Harnessing the Power of P450 Enzymes Aaron T. Larsen

Home , Lisofylline

Harnessing the Power of P450 Enzymes A Chemical Auxiliary-Based Approach to Predictable P450 Oxidations at Inactivated C-H Bonds

Aaron T. Larsen

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 2K6 Submitted: November, 2011

Abstract

Enantioselective hydroxylation of one specific methylene in the presence of many similar groups is debatably the most challenging chemical transformation. Although chemists have recently made progress towards the hydroxylation of inactivated C-H bonds, enzymes like P450s

(CYPs) remain unsurpassed in specificity and scope. The substrate promiscuity of many P450s is desirable for synthetic applications; however, the inability to predict the products of these enzymatic reactions and the poor activity and stability of these enzymes is impeding advancement.

In chapter 2 of this thesis, we evaluate several strategies to improve the activity and stability of CYP3A4. These strategies include the immobilization of CYP3A4 inside molecular hydrogels and silica in addition to the chemical modification of CYP3A4 using various anhydrides. Although none of the strategies we investigate here greatly enhance the catalytic utility of the enzyme, CYP3A4 is shown to be highly tolerant to functionalization at a large number of surface residues and to the presence of extremely high concentrations of silica during catalysis.

Recognizing the potential for enzymes containing small, hydrophobic active sites to catalyze Diels-Alder reactions, chapter 3 describes the design and application of several assays to evaluate the Diels-Alderase activity of CYP2E1. Although the presence of CYP2E1 is not found to increase the rates of the reactions we investigate here, the results do demonstrate that there is

Abstract an interaction between one or more of the substrates and the enzyme at either the active site or at another binding pocket.

In chapter 4, we evaluate 4 auxiliaries for their ability to direct CYP3A4 oxidations.

When linked to substrates, several of these auxiliaries are shown to direct CYP3A4 oxidations at specific C-H bonds. Although the auxiliaries we explore here are found to be limited in utility, several important lessons are learned which we apply to the design of a next generation auxiliary to be discussed in the following chapter.

In chapter 5, we demonstrate the utility of theobromine as a chemical auxiliary to control the selectivity of CYP3A4 reactions. When linked to substrates, inexpensive, achiral theobromine directs the reaction to produce hydroxylation or epoxidation at the fourth carbon from the auxiliary with pro-R facial selectivity. This strategy provides a versatile yet controllable system for regio-, chemo- and stereo-selective oxidations at inactivated C-H bonds and establishes the utility of directing auxiliaries to mediate the activity of highly promiscuous enzymes.

Recognizing the importance of product recovery, chapter 6 evaluates molecularly imprinted polymers for the selective purification of theobromine-containing molecules. When used for the solid-phase extraction, these materials allow for the near complete recovery of theobromine-containing products and starting materials from biocatalytic mixtures. This strategy represents an easily-tailored, effective, and reusable method of improving the recovered yield of theobromine-directed CYP3A4 oxidations.

iii

Résumé

L’hydroxylation énantiosélective d'un méthylène spécifique en présence de nombreux autres groupes semblables est défendablement la transformation chimique la plus difficile. Bien que les chimistes aient récemment fait du progrès vers l'hydroxylation de liaisons C-H inactivées, il existe des enzymes comme les cytochromes P450 (CYP) qui demeurent inégalées par rapport à leur spécificité et leur portée. La promiscuité de substrat démontré par plusieurs

P450 est souhaitable pour certaines applications de synthèse, mais une prévisibilité des produits difficiles, en plus de leur faible activité et stabilité empêchent l'avancement dans ce domaine.

Dans le chapitre 2 de cette thèse, nous évaluons plusieurs stratégies pour améliorer l'activité et la stabilité de CYP3A4. Ces stratégies comprennent son immobilisation à l'intérieur d’hydrogels moléculaire et de silice, en plus de sa modification chimique avec une variété d’anhydrides. Bien qu'aucunes des stratégies étudiées ici n’aient grandement amélioré l'utilité catalytique de l'enzyme, elles démontrent quand-même que CYP3A4 est très tolérantes envers la fonctionnalisation d’un grand nombre de ses résidus de surface et à la présence de concentrations extrêmement élevées de silice pendant la catalyse.

Reconnaissant le potentiel des enzymes possédant de petits sites actifs hydrophobes de catalyser des réactions Diels-Alder, chapitre 3 décrit la conception et l'application de plusieurs tests pour évaluer l'activité Diels-Alderase de CYP2E1. Bien que la présence de CYP2E1 n’a pas augmenter le taux des réactions étudiées ici, les résultats démontrent qu'il existe une interaction entre un ou plusieurs des substrats et l'enzyme, soit au site actif ou à un autre poche de liaison.

Résumé

Dans le chapitre 4, nous évaluons quatre auxiliaires pour leur capacité de diriger des oxydations par CYP3A4. Quand ils sont reliés à des substrats, nous avons trouvé que plusieurs de ces auxiliaires dirigent les oxydations par CYP3A4 à des liaisons C-H spécifiques. Bien que les auxiliaires explorés ici se trouvent à être limitées dans leur utilité, nous avons appris plusieurs leçons importantes que nous avons appliqué envers la conception d'une nouvelle génération d’auxiliaires à être discutés dans le chapitre suivant.

Dans le chapitre 5, nous démontrons l'utilité de la théobromine en tant qu’auxiliaires chimique pour contrôler la sélectivité des réactions du CYP3A4. Quand il est relié à des substrats peu coûteux, la théobromine achiral dirige la réaction, donnant naissance à des produits hydroxylés ou époxydés au niveau du quatrième carbone à partir de l'auxiliaire avec une sélectivité faciale pro-R. Cette stratégie fournit un système versatile et contrôlable offrant des produits oxydés de façon regio-, chimio- et stéréo-sélective à des liaisons CH inactivé. Elle démontre aussi l'utilité des auxiliaires par rapport à leur habileté de contrôler l'activité des enzymes hautement promiscues.

Reconnaissant l'importance de la récupération du produit, le chapitre 6 évalue des polymères à empreintes moléculaires pour la purification sélective de molécules contenant la théobromine. Quand ils sont utilisés pour l'extraction en phase solide, ces matériaux permettent la récupération quasi-complète de produits et de substrats contenant la théobromine à partir de mélanges biocatalytiques. Cette stratégie représente une méthode adaptable, efficace et réutilisable permettant d'améliorer le rendement récupéré des produits d’oxydations dirigée par la théobromine.

Acknowledgements

I would like to thank: My parents for granting me support and independence; my brother,

Leif, for always looking up to me; Kelly Livesley for her patience; Dr. Michael Eze for taking a chance on me; Dr. Désirée Vanderwel for making me love enzymes; Dr. Kayode Akinnusi for teaching me how to be a synthetic chemist; Amélie Ménard for more useful discussion than can be remembered; Erin May for helping me make the breakthrough; Dr. Lee Frieburger for testing the waters before me; Vanja Polic for being the ideal protégé; Dr. Faisal Aldaye for his continued kindness and inspiration; Nadim Saaed for preferring me; Dr. Eric Therrien for tolerating my questions; Omar Zahr for deactivating the lethargy field; Siqi Zhu for brightening the lab; Dr. Karine Auclair for teaching me what it takes to be a great professor and adviser; and

Mandy Slávik for being my best friend.

Table of Contents

Abstract…………………………………………………………………………………...…….ii

Résumé………………………………………………………………………………………....iv

Acknowledgements………………………………………………………………………….…vi

Table of Contents………………………………………………………………………...…....vii

List of Figures……………………………………………………………………………….…xv

List of Schemes…………………………….………………………………………………..…xx

List of Tables……………………………………………………………………………….….xxi

Abbreviations……………………………………………………………………………..…..xxii

References………………………………………………………………………………….…204

Select NMR spectra………………………………………………………………………………...

Select HPLC spectra………………………………………………………………………………..

Chapter 1: P450 enzymes and biocatalysis

1.1 Biocatalysis...... 2

1.1.1 A case for biocatalysis……………………………..…………………………..…..2

1.1.2 Biocatalysis: an overview………………………………………………………….4

1.1.3 Advantages of biocatalysis……………………………………………………...... 5

1.1.4 Techniques to improve existing biocatalysts…………...……………………….....7

1.2 P450 enzymes……………………………………………………………………………….8

1.2.1 P450 structural characteristics……………………………………………………..8

vii

Table of Contents

1.3 P450 catalytic mechanism………………………………………………………………...13

1.4 P450-catalyzed reactions…………………………………………..……………………...15

1.4.1 P450 catalyzed hydrocarbon hydroxylation……………………………………...15

1.4.2 P450-catalyzed hydroxylations of aromatic rings……………………………..…17

1.4.3 P450-catalyzed oxidation of heteroatoms……………………………………...... 18

1.4.4 P450-catalyzed epoxidation……………………………………………………...20

1.5 Chemical catalysts currently used for inactivated C-H bond oxidation…………….....22

1.5.1 Porhyrin-based chemical catalysts for inactivated C-H bond oxidation…………22

1.5.2 Non-porphyrin, metal-based catalysts for C-H oxidation……………………..…25

1.5.3 Non-metal containing catalysts/reagents for C-H bond oxidation…………….....26

1.5.4 General limitations of existing catalysts……………………………………...…..27

1.6 P450s as biocatalysts……………………………………………………………………....28

1.6.1 Precedent for P450s as biocatalysts…………………………………………....…28

1.6.2 Potential advantages of P450s over existing catalysts………………………...…29

1.6.3 CYP3A4 as a versatile biocatalyst…………………………………………….…29

1.7 Overall objective and general preface…………………………………………...…30

Chapter 2: Strategies for the Immobilization and Chemical Modification of P450s

2.0 Preface……………………………………………………………………………………..32

2.1 Introduction………………………………………………………...…………………..…33

2.1.1 P450 enzyme stability………………………………………………………...….33

2.1.2 Immobilization of P450s………………………………………………………....34

2.1.3 Multiple generic modifications of P450s…………………………………….…..35 viii

Table of Contents

2.2 Molecular hydrogel immobilization of CYP3A4…………………………………….…36

2.2.1 Assay……………………………………………………………………………..36

2.2.2 Results and discussion……………..…………………………………………..…38

2.3 Silica immobilization of CYP3A4………………………………………………………..40

2.3.2 Assay…………………………………………………………………………..…40

2.3.3 Results and discussion……………………………………………………………41

2.4 Chemical modification of CYP3A4 with citraconic anhydride and………………...…43 maleic anhydride

2.4.2 Assay……………………………………………………………………………..45

2.4.3 Results and discussion……………………………………………………………46

2.5 Conclusions………………………………………………………………………………..47

Chapter 3: Exploring CYP2E1 as a Biocatalyst for Diels-Alder Cycloadditions

3.0 Preface……………………………………………………………………………………..49

3.1 Introduction……………………………………………………………………………….50

3.1.1 Introduction to the Diels-Alder reaction…………………………….………...…50

3.1.2 Mechanism, tolerance, selectivity, and variations of the…………………….…..51 Diels-Alder reaction

3.1.3 Enzyme-catalyzed Diels-Alder reactions……………………………………...…54

3.2 CYP2E1 as a potential Diels-Alderase………………………………………………...…56

3.3 Substrate selection……………………………………………………………….……..…57

3.3.1 Dienes…………………………………………………………….…………...…61

3.3.2 Dienophiles……………………………………………………………….………62

3.4 Assays, results, and discussion…………………………………………………….……..63

Table of Contents

3.4.1 Synthesis of authentic standards of Diels-Alder products…………………….…63

3.4.2 GC-MS assay…………………………………………………………………..…63

3.4.3 Results of GC-MS Assay……………………………………………………...….64

3.4.4 UV assay……………………………………………………………………….…66

3.4.5 Results of UV Assay…………………………………………………………..…66

3.5 Conclusions……………………………………………………………………………..…67

Chapter 4: The Search for a CYP3A4-Directing Chemical Auxiliary

4.0 Preface…………………………………………………………………………………..…70

4.1 Introduction…………………………………………………………………………….…70

4.1.1 Chemical auxiliaries used in non-enzymatic reactions as……………………..…70 Inspiration

4.1.2 General strategy for design of a chemical auxiliary to control………………..…72 the selectivity of enzymes

4.2 4-(Trifluoromethyl)coumarin as a potential CYP3A4-directing……………………....75 chemical auxiliary

4.2.1 Synthesis of the 4-(trifluoromethyl)coumarin-based auxiliary-substrates………..76

4.2.2 Assay……………………………………………………………………………..77

4.2.3 Results and discussion……………………………………………………………77

4.3 Fluorescein as a potential CYP3A4-directing chemical auxiliary…………...…………78

4.3.1 Synthesis of fluorescein-based auxiliary-substrates……………………………...79

4.3.3 Results and discussion……………………………………………………………80

4.4 Theophylline as a CYP3A4-directing chemical auxiliary………………………………80

4.4.1 Synthesis of a theophylline-based auxiliary-substrate……………………………82

4.2.2 Results and discussion……………………………………………………………82 x

Table of Contents

4.5 Phthalamide-protected amines as interesting auxiliary-substrates for…………….…84 directing CYP3A4 reactivity

4.5.1 Synthesis of a phthalamide-based auxiliary-substrate……………………………85

4.5.2 Enzymatic reaction of 4.21 with CYP3A4……………………………..………..86

4.5.3 Results and discussion……………………………………………………………86

4.5 Conclusions………………………………………………………………………………...90

Chapter 5: Theobromine as a CYP3A4-directing auxiliary

5.0 Preface…………………………………………………………………………………..…93

5.1 Introduction……………………………………………………………………………..…94

5.2 Synthesis of theobromine-based auxiliary-substrates………………………………...…96

5.3 CYP3A4-catalyzed oxidation assays…………………………………………………..…100

5.3.1 Comparison of natural and non-natural cofactors……….…………………….…101

5.3.2 UV absorption properties of theobromine derivatives…………………………...102

5.3.3 Optimization of enzyme activity using purified CYP3A4 and…………………103 CYP3A4-containing membranes 5.4 Structural determination of CYP3A4-oxidized theobromine-based……………….....103 auxiliary-substrates

5.4.1 Structural determinations by mass spectroscopy fragmentation………………..104

5.4.2 Structural determination by further oxidation of the auxiliary-products……..…105

5.4.3 Structural determination by comparison to authentic standards………………..106

5.4.4 Structural determination by NMR……………………………………….…..…109

5.5 Products of the transformation of auxiliary-substrate-s by CYP3A4…………….....109

Table of Contents

5.5.1 Hydroxylation of theobromine-based auxiliary-substrates containing………....110 aliphatic substrates

5.5.2 Enzymatic reactions of theobromine-based auxiliary-substrates……………..…115 containing aromatic substrates

5.5.3 Functional group tolerance of CYP3A4 with various auxiliary-substrates….…117

5.5.4 CYP3A4-catalyzed oxidation of heteroatom containing auxiliary-substrates...... 120

5.5.5 CYP3A4-catalyzed oxidation of olefin-containing auxiliary-substrates………..122

5.5.6 Large scale (100 mg) CYP3A4-catalyzed oxidation of compound 5.5………....124

5.5.7 Verification of the role and the need of the theobromine auxiliary…….………124 5.6 CYP2D6–catalyzed oxidation of auxiliary-substrates…………………………....……125 5.7 In silico studies on the selectivity of CYP3A4–catalyzed oxidation of…………….....126 theobromine-based auxiliary-substrates

5.8 Cleavage of products from auxiliary-products……………………………………...…127

5.9 Conclusions………………………………………………………………………………128

Chapter 6: Molecularly Imprinted Polymers for Selective Purification of Theobromine-Containing Molecules

6.0 Preface…………………………………………………………………………………....131

6.1 Introduction………………………………………………………………………………131

6.2 MIP Preparation…………………………………………………………………………134

6.3 Assay…………………………...…………………………………………………………136

6.4 Results and discussion…………………………………………………………………...137

6.4.1 MIP efficacy………………………………………………………………………….…137

6.4.2 MIP Reusability…………………………………………………………………………141

6.4.3 MIP Selectivity……………………………………………………………………….…142

6.4.4 MIP purification of theobromine-containing compounds from dilute solutions………..144 xii

Table of Contents

6.4.5 MIP purification of theobromine-containing compounds from……………………...…146 CYP3A4-containing membrane-catalyzed oxidations.

6.5 Conclusions……………………………………………………………………………....149

Chapter 7: Contributions and Future Directions

7.1 Contributions...... 152

7.1.1 Contributions towards the design of enzyme-directing auxiliaries……………..152

7.1.2 Contributions towards the improved predictability of CYP3A4…………….....153

7.1.3 Contributions towards the biosynthesis of valuable chiral alcohols and…….....153 epoxides

7.1.4 Contributions towards the recovery of valuable molecules from biocatalytic….154 mixtures

7.2 Publications……………………………………………………………………………....155

7.3 Future Directions………………………………………………………………………...156

7.3.1 The development of new P450-directing auxiliaries…………………………....156 7.3.2 Enzyme engineering to modify CYP3A4 binding to…………………….……158 theobromine-substrates

7.3.3 Engineering more stable, selective, and predictable P450s…………………….158

7.3.4 P450s as Diels-Alderases…………………………………………………..…...159 7.4.5 Conclusions to future directions………………………………………………..160

Chapter 8: Experimental Protocols

8.1 General Methods……………………………………………………………………...…162

8.2 Biological studies………………………………………………………………..…….…165

8.1.1 Expression and purification of enzymes………………………………….….…165

xiii

Table of Contents

8.1.2 Enzymatic assays……………………………………………………………..…165

8.1.2.1 Optimized enzymatic transformation assay using purified enzyme…..165

8.1.2.2 Optimized enzymatic transformation assay using………………….....166 CYP3A4-containing membranes

8.1.2.3 100 mg scale CYP3A4-catalyzed transformation of compound 5.5….167

8.2 Assay of Diels-Alderase activity of CYP2E1…………………………………………..168

8.2.1 Synthesis of authentic standards of Diels-Alder products…………………...…168

8.2.2 GC-MS assay of Diels-Alderase activity of CYP2E1……………………….…169

8.2.3 UV assay of Diels-Alderase activity of CYP2E1……………………………....170

8.3 Synthetic protocols……………………………………………………………………....170

8.4 In Silico studies………………………………………………………………………..…206

8.5 MIP studies…………………………………………………………………………….....206

8.5.1 MIP synthesis…………………………………………………………………...206

8.5.2 Small volume MIP assay………………………………………………………..207

8.5.3 Large volume MIP assay to separate theobromine-containing compounds….....207 from complex mixtures.

8.5.4 Large volume MIP assay to separate theobromine-containing compounds….…208 from a reaction mixture of CYP3A4-catalyzed oxidation of 5.6.

xiv

List of Figures

1.1: The structure of taxol (paclitaxel)…………………………………………………………………...…3

1.2: Examples of the 3 general strategies used by biocatalysts to produce enantio0enriched …………..…5 products.

1.3: The 12 principles of green chemistry…………………………………………………………………...…………….6

1.4: The structures of various P450s demonstrating the extent to which the overall fold ……..………..…9 conserved.

1.5: The main loops that surround the heme in the structure of P450CAM…………………………………10

1.6: Important amino acids influencing the substrate-selectivity of CYP3A4……………………………12

1.7: An illustration demonstrating the flexibility of the CYP3A4 active site by a comparison of …….…13 the overall structures of CYP3A4 in complex with ketoconazole (A) and erythromycin (B)

1.8: The accepted general mechanism of P450 oxidation. The peroxide shunt is rendered in red1………14

1.9: The relationship connecting electron density and C-H bond strength………………………….……16

1.10: The CYP2A6 hydroxylation of an aromatic C-H bond of coumarin…………………………….…17

1.11: The CYP1A1 hydroxylation of an aromatic C-H bond of esterone…………………………………17

1.12: The reaction manifold of the P450 catalyzed oxidation of C-H bonds adjacent to…..………….….19 heteroatoms, in this case, nitrogen.

1.13: Examples of chiral heteroatom containing drugs………………….……………………………..…20

1.14: The P450 catalyzed epoxidations of various olefins demonstrates that the stereochemistry…….…21 about the olefin is retained in the epoxide product.

1.15: The strcutre of 5,10,15,20-Tetrakis(pentafluorophenyl)-21H,23H-porphine iron(III) chloride, …..23 a porphyrin-based catalyst for C-H bond oxidation.

1.16: A metalloporphyrin-catalyzed oxidation of a steroid derivative, demonstrating the difficulty….….24 inherent in predicting selectivity.

1.17: Christina White's iron-based catalysts and typical reaction outcomes………………………….…..25

1.18: The non-porphyrin, metal-based catalyzed oxidation of an electronically and sterically…….….…26 unique methylene.

1.19: The general structures of dioxeranes and oxizeradines…………………………………….…….…26 xv

List of Figures

1.20: The use of chiral oxaziridines in the enantioselective hydroxylation of tertiary C-H bonds. ………27 The catalyst is regenerated using hydrogen peroxide.

2.1: General strategies for the immobilization of enzymes……………………………………………….32

2.2: The chemical modification of an enzyme with R-X, where X is a leaving group and R can ……..…33 be any desired functionality.

2.3: CYP3A4 hydroxylates the benzyl group of BFC, resulting in a fluorescently active product…….…37

2.4: The relative activity of CYP3A4 immobilized in molecular hydrogel compared to that of free…….39 CYP3A4.

2.5: 6ß-Hydroxylation of testosterone by CYP3A4……………………………………………….………41

2.6: Structural comparison of testosterone and cortexolone………………………………………………41

2.7: The % activity of CYP3A4 with and without silica as a lyoprotectant………………………………42

2.8: Chemical modification of proteins using anhydrides………………………………………………...44

2.9: The location of all lysine (green) and cysteine (red) residues on CYP3A4…………………..….45

2.10: The % activity of CA and MA modified CYP3A4 compared to that of the unmodified……..…….46 enzyme.

3.1: DA catalyzed intramolecular [4+2] cycloaddition of cytochalasin B…………………………..……49

3.2: The [4+2] cycloaddition of cyclopentadiene and 1,4-Benzoquinone……………………………...…50

3.3: The oxo and aza variations of the Diels-Alder reaction………………………………………….…..52

3.4: FMOT interpretations of variations on Diels-Alder reactions………………………………….….…53

3.5: A simplified example of the Alder Endo rule……………………………………..………….…54

3.6: The proposed diesalderase-catlayzed cycloaddition towards the biosynthesis of lovastatin…..56

3.7: Binding site of CYP2E1 bound to 10-(1H-imidazol-1-yl)decanoic acid (from crystal structure)...…57

3.8: Maestro predicted binding mode of bicyclo[2.2.1]hept-5-ene-2-carboxylic acid in the active site.…61 of CYP2E1.

3.9: Selection of representative dienes for testing the Diels-Alderase activity of CYP2E1………………62

3.10: Selection of representative dienophiles for testing the Diels-Alderase activity of CYP2E1…..…...62

xvi

List of Figures

3.11: DAs used to probe CYP2E1 for DAase activity……………………………………………………64

3.12: The effect of the addition of CYP2E1 to the rates of DA cycloaddition reactions of …………..….65 representative dienes and dienophiles.

3.13: The effect of CYP2E1 on the rate of consumption of 3.4………………………………………..…67

4.1: Scheme demonstrating the use of an Evans’ auxiliary…………………………………………….…71

4.2: Examples of commercially available chemical auxiliaries………………………………………..…72

4.3: An ideal auxiliary binding to the enzyme and projecting a single C-H bond towards the ……….….74 reactive heme iron with facial selectivity.

4.4: Oxidative cleavage of 4.1 by CYP3A4………………………………………………………………75

4.5: A series of auxiliaries-substrates featuring 4.2 as the auxiliary………………………………………76

4.6: Oxidation of compound 4.13 by CYP3A4……………………………………………………………78

4.7: An auxiliary-substrate complexes featuring fluorescein as the auxiliary…………………………….79

4.8: Structural similarities of theophylline and theobromine with lisofylline, a natural substrate……..…81 of CYP3A4.

4.9: HPLC trace of the reaction mixture for CYP3A4-catalyzed transformation of 4.18 with CHP…… .83 as a cofactor surrogate.

4.10: Structural similarities of phthalamide-based auxiliary-substrates and lisofylline, a known ………85 natural substrate of CYP3A4.

4.11: HPLC trace for the reaction mixture of 4.21 with CYP3A4………………………………….….…87

4.12: The CYP3A4-catalyzed hydroxylation of compound 4.21 results in hydroxylation at C-4……..…88

4.13: The sites of CYP3A4-catalyzed hydroxylation on compound 4.21 and lisofylline, two………...…89 structurally similar molecules.

5.1: The structure of theobromine…………………………………………………………………………93

5.2: The derivatization of theobromine into auxiliary-substrate complexes which resemble…………….94 lisofylline.

5.3: The expected site of CYP3A4 oxidation on a theobromine-based auxiliary-substrate………………95 complexes compared to lisofylline.

5.4: Removal of the oxidized substrate from the theobromine auxiliary using similar conditions…….…95 used to remove phthalamide protecting groups from nitrogens

xvii

List of Figures

5.5: The molar absorption coefficient of caffeine, theobromine-substrate 5.11 and the……………...…102 corresponding theobromine-product, 5.47.

5.6: The ESI-MS fragmentation pattern of the CYP3A4-oxidized auxiliary-products suggest…………104 that oxidation is occurring at C-4 or further from the theobromine auxiliary.

5.7: The possible sites of CYP3A4-catalyzed oxidation of compounds 5.11 and 5.12 from ESI-MS…..105 fragmentation.

5.8: The method used to assign the absolute stereochemistry of 5.52-5.54, corresponding to the………109 CYP3A4-oxidized products of 5.5, 5.6, and 5.37, respectively.

5.9: The HPLC trace for analysis of the reaction mixture for the CYP3A4-catalyzed hydroxylation….112 of compound 5.5.

5.10: Chiral HPLC trace of 5.52 obtained by enzymatic reaction from reaction of 5.5 with CYP3A4…113

5.11: The CYP3A4-catalyzed oxidation of racemic 5.15, resulting in racemic 5.71……………………114

5.12: The CYP3A4-catalyzed oxidation of 5.7 and 5.8 afforded at least 3 oxidized products each….…115

5.13: The HPLC trace obtained for analysis of the reaction mixture for the CYP3A4-catalyzed…….…117 hydroxylation of compound 5.21.

5.14: The HPLC spectra of the CYP3A4-catalyzed hydroxylation of compounds 5.28 and 5.29………120

5.15: The HPLC strace of the CYP3A4-catalyzed hydroxylation of compound 5.34…………………..121

5.16: The HPLC trace for the reaction mixture of the CYP3A4-catalyzed hydroxylation of ………..…123 compound 5.36.

5.17: None of the compounds shown were transformed by CYP3A4 under catalytic conditions………125 Results for the last 7 compounds are from Chefson et. al.

5.18: Comparison of the structures of CYP3A4 and CYP2D6………………………………………..…126

5.19: Figure obtained from docking studies. Proposed orientation of compound 5.5 relative to……..…127 the heme group in the enzyme pocket of CYP3A4.

6.1: A simplified cartoon depicting the synthesis of an MIP specific for theobromine derivatives……133

6.2: The structural relationship between theobromine and caffeine……………………………….……134

6.3: The structures of the components used to construct the MIPs: methacrylic acid, 6.2, and ……...…135 ethylene glycol dimethacylate, 6.3.

6.4: The benzyleroxide initialized radical polymerization of 6.2 (R = H) and 6.3 (R = C6O2H8)……….135

xviii

List of Figures

6.5: The theobromine-containing compounds evaluated as templates for MIP preparation………….…136

6.6: The efficacy of 5 different MIPs when used to recover the compounds for which they were…...…138 designed from a concentrated solution (1 mM) in water.

6.7: Scatchard plot for 7 and the MIP designed for 7……………………………………………………137

6.8: The reusability of 5 different MIPs when used to recover the compounds for which they were…...141 designed from a concentrated solution (1 mM) in water.

6.9: The selectivity of 2 different MIPs when used to recover compounds structurally related to…...…143 those for which they were designed from a concentrated solution (1 mM) in water.

6.10: The selectivity of the MIP designed for 5.53 when used to recover 5.53 from relatively dilute.…145 (50 μM) solutions of water and LB broth.

6.11: The recovery of 5.53 and 5.6 from a ‘spiked’ CYP3A4-containing membrane solution and….…147 from a CYP3A4-containing membrane catalyzed oxidation of 5.6.

6.12: HPLC traces of different fractions following MIP purification of a CYP3A4-containing………..148 membrane-catalyzed oxidation of 5.6.

6.13: Fractions from the MIP purification of the CYP3A4-containing membrane-catalyzed ………..…149 oxidation of 5.6

7.1: A comparison of the prices for racemic and enantiopure materials……………………………..….154

7.2: The binding mode of metoprolol in the active site of CYP2D6 as predicted by FITTED…….……157

7.3: Binding site of CYP2E1 bound to 10-(1H-imidazol-1-yl)decanoic acid (from crystal structure2)…159

xix

List of Schemes

4.1: Synthesis of 4-(trifluoromethyl)coumarin based substrate-auxiliary complex…………………….…76

4.2: Synthesis of fluorescein-based auxiliary-substrate complex…………………………………………79

4.3: The synthesis of a theophylline-based auxiliary-substrate………………………………………..….82

4.4: The synthesis of a phthalamide-containing auxiliary-substrate………………………………………86

5.1: The general protocols for the synthesis of theobromine-substrates…………………………….….…97

5.2: Synthetic routes for the synthesis of theobromine-substrates 5.25-5.29 and 5.31...... 99

5.3: The synthesis of authentic standards for auxiliary-products 5.47, 5.52, 5.53 and 5.54 ……………107 corresponding to enzymatic reaction products from auxiliary substrates 5.11, 5.5, 5.6, 5.35 respectively.

5.4: The synthesis of authentic standards for auxiliary-products 5.65, 5.66, and 5.67……………….….108 corresponding to enzymatic reaction products from auxiliary substrates 5.35, 5.63 and 5.34 respectively.

5.5: Successful conditions for the cleavage of the theobromine auxiliary from auxiliary-products……..128

List of Tables

3.1: Selected dienophiles and dienes and several examples of possible DA products…..………58

3.2: Relative binding strength of DA products in the active site of CYP2E1 as predicted……...60 using the Maestro software suite.

5.1: The scope of auxiliary-substrates designed and synthesized to evaluate the…………….…96 CYP3A4-directing capability of theobromine.

5.2: Auxiliary-product obtained for the enzymatic reaction of each auxiliary-substrate C-4….111 regioselectivity, enantiomeric ratios, and isolated yields are listed.

5.3: Auxiliary-product obtained for the enzymatic reaction of auxiliary-substrates 5.18-5.21...116

5.4: Auxiliary-product obtained for the enzymatic reaction of auxiliary-substrates 5.24-5.33...118

5.5: Auxiliary-product obtained for the enzymatic reaction of auxiliary-substrate 5.34…….…121 5.6: Auxiliary-product obtained for the enzymatic reaction of auxiliary-substrates 5.35-5.38...122 8.1: Flow rates and linear gradient profiles used for HPLC analyses and purifications………..163

8.2: Diels-Alder 4+2 cycloaddition reactions of test dienes and dienophiles………………..…169

xxi

Abbreviations

BFC = 7-Benzyloxy-4-trifluoromethylcoumarin

BCE = Before Common Era

CA = Citraconic Anhydride

C-C = Carbon-Carbon

C-H = Carbon-Hydrogen

C-O = Carbon-Oxygen

CPR = Cytochrome P450 Reductase

CPO = chloroperoxidase

CYP = Cytochrome P450

DA = Diels-Alder DMP = Dess-Martin Periodinane DCM = Dichloromethane DMAP = 4-Dimethylaminopyridine DMSO = Dimethyl Sulfoxide DNA = Deoxyribonucleic Acid er = Enantiomeric Ratio ESI-MS = Electrospray Ionization Mass Spectrometry FMOT = Frontier Molecular Orbital Theory HOMO = Highest Occupied Molecular Orbital LAH = Lithium Aluminium Hydride LB = Lurial Broth LC = Liquid Chromatography LUMO = Lowest Unoccupied Molecular Orbital

xxii

Abbreviations

MA = Maleic Anhydride

MIPs = Molecularly Imprinted Polymers

MS = Mass Spectrometry

NAD(P)H = Nicotinamide Adenine Dinucleotide Phosphate ND = None Detected NMR = Nuclear Magnetic Resonance NOS = Nitric Oxide Synthase ON = Over Night PCC = Pyridinium Chlorochromate Rf = Retention Factor RP-HPLC = Reverse-Phase High Pressure Liquid Chromatography RT = Retention Time Tb = Theobromine TBAF = Tetra-N-Butylammonium Fluoride TBSCl = Tert-Butyldimethylsilyl Chloride TEA = Triethylamine THF = Tetrahydrofuran TsCl = 4-Toluenesulfonyl Chloride

UV = Ultraviolet

xxiii

Chapter 1

Chapter 1:

P450 enzymes and biocatalysis

Chapter 1

1.1 Biocatalysis

Biochemical reactions have existed long before humans first extracted metals from ores, formulated glazes for pottery and even before we first harnessed the wonders of fermentation.

Well before such activities, biocatalysts were busing themselves in the construction of molecules which, once discovered by natural product chemists, would become the modern paradigms of synthetic complexity. In many ways, the pursuit of asymmetric chemical catalysts, one of the overarching interests which defined the previous century of chemistry, can be thought of as an attempt to mimic the catalytic power of biocatalysts. Although chemical catalysts are capable of excellent versatility and convenience, it remains clear that biological catalysts are vastly superior in many respects. This fact is exemplified by considering the synthesis of Taxol.

1.1.1 A case for biocatalysis

The first successful synthesis of taxol (Figure 1.1; also referred to as paclitaxel) was reported in 1994 following 20 years of effort by over 30 research groups3. The successful synthesis required 55 chemical steps, many of them necessitating protections/deprotections, elevated temperatures, and large amounts of organic solvents. Ultimately, this route produced a total yield of less than 0.5%, a value which precludes industrial utility of this, or any other, total synthesis of taxol to date. Although Wender has achieved moderate improvements over this yield4, taxol must still be produced for commercial use by a semi-synthetic route involving plant cell fermentation5.

Chapter 1

Figure 1.1: The structure of taxol (paclitaxel).

Perhaps the most striking difference between the chemical and biological syntheses of taxol is the method by which the chiral C-O bonds are installed. In the chemical synthesese, the

C-O bonds at the chiral centers are mostly generated in the early steps and are subsequently protected and carried through the remaining steps before deprotection. These operations greatly increase the complexity of the synthesis as they each introduce a minimum of 2 chemical steps and associated issues with selectivity and reactivity. In contrast, the pacific yew tree, the organism from which taxol was first isolated6, takes a very different approach. In the biosynthetic route, the carbon skeleton is first constructed in approximately 12 enzymatic steps before a series of P450 enzymes oxidize 8 C-H bonds (mostly of methylenes) with stereo-, chemo- and regiocontrol7. Each step is performed at ambient temperature/pressure under aqueous conditions.

In the synthesis of complex and simple molecules, a route requiring only water as solvent and having fewer steps/reduced energy cost, while simultaneously improving the yield (and

Chapter 1 reducing waste) is preferred. Performing chemistry in this manner is the ultimate objective of harnessing P450 enzymes as biocatalytic alternatives to chemical catalysts.

1.1.2 Biocatalysis: an overview

Biocatalysis is the use of biomolecules, both naturally occurring and engineered, to catalyze chemical reactions. Surprisingly, this technology was involved in one of the very first of human inventions, ie. the brewing of beer, which has been dated to no more recent than the 6th millennium BCE8. However, the purposeful use of biocatalysis in vitro was first reported in 1897 when Eduard Buchner demonstrated that liquid removed from yeast cells could catalyze reactions (fermentation in this case). Since the late 1970s, the development of recombinant DNA techniques has given researchers access to more enzymes and in larger amounts, leading to the dramatic development of the modern field of biocatalysis. Today, biocatalysis is an important toll which is used, especially at the industrial scale, to perform reactions which are otherwise inaccessible or impractical by other means9. To a great extent, this progress is owed to the increased predictability of biocatalysis which has occurred in the past several decades10.

At the research laboratory scale, one of the most commonly used biocatalysts is baker’s yeast. This whole-cell reagent is ideal for the stereoselective reduction of ketones to alcohols11.

While this reaction is normally performed in water, it has been extended for use in ionic liquids12 and organic solvents13. Baker’s yeast lends itself well to the research lab scale as the catalytic agent; the dried yeast itself is as easy to handle, purchase, and quantify as most chemical catalysts. Additionally, upon completion of the reaction, the product can be easily extracted from the reaction mixture.

Chapter 1

The use of biocatalysts to produce enantio-enriched products generally uses 3 strategies: i) asymmetric reactivity, as exemplified by the 6ß-hydroxylation of testosterone by CYP3A414; ii) kinetic resolution, as exemplified by the enantioselective deacetylation of alkynyl esters by

Baker’s yeast esterase15; and iii) desymmetrization, as exemplified by the asymmetric demethylation of select meso compounds by pigs liver esterase (PLE)16 (Figure 1.2).

Figure 1.2: Examples of the 3 general strategies used by biocatalysts to produce enantio-enriched products.

1.1.3 Advantages of biocatalysis

Many of the advantages of biocatalysts over chemical catalysts can be illustrated with reference to the principles of green chemistry17 (figure 1.3).

Chapter 1

Figure 1.3: The 12 principles of green chemistry17.

Concerning the production of less hazardous chemical syntheses/the design of safer chemicals, toxicity is a non-issue when considering the enzymes commonly used in biocatalysis

(at least to the extent that finding evidence to the contrary has proved fruitless). Concerning the use of safer solvents, the vast majority of biosynthetic chemistry is performed in water, although biosynthesis can be performed in ionic liquids12 and organic solvents13. Concerning the issue of energy efficiency, biocatalysis is almost always performed at ambient pressures (although exceptions exist18) and, due to inherent issues with operational stability, biosynthesis is most commonly performed at near ambient temperatures. As discribed above by the example of the in vivo synthesis of taxol, biosynthesis can eliminate the requirements of protecting group

Chapter 1 manipulations through improved regio- and chemoselectivity, satisfying the requirement to reduce derivatization. Finally, biocatalysts are, by their very nature, derived from renewable feedstocks and are biodegradable, meeting a further 2 requirements.

Perhaps the greatest advantage of biocatalysts is that of selectivity19. Although excellent selectivity is somewhat unsurprising when considering examples of enzymes which are specific for one substrate, there are many example of enzymes which demonstrate excellent regio-, chemo-, and stereoselectivity while simultaneously having an immense substrate scope20. These enzymes are the best candidates for versatile biocatalysts with synthetic applications.

1.1.4 Techniques to improve existing biocatalysts

Although there are many biocatalysts known and many more that remain un/underutilized or undiscovered, there is always a desire to improve the activity, stability, solvent tolerance, and scope of biocatalysts for synthetic applications.

Examples abound of techniques used to improve the activity and stability of enzymes, including the incorporation of unnatural amino acids21, immobilization22, chemical modifications23, directed evolution24, logical mutations25, and the generation of fusion proteins by the incorporation of foreign domains26 and by linking enzymes to their protein cofactors27.

Although some enzymes are naturally tolerant to organic solvents (especially those of solvent-tolerant bacteria, thermophiles, halophiles and mesophiles)28, the incompatibility of most enzymes with non-aqueous conditions limits their utility in synthetic applications. Therefore, improving the tolerance of enzymes for organic solvents, supercritical fluids, ionic solvents, and biphasic systems are important pursuits. The solvent tolerance of enzymes has been improved by 7

Chapter 1 various techniques including immobilization29, chemical modification30, freeze-drying in the presence of lyoprotectants31, and the molecular imprinting of the substrate on the enzyme31.

