Post-Synthetic Functionalization of Bioactive Compounds for Rapid Anticancer

Post-synthetic Functionalization of Bioactive Compounds for Rapid Anticancer

Library Expansion and Mechanistic Probe Development for Antimicrobial

Resistance

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Chido M. Hambira

Graduate Program in Pharmaceutical Sciences

The Ohio State University

2018

Dissertation Committee:

Professor James Fuchs Advisor

Professor David Nagib Co-Advisor

Professor Karl Werbovetz

Professor Mark Mitton-Fry Copyright by

Chido M. Hambira

2018 Abstract

Within the realm of medicinal chemistry, not only is it important to optimize for target potency and physicochemical properties of bioactive compounds, but development of enabling technologies that aid in elucidation of biochemical pathways is of equal importance. The body of work contained within this dissertation is a summary of my efforts toward the development of a new C-H functionalization method to facilitate late-stage derivatization of complex bioactive molecules, and post-synthetic modifications of known bioactive compounds of pharmaceutical agents for the development of mechanistic probes.

By taking advantage of the versatile reactivity of hypervalent iodine, aided by the labile nature of the ligands, we have harnessed the mild yet enhanced reactivity of iodosobenzene chloroacetate for the chlorination of (hetero)arenes. My key contributions to this project were the discovery of methods amenable to the halogenation of challenging substrates of the (iso)quinoline class of compounds, and the cytotoxic natural product derivate 2”-acetyl phyllanthusmin D, in addition to conducting mechanistic investigations.

A significant portion of my doctoral research has been spent developing tool compounds targeting two aspects contributing to the looming public health issue of antimicrobial resistance. The first effort was aimed at the investigation of small molecules transport across Gram-negative bacterial membranes. This work was conducted during a

12-month co-op at GlaxoSmithKline. We explored the feasibility of hijacking a native

ii nutrient uptake system, the FadL/FadD-mediated uptake of long chain fatty acids to enable uptake of antimicrobial agents across the Gram-negative membranes. These investigations demonstrated the first proof-of-principle that a post-synthetic attachment of a long fatty acid chain to an antimicrobial agent could facilitate its uptake and cellular accumulation in a FadD-dependent manner. The potential impact of this work lies in conferring broad spectrum activity to Gram-positive only agents by simply attaching a long chain fatty acid recognition element.

In addition to the significant drug development challenge imposed by the Gram- negative membranes, resistance mechanisms can develop and propagate within persistent bacterial biofilms impervious to therapeutic intervention. One example of an infection that relies very heavily on the biofilm lifestyle is that caused by S. Typhi, the causative agent of typhoid fever. In an attempt to find new ways to prevent formation of biofilms caused by S. Typhi, structurally distinct compounds, JK-1 and T315, were identified from compound libraries as antibiofilm agents. Several analogues to JK-1 were synthesized in an attempt to build initial structure activity relationships and improve the potency. In addition to analogue development of this class, two T315 mechanistic probes bearing a biotin tag were synthesized and subjected to affinity-based proteomics. These investigations led to the identification of a NADH-dependent oxidoreductase, WrbA as one of the protein targets implicated in the biofilm lifestyle of S. Typhi. This discovery is expected to inform rational analogue design to transform these initial hit compounds into viable drug development leads for the eradication of biofilm-related chronic infections.

iii Dedication

This document is dedicated to my family. To God be the Glory.

iv Acknowledgments

I would like to express my deepest gratitude to my advisor Dr. James R. Fuchs for his continued guidance and unwavering support throughout the years. Thank you for challenging me intellectually, encouraging my creativity, and helping me develop into the medicinal chemist that I am today. It has been an absolute joy working with you and I will forever be grateful for laying the foundation necessary for me to address tough research challenges in my professional career moving forward. I am also grateful to my dissertation committee, Dr. David A. Nagib, Dr. Karl Werbovetz, Dr. Mark Mitton-Fry, for their tremendous guidance and supervision during my graduate career and for helping me ensure that this document was a correct and accurate representation of my work and of myself.

I am additionally very grateful to both past and present (2012-2018) colleagues in the Fuchs lab for providing an ideal work environment for intellectual stimulation and for all your support in various ways. I will never forget the human touch in this lab. I would specifically like to acknowledge Dr. Nivedita (Nivi) Jena for your supervision in the lab and for being an amazing mentor. I would like to express my gratitude to Dr. John S. Gunn, and his lab, particularly Jasmine Moshiri, and Darpan Kaur for an amazing collaboration for the last 3.5 years.

v My sincere gratitude goes to Dr. David A. Nagib for allowing me to learn a new skill-set in your lab. I am grateful for incorporating me into your lab culture and for guiding me throughout the project. The collaborative project in your lab became a labor of love and

I am thankful to my colleagues in the Nagib lab for all their support and energetic welcome.

I am especially grateful to Stacy Fosu, and Andrew Chen for all their hard-work on the aryl

C-H functionalization project and for their patience and support throughout the project. A special thanks goes out to Dr. Lu Wang – you’re the best. Thank you to the Department of

Chemistry and Biochemistry for providing a safe and well equipped lab home and for treating me like one of your own.

I would like to thank Dr. Fuchs and Dr. Werbovetz and for supporting my one year research experience at GlaxoSmithKline (GSK), PA. This opportunity would not have been possible with the collective effort of you, including various administrators at The Ohio

State University College of Pharmacy. I am incredibly grateful to my manager at GSK, Dr.

Rob Stavenger, for guiding me throughout this internship, and giving me many opportunities to further develop as a well-rounded scientist. I am very thankful to my coworkers at GSK, Dr. Xiangmin Liao, Dr. Pan Chan and Dr. Nabil Abraham for supporting me during this time.

To all my friends at The Ohio State University, my sincere gratitude goes out to you for helping me get through the tough days and the good days. I am especially thankful to Jason Fang, Julian Richard, Janet Antwi, Tehane Ali, Andrew Chen, and Jeremy Lear. I could not have done this without you. You are all very special to me in ways I could never capture with words. From the depth of my heart, thank you.

vi To the Ohio State University College of Pharmacy, thank you for providing me with the necessary tools required for the successful completion of this dissertation and for providing financial support throughout my graduate career in the form of the Albeit H.

Soloway Graduate Endowed Fund in Pharmacy and Cancer Research, and the Jane Chen

Fellowship in Medicinal Chemistry and Pharmacognosy. I am grateful to Dr. Craig

McElroy for all your help with instrumental analysis and challenging characterizations

Lastly, I would like to thank my family for the inspiration, love, support, prayers, comfort, and well wishes. I hope I have made you proud. Thank you to my extended family members Nancy Ryan and Debbie Splaingard for supporting me and looking out for me throughout the years. I am forever grateful. Thank you to all my friends from back home in Zimbabwe, and from Berea College for keeping me grounded and keeping me laughing.

Thank you Jesus for sustaining me throughout this experience.

vii Vita

2012...... B.A. Chemistry, Berea College

2014...... M.Sc. Pharmaceutical Chemistry, The Ohio

...... State University

2012 to present ...... Graduate Teaching Associate, Graduate

Research Fellow, and Graduate Research

Associate, Division of Medicinal Chemistry

and Pharmacognosy, The Ohio State

University

Publications

1. Fosu, S.;† Hambira, C. M.;† Chen, A. D.; Fuchs, J. R.; Nagib, D. A. Site-selective C-H functionalization of (hetero)arenes via transient, non-symmetric iodanes. †Authors contributed equally. Chem – Cell Press, Accepted November 8th, 2018. 2. Moshiri, J. S.; Kaur, D.; Hambira, C. M.; Sandala, J. L.; Koopman, J. A.; Fuchs, J. R.; Gunn, J. S. Identification of a small molecule anti-biofilm agent against Salmonella enterica. Front. Microbiol. 2018, 9, 2804.

Fields of Study

Major Field: Pharmaceutical Sciences

viii Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgments...... v

Vita ...... viii

Publications ...... viii

Fields of Study ...... viii

Table of Contents ...... ix

List of Tables ...... xiv

List of Figures ...... xv

List of Schemes ...... xviii

Chapter 1. Late-stage Functionalization of Bioactive Molecules ...... 1

1.1. Introduction ...... 2 ix 1.1.1 Late-stage functionalization ...... 3

1.2 Developing a late-stage functionalization strategy ...... 7

1.2.1 Directed C-H functionalization ...... 8

1.2.2 Innate reactivity C-H functionalization ...... 10

1.3 Synthesis of tool compounds by late-stage C-H Functionalization ...... 12

1.4 Late-stage C-H Functionalization for Bioconjugation ...... 15

1.5 Conclusion ...... 21

Chapter 2. Iodane-mediated late-stage (hetero)aryl C-H functionalization ...... 22

2.1 Natural product-derived anticancer agents ...... 23

2.2 Arylnaphthalene lignan lactones in nature ...... 23

2.3 Phyllanthusmins as novel anticancer agents ...... 25

2.3.1 Isolation of phyllanthusmins A – E ...... 26

2.3.2 Phyllanthusmin structural considerations ...... 27

2.4 Strategies toward the access of arylnaphthalene cores ...... 30

2.4.1 Narender’s silver (I)-catalyzed radical cyclization ...... 31

2.4.2 Vidal’s Brønsted acid-catalyzed benzannulation ...... 33

2.5 Iodane activation enables direct C-H functionalization of arenes ...... 38

2.5.1 Activation of hypervalent iodine reagents ...... 38

2.5.2 Development of aryl-C-H chlorination conditions ...... 42

2.6 Structural considerations of halogenated Phy-4 analogues ...... 52

2.7 Mechanistic investigations ...... 54

2.8 Conclusion ...... 62

Chapter 3. Hijacking the Fatty Acid Transport System to Facilitate Drug Penetration

Across Bacterial Membranes ...... 63

3.1 Penetration across bacterial membranes ...... 65

3.1.1 Introduction ...... 65

3.1.2 Bacterial membrane structural considerations ...... 69

3.1.3 Strategies to permeate the outer membrane ...... 71

3.2 Iron as an essential micronutrient for bacteria ...... 75

3.3 Iron-siderophore uptake by Gram-negative bacteria: structural considerations ..... 76

3.4 Identification of fatty acid transport system as a promising uptake pathway ...... 80

3.4.1 Fatty acid uptake in E. coli ...... 81

3.4.2 FadL structural considerations ...... 82

3.5 Hijacking the fatty acid uptake system: experimental design ...... 85

3.5.1 In vitro fluorescence imaging of E. coli by BODIPY-fatty acid conjugates .... 85

3.5.2 Fatty acid conjugate synthesis and biological evaluation (SAR) ...... 87

3.5.3 In vitro evaluation: discussion ...... 92

3.5.4 In vivo activity of cipro-fatty acid conjugates ...... 92

3.6 Conclusion ...... 94

Chapter 4. Combatting Antimicrobial Resistance through Biofilm Interference ...... 96

4.1 Formation of bacterial biofilms ...... 97

4.1.1 Impact of biofilms on development of S. Typhi chronic carriage state ...... 98

4.2 Development of new anti-biofilm agents ...... 99

4.2.1 Anti-biofilm natural products ...... 100

4.2.2 Discovery of JK-1 and primary activity profile...... 102

4.2.3 Retrosynthetic strategy to JK-1 and pyrazole analogs ...... 103

4.2.4 Optimization of arylisothiocyanate synthesis ...... 104

4.2.5 Synthesis of JK-1 and pyrazole analogs ...... 106

4.3 Biological activity evaluation of JK-1 ...... 108

4.4 Identification of a small molecule (OSU-T315) anti-biofilm agent against

Salmonella ...... 110

4.4.1 Activity profile of T315 against S. Typhi, S. Typhimurium and A. baumannii

...... 111

4.4.2 Primary biological data for T315 anti-biofilm activity ...... 112

4.4.3 Effect on biofilm formation of T315 and ciprofloxacin combination ...... 113

4.5 Development of T315 mechanistic probes for target identification ...... 114

4.5.1 Synthesis of biotinylated probe T315-S1 ...... 115

xii

4.5.2 Synthesis of biotinylated probe T315-S2 ...... 117

4.6 Results of affinity-based pull-down and proteomic analysis ...... 120

4.7 WrbA in biofilm development: discussion ...... 123

4.8 Conclusion ...... 124

Chapter 5: Experimental Section ...... 125

5.1 Synthesis and characterization of C-H functionalization substrates ...... 126

5.2 Synthesis and characterization of ciprofloxacin-fatty acid conjugates ...... 159

5.3 Synthesis and characterization of anti-biofilm compounds ...... 179

References ...... 209

Appendix: NMR Spectra of Selected Compounds ...... 225

xiii

List of Tables

Table 2.1 Biological activity evaluation of phyllanthusmins against HT-29 cell line ...... 28

Table 2.2 Pivanilide C-H functionalization using Bronsted acids ...... 42

Table 2.3 Chlorination of arenes-substrate scope ...... 44

Table 2.4 Chlorination of heteroarenes ...... 46

Table 2.5 Iodane-mediated chlorination of PHY-4 ...... 50

Table 2.6 Investigation of PHY-4 bromination conditions ...... 52

Table 2.7 Comparison to known aryl C-H chlorinating reagents ...... 56

Table 2.8 Investigation of N-donor iodonium as possible reaction intermediate ...... 58

Table 3.1 Minimum inhibitory concentrations of alkyl-linked cipro-FA conjugates ...... 88

Table 3.2 Minimum inhibitory concentrations of triazole-linked cipro-FA conjugates ... 90

Table 3.3 Minimum inhibitory concentrations of amide-linked cipro-FA conjugates ..... 91

Table 4.1 Biological activity of JK-1 pyrazole analogs ...... 109

Table 4.2 Carboxybenzyl deprotection of T315-S2 ...... 119

xiv

List of Figures

Figure 1.1 Conceptual C-H bond functionalization ...... 4

Figure 1.2 Step-count reduction in total syntheses enabled by C-H functionalization ...... 6

Figure 1.3 Directing-group mediated C-H functionalization of Celecoxib ...... 9

Figure 1.4 C-H functionalization of quinine by innate reactivity ...... 11

Figure 1.5 Development of photo-crosslinking functionalization tags ...... 16

Figure 1.6 Development of chemical probes from natural products ...... 18

Figure 2.1 Lignan skeleton ...... 24

Figure 2.2 Biosynthetic origins of diphyllin and podophyllotoxin ...... 25

Figure 2.3 Structures of phyllanthusmins A-C ...... 27

Figure 2.4 General structure of the phyllanthusmins ...... 28

Figure 2.5 Structural similarity between PHY-4 and etoposide ...... 29

Figure 2.6 Antiproliferative activity in HT-29 cell line (IC50 in M)...... 36

Figure 2.7 Structure-activity relationship summary of phyllanthusmin analogues ...... 37

Figure 2.8 Generation of non-symmetrical iodane hypothesis ...... 39

1 Figure 2.9 H NMR (400 MHz, CDCl3) analysis of ligand exchange after 15 minutes ... 41

Figure 2.10 Chloride source investigation for heteroaryl chlorination ...... 45

xv Figure 2.11 Halogenation of pharmaceuticals and natural products ...... 47

Figure 2.12 Chlorination of arenes and heteroarenes competition experiment ...... 48

Figure 2.13 Biological activity evaluation of brominated PHY-4 ...... 53

Figure 2.14 DFT Computational analysis of regioselectivity ...... 61

Figure 3.1 Cell wall penetration of antibiotics in GP and GN bacteria...... 68

Figure 3.2 Cell wall structure of representative GN bacteria E. coli...... 70

Figure 3.3 Effect of primary amines on accumulation in GN bacteria ...... 73

Figure 3.4 Effect of rigidity and globularity on accumulation in GN bacteria ...... 75

Figure 3.5 Common classes of naturally occurring siderophores ...... 78

Figure 3.6 Iron-siderophore uptake in E. coli ...... 79

Figure 3.7 Iron-binding antibiotics in recent clinical development ...... 80

Figure 3.8 Overview fatty acid uptake in Gram-negative bacteria ...... 81

Figure 3.9 Crystal structure of FadL ...... 83

Figure 3.10 Schematic mechanism for transport through FadL ...... 84

Figure 3.11 Fluorescently-labeled fatty acids in E. coli ...... 86

Figure 3.12 C8-Cipro in vivo efficacy in immunocompromised rodent urinary tract

infection model...... 93

Figure 3.13 Revised model of Cipro-FA conjugates ...... 94

Figure 4.1 Stages of bacterial biofilm formation ...... 97

Figure 4.2 Anti-biofilm natural products ...... 101

Figure 4.3 Identification of JK-1 from ATP mimetic library ...... 102

Figure 4.4 OSU-315 anti-biofilm activity against S. Typhimurium ...... 111

xvi Figure 4.5 Activity profile of T315 against S. Typhimurium. (A) Cell viability. (B)

Temporal addition assay...... 112

Figure 4.6 Effect of combination therapy on biofilm development. (A) S. Typhimurium.

(B) S. Typhi...... 114

Figure 4.7 Strategy toward the synthesis of T315-S1 probe ...... 115

Figure 4.8 T315 pulldown flow-through analysis by SDS-PAGE ...... 121

Figure 4.9 (A) S. Typhimurium wrbA biofilm formation. (B) Effect of T315 on WrbA.

...... 122

xvii

List of Schemes

Scheme 1.1 Total syntheses enabled by palladium-catalyzed cross-coupling reactions ... 5

Scheme 1.2 Tritiation of bioactive compounds for metabolism studies ...... 12

Scheme 1.3 18F-labelling of bioactive compounds for PET studies ...... 13

Scheme 1.4 Aldehyde oxidase mechanistic probe developed by the Baran lab ...... 14

Scheme 1.5 Site-selective allylic C-H amination of euplamerin developed by the Romo

lab ...... 19

Scheme 1.6 Native chemical tagging of antibodies ...... 20

Scheme 2.1 Methods to construct arylnaphthalene cores I ...... 30

Scheme 2.2 Silver (I)-catalyzed regioselective synthesis of lignan cores ...... 32

Scheme 2.3 Methods to construct arylnaphthalene cores II ...... 34

Scheme 2.4 Charlton’s diphyllin synthesis ...... 35

Scheme 2.5 Activation of iodobenzene diacetate via desymmetrization ...... 40

Scheme 2.6 Iodane-mediated chlorination of PHY-4 ...... 49

Scheme 2.7 Iodane-mediated chlorination of acetyl-diphyllin ...... 50

Scheme 2.8 Iodane-mediated chlorination of PHY-4 ...... 51

Scheme 2.9 Halogenation mechanistic proposals ...... 54

Scheme 2.10 Desymmetrization of iodobenzene dichloride by trifluoroacetic acid ...... 55

xviii Scheme 2.11 Diaryliodonium as possible reaction intermediate ...... 57

Scheme 2.12 Divergent regioselectivity of PHY-4 halogenation ...... 59

Scheme 3.1 Synthesis of alkyl-linked Cipro-FA conjugates ...... 88

Scheme 3.2 Synthesis of triazole-linked Cipro-FA conjugates ...... 90

Scheme 3.3 Synthesis of amide-linked Cipro-FA conjugates ...... 91

Scheme 4.1 Retrosynthetic analysis of JK-1 ...... 103

Scheme 4.2 Investigation into conversion of arylamine into isothiocyanate ...... 105

Scheme 4.3 Synthesis of JK-1 ...... 107

Scheme 4.4 Hydroxypyrazole tautomeric forms ...... 107

Scheme 4.5 Synthesis of T315-S1 probe ...... 116

Scheme 4.6 Synthesis of CBz-protected T315-S2 probe ...... 118

Scheme 4.7 Successful synthesis of T315-S2 from Boc-protected intermediated ...... 120

xix Chapter 1. Late-stage Functionalization of Bioactive Molecules

1.1. Introduction

The diversity of new chemical entities provided by natural products and synthetic compounds has propelled the field of new reaction discovery to facilitate bond formation in complex molecule synthesis. This diversity provides an opportunity to exploit these molecules in various disease states. One potential limitation, however, is the lack of ability to affect late-stage functionalization of these molecules to quickly optimize their properties, exploit potential interactions, or develop probe molecules without embarking on an entirely new synthetic sequence. In this thesis, methods for late-stage functionalization will be presented that allow for rapid analogue generation, and for the synthesis of tool molecules investigating key pathways implicated in the development of antimicrobial resistance. Chapter 2 discusses the development of a new hypervalent iodine- mediated strategy for the late-stage C-H functionalization of phyllanthusmin natural product derivatives. In chapter 3 traditional functional group interconversions are discussed to synthesize tool compounds through the post-synthetic functionalization of a known antibiotic to investigate an uptake pathway in Gram-negative bacteria. This thesis will conclude with a discussion on the power of late-stage functionalization for the construction of mechanistic probes employed in protein target identification studies. First, general C-H functionalization strategies will be discussed.

1.1.1 Late-stage functionalization

The last decade has seen an explosion in the number of new methods reported to afford direct C-H functionalization. In addition to new method development, several methods have been reported improving existing strategies. These attempts are aimed at increasing functional group tolerance and enhancing the substrate scope to allow for application toward complex molecule synthesis and derivatization of bioactive natural products bypassing de novo synthesis. This increase in C-H functionalization method development could be attributed to the change in perception of C-H bonds from latent backbone components to functional groups. Due to the ubiquitous nature of C-H bonds in a molecule, one can envision numerous opportunities to transform seemingly inert C-H bonds to functional moieties for various purposes within a synthetic sequence or drug development program.

The traditional synthetic approach to functionalize C-H bonds involves the conversion of one functional group to another in an iterative manner, or the combination of two functional groups to form a carbon-carbon bond (Figure 1A).1 On the other hand,

C-H functionalization directly converts a C-H bond into a functional group, which could in turn facilitate C-C bond formation (Figure 1.1B).1

Figure 1.1 Conceptual C-H bond functionalization

This direct way of transforming C-H bonds has revolutionized the way chemists approach challenges in natural product synthesis and analog generation in medicinal chemistry programs, however not without its limitations.

For the longest time a major limitation in the realm of C-H functionalization was that of selectivity in the C-H bond undergoing chemistry among the abundant C-H bonds in a complex molecule. This selectivity challenge highlighted immaturity in the methodologies resulting in limited application especially in natural product syntheses, and late-stage derivatization of lead drug compounds bearing dense functionality.

The suite of palladium-catalyzed cross-coupling methods is arguably one of the most useful reactions in organic chemistry.2 Despite having a long historical footprint,3 the progression of palladium-catalyzed cross-coupling C-C bond forming reactions was propelled out of necessity for the total synthesis of key anti-cancer natural product taxol

(1.3),4 enediyne antibiotic dynemicin A (1.6),5 and potent neurotoxin brevetoxin A.6

Moreover, the demonstration of these strategies on the large industrial scale toward the

4 syntheses of naproxen, singulair, and the herbicide prosulforon, led to a broader realization of the potential and utility of C-H functionalization as a synthetic tool.7

Scheme 1.1 Total syntheses enabled by palladium-catalyzed cross-coupling reactions

Likewise, it is possible that the adoption of C-H functionalization as a powerful tool in synthetic chemistry is attributed to the demonstration of more efficient total syntheses of key natural product targets. For example, the first synthesis of racemic austamide (1.7) was completed by the Kishi lab in 1979 over 29 steps.8 In 2002, Baran and

Corey disclosed the first enantioselective synthesis of austamide (1.7) in 5 steps (Figure

1.2). This significant achievement in efficiency was enabled by a palladium-mediated C-H alkylation of a N-prenylated tryptophan derivative to form a dihydroindoazocine tricycle as the key transformation.9 Toward the completion of dragmacidin D (1.8), Yamaguchi and 5 coworkers reported a 15-step synthesis of the complex marine natural product in 2011 facilitated by a series of direct C-H coupling reactions.10 The same natural product had previously been synthesized by Stolz et al. in 25 steps.11

Figure 1.2 Step-count reduction in total syntheses enabled by C-H functionalization

Finally, in another remarkable demonstration of the power of C-H functionalization to streamline synthesis of natural products, the 67 step synthetic sequence toward the completion of tetrodotoxin (1.9) by Isobe and coworkers in 200312 was cut down to 32 steps by the Du Bois group.12 The key transformations contributing to the massive step- count reduction was a Rh-catalyzed carbene C-H insertion reaction, and a late-stage stereospecific C-H amination mediated by a Rh-nitrene to afford the 6-membered quanidinium-containing ring. These examples showcase the potential to fundamentally improve upon classic strategies and streamline synthesis of complex molecules using C-H functionalization.

1.2 Developing a late-stage functionalization strategy

Toward the development of a late-stage functionalization strategy Cernak et al. recommend a four-step process when designing new methods or utilizing existing ones for a specific purpose within a research program.13 Following identification of the target bonds to functionalize based on prior SAR, stability studies, or computational chemistry, the potential reaction types corresponding to innate reactivity and selectivity patterns in the molecule should be mapped out. Additionally, computational models can be employed to assist in predicting the regio- and stereo-selectivity outcomes of potential transformations.13 Step 3 entails consideration of functional groups to be installed based on the objective of a program. For example, if the ultimate goal is to synthesize a common intermediate to facilitate subsequent generation of multiple analogs, then borylation or halogenation reactions are ideal. However, if the goal is to affect late-stage correction of

ADME properties in a drug development program, one single transformation e.g. trifluoromethylation, could be sufficient. Finally, it is essential to consider properties of the functionalized product particularly with respect to drug design principles, as well as objectively assess the ease and speed with which the product could be accessed by de novo synthesis using traditional robust approaches. Following these guiding principles could aid in providing a starting point for selection of reaction types or developing new chemistry to afford desired C-H functionalization.

1.2.1 Directed C-H functionalization

C-H functionalization reactions on complex molecules can be notorious for producing complex mixtures of products, resulting in challenging purifications and analysis. One of the more reliable ways to ensure site-selectivity is to guide the functionalization by way of a directing group. In some cases, the directing group can be installed onto the molecule and later removed, or optimally it can be a functional group native to the compound itself. Taking advantage of the latter directing approach, the Yu group has developed a suite of C-H functionalization reactions employing directing groups to afford site-selective transformations (Figure 1.3).14 One elegant and powerful example of this strategy and was the exploitation of the sulfonamide group on the blockbuster anti- inflammatory drug celecoxib (1.12a) as a directing group to facilitate ortho-selective aryl

C-H activation by a palladium catalyst system (Figure 1.3).14 Following ortho-C-H activation, six transformations (carboxylation (1.13), carbonylation (1.14), iodination

(1.15), arylation (1.16), methylation (1.17), and olefination (1.18), were independently developed to afford an array of celecoxib analogs from a single intermediate in a highly divergent manner. The key to this transformation was the conversion of the sulfonamide group to a more acidic perfluoroaryl sulfonamide facilitating C-H activation through a weakly coordinated palladium species resulting in C-H cleavage.14

Although this example beautifully highlights the diversity-generating power of C-

H functionalization by accessing one C-H bond selectively via a directing group, most substrates have multiple C-H bonds, and selectivity principles must be considered to ensure

8 the desired outcome is achieved. In the event that a directing group strategy cannot be employed, strategically utilizing innate reactivity of C-H bonds can lead to an array of differentially functionalized pharmacophores as seen with the derivatization of the cinchona alkaloids.

Figure 1.3 Directing-group mediated C-H functionalization of Celecoxib

1.2.2 Innate reactivity C-H functionalization

Careful consideration of the innate reactivity of unsubstituted C-H bonds in a molecule has also led to an elaborate family of cinchona alkaloid derivatives by several independent groups.15–19 The polarization of selected atoms for quinine (1.19), an indication of the innate reactivity of the corresponding aryl C-H bonds, is highlighted in

Figure 1.4. Electrophilic radicals are polarity matched with the - C5 and C7 positions.

Exploiting this reactivity, Baran and coworkers reported the addition of a trifluoromethyl radical to the dihydroquinine at the C7 position selectively (1.20).19 On the other hand, anionic methyl groups have been shown to add into the C2 position of the quinoline core to yield methylated derivate (1.21) in work reported in 2006 by the Gaunt group.16 The C2 position of the quinoline core is also acidic and be deprotonated by the hindered base,

Figure 1.4 C-H functionalization of quinine by innate reactivity

TMPMgCl●LiCl, when the alcohol is protected as the tert-butyl-dimethylsilyl ether, preventing coordination of the base to the amine. Knochel et al. demonstrated the trapping of the resulting nucleophile with electrophilic reagents such as allyl bromide to afford analogue 1.22 in moderate yield (41%).18 Furthermore, nucleophilic aryl or alkyl radicals have been shown to be selective for the C2 position.15 Alternatively, in the presence of the free alcohol the BF3●Et2O/TMPMgCl●LiCl system affords regioselective metalation at the

C3 position.18 In this care, the tertiary amine on quinine 1.19 can act as a directing group, facilitating the deprotonation event. This directed deprotonation thus allows for the trapping of the resulting species by electrophiles such as 2-dibromo-1,1,2,2- tetrachloroethane (Br+ source) to afford brominated alkaloid 1.23.

1.3 Synthesis of tool compounds by late-stage C-H Functionalization

The ability to rapidly access compounds for development of structure-activity relationships (SAR), target potency and selectivity optimization, and modulation of absorption–distribution–metabolism–excretion (ADME) properties is highly sought after in pharmaceutical research. Furthermore, this efficiency has the potential to enable optimization of physicochemical properties of drug candidates and provide access to new intellectual property opportunities by exploring innovative chemical space.

Scheme 1.2 Tritiation of bioactive compounds for metabolism studies

En route to candidate selection in a typical drug discovery program, several compounds are synthesized some of which serve as tool compounds to help in elucidation of biochemical mechanisms and investigating ADME properties. Late-stage functionalization can be used to convert lead compounds into tool compounds for example.

There is a long history of converting C-H bonds to C-T bonds to facilitate in vivo metabolism and distribution studies. Historically, Crabtree’s catalyst has been employed to afford this transformation,20 as well as palladium-mediated reduction of aryl halides in 12 the presence of T2 gas. Taking inspiration from Lockley’s work on the deuteration of aromatic compounds by an iridium catalyst,20 a collaborative effort by scientists from

Novartis and Bristol-Myers Squibb led to the tritiation of aristolochic acid II (1.24) with tritiated water (Scheme 1.2).21

Scheme 1.3 18F-labelling of bioactive compounds for PET studies

Similarly, Positron Emission Tomography (PET) is employed in drug development for in vivo distribution studies and visualizing target engagement. The most desirable radioisotope for this purpose is 18F due to its relatively longer half-life (110 min), aiding in its practical use. Although many methods have been disclosed addressing synthesis of 18F labeled compounds in the last decade, these methods are generally not compatible basic nitrogen-containing substrates. Hooker and Groves described a new heme-inspired method to directly convert benzylic C-H bonds to C-18F mediated by Mn(salen)OTs (Scheme

1.3).22 The utility of this strategy was demonstrated by the rapid labeling of a broad scope of bioactive compounds with 18F in which the Mn(salen)OTs concomitantly facilitates the activation of the C-H bond and 18F transfer to afford 18F-labeled analogs (1.27).22

Many of the most common metabolites observed for drug candidates in vivo are formed through oxidation of C-H bonds. Authentic samples of the metabolites are required

13 to validate pooled metabolites from in vivo experiments. Late-stage functionalization offers the advantage of functionalizing the candidate to provide samples required as standards for metabolism studies in an efficient manner from an advanced substrate as opposed to synthesizing the desired compound from the beginning.

Scheme 1.4 Aldehyde oxidase mechanistic probe developed by the Baran Lab

Me Me Me Me Me O Me O Zn(SO CF H) (DFMS) N 2 2 2 N NC N NC Me H N Me H Me Me N N CF3COOH, TBHP N DMSO, RT N 1.28 2014, Baran 1.29 H CF2H

To showcase the creativity with which one can employ late-stage C-H functionalization chemistry, the Baran group reported an ingenious application aimed to serve as a litmus test for aldehyde oxidase (AO) activity. The CF2H radical generated from

(((difluoromethyl)sulfinyl)oxy)zinc (DFMS) at ambient temperature added to electrophilic heteroarenes (1.28), albeit non-selectively (Scheme 1.4).23 This system is presumed to mimic the activity of the metabolic enzyme aldehyde oxidase, AO a major contributor to the pharmacokinetic profile of drugs in humans. The active site of AO has been shown to contain an oxidized high-valent molybdenum species responsible for attacking electrophilic carbons on heterocycles.24 This reactivity strongly resembles that observed in the electrophilic attack of heteroarenes by nucleophilic radicals. As a result, the addition of this CF2H radical into heteroarenes serves as a qualitative probe to identify metabolically susceptible sites on a drug molecule generated by AO.23 14

1.4 Late-stage C-H Functionalization for Bioconjugation

Natural products hold the potential to discover novel cellular targets in drug discovery. The new mechanisms of action provided by the natural product pharmacophores can be attributed to the unique core structures beautifully decorated with functional groups engaged in target binding. In order to develop probes that allow for elucidation of these mechanisms, synthesis of conjugates bearing reporter tags (biotin/fluorophore) attached at a position that does not affect bioactivity is a commonly employed strategy. However, functionalization natural products is typically not a small feat due to their structural complexity, limited quantities, and chemo- and regioselectivity challenges in a densely functionalized molecule. In most cases, developing chemical probes can be a shot in the dark. Ideally, functionalization of the natural product should occur at a site that does not adversely affect biological activity. However, sites that tolerate structural changes are not always obvious. Due to the uncertainty in determining appropriate sites of derivatization,

Kanoh developed a series of photo-generated carbene precursors that immobilize small molecules onto an affinity matrix in a chemo-selective manner (Figure 1.5).25 The advantage of this approach is that upon irradiation, the resulting carbene can add into a number of functional groups (alcohols 1.27, acids 1.28, thiols 1.29, amines 1.30, alkyl 1.31, aryl 1.33), on the bioactive molecule, increasing the likelihood that one of those analogs will retain activity and enable target identification.25

Figure 1.5 Development of photo-crosslinking functionalization tags

Although this non-site-selective approach to developing chemical probes can be applied to natural products, the method generally works best for derivatization of small molecules for which availability of the bioactive agent is not a limiting factor. In addition, these methods do not provide reliable SAR, as well as no way of pinpointing the site of attachment with certainty in the event that no targets are identified.

Strategic placement of linkers and tags is required in generating natural product- based probes using mild chemistry that allows for site-selective functionalization on a micro-scale. Traditionally, these probes have assisted the biological community in identifying putative targets and an increased understanding of the mechanisms of action by 16 which they exert their effects. One way to afford such conjugates, is to exploit unsubstituted C-H bonds on the molecule. Recently, Romo and co-workers reported several micro-scale strategies to arm natural products with functional groups enabling attachment of tags, as well as SAR studies.26 Among the disclosed methods are RhII -catalyzed OH and

NH insertions on alcohol- or amine-bearing natural products, InIII-mediated iodination on arenes, and cyclopropanation reactions on a broad alkene substrate scope.26 Since the methods are guided by existing functional groups, selectivity is usually predictable.

However, this can also be deemed a liability since the functional groups present on a natural product are normally involved in binding and may not be tolerant of changes.

Inspired by the C-H amination work of Du Bois and coworkers, as well as the mild methods developed for the functionalization of C-H bonds alpha to heteroatoms, alkenes and aryl groups, the Romo lab demonstrated an elegant application of this approach through installation of an alkyne moieties via a metallo-nitrenoid to natural products (1.35) to simultaneously arm the bioactive agent for probe development as well as enable SAR studies (Figure 1.6).27 Furthermore, this method provides the potential for rearrangements to occur leading to new structures (1.38).

