and a loss of catalytic functions. An increase of enzyme stability is often accompanied by rigidifying protein conformation especially on the surface to stabilize the overall structure when exposed to higher temperatures.[53] During the generation of a thermostable RebH variant similar motifs were proposed by Poor et al. in 2014. It was assumed that beneficial mutations on the protein surface especially contribute to increased thermostability. In addition, it was supposed that increased rigidity of the protein, e.g. due to a proline residue, might stabilize the protein and impede thermal unfolding. However, the increased rigidity might hinder the enzyme's activity.[42] In order to estimate the effect of the altered amino acid residues on Thal-E2R's thermostability, it was helpful to reveal the residue's location and orientation within the protein scaffold and its interaction with surrounding residues. As no crystal structure was available at present, a homology model of Thal-E2R was calculated using “SwissModel”.[54] The tryptophan 7-halogenase RebH served as template for modeling due to its highest homology to Thal as predicted by the database.[25] This gave a rough clue about the protein structure and allowed some speculations about the thermostability. This model revealed that Arg-374 is surface exposed and directly located at the proposed contact area between the two Thal monomers (Figure 18).[54] Arginine might strengthen the interaction between these monomers, as it tends to the formation of salt bridges, in comparison to lysine with weaker basicity. This exchange probably contributes to the dimer's stability as temperature rises as well as leading to a prolonged catalytic activity before denaturation. Additionally, examinations on thermophilic proteins revealed that arginine is highly abundant in these proteins and contributes towards protein stability, as its large side chain may be able to take part in both long and short range interactions.[53] Because the proposed structure model suggests that Arg-374 is part of an α-helix, this feature could additionally stabilize the protein's secondary structure that contributes to Thal- E2R's enhanced thermostability. Regarding Gly-359 this amino acid residue is located in a cavity and probably not directly exposed to the protein's surface. Despite its function could not be unambiguously detected

86

this substitution against serine is proposed to be involved in Thal-E2R's thermostability as serine is rarely detected in thermophilic proteins.[53]

Figure 18

Excerpt of Thal-E2R's proposed structure shown as monomer based on the crystal structure of RebH (PDB: 2OA1). The amino acid residues Arg-374 and Gly-359, which are mutated in comparison to the WT are highlighted as red stick model. The remaining amino acid residues are shown as blue ribbon model (structure was generated with Swiss-Model).[54]

Even though the obtained structure of Thal-E2R is only a proposed model and not absolutely depicting reality, the mentioned observations support Thal-E2R's increased thermostability. However, to gain profound insight into the effects of the mutations on the thermal stability of Thal, crystal structures for both WT and E2R are required. They allow more precise speculations on molecular level, how these residues influence protein conformation.

87

6

Establishment of a marine brominase: Results

6.1

Establishment of the marine brominase Bmp5

6.1.1

Subcloning of bmp5 into pET28a expression vector and transformation into E. coli DH5α

The gene bmp5 was obtained as codon-optimized sequence deduced from the amino acid sequence of Bmp5 from Pseudoalteromonas luteoviolacea for overexpression in E. coli. The nucleotide sequence was flanked by NdeI and BamHI restriction sites enabling subcloning into an equally digested pET28a vector for coexpression as N-terminal His6tagged fusion protein. Therefore, the obtained pMK-RQ-bmp5 vector was transformed via heat-shock into E. coli DH5α, plasmid DNA was isolated and the bmp5 insert restricted by NdeI and BamHI digestion. The separation of 1.6 kb bmp5 insert from residual pMKRQ vector backbone with a size of 2.3 kb was performed by agarose gel electrophoresis (Figure 19). To avoid mutations in the bmp5 gene introduced by ultraviolet irradiation only a small amount of digested DNA was stained with ethidium bromide for UV detection. The position of the corresponding DNA band in the agarose gel was detected with UV irradiation and served as template for the remaining DNA samples that were excised from gel without previous staining and visualization.

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_6

89

Figure 19

Separation of bmp5 from residual pMK-RQ vector by agarose gel electrophoresis. The insert was excised from pMK-RQ vector by restriction digest with NdeI and BamHI and subsequently separated from residual vector backbone on 1 % agarose gel and stained via ethidium bromide.

As the expected bands corresponding to digested vector (2.3 kb) and insert (1.6 kb) were observed and confirmed successful restriction digest, bmp5 was cut out of the agarose gel and subsequently purified via gel extraction kit. After the target vector pET28a was digested with the same restriction enzymes NdeI and BamHI, both components were ligated and resulting pET28a-bmp5 transformed via heat-shock into E. coli DH5α with primary selection towards kanamycin resistance. Positive clones containing the pET28abmp5 plasmid were identified by colony-PCR, as they revealed the expected DNA band around 1.9 kb (Figure 20). The PCR product was larger than bmp5 due to flanking regions at both termini that were additionally amplified using T7 promoter and terminator as primers. Plasmid DNA was isolated fro m three positive clones and the correct bmp5 sequence of transformant no. 4 confirmed by Sanger sequencing.

Figure 20

90

Identification of bmp5 containing colonies via colony-PCR. 14 clones were picked and screened for the presence of the bmp5 insert by employing T7P and T7T primer. Positive clones were identified by the expected DNA band at 1.9 kb. pET28a-rebH plasmid with similar size served as positive control for success of PCR. The DNA samples were separated on 1 % agarose gel and stained via ethidium bromide.

6.1.2

Expression of Bmp5 in E. coli BL21 (DE3) pGro7

To study enzyme activity of the halogenase Bmp5 it was desired to overexpress and purify the enzyme for subsequent biocatalytic reaction approaches. During the initial attempt of Bmp5 expression in E. coli no data were available whether the coexpression of chaperones GroEL / ES would enhance proper folding of His6-Bmp5, therefore the pET28a-bmp5 plasmid was initially transformed via heat-shock into the strain E. coli BL21 (DE3) pGro7. As previous overexpression of other halogenases such as Thal in E. coli BL21 (DE3) pGro7 revealed satisfying protein concentrations, this E. coli strain was employed for first expression attempts for Bmp5. 6.1.3

Protein purification via HisTALON affinity chromatography

His6-Bmp5 was purified from 1.5 L E. coli lysate via HisTALON affinity chromatography and therefore initially lysed by French press. Success of purification and purity grade was checked by separation of the resulting fractions by SDS-PAGE (Figure 21).

Figure 21

SDS-PAGE analysis of Bmp5 expression in E. coli BL21 (DE3) pGro7 pET28a-bmp5. The cleared lysate was separated in a 12 % polyacrylamide gel along with the collected flow-through, washing fraction 1, washing fraction 2 and elution fractions after HisTALON affinity chromatography (10 μL of sample was applied in each lane). The cleared lysate, FT and WF1 revealed very intense bands around 60 kDa, corresponding to the molecular weight of His6-tagged Bmp5 and the chaperone GroEL.

The separation of the cleared lysate resulted in several bands differing in molecular weight, referring to various E. coli proteins. In addition, a very dominant but blurred band could be observed between the 46 and 58 kDa protein standard. As this lane contained too much protein, the precise assignment of bands to their corresponding weight was difficult. Still, this intensive band probably corresponded to the molecular weight of the overexpressed protein components Bmp5 as well as GroEL. This band was observed in all further fractions, whereas it was most dominant in the flow-through and first washing 91

fraction. Nevertheless, the elution fractions showed the desired band at 61 kDa. However, Bmp5-containing elution fractions contained a very low protein concentration of 1 mg mL-1. Eventually the majority of Bmp5 could not bind via its His6-tag to the matrix and therefore it was mainly present in the flow-through and washing fractions as indicated by the dominant protein bands. In order to assign unambiguously the obtained protein band at 61 kDa to overexpressed Bmp5 and not to the chaperone GroEL that has a similar weight, the expression of Bmp5 was repeated in the strain E. coli BL21 (DE3) in absence of chaperones. 6.1.4

Bmp5 expression in E. coli BL21 (DE3)

According to the expectations a protein band around 60 kDa was not observed prior induction, instead several bands of major and minor size appeared. Apart from broadly distributed protein bands the cleared lysate revealed an intensive band at 60 kDa, probably corresponding to Bmp5's molecular weight that was overexpressed successfully. This band was also clearly visible in the flow-through and first washing fraction, indicating that His6-Bmp5 did not bind efficiently enough to the matrix. This assumption was supported by the lack of a dominant 60 kDa band in the elution fractions that merely showed several differently sized bands of lower molecular weight. As indicated by SDSPAGE the protein concentration of combined elution fractions containing His6-Bmp5 revealed a protein concentration of less than 1 mg mL-1.

Figure 22

92

SDS-PAGE analysis of Bmp5 expression in E. coli BL21 (DE3) pET28a-bmp5. The cleared lysate as well as a sample prior induction were separated in a 12 % polyacrylamide gel along with the collected flow-through, washing fraction 1, washing fraction 2 and collected elution fractions during HisTALON affinity chromatography (10 μL for WF1, WF2 and elution fractions, 5 μL for remaining fractions). The expected molecular weight of 61 kDa for Bmp5 could not be observed in the elution fractions, but was present in the cleared lysate, FT and WF1.

6.2

Identification of Bmp5

6.2.1

Detection of His6-tagged Bmp5 by Western blot analysis

Western blot analysis of recombinant N-terminal His6-tagged Bmp5 was performed to prove successful expression in E. coli BL21 (DE3) pET28a-bmp5 crude lysate due to challenging purification by metal ion affinity chromatography. Different amounts of lysate were initially separated by SDS-PAGE and His6-Bmp5 subsequently detected by incubation with specific anti-His6 primary and HRP-conjugated secondary antibody.

Figure 23

Detection of His6-Bmp5 by Western blot analysis of E. coli BL21 (DE3) pET28abmp5 crude lysate employing anti-His6 and HRP-conjugated antibodies. Presence of Bmp5 was proven by a protein band around 60 kDa. Two additional bands below 58 kDa as well as 46 kDa referred to degraded His6-tagged protein.

