Developing Biocatalytic Processes for the Synthesis of Nitro Compounds

DEVELOPING BIOCATALYTIC PROCESSES FOR THE SYNTHESIS OF NITRO COMPOUNDS

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2018

To my mother, my father and my wife

ACKNOWLEDGMENTS

First, I would like to thank my committee chair, Dr. Yousong Ding, for his support and direction through these five years in graduate school. I appreciate his immense support, guidance and patience throughout my graduate studies. His constant advice and encouragement were a tremendous source of inspiration for me to complete my dissertation.

I would also like to extend my appreciation to Dr. Robert Huigens III, Dr. Hendrik

Luesch, Dr. Jon Stewart, and Dr. Steven Bruner for graciously agreeing to be members of my dissertation committee. I am grateful for their advice, encouragement and constructive comments on my research over the past several years.

I would also like to thank our collaborators: Dr. Rosemary Loria, Dr. Yucheng

Zhang and Dr. Jose Huguet-Tapia, for providing Streptomyces strains and lending their expertise in the thaxtomin biosynthesis and extraction; Dr. Steven Bruner and Dr. Matt

Burg for the help in the TxtE crystallization; Dr. Robert Huigens III, Ms. Yasmeen

Abouelhassan, Dr. Aaron Garrison and Dr. Akash Basak for providing a collection of antimicrobial compounds and guidance in the anti-biofilm assay. I thank them for their support and fruitful discussions.

I also thank the former and current staff of the Medicinal Chemistry Department -

Ms. Nancy Burgos, Mr. David Jenkins and Ms. Jan Kallman - for their support during my graduate study. I am also grateful to Ms. Laura Faux for providing LC-MS training and troubleshooting.

Finally, I must thank the former and current members in the Ding research lab for their assistance and friendship. I thank Dr. Guang Yang for expanding my repertoire of molecular biology, Dr. Yu Sha for the help in the mass spectrum analysis, Mr. Yi Zhang

and Mr. Guangde Jiang for the help in the NMR analysis, and Ms. Peilan Zhang for the help in the bioassay. I appreciate the technical expertise that everyone has shared as well as the insightful discussions we had. I hope our friendships continue for the years to come.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 9

LIST OF FIGURES ...... 10

LIST OF ABBREVIATIONS ...... 13

ABSTRACT ...... 15

CHAPTER

1 INTRODUCTION TO BIOCATALYSIS...... 17

Biocatalysis: a Practical Alternative to Traditional Catalysis in Chemical Synthesis ...... 17 Methodology Development in the Engineering of Biocatalysts ...... 19 Directed Evolution ...... 19 Rational Design ...... 22 Semi-rational Design ...... 23 De novo Design ...... 25 Whole-cell Based Biocatalysis ...... 26 Cytochromes P450 as Biocatalysts ...... 27 Generalized Catalytic Cycle of P450 ...... 28 Electron Transfer Partners of P450s ...... 29 Development and Applications of P450s as Biocatalysts ...... 30 Developing Biocatalytic Processes for the Nitration Reaction ...... 33 Chemical Nitration Methods ...... 35 Natural Nitration Strategies ...... 36 Putative Catalytic Cycle of TxtE in Aromatic Nitration ...... 38 Research Aims ...... 38

2 CREATION AND BIOCHEMICAL ANALYSIS OF AN ARTIFICIAL SELF- SUFFICIENT CYTOCHROME P450 DIRECTLY NITRATES TRYPTOPHAN* ...... 52

Introduction ...... 52 Results ...... 56 Preparation of Self-sufficient TxtE Variants ...... 56 Catalytic Activity of Fusion Enzymes ...... 57 Biochemical Characterization of TxtE and TxtE-BM3R ...... 58 Design and Production of Chimeric TxtE-BM3R Variants ...... 59 Nitration Performance of Chimeric TxtE-BM3R Variants ...... 61 Conclusion and Discussion ...... 63 Methods and Materials...... 65

General Chemicals, DNA sub-cloning, and Bacterial Strains ...... 65 Construction of Self-sufficient TxtE Variants ...... 66 Construction of TxtE-BM3R Variants ...... 67 Heterologous Expression and Purification of Recombinant Proteins ...... 68 Analytical HPLC Analysis ...... 68 Biochemical Characterization of Self-sufficient TxtE Variants ...... 68 Spectral Analysis of Chimeric TxtE-BM3R Variants ...... 69 Catalytic Activities of Chimeric TxtE-BM3R Variants ...... 70

3 DEVELOPING TXTE AS NITRATION BIOCATALYSTS BY ENGINEERING ITS BINDING POCKET ...... 85

Introduction ...... 85 Results ...... 86 Substrate Scope of TxtE and TB14 ...... 86 Structural Characterization of TB14 Catalyzed Tryptophan Analogues Nitration Products ...... 89 Structural Characterization of TB14 Catalyzed Styrene-analogues Conversion Products ...... 94 Identification of Key Residues for Tailoring TB14 Substrate Scope ...... 95 Preparation and Functional Characterization of TB14 Binding Pocket Mutant Library ...... 96 Substrate Scope of TB14 Variants ...... 98 Conclusion and Discussion ...... 99 Methods and Materials...... 102 General Chemicals, DNA Sub-cloning, and Bacterial Strains ...... 102 Large-scale Enzymatic Synthesis of Nitrated Products ...... 103 Analytical and Semi-preparative HPLC Analysis ...... 103 LC-MS and LC-MS/MS Analysis of Nitrated Products ...... 104 NMR Analysis of Nitrated Tryptophan and Styrene Analogues ...... 104 Marfey’s Derivatization ...... 105

4 DEVELOPING E. COLI CELL FACTORIES FOR THE PRODUCTION OF NITRO AROMATICS ...... 136

Introduction ...... 136 Results ...... 138 Creation of E. coli Strain for the Whole Cell Nitration Reaction ...... 138 Optimization of Heterologous Enzyme Expression ...... 140 Optimization of Fermentation Conditions ...... 140 Production of Nitrated Tryptophan Analogues by Whole Cell Nitration System ...... 142 Conclusion and Discussion ...... 143 Methods and Materials...... 144 General Chemicals, DNA sub-cloning, and Bacterial Strains ...... 144 Construction of Plasmids for Whole Cell Transformation ...... 145 Heterologous Expression and Purification of Recombinant Proteins ...... 146

Analytical HPLC Analysis ...... 146 Whole-cell Biotransformation ...... 147

5 GENERAL CONCLUSION ...... 155

LIST OF REFERENCES ...... 158

BIOGRAPHICAL SKETCH ...... 158

LIST OF TABLES

Table page

1-1 Comparison of the different protein engineering strategies...... 51

2-1 Binding affinity toward L-Trp, coupling efficiency and total turnover numbers of TxtE and its chimeras...... 82

2-2 Primers for construction of self-sufficient TxtE variants...... 83

2-3 Primers for construction of TxtE-BM3R variants with altered linker regions...... 84

3-1 Binding affinities of 20 Trp analogues toward TxtE and TB14...... 131

3-2 Binding affinities of styrene analogues toward TB14...... 132

13 1 3-3 C and H NMR data for 5-F-4-NO2-L-Trp and 4-F-7-NO2- L-Trp...... 133

13 1 1 3-4 C and H NMR data of 4-Me-5-NO2-L-Trp and 4-Me-7-NO2-L-Trp and H NMR data of 4-Me-7-NO2-L-Trp...... 134

3-5 Nitration activities of TB14 Alanine mutants...... 135

4-1 Primers for construction of whole cell transformation plasmids...... 154

LIST OF FIGURES

Figure page

1-1 Selected examples of biocatalysts in industrial applications...... 40

1-2 Overview of approaches for protein engineering...... 41

1-3 Biosynthetic strategies of artemisinin and paclitaxel using whole-cell biocatalysts...... 42

1-4 The P450 structural fold...... 43

1-5 The paradigm for P450 catalyzed hydroxylation and peroxide shunt pathways...... 44

1-6 Selected classes of P450s based on their electron transfer partner proteins. .... 45

1-7 Selected examples of engineered P450s catalyzed chemical transformations. . 46

1-8 Selected nitro chemicals and their applications...... 47

1-9 Selected chemical aromatic nitration approaches...... 48

1-10 Three methods to produce nitro compounds evolved in nature...... 49

1-11 General paradigm for P450 catalyzed hydroxylation and putative mechanism for TxtE-promoted nitration...... 50

2-1 TxtE catalyzes an aromatic nitration reaction on the C4 of L-Trp indole ring...... 71

2-2 Characterization of recombinant TxtE variants...... 72

2-3 Thermostability and pH dependence of TxtE and TxtE-BM3R...... 73

2-4 The pH stability of TxtE and TxtE-BM3R...... 74

2-5 Schematic depiction of chimeric TxtE-BM3R constructs with variable linker length or a swapped loop...... 75

2-6 Construction of TxtE chimeras with variable linker lengths...... 76

2-7 Engineering design of TxtE loop connecting J and K helices...... 77

2-8 SDS-PAGE analysis of purified recombinant proteins...... 78

2-9 CO-reduced difference spectra of chimeric TxtE-BM3R variants and TxtE...... 79

2-10 Relative nitration activity of TxtE and chimeric TxtE-BM3R variants...... 80

2-11 HPLC spectrum of TxtE and chimeric TxtE-BM3R variants relative nitration activity assay...... 81

3-1 Chemicals used in the TxtE and its variants substrate scope screening...... 107

3-2 TxtE and TB14 nitrated Trp and its analogues to varying degrees...... 108

3-3 LCMS analysis of styrene nitration reaction catalyzed by TB14...... 109

3-4 Styrene analogue library screened in the current research...... 110

3-5 Relative nitration activities of TB14 towards styrene analogues...... 111

3-6 The MS2 spectra of 4-NO2-L-Trp, nitrated 5-F-L-Trp, and nitrated 4-F-DL- Trp...... 112

3-7 Putative fragmentation pathways of 4-NO2-L-Trp, 4-NO2-5-F-L-Trp, and 7- NO2-4-F-L-Trp...... 113

3-8 HRMS spectra of nitrated 5-F-Trp and 4-F-Trp in TB14 reactions...... 114

3-9 1H NMR spectra of nitrated 5-F-Trp and 4-F-Trp products...... 115

3-10 13C NMR spectra of nitrated 5-F-Trp and 4-F-Trp products...... 116

3-11 HSQC NMR spectra of nitrated 5-F-Trp and 4-F-Trp products...... 117

3-12 HMBC NMR spectra of nitrated 5-F-Trp and 4-F-Trp products...... 118

3-13 LC-MS analysis of Marfey’s derivatization of 4-Me-DL-Trp and its nitro product...... 119

3-14 LC-HRMS analysis of the isolated 4-Me-NO2-L-Trp product...... 120

3-15 Both 4-Me-5-NO2-L-Trp and 4-Me-7-NO2-L-Trp were produced in the TB14 reaction with 4-Me-DL-Trp as substrate...... 121

1 3-16 H NMR spectrum of the isolated 4-Me-NO2-L-Trp product...... 122

3-17 COSY spectrum of the isolated 4-Me-NO2-L-Trp product...... 122

13 3-18 C NMR spectrum of the isolated 4-Me-NO2-L-Trp product...... 123

3-19 HSQC spectrum of the isolated 4-Me-NO2-L-Trp product...... 123

3-20 HMBC spectrum of the isolated 4-Me-NO2-L-Trp product...... 124

3-21 Expansion of HMBC spectrum in aromatic and aliphatic regions and the representative correlations on 4-Me-5-NO2-L-Trp...... 124

3-22 1H NMR spectrum of the isolated 4-methoxystyrene converted product and proposed product structure...... 125

3-23 1H NMR spectrum of the isolated 4-vinylpyridine converted product and proposed product structure...... 126

3-24 Selected 20 amino acid residues lining the TxtE L-Trp binding pocket...... 127

3-25 SDS-PAGE analysis of purified TB14 mutants...... 128

3-26 Relative nitration activities of TB14 variants...... 129

3-27 Relative nitration activities of TB14 variants with styrene as the substrate...... 130

4-1 Schematic overview of bacterial cell factories for the production of nitro- chemicals...... 148

4-2 Creation and initial activity test of the whole cell nitration system...... 149

4-3 Creation and analysis of different whole cell nitration systems...... 150

4-4 Nitrated tryptophan concentration in the whole cell nitration system supported by different types of medium...... 151

4-5 Nitrated tryptophan concentration in the whole cell nitration system fermented at different temperatures...... 152

4-6 Nitrated tryptophan analogues production by the whole cell nitration system. . 153

LIST OF ABBREVIATIONS

ALA 5-Aminolevulinic acid

BM3R Cytochrome P450 reductase domain from Bacillus megaterium (CYP102A1)

COSY Homonuclear correlation spectroscopy

CPR Cytochrome P450 reductase

CYP Cytochrome P450 monooxygenase dNTPs Deoxynucleotide triphosphates

DMSO Dimethyl sulfoxide

ESI Electrospray ionization

FAD Flavin adenine dinucleotide

FeS Iron sulfur cluster

FMN Flavin mononucleotide

Fer Ferredoxin

Frd Ferredoxin reductase

HRMS High-resolution mass spectrometry

HMBC Heteronuclear multiple-bond correlation spectroscopy

HSQC Heteronuclear single quantum coherence spectroscopy

HPLC High performance liquid chromatography

IPTG Isopropyl β-D-1-thiogalactopyranoside

LCMS Liquid chromatography-mass spectrometry

MW Molecular weight

NADH Nicotinamide adenine dinucleotide, reduced

NADPH Nicotinamide adenine dinucleotide phosphate, reduced

NMR Nuclear magnetic resonance

NOC-5 3-(Aminopropyl)-1-hydroxy-3-isopropyl-2-oxo-1-triazene

P450 Cytochrome P450

PCR Polymerase chain reaction

PDB Protein Data Bank

Pdr Putidaredoxin reductase

Pdx Putidaredoxin

PFOR Phthalate family oxygenase reductase

RhF P450RhF (CYP116B2)

RhFRed Reductase domain from P450RhF

SDS-PAGE Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

TTN Total turnover number

Trp Tryptophan

WT Wild type

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

DEVELOPING BIOCATALYTIC PROCESSES FOR THE SYNTHESIS OF NITRO COMPOUNDS

Ran Zuo

May 2018

Chair: Yousong Ding Major: Pharmaceutical Sciences

Nitro chemicals are among the largest and most important families of industrial chemicals and have broad applications in industry, agriculture, and health. However, industrial nitration processes are notorious sources of pollution and pose a number of selectivity, environmental and safety issues. Previous attempts to develop practical and advanced solutions illuminate the fundamental challenges of green nitration and highlight the long-standing needs of the field. By contrast, enzymes as biocatalysts offer many notable advantages over traditional chemical catalysts because of their high activities, impressive regio- and stereo-selectivity, and environmental benignness.

However, the nitration reaction, especially aromatic nitration, has benefited little to none from biocatalysis research yet. The overall objective of this research project is to develop biocatalytic processes for the synthesis of nitro compounds, therefore filling the lasting deficiency in the nitration field. Specifically, we aim to harness the power of a distinct bacterial cytochrome P450 TxtE. TxtE is able to nitrate at C4 of L-Trp within the biosynthetic pathway of phytotoxin thaxtomin. To develop this enzyme for chemical nitration, we first fused TxtE with the reductase domains of P450BM3 and P450RhF to create artificial self-sufficient biocatalysts. Biochemical characterization identified one

chimera, TxtE-BM3R, was slightly more active than wild type TxtE supplemented with spinach ferredoxin and ferredoxin reductase. To further improve the catalytic performance of TxtE-BM3R, we then created fifteen new chimeras by rationally replacing a catalytically important loop and varying the length of the linker connecting

TxtE and BM3R domains. Among them, TB14 showed significantly improved nitration activity, coupling efficiency and total turnover number. We further assessed substrate scope of TxtE and TB14, and revealed that enzyme’s regio-selectivity was controlled by unnatural tryptophan analogues. To further improve TB14 properties as nitration biocatalysts, we identified 19 residues lining its substrate binding pocket, and mutated each with alanine. Several variants showed altered substrate specificity. Unexpectedly, some variants demonstrated substantial catalytic activity in nitrating styrene and its analogues, whose structures are not similar to tryptophan. Finally, we employed TB14, nitric oxide synthase and glucose dehydrogenase to construct an artificial metabolic pathway for producing nitrated L-Trp and L-Trp analogues in a cell factory.

CHAPTER 1 INTRODUCTION TO BIOCATALYSIS

Biocatalysis: A Practical Alternative to Traditional Catalysis in Chemical Synthesis

Biocatalysis is the application of enzymes and microbes in synthetic chemistry1.

Microbes, crude cell extracts, and isolated enzymes all can be applied as biocatalysts for useful chemical transformations. Biocatalysts are considered as an important alternative to traditional chemical catalysts since they offer attractive advantages over conventional catalysts. Similar to other catalysts, biocatalysts increase the speed in which a reaction takes place but do not affect the thermodynamics of the reaction. The most important advantage of biocatalysts is their high selectivity, including chemo-

selectivity, regio-selectivity and stereo-selectivity2. The high selectivity allows

biocatalytic process to avoid the lengthy protection/deprotection steps and the

generation of side-products and impurities, which also ease the separation step.

Another significant advantage of biocatalysts is their environmentally-friendly nature.

They are completely biodegradable in the environment, operated under milder reaction conditions, non-toxic and leave no residual heavy metal contamination. In addition, there is an enzyme-catalyzed reaction equivalent to the majority of known organic

reactions. These properties have resulted in their increasing industrial applications and

led to the development of new biocatalysts for unmet needs.

Biocatalysis development can be divided into several stages. Since hundreds of

years ago, biocatalysts have been used to catalyze natural substrates into natural

products as their native functions. For example, they have long been used in the

fermentation process to produce alcohol and cheese. It is considered as the first stage

of biocatalysis1. From 1980s to 1990s, the development of protein engineering

technologies allowed the transformation of non-natural substrates by using engineered biocatalysts. This stage of biocatalysis development was based on the observation that many enzymes catalyze a set of substrates with similar or diverse structures, therefore possessing relaxed substrate requirement. The activities of enzymes used in the

biocatalytic processes were usually not altered. A number of biocatalysts developed in

this stage have found industrial applications. For example, a lipase from Serratia

marcescens was used to synthesize a key chiral intermediate of diltiazem

hydrochloride, a coronary vasodilator (Figure 1-1A), and the biocatalytic process was

more efficient than the conventional chemical route3. Oxynitrilases were also used to

synthesize enantiomerically pure cyanohydrins (Figure 1-1B), which are important

synthetic intermediates for pharmaceuticals and agrochemicals4.

The development of advanced protein engineering technologies in the mid and

late 1990s (including directed evolution, rational design, semi-rational design, and de

novo design) has further expanded the uses of biocatalysts in the manufacture of a

series of fine chemicals, especially pharmaceuticals and their intermediates. New

desirable traits of biocatalysts have been achieved in this new stage. Novel biocatalysts

have the ability to remain stable in industrial manufacturing conditions, can accept new

substrates and catalyze new reactions with efficiency and selectivity. For example, a

ketoreductase (KRED) engineered via directed evolution technologies has been used to

asymmetrically reduce (E)-methyl 2-(3-(3-(2-(7-chloroquinolin-2-yl)vinyl)phenyl)-3-

oxopropyl)benzoate to the corresponding (S)-alcohol, a key intermediate in the

synthesis of montelukast sodium (Singulair), with very high enantioselectivity (>99.9%

ee) and at >200 kg scale5. (Figure 1-1C) Another successful example is the biocatalytic

process developed by researchers from Merck, Solvias and Codexis for the industrial-

scale manufacture of the antidiabetic drug sitagliptin. (Figure 1-1D) Starting from an

enzyme that had the catalytic machinery to perform the desired chemistry but lacked

any activity toward the prositagliptin ketone, they applied a substrate walking, modeling,

and mutation approach to create a transaminase with marginal activity for the synthesis

of the chiral amine. This variant was then further engineered via directed evolution for

practical application in a manufacturing setting6, 7.

Methodology Development in the Engineering of Biocatalysts

Despite the impressive potential of biocatalysts in chemical synthesis, the low

activity, narrow substrate scope, and instability of native enzymes under the process

conditions for industrial production are often considered as the drawbacks of

biocatalysts8. In this regard, protein engineering approaches have been developed to

tailor characteristics of native enzymes, including but not limited to activity, stability, selectivity and substrate scope. Directed evolution, rational design, semi rational design

and computational design are the main strategies that have been developed for protein

engineering (Figure 1-2). These methods can be applied to engineer single enzymes,

whole-cell biocatalysts, or even metabolic pathways for industrial, especially

pharmaceutical, manufacture processes. For each protein engineering strategy, the

advantages and disadvantages are summarized in Table 1-1.

Directed Evolution

The directed evolution strategy combines the generation of a large and random

mutant library with the identification of mutant/mutants with desired property. The most

significant advantage of the directed evolution strategy is that it does not need prior

information of the structures, functions or sequence-function-relationship of the parent

protein. In this strategy, the natural evolution is mimicked as random mutations are introduced into the gene sequence of the target enzyme9.

After choosing the suitable parent enzyme, a mutant library is created from its

gene. In order to achieve the library diversity, random mutagenesis or DNA

recombination is usually applied. For random mutagenesis, the error-prone PCR

(epPCR) is the most widely used method. Using low-fidelity DNA polymerases,

replacing Mg2+ with different concentrations of Mn2+ and adding uneven concentrations

of four deoxynucleotide triphosphates (dNTPs) are the common ways to introduce and

adjust the mutation rate of the parent gene. By epPCR, only a few mutations are usually

introduced to the parent gene. DNA shuffling is another method developed for the

generation of in vitro DNA recombination libraries. It can be used to create gene

swapping segment and/or multiple mutations10. To perform DNA shuffling, the first step

is to fragment the target gene and its homologous or mutant gene/genes with DNaseI.

The gene fragments are then pooled together to assemble the full-length gene by

homologous region annealing between these fragments. Additions to these above two

main approaches, their variations have also been developed for protein engineering.

For example, sequence saturation mutagenesis is one widely used variation of random

mutagenesis11, while incremental truncation for the creation of hybrid enzymes12, non-

homologous random recombination13 and random multi-recombinant PCR14 are three

reported technologies that eliminate the requirement for homogenous genes. In vivo

directed evolution has also been developed in complementary to the above mentioned

in vitro directed evolution methods. For example, phage-assisted continuous evolution

was developed in E. coli15, and heritable recombination was developed in yeast16.

The development of an efficient and convenient screening protocol to identify desired mutants is equally, if not more, important prerequisite for the success of the directed evolution strategy. Traditionally, the spectral analysis of reaction mixtures in

96-well microplates or even direct observation has been used for screening, and is still commonly used today. Recently, researchers have made significant advancement in other high-throughput screening or even ultra high-throughput screening protocols.

These achievements tremendously promoted the progress of directed evolution

methods. For example, Becker et al.17 developed a single cell high-throughput

screening method by labeling two enantiomers with different fluorescent dyes. Highly

enantioselective hydrolytic enzymes displayed on cell surfaces were identified by real- time analysis of the ratio of green and red single-cell fluorescence. Another versatile ultra high-throughput method developed by Weitz et al.18 combined drop-based

microfluidics screening with fluorescence-activated cell sorting. This super-efficient

method used aqueous drops dispersed in oil as picoliter-volume reaction vessels and

screened them at the rate of thousands per second. It has been used to screen

horseradish peroxidase random mutagenesis libraries and identified mutants exhibiting

catalytic rates more than 10-fold faster than their parents in only 10 h, using less than

150 μL of total reagent volume.

Coupling the targeted enzymatic reactions with cell viability, the selection

appears to be another convenient and efficient strategy. In order to screen an

enantioselective esterase mutant library generated by directed evolution, Fernandez-

Alvaro et al.19 coupled one enantiomer of 3-phenyl butyric acid to 2,3-dibromopropanol

and the other one to glycerol. The esterase mutants that produce the desired

enantiomer will release glycerol, which serves as carbon source and supports the host

growth. In the meantime, the esterase mutants that produce the undesired enantiomer

will release 2,3-dibromopropanol, which serves as toxic compound and kills the host.

Combining this in vivo selection method with cell sorting, E. coli clones harboring the

desired enantioselective esterases were identified.

Rational Design

Unlike the directed evolution strategy, the rational design method relies on the

knowledge of the parent enzymes, especially on the 3D structures. Enzymes have

evolved different biochemical properties while still possess similar 3D structures.

Structural superposition and amino acid sequence alignment could help researchers to

rationally design mutations on the parent enzymes. For example, based on the

structural superposition and amino acid sequence alignment of cancer-associated

mutations from isocitrate dehydrogenases with homoisocitrate dehydrogenases, Yan et

al.20 identified homologous residues to the human mutational hotspots. R114Q, R143C,

R143H and R143K were then designed to catalyze the conversion of 2-oxoadipate to

(R)-2-hydroxyadipate. After characterization, three of them, R143C, R143H and R143K,

were found to be highly enantioselective (R)-2-hydroxyadipate dehydrogenases. In

another example, although human AKR1D1 catalyzes the 5β-reduction of D4-3-

ketosteroids, while AKR1C is a hydroxysteroid dehydrogenase, these enzymes possess

high similar amino acid sequences. Based on the sequence alignment, one single

residue of AKR1D1, His120, was changed to glutamate and the resulting enzyme was

switched as a 3β-hydroxysteroid dehydrogenase21.

Besides residues identified by homologous comparison, the residues in the

enzyme substrate binding pockets may play important roles in the substrate selectivity

and catalytic activity, making them excellent targets for engineering. For example, transketolases can catalyze the transfer of a two-carbon keto group to the aldose. In order to engineer an E. coli transketolase to accept aromatic aldehydes as substrates,

Dalby et al.22 used 3-formylbenzoic acid and 4-formylbenzoic acid as molecular probes

along with transketolase mutants to reveal the factors governing the rate of reaction between transketolase and aromatic aldehydes. Residues around the phosphate binding pocket were chosen and subjected to mutagenesis. After characterization, one of the mutants exhibited up to 250-fold higher specific activities toward benzaldehyde.

Semi-rational Design

The directed evolution strategy has the almost unlimited potential to create desirable mutants. However, the screening of a large library can be very time and cost consuming. On the other hand, the rational design generates only a small library and thus requires much less screening efforts. But the design may not be perfect which could in turn lead to failure in obtaining desirable mutants. The semi-rational design strategy is a combination of directed evolution and rational design. It aims to minimize the library size based on rational design, while at the same time maintain the library diversity with the help of directed evolution. Semi-rational design is to date the most successful method in engineering biocatalysts for pharmaceutical processes.

SCHEMA, the protein Sequence Activity Relationships (PROSAR), and iterative saturation mutagenesis are the three most widely used semi-rational design methods.

As a random mutagenesis approach, the DNA shuffling generates a great proportion of misfolded mutants in the libraries. The SCHEMA algorithm was designed by the Arnold group in order to reduce the library size created by DNA shuffling23. This algorithm

requires structural information of the parent enzymes. It reduces the proportion of

misfolded mutants by locating side-chain interactions and avoids generating chimeras

with damaged interactions during enzyme recombination23. It has been employed in the

engineering of fungal cellulose and cellobiohydrolase, and mutants with enhanced

thermostability have been obtained from a library with a smaller size and larger fraction

of properly folded enzymes24, 25.

The proSAR method incorporates statistical analysis into recombination-based

directed evolution26. By statistical analysis, it assigns mutations as ‘beneficial’,

‘potentially beneficial’, ‘neutral’ or ‘deleterious’, and aims to accumulate only the

beneficial mutations. It is a mutation-oriented enzyme optimization strategy based on

the analysis of sequence-activity data. Even beneficial mutations in variants with

reduced function can be identified by this method. The authors evolved a bacterial

halohydrin dehalogenase by this method and obtained variants that improve the

volumetric productivity of a cyanation process approximately 4,000-fold26.

