P450 Enzymes in Biocatalysis

P450 enzymes in biocatalysis

Exploration of chemical auxiliaries, macromolecular crowding, bioconjugation and oriented-immobilization

Amélie Ménard

A thesis submitted to McGill University in partial fulfillment of the requirements of the degree of Doctor of Philosophy

Department of Chemistry McGill University Montreal, Quebec, Canada H3A 0B8 Submitted: December 2012

Abstract

Cytochrome P450 enzymes (CYPs or P450s) form a ubiquitous family of heme- dependent monooxygenases known mainly for their role in xenobiotic metabolism and their remarkable ability to regio- and stereoselectively oxidize inactivated C-H bonds, a feat that is difficult to achieve by chemical methods. Unfortunately, our ability to study and exploit these enzymes as in vitro biocatalysts has been limited by their low activity, low stability and poor product predictability. This thesis focuses on the study of human drug metabolizing P450 isoforms, namely CYP2E1, CYP3A4 and CYP2D6 because of their exceptional ability to accept a large variety of substrates.

In Chapter 2, we demonstrate the utility of “type II ligands” as chemical auxiliaries for biocatalysis with human CYP2E1. We show that linking the chemical auxiliary nicotinate to a variety of short hydrocarbon substrates can promote their oxidation with predictable regioselectivity at the secondary aliphatic or alkenyl C-H bond furthest from the auxiliary. The origin of this selectivity was rationalized through docking studies of our auxiliary-substrate compounds with reported X-ray crystals structures of

CYP2E1. These results not only confirm the general utility of the chemical auxiliary approach pioneered by our lab to direct the predictable oxidation of inactivated C-H bonds by P450 enzymes, but also provide a system with complementary regioselectivity.

A short study of the effects of macromolecular crowding on the activity of human

CYP3A4 and CYP2D6 is presented in Chapter 3. We found that certain crowding agents were not detrimental to enzyme activity while others had a negative effect. Moreover, certain conditions (initially tested as controls) that improved enzymatic activity were uncovered.

Abstract

In Chapter 4, the non-covalent oriented-immobilization of CYP3A4 via its C- terminal histidine-tag is described. We show that immobilization on Ni-NTA agarose resin via this strategy has no detrimental effect on enzyme activity or stability. The lyoprotectant properties of Ni-NTA were also investigated.

In Chapter 5, we designed and characterized a mutant of CYP3A4 that retains its enzymatic activity upon modification with a variety of fluorescent maleimide dyes via a single cysteine residue on its surface, namely C64. We also show that the activity of this mutant is preserved upon immobilization onto solid supports via this same cysteine residue. Finally, results of a preliminary feasibility study towards applying this immobilization strategy to eventual single-molecule fluorescence microscopy studies are presented.

iii

Résumé

Les enzymes cytochrome P450 (CYPs ou P450) forment une famille omniprésente de mono-oxygénases possédant un noyau hème au site-actif. Ces enzymes sont surtout connues pour leur rôle dans le métabolisme de produits pharmaceutiques et pour leur capacité remarquable à oxyder les liens C-H non-activés de façon régio- et stéréosélective. Malheureusement, notre capacité d’étudier et d’utiliser ces enzymes comme biocatalyseurs in vitro est limitée par leur faible activité, instabilité et une incapacité des connaissances actuelles à prédire leurs produits.

Dans le chapitre 2, nous démontrons l’utilité des ligands de type II comme auxiliaires chimiques pour la biocatalyse avec CYP2E1 humaine. Nous démontrons que l’auxiliaire chimique nicotinate, lorsque lié à une variété d’hydrocarbures courts, peux promouvoir leur oxydation avec une régiosélectivité prévisible pour le lien C-H secondaire aliphatique ou alcényle le plus éloigné de l’auxiliaire. L’origine de cette sélectivité a été rationalisée à l’aide de «docking» moléculaire de nos composés auxiliaire-substrats à l’intérieure de structures cristallines de CYP2E1 publiées par d’autres chercheurs. L’utilité d’auxiliaires chimiques pour contrôler la régiosélectivité des enzymes P450 avait déjà été démontrée par notre groupe de recherche. Les résultats présentés dans ce chapitre offrent non seulement une confirmation du potentiel de cette stratégie, mais aussi un système complémentaire pour l’oxydation prévisible de liens C-H non-activés par les enzymes P450.

Ces résultats confirment également la généralité de l’approche mis au point dans notre laboratoire qui décrit l’utilisation d’une auxiliaire chimique pour diriger l’oxydation prévisible de liens C-H non-activés par les enzymes P450.

Résumé

Une étude des effets de l’encombrement macromoléculaire sur l’activité enzymatique des CYP3A4 et CYP2D6 humaines est présentée dans le chapitre 3. Nous avons trouvé que leur activité demeure inchangée par la présence de certains agents encombrants alors que d’autres ont un effet négatif. De plus, certaines conditions (testées initialement comme contrôle) qui améliorent l’activité enzymatique ont été découvertes.

Dans le chapitre 4, l’immobilisation orienté non-covalente de CYP3A4 par son

étiquette de type his-tag C-terminale est décrite. Nous démontrons que son immobilisation sur une résine Ni-NTA à base d’agarose via cette stratégie n’a aucun effet négatif sur l’activité ou la stabilité de l’enzyme. Les propriétés lyoprotectrices de cette résine ont aussi été investiguées.

Dans le chapitre 5, nous concevons et caractérisons un mutant actif de CYP3A4 lors de modifications avec une variété de maléimides fluorescentes à l’endroit d’un unique résidu cystéinique à sa surface, soit le C64. Nous démontrons aussi que ce mutant préserve son activité lorsqu’immobilisé sur des supports solides par ce même résidu cystéinique. Finalement, les résultats d’études préliminaires sont présentés qui envisagent l’application de cette stratégie d’immobilisation envers des études éventuelles de spectroscopie de fluorescence à la résolution d’une seule molécule.

Acknowledgments

First, I would like to thank my loving husband for his inexhaustible patience and devoted support particularly during these last five years. I would also like to thank my parents for their encouragement and for providing me with the opportunities, experiences and moral guidance that have made me who I am today. Words cannot express how much

I appreciate all they’ve done for me. A thousand thanks to them and to all my family and friends for their prayers and moral support. A special thanks also to my supervisor Prof.

Karine Auclair without whose expertise, mentorship and encouragement I could not have completed this work.

Thank you to all my past and present colleagues: Jin, Xuxu, Lee, Eric, Kayode,

Emelia, Matthew, Annabelle, Vanja, and especially Kenward for being a best friend, Siqi for her constant flattery and Aaron for his willingness to teach, share and exchange knowledge. I am equally grateful to our collaborator Prof. Cosa and his students Pierre,

Christina, Hsiao-Wei, and Wayne. I am also indebted to the undergrads I’ve had the pleasure to mentor for their helpful contributions: Jessica, Rym, Ronan, Camilo and

Yolanda. Thanks also to Dr. Eric Therrien for his help with docking experiments.

Thank you to all the professors who have sat on my review committee for their valuable insights and suggestions: Prof. Cosa, Prof. Damha and Prof. Sleiman, and to the members of the staff for their patient assistance, especially Chantal Marotte, Dr. Nadim

Saadeh, Norman Trempe, Sameer Al-Abdul-Wahid and Mario Perrone.

Finally, I am grateful for all the financial assistance I have received, namely from

NSERC, the Department of Chemistry at McGill (Tak-Hang (Bill) and Christina Chan

Fellowship in Chemistry) and the Centre in Green Chemistry and Catalysis.

Table of contents

Abstract ...... ii

Résumé ...... iv

Acknowledgments ...... vi

Table of contents ...... vii

List of figures ...... xiv

List of schemes ...... xviii

List of tables ...... xix

Abbreviations ...... xx

Chapter 1 – Introduction ...... 1

1.1 Preface ...... 1

1.2 Cytochrome P450 enzymes ...... 1

1.1.1 Cytochrome P450 catalytic mechanism ...... 2

1.1.2 Structural features of P450 enzymes ...... 5

1.1.3 Human cytochrome P450 enzymes ...... 8

1.3 Protein bioconjugation – methods and applications ...... 10

1.3.1 The challenge ...... 10

1.3.2 The advantage of site-specificity ...... 11

1.3.3 Site-specific protein bioconjugation strategies ...... 12

1.3.4 Bioconjugation applied to cytochrome P450 enzymes ...... 18

1.4 Biocatalysis ...... 23

vii

Table of contents

1.4.1 Xenobiotic-metabolizing P450 enzymes as biocatalysts ...... 24

1.4.1.1 Improving recombinant expression of P450s ...... 26

1.4.1.2 The cofactor complication ...... 27

1.4.1.3 The stability problem ...... 30

1.4.1.4 The promiscuity paradox ...... 33

1.5 Research goals ...... 34

Chapter 2 – Type II binders as chemical auxiliaries for biocatalysis with CYP2E1 36

2.1 Preface ...... 36

2.2 Introduction ...... 38

2.2.1 Types of P450 ligands ...... 38

2.2.2 Applications of type II ligands ...... 39

2.3 Objectives ...... 42

2.4 Results and discussion ...... 42

2.4.1 Selection of an appropriate chemical auxiliary ...... 43

2.4.2 Effects of varying substrate structures ...... 48

2.4.3 Molecular modeling studies ...... 51

2.4.4 Necessity of using the chemical auxiliary ...... 55

2.4.5 Nicotinate as a chemical auxiliary ...... 56

2.5 Conclusion ...... 57

Chapter 3 – Effects of macromolecular crowding on the enzymatic activity of CYP3A4 and CYP2D6 ...... 59

3.1 Preface ...... 59

3.2 Introduction ...... 60

viii

Table of contents

3.3 Objectives ...... 66

3.4 Activity assays ...... 68

3.5 Results ...... 69

3.5.1 Effects of macromolecular crowding agents on CYP3A4 activity ...... 69

3.5.2 Effects of macromolecular crowding agents on CYP2D6 activity ...... 71

3.6 Discussion ...... 73

3.6.1 Differential effect of cofactors: CPR and NADPH vs. CHP ...... 73

3.6.2 P450 activity under crowded conditions ...... 75

3.6.3 Effects of the differing crowding agent properties ...... 78

3.6.4 The effect of adding small sugars on P450 activity ...... 80

3.7 Conclusion ...... 81

Chapter 4 – Non-covalent immobilization of CYP3A4 via C-terminal His-tag ...... 82

4.1 Prelude ...... 82

4.2 Introduction ...... 84

4.2.1 Non-covalent immobilization of CYP3A4...... 84

4.2.2 Immobilization of His-tagged proteins ...... 85

4.3 Results and discussion ...... 87

4.3.1 Immobilization of His-tagged CYP3A4 onto Ni-NTA agarose resin ...... 87

4.3.2 Effect of immobilization on the end-point activity of CYP3A4 ...... 88

4.3.3 Effect of immobilization on the kinetic stability of CYP3A4 ...... 89

4.3.4 Effect of immobilization on the thermodynamic stability of CYP3A4 ...... 90

4.3.5 Assessing the suitability of Ni-NTA agarose as a lyoprotectant for CYP3A490

4.4 Conclusion ...... 91

Table of contents

Chapter 5 – Site-specific fluorescent labeling and oriented-immobilization of a triple mutant of CYP3A4 via C64 ...... 92

5.1 Preface ...... 92

5.2 Introduction ...... 93

5.3 Maleimide preparation and their reaction with CYP3A4 ...... 95

5.4 Results and discussion ...... 96

5.4.1 Analysis of fluorescently labeled wild-type CYP3A4 ...... 96

5.4.2 Design of cysteine-depleted CYP3A4 mutants for site-specific cysteine modification ...... 100

5.4.3 Effect of varying the maleimide label on enzyme activity ...... 104

5.4.4 Immobilization of CYP3A4 mutant 3 onto maleimide-functionalized CarboxyLink resin ...... 106

5.4.5 Immobilization of CYP3A4 mutant 3 onto maleimide-functionalized silica microspheres ...... 107

5.4.6 Preliminary results towards single-molecule fluorescence microscopy studies of CYP3A4 mutant 3 on silica microspheres ...... 108

5.5 Conclusion ...... 115

Chapter 6 – Contributions ...... 116

6.1 Selective aliphatic C-H oxidations ...... 116

6.1.1 Novel use of type II binders as P450 enzyme-targeting chemical auxiliaries ...... 117

6.2 Understanding activity and stability ...... 118

6.3 Progress towards applications in biotechnology and biophysical studies ...... 119

6.3.1 Site-specific fluorescent labeling of CYP3A4 ...... 119

6.3.2 Oriented-immobilization of CYP3A4 ...... 119

6.4 Publications ...... 120

Table of contents

6.5 Presentations ...... 120

Chapter 7 – Experimental protocols ...... 123

7.1 General methods ...... 123

7.1.1 Chemicals ...... 123

7.1.2 Spectroscopy ...... 123

7.2 Biological studies ...... 124

7.2.1 Molecular biology ...... 124

7.2.1.1 Site-directed mutagenesis of CYP3A4 ...... 124

7.2.1.2 Addition of His6-tag to CYP2E1 ...... 125

7.2.2 Expression and purification of enzymes ...... 125

7.2.3 Chemical auxiliary studies with CYP2E1 ...... 126

7.2.3.1 Small scale biotransformation of auxiliary-substrates 2.1a-2.7a and 2.9a-2.23a with CYP2E1 for yield and regioselectivity estimates ...... 126

7.2.3.2 Biotransformation of 1-hexanol (2.8a), 3-methyl-1-pentanol and cyclopentylmethanol by CYP2E1 ...... 127

7.2.3.3 Comparing extinction coefficients of 2.1a, 2.12a and 2.12b ...... 128

7.2.3.4 Spectral binding studies with CYP2E1 ...... 129

7.2.3.5 Removal of the nicotinate auxiliary ...... 129

7.2.3.6 Docking studies with CYP2E1 ...... 130

7.2.4 Macromolecular crowding studies with CYP3A4 and CYP2D6 ...... 130

7.2.4.1 Preparation of crowder and small sugar stock solutions ...... 130

7.2.4.2 Effect of macromolecular crowders on CYP3A4 activity ...... 130

7.2.4.3 Effect of macromolecular crowders on CYP2D6 activity ...... 131

7.2.5 C-terminal His-tag immobilization of CYP3A4 ...... 132

7.2.5.1 Preparation of Ni-NTA and Ni-free NTA agarose resin ...... 132

Table of contents

7.2.5.2 Yield of CYP3A4 immobilization on Ni-NTA agarose resin ...... 132

7.2.5.3 Kinetic stability of CYP3A4 on Ni-NTA agarose resin ...... 133

7.2.5.4 Thermodynamic stability of CYP3A4 on Ni-NTA resin ...... 134

7.2.5.5 Effect of immobilization of CYP3A4 on Ni-NTA on its tolerance to lyophilization ...... 134

7.2.6 Site-specific bioconjugation of CYP3A4 mutant 3 ...... 134

7.2.6.1 Cysteine-specific protein labeling with maleimides ...... 134

7.2.6.2 Generation of maleimide labels from their corresponding amines ..... 135

7.2.6.3 Assessing the effect of quenched maleimide labels on enzymatic activity...... 136

7.2.6.4 Estimating the labeling yield of CYP3A4 mutant 3 with the maleimide DyLight 549 (5.10) ...... 136

7.2.6.5 Sample preparation for single-molecule photobleaching analysis ...... 137

7.2.6.6 Quantification of labeling via single-molecule photobleaching analysis ...... 138

7.2.6.7 Quantification of labeling via single-molecule photobleaching analysis: Intensity analysis ...... 138

7.2.6.8 In-gel BrCN digestion of labeled protein...... 139

7.2.6.9 Preparation of maleimide-functionalized resins ...... 139

7.2.6.10 Immobilization of CYP3A4 mutant 3 onto maleimide-functionalized resins ...... 140

7.2.6.11 Preparation of maleimide-functionalized silica microspheres ...... 141

7.2.6.12 Immobilization of CYP3A4 mutant 3 onto maleimide-functionalized silica microspheres ...... 142

7.2.6.13 Silica microsphere immobilized-CYP3A4 mutant 3 activity assays with confocal fluorescence microscopy: towards single-molecule studies 143

7.2.6.14 CYP3A4 activity assays with testosterone ...... 145

7.2.6.15 Determination of initial rates for CYP3A4 using the substrate 7- benzyloxy-4-trifluoromethylcoumarin (BFC) ...... 146

xii

Table of contents

7.2.6.16 End-point activity determination for CYP3A4 using the substrate 7- benzyloxy-4-trifluoromethylcoumarin (BFC) ...... 146

7.3 Synthetic procedures ...... 147

7.3.1 Synthesis and characterization of auxiliary-substrates ...... 147

7.3.2 Synthesis and characterization of CYP2E1 oxidation products ...... 157

7.3.3 Synthesis and characterization of maleimides ...... 164

References ...... 169

Appendix I – NMR spectra ...... 189

Appendix II – Type II ligands as substrates of CYP2E1 ...... 229

xiii

List of figures

Figure 1.1 Cartoon representation of the different redox partners involved in the P450 catalytic cycle ...... 3

Figure 1.2 Catalytic cycle of cytochrome P450 where R-H is the substrate and R-OH is the product ...... 4

Figure 1.3 Structure of P450 BM3 heme domain representative of the conserved three- dimensional fold of P450s ...... 5

Figure 1.4 A) Conserved Cys ligand loop region in P450cam. B) Conserved I helix region in P450cam ...... 6

Figure 1.5 Comparison of the general fold adopted by soluble CYP101 vs. membrane- bound CYP2E1 ...... 7

Figure 1.6 Examples of human P450-catalyzed reactions...... 9

Figure 1.7 Examples of functional groups used in the bioconjugation of lysine and cysteine residues ...... 12

Figure 1.8 A) N-terminal-specific bioconjugation via a biomimetic transamination reaction. B) N-terminal specific bioconjugaion via nucleophilic addition to ketenes ...... 13

Figure 1.9 A) Mutated intein approach to C-terminal specific protein modification with a nucleophile H-Nu. B) Site-specific modification of a cell surface fused to an acceptor peptide...... 16

Figure 1.10 Synthetic scheme used by Uvarov et al.67 to modify FMN for covalent linkage to CYP2B4 via a cysteine residue ...... 20

Figure 1.11 Suite of activity-based probes for the P450 family of enzymes ...... 22

Figure 1.12 Warheads of mechanism-based inhibitiors in activity-based probes designed for P450 activity profiling ...... 22

Figure 1.13 Chemical auxiliary approach for improved product predictability with CYP3A4...... 34

xiv

List of figures

Figure 2.1 Types of cytochrome P450 ligands ...... 39

Figure 2.2 Chemical structures of antifungal azoles in clinical use ...... 40

Figure 2.3 Chemical structures of CYP17 inhibitors in clinical use or clinical trials for the treatment of prostate cancer ...... 40

Figure 2.4 Chemical structures of antihistamines cimetidine and ranitidine and the S- nitrosoglutathione reductase inhibitor N6022 ...... 41

Figure 2.5 General structure of pyridine quinoline-4-carboxamides ...... 42

Figure 2.6 Oxidation of nicotine-derived nitrosamine ketone (NNK) to 4-hydroxy-1-(3- pyridyl)-1-butanone by CYP2E1 ...... 43

Figure 2.7 Spectral binding characteristics and biotransformation of 2.1a with CYP2E1...... 44

Figure 2.8 Spectral binding characteristics of 2.6a with CYP2E1 ...... 47

Figure 2.9 Molecular modeling of compounds 2.1a and 2.9-2.23a with CYP2E1 ...... 53

Figure 3.1 Illustration of the excluded volume effect ...... 61

Figure 3.2 A) Thermodynamic cycle showing the relationship between the free energy of transfer of monomers and dimers from a dilute solution to a crowded medium and the free energy of association. B) Influence of volume fraction of crowding agent and dimer shape at constant dimer volume on the equilibrium constant for the formation of a dimer from two monomers ...... 63

Figure 3.3 Relationship between the reaction rate constant of a bimolecular association with respect to the concentration of crowding agent for diffusion-controlled and transition state-controlled regimes ...... 64

Figure 3.4 Chemical composition of crowding agents ...... 67

Figure 3.5 Structures of small sugars: sucrose, glucose, trehalose and melezitose ...... 67

Figure 3.6 Substrates used to measure the activity of CYP3A4 and CYP2D6 for end- point type assays in the presence of macromolecular crowding agents ...... 69

List of figures

Figure 3.7 Effect of the addition of macromolecular crowding agents or small sugars on the activity of CYP3A4 with the substrate testosterone ...... 70

Figure 3.8 Effect of varying small sugar concentrations on activity of CYP3A4 with the substrate testosterone and CHP as the cofactor surrogate ...... 71

Figure 3.9 Effect of the addition of macromolecular crowders or small sugars (20% w/v) on the activity of CYP2D6 with the substrate dextromethorphan ...... 72

Figure 3.10 Effect of varying small sugar concentrations on the activity of CYP2D6 with the substrate dextromethorphan and CHP as the cofactor surrogate ...... 73

Figure 3.11 Model of the “open-closed” equilibrium in CPR ...... 74

Figure 3.12 “Three-site sequential ping-pong” mechanism proposed by Davydov and Halpert to explain the cooperativity observed in CYP3A4 ...... 76

Figure 3.14 Analysis of the conformational flexibility of CYP3A4 ...... 77

Figure 4.1 Ni-NTA affinity support for binding His-tagged proteins ...... 86

Figure 4.2 Debenzylation of fluorogenic substrate 7-benzyloxy-4- trifluoromethylcoumarin (BFC) to 7-hydroxy-4-trifluoromethylcoumarin (HFC) by CYP3A4 ...... 87

Figure 4.3 Immobilization efficiency of His-tagged CYP3A4 on Ni-NTA agarose resin ...... 88

Figure 4.4 Kinetic stability of immobilized CYP3A4 ...... 89

Figure 4.5 Thermodynamic stability of immobilized CYP3A4 ...... 90

Figure 5.1 Thermodynamic stability of CYP3A4 at 25°C in 0.1 M potassium phosphate buffer, pH 7.4 ...... 94

Figure 5.2 Structures for molecules 5.1-5.9 and 5.14 ...... 96

Figure 5.3 Crystal structure of CYP3A4 with cysteine residues highlighted in red ..... 97

Figure 5.4 Single-molecule photobleaching of wild-type CYP3A4 labeled with DyLight 549 maleimide (5.10)...... 98

xvi

List of figures

Figure 5.5 A) Primary sequence of human CYP3A4 used here listed with the one letter code for amino acids. B) Crystal structure of CYP3A4 with cysteine residues highlighted in red...... 99

Figure 5.6 Tricine-SDS-PAGE analysis of BrCN digestion peptides of CYP3A4 wild- type (wt) labeled with DyLight 549 maleimide (5.10) ...... 100

Figure 5.7 Tricine-SDS-PAGE analysis of BrCN digestion peptides of CYP3A4 variants labeled with DyLight 549 maleimide (5.10) ...... 101

Figure 5.8 Absorbance spectra of CYP3A4 mutant 3 labeled with DyLight 549 maleimide (5.10) ...... 103

Figure 5.9 Relative activity of CYP3A4 wild-type (in blue) and mutant 3 (in purple) after labeling with maleimides 5.1-5.13 ...... 104

Figure 5.10 Effect of free fluorophores on the enzymatic activity of non-labelled CYP3A4 mutant 3...... 105

Figure 5.11 Effect of labeling with maleimides 5.9d and 5.2d on the CO-difference spectrum of CYP3A4 mutant 3 compared to the non-labeled control ...... 106

Figure 5.12 Immobilization of CYP3A4 mutant 3 on silica microspheres ...... 108

Figure 5.13 Debenzylation of fluorogenic substrate 7-benzyloxy-4- trifluoromethylcoumarin (BFC) to 7-hydroxy-4-trifluoromethylcoumarin (HFC) by CYP3A4 ...... 109

Figure 5.14 Preparation of doubly modified silica microspheres ...... 110

Figure 5.15 Example of an intensity vs. time trajectory acquired upon 406 nm excitation of a single silica microsphere bearing CYP3A4 in the presence of the substrate BFC and the surrogate cofactor CHP ...... 111

Figure 5.16 Fluorescence scanning confocal images of CYP3A4 activity on single silica microspheres ...... 112

Figure 5.17 Mechanisms of peroxo O-O bond cleavage of cumene hydroperoxide (CHP) by the heme-iron in P450 enzymes ...... 113

xvii

List of schemes

Scheme 1.1 Overall reaction of P450 mono-oxygenation of a substrate RH ...... 2

Scheme 5.1 A) General reaction used to label wild-type and mutant CYP3A4 enzymes with maleimides 5.1-5.14. B) General approach to preparing the non- commercial maleimide labels...... 95

Scheme 5.2 Cleavage of the peptide backbone after a methionine residue by cyanogen bromide (BrCN) ...... 99

xviii

List of tables

Table 1.1 Reports of P450-catalyzed reactions in whole cells supported by co- expression of CPR ...... 28

Table 2.1 Characterization of auxiliary-substrates 2.1a-2.8a with respect to binding mode and spectral binding constant (Ks) for the corresponding CYP2E1 complex, as well as ratio of different regioisomeric products and yield of transformation by CYP2E1 ...... 45

Table 2.2 Characterization of auxiliary-substrates 2.9a-2.23a with respect to binding mode and spectral binding constant (Ks) for the corresponding CYP2E1 complex, as well as ratio of different regioisomeric products and yield of transformation by CYP2E1 ...... 49

Table 2.3 Structures of the major CYP2E1 oxidation products for auxiliary-substrates 2.10a-2.14a, 2.18a, 2.20a and 2.22a ...... 51

Table 4.1 General protein immobilization strategies: advantages and disadvantages .. 84

xix

Abbreviations

A – alanine

ABPP – activity-based protein profiling

ADP – adenosine diphosphate

AP – acceptor peptide

BFC – 7-benzyloxy-4-trifluoromethylcoumarin

BMR – reductase domain of P450 BM3

BSA – bovine serum albumin

C – cysteine cDNA – coding deoxyribonucleotide

C-H – carbon-hydrogen

CHP – cumene hydroperoxide

CPR – cytochrome P450 reductase

CYP or P450 – cytochrome P450 enzyme

CYP3A4 – human cytochrome P450 3A4

CYP2D6 – human cytochrome P450 2D6

CYP2E1 – human cytochrome P450 2E1

DCM - dichloromethane

DEX 70 – dextran 70K

DEX 500 – dextran 500K

DIEA – diisopropylethylamine

DLPC – 1,2-dilauroyl-sn-glycero-3-phosphocholine

DTT – dithiothreitol

EDC – 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide

EDTA – ethylenediaminetetraacetic acid

Abbreviations

F – phenylalanine

FAD – flavin dinucleotide

FdR – ferredoxin reductase

Fdx – ferrodoxin

FIC 70 – Ficoll® 70K

Fe-O – iron-oxygen

FMN – flavin mononucleotide

FRET – Förster resonance energy transfer

G – glycine

GC-MS – gas chromatography-mass spectrometry

GMBS – N-(γ-maleimidobutyryloxy)succinimide ester

GST – glutathione S-transferase

HFC – 7-hydroxy-4-trifluoromethylcoumarin

His – histidine

HOBt – hydroxybenzotriazole

I – isoleucine

IMER – immobilized enzyme reactor

Ks – spectral dissociation constant

LC-MS – liquid chromatography-mass spectrometry

LC-UV-MS – liquid chromatography-ultraviolet-mass spectrometry

LSPR – localized surface plasmon resonance

NAD(P)H – nicotinamide adenine dinucleotide (phosphate)

NHS – N-hydroxysuccinimide

NMR – nuclear magnetic resonance

NNK – nicotine-derived nitrosamine ketone

xxi

Abbreviations

NTA – nitrilotriacetic acid

P – proline

PCR – polymerase chain reaction

PDB – protein data bank

PEG – polyethylene glycol

PEG 6 – polyethylene glycol 6K

Phe – phenylalanine

PLP – pyridoxal phosphate

Prep-TLC – preparatory thin layer chromatography

PVP 40 – polyvinylpyrrolidone 40K

RMSD – root-mean-square deviation

S – serine

SAM – self-assembled monolayer

SDS-PAGE – sodium dodecyl sulfate-polyacrylamide gel electrophoresis

SPR – surface plasmon resonance

TCEP – tris(2-carboxyethyl)phosphine

TFA – trifluoroacetic acid

TIRFM – total internal reflection fluorescence microscopy

TLC – thin layer chromatography tRNA – transfer ribonucleic acid

UAG – uracil-adenine-guanine

UV – ultraviolet

V – valine

W – tryptophan

ω-1 – omega minus 1

xxii

Chapter 1

Chapter 1 – Introduction

1.1 Preface

The purpose of this chapter is to introduce various aspects of cytochrome P450 enzymes related to the work detailed in subsequent chapters. The first section is a general introduction of P450 enzymes and includes discussions of their catalytic mechanism, of some important structural features and of the differences between P450 enzymes of human versus bacterial origins. The second section provides the background information and context specifically required in order to better appreciate the results presented in

Chapter 5: the bioconjugation of CYP3A4. Here, general challenges associated with protein bioconjugation are discussed. We argue that site-specificity presents many advantages when choosing a bioconjugation strategy and such site-specific approaches are examined. A review of the literature reporting instances of bioconjugation with P450 enzymes is also presented. In the third section, the challenges associated with using P450 enzymes as biocatalysts are discussed. Several aspects of this issue are presented and include discussions of the recombinant expression of these enzymes, their dependence on cofactors, their generally poor stability and the challenges associated with product predictability.

1.2 Cytochrome P450 enzymes

Cytochrome P450 enzymes or CYPs form a large family of heme-dependent monooxygenases that are found in all domains of life. This type of cytochrome was first characterized in 1962 by Omura and Sato as a new pigment from rabbit liver microsomes

Chapter 1

that, when reduced and complexed with carbon monoxide, absorbs strongly at 450 nm; hence the name cytochrome P4501. CYPs are involved in the biosynthesis of secondary metabolites such as steroids, lipids and vitamins, and the metabolism of xenobiotics such as drugs, pesticides and carcinogens. Of note is their ability to oxidize inactivated C-H bonds, often with remarkable regio- and stereoselectivity.

1.1.1 Cytochrome P450 catalytic mechanism

The overall reaction of C-H oxidation by CYPs involves the activation of molecular oxygen by the heme-iron center such that one oxygen atom is inserted into the substrate while the other is reduced to water (Scheme 1.1). The electrons are derived from

+ - RH + O2 + 2 H + 2 e ROH + H2O

Scheme 1.1 – Overall reaction of P450 monooxygenation of a substrate RH. Activation of molecular oxygen by the enzyme leads to one oxygen being inserted into the substrate to produce ROH and the other being reduced to water. nicotinamide adenine dinucleotide (phosphate) (NAD(P)H) and are delivered to the P450 by enzymatic redox partners2. The latter are usually expressed independently like the iron-sulfur containing ferredoxin (Fdx) and flavin adenine dinucleotide (FAD)-containing ferredoxin reductase (FdR) that shuttle electrons to bacterial and mitochontrial P450s, or the FAD- and flavin mononucleotide (FMN)-containing NADPH cytochrome P450 reductase (CPR) used by mammalian microsomal P450s. However, there are examples of

“self-sufficient” P450s in which the P450 and reductase domains are fused like in P450

BM3 (Figure 1.1)3.

Chapter 1

A) B)

C) D)

Figure 1.1 – Cartoon representation of the different redox partners involved in the P450 catalytic cycle. A) A common system in which redox enzymes ferredoxin (Fdx) and ferredoxin reductase (FdR) shuttle electrons from NAD(P)H to soluble bacterial P450s. B) Mitochondrial P450s are anchored to the inner mitochondrial membrane and are also reduced by FdR and Fdx. C) Microsomal P450s are reduced by the redox enzyme cytochrome P450 reductase (CPR); both are on the outer membrane of the endoplasmic reticulum (ER). D) Self-sufficient P450 BM3: the reductase and P450 domains are on the same polypeptide. (Image taken from Bernhardt et al.3)

A great deal of time and effort has been dedicated to the complete elucidation of the mechanism by which CYPs activate and transfer oxygen to their substrates. A scheme that summarizes the current consensus of the catalytic cycle for P450 hydroxylation can be seen in Figure 1.2.4In its resting state, the heme-iron is coordinated to four nitrogens from the porphyrin ring, to a cysteinate proximal ligand and to a water molecule distal ligand. Iron is in the +3 oxidation state (ferric). The first step of the cycle involves the displacement of the distal water ligand by the substrate R-H resulting, in most cases, in a shift in the spin-state of the iron from low-spin to high-spin (step 1). Next, the ferric heme enzyme is reduced to the ferrous (+2) state by the transfer of a first electron from NADPH via the redox partner enzyme(s) (step 2). The resulting species then binds molecular

Chapter 1

oxygen (step 3) and is reduced again by the transfer of the second electron from the redox partner (step 4), before protonation to form the iron(III)-hydroperoxy complex, also known as Compound 0 (step 5).2

Subsequent protonation and heterolytic cleavage of the O-O bond leads to the formation of Compound I (step 6). For many decades, the evidence towards Compound I being the catalytically relevant species of the cycle has been mounting, but only recently

(in 2010) has it finally been directly observed and characterized spectroscopically5. This iron(IV)-oxo intermediate with a radical cation delocalized over the porphyrin and the

Figure 1.2 – Catalytic cycle of cytochrome P450 where R-H is the substrate and R-OH is the product. Steps of the main cycle are numbered in a clockwise fashion and shunt pathways are indicated with letters. For simplification purposes, the P450 active site is depicted by a cartoon representation of heme b (a protoporphyrin IX-iron complex) ligated to the cysteine thiolate ligand (only the sulfur atom is shown for clarity). Note: the substrate R-H binds in the active site on the distal side of the heme (opposite the proximal thiolate ligand), but does not coordinate directly to the heme-iron. The binding of R-H affects the conformation of the enzyme, and therefore its interaction with the heme prosthetic group. This in turn affects the spin state of the iron. R-H remains bound throughout the oxygen activation process until its oxidation and release as R-OH. (Image taken from Munro et al.4) 4

Chapter 1

thiolate ligand, is responsible for abstracting a hydrogen atom and generating the substrate radical (R•) (step 7). Finally, oxygen-transfer from the iron-oxo to the substrate radical occurs, and the hydroxylated product R-OH is released (step 8).2

As depicted in Scheme 1.1, the overall productive pathway of CYP substrate oxidation requires two electrons, two protons and one molecule of oxygen. However, in some cases, this stoichiometry is altered and the cycle is interrupted or “uncoupled”.

There are three uncoupling pathways in the catalytic cycle as illustrated in Figure 1.2. The first so-called oxidase shunt pathway involves the breakdown of Compound I, consuming two additional redox equivalents and releasing H2O (A). The second involves the protonation and breakdown of Compound 0 releasing H2O2; the peroxide shunt (B). Last is the autoxidation shunt pathway whereby the cycle is interrupted by the release of

•- 2 superoxide (O2 ) by the ferrous dioxy intermediate (C).

The manner in which CYPs control and use these uncoupling pathways to fine- tune their activity is not yet fully understood. However, the degree of uncoupling can vary greatly between CYPs or between different substrates for a given CYP. Also, mammalian

CYPs tend to display lower coupling efficiencies than bacterial ones in vitro. Some factors known to play an important role are the H-bond network at the P450 active site6

(see section 1.1.2), the nature of the substrate and active site water access7.

1.1.2 Structural features of P450 enzymes

There are now over 70 X-ray structures of P450 enzymes available in the Protein

Data Bank, all of which display the same general fold8 that is unique to this class of enzymes (Figure 1.3). This is quite remarkable for such a diverse class of enzymes in terms of size, cellular localization, function and sequence (<20% overall identity)9.

Chapter 1

Nevertheless, there are some highly conserved residues and structural elements. These are located close to the heme group and are essential to enzymatic catalytic function.

Variations in other regions account for substrate diversity and cellular localization.

The first and most important conserved secondary structural element is the cysteine ligand loop found in all P450s (Figure 1.4A)2. It is stabilized by a peptide backbone H-bond between the carbonyl oxygen of a phenylalanine residue and the N-H of the cysteine ligand. This architecture is responsible for maintaining a proper heme-iron redox potential. More recently, it has been suggested that CYP119, a thermophilic P450, may derive some of its thermal stability by constraining this cysteine ligand loop8. The second highly conserved region is responsible for proper proton delivery to the heme-iron center during the catalytic cycle and is therefore crucial to the formation of the active iron-oxo complex, Compound I2. It is found near the heme group and, as seen in Figure

1.4B, consists of an H-bond between the side chain OH of a conserved threonine residue and a peptide carbonyl oxygen, both in the I helix (see Figure 1.3).

Figure 1.4 – A) Conserved Cys ligand loop region in P450cam. B) Conserved I helix region in P450cam. Black lines represent the peptide backbone, grey lines represent the heme and dashed lines represent H-bonds. (Image taken from Ortiz de Montellano2)

Chapter 1

Most mammalian P450s are membrane-bound and therefore have an added transmembrane domain that is lacking in bacterial P450s. It is located at the N-terminus, is composed of 30-50 amino acids and is responsible for anchoring the enzyme into the cytoplasmic side of endoplasmic reticulum membrane. This domain is not strictly required for function however and its modification has proven essential in the successful recombinant expression of human CYPs in E. coli10, greatly facilitating their purification and study.

More recently, another more subtle structural difference between soluble and membrane-bound CYPs has been observed. Upon performing a systematic comparison of the X-ray structures of 25 membrane-bound CYPs and 45 soluble ones, Denisov et al. found that the β-domain of membrane-bound CYPs is shifted towards the proximal face of the porphyrin ring (Figure 1.5)9. (The proximal face is that to which the cysteine thiolate ligand binds.) This changes the location of the main substrate access channel and increases the size of the binding pocket allowing access to lipophilic substrates directly from the membrane.

Figure 1.5 – Comparison of the general fold adopted by soluble CYP101 vs. membrane-bound CYP2E1. Note the difference in position of the β-domain (yellow) relative to the heme plane (red).

Chapter 1

1.1.3 Human cytochrome P450 enzymes

In humans, P450 enzymes expressed in the liver play a major role in the oxidative metabolism of xenobiotics. What’s more, most of the known CYP-mediated metabolic reactions are performed by only twelve isoforms: CYP1A1, 1A2, 1B1, 2A6, 2B6,

CYP2C8, 2C9, 2C19, 2D6, 2E1, 3A4 and 3A5. Thus, to accomplish their function, these enzymes must have the ability to recognize a great variety of molecules. In fact, ~75% of all known drugs are metabolised by CYPs5, and of these ~50% by CYP3A4 alone11. As mentioned above, aliphatic C-H bond hydroxylation is likely the most intriguing P450- catalyzed reaction. This type of reaction is difficult to achieve by chemical means, especially with comparable regio- and stereoselectivities. There are over 60 other types of

P450-catalyzed reactions12 including aromatic hydroxylation, N- and O-dealkylation, N- and S-oxidation, epoxidation of alkenes and arenes, oxidation of alcohols and unsaturated ring systems, isomerizations and carbon-carbon bond cleavage among others3, 13 (see

Figure 1.6 for examples).

It is this combination of high substrate promiscuity, wide scope of reactivities and high regio- and stereoselectivity that has attracted the attention of scientists from a variety of disciplines. Many now study these remarkable enzymes in hopes of better mimicking their abilities chemically or of exploiting them to develop new applications in biotechnology. Because of their versatility, some claim that human CYPs possess more advantageous attributes with respect to such applications compared to bacterial CYPs12, 14-

15. Consequently, we have selected three human isoforms to be the subject of our investigations with respects to applications in protein bioconjugation and biocatalysis:

Chapter 1

CYP3A4, CY2D6 and CYP2E1. Discussions below (sections 1.2 and 1.3) will be limited to human CYPs with a few relevant examples from their mammalian homologues.

Figure 1.6 – Examples of human P450-catalyzed reactions.

Chapter 1

1.3 Protein bioconjugation – methods and applications

1.3.1 The challenge

Bioconjugation refers to the covalent linkage of biomolecules to small molecules, solid matrices or other biomolecules. In recent years, the desire to generate protein and enzyme bioconjugates has exploded with the rise of more sophisticated biochemical and biophysical analytical techniques and applications such as single-molecule spectroscopy, the fabrication of protein microarrays, biosensors and bioreactors, diagnostics, biocatalysis and the in vivo visualization, tracking and targeting of biomolecules. As a result, the branch of chemical biology concerned with the development of bioconjugation strategies and “bioorthogonal” chemistry16-17 has grown tremendously in order to meet the specific criteria required by these diverse applications.

Despite the expanding tool box, each strategy has its limitations and most have only been tested within a limited context. Therefore, the difficulty still remains in choosing the appropriate strategy that will provide bioconjugates that retain all the characteristics, including activity, functionality and stability, of the native protein of interest18. However, the complexity of the chemical space found on protein surfaces can lead to unexpected reactivities (i.e. due to local pH differences, conformation or dynamic effects). Compounding the problem is the fact that many methods suffer from slow, second-order kinetics and therefore often require high reagent concentrations, harsh conditions (i.e. high temperatures, extreme pHs) and long reaction times that are incompatible with many proteins. Ultimately, meeting all these demands is often quite challenging and involves a great deal of trial and error.

Chapter 1

1.3.2 The advantage of site-specificity

Likely the most straightforward and widely applied protein bioconjugation strategies exploit the reactivities of functional groups found in the canonical amino acids.

The most commonly used are the highly abundant lysine residues. They are typically derivatized via amide formation with N-hydroxysuccinimide (NHS)-esters. Lysine residues can also be modified with isocyanates, isothiocyanates, epoxides, aldehydes, ketones, sulfonyl chlorides and carboxylic acids in the presence of a coupling agent

(Figure 1.7). Note that in all cases, the N-terminus is likely to be modified as well.

Although less abundant, cysteine residues are the next most popular target. At neutral pH,

>99.9% of lysine residues are protonated and thus, cysteine residues (of which ~10% are thiolates) can be selectively modified via conjugate addition with maleimides, monobromo- and dibromo-maleimides, iodoacetamides, acrylates and vinyl sulfones yielding thioethers (Figure 1.7). Although cysteine and lysine are the usual targets, tyrosine- and tryptophan-specific reactions have also been described.16

While these strategies hold the advantage that they do not require any prior modification of your protein of interest (further discussed below) and are therefore generally applicable and easily executed, they can suffer from several pitfalls. This is because, since most proteins contain multiple copies of any given amino acid, these methods inherently lack control and specificity. This can disrupt important residues

(involved in catalysis or proper folding) and consequently result in altered functionality, structure, activity and/or stability. Moreover, even if these properties are unaffected, the resulting bioconjugate population will often lack uniformity, a property that is especially important, for example, in single-molecule experiments in which one wishes to detect

Chapter 1

biologically significant heterogeneities within a given population. Thus, it is now widely acknowledged that the best way to avoid these drawbacks is by making use of site- specific bioconjugation strategies, especially with regards to protein immobilization.19,20-

Figure 1.7 – Examples of functional groups used in the bioconjugation of lysine (left, purple protein) and cysteine (right, green protein) residues.

1.3.3 Site-specific protein bioconjugation strategies

When it comes to such site-specific strategies, they can be divided into two categories: those that only seek to make use of the functional groups found in the canonical amino acids and those that require the introduction of unnatural functional groups into proteins. In the first category, the most common strategy has been to use site-

Chapter 1

directed mutagenesis in order to remove unwanted reactive cysteines and/or introduce a new one at a desired location such that the resulting mutant has only one reactive cysteine

18 residue . Others have exploited the lower pKa of N-terminal amino groups (~8) relative to those of lysine residues (~10) to specifically target the N-terminus for bioconjugation.

This has been achieved via a biomimetic transamination reaction developed by the

Francis lab22 (Figure 1.8A) or more recently by reacting with ketenes23 (Figure 1.8B).