The blot revealed a very dominant but blurred band between the 58 and 80 kDa protein standard, probably corresponding to His6-Bmp5 with an expected mass of 61 kDa. Different lysate volumes were loaded, because direct determination of Bmp5 concentration was not possible in lysate. As expected the intensity of observed bands increased with rising amount of applied sample, even though protein loading was too high on the polyacrylamide gel. Assumedly, His6-Bmp5 degradation products were detected in each protein sample by two additional bands around 50 and 40 kDa. 6.2.2

Identification of Bmp5 by peptide mass fingerprint

Even though His6-Bmp5's presence was confirmed by Western blot analysis, the halogenase was additionally identified by peptide mass fingerprint subsequent to tryptic digestion and MALDI-ToF mass spectrometry analysis. Therefore, Bmp5's expected protein band at 61 kDa was excised from the SDS gel, digested with the protease trypsin and the resulting characteristic peptide fragments were analyzed by MALDI-ToF MS. In

93

silico digestion of the His6-bmp5 sequence resulted in a list of theoretical peptides that were compared with acquired data, thereby revealing a sequence coverage of 41 %. 6.3

Enzymatic bromination of 4-HBA after purification of His6-Bmp5

After purification of Bmp5 via HisTALON affinity chromatography from 1.5 L expression culture of E. coli BL21 (DE3) pGro7 SDS-PAGE analysis indicated that His6Bmp5 could not be purified successfully resulting in a low overall protein concentration after elution. As His6-Bmp5 did not bind efficiently enough to the matrix a significant high enzyme concentration was expected to occur in the flow-through. To further examine Bmp5 activity, collected samples from HisTALON purification were taken from the combined elution fractions and the flow-through. The first reaction approach was conducted on a 200 μL scale without additional cofactor regeneration system (Scheme 7). The regeneration of the essential cofactor FADH2 was assumed to be performed by Bmp5 that internally reduces FAD by oxidizing NADPH. The reducing agent was added in 2.5 fold excess referred to the starting material 19.

Scheme 7

To convert 19 into final product 23, Bmp5 also regenerates the essential cofactor FADH2 by oxidizing NADPH to NADP+.

The conversion of 19 was monitored by RP-HPLC and the reaction mixture containing Bmp5 present in the flow-through revealed 17 % decrease of starting material after a reaction time of 1 h. After 24 h 50 % conversion of 19 was achieved, however the reaction did not proceed further. The intermediate 3-bromo-4-hydroxybenzoic acid (21) and the product 2,4-dibromophenol (23) (Scheme 8) could be assigned by comparison of HPLC elution signals with authentic standards. Small traces of 4-bromophenol (27) and 2,4,6-tribromophenol (28) were additionally observed. The same reaction approach performed with Bmp5 from the elution fractions barely showed any conversion of 19 and therefore supported the assumption that Bmp5 could not be purified successfully by HisTALON chromatography, but was only present in the flow-through in which activity of the brominase could be confirmed.

94

Scheme 8

Bmp5-catalyzed conversion of 19 is proposed to lead via intermediate 21 to formation of product 23.

In order to examine Bmp5's catalytic activity in presence of an additional cofactor regeneration system, FADH2 was directly supplied by the flavin reductase PrnF and NADH regenerated by RR-ADH (Figure 24). This approach should examine whether the intrinsic function of Bmp5 as flavin reductase can be simply circumvented by external supply of FADH2 as established for tryptophan halogenases.

Figure 24

Approach for Bmp5-catalyzed bromination of 19 to 23 with cofactor regeneration using an external flavin reductase, PrnF, as well as an alcohol dehydrogenase from Rhodococcus spp. to supply NADH by oxidation of iso-propanol.

The reaction approach containing Bmp5 from the flow-through revealed 9 % conversion of 19 after one hour and did not continue further, indicating that an additional flavin reductase in combination with ADH is inappropriate for continuous cofactor regeneration in Bmp5-catalyzed bromination. The same reaction performed with an enzyme sample from the elution fractions did not show any conversion at all as observed previously. To support the assumption that His6-Bmp5 could not be sufficiently purified via HisTALON affinity chromatography, the same reaction approach was repeated with Bmp5 expressed in E. coli BL21 (DE3), lacking chaperones (6.1.4). The product 23, intermediate 21 as well as the mentioned side products were observed as expected in reaction approaches employing Bmp5 from the flow-through. However, the obtained protein concentrations in the elution fractions were again too low to continue reliable 95

experiments with His6-Bmp5. Therefore, different purification approaches were considered, leading to the strategy of purifying Bmp5 via GST-tag. 6.4

Generation of GST-tagged Bmp5

6.4.1

Subcloning of bmp5 into pETM30 vector

In order to facilitate expression of fusion protein GST-Bmp5, the gene bmp5 was subcloned from the pET28a vector into the pETM30 vector, encoding the sequence for an N-terminal GST-tag. As the obtained pETM30 vector still included a foreign insert, it was restricted by NcoI and NotI digestion and the vector backbone separated from the undesired insert by agarose gel electrophoresis (Figure 25). The 6.0 kb DNA band corresponding to the residual pETM30 vector was excised from the agarose gel and purified via gel extraction kit.

Figure 25

Separation of the 1 kb insert from desired 6 kb pETM30 vector by digestion with NcoI and NotI. Both components were separated on 1 % agarose gel and stained via ethidium bromide.

In order to enable direct subcloning of the bmp5 insert into the NcoI/NotI digested pETM30 vector, bmp5 had to be flanked by identical restriction sites. The required recognition sequences were introduced by PCR during amplification of bmp5 employing NcoI-bmp5-FW and NotI-bmp5-REV primer that encoded the corresponding restriction site for either NcoI or NotI. Purified pMK-RQ-bmp5 plasmid served as template for the PCR that was performed in two separate approaches, containing either Phusion HF or GC buffer. Both reactions revealed the successful amplification of 1.6 kb bmp5 by agarose gel electrophoresis and the PCR product amplified in HF buffer was chosen for following steps. To create the required NcoI/NotI sticky ends the NcoI-bmp5-NotI amplificate was digested with the corresponding restriction enzymes and DpnI was added to remove the 96

methylated template plasmid. Subsequently the digested bmp5 insert was ligated with pETM30 vector and transformed via heat-shock into E. coli DH5α with primary selection towards kanamycin resistance. 6.4.1.1

Identification of positive pETM30-GST-bmp5 clones by colony-PCR

After successful transformation of pETM30-bmp5 into E. coli DH5α positive clones were identified by colony-PCR showing the expected band at 2.3 kb corresponding to the sequence of GST-bmp5. As the presence of an additional DNA band with a size of 3.3 kb was unexpected, plasmid DNA was isolated from two positive clones and the presence of bmp5 insert additionally confirmed by NcoI/NotI digestion. The 1.6 kb insert was separated from residual vector backbone by agarose gel electrophoresis and clarified the presence of bmp5 in these colonies. The unexpected 3.3 kb insert might be formed by unspecific primer annealing. As transformants no. 2 and 9 contained the desired bmp5 insert their sequence was additionally confirmed by Sanger sequencing.

Figure 26

6.4.2

Identification of GST-bmp5 containing colonies via colony-PCR. Positive transformants were identified by the 2.3 kb band employing the T7P and T7T primer. pET28a-bmp5 plasmid served as positive control to confirm PCR success. DNA samples were separated on 1% agarose gel and stained via ethidium bromide.

Expression and purification of GST-Bmp5

For subsequent overexpression and purification of GST-Bmp5, pETM30-bmp5 was transformed via heat-shock into E. coli BL21 (DE3). After cell harvest and lysis of 1.5 L E. coli culture GST-Bmp5 should be purified via GST-affinity chromatography. Success of purification and purity grade were examined by separation of the resulting fractions by SDS-PAGE. A sample prior induction, cleared lysate as well as components of the E. coli BL21 (DE3) pETM30-bmp5 cell debris were analyzed in addition (Figure 27).

97

Figure 27

SDS-PAGE analysis of GST-Bmp5 expression in E. coli BL21 (DE3) pETM30-bmp5 after cell disruption and purification via GST-affinity chromatography. The flowthrough, three washing- and elution fractions were collected and separated in a 12 % polyacrylamide gel. Additionally, a sample prior induction, the cleared lysate as well as the cell pellet resolubilized in urea were analyzed (10 μL for elution fractions, 5 μL for remaining fractions). The expected band at 85 kDa, corresponding to GST-Bmp5 could not be observed in the elution fractions, but a similar sized band was obtained in cell debris pointing to the formation of inclusion bodies. Unexpectedly, a protein around 27 kDa was observed in the elution fractions.

As expected the sample prior induction revealed several differently sized bands referring to a variety of E. coli proteins. A similar distribution of protein bands was observed in the cleared lysate and flow-through, whereas these samples lacked the expected band at 85 kDa, referring to overexpressed GST-Bmp5. The washing fractions revealed a decreasing amount of differently sized faint bands. Unexpectedly, the elution fractions did not show the desired 85 kDa band but instead an intensive protein around 27 kDa was visible. In contrast, the analysis of the solubilized components of the cell debris revealed numerous bands of minor size. In addition, an intensive band in the range of 80 kDa was recognized, probably corresponding to GST-Bmp5. This observation and the lack of a GST-Bmp5 corresponding band in the elution fractions suggested the aggregation of GST-Bmp5 as inclusion bodies that formed within the insoluble protein fraction. Determination of protein concentration of the combined elution fractions revealed a value of 0.7 mg mL-1, confirming that no soluble GST-Bmp5 was isolated and probably accumulated as insoluble aggregates. 6.4.3

Attempts for enzymatic bromination of 4-HBA after purification of GST-Bmp5

In accordance with the results of expression analysis the isolated elution fractions as well as the flow-through of GST-affinity chromatography showed no significant halogenase 98

activity for GST-Bmp5 as evidenced in 200 μL scale reactions employing 4-HBA (19) as substrate. These data confirmed that expression of GST-tagged Bmp5 was not suitable to obtain an active brominase. Additionally, the catalytic activity was further investigated directly in E. coli BL21 (DE3) pETM30-bmp5 crude lysate. As expected this reaction approach did not reveal notable conversion of 19 after 24 h. 6.5

Establishment of cofactor regeneration for Bmp5 in E. coli lysate

6.5.1

Investigation of glucose dehydrogenase

Due to inactivity of GST-Bmp5 focus was returned on His6-Bmp5 as this fusion protein already revealed to successfully convert 4-HBA (19) to the final product 23. As Bmp5 is able to reduce FAD by oxidizing NADPH with its intrinsic flavin reductase activity, a new cofactor regeneration system was tested to enable the continuous supply of NADPH. Therefore, the GDH from Bacillus megaterium was employed to oxidize D-glucose to + D-glucono-1,5-lactone and thereby catalyzes the reduction of NADP to NADPH [55–57] (Scheme 9). The following hydrolysis of D-glucono-1,5-lactone to gluconic acid was considered as positive effect, as the removal of glucono-1,5-lactone would shift the equilibrium in favor of product formation.