Figure 1.6 Development of chemical probes from natural products

The utility of Romo’s method was demonstrated by facilitating the identification of the binding targets of eupalmerin acetate (EuPA, 1.40), a cytotoxic cembranolide diterpene natural product. Despite possessing potent anticancer activity against several cell lines including in vivo efficacy, the biological target of this compound was unknown. 18

Scheme 1.5 Site-selective allylic C-H amination of eupalmerin developed by the Romo lab

Arming this natural product with a linker-alkyne moiety (1.41) using Romo’s C-H amination strategy selectively afforded the allylic functionalized analog, EuPayne (1.42)

27 as a 1:1 mixture of diastereomers (Scheme 1.5). The biological activity of 1.42 (IC50 =

15 M) was only slightly attenuated compared to that of the natural product 1.40 (IC50 =

3.0 M) against the HL-60 human acute myeloid leukaemia cell line, and making it a good candidate for proteomic analysis. Following competitive activity-based protein profiling

(ABPP), a technique often employed to discover new enzymes, and multi-dimensional protein identification technology using the biotinylated adduct (MudPIT) led to the identification three targets. EuPa (1.40) was found to target, derlin-1 (DERL1), cytochrome b5 type B (CYB5B), and thromboxane A synthase (TBXAS1). DERL1 is known to be associated with cell proliferation. The other two targets CYB5B and TBXAS1 are proteins overexpressed in cancer.27

Scheme 1.6 Native chemical tagging of antibodies

One example of applying mild C-H functionalization chemistry toward the chemical tagging of small molecules for bioconjugation was reported by the Baran group in 2016. To facilitate late-stage functionalization for bioconjugation experiments the Baran group developed a series of sulfinate reagents known as diversinates that afford C-H functionalization of heteroarenes by exploiting innate reactivity.28 The Baran group demonstrated the power of this strategy by attaching azidoalkyl chains to a heteroarene that could then be conjugated to biomolecules such as a monoclonal antibody through a copper- free 1,3-dipolar cycloaddition with dibenzylazacyclooctyne (Scheme 1.6).29 This strategy has the potential to expand the scope of small molecules that can be conjugated to macromolecules. Beyond this report, the potential of this method lies in appending cytotoxic small molecules to antibodies to form antibody-drug conjugates conferring cell- type selectivity. Though this chemistry provides great promise for further bioconjugation studies, it is not without its limitations. In addition to low reaction yields and formation of regioisomeric products leading to challenging purification, acid labile groups and oxidation-prone functional groups are not well tolerated under these reaction conditions. 20 1.5 Conclusion

The rapid pace at which C-H functionalization chemistry is progressing is telling of the realization of its potential to enable drug discovery and development. This explosion seems to have been fueled out of necessity to streamline the synthesis of complex bioactive natural products and to generate a large number of compounds rapidly from common advanced intermediates for exploration of SAR vectors and enhancement of intellectual property positions. Once its utilization was demonstrated on complex molecules, the scope of application has been steadily increasing to facilitate biodistribution studies, metabolism investigations, and biochemical probe development. With respect to utilizing existing strategies, or developing new ones, careful consideration of the inherent reactivity of the

C-H bonds is imperative. Employing directing group strategies can also provide enhanced selectivity. Although methods for the direct C-H functionalization have been enabling to fields of synthetic and medicinal chemistry, traditional functional group interconversions are still widely valuable. The presence of heteroatoms (e.g. O, N, and S) in bioactive molecules, provides handles for various synthetic transformations, also facilitating the development of tool compounds and mechanistic probes in a post-synthetic fashion. The remaining chapters will highlight the value of both new direct C-H functionalization strategies, as well as traditional synthetic approaches for the purposes of rapid library expansion, and the synthesis of tool compounds for the elucidation of biochemical pathways of relevance to human health and disease.

Chapter 2. Iodane-mediated late-stage (hetero)aryl C-H functionalization

This project was a collaborative effort between the labs of Professor Fuchs in the Division of

Medicinal Chemistry and Professor Nagib in the Department of Chemistry and Biochemistry.

Contributions other than myself:

1. Stacy Fosu was responsible for experimental execution and characterization of some

substrates contained within the body of this chapter but absent from experimental

section.

2. Andrew Chen was responsible for DFT computational calculations.

2.1 Natural product-derived anticancer agents

Due to the worldwide prevalence cancer in modern society, as well as the morbidity and mortality that results, there is undoubtedly an unmet need for more effective treatment and prevention strategies. Nature has been one of the most valuable sources of new anticancer agents. The clinical success of paclitaxel,30,31 camptothecin,32,33 etoposide,34 and vinblastine,35 among many others,36 have demonstrated the potential of developing effective and innovative medicines inspired by natural products. Despite previous successes in this realm of chemotherapy, the multifactorial nature of cancer demands continued efforts in the discovery of new cytotoxic agents with unique modes of action.

Moreover, the destructive mechanisms by which most anti-cancer agents exert therapeutic effects require a need for optimization of agents to alleviate toxicity while still providing superior efficacy. Toward this effort, we have taken inspiration from nature to develop a new class of potent anti-cancer agents derived from the arylnaphthalene lignan class of natural products with a potentially novel mechanism of action.

2.2 Arylnaphthalene lignan lactones in nature

Lignans are a large class of naturally occurring phenols originally classified by

Harworth in 1948 as a dimer of monomeric phenylpropanoid units (Figure 2.1).37 The monomers (2.1) are connected through a -‘linkage of the propyl side chains and contain one or more asymmetric carbons (2.2). The resulting skeleton is oxygenated with hydroxyl, methoxyl, or methylenedioxy groups and further decorated with side chains carrying

23 various degrees of oxidation or cyclization to afford tetrahydrofurans (2.3) or tetrahydronaphthalenes (2.4). Historically lignans have been isolated from roots, heartwood, foliage, fruits, or resinous exudates of plants and display a wide range of biological activities.38

Figure 2.1 Lignan skeleton

Biosynthetically, ligans arise from cinnamic acid derivatives and are related to biochemical metabolism of phenylalanine through the Shikimate pathway.39 The product of the well characterized Shikimate pathway is coniferyl alcohol, a product of couramic acid reduction. Coniferyl alcohol is a valuable starting material in subsequent pathway ultimately leading to the generation secoisolariciresinol (2.5), a key advanced intermediate leading to the generation of both aryltetralin (2.6), and arylnaphthalene (2.7) classes of compounds (Figure 2.2). One prominent example of a natural product possessing an

24 aryltetralin skeleton (2.7) is the cytotoxic agent, podophyllotoxin (2.8). Likewise, several natural products have been isolated possessing the arylnaphthalene core, diphyllin (2.9).

Figure 2.2 Biosynthetic origins of diphyllin and podophyllotoxin

2.3 Phyllanthusmins as novel anticancer agents

The plant genus Phyllanthus constitutes one of the largest Euphorbiaceae plant families with over 600 identified species over a wide geographical distribution of South

America, Asia and Africa. In addition to the wide distribution, this plant genus has been shown to produce a variety of natural products such as flavonoids,40,41 alkaloids,42,43 terpenes44,45 and lignans,46–48 all showing diverse biological activities. An array of uses such as treatments for headaches, warts, diabetes and diarrhea has been found in Chinese medicine from many species of this plant genus.49 This diversity in biological activity and

25 ethnopharmacology has prompted intense research efforts toward the collection, isolation, and identification of the constituent agents from plants of this genus.

2.3.1 Isolation of phyllanthusmins A – E

Phyllanthus oligospermus is one of the most common plant species found in

Taiwan.50 Employing bioactivity guided fractionation, plant extracts of the stems and roots of the P. oligospermus shrub collected in Taiwan showed growth inhibition of the KB human epidermoid cell line and the P-388 mouse leukemia cell line.51 This finding resulted in further investigation into the active constituents of this chloroform extract leading to the identification of nine known compounds and three new arylnaphthalene lignans, phyllanthusmins A-C (2.10 – 2.12, Figure 2.3). The biological activity of the three new compounds was evaluated using a cytotoxicity assay. Phyllanthusmin A was found to exhibit noteworthy cytotoxicity against the KB and P-388 cell lines with IC50 values of

2.24 g/mL and 0.13 g/mL respectively.51

Figure 2.3 Structures of phyllanthusmins A-C

In 2014, the Kinghorn lab isolated a group of compounds from the phyllanthus plant genus collected in Vietnam. This class of compounds belongs to the arylnaphthalene lignan class of natural products and demonstrated potent cytotoxic activity against the HT-29 colon cancer cell line.

2.3.2 Phyllanthusmin structural considerations

The general structure of the phyllanthusmins B-E (2.11 – 2.14) is shown in Figure

2.4. The phyllanthusmins consist of an aglycone, diphyllin, glycosylated at the C7 phenolic position with an arabinose derivate with varying degrees of acetylation.52 A simple acylation of phyllanthusmin D resulted in the semi-synthetic derivative 2”-acetyl phyllanthusmin D (2.15). All five compounds were tested for anticancer activity in a cell- based cytotoxicity assay against the HT-29 colon cancer cell line. The general activity trend observed was that of increased potency with increased acetylation (Table 2.1). The semi- synthetic derivative 2”-acetyl phyllanthusmin D (2.15) was the most potent compound and

27 was initially as lead for further development into viable anti-cancer agents. Additionally, phyllanthusmin D (2.14) was shown to be active in a hollow fiber assay in which HT-29 cells were implanted into immunodeficient NCr nu/nu mice. These data combined suggested promising anticancer activity that required further explorations.52

Figure 2.4 General structure of the phyllanthusmins

1 2 3 Compound R R R IC50 (M: HT-29) Phyllanthusmin B 2.11 OH OH OAc 1.8 Phyllanthusmin C 2.12 OH OH OH 3.2 Phyllanthusmin D 2.13 OH OAc OAc -0.17 Phyllanthusmin E 2.14 OH OAc OH 1.8 2”-Acetyl phyllanthusmin D 2.15 OAc OAc OAc 0.11

Table 2.1 Biological activity evaluation of phyllanthusmins against HT-29 cell line

The potent anticancer activity observed in primary cell-based assays and hollow fiber assay prompted further inquiry into the affected biochemical pathway in cancer cells.

The structural similarities between the phyllanthusmins and etoposide a cancer chemotherapeutic agent of broad clinical applications prompted further investigation into

28 the mechanism of action (Figure 2.5). The podophyllotoxin etoposide belongs to the aryltetralin (2.7) class of lignans. Furthermore, biosynthetically both the phyllanthusmins and the podophyllotoxins are derived from a common intermediate secoisolariciresinol

(2.5, see Section 2.2). Etoposide (2.16) is a potent topoisomerase II inhibitor, preventing replication of cancer cells by interfering with DNA replication. The most potent compound from the Kinghorn isolation, phyllanthusmin D (2.14), was evaluated for topoisomerase activity and found to exhibit no activity against topoisomerase up to 100 M (Figure 2.5).52

Figure 2.5 Structural similarity between PHY-4 and etoposide

One of the key factors in the process of programmed cell death is a protease known as caspase-3, an anticancer drug target of high interest due to its pivotal role in apoptosis and inflammation.53–55 Phyllanthusmin D and etoposide were evaluated for activation of caspase-3 activity in the HT-29 cell line. At a dose as low as 1 M, phyllanthusmin D was shown to activate caspase-3 in the HT-29 cell line, whereas etoposide did not activate

29 caspase-3 at doses up to 10 M.52 Additionally, phyllanthusmin C was shown to activate

NK cell IFN- production. IFN- plays the pivotal role in the activation of innate and adaptive immunity against tumor cells and viral infections.56 The interesting activity profile of the phyllanthusmins warranted further analogue generation.

2.4 Strategies toward the access of arylnaphthalene cores

In recent years, several research groups have developed methods for the construction of polysubstituted naphthalene cores (2.19 – 2.23) due to their value in bioactive compounds,57–59 materials, and synthetic applications.60,61 Previous methods describing new methods to construct this valuable motif include transition metal catalyzed cycloadditions (Scheme 2.1A)62 and Lewis-acid mediated annulations of prefunctionalized substrates.63

Scheme 2.1 Methods to construct arylnaphthalene cores I

Additionally, there is an increasing number of reports of radical methods aimed at the synthesis of aryl naphthols such as strategies developed by the Wirth and coworkers utilizing fully loaded substrates (2.20) to afford the desired substitution (Scheme 2.1B).64

One major disadvantage of the methods described above is the use of prefunctionalized substrates, ultimately leading to limited scope and utility.

2.4.1 Narender’s silver (I)-catalyzed radical cyclization

To address this problem, Narender and coworkers developed an atom economical method for the regioselective construction of substituted -naphthols (2.26) from accessible substrates (Scheme 2.2). They discovered that under mild oxidation conditions, employing sodium persulfate, -ketoesters (2.24) undergo radical cyclization with a alkynes (2.25) in a single step using a silver (I) salt to generate naphthalenes (2.26).65

Scheme 2.2 Silver (I)-catalyzed regioselective synthesis of lignan cores

The proposed radical-mediated mechanism for this transformation is highlighted in

Scheme 2.2. The disproportionation of the persulfate anion into the sulfate dianion and the sulfate radical is facilitated by a silver (I) salt to afford the active Ag (II) species, which can then oxidize the -ketoester substrate (2.24). The resulting stabilized radical (2.27) adds into the alkyne (2.25) to afford a vinyl radical (2.28) which then undergoes intramolecular cyclization to afford the delocalized high energy system (2.29) that is prone to immediate aromatization following single-electron transfer and deprotonation. The regioselectivity is presumably determined by the stability of the resulting vinylic radical

(2.28). The -naphthols resulting from this protocol can then be converted in one step to biologically lignans.

2.4.2 Vidal’s Brønsted acid-catalyzed benzannulation

Another commonly employed strategy to generate polysubtituted naphthols involves the engagement of catalytic approaches to enable a formal [4+2] cyclization to construct the second aromatic ring of the core (Scheme 2.3). Typical examples of such methods are Larock’s palladium (0)-mediated cyclizations (I),66 -Lewis acid-catalyzed benzannulations of alkynyl benzaldehydes (2.36, II),67 or condensation reactions of phenylacetaldehydes (2.34) with alkynes (2.31) under Ga (III)68 or Au (III)69 catalytic systems (III). In 2015, Ponra and coworkers disclosed the first Brønsted acid-catalyzed benzannulations (IV). They discovered the efficient benzannulation of phenylacetaldehyde

(2.38) with alkynes (2.31) employing triflimide as an organocatalyst.70 Although several methods have been developed to construct these arylnaphthalene cores, one major disadvantage is the required prefunctionalized non-commercially available synthons. This limitation results in a decrease in modularity allowing for introduction of variability toward analog generation.

Scheme 2.3 Methods to construct arylnaphthalene cores II

Although multiple methods are available for the preparation of diphyllin, the method of Charlton and coworkers provides the most efficient access to the aglycone

(Scheme 2.4).71 This strategy utilizes a five step sequence to access the naphthalene ring system through a key Diels-Alder cycloaddition reaction. A final, directed reduction leads to the synthesis of diphyllin (2.45). Overall, the process requires only a single chromatographic step and generates the desired product in an approximately 40% yield from the starting veratraldehyde (2.40) on gram-scale. The synthetic sequence is then completed by a phase transfer glycosylation of diphyllin (2.45) by a triacetylated-L- arabinose glycosyl donor (2.46) to afford PHY-4 (2.15, Scheme 2.4).

Scheme 2.4 Charlton’s diphyllin synthesis

2.2.4 Structure-activity relationships of phyllanthusmin derivatives

The synthesis of analogues of the phyllanthusmins can essentially be reduced to the glycosylation of diphyllin. The key is the ability to synthesize diphyllin in appropriate quantities. Utilizing Charlton’s synthesis of diphyllin, several glycone analogs were generated and evaluated for antiproliferative activity against the colon carcinoma HT-29 cell line (Figure 2.6). Analysis of the data revealed structure-activity relationships that were consistent with functionalized hydroxyl groups on the sugar resulting in more potent analogs

Figure 2.6 Antiproliferative activity in HT-29 cell line (IC50 in M).

ranging in activity from 0.05 to 2 M. The most potent analog was a methylated-xylose derivative (2.50) with an IC50 of 0.0018 M. Interestingly, the analog possessing a triacetylated D-arabinose sugar was slightly more potent than its L-arabinose counterpart found in the natural products (data not shown).72 In addition, the larger disaccharide, lactose (2.55), showed the most significant loss in potency highlighting a possible size constraint with respect to the sugar binding pocket in the target protein(s). A handful of C- ring analogs have been synthesized in our lab. These analogs demonstrate the C-ring is required for activity.72 In total more than 50 analogs have been synthesized, mostly targeting derivatization of the glycone due to the late-stage installation in the synthetic sequence. A rather wide distribution of potencies has been observed in the tested cell lines

36 ranging from 0.110 M to 0.0002 M. Figure 2.7 shows a summary of the elements that have been explored and the corresponding biological activity ranges.

Figure 2.7 Structure-activity relationship summary of phyllanthusmin analogues

The design strategy to afford phyllanthusmins analogues relied on introduction of prefunctionalized substrates into a linear synthetic sequence in a highly modular fashion.

However, the limitation to this approach was the need to repeat the entire sequence for the generation of A- or D-ring analogs. A late-stage functionalization of A- or D-rings would potentially increase the efficiency of this approach. In an attempt to probe the structure- activity relationships with respect to the unsubstituted aryl positions, we sought to develop a method for the late-stage functionalization of the diphyllin core. These changes were anticipated to modulate the potency, solubility, and metabolism of the analogs. This strategy was envisioned to allow for the rapid generation of derivatives from a common advanced intermediate.

2.5 Iodane activation enables direct C-H functionalization of arenes

2.5.1 Activation of hypervalent iodine reagents

Chapter 1 of this thesis briefly discussed the great leaps in the development of C-H functionalization methods. While these advancements have revolutionized the way medicinal chemists approach synthetic challenges in the multi-step synthesis of bioactive molecules, the methodologies required for transformation of more complex substrates such as densely functionalized natural products and pharmaceutical agents are somewhat limited. The harsh reaction conditions typically employed (strong oxidants and heat) result in limited applicability toward the functionalization of advanced intermediates. Inspired by the versatile reactivity of hypervalent iodine due to the labile nature of the ligands on the iodine (III) center we wondered if we could exploit this characteristic for the development of a mild aryl C-H functionalization ‘toolbox’ toward the late-stage functionalization of the diphyllin core or optimally for the derivatization of the phyllanthusmins themselves. In order to develop the necessary methodology, a more thorough investigation of the iodane system and substrate scope would be required.

Encouraged by the utility of hypervalent iodine as a mild oxidant, we wondered if we could further exploit the labile nature of the ligands on the electrophilic iodine center

3 to generate an asymmetrical  -iodane 2.57 from PhI(OAc)2 (PIDA, 2.56, Figure 2.8). We hypothesized that in the presence of dilute Brønsted acid the acetate ligands on iodobenzene diacetate (2.56) would be sequentially replaced by the corresponding anion

- (X ) ultimately forming bis-substituted iodane, PhIX2 (2.58). En route to formation of

PhIX2 (2.58), we were curious if the high reactivity of the short-lived hybrid iodane (2.57)

38 could be harnessed in the presence of an arene, to afford selective functionalization in good yield and innocuous iodobenzene which could be recycled and re-oxidized.

Figure 2.8 Generation of non-symmetrical iodane hypothesis

Iodosobenzene chlorotrifluoroacetate [PhICl(OCOCF3)], which presumably arises from ligand exchange of iodobenzene dichloride in trifluoroacetic acid, was reported by

Keefer and coworkers in 1962 to be a more reactive chlorinating agent than iodobenzene

73 dichloride, PhICl2 (2.61, see Section 2.7). We wondered if an analogous chloroacetate species would facilitate efficient chlorination reactions bypassing the need to pre-form the active oxidant, and perhaps giving rise to new reactivity. We envisioned stirring (PIDA,

2.56) with a mineral acid having a pKa lower than that of acetic acid would favor sequential ligand exchange on the iodine (III) center (Figure 2.8). We postulated that the elongated I-

X bond (where X = anion) would be highly reactive due to a low-lying LUMO. Evidence of this phenomenon has been disclosed by the Shafir group (Scheme 2.5), and they suggest the activation of hypervalent iodine reagents such as PIDA (2.56) occurs by way of desymmetrization of the ligands on the iodine (III) center to afford 2.59.74 39

Scheme 2.5 Activation of iodobenzene diacetate via desymmetrization

This desymmetrization is thought to be the reason activated hypervalent iodine reagents from either PIDA (2.56) or PhI(OCOCF3)2 (PIFA) are more reactive, possibly due to the generation of the strongly oxidizing PhIOAc+. Indeed to our delight, NMR analysis of a mixture of PIDA (2.56) and dilute HCl in methylene chloride showed an intermediary species, which was presumed to be iodosobenzene chloroacetate (2.60) by tracking the ortho aryl C-H protons (Figure 2.9).

1 Figure 2.9 H NMR (400 MHz, CDCl3) analysis of ligand exchange after 15 minutes

Having demonstrated successful ligand exchange with HCl we explored the feasibility of enabling a similar transformation by way of other simple dilute mineral acids (Table 2.2).

To our delight, using 2-methyl pivanilide (2.62) as a model substrate we demonstrated the expansion of this methodology to afford both oxygenations (2.63 – 2.65), and halogenations (2.66 and 2.67) highlighting the capabilities to develop a C-H functionalization tool-box using PIDA (2.56) and simple acids. Although all these transformations are potentially useful, the prevalence of halogenated arenes in pharmaceutically relevant compounds, and their potentially for subsequent rapid functionalization focused our attention on the latter transformation for further optimization.

Compound HX functionalization %Yield 2.63 HOTf oxytriflation 53% 2.64 HOTs oxytosylation 65% 2.65 HOMs oxymesylation 66% 2.66 HBr bromination 92% 2.67 HCl chlorination 88%

Table 2.2 Pivanilide C-H functionalization using Bronsted acids

2.5.2 Development of aryl-C-H chlorination conditions

For decades, the synthetic community has realized the versatile value of halogenated arenes and heteroarenes through a collection of well-established coupling reactions as well as classical functional group interconversions. In addition to behaving as robust functional group handles, halogenated arenes can be found in core structures of natural products, as well as many drug molecules. In recent years, a study conducted by the Njardarson lab analyzing the elemental composition of U.S. FDA approved drugs highlighted the prevalence of chlorine as the most incorporated halogen in pharmaceutical agents (Cl > F >> Br > I).75 In fact, this analysis placed chlorine as the 6th most prevalent element following C, H, N, O and S.75 The chlorine atoms in these molecules serve the purpose of modulating protein interactions, tuning lipophilicity, and participating in halogen bonding during target engagement. Furthermore, chlorinated arenes and

42 heteroarenes are appearing more frequently in material science76,77 and as imaging agents,78 signifying their diverse utility.

The exciting confirmation of the in-situ generation of PhICl(OAc) (2.60, Figure

2.9) led to the exploration of the substrate scope of chlorinated arenes. We were pleased to observe regioselective monochlorination of arenes in good preparative yields (Table 2.3).

In addition to bypassing the synthesis of the chlorinating reagent, the use of ubiquitous dilute HCl, typically employed as a work-up solution, and the bench stable PIDA (2.56) provides easy and efficient conditions for aromatic chlorination tolerant of benzylic positions (2.72, 2.73, 2.76), -keto positions (2.71, 2.73), and also allowed for the generation of bis-halogenated arenes (2.68, 2.78, 2.79). In addition, most substrates highlighted a strong para-selectivity. Halogenated arenes are prominent moieties in drug discovery and development. In addition to providing useful synthetic handles in multi-step syntheses, halogens can be intentionally introduced onto aromatic rings to improve metabolic stability by attenuating electron density via para-substitution.

Table 2.3 Chlorination of arenes-substrate scope

Having an awareness of the prevalence of heteroarenes in small molecule drug discovery, we explored chlorination of a few structurally unique heteroarenes. Using isoquinoline (2.80) as an initial model substrate, protonation (2.81) of the nitrogen under the previously established PIDA/HCl conditions led to deactivation of the ring preventing functionalization (Figure 2.10).

Figure 2.10 Chloride source investigation for heteroaryl chlorination

Therefore, to facilitate chlorination of this ring system we investigated several Lewis acid chloride sources, as well as acyl chlorides (Figure 2.10). Interestingly, acyl chlorides

(AcCl, and EtO2CCl) gave excellent yields of 4-chloroisoquinoline (2.82), with ethyl chloroformate providing the best yield, 92%. The surprising selectivity of this reaction for the C4 position on isoquinoline (2.80) is unique in that there are not many methods to generate this halogenated species.79 In fact, C-H functionalization methods, devoid of a directing group strategy, selective for the C4-isoquinoline or C3-quinoline position are quite rare,80 which could explain the high cost associated with commercially acquiring 4- chloroisoquinoline (~$500/g Matrix Scientific). This unexpected discovery of efficient

PIDA/acyl chloride conditions prompted us to investigate the composition of this reaction mixture by NMR. Spectral analysis showed a slower formation of PhICl2 (2.61), compared to the HCl conditions (see Chapter 5: Experimental Section). Optimization of this reaction led to a set of conditions that could be efficiently applied to the chlorination of a diverse array of heteroarenes (Table 2.4). In addition to isoquinoline (2.82, 2.86-2.88) and quinoline (2.89), quinoxaline (2.90), indoles (2.83 and 2.84), pyrazoles (2.85), and 45 pyridines (2.91-2.93), and pyrimidine (2.94) were all tolerant of the reaction conditions, demonstrating that both electron rich and electron deficient systems could be functionalized in this way.

Table 2.4 Chlorination of heteroarenes .

Lastly, to further demonstrate the functional group tolerance, as well as highlight the utility of this method for late-stage halogenation of medicinally relevant compounds, we were pleased to chlorinate lidocaine (2.95), caffeine (2.96), naproxen ester (2.97 and

2.98), dimethyl uracil (2.99 and 2.100), and papaverine (2.101 and 2.102) with good to

46 excellent yield and complete regioselectivity (Figure 2.11). Mechanistic studies on the acyl chloride-facilitated reaction should aid in understanding this selectivity.

Figure 2.11 Halogenation of pharmaceuticals and natural products

To determine if any selectivity for different aromatic rings existed, a competition experiment (Figure 2.12) between 2-iodoanisole (2.118) and isoquinoline (2.80) was conducted employing either the standard optimized HCl conditions or ethyl chloroformate as the halogen source. To our surprise, analysis of these reaction mixtures showed no cross reactivity. Under the HCl conditions, only chlorinated 2-iodoanisole (2.103) was produced.

In fact, an almost quantitative retention of isoquinoline starting material was also recovered from the reaction. In the presence of ethyl chloroformate, however, the isoquinoline was selectively chlorinated to produce (2.82) and the iodoanisole (2.118) remained untouched.

Mechanistic studies on the acyl chloride facilitated reaction should aid in understanding this selectivity.

Figure 2.12 Chlorination of arenes and heteroarenes competition experiment

One can certainly envision the utility of such selectively, however, when functionalizing heteroarenes in the presence of arenes and vice versa as demonstrated in the functionalization of pharmaceutical agents and natural products like those in Figure 14.

Having established the reactivity and selectivity parameters of the methodology, functionalization of PHY-4 (2.15), a semi-synthetic analogue of phyllanthusmin D, was attempted (Scheme 2.6). This structurally complex, advanced intermediate was subjected to our aqueous HCl chlorination conditions, but unfortunately these conditions resulted in deglycosylation to afford the inactive aglycone diphyllin (2.45, Scheme 2.6).

Scheme 2.6 Attempted iodane-mediated chlorination of PHY-4

Preservation of the carbohydrate moiety was vital to maintaining biological activity. In order to investigate other potential suitable partners for this reaction, a model substrate, acetyl-protected diphyllin (2.104), was employed. Results of this investigation are shown in Table 2.5. While Me3SiCl only afforded a trace amount of chlorinated acetyl- diphyllin 2.105, LiCl provided complete conservation of starting material. Interestingly, a freshly prepared sample of PhICl2 resulted in complete decomposition of the substrate. This degradation was ascribed to the residual amount of acid inherently present in the reagent as a photo- or heat decomposition product. Although ICl gave approximately 6% yield of the target molecule, our best results were achieved upon the utilization of Bu4NCl as the chloride source and 1.5 equivalents of PhI(OAc)2, to afford 40% of target molecule 2.105.

Doubling the amount of oxidant led to a boost in isolated yield to 63%.

Scheme 2.7 Iodane-mediated chlorination of acetyl-diphyllin

63%

42%

3% 8%

*2x equiv. of [O] Table 2.5 Iodane-mediated chlorination of PHY-4

The use of a non-acidic chloride source Bu4NCl, gave us confidence to attempt chlorination of PHY-4 (2.15). To our delight, Bu4NCl cleanly furnished the chlorinated analog 6-chloro-

PHY-4 (2.106) in 54% yield without loss of the carbohydrate moiety (Scheme 2.8).

Scheme 2.8 Iodane-mediated chlorination of PHY-4

To expand the library of analogues, bromination was attempted on PHY-4 (2.15).

Remarkably, following investigation of several bromide reagents, quantitative conversion to bromo-PHY-4 (2.107) was achieved by LiBr, where other bromide salts afforded no product (Table 2.6). Structural assignment of the position of functionalization was enabled by a suite of 2D-NMR techniques. Interestingly, the target molecule was found to be 6’- bromo-PHY-4 (2.107) exclusively. We hypothesize that the switch in regioselectivity was attributed to the adverse steric interaction between the larger bromide and the arabinose sugar. Furthermore, investigations into the optimum salts to afford the required transformations led to a striking counter ion effect (see Table 2.5 and Table 2.6). We hypothesize that the more dissociated the ion pair of the respective salt, the greater the ease of formation of the hybrid-iodane 2.57.

Table 2.6 Investigation of PHY-4 bromination conditions

2.6 Structural considerations of halogenated Phy-4 analogues

The initial report of the isolation of the phyllanthusmins by the Kinghorn lab indicated the existence of these compounds as relatively kinetically stable rotamers due to hindered rotation about the C1’ - C7’ bond.52 The Charlton group also made the same observation independently.71 As a result, the 13C NMR spectra of these halogenated derivatives had pronounced doubling and splitting of specific peaks. The bromination reaction led to the formation of atropisomers and this mixture was submitted for biological evaluation (see Chapter 5.1). The cytotoxicity was evaluated against an HT-29 colon cancer cell line, three ovarian cancer cell lines OVCAR4, OVCAR8, and KURA (Figure

2.13). In all cases 6’-bromo-PHY-4 (2.107) showed a significant decrease in activity compared to the parent compound PHY-4 (2.15). Despite the attenuated activity, marked selectivity among the cell lines was observed, particularly in the ovarian cancer cell lines. 52

This striking differential activity profile in the ovarian cancer cell lines by simply adding a bromine to the D-ring warrants further investigation. Due to the fact that this brominated derivative had decreased cytotoxic activity (Figure 2.13) compared to the parent compound, we suspect the hindered rotation locks the benzodioxole ring in a conformation that is not conducive to target protein binding. Alternatively, this data could suggest substitution of the C6’’ position is generally not well tolerated. This seemingly simple transformation and the subsequent data generated highlights the potential of this approach to facilitate efficient analogue library expansion. In addition, installation of such functional handles allows for further library expansion by exploiting the versatile reactivity of halogens in synthetic transformations.

Figure 2.13 Biological activity evaluation of brominated PHY-4

2.7 Mechanistic investigations

With respect to the mechanism by which the halogenation reaction occurs, we propose two possible mechanisms. Our general hypothesis is that either (A) single electron oxidation of the arene generates of a radical cation that is attacked by the respective anion or (B) the generation of a cationic iodonium species affording electrophilic Cl+ species prone to attack by an arene nucleophile. The hybrid iodane 2.60 is prone to ligand dissociation to generate the cationic iodonium species with the chloride counterion 2.109, which is presumably in equilibrium in the iodonium ion with an acetate counterion 2.110.

Scheme 2.9 Halogenation mechanistic proposals

As briefly mentioned in section Section 2.5.1, more than 50 years ago Keefer and coworkers reported the inhibitory effect of HCl on the iodobenzene dichloride-mediated 54

chlorination of durene in trifluoracetic acid as a catalyst (Scheme 2.10).73 They hypothesized that the catalytic activity was due to the ability to form one highly electrophilic halogen (Cla) atom of iodobenzene dichloride, in a manner that strongly resembles Shafir’s desymmetrization notion (cf Section 2.5.1). Hydrogen bonding between trifluoroacetic acid to the other halogen (Clb) promotes the increased electrophilicity on Cla.

Scheme 2.10 Desymmetrization of iodobenzene dichloride by trifluoroacetic acid

One can then propose that under our HCl conditions, PhICl2 (2.61) is formed, however in the presence of acetic acid, you can similarly achieve enhanced electrophilicity by way of desymmetrization through hydrogen bonding as shown in Scheme 2.10. NMR analysis of reaction mixtures of PhI(OAc)2 (2.56) and HCl does indeed show formation of

PhICl2 (2.61) and the hybrid iodane 2.60 (Figure 2.9). We cannot conclude with absolute certainly whether the formation of the hybrid iodane occurs through desymmetrization of

PhICl2 (2.61) as in Scheme 2.5, or a true sequential ligand substitution of PhI(OAc)2 (2.56).

However we can make reasonable data-driven conclusions. Although a mixture of

PhI(OAc)2 and AcCl spectroscopically shows formation of the hybrid iodane 2.60, other

acyl chlorides that afford the same product (e.g. 4-Cl-isoquinoline, 2.82) such as mixtures of ethyl chloroformate and PhI(OAc)2 do not show presence of PhICl2 by NMR analysis

(see Section 5.2). This data supports the sequential ligand substitution hypothesis to form the highly electrophilic PhICl(OAc) 2.60. With this in mind, we compared our chlorination conditions to classical electrophilic chlorinating reagents, and new state-of-the-art reagents such as Palau’chlor and IBA-Cl (Table 2.7). To our delight, target molecules were formed only under our conditions further demonstrating the unique and heightened reactivity of the non-symmetric iodane (Table 2.7).

Reagents %Yield of 2.82 %Yield of 2.89 tBuOCl 0% 0% NCS 0% 0% Palau’chlor 0% 0% IBA-Cl 0% 0% RCl, PhI(OAc)2 92% 63% ( R = EtO2CCl) (R = C6F5COCl)

Table 2.7 Comparison to known aryl C-H chlorinating reagents

To investigate plausibility of a single-electron oxidation of the heteroarenes to generate a radical cation, oxidation potentials of isoquinoline and quinoline were measured by cyclic voltammetry in acetonitrile (see experimental section). We did not observe redox

potentials of the hypervalent iodine reagents relevant to our chemistry that supports oxidation of (iso)quinoline within the wide solvent window of MeCN; therefore, we expect this pathway is unlikely.

To investigate the plausibility of heteroarene functionalization via an iodonium intermediate the pyridinyl diaryliodonium salt (2.114) was synthesized and subjected to acetyl chloride-mediated conditions with and without iodobenzene diacetate (2.56). Under both conditions, neither 3-chloropryridine (2.115) nor 4-chloroanisole (2.116) was observed (Scheme 2.11). This finding suggests that the pyridinyl diaryliodonium salt

(2.114) is most likely not a relevant intermediate in the reaction mechanism.

Scheme 2.11 Diaryliodonium as possible reaction intermediate

To probe the plausibility of an N-donor hypervalent iodine species as being an active intermediate in the reaction pathway the isoquinoline variant (2.117) was synthesized according to a procedure reported by Wengryniuk (Table 2.8).81 The dicationic isoquinolinium substrate (2.117) was subjected to chlorinating conditions highlighted in

Table 2.8. In the case of acyl chloride sources AcCl, and EtO2CCl, no target molecule

(2.82) was attained. In the presence of PhICl2 and the AcCl/PhIOAc2 combination, only a

trace amount of target materials was observed. This data indicates that this N-donor complex is not likely an intermediate in the reaction pathway.

% Recovered Reagent %Yield of 2.82 Comment Iodonium AcCl 0% 0% 4% isoquinoline recovered EtO2Cl 0% 0% decomposition PhICl2 3% 0% 78% isoquinoline recovered AcCl, PhI(OAc)2 1% 0% 100% isoquinoline recovered

Table 2.8 Investigation of N-donor iodonium as possible reaction intermediate

In an attempt to understand the divergent regioselectivity observed in the halogenation of PHY-4 (2.15, Scheme 2.12), as well as the unique regioselectivity observed for the (iso)quinolines density functional theory (DFT) calculations were conducted according to Ritter’s procedure in his elegant work on predicting selectivity in charge transfer-directed C-H functionalizations.82

Scheme 2.12 Divergent regioselectivity of PHY-4 halogenation

The DFT calculations were performed using the Gaussian 16 (revision A.03) suite of programs using Becke’s three-parameter hybrid exchange functional and the Lee-Yang-

Parr correlation functional (B3LYP) DFT method. Population analysis on both the N and

N-1 systems were carried out with Weinhold’s Natural Bond Order (NBO) program

(version 3.1), also included in Gaussian 16 suite of programs. Population analysis on a representatives set of arenes and heteroarenes was conducted following optimization of geometries to provide values for the N electron system. This data provides a measure of charge density at each atom. From population values, Fukui index values were determined by substracting the subtracting the charge densities of the N-1 system (one electron oxidation) from the N system. The Fukui index for electrophilic attack provides a measure of electron density loss at each atom following a global one electron global oxidation. The 59

position with the greatest loss in electron density is presumably the site that is most prone to electrophilic attack. Following careful data analysis of the DFT calculations we discovered that population analysis and Fukui index predicted differential regioselectivity depending on the arene type (Figure 2.14). Fukui index values matched experimental observations with respect to the electron-rich arenes. For example in the case of PHY-4

(2.15), the Fukui index predicts the two most susceptible positions to electrophilic attack as the C6 and the C6’ positions (2.106 and 2.107). Experimental results support this prediction for Cl-pivanilide (2.67), and Cl- papaverine (2.101) (Scheme 2.12 and Figure

2.14).