Iterative saturation mutagenesis method focuses on residues which play

important roles in the activity of the parent enzymes, such as substrate binding or

thermostability. After identifying such amino acid residues, saturation mutagenesis is performed on them and mutants obtained are subjected to screening27, 28. Hoffmann et

al. selected nine key amino acid residues out of 13 residues in the active site of

P450cam monooxygenase29. After performing mutagenesis on these residues, a

focused library comprising 300,000 protein variants was generated and screened for activity on diphenylmethane.∼ Five mutants were found to be able to transform diphenylmethane with a specific activity of up to 75% of the wild-type activity on D- camphor.

Other semi-rational design methods are available. For example, iterative directed evolution of enzymes can reach a plateau and fail to further improve the enzymes.

Whitehead et al. obtained comprehensive sequence-function maps by deep sequencing and combined many individually small contributions not detectable in conventional approaches30. They used this approach to optimize two computationally designed

inhibitors against an influenza hemagglutinin and obtained variants with sub-nanomolar

binding affinity.

De novo Design

Besides the above-mentioned enzyme engineering strategies using natural

enzymes as starting materials, de novo design of enzymes is also a feasible strategy31.

The most successful and widely used de novo design method is the Rosetta method. It

generally requires four steps: the first step is to characterize the minimal active site for

the chosen catalytic mechanism; the second step is to find scaffold proteins and the specific sites where the minimal active site can be incorporated; the third step is to optimize surrounding residues to stabilize interactions with the transition state and primary catalytic residues; the fourth step is to evaluate and rank the variants obtained from the design. Finally, the variants with the best scores will be subjected to experimental evaluation32. The biochemical properties such as the activity of the de

novo designed enzyme are often not ideal, so it is very common that a directed

evolution is subsequently performed to improve the enzyme. This Rosetta method has

successfully been employed to create retro-aldol enzymes, triosephosphate isomerase,

stereoselective Diels-Alder enzymes and so on32-34.

Whole-cell Based Biocatalysis

Whole-cell biocatalysis uses entire microorganisms instead of purified enzyme to catalyze desired reactions35. Compared to in vitro reactions, the whole-cell biocatalysis minimizes the usage of cofactors, simplifies the catalysis process, and has advantages on catalyzing multi-step reactions35. What’s more, metabolic engineering in combination

with protein engineering greatly facilitates the development of whole-cell biocatalysts.

Whole-cell biocatalysts are particularly useful for the production of

pharmaceuticals derived from the natural source due to complicated multi-step

biosynthetic routes, structure complexity, and limited natural supplies of these

compounds. One of the most successful examples of the whole-cell biocatalysis in the

pharmaceutical manufacture is the production of semisynthetic anti-malarial artemisinin,

developed by the Jay Keasling’s group. Artemisinin can be extracted from plant

Artemisia annua. The supply and price of artemisinin thus fluctuated greatly.

Researchers from the Jay Keasling’s group transplanted the artemisinin biosynthesis

genes to yeast. After metabolic engineering to redirect the carbon flux to the artemisinin

biosynthesis pathway, the production of artemisinic acid, the immediate precursor of

artemisinin, achieved 1 g/L using glucose as starting material36, 37. Further metabolic

and process engineering provided a complete process for the production of

semisynthetic artemisinin, and the production amorpha-4,11-diene, another artemisinin precursor, was increased to 40 g/L38 (Figure 1-3). In addition, the development of

continuous-flow conversion of artemisinic acid to artemisinin made the industry

application more efficient and economic39. Protein engineering also helped the

artemisinin synthesis development. P450BM3 from Bacillus megaterium was semi-

rationally engineered to convert amorpha-4,11-diene to artemisinic-11S,12-epoxide,

providing an alternative route for the semisynthetic artemesinin production40 (Figure 1-

3). Researchers from the Rudi Fasan’s lab also engineered P450BM3 by active site mutagenesis. The selected mutants possessed refined regio- and stereoselectivity for the hydroxylation of unactivated C-H sites in artemisinin at preparative scales and high yields41.

Similar efforts have been made to produce taxadiene, a precursor of paclitaxel

(Taxol). Researchers from the Stephanopoulos’ group introduced a heterologous

terpenoid biosynthetic pathway to E. coli, and after metabolic engineering, the strain

could produce taxadiene at approximately 1 g/L42 (Figure 1-3). Protein engineering has

also been done on Taxus cytochrome P450 for the hydroxylation of taxadiene to

taxadien-5α-ol. The mutant yielded by fusions with redox partners and transmembrane

optimization showed higher efficiency. However, as the taxol biosynthetic pathway is

still not fully elucidated, there is no complete biosynthetic or semisynthetic route

developed yet. Currently the plant cell fermentation and direct extraction methods are

still more advantageous for the taxol production43.

In summary, the whole-cell based biocatalysis development includes three

aspects: the first is to transplant the biosynthetic pathway genes into suitable hosts;

secondly, metabolic engineering is applied to optimize the metabolic flux and improve

the productivity; thirdly, protein engineering is often incorporated when necessary to

improve the productivity, generate new functions and/or expand the substrate scope8.

Cytochromes P450 as Biocatalysts

Cytochromes P450 are a superfamily of heme-thiolate monooxygenases. The

name P450 is derived from the spectrophotometric absorption peak at the wavelength

of 450 nm when it is exposed to CO compared with the enzyme in the reduced state.

P450s share a conserved heart-shaped fold (Figure 1-4). The heme region and I-helix

are the two most conserved substructures. They form part of the substrate-binding

pocket of the enzymes and facilitate dioxygen activation at the central heme iron. The

FG-helices and B region are generally flexible; they form the dome of the substrate-

binding pocket and determine the binding and selectivity of substrates. The substrate

access channels to the active site pocket are also formed by the movement of these

variable B regions44.

P450s are widely spread in all domains of life including bacteria, archaea and

eukarya, and genome sequencing even revealed two putative virally encoded P450s45.

They fulfill a wide range of physiological roles in the host cells such as natural product biosynthesis, metabolism of xenobiotics and assimilation of carbon sources for growth46.

The wide array of natural P450s catalyzed reactions is exemplified by hydroxylation,

sulfoxidation, epoxidation, reductive dehalogenation and aryl-aryl coupling, oxidation47,

48. The versatile functions and wide distribution make P450s a rich source for

biocatalysts which can be applied to various synthetic applications. The most important

feature of P450s, from the biocatalysis perspective, is that they can catalyze the inert C-

H bond oxidation with high selectivity, a reaction in most cases difficult for chemical

catalysts49. P450s thus have attracted many researchers’ interests and been

extensively studied and engineered in the past two decades.

Generalized Catalytic Cycle of P450

P450s generally use O2 and a reducing cofactor (NADPH/NADH) to insert one

oxygen atom into an organic substrate, often the inert C-H bond. Remarkably, P450s’ substrates exhibit the most structural diversity of all known enzyme families and have consistently expanded over the past 50 years50. General mechanisms of the P450s

catalytic cycles have been well investigated and some key, highly reactive intermediates have been spectroscopically confirmed51-54 (Figure 1-5). In brief, substrate binding to

the resting low-spin ferric (Fe3+) enzyme (1) is the first step. Binding leads to the

dissociation of one water molecule distally coordinated to heme iron, the conversion of

the heme from low to high spin (2), and allowance of an associated reductase to reduce

2+ the heme to the ferrous (Fe ) state (3). Next, O2 binds to the ferrous heme, forming

ferric superoxide complex (4). A ferric peroxo species (5) is then formed after reduction

with the second electron, and subsequent protonation of its distal oxygen generates a

ferric hydroperoxo complex (6). Delivery of an additional proton to the distal oxygen

cleaves the O-O bond, yielding compound I (7). A highly conserved Thr residue is

widely believed to facilitate two protonation steps. Compound I is able to abstract

hydrogen from the substrate to yield both compound II (8) and a substrate radical. The

rapid recombination of two radicals yields a hydroxylated product. The resting ferric

enzyme (1) is regenerated after product dissociation and then water coordination to the

heme. This P450 cycle leads to formation of hydroxylated or epoxidized products, but

also promotes a number of uncommon processes.

Electron Transfer Partners of P450s

P450s need two electrons supplied from NAD(P)H in a stepwise manner to

activate O2. The electron transfer systems are diverse among P450s and usually

contain one or two proteins. Hannemann et al.55 classified P450s into ten different

classes by the electron transfer systems and here we describe four most important and

common systems as shown in Figure 1-6. Most of the bacterial P450s, for example,

P450cam and TxtE, belong to Class I. In this class, the electrons flow from NAD(P)H to

ferredoxin reductase (FAD redox center), then transferred to ferredoxin (FeS redox

center), and finally to the heme prosthetic group of P450s. All three components of this class are separated proteins. Class II P450s are membrane bound, which are mostly comprised of non-mitochondrial mammalian enzymes. Class II system includes two components: the P450 and cytochrome P450 di-flavin reductase (CPR). Class VII and class VIII are catalytically self-sufficient P450s. P450s of this class are featured with electron transfer components fused to the P450 catalytic domain, forming a single

polypeptide. P450BM3 of Class VIII was discovered from Bacillus megaterium, and is

the first self-sufficient P450s found in nature56, 57. The reductase domain of this class of

P450s is a di-flavin reductase, similar with eukaryotic CPR. Class VII is another self-

sufficient P450s system with flavin-ferredoxin reductase fused with the P450 catalytic

domain. The prototypical Class VII enzyme P450RhF was isolated from a Rhodococcus

species58. Three components of Class VII enzyme, P450 catalytic domain - ferredoxin reductase domain bearing a flavin mononucleotide (FMN)-ferredoxin domain bearing a

2Fe-2S cluster, are linked in the order of N to C terminus55, 59 (Figure 1-6).

Inspired by natural P450-redox fusion proteins, many investigators have sought

to generate new artificial fusions, an area of research that has been reviewed59-61. By

fusing P450s to a redox partner the need to express and process multiple proteins

separately is removed. The control of relative expression levels is also simplified, and

P450 and reductase are produced in equivalent amounts. Fusion may also improve

catalytic performance as suggested by the unmatched speed and efficiency of

P450BM3.

Development and Applications of P450s as Biocatalysts

Using the above catalytic cycle, P450s achieve regio- and stereo-selective oxidations of a wide range of organic compounds (e.g., natural products and

xenobiotics), and have attracted increasing interest in engineering and applying them for the industrial synthesis of value-added products62.

Besides the aforementioned P450 engineering aimed to catalyze three

successive oxidation reactions to produce artemisinic acid, another successful industrial

application of P450 is the biocatalytic process for pravastatin synthesis63. Pravastatin is

a potent cholesterol-lowering drug and can be produced from the fungal metabolite

compactin after 6β-hydroxylation. CYP105AS1 from Amycolatopsis orientalis was

isolated to catalyze the final compactin hydroxylation step, but the predominant

compactin binding modes lead mainly to the ineffective epimer 6-epi-pravastatin. The

enzyme was then evolved to invert stereoselectivity (96:4), and the optimized mutant

P450Prava fused to a redox partner in compactin-producing P. chrysogenum yielded

more than 6 g/L pravastatin at a pilot production scale63 (Figure 1-7). Furthermore,

1α,25-dihydroxyvitamin D3 is an important drug for hypothyroidism, osteoporosis, and

chronic renal failure. CYP107 and CYP105A1 have been engineered to replace the

lengthy and low-efficient chemical synthesis to produce this chemical from vitamin D364.

Besides natural reactions, novel chemistries which are not found in nature have

been introduced to the engineered P450s65 (Figure 1-7). Inspired by the capacity of

metalloporphyrin complexes to catalyze carbene and nitrene transfer reactions, Coelho

and coworkers used cytochrome P450BM3 and other heme-containing proteins to

catalyze the cyclopropanation of styrene66, with ethyl diazoacetate (EDA) as a carbene

precursor (Figure 1-7). They further demonstrated that the abiotic activity and

stereoselectivity of the cyclopropanation biocatalyst could be evolved and improved by

replacement of the conserved P450 axial cysteine residue with serine67 or histidine68.

Notably, one P450BM3 variant, HStar, was applied in whole-cell reactions to produce the cyclopropane core of levomilnacipran with 86% isolated yield and 96% de and 92%

ee68. Besides P450BM3, P450s involved in bacterial thiopeptide biogenesis pathway

were also engineered to selectively cyclopropanate dehydroalanines in several complex

thiopeptide-based substrates69, 70. For example, several internal, aliphatic and electron-

deficient alkenes not accepted by Fe-containing enzymes could undergo

cyclopropanation when catalyzed by Ir-substituted CYP119 variants71.

Carbene C-H insertion was also realized by replacing the heme with Ir(Me)-

mesoporphyrin IX in CYP11972-74. Variants of CYP119 were found to catalyze

intramolecular carbene C-H insertion of various ethyl 2-diazo-2-(2-methoxyphenyl) acetate derivatives and intermolecular carbene C-H insertion reaction with phthalan and

EDA (Figure 1-7).

P450s were demonstrated to be capable of activating arylsulfonyl azide and carbonazidates via intramolecular C-H amination reaction, yielding sultam products or aminoalcohols (Figure 1-7)75-79. Recently, P450-mediated intermolecular amination of

benzylic C-H bonds was also realized (Figure 1-7). Starting from a cytochrome P411

variant with activity of amination of 4-ethylanisole with tosyl azide, the authors obtained

variant P411CHA with up to 1300 TTN and 99% ee for benzylic C-H amination of a

range of alkylarene substrates after several rounds of directed evolution80, 81. In another study, the serine-ligated cytochrome ‘P411’ variants were also found to be able to catalyze intermolecular nitrene addition to alkenes (aziridination) (Figure 1-7)82.

Although not based on P450s, another worth mentioning breakthrough is the

development of heme proteins to catalyze the formation of organosilicon compounds via

carbene insertion into silicon–hydrogen bonds. Directed evolution of cytochrome c from

Rhodothermus marinus achieve >15-fold higher turnover than chemical catalysts while

accepting a wide range of substrates with high chemo- and enantioselectivity83. Later,

this enzyme was also engineered to form carbon-boron bonds in the presence of

borane-Lewis base complexes, through carbene insertion into boron-hydrogen bonds.

Directed evolution yielded variants with up to 15,300 turnovers, turnover frequency of

6,100 h-1, 99:1 enantiomeric ratio and 100% chemoselectivity. A total of 16 novel chiral

organoboranes were produced by these engineered biocatalysts84.

Considering recent major advances in P450 applications and engineering

methodologies, together with increasing needs from synthetic industry, more novel

pharmaceutical and biotechnological applications with engineered P450s are expected

in the coming years.

Developing Biocatalytic Processes for the Nitration Reaction

Compound containing one or more nitro groups (-NO2) are among the largest

and most important families of industrial chemicals, as evidenced by their annual global

production of greater than 108 tons85. These chemicals have broad applications as

explosives, dyes, polymers, plastics, perfumes, food additives, pesticides, herbicides,

and pharmaceuticals86 (Figure 1-8A). Easy transformation of the nitro group to other

functionalities merits the wide application of nitro chemicals as building blocks for the

synthetic generation of various complex molecules86. In addition, nitro aromatic

prodrugs have very important applications in gene-directed enzyme prodrug therapy

(GDEPT) for cancer treatment. GDEPT operates via three key steps: (1) tumor-specific

delivery and expression of a therapeutic gene; (2) conversion of inert prodrugs into

cytotoxins by the expressed enzyme within target cells; and (3) diffusion of activated

cytotoxins to neighboring tumor cells. One of the GDEPT strategies utilizes nitroreductase enzymes as they can activate nitro aromatic prodrugs to DNA-damaging products that are cytotoxic to tumor cells87. To date, there are already a number of nitro

aromatic prodrugs with potential utility in GDEPT which have been extensively studied

(Figure 1-8B). Some have undergone clinical development as hypoxia-targeting

chemotherapeutics or have even achieved clinical approval as antibiotics

(metronidazole and tinidazole, for example). The most widely-studied nitro aromatic

prodrug for GDEPT is CB1954 (5-(aziridin-1-yl)-2,4-dinitrobenzamide) which has

undergone phase II clinical trial. CB1954 was converted into 2- and the 4-hydroxylamine

derivatives by nitroreductase. The 2-hydroxylamine derivatives subsequently disproportionate to yield the 2-amine and 2-nitroso derivatives which are capable of forming DNA mono-adducts. The 4-hydroxylamine derivatives, on the other hand,

undergo a spontaneous reaction with acetyl-CoA and form a bi-functional alkylating agent capable of cross-linking DNA and thus disrupt the replication fork upon mitosis88-

90. Dinitrobenzamidemustards (DNBMs) have also received much attention as next- generation nitro aromatic prodrugs. PR-104 [2-((2-bromoethyl)(2,4-dinitro-6-((2-

(phosphonooxy)ethyl)carbamoyl)phenyl)amino)ethyl methanesulfonate] of this class has advanced to clinical trials. PR-104 is firstly converted into the DNBM alcohol PR-104A

[2-((2-bromoethyl)(2-((2-hydroxyethyl) carbamoyl)-4,6-dinitrophenyl)amino)ethyl methanesulfonate]. PR-104A was then reduced by nitroreductase to yield hydroxylamine and amine products which can cause inter-strand DNA cross-links91, 92

(Figure 1-8B).

Chemical Nitration Methods

Given the paramount importance of nitro chemicals, particularly nitro aromatics,

nitration is one of the most widely studied organic reactions93. Principles of industrial

nitration, however, have literally remained unchanged since their implementation more

than 100 years ago93. Aromatic nitration still relies on the classical electrophilic nitration

method that involves an excess of nitric acid and other acid catalysts and /or dinitrigen

+ pentoxide (N2O5). The nitronium ion, NO2 , is believed to be the active species for this

aromatic nitration, albeit with the potential minor contribution of a radical (NO2•)

mechanism93. However, industrial nitration approaches are notorious for their polluting

process, which generate large quantities of acidic waste and nitrogen oxide fume.

Additional persistent issues include poor regio-selectivity, especially for mono-

substituted aromatics; low chemo-selectivity due to over-nitration and oxidation; and limited functional group and/or substitute compatibility. These issues prevent industrial sectors from expanding the uses of existing and new nitro compounds and highlight pressingly industrial and societal need for green chemical processes leading to high quality products.

In recent years, several advanced nitration methods93-97 have been developed to

address the above issues in aromatic nitration. The first strategy relies on the prefunctionalization of C-H bonds in substrates and includes approaches of (1) ipso- nitration by the nitrodemetalation of an aryl C-M bond98, nitrodecarboxylation of an aryl

carboxylic acid99, and transition metal-mediated cross coupling of aryl halides, pseudohalides, or aryl boronic acids100; and (2) the ipso-oxidation of an amino or azide group101. The second focuses on the use of transition metals for direct regio-selective

aromatic C-H nitration with the assist of directing groups in the substrates102-105. This

strategy is advanced concurrently with the development and use of various green

nitrating agents, which is the third strategy. Both inorganic nitrating salts such as

106 107 AgNO3 and organic nitrating compounds have successfully nitrated aromatic

substrates108. Finally, the fourth strategy features the application of novel reaction components, such as recyclable and less pollutant solid acid catalysts109-111 and ionic liquids94, 95, 97. These advanced strategies have succeeded to some extend in improving

some criteria of a green nitration reaction, but have often faced new problematic issues,

including expensive catalysts and/or reagents, lengthy synthesis of substrates, low

reaction rates, and new sources of environmental pollution. These challenges illuminate the fundamental difficulties in developing green chemical nitration processes (Figure 1-

9).

Natural Nitration Strategies

Developing efficient catalysts that can sustainably produce a variety of chemicals represents a major challenge in green chemistry. Enzymes as biocatalysts are of great synthetic significance as their typical high stereo-, regio-, and chemo-selectivity avoids lengthy protection and deprotection steps and the generation of impurities. As described above, biocatalysts enable chemical transformations spanning almost all key aspects in organic chemistry112, 113. Among the highest-ranking reactions in organic chemistry is

nitration, especially aromatic nitration. Surprisingly, this important but environmentally

damaging reaction has benefitted little to none from biocatalysis research as of yet.

Although the majority of nitro aromatics are man-made, nature has evolved at least three biological strategies to produce hundreds of nitro compounds with a great degree of structural diversity114, 115 (Figure 1-10).The first strategy has been observed

only in the biosynthesis of dioxapyrrolomycin in Streptomyces fumanus116 and

resembles chemical electrophilic nitration, thus having limited applications in green

chemistry. Enzymatic N-oxidation, the second strategy and a dominant nitration-

formation mechanism, converts a preinstalled substrate amine group into the nitro

group114. Several types of N-oxidase with diverse cofactors have been identified to

enable this conversion on a sugar or an aromatic ring system31, 117-119. The third

mechanism was uncovered by Dr. Loria’s group at the University of Florida during an

investigation of thaxtomin biosynthesis120, 121. Thaxtomins, thaxtomin A in particular, are

key virulence molecules produced by several pathogenic Streptomyces species that aid

in the infection of plants121-123. Importantly, thaxtomin A is being developed as a

bioherbicide to control weed growth. TxtE, a unique P450 enzyme involved in the

thaxtomin A biosynthesis, promotes an unprecedented, direct nitration reaction on the

120 C4 of the L-Trp indole ring with oxygen (O2) and nitric oxide (NO) as co-substrates .

Substrate NO is biologically produced from amino acid L-Arg by NO synthase (Figure 1-

10). Recently, another P450 enzyme, RufO, was identified from the Streptomyces

atratus rufomycin biosynthesis pathway124. This enzyme catalyzes direct nitration

reaction on the C3 of tyrosine and resembles the TxtE catalyzed nitration reaction. It is also worth noting that SyrB2, a non-heme iron-dependent halogenase, was able to

catalyze aliphatic nitration and azidation with anions NO2 and N3 as donors,

respectively. Engineering of this enzyme reduced native halogen anion binding and

improved nitration and azidation efficiency by 2.5-fold and 13-fold125.

Despite the discoveries of these natural nitration strategies and remarkable

nitrating enzymes, they have only rarely been applied to nitro aromatic production126.

Nonetheless, these enzymes offer compelling opportunities for engineering nitration

biocatalysts. Compared with N-oxidase, TxtE is particularly attractive due to its direct

nitration (avoiding pre-installation of amine), considerable substrate flexibility, controllable regio-selectivity, and distinguishing chemistry among all P450-promoted

reactions47, 127. Researchers of the Arnold group have discovered a single mutation in

the F/G loop of TxtE that simultaneously controls loop dynamics and completely shifts

the enzyme’s regio-selectivity from the C4 to the C5 position of L-Trp. This mutation is

also found naturally present in a subset of TxtE homologues which exclusively produce

127 5-NO2-L-Trp .

Putative Catalytic Cycle of TxtE in Aromatic Nitration

The aromatic nitration reaction catalyzed by TxtE can be distinguished from all other known P450 reactions51, 120. This reaction presumably involves diversion of the

ferric superoxide (4) to a Fe3+-peroxynitrite (PN, ONOO-) intermediate after reacting with

3+ NO. The Fe -PN species may then be decomposed (i) homolytically to yield NO2• and

+ compound II (8), or (ii) heterolytically to produce the resting ferric state (1) and NO2 .

+ Depending on the identity of the nitrating agent (NO2• or NO2 ), TxtE may then undergo

either radical or electrophilic substitution (Figure 1-11).

Research Aims

Nitro chemicals are among the largest and most important families of industrial

chemicals with annual global production of greater than 108 tons85. These chemicals

have broad applications in industry, agriculture, and health. The easy transformation of

the nitro group to other functionalities such as amine group also merits the wide

application of nitro chemicals as building blocks in the organic synthesis93. However,

current industrial nitration processes are notorious sources of pollution and pose a number of selectivity and safety issues that conflict with industrial and societal demands for high quality products and earth-friendly chemical processes. Biocatalysts could serve as an alternative to the traditional chemical catalysts because of their advantages in environmental benignness, selectivity, and activities. The objective of this research is

to develop biocatalytic processes for the synthesis of nitro compounds using

cytochrome P450 TxtE as the parent enzyme. First, as mentioned in Chapter 1, TxtE belongs to Class I P450 which requires two additional proteins (ferredoxin reductase

and ferredoxin) to transfer the electrons from NADPH to the heme prosthetic group. In

order to ease the usage of TxtE in both research and industrial applications, this study

aims to create artificial self-sufficient TxtE biocatalysts by fusing TxtE with the reductase

domains of P450BM3 and P450RhF. The chimera should also be optimized to further

improve the catalytic performance. The second goal of this study is to biochemically

characterize the self-sufficient nitration biocatalyst in terms of its stability, activity,

selectivity, substrate promiscuity, and the key residues of the enzyme binding pocket.

The third aim is to engineer the nitration biocatalyst to create more versatile variants

and improve its activity on unnatural substrates. Finally, to reduce the labor and

economic cost and facilitate the application of the nitration biocatalyst, this study also

sought to develop a whole cell based biotransformation system by employing nitric

oxide synthase and glucose dehydrogenase as the co-substrate producers.

Figure 1-1. Selected examples of biocatalysts in industrial applications.

Figure 1-2. Overview of approaches for protein engineering.

Figure 1-3. Biosynthetic strategies of artemisinin and paclitaxel using whole-cell biocatalysts. Figure modified from Mingzi Zhang et al.8

Figure 1-4. The P450 structural fold44. The figure shows overall P450 structural topology, as exemplified by the Mycobacterium tuberculosis cholesterol oxidase CYP125 (PDB 3IW0). The heme is in pink sticks in the center of the molecule, with iron as an orange sphere. Helices are in yellow and sheets in green. The important structural regions, the I-helix, FG-helices, and B region, are highlighted in cyan, red, and blue, respectively. The heme region and adjacent section of the I-helix form part of the substrate-binding pocket and are generally conserved among P450 isozymes, facilitating dioxygen activation at the central heme iron. The variable FG-helices and B region (which comprises the short B0-helix, the long and diverse loops between the B- and C-helices and strand 5 of β-sheet 1) form the dome of the active site pocket and modulate the binding and selectivity of different substrates. Movement of these variable regions allows the formation of access channel(s) to the heme pocket.

Figure 1-5. The paradigm for P450 catalyzed hydroxylation and peroxide shunt pathways.

Figure 1-6. Selected classes of P450s based on their electron transfer partner proteins. Figure modified from Roberts et al.58

Figure 1-7. Selected examples of engineered P450s catalyzed chemical transformations.

Figure 1-8. Selected nitro chemicals and their applications. (A): selected nitro chemicals and their applications in industry, agriculture et al. (B): selected nitro aromatic prodrugs in GDEPT.

Figure 1-9. Selected chemical aromatic nitration approaches. Figure modified from Yan et al.93

Figure 1-10. Three methods to produce nitro compounds evolved in nature.

Figure 1-11. General paradigm for P450 catalyzed hydroxylation and putative mechanism for TxtE-promoted nitration. Dashed arrows denote putative ● + pathways for ferric-PN decomposition that lead to NO2 or NO2 .

Table 1-1. Comparison of the different protein engineering strategies. Table modified from Mingzi Zhang et al.8 Strategy Advantages Disadvantages

Can be labor-intensive. Throughput is Knowledge of the enzyme Directed evolution highly dependent on the screening or is not required selection method used

Knowledge about the enzyme, particularly the structural information, is Rational design Small set of mutants to test required. A limited sequence space is explored, and beneficial mutations may be consequently missed

Less labor-intensive than directed evolution, more Structural information of enzyme or Semi-rational design diversity than rational several rounds of screening is required design

Can be used to create The designed biocatalyst usually has De novo design novel biocatalyst not found relatively low activity and need further in nature optimization

CHAPTER 2 CREATION AND BIOCHEMICAL ANALYSIS OF AN ARTIFICIAL SELF-SUFFICIENT CYTOCHROME P450 DIRECTLY NITRATES TRYPTOPHAN*

Introduction

Nitro (-NO2) compounds, particularly nitro aromatic and heterocyclic derivatives,

are important industrial chemicals, with an estimated annual production of greater than

108 tons85. Their applications span a broad range such as food additives, pesticides,

herbicides, polymers, explosives, and dyes86. Despite its association with several

toxicity issues, the nitro group is also an important functional unit in pharmaceuticals

such as chloramphenicol, nilutamine, tolcapone, metronidazole, and anti-tuberculosis drug delamanid128, 129. In 2017, two nitro-containing drugs, benznidazole and

secnidazole were approved by FDA to treat children ages 2 to 12 years old with Chagas

disease and bacterial vaginosis respectively. In addition, the therapeutic relevance of

nitro group is further illustrated by lead drug candidates such as 9-NO2-noscapine for

130 the treatment of multidrug resistant cancers and 5-NO2-2-furancarboxylamides in treating neglected parasitic protozoa infections131.