Scheme 9

Bmp5-catalyzed conversion of 19 to 23 employing GDH for regeneration of NADPH by oxidation of D-glucose.

The enzymatic bromination of 4-HBA was directly performed in E. coli BL21 (DE3) pET28a-bmp5 crude lysate on 30 mL scale. After a reaction time of 20 h RP-HPLC analysis revealed 15 % conversion of starting material (19), confirming the general functionality of GDH for the regeneration of NADPH. However, further incubation did not lead to increasing conversion of 19, as indicated by merely 16 % conversion after 26 h. In addition, the pH value was determined to 4.9 after that duration. The accumulation of gluconic acid due to GDH activity resulted in an acidification of the reaction mixture and subsequent denaturation of enzyme components. The declining pH value was considered as reason for the decreasing enzymatic activity of Bmp5. Thus, a suitable alcohol dehydrogenase was tested in another attempt for cofactor regeneration to enable preparative enzymatic bromination with Bmp5. 99

6.5.2

Optimization of cofactor regeneration in Bmp5-catalyzed bromination employing alcohol dehydrogenase

In order to examine catalytic activity of Bmp5 in E. coli lysate, two different approaches were performed. The first approach, employing an excess of NADPH, intended to examine the general activity of Bmp5 in E. coli BL21 (DE3) pET28a-bmp5 crude lysate, as previous experiments already revealed the brominase's activity when supplying an excess of NADPH. After 22 h this reaction approach revealed 60 % conversion of the starting material 19. The presence of intermediate 21, 4-bromophenol (27) as well as the final product 2,4-dibromophenol (23) could be initially assigned by comparison of RPHPLC elution signals with authentic standards. As these observations confirmed His6Bmp5's bromination activity in crude lysate, another cofactor regeneration system was tested striving for full conversion of 19 in context of preparative purposes and avoiding the less efficient supply with excess of NADPH. In this reaction approach NADP+dependent alcohol dehydrogenase LK-ADH from Lactobacillus kefir was employed as additional enzyme component, as this oxidoreductase catalyzes the reduction of NADP+ to NADPH by oxidizing iso-propanol (Scheme 10)[58]. Therefore, LK-ADH was overexpressed on a 1.5 L scale in E. coli BL21 (DE3) pET21LK-ADH, the cells were harvested and subsequently lysed by French press. LK-ADH activity was determined in crude lysate and directly employed for cofactor regeneration.

Scheme 10

Approach for Bmp5-catalyzed bromination of 19 to 23 in presence of continuous cofactor regeneration using an external alcohol dehydrogenase from Lactobacillus kefir to supply NADPH by oxidation of iso-propanol.

The reaction was performed in E. coli BL21 (DE3) pET28a-bmp5 crude lysate on 30 mL scale and the conversion of 2 mM 19 (tr=116 s) monitored by RP-HPLC for 22 h (Figure 28). After a reaction time of 45 min the formation of intermediate 21 (tr= 158 s) as well as a minor amount of final product 23 (tr= 225 s) was detected. The total conversion of starting component 19 and intermediate 21 was reached after 2 h. Further reaction time revealed the additional formation of 28 (tr= 255 s) after 22 h, whereas a signal corresponding to 4-bromophenol (27, tr= 197 s) could be observed right from the 100

beginning. The detection of intermediate 21 at the start of the reaction indicated that the conversion of 19 proceeded rapidly.

Figure 28

6.5.3

Progress of Bmp5-catalyzed conversion of 19 via RP-HPLC at 220 nm in E. coli lysate employing LK-ADH for NADPH regeneration. The corresponding brominated products were identified by authentic standards and GC-MS as depicted below. The time-dependent progress revealed the conversion of 19 to the final product 23 via intermediate 21. The presence of 27 and formation of 28 as possible byproducts was observed in addition.

Identification of reaction components by GC-MS analysis

As previously described the Bmp5-catalyzed conversion of 19 was performed in two different reaction approaches regarding cofactor regeneration. Initially, the assignment of RP-HPLC elution signals was performed by comparison with authentic standards, but the identity of reaction components was clarified by GC-MS analysis. Therefore, phenol derivatives were extracted with ethyl acetate from the acidified reaction solution for GCMS. As the first reaction approach, employing an excess of NADPH, did not result in full conversion of 19, this sample enabled the identification of starting material 19 and 101

intermediate 21 by comparison of calculated with observed masses. In addition, the typical isotopic pattern helped to identify brominated compounds. Figure 29 shows a representative mass spectrum for 21. The calculated mass corresponded to the measured m/z value, whereas the deviation of more than m/z 1 resulted from calibration inaccuracies. Still, the typical isotopic distribution for one bromine substituent could be identified as 79Br/81Br, 1:1. The mass spectra also revealed typical fragmentation pattern of 19 and 21 due to a loss of hydroxyl radical, thereby allowing the explicit assignment of both reaction components.

Figure 29

GC-MS analysis of Bmp5-catalyzed conversion of 19 to 23 via intermediate 21. The EI mass spectrum confirmed successful formation of 21 as indicated by the isotopic pattern of monobrominated compounds. (tr = 20.3 min / positive mode, m/z = [M+∙(79Br)] obs. 217.0 calc. 215.9; [M+∙(81Br)] obs. 219.0 calc. 217.9).

The second approach regarding Bmp5-catalyzed conversion of 19 was performed with LK-ADH for regeneration of the reducing agent NADPH. After a reaction time of 22 h product 23 as well as components 27 and 28 could be detected by GC-MS analysis. GC separation of the reaction crude extract revealed a retention time of 14.3 min for product 23 and its identity was confirmed by observed masses corresponding to calculated values (Figure 30). In addition, the mass spectrum showed the typical isotopic pattern of a dibrominated compound (1:2:1).

102

Figure 30

GC-MS analysis of Bmp5-catalyzed conversion of 19 to 2,4-dibromophenol (23). The EI mass spectrum proved successful formation of 23 as indicated by the isotopic pattern of dibrominated compounds. (tr = 14.3 min / positive mode, m/z = ([M+∙(79Br/79Br)] obs. 251.8 calc. 249.9; [M+∙(79Br/81Br)] obs. 253.3 calc. 251.9; [M+∙(81Br/81Br)] obs. 255.2 calc. 253.9).

Apart from these expected reaction components, byproduct 27 and 28 could be unambiguously detected.

(a)

103

(b)

Figure 31

GS-MS identification of side products formed during Bmp5-catalyzed bromination. (a) 4-bromophenol (27) (tr = 20.9 min / positive mode, m/z = ([M+∙(79Br)] obs. 171.7 calc. 171.9; [M+∙(81Br)] obs. 173.7 calc. 173.9). (b) 2,4,6-tribromophenol (28) (tr = 21.2 min / positive mode, m/z = ([M+∙(79Br/79Br/79Br)] obs. 329.7 calc. 327.8; [M+∙(79Br/79Br/81Br)] obs. 331.8 calc. 329.8; [M+∙(79Br /81Br/81Br)] obs. 333.5 calc. 331.8, [M+∙(81Br /81Br/81Br)] obs. 335.5 calc. 333.8).

Employing LK-ADH for cofactor regeneration enabled detailed study of Bmp5 activity. Full conversion of 19 was achieved leading to intermediate 21 as well as final product 23. Moreover, the occurrence of mono- and tribrominated phenol derivatives as by-products was observed by means of this approach. 6.5.4

Study of Bmp5’s bromination ability in presence of chloride

Bmp5's proposed ability to catalyze the bromination of its substrate 19 even in presence of NaCl was investigated in E. coli BL21 (DE3) pET28a-bmp5 lysate. Five separate bromination mixtures, each containing different concentrations of NaCl together with 30 mM NaBr were prepared and the reaction progress was monitored for 24 h by RPHPLC. The observed elution signals corresponded to known brominated reaction products, indicating that Bmp5 accepted bromide as co-substrate rather than chloride. In order to confirm this assumption, selected samples were analyzed by GC-MS. Even though not all measurements were successful, analysis of the reaction mixture containing the highest concentration of 300 mM NaCl that corresponded to a tenfold chloride excess led to the identification of intermediate 21 as well as 27. As no chlorinated compounds could be identified in the analyzed samples, Bmp5 was probably able to continue bromination of 19 even in presence of a large excess of chloride ions.