With respect to the heteroarenes, population analysis highlights the positions of greatest electron density as C4 on isoquinoline (2.82), C3 on N-pivaloylindole (2.84), and

C3 on quinoline (2.89) (Figure 2.14). Experimentally, chlorination of (iso)quinoline using our conditions reflects the prediction with the generation of 3-choroquinoline and 4- chloroisoquinoline.

Figure 2.14 DFT computational analysis of regioselectivity

61 2.8 Conclusion

In summary, we have developed a practical method to functionalize arenes using a readily available hypervalent iodine source, iodosobenzene diacetate, and ubiquitous dilute

1M HCl under mild conditions, with excellent selectivity. This method has been further expanded to afford the chlorination of heteroarenes in good to excellent yields, providing one of the few existing examples of chlorination of unsubstituted isoquinoline at the C4 position and quinoline at the C3 position using mild conditions. Mechanistic evaluation has led to the exclusion of an iodonium-mediated transformation and highlights either single electron oxidation of the arene to generate of a radical cation that is attacked by the respective anion or the generation of the cationic iodonium species affording electrophilic

Cl+ prone to attack by an arene nucleophile.

This methodology was ultimately developed as a general tool for the late-stage functionalization of natural products and potential drug candidates. Application of this methodology successfully resulted in the chlorination of a series of pharmaceutically relevant compounds and natural products. In fact, the chlorination and bromination of

PHY-4 provided an excellent proof-of-concept that this methodology could be tolerated by a highly functionalized natural product derivative, demonstrating the potential for this methodology to be employed for a wide range of substrates. Continued application in other systems would ultimately be expected to lead to active analogues based on the structural requirements of the biological target.

Chapter 3. Hijacking the Fatty Acid Transport System to Facilitate Drug

Penetration Across Bacterial Membranes

The data in this chapter was generated in the Antibacterial Discovery Performance Unit at GlaxoSmithKline (GSK), Collegeville PA under the leadership of Robert Stavenger

(medicinal chemistry) and David Holmes (microbiology). Portions of this work was supported by the TRANSLOCATION consortium (www.translocation.eu) under the

Innovative Medicines Initiative Joint Undertaking grant agreement no. 115525, resources of which are composed of financial contribution from the European Union’s seventh framework program (FP/2007-2013) and EFPIA companies in kind contributions.

Contributions other than myself:

1. Xiangmin Liao (GSK) was responsible for synthesis of initial series of alkyl-linked

analogues (3.26, 3.27, 3.29, 3.31, 3.32) and alkyne functionalized ciprofloxacin

(3.33).

2. Supporting imaging data was generated by Maggie Truong (GSK) and

collaborators in the Winterhalter Lab at Jacobs University, Bremen.

3. Biological data was generated by microbiologists Pan Chan and Nabil Abraham

(both from GSK).

3.1 Penetration across bacterial membranes

3.1.1 Introduction

The discovery and development of new antibiotics has drastically slowed down in recent decades due to a complex network of factors involving regulatory hurdles, low return on investment, and significant scientific challenges. As a result, many pharmaceutical companies have abandoned antibiotic drug discovery efforts. In the last 30 years, almost all ‘new’ antibiotics have been derivatives of existing drugs.83a In fact, each antibiotic available to date is a part of a class that was discovered between 1900 and 1984, since then there have been no new classes brought to market.84a Concurrently there is increasing concern as we enter an era of growing antimicrobial resistance (AMR) with what appears to be no simple solution in sight. Although the first antimicrobial agents were synthetic agents (sulfa drugs), the serendipitous discovery of natural product antimicrobial agents, hallmarked by penicillin (1929)85 for Gram-positive infections caused by

Staphylococcus and Streptococcus, and streptomycin (1943)86 for the treatment of

Mycobacterium tuberculosis introduced what is known as the golden era of antibiotic discovery (1950 – 1960).83a

This era was characterized by the mining of metabolites of bacteria and fungi, which was sometimes followed by post-isolation chemical modification. Some examples of successful antibiotics discovered using this approach are the aminoglycosides, cephalosporins and tetracyclines.83b Following these key discoveries, changes in the scaffolds led to the discovery of second, third, etc. generation compounds that were 65 effective in treating bacterial infections due to their expanded spectrum, and safety profile.87 These new agents revolutionized the treatment of infectious diseases, and helped to enable invasive surgeries such as organ transplantation, and cancer chemotherapy.87

Many breakthroughs in transformative medicine and recent improvements in quality of life that have been enabled by antibiotics are now at risk because bacteria are developing resistance, and no new mechanism antibiotics are replacing the current antibiotic arsenal.

The effectiveness provided by antibiotics led to misuse in the control and treatment of human and animal disease. Overuse and inappropriate use of the drugs resulted in the development of resistance among bacterial populations. Due to the rapid rate of replication of bacteria, this decades-long mobilization of resistance has led to the emergence of multidrug resistant (MDR) bacteria undermining our ability to control bacterial infections.87

Despite the many successes during the golden era, there was a considerable gap in the discovery of antibiotics with a new mechanism of action from 1960s to early 1990s. The scientific community attempted to respond by employing innovative discovery approaches.

Technological advancements, including high throughput screening techniques enabled by robotic liquid handling, the creation large libraries of compounds using combinatorial chemistry techniques, and identification of new antibacterial targets through genomic technologies, held promise as new approaches in antibacterial drug discovery.

Computational chemistry aided by improvements in techniques to determine and study protein structure, was also critical through facilitation of rational drug design. However, despite the promise of genomics, and an increased understanding of the biosynthesis of

natural products, these approaches have generally been unsuccessful in identifying new compound classes.87,(see also 94,102)

If no progress is made, we may enter a post-antibiotic era where bacterial infections become a leading cause of death,84b which is fearfully reminiscent of pre-antibiotic times.

For this reason, it is imperative that the use of existing antibiotics be managed more effectively.84 There is hope, however, that new antibiotic targets, chemotypes, and therapeutic strategies may still be discovered. The improvements in laboratory cultivation of challenging microorganisms is expected to facilitate the discovery of new scaffolds for antibiotics. An increased understanding of underlying complex mechanisms presented by existing antibiotics could help uncover possible combination therapy approaches.

Development of narrow spectrum agents via mining of discarded antibiotics (e.g. discovery of daptomycin) has potential to expand the available target space beyond cell wall synthesis inhibitors, DNA and protein synthesis inhibition, or membrane integrity disruption.

Currently infections are treated empirically and that requires broad spectrum agents.

Diagnostics would transform the field through the effective use of single-pathogen antibiotics. One advantage of this therapeutic approach is the reduced impact on the patient microbiome.87

In addition to efforts to reinvigorate antibacterial research by incentivizing the process from a policy and financial perspective, there is still much to do from a scientific perspective. For example, one specific obstacle in the discovery of new agents is that of the dual membrane barrier presented by Gram-negative (GN) bacteria (Figure 3.1),84a as

opposed to the single membrane found in the Gram-positive (GP) bacteria. The GN cell wall presents a significant challenge because many drug targets are intracellular. These two membranes in GN bacteria possess orthogonal properties (discussed in further detail in

Section 3.12) increasing the challenge in designing compounds able to effectively penetrate both membranes (Figure 3.1).

Figure 3.1 Cell wall penetration of antibiotics in GP and GN bacteria.84a

Gram-negative bacteria also possess a number of efflux pumps actively driving out toxic substances. Furthermore, both GP and GN bacteria have developed resistance mechanisms decreasing efficacy of antibiotics.83 Some common resistance mechanisms include metabolic resistance via -lactamases, upregulation of bacterial target in the case of vancomycin resistance, and target-mediated resistance with the fluoroquinolone class of antibiotics.83,87 An important route of entry for small molecules through GN bacterial membranes is porins.83 Mutations to these porins can further increase resistance through 68

reduced uptake.83 This combination of defenses makes penetration a formidable challenge and a major bottleneck in the development of new agents. While high target potency is often achievable, attaining whole-cell activity due to these barriers is difficult and can contribute to the high doses of antibiotics in the clinic in order to achieve activity. Several groups have approached the penetration problem from various angles with variable levels of success.83,88–95

3.1.2 Bacterial membrane structural considerations

The GN cell wall is a complex multi-layered structure which prevents the entry of toxic substances. It is comprised of two membranes with orthogonal characteristics, and numerous efflux pumps lining the inner membrane (Figure 3.2).96 From the extracellular space, basic components of the cell envelope are an outer membrane (OM), a peptidoglycan

(PG) layer providing rigidity to the cell, and an inner membrane (IM). The OM bilayer is asymmetric possessing an outer lipopolysaccharide (LPS) leaflet and an inner phospholipid leaflet, whereas the IM is a symmetric phospholipid bilayer. The outer leaflet of the OM is comprised of lipid A, which is attached to a network of sugars that are well hydrated resulting in the formation of a significant barrier to penetration of hydrophobic molecules.

Simultaneously the LPS presents a barrier to hydrophilic compounds by preventing diffusion, making the OM relatively impervious to both hydrophobic and hydrophilic molecules.96 Entry of nutrients is therefore controlled and mediated by water-filled -barrel proteins called porins.

Figure 3.2 Cell wall structure of representative GN bacteria E. coli.96

General porins in E. coli are OmpF and OmpC. In addition to showing a preference for hydrophilic substrates, these porins are size restricted. Furthermore, there are substrate specific porins, characteristically found in Pseudomonas and Acinetobacter. In the case of

Pseudomonas, one example of a substrate specific porin that is responsible for the uptake of amino acid building blocks through facilitated diffusion is OprD.96 Interestingly, these solute specific porins can be employed to facilitate transport of structurally similar substrates as in the case of OprD for the uptake of carbapenems.97,98 Lastly, there are energy-dependent uptake systems, most well characterized being the FhuA/TonB system responsible for the uptake of iron-siderophores spanning the OM and associated with IM

96 proteins for energy transduction.

The IM of GN bacteria behaves in a manner much like standard phospholipid bilayers, however more selective against hydrophilic molecules.96 Transport of highly

70 polar or charged species through the IM is mediated by substrate specific transporters that use an energy gradient to facilitate movement of molecules through the membrane. The proton gradient of the proton motive force (PMF) aids in the transport of weakly charged species with relative lipophilicity. The role of removing toxic substances from the

99–101 periplasmic space is fulfilled by the RND efflux pumps which span the IM and OM.

The AcrAB/TolC system of E. coli serves as a model for the RND pumps. These pumps have 3 components, the first of which is the AcrB, reaching from the IM into the periplasm, followed by AcrA, a membrane fusion protein and lastly, the OM component TolC. AcrB has been shown to possess broad specificity making it difficult to predict if an investigational drug will act as a substrate for the efflux pump. Although there is evidence suggesting a preference for amphiphilic compounds.99

3.1.3 Strategies to permeate the outer membrane

Aided by a concentration gradient, hydrophobic agents can penetrate the outer membrane of GN bacteria by diffusion through the lipid bilayer.96 Hydrophilic compounds, on the other hand, pass through the outer membrane by way of porins. Outer membrane permeation is then followed by diffusion through the periplasmic space and then passage through the inner membrane resulting in cytosolic accumulation. In addition to the RND efflux pumps, the inner membrane is lined with pumps that can drive compounds back into the periplasmic space.

Attempts have been made to retroactively analyze physicochemical properties of known antibiotics with high accumulation with the goal of establishing guidelines for future R&D efforts.95, 102–106 Thus far the community as a whole has not been successful in determining characteristics that generally afford small molecule accumulation in bacteria.

Over 50 years ago it was discovered that converting penicillin G into ampicillin conferred broad spectrum activity by incorporation of a primary amine to the benzylic position of the side chain. This discovery however has not been generalizable to other compound classes.

Hergenrother and co-workers recently reported an elegant analysis of an unbiased set of structurally diverse compounds and quantified their ability to accumulate in E. coli.

Although this research effort was aimed at developing guidelines to covert GP-only agents into GN agents, and not necessarily a universal ground-up approach to creating GN active agents, the rules are generally applicable in considering accumulation of antibiotics in GN bacteria. Following a careful analysis of whole cell accumulation data, SAR and computational studies led to the development of predictive recommendations to afford appreciable concentrations of antibacterial agents inside GN bacteria.107

72 Figure 3.3 Effect of primary amines on accumulation in GN bacteria (relative to controls).

One of the key conclusions of that study was the effect of amines on the accumulation of a compound (Figure 3.3). In particular, compounds containing primary amines showed the most significant increases in penetration into GN bacteria. Subsequent substitutions on the amine nitrogen were shown to result in reduced accumulation as shown in 3.1 – 3.5 (Figure 3.3). Of the high accumulating compounds containing primary amines,

107 no correlation was observed for ClogD7.4 nor molecular weight. Furthermore, the steric hindrance surrounding the amine had an impact on accumulation. The more exposed or separated an amine was from the sterically encumbered core, the greater the accumulation

(comparison of 3.6 – 3.8, Figure 3.3). Although the presence of amines, especially primary amines, seemed to result in a high propensity for accumulation, this alone was not a sufficient characteristic. Hergenrother’s work revealed the importance of flexibility in a

73 molecule and its global shape in determining accumulation in GN bacteria. Comparing two compounds with similar molecular weight, the more rigid backbone having less rotatable bonds was found to exhibit higher accumulation in E. coli (Figure 4). Globularity was also found to be an important factor in determining accumulation in these studies. In this instance, globularity refers to a descriptor of the three-dimensional shape of a compound relative to a standard. The general trend observed was that low globularity corresponded to high accumulation indicating a preference for compounds with a relatively fixed linear shape for easier passage through porins (3.9 – 3.12, Figure 3.4).107 Although this work is an impressive and potentially useful predictive model for accumulation of antibiotics into GN bacteria, it has yet to be fully applied to an active drug discovery effort.

An alternative approach scientists have been using to circumvent the penetration barrier problem is to exploit native nutrient uptake systems to facilitate uptake of drug across the

GN cell wall. The remainder of this chapter will discuss this approach.

74 Figure 3.4 Effect of rigidity and globularity on accumulation in GN bacteria (relative to controls).

3.2 Iron as an essential micronutrient for bacteria

The micronutrient iron is essential for both mammalian and bacterial cell function and survival. Iron is required for efficient infectivity of bacteria and progression of an infection making its uptake from the surrounding microenvironment critical. Iron (III) is the biologically available form of iron, despite the fact that it is highly insoluble and free levels are present in very low concentrations. As a result, bacteria have developed highly specific methods to sequester iron from the environment. This sequestration is mediated by iron chelators called siderophores. The siderophores form complexes with Fe3+ followed by recognition by specific outer membrane transporters initiating an energy dependent process that results in the translocation of the iron-siderophore complex through the membrane and into the cytoplasm.108 Siderophores are designed to give a selective growth

75 advantage to the producing organism and more than 500 unique siderophore structures have been discovered and reported, although some overlap does exist in siderophore activity in different microbes.109,110 Interestingly, some bacteria produce natural siderophore-antibiotic complexes known as sideromycins. Representative compounds in this class are the salmycins111 and albomycins112 whose warheads are a potent aminoglycoside and a tRNA synthetase inhibitor respectively. The activity of the antibiotic is exposed upon release from the siderophore-complex following reduction of the Fe3+ to

Fe2+ via a reductase enzyme, and subsequent binding to the intracellular cytoplasmic target.116

Over many decades several research efforts have been aimed at mimicking the sideromycin strategy in antimicrobial therapy for transport of antibiotics across the Gram- negative cell wall. Siderophore--lactam conjugates have been shown to be active without the need for a release mechanism, however they may be prone to inactivation by -lactam.

Other conjugates with a warhead that has a cytoplasmic target like ciprofloxacin have generally been unsuccessful, likely due to poor passage through the IM of failure to achieve an appropriate release mechanism.113

3.3 Iron-siderophore uptake by Gram-negative bacteria: structural considerations

One example of exploiting native nutrient uptake pathways to facilitate transport of antibiotics across bacterial membranes is exemplified by the iron-siderophore uptake system. The uptake of iron can occur through free diffusion under conditions of high iron 76

concentration in a low affinity mechanism. However, under iron deficient conditions, which is considered to be the state in an active infection, siderophores and high affinity membrane transporters are engaged to facilitate uptake. Siderophores are low molecular weight organic compounds that form complexes with Fe3+ of high thermodynamic stability.114,115 Despite the high structural diversity encountered in natural siderophores, generally only a few functional groups are used to actually bind iron. Siderophores can be classified based on the iron-binding motif. The three main classes are hydroxamates 3.13, catecholates 3.14, and -hydroxycarboxylates 3.15 (Figure 3.5).116 These compound classes contain oxygen-donor moieties that behave as hexadentate ligands forming octahedral complexes with the iron.114 Due to the chelating ability of these functional groups, it has been shown that siderophores can also bind other heavy metal ions such as

Al3+, Zn2+, Ga3+, Cr3+, Pu3+, and Pu4+.115,117

77 Figure 3.5 Common classes of naturally occurring siderophores.

The iron transport system consists of outer membrane receptors such as FhuA,

FecA, FepA, in E. coli and FptA, FpvA in Pseudomonas aeruginosa, all showing striking similarities by X-ray crystal structure analysis.116 These proteins are responsible for initial recognition and binding of the iron-siderophore complex, and are associated with the TonB complex through the inner leaflet of the OM. This TonB complex is responsible for transducing the proton motive force required for active transport of the iron-siderophore complex (Figure 3.6).109,117–119 Once in the periplasm, the iron-siderophore complex is recognized and bound to its respective periplasmic binding protein which is then transported through the inner membrane by an ATP-binding cassette transporter system into the cytoplasm. These transporters use energy from ATP hydrolysis to pump substrates into the cytoplasm. 120 In the cytoplasm, the iron is released by reduction of Fe3+ to Fe2+ by 78 an iron reductase enzyme releasing the free siderophore which is either degraded or effluxed out and recycled.116

Figure 3.6 Iron-siderophore uptake in E. coli 116

This approach to smuggle antimicrobial agents into the cytoplasm by way of attaching an iron chelating group has been explored for decades. Of note is the group of

M. J. Miller making significant contributions in this field by developing high affinity synthetic siderophores bound to daptomycin and demonstrating potent activity against multidrug resistant strains of A. baumannii.108 Several compounds using this approach have progressed to clinical development recently (MC-1 by Pfizer,121 S-649266 3.20 by

Shionogi,122 and BAL30072 by Basilea Pharmaceutica123) and S-649266 continues to show promise in utilizing nutrient uptake systems to enable movement of antibiotics through the relatively impervious membranes of Gram-negative bacteria (Figure 3.7).

79 Figure 3.7 Iron-binding antibiotics in recent clinical development

3.4 Identification of fatty acid transport system as a promising uptake pathway

The last few decades have seen a steady decline in the number of novel antibiotics, especially those derived from traditional target screens for the treatment of MDR pathogens. This failure to address the growing unmet need for new agents requires innovative and disruptive drug-discovery approaches. Similar to the iron uptake approach, we proposed the possibility of employing bacterial transporters involved in uptake of nutrients to facilitate movement of antibiotic across the GN cell wall. Following careful genomic analysis of known nutrient uptake systems employed by bacteria, the long chain fatty acid uptake system presented a well-characterized system for potential exploitation.

One impact of this approach is the potential to confer broad spectrum activity to GP-only agents.

3.4.1 Fatty acid uptake in E. coli

Exogenous long chain fatty acids (LCFAs) are required for intracellular signaling, gene expression, and as a source of metabolic energy. E. coli utilize the Fad-system to shuttle long chain fatty acids into the cell.124b In E. coli FadL shuttles LCFAs across the outer membrane (OM). Following diffusion through the inner membrane (IM), conversion of the fatty acid into its corresponding CoA-thioester begins via the membrane bound CoA- synthetase FadD. This activation step of the acid may render it incapable of diffusing back out through the inner membrane. The intracellular trapping of the LCFA creates a concentration gradient in favoring uptake from the extracellular space (Figure 3.8).

Figure 3.8 Overview fatty acid uptake in Gram-negative bacteria.

3.4.2 FadL structural considerations

In 2004 Bert van den Berg and coworkers disclosed the structure of the outer membrane of the LCFA transport protein, FadL (Figure 3.9).124 This important report described the structural properties on the monomeric protein composed of 14 antiparallel strands (A) making up a hollow barrel approximately 50 Å long.124 This barrel is plugged by a hatch domain at the N-terminus containing three alpha helices. Additionally, the N- terminus extends into the extracellular space through the barrel in a manner unique to FadL.

Strand S3 consists of an atypical kink pointing inwards resulting in the disruption of the hydrogen-bonding network between the -sheets (Figure 3.9 A and B). The stabilization of this kink (orange) is afforded by a hydrogen bonding network between S3 and the N- terminus resides (purple), and ultimately leads to a gap in the wall (Figure 3.9B). The monoclinic crystals also showed a solvent exposed groove of hydrophobic residues between loops L3 and L4, as well as a hydrophobic pocket (P) located on the extracellular portion of the membrane whose entrance is solvent accessible near L3 and L4.

Figure 3.9 Crystal structure of FadL. (A) Side view of FadL. (B) Side view of FadL rotated 45° relative to (A). Presumed position of the membrane boundaries (M) represented by 124b horizontal lines. Extracellular space (E), Periplasm (P), Extracellular loops (L3 and L4).

The model for entry of LCFAs is depicted in Figure 3.10.124 The process begins by entrance of the substrate into the hydrophobic groove between loops L3 and L4 followed by diffusion into the binding pocket. The N-terminus undergoes conformational changes that result in decreased binding affinity for the pocket. This diminished affinity for the substrate, and movements in the hatch and kink region drive the substrate toward the barrel wall. These changes facilitate the movement of the substrate directly into the periplasm.

From the periplasm it is assumed that LCFAs, in their protonated form, diffuse readily through the IM into the cytoplasm.

Figure 3.10 Schematic mechanism for transport through FadL

Unlike other outer membrane transport proteins, such as the siderophore transporter

FhuA, FadL does not require energy input for its function. While FhuA is associated with the inner membrane protein TonB, responsible for energy output derived from the PMF that induces conformational changes in the transporter facilitating release of the substrate from a high affinity binding site, FadL is TonB independent. Binding of substrates to FadL is low affinity and the PMF is not required for transporter function. Movement is therefore presumed to be dependent on the overall concentration gradient of LCFAs from outside to inside the cell and the spontaneous conformational changes in the channel. We hypothesized that we could exploit this nutrient uptake pathway to facilitate the transport of payloads across the cell envelope by appending a long chain fatty acid as the recognition element to known antibiotic payloads.

3.5 Hijacking the fatty acid uptake system: experimental design

To probe the feasibility of exploiting the LCFA uptake system for the transport of drugs across the GN cell wall, we employed a late-stage functionalization strategy to synthesize tool compounds to serve as mechanistic probes. The general approach involved attaching a fatty acid recognition element to either a (i) fluorescent payload for imaging studies, or (ii) a known antibiotic to assess antimicrobial activity post derivatization. The goal of these investigations was to ascertain if we could achieve active uptake of antibiotic- fatty acid derivatives via the LCFA system to afford cytosolic accumulation of the antimicrobial agent and the desired therapeutic response of bacterial cell death. Antibiotic- fatty acid analogues were synthesized utilizing traditional functional group interconversions from an advanced intermediate to rapidly assess probability of success as a general approach.

3.5.1 In vitro fluorescence imaging of E. coli by BODIPY-fatty acid conjugates

The first payload investigated was a fluorescent probe, boron-dipyrromethene

(BODIPY) moiety, conjugated to a C5 or C12 fatty acid. Confocal fluorescence microscopy was employed to investigate uptake of the probe by bacteria and determine the level of intracellular accumulation. Using the BW25113 wild type E. coli strain, a chain length dependent uptake was observed with preferential uptake of the C12-probe over the

C5 probe (Figure 3.11). This uptake was also demonstrated to be due to intracellular

accumulation of the payload, rather than solely association with the cell membrane(s), through cross-sectional analyses (data not shown). Surprisingly, when an isogenic fadL knockout strain was subjected to the same two probes, the C12 conjugate showed comparable fluorescence to the wild-type control, suggesting a limited role for FadL in the uptake of the C12 conjugate. Interestingly, the fadD knockout strain showed significantly lower fluorescence. In addition to demonstrating a chain-length dependency for the uptake, the fluorescence imaging data revealed the necessity of FadD to facilitate this process. This initial finding prompted further exploration of this system to facilitate uptake of antimicrobial agents across the bacterial cell membranes.

Figure 3.11 Fluorescently-labeled fatty acids in E. coli.

3.5.2 Fatty acid conjugate synthesis and biological evaluation (SAR)

Utilizing a known bacterial topoisomerase inhibitor, ciprofloxacin (cipro), we hypothesized that in a manner similar to the fluorescence probes, we could attach fatty acids of varying lengths to the piperazine NH and evaluate the antimicrobial activity of the conjugates. A survey of fluoroquinolone literature demonstrated tolerability of functionalization or substitution of the piperazine ring.125a In addition to investigating the chain length dependency, we sought to evaluate the optimal type of linker between the fatty acid and the payload. This strategy allowed us to rapidly generate structure-activity relationships that would inform subsequent experiments.

To that end, several alkyl-linked analogues were generated by subjecting cipro to bromo-alkyl ester electrophiles of different chain-lengths, followed by saponification under basic conditions, and treatment with trifluoroacetic acid to afford the desired analogues as the trifluoroacetate salts (Scheme 3.1). Previous data collected on analogues

3.26, 3.27, 3.29, 3.31, and 3.32 demonstrated a strong dependence on fatty acid chain length. An increase in fatty acid chain length corresponded to a more potent antibacterial activity evidenced by the drop in minimum inhibitory concentration (MIC) going from C5 to C10, with the exception of C7. This observation prompted us to synthesize C8 (3.28) and C10 (3.30) analogues using conditions highlighted in Scheme 3.1. Biological activity of both the acids and the corresponding esters was evaluated in the wild-type E. coli strain as well as the fadD, and fadL mutants.

Scheme 3.1 Synthesis of alkyl-linked cipro-FA conjugates

MIC (g/mL); E. coli Fatty acid- Fatty acid ciprofloxacin chain conjugate length WT BW25113 FadD FadL

3.26 C5 4 4 4

3.27 C7 0.5 1 0.5

3.28 C8 0.25 1 0.25

3.29 C9 1 4 1

3.30 C10 2 8 2

3.31 C11 4 16 4

3.32 C12 8 16 8

Cipro, 3.25 - 0.016 0.031 0.016

Table 3.1 Minimum inhibitory concentrations of alkyl-linked cipro-FA conjugates.

As shown in Table 3.1, the analogues demonstrated a chain length dependency consistent with the FA transport system. In addition, they also showed dependence on

FadD, but not FadL, consistent with imaging data. C8-cipro (3.28) became the lead compound in the study showing good MICs and FadD dependence in E. coli. Starting from a chain length of C5, the antimicrobial activity improved with an increase in chain length up to C8. Although the best compound in this series C8-cipro (WT MIC 0.25) had a 16- fold greater MIC compared to the positive control, ciprofloxacin, the activity observed was nonetheless impressive considering the attachment of a greasy tail to the compound, which could have a detrimental effect on the permeability relative to a highly optimized cipro.

To investigate the type of linkage between the fatty acid and the piperazine of cipro, several triazole and amide containing conjugates were synthesized under standard copper catalyzed cycloaddition conditions (Scheme 3.2), and T3P mediated amidation conditions respectively (Scheme 3.3). The biological activity was evaluated in the same E. coli strains as in the above studies with the alkyl linkers. The triazole conjugates showed chain length dependency in E. coli MICs: C11

89 Scheme 3.2 Synthesis of triazole-linked cipro-FA conjugates

MIC (g/mL); E. coli Fatty acid- Fatty acid ciprofloxacin chain conjugate length WT BW25113 FadD FadL

3.34 C8 32/64 128 32

3.35 C9 32 128 32

3.36 C10 32 128 32

3.37 C11 16/32 64 16/32

Table 3.2 Minimum inhibitory concentrations of triazole-linked cipro-FA conjugates.

For the amide series, only C5 and C9 analogues were synthesized based on previously established SAR. Using T3P to activate the fatty acid to nucleophilic attack by the piperazine nitrogen, C5 and C9 analogues were generated (Scheme 3.3). These conjugates gave weak wild-type MICs, and due to this diminished activity profile (64-fold loss in activity between C9 alkyl-linked and C9 amide-linked), FadD dependence could not be determined conclusively (Table 3.3). 90

Scheme 3.3 Synthesis of amide-linked cipro-FA conjugates

MIC (g/mL); E. coli Fatty acid- Fatty acid ciprofloxacin chain conjugate length WT BW25113 FadD FadL

3.38 C5 acid 64 >64 64

3.39 C9 acid 64 >64 64

Table 3.3 Minimum inhibitory concentrations of amide-linked cipro-FA conjugates.

3.5.3 In vitro evaluation: discussion

In general, a chain-length dependence was observed for the alkyl-linked and trizole- linked fatty acid conjugates. Interestingly, of the chain-lengths evaluated, derivatives consisting of what would be characterized as a medium chain length (C8 and C9), demonstrated greatest antimicrobial activity. The flexibility in the longer chains presents a significant entropic cost to binding, while the hydrophobicity may lead to the formation of more globular structures in an aqueous environment. With respect to the linker chemotype, alkyl-linked analogues were more potent than triazole- and amide-linked compounds. The triazole ring system confers some level of rigidity resulting in a defined orientation of the fatty acid arm. The reduction in activity observed for the triazole series may suggest strict conformational poses allowed for optimum substrate-protein interactions.

3.5.4 In vivo activity of cipro-fatty acid conjugates

Having established C8-cipro as the most active compound, the activity was evaluated in vivo using an immunocompromised rodent urinary tract infection model.

Results are depicted in Figure 3.12. Upon administration of C8-ciprofloxacin as the disodium salt, dose-dependent efficacy was observed with a reduction in bacterial burden.

This effect was attenuated in the fadD knockout mutant compared to wild type. Although the efficacious dose is higher than that of ciprofloxacin, the C8-cipro conjugate only had

6% free fraction compared to the 70% of ciprofloxacin alone, and it is well accepted that antibacterial activity in vivo depends primarily on the free fraction of the drug.125b In

addition, the MIC of the conjugate is 16-fold higher than that parent drug. This data is consistent with the fluorescence, MIC data, all of which are consistent with FadD- dependent uptake of C8-cipro conjugate.

Figure 3.12 C8-Cipro in vivo efficacy in immunocompromised rodent urinary tract infection model.

All studies were conducted in accordance with the GSK Policy on the Care, Welfare and

Treatment of Laboratory Animals and were reviewed the Institutional Animal Care and

Use Committee either at GSK or by the ethical review process at the institution where the work was performed.

3.6 Conclusion

A model for the uptake of the cipro-fatty acid conjugates is shown in Figure 3.13.

Cipro-fatty acid conjugates (3.42) most likely passes through the outer membrane by way of the general porin, OmpF (unpublished data, Jacobs University, Bremen). OmpF is known to be responsible for the uptake of ciprofloxacin. Following passage through OmpF, the conjugate partition through the inner membrane and is successively converted to its

CoA analogue (3.43) by the fatty acyl CoA synthetase FadD, with concomitant ATP hydrolysis. The activation of the FA drives further uptake of more conjugate by creating a concentration gradient in substrate sink effect (Figure 3.13). The CoA, being highly negatively charged is unable to diffuse out of the cell and thus accumulates intracellularly.

Figure 3.13 Revised model of Cipro-FA conjugates

There is a long history of incorporating iron-chelating motifs into drug candidates to aid in transport across the outer membrane using the iron-siderophore pathways. Here we have demonstrated that a new nutrient uptake pathway, the long chain fatty acid transport system, can be used to facilitate uptake of antimicrobials agents. Using a

BODIPY fluorescent probe and ciprofloxacin conjugated to fatty acid chains, chain length and FadD-dependent uptake was demonstrated. In addition to showing in vitro activity, in vivo efficacy was shown in a rodent urinary tract infection model. In these investigations ciprofloxacin was chosen as the test agent due to the large literature on the fluoroquinolones suggesting modification at the piperazine NH is generally well tolerated.

Cipro, however, is a broad-spectrum antibiotic with intrinsically good OM and IM penetration characteristics. This approach could ultimately demonstrate its utility for agents that lack GN activity. Ideally, future studies would employ a known GP-only agent and evaluate feasibility of conversion into an effective broad-spectrum antibiotic by simply linking fatty acids to the drug of interest. It would also be interesting to explore cleavable linkers. One can envision fatty acid-mediated uptake, followed by removal of the long fatty acid chain to release free payload into the cytosol for protein target engagement in a manner that resembles a prodrug approach. The impact of this proof-of-principle work lies in providing insight into the potential of exploring other native nutrient uptake pathways to facilitate uptake of antimicrobial agents through the GN membranes. In addition, these findings could propel new directions of the conversion of GP only agents into effective GN agents.

Chapter 4. Combatting Antimicrobial Resistance through Biofilm Interference

This project was a collaborative effort between the lab of Professor Fuchs in the Division of Medicinal Chemistry and Professor Gunn in the Department of Microbial Infection and Immunity, Infectious Diseases Institute.

Contributions other than myself:

1. Biological evaluation was done by Jasmine S. Moshiri, Darpan Kaur, and Jenna

Sandala (Gunn Lab).

4.1 Formation of bacterial biofilms

Bacterial biofilms are colonies of bacteria embedded in a polymeric matrix comprised of DNA, polysaccharides and proteins attached to inert or biotic surfaces.126

Biofilms can be comprised of a single species of bacteria or multiple species. The formation of biofilms is a multi-step process that is initiated by bacteria adhering to a surface, followed by the formation of microcolonies and subsequent exopolysaccharide (EPS) production (Figure 4.1). The production of this extracellular matrix results in a firmer attachment. Following maturation, the biofilm can burst open, releasing phenotypically distinct planktonic cells, or form a new biofilm (Figure 4.1).126

Figure 4.1 Stages of bacterial biofilm formation

Cells contained within a mature biofilm, sessile cells, require higher doses of antimicrobial drugs compared to planktonic cells. The EPS itself represents a high penetration barrier to antibiotics, and due to the slow entry of drugs through the matrix, resistance mechanisms are presumed to develop within the biofilm. Moreover, this chemical environment can also 97

affect growth phases of bacteria having an additional impact on compound efficacy. The containment of bacteria within the biofilm allows for adaptive stress responses to develop, including the growth of a small population of highly tolerant cells known as persister cells.125 Considering the fact that 65-80% of infections are biofilm-related, this presents a significant challenge toward the treatment and eradication of infectious diseases. There is a great need for agents that can either disperse existing biofilms, thus making them susceptible to antimicrobial agents, or inhibit the formation of biofilms (Figure 4.1).