Many biocatalysts used in industrial processes are highly tuned to perform a single reaction. At the industrial scale, this specificity is appropriate; however, at the scale of the research lab, versatility is of paragon importance. Thus, expanding the scope of existing biocatalysts is an important pursuit. Advancements towards this goal have been made mostly by using site-directed mutagenesis32 and directed evolution33. Another way to increase the substrate scope of biocatalysts relates to the improvement of solvent tolerance. As many substrates are not soluble or stable in aqueous solutions, improving the compatibility of a biocatalyst with different solvent systems may allow the application of the enzyme to entire classes of substrates which it would otherwise never encounter.

1.2 P450 enzymes

P450s are a ubiquitous class of heme-containing enzymes which participate in metabolic processes ranging from the biosynthesis of natural products in plants to the clearance of toxins in mammals34. Human P450s are most commonly expressed in the liver and intestines and function primarily in xenobiotic metabolism. Due to the inherent reactivity of the heme-iron, P450s catalyze a large range of reactions as diverse as C-H oxidation and C-C bond cleavage35.

1.2.1 P450 structural characteristics

While it‟s immediately apparent that all P450s share one defining structural characteristic, a heme prosthetic group bound in the active site, what is less obvious and somewhat surprising is the extent to which P450s share a highly conserved fold34, 36 (Figure 1.4). 8

Chapter 1

This is true for P450s found in mammals, plants, microbes, electron-transport chains and thermophilic bacteria37.

Figure 1.4: The structures of various P450s demonstrating the extent to which the overall fold is conserved34.

Chapter 1

Although the overall fold is maintained, the location of certain structural elements often varies. As a general rule, however, the closer that elements are to the heme, the more conserved they tend to be34. This is especially true of elements that are in direct contact with the heme, such as helices I and L34 (Figure 1.5). The regions which are least conserved are those involved in substrate access and binding; a result which explains the great variation in substrate specificity between isoforms.

Figure 1.5: The main loops that surround the heme in the structure of P450CAM.

Another striking similarity common to P450s, and indeed other enzymes such as nitric oxide synthase (NOS) and chloroperoxidase (CPO), is the rigid architecture used to position and protect the heme-iron proximal cysteine residue, which is partially responsible for positioning/orienting the heme and tuning the redox potential of the heme iron38.

Chapter 1

The first thermophilic P450 to be characterized by X-ray crystallography was CYP119, the 3-D structure of which melts at near 90oC37, an increase of approximately 40oC over most non-thermophilic P450s. The main structural differences between this P450 and its less thermodynamically stable counterparts are a series of aromatic residues that stack down one side of the protein. Disrupting the π-stacking interactions of the series of aromatic residues by point mutations can lower the melting point by 10-15 degrees per mutation, indicating that this structural feature contributes to the thermal stability39. Moreover, due to shorter surface loops and a truncated N-terminus, CYP119 is more compact and less flexible than most P450s34.

The P450 which is the primary focus of this thesis, CYP3A4, is a fairly typical P450 in terms of structure. In this isoform, site-directed mutagenesis studies have demonstrated that residues Phe30440 and Ala30541 control access to the catalytic centre, Phe304 and Asn206 control the positioning of substrates41, and Leu211 controls the size of the active site42 (Figure

1.6).

Chapter 1

Figure 1.6: Important amino acids influencing the substrate-selectivity of CYP3A4.

Finally, it is well established that CYP3A4 has a relatively open and flexible active site

(Figure 1.7), a feature which is thought to contribute to both its prolific substrate scope and to the tremendous difficulty of predicting its reactivity34.

Chapter 1

Figure 1.7: An illustration demonstrating the flexibility of the CYP3A4 active site by a comparison of the overall structures of CYP3A4 in the apo form (lighter coloured structures) and in complex (darker coloured structures) with ketoconazole (A, PDB# 2V0M) and erythromycin (B, PDB# 2J0D)34. Arrows indicate direction of domain movement upon binding.

1.3 P450 catalytic mechanism

For all known P450s, catalysis occurrs at the heme iron. The heme iron is in the +3 oxidation state in the resting state of the enzyme and is ligated to the sulfur of a proximal cysteine residue as well as to 4 porphyrin nitrogens and a water molecule.

The currently accepted general mechanism for P450 hydroxylation (Figure 1.8) begins with the substrate (R-H) binding to the enzyme near the heme iron (step 1). This is quickly followed by the first electron transfer from the redox partner co-enzyme to the iron, leading to a shift from the 6-coordinate species to the 5-coordinate species (step 2). Oxygen then binds to the reduced iron (step 3), facilitating the transfer of the second electron from the redox partner (step

4). This second electron transfer results in the formation of the iron-peroxo species which, in turn, loses water to form the reactive iron-oxo intermediate with a radical cation delocalized on

Chapter 1 the porphyrin (step 5). This reactive species is known as compound 134 and was first observed directly in 201043. Compound 1 is believed to be the intermediate which reacts with the bound substrate to abstract H radical. The proposed rebound mechanism of Groves44 next involves recombination of the resulting OH radical with the substrate radical (steps 6 and 7). After oxidation, the product (R-OH) is released and the iron coordinates with water, ready for another catalytic cycle (step 8).

Figure 1.8: The accepted general mechanism of P450 oxidation. Dashed lines indicate uncoupling pathways. RH is the substrate and CPR is the redox partner enzyme. RH is the substrate and CPR is the redox partner enzyme.

The identity of the redox partner(s) responsible for transferring electrons varies by isoforms and ranges from cytochrome P450 reductase (CPR) reduced by NADPH, to a 14

Chapter 1 ferredoxin reduced by ferredoxin reductase using NAD(P)H45. Sometimes, the 2 or 3 protein redox partenrs are fused into a single polypeptide, as with BM3 and thromboxane-A synthase46.

Furthermore, the peroxide shunt, a mechanistic ‘short-cut’, can eliminate the need for redox partners in many other P450s by converting iron(II) directly into the reactive iron-oxo upon reaction with a peroxide47.

As indicated in figure 1.8, the introduction of electrons from the redox partners into the

P450 hydroxylation mechanism can be uncoupled from substrate oxidation by the premature release of water or hydrogen peroxide. This uncoupling has been taken advantage of synthetically as a metabolic ‘short cut’ used to replace the expensive natural cofactors with inexpensive and convenient surrogates such as cumene hydroperoxide (CHP)48.

1.4 P450-catalyzed reactions

The hydroxylation of C-H bonds is the most commonly discussed P450-catalyzed reaction. However, P450s are known to catalyze a wide range of reactions including epoxidations, rearrangements, dehydrogenations, S-oxidations, N-oxidations, Baeyer-Villiger oxidations and even carbon-carbon bond cleavage reactions32. Because the thesis work described herein takes advantage of hydroxylation, epoxidation and heteroatom oxidations, only these reactions will be described here.

1.4.1 P450 catalyzed hydrocarbon hydroxylation

The P450 catalyzed hydroxylation of hydrocarbons is the archetypal P450 reaction.

P450s have been shown to demonstrate impressive selectivity towards the oxidation of a single

C-H bond in the presence of other sterically and electronically similar C-H bonds49. 15

Chapter 1

In addition to being highly selective, P450s are simultaneously extremely reactive. It is well established that C-H bond strength decreases as a function of the electron density surrounding the carbon in question50. As breaking the C-H bond is a necessary step in C-H oxidation, the stronger the bond connecting the carbon and hydrogen atoms, the more difficult the oxidation reaction51. Thus, in an electronic sense, tertiary C-H bonds are more easily oxidized than secondary C-H bonds, which in turn are more easily oxidized than primary C-H bonds (Figure 1.9). This trend is partially responsible for the chemoselectivity observed with most C-H activating chemical catalysts, which typically react most readily with tertiary C-H bonds52. Furthermore, the radical formed during many oxidation mechanisms is more stable on more highly substutituded carons. In contrast, P450 enzymes have been demonstrated to oxidize tertiary, secondary, and even some primary C-H bonds. Even more impressive is their ability to oxidize at energetically disfavored secondary or primary C-H bonds in the presence of much more energetically favorable positions while simultaneously maintaining stereocontrol (in the case of secondary C-H bonds), a feat which has remained unmatched by chemical catalysts53.

Figure 1.9: The relationship connecting electron density and C-H bond strength.

Chapter 1

As chiral C-O bonds are a common motif among natural products and the installation of such functionalities is often a limiting factor in synthetic efficacy, this is perhaps the most synthetically relevant reaction catalyzed by P450s.

1.4.2 P450-catalyzed hydroxylations of aromatic rings

P450s are known to catalyze the oxidation of aromatic C-H bonds in many compounds including the CYP2A6-catalyzed hydroxylation of coumarin32 (Figure 1.10), and the Oxy B catalyzed aromatic oxidation of vancomycin54. In general, P450-catalyzed hydroxylations of aromatic rings proceed by the oxidation of a π-bond and not by the direct insertion of the oxygen into a C-H bond34. The resulting epoxide can collapse into a ketone intermediate which then tautomerizes, forming the phenolic alcohol. However, the mechanism becomes much more complex in the presence of other aromatic C-O bonds, as exemplified by the CYP1A1 catalyzed aromatic oxidation of estrone55 (Figure 1.11). Because of the complex mechanism involved, product prediction is especially challenging. Nonetheless, it remains clear that P450s represent a powerful tool for the oxidation of aromatic rings and may enable orthogonal selectivities to current methods.

Figure 1.10: The CYP2A6 hydroxylation of an aromatic ring of coumarin.

Chapter 1

Figure 1.11: The CYP1A1 hydroxylation of an aromatic ring of esterone.

1.4.3 P450-catalyzed oxidation of heteroatoms

Although P450 oxidations of C-H bonds adjacent to heteroatoms are similar in many respects to normal P450-catalyzed hydrocarbon hydroxylations, the former reaction often results in the elimination of the alkyl group as an aldehyde34. Furthermore, the mechanisms of these two reactions differ: P450 catalyzed hydroxylations adjacent to heteroatoms proceed by an initial abstraction of an electron from the heteroatom, forming a radical cation. This step can be followed directly by the oxidation of the heteroatom (in much the same way as described for the hydroxylation of aliphatic C-H bonds) or by the deprotonation of the carbon adjacent to the heteroatom. If the carbon adjacent to the heteroatom is deprotonated, then the mechanism can result in either dealkylation or carbon-heteroatom bond cleavage (Figure 1.12).

Chapter 1

Figure 1.12: The reaction manifold of the P450 catalyzed oxidation of C-H bonds adjacent to heteroatoms, in this case, nitrogen34.

N- and S-oxides are not common functionalities found in natural or commercial products.

Important examples include esomeprazole (Nexium) and armodafinil (Nuvigil) (Figure 1.13).

When the need arises, these functionalities can be difficult to install, especially with stereocontrol.

Chapter 1

Figure 1.13: Examples of chiral heteroatom containing drugs.

As drug development moves away from natural products in their original forms and towards natural product derivatives and drugs designed de novo, the demand for reactions capable of selectively oxidizing hetereoatoms such as sulfur and nitrogen (ie. the installation of sulfoxides with stereocontrol) will increase. The availability of biocatalysts, such as P450s, that can be called upon to facilitate such oxidations will only accelerate the exploration of this relatively unexploited chemical space.

1.4.4 P450-catalyzed epoxidation

There are many examples of P450 catalyzed oxidations of olefins, resulting in epoxides.

Both X-ray56 and computational studies57 suggest that epoxidation is achieved, at least in part by the hydroperoxoferric intermediate. P450 epoxidation products are known to retain the original configuration about the olefin substrate, as exemplified by the epoxidations of oleic acid, cis- stilbene, and trans-[1-2H]-1-octene34 (Figure 1.14). This observation suggests that the epoxidation proceeds via a concerted mechanism.

Chapter 1

Figure 1.14: The P450 catalyzed epoxidations of various olefins demonstrates that the stereochemistry about the olefin is retained in the epoxide product.

Although epoxides are less common in natural and commercial products than alcohols or ethers, they represent important synthons in chemical synthesis. Thus, the ability to generate epoxides with regio- and sterocontrol is an important pursuit in organic synthesis. This is exemplified by the importance and general utility of the Sharpless epoxidation58. Most methods used to achieve chiral epoxides rely upon the proximity of directing groups, such as the allylic alcohol group in the Sharpless epoxidation. For this reason, the epoxidation of inactivated, terminal olefins has proved a difficult challenge. Indeed, although there exist chemical catalysts capable of epoxidizing terminal olefins59, no such catalyst has achieved a significant measure of stereoselectivity. Currently, the best chemical method to access chiral terminal olefins is the kinetic resolution of racemic epoxides with catalysts such as those developed by Jacobsen60. In

Chapter 1 contrast, P450 enzymes are capable of performing asymmetric epoxidation reactions with excellent stereocontrol61.

1.5 Chemical catalysts currently used for inactivated C-H bond oxidation

Before discussing how P450s can revolutionize the oxidative toolkit available to chemists, it’s first necessary to survey the catalysts currently available for C-H bond oxidations.

As monohydroxylation has the greatest immediate synthetic utility of any P450 catalyzed reaction (see section 1.2.3.1), the discussion will be focused on existing reagents which catalyze this reaction and their limitations.

1.5.1 Porphyrin-based chemical catalysts for inactivated C-H bond oxidation

There exist various classes of metal-containing catalysts for hydroxylations at inactivated

C-H bonds. Commonly, an iron metal is found at the centre of these catalysts; inspired by and mimicking the function of the iron found in the heme group of P450s and similar enzymes (such as the chlorophyll containing enzymes of the electron transport chain found in plant cells). By attenuating the redox potential of the iron through ligation of various functionalities, the reactivity can be tuned to achieve oxidations at inactivated C-H bonds. The surrounding functionalities also serve to introduce some degree of chemo-, regio-, and stereoselectivity.

Metalloporphyrins are catalysts designed with inspiration from the catalytic center of

P450s. These catalysts consist of porphyrin derivatives functionalized in an effort replace the amino acid residues found in the P450 active site. Although, metalloporphyrins can contain a range of metals including cobalt, copper, and vanadium, only those containing iron, ruthenium, and manganese are commonly used for oxidations35. 22

Chapter 1

Several commercially available metalloporphyrins are commonly used in alkane oxidation chemistry, such as 5,10,15,20-Tetrakis(pentafluorophenyl)-21H,23H-porphine iron(III) chloride (Figure 1.15). However, as this molecule is achiral, stereoselectivity is not observed.

Although such reagents have been used to achieve transformations such as the oxidative cleavage of DNA62, typically, numerous hydroxylation and epoxidation products are observed.

Figure 1.15: The structure of 5,10,15,20-Tetrakis(pentafluorophenyl)-21H,23H-porphine iron(III) chloride, a porphyrin-based catalyst for C-H bond oxidation.

Chapter 1

To impart a greater degree of selectivity, the effects of different functionalizations of metalloporphyrin have been explored. The first report of the use of a chiral metalloporphyrin to produce an enantio-enriched product by oxidation was made in 198363. Since this time, the functionalization of the porphyrin derivatives has given access to a number of chiral, manganese and rhodium-containing metalloporphyrins which have been shown to be synthetically useful

(Figure 1.16).

Figure 1.16: A metalloporphyrin-catalyzed oxidation of a steroid derivative, demonstrating the difficulty inherent in predicting selectivity. TPFPP = meso-tetrakis(pentafluorophenyl)porphinato dianion.

However, poor selectivity and predictability are still problematic. In most cases, selectivity, even regioselectivty, cannot be predicted and can only be determined empirically.

Furthermore, multiple oxidation products including aliphatic alcohols, aromatic alcohols, and carbonyls are often observed64.

Chapter 1

1.5.2 Non-porphyrin, metal-based catalysts for C-H oxidation

A second class of metal-containing catalysts used to achieve oxidations at inactivated C-

H bonds are non-porphyrin, metal-based catalysts developed by Christina White52a and others.

While these catalysts are capable of delivering high yields and are reasonably predictable, their regio- and chemoselectivities are almost completely dictated by the electronic and steric nature of the substrate: oxidations at tertiary C-H bonds are largely favored over secondary C-H bonds and when these catalysts have been tuned for methylene oxidation, the resulting reactions suffer from over-oxidation, giving ketones and the loss of chirality (Figure 1.17).

Figure 1.17: Christina White's iron-based catalysts and typical reaction outcomes52a.

There are cases in which these catalysts have achieved hydroxylations at inactivated, secondary C-H bonds, but these examples are rare and limited to specific cases involving the

Chapter 1 oxidation of electronically and sterically unique secondary C-H bonds65 (Figure 1.18).

Furthermore, while these catalysts have been shown to exhibit diastereoselectivity, no examples of enantioselective catalysts have been reported so far. Although this chemistry represents a large achievement in the field of inactivated C-H bond oxidation, it seems unlikely that such catalysts will be capable of achieving hydroxylation at a single specific methylene in the presence of other electronically and sterically similar groups, as is possible with P450s.

Figure 1.18: The non-porphyrin, metal-based catalyzed oxidation of an electronically and sterically unique methylene52a.

1.5.3 Non-metal containing catalysts/reagents for C-H bond oxidation

Both dioxiranes and oxaziridines (Figure 1.19) have been shown to oxidize inactivated C-

H bonds.

Figure 1.19: The general structures of dioxiranes and oxaziridines.

Chapter 1

In these reactions, dioxiranes are non-catalytic, achiral, and oxidize at tertiary C-H bonds or nitrogens66. While there have not been significant efforts to produce chiral dioxiranes, the use of chiral oxaziridine-containing catalysts for C-H oxidation has been well explored.

The group of Justin DuBois has developed a number of chiral oxaziridine derivatives and demonstrated their use in the enantioselective oxidation of inactivated C-H bonds67 (Figure

1.20).

Figure 1.20: The use of chiral oxaziridines in the enantioselective hydroxylation of tertiary C-H bonds67. The catalyst is regenerated using hydrogen peroxide.

However, this catalyst appears completely unable to catalyze hydroxylations at any energetically less favorable C-H bonds such as those of methylene or methyl groups and appears unable to distinguish among multiple sterically and electronically similar methines.

1.5.4 General limitations of existing catalysts

As demonstrated by the above examples, there exist many types of catalysts for oxidations at inactivated C-H bonds. However, even the best of these suffer from critical

Chapter 1 limitations. The few catalysts, such as metalloporphyrins, which are capable of achieving oxidations at inactived methylenes are difficult to predict, poor yielding, prone to giving multiple products and over-oxidation, and are difficult to prepare. While the reactivities of the remaining catalysts are easier to predict, they almost exclusively oxidize tertiary C-H bonds and the regio- and chemoselectivities are dictated by electronic and steric effects, precluding discrimination between groups which are similar in these regards. An ideal catalyst would combine the strengths from each group while avoiding each limitation; it would enable monooxidations at methines, methylenes, and methyl groups with predictable regio-, chemo-, and stereoselectivity while avoiding over-oxidation.

1.6 P450s as biocatalysts

1.6.1 Precedent for P450s as biocatalysts

As discussed in section 1.2, P450s catalyze a wide variety of reactions including hydroxylations, epoxidations, heteroatom oxidations, N- or O-dealkylations, dehydrogenations, rearrangements, and even carbon-carbon bond cleavage reactions. These enzymes act upon thousands of substrates with exceptional regio-, chemo-, and stereoselectivities. Despite these impressive attributes, the P450s which have so far been exploited as biocatalysts are highly specific for the transformation of single substrates, limiting their utilty as versatile synthetic tools68. The most notable examples of P450s in industrial biocatalysis are the 11α-hydroxylation of progesterone by Rhizopus, and the 6 ß-hydroxylation of compactin by Mucor hiemalis69.

However, the exact enzymes responsible for these reactions have not been characterized or this information has not been made public. Select P450s are also used in industy for the preparation of artimisinin70 and, as previously mentioned, various taxols71. However, the predictability and 28

Chapter 1 versatility of P450s must be improved in order to transform these enzymes from highly specialized catalysts into common tools which research chemists would consider employing over the various catalyses discussed in the previous section. Although this challenge is considerable, the potential advantages of such a strategy are very attractive.

1.6.2 Potential advantages of P450s over existing catalysts

P450s combine two attributes highly desired in versatile catalysts: scope and selectivity.

These enzymes account for over 75% of all reactions involved in drug metabolism72 and are capable of discerning specific inactivated methylenes in compounds with dozens of electronically and sterically similar groups. P450 enzymes are simultaneously capable of both power and mildness; they oxidize most varieties of C-H bonds across a wide range of bond strengths, even those of methyl groups, while avoiding over-oxidation of relatively weaker C-H bonds. Unlike chemical catalysts, P450s can be used in vitro (in both water and organic solvents to some extent31) and in vivo (commonly over-expressed in bacterial cells), enabling not only a wider variety of applications but also the ability to perform fermentive oxidations at industrial scales.

Poor activity/stability and poor predictability are the main limitations of P450s which are hindering their utility as widely used biocatalysts.

1.6.3 CYP3A4 as a versatile biocatalyst

CYP3A4 is an ideal candidate for development as a versatile P450 biocatalyst. In addition to being responsible for the metabolism of over half of all known drugs, this isoform is one the most studied and well characterized of all human P450s. Furthermore, as the staggering 29

Chapter 1 substrate promiscuity of CYP3A4 arises from its large, flexible active site (see section 1.2.3), there is much opportunity for the tuning of reactivity/selectivity by logical mutations, directed evolution or by other means.

The Auclair lab has previously addressed the issues hindering the use of P450s including the need for expensive, natural cofactors48, solvent tolerance31, and poor stability14 (to some extent). However, poor predictability remains a central issue which must be addressed.

1.7 Overall objective and general preface

The goal of this research is to harness of the catalytic power of P450s for use as powerful and versatile biocatalysts. This goal was approached by attempting to overcome the most precarious limitations of P450s; their inherently poor activity/stability in vitro and, most importantly, their poor predictability. The issues of activity/stability were addressed by several methods including immobilization and chemical modification. The central issue of predictability was addressed by the logical design of several P450-directing chemical auxiliaries capable of forcing the enzyme to act at a specific site. Finally, molecularly imprinted polymers were used to enhance the recoverability of the products and starting materials of CYP3A4 oxidations from complex mixtures.

Chapter 2

Chapter 2:

Strategies for the Immobilization and Chemical Modification of P450s

Chapter 2

2.0 Preface

Enzyme immobilization is a strategy commonly used to improve the utility of enzymes as biocatalysts. Enzymes are immobilized either covalently or not, in or onto an insoluble and inert material (Figure 2.1). Recognizing the need to improve the otherwise poor stability of P450 enzymes, 2 techniques are explored in this chapter to immobilize CYP3A4.

Figure 2.1: General strategies for the immobilization of enzymes.

A second strategy commonly used to increase enzyme activity/stability involves multiple generic modifications (Figure 2.2). Thus, the derivatization of surface lysines using various reactive anhydrides was explored.

Chapter 2

Figure 2.2: The chemical modification of an enzyme with R-X, where X is a leaving group and R can be any desired functionality.

The experimental work described in this chapter was performed by me and Andrea Hill, an undergraduate student in the Auclair lab: specifically, I performed all of the experiments related to molecular hydrogel immobilization and chemical modification and Andrea Hill assisted with the silica immobilization experiments.

2.1 Introduction

2.1.1 P450 enzyme stability

P450 enzymes are well known to suffer from poor stability in vitro and this is especially true of the mammalian isoforms73. In particular, CYP3A4 has a relatively flexible structure which is highly susceptible to denaturation, even under physiological conditions32. Problems also exist with its kinetic stability, mostly arising from the generation of reactive peroxides inside the active site as a side-reaction74. Due to the local concentration of these reactive oxygen bi- products near the heme and the most sensitive residues of the protein, it is difficult to effectively quench this process using antioxidants, although the use of superoxide dismutase and catalase

Chapter 2 has been shown to help somewhat75. Overcoming poor in vitro stability is of key importance to the development of P450s as usefull biocatalysts76.

It is difficult to decouple the contributions of stability and activity when performing assays on enzymes having such a short catalytic lifespan (in our experience, reactions using

CYP3A4 under catalytic conditions are quiescent after periods as short as 1-1.5 hours).

Furthermore, as the reaction yield is, by definition, the mathematical product of activity

(molecules turned over per unit time) and stability (which, in practical terms, is the time for which the enzyme is active), it stands to reason that if the enzyme activity were significantly increased without also increasing the kinetic stability, the proportional decrease in catalytic lifespan would result in roughly the same overall product yield. As increased product yield

(turnover) is ultimate goal of improving P450 stability, the assays performed will be evaluated by the final enzymatic reaction yield.

2.1.2 Immobilization of P450s

Non-covalent immobilization is a facile and operationally gentle strategy to improve the utility of enzymes as biocatalysts77. In addition to often improving enzyme stability78, immobilization enhances the convenience and economics of enzymatic reactions as, upon completion of the reaction, the enzyme can be easily removed from the mixture and potentially recycled for further reactions79. These features are highly desirable for the development of human P450s as biocatalysts80.

There exist many examples where the immobilization of P450 enzymes has been successful. Recent efforts have included the immobilization of P450s on coated silver

Chapter 2 electrodes81, on gold82, on patterned lipid membranes83, and in reverse micelles84. However, the majority of successful techniques have required the use of protein engineering, highly specialized equipment, or have generated an enzyme system that is better suited to microscopy than biocatalysis. The techniques best suited to improve the utility of mammalian P450s in biocatalysis are those that are compatible with the already unstable nature of these enzymes (in terms of temperature, pH, pressure, and solvent tolerance), are facile/efficient, require a minimum of manipulations, and are compatible with already established P450 activity assays.

2.1.3 Multiple generic modifications of P450s

The chemical modification of P450 side chains is a potentially useful strategy to improve the activity/stability of P450s. However, to insure that the enzymes remain active and are modified homogenously, the reaction must show some selectivity, be efficient, and be compatible with whatever solvent the enzyme is suspended in85. One such example used in other enzyme systems is the covalent modification of lysine residues with activated molecules such as citraconic anhydride or maleic anhydride86. This type of modification has been shown to significantly improve the catalytic activities, thermal stabilities, and tolerance to organic solvents of enzymes, such as in the cases of a lipase from Rhizomucor miehei30 and horseradish peroxidase87. Additionally, this reaction has been shown to proceed with high efficiency and selectivity under gentle conditions (atmospheric temperatures and pressures in water or buffer).

Although several groups have reported on the chemical modification of P450s lysine residues, none were performed for the purpose of improving activity/stability. Furthermore, none of the previous studies have featured the enzyme of interest here, CYP3A4. Instead, the majority of modification studies have focused on the functionalization of lysine residues in other P450 35

Chapter 2 isoforms to probe reactivity and selectivity of an unidentified microsomal P45088, to better understand enzyme-cofactor interactions of CYP2B189, and to determine specific sites of substrate binding in CYPSCC 90. Other strategies commonly used to chemically modify P450s include cross-linking CYPSCC to its redox parter91, the use of mechanism-based inhibitors to probe active-site-substrate interactions of CYP1A292, and the oligomerization of the CYP3A493.

2.2 Molecular hydrogel immobilization of CYP3A4

A molecular hydrogel is a 3-dimentional chemical network assembled from a hydrophilic polymer and is capable of containing relatively large amounts of water, causing the network to swell when exposed to aqueous solutions94. Some enzymes immobilized inside of molecular hydrogels exhibit super-activity and improved stability in both water and organic solvents compared to their unimmobilized counterparts78. The assembly of molecular hydrogels is straight-forward, requiring a minimum of specialized equipment, and because the enzyme- containing gels are added directly to reaction mixtures along with substrate and cofactor, modifying existing assays is facile. Different types of molecular hydrogels have been described but for the purposes of these investigations, a hydrogel made from sodium carbonate, Fmoc-L- lysine, and Fmoc-L-phenylalanine was chosen for ease of handling and because this hydrogel is among the best studied78. CYP3A4 was encapsulated in this molecular hydrogel and the activity of the immobilized enzyme was compared to that of free CYP3A4.

2.2.1 Assay

Ideal substrates for enzyme activity assays undergo a strong change in fluorescent or UV properties upon enzymatic reaction, are highly soluble in the reaction mixture, and are stable

Chapter 2 when not under catalytic conditions. Substrates for this assay were chosen based upon previous experience in the Auclair lab14. For CYP3A4, 7-benzyloxy-4-trifluoromethylcoumarin (BFC) is an ideal substrate. This molecule gives a distinctive and easily quantifiable fluorescent emission at 530 nm upon oxidative cleavage of the benzyl group by CYP3A4 (Figure 2.3).

Figure 2.3: CYP3A4 hydroxylates the benzyl group of BFC, resulting in a fluorescently active product.

Originally, it was envisaged to monitor the enzyme activity in real-time as this type of assay can give information about both the activity and stability in enzymatic systems, however the translucent nature of the molecular hydrogel complicated taking fluorescence measurements in real-time and necessitated the use of an endpoint measurement instead. End points of 2 hours were selected as CYP3A4 reactions are typically > 90% complete in 1.5 hours. Thus, an increase in yield at the endpoint suggests an increase in either the activity or stability (or some combination thereof) of the enzyme.

The enzyme, BFC, molecular hydrogel and the cofactor surrogate, CHP, were mixed in buffer and stirred for 2 hours. Aliquots of the reaction mixtures were analyzed by fluorescence to

Chapter 2 determine yield. To evaluate the ability of molecular hydrogels to improve the tolerance of enzymes to organic solvents, toluene was used in place of buffer in a set of homologous assays.

2.2.2 Results and discussion

Regardless of whether the reaction was performed in water or toluene, CYP3A4 immobilized in molecular hydrogels produced significantly less product than free CYP3A4

(Figure 2.4). In each case, the yield from the immobilized enzyme was roughly half that of the free enzyme.

Figure 2.4: The relative activity of CYP3A4 immobilized in molecular hydrogel compared to that of free CYP3A4. The enzyme (50 nM), BFC (15 μM), molecular hydrogel (40 μL) and CHP (0.1 mM) 38

Chapter 2 were mixed in buffer (potassium phosphate buffer, 0.1 M, pH 7.4, 300 μL total volume) or toluene and stirred at 250 RPM and 37oC for 2 hours. Aliquots of the reaction mixtures were analyzed by fluorescence (excitation at 409 nm and emission at 530 nm) to determine the yield. Reactions were performed in duplicate.

One possible explanation for the observed decrease in yield could involve the potential decrease in permeability of the hydrogels to the substrate, limiting the rate at which the substrate can diffuse into (or the product out of) the CYP3A4 containing hydrogels. It is possible that different hydrogels may show more promising results.

2.3 Silica immobilization of CYP3A4

There are many examples of enzymes being immobilized in/on silica for the purpose of enhancing activity/stability95. There has even been a report of the immobilization of an enzyme via the genetic engineering of a silica producing diatom96. However, the inspiration for the research described here was a report describing the use of silica as a lyoprotectant, effectively immobilizing the enzyme to increase stability and solvent tolerance97. This report described how the addition of sucrose, a common lyoprotectant, and silica greatly increased the tolerance of lyophilized enzymes to organic solvents and improved the enzyme‟s stability in general. The

Auclair lab has previously had success with the lyoprotection of P450s with sucrose14 and the possibility of a single additive further enhancing the protective effects was enticing.

2.3.2 Assay

Testosterone was chosen as the substrate of this assay as the CYP3A4-hydroxylated product is easy to resolve by RP-HPLC with UV absorption detection (Figure 2.5). CYP3A4 was 39

Chapter 2 combined with sucrose, testosterone and silica in buffer and the mixture was lyophilized until dry. The residue was redissolved in buffer with the cofactor surrogate, CHP, and shaken for 2 hours at 37oC. The reaction product was extracted into chloroform (2 x 1 mL) before quantification as the presence of insoluble silica and sucrose in the aqueous reaction mixture would obscure UV absorbance measurements.

Figure 2.5: 6ß-Hydroxylation of testosterone by CYP3A4.

Cortexolone, an internal standard with similar retention and partitioning properties to testosterone (Figure 2.6), was added to the mixture before extraction to eliminate errors arising from differential extractions. The yield of each reaction was measured by RP-HPLC using UV detection.

Figure 2.6: Structural comparison of testosterone and cortexolone.

Chapter 2

2.3.3 Results and discussion

The addition of silica as a lyoprotectant did not appear to have any beneficial effect on the activity/stability of CYP3A4 in buffer (Figure 2.7). This result is surprising as it seems reasonable that such large concentrations of additives should be expected to have a significant effect on the enzymes‟ catalytic performance, whether it is beneficial or not.

Figure 2.7: The % activity of CYP3A4 with and without silica as a lyoprotectant. CYP3A4 (100 nM) was combined with sucrose (180 mM), testosterone (667 μM) and silica (15 mg) in potassium phosphate buffer (417 μL, 0.1 M, pH 7.4) and the mixture was lyophilized until dry. The residue was redissolved in phosphate buffer (700 μL, 0.1 M, pH 7.4, 300 μL total volume) with CHP (1.7 μM) and was shaken at 250 RPM at 37oC. Experiments were run in triplicate.

Chapter 2

Although the experimental ratio of silica to sucrose was taken directly from the report that this experiment was based upon, it’s possible that other ratios may lead to a different experimental outcome. It should be noted that the effect of silica as a lyoprotectant was only probed in buffer and that it’s possible that the protective qualities of silica may be better suited to catalysis in organic solvent.

2.4 Chemical modification of CYP3A4 with citraconic anhydride and maleic anhydride

The covalent functionalization of amino acid side chains by chemical ligation has been shown to increase the stability of enzymes85, 88b, 90-91. Chemical modification of enzymes using citraconic anhydride (CA) or maleic anhydride (MA) can be easily achieved under gentle reaction conditions90, an important factor when using relatively unstable mammalian P450s.

Upon incubation with the anhydride in buffer, enzymes are covalently modified at the ε-amino groups of lysine residues via a simple substitution reaction86 (Figure 2.8). The observed improvements in activity/stability are commonly explained by beneficial effects of exchanging the positive charges (at physiological pH) of the lysine amines to the negative charges of the carboxyl groups added with this reaction85. These negative charges are thought to cause beneficial reorientation of the modified side chains, although the exact mechanism by which the enzyme is stabilized remains poorly understood.

Chapter 2

Figure 2.8: Chemical modification of proteins using anhydrides.

CYP3A4 has no fewer than 30 lysine residues on the enzyme surface (Figure 2.9), the region most accessible to chemical modifications. It was expected that the modification of most of these residues (making up more than 5% of the total number of residues of the protein) would have a drastic effect on the activity of the enzyme.

Chapter 2

Figure 2.9: The location of all lysine (green) and cysteine (red) residues on CYP3A4. Almost all lysine residues are found on the surface while all cysteine residues are internal.

2.4.2 Assay

CYP3A4 modified with CA or MA was reacted with BFC and CHP (as in previous assay described in section 2.2.1). The enzymatic yield was measured as an end-point by fluorescence.

Chapter 2

2.4.3 Results and discussion

The modification of surface lysine residues with citraconic anhydride or maleic anhydride had negligible effect on the activity/stability of CYP3A4 in buffer (Figure 2.10). This result is surprising as the extent to which the protein is being modified is considerable.

Figure 2.10: The % activity of CA and MA modified CYP3A4 compared to that of the unmodified enzyme. After modification, the enzymes (100 nM) were combined with BFC (15 μM) and CHP (0.1 mM) in potassium phosphate buffer (0.1 M, pH 7.4, 300 μL total volume) and stirred at 250 RPM and 37oC for 2 hours. Aliquots of the reaction mixtures were analyzed by fluorescence (excitation at 409 nm and emission at 530 nm) to determine the enzymatic reaction yield. Reactions were performed in duplicate and the enzymatic modifications were repeated separately each time.

Chapter 2

It seems likely from these results that surface lysines have only minimal control over enzyme activity and stability of CYP3A4. However, the knowledge that surface lysine residues can be modified without detriment to the enzyme’s catalytic capacity may be of interest to researchers hoping to chemically conjugate CYP3A4 (or other P450s) to other chemical species.

Because of their lower absorbance, cysteines are normally the preferred residue upon which to functionalize enzymes. However, many proteins, including CYP3A4, lack easily accessible surface cysteines. For such proteins, the ability to conjugate at lysine residues without determent to the activity of the enzyme may be of special interest.

2.5 Conclusions

This chapter has described efforts towards finding a facile and convenient method to improve the activity of CYP3A4. Towards this goal, 3 widely used and relatively robust methods were tested. None of these methods resulted in the greatly enhanced catalytic utility of the enzyme. Interestingly, however, CYP3A4 appears to be highly tolerant to functionalization at a large number of surface residues and also to the presence of extremely high concentrations of silica during catalysis. Either of these findings could certainly inform future efforts towards the use of P450s as biocatalysts. The results described in this chapter also demonstrate the difficulty inherent in overcoming the poor activity and stability of mammalian P450s and confirm the importance of giving attention to this problem in the future. While this work was progressing, the efforts of Amelie Menard, a graduate student in the Auclair group, towards the selective functionalization and cross-linking of CYP3A4 began to generate very encouraging results and, ultimately, her approach was prioritized.

Chapter 3

Chapter 3:

Exploring CYP2E1 as a Biocatalyst for Diels-Alder Cycloadditions

Chapter 3

3.0 Preface

The Diels-Alder reaction (DA) is perhaps the most powerful chemical synthetic tool in use today. The ability to form two C-C bonds and up to four new stereo-centers in a single step provides expedient access to a variety of complex, ring-containing structures (Figure 3.1)98.

Conveniently, many natural products contain functionalities which resemble DA products, bridging the gap between elegant reactivity and remarkable utility.

Figure 3.1: DA catalyzed intramolecular [4+2] cycloaddition of cytochalasin B.

The scalability of the DA lends itself to use in both a small scale for research purposes and an industrial scale for the manufacture of drugs and other valuable materials. There exist several asymmetric DA conditions in which the stereochemistry is controlled by small organic catalysts and auxiliaries99. However, examples of enzyme-catalyzed DA reactions are rare, especially for bimolecular DA reactions100. Recognizing the potential for enzymes containing small, hydrophobic active sites to catalyze DA reactions, it was envisaged that such enzymes may find utility as general DA catalysts. This chapter will describe several assays which I designed, performed and analyzed to probe the Diels-Alderase (DAase) activity of CYP2E1.

Chapter 3

I performed all of the docking studies described and a graduate student in the Auclair

Lab, Amelie Menard, assisted with the activity assays and produced several authentic standards of the expected DA products

3.1 Introduction

3.1.1 Introduction to the Diels-Alder reaction

Professor Otto Diels and his student, Kurt Alder, discovered the [4+2] cycloaddition of cyclopentadiene and quinone (Figure 3.2) in 1928; a finding which was published in their landmark paper101. Despite a warning to other chemists that they “reserve for [themselves] the application of the reaction”, the DA has been greatly exploited since that time and its utility was elegantly demonstrated in the total synthesis of cantharidin in 1951102 and the first total synthesis of morphine in the following year103.

Figure 3.2: The [4+2] cycloaddition of cyclopentadiene and 1,4-benzoquinone.

The delay between the invention of the reaction and its use in total synthesis can probably be attributed to a limited understanding of the reaction scope104. Rules which dictate the outcome of DA, such as the Alder endo rule, were only solidified six years after its first report105. For the development of their eponymous reaction, Diels and Alder shared the 1950 Nobel Prize in

Chemistry. The ability to generate complex unsaturated cycles with excellent stereo- and regio-

Chapter 3 control under mild conditions would cement the DA as one of the most ubiquitous in synthesis, particularly in the total synthesis of complex molecules.