104

6.5.5

Examination of iodination ability

As Bmp5 was proposed to catalyze the formation of iodophenols in vivo[8], this assumption was examined in vitro with 4-HBA (19) as substrate in E. coli lysate. The reaction progress was monitored for 20 h by RP-HPLC, whereby no conversion of starting material was observed. Therefore, no iodinated reaction compounds could be detected by GC-MS analysis. The iodination attempt of 4-HBA was additionally repeated on a preparative scale using Bmp5-ADH combiCLEAs. As before this approach revealed no conversion of starting material after a reaction time of 40 h, thereby supporting the assumption that Bmp5 is not able to catalyze the iodination of 19 in vitro. 6.6

Investigation of Bmp5's substrate scope

Since the Bmp5-catalyzed bromination of 4-HBA (19) proved to be successful on an analytical scale, further potential substrates were screened in order to analyze Bmp5's substrate scope. The reactions were performed in E. coli lysate containing overexpressed His6-Bmp5 with combination of LK-ADH for continuous cofactor regeneration. The reactions, performed on 6 mL scale, were incubated for 24 h at 25 °C and the reaction progress was monitored by RP-HPLC. The following table summarizes whether the potential substrates revealed a conversion as indicated by RP-HPLC measurement and if these brominated substrates could be detected by GC-MS analysis. Table 7

Listing of potential substrates for Bmp5 and their observed conversion by RP-HPLC. Legend: 3: RP-HPLC spectra gave an indication that bromination of substrate might occur. 2 RP-HPLC spectra gave no indication for bromination.

potential substrate 4-HBA Phenol Benzoic acid Aniline Indole Tyrosine Phenylalanine

indication for bromination in RP-HPLC chromatogram 3 3 2 2 2 2 2

observed bromination in GC-MS spectrum 3 3 2 2 2 2 2

The conversion of 4-HBA (19) by Bmp5 was observed as expected and served as positive control, as it proved the enzyme's catalytic activity. RP-HPLC analysis revealed 14 % conversion of phenol within the reaction mixture indicating that this substrate might be converted by Bmp5. Therefore, the crude mixture was further examined by GC-MS confirming the presence of bromophenol as well as dibromophenol as proven by m/z values. In addition, both components revealed the typical isotopic distribution for a monoor dibrominated compound (Figure 32). 105

As already indicated by RP-HPLC measurement the remaining substrates did not reveal brominated compounds or fragment ions in the mass spectra, indicating that these substrates were not accepted by Bmp5. (a)

(b)

Figure 32

106

GC-MS results of phenol bromination catalyzed by Bmp5. Substrate screening revealed the Bmp5-catalyzed conversion of phenol to bromophenol and dibromophenol. (a) bromophenol, (tr = 13.1 min / positive mode, m/z = ([M+∙(79Br)] obs. 171.8 calc. 171.9; [M+∙(81Br)] obs. 173.8 calc. 173.9). (b) dibromophenol, (tr = 14.4 min / positive mode, m/z = ([M+∙(79Br/79Br)] obs. 249.6 calc. 249.9; [M+∙(79Br/81Br)] obs. 251.6 calc. 251.9; [M+∙(81Br/81Br)] obs. 253.6 calc. 253.9).

6.7

Preparative synthesis of 2,4-dibromophenol using Bmp5-ADH combiCLEAs

The preparative synthesis of 2,4-dibromophenol (23) was performed as combiCLEAs by immobilization of overexpressed Bmp5 from 1.5 L E. coli lysate with LK-ADH for continuous cofactor regeneration. 1 mM substrate 19 served for Bmp5-catalyzed bromination on 0.5 L scale and the reaction progress was monitored by RP-HPLC.

Figure 33

Bmp5-ADH combiCLEAs-catalyzed conversion of 1 mM 19 proceeded via intermediate 21 to the formation of final product 23. After 63 h no further conversion of 19 was observed resulting in a mixture of 21 and 23.

After a reaction time of 16 h immobilized Bmp5 showed 52 % conversion of starting material 19. Here the chromatogram revealed the formation of intermediate 21 and initial formation of the final product 23. While the reaction proceeded, the amount of 19 as well as 21 decreased resulting in 80 % conversion of starting material after 63 h. As the conversion of 19 barely increased in the last 24 h of incubation, the solid biocatalyst was 107

removed by filtration as no total conversion of starting material was expected. Subsequently, 23 was purified by extraction and column chromatography leading to 40 % final yield (51 mg). Needle shaped crystals were obtained in high purity.

Figure 34

Structural elucidation of 23 via confirmed product identity.

1

H-NMR spectroscopy (500 MHz, DMSO-d6)

Product identity was confirmed by NMR spectroscopy, RP-HPLC as well as ESI-MS. The shown excerpt of 1H-NMR spectrum pointing to the aromatic protons and the additional interpretation of the 13C-NMR proved the Bmp5-catalyzed synthesis of 23. The C3 proton, located between the bromine substituents, resonated as a doublet (7.66 ppm) due to 4J coupling with the proton of C5. For this proton a doublet of doublets was observed at 7.35 ppm, resulting from the mentioned 4J coupling with C3H as well as 3 J coupling to C6H with a coupling constant of 8.7 Hz. A doublet signal (6.91 ppm) was obtained for the C6H caused by 3J coupling to adjacent C5H. The hydroxyl proton resulted in a broad signal at 10.57 ppm. These results confirmed the successful application of Bmp5-ADH combiCLEAs for the preparative synthesis of 2,4-dibromophenol (23). However, the reaction approach should be optimized regarding full conversion of starting material and increasing reaction yield. Furthermore, product identity of 23 was confirmed by ESI-MS, as the observed m/z values were in good consistence to calculated values. Apart from the [M-H]- ions, a chloride adduct ion was additionally detected, thereby supporting the formation of final product 23. 108

7

Establishment of a marine brominase

7.1

Establishment of the brominase Bmp5 for heterologous expression

This project focused on the characterization, establishment and chemoenzymatic application of the FAD-dependent halogenase Bmp5 from Pseudoalteromonas luteoviolacea. Subsequent to subcloning to enable heterologous expression in E. coli the enzyme expression, catalytic activity, substrate scope as well as Bmp5's ability towards immobilization and suitable cofactor regeneration systems was examined to gain an overview of the brominase's area of application. 7.1.1

Successful subcloning of Bmp5 into pET28a

Codon-optimization of bmp5, meaning the change of codons to match the most prevalent tRNAs, was recommended to ensure successful protein translation in E. coli, due to considerable variations of codon preference between different species[51]. Therefore, bmp5 was obtained as codon-optimized sequence deduced from the amino acid sequence of Bmp5 from Pseudoalteromonas luteoviolacea.[8] The halogenase was subcloned from the commercially obtained pMK-RQ vector into the pET28a expression vector and subsequently transformed into E. coli DH5α. Positive clones containing pET28a-bmp5 plasmid were identified by colony-PCR and the correct bmp5 sequence confirmed by Sanger sequencing. To sum up, bmp5 was successfully subcloned for heterologous protein expression in E. coli which was an essential precondition for biocatalytic purposes. 7.1.2

His6-Bmp5 was not obtained in elution fractions after purification via HisTALON affinity chromatography

The halogenase was overexpressed in E. coli BL21 (DE3) pGro7 pET28a-bmp5 and the lysate subsequently employed for purification via His-TALON affinity chromatography. Even though SDS-PAGE analysis indicated successful protein purification as the elution fractions revealed Bmp5's expected band at 61 kDa, protein concentration was determined to merely 1 mg mL-1. Unexpectedly, a very intense band corresponding to Bmp5's and GroEL's molecular mass was observed in the flow-through and first washing fraction,

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_7

109

indicating on the one hand successful protein expression, on the other hand that Bmp5 could not bind via its His6-tag to the matrix. As the obtained bands could correspond to overexpressed chaperone GroEL, the expression of Bmp5 was repeated in E. coli BL21 (DE3) lacking the chaperones. As presumed Bmp5 was only present in the flow-through and not obtained in the elution fractions. These data indicated that despite soluble expression of His6-Bmp5 was possible in E. coli without supporting chaperones, its purification by IMAC was not successful. This suggestion was additionally supported by findings of Agarwal et al. Even though they employed a Ni-NTA column for purification of His6-Bmp5, they also reported that a majority of Bmp5 did not bind to the matrix and was obtained in flow-through and washing fractions.[8] A reason that might cause Bmp5's insufficient interaction with the matrix is that the N-terminal His6-tag is inaccessible, as it might fold inside the protein and is not exposed to the protein surface. To circumvent this difficulty, Bmp5 might be purified under denaturating conditions. However, the enzyme has to be refolded afterwards, what could prove problematic regarding its proper structure and remaining catalytic activity. Another approach to increase Bmp5 purification by IMAC and facilitating the access to the tag is a transfer of the N-terminal His6-tag to the C-terminus. 7.2

Expression of GST-Bmp5 fusion protein

As the purification of His6-Bmp5 via HisTALON affinity chromatography proved challenging, the employment of another suitable tag for the expression of Bmp5 as fusion protein was worth considering. The GST-tag was assumed to be appropriate, as GST-fusion proteins are in the majority of cases soluble in aqueous solution and can easily be purified under native conditions by affinity chromatography employing a glutathione-sepharose matrix. In addition, the GSTtag is reported to stabilize the folding of recombinant proteins and assists to protect against intracellular protease cleavage.[59,60] 7.2.1

Successful subcloning of bmp5 into pETM30 vector

The desired GST-Bmp5 fusion protein was generated by subcloning of bmp5 from pET28a into pETM30 vector that contained the sequence encoding the N-terminal GSTtag. The required restriction sites were introduced into bmp5 by PCR to enable direct subcloning with equally digested pETM30 vector. After successful transformation of pETM30-bmp5 into E. coli DH5α positive clones were identified by colony-PCR and the correct GST-bmp5 sequence confirmed by Sanger sequencing. 7.2.2

GST-Bmp5 aggregated as inclusion bodies and could not be purified via affinity chromatography

To investigate, whether the purification of Bmp5 proceeded successfully with an Nterminal GST-tag, pETM30-bmp5 was overexpressed in E. coli BL21 (DE3) and purified via GST-affinity chromatography. As revealed by SDS-PAGE analysis, no GST-Bmp5 was obtained in the elution fractions, but instead an intensive protein around 27 kDa, 110

corresponding to the GST-tag appeared. As this tag was expressed at the N-terminus of the fusion protein and was therefore synthesized prior to the remaining protein, its size might hinder proper folding and interrupt further protein translation, thereby explaining its presence and the lack of complete fusion protein in the elution fractions. Further evidence towards an incorrect folding was given by the observation that a protein band corresponding to GST-Bmp5 appeared in resolubilized cell debris. This supported the suggestion that the fusion protein formed aggregates and consequently, the protein occurred in the insoluble fraction. Probably aggregation leads to an interruption of protein biosynthesis as well as degradation of Bmp5 and therefore merely GST was found in the soluble fraction. Following attempts for enzymatic bromination of 4-HBA with GST-Bmp5 from flowthrough and in E. coli lysate revealed no significant enzymatic activity. This observation confirmed the suggestion that the GST-tag is neither suitable to purify Bmp5 nor to obtain an active brominase. In order to prevent the accumulation of GST-Bmp5 as inclusion bodies, a recommended method is to lower the expression temperature to 20 or 15 °C, thereby trying to increase the expression of soluble protein. Another approach is the extraction of the aggregated proteins from the inclusion bodies with chaotropic reagents such as urea and afterwards refolding the proteins so that the enzyme regains its catalytic activity.[60] As the GST-tagged Bmp5 did not reveal any catalytic activity and as the success of the mentioned optimization steps was uncertain, enzymatic approaches using GST-Bmp5 were not further investigated but therefore the focus returned to His6-Bmp5. 7.3

His6-Bmp5 revealed sufficient enzymatic activity after purification

Initially, Bmp5's enzymatic activity regarding the conversion of 4-HBA (19) was investigated by employing Bmp5 collected from the flow-through after purification attempts via His-TALON affinity chromatography. Concerning the flow-through a moderate conversion of 50 % starting material was observed after 24 h and the detection of intermediate 21 as well as the final product 23 proved Bmp5's catalytic activity. As expected samples from the elution fractions lacked catalytic activity. These data were in consistence to published results of Agarwal et al.[8] As Bmp5 obtains an intrinsic flavin reductase domain, it is able to generate the essential cofactor FADH2 autonomously in situ and merely requires NADPH as reducing agent. Even though the described Bmp5catalyzed conversion of 4-HBA (19) proved Bmp5's intrinsic ability to generate FADH2, it did not result in full conversion of the starting material. As the amount of NADPH was limiting in this approach due to oxidation by different oxidoreductases and further degrading enzymes co-eluting in the flow-through, no full conversion of 19 was expected even though an excess of reducing agent was supplied.