Aromatic nitration is one of most widely studied and used organic reactions93.

Industrial scale reactions usually include a mixture of nitric acid and sulfuric acid or

+ sometimes nitric acid with other acids. In these reactions, the nitronium ion, NO2 , is

believed to be the active species, albeit the potential minor contribution of a radical

mechanism93. The practicability and economics of this nitration method have made it predominant in producing nitro aromatics. However, several common issues such as poor selectivity, low yield, generation of multiple isomers and by-products, and low

*Reprinted with permission from Zuo, R. et al. An artificial self‐sufficient cytochrome P450 directly nitrates fluorinated tryptophan analogs with a different regio‐selectivity. Biotechnology journal 11, 624-632 (2016). Copyright © 2016 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

functional group tolerance frequently occur and limit their uses in generating products with specific requirements. Two advanced strategies have recently been developed to address these issues. The first strategy relies on prefunctionalization of substrate precursors and includes approaches of ipso-nitration by the nitrodemetalation of an aryl

C-M bond98, nitrodecarboxylation of an aryl carboxylic acid99, and transition-metal (Pd or

Cu) mediated cross coupling of aryl halides, pseudohalides, or aryl boronic acids100, and

the ipso-oxidation of an amino or azide group101. On the other hand, the second focuses

on the use of transition metals (Pd, Rh, Cu, Ag) for direct regioselective aromatic C-H

nitration with the assist of directing groups in the substrates102. Although these

strategies have succeeded in solving the above issues to some extent, they

experienced their own disadvantages such as high cost, lengthy synthesis of

substrates, and potential environmental pollution. Therefore, there is an increasing need

to develop environmentally benign, selective, practical and efficient direct aromatic

nitration approaches.

Compared with chemical catalysis, biocatalysis based on enzymes or cells

harboring desirable enzymes offer a number of distinct advantages such as high

efficiency, high degree of selectivity, mild reaction conditions, and environmental

friendliness. These features are driving the rapid development of biocatalysts to fulfill

consumer demands for new products, industrial needs for cost reduction, and societal

and government pressures for “greener” technologies. For instance, a transaminase has

been developed and applied in the production of the type II diabetes drug Januvia® on

an industrial scale8. Although the majority of nitro aromatics are man-made, nature has

evolved at least three different biological strategies to produce hundreds of nitro-

containing compounds with a great degree of structural diversity132, 133. The first strategy

was observed only in the biosynthesis of dioxapyrrolomycin in Streptomyces

fumanus116, and resembles chemical electrophilic nitration. Enzymatic N-oxygenation, the second mechanism, converts a preinstalled substrate amine group into the nitro group by several types of metalloenzymes114. This strategy has predominantly been

used to biosynthesize identified nitro-containing natural products such as pyrrolnitrin117

and aureothin118. The third strategy has been employed by a unique cytochrome P450

(CYP), TxtE, in the biosynthesis of phytotoxin thaxtomin121. This enzyme promotes an

unprecedented, direct nitration reaction on the L-Trp indole ring with O2 and NO as co-

substrates (Figure 2-1)120. Despite the discoveries of these natural nitration strategies

and remarkable nitrating enzymes, their applications in producing nitro aromatics,

especially nitro-indole compounds, have barely been reported126, 134.

CYPs form a super-family of heme-thiolate containing enzymes. These versatile enzymes regio/stereo-selectively catalyze a variety of chemical reactions that are generaly driven by the consumption of a reducing agent NAD(P)H. Effectivly transferring electrons from the reducing agent to heme center requires a proper interaction between the CYP and suitable auxiliary redox proteins. Based on the types of redox proteins,

CYPs have been classified into several classes135. For instance, TxtE is a class I CYP

whose catalytic activity depends on a small redox 2Fe-2S iron-sulfur ferredoxin (Fer)

and a flavin adenine dinucleotide containing reductase (Frd)120. The class III consists of

self-sufficient CYPs, in which the heme domains are fused with reductase domains as

single polypeptides136. Compared to other classes of CYPs, the fusion nature enhances

catalytic activities of the class III enzymes and significantly eases their preparation, two critical features attractive in biotechnological applications135, 137-140.

Inspired by natural self-sufficient CYPs, multiple researchs have been done to

generate new artificial fusions60, 61, 141-147. The first strategy is to fuse the P450s to their natural redox partners. Sibbesen et al.147 fused P450cam with its natural redox partners

Pdx and Pdr with different order and P450cam-Pdr linker lengths. The best variants fused in the order Pdr-Pdx-P450cam showed comparable activity with other

reconstituted non-fused P450 systems.

However, in most cases, the natural redox partners are unknown for one certain

P450. So non-natural redox partners are more often employed. The CPR-like reductase

of P450BM3 and the PFOR-like reductase of P450RhF are the two most commonly

used redox partners in this research field61. For exampole, Gilardi et al.148 employed a

Molecular Lego method to generate mammalian P450BM3 reductase fusion enzymes.

The resulting chimeras showed comparable catalytic activity with the membrane bound

parental CYP, and the solubility was also greatly improved.

Researchers in the Misawa group144 are the first to use P450RhF reductase to

construct artificial fusion enzymes. They successfully fused the reductase from

P450RhF to P450cam, P450bzo, P450balk and CYP153A-family P450s to generate chimeras that active on the natural substrates. The P450balk-RhFRed chimera was further characterized. The fusion enzyme showed better catalytic activity, selectivity and

stability compared to many other alkane hydroxylases149. PikC (CYP107L) is a type III

P450 found in the pikromycin biosynthesis pathway of Streptomyces venezuelae.

Researchers in the Sherman group fused PikC to RhFRed and the catalytic activity of

the chimera was four-fold higher than the reconstructed system using spinach ferredoxin and ferredoxin reductase141.

To develop novel biological nitration approaches for producing nitro aromatics in

a highly selective and environmentally benign manner, we herein created self-sufficient

TxtEs as the nitration biocatalysts. We first fused the TxtE with the reductase domains of CYP102A1 (P450BM3) from Bacillus megatorium137 and of P450RhF from

Rhodococcus species150. Spectral analysis and catalytical activity evaluation identified

TxtE-BM3R as the best self-sufficient candidate. We then biochemically characterized

TxtE and the fusion enzyme including thermostability, pH stability and dependence.

This chapter further describes the characterization of 15 new chimeric TxtE-

BM3R biocatalysts. These chimeras were developed by varying the length of a linker

connecting TxtE and BM3R and swapping a putative interfacial loop on the TxtE to

improve interactions with the reductase domain. These studies have yielded TxtE-

BM3R constructs with improved catalytic turnover, coupling efficiency, and broad

substrate specificity.

Results

Preparation of Self-sufficient TxtE Variants

TxtE promotes a regio-selective nitration on the C4 of L-Trp indole ring using O2

and NO as co-substrates and consuming NADPH (Figure 2-1)120. Since the native redox

partners of TxtE remain unidentified, spinach Fer and Frd were used to support the

reaction. To use TxtE as a broadly applicable biocatalyst for aromatic nitration, we

designed three artificial self-sufficient TxtE fusion enzymes, TxtE-BM3R, TxtE-RhFRed, and TxtE-RhFRed*, by appending NADPH-dependent reductase domains of P450BM3 and of P450RhF to C-terminus of TxtE. The linker of TxtE-BM3R was predicted from

P450BM3 using software Domcut151. Due to proven effects of linker lengths on catalytic activities of RhFRed fusion enzymes141, 142, we created two fusion enzymes. TxtE-

RhFRed contained the native linker length, while TxtE-RhFRed* has eight additional

residues: this design offered the highest activities in previous studies142. All fusion

enzymes were expressed in E. coli and purified to homogeneity with over 85% purity by

a single nickel affinity chromatography (Figure 2-2A). All recombinant proteins showed

calculated molecular weights, 112kD for TxtE-BM3R and approximately 82kD for both

TxtE-RhFRed and TxtE-RhFRed*, in SDS-PAGE analysis (Figure 2-2A). To assess the

functional folding of recombinant fusion proteins, we used UV/Vis spectroscopy to

record their absorption spectra (Figure 2-2B). CO-bound oxidized form (green lines) and

reduced form (blue lines) of these enzymes resembled similar features to wild type TxtE

and other bacterial CYPs. Soret peaks were shifted from approximately 419 nm in the

oxidized forms to approximately 449 nm in the reduced-CO difference forms (red lines),

indicating the proper folding of all fusion enzymes. The concentrations of functional

heme-enzymes were accurately determined by this spectral approach, following the

previously published protocols152, 153.

Catalytic Activity of Fusion Enzymes

Following the successful enzyme expression and purification, we assessed

catalytic activities of all three fusion enzymes along with NADPH, the NO donor NOC-5

and L-Trp. As a control, wild type TxtE was reconstructed with spinach Fer and Frd.

HPLC analysis of reaction mixtures revealed that all fusion enzymes enabled the L-Trp nitration reaction to a different extent (Figure 2-2C). TxtE-BM3R exhibited slightly higher conversion (109%) than the control, while both TxtE-RhFRed and TxtE-RhFRed* only reached 13% and 16% of the conversion level of the control, respectively. To examine

the extent to which the fusion arrangement influenced the substrate-enzyme interaction, which might induce the observed variation of enzyme activity, we measured the binding affinity of L-Trp toward all fusion enzymes. TxtE-RhFRed showed the highest binding affinity with the Kd value of 18.21 ± 1.38 μM, followed by TxtE-BM3R (Kd = 20.83 ± 0.35

μM) and TxtE-RhFRed* (Kd = 24.34 ± 1.21 μM). These values remained in the same

range as wild type TxtE (Kd = 24.77 ± 1.07 μM). Therefore, activity differences of fusion enzymes may be originated from electron transfer efficiency between TxtE and reductase domains. Interestingly, no nitrated product was detected by LC-MS analysis when TxtE was incubated with a standalone BM3R (data not shown), indicating the necessity of covalently linking two domains to promote the catalytically active electron transfer.

Biochemical Characterization of TxtE and TxtE-BM3R

Despite the promising biotechnological potential of TxtE, optimal reaction conditions of TxtE remained uncharacterized. Here we first examined thermostability of both TxtE and TxtE-BM3R (Figure 2-3A). These enzymes were incubated under different temperatures (4 to 65 °C) for 15 min and then used in L-Trp nitration reaction at 20 °C. Both enzymes showed a similar level of thermostability with the T50 of

approximately 45 °C (Figure 2-3A). After incubation at 65 °C for 15 min, their activity

was completely lost, indicating irreversible conformation changes at high temperature.

Next, we examined the pH dependence of TxtE and TxtE-BM3R using NOC-5 as the

NO donor (Figure 2-3B). This reagent is stable over a broad pH range (data not shown).

Both enzymes remained <5% activity at buffers with pH below 7.0. TxtE-BM3R showed over 50% activity from pH 7.5 to pH 9.5 and its optimal pH range was 8.0 to 8.5. TxtE’s activity depended on a narrower pH range, and its optimal activity preferred to pH 8.0.

We further examined the extent to which the stability of both enzymes was affected by buffers with different pH values. After incubated in these buffers for 15 min, enzymes were then used to nitrate L-Trp at pH 8.0, 20 °C (Figure 2-4). HPLC analysis revealed that enzyme activities were only minimally affected by the incubation in different buffers.

This result suggested that the pH dependence of enzyme activity (Figure 2-3B) was not

associated with the enzyme pH stability.

Design and Production of Chimeric TxtE-BM3R Variants

In previous studies, we created three self-sufficient TxtE constructs154. Of the two

reductase domains that we evaluated, the di-flavin reductase BM3R homologous to eukaryotic cytochrome P450 reductase155 conferred superior TxtE nitration activity

compared with the P450RhF reductase domain (RhFRED), a natural fusion of Frd and

Fer150. In the latter case, the catalytic performance was notably dependent on the length

of the linker connecting TxtE and RhFRED154. Indeed, profound impact of linker length on the performance of artificial fusions has appeared in several recent studies142, 156-158.

Inspired by these results, we sought to create serial chimeric TxtE-BM3R variants by

varying linker length in this work (Figure 2-5). Using a stepwise cloning approach,

chimeric TxtE-BM3R variants with linker lengths of 3, 6, 9, 11, 14, 17, 19, 22, 24, and 27

amino acids (AA) were first constructed (Figure 2-6) to quickly assess potential

influences of linker length on enzyme nitration performance. To investigate the optimal

length, a second set of variants with linker lengths of 12, 13, 15, and 16 was latter

prepared by the same cloning method. Hereafter, these TxtE-BM3R variants are

designated as TB-linker length, so for example the TxtE-BM3R variant connected by a

3-AA linker is denoted TB3. In all variants, two amino acids glutamate (E) and leucine

(L) were appended to N-termini of the linkers as a result of the SacI digestion site

facilitating molecular cloning. For instance, the linker of TB27 comprised EL and the entire 25-AA linker of P450BM3 (Figure 2-6). The previous TxtE-BM3R construct contains a 13-AA linker that bears EQ at its N-terminus154 (Figure 2-6), and is denoted

TB13-Q in this study.

Superimposition of the crystal structure of TxtE (PDB: 4TPO) with the P450

heme domain of P450BM3 (PDB: 1BVY)155 revealed an additional opportunity to

potentially improve chimera’s catalytic activity. A loop connecting two helices (J and K,

Figure 2-5, Figure 2-7) of P450BM3 putatively contributes to the interface of heme and

reductase domains. This basic-residue-rich loop (A291 to Y313) is approximately 7 Å

away from the acidic residues of the FMN-binding domain and can provide a specific

protein-protein interaction for efficient electron transfer (Figure 2-7)155. This stretch of

amino acids is significantly different from those in TxtE (Figure 2-7). We thereby

hypothesized that replacing those residues in TxtE with the corresponding residues in

P450BM3 could increase the coupling efficiency between the domains (Figure 2-5). The

rationally swapped chimera (TB13S) carrying the JK loop from P450BM3 was

constructed with overlapping PCR using TB13-Q and P450BM3 as the templates. All

constructs were expressed in E. coli and purified to homogeneity by a single Ni2+-NTA

affinity chromatography154 (Figure 2-8). All chimeras were similarly soluble, indicating

the minimal effect of variable linker length on the protein solubility. Except TB3, TB6 and

TB9, all chimeras showed a peak at approximately 450 nm in their reduced-carbon

monoxide (CO) difference spectra152 (Figure 2-9), indicating the production of active

P450s. The concentrations of active chimeras and TxtE in purified proteins were

-1 -1 quantitated using an extinction coefficient of Δε450-490nm of 91 mM cm for the ferrous

P450-CO complex.

Nitration Performance of Chimeric TxtE-BM3R Variants

We first measured the catalytic performance of TB11, TB13-Q, TB13S, TB14,

TB17, TB19, TB22, TB24, and TB27 to quickly assess the potential effects of linker length (Figure 2-10). TxtE coupled with spinach Fer and Frd served as the control.

HPLC analysis (Figure 2-11) revealed that all fusion enzymes nitrated Trp to a different extent (Figure 2-10). After 30 min, TxtE (1.5 µM) converted 6.1% of Trp (0.5 mM) into 4-

NO2-L-Trp, and this rate was lower than all tested chimeric constructs (1.5 µM).

Compared with TxtE, the nitration activity of the most active chimera, TB14, was

improved by 2.4-fold to reach 14.5% conversion in 30 min. The same was true when the

reaction time was extended to 60 min (12.5% for TxtE vs. 28.8% for TB14). The second

most active enzyme was TB11 (10.2%), followed by TB13-Q (8.9%). The nitration

activity of chimeras with longer linker length from 17 to 27 AAs were similar (6.2 to

7.5%) and approximately 2-fold lower than TB14 (Figure 2-10). Next, we characterized

TB12, TB13, TB15, and TB16 to finely examine the optimal length from 11 to 17 AAs

(Figure 2-6B). The activity of both TB15 and TB16 was higher than TB11 and TB17 and was only slightly lower than TB14 (Figure 2-10). On the other hand, TB12 and TB13 retained approximately 60% of TB14’s activity. These results demonstrated the optimal linker length to be 14 to 16 AAs. Moreover, the similar activity of TB13 and TB13-Q

(Figure 2-6) suggested that linker content played a minor role in determining activity

(Figure 2-10), consistent with the conclusions of several previous reports156-158.

Additionally, standalone BM3R (Figure 2-10) in solution was unable to support TxtE for

nitration, indicating the necessity of the linker in modulating proper interactions between

TxtE and BM3R.

The nitration activity of rationally designed TB13S (4.8%) was lower than TxtE and TB13-Q (Figure 2-10). The observed catalytic activity, however, for the first time demonstrated that TxtE is a robust scaffold for chimerogenesis engineering that tolerates foreign structural elements and has potential to greatly expand fitness and improve catalytic properties of nitration biocatalysts159, 160.

To yield mechanistic insights to account for the observed activity changes, we

first assessed the binding affinities of L-Trp toward all active chimeras by UV-Vis

difference spectroscopy (Table 2-1). All fusions showed Type I spectral shift of the Soret

peak from 420 nm to approximately 390 nm after adding L-Trp and demonstrated the

overly similar level of binding affinity (Kd = ~20 μM). This result indicated that linker

length or structural swapping minimally influenced substrate binding. We next

determined coupling efficiency to evaluate the extent to which linker length and

structural swapping may affect electron transfer compatibility during nitration reaction

(Table 2-1). Coupling efficiency of TB11 (5.2%) and TB14 (5.3%) were approximately

2.2-fold higher than TxtE (2.4%), possibly indicating that improvement in their activity was driven by more effective electron transfer. TB15 (3.9%) and TB16 (2.7%) also showed improved coupling efficiency in comparison with TxtE. Coupling efficiency of the other fusions was similar to that of TxtE. Of note, the positive influence of the swapped

JK loop from P450BM3 was reflected by 1.3-time improvement in electron transfer efficiency of TB13S (2.6% vs 2.0% of TB13-Q). However, the coupling efficiency of even the most active chimera is comparably low56, 156, indicating the great potential of

further advancing nitration biocatalysts. Additionally, we determined total turnover number (TTN) as nmol product per nmol P450 in all reactions (Table 3-1). TB14 had the highest TTN at 707 that was 1.9-fold and 2.3-fold higher than the values of TxtE and

TB13-Q, respectively. TTNs of TB11, TB15, and TB16 were also substantially higher than TxtE, presumably reflecting their higher nitration activity (Figure 2-10). We also

observed a high TTN of TB27. On the other hand, TTN of TB13S decreased from 308 of

TB13-Q to 202, indicating the less efficient use of electrons in its nitration.

Conclusion and Discussion

Nitro aromatics have a broad range of applications. Classical electrophilic acid-

based nitration has long been used to produce a number of nitro-containing aromatics.

As well, many advanced methods relying on either prefunctionalization of substrate

precursors or transition metal-mediated nitration have been developed in recent years93.

However, green routes centered on the use of nitrating biocatalysts have not been

reported. In this chapter, we engineered and characterized self-sufficient biocatalysts from a unique nitration-enabling P450 TxtE. These studies laid a basis for implementing this biocatalyst in the nitration of substrates carrying an indole moiety and opened up the possibilities in developing other direct nitration biocatalysts for the production of additional types of nitro aromatics.

In comparison with class I CYPs that includes the majority of bacterial P450s, the class III self-sufficient enzymes generally execute higher catalytic efficiency161, 162.

Attractive features of fusion CYPs such as higher activity and easier preparation have

prompted attempts to create artificial self-sufficient systems by fusing the CYP heme domain with the redox domains of natural systems. For instance, human CYP3A4-

BM3R was created and showed a similar level of catalytic activity to the native enzyme

in a reconstructed system143. In addition, PikC and many other bacterial P450s have

been fused to the RhFRED to increase catalytic efficiency141, 142, 144-146. In this study, we

selected both BM3R and RhFRED to support the TxtE nitration reaction as the fusion

enzymes. BM3R is a di-flavin reductase similar to the eukaryotic membrane bound

cytochrome P450 reductase155, and it has been widely used to create self-sufficient

enzymes. RhFRED is a natural fusion of ferredoxin reductase and ferredoxin (N to C

terminus)150. Interestingly, it contains a 2Fe-2S cluster and a flavin domain using FMN instead FAD typical in the bacterial systems, resembling the domain identities of phthalate dioxygenase reductases55. Although the usefulness of RhFRED as a

surrogate reductase in the fusion format has well been demonstrated in characterizing

greater than 30 bacterial P450s141, 142, 144-146, TxtE-RhFRED showed only less than 15%

conversion rate of the TxtE in the reconstructed system (Figrue 2-2). Optimization of the

linker length in TxtE-RhFRED* yielded only a limited level of activity improvement. In

contract, the catalytic activity of TxtE-BM3R was slightly higher than the TxtE

supplemented with spinach Fer and Frd. Since the fusion arrangement induced a

minimal change on binding interactions between TxtE domain and L-Trp, proper heme

domain-reductase domain interactions may determine the overall activity of fusion

enzymes via affecting electron transfer.

To further improve the catalytic activity of TxtE-BM3R, we created 15 new

chimeric TxtE fusion constructs by employing two protein engineering strategies.

Rationally swapping the JK loop of TxtE with that of P450BM3 noticeably improved the

coupling efficiency of resulted variant (TB13S) but not its overall nitration performance.

Varying the linker length between TxtE and BM3R led to the creation and identification

of multiple chimeras, particularly TB14, which showed significantly higher catalytic activity than TxtE and TB13-Q23. Remarkably, TB14 is the most active aromatic nitration biocatalyst ever developed.

Previously, a few enzymes have been explored in nitration reactions but only minor progress was made. Although peroxidases and superoxide dismutases catalyzed aromatic nitration reactions, they lacked required regio-selectivity and were unstable to

the NO donor peroxynitrite163-165. The same instability issue virtually eliminated the use

of both P450BM3 and P450cam originally from Pseudomonas putida for aromatic

nitration57, 166. TxtE is a special enzyme dedicated to the direct nitration reaction. Its

development as shown in this report offers an efficient, green approach to elaborate

thousands of indole alkaloids and peptidic compounds containing L-Trp moiety.

Methods and Materials

General Chemicals, DNA sub-cloning, and Bacterial Strains

Molecular biology reagents and enzymes were purchased from Fisher Scientific.

Primers were ordered from Sigma-Aldrich. 4-Me-DL-Trp was from MP Biomedical

(Santa Ana, CA), while NOC-5 (3-(Aminopropyl)-1-hydroxy-3-isopropyl-2-oxo-1-triazene) was purchased from EMD Millipore. Other chemicals and solvents were purchased from

Sigma-Aldrich and Fisher Scientific. Escherichia coli DH5α (Life Technologies) was used for cloning and plasmid harvesting. E. coli BL21-GOLD (DE3) (Agilent) was used for protein overexpression. E. coli strains were grown in Luria-Bertani broth or Terrific broth. DNA sequencing was performed at Eurofins. A Shimadzu Prominence UHPLC system (Kyoto, Japan) fitted with an Agilent Poroshell 120 EC-C18 column (2.7 µm, 3.0 x 50 mm), coupled with a PDA detector was used for HPLC analysis. Student's t-test

was used to determine if the nitration activity of TxtE and its chimeras are significantly different from each other.

Construction of Self-sufficient TxtE Variants

TxtE gene (Genbank: FN554889 REGION: 3613916-3615136) was amplified

from genomic DNA of S. scabies 87.22 (NRRL B-24449) using a pair of TxtEFN and

TxtERH primers (Table 2-2) in PCR reaction. The PCR mixture (50 μL) contained 50 ng template, 2 μM of each primer, 0.1 mM of dNTP, 3% dimethyl sulfoxide, and 0.5 μL

Phusion high fidelity DNA polymerase in 1XGC reaction buffer. Reaction conditions consisted of an initial denaturation step at 98 °C for 30 s followed by 30 cycles of 98 °C for 10 s, 70 °C for 20 s, and 72 °C for 30 s, and a final extension of 72 °C for 5 min. The

PCR product was analyzed by agarose gel and extracted with a GeneJET Gel

Extraction Kit (Thermo) following a manufacture’s protocol. To create the TxtE-

P450BM3 reductase (BM3R) domain fusion gene, TxtE gene was amplified using a pair of TxtEFN and TxtEBRR primers while TxtEBRF and BRRS primers were used to amplify BM3R gene (GenBank: J04832.1) from the genome of B. megaterium ATCC

14581, which was then followed by an overlapping PCR167. Similarly, TxtE-RhFRed and

TxtE-RhFRed* fusion genes were generated by fusing TxtE gene with P450RhF

reductase domain (RhFRed) gene (GenBank: AF459424.1) amplified from the template

of pET21b-RhFRED kindly provided by Professor David H. Sherman (University of

Michigan). Corresponding primers were included in Table 2-2. Purified PCR products

and pET28b were digested with the same sets of restriction enzymes and

corresponding linear DNAs were ligated to generate expression constructs. All inserts in

the constructs were sequenced to exclude mutations introduced during PCR

amplification and gene manipulation.

Construction of TxtE-BM3R Variants

TxtE gene was amplified from genomic DNA of S. scabies 87.22 (NRRL B-

24449) using a pair of SELKnco-F and SELKsac-R primers in PCR reactions (Table 2-

3). The PCR product was analyzed by agarose gel and extracted with a GeneJET Gel

Extraction Kit (Thermo). Purified PCR products and pET28b were digested with the restriction enzymes NcoI and SacI, and corresponding linear DNAs were ligated to generate expression construct pET28b-TxtEs. To further create the TxtE-BM3R variants with variable linker length, BM3R domain with selected linker lengths was amplified from

P450BM3 gene by a set of primer pairs (Table 2-3). Purified PCR products and

pET28b-TxtE construct were then digested with the restriction enzymes SacI and XhoI,

and corresponding linear DNAs were ligated to generate pET28b-TxtE-BM3R

expression constructs (Figure 2-6). To create the P450BM3 standalone reductase

(BM3R) expression constructs, BM3R gene was amplified using a pair of BM3R-F and

BM3R-R primers. Purified PCR products and pET28b were digested with the restriction enzymes NdeI and XhoI, and corresponding linear DNAs were ligated to create pET28b-BM3R. To create TB13S fusion variant, we used TB13-Q as the template and amplified the TxtES fragment with primers of TxtESF and TxtESR (Table 2-3), and the

S13BM3R fragment with primers of FW-V1 and RV (Table 2-3) in PCR reactions.

P450BM3 gene was used as the template to amplify JKL insert using primers of V1-3-F

and V1-3-R. After purification, these fragments were fused by overlapping PCR

technology. The full-length TB13S was cloned into the expression vector as described

above. All inserts in the constructs were sequenced to exclude mutations introduced

during PCR amplification and gene manipulation.

Heterologous Expression and Purification of Recombinant Proteins

Protein expression and purification followed the previous protocols154. The

purified proteins were exchanged into storage buffer (25 mM Tris-HCl, pH 8.0, 100 mM

NaCl, 3 mM βME, and 10% glycerol) by PD-10 column, aliquoted and stored at -80 °C

until needed. CO difference spectroscopy was used to measure the concentrations of

functional P450s152.