Figure 1.8 – A) N-terminal-specific bioconjugation via a biomimetic transamination reaction. Transamination occurs in the first step by reaction between the N-terminal amine and pyridoxal phosphate (PLP) to generate bioorthogonal carbonyl functionality at the protein N-terminus and pyridoxamine as the byproduct. The resulting carbonyl can then be derivatized with alkoxyamines through an oxime linkage. B) N-terminal specific bioconjugaion via nucleophilic addition to ketenes.

Chapter 1

Yet another strategy has been the activity-based labeling of enzymes based on their catalytic mechanism16, 18. This involves the use of activity-based probes that react specifically with a group of mechanistically related enzymes (e.g. kinases or proteases).

These probes consist of a warhead linked to a reporter tag. The warhead is often a mechanism-based inhibitor. Its modification by the target enzymes uncovers an electrophile that reacts with nearby nucleophiles in the active site. The reporter tag serves to facilitate retrieval and visualization of the target enzyme once it becomes covalently linked to the warhead. These can be fluorescent tags, affinity tags or latent bioorthogonal chemical reporters (e.g. alkynes, azides). Since this approach inherently leads to enzyme inactivation, its applications are limited to the proteomics field of activity-based protein profiling (ABPP) which involves the quantification of enzyme activity in complex biological mixtures. A specific example involving P450 enzymes will be presented in the next section (1.2.4).

The second type of site-specific bioconjugation strategies comprises those that require the incorporation of new functional groups into proteins. The number of strategies that fit into this category has multiplied during the last few years. And, in the view of many, these ingenious approaches based on what is now termed “bioorthogonal” chemistry16-17, have truly revolutionized the field of bioconjugation. In order to develop a bioconjugation method based on new functional groups, one must first find a reaction that can take place selectively in the presence of other functional groups found in native biomolecules, and under “protein-friendly” conditions (i.e. aqueous buffered conditions, neutral pH and room temperature, potentially in cell lysates and ultimately, in vivo). It follows that such reactions must involve functional groups not naturally present in

Chapter 1

proteins such as azides, alkynes, alkenes, ketones, aldehydes, phosphines, tetrazoles, tetrazines, alkoxyamines, hydrazides, anilines and thioesters. To date, many such reactions have been described and improvements (especially in terms of kinetics) are constantly being sought. These include variations of the alkyne-azide cycloaddition24-28, the Staudinger ligation29-32, the Diels-Alder cycloaddition33-36, oxime or hydrazone ligations37-42, olefin metathesis43-44, palladium-catalyzed cross-couplings45-48, the quadricyclane ligation49, and the oxidative coupling of anilines and aminophenols50.

Equally essential to the puzzle, are the variety of methods that have been developed by which functional groups required to perform the above mentioned reactions have been introduced into proteins. These rely largely on our ability to genetically manipulate protein sequences through molecular biology techniques. An important such method has been the in vivo incorporation of unnatural amino acids into proteins (during their heterologous expression in E. coli or S. cerevisiae) pioneered by the Schultz lab at

The Scripps Research Institute. This method has been used for the site-specific incorporation of alkyne-, azide-, alkene- and ketone-containing unnatural amino acids among others51. This is achieved by hijacking protein synthesis with an engineered tRNA/tRNA synthetase pair that orthogonally recognizes and incorporates a given unnatural amino acid at the amber codon (UAG). Thus, the location of this codon, which can easily be introduced at a desired location by site-directed mutagenesis, will dictate the location of the unnatural amino acid bearing bioorthogonal functionality. The latter can then serve as a handle for site-specific bioconjugation via the corresponding bioorthogonal reaction16.

Chapter 1

Others have developed methods that make use of genetic fusions to modified intein domains (Figure 1.9A) or enzyme modified peptide-tags (Figure 1.9B) to add unnatural functionality to proteins. One limitation of these strategies is that they limit bioconjugation to the protein termini. In the former, a modified intein domain is fused to the C-terminus of a protein of interest. An intein is sometimes referred to as a “protein intron”. It is a domain that self-catalyses its excision from a protein sequence and rejoins the ends. In this strategy, the intein is mutated such that the excision reaction is

Figure 1.9 – A) Mutated intein approach to C-terminal specific protein modification with a nucleophile H-Nu. Here the intein is fused to an affinity tag to facilitate purification of the final bioconjugate. (Image taken from Kalia et al.54). B) Site-specific modification of a cell surface fused to an acceptor peptide (AP). AP is an amino acid sequence recognized by biotin ligase. Biotin ligase was used to add Ketone 1, a derivative of biotin, to a specific lysine in the AP. Subsequent reaction with a fluorescent dye (green cycle) bearing a hydrazine group results in fluorescent labeling of the cell surface protein. (Image adapted from Chen et al.55)

Chapter 1

incomplete, and a bioorthogonal C-terminal thioesters linkage is generated (Figure

1.9A)52-54. Such thioesters have been displaced with nitrogen nucleophiles and cysteine- modified small molecule nucleophiles as a means to achieve C-terminal specific bioconjugates. On the other hand, the peptide tag approach exploits Nature’s collection of enzymes that recognize and modify specific amino acid sequences or tags. Many groups have engineered such enzymes to accept modified substrates, and have used them to site- specifically incorporate bioorthogonal functional groups into proteins that have been fused to the corresponding peptide-tag16. The example shown in Figure 1.9B involves the site-selective modification of a cell surface protein with a modified biotin molecule using biotin ligase from BirA E. coli55. In this case, the cell surface protein is fused to a 15 amino acid acceptor peptide (AP) recognized by BirA, which subsequently adds the modified biotin to a lysine residue within this specific peptide sequence.

Although, a great variety of bioconjugation strategies are now available, careful consideration must be taken when choosing one18. Many methods still suffer from slow kinetics and therefore require harsh conditions and long reaction times, which may be problematic in certain situations. But improvements are on the horizon. The fastest bioorthogonal reaction known thus far was developed in 2008: the tetrazine ligation34. It entails an inverse electron demand Diels-Alder reaction between trans-cyclooctenes and

-1 -1 tetrazines with rate constants (k2) ranging from of 2000-6000 M •s . Earlier this year, this reaction was adapted for in vivo site-specific bioconjugation of proteins by incorporating a tetrazine-bearing unnatural amino acid into a protein for the first time56.

Advancements like these will undoubtedly continue such that ultimately, the site-specific bioconjugation of even the most delicate and unpredictable of proteins will be possible.

Chapter 1

1.3.4 Bioconjugation applied to cytochrome P450 enzymes

The vast field of P450 bioelectrochemistry has produced some interesting examples of P450 bioconjugates. Most studies in this area seek to take advantage of the remarkable properties of these enzymes (their ability to specifically recognize a vast scope of pharmaceuticals and xenobiotics as well as their unique catalytic capabilities as powerful yet controlled oxidants) to generate electrochemically driven biosensors and bioreactors. These types of P450-based devices could be used for bioremediation, to monitor drug levels in blood plasma, to detect environmental or food contaminants such as pesticides or in the production of fine chemicals, drugs or drug-metabolites15. Being electrochemically driven, such devices should be catalytically self-sufficient. That is, the electrons necessary for catalytic activity would be supplied directly by the electrode, bypassing the usual mediation by a redox partner enzyme. This type of electrode-driven catalysis has proven challenging to achieve efficiently with P450s. Thus, particular attention has been paid to the procedures used to immobilize them onto the electrode surface in order to maintain function and reduce uncoupling, ensuring that electron consumption is strictly linked to product formation57. While most examples to date involve their non-specific entrapment within conducting polymer or surfactant films at the electrode surface, there are some successful examples with covalently immobilized

CYPs57-58.

The first was reported in 2004 with CYP2E159. In this study, CYP2E1 was immobilized to a maleimide-modified gold electrode via two available surface cysteines and electrode-driven conversion of the substrate 4-nitrophenol to 4-nitrocatechol was observed. CYP2C9 has also been immobilized to a gold electrode through a self-

Chapter 1

assembled monolayer (SAM) composed of 11-mercaptoundecanoic acid and octanethiol.

It is claimed to have been linked in a specific fashion via its N-terminal lysine by amide bond formation to the carboxyl groups exposed on the surface of the SAM60. The resulting biosensor-like construct afforded the electrocatalytic conversion of warfarin to

7-hydroxywarfarin. CYP2B4 (the rabbit homologue of CYP2B6) has also been covalently attached to an electrode through a SAM, this one composed of mercaptopropionic acid61.

This was achieved by amide bond formation between surface amino groups on the enzyme and the carboxyl groups on the SAM. Detection of the antiepileptic drug phenobarbitol was demonstrated by this CYP2B4-based device. A later report demonstrated its ability to also detect cocaine62. Similarly, mercaptopropionic acid capped ZnSe quantum dots have also been used to fabricate a highly sensitive CYP3A4- based biosensor for 17β-estradiol63.

In some cases, efficient electron transfer from the electrode to the P450 has required the presence of an enzymatic or chemical mediator58. In one example, CYP3A4 was fused to the reductase domain of P450 BM3 (BMR) or to flavodoxin from

Desulfovibrio vulgaris64. These chimeras were then covalently immobilized to cystamine- maleimide-coated gold electrodes by reacting with surface cysteine thiols. Fusion to these domains was found to improve product formation by increasing coupling efficiency compared to CYP3A4 alone. Interestingly, covalent modification of CYP2B4 and

CYP1A2 with the redox cofactor riboflavin has also helped in rendering them electrochemically competent65. CYP bioconjugation to riboflavin was achieved by first activating the latter with carbonyldiimidazole and then reacting with amine groups on the enzyme surface66. These two semi-synthetic flavocytochromes were able to

Chapter 1

electrochemically catalyze aminopyrine N-demethylation, aniline p-hydroxylation, and 7- ethoxyresorufin and 7-pentoxyresorufin O-dealkylation when interfaced with a rhodium- graphite electrode with rates comparable to NADPH-supported reactions.

In a non-electrochemically related example, CYP2B4 covalently linked to FMN was catalytically active in the presence of NADPH67. As seen in Figure 1.10, FMN was first modified with 1,8-diaminooctane by formation of a phosphonamide linkage at its 5’- phosphate group. Covalent attachment of the resulting flavin was achieved by means of the heterobifunctional crosslinking agent N-(γ-maleimidobutyryloxy)succinimide ester

(GMBS). Authors claim the site of modification to be Cys-152.

Figure 1.10 – Synthetic scheme used by Uvarov et al.67 to modify FMN for covalent linkage to CYP2B4 via a cysteine residue. Other types of P450-based bioreactors that are not electrochemically driven have also been described for potential use in drug metabolism studies. These therefore require the use of biological cofactors and often incorporate cofactor regeneration systems. In a report by Nicoli et al., CYP3A4 and CYP2D6 bioconjugates were used to generate

Chapter 1

immobilized enzyme reactors (IMERs)68. These devices consisted of reconstituted liposomal membranes containing both a CYP and their redox partner CPR. The CYPs were biotinylated on surface amino groups with NHS-PEG12-biotin enabling their attachment to the NeutrAvidin coated reactor surface. When coupled to an LC-MS detector, the metabolism of dextromethorphan and midazolam (by CYP2D6 and CYP3A4 respectively) was successfully observed. Incorporation into these IMERs also improved their kinetic stability from less than an hour in solution to two days. Another example described by Tanvi et al. involved CYP2E169. In this case, the CYP was immobilized to a silane-maleimide modified porous alumina membrane through surface cysteines as part of a reconstituted microsome containing CPR. This was then inserted into a fluidics device, preceded by a membrane containing glucose-6-phosphate dehydrogenase. The latter converts NADP+ and glucose-6-phosphate to NADPH and glucolactone and therefore acts as a cofactor regenerating system, supplying NADPH to the CYP. The reactor was tested with a fluorogenic substrate 3-cyano-7-ethoxycoumarin and was found to retain 100% of the activity observed for the same system in solution.

As mentioned in the previous section (1.2.3), one strategy that has been used to generate enzyme bioconjugates involves reaction with activity-based probes for activity- based protein profiling (ABPP). In fact, a suite of P450-based affinity probes has been developed by the Cravatt lab at The Scripps Research Institute (Figure 1.1)70-71. These are composed of a mechanism-based inhibitor linked to a terminal alkyne for click chemistry conjugation to a reporter tag.

Chapter 1

Figure 1.11 – Suite of activity-based probes for the P450 family of enzymes. Each probe consists of a warhead and a latent bioorthogonal alkyne reporter tag (R).

Within these molecules, there are three types of warheads that, once transformed by the P450, become highly electrophilic and react immediately with nucleophilic residues in the active site, thus inactivating the enzyme. The first is an aryl alkyne that is oxidized to a ketene (Figure 1.12A); the second is a propynyl group that is converted to a

Michael acceptor (Figure 1.2B); and the third is a furanocoumarin that is oxidized to an epoxide (Figure 1.2C). These probes were successfully used to profile activity changes in a panel of human CYPs when exposed to the aromatase inhibitor anastrozole, a drug currently used in the treatment of breast cancer.

Figure 1.12 – Warheads of mechanism-based inhibitiors in activity-based probes designed for P450 activity profiling. A) An aromatic alkyne is oxidized to a ketene. B) A propyne group is converted to a Michael acceptor. C) A furanocoumarin is oxidized to an epoxide. Once formed, these electrophiles react with nucleophilic groups in the P450 active site to form covalent adducts.

Chapter 1

A similar approach uses photo-affinity labels to probe and identify important residues in the active sites of enzymes. Such labels are designed to be recognized by a given enzyme. Once bound, the probes are excited with light to generate reactive intermediates (e.g. carbenes or nitrenes) that react with active site residues resulting in covalent adducts. Lapachenole and 7-azido-4-methylcoumarin have been used to identify important residues in the active sites of CYP3A472 and CYP2E173 respectively.

Finally, bioconjugation of proteins with fluorescent dyes has found its way into many biological applications. A popular example is their use to quantify protein dynamics using Förster resonance energy transfer (FRET). To this end, Tsalkova et al. have described the site-specific fluorescent labeling of a cysteine-depleted mutant of

CYP3A474. This mutant had only one remaining reactive cysteine such that, when attached at this position, fluorescent dyes were able to communicate with the nearby heme group through FRET. The resulting bioconjugates were used to gain insight into conformational changes that occur upon binding to various substrates and inhibitors. A similar cysteine-depleted mutant of CYP3A4 was also used to monitor its oligomerization in proteoliposomes by FRET75.

1.4 Biocatalysis

When looking to design a new catalyst, chemists seek properties such as high chemo-, regio- and stereoselectivity, broad substrate scope, high efficiency and stability, easy and low-cost production as well as low toxicity. Interestingly, the immense variety of enzymes that Nature uses to catalyze the reactions that sustain life can possess many of these qualities. In fact, there exists an enzymatic manifestation of almost every type of

Chapter 1

reaction in organic chemistry76. The most attractive advantage of using enzymes as biocatalysts is their exquisite chemo-, regio- and enantioselectivity77. Nowadays, their ability to catalyze reactions under mild conditions using water as a solvent is also considered as an advantage due to rising environmental concerns and an increased desire to adhere to the tenets of “green chemistry”.

Many scientists specializing in a variety of areas have recognized these advantages and, over the last few decades, the field of biocatalysis has expanded dramatically into various biotechnological applications. Currently, over 500 commercial products are prepared with the help of enzymes and in 2009, the industrial enzymes market reached 5.1 billion dollars US78. However, in order to get to where we are today, many hurdles had to be overcome. That is because many enzymes suffer from low stability, high production cost, narrow substrate scope and a requirement for expensive cofactors. Thanks to a combination of molecular biology tools, protein engineering, enzyme immobilization and improved reactor design, many of these obstacles are being addressed, leading to the development of improved biocatalysts78. Specific examples with respect to P450 enzymes will be discussed in the following sections.

1.4.1 Xenobiotic-metabolizing P450 enzymes as biocatalysts

While on the one hand microbial P450s have a narrower substrate scope and are more substrate selective than human xenobiotic-metabolizing P450s, they are also easy to express in high yields in E. coli, are soluble (as opposed to membrane-bound), are highly active and have high coupling efficiencies. Thus, most successful P450 biocatalysts are microbial in origin. P450 BM3 is of particular interest because of its catalytic self- sufficiency and because it exhibits the highest turnover of all known P450-catalyzed

Chapter 1

-1 79 reactions (kcat of 17100 min ) with arachidonate . The usefulness of this P450 isoform was demonstrated for example with the enantioselective laboratory scale preparation of

14(S),15(R)-epoxyeicosatrienoic acid from arachidonic acid and (+)-leukotoxin B from linoleic acid using wild-type P450 BM3 and its F87V mutant80. In fact, the few examples of industrial uses of P450 enzymes are limited to oxidation of substrates by microbially expressed P450s in fermentation reactors81. In this way, hydrocortisone is produced via

82-83 the 11β-hydroxylation of cortexolone by P450lun expressed in Curvularia lunata at a scale of 100 tons per year15. Another example is the use of P450sca in Streptomyces carbophilus as part of a two-step fermentation process involving the hydroxylation of mevastatin to the cholesterol lowering drug pravastatin12, 84-85. Another statin, simvastatin, is converted to a clinically relevant metabolite 6β-hydroxymethyl simvastatin by an unknown P450 in Nocardia autotropica86. In another example, α,ω-dicarboxylic acids are produced from long chain alkanes by CYP52A1 in Candida tropicalis87. Finally, a

Sacharomyces cerevisiae strain has been engineered to produce artemisinic acid, a key precursor in the synthesis of the anti-malarial drug artemisinin, from simple sugars88. The engineered biosynthetic pathway includes a three-step hydroxylation by CYP71AV1, which, in this case, is a plant-derived P450 enzyme.

On the other hand, human P450s display high substrate promiscuity and recognize a much wider diversity of substrates compared to microbial P450s. They also perform a large variety of chemical transformations often with high chemo-, regio- and stereoselectivity. Thus, they and their mammalian homologues have a much broader scope of potential applications ranging from the production of drug metabolites, the synthesis of drugs and fine chemicals, to bioremediation and diagnostics. However,

Chapter 1

exploiting these attributes has proven difficult because of low activity, low stability, the requirement for expensive cofactors and poor product predictability. Yet, recognizing their tremendous potential for these applications, many researchers are hard at work to resolve these issues. Strategies that have been used to improve the biocatalytic properties of human CYPs will be discussed below.

1.4.1.1 Improving recombinant expression of P450s

In light of the current limitations hindering the use of P450 enzymes as biocatalysts, their use in whole cells is desirable since it can solve the problem of low stability and cofactor requirements. In fact, as seen in the previous section, all current industrial processes that use P450s as biocatalysts are performed in whole cells. Thus, improving recombinant expression of P450s is an important part of optimizing their biocatalytic properties. This has involved extensive optimization of expression plasmids89 as well as N-terminal modifications. The latter has been especially crucial in obtaining high recombinant expression yields of eukaryotic P450s in E. coli. In 1991, Barnes et al. described modifications to the first 7 codons in the coding sequence of bovine CYP17A that provided the first example of high level expression of a eukaryotic P450 in E. coli with expression levels reaching 16 mg per litre of culture90. These 7 codons were rationally modified in order to optimize mRNA recognition by E. coli ribosomes and minimize secondary structure formation at the 5’-end. Moreover, truncation of the N- terminal hydrophobic membrane anchor domain was found to be important in the expression of CYP2E191. These reports helped establish the minimum requirements to obtain high expression levels and were applied soon after to express many drug- metabolising P450s in E. coli. These include CYP3A492, CYP2D693, CYP2E194,

Chapter 1

CYP1A295, CYP2B696, CYP3A597 and CYPs from the 2C subfamily98-99. The expression of CYP2B6 was further improved by co-expression with chaperone proteins

GroES/EL100.

While N-terminal modifications were not deleterious to enzymatic activity,

Pritchard et al., in an attempt to improve expression while maintaining the wild-type sequences of these enzymes, developed a strategy involving N-terminal fusion to bacterial signal peptides101. In this previously established strategy, the leader peptides direct the growing polypeptide chain to the bacterial inner membrane and are then cleaved by endogenous proteolytic enzymes. Another advantage of this method is that no optimization of the 5’-end sequence is needed since E. coli ribosomes already recognize the leader sequences. This approach was found to be generally applicable to the expression of CYP3A4, CYP2E1 and CYP2A6. Cytochromes P450 have also been expressed in yeast, insect and mammalian cells. However, expression in E. coli is much more common because of its low cost, reliability and shorter incubation times.

1.4.1.2 The cofactor complication

Coexpression of CPR is necessary in order to sustain human P450 activity in E. coli cells. While some reductases are endogenously expressed by E. coli, they only support extremely low levels of human P450 activity102-103. Thus, co-expression with

CPR can be achieved by co-transformation of two separate cDNAs or by engineering a bicistronic plasmid that codes for both enzymes. Both approaches have proven useful in reconstituting the activity of many P450 isoforms in whole cells with rates comparable to those optimized in equivalent cell-free systems and without the need for external addition of the expensive cofactor NADPH (see Table 1.1). In some cases, metabolites were even

Chapter 1

isolated on a milligram scale. Being able to generate P450-derived drug metabolites is a valuable asset to the drug development process since it allows their characterization with respect to structural and toxicological properties.

Table 1.1 – Reports of P450-catalyzed reactions in whole cells supported by co-expression of CPR. Entry Isoform Reaction catalyzed Comment

1- CYP3A4 testosterone 6β-hydroxylation First examples of mammalian xenobiotic- nifedipine oxidation metabolising P450 activity in E. coli.104

2- CYP1A1 phenacetin O-deethylation Demonstrated general applicability of the 7-ethoxyresorufin O-deethylation co-expression strategy with six of the most CYP1A2 phenacetin O-deethylation important human drug-metabolising 7-ethoxyresorufin O-deethylation P450s.105 CYP2C9 tolbutamide methyl hydroxylation CYP2D6 bufuralol 1’-hydroxylation CYP2E1 chlorzoxazone 6-hydroxylation CYP3A4 testosterone 6β-hydroxylation

3- CYP1A1 7-ethoxyresorufin O-deethylation Similar to above but with some different 7-ethoxycoumarin O-deethylation substrates.106 CYP1A2 7-ethoxyresorufin O-deethylation 7-ethoxycoumarin O-deethylation CYP2A6 coumarin 7-hydroxylation CYP2C8 taxol 6-hydroxylation CYP2C9 tolbutamide methyl hydroxylation CYP2C19 (S)-mephenytoin 4’-hydroxylation CYP2D6 burfuralol 1’-hydroxylation CYP2E1 aniline 4-hydroxylation 4-nitrophenol hydroxylation CYP3A4 testosterone 6β-hydroxylation

4- CYP3A4 testosterone 6β-hydroxylation First purification of P450-generated CYP2C9 diclofenac 4’-hydroxylation metabolites on a milligram scale in a bioreactor. Used insect cells co-infected with separate cDNAs coding for P450 and CPR. Obtained 4’-hydroxydiclofenac in 28% yield (2.2 mg).107

5- CYP3A4 testosterone 6β-hydroxylation First preparative synthesis of drug CYP2C9 diclofenac 4’-hydroxylation metabolites in E. coli co-expressing P450s CYP1A2 phenacetin O-deethylation and CPR. Conversion yields ranged from 29-93%. Obtained 59 mg of 6β- hydroxytestosterone, 110 mg of 4’-hydroxy diclofenac and 88 mg of acetaminophen.108

Another effective way to simplify the electron transfer system required by P450s is by engineering fusion proteins with CPR109. As mentioned before, this type of configuration exists naturally with some P450 enzymes110, the most famous example

Chapter 1

being the bacterial P450 BM3. Eukaryotic examples include fungal enzymes CYP505A1 and CYP505B1. These have served as models for the design of artificial mammalian

P450-CPR fusions. Rat CYP1A1 was the first catalytically active example of such an artificial fusion111. Interestingly, a similar construct was later used in a light-driven bioreactor that coupled NADPH-production by chloroplasts to P450-mediated deethylation of the fluorogenic substrate 7-ethoxycoumarin to 7-hydroxycoumarin112.

Since then, there have been several examples of catalytically competent fusions of human CYPs with CPR or BMR (the reductase domain of P450 BM3). These include

CYP3A464, 113-116, CYP2D6117, CYP2C9115, CYP2C19115 and CYP2E1118. While most of these chimeras display rates that are comparable or slightly improved compared to their parent P450s, none reach even 10% of the turnover number of 17110 min-1 observed for arachidonate metabolism by wild-type P450 BM351. Evidently, having its P450 and reductase domains as part of the same polypeptide is not the only feature responsible for

BM3’s high catalytic activities. This enzyme is highly adapted to ensure proper orientation and flexibility between domains, a feature that was likely not achieved in the above-mentioned human CYP-reductase fusions. Nonetheless, these fusions could serve as helpful tools in the study of drug metabolism not only in cells but also in vitro.

While the use of whole cells has its advantages, isolated enzymes are more practical for certain applications, especially those that require immobilized enzyme.

Consequently, to address the cofactor problem in vitro, many have sought ways to replace

CPR and NADPH with more cost-effective alternatives. This has been achieved with various bacterial P450s through direct chemical or photochemical heme-reduction or by taking advantage of the peroxide shunt pathway with peroxide oxidants81, 109. In terms of

Chapter 1

human CYPs, our lab has reported the use of cumene hydroperoxide (CHP) and sodium percarbonate as effective cofactors surrogate for CYP3A4 and CYP2D6119. In fact, initial rate improvements of 210 and 130% compared to NADPH-supported reactions were achieved, respectively, without protein engineering. In a later study, Kumar et al. used a semi-rational approach to generate a CYP3A4 mutant (F228I/T309A) with 11-fold higher

CHP-driven catalytic efficiency (kcat/Km) than the wild-type enzyme with the substrate 7- benzyloxyquinoline120. However, CHP-supported activity was still ~ 10 times lower than if NADPH was used. Design of this mutant was based partly on the observation that the

T309V mutant of CYP2D6 displayed a 75-fold increase in activity using CHP compared to the wild-type enzyme121. Ultimately, this approach is not universally applicable since the ability of P450s to catalyze reactions via the peroxide shunt pathway varies between isoforms and between substrates for a given isoform.

A final strategy to do away with the tedious reconstitution of P450 catalytic activity in vitro has been through electrochemical reduction. As discussed in section

1.3.4, this strategy has proven difficult and has not yet been widely applied. Furthermore, it is still limited by the poor stability of human CYPs. However, further investigations are likely to be well rewarded, since this may well be the simplest and most cost-effective approach to sustain P450 activity in vitro for biocatalytic applications.

1.4.1.3 The stability problem

There are two types of stability to speak of when it comes to enzymes: thermodynamic and kinetic. Thermodynamic stability refers to the enzyme’s ability to maintain its integrity while at rest. Exposure to high temperatures, extreme pHs, organic solvents and long term storage above -80°C are usually not well tolerated. Kinetic

Chapter 1

stability refers to the enzyme’s ability to maintain its integrity during catalysis. Poor kinetic stability in P450 enzymes is thought to stem from their low coupling efficiencies.

As depicted in Figure 1.2, uncoupling generates reactive oxygen species that cause damage to the enzyme and/or its heme group during catalysis, resulting in enzyme inactivation. Thus, stability and activity are inextricably linked.

Purified P450 enzymes are typically stored in buffered solutions at -80°C.

Improving the storage stability of P450 enzymes would greatly facilitate their use in synthesis. To this end, our group has shown that it is possible to store CYP3A4 and

CYP2D6 in a dehydrated form without loss of activity when colyophilized with the disaccharides sucrose or trehalose respectively122. A subsequent study demonstrated that when dehydrated in this manner, CYP2D6 activity in water saturated n-decane was 125% higher than in buffer123. However, CYP3A4 only retained 11% of its activity under similar conditions124. This was not all that surprising since most enzymes125, including

P450s123, generally do not tolerate even low percentages of commonly used organic solvents (vide infra).

One of the advantages of using biocatalysts in synthesis is that they usually exhibit optimal activity under mild aqueous conditions, which are considered very “green”.

However, some reactions are not compatible with aqueous conditions and must be performed in organic solvents. Improved substrate and/or product solubility, stability and recovery are other advantages125. Our group has investigated the effects of other organic solvents on CYP3A4124 and CYP2D6123, 126. More specifically, detrimental effects on

CYP3A4 activity were observed with common water-miscible co-solvents present in levels as low as 1% and levels of 15% caused complete inactivation. While nearly

Chapter 1

anhydrous organic solvents were also quite detrimental to enzymatic activity, CYP3A4 retained 85% of its activity in a 50-50 hexanes-buffer biphasic system. In a similar system comprised of 2:1 isooctane:buffer, CYP2D6 reached 76% of its activity measured in buffer with the substrate dextromethophan. Moreover, it was able to debenzylate a new hydrophobic substrate 7-benzyloxy-4-N,N-diethylaminomethyl coumarin. Ultimately, best retention of CYP2D6 activity was observed in water saturated n-decane as mentioned in the previous paragraph.

Many strategies have been used to improve enzyme tolerance towards organic solvents including immobilization and protein engineering125. The only such study with mammalian P450s used CYP2B1, the rat homologue of human CYP2B6. It was engineered by directed evolution for improved tolerance to DMSO127. The mutant was

≥2-fold more tolerant with most DMSO concentrations tested. However, it still only retained ~35% of its activity when DMSO concentrations reached 30%.

While organic solvents can pose serious problems, high temperatures are generally not well tolerated either. Although protein engineering has proven successful in improving the thermal stability of many enzymes128-129, not much progress has been made with P450s in this respect, except for a few studies with the CYP2B family of mammalian homologues127, 130-131. Most recently, by comparing the sequences of two more stable homologues (rat CYP2B1 and rabbit CYP2B4) to two less stable ones (human CYP2B6 and canine CYP2B11), Kumar et al. identified seven residues that were conserved within the first pair but not the second131. They hypothesized that these could potentially play a role in stability. By mutating these residues in CYP2B6 to those found at equivalent

Chapter 1

positions in CYP2B1 or CYP2B4, they identified a mutation (P334S) that resulted in a

6°C increase in melting temperature without affecting kcat or Km.

1.4.1.4 The promiscuity paradox

Whereas on the one hand, the substrate promiscuity of xenobiotic-metabolizing cytochromes P450 may be their most attractive attribute when it comes to biocatalytic applications, it may also be their greatest drawback. This promiscuity is an essential adaptation with respect to their role in vivo. However, it means that their active sites are not specifically adapted towards a given substrate or groups of highly related substrates

(unlike most other enzymes). Thus, substrates generally do not fit perfectly into the active site and can often have multiple binding modes. This leads to uncontrolled water access and high uncoupling rates. In this way, promiscuity comes at the cost of enzymatic activity and product predictability. Consequently, human CYPs are not generally thought of as “ready-to-use” biocatalysts. Rather, as the reader has likely already realized, they need to be improved, optimized and tamed. Many have opted to use protein engineering approaches (random or rational) to improve enzymatic activity, change substrate specificity and alter product distributions14-15.

Our group on the other hand, has pioneered a substrate engineering strategy in an attempt to tackle concerns of product predictability. A proof of concept study with

CYP3A4 has demonstrated that predictable hydroxylations of small hydrocarbon-based substrates are possible when they are covalently linked to a suitable chemical auxiliary, in this case theobromine132. Larsen et al. found that theobromine was able to orient the substrate in the active site of CYP3A4 in such a way as to promote consistent hydroxylation at the fourth carbon from the auxiliary with Pro-R facial selectivity (Figure

Chapter 1

1.13). Regioselectivities for hydroxylations were >95% with most substrates while enantiomeric ratios were modest, reaching 75:25 at best. However, when a terminal double bond was placed between the third and fourth carbons from the auxiliary, a terminal epoxide was formed in an exceptional enantiomeric ratio of >99:1. Such enantioselectivity is especially challenging to attain synthetically with terminal epoxides and is best achieved through hydrolytic kinetic resolution133.

Figure 1.103 – Chemical auxiliary approach for improved product predictability with CYP3A4.

1.5 Research goals

The research detailed in this thesis is concerned with studying and improving the in vitro biocatalytic properties of human CYPs with respect to product predictability, activity and stability. These issues were investigated using several diverse strategies ranging from substrate and enzyme engineering to enzyme immobilization. First, the issue of product predictability was examined with CYP2E1 in Chapter 2. By way of a chemical auxiliary approach similar to that described above (section 1.3.1.4), the oxidation of small hydrocarbon substrates was achieved with improved yields and predictable regioselectivity. Here, type II ligands were used as chemical auxiliaries to target substrates to the P450 active site. Type II ligands coordinate to the P450 heme-

Chapter 1

irons via an aromatic nitrogen2. They are generally thought of as P450 inhibitors and have been used in drug development to inhibit P450 enzyme targets134-136. Recent reports suggesting that type II ligands can also be P450 substrates137-139 prompted us to investigate their utility as chemical auxiliaries for biocatalysis with this P450 isoform.

In their native environment, CYP3A4 and CYP2D6 are membrane-bound and surrounded by a large number of macromolecules. Many researchers have demonstrated that such crowded environments can have significant thermodynamic and kinetic effects on macromolecules140-141. These conditions are not well mimicked by the dilute aqueous buffered solutions used in most enzymology studies. Thus, a short study on the effect of macromolecular crowding on the activity of these two isoforms was conducted and is presented in Chapter 3.

We also became interested in the immobilization of CYP3A4 in hopes of opening up opportunities to study and make use of its biocatalytic potential in new ways. Thus, seeking methods that would retain or improve enzyme activity, a variety of immobilization methods were tested (two of which are described in this thesis) that would allow enzyme immobilization in different orientations. The first strategy, detailed in

Chapter 4, involved non-covalent C-terminal immobilization of CYP3A4 via its C- terminal His-tag. The second involved the covalent immobilization of a triple mutant of

CYP3A4 via a cysteine thiol group on its surface as described in Chapter 5. Both methods were assessed with respect to their effects on enzyme activity and stability.

Chapter 2

Chapter 2 – Type II binders as chemical auxiliaries for biocatalysis with CYP2E1

2.1 Preface

As explained in Chapter 1, cytochrome P450 enzymes are heme monooxygenases involved in the biosynthesis of secondary metabolites (e.g. steroids, lipids, vitamins) and play a central role in the metabolism of xenobiotics. In fact, approximately 75% of all known pharmaceuticals are metabolized by P450s5. Of note is their remarkable ability to oxidize inactivated C-H bonds with high regio- and stereoselectivities, a feat that is difficult to achieve by chemical means. Despite their tremendous potential to be used as biocatalysts14-15, 142, our ability to exploit these qualities in vitro has been hampered by the enzyme’s low stability, low activity, and poor product predictability, as well as the requirement for expensive cofactors.

Over the last several years, our lab has been addressing these issues with a special focus on CYP3A4119, 122, 124, 132, 143 because of its highly promiscuous behaviour towards substrates, which enhance its potential as a useful and versatile biocatalyst. Most recently, our lab has pioneered a substrate engineering strategy that has proven useful in taming the promiscuity of CYP3A4. In this type of approach, a chemical auxiliary is covalently linked to the substrate. The role of the chemical auxiliary is to orient the substrate in the

P450 active site in such a way as to promote its oxidation with consistent selectivity. The auxiliary is then removed to yield the desired product. In a proof of concept study by

Chapter 2

Larsen et al., it was found that when using theobromine as a chemical auxiliary, small hydrocarbon-based substrates were reliably hydroxylated at the fourth carbon from the auxiliary with Pro-R facial selectivity132. This was the first example of hydroxylation at an inactivated secondary C-H bond in which the catalyst was able to discriminate between methylene groups with similar electronic properties (see section 1.3.1.4). It was also an important step in improving product predictability of these enzymes, a significant hurdle preventing their use as biocatalysts.

In this chapter, a new chemical auxiliary for biocatalysis with CYP2E1 is described which has type II binding characteristics (vide infra). A small scope of auxiliary-substrates was synthesized in order to assess the role of the auxiliary. Then, the substrate portion was varied to investigate effects on yield and regioselectivity. Many of the CYP2E1 oxidation products were produced enzymatically in amounts large enough for characterization by NMR. More specifically, I designed and performed most of the experimental work with the help of two undergraduate students in the Auclair lab, Camilo

Fabra and Yue Huang. Camilo Fabra synthesized the auxiliary-substrates (2.4a, 2.9a,

2.10a and 2.15a-2.23a) as designed by myself. Yue Huang prepared and purified

CYP2E1-oxidation products 2.10b and 2.11b and performed a few activity assays with

2.1a. I also performed the computational studies with the guidance of Dr. Eric Therrien.

The results presented in this chapter were published as Ménard, M., Fabra, C., Huang, Y. and Auclair, K. (2012) Type II ligands as chemical auxiliaries to favor enzymatic transformations by P450 2E1. Chembiochem DOI: 10.1002/cbic.201200524.

Chapter 2

2.2 Introduction

2.2.1 Types of P450 ligands

Upon ligand binding, P450 enzymes exhibit interesting behaviours that can be monitored spectroscopically (Figure 2.1144). In their resting state, P450s have a water molecule coordinated to the active site heme-iron (6-coordinate iron). Interaction of the enzyme with ligands usually favours 5-coordinate iron with release of the water molecule.

Broadly speaking, ligands can lead to three types of UV spectral shifts of the Soret absorption band: type I, type II or reverse type I. Type I binders cause a shift in the iron spin equilibrium towards the high-spin species, resulting in an increase in absorbance around 390 nm and a decrease around 420 nm. Type II binders generally possess an aromatic nitrogen which coordinates to the heme-iron and stabilizes the low-spin state.

They cause a decrease in absorbance around 390 nm and an increase in absorbance around 425 nm. A reverse type I spectral shift is similar to type II except without a red shift of the Soret peak. It is often observed with ligands containing an oxygen atom that coordinates to the heme-iron2.

The type II interaction has often been called a dead-end complex and is believed to inhibit P450 catalysis in two ways2. The first way is by trapping the enzyme in the low- spin state, making heme-iron reduction by cytochrome P450 reductase (CPR) more difficult. The second way is by competitive exclusion of other potential ligands from the active site. Moreover, it has been demonstrated that type II ligands have a much higher affinity for P450 enzymes than similar type I binders137, 145-147.

Chapter 2

Figure 2.1 – Types of cytochrome P450 ligands. A. In its resting state, the heme-iron in the P450 active site is coordinated to four equatorial nitrogens (from the porphyrin ring depicted by a square), to a proximal cysteine thiolate and to a distal water molecule. B. Type I ligands bind near the heme in the P450 active site but do not coordinate to the iron. Note also that the distal water ligand is absent. C. Type II ligands (e.g. pyrazole) possess an aromatic nitrogen that coordinates to the heme-iron. D. Reverse type I ligands often contain an oxygen atom that is capable of coordination to the heme-iron. E. Comparison of typical type I and type II difference spectra of cytochrome P450s (i.e. the difference between the ligand-free and ligand bound absorbance spectra). Type I ligands produce a peak at around 390 nm and a trough at around 420 nm (black line). Type II binders produce a peak around 425 nm and a trough around 390 nm (red line). (Image E. taken from Segall144.)

2.2.2 Applications of type II ligands

These high affinity inhibitory interactions of type II binders have been exploited in the design of P450-targeted therapeutic agents by incorporation of nitrogen-containing aromatic functional groups such as imidazoles and triazoles. For example, this strategy has led to the successful development of inhibitors against the P450 enzyme lanosterol

Chapter 2

14-demethylase (CYP51A1) as antifungal agents134-135. These include miconazole and fluconazole which are currently in clinical use (Figure 2.2). Also, inhibitors of CYP17 in clinical use or in clinical trials for the treatment of prostate cancer all possess aromatic nitrogens capable of type II binding136 (Figure 2.3). Others have observed that in some cases, type II binding can also improve metabolic stability i.e. prevent transformation by

P450 enzymes in humans148, an important criterion in drug development.

Figure 2.2 – Chemical structures of antifungal azoles in clinical use.

Figure 2.3 – Chemical structures of CYP17 inhibitors in clinical use or clinical trials for the treatment of prostate cancer.

Chapter 2

On the other hand, many other drugs that have not been designed to inhibit P450s possess nitrogen-containing aromatic heterocycles. Of particular concern to drug developers is their potential to cause unwanted drug-drug interactions by inadvertent inhibition of P450 enzymes, resulting in the altered metabolism of co-administered drugs.

For instance, many azole antifungals have been found to inhibit several of the most important drug metabolizing P450s, and thus cause clinically relevant drug interactions2,

149-151. For example, the antihistamine cimetidine exhibits unwanted drug-drug interactions resulting from coordination of one imidazole nitrogen in its structure to the heme-iron of hepatic P450 enzymes152. Replacement of the imidazole group with a furan in the structurally related ranitidine abolishes these undesirable effects while maintaining its antihistaminic properties (Figure 2.4). Similarly, Sun et al. found that adding a methyl substituent at the 2-position on the imidazole group in N6022 (an S-nitrosoglutathione reductase inhibitor in clinical trials for the treatment of acute asthma, see Figure 2.4) reduced P450 inhibition, presumably by weakening its type II binding abilities through steric effects153.

Figure 2.4 – Chemical structures of antihistamines cimetidine and ranitidine and the S- nitrosoglutathione reductase inhibitor N6022.

Chapter 2

2.3 Objectives

While type II P450 ligands are generally inhibitors, further scrutiny has revealed that some are also substrates of these enzymes137-139, 154-155. A recent systematic study of pyridinyl quinoline-4-carboxamide-derived type II ligands of CYP3A4 by Peng et al.

(Figure 2.5) showed that type II binding can increase binding affinity by up to 1200- fold137. Moreover, while the type II ligands studied are competitive inhibitors of testosterone metabolism by CYP3A4, they are also extensively metabolized themselves, in some cases to a greater extent than similar type I binders. These results have inspired us to investigate the utility of type II ligands as chemical auxiliaries to target substrates to the active site of P450 enzymes. Although type II ligands have been introduced into molecules with the goal of generating P450 inhibitors, the focus of this report is to investigate the possibility of harnessing type II ligation to favour biocatalysis and afford oxidations with predictable regioselectivity.

Figure 2.5 – General structure of pyridine quinoline-4-carboxamides studied by Peng et al.137.

2.4 Results and discussion

Intrigued by the seemingly contradictory properties of P450 type II ligands, we wondered if they could be useful in the design of a chemical auxiliary to promote biocatalysis. Others in the Auclair lab have already described theobromine as an effective

Chapter 2

chemical auxiliary to improve the predictability of CYP3A4 oxidations132. Thus, we turned our attention towards another isoform, CYP2E1, in hopes of achieving complimentary regioselectivity. We envisaged taking advantage of type II ligation built into a chemical auxiliary to target substrates to the CYP2E1 active site and thus improve turnover. Moreover, the scarcity of examples of transformed P450 type II ligands clearly begged for more studies of these systems.

2.4.1 Selection of an appropriate chemical auxiliary

We were inspired to use nicotinate as a chemical auxiliary for CYP2E1 by one of its known substrates, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (more commonly known as nicotine-derived nitrosamine ketone or NKK)156, a nitrosamine present in tobacco and metabolized by CYP2E1 as shown in Figure 2.6. Moreover, both pyridine and methyl nicotinate are known type II binders of CYP2E1147. The hexyl ester of nicotinic acid (2.1a) was synthesized as the benchmark of this auxiliary-substrate series because of its similarity to NKK. As hoped, preliminary spectral binding studies indicate

Figure 2.6 – Oxidation of nicotine-derived nitrosamine ketone (NNK) to 4-hydroxy-1-(3-pyridyl)-1- butanone by CYP2E1. that 2.1a is a type II ligand of CYP2E1 (Figure 2.7A) with a Ks of 1.4 ± 0.1 µM (Figure

2.7B). We were also encouraged by LC-UV-MS results suggesting that 2.1a is transformed by CYP2E1 in the presence of its natural cofactors (CPR/NADPH) or the surrogate cofactor cumene hydroperoxide (CHP). MS analysis of the major product (later

Chapter 2

characterized as 2.1b) indicates the addition of 16 mass units compared to the starting material 2.1a (Figure 2.7C). Moreover, inspection of the fragmentation pattern confirms that oxidation occurred on the alkyl chain (the substrate portion) rather than on the aromatic ring (the auxiliary portion) (Figure 2.7D).