4.1.1 Impact of biofilms on development of S. Typhi chronic carriage state

Chronic infections are often difficult to eradicate due to the presence of resistant bacteria embedded within a biofilm that is impervious to antimicrobial agents.125 Bacteria within a biofilm also experience physiological changes enabling emergence into persister cells.127 Salmonella enterica serovar Typhi (S. Typhi), is one example of a bacterial infection reliant on biofilms to generate and propagate persister cells, resulting in a chronic carriage state. The human specific pathogen S. Typhi is the causative agent for typhoid fever, affecting millions of people, mainly in South-eastern Asia, South-central Asia, and

Southern Africa.128 The infection of new hosts is facilitated by contaminated food and water. An estimated 11.9 million cases of S. Typhi were reported in 2010 and a corresponding 129,000 deaths associated with the infection according to a 2010 study.129

An acute S. Typhi infection manifests in the small intestine first, followed by systemic progression via phagocytosis by macrophages to afford infections in the bone

marrow, liver, spleen, ileum, and gallbladder.130 In the event that the acute infection is resolved, 2-5% of hosts will become asymptomatic chronic carriers of persistent bacteria in the gallbladder.131,132 During this pathophysiological state of S. Typhi carriage, biofilms form on the surface of gallstones and enter the epithelium.133 In a study by Schioler, 88% of carriers were found to have gallstones, providing more evidence of the role of gallstone biofilms in sustaining the chronic carriage state.134 Persistence of bacteria in this state allows for intermittent shedding into feces, propagating transmission of new infections if the fecal matter contaminates water systems.135 The current standard of care for typhoid carriage involves fluoroquinolone treatment.136 However, the emergence of multidrug resistant strains of Salmonella presents a significant dilemma in combatting chronic carriage S. Typhi infections.137 More recently, Gunn and coworkers demonstrated the difficulty in clearing aggregate Salmonella infections on gallstone surfaces using antibiotic therapy in a murine model.138

4.2 Development of new anti-biofilm agents

Several anti-biofilm agents derived or inspired by natural products have been identified and are under investigation as this holds the potential to eradicate the carriage state of many bacterial infections including those caused by S. Typhi. Many recently developed anti-biofilm agents do not affect cell survival which is considered beneficial in that development of resistance mechanisms to these agents is not anticipated.139 Ideal agents would be those that can disrupt existing biofilms, releasing planktonic cells that can

be eradicated with standard antimicrobial therapy. Alternatively, agents that prevent formation of bacterial biofilms are advantageous as a potential co-therapeutic option for infections that are known to rely on the biofilm lifestyle.

4.2.1 Anti-biofilm natural products

Several marine natural products consist of 2-aminoimidazole functionality which is deemed a guanidine mimetic. An example of this class is a natural product isolated from

Agelas conifer known as oroidin (4.1, Figure 4.2).139 This 2-aminoimidazole moiety has been found to be required for the anti-biofilm activity of this class. Numerous 2- aminoimidazole derived anti-biofilm agents developed by the Melander group have shown efficient anti-biofilm activity profiles.140–144 Another marine-derived class of natural products secreted by sponges to protect against predators are the bromopyrrole alkaloids.

4-Thiazolidinone bromopyrrole derivatives, exemplified by 4.2, are potent anti-biofilm agents against S. aureus biofilms with an MIC of 0.78 g/mL (Figure 4.2).145

Inhibition of quorum sensing, the chemical communication process bacteria use to signal biochemical processes necessary for survival, and replication, has also been shown to have an effect on biofilm formation. For example, the bromofuranone class of natural products (4.3) has been found to inhibit this quorum sensing process, thus having an effect of the biofilm lifestyle of certain bacterial strains.139

100

Figure 4.2 Anti-biofilm natural products

Indole derivatives have also shown promising anti-biofilm activity linked to intracellular signaling events. The degradation of the amino acid tryptophan by tryptophanase produces indole, which is a known intracellular signaling molecule in some bacterial species.146 In the clinically relevant E. coli strain O157:H7, indole was found to inhibit biofilm formation at 500 M.147,148 Furthermore, upon investigation of gene expression, biofilms treated with indole were found to have up to four-fold reduction of expression of cold shock regulator genes (cspGH) and modulation of phosphate related genes.147 It has also been demonstrated that oxidized derivatives of indole such as 5-hydroxyindole (4.4, Figure 4.2), and 7- hydroxyindole are more efficient at reducing biofilm formation in E. coli (up to 10-fold compared to indole at 6-fold).149 Natural products are proving to be viable leads for new anti-biofilm agents, whose continued development holds promise.

101

4.2.2 Discovery of JK-1 and primary activity profile

In efforts to identify novel biofilm dispersal agents, the Gunn lab employed a high- throughput assay developed in-house to screen a library of 3000 adenosine mimetics purchased from Cambridge. A flowchart depicting the screening process and prioritization of hits from this library is provided in Figure 4.3. This investigation led to the identification of a thienopyrimidinone compound (7955004), which was subsequently referred to as JK-

1. In these studies, JK-1 was shown to inhibit biofilms at an EC50 of 7.27 M in a rapid attachment assay against the wild-type S. Typhimurium strain JSG210.150 Time-dosing analysis indicated that the activity was based on early stage inhibition of biofilm formation.

Figure 4.3 Identification of JK-1 from ATP mimetic library

In addition, JK-1 was also found to inhibit biofilm formation in Acinetobacter baumannii at an EC50 of 6.05 M. Both phenotypic activities were not dependent on bacterial killing or bacteriostatic effects. Furthermore, JK-1 did not exhibit acute toxicity toward host

102

150 mammalian HepG2 cells at a concentration of 50 M, 6-fold greater than the EC50. As a result of this promising initial activity, JK-1 was chosen as the hit compound for development into a viable lead compound.

4.2.3 Retrosynthetic strategy to JK-1 and pyrazole analogs

In order to validate the result from the initial high throughput screen, JK-1 would need to be re-synthesized and characterized to confirm purity and structural identity and to generate the compound in appreciable quantities. To this end, a retrosynthetic analysis was performed on JK-1 (4.5), and a synthetic plan was developed (Scheme 4.1). The synthetic approach relied on key disconnections that would allow for introduction of variable functional groups from a common intermediate, thienopyrimidinone 4.6.

Scheme 4.1 Retrosynthetic analysis of JK-1

103

The synthesis of JK-1 (4.5) was envisioned to arise from a Knorr-type cyclization to construct the 5-hydroxypyrazole. The precursor to the Knorr-type cyclization, a -keto ester, would be introduced through alkylation of the sulfur atom of thienopyrimidinone

(4.6). The required thienopyrimidinone (4.6) would be prepared from isothiocyanate (4.7) in a 2-step process involving formation of a furanyl-bearing thiourea, followed by a base- mediated cyclization form the pyrimidinone ring system. The key transformation in this analysis was expected to be the formation of the isothiocyanate directly from the aminothiophene 4.8, which in turn would arise from cyclopentanone using Gewald’s method of constructing substituted thiophene rings (Scheme 4.1). The modular strategy employed in constructing JK-1 would facilitate the introduction of variable moieties at key stages in the sequence providing a method to rapidly generate analogs in a highly divergent fashion from common intermediates. Toward the generation of analogues, three key areas for introduction of variability were identified, the hydroxypyrazole, furan, and cyclopentathiophene, while maintaining the thienopyrimidinone core.

4.2.4 Optimization of arylisothiocyanate synthesis

Utilizing the method described by Gewald and coworkers in 1966 for the construction of substituted thiophene rings and optimized by Mojtahedi and coworkers. 2- aminothiophene (4.8) was prepared in aqueous media from cyclopentanone 4.9.151,152

Unbeknownst to us at the time, the direct conversion of an electron deficient amine to the

104

corresponding isothiocyanate is a challenging transformation to carry-out on scale in an environmentally-friendly and health-conscious manner. On the other hand, several methods have been reported for the conversion of electron rich aryl amines into isothiocyanates with excellent efficiency.

Scheme 4.2 Investigation into conversion of arylamine into isothiocyanate

The presence of an electron deficient group in conjugation with the -system of the arene, however, significantly reduces the efficiency of the transformation, presumably due to the decreased electron density on the amine nitrogen. In an attempt to find the most suitable method for the large-scale preparation of the isothiocyanate, several conditions were investigated (Scheme 4.2).

Avoiding the use of highly toxic thiophosgene (Scheme 4.2, equation I), the most environmentally friendly route was the generation of a dithiocarbamate intermediate, followed by desulfurylating with a suitable electrophile. The method reported by Boas and 105

co-workers utilizing di-tert-butyl dicarbonate as the desulfurizing reagent (Scheme 4.2, equation II)153 afforded no target material. Efforts to decompose this intermediate to desired isothiocyanate were unsuccessful. Making use of tosyl chloride as the desulfurylating reagent afforded desired material in 40% yield (Scheme 4.2, equation

III).154 Following optimization, the maximum yield obtained for this reaction was 52%.

Although this was a significant improvement as it provided material for the progression of the synthesis toward JK-1, the significant loss of material at this step made it a bottleneck in the sequence warranting further investigation. With the thienopyrimidinone (4.6) being a key intermediate as the branching point for the generation of analogs it was imperative that a more efficient way to carry out this transformation be developed. After several investigations shown in Scheme 4.2, a 2-step protocol in which the aminothiophene was converted to an iminophosphorane in quantitative yield, followed by an Aza-Wittig type reaction with carbon disulfide afforded the isothiocyanate 4.7 in 80% yield over two steps

(Scheme 4.2, equation IV).155

4.2.5 Synthesis of JK-1 and pyrazole analogs

Following optimization of the conversion of arylamine 4.8 into isothiocyanate 4.7 the synthesis of JK-1 was completed according to Scheme 4.3. The final target molecule was purified by trituration, first with ethyl acetate and then with methanol, to afford a white powder that was insoluble in most organic solvents.

106

Scheme 4.3 Synthesis of JK-1

Structural confirmation of the target molecule was not a trivial task. Confirmation of the synthetic JK-1 material was made challenging by the ability of the hydroxypyrazole to exist as a mixture of pH dependent tautomers (Scheme 4.4). NMR analysis was performed in deuterated DMSO, and 2D spectroscopic techniques were utilized to confirm the structure in addition to comparison data to commercially acquired material.

Scheme 4.4 Hydroxypyrazole tautomeric forms

107

4.3 Biological activity evaluation of JK-1

Synthetic JK-1 showed similar activity to that observed in the primary assay at 5

M. Targeting the lowest hanging fruit, we investigated the importance of the hydroxypyrazole moiety. Toward that goal, several analogs were synthesized by subjecting pyrimidinone 4.6 to different electrophiles. Biofilm activity of JK-1 and corresponding pyrazole analogs was evaluated using the rapid attachment assay. The optical density of the plate at 570 nm, which corresponds to the amount of biofilm formed was measured for

S. Typhimurium exposed to 5 M and 10 M of compound. In all cases, pyrazole analogs had less ability to inhibit formation of biofilms compared to the parent compound JK-1

(4.5) (Table 4.1). These analogues showed that the pyrazole moiety was deemed crucial for activity due to the significant decrease in activity of all derivatives when this group was removed, including the thienopyrimidinone 4.6 (Table 4.1). Loss of hydrogen-bonding capabilities by replacement of pyrazole ring is assumed to be responsible for the drop in activity.

108

Analog Structure Biofilm inhibition at 5 M Biofilm inhibition at 10 M 4.5 JK-1 25.4, 22.9, 34.2 11.4, 10.6, 21.9

4.15 JK-2 5.5, -4.2, -5.3 -26.3, -9.8, 6.9

4.16 JK-3 11.4, -1.3, 0.3 -12.9, -2.9, -6.1

4.17 JK-4 3.0, -5.3, -1.2 -19.5, -9.4, -3.2 4.18 JK-5 -0.5, 2.4, -5.1 -10.9, -1.4, -16.8

4.19 JK-6 -0.3, -3.5, -10.8 -22.9, -2.1, -11.8

4.20 JK-7 6.1, -4.2, -1.5 -3.4, -0.6, 7.5

4.21 JK-8 7.6, -27.5 4.22 JK-10 12.3 -13.8 4.23 JK-11 7.6 -11.4

4.24 JK-12 7.2 -13.7 4.6 JK-13 5.3 3.1

4.25 JK-14 ND 8, 2

4.26 JK-15 ND 5, -15,

4.27 JK-16 ND 10, 10

Table 4.1 Biological activity of JK-1 pyrazole analogs

109

Solubility challenges were observed with all compounds including JK-1in the biochemical assay (1% DMSO solution). Even within the series, JK-1 to JK- 7 were tested in triplicate and significant variability was noted in the results. This variation was attributed to poor water solubility of the compounds. Manipulation of the pH conditions of the assay had to be employed in some cases to aid in the solubility. It became increasingly clear that more pronounced structural changes would be necessary to disrupt the planarity in the core, presumably allowing for tight packing of molecules in solution. In addition to addressing poor physicochemical properties of this compound series, rational analog design was hindered by the lack of an identified protein. While working on addressing the solubility challenges with the JK-1 series of compounds, the Gunn lab identified another compound

T315 with promising anti-biofilm activity. The remainder of the chapter will discuss our mechanistic investigations into the activity of T315.

4.4 Identification of a small molecule (OSU-T315) anti-biofilm agent against

Salmonella

A high-throughput screen of 90 kinase inhibitors previously synthesized at OSU and provided to the Gunn lab led to the identification of T315 as an early stage anti-biofilm agent against Salmonella.156 T315 had originally been designed and synthesized for the inhibition of integrin-linked kinase (ILK), implicated in the promotion of oncogenesis and tumor progression. This compound reduced biofilm formation in a dose-dependent manner with 59.4% reduction at 5 M (Figure 4.4). In addition, between concentrations of 1-5 M,

110

T315 was shown by Lee and co-workers to be non-cytotoxic to prostate and mammary epithelial cells.156 The promising anti-biofilm activity and lack of host cell toxicity prompted further investigation into its potential as a feasible small-molecule lead for drug development as an inhibitor of biofilms.

Figure 4.4 OSU-T315 anti-biofilm activity against S. Typhimurium

4.4.1 Activity profile of T315 against S. Typhi, S. Typhimurium and A. baumannii

To determine if the anti-biofilm activity of T315 (4.28) was due to diminshed bacterial growth, planktonic cell viability upon exposure to T315 was evaluated. Planktonic bacterial cell viability of S. Typhimurium in the presence of 10 M T315 (black squares) was not affected over a 24-hour period compared to the no-drug DMSO control (white circles) indicating the inhibition of biofilms devoid of bacteriostatic or bactericidal effects

(Figure 4.5 A). From a time-of-addition study, T315 was found to inhibit early stages of biofilm formation. When T315 was added 6 hours post initiation of assay, there was a significant decrease in inhibitory activity compared to early addition time points. A 5 M

111

concentration of T315 added 1-hour post-initial incubation resulted in 42% decrease in

biofilm formation, whereas administration at the 3-hour time-point resulted in 30%

reduction, and addition after 6 hours of incubation resulted in only 20% reduction in

biofilm formation (Figure 4.5B). T315 was also shown to be ineffective a dispersing

existing biofilm (data not shown).

A BA1011 C B

) 11 L

10 m 50 / 10

u 10 *** ** *

)

(

)

10 o

u 10 i 40

t 9 i

a 10 b

(

t 9 c

a 30

10 n

8 f

n 10

o 8 o

e 7

10 (

i e 7

l 10 t 10

a 6

A i 10

V 0 6 12 18 24 106 Time (h) 0 0 6 12 18 24 1 3 6 Time (h) Time of Drug Addition (h since initial incubation)

Figure 4.5 Activity profile of T315 against S. Typhimurium. (A) Cell viability. (B) Temporal addition assay.

4.4.2 Primary biological data for T315 anti-biofilm activity

Interestingly, initial evaluation of T315 demonstrated its ability to significantly

inhibit A. baumannii biofilm formation 61.2% at 10 M. However, T315 did not show any

anti-biofilm activity up to 20 M in P. aeruginosa PAO1. Further investigation into the

activity profile of T315 revealed its potential use in targeting the carriage state of typhoid

fever. The carriage state of typhoid cholesterol gallstones is mimicked in vitro by growing

bacterial biofilms on cholesterol-coated plates.135 Upon administration of T315 to this

simulated chronic typhoid state, biofilm formation was inhibited dose-dependently, with a 112

maximal effect of 98.2% at 100 M, the highest concentration tested (data not shown). The

EC50 of T315 against S. Typhi biofilm development was 21 M. Lastly, our collaborators in the Gunn lab investigated the potential of using T315 in combination with a clinical antimicrobial agent, ciprofloxacin, for the eradication of the typhoid carriage state.

4.4.3 Effect on biofilm formation of T315 and ciprofloxacin combination

Recently, Majtan and co-workers showed that sub-inhibitory doses of certain antibiotics have a protective effect against the formation of biofilms from S. enterica clinical isolates.157 The data generated by the Gunn lab suggested that a combination of sub-lethal doses of ciprofloxacin and 20 M T315 treatment resulted in a significant decrease of Salmonella biofilm formation, highlighting the potential for its use with low dose antibiotics (Figure 4.6A and 4.6B). It is important to note that this data supports the notion that sub-MIC doses of ciprofloxacin only exhibit anti-biofilm activity, although this effect was markedly increased upon co-administration with 20 M of T315 (Figure 4.6).

The biological activity data profile for T315 against Salmonella and Acinetobacter prompted further evaluation into the mechanism of action. To identify the putative target of T315 in biofilm development, a direct pull-down approach followed by a proteomics analysis was employed.

113

A B

)

100 f

)

100

n i 80

i f 80

i 60

( 60

(

40 c

A 40

20 20

3 i

0 r 0

0 1 5 o -7 -6 -5 -4

10 15 20 25 30 40 50 60 70 80 90 N 2.5 100 log10[T315] (M) [T315] (mM) C D A. S. Typhimurium B. S. Typhi 1.5 1.5 Ciprofloxacin Alone Ciprofloxacin Alone Ciprofloxacin + 20 µM T315 Ciprofloxacin + 20 µM T315

) )

m 1.0 m 1.0

0 0

7 7

D D

(

l l

o o

i 0.5 i 0.5

0.0 0.0 No CIP 1 2 3.9 7.8 15.6 31.3 10000 No CIP 1 2 3.9 7.8 15.6 31.3 10000 [Ciprofloxacin] (ng/mL) [Ciprofloxacin] (ng/mL) Figure 4.6 Effect of combination therapy on biofilm development. (A) S. Typhimurium. (B). S. Typhi

4.5 Development of T315 mechanistic probes for target identification

Inspired by the development of pull-down probes for epi-silvesterol by Rizzacasa and coworkers,158 two sites on T315 were identified for derivatization by attaching a biotin tag through a linker. Two probes were designed that would take advantage of the functionalization of the terminal piperazine NH to generate probe T315-S1, and the terminal amide to generate T315-S2 (Figure 4.7). The polyethylene glycol linker chain length was chosen to minimize any adverse steric interactions between the biotin- streptavidin complex and drug-protein complex isolated through affinity-based chromatography. Unfortunately, based on the design of these compounds and the

114

availability of T315, the compound would need to be re-synthesized with modifications in the route to accommodate the probe linkages.

Figure 4.7 Strategy toward the synthesis of T315-S1 probe

4.5.1 Synthesis of biotinylated probe T315-S1

The parent compound T315 was synthesized according to the original disclosure by Lee and co-workers.156 Starting from 4-bromoacetophenone, T315 was synthesized in

6 linear steps as a white free-flowing powder (13% overall yield). An advanced T315-S1 intermediate bearing a pendant alkyne (4.30) was synthesized through an alkylation of the piperazine nitrogen with propargyl bromide.

115

Scheme 4.5 Synthesis of T315-S1 probe

The resulting alkyne was ‘clicked’ onto a biotin-PEG-azide (4.31) to afford T315-S1

(0.52% overall yield, 11.3 mg, 8 linear steps) as a clear residue (Scheme 4.4).

116

4.5.2 Synthesis of biotinylated probe T315-S2

Toward the synthesis of T315-S2, the piperazine nitrogen was protected by a carboxybenzyl (CBz) group to prevent any undesired side reactions. Following formation of T315-alkyne 4.33, copper-(I)-catalyzed cycloaddition with the biotin tag (4.31) afforded the CBz-protected T315-S1 probe (4.36). Unbeknownst to us at the time, late-stage deprotection of the carboxybenzyl group would prove to be a significant challenge. Under reductive atmosphere, in the presence of a catalytic amount of Pd/C, the N-methylated adduct 4.34 was exclusively obtained (Scheme 4.5). A careful literature survey of this phenomenon highlighted a similar observation in the CBz deprotection of amino acids in methanol as the solvent. In fact, it is the methanol itself that serves as the carbon source via the palladium-mediated formation of formaldehyde. Despite the low levels of formaldehyde generated, any amount present will immediately condense with the free amine, driving the equilibrium in favor of more formaldehyde formation.

117

Scheme 4.6 Synthesis of CBz-protected T315-S2 probe

One can envision employing another solvent such as ethyl acetate as a reasonable way to avoid formation of this adduct. However, simply changing the solvent to ethyl acetate resulted in complete preservation of starting material, even at 70 psi (Table 4.2).

Another conceivable way to avoid this alkylation is to employ an acidic solution for the reduction. The immediate protonation of the free-amine would presumably eliminate its nucleophilicity, and corresponding side reactions. Unfortunately, when the reaction was carried out in 10% acetic acid, a 1:1 mixture of starting material:target compound was obtained. Separation of target material from this mixture was not a small feat. Despite many

118

attempts to drive reaction to completion, the reaction mixtures always contained both starting material (4.35) and product (4.36), so did employing Me3SiI (Table 4.2).

Table 4.2 Carboxybenzyl deprotection of T315-S2

Given the purification challenges encountered for the final compound, a different protecting group strategy was employed. Switching to a Boc protecting group led to the straightforward formation of T315-S2 (0.91% overall yield, 6.6 mg, MW 1002.17 g/mol) as an off-white film following deprotection in 20% TFA in dichloromethane.

119

Scheme 4.7 Successful synthesis of T315-S2 from Boc-protected intermediated

4.6 Results of affinity-based pull-down and proteomic analysis

S. Typhimurium cells were treated with both T315-biotin probes independently.

Cell lysates were subjected to affinity-based chromatography using streptavidin containing beads as the stationary phase. The flow-through was analyzed by SDS-PAGE (Figure 4.8) and a 20kDa band was observed in both T315-biotin probes lanes, yet absent in the no- drug control.

120

Bait Lysate flow-through flow-through Elution 1 Elution 2

g g g g

u u u u

r r r r

D D D D

o o o o

1 1 1 1

2 2 2 2

S S S S L

S S S S

N N N N

Figure 4.8 T315 pulldown flow-through analysis by SDS-PAGE

This band was then excised and, processed to elute and digest the protein, followed by tandem LC-MS/MS analysis. The proteomics analysis revealed the identity of the band as resulting from the flavin mononucleotide oxidoreductase WrbA. This finding was also confirmed by the diminished ability of S. Typhimurium to develop biofilms in wrbA mutant strain of S. Typhimurium 22.4% less than wildtype with no effect on cell viability.

In addition, wrbA mutant strain of S. Typhimurium was less responsive to the anti-biofilm effects of T315 compared to wildtype.

121

Figure 4.9 (A) S. Typhimurium wrbA biofilm formation. (B) Effect of T315 on WrbA.

Compared to wildtype, S. Typhimurium wrbA was shown to have a 27.3% reduced ability to form biofilms (Figure 4.9A). In addition, 20 M of T315 had a 30.3% reduction in activity in wrbA S. Typhimurium compared to wildtype highlighting the dependence on

WrbA for maximum inhibition of biofilm formation (Figure 9B). However, this same concentration of T315, 20 M, was still able to inhibit biofilm formation in the wrbA S.

Typhimurium strain at an average of 65.2% (Figure 4.9B). This data suggests that WrbA is likely one of the targets of T315, and activity is also mediated by at least one other protein or compensatory pathways.

122

4.7 WrbA in biofilm development: discussion

The 20.8 kDa protein WrbA [tryptophan (W) repressor binding protein] was originally named for its proposed ability to directly bind to the tryptophan repressor TrpR- operator DNA complex in E. coli.159 Although this direct binding is yet to be proven, E. coli WrbA has been shown to possess flavin mononucleotide (FMN)-dependent

NADH:quinone oxidoreductase activity.160 NCBI Protein BLAST analysis indicated a 94% amino sequence homology of E. coli WrbA to that of S. Typhimurium. With respect to activity, Kishko and coworkers proposed the mutually exclusive binding of both quinone and NADH exhibiting NAD(P)H:quinone oxidoreductase activity in a flavo- mononucleotide manner (FMN).161 WrbA has previously been implicated in P. aeruginosa response to oxidative stress.162,163 Furthermore, the natural product zosteric acid, was found to inhibit biofilm formation in E. coli in a mechanism mediated by WrbA through pull- down affinity-based proteomics.164

There is reasonable evidence in the literature to suggest that the oxidoreductase activity of WrbA is responsible for its role in biofilm formation. Accumulation of reactive oxygen species (ROS) in a cell can lead to DNA, lipid, and protein damage, ultimately causing irreparable damage to a cell’s essential macromolecules leading to death.165 The oxidoreductase activity of WrbA could prevent accumulation of ROS in a cell whose buildup could affect formation of healthy biofilms.166 An alternative mechanism could involve WrbA’s FMN cofactor that acts as a redox sensor affecting downstream processes that correct for oxidative stress.162,167,168 The exact mechanism by which WrbA facilitates

123

biofilm formation is yet to be fully elucidated. Future work in this area could highlight novel opportunities to exploit for new drug development targeting biofilm formation and disruption.

4.8 Conclusion

The data presented in this chapter demonstrate the potential for using small molecules to inhibit formation of bacterial biofilms that contribute to chronic carriage state of S. Typhi infections. JK-1 and T315 were identified from adenosine mimetic and kinase inhibitor compound libraries respectively and exhibit early stage biofilm inhibition activity. Employing T315 as the parent compound, affinity-based proteomics from two biotinylated mechanistic probes led to the identification of WrbA as one of the protein targets. Further mechanistic studies into the role of WrbA in biofilm formation are ongoing.

These findings are anticipated to highlight new areas of opportunity in the development of novel agents with the ability to inhibit biofilm formation. Since identification of WrbA as one of the putative targets of JK-1 in knockdown experiments, computational analysis of potential binding modes of JK-1 in either the FMN binding site or the NADH binding site is expected to enable rational analogue design to improve the potency of this class of compounds. Development of JK-1 derivatives into potent antibiofilm agents with optimal physicochemical properties has the potential to provide new weapons to combat the chronic carriage state of S. Typhi and many other infections that rely heavily on the biofilm lifestyle.

124

Chapter 5: Experimental Section

125

5.1 Synthesis and characterization of C-H functionalization substrates

All chemicals and reagents were purchased from Sigma-Aldrich, Alfa Aesar, Acros, TCI, or ChemImpex. Silicycle F60 (230-400 mesh) silica gel was used for column chromatography unless otherwise stated. Thin layer chromatography (TLC) analyses were performed using Merck silica gel 60 F254 plates and visualized under UV, KMNO4 or iodine stain. Melting points were determined using a Thermo Scientific Mel-Temp or a Thomas

Hoover Uni-melt capillary melting point apparatus. 1H, 19F, 13C NMR spectra were recorded using a Bruker AVIII 400 MHz, AVIII 600 MHz, or AVIII 700 MHz NMR spectrometer. 1H NMR and 13C NMR chemical shifts are reported in parts per million and

1 13 referenced with respect to CDCl3 ( H: residual CHCl3 at δ 7.26, C: CDCl3 triplet at δ

77.16). 1H NMR data are reported as chemical shifts (δ ppm), multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet), coupling constant (Hz), relative integral.

13C and 19F NMR data are reported as chemical shifts (δ ppm). High resolution mass spectra were obtained using Bruker MicrOTOF (ESI) or Thermo LTQ Orbitrap. IR spectra were recorded using a Thermo Fisher Nicolet iS10 FT-IR or Thermo Scientific Nicolet 6700 FT-

IR and are reported in terms of frequency of absorption (cm–1). Unless otherwise indicated, all hydrochloric acid solutions are in water.

126

General Procedure for Chlorination of Arenes (GP1)

To an 8 mL dram vial was added iodobenzene diacetate (0.6 mmol, 1.5 equiv), arene (0.4 mmol, 1 eq.), dichloroethane (2 mL), then 1 M hydrochloric acid (2 mL, 5 equiv). The solution was allowed to stir (1000 rpm) at 50 °C for the indicated amount of time, after which the solution was washed with saturated sodium bicarbonate, followed by saturated sodium thiosulfate, and concentrated. The crude mixture was then purified by column chromatography.

General Procedure for Chlorination of Heteroarenes (GP2)

To an 8 mL dram vial was added iodobenzene diacetate (0.6 mmol, 1.5 equiv), and heteroarene (0.4 mmol, 1 eq.), anhydrous dichloroethane (1 mL), then chloride source (5 equiv). The solution was allowed to stir (1000 rpm) at 50 °C for the indicated amount of time, after which the solution was washed with saturated sodium bicarbonate, followed by

127

saturated sodium thiosulfate, and concentrated. The crude mixture was then purified by column chromatography.

Substrate synthesis:

1,3-dimethyl-1,3-dihydro-2H-benzo[d]imidazol-2-one (S1). Prepared according literature.169 2-hydroxybenzimidazole (2.68 g, 20 mmol) was dissolved in DMF (15 mL), and potassium tert-butoxide (4.49 g, 40 mmol) was added and stirred for 30 minutes.

Iodomethane (2.5 mL, 40 mmol) was added slowly and stirred further for 30 minutes.

Additional potassium tert-butoxide (1.12 g, 20 mmol) was added along with iodomethane

(600 µL, 10 mmol) and heated to 60 °C for 1.5 hours. The mixture was cooled to room temperature, followed by the addition of water (10 mL), and extracted with ethyl acetate

(20 mL x 3). The organic layers were combined and washed with brine (50 mL) and dried over sodium sulfate, filtered, and concentrated. The product was purified by column chromatography eluting with 35% ethyl acetate/hexanes to yield S1 (2.17g, 67%) as a light

1 yellow solid. Rf: 0.15 (35% ethyl acetate/hexanes). H NMR (600 MHz, CDCl3): δ = 7.10

(dd, J = 5.7, 3.2 Hz, 2H), 6.97 (dd, J = 5.7, 3.2 Hz, 2H), 3.42 (s, 6H). 13C NMR (151 MHz,

CDCl3): δ = 154.83, 130.22, 121.33, 107.43, 27.27. Spectral data were consistent with literature.170

128

Substrate characterization:

N-(4-chloro-2-methylphenyl)pivalamide (2.67). Prepared according to GP1. Anilide (50 mg, 0.26 mmol) was reacted with iodobenzene diacetate (126.1 mg, 0.39 mmol) and 1 M

HCl (1.31 mL, 1.31 mmol) in dichloroethane (1 mL) for 4 hours. The crude product was purified by column chromatography eluting with 10% ethyl acetate/hexanes to yield 2.67

1 (51.8 mg, 88%) as a white solid. Rf: 0.65 (10% ethyl acetate/hexanes). Mp: 112 ºC. H

NMR (600 MHz, CDCl3): δ 7.77 – 7.72 (m, 1H), 7.21 (s, 1H), 7.17 – 7.12 (m, 2H), 2.20

13 (s, 3H), 1.32 (s, 9H). C NMR (151 MHz, CDCl3): δ 176.63, 134.64, 130.98, 130.22,

130.02, 126.81, 124.42, 39.84, 27.78, 17.59. IR (film) cm–1: 3314, 1646, 1505, 811. HRMS

+ (ESI) m/z: calc’d for C12H16ClNO [M+H] 226.0993, found 226.0991. Spectral data consistent with literature.171

5-chloro-2-methyl-1H-benzo[d]imidazole (2.76): To an 8 mL dram vial was added iodobenzene diacetate (0.6 mmol, 1.5 equiv), 2-methylbenzimidazole (50.0 mg, 0.38 mmol), dichloroethane (2 mL), then tetrabutylammonium chloride (526 mg, 1.89 mmol).

The solution was allowed to stir (1000 rpm) at 50 °C for 1.5 hours, after which the solution

129

was washed with saturated sodium bicarbonate, followed by saturated sodium thiosulfate, and concentrated. The crude mixture was then purified by column chromatography eluting with 5% methanol/dichloromethane to provide the 2.76 (45.6 mg, 72% yield) as a white

1 amorphous solid. Rf: 0.3 (5% methanol/dichloromethane). H NMR (400 MHz, CDCl3): δ

7.61 (s, 1H), 7.52 (m, 1H), 7.44 (d, J = 8.5 Hz, 1H), 7.20 (dd, J = 8.5, 2.0 Hz, 1H), 2.63 (s,

13 3H). C NMR (101 MHz, CDCl3): δ 152.17, 128.20, 123.11, 115.36, 114.65, 15.13. IR

–1 + (film) cm : 3315, 1647, 1506, 811. HRMS (ESI-TOF) m/z: calc’d for C8H7ClN2 [M+H] m/z 167.0371, found 167.0369. Spectral data were consistent with literature.172

5-chloro-1,3-dimethyl-1,3-dihydro-2H-benzo[d]imidazol-2-one (2.77). Prepared according to GP1. Benzoxindolidinone S1 (64.9 mg, 0.4 mmol) was reacted with iodobenzene diacetate (142 mg, 0.44 mmol) and 1 M HCl (2 mL, 2 mmol) in dichloroethane (2 mL) for 1 hour to give 81% of 2.77 (by crude 1H NMR). An analytical sample was purified by preparatory thin layer chromatography eluting with 2% methanol/dichloromethane to yield 2.77 as a white solid. Rf: 0.16 (2%

1 methanol/dichloromethane). Mp: 160.8 – 162.5°C. H NMR (600 MHz, CDCl3): δ = 7.07

(dd, J = 8.3, 2.0 Hz, 1H), 6.97 (m, 1H), 6.87 (d, J = 8.3 Hz, 1H), 3.41 (s, 3H), 3.40 (s, 3H).

13 C NMR (151 MHz, CDCl3): δ = 154.75, 131.05, 128.81, 126.97, 121.23, 108.08, 107.99,

27.42. IR (film) cm–1: 1703, 1655, 1446, 1394, 1342, 1228, 912, 744. HRMS (ESI) m/z:

130

+ calc’d for C9H9ClN2NaO [M+Na] 219.0301, found 219.0305. Spectral data were consistent with literature.173

4-chloroisoquinoline (2.82). Prepared according to GP2. Isoquinoline (51.7 mg, 0.4 mmol) was reacted with iodobenzene diacetate (193 mg, 0.6 mmol) and ethylchloroformate

(191 µL, 2 mmol) for 3 hours. The reaction mixture was purified by column chromatography eluting with 0.5% methanol/dichloromethane to yield 2.82 (60.5 mg, 92%

1 yield) as a clear oil. Rf: 0.05 (0.5% methanol/dichloromethane). H NMR (400 MHz,

CDCl3): δ = 9.13 (s, 1H), 8.56 (s, 1H), 8.17 (d, J = 8.5 Hz, 1H), 7.97 (d, J = 8.2 Hz, 1H),

13 7.80 (t, J = 7.6 Hz, 1H), 7.66 (t, J = 7.5 Hz, 1H). C NMR (101 MHz, CDCl3): δ = 151.21,

141.95, 133.64, 131.53, 129.51, 128.55, 128.28, 127.84, 123.41. IR (film) cm–1: 1572,

+ 1379, 1254, 979, 888, 794. HRMS (ESI) m/z: calc’d for C9H6ClN [M+H] 164.0262, found

164.0260. Spectral data were consistent with literature.174

131

position 1H 13C HMBC correlations on 2.82 1 9.12 151.22 3, 4, 4a, 5, 8, 8a 3 8.56 141.96 1, 4, 4a, 5, 6, 8a, 4 128.56 4a 133.65 5 8.17 123.42 4, 4a, 7, 8a 6 7.80 131.54 4, 4a, 5, 7, 8, 8a 7 7.66 128.29 4a, 5, 6, 8, 8a 8 7.97 127.85 1, 4, 4a, 6, 7, 8a 8a 129.52

4-bromoisoquinoline (2.86): Prepared according to GP2. Isoquinoline (51.7 mg, 0.4 mmol) was reacted with iodobenzene diacetate (193 mg, 0.6 mmol) and dried fine KBr powder (238 mg, 2 mmol). After 14 hours, the reaction mixture was purified by column chromatography eluting with 0.5% methanol/dichloromethane to yield 2.86 (58.3 mg, 70%

1 yield) as a brown oil. Rf: 0.2 (0.25% methanol/dichloromethane). H NMR (600 MHz,

CDCl3): δ 9.18 (s, 1H), 8.72 (s, 1H), 8.17 (d, J = 8.5 Hz, 1H), 7.99 (d, J = 8.2 Hz, 1H), 7.84

13 (ddd, J = 8.3, 6.9, 1.2 Hz, 1H), 7.65 – 7.74 (m, 1H). C NMR (151 MHz, CDCl3): δ 151.66,

144.43, 135.06, 132.04, 129.86, 128.13, 128.52, 126.12, 119.91. IR (film) cm–1: 1375,

+ 1215, 958, 772. HRMS (ESI) m/z: calc’d for C9H6BrN [M+H] 207.9756, found 207.9747.