Analytical HPLC Analysis

For analytical analysis, the HPLC column kept at 40 °C, water with 0.1% formic

acid was used as solvent A and acetonitrile with 0.1% formic acid was used as solvent

B. For analytical analysis of Trp analogues and pyridine enzymatic reactions, the

column was eluted first with 1% solvent B for 1 min and then with a linear gradient of 1-

20% solvent B in 8 min, followed by another linear gradient of 20-99% solvent B in 2

min. The column was further cleaned with 99% solvent B for 2 min and then re-

equilibrated with 1% solvent B for 2 min. The flow rate was set as 1 mL/min, and the

products were detected at 211 nm with a PDA detector.

Biochemical Characterization of Self-sufficient TxtE Variants

The stability of NO donor NOC-5 was first examined. Its solution was incubated

at different pH value (4.5 to 9.5) and temperatures (4 to 65 °C) for 30 min. It was then

used as NO donor in the P450 nitration reactions. NOC-5 was stable in all tested pH

value but was decomposed quickly and significantly at the temperature higher than

25 °C. To determine pH effects on the activity of TxtE and TxtE-BM3R, enzyme (1.5 µM)

reactions were performed in 100 mM Tris-Cl or sodium phosphate with various pH

values (4.5 to 9.5) at 20 °C, 300 rpm for 30 min. To determine enzyme pH stability, 5 µL

of 30 µM enzyme solutions were incubated at buffers with different pH value (4.5 to 9.5).

After 15 minutes, other reaction components (95 µL) were mixed to initiate nitration reactions as described above. To test enzyme thermostability, we incubated TxtE and

TxtE-BM3R in 100 mM Tris-HCl (pH 8.0) at different temperature (4 °C, 15 °C, 20 °C,

25 °C, 30 °C, 35 °C, 40 °C, 45 °C, 55 °C, and 65 °C) for 15 min. After cooling on ice for

5 min, enzyme solutions were centrifuged and then used to initiate reactions at 20 °C,

300 rpm for 30 min. Products were quantitated by HPLC as described above. All

experiments were performed at least in duplication. In this study, the T50 is defined as

the temperature at which a 15-minute incubation of the enzyme causes the loss of one- half of the enzyme activity, relative to a 100% activity reference enzyme that does not undergo incubation.

Spectral Analysis of Chimeric TxtE-BM3R Variants

Purified TxtE and its chimeric fusions were spectrally analyzed following a previous protocol154. Briefly, the absorbance spectra (400-600 nm) of TxtE and its

chimeric fusions in Tris-HCl (25 mM, pH 8) buffer were recorded by a Shimadzu

UV2700 dual beam UV-Vis spectrophotometer. The ferric heme of enzymes was then

saturated by bubbling carbon monoxide (Airgas) and the spectra of the saturated

enzyme solutions were recorded. Sodium dithionite solution (30 μL, 0.5 M) was then

added to reduce ferric ion, and reduced spectra were taken subsequently. CO reduced

difference spectra of all P450s were created by subtracting the CO binding spectra from

the reduced spectra. Data were further analyzed by Excel. Substrate binding affinities to

P450s were measured using 1.5 μM of enzyme solutions in 25 mM Tris-HCl, pH 8.0.

The changes in absorbance (ΔA) were determined by subtracting the absorbance at

~420 nm from that at ~390 nm. Data were then fitted to the equation of ΔA =

ΔAmax[L]/(Kd + [L]) using GraphPad Prism 4.

Catalytic Activities of Chimeric TxtE-BM3R Variants

P450 reactions (100 μL) contained 0.5 mM substrate, 1 mM NADP+, 1 mM

glucose, ~10 units/mL self-prepared glucose dehydrogenase crude extract, 1 mM NOC-

5 in 100 μL of Tris-HCl buffer (100 mM, pH 8.0). As the positive control, the TxtE reaction was also re-constructed in the above mixture further supplemented with 0.43

µM spinach Fer and 0.33 µM Frd. The reactions were initiated by adding 1.5 µM P450s,

and incubated at 20 °C, 300 rpm on a thermostat (Eppendorf) for 30 or 45 minutes.

Methanol (200 μL) was then added to stop the reactions. After centrifugation, 10 μL

solutions were analyzed by HPLC. The 4-NO2-L-Trp was synthesized in a large-scale

enzymatic reaction to establish a standard curve for product quantification. To

determine the coupling efficiency, NADPH (2 mM) was used to replace the NADPH

regeneration system in the reaction mixture. NADPH consumption in enzyme reactions

was measured at 340 nm (ε = 6.22 mM-1cm-1) with a Biotek Synergy HT Multi-Detection

Microplate Reader. Non-enzymatic oxidation of NADPH was subtracted as the

background. The quantity of nitrated product was determined by HPLC analysis as

described above. Coupling efficiency (%) was determined as product (nmol) / consumed

NADPH (nmol) X 100%. All reactions were independently repeated at least three times.

In TTN and coupling efficiency studies, 0.5 µM P450s and 0.5 mM substrate were used.

Total turnover number (TTN) was reported as nmol product per nmol P450 Conversion

rate (%) was calculated as product (nmol) / (product + substrate) (nmol) X 100%. All

experiments were performed at least in triplication.

1'COOH 1. TxtE + Fer +Frd COOH 2. TxtE fusions 3' NO2 4 2' NO, O2 3a NH2 NH2 5 3 2 + 6 NADPH NADP 7a N N 7 H GDH H

L-Trp 4-Nitro-L-Trp Glucose Gluconate

Figure 2-1. TxtE catalyzes an aromatic nitration reaction on the C4 of L-Trp indole ring. O2 and NO act as co-substrates. NADPH is consumed and recycled with glucose dehydrogenase (GDH) in the reaction. Wild-type TxtE requires spinach Fer and Frd in this reaction while created artificial TxtE variants are self-sufficient.

Figure 2-2. Characterization of recombinant TxtE variants. (A): SDS-PAGE analysis of TxtE and its self-sufficient variants. Recombinant proteins were purified with a single Ni-NTA affinity column and showed expected molecular weights (MW). Lane M: protein marker; lane 1, TxtE [calculated MW (cal. MW): 46.3 kD]; lane 2, TxtE-BM3R (cal. MW: 112.1 kD); lane 3: TxtE-RhFRed (cal. MW: 81.8 kD); lane 4: TxtE-RhFRed* (cal. MW: 82.7 kD). (B): Spectroscopic analysis of TxtE, TxtE-BM3R, TxtE-RhFRed, and TxtE-RhFRed*. (C): Relative catalytic activities of recombinant TxtE self-sufficient variants in nitrating L-Trp. TxtE was used as the control and its activity was set as 100%. TxtE-BM3R activity was slightly higher than the control, while both TxtE-RhFRed and TxtE- RhFRed* only retained less than 15% of TxtE activity.

Figure 2-3. Thermostability (A) and pH dependence (B) of TxtE and TxtE-BM3R. Both enzymes showed the similar T50 at approximately 45 °C and exhibited the best activity at a narrow range of around pH 8.0.

Figure 2-4. The pH stability of TxtE and TxtE-BM3R. Both enzymes were incubated in the buffers with pH from 4.5 to 9.5 for 15 min and then used in the reactions with 0.5 mM L-Trp, 1 mM NADP+, 1 mM glucose, ~10 units/mL self-prepared glucose dehydrogenase crude extract, 1 mM NOC-5 in 100 μL of Tris-HCl buffer (100 mM, pH 8.0). For TxtE reactions, 0.43 µM spinach ferredoxin and 0.33 µM spinach ferredoxin-NADP+ reductase were included. The reactions were incubated at 20 °C, 300 rpm on a thermostat (Eppendorf) for 30 minutes.

Figure 2-5. Schematic depiction of chimeric TxtE-BM3R constructs with variable linker length or a swapped loop (yellow). The structure of human NADPH- cytochrome P450 reductase (PDB: 3QE2, right) represented not-available BM3R structure along with TxtE (PDB: 4TPO, left). The 25-AA linker of P450BM3 is shown as a green dash line (middle) along with the amino acid sequence.

Figure 2-6. Construction of TxtE chimeras with variable linker lengths. (A): Schematic depiction of stepwise construction of TxtE chimeras with variable linker lengths. The expression vector backbone is pET28b. (B): Detailed description of TxtE chimeras with variable linker lengths.

Figure 2-7. Engineering design of TxtE loop connecting J and K helices. (A): Notable length difference at the loop connecting J and K helices was identified after superimposing crystal structures of TxtE (PDB: 4TPO) in purple and P450BM3 heme domain (PDB: 1ZO9) in green. (B): The interface between BM3 heme and FMN-binding domains. The loop connecting J and K helices is labeled in green and its basic residues are shown as sticks. Acidic residues in the loop motif of FMN-binding domain are also shown.

Figure 2-8. SDS-PAGE analysis of purified recombinant proteins. M: protein marker; Lane 1-14: TxtE, BM3R, TB3, TB6, TB9, TB11, TB13-Q, TB14, TB17, TB19, TB22, TB24, TB27, and TB13S. Lane 15-18: TB12, TB13, TB15, and TB16.

Figure 2-9. CO-reduced difference spectra of chimeric TxtE-BM3R variants and TxtE. A peak at approximately 450 nm indicates the properly folded, active P450.

Figure 2-10. Relative nitration activity of TxtE and chimeric TxtE-BM3R variants. All reactions contained 0.5 mM Trp and 1.5 µM P450. The TxtE reaction was further supplemented with 0.43 µM spinach Fer and 0.33 µM Frd. The reactions were incubated at 20 °C, 300 rpm for 30 minutes. All experiments were repeated at least three times. The results of TB12, TB13, TB15, and TB16 were shown as black bars. Chimeras showed significant difference in nitration activity compared to the TxtE were indicated with asterisks, with single or double asterisks indicating a significant difference of P < 0.05 or P < 0.01 respectively.

Figure 2-11. HPLC spectrum of TxtE and chimeric TxtE-BM3R variants relative nitration activity assay. All reactions contained 0.5 mM Trp and 1.5 µM P450. The TxtE reaction was further supplemented with 0.43 µM spinach Fer and 0.33 µM Frd. The reactions were incubated at 20 °C, 300 rpm for 30 minutes. All experiments were repeated at least three times.

Table 2-1. Binding affinity toward L-Trp, coupling efficiency and total turnover numbers of TxtE and its chimeras. All experiments were performed at least three times. Chimeras Kd (µM) Coupling (%) NADPH consumption (µM/min) TTN TxtE 24.8 ± 1.1 2.4 ± 0.3 42.3 ± 4.2 378 ± 17 TB11 16.2 ± 0.6 5.2 ± 0.5 32.8 ± 5.0 535 ± 28 TB12 22.3 ± 1.4 1.8 ± 0.2 86.3 ± 3.9 332 ± 40 TB13-Q 21.5 ± 0.8 2.0 ± 0.1 73.9 ± 3.8 308 ± 18 TB13S 19.3 ± 0.6 2.6 ± 0.2 30.8 ± 3.1 202 ± 20 TB13 21.3 ± 0.4 2.2 ± 0.3 65.7 ± 4.1 342 ± 29 TB14 17.4 ± 1.0 5.3 ± 0.5 45.5 ± 4.8 707 ± 16 TB15 21.4 ± 0.9 3.9 ± 0.4 53.6 ± 5.1 658 ± 33 TB16 18.9 ± 1.0 2.7 ± 0.2 80.5 ± 4.3 351 ± 39 TB17 15.5 ± 0.6 2.1 ± 0.3 50.1 ± 4.6 348 ± 24 TB19 16.0 ± 0.8 2.4 ± 0.5 52.1 ± 6.7 464 ± 21 TB22 27.1 ± 1.3 2.4 ± 0.5 46.3 ± 6.3 408 ± 32 TB24 14.9 ± 0.3 2.3 ± 0.1 46.4 ± 3.4 381 ± 15 TB27 16.2 ± 1.1 2.0 ± 0.6 51.1 ± 7.2 548 ± 17

Table 2-2. Primers for construction of self-sufficient TxtE variants. Name Sequence (5’→3’) Function TxtE-FN CACCCATGGTGACCGTCCCCTCGC TxtE cloning TxtE-RH ATATAAGCTTGCGGAGGCTGAGCGGCAG TxtE cloning TxtEBRF GCCGCTCAGCCTCCGCTCTGCTAAAAAAGTACGC TxtE-BM3R fusion TxtEBRR GCGTACTTTTTTAGCAGAGCGGAGGCTGAGCGGC TxtE-BM3R fusion BRRS ATCGAGCTCGACCCAGCCCACACGTCTTTTGC TxtE-BM3R fusion TxtERedF CCGCTCAGCCTCCGCGTGCTGCACCGCCATC TxtE-RhFRed fusion TxtERedR GATGGCGGTGCAGCACGCGGAGGCTGAGCGG TxtE-RhFRed fusion TxtE8ARedF GCCGCTCAGCCTCCGCCATGTGCGATTGGCGTC TxtE-RhFRed* fusion TxtE8ARedR GACGCCAATCGCACATGGCGGAGGCTGAGCGGC TxtE-RhFRed* fusion RedRH CTCAAGCTTGAGGCGCAGGGCCAGGCG TxtE-RhFRed fusion

Table 2-3. Primers for construction of TxtE-BM3R variants with altered linker regions. Name Sequence (5’→3’) Function SELKnco-F ATACCATGGTGACCGTCCCCTCGCCG TxtE cloning SELKsac-R ATAGAGCTCGCGGAGGCTGAGCGGCAG TxtE cloning BM3R-F CTACATATGTCTGCTAAAAAAGTACGCAA BM3R fusion BM3R-R ATCCTCGAGCCCAGCCCACACGTCTTTTG BM3R fusion BM3LKsac3-F tctGAGCTCAACGCTCATAATACGCCGCTG TxtE-BM3R fusion F primer BM3LKsac6-F tctGAGCTCAAGGCAGAAAACGCTCATAATACG TxtE-BM3R fusion F primer BM3LKsac9-F tctGAGCTCGTACGCAAAAAGGCAGAAAACG TxtE-BM3R fusion F primer BM3LKsac11- tctGAGCTCAAAAAAGTACGCAAAAAGGCAG TxtE-BM3R fusion F primer F BM3LKsac12- tctGAGCTCGCTAAAAAAGTACGCAAAAAGGCA TxtE-BM3R fusion F primer F G BM3LKsac13- tctGAGCTCTCTGCTAAAAAAGTACGCAAAAAG TxtE-BM3R fusion F primer F GCAG BM3LKsac14- tctGAGCTCCAGTCTGCTAAAAAAGTACGCAAAA TxtE-BM3R fusion F primer F AG BM3LKsac15- tctGAGCTCGAACAGTCTGCTAAAAAAGTAC TxtE-BM3R fusion F primer F BM3LKsac16- tctGAGCTCACTGAACAGTCTGCTAAAAAAG TxtE-BM3R fusion F primer F BM3LKsac17- tctGAGCTCAGCACTGAACAGTCTGCTAAAAAA TxtE-BM3R fusion F primer F G BM3LKsac19- tctGAGCTCTCACCTAGCACTGAACAGTCTGC TxtE-BM3R fusion F primer F BM3LKsac22- tctGAGCTCGGTATTCCTTCACCTAGCACTGAAC TxtE-BM3R fusion F primer F BM3LKsac24- tctGAGCTCCTTGGCGGTATTCCTTCACCTAG TxtE-BM3R fusion F primer F BM3LKsac27- tctGAGCTCAAAATTCCGCTTGGCGGTATTC TxtE-BM3R fusion F primer F BM3LKxho-R atcCTCGAGCCCAGCCCACACGTCTTTTGC TxtE-BM3R fusion R primer TxtESF CACCCATGGTGACCGTCCCCTCGCCGCTC TxtES fusion F primer TxtESR CGGGTTGCGGGCGAACGC TxtES fusion R primer V1-3-F GCGTTCGCCCGCAACCCGCATGTATTACAAAA JKLoop insert F primer AGCAGCAGAAGAAGC V1-3-R GGCCGCGACGCGCCAAGGAGCAGTTGGCCAT JKLoop insert R primer AAGCG FW-V1 ACCTGGCGCGTCGCGGC S13BM3R fusion F primer RV GACCCAGCCCACACGTCTTTTGC S13BM3R fusion R primer

CHAPTER 3 DEVELOPING TxtE AS NITRATION BIOCATALYSTS BY ENGINEERING ITS BINDING POCKET*

Introduction

Wild type enzyme must be tailored to realize the enticing potential of biocatalysts.

Nature enzymes seldom fulfill the criteria required by industrial process, and further

engineering and optimization is required to convey properties such as substrate scope

and stereo- and regio- selectivity. Protein engineering has been extensively explored as

a means for biocatalyst discovery and development, mainly via rational design, directed

evolution and de novo computational design8, 168, 169. In the previous studies, we have taken significant steps to create nitration biocatalyst using TxtE as engineering template.

Auxiliary redox proteins are required to deliver electrons for TxtE catalyzed nitration reactions but can be cost and time consuming in preparation. To alleviate this issue, we created self-sufficient TxtE variants by fusing TxtE with BM3R and RhFRED154. BM3R

was proven to better support the TxtE nitration and further optimization of the linker

region connecting TxtE and BM3R yielded TB14 as the best variant regarding the

catalytic activity. Furthermore, the fusion organization had no effect on TxtE’s substrate

scope; TB14 showed similar activity to the same set of nine chemicals as TxtE. These

results suggest that the single TxtE-BM3R fusion is a viable replacement of the TxtE-

Fer-Frd three-enzyme system and thus has been used as template in our ongoing

engineering effort.

In this chapter, the substrate promiscuities of TxtE and TB14 were assessed with a library of 33 chemicals that carry substitutions and moderate alternations on amine,

carboxylate, or indole moieties of the physiological substrate L-Trp. Substrates with

changes on amine or carboxylate moieties showed significantly weakened binding

affinities, while substitutions on the indole ring did not produce as dramatic an effect.

These results suggest that TxtE tolerates changes on the indole ring more than

changes on other moieties. One of our research goals is to modulate TxtE’s substrate

scope to nitrate simple aromatics. These nitro compounds are commonly used solvents

and building blocks of chemical synthesis. To achieve this goal, we are employing semi-

rational design approach to target residues lining TxtE’s substrate binding pocket. The

same approach has demonstrated numerous successes in biocatalyst development.

Results

Substrate Scope of TxtE and TB14

To characterize the substrate scope of TxtE and its most active self-sufficient variant TB14, we first attempted to screen the enzyme nitration activities with a library of

33 chemicals, including 19 tryptophan analogues that carry substitutions and moderate alterations on the amine, carboxylate, or indole moieties (Figure 3-1) (part of the chemical library was also screened by Dodani et al.126).

In spectroscopic analysis, all 19 Trp analogues (Figure 3-1A) induced the Type I

spectral shift of TB14, indicating substrate-type binding. There was no significant

difference between TB14 and TxtE in terms of binding affinities (Table 3-1), further

confirming the minimal effect of fused BM3R on substrate binding154. Compounds with

modified amine or carboxylate moiety generally showed weaker interactions than those

with substitutions on the indole ring, the same observed previously126. We then

examined the nitration activity of TB14 and TxtE toward all library members. To

generate sufficient products from less favourable substrates for HPLC detection, the

reactions were incubated for 45 min. Besides Trp and α-Me-DL-Trp, both enzymes nitrated seven Trp analogues with substitutions on the C4, C5, C6, and C7 of the indole

(Figure 3-2). TB14 showed the higher activity toward all substrates. After 45 minutes,

both enzymes nitrated a significant amount of 4-Me-DL-Trp, 5-Me-DL-Trp, 5-F-L-Trp and 6-F-DL-Trp but not α-Me-DL-Trp, 4-F-DL-Trp, 5-MeO-DL-Trp and 7-Me-DL-Trp. 5-

F-L-Trp emerged as the best substrate of TxtE and was comparable to Trp as the best in the TB14 reaction. Since enzymes nitrate only the L-conformers of racemic substrates, a higher enzyme activity can be expected with enantiopure substrates. In a recent report, the substrate scope of wild type TxtE was qualitatively assessed with a substrate library that contains 19 substances used in the current work126. The quantitative measurement of both TxtE and TB14 toward the select analogues supported improved performance of TB14 across a variety of substrate analogues and underlined its immediate applications to produce unnatural nitro-Trp compounds. These compounds can have broad applications in synthesizing numerous bioactive peptidic compounds and in analysing macromolecule structures and dynamics170.

Next, we probed the substrate promiscuity of TxtE and TB14 with 14 other

compounds, including L-tyrosine, L-phenylalanine, L-histidine, isoquinoline, quinoline, indole, imidazole, benzoxazole, coumarin, styrene, D-camphor, 1,2,3,4-tetrahydro-9H- pyrido[3,4-b]indole, tetrahydroharmine and deformylflustrabromine (Figure 3-1B). Some

of these compounds carry the indole moiety but varied modifications while styrene and

D-camphor represented simple aromatics and bulky nonaromatic cyclic compounds.

Notably, for reactions with styrene as substrate, LC-MS analysis with the positive mode

revealed two possible products both with m/z value of 150 (Figures 3-3). Peak contents

showed identical MS/MS fragmentation patterns, albeit slight changes of signal intensities of two most abundant fragments (Figure 3-3B and 3-3C). Neither of these peaks appeared in the control reactions with boiled inactive TxtE/TB14 or without NOC-

5 or NADPH regeneration system (Figure 3-3A). Collectively, these results demonstrated that, besides tryptophan and tryptophan analogues, TxtE and TB14 also promote reactions on styrene.

To further probe the reactions, we selected a library comprising 24 styrene analogues (Figure 3-4). The first group of six styrene analogues differs with styrene at aliphatic moieties as length, the location of double bond, saturation status, and substitutions. Although most of them could bind to the TB14 binding pocket (Table 3-2),

LC-MS analysis indicated that none was converted by the enzyme: Neither ethylbenzene nor toluene was converted by TB14, indicating the strict requirement of the vinyl group in enzymatic reactions. Furthermore, compounds with modified vinyl groups, including α-methyl, β-methyl and β-nitro substitutions, didn’t act as the substrates of TB14 either. Similarly, TB14 failed to convert allylbenzene as well. These data confirmed the essential of the intact vinyl group in enzymatic reaction of styrene.

On the other hand, TB14 exhibited considerable tolerance to minor to modest changes on the phenyl ring of styrene (Figure 3-5). LC-MS analysis found that 2- vinylnaphthalene, 2-vinylpyridine and 4-vinylpyridine were successfully converted.

Moreover, styrene analogues with methyl, halogen, or nitro substitution at C3 or C4 were accepted as substrates by TB14. Remarkably, all of these active substrates demonstrated higher conversion rates than styrene (Figure 3-5). By contrast, LC-MS analysis did not detect any products from the TB14 reactions containing nonaromatic

vinylcyclohexane or relatively hydrophilic 4-hydroxystyrene. These results indicated that

both aromatic and vinyl moieties are critical for the reaction and the reaction might

require a relatively hydrophobic environment. Indeed, the hydrophilic 5-

hydroxytryptophan was not the substrate of TxtE or TB14 either. The 2,3,4,5,6-

pentafluorostyrene was not converted due to the loss of the binding affinity to the

enzyme (Table 3-2).

Interestingly, although most reactions yielded one major product and one minor

product, the reactions with four compounds, 2-vinylpyridine, 4-acetoxystyrene, 2,4-

dimethylstyrene and 2,4,6-trimethylstyrene, yielded only one product (Figure 3-4, Figure

3-5).

Structural Characterization of Nitrated Tryptophan Analogues

With the availability of highly active TB14 as well as the characterization of

enzyme optimal reaction conditions, we next employed it to produce nitro-Trp

analogues. In this study, we chose commercially available racemic 4-F-DL-Trp and 5-F-

L-Trp as unnatural substrates because fluorine substitution is a common strategy used

by medicinal chemists to generate drug molecules with improved properties171.

Similar to L-Trp, two fluorinated substrates induced type I spectral changes in

both enzyme solutions. Compared with TxtE, the binding affinities between these

substrates and TB14 were approximately 60% tighter, indicating the BM3R might

facilitate substrate binding (Table 3-1). In previous studies, D-Trp was unable to induce

spectral changes in TxtE solution120, 126. This observation may suggest that 4-F-L-Trp of the racemic mixture is the actual ligand bound to TxtE and TB14. With current inaccessibility to optically pure 4-F-L-Trp, we used the total concentration of the racemic mixture to calculate Kd values, which thus underestimated the accurate binding

affinities. Nonetheless, compared with native substrate L-Trp (Kd = 20.83 ± 0.35 μM),

the binding affinities between TB14 and 4-F-DL-Trp (Kd = 189.20 ± 11.14 μM) and 5-F-

L-Trp (Kd = 84.18 ± 4.37 μM) were lowered by approximately 8- and 3-fold, respectively,

reflecting the binding interferences induced by the F-substitution at different positions.

Next, we examined the influences of the fluorination substitution on enzyme activity.

Remarkably, both TxtE and TB14 slightly preferred to 5-F-L-Trp over L-Trp (1.2 : 1) in

the nitration reaction. In addition, although the C4 in 4-F-L-Trp is occupied by an F substitution, both enzymes were able to nitrate this substrate as characterized by HPLC and LC-MS/MS analysis (Figure 3-6). The overall conversion rate was, however, only approximately 20 % of L-Trp. As described above, only 4-F-L-Trp in the racemic mixture may be nitrated120, 126.

Structural characterization of nitrated fluoro tryptophan products was first

performed by LC-MS/MS (Figure 3-6). Nitrated 5-F-L-Trp was fragmented in the same pattern as that of 4-NO2-L-Trp in MS2 spectra. However, the C5-F substitution not only

increased the m/z values of all corresponding ions by 18 Da but obviously affected the

distribution of different ions (Figure 3-6A and 3-6B). The most abundant ion in the MS2

spectrum of 4-NO2-L-Trp had the m/z values of 159.0. It was switched to 174.2 in the

MS2 spectrum of nitrated 5-F-L-Trp, corresponding to the non-fluorinated ion of 156.2.

The most abundant ion in the MS2 spectrum of nitrated product with 4-F-DL-Trp as the

substrate had an m/z value of 209.0 (Figure 3-6C). Importantly, its overall fragmentation pattern was notably different with that of nitrated 5-F-L-Trp. Putative chemical structures of red-labeled ions in these MS2 spectra were shown in Figure 3-7. This result suggested that 5-F-L-Trp may be nitrated at the same site, the C4, as L-Trp but the

nitration site at 4-F-L-Trp as the assumed real substrate in the racemic mixture is different.

To further elucidate the nitro position in nitrated products, we performed the large scale enzymatic reactions. Nearly 90% of 5-F-L-Trp was nitrated and approximately 2 milligrams of the nitro product as a yellow powder were purified by a semi-preparative

HPLC. Similarly, less than 0.2 mg of putative nitro-4-F-L-Trp as a light beige solid was isolated. Both products carried a single nitro group as revealed in HRMS analysis

(Figure 3-8).