3.1.2 Mechanism, tolerance, selectivity, and variations of the Diels-Alder reaction

In all cases, a diene and a dienophile associate through a single transition state having a smaller volume than either the sum of the starting materials or the product itself106. The choice of diene is relatively unrestricted and can be either an open chain or a cyclic compound with many decorating functionalities tolerated. One critical limitation is the ability of the diene to exist in the s-cis conformation104. Large substituents can also introduce steric hindrance, slowing the reaction rate. Diene stability is not a strict necessity as unstable dienes can be generated in situ and gradually reacted with the already present dienophile104.

Typically, the dienophile will feature an electron withdrawing group conjugated to a C-C

π-bond. It is not strictly necessary that the dienophile contain an olefin functionality as dienophiles which contain heteroatoms, typically S, O, and N double bonded to C, can produce the corresponding heterocycle products, as exemplified in the aza107 and oxo108 DA (Figure 3.3).

Dienophiles that contain temporarily „masked functionalities‟ can be employed to access DA products containing functionalities that would otherwise reduce or abolish the reactivity of the dienophile109. Once the product is formed, the masking group can be removed or otherwise modified to access the desired functionality.

Chapter 3

Figure 3.3: The oxo and aza variations of the Diels-Alder reaction.

A variety of isomers can result from DA but simple rules (dictated by frontier molecular orbital theory) can reliably predict the major products. In the most common variation of the reaction, the diene contains an electron donating group and the dieneophile an electron withdrawing group, encouraging a strong interaction between the highest occupied molecular orbital (HOMO) of the diene and the lowest unoccupied molecular orbital (LUMO) of the dienophile110. This interaction dictates which carbons will associate and therefore directs the orientation of the starting materials in the transition state, controlling the final structure of the major product. A simplification of this model predicts that the diene carbon with the greatest partial positive charge will bond to the carbon of the dienophile with the greatest partial negative charge. This rule is reversed in the inverse-electron-demand variation of the reaction (Figure

3.4); a variation in which the diene contains the electron-withdrawing group and the dienophile contains the electron-donating group, resulting in the LUMO of the diene overlapping with the

HOMO of the dienophile111.

Chapter 3

Figure 3.4: FMOT interpretations of variations on Diels-Alder reactions.

In the years following the initial work of Diels and Alder, several rules were solidified to predict the regio- and stereo-chemical outcome of the reaction. The cis principle originated from work of Alder and Stein in 1937 and states that the stereochemistry of starting materials will be retained in the DA product112. Thus, a cis-dienophile will result in both substituents projecting from the same face of the product ring and vice versa for the trans-dienophile. The same principle holds true in the case of the diene. The endo addition rule (Figure 3.5) states that the most stable transition state will arise from whichever orientation of substrates allows maximum overlap of pi orbitals; that is to say the orientation in which the orbitals of the diene have the most ideal symmetry for bonding with their aligned counterparts in the dieneophile. Such an orientation most commonly results in the formation of endo ring systems. Although this rule is sometimes violated, it applies most reliably to cyclic dienes and dienophiles. In addition to these rules, the stereochemistry of resulting DA products can also be controlled through a variety of asymmetric reaction conditions including the use of catalysts such as the MacMillan catalyst113 and the use of chiral auxiliaries114.

Chapter 3

Figure 3.5: A simplified example of the Alder Endo rule.

It should be noted that DAs are reversible and retro-DAs can occur, resulting in the diene and dienophile being reformed from an unsaturated, 6-membered ring, usually in the presence of heat115.

Several operational variations have been shown to increase the reaction rate; perhaps the best known example being the acceleration of the reaction observed when using water as the solvent, an effect which is believed to arise from the increase of the the local relative concentration of diene and dienophile by the hydrophobic effect116. Furthermore, the addition of a Lewis acid is well established to speed the reaction rate via activation of the diene117.

3.1.3 Enzyme-catalyzed Diels-Alder reactions

The prospect that synthetic chemistry‟s best reactions were long ago harnessed by enzymes is extremely attractive and potentially valuable if the activity of such enzymes could be exploited. Perhaps, nowhere is this truer than with the DA. The first purified enzyme with DAase activity was discovered by Auclair and Vederas in 2000118. The authors concluded that lovastatin 53

Chapter 3 nonaketide synthase catalyzed an intramolecular cyclization based upon results with a model reaction where the product produced in the presence of the enzyme is different from the major product of the non-catalyzed or chemically-catalyzed reaction (Figure 3.6). The investigators proposed that lovastatin nonaketide synthase catalyzes DA by binding to substeates with the correct proximity and orientation necessary to promote cyclization. Several other intramolecular

DA-catalyzing enzymes were reported119, including solanapyrone synthase120 and macrophomate synthase121, before the first successful attempt to design (de novo) an intermolecular DA catalyzing enzyme100. This enzyme features a small, mostly hydrophobic active site with a single hydrogen bond donor and acceptor and functionalities known to stabilize the partial positive and negative charges that accumulate in the DA transition state116. Using this enzyme, yields of over

80% were achieved and a single stereoisomer (among the 8 possible stereo- and structural isomers) accounted for 97% of the yield; a 13 fold improvement over the uncatalyzed reaction.

This study has laid the foundation for the engineering of new DA-catalyzing enzymes and suggests possible features to look for in the search for DAase activity among other naturally occurring enzymes.

Chapter 3

Figure 3.6: The proposed Diels-Alderase-catalyzed cycloaddition towards the biosynthesis of lovastatin122.

3.2 CYP2E1 as a potential Diels-Alderase

Althogh there are enzymes known to catalyze DA, these enzymes tend to be highly selective for single substrates and are also known to catalyze many reactions other than DA.118

For these reasons, new DA-catlayzing enzymes are highly desirable. CYP2E1 may be an ideal candidate as a potential DAase. In addition to having a relatively compact active site (Figure

3.7), a feature common among natural and engineered DAases, CYP2E1 is known to accommodate a variety of small ring-containing substrates that resemble DA products. Perhaps even more promising is the fact that the active site features a majority of hydrophobic residues (a 55

Chapter 3 variety of valine, leucine, isoleucine and phenylalanine residues) and also several hydrogen bond donors and acceptors (in the form of threonine/arginine residues and backbone amides) to activate the electron withdrawing and donating groups of potential DA precursors. It seems reasonable that the iron metal at the heart of the active site may serve to replace the Lewis acid component so often used to accelerate DAs. Finally, enzymatic catalysis is typically performed in water or buffer, conditions known to increase the rate of DA reactions116 and to favor the binding of hydrophobic substrates in the active site of CYP2E1.

Figure 3.7: Binding site of CYP2E1 bound to 10-(1H-imidazol-1-yl)decanoic acid (from crystal

structure2).

3.3 Substrate selection

A series of typical dienophiles and dienes were selected as potential test substrates on the basis of size, solubility in water, and commercial availability. Only substrates which have been

Chapter 3 previously shown to spontaneously cyclize into easily elucidated DA products were considered.

Ten dienophiles and five dienes were chosen based upon these criteria, resulting in a list of fifty potential DA products (Table 1.1).

Table 3.1: Selected dienophiles and dienes and several examples of possible DA products.

As previously stated, transition state stabilization is a determining factor in the rate of

DA, and DA products closely resemble the transition states by which they were formed; therefore, it stands to reason that the more tightly a DA product is bound by the active site of

CYP2E1, the more likely it is that formation of the corresponding product will be accelerated by the presence of the enzyme. With this logic in mind, the DA products in Table 1 were docked in 57

Chapter 3 silico to the active site of CYP2E1 using Maestro, a popular docking and scoring program, and ranked in order of binding score, a prediction of how tightly each molecule binds. These compounds produced binding scores between -10 and -15, similar values to those produced for natural substrates of the enzyme such as 3-(1-Benzyl-1H-imidazol-4-yl)-pyridine and 3-(1- methyl-1H-imidazol-4-yl)pyridine. Table 2 highlights the six DA products which were predicted to bind most tightly to the active site of CYP2E1. The results of this study predict a clear preference for several substrates and dictated which DAs would be investigated.

Table 3.2: Relative binding strength of DA products in the active site of CYP2E1 as predicted using the Maestro software suite. Ligand structures were prepared using LigPrep, the protein (pdb

Chapter 3

3E4E) was prepared from using the Protein Preperation Wizard, and the docking was performed using Glide.

Figure 3.7 illustrates the predicted binding mode of a DA product bound to the active site of CYP2E1. In this representation, the DA product binds closely to the reactive iron of the porphyrin ring and is almost equidistant to the 4 phenylalanine residues that flank the active site.

Furthermore, the ligand’s hydrogen bond accepting functionality is oriented towards the highest concentration of hydrogen bonding side-chains; supporting the proposed mechanism of binding.

It should be noted that this binding position is similar to that of the unsaturated cyclic moiety of the substrate in the crystal structure illustrated in figure 3.8. The remaining proposed DA products were all predicted to bind in a similar position and ordination.

Chapter 3

Figure 3.8: Maestro-predicted binding mode of bicyclo[2.2.1]hept-5-ene-2-carboxylic acid in the active site of CYP2E1.

3.3.1 Dienes

In addition to being synthons of the highest scoring DA products from the docking studies, cyclopentadiene, 3.1, cyclohexadiene, 3.2, and 2,3-dimethyl-1,3-butadiene, 3.3, each have an almost ubiquitous presence in experiments seeking to probe new DA methodologies.

Furthermore, these compounds represent both open- and closed-chain dienes, affording structural

Chapter 3 variety in the resulting products and the potential to probe the endo/exo selectivity of CYP2E1 as a DAase.

Figure 3.9: Selection of representative dienes for testing the Diels-Alderase activity of CYP2E1.

3.3.2 Dienophiles

Only two dienophiles were represented in the top 6 highest scoring DA products from the docking studies, N-methyl maleimide, 3.4, and but-3-ene-2-one, 3.5. Both compounds are commercially available and well studied as DA substrates123.

Figure 3.10: Selection of representative dienophiles for testing the Diels-Alderase activity of

CYP2E1.

3.4 Assays, results, and discussion

3.4.1 Synthesis of authentic standards of Diels-Alder products

Chapter 3

To confirm the identity of DA products resulting from the reaction of each diene and dienophile, authentic standards were synthesized by Amelie Menard, a graduate student in the

Auclair group, using AlCl3 as the catalyst prior to GC-MS analysis.

3.4.2 GC-MS assay

GC-MS was selected as the ideal instrument to determine the identity (by mass/charge ratio) and the relative yields (by reference to an internal standard, acetophenone) of products formed during the DA investigated. CYP2E1 was mixed in buffer with selected dienes and dienophiles and left to react before the addition of internal standard and quenching by organic extraction. The uncatalyzed rates of two DAs were evaluated (Figure 3.11) at substrate concentrations ranging from 0.1 mM to 10 mM to determine which concentration would afford product formation at a convenient level of detection. Although product formation was observable at all substrate concentrations tested, 1 mM was selected as the best experimental concentration in order to minimize the amount of enzyme used while maintaining a relatively high level of detection. As GC-MS is primarily a „snap shot‟ of reaction progress, reactions were quenched and sampled at both 2 hours and 24 hours after initiation; time points chosen based upon the previously mentioned uncatalyzed studies.

Chapter 3

Figure 3.11: DAs used to probe CYP2E1 for DAase activity.

3.4.3 Results of GC-MS Assay

In all cases, the addition of CYP2E1to the DA decreased the resulting yield at each time point (Figure 3.12). This result is surprising as it implies that the addition of CYP2E1 not only fails to accelerate the DA reaction but also somehow hinders the capability of the dienes and dienophiles to attain the proper position, orientation, or electronic environment necessary for reaction.

Chapter 3

Figure 3.12: The effect of the addition of CYP2E1 to the rates of DA cycloaddition reactions of representative dienes and dienophiles. Reactions were performed in buffer (potassium phosphate, 0.1 M, pH 7.4) using 1 mM concentrations of all substrates and 50 nM of CYP2E1.

The observed decrease in yield may result from the substrates binding in the active site in positions or orientations that don’t favor reaction, slowing the overall reaction rate. Other explanations include the possibility that the substrates may bind to areas outside the active site and would therefore be sequestered into positions or electronic environments unfavorable for reaction. These hypotheses, if correct, would explain the resulting decrease in yield and may inform the design of DAs that are more likely to be catalyzed by CYP2E1. For example, Figure 6 64

Chapter 3 illustrates that the enzyme binding pocket is flanked by 4 phenylalanine residues which likely interact with many natural substrates via π-stacking interactions. Thus, it seems likely that dienes and dienophiles decorated with aromatic groups (such as benzyl and benzoyl moieties) may bind more tightly to the active site and orient themselves in a way that is more inductive to DA.

3.4.4 UV assay

Although GC-MS can be used to determine product identity and yield at an „end point‟, it is generally unable to monitor reaction progress in real-time. In contrast, UV spectroscopy is ideal for this purpose and can provide information about the initial reaction rate, which is of special importance to investigations concerned with catalysis. As 3.4 contains a chromophore with a convenient λ max at 300 nm, UV spectroscopy was selected to measure the rate at which it is consumed while reacting with 3.1 to form the corresponding DA product, 3.6 (Figure 3.13).

3.4.5 Results of UV Assay

The initial reaction rate upon mixing 3.1 and 3.4 in the presence and absence of CYP2E1 was measured by tracking the consumption of 3.4 using the experimentally derived molar extinction coefficient of 620 M-1cm-1.

Again, the reaction rate was significantly retarded by the addition of CYP2E1 (Figure

11), further validating the results of the GC-MS assay (Figure 10). This result supports the conclusion that the substrates are unproductively attracted to the active site or other areas of the enzyme in such a way that does not facilitate DA reactivity.

Chapter 3

Figure 3.13: The effect of CYP2E1 on the rate of consumption of compound 3.4. Reactions were performed in buffer (potassium phosphate, 0.1 M, pH 7.4) using 1 mM concentrations of all substrates and 50 nM of CYP2E1. Reactions were performed in duplicate.

3.5 Conclusions

The results presented here demonstrate that the expected DA products were generated both in the presence and in the absence of CYP2E1. In each experiment, the rates and yields of the DA tested appear to be hindered by the addition of CYP2E1. This surprising result suggests that there is an interaction between the enzyme and one or more of the substrates but that this interaction serves to hinder the reaction by failing to position, orient, or activate the substrates in such a way to encourage product formation. Importantly, the identical retention times of the

Chapter 3 products and standards suggest that the presence of CYP2E1 resulted in no change in the product identity (including the exo/endo selectivity) compared to catalysis by AlCl3 or the uncatalyzed reaction.

Although DAase activity was not observed in this system, it is encouraging that the addition of CYP2E1 had any effect, even a negative one, on the rate of DA reactions. The implication of this result is clear: there is an interaction between the enzyme and one or more of the substrates. This interaction may take place in the active site or at another binding pocket.

Further docking studies or NMR experiments may help determine the positions and orientations with which the substrates bind and may lead to rational design solutions (such as modifications to the enzyme or substrates) which could facilitate DAase activity.

Chapter 4

Chapter 4:

The Search for a CYP3A4-Directing Chemical Auxiliary

Chapter 4

4.0 Preface

A chemical auxiliary is a compound which, when bound to a substrate, helps orient the substrate during a chemical reaction (often by restricting access to some conformations) for the purpose of predictably directing chemo-, regio- or stereoselectivity. Recognizing the need for improved P450 product predictability, it was envisaged that an auxiliary capable of directing the oxidative power of CYP3A4 would be highly useful and could potentially effect a paradigm shift in the way that chemists approach difficult oxidations.

The experimental work described in this chapter was performed by me, with some help from Julian Ferras, a visiting M.Sc. student. Specifically, compounds 4.3-4.12 were synthesized and characterized by Julian Ferras. I synthesized the remaining compounds and performed and analyzed all the enzyme assays described here.

4.1 Introduction

4.1.1 Chemical auxiliaries used in non-enzymatic reactions as inspiration

Typically, chemical auxiliaries are chiral molecules employed to improve the stereoselective outcome of a given reaction. The Evans‟ auxiliaries are classic examples of chemical auxiliairies and consist of oxazolidinones with a chiral carbon at C-5 (Figure 4.1)124.

The Evans‟ auxiliaries control the stereoselectivity of aldol reactions by restricting access to one side of the enolate, resulting in asymmetric reactivity. In essence, the chemical auxiliary directs the stereoselectivity by converting the normally pro-enantiomeric reaction site of the substrate into a pro-diastereomeric site.

Chapter 4

Figure 4.1: Scheme demonstrating the use of an Evans’ auxiliary. The auxiliary is attached to an activated carbonyl, directs the sterochemical outcome of the reaction, and is then removed.

The Evans‟ auxiliaries have been successful mainly because of their low cost, ease of attachment and removal. Moreover, they also address an important need by providing selective aldol reaction products125.

Other examples of chiral chemical auxiliaries include ephedrine derivatives126, a diverse range of sulfonamides127, and a series of thioglycolate lactams developed at McGill by the

Gleason Lab128 (Figure 4.2). It should be noted that while auxiliaries are typically covalently bound to the substrate, this is not always the case and there exist auxiliaries which direct

Chapter 4 reactions through non-covalent interactions alone such as in the case of the auxiliary-directed, enantioselective synthesis of propargylic amines129.

Figure 4.2: Examples of commercially available chemical auxiliaries. An ephedrine derivative- based auxiliary, a sulfonamide-containing auxiliary, and a Gleason auxiliary, respectively.

4.1.2 General strategy for the design of a chemical auxiliary to control the selectivity of enzymes

Although there are many examples of chemical auxiliaries which assist in chemical reactions124, 126-127, to our knowledge, at the time this research was initiated, there were no examples of auxiliaries capable of controlling the regio- and stereoselectivity of enzyme- catalyzed reactions. This may be attributed to both the difficulty inherent in their design and to the fact that most enzyme-catalyzed reactions are already relatively selective by nature. Enzyme- directing auxiliaries should, however, find utility with enzymatic reactions which are sought attractive for their potential synthetic utility but are hindered poor predictability, such as in the case with CYP3A4. Ideally, an enzyme-directing auxiliary should be i) inexpensive; ii) easy to functionalize with a large range of substrates; iii) capable of significantly improving the selectivity of a specific chemical reaction; and iv) easily removed from the substrate upon completion of the reaction. Other important, practical considerations include spectral properties

Chapter 4 of the auxiliary (auxiliaries which have chromophores will enable easier recovery and tracking of reaction progress) and the ability of the auxiliary to assist in transport of a substrate across biological membranes, an important concern when considering whole-cell fermentations.

Several groups have reported the use of substrate engineering in biocatalysis, but mostly to increase recognition of non-natural substrates and not for rational control of selectivity130. To our knowledge, the only example with P450 enzymes used a carbolide to „anchor‟ macrolides into P450 PikC131. This method did not however control the site of oxidation and generated multiple products without predictability. Instead, the auxiliary desired here must anchor the substrate and also allow predictable control of the reaction chemo-, regio-, and stereoselectivity.

For the purpose of directing CYP3A4-catalyzed oxidations, an auxiliary should fit in a binding pocket near the enzyme’s heme group and display one specific C-H bond of the substrate towards the reactive iron species (Figure 4.3). Like a ruler, the distance between the auxiliary and the iron would control the regioselectivty of hydroxylation. Facial selectivity will arise from the orientation of the auxiliary relative to the heme prosthetic group.

Chapter 4

Figure 4.3: An ideal auxiliary binding to the enzyme and projecting a single C-H bond towards the reactive heme iron with facial selectivity.

As with chemical auxiliaries, the auxiliary should be easy to attach and remove from the substrate. Furthermore, the auxiliary should be larger and more highly functionalized than the substrate, ensuring that the auxiliary-enzyme interactions dominate the control of the binding orientation.

As previously stated, predicting the site of CYP3A4-catalyzed oxidation is difficult. To mediate this problem, natural CYP3A4 substrates with known reactivity were used here as inspiration for the design of CYP3A4-directing auxiliaries. With this model in mind, we selected several natural substrates of CYP3A4 which are oxidized on a part of the molecule remote from the core of the substrate, with the aim of using the core of these substrates as potential auxiliaries. Ideally, the auxiliary would be commercially available and easily functionalized.

Chapter 4

4.2 4-(Trifluoromethyl)coumarin as a potential CYP3A4-directing chemical auxiliary

7-Benzyl-4-(trifluoromethyl)coumarin (4.1) is a known substrate of many human P450 isoforms including CYP3A4, CYP1A2, CYP3A, and others132. CYP3A4 hydroxylates compound

4.1 at the benzyl methylene, forming a hemi-acetal which quickly cleaves to yield the fluorescent product 7-hydroxy-4-(trifluoromethyl)coumarin and benzaldehyde (4.2) (Figure 4.4).

Figure 4.4: Oxidative cleavage of compound 4.1 by CYP3A4.

Because the corresponding ethoxy derivative is also modified by CYP3A4 in a similar way, it was envisaged that the 4-(trifluoromethyl)coumarin ring of compound 4.1 might be responsible for binding and orienting the molecule in the active site of CYP3A4 and, therefore, oxidations might occur on most substrates attached to the 4-(trifluoromethyl)coumarin ring in place of the benzyl group. Thus, 4.2 could serve as a chemical auxiliary. However, an ether functionality is not desirable for easy attachment/removal of an auxiliary. To overcome this limitation, substrates could be attached to 4.2 via an ester linkage.

Chapter 4

4.2.1 Synthesis of the 4-(trifluoromethyl)coumarin-based auxiliary-substrates

Nucleophilic substitution chemistry (Scheme 4.1) was used by Julian Ferras to synthesize the 4.2-based auxiliary-substrates shown in figure 4.5.

Scheme 4.1: Synthesis of 4-(trifluoromethyl)coumarin based substrate-auxiliary complex.

Figure 4.5: A series of auxiliaries-substrates featuring 4.2 as the auxiliary. The moiety in green is the auxiliary and the moiety in blue is the substrate.

Chapter 4

These compounds were combined with CYP3A4 under catalytic conditions and tested for oxidation by myself.

4.2.2 Assay

Each auxiliary-substrate, 4.3-4.12, was combined with CYP3A4 in the presence of the cofactor surrogate cumene hydroperoxide (CHP) in potassium phosphate buffer and left to react for 1.5 hours before termination by organic extraction. The extent of oxidation was quantified by

RP-HPLC with UV detection (using a diode array monitoring 200-400 nm). Any significant peaks (>5% of the substrate peak) which could correspond to oxidized products were collected and characterized by ESI-MS.

4.2.3 Results and discussion

Each auxiliary-substrate was easily resolved by RP-HPLC but no oxidized products were detected for any of the compounds. It was therefore concluded that compounds 4.3-4.12 are not substrates of CYP3A4.

This result is not all together unexpected as the known site of oxidation of the substrate,

4.1, is the methylene of the benzyl group and in the auxiliary-substrate, this position is fully oxidized. It was hoped that CYP3A4, being a highly promiscuous enzyme, would instead oxidize the next position along the chain, however, this has not been observed here.

Chapter 4

4.3 Fluorescein as a potential CYP3A4-directing chemical auxiliary

Fluorescein benzyl ether (compound 4.13) is a known substrate of many human P450 isoforms including CYP3A4 and CYP2C19133. As with 4.1, 4.13 is hydroxylated at the benzyl methylene by CYP3A4 to yield a hemi-acetal which quickly cleaves to yield a fluorescent product (compound 4.14).

Figure 4.6: Oxidation of compound 4.13 by CYP3A4.

It was envisaged that the fluorescein ring of 4.13 is primarily responsible for binding and orienting the molecule in the active site of CYP3A4 and, therefore, oxidations may occur on substrates attached to the fluorescein ring in place of one or both of the benzyl groups. However, here again the ether functionality attaching the fluorescein ring to the benzyl groups is not ideal for reversible functionalization with other substrates. With this in mind, it was envisaged to link substrates to the fluorescein group via an ester linkage.

Chapter 4

4.3.1 Synthesis of fluorescein-based auxiliary-substrates

Fluorescein-based auxiliary-substrates were synthesized using nucleophilic substitution chemistry (Scheme 4.2).

Scheme 4.2: Synthesis of fluorescein-based auxiliary-substrate complex.

Using this chemistry, a single auxiliary-substrate complex was synthesized (Figure 4.7).

Figure 4.7: An auxiliary-substrate complexes featuring fluorescein as the auxiliary.

Because long, flexible substrates can potentially sample a larger portion of the active site, palmitic acid was chosen as the single substrate to be tested. Compound 4.15 was reacted with

CYP3A4 as described in section 4.2.3.

Chapter 4

4.3.3 Results and discussion

The auxiliary-substrate, 4.15, was easily resolved by RP-HPLC but again, no oxidation products were detected. This result suggests that compound 4.15 is not a substrate of CYP3A4.

The natural site of oxidation of compound 4.13 by CYP3A4 is the methylene of the benzyl group and again, this position is fully oxidized as the ester in compound 4.15. This result, along with the results for the previous auxiliary candidate, 4.2, suggest that a successful auxiliary might be better modeled after a substrate which is oxidized more distally from the site of substrate attachment.

4.4 Theophylline as a CYP3A4-directing chemical auxiliary

Theophylline (4.16), lisofylline (4.17), and theobromine (5.1, to be discussed more in the next chapter) each contain a xanthine ring. The side chain of lisofylline is hydroxylated by

CYP3A4 and one expects that a different side chain (now the substrate) added to either theophylline or theobromine might also get hydroxylated (Figure 4.8).

Chapter 4

Figure 4.8: Structural similarities of theophylline and theobromine with lisofylline, a natural substrate of CYP3A4. One may expect that the side chain substrate of the derivatives would be hydroxylated by CYP3A4, much like the side chain of lisofylline.

Theophylline is easily modified at N-7 and/or N-9 and may therefore be a useful auxiliary. Assuming that xanthine-containing molecules bind to CYP3A4 in a similar position and orientation as lisofylline, a substrate attached to theobromine would project in the same position as the side-chain of lisofylline yet in a different orientation for a substrate linked to theophylline. This has the potential to afford different products from similar auxiliaries. This system would have the added advantage that xanthine-containing molecules have high extinction coefficients and are known to assist in the crossing of biological membranes134, an important factor in the consideration of whole-cell fermentation techniques common to biocatalysis. The use of theobromine as a chemical auxiliary for CYP3A4 will be discussed in the next chapter.

Chapter 4

4.4.1 Synthesis of a theophylline-based auxiliary-substrate

A theophylline-based auxiliary-substrate was synthesized from theophylline using nucleophilic substitution on a chloride under basic conditions. After washing with water to remove unreacted theophylline, pure compound 4.18 was recovered with an 80% yield. A

NOESY 1D spectra was taken to confirm that functionalization was limited to the nitrogen at the

9-position of the ring (Scheme 4.3).

Scheme 4.3: The synthesis of a theophylline-based auxiliary-substrate.

This auxiliary-substrate, 4.18, was reacted with CYP3A4 as described in section 4.2.3.

4.2.2 Results and discussion

Following organic extraction of the enzymatic reaction mixture with 4.18 and RP-HPLC separation, 2 small peaks were observed at lower retention times than the substrate (Figure 4.9).

The mass-to-charge ratio of each new peak corresponded to a gain of 16 mass units compared to the unoxidized substrate, a change which corresponds exactly to the increase expected from monohydroxylation of compound 4.18. The areas of each peak indicated conversion yields of

Chapter 4 approximately 5%. The different orientations of the substrate portion of this substrate-auxiliary complex compared to the aliphatic chain of lisofylline may partially explain these low yields.

Figure 4.9: HPLC trace of the reaction mixture for CYP3A4-catalyzed transformation of 4.18 with CHP as a cofactor surrogate. The mass-to-charge ratio of each product corresponds to a gain of 16 mass units.

A peak corresponding to the mass of the unchanged theophylline ring and a peak corresponding to a loss of water (-18 mass units) were both simultaneously observed in the mass spectra of the two new peaks, suggesting the sites of oxidation are on the substrate and not on the theophylline auxiliary moiety. The low yields of product precluded structural elucidation by

NMR.

This result is significant; it represents the successful design of a CYP3A4-directing auxiliary which was logically designed with inspiration from a natural substrate. However, as the

Chapter 4 reaction generates 2 different products (hence low regioselectivty) and the yields of both oxidized products were low, it seems unlikely that theophylline will meet the requirements of an ideal CYP3A4-directing auxiliary. Furthermore, the challenge of removing the oxidized substrates from the theophylline ring is non-trivial. Ultimately, exploring this auxiliary was discontinued in favor of a potentially more valuable CYP3A4-directing auxiliary, theobromine, which is discussed in the next chapter.

4.5 Phthalamide-protected amines as interesting auxiliary-substrates for directing CYP3A4 reactivity

Phthalamide-protected amines, whether resulting from the nucleophilic attack of an amine on phthalic anhydride under acidic conditions or from functionalization of phthalamide at the nucleophilic amine, share some structural similarities with lisofylline, a known substrate of

CYP3A4 that served as inspiration in the design of the theophylline-based auxiliary (Figure

4.10).

Chapter 4

Figure 4.10: Structural similarities of phthalamide-based auxiliary-substrates and lisofylline, a known natural substrate of CYP3A4.

More importantly, this system would benefit from the ease with which phthalamides are functionalized to and removed from substrates. While theophylline (4.16) can only easily be functionalized with electrophiles, phthalamide derivatives can be generated from either electrophiles or nucleophiles, further increasing utility135.

4.5.1 Synthesis of a phthalamide-based auxiliary-substrate

A phthalamide-based substrate-auxiliary was synthesized from phthalic anhydride using nucleophilic substitution under acidic conditions (Scheme 4.4), giving a 70% isolated yield of

4.21 after purification. In this reaction, the phthalic anhydride is first combined with the amine, generating an amide and a carboxylate. Acid is then added to promote the loss of H20 during intramolecular attack by the resulting amide nitrogen.

Chapter 4

Scheme 4.4: The synthesis of a phthalamide-containing auxiliary-substrate.

4.5.2 Enzymatic reaction of 4.21 with CYP3A4

Compound 4.21 was combined with CYP3A4 under catalytic conditions. The assay used to detect product formation was performed as in section 4.2.3. A second, larger scale reaction was performed with 2 mg of compound 4.21, with the concentrations of the other components unchanged from section 4.2.3.

4.5.3 Results and discussion

Upon RP-HPLC separation of the reaction mixture resulting from the combination of compound 4.21, CYP3A4, and CHP in buffer, a large peak was observed at a lower RT than the auxiliary-substrate (Figure 4.11). The mass-to-charge ratio of this new peak corresponds to a gain of 16 mass units compared to the unoxidized auxiliary-substrate, a gain suggesting monohydroxylation of 4.21. The area of this peak indicated a conversion yield of 75% and, with the exception of the peaks corresponding to the auxiliary-product and auxiliary-substrate, no other peaks corresponding to oxidized products were observed.

Chapter 4

Figure 4.11: HPLC trace for the reaction mixture of compound 4.21 with CYP3A4. The auxiliary- product, 4.22, appears at a lower retention time and the mass-to-charge ratio of this new peak corresponds to a gain of 16 mass units compared to that of the auxiliary-substrate.

The auxiliary-product 4.22 was isolated and re-injected on a chiral HPLC column and a single sharp peak was observed. It is possible that enantiomers of 4.22 would not separate on this column. The production of a diastereosomer by derivatization of the alcohol with a second chiral molecule, such as a Mosher's acid136, would allow a more certain determination of the enantiomeric ratio (er). The limited amount of auxiliary-product however, (approximately 1 mg) impeded the practicality of this approach.

Chapter 4

The structure of 4.22 was solved by 1H NMR (see section 8.7), indicating that CYP3A4 catalyzed hydroxylation at the 4th carbon (the C-4 position) from the amide of the phthalamide auxiliary (Figure 4.12).

Figure 4.12: The CYP3A4-catalyzed hydroxylation of compound 4.21 results in hydroxylation at C- 4.

Interestingly, the CYP3A4-catalyzed hydroxylation of 4.21 takes place at the same distance from the phthalate auxiliary as the CYP3A4-catalyzed hydroxylation of lisofylline, 4.17, if one were to consider theobromine as the auxiliary (Figure 4.13). This suggests that theobromine might serve as an effective auxiliary, a possibility explored further in the next chapter.

Chapter 4

Figure 4.13: The sites of CYP3A4-catalyzed hydroxylation on compound 4.21 and lisofylline, two structurally similar molecules.

As stated previously, phthalamide, 4.21, is a well characterized protecting group and, as such, is designed to be easily removed, giving the free amine. Typically, phthalamide-type protecting groups are removed by substitution with nucleophiles such as hydrazine (and derivatives), however other methods using reducing agents have also been reported135.

This result is significant; it represents the successful design of a new CYP3A4-directing auxiliary which was logically designed with inspiration from a known substrate. However, as only one substrate using this auxiliary was tested, it remains unknown whether the phthalate group will lead to C-4 selectivity with other substrates.

An advantage of this auxiliary is that phthalamide-protected amines can be generated from both electrophiles (using phthalmide) or nucleophiles (using phthalic anhydride). This

Chapter 4 flexibility may greatly increase the scope of substrates compatible with the strategy.

Furthermore, there exist multiple methods to easily cleave/remove phthalamide groups, enabling greater tolerance to easily reducible or otherwise fragile functionalities.

4.5 Conclusions

The results of the first 4 studies discussed here illustrate the difficulty associated with predicting the catalytic behavior of CYP3A4. The auxiliary-substrate selected, to one extent or another, resembled known substrates of CYP3A4; yet, except for theophylline- and phthalamide- based auxiliary-substrates, none were transformed by the enzyme under catalytic conditions.

Important lessons can be learned from these studies for the design of successful CYP3A4- directing auxiliaries. Firstly, the structures of successful auxiliaries should deviate as little as possible from those of natural substrates of CYP3A4; this will ensure the best chance of binding to the enzyme in the same mode. Secondly, the expected site of oxidation on the auxiliary- substrate, estimated from the site of oxidation on the natural substrate, should be distal (several carbons) from the core of the molecule, ie. the proposed auxiliary. Furthermore, this site should not be fully oxidized because, as we have seen here with compounds 4.3-4.12, and 4.15, the promiscuity of the enzyme may not be depended on to extend beyond the most favorable sites of oxidation.

The relative success of the theophylline-based auxiliary, 4.18, can be explained in the context of the following lessons. In contrast to the 4-(trifluoromethyl)coumarin and fluorescein auxiliaries, the expected site of oxidation of the theophylline complex was chemically, sterically, and electronically similar to that of he natural CYP3A4 substreate lisofylline. Unfortunately, the theophylline auxiliary did not afford the yields of oxidized products desired; however, the 89

Chapter 4 lessons learned here were instrumental to the design of the next generation of CYP3A4-directing auxiliaries.

Chapter 5

Chapter 5:

Theobromine as a CYP3A4-directing auxiliary

Chapter 5

5.0 Preface

In the previous chapter, the design, synthesis and evaluation of several potential

CYP3A4-directing auxiliaries was discussed. The lessons learned from these ‘first generation’ attempts were applied to the evaluation of another auxiliary, theobromine, compound 5.1 (Figure

5.1). In this chapter, we will see that theobromine can be used to direct CYP3A4 oxidations with predictable regio-, chemo-, and stereoselectivity.

Figure 5.1: The structure of theobromine.

I performed the majority of the experimental work described in this chapter except for a section which was performed by Erin May, a previous undergraduate student in the Auclair lab.

Specifically, Erin May assisted with the synthesis of 5.50-5.57. I synthesized and characterized the remaining molecules and performed all of the biological and structural elucidation studies.

Most of this work was published as Larsen, A. T.; May, E. M.; Auclair, K., Journal of the

American Chemical Society 2011, 133 (20), 7853-8. The rest will be submitted for publication in the near future.

5.1 Introduction 92

Chapter 5

Theobromine, or 5.1, named for its association with chocolate (the name translates from

Greek into ‘food of the gods’), is an inexpensive, commercially available, achiral, and easily functionalized compound. As mentioned in an earlier chapter, we were inspired to use 5.1 as an auxiliary from lisofylline, a natural CYP3A4 substrate hydroxylated at the side chain. Due to the nucleophilic N-1 of its xanthine ring, theobromine can be easily derivatized into analogues of lisofylline (Figure 5.2), allowing access to a wide scope of theobromine-based substrate- auxiliary complexes.

Figure 5.2: The derivatization of theobromine into auxiliary-substrate complexes which resemble lisofylline.

There are additional benefits to using theobromine as an auxiliary including a high extinction coefficient137, which can assist in the recovery of oxidized products and unoxidized starting material, and the ability to cross biological membranes134, an important factor when considering whole-cell fermentations common to biocatalysis. Furthermore, the expected site of oxidation is sufficiently distal from the auxiliary to avoid the issues encountered previously

(Figure 5.3).

Chapter 5

Figure 5.3: The expected site of CYP3A4 oxidation on a theobromine-based auxiliary-substrate complexe compared to lisofylline.

Finally, cleavage of the auxiliary should be achieved using conditions similar to those used for phthalamide deprotection135, affording the free amine (Figure 5.4).

Figure 5.4: Removal of the oxidized substrate from the theobromine auxiliary using similar conditions used to remove phthalamide protecting groups from nitrogens135.

With these considerations in mind, a series of theobromine-based complexes were synthesized and tested.

5.2 Synthesis of theobromine-based auxiliary-substrates

Chapter 5

A series of theobromine-based auxiliary-substrates were synthesized to evaluate theobromine as a CYP3A4-directing auxiliary (Table 5.1).

Table 5.1: The scope of auxiliary-substrates designed and synthesized to evaluate the CYP3A4- directing capability of theobromine.

Chapter 5

The majority of these substrates were directly synthesized by functionalizing the theobromine N-1 using various alkyl halides or tosylates under basic conditions (Scheme 5.1).

Scheme 5.1: The general protocols for the synthesis of theobromine-substrates. A) Nucleophilic substitution of halides, where X = Cl or Br; B) Nucleophilic substitution of tosylated alcohols. Theobromine is shown in green and the substrates in blue.

In each case, functionalization of theobromine, 5.1, under basic conditions gave an average isolated yield of approximately 80% and required no purification besides a water wash to remove unreacted theobromine and other side products.

Several of the theobromine substrates containing more complex functionalities could not be directly synthesized using the chemistry illustrated in Scheme 1 and were therefore synthesized using somewhat more intricate routes (Scheme 5.2). For example, in the synthesis of

5.25 first required the coupling of theobromine to a tosylated and protected diol, followed by removal of the protecting group, oxidation of the resulting alcohol to the aldehyde, functionalization by Grignard addition and further oxidation to the ketone. Compounds 5.22,

Chapter 5

5.23, and 5.26 are all intermediates in the synthesis of 5.25. First, 1,3-propanediol was mono- protected and activated before functionalization with theobromine. Subsequent deprotection and oxidation afforded 5.22. Treatment of 5.22 with Dess-martin periodinane afforded 5.23. The functionalization of 5.23 with ethylmagnesium bromine afforded 5.26. Finally, oxidation of resulting alcohol gave access to 5.25. Alternatively, bromination of 5.22 gave access a substrate containing sulfur at C-4 (Scheme 5.2A). Functionalization of theobromine with 5-bromopentyl ethyl carbonate directly provided target 5.31. Its reduction yielded 5.28, which was alkylated to

5.29 (Scheme 5.2B). Thus, this synthetic plan provided access to compounds with electron-poor functionalities at the C-3 and C-5 positions, as in theobromine-substrates 5.25-5.28, and 5.31, and an electron-rich functionality at C-5 in the case of 5.29.