Figure 2 .7 – Spectral binding characteristics and biotransformation of 2.1a with CYP2E1. A) Difference binding spectrum of CYP2E1 with 2.1a displaying a peak at ~423 nm and a trough at ~390 nm characteristic of a type II ligand. B) Spectral binding curve of CYP2E1 titrated with 2.1a. The difference in absorbance at 423 and 390 nm was plotted against concentration of 2.1a yielding a spectral binding constant (Ks) of 1.4 ± 0.1 µM (fitting in GraphPad). C) UV-vis chromatograph obtained from the LC-MS analysis of the extracts obtained from reacting 2.1a with CYP2E1. Major and minor oxidation products were identified from the corresponding mass spectrum (data not shown). D) Example of the mass spectrum of the major product following CYP2E1 oxidation of 2.1a yielding 2.1b (m/z = 224.1). Fragment masses indicate that oxidation has taken place on the alkyl chain R rather than on the pyridine ring.

To study the effect of the chemical auxiliary, several other auxiliary-substrates with a hexyl group as the substrate part were prepared (see Table 2.1 for structures). Each auxiliary-substrate was characterized with respect to its binding mode and spectral dissociation constant (Ks) in the CYP2E1 complex. The latter was extracted from difference spectra titration curves. Also, the ratio and yield of oxidized products formed

Chapter 2

by reaction with CYP2E1 was estimated by LC-UV-MS analysis. Results are summarized in Table 2.1.

Table 2.1 – Characterization of auxiliary-substrates 2.1a-2.8a with respect to binding mode and spectral binding constant (Ks) for the corresponding CYP2E1 complex, as well as ratio of different regioisomeric products and yield of transformation by CYP2E1. The numbers in parentheses correspond to the standard error in the last digit ascribed by the curve fitting algorithm in GraphPad.

Product Compound Structure of R Binding mode K (µM)a Yield (%) s ratiob

2.1a type II 1.4 (0.1) 95:2:2:1 60

2.2a type I 120 (30) 50:35:15 30

2.3a type II 0.39 (0.02) 80:10:5:5 30

2.4a type I 1.6 (0.2) 95:5 20

2.5a type I >50 95:5 25

2.6ac type I 3.0 (0.4) >99:1 2

2.7ac type II 2.1 (0.1) >99:1 10

2.8a reverse type I 42 (7) - 15

aThe numbers in parentheses correspond to the standard error in the last digit ascribed by the curve-fitting algorithm in GraphPad. bProduct formation of <1% falls below the detection limit of our LC-MS. The ratio refers to the relative abundance of product regioisomers calculated from the areas under the product peaks in the UV chromatograms that were identified as having the appropriate molecular weight (that of the substrate+16 mass units) from the corresponding mass spectra. cOxidation products were not characterized.

Chapter 2

The relationships observed between binding mode and spectral binding affinity

137, 145-147 correlate well with other reports in the literature . As expected, Ks values of type I binders were in most cases at least an order of magnitude higher than those of type II binders except for compounds 2.4a and 2.6a. The presence of an aromatic nitrogen was necessary for type II binding to occur; 2.5a and 2.8a lacked such a nitrogen and exhibited type I and reverse type I behaviour respectively. On the other hand, not all structures containing an aromatic nitrogen exhibited type II binding characteristics as seen for compounds 2.2a, 2.4a and 2.6a. This has been attributed by others to steric interactions restricting access of the heme-iron for the nitrogen145, 147-148, 157-158. Also, changing the position of the nitrogen with respect to the ester group in the aromatic ring from meta

(2.1a) to para (2.3a) decreased the Ks value by an order of magnitude, indicating a higher affinity interaction. This is consistent with observations made by Peng et al. that para substituted quinoline-4-carboxamide analogs had a higher affinity towards CYP3A4 than their meta substituted counterparts137. Also in agreement with the results of this study is our observation that the ortho substituted auxiliary substrate 2.2a exhibited type I spectral binding characteristics, and thus was not able to coordinate to the heme-iron through its sp2 nitrogen. Others have also observed similar steric restrictions to heme-iron coordination caused by ortho substitution which also explains why compound 2.4a is not

137, 145, 148, 157-158 a type II binder . The low Ks value observed for 2.4a despite its type I binding character is likely due to the additional aromatic ring. Its presence probably increases the affinity of 2.4a towards the hydrophobic active site of CYP2E1 which is enclosed on the distal side by five phenylalanine residues159.

Interestingly, changing the linking group between the auxiliary and the substrate from an ester (2.1a) to an amide (2.6a) was accompanied by a change in binding mode

Chapter 2

(see Figure 2.8). In some cases, sterically unhindered meta substituted compounds have been reported to be type I binders instead of type II 137, 145. Retention of binding affinity can be explained at least in part by the fact that amide carbonyl oxygens are better hydrogen bond acceptors than their ester equivalents. Thus, 2.6a can form a stronger bond with the backbone N-H of Thr-303 which is at an appropriate distance of 2.8 Å.

Figure 2.8 – Spectral binding characteristics of 2.6a with CYP2E1. A) Difference binding spectrum of CYP2E1 with 2.6a displaying a peak at ~390 nm and a trough at ~423 nm characteristic of a type I ligand. B) Spectral binding curve of CYP2E1 titrated with 2.6a. The difference in absorbance at 390 and 423 nm was plotted against concentration of 2.6a yielding a spectral binding constant (Ks) of 3.0 ± 0.4 µM (fitting in GraphPad).

LC-UV-MS analysis revealed that CYP2E1 oxidation yielded more than one product for many of the auxiliary-substrates, yet in most cases one product was dominant.

Only products with molecular weights corresponding to that of the substrate + 16 mass units were observed, indicating that hydroxylation had taken place. The different products detected were typically regioisomers, and their relative abundance (reported as product ratio) was calculated from the areas under the product peaks in the UV chromatograms identified as having the appropriate molecular weight from the corresponding mass spectra. This is a reasonable measure of regioselectivity in as much as the substrate and corresponding products have similar extinction coefficients. This was in fact confirmed to be the case for 2.1a, 2.12a and 2.12b (see Experimental Protocols section 7.2.3.3). As

Chapter 2

seen in Table 2.1, there is no obvious relationship between binding mode and regioselectivity, and values of ≥95% can be achieved by both type I and type II binders.

While transformation of 2.6a and 2.7a by CYP2E1 occurred with excellent regioselectivity (within detection limits of the MS), product yields were quite poor. The nicotinate-based auxiliary-substrate 2.1a stood out as it was oxidized in very good yield

(60%) and excellent regioselectivity (95%).

The major CYP2E1 oxidation products for the reactions with higher yields (i.e. reactions of 2.1a-2.5a) were prepared biosynthetically, purified and characterized by

NMR and high resolution mass spectrometry. Results revealed that the major site of oxidation was always at the furthest methylene group from the auxiliary (Table 2.1), unaffected by the binding mode of the parent substrate. Comparison to an authentic standard via GC-MS confirmed this same preferred regioselectivity for oxidation of 1- hexanol (2.8a) to produce 1,5-hexanediol. Ultimately, nicotinate was retained as the most promising auxiliary in terms of yield and regioselectivity for subsequent studies.

2.4.2 Effects of varying substrate structures

Having selected nicotinate as our chemical auxiliary of choice, we proceeded to investigate its ability to direct the hydroxylation of various other hydrocarbon substrates.

Thus, a variety of nicotinate esters were synthesized from nicotinic acid and various alcohols containing branching (2.9a-2.11a), alkenes (cis and trans, 2.15a-2.19a), and cycles (saturated and unsaturated, 2.20a-2.23a). See Table 2.2 for structures.

Chapter 2

Table 2.2 – Characterization of auxiliary-substrates 2.9a-2.23a with respect to binding mode and spectral binding constant (Ks) for the corresponding CYP2E1 complex, as well as ratio of different regioisomeric products and yield of transformation by CYP2E1. The numbers in parentheses correspond to the standard error in the last digit ascribed by the curve fitting algorithm in GraphPad.

Binding Product Compound Structure of R K (µM) Yield (%) mode s ratioa

2.9a - - 50:50 15

2.10a type II 2.6 (0.2) 70:30 15

2.11a type II 1.8 (0.1) 85:5 25

2.12a type II 5.9 (0.4) 80:20 40

2.1a type II 1.4 (0.1) 95:2:2:1 60

2.13a - - 80:20 35

2.14a type II 2.1 (0.1) 95:5 40

2.15a - - 55:30:15 30

2.16a type II 1.3 (0.1) 40:40:10:10 70

2.17a - - 60:25:10:5 55

2.18a type II 12 (1) 90:10 20

2.19a - - 55:30:15 50

2.20a type II 1.6 (0.2) 70:25:5 20

2.21a - - 60:40 20

2.22a - - 75:25 15

2.23a - - 55:25:10:10 10

Chapter 2

All compounds were oxidized to a certain extent by CYP2E1, producing again at least two products, with one being predominant in most cases. Also, spectral binding studies with a subset of compounds (2.10a-2.12a, 2.14a, 2.16a, 2.18a, 2.20a; sampling each substrate type) confirmed their ability to bind CYP2E1 in a type II fashion with Ks values for their CYP2E1 complex similar to that of 2.1a. Unfortunately, in many cases yields were significantly lower than with 2.1a. Since commercial applications using P450 enzymes are currently limited to using whole cells81, and since this enzyme is naturally membrane-bound and less stable in vitro, it is reasonable to expect that yields can be improved with whole-cell transformations. Regioselectivities were lower with most substrates than with 2.1a, yet very good in most cases. Substrates that were transformed with the best regioselectivities, 2.10a-2.14a, 2.18a, 2.20a and 2.22a, were retained for characterization of the major product. Table 3 lists the corresponding product structures

2.10b-2.14b, 2.18b, 2.20b, and 2.22b. Because of the lower stability of 2.18b compared to the other products, an authentic standard of the suspected epoxide product was synthesized. Comparison of LC-MS profiles confirmed that this was in fact the correct product structure. Taken together, these results demonstrate a preference by CYP2E1 to oxidize the aliphatic or alkenyl secondary C-H bond furthest away from the auxiliary.

This is similar to the regioselectivity observed for CYP2E1 hydroxylation of some medium and long chain fatty acids160. These include saturated fatty acids lauric, myristic, palmitic and stearic acids which are hydroxylated at the ω-1 position.

Chapter 2

Table 2.3 – Structures of the major CYP2E1 oxidation products for auxiliary-substrates 2.10a-2.14a, 2.18a, 2.20a and 2.22a.

Substrate Structure of R Product Structure of R’

2.10a 10b

2.11a 11b

2.12a 12b

2.13a 13b

2.14a 14b

2.18a 18b

2.20a 20b

2.22a 22b

2.4.3 Molecular modeling studies

To gain insight into the origins of the regioselectivity observed with respect to

CYP2E1 oxidation of these auxiliary-substrates, we turned our attention towards molecular modeling. Docking of molecules 2.1a-2.7a and 2.9a-2.23a was performed using FITTED via the Web-based platform FORECASTER, both developed by N. Moitessier

161 and his research group . Preliminary docking with FITTED in flexible protein mode

Chapter 2

allowed us to select the most appropriate X-ray crystal structure. PDB codes 3E4E

(CYP2E1 in complex with 4-methylpyrrazole) and 3LC4 (CYP2E1 in complex with omega-imidazolyl-dodecanoic acid, a fatty acid mimic) were compared for their ability to accommodate 2.1a-2.7a and 2.9a-2.23a in the active site. These two structures differ in active site volume (190 vs. 440 Å respectively). In 3LC4, Phe-298 is rotated toward the

B’ helix, opening up the active site to a void above the I helix that is closed off in

3E4E162. Perhaps not surprisingly, 3LC4 was identified as the better candidate for docking of our compounds. This is most likely due to their elongated shape, somewhat resembling that of a fatty acid. A protein RMSD value of 0 was obtained for 3LC4 with all molecules except 2.6a, indicating that the favoured structure for docking these molecules corresponds to that of 3LC4. (Alternatively, an RMSD value greater than 0 for both PDB files would indicate that the favoured structure combines features from both parent input files). Thus, 3LC4 was selected and used for rigid docking studies. All default docking parameters were used.

When docked into 3LC4 in rigid protein mode, each compound (2.1a-2.7a and

2.9a-2.23a) adopted several binding poses of similar energy. Thus, it was not possible to distinguish between them to identify which was favoured. However, careful inspection of the docking results revealed that molecules 2.1a-2.7a and 2.9a-2.23a clustered in three separate groupings (Clusters A, B and C, shown in Figure 2.9A, 9B, and 9C respectively) with respect to the proximal face of the heme prosthetic group in the active site of

CYP2E1. In the first two groupings (Clusters A and B), the auxiliary portion binds almost directly above the porphyrin ring at two discrete distances with the substrate portion directed towards the heme-iron. Another orientation and third grouping (Cluster C) was

Chapter 2

also observed in which the docked molecules were oriented with the auxiliary portion closest to the heme in a manner akin to other known type II binders.

Figure 2.9 – Molecular modeling of compounds 2.1a and 2.9-2.23a with CYP2E1 (PDB 3LC4). A) Cluster A: overlap of compounds 2.1a, 2.9a, 2.10a, 2.13a, 2.14a, 2.16a-2.18a, 2.22a and 2.23a (purple) with the auxiliary moiety docking closer to the G helix and further from the heme (black). B) Cluster B: overlap of compounds 2.2a, 2.10a, 2.12a, 2.15a, 2.16a and 2.18a-2.20a (purple) with the auxiliary moiety binding further from the G helix and closer to the heme (black). C) Cluster C: overlap of compounds 2.1a and 2.9a-2.22a (purple) docking in a type II mode with their aromatic nitrogen atom ~3.1 Å from the heme-iron (green sphere). D) Overlap of 2.18a poses in Cluster A (purple) and Cluster B (orange). Amino acids within 4 Å of either structure are highlighted and labeled.

Chapter 2

Compounds containing different auxiliaries with the same substrate (as in 2.1a-

2.7a) yielded binding poses of similar energy that grouped well into Cluster A and/or B, and C. As expected, binding poses of most type I binders (2.2a, 2.4a and 2.5a) belonging to Cluster C have their aromatic nitrogen further and pointed away from the heme-iron.

Interestingly, this was not the case for compound 2.6a whose Cluster C binding pose suggests type II binding.

The 3 clusters are especially useful to explain our results when different substrates with the same auxiliary (nicotinate) are compared. Binding poses for 2.1a, 2.9a, 2.10a,

2.13a, 2.14a, 2.16a-2.18a, 2.22a and 2.23a with the auxiliary portion binding further away (Cluster A) from the heme are shown overlapped in Figure 2.9A. A similar overlap of molecules 2.10a, 2.12a, 2.15a, 2.16a and 2.18a-2.20a binding closer to the heme

(Cluster B) is rendered in Figure 2.9B. In general, longer molecules were found in Cluster

A and shorter ones in Cluster B. It is interesting to note that some molecules (2.10a, 2.16a and 2.18a) were found to have binding poses of similar energy in both clusters (Figure

2.9D). This suggests that shorter auxiliary-substrates can navigate between both binding sites, allowing them to be oxidized while residing closer to the heme (as in Cluster B).

For the observed binding mode predicted to give the major product (belonging either to

Cluster A or B), distances from the heme-iron to the closest non-aromatic secondary C-H bond (Fe-H distance) range from 2.70-7.51 Å, with an average distance of 4.50 Å. Such distances indicate that these binding poses could correspond to the true productive binding mode for these compounds within the active site of CYP2E1. In fact, only four distances (2.5a, 2.7a, 2.9a, and 2.22a from Cluster A) fell outside of the 6 Å cutoff suggested by Tarcsay et al. to distinguish between productive and non-productive binding modes during computational studies with CYP2C9163. Similarly, Prasad et al. used a

Chapter 2

cutoff distance of 4 Å from the substrate to the iron-oxo with an Fe-O bond length of 1.9

Å (a total of ~6 Å) in their modeling studies with CYP1A homologues164. Since those molecules that fell outside the 6 Å range were also transformed by CYP2E1, it is likely that experimentally, they can adopt a binding mode similar to other molecules in Cluster

B (even though it was not observed computationally). Thus, CYP2E1 simply oxidizes the

C-H bond with the lowest bond dissociation energy in its vicinity. Greater preference is given to the oxidation of more electron rich secondary and tertiary non-aromatic C-H bonds at the sub-terminal position rather than the terminal primary position.

Finally, binding poses for nicotinate-derived substrates 2.1a and 2.9a-2.22a in

Cluster C are shown overlapped in Figure 2.9C. The average distance between the heme- iron and the aromatic nitrogen of molecules in this Cluster was 3.1 Å. This is the same as was observed for the known type II binding indazole when co-crystallized with

CYP2E1159. Other type II binders 2.3a and 2.7a also adopted a similar pose (data not shown). Taken together, our molecular modeling and experimental results are in good agreement and help explain how our compounds can be both type II binders and substrates of CYP2E1 by adopting multiple binding modes within the active site.

2.4.4 Necessity of using the chemical auxiliary

In order to confirm the importance of the nicotinate auxiliary, control enzymatic reactions were performed with three substrates without the chemical auxiliary portion.

Thus, 1-hexanol (2.8a), the substrate part of molecules 2.1a-2.7a, as well as 3-methyl-1- pentanol, the substrate part of 2.10a, and cyclopentylmethanol, the substrate part of 2.20a, were treated with CYP2E1 under the same conditions as above. Reaction mixtures were analyzed by GC-MS. As seen in Table 2.1, 2.8a is oxidized by CYP2E1 but is a poorer

Chapter 2

substrate than compounds 2.1a-2.5a (15% yield observed for the transformation of 2.8a compared to 60% for that of 2.1a). On the other hand, neither 3-methyl-1-pentanol nor cyclopentylmethanol were transformed by CYP2E1 at a detectable level (<1%). These results support our hypothesis that type II ligation can be used to favour biocatalysis by

P450 enzymes.

2.4.5 Nicotinate as a chemical auxiliary

Ultimately, the type II binder nicotinate, when used as a chemical auxiliary was found to improve the yield of 1-hexanol (2.1a vs. 2.8a, Table 2.1) oxidation by CYP2E1 without changing the major site of oxidation. Moreover, covalent binding of nicotinate made possible the transformation of some small alcohols that are normally not substrates of this enzyme. Not all type II binding auxiliaries had the same effect and optimization is likely necessary with each P450 enzyme. It is possible that some ligands favour the non- productive type II orientation within the P450 active site to a greater extent than nicotinate. Thus, 2.1a exhibits a balance between type II binding which increases its affinity for CYP2E1 and productive binding that leads to better oxidation yields.

Interestingly, in all cases the major product (Tables 2.1 and 2.2) is the result of preferential oxidation at the secondary or tertiary aliphatic or alkenyl C-H furthest from the auxiliary. This type of regioselectivity is in contrast to what was observed by Larsen et al. with CYP3A4 and the chemical auxiliary theobromine132. In this case, the site of oxidation always occurs at a set distance from the auxiliary (namely the fourth secondary

C-H bond) regardless of substrate length. Nicotinate’s lack of influence on the major site of oxidation is likely due to the topology of the CYP2E1 active site and the resemblance of our substrate auxiliaries to short fatty acids. Consequently, the auxiliary can bind at

Chapter 2

more than one discrete distance from the heme as was observed with fatty acid analogues of different lengths co-crystallized with CYP2E162. This leads to an oxidation regioselectivity that is controlled by electronics and mobility and not by the distance of a given C-H bond from the chemical auxiliary (as was observed with CYP3A4 and theobromine132). Differences in product ratios may perhaps be explained by differences in the mobility of various auxiliary-substrates within the active site of CYP2E1. In other words, more mobile auxiliary-substrates would expose C-H bonds with similar bond dissociation energies within a productive range from the heme-iron in a manner that would result in an increased number or proportion of minor oxidation products.

Finally, using nicotinate as a chemical auxiliary has several practical advantages.

It is achiral, commercially available and inexpensive. Moreover, it contains one easily derivatized group, the carboxylate, and is therefore readily linked to various substrates.

Importantly, this auxiliary also contains a chromophore for easy detection and is easily cleaved off after enzymatic reactions (see Experimental Protocols section 7.2.3.5), with the generation of an alcohol, or other functional groups depending on the conditions used.

Additionally, many nicotinate derivatives are known to cross cell membranes, which is an important consideration for whole-cell transformations.

2.5 Conclusion

In this chapter, we show that linking the type II binding chemical auxiliary nicotinate to a variety of hydrocarbon substrates has been successful at promoting

CYP2E1 biocatalysis with predictable regioselectivity. CYP2E1’s preferential oxidation of the secondary or tertiary, non-aromatic C-H bond furthest from the auxiliary was also

Chapter 2

rationalized with the help of computational docking studies. FITTED was able to predict both the type II and the productive binding modes. This study not only confirms the recently reported oxidation of type II ligands by yet another P450 enzyme but also demonstrates for the first time that type II ligands may find use in targeting various substrates to the P450 active site for biocatalytic applications, as opposed to the traditional goal of inhibiting transformation. We expect that this concept might find use with other chemical auxiliaries and with other P450 isoforms to potentially afford complementary products, after identification of the optimal type II ligands.

Chapter 3

Chapter 3 – Effects of macromolecular crowding on the enzymatic activity of CYP3A4 and CYP2D6

3.1 Preface

One of the main drawbacks of using P450 enzymes as in vitro biocatalysts is their poor activity. In cells, these enzymes are associated with the outer membrane of the endoplasmic reticulum and surrounded by a dense soup of macromolecules. In fact, it is estimated that macromolecules can occupy 5-40% of the total volume inside a cell165 with typical values ranging from 20-30%166. For example, the total concentration of protein and RNA in E. coli cells ranges from 300 to 400 g/L167. This type of environment is referred to as “crowded” rather than concentrated because it is comprised of a large number of hugely varied macromolecules each present at low concentrations. Enzymes are adapted to function optimally under such conditions; conditions that are poorly mimicked by the simple “uncrowded” buffered solutions that are typically used in in vitro studies. It is for this reason that a growing appreciation for macromolecular crowding is emerging and the profound effects that it can have on a variety of biochemical and biophysical events are being uncovered141.

In light of these facts, we became interested in how P450 activity would be impacted if measured in a highly crowded environment compared to buffer. We hypothesized that perhaps, by mimicking the highly crowded intracellular environment encountered in vivo, higher in vitro activities could be achieved. Therefore, Jessica

Chapter 3

Douglas (an undergraduate honours student in Prof. Auclair’s lab) and I conducted a short study to investigate the effects of a variety of macromolecular crowding agents on two

P450 isoforms: CYP3A4 and CYP2D6. More specifically, I performed most of the experiment planning, assay optimization and data analysis. Data collection was for the most part performed by Jessica Douglas. The results presented in this chapter are not published.

3.2 Introduction

Macromolecular crowding can also be referred to as the “excluded volume effect”.

In a sense, this term is more accurate since it better describes the origins of its effects: the exclusion of volume by one solute to another within a solution of a given volume141. This occurs as a result of nonspecific steric repulsions that are always present between individual solutes. However, the magnitude of this effect is highly dependent on the size of the solutes and is most prominent when large molecules of similar size occupy a significant fraction of a given volume. Take for example, as illustrated in Figure 3.1A, a box (or a cell) with 30% of its volume occupied by black spheres (macromolecules).

There is virtually no excluded volume effect to introducing a small molecule into this cell since it has access to all of the remaining available volume (yellow area). However, when a macromolecule of similar size is introduced, the excluded volume effect becomes obvious (Figure 3.1B). In other words, the available volume becomes much less than 70% because the centre of the macromolecule being introduced cannot approach the others by less than the distance delineated by the open circles. This of course is the distance at which the surfaces of two macromolecules meet, assuming their mutual impenetrability.

Chapter 3

Figure 3.1 – Illustration of the excluded volume effect. The box and black spheres represent a cell with 30% of its total volume occupied by mutually impenetrable macromolecules. A) There is virtually no excluded volume effect to introducing a small molecule into this cell since it has access to all of the remaining available volume (i.e. 70%) shown in yellow. B) The excluded volume becomes obvious rather, when a macromolecule of similar size is introduced. In this case, the available volume becomes much less than 70% because the centre of the macromolecule being introduced cannot approach the others by less than the distance delineated by the open circles: the distance at which the surfaces of two macromolecules meet (indicated for one pair of molecules by a red arrow). (Image adapted from Ellis166)

Macromolecular crowding effects can have both thermodynamic and kinetic repercussion on intermolecular interactions166. The thermodynamic component affects the equilibrium constant of macromolecular interactions (Figure 3.2A). In theory, the effect on the association constant of two associating biomolecules can either be positive or a negative depending on the size and shape of the resulting dimer. As seen in an illustration borrowed from a review by Zhou et al.141 (Figure 3.2B), for a constant dimer volume, the association constant in the crowded medium ( ) increases with the fraction of the total

volume occupied by macromolecules (Φ) relative to that in dilute conditions ( ) for more spherically shaped dimers. However, as the dimer shape becomes more oblong, the excluded volume effect is such that the association constant is negatively affected as the fractional volume occupied increases141. This is because more oblong dimers exclude more volume than two individual monomers. Despite these conflicting theoretical

Chapter 3

outcomes, the main effect of macromolecular crowding observed experimentally is typically to promote associations between macromolecules166.

Marcromolecular crowding can also have kinetic effects on the rate of intermolecular associations. Here again, the effect can be positive or negative depending this time on whether the association is diffusion-controlled or transition state-controlled.

Take a simple case of the bimolecular association of two monomers (A + B) to form a dimer (AB) with a transition state (AB*). The reaction is transition state-controlled if the rate is limited by conformational changes that must occur in the transition state before successful dimer formation can occur. These reactions are relatively slow. On the other hand, fast reactions are diffusion-controlled and are only limited by the diffusion rates of the reactants. Since macromolecular crowders tend to promote macromolecular associations but decrease diffusion rates, they are expected to accelerate transition state- controlled reaction rates and decelerate diffusion-controlled ones (Figure 3.3).166

It follows that macromolecular crowding can also have intramolecular effects.

Thermodynamically speaking, crowding will favour more compact macromolecular conformations thus shifting conformational equilibria towards conformations that occupy less volume. This could have a negative effect on macromolecular interactions if the optimal orientation for maximum interaction cannot be achieved due to conformational restrictions. In terms of kinetic effects, macromolecular crowding should decelerate diffusion-controlled conformational changes such as those that involve the movement of solvent exposed loops. On the other hand, macromolecular crowding can accelerate conformational changes if the transition state is more compact (lower in energy) under crowded conditions. Thus, macromolecular crowding is generally expected to accelerate protein folding and have a stabilizing effect against temperature induced denaturation.

Chapter 3

z A)

z B)

Figure 3.2 – A) Thermodynamic cycle showing the relationship between the free energy of transfer of monomers and dimers from a dilute solution to a crowded medium ( and the free energy of association ( , where R is the molar gas constant and T is the absolute temperature. B) Influence of volume fraction of crowding agent (Φ) and dimer shape at constant dimer volume on the equilibrium constant KAB for the formation of a dimer from two monomers with radius = R1 = 1. Plotted are the solutions to an equation derived from scaled particle theory which calculates the free energy required to create a cavity in a solution of hard spheres with a) L2 = 0, R2 = 1.26; b) L2 = 2/3, R2 = 1; c) L2 = 1, R2 = 0.928, d) L2 = 1.5, R2 = 0.851; and e) L2 = 2, R2 = 0.794. (Images adapted from Zhou et al.141)

Chapter 3

Figure 3.3 – Relationship between the reaction rate constant of a bimolecular association with respect to the concentration of crowding agent for diffusion-controlled (green) and transition state-controlled (blue) regimes. In practice, reactions are transition state-controlled at low crowder concentrations but become diffusion-controlled at higher concentrations (red). (Image taken from Ellis166)

One might easily envisage that the effects of macromolecular crowding discussed above can potentially influence enzymatic reactions. While positive effects on reaction rates are theoretically possible and have been observed experimentally, they are not guaranteed. Not only do the thermodynamic and kinetic parameters of the enzyme system come into play, but factors such as the nature and concentration of crowding agent can also influence the net outcome of crowders on enzymatic activity. This is because, in practice, macromolecules are not inert spheres; they can interact with each other via multiple weak non-covalent interactions that, taken together, can affect protein stability and diffusion168-169. For example, when studying multimeric decarboxylating enzymes

(urease, pyruvate decarboxylase and glutamate decarboxylase), Derham and Harding found that the effect of a crowded environment on reaction rates depended greatly on the nature of the crowding agent used170. With increasing concentrations (0-60% w/v) of globular protein crowding agents (haemoglobin and lysozyme) reaction rates increased at

Chapter 3

first and then decreased. With polymeric crowding agents (polyethylene glycol (PEG) 8K and the glucose polymer dextran 10, 70, and 120K) however, reaction rates decreased at all concentrations. Authors attribute this discrepancy to the elongated shape of the polymers that result in an increase in viscosity, and disproportionate impedance of diffusion compared to globular shapes. Dextran and Ficoll® (a non-ionic, highly branched polymer of sucrose) of various molecular weights also reduced the activity of alkaline phosphatase in a molecular-weight and concentration dependent manner (0-30% w/v)171. In this case, Ficoll was less detrimental than dextran, possibly due to its more compact nature compared to dextran.

These observations cannot be generalized to all systems however. For example, unlike with the decarboxylases, others have observed that the addition of dextran and

Ficoll produced non-linear effects on the enzymatic activity of yeast multi-copper oxidase

Fet3p172. Moreover, the presence of polymeric crowder PEG 6K had a positive effect on the enzymatic activity of adenosine diphosphate (ADP)-sugar pyrophosphatase (AspP)173.

Addition of PEG 6K and other polymeric crowders such as dextran 70K and Ficoll 70K also enhanced the enzymatic activity isochorismate synthase (EntC)174. This effect was attributed to a conformational change promoted by increasing crowder concentration

(observed by circular dichroism with Ficoll 70K) which correlated with a decrease in Km.

Moreover, bovine serum albumin (BSA) had a net negative effect on the catalytic

175 efficiency (kcat/Km) of yeast hexokinase . While increasing concentrations of BSA were accompanied by a decrease in Km, they also caused a disproportionate decrease in kcat.

Chapter 3

3.3 Objectives

Intrigued by the possibility of improving enzymatic activity by macromolecular crowding, Jessica Douglas (an undergraduate honours student in the Auclair lab) and I embarked on a short study to investigate the effects of a variety of macromolecular crowding agents on the activity of two P450 isoforms: CYP3A4 and CYP2D6. Crowding agents commonly used in such experiments were selected such as to cover a wide range of sizes, and chemical compositions (Figure 3.4). These included polysaccharides dextran

70K (DEX 70), dextran 500K (DEX 500) and Ficoll® 70K (FIC 70), the non-biologically derived polymers polyethylene glycol 6K (PEG 6) and polyvinylpyrrolidone 40K (PVP

40) as well as the globular protein BSA. Small sugars sucrose, trehalose, glucose and melezitose were also included as controls having similar polarity to the crowders without exhibiting volume exclusion effects (Figure 3.5).

Chapter 3

C) B)

Figure 3.4 – Chemical composition of crowding agents. A) Dextran: a branched polymer of glucose isolated from Leuconostoc spp. B) Ficoll®: a highly branched copolymer of sucrose and epichlorohydrin drawn as the individual monomers because the exact structure is not disclosed. C) Polyethylene glycol (PEG). D) Polyvinylpyrrolidone (PVP).

Figure 3.5 – Structures of small sugars: sucrose (A), glucose (B), trehalose (C) and melezitose (D).

Chapter 3

3.4 Activity assays

Initially, the use of fluorogenic substrates was investigated to assay the activity of

P450 enzymes under crowded conditions. These types of substrates are advantageous because of their high sensitivity (low product concentrations can easily be detected) and the ease with which reaction rates can be measured by directly monitoring product formation over time in a high throughput manner using a fluorescence plate reader.

Unfortunately, many of the crowders were found to interfere with the fluorescence intensity of the products to be observed in a concentration-dependent manner, making this approach to measuring initial rates impractical (data not shown). We therefore opted to measure the end-point activities of CYP3A4 and CYP2D6 with standard substrates testosterone and dextromethorphan respectively (Figure 3.6). This type of approach involves measuring the total amount of product formed by the enzyme as a means of estimating and comparing enzymatic activity under differing conditions. Such a strategy is possible with P450s since conversion to product rarely goes to completion on the account of their intrinsically high uncoupling rates that lead to enzyme inactivation by reactive oxygen species.

Chapter 3

Figure 3.6 – Substrates used to measure the activity of CYP3A4 and CYP2D6 for end-point type assays in the presence of macromolecular crowding agents. A) 6β-hydroxylation of testosterone by CYP3A4. B) O-demethylation of dextromethorphan by CYP2D6.

3.5 Results

3.5.1 Effects of macromolecular crowding agents on CYP3A4 activity

The activity of CYP3A4 was first assayed in the presence of 20% w/v (g/mL) of crowding agents. Small sugar molecules were also tested to control for polarity effects.

As seen in Figure 3.7, enzymatic activity was reduced to ≤50% by the addition of all crowding agents tested when CPR and NADPH were used as the cofactors. Except with

PVP 40, loss of activity was equal or less pronounced when the surrogate cofactor CHP was used instead of CPR and NADPH. To our surprise, addition of sucrose or trehalose resulted in a significant increase of CYP3A4 activity with CHP by 140 ± 10% and 150 ±

20% respectively. On the other hand, this enhancement in activity was absent with the addition of glucose (110 ± 10%) and reversed with melezitose (33 ± 3%). Interestingly,

Chapter 3

with CPR and NADPH as the cofactors, activity in the presence of sucrose was similar to that with most crowders (52 ± 3%).

Macromolecular crowders Small sugar controls 180 160

140

120

100

80 60

Activity ofcontrol) (%Activity 40 20 0

Figure 3.7 – Effect of the addition of macromolecular crowding agents or small sugars (20% w/v) on the activity of CYP3A4 with the substrate testosterone. Percent activity is reported relative to that observed in the absence of additive with CPR and NADPH (blue) or CHP (red) as the average of duplicates or triplicates ± one standard deviation. Note: P450 activity was not tested in the presence of trehalose, glucose and melezitose with CPR and NADPH.

Intrigued by the positive effect of certain small sugars, the possibility of a concentration-dependence was investigated (Figure 3.8). CYP3A4 activity with CHP improved with increasing concentrations of sucrose and trehalose with maximal activities of 164 ± 4% (at 40% w/v) and 167 ± 3% (at 30% w/v) respectively. Slight improvements in activity (~10%) were observed in the presence of glucose at 10, 20 and 30% w/v. As for melezitose, it caused a decrease in CYP3A4 activity of ≥60% at all concentrations tested.

Chapter 3

Figure 3.8 – Effect of varying small sugar concentrations on activity of CYP3A4 with the substrate testosterone and CHP as the cofactor surrogate. Percent activity is reported relative to that observed in the absence of small sugar sucrose (blue), trehalose (red), glucose (green) or melezitose (purple) with CHP as the average of triplicates ± one standard deviation.

3.5.2 Effects of macromolecular crowding agents on CYP2D6 activity

As with CYP3A4, the effect of crowders at a concentration of 20% w/v on

CYP2D6 activity was investigated with a focus on CHP as the cofactor. As seen in Figure

3.9, the addition of PVP 40, BSA and DEX 500 reduced enzymatic activity to ≤20%. On the other hand, addition of PEG 6, FIC 70 and DEX 70 produced no significant effect. As for the small sugar controls, addition of sucrose caused no significant effect, trehalose and melezitose decreased enzymatic activity by ~40% and addition of glucose improved activity by 130 ± 20%. In general, as with CYP3A4, when CPR and NADPH were used instead of CHP, CYP2D6 activity was equal or less in the presence of a given crowding agent or small sugar control (data not shown).

Chapter 3

Small sugar controls 160 Macromolecular crowders

140

120

100

Activity ofcontrol) (%Activity 40 20

Figure 3.9 – Effect of the addition of macromolecular crowders or small sugars (20% w/v) on the activity of CYP2D6 with the substrate dextromethorphan. Percent activity is reported relative to that observed in the absence of additive with the cofactor CHP as the average of ≥3 replicates ± one standard deviation.

Further investigation of the effect of adding small sugars to the enzymatic reactions once more revealed a concentration-dependence in certain cases (Figure 3.10).

While at the original sugar concentration of 20% w/v, trehalose had a detrimental effect on the activity of CYP2D6, increasing trehalose concentration resulted in a linear increase in activity peaking at 113 ± 9% with 40% w/v trehalose. The presence of sucrose had no significant effect at most concentrations tested but a slight increase was observed at 50% w/v (122 ± 9%). Melezitose consistently reduced CYP2D6’s activity by about 40%.

Finally, optimal CYP2D6 activity in the presence of glucose was observed at a concentration of 20% w/v (130 ± 20%) with further increases in concentration eventually leading to decreased activity.

Chapter 3

Figure 3.1 – Effect of varying small sugar concentrations on the activity of CYP2D6 with the substrate dextromethorphan and CHP as the cofactor surrogate. Percent activity is reported relative to that observed in the absence of small sugar sucrose (A), glucose (B), trehalose (C) or melezitose (D) with CHP as the average of duplicates or triplicates ± one standard deviation.

3.6 Discussion

3.6.1 Differential effect of cofactors: CPR and NADPH vs. CHP

In general, the activity of both CYPs was better under crowded conditions when

CHP was used as the cofactor instead of CPR and NADPH (Figures 3.7 and 3.9). In order for proper delivery of electrons to occur, CYP and CPR must associate. Therefore, this observation suggests that, at this crowder concentration of 20% w/v, the association was diffusion-limited. On the other hand, because of its small size (152 g/mol), CHP was able to diffuse freely despite the highly crowded environment and reach the P450 enzyme relatively unimpeded to feed into its catalytic cycle.

Also likely contributing to the lower rates observed with CPR, are the large domain motions involved with its reduction by NADPH followed by electron transfer to

Chapter 3

the P450176. NMR and small angle x-ray scattering experiments performed by Ellis et al. revealed that, in solution, CPR can adopt two distinct conformations: open or closed

(Figure 3.21)177. When CPR is oxidized, there is a 1:1 ratio of these two states. The equilibrium shifts to 85:15 in favour of the closed form with NADP+-bound, 4-electron reduced CPR. This corresponds to the hydroquinone form of CPR where both FAD and

FMN are reduced to FADH2 and FMNH2 respectively.

Figure 3 .2 – Model of the “open-closed” equilibrium in CPR. CPR is shown here associated to a lipid bilayer through its N-terminus. Its FMN (blue) and FAD (green) domains are linked through a hinge domain (magenta). Directional electron transfer from NADPH (blue sticks) to FAD (orange sticks) to FMN (yellow spheres) is facilitated by large scale domain reorientations. In its resting oxidized state, CPR adopts a 1:1 equilibrium of these two conformations. NADPH binding causes a shift towards the closed state (right). (Image taken from Ellis et al.177)

More detailed observations of these redox-linked conformational changes in CPR have also been observed through FRET studies coupled to stopped-flow kinetic electron transfer experiments178. In these investigations, NADPH binding was also observed to favour the closed conformation. However, upon 2- and 4-electron reduction of the enzyme, a transient shift to the open conformation was observed, followed by a return to the closed form. (Two-electron reduction leads to the semiquinone form of CPR). Thus

Chapter 3

upon NADPH binding, the FAD and FMN domains come into closer proximity for more efficient intramolecular electron transfer followed by a conformational opening which exposes the FMN domain again, allowing it to come into contact with the P450 enzyme.

It is possible that, in the presence of molecular crowding agents, these large scale domain motions in CPR were impeded. In fact, there is evidence that the rate of intramolecular electron transfer from FAD to FMN is limited by this conformational change. Particularly interesting is the observation that increasing the viscosity of the environment with glycerol decreased this electron transfer rate179. Due to similarities in rates calculated for this conformation change and that observed for the first electron transfer from CPR to CYP3A4, some have even suggested that CPR domain reorientations may contribute to the rate limiting step of this process180. Thus, conformational changes in CPR were highly likely diffusion-limited under the crowded environments tested here, resulting in lower CYP3A4 and CYP2D6 activity with CPR and NADPH compared to CHP.

3.6.2 P450 activity under crowded conditions

Despite CHP being a better choice of cofactor, the residual activity of CYP3A4 under crowded conditions was ≤70% regardless of the crowding agent used. This suggests that, under these conditions, there may have been one or more other diffusion- limited components to its catalytic cycle. CYP3A4 (and other CYPs) displays atypical kinetic behaviour with many substrates, including testosterone. Together, the results of many studies aimed at elucidating the mechanisms responsible for this behaviour have led

Davydov and Halpert to suggest a model that involves a role for both substrate-induced conformational change and multiple substrate binding sites (Figure 3.32)181. In this

Chapter 3

model, the first substrate binding event occurs peripherally and engenders a conformational change that allows for binding of a second substrate molecule. The peripherally bound substrate is then released and a third and final substrate binding event follows to form a fully active P450-substrate complex.

Figure 3.32 – “Three-site sequential ping-pong” mechanism proposed by Davydov and Halpert to explain the cooperativity observed in CYP3A4181. Substrate (orange hexagon) initially binds to a peripheral site on the enzyme in its resting conformation (magenta) but does not induce a spin shift of the heme-iron (A). This binding event induces a conformational change to a state with decreased water accessibility to the active site (green, B). Next, another substrate molecule binds, yielding an unstable ternary complex (C) from which the peripherally bound substrate molecule is released (D) resulting in some shift to high-spin. Finally, binding of another substrate molecule leads to the final ternary complex (E) accompanied by a full-amplitude shift to high-spin.

Comparison of the available X-ray crystal structures of CYP3A4 either unbound or bound to ligands of various sizes have contributed to this mechanistic model, demonstrating that significant conformational changes are possible and necessary to accommodate larger and/or multiple substrates (Figure 3.43)182-183. In fact, the volume of its active site can increase by >80% upon substrate binding. Also consistent with this model, and particularly relevant here is the fact that CYP3A4 is able to bind up to three

Chapter 3

184 molecules of testosterone concurrently with individual Kd values of 19, 37 and 56 µM .

In fact, CYP3A4 must bind at least two molecules before catalytic activity is observed.

The third binding event, which is likely occurring at an allosteric effector site away from the active site, increases the coupling efficiency of substrate turnover. Thus, if these conformational changes happened to be diffusion-limited under the crowded conditions that were tested, it would explain their general negative effect on CYP3A4 activity. On the other hand, it has been suggested that due to its smaller active site volume and lower flexibility, conformational changes in CYP2D6 may be less prominent185, offering a clue as to why its enzymatic activity was maintained in the presence of some crowding agents

(PEG 6, DEX 70 and FIC 70).

Figure 3.4 – Analysis by Ekroos and Störgen 2006182 of the conformational flexibility of CYP3A4. The most compact X-ray crystal structure of CYP3A4 (1TQN-ligand free) was superimposed with all others available at the time (2J0C-ketoconazole and 2J0D-erythromycin) and coloured according to observed differences is Cα positions with maximal differences shown in red.

Chapter 3

3.6.3 Effects of the differing crowding agent properties

Earlier studies of macromolecular crowding led many to assume that volume exclusion was its dominant effect and that comparing macromolecules to inert spheres interacting only through steric repulsions was a justifiable simplification. While there were caveats concerning the use of PEG as a crowding agent due to its propensity to interact with hydrophobic amino acid residues, the use of other artificial and proteinaceous polymers was recommended for their strict adherence to the excluded volume models141. More recently however, evidence has emerged suggesting that many of the most commonly used crowding agents are not completely inert. These include

BSA169, PVP168 and possibly even dextran and Ficoll186. (The latter two however are still considered sufficiently inert so as to exert purely volume exclusion effects.)