Isolated products were further structurally characterized by 1H and 13C and 2D

NMR analysis (Figure 3-9 - Figure 3-12) (Table 3-3). Examining the NMR data made it

clear that the C4 and the C7 of 5-F-L-Trp and 4-F-L-Trp, respectively, were nitrated in

TB14 reactions. From the 1H NMR spectrum of the nitrated 5-F-L-Trp product (Figure 3-

9), the large coupling constant (J = 10.2 Hz) of the triplet-like peak at δ 6.95 ppm (C6)

suggested a single vicinal coupling with the fluorine atom. Furthermore, a neighboring

doublet peak at δ 7.52 ppm (C7) with a coupling constant J = 8.8 Hz defined an ortho

substitution pattern of the two aromatic protons. The aforementioned multiplicity and

coupling constants therefore determined the C4 nitro substitution in the 5-F-L-Trp

substrate, which was further confirmed by HSQC and HMBC analysis (Figure 3-11,

Figure 3-12). Although < 0.2 mg of nitro-4-F-L-Trp were isolated, interpretation of the nitro position in this product was significantly eased using a 1.5 mm High Temperature

Superconductor Probe that is only available in the AMRIS facility at the University of

Florida. From its 1H NMR spectrum, a triplet-like peak at δ 7.80 ppm (C5) displayed a

doublet of doublet split with two approximately equal coupling constants of 8.1 Hz,

suggesting a vicinal coupling with the fluorine atom. An ortho substitution pattern of the two aromatic protons was further defined by a large coupling constant (J = 9.1 Hz) of neighboring doublet peak at δ 7.52 ppm (C6). Together, the nitro site was determined to be the C7 of 4-F-L-Trp, which was further confirmed by HSQC and HMBC analysis

(Figure 3-11, Figure 3-12). These results therefore revealed TB14 as a versatile nitrating biocatalyst with remarkable regio-selectivity and substrate promiscuity.

We have observed remarkable, substrate-tuned regio-specificity of TB14 that selectively nitrates C7 when C4 of its substrate indole is occupied by an F substitution154. To further probe the extent to which physiochemical properties of the

substituted group at C4 impact enzyme regio-selectivity, we employed TB14 in a

scaled-up reaction to nitrate commercially available racemic 4-Me-DL-Trp. Compared

with the fluorine replacement, the methyl group is larger in size and is electron donating

in nature. Interestingly, the binding affinity of 4-Me-DL-Trp was 13-fold higher than 4-F-

DL-Trp (Table 3-1), and TB14 favoured 4-Me-DL-Trp approximately 8-fold more over 4-

F-DL-Trp in nitration (Figure 3-2).

Despite technical challenges associated with production purification, we isolated

approximately 80 µg of nitrated 4-Me-DL-Trp as a yellow powder by semi-preparative

HPLC. By Marfey’s derivatization, we observed the significant consumption (~85%) of 4-

Me-L-Trp in the reaction and identified the nitro product with L-configuration (Figure 3-

13).

This result further confirmed the strict stereo-selectivity of TB14 in nitration. LC- high resolution (HR) MS analysis of the isolated product confirmed the nitration on the substrate by giving one major peak with the expected molecular weight of nitro product

(m/z = 264.0892). Interestingly, one minor peak with the same molecular weight was eluted right after the major one (Figure 3-14), suggesting the coexistence of two structural isomers in the product sample.

Next, we structurally characterized the isolated product by 1H and 13C and 2D

NMR analysis (Figure 3-15 - Figure 3-21). From the 1H and COSY NMR spectra (Figure

3-15 - Figure 3-17), we noticed two well separated AX coupling systems in the aromatic region, presumably indicating two nitration sites in the indole. One AX coupling system involved two large doublet peaks at δ 7.22 ppm and 6.84 ppm while the other was from another set of two small doublet peaks at δ 6.80 ppm and 6.62 ppm (Figure 3-15). The large vicinal coupling constants (8-9 Hz) of the two AX coupling systems suggested the

C5 and C7 of the indole as the nitration sites (Table 3-4, Figure 3-15). Two additional

AMX coupling systems occurring in the aliphatic region further provided details about the existence of two sets of α and β protons of Trp derivatives (Figure 3-16, Figure 3-

17). Given the larger deshielding effect of C5-NO2 than C7-NO2 on the C4-Me group,

the methyl signal of 4-Me-5-NO2-L-Trp was in the lower field (δ 2.22 ppm vs 2.10 ppm in

4-Me-7-NO2-L-Trp, Table 3-4). Using the integration values of two methyl groups, we

determined the molar ratio of 4-Me-5-NO2-L-Trp: 4-Me-7-NO2-L-Trp to be approximately

10:1 (Figure 3-15, Figure 3-16). We were also able to observe the chemical shift signals

13 of 4-Me-5-NO2-L-Trp in the C NMR spectrum (Table 3-4, Figure 3-18). The

determination of the product structure was then assisted by HSQC and HMBC spectra

(Figure 3-19, Figure 3-20). In the HMBC spectrum, the C4-Me group (2.22 ppm)

correlated with a significantly deshielded aromatic carbon (142.18 ppm) that became

possibly only by the C5-NO2 group (Figure 3-20, Figure 3-21). Therefore, TB14 carried

a striking regio-flexibility in nitrating 4-Me-DL-Trp to produce predominantly 4-Me-5-

NO2-L-Trp and 4-Me-7-NO2-L-Trp as a minor product. These results uncovered that

different types of substituted groups at C4 of the indole ring can affect key parameters

(activity and regio-selectivity) of TxtE derived biocatalysts. Interestingly, a single residue

His176 in the F/G loop of TxtE was computationally identified as a potential determinant

of enzyme regio-selectivity, and engineering of this site to some residues (e.g., Phe and

127 Tyr) indeed created TxtE mutants that predominantly produce 5-NO2-L-Trp . Our

results provided new insights into intriguing and synthetically important TxtE’s regio-

selectivity.

Structural Characterization of Nitrated Styrene Analogues

To isolated sufficient amount of the products for NMR analysis, we used TB14 to

convert two substrates with highest conversion rate, 4-vinylpyridine and 4-

methoxystyrene, in 20 mL scale enzymatic reactions. The major products of the two

reactions were subsequently isolated by semi-prep HPLC. Due to the limited amount of

the products, only 1H NMR was performed.

For the product of 4-methoxystyrene, The J value of the first doublet peak at δH

7.6 is 9.1 Hz while the J value of the second doublet peak at δH 6.9 is also 9.1 Hz which

indicated they come from the symmetry benzene ring. We could see the doublet at δH

6.9 is quite smaller than the typical chemical shift of benzene proton which is

approximately 7.3 ppm, so it is highly possible they are next to a methoxy group which

could exert a shielding effect to the benzene ring and cause a decrease of chemical

shift. At the same time, the singlet at δH 3.8 matches the chemical shift of typical

methoxy group attached to benzene ring. The quartet at δH 2.5 and the triplet at δH 1.0

also have similar J value is approximately 7.6 which indicated the presence of the non-

typical ethyl group (Figure 3-22). Similarly, for the product of 4-vinylpyridine, the AM coupling system in aromatic region and A2M3 coupling system in aliphatic region clearly

indicate the existence of four intact pyridine protons and an ethyl group respectively

(Figure 3-23).

Both 1H-NMR spectrums clearly showed that the four H atoms of the pyridine/phenol moiety of the isolated product remain intact, providing the strong supportive evidence that the reaction occurred on the aliphatic side-chain in the TB14

catalyzed styrene conversion reaction. In addition, both 1H-NMR spectrums suggested

the product contained one ethyl group. The HRMS (posotive mode: 180.1017) was also

performed on the 4-methoxystyrene product. A potential chemical formula of C10H14NO2

was deduced based on the HRMS result (HRMS (ESI+): calc. 180.1025, found:

180.1017). A tentative structure was proposed but the structure remained obscure

(Figure 3-22 and Figure 3-23).

Identification of Key Residues for Tailoring TB14 Substrate Scope

TxtE and TB14 have modest tolerance toward substitutions and modifications on

the indole moiety but not on the amine or acid parts of L-Trp. Our goal is to employ a

semi-rational design approach to explore the active site of TxtE for altering its substrate

scope. Our engineering efforts focused on the selected amino acid sites in close

proximity to the L-Trp binding pocket126, 172. Using the L-Trp bound TxtE crystal structure

(PDB: 4TPO)126 as the model template in our computation analysis, we identified 20

residues lining the L-Trp binding pocket (Figure 3-24) that may be important for controlling enzyme substrate promiscuity as well as regioselectivity: R59, V63, W82,

M88, Y89, M173, L241, I244, A245, A248, P249, T250, N293, F295, T296, W297, R298,

L321, E394, and F395.

Among these residues, M88 corresponds to the F87 residue in P450BM3. F87 possesses powerful and well-validated abilities to influence selectivity and regulate activity, and it is the most commonly mutated residue of P450BM357. T250 corresponds

to T268 in P450BM3. T268 has many potential roles in catalysis, including proton

delivery, dioxygen activation, and stabilization of oxy-ferrous or hydroperoxy catalytic intermediates57, but none of these roles seems relevant to the TxtE catalytic cycle. Y89

interacts with the substrate carboxylate moiety, but its aromatic nature and proximity to

the indole binding pocket also make it possible for altering regio-selectivity. F395 lies on

the left side of the indole binding pocket. The size of its aromatic side chain may

influence the indole orientation for nitration, thereby affecting regio-selectivity. N293

forms H-bonds with amine group and indole nitrogen of substrate L-Trp. T296 also

interact with the substrate amine group, while R59 and E394 interact with carboxylate

group via H-bonds. All other residues presumably provide hydrophobic environment for

substrate binding and nitration. All these residues were subjected to alanine mutation in

the following study.

Preparation and Functional Characterization of TB14 Binding Pocket Mutant Library

Alanine scanning is a commonly applied approach to quickly identify critical sites

affording desirable enzyme properties173. This approach was also employed in this work to mutate all identified sites of TxtE binding pocket, except A245 and A248. In total, 18 alanine mutants were created by quick-change PCR reactions using TB14 chimera gene as template and shown as the following: R59A, V63A, W82A, M88A, Y89A,

M173A, L241A, I244A, P249A, T250A, N293A, F295A, T296A, W297A, R298A, L321A,

E394A, and F395A (Table 3-5).

We prepared all 18 recombinant mutants in E. coli (Figure 3-25). Among them,

13 mutants were identified as properly folded P450 enzymes via a CO difference spectrum assay174 (Table 3-5). Recombinant W82A, Y89A, F295A, R298A, and L321A

were soluble but did not show a Soret peak at ~450 nm in their reduced-CO difference

spectra. This result indicated that these five residues may play important roles in

properly incorporating and interacting with the heme cofactor.

The nitration activities of all 13 properly folded variants were further examined

with L-Trp as substrate (Table 3-5). HPLC and LC-MS analysis indicated that nine

variants retained the L-Trp nitration activity (Table 3-5) but their activities were decreased to 21 to 38% of the wild type (Figure 3-26). R59A, M88A, T250A, and F395A completely lost the nitration activity towards the natural substrate L-Trp. To dissect the underlying basis of the lost function of these variants, we quantitated their binding with

L-Trp. R59A, M88A and F395A showed no spectral changes with serial concentrations of L-Trp, indicating the lost substrate interaction. This result indicated the critical roles of

R59, M88 and F395 in enzyme-tryptophan binding. Indeed, R59 is expected to interact with carboxylate group of L-Trp and an early report also confirmed its role172. L-Trp bound TxtE structure further suggested that M88 and F395 can interact with the indole ring in a sandwich manner (Figure 3-24). Replacing these residues with a small side- chain can significantly enlarge the binding pocket, leading to weak-to-no binding of original substrate. To further explore the functions of M88 and F395, we created a M88 saturation mutagenesis library and mutated F395 into aromatic amino acids Trp and

Tyr. Mutating M88 resulted in only three variants (M88A, M88L and M88K) that were properly folded based on CO difference spectrum assays, although many recombinant

mutants were soluble. This result further indicated that M88 can also contribute to the incorporation of the heme cofactor. M88L retained <5% of wild type’s activity toward L-

Trp while M88K was inactive in nitrating this substrate (Figure 3-26), likely suggesting that both the size and physiochemical properties of side-chains at this site are important to enzyme catalysis. Both F395W and F395Y were properly folded but their nitration activity toward L-Trp was significantly decreased (~5%) in comparison with wild type

(Figure 3-26). In contrast to M88A and F395A, the binding affinity between L-Trp and

T250A was approximately eight-fold higher than wild type (2 µM vs. 17 µM). This result suggested that T250 may play important catalytic roles, or the substrate is not in the catalytically active binding position. Additionally, TxtE contains a Pro residue N-terminal to T250 but an acidic residue commonly appears in this position69. To further assess the potential contributions of the Pro to the TxtE nitration activity, we mutated it into Asp and

Glu. Both P249D and P249E were properly folded and showed comparable or slightly higher nitration activity toward L-Trp in comparison to wild type (Figure 3-26). Therefore, the acidic side-chain at the position of 249 is not required for TxtE catalysis and the replacement of Pro might not induce significant changes on the role of T250.

Nonetheless, these data confirmed the roles of identified residues in the substrate binding and likely suggested the small side-chains in the binding pocket may provide unfavorable interactions with the native substrate.

Substrate Scope of TB14 Variants

Next, we examined the catalytic activity of all properly folded TxtE mutants with the chemical library we used to assess TB14 (Figure 3-1). TxtE and TB14 showed nitration activity toward eight L-Trp analogues including α-Me-Trp, 4-F-Trp, 4-Me-Trp, 5-

F-L-Trp, 5-Me-Trp, 5-MeO-Trp, 6-F-Trp and 7-Me-Trp. HPLC and LC-MS analysis

indicated that all alanine mutants completely lost their activity toward any substrates except for L-Trp and 7-Me-Trp (Figure 3-26). Seven alanine mutants showed substantial activities toward 7-Me-Trp. Among them, F395A nitrated this substrate to the comparable level of the wild type, while the two non-alanine mutants at the same residue, F395W and F395Y, were inactive toward 7-Me-Trp. How the methyl group of 7-

Me-Trp compensates the vacancy created by the F395A mutation in the binding and catalysis remains unknown. F395W, F395Y, M88L and M88K showed no nitration activity toward any additional substrates besides tryptophan in the test, while P249D and P249E nitrated tryptophan, 4-Me-Trp, 5-F-L-Trp and 7-Me-Trp to a significant extent (Figure 3-26). No nitrated product in the enzyme reactions carried altered regio- selectivity from wild type based on HPLC analysis.

We also probed the substrate promiscuity of all properly folded variants with the other 14 compounds (Figure 3-1B). Similar with TB14, none of the enzymes showed detectable activity toward these compounds except for styrene based on LC-MS analysis.

Importantly, N293A, I244A, and L241A mutants showed up to three folds higher activity than the wild type (Figure 3-27). We further combined these beneficial mutations to generate double and triple mutants, but none was more active than the single mutants, indicating that the beneficial effect of these three residues was not accumulative (Figure 3-27).

Conclusion and Discussion

The physiological substrate of TxtE was characterized as L-Trp120. Remarkably,

this enzyme showed a considerable substrate tolerance, particularly towards small

modifications on the indole ring126, and nitrated tryptophans derivatized with a single -F

or -Me substitution on their C5, C6, or C7. Dodani and coworkers also found that

racemic 4-Me-DL-Trp was an effective substrate of TxtE126, although the nitration is

anticipated to occur on the C4 of the indole ring of L-Trp120. In this report, we observed type I spectral changes of TxtE and TB14 induced by both 5-F-L-Trp and 4-F-DL-Trp.

Both enzymes exhibited improved activities toward 5-F-L-Trp over L-Trp, even though the binding of the former was weaker. TxtE and TB14 remained 15 to 20% of activities on 4-F-DL-Trp. Barry et al. reported that D-Trp was not the substrate of TxtE120, suggesting observed substrate binding and conversion might be ascribed only to 4-F-L-

Trp in the racemic mixture.

To ascertain enzyme regio-selectivity in nitrating L-Trp analogues, we isolated

nitrated products from large-scale reactions with both 4-F-DL-Trp and 5-F-L-Trp as

substrates. Nitrated products showed different fragmentation patterns in LC-MS/MS

analysis, suggesting different nitration sites in two F-L-Trp analogues. NMR analysis

clearly identified the nitro substitution on the C7 of 4-F-Trp and on the C4 of 5-F-L-Trp,

respectively, illustrating the remarkable regio-flexibility of TxtE in nitrating different

substrates. Regio-flexibility has previously been observed in several other P450

reactions. PikC hydroxylates both C12 and C10 of YC-17 in pikromycin biosynthesis175,

176, while TamI promotes two hydroxylation reactions at C10 and C18, and one

epoxidation reaction on C11=C12 of its substrate in tirandamycin biosynthesis146.

Furthermore, P450BM3 and engineered variants can hydroxylate different positions of

fatty acids and unnatural substrates57, 177, 178. However, these P450-pomoted regio- flexible reactions are commonly occurred on the same substrate. TxtE is distinguished from these P450s by the different regio-selectivity toward various substrates, an

100

intriguing feature shared by some recently characterized prenyltransferases179-181. To

structurally understand enzyme regio-selectivity, we docked 4-F-D/L-Trp and 5-F-L-Trp into the active site of TxtE structure (4TPO)126. 4-F-L-Trp, 4-F-D-Trp and 5-F-L-Trp adapted a similar orientation and binding pattern as the co-crystallized L-Trp (data not shown). The average distance between the C4 of the top 5 docked 5-F-L-Trp ligands and the heme iron was 6.3 Å (7.2 Å for L-Trp), making this site suitable for a P450 reaction175, 176, 182-184. However, this distance from the C7 of the top 5 docked 4-F-D/L-

Trp and the heme iron was reduced to approximately 3.1 Å, too close to allow P450- promoted reaction. Co-crystallization of 4-F-DL-Trp with TxtE is expected to provide precise structural explanation to the observed TxtE regio-flexibility. Nonetheless, this study described that a fluorine substitution in the substrate dramatically affected enzyme performance and illuminated the possibilities to create TxtE variants with altered regio-selectivity toward different aromatic substrates.

Furthermore, we demonstrated the production of both 4-Me-5-NO2-L-Trp and 4-

Me-7-NO2-L-Trp by the TB14-driven process. The demonstrated ability of TB14 to nitrate two sites of 4-Me-L-Trp marked TxtE as the promising material for developing

nitration biocatalysts to synthesize structurally diverse nitro aromatics via protein

engineering.

Notably, styrene and styrene analogues were also converted by TxtE and its

variants. The preliminary information of the product structure indicated the reaction was

not simply the nitration because the aromatic moiety remained intact and an ethyl group

was identified. Based on the 1H-NMR and HRMS analysis, a tentative structure was

deduced. However, if this deduced structure was true, the reaction must undergo very

101

unusual C-C bond break and reform. Structural characterization by 13C NMR and 2D-

NMR are under way to clarify the product structure. Future mechanistic studies of this unusual reaction will expand the fundamental understanding of enzyme reactions and guide following engineering efforts.

Methods and Materials

General Chemicals, DNA Sub-cloning, and Bacterial Strains

Molecular biology reagents and enzymes were purchased from Fisher Scientific.

Primers were ordered from Sigma-Aldrich. 4-Me-DL-Trp was from MP Biomedical

(Santa Ana, CA), while NOC-5 (3-(Aminopropyl)-1-hydroxy-3-isopropyl-2-oxo-1-triazene) was purchased from EMD Millipore. Other chemicals and solvents were purchased from

a 5 mm TXI Cryoprobe in the AMRIS facility at the University of Florida, Gainesville, FL,

USA. The instrument was operated at 600.17 MHz for 1H NMR and 150.9 MHz for 13C

NMR. Spectroscopy data were collected using Topspin 3.5 software. HRMS data were

obtained using a Thermo Fisher Q Exactive Focus mass spectrometer equipped with

electrospray probe on Universal Ion Max API source.

102

Large-scale Enzymatic Synthesis of Nitrated Products

To isolate sufficient amounts of nitrated products for structural determination, 18

μM TxtE-BM3R was used in a 10-mL reaction mixture containing 1.5 mM substrate, 3

mM NADP+, 3 mM glucose, ~30 units/mL self-prepared glucose dehydrogenase crude

extract, 3 mM NOC-5 in 100 mM Tris-HCl buffer (pH 8.0). The reactions in a 200-mL flask were incubated at 20 °C, 250 rpm overnight, and then terminated by 20 mL methanol or acidification to pH 1.0 with 6 M HCl. After centrifugation, the supernatant was concentrated in vacuo and then freeze-dried. The powders were redissolved in 3 mL methanol. Semi-preparation was performed by HPLC (Shimazu) with a semi-prep

C18 column (Agilent ZORBAX SB-C18, 5 µm, 9.4 x 250 mm).

Analytical and Semi-preparative HPLC Analysis

For analytical analysis, the HPLC column kept at 40 °C, water with 0.1% formic acid was used as solvent A and acetonitrile with 0.1% formic acid was used as solvent

B. For analytical analysis of Trp analogues and pyridine enzymatic reactions, the column was eluted first with 1% solvent B for 1 min and then with a linear gradient of 1-

20% solvent B in 8 min, followed by another linear gradient of 20-99% solvent B in 2 min. The column was further cleaned with 99% solvent B for 2 min and then re- equilibrated with 1% solvent B for 2 min. The flow rate was set as 1 mL/min, and the products were detected at 211 nm with a PDA detector. For analytical analysis of styrene analogue enzymatic reactions, the column was eluted first with 20% solvent B for 2 min and then with a linear gradient of 1-99% solvent B in 8 min. The column was further cleaned with 99% solvent B for 2 min and then re-equilibrated with 20% solvent

B for 2 min. The flow rate was set as 1 mL/min, and the products were detected at 211 nm with a PDA detector. For semi-preparative analysis, the column was first eluted with

103

20% solvent B (acetonitrile with 0.1% formic acid) for 3 min and then with a linear

gradient of 20-40% solvent B for 3 min, followed by a linear gradient of 40-99% solvent

B for 6 min. The column was then cleaned by 99% solvent B for 2 min and re-

equilibrated with 20% solvent B for 3 min. The flow rate was set at 3 mL/min, and the

products were detected at 211 nm with a PDA detector. All isolates were combined,

concentrated, freeze-dried, and then weighed.

LC-MS and LC-MS/MS Analysis of Nitrated Products

A SHIMADZU Prominence UPLC system fitted with an Agilent Poroshell 120 EC-

C18 column (2.7 µm, 3.0 x 50 mm) coupled with a Linear Ion Trap Quadrupole

LC/MS/MS Mass Spectrometer system was used in the studies. For LC, the methods

used were identical with the HPLC analysis. For MS detection, the turbo spray conditions were identical for all chemicals (curtain gas: 30 psi; ion spray voltage: 5500 V; temperature: 750 °C; ion source gas 1: 60 psi; ion source gas 2: 70 psi). For MS/MS analysis, the collision energy was 20 eV.

NMR Analysis of Nitrated Tryptophan and Styrene Analogues

In NMR analysis, chemical shifts were reported in parts per million (ppm)

downfield from tetramethylsilane. Proton coupling patterns were described as singlet

(s), doublet (d), double doublet (dd), triplet (t), and multiplet (m).

1 4-Me-5-NO2-L-Trp: H NMR (600 MHz, 100mM DCl in D2O) δ 7.22 (d, J = 9.0 Hz,

1H), 6.88 (s, 1H), 6.84 (d, J = 8.9 Hz, 1H), 3.80 (dd, J = 10.1, 5.1 Hz, 1H), 3.22 (dd, J =

15.6, 5.1 Hz, 1H), 2.84 (dd, J = 15.7, 10.1 Hz, 1H), 2.23 (s, 3H). 13C NMR (151 MHz,

D2O) δ 170.56, 154.71, 142.20, 138.27, 128.39, 127.74, 124.17, 118.79, 109.77,

+ + 109.69, 58.96, 53.66, 27.77, 15.15. HRMS (ESI ): calc. for C12H13N3O4 [M+H] :

264.0906, found: 264.0892.

104

1 4-Me-7-NO2-L-Trp: H NMR (600 MHz, 100mM DCl in D2O) δ 6.80 (d, J = 8.3 Hz,

1H), 6.63 (d, J = 8.1 Hz, 1H), 3.92 (dd, J = 10.7, 5.4 Hz, 1H), 3.22 (dd, J = 16.2, 5.4 Hz,

+ 1H), 2.98 (dd, J = 16.2, 10.8 Hz, 1H), 2.10 (s, 3H). HRMS (ESI ): calc. for C12H13N3O4

[M+H]+: 264.0906, found: 264.0893.

1 5-F-4-NO2-L-Trp: H NMR (600 MHz, D2O) δ 7.52 (d, J = 8.8 Hz, 1H), 7.35 (s,

1H), 6.95 (t, J = 10.2 Hz, 1H), 4.04 – 3.93 (m, 1H), 3.23 (dd, J = 15.3, 5.6 Hz, 2H), 3.05

13 (dd, J = 15.3, 8.4 Hz, 2H); C NMR (151 MHz, D2O) δ 171.29, 151.69, 150.03, 134.52,

131.66, 129.49, 129.41, 118.41, 118.34, 117.89, 110.47, 110.30, 105.89, 105.86, 72.01,

+ + 62.46, 59.31, 53.65, 27.09. HRMS (ESI ): calc. for C11H11FN3O4 [M+H] : 268.0728, found: 268.0728.

1 4-F-7-NO2-L-Trp: H NMR (600 MHz, D2O) δ 7.80 (t, J = 8.1 Hz, 1H), 7.29 (s,

1H), 7.22 (d, J = 9.1 Hz, 1H), 4.29 - 4.23 (m, 1H), 3.43 (dd, J = 11.7, 6.5 Hz, 2H), 3.32

13 (dd, J = 15.2, 8.2 Hz, 1H); C NMR (151 MHz, D2O) δ 171.29, 152.41, 150.65, 142.34,

142.24, 128.59, 128.48, 119.38, 115.29, 115.18, 108.40, 108.35, 72.00, 62.45, 59.30,

- - 53.73, 38.70, 26.72. HRMS (ESI ): calc. for C11H11FN3O4 [M-H] 266.0583, found:

266.0577.

1 Product of 4-vinylpyridine: H NMR (400 MHz, CDCl3) δ 8.64 (d, J = 6.2 Hz, 2H),

7.54 (d, J = 6.1 Hz, 2H), 2.80 (q, J = 7.7 Hz, 2H), 1.17 (t, J = 7.6 Hz, 3H).

1 Product of 4-methoxystyrene: H NMR (400 MHz, CDCl3) δ 7.57 (d, J = 9.1 Hz,

2H), 6.90 (d, J = 9.1 Hz, 2H), 3.80 (s, 3H), 2.78 (q, J = 7.6 Hz, 2H), 1.01 (t, J = 7.6 Hz,

3H). HRMS (ESI+): calc. 180.1025, found: 180.1017.

Marfey’s Derivatization

4-Me-DL-Trp and nitrated 4-Me-DL-Trp from enzyme reactions were reacted with

Marfey’s reagent following manufacture manual (Thermo Scientific). Derivatized

105

products were analyzed by LC-MS with A SHIMADZU Prominence UPLC system fitted with a Waters SymmetryShield TM RP-C18 column (3.5 µm, 4.6 x 100 mm) and a

Linear Ion Trap Quadrupole LC/MS/MS Mass Spectrometer system. The flow rate was

0.5 mL/min. The column was eluted with 90% solvent A (0.05 M triethylammonium acetate, pH 3.0), 10% solvent B (acetonitrile) for 2 min and then with a linear gradient of

10-50% solvent B for 60 min. The column was then cleaned by 50% solvent B for 5 min followed by a re-equilibration with 10% solvent B for 2 min. For MS detection, the turbo spray conditions used were: curtain gas: 30 psi; ion spray voltage: 5000 V; temperature:

550 °C; ion source gas 1: 30 psi; ion source gas 2: 20 psi.

106

Figure 3-1. Chemicals used in the TxtE and its variants substrate scope screening. (A): Chemical structures of Trp and its analogues that induced type I spectral shift of TxtE and TB14. (B): Chemical structures of other compounds screened against TxtE and TB14 in this study.

107

Figure 3-2. TxtE and TB14 nitrated Trp and its analogues to varying degrees. The reactions were prepared as described previously and incubated at 20 °C, 300 rpm for 45 minutes. All experiments were repeated at least three times. Groups showed significant difference in nitration activity between TxtE and TB14 were indicated with asterisks, with single or double asterisks indicating a significant difference of P < 0.05 or P < 0.01 respectively.