Chapter 5

Scheme 5.2: Synthetic routes for the synthesis of theobromine-substrates 5.25-5.29 and 5.31. Tb is shown in green, and substrates in blue. 98

Chapter 5

5.3 CYP3A4-catalyzed oxidation assays

Each theobromine-substrate (5.3-5.38) was combined with CYP3A4 under catalytic conditions and the products were separated and characterized using LC-UV-MS.

A cofactor surrogate, cumene hydroperoxide (CHP)48, was used in place of the expensive and cumbersome natural cofactors, cytochrome P450 reductase (CPR) and NADPH. This substitution eliminated the need to express and purify CPR, greatly reduced the cost of the assay, and enabled more reproducible results. Additionally, it was verified that product ratios and product regio- and stereochemistries were identical whether using natural cofactors or CHP.

Most enzymatic reactions were performed at two scales; small scale reactions to search for CYP3A4-oxidized products and large scale reactions to generate sufficient amounts of oxidized auxiliary-products for characterization. Small scale reactions were performed using 20

µg of auxiliary-substrate in approximately 100 µL of buffer using purified CYP3A4. Large scale reactions were performed using 2 mg of auxiliary-substrate in 10 mL of buffer using CYP3A4- containing membranes (a partially purified enzyme preparation). Detailed protocols for all experiments appear in chapter 8. Finally, a single 100 mg scale reaction was performed on compound 5.5 using purified enzyme to demonstrate the practical synthetic utility of the method.

Regardless of scale, the assays were performed using the same procedure: the enzyme and auxiliary-substrate were combined in buffer and pre-incubated at 30oC for 5 minutes before the addition of the cofactor, CHP. The enzymatic reaction was >90% complete after approximately 1.5 hours and was terminated by extraction with chloroform. The organic layer

Chapter 5 was evaporated and the residue redissolved in acetonitrile before RP-HPLC separation with detection by UV and ESI-MS.

For each auxiliary-substrate, separate control reactions were ran without enzyme, cofactor, or auxiliary-substrate, to ensure that the observed auxiliary-products were not being generated in any reaction other than a CYP3A4-catalyzed oxidation.

Peaks appearing at a lower retention times than the auxiliary-substrate (possibly indicating an increase in polarity resulting from oxidation) and having mass/charge gains corresponding to oxidations (typically +16 mass units) were investigated as possible CYP3A4- oxidized auxiliary-products. The peaks corresponding to auxiliary-products were collected and reinjected to calculate isolated yields by the area of UV absorbance. Enantiomeric ratios (er) were determined by the reinjection of oxidized products onto a chiral HPLC column. To ensure that yields could be evaluated by UV detection, the λmax and molar absorption coefficients of theobromine-substrates and their auxiliary-products were determined as described below.

5.3.1 Comparison of natural and non-natural cofactors

For selected auxiliary-substrates, assays were performed using the natural cofactors for comparison with the non-natural surrogate, CHP. In each case, whether natural cofactors, CPR and NADPH, or the non-natural cofactor, CHP, were used, identical product identities, yields and enantiomeric ratios were observed. This result demonstrates that the use of CHP did not modify the natural reactivity of the enzyme in terms of activity or selectivity.

100

Chapter 5

5.3.2 UV absorption properties of theobromine derivatives

The molar absorption coefficients (ε) of 5.11 and the corresponding CYP3A4-oxidized alcohol, 5.47, were compared to each other and to caffeine by UV absorption at 273 nm (the wavelength used for HPLC quantification). Solutions ranging in concentration between 4.4 µM and 26 µM were prepared and used to calculate values of ε based on the Beer–Lambert law (all concentrations used were in the linear range). In all cases, the λmax was observed at 273 nm. As shown below, hydroxylation of the substrate portion of the auxiliary-substrate complex did not significantly affect ε (Figure 5.5), nor did it affect the UV-Vis spectrum. Hence the same ε (8.81

E+03 M-1 cm-1) was used in concentration determinations for all theobromine derivatives.

Figure 5.5: The molar absorption coefficient of caffeine, theobromine-substrate 5.11 and the corresponding theobromine-product, 5.47.

101

Chapter 5

5.3.3 Optimization of enzyme activity using purified CYP3A4 and CYP3A4-containing membranes

The concentrations of auxiliary-substrate, enzyme, and cofactor were optimized to maximize the yield of the CYP3A4-catalyzed oxidation of theobromine-containing auxiliary- substrates. The conditions tested included concentrations of CYP3A4 ranging from 2 μM to 24

μM, concentrations of CHP ranging from 150 μM to 1.2 mM, and concentrations of substrate ranging from 100 μM to 2 mM. For both purified CYP3A4 and CYP3A4-containing membranes, the ideal concentration of enzyme was found to be 12 μM. The ideal concentration of both theobromine-substrate and CHP were determined to be 600 μM. Although CYP3A4 is a mammalian protein, the ideal temperature for the reaction was determined to be 30oC. This result may be interpreted in terms of stability: it is well known that CYP3A4, like other mammalian

P450s, is unstable at physiological temperatures in vitro138 and therefore, performing the reaction below 37oC may help improve the operational stability of the enzyme, affording greater product yields.

5.4 Structural determination of CYP3A4-oxidized theobromine-based auxiliary-substrates

The regio- and stereoselectivites of the CYP3A4-catalyzed oxidations of compounds 5.3-

5.38 were evaluated using several methods: i) MS fragmentation; ii) comparison to authentic standards; iii) modification of the auxiliary-products by chemical oxidation; and iv) NMR analysis.

5.4.1 Structural determinations by mass spectroscopy fragmentation

102

Chapter 5

ESI-MS fragmentation was used to evaluate the structures of all CYP3A4-oxidized products. As mentioned in section 5.3, the gain in charge-to-mass ratio of most CYP3A4- oxidized products was +16 mass units, which corresponds to monooxidation. Although ESI-MS fragmentation alone did not allow definitive structural determination, several important structural characteristics were established with this method. For example, for all theobromine-containing molecules tested, oxidized or not, a distinct peak at 181 m/z was observed in the positive mode.

This peak corresponds to the mass of theobromine. From this observation, it is possible to conclude that the theobromine portion of the auxiliary-substrate remained unchanged in all cases.

Secondly, for many auxiliary-products, peaks were observed at 195, 209, and 223 m/z, corresponding to theobromine with 1, 2, or 3 methylene groups respectively (Figure 5.6).

Figure 5.6: The ESI-MS fragmentation pattern of the CYP3A4-oxidized auxiliary-products suggests that oxidation is occurring at C-4 or further from the theobromine auxiliary. This type of information was used to estimate the site of CYP3A4 oxidation for auxiliary-substrates 5.3-5.38.

Finally, the fact that a peak corresponding to a m/z of 237 (corresponding to

th TbCH2CH2CH2CH2) was never observed suggests that the site of oxidation may be the 4 carbon from the theobromine auxiliary, the C-4 position. This information was critical in the design of

103

Chapter 5 authentic standards used to further confirm the regio- and stereoselectivities of the CYP3A4- catalyzed reactions.

5.4.2 Structural determination by further oxidation of the auxiliary-products

In the case of 5.11 and 5.12, more structural information was gained by further oxidizing the CYP3A4-oxidized products with an excess of pyridinium chlorochromate (PCC) followed by re-analysis by ESI-MS (Figure 5.7).

Figure 5.7: The possible sites of CYP3A4-catalyzed oxidation of compounds 5.11 and 5.12 from ESI-MS fragmentation.

If oxidation to the ketone or aldehyde is observed, then the site of oxidation must be a methyl or methylene position. If oxidation is not observed, then this suggests that the site of oxidation is a methine.

5.4.3 Structural determination by comparison to authentic standards

Comparison to an authentic standard is the most reliable method to elucidate the structure of an unknown molecule. Using the information gathered from the ESI-MS fragmentation

104

Chapter 5 patterns, authentic standards of CYP3A4-oxidized products were designed and synthesized. The most common route was the Grignard derivatization of a theobromine-containing molecule having a carbonyl at C-4. 5.52-5.54 were directly obtained this way (Scheme 5.3). To make standards for auxiliary-products containing secondary alcohols, such as the expected CYP3A4- oxidized products of 5.5, 5.6, and 5.37, theobromine was functionalized with an ester followed by reduction to the alcohol and oxidation to the aldehyde. To make standards for auxiliary- products containing secondary tertiary or alcohols, such as the expected CYP3A4-oxidized product of 5.11, thebromine was functionalized with a ketone-containing electrophile (Scheme

5.3A). The resulting racemic alcohols were separated by chiral HPLC and the absolute stereochemistries were assigned via derivatization to two corresponding Mosher esters, followed by NMR analysis139 (Figure 5.8).

Scheme 5.3: The synthesis of authentic standards for auxiliary-products 5.47, 5.52, 5.53 and 5.54, corresponding to enzymatic reaction products from auxiliary substrates 5.11, 5.5, 5.6, 5.35 respectively.

105

Chapter 5

The synthesis of chiral, theobromine-containing epoxides, 5.65 and 5.66, began with the epoxidation of unsaturated alkyl-halides, affording 5.61 and 5.62, followed by hydrolytic kinetic resolution using the Jacobsen’s catalyst, affording 5.63 and 5.64. Theobromine was functionalized with the resulting enantio-enriched epoxides under basic conditions, affording

5.65 and 5.66 (Scheme 5.4A). The authentic standard of the auxiliary-product 5.67 was produced with a chemical oxidant, ammonium persulfate, used previously in our lab to selectively access sulfoxides from sulfides140 (Scheme 5.4B).

Scheme 5.4: The synthesis of authentic standards for auxiliary-products 5.65, 5.66, and 5.67, corresponding to enzymatic reaction products from auxiliary substrates 5.35, 5.63 and 5.34, respectively.

The authentic standards were compared to CYP3A4-oxidized products by HPLC co- injection using 2 or 3 different elution gradients and by comparison of ESI-MS fragmentation patterns. The enantiomeric ratio (er) of CYP3A4-oxidized products was evaluated by the co-

106

Chapter 5 injection of CYP3A4-oxidized products and enantioenriched authentic standards on a chiral

HPLC column. The use of enantioenriched standards allowed us to verify that the elution conditions led to a separation of both enantiomers.

Figure 5.8: The method used to assign the absolute stereochemistry of 5.52-5.54, corresponding to the CYP3A4-oxidized products of 5.5, 5.6, and 5.37, respectively.

5.4.4 Structural determination by NMR

In the case of compound 5.34, a sufficient amount of CYP3A4-oxidized product was produced to also allow for structural elucidation by 1H NMR, further supporting the validity of our methods.

5.5 Products of the transformation of auxiliary-substrate by CYP3A4

Without exception, the major theobromine-products obtained were oxidized at the 4th carbon from the theobromine auxiliary, the C-4 position, as expected from the CYP3A4 oxidation of lisofylline (Table 5.2 and 5.5). For each theobromine-substrate with a prochiral C-H bond at the C-4 position, oxidations proceeded with R-facial selectivity, giving predominantly R-

107

Chapter 5 enantiomers. The results are discussed below by groups of structurally related auxiliary- substrates.

5.5.1 Hydroxylation of theobromine-based auxiliary-substrates containing aliphatic substrates

To evaluate the ability of CYP3A4 to reliably distinguish between electronically and sterically similar C-H bonds, a series of simple, aliphatic theobromine-based auxiliary-substrates were reacted with the enzyme. In all cases, CYP3A4 catalyzed the hydroxylation at the C-4 position (Table 5.2). When no CH2 was available at C-4, (5.3, 5.4, 5.13, 5.14), no reaction was observed.

108

Chapter 5

Table 5.2: Auxiliary-product obtained for the enzymatic reaction of each auxiliary-substrate. C-4 regioselectivity, enantiomeric ratios, and isolated yields are listed. aAs determined by ESI-MS fragmentation, oxidation studies, and comparison to authentic standards; b% oxidation at C-4 position compared to all other products. 109

Chapter 5

In all cases where oxidation was observed, HPLC-MS analysis revealed that the mass-to- charge ratio of the new peaks corresponded to a gain of 16 mass units compared to the unoxidized substrate, as in the example of 5.5 (Figure 5.9).

Figure 5.9: The RP-HPLC trace for analysis of the reaction mixture for the CYP3A4-catalyzed hydroxylation of compound 5.5. The auxiliary-product, 5.52, appears at a lower retention time than 5.5, and the mass-to-charge ratio of the auxiliary-product peak corresponds to a gain of 16 mass units compared to that of the unoxidized auxiliary-substrate 5.5.

The er was determined by re-injection of the auxiliary-products onto a chiral RP-HPLC column and comparison with injection of the authentic enantioenriched standard, and, as a separate experiment, co-injection of the auxiliary-product with the corresponding standard

(Figure 5.10).

110

Chapter 5

Figure 5.10: Chiral RP-HPLC trace of 5.52 obtained by enzymatic reaction from reaction of 5.5 with CYP3A4.

As discussed in section 4.1.2, like a ruler, the distance between the auxiliary and the reactive heme iron was expected to control the position of hydroxylation. From the results illustrated in Table 2, it is clear that this auxiliary-enzyme system has a ‘ruler’ length of 4 carbons from the auxiliary, assuming typical straight-chain bond angles. This guideline explains why compound 5.5 is oxidized by CYP3A4 while compounds 5.3 and 5.13 are not turned over by the enzyme. Compound 5.14 has a methylene C-H bond at the C-4 position, however the more acute bond angles demanded by the 3- or 4-membered rings reduces the distance between the C-4 position and the auxiliary, shortening the distance below the critical point required to reach the heme iron. In the case of compound 5.4, the terminal methyl group of the substrate satisfies the length requirement but the substrate is not oxidized due to unfavorable electronics ie. the high bond strength of the methyl C-H bonds appears to preclude reactivity compared to the relatively weaker bond strength of the methylene C-H bond at C-4 of 5.5. The remaining

111

Chapter 5 compounds in Table 2 each have methylene or methine C-H bonds at the C-4 position and each are reliably hydroxylated with identical regio-selectivity. Importantly, compounds 5.10 and 5.12 were predominantly oxidized at the C-4 position, despite the proximity of much weaker methine

C-H bonds at the C-3 and C-5 positions, respectively. These results indicate that the positioning and orientation effects of the auxiliary are sufficient to overcome the electronic preference to oxidize the most electron rich C-H bond.

The CYP3A4-catalyzed oxidation of 5.15 was meant to verify the tolerance of the enzyme for enantiomeric pairs. When racemic 5.15 was treated with CYP3A4, the resulting hemiacetal spontaneously cleaved to give an alcohol and aldehyde (Figure 5.11). Despite a high isolated yield, the oxidized product was racemic, indicating that both enantiomers of 5.15 were oxidized by CYP3A4 equally efficiently.

Figure 5.11: The CYP3A4-catalyzed oxidation of racemic 5.15, resulting in racemic 5.71 following spontaneous cleavage of the resulting hemiacetal.

With the exception of 5.11, which lacks a prochiral C-H bond at the C-4 position, and

5.15, which is was synthesized as a racemic mixture, all substrate-auxiliaries in Table 2 were predominantly hydroxylated to the R enantiomers. This result suggests that the theobromine

112

Chapter 5 portion of the auxiliary-substrate binds in the active site of CYP3A4 in such a way that the pro-R face of the substrate is oriented more favorably towards the iron than the pro-S face.

Compounds such as 5.7 and 5.8 gave at least 3 CYP3A4-oxidized products (Figure 5.12), likely due to a high degree of conformational flexibly which may allow more C-H bonds to come within proximity of the heme iron.

Figure 5.12: The CYP3A4-catalyzed oxidation of 5.7 and 5.8 afforded at least 3 oxidized products each.

5.5.2 Enzymatic reactions of theobromine-based auxiliary-substrates containing aromatic substrates

P450s, including CYP3A4, are well known to oxidize aromatic substrates32. Surprisingly, none of the aromatic auxiliary-substrates shown in Table 5.3 were oxidized by CYP3A4 in significant quantities (Table 5.3), making the determination of the structural identities of products impractical.

113

Chapter 5

Table 5.3: Auxiliary-product obtained for the enzymatic reaction of auxiliary-substrates 5.18-5.21. C-4 regioselectivity, enantiomeric ratios, and isolated yields are listed. aAs determined by ESI-MS fragmentation, oxidation studies, and comparison to authentic standards; b% oxidation at C-4 position compared to all other products.

It should be noted that compounds 5.20 and 5.21 gave trace products compared to the other substrates (Figure 5.13). This enhanced reactivity is likely due to the mesomeric effect of the methoxy groups activating the aromatic C-H bonds of these compounds towards electrophiles such as the iron-oxo species in P450s.

It is possible the reduced reactivity of aromatic theobromine-substrates may be due to π- stacking interactions of the aromatic groups of the substrate with nearby aromatic residues. Such interactions may prevent the substrate from coming within proximity of the iron.

114

Chapter 5

Figure 5.13: The RP-HPLC trace obtained for analysis of the reaction mixture for the CYP3A4- catalyzed hydroxylation of compound 5.21. The oxidized auxiliary-product appears at a lower retention time and the mass-to-charge ratio of this new peak corresponds to a gain of 16 mass units compared to the unoxidized substrate. The remaining peaks were present in the control reaction.

5.5.3 Functional group tolerance of CYP3A4 with various auxiliary-substrates

Additional auxiliary-substrates functionalized near the C-4 position were evaluated to further confirm the regio- and chemoselectivity of CYP3A4-catalyzed oxidations. Results for these enzymatic reactions are shown in Table 5.4.

These results suggest that CYP3A4 is sensitive to electronic effects. Interestingly, functionalization with electron-withdrawing groups alpha to the C-4 position is detrimental to enzymatic activity (Table 5.4). It is well understood that removing electron density around C-H

115

Chapter 5 bonds increases the bond-strength, slowing reactions such as those catalyzed by the electrophilic reactive intermediate of CYP3A4.

Table 5.4: Auxiliary-product obtained for the enzymatic reaction of auxiliary-substrates 5.24-5.33. C-4 regioselectivity, enantiomeric ratios, and isolated yields are listed. aAs determined by ESI-MS fragmentation, oxidation studies, and comparison to authentic standards; b% oxidation at C-4 position compared to all other products.

116

Chapter 5

Interestingly, replacing highly electron-withdrawing groups with groups which are less electron-withdrawing appears to rescue a small amount of reactivity. This effect is illustrated by comparing the reactivity of compound 5.28 vs. 5.29 (Table 5.4, Figure 5.14), compound 5.30 vs.

5.31 (Table 5.4), and compound 5.32 vs. 5.33 (Table 5.4). In each case, the slight increase of electron density around the C-H bond at the C-4 position marginally increases the reactivity with

CYP3A4.

117

Chapter 5

Figure 5.14: The RP-HPLC spectra of the CYP3A4-catalyzed hydroxylation of compounds 5.28 and 5.29. The auxiliary-product appears at a lower retention time and the mass-to-charge ratio of this new peak corresponds to a gain of 16 mass units compared to the unoxidized auxiliary-substrate.

5.5.4 CYP3A4-catalyzed oxidation of heteroatom containing auxiliary-substrates

As described in section 1.2.3, P450s are well known to catalyze many reactions other than C-H oxidation, including S-oxidation. Compound 5.34 was designed to evaluate the ability of the theobromine auxiliary to direct CYP3A4-catalyzed oxidations at sulfur atoms. We were excited to find that CYP3A4 catalyzed the oxidation of the sulfur atom with total regioselectivity (to the limitation of our detection) and a moderate yield, resulting in a sulfoxide (Table 5.5).

118

Chapter 5

Table 5.5: Auxiliary-product obtained for the enzymatic reaction of auxiliary-substrate 5.34. C-4 regioselectivity, enantiomeric ratios, and isolated yields are listed. aAs determined by ESI-MS fragmentation, oxidation studies, and comparison to an authentic standard.

Importantly, none of the sulfone product was detected; once again demonstrating the ability of the CYP3A4 to catalyze difficult oxidations while avoiding over-oxidation.

Figure 5.15: The RP-HPLC trace of the CYP3A4-catalyzed hydroxylation of compound 5.34. The auxiliary-product appears at a lower retention time and the mass-to-charge ratio of this new peak corresponds to a gain of 16 mass units compared to the unoxidized auxiliary-substrate.

119

Chapter 5

5.5.5 CYP3A4-catalyzed oxidation of olefin-containing auxiliary-substrates

A series of compounds containing olefins at or around the C-4 position were designed and evaluated to determine which products would result from CYP3A4-catalyzed oxidation

(Table 5.6).

Table 5.6: Auxiliary-product obtained for the enzymatic reaction of auxiliary-substrates 5.35-5.38., C-4 regioselectivity, enantiomeric ratios, and isolated yields are listed. aAs determined by ESI-MS fragmentation, oxidation studies, and comparison to authentic standards; b% oxidation at C-4 position compared to all other products.

As demonstrated with 5.35 and 5.36, CYP3A4-catalyzed oxidation of theobromine- containing molecules with olefins at C-4 results in the formation of epoxides. In the case of 5.35, the resulting epoxide is racemic, however, in the case of 5.36 (Table 5.6), the epoxide is formed with an er greater than 99:1. This result represents, to the best of our knowledge, the greatest enantioselectivity observed for the terminal epoxidation of an inactivated olefin to date.

Although there is not currently a full mechanistic understanding of P450-catalyzed epoxidation34, 120

Chapter 5 it seems reasonable that these results can be interpreted by considering the pro-chirality of the C-

H bonds at the C-4 positions of compounds 5.35 and 5.36. This difference supports our hypothesis that the enzyme works at C-4. Indeed, the C-H bond at the C-4 position of compound

5.35 is not prochiral. Conversely, the same C-H bond of compound 5.36 is prochiral.

Figure 5.16: The RP-HPLC trace for the reaction mixture of the CYP3A4-catalyzed hydroxylation of compound 5.36. The oxidized product, 5.66, appears at a lower retention time and the mass-to- charge ratio of this new peak corresponds to a gain of 16 mass units compared to the unoxidized auxiliary-substrate.

The presence of an olefin alpha to the C-4 position, as in 5.37 and 5.38, does not perturb the chemo-, regio- or stereoselectivity of CYP3A4, a result which further supports the robust selectivity of this strategy.

5.5.6 Large scale (100 mg) CYP3A4-catalyzed oxidation of compound 5.5 121

Chapter 5

It is generally accepted that a reaction must accommodate a minimum scale of 100 mg of substrate to be considered synthetically practical. In order to demonstrate the synthetic utility of our strategy, scaled up experiments were carried out. We were pleased to find that the transformation of 100 mg of 5.5 affords compound 5.52 in 63% isolated yield after purification by flash chromatography. Moreover, the regioselectivity and enantioselectivity of the reaction are identical to those of reactions carried out at all other scales tested including 2 mg and 20 μg.

5.5.7 Verification of the role and the need of the theobromine auxiliary.

The necessity of the auxiliary was verified by reacting CYP3A4 with substrates lacking the theobromine group. A number of commercially available molecules resembling the substrate moieties of 5.4, 5.11, 5.15, and 5.16 were tested for transformation by CYP3A4 (Figure 5.18).

Also included in figure 5.17, are compounds tested in a previous study in our research group31.

None of the molecules were transformed by the enzyme to any detectable level (< 0.1%) without the theobromine auxiliary. Theobromine was also not transformed.

122

Chapter 5

Figure 5.17: None of the compounds shown were transformed by CYP3A4 under catalytic conditions. Results for the last 7 compounds are from Chefson et. al.31

5.6 CYP2D6 –catalyzed oxidation of auxiliary-substrates

To investigate if the ability of theobromine to direct P450 oxidations was unique to

CYP3A4, compounds 5.3-5.8 were combined with CYP2D6 under identical catalytic conditions as before. The HPLC traces of the reaction mixtures indicated that different products were formed than those observed for CYP3A4 and in some cases, no products were formed at all.

These results demonstrate the orthogonal reactivity of different P450s. This result was expected as the majority of structural differences which distinguish CYP3A4 from CYP2D6 are found in the residues involved in substrate access or binding (Figure 5.18).

123

Chapter 5

Figure 5.18: Comparison of the structures of CYP3A4 and CYP2D6.

5.7 In silico studies on the selectivity of CYP3A4–catalyzed oxidation of theobromine-based auxiliary-substrates

In an attempt to better understand the reasons for the observed regio- and stereoselectivites, we performed in silico docking studies using compounds 5.3-5.38 and one of the crystal structures of CYP3A4 available in the protein database (pdb ID#:ITQN)141. Docking studies were performed using FITTED, a software suite developed by the Moitessier lab at

McGill. Although P450 enzymes are known to be poorly compatbile with most in silico applications142 (partially due to the fact that many P450s feature large, flexible active sites) this

124

Chapter 5 docking survey led to a proposed binding mode consistent with our results. Figure 5.19 shows

5.5 bound in proximity to the CYP3A4 heme.

Figure 5.19: Figure obtained from docking studies. Proposed orientation of compound 5.5 relative to the heme group in the enzyme pocket of CYP3A4.

This binding mode shows that the C-H bond at the pro-R face of the C-4 position is the most likely bond to be oxidized, based upon proximity to the heme iron.

5.8 Cleavage of products from auxiliary-products

For use in the next step of a chemical synthesis, it will be generally desirable to remove the oxidized substrate from the auxiliary. Several of the most common sets of conditions reported for the deprotection of phthalamide-protected amines were evaluated with 5.53, to cleave the auxiliary from the theobromine-products. Nucleophilic conditions, such those which use hydrazine135, were unsuccessful; likely due to the aromatic nature of the theobromine group.

However, it was found that reducing conditions effectively released the amine from the amide

125

Chapter 5 bond (Scheme 5.4). Sodium borohydride was effective but only at elevated temperatures

(Scheme 5.5A). On the other hand, excess DIBAL-H at -78oC was highly effective and relatively gentle (Scheme 5.5B).

Scheme 5.5: Successful conditions for the cleavage of the theobromine auxiliary from auxiliary- products.

5.9 Conclusions

In summary, we have established a proof-of-concept demonstrating the utility of chemical auxiliaries for directing the selectivity of enzymatic reactions. We have demonstrated that inexpensive achiral theobromine is a powerful chemical auxiliary for CYP3A4 transformations, enabling the prediction of both the oxidation site and the facial selectivity. In all cases, oxidation at the C-4 position and pro-R facial selectivity are favored. The isolated yields were in the order of ~70% with purified enzyme at synthetically useful scales, and the use of membranes and whole-cell systems should greatly increase the scale at which the reaction can be

126

Chapter 5 effectively performed. This strategy provides easy access to small, enantioenriched or enantiopure, alcohols and epoxides that are otherwise difficult to access, often requiring either expensive enantioselective separation, or low yielding hydrolytic kinetic resolution. Although limited to substrates not considerably larger than the auxiliary, this method has the added advantages of not yielding over-oxidation products. Chemical methods for selective oxidations at methylene groups52a, 65, 67 preferentially lead to ketone formation and show poor functional group tolerance. Conversely, no carbonyl-containing products were detected here (<0.1%), and this enzyme has a well established extended functional group tolerance. This was confirmed here with the selective methylene hydroxylation in the presence of a nearby double bond and of the different functionalities present on the auxiliary itself. P450 enzymes, including CYP3A4, are known to catalyze N- and O-demethylation, yet none of the N-methyl groups of the theobromine auxiliary were affected during our transformations. We believe that new auxiliaries can be designed and exploited to afford different regio- and stereo-selectivities for CYP3A4 transformations. This approach should also apply to other P450s, and to enzymes from other families, to produce a series of biocatalyst/auxiliary systems, each with complementary selectivities. This strategy should therefore find broad application in synthesis.

127

Chapter 6

Chapter 6:

Molecularly Imprinted Polymers for the Selective Purification of

Theobromine-Containing Molecules

128

Chapter 6

6.0 Preface

In addition to a system which predictably directs P450 oxidations, we desired a strategy to maximize the recovery of the oxidized products.

Molecularly imprinted polymers (MIPs) are highly selective solid phases tailored to selectively bind to a single molecule from a complex mixture. Recognizing the need to selectively purify theobromine-containing molecules from complex mixtures, MIPs were synthesized and evaluated for this purpose.

The experimental work described in this chapter was performed by me; Tiffany Lai, an undergraduate student in the Auclair lab; and Vanja Polic, a graduate student in the Auclair lab.

Tiffany Lai and Vanja Polic performed some of the experimental manipulations including the cleaning, loading, washing, and elution of the MIPs for the experiments described in section

6.4.1 and 6.4.2. I designed all of the experiments, synthesized the MIPs, performed the HPLC studies, trained each student, and contributed to the experimental manipulations.

6.1 Introduction

Enabling the easy recovery of both the products and starting materials of chemical reactions can significantly improve the utility of a synthetic method. Most commonly, the purification of valuable components in complex mixtures is performed using chromatography.

While chromatography is valued for its versatility, it can be time-consuming, expensive (often requiring large amounts of organic solvents and costly equipment), and compound detection is limited by the detectors available (typically UV/vis). While versatility is important for methods tasked with the purification of diverse classes of molecules from assorted mixtures, it is less 129

Chapter 6 important for methods limited to the purification of a single class of molecules from a single type of mixture. We required a strategy for purifying theobromine-containing molecules from complex aqueous mixtures; this scenario is ideally suited to the use of MIPs.

MIPs are often compared to plastic antibodies. These materials are made by first adding polymer subunits to a solution containing the template ligand. The template ligand and the polymer subunits associate non-covalently, followed by the addition of a second polymer subunit and a radical initiator. Subunits polymerize to themselves and each other in the presence of the template ligand (Figure 6.1). After grinding the polymer into a fine powder and the removal of the template ligand, the resulting polymer will contain numerous pockets which are ideally suited in terms of size and/or electrostatic interactions, for binding the original template ligand. As in affinity chromatography, MIPs can be used to „capture‟ the template molecules in a complex mixture143.

130

Chapter 6

Figure 6.1: A simplified cartoon depicting the synthesis of an MIP specific for theobromine derivatives. A) The template ligand is exposed to monomeric units (circles) designed to bind non- covalently to the ligand. B) The second subunit (rods) are introduced. This subunit is designed to bind to itself and to the first monomer, solidifying the network around the ligand. C) The template is removed, leaving cavities ideally suited to binding molecules resembling the template.

MIPs have been used as for sensing specific compounds in complex mixtures144, the extraction of various biological molecules from macerated samples143, 145, the gas-phase detection of explosives146, and as catalysts by binding multiple substrates in orientations resembling the transition state of their product147. The study which attracted our attention was a report describing the use of MIPs for the purification of caffeine and related molecules from tea148. As xanthine-containing molecules, caffeine and theobromine are structurally related

131

Chapter 6

(Figure 6.2) and thus, we expect this method to apply towards the purification of theobromine- products or theobromine-substrates (described in chapter 5) from complex mixtures such as those resulting from catalysis using CYP3A4-containing membranes or even whole-cell fermentations.

Figure 6.2: The structural relationship between theobromine and caffeine.

6.2 MIP Preparation

Inspired from previous work with caffeine137, the MIPs used in this investigation were prepared using methacrylic acid and ethylene glycol dimethacrylate (Figure 6.3). 6.2 is expected to associate with the target ligand non-covalently, most likely via hydrogen bonds resulting from the association of the hydrogen bond donors/acceptors of 5.1 or the chosen theobromine- substrate and those of 6.2. Benzoyl peroxide, 6.3, will initiate a radical polymerization reaction in the presence of heat (Figure 6.4) which will result in the cross-linking of 6.4 to 6.2 and to itself, forming a rigid network around the bound template ligand.

132

Chapter 6

Figure 6.3: The structures of the components used to construct the MIPs: methacrylic acid, 6.2, Benzoyl peroxide (the radical initiator) 6.3, and ethylene glycol dimethacylate, 6.4.

Figure 6.4: The benzoyl peroxide initialized radical polymerization of 6.2 (R = H) and 6.3 (R = C6O2H8).

Using this protocol of Theodoridis and Manesiotis148, a series of MIPs were designed for the purification of 5 structurally similar theobromine derivatives. Each MIP was designed to be selective for either 6.1, 6.2, 5.3, 5.6, or 5.53 (Figure 6.5). Thus, 6.1, 6.2, 5.3, 5.6, or 5.53 were each used separately to generate different MIPs using the same polymerization reaction. 133

Chapter 6

Methacylic acid was combined with one theobromine-containing compound in acetonitrile and mixed thoroughly to encourage the association of compounds by non-covalent interactions before the addition of ethylene glycol dimethacrylate and the radical initiator, benzyl peroxide.

The neat mixture was heated at 60oC overnight, resulting in a solid, off-white polymer. The solid polymer was removed from the reaction vessel and ground using a mortar and pestle, resulting in a fine powder. The finest particles were removed by first suspending the powder in water and carefully decanting the water once the majority of the powder had settled. Finally, the template molecules were removed from the solvent accessible regions of the powder by repeated washings with a 9:1 mixture (v/v) of methanol and acetic acid before loading the powder into empty solid- phase extraction tubes or empty flash chromatography columns.

Figure 6.5: The theobromine-containing compounds evaluated as templates for MIP preparation.

6.3 Assay

Before assays were performed, the MIPs were cleaned with 10 column volumes of a 9:1 mixture (v/v) of methanol and acetic acid. This mixture was previously reported to effectively elute molecules such as caffeine or 5.1 from an MIP similar to those prepared here137. This mixture likely disrupts the hydrogen bonding networks responsible for the association of the template and the MIP, thus releasing the template and leaving the MIP clean for use. The last

134

Chapter 6 fraction of this cleaning eluate was kept and analyzed (by LC-MS) to ensure that the solvent accessible region of the MIP was free of the template compound. After ensuring that the MIP was clean, aqueous solutions of theobromine derivatives 6.1, 6.2, 5.3, 5.6, or 5.53 were loaded onto the column using centrifugation or pumping and the entirety of the flow-through was collected, leaving the solid-phase dry. The flow-through was analyzed by LC-UV-MS to account for ligand that did not bind to the MIP. Next, the MIP was washed with several column volumes of water and the flow-through was again collected and analyzed to account for non-specifically bound ligand. Finally, the specifically bound ligand was eluted from the MIP with several column volumes of a 9:1 mixture (v/v) of methanol and acetic acid for the same reasons as previously mentioned. The amount of ligand in each fraction was determined by comparison to an external standard using LC-UV-MS.

6.4 Results and discussion

6.4.1 MIP efficacy

First, the MIPs were evaluated for their ability to retain the compounds for which they were designed from a solution of water. This most basic test of the MIP efficacy was performed by using relatively concentrated aqueous solutions (1 mM) of ligands (Figure 6.6).

135

Chapter 6

Figure 6.6: The efficacy of 5 different MIPs when used to recover the compounds for which they were designed from a concentrated solution (1 mM) in water. Experiments were performed in duplicate. The standard deviation for each elution was 5% on average.

These results (Figure 6.6) support several important conclusions. First, the near lack of ligand in the ‘flow through’ fraction indicates that the ligand was effectively removed from the water solution and bound to the MIP. Secondly, the lack of ligand in the ‘wash’ fractions indicates that the ligand was tightly bound to the MIP solid-phase. Finally, the near total recovery of ligand in the ‘elution’ fraction indicates that the ligand was released from the MIP

136

Chapter 6 when the non-covalent interactions (hydrogen bonding, most importantly) were disrupted by the addition of the elution solvent.

In each case, there was approximately 10% of ligand missing from the combined totals of each fraction. This consistent deficit can be partially explained by the attrition of product in the manipulations required by the analysis such as transfers of solutions from vessel to vessel.

This experiment demonstrates the ability of the MIPs to recover theobromine-containing molecules from relatively concentrated and otherwise pure aqueous solutions of water.

The affinity of 5.6 for the MIP prepared for 5.6 was determined using Scatchard analysis149. The Scatchard equation is:

B/U = (Bmax – B) / KD

Where: B is the amount of ligand bound per gram of MIP used. This value was measured by the HPLC absorbance (using the previously determined value ε = 8.81 E+03 M-1 cm-1) of the elution fraction, indicating the amount of ligand released from the MIP during elution. U is the equilibrium concentration of unbound ligand, measured by subtracting B from the total amount of ligand initially loaded onto the MIP. Bmax is the maximum binding capacity and KD is the equilibrium dissociation constant.

Figure 6.7 is a plot of these parameters where KD is the slope and Bmax is the intercept.

137

Chapter 6

Figure 6.7: Scatchard plot for 7 and the MIP designed for 7.

This plot features two distinct linear regions, as is expected when using Scatchard analysis to evaluate an MIP150. It is known that MIPs often contain two heterogeneous binding sites on the polymer surface, high affinity sites and low affinity sites. The first linear portion describes the binding parameters of the high affinity sites, and the second linear portion describes the binding parameters of the low affinity sites. The relevant binding parameters used for analysis are determined by the amount of ligand loaded onto the MIP. When using amounts of ligand below the Bmax of the high affinity region, the KD of the high affinity region may be considered the relevant equilibrium dissociation constant. From the slope and intercept of the above plot, the equilibrium dissociation constant and maximum binding capacity of 5.6 on the

-4 -1 MIP designed for 5.6 were KD = 1.42×10 M , Bmax = 3.7 μmol/g for the high affinity binding

138

Chapter 6

-4 -1 sites and KD = 8.33×10 M , Bmax = 12.3 μmol/g for the high affinity binding sites. These values are in good agreement with previous evaluations of similar MIPs for similar ligands.149-150

6.4.2 MIP Reusability

Although efficacy is a necessary pre-requisite of any solid-phase extraction technique, reusability is also of great importance. To evaluate reusability, the MIPs were cleaned and the

MIP efficacy experiment was repeated (Figure 6.7).

Figure 6.7: The reusability of 5 different MIPs when used to recover the compounds for which they were designed from a concentrated solution (1 mM) in water. Experiments were performed in duplicate. The standard deviation for each elution was 5% on average.

139

Chapter 6

In all cases, the results mirrored those of the MIP efficacy experiment with no loss in the affinity of the MIPs for the ligands. This result demonstrates that the MIPs are reusable, an important quality for a practical solid-phase extraction technique. The MIPs were each reused several more times with no significant loss in efficacy.

6.4.3 MIP Selectivity

As discussed above, selectivity is an important advantage of MIPs over other purification techniques. Having established that the MIPs were effective for the recovery of the compounds from which they were designed, they were next evaluated for their selectivity by probing their ability to recover molecules structurally related to those from which they were designed. 5.6 and

5.53 were chosen for this experiment as they had already been established to be a substrate and product, respectively, of CYP3A4-catalyzed oxidation (see section 5.5.1) and are highly similar in terms of structure (Figure 6.4). The MIPs designed for 5.6 and 5.53 were evaluated with ligands for which they were not designed: thus 5.6 was purified using an MIP that was prepared for 5.53 as a template and vice versa (Figure 6.7).

140

Chapter 6

Figure 6.8: The selectivity of 2 different MIPs when used to recover compounds structurally related to those for which they were designed from a concentrated solution (1 mM) in water. Experiments were performed in duplicate. The standard deviation for each fraction was less than 5%.

The results (Figure 6.8) illustrate that the MIP designed to bind to 5.53 binds with significantly less affinity to 5.6 despite the high degree of structural similarity shared between these molecules. This selectivity may be due to the additional hydrogen bond donor of 5.53 forming additional interactions with methacrylic acid. These results suggest that the MIPs may be capable of selectively purifying their target compounds from complex mixtures.

141

Chapter 6

6.4.4 MIP purification of theobromine-containing compounds from dilute solutions.

Biocatalysis often requires the purification of small amounts of products from dilute, complex mixtures containing many classes of small molecules as well as large biological molecules including proteins and DNA. Although the MIPs have been demonstrated to be highly effective for the recovery of their target molecules dissolved in pure, concentrated solutions, it was important to establish their efficacy with complex, dilute solutions. First, the effect of dilution was evaluated. The MIP designed using 5.53 as a template was evaluated for its ability to purify 5.53 from 50 mL solutions containing template (50 μM) dissolved in water or Luria broth (LB) (Figure 6.9), a complex and nutritionally rich mixture.