111

7.3.1

Bmp5's intrinsic flavin reductase domain was required for sufficient enzymatic activity

In order to investigate the requirement of intrinsic cofactor regeneration for Bmp5, an external cofactor regeneration system consisting of PrnF and an alcohol dehydrogenase was employed. This approach should circumvent Bmp5's internal flavin reductase ability by in situ generation of FADH2. Even though the employed cofactor regeneration system was well-established for FAD-dependent tryptophan halogenases, this approach gave merely 9 % conversion of starting material 19. Therefore, the additional flavin reductase in combination with ADH was considered as inappropriate for Bmp5-catalyzed bromination and was not followed in detail for further investigations. One speculation might be that the cofactor FAD is deeply buried in the enzyme’s active site facilitating simple reduction to FADH2 by NADPH to enable further reaction progress without release of cofactor from the enzyme scaffold. Probably this mechanism facilitates accelerated bromination, because the overall reaction does not depend on an external flavin reductase. Moreover, FADH2 released into the medium can undergo side reactions that impede halogenation, e.g. with O2 to form flavin hydroperoxide. 7.4

Establishment of cofactor regeneration system for Bmp5-catalyzed reactions

Before investigations on a new cofactor regeneration system begun, the general activity of Bmp5 in E. coli BL21 (DE3) pET28a-bmp5 crude lysate was proven by achieving 60 % final conversion of starting material after supplying an excess of NADPH. In addition, the presence of intermediate 21, 4-bromophenol (27) as well as the final product 2,4-dibromophenol (23) was identified by comparison of RP-HPLC elution signals with authentic standards and product identity was further confirmed by GC-MS analysis. Full conversion of starting material was probably not achieved, as NADPH was oxidized by other enzymes present in E. coli lysate, thereby impeding the generation of essential cofactor FADH2. 7.4.1

GDH led to acidification of reaction mixture and denaturation of enzymes

As revealed by previous examinations the intrinsic function of Bmp5 as flavin reductase could not be efficiently circumvented by external supply of FADH2. Therefore, another approach for cofactor regeneration was performed to assure the steady supply of NADPH required for the reduction of FAD by Bmp5 to promote the bromination. Here glucose dehydrogenase from Bacillus megaterium was considered as suitable supplier for NADPH, as it oxidizes D-glucose, a low-cost chemical, to D-glucono-1,5-lactone.[55–57] The general success of GDH for the regeneration of NADPH could be demonstrated by only 15 % conversion of 4-HBA (19). On the one hand, the hydrolysis of D-glucono-1,5lactone to gluconic acid was considered as positive effect due to the shift of equilibrium in favor of product formation. On the other hand, this byproduct led to an acidification of the reaction mixture and following denaturation of enzyme components. In order to circumvent the denaturation of enzymes and resulting catalytic inactivity, the pH value of the reaction mixture should be kept between pH 7 and 7.5 by addition of appropriate 112

volumes of base. This control could be conducted manually or automated by employing an automatic titrator. However, as a simple and robust bromination system was desired, GDH was not further employed as cofactor regeneration system and therefore focus was laid on ADH from Lactobacillus kefir. 7.4.2

NADP+-dependent ADH from Lactobacillus kefir was suitable for effective cofactor regeneration for Bmp5-catalyzed reactions

The NADP+-dependent alcohol dehydrogenase LK-ADH from Lactobacillus kefir was employed to catalyze the reduction of NADP+ to NADPH by oxidizing iso-propanol and turned out to be a suitable oxidoreductase, as it enabled full conversion of Bmp5catalyzed bromination of 4-HBA (19).[58] Furthermore, the addition of iso-propanol was beneficial to dissolve the less polar reaction components, e.g. 4-HBA, in aqueous buffer. The conversion of 19 was directly performed in E. coli lysate and total conversion of starting material and intermediate 21 was reached within 2 h of reaction time, confirming an efficient supply of NADPH by LK-ADH. Further incubation led to the formation of the final product 23 as proven by GC-MS analysis. To sum up, a successful and efficient cofactor regeneration system was established by employing LK-ADH as auxiliary enzyme that presents an essential precondition for preparative application of the brominase. 7.5

Evidence of Bmp5 expression by Western blot analysis and peptide mass fingerprint

Even though the previous experiments clearly indicated Bmp5's successful expression, as they revealed brominase activity by conversion of 4-HBA and detection of expected reaction components, the halogenase's presence was unambiguously confirmed by Western blot analysis and peptide mass fingerprint. Regarding Western blot analysis His6-Bmp5 was successfully detected at its molecular mass of 61 kDa. Bmp5 degradation products still containing the His6-tag were detected in addition. To ensure that these bands were certainly degradation products, new protein samples containing a universal protease-inhibitor mix could be analyzed that should reveal less intense bands of minor size. As purification of His6-Bmp5 proved problematic but could not be optimized so far, the protein samples were obtained from E. coli lysate. Consequently, the blot showed a high background what could be minimized by increased and prolonged washing steps. In addition, higher dilution of lysate for immunoblot detection is recommended for future analysis. Nevertheless, the obtained results were sufficient to confirm Bmp5's expression in E. coli. Additional identification of Bmp5 was performed by peptide mass fingerprint subsequent to tryptic digestion and MALDI-ToF mass spectrometry analysis. Comparison of in silico digestion resulted in a sequence coverage of 41 %. As over 25 % of theoretical fragments from trypsin digestion have a length of only one amino acid, many peptide sequences are too short to be assigned uniquely to the amino acid sequence. Even though the theoretical 113

coverage limit of trypsin digestion is 74 %, the obtained coverage of 41 % is still satisfying to clearly identify Bmp5.[61] 7.6

Detected reaction compounds confirmed proposed Bmp5-catalyzed reaction mechanism

So far the detected reaction compounds of the Bmp5-catalyzed conversion of 4-HBA were in agreement to findings reported by Agarwal et al.[8] However, some differences arose concerning the detection of 2,4,6-tribromophenol (28). Even though Agarwal et al. examined the presence of compound 28 in vivo when Bmp5 is heterologously expressed in E. coli, they were not able to detect the compound in vitro. However, previous experiments as part of this work that were also conducted in vitro, proved the presence of 28 by GC-MS analysis. Compound 28 is supposed to appear as side product in the Bmp5-catalyzed conversion of 4-HBA that mainly results in the formation of dibrominated phenol 23.[8] As shown in Scheme 11 two reaction mechanisms, (a) and (b), were suggested by Agarwal et al. to lead to the formation of 23 as well as compound 28. For both reaction pathways 4-HBA (19) is initially brominated in ortho position due to the ortho/para directing effect of the hydroxyl group. As the second ortho bromination (b) is sterically hindered by Bmp5's active site what is in accordance to Agarwal, a decarboxylative bromination in para position (a) seems to be the preferred second reaction step, thereby resulting in formation of 23.[8] In addition, the release of CO2 might drive the equilibrium for electrophilic substitution at this position. Even though an additional bromination in ortho position to the hydroxyl group of intermediate 21 is disfavored, it still occurs to a small extent and leads to formation of 30 as described in pathway (b). Following decarboxylative bromination in para position explains the formation of 28. As its precursor 30 is reported to be a poor substrate for Bmp5 in comparison to 4-HBA and intermediate 21, conversion of 30 will just take place to a small extent what corresponds to minor amounts of 28 detected in Bmp5-catalyzed reactions. However, the observation that compound 30 was neither detected by RP-HPLC nor by GC-MS analysis raises doubts that the reaction pathway (b), proposed by Agarwal et al., is depicting reality. The synthesis of compound 28 could also proceed via pathway (c), whereby the final product 23 is additionally brominated in second ortho position to the hydroxyl group. Even though further substitution of the dibrominated phenol derivative might be hindered due to high deactivation of the aryl compound and sterical effects, the proposed reaction pathway would correspond to all detected reaction compounds. Therefore, in contrast to previous suggestions by Agarwal et al., a small extent of final product 23 is supposed to be directly converted into 28. Based on these data it can be assumed that the synthesis of 28 according to route (b) is less likely to take place. In order to gain insight into the pathway of Bmp5-catalyzed bromination substrate analogues, e.g. ortho-substituted phenols, can be used for stepwise identification of reaction intermediates.

114

Scheme 11

Proposed reaction mechanism for Bmp5-catalyzed conversion of 4-HBA (19). The main product 23 is synthesized via reaction pathway (a), whereas compound 28, formed via pathway (b) or (c), occurs merely a side product.

Agarwal et al. supported their proposed ortho bromination of 4-HBA as initial reaction step, as they could not detect 4-bromophenol (27) via HPLC-analysis and did not suggest it as intermediate or product of the Bmp5-catalyzed bromination. However, these investigations could prove the presence of 27 by RP-HPLC measurements as well as GCMS analysis. Even though this observation appeared controversial to Agarwal et al., 27 could be detected in every approach rapidly after reaction start and did not show any further conversion during reaction progress, as signal intensity did not decrease during the reaction. Therefore, it is in agreement with Agarwal's suggestions that 27 is no intermediate or product of Bmp5-catalyzed reactions. Summing up, the detected reaction components and detailed evaluation of their conversion generally supported Agarwal's assumptions about Bmp5's reaction pathway that mainly results in the formation of 23. Still additional compounds such as 4-bromophenol (27) and 28 were detected in vitro that was so far not observed. However, a new synthesis route was proposed for tribrominated phenol compound 28 as side product thereby questioning the suggestions made by Agarwal et al.