Consequently, theoreticians and experimentalists alike are calling for a more accurate model of crowding effects, one that takes into account the competition between volume exclusion and so-called “soft” interactions that arise due to the chemical nature of the interacting macromolecules168. Soft interactions can be non-specific (interactions between the crowder and amino acid residues or the peptide backbone) or specific

(interactions between the crowder and a certain motif or domain on the protein in its native state). Both types of soft interactions can either be stabilizing or destabilizing.

Unfortunately, such a model that has robust predictive powers is hard to come by, since these non-specific interactions are difficult to quantify and are likely to vary from system to system. Even with relatively inert dextran and Ficoll, there is evidence that current structural models may be inadequate186.

As seen in Figure 3.9, the addition of different crowding agents has vastly different effects on CYP2D6 activity. In light of the above facts, it is likely that these

Chapter 3

differences arise from soft interactions with some of the crowding agents. Attempts to explain the observed difference between them would be highly speculative at best.

However, it may be noteworthy to mention that adding PEG 6, DEX 70 or FIC 70 had no detrimental effect on CYP2D6 activity. Due to its propensity to interact with hydrophobic amino acid residues141, one would have expected PEG 6 to destabilize the tertiary structure of CYP2D6 leading to decreased enzyme activity in its presence. This was not observed however, possibly due to compensating volume exclusion effects. Further evidence for this is the fact that addition of the more inert and larger DEX 70 and FIC 70 also had no effect on enzymatic activity. In this case, one would expect soft interactions to play a lesser role and for volume exclusion effects to dominate. But, because of their larger size, positive effects from volume exclusion were likely cancelled by diffusion- limiting effects. In fact, the impact of hindered diffusion rates was clearly observed when the size of DEX was increased to 500 kDa. It follows that the net effect of PVP 40 and

BSA on CYP2D6 activity is likely the result of overwhelming negative soft interaction effects on diffusion or possibly protein stability.

As for CYP3A4, it is unlikely to have benefitted at all from excluded volume effects under the conditions that were tested since even the addition DEX 70 and FIC 70 was detrimental to its activity to a similar extent as the other crowding agents (Figure

3.5). However, it is possible that had a lower crowder concentration been tested, beneficial effects could have been observed by shifting the system out of a diffusion- limited regime into a transition state-limited one (Figure 3.3). (And the same could be said for CYP2D6.) Nonetheless, competing negative effects arising from soft interactions with the crowders are also possible and cannot be ruled out.

Chapter 3

3.6.4 The effect of adding small sugars on P450 activity

At first, CYP activity was assayed in the presence of sucrose because of its suitability as a control to distinguish crowding effects by dextran and Ficoll from those that may arise from increased polarity of the environment174. Surprisingly, a significant improvement in CYP3A4 activity (140 ± 10%) was consistently observed when 20% w/v sucrose was added, but only if CHP was used as the cofactor. Fearing possible interference by non-specific polarity effects, this led us to test the effects of other small sugars such as glucose, trehalose and melezitose. To our dismay, we found that the effects of these small sugars differed between sugars, between P450 isoforms, and between the cofactor systems used. Moreover, some of the sugars tested displayed a concentration- dependence with no obvious correlation to chemical structure.

While further experimentation is warranted to determine the mechanisms responsible for these effects (of which there may be several), some tentative conclusions can be drawn. First, these results seem to confirm the general observation that CPR domain motions are strongly diffusion-limited. While high sugar concentrations increase the viscosity of a solution, they do so to a lesser extent than crowding agents.

Nonetheless, even small increases in viscosity seem to suppress these important domain motions, cancelling any positive effects derived from the presence of the small sugar.

Second, while all sugars likely exert a negative effect on P450 activity by increasing viscosity, some have the ability to compensate for these negative effects through certain undetermined soft interactions. Finally, in the case of CYP2D6, sucrose was more likely to be an appropriate polarity control since it had no significant effect on its activity when added in the same concentration as the crowders that were tested. However, this outcome

Chapter 3

is undistinguishable from one that would result from the competition of compensating soft interactions, resulting in no net effect.

3.7 Conclusion

While it is difficult to conclusively distinguish between the effects of volume exclusion and soft interactions, conditions that afforded significant improvements in both

CYP2D6 and CYP3A4 activity were uncovered. Interestingly, these were likely not due to macromolecular crowding effects since they were observed only in the presence of small sugars. The addition of 30% w/v trehalose to CYP3A4 and 20% w/v glucose to

CYP2D6 afforded the best improvements in activity of 167 ± 3% and 130 ± 20% respectively when using CHP as the cofactor. No crowding agents had a positive effect on the enzyme activity. At best, CYP2D6 in presence of PEG 6, DEX 70 and FIC 70,

CYP2D6 had unchanged activity. On the other hand, all crowding agents tested decreased

CYP3A4 activity (by 30-85%). Thus, CYP3A4 seemed to be more diffusion-limited under the conditions that were studied, consistent with larger conformational changes182-

183 required for efficient catalytic activity compared to CYP2D6. Our results are also generally consistent with previous reports that CPR domain motions are diffusion- limited179.

Chapter 4

Chapter 4 – Non-covalent immobilization of CYP3A4 via C-terminal His-tag

4.1 Prelude

As discussed in Chapter 1, protein immobilization finds many applications including in the generation of protein microarrays, biosensors, bioreactors and in a variety of biophysical techniques like single-molecule and surface plasmon resonance (SPR) spectroscopy. More importantly, it has also become an important and successful strategy in the improvement of many biocatalysts. Properties such as stability, activity and selectivity can all potentially be improved by immobilization187. What’s more, biocatalyst immobilization facilitates recovery and reuse of the biocatalyst as well as product isolation. Unfortunately, an immobilization strategy that is generally tolerated by all enzymes and compatible with all potential applications has yet to be described. Instead, choosing from the wide array of available immobilization methods is much like choosing a bioconjugation strategy (see section 1.2). These reported methods can be classified into three broad categories: covalent, non-covalent and entrapment (Table 4.1)19, 78, 187-189.

Chapter 4

Table 4.1 – General protein immobilization strategies: advantages and disadvantages

Description Advantages Disadvantages

Covalent:

Random Usually via covalent Simple to implement, Random orientation binding through surface high yielding, mild not compatible with all amino acids like Cys and conditions, multi-site applications, often Lys/N-terminus attachment can leads to enzyme improve stability denaturation and/or inactivation, lacks orientation control

Site-specific Makes use of site-specific Better control of Requires prior protein and bioorthogonal orientation means modification via bioconjugation chemistry improved retention of molecular biology (see section 1.2) enzymatic properties, techniques allows immobilization in specific and uniform orientation, can potentially immobilize target protein directly from cell lysate Non-covalent:

Random Non-specific adsorption Same as random Same as random onto solid matrices (e.g. covalent above, covalent, biocatalyst nitrocellulose, hydrogel, requires no additional leaching is likely polylysine surface) or reagents, reversibility lipid membranes via allows reuse of hydrophobic and/or enzyme and/or solid electrostatic interactions support

Site-specific Most commonly via high Same as site-specific Same as site-specific affinity peptide tags or covalent, reversibility covalent, many domains (e.g. His-tag, allows reuse of proteins have already GST-tag, streptavidin, enzyme and/or solid been engineered with amtibodies) support affinity tags for purification purposes, usually limited to immobilization through protein termini

Entrapment: Physical confinement in No prior protein Can hinder substrate polymeric matrices such modification, creates access and product as polyacrylamide, protective egress agarose, alginate microenvironment, decreased leaching

Chapter 4

In this chapter, the oriented, non-covalent immobilization of CYP3A4 onto Ni-

NTA (nitrilotriacetic acid) resin via its C-terminal His-tag is described. The effects of immobilization on enzymatic activity as well as kinetic and thermodynamic stability of

CYP3A4 were assessed. Finally, the effects of immobilization on enzymatic activity after lyophilisation were also investigated. These experiments were performed with the help of two undergraduate students in the Auclair lab, Ronan Hanley and Yue Huang. More specifically, I preformed most of the experimental planning and some data analysis.

Ronan Hanley performed preliminary studies to confirm successful immobilization and assess its effects on enzymatic activity and thermodynamic stability. Yue Huang repeated these experiments to confirm their consistency. Kinetic stability and lyophilization studies were performed by Yue Huang. The results presented in this chapter are not published.

4.2 Introduction

4.2.1 Non-covalent immobilization of CYP3A4

A review of the literature concerning covalent immobilization of human CYPs including CYP3A4 was provided in section 1.2.4 in the context of bioconjugation.

However, this isoform has also been successfully immobilized by non-covalent means on a variety of solid matrices. These include gold electrodes for electrochemically driven drug metabolism190, glass slides for analysis of allosteric mechanisms by total internal reflection microscopy (TIRFM)191 and silver nanoparticles in the development of a localized surface plasmon resonance (LSPR) spectroscopy-based biosensor192. Various other surfaces have also been employed in the development of CYP3A4-based drug metabolism biosensors193-199. In most of the above cases, CYP3A4 was placed either in

Chapter 4

biological membranes (e.g. liver microsomes or E. coli-derived) or in artificial membranes (e.g. Nanodiscs191 or didodecyldimethylammonium bromide vesicles) before its immobilization. These types of systems are meant to mimic this enzyme’s native environment thereby allowing it to retain its catalytic properties upon immobilization.

Alternatively, CYP3A4 was entrapped in polyionic synthetic polymer films.

4.2.2 Immobilization of His-tagged proteins

Perhaps the most well known and widely exploited non-covalent immobilization method involves the polyhistidine affinity tag. This small tag of 4-8 amino acids can be easily fused to a protein of interest (usually at its termini but also in exposed loops) by established molecular biology techniques such as cloning into commercially available plasmid vectors designed for this purpose or PCR-based insertion mutagenesis. Also, His- tags are generally well tolerated by most proteins because of its small size. Proteins bearing a His-tag are immobilized by chelation to transition metals like Ni, Cu, Co and

Zn exposed on the surface of solid supports. The most common His-tag affinity support is functionalized with Ni-NTA as shown in Figure 4.1.19

While immobilization by this method is most often used for recombinant protein purification purposes, because of its ease of use and general versatility, it has also found its way into numerous other applications. Oriented non-covalent immobilization of His- tagged proteins onto electrodes via NTA-modified self-assembled monolayers (SAMs) or conductive polymers has been especially successful200. His-tagged proteins have also been attached to the surface of liposomes or micelles containing NTA-modified lipids201-

202 and onto microarray chips203-205. Moreover, many metal chelating surfaces are

Chapter 4

commercially available including polymeric resins, SPR gold sensor chips, glass microscope slides, microarray slides and 96-well microtiter plates.

Figure 4.1 – Ni-NTA affinity support for binding His-tagged proteins. The NTA support is first treated with Ni2+ ions. Protein immobilization occurs by chelation of the Ni2+ to the imidazole groups on the His-tag. Image taken from Wong et al.19.

During our lab’s investigations with respect to the biocatalytic properties of cytochrome P450 enzymes, especially CYP3A4119, 122, 124, 132, we became interested in their immobilization. Thus, we investigated several strategies (some of which are not described in this thesis) that would allow for their immobilization in several different orientations either non-covalently or covalently (see Chapter 5). Our goal was to find strategies that would retain or improve enzyme activity/stability in order to open up opportunities to study and apply these enzymes in new ways. In this chapter, the effects of non-covalent C-terminal immobilization of CYP3A4 on enzyme activity and stability are described.

Chapter 4

4.3 Results and discussion

4.3.1 Immobilization of His-tagged CYP3A4 onto Ni-NTA agarose resin

The CYP3A4 expression vector used in this thesis work produces CYP3A4 fused

206 to a C-terminal tetra-histidine tag . The His4-tag is used to facilitate enzyme purification by affinity chromatography with Ni-NTA agarose resin. We sought to take advantage of this high affinity interaction between the His-tag and Ni-NTA to assess the effects of C- terminal immobilization of CYP3A4 on its activity and stability. But first, it was desirable to gain some insight into the extent to which CYP3A4 could be immobilized on Ni-NTA agarose resin. To this end, CYP3A4 (45 pmol) was gently mixed with NTA agarose resin

(10 µL) loaded with Ni2+ (as received from the vendor Qiagen) or stripped of Ni2+ with ethylenediaminetetraacetic acid (EDTA). The resin was settled by centrifugation and the supernatant was assayed for residual CYP3A4 activity by monitoring the conversion of the substrate 7-benzyloxy-4-trifluoromethylcoumarin (BFC) to the fluorescent product 7- hydroxy-4-trifluoromethylcoumarin (HFC) using a fluorescence plate reader (Figure 4.2).

Figure 4.2 – Debenzylation of fluorogenic substrate 7-benzyloxy-4-trifluoromethylcoumarin (BFC) to 7-hydroxy-4-trifluoromethylcoumarin (HFC) by CYP3A4. Formation of the fluorescent product HFC can be monitored fluorometrically over time to assay enzymatic activity.

As seen in Figure 4.3, the residual CYP3A4 activity in the supernatant reserved from the mixture with Ni-NTA agarose resin was only ~15% of that found when the resin

Chapter 4

was devoid of Ni2+. Moreover, the activity found in the supernatant of the mixture with

Ni-free resin was the same as expected from the same amount of CYP3A4 in solution without any resin present (data not shown). This confirms that in the presence of Ni-NTA,

~85% of total CYP3A4 equilibrates towards immobilization on the resin. On the other hand, CYP3A4 has no detectable affinity for the resin when it has been stripped of Ni2+.

Also, the presence of Ni-free NTA is not detrimental to enzyme activity (data not shown).

160

140

120

100

Fluorescence (RFU) Fluorescence 40

0 0 500 1000 1500 2000 2500 -20 Time (s)

Figure 4.3 – Immobilization efficiency of His-tagged CYP3A4 on Ni-NTA agarose resin. CYP3A4 was mixed with Ni-NTA resin or Ni-free resin. Enzyme activity in the supernatant was assayed (blue and red lines respectively) by monitoring the appearance of the fluorescent product HFC from the substrate BFC over time. The immobilization yield was estimated by comparing the initial slope (black line) of each curve: a measure of the initial rate of each reaction. This experiment was repeated three times.

4.3.2 Effect of immobilization on the end-point activity of CYP3A4

Next, we assayed the activity of CYP3A4 immobilized on Ni-NTA agarose using testosterone as the substrate (see section 3.4). Each reaction tube contained ca. ~10 μL of resin in a total volume of 75 μL. Tubes were gently agitated in order to maintain the resin in suspension during the enzymatic reaction. To our satisfaction, the extent of product

Chapter 4

formation by CYP3A4 on Ni-NTA was statistically indistinguishable from that of

CYP3A4 in solution.

4.3.3 Effect of immobilization on the kinetic stability of CYP3A4

Testosterone was also used as a substrate to assay the kinetic stability of CYP3A4.

This time, the extent of product formation was measured at various time points using the same reaction conditions as described in section 4.3.2. As seen in Figure 4.4, the amount of product formed by CYP3A4 at each time point is generally not perturbed by immobilization on Ni-NTA resin. This suggests that immobilization has no detrimental effect on the kinetic stability of CYP3A4.

0.4

0.35

0.3

0.25

0.2

0.15 Product formationProduct 0.1

0.05

0 0 5 10 15 20 25 30 35 Time (min)

Figure 4.4 – Kinetic stability of immobilized CYP3A4. CYP3A4 catalyzed bioconversion of testosterone was quenched at various time points. Product formation was quantified by HPLC analysis and plotted (as a ratio to the internal standard) against reaction time, for CYP3A4 immobilized on Ni-NTA resin (red) or in solution (blue).

Chapter 4

4.3.4 Effect of immobilization on the thermodynamic stability of CYP3A4

To assay thermodynamic stability i.e. stability upon storage, CYP3A4 was incubated at room temperature in the absence of substrate and cofactors for various lengths of time. The enzyme activity was measured using testosterone as the substrate using the same reaction conditions as described in section 4.3.2. As seen in Figure 4.5, the amount of product formed by CYP3A4 at each time point is generally not perturbed by immobilization on Ni-NTA resin. This suggests that immobilization has little to no effect on the thermodynamic stability of CYP3A4.

1.2

0.8

0.6

Relative activity Relative 0.4

0.2

0 5 10 15 20 25

Incubation time (h)

Figure 4.5 – Thermodynamic stability of immobilized CYP3A4. The activity of CYP3A4 immobilized on Ni-NTA (red) or in solution (blue) using testosterone as the substrate after incubation at room temperature. Product formation (relative to that observed with incubation length t = 0) is plotted against incubation time.

4.3.5 Assessing the suitability of Ni-NTA agarose as a lyoprotectant for CYP3A4

CYP3A4 was lyophilized in solution, after immobilization on Ni-NTA resin or in the presence of Ni-free NTA resin. End-point activities of 41 ± 4%, 50 ± 10% and 51 ±

10% were observed relative to non-lyophilized enzyme respectively. Thus,

Chapter 4

immobilization on Ni-NTA resin seems to provide no protection against the damaging effects of freeze-drying.

4.4 Conclusion

Non-covalent immobilization of enzymes via His-tags is relevant to a wide variety of applications. Here we have demonstrated for the first time that non-covalent C- terminal immobilization of CYP3A4 through chelation to Ni-NTA agrose resin is generally tolerated with respect to activity as well as kinetic and thermodynamic stability.

This suggests that the C-terminus of CYP3A4 is accessible and that presumably, covalent

C-terminal immobilization would be equally well tolerated by this P450 isoform. While this work was progressing, single-molecule fluorescence microscopy feasibility studies were undertaken with a mutant of CYP3A4 immobilized via a cysteine residue onto silica microspheres (see Chapter 5). It is our intention to also apply the method of immobilization described in the present chapter to these experiments in order to test whether the orientation of the enzyme with respect to the solid support has any effect. We plan to publish the outcome of our studies with both orientations as soon as possible.

While preliminary studies look promising, certain technical details remain to be optimized (see section 5.4.6) before testing with the C-terminal His-tag immobilized

CYP3A4 can proceed. The results presented in this chapter may also be useful in other applications including the production of P450 microarrays, biosensors or bioreactors.

Chapter 5

Chapter 5 – Site-specific fluorescent labeling and oriented-immobilization of a triple mutant of CYP3A4 via C64

5.1 Preface

In recent years, the desire to generate protein and enzyme bioconjugates has exploded with the rise of more sophisticated biochemical and biophysical analytical techniques and applications such as single-molecule spectroscopy, the fabrication of protein microarrays, biosensors and bioreactors, diagnostics, biocatalysis and the in vivo visualization, tracking and targeting of biomolecules. As discussed in section 1.2.2, site- specific bioconjugation methods hold the most promise with respect to preserving native enzyme structure and function. During our lab’s various investigations geared towards the use of human cytochrome P450 enzymes as in vitro biocatalysts, it was envisaged that the successful generation of functional and site-specific covalent conjugates of these enzymes could potentially open up opportunities to study and apply these catalysts in new ways.

Once again, our attention was focused on CYP3A4 because of its high substrate promiscuous which boosts this P450 enzyme’s potential as a useful and versatile biocatalyst. It is worth repeating however, that despite its potential, many obstacles still prevent the use this enzyme (and other human P450s) in biocatalytic applications.

In this chapter, the site-specific modification of CYP3A4 is described. Fluorescent bioconjugates were first tested to facilitate their characterization with respect to yield and position of labeling. Activity assays were performed to assess the effects of

Chapter 5

bioconjugation on P450 activity. Finally, the optimized bioconjugation strategy was applied to the oriented-immobilization of CYP3A4 on solid supports. The work detailed in this chapter was performed in collaboration with Pierre Karam and Christina Calver

(graduate students in the lab of Prof. Gonzalo Cosa) and with the help of Yue Huang (an undergraduate student in the lab of Prof. Karine Auclair). Specifically, I performed the majority of the experimental work, Yue Huang helped in assaying the activity of some of the fluorescent bioconjugates, Pierre Karam performed the photobleaching studies and

Christina Calver performed the fluorescence confocal microscopy work. Most of these results have been published as Ménard, A., Huang, Y., Karam, P., Cosa, G. and Auclair,

K. (2012) Site-specific fluorescent labeling and oriented immobilization of a triple mutant of CYP3A4 via C64 Bioconj. Chem. 23, 826-836.

5.2 Introduction

At the time of undertaking this work, most of the bioorthogonal reactions that had been described suffered from slow, second-order kinetics and therefore often required high reagent concentrations, harsh conditions (i.e. high temperatures and extreme pHs) and long reaction times that would result in loss of P450 activity. A few improvements in this area have been reported recently however, namely for the CuI-catalyzed azide-alkyne cycloaddition reaction25, the hydrazone-ligation42 and the oxidative coupling of anilines to o-aminophenols50. Considering our observations on the stability of CYP3A4 at room temperature (Figure 5.1), we thought it prudent to attempt bioconjugation using cysteine- specific maleimide chemistry which can be achieved at room temperature and neutral pH, and proceeds to completion within a few hours. There is also precedent in generating

Chapter 5

functional, partially cysteine-depleted mutants of CYP3A4. Reactions of a quadruple mutant (C98S/C239S/C377S/C468A) with cysteine-reactive fluorescent probes have generated C64-specific labeled CYP3A4 used to study the mechanism of α- naphthoflavone binding by FRET74.

Figure 5.1 – Thermodynamic stability of CYP3A4 at 25°C in 0.1 M potassium phosphate buffer, pH 7.4. The enzyme was incubated in solution at 25°C and samples were assayed for residual P450 activity with testosterone as the substrate and CHP as the cofactor. All values represent an average of three separate experiments.

In this chapter, a similar cysteine-depletion approach to site-specific labeling of

CYP3A4 is described. Several mutants were prepared and labeled with fluorescent maleimides in order to facilitate the characterization of the resulting bioconjugates by a variety of methods. These included single-molecule fluorescence spectroscopy, electrophoresis, UV and fluorescence spectroscopy, and enzymatic assays. Thus, a triple mutant of CYP3A4 was identified that is singly modified at one of its surface cysteine residues. Finally, we demonstrate the functional immobilization of this mutant on maleimide-functionalized agarose resin and silica microspheres. This is, to our knowledge, the first functional immobilization of CYP3A4 in a covalent and oriented

Chapter 5

fashion. Covalent and non-covalent immobilization of CYP3A4 enzymes were reviewed in sections 1.2.4 and 4.2.1 respectively.

5.3 Maleimide preparation and their reaction with CYP3A4

During these investigations, CYP3A4 and its mutants were modified with a variety of soluble and solid-supported maleimide derivatives by reacting with available cysteine thiols on its surface (Scheme 5.1A). Treatment of the enzyme with tris(2- carboxyethyl)phosphine (TCEP) ensures the availability of these thiols to react with the maleimide by reducing any disulfide bonds. Several of the maleimides were prepared, as shown in Scheme 5.1B, by the addition of an amine to the NHS group of either of two heterobifunctional crosslinkers 5.15b or 5.15c. Others were commercially available and used as is. Detailed structures are shown in Figure 5.2.

Scheme 5.1 – A) General reaction used to label wild-type and mutant CYP3A4 enzymes with maleimides 5.1-5.14. B) General approach to preparing the non-commercial maleimide labels. See Figure 5.2 for detailed structures.

Chapter 5

Figure 5.2 – Structures for molecules 5.1-5.9 and 5.14. The structures for commercial compounds 5.10- 5.13 are not disclosed by the vendor. The larger sphere represents CarboxyLink resin in 5.4a- 5.4c. The smaller sphere represents the silica microsphere in 5.14a and 5.14c.

5.4 Results and discussion

5.4.1 Analysis of fluorescently labeled wild-type CYP3A4

Information about the number and identity of CYP3A4 cysteine residues available for reaction with maleimides is desirable to achieve regioselective conjugation of

CYP3A4. The primary sequence of CYP3A4 lists 7 cysteine residues (Figure 5.3). Upon careful inspection of the crystal structure of CYP3A4, one notes that C442 is coordinated

Chapter 5

to the heme prosthetic group and buried at the center of the protein, making it inaccessible to maleimide modification when CYP3A4 is in its correctly folded holo-form. After chemical reaction of wild-type CYP3A4 with maleimide-containing molecules 5.1b and

5.3b (Scheme 5.1, Figure 5.2), MS analysis gave ambiguous results with poor reproducibility (data not shown). Failure of MS to provide the desired information is likely due to the low stability of CYP3A4 and its notorious tendency to form aggregates.

Figure 5.3 – Crystal structure of CYP3A4 (PDB: 1TQN) with cysteine residues highlighted in red.

A combination of methods was used here to characterize the CYP3A4 conjugates.

These include single-molecule fluorescence spectroscopy, electrophoresis, UV and fluorescence spectroscopy, and enzymatic assays. For detailed characterization of the conjugates, wild-type CYP3A4 was reacted with an excess of fluorescent maleimide

DyLight 549 (5.10, structure not disclosed by vendor) as shown in Scheme 5.1A. A sample of CYP3A4 labeled with 5.10 was imaged with a single-molecule total internal reflection fluorescence microscopy setup (TIRFM, see Experimental protocols section

7.2.6.6 for details). Analysis of the images acquired allowed us to count the number of

Chapter 5

fluorescent moieties present per labeled protein molecule (Figure 5.4). Results indicated that modification with 5.10 gave predominantly singly modified CYP3A4 (76.9%), with a significant proportion being doubly and triply modified (18.6% and 4.2% respectively).

Figure 5.4 – Single-molecule photobleaching of wild-type CYP3A4 labeled with DyLight 549 maleimide (5.10). A) Representative example of a wide-field single-molecule total internal reflection fluorescence (TIRF) microscope image obtained for fluorescently labeled CYP3A4 upon excitation at 532 nm. B) Representative intensity vs. time trajectories observed when the protein has one fluorophore/protein molecule (one step), two fluorophores/protein molecule (two steps) and three fluorophores/protein molecule (three steps). C) Tallied trajectories for wild-type CYP3A4 (n = 2212). Pie colours used: blue = one step, red = two steps, green = three steps, purple = four steps.

On the down side, this same conjugate only retained ~5% of its original enzymatic activity when assayed with testosterone as the substrate and cytochrome P450 reductase

(CPR)/NADPH as the cofactors. Replacing the native cofactors with cumene hydroperoxide (CHP) marginally improved the retained activity to ~10%. This deleterious effect on the activity hints at C98 being the main site of modification, based on recent studies indicating that large aromatic substitutions at this residue result in loss

Chapter 5

of activity207. The C98W and C98F mutations were indeed reported to result in a conformational shift that led to a subsequent loss in catalytic activity, due in part to a reduction in the ability of CYP3A4 to properly interact with its reductase CPR. The C98S

CYP3A4 mutant on the other hand was reported to display unchanged activity.

These results do not eliminate the possibility of different singly labeled regioisomers being formed concurrently. To gain a better handle on the identity of the residues being modified, we turned to an approach combining SDS-PAGE and BrCN digestion of fluorescently labeled P450 samples. BrCN is known to cleave proteins selectively after methionine residues (Scheme 5.2). Sequence and molecular weight of the cysteine- containing peptides expected from BrCN digestion of CYP3A4 are shown in Figure 5.5.

All of the expected digestion peptides contain a maximum of one cysteine residue.

Figure 5.5 – A) Primary sequence of human CYP3A4 used here (referred to as the wild-type) listed with the one letter code for amino acids. Theoretical BrCN digestion peptides are underlined, and their approximate molecular weight when labelled with compound 5.10 is shown with the sequence number of the cysteine residue included in each peptide. B) Crystal structure of CYP3A4 (PDB: 1TQN) with cysteine residues highlighted in red.

Scheme 5.2 – Cleavage of the peptide backbone after a methionine residue by cyanogen bromide (BrCN).

Chapter 5

Digestion and peptide analysis of 5.10-labeled wild-type CYP3A4 revealed that at least 5 of the cysteine positions were accessible for modification (Figure 5.6). Indeed, SDS-

PAGE of the digested protein revealed 5 peptide bands exhibiting fluorescence emission between 565 and 595 nm (580 nm band-pass filter) when excited at 532 nm (green- excited mode), as suggested by vendor for DyLight 549 (5.10). The bands corresponding to the C239- and C377-containing peptides are significantly fainter than the other ones, possibly suggesting minor reactivity at these sites.

Figure 5.6 – Tricine-SDS-PAGE analysis of BrCN digestion peptides of CYP3A4 wild-type (wt) labeled with DyLight 549 maleimide (5.10). The image was acquired with a Typhoon Trio + scanner using the Green-excited mode (λex = 532 nm), with a 580 nm band-pass filter (580 BP 30, transmits light between 565 and 595 nm with a transmission peak centered on 580 nm).

5.4.2 Design of cysteine-depleted CYP3A4 mutants for site-specific cysteine modification

As alluded to above, a C98S substitution was desirable in order to prevent conjugation at C98 which is expected to incur a loss in activity. It was also desirable to minimize the number of mutations in order to increase the likelihood of obtaining fully active and stable mutants. Thus, mutant 1 contained only one mutation, C98S. Mutant 2

(C98S/C377S/C468G) and mutant 3 (C98S/C239S/C468G) were also generated to confirm the sites of modifications (especially since some peptides co-migrate on the gel),

100

Chapter 5

and in the hopes of producing variants that would be singly modified by maleimides. We chose to mutant C468 to glycine because a sequence alignment performed by Tsalkova et al. suggested that such a substitution should be more appropriate than C468S74.

The photobleaching profile of mutant 1 labeled with 5.10 was very similar to that of labeled wild-type enzyme (data not shown). The key difference was revealed by gel analysis of the digested protein (Figure 5.7) which shows a decrease in the fluorescence intensity of the band with a molecular weight corresponding to the C98-containing peptide. Because of their very similar predicted molecular weights (4.0 kDa and 3.8 kDa respectively), the C64- and C98-containing peptides are not resolved under the electrophoretic conditions used. Since C98 was mutated to Ser in mutant 1, we do not expect the 98th-residue-containing peptide to have reacted with maleimide 5.10, thus reducing the fluorescence of the band at ~3.9 kDa. The remainder of fluorescence at this

MW implies that C64 is modified in mutant 1, and both C64 and C98 are labeled in the wild-type enzyme. Taken together these results suggest that under our conditions, most of the cysteine residues can be modified, although only up to 2 (or 3, but rarely) are labeled at a time. This is consistent with previous reports suggesting that more than 2 cysteine-

Figure 5.7 – Tricine-SDS-PAGE analysis of BrCN digestion peptides of CYP3A4 variants labeled with DyLight 549 maleimide (5.10). The image was acquired with a Typhoon Trio + scanner using the Green-excited mode (λex = 532 nm), with a 580 nm band-pass filter (580 BP 30, transmits light between 565 and 595 nm with a transmission peak centered on 580 nm).

101

Chapter 5

modifications per CYP3A4 protein resulted in protein denaturation74.

As for mutants 2 (C98S/C377S/C468G) and 3 (C98S/C239S/C468G), however, our data suggest that both mutants are modified primarily at C64 (Figure 5.7). They cannot be labeled at the co-migrating peptide containing C98 because of the C98S mutation. In the case of mutant 2, a minor band corresponding to labeling at C239 and a significant band suggesting modification at C442 are observed. The latter is likely due to partial denaturation of the enzyme resulting in exposure of the heme-ligated cysteine residue. Since this band is not observed with mutant 3, a detrimental effect of C377S is inferred. On the other hand, we were delighted to see one major band with labeled mutant

3, suggesting very specific modification by 5.10 at the C64 position. For reasons that we do not understand, C377 and C58 were not modified despite sequence analysis confirming that they had not been mutated. This is likely due to a conformational difference found in mutant 3 rendering these residues unavailable for reactions with maleimides.

Corroborating results were also obtained by UV-Vis spectroscopy indicating the presence of ~0.8 DyLight groups (fluorophore of 5.10) per protein molecule under saturating conditions (Figure 5.8), and by photobleaching analysis which indicated that

96.9% of the mutant 3 conjugates were singly labeled. According to the analysis of the gel (Figure 5.7), the remaining 3-4% that was doubly labeled is likely to result from a small proportion of the protein being denatured and exposing C442 for labeling.

Moreover, this high regioselectivity was consistently observed with all other fluorescent maleimide-conjugates of mutant 3 that were characterized, namely those of 5.1b, 5.5d,

5.8d, 5.11, 5.12 and 5.13 (data not shown).

102

Chapter 5

Figure 5.8 – Absorbance spectra of CYP3A4 mutant 3 labeled with DyLight 549 maleimide (5.10). In blue, 10 maleimide equivalents were used resulting in ca. ~0.4 labels/protein molecule. In purple, 100 - maleimide equivalents were used resulting in ca. ~0.8 labels/protein molecule. An ε280 nm of 40,340 M 1•cm-1 was used to calculate the concentration of CYP3A4.

The identity of the main residue being modified in mutant 3 was further verified with mutant 4, consisting of C64S and C98S mutations. As seen in Figure 5.7, characterization of 10-labeled mutant 4 shows that the band corresponding to the C64- containing peptide does not fluoresce anymore. This confirms that mutant 3 is modified almost exclusively at C64.

In addition, we were gratified to find that mutant 3 is as stable and even more active (150 ± 30%) than wild-type CYP3A4. This increase in activity compared to the wild-type is possibly linked to conformational differences (such as those proposed to explain the labeling selectivity). These could somehow be responsible for increasing the kinetic stability of the enzyme (i.e. its efficiency) or improving substrate binding.

Moreover, whereas labeled wild-type CYP3A4 lost >90% of its activity, 5.10-labeled mutant 3 remained at least as active, if not more, than unlabeled wild-type CYP3A4 (120

± 20%).

103

Chapter 5

5.4.3 Effect of varying the maleimide label on enzyme activity

Next we wanted to expand our scope of maleimides used to label the CYP3A4 mutant 3. This mutant was therefore reacted with a variety of structurally diverse maleimides, either commercially available or synthesized by our group (Scheme 5.1,

Figure 5.2). Unlike the wild-type enzyme, mutant 3 retained some degree of activity upon labeling with most maleimides that were tested (Figure 5.9, labeled mutant in purple).

DyLight maleimides 5.10-5.13 were especially interesting, producing mutant 3 conjugates with 80-120% of the activity of non-modified wild-type. On the other hand, addition of maleimides 5.1b-c, 5.2c-d, and 5.8d proved quite noxious to enzymatic activity.

Figure 5.9 – Relative activity of CYP3A4 wild-type (in blue) and mutant 3 (in purple) after labeling with maleimides 5.1-5.13. Activity was measured using the testosterone assay similar to that described in section 3.4. Percent activity is reported relative to that of non-labeled wild-type CYP3A4 (which was normalized to 100%, red line). All values are the average of duplicates or triplicates ± 1 standard deviation.

To assess the role that the linker portion (as opposed to the fluorophore portion) plays in the observed loss of activity, maleimides 5.3b and 5.3c were prepared and coupled to mutant 3. As seen in Figure 5.9, when modified with these maleimides, mutant

104

Chapter 5

3 retains most of its activity. This suggests that the fluorophore portions are likely the main culprits causing the loss of activity observed upon labeling of mutant 3. It was envisaged that the fluorophores themselves might inhibit the activity of CYP3A4. This hypothesis was verified by monitoring the enzyme activity of unlabeled wild-type and mutant enzyme in the presence of the free fluorophores. As depicted in Figure 5.10, in general, the degree of inhibition observed in the presence of free fluorophores correlates with the loss of activity seen upon labeling (Figure 5.9).

Figure 5.10 – Effect of free fluorophores (dark purple: 100 µM, light purple: 1 mM) on the enzymatic activity of non-labelled CYP3A4 mutant 3. The data is normalized against the activity of non-labeled mutant 3 in the absence of free fluorophore. All values are the average of duplicates or triplicates ± 1 standard deviation.

We can further explain the effect of modifying mutant 3 on its activity by observing changes in its reduced-CO spectrum. As seen in Figure 5.11, labeling with maleimides 5.2d and 5.9d caused an observable loss in intensity of the band at 450 nm and an increase in absorbance at 420 nm. This is characteristic of a switch to the non- functional P420 form of the enzyme. These results allow us to rule out steric effects and to attribute the loss of activity observed with selected labels to varying degrees of enzyme

105

Chapter 5

inhibition by the fluorophores themselves. In the end, it would seem prudent to test a variety of molecular structures when attempting to label CYP3A4.

Figure 5.11 – Effect of labeling with maleimides 5.9d (A) and 5.2d (B) on the CO-difference spectrum of CYP3A4 mutant 3 compared to the non-labeled control (C). Each plot compares the CO- difference spectrum before (in blue) and after (in purple) labeling. The CO-difference spectra were acquired in the manner reported by Omura and Sato1. Note that 5.2d absorbs maximally in the region around 500 nm.

5.4.4 Immobilization of CYP3A4 mutant 3 onto maleimide-functionalized CarboxyLink resin

As an extension of our work with soluble fluorescent labels, we sought to modify mutant 3 with solid supports derivatized with maleimides. Thus, commercial

CarboxyLink resin (5.4a) was modified to display a maleimide group at the end of a PEG chain of varying length (see 5.4b and 5.4c). Maleimide coverage of the resin was calculated to be ~0.6 nmol/µL resin. This derivatization was followed by reaction with

CYP3A4 mutant 3. No attempts were made to immobilize the wild-type enzyme in light

106

Chapter 5

of the fact that attachment of the linker alone (5.3b or 5.3c) to wild-type CYP3A4 suffices to bring the enzymatic activity below 5% (Figure 5.9). Attachment of 5.3b or

5.3c to mutant 3 on the other hand did not affect the enzyme activity significantly, and to our satisfaction, covalent bond formation with the modified resins 5.4b and 5.4c did not cause significant decrease in activity either. To assess the efficiency of CYP3A4 immobilization, the enzyme activity present in the supernatant was assayed using BFC as the substrate. The activity detected in the supernatants was <5% compared to 100% for a control containing 5.4a, indicating an immobilization yield of ≥95%. Moreover, once immobilized on 5.4b and 5.4c, the activity of CYP3A4 mutant 3 was retained (90 ± 30% and 80 ± 30% respectively, relative to non-labeled wild-type) and no penalty was incurred on the thermodynamic stability of the mutant over the 16 h observation period. These results were promising with respect to applications that require oriented-immobilization of enzymes onto solid surfaces such as single-molecule spectroscopy and the production of microarrays.

5.4.5 Immobilization of CYP3A4 mutant 3 onto maleimide-functionalized silica microspheres

Finally, we sought to assess the activity of mutant 3 upon immobilization to a surface that would be more appropriate for applications such as those mentioned above.

Thus, 0.1 µm colloidal silica microspheres were aminosilanized and treated with crosslinker 5.15c (Scheme 5.1) to produce a PEG-maleimide surface analogous to that used to immobilize biomolecules on glass slides for single-molecule experiments208 or the production of microarrays. In fact, biomolecule-modified microspheres are often used directly in single-molecule experiments209-210 and in the production of liquid bead arrays.

107

Chapter 5

CYP3A4 mutant 3 was then immobilized onto these beads as described above for the

CarboxyLink agarose resin. Under the conditions used here (see Experimental protocols section 7.2.6.12 Method A for details), the immobilization yield was ~70%. Once immobilized, mutant 3 retained 60 ± 10% of its activity compared to the activity of mutant 3 in solution when assayed with testosterone as the substrate, and 90 ± 10% activity retained versus the wild-type enzyme (Figures 5.12). Activity of the microsphere- immobilized mutant 3 was also observed with BFC as the substrate (60 ± 10% versus wild-type and 40 ± 10% versus mutant 3).

Figure 5.12 – Immobilization of CYP3A4 mutant 3 on silica microspheres. A) Control: representative HPLC UV-chromatogram of 6β-hydroxytestosterone (P, 7.1 min) formation by CYP3A4 mutant 3 in solution. B) Representative HPLC UV-chromatogram of 6β- hydroxytestosterone (P, 7.1 min) formation by microsphere-immobilized CYP3A4 mutant 3. The peaks at 10.1 min and 15.8 min correspond to the internal standard cortexolone (IS) and the substrate testosterone (T) respectively. Note: a very large excess of testosterone was used in these assays.

5.4.6 Preliminary results towards single-molecule fluorescence microscopy studies of CYP3A4 mutant 3 on silica microspheres

With active CYP3A4 mutant 3 (referred to as CYP3A4 for the remainder of this chapter) on silica microspheres in hand, feasibility studies were undertaken towards the goal of using our system to visualize single enzyme turnovers and perform single-

108

Chapter 5

molecule kinetics studies with this P450 isoform. We envisaged using BFC as an ideal substrate for such experiments because of its fluorogenic properties. In other words, it only becomes fluorescent upon debenzylation to 7-hydroxy-4-trifluoromethylcoumarin

(HFC) by CYP3A4 (Figure 5.13). The end goal then is to place a single CYP3A4 enzyme per bead in order to observe the production of HFC by CYP3A4 at a single-molecule level, using a confocal fluorescence or TIRF microscope. This experimental design is similar to that employed by English et al. to visualize single turnovers of the fluorogenic substrate resorufin-β-D-galactopyranoside to resorufin by the enzyme β-galatosidase209.

Figure 5.13 – Debenzylation of fluorogenic substrate 7-benzyloxy-4-trifluoromethylcoumarin (BFC) to 7-hydroxy-4-trifluoromethylcoumarin (HFC) by CYP3A4. Formation of the fluorescent product HFC can be monitored fluorometrically over time to assay enzymatic activity.

Because of the low turnover number of the CYP3A4-catalyzed conversion of BFC in solution (~0.7 min-1 at 22°C), we thought it prudent to first attempt these experiments with several active CYP3A4 enzymes per bead (ca. ~40, see Experimental protocols section 7.2.6.12 Method B for details) in order to ensure that HFC production could be visualized. The beads were also labeled with the fluorescent dye DyLight 650 maleimide

(5.13) in order to easily locate them under the microscope using a dual-colour setup (see

Figure 5.14 and Experimental protocols section 7.2.6.13 for details). With our particular setup, the samples can be excited with either 406 nm or 633 nm laser light. The emission is split into two channels: the red channel (>640 nm) and the green channel (<640 nm).

109

Chapter 5

Emission is monitored and recorded in both channels concurrently regardless of the laser being used.

Figure 5.14 – Preparation of doubly modified silica microspheres. The maleimide-modified microspheres (5.14c, only two maleimides are shown and the remainder of crosslinker 5.15c is replaced by a wavy line for clarity). First, CYP3A4 is immobilized onto the beads via its C64 thiol (ca. ~40 per bead) and the remaining maleimides are quenched with DTT. Finally, the beads are fluorescently labeled with DyLight maleimide 650 (5.13) by reacting with the other thiol on the DTT moiety (ca. ~10 per bead).

To visualize the appearance of HFC, the doubly modified beads were adsorbed onto the surface of a glass microscope slide. Single beads were located by raster scanning a 10 × 10 µm area with the 633 nm laser and with emission from DyLight 650 detected in the red channel (>640 nm). (The vendors of 5.13 report a maximum excitation wavelength of 652 nm and a maximum emission wavelength of 672 nm). This tells us where CYP3A4 is localized and where to look for the appearance of HFC. The substrate

BFC and the cofactor CHP were added and the appearance of HFC was monitored over

110

Chapter 5

time by exciting each bead continuously for 10 minutes (one bead at a time) with the 406 nm laser and with emission detected in the green channel (<640 nm). (When excited at

406 nm, the maximum emission wavelength of the enzymatic product HFC is ~505 nm).

Gratifyingly, increased fluorescence above background levels was observed at locations corresponding to those of the beads and thus of CYP3A4. However, two unanticipated observations were made. First, as seen in Figure 5.15, instead of seeing brief increases in fluorescence intensity as would be expected if HFC diffused away from the bead soon after being formed, we observed a linear increase in fluorescence intensity over time, preceded by a lag period. Second, this increase in fluorescence intensity does not occur

(at a detectable level) concurrently on all beads even though they are all exposed to BFC and CHP at the same time. Instead, it appears as though excitation by the 406 nm laser is necessary to initiate this fluorescence increase (Figure 5.16). In other words, the fluorescence intensity at the location of each bead starts off at background levels and only increases to a significant extent once excited.