142

Chapter 6

Figure 6.9: The selectivity of the MIP designed for 5.53 when used to recover 5.53 from relatively dilute (50 μM) solutions of water and LB broth. Experiments were performed in duplicate. The standard deviation for each fraction was less than 5%.

The results from experiments in water demonstrate that dilution does not significantly affect the ability of the MIP to selectivity purify the target compound. Furthermore, the complexity of the mixture, as demonstrated with the LB matrix, did not significantly affect the ability of the MIP to selectively purify compound 5.53, a result which demonstrates that the MIP is specifically binding to the molecule for which it was designed.

6.4.5 MIP purification of theobromine-containing compounds from CYP3A4-containing membrane-catalyzed oxidations. 143

Chapter 6

In general, CYP3A4-catalyzed oxidations of theobromine-substrates results in yields ranging between 40-80% (see section 5.5), with starting material comprising the remainder. The ability to recover product as well as starting material is an important consideration in the pursuit of highly efficient derrivative syntheses17. Although the ability of the MIPs to selectively bind to a single theobromine-containing compound in simple and complex mixtures has been evaluated so far, it was important to establish the ability of MIPs to recover theobromine-products from actual biocatalytic mixtures. Thus, the MIP using 5.53 as template was evaluated for its ability to selectively recover 5.53 (product) and 5.6 (starting material) from 50 μM aqueous solution containing 75% 5.53 and 25% 5.6. This ratio reflects a typical unrecovered yield of a CYP3A4- catalyzed oxidation (see section 5.5.1). Additionally, the ability of the MIP to selectively recover, 5.53, and 5.6, from a CYP3A4-containing membrane-catalyzed oxidation of 5.6 was evaluated (Figure 6.10). The reaction was performed as described in section 5.3.

144

Chapter 6

Figure 6.10: The recovery of 5.53 and 5.6 from a ‘spiked’ CYP3A4-containing membrane solution and from a CYP3A4-containing membrane catalyzed oxidation of 5.6. Experiments were performed in duplicate. The standard deviation for each fraction was less than 5%.

In each case, the MIP was able to effectively recover both product and starting material from the mixtures. Although there is a small reduction of selectivity towards 5.6 (starting material) compared to 5.53 (product) for the CYP3A4-containing membrane-catalyzed conversion experiment, both compounds appear to be recovered equally from the spiked

CYP3A4 containing membrane experiment. This higher than expected recovery of 5.6 is likely due to the fact that only 33% as much 5.6 is being loaded onto column as compared to 5.53 and only 25% as in in previous experiments using 5.6. It is well accepted that solid phase extraction

145

Chapter 6 techniques are more efficient as the ratio to solid phase to target compound increases and this effect may be responsible for the higher than expected selectivity towards 5.6 observed here.

The HPLC traces of each fraction show that a large amount of non-specifically bound material was eluted from the MIP during the flow-through stage (Figure 6.11).

Figure 6.11: HPLC traces of different fractions following MIP purification of a CYP3A4-containing membrane-catalyzed oxidation of 5.6. Panel A: The flow-through fraction contains a large amount of material which elutes near the solvent front but contains no auxiliary-product (5.53) or auxiliary- substrate (5.6). Panel B: The elution fraction contains approximately 90% of both the auxiliary- product (5.53) and the auxiliary-substrate (5.6) expected from the reaction.

146

Chapter 6

Indeed, even a simple visual inspection of each fraction shows that the majority of the material present in the mixture other than product and starting material passed through the MIP without binding (Figure 6.12).

Figure 6.12: Fractions from the MIP purification of the CYP3A4-containing membrane-catalyzed oxidation of 5.6. A: The flow-through fraction containing most of the material other than the no auxiliary-product (5.53) or auxiliary-substrate (5.6); B: The wash fraction is significantly less turbid than the flow-through; C: The elution fraction, is clear.

6.5 Conclusions

The experiments and results presented in this chapter demonstrate that MIPs represent an effective and versatile method for the purification of theobromine-containing molecules from complex mixtures such as CYP3A4-catalyzed hydroxylations. In addition to recovering the

147

Chapter 6 product, the MIP was also able to recover unreacted starting material, an important consideration towards ‘greener’ synthesis by improving atom economy (the total amount of materials entering a reaction versus total amounts of products and starting materials recovered). In addition to being highly effective for purification, the MIPs were easy to customize to chosen ligands, highly specific, and reusable. Most importantly, the MIPs tested here performed well when used to purify products and starting materials from real reaction mixtures, demonstrating practical efficacy.

148

Chapter 7

Chapter 7:

Contributions and Future Directions

149

Chapter 7

7.1 Contributions

This thesis contributes towards science in four ways. This first is the pioneering of chemical auxiliaries capable of directing the reactivity of highly promiscuous enzymes. The second is improving the predictability of P450 enzymes, CYP3A4 in particular. The third is enabling the new biocatalytic routes to valuable chiral alcohols and epoxides. The fourth is the application of a known method towards the recovery of target molecules from biocatalytic mixtures.

7.1.1 Contributions towards the design of enzyme-directing auxiliaries

As the need for drugs and other complex molecules increases, so does the need for new methods to catalyze reactions151. In particular, biocatalysts represent a powerful and green alternative to chemical catalysts and will likely continue to gain prominence in the years and decades to come. However, the biocatalytic promise of many enzymes is hindered by the inability to control the regio- and stereoselectivity of the reactions they catalyze152. Chapter 4 and 5 showed that it is possible to design chemical auxiliaries which can control the selectivity of enzymes, giving access to biocatalyzed oxidations with desired regio- and stereoselectivity. This work is the first of its kind and may represent a paradigm shift in biocatalysis. Previous efforts in biocatalysis have generally favored the modification of enzymes to accommodate select substrates153. The work described here may enable an orthogonal approach: The modification of substrates to accommodate the use of select enzymes. It is possible that this work may represent

150

Chapter 7 be the first of many efforts to expand the use of auxiliaries to direct biocatalysis using P450s and potentially other families of enzymes.

7.1.2 Contributions towards the improved predictability of CYP3A4

Drug-metabolizing P450 enzymes are well known for the difficulty inherent in predicting their reactivity34. The selectivity of CYP3A4, in particular, has proved especially difficult to predict. Indeed, product prediction has met with limited success in vitro32, in vivo154, and due to the large and flexible active site, in silico155. The work in chapters 3, 4, and 5 contributes directly to improving the predictability of CYP3A4 catalysis. This work demonstrates that interactions between the enzyme and a chemical auxiliary can be exploited to predict the site of catalysis on substrates attached to the auxiliary. While a previous study demonstrated the use of a carbolide to „anchor‟ macrolides into P450 PikC131, this method did not however control the site of oxidation and generated multiple products without predictability. Instead, the auxiliaries described here anchor the substrate and also allow for predictable control of the chemo-, regio-, and stereoselectivity of the catalyzed reaction. Furthermore, the knowledge gained from the computational studies used to justify the observed selectivities may be used to inform protein engineering studies which may contribute further to understanding of CYP3A4 catalysis.

7.1.3 Contributions towards the semi-synthesis of valuable chiral alcohols and epoxides

Small, chiral alcohols and epoxides are highly valuable as starting materials and synthons in the synthesis of complex molecules and drugs. A simple comparison of the prices of 2-hexanol and (R)-(-)-2-hexanol of similar purities from Sigma-Aldrich demonstrates the value of enantiopure materials (Figure 7.1). A similar comparison of 1,2-epoxyheptane and (R)-(+)-1,2-

151

Chapter 7 epoxyheptane further illustrates the point (Figure 7.1). It is clear that improving access to such materials will accelerate and reduce the cost of drug development and manufacture.

Figure 7.1: A comparison of the prices for racemic and enantiopure materials.

The work in chapter 5 demonstrates a new biocatalytic route to small, chiral alcohols and epoxides. This strategy was shown to be compatible with easy-to-prepare CYP3A4-containing membranes as well as the purified enzyme. It is likely that this technique can be modified to work with whole cell fermentations to enable the synthesis of these materials at industrial scales.

7.1.4 Contributions towards the recovery of valuable molecules from complex matrices

Biocatalysis at large scales often requires the use of matrices containing complex mixtures of biomolecules, enzyme containing membranes, or even whole cells. For the best isolated yield, a biocatalytic method should couple high conversion efficiency with a method to achieve near complete product recovery. Chapter 5 demonstrated that theobromine-directed

CYP3A4 oxidations can achieve high conversion efficiently and the work in chapter 6 addressed product recovery. Specifically, the work in chapter 6 expands on previous efforts to isolate

152

Chapter 7 xanthine-containing molecules from complex mixtures137. We applied this method to the recovery of theobromine-substrates and theobromine-products from complex biocatalytic mixtures. This work is important for two reasons: First, It addresses the need for near complete product recovery in the biocatalytic methods discussed in chapter 5, increasing the utility of our strategy. Secondly, it demonstrates the viability of the general use of MIPs as easily tailored solid-phases for the purification of desired compounds diluted in complex mixtures.

7.2 Publications

1) Gao, F.; Yan, X.; Zahr, O.; Larsen, A.; Vong, K.; Auclair, K., Synthesis and use of sulfonamide-, sulfoxide-, or sulfone-containing aminoglycoside-CoA bisubstrates as mechanistic probes for aminoglycoside N-6'-acetyltransferase. Bioorg Med Chem Lett 2008, 18 (20), 5518-

22.

2) 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), 7853-8.

3) Yan, X.; Akinnusi, T. O.; Larsen, A. T.; Auclair, K., Synthesis of 4'-aminopantetheine and derivatives to probe aminoglycoside N-6'-acetyltransferase. Org Biomol Chem 2011, 9 (5),

1538-46.

4) Tewari, B. B.; Beaulieu-Houle, G.; Larsen, A.; Kengne-Momo, R.; Auclair, K.; Butler, I.

T., An overview of molecular spectroscopic studies on theobromine and related alkaloids. Appl

Spectrosc Rev Accepted 18 Oct 2011, in press.

153

Chapter 7

5) Larsen, A. T.; Polic, V.; Lai, T.; Auclair, K., The use of molecularly imprinted polymers for the recovery of theobromine-containing compounds from biocatalytic mixtures. Manuscript to be submitted in January 2012.

6) Larsen, A. T.; Auclair, K., The functional group tolerance of P450 3A4-catalyzed hydroxylations of theobromine-containing molecules. Manuscript to be submitted in January

2012.

7.3 Future Directions

The ultimate objective of this project was to improve the predictability, stability/activity and general utility of P450s, especially CYP3A4, as biocatalysts. Previously, the Auclair lab addressed several barriers impeding the utility of CYP3A4 and CYP2D6 such as the need for expensive, cumbersome cofactors32, and the issue of solvent tolerance31. The research described in this thesis addressed remaining barriers including stability/activity, predictability, and product recovery. Although major advancements were made towards overcoming several of these barriers, many challenges still remain.

7.3.1 The development of new P450-directing chemical auxiliaries

The work described in chapters 4 and 5 can be considered a proof of concept: This is the first example of the use of a chemical auxiliary to predictably direct enzymatic catalysis. The foundations have been laid for the exploration of myriad new auxiliaries to direct CYP3A4 towards other selectivities, other P450 isoforms or any other biocatalyst.

154

Chapter 7

We believe that the strategies used here to identify and develop potential auxiliaries are sound and a next step would be to apply them in the search for new auxiliaries. By docking a small library of known substrates, a few potential new CYP2D6-directing auxiliaries have already been identified. Metoprolol, a selective β1 receptor blocker used in treatment hypertension156 is one example. This compound is predicted by docking studies to bind to the active site of CYP2D6, projecting an easily functionalized group towards the reactive iron-oxo species (Figure 7.2).

Figure 7.2: The binding mode of metoprolol in the active site of CYP2D6 as predicted by FITTED.

155

Chapter 7

While the predictable oxidation of small molecules is an important goal, it would be valuable to develop auxiliaries capable of directing P450 oxidations to specific sites on large, drug-like molecules, giving access to late-stage oxidations which would greatly streamline the synthesis of highly valuable drugs. We predict that auxiliaries larger than theobromine would be necessary for such efforts. It stands to reason that the larger the substrate, the larger the auxiliary necessary to ensure that the enzyme-auxiliary interactions alone controls the binding mode of the auxiliary-substrate.

7.3.2 Enzyme engineering to modify CYP3A4 binding to theobromine-substrates.

Although the P450-directing auxiliaries developed here have been demonstrated to control the selectivity of P450 oxidations with acceptable regio- and stereoselectivity, there is room for improvement. It may be valuable to mutate residues involved in substrate binding, either rationally or randomly to generate a library of recombinant CYP3A4s and to compare their transformation of the theobromine-substrate reported in this thesis.

7.3.3 Engineering more stable, selective, and predictable P450s

The efforts described herein towards improving the activity/stability of P450s have, so far, been met with limited success. As described in chapter 1, there exist P450 isoforms from extremophiles which exhibit increased stability compared to most mammalian isoforms.

However, these P450s, in general, have greatly reduced substrate scopes compared to their mammalian counterparts. It is my opinion that generating chimeras incorporating the active site architectures CYP3A4 with the structural features thought to be responsible for increased stability in extremophile P450s may yield a stable P450 which would remain compatible with the auxiliaries reported in this thesis. This effort would begin with computational studies to predict 156

Chapter 7 the structure and thermal stability of potential chimeras157 followed by the expression of the best candidates and the evaluation of their stabilities and selectivities.

7.3.4 P450s as Diels-Alderases

Chapter 3 explored the possibility of using CYP2E1 as a Diels-Alder-catalyzing enzyme.

Although the several Diels-Alder reactions probed in this chapter were not accelerated by

CYP2E1, it was determined that the enzyme was binding to most of the substrates. As described in the chapter, the active site of CYP2E1 is flanked by several aromatic residues (Figure 7.3). It is our belief that Diels-Alder substrates containing aromatic groups may be better potential candidates for CYP2E1 catalysis than the substrates tested so far.

Figure 7.3: Binding site of CYP2E1 bound to 10-(1H-imidazol-1-yl)decanoic acid (from crystal structure2). Red arrows indicate aromatic residues.

157

Chapter 7

7.3.5 Conclusions to future directions

Although many questions were satisfied in the course of this research project, many more questions have arisen. It is certain that there are various important directions in which future research can proceed. We believe that through such efforts, P450s will be harnessed as versatile, robust, and effective biocatalysts capable of achieving otherwise inaccessible reactions for both research and industrial activities.

158

Chapter 8

Chapter 8:

Experimental Protocols

159

Chapter 8

8.1 General Methods

Unless otherwise noted, all reagents were purchased from Sigma-Aldrich Canada, Ltd.

(Oakville, Ontario, Canada). Reagents and solvents were used without further purification except where stated. Flash chromatography and TLC analyses were performed with 60 Å silica gel from

Silicycle (Quebec, Canada). Preparative TLC was performed with type 60 F254 plates from EMD

(Gibbstown, NJ). Separations by reverse-phase RP-HPLC was achieved using an Agilent 1100 modular system equipped with an autosampler, a quaternary pump system, a photodiode array detector, a thermostatted column compartment and a ChemStation (for LC 3D A.09.03) data system. When MS detection was used, the MS was an Agilent 6120 quadrupole LC/MS. The columns used were an analytical 4.6  250 mm, 4 µm SYNERGI 4µ Hydro-RP 80 A

(Phenomenex, Torrance, CA), an analytical 2.6  100 mm KINETEX 2.6 µ C-18 100 A

(Phenomenex, Torrance, CA), a chiral analytical 4.6  250 mm, 5 µm Lux 5µ Cellulose-1

(Phenomenex, Torrance, CA), and a chiral semi-preparative 10  250 mm, 5 µm Lux 5µ

Cellulose-1 (Phenomenex, Torrance, CA). Samples were eluted using a combination of mobile phase A (0.1% aqueous formic acid), mobile phase B (acetonitrile containing 0.1% formic acid) and/or mobile phase C (methanol containing 0.1% formic acid. The detector was set to 273 nm.

The different HPLC elution gradients and flow rates are detailed in Table 1 below.

160

Chapter 8

Table 8.1: Flow rates and linear gradient profiles used for HPLC analyses and purifications. A is water and B is acetonitrile. All gradients were linear.

Non-Chiral Method A: 0.5 ml min-1, with the analytical SYNERGI 4µ Hydro-

RP 80A

Time %A %B

(min)

0 50 50

20 5 95

Non-Chiral Method B: 0.5 ml min-1, with the analytical Kinetex 2.6 µ C-18

100 A

Time %A %B

(min)

0 35 65

8 35 65

Chiral Method A : 0.5 ml min-1, with the chiral analytical Lux 5µ Cellulose-1

Time %A %B

(min)

0 50 50

20 5 95

161

Chapter 8

Chiral Method B: 1.5 ml min-1, with the semi-preparative chiral Lux 5µ

Cellulose-1

Time %A %C

(min)

0 50 50

60 5 95

All chemical reactions were performed under an atmosphere of nitrogen using standard anhydrous techniques. Anhydrous THF was obtained from distillation over sodium. Anhydrous

DCM was obtained from distillation over CaH2. High resolution mass spectra (HR-MS) were recorded on a Thermo Fisher Scientific Inc. Exactive Orbitrap system. Liquid chromatography- mass spectra (LC-MS) were recorded on a Thermo Fisher Scientific Inc. Finnigan LCQ Duo or in the lab using an Agilent 6120 quadrupole LC/MS. 1H NMR spectra were obtained with Varian

Mercury 400 or 500 MHz NMR spectrometers. Chemical shifts are reported in parts per million

(ppm) downfield from tetramethylsilane. The peak patterns are indicated as follows: s, singlet; d, doublet; t, triplet; dd; doublet of doublet; dt, doublet of triplet; ddd, doublet of doublet of doublet; td, triplet of doublet; m, multiplet; q, quartet; p, and br, broad singlet. Coupling constants are reported in Hertz (Hz). 13C NMR spectra were obtained with a Varian Mercury 300

MHz NMR spectrometer. Chemical shifts are reported in ppm relative to the center peak of the triplet at 77.0 ppm corresponding to deuteriochloroform. 13C NMR assignments for xanthine- containing compounds were made by comparison to the structure of caffeine158 and by the cross- comparison of all structurally related compounds reported here. HSQC spectra were collected to derive 13C NMR chemical shifts and were recorded on a Varian Mercury 500 MHz NMR 162

Chapter 8 spectrometer. GC-MS spectra were recorded on an Agilent Technologies 6890N Network GC system coupled to a 5973 Inert Mass Selective Detector.

8.2 Biological studies

8.1.1 Expression and purification of enzymes

Purified CYP3A4, CYP2E1 and CPR were expressed and purified according to previously reported procedures48. The average yield of protein is about 15 nmol, 7 nmol, and 10 nmol per liter of culture for CYP3A4, CYP2E1, and CPR, respectively.

A variation of the above procedure was used to prepare CYP3A4-containing membranes.

Following the sonication step, the resulting viscous mixture was centrifuged at 10,000  g for 20 minutes. The DNA-containing pellet was discarded and the supernatant was centrifuged at 53000

 g for 65 minutes. The new pellet was resuspended in potassium phosphate buffer (15 mL,

0.1M, pH 7.4) per liter of culture, affording a solution of CYP3A4-containing membranes. The concentration of CYP3A4 was determined as previously described, affording approximately 1 mL of 15 μM per liter. The solution was frozen at -80oC until use.

8.1.2 Enzymatic assays

8.1.2.1 Optimized enzymatic transformation assay using purified enzyme

The desired substrate (60 nmol, 600 µM final concentration) and CYP3A4 (1.2 nmol, 12

µM final concentration) were mixed in potassium phosphate buffer (100 µL, 0.1 M, pH = 7.4) and pre-incubated at 27oC for 5 minutes. Reactions were initiated by the addition of the cofactor 163

Chapter 8 surrogate, CHP (60 nmol, 600 µM final concentration). Additional CHP was added at 45 minutes

(60 nmol, 600 µM final concentration). Controls were run in the absence of CHP. Reactions were allowed to proceed at 27oC with orbital shaking (250 RPM). For convenience, the reactions were generally terminated after 12 hours with the addition of DCM (1.5 mL); however, quenching after 2-3 hours affords similar yields (± 10%). The aqueous fraction was further extracted with DCM (2 x 1.5 mL). The organic fractions were obtained and concentrated. The organic residue was redissolved in acetonitrile (200 µL) for HPLC analysis using Non-Chiral

Method A. Products were collected by HPLC and product identities were confirmed by comparison with synthetic standards on LC-MS using Non-Chiral Method A. Purified products were re-injected on HPLC using Non-Chiral Method A and isolated yields were quantified by

UV absorbance at 273 nm using the extinction coefficient of caffeine (8750 M-1cm-1).

Calibration curves of 5.6 and 5.53 demonstrated that this extinction coefficient was appropriate for the theobromine-containing molecules described here.

8.1.2.2 Optimized enzymatic transformation assay using CYP3A4-containing membranes

The desired substrate (60 nmol, 600 µM final concentration) and CYP3A4-containing membranes (1.2 nmol CYP3A4, 12 µM final concentration of CYP3A4) were mixed in potassium phosphate buffer (10 mL, 0.1 M, pH = 7.4) and pre-incubated at 27oC for 5 minutes.

Reactions were initiated by the addition of CHP (60 nmol, 600 µM final concentration).

Additional CHP was added after 45 minutes (60 nmol, 600 µM final concentration). Controls were run in the absence of CHP. Reactions were allowed to proceed at 27oC with orbital shaking

(250 RPM). For convenience, the reactions were generally terminated after 12 hours with the addition of DCM (1.5 mL); however, quenching after 2-3 hours affords similar yields (± 10%).

164

Chapter 8

The organic fractions were combined and concentrated. The organic residue was redissolved in acetonitrile (200 µL) for HPLC analysis using non-chiral method A. Products were collected by

HPLC and product identities were confirmed by comparison with synthetic standards on LC-MS using Non-Chiral Method A. Purified products were re-injected on HPLC using non-chiral method A and isolated yields were obtained using UV absorbance at 273 nm.

8.1.2.3 100 mg scale CYP3A4-catalyzed transformation of compound 5.5

Compound 5.5 (100 mg, 400 µmol, 800 µM final concentration) and CYP3A4 (6 µmol,

12 µM final concentration) were mixed in potassium phosphate buffer (500 mL, 0.1 M, pH =

7.4) and pre-incubated at 27oC for 5 minutes. Reactions were initiated by the addition of CHP

(400 µmol, 800 µM final concentration). Additional CHP was added after 1.5 hours (400 µmol,

800 µM final concentration). Controls were run in the absence of CHP. Reactions were allowed to proceed at 27oC with orbital shaking (250 RPM). After 12 hours, reactions were terminated by extraction with DCM (3 x 500 mL). The organic fractions were combined and concentrated in vacuo and purified by flash chromatography (silica gel, 10:90 chloroform/methanol, Rf = 0.3) to afford the product as a yellow solid (67.1 mg, 63%). Product identity was confirmed by NMR and comparison with the synthetic standard on LC-MS using non-chiral method A.

Stereoselectivity was determined to match that obtained with the general enzymatic transformation assay by re-injection using chiral method A.

165

Chapter 8

8.2 Assay of Diels-Alderase activity of CYP2E1

8.2.1 Synthesis of authentic standards of Diels-Alder products

To confirm the identity of Diels-Alder reaction products resulting from the combination of each diene and dienophile, authentic standards were synthesized using standard conditions.

GC-MS analysis was used to compare the enzymatic and chemical reaction products. In each case, the diene and dienophile (1 mmol each) were combined in DCM (3 mL) containing a catalytic amount of AlCl3 (2 mol%). The reaction mixtures were stirred at room temperature for

1 hour before filtration through a plug of silica and evaporation to dryness in vacuo. Yields were excellent and the endo:exo ratio was in excess of 90% in all cases.

166

Chapter 8

Table 8.2: Diels-Alder 4+2 cycloaddition reactions of test dienes and dienophiles.

8.2.2 GC-MS assay of Diels-Alderase activity of CYP2E1

Dienes (1 to 10 mM) and dienophiles (1 to 10 mM) were combined with purified

CYP2E1 (50 nmol) in potassium phosphate buffer (500 μL, 0.1 M at pH 7.4). The mixtures were shaken at 250 RPM and 37oC for up to 24 hours. Each reaction was terminated by the addition of chloroform (2 x 1 mL). Acetophenone (1 mM) was added as an internal standard and the mixture was shaken vigorously. An aliquot of the organic layer was removed and analyzed by GC-MS.

167

Chapter 8

Products were injected on an Agilent J&W GC Column (HP-5MS UI, 30 m length, 0.250 mm diameter, 0.25 μm film). Upon injection, the temperature was increased from 100oC to

350oC at a constant rate over 25 minutes. Product formation was quantified by area in relation to the peak at 3.40 minutes corresponding to acetophenone, the internal standard.

8.2.3 UV assay of Diels-Alderase activity of CYP2E1

Dienes (1 mM) and dienophiles (1 mM) were combined with purified CYP2E1 (50 nM) in potassium phosphate buffer (0.1 M at pH 7.4) and the temperature was maintained at 19.5oC.

Reactions were monitored by the disappearance of substrates by UV at 300 nm, a convenient wavelength close to the lambda max values of 2 and 4.

8.3 Synthetic protocols

General protocol for synthesis of compounds 4.15, 4.18, 5.3-5.8, 5.11-5.21, 5.30-5.33, 5.35-

5.38, 5.42, 5.50, 5.55, 5.66, 5.73, and 5.74 NaH (11 mmol, 60% in mineral oil) was washed with hexanes (2  8 mL), dried under nitrogen, and suspended in DMSO (30 mL). The NaH solution was heated to 60oC with stirring before addition of theobromine (5.1, 3.1 g, 11 mmol), fluorescein (4.14, 3.7 g, 11 mmol), or theophylline (4.17, 3.1 g, 11 mmol) as a solid in a single addition. After 20 minutes, the desired alkylhalide or tosylated alcohol (11 mmol) was added in a single addition, and the reaction was allowed to proceed overnight before cooling to room temperature and quenching with water (2 mL). After concentration of the solution in vacuo, the resulting solid was re-suspended in CHCl3 (50 mL). This solution was washed with saturated ammonium chloride (20 mL) then brine (20 mL) and concentrated, to afford the desired product.

168

Chapter 8

3'-oxo-3'H,10H-spiro[anthracene-9,1'-isobenzofuran]-3,6-diyl dipalmitate (4.15): Yield 7.0 g of known yellowish solid, 80%. λmax (MeCN) 273 nm; HPLC RT = 18.5 min (non-chiral

1 method B); H NMR (300 MHz, CDCl3) δ 6.6-7.7 (m, 10H, H-1,2,4,10,11,12,13), 3.94 (s, 2H, H-

7), 2.35 (t, J = 7.2, 4H, H-17), 1.55-1.63 (m, 4H, H-18), 1.30-1.42 (m, 48H, H-

13 19,20,21,22,23,24,25,26,27,28,29,30), 0.95 (t, J = 7.2, 6H, C-31); C NMR (300 MHz, CDCl3)

δ 175.3, 168.6, 152.7, 149.4, 140.3, 138.8, 133.5, 125.8, 126.2, 125.6, 123.1, 124.6, 119.9, 119.4,

91.6, 41.6, 35.5, 30.1, 29.8, 29.6, 29.4, 29.1, 28.5, 27.8, 28.8, 27.5, 26.5, 25.2, 23.9, 16.4, 14.8;

HRMS (EI): Calcd. For C53H74O6 (M+): 806.54854, Found: 806.55127.

1,3-Dimethyl-7-pentyl-1H-purine-2,6(3H,7H)-dione (4.18): Yield

2.2 g of yellowish solid, 80%. λmax (MeCN) 273 nm; HPLC RT = 2.9

1 min (non-chiral method B); H NMR (300 MHz, CDCl3) δ 7.59 (s,

1H, H-4), 4.27 (t, J = 7.2, 2H, H-8), 3.59 (s, 3H, H-2), 3.41 (s, 3H, H-

7), 1.86 (p, J = 7.2, 2H, H-9), 1.35-1.26 (m, 4H, H-10,11), 0.87 (t, 3H, J = 6.6, C-12); 13C NMR

(300 MHz, CDCl3) δ 156.2 (C-6), 150.3 (C-3), 146.7 (C-1), 142.8 (C-4), 104.3 (C-5), 49.5 (C-8),

31.6 (C-2), 29.6 (C-7), 28.7 (C-8), 28.1 (C-10), 25.2 (C-9), 21.3 (C-11), 14.2 (C-12); HRMS

(EI): Calcd. For C12H18O2N4 (M+): 250.14298, Found: 250.14325. 169

Chapter 8

2-Pentylisoindoline-1,3-dione (4.21): Pentane-1-amine (78 μL, 0.68

mmol) and phthalic anhydride (100 mg, 0.68 mmol) were combined

in chloroform (5 mL) and stirred for 1 hour at room temperature before the addition of TFA (57 μL, 0.75 mmol) and stirring at reflux for 3 hours. The mixture was washed with water (3 x 5 mL) and filtered through Celite before purification by silica gel chromatography (hexanes/EtOAc, 1:1, Rf = 0.3) affording the product as a white solid (100 mg,

1 70%). λmax (MeCN) 273 nm; HPLC RT = 18.5 min (non-chiral method A); H NMR (300 MHz,

CDCl3) δ 7.80-7.83 (m, 2H, H-3), 7.66-7.69 (m, 2H, H-4), 3.65 (t, J = 7.2, 2H, H-5), 1.61-1.71

13 (m, 2H, H-6), 1.27-1.35 (m, 4H, H-7,8), 0.87 (t, J = 7.2, 3H, H-9); C NMR (300 MHz, CDCl3)

δ 168.3 (C-1), 133.7 (C-2), 131.2 (C-3), 123.0 (C-4), 37.9 (C-5), 28.8 (C-6), 28.2 (C-7), 22.2 (C-

8), 13.9 (C-9); HRMS (ESI): Calcd. for C13H15NO2 (M+1): 218.11756; Found: 218.11719.

2-(4-Hydroxypentyl)isoindoline-1,3-dione (4.22): Was prepared

using CYP3A4-containing membranes as described in section 8.1.2.2

for protocol. λmax (MeCN) 273 nm; HPLC RT = 8.3 min (non-chiral

1 method A); H NMR (300 MHz, CDCl3) δ 7.83-7.85 (m, 2H, H-3), 7.70-7.72 (m, 2H, H-4), 3.85

(m, 1H, H-8), 3.72 (t, J = 7.0, 2H, H-5), 1.71-1.82 (m, 2H, H-7), 1.47-1.52 (m, 2H, H-6), 1.19 (d,

J = 6.0, 3H, H-9); LC-MS (ESI): Calcd. for C13H15NO3 (M+1): 233.1; Found: 233.1.

170

Chapter 8

3,7-Dimethyl-1-propyl-1H-purine-2,6-(3H,7H)-dione (5.3): Yield 2.3

159 g of known yellowish solid , 94%. λmax (MeCN) 273 nm; HPLC RT =

1 14.9 min (non-chiral method A); H NMR (400 MHz, CDCl3) δ 7.59 (s,

1H, H-4), 3.94-3.99 (m, 5H, H-5,8), 3.58 (s, 3H, H-2), 1.62-1.71 (m, 2H, H-9), 0.95 (t, J = 8.0,

13 3H, H-10); C NMR (300 MHz, CDCl3) δ 155.2 (C-7), 151.3 (C-3), 148.5 (C-1), 141.2 (C-4),

107.5 (C-6), 42.7 (C-8), 33.5 (C-5), 29.6 (C-2), 21.2 (C-9), 11.2 (C-10); HRMS (EI): Calcd. For

C10H15O2N4 (M+): 223.11895, Found: 223.11898.

1-Butyl-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione (5.4): Yield

160 2.2 g of known yellowish solid , 84%. λmax (MeCN) 273 nm;

HPLC RT = 11.6 min (non-chiral method A); 1H NMR (300 MHz,

CDCl3) δ 7.49 (s, 1H, H-4) 3.96-4.01 (m, 5H, H-5,8), 3.56 (s, 3H, H-2), 1.47-1.67 (m, 2H, H-9),

13 1.32-1.44 (m, 2H, H-10), 0.93 (t, J = 7.5, 3H, H-11); C NMR (300 MHz, CDCl3) δ 155.6 (C-7),

151.7 (C-3), 143.9 (C-1), 141.6 (C-4), 107.9 (C-6), 41.5 (C-8), 33.8 (C-5), 30.4 (C-2), 29.9 (C-

9), 20.4 (C-10), 14.0 (C-11); HRMS (EI): Calcd. for C11H16N4O2 (M+): 236.12733, Found:

236.12701.

3,7-Dimethyl-1-pentyl-1H-purine-2,6(3H,7H)-dione (5.5):

Recovered yield = 2.3 g of the known yellowish solid160, 84%.

λmax (MeCN) 273 nm; HPLC RT = 14.4 min (non-chiral method

1 A); H NMR (400 MHz, CDCl3) δ 7.49 (s, 1H, H-4), 3.95-4.00 (m, 5H, H-5,8), 3.56 (s, 3H, H-2),

1.59-1.68 (m, 2H, H-9), 1.31-1.38 (m, 4H, H-10,11), 0.88 (t, J = 6.9, 3H, H-12); 13C NMR (300 171

Chapter 8

MHz, CDCl3) δ 155.5 (C-7), 151.7 (C-3), 148.9 (C-1), 141.5 (C-4), 107.9 (C-6), 41.7 (C-8), 33.8

(C-5), 29.9 (C-2), 29.3 (C-9), 28.0 (C-10), 22.7 (C-11), 14.2 (C-12); HRMS (EI): Calcd. for

C12H18N4O2 (M+): 250.14298; Found: 250.14265.

1-Hexyl-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione (5.6):

Recovered yield = 2.6 g of yellowish solid, 89%. λmax

(MeCN) 273 nm; HPLC RT = 18.0 min (non-chiral method

1 A); H NMR (300 MHz, CDCl3) 8.00 (s, 1H, C-4), 3.85 (s, 3H, H-5), 3.79-3.83 (m, 2H, H-8),

3.38 (s, 3H, H-2), 1.44-1.54 (m, 2H, H-9), 1.24 (s, 6H, H-10,11,12), δ 0.83 (t, J = 6.40, 3H, H-

13 13); C NMR (300 MHz, CDCl3) δ 155.2 (C-7), 151.4 (C-3), 148.5 (C-1), 141.3 (C-4), 107.6

(C-6), 40.9 (C-8), 33.6 (C-5), 31.4 (C-2), 29.6 (C-9), 27.9 (C-10), 26.6 (C-11), 22.5 (C-12), 14.0

(C-13); HRMS (EI): Calcd. for C13H20N4O2 (M+): 264.15863; Found: 264.15892.

1-Heptyl-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione

(5.7): Recovered yield = 2.4 g of yellowish solid, 80%. λmax

(MeCN) 273 nm; HPLC RT = 19.2 min (non-chiral method

1 A); H NMR (300 MHz, CDCl3) δ 7.50 (s, 1H, H-4), 3.95-4.01 (m, 5H, H-5,8), 3.57 (s, 3H, H-2),

1.61-1.68 (m, 2H, H-9), 1.22-1.38 (m, 8H, H-10-13), 0.86 (t, J = 6.9, 3H, H-14); 13C NMR (300

MHz, CDCl3) δ 155.3, (C-7), 151.4 (C-3), 148.6 (C-1), 141.3 (C-4), 107.7 (C-6), 41.4 (C-8),

33.6 (C-5), 28.9 (C-2), 28.0 (C-9), 26.9 (C-10), 22.5 (C-11), 21.2 (C-12), 20.8 (C-13), 14.0 (C-

14); HRMS (EI): Calcd. for C14H22N4O2 (M+): 278.17428; Found: 278.27466.

172

Chapter 8

1-Octyl-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione

(5.9): Recovered yield = 2.6 g of yellowish solid, 80%.

λmax (MeCN) 273 nm; HPLC RT = 19.7 min (non-chiral

1 method A); H NMR (300 MHz, CDCl3) δ 7.48 (s, 1H, H-4), 3.94-3.99 (m, 5H, H-5,8), 3.55 (s,

3H, H-2), 1.57-1.67 (m, 2H, H-9), 1.18-1.36 (m, 10H, H-10,11,12,13,14), 0.84 (t, J = 6.9, 3H, H-

13 15); C NMR (300 MHz, CDCl3) δ 155.6 (C-7), 151.7 (C-3), 148.9 (C-1), 141.5 (C-4), 41.2 (C-

6), 33.8 (C-8), 32.0 (C-5), 29.9 (C-2), 29.5 (C-9), 28.3 (C-10), 27.2 (C-11), 22.9 (C-12), 22.0 (C-

13), 21.2 (C-14), 14.3 (C-15); HRMS (EI): Calcd. for C15H24N4O2 (M+): 292.18993; Found:

292.18945.

3,7-Dimethyl-1-(2-methylpentyl)-1H-purine-2,6(3H,7H)-dione

(5.9): 2-Methylpentan-1-ol (1.0 g, 10 mmol) was combined with

4-toluenesulfonyl chloride (1.9 g, 10 mmol, 1.0 equiv.) in pyridine (10 mL) at 0oC and allowed to warm to room temperature with stirring. The solution was stirred for an extra 3 hours before concentration of the solution in vacuo. The residue was dissolved in ethyl acetate (50 mL) before washing with brine (50 mL) and saturated ammonium chloride (50 mL). The organic layer was concentrated in vacuo. The resulting solid was dissolved in minimal DMSO (30 mL) and added to a mixture of NaH (256 mg, 6 mmol) and theobromine (5.1, 1.2 g, 6 mmol) at 600C. The reaction was allowed to proceed overnight before cooling to room temperature and quenching with water (2 mL). After concentration of the solution in vacuo, the resulting solid was dissolved in CHCl3 (50 mL). This solution was washed

173

Chapter 8 with saturated ammonium chloride (20 mL) followed by brine (20 mL) and concentrated, to afford the desired product. Recovered yield = 1.4 g of yellowish solid, 54%. λmax (MeCN) 273

1 nm; HPLC RT = 13.5 min (non-chiral method A); H NMR (400 MHz, CDCl3) δ 7.50 (s, 1H, H-

4), 3.98 (s, 3H, H-5), 3.86 (d, J = 7.2, 2H, H-8), 3.56 (s, 3H, H-2), 1.98-2.07 (m, 1H, H-9), 1.38-

1.48 (m, 1H, H-10a), 1.23-1.37 (m, 2H, H-10b), 1.11-1.21 (m, 2H, H-11), 0.86-0.90 (m, 6H, H-

13 12,13); C NMR (300 MHz, CDCl3) δ 155.5 (C-7), 151.6 (C-3), 148.6 (C-1), 141.4 (C-4), 107.4

(C-6), 47.0 (C-8), 36.6 (C-5), 33.5 (C-2), 31.6 (C-9), 29.6 (C-10), 19.7 (C-11), 17.2 (C-13), 14.2

(C-12); HRMS (EI): Calcd. For C13H20O2N4 (M+): 264.15863; Found: 264.15798.