Spectral data were consistent with literature.175

132

1-(3-chloro-1H-indol-1-yl)-2,2-dimethylpropan-1-one (2.83). Prepared according to

GP1. 1-(1H-indol-1-yl)-2,2-dimethylpropan-1-one (80.5 mg, 0.4 mmol) was reacted with iodobenzene diacetate (193 mg, 0.6 mmol) and 1M HCl (2 mL, 2.0 mmol) for 55 minutes.

The reaction mixture was purified by column chromatography with 5% ethyl acetate/hexanes to yield 2.83 (66.7 mg, 71%) as a clear oil. Rf: 0.65 (5% ethyl

1 acetate/hexanes). H NMR (600 MHz, CDCl3): δ = 8.52 (d, J = 8.4 Hz, 1H), 7.73 (s, 1H),

7.59 (m, 1H), 7.44 – 7.39 (m, 1H), 7.38 – 7.33 (m, 1H), 1.52 (s, 9H); 13C NMR (151 MHz,

CDCl3) δ 176.59, 136.11, 127.23, 126.55, 124.20, 122.07, 118.25, 117.64, 113.26, 41.41,

28.79; IR (thin film): 752.18, 1171.42, 1307.85, 1446.54, 1700.17 cm-1. HRMS (ESI) m/z:

+ calc’d for C13H15ClNO [M+H] 236.0842, found 236.0841. Spectral data were consistent with literature.176

4-chloro-3,5-dimethyl-1H-pyrazole (2.85). Prepared according to GP2. 3,5-

Dimethylpyrazole (50.0 mg, 0.52 mmol) was reacted with tetrabutylammonium chloride

(723 mg, 2.6 mmol), and iodobenzene diacetate (503 mg, 1.56 mmol) for 2 hours. The

133

crude product was purified by column chromatography eluting with 50% ethyl acetate/hexanes to yield 2.83 (46.5 mg, 69%) as a white solid. Rf: 0.9 (2%

1 methanol/dichloromethane) Mp: 88.2 – 90.0 °C. H NMR (600 MHz, CDCl3): δ = 9.45 (s,

13 1H), 2.22 – 2.67 (s, 6H). C NMR (151 MHz, CDCl3): δ = 141.17, 108.07, 10.47. IR (film) cm–1: 3201, 3122, 3059, 1654, 1597, 1479, 1122, 1041, 912, 829, 742. HRMS (ESI) m/z:

+ calc’d for C5H8ClN2 [M+H] 131.0376, found 131.0376. Spectral data were consistent with literature.177

3-chloroquinoline (2.89): Prepared according to GP2. Quinoline (47 µL, 0.4 mmol) was reacted with iodobenzene diacetate (193 mg, 0.6 mmol) and pentafluorobenzoyl chloride

(288 µL, 2.0 mmol) in anhydrous dichloroethane (0.5 ml, 0.8 M) for 17 hours. The reaction mixture was quenched with saturated sodium bicarbonate and extracted using dichloromethane. The organic layer was further washed with 1M sodium hydroxide followed by saturated sodium thiosulfate, then concentrated. The crude mixture was purified by column chromatography eluting with 100% dichloromethane to yield 2.89 (41.2

1 mg, 63% yield) as a clear oil. Rf: 0.3 (100% dichloromethane). H NMR (600 MHz,

CDCl3): δ = 8.83 (m, 1H), 8.14 (m, 1H), 8.10 (dd, J = 8.5, 0.4 Hz, 1H), 7.76 (d, J = 8.2 Hz,

1H), 7.72 (ddd, J = 8.4, 6.9, 1.4 Hz, 1H), 7.58 (ddd, J = 8.1, 6.9, 1.1 Hz, 1H). 13C NMR

(151 MHz, CDCl3): δ = 149.83, 146.90, 146.43, 134.11, 129.78, 129.64, 128.59, 127.86,

134

–1 127.16. IR (film) cm : 2919, 2850, 2359, 953, 751. HRMS (ESI) m/z: calc’d for C9H7ClN

[M+H]+ 164.0267, found 164.0265.

8-chloro-1,3,7-trimethyl-3,7-dihydro-1H-purine-2,6-dione (2.96). Prepared according to GP2. Caffeine (77.7 mg, 0.4 mmol) was reacted with iodobenzene diacetate (193 mg,

0.6 mmol) and pentafluorobenzoyl chloride (288 µL, 2.0 mmol) for 4 hours. The reaction mixture was quenched with saturated sodium bicarbonate and extracted using dichloromethane. The organic layer was further washed with 1M sodium hydroxide followed by saturated sodium thiosulfate, then concentrated. The reaction mixture was purified by column chromatography eluting with 1% methanol/dichloromethane to yield

2.96 (60.1 g, 65%) as a white solid. Rf: 0.5 (3% methanol/dichloromethane). Mp: 188 ºC.

1 13 H NMR (400 MHz, CDCl3): δ = 3.95 (s, 3H), 3.55 (s, 3H), 3.40 (s, 3H). C NMR (101

MHz, CDCl3): δ = 154.71, 151.42, 147.25, 139.07, 108.40, 32.79, 29.92, 28.09. IR (film)

–1 + cm : 1707, 1664, 1369, 755. HRMS (ESI) m/z: calc’d for C8H9ClN4O2 [M+H] m/z

229.0487, found 229.0474. Spectral data were consistent with literature.178

135

8-bromo-1,3,7-trimethyl-3,7-dihydro-1H-purine-2,6-dione (S2). Prepared according to

GP2. Caffeine (77.7 mg, 0.4 mmol) was reacted with iodobenzene diacetate (193 mg, 0.6 mmol) and lithium bromide (173.7 mg, 2.0 mmol) for 3 hours. The reaction mixture was quenched with saturated sodium bicarbonate and extracted using dichloromethane. The organic layer was further washed with saturated sodium thiosulfate, then concentrated. The reaction mixture was purified by column chromatography eluting with 1% methanol/dichloromethane to yield S2 (31.1 mg, 29%) as a white powder. Rf: 0.2 (1%

1 methanol/dichloromethane). Mp: 205.1 – 207.0 ºC. H NMR (400 MHz, CDCl3): δ = 3.95

13 (s, 1H), 3.54 (s, 1H), 3.39 (s, 1H). C NMR (101 MHz, CDCl3): δ = 154.60, 151.43,

148.19, 128.27, 109.53, 34.10, 29.97, 28.15. IR (film) cm–1: 1707, 1664, 1454, 1457, 1353,

+ 743. HRMS (ESI-TOF) m/z: calc’d for C8H10BrN4O2 [M+H] m/z 272.9987, found

272.9984. Spectral data were consistent with literature.178

5-chloro-1,3-dimethylpyrimidine-2,4(1H,3H)-dione (2.99). Prepared according to GP1.

1,3-dimethyluracil (56.1 mg, 0.4 mmol) was reacted with iodobenzene diacetate (193 mg,

0.6 mmol) and 2M HCl in diethyl ether (1.0 ml, 2.0 mmol) for 40 minutes. The reaction 136

mixture was purified by chromatography eluting with 1% methanol/dichloromethane to yield 2.99 (59.8 mg, 86%) as a white amorphous solid. Rf: 0.3 (2%

1 methanol/dichloromethane). Mp: 142 ºC. H NMR (400 MHz, CDCl3): δ = 7.42 (s, 1H),

13 3.41 (s, 3H), 3.38 (s, 3H). C NMR (101 MHz, CDCl3): δ = 159.64, 150.97, 140.03,

108.05, 37.37, 29.09. IR (film) cm–1: 1717, 1661, 1445, 1340, 757. HRMS (ESI) m/z: calc’d

+ for C6H7ClN2O2Na [M+Na] 197.0088, found 197.0077. Spectral data were consistent with literature.179

1-(2-chloro-4,5-dimethoxybenzyl)-6,7-dimethoxyisoquinoline (2.101): Prepared according to GP1 using HCl (67.9 mg, 2 mmol). After 4 hours, the reaction mixture was purified by column chromatography eluting with 1% methanol/dichloromethane to yield

2.101 (70.4 g, 94% yield) as an off-white foamy solid. Rf: 0.5 (1%

1 methanol/dichloromethane). H NMR (400 MHz, CDCl3): δ 8.36 (d, J = 5.7 Hz, 1H), 7.43

(d, J = 5.7 Hz, 1H), 7.36 (s, 1H), 7.03 (s, 1H), 6.87 (s, 1H), 6.69 (s, 1H), 4.63 (s, 2H), 3.99

13 (s, 3H), 3.95 (s, 3H), 3.82 (s, 3H), 3.61 (s, 3H). C NMR (101 MHz, CDCl3): δ 157.47,

152.78, 150.24, 148.31, 148.19, 140.71, 133.52, 128.99, 123.88, 123.04, 119.02, 112.95,

112.16, 105.34, 104.21, 56.30, 56.20, 56.13, 56.02, 38.53. IR (film) cm–1: 2360, 1508,

137

+ 1272, 1235, 1160, 858. HRMS (ESI) m/z: calc’d for C20H20ClNO4 [M+H] 374.1154, found 374.1129. Spectral data were consistent with literature.180

1-(2-bromo-4,5-dimethoxybenzyl)-6,7-dimethoxyisoquinoline (2.102): Prepared according to GP1 using 48.8% aqueous HBr (111 µL, 2 mmol). After 2.5 hours, the reaction mixture was purified by column chromatography eluting with 1.5% methanol/dichloromethane to yield 2.102 (82.8 g, 99%) as a foamy brown solid. Rf: 0.5

1 (100% dichloromethane). H NMR (600 MHz, CDCl3): δ 8.37 (d, J = 5.6 Hz, 1H), 7.43 (d,

J = 5.6 Hz, 1H), 7.33 (s, 1H), 7.04 (s, 2H), 6.66 (s, 1H), 4.64 (s, 2H), 3.99 (s, 3H), 3.97 (s,

13 3H), 3.83 (s, 3H), 3.59 (s, 3H). C NMR (151 MHz, CDCl3): δ 157.64, 152.78, 150.29,

148.83, 148.47, 141.01, 133.53, 131.16, 123.17, 118.98, 115.32, 113.82, 113.17, 105.40,

104.45, 56.51, 56.29, 56.13, 55.97, 41.58. IR (film) cm–1: 1508, 1235, 1159, 1030, 857,

+ 731. HRMS (ESI) m/z: calc’d for C20H20BrNO4 [M+H] 418.0648, found 418.0627.

Spectral data were consistent with literature.181

138

4-chloro-2-iodo-1-methoxybenzene (2.103). Prepared according to GP1. 2-iodoanisole

(52 µL, 0.4 mmol) was reacted with iodobenzene diacetate (193 mg, 1.5 eq) and 1 M HCl

(2 mL, 5 equiv) in dichloroethane (2 mL) for 5 hours. Crude product was purified by column chromatography eluting with hexanes to yield 2.103 (79.1 mg, 74%) as colorless

1 oil. Rf: 0.32 (100% hexanes). H NMR (400 MHz, CDCl3): δ = 7.75 (d, J = 2.5 Hz, 1H),

7.28 (dd, J = 8.8, 2.6 Hz, 1H), 6.73 (d, J = 8.8 Hz, 1H), 3.86 (s, 3H). 13C NMR (101 MHz,

CDCl3): δ = 138.78, 129.38, 126.52, 111.53, 56.79. Spectral data consistent with literature.182

6-Chloro-2”-acetyl phyllanthusmin D (2.106). Synthesis of 2.106 began from a known natural product derivative 2”-acetyl phyllanthusmin D52 prepared according to literature procedures.71 To an 8 mL dram vial was added iodobenzene diacetate (96.6 mg, 0.3 mmol), 139

2”-acetyl phyllanthusmin D (65.6 mg, 0.1 mmol), dichloroethane (1 mL), and tetrabutylammonium chloride (139.0 mg, 0.5 mmol). The solution was allowed to stir at

1000 rpm at room temperature for 48 hours. After which the solution was washed with saturated sodium bicarbonate, followed by saturated sodium thiosulfate and concentrated.

The crude mixture was purified by column chromatography eluting with 10% acetonitrile in toluene, followed by 20% acetone in hexanes to yield 2.106 (12.4 mg, 18%) as a white

1 solid. Rf: 0.3 (10% acetonitrile/90% toluene). H NMR (700 MHz, CDCl3): δ = 7.13 (s,

1H), 6.96 (m, 1H), 6.81 – 6.74 (m, 2H), 6.11 – 6.05 (m, 2H), 5.62 (m, 1H), 5.53 (m, 1H),

5.45 (m, 1H), 5.30 (m, 1H), 5.21 (dd, J = 7.3, 2.3 Hz, 1H), 5.16 (dd, J = 9.8, 3.4 Hz, 1H),

4.00 – 3.98 (m, 1H), 3.98 (s, 3H), 3.77 (m, 3H), 3.54 (m, 1H), 2.22 (s, 3H), 2.14 (s, 3H),

2.07 (s, 3H). Doubling and splitting of specific peaks has been previously and independently reported in structurally similar compounds by the Charlton71 and Kinghorn52 groups. This effect is attributed to the hindered rotation about the C1’-C7’ bond. In the characterization data below major peaks are listed (with all signals observed in

13 parentheses). C NMR (176 MHz, CDCl3): δ =170.42, 170.27, 169.7 (169.64, 169.65),

169.26, 152.45, 149.26, 147.92, 147.84, 143.6 (143.56, 143.57), 137.14, 134.7 (134.61,

134.68), 134.11, 128.10, 124.07, 123.79, 123.56, 122.00, 121.95, 121.38, 110.8 (110.75,

110.76), 110.6 (110.56, 110.58), 108.51, 107.3 (107.24, 107.25, 107.26), 102.2 (102.14,

102.15), 101.50, 70.35, 69.69, 68.1 (68.12, 68.14), 67.82, 64.53, 60.91, 55.9 (55.93, 55.95),

31.73, 29.85, 21.1 (21.08, 21.09, 21.11, 21.13), 20.8 (20.81, 20.82). IR (film) cm–1: 2921,

140

2850, 1749, 1488, 1457, 1417, 1224, 1068, 1037. HRMS (ESI) m/z: calc’d for

+ C32H29ClNaO14 [M+Na] 695.11380 found 695.11267.

6’-Bromo-2”-acetyl phyllanthusmin D (2.107). Synthesis of 2.107 began from a known natural product derivative 2”-acetyl phyllanthusmin D52 prepared according to literature procedures.71 To an 8 mL dram vial was added iodobenzene diacetate (24.2 mg, 0.075 mmol), 2”-acetyl phyllanthusmin D (34.5 mg, 0.05 mmol), dichloroethane (1 mL), and lithium bromide (23.5 mg, 0.27 mmol). The solution was allowed to stir at 1000 rpm at room temperature for 1 hour. After which the solution was washed with saturated sodium bicarbonate, followed by saturated sodium thiosulfate and concentrated. The crude mixture was purified by column chromatography eluting with 50% acetone in hexanes to yield

1 2.107 (35.6 mg, 99%) as an orange solid. Rf: 0.5 (50% acetone/50% hexanes). H NMR

(400 MHz, CDCl3): δ = 7.56 (m, 1H), 7.20 (m, 1H), 6.83 (s, 1H), 6.71 (d, J = 9.8 Hz, 1H),

6.10 (m, 2H), 5.70 (m, 1H), 5.47 (m, 3H), 5.39 (m, 1H), 5.19 (dd, J = 9.5, 3.5 Hz, 1H),

141

5.13 (m, 1H), 4.21 (m, 1H), 3.82 (m, 3H), 4.09 (s, 3H), 3.75 (m, 1H), 2.11 (s, 3H), 2.22 (s,

3H), 2.09 (s, 3H). Doubling and splitting of specific peaks has been previously and independently reported in structurally similar compounds by the Charlton71 and Kinghorn52 groups. This effect is attributed to the hindered rotation about the C1’-C7’ bond. In the characterization data below major peaks are listed (with all signals observed in

13 parentheses). C NMR (101 MHz, CDCl3): δ = 170.3 (170.33, 170.34), 170.2 (170.19,

170.23), 169.68, 169.60, 169.2 (169.17, 169.20), 152.3 (152.23, 152.27), 151.0 (150.95,

150.96), 148.8 (148.75, 148.76), 147.6 (147.58, 147.62), 144.8 (144.78, 144.80), 134.5

(134.47, 134.57), 134.5 (134.47, 134.57), 130.3 (130.24, 130.32), 128.9 (128.90, 128.91),

127.09, 126.34, 126.2 (126.23, 126.25), 120.2 (120.17, 120.18), 114.8 (114.78, 114.80),

113.0 (112.94, 113.01), 111.1 (111.05, 111.16), 105.6 (105.63, 105.69), 102.23, 101.0

(100.83, 100.99, 101.21, 101.40), 77.36, 70.2 (70.17, 70.21), 69.5 (69.49, 69.54), 67.3

(67.31, 67.33, 67.36, 67.37), 64.0 (63.98, 64.10), 56.4 (56.40, 56.42), 56.09, 34.82, 34.67,

31.73, 29.21, 27.07, 25.43, 22.80, 21.1 (21.08, 21.10, 21.13), 20.8 (20.82, 20.83, 20.84).

IR (film) cm–1: 2922, 1749, 1506, 1488, 1475, 1433, 1217, 1031, 669. HRMS (ESI) m/z:

+ calc’d for C32H29ClNaO14 [M+Na] 717.08134 found 717.08371.

142

Supporting information:

I. Anion Source Investigation for electron-rich arenes

Effect of chloride sources on functionalization of model arene Chloride % Recovered Starting % Yield Source Material HCl 88% 0%

ZnCl2 85% 0% MgCl 73% 0% 2 LiCl 68% 0% CuCl2 60% 0% AcCl 56% 0% KCl 0% <52% Me3SiCl 0% <55% NaCl 0% 63% NH4Cl 0% 73% Bu4NCl 0% 84% FeCl Trace 89% 2

All reactions were carried out according to GP1 using N-(o-tolyl)pivalamide 2.62 as the test substrate (50 mg, 0.26 mmol), and the respective chloride sources (1.3 mmol) indicated in the table for 3 hours. The crude mixtures were then purified by column chromatography eluting with 10% ethyl acetate/hexanes to provide 2.67.

143

II. Anion Source Investigation for Heteroarenes

Chloride % Recovered % Yield Source Starting Material

MgCl 0% 91% 2 Bu4NCl 0% 65% NaCl 0% 64%

ZnCl 0% 45% 2 Me3SiCl 28% 35% C6F5COCl 4% 0% AcCl 86% 0% EtOCOCl 92% 0%

Effect of chloride sources on functionalization of model heteroarene

All reactions were carried out according to GP2 using the respective chloride sources (2 mmol, 5 equiv.) indicated in the table for 3 hours. The crude mixtures were then purified by column chromatography eluting with 0.5% methanol in dichloromethane to provide

2.82.

144

III. Effect of Various Acyl Chlorides on Heteroarene Chlorination

Chloride Source

C6F5COCl 65% 55% 52% 4% 63% AcCl 63% 54% 57% 86% 0% EtOCOCl 5% 0% 36% 92% 24%

All reactions were carried out according to GP2 using the respective chloride sources (2 mmol, 5 equiv.) indicated in the table. Reaction progress was monitored by TLC and mass spectrometry. The crude mixtures were then purified by column chromatography. Isolated yields. *Reaction times and purification conditions can be found under corresponding experimental data section for each substrate.*

IV. Effect of Various Acyl Chlorides on Arene Chlorination

Chloride % Recovered % Yield Source Starting Material

C6F5COCl 41% 0% AcCl 56% 0% EtOCOCl 0% 75%

145

To an 8 mL dram vial was added iodobenzene diacetate (96.6 mg, 0.3 mmol), 2-iodoanisole

(46.8 mg, 0.2 mmol), dichloroethane (1 mL), and respective chloride sources (1 mmol).

The solution was allowed to stir (1000 rpm) at 50 °C for 30 minutes. After which the solution was washed with saturated sodium bicarbonate, followed by saturated sodium thiosulfate and concentrated. Isopropyl acetate (0.2 mmol) was then added to the crude

1 product followed by CDCl3. Yields are based on H NMR using isopropyl acetate as a standard.

V. Competitive Chlorination Experiments

% Recovered Target % Yield Starting Material 4-chloroisoquinoline, 2.82 63% 0% 4-chloroiodoanisole, 2.103 0% 93%

To an 8 mL dram vial was added iodobenzene diacetate (96.6 mg, 0.3 mmol), isoquinoline

(25.8 mg, 0.2 mmol), 2-iodoanisole (46.8 mg, 0.2 mmol), dichloroethane (1 mL), and ethyl chloroformate (109 mg, 1 mmol). The solution was allowed to stir (1000 rpm) at 50 °C for

2 hours. After which the solution was washed with saturated sodium bicarbonate, followed 146

by saturated sodium thiosulfate and concentrated. Isopropyl acetate (0.2 mmol) was then

1 added to the crude product followed by CDCl3. Yields are based on H NMR using isopropyl acetate as an internal standard

% Recovered Target % Yield Starting Material 4-chloroisoquinoline, 2.82 0% 97% 4-chloroiodoanisole, 2.103 65% 10%

To an 8 mL dram vial was added iodobenzene diacetate (96.6 mg, 0.3 mmol), isoquinoline

(25.8 mg, 0.2 mmol), 2-iodoanisole (46.8 mg, 0.2 mmol), dichloroethane (1 mL), and 1 M hydrochloric acid (1 mL, 1 mmol). The solution was allowed to stir at 1000 rpm at 50 °C for 2 hours. After which the solution was washed with saturated sodium bicarbonate, followed by saturated sodium thiosulfate and concentrated. Isopropyl acetate (0.2 mmol)

1 was then added to the crude product followed by CDCl3. Yields are based on H NMR using isopropyl acetate as an internal standard.

147

VI. Comparison of Chlorinating Reagents on Isoquinoline and Quinoline

Effect of various chlorinating reagents on functionalization of isoquinoline

% Yield of % Recovered Chlorinating Reagent Conditions 2.82 Starting Material tBuOCl A 20% 24% NCS B 0% 100% Palau’chlor C 0% 97 Palau’chlor D 0% 91 IBA-Cl E 0% 60% IBA-Cl F 0% 41%

RCl, PhI(OAc)2 GP2 92% 0%

148

Effect of various chlorinating reagents on functionalization of quinoline

Chlorinating % Yield of % Recovered Conditions Reagent 2.89 Starting Material tBuOCl G 0% 74% NCS H 0% 81% Palau’chlor C 0% 78% Palau’chlor D 0% 86% IBA-Cl E 0% 92% IBA-Cl F 0% 89%

RCl, PhI(OAc)2 GP2 63% 0%

Conditions A (tBuOCl). To a solution of isoquinoline (0.1 mmol), in acetonitrile (1 ml) was added tert-butyl hypochlorite (12.6 mg, 0.1 mmol). The solution was left to stir at room temperature for 16 hours. 1H NMR analysis of reaction showed 20% conversion to

4-chloroisoquinoline, 24% starting material remaining, and formation of multiple unidentifiable product.

Conditions B (NCS). To a solution of isoquinoline (0.1 mmol), in acetonitrile (1 ml) was added N-chlorosuccinimide (NCS) (18.6 mg, 0.14 mmol). The solution was left to stir at

149

room temperature for 4 hours. 1H NMR analysis of reaction showed complete retention of starting material using isopropyl acetate as an internal standard.

Conditions C (Palau’chlor). To a solution of quinoline or isoquinoline (0.1 mmol), in chloroform (1 ml) was added chlorobis(methoxycarbonyl)guanidine (CBMG or

Palau’chlor) (31.4 mg, 0.15 mmol). The solution was left to stir at room temperature for

12 hours. After which the reactions were concentrated and analyzed by 1H NMR spectroscopy using isopropyl acetate as an internal standard.

Conditions D180 (Palau’chlor). To a solution of quinoline or isoquinoline (0.1 mmol), in acetonitrile (1 ml) was added chlorobis(methoxycarbonyl)guanidine (CBMG or

Palau’chlor) (31.4 mg, 0.15 mmol). The solution was left to stir at room temperature for

12 hours. After which the reactions were concentrated and analyzed by 1H NMR spectroscopy using isopropyl acetate as an internal standard.

Conditions E (IBA-Cl). To a solution of (iso)quinoline (0.3 mmol) in dichloroethane (1 mL) was added 1- chloro-1,2-benziodoxol-3-one (102 mg, 0.36 mmol) and stirred at 50 °C for 12 hours (quinoline) or 24 hours (isoquinoline). Yield determined by crude 1H NMR spectroscopy using mesitylene as an internal standard.

Conditions F176 (IBA-Cl). To a solution of (iso)quinoline (0.3 mmol) in DMF (1 mL) was added 1-chloro-1,2-benziodoxol-3-one (102 mg, 0.36 mmol) and stirred at 23 °C 12

150

hours (quinoline) or 24 hours (isoquinoline). Yield determined by crude 1H NMR spectroscopy using mesitylene as an internal standard.

Conditions G (tBuOCl). To a solution of quinoline (47 µL, 0.4 mmol) in dichloroethane

(1 mL) was added tBuOCl (68 µL, 0.4 mmol) and stirred at 50 °C for 4 hours. Yield determined by crude 1H NMR spectroscopy using mesitylene as an internal standard.

Conditions H (NCS). To a solution of quinoline (47 µL, 0.4 mmol) in dichloroethane (1 mL) was added N-chlorosuccinimide (80 mg, 0.6 mmol) and stirred at 50 °C for 4 hours.

Yield determined by crude 1H NMR spectroscopy using mesitylene as an internal standard.

151

VII. NMR Analysis of Active Oxidant

Chlorinating reagent PhICl(X), S4 PhI(OAc)2, 2.56 PhI, S5

HCl (aq) 100 0 55

HCl (org) 98.5 0 8

AcCl 62 32 0

EtOCOCl 0 91 0

C6F5COCl Broad peaks - 9

To an 8 mL dram vial was added iodobenzene diacetate (25 mg, 0.08 mmol), deuterated chloroform (1 mL), and the respective chloride sources (0.39 mmol). The solution was allowed to stir at 1000 rpm at 50 °C for 30 minutes. Upon cooling to room, the reactions were analyzed by 1H NMR using isopropyl acetate as a standard. Interestingly, ethyl chloroformate did not form any observable amount of the active oxidant.

152

VIII. Investigation of Isoquinoline Pre-activation with Acyl Chloride

% Yield of % Recovered Conditions 2.82 Starting Material

2.80 + EtO2CCl 10 min. prestir 72% 0%

No prestir 89% 0%

4-chloroisoquinoline (6). Prepared according to GP3. Isoquinoline (100.0 mg, 0.8 mmol) was reacted with iodobenzene diacetate (374.0 mg, 1.16 mmol) and ethylchloroformate

(370.0 µL, 3.87 mmol) in dichloroethane (1 mL) for 4 hours. The reaction mixture was purified by column chromatography eluting with 0.5% methanol/dichloromethane to yield

2.82 as a clear oil. Rf: 0.05 (0.5% methanol/dichloromethane).

IX. Investigation of Heteroarene Chlorination via Iodonium Intermediate

To investigate the plausibility of product formation (functionalized heteroarene) via an iodonium intermediate the following investigations were carried out.

153

Synthesis of diaryliodonium salt

4-methoxyphenyl(3-pyridyl)iodonium triflate (2.114). To a solution of 3-iodopyridine

(100 mg, 0.5 mmol) in dichloromethane (3.2 mL) was added TfOH (294.2 mg, 1.96 mmol) and the mixture was stirred at room temperature for 5 minutes. mCPBA (148 mg, 0.86 mmol) was added and the reaction was heated at 60 °C in a sealed tube. After stirring for

30 minutes the reaction was cooled down to 0 °C, followed by the addition of water (18 mg, 1 mmol). A solution of anisole (106 mg, 0.98 mmol) in dichloromethane (1.3 mL) was added dropwise and the reaction mixture was stirred at 0 °C for 15 minutes then concentrated. Diethyl ether (1 mL) was added and the resulting mixture was stirred at 0 °C for 30 minutes. A precipitate formed which was filtered and washed with diethyl ether (2 mL x 2) to afford 4-methoxyphenyl(3-pyridinium)iodonium bistriflate upon drying as an off-white solid. 1H NMR (400 MHz, MeOD) δ 9.25 (m, 1H), 8.86 (m, 1H), 8.66 (m, 1H),

8.19 – 8.13 (m, 2H), 7.65 (m, 1H), 7.12 – 7.06 (m, 2H), 3.86 (s, 3H).

The resulting bistriflate salt was dissolved in dichloromethane:methanol (0.5 mL, 5:1) and loaded onto a column packed with Al2O3 (500 mg). The column was eluted with a solution of dichloromethane:methanol (15 mL, 20:1), and combined fractions were concentrated to afford 2.114 as an oily residue (172 mg, 76% yield). 1H NMR (400 MHz, MeOD) δ 9.19

154

(m, 1H), 8.82 (m, 1H), 8.56 (m, 1H), 8.17 – 8.12 (m, 2H), 7.56 (m, 1H), 7.11 – 7.06 (m,

2H), 3.86 (s, 3H). Spectral data consistent with literature.183

Functionalization of diaryliodonium salt

To a solution of 2.114 (17.4 mg, 0.038 mmol) in dichloroethane (0.075 M) was added acetyl chloride (14.9 mg, 0.19 mmol). The solution was stirred at 50 °C for 3 hours and then concentrated. Mass spectrometric analysis of crude mixture did not indicate formation of 3-chloropyridine (2.115) or 4-chloroanisole (2.116), but rather oxysulfonylated anisole.

To a solution of 2.114 26.3 mg, 0.057 mmol) in dichloroethane (0.075 M) was added iodobenzene diacetate (27.5 mg, 0.09 mmol) then acetyl chloride (22.4 mg, 0.29 mmol).

The solution was stirred at 50 °C for 3 hours and then concentrated. Mass spectrometric analysis of crude mixture did not indicate formation of 3-chloropyridine (2.115) or 4-

155

chloroanisole (2.116), but rather oxysulfonylated anisole, iodobenzene (S5) and unreacted

2.114.

X. Investigation of Heteroarene Chlorination via N-donor Polycationic Hypervalent Iodine Intermediate

Reagent %Yield of % Recovered Comment 2.82 Iodonium 2.117 AcCl 0% 0% 4% isoquinoline recovered EtO2Cl 0% 0% decomposition PhICl2 3% 0% 78% isoquinoline recovered AcCl, PhI(OAc)2 1% 0% 100% isoquinoline recovered

1,1'-(phenyl-l3-iodanediyl)bis(isoquinolin-2-ium) trifluoromethanesulfonate (2.117) was prepared according to literature procedure81 and immediately subjected to chlorination conditions (below).

Conditions A (AcCl). To a solution of isoquinolinium hypervalent iodine reagent (0.041 mmol), in anhydrous dichloromethane (0.5 ml) was added acetyl chloride (32.2 mg, 0.41 mmol). The solution was left to stir at 50 °C for 3 hours. After cooling to room temperature, dichloromethane (1 mL) was added and the reaction was extracted with saturated sodium bicarbonate (1 ml x 3), followed by a 10% sodium thiosulfate solution (2 mL). The organic

156

layer was concentrated and analyzed by 1H NMR analysis using isopropyl acetate as an internal standard.

Conditions B (EtO2Cl). To a solution of isoquinolinium hypervalent iodine reagent (0.041 mmol), in anhydrous dichloromethane (0.5 ml) was added ethyl chloroformate (44.5 mg,

0.41 mmol). The solution was left to stir at 50 °C for 3 hours. After cooling to room temperature, dichloromethane (1 mL) was added and the reaction was extracted with saturated sodium bicarbonate (1 mL x 3), followed by a 10% sodium thiosulfate solution

(2 mL). The organic layer was concentrated and analyzed by 1H NMR analysis using isopropyl acetate as an internal standard.

Conditions C (PhICl2). To a solution of isoquinolinium hypervalent iodine reagent (0.041 mmol), in anhydrous dichloromethane (0.5 ml) was added iodobenzene dichloride (56.4 mg, 0.41 mmol). The solution was left to stir at 50 °C for 3 hours. After cooling to room temperature, dichloromethane (1 mL) was added and the reaction was extracted with saturated sodium bicarbonate (1 mL x 3), followed by a 10% sodium thiosulfate solution

(2 mL). The organic layer was concentrated and analyzed by 1H NMR analysis using isopropyl acetate as an internal standard.

Conditions D (AcCl, PhI(OAc)2). To a solution of isoquinolinium hypervalent iodine reagent (0.041 mmol), in anhydrous dichloromethane (0.5 mL) was added acetyl chloride

(32.2 mg, 0.41 mmol) followed by iodobenzene diacetate (13.2 mg, 0.08 mmol). The solution was left to stir at 50 °C for 3 hours. After cooling to room temperature, dichloromethane (1 mL) was added and the reaction was extracted with saturated sodium

157

bicarbonate (1 mL x 3), followed by a 10% sodium thiosulfate solution (2 mL). The organic layer was concentrated and analyzed by 1H NMR analysis using isopropyl acetate as an internal standard.

XI. Redox Potential Measurements by Cyclic Voltammetry

A Biologic® SP-50 potentiostat was used for all electrochemical experiments. A three- electrode system was employed consisting of a Ag/AgNO3 reference electrode (BASi), a

Pt wire counter electrode (BASi), and the glassy carbon working electrode (BASi). The sweep rate for CV was 100 mV/s. All CV experiments were performed in acetonitrile using

Bu4NPF6 as supporting electrolyte (0.1 M) under argon atmosphere. Ferrocene used as reference electrode. Values reported for peaks.

Potential (V) Potential (V) Substrate vs Ag/Ag+ vs SCE (+0.298 V)

Isoquinoline, Ep,a 1.65 1.95

Quinoline, Ep,a 1.72 2.02

PhI(OAc)2, Ep,c -1.46 -1.16

PhICl2, Ep,c -0.30 -0.01

Ac2O, PhICl2, Ep,c -0.34 -0.03

Table 1. Voltammetric Peak Potentials (Ep, V)

158

5.2 Synthesis and characterization of ciprofloxacin-fatty acid conjugates

All commercial reagents, anhydrous solvents, and reagent-grade solvents were purchased from commercial suppliers as the highest available purity and used without further purification. Compound 3.33 was prepared by Xiangmin Liao. Unless otherwise indicated, all reactions were conducted under an inert atmosphere at ambient temperature. All temperatures are expressed in ºC (degrees Centigrade). 1H NMR (hereinafter also "NMR") spectra were recorded on Brucker AVANCE-400 spectrometers and calibrated using the

1 13 residual solvent peak (CDCl3: δ 7.26 ppm H NMR, 77.16 ppm C NMR; acetone-d6: δ

1 13 1 2.05 ppm H NMR, 206.26 ppm C NMR; DMSO-d6: δ 2.50 ppm H NMR, 39.52 ppm

13 1 13 1 C NMR; CD3OD: δ 3.31 ppm H NMR, 49.00 ppm C NMR). H NMR data are reported as chemical shifts (δ ppm), multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, quint = quintet, m = multiplet, br = broad), coupling constant (Hz), relative integral. 13C

NMR data are reported as chemical shifts (δ ppm). Mass spectra were run on open access

LC-MS systems, either a PE Sciex Single Quadrupole LC/MS API-150 or a Waters using a reverse phase column, e.g., Xbridge-C18, Sunfire-C18, Thermo Aquasil/Aquasil C18,

Acquity HPLC C18, Thermo Hypersil Gold eluted using an acetonitrile and water gradient with a low percentage of an acid modifier such as 0.02% TFA. Analytical HPLC was run using an Agilent system (1100 series) with variable wavelength UV detection using

Sunfire-C18 analytical columns and reverse phase chromatography with acetonitrile and water gradient with a 0.05 or 0.1 % TFA modifier (added to each solvent). Preparative

159

HPLC was performed using a Gilson Preparative System with variable wavelength UV detection and SunFire C18 preparative columns. The compounds were eluted using a gradient of acetonitrile and water. Neutral conditions used an acetonitrile and water gradient with no additional modifier, acidic conditions used an acid modifier, usually 0.05

% or 0.1 % TFA (added to both the acetonitrile and water). Flash chromatography was run using a Teledyne Isco Combiflash RF or Companion, with normal or reverse phase, disposable Redi-Sep flash columns, and a detector with UV wavelength at 254 nm and a variety of solvents or solvent combinations. Heating of reaction mixtures with microwave irradiations was carried out on a Biotage Microwave. All compounds submitted for biological evaluation were greater than 95% pure by LCMS.