108

Figure 3-3. LCMS analysis of styrene nitration reaction catalyzed by TB14. (A): Ion extraction (XIC) spectrum of m/z value corresponding to the nitrated styrene (m/z=150), two peaks with retention time 3.1 min and 4.0 min were identified. Blue trace: complete TB14 catalyzed nitration reaction; red trace: reaction with heat-inactivated TB14; yellow trace: reaction without NOC5; black trace: reaction without NADPH regeneration system; green trace: reaction without styrene. (B) and (C): MSMS spectra of the peaks with retention time 3.1 min and 4.0 min. (D): Proposed scheme of the TB14 catalyzed styrene nitration reaction.

109

Figure 3-4. Styrene analogue library screened in the current research. Compounds in red and blue could be converted by TxtE and its variants; while compounds in black could not. Reactions with compounds in red yielded two product peaks, and reactions with compounds in blue yielded one product peak.

110

Figure 3-5. Relative nitration activities of TB14 towards styrene analogues. Blue bar: relative peak area with shorter retention time; red bar: relative peak area with longer retention time. The nitration activities were determined by the peak areas of XIC spectrum, with peak area (R.T.= 4.0 min) of wild type TxtE- BM3R set as 1. All experiments were repeated at least three times.

111

Figure 3-6. The MS2 spectra of 4-NO2-L-Trp (A), nitrated 5-F-L-Trp (B), and nitrated 4- F-DL-Trp (C). The fragmentation pattern in (A) and (B) was the same but it was different between (B) and (C). Compared with those in (A), the C5-F substitution increased the m/z values of most ions in (B) by 18 Da. Putative chemical structures of ions in red labeled peaks were shown in Figure 3-7.

112

COOH COOH NO2 F F/H NH3 NH3

N N H H Exact Mass: 250.08/268.08 NO2 Exact Mass: 268.07

-NH3 -HCOOH -HCOOH -NH3 HOOC COOH F NO2 NO2 F NH F/H F/H 2 H2N N N N + + N H H H H NO Exact Mass: 204.08/222.07 2 Exact Mass: 233.06/251.05 NO2 Exact Mass: 251.05 Exact Mass: 222.07 -H2O -CH2CO -H2O O O F +HO F NO2 F/H

N N N H H H NO NO2 Exact Mass: 215.05/233.04 2 Exact Mass: 209.04 Exact Mass: 233.04

Figure 3-7. Putative fragmentation pathways of 4-NO2-L-Trp, 4-NO2-5-F-L-Trp, and 7- NO2-4-F-L-Trp.

113

Figure 3-8. HRMS spectra of nitrated 5-F-Trp and 4-F-Trp in TB14 reactions.

114

YD-YZ-nitro-5f-lTrp-1H nitro-5f-lTrp 7.5263 7.5117 7.3550 6.9693 6.9512 6.9354 4.0040 3.9912 3.9803 3.6472 3.6063 3.5275 3.5183 3.5062 3.5012 3.4377 3.4268 3.4181 3.4073 3.2458 3.2363 3.2201 3.2108 3.0745 3.0605 3.0490 3.0349 2.5877

O 18 OH 17 - 19 O O 16 + 15 N 11 10 14 NH F 6 2 9 13 12 1 5 8 2 4 NH 3 7

B (s) F (dd) 7.35 3.05

A (d) C (t) D (m) E (dd) 7.52 6.95 3.99 3.23

1.02 1.10 0.99 1.00 0.97 0.97 0.18 0.24 0.31 0.28 .0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1 f1 (ppm)

YD-YZ-nitro-4f-ltrp-proton nitro-4f-lTrp 7.8170 7.8025 7.7899 7.2911 7.2299 7.2148 4.2701 4.2587 4.2471 3.6572 3.6158 3.5383 3.5315 3.5235 3.5189 3.5115 3.5008 3.4487 3.4378 3.4292 3.4183 3.3376 3.3240 3.3122 3.2986 120

O 110 15 OH 14 16

F 11 100 10 13 NH 6 2 9 12 1 5 90 8 2 4 NH 3 7 80 + N - O 17 O 18 19 70 C (d) F (dd) 7.22 3.32 60 A (t) B (s) D (m) E (dd) 7.80 7.29 4.26 3.43 50

-10 0.99 0.95 0.95 1.17 0.95 2.25 3.13 1.67 1.19

.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 f1 (ppm)

Figure 3-9. 1H NMR spectra of nitrated 5-F-Trp and 4-F-Trp products.

115

YD-YZ-nitro-5f-lTrp-C13 nitro-5f-lTrp

O 18 OH 17 - 19 O O 16 + 15 N 11 10 14 NH F 6 2 9 13 12 1 5 8 2 4 NH 3 7 53.6512 131.6586 27.0945 134.5248 117.8859 171.2859 118.4129 118.3445 110.4672 110.3012 105.8868 105.8610 150.0279 151.6903 62.4598 129.4856 129.4078 59.3128 72.0111

230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 -10 f1 (ppm) YD-YZ-nitro-4f-ltrp-C13 nitro-4f-lTrp

O 15, OH 14, 16,

F 11, 10, 13, NH 6, 2 9, 12, 1, 5, 8, 2, 4, NH 3, 7, 62.45 + N - O 17, O 18, 19, 59.30 128.48 53.73 108.35 72.00 26.72 119.38 171.29 108.40 142.24 150.65 152.41 142.34 38.70 128.59 115.18 115.29

180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 f1 (ppm) Figure 3-10. 13C NMR spectra of nitrated 5-F-Trp and 4-F-Trp products.

116

YD-YZ-nitro-5f-lTrp-HSQC nitro-5f-lTrp 0

20 {3.22,27.08} {3.25,27.07}

30 {3.06,27.12}

{4.00,53.68} 50

{3.61,59.41} 60

{8.87,74.61} 70

80 f1 (ppm)

100

{6.96,110.64} 110 {7.52,118.61} O 120 15, OH - 14, 16, {7.36,131.98} O O + 19, 18, 130 N 11, 10, 17, NH2 F 6, 140 9, 13, 12, 1, 5, 8, 150 2, 4, NH 3, 7,

9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 f2 (ppm)

YD-YZ-nitro-4f-ltrp-HSQC nitro-4f-lTrp 0

{4.26,53.72} 50 {3.61,59.41} 60

80 f1 (ppm)

O 100 {7.22,108.47} 15, OH 14, 16, 110

{7.81,119.34} F 11, 10, 13, 120 NH2 6, {7.29,128.43} 9, 12, 1, 5, 130 8, 2, 4, NH 140 3, 7, + N 150 - O 17, O 18, 19,

9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 f2 (ppm) Figure 3-11. HSQC NMR spectra of nitrated 5-F-Trp and 4-F-Trp products.

117

YD-YZ-nitro-5f-lTrp-HMBC-10hz nitro-5f-lTrp 0

20 {4.00,26.99}

{3.06,53.67} 50 {3.61,59.51} 60

90 f1 (ppm) {7.36,106.01} {3.06,105.99} {3.22,106.02} 100

110 {7.36,118.05} {7.52,118.09} 120 {6.96,129.61} {3.06,131.82} {3.22,131.84} {6.95,134.66} 130

{7.36,134.70} O 140 15, H {7.52,151.03} O - 14, 16, O O 150 + 19, 18, N 11, 10, 17, 160 NH2 F 6, {3.99,171.43} 9, 13, 12, 1, 5, 170 8, 2, 4, 180 NH 3, 7,

9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 f2 (ppm)

YD-YZ-nitro-4f-ltrp-HMBC nitro-4f-lTrp 0

{3.53,53.67} 50 {3.62,59.29} 60

100 f1 (ppm) {7.29,108.46} {3.32,108.39}{3.53,108.44} {7.29,115.25} 110

120 {7.22,128.56} {3.53,128.52} O 15, OH 130 14, 16, {7.30,142.34} 140 F 11, 10, 13, NH2 6, 150 9, 12, 1, 5, 8, 160 2, 4, NH 3, 170 7, + N 180 - O 17, O 18, 19,

9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 f2 (ppm) Figure 3-12. HMBC NMR spectra of nitrated 5-F-Trp and 4-F-Trp products.

118

Figure 3-13. LC-MS analysis of Marfey’s derivatization of 4-Me-DL-Trp and its nitro product. Blue: ion extract spectrum of Marfey’s derivatized 4-Me-DL-Trp; Green: ion extract spectrum of Marfey’s derivatized 4-Me-DL-Trp after the enzyme reaction; and Red: ion extract spectrum of Marfey’s derivatized nitro product.

119

Figure 3-14. LC-HRMS analysis of the isolated 4-Me-NO2-L-Trp product. Two peaks with the retention times of 9.08 min and 9.67 min showed the same m/z value of approximately 264.0892.

120

Figure 3-15. Both 4-Me-5-NO2-L-Trp and 4-Me-7-NO2-L-Trp were produced in the TB14 reaction with 4-Me-DL-Trp as substrate. The aromatic region of 1H NMR spectrum of isolated product showed the chemical shifts of two sets of aromatic protons. Integrated values of these protons were also included.

121

1 Figure 3-16. H NMR spectrum of the isolated 4-Me-NO2-L-Trp product.

Figure 3-17. COSY spectrum of the isolated 4-Me-NO2-L-Trp product.

122

13 Figure 3-18. C NMR spectrum of the isolated 4-Me-NO2-L-Trp product.

Figure 3-19. HSQC spectrum of the isolated 4-Me-NO2-L-Trp product.

123

Figure 3-20. HMBC spectrum of the isolated 4-Me-NO2-L-Trp product.

90 {3.81,27.59} 40 {3.20,53.64} {2.84,53.74} 100 60 {6.88,109.70} 110 80 {6.89,124.22} 120 {7.21,127.68} {3.80,109.74} 100 {6.83,124.69} 130 {7.21,138.10} {6.83,142.36} f1 (ppm) {3.20,124.85} {3.19,109.94} 120 140 {6.88,138.19} {2.84,124.30} 150 {3.20,128.24} 140 {2.22,142.18} 160 {3.81,170.55} 160 170 {3.20,170.59} 180 8.0 7.0 6.0 4.0 3.0 2.0 f2 (ppm) COOH

15.15 27.70 53.60

NH 127.74 2 O2N 109.80 142.20 124.19 128.39 138.27 118.79 109.69 N H

Figure 3-21. Expansion of HMBC spectrum in aromatic and aliphatic regions and the representative correlations on 4-Me-5-NO2-L-Trp.

124

Figure 3-22. 1H NMR spectrum of the isolated 4-methoxystyrene converted product and proposed product structure.

125

Figure 3-23. 1H NMR spectrum of the isolated 4-vinylpyridine converted product and proposed product structure.

126

Figure 3-24. Selected 20 amino acid residues lining the TxtE L-Trp binding pocket.

127

Figure 3-25. SDS-PAGE analysis of purified TB14 mutants. M: protein marker; Lane 1- 25: wild type TB14, R59A, V63A, W82A, M88A, Y89A, M173A, L241A, I244A, P249A, T250A, N293A, F295A, T296A, W297A, R298A, L321A, E394A, F395A, F395W, F395Y, M88L, M88K, P249D and P249E.

128

Figure 3-26. Relative nitration activities of TB14 variants. The wild type TB14 nitration activity on L-Trp was set as 100. Only substrates could be nitrated by TB14 wild type or variants were shown. Standard error of each data was less than 10%.

129

Figure 3-27. Relative nitration activities of TB14 variants with styrene as the substrate. The nitration activities were determined by peak areas of XIC spectra, with peak area at 4.0 min from the wild type TB14 reaction as 100. Blue bar: relative peak area with retention time 4.0 min; red bar: relative peak area with retention time 3.1 min. All experiments were repeated at least three times.

130

Table 3-1. Binding affinities of 20 Trp analogues toward TxtE and TB14a. Substrate analogues TxtE Kd (µM) TB14 Kd (µM) L-Trp 25 ± 1 17 ± 1 D-Trp 430 ± 17 470 ± 25 Indole-3-pyruvic acid 390 ± 43 480 ± 32 Indole-3-lactic acid > 7 mM > 7 mM 3-Indoleacetic acid > 17 mM > 17 mM Tryptamine 68 ± 6 60 ± 8 Serotonin 320 ± 31 310 ± 39 Indole-3-carboxylic acid > 9 mM > 9 mM 2,3,4,9-Tetrahydro-1H-β- 580 ± 44 730 ± 86 carboline-3-carboxynate L-tryptophanol 220 ± 32 280 ± 43 α-Me-Trp 100 ± 4 95± 9 7-Azatryptophan 17 ± 1 26 ± 1 4-F-Trp 190 ± 11 180 ± 13 4-Me-Trp 10 ± 1 14 ± 1 5-OH-L-Trp 340 ± 26 450 ± 16 5-MeO-Trp 29 ± 1 26 ± 2 5-Me-Trp 35 ± 1 30 ± 1 5-F-L-Trp 84 ± 4 78 ± 6 6-F-Trp 73 ± 6 65 ± 5 7-Me-Trp 13 ± 1 11 ± 1 a: Binding affinities of chemicals to P450s were measured using 1.5 μM of enzymes in 25 mM Tris-HCl, pH 8.0. All experiments were repeated at least three times. Data were fitted to the Michaelis-Menten equation.

131

Table 3-2. Binding affinities of styrene analogues toward TB14a. Substrate analogues TB14 Kd (µM) Binding type Styrene 246 ± 10 Type I 2-Vinylnaphthalene 25 ± 1 Type I 4-Vinylpyridine 4 ± 0.2 Type II 2-Vinylpyridine 367 ± 33 Type I 4-Methylstyrene 152 ± 7 Type I 4-Fluorostyrene 214 ± 12 Type I 4-Chlorostyrene 152 ± 6 Type I 3-nitrostyrene 482 ± 9 Type I Ethylbenzene 247 ± 12 Type I trans-β-Methylstyrene 128 ± 11 Type I 4-Vinylphenol N.A. No binding Allylbenzene 280 ± 12 Type I α-Methylstyrene 122 ± 12 Type I Vinylcyclohexane N.A. No binding beta-nitrostyrene N.A. No binding 4MeO-styrene 35 ± 6 Type I 4-trifluoromethylstyrene 35 ± 8 Type I 2,4,6-Trimethylstyrene 51 ± 10 Type I 4-Vinylbiphenyl 55 ± 5 Type I 2,3,4,5,6-Pentafluorostyrene 156 ± 59 Type I 4MeO-styrene 35 ± 6 Type I 4-trifluoromethylstyrene 35 ± 8 Type I 2,4,6-Trimethylstyrene 51 ± 10 Type I 4-Vinylbiphenyl 55 ± 5 Type I 2,3,4,5,6-Pentafluorostyrene N.A. No binding aBinding affinities of chemicals to P450s were measured using 1.5 μM of enzymes in 25 mM Tris-HCl, pH 8.0. All experiments were repeated at least three times. Data were fitted to the Michaelis-Menten equation.

132

13 1 Table 3-3. C and H NMR data for 5-F-4-NO2-L-Trp and 4-F-7-NO2- L-Trp (recorded in D2O). Atom 5-F-4-NO2-L-Trp 4-F-7-NO2-L-Trp a b a b δC , type δH (J in Hz) δC , type δH (J in Hz) 2 131.7, CH 7.35 s 128.5, CH 7.29 s 3 105.9, C 108.3, C 3a 117.9, C 115.2, C 4 129.5, C 151.5, C 5 150.9, C 119.4, CH 7.80 dd (8.1, 8.1) 6 110.4, CH 6.95 dd (10.2,10.2) 108.4, CH 7.22 d (9.1) 7 118.4, CH 7.52 d (8.8) 128.6, C 7a 134.5, C 142.3, C 1’ 171.3, C 171.3, C 2’ 53.7, CH 3.99 m 53.7, CH 4.26 m 3’ 27.1, CH2 3.23 dd (15.3, 5.6) 26.7, CH2 3.43 dd (15.2, 6.5) 3.05 dd (15.3, 8.4) 3.32 dd (15.2, 8.2) aData recorded at 151 MHz. bData recorded at 600 MHz

133

13 1 1 Table 3-4. C and H NMR data of 4-Me-5-NO2-L-Trp and 4-Me-7-NO2-L-Trp and H NMR data of 4-Me-7-NO2-L-Trp. c Atom 4-Me-5-NO2-L-Trp 4-Me-7-NO2-L-Trp 4-F-7-NO2-L-Trp a b δC , type δH (J in Hz) δC, type δH (J in Hz) δC, type δH (J in Hz) 2 128.4, CH 6.88 s -, CH 128.5, CH 7.29 s 3 109.8, C -, C 108.3, C 3a 124.2, C -, C 115.2, C 4 127.7, C -, C 151.5, C 5 142.2, C -, CH 6.63 d (8.1) 119.4, CH 7.80 dd (8.1, 8.1) 6 118.7, CH 7.22 d (9.0) -, CH 6.80 d (8.3) 108.4, CH 7.22 d (9.1) 7 109.7, CH 6.84 d (8.9) -, C 128.6, C 7a 138.3, C -, C 142.3, C 1’ 171.4, C -, C 171.3, C 2’ 53.6, CH 3.80 dd (10.1, -, CH 3.92 dd (10.7, 53.7, CH 4.26 m 5.1) 5.4) 3’ 27.7, CH2 3.22 dd (15.6, -, CH2 3.22 dd (16.2, 26.7, CH2 3.43 dd (15.2, 5.1) 5.4) 6.5) 2.84 dd (15.7, 2.98 dd (16.2, 3.32 dd (15.2, 10.8) 8.2) 10.1)

4-Me 15.15, CH3 2.23 s -, CH3 2.10 s - - aData recorded at 151 MHz; bdata recorded at 600 MHz;

134

Table 3-5. Nitration activities of TB14 Alanine mutants. TB14 variants Active P450 Trp Nitration activity Styrene Nitration activity Wild type √ √ √ R59A √ X √ V63A √ √ √ W82A X X X M88A √ X X Y89A X X X M173A √ √ √ L241A √ √ √ I244A √ √ √ P249A √ √ √ T250A √ X √ N293A √ √ √ F295A X X X T296A √ √ √ W297A √ √ √ R298A X X X L321A X X X E394A √ √ √ F395A √ X √

135

CHAPTER 4 DEVELOPING E. coli CELL FACTORIES FOR THE PRODUCTION OF NITRO AROMATICS

Introduction

The nitro (-NO2) group acts as an essential unit in a number of pharmaceuticals185, exemplified by anticancer drug nilutamine, antiparkinson agent tolcapone, and anti-infective agents chloramphenicol and the recently

approved delamanid186 and nifurtimox-eflornithine combination187. Drug

candidates bearing the -NO2 group also commonly appear in drug pipelines for treating a variety of existing and emerging diseases130, 131, 188. Additionally, the nitro group in particular is a versatile synthetic handle present in numerous building blocks in the synthesis of complex drug molecules189-191. The

fundamental importance of the nitro group in pharmaceutical industry has driven

the development of chemical nitration methods93. Indeed, syntheses of nitro chemicals, particularly nitro aromatics, are one of the most studied organic reactions, and classical electrophilic nitration methods with nitric acid as the nitrating reagent dominate current industrial processes93. The limitations of the

electrophilic method, however, is that it is notoriously non-selective, poorly

tolerates other functional groups, poses safety concerns, and generates large

quantities of acidic waste. Advanced nitration methods93-97 have recently been developed to account for these issues but none of them was successful on the scales required for pharmaceutical production.

Developing efficient catalysts that can sustainably produce a broad range of

chemicals represents a major challenge in modern organic chemistry. Enzymes as

biocatalysts are of great synthetic interest because their typically high stereo-, regio-

136

and chemo-selectivity avoids lengthy protection/deprotection steps and the generation

of impurities. Furthermore, biocatalytic processes are often non-toxic, require generally

mild reaction conditions, and leave no residual heavy metal contamination1, 192, 193. As a

result, the footprint of enzymes in the industrial production of chemicals is ever-

increasing, exemplified by biocatalytic manufacturing of anti-diabetic drug sitagliptin7. In

these applications, enzymes are capable of catalyzing a breadth of organic

transformations113. However, despite Nature’s successful application of several

strategies to synthesize nitro-containing compounds132, there has been comparably little

effort to exploit them in chemical industry.

The use of biocatalysts for the industrial synthesis of chemicals has been attracting much attention due to their environment compatibility and high selectivity and specificity8, 193. Microbial cells as biocatalyst play a leading role in “chemo-enzymatic

synthesis” because of their incredible diversity and validated engineering potential.

Compared with purified enzymes, using whole cells is more cost-effective, provides a natural environment to prevent the loss of enzyme activity during purification and reactions, and is able to efficiently regenerate co-factors, particularly for redox enzymes including P450s. Recent advances in the emerging field of synthetic biology further expand the tools for the design, engineering, and optimization of microbial cell factories for the production of chemicals194, 195. The industrial nitration process is devastatingly

polluted and often generates a significant amount of unwanted side products, which

lowers the quality of desirable products and increases the production cost. TB14 is the

most active aromatic nitration biocatalyst ever developed. TB14 is also noteworthy in

having relatively broad substrate scope, suggesting its promising uses to synthesize a

137

variety of nitro Trp analogues for biomedical and biological applications. An important goal of this research is to develop a bacterial cell manufacturing platform that leverages

TxtE to produce nitro compounds.

Results

Creation of E. coli Strain for the Whole Cell Nitration Reaction

In this chapter, an E. coli based biotransformation system for the production of nitrated L-Trp was developed (Figure 4-1). The engineered E. coli strain contained three functional genes, TB14, nitro oxide synthase (NOS), and GDH. As described above,

TB14 is a self-sufficient nitration biocatalyst and soluble in E. coli. The co-substrate NO is indispensable for TxtE nitration reaction. In the in vitro assays, NO was derived from the NO precursor NOC-5 that is expensive, short-life and incompatible with bacterial cells (NO at high concentration is toxic). The thaxtomin biosynthetic gene cluster in

Streptomyces scabies contains a TxtD gene encoding a nitro oxide synthase that converts L-Arg into L-citrulline and NO along with the consumption of NADPH120. The expression of the NOS gene in E. coli can provide a sustainable and environment- friendly approach to eliminate the dependence of the high-cost and unstable NO precursors in whole cell nitration biotransformation. However, we expressed the TxtD gene from two thaxtomin-producing Streptomyces strains (Streptomyces scabies and

Streptomyces turgidiscabies) and yielded only insoluble proteins after optimizing expression conditions. Instead, a codon-optimized NOS gene from Bacillus subtilis

(BsNOS, provided by Prof. Thomas Poulos, UC-Irvine)196 led to the production of soluble protein in E. coli and was used in the subsequent research. The reaction of

NOS requires redox partners for transferring electrons from NADPH. Fortunately, non- specific redox partners of E. coli effectively support the BsNOS reaction197, making

138

BsNOS containing E. coli strain a viable biosystem to supply NO for the nitration

reaction. Finally, as insufficient supply of NADPH could limit the productivity of the

biotransformation, the GDH gene from Bacillus subtilis198 was also engineered into E.

coli to regenerate NADPH that is consumed in both TB14 and BsNOS reactions. GDH

catalyzes the oxidation of β-D-glucose to β-D-glucono-1,5-lactone with simultaneous

reduction of the cofactor NADP+ to NADPH, and it is commonly applied in biocatalysis

procedures to regenerate NADPH.

The availability of the three functional genes enables engineering bacterial nitro

chemical factories. In the first attempt, we co-expressed TB14 and BsNOS genes using

vector pETDuet, while the GDH gene was separately expressed in the vector pET28b.

Both vectors have the same, medium copy numbers (15 to 60) in the host and drive the

expression of each gene with a strong inducible promoter T7. In addition, the different

antibiotic resistant markers (ampicillin R and kanamycin R) in these two vectors make

them suitable for simultaneous expression of three genes in the same host. The two

constructs described above were transformed into E. coli BL21-GOLD strain. The

overexpression of the three enzymes was induced by IPTG. SDS-PAGE analysis of the

soluble crude extract (Figure 4-2A, lane 1) revealed the successful overexpression of

BsNOS (42 kD) and GDH (28 kD). By contrast, only a small quantity of soluble TB14

(110 kD) was observed. Compared with our previous results these data indicated that

the co-expression of BsNOS and/or GDH negatively influenced the expression of TB14.

Nevertheless, the engineered E. coli strain was used in the whole-cell biotransformation

to produce 4-NO2-L-Trp from fed L-Trp. After 20-h incubation, the successful production of 4-NO2-L-Trp was confirmed by LC-MS (Figure 4-2B).

139

Optimization of Heterologous Enzyme Expression

Previous results indicated the very low expression level of TB14 in the E. coli

strain engineered with pETDuet-TB14-BsNOS and pET28b-GDH, which can likely be

the bottleneck of the transformation efficiency. In order to increase the expression level

of TB14, we sought to vary the copy numbers of these three genes and to improve

plasmid stability. Specifically, we included pACYCDuet for the co-expression of two

target genes. The pACYCDuet plasmid that includes two T7 promoters to drive the

proteins expression carries the P15A replicon instead of pBR322-derived ColE1

replicon as in pETDuet and pET28b, which can provide higher plasmid stability when

two plasmids are used. Subsequently, we created three new pairs of expression

constructs, pETDuet-GDH -BsNOS+pET28b-TB14, pACYCDuet-TB14-

BsNOS+pET28b-GDH, and pACYCDuet-GDH-BsNOS+pET28b-TB14, and the corresponding engineered E. coli strains (Figure 4-3A). We then examined the protein expression levels in these strains by SDS-PAGE (Figure 4-2A). The strain transformed with the pair of pETDuet-GDH-BsNOS and pET28b-TB14 plasmids showed most significantly increased TB14 expression level, and it demonstrated the highest nitration activity (Figure 4-3B). This strain thus was used in the following optimization efforts.

Optimization of Fermentation Conditions

To further improve the activity of the whole cell nitration system, we aimed to optimize the fermentation conditions, including medium, temperature, substrate supplement and harvesting time. We used the minimal medium M9 in the previous fermentation process because it is the most widely used medium for the whole cell transformations68, 81, 84. As M9 medium is nutritiously poor, we then examined whether

the nutrition availability could influence the whole cell nitration efficiency. Three nutrition

140

rich media, including LB medium, SOC medium and TB medium, were tested along with the M9 medium in the whole cell nitration system. As shown in (Figure 4-4), the transformations supported by nutrition richer media generally had higher efficiency than those supported by the M9. Notably, the TB medium supported transformation yielded as high as 600 µM of nitrated tryptophan (149 mg/L) after 20-hour fermentation.

We further tested the time profile of the product formation (Figure 4-4). In all the nutrition rich media, the highest productivity was observed at approximately 20 hours.

By contrast, it required 3 days in the M9 medium, clearly suggest the use of nutrition rich media for the whole-cell nitration. Notably, the concentration of nitrated tryptophan in the rich media started to decrease after 30-hour fermentation. We speculated that the

E. coli endogenous tryptophanase (EC 4.1.99.1), which could convert the tryptophan to

199 indole, pyruvate and NH3 , might be responsible for the production decomposition. We then cloned E. coli tryptophanase gene tnaA and prepared the recombinant enzyme.

However, our in vitro biochemical studies demonstrated that TnaA was not able to decompose 4-NO2-Trp (data not shown). The pathway for the product decomposition therefore remained unknown.

In the whole cell transformation, the key cosubstrate NO is generated from L- arginine by BsNOS. In next step, we further tested whether increasing the concentration of L-Arg would be beneficial for the nitrated tryptophan production (Figure 4-4).

However, no significant change of the production was observed when 5 mM of L-Arg was added to each of the transformation system. This result likely suggests that L-Arg or NO is not the limiting factor in the whole cell transformation. We also tested the effect of increased concentration of the substrate L-Trp (Figure 4-4). No significant change of

141

the production was observed when 5 mM of L-Trp was added to the transformation

system.