Figure 5.15 – Example of an intensity vs. time trajectory acquired upon 406 nm excitation of a single silica microsphere bearing CYP3A4 in the presence of the substrate BFC and the surrogate cofactor CHP. Fluorescence intensity was recorded in the green channel (<640 nm).

111

Chapter 5

Figure 5.16 – Fluorescence scanning confocal images of CYP3A4 activity on single silica microspheres. A) The beads are localized by scanning a region with 633 nm laser light (0.7 µW). The resulting emission from the DyLight 650 label (5.13) was monitored in the red channel (>640 nm). B) After 406 nm excitation (8 µW) of single beads, the same region is scanned with 406 nm laser light (2 µW). The resulting emission was monitored in the green channel (<640 nm). Note that beads from the red channel (A) that do not show fluorescence at a corresponding location in the green channel (B) were not excited with the 406 nm laser. The area scanned was 10 × 10 microns or 250 × 250 pixels as indicated on the scale, giving a resolution of 0.04 microns per pixel.

Intrigued by these observations, several control experiments were performed to rule out any non-CYP3A4 mediated conversion of BFC to HFC by the laser. Results of these experiments confirm that indeed BFC, CHP and active CYP3A4 are all necessary for the fluorescence increase to occur. Neither CHP nor BFC alone was sufficient to produce a detectable increase in fluorescence. Likewise, no increase above background was observed with denatured enzyme (in the presence of BFC and CHP) or when the non- fluorogenic substrate testosterone was used instead of BFC. Thus, it appears as though the observed fluorescence increase is indeed due to the CYP3A4-catalyzed production of

HFC from the substrate BFC.

112

Chapter 5

In order to better understand our observations, it may be important to take a closer look into the mechanism by which CHP activates P450 enzymes during catalysis (Figure

5.17). When CHP is used as an oxygen donor, it first coordinates to the ferric heme.

Figure 5.17 – Mechanisms of peroxo O-O bond cleavage of cumene hydroperoxide (CHP) by the heme-iron in P450 enzymes. A) Heterolytic cleavage leads to the catalytically competent iron(IV)-oxo complex (Compound I). B) Homolytic cleavage initiates oxidative damage of the P450 active site by formation of the cumoxyl radical and the dead-end complex Compound II. The porphyrin ring is represented by a square around the iron. Then, the iron catalyzes O-O bond scission via heterolytic or homolytic cleavage. Both mechanisms are thought to occur concurrently to a significant extent. While heterolytic cleavage leads to the catalytically competent iron(IV)-oxo intermediate (Compound I)

(Figure 5.17A), the competing homolytic cleavage mechanism leads to the formation of the catalytically incompetent Compound II and the cumoxyl radical (Figure 5.17B) which is thought to initiate oxidative degradation of the P450.211-212

Further insights may also be gained from known photochemistry of heme proteins to explain the increased HFC production observed upon irradiation. Flash photolysis

113

Chapter 5

experiments have demonstrated that Compound I can be generated from Compound II, albeit by irradiation with 355 nm laser light213 (here we used a 406 nm laser which is also within Soret band of P450 enzymes). If this same photochemical process is taking place during our experiments, it would increase the production of Compound I and consequently increase the rate of conversion of BFC to HFC by CYP3A4 by rescuing the enzyme from the dead-end production of Compound II. While others have also activated

P450 enzymes photochemically via direct reduction109 (with or without sensitizers/sacrificial electron donors), our control experiments confirm that this is likely not the primary mechanism observed in these experiments.

Another possibility is that irradiation with the 406 nm laser simply heats up the sample cell to a temperature that favours a higher CYP3A4 catalytic activity. During ensemble experiments with CYP3A4 in solution, turnover numbers were almost 6 times higher at 37°C compared to 22°C (data not shown). Thus, even increasing the temperature by a few degrees could have an effect and may help explain our observation, if only in part. Further experiments may help clarify this point.

Ultimately, further investigations will be required to fully explain our observations and adapt our conditions to permit the visualization of single enzyme turnovers. Others have studied allosteric mechanisms of CYP3A4 ligand binding using single-molecule spectroscopy191. To our knowledge, if these experiments are successful, this will be the first time single-molecule turnovers are observed with a P450 enzyme. Future investigations will potentially lead to interesting single-molecule kinetics experiments that could provide valuable insights into the hidden intricacies of P450-catalyzed reactions that cannot be detected by ensemble methods.

114

Chapter 5

5.5 Conclusion

In summary, we have generated a triple mutant of CYP3A4 (C98S/C239S/C468G) that is singly modified at C64 by a variety of soluble and solid-supported maleimides.

While the actual chemical modification at C64 incurs a minor cost to activity, this is made up for in many cases by the fact that this mutant is more active (150 ± 30%) than the wild-type enzyme. Bioconjugation of mutant 3 with some dyes (5.1, 5.2 and 5.8) had a detrimental effect on the enzyme activity. We were able to rule out the implication of steric effects in this observed loss in activity and demonstrated instead that enzyme inhibition by some fluorophore moieties on the maleimide labels played a predominant role. Finally, our bioconjugation strategy was successfully applied toward the immobilization of this CYP3A4 mutant on maleimide-functionalized CarboxyLink resin and silica microspheres. In fact the enzyme, attached via C64, remained active on these solid matrices and enzymatic activity was visualized by fluorescence confocal microscopy. These results are promising with respect to future kinetic studies on the single-molecule level and suggest that this mutant of CYP3A4 would also be an appropriate choice for other applications that favour oriented-immobilization e.g. in the development of protein microarrays, biosensors or bioreactors.

115

Chapter 6

Chapter 6 – Contributions

The aim of this thesis work was to study human cytochrome P450 enzymes in terms of their biocatalytic properties with a focus on the issues of predictability, activity and stability. The results presented here contribute to scientific knowledge in several ways. First, the novel use of type II ligands as chemical auxiliaries for CYP2E1 further reinforces the potential of P450 enzymes as selective and predictable C-H oxidation biocatalysts. Next, explorations in macromolecular crowding, enzyme bioconjugation and oriented-immobilization has contributed to a greater understanding of factors that affect the activity and stability of CYP3A4 and CYP2D6. Finally, the oriented-immobilization of CYP3A4 has opened up opportunities toward applications in biotechnology and biophysical studies.

6.1 Selective aliphatic C-H oxidations

One of the great challenges in synthetic chemistry is that of achieving selective and predictable oxidation of inactivated sp3 C-H bond214. Likely the most important advances in this field thus far have come out of the lab of M. Christina White. Their highly electrophilic and bulky oxidant [(Fe(S,S-PDP)(CH3CN)2](SbF6)2 (also known as the White-Chen catalyst) is able to oxidize the more facile 3° C-H bonds215 as well as 2°

C-H bonds216 with predictable reactivity trends governed by substrate electronics, sterics and stereoelectronics. More recently, they have demonstrated that this substrate bias can be overridden by the judicious placement of a substrate directing group217. Despite this exciting progress, most examples are limited to very specific contexts, show only modest

116

Chapter 6

regioselectivity and methylene bond oxidations usually result in over-oxidation to the ketone. Our group’s interest in P450 enzymes stems from the potential role they could play in tackling this important challenge. Despite their remarkable ability to hydroxylate inactivated C-H bonds with high regio- and stereoselectivity, the use of P450 enzymes as in vitro C-H oxidation biocatalysts has been limited by, among other factors, poor predictability. Thus, part of the research in this thesis focuses on the predictability of

CYP2E1 biotransformations.

6.1.1 Novel use of type II binders as P450 enzyme-targeting chemical auxiliaries

The results presented in Chapter 2 contribute to addressing the issue of poor product predictability using nicotinate as a chemical auxiliary with type II binding properties. While substrate promiscuity is an important asset of these enzymes with respect to their potential as versatile biocatalysts, it also makes predicting the site of oxidation difficult, a problem that is especially important in the field of drug design and development218. Here, we showed that attaching nicotinate to a variety of short hydrocarbon molecules that are not CYP2E1 substrates, can promote their oxidation with predictable regioselectivity. Moreover, nicotinate improved CYP2E1 oxidation yields, presumably by directing the substrates to the P450 active site via high affinity type II interaction with the heme-iron. This result also describes a novel application of these type

II ligands that differs greatly from their traditional use as P450 inhibitors2, further reinforcing the bourgeoning realization that type II ligands are sometimes also P450

137-139, 154 substrates . Finally, results from docking studies performed with FITTED were in good agreement with our experimental findings, providing further validation of its accuracy in predicting genuine binding modes161, 219.

117

Chapter 6

6.2 Understanding activity and stability

Other factors that have hindered the use of P450 enzymes as in vitro biocatalysts are their poor activity and poor stability. Several aspects of the research presented in this thesis contribute to a better understanding of different factors that can affect (or not) the activity and stability of CYP3A4 and CYP2D6. The results discussed in Chapter 3 demonstrate that the activity of these two isoforms can be negatively affected or unaffected by the presence of various macromolecular crowding agents by a combination of volume exclusion effects and soft interactions. These results provide insight into the diffusion-limited processes that take place during the P450 catalytic cycle. Conditions that improve enzymatic activity (i.e. the addition of small sugar molecules) were also revealed. In Chapter 4, we find that CYP3A4 can be non-covalently immobilized via its

C-terminal His-tag with no detrimental effect on its activity or stability (kinetic and thermodynamic). In Chapter 5, a triple mutant of CYP3A4 was discovered that is as stable as the wild-type enzyme and 1.5 times more active. Thanks to this mutant, we showed that CYP3A4 can be modified with a variety of small fluorescent maleimide dyes while conserving enzymatic activity. This was made possible because of the site-specific nature of the bioconjugation with this mutant that has only one remaining reactive cysteine thiol on its surface, confirming once again, the advantage of site-selectivity in protein bioconjugation. Finally, also in Chapter 5, we demonstrate that this mutant retains its activity when immobilized on two types of maleimide-functionalized solid supports.

Together, these results will help future researchers appreciate conditions and modifications that can and cannot be well tolerated by these enzymes.

118

Chapter 6

6.3 Progress towards applications in biotechnology and biophysical studies

6.3.1 Site-specific fluorescent labeling of CYP3A4

Finding methods for site-specific labeling of proteins is an important bottleneck hindering single-molecule FRET studies in particular220-221. Moreover, achieving site- selectivity is of supreme importance in order to detect biologically relevant heterogeneities (i.e. conformational distributions or stochastic dynamics) that cannot be measured by similar ensemble methods. During the course of our investigations towards the site-specific modification of CYP3A4 (Chapter 5), we noticed that certain dyes were well tolerated by the enzyme (in terms of retention of activity) while others were quite noxious to the enzyme. This is likely to be the case for other proteins and enzymes as well, and should serve as a caution to other researchers attempting to site-specifically label their protein of interest; not only is the location of labeling important but also the chemical nature of the label. These results also provide the ground work for future experiments that could involve single-molecule FRET studies of this fluorescently labeled

P450 isoform and its interaction with CPR.

6.3.2 Oriented-immobilization of CYP3A4

Realizing that the biocatalytic potential of P450 enzymes goes beyond their use as

C-H oxidation catalysts, an important part this thesis was devoted to their oriented- immobilization. While protein immobilization has been used to improve biocatalysts with respect to stability, activity and selectivity187, it also finds its way into many applications that include the generation of protein microarrays, biosensors, bioreactors and in a variety of biophysical techniques like single-molecule and surface plasmon resonance

119

Chapter 6

spectroscopy. In Chapter 4 and 5, strategies for the non-covalent and covalent oriented- immobilization of CYP3A4 were described respectively. These will undoubtedly open up new opportunities to exploit and study this P450 isoform in some of the applications mentioned above. In fact, biophysical investigations toward eventual single-molecule fluorescence microscopy experiments that require oriented-immobilization onto silica microspheres are already underway in collaboration with the research group of Prof.

Gonzalo Cosa. These were described in Chapter 5.

6.4 Publications

1) Ménard, A., Huang, Y., Karam, P., Cosa, G. and Auclair, K. (2012) Site-specific

fluorescent labeling and oriented immobilization of a triple mutant of CYP3A4 via

C64. Bioconj. Chem. 23, 826.

2) Ménard, M., Fabra, C., Huang, Y. and Auclair, K. (2012) Type II ligands as chemical

auxiliaries to favor enzymatic transformations by P450 2E1. Chembiochem DOI:

10.1002/cbic.201200524.

6.5 Presentations

1) Ménard, A., Huang, Y., Fabra C. and Auclair, K. (2012) Improving biocatalysis with

P450 enzymes: bioconjugation and chemical auxiliary. Chemistry and Biochemistry

Graduate Research Conference (CBGRC), Montreal, QC, Canada, November 2011.

(Oral)

120

Chapter 6

2) Ménard, A., Karam, P., Cosa, G. and Auclair, K. Improving biocatalysis with P450

enzymes: bioconjugation and chemical auxiliary. Canadian Society for Chemistry

(CSC) Conference, Montreal, QC, Canada, June 2011. (Poster)

3) Ménard, A., Karam, P., Cosa, G. and Auclair, K. Orthogonal site-specific

bioconjugation of P450 enzymes and their reductase. CBGRC, Montreal, QC,

Canada, November 2010. (Poster)

4) Ménard, A., Karam, P., Cosa, G. and Auclair, K. Orthogonal site-specific

bioconjugation of P450 enzymes and their reductase. Quebec-Ontario Mini-

Symposium on Bioorganic and Organic Chemistry Conference (QOMSBOC), St.

Catharines, ON, Canada, October 2010. (Poster)

5) Ménard, A., Karam, P., Cosa, G. and Auclair, K. Site-specific bioconjugation of

P450 enzymes for applications in bionanotechnology. CSC Conference, Toronto,

ON, Canada, June 2010. (Poster)

6) Ménard, A. and Auclair, K. Site-specific bioconjugation of P450 enzymes for

applications in bionanotechnology. PROTEO Conference, Montreal, QC, Canada,

June 2010. (Poster)

7) Ménard, A. and Auclair, K. Site-specific bioconjugation of P450 enzymes for

applications in bionanotechnology. McGill Biophysical Chemistry Symposium,

Montreal, QC, Canada, May 2010. (Poster)

8) Ménard, A. and Auclair, K. Towards the site-specific bioconjugation of P450

enzymes. Centre in Green Chemistry and Catalysis (CGCC) Annual Meeting,

Montreal, QC, Canada, December 2009. (Poster)

121

Chapter 6

9) Ménard, A. and Auclair, K. Site-specific bioconjugation of P450 enzymes for

applications in bionanotechnology. CBGRC, Montreal, QC, Canada, November 2009.

(Poster)

10) Ménard, A. and Auclair, K. Site-specific bioconjugation of P450 enzymes for

applications in bionanotechnology. QOMSBOC, Quebec, QC, Canada, October 2009.

(Poster)

11) Ménard, A. and Auclair, K. Towards the site-specific bioconjugation of P450

enzymes for nanotechnological applications. McGill Biophysical Chemistry

Symposium, Montreal, QC, Canada, May 2009. (Poster)

12) Ménard, A. and Auclair, K. Towards the site-specific bioconjugation of P450

enzymes for nanotechnological applications. CBGRC, Montreal, QC, Canada,

November 2008. (Poster)

13) Ménard, A. and Auclair, K. Towards the site-specific bioconjugation of P450

enzymes for nanotechnological applications. QOMSBOC, Toronto, ON, Canada,

October 2008. (Poster)

122

Chapter 7

Chapter 7 – Experimental protocols

7.1 General methods

7.1.1 Chemicals

All chemicals were purchased from Sigma-Aldrich or Alfa Aesar unless otherwise noted. 1,2-Dilauroyl-sn-glycero-3-phosphocholine (DLPC) was purchased from Avanti

Polar Lipids, Inc. (Alabaster, Alabama). Flash chromatography was performed on a

CombiFlash Rf system (Teledyne ISCO, Lincoln, NE) using RediSep Rf Gold columns.

Thin layer chromatography (TLC, analytical and preparatory) was performed using 60

F254 plates from EMD (Gibbstown, NJ).

7.1.2 Spectroscopy

The following abbreviations were used to describe NMR coupling patterns: s, singlet; br s, broad singlet; d, doublet; t, triplet; q, quartet; p, pentet; dd, doublet of doublet; dt, doublet of triple; ddd, doublet of doublet of doublet and m, multiplet.

Coupling patterns for certain spectra acquired on the 300 or 400 MHz instruments were sometimes deciphered from the spectra of similar compounds acquired on the 500 MHz instrument. Coupling constants are reported in Hertz (Hz). Chemical shifts are reported in parts per million (ppm). High resolution mass spectra (HRMS) were acquired on a

Thermo Fisher Scientific Inc. Exactive Orbitrap system. HPLC analysis was performed using an Agilent 1100 series modular system equipped with a vacuum degasser, an autosampler, a quaternary pump system, a UV-vis detector and in some cases a fluorescence detector. For LC-MS analysis, the modules were connected in-line with an

123

Chapter 7

Agilent 6120 quadrupole MS detector. Analytical separations were achieved with a

SYNERGI 4 μm HydroRP 80 Å (4.6 × 250 mm) analytical column purchased from

Phenomonex (Torrance, CA). Gas chromatography-mass spectra (GC-MS) were acquired on an Agilent 7890A GC system in line with an Agilent 5975C MS detector.

7.2 Biological studies

7.2.1 Molecular biology

7.2.1.1 Site-directed mutagenesis of CYP3A4

Cysteine mutants of CYP3A4 were generated using the QuikChangeTM Multi Site-

Directed Mutagenesis Kit (Stratagene, Agilent) and pSE3A4His encoding N-terminally truncated and tetrahistidine-tagged CYP3A4 wild-type cDNA as the template206. This plasmid was donated to us by Prof. J. R. Halpert. The following PCR primers were used: for C64S, 5’-GCTTTTGTATGTTTGACATGGAAAGTCATAAAAAGTATGGA

AAAGTG-3’; for C98S, 5’-CAAAACAGTGCTAGTGAAAGAAAGTTATTCTGTCTT

CACAAACCG-3’; for C239S, 5’-TCCCAATTCTTGAAGTATTAAATATCAGTGTG

TTTCCAAGAGAAGT-3’; for C337S, 5’-CTATGAGACTTGAGAGGGTCAGCAAAA

AAGATGTTGAGATC-3’ and for C468G, 5’-GAACTTCTCCTTCAAACCTGGTA

AAGAAACACAGATCCC-3’. The following mutants were constructed: C98S (mutant

1), C98S/C377S/C468G (mutant 2), C98S/C239S/C468G (mutant 3) and C64S/C98S

(mutant 4).

124

Chapter 7

7.2.1.2 Addition of His6-tag to CYP2E1

A C-terminal hexahistidine tag was generated by insertion mutagenesis and using pCW-2E194 as the template. This plasmid was kindly donated to us by Prof. F. Peter

Guengerich. The following primers were used along with the QuikChangeTM II XL mutagenesis kit (Stratagene, Agilent): 5’-CATTCCCCGCTCACATCACCATCATCATC

ATTGAGTGTGTGGAGGACACCCTGAACC-3’ and 5’-ATGTGAGCGGGGAATG

ACACAGAGTTTGTAACGTGGTGGGATACAG-3’. These primers contain non- overlapping regions and were designed according to Liu and Naismith222. Primers designed as per the instructions provided with the QuikChangeTM kit did not produce the desired insertion.

7.2.2 Expression and purification of enzymes

For the macromolecular crowding studies, CYP3A4, CYP2D6 and CPR were expressed and purified according to previously published procedures119. In all other studies, CYP3A4 purification was slightly modified143. CYP2E1 was expressed and purified very similarly to CYP3A4143 with a few minor changes. During the expression of

CYP2E1, the growth media was supplemented with thiamine-HCl (10 mM) and trace elements (188 µL per 750 mL of culture media). The trace elements solution was composed of 100 mM FeCl3, 10 mM ZnCl3, 10 mM Na2MoO4, 14.3 mM CaCl2, 7.4 mM

CuCl2 and 10 mM H3BO3 in 1/9 (v/v) conc. HCl/Milli-Q water. During the purification,

CYP2E1 was eluted from Ni-NTA with 500 mM imidazole (instead of 200 mM as reported in Ménard et al.143 for CYP3A4). Also, the ion-exchange chromatography step was omitted. For all above mentioned enzymes, the fractions of interest were identified by SDS-PAGE analysis. Desired fractions were pooled and dialyzed against 0.1 M

125

Chapter 7

potassium phosphate (KPi) at pH 7.4 containing 10% glycerol, except when macromolecular crowding studies where planned, no glycerol was added. Finally, the enzymes were stored at -80ºC for future use. The concentration of P450 was calculated from its reduced-CO spectrum1. The concentration of holo-CPR was estimated from its flavin content using an extinction coefficient of 21.4 mM-1•cm-1 at 456 nm223. Protein concentration was measured by the Bradford method (Pierce, 23236).

7.2.3 Chemical auxiliary studies with CYP2E1

7.2.3.1 Small scale biotransformation of auxiliary-substrates 2.1a-2.7a and 2.9a-2.23a with CYP2E1 for yield and regioselectivity estimates

CYP2E1 (100 pmol, 34.3 µL from a 2.9 µM stock in storage buffer, final concentration 1.9 µM) was diluted in Buffer B (0.1 M KPi at pH 7.4 containing 10% v/v glycerol, 15.6 µL) and pre-incubated at 37°C and 250 RPM for 5 min (reaction tubes were placed on their side in an orbital shaker/incubator). Inclusion of DLPC (20 µM) made no significant difference. Reactions were initiated by sequential addition of the substrate (45 µM, 1.2 µL from a 2 mM stock in acetonitrile) and cumene hydroperoxide

(CHP, 400 µM, 2.1 µL from a 10 mM stock in buffer containing 10% MeOH) and incubated for 20 min as above. Then, more CYP2E1 (100 pmol) and CHP (2.1 µL from a

10 mM stock) were added and the reaction was incubated for a further 25 min after which the products were extracted in EtOAc (3 × 500 µL). The extracts were combined and evaporated. The residue was redissolved in acetonitrile (30 µL) and analyzed by LC-MS.

Standards were also prepared by mixing each substrate (80 µM, 1.2 µL from a 2 mM stock in acetonitrile) with acetonitrile (28.8 µL) before LC-MS analysis. Analytical separation and quantification of the products was achieved using mobile phases A (Milli-

126

Chapter 7

Q water) and B (acetonitrile) at a flow rate of 0.5 mL/min with the UV detector set to monitor at 256 nm. Elution consisted of a linear increase from 50% phase B to 95% phase

B over a 20 min period followed by an isocratic step at 95% phase B. Yields were estimated by taking the ratio of the area under the product peak(s) to the area under the substrate peak (injected separately as a “no-reaction” standard) from their respective UV- vis spectra. Regioselectivity was estimated from the area under the product peaks from the UV-vis spectra, except for compound 2.18a in which case the SIM spectrum was used due to poor resolution in the UV-vis spectrum, and 2.2a in which case the NMR spectrum was also used since the major product 2.2b could not be separated from one of the minor products (2.2c).

7.2.3.2 Biotransformation of 1-hexanol (2.8a), 3-methyl-1-pentanol and cyclopentylmethanol by CYP2E1

Reaction conditions used were as described above for substrates 2.1a-2.7a and

2.9a-2.23a but scaled up 3 times. At the end of the incubation period, NaCl (~10 mg) was added to each reaction before extraction in chloroform (3 × 500 µL). Extracts were combined, evaporated, redissolved in chloroform (50 µL for 2.8a and 30 µL for the other two alcohols) and analyzed by GC-MS. Reaction yield for 2.8a was estimated by comparing the area under the product peak to that of an authentic standard (1,5 hexanediol) injected separately in the same amount as the substrate initially added to the reaction. The area of the latter was taken as that expected for a reaction with 100% yield.

Separations were achieved using an Agilent CYCLODEX-B column (60 m × 0.250 mm ×

0.25 µm). The temperature program consisted of an initial isocratic step at 100°C for 3 min, followed by a ramp to 150°C (20°C/min) and a final isocratic step at 150°C for 15

127

Chapter 7

min. The inlet temperature was set to 250°C and splitless injections were performed (1 µL for 2.8a and 3 µL for the other two alcohols). The carrier gas (helium) flow rate was

16.3316 cm/s (0.5 mL/min) and the detector temperature was 250°C.

7.2.3.3 Comparing extinction coefficients of 2.1a, 2.12a and 2.12b

The absorbance spectrum of each compound 2.1a, 2.12a and 2.12b was acquired at four different concentrations (15.6, 31.3, 62.5 and 125 µM) in KPi buffer (0.1 M at pH

7.4) in duplicate. This was repeated twice, each time with a different stock solution (20 mM in acetonitrile). Maximum absorbance at 263 nm was plotted against compound concentration. Slopes were calculated in Excel. The average slope values for these three compounds do not differ by more than ± 1 standard deviation.

2.1a 2.12a 2.12b slope 1 (M-1) 2813 3123 2567 slope 2 (M-1) 2658 2704 2760 average (M-1) 2700 2900 2700 standard deviation (M-1) 100 300 100

128

Chapter 7

7.2.3.4 Spectral binding studies with CYP2E1

All UV-Vis spectra of the enzyme-ligand complexes were obtained on a Carry

5000 UV-vis spectrophotometer. CYP2E1 (145 pmol, 50 µL from a 2.9 µM stock), KPi buffer (445 µL, 0.1 M KPi at pH 7.4 with 10% glycerol) and DLPC (20 µM, 5 µL from a

2 mM stock) were added to both the reference and sample cuvettes and a blank was taken.

Substrates 2.1a-2.8a, 2.10a-2.12a, 2.14a, 2.16a, 2.18a and 2.20a were titrated into the sample cuvette from stock solutions in acetonitrile such that the final acetonitrile concentration never exceeded 3% (v/v). These acetonitrile concentrations had no detectable effect on the P450 spectra. Equal volumes of acetonitrile were added to the reference cuvette and the difference spectra were acquired from 350-500 nm. The difference in absorbance at the peak and the trough was plotted against substrate concentration to obtain a binding curve. Spectral dissociation constants (Ks) were extracted from the binding curves by fitting to the following equation:

, where ΔA is the difference in absorbance between the peak and the trough,

ΔAmax is the maximum reachable value of ΔA at saturating substrate concentrations, [S] is the substrate concentration, and Ks is the substrate concentration at half saturation. Fitting was performed using the GraphPad Software.

7.2.3.5 Removal of the nicotinate auxiliary

Potassium hydroxide (10 mg, 100 µL from a 100 mg/mL solution in MeOH) was added to compound 2.12b (1.3 mg) and stirred at 37°C for 1 h. Reaction completion was confirmed by TLC (100% EtOAc).

129

Chapter 7

7.2.3.6 Docking studies with CYP2E1

Docking was performed using FITTED via the Web-based platform FORECASTER, both developed by Prof. N. Moitessier and his research group161. The PDB files (3E4E and 3LC4) were prepared for docking using the programs PREPARE and PROCESS, whereas the ligands were prepared using SMART. Docking was performed using FITTED in both flexible and rigid protein mode with default settings. All programs are freely available.

7.2.4 Macromolecular crowding studies with CYP3A4 and CYP2D6

7.2.4.1 Preparation of crowder and small sugar stock solutions

Stock solutions of polyvinylpyrrolidone 40 kDa (PVP 40), polyethylene glycol 6 kDa (PEG 6), dextran from Leuconostoc mesenteroides 70 kDa (DEX 70), dextran sodium sulfate from Leuconostoc spp. 500 kDa (DEX 500), bovine serum albumin 66 kDa (BSA), Ficoll® 70 kDa (FIC 70), D-(+)-sucrose, D-(+)-trehalose dihydrate, D-(+)- glucose, and D-(+)-melezitose monohydrate (Fisher Scientific) were made up in KPi buffer (0.1 M at pH 7.4) to a concentration of 50% (w/v). Complete dissolution was best achieved by applying low heat and slow stirring in a small beaker.

7.2.4.2 Effect of macromolecular crowders on CYP3A4 activity

Each enzymatic reaction contained CYP3A4 (0.4 μM), testosterone (50 μM) the desired macromolecular crowder (PVP 40, PEG 6, DEX 70, DEX 500, BSA or FIC 70) or small sugar (glucose, sucrose, trehalose, melezitose) and either CPR (1.6 μM)/NADPH (1 mM) or CHP (167 μM) in KPi buffer (0.1 M at pH 7.4) in a final volume of 300 μL. First,

CYP3A4, testosterone, the desired macromolecular crowder or small sugar, and CPR (if

130

Chapter 7

required) were combined and pre-incubated for 5 min at 37ºC. The reactions were initiated by the addition of NADPH or CHP and incubated at 37ºC and 250 RPM for 2 h.

After incubation, the samples were spiked with the internal standard cortexolone (10 μM,

15 μL from a 200 μM solution in MeOH) and extracted with EtOAc (1 mL). The solvent was then evaporated and the samples were redissolved in acetonitrile (300 μL) and analyzed by HPLC. Analytical separation and quantification of the product (relative to the internal standard) was achieved using mobile phases A (Milli-Q water + 0.05% TFA) and

B (acetonitrile + 0.05% TFA) at a flow rate of 1 mL/min with the UV detector set to monitor at 244 nm. Elution consisted of an initial isocratic step at 15% phase B for 4 min followed by a linear increase to 50% phase B over the next 20 min and return to 15% in

10 min. The retention times were 19.1 min for the product 6β-hydroxytestosterone, 24.5 min for the internal standard cortexolone and 30.6 min for the substrate testosterone.

7.2.4.3 Effect of macromolecular crowders on CYP2D6 activity

Each enzymatic reaction contained CYP2D6 (0.2 μM), dextromethorphan (100

μM) the desired macromolecular crowder (PVP 40, PEG 6, DEX 70, DEX 500, BSA or

FIC 70) or small sugar (glucose, sucrose, trehalose, melezitose) and either CPR (0.6

μM)/NADPH (1.43 mM) or CHP (167 μM) in KPi buffer (0.1 M and pH 7.4) in a final volume of 300 μL. Incubations were performed as described in section 7.2.4.2 above with

CYP3A4. After incubation, the samples were spiked with the internal standard levallorphan (10 μM, 17 μL from a 178 μM solution in MeOH) and extracted with 95% hexanes-5% n-butanol (1 mL). The solvent was then evaporated and the samples were redissolved in acetonitrile (100 μL) and analyzed by HPLC. Analytical separation and quantification of the product (relative to the internal standard) was achieved using mobile

131

Chapter 7

phases A (Milli-Q water + 0.05% TFA) and B (acetonitrile + 0.05% TFA) at a flow rate of 0.5 mL/min and the fluorescence detector set to monitor at 310 nm with an excitation wavelength of 280 nm. Elution consisted of a linear gradient from 40% phase B to 60% phase B in 7 min, followed by an isocratic step for 5 min and a linear increase to 90% phase B in 10 min. The retention times were 6.7 min for the product dextrorphan, 8.4 min for the internal standard levallorphan and 10.4 min for the substrate dextromethorphan.

7.2.5 C-terminal His-tag immobilization of CYP3A4

7.2.5.1 Preparation of Ni-NTA and Ni-free NTA agarose resin

An aliquot of Ni-NTA agarose (Qiagen) was equilibrated with KPi buffer (0.1 M at pH 7.4). A different aliquot, to be used as a control, was stripped of nickel by washing with a solution of EDTA (0.1 M at pH 8 in Milli-Q water) followed by copious amounts of Milli-Q water and finally KPi buffer (0.1 M at pH 7.4).

7.2.5.2 Yield of CYP3A4 immobilization on Ni-NTA agarose resin

To assess the yield of CYP3A4 immobilization onto the Ni-NTA resin, the enzyme activity present in the supernatant was assayed using 7-benzyloxy-4- trifluoromethylcoumarin (BFC). The debenzylation of BFC to the fluorescent product 7- hydroxy-4-trifluoromethylcoumarin (HFC) was monitored over time with a microtiter plate fluorimeter (SpectraMax Gemini XS). First, CYP3A4 (0.3 µM, 3 µL from 15 µM stock), CPR (0.9 µM, 18 µL from 8 µM stock), BFC (45 µM, 30 µL from 226 µM stock) and Ni-NTA agarose resin (20 µL from a 50-50 slurry (v/v) in KPi buffer (0.1 M at pH

7.4) were combined in same buffer to give a final volume of 150 µL, mixed gently and centrifuged for 5 min at 5000 RPM. (Note: the substrate BFC is not required to be present

132

Chapter 7

for immobilization to occur). An aliquot of the supernatant (100 µL) was removed and transferred to a 96-well plate and diluted with KPi buffer (44 µL, 0.1 M and pH 7.4).

After a 5 min pre-incubation period at 37°C, the reaction was initiated by addition of

NADPH (1 mM, 6 µL from 25 mM stock). Product formation was monitored over time with an excitation wavelength of 410 nm and an emission wavelength of 530 nm.

Controls were also performed using Ni-free resin, no resin and no CYP3A4.

7.2.5.3 Kinetic stability of CYP3A4 on Ni-NTA agarose resin

Each enzymatic assay contained CYP3A4 (0.4 µM), testosterone (50 µM), Ni-

NTA or Ni-free agarose (20 µL from a 50-50 slurry) and CHP (100 µM) in KPi buffer

(0.1 M at pH 7.4) in a total volume of 75 µL. CYP3A4, testosterone, resin and buffer were combined and pre-incubated for 5 min at 37°C. The reactions were initiated with the addition of CHP, incubated at 37°C and 225 RPM and quenched after 2, 5, 10, 15, 20, 25 or 30 min by the addition of the internal standard cortexolone (15 µL from 200 µM stock in MeOH) followed immediately by extraction with EtOAc (500 µL). The extracts were evaporated, redissolved in acetonitrile (75 µL) and analyzed by HPLC. Separation and quantification of the product (relative to the internal standard cortexolone) was achieved using mobile phases A (Milli-Q water) and B (acetonitrile) at a flow rate of 0.5 mL/min.

Elution consisted of an initial isocratic step at 50% phase B for 5 min followed by an increase to 70% phase B over the next 7 min and finally to 95% phase B over the final 8 min. The retention times were 7.1 min for the product 6β-hydroxytestosterone, 10.1 min for the internal standard cortexolone and 15.8 min for the substrate testosterone.

133

Chapter 7

7.2.5.4 Thermodynamic stability of CYP3A4 on Ni-NTA resin

Each reaction contained the same components and in the same amounts as those described in the previous section 7.2.5.3. CYP3A4, Ni-NTA or Ni-free agarose resin and buffer were combined first and incubated for 0, 2, 6, 18 and 24 h at 25°C. Then, testosterone was added followed by CHP after a 5 min pre-incubation at 37°C. Reactions were then quenched, extracted and analyzed by HPLC as in the previous section (7.2.5.3).

7.2.5.5 Effect of immobilization of CYP3A4 on Ni-NTA on its tolerance to lyophilization

The procedure was the same as in the previous section 7.2.5.4, except that instead of being incubated at 25°C, samples were frozen and lyophilized overnight. CYP3A4 activity post-lyophilization was assayed as in the previous section 7.2.5.4 after resolubilization in KPi buffer (0.1 M at pH 7.4).

7.2.6 Site-specific bioconjugation of CYP3A4 mutant 3

7.2.6.1 Cysteine-specific protein labeling with maleimides

Maleimide labels were purchased or generated from the corresponding amine as described below (section 7.2.6.2). Stock solutions (20 mM) of the maleimides were prepared in anhydrous DMSO. In a small glass vial, wild-type or mutant CYP3A4

(typically ~10 μM, based on Bradford assay) was combined with 9 equivalents of TCEP

(10 mM stock in Milli-Q water) in KPi buffer (0.1 M at pH 7.4), and incubated at 25ºC and 250 RPM for 20 min. Then, 100 equivalents of the desired maleimide label were added, and the reaction was incubated for 2 h under the same conditions. Note, equivalents are reported relative to the concentration of CYP3A4. The reaction was

134

Chapter 7

quenched by the addition of DTT (2 mM, from a 0.1 M stock solution in Milli-Q water) and incubation for 5-10 min. Control reactions were also performed in which pure anhydrous DMSO was added instead of the maleimide label with and without the 2 h incubation period.

7.2.6.2 Generation of maleimide labels from their corresponding amines

Maleimides 5.1b, 5.1c, 5.2c, 5.3b, and 5.3c were prepared as follows (see synthetic procedures in section 7.3.3 for full product characterization). All other maleimides were purchased and used as is. In a typical reaction, heterobifunctional crosslinker 5.15b or 5.15c (6-23 nmol) was combined with 1.05 equivalents of 5.1a, 5.2a, or 5.3a in anhydrous DMSO (final concentration of ~45 mM) and stirred at 25°C. The reaction was monitored by ESI-MS and was usually complete within 4 h after which the crosslinker (15b or 15c) had been consumed and the major peak corresponded to the desired amide bond formation product between the amine and the NHS-ester. A small amount (10-15%) of double addition product (conjugate addition of the amine to the maleimide) was also observed. The products were stored at -20°C and used directly without purification. The final maleimide concentration was quantified spectrophotometrically from the difference in the amount of 4-thiopyridone product formed when 2-mercaptoethanol reacts with 4,4’-dithiodipyridine in the presence and absence of the maleimide reaction mixture (maleimides react with 2-mercaptoethanol but not with 4,4’-dithiodipyridine)224. Briefly, an aliquot (5 µL) of the reaction mixture in

DMSO was diluted into KPi buffer (195 µL of 0.1 M at pH 7.4). A sample of this diluted stock (2-5 µL) was combined with 2-mercaptoethanol (10 µL from a 1.5 mM stock solution prepared fresh in Milli-Q water) and KPi buffer (0.1 M at pH 7.4), to a final

135

Chapter 7

volume of 450 µL. Control reactions containing only 2-mercaptoethanol were also prepared. These were incubated 10 min at room temperature after which a saturated solution of 4,4’-dithiodipyridine (50 µL in Milli-Q water) was added before incubation for another 15 min. Absorbance was read at 324 nm against a blank containing only 4,4’- dithiodipyridine. The concentration of the 4-thiopyridone formed from the reaction between 2-mercaptoethanol and 4,4’-dithiodipyridine in the presence and absence of the

-1 -1 maleimide was calculated using ε224nm = 19,800 M •cm . The difference gave the maleimide concentration in the reaction mixture.

7.2.6.3 Assessing the effect of quenched maleimide labels on enzymatic activity

Samples were prepared in the same manner described above for the cysteine- specific labeling (section 7.2.6.1) except that the maleimides were quenched with DTT prior to mixing with enzyme. The enzymatic activity was then assayed immediately using testosterone as the substrate and CPR/NADPH as the cofactors (end-point assay, see section 7.2.6.14), omitting the 2 h incubation period usually required for complete labeling. SDS-PAGE analysis followed by fluorescence visualization of the gels confirmed that CYP3A4 remained essentially label-free when the maleimides were pre- quenched in this manner.

7.2.6.4 Estimating the labeling yield of CYP3A4 mutant 3 with the maleimide DyLight 549 (5.10)

CYP3A4 mutant 3 was labeled with 10 or 100 equivalents of DyLight 549 maleimide as described above. The excess label was removed by size-exclusion using

Zeba Spin Desalting Columns (3 × 0.5 mL, Pierce). The absorbance spectrum of the flow through containing the labeled protein was obtained and the degree of labeling was

136

Chapter 7

-1 -1 calculated as per the manufacturer’s instructions using ε280nm = 40,340 M •cm (ExPASy

ProtParam tool) to determine the protein concentration.

7.2.6.5 Sample preparation for single-molecule photobleaching analysis

An aliquot of DyLight 549 maleimide-labeled wild-type or mutant CYP3A4 (see section 7.2.6.1) was combined with SDS sample buffer (4 μL of 0.13 M Tris pH 6.8, 20% glycerol, 4% SDS, 1.25% bromophenol blue, 1.4 M 2-mercaptoethanol) and boiled for 1 min. The sample was then loaded onto a precast 12.5% polyacrylamide gel and subjected to SDS-PAGE (with PhastSystem, Pharmacia) in order to ensure complete removal of excess label. The fluorescent band of appropriate molecular weight (ca. 57 kDa) was excised and subjected to syringe-maceration-extraction225 in order to extract the labeled protein. Coverslips were cleaned in Piranha solution following a procedure reported earlier and used without any further modification of the surface226. Predrilled polycarbonate films with an adhesive gasket (Grace Bio-Labs, Bend, OR) were assembled on top of the cleaned coverslips yielding a chamber with a total volume of 10

µL. Inlet and outlet ports (Nanoport, Upchurch Scientific, Oak Harbor, WA) were glued on top of the chambers226-227. In order to increase the dye photostability, all experiments were run under a constant flow of an oxygen scavenger solution consisting of 2- mercaptoethanol (1% v/v), β-D(+)glucose (3% w/v), glucose oxidase (0.1 mg/mL), and catalase (0.02 mg/mL) 10,228-229. All experiments were conducted at room temperature (22-

23°C). Purified labeled proteins were diluted (ca. 1000 fold) and then injected into a clean coverslip until a satisfactory surface density was observed.

137

Chapter 7

7.2.6.6 Quantification of labeling via single-molecule photobleaching analysis

Total internal reflection fluorescence microscopy (TIRFM) was utilized to quantify the number of fluorescent dyes per single protein. The setup consisted of an inverted microscope (Olympus IX71) equipped with a laser-based TIRFM illumination module (IX2-RFAEVA-2, Olympus) coupled to a diode-pumped solid-state green laser with 532 nm output (CrystaLaser). The beam position was adjusted using the illuminator to attain total internal reflection through a 60×, N.A. 1.45, oil-immersion objective

(Olympus U PLAN SAPO). Fluorescence emission was collected through the objective and images were captured with an electron multiplied back illuminated charged coupled device (EMCCD) camera (CascadeII:512, Photometrics, Roper Scientific). The camera was controlled using Image-Pro Plus 5.1 (Media Cybernetics), capturing 8-bit 512 × 512 pixel images with an exposure time of 200 ms per frame, a conversion gain of 3, and a multiplication gain of 4095. Excitation was carried out at a full power setting (25 mW) with a power output of 6.6 mW measured at the objective.

7.2.6.7 Quantification of labeling via single-molecule photobleaching analysis: Intensity analysis

A Matlab algorithm was utilized which first maps the ca. hundreds of single- molecule spots and following background subtraction it then extracts intensity vs. time trajectories from the stack of images for each individual molecule. Upon inspecting individual trajectories, photobleaching events were observed as steps. The number of fluorophores per protein was next calculated from the number of photobleaching steps recorded. The total number of trajectories analyzed was n = 2212, 1778 and 703 for

CYP3A4 wild-type, mutant 1 and mutant 3 respectively.

138

Chapter 7

7.2.6.8 In-gel BrCN digestion of labeled protein

An aliquot of the dye-labeled sample (20 μL) was combined with SDS sample buffer (4 μL, see section 7.2.6.5). Exceptionally, for dansyl-labeled (5.1b) samples, an aliquot (500 μL) was first concentrated (to 20 μL) with a centrifugal filter (10 kDa

MWCO, Millipore) before combination with SDS sample buffer (4 μL). Samples were boiled for 1 min and separated by SDS-PAGE using a classical vertical separation system.

The appropriate bands were excised, washed and treated with cyanogen bromide230. The resulting peptides were analysed by tricine-SDS-PAGE231. Optimal resolution was achieved with a 1.7 mm thick gel when a 10% spacer gel was included between the 4% stacking gel above and the 16%/6 M urea separating gel below. Gels were imaged with a

Typhoon Trio + scanner using the Green-excited mode (λex = 532 nm) and the 580 nm band-pass filter (580 BP 30, which transmits light between 565 and 595 nm with a transmission peak centered on 580 nm).