160

Methyl 8-bromooctanoate (3.44). To a mixture of 8-bromooctanoic acid (0.531 g, 2.380 mmol) in methanol (11.27 mL, 278 mmol) was added thionyl chloride (0.347 mL, 4.76 mmol) slowly. The mixture was stirred at room temperature overnight. The mixture was then concentrated under vacuum and the residue was taken up in EtOAc (30 mL) and washed with aqueous NaHCO3 (10 mL), and water (10 mL). The organic extract was dried over Na2SO4, filtered and concentrated to afford 3.44 (0.5506 g, 2.322 mmol, 98% yield)

1 as a clear oil. H NMR (400 MHz, CD3OD) δ ppm 1.30 - 1.40 (m, 4H), 1.43 - 1.52 (m, 2H)

1.58 - 1.68 (m, 2H), 1.86 (quin, J = 7.1 Hz, 2H), 2.35 (t, J = 7.5 Hz, 2H), 3.46 (t, J = 6.8

Hz, 2H), 3.64 - 3.69 (s, 3H).

161

1-Cyclopropyl-6-fluoro-7-(4-(8-methoxy-8-oxooctyl)piperazin-1-yl)-4-oxo-1,4- dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.45). A mixture of 3.25

(205.4 mg, 0.620 mmol), methyl 8-bromooctanoate 3.44 (147 mg, 0.620 mmol) and sodium bicarbonate (52.1 mg, 0.620 mmol) in N,N-dimethylformamide (DMF) (1 mL) in a sealed tube was heated to 100 ºC overnight. LCMS indicated formation of desired product (7:84

SM/P). The mixture was diluted with DMF (3 mL) and acidified by addition of TFA. The mixture was filtered and the clear solution was purified by Gilson automated Prep HPLC

(10-65% organic, 0.1% TFA) to afford 3.45 (315.6 mg, 0.498 mmol, 80% yield) as a white

1 solid. H NMR (400 MHz, CD3OD) δ ppm 1.26 (m, 2H), 1.36 - 1.51 (m, 8H) 1.66 (s, 2H),

1.80 - 1.88 (m, 2H), 2.37 (t, J = 7.4 Hz, 2H), 3.25 - 3.32 (m, 4H), 3.38 - 3.45 (m, 2H), 3.64

- 3.71 (s, 3H), 3.77 - 3.84 (m, 2H), 3.97 - 4.07 (m, 2H), 7.69 (d, J = 7.4 Hz, 1H), 8.03 (d, J

= 12.9 Hz, 1H), 8.85 (s, 1H). LCMS: [M+H]+: 488.2.

162

7-(4-(7-Carboxyheptyl)piperazin-1-yl)-1-cyclopropyl-6-fluoro-4-oxo-1,4- dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.28). To a suspension of 3.45 (233.7 mg, 0.479 mmol) in tetrahydrofuran (THF) (9.215 mL) and water (2.77 mL) was added sodium hydroxide (5M, 0.431 mL, 2.157 mmol). The mixture was stirred at room temperature. LCMS indicated completion of the reaction after 2 h. The mixture was neutralized by addition of TFA and then concentrated to dryness. The crude product was purified by Gilson automated Prep HPLC (10-50% organic, 0.1% TFA) to afford 3.28

1 (169.1 mg, 0.288 mmol, 60.0% yield) as a white foam. HNMR (400 MHz, CD3OD) δ ppm

1.27 (m, 2H), 1.45 (m, 8H), 1.60 - 1.71 (m, 2H), 1.78 - 1.93 (m, 2H) 2.33 (t, J = 7.4 Hz,

2H), 3.25 - 3.32 (m, 4H), 3.37 - 3.44 (m, 2H), 3.75 - 3.83 (m, 2H), 3.96 - 4.08 (m, 2H),

7.65 - 7.75 (m, 1H) 8.04 (d, J = 12.93 Hz, 1H), 8.85 (s, 1 H). LCMS [M+H]+: 474.2.

163

1-Cyclopropyl-6-fluoro-7-(4-(10-methoxy-10-oxodecyl)piperazin-1-yl)-4-oxo-1,4- dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.46). A mixture of 3.25

(205.4 mg, 0.620 mmol), methyl 10-bromodecanoate (164.3 mg, 0.620 mmol) and sodium bicarbonate (52.1 mg, 0.620 mmol) in N,N-dimethylformamide (DMF) (1 mL) in a sealed tube was heated to 100 ºC overnight. LCMS indicated formation of desired product (8:70

SM/P). The mixture was diluted with DMF (3 mL) and acidified by addition of TFA (0.096 mL, 1.240 mmol). The mixture was filtered and the clear solution was purified by Gilson automated Prep HPLC (10-65% organic, 0.1% TFA) to afford 3.46 (267.3 mg, 0.403 mmol,

1 65.1% yield) as a white solid. H NMR (400 MHz, CD3OD) δ ppm 1.23 - 1.28 (m, 2H),

1.35 - 1.48 (m, 12H), 1.56 - 1.70 (m, 2H), 1.77 - 1.89 (m, 2H), 2.35 (s, 2H), 3.25 - 3.29 (m,

4H), 3.37 - 3.43 (m, 2H), 3.67 (s, 3H), 3.76 - 3.83 (m, 2H), 3.96 - 4.09 (m, 2H), 7.67 - 7.76

(m, 1H), 8.03 - 8.09 (m, 1H), 8.88 (s, 1 H). LCMS: [M+H]+: 516.3.

164

7-(4-(9-Carboxynonyl)piperazin-1-yl)-1-cyclopropyl-6-fluoro-4-oxo-1,4- dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.30). To a suspension of 3.46 (217.2 mg, 0.421 mmol) in tetrahydrofuran (THF) (8.095 mL) and water (2.431 mL) was added sodium hydroxide (5M, 0.337 mL, 1.685 mmol). The mixture was stirred at rt. LCMS indicated completion of the reaction after stirring overnight. The mixture was neutralized by addition of TFA and then concentrated to dryness. The crude product was purified by Gilson automated Prep HPLC (10-50% organic, 0.1% TFA) to afford 3.30

1 (159.6 mg, 0.259 mmol, 61.5% yield) as an off-white solid. H NMR (400 MHz, CD3OD)

δ ppm 1.24 - 1.29 (m, 2H), 1.36 - 1.49 (m, 12H), 1.58 - 1.68 (m, 2H), 1.79 - 1.89 (m, 2H),

2.31 (s, 2H), 3.31 (m, 4H), 3.37 - 3.45 (m, 2H), 3.75 - 3.83 (m, 2H), 3.94 - 4.08 (m, 2H),

7.64 - 7.75 (m, 1H), 7.99 - 8.11 (m, 1H), 8.87 (s, 1H). LCMS [M+H]+: 502.2.

165

Methyl 8-azidooctanoate (3.47). A mixture of methyl 8-bromooctanoate (0.274 g, 1.155 mmol) and sodium azide (0.376 g, 5.78 mmol) in N,N-dimethylformamide (DMF) (1 mL) in a sealed tube was heated to 80 ºC for 48 h. The mixture was diluted with EtOAc/hexanes

(4:1, 25 mL) and washed with NaHCO3 (2 x 15 mL), water (2 x 15 mL). The mixture was filtered and concentrated under vacuum to afford 3.47 (231.1 mg, 1.160 mmol, 100% yield)

1 as a clear oil. H NMR (400 MHz, CD3OD) δ ppm 1.32 - 1.45 (m, 6H), 1.56 - 1.68 (m,

4H), 2.35 (t, J = 7.5 Hz, 2H), 3.27 - 3.32 (m, 2H), 3.64 - 3.69 (s, 3H).

1-Cyclopropyl-6-fluoro-7-(4-((1-(8-methoxy-8-oxooctyl)-1H-1,2,3-triazol-4- yl)methyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid,

Trifluoroacetic acid salt (3.48). To a mixture of 3.33 (0.238 g, 0.644 mmol), methyl 8- azidooctanoate 3.47 (0.128 g, 0.644 mmol) and N-ethyl-N-isopropylpropan-2-amine

(0.315 mL, 1.804 mmol) in dimethyl sulfoxide (DMSO) (2.078 mL) under N2 was added copper(I) iodide (0.061 g, 0.322 mmol). The mixture was stirred at room temperature overnight. LCMS indicated 75% completion of the reaction. The mixture was diluted with 166

DCM (40 mL) and washed with NH4OH/sat. NH4Cl (1:4, 3 x 30 mL) and saturated NH4Cl

(30 mL). The organic layer was dried over Na2SO4, filtered and concentrated. The residue was purified by Gilson automated prep HPLC (10-70% organic, 0.1% TFA) to afford 3.34

1 (235.7 mg, 0.338 mmol, 52.5% yield) as an off-white solid. H NMR (400 MHz, CD3OD)

δ ppm 1.26 (br s, 2H), 1.33 - 1.45 (m, 8H), 1.56 - 1.66 (m, 2H), 1.92 - 2.02 (m, 2H), 2.33

(t, J = 7.5 Hz, 2H), 3.56 -3.72 (m, 9H), 3.75 - 3.84 (m, 2H), 4.50 (t, J = 7.1 Hz, 2H), 4.64

(s, 2H), 7.64 - 7.72 (m, 1H), 8.03 (d, J = 12.9 Hz, 1H), 8.26 (s, 1H), 8.85 (s, 1H). LCMS:

[M+H]+: 569.3.

7-(4-((1-(7-Carboxyheptyl)-1H-1,2,3-triazol-4-yl)methyl)piperazin-1-yl)-1- cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.34). To a suspension of 3.48 (205.6 mg, 0.362 mmol) in tetrahydrofuran (THF)

(6.951 mL) and water (2.088 mL) was added sodium hydroxide (5M. 0.289 mL, 1.446 mmol). The mixture was stirred at room temperature. LCMS indicated completion of the reaction after 1 h. The mixture was neutralized by addition of TFA and then concentrated to dryness. The crude product was purified by Gilson automated Prep HPLC (10-50% organic, 0.1% TFA) to afford 3.34 (120.9 mg, 0.181 mmol, 50.0% yield), as a white solid. 167

1 H NMR (400 MHz, CD3OD) δ ppm 1.25 (m, 2H), 1.34 - 1.46 (m, 8H), 1.55 - 1.67 (m,

2H), 1.92 - 2.03 (m, 2H), 2.29 (t, J = 7.4 Hz, 2H), 3.64 (br s, 6H), 3.75 - 3.85 (m, 2H), 4.51

(t, J = 7.1 Hz, 2H), 4.64 (s, 2H), 7.69 (d, J = 7.4 Hz, 1H), 8.04 (d, J = 12.9 Hz, 1H), 8.26

(s, 1H), 8.86 (s, 1H). LCMS [M+H]+: 555.3.

Methyl 9-azidononanoate (3.49). A mixture of methyl 9-bromononanoate (0.0708 g,

0.282 mmol) and sodium azide (0.092 g, 1.409 mmol) in N,N-dimethylformamide (DMF)

(0.287 mL) in a sealed tube was heated to 80 ºC for 30 h. The mixture was diluted with

EtOAc/hexanes (4:1, 10 mL) and washed with NaHCO3 (2 x 5 mL), water (2 x 5 mL). The mixture was filtered and concentrated under vacuum to afford 3.49 (56 mg, 0.263 mmol,

1 93% yield) as a clear oil. H NMR (400 MHz, CD3OD) δ ppm 1.30 - 1.43 (m, 8H), 1.55 -

1.68 (m, 4H), 2.34 (t, J = 7.4 Hz, 2H), 3.30 (t, J = 6.8 Hz, 2H), 3.67 (s, 3H).

168

1-Cyclopropyl-6-fluoro-7-(4-((1-(9-methoxy-9-oxononyl)-1H-1,2,3-triazol-4- yl)methyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid,

Trifluoroacetic acid salt (3.50). To a mixture of 3.33 (0.097 g, 0.263 mmol), 3.49 (0.056 g, 0.263 mmol) and N-ethyl-N-isopropylpropan-2-amine (0.129 mL, 0.736 mmol) in dimethyl sulfoxide (DMSO) (0.848 mL) under N2 was added copper(I) iodide (0.025 g,

0.132 mmol). The mixture was stirred at room temperature overnight. LCMS indicated completion of the reaction. The mixture was diluted with DCM (20 mL) and washed with

NH4OH/sat. NH4Cl (1:4, 3 x 15 mL) and saturated NH4Cl (15 mL). The organic layer was dried over Na2SO4, filtered and concentrated. The residue was purified by Gilson automated prep HPLC (10-70% organic, 0.1% TFA) to afford 3.50 (77.5 mg, 0.111 mmol,

1 42.3% yield), trifluoroacetic acid salt as an off-white solid. H NMR (400 MHz, CD3OD)

δ ppm 1.21 - 1.29 (m, 2H), 1.31 - 1.40 (m, 8H), 1.43 (d, J = 6.3 Hz, 2H), 1.56 - 1.66 (m,

2H), 1.91 - 2.03 (m, 2H), 2.32 (t, J = 7.4 Hz, 2H), 3.53 - 3.71 (m, 9H), 3.76 - 3.83 (m, 2H),

4.50 (m, 2H) 4.64 (s, 2H), 7.63 - 7.76 (m, 1H), 8.05 (d, J = 12.9 Hz, 1H), 8.25 (s, 1H), 8.87

(s, 1H). LCMS: [M+H]+: 583.4.

169

7-(4-((1-(8-Carboxyoctyl)-1H-1,2,3-triazol-4-yl)methyl)piperazin-1-yl)-1- cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.35). To a suspension of 3.50 (66.1 mg, 0.113 mmol) in tetrahydrofuran (THF)

(2.181 mL) and water (0.655 mL) was added sodium hydroxide (5M, 0.091 mL, 0.454 mmol). The mixture was stirred at room temperature. LCMS indicated completion of the reaction after 1.5 hours. The mixture was neutralized by addition of TFA and then concentrated to dryness. The crude product was purified by Gilson automated Prep HPLC

(10-50% organic, 0.1% TFA) to afford 3.35 (51 mg, 0.075 mmol, 65.9% yield) as an off-

1 white solid. H NMR (400 MHz, CD3OD) δ ppm 1.25 (br s, 2H), 1.33 - 1.45 (m, 10H),

1.60 (t, J = 7.1 Hz, 2H), 1.90 - 2.03 (m, 2H), 2.28 (t, J = 7.4 Hz, 2H), 3.54 - 3.87 (m, 8H),

4.50 (t, J = 7.1 Hz, 2H), 4.64 (s, 2H), 7.65 (d, J = 7.1 Hz, 1H), 7.94 (d, J = 12.9 Hz, 1H),

8.26 (s, 1H), 8.78 (s, 1H). LCMS [M+H]+: 569.4.

Methyl 10-azidodecanoate (3.51). A mixture of methyl 10-bromodecanoate (0.263 g,

0.992 mmol) and sodium azide (0.322 g, 4.96 mmol) in N,N-dimethylformamide (DMF)

(1 mL) in a sealed tube was heated to 80 ºC for 30 h. The mixture was diluted with

170

EtOAc/hexanes (4:1, 25 mL) and washed with NaHCO3 (2 x 15 mL), water (2 x 15 mL).

The mixture was filtered and concentrated under vacuum to afford 3.51 (220.5 mg, 0.970

1 mmol, 98% yield) as a clear oil. H NMR (400 MHz, CD3OD) δ ppm 1.34 - 1.42 (m, 10H),

1.61 (m, 4H), 2.34 (t, J = 7.5 Hz, 2H), 3.30 (t, J = 6.8 Hz, 2H), 3.67 (s, 3H).

1-Cyclopropyl-6-fluoro-7-(4-((1-(10-methoxy-10-oxodecyl)-1H-1,2,3-triazol-4- yl)methyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid,

Trifluoroacetic acid salt (3.52). To a mixture of 3.33 (300 mg, 0.813 mmol), methyl 10- azidodecanoate 3.51 (185 mg, 0.813 mmol) and N-ethyl-N-isopropylpropan-2-amine

(0.398 mL, 2.276 mmol) in dimethyl sulfoxide (DMSO) (2.623 mL) under N2 was added copper(I) iodide (77 mg, 0.407 mmol). The mixture was stirred at room temperature overnight. LCMS indicated completion of the reaction. The mixture was diluted with DCM

(40 mL) and washed with NH4OH/sat. NH4Cl (1:4, 3 x 30 mL) and saturated NH4Cl (30 mL). The organic layer was dried over Na2SO4, filtered and concentrated. The residue was purified by Gilson automated prep HPLC (10-70% organic, 0.1% TFA) to afford 3.52

1 (260.5 mg, 0.359 mmol, 44.2% yield), as a brownish solid. H NMR (400 MHz, CD3OD) 171

δ ppm 1.25 (m, 2H), 1.31 - 1.39 (m, 10H), 1.43 (m, 2H), 1.60 (br s, 2H), 1.96 (m, 2H), 2.32

(t, J = 7.5 Hz, 2H), 3.55 - 3.73 (m, 9H), 3.76 - 3.83 (m, 2H), 4.50 (t, J = 7.1 Hz, 2H), 4.64

(s, 2H), 7.68 (d, J = 7.4 Hz, 1H), 8.02 (d, J = 12.9 Hz, 1H), 8.26 (s, 1H), 8.84 (s, 1H).

LCMS [M+H]+: 597.5.

7-(4-((1-(9-Carboxynonyl)-1H-1,2,3-triazol-4-yl)methyl)piperazin-1-yl)-1- cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.36). To a suspension of 3.52 (215.2 mg, 0.361 mmol) in tetrahydrofuran (THF)

(6.934 mL) and water (2.082 mL) was added sodium hydroxide (5M, 0.289 mL, 1.443 mmol). The mixture was stirred at room temperature overnight. The mixture was neutralized by addition of TFA and then concentrated to dryness. The crude product was purified by Gilson automated Prep HPLC (10-50% organic, 0.1% TFA) to afford 3.36

1 (155.7 mg, 0.223 mmol, 62.0% yield) as a white solid. H NMR (400 MHz, CD3OD) δ ppm 1.25 (m, 2H), 1.30 - 1.46 (m, 12H), 1.53 - 1.65 (m, 2H), 1.91 - 2.02 (m, 2H), 2.28 (t,

J = 7.5 Hz, 2H), 3.63 (br s, 8H), 4.50 (t, J = 7.1 Hz, 2H), 4.62 (s, 2H), 7.69 (d, J = 7.1 Hz,

1H), 8.04 (d, J = 12.9 Hz, 1H), 8.25 (s, 1H), 8.86 (s, 1H). LCMS [M+H]+: 583.4. 172

Methyl 11-azidoundecanoate (3.53). A mixture of methyl 11-bromoundecanoate (0.274 g, 0.981 mmol) and sodium azide (0.319 g, 4.91 mmol) in N,N-dimethylformamide (DMF)

(1 mL) in a sealed tube was heated to 80 ºC for 64 h. The mixture was diluted with

EtOAc/hexanes (4:1, 25 mL) and washed with NaHCO3 (2 x 15 mL), water (2 x 15 mL).

The mixture was filtered and concentrated under vacuum to afford 3.53 (233.3 mg, 0.967

1 mmol, 99% yield) as a clear oil. H NMR (400 MHz, CDCl3) δ ppm 1.26 - 1.44 (m, 12H),

1.58 - 1.73 (m, 4H), 2.34 (t, J = 7.5 Hz, 2H), 3.30 (t, J = 7.0 Hz, 2H), 3.71 (s, 3H).

1-Cyclopropyl-6-fluoro-7-(4-((1-(11-methoxy-11-oxoundecyl)-1H-1,2,3-triazol-4- yl)methyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid,

Trifluoroacetic acid salt (3.54). To a mixture of 3.33 (344 mg, 0.930 mmol), methyl 11- azidoundecanoate 3.53 (224.5 mg, 0.930 mmol) and N-ethyl-N-isopropylpropan-2-amine

(0.455 mL, 2.60 mmol) in dimethyl sulfoxide (DMSO) (3 mL) under N2 was added copper(I) iodide (89 mg, 0.465 mmol). The mixture was stirred at room temperature 173

overnight. LCMS indicated completion of the reaction. The mixture was diluted with DCM

(403.7 mg, 0.557 mmol, 59.9% yield), trifluoroacetic acid salt as a brownish solid. 1H NMR

(400 MHz, CDCl3) δ ppm 1.23 - 1.27 (m, 2H), 1.28 - 1.42 (m, 14H), 1.43 - 1.49 (m, 2H),

1.65 (t, J = 7.2 Hz, 2H), 1.96 (t, J = 6.8 Hz, 2H), 2.34 (t, J = 7.6 Hz, 2H), 3.51 - 3.88 (m,

12H), 4.43 (t, J = 7.4 Hz, 2H), 4.52 (s, 2H), 7.48 (d, J = 6.8 Hz, 1H), 7.93 (s, 1H), 8.07 (d,

J = 12.4 Hz, 1H), 8.82 (s, 1H). LCMS: [M+H]+: 611.4.

7-(4-((1-(10-Carboxydecyl)-1H-1,2,3-triazol-4-yl)methyl)piperazin-1-yl)-1- cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.37). To a suspension of 3.54 (319.7 mg, 0.523 mmol) in tetrahydrofuran (THF)

(10 mL) and water (3.00 mL) was added sodium hydroxide (5M, 0.419 mL, 2.094 mmol).

The mixture was stirred at room temperature. LCMS indicated completion of the reaction after 3 h. The mixture was neutralized by addition of TFA and then concentrated to dryness.

The crude product was purified by Gilson automated Prep HPLC (10-50% organic, 0.1% 174

TFA) to afford 3.37 (279.4 mg, 0.389 mmol, 74.3% yield) as an off-white solid. 1H NMR

(400 MHz, CD3OD) δ ppm 1.26 (br s, 2H), 1.32 - 1.38 (m, 12H), 1.43 (m, 2H), 1.60 (t, J =

7.2 Hz, 2H), 1.92 - 2.01 (m, 2H), 2.28 (t, J = 7.5 Hz, 2H), 3.57 - 3.85 (m, 9H), 4.50 (t, J =

7.1 Hz, 2H), 4.64 (s, 2H), 7.68 (d, J = 7.4 Hz, 1H), 8.02 (d, J = 12.9 Hz, 1H), 8.25 (s, 1H),

8.84 (s, 1H). LCMS: [M+H]+: 597.4.

1-Cyclopropyl-6-fluoro-7-(4-(5-methoxy-5-oxopentanoyl)piperazin-1-yl)-4-oxo-1,4- dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.40). To a mixture of

3.25 (200 mg, 0.604 mmol), 5-methoxy-5-oxopentanoic acid (88 mg, 0.604 mmol) and triethylamine (0.084 mL, 0.604 mmol) in N,N-dimethylformamide (DMF) (5.651 mL)

º under N2 at 0 C was added 2,4,6-tripropyl-1,3,5,2,4,6-trioxatriphosphinane 2,4,6-trioxide,

50% wt in ethyl acetate (0.180 mL, 0.604 mmol) dropwise. The mixture was stirred at room temperature for 24 hours. LCMS indicated formation of target molecule. The mixture was concentrated in vacuo and the resulting residue was diluted with water (20 mL) and extracted with DCM (3 x 20 mL). The organic phase was washed with saturated NaHCO3

(2 x 30 mL), brine (1 x 60 mL) and dried over Na2SO4, filtered then concentrated. The

175

crude product was purified by Gilson automated Prep HPLC (10-70% organic, 0.1% TFA) to afford 3.40 (91 mg, 0.157 mmol, 26.0% yield), as a white powder. 1H NMR (400 MHz,

CD3OD) δ ppm 1.20 - 1.29 (m, 2H), 1.39 - 1.46 (m, 2H), 1.91 - 2.00 (m, 2H), 2.46 (t, J =

7.4 Hz, 2H), 2.51 - 2.59 (t, J = 7.6 Hz, 2H), 3.37 - 3.41 (m, 2H), 3.42 - 3.48 (m, 2H), 3.70

(s, 3H), 3.79 - 3.86 (m, 4H), 7.60 - 7.70 (m, 1H), 7.97 - 8.05 (m, 1H), 8.81 - 8.90 (s, 1H).

LCMS: [M+H]+: 460.1.

7-(4-(4-Carboxybutanoyl)piperazin-1-yl)-1-cyclopropyl-6-fluoro-4-oxo-1,4- dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.38). To a suspension of 3.40 (78.7 mg, 0.171 mmol) in tetrahydrofuran (THF) (3.293 mL) and water (0.989 mL) was added sodium hydroxide (1M, 0.685 mL, 0.685 mmol). The mixture was stirred at room temperature. LCMS indicated completion of the reaction after 1 h. The mixture was neutralized by addition of TFA (0.066 ml, 0.856 mmol) and then concentrated to dryness.

The crude product was purified by Gilson automated Prep HPLC (10-50% organic, 0.1%

TFA) to afford 3.38 (56.1 mg, 0.100 mmol, 58.5% yield) as a white solid. 1H NMR (400

MHz, CD3OD) δ ppm 1.20 - 1.28 (m, 2H), 1.38 - 1.46 (m, 2H), 1.89 - 2.00 (m, 2H), 2.43

(t, J = 7.4 Hz, 2H), 2.51 - 2.60 (t, J = 7.6 Hz, 2H), 3.39 - 3.50 (m, 4H), 3.80 - 3.87 (m, 4H),

7.60 - 7.72 (m, 1H) 7.98 - 8.07 (m, 1H), 8.86 (s, 1H). LCMS: [M+H]+: 446.1.

176

1-Cyclopropyl-6-fluoro-7-(4-(9-methoxy-9- oxononanoyl)piperazin-1-yl)-4-oxo-1,4- dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.41). To a mixture of

3.25 (200 mg, 0.604 mmol), 9-methoxy-9-oxononanoic acid (122 mg, 0.604 mmol) and triethylamine (0.084 mL, 0.604 mmol) in N,N-dimethylformamide (DMF) (5.651 mL)

º under N2 at 0 C was added 2,4,6-tripropyl-1,3,5,2,4,6-trioxatriphosphinane 2,4,6-trioxide,

50% wt in ethyl acetate (0.180 mL, 0.604 mmol) dropwise. The mixture was stirred at room temperature for 25 hours. LCMS indicated formation of target molecule. The mixture was concentrated in vacuo and the resulting residue was diluted with water (20 mL) and extracted with DCM (3 x 20 mL). The organic phase was washed with saturated NaHCO3

(2 x 30 mL), brine (1 x 60 mL) and dried over Na2SO4, filtered then concentrated. The crude product was purified by Gilson automated Prep HPLC (10-70% organic, 0.1% TFA) to afford 3.41 (252.1 mg, 0.396 mmol, 65.7% yield) as a yellow solid. 1H NMR (400 MHz,

CD3OD) δ ppm 1.23 - 1.28 (m, 2H), 1.37 - 1.46 (m, 8H), 1.59 - 1.71 (m, 4H), 2.34 (t, J =

7.4 Hz, 2H), 2.50 (t, J = 7.6 Hz, 2H), 3.37 - 3.42 (m, 2H), 3.42 - 3.47 (m, 2H), 3.65 - 3.68

177

(s, 3H), 3.83 (m, 4H), 7.56 - 7.66 (m, 1H), 7.87 - 7.97 (m, 1H), 8.29 (s, 1H), 8.81 (s, 1H).

LCMS [M+H]+: 516.2.

7-(4-(8-Carboxyoctanoyl)piperazin-1-yl)-1-cyclopropyl-6-fluoro-4-oxo-1,4- dihydroquinoline-3-carboxylic acid, Trifluoroacetic acid salt (3.39). To a suspension of 3.41 (100 mg, 0.194 mmol) in tetrahydrofuran (THF) (3.729 mL) and water (1.120 mL) was added sodium hydroxide (5M, 0.155 mL, 0.776 mmol). The mixture was stirred at room temperature. LCMS indicated completion of the reaction after 1 h. The mixture was neutralized by addition of TFA and then concentrated to dryness. The crude product was purified by Gilson automated Prep HPLC (10-50% organic, 0.1% TFA) to afford 3.39 (48.6

1 mg, 0.078 mmol, 40.3% yield) as an off-white solid. H NMR (400 MHz, CD3OD) δ ppm

1.25 (m, 2H) 1.36 - 1.46 (m, 8H) 1.65 (m, 4H) 2.31 (t, J = 7.4 Hz, 2H), 2.49 (t, J = 7.6 Hz,

2H), 3.39 (m, 4H), 3.83 (m, 4H), 7.60 - 7.71 (m, 1H), 8.01 (d, J = 13.2 Hz, 1H), 8.85 (s,

1H). LCMS: [M+]+: 502.1.

178

5.3 Synthesis and characterization of anti-biofilm compounds

All reactions were carried out under an argon atmosphere with dry solvents under anhydrous conditions unless otherwise stated. Oven-dried syringes were used to transfer air and moisture sensitive liquids. All commercial reagents, anhydrous solvents, and reagent-grade solvents were purchased from Sigma-Aldrich, Fisher Scientific, VWR, or

PurePEG and used as received without further purification, unless otherwise stated.

Reactions were monitored by thin layer chromatography (TLC) using aluminum backed pre-coated silica gel plates from (TLC Silica Gel F-254, 200 µm, Dynamic Adsorbents) using UV light as the visualizing agent. Flash chromatography was performed using silica gel (60Å, pore size 32-63 µm, Dynamic Adsorbents). Silica gel was deactivated by first washing with a 10% triethylamine solution in the eluent and then washing with the eluent itself (3x) unless otherwise stated. Deuterated solvents for NMR were purchased from

Cambridge Isotope Labs and used as received. NMR spectra were recorded on Bruker

DPX250, AV300, or DRX400 MHz spectrometers and calibrated using the residual

1 13 undeuterated solvent peak (CDCl3: δ 7.26 ppm H NMR, 77.16 ppm C NMR; acetone-

1 13 1 d6: δ 2.05 ppm H NMR, 206.26 ppm C NMR; DMSO-d6: δ 2.50 ppm H NMR, 39.52

13 1 13 1 ppm C NMR; CD3OD: δ 3.31 ppm H NMR, 49.00 ppm C NMR). Proton ( H) NMR data is reported as follows: chemical shift in ppm (multiplicity [as: s = singlet, d = doublet, t = triplet, q = quartet, quint = quintet, sext = sextet, sept = septet, m = multiplet, br = broad], coupling constant(s) in Hz, relative integration). Carbon (13C) NMR data was reported as chemical shift (δ) in ppm. High resolution mass spectra (HRMS) were recorded

179

on Thermo LTQ Orbitrap by electrospray ionization (ESI) time of flight experiments and reported as m/z.

180

Methyl 2-amino-5,6-dihydro-4H-cyclopenta[b]thiophene-3-carboxylate (4.8). To 4 test-tubes each containing biphasic solution of cyclopentanone 4.9 (1 g, 11.88 mmol), methyl 2-cyanocetate 4.42 (1.177 g, 11.88 mmol), and diisoproylamine (2.303 g, 17.82 mmol) was added sulfur (419.3 mg, 13.07 mmol). The mixture was sonicated for 30 minutes, in 5-minute intervals with stirring. Upon completion, the reaction was transferred to a separatory funnel, and extracted with ethyl acetate (30 ml x 3). The combined organic phase was washed with brine (50 ml x 2), dried over anhydrous sodium sulfate, and concentrated. The crude material was dry loaded onto a column and purified by normal phase silica gel chromatography (EtOAc:hexanes, 1:10) to afford 4.8 (3.534 g, 17.92

1 mmol, 38%), as an off-white solid. H NMR (400 MHz, CDCl3) δ 5.87 (s, 2H), 3.78 (s,

13 3H), 2.80 (m, 2H), 2.73 – 2.67 (m, 2H), 2.35 – 2.22 (m, 2H). C NMR (101 MHz, CDCl3)

δ 166.65, 166.30, 142.66, 121.47, 102.75, 50.87, 30.83, 29.00, 27.36. IR (film) cm–1: 3407,

3288, 3161, 2855, 1649, 1595, 1574, 1487, 1446, 1351, 1289, 1268, 1037, 779.

181

Methyl 2-isothiocyanato-5,6-dihydro-4H-cyclopenta[b]thiophene-3-carboxylate

(4.7). To an oven dried round bottom flask containing a solution of 4.8 (1.433 g, 7.26 mmol), hexachloroethane (2.581 g, 10.90 mmol), and triphenyl phosphine (2.858 g, 10.90 mmol) in dry acetonitrile (35 mL) was added triethylamine (2.204 g, 21.78 mmol) dropwise. The solution was left to stir at room temperature for 2 hours. The solvent was then removed under reduced pressure, and upon confirmation of target molecule by 1H

NMR, the resulting iminophosphorane was carried onto the next reaction without further purification. To the round bottom flask containing iminophosphorane (2.525 g, 5.518 mmol) was added excess carbon disulfide (4.201g, 55.18 mmol). After stirring at room temperature overnight, the excess CS2 was distilled off, followed by addition of ether (10 mL) to precipitate out triphenylphosphine sulfide, which was filtered off. The crude material was purified by normal phase silica gel chromatography 100%-10% ethyl acetate in hexanes to afford aryl isothiocyanate 4.7 (1.056 g, 4.41 mmol, 80%), as a clear oil. 1H

NMR (300 MHz, CDCl3) δ 3.88 (s, 3H), 2.98 – 2.79 (m, 4H), 2.45 – 2.29 (m, 2H).

182

3-(furan-2-ylmethyl)-2-thioxo-1,2,3,5,6,7-hexahydro-4H-cyclopenta[4,5]thieno[2,3- d]pyrimidin-4-one (4.6). To an oven dried round bottom flask containing iosthiocyanate

4.7 (189.2 mg. 0.791 mmol) in anhydrous dichloromethane (8 mL) was added furfurylamine. After allowing the reaction to stir at room temperature for 2 hours, the solvent was evaporated under reduced pressure to afford an intermediate thiourea, which was carried onto the next reaction without further purification.

To a solution of intermediate thiourea (266.1 mg, 0.791 mmol) in tert-butanol (8 mL) was added potassium tert-butoxide (177.5 mg, 1.58 mmol). After stirring at 75 °C for 4 hours, the reaction was allowed to cool down to room temperature and concentrated. Water (5 mL) was added and the crude mixture was extracted with ethyl acetate (5 mL x 3), washed with brine (15 mL) and concentrated to afford 4.6 as a white powder (230.5 mg, 0.76 mmol,

1 96%) as an off-white powder. H NMR (400 MHz, CDCl3) δ 7.35 (m, 1H), 6.51 (m, 1H),

6.31 (m, 1H), 5.72 (s, 2H), 3.01 (m, 2H), 2.87 (m, 2H), 2.47 (m, 2H). 13C NMR (176 MHz,

CDCl3) δ 174.22, 156.55, 152.41, 149.44, 143.05, 142.29, 141.80, 140.59, 134.17, 110.81,

110.43, 110.27, 42.67, 41.21, 29.07, 28.69, 28.45.

183

3-(furan-2-ylmethyl)-2-(((5-hydroxy-1H-pyrazol-3-yl)methyl)thio)-3,5,6,7- tetrahydro-4H-cyclopenta[4,5]thieno[2,3-d]pyrimidin-4-one (4.5). To a solution of 4.6

(151.1 mg, 0.496 mmol) in acetonitrile (5 mL) was added ethyl 4-bromo-3-oxobutanoate

4.11 (103.7 mg, 0.496 mmol) followed by potassium carbonate (164.1 mg, 0.993 mmol).

The reaction was left to stir at room temperature for 4 hours, and subsequently concentrated under reduced pressure. Water (5 mL) was added followed by extraction with ethyl acetate

(5 mL x 3), washing with brine (15 mL) and drying over sodium sulfate. Upon concentration, the crude material was carried onto the next reaction without further purification. 1H NMR (400 MHz, DMSO) δ 7.35 (m, 1H), 6.43 (m, 1H), 6.31 (m, 1H), 5.32

(s, 2H), 4.21 (q, J = 7.1 Hz, 2H), 4.12 (s, 2H), 3.70 (s, 2H), 3.05 – 2.99 (m, 2H), 2.92 –

2.86 (m, 2H), 2.43 – 2.39 (m, 2H), 1.28 (t, J = 7.1 Hz, 3H). To a solution of -ketoester intermediate (108.8 mg, 0.252 mmol) in acetic acid (3 mL) was added hydrazine monohydrate (15.7 mg, 0.314 mmol). The reaction was refluxed at 85 °C for 4 hours. The acetic acid was concentrated under vacuum to a minimum volume. Water (5 mL) was added followed extraction with ethyl acetate (5 ml x 3). The combined organic phase was washed with brine (20 mL x 3), dried over anhydrous sodium sulfate, and concentrated.