Finally, we examined the temperature effects on the whole cell nitration. All

previous experiments were performed at 20 °C but the optimal temperature range for

the TB14 activity in in vitro studies was 10 to 30 °C200. We therefore examined the productivity of the whole cell system at four different fermentation temperatures (15 °C,

20 °C, 28 °C and 37 °C) at different time points (Figure 4-5). The transformations at

15 °C, 20 °C and 28 °C have good product yield, while the optimal temperature was

28 °C with the product concentration of 700 µM after 24-hour fermentation. Interestingly, the optimal growth temperature of E. coli, 37 °C, almost completely abolished the nitration transformation. In conclusion, we identified the optimal transformation conditions as TB medium at 28 °C, which were used in the following study.

Production of Nitrated Tryptophan Analogues by Whole Cell Nitration System

In Chapter 3, we have identified a series of tryptophan analogues that can be nitrated by TxtE and its variants in vitro. In this section, we sought to examine the substrate scope of the whole cell system with these tryptophan analogues. Of note, these unnatural analogues should compete with the native substrate L-Trp abundant in the TB medium in the whole cell transformation. Remarkably, under the optimal conditions, α-Me-Trp, 4-F-Trp, 4-Me-Trp, 5-MeO-Trp, 5-Me-Trp, 5-F-Trp, 6-F-Trp and 7-

Me-Trp all were successfully nitrated using the whole cell nitration system. Similar to the in vitro enzymatic reactions (Chapter 3), the whole cell nitration demonstrated the highest conversion rates with 4-Me-Trp, 5-Me-Trp and 5-F-Trp (Figure 4-6). The best substrate was 5-Me-Trp, and the product concentration reached approximately 250 µM after 24-hour fermentation, along with approximately 320 µM of nitrated tryptophan

142

(Figure 4-6). To eliminate the production of nitro-tryptophan, we cultured the strain in TB

and then resuspended the cell pellets in Synthetic Complete medium without L-Trp for

the whole cell transformation. Unfortunately, this approach led to only low product

concentrations (0-50 µM).

Conclusion and Discussion

Biocatalysis is recognized as a valuable addition to traditional chemical synthesis

routes. The economic costs, activities and stabilities are the most considered factors in

the application of chemical manufacturing35. Compared to the in vitro enzyme catalyzed

reactions, the whole cell catalysts circumvent the cell lysis and enzyme purification

steps and hence significantly cut the cost and labor. Whole cell catalysts are the

cheapest catalyst possible. In a previous study, the cost of whole cell catalysts in

biocatalyst production was found to be ten-fold lower than the purified enzymes201.

Whole cell biocatalysts are more preferred in the case of cofactor-dependent reactions. The host microorganism may contain inherent universal cofactors which support target reactions and circumvent the requirement of expensive external cofactors supplement. For example, reactions catalyzed by oxidoreductases (including P450s) were often utilized in the form of whole cell catalysts because these reactions usually depend on NADH or NADPH cofactors and sometimes redox partners. In our case,

Both TB14 and BsNOS require NADPH. In addition to endogenous NADPH in the E. coli host, we further provided GDH to regenerate NADPH for the whole cell biotransformation. More importantly, although the TB14 was already engineered to a self-sufficient P450 enzyme (Chapter 2), the BsNOS still requires redox partners to transfer electrons from NADPH to itself. Using the whole cell biocatalysis, we

143

circumvented this barrier because the E. coli cells could produce universal redox

partners which effectively support the BsNOS reaction197.

The most significant advantage of this whole cell nitration system compared to

the in vitro reaction is the in situ production of NO from L-Arg. The expensive NO

donors are the major barrier for the industrial application of TB14 as nitration

biocatalyst. With the help of functional BsNOS in the whole cell nitration system, we

made the cell produce NO from L-Arg, which could also be synthesized by the E. coli

cell from cheap carbon and nitrogen source, and hence greatly lowered the cost of the

biocatalytic nitration process.

While optimizing the fermentation conditions, we also noticed that only nutrition rich medium (TB medium) could support good product yield (up to 700 µM for nitrated L-

Trp and up to 250 µM for nitrated L-Trp analogues). It might due to the high metabolic

burden of the cells. Indeed, three heterologous proteins were over-expressed in the host cell and both BsNOS and TB14 reactions required NADPH generated from the carbon source.

In conclusion, we have successfully constructed a whole cell nitration system for the production of nitrated tryptophan and tryptophan analogues. After optimization of the fermentation procedures, the production could reach as high as 700 µM in the reaction mixture. The significant economic and labor cost reduction laid a solid foundation for the industrial application of the nitration biocatalysts.

Methods and Materials

General Chemicals, DNA sub-cloning, and Bacterial Strains

Molecular biology reagents and enzymes were purchased from Fisher Scientific.

Primers were ordered from Sigma-Aldrich. 4-Me-DLTrp was from MP Biomedical (Santa

144

Ana, CA), while NOC-5 (3-(Aminopropyl)-1-hydroxy-3-isopropyl-2-oxo-1-triazene) was

purchased from EMD Millipore. Other chemicals and solvents were purchased from

Sigma-Aldrich and Fisher Scientific. Escherichia coli DH5α (Life Technologies) was

used for cloning and plasmid harvesting. E. coli BL21-GOLD (DE3) (Agilent) was used for protein overexpression. E. coli strains were grown in Luria-Bertani broth or Terrific broth. DNA sequencing was performed at Eurofins. A Shimadzu Prominence UHPLC system (Kyoto, Japan) fitted with an Agilent Poroshell 120 EC-C18 column (2.7 µm, 3.0 x 50 mm), coupled with a PDA detector was used for HPLC analysis.

Construction of Plasmids for Whole Cell Transformation

TB14 gene was amplified from TB14/pET28b constructed in Chapter 2 using a

pair of TB14FN and TB14RH primers in PCR reactions (Table 4-2). The PCR product

was analyzed by agarose gel and extracted with a GeneJET Gel Extraction Kit

(Thermo). Purified PCR products, pACYCDuet and pETDuet were digested with the

restriction enzymes NcoI and HindIII, and corresponding linear DNAs were ligated to

generate expression constructs. GDH gene was amplified from GDH /pET21b (Chapter

2) using a pair of GDHFB and GDHRH primers in PCR reactions (Table 4-2). The PCR

product was analyzed by agarose gel and extracted with a GeneJET Gel Extraction Kit

(Thermo). Purified PCR products, pET28b, pACYCDuet and pETDuet were digested

with the restriction enzymes BamHI and HindIII, and corresponding linear DNAs were

ligated to generate expression constructs. BsNOS gene was amplified from BsNOS

/pET15b kindly provided by Professor Dennis J. Stuehr (Lerner Research Institute)

using a pair of BsNOSFN and BsNOSRH primers in PCR reactions (Table 4-2). The

PCR product was analyzed by agarose gel and extracted with a GeneJET Gel

Extraction Kit (Thermo). Purified PCR products, pET28b, pACYCDuet and pETDuet

145

were digested with the restriction enzymes NdeI and HindIII, and corresponding linear

DNAs were ligated to generate expression constructs. SsTxtD and StTxtD were amplified from genomic DNA of S. scabies 87.22 (NRRL B-24449) and S. turgidiscabies

Car8 using a pair of SsTxtDFN/ SsTxtDRH and StTxtDFN/ StTxtDRH primers in PCR reactions (Table 4-2). The PCR product was analyzed by agarose gel and extracted with a GeneJET Gel Extraction Kit (Thermo). Purified PCR products and pET28b were digested with the restriction enzymes NdeI and HindIII, and corresponding linear DNAs were ligated to generate expression constructs. All inserts in the constructs were sequenced to exclude mutations introduced during PCR amplification and gene manipulation.

Heterologous Expression and Purification of Recombinant Proteins

Protein expression and purification followed the previous protocols154. The purified proteins were exchanged into storage buffer (25 mM Tris-HCl, pH 8.0, 100 mM

NaCl, 3 mM βME, and 10% glycerol) by PD-10 column, aliquoted and stored at -80°C until needed. CO difference spectroscopy was used to measure the concentrations of functional P450s152.

Analytical HPLC Analysis

For analytical analysis, the HPLC column kept at 40 °C, water with 0.1% formic acid was used as solvent A and acetonitrile with 0.1% formic acid was used as solvent

B. The column was eluted first with 1% solvent B for 1 min and then with a linear gradient of 1-20% solvent B in 8 min, followed by another linear gradient of 20-99% solvent B in 2 min. The column was further cleaned with 99% solvent B for 2 min and then re-equilibrated with 1% solvent B for 2 min. The flow rate was set as 1 mL/min, and the products were detected at 211 nm with a PDA detector.

146

Whole-cell Biotransformation

E. coli BL21 Gold cells containing pETDuet and pET28b derived plasmids were

grown from glycerol stock overnight in 5 mL Luria broth with 0.1 mg/mL ampicillin and

0.05 mg/mL kanamycin (37 °C, 250 rpm). The pre-culture was used to inoculate 100 mL

of Terrific broth medium (0.1 mg/mL ampicillin and 0.05 mg/mL kanamycin) in a 500 mL

flask; this culture was incubated at 37 °C, 250 rpm to OD600 = 0.6 - 0.8. The cultures

were cooled on water-ice mixture and induced with 0.5 mM IPTG. Expression was

conducted at 18 °C, 250 rpm, for 20 h. For the culture of E. coli BL21 Gold cells

containing pACYCDuet and pET28b derived plasmids, 0.05 mg/mL chloramphenicol

was used instead of 0.05 mg/mL kanamycin. The cultures were then harvested and

resuspended to OD600 = 30 in test medium. Aliquots of the cell suspension were used in

the whole cell transformation. To a test tube was added 5mL cell suspension, 25 µL

100mM L-Trp or L-Trp analogues, and 25 µL 100 mM L-Arg when necessary. The mixture was then incubated at different conditions. The reactions were quenched by adding equal volume of methanol and the resulting mixture was aliquoted and transferred to a microcentrifuge tube and centrifuged at 14,000 rpm for 10 minutes. The supernatant was transferred to an HPLC vial and analyzed by LC-MS.

147

Figure 4-1. Schematic overview of bacterial cell factories for the production of nitro- chemicals.

148

Figure 4-2. Creation and initial activity test of the whole cell nitration system. (A): SDS- PAGE analysis of whole cell nitration systems. M, marker; 1, pETDUET- TB14-BsNOS+pET28b-GDH; 2, pETDUET-GDH-BsNOS+pET28b-TB14; 3, pACYCDUET-TB14-BsNOS+pET28b-GDH; 4, pACYCDUET-GDH- BsNOS+pET28b-TB14. Three soluble recombinant proteins were highlighted with red arrows. (B): LCMS analysis (positive mode) of products in the whole cell nitration. The reaction was quenched with twice the volume of methanol. After centrifugation, the crude extracts were subjected to LCMS analysis.

149

Figure 4-3. Creation and analysis of different whole cell nitration systems. (A): Schematic representation of plasmids combination used in the whole cell nitration systems. (B): Nitrated tryptophan concentration in different whole cell nitration systems supported by M9 medium.

150

Figure 4-4. Nitrated tryptophan concentration in the whole cell nitration system supported by different types of medium.

151

Figure 4-5. Nitrated tryptophan concentration in the whole cell nitration system fermented at different temperatures.

152

Figure 4-6. Nitrated tryptophan analogues production by the whole cell nitration system.

153

Table 4-1. Primers for construction of whole cell transformation plasmids. Name Sequence (5’→3’) Function TB14FN ATACCATGGTGACCGTCCCCTCGCCG TB14 cloning TB14RH ATCAAGCTTCCCAGCCCACACGTCTTTTGC TB14 cloning GDHFB CAGGATCC GATGTATAAAGATCTGGAAGGTAAAGTGGTG GDH cloning GDHRH CAAAGCTTTTAGCCACGACCTGCCTGAAAG GDH cloning BsNOSFN ACTCATATGATGGAAGAAAAAGAAATC BsNOS cloning BsNOSRH ACTAAGCTT CTATTCATACGGTTTGTC BsNOS cloning SsTxtDFN CTACATATGGTGACTTTCGAAGTCGC SsTxtD cloning SsTxtDRH CTCAAGCTTCTGATGAGGGTAAAAGTTG SsTxtD cloning StTxtDFN ACTCATATGGTGACTTTCGAAGTCGCCCTG StTxtD cloning StTxtDRH ACTAAGCTTCTGATGAGGGTAAAAGTTGGGG StTxtD cloning

154

CHAPTER 5 GENERAL CONCLUSION

The use of biocatalysts for regio- and stereo-selective biotransformation reflects an efficient, cost-effective, and environmentally benign alternative to classical organic chemistry. Aromatic nitration is an immensely important industrial process to produce chemicals for a variety of applications. However, this classic acid-based electrophilic reaction often suffers from multiple challenges including poor selectivity, unwanted multiple impurities, and the generation of a large amount of acidic waste. Recently, a unique cytochrome P450 enzyme, TxtE, was discovered to nitrate the indole of L-Trp within the biosynthetic pathway of phytotoxin thaxtomin.

The objective of this research is to develop biocatalytic processes for the synthesis of nitro compounds using cytochrome P450 TxtE as the parent enzyme. In

Chapter 2, three artificial self-sufficient enzymes were created by fusing TxtE with the reductase domains of CYP102A1 (P450BM3) from Bacillus megatorium or P450RhF from Rhodococcus sp. Purified recombinant fusion proteins were properly folded and exhibited various levels of nitration activities. TxtE-BM3R showed the best activity and was slightly more active than wild type TxtE supplemented with spinach ferredoxin and ferredoxin reductase. This enzyme and TxtE were then biochemically characterized in terms of thermostability, pH stability, and pH dependence. We further created fifteen new chimeras by rationally replacing a catalytically important loop and varying the lengths of linkers connecting TxtE and BM3R. Among them, TB14 showed most significant improvement in nitration activity, coupling efficiency and total turnover numbers. Although a few enzymes have been explored in nitration reactions, but only minor progress was made due to the lack of regio-selectivity and stability. To the best of

155

our knowledge, TB14 is the most active aromatic nitration biocatalyst ever developed.

More importantly, TB14 catalyzed nitration reactions circumvented the requirement of heterologous redox partners, thus not only significantly reduced the cost but also eased the reaction preparation.

In the following research, substrate promiscuities of TxtE and TB14 were further assessed with a chemical library. This biocatalyst showed a considerable substrate tolerance, particularly towards small modifications on the indole ring. The remarkable regio-flexibility in nitrating 4 position pre-occupied L-Trp analogues were also identified.

Interestingly, styrene and its analogues could also be converted by TxtE and its variants. Based on 1H-NMR and HRMS analysis, we concluded that the reaction was not simply nitration, but the modification occurred on the aliphatic chain of styrene and its analogues. The product structure remained obscure due to the extremely low conversion rate. For future work, 13C NMR and 2D-NMR are necessary to determine the product structure. Mechanism of this unusual reaction can be quite interesting and may even expand the fundamental understanding of enzyme reactions. These studies will further guide the engineering efforts to improve the efficiency of this unique reaction and expand the application of TB14 as a versatile biocatalyst.

In the last chapter, we develop a whole cell biotransformation system relying on co-expression of TB14, BsNOS, and GDH genes in E. coli. The whole cell catalysts circumvent the cell lysis and enzyme purification steps and hence significantly cut the cost and labor. More importantly, the requirement of chemical NO donors was also eased compared to the previous established in vitro enzymatic reactions. The fermentation conditions were subsequently optimized, and good nitration product yield

156

was achieved (up to 700 µM for L-Trp and up to 250 µM for L-Trp analogues). Metabolic engineering of the host strain may further improve the efficiency of the whole cell nitration system and is currently underway in our lab.

To our knowledge, the studies in this dissertation represent the first practice in developing biological nitration approaches and lay a solid basis to the use of TxtE- based biocatalysts for the production of valuable nitro compounds. Further engineering efforts are necessary in order to promote the industrial application.

157

LIST OF REFERENCES

1. Bornscheuer, U.T. et al. Engineering the third wave of biocatalysis. Nature 485, 185-194 (2012).

2. Hughes, G. & Lewis, J.C. Introduction: Biocatalysis in industry. Chem Rev 118, 1-3 (2017).

3. Matsumae, H., Furui, M. & Shibatani, T. Lipase-catalyzed asymmetric hydrolysis of 3-phenylglycidic acid ester, the key intermediate in the synthesis of diltiazem hydrochloride. J Ferment Bioeng 75, 93-98 (1993).

4. Griengl, H., Schwab, H. & Fechter, M. The synthesis of chiral cyanohydrins by oxynitrilases. Trends Biotechnol 18, 252-256 (2000).

5. Liang, J. et al. Development of a biocatalytic process as an alternative to the (-)- DIP-Cl-mediated asymmetric reduction of a key intermediate of montelukast. Org Process Res Dev 14, 193-198 (2010).

6. Desai, A.A. Sitagliptin manufacture: a compelling tale of green chemistry, process intensification, and industrial asymmetric catalysis. Angew Chem Int Edit 50, 1974-1976 (2011).

7. Savile, C.K. et al. Biocatalytic asymmetric synthesis of chiral amines from ketones applied to sitagliptin manufacture. Science 329, 305-309 (2010).

8. Zhang, M.M., Su, X., Ang, E.L. & Zhao, H. Recent advances in biocatalyst development in the pharmaceutical industry. Pharma Bioprocess 1, 179-196 (2013).

9. Arnold, F.H., Wintrode, P.L., Miyazaki, K. & Gershenson, A. How enzymes adapt: lessons from directed evolution. Trends Biochem Sci 26, 100-106 (2001).

10. Zhang, J.H., Dawes, G. & Stemmer, W.P.C. Directed evolution of a fucosidase from a galactosidase by DNA shuffling and screening. P Natl Acad Sci USA 94, 4504-4509 (1997).

11. Wong, T.S., Tee, K.L., Hauer, B. & Schwaneberg, U. Sequence saturation mutagenesis (SeSaM): a novel method for directed evolution. Nucleic Acids Res 32 (2004).

12. Ostermeier, M., Shim, J.H. & Benkovic, S.J. A combinatorial approach to hybrid enzymes independent of DNA homology. Nat Biotechnol 17, 1205-1209 (1999).

13. Bittker, J.A., Le, B.V., Liu, J.M. & Liu, D.R. Directed evolution of protein enzymes using nonhomologous random recombination. P Natl Acad Sci USA 101, 7011- 7016 (2004).

158

14. Tsuji, T., Onimaru, M. & Yanagawa, H. Random multi-recombinant PCR for the construction of combinatorial protein libraries. Nucleic Acids Res 29, e97 (2001).

15. Esvelt, K.M., Carlson, J.C. & Liu, D.R. A system for the continuous directed evolution of biomolecules. Nature 472, 499-503 (2011).

16. Romanini, D.W., Peralta-Yahya, P., Mondol, V. & Cornish, V.W. A heritable recombination system for synthetic darwinian evolution in yeast. Acs Synth Biol 1, 602-609 (2012).

17. Becker, S. et al. Single-cell high-throughput screening to identify enantioselective hydrolytic enzymes. Angew Chem Int Edit 47, 5085-5088 (2008).

18. Agresti, J.J. et al. Ultrahigh-throughput screening in drop-based microfluidics for directed evolution. P Natl Acad Sci USA 107, 4004-4009 (2010).

19. Fernandez-Alvaro, E. et al. A combination of in vivo selection and cell sorting for the identification of enantioselective biocatalysts. Angew Chem Int Edit 50, 8584- 8587 (2011).

20. Reitman, Z.J. et al. Enzyme redesign guided by cancer-derived IDH1 mutations. Nat Chem Biol 8, 887-889 (2012).

21. Chen, M., Drury, J.E., Christianson, D.W. & Penning, T.M. Conversion of human steroid 5beta-reductase (AKR1D1) into 3beta-hydroxysteroid dehydrogenase by single point mutation E120H: example of perfect enzyme engineering. J Biol Chem 287, 16609-16622 (2012).

22. Payongsri, P. et al. Rational substrate and enzyme engineering of transketolase for aromatics. Org Biomol Chem 10, 9021-9029 (2012).

23. Meyer, M.M., Hochrein, L. & Arnold, F.H. Structure-guided SCHEMA recombination of distantly related beta-lactamases. Protein Eng Des Sel 19, 563- 570 (2006).

24. Heinzelman, P. et al. SCHEMA recombination of a fungal cellulase uncovers a single mutation that contributes markedly to stability. J Biol Chem 284, 26229- 26233 (2009).

25. Heinzelman, P. et al. Efficient screening of fungal cellobiohydrolase class I enzymes for thermostabilizing sequence blocks by SCHEMA structure-guided recombination. Protein Eng Des Sel 23, 871-880 (2010).

26. Fox, R.J. et al. Improving catalytic function by ProSAR-driven enzyme evolution. Nat Biotechnol 25, 338-344 (2007).

159

27. Chockalingam, K., Chen, Z.L., Katzenellenbogen, J.A. & Zhao, H.M. Directed evolution of specific receptor-ligand pairs for use in the creation of gene switches. P Natl Acad Sci USA 102, 5691-5696 (2005).

28. Reetz, M.T., D Carballeira, J. & Vogel, A. Iterative saturation mutagenesis on the basis of B factors as a strategy for increasing protein thermostability. Angew Chem Int Edit 45, 7745-7751 (2006).

29. Hoffmann, G., Bonsch, K., Greiner-Stoffele, T. & Ballschmiter, M. Changing the substrate specificity of P450cam towards diphenylmethane by semi-rational enzyme engineering. Protein Eng Des Sel 24, 439-446 (2011).

30. Whitehead, T.A. et al. Optimization of affinity, specificity and function of designed influenza inhibitors using deep sequencing. Nat Biotechnol 30, 543-548 (2012).

31. DeGrado, W.F., Summa, C.M., Pavone, V., Nastri, F. & Lombardi, A. De novo design and structural characterization of proteins and metalloproteins. Annu Rev Biochem 68, 779-819 (1999).

32. Richter, F., Leaver-Fay, A., Khare, S.D., Bjelic, S. & Baker, D. De Novo enzyme design using Rosetta3. Plos One 6 (2011).

33. Jiang, L. et al. De novo computational design of retro-aldol enzymes. Science 319, 1387-1391 (2008).

34. Siegel, J.B. et al. Computational design of an enzyme catalyst for a stereoselective bimolecular Diels-Alder reaction. Science 329, 309-313 (2010).

35. Wachtmeister, J. & Rother, D. Recent advances in whole cell biocatalysis techniques bridging from investigative to industrial scale. Curr Opin Biotech 42, 169-177 (2016).

36. Ro, D.K. et al. Production of the antimalarial drug precursor artemisinic acid in engineered yeast. Nature 440, 940-943 (2006).

37. Ro, D.K. et al. Induction of multiple pleiotropic drug resistance genes in yeast engineered to produce an increased level of anti-malarial drug precursor, artemisinic acid. BMC Biotechnol 8 (2008).

38. Westfall, P.J. et al. Production of amorphadiene in yeast, and its conversion to dihydroartemisinic acid, precursor to the antimalarial agent artemisinin. P Natl Acad Sci USA 109, 111-118 (2012).

39. Levesque, F. & Seeberger, P.H. Continuous-flow synthesis of the anti-malaria drug artemisinin. Angew Chem Int Edit 51, 1706-1709 (2012).

160

40. Dietrich, J.A. et al. A novel semi-biosynthetic route for artemisinin production using engineered substrate-promiscuous P450BM3. ACS Chem Biol 4, 261-267 (2009).

41. Zhang, K., Shafer, B.M., Demars, M.D., Stern, H.A. & Fasan, R. Controlled oxidation of remote sp3 C–H bonds in artemisinin via P450 catalysts with fine- tuned regio-and stereoselectivity. J Am Chem Soc 134, 18695-18704 (2012).

42. Ajikumar, P.K. et al. Isoprenoid pathway optimization for taxol precursor overproduction in Escherichia coli. Science 330, 70-74 (2010).

43. Yue, W. et al. Medicinal plant cell suspension cultures: pharmaceutical applications and high-yielding strategies for the desired secondary metabolites. Crit Rev Biotechnol 36, 215-232 (2016).

44. Munro, A.W., Girvan, H.M., Mason, A.E., Dunford, A.J. & McLean, K.J. What makes a P450 tick? Trends Biochem Sci 38, 140-150 (2013).

45. Lamb, D.C. et al. The first virally encoded cytochrome P450. J Virol 83, 8266- 8269 (2009).

46. Nelson, D.R. A world of cytochrome P450s. Philos T R Soc B 368 (2013).

47. Guengerich, F.P. & Munro, A.W. Unusual cytochrome P450 enzymes and reactions. J Biol Chem 288, 17065-17073 (2013).

48. Podust, L.M. & Sherman, D.H. Diversity of P450 enzymes in the biosynthesis of natural products. Nat Prod Rep 29, 1251-1266 (2012).

49. Fasan, R. Tuning P450 enzymes as oxidation catalysts. ACS Catal 2, 647-666 (2012).

50. Guengerich, F.P. Fifty Years after the discovery of cytochrome P450: what have we learned and what challenges are left regarding oxidative metabolism? Drug Metab Rev 40, 1-1 (2008).

51. Krest, C.M. et al. Reactive intermediates in cytochrome P450 catalysis. J Biol Chem 288, 17074-17081 (2013).

52. Rittle, J. & Green, M.T. Cytochrome P450 compound I: capture, characterization, and C-H bond activation kinetics. Science 330, 933-937 (2010).

53. Guengerich, F.P. Mechanisms of cytochrome p450 substrate oxidation: minireview. J Biochem Mol Toxic 21, 163-168 (2007).

161

54. Denisov, I.G., Makris, T.M., Sligar, S.G. & Schlichting, I. Structure and chemistry of cytochrome P450. Chem Rev 105, 2253-2277 (2005).

55. Hannemann, F., Bichet, A., Ewen, K.M. & Bernhardt, R. Cytochrome P450 systems - biological variations of electron transport chains. Bba-Gen Subjects 1770, 330-344 (2007).

56. Munro, A.W., Daff, S., Coggins, J.R., Lindsay, J.G. & Chapman, S.K. Probing electron transfer in flavocytochrome P-450 BM3 and its component domains. Eur J Biochem 239, 403-409 (1996).

57. Whitehouse, C.J., Bell, S.G. & Wong, L.L. P450(BM3) (CYP102A1): connecting the dots. Chem Soc Rev 41, 1218-1260 (2012).

58. Roberts, G.A., Grogan, G., Greter, A., Flitsch, S.L. & Turner, N.J. Identification of a new class of cytochrome P450 from a Rhodococcus sp. J Bacteriol 184, 3898- 3908 (2002).

59. Munro, A.W., Girvan, H.M. & McLean, K.J. Cytochrome P450 - redox partner fusion enzymes. Bba-Gen Subjects 1770, 345-359 (2007).

60. McLean, K.J., Girvan, H.M. & Munro, A.W. Cytochrome P450/redox partner fusion enzymes: biotechnological and toxicological prospects. Expert Opin Drug Met 3, 847-863 (2007).

61. Sadeghi, S.J. & Gilardi, G. Chimeric P450 enzymes: activity of artificial redox fusions driven by different reductases for biotechnological applications. Biotechnol Appl Bioc 60, 102-110 (2013).

62. Wei, Y., Ang, E.L. & Zhao, H. Recent developments in the application of P450 based biocatalysts. Curr Opin Chem Biol 43, 1-7 (2017).

63. McLean, K.J. et al. Single-step fermentative production of the cholesterol- lowering drug pravastatin via reprogramming of Penicillium chrysogenum. P Natl Acad Sci USA 112, 2847-2852 (2015).

64. Sakaki, T. Practical Application of Cytochrome P450. Biol Pharm Bull 35, 844- 849 (2012).

65. Brandenberg, O.F., Fasan, R. & Arnold, F.H. Exploiting and engineering hemoproteins for abiological carbene and nitrene transfer reactions. Curr Opin Biotech 47, 102-111 (2017).