7.2.6.9 Preparation of maleimide-functionalized resins

Maleimide-functionalized resins were generated through modification at the terminal amine of CarboxyLink resin (5.4a, Pierce) with heterobifunctional crosslinkers

5.15b or 5.15c. Aliquots of the resin (1 mL) were washed with KPi buffer (3 × 2 mL, 0.1

M at pH 7.4) through centrifugation/resuspension cycles resulting ultimately in removal of as much supernatant as possible. The crosslinkers (5.15b or 5.15c, 22 µmol) were dissolved in anhydrous DMSO (0.5 mL), diluted with the same KPi buffer (0.5 mL) and immediately added to the resin. After brief vortexing, the resin was incubated at room

139

Chapter 7

temperature and ~100 RPM for 1.5 h to give 5.4b or 5.4c. Reaction completion was confirmed with a ninhydrin test232 as follows. A small aliquot of the resin was transferred to a glass vial and the supernatant was removed. One or two drops each of pyridine, 4:1 phenol:ethanol, and 4% (w/v) ninhydrin in EtOH were added to the resin which was then heated to 100°C for 2 min. No colour change indicated that all free amines on the resin had reacted with the crosslinkers. Unmodified resin turned deep blue under the same conditions. The resin was then washed with buffer (5 × 2 mL) to remove excess crosslinker. The degree of maleimide loading was quantified from mixtures of resin aliquots (20 µL, supernatant removed) and 4,4’-dithiodipyridine as described above in section 7.2.6.2. The degree of loading was generally ~12 nmol maleimide/20 µL resin.

7.2.6.10 Immobilization of CYP3A4 mutant 3 onto maleimide-functionalized resins

CYP3A4 (~1 nmol based on Bradford assay, ~0.18 nmol active P4501, from a stock in Buffer A) was combined with TCEP (9 nmol, 0.9 µL from a 10 mM stock in

Milli-Q water). The volume was adjusted to 100 µL with KPi buffer (0.1 M at pH 7.4) before incubation at 25°C and 250 RPM for 20 min. This TCEP-reduced enzyme solution was then added to an aliquot of resin (200 µL, supernatant removed, contained ca. 120 nmol maleimide) and incubated for 2 h. The enzyme activity was measured both in the supernatant and on the resin. To evaluate the immobilization yield, an aliquot of the supernatant collected after the immobilization reaction was assayed for P450 activity using BFC as the substrate (initial rate measurement, see section 7.2.6.15). The immobilization yield was calculated from the difference in P450 activity found in the supernatant versus the control containing unmodified resin 5.4a mixed with free mutant 3

140

Chapter 7

enzyme. This was also compared to the standard enzyme activity of mutant 3 in solution to rule out non-specific interaction of the enzyme with the resin. The activity of the enzyme immobilized on the resin was measured using the testosterone assay (end-point assay, see section 7.2.6.14) after washing the resin with KPi buffer (0.1 M at pH 7.4).

7.2.6.11 Preparation of maleimide-functionalized silica microspheres

To an aliquot (427 µL) of silica microspheres (0.1 µm, colloidal, Polysciences

Inc., ca. ~90 nM in beads) was added 3-(aminopropyl)trimethoxysilane (2 µL). The mixture was vortexed briefly and incubated at room temperature for 5 min. The reaction was then quenched with the addition of HCl (conc., 8 µL). The resulting aminosilanized beads (5.14a) were then washed with KPi buffer (3 × 1 mL, 0.1 M at pH 7.4) using centrifugation and resuspension cycles. The beads were finally resuspended in the same buffer (500 µL). The beads were then mixed with crosslinker 5.15c (22 µmol in 500 µL

DMSO) and incubated at 25°C and 250 RPM for 1.5 h. Reaction completion was confirmed using the ninhydrin test (see section 7.2.6.9). The beads were washed again to remove excess crosslinker 5.15c and resuspended in buffer (500 µL). The degree of maleimide loading was quantified from mixtures of bead aliquots (20 µL washed, supernatant included) and 4,4’-dithiodipyridine as described in section 7.2.6.2 above. The degree of loading was generally ~5 nmol maleimide/20 µL beads (supernatant included).

Finally, the microspheres were concentrated by centrifugation and resuspension in KPi buffer (200 µL).

141

Chapter 7

7.2.6.12 Immobilization of CYP3A4 mutant 3 onto maleimide-functionalized silica microspheres

Method A – for ensemble experiments. An aliquot of TCEP-reduced CYP3A4 mutant 3 (43 µL, see section 7.2.6.10) was combined with the maleimide-functionalized microsphere suspension prepared as in section 7.2.6.11 above (100 µL, contained ca. ~54 nmol maleimide) and incubated at 25°C and 250 RPM for 2 h. Controls were also performed in which the beads were quenched with DTT (2 mM from a 0.1 M stock solution in Milli-Q water) prior to combining with mutant 3, or in which no beads were included. At the end of the incubation period, the reactions were quenched with DTT and the enzyme activity in the supernatant was assayed with BFC (initial rate measurement, see section 7.2.6.15) in order to estimate the immobilization yield. The immobilization

(~70%) yield was calculated from the difference in P450 activity found in the supernatant versus the control containing pre-quenched resin. The beads were washed (2 × 1 mL) and resuspended in KPi buffer (final volume ~150 µL, 0.1 M at pH 7.4) before the activity of the immobilized enzyme was assayed using either testosterone or BFC (end-point assays, see sections 7.2.6.14 and 7.2.6.16 respectively).

Method B – for single-molecule feasibility studies with confocal fluorescence microscopy. Silica microspheres were aminosilanized and derivatized with crosslinker

5.15c as in section 7.2.6.11 above. CYP3A4 mutant 3 (1 mL, ~38 nmol based on

Bradford assay, ~6.8 nmol active P4501) was combined with TCEP (34 nmol, 3.4 µL from 10 mM stock in Milli-Q water) and incubated at 25°C and 250 RPM for 20 min. The

TCEP-reduced CYP3A4 was then combined with an aliquot of maleimide-functionalized microspheres (5.14c, 100 µL) and incubated for at 25°C and 250 RPM for 2 h. Remaining maleimides were quenched with DTT (2 mM from a 0.1 M stock in Milli-Q water) and

142

Chapter 7

the immobilization yield was ~10% (estimated as in Method A). The beads were then washed with KPi buffer (3 × 1 mL, 0.1 M at pH 7.4) and resuspended in same buffer (200

µL). Free thiols on the bead (from the DTT used to quench the remaining maleimides) were then labeled with DyLight maleimide 650 (5.13, 200 pmol, 1 µL from 200 µM stock, ca. ~10 per bead). Tubes were incubated at 25°C and 250 RPM for 20 min, washed with KPi buffer (3 × 1 mL, 0.1 M at pH 7.4) and resuspended in same buffer (100 µL).

Activity of the enzyme on the beads was assayed as in section 7.2.6.15 except that each reaction contained CYP3A4-modified beads (10 µL, ca. ~70 pmol P450). Finally, the beads were diluted 1000 times in KPi buffer (0.1 M at pH 7.4) before analysis on the confocal fluorescence microscope (see section 7.2.6.13).

7.2.6.13 Silica microsphere immobilized-CYP3A4 mutant 3 activity assays with confocal fluorescence microscopy: towards single-molecule studies

Coverslips were cleaned, and the chambers and ports assembled as in section

7.2.6.5. The confocal setup was built using an Olympus IX-71 inverted microscope equipped with a closed-loop sample scanning stage (Nano LP100, Mad City Labs,

Madison, WI) with a movement range of 100 µm controlled by a home-written Labview routine. Samples were excited continuously using either a 406 nm laser (WSTech) or a

633 nm HeNe laser (CVI Melles Griot) source. The circularly polarized laser beams were introduced via a single mode fiber optic and transmitted by a dichroic beamsplitter (z406-

633rdc, Chroma, Rockingham, VT) to the sample through a high numerical aperture

(N.A. = 1.40) oil immersion objective (Olympus U PLAN SAPO 100X). Fluorescence emission was collected through the same objective and passed through a long pass filter

(430LP, Chroma, Rockingham, VT) and a 633 nm holographic Raman notchplus filter

143

Chapter 7

(Kaiser Optical Systems, Ann Arbor, MI) to remove residual 406 nm and 633 nm laser light, respectively. The emission was then transmitted or reflected by a beam splitter

(640DCXR Chroma, Rockingham, VT) onto two avalanche photodiode detectors (Perkin

Elmer Optoelectronics SPCM-AQR-14, Vaudreuil, QC). Images consisting of 256 × 256 pixels were acquired by collecting the emission for 1 ms dwell times at each pixel.

Intensity vs. time trajectories were acquired by positioning the stage on single beads and recording the photon counts during every 1 ms time interval. A National Instruments NI-

PCI-6602 board was used as a counter board.

A solution containing the silica microspheres bearing CYP3A4 mutant 3 and the fluorescent DyLight label (5.13) was injected into the sample cell on the glass coverslip

(total volume 10 µL) and washed with KPi buffer. The buffer was then replaced by a solution containing BFC (130 µM) and CHP (1 mM). The beads were located by raster scanning an area (10 × 10 µm) with the 633 nm laser in order to visualize emission from the 5.13 label in the red channel (>640 nm) using a laser power of 0.7 µW. Single beads were then excited continuously with the 406 nm laser (8 µW) and intensity vs. time trajectories acquired for a total of 10 min per bead. A total of 7 beads were excited one at a time. The same area was then scanned with the 406 nm laser with a laser power of 2

µW to visualize HFC emission in the green channel (<640 nm). As control experiments, intensity vs. time transients were also acquired in the absence of BFC and/or CHP, in the presence of denatured CYP3A4 instead of active enzyme, and with BFC replaced by the non-fluorogenic CYP3A4 substrate testosterone. All experiments were performed at room temperature (22-23°C).

144

Chapter 7

7.2.6.14 CYP3A4 activity assays with testosterone

End-point CYP3A4 activity was assayed from the conversion of testosterone to

6β-hydroxytestosterone. To assay the activity of CYP3A4 in solution, each enzymatic reaction contained P450 (0.4 μM), testosterone (50 μM) and either CPR (1.6

µM)/NADPH (2 mM), or cumene hydroperoxide (CHP, 100 µM) in KPi buffer (0.1 M at pH 7.4) in a final volume of 75 μL. For agarose resin-immobilized P450, each reaction contained P450 (ca. ~0.4 µM, 70 μL resin, supernatant removed), testosterone (667 μM) and either CPR (1.6 µM) and NADPH (1 mM) or CHP (1 mM) in KPi buffer (0.1 M at pH 7.4) in a final volume of 150 μL. Similarly, for microsphere-immobilized P450, each reaction contained P450 (ca. ~0.4 µM, 150 µL microsphere suspension), testosterone (667

µM) and CHP (1 mM) in KPi buffer (0.1 M at pH 7.4) in a final volume of 300 µL. The buffer, testosterone, P450 and CPR (if required) were combined and pre-incubated for 5 min at 37ºC. The reactions were initiated by the addition of NADPH or CHP, and the reaction was allowed to proceed for 1 h at 37ºC with shaking at 250 RPM. At the end of the reaction, the tubes were spiked with internal standard cortexolone (5 μL from a 200

μM stock solution in MeOH) and extracted with EtOAc (0.5 mL). The extracts were evaporated and redissolved in acetonitrile (75 µL for the resin or 150 µL for the microspheres) before HPLC analysis. Separation and quantification of the product

(relative to the internal standard cortexolone) was achieved using mobile phases A (Milli-

Q water) and B (acetonitrile) at a flow rate of 0.5 mL/min with the UV detector set to monitor at 244 nm. Elution consisted of an initial isocratic step at 50% phase B for 5 min, followed by an increase to 70% phase B over the next 7 min and finally to 95% phase B over the final 8 min. The retention times were 7.1 min for the product 6β-

145

Chapter 7

hydroxytestosterone, 10.1 min for the internal standard cortexolone and 15.8 min for the substrate testosterone.

7.2.6.15 Determination of initial rates for CYP3A4 using the substrate 7- benzyloxy-4-trifluoromethylcoumarin (BFC)

To assay initial enzyme activity rates, the debenzylation of BFC to the fluorescent product HFC was monitored over time with a microtiter plate fluorimeter (SpectraMax

Gemini XS). CYP3A4 (~50 nM), CPR (150 nM) and BFC (30 μM) were first combined in KPi buffer (0.1 M at pH 7.4, used to adjust the final reaction volume to 150 μL) and pre-incubated for 5 min at 37ºC. The reaction was initiated with the addition of NADPH

(1 mM). Product formation was monitored over time with an excitation wavelength of

410 nm and an emission wavelength of 530 nm.

7.2.6.16 End-point activity determination for CYP3A4 using the substrate 7- benzyloxy-4-trifluoromethylcoumarin (BFC)

End-point activity of microsphere-immobilized CYP3A4 mutant 3 was also assayed with BFC. Each enzymatic reaction contained P450 (ca. ~0.4 µM from a 150 µL microsphere suspension), BFC (130 µM) and CHP (1 mM) in KPi buffer (0.1 M at pH

7.4) in a final volume of 300 µL. The buffer, BFC and P450 were combined and pre- incubated for 5 min at 37ºC. The reactions were initiated by the addition of CHP, and the reaction was allowed to proceed for 1 h at 37ºC with shaking at 250 RPM. At the end of the reaction, the tubes were spiked with the internal standard 7-hydroxycoumarin (5 μL from a 200 μM stock solution in EtOAc) and extracted with EtOAc (0.5 mL). The extracts were evaporated and redissolved in acetonitrile (150 µL) before HPLC analysis.

Separation and quantification of the product (relative to the internal standard 7-

146

Chapter 7

hydroxycoumarin) was achieved using using mobile phases A (Milli-Q water) and B

(acetonitrile) at a flow rate of 0.5 mL/min with the UV detector set to monitor at 335 nm.

Elution consisted of a linear gradient from 50% phase B to 95% phase B over 20 min followed by an isocratic step at 95% phase B for 5 min. The retention times were 10.1 min for the product HFC, 6.5 min for the internal standard 7-hydroxycoumarin and 20.8 min for the substrate BFC.

7.3 Synthetic procedures

7.3.1 Synthesis and characterization of auxiliary-substrates

General procedure for synthesis of compounds 2.1a-2.5a and 2.9a-2.23a. Nicotinic acid (123 mg, 1 mmol) was dissolved in DMF (5 mL). EDC-HCl (230 mg, 1.2 mmol), the desired alcohol (1.1 mmol) and DMAP (~5 mg, catalytic) were added and the reaction was stirred at room temperature overnight. DMF was then evaporated in vacuo, the crude mixture was redissolved in EtOAc (~20 mL) and washed with NaHCO3 (sat., 3 × 20 mL) and brine (1 × 20 mL). Finally, the organic layer was dried over anhydrous sodium sulfate and concentrated to afford the desired product. Some crude products required further purification (see below for each specific case) others were >95% pure as estimated by

NMR and HPLC and used directly.

Hexyl nicotinate (2.1a). Light yellow oil (173 mg, 84%).

1 Rf = 0.16 (90% hexanes-EtOAc); H NMR (300 MHz,

CDCl3): δ 9.23 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.77 (dd, 1H, J = 1.7, 4.9 Hz, H-7), 8.31

(ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-9), 7.40 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8), 4.35 (t, 2H,

147

Chapter 7

J = 6.7 Hz, H-1), 1.77 (m, 2H, H-2), 1.28-1.50 (m, 6H, H-3,4,5), 0.90 (t, 3H, J = 7.0 Hz,

13 H-6); C NMR (125 MHz, CDCl3): δ 165.2, 153.5, 150.6, 137.3, 126.5, 123.4, 65.7,

+ 31.4, 28.6, 25.6, 22.5, 14.0; HRMS (ESI): Calculated for [C12H17NO2 + H] 208.13321;

Found 208.13269.

Hexyl picolinate (2.2a). Yellow oil (130 mg, 63%). Rf =

1 0.38 (80% hexanes-EtOAc); H NMR (300 MHz, CDCl3):

δ 8.76 (ddd, 1H, J = 0.9, 1.7, 4.8 Hz, H-7), 8.12 (ddd, 1H, J = 0.9, 1.2, 7.8 Hz, H-10),

7.84 (dt, 1H, J = 1.7, 7.8 Hz, H-9), 7.47 (ddd, 1H, J = 1.2, 4.8, 7.8 Hz, H-8), 4.41 (t, 2H, J

= 7.0 Hz, H-1), 1.82 (m, 2H, H-2), 1.27-1.50 (m, 6H, H-3,4,5), 0.88 (t, 3H, J = 7.0 Hz, H-

13 6); C NMR (125 MHz, CDCl3): δ 165.2, 149.8, 148.3, 137.0, 126.7, 125.1, 66.1, 31.4,

+ 28.6, 25.5, 22.4, 14.0; HRMS (ESI): Calculated for [C12H17NO2 + H] 208.13321; Found

208.13234.

Hexyl isonicotinate (2.3a). Crude product was purified

by flash chromatography (gradient from 100% hexanes to

70% hexanes-EtOAc) to afford a light yellow oil (147 mg, 71%). Rf = 0.18 (75%

1 hexanes-EtOAc); H NMR (300 MHz, CDCl3): δ 8.78 (d, 2H, J = 4.4 Hz, H-7), 7.85 (d,

2H, J = 4.4 Hz, H-8), 4.35 (t, 2H, J = 6.7 Hz, H-1), 1.77 (m, 2H, H-2), 1.28-1.50 (m, 6H,

13 H-3,4,5), 0.90 (t, 3H, J = 6.9 Hz, H-6); C NMR (75 MHz, CDCl3): δ 165.1, 150.5,

137.6, 122.8, 65.9, 31.3, 28.5, 25.6, 22.5, 13.9; HRMS (ESI): Calculated for [C12H17NO2

+ H]+ 208.13321; Found 208.13262.

148

Chapter 7

Hexyl quinoline-3-carboxylate (2.4a). Crude

product was purified by prep-TLC (75% hexanes-

EtOAc) to afford light yellow crystals (223 mg, 87%). Rf = 0.26 (90% hexanes-EtOAc);

1 H NMR (300 MHz, CDCl3): δ 9.44 (s, 1H, H-8), 8.83 (s, 1H, H-10), 8.15 (d, 1H, J = 8.5

Hz, H-15), 7.93 (d, 1H, 7.8 Hz, H-12), 7.82 (m, 1H, H-14), 7.61 (m, 1H, H-13), 4.40 (t,

2H, J = 6.6 Hz, H-1), 1.81 (m, 2H, H-2), 1.47 (m, 2H, H-3), 1.35 (m, 4H, H-4,5), 0.90 (t,

13 3H, J = 6.4 Hz, H-6); C NMR (75 MHz, CDCl3): δ 165.4, 150.0, 149.7, 138.7, 131.8,

129.4, 129.1, 127.4, 126.8, 123.3, 65.7, 31.4, 28.6, 25.7, 22.5, 14.0; HRMS (ESI):

+ Calculated for [C16H19NO2 + H] 258.14886; Found 258.14821.

Hexyl benzoate (2.5a). Crude product was purified by

flash chromatography (gradient from 100% hexanes to

70% hexanes-EtOAc) to afford a colourless oil (194 mg, 94%). Rf = 0.56 (90% hexanes-

1 EtOAc); H NMR (300 MHz, CDCl3): δ 8.05 (m, 2H, H-8), 7.55 (m, 1H, H-10), 7.43 (m,

2H, H-9), 4.31 (t, 2H, J = 6.7 Hz, H-1), 1.76 (m, 2H, H-2), 1.28-1.52 (m, 6H, H-3,4,5),

13 0.90 (t, 3H, J = 7.0 Hz, H-6); C NMR (75 MHz, CDCl3): δ 166.7, 133.8, 130.5, 129.5,

+ 128.3, 65.1, 31.5, 28.7, 25.7, 22.6, 14.0; HRMS (ESI): Calculated for [C13H18O2 + H]

207.13796; Found 207.13776. The 1H NMR and 13C NMR spectra for this known compound were in agreement with those found in the AIST database233.

N-Hexylnicotinamide (2.6a). Nicotinic acid (615 mg, 5

mmol), EDC-HCl (1.15 g, 6 mmol), HOBt (811 mg, 6 mmol) and DIEA (3 mL, 15 mmol) were combined in DMF (20 mL). Hexylamine (793

149

Chapter 7

uL, 6 mmol) was added and the reaction was stirred overnight at room temperature. DMF was then evaporated in vacuo, the crude mixture was redissolved in EtOAc (~40 mL) and washed with NaHCO3 (sat., 3 × 40 mL) and brine (1 × 40 mL). Finally, the organic layer was dried over anhydrous sodium sulfate and concentrated. The product was purified by passing through a short silica plug (100% EtOAc) to afford a dark yellow wax (789 mg,

1 76%). Rf = 0.18 (40% hexanes-EtOAc); H NMR (300 MHz, CDCl3): δ 8.95 (dd, 1H, J =

0.9, 2.3 Hz, H-11), 8.68 (dd, 1H, J = 1.7, 4.8 Hz, H-7), 8.10 (ddd, 1H, J = 1.7, 2.3, 7.9

Hz, H-9), 7.35 (ddd, 1H, J = 0.9, 4.8, 7.9 Hz, H-8), 6.54 (br s, 1H, NH), 3.43 (m, 2H, H-

1), 1.60, (m, 2H, H-2), 1.22-1.41 (m, 6H, H-3,4,5), 0.87 (t, 3H, J = 7.0 Hz, H-6); 13C

NMR (75 MHz, CDCl3): δ 165.7, 151.8, 148.0, 135.1, 130.5, 123.4, 40.2, 31.4, 29.4,

+ 26.6, 22.5, 14.0; HRMS (ESI): Calculated for [C12H18N2O + H] 207.14919 ; Found

207.14877.

Hexyl-1H-imidazole-4-carboxamide (2.7a). Imidazole-

4-carboxylic acid (112 mg, 1 mmol) was dissolved in

DMF (5 mL). EDC-HCl (230 mg, 1.2 mmol), the desired alcohol (1.1 mmol) and DMAP

(catalytic) were added and the reaction was stirred at room temperature overnight. DMF was then evaporated in vacuo, the crude mixture was redissolved in EtOAc (~20 mL) and washed with NaHCO3 (sat., 3 × 20 mL) and brine (1 × 20 mL). Finally, the organic layer was dried over anhydrous sodium sulfate and concentrated. The crude product was purified by flash chromatography (gradient from 0-10% MeOH in DCM) to afford a

1 white powder (147 mg, 75%). Rf = 0.1 (40% hexanes-EtOAc); H NMR (300 MHz,

CDCl3): δ 8.16 (s, 1H, H-7), 7.77 (s, 1H, H-10), 5.17 (br s, 1H, NH), 4.30 (t, 2H, J = 6.8

Hz, H-1), 1.73 (m, 2H, H-2), 1.24-1.46 (m, 6H, H-3,4,5), 0.89 (t, 3H, J = 6.8 Hz, H-6);

150

Chapter 7

13 CNMR (75 MHz, CDCl3): δ 161.7, 137.0, 129.0, 126.0, 65.1, 31.4, 28.7, 25.6, 22.5,

+ 14.0; HRMS (ESI): Calculated for [C10H16N2O2 + H] 197.12845; Found 197.12829.

4-Methylpentyl nicotinate (2.9a). Light yellow oil (126 mg,

1 61%). Rf = 0.13 (90% hexanes-EtOAc); H NMR (300 MHz,

CDCl3): δ 9.23 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.77 (dd, 1H, J = 1.7, 4.9 Hz, H-7), 8.30

(ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-9), 7.40 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8), 4.34 (t, 2H,

J = 6.7 Hz, H-1), 1.71-1.84 (m, 2H, H-2), 1.53-1.69 (m, 1H, H-4), 1.26-1.37 (m, 2H, H-

13 3), 0.92 (d, 6H, J = 6.6 Hz, H-5); C NMR (125 MHz, CDCl3): δ 165.3, 153.2, 150.8,

137.1, 126.4, 123.3, 65.9, 35.0, 27.7, 26.5, 22.5; HRMS (ESI): Calculated for [C12H17NO2

+ H]+ 208.13321; Found 208.13283.

3-Methylpentyl nicotinate (2.10a). Crude product was

purified by prep-TLC (10% hexanes-EtOAc) to afford a light

1 yellow oil (151 mg, 73%). Rf = 0.27 (85% hexanes-EtOAc); H NMR (300 MHz, CDCl3):

δ 9.21 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.75 (dd, 1H, J = 1.7, 4.9 Hz, H-7), 8.27 (ddd, 1H,

J = 1.7, 2.1, 8.0 Hz, H-9), 7.37 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8), 4.37 (m, 2H, H-1),

1.79 (m, 1H, H-3), 1.47-1.64 (m, 2H, H-2), 1.14-1.47 (m, 2H, H-4), 1.94 (d, 3H, J = 6.4

13 Hz, H-6), 0.89 (t, 3H, J = 7.4 Hz, H-5); C NMR (75 MHz, CDCl3): δ 165.3, 153.2,

150.8, 137.0, 126.4, 123.3, 64.0, 35.0, 31.5, 29.4, 19.0, 11.2; HRMS (ESI): Calculated for

+ [C12H17NO2 + H] 208.13321; Found 208.13268.

151

Chapter 7

5-Methylhexyl nicotinate (2.11a). Light yellow oil (106

1 mg, 48%). Rf = 0.26 (85% hexanes-EtOAc); H NMR

(300 MHz, CDCl3): δ 9.22 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.77 (dd, 1H, J = 1.7, 4.9 Hz,

H-7), 8.30 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-9), 7.40 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8),

4.35 (t, 2H, J = 6.7 Hz, H-1), 1.75 (m, 2H, H-2), 1.36-1.63 (m, 3H, H-3,5), 1.18-1.29 (m,

13 2H, H-4), 0.88 (d, 6H, J = 6.6 Hz, H-6); C NMR (75 MHz, CDCl3): δ 165.3, 153.2,

150.8, 136.9, 126.3, 123.2, 65.5, 38.4, 28.8, 27.8, 23.7, 22.4; HRMS (ESI): Calculated for

+ [C13H19NO2 + H] 222.14886; Found 222.14820.

Pentyl nicotinate (2.12a). Light yellow oil (168 mg, 87%).

1 Rf = 0.25 (85% hexanes-EtOAc); H NMR (300 MHz,

CDCl3): δ 9.22 (dd, 1H, J = 0.9, 2.1 Hz, H-10), 8.77 (dd, 1H, J = 1.7, 4.9 Hz, H-6), 8.30

(ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-8), 7.40 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-7), 4.34 (t, 2H,

J = 6.7 Hz, H-1), 1.77 (m, 2H, H-2), 1.31-1.49 (m, 4H, H-3,4), 0.92 (t, 3H, J = 7.3 Hz, H-

13 5); C NMR (125 MHz, CDCl3): δ 165.2, 153.2, 150.8, 137.0, 126.3, 123.2, 65.5, 28.3,

+ 28.1, 22.3, 13.9; HRMS (ESI): Calculated for [C11H15NO2 + H] 194.11756; Found

194.11730.

Heptyl nicotinate (2.13a). Light yellow oil (169 mg,

1 76%). Rf = 0.26 (85% hexanes-EtOAc); H NMR (300

MHz, CDCl3): δ 9.23 (dd, 1H, J = 0.9, 2.1 Hz, H-12), 8.78 (dd, 1H, J = 1.7, 4.9 Hz, H-8),

8.32 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-10), 7.41 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-9), 4.35

(t, 2H, J = 6.7 Hz, H-1), 1.78 (m, 2H, H-2), 1.23-1.50 (m, 8H, H-3,4,5,6), 0.89 (t, 3H, J =

13 6.8 Hz, H-7); C NMR (75 MHz, CDCl3): δ 165.2, 153.2, 150.8, 137.0, 126.3, 123.2,

152

Chapter 7

+ 65.5, 31.6, 28.9, 28.6, 25.9, 22.5, 14.0; HRMS (ESI): Calculated for [C13H19NO2 + H]

222.14886; Found 222.14824.

Octyl nicotinate (2.14a). Light yellow oil (186 mg,

1 79%). Rf = 0.20 (85% hexanes-EtOAc); H NMR

(300 MHz, CDCl3): δ 9.23 (dd, 1H, J = 0.9, 2.1 Hz, H-13), 8.77 (dd, 1H, J = 1.7, 4.9 Hz,

H-9), 8.31 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-11), 7.40 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-

10), 4.35 (t, 2H, J = 6.7 Hz, H-1), 1.78 (m, 2H, H-2), 1.21-1.50 (m, 10H, H-3,4,5,6,7),

13 0.88 (t, 3H, J = 6.7 Hz, H-8); C NMR (75 MHz, CDCl3): δ 165.2, 153.2, 150.8, 136.9,

126.3, 123.2, 65.5, 31.7, 29.1 (2C), 28.5, 25.9, 22.5, 14.0; HRMS (ESI): Calculated for

+ [C14H21NO2 + H] 236.16451; Found 236.16382.

Pent-4-en-1-yl nicotinate (2.15a). Crude product was

purified by prep-TLC (40% hexanes-EtOAc) to afford a light

1 yellow oil (157 mg, 82%). Rf = 0.27 (85% hexanes-EtOAc); H NMR (300 MHz, CDCl3):

δ 9.20 (dd, 1H, J = 0.9, 2.1 Hz, H-10), 8.75 (dd, 1H, J = 1.7, 4.9 Hz, H-6), 8.27 (ddd, 1H,

J = 1.7, 2.1, 8.0 Hz, H-8), 7.37 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-7), 5.82 (m, 1H, H-4),

4.94-5.10 (m, 2H, H-5), 4.35 (t, 2H, J = 6.6 Hz, H-1), 2.20 (m, 2H, H-2), 1.86 (m, 2H, H-

13 3); C NMR (75 MHz, CDCl3): δ 165.1, 153.2, 150.7, 137.1, 136.9, 126.2, 123.2, 115.4,

+ 64.7, 30.0, 27.7; HRMS (ESI): Calculated for [C11H13NO2 + H] 192.10191; Found

192.10143.

153

Chapter 7

(Z)-Hex-3-en-1-yl nicotinate (2.16a). The starting alcohol

3-hexen-1-ol was purchased from Sigma-Aldrich as the 98% pure Z-isomer. product was purified by prep-TLC (40% hexanes-EtOAc) to afford a light

1 yellow oil (166 mg, 81%). Rf = 0.6 (40% hexanes-EtOAc); H NMR (400 MHz, CDCl3):

δ 9.21 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.76 (dd, 1H, J = 1.7, 4.9 Hz, H-7), 8.28 (ddd, 1H,

J = 1.7, 2.1, 8.0 Hz, H-9), 7.38 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8), 5.49-5.58 (m, 1H, H-

3), 5.33-5.43 (m, 1H, H-4), 4.34 (t, 2H, J = 6.9 Hz, H-1), 2.52 (m, 2H, H-2), 2.08 (m, 2H,

13 H-5), 0.95 (t, 3H, J = 7.5 Hz, H-6); C NMR (75 MHz, CDCl3): δ 165.3, 153.3, 150.9,

137.0, 134.9, 126.2, 123.4, 123.3, 65.9, 26.7, 20.6, 14.2; HRMS (ESI): Calculated for

+ [C12H15NO2 + H] 206.11756; Found 206.11705.

(Z)-Pent-2-en-1-yl nicotinate (2.17a). The starting alcohol

2-penten-1-ol was purchased from Sigma-Aldrich as the

95% pure Z-isomer. Crude product was purified by prep-TLC (40% hexanes-EtOAc) to

1 afford a light yellow oil (162 mg, 85%). Rf = 0.24 (85% hexanes-EtOAc); H NMR (300

MHz, CDCl3): δ 9.22 (dd, 1H, J = 0.9, 2.1 Hz, H-10), 8.76 (dd, 1H, J = 1.7, 4.9 Hz, H-6),

8.30 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-8), 7.38 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-7), 5.56-

5.77 (m, 2H, H-2,3), 4.90 (d, 2H, J = 6.6 Hz, H-1), 2.19 (m, 2H, H-4), 1.02 (t, 3H, J = 7.5

13 Hz, H-5); C NMR (75 MHz, CDCl3): δ 165.2, 153.3, 150.9, 137.8, 137.1, 126.2, 123.3,

+ 123.2, 61.2, 21.0, 14.1; HRMS (ESI): Calculated for [C11H13NO2 + H] 192.10056;

Found 192.10152.

154

Chapter 7

But-3-en-1-yl nicotinate (2.18a). Crude product was purified

by prep-TLC (40% hexanes-EtOAc) to afford a light yellow oil

1 (111 mg, 63%). Rf = 0.21 (85% hexanes-EtOAc); H NMR (300 MHz, CDCl3): δ 9.20

(dd, 1H, J = 0.9, 2.1 Hz, H-9), 8.75 (dd, 1H, J = 1.7, 4.9 Hz, H-5), 8.26 (ddd, 1H, J = 1.7,

2.1, 8.0 Hz, H-7), 7.36 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-6), 5.84 (m, 1H, H-3), 5.06-5.20

(m, 2H, H-4), 4.38 (t, 2H, J = 6.7 Hz, H-1), 2.51 (m, 2H, H-2); 13C NMR (75 MHz,

CDCl3): δ 165.2, 153.4, 150.9, 137.0, 133.7, 126.1, 123.2, 117.6, 64.3, 33.0; HRMS

+ (ESI): Calculated for [C10H11NO2 + H] 178.08626; Found 178.08643.

(E)-Hex-4-en-1-yl nicotinate (2.19a). The starting

alcohol 4-hexen-1-ol was purchased from Sigma-Aldrich as the 97% pure E-isomer. Crude product was purified by prep-TLC (40% hexanes-

1 EtOAc) to afford a light yellow oil (174 mg, 85%). Rf = 0.18 (85% hexanes-EtOAc); H

NMR (400 MHz, CDCl3): δ 9.13 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.67 (dd, 1H, J = 1.7,

4.9 Hz, H-7), 8.20 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-9), 7.29 (ddd, 1H, J = 0.9, 4.9, 8.0

Hz, H-8), 5.28-5.45 (m, 2H, H-4,5), 4.26 (t, 2H, J = 6.7 Hz, H-1), 2.05 (m, 2H, H-2), 1.74

13 (m, 2H, H-3), 1.54 (d, 3H, J = 4.9 Hz, H-6); C NMR (100 MHz, CDCl3): δ 165.1, 153.2,

150.8, 136.9, 129.7, 126.2, 125.9, 123.2, 64.8, 28.8, 28.3, 17.8; HRMS (ESI): Calculated

+ for [C12H15NO2 + H] 206.11756; Found 206.11699.

Cyclopentylmethyl nicotinate (2.20a). Light yellow oil (132

1 mg, 64%). Rf = 0.24 (85% hexanes-EtOAc); H NMR (400

MHz, CDCl3): δ 9.23 (dd, 1H, J = 0.9, 2.1 Hz, H-8), 8.78 (dd, 1H, J = 1.7, 4.9 Hz, H-4),

8.31 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-6), 7.40 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-5), 4.25

155

Chapter 7

(d, 2H, J = 7.5 Hz, H-1), 2.36 (m, 1H, H-2), 1.79-1.89 (m, 2H), 1.54-1.73 (m, 4H), 1.30-

13 1.41 (m, 2H); C NMR (125 MHz, CDCl3): δ 165.3, 153.2, 150.8, 137.1, 126.4, 123.3,

+ 69.3, 38.6, 29.4, 25.3; HRMS (ESI): Calculated for [C12H15NO2 + H] 206.11756; Found

206.11696.

3-Cyclopentylpropyl nicotinate (2.21a). Crude product

was purified by prep-TLC (40% hexanes-EtOAc) to afford

1 a light yellow oil (210 mg, 90%). Rf = 0.26 (85% hexanes-EtOAc); H NMR (300 MHz,

CDCl3): δ 9.21 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.76 (dd, 1H, J = 1.7, 4.9 Hz, H-7), 8.28

(ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-9), 7.39 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8), 4.33 (t, 2H,

J = 6.7 Hz, H-1), 1.70-1.85 (m, 5H, H-2,3,4), 1.37-1.66 (m, 6H), 1.00-1.17 (m, 2H); 13C

NMR (100 MHz, CDCl3): δ 165.3, 153.2, 150.8, 137.1, 126.4, 123.3, 65.8, 39.8, 32.6,

+ 32.3, 27.9, 25.1; HRMS (ESI): Calculated for [C14H19NO2 + H] 234.14886; Found

234.14798.

3-Phenylpropyl nicotinate (2.22a). Crude product was

purified by prep-TLC (40% hexanes-EtOAc) to afford a

1 light yellow oil (207 mg, 86%). Rf = 0.16 (85% hexanes-EtOAc); H NMR (400 MHz,

CDCl3): δ 9.21 (dd, 1H, J = 0.9, 2.1 Hz, H-12), 8.78 (dd, 1H, J = 1.7, 4.9 Hz, H-8), 8.26

(ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-10), 7.38 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-9), 7.28 (m,

2H, H-5), 7.20 (m, 3H, H-6,7), 4.37 (t, 2H, J = 6.4 Hz, H-1), 2.78 (m, 2H, H-2), 2.11 (m,

13 2H, H-3); C NMR (75 MHz, CDCl3): δ 165.2, 153.3, 150.8, 141.0, 137.1, 128.5, 128.4,

+ 126.3, 126.1, 123.3, 64.8, 32.3, 30.1; HRMS (ESI): Calculated for [C15H15NO2 + H]

242.11756; Found 242.11678.

156

Chapter 7

3-Cyclohexylpropyl nicotinate (2.23a). Crude product

was purified by prep-TLC (40% hexanes-EtOAc) to

1 afford a light yellow oil (67 mg, 27%). Rf = 0.27 (85% hexanes-EtOAc); H NMR (400

MHz, CDCl3): δ 9.22 (dd, 1H, J = 0.9, 2.1 Hz, H-12), 8.77 (dd, 1H, J = 1.7, 4.9 Hz, H-8),

8.29 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-10), 7.38 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-9), 4.32

(t, 2H, 6.7 Hz, H-1), 1.58-1.85 (m, 7H), 1.08-1.36 (m, 6H), 0.78-0.98 (m, 2H); 13C NMR

(75 MHz, CDCl3): δ 165.3, 153.1, 150.7, 137.1, 126.4, 123.3, 65.9, 37.3, 33.6, 33.3, 26.6,

+ 26.3, 26.0; HRMS (ESI): Calculated for [C15H21NO2 + H] 248.16451; Found 248.16385.

7.3.2 Synthesis and characterization of CYP2E1 oxidation products

General procedure for biosynthesis of compounds 2.1b-2.5b, 2.9b, 2.11b-14b, 2.18b and 2.20b-2.22b. CYP2E1 (0.9-1.9 µM, 10 mL, from a 7-14 µM stock) was diluted in

KPi buffer (57 mL, 0.1 M at pH 7.4) and pre-incubated at 37°C and 250 RPM (orbital shaking) for 15 min. Reactions were initiated by sequential addition of the substrate (500

µM, 1.9 mL, 20 mM stock in acetonitrile) and CHP (800 µM, 6 mL from a 10 mM stock in buffer containing 1% MeOH) before allowing to react for 45 min at 37°C and 250

RPM. The products were extracted in EtOAc (3 × 50 mL). The extracts were combined, dried over anhydrous sodium sulfate and evaporated under reduced pressure. The residues were redissolved in DCM (~500 µL) and the products purified by prep-TLC with hexanes-EtOAc as the mobile phase. Some products required further purification by reverse phase semi-preparatory HPLC. A Zorbax 300 SB-C8 (9.4 mm × 25 cm) column was used with phases A (Milli-Q water) and B (acetonitrile) with a flow rate of 3

157

Chapter 7

mL/min. Elution consisted of a linear increase from 50% phase B to 95% phase B over a

20 min period.

5-Hydroxyhexyl nicotinate (2.1b). Crude product was

purified by prep-TLC (100% EtOAc). Rf = 0.35 (100%

1 EtOAc); H NMR (300 MHz, CD3OD): δ 9.11 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.74 (dd,

1H, J = 1.7, 4.9 Hz, H-7), 8.40 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-9), 7.57 (ddd, 1H, J =

0.9, 4.9, 8.0 Hz, H-8), 4.37 (t, 2H, J = 6.5 Hz, H-1), 3.65-3.78 (m, 1H, H-5), 1.81 (m, 2H,

H-2), 1.57-1.37 (m, 4H, H-3,4), 1.14 (d, 3H, J = 6.2 Hz, H-6); 13C NMR from HSQC

(125 MHz, CD3OD): δ 152.1, 149.3, 136.9, 123.6, 67.1, 64.7, 38.1, 28.1, 26.1, 22.2;

+ HRMS (ESI): Calculated for [C12H17NO3 + H] 224.12812; Found 224.12748.

5-Hydroxyhexyl picolinate (2.2b) and 6-

hydroxyhexyl picolinate (2.2c). Crude product was

purified by prep-TLC (100% EtOAc) as a 60:40

mixture of regioisomers (several attempts to separate

1 by HPLC were fruitless). Rf = 0.25 (100% EtOAc); H NMR (300 MHz, CDCl3): δ for

2.2b 8.78 (ddd, 1H, J = 0.7, 1.7, 4.8 Hz, H-7), 8.13 (dt, 1H, J = 0.9, 7.8 Hz, H-10), 7.85

(dt, 1H, J = 1.7, 7.8 Hz, H-9), 7.47 (ddd, 1H, J = 1.2, 4.8, 7.8 Hz, H-8), 4.433 (t, 2H, J =

6.8 Hz, H-1), 3.83 (m, 1H, H-5), 1.81-1.90 (m, 2H, H-2), 1.41-1.66 (m, 4H, H-3,4), 1.20

(d, 3H, J = 6.2 Hz, H-6); δ for 2.2c 8.78 (ddd, 1H, J = 0.7, 1.7, 4.8 Hz, H-7), 8.13 (dt, 1H,

J = 0.9, 7.8 Hz, H-10), 7.85 (dt, 1H, J = 1.7, 7.8 Hz, H-9), 7.47 (ddd, 1H, J = 1.2, 4.8, 7.8

Hz, H-8), 4.426 (t, 2H, J = 6.8 Hz, H-1), 3.65 (t, 2H, J = 6.5 Hz, H-6), 1.81-1.90 (m, 2H,

13 H-2), 1.41-1.66 (m, 6H, H-3,4,5); C NMR from HSQC (125 MHz, CD3OD): δ 149.2,

158

Chapter 7

125.1, 137.9, 127.4, 66.9(2.2c), 65.7(2.2b, 2.2c), 61.4(2.2b), 38.3(2.2b, 2.2c), 32.1(2.2c),

28.3(2.2b, 2.2c), 25.4 (2.2b), 22.1(2.2b), 21.9(2.2c); HRMS (ESI): Calculated for

+ [C12H17NO3 + H] 224.12812; Found 224.12764.

5-Hydroxyhexyl isonicotinate (2.3b). Crude product was

purified by prep-TLC (100% EtOAc). Rf = 0.26 (100%

1 EtOAc). H NMR (300 MHz, CD3OD): δ 8.74 (d, 2H, J = 4.8 Hz, H-7), 7.94 (d, 2H, J =

4.8 Hz, H-8), 4.38 (t, 2H, J = 6.6 Hz, H-1), 3.69-3.79 (m, 1H, H-5), 1.76-1.88 (m, 2H, H-

2), 1.44-1.64 (m, 4H, H-3,4), 1.14 (d, 3H, J = 6.2 Hz, H-6); 13C NMR from HSQC (125

MHz, CD3OD): δ 152.7, 125.8, 68.5, 50.8, 41.1, 32.1, 31.1, 25.1; HRMS (ESI):

+ Calculated for [C12H17NO3 + H] 224.12812; Found 224.12758.

5-Hydroxyhexyl quinoline-3-carboxylate (2.4b).