184

The crude material was triturated with ethyl acetate (3 mL x 2) and methanol (2 mL) to afford 4.5 (52.0 mg, 0.13 mmol 52%), as an off-white solid. 1H NMR (400 MHz, DMSO)

δ 7.58 (m, 1H), 6.40 (m, 1H), 6.37 (m, 1H), 5.43 (s, 1H), 5.25 (s, 2H), 4.34 (s, 2H), 2.91

(t, J = 7.3 Hz, 4H), 2.42 – 2.34 (m, 2H). 13C NMR (101 MHz, DMSO) δ 172.88, 171.21,

166.86, 160.34, 157.65, 156.61, 149.24, 143.66, 140.34, 137.54, 116.52, 111.57, 110.02,

+ 89.98, 40.93, 29.92, 29.47, 28.72. HRMS m/z calc’d for C18H17N4O3S2 [M+H] : 401.0737, found 401.0415.

3-(furan-2-ylmethyl)-2-((4-methoxybenzyl)thio)-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-4-one (4.15). To a solution of 4.6 (21.1 mg, 0.069 mmol) in acetonitrile (1 mL) was added 1-(bromomethyl)-4-methoxybenzene 4.43 (13.87 mg, 0.069 mmol), followed by potassium carbonate (22.91 mg, 0.139 mmol). The reaction was left to stir at room temperature for 6 hours, and subsequently concentrated under reduced pressure. Water (2 mL) was added followed by extraction with ethyl acetate (3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography 10% - 20% ethyl acetate in hexanes to afford 4.15 (25.60 mg, 0.060 mmol, 87%), as an orange solid. 1H

185

NMR (400 MHz, CDCl3) δ 7.37 – 7.32 (m, 3H), 6.87 – 6.83 (m, 2H), 6.37 (m, 1H), 6.29

(m, 1H), 5.30 (s, 2H), 4.44 (s, 2H), 3.79 (s, 3H), 3.10 – 3.01 (m, 2H), 2.92 (m, 2H), 2.44

13 (m, 2H). C NMR (101 MHz, CDCl3) δ 167.10, 159.31, 158.21, 155.76, 148.84, 142.52,

140.46, 136.97, 130.73, 127.81, 116.61, 114.20, 110.55, 109.86, 55.41, 40.39, 36.98,

+ 29.69, 29.12, 28.07. HRMS m/z calc’d for C22H20N2NaO3S2 [M+Na] : 447.0813, found

447.0818.

2-(allylthio)-3-(furan-2-ylmethyl)-3,5,6,7-tetrahydro-4H-cyclopenta[4,5]thieno[2,3 d]pyrimidin-4-one (4.16). To a solution of 4.6 (40.3 mg, 0.132 mmol) in acetonitrile (1.3 mL) was added allyl bromide 4.44 (15.97 mg, 0.132 mmol), followed by potassium carbonate (43.76 mg, 0.265 mmol). The reaction was left to stir at room temperature for 4 hours, and subsequently concentrated under reduced pressure. Water (2 mL) was added followed by extraction with ethyl acetate (3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography 10% ethyl acetate in hexanes to afford 4.16 (34.8 mg,

1 0.101 mmol, 77%), as a flakey white solid. H NMR (400 MHz, CDCl3) δ 7.37 – 7.34 (m,

1H), 6.40 (m, 1H), 6.31 (m, 1H), 5.96 (m, 1H), 5.38 (m, 1H), 5.33 (s, 2H), 5.19 (m, 1H),

3.97 – 3.87 (m, 2H), 3.13 – 3.00 (m, 2H), 2.97 – 2.86 (m, 2H), 2.52 – 2.36 (m, 2H). 13C

186

NMR (101 MHz, CDCl3) δ 167.08, 158.19, 155.34, 148.86, 142.52, 140.39, 136.98,

132.32, 119.30, 116.55, 110.57, 109.86, 40.38, 35.63, 29.81, 29.66, 29.10, 28.05. HRMS

+ m/z calc’d for C18H17N2NaO2S2 [M+Na] : 367.0551, found 367.0559.

2-(benzylthio)-3-(furan-2-ylmethyl)-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-4-one (4.17). To a solution of 4.6 (16.4 mg, 0.054 mmol) in acetonitrile (1 mL) was added (bromomethyl)benzene 4.45 (9.24 mg, 0.054 mmol), followed by potassium carbonate (17.8 mg, 0.108 mmol). The reaction was left to stir at room temperature for 1 hours, and subsequently concentrated under reduced pressure. Water (2 mL) was added followed by extraction with ethyl acetate (3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography 10% ethyl acetate in hexanes to afford 4.17 (19.6 mg, 0.05 mmol, 92%), as a white solid. 1H NMR (400 MHz,

CDCl3) δ 7.42 (m, 2H), 7.36 – 7.27 (m, 4H), 6.38 (m, 1H), 6.30 (m, 1H), 5.31 (s, 2H), 4.49

13 (s, 2H), 3.05 (m, 2H), 2.92 (m, 2H), 2.50 – 2.39 (m, 2H). C NMR (101 MHz, CDCl3) δ

167.05, 158.20, 155.57, 148.83, 142.55, 140.48, 137.05, 136.06, 129.54, 129.53, 128.78,

128.77, 127.84, 116.65, 110.58, 109.89, 40.41, 37.34, 29.70, 29.12, 28.09. HRMS m/z

+ calc’d for C21H18N2NaO2S2 [M+Na] : 417.0707, found 417.0717. 187

2-((2-(1H-indol-2-yl)ethyl)thio)-3-(furan-2-ylmethyl)-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-4-one (4.18). To a solution of 4.6 (43.5 mg, 0.143 mmol) in acetonitrile (1.5 mL) was added 3-(2-bromoethyl)-1H-indole 4.46 (32.0 mg,

0.143 mmol), followed by potassium carbonate (47.26 mg, 0.286 mmol). The reaction was left to stir at room temperature overnight, and subsequently concentrated under reduced pressure. Water (4 mL) was added followed by extraction with ethyl acetate (3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, The crude material was triturated with ethyl acetate (3 mL x 3) to afford 4.18 (46.6 mg, 0.104

1 mmol, 73%), as an off-white solid. H NMR (400 MHz, CDCl3) δ 8.01 (s, 1H), 7.80 (m,

1H), 7.42 – 7.33 (m, 2H), 7.25 – 7.16 (m, 2H), 7.07 (m, 1H), 6.43 – 6.37 (m, 1H), 6.31 (m,

1H), 5.34 (s, 2H), 3.55 (m, 2H), 3.28 – 3.18 (m, 2H), 3.07 (m, 2H), 2.94 (m, 2H), 2.52 –

2.38 (m, 2H). 13C NMR (101 MHz, DMSO) δ 166.31, 156.87, 156.59, 148.51, 142.68,

139.45, 136.35, 136.22, 126.90, 123.09, 121.02, 118.42, 118.32, 115.55, 112.41, 111.43,

110.66, 108.99, 32.78, 29.03, 28.61, 27.35, 24.83. HRMS m/z calc’d for C24H21N3NaO2S2

[M+Na]+: 470.0973, found 470.0976.

188

3-(furan-2-ylmethyl)-2-((pyridin-4-ylmethyl)thio)-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-4-one (4.19). To a solution of 4.6 (44.1 mg, 0.145 mmol) in acetonitrile (1.5 mL) was added 4-(bromomethyl)pyridine 4.47 (36.7 mg, 0.145 mmol), followed by potassium carbonate (47.9 mg, 0.29 mmol). The reaction was left to stir at room temperature for 6 hours, and subsequently concentrated under reduced pressure. Water (3 mL) was added followed by extraction with ethyl acetate (3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography 15% ethyl acetate in hexanes to afford 4.19 (21.9 mg, 0.055 mmol, 38%), as an off-white solid. 1H NMR (400

MHz, CDCl3) δ 8.56 (s, 2H), 7.44 – 7.31 (m, 3H), 6.41 – 6.37 (m, 1H), 6.31 (m, 1H), 5.31

(s, 2H), 4.44 (s, 2H), 3.11 – 3.00 (m, 2H), 2.91 (m, 2H), 2.52 – 2.36 (m, 2H). 13C NMR

(101 MHz, CDCl3) δ 166.58, 157.98, 154.28, 149.79, 148.57, 146.20, 142.61, 140.52,

137.42, 124.41, 116.78, 110.64, 109.97, 40.42, 35.54, 29.65, 29.04, 28.06. HRMS m/z

+ calc’d for C20H17N3NaO2S2 [M+Na] : 418.0660, found 418.0671.

189

3-(furan-2-ylmethyl)-2-(prop-2-yn-1-ylthio)-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-4-one (4.21). To a solution of 4.6 (58.5 mg, 0.192 mmol) in acetonitrile (2 mL) was added 3-bromoprop-1-yne 4.48 (27.4 mg, 0.0.230 mmol), followed by potassium carbonate (26.7 mg, 0.192 mmol). The reaction was left to stir at room temperature for 6 hours, and subsequently concentrated under reduced pressure.

Water (3 mL) was added followed by extraction with ethyl acetate (3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography 10% ethyl acetate in hexanes to afford 4.21 (41.3 mg, 0.121 mmol, 63%), as an off-white solid. 1H NMR (400 MHz,

CDCl3) δ 7.36 (m, 1H), 6.43 (m, 1H), 6.32 (m, 1H), 5.32 (s, 2H), 4.03 (m, 2H), 3.09 – 3.01

(m, 2H), 2.97 – 2.87 (m, 2H), 2.51 – 2.38 (m, 2H), 2.25 (m, 1H). 13C NMR (101 MHz,

CDCl3) δ 166.81, 158.07, 154.07, 148.55, 142.72, 140.45, 137.52, 116.78, 110.65, 110.11,

78.13, 72.13, 40.43, 29.71, 29.10, 28.10, 21.55. HRMS m/z calc’d for C17H14N2NaO2S2

[M+Na]+: 365.0394, found 365.0403.

190

Methyl-2-((3-(furan-2-ylmethyl)-4-oxo-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-2-yl)thio)acetate (4.22). To a solution of 4.6

(78.3 mg, 0.257 mmol) in acetonitrile (2.5 mL) was added methyl 2-bromoacetate 4.49

(39.31 mg, 0.257 mmol), followed by potassium carbonate (42.62 mg, 0.308 mmol). The reaction was left to stir at room temperature for 6 hours, and subsequently concentrated under reduced pressure. Water (3 mL) was added followed by extraction with ethyl acetate

(3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography

20% ethyl acetate in hexanes to afford 4.22 (72.7 mg, 0.193 mmol, 75%), as an off-white

1 solid. H NMR (400 MHz, CDCl3) δ 7.36 (m, 1H), 6.44 (m, 1H), 6.32 (m, 1H), 5.34 (s,

2H), 4.02 (s, 2H), 3.77 (s, 3H), 3.09 – 2.99 (m, 2H), 2.95 – 2.85 (m, 2H), 2.52 – 2.36 (m,

13 2H). C NMR (101 MHz, CDCl3) δ 168.92, 166.63, 157.99, 154.24, 148.54, 142.70,

140.45, 137.42, 116.76, 110.66, 110.11, 53.05, 40.56, 34.74, 29.67, 29.06, 28.08. HRMS

+ m/z calc’d for C17H16N2NaO4S2 [M+Na] : 339.0449, found 399.0459.

191

methyl 3-((3-(furan-2-ylmethyl)-4-oxo-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-2-yl)thio)propanoate (4.23). To a solution of

4.6 (83.1 mg, 0.273 mmol) in acetonitrile (2.7 mL) was added methyl 3-bromopropanoate

4.50 (45.59 mg, 0.273 mmol), followed by potassium carbonate (42.28 mg, 0.328 mmol).

The reaction was left to stir at room temperature for 6 hours, and subsequently concentrated under reduced pressure. Water (3 mL) was added followed by extraction with ethyl acetate

(3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography

20% ethyl acetate in hexanes to afford 4.23 (90.3 mg, 0.231 mmol, 85%), as a white solid.

1 H NMR (400 MHz, CDCl3) δ 7.34 (m, 1H), 6.39 (m, 1H), 6.30 (m, 1H), 5.30 (s, 2H), 3.72

(s, 3H), 3.49 (t, J = 6.9 Hz, 2H), 3.09 – 3.00 (m, 2H), 2.94 – 2.88 (m, 2H), 2.84 (t, J = 6.9

13 Hz, 2H), 2.48 – 2.38 (m, 2H). C NMR (101 MHz, CDCl3) δ 172.28, 167.01, 158.16,

155.23, 148.76, 142.55, 140.42, 137.11, 116.63, 110.57, 109.92, 52.04, 40.33, 33.86,

+ 29.68, 29.10, 28.07, 27.56. HRMS m/z calc’d for C18H18N2NaO4S2 [M+Na] : 413.0606, found 413.0615.

192

3-(furan-2-ylmethyl)-2-methoxy-3,5,6,7-tetrahydro-4H-cyclopenta[4,5]thieno[2,3- d]pyrimidin-4-one (4.24). To a solution of 4.6 (57.3 mg, 0.152 mmol) in methanol:tetrahydrofuran (1:1, 1 mL) was added sodium hydroxide (30.4 mg, 0.761 mmol.

The reaction was left to stir at room temperature for 1 hour, and subsequently concentrated under reduced pressure. Water (3 mL) was added followed by extraction with ethyl acetate

(3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography

10% ethyl acetate in hexanes to afford 4.24 (30.2 mg, 0.1 mmol, 66%), as a white solid. 1H

NMR (400 MHz, CDCl3) δ 7.32 (s, 1H), 6.38 – 6.26 (m, 2H), 5.24 (s, 2H), 4.07 (s, 3H),

13 3.03 (m, 2H), 2.88 (m, 2H), 2.49 – 2.35 (m, 2H). C NMR (101 MHz, CDCl3) δ 166.92,

158.27, 153.32, 149.90, 142.35, 140.22, 134.98, 115.28, 110.54, 109.19, 56.11, 37.43,

+ 29.54, 29.07, 27.98. HRMS m/z calc’d for C15H14N2NaO3S [M+Na] : 325.0623, found

325.0630.

193

3-((3-(furan-2-ylmethyl)-4-oxo-3,5,6,7-tetrahydro-4H-cyclopenta[4,5]thieno[2,3- d]pyrimidin-2-yl)thio)propanenitrile (4.25). To a solution of 4.6 (361.1 mg, 1.186 mmol) in acetonitrile (10 mL) was added 3-bromopropanenitrile 4.51 (158.95 mg, 1.186 mmol), followed by potassium carbonate (327.83 mg, 2.372 mmol). The reaction was left to stir at 50 °C for 3 hours, and subsequently concentrated under reduced pressure. Water

(3 mL) was added followed by extraction with ethyl acetate (3 mL x 3), washing with brine

(10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography 25% ethyl acetate in hexanes to afford

1 4.25 (247.1 mg, 0.231 mmol, 69%), as a white solid. H NMR (400 MHz, CDCl3) δ 7.36

(m, 1H), 6.43 (m, 1H), 6.32 (m, 1H), 5.31 (s, 2H), 3.47 (t, J = 6.9 Hz, 2H), 3.09 – 3.01 (m,

13 2H), 2.93 (m, 4H), 2.45 (m, 2H). C NMR (101 MHz, CDCl3) δ 166.52, 157.94, 153.82,

148.40, 142.75, 140.57, 137.71, 118.06, 116.98, 110.69, 110.23, 40.43, 29.70, 29.08,

+ 28.10, 28.08, 18.32. HRMS m/z calc’d for C17H15N3NaO2S2 [M+Na] : 380.0503, found

380.0513.

194

2-((2-(1H-tetrazol-5-yl)ethyl)thio)-3-(furan-2-ylmethyl)-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-4-one (4.26). To a solution of 4.25 (34.2 mg,

0.096 mmol) in butanol (1 mL) was added sodium azide (7.5 mg, 0.115 mmol), followed by zinc bromide (21.55 mg, 0.096 mmol). The reaction was left to stir at 1100 °C for 4 hours, and subsequently concentrated under reduced pressure. Water (3 mL) was added followed by extraction with ethyl acetate (3 mL x 3), washing with brine (10 mL) and drying over sodium sulfate. Upon concentration, the crude material was purified by normal phase silica gel chromatography 50% ethyl acetate in hexanes to afford 4.26 (12.7 mg,

1 0.032 mmol, 33%), as a yellow solid. H NMR (400 MHz, CDCl3) δ 7.35 (m, 1H), 6.43

(m, 1H), 6.30 (m, 1H), 5.18 (s, 2H), 3.74 – 3.67 (m, 2H), 3.61 (m, 2H), 2.94 (t, J = 7.1 Hz,

13 2H), 2.86 (t, J = 7.3 Hz, 2H), 2.49 – 2.37 (m, 2H). C NMR (101 MHz, CDCl3) δ 167.13,

158.13, 154.49, 147.83, 143.07, 139.86, 138.16, 116.37, 110.82, 110.81, 40.57, 30.87,

29.84, 29.65, 29.04, 27.92, 23.27.

195

2-((2-(1,3-dioxan-2-yl)ethyl)thio)-3-(furan-2-ylmethyl)-3,5,6,7-tetrahydro-4H- cyclopenta[4,5]thieno[2,3-d]pyrimidin-4-one (4.27). To a solution of 4.6 (38.6 mg, 0.127 mmol) in acetonitrile (1.5 mL) was added 2-(2-bromoethyl)-1,3-dioxane 4.52 (49.35 mg,

0.253 mmol), followed by potassium carbonate (35.05 mg, 0.253 mmol). The reaction was left to stir at 50 °C for 4 hours, and subsequently concentrated under reduced pressure.

1 afford 4.27 (42.2 mg, 0.101 mmol, 79%), as a white solid. H NMR (400 MHz, CDCl3) δ

7.34 (m, 1H), 6.39 (m, 1H), 6.31 (m, 1H), 5.32 (s, 2H), 4.68 (t, J = 5.1 Hz, 1H), 4.15 – 4.05

(m, 2H), 3.77 (m, 2H), 3.32 (t, J = 7.2 Hz, 2H), 3.10 – 3.00 (m, 2H), 2.99 – 2.84 (m, 2H),

13 2.46 – 2.39 (m, 2H), 2.15 – 1.97 (m, 4H). C NMR (101 MHz, CDCl3) δ 167.20, 158.22,

155.76, 148.94, 142.43, 140.32, 136.87, 129.72, 127.23, 116.45, 110.54, 109.73, 100.74,

40.33, 34.15, 29.64, 29.08, 28.02, 27.47, 25.84. HRMS m/z calc’d for C20H22N2NaO4S2

[M+Na]+: 441.0919, found 441.0925.

196

Synthesis of Azide biotin linker

N-(2-(2-(2-(2-azidoethoxy)ethoxy)ethoxy)ethyl)-5-((3aS,4S,6aR)-2-oxohexahydro-

1H-thieno[3,4-d]imidazol-4-yl)pentanamide (4.31). The azide biotin linker 4.31 was prepared according to the literature procedure.158 A solution of biotin (250 mg, 1.02 mmol) in DMF (5 mL) was heated to 70ºC for 10 minutes and allowed to cool to room temperature.

Triethylamine (181.2 mg, 1.79 mmol) was added followed by pentafluorophenyl trifluoroacetate (405.60 mg, 1.45 mmol). The reaction was left to stir at room temperature for 30 minutes at which point it became pink. The reaction was concentrated under vacuum and the resulting residue was triturated with diethyl ether (10 mL) to afford biotin-pFp ester

(382 mg, 0.931 mmol) as a white solid that was immediately carried forward without further purification. To a flask containing 11-azido-3,6,9-trioxaundecan-1-amine (203.5 mg, 1.02 mmol) and triethylamine (206.4 mg, 2.04 mmol) in DMF (13.3 mL) cooled to

0ºC was added a cooled solution of biotin-pFp ester in DMF (9 mL) dropwise. The reaction was left to stir at room temperature for 1 hour, then concentrated under vacuum. The resulting residue was triturated with diethyl ether (10 mL x 2) to afford azide biotin linker

4.31 (275 mg, 0.62 mmol, 66%) as a white solid that was carried forward without further

197

purification. 1H NMR (400 MHz, MeOD) δ 4.49 (dd, J = 7.9, 4.3 Hz, 1H), 4.31 (dd, J =

7.9, 4.5 Hz, 1H), 3.70 – 3.64 (m, 8H), 3.64 – 3.60 (m, 2H), 3.55 (t, J = 5.5 Hz, 2H), 3.40 –

3.34 (m, 4H), 3.24 – 3.18 (m, 1H), 2.93 (dd, J = 12.7, 5.0 Hz, 1H), 2.71 (d, J = 12.7 Hz,

1H), 2.22 (t, J = 7.4 Hz, 2H), 1.79 – 1.55 (m, 4H), 1.45 (p, J = 7.4 Hz, 2H). 13C NMR (101

MHz, MeOD) δ 176.14, 166.11, 71.67, 71.63, 71.53, 71.27, 71.13, 70.60, 63.37, 61.63,

56.99, 51.79, 41.04, 40.37, 36.73, 29.75, 29.50, 26.84. HRMS m/z calc’d for

+ C18H32N6NaO5S [M+Na] : 467.2047, found 467.2040.

198

Synthesis of T315-S1 Probe

N-methyl-3-(1-(4-(4-(prop-2-yn-1-yl)piperazin-1-yl)phenyl)-5-(4'-(trifluoromethyl)-

[1,1'-biphenyl]-4-yl)-1H-pyrazol-3-yl)propenamide (4.30). Synthesis of T315-S1 Probe

(4.32) began from a known intermediate OSU-T315 prepared according to literature procedure.156 To a solution of T315 4.28 (31.1 mg, 0.06 mmol) in THF (1 mL) was added propargyl bromide (7.9 mg, 0.066 mmol) followed by diisopropylethylamine (8.53 mg,

0.066 mmol). The reaction was left to stir at room temperature for 4 hours and then concentrated under vacuum. Resulting reside was purified by normal phase silica gel chromatography (EtOAc:CH3CN:MeOH , 7:2.5:0.5, 0.1% Et3N) to afford 4.30 (19 mg,

0.033 mmol, 55%) as an off-white solid. 1H NMR (400 MHz, MeOD) δ 7.94 – 7.89 (m,

2H), 7.85 (d, J = 8.1 Hz, 2H), 7.76 – 7.71 (m, 4H), 7.40 – 7.34 (m, 2H), 7.14 – 7.08 (m,

2H), 6.69 (s, 1H), 3.39 (m, 2H), 3.35 – 3.32 (m, 4H), 2.94 (t, J = 7.5 Hz, 2H), 2.81 – 2.75

(m, 4H), 2.72 (m, 1H), 2.70 (s, 3H), 2.52 (t, J = 7.5 Hz, 2H). 13C NMR (101 MHz, MeOD)

199

δ 174.71, 152.78, 152.04, 146.15, 140.18, 134.43, 132.15, 128.52, 128.38, 128.13, 127.38,

126.82, 126.78, 117.08, 103.63, 78.96, 75.28, 52.77, 35.50, 26.35, 23.13. HRMS m/z

+ calc’d for C33H33F3N5O [M+H] : 572.2637, found 572.2640.

N-(2-(2-(2-(2-(4-((4-(4-(3-(3-(methylamino)-3-oxopropyl)-5-(4'-(trifluoromethyl)-

[1,1'-biphenyl]-4-yl)-1H-pyrazol-1-yl)phenyl)piperazin-1-yl)methyl)-1H-1,2,3- triazol-1-yl)ethoxy)ethoxy)ethoxy)ethyl)-5-((3aS,4S,6aR)-2-oxohexahydro-1H- thieno[3,4-d]imidazol-4-yl)pentanamide (4.32). Azide biotin linker 4.31 (8.4 mg, 0.02 mmol) was added to a solution of T315 alkyne 4.30, (10.9 mg, 0.02 mmol) in CH3CN (1 mL), followed by copper iodide (0.4 mg, 0.0022 mmol) and diisopropylethylamine (2.8 mg, 0.022 mmol). The reaction was left to stir at 40ºC for 12 hours under nitrogen. Upon completion, the reaction was filtered over a pad of Celite and concentrated. The residue

200

was taken up in a small amount of acetonitrile and purified by normal phase silica gel chromatography (EtOAc:CH3CN:MeOH, 3:6.5:0.5, 0.2% Et3N). Pooled fractions were collected and concentrated under vacuum to afford 4.32 as a clear film (11.3 mg, 0.01 mmol, 56%). 1H NMR (300 MHz, MeOD) δ 8.76 (s, 1H), 7.96 – 7.84 (m, 5H), 7.78 – 7.71

(m, 4H), 7.43 (m, 2H), 7.18 (d, J = 8.7 Hz, 2H), 6.71 (s, 1H), 3.99 (t, J = 6.6 Hz, 2H), 3.64

(m, 6H), 3.54 (m, 2H), 3.50 – 3.35 (m, 11H), 3.18 (m, 3H), 3.07 (m, 1H), 2.98 – 2.93 (m,

2H), 2.70 (s, 3H), 2.56 – 2.50 (m, 2H), 1.64 (m, 10H), 1.14 – 1.07 (m, 4H). HRMS m/z

+ calc’d for C51H64F3N11NaO6S [M+Na] : 1038.4612, found 1038.4670.

201

Synthesis of T315-S2 Probe

Ethyl 3-(1-(4-aminophenyl)-5-(4'-(trifluoromethyl)-[1,1'-biphenyl]-4-yl)-1H-pyrazol-

3-yl)propanoate (4.39). Synthesis of T-315 S2 began from a known intermediate ethyl 3-

(1-(4-nitrophenyl)-5-(4'-(trifluoromethyl)-[1,1'-biphenyl]-4-yl)-1H-pyrazol-3- yl)propanoate (4.38) that was prepared following literature procedure.156 To a solution of

4.38 (464.4 mg, 0.91 mmol) in MeOH:EtOAc (1:3, 10 mL) was added Pd/C (14 mg). The reaction was stirred under H2 (40 psi, room temperature) for 1 hour. The reaction mixture was filtered over Celite and concentrated to afford 4.39 (426.1 mg, 0.89 mmol, 98%) as a brown foam that was carried forward without further purification. 1H NMR (400 MHz,

CDCl3) δ 7.96 (d, J = 8.5 Hz, 2H), 7.73 (m, 4H), 7.66 (d, J = 8.5 Hz, 2H), 7.27 (d, J = 8.0

Hz, 2H), 6.77 (d, J = 8.7 Hz, 2H), 6.56 (s, 1H), 4.16 (q, J = 7.1 Hz, 2H), 2.99 (t, J = 7.6

13 Hz, 2H), 2.62 (t, J = 7.6 Hz, 2H), 1.23 (t, J = 7.1 Hz, 3H). C NMR (176 MHz, CDCl3) δ

172.14, 150.84, 144.32, 144.02, 139.17, 127.58, 127.38, 127.30, 127.10, 126.40, 125.85,

202

+ 125.83, 102.92, 60.93, 33.19, 21.79, 14.34. HRMS m/z calc’d for C27H25F3N3O2 [M +H] :

480.1899, found 480.1869.

tert-Butyl 4-(4-(3-(3-ethoxy-3-oxopropyl)-5-(4'-(trifluoromethyl)-[1,1'-biphenyl]-4- yl)-1H-pyrazol-1-yl)phenyl)piperazine-1-carboxylate (4.40). To a flask containing a solution of 4.39 (517.9 mg, 1.08 mmol) in xylenes (11 mL) was added bis(2- chloroethyl)amine (184.1 mg, 1.30 mmol). After heating at 130ºC for 72 hours, the reaction was cooled to room temperature and concentrated under vacuum to afford ethyl

3-(1-(4-(piperazin-1-yl)phenyl)-5-(4'-(trifluoromethyl)-[1,1'-biphenyl]-4-yl)-1H-pyrazol-

3-yl)propanoate HRMS m/z calc’d for C31H32F3N4O2 [M+H]+: 549.2477, found 549.5000.

The resulting residue (75.5 mg, 0.14 mmol) was taken up in methylene chloride (13.8 mL) followed by the addition of di-tert-butyl dicarbonate (31.5 mg, 0.144 mmol), then 4- dimethylaminopyridine (1.67 mg, 0.014 mmol). The reaction was left to stir at room temperature for 3 hours and then concentrated under vacuum. The residue was purified by normal phase silica gel chromatography (1% - 5% MeOH in CH2Cl2, 0.1% Et3N) to afford

1 4.40 as a colorless foam (47.4 mg, 0.073 mmol, 53%). H NMR (400 MHz, CDCl3) δ 7.94 203

(d, J = 8.4 Hz, 2H), 7.71 (m, 4H), 7.63 (d, J = 8.4 Hz, 2H), 7.39 (d, J = 8.8 Hz, 2H), 7.03

(d, J = 8.6 Hz, 2H), 6.55 (s, 1H), 4.14 (q, J = 7.1 Hz, 2H), 3.67 – 3.58 (m, 4H), 3.26 – 3.17

(m, 4H), 2.99 (t, J = 7.6 Hz, 2H), 2.65 (t, J = 7.6 Hz, 2H), 1.50 (s, 9H), 1.24 (t, J = 7.1 Hz,

13 3H). C NMR (101 MHz, CDCl3) δ 172.25, 154.79, 143.70, 138.92, 133.42, 127.53,

127.31, 126.81, 126.33, 125.84, 125.80, 102.58, 80.21, 60.86, 33.33, 28.56, 21.80, 14.33.

+ HRMS m/z calc’d for C36H40F3N4O4 [M+H] : 649.3002, found 649.2995.

tert-Butyl 4-(4-(3-(3-oxo-3-(prop-2-yn-1-ylamino)propyl)-5-(4'-(trifluoromethyl)-

[1,1'-biphenyl]-4-yl)-1H-pyrazol-1-yl)phenyl)piperazine-1-carboxylate (4.37). To a solution of 4.40 (25.1 mg, 0.039 mmol) in a solution of THF:H2O (3:1, 1 mL) was added

3 M NaOH (4.64 mg, 0.116 mmol). The reaction was left to stir at room temperature for 3 hours. TLC indicated complete hydrolysis of ester. The reaction was concentrated to remove THF. Water (3mL) was added and the solution was extracted with ethyl acetate (4

204

ml x 3), dried over sodium sulfate, filtered and concentrated to afford 3-(1-(4-(4-(tert- butoxycarbonyl)piperazin-1-yl)phenyl)-5-(4'-(trifluoromethyl)-[1,1'-biphenyl]-4-yl)-1H- pyrazol-3-yl)propanoic acid. The residue was immediately dissolved in anhydrous acetonitrile followed by the addition of pentafluorophenyl trifluoroacetate (12.01 mg,

0.043 mmol) and diisopropylethylamine (5.56 mg, 0.043mmol). The reaction was allowed to stir at room temperature for 2 hours, followed by the dropwise addition of propargyl amine (2.37 mg, 0.043 mmol). After 2 hours, the reaction was concentrated and purified by normal phase chromatography (0.5% - 1% MeOH in CH2Cl2, 0.1% Et3N) to afford 4.37

1 (20.2 mg, 0.031 mmol, 78%). as a colorless foam. H NMR (300 MHz, CDCl3) δ 7.95 –

7.89 (m, 2H), 7.66 (m, 7H), 7.41 (d, J = 8.9 Hz, 2H), 7.13 (d, J = 7.5 Hz, 2H), 6.56 (s, 1H),

4.03 (m, 2H), 3.73 – 3.61 (m, 4H), 3.28 – 3.20 (m, 4H), 3.06 – 2.98 (m, 2H), 2.56 – 2.46

13 (m, 2H), 2.21 (m, 1H), 1.50 (s, 9H). C NMR (75 MHz, CDCl3) δ 170.65, 154.57, 150.78,

144.23, 143.79, 138.99, 133.01, 127.43, 127.18, 126.72, 126.26, 126.10, 125.73, 125.68,

102.83, 80.35, 79.32, 77.44, 77.22, 77.02, 76.59, 71.76, 34.93, 29.27, 28.41, 21.89. HRMS

+ m/z calc’d for C37H39F3N5O3 [M+H] : 658.3000, found 658.3014.

205

tert-Butyl 4-(4-(3-(3-oxo-3-(((1-(13-oxo-17-((3aS,4S,6aR)-2-oxohexahydro-1H- thieno[3,4-d]imidazol-4-yl)-3,6,9-trioxa-12-azaheptadecyl)-1H-1,2,3-triazol-4- yl)methyl)amino)propyl)-5-(4'-(trifluoromethyl)-[1,1'-biphenyl]-4-yl)-1H-pyrazol-1- yl)phenyl)piperazine-1-carboxylate) (4.41). To a solution of 4.37 (11.3 mg, 0.017 mmol) in CH3CN (1 mL) stirring under argon was added biotin azide 4.31 (7.6 mg, 0.017 mmol), copper iodide (0.36 mg, 0.0019 mmol), followed by diisopropylethylamine (2.4 mg, 0.019 mmol. The reaction was left to stir at 40 ºC for 2 hours then filtered over Celite and concentrated under vacuum. The residue was purified by normal phase silica gel chromatography (EtOAc:CH3CN:MeOH, 5:4:1, 0.1% Et3N) to afford 4.41 (13.3 mg, 0.012 mmol, 71%) as a light brown film. 1H NMR (400 MHz, MeOD) δ 7.89 (s, 1H), 7.82 (d, J

206

= 8.2 Hz, 2H), 7.73 (d, J = 8.4 Hz, 2H), 7.65 (d, J = 8.5 Hz, 2H), 7.33 (d, J = 8.4 Hz, 2H),

7.15 (d, J = 9.0 Hz, 2H), 6.99 (d, J = 9.0 Hz, 2H), 6.46 (s, 1H), 4.48 – 4.44 (m, 4H), 4.27

(m, 1H), 3.84 – 3.79 (m, 2H), 3.57 (m, 10H), 3.50 (m, 2H), 3.20 (m, 6H), 3.04 (t, J = 7.2

Hz, 2H), 2.91 (m, 1H), 2.67 (m, 3H), 2.18 (m, 2H), 1.74 – 1.52 (m, 8H), 1.48 (s, 9H), 1.42

+ (m, 4H), 1.11 (t, J = 7.3 Hz, 2H). HRMS m/z calc’d for C55H70F3N11NaO8S [M+Na] :

1124.4979, found 1124.4975.

5-((3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl)-N-(2-(2-(2-(2-(4-((3-

(1-(4-(piperazin-1-yl)phenyl)-5-(4'-(trifluoromethyl)-[1,1'-biphenyl]-4-yl)-1H- pyrazol-3-yl)propanamido)methyl)-1H-1,2,3-triazol-1- yl)ethoxy)ethoxy)ethoxy)ethyl)pentanamide (4.36). To a vial containing 4.41 (13 mg,

0.012 mmol) was added a 20% solution of trifluoroacetic acid in methylene chloride (0.5 mL) slowly at 0ºC. The reaction was allowed to gradually warm up to room temperature

207

and left to stir for 3 hours. The reaction was then concentrated to dryness and the resulting residue was taken up in a minimal amount of methylene chloride and purified by normal phase silica gel chromatography (EtOAc:CH3CN:MeOH, 3:6.5:0.5, 0.2% Et3N) to afford

T315-S2 Probe, 4.36 (6.6 mg, 0.007 mmol 55%), as an off-white film. 1H NMR (400 MHz,

MeOD) δ 7.88 (s, 1H), 7.81 (d, J = 8.2 Hz, 2H), 7.73 (d, J = 8.4 Hz, 2H), 7.64 (d, J = 8.2

Hz, 2H), 7.33 (d, J = 8.3 Hz, 2H), 7.19 (d, J = 8.9 Hz, 1H), 7.16 – 7.11 (m, 1H), 7.03 (d, J

= 8.9 Hz, 1H), 7.01 – 6.94 (m, 1H), 6.45 (s, 1H), 4.50 – 4.42 (m, 5H), 4.27 (m, 1H), 3.81

(m, 2H), 3.56 (m, 7H), 3.52 – 3.47 (m, 3H), 3.44 – 3.40 (m, 2H), 3.19 (m, 3H), 3.03 (t, J =

7.5 Hz, 2H), 2.89 (m, 1H), 2.66 (m, 3H), 2.18 (t, J = 7.3 Hz, 2H), 1.76 – 1.35 (m, 10H),

1.17 (t, J = 7.0 Hz, 2H), 1.10 (t, J = 7.3 Hz, 2H). HRMS m/z calc’d for C50H62F3N11NaO6S

[M+Na]+: 1024.4455, found 1024.4470.