115

7.6.1

Bmp5 preferred bromination to chlorination

In order to analyze Bmp5's specificity regarding the incorporation and oxidation of bromide in comparison to chloride, different mixtures of 4-HBA were prepared that contained up to a 10-fold excess of NaCl compared to NaBr. Still, no chlorinated reaction products could be identified by GC-MS analysis. The mentioned results indicate that exclusively bromide serves as co-substrate of Bmp5 that is oxidized into hypohalous acid for subsequent electrophilic substitution even if NaCl is supplied in large excess. Initial investigations that Bmp5 could not use chloride in vivo were performed by Agarwal et al. However, the systematic elevation of chloride concentration described in this work proved the clear preference of Bmp5 to catalyze solely bromination of 4-HBA.[8] As chloride's radius is smaller than bromide's (181 Ǻ vs. 196 Ǻ)[62], it was assumed that it easily fits into Bmp5's active pocket and is consequently employed for the chlorination of substrates. In addition, other FAD dependent halogenases such as the tryptophan halogenases RebH, PyrH and Thal catalyze both bromination and chlorination, whereas the latter reaction type is preferred, as they are probably unable to distinguish between the mentioned ions.[19,20] Therefore, the ion radius is proposed to be crucial for reaction rate. As no crystal structure of Bmp5 or similar brominases is available so far, it is difficult to give profound suggestions about the structure of the halogenase's active site. However, it can be proposed that Bmp5's active pocket is made up of amino acid residues that exclusively interact with bromide ions and are not able to position chloride ions correctly. Maybe the solvation shell of the halide ions influences the uptake into the active site. In this context, shell size as well as enthalpic effects might play a role to enable the clear discrimination. Bmp5 is originally expressed in the marine organism Pseudoalteromonas luteoviolacea and part of the bmp-gene locus that is proposed to be involved in the synthesis of polybrominated diphenyl ethers (PBDEs). It seems likely that these organisms must be able to ensure the synthesis of brominated compounds even in presence of high NaCl concentrations in seawater. Because brominated metabolites, such as PBDEs represent a ubiquitous group of marine natural products that are broadly distributed in marine microorganisms, the highly specific introduction of bromide rather than chloride might be an essential feature of Bmp5 that ensures the successful synthesis of PBDEs as secondary metabolites.[44] As PBDEs are highly toxic for other organisms such as mammals, these compounds might serve as protection against predators. 7.6.2

Bmp5 was unable to catalyze the iodination of 4-HBA

As indicated by previous suggestions regarding Bmp5's acceptance of halide ions for the conversion of 4-HBA, it was proposed that Bmp5 exclusively catalyzes the incorporation of bromide rather than chloride. Focusing on iodide it was consequently assumed that Bmp5 would not incorporate this ion, as its radius of 220 Ǻ[62] is significantly larger than bromide's and will therefore not fit into the enzyme's active pocket. Consistent with these assumptions several attempts concerning the iodination of 4-HBA in E. coli lysate as well as on a preparative scale with Bmp5-ADH combiCLEAs did not lead to conversion of 116

starting material and iodinated reaction products could not be detected by GC-MS analysis. Even though Agarwal et al. reported that Bmp5 can accept iodide for incorporation to 4HBA in vivo, a closer look at their experimental setup revealed that the production of iodinated compounds was only detected in E. coli cells expressing the whole gene cluster Bmp1-8 and not just Bmp5.[8] Probably the interaction between the different enzymes within the in vivo system is crucial to facilitate this promiscuous enzyme function. The iodination approaches of 4-HBA should still be repeated under different reaction conditions to make an unambiguous statement that Bmp5 is indeed unable to incorporate iodide. Therefore, Bmp5's sole acceptance of bromide is a remarkable trait that stands out from other known FAD-dependent halogenases and could be very useful regarding the specific bromination of organic compounds. 7.6.3

Investigations on Bmp5's substrate scope revealed synthesis of mono- and dibrominated phenol

As implied by Bmp5's exclusive specificity for bromide as halogenating species, it can be suggested that Bmp5 has a quite narrow substrate scope and only accepts derivatives that are structurally related to 4-HBA. Therefore, the next step was the detailed investigation of Bmp5's substrate specificity regarding six potential substrates representing a broad range of aromatic compounds. The conversion of phenol (31) was already indicated by RP-HPLC analysis and the presence of bromophenol (32) as well as dibromophenol (33) confirmed by GC-MS. As 31 is structurally very similar to 4-HBA and merely lacks the carboxyl function, it is likely that this compound fits well into Bmp5's active pocket. It is proposed that bromination occurs in ortho position as observed for the natural substrate 4-HBA due to the electron donating, ortho/para directing effect of the hydroxyl group. The proposed product 2-bromophenol (32) is structurally related to 21 and subsequently serves as additional substrate to enable second bromination. It can be assumed that the second substitution resulting in dibromophenol (33) takes place in second ortho or para position relative to the hydroxyl group. Nevertheless, due to the ortho/para-directing effects of the hydroxyl as well as bromine substituent in combination with steric properties the substitution pattern of 33 cannot be predicted precisely (Figure 35).

Figure 35

Analysis of Bmp5's substrate scope identified phenol (31) as potential substrate. 31 is initially brominated in ortho position and subsequently converted into the dibrominated compound 33. 117

Certainty about the exact position of bromination will be revealed by NMR spectroscopy. To obtain the required amount of 33 for this analysis, the conversion of phenol should be conducted on a preparative scale. Even though tyrosine carries a phenol moiety and was expected to serve as substrate, no conversion was observed as the additional amino acid backbone probably hinders correct positioning in Bmp5's active site. Surprisingly, neither aniline nor benzoic acid that are structurally similar to 4-HBA were tolerated as substrates. It is possible that Bmp5 requires a hydroxyl group for correct positioning of the substrate as this functional group is present in 4-HBA as well as phenol. As aniline and benzoic acid lack this functionality, they probably cannot serve as substrate for Bmp5. Concerning benzoic acid it should be noted that the electron density of the aromatic moiety is lowered because of the electron withdrawing effect of carboxylate, thereby impeding electrophilic substitution. As the remaining tested substrates indole and phenylalanine have an even more complex structure and are bigger in size than the known accepted substrates, it is not surprising that these compounds were not converted by Bmp5. Summing up, Bmp5 exhibits a narrow substrate scope and does not accept all tested structurally similar compounds. This finding is in consistence to flavin-dependent tryptophan halogenases that are characterized by a considerable narrow substrate scope, although Bmp5 has only weak relationship to these enzymes. Especially in consideration of the application of Bmp5 in organic chemistry, a broadened substrate spectrum would be recommended to brominate a variety of compounds. Therefore, Bmp5's substrate spectrum could be modified by directed evolution, screening for a mutant that accepts additional substrates or is even able to efficiently incorporate iodine substituents. 7.6.4

Bmp5-ADH combiCLEAs enabled preparative scale synthesis of 2,4-dibromophenol

As complete conversion of 4-HBA to 2,4-dibromophenol (23) was achieved on an analytical scale and product identity confirmed by GC-MS analysis, the next step was the preparative scale synthesis of 23. The co-immobilization of Bmp5 with LK-ADH for cofactor regeneration as combiCLEAs has never been performed so far, but good results were expected, as other FAD-dependent halogenases such as Thal could be immobilized successfully for simple upscaling of enzymatic bromination on large scale. After a reaction time of 63 h the combiCLEAs revealed 80 % conversion of 1 mM 4-HBA in 0.5 L reaction volume, thereby proving the successful immobilization of Bmp5 using the CLEA methodology. After subsequent separation and purification from remaining reaction products 2,4-dibromophenol (23) was obtained in 40 % yield and product identity was confirmed by NMR spectroscopy. Even though the increasing degradation of both Bmp5 and the auxiliary enzyme ADH together with essential cofactors might be a reason that prevented the reaction mixture from reaching full conversion, the obtained yield was satisfying as the Bmp5-ADH combiCLEAs were the first attempt for preparative scale synthesis of 23. As this compound tended to form crystals after purification, re-crystallization as simple workup step can be considered. In order to 118

optimize the reaction process, influencing factors such as substrate loading or glutaraldehyde concentration could still be improved in the future. Especially the employed glutaraldehyde concentration of 0.7 %, that was initially chosen with respect to tryptophan halogenases might be too high, as Bmp5 is merely cross-linked with ADH and lacking an additional flavin reductase as required for the immobilization of tryptophan halogenases. To avoid that the reaction ends up in a mixture of intermediate and dibrominated compound 23, drop-wise, retarded addition of substrate 19 could be tried out. This should result in a low substrate concentration that is immediately converted into the intermediate and further reacts to 23, when no additional substrate is available. All in all it was shown that Bmp5 could be successfully immobilized as combiCLEAS and enabled the preparative scale synthesis of 23.

119

8

Outlook

This thesis focused on two distinct projects, whereas the first part dealt with the generation of a thermostable tryptophan 6-halogenase and the second project referred to the characterization and establishment of the marine brominase Bmp5. In the course of this thesis a thermostable Thal variant, Thal-E2R, that revealed a significantly improved thermostability and catalytic activity in comparison to the WT was generated by random mutagenesis using epPCR for diversification. As the purification of the Thal mutant proved difficult and impeded further chemoenzymatic approaches with purified enzyme, the purification strategy should be reconsidered and optimized as an initial step in the future. High protein expression levels would also facilitate and accelerate preparative scale synthesis of L-6-bromotryptophan, thereby paving the way towards the application of the improved Thal variant for synthetic purposes. Especially employing Thal-E2R in tandem reaction cascades, e.g. multistep biotransformations and Pd-catalyzed cross-couplings, is expected to be advantageous due to its big potential for large-scale applications. A more efficient tryptophan 6-halogenase could also serve as catalyst for the bromination of several non-native substrates. As greater amounts of pure protein are also required for the elucidation of a protein's crystal structure, this aspect would be an additional motivation to improve the current purification strategy. Because Thal's crystal structure has not been elucidated yet, crystallization would be a desired goal, in order to get precise information about the protein's three-dimensional structure. A long-term goal would thereby be the elucidation of a crystal structure for the Thal mutant, allowing more precise speculations, how the mutated amino acid residues influence the protein conformation and consequently its thermostability. In addition, this information enables evaluations on Thal's halogenation mechanism that is currently based on other tryptophan halogenases, such as PrnA. To further improve the application scope of Thal mutant E2R, additional rounds of directed evolution should be performed to stepwise optimize the enzyme's thermal resistance. Apart from the generation of an even more thermostable variant, Thal-E2R could also serve as parent for screening for the acceptance of additional substrates or organic solvents, thereby increasing the enzyme's applicability for organic synthesis.