Crude product was purified by prep-TLC (25%

1 hexanes-EtOAc). Rf = 0.38 (25% hexanes-EtOAc); H NMR (300 MHz, CD3OD): δ 9.34

(s, 1H, H-8), 9.02 (s, 1H, H-10), 8.12 (m, 2H, H-12,15), 7.92 (m, 1H, H-14), 7.72 (m, 1H,

H-13), 4.45 (t, 2H, J = 6.6 Hz, H-1), 3.71-3.82 (m, 1H, H-5), 1.80-1.94 (m, 2H, H-2),

1.47-1.70 (m, 4H, H-3,4), 1.17 (d, 3H, J = 6.2 Hz, H-6); 13C NMR from HSQC (125

MHz, CD3OD): δ 152.2, 142.0, 135.2, 132.2, 130.9, 130.6, 69.9, 68.3, 41.2, 31.3, 24.8,

+ 24.9, 25.0; HRMS (ESI): Calculated for [C16H19NO3 + H] 274.14377; Found 274.14337.

5-Hydroxyhexyl benzoate (2.5b). Crude product was

purified by prep-TLC (75% hexanes-EtOAc). Rf = 0.16

1 (75% hexanes-EtOAc); H NMR (500 MHz, CD3OD): δ 8.01 (m, 2H, H-8), 7.57-7.62 (m,

159

Chapter 7

1H, H-10), 7.45-7.49 (m, 2H, H-9), 4.33 (t, 2H, J = 6.6 Hz, H-1), 3.74 (m, 1H, H-5),

1.72-1.85 (m, 2H, H-2), 1.43-1.65 (m, 4H, H-3,4), 1.16 (d, 3H, J = 6.2 Hz, H-6) ; 13C

NMR from HSQC (125 MHz, CD3OD): δ 135.7, 131.9, 131.1, 69.9, 67.7, 41.3, 31.3,

+ 25.1, 24.9; HRMS (ESI): Calculated for [C13H18O3 + H] 223.13287; Found 223.13251.

4-Hydroxy-3-methylpentyl nicotinate (2.10b). Crude

product was purified by prep-TLC (40% hexanes-EtOAc). Rf

1 = 0.29 (40% hexanes-EtOAc); H NMR (500 MHz, CD3OD): δ 9.15 (dd, 1H, J = 0.9, 2.1

Hz, H-11), 8.77 (dd, 1H, J = 1.7, 4.9 Hz, H-7), 8.43 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-9),

7.60 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8), 4.41-4.53 (t, 2H, H-1), 3.62-3.69 (m, 1H, H-4),

2.03-2.13 (m, 1H, H-3), 1.54-1.76 (m, 2H, H-2), 1.18 (d, 3H, J = 6.2 Hz, H-5), 1.00 (d,

13 3H, J = 6.8 Hz, H-6); C NMR from HSQC (125 MHz, CD3OD): δ 155.4, 152.5, 140.2,

+ 126.7, 73.6, 66.8, 34.0, 32.1, 21.3, 16.7; HRMS (ESI): Calculated for [C12H17NO3 + H]

224.12812; Found 224.12769.

5-Hydroxy-5-methylhexyl nicotinate (2.11b). Crude

product was purified by prep-TLC (40% hexanes-

1 EtOAc). Rf = 0.26 (40% hexanes-EtOAc); H NMR (500 MHz, CD3OD): δ 9.12 (dd, 1H,

J = 0.9, 2.1 Hz, H-11), 8.74 (dd, 1H, J = 1.7, 4.9 Hz, H-7), 8.40 (ddd, 1H, J = 1.7, 2.1, 8.0

Hz, H-9), 7.56 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8), 4.39 (t, 2H, J = 6.6 Hz, H-1), 1.76-

1.83 (m, 2H, H-2), 1.51-1.57 (m, 2H, H-3,4), 1.19 (s, 6H, H-6); 13C NMR from HSQC

(125 MHz, CD3OD): δ 155.6, 152.6, 140.3, 126.7, 68.2, 45.8, 32.7, 30.6, 23.4; HRMS

+ (ESI): Calculated for [C13H19NO3 + H] 238.14377; Found 238.14314.

160

Chapter 7

4-Hydroxypentyl nicotinate (2.12b). When prepared

synthetically from nicotinic acid and 1,5-hexanediol, the crude product was passed through a short silica plug eluting with 100% EtOAc to give a light yellow oil (155 mg, 74%). When prepared enzymatically, the crude product was

1 purified by prep-TLC (100% EtOAc). Rf = 0.32 (100% EtOAc); H NMR (500 MHz,

CD3OD): δ 9.12 (dd, 1H, J = 0.9, 2.1 Hz, H-10), 8.74 (dd, 1H, J = 1.7, 4.9 Hz, H-6), 8.40

(ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-8), 7.57 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-7), 4.39 (t, 2H,

J = 6.6 Hz, H-1), 3.79 (m, 1H, H-4), 1.78-1.98 (m, 2H, H-2), 1.53-1.62 (m, 2H, H-3),

13 1.19 (d, 3H, J = 6.2 Hz, H-5); C NMR (75 MHz, CD3OD): δ 164.8, 152.6, 149.7, 137.4,

+ 126.7, 123.8, 66.7, 65.4, 35.0, 24.8, 22.2; HRMS (ESI): Calculated for [C11H15NO3 + H]

210.11247; Found 210.11220.

6-Hydroxyheptyl nicotinate (2.13b). Crude product

was purified by prep-TLC (100% EtOAc). Rf = 0.38

1 (100% EtOAc); H NMR (300 MHz, CD3OD): δ 9.11 (dd, 1H, J = 0.9, 2.1 Hz, H-12),

8.74 (dd, 1H, J = 1.7, 4.9 Hz, H-8), 8.40 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-10), 7.57 (ddd,

1H, J = 0.9, 4.9, 8.0 Hz, H-9), 4.37 (t, 2H, J = 6.6 Hz, H-1), 3.65-3.77 (m, 1H, H-6), 1.75-

1.89 (m, 2H, H-2), 1.56-1.38 (m, 6H, H-3,4,5), 1.15 (d, 3H, J = 6.2 Hz, H-7); 13C NMR from HSQC (125 MHz, CD3OD): δ 152.1, 149.3, 137.1, 123.6, 66.7, 65.0, 38.4, 27.9,

+ 26.9, 25.2, 22.7; HRMS (ESI): Calculated for [C13H19NO3 + H] 238.14377; Found

238.14333.

161

Chapter 7

7-Hydroxyoctyl nicotinate (2.14b). Crude product

was purified by prep-TLC (100% EtOAc). Rf = 0.42

1 (100% EtOAc); H NMR (300 MHz, CD3OD): δ 9.11 (dd, 1H, J = 0.9, 2.1 Hz, H-13),

8.74 (dd, 1H, J = 1.7, 4.9 Hz, H-9), 8.40 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-11), 7.57 (ddd,

1H, J = 0.9, 4.9, 8.0 Hz, H-10), 4.37 (t, 2H, J = 6.6 Hz, H-1), 3.65-3.76 (m, 1H, H-7),

1.74-1.87 (m, 2H, H-2), 1.56-1.32 (m, 8H, H-3,4,5,6), 1.13 (d, 3H, J = 6.2 Hz, H-8); 13C

NMR from HSQC (125 MHz, CD3OD): δ 152.1, 149.3, 137.1, 123.6, 66.7, 65.0, 38.4,

+ 27.9, 26.9, 25.2, 24.8, 22.7; HRMS (ESI): Calculated for [C14H21NO3 + H] 252.15942;

Found 252.15888.

2-(Oxiran-2-yl)ethyl nicotinate (2.18b). But-3-en-1-ol (1.44

g, 20 mmol) was dissolved in DCM (50 mL) and cooled to 4°C in an ice bath. To this was added meta-chloroperbenzoic acid (ca. 1 eq) in small portions.

The reaction was removed from the ice and stirred at room temperature overnight. A white precipitate formed and was filtered off. The mother liquor was collected and washed with saturated NaHCO3 (2 × 20 mL) and brine (1 × 20 mL). The aqueous layers were combined and back extracted with DCM (10 × 50 mL). The organic layer was dried over anhydrous sodium sulfate and evaporated in vacuo to afford 2-(oxiran-2-yl)ethanol as a light yellow oil with ~ 80% purity (1.1 g, 50%). Rf = 0.24 (50% hexanes-EtOAc).

Without further purification, 2-(oxiran-2-yl)ethanol (187 mg, 1.7 mmol) was combined with nicotinic acid (155 mg, 1.26 mmol), EDC-HCl (280 mg, 1.46 mmol) and DMAP (~5 mg, cat.) in DCM (10 mL) and stirred overnight at room temperature. The reaction was washed with saturated NaHCO3 (2 × 20 mL) and brine (1 × 20 mL) and dried over sodium sulfate anhydrous. The product was then purified by flash chromatography with a

162

Chapter 7

gradient from 100% hexanes to 0% hexanes-EtOAc to afford the product as a light yellow

1 oil (141 mg, 58%). Rf = 0.33 (50% hexanes-EtOAc); H NMR (500 MHz, CDCl3): δ 9.18

(dd, 1H, J = 0.9, 2.1 Hz, H-9), 8.74 (dd, 1H, J = 1.7, 4.9 Hz, H-5), 8.26 (ddd, 1H, J = 1.7,

2.1, 8.0 Hz, H-7), 7.36 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-6), 4.48 (m, 2H, H-1), 3.03-3.08

(m, 1H, H-3), 2.78 (m, 1H, H-4), 2.52 (m, 1H, H-4), 2.03-2.11 (m, 1H, H-2), 1.86-1.94

13 (m, 1H, H-2); C NMR (125 MHz, CD3OD): δ 165.1, 153.5, 150.8, 137.0, 125.9, 123.3,

+ 62.3, 49.5, 46.8, 31.9; LC-MS (ESI): Calculated for [C10H11NO3 + H] 194.1; Found.

194.0.

(3-Hydroxycyclopentyl)methyl nicotinate (2.20b). Crude

product was purified by prep-TLC (100% EtOAc) to give a

1 70:30 mixture of diastereomeriomers. Rf = 0.46 (100% EtOAc); H NMR major isomer

(500 MHz, CDCl3): δ 9.22 (dd, 1H, J = 0.9, 2.1 Hz, H-11), 8.78 (dd, 1H, J = 1.7, 4.9 Hz,

H-7), 8.29 (ddd, 1H, J = 1.7, 2.1, 8.0 Hz, H-9), 7.40 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-8),

4.43-4.47 (m, 1H, H-3), 4.28 (m, 2H, H-6), 2.65-2.75 (m, 1H, H-1), 1.40-2.20 (m, 6H, H-

13 2,4,5), assignments and integrations were also confirmed in CD3OD; C NMR from

HSQC for the major isomer (125 MHz, CDCl3): δ 154.8, 152.5, 141.2, 126.8, 76.5, 72.2,

+ 42.1 (2C), 39.3, 37.8; HRMS (ESI): Calculated for [C12H15NO3 + H] 222.11247; Found

222.11226.

3-Hydroxy-3-phenylpropyl nicotinate (2.22b). Crude

product was purified by prep-TLC (25% hexanes-EtOAc).

1 Rf = 0.32 (40% hexanes-EtOAc); H NMR (500 MHz, CDCl3): δ 9.18 (dd, 1H, J = 0.9,

2.1 Hz, H-12), 8.79 (dd, 1H, J = 1.7, 4.9 Hz, H-8), 8.26 (ddd, 1H, J = 0.9, 4.9, 8.0 Hz, H-

163

Chapter 7

10), 7.28-7.45 (m, 6H, H-5,6,7,9), 4.90 (dd, 1H, J = 5.2, 8.0 Hz, H-3), 4.57-4.64 (m, 1H,

H-1), 4.40-4.47(m, 1H, H-1), 2.16-2.31 (m, 2H, H-2); 13C NMR from HSQC (125 MHz,

CDCl3): δ 156.3, 153.6, 140.1, 131.5, 130.8, 128.6, 126.1, 74.4, 65.4, 65.3, 40.8; HRMS

+ (ESI): Calculated for [C15H15NO3 + H] 258.11247; Found 258.11189.

7.3.3 Synthesis and characterization of maleimides

Amines 5.1a and 5.2a were purchased from Invitrogen (La Jolla, CA).

Heterobifunctional crosslinkers 5.15b and 5.15c were purchased from Pierce Thermo

Fisher Scientific, Inc. (Rockford, IL).

N-(2-(5-(Dimethylamino)naphthalene-1-

sulfonamido)ethyl)-4-(2,5-dioxo-2,5-dihydro-1H-

pyrrol-1-yl)butanamide (5.1b). Heterobifunctional crosslinker 5.15b (9.1 mg, 0.032 mmol) was combined with amine 5.1a (10.0 mg, 0.034 mmol) in DMSO-d6 (1 mL) and stirred at room temperature for 2 h. The reaction mixture was diluted with water (~10 mL) and extracted with EtOAc (3 × 20 mL). The organic layers were combined, dried over anhydrous sodium sulfate and evaporated under reduced pressure. The product was purified by flash chromatography (gradient from 50%

EtOAc-hexanes to 100% EtOAc) to afford a yellow-green oil (11 mg, 75%). Rf = 0.3

1 (100% EtOAc); H NMR (500 MHz, CDCl3): δ 8.54 (d, 1H, J = 8.5 Hz), 8.29 (d, 1H, J =

8.6 Hz), 8.23 (dd, 1H, J = 7.3, 1.3 Hz), 7.53 (dd, 1H, J = 8.6, 7.6 Hz) 7.52 (dd, 1H, J =

8.5, 7.3 Hz), 7.18 (d, 1H, J = 7.6 Hz), 6.70 (s, 2H, H-6), 6.04 (t, 1H, J = 5.8, NH), 5.85 (t,

1H, J = 6.1, NH), 3.47 (t, 2H, J = 6.3 Hz, H-5), 3.27 (m, 2H), 3.06 (m, 2H), 2.89, (s, 6H,

H-13), 1.99-2.03 (m, 2H, H-3), 1.85 (m, 2H, H-4); 13C NMR (125 MHz, CDCl3): δ 172.5,

164

Chapter 7

171.1, 152.0, 134.7, 134.2, 130.5, 129.9, 129.6, 129.5, 128.4, 123.3, 118.7, 115.3, 45.4,

+ 43.1, 39.5, 36.9, 32.9, 24.3; HRMS (ESI): Calculated for [C22H26N4O5S + H] 459.16967;

Found 459.16895.

N-(2-(5-

(Dimethylamino)naphthalene

-1-sulfonamido)ethyl)-1-(3-

(2,5-dioxo-2,5-dihydro-1H-pyrrol-1-yl)propanamido)-3,6,9,12,15,18- hexaoxahenicosan-21-amide (5.1c). Heterobifunctional crosslinker 5.15c (0.0353 mmol,

141 µL from a 250 mM stock in DMSO-d6) was combined with amine 5.1a (10.3 mg,

0.0351 mmol) in DMSO-d6 (1 mL) and stirred at room temperature for 2 h. The reaction mixture was diluted with water (~10 mL) and extracted with EtOAc (4 × 20 mL). The organic layers were combined, dried over anhydrous sodium sulfate and evaporated under reduced pressure. The product was purified by prep-TLC (8% MeOH-DCM) to afford a

1 yellow-green oil (5.8 mg, 36%). Rf = 0.4 (8% MeOH-DCM); H NMR (500 MHz,

DMSO-d6): δ 8.44 (d, 1H, J = 8.5 Hz,), 8.24 (d, 1H, J = 8.6 Hz), 8.07 (dd, 1H, J = 7.3, 1.2

Hz), 8.00 (t, 1H, J = 5.4 Hz, NH), 7.96 (t, 1H, J = 6.0 Hz, NH), 7.77 (t, 1H, J = 5.6 Hz,

NH), 7.61 (dd, 1H, J = 8.5, 7.3 Hz), 7.58 (dd, 1H, J =8.6, 7.6 Hz), 7.24 (d, 1H, J = 7.6

Hz), 6.98 (s, 2H, H-9), 3.57 (t, 2H, J = 7.3 Hz, H-8) 3.39-3.51 (m, 22H), 3.34 (m, 2H),

3.12 (m, 2H), 3.01 (m, 2H), 2.81 (s, 6H, H-16), 2.76 (m, 2H), 2.30 (t, 2H, J = 7.3 Hz, H-

13 7), 2.16 (t, 2H, J = 6.5 Hz); C NMR (125 MHz, DMSO-d6): δ 171.2, 170.7, 169.9,

151.8, 136.2, 135.0, 129.9, 129.5, 129.4, 128.7, 128.3, 124.1, 119.5, 115.6, 70.1-70.2

(7C), 70.1, 70.0, 69.9, 69.4, 67.0, 45.5, 42.3, 39.0, 38.9, 36.4, 34.5, 34.4; HRMS (ESI):

+ Calculated for [C36H53N5O12S + H] : 780.34842; Found: 780.34764.

165

Chapter 7

2-(6-(Dimethylamino)-

3-(dimethyliminio)-3H-

xanthen-9-yl)-5-((1-(2,5-dioxo-

2,5-dihydro-1H-pyrrol-1-yl)-

3,25-dioxo-7,10,13,16,19,22-hexaoxa-4,26-diazahentriacontan-31- yl)carbamoyl)benzoate (5.2c). Heterobifunctional crosslinker 5.15c (0.0088 mmol, 35.2

µL from a 250 mM stock in DMSO-d6) was combined with amine 5.2a (4.5 mg, 0.0087 mmol) in DMSO-d6 (1 mL) and stirred at room temperature for 2 h. The reaction mixture was diluted with water (~10 mL) and extracted with DCM (4 × 20 mL). The organic layers were combined, dried over anhydrous sodium sulfate and evaporated under reduced pressure. The product was purified by prep-TLC (20% MeOH-DCM) to afford a

1 dark purple solid (3.1 mg, 36%). Rf = 0.5 (20% MeOH-DCM); H NMR (500 MHz,

CDCl3): δ 8.50 (br s, 1H), 8.25 (dd, 1H, J = 7.9, 1.2 Hz), 6.84 (br s, 1H), 6.68 (s, 2H),

6.60-6.66 (m, 3H), 6.51 (d, 2H, J = 2.5 Hz), 6.43 (dd, 2H, J = 8.8, 2.1 Hz), 3.82 (t, 2H, J =

7.2 Hz), 3.71 (t, 2H, J = 5.7 Hz), 3.57-3.66 (m, 20H), 3.54-3.49 (m, 4H), 3.39 (m, 2H),

3.27-3.32 (m, 2H), 3.02 (s, 12H), 2.48-2.53 (m, 4H), 1.69-1.76 (m, 2H), 1.58 (p, 2H, J =

13 7.0 Hz), 1.43-1.51 (m, 2H); C NMR (125 MHz, CDCl3): δ 172.1, 170.5, 169.9, 166.2,

136.5, 134.2, 134.1, 129.0, 125.2, 123.9, 109.3, 98.3, 70.6-70.4 (6C), 70.3, 70.2, 70.1,

69.7, 67.3, 40.3, 40.1, 39.2, 38.7, 36.9, 34.5, 34.4, 29.7, 29.2, 28.4, 23.8 (signals for 7

+ quaternary carbons are missing); HRMS (ESI): Calculated for [C52H68N6O14 + H] :

1001.48663; Found: 1001.48613.

166

Chapter 7

N-(3-(Dimethylamino)propyl)-4-(2,5-

dioxo-2,5-dihydro-1H-pyrrol-1-yl)butanamide

(5.3b). Heterobifunctional crosslinker 5.15b (26 mg, 0.093 mmol) was combined with amine 5.3a (10 mg, 0.098 mmol) in DMSO-d6 (1 mL) and stirred at room temperature for 2 h. The crude reaction was concentrated under reduced pressure and purified by flash chromatography (2% conc. NH4OH/14%

MeOH/84% DCM) affording a pink oil (6.6 mg, 27%). Rf = 0.3 (2% conc. NH4OH/14%

1 MeOH/84% DCM); H NMR (500 MHz, CDCl3): δ 6.91 (br s, 1H, NH), 6.69 (s, 2H, H-

5), 3.56 (t, 2H, J = 6.5 Hz, H-4), 3.31 (m, 2H, H-6), 2.37 (t, 2H, J = 6.5 Hz, H-2), 2.22 (s,

6H, H-9), 2.13 (t, 2H, J = 7.4 Hz, H-8), 1.92 (m, 2H, H-7), 1.65 (p, 2H, J = 6.5 Hz, H-3);

13 C NMR (125 MHz, CDCl3): δ 171.5, 170.9, 134.1, 58.4, 45.4, 39.1, 37.3, 33.8, 26.2,

+ 24.7; HRMS (ESI): Calculated for [C13H21N3O3 + H] : 268.16557; Found: 268.16481.

N-(3-

(Dimethylamino)propyl)-1-(3-(2,5-

dioxo-2,5-dihydro-1H-pyrrol-1- yl)propanamido)-3,6,9,12,15,18-hexaoxahenicosan-21-amide (5.3c).

Heterobifunctional crosslinker 15c (0.0166 mmol, 66.4 µL from 250 mM stock in

DMSO-d6) was combined with amine 3a (1.67 mg, 0.0164 mmol) in DMSO-d6 (1 mL) and stirred at room temperature for 2 h. Attempts to isolate the product by prep-TLC or

HPLC were unsuccessful due to decomposition. Purity was >70%. 1H NMR (500 MHz,

DMSO-d6): δ 8.00 (t, 1H, J = 5.5 Hz, NH), 7.78 (t, 1H, J = 5.4 Hz, NH), 6.98 (s, 2H, H-

7), 3.54-3.59 (m, 2H), 3.44-3.52 (m, 22H), 3.34 (t, 2H, J = 5.9 Hz), 3.13 (m, 2H), 3.02

167

Chapter 7

(m, 2H), 2.31 (t, J = 7.3 Hz, 2H), 2.26 (t, 2H, 6.5 Hz), 2.17 (t, 2H, J = 7.2 Hz.), 2.09 (s,

13 6H, H-11), 1.48 (p, 2H, 7.2 Hz, H-9); C NMR (125 MHz, DMSO-d6): δ 171.2, 170.3,

169.9, 135.0, 70.2 (5C), 70.15, 70.12, 70.00, 69.96, 69.4, 67.3, 57.1, 45.6, 38.9, 37.2,

+ 36.6, 34.5, 34.4, 27.6, 25.9; HRMS (ESI): Calculated for [C27H48N4O10 + H] : 589.34432;

Found: 589.34373.

168

References

(1) Omura, T.; Sato, R., A new cytochrome in liver microsomes. J. Biol. Chem. 1962, 237 (4), PC1375-PC1376.

(2) Ortiz de Montellano, P. R., Cytochrome P450: structure, mechanism, and biochemistry. 3rd ed.; Kluwer Academic/Plenum Publishers, New York: 2005; p 689.

(3) Bernhardt, R., Cytochromes P450 as versatile biocatalysts. J. Biotechnol. 2006, 124 (1), 128-145.

(4) Munro, A. W.; Girvan, H. M.; McLean, K. J., Variations on a (t)heme-novel mechanisms, redox partners and catalytic functions in the cytochrome P450 superfamily. Nat. Prod. Rep. 2007, 24 (3), 585-609.

(5) Rittle, J.; Green, M. T., Cytochrome P450 compound I: capture, characterization, and C-H bond activation kinetics. Science 2010, 330 (6006), 933-937.

(6) Imai, M.; Shimada, H.; Watanabe, Y.; Matsushima-Hibiya, Y.; Makino, R.; Koga, H.; Horiuchi, T.; Ishimura, Y., Uncoupling of the cytochrome P-450cam monooxygenase reaction by a single mutation, threonine-252 to alanine or valine: a possible role of the hydroxy amino acid in oxygen activation. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 7823-7827.

(7) Loida, P. J.; Sligar, S. G., Engineering cytochrome P-450cam to increase the stereospecificity and coupling of aliphatic hydroxylation. Protein Eng. 1993, 6 (2), 207-212.

(8) Meharenna, Y. T.; Poulos, T. L., Using molecular dynamics to probe the structural basis for enhanced stability in thermal stable cytochromes P450. Biochemistry 2010, 49 (31), 6680-6686.

(9) Denisov, I. G.; Shih, A. Y.; Sligar, S. G., Structural differences between soluble and membrane bound cytochrome P450s. J. Inorg. Biochem. 2012, 108 (0), 150- 158.

(10) Guengerich, F. P.; Martin, M. V.; Guo, Z.; Chun, Y.-J., Purification of functional recombinant P450s from bacteria. In Methods Enzymol., Eric, F. J.; Michael, R. W., Eds. Academic Press: 1996; Vol. Volume 272, pp 35-44.

(11) Redlich, G.; Zanger, U. M.; Riedmaier, S.; Bache, N.; Giessing, A. B. M.; Eisenacher, M.; Stephan, C.; Meyer, H. E.; Jensen, O. N.; Marcus, K., Distinction between human cytochrome P450 (CYP) isoforms and identification of new phosphorylation sites by mass spectrometry. J. Proteome Res. 2008, 7 (11), 4678- 4688.

169

References

(12) Gillam, E. M. J., Exploring the potential of xenobiotic-metabolising enzymes as biocatalysts: evolving designer catalysts from polyfunctional cytochrome P450 enzymes. Clin. Exp. Pharmacol. P. 2005, 32 (3), 147-152.

(13) Guengerich, F. P., Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chem. Res. Toxicol. 2001, 14 (6), 611-650.

(14) Gillam, E. M. J., Extending the capabilities of nature’s most versatile catalysts: directed evolution of mammalian xenobiotic-metabolizing P450s. Arch. Biochem. Biophys. 2007, 464 (2), 176-186.

(15) Kumar, S., Engineering cytochrome P450 biocatalysts for biotechnology, medicine and bioremediation. Expert Opin. Drug Metab. Toxicol. 2010, 6 (2), 115-131.

(16) Sletten, E. M.; Bertozzi, C. R., Bioorthogonal chemistry: fishing for selectivity in a sea of functionality. Angew. Chem. Int. Ed. 2009, 48 (38), 6974-6998.

(17) Lim, R. K. V.; Lin, Q., Bioorthogonal chemistry: recent progress and future directions. Chem. Commun. 2010, 46 (10), 1589-1600.

(18) Stephanopoulos, N.; Francis, M. B., Choosing an effective protein bioconjugation strategy. Nat. Chem. Biol. 2011, 7 (12), 876-884.

(19) Wong, L. S.; Khan, F.; Micklefield, J., Selective covalent protein immobilization: strategies and applications. Chem. Rev. 2009, 109 (9), 4025-4053.

(20) Camarero, J. A., Recent developments in the site-specific immobilization of proteins onto solid supports. Pept. Sci. 2008, 90 (3), 450-458.

(21) Kalia, J.; Raines, R. T., Advances in bioconjugation. Curr. Org. Chem. 2010, 14, 138-147.

(22) Scheck, R. A.; Dedeo, M. T.; Iavarone, A. T.; Francis, M. B., Optimization of a biomimetic transamination reaction. J. Am. Chem. Soc. 2008, 130 (35), 11762- 11770.

(23) Chan, A. O.-Y.; Ho, C.-M.; Chong, H.-C.; Leung, Y.-C.; Huang, J.-S.; Wong, M.- K.; Che, C.-M., Modification of N-terminal α-amino groups of peptides and proteins using ketenes. J. Am. Chem. Soc. 2012, 134 (5), 2589-2598.

(24) Agard, N. J.; Prescher, J. A.; Bertozzi, C. R., A strain-promoted [3 + 2] azide−alkyne cycloaddition for covalent modification of biomolecules in living systems. J. Am. Chem. Soc. 2004, 126 (46), 15046-15047.

(25) Besanceney-Webler, C.; Jiang, H.; Zheng, T.; Feng, L.; Soriano del Amo, D.; Wang, W.; Klivansky, L. M.; Marlow, F. L.; Liu, Y.; Wu, P., Increasing the

170

References

efficacy of bioorthogonal click reactions for bioconjugation: a comparative study. Angew. Chem. Int. Ed. 2011, 50 (35), 8051-8056.

(26) de Almeida, G.; Sletten, E. M.; Nakamura, H.; Palaniappan, K. K.; Bertozzi, C. R., Thiacycloalkynes for copper-free click chemistry. Angew.Chem. Int. Ed. 2012, 51 (10), 2443-2447.

(27) Lallana, E.; Riguera, R.; Fernandez-Megia, E., Reliable and efficient procedures for the conjugation of biomolecules through Huisgen azide–alkyne cycloadditions. Angew. Chem. Int. Ed. 2011, 50 (38), 8794-8804.

(28) Lin, P.-C.; Ueng, S.-H.; Tseng, M.-C.; Ko, J.-L.; Huang, K.-T.; Yu, S.-C.; Adak, A. K.; Chen, Y.-J.; Lin, C.-C., Site-specific protein modification through CuI- catalyzed 1,2,3-triazole formation and its implementation in protein microarray fabrication. Angew. Chem. 2006, 118 (26), 4392-4396.

(29) van Berkel, S. S.; van Eldijk, M. B.; van Hest, J. C. M., Staudinger ligation as a method for bioconjugation. Angew. Chem. Int. Ed.2011, 50 (38), 8806-8827.

(30) Nilsson, B. L.; Kiessling, L. L.; Raines, R. T., Staudinger ligation: a peptide from a thioester and azide. Org. Lett. 2000, 2 (13), 1939-1941.

(31) Saxon, E.; Armstrong, J. I.; Bertozzi, C. R., A “traceless” staudinger ligation for the chemoselective synthesis of amide bonds. Org. Lett. 2000, 2 (14), 2141-2143.

(32) Saxon, E.; Bertozzi, C. R., Cell surface engineering by a modified staudinger reaction. Science 2000, 287 (5460), 2007-2010.

(33) de Araújo, A. D.; Palomo, J. M.; Cramer, J.; Köhn, M.; Schröder, H.; Wacker, R.; Niemeyer, C.; Alexandrov, K.; Waldmann, H., Diels–Alder ligation and surface immobilization of proteins. Angew. Chem. 2006, 118 (2), 302-307.

(34) Blackman, M. L.; Royzen, M.; Fox, J. M., Tetrazine ligation: fast bioconjugation based on inverse-electron-demand Diels−Alder reactivity. J. Am. Chem. Soc. 2008, 130 (41), 13518-13519.

(35) Song, W.; Wang, Y.; Qu, J.; Madden, M. M.; Lin, Q., A photoinducible 1,3- dipolar cycloaddition reaction for rapid, selective modification of tetrazole- containing proteins. Angew. Chem. Int. Ed. 2008, 47 (15), 2832-2835.

(36) Liu, D. S.; Tangpeerachaikul, A.; Selvaraj, R.; Taylor, M. T.; Fox, J. M.; Ting, A. Y., Diels–Alder cycloaddition for fluorophore targeting to specific proteins inside living cells. J. Am. Chem. Soc. 2011, 134 (2), 792-795.

(37) Rideout, D., Self-assembling cytotoxins. Science 1986, 233 (4763), 561-563.

(38) Dirksen, A.; Hackeng, T. M.; Dawson, P. E., Nucleophilic catalysis of oxime ligation. Angew. Chem. 2006, 118 (45), 7743-7746.

171

References

(39) Dirksen, A.; Dirksen, S.; Hackeng, T. M.; Dawson, P. E., Nucleophilic catalysis of hydrazone formation and transimination: implications for dynamic covalent chemistry. J. Am. Chem. Soc. 2006, 128 (49), 15602-15603.

(40) Dirksen, A.; Dawson, P. E., Rapid Oxime and Hydrazone Ligations with Aromatic Aldehydes for Biomolecular Labeling. Bioconj. Chem. 2008, 19 (12), 2543-2548.

(41) Nauman, D. A.; Bertozzi, C. R., Kinetic parameters for small-molecule drug delivery by covalent cell surface targeting. Biochim. Biophys. Acta, Gen. Subj. 2001, 1568 (2), 147-154.

(42) Blanden, A. R.; Mukherjee, K.; Dilek, O.; Loew, M.; Bane, S. L., 4- Aminophenylalanine as a biocompatible nucleophilic catalyst for hydrazone ligations at low temperature and neutral pH. Bioconj. Chem. 2011, 22 (10), 1954- 1961.

(43) Lin, Y. A.; Chalker, J. M.; Floyd, N.; Bernardes, G. a. J. L.; Davis, B. G., Allyl sulfides are privileged substrates in aqueous cross-metathesis: application to site- selective protein modification. J. Am. Chem. Soc. 2008, 130 (30), 9642-9643.

(44) Lin, Y. A.; Chalker, J. M.; Davis, B. G., Olefin metathesis for site-selective protein modification. ChemBioChem 2009, 10 (6), 959-969.

(45) Kodama, K.; Fukuzawa, S.; Nakayama, H.; Sakamoto, K.; Kigawa, T.; Yabuki, T.; Matsuda, N.; Shirouzu, M.; Takio, K.; Yokoyama, S.; Tachibana, K., Site- specific functionalization of proteins by organopalladium reactions. ChemBioChem 2007, 8 (2), 232-238.

(46) Chalker, J. M.; Wood, C. S. C.; Davis, B. G., A convenient catalyst for aqueous and protein Suzuki−Miyaura cross-coupling. J. Am. Chem. Soc. 2009, 131 (45), 16346-16347.

(47) Davis, B. G.; Chalker, J. M.; Lin, Y. A., Olefin metathesis for site-selective protein modification. ChemBioChem 2009, 10, 959-969.

(48) Li, N.; Lim, R. K. V.; Edwardraja, S.; Lin, Q., Copper-free Sonogashira cross- coupling for functionalization of alkyne-encoded proteins in aqueous medium and in bacterial cells. J. Am. Chem. Soc. 2011, 133 (39), 15316-15319.

(49) Sletten, E. M.; Bertozzi, C. R., A bioorthogonal quadricyclane ligation. J. Am. Chem. Soc. 2011, 133 (44), 17570-17573.

(50) Behrens, C. R.; Hooker, J. M.; Obermeyer, A. C.; Romanini, D. W.; Katz, E. M.; Francis, M. B., Rapid chemoselective bioconjugation through oxidative coupling of anilines and aminophenols. J. Am. Chem. Soc. 2011, 133 (41), 16398-16401.

(51) Liu, C. C.; Schultz, P. G., Adding new chemistries to the genetic code. Annu. Rev. Biochem. 2010, 79 (1), 413-444.

172

References

(52) Xu, M.-Q.; Evans Jr, T. C., Intein-mediated ligation and cyclization of expressed proteins. Methods 2001, 24 (3), 257-277.

(53) Tan, L.-P.; Lue, R. Y. P.; Chen, G. Y. J.; Yao, S. Q., Improving the intein- mediated, site-specific protein biotinylation strategies both in vitro and in vivo. Bioorg. Med. Chem. Lett. 2004, 14 (24), 6067-6070.

(54) Kalia, J.; Raines , R. T., Reactivity of intein thioesters: appending a functional group to a protein. ChemBioChem 2006, 7 (9), 1375-1383.

(55) Chen, I.; Howarth, M.; Lin, W.; Ting, A. Y., Site-specific labeling of cell surface proteins with biophysical probes using biotin ligase. Nat. Meth. 2005, 2 (2), 99- 104.

(56) Seitchik, J. L.; Peeler, J. C.; Taylor, M. T.; Blackman, M. L.; Rhoads, T. W.; Cooley, R. B.; Refakis, C.; Fox, J. M.; Mehl, R. A., Genetically encoded tetrazine amino acid directs rapid site-specific in vivo bioorthogonal ligation with trans- cyclooctenes. J. Am. Chem. Soc. 2012, 134 (6), 2898-2901.

(57) Sadeghi, S. J.; Fantuzzi, A.; Gilardi, G., Breakthrough in P450 bioelectrochemistry and future perspectives. Biochim. Biophys. Acta, Proteins Proteomics 2011, 1814 (1), 237-248.

(58) Bistolas, N.; Wollenberger, U.; Jung, C.; Scheller, F. W., Cytochrome P450 biosensors—a review. Biosens. Bioelectron. 2005, 20 (12), 2408-2423.

(59) Fantuzzi, A.; Fairhead, M.; Gilardi, G., Direct electrochemistry of immobilized human cytochrome P450 2E1. J. Am. Chem. Soc. 2004, 126 (16), 5040-5041.

(60) Yang, M.; Kabulski, J. L.; Wollenberg, L.; Chen, X.; Subramanian, M.; Tracy, T. S.; Lederman, D.; Gannett, P. M.; Wu, N., Electrocatalytic drug metabolism by CYP2C9 bonded to a self-assembled monolayer-modified electrode. Drug Metab. Dispos. 2009, 37 (4), 892-899.

(61) Alonso-Lomillo, M. A.; Yardimci, C.; Domínguez-Renedo, O.; Arcos-Martínez, M. J., CYP450 2B4 covalently attached to carbon and gold screen printed electrodes by diazonium salt and thiols monolayers. Anal. Chim. Acta 2009, 633 (1), 51-56.

(62) Asturias-Arribas, L.; Alonso-Lomillo, M. A.; Domínguez-Renedo, O.; Arcos- Martínez, M. J., CYP450 biosensors based on screen-printed carbon electrodes for the determination of cocaine. Anal. Chim. Acta 2011, 685 (1), 15-20.

(63) Ndangili, P. M.; Jijana, A. M.; Baker, P. G. L.; Iwuoha, E. I., 3- Mercaptopropionic acid capped ZnSe quantum dot-cytochrome P450 3A4 enzyme biotransducer for 17β-estradiol. J. Electroanal. Chem. 2011, 653 (1–2), 67-74.

173

References

(64) Dodhia, V. R.; Sassone, C.; Fantuzzi, A.; Nardo, G. D.; Sadeghi, S. J.; Gilardi, G., Modulating the coupling efficiency of human cytochrome P450 CYP3A4 at electrode surfaces through protein engineering. Electrochem. Commun. 2008, 10 (11), 1744-1747.

(65) Shumyantseva, V. V.; Bulko, T. V.; Bachmann, T. T.; Bilitewski, U.; Schmid, R. D.; Archakov, A. I., Electrochemical reduction of flavocytochromes 2B4 and 1A2 and their catalytic activity. Arch. Biochem. Biophys. 2000, 377 (1), 43-48.

(66) Shumyantseva, V. V.; Uvarov, V. Y.; Byakova, O. E.; Archakov, A. I., Semisynthetic flavocytochromes based on cytochrome P450 2B4: reductase and oxygenase activities. Arch. Biochem. Biophys. 1998, 354 (1), 133-138.

(67) Uvarov, V. Y.; Shumyantseva, V. V.; Bykhovskaya, E. A.; Kolyada, L. N.; Archakov, A. L., Semi artificial hydroxylating enzymes created by flavins binding to cytochrome P450 2B4 and by bleomycin binding to NADPH-cytochrome P450 reductase. Biochem. Biophys. Res. Commun. 1994, 200 (2), 722-725.

(68) Nicoli, R.; Bartolini, M.; Rudaz, S.; Andrisano, V.; Veuthey, J. L., Development of immobilized enzyme reactors based on human recombinant cytochrome P450 enzymes for phase I drug metabolism studies. J. Chromatogr. A 2008, 1206 (1), 2- 10.

(69) Tanvir, S.; Pantigny, J.; Boulnois, P.; Pulvin, S., Covalent immobilization of recombinant human cytochrome CYP2E1 and glucose-6-phosphate dehydrogenase in alumina membrane for drug screening applications. J. Membr. Sci. 2009, 329 (1–2), 85-90.

(70) Wright, A. T.; Cravatt, B. F., Chemical proteomic probes for profiling cytochrome P450 activities and drug interactions in vivo. Chem. Biol. 2007, 14 (9), 1043- 1051.

(71) Wright, A. T.; Song, J. D.; Cravatt, B. F., A suite of activity-based probes for human cytochrome P450 enzymes. J. Am. Chem. Soc. 2009, 131 (30), 10692- 10700.

(72) Wen, B.; Doneanu, C. E.; Gartner, C. A.; Roberts, A. G.; Atkins, W. M.; Nelson, S. D., Fluorescent photoaffinity labeling of cytochrome P450 3A4 by lapachenole: identification of modification sites by mass spectrometry. Biochemistry 2005, 44 (6), 1833-1845.

(73) Collom, S. L.; Jamakhandi, A. P.; Tackett, A. J.; Radominska-Pandya, A.; Miller, G. P., CYP2E1 active site residues in substrate recognition sequence 5 identified by photoaffinity labeling and homology modeling. Arch. Biochem. Biophys. 2007, 459 (1), 59-69.

(74) Tsalkova, T. N.; Davydova, N. Y.; Halpert, J. R.; Davydov, D. R., Mechanism of interactions of α-naphthoflavone with cytochrome P450 3A4 explored with an

174

References

engineered enzyme bearing a fluorescent probe. Biochemistry 2007, 46 (1), 106- 119.

(75) Davydov, D. R.; Sineva, E. V.; Sistla, S.; Davydova, N. Y.; Frank, D. J.; Sligar, S. G.; Halpert, J. R., Electron transfer in the complex of membrane-bound human cytochrome P450 3A4 with the flavin domain of P450BM-3: the effect of oligomerization of the heme protein and intermittent modulation of the spin equilibrium. Biochim. Biophys. Acta 2010, 1797 (3), 378.

(76) Bommarius, A. S.; Riebel, B. R., Biocatalysis. In Biocatalysis, WILEY-VCH Verlag GmbH & Co. KGaA: Weinheim, 2004.

(77) Ahmed, M.; Kelly, T.; Ghanem, A., Applications of enzymatic and non-enzymatic methods to access enantiomerically pure compounds using kinetic resolution and racemisation. An overview. Tetrahedron 2012, 68 (34), 6781-6802.

(78) Illanes, A.; Cauerhff, A.; Wilson, L.; Castro, G. R., Recent trends in biocatalysis engineering. Bioresour. Technol. 2012, 115 (0), 48-57.

(79) Noble, M. A.; Miles, C. S.; Chapman, S. K.; Lysek, D. A.; MacKay, A. C.; Reid, G. A.; Hanzlik, R. P.; Munro, A. W., Roles of key active-site residues in flavocytochrome P450 BM3. Biochem. J. 1999, 339, 371-379.

(80) Falck, J. R.; Reddy, Y. K.; Haines, D. C.; Reddy, K. M.; Krishna, U. M.; Graham, S.; Murry, B.; Peterson, J. A., Practical, enantiospecific syntheses of 14,15-EET and leukotoxin B (vernolic acid). Tet. Lett. 2001, 42 (25), 4131-4133.

(81) Julsing, M. K.; Cornelissen, S.; Bühler, B.; Schmid, A., Heme-iron oxygenases: powerful industrial biocatalysts? Curr. Opin. Chem. Biol. 2008, 12 (2), 177-186.

(82) Petzoldt, K.; Annen, K.; Leurent, H.; Wiechert, R., Process for the preparation of 11-beta-hydroxy steroids. US Patent 1982, 4,353,985, 12 October 1982.

(83) Sonomoto, K.; Hog, M. M.; Tanaka, A.; Fukui, S., 11β-hydroxylation of cortexolone (Reichstein Compound S) to hyrocortisone by Curvularia lunata entrapped in photo-cross-linked resin gels. Appl. Environ. Microbiol. 1983, 45 (2), 436-443.

(84) Matsuoka, T.; Miyakoshi, S.; Tanzawa, K.; Nakahara, K.; Hosobuchi, M.; Serizawa, N., Purification and characterization of cytochrome P-450sca from Streptomyces carbophilus. Eur. J. Biochem. 1989, 184 (3), 707-713.

(85) Serizawa, N., Biochemical and molecular approaches for production of pravastatin, a potent cholesterol-lowering drug. In Biotechnol. Annu. Rev., El- Gewely, M. R., Ed. Elsevier: 1996; Vol. Volume 2, pp 373-389.

175

References

(86) Gbewonyo, K.; Buckland, B. C.; Lilly, M. D., Development of a large-scale continuous substrate feed process for the biotransformation of simvastatin by Nocardia s.p. Biotechnol. Bioeng. 1991, 37 (11), 1101-1107.

(87) Liu, S.; Li, C.; Fang, X.; Cao, Z. a., Optimal pH control strategy for high-level production of long-chain α,ω-dicarboxylic acid by Candida tropicalis. Enzyme Microb. Technol. 2004, 34 (1), 73-77.