3,7-Dimethyl-1-(3-methylpentyl)-1H-purine-2,6(3H,7H)-dione

(5.10): 3-Methylpentan-1-ol (1.0 g, 10 mmol) was combined with

4-toluenesulfonyl chloride (1.9 g, 10 mmol, 1.0 equiv.) in

pyridine (10 mL) at 0oC and allowed to warm to room temperature with stirring. The solution was stirred for 3 extra hours before concentration of the solution in vacuo. The residue was redissolved in ethyl acetate (50 mL) before washing with brine (50 mL) and saturated ammonium chloride (50 mL). The organic layer was separated and concentrated in vacuo. The resulting solid was dissolved in minimal DMSO (30 mL) and added to a mixture of NaH (164 mg, 4 mmol) and theobromine (5.1, 738 mg, 4 mmol) at 60oC. The reaction was allowed to proceed overnight before cooling to room temperature and quenching with water (2 mL). After concentration of the solution in vacuo, the resulting solid was dissolved in CHCl3 (50 mL). This solution was washed with saturated ammonium chloride (20 mL) followed by brine (20 mL) and concentrated, to afford the desired product. Recovered yield =

174

Chapter 8

1.08 g of yellowish solid, 41%. λmax (MeCN) 273 nm; HPLC RT = 13.7 min (non-chiral method

1 A); H NMR (400 MHz, CDCl3) δ 7.48 (s, 1H, H-4), 3.97-4.02 (m, 5H, H-5,8), 3.56 (s, 3H, H-2),

1.59-1.66 (m, 1H, H-10), 1.36-1.49 (m, 3H, H-9,11a), 1.15-1.24 (m, 1H, H-11b), 0.96 (d, J = 6.0,

13 3H, H-13), 0.87 (t, J = 7.2, 3H, H-12); C NMR (300 MHz, CDCl3) δ 155.2 (C-7), 151.4 (C-3),

148.6 (C-1), 141.3 (C-4), 107.6 (C-6), 39.9 (C-8), 34.4 (C-5), 33.5 (C-2), 32.6 (C-10), 29.6 (C-

9), 29.3 (C-11), 19.0 (C-13), 11.2 (C-12); HRMS (EI): Calcd. For C13H20O2N4 (M+): 264.15863;

Found: 264.15806.

3,7-Dimethyl-1-(4-methylpentyl)-1H-purine-2,6(3H,7H)-dione

(5.11): Recovered yield = 2.4 g of yellowish solid, 82%. λmax

(MeCN) 273 nm; HPLC RT = 13.8 min (non-chiral Method A);

1 H NMR (400 MHz, CDCl3) δ 7.52 (s, 1H, H-4), 3.98 (s, 3H, H-5), 3.93-3.97 (m, 2H, H-8), 3.68

(s, 3H, H-2), 1.51-1.69 (m, 3H, H-9,11), 1.21-1.28 (m, 2H, H-10), 0.87 (d, J = 6.0, 6H, H-12);

13 C NMR (300 MHz, CDCl3) δ 155.3 (C-7), 151.4 (C-3), 148.5 (C-1), 141.2 (C-4), 107.7 (C-6),

41.7 (C-8), 35.9 (C-5), 33.6 (C-2), 29.7 (C-11), 27.8 (C-9), 25.9 (C-10), 22.6 (C-12); HRMS

(EI): Calcd. For C13H19N4O2 (M+): 263.15025; Found: 263.15031.

3,7-Dimethyl-1-(5-methylhexyl)-1H-purine-2,6(3H,7H)-

dione (5.12): Recovered yield = 2.4 g of yellowish solid,

80%. λmax (MeCN) 273 nm; HPLC RT = 15.1 min (non-chiral

1 method A); H NMR (800 MHz, CDCl3) δ 7.50 (s, 1H, H-4),

3.99-4.01 (m, 5H, H-5,8), 3.58 (s, 3H, H-2), 1.64 (p, J = 8.0, 2H, H-9), 1.52-1.57 (m, 1H, H-12), 175

Chapter 8

1.32-1.41 (m, 2H, H-10), 1.18-2.17 (m, 2H, H-11), 1.87 (d, J = 7.2, 6H, H-13); 13C NMR (300

MHz, CDCl3) δ 156.4 (C-7), 149.6 (C-3), 145.1 (C-1), 142.5 (C-4), 108.6 (C-6), 41.2 (C-8), 39.9

(C-5), 34.4 (C-2), 30.7 (C-12), 30.1 (C-9), 22.9 (C-10), 28.1 (C-11), 23.5 (C-13) ; HRMS (EI):

Calcd. for C14H22N4O2 (M+): 278.17428; Found: 278.17398.

1-(Cyclopropylmethyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione

(5.13): Recovered yield = 2.0 g of yellowish solid, 80%. λmax (MeCN)

273 nm; HPLC RT = 12.7 min (non-chiral method A); 1H NMR (300

MHz, CDCl3) δ 7.42 (s, 1H, H-4), 3.84 (s, 3H, H-5), 3.70 (d, J = 7.2, 2H, H-8), 3.39 (s, 3H, H-2),

13 1.04-1.13 (m, 1H, H-9), 0.24-0.307 (m, 4H, H-10); C NMR (300 MHz, CDCl3) δ 155.2 (C-7),

151.5 (C-3), 148.5 (C-1), 141.4 (C-4), 107.5 (C-6), 45.5 (C-8), 33.4 (C-5), 29.5 (C-2), 9.9 (C-9),

3.7 (C-10); HRMS (ESI): Calcd. for C11H15N4O2 (M+1): 235.11895; Found: 235.11862.

1-Octyl-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione (5.14):

Recovered yield = 2.2 g of yellowish solid, 80%. λmax (MeCN) 273

nm; HPLC RT = 13.1 min (non-chiral method A); 1H NMR (300

MHz, CDCl3) δ 7.38 (s, 1H, H-4), 3.79-3.82 (m, 5H, H-5,8), 3.32 (s, 3H, H-2), 2.46-2.24 (m, 1H,

13 H-9), 1.70-1.80 (m, 2H, H-10), 1.56-1.65 (m, 4H, H-11); C NMR (300 MHz, CDCl3) δ 155.2

(C-7), 151.4 (C-3), 148.4 (C-1), 141.4 (C-4), 107.3 (C-6), 45.63 (C-8), 29.4 (C-9), 26.0 (C-10),

18.1 (C-11); HRMS (ESI): Calcd. for C12H17N4O2 (M+1): 249.13460; Found: 249.13446.

176

Chapter 8

3,7-Dimethyl-1-((tetrahydrofuran-2-yl)methyl)-1H-purine-

2,6(3H,7H)-dione (5.15): Recovered yield = 2.4 g of yellowish

solid, 81%. λmax (MeCN) 273 nm; HPLC RT = 6.2 min (non-chiral

1 method A); H NMR (400 MHz, CDCl3) δ 7.49 (s, 1H, H-4), 4.25-4.38 (m, 2H, H-8), 3.92-3.98

(m, 4H, H-5, H-9a), 3.85-3.89 (m, 1H, H-12a), 3.71-3.76 (m, 1H, H-12b), 3.57 (s, 3H, H-2),

1.97-2.08 (m, 2H, H-11), 1.84-1.94 (m, 1H, H-10a), 1.67-1.72 (m, 1H, H-10b); 13C NMR (300

MHz, CDCl3) δ 155.4 (C-7), 151.6 (C-3), 148.8 (C-1), 141.4 (C-4), 107.7 (C-6), 76.0 (C-8), 67.8

(C-9), 44.5 (C-12), 33.6 (C-5), 29.7 (C-2), 29.2 (C-10), 25.2 (C-11); HRMS (EI): Calcd. For

C12H16O2N4 (M+): 264.12224; Found: 264.12178.

1-(Cyclohexylmethyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione

(5.16): Recovered yield = 3.4 g of yellowish solid, 89%. λmax

(MeCN) 273 nm; HPLC RT = 17.0 min (non-chiral method A); 1H

NMR (400 MHz, CDCl3) δ 7.49 (s, 1H, H-4), 3.97 (s, 3H, H-5), 3.84 (d, J = 7.2, 2H, H-8), 3.36

(s, 3H, H-2), 2.15 (m, 1H, H-9), 1.62-1.71 (m, 4H, H-9,10a,11a,11b), 1.02-1.21 (m, 6H, H-

13 10b,12a,12b); C NMR (300 MHz, CDCl3) δ 155.5 (C-7), 151.7 (C-3), 148.6 (C-1), 141.42 (C-

4), 107.5 (C-6), 47.0 (C-8), 40.9 (C-9), 36.5 (C-5), 33.5 (C-2), 30.7 (C-10), 29.6 (C-11), 25.7 (C-

12); HRMS (EI): Calcd. for C14H20N4O2 (M+): 276.15863; Found: 276.15881.

177

Chapter 8

1-(2-Cyclopropylethyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-

dione (5.17): Recovered yield = 2.3 g of yellowish solid, 85%.

λmax (MeCN) 273 nm; HPLC RT = 16.5 min (non-chiral method

1 A); H NMR (800 MHz, CDCl3) δ 7.47 (s, 1H, H-4), 4.09 (t, J =

7.2, 2H, H-8), 3.97 (s, 3H, H-5), 3.56 (s, 3H, H-2), 1.54 (dt, J1 = 7.2, J2 = 7.8, 2H, H-9) 0.71-

13 0.75 (m, 1H, H-10), 0.40-0.43 (m, 4H, H-11) ; C NMR (300 MHz, CDCl3) δ 156.7 (C-7), 148.1

(C-3), 145.3 (C-1), 134.1 (C-4), 108.0 (C-6), 41.3 (C-8), 35.0 (C-5), 33.8 (C-2), 21.6 (C-9), 7.2

(C-10), 4.1 (C-11); HRMS (ESI): Calcd. for C12H17N4O2 (M+1): 249.13460; Found: 249.13411.

1-Benzyl-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione (5.18):

Recovered yield = 2.4 g of yellowish solid, 82%. λmax (MeCN) 273

nm; HPLC RT = 11.8 min (non-chiral method A); 1H NMR (300

MHz, CDCl3) δ 7.11-7.41 (m, 6H, H-4,10,11,12), 5.09 (s, 2H, H-8), 3.83 (s, 3H, H-5), 3.46 (s,

13 3H, H-2); C NMR (300 MHz, CDCl3) δ 155.1 (C-7), 151.4 (C-3), 148.6 (C-1), 141.7 (C-4),

137.3 (C-9), 130.5 C-11), 128.6 (C-10), 128.3 (C-12), 107.5 (C-6), 44.31 (C-8), 33.5 (C-5), 29.7

(C-2); HRMS (EI): Calcd. for C14H14N4O2 (M+): 270.11168; Found: 270.11129.

3,7-Dimethyl-1-(4-nitrobenzyl)-1H-purine-2,6(3H,7H)-dione

(5.19): Recovered yield = 2.7 g of yellowish solid, 78%. λmax

(MeCN) 273 nm; HPLC RT = 9.2 min (non-chiral method A);

1 H NMR (300 MHz, CDCl3) δ 8.15 (d, J = 6.9, 2H, H-11), 7.61 (d, J = 6.9, 2H, H-10), 7.54 (s,

13 1H, H-4), 5.26 (s, 2H, H-8), 3.99 (s, 3H, H-5), 3.58 (s, 3H, H-2); C NMR (300 MHz, CDCl3) δ 178

Chapter 8

155.1 (C-7), 151.0 (C-3), 148.3 (C-1), 144.5 (C-12), 143.3 (C-4), 129.5 (C-9), 123.7 (C-11),

113.6 (C-10), 107.2 (C-6), 43.8 (C-8), 33.7 (C-5), 29.9 (C-2); HRMS (EI): Calcd. for

C14H13N5O4 (M+): 315.09675; Found: 315.09626.

1-(4-Methoxybenzyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-

dione (5.20): Recovered yield = 2.5 g of yellowish solid, 75%.

λmax (MeCN) 273 nm; HPLC RT = 9.1 min (non-chiral method

1 A); H NMR (300 MHz, CDCl3) δ 7.45-7.49 (m, 3H, H-4,10), 6.81-6.84 (m, 2H, H-11), 5.12 (s,

13 2H, H-8), 3.90 (s, 3H, H-15), 3.78 (s, 3H, H-5), 3.50 (s, 3H, H-2); C NMR (300 MHz, CDCl3)

δ 160.3 (C-12), 155.6 (C-7), 152.0 (C-3), 148.3 (C-1), 143.3 (C-4), 131.1 (C-9), 126.6 (C-10),

114.2 (C-11), 106.2 (C-6), 56.3 (C-15), 43.2 (C-8), 33.2 (C-5), 29.4 (C-2); HRMS (ESI): Calcd. for C15H17N4O2 (M+1): 301.12952; Found: 301.12927.

1-(3-Methoxybenzyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-

dione (5.21): Recovered yield = 2.5 g of yellowish solid, 75%.

λmax (MeCN) 273 nm; HPLC RT = 8.9 min (non-chiral method

1 A); H NMR (400 MHz, CDCl3) δ 7.51 (s, 1H, H-4), 7.19-7.24 (m, 1H, H-13), 6.99-7.07 (m, 2H,

H-10,12), 6.75-6.88 (m, 1H, H-14), 5.16 (s, 2H, H-8), 3.98 (s, 3H, H-5), 3.77 (s, 3H, H-15), 3.56

13 (s, 3H, H-2); C NMR (300 MHz, CDCl3) δ 160.9 (C-11), 154.8 (C-7), 152.0 (C-3), 148.1 (C-1),

146.3 (C-9), 143.3 (C-4), 142.2 (C-10), 131.1 (C-12), 113.8 (C-13), 112.1 (C-14), 109.5 (C-6),

60.1 (C-15), 47.5 (C-8), 35.5 (C-5), 31.6 (C-2); HRMS (ESI): Calcd. for C15H17N4O3 (M+1):

301.12952; Found: 301.12914. 179

Chapter 8

1-(3-Hydroxypropyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione

(5.22): Compound 5.42 (2.4 g, 6.8 mmol) was dissolved in THF

(10 mL) and tetra-N-butylammonium fluoride (2.61 g, 10 mmol,

1.5 equiv.) was added drop-wise. The reaction progress was monitored by TLC and upon completion (12 hours), the reaction was terminated by the addition of saturated aqueous NaHCO3

(20 mL). The organic layer was separated, washed with brine (20 mL) and filtered through Celite before being concentrated in vacuo to afford the product as a slightly yellow oil (1.37 g, 90%).

1 H NMR (300 MHz, CDCl3) δ 7.49 (s, 1H, H-4), 4.02 (t, J = 7.2, 2H, H-8), 3.98 (s, 3H, H-5),

3.65 (t, J = 6.8, 2H, H-10), 3.49 (s, 3H, H-2), 1.38-1.48 (m, 2H, H-9); 13C NMR (300 MHz,

CDCl3) δ 155.8 (C-7), 151.9 (C-3), 148.9 (C-1), 141.8 (C-4), 107.5 (C-6), 58.5 (C-10), 37.7 (C-

5), 33.7 (C-8), 30.8 (C-2), 29.9 (C-9); HRMS (ESI): Calcd. for C10H15N4O3 (M+1): 239.11387;

Found: 239.11335.

3-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-1-

yl)propanal (5.23): Compound 5.22 (1.37 g, 6.1 mmol) was

dissolved in DCM (10 mL) and Dess-Martin periodinane (2.85 g, 6.7 mmol, 1.1 equiv.) was added as a solid. The solution was stirred for 2 hours before termination by the addition of saturated aqueous NaHCO3 (5 mL) and saturated aqueous Na2S2O3 (5 mL) followed by stirring at 40oC for 30 minutes. The organic layer was separarated and washed with brine (10 mL) before filtration though Celite and concentration in vacuo, affording the product as

1 a slightly yellow oil (1.23 g, 90 %). H NMR (500 MHz, CDCl3) δ 9.83 (s, 1H, H-10), 7.51 (s,

180

Chapter 8

1H, H-4), 4.38 (t, J = 7.0, 2H, H-8), 3.97 (s, 3H, H-5), 3.57 (s, 3H, H-2), 2.78 (t, J = 7.0, 2H, H-

13 9); C NMR (300 MHz, CDCl3) δ 200.2 (C-10), 156.0 (C-7), 149.1 (C-3), 144.9 (C-1), 143.6

(C-4), 107.7 (C-6), 39.9 (C-8), 39.5 (C-5), 33.3 (C-2), 30.0 (C-9); HRMS (ESI): Calcd. for

C10H13N4O3 (M+1): 237.09877; Found: 237.09983.

3,7-Dimethyl-1-(3-oxopentyl)-1H-purine-2,6(3H,7H)-dione

(5.25): Compound 5.26 (100 mg, 0.4 mmol) was dissolved in

DCM (5 mL) and Dess-Martin periodinane (177 mg, 0.5 mmol,

1.1 equiv.) was added as a solid. The solution was stirred for 2 hours before termination by the addition of saturated aqueous NaHCO3 (5 mL) and saturated aqueous Na2S2O3 (5 mL) followed by stirring at 40oC for 30 minutes. The organic layer was separated and washed with brine (5 mL) before filtration though Celite and concentration in vacuo, affording the product as a slightly yellow oil (90 mg, 90 %). λmax (MeCN) 273 nm; HPLC RT = 6.6 min (non-chiral

1 method A); H NMR (300 MHz, CDCl3): δ 7.50 (s, 1H, H-4), 4.23-4.32 (t, J = 7.5, 2H, H-8),

3.98 (s, 3H, H-5), 3.56 (s, 3H, H-2), 2.79 (t, J = 7.5, 2H, H-9), 2.47 (q, J = 7.5, 2H, H-11), 1.06

13 (t, J = 7.5, 3H, H-12); C NMR (300 MHz, CDCl3) δ 210.8 (C-10), 156.6 (C-7), 148.1 (C-3),

145.2 (C-1), 143.1 (C-4), 108.2 (C-6), 42.0 (C-8), 39.8 (C-5), 35.1 (C-2), 34.7 (C-9), 30.7 (C-

11), 8.5 (C-12); HRMS (ESI): Calcd. for C12H17N4O3 (M+1): 265.12952; Found: 265.12902.

1-(3-Hydroxypentyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-

dione (5.26): Compound 5.23 (200 mg, 0.75 mmol) was

181

Chapter 8

o dissolved in dry THF (10 mL) and cooled to -78 C under N2. A solution of methylmagnesium bromide was added dropwise (800 µL, 3.0 M in ether, 3 equiv.) over 20 minutes. The reaction was terminated by the addition of aqueous ammonium chloride (0.5 mL). The solution was filtered through Celite, dried over anhydrous sodium sulfate, and concentrated in vacuo before purification by preparative TLC (chloroform:methanol 9:1, Rf = 0.30) to afford the product as a slightly yellow solid (130 mg, 60%). λmax (MeCN) 273 nm; HPLC RT = 5.4 min (non-chiral

1 method A); H NMR (300 MHz, CDCl3): δ 7.53 (s, 1H, H-4), 4.09-4.30 (m, 2H, H-8), 4.00 (s,

3H, H-5), 3.73-3.76 (m, 1H, H-10), 3.60 (s, 3H, H-2), 1.58-1.65 (m, 2H, H-9), 1.42-1.53 (m, 2H,

13 H-11), 0.93 (t, J = 7.5, 3H, H-12); C NMR (300 MHz, CDCl3) δ 156.2, (C-7), 148.2 (C-3),

144.8 (C-1), 143.9 (C-4), 108.7 (C-6), 71.1 (C-10), 37.4 (C-8), 34.4 (C-5), 34.0 (C-2), 30.2 (C-

9), 25.6 (C-11), 10.6 (C-12); HRMS (ESI): Calcd. for C12H19N4O3 (M+1): 267.14517; Found:

267.14461.

1-Octyl-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione (5.27):

Recovered yield = 2.2 g of yellowish solid, 77%. λmax (MeCN)

273 nm; HPLC RT = 6.4 min (non-chiral method A); 1H NMR

(200 MHz, CDCl3) δ 7.51 (s, 1H, H-4), 4.03 (t, J = 7.2, 2H, H-8), 3.99 (s, 3H, H-5), 3.57 (s, 3H,

H-2), 2.51 (t, J = 7.2, 2H, H-10), 2.14 (s, 3H, H-12), 1.88-2.03 (m, 2H, H-9); 13C NMR (300

MHz, CDCl3) δ 172.8 (C-11), 155.3 (C-3), 151.4 (C-1), 141.4 (C-4), 107.6 (C-6), 41.7 (C-8),

40.5 (C-10), 34.9 (C-5), 33.6 (C-2), 30.9 (C-12), 22.1 (C-9); HRMS (ESI): Calcd. for

C12H17N4O3 (M+1): 265.13007; Found: 265.12952.

182

Chapter 8

1-(5-Hydroxypentyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-

dione (5.28): Compound 5.32 (200 mg, 0.75 mmol) was

dissolved in THF (5 mL) and added dropwise to a 0oC solution of lithium aluminium hydride (86 mg, 19.2 mmol, 3 equiv.) in THF (10 mL). The solution was allowed to warm to room temperature and the progress was monitored by TLC.

Upon completion (~3 hours), the reaction was terminated by the slow, dropwise addition of saturated aqueous NaOH (200 µL). The resulting yellow liquid was filtered through Celite and dried over anhydrous sodium sulfate before being concentrated in vacuo to afford the product as a slightly yellow solid (110 mg, 64 %). λmax (MeCN) 273 nm; HPLC RT = 5.1 min (non-chiral

1 method A); H NMR (300 MHz, CDCl3) δ 7.49 (s, 1H, H-4), 3.95-3.99 (m, 5H, H-5,8), 3.6 (t, J

= 6.3, 2H, H-12), 3.53 (s, 3H, H-2), 1.55-1.70 (m, 4H, H-9,11), 1.40 (m, 2H, H-10); 13C NMR

(300 MHz, CDCl3) δ 155.3 (C-7), 151.5 (C-3), 148.7 (C-1), 141.3 (C-4), 107.6 (C-6), 62.5 (C-

12), 41.2 (C-8), 33.6 (C-5), 32.3 (C-2), 29.7 (C-11), 27.7 (C-10), 23.0 (C-9); HRMS (ESI):

Calcd. for C12H19N4O3 (M+1): 267.14517; Found: 267.14474.

1-(5-Methoxypentyl)-3,7-dimethyl-1H-purine-

2,6(3H,7H)-dione (5.29): Compound 5.28 (100 mg, 0.4

mmol) was added to a solution of NaH (9 mg, 0.4 mg, 1 equiv.) in THF (5 mL) at 0oC. MeI (25 μL, 0.4 mmol, 1.0 equiv.) was added to this solution dropwise and the mixture was allowed to warm to room temperature and then stirred for 2 hours.

The reaction was terminated by the addition of water and the organic fraction was filtered through Celite and anhydrous sodium sulfate. The organic fraction was evaporated in vacuo to

183

Chapter 8

afford the product as a slightly yellow solid (84 mg, 80 %). λmax (MeCN) 273 nm; HPLC RT =

1 7.4 min (non-chiral method A); H NMR (300 MHz, CDCl3) δ 7.48 (s, 1H, H-4), 3.95-3.99 (m,

5H, H-5, 8), 3.54 (s, 3H, H-2), 3.34 (t, J = 6.6, 2H, H-12), 3.28 (s, 3H, H-13), 1.54-1.69 (m, 4H,

13 H-10,11), 1.34-1.44 (m, 2H, H-9); C NMR (300 MHz, CDCl3) δ 155.2 (C-7), 151.4 (C-3),

148.7 (C-1), 141.4 (C-4), 107.7 (C-6), 72.6 (C-12), 58.5 (C-13), 41.3 (C-8), 33.6 (C-5), 29.7 (C-

2), 29.3 (C-11), 27.9 (C-9), 23.5 (C-10); HRMS (ESI): Calcd. for C13H21N4O3 (M+1):

281.16082; Found: 281.16062.

4-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-

1-yl)butyl acetate (5.30): Recovered yield = 2.7 g of

yellowish solid, 82%. λmax (MeCN) 273 nm; HPLC RT =

1 7.1 min (non-chiral method A); H NMR (400 MHz, CDCl3) δ 7.49 (s, 1H, H-4), 3.96-4.04 (m,

7H, H-5,8,11), 3.54 (s, 3H, H-2), 2.59 (s, 3H, H-13), 1.61-1.70 (m, 4H, H-9,10); 13C NMR (300

MHz, CDCl3) δ 171.1 (C-12), 155.2 (C-7), 151.4 (C-3), 148.7 (C-1), 141.4 (C-4), 107.6 (C-6),

64.3 (C-11), 41.0 (C-8), 33.5 (C-5), 29.6 (C-2), 28.2 (C-13), 27.6 (C-9), 23.3 (C-10); HRMS

(ESI): Calcd. For C13H19N4O4 (M+1): 295.14008; Found: 295.13983.

Ethyl 5-(3,7-dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-

purin-1-yl)pentanoate (5.31): Recovered yield = 2.7 g of

yellowish solid, 81%. λmax (MeCN) 273 nm; HPLC RT =

1 7.6 min (non-chiral method A); H NMR (400 MHz, CDCl3) δ 7.39 (s, 1H, C-4), 3.87 (t, J = 7.2,

2H, H-8), 3.71-3.81 (m, 5H, H-5,13), 3.30 (s, 3H, H-2), 2.10 (t, J = 7.2, 2H, H-11), 1.37-1.51 (m, 184

Chapter 8

13 4H, H-9,10), 1.00 (t, J = 7.2, 3H, H-14); C NMR (300 MHz, CDCl3) δ 173.0 (C-12), 154.9 (C-

7), 151.1 (C-3), 148.4 (C-1), 141.6 (C-4), 107.3 (C-6), 60.0 (C-13), 40.5 (C-8), 33.7 (C-5), 33.3

(C-11), 29.4 (C-2), 27.3 (C-9), 22.0 (C-10), 14.0 (C-14); HRMS (ESI): Calcd. For C14H21O4N4

(M+1): 309.15628; Found: 309.15689.

5-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-

purin-1-yl)pentyl acetate (5.32) Recovered yield = 2.8

g of yellowish solid, 84%. λmax (MeCN) 273 nm; HPLC

1 RT = 7.4 min (non-chiral method A); H NMR (400 MHz, CDCl3) δ 7.52 (s, 1H, H-4), 3.94-4.21

(m, 7H, H-5,8,12), 3.51 (s, 3H, H-2), 2.60 (s, 3H, H-14), 1.55-1.71 (m, 4H, H-9,11), 1.25-1.33

13 (m, 2H, H-10); C NMR (300 MHz, CDCl3) δ 171.1 (C-13), 155.2 (C-7), 151.4 (C-3), 148.7 (C-

1), 141.4 (C-4), 101.6 (C-6), 64.1 (C-11), 41.0 (C-8), 33.5 (C-5), 31.2 (C-2) 29.6 (C-12), 28.2

(C-9), 27.6 (C-11), 23.3 (C-10), 20.9 (C-14); HRMS (ESI): Calcd. For C14H21O4N4 (M+1):

309.15573; Found: 309.15556.

Ethyl 6-(3,7-dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-

1H-purin-1-yl)hexanoate (5.33) Recovered yield = 2.8

g of yellowish solid, 82%. λmax (MeCN) 273 nm; HPLC

1 RT = 7.8 min (non-chiral method A); H NMR (400 MHz, CDCl3) δ 7.41 (s, 1H, H-4), 3.91 (q, J

= 7.2, 2H, H-14), 3.76-3.81 (m, 5H, H-5,8), 3.36 (s, 3H, H-2), 2.11 (t, J = 7.5, 2H, H-12), 1.41-

13 1.53 (m, 6H, H-9,10,11), 1.05 (t, J = 7.2, 3H, H-15); C NMR (300 MHz, CDCl3) δ 173.4 (C-

13), 155.0 (C-7), 151.2 (C-3), 148.5 (C-1), 141.5 (C-4), 107.4, C-6), 59.9 (C-14), 40.9 (C-8), 185

Chapter 8

33.9 (C-5), 33.4 (C-2), 29.5 (C-12), 27.5 (C-9), 26.2 (C-10), 24.5 (C-11), 14.0 (C-15); HRMS

(ESI): Calcd. For C13H21N4O7 (M+1): 323.17193; Found: 323.18031.

1-(3-(Ethylthio)propyl)-3,7-dimethyl-1H-purine-

2,6(3H,7H)-dione (5.34) Compound 5.44 (100 mg, 0.33

mmol), ethanethiol (36 μL, 0.33 mmol), cesium carbonate (107

o mg, 0.33 mmol), and NaH (12 mg, 0.33 mmol) were combined in DMF (20 mL) at 0 C under N2 and stirred for 1 hour. The mixture was allowed to come to room temperature and stirred for an additional 2 hours. Upon completion of the reaction, the mixture was concentrated in vacuo. The residue was redissolved in chloroform and washed with 1 M HCl (50 mL) and brine (50 mL).

The organic fraction was filtered through Celite and concentrated in vacuo, affording the pure product as a slightly yellow oil (75 mg, 80%). λmax (MeCN) 273 nm; HPLC RT = 9.1 min (non-

1 chiral method A); H NMR (400 MHz, CDCl3) δ 7.46 (s, 1H, H-4), 4.00 (t, J = 7.2, 2H, H-8),

3.90 (s, 3H, H-5), 3.47 (s, 3H, H-2), 2.48 (m, 4H, H-10,11) 1.81-1.91 (m, 2H, H-9), 1.15 (t, J =

13 7.5, 3H, H-12); C NMR (300 MHz, CDCl3) δ 155.4 (C-7), 151.4 (C-3), 148.8 (C-1), 141.5 (C-

4), 107.5 (C-6), 40.5 (C-8), 36.4 (C-5), 33.5 (C-2), 29.6 (C-10), 27.7 (C-11), 25.6 (C-9), 14.6 (C-

12); HRMS (ESI): Calcd. For C12H22O2N4S (M+1): 238.12126; Found: 283.12134.

1-(But-3-enyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione (5.35)

Recovered yield = 2.2 g of yellowish solid, 85%. λmax (MeCN) 273

nm; HPLC RT = 7.8 min (non-chiral method A); 1H NMR (400

MHz, CDCl3) δ 7.49 (s, 1H, H-4), 5.75-5.86 (m, 1H, H-10), 4.97-5.09 (m, 2H, H-11), 4.05 (t, J = 186

Chapter 8

7.5, 2H, H-8), 3.96 (s, 3H, H-5), 3.54 (s, 3H, H-2), 2.36-2.43 (m, 2H, H-9); 13C NMR (300 MHz,

CDCl3) δ 155.2 (C-7), 151.4 (C-3), 148.6 (C-1), 141.3 (C-4), 134.9 (C-10), 116.9 (C-11), 107.6

(C-6), 40.9 (C-8), 33.6 (C-5), 32.4 (C-2), 29.7 (C-9); HRMS (EI): Calcd. for C11H14N4O2 (M+):

234.11168; Found: 234.11129.

3,7-Dimethyl-1-(pent-4-enyl)-1H-purine-2,6(3H,7H)-dione

(5.36) Recovered yield = 2.2 g of yellowish solid, 81%. λmax

(MeCN) 273 nm; HPLC RT = 9.5 min (non-chiral method A); 1H

NMR (400 MHz, CDCl3) δ 7.49 (s, 1H, H-4), 5.74-5.88 (m, 1H, H-11), 4.92-5.06 (m, 2H, H-12),

3.96-4.00 (m, 2H, H-8), 3.95 (s, 3H, H-5), 3.53 (s, 3H, H-2), 2.00-2.14 (m, 2H, H-10), 1.67-1.77

13 (m, 2H, H-9); C NMR (300 MHz, CDCl3) δ 155.5 (C-7), 151.7 (C-3), 148.9 (C-1), 141.7 (C-4),

138.0 (C-11), 115.2 (C-12), 100.0 (C-6), 41.1 (C-8), 33.8 (C-5), 31.3 (C-2), 29.9 (C-10), 27.2 (C-

9); HRMS (EI): Calcd. for C12H16N4O2 (M+): 248.12733; Found: 248.12763.

1-(Hex-5-enyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione

(5.37) Recovered yield = 2.5 g of yellowish solid, 87%. λmax

(MeCN) 273 nm; HPLC RT = 14.5 min (non-chiral method

1 A); H NMR (400 MHz, CDCl3) δ 7.50 (s, 1H, H-4), 5.74-5.87 (m, 1H, H-12), 4.92-5.04 (m, 2H,

H-13), 3.99-4.03 (m, 2H, H-8), 3.99 (s, 3H, H-5), 3.57 (s, 3H, H-2), 2.07-2.14 (m, 2H, H-11),

13 1.62-1.72 (m, 2H, H-9), 1.41-1.51 (m, 2H, H-10); C NMR (300 MHz, CDCl3) δ 155.6 (C-7),

151.7 (C-3), 149.0 (C-1), 141.6 (C-4), 138.8 (C-12), 114.9 (C-13), 100.0 (C-6), 41.5 (C-8), 41.2

187

Chapter 8

(C-11), 33.8 (C-5), 29.9 (C-2), 27.8 (C-9), 26.5 (C-10); HRMS (EI): Calcd. for C13H18N4O2

(M+): 262.14298; Found: 262.14260.

(Z)-3,7-Dimethyl-1-(pent-2-enyl)-1H-purine-2,6(3H,7H)-dione

(5.38): (Z)-Pent-2-en-1-ol (1.0 g, 12 mmol) was combined with

NaH (422 mg, 12 mmol) in THF (20 mL) at 0oC. 4-Toluenesulfonyl

chloride (2.21 g, 12 mmol) dissolved in THF (5 mL) was added dropwise. The solution was allowed to warm to room temperature with stirring and was stirred for an extra 3 hours before concentration of the solution in vacuo. The residue was dissolved in ethyl acetate (50 mL) before washing with brine (50 mL) and saturated ammonium chloride (50 mL). The organic layer was separated and concentrated in vacuo. The resulting solid was dissolved in minimal DMSO and added to a mixture of NaH (277 mg, 7 mmol) and theobromine

(5.1, 1.4 g, 7 mmol) in DMSO (30 mL) at 600C. The reaction was allowed to proceed overnight before cooling to room temperature and quenching with water (2 mL). After concentration of the solution in vacuo, the resulting solid was dissolved in CHCl3 (50 mL). This solution was washed with saturated ammonium chloride (20 mL) followed by brine (20 mL) and concentrated, to afford the desired product. Yield = 1.4 g of yellowish solid, 50%. λmax (MeCN) 273 nm; HPLC

1 RT = 9.5 min (non-chiral method A); H NMR (400 MHz, CDCl3) δ 7.49 (s, 1H, H-4), 5.55-5.62

(m, 1H, H-9), 5.37-5.44 (m, 1H, H-10), 4.67 (d, J = 6.4, 2H, H-8), 3.99 (s, 3H, H-5), 3.58 (s, 3H,

13 H-2), 1.99-2.06 (m, 2H, H-11), 1.04 (t, J = 7.2, 3H, H-12); C NMR (300 MHz, CDCl3) δ 155.2

(C-7), 151.6 (C-3), 148.7 (C-1), 141.5 (C-4), 136.7 (C-9), 122.6 (C-10), 107.7 (C-6), 42.8 (C-8),

188

Chapter 8

40.9 (C-5), 33.6 (C-2), 29.7 (C-11), 13.1 (C-12); HRMS (EI): Calcd. For C12H16N4O2 (M+):

248.12733; Found: 248.12688.

3-(Tert-butyldimethylsilyloxy)propan-1-ol (5.40) Propane-1,3-diol

(5.0 g, 66 mmol) was combined with NaH (4.2 g, 66 mmol) in dry

THF (400 mL) and stirred for 45 minutes at room temperature before the addition of TBSCl (9.9 g, 66 mmol). The mixture was stirred for 90 minutes before filtration through Celite followed by concentration in vacuo. The resulting oil was resuspended in chloroform , washed with water (3 x 400 mL) and dried over anhydrous sodium sulfate before concentration in vacuo. The residue was purified by silica gel chromatography (hexanes/EtOAc

4:1, Rf = 0.4) to afford the product as a clear oil (6.0 g, 48%). NMR spectra for this known compound were in agreement with literature sources161.

3-(Tert-butyldimethylsilyloxy)propyl 4-

methylbenzenesulfonate (5.41) Compound 5.40 (6.0 g,

31 mmol), DMAP (192 mg, 1.5 mmol, 5 mol%), tosylchloride (7.2 g, 37 mmol) and TEA (6.1 mL, 31 mmol) were combined in DCM (200 mL) and stirred at room temperature for 12 hours. The reaction mixture was washed with water (3 x

200 mL) and the organic fraction was dried over anhydrous sodium sulfate before concentration in vacuo. The residue was purified by silica gel chromatography (hexanes/EtOAc 9:1, Rf = 0.5)

1 to afford the product as a clear oil (9.3 g, 87 %). H NMR (400 MHz, CDCl3) δ 7.75 (d, J = 8.1,

2H, H-4), 7.30 (d, J = 8.0, 2H, H-3), 4.11 (t, J = 6.6, 2H, H-6), 3.60 (t, J = 6.6, 2H, H-8), 2.41 (s, 189

Chapter 8

3H, H-1), 1.76-1.84 (m, 2H, H-7), 0.79 (s, 9H, H-11), -0.041 (s, 6H, H-9); 13C NMR (300 MHz,

CDCl3) δ 144.7 (C-5), 133.0 (C-2), 129.8 (C-3), 127.8 (C-4), 67.5 (C-8), 58.3 (C-6), 31.9 (C-7),

25.7 (C-10), 21.6 (C-11), 18.1 (C-1), -5.6 (C-9); HRMS (EI): Calcd. For C16H28O2SSi (M+):

344.14776; Found: 344.15102.

1-(3-(Tert-butyldimethylsilyloxy)propyl)-3,7-

dimethyl-1H-purine-2,6(3H,7H)-dione (5.42):

Recovered yield = 2.9 g of yellowish solid, 74%. 1H

NMR (400 MHz, CDCl3) δ 7.43 (s, 1H, H-4), 3.98 (t, J = 7.2, 2H, H-8), 3.87 (s, 3H, H-5), 3.60

(m, 2H, H-10), 3.44 (s, 3H, H-2), 1.71-1.81 (m, 2H, H-9), 0.75 (s, 9H, H-13), -0.9 (s, 6H, H-11);

13 C NMR (300 MHz, CDCl3) δ 155.1 (C-7), 151.3 (C-3), 148.6 (C-1), 141.3 (C-4), 107.5 (C-6),

61.3 (C-10), 40.9 (C-8), 39.0 (C-5), 33.5 (C-2), 31.0 (C-9), 25.8 (C-13), 18.1 (C-12), -5.5 (C-11);

HRMS (EI): Calcd. For C16H28N4O3Si (M+): 352.19307; Found: 352.19332.

1-(3-Bromopropyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-dione

(5.44): Compound 5.22 (1.45 g, 6.1 mmol) and carbon tetrabromide

(1.2 g, 6.7 mmol) were combined in DCM (100 mL) and stirred at

0oC for 10 minutes before the addition of triphenylphosphine (1.7 g, 6.7 mmol). The mixture was allowed to come to room temperature and was stirred for 3 hours before being quenched by repeated washes with water (3 x 100 mL). The organic fraction was filtered through Celite and concentrated in vacuo to afford the pure product as a slightly yellow solid (1.4 g, 90 %). 1H

NMR (300 MHz, CDCl3): δ 7.45 (s, 1H, H-4), 3.99 (t, J = 6.9, 2H, H-8) 3.85 (s, 3H, H-5), 3.42 190

Chapter 8

13 (s, 3H, H-2), 3.31 (t, J = 6.9, 2H, H-10), 2.02-2.13 (m, 2H, H-9); C NMR (300 MHz, CDCl3) δ

154.9 (C-7), 151.2 (C-3), 148.6 (C-1), 141.6 (C-4), 107.4 (C-6), 38.9 (C-10), 35.8 (C-5), 33.4 (C-

2), 27.1 (C-9); HRMS (EI): Calcd. for C10H13BrN4O2 (M+): 300.00219; Found: 300.00158.