66. Coelho, P.S., Brustad, E.M., Kannan, A. & Arnold, F.H. Olefin cyclopropanation via carbene transfer catalyzed by engineered cytochrome P450 enzymes. Science 339, 307-310 (2013).

162

67. Coelho, P.S. et al. A serine-substituted P450 catalyzes highly efficient carbene transfer to olefins in vivo. Nat Chem Biol 9, 485-487 (2013).

68. Wang, Z.J. et al. Improved cyclopropanation activity of histidine-ligated cytochrome P450 enables the enantioselective formal synthesis of levomilnacipran. Angew Chem Int Edit 53, 6810-6813 (2014).

69. Gober, J.G. et al. Mutating a highly conserved residue in diverse cytochrome P450s facilitates diastereoselective olefin cyclopropanation. Chembiochem 17, 2099-2099 (2016).

70. Gober, J.G. et al. P450-mediated non-natural cyclopropanation of dehydroalanine-containing thiopeptides. ACS Chem Biol 12, 1726-1731 (2017).

71. Key, H.M. et al. Beyond iron: iridium-containing P450 enzymes for selective cyclopropanations of structurally diverse alkenes. ACS Central Sci 3, 302-308 (2017).

72. Sreenilayam, G., Moore, E.J., Steck, V. & Fasan, R. Metal substitution modulates the reactivity and extends the reaction scope of myoglobin carbene transfer catalysts. Adv Synth Catal 359, 2076-2089 (2017).

73. Dydio, P. et al. An artificial metalloenzyme with the kinetics of native enzymes. Science 354, 102-106 (2016).

74. Key, H.M., Dydio, P., Clark, D.S. & Hartwig, J.F. Abiological catalysis by artificial haem proteins containing noble metals in place of iron. Nature 534, 534-537 (2016).

75. Singh, R., Kolev, J.N., Sutera, P.A. & Fasan, R. Enzymatic C(sp(3))-H amination: P450-catalyzed conversion of carbonazidates into oxazolidinones. ACS Catal 5, 1685-1691 (2015).

76. Singh, R., Bordeaux, M. & Fasan, R. P450-catalyzed intramolecular sp(3) C-H amination with arylsulfonyl azide substrates. ACS Catal 4, 546-552 (2014).

77. McIntosh, J.A. et al. Enantioselective intramolecular C-H amination catalyzed by engineered cytochrome P450 enzymes in vitro and in vivo. Angew Chem Int Edit 52, 9309-9312 (2013).

78. Hyster, T.K., Farwell, C.C., Buller, A.R., McIntosh, J.A. & Arnold, F.H. Enzyme- controlled nitrogen-atom transfer enables regiodivergent C-H amination. J Am Chem Soc 136, 15505-15508 (2014).

163

79. Bordeaux, M., Singh, R. & Fasan, R. Intramolecular C(sp(3))-H amination of arylsulfonyl azides with engineered and artificial myoglobin-based catalysts. Bioorgan Med Chem 22, 5697-5704 (2014).

80. Prier, C.K., Hyster, T.K., Farwell, C.C., Huang, A. & Arnold, F.H. Asymmetric enzymatic synthesis of allylic amines: a sigmatropic rearrangement strategy. Angew Chem Int Edit 55, 4711-4715 (2016).

81. Prier, C.K., Zhang, R.J.K., Buller, A.R., Brinkmann-Chen, S. & Arnold, F.H. Enantioselective, intermolecular benzylic C-H amination catalysed by an engineered iron-haem enzyme. Nat Chem 9, 629-634 (2017).

82. Farwell, C.C., Zhang, R.K., McIntosh, J.A., Hyster, T.K. & Arnold, F.H. Enantioselective enzyme-catalyzed aziridination enabled by active-site evolution of a cytochrome P450. ACS Central Sci 1, 89-93 (2015).

83. Kan, S.B.J., Lewis, R.D., Chen, K. & Arnold, F.H. Directed evolution of cytochrome c for carbon-silicon bond formation: bringing silicon to life. Science 354, 1048-1051 (2016).

84. Kan, S.B.J., Huang, X.Y., Gumulya, Y., Chen, K. & Arnold, F.H. Genetically programmed chiral organoborane synthesis. Nature 552, 132-136 (2017).

85. Kulkarni, M. & Chaudhari, A. Microbial remediation of nitro-aromatic compounds: an overview. J Environ Manage 85, 496-512 (2007).

86. Ju, K.S. & Parales, R.E. Nitroaromatic compounds, from Synthesis to biodegradation. Microbiol Mol Biol R 74, 250-272 (2010).

87. Williams, E.M. et al. Nitroreductase gene-directed enzyme prodrug therapy: insights and advances toward clinical utility. Biochem J 471, 131-153 (2015).

88. Helsby, N.A., Ferry, D.M., Patterson, A.V., Pullen, S.M. & Wilson, W.R. 2-Amino metabolites are key mediators of CB 1954 and SN 23862 bystander effects in nitroreductase GDEPT. Brit J Cancer 90, 1084-1092 (2004).

89. Bridgewater, J.A., Knox, R.J., Pitts, J.D., Collins, M.K. & Springer, C.J. The bystander effect of the nitroreductase CB 1954 enzyme prodrug system is due to a cell-permeable metabolite. Hum Gene Ther 8, 709-717 (1997).

90. Hunt, M.A. et al. Characterisation of enzyme prodrug gene therapy combinations in coated spheroids and vascular networks in vitro. J Gene Med 14, 62-74 (2012).

164

91. Singleton, R.S. et al. DNA cross-links in human tumor cells exposed to the prodrug PR-104A: relationships to hypoxia, bioreductive metabolism, and cytotoxicity. Cancer Res 69, 3884-3891 (2009).

92. Gu, Y.C. et al. Roles of DNA repair and reductase activity in the cytotoxicity of the hypoxia-activated dinitrobenzamide mustard PR-104A. Mol Cancer Ther 8, 1714-1723 (2009).

93. Yan, G. & Yang, M. Recent advances in the synthesis of aromatic nitro compounds. Org Biomol Chem 11, 2554-2566 (2013).

94. Zolfigol, M.A. et al. Design of ionic liquid 3-methyl-1-sulfonic acid imidazolium nitrate as reagent for the nitration of aromatic compounds by in situ generation of NO2 in acidic media. J Org Chem 77, 3640-3645 (2012).

95. Aridoss, G. & Laali, K.K. Ethylammonium nitrate (EAN)/Tf2O and EAN/TFAA: ionic liquid based systems for aromatic nitration. J Org Chem 76, 8088-8094 (2011).

96. Lancaster, N.L. & Llopis-Mestre, V. Aromatic nitrations in ionic liquids: the importance of cation choice. Chem Commun, 2812-2813 (2003).

97. Laali, K.K. & Gettwert, V.J. Electrophilic nitration of aromatics in ionic liquid solvents. J Org Chem 66, 35-40 (2001).

98. Tani, K., Lukin, K. & Eaton, P.E. Nitration of organolithiums and Grignards with dinitrogen tetroxide: success at melting interfaces. J Am Chem Soc 119, 1476- 1477 (1997).

99. Das, J.P., Sinha, P. & Roy, S. A nitro-Hunsdiecker reaction: From unsaturated carboxylic acids to nitrostyrenes and nitroarenes. Org Lett 4, 3055-3058 (2002).

100. Fors, B.P. & Buchwald, S.L. Pd-catalyzed conversion of aryl chlorides, triflates, and nonaflates to nitroaromatics. J Am Chem Soc 131, 12898-12899 (2009).

101. Rozen, S. & Carmeli, M. From azides to nitro compounds in a few seconds using HOF-CH3CN. J Am Chem Soc 125, 8118-8119 (2003).

102. Zhang, L. et al. Copper-mediated chelation-assisted ortho nitration of (hetero)arenes. Org Lett 13, 6536-6539 (2011).

103. Zhang, W., Lou, S.J., Liu, Y.K. & Xu, Z.Y. Palladium-catalyzed chelation-assisted aromatic C-H nitration: regiospecific synthesis of nitroarenes free from the effect of the orientation rules. J Org Chem 78, 5932-5948 (2013).

165

104. Andrews, M.A. et al. Nitration of alkenes by palladium nitro complexes. Organometallics 3, 1479-1484 (1984).

105. Kasahara, A., Katsuyas.Ui & Tanaka, K. Nitration and halogenation of palladium acetylacetonate. B Chem Soc Jpn 39, 2227-2229 (1966).

106. Nowrouzi, N., Mehranpour, A.M., Bashiri, E. & Shayan, Z. Aromatic nitration under neutral conditions using N-bromosuccinimide/silver(I) nitrate. Tetrahedron Lett 53, 4841-4842 (2012).

107. Koley, D., Colon, O.C. & Savinov, S.N. Chemoselective nitration of phenols with tert-butyl nitrite in solution and on solid support. Org Lett 11, 4172-4175 (2009).

108. Vekariya, R.H. & Patel, H.D. Selective nitration of phenolic compounds by green synthetic approaches. Synthetic Commun 44, 2313-2335 (2014).

109. Umbarkar, S.B. et al. Vapor phase nitration of benzene using mesoporous MoO3/SiO2 solid acid catalyst. Green Chem 8, 488-493 (2006).

110. Ma, X.M. et al. An efficient and eco-friendly MoO3-SiO2 solid acid catalyst for electrophilic aromatic nitration with N2O5. Catal Lett 141, 1814-1820 (2011).

111. Khder, A.S. & Ahmed, A.I. Selective nitration of phenol over nanosized tungsten oxide supported on sulfated SnO2 as a solid acid catalyst. Appl Catal a-Gen 354, 153-160 (2009).

112. Roiban, G.D. & Reetz, M.T. Expanding the toolbox of organic chemists: directed evolution of P450 monooxygenases as catalysts in regio- and stereoselective oxidative hydroxylation. Chem Commun 51, 2208-2224 (2015).

113. Clouthier, C.M. & Pelletier, J.N. Expanding the organic toolbox: a guide to integrating biocatalysis in synthesis. Chem Soc Rev 41, 1585-1605 (2012).

114. Kersten, R.D. & Dorrestein, P.C. Metalloenzymes natural product nitrosation. Nat Chem Biol 6, 636-637 (2010).

115. Johnson, H.D. & Thorson, J.S. Characterization of CalE10, the N-oxidase involved in calicheamicin hydroxyaminosugar formation. J Am Chem Soc 130, 17662-17663 (2008).

116. Carter, G.T. et al. Direct biochemical nitration in the biosynthesis of dioxapyrrolomycin - a unique mechanism for the introduction of nitro-groups in microbial products. J Chem Soc Chem Comm, 1271-1273 (1989).

166

117. Lee, J.K., Simurdiak, M. & Zhao, H.M. Reconstitution and characterization of aminopyrrolnitrin oxygenase, a Rieske N-oxygenase that catalyzes unusual arylamine oxidation. J Biol Chem 280, 36719-36728 (2005).

118. Winkler, R. et al. A binuclear manganese cluster that catalyzes radical-mediated N-oxygenation. Angew Chem Int Edit 46, 8605-8608 (2007).

119. Hu, Y., Al-Mestarihi, A., Grimes, C.L., Kahne, D. & Bachmann, B.O. A unifying nitrososynthase involved in nitrosugar biosynthesis. J Am Chem Soc 130, 15756- 15757 (2008).

120. Barry, S.M. et al. Cytochrome P450-catalyzed L-tryptophan nitration in thaxtomin phytotoxin biosynthesis. Nat Chem Biol 8, 814-816 (2012).

121. Loria, R. et al. Thaxtomin biosynthesis: the path to plant pathogenicity in the genus Streptomyces. Anton Leeuw Int J G 94, 3-10 (2008).

122. Johnson, E.G. et al. 4-Nitrotryptophan is a substrate for the non-ribosomal peptide synthetase TxtB in the thaxtomin A biosynthetic pathway. Mol Microbiol 73, 409-418 (2009).

123. Krasnoff, S.B., Lobkovsky, E.B., Wach, M.J., Loria, R. & Gibson, D.M. Chemistry and phytotoxicity of thaxtomin A alkyl ethers. J Agr Food Chem 53, 9446-9451 (2005).

124. Tomita, H., Katsuyama, Y., Minami, H. & Ohnishi, Y. Identification and characterization of a bacterial cytochrome P450 monooxygenase catalyzing the 3-nitration of tyrosine in rufomycin biosynthesis. J Biol Chem 292, 15859-15869 (2017).

125. Matthews, M.L. et al. Direct nitration and azidation of aliphatic carbons by an iron-dependent halogenase. Nat Chem Biol 10, 209-215 (2014).

126. Dodani, S.C. et al. Structural, functional, and spectroscopic characterization of the substrate scope of the novel nitrating cytochrome P450 TxtE. Chembiochem 15, 2259-2267 (2014).

127. Dodani, S.C. et al. Discovery of a regioselectivity switch in nitrating P450s guided by molecular dynamics simulations and Markov models. Nat Chem 8, 419-425 (2016).

128. Padda, R.S., Wang, C., Hughes, J.B., Kutty, R. & Bennett, G.N. Mutagenicity of nitroaromatic degradation compounds. Environ Toxicol Chem 22, 2293-2297 (2003).

167

129. Martino, P.D., Fursy, R., Bret, L., Sundararaju, B. & Phillips, R.S. Indole can act as an extracellular signal to regulate biofilm formation of Escherichia coli and other indole-producing bacteria. Can J Microbiol 49, 443-449 (2003).

130. Aneja, R. et al. Development of a novel nitro-derivative of noscapine for the potential treatment of drug-resistant ovarian cancer and T-cell lymphoma. Mol Pharmacol 69, 1801-1809 (2006).

131. Zhou, L., Stewart, G., Rideau, E., Westwood, N.J. & Smith, T.K. A class of 5- nitro-2-furancarboxylamides with potent trypanocidal activity against Trypanosoma brucei in vitro. J Med Chem 56, 796-806 (2013).

132. Parry, R., Nishino, S. & Spain, J. Naturally-occurring nitro compounds. Nat Prod Rep 28, 152-167 (2011).

133. Winkler, R. & Hertweck, C. Biosynthesis of nitro compounds. Chembiochem 8, 973-977 (2007).

134. Winkler, R., Richter, M.E., Knupfer, U., Merten, D. & Hertweck, C. Regio- and chemoselective enzymatic N-oxygenation in vivo, in vitro, and in flow. Angew Chem Int Ed Engl 45, 8016-8018 (2006).

135. Urlacher, V.B. & Girhard, M. Cytochrome P450 monooxygenases: an update on perspectives for synthetic application. Trends Biotechnol 30, 26-36 (2012).

136. De Mot, R. & Parret, A.H.A. A novel class of self-sufficient cytochrome P450 monooxygenases in prokaryotes. Trends Microbiol 10, 502-508 (2002).

137. Narhi, L.O. & Fulco, A.J. Characterization of a catalytically self-sufficient 119,000-dalton cytochrome P-450 monooxygenase induced by barbiturates in Bacillus megaterium. J Biol Chem 261, 7160-7169 (1986).

138. Mclntosh, J.A., Farwell, C.C. & Arnold, F.H. Expanding P450 catalytic reaction space through evolution and engineering. Curr Opin Chem Biol 19, 126-134 (2014).

139. Gillam, E.M.J. & Hayes, M.A. The evolution of cytochrome P450 enzymes as biocatalysts in drug discovery and development. Curr Top Med Chem 13, 2254- 2280 (2013).

140. Caswell, J.M., O'Neill, M., Taylor, S.J.C. & Moody, T.S. Engineering and application of P450 monooxygenases in pharmaceutical and metabolite synthesis. Curr Opin Chem Biol 17, 271-275 (2013).

168

141. Li, S.Y., Podust, L.M. & Sherman, D.H. Engineering and analysis of a self- sufficient biosynthetic cytochrome P450 PikC fused to the RhFRED reductase domain. J Am Chem Soc 129, 12940-12941 (2007).

142. Robin, A. et al. Engineering and improvement of the efficiency of a chimeric [P450cam-RhFRed reductase domain] enzyme. Chem Commun, 2478-2480 (2009).

143. Dodhia, V.R., Fantuzzi, A. & Gilardi, G. Engineering human cytochrome P450 enzymes into catalytically self-sufficient chimeras using molecular Lego. J Biol Inorg Chem 11, 903-916 (2006).

144. Nodate, M., Kubota, M. & Misawa, N. Functional expression system for cytochrome P450 genes using the reductase domain of self-sufficient P450RhF from Rhodococcus sp NCIMB 9784. Appl Microbiol Biot 71, 455-462 (2006).

145. Sabbadin, F. et al. LICRED: a versatile drop-in vector for rapid generation of redox-self-sufficient cytochrome P450s. Chembiochem 11, 987-994 (2010).

146. Carlson, J.C. et al. Tirandamycin biosynthesis is mediated by co-dependent oxidative enzymes. Nat Chem 3, 628-633 (2011).

147. Sibbesen, O., DeVoss, J.J. & deMontellano, P.R.O. Putidaredoxin reductase- putidaredoxin-cytochrome P450(cam) triple fusion protein - construction of a self- sufficient Escherichia coli catalytic system. J Biol Chem 271, 22462-22469 (1996).

148. Gilardi, G. et al. Molecular Lego: design of molecular assemblies of P450 enzymes for nanobiotechnology. Biosens Bioelectron 17, 133-145 (2002).

149. Bordeaux, M., Galarneau, A., Fajula, F. & Drone, J. A Regioselective biocatalyst for alkane activation under mild conditions. Angew Chem Int Edit 50, 2075-2079 (2011).

150. Roberts, G.A. et al. A self-sufficient cytochrome P450 with a primary structural organization that includes a flavin domain and a [2Fe-2S] redox center. J Biol Chem 278, 48914-48920 (2003).

151. Suyama, M. & Ohara, O. DomCut: prediction of inter-domain linker regions in amino acid sequences. Bioinformatics 19, 673-674 (2003).

152. Omura, T. & Sato, R. The carbon monoxide-binding pigment of liver microsomes. ii. solubilization, purification, and properties. J Biol Chem 239, 2379-2385 (1964).

169

153. Ding, Y., Seufert, W.H., Beck, Z.Q. & Sherman, D.H. Analysis of the cryptophycin P450 epoxidase reveals substrate tolerance and cooperativity. J Am Chem Soc 130, 5492-5498 (2008).

154. Zuo, R. et al. An artificial self-sufficient cytochrome P450 directly nitrates fluorinated tryptophan analogs with a different regio-selectivity. Biotechnol J 11, 624-632 (2016).

155. Sevrioukova, I.F., Li, H.Y., Zhang, H., Peterson, J.A. & Poulos, T.L. Structure of a cytochrome P450-redox partner electron-transfer complex. P Natl Acad Sci USA 96, 1863-1868 (1999).

156. Hoffmann, S.M. et al. The impact of linker length on P450 fusion constructs: activity, stability and coupling. ChemCatChem 8, 1591-1597 (2016).

157. Belsare, K.D. et al. P-LinK: a method for generating multicomponent cytochrome P450 fusions with variable linker length. Biotechniques 57, 13-20 (2014).

158. Govindaraj, S. & Poulos, T.L. Role of the linker region connecting the reductase and heme domains in cytochrome P450BM-3. Biochemistry 34, 11221-11226 (1995).

159. Rosic, N.N., Huang, W., Johnston, W.A., DeVoss, J.J. & Gillam, E.M. Extending the diversity of cytochrome P450 enzymes by DNA family shuffling. Gene 395, 40-48 (2007).

160. Lautier, T. et al. Ordered chimerogenesis applied to CYP2B P450 enzymes. Biochim Biophys Acta 1860, 1395-1403 (2016).

161. Loida, P.J. & Sligar, S.G. Molecular recognition in cytochrome-P-450 - mechanism for the control of uncoupling reactions. Biochemistry 32, 11530- 11538 (1993).

162. Noble, M.A. et al. Roles of key active-site residues in flavocytochrome P450 BM3. Biochem J 339, 371-379 (1999).

163. Sakihama, Y. et al. Enzymatic nitration of phytophenolics: evidence for peroxynitrite-independent nitration of plant secondary metabolites. FEBS Lett 553, 377-380 (2003).

164. Casella, L. et al. Formation of reactive nitrogen species at biologic heme centers: a potential mechanism of nitric oxide-dependent toxicity. Environ Health Perspect 110 Suppl 5, 709-711 (2002).

165. Budde, C.L., Beyer, A., Munir, I.Z., Dordick, J.S. & Khmelnitsky, Y.L. Enzymatic nitration of phenols. J Mol Catal B Enzym 15, 55-64 (2001).

170

166. Bell, S.G., Chen, X., Xu, F., Rao, Z. & Wong, L.L. Engineering substrate recognition in catalysis by cytochrome P450cam. Biochem Soc Trans 31, 558- 562 (2003).

167. Higuchi, R., Krummel, B. & Saiki, R.K. A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res 16, 7351-7367 (1988).

168. Huisman, G.W. & Collier, S.J. On the development of new biocatalytic processes for practical pharmaceutical synthesis. Curr Opin Chem Biol 17, 284-292 (2013).

169. Reetz, M.T. Biocatalysis in organic chemistry and biotechnology: past, present, and future. J Am Chem Soc 135, 12480-12496 (2013).

170. Royer, C.A. Probing protein folding and conformational transitions with fluorescence. Chem Rev 106, 1769-1784 (2006).

171. Ilardi, E.A., Vitaku, E. & Njardarson, J.T. Data-mining for sulfur and fluorine: an evaluation of pharmaceuticals to reveal opportunities for drug design and discovery. J Med Chem 57, 2832-2842 (2014).

172. Yu, F. et al. Structural insights into the mechanism for recognizing substrate of the cytochrome P450 enzyme TxtE. Plos One 8 (2013).

173. Bian, F. et al. A comprehensive alanine-scanning mutagenesis study reveals roles for salt bridges in the structure and activity of Pseudomonas aeruginosa elastase. Plos One 10 (2015).

174. Guengerich, F.P., Martin, M.V., Sohl, C.D. & Cheng, Q. Measurement of cytochrome P450 and NADPH-cytochrome P450 reductase. Nat Protoc 4, 1245- 1251 (2009).

175. Sherman, D.H. et al. The structural basis for substrate anchoring, active site selectivity, and product formation by P450 PikC from Streptomyces venezuelae. J Biol Chem 281, 26289-26297 (2006).

176. Xue, Y.Q., Wilson, D., Zhao, L.S., Liu, H.W. & Sherman, D.H. Hydroxylation of macrolactones YC-17 and narbomycin is mediated by the pikC-encoded cytochrome P450 in Streptomyces venezuelae. Chem Biol 5, 661-667 (1998).

177. Boddupalli, S.S., Pramanik, B.C., Slaughter, C.A., Estabrook, R.W. & Peterson, J.A. Fatty-acid monooxygenation by P450BM-3 - product identification and proposed mechanisms for the sequential hydroxylation reactions. Arch Biochem Biophys 292, 20-28 (1992).

171

178. Sawayama, A.M. et al. A panel of cytochrome P450 BM3 variants to produce drug metabolites and diversify lead compounds. Chemistry 15, 11723-11729 (2009).

179. Rudolf, J.D. & Poulter, C.D. Tyrosine O-prenyltransferase SirD catalyzes S-, C-, and N-prenylations on tyrosine and tryptophan derivatives. ACS Chem Biol 8, 2707-2714 (2013).

180. Yu, X., Xie, X. & Li, S.M. Substrate promiscuity of secondary metabolite enzymes: prenylation of hydroxynaphthalenes by fungal indole prenyltransferases. Appl Microbiol Biotechnol 92, 737-748 (2011).

181. Walsh, C.T. Biological matching of chemical reactivity: pairing indole nucleophilicity with electrophilic isoprenoids. ACS Chem Biol 9, 2718-2728 (2014).

182. Truan, G., Komandla, M.R., Falck, J.R. & Peterson, J.A. P450BM-3: absolute configuration of the primary metabolites of palmitic acid. Arch Biochem Biophys 366, 192-198 (1999).

183. Li, H.Y. & Poulos, T.L. The structure of the cytochrome P450BM-3 haem domain complexed with the fatty acid substrate, palmitoleic acid. Nat Struct Biol 4, 140- 146 (1997).

184. Yang, Y., Wong, S.E. & Lightstone, F.C. Understanding a substrate's product regioselectivity in a family of enzymes: a case study of acetaminophen binding in cytochrome P450s. Plos One 9, e87058 (2014).

185. Patterson, S. & Wyllie, S. Nitro drugs for the treatment of trypanosomatid diseases: past, present, and future prospects. Trends Parasitol 30, 289-298 (2014).

186. Ryan, N.J. & Lo, J.H. Delamanid: first global approval. Drugs 74, 1041-1045 (2014).

187. Priotto, G. et al. Nifurtimox-eflornithine combination therapy for second-stage African Trypanosoma brucei gambiense trypanosomiasis: a multicentre, randomised, phase III, non-inferiority trial. Lancet 374, 56-64 (2009).

188. Diacon, A.H. et al. Phase II dose-ranging trial of the early bactericidal activity of PA-824. Antimicrob Agents Chemother 56, 3027-3031 (2012).

189. Ding, H.X. et al. Synthetic approaches to the 2013 new drugs. Bioorgan Med Chem 23, 1895-1922 (2015).

172

190. Flick, A.C. et al. Synthetic approaches to the 2014 new drugs. Bioorgan Med Chem 24, 1937-1980 (2016).

191. Flick, A.C. et al. Synthetic approaches to the new drugs approved during 2015. J Med Chem 60, 6480-6515 (2017).

192. Choi, J.M., Han, S.S. & Kim, H.S. Industrial applications of enzyme biocatalysis: current status and future aspects. Biotechnol Adv 33, 1443-1454 (2015).

193. Yang, G. & Ding, Y.S. Recent advances in biocatalyst discovery, development and applications. Bioorgan Med Chem 22, 5604-5612 (2014).

194. Jullesson, D., David, F., Pfleger, B. & Nielsen, J. Impact of synthetic biology and metabolic engineering on industrial production of fine chemicals. Biotechnol Adv 33, 1395-1402 (2015).

195. Keasling, J. Synthetic biology for synthetic chemistry. New Biotechnol 31, S9-S9 (2014).

196. Holden, J.K. et al. Structural and biological studies on bacterial nitric oxide synthase inhibitors. P Natl Acad Sci USA 110, 18127-18131 (2013).

197. Gusarov, I. et al. Bacterial nitric-oxide synthases operate without a dedicated redox partner. J Biol Chem 283, 13140-13147 (2008).

198. Lampel, K.A., Uratani, B., Chaudhry, G.R., Ramaley, R.F. & Rudikoff, S. Characterization of the developmentally regulated Bacillus subtilis glucose- dehydrogenase gene. J Bacteriol 166, 238-243 (1986).

199. Newton, W.A. & Snell, E.E. Catalytic properties of tryptophanase, a multifunctional pyridoxal phosphate enzyme. Proc Natl Acad Sci U S A 51, 382- 389 (1964).

200. Zuo, R. et al. Engineered P450 biocatalysts show improved activity and regio- promiscuity in aromatic nitration. Sci Rep 7 (2017).

201. Tufvesson, P., Lima-Ramos, J., Nordblad, M. & Woodley, J.M. Guidelines and cost analysis for catalyst production in biocatalytic processes. Org Process Res Dev 15, 266-274 (2011).

173

BIOGRAPHICAL SKETCH

Ran Zuo was born in Tai’an City, China. He received his Bachelor of Science in

Biotechnology at the MInzu University in 2009. On the same year, he was accepted by the M.S. program of Qingdao Institute of Bioenergy and Bioprocess Technology,

Chinese Academy of Sciences, under the supervision of Dr. Gongke Zhou. In 2012, he obtained his Master of Science in Biochemistry and Molecular Biology. In 2013, he joined the PhD program in the Department of Medicinal Chemistry at University of

Florida, under the supervision of Dr. Yousong Ding. He received his Doctor of

Philosophy in Pharmaceutical Sciences from the University of Florida in the spring of

2018.

174