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_8

121

Regarding the second project, focusing on the establishment of the FAD-dependent halogenase Bmp5 from Pseudoalteromonas luteoviolacea, the corresponding gene bmp5 was successfully subcloned and subsequently overexpressed in E. coli. Similar to ThalE2R, the purification of His6-tagged Bmp5 was challenging, as the protein tag did not bind sufficiently to the matrix. Therefore, subcloning of the N-terminal His6-tag to the Cterminus or the employment of another protein tag should be considered in the future as promising purification strategy. Based on successful bromination of 4-HBA and phenol further examinations of Bmp5's substrate scope can be carried out by testing different substituted phenol derivatives. Thereby, results about the substrate-binding mode of Bmp5 would be clarified. It is of special interest to study the regioselectivity of the biocatalyst with regard to a possible influence of directing effects. In order to broaden Bmp5's substrate scope, the brominase could be additionally modified by directed evolution with special emphasis to more complex phenol derivatives or improved thermostability as performed for Thal. For this purpose the establishment of a high-throughput screening strategy suitable for Bmp5 is recommended. Even though the preparative scale synthesis of Bmp5's final product 2,4-dibromophenol proved successful in an initial approach, the application of Bmp5 in whole-cell catalysis might be a versatile alternative to immobilization using combiCLEAs. To enable cofactor regeneration Bmp5 and the alcohol dehydrogenase can be simultaneously overexpressed in E. coli providing a suitable reaction compartment for bromination. The substrate 4HBA could enter the permeabilized cell, where bromination should occur and subsequently lead to release of the brominated product. Perhaps Bmp5's activity is increased in this approach to enable conversion of higher substrate loadings. The elucidation of Bmp5's crystal structure is a challenging, but a desirable aim, as this brominase has not been crystallized until now. The obtained structure would give valuable information on the enzyme's active pocket and would allow to examine Bmp5's proposed bromination mechanism in more detail, whereby special importance can be laid on the remarkable preference of bromide to chloride ions as cosubstrate. To sum up, both studied halogenases, Thal-E2R and Bmp5, obtain a substantial potential for the application in bioorganic chemistry and are worth for further investigations to optimize both enzymes for chemoenzymatic approaches leading to the sustainable synthesis of haloarenes.

122

9

Summary

This thesis was based on the investigation and evaluation of two FAD-dependent halogenases, in order to analyze and improve their applicability for chemoenzymatic approaches in organic chemistry. In contrast, conventional strategies for halogenation require hazardous reagents, e. g. molecular halogens, and suffer from low selectivity. To circumvent this issue establishment of enzymatic halogenation arises as a promising alternative. Therefore, the first project focused on the generation of a thermostable variant of the tryptophan 6-halogenase Thal by means of directed evolution. A thermo resistant halogenase promised to exhibit an improved lifetime and catalytic activity even at elevated temperatures, thereby facilitating the enzyme's application for preparative and industrial purposes. Concerning the establishment of a thermostable Thal variant, the wild type's thermostability was initially investigated in detail, in order to determine the enzyme's temperature threshold above which its catalytic activity significantly decreases due to denaturation. By incubation of purified Thal WT at different temperatures and subsequent bromination of L-tryptophan at 25 °C the enzyme's activity profile was determined and revealed no bromination activity after heat treatment above 49.1 °C. Consequently, this temperature served as screening criterion for the identification of a significantly more temperature stable mutant by means of directed evolution. Therefore, mutant libraries containing randomly introduced mutations in the thal gene were generated by epPCR, after reaction conditions concerning mutagenesis were optimized with regard to the desired mutation rate of 2 mutations per kb. After cultivation and protein expression in a 96-well plate the cells were lysed and the lysate containing overexpressed Thal mutant was heat-shocked at 49.1 °C for 20 min. The remaining catalytic efficiency of the mutants was tested by bromination of L-tryptophan with continuous cofactor regeneration, whereby the cofactors as well as auxiliary enzymes were added after the heat-shock, as these components should not be exposed to elevated temperatures. High-throughput screening of the libraries to quantify the final conversion of L-tryptophan to L-6-bromotryptophan was monitored by fluorogenic Suzuki-Miyaura

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0_9

123

coupling resulting in a fluorescent biaryl. Subsequently, positive mutants were rescreened by fluorescence detection and HPLC measurement. After screening of 240 Thal mutants for the desired elevated thermal stability, three potential mutants revealed an up to 2.7-fold increased fluorescence, directly corresponding to increased conversion of L-tryptophan to L-6-bromotryptophan. After rescreening of these three outliers, one mutant, referred to as Thal-E2, proved the previously observed enhanced thermostability in comparison to the WT. Sequencing of Thal-E2 revealed three mutations, S359G and K374R whereby the third one was a silent mutation of K510. This mutation was initially considered as reason for low protein concentrations of His6-Thal after overexpression and purification via His-TALON affinity chromatography due to E. coli's codon usage. Therefore, the rarely employed AAA codon encoding lysine in E. coli was reversely mutated into the wild type sequence at this position, thereby striving for an increase of Thal-E2's protein expression, that was now referred to as Thal-E2R. Unfortunately, this optimization attempt did not result in the desired increased protein levels. Another approach for the improvement of protein expression focused on the codon usage of the mutation K374R, that resulted in a codon that merely represented 4 % of the employed triplets for arginine. Still, expression of the Thal mutant E2R in E. coli strain containing additional copies of tRNAs recognizing this rare arginine codon did not lead to elevated protein expression. Comparison of the Thal mutants to the WT revealed significantly improved enzymatic activity despite thermal treatment at 49.1 °C and subsequent bromination of L-tryptophan at 25 °C. Although the WT's activity dropped to zero above 49.1 °C, the mutant showed catalytic activity up to 58 °C, thereby exhibiting an elevation of enzyme stability of 9 °C. Thal-E2R's improved thermal resistance was further investigated and compared to the WT in E. coli lysate, what revealed up to 3-times higher conversion rates of substrate and retained enzymatic activity despite thermal treatment prior bromination reaction. Preparative synthesis of L-6-bromotryptophan was successfully performed as combiCLEAs by co-immobilization of Thal-E2R with a flavin reductase and alcohol dehydrogenase to enable continuous cofactor regeneration. Even though the equal CLEAs preparation with Thal WT resulted in a more efficient conversion of substrate due to a higher expression level, the Thal-E2R-PrnF-ADH combiCLEAs approach revealed a sufficient amount of L-6-bromotryptophan and the mutant's retained regioselectivity was proven by one- and two-dimensional NMR spectroscopy. The beneficial effect of Thal-E2R's mutation of serine against glycine and lysine against arginine regarding its thermostability was supported by a proposed homology model of Thal-E2R, as no crystal structure was elucidated so far. It was assumed that the mutated arginine is exposed to the surface and directly located between the proposed contact area of the two Thal-E2R monomers, thereby strengthening their interaction and contributing to the dimer's stability as temperatures rise. The substitution of serine against glycine was proposed to additionally contribute to the mutant's stability, as serine was reported to be rarely employed in thermophilic proteins. 124

The second project of this thesis dealt with the establishment and investigation of the marine FAD-dependent halogenase Bmp5. The bmp5 gene was deduced from the marine organism Pseudoalteromonas luteoviolacea and obtained as codon-optimized gene for the overexpression in E. coli. After successful subcloning into pET28a expression vector and following overexpression the purification of His6-Bmp5 via HisTALON affinity chromatography proved challenging, as Bmp5 could not be obtained in the elution fractions. Misfolding or inaccessibility of the N-terminal protein tag was considered as reason for the insufficient interaction with the matrix, thereby leading to the detection of Bmp5 in the flow-through. Following subcloning of bmp5 into pETM30 vector, that encoded the sequence for an N-terminal GST-tag, aimed to improve the halogenase's purification procedure via GST-affinity chromatography. However, purification of GSTBmp5 proved even more challenging, as the fusion protein aggregated to form inclusion bodies and could not be obtained in the soluble protein fractions. In addition, the GST-tag was not suitable for purification of Bmp5, because its huge size probably hindered proper folding of Bmp5 or impeded complete protein translation. Therefore, the focus was returned to His6-Bmp5 that still revealed sufficient enzymatic activity after purification, even though it was obtained in the flow-through. Furthermore, Bmp5's expression in E. coli was proven by Western blot analysis as well as peptide mass fingerprint, resulting in a sequence coverage of 41 % in comparison to in silico digestion. In order to catalyze the conversion of Bmp5's substrate 4-hydroxybenzoic acid to 2,4-dibromophenol, Bmp5's intrinsic flavin reductase domain is able to autonomously generate the essential cofactor FADH2 by oxidizing NADPH. An attempt to circumvent the in situ generation of FADH2 by an external regeneration of FADH2 revealed a significantly reduced enzymatic activity, thereby underlining the necessity of the intrinsic flavin reductase domain for sufficient enzymatic activity of Bmp5. In order to assure the steady supply of NADPH, glucose dehydrogenase from Bacillus megaterium was considered as suitable regeneration system for NADPH. However, GDH gave no satisfying results for the desired application, as degradation products led to the acidification of reaction mixture and subsequent denaturation of proteins. Finally, NADP+-dependent alcohol dehydrogenase LK-ADH from Lactobacillus kefir was chosen as appropriate cofactor regeneration system, as it catalyzes the reduction of NADP+ to NADPH by oxidizing iso-propanol. This approach enabled full conversion of 4-HBA into intermediary formed 3-bromo-4-hydroxybenzoic acid and the final product 2,4-dibromophenol. The presence of intermediates and side products was proven by GCMS analysis. In contrast to previous examinations of Bmp5's bromination mechanism, a modified reaction pathway for the synthesis of 2,4,6-tribromophenol as side products was established. Apart from a modified reaction progress it was examined that Bmp5 exclusively catalyzes the bromination of its substrates and is unable to introduce chloride as well as iodide. Probably Bmp5's active pocket only allows the sole introduction of bromine rather than other halide ions, even if they are smaller in size. In addition, the brominase revealed a narrow substrate scope, as it merely accepted phenol as substrate. 125

Finally, Bmp5's preparative synthesis of 2,4-dibromophenol was performed as combiCLEAs, thereby co-immobilizing the brominase with LK-ADH for continuous cofactor regeneration. This first attempt of immobilization revealed sufficient conversion rates of starting material and proved the enzyme's applicability for combiCLEAs leading to an acceptable overall yield.