1-(4-Hydroxy-4-methylpentyl)-3,7-dimethyl-1H-purine-

2,6(3H,7H)-dione (5.47): Compound 5.27 (50 mg, 0.17 mmol)

was dissolved in THF (2 mL) and cooled to -78oC.

Methylmagnesium bromide (567 µL, 3 M in ether, 10 equiv.) was added dropwise over 20 minutes. The reaction was terminated after 2 hours by the addition of aqueous ammonium chloride (0.5 mL). The solution was filtered through Celite, dried over anhydrous sodium sulfate, and concentrated in vacuo before purification by preparative TLC (chloroform:methanol 9:1, Rf

= 0.35) to afford the product as a slightly yellow solid (41 mg, 87%). λmax (MeCN) 273 nm;

1 Purity > 85%, HPLC RT = 2.1 min (non-chiral method B); H NMR (400 MHz, CDCl3) δ 7.56

(s, 1H, H-4), 4.00-4.06 (m, 5H, H-5,8), 3.59 (s, 3H, H-2), 1.70-1.80 (m, 2H, H-10), 1.52-1.56 (m,

13 2H, H-9), 1.22 (s, 6H, H-12); C HSQC NMR (300 MHz, CDCl3) δ 155.6 (C-7), 151.7 (C-3),

143.9 (C-1), 141.6 (C-4), 107.6 (C-6), 41.5 (C-11), 40.5 (C-8), 33.6 (C-10), 29.8 (C-12), 18.0

(C-9); HRMS (EI): Calcd. For C13H20N4O3 (M+Na): 303.14276; Found 303.14280.

Ethyl 4-(3,7-dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-

purin-1-yl)butanoate (5.49): NaH (444 mg, 11 mmol, 60%

in mineral oil) was washed with hexanes (2  8 mL), dried under nitrogen, and suspended in DMSO (30 mL). The NaH solution was heated to 60oC with 191

Chapter 8 stirring before addition of theobromine (5.1, 2 g, 11 mmol) as a solid in a single addition. After

20 minutes, ethyl 4-chlorobutanoate (1.5 mL, 11 mmol) was added in a single addition, and the reaction was allowed to proceed overnight before cooling to room temperature and quenching with water (2 mL). After concentration of the solution in vacuo, the resulting solid was redissolved in CHCl3 (50 mL). This solution was washed with saturated ammonium chloride (20 mL) then brine (20 mL) and concentrated, to afford the desired product as an orange solid (2.5 g,

1 78%). H NMR (400 MHz, CDCl3) δ 7.50 (s, 1H, H-4), 4.08 (m, 4H, H-12,8), 3.98 (s, 3H, H-5),

3.57 (s, 3H, H-2), 2.38 (t, J = 8.0, 2H, H-10), 1.93-2.02 (m, 2H, H-9), 1.24 (t, J = 7.2, 3H, H-13);

13 C NMR (300 MHz, CDCl3) δ 172.7 (C-11), 155.1 (C-7), 151.3 (C-3), 148.6 (C-1), 141.5 (C-4),

107.4 (C-6), 60.2 (C-12), 40.3 (C-8), 33.5 (C-5), 31.6 (C-2), 29.5 (C-10), 23.3 (C-9), 14.1 (C-

13); HRMS (EI): Calcd. for C13H19N4O4 (M+): 295.14008; Found: 295.14008.

1-(4-Hydroxybutyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-

dione (5.50): Compound 5.49 (1.9 g, 6.4 mmol) was dissolved

in THF (10 mL) and added dropwise to a 0oC solution of lithium aluminum hydride (728 mg, 19.2 mmol) in THF (30 mL). The solution was allowed to warm to room temperature and the progress was monitored by TLC. Upon completion (~3 hours), the reaction was terminated by the addition of saturated aqueous NaOH (200 µL). The resulting yellow liquid was filtered through Celite and dried over anhydrous sodium sulfate before being concentrated in vacuo to afford the product as a slightly yellow solid (1.14 g, 64 %). 1H NMR

(300 MHz, CDCl3): δ 7.51 (s, 1H, H-4), 4.04 (t, J = 7.2, 2H, H-8), 3.98 (s, 3H, H-5), 3.66-3.71

(m, 2H, H-11), 3.56 (s, 3H, H-2), 1.99 (bs, 1H, OH), 1.57-1.78 (m, 4H, H-10,9); 13C NMR (300

192

Chapter 8

MHz, CDCl3) δ 155.6 (C-7), 151.7 (C-3), 149.0 (C-1), 141.7 (C-4), 104.1 (C-6), 62.6 (C-11),

41.10 (C-8), 35.2 (C-5), 31.1 (C-2), 29.9 (C-10), 24.6 (C-9); HRMS (EI): Calcd. for C11H16N4O3

(M+): 253.12952; Found: 253.12939.

4-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-1-

yl)butanal (5.51): Compound 5.50 (1.14 mg, 4.6 mmol) and

TEMPO suspended on silica (6.54 g, 4.6 mmol) were combined in

DCM (30 mL) followed by the addition of iodobenzene diacetate (1.48 g, 1.5 equiv.). The solution was allowed to stir at room temperature for 2 hours before filtration over Celite. The filtrate was concentrated in vacuo and was resuspended in a minimum volume of CHCl3.

Purification by flash chromatography (chloroform:methanol 9:1, Rf = 0.40) afforded the product

1 as a slightly orange oil (828 mg, 72%). H NMR (300 MHz, CDCl3) δ 9.78 (s, 1H, H-11), 7.51 (s,

1H, H-4), 3.91-4.02 (m, 5H, H-5,8), 3.57 (s, 3H, H-2), 2.44-2.52 (m, 2H, H-10), 1.95-2.04 (m,

13 2H, H-9); C NMR (300 MHz, CDCl3) δ 201.5 (C-11), 155.2 (C-7), 151.5 (C-3), 148.7 (C-1),

141.6 (C-4), 107.6 (C-6), 41.2 (C-8), 40.5 (C-10), 20.6 (C-9); HRMS (EI): Calcd. for

C11H15N4O3 (M+): 251.11387; Found: 251.11363.

1-(4-Hydroxypentyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-

dione (5.52): Compound 5.51 (50 mg, 0.20 mmol) was dissolved

in THF (2 mL) and cooled to -78oC. Methylmagnesium bromide

(200 µL, 3.0 M in ether, 3 equiv.) was added dropwise over 20 minutes. The reaction was terminated by the addition of aqueous ammonium chloride (0.5 mL). The solution was filtered 193

Chapter 8 through Celite, dried over anhydrous sodium sulfate, and concentrated in vacuo before purification by preparative TLC (chloroform:methanol 9:1, Rf = 0.35) to afford the product as a slightly yellow solid (47 mg, 88%). Pure compound 5.52 was separated into R and S enantiomers using HPLC chiral method B. λmax (MeCN) 273 nm; Purity > 95%, HPLC RT = 2.0 min (non-chiral method B), RT of enantiomers = 7.9 and 8.2 min (chiral method A), RT of

1 enantiomers = 20.0 and 21.2 min (chiral method B); H NMR (400 MHz, CDCl3) δ 7.52 (s, 1H,

H-4), 4.04 (t, J = 7.2, 2H, H-8), 3.99 (s, 3H, H-5), 3.89 (m, 1H, H-11), 3.58 (s, 3H, H-2), 1.99

(bs, 1H, OH), 1.67-1.84 (m, 2H, H-10), 1.52 (m, 2H, H-9), 1.20 (d, J = 6.4, 3H, H-12); 13C

HSQC NMR (500 MHz, CDCl3) δ 155.1 (C-7), 151.3 (C-3), 148.6 (C-1), 143.2 (C-4), 107.4 (C-

6), 67.4 (C-11), 40.5 (C-8), 35.2 (C-5), 34.8 (C-2), 29.3 (C-10), 24.0 (C-9), 23.2 (C-12); HRMS

(EI): Calcd. for C12H18N4O3 (M+Na): 289.12711; Found: 289.12687

1-(4-Hydroxyhexyl)-3,7-dimethyl-1H-purine-2,6(3H,7H)-

dione (5.53): Compound 5.51 (50 mg, 0.20 mmol) was

dissolved in THF (2 mL) and cooled to -78oC.

Ethylmagnesium bromide (200 µL, 3.0 M in ether, 3 equiv.) was added dropwise over 20 minutes. The reaction was terminated by the addition of aqueous ammonium chloride (0.5 mL).

The solution was filtered through Celite, dried over anhydrous sodium sulfate, and concentrated in vacuo before purification by preparative TLC (9:1 chloroform:methanol, Rf = 0.35) to afford the product as a slightly yellow solid (47 mg, 84%). Pure compound 5.53 was separated into R and S enantiomers using HPLC chiral method B. λmax (MeCN) 273 nm; Purity > 95%, HPLC RT

= 2.2 min (non-chiral Method B), RT of enantiomers = 8.6 and 9.4 min (chiral method A), RT of

194

Chapter 8

1 enantiomers = 35.2 and 39.8 min (chiral method B); H NMR (400 MHz, CDCl3): δ 7.50 (s, 1H,

H-4), 4.03 (t, J = 8.0, 2H, H-8), 3.97 (s, 3H, H-5), 3.54-3.62 (m, 3H, H-11,2), 1.40-1.84 (m, 6H,

13 H-12,10,9), 0.93 (t, J = 7.2, 3H, H-13); C HSQC NMR (500 MHz, CDCl3) δ 155.4 (C-7), 151.7

(C-3), 149.4 (C-1), 141.2 (C-4), 105.4 (C-6), 73.3 (C-11), 41.1 (C-8), 36.2 (C-5), 35.3 (C-2),

34.5 (C-10), 32.8 (C-12), 29.9 (C-9), 20.8 (C-13); HRMS (EI): Calcd. for C13H20N4O3 (M+23):

303.1428; Found: 303.1423.

1-(4-Hydroxyhex-5-enyl)-3,7-dimethyl-1H-purine-

2,6(3H,7H)-dione (5.54): Compound 5.51 (50 mg, 0.20 mmol)

was dissolved in THF (2 mL) and cooled to -78oC.

Vinylmagnesium bromide (600 µL, 1.0 M in THF, 3 equiv.) was added dropwise over 20 minutes. The reaction was terminated by the addition of aqueous ammonium chloride (0.5 mL).

The solution was filtered through Celite, dried over anhydrous sodium sulfate, and concentrated in vacuo before purification by preparative TLC (chloroform:methanol 9:1, Rf = 0.30) to afford the product as a slightly yellow solid (48 mg, 87%). Pure compound 5.54 was separated into R and S enantiomers using HPLC chiral method B. λmax (MeCN) 273 nm; Purity >95%, HPLC RT

= 2.1 min (non-chiral method B), RT of enantiomers = 8.5 and 9.1 min (chiral method A), RT of

1 enantiomers = 27.0 and 30.8 min (chiral method B); H NMR (400 MHz, CDCl3): δ 7.50 (s, 1H,

H-4), 5.86 (m, 1H, H-12), 5.23 (m, 1H, H-13a), 5.08 (m, 1H, H-13b), 4.17 (m, 1H, H-11), 4.04

(t, 2H, H-8), 3.97 (s, 3H, H-5), 3.56 (s, 3H, H-2), 2.17 (bs, 1H, OH), 1.66-1.82 (m, 2H, H-10),

13 1.60 (m, 2H, H-9); C NMR (300 MHz, CDCl3) δ 155.5 (C-7), 151.7 (C-3), 149.0 (C-1), 141.7

195

Chapter 8

(C-4), 141.2 (C-12), 114.9 (C-13), 100.0 (C-6), 72.9 (C-11), 41.2 (C-8), 29.9 (C-9), 24.2 (C-10);

HRMS (EI): Calcd. For C11H16N4O3 (M+Na-CH2CH2): 275.11146; Found 275.11078.

(R)-((R)-5-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-

1H-purin-1-yl)pentan-2-yl) 2-methoxy-2-phenylacetate

(5.55a): The optically pure R isomer of compound 5.52 (5

mg, 18.0 µmol) was added to a solution of (R)-(−)-α-

methoxyphenylacetic acid (3.7 mg, 22 µmol), 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide (3.5 mg, 22 µmol), and 4-dimethylaminopyridine (1 mg, 3.2

µmol) in DCM (1 mL). The solution was allowed to stir at room temperature overnight before concentration in vacuo and purification by preparative TLC (chloroform:methanol 9:1, Rf =

1 0.50) to afford the product as a slightly yellow oil (1.2 mg, 16%). H NMR (300 MHz, CDCl3) δ

7.21-7.52 (m, 6H, H-4,17,18,19), 4.78 (s, 1H, H-14), 3.98 (s, 3H, H-5), 3.86 (t, J = 6.8, 2H, H-8),

3.74 (m, 1H, H-11), 3.58 (s, 3H, H-2), 3.40 (s, 3H, H-15), 1.40-1.51 (m, 4H, H-10,9), 1.23 (d, J

13 = 6.4, 3H, H-12); C HSQC NMR (500 MHz, CDCl3) δ 170.4, 155.5, 151.5, 149.9, 141.3,

130.6, 125.5, 120.6, 115.9, 105.8, 89.9, 72.6, 57.3, 29.5, 35.8, 33.1, 29.8, 20.5, 18.4; HRMS (EI):

Calcd. for C21H26N4O5 (M+Na): 437.1795; Found: 437.1788.

(S)-((R)-5-(3,7-dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-

purin-1-yl)pentan-2-yl) 2-methoxy-2-phenylacetate

(5.55b): The optically pure R isomer of compound 5.52 (6

196

Chapter 8 mg, 18.0 µmol) was added to a solution of (S)-(+)-α-methoxyphenylacetic acid (3.7 mg, 22

µmol), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (3.5 mg, 22 µmol), and 4- dimethylaminopyridine (1 mg, 3.2 µmol) in DCM (1 mL). The solution was allowed to stir at room temperature overnight before concentration in vacuo and purification by preparative TLC

(chloroform:methanol 9:1, Rf = 0.50) to afford the product as a slightly yellow oil (1.4 mg,

1 19%). H NMR (300 MHz, CDCl3) δ 7.51 (s, 1H, H-4), 7.30-7.45 (m, 5H, H-17,18,19), 4.50 (m,

1H, H-11), 4.74 (s, 1H, H-14), 3.99-4.01 (m, 5H, H-8,5), 3.73 (s, 3H, H-15), 3.58 (s, 3H, H-2),

1.64-1.68 (m, 4H, H-10,9), 1.07 (d, J = 6.4, 3H, H-12). HRMS (EI): Calcd. for C21H26N4O5

(M+23): 437.1795; Found: 437.1765.

(R)-((R)-6-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-

purin-1-yl)hexan-3-yl) 2-methoxy-2-phenylacetate

(5.56a): The optically pure R isomer of compound 5.53 (5

mg, 18.0 µmol) was added to a solution of (R)-(−)-α-

methoxyphenylacetic acid (3.7 mg, 22 µmol), 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide (3.5 mg, 22 µmol), and 4-dimethylaminopyridine (1 mg, 3.2

µmol) in DCM (1 mL). The solution was allowed to stir at room temperature overnight before concentration in vacuo and purification by preparative TLC (chloroform:methanol 9:1, Rf =

1 0.50) to afford the product as a slightly yellow oil (1.8 mg, 23%). H NMR (400 MHz, CDCl3): δ

7.21-7.52 (m, 6H, H-4,18,19,20), 4.73 (s, 1H, H-15), 3.91-4.00 (m, 4H, H-5,11), 3.38 (t, J = 7.2,

2H, H-8), 3.56 (s, 3H, H-2), 3.41 (s, 3H, H-16), 0.96-1.02 (m, 2H, H-10), 0.87-0.92 (m, 2H, H-

13 12), 0.81-0.85 (m, 5H, H-9,13); C HSQC NMR (500 MHz, CDCl3) δ 170.9, 155.4, 151.9,

197

Chapter 8

149.8, 141.5, 130.5, 125.2, 120.7, 115.1, 105.0, 90.2, 72.0, 56.4, 29.4, 35.8, 33.1, 31.2 29.9, 20.1,

19.1; HRMS (EI): Calcd. For C22H28O5N4 (M+Na): 451.19519; Found: 451.19581.

(S) -((R)-6-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-

1H-purin-1-yl)hexan-3-yl) 2-methoxy-2-phenylacetate

(5.56b): The optically pure R isomer of compound 5.53 (5

mg, 18.0 µmol) was added to a solution of (S)-(+)-α-

methoxyphenylacetic acid (3.7 mg, 22 µmol), 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide (3.5 mg, 22 µmol), and 4-dimethylaminopyridine (1 mg, 3.2

µmol) in DCM (1 mL). The solution was allowed to stir at room temperature overnight before concentration in vacuo and purification by preparative TLC (chloroform:methanol 9:1, Rf =

1 0.50) to afford the product as a slightly yellow oil (1.7 mg, 22%). H NMR (400 MHz, CDCl3): δ

7.34-7.51 (m, 6H, H-4,18,19,20), 4.75 (s, 1H, H-15), 3.99 (m, H, H-5,8), 3.57 (s, 3H, H-2), 3.40

(s, 3H, H-16), 0.79-0.97 (m, 6H, H-9,10,12), 0.54 (t, J = 7.2, 3H, H-13); HRMS (EI): Calcd. For

C22H28O5N4Na (M+): 451.19519; Found: 451.19551.

(R)- ((S)-6-(3,7-dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-

1H-purin-1-yl)hex-1-en-3-yl) 2-methoxy-2-phenylacetate

(5.57a): The optically pure S isomer of compound 5.54 (5

mg, 18.0 µmol) was added to a solution of (R)-(−)-α-

methoxyphenylacetic acid (3.7 mg, 22 µmol), 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide (3.5 mg, 22 µmol), and 4-dimethylaminopyridine (1 mg, 3.2 198

Chapter 8

µmol) in DCM (1 mL). The solution was allowed to stir at room temperature overnight before concentration in vacuo and purification by preparative TLC (chloroform:methanol 9:1, Rf =

1 0.50) to afford the product as a slightly yellow oil (1.5 mg, 19%). H NMR (300 MHz, CDCl3): δ

7.31-7.52 (m, 6H, H-4,18,19,20), 5.70-5.85 (m, 1H, H-12), 5.14-5.24 (m, 3H, H-11,13), 4.76 (s,

1H, H-15), 3.99 (s, 3H, H-5), 3.88-3.94 (m, 2H, H-8), 3.58 (s, 3H, H-16), 3.49 (s, 3H, H-2),

13 0.790-0.90 (m, 4H, H-10,9); C HSQC NMR (500 MHz, CDCl3) δ 169.9, 155.0, 151.0, 149.4,

141.6, 135.4, 130.1, 125.0, 120.4, 116.6, 114.8, 105.5, 90.5, 72.4, 56.6, 29.1, 35.6, 32.8, 29.5,

20.4; HRMS (EI): Calcd. for C22H26N4O5 (M+Na): 449.1795; Found: 449.1786.

(S)-((S)-6-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-

purin-1-yl)hex-1-en-3-yl) 2-methoxy-2-phenylacetate

(5.57b): The optically pure R isomer of compound 5.54 (5

mg, 18.0 µmol) was added to a solution of (S)-(+)-α-

methoxyphenylacetic acid (3.7 mg, 22 µmol), 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide (3.5 mg, 22 µmol), and 4-dimethylaminopyridine (1 mg, 3.2

µmol) in DCM (1 mL). The solution was allowed to stir at room temperature overnight before in vacuo concentration and purification by preparative TLC (chloroform:methanol 9:1, Rf = 0.50)

1 to afford the product as a slightly yellow oil (2.0 mg, 25%). H NMR (300 MHz, CDCl3): δ 7.30-

7.52 (m, 6H, H-4,18,19,20), 5.46-5.61 (m, 1H, H-12), 4.85-4.98 (m, 2H, H-13), 4.76 (s, 1H, H-

15), 3.98 (s, 3H, H-5), 3.88 (t, J = 7.6, 2H, H-8), 3.56 (s, 3H, H-16), 3.41 (s, 3H, H-2), 0.76-0.90

(m, 4H, H-9,10); HRMS (EI): Calcd. for C22H26N4O5 (M+Na): 449.1795; Found: 449.17168.

199

Chapter 8

(2-Bromoethyl)oxirane (5.61): 4-Bromo-1-butene (750 µL, 4.7 mmol) was

added to a solution of meta-chloroperoxybenzoic acid (4.2 g, 77% maximum purity, ~4 equiv.) in DCM (40 mL). The solution was allowed to stir overnight before purification by in vacuo distillation to afford the desired known compound as a colorless oil (600 mg, 85%). B.p. 65oC at 240 mmHg; NMR spectra for this known compound were in agreement with literature sources162.

(3-Bromopropyl)oxirane (5.62): 4-Bromo-1-pentene (793 µL, 7.4 mmol)

was added to a solution of meta-chloroperoxybenzoic acid (6.7 g, 77% maximum purity, 4 equiv.) in DCM (40 mL). The solution was allowed to stir overnight before purification by in vacuo distillation to afford the desired known compound as a colorless oil (600 mg, 85%). B.p. 65oC at 240 mmHg; NMR spectra for this known compound were in agreement with literature sources162.

(R)-(2-Bromomethyl)oxirane (5.63): (S,S)-Co(OAc)-Salen was prepared

according to Jacobsen‟s protocol163. Water (15 µL, 0.83 mmol) was added to a 0oC mixture of (S,S)-Co(OAc)-Salen (2 mol%, 30 µmol), and compound 5.61 (197 mg, 1.5 mmol) in THF (200 µL). The solution was allowed to warm to room temperature and was left to stir for 3 hours before purification by in vacuo distillation to afford the desired compound as a colorless oil (101 mg, 45%). B.p. 65oC at 240 mmHg; NMR spectra for this known compound were in agreement with literature sources162.

200

Chapter 8

(R)-(3-Bromopropyl)oxirane (5.64): (S,S)-Co(OAc)-Salen was

prepared according to Jacobsen‟s protocol163. Water (15 µL, 0.83 mmol) was added to a 0oC mixture of (S,S)-Co(OAc)-Salen (2 mol%, 30 µmol), and compound 5.62

(247 mg, 1.5 mmol) in THF (200 µL). The solution was allowed to warm to room temperature and was left to stir for 3 hours before purification by in vacuo distillation to afford the desired compound as a colorless oil (103 mg, 42%). B.p. 65oC at 240 mmHg; NMR spectra for this known compound were in agreement with literature sources162.

(R)-3,7-Dimethyl-1-(2-(oxiran-2-yl)ethyl)-1H-purine-

2,6(3H,7H)-dione (5.65): NaH (16 mg, 0.67 mmol, 60% in mineral

oil) was washed with hexanes (1 mL), dried under nitrogen, and

suspended in DMSO (6 mL). The NaH suspension was heated to

70oC with stirring before addition of theobromine (5.1, 120 mg, 0.67 mmol) as a solid in a single addition. After 20 minutes, compound 5.63 (101 mg, 0.67 mmol) was added in a single addition, and the reaction was allowed to proceed overnight before cooling to room temperature and quenching with water (200 µL). After concentration of the solution in vacuo, the resulting solid was redissolved in CHCl3 (5 mL). This solution was washed with saturated ammonium chloride

(5 mL), brine (5 mL), and water (5 mL). The organic fraction was concentrated in vacuo and purified by flash chromatography (chloroform:methanol 9:1, Rf = 0.40) to afford the product as a slightly orange oil (125 mg, 75%). λmax (MeCN) 273 nm; Purity >90%, HPLC RT = 5.5 min

(non-chiral method A), RT of enantiomers = 7.4 and 8.6 min (chiral method A); 1H NMR (400

MHz, CDCl3) δ 7.50 (s, 1H, H-4), 4.14-4.31 (m, 2H, H-8), 3.98 (s, 3H, H-5), 3.58 (s, 3H, H-2),

201

Chapter 8

3.02-3.07 (m, 1H, H-10), 2.72 (m, 1H, H-11a), 2.45 (m, 1H, H-11b) 1.84-2.00 (m, 2H, H-9); 13C

HSQC NMR (500 MHz, CDCl3): δ 155.1 (C-7), 151.4 (C-3), 148.7 (C-1), 141.6 (C-4), 107.5 (C-

6), 51.8 (C-10), 46.9 (C-11), 42.6 (C-8), 29.6 (C-9); HRMS (EI): Calcd. for C11H14N4O3

(M+Na): 273.09581; Found: 273.09558.

(R)-3,7-Dimethyl-1-(3-(oxiran-2-yl)propyl)-1H-purine-

2,6(3H,7H)-dione (5.66): NaH (15 mg, 0.62 mmol, 60% in

mineral oil) was washed with hexanes (1 mL), dried under

nitrogen, and suspended in DMSO (6 mL). The NaH suspension was heated to 70oC with stirring before addition of theobromine (5.1, 111 mg, 0.62 mmol) as a solid in a single addition. After 20 minutes, compound 5.64 (103 mg, 0.62 mmol) was added in a single addition, and the reaction was allowed to proceed overnight before cooling to room temperature and quenching with water (200 µL). After concentration of the solution in vacuo, the resulting solid was redissolved in CHCl3 (5 mL). This solution was washed with saturated ammonium chloride (5 mL), brine (5 mL), then water (5 mL). The organic fraction was concentrated in vacuo and purified by flash chromatography (chloroform:methanol 9:1, Rf =

0.40) to afford the product as a slightly orange oil (118 mg, 72%). λmax (MeCN) 273 nm; Purity >

95%, HPLC RT = 5.9 min (non-chiral method A), RT of enantiomers = 7.0 and 7.4 min (chiral

1 method A); H NMR (400 MHz, CDCl3): δ 7.48 (s, 1H, H-4), 3.94-3.99 (m, 2H, H-8), 3.90 (s,

3H, H-5), 3.42 (s, 3H, H-2), 2.86-2.90 (m, 1H, H-11), 2.65-2.68 (m, 1H, H-12a), 2.42-2.43 (m,

13 1H, H-12b), 1.64-1.81 (m, 2H, H-9), 1.50-1.58 (m, 2H, H-10); C NMR (300 MHz, CDCl3) δ

155.3 (C-7), 151.3 (C-3), 148.5 (C-1), 141.7 (C-4), 107.5 (C-6), 51.8 (C-11), 46.9 (C-12), 42.6

202

Chapter 8

(C-8), 36.5 (C-5), 33.5 (C-2), 30.2 (C-10), 29.6 (C-9); HRMS (EI): Calcd. for C12H17N4O3 (M+):

265.12952; Found: 265.12947.

(R)-3,7-Dimethyl-1-(3-(oxiran-2-yl)propyl)-1H-purine-

2,6(3H,7H)-dione (5.67): Compound 5.34 (10 mg, 35 μmol) and

ammonium persulfate (12.2 mg, 53 μmol) were combined in water/THF (50/50 v/v, 5 mL) and stirred at room temperature for 1 hour. The mixture was concentrated in vacuo before purification by preparative TLC (chloroform:methanol 9:1, Rf =

0.40) affording the product as a slightly yellow solid (6 mg, 56%). λmax (MeCN) 273 nm; Purity

1 > 95%, HPLC RT = 5.0 min (non-chiral method A); H NMR (500 MHz, CDCl3): δ 7.53 (s, 1H,

H-4), 4.14-4.20 (m, 2H, H-8), 3.98 (s, 3H, H-5), 3.58 (s, 3H, H-2), 2.69-2.77 (m, 4H, H-10,11),

13 2.12-2.20 (m, 2H, H-9), 1.34 (t, J = 7.5, H-12); C NMR (300 MHz, CDCl3) δ 155.7 (C-7),

151.3 (C-3), 146.7 (C-1), 144.0 (C-4), 107.4 (C-6), 48.9 (C-8), 44.9 (C-10), 42.5 (C-11), 35.2 (C-

5), 29.9 (C-2), 25.8 (C-9), 7.1 (C-12); HRMS (EI): Calcd. for C12H18N4O3S (M+): 298.10996;

Found: 268.10937.

6-Aminohexan-3-ol (5.74): 1.0 M DIBAL-H in THF (1.07 mL,

1.07 mM, 10 equiv.) was added dropwise to a solution of compound 5.53

(30 mg, 107 µM), in THF (2 mL) at -78oC. The mixture was left to stir overnight before dilution with THF at 0oC followed the addition of water (50 µL), 15% NaOH (50 µL), and water (125

µL) to quench the reaction. The organic phase was dried over magnesium sulfate and evaporated

203

Chapter 8 in vacuo before purification by preparative TLC (chloroform:methanol:ammonium hydroxide

4:3:1, Rf = 0.3) to afford a colorless oil (80%). NMR spectra for this known compound were in

164 agreement with literature sources ; LC-MS (ES): Calcd. for C6H15NO (M+1): 118.1; Found:

118.1.

8.4 In Silico studies

Docking studies were performed using the default settings of the FITTED V3.0 suite of docking softwares developed at and licensed by McGill University.The pdb file of the crystal structure for CYP3A4 (1w0g) was prepared by PROCESS and the ligands to be docked were prepared by SMART; both modules within the FITTED suite.

8.5 MIP studies

8.5.1 MIP synthesis

The MIPs were synthesized according to the procedure of Theodoridis and Chromatigr148.

The radical initiator, benzyl peroxide (16 mg, 61 μmol) and the desired template ligand (~2-3 mg, 0.5 mmol) were dissolved in ACN (2 mL). Methacrylic acid (169 μL, 2 mmol) and ethylene glycol dimethacrylate (1.2 mL, 6 mmol) were added and the mixture was degassed with nitrogen for 2 minutes before being left to stir at 60°C overnight. The resulting solid was ground into a fine powder and the finest particles were removed by repeatedly suspending the powder in water

(3 x 50 mL) followed by decanting the aqueous layer above the precipitate. The powder was then oven-dried, loaded into empty solid-phase extraction cartridges (empty columns) and washed

204

Chapter 8 with 5 column volumes of methanol:acetic acid (9:1). The final wash was analyzed by HPLC to verify that no template ligand was present in the MIP.

8.5.2 Small volume MIP assay

A solution of the desired ligand (~0.7 mg, 2.5 mL, 1.03 mM) was loaded onto the MIPs

(2.5 mL) by centrifugation (3000g for 5 minutes) and the flow through was collected, leaving a dry solid phase. The MIPs were then washed with 2 column volumes (5 mL) of water by centrifugation (3000 × g for 5 minutes each), leaving a dry solidphase. Finally, the MIPs were eluted with 2 column volumes (5 mL) of methanol:acetic acid (9:1) and all fractions were analyzed by LC-UV-MS after dilution to 10 mL in water.

8.5.3 Large volume MIP assay to separate theobromine-containing compounds from complex mixtures.

Solutions containing the theobromine derivatives (50 mL, 20 column volumes, 8.50 μM) were dissolved in water, Lysogeny Broth, or a solution of CYP3A4-containing membranes (see section 8.1.2.2) and the resulting „spiked‟ mixtures were loaded onto the MIPs (2.5 mL) using a peristaltic pump. The flow-through was collected, leaving a dry solid phase. The MIP was then washed with 2 column volumes (5 mL) of water by centrifugation (3000 × g for 5 minutes), leaving a dry solid phase. Finally, the MIPs were eluded with 2 column volumes (5 mL) of methanol/acetic acid (9:1) and all fractions were analyzed by LC-UV-MS after dilution to 100 mL in water.

205

Chapter 8

8.5.4 Large volume MIP assay to separate theobromine-containing compounds from a reaction mixture of CYP3A4-catalyzed oxidation of 5.6.

Compound 5.6 (0.7 mg, 2.6 mmol) was added to a solution of CYP3A4-containing membranes (5 mL, 15 μM, see section 8.1.2.2) and the mixture was pre-incubated at 27oC for 5 minutes before the addition of CHP (481 μL, 2.6 mmol). The mixture was shaken at 250 RPM and 27oC for 2 hours before dilution to 50 mL in water. The diluted solution was loaded onto the

MIPs (2.5 mL) using a peristaltic pump with minimal back-pressure. The flow-through was collected, leaving a dry solid phase. The MIPs were then washed with 20 column volumes (50 mL) of water by centrifugation (3000 × g for 5 minutes), leaving a dry solid phase. Finally, the

MIPs were eluded with 20 column volumes (50 mL) of methanol:acetic acid (9:1) and all fractions were analyzed by LC-UV-MS after dilution to 100 mL in water.

206

References

1. Wei, L.; Locuson, C. W.; Tracy, T. S., Polymorphic Variants of CYP2C9: Mechanisms Involved in Reduced Catalytic Activity. Molecular Pharmacology 2007, 72 (5), 1280-1288. 2. Porubsky, P. R.; Battaile, K. P.; Scott, E. E., Human Cytochrome P450 2E1 Structures with Fatty Acid Analogs Reveal a Previously Unobserved Binding Mode. Journal of Biological Chemistry 2010, 285 (29), 22282-22290. 3. Nicolaou, K. C.; Ueno, H.; Liu, J. J.; Nantermet, P. G.; Yang, Z.; Renaud, J.; Paulvannan, K.; Chadha, R., Total Synthesis of Taxol. 4. The Final Stages and Completion of the Synthesis. J Am Chem Soc 1995, 117 (2), 653-659. 4. Wender, P. A.; Badham, N. F.; Conway, S. P.; Floreancig, P. E.; Glass, T. E.; Gränicher, C.; Houze, J. B.; Jänichen, J.; Lee, D.; Marquess, D. G.; McGrane, P. L.; Meng, W.; Mucciaro, T. P.; Mühlebach, M.; Natchus, M. G.; Paulsen, H.; Rawlins, D. B.; Satkofsky, J.; Shuker, A. J.; Sutton, J. C.; Taylor, R. E.; Tomooka, K., The Pinene Path to Taxanes. 5. Stereocontrolled Synthesis of a Versatile Taxane Precursor. J Am Chem Soc 1997, 119 (11), 2755-2756. 5. Schuler, M. L., Measuring Taxol production. Nat Biotech 1996, 14 (13), 1635-1635. 6. Goodman, J.; Walsh, V., The story of taxol: nature and politics in the pursuit of an anti- cancer drug. Cambridge University Press: 2001. 7. Ajikumar, P. K.; Xiao, W.-H.; Tyo, K. E. J.; Wang, Y.; Simeon, F.; Leonard, E.; Mucha, O.; Phon, T. H.; Pfeifer, B.; Stephanopoulos, G., Isoprenoid Pathway Optimization for Taxol Precursor Overproduction in Escherichia coli. Science 2010, 330 (6000), 70-74. 8. Arnold, J. P., Origin and history of beer and brewing : from prehistoric times to the beginning of brewing science and technology ; a critical essay. Reprint Edition by BeerBooks: Cleveland, Ohio, 2005. 9. Aleksey, Z., Industrial biocatalysis. Current Opinion in Chemical Biology 2001, 5 (2), 130-136. 10. Straathof, A. J. J.; Adlercreutz, P., Applied biocatalysis. Harwood Academic Publishers: 2000. 11. Nakamura, K.; Kondo, S.-i.; Kawai, Y.; Ohno, A., Asymmetric reduction of ketopantolactone by baker's yeast. Tetrahedron: Asymmetry 1993, 4 (6), 1253-1254. 12. Howarth, J.; James, P.; Dai, J., Immobilized baker's yeast reduction of ketones in an ionic liquid, [bmim]PF6 and water mix. Tetrahedron Letters 2001, 42 (42), 7517-7519. 13. Nakamura, K.; Kondo, S.-i.; Kawai, Y.; Ohno, A., Reduction by bakers' yeast in benzene. Tetrahedron Letters 1991, 32 (48), 7075-7078. 14. Chefson, A.; Zhao, J.; Auclair, K., Sugar-mediated lyoprotection of purified human CYP3A4 and CYP2D6. Journal of Biotechnology 2007, 130 (4), 436-440. 15. Glänzer, B. I.; Faber, K.; Griengl, H., Enantioselective hydrolyses by baker's yeast - III: Microbial resolution of alkynyl esters using lyophilized yeast. Tetrahedron 1987, 43 (24), 5791- 5796. 16. Böhm, C.; Austin, W. F.; Trauner, D., Enzymatic desymmetrization of a centrosymmetric diacetate. Tetrahedron: Asymmetry 2003, 14 (1), 71-74. 17. Manley, J. B.; Anastas, P. T.; Cue Jr, B. W., Frontiers in Green Chemistry: meeting the grand challenges for sustainability in R&D and manufacturing. Journal of Cleaner Production 2008, 16 (6), 743-750. 207

References

18. Kamat, S.; Jacob, J.; Beckman, E. J.; Russell, A. J., Biocatalytic synthesis of acrylates in supercritical fluids. Enzyme and Microbial Technology 1991, 13 (6), 519. 19. Ramesh N, P., Synthesis of chiral pharmaceutical intermediates by biocatalysis. Coordination Chemistry Reviews 2008, 252 (5-7), 659-701. 20. Jung, S. T.; Lauchli, R.; Arnold, F. H., Cytochrome P450: taming a wild type enzyme. Current Opinion in Biotechnology (0). 21. Ugwumba, I. N.; Ozawa, K.; Xu, Z.-Q.; Ely, F.; Foo, J.-L.; Herlt, A. J.; Coppin, C.; Brown, S.; Taylor, M. C.; Ollis, D. L.; Mander, L. N.; Schenk, G.; Dixon, N. E.; Otting, G.; Oakeshott, J. G.; Jackson, C. J., Improving a Natural Enzyme Activity through Incorporation of Unnatural Amino Acids. J Am Chem Soc 2010, 133 (2), 326-333. 22. 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 Microbial Technology 2007, 40 (6), 1451-1463. 23. Sadana, A.; Henley, J. P., Effects of chemical modification on enzymatic activities and stabilities. Biotechnology and Bioengineering 1986, 28 (2), 256-268. 24. Tao, H.; Cornish, V. W., Milestones in directed enzyme evolution. Current Opinion in Chemical Biology 2002, 6 (6), 858-864. 25. Morley, K. L.; Kazlauskas, R. J., Improving enzyme properties: when are closer mutations better? Trends in Biotechnology 2005, 23 (5), 231-237. 26. Wheeldon, I. R.; Campbell, E.; Banta, S., A Chimeric Fusion Protein Engineered with Disparate Functionalities—Enzymatic Activity and Self–assembly. Journal of Molecular Biology 2009, 392 (1), 129-142. 27. McLean, K. J.; Girvan, H. M.; Munro, A. W., Cytochrome P450/redox partner fusion enzymes: biotechnological and toxicological prospects. Expert Opinion on Drug Metabolism & Toxicology 2007, 3 (6), 847-863. 28. Doukyu, N.; Ogino, H., Organic solvent-tolerant enzymes. Biochemical Engineering Journal 2010, 48 (3), 270-282. 29. Qun, J.; Shanjing, Y.; Lehe, M., Tolerance of immobilized baker’s yeast in organic solvents. Enzyme and Microbial Technology 2002, 30 (6), 721-725. 30. Koops, B. C.; Verheij, H. M.; Slotboom, A. J.; Egmond, M. R., Effect of chemical modification on the activity of lipases in organic solvents. Enzyme and Microbial Technology 1999, 25 (7), 622-631. 31. 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. 32. Chefson, A.; Auclair, K., Progress towards the easier use of P450 enzymes. Molecular BioSystems 2006, 2 (10), 462-469. 33. Farinas, E. T.; Bulter, T.; Arnold, F. H., Directed enzyme evolution. Current Opinion in Biotechnology 2001, 12 (6), 545-551. 34. Montellano, P. R. O., Cytochrome P450: structure, mechanism, and biochemistry. Kluwer Academic/Plenum Publishers: 2005. 35. Che, C.-M.; Lo, V. K.-Y.; Zhou, C.-Y.; Huang, J.-S., Selective functionalisation of saturated C-H bonds with metalloporphyrin catalysts. Chemical Society Reviews 2011, 40 (4), 1950-1975. 36. Hasemann, C. A.; Kurumbail, R. G.; Boddupalli, S. S.; Peterson, J. A.; Deisenhofer, J., Structure and function of cytochromes P450:a comparative analysis of three crystal structures. Structure 1995, 3 (1), 41-62. 208