(88) Ro, D.-K.; Paradise, E. M.; Ouellet, M.; Fisher, K. J.; Newman, K. L.; Ndungu, J. M.; Ho, K. A.; Eachus, R. A.; Ham, T. S.; Kirby, J.; Chang, M. C. Y.; Withers, S. T.; Shiba, Y.; Sarpong, R.; Keasling, J. D., Production of the antimalarial drug precursor artemisinic acid in engineered yeast. Nature 2006, 440 (7086), 940-943.

(89) Barnes, H. J., Maximizing expression of eukaryotic cytochrome P450s in Escherichia coli. Methods Enzymol. 1996, 272, 3-14.

(90) Barnes, H. J.; Arlotto, M. P.; Waterman, M. R., Expression and enzymatic activity of recombinant cytochrome P450 17 alpha-hydroxylase in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 1991, 88 (13), 5597-5601.

(91) Larson, J. R.; Coon, M. J.; Porter, T. D., Purification and properties of a shortened form of cytochrome P-450 2E1: deletion of the NH2-terminal membrane-insertion signal peptide does not alter the catalytic activities. Proc. Natl. Acad. Sci. U.S.A. 1991, 88 (20), 9141-9145.

(92) Gillam, E. M. J.; Baba, T.; Kim, B. R.; Ohmori, S.; Guengerich, F. P., Expression of modified human cytochrome P450 3A4 in Escherichia coli and purification and reconstitution of the enzyme. Arch. Biochem. Biophys. 1993, 305 (1), 123-131.

(93) Gillam, E. M. J.; Guo, Z. Y.; Martin, M. V.; Jenkins, C. M.; Guengerich, F. P., Expression of cytochrome P450 2D6 in Escherichia coli, purification, and spectral and catalytic characterization. Arch. Biochem. Biophys. 1995, 319 (2), 540-550.

(94) Gillam, E. M. J.; Guo, Z. Y.; Guengerich, F. P., Expression of modified human cytochrome P450 2E1 in Escherichia coli, purification, and spectral and catalytic properties. Arch. Biochem. Biophys. 1994, 312 (1), 59-66.

(95) Sandhu, P.; Guo, Z. Y.; Baba, T.; Martin, M. V.; Tukey, R. H.; Guengerich, F. P., Expression of modified human cytochrome P450 1A2 in Escherichia coli: stabilization, purification, spectral characterization, and catalytic activities of the enzyme. Arch. Biochem. Biophys. 1994, 309 (1), 168-177.

(96) Hanna, I. H.; Reed, J. R.; Guengerich, F. P.; Hollenberg, P. F., Expression of human cytochrome P450 2B6 in Escherichia coli: characterization of catalytic activity and expression levels in human liver. Arch. Biochem. Biophys. 2000, 376 (1), 206-216.

176

References

(97) Gillam, E. M. J.; Guo, Z. Y.; Ueng, Y. F.; Yamazaki, H.; Cock, I.; Reilly, P. E. B.; Hooper, W. D.; Guengerich, F. P., Expression of cytochrome-P450-3A5 in Escherichia Coli: effects of 5′ modification, purification, spectral characterization, reconstitution conditions, and catalytic activities. Arch. Biochem. Biophys. 1995, 317 (2), 374-384.

(98) Sandhu, P.; Baba, T.; Guengerich, F. P., Expression of modified cytochrome P450 2C10 (2C9) in Escherichia coli, purification, and reconstitution of catalytic activity. Arch. Biochem. Biophys. 1993, 306 (2), 443-450.

(99) Richardson, T. H.; Jung, F.; Griffin, K. J.; Wester, M.; Raucy, J. L.; Kemper, B.; Bornheim, L. M.; Hassett, C.; Omiecinski, C. J.; Johnson, E. F., A Universal approach to the expression of human and rabbit cytochrome P450s of the 2C subfamily in Escherichia coli. Arch. Biochem. Biophys. 1995, 323 (1), 87-96.

(100) Mitsuda, M.; Iwasaki, M., Improvement in the expression of CYP2B6 by co- expression with molecular chaperones GroES/EL in Escherichia coli. Protein Expression and Purif. 2006, 46 (2), 401-405.

(101) Pritchard, M. P.; Ossetian, R.; Li, D. N.; Henderson, C. J.; Burchell, B.; Wolf, C. R.; Friedberg, T., A general strategy for the expression of recombinant human cytochrome P450s in Escherichia coli using bacterial signal peptides: expression of CYP3A4, CYP2A6, and CYP2E1. Arch. Biochem. Biophys. 1997, 345 (2), 342- 354.

(102) Jenkins, C. M.; Waterman, M. R., Flavodoxin and NADPH-flavodoxin reductase from Escherichia coli support bovine cytochrome P450c17 hydroxylase activities. J. Biol. Chem. 1994, 269 (44), 27401-27408.

(103) Yamazaki, H.; Ueng, Y.-F.; Shimada, T.; Guengerich, F. P., Roles of divalent metal ions in oxidations catalyzed by recombinant cytochrome P450 3A4 and replacement of NADPH-cytochrome P450 reductase with other flavoproteins, ferredoxin, and oxygen surrogates. Biochemistry 1995, 34 (26), 8380-8389.

(104) Blake, J. A. R.; Pritchard, M.; Ding, S.; Smith, G. C. M.; Burchell, B.; Wolf, C. R.; Friedberg, T., Coexpression of a human P450 (CYP3A4) and P450 reductase generates a highly functional monooxygenase system in Escherichia coli. FEBS Letters 1996, 397 (2–3), 210-214.

(105) Parikh, A.; Gillam, E. M. J.; Guengerich, F. P., Drug metabolism by Escherichia coli expressing human cytochromes P450. Nat. Biotech. 1997, 15 (8), 784-788.

(106) Iwata, H.; Fujita, K.-i.; Kushida, H.; Suzuki, A.; Konno, Y.; Nakamura, K.; Fujino, A.; Kamataki, T., High catalytic activity of human cytochrome P450 co- expressed with human NADPH-cytochrome P450 reductase in Escherichia coli. Biochem. Pharmacol. 1998, 55 (8), 1315-1325.

177

References

(107) Rushmore, T. H.; Reider, P. J.; Slaughter, D.; Assang, C.; Shou, M., Bioreactor systems in drug metabolism: synthesis of cytochrome P450-generated metabolites. Metab. Eng. 2000, 2 (2), 115-125.

(108) Vail, R. B.; Homann, M. J.; Hanna, I.; Zaks, A., Preparative synthesis of drug metabolites using human cytochrome P450s 3A4, 2C9 and 1A2 with NADPH- P450 reductase expressed in Escherichia coli. J. Ind. Microbiol. Biotechnol. 2005, 32 (2), 67-74.

(109) Hlavica, P., Assembly of non-natural electron transfer conduits in the cytochrome P450 system: A critical assessment and update of artificial redox constructs amenable to exploitation in biotechnological areas. Biotechnol. Adv. 2009, 27 (2), 103-121.

(110) Munro, A. W.; Girvan, H. M.; McLean, K. J., Cytochrome P450–redox partner fusion enzymes. Biochim. Biophys. Acta, Gen. Subj. 2007, 1770 (3), 345-359.

(111) Murakami, H.; Yabusaki, Y.; Sakaki, T., A genetically engineered P450 monooxygenase: Construction of the functional fused enzyme between rat cytochrome P450c and NADPH-cytochrome P450 reductase. DNA 1987, 6 (3), 189-197.

(112) Hara, M.; Iazvovskaia, S.; Ohkawa, H.; Asada, Y.; Miyake, J., Immobilization of P450 monooxygenase and chloroplast for use in light-driven bioreactors. J. Biosci. Bioeng. 1999, 87 (6), 793-797.

(113) Shet, M. S.; Fisher, C. W.; Holmans, P. L.; Estabrook, R. W., Human cytochrome P450 3A4: enzymatic properties of a purified recombinant fusion protein containing NADPH-P450 reductase. Proc. Natl. Acad. Sci. U.S.A. 1993, 90 (24), 11748-11752.

(114) Hayashi, K.; Sakaki, T.; Kominami, S.; Inouye, K.; Yabusaki, Y., Coexpression of genetically engineered fused enzyme between yeast NADPH–P450 reductase and human cytochrome P450 3A4 and human cytochrome b5 in yeast. Arch. Biochem. Biophys. 2000, 381 (1), 164-170.

(115) Dodhia, V.; Fantuzzi, A.; Gilardi, G., Engineering human cytochrome P450 enzymes into catalytically self-sufficient chimeras using molecular Lego. J. Biol. Inorg. Chem. 2006, 11 (7), 903-916.

(116) Inui, H.; Maeda, A.; Ohkawa, H., Molecular characterization of specifically active recombinant fused enzymes consisting of CYP3A4, NADPH-cytochrome P450 oxidoreductase, and cytochrome b5. Biochemistry 2007, 46 (35), 10213-10221.

(117) Deeni, Y. Y.; Paine, M. J. I.; Ayrton, A. D.; Clarke, S. E.; Chenery, R.; Wolf, C. R., Expression, purification, and biochemical characterization of a human cytochrome P450 CYP2D6-NADPH cytochrome P450 reductase fusion protein. Arch. Biochem. Biophys. 2001, 396 (1), 16-24.

178

References

(118) Fairhead, M.; Giannini, S.; Gillam, E.; Gilardi, G., Functional characterisation of an engineered multidomain human P450 2E1 by molecular Lego. J. Biol. Inorg. Chem. 2005, 10 (8), 842-853.

(119) Chefson, A.; Zhao, J.; Auclair, K., Replacement of natural cofactors by selected hydrogen peroxide donors or organic peroxides results in improved activity for CYP3A4 and CYP2D6. ChemBioChem 2006, 7 (6), 916-919.

(120) Kumar, S.; Liu, H.; Halpert, J. R., Engineering of cytochrome P450 3A4 for enhanced peroxide-mediated substrate oxidation using directed evolution and site- directed mutagenesis. Drug Metab. Dispos. 2006, 34 (12), 1958-1965.

(121) Keizers, P. H. J.; Schraven, L. H. M.; de Graaf, C.; Hidestrand, M.; Ingelman- Sundberg, M.; van Dijk, B. R.; Vermeulen, N. P. E.; Commandeur, J. N. M., Role of the conserved threonine 309 in mechanism of oxidation by cytochrome P450 2D6. Biochem. Biophys. Res. Commun. 2005, 338 (2), 1065-1074.

(122) Chefson, A.; Zhao, J.; Auclair, K., Sugar-mediated lyoprotection of purified human CYP3A4 and CYP2D6. J. Biotechnol. 2007, 130 (4), 436-440.

(123) Zhao, J.; Auclair, K., The activity of human CYP2D6 in low water organic solvents. Biotechnol. Bioeng. 2009, 102 (4), 1268-1272.

(124) Chefson, A.; Auclair, K., CYP3A4 activity in the presence of organic cosolvents, ionic liquids, or water-immiscible organic solvents. ChemBioChem 2007, 8 (10), 1189-1197.

(125) Doukyu, N.; Ogino, H., Organic solvent-tolerant enzymes. Biochem. Eng. J. 2010, 48 (3), 270-282.

(126) Zhao, J.; Tan, E.; Ferras, J.; Auclair, K., Activity of human P450 2D6 in biphasic solvent systems. Biotechnol. Bioeng. 2007, 98 (2), 508-513.

(127) Kumar, S.; Sun, L.; Liu, H.; Muralidhara, B. K.; Halpert, J. R., Engineering mammalian cytochrome P450 2B1 by directed evolution for enhanced catalytic tolerance to temperature and dimethyl sulfoxide. Protein Eng. Des. Sel. 2006, 19 (12), 547-554.

(128) Eijsink, V. G. H.; Bjørk, A.; Gåseidnes, S.; Sirevåg, R.; Synstad, B.; Burg, B. v. d.; Vriend, G., Rational engineering of enzyme stability. J. Biotechnol. 2004, 113 (1–3), 105-120.

(129) Eijsink, V. G. H.; Gåseidnes, S.; Borchert, T. V.; van den Burg, B., Directed evolution of enzyme stability. Biomol. Eng. 2005, 22 (1–3), 21-30.

(130) Kumar, S.; Chen, C. S.; Waxman, D. J.; Halpert, J. R., Directed evolution of mammalian cytochrome P450 2B1. J. Biol. Chem. 2005, 280 (20), 19569-19575.

179

References

(131) Talakad, J. C.; Wilderman, P. R.; Davydov, D. R.; Kumar, S.; Halpert, J. R., Rational engineering of cytochromes P450 2B6 and 2B11 for enhanced stability: insights into structural importance of residue 334. Arch. Biochem. Biophys. 2010, 494 (2), 151-158.

(132) Larsen, A. T.; May, E. M.; Auclair, K., Predictable stereoselective and chemoselective hydroxylations and epoxidations with P450 3A4. J. Am. Chem. Soc. 2011, 133 (20), 7853-7858.

(133) Schaus, S. E.; Brandes, B. D.; Larrow, J. F.; Tokunaga, M.; Hansen, K. B.; Gould, A. E.; Furrow, M. E.; Jacobsen, E. N., Highly selective hydrolytic kinetic resolution of terminal epoxides catalyzed by chiral (salen)CoIII complexes. practical synthesis of enantioenriched terminal epoxides and 1,2-diols. J. Am. Chem. Soc. 2002, 124 (7), 1307-1315.

(134) Mercer, E., Sterol biosynthesis inhibitors: their current status and modes of action. Lipids 1991, 26 (8), 584-597.

(135) Vanden Bossche, H.; Lauwers, W.; Willemsens, G.; Marichal, P.; Cornelissen, F.; Cool, W., Molecular basis for the antimycotic and antibacterial activity of N- substituted imidazoles and triazoles: the inhibition of isoprenoid biosynthesis. Pestic. Sci. 1984, 15, 188-198.

(136) Vasaitis, T. S.; Bruno, R. D.; Njar, V. C. O., CYP17 inhibitors for prostate cancer therapy. J. Steroid Biochem. Mol. Biol. 2011, 125 (1–2), 23-31.

(137) Peng, C.-C.; Pearson, J. T.; Rock, D. A.; Joswig-Jones, C. A.; Jones, J. P., The effects of type II binding on metabolic stability and binding affinity in cytochrome P450 CYP3A4. Arch. Biochem. Biophys. 2010, 497 (1–2), 68-81.

(138) Pearson, J. T.; Hill, J. J.; Swank, J.; Isoherranen, N.; Kunze, K. L.; Atkins, W. M., Surface plasmon resonance analysis of antifungal azoles binding to CYP3A4 with kinetic resolution of multiple binding orientations. Biochemistry 2006, 45 (20), 6341-6353.

(139) Kunze, K. L.; Nelson, W. L.; Kharasch, E. D.; Thummel, K. E.; Isoherranen, N., Stereochemical aspects of itraconazole metabolism in vitro and in vivo. Drug Metab. Dispos. 2006, 34 (4), 583-590.

(140) Minton, A. P., How can biochemical reactions within cells differ from those in test tubes? J. Cell Sci. 2006, 119 (14), 2863-2869.

(141) Zhou, H.-X.; Rivas, G.; Minton, A. P., Macromolecular crowding and confinement: biochemical, biophysical and potential physiological consequences. Annu. Rev. Biophys. 2008, 37, 375-397.

(142) O'Reilly, E.; Kohler, V.; Flitsch, S. L.; Turner, N. J., Cytochromes P450 as useful biocatalysts: addressing the limitations. Chem. Commun. 2011, 47 (9), 2490-2501.

180

References

(143) Ménard, A.; Huang, Y.; Karam, P.; Cosa, G.; Auclair, K., Site-specific fluorescent labeling and oriented immobilization of a triple mutant of CYP3A4 via C64. Bioconj. Chem. 2012, 23 (4), 826-836.

(144) Segall, M., Physical properties of P450. 1997, (Accessed July 21, 2012).

(145) Peng, C.-C.; Cape, J. L.; Rushmore, T.; Crouch, G. J.; Jones, J. P., Cytochrome P450 2C9 type II binding studies on quinoline-4-carboxamide analogues. J. Med. Chem. 2008, 51 (24), 8000-8011.

(146) Peng, C.-C.; Rushmore, T.; Crouch, G. J.; Jones, J. P., Modeling and synthesis of novel tight-binding inhibitors of cytochrome P450 2C9. Bioorg. Med. Chem. 2008, 16 (7), 4064-4074.

(147) Jones, J. P.; Joswig-Jones, C. A.; Hebner, M.; Chu, Y.; Koop, D. R., The effects of nitrogen-heme-iron coordination on substrate affinities for cytochrome P450 2E1. Chem. Biol. Interact. 2011, 193 (1), 50-56.

(148) Chiba, M.; Jin, L.; Neway, W.; Vacca, J. P.; Tata, J. R.; Chapman, K.; Lin, J. H., P450 interaction with HIV protease inhibitors: relationship between metabolic stability, inhibitory potency, and P450 binding spectra. Drug Metab. Dispos. 2001, 29 (1), 1-3.

(149) Zhang, W.; Ramamoorthy, Y.; Kilicarslan, T.; Nolte, H.; Tyndale, R. F.; Sellers, E. M., Inhibition of cytochromes P450 by antifungal imidazole derivatives. Drug Metab. Dispos. 2002, 30 (3), 314-318.

(150) Gubbins, P. O., Triazole antifungal agents drug-drug interactions involving hepatic cytochrome P450. Expert Opin. Drug Metab. Toxicol. 2011, 7 (11), 1411- 1429.

(151) Dvorak, Z., Drug-drug interactions by azole antifungals: beyond a dogma of CYP3A4 enzyme activity inhibition. Toxicol. Lett. 2011, 202, 129-132.

(152) Gerber, M. C.; Tejwani, G. A.; Gerber, N.; Bianchine, J. R., Drug interactions with cimetidine: an update. Pharmacol. Ther. 1985, 27 (3), 353-370.

(153) Sun, X.; Qiu, J.; Strong, S. A.; Green, L. S.; Wasley, J. W. F.; Blonder, J. P.; Colagiovanni, D. B.; Mutka, S. C.; Stout, A. M.; Richards, J. P.; Rosenthal, G. J., Discovery of potent and novel S-nitrosoglutathione reductase inhibitors devoid of cytochrome P450 activities. Bioorg. Med. Chem. Lett. 2011, 21 (19), 5849-5853.

(154) Dahal, U. P.; Joswig-Jones, C.; Jones, J. P., Comparative study of the affinity and metabolism of type I and type II binding quinoline carboxamide analogues by cytochrome P450 3A4. J. Med. Chem. 2012, 55 (1), 280-290.

181

References

(155) Conner, K. P.; Vennam, P.; Krzyaniak, M. D.; Woods, C. M.; Bowman, M. K.; Atkins, W. M., 1,2,3-triazole-heme interactions in cytochrome P450: functionally competent triazole-water-heme complexes. Biochemistry 2012.

(156) Krishnan, S.; Wasalathanthri, D.; Zhao, L.; Schenkman, J. B.; Rusling, J. F., Efficient bioelectronic actuation of the natural catalytic pathway of human metabolic cytochrome P450s. J. Am. Chem. Soc. 2011, 133 (5), 1459-1465.

(157) Ahlström, M. M.; Zamora, I., Characterization of type II ligands in CYP2C9 and CYP3A4. J. Med. Chem. 2008, 51 (6), 1755-1763.

(158) Chiba, M.; Tang, C.; Neway, W. E.; Williams, T. M.; Jane Desolms, S.; Dinsmore, C. J.; Wai, J. S.; Lin, J. H., P450 interaction with farnesyl-protein transferase inhibitors metabolic stability, inhibitory potency, and P450 binding spectra in human liver microsomes. Biochem. Pharmacol. 2001, 62 (6), 773-776.

(159) Porubsky, P. R.; Meneely, K. M.; Scott, E. E., Structures of human cytochrome P- 450 2E1. J. Biol. Chem. 2008, 283 (48), 33698-33707.

(160) Adas, F.; Salaün, J. P.; Berthou, F.; Picart, D.; Simon, B.; Amet, Y., Requirement for ω and (ω–1)-hydroxylations of fatty acids by human cytochromes P450 2E1 and 4A11. J. Lipid Res. 1999, 40 (11), 1990-1997.

(161) Therrien, E.; Englebienne, P.; Arrowsmith, A. G.; Mendoza-Sanchez, R.; Corbeil, C. R.; Weill, N.; Campagna-Slater, V.; Moitessier, N., Integrating medicinal chemistry, organic/combinatorial chemistry, and computational chemistry for the discovery of selective estrogen receptor modulators with Forecaster, a novel platform for drug discovery. J. Chem. Inf. Model. 2011, 52 (1), 210-224.

(162) Porubsky, P. R.; Battaile, K. P.; Scott, E. E., Human cytochrome P450 2E1 structures with fatty acid analogs reveal a previously unobserved binding mode. J. Biol. Chem. 2010, 285 (29), 22282-22290.

(163) Tarcsay, Á.; Kiss, R.; Keserű, G., Site of metabolism prediction on cytochrome P450 2C9: a knowledge-based docking approach. J. Comput. Aid. Mol. Des. 2010, 24 (5), 399-408.

(164) Prasad, J. C.; Goldstone, J. V.; Camacho, C. J.; Vajda, S.; Stegeman, J. J., Ensemble modeling of substrate binding to cytochromes P450: analysis of catalytic differences between CYP1A orthologs. Biochemistry 2007, 46 (10), 2640-2654.

(165) Ellis, R. J.; Minton, A. P., Cell biology: join the crowd. Nature 2003, 425 (6953), 27-28.

(166) Ellis, R. J., Macromolecular crowding: obvious but underappreciated. Trends Biochem. Sci. 2001, 26 (10), 597-604.

182

References

(167) Zimmerman, S. B.; Trach, S. O., Estimation of macromolecule concentrations and excluded volume effects for the cytoplasm of Escherichia coli. J. Mol. Biol. 1991, 222 (3), 599-620.

(168) Miklos, A. C.; Li, C.; Sharaf, N. G.; Pielak, G. J., Volume exclusion and soft interaction effects on protein stability under crowded conditions. Biochemistry 2010, 49 (33), 6984-6991.

(169) Wang, Y.; Li, C.; Pielak, G. J., Effects of proteins on protein diffusion. J. Am. Chem. Soc. 2010, 132 (27), 9392-9397.

(170) Derham, B. K.; Harding, J. J., The effect of the presence of globular proteins and elongated polymers on enzyme activity. Biochim. Biophys. Acta, Proteins and Proteomics 2006, 1764 (6), 1000-1006.

(171) Homchaudhuri, L.; Sarma, N.; Swaminathan, R., Effect of crowding by dextrans and Ficolls on the rate of alkaline phosphatase–catalyzed hydrolysis: a size- dependent investigation. Biopolymers 2006, 83 (5), 477-486.

(172) Pozdnyakova, I.; Wittung-Stafshede, P., Non-linear effects of macromolecular crowding on enzymatic activity of multi-copper oxidase. Biochim. Biophys. Acta, Proteins and Proteomics 2010, 1804 (4), 740-744.

(173) Morán-Zorzano, M. T.; Viale, A. M.; Muñoz, F. J.; Alonso-Casajús, N.; Eydallín, G. G.; Zugasti, B.; Baroja-Fernández, E.; Pozueta-Romero, J., Escherichia coli AspP activity is enhanced by macromolecular crowding and by both glucose-1,6- bisphosphate and nucleotide-sugars. FEBS Letters 2007, 581 (5), 1035-1040.

(174) Jiang, M.; Guo, Z., Effects of macromolecular crowding on the intrinsic catalytic efficiency and structure of enterobactin-specific isochorismate synthase. J. Am. Chem. Soc. 2007, 129 (4), 730-731.

(175) Olsen, S. N., Applications of isothermal titration calorimetry to measure enzyme kinetics and activity in complex solutions. Thermochim. Acta 2006, 448 (1), 12- 18.

(176) Laursen, T.; Jensen, K.; Møller, B. L., Conformational changes of the NADPH- dependent cytochrome P450 reductase in the course of electron transfer to cytochromes P450. Biochim. Biophys. Acta, Proteins and Proteomics 2011, 1814 (1), 132-138.

(177) Ellis, J.; Gutierrez, A.; Barsukov, I. L.; Huang, W.-C.; Grossmann, J. G.; Roberts, G. C. K., Domain motion in cytochrome P450 reductase: conformational equilibria revealed by NMR and small-angle X-ray scattering. J. Biol. Chem. 2009, 284 (52), 36628-36637.

183

References

(178) Pudney, C. R.; Khara, B.; Johannissen, L. O.; Scrutton, N. S., Coupled motions direct electrons along human microsomal P450 chains. PLoS Biology 2011, 9 (12), 1-10.

(179) Gutierrez, A.; Paine, M.; Wolf, C. R.; Scrutton, N. S.; Roberts, G. C. K., Relaxation kinetics of cytochrome P450 reductase: internal electron transfer is limited by conformational change and regulated by coenzyme binding. Biochemistry 2002, 41 (14), 4626-4637.

(180) Farooq, Y.; Roberts, G. C., Kinetics of electron transfer between NADPH- cytochrome P450 reductase and cytochrome P450 3A4. Biochem. J. 2010, 432 (3), 485-93.

(181) Davydov, D. R.; Halpert, J. R., Allosteric P450 mechanisms: multiple binding sites, multiple conformers or both? Expert Opinion on Drug Metabolism & Toxicology 2008, 4 (12), 1523-1535.

(182) Ekroos, M.; Sjögren, T., Structural basis for ligand promiscuity in cytochrome P450 3A4. Proc. Natl. Acad. Sci. U.S.A. 2006, 103 (37), 13682-13687.

(183) Sun, H.; Bessire, A. J.; Vaz, A., Dirlotapide as a model substrate to refine structure-based drug design strategies on CYP3A4-catalyzed metabolism. Bioorg. Med. Chem. Lett. 2012, 22 (1), 371-376.

(184) Denisov, I. G.; Baas, B. J.; Grinkova, Y. V.; Sligar, S. G., Cooperativity in cytochrome P450 3A4: linkages in substrate binding, spin state, uncoupling, and product formation. J. Biol. Chem. 2007, 282 (10), 7066-76.

(185) Subramanian, M.; Low, M.; Locuson, C. W.; Tracy, T. S., CYP2D6-CYP2C9 Protein-protein interactions and isoform-selective effects on substrate binding and catalysis. Drug Metab. Dispos. 2009, 37 (8), 1682-1689.

(186) Elcock, A. H., Models of macromolecular crowding effects and the need for quantitative comparisons with experiment. Curr. Opin. Struct. Biol. 2010, 20 (2), 196-206.

(187) Mateo, C.; Palomo, J. M.; Fernandez-Lorente, G.; Guisan, J. M.; Fernandez- Lafuente, R., Improvement of enzyme activity, stability and selectivity via immobilization techniques. Enzyme and Microb. Technol. 2007, 40 (6), 1451- 1463.

(188) Polizzi, K. M.; Bommarius, A. S.; Broering, J. M.; Chaparro-Riggers, J. F., Stability of biocatalysts. Curr. Opin. Chem. Biol. 2007, 11 (2), 220-225.

(189) Hernandez, K.; Fernandez-Lafuente, R., Control of protein immobilization: coupling immobilization and site-directed mutagenesis to improve biocatalyst or biosensor performance. Enzyme Microb. Technol. 2011, 48 (2), 107-22.

184

References

(190) Mie, Y.; Suzuki, M.; Komatsu, Y., Electrochemically driven drug metabolism by membranes containing human cytochrome P450. J. Am. Chem. Soc. 2009, 131 (19), 6646-6647.

(191) Nath, A.; Koo, P. K.; Rhoades, E.; Atkins, W. M., Allosteric effects on substrate dissociation from cytochrome P450 3A4 in nanodiscs observed by ensemble and single-molecule fluorescence spectroscopy. J. Am. Chem. Soc. 2008, 130 (47), 15746-15747.

(192) Das, A.; Zhao, J.; Schatz, G. C.; Sligar, S. G.; Van Duyne, R. P., Screening of type I and II drug binding to human cytochrome P450-3A4 in nanodiscs by localized surface plasmon resonance spectroscopy. Anal. Chem. 2009, 81 (10), 3754-3759.

(193) Shumyantseva, V. V.; Bulko, T. V.; Kuznetsova, G. P.; Samenkova, N. F.; Archakov, A. L., Electrochemistry of cytochromes P450: analysis of current- voltage characteristics of electrodes with immobilized cytochromes P450 for the screening of substrates and inhibitors. Biochemistry (Mosc) 2009, 74 (4), 438-444.

(194) Shumyantseva, V. V.; Shikh, E. V.; Makhova, A. A.; Bylko, T. V.; Kukes, V. G.; Sizova, O. S.; Ramenskaia, G. V.; Ursanov, S. A.; Archakov, A. L., The influence of vitamin B group on monooxygenase activity of cytochrome P450 3A4: pharmacokinetics and electro analysis of catalytic properties. Biomed. Khim. 2011, 57 (3), 343-354.

(195) Makhova, A.; Shumyantseva, V.; Shich, E.; Bulko, T.; Kukes, V.; Sizova, O.; Ramenskaya, G.; Usanov, S.; Archakov, A., Electroanalysis of cytochrome P450 3A4 catalytic properties with nanostructured electrodes: the influence of vitamin B group on diclofenac metabolism. BioNanoScience 2011, 1 (1), 46-52.

(196) Ignaszak, A.; Hendricks, N.; Waryo, T.; Songa, E.; Jahed, N.; Ngece, R.; Al- Ahmed, A.; Kgarebe, B.; Baker, P.; Iwuoha, E. I., Novel therapeutic biosensor for indinavir—A protease inhibitor antiretroviral drug. J. Pharm. Biomed. Anal. 2009, 49 (2), 498-501.

(197) Joseph, S.; Rusling, J. F.; Lvov, Y. M.; Friedberg, T.; Fuhr, U., An amperometric biosensor with human CYP3A4 as a novel drug screening tool. Biochem. Pharmacol. 2003, 65 (11), 1817-1826.

(198) Hendricks, N. R.; Waryo, T. T.; Arotiba, O.; Jahed, N.; Baker, P. G. L.; Iwuoha, E. I., Microsomal cytochrome P450-3A4 (CYP3A4) nanobiosensor for the determination of 2,4-dichlorophenol—an endocrine disruptor compound. Electrochim. Acta 2009, 54 (7), 1925-1931.

(199) Chang, G.; Morigaki, K.; Tatsu, Y.; Hikawa, T.; Goto, T.; Imaishi, H., Vertically integrated human P450 and oxygen sensing film for the assays of P450 metabolic activities. Anal. Chem. 2011, 83 (8), 2956-2963.

185

References

(200) Ley, C.; Holtmann, D.; Mangold, K.-M.; Schrader, J., Immobilization of histidine- tagged proteins on electrodes. Colloids Surf., B 2011, 88 (2), 539-551.

(201) Mašek, J.; Bartheldyová, E.; Korvasová, Z.; Škrabalová, M.; Koudelka, Š.; Kulich, P.; Kratochvílová, I.; Miller, A. D.; Ledvina, M.; Raška, M.; Turánek, J., Immobilization of histidine-tagged proteins on monodisperse metallochelation liposomes: Preparation and study of their structure. Anal. Biochem. 2011, 408 (1), 95-104.

(202) Lim, Y. T.; Lee, K. Y.; Lee, K.; Chung, B. H., Immobilization of histidine-tagged proteins by magnetic nanoparticles encapsulated with nitrilotriacetic acid (NTA)- phospholipids micelle. Biochem.Biophys. Res. Commun. 2006, 344 (3), 926-930.

(203) Sinchaikul, S.; Pan, F.-M.; Cheng, C.-W.; Wong, C.-H.; Fang, J.-M.; Tseng, M.- J.; Chen, S.-T., Protein microarray using α-amino acids as metal tags on chips. Bioorg. Med. Chem. Lett. 2005, 15 (5), 1413-1416.

(204) Chang, Y.-J.; Chang, C.-H., Protein microarray chip with Ni–Co alloy coated surface. Biosens. Bioelectron. 2010, 25 (7), 1748-1754.

(205) Zhu, H.; Bilgin, M.; Bangham, R.; Hall, D.; Casamayor, A.; Bertone, P.; Lan, N.; Jansen, R.; Bidlingmaier, S.; Houfek, T.; Mitchell, T.; Miller, P.; Dean, R. A.; Gerstein, M.; Snyder, M., Global analysis of protein activities using proteome chips. Science 2001, 293 (5537), 2101-2105.

(206) Domanski, T. L.; Liu, J.; Harlow, G. R.; Halpert, J. R., Analysis of four residues within substrate recognition site 4 of human cytochrome P450 3A4: role in steroid hydroxylase activity and α-naphthoflavone stimulation. Arch. Biochem. Biophys. 1998, 350 (2), 223-232.

(207) Wen, B.; Lampe, J. N.; Roberts, A. G.; Atkins, W. M.; David Rodrigues, A.; Nelson, S. D., Cysteine 98 in CYP3A4 contributes to conformational integrity required for P450 interaction with CYP reductase. Arch. Biochem. Biophys. 2006, 454 (1), 42-54.

(208) Zimmermann, J. L.; Nicolaus, T.; Neuert, G.; Blank, K., Thiol-based, site-specific and covalent immobilization of biomolecules for single-molecule experiments. Nat. Protoc. 2010, 5 (6), 975-985.

(209) English, B. P.; Min, W.; van Oijen, A. M.; Lee, K. T.; Luo, G.; Sun, H.; Cherayil, B. J.; Kou, S. C.; Xie, X. S., Ever-fluctuating single enzyme molecules: Michaelis-Menten equation revisited. Nat. Chem. Biol. 2006, 2 (2), 87-94.

(210) Neuman, K. C.; Nagy, A., Single-molecule force spectroscopy: optical tweezers, magnetic tweezers and atomic force microscopy. Nat. Meth. 2008, 5 (6), 491-505.

(211) Lee, W. A.; Bruice, T. C., Homolytic and heterolytic oxygen-oxygen bond scissions accompanying oxygen transfer to iron(III) porphyrins by percarboxylic

186

References

acids and hydroperoxides. A mechanistic criterion for peroxidase and cytochrome P-450. J. Am. Chem. Soc. 1985, 107 (2), 513-514.

(212) Barr, D. P.; Martin, M. V.; Guengerich, F. P.; Mason, R. P., Reaction of cytochrome P450 with cumene hydroperoxide: ESR spin-trapping evidence for the homolytic scission of the peroxide O−O bond by ferric cytochrome P450 1A2. Chem. Res. Toxicol. 1996, 9 (1), 318-325.

(213) Zhang, R.; Chandrasena, R. E. P.; Martinez II, E.; Horner, J. H.; Newcomb, M., Formation of compound I by photo-oxidation of compound II. Org. Lett. 2005, (7) (6), 1193-1195.

(214) White, M. C., Adding aliphatic C–H bond oxidations to synthesis. Science 2012, 335 (6070), 807-809.

(215) Chen, M. S.; White, M. C., A predictably selective aliphatic C–H oxidation reaction for complex molecule synthesis. Science 2007, 318 (5851), 783-787.

(216) Chen, M. S.; White, M. C., Combined effects on selectivity in Fe-catalyzed methylene oxidation. Science 2010, 327 (5965), 566-571.

(217) Bigi, M. A.; Reed, S. A.; White, M. C., Directed metal (oxo) aliphatic C–H hydroxylations: overriding substrate bias. J. Am. Chem. Soc. 2012, 134 (23), 9721-9726.

(218) Jones, B. C.; Middleton, D. S.; Youdim, K., Cytochrome P450 metabolism and inhibition: analysis for drug discovery. In Progress in Medicinal Chemistry, Lawton, G.; Witty, D. R., Eds. Elsevier: 2009; Vol. Volume 47, pp 239-263.

(219) Corbeil, C. R.; Englebienne, P.; Moitessier, N., Docking ligands into flexible and solvated macromolecules. 1. Development and validation of FITTED 1.0. J. Chem. Inf. Model. 2007, 47 (2), 435-449.

(220) Brustad, E. M.; Lemke, E. A.; Schultz, P. G.; Deniz, A. A., A General and efficient method for the site-specific dual-labeling of proteins for single molecule fluorescence resonance energy transfer. J. Am. Chem. Soc. 2008, 130 (52), 17664- 17665.

(221) Milles, S.; Tyagi, S.; Banterle, N.; Koehler, C.; VanDelinder, V.; Plass, T.; Neal, A. P.; Lemke, E. A., Click strategies for single-molecule protein fluorescence. J. Am. Chem. Soc. 2012, 134 (11), 5187-5195.

(222) Liu, H.; Naismith, J. H., An efficient one-step site-directed delection, insertion, single and multiple-site plasmid mutagenesis protocol. BMC Biotech. 2008, 8 (91).

187

References

(223) Oprian, D. D.; Coon, M. J., Oxidation-reduction states of FMN and FAD in NADPH-cytochrome P-450 reductase during reduction by NADPH. J. Biol. Chem. 1982, 257 (15), 8935-44.

(224) Grassetti, D. R.; Murray Jr, J. F., Determination of sulfhydryl groups with 2,2′- or 4,4′-dithiodipyridine. Arch. Biochem. Biophys. 1967, 119 (0), 41-49.

(225) Scheer, J. M.; Ryan, C. A., A method for the quantitative recovery of proteins from polyacrylamide gels. Anal. Biochem. 2001, 298 (1), 130-132.

(226) Karam, P.; Ngo, A. T.; Rouiller, I.; Cosa, G., Unraveling electronic energy transfer in single conjugated polyelectrolytes encapsulated in lipid vesicles. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 17480-17485.

(227) Ngo, A. T.; Karam, P.; Fuller, E.; Burger, M.; Cosa, G., Liposome encapsulation of conjugated polyelectrolytes: toward a liposome beacon. J. Am. Chem. Soc. 2007, 130 (2), 457-459.

(228) Ha, T., Single-molecule fluorescence resonance energy transfer. Methods 2001, 25 (1), 78-86.

(229) Harada, Y.; Sakurada, K.; Aoki, T.; Thomas, D. D.; Yanagida, T., Mechanochemical coupling in actomyosin energy transduction studied by in vitro movement assay. J. Mol. Biol. 1990, 216 (1), 49-68.

(230) Córdoba, O. L.; Linskens, S. B.; Dacci, E.; Santomé, J. A., ‘In gel’ cleavage with cyanogen bromide for protein internal sequencing. J. Biochem. Bioph. Methods 1997, 35 (1), 1-10.

(231) Schagger, H., Tricine-SDS-PAGE. Nat. Protoc. 2006, 1 (1), 16-22.

(232) Moss, J. A., Guide for resin and linker selection in solid-phase peptide synthesis. In Curr. Protoc. Protein Sci., John Wiley & Sons, Inc.: 2001.

(233) SDBSWeb: http://riodb01.ibase.aist.go.jp/sdbs/ (National Institute of Advanced Industrial Science and Technology, July 6 2012). [searched hexyl benzoate]

(234) Pearson, J.; Dahal, U. P.; Rock, D,; Peng, C.-C.; Schenk, J. O.; Joswig-Jones, C.; Jones, J. P., The kinetic mechanism of cytochrome P450 metabolism of Type II binding compounds: Evidence supporting direct reduction. Arch. Biochem. Biophys. 2011, 511, 69-79.

188

Appendix I

Appendix I – NMR spectra

189

Appendix I

190

Appendix I

191

Appendix I

192

Appendix I

193

Appendix I

194

Appendix I

195

Appendix I

196

Appendix I

197

Appendix I

198

Appendix I

199

Appendix I

200

Appendix I

201

Appendix I

202

Appendix I

203

Appendix I

204

Appendix I

205

Appendix I

206

Appendix I

207

Appendix I

208

Appendix I

209

Appendix I

210

Appendix I

211

Appendix I

212

Appendix I

213

Appendix I

214

Appendix I

215

Appendix I

216

Appendix I

217

Appendix I

218

Appendix I

219

Appendix I

220

Appendix I

221

Appendix I

222

Appendix I

223

Appendix I

224

Appendix I

225

Appendix I

226

Appendix I

227

Appendix I

228

Appendix II

Appendix II – Type II ligands as substrates of CYP2E1

In this appendix, I will propose answers to two important questions the reader may have after reading Chapter 2 of this thesis and propose experiments that would be helpful in confirming these hypotheses.

Question 1 – How does type II binding character contribute to improved oxidation yields?

In order to answer this question, one must look to the enzyme kinetic parameters kcat and Km and the ratio of these two (kcat/Km) which is a measure of enzyme catalytic efficiency. Thus, higher kcat values (higher turnover numbers) and lower Km values

(higher substrate affinities) translate into higher catalytic efficiencies and improved reaction yields. While type II binding character likely lowers turnover rates (lower kcat), this is more than compensated for by its ability to improve binding affinity (lower Km) resulting in a net increase in catalytic efficiency. This phenomenon was indeed generally observed by Dahal et al.154 and Peason et al.234 with a series of pyridine quinoline-4- carboxamide analogues metabolized by CYP3A4 (see Figure 2.5). Further kinetic experiments to determine the kcat and Km parameters of auxiliary-substrates with CYP2E1 would be helpful in supporting this hypothesis and perhaps also in explaining why lower

Ks values (also a measure of binding affinity) did not strictly correlate with higher yields

(see Tables 2.1-2.3). Note that this discrepancy is not all that surprising since Ks and Km are determined under very different conditions. The latter is measured while the enzyme is actively turning over whereas the former is determined by observing spectral changes upon substrate binding to the resting enzyme (and therefore does not distinguish between productive and non-productive binding modes). Note also that interactions other than type

II coordination to the heme-iron are likely involved in determining binding affinity.

229

Appendix II

Question 2 – How can some type II ligands of P450 enzymes also be substrates?

The underlying question here is twofold. First, does a type II bound substrate have to unbind (exit the active site) and rebind in a type I manner in order to be oxidized by the enzyme. If not, does reorientation within the active site occur before or after the cofactor

CHP binds to the heme-iron? Since our type II binding auxiliary-substrates generally gave slightly higher oxidation yields than similar type I binders, the reorientation from the unproductive (type II) to the productive (type I) binding mode is likely not achieved by unbinding and rebinding of the substrate137. Because of the higher binding affinity of type

II ligands, this would always result in a lower type I to type II binding ratio and lead to lower oxidation yields instead. One could confirm this hypothesis by measuring auxiliary- substrate koff and kon rates from CYP2E1 by SRP and comparing them to the corresponding catalytic rates (kcat). Lower unbinding rates would suggest that once bound, the auxiliary-substrate very rarely unbinds and is committed to catalysis. This was indeed found to be the case with the pyridine quinoline-4-carboxamide analogues studies by

Pearson et al.234 (see Figure 2.5).

Further experiments would also be helpful in gaining insight into the timing of auxiliary-substrate reorientation within the P450 active site with respect to CHP binding.

According to our findings that reaction rates were slower with CPR and NADPH as the cofactors than with CHP (data not shown), I would venture to say that reorientation occurs only after CHP binds to the heme. In other words, CHP actively displaces the type

II-bound auxiliary-substrate, allowing it to reorient to the productive mode. Results by

Pearson et al. indicate that reduction rates of type II-ligated heme-iron is up to 450-fold

230

Appendix II

faster than the rate of N-Fe3+ bond breaking234. This suggests that heme reduction occurs mostly while the substrate is bound in a type II manner and explains the higher transformation rates observed with type II binders. Assuming that our auxiliary-substrates exhibit similar N-Fe3+ bond breaking rates, a mechanism requiring breaking of this bond before CHP heme-iron activation would be inconsistent with our findings. Of course, one would have to determine the N-Fe3+ bond breaking rates of our auxiliary-substrates with

CYP2E1 as well as the CHP activation rate of nitrogen-coordinated heme in order to confirm this hypothesis.

The mechanism proposed above would require CHP and the auxiliary-substrates to occupy the P450 active site concurrently and in close proximity to the heme-iron. This should not be a problem due to the highly hydrophobic and flexible nature of CYP2E1’s active site159, 162. In fact, the situation described above would not be all that different than that expected for other CHP-supported P450-catalysed reactions.

231