126

Literature [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40]

C.-H. Wong, G. M. Whitesides, Enzymes in Synthetic Organic Chemistry, Academic Press, 1994. “The Central Role of Enzymes as Biological Catalysts - The Cell - NCBI Bookshelf,” R. Wohlgemuth, Curr. Opin. Biotechnol. 2010, 21, 713–724. D. R. M. Smith, S. Grueschow, R. J. M. Goss, Curr. Opin. Chem. Biol. 2013, 17, 276–283. F. H. Vaillancourt, E. Yeh, D. A. Vosburg, S. Garneau-Tsodikova, C. T. Walsh, Chem. Rev. 2006, 106, 3364–3378. J. A. Bush, B. H. Long, J. J. Catino, W. T. Bradner, K. Tomita, J. Antibiot. 1987, 40, 668–678. M. Piraee, R. L. White, L. C. Vining, Microbiol. Read. 2004, 150, 85–94. V. Agarwal, A. A. El Gamal, K. Yamanaka, D. Poth, R. D. Kersten, M. Schorn, E. E. Allen, B. S. Moore, Nat. Chem. Biol. 2014, 10, 640–647. G. W. Gribble, J. Chem. Educ. 2004, 81, 1441. H. Deng, L. Ma, N. Bandaranayaka, Z. Qin, G. Mann, K. Kyeremeh, Y. Yu, T. Shepherd, J. H. Naismith, D. O’Hagan, ChemBioChem 2014, 15, 364–368. D. O’Hagan, H. Deng, Chem. Rev. 2015, 115, 634–649. F. Diederich, P. J. Stang, Metal-Catalyzed Cross-Coupling Reactions, John Wiley & Sons, 2008. M. Z. Hernandes, S. M. T. Cavalcanti, D. R. M. Moreira, W. F. de Azevedo Junior, A. C. L. Leite, Curr. Drug Targets 2010, 11, 303–314. P. Jeschke, Pest Manag. Sci. 2010, 66, 10–27. K. Smith, G. A. El-Hiti, Curr. Org. Synth. 2004, 1, 253–274. C. D. Murphy, B. R. Clark, in Ster. Synth. Drugs Nat. Prod., John Wiley & Sons, Inc., 2013. S. Keller, T. Wage, K. Hohaus, M. Hölzer, E. Eichhorn, K.-H. van Pée, Angew. Chem. Int. Ed. 2000, 39, 2300–2302. S. Flecks, E. P. Patallo, X. Zhu, A. J. Ernyei, G. Seifert, Alexander, C. Dong, J. H. Naismith, K.-H. van Pée, Angew. Chem. Int. Ed 2008, 47, 9533–9536. E. Yeh, S. Garneau, C. T. Walsh, Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 3960–3965. S. Zehner, A. Kotzsch, B. Bister, R. D. Süssmuth, C. Méndez, J. A. Salas, K.-H. van Pée, Chem. Biol. 2005, 12, 445–452. D. Milbredt, E. P. Patallo, K.-H. van Pée, ChemBioChem 2014, 15, 1011–1020. J. R. Heemstra, C. T. Walsh, J. Am. Chem. Soc. 2008, 130, 14024–14025. C. Dong, S. Flecks, S. Unversucht, C. Haupt, K.-H. van Pée, J. H. Naismith, Science 2005, 309, 2216–2219. C. Dong, A. Kotzsch, M. Dorward, K. H. van Pée, J. H. Naismith, Acta Crystallogr. D Biol. Crystallogr. 2004, 60, 1438–1440. E. Bitto, Y. Huang, C. A. Bingman, S. Singh, J. S. Thorson, G. N. Phillips, Proteins Struct. Funct. Bioinforma. 2008, 70, 289–293. X. Zhu, W. De Laurentis, K. Leang, J. Herrmann, K. Ihlefeld, K.-H. van Pée, J. H. Naismith, J. Mol. Biol. 2009, 391, 74–85. E. Yeh, L. C. Blasiak, A. Koglin, C. L. Drennan, C. T. Walsh, Biochemistry 2007, 46, 1284–1292. N. M. Kamerbeek, D. B. Janssen, W. J. H. van Berkel, M. W. Fraaije, Adv. Synth. Catal. 2003, 345, 667–678. E. Yeh, L. J. Cole, E. W. Barr, J. M. Bollinger, D. P. Ballou, C. T. Walsh, Biochemistry 2006, 45, 7904–7912. C. Seibold, H. Schnerr, J. Rumpf, A. Kunzendorf, C. Hatscher, T. Wage, A. J. Ernyei, C. Dong, J. H. Naismith, K.-H. van Pée, Biocatal. Biotransformation 2006, 24, 401–408. M. Frese, N. Sewald, Angew. Chem. Int. Ed. 2015, 54, 298-301 W. S. Glenn, E. Nims, S. E. O’Connor, J. Am. Chem. Soc. 2011, 133, 19346–19349. M. Frese, P. H. Guzowska, H. Voß, N. Sewald, ChemCatChem 2014, 6, 1270–1276. M. Hölzer, W. Burd, H.-U. Reißig, K.-H. van Pée, Adv. Synth. Catal. 2001, 343, 591–595. J. T. Payne, M. C. Andorfer, J. C. Lewis, Angew. Chem. Int. Ed. 2013, 52, 5271-5274 A. Lang, S. Polnick, T. Nicke, P. William, E. P. Patallo, J. H. Naismith, K.-H. van Pée, Angew. Chem. Int. Ed. 2011, 50, 2951–2953. C. K. Savile, J. M. Janey, E. C. Mundorff, J. C. Moore, S. Tam, W. R. Jarvis, J. C. Colbeck, A. Krebber, F. J. Fleitz, J. Brands, P.Devine, G.Huisman, G.Hughes, Science 2010, 329, 305–309. H. Renata, Z. J. Wang, F. H. Arnold, Angew. Chem. Int. Ed. 2015, 54, 3351–3367. R. J. Kazlauskas, U. T. Bornscheuer, Nat. Chem. Biol. 2009, 5, 526–529. N. J. Turner, Nat. Chem. Biol. 2009, 5, 567–573.

© Springer Fachmedien Wiesbaden GmbH 2017 H. Minges, Engineering of Halogenases towards Synthetic Applications, BestMasters, DOI 10.1007/978-3-658-18410-0

127

[41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62]

128

M. Alexeeva, R. Carr, N. J. Turner, Org. Biomol. Chem. 2003, 1, 4133–4137. C. B. Poor, M. C. Andorfer, J. C. Lewis, ChemBioChem 2014, 15, 1286–1289. J. T. Payne, C. B. Poor, J. C. Lewis, Angew. Chem. Int. Ed. 2015, 54, 4226–4230. V. Agarwal, J. Li, I. Rahman, M. Borgen, L. I. Aluwihare, J. S. Biggs, V. J. Paul, B. S. Moore, Environ. Sci. Technol. 2015, 49, 1339–1346. M.-I. Aguilar, HPLC of Peptides and Proteins: Methods and Protocols, Springer Science & Business Media, 2004. N. D. Rawlings, A. J. Barrett, A. Bateman, Nucleic Acids Res. 2012, 40, 343–350. S. Harper, D. W. Speicher, Methods Mol. Biol. Clifton NJ 2011, 681, 259–280. Agilent Technologies, Gene morph II random mutagenesis kit, Instruction Manual. S. R. Maloy, V. J. Stewart, R. K. Taylor, Genetic Analysis of Pathogenic Bacteria: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Plainview, New York, 1996. K. Nishihara, M. Kanemori, M. Kitagawa, H. Yanagi, T. Yura, Appl. Environ. Microbiol. 1998, 64, 1694–1699. P. M. Sharp, E. Cowe, D. G. Higgins, D. C. Shields, K. H. Wolfe, F. Wright, Nucleic Acids Res. 1988, 16, 8207–8211. Agilent Technologies, BL21-CodonPlus Competent Cells, Instruction Manual. S. Kumar, C.-J. Tsai, R. Nussinov, Protein Eng. 2000, 13, 179–191. M. Biasini, S. Bienert, A. Waterhouse, K. Arnold, G. Studer, T. Schmidt, F. Kiefer, T. G. Cassarino, M. Bertoni, L. Bordoli, T.Schwede, Nucleic Acids Res. 2014, 42, 252–258. M. Kataoka, L. P. S. Rohani, M. Wada, K. Kita, H. Yanase, I. Urabe, S. Shimizu, Biosci. Biotechnol. Biochem. 1998, 62, 167–169. Z. Xu, K. Jing, Y. Liu, P. Cen, J. Ind. Microbiol. Biotechnol. 2006, 34, 83–90. S. A. Parke, G. G. Birch, D. B. MacDougall, D. A. Stevens, Chem. Senses 1997, 22, 53–65. A. Weckbecker, W. Hummel, Biocatal. Biotransform. 2006, 24, 380–389. K. Terpe, Appl. Microbiol. Biotechnol. 2003, 60, 523–533. S. Harper, D. W. Speicher, Curr. Protoc. Protein Sci. Editor. Board John E Coligan Al 2001, Chapter 6, Unit 6.6. J. G. Meyer, Hindawi Publishing Corporation, Computational Biology 2014. E. Riedel, C. Janiak, Anorganische Chemie, Walter De Gruyter GmbH & Co KG, 2012.