References

37. Park, S.-Y.; Yamane, K.; Adachi, S.-i.; Shiro, Y.; Weiss, K. E.; Maves, S. A.; Sligar, S. G., Thermophilic cytochrome P450 (CYP119) from Sulfolobus solfataricus: high resolution structure and functional properties. Journal of Inorganic Biochemistry 2002, 91 (4), 491-501. 38. (a) Sundaramoorthy, M.; Terner, J.; Poulos, T. L., The crystal structure of chloroperoxidase: a heme peroxidase–cytochrome P450 functional hybrid. Structure 1995, 3 (12), 1367-1378; (b) Gorren, A. C. F.; Mayer, B., Nitric-oxide synthase: A cytochrome P450 family foster child. Biochimica et Biophysica Acta (BBA) - General Subjects 2007, 1770 (3), 432-445. 39. Puchkaev, A. V.; Koo, L. S.; Ortiz de Montellano, P. R., Aromatic stacking as a determinant of the thermal stability of CYP119 from Sulfolobus solfataricus. Archives of Biochemistry and Biophysics 2003, 409 (1), 52-58. 40. Domanski, T. L.; He, Y.-A.; Harlow, G. R.; Halpert, J. R., Dual Role of Human Cytochrome P450 3A4 Residue Phe-304 in Substrate Specificity and Cooperativity. Journal of Pharmacology and Experimental Therapeutics 2000, 293 (2), 585-591. 41. Xue, L.; Wang, H. F.; Wang, Q.; Szklarz, G. D.; Domanski, T. L.; Halpert, J. R.; Correia, M. A., Influence of P450 3A4 SRS-2 Residues on Cooperativity and/or Regioselectivity of Aflatoxin B1 Oxidation. Chemical Research in Toxicology 2001, 14 (5), 483-491. 42. Fowler, S. M.; Taylor, J. M.; Friedberg, T.; Wolf, C. R.; Riley, R. J., CYP3A4 Active Site Volume Modification by Mutagenesis of Leucine 211. Drug Metabolism and Disposition 2002, 30 (4), 452-456. 43. Rittle, J.; Green, M. T., Cytochrome P450 Compound I: Capture, Characterization, and C-H Bond Activation Kinetics. Science 2010, 330 (6006), 933-937. 44. Groves, J. T.; Adhyam, D. V., Hydroxylation by cytochrome P-450 and metalloporphyrin models. Evidence for allylic rearrangement. J Am Chem Soc 1984, 106 (7), 2177-2181. 45. Lewis, D. F. V.; Hlavica, P., Interactions between redox partners in various cytochrome P450 systems: functional and structural aspects. Biochimica et Biophysica Acta (BBA) - Bioenergetics 2000, 1460 (2-3), 353-374. 46. Wang, L.-H.; Kulmacz, R. J., Thromboxane synthase: structure and function of protein and gene. Prostaglandins & Other Lipid Mediators 2002, 68-69 (0), 409-422. 47. Prasad, S.; Mitra, S., Substrate modulates compound I formation in peroxide shunt pathway of Pseudomonas putida cytochrome P450cam. Biochemical and Biophysical Research Communications 2004, 314 (2), 610-614. 48. 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. 49. F, P. G., Catalytic selectivity of human cytochrome P450 enzymes: relevance to drug metabolism and toxicity. Toxicology Letters 1994, 70 (2), 133-138. 50. (a) Van der Plaats, G.; Smit, W. M. A., C-H stretching frequencies and C-H bond strengths of CH3X compounds. Journal of Molecular Structure 1976, 32 (2), 325-337; (b) Novoa, J. J.; Mota, F., The C–H⋯π bonds: strength, identification, and hydrogen-bonded nature: a theoretical study. Chemical Physics Letters 2000, 318 (4-5), 345-354. 51. Newcomb, M.; Hollenberg, P. F.; Coon, M. J., Multiple mechanisms and multiple oxidants in P450-catalyzed hydroxylations. Archives of Biochemistry and Biophysics 2003, 409 (1), 72-79. 52. (a) Chen, M. S.; White, M. C., A Predictably Selective Aliphatic C–H Oxidation Reaction for Complex Molecule Synthesis. Science 2007, 318 (5851), 783-787; (b) Litvinas, N. 209

References

D.; Brodsky, B. H.; Du H2O2. Angewandte Chemie International Edition 2009, 48 (25), 4513-4516. 53. Johnston, J. B.; Ouellet, H.; Podust, L. M.; Ortiz de Montellano, P. R., Structural control of cytochrome P450-catalyzed ω-hydroxylation. Archives of Biochemistry and Biophysics 2011, 507 (1), 86-94. 54. Zerbe, K.; Pylypenko, O.; Vitali, F.; Zhang, W.; Rouset, S.; Heck, M.; Vrijbloed, J. W.; Bischoff, D.; Bister, B.; Süssmuth, R. D.; Pelzer, S.; Wohlleben, W.; Robinson, J. A.; Schlichting, I., Crystal Structure of OxyB, a Cytochrome P450 Implicated in an Oxidative Phenol Coupling Reaction during Vancomycin Biosynthesis. Journal of Biological Chemistry 2002, 277 (49), 47476-47485. 55. Zhu, B. T.; Lee, A. J., NADPH-dependent metabolism of 17β-estradiol and estrone to polar and nonpolar metabolites by human tissues and cytochrome P450 isoforms. Steroids 2005, 70 (4), 225-244. 56. Kells, P. M.; Ouellet, H.; Santos-Aberturas, J.; Aparicio, J. F.; Podust, L. M., Structure of Cytochrome P450 PimD Suggests Epoxidation of the Polyene Macrolide Pimaricin Occurs via a Hydroperoxoferric Intermediate. Chemistry & Biology 2010, 17 (8), 841-851. 57. Shaik, S.; de Visser, S. P.; Ogliaro, F.; Schwarz, H.; Schröder, D., Two-state reactivity mechanisms of hydroxylation and epoxidation by cytochrome P-450 revealed by theory. Current Opinion in Chemical Biology 2002, 6 (5), 556-567. 58. Katsuki, T.; Sharpless, K. B., The first practical method for asymmetric epoxidation. J Am Chem Soc 1980, 102 (18), 5974-5976. 59. Dubois, G.; Murphy, A.; Stack, T. D. P., Simple Iron Catalyst for Terminal Alkene Epoxidation. Organic Letters 2003, 5 (14), 2469-2472. 60. Zhang, W.; Loebach, J. L.; Wilson, S. R.; Jacobsen, E. N., Enantioselective epoxidation of unfunctionalized olefins catalyzed by salen manganese complexes. J Am Chem Soc 1990, 112 (7), 2801-2803. 61. Lin, H.; Liu, J.-Y.; Wang, H.-B.; Ahmed, A. A. Q.; Wu, Z.-L., Biocatalysis as an alternative for the production of chiral epoxides: A comparative review. Journal of Molecular Catalysis B: Enzymatic 2011, 72 (3-4), 77-89. 62. Meunier, B., Metalloporphyrins as versatile catalysts for oxidation reactions and oxidative DNA cleavage. Chemical Reviews 1992, 92 (6), 1411-1456. 63. Groves, J. T.; Myers, R. S., Catalytic asymmetric epoxidations with chiral iron porphyrins. J Am Chem Soc 1983, 105 (18), 5791-5796. 64. Miquel, C., Selective C–H oxidation catalyzed by metalloporphyrins. Coordination Chemistry Reviews (0). 65. Chen, M. S.; White, M. C., Combined Effects on Selectivity in Fe-Catalyzed Methylene Oxidation. Science 2010, 327 (5965), 566-571. 66. Rella, M. R.; Williard, P. G., Oxidation of Peptides by Methyl(trifluoromethyl)dioxirane: The Protecting Group Matters. The Journal of Organic Chemistry 2006, 72 (2), 525-531. 67. Brodsky, B. H.; Du Bois, J., Oxaziridine-Mediated Catalytic Hydroxylation of Unactivated 3° C−H Bonds Using Hydrogen Peroxide. J Am Chem Soc 2005, 127 (44), 15391- 15393. 68. Gideon, G., Cytochromes P450: exploiting diversity and enabling application as biocatalysts. Current Opinion in Chemical Biology 2011, 15 (2), 241-248. 69. Guengerich, F. P., Cytochrome p450 enzymes in the generation of commercial products. Nat Rev Drug Discov 2002, 1 (5), 359-366. 210

References

70. Teoh, K. H.; Polichuk, D. R.; Reed, D. W.; Nowak, G.; Covello, P. S., Artemisia annua L. (Asteraceae) trichome-specific cDNAs reveal CYP71AV1, a cytochrome P450 with a key role in the biosynthesis of the antimalarial sesquiterpene lactone artemisinin. FEBS Letters 2006, 580 (5), 1411-1416. 71. Jennewein, S.; Long, R. M.; Williams, R. M.; Croteau, R., Cytochrome P450 Taxadiene 5α-Hydroxylase, a Mechanistically Unusual Monooxygenase Catalyzing the First Oxygenation Step of Taxol Biosynthesis. Chemistry & Biology 2004, 11 (3), 379-387. 72. Guengerich, F. P., Cytochrome P450 and Chemical Toxicology. Chemical Research in Toxicology 2007, 21 (1), 70-83. 73. Anzenbacherová, E.; Bec, N.; Anzenbacher, P.; Hudeček, J.; Souček, P.; Jung, C.; Munro, A. W.; Lange, R., Flexibility and stability of the structure of cytochromes P450 3A4 and BM-3. European Journal of Biochemistry 2000, 267 (10), 2916-2920. 74. Strolin-Benedetti, M.; Brogin, G.; Bani, M.; Oesch, F.; Hengstler, J. G., Association of cytochrome P450 induction with oxidative stress in vivo as evidenced by 3-hydroxylation of salicylate. Xenobiotica 1999, 29 (11), 1171-1180. 75. Karuzina, I. I.; Archakov, A. I., The oxidative inactivation of cytochrome P450 in monooxygenase reactions. Free Radical Biology and Medicine 1994, 16 (1), 73-97. 76. Rita, B., Cytochromes P450 as versatile biocatalysts. Journal of Biotechnology 2006, 124 (1), 128-145. 77. Ansari, S. A.; Husain, Q., Potential applications of enzymes immobilized on/in nano materials: A review. Biotechnology Advances (0). 78. Wang, Q.; Yang, Z.; Wang, L.; Ma, M.; Xu, B., Molecular hydrogel-immobilized enzymes exhibit superactivity and high stability in organic solvents. Chemical Communications 2007, (10), 1032-1034. 79. Cancilla, M. T.; Leavell, M. D.; Chow, J.; Leary, J. A., Mass spectrometry and immobilized enzymes for the screening of inhibitor libraries. Proceedings of the National Academy of Sciences 2000, 97 (22), 12008-12013. 80. Yim, S.-K.; Ahn, T.; Jung, H.-C.; Pan, J.-G.; Yun, C.-H., Temperature effect on the functional expression of human cytochromes P450 2A6 and 2E1 in <i>Escherichia coli</i>. Archives of Pharmacal Research 2005, 28 (4), 433-437. 81. Bonifacio, A.; Millo, D.; Keizers, P.; Boegschoten, R.; Commandeur, J.; Vermeulen, N.; Gooijer, C.; van der Zwan, G., Active-site structure, binding and redox activity of the heme– thiolate enzyme CYP2D6 immobilized on coated Ag electrodes: a surface-enhanced resonance Raman scattering study. Journal of Biological Inorganic Chemistry 2008, 13 (1), 85-96. 82. Ferrero, V. E. V.; Andolfi, L.; Di Nardo, G.; Sadeghi, S. J.; Fantuzzi, A.; Cannistraro, S.; Gilardi, G., Protein and Electrode Engineering for the Covalent Immobilization of P450 BMP on Gold. Analytical Chemistry 2008, 80 (22), 8438-8446. 83. Ueda, Y.; Morigaki, K.; Tatsu, Y.; Yumoto, N.; Imaishi, H., Immobilization and activity assay of cytochrome P450 on patterned lipid membranes. Biochemical and Biophysical Research Communications 2007, 355 (4), 926-931. 84. Ichinose, H.; Michizoe, J.; Maruyama, T.; Kamiya, N.; Goto, M., Electron-Transfer Reactions and Functionalization of Cytochrome P450cam Monooxygenase System in Reverse Micelles. Langmuir 2004, 20 (13), 5564-5568. 85. Davis, B. G., Chemical modification of biocatalysts. Current Opinion in Biotechnology 2003, 14 (4), 379-386.

211

References

86. Liu, J.-Z.; Wang, M., Improvement of activity and stability of chloroperoxidase by chemical modification. BMC Biotechnology 2007, 7 (1), 23. 87. Wirth, P.; Souppe, J.; Tritsch, D.; Biellmann, J.-F., Chemical modification of horseradish peroxidase with ethanal-methoxypolyethylene glycol: Solubility in organic solvents, activity, and properties. Bioorganic Chemistry 1991, 19 (2), 133-142. 88. (a) Kunz, B. C.; Richter, C., Chemical modification of microsomal cytochrome P450: role of lysyl residues in hydroxylation activity. FEBS Letters 1983, 161 (2), 311-314; (b) Golly, I.; Hlavica, P., Chemical modification of lysine residues in cytochrome P450LM2 (P450IIB4): Influence on heme liganding of arylamines. Archives of Biochemistry and Biophysics 1992, 292 (1), 287-294. 89. Shen, S. J.; Strobel, H. W., Role of Lysine and Arginine Residues of Cytochrome P450 in the Interaction between Cytochrome P4502B1 and NADPH-Cytochrome P450 Reductase. Archives of Biochemistry and Biophysics 1993, 304 (1), 257-265. 90. Adamovich, T. B.; Pikuleva, I. A.; Chashchin, V. L.; Usanov, S. A., Selective chemical modification of cytochrome P-450SCC lysine residues. Identification of lysines involved in the interaction with adrenodoxin. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1989, 996 (3), 247-253. 91. Chashchin, V. L.; Turko, I. V.; Akhrem, A. A.; Usanov, S. A., Cross-linking studies of adrenocortical cytochrome P-450scc. Evidence for a covalent complex with adrenodoxin and localization of the adrenodoxin-binding domain. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1985, 828 (3), 313-324. 92. Yun, C. H.; Hammons, G. J.; Jones, G.; Martin, M. V.; Hopkins, N. E.; Alworth, W. L.; Guengerich, F. P., Modification of cytochrome P450 1A2 enzymes by the mechanism-based inactivator 2-ethynylnaphthalene and the photoaffinity label 4-azidobiphenyl. Biochemistry 1992, 31 (43), 10556-10563. 93. 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. Biochimica et Biophysica Acta (BBA) - Bioenergetics 2010, 1797 (3), 378-390. 94. Li, J.; Li, X.; Ni, X.; Wang, X.; Li, H.; Leong, K. W., Self-assembled supramolecular hydrogels formed by biodegradable PEO–PHB–PEO triblock copolymers and α-cyclodextrin for controlled drug delivery. Biomaterials 2006, 27 (22), 4132-4140. 95. (a) Luckarift, H. R.; Spain, J. C.; Naik, R. R.; Stone, M. O., Enzyme immobilization in a biomimetic silica support. Nat Biotech 2004, 22 (2), 211-213; (b) Ispas, C.; Sokolov, I.; Andreescu, S., Enzyme-functionalized mesoporous silica for bioanalytical applications. Analytical and Bioanalytical Chemistry 2009, 393 (2), 543-554. 96. Poulsen, N.; Berne, C.; Spain, J.; Kröger, N., Silica Immobilization of an Enzyme through Genetic Engineering of the Diatom Thalassiosira pseudonana. Angewandte Chemie 2007, 119 (11), 1875-1878. 97. Cassimjee, K. E.; Trummer, M.; Branneby, C.; Berglund, P., Silica-immobilized His6- tagged enzyme: Alanine racemase in hydrophobic solvent. Biotechnology and Bioengineering 2008, 99 (3), 712-716. 98. Thomas, E. J., Cytochalasan synthesis: macrocycle formation via intramolecular Diels- Alder reactions. Accounts of Chemical Research 1991, 24 (8), 229-235.

212

References

99. Liu, D.; Canales, E.; Corey, E. J., Chiral Oxazaborolidine−Aluminum Bromide Complexes Are Unusually Powerful and Effective Catalysts for Enantioselective Diels−Alder Reactions. J Am Chem Soc 2007, 129 (6), 1498-1499. 100. Siegel, J. B.; Zanghellini, A.; Lovick, H. M.; Kiss, G.; Lambert, A. R.; St.Clair, J. L.; Gallaher, J. L.; Hilvert, D.; Gelb, M. H.; Stoddard, B. L.; Houk, K. N.; Michael, F. E.; Baker, D., Computational Design of an Enzyme Catalyst for a Stereoselective Bimolecular Diels-Alder Reaction. Science 2010, 329 (5989), 309-313. 101. Diels, O.; Alder, K., Synthesen in der hydroaromatischen Reihe. Justus Liebigs Annalen der Chemie 1928, 460 (1), 98-122. 102. Stork, G.; Tamelen, E. E. V. A. N.; Friedman, L. J.; Burgstahler, A. W., CANTHARIDIN. A STEREOSPECIFIC TOTAL SYNTHESIS. J Am Chem Soc 1951, 73 (9), 4501-4501. 103. Gates, M.; Tschudi, G., THE SYNTHESIS OF MORPHINE. J Am Chem Soc 1952, 74 (4), 1109-1110. 104. Nicolaou, K. C.; Snyder, S. A.; Montagnon, T.; Vassilikogiannakis, G., The Diels–Alder Reaction in Total Synthesis. Angewandte Chemie International Edition 2002, 41 (10), 1668- 1698. 105. Alder, K.; Stein, G., Über den sterischen Verlauf von Additions- und Substitutions- reaktionen. I. Zur Stereochemie der Dien-synthese. Gemeinsam mit Dr. Frhr. v. Buddenbrock, Dr. W. Eckardt, Dr. W. Frercks und Dr. St. Schneider. Justus Liebigs Annalen der Chemie 1934, 514 (1), 1-33. 106. Woodward, R. B.; Katz, T. J., The mechanism of the Diels-Alder reaction. Tetrahedron 1959, 5 (1), 70-89. 107. Sundén, H.; Ibrahem, I.; Eriksson, L.; Córdova, A., Direct Catalytic Enantioselective Aza-Diels–Alder Reactions. Angewandte Chemie International Edition 2005, 44 (31), 4877- 4880. 108. Hélène, P., Asymmetric hetero-Diels–Alder reactions of carbonyl compounds. Tetrahedron 2009, 65 (15), 2839-2877. 109. Lin, K.-C.; Liao, C.-C., The hetero Diels-Alder reactions of masked -benzoquinones with nitroso compounds. Chemical Communications 2001, (17), 1624-1625. 110. Fleming, I., Molecular orbitals and organic chemical reactions. John Wiley & Sons: 2010. 111. Esquivias, J.; Arrayás, R. G.; Carretero, J. C., Catalytic Asymmetric Inverse-Electron- Demand Diels−Alder Reaction of N-Sulfonyl-1-Aza-1,3-Dienes. J Am Chem Soc 2007, 129 (6), 1480-1481. 112. Alder, K.; Stein, G., Untersuchungen über den Verlauf der Diensynthese. Angewandte Chemie 1937, 50 (28), 510-519. 113. Ahrendt, K. A.; Borths, C. J.; MacMillan, D. W. C., New Strategies for Organic Catalysis: The First Highly Enantioselective Organocatalytic Diels−Alder Reaction. J Am Chem Soc 2000, 122 (17), 4243-4244. 114. Le, T. X. H.; Bussolari, J. C.; Murray, W. V., 2,5-Diketopiperazines, new chiral auxiliaries for asymmetric Diels-Alder reactions. Tetrahedron Letters 1997, 38 (22), 3849-3852. 115. Rickborn, B., The Retro–Diels– Organic Reactions, John Wiley & Sons, Inc.: 2004.

213

References

116. Otto, S.; Blokzijl, W.; Engberts, J. B. F. N., Diels-Alder Reactions in Water. Effects of Hydrophobicity and Hydrogen Bonding. The Journal of Organic Chemistry 1994, 59 (18), 5372- 5376. 117. Han, L.-B.; Zhao, C.-Q., Stereospecific Addition of H−P Bond to Alkenes: A Simple Method for the Preparation of (RP)-Phenylphosphinates. The Journal of Organic Chemistry 2005, 70 (24), 10121-10123. 118. Auclair, K.; Sutherland, A.; Kennedy, J.; Witter, D. J.; Van den Heever, J. P.; Hutchinson, C. R.; Vederas, J. C., Lovastatin Nonaketide Synthase Catalyzes an Intramolecular Diels−Alder Reaction of a Substrate Analogue. J Am Chem Soc 2000, 122 (46), 11519-11520. 119. Kirkham, J. E. D.; Lee, V.; Baldwin, J. E., Biomimetic synthesis of marine sponge metabolite spiculoic acid A and establishment of the absolute configuration of the natural product. Chemical Communications 2006, (27), 2863-2865. 120. Katayama, K.; Kobayashi, T.; Oikawa, H.; Honma, M.; Ichihara, A., Enzymatic activity and partial purification of solanapyrone synthase: first enzyme catalyzing Diels–Alder reaction. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1998, 1384 (2), 387-395. 121. Ichihara, A., Discoveries of the Diels-Alderases. In Studies in Natural Products Chemistry, Atta ur, R., Ed. Elsevier: 2008; Vol. Volume 34, pp 129-171. 122. Witter, D. J.; Vederas, J. C., Putative Diels−Alder-Catalyzed Cyclization during the Biosynthesis of Lovastatin. The Journal of Organic Chemistry 1996, 61 (8), 2613-2623. 123. (a) Marchand, A. P.; Dong, E. Z.; Shukla, R.; Prasad, A. D.; Bott, S. G., Stereoselectivities of Diels-Alder cycloadditions of tricyclic dienes to MTAD, PTAD, and N- methylmaleimide. Tetrahedron 1997, 53 (44), 14895-14902; (b) Gilchrist, T. L.; d, A. R.; Gonsalves, A. M.; Pinbo; Melo, T. M. V. D., Diels-Alder reactions of 2-azadienes derived from cysteine and serine methyl esters and aldehydes. Tetrahedron 1994, 50 (48), 13709-13724. 124. Evans, D. A.; Bartroli, J.; Shih, T. L., Enantioselective aldol condensations. 2. Erythro- selective chiral aldol condensations via boron enolates. J Am Chem Soc 1981, 103 (8), 2127- 2129. 125. (a) Yadav, J. S.; Bezawada, P.; Chenna, V., Formal total synthesis of borrelidin: synthesis of C1–C11 fragment via desymmetrization strategy. Tetrahedron Letters 2009, 50 (27), 3772- 3775; (b) Sasmal, P. K.; Abbineni, C.; Iqbal, J.; Mukkanti, K., Stereoselective synthesis of an advanced seco ester intermediate as a precursor toward the synthesis of amphidinolides T1, T3, and T4. Tetrahedron 2010, 66 (27-28), 5000-5007; (c) Hiebel, M.-A.; Pelotier, B.; Piva, O., Stereoselective synthesis of the C1–C13 fragment of bistramide A. Tetrahedron Letters 2010, 51 (38), 5091-5093. 126. Benson, S. C.; Snyder, J. K., Optically pure chiral sulfoxides using ephedrine as a chiral auxiliary. Tetrahedron Letters 1991, 32 (42), 5885-5888. 127. Swindell, C. S.; Blase, F. R.; Carroll, P. J., Evaluation of a chiral aryl auxiliary design for sulfur: construction of auxiliary-modified reagents and stereoselection in sulfoxide formation. Tetrahedron Letters 1990, 31 (38), 5405-5408. 128. Tiong, E. A.; Gleason, J. L., Stereoselective Formation of α-Quaternary Stereocenters in the Mannich Reaction. Organic Letters 2009, 11 (8), 1725-1728. 129. Zani, L.; Eichhorn, T.; Bolm, C., Dimethylzinc-Mediated, Enantioselective Synthesis of Propargylic Amines. Chemistry – A European Journal 2007, 13 (9), 2587-2600. 130. (a) Lairson, L. L.; Watts, A. G.; Wakarchuk, W. W.; Withers, S. G., Nat. Chem. Biol. 2006, 2 (null), 724; (b) Braunegg, G.; de Raadt, A.; Feichtenhofer, S.; Griengl, H.; Kopper, I. I.; 214

References

Lehmann, A.; Weber, H. J., Angew. Chem., Int. Ed. 1999, 38 (null), 2763; (c) de Raadt, A.; Griengl, H.; Weber, H., Chem.—Eur. J. 2001, 7 (null), 27; (d) de Raadt, A.; Griengl, H., Curr. Opin. Biotechnol. 2002, 13 (null), 537. 131. Li, S.; Chaulagain, M. R.; Knauff, A. R.; Podust, L. M.; Montgomery, J.; Sherman, D. H., Proc. Natl. Acad. Sci. U.S.A. 2009, 106 (null), 18463. 132. Renwick, A. B.; Surry, D.; Price, R. J.; Lake, B. G.; Evans, D. C., Metabolism of 7- benzyloxy-4-trifluoromethylcoumarin by human hepatic cytochrome P450 isoforms. Xenobiotica 2000, 30 (10), 955-969. 133. Salminen, K. A.; Leppänen, J.; Venäläinen, J. I.; Pasanen, M.; Auriola, S.; Juvonen, R. O.; Raunio, H., Simple, Direct, and Informative Method for the Assessment of CYP2C19 Enzyme Inactivation Kinetics. Drug Metabolism and Disposition 2011, 39 (3), 412-418. 134. Wilkinson, J. M.; Pollard, I., Accumulation of theophylline, theobromine and paraxanthine in the fetal rat brain following a single oral dose of caffeine. Developmental Brain Research 1993, 75 (2), 193-199. 135. Greene, T. W.; Wuts, P. G. M., Protective groups in organic synthesis. Wiley: 1991. 136. Dale, J. A.; Dull, D. L.; Mosher, H. S., .alpha.-Methoxy-.alpha.- trifluoromethylphenylacetic acid, a versatile reagent for the determination of enantiomeric composition of alcohols and amines. The Journal of Organic Chemistry 1969, 34 (9), 2543-2549. 137. Aresta, A.; Palmisano, F.; Zambonin, C. G., Simultaneous determination of caffeine, theobromine, theophylline, paraxanthine and nicotine in human milk by liquid chromatography with diode array UV detection. Food Chemistry 2005, 93 (1), 177-181. 138. Maves, S. A.; Sligar, S. G., Understanding thermostability in cytochrome P450 by combinatorial mutagenesis. Protein Science 2001, 10 (1), 161-168. 139. Hoye, T. R.; Jeffrey, C. S.; Shao, F., Mosher ester analysis for the determination of absolute configuration of stereogenic (chiral) carbinol carbons. Nat. Protocols 2007, 2 (10), 2451-2458. 140. Akinnusi, T. O.; Vong, K.; Auclair, K., Geminal dialkyl derivatives of N-substituted pantothenamides: Synthesis and antibacterial activity. Bioorganic & Medicinal Chemistry 2011, 19 (8), 2696-2706. 141. Yano, J. K.; Wester, M. R.; Schoch, G. A.; Griffin, K. J.; Stout, C. D.; Johnson, E. F., The Structure of Human Microsomal Cytochrome P450 3A4 Determined by X-ray Crystallography to 2.05-Å Resolution. Journal of Biological Chemistry 2004, 279 (37), 38091- 38094. 142. de Graaf, C.; Vermeulen, N. P. E.; Feenstra, K. A., Cytochrome P450 in Silico: An Integrative Modeling Approach. Journal of Medicinal Chemistry 2005, 48 (8), 2725-2755. 143. He, C.; Long, Y.; Pan, J.; Li, K.; Liu, F., Application of molecularly imprinted polymers to solid-phase extraction of analytes from real samples. Journal of Biochemical and Biophysical Methods 2007, 70 (2), 133-150. 144. (a) Izenberg, N. R.; Murrray, G. M.; Pilato, R. S.; Baird, L. M.; Levin, S. M.; Van Houten, K. A., Astrobiological molecularly imprinted polymer sensors. Planetary and Space Science 2009, 57 (7), 846-853; (b) Yang, Y.; Lai, E. P. C., An investigation of porous structure in molecularly imprinted polymer for sensor development: Non-linear fluorescence quenching of 17β-estradiol bound inside MIP submicron particles by sodium nitrite and methacrylamide. Journal of Photochemistry and Photobiology A: Chemistry 2010, 213 (2-3), 123-128.

215

References

145. Cirillo, G.; Curcio, M.; Parisi, O. I.; Puoci, F.; Iemma, F.; Spizzirri, U. G.; Restuccia, D.; Picci, N., Molecularly imprinted polymers for the selective extraction of glycyrrhizic acid from liquorice roots. Food Chemistry 2011, 125 (3), 1058-1063. 146. Bunte, G.; Hürttlen, J.; Pontius, H.; Hartlieb, K.; Krause, H., Gas phase detection of explosives such as 2,4,6-trinitrotoluene by molecularly imprinted polymers. Analytica Chimica Acta 2007, 591 (1), 49-56. 147. (a) Sergeyeva, T. A.; Slinchenko, O. A.; Gorbach, L. A.; Matyushov, V. F.; Brovko, O. O.; Piletsky, S. A.; Sergeeva, L. M.; Elska, G. V., Catalytic molecularly imprinted polymer membranes: Development of the biomimetic sensor for phenols detection. Analytica Chimica Acta 2010, 659 (1-2), 274-279; (b) Ramström, O.; Mosbach, K., Synthesis and catalysis by molecularly imprinted materials. Current Opinion in Chemical Biology 1999, 3 (6), 759-764. 148. Theodoridis, G.; Manesiotis, P., Selective solid-phase extraction sorbent for caffeine made by molecular imprinting. Journal of Chromatography A 2002, 948 (1-2), 163-169. 149. Wang, D.; Hong, S.; Yang, G.; Row, K., Caffeine Molecular Imprinted Microgel Spheres by Precipitation Polymerization. Korean Journal of Chemical Engineering 2003, 20 (6), 1073- 1076. 150. Shimizu, K. D.; CHEMISTRY, S. C. U. C. D. O.; BIOCHEMISTRY., Characterization of MIPs Using Heterogeneous Binding Models. Defense Technical Information Center: 2002. 151. H.-J, F., 2.17 - Chiral Drug Discovery and Development – From Concept Stage to Market Launch. In Comprehensive Medicinal Chemistry II, Editors-in-Chief: John, B. T.; David, J. T., Eds. Elsevier: Oxford, 2007; pp 713-736. 152. Faber, K.; Franssen, M. C. R., Prospects for the increased application of biocatalysts in organic transformations. Trends in Biotechnology 1993, 11 (11), 461-470. 153. Beilen, J. B. v.; Li, Z., Enzyme technology: an overview. Current Opinion in Biotechnology 2002, 13 (4), 338-344. 154. Thörn, H. A.; Lundahl, A.; Schrickx, J. A.; Dickinson, P. A.; Lennernäs, H., Drug metabolism of CYP3A4, CYP2C9 and CYP2D6 substrates in pigs and humans. European Journal of Pharmaceutical Sciences 2011, 43 (3), 89-98. 155. Sun, H.; Bessire, A. J.; Vaz, A., Dirlotapide as a model substrate to refine structure-based drug design strategies on CYP3A4-catalyzed metabolism. Bioorganic & Medicinal Chemistry Letters (0). 156. Antman, E. M.; Dupont, W. D.; Bonalsky, J.; Califf, R. M.; Corwin, S.; Fink, L.; Hansen, D. E.; Kawanishi, D. T.; Kronenberg, M. W.; McKay, C. R.; Robertson, G. C.; Tenzer, M. L.; Touchon, R.; Willerson, J. T., Early treatment with intravenous metoprolol for suspected acute myocardial infarction: a Phase IV United States trial. International Journal of Cardiology 1989, 23 (2), 185-197. 157. Kamondi, S.; Szilágyi, A.; Barna, L.; Závodszky, P., Engineering the thermostability of a TIM-barrel enzyme by rational family shuffling. Biochemical and Biophysical Research Communications 2008, 374 (4), 725-730. 158. Weast, R. C.; Grasselli, J. G., Handbook of data on organic compounds. CRC Press: 1992. 159. Daly, J. W.; Padgett, W. L.; Shamim, M. T., Analogs of caffeine and theophylline: effect of structural alterations on affinity at adenosine receptors. J. Med. Chem. 1986, 29 (Copyright (C) 2011 American Chemical Society (ACS). All Rights Reserved.), 1305-8.

216

References

160. Madyastha, K. M.; Sridhar, G. R., Highly efficient C-8 oxidation of substituted xanthines with substitution at the 1-, 3-, and 7- positions using biocatalysts. J. Chem. Soc., Perkin Trans. 1 1999, (Copyright (C) 2011 American Chemical Society (ACS). All Rights Reserved.), 677-680. 161. Smith, A. B., III; Fox, R. J.; Vanecko, J. A., (+)-Sorangicin A Synthetic Studies. Construction of the C(1-15) and C(16-29) Subtargets. Org. Lett. 2005, 7 (Copyright (C) 2011 American Chemical Society (ACS). All Rights Reserved.), 3099-3102. 162. Cruickshank, P. A.; Fishman, M., Studies in alkylation. II. Reactions of epoxyalkyl bromides. The Journal of Organic Chemistry 1969, 34 (12), 4060-4065. 163. Tokunaga, M.; Larrow, J. F.; Kakiuchi, F.; Jacobsen, E. N., Asymmetric Catalysis with Water: Efficient Kinetic Resolution of Terminal Epoxides by Means of Catalytic Hydrolysis. Science 1997, 277 (5328), 936-938. 164. Scheideman, M.; Wang, G.; Vedejs, E., Amine-Directed Hydroboration: Scope and Limitations. J Am Chem Soc 2008, 130 (27), 8669-8676.

217

Appendix

Apendix: Select NMR and HPLC Spectra

218

Select HPLC Spectra

219

Select HPLC spectra

Compound 4.18

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 4.18 with CHP: 2 products detected.

220

Select HPLC spectra

Compound 5.5

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.5 with CHP: 1 product detected.

Reinjection on a chiral column of the major product (5.52) purified after enzymatic transformation of substrate 5.5 (Chiral Method A).

DAD1 A, Sig=273,16 Ref =360,100 (ALS2-51\51P1C3A4.D) mAU

160 7.870

140 Area: 1843.14

120

100

8.227 Area: 825.004

0 5 10 15 20 25 min

221

Select HPLC spectra

Compound 5.6

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.6 with CHP: 2 products detected.

Reinjection on a chiral column of the major product (5.53) purified after enzymatic transformation of substrate 5.6 (Chiral Method A).

DAD1 A, Sig=273,16 Ref =360,100 (AL13-17\AL121-PC.D)

mAU 8.559 80

Area: 1273.84

30 9.371 Area: 459.032

0 5 10 15 20 25 min

222

Select HPLC spectra

Compound 5.9 HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.9 with CHP: 1 major and 1 minor product detected.

Reinjection on a chiral column of the major product (5.68) purified after enzymatic transformation of substrate 5.9 (Chiral Method A).

223

Select HPLC spectra

Compound 5.10

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.10 with CHP: 1 product detected.

Reinjection on a chiral column of the major product (5.69) purified after enzymatic transformation of substrate 5.10 (Chiral Method A).

224

Select HPLC spectra

Compound 5.11

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.11 with CHP: 1 product detected.

225

Select HPLC spectra

Compound 5.12

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.12 with CHP: 2 products detected.

226

Select HPLC spectra

Compound 5.15

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.15 with CHP: 1 product detected.

Reinjection on a chiral column of the major product (5.71) purified after enzymatic transformation of substrate 5.15 (Chiral Method A).

227

Select HPLC spectra

Compound 5.34

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.34 with CHP: 1 product detected.

228

Select HPLC spectra

Compound 5.35

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.35 with CHP: 1 product detected.

229

Select HPLC spectra

Compound 5.36

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.36 with CHP: 1 product detected.

Reinjection on a chiral column of the major product (5.66) purified after enzymatic transformation of substrate 5.35 (Chiral Method A).

DAD1 A, Sig=273,16 Ref =360,100 (AL2-104F\3A4PURE.D) mAU

7.678Area: 109.82 5

-5 0 2 4 6 8 10 12 14 min

230

Select HPLC spectra

HPLC trace for the synthetic enantioenriched standard of 5.56 (Chiral Method A).

DAD1 A, Sig=273,16 Ref =360,100 (AL2-104F\3-132P.D) mAU

2 6.973 Area: 48.926

7.443 Area: 27.0801 0

-2 0 2 4 6 8 10 12 14 min

Co-injection of enzymatic transformation mixture with synthetic standard (Chiral Method A)

231

Select HPLC spectra

Compound 5.37

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.37 with CHP: 2 products detected.

Reinjection on a chiral column of the major product (5.54) purified after enzymatic transformation of substrate 5.37 (Chiral Method A).

DAD1 A, Sig=273,16 Ref =360,100 (ALS2-103\P1C.D) mAU

16 8.448

Area: 208.03 14

9.056 Area: 102.003

-2 0 5 10 15 20 25 min 232

Select HPLC spectra

Compound 5.38

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.38 with CHP: 1 product detected.

Reinjection on a chiral column of the major product (5.72) purified after enzymatic transformation of substrate 5.38 (Chiral Method A).

233

Select HPLC spectra

Compound 5.75

HPLC trace from LC-MS analysis (Non-Chiral Method B, MW observed is given) of the reaction mixture for the enzymatic transformation of 5.75 with CHP: 1 product detected.

234

Select NMR Spectra

Select NMR spectra

235

Select NMR spectra

Compound 4.20

236

Select NMR spectra

Compound 4.21

237

Select NMR spectra

Compound 5.5

Compound 5.6

238

Select NMR spectra

239

Select NMR spectra

Compound 5.9

240

Select NMR spectra

Compound 5.10

241

Select NMR spectra

Compound 5.11

242

Select NMR spectra

Compound 5.12

13C NMR signals derived from HSQC

243

Select NMR spectra

Compound 5.15

244

Select NMR spectra

Compound 5.20

13C NMR signals derived from HSQC

245

Select NMR spectra

Compound 5.21

13C NMR signals derived from HSQC

246

Select NMR spectra

Compound 5.29

247

Select NMR spectra

Compound 5.31

248

Select NMR spectra

Compound 5.34

249

Select NMR spectra

Compound 5.35

250

Select NMR spectra

Compound 5.36

251

Select NMR spectra

Compound 5.37

252

Select NMR spectra

Compound 5.38

13C NMR signals derived from HSQC

253

Select NMR spectra

Compound 5.47

13C NMR signals are from HSQC

254

Select NMR spectra

Compound 5.52

13C NMR signals derived from HSQC

255

Select NMR spectra

Compound 5.53

13C NMR signals are from HSQC

256

Select NMR spectra

Compound 5.54

257

Select NMR spectra

Compound 5.65

13C NMR signal from HSQC

258

Select NMR spectra

Compound 5.66

259

Select NMR spectra

Compound 5.67

260

Select NMR spectra

261