208

References

(1) Godula, K.; Sames, D. Science. 2006, 312, 67-72.

(2) Nicolaou, K. C.; Bulger, P. G.; Sarlah, D. Angew. Chemie - Int. Ed. 2005, 44,

4442–4489.

(3) Handbook of Organopalladium Chemistry for Organic Synthesis; Negishi, E., Ed.;

Wiley Interscience: New York, 2002.

(4) Masters, J. J.; Link, J. T.; Snyder, L. B.; Young, W. B.; Danishefsky, S. J. Angew.

Chemie 2018, 107, 1886–1888.

(5) Shair, M. D.; Yoon, T. Y.; Mosny, K. K.; Chou, T. C.; Danishefsky, S. J. J. Am.

Chem. Soc. 1996, 118, 9509–9525.

(6) Nicolaou, K. C.; Yang, Z.; Shi, G.; Gunzner, J. L.; Agrios, K. A.; Gärtner, P.

Nature 1998, 392, 264-269.

(7) Yamaguchi, J.; Yamaguchi, A. D.; Itami, K. Angew. Chemie - Int. Ed. 2012, 51,

8960–9009.

(8) Hutchison, A. J.; Kishi, Y. J. Am. Chem. Soc. 1979, 101, 6786–6788.

(9) Baran, P. S.; Corey, E. J. J. Am. Chem. Soc. 2002, 124, 7904–7905.

(10) Mandal, D.; Yamaguchi, A. D.; Yamaguchi, J.; Itami, K. J. Am. Chem. Soc. 2011,

133, 19660–19663.

(11) Garg, N. K.; Sarpong, R.; Stoltz, B. M. J. Am. Chem. Soc. 2002, 124, 13179–

13184.

(12) Ohyabu, N.; Nishikawa, T.; Isobe, M. J. Am. Chem. Soc. 2003, 125, 8798–8805.

209

(13) Cernak, T.; Dykstra, K. D.; Tyagarajan, S.; Vachal, P.; Krska, S. W. Chem. Soc.

Rev. 2016, 45, 546–576.

(14) Dai, H. X.; Stepan, A. F.; Plummer, M. S.; Zhang, Y. H.; Yu, J. Q. J. Am. Chem.

Soc. 2011, 133, 7222–7228.

(15) Seiple, I. B.; Su, S.; Rodriguez, R. A.; Gianatassio, R.; Fujiwara, Y.; Sobel, A. L.;

Baran, P. S. J. Am. Chem. Soc. 2010, 132, 13194–13196.

(16) Johansson, C. C. C.; Bremeyer, N.; Ley, S. V; Owen, D. R.; Smith, S. C.; Gaunt,

M. J. Angew. Chemie Int. Ed. 2006, 45, 6024–6028.

(17) Hintermann, L.; Schmitz, M.; Englert, U. Angew. Chemie Int. Ed. 2007, 46, 5164–

5167.

(18) Jaric, M.; Haag, B. A.; Manolikakes, S. M.; Knochel, P. Org. Lett. 2011, 13, 2306–

2309.

(19) Ji, Y.; Brueckl, T.; Baxter, R. D.; Fujiwara, Y.; Seiple, I. B.; Su, S.; Blackmond,

D. G.; Baran, P. S. Proc. Natl. Acad. Sci. 2011, 108, 14411-14415.

(20) Lockley, W. J. S. J. Label. Compd. Radiopharm. 2007, 50, 779–788.

(21) Ray, T.; Wu, A.; Allentoff, A. J. J. Label. Compd. Radiopharm. 2007, 50, 433–

434.

(22) Huang, X.; Liu, W.; Ren, H.; Neelamegam, R.; Hooker, J. M.; Groves, J. T. J. Am.

Chem. Soc. 2014, 136, 6842–6845.

(23) O’Hara, F.; Burns, A. C.; Collins, M. R.; Dalvie, D.; Ornelas, M. A.; Vaz, A. D.

N.; Fujiwara, Y.; Baran, P. S. J. Med. Chem. 2014, 57, 1616–1620.

210

(24) Alfaro, J. F.; Jones, J. P. J. Org. Chem. 2008, 73, 9469–9472.

(25) Kanoh, N. Nat. Prod. Rep. 2016, 33, 709–718.

(26) Peddibhotla, S.; Dang, Y.; Liu, J. O.; Romo, D. J. Am. Chem. Soc. 2007, 129,

12222–12231.

(27) Li, J.; Cisar, J. S.; Zhou, C.-Y.; Vera, B.; Williams, H.; Rodríguez, A. D.; Cravatt,

B. F.; Romo, D. Nat. Chem. 2013, 5, 510-517.

(28) Yan, M.; Lo, J. C.; Edwards, J. T.; Baran, P. S. J. Am. Chem. Soc. 2016, 138,

12692–12714.

(29) Zhou, Q.; Gui, J.; Pan, C. M.; Albone, E.; Cheng, X.; Suh, E. M.; Grasso, L.;

Ishihara, Y.; Baran, P. S. J. Am. Chem. Soc. 2013, 135, 12994–12997.

(30) Donehower, R. C. Oncologist 1996, 1, 240–243.

(31) Khanna, C.; Rosenberg, M.; Vail, D. M. A. J. Vet. Intern. Med. 2015, 29, 1006–

1012.

(32) Pommier, Y. Nat. Rev. Cancer 2006, 6, 789-802.

(33) Li, Q. Y.; Zu, Y. G.; Shi, R. Z.; Yao, L. P. Curr. Med. Chem. 2006, 13, 2021–

2039.

(34) Hande, K. R. Eur. J. Cancer 1998, 34, 1514–1521.

(35) Moudi, M.; Go, R.; Yien, C. Y. S.; Nazre, M. Int. J. Prev. Med. 2013, 4, 1231–

1235.

(36) Demain, A. L.; Vaishnav, P. Microb. Biotechnol. 2011, 4, 687–699.

(37) Cunha, W. R.; Silva, M. L.; Sola Veneziani, R. C.; Ambrósio, S. R.; Astos, J. K.

211

Phytochemicals - A Global Perspective of Their Role in Nutrition and Health; Rao,

V., Ed.; IntechOpen, 2012; 213–234.

(38) Suzuki, S.; Umezawa, T. J. Wood Sci. 2007, 53, 273–284.

(39) Teponno, R. B.; Kusari, S.; Spiteller, M. Nat. Prod. Rep. 2016, 33, 1044–1092.

(40) Zhang, Y.-J.; Abe, T.; Tanaka, T.; Yang, C.-R.; Kouno, I. J. Nat. Prod. 2001, 64,

1527–1532.

(41) Ishimaru, K.; Yoshimatsu, K.; Yamakawa, T.; Kamada, H.; Shimomura, K.

Phytochemistry 1992, 31, 2015–2018.

(42) Houghton, P. J.; Woldemariam, T. Z.; O’Shea, S.; Thyagarajan, S. P.

Phytochemistry 1996, 43, 715–717.

(43) Huang, Y.-L.; Chen, C.-C.; Hsu, F.-L.; Chen, C.-F.. J. Nat. Prod. 1998, 61, 1194–

1197.

(44) Youkwan, J.; Srisomphot, P.; Sutthivaiyakit, S. J. Nat. Prod. 2005, 68, 1006–1009.

(45) Tanaka, R.; Matsunaga, S. Phytochemistry 1988, 27, 2273–2277.

(46) Sutthivaiyakit, S.; Nakorn, N. N.; Kraus, W.; Sutthivaiyakit, P. Tetrahedron 2003,

59, 9991–9995.

(47) Bachmann, T. L.; Ghia, F.; Torssell, K. B. G. Phytochemistry 1993, 33, 189–191.

(48) Lin, M.-T.; Lee, S.-S.; Liu, K. C. S. C. J. Nat. Prod. 1995, 58, 244–249.

(49) Kan, W. S. Manual of Medicinal Plants in Taiwan; Republic of China, 1985; Vol.

(50) Hsieh, C. F.; Chaw, S. M.; Wang, J. C. Flora of Taiwan, 2nd ed.; Taiwan, E. C. of

212

the F. of, Ed.; Taipei, 1993.

(51) Wu, S.-J.; Wu, T.-S. Chem. Pharm. Bull. 2006, 54, 1223–1225.

(52) Ren, Y.; Lantvit, D. D.; Deng, Y.; Kanagasabai, R.; Gallucci, J. C.; Ninh, T. N.;

Chai, H.-B.; Soejarto, D. D.; Fuchs, J. R.; Yalowich, J. C.; et al. J. Nat. Prod.

2014, 77, 1494–1504.

(53) Poręba, M.; Strózyk, A.; Salvesen, G. S.; Drag, M. Cold Spring Harb. Perspect.

Biol. 2013, 5, a008680.

(54) Li, P.; Nijhawan, D.; Wang, X. Cell 2004, S116, S57–S59.

(55) Joseph, B.; Ekedahl, J.; Lewensohn, R.; Marchetti, P.; Formstecher, P.;

Zhivotovsky, B. Oncogene 2001, 20, 2877-2888.

(56) Deng, Y.; Chu, J.; Ren, Y.; Fan, Z.; Ji, X.; Mundy, B.; Yuan, S.; Hughes, T.;

Zhang, J.; Cheema, B.; et al. J. Immunol. 2014, 193, 2994–3002.

(57) Huang, X.; Xue, J. J. Org. Chem. 2007, 72, 3965–3968.

(58) Vasilev, N.; Elfahmi; Bos, R.; Kayser, O.; Momekov, G.; Konstantinov, S.;

Ionkova, I. J. Nat. Prod. 2006, 69, 1014–1017.

(59) Yeo, H.; Li, Y.; Fu, L.; Zhu, J.-L.; Gullen, E. A.; Dutschman, G. E.; Lee, Y.;

Chung, R.; Huang, E.-S.; Austin, D. J.; et al. J. Med. Chem. 2005, 48, 534–546.

(60) Parmar, D.; Sugiono, E.; Raja, S.; Rueping, M. Chem. Rev. 2014, 114, 9047–9153.

(61) Brunel, J. M. Chem. Rev. 2005, 105, 857–898.

(62) Patel, R. M.; Argade, N. P. Org. Lett. 2013, 15, 14–17.

(63) Kim, H. Y.; Oh, K. Org. Lett. 2014, 16, 5934–5936.

213

(64) Shahzad, S. A.; Vivant, C.; Wirth, T. Org. Lett. 2010, 12, 1364–1367.

(65) Naresh, G.; Kant, R.; Narender, T. Org. Lett. 2015, 17, 3446–3449.

(66) Larock, R. C.; Doty, M. J.; Tian, Q.; Zenner, J. M. J. Org. Chem. 1997, 62, 7536–

7537.

(67) Asao, N.; Takahashi, K.; Lee, S.; Kasahara, T.; Yamamoto, Y. J. Am. Chem. Soc.

2002, 124, 12650–12651.

(68) Viswanathan, G. S.; Wang, M.; Li, C.-J. Angew. Chem., Int. Ed. 2002, 41, 2138–

2141.

(69) Balamurugan, R.; Gudla, V. Org. Lett. 2009, 11, 3116–3119.

(70) Ponra, S.; Vitale, M. R.; Michelet, V.; Ratovelomanana-Vidal, V. J. Org. Chem.

2015, 80, 3250–3257.

(71) Charlton, J. L.; Oleschuk, C. J.; Chee, G.-L. J. Org. Chem. 1996, 61, 3452–3457.

(72) Woodard, J. L.; Huntsman, A. C.; Patel, P. A.; Chai, H.-B.; Kanagasabai, R.;

Karmahapatra, S.; Young, A. N.; Ren, Y.; Cole, M. S.; Herrera, D.; et al. Bioorg.

Med. Chem. 2018, 26, 2354–2364.

(73) Cotter, J. L.; Andrews, L. J.; Keefer, R. M. J. Am. Chem. Soc. 1962, 84, 793–797.

(74) Izquierdo, S.; Essafi, S.; del Rosal, I.; Vidossich, P.; Pleixats, R.; Vallribera, A.;

Ujaque, G.; Lledós, A.; Shafir, A. J. Am. Chem. Soc. 2016, 138, 12747–12750.

(75) Smith, B. R.; Eastman, C. M.; Njardarson, J. T. J. Med. Chem. 2014, 57, 9764-

9773.

(76) Tang, M. L.; Bao, Z. Chem. Mater. 2011, 23, 446–455.

214

(77) Tang, M. L.; Oh, J. H.; Reichardt, A. D.; Bao, Z. J. Am. Chem. Soc. 2009, 10,

3733–3740.

(78) Schlyer, D. J. Ann. Acad. Med. Singapore 2004, 33, 146–154.

(79) Kress, T. J.; Costantino, S. M. J. Heterocycl. Chem. 2018, 10, 409–410.

(80) Iwai, T.; Sawamura, M. ACS Catal. 2015, 5, 5031-5040.

(81) Kelley, B. T.; Walters, J. C.; Wengryniuk, S. E. Org. Lett. 2016, 18, 1896–1899.

(82) Boursalian, G. B.; Ham, W. S.; Mazzotti, A. R.; Ritter, T. Nat. Chem. 2016, 8,

810–815.

(83) (a) Silver, L. L. Clin. Microbiol. Rev. 2011, 24, 71–109; (b) Zhanel, G. G.; Walkty,

A.; Vercaigne, L.; Karlowsky, J. A.; Embil, J.; Gin, A. S.; Hoban, D. J. Can. J.

Infect. Dis. 1999, 10, 207-238.

(84) Urahn, Susan, K.; Coukell, A.; Jungman, E. A Scientific Roadmap for Antibiotic

Discovery; 2016. (b) O’Neill, J. The Review on Antimicrobial Resistance, chaired

by Jim O’Neill. Antimicrobial Resistance: Tackling a crisis for the health and

wealth of nations; 2014.

(85) Fleming, A. 1929. Br. J. Exp. Pathol. 1929, 10, 226–236.

(86) Comroe, J. H. J. Am. Rev. Respir. Dis. 1978, 117, 773–781.

(87) Brown, E. D.; Wright, G. D. Nature 2016, 529, 336-343.

(88) Hancock, R. E. W.; Bell, A. Eur. J. Clin. Microbiol. Infect. Dis. 1988, 7, 713–720.

(89) Lepore, B. W.; Indic, M.; Pham, H.; Hearn, E. M.; Patel, D. R.; van den Berg, B.

Proc. Natl. Acad. Sci. 2011, 108, 10121–10126.

215

(90) Eren, E.; Parkin, J.; Adelanwa, A.; Cheneke, B.; Movileanu, L.; Khalid, S.; Van

Den Berg, B. J. Biol. Chem. 2013, 288, 12042–12053.

(91) Ceccarelli, M.; Danelon, C.; Laio, A.; Parrinello, M. Biophys. J. 2004, 87, 58–64.

(92) Weingart, H.; Petrescu, M.; Winterhalter, M. Curr. Drug Targets 2008, 9, 789–

796.

(93) Nikaido, H.; Thanassi, D. G. Antimicrob. Agents Chemother. 1993, 37, 1393-1399.

(94) Tommasi, R.; Brown, D. G.; Walkup, G. K.; Manchester, J. I.; Miller, A. A. Nat.

Rev. Drug Discov. 2015, 14, 529-542.

(95) O’Shea, R.; Moser, H. E. J. Med. Chem. 2008, 51, 2871–2878.

(96) Silver, L. L. A. Bioorganic Med. Chem. 2016, 24, 6379–6389. For OM

penetration of hydrophobic agents see: Zgurskaya, H. I.; Lopez, C. A.;

Gnanakaran, S. ACS Infect. Dis. 2015, 1, 512–522; Delcour, A. H. Biochim

Biophys Acta. 2009, 1794, 808–816.

(97) Trias, J.; Nikaido, H. Antimicrob. Agents Chemother. 1990, 34, 52–57.

(98) Trias, J.; Nikaido, H. J. Biol. Chem. 1990, 265, 15680–15684.

(99) Zgurskaya, H. I.; López, C. A.; Gnanakaran, S. ACS Infect. Dis. 2015, 1, 512–522.

(100) Nikaido, H.; Pages, J.-M. FEMS Microbiol. Rev. 2012, 36, 340–363.

(101) Li, X.-Z.; Plésiat, P.; Nikaido, H. Clin. Microbiol. Rev. 2015, 28, 337-418.

(102) Payne, D. J.; Gwynn, M. N.; Holmes, D. J.; Pompliano, D. L. Nat. Rev. Drug

Discov. 2006, 6, 29-40.

(103) Bisacchi, G. S.; Manchester, J. I. ACS Infect. Dis. 2015, 1, 4–41.

216

(104) Bazile, S.; Moreau, N.; Bouzard, D.; Essiz, M. Antimicrob. Agents Chemother.

1992, 36, 2622-2627.

(105) Davis, T. D.; Gerry, C. J.; Tan, D. S. ACS Chem. Biol. 2014, 9, 2535–2544.

(106) Zhou, Y.; Joubran, C.; Miller-Vedam, L.; Isabella, V.; Nayar, A.; Tentarelli, S.;

Miller, A. Anal. Chem. 2015, 87, 3579–3584.

(107) Richter, M. F.; Drown, B. S.; Riley, A. P.; Garcia, A.; Shirai, T.; Svec, R. L.;

Hergenrother, P. J. Nature 2017, 545, 299-304.

(108) Ghosh, M.; Miller, P. A.; Möllmann, U.; Claypool, W. D.; Schroeder, V. A.;

Wolter, W. R.; Suckow, M.; Yu, H.; Li, S.; Huang, W.; et al. J. Med. Chem. 2017,

60, 4577–4583.

(109) Andrews, S. C.; Robinson, A. K.; Rodriguez-Quinones, F. FEMS Microbiol. Rev.

2003, 27, 215–237.

(110) Hider, R. C.; Kong, X. Chemistry and Biology of Siderophores. Nat. Prod. Rep.

2010, 27, 637–657.

(111) Vertesy, L.; Aretz, W.; Fehlhaber, H.-W.; Kogler, H. Helv. Chim. Acta 1995, 78,

46–60.

(112) Benz, G.; Schroder, T.; Kurz, J.; Wünsche, C.; Karl, W.; Steffens, G.; Pfitzner, J.;

Schmidt, D. Angew. Chem., Int. Ed. Engl. 1982, 21, 527–528.

(113) Ji, C.; Juarez-Hernandez, R. E.; Miller, M. J. Futur. Med. Chem. 2012, 4, 297–

313.

(114) Dhungana, S.; Harrington, J. M.; Gebhardt, P.; Möllmann, U.; Crumbliss, A. L..

217

Inorg. Chem. 2007, 46, 8362–8371.

(115) Neu, M. P. Los Alamos Sci. 2000, Chemical I, 416–417.

(116) Górska, A.; Sloderbach, A.; Marszałł, M. P. Trends Pharmacol. Sci. 2014, 35,

442–449.

(117) Miethke, M.; Marahiel, M. A. Microbiol. Mol. Biol. Rev. 2007, 71, 413–451.

(118) Schauer, K.; Rodionov, Amitry, A.; de Reuse, H. Cell 2008, 33, 330–338.

(119) Krewulak, K. D.; Vogel, H. J.. Biochim. Biophys. Acta - Biomembr. 2008, 1778,

1781–1804.

(120) Lubelski, J.; Konings, W. N.; Driessen, A. J. M. Microbiol. Mol. Biol. Rev. 2007,

71, 463–476.

(121) McPherson, C. J.; Aschenbrenner, L. M.; Lacey, B. M.; Fahnoe, K. C.; Lemmon,

M. M.; Finegan, S. M.; Tadakamalla, B.; O’Donnell, J. P.; Mueller, J. P.; Tomaras,

A. P. Antimicrob. Agents Chemother. 2012, 56, 6334–6342.

(122) Kohira, N.; West, J.; Ito, A.; Ito-Horiyama, T.; Nakamura, R.; Sato, T.;

Rittenhouse, S.; Tsuji, M.; Yamano, Y. Antimicrob. Agents Chemother. 2016, 60,

729–734.

(123) Page, M. G. P.; Dantier, C.; Desarbre, E. Antimicrob. Agents Chemother. 2010, 54,

2291–2302.

(124) (a) Black, P. N.; Dirusso, C. C. Microbiol. Mol. Biol. Rev. 2003, 67, 454–472. (b)

van den Berg, B.; Black, P. N.; Clemons, W. M.; Rapoport, T. A. Science, 2004,

304, 1506-1509.

218

(125) (a) Zhanel, G. G.; Walkty, A.; Vercaigne, L.; Karlowsky, J. A.; Embil, J.; Gin, A.

S.; Hoban, D. J. Can. J. Infect. Dis. 1999, 10, 207-238; (b) Zeitlinger, M. A.;

Sauermann, R.; Traunmüller, F.; Georgopoulos, A.; Müller, M.; Joukhadar, C. J.

Antimicrob. Chemother. 2004, 54, 876–880.

(126) Jolivet-Gougeon, A.; Bonnaure-Mallet, M. Drug Discov. Today Technol. 2014, 11,

49–56.

(127) Singh, R.; Ray, P.; Das, A.; Sharma, M. J. Med. Microbiol. 2009, 58, 1067–1073.

(128) Crump, J. A.; Luby, S. P.; Mintz, E. D. T. Bull. World Health Organ. 2004, 82,

346–353.

(129) Mogasale, V.; Maskery, B.; Ochiai, R. L.; Lee, J. S.; Mogasale, V. V; Ramani, E.;

Kim, Y. E.; Park, J. K.; Wierzba, T. F. Lancet Glob. Heal. 2014, 2, e570–e580.

(130) Gonzalez-escobedo, G.; Marshall, J. M.; Gunn, J. S. Nat. Rev. Microbiol. 2012, 9 ,

9–14.

(131) Merselis, J. G.; Kaye, D.; Connolly, S.; Hook, E. Am. J. Trop. Med. Hyg. 1964, 13

, 425–429.

(132) Levine, M. M.; Black, R. E.; Lanata, C. J. Infect. Dis. 1982, 146, 724–726.

(133) Gonzalez-Escobedo, G.; Marshall, J. M.; Gunn, J. S. Nat. Rev. Microbiol. 2011, 9 ,

9–14.

(134) Schiøler, H.; Christiansen, E. D.; Høybye, G.; Rasmussen, S. N.; Greibe, J. Scand.

J. Infect. Dis. 1983, 15, 17–19.

(135) Crawford, R. W.; Rosales-Reyes, R.; Ramírez-Aguilar, M. de la L.; Chapa-Azuela,

219

O.; Alpuche-Aranda, C.; Gunn, J. S. GProc. Natl. Acad. Sci. 2010, 107, 4353-

4358.

(136) Gotuzzo, E.; Guerra, J. G.; Benavente, L.; Palomino, J. C.; Carrillo, C.; Lopera, J.;

Delgado, F.; Nalin, D. R.; Sabbaj, J. J. Infect. Dis. 1988, 157, 1221–1225.

(137) Pratap, C. B.; Patel, S. K.; Shukla, V. K.; Tripathi, S. K.; Singh, T. B.; Nath, G.

Int. J. Antimicrob. Agents 2012, 40, 279–280.

(138) González, J. F.; Alberts, H.; Lee, J.; Doolittle, L.; Gunn, J. S. Sci. Rep. 2018, 8,

222.

(139) Rabin, N.; Zheng, Y.; Opoku-Temeng, C.; Du, Y.; Bonsu, E.; Sintim, H. O. Future

Med. Chem. 2015, 7, 647–671.

(140) Huigens, R. W.; Richards, J. J.; Parise, G.; Ballard, T. E.; Zeng, W.; Deora, R.;

Melander, C. J. Am. Chem. Soc. 2007, 129, 6966–6967.

(141) Richards, J. J.; Ballard, T. E.; Melander, C. Org. Biomol. Chem. 2008, 6, 1356–

1363.

(142) Richards, J. J.; Ballard, T. E.; Huigens, R. W.; Melander, C. ChemBioChem 2008,

9, 1267–1279.

(143) Rogers, S. A.; Melander, C. Angew. Chemie Int. Ed. 2008, 47, 5229–5231.

(144) Ballard, T. E.; Richards, J. J.; Aquino, A.; Reed, C. S.; Melander, C. J. Org. Chem.

2009, 74, 1755–1758.

(145) Rane, R. A.; Sahu, N. U.; Shah, C. P. Bioorg. Med. Chem. Lett. 2012, 22, 7131–

7134.

220

(146) Di Martino, P.; Fursy, R.; Bret, L.; Sundararaju, B.; Phillips, R. S. Can. J.

Microbiol. 2003, 49, 443–449.

(147) Bansal, T.; Englert, D.; Lee, J.; Hegde, M.; Wood, T. K.; Jayaraman, A. Infect.

Immun. 2007, 75, 4597–4607.

(148) Domka, J.; Lee, J.; Wood, T. K. Appl. Environ. Microbiol. 2006, 72, 2449-2459.

(149) Lee, J.; Bansal, T.; Jayaraman, A.; Bentley, W. E.; Wood, T. K. Appl. Environ.

Microbiol. 2007, 73, 4100–4109.

(150) Koopman, J. A; Marshall, J. M.; Bhatiya, A.; Eguale, T.; Kwiek, J. J.; Gunn, J. S.

Adenosine Mimetics. Antimicrob. Agents Chemother. 2015, 59, 76–84.

(151) Gewald, K.; Schinke, E.; Böttcher, H. Chem. Ber. 2018, 99, 94–100.

(152) Mojtahedi, M. M.; Abaee, M. S.; Mahmoodi, P.; Adib, M. Synth. Commun. 2010,

40, 2067–2074.

(153) Munch, H.; Hansen, J. S.; Pittelkow, M.; Christensen, J. B.; Boas, U. Tetrahedron

Lett. 2008, 49, 3117–3119.

(154) Wong, R.; Dolman, S. J. J. Org. Chem. 2007, 72, 3969–3971.

(155) Zeng, X.-H.; Liu, M.; Ding, M.-W.; He, H.-W. Synth. Commun. 2010, 40, 1453–

1460.

(156) Lee, S.-L.; Hsu, E.-C.; Chou, C.-C.; Chuang, H.-C.; Bai, L.-Y.; Kulp, S. K.; Chen,

C.-S. J. Med. Chem. 2011, 54, 6364–6374.

(157) Majtán, J.; Majtánová, Ľ.; Xu, M.; Majtán, V. J. Appl. Microbiol. 2007, 104,

1294–1301.

221

(158) Chambers, J. M.; Lindqvist, L. M.; Webb, A.; Huang, D. C. S.; Savage, G. P.;

Rizzacasa, M. A. Org. Lett. 2013, 15, 1406–1409.

(159) Yang, W.; Ni, L.; Somerville, R. L. Proc. Natl. Acad. Sci. 1993, 90, 5796-5800.

(160) Gorman, J.; Shapiro, L. C Protein Sci. 2009, 14, 3004–3012.

(161) Kishko, I.; Harish, B.; Zayats, V.; Reha, D.; Tenner, B.; Beri, D.; Gustavsson, T.;

Ettrich, R.; Carey, J. BPLoS One 2012, 7, e43902.

(162) Green, J.; Paget, M. S. Nat. Rev. Microbiol. 2004, 2, 954-966.

(163) Green, L. K.; La Flamme, A. C.; Ackerley, D. F. J. Microbiol. 2014, 52, 771–777.

(164) Catto, C.; Dell’Orto, S.; Villa, F.; Villa, S.; Gelain, A.; Vitali, A.; Marzano, V.;

Baroni, S.; Forlani, F.; Cappitelli, F. PLoS One 10, e0131519.

(165) Imlay, J. A. Nat. Rev. Microbiol. 2013, 11, 443-454.

(166) Gambino, M.; Cappitelli, F. Biofouling 2016, 32, 167–178.

(167) Petrova, O. E.; Sauer, K. J. Bacteriol. 2012, 194, 5817-5828.

(168) Martinez-Argudo, I.; Little, R.; Shearer, N.; Johnson, P.; Dixon, R. J. Bacteriol.

2004, 186, 601-610.

(169) Lin, S.-Y.; Yeh, T.-K.; Kuo, C.-C.; Song, J.-S.; Cheng, M.-F.; Liao, F.-Y.; Chao,

M.-W.; Huang, H.-L.; Chen, Y.-L.; Yang, C.-Y.; et al. J. Med. Chem. 2016, 59,

419–430.

(170) Jin, Y.; Ou, L.; Yang, H.; Fu, H. J. Am. Chem. Soc. 2017, 139, 14237–14243.

(171) Fuhrer, W.; Gschwend, H. W. J. Org. Chem. 1979, 44, 1133–1136.

(172) Cai, H.; Liu, Q.; Gao, D.; Wang, T.; Chen, T.; Yan, G.; Chen, K.; Xu, Y.; Wang,

222

H.; Li, Y.; et al. Eur. J. Med. Chem. 2015, 90, 241–250.

(173) Chiarino, D.; Ferrario, F.; Pellacini, F.; Sala, A. J. Heterocycl. Chem. 2018, 26,

589–593.

(174) van Veldhuizen, A.; van Dijk, M.; Sanders, G. M. Org. Magn. Reson. 2018, 13,

105–109.

(175) Arjunan, V.; Thillai Govindaraja, S.; Jayapraksh, A.; Mohan, S. Spectrochim. Acta

Part A Mol. Biomol. Spectrosc. 2013, 107, 62–71.

(176) Wang, M.; Zhang, Y.; Wang, T.; Wang, C.; Xue, D.; Xiao, J. Org. Lett. 2016, 18,

1976–1979.

(177) Stefani, H. A.; Pereira, C. M. P.; Almeida, R. B.; Braga, R. C.; Guzen, K. P.;

Cella, R. A Tetrahedron Lett. 2005, 46, 6833–6837.

(178) Maddox, S. M.; Nalbandian, C. J.; Smith, D. E.; Gustafson, J. L. A Org. Lett.

2015, 17, 1042–1045.

(179) Asakura, J.; Robins, M. J. J. Org. Chem. 1990, 55, 4928–4933.

(180) Rodriguez, R. A.; Pan, C.; Yabe, Y.; Kawamata, Y.; Eastgate, M. D.; Baran, P. S.

J. Am. Chem. Soc. 2014, 136, 6908–6911.

(181) Martı́nez, E.; Martı́nez, L.; Treus, M.; Estévez, J. C.; Estévez, R. J.; Castedo, L.

Tetrahedron 2000, 56, 6023–6030.

(182) Conway, B.; Crosbie, E.; Kennedy, A. R.; Mulvey, R. E.; Robertson, S. D. Chem.

Commun. 2012, 48, 4674–4676.

(183) Bielawski, M.; Malmgren, J.; Pardo, L. M.; Wikmark, Y.; Olofsson, B.

223

ChemistryOpen 2014, 3, 19–22.

224

Appendix: NMR Spectra of Selected Compounds

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 0.9 2.0 3.0 9.0 1 400 MHz H NMR spectrum of 2.67 in CDCl3

13 101 MHz C NMR spectrum of 2.67 in CDCl3 225

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 1.0 1.0 3.0 1 400 MHz H NMR spectrum of 2.76 in CDCl3

13 101 MHz C NMR spectrum of 2.76 in CDCl3

226

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 1.0 1.0 3.0 3.0

1 600 MHz H NMR spectrum of 2.77 in CDCl3

13 151 MHz C NMR spectrum of 2.77 in CDCl3

227

1 400 MHz H NMR spectrum of 2.82 in CDCl3

1 400 MHz H NMR spectrum of 6 in CDCl3

13 101 MHz C NMR spectrum of 2.82 in CDCl3

228

nOe NMR spectrum of 2.82 in CDCl3

229

Selective Gradient 1D NOESY of 2.82 (400 MHz in CDCl3)

230

HSQC NMR spectrum of 2.82 in CDCl3

231

HMBC NMR spectrum of 2.82 in CDCl3 (HSQC cross peaks in green)

232

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 1.0 1.0 1.0 1.0 1.0

1 600 MHz H NMR spectrum of 2.86 in CDCl3

13 151 MHz C NMR spectrum of 2.86 in CDCl3

233

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 1.0 1.0 1.2 1.1 8.9

1 400 MHz H NMR spectrum of 2.83 in CDCl3

13 101 MHz C NMR spectrum of 2.83 in CDCl3

234

11 10 9 8 7 6 5 4 3 2 1 ppm

1.0 6.0

1 600 MHz H NMR spectrum of 2.85 in CDCl3

13 151 MHz C NMR spectrum of 2.85 in CDCl3

235

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 1.0 1.0 1.0 1.0 1.0 1 400 MHz H NMR spectrum of 2.89 in CDCl3

13 101 MHz C NMR spectrum of 2.89 in CDCl3

236

HSQC spectrum of 2.89 in CDCl3

HMBC spectrum of 2.89 in CDCl3

237

1 400 MHz H NMR spectrum of 2.96 in CDCl3

1 101 MHz H NMR spectrum of 2.96 in CDCl3 238

1 400 MHz H NMR spectrum of S2 in CDCl3

13 101 MHz C NMR spectrum of S2 in CDCl3

239

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 3.0 3.0 1 400 MHz H NMR spectrum of 2.99 in CDCl3

13 101 MHz C NMR spectrum of 2.99 in CDCl3

240

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 1.0 1.0 1.0 1.0 0.9 2.0 3.0 3.0 3.0 3.0

1 400 MHz H NMR spectrum of 2.101 in CDCl3

13 101 MHz C NMR spectrum of 2.101 in CDCl3

241

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm

1.0 1.0 1.0 2.0 1.0 2.0 6.0 3.0 3.0 1 400 MHz H NMR spectrum of 2.102 in CDCl3

13 151 MHz C NMR spectrum of 2.102 in CDCl3

242

1 400 MHz H NMR spectrum of 2.103 in CDCl3

13 101 MHz C NMR spectrum of 2.103 in CDCl3

243

1 700 MHz H NMR spectrum of 2.106 in CDCl3

13 176 MHz C NMR spectrum of 2.106 in CDCl3

244

1 4 00 MHz H NMR spectrum of 2.107 in CDCl3

13 101 MHz C NMR spectrum of 2.107 in CDCl3 245

1 400 MHz H NMR spectrum of 4.5 in DMSO-d6

13 101 MHz C NMR spectrum of 4.5 in DMSO-d6

246

1 400 MHz H NMR spectrum of 4.16 in CDCl3

13 101 MHz C NMR spectrum of 4.16 in CDCl3

247

1 400 MHz H NMR spectrum of 4.18 in CDCl3

13 101 MHz C NMR spectrum of 4.18 in DMSO-d6

248

1 400 MHz H NMR spectrum of 4.19 in CDCl3

13 101 MHz C NMR spectrum of 4.19 in CDCl3

249

1 400 MHz H NMR spectrum of 4.21 in CDCl3

13 101 MHz C NMR spectrum of 4.21 in CDCl3

250

1 400 MHz H NMR spectrum of 4.22 in CDCl3

13 101 MHz C NMR spectrum of 4.22 in CDCl3

251

1 400 MHz H NMR spectrum of 4.24 in CDCl3

13 101 MHz C NMR spectrum of 4.24 in CDCl3

252

1 400 MHz H NMR spectrum of 4.25 in CDCl3

13 101 MHz C NMR spectrum of 4.25 in CDCl3

253

1 400 MHz H NMR spectrum of 4.26 in CDCl3

13 101 MHz C NMR spectrum of 4.26 in CDCl3

254

1 400 MHz H NMR spectrum of 4.27 in CDCl3

13 101 MHz C NMR spectrum of 4.27 in CDCl3 255

400 MHz 1H NMR spectrum of 4.31 in MeOD

101 MHz 13C NMR spectrum of 4.31 in MeOD 256

400 MHz 1H NMR spectrum of 4.30 in MeOD

101 MHz 13C NMR spectrum of 4.30 in MeOD

257

1 300 MHz H NMR spectrum of 4.32 in MeOD

258

1 400 MHz H NMR spectrum of 4.39 in CDCl3

13 176 MHz C NMR spectrum of 4.39 in CDCl3

259

1 400 MHz H NMR spectrum of 4.40 in CDCl3

13 101 MHz C NMR spectrum of 4.40 in CDCl3

260

1 300 MHz H NMR spectrum of 4.37 in CDCl3

13 75 MHz C NMR spectrum of 4.37 in CDCl3

261

1 400 MHz H NMR spectrum of 4.41 in MeOD

262

400 MHz 1H NMR spectrum of 4.36 in MeOD

263