Citation
Structure-based design and synthesis of selective small molecule inhibitors for MCL-1, an important cancer target

Material Information

Title:
Structure-based design and synthesis of selective small molecule inhibitors for MCL-1, an important cancer target
Creator:
McNulty, Oren ( author )
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electrnic file (139 pages). : ;

Subjects

Subjects / Keywords:
Cancer -- Molecular aspects ( lcsh )
Protein binding ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Mcl-1, a member of the Bcl-2 family of proteins responsible for regulating apoptosis, is an important target for cancer research. Overexpression of Mcl-1 has been linked to many cancers, including multiple myeloma and chronic myeloid leukemia. Additionally, recent studies have shown that resistance to many chemotherapeutic drugs often is correlated with Mcl-1 overexpression. The apical anti-apoptotic function of Mcl-1 in comparison to other anti-apoptotic Bcl-2 family members signal the importance of finding a pharmacophore to counteract this overexpression. Recent success at finding potent inhibitors of other Bcl-2 family members has heightened the search for a small molecule capable of performing the crucial Mcl-1 inhibitory function. By using molecular modeling to predict protein interactions of known Mcl-1 inhibitor obatoclax (1) as a starting point, novel synthetic targets for an Mcl-1 inhibitor were formulated. Obatoclax (1) is a known potent pan-inhibitor of the anti-apoptotic members of the Bcl-2 family, although its pharmacologic efficacy is poor due to neurotoxicity and poor solubility. According to molecular docking experiments performed, obatoclax (1) has two crucial moieties that enable the molecule to interact with the binding groove of Mcl-1. The indole nitrogen acts as a hydrogen bond donator when protonated and the central pyrrole nitrogen acts as a hydrogen bond acceptor. Therefore, a structure-based design target (2) was devised to emulate these interactions. From this target, a novel indole pyridine core (3) was established that led to the synthesis of eight novel compounds, seven of which were tested for binding affinity for Mcl-1. Two of these synthesized derivatives had binding constants (Kd) between 10 and 50 &mgr;M for Mcl-1 using a surface plasmon resonance (SPR) assay, prompting further structure-activity relationship (SAR) studies and optimization. In collaboration with Scripps Florida, a hybrid compound was synthesized using the indole pyridine core with a trifluoromethyl (CF3) group on the indole ring that achieved a 6.6 &mgr;M (Kd) affinity for Mcl-1. This compound and two others that contain the trifluoromethyl group (Kd between 1 and 10 &mgr;M) have prompted a SAR study surrounding this moiety to probe if it is forming a key interaction with the Mcl-1 protein, and thus increasing affinity for binding. The indole pyridine core, especially derivatives containing the trifluoromethyl group, appears to be a promising new pharmacophore for future development as selective small-molecule Mcl-1 inhibitors.
Thesis:
Thesis (M.S.)--University of Colorado Denver. Department of Chemistry
Bibliography:
Includes bibliographic references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Oren McNulty.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
913224691 ( OCLC )
ocn913224691

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

STRUCTURE BASED DESIGN AND SYNTHESIS OF SELECTIVE SMALL MOLECULE INHIBITORS FOR MCL 1, AN IMPORTANT CANCER TARGET by OREN MCNULTY B.S. Chemistry, University of Colorado Denver 2002 B.S. Biology, University of Colorado Denver, 2002 A thesis su bmitted to the Faculty of the Graduate School of the University of Colorado Denver in partial fulfillment of the requirements for the degree of Master of Science Chemistry 2015

PAGE 2

! "" 2015 OREN MCNULTY ALL RIGHTS RESERVED

PAGE 3

! """ This thesis for the Master of Science degree by Oren McNulty Has been approved for the Department of Chemistry by Lisa Julian Chair Jefferson Knight Hai Lin Scott Reed April 23rd 2015

PAGE 4

! "# McNulty, Oren (M.S., Chemistry) Structure based design and synthesis of se lective small molecule inhibitors for Mcl 1, an important cancer target. Thesis directed by Professor Lisa Julian. ABSTRACT Mcl 1, a member of the Bcl 2 family of proteins responsible for regulating apoptosis, is an important target for cancer research. Overexpression of Mcl 1 has been linked to many cancers, including multiple myeloma and chronic myeloid leukemia. Additionally, r ecent studies have shown that resistance to many chemotherapeutic drugs often is correlated with Mcl 1 overexpression. The apic al anti apoptotic function of Mcl 1 in comparison to other anti apoptotic Bcl 2 family members signal the importan ce of find ing a pharmac o phore to counteract this overexpression. Recent success at finding potent inhibitors of other Bcl 2 family members has heightened the search for a small molecule capable of performing the crucial Mcl 1 inhibitory function. B y using molecular modeling to predict protein interactions of known Mcl 1 inhibitor obatoclax ( 1 ) as a starting point novel synthetic targets for an Mcl 1 inhibitor were formulated. Obatoclax ( 1 ) is a known potent pan inhibitor of the anti apoptotic members of the Bcl 2 family although its pharmacologic efficacy is poor due to neurotoxicity and poor solubility According to molecular docking experim ents performed obatoclax ( 1 ) has two crucial moieties that enable the molecule to interact with the binding groove of Mcl 1. The indole nitrogen acts as a hydrogen bond donator when protonated and the central pyrrole nitrogen acts as a hydrogen bond accep tor Therefore, a structure based NH N HN O NH N HN NH N X Obatoclax (1) Structure-based design target ( 2 ) Indole pyridine core ( 3 )

PAGE 5

! # design target ( 2 ) was devised to emulate these interactions. From this target, a novel indole pyridine core ( 3 ) was established that led to the synthesis of eight novel compounds seven of which were tested for binding aff inity for Mcl 1. Two of the se synthesized derivatives had binding constants (K d ) between 10 a nd 5 0 !M for Mcl 1 using a surfac e plasmon resonance (SPR) assay prompting further structure activity relationship (SAR) studies and optimization In collaborati on with Scripps Florida a hybrid compound was synthesized using the indole pyridine core with a trifluoromethyl (CF 3 ) group on the indole ring that achieve d a 6.6 !M (K d ) affinity for Mcl 1. This compound and two others that contain the trifluoromethyl gr oup (K d between 1 and 10 M ) have prompted a SAR study surrounding this moiety to probe if it is forming a key interaction with the Mcl 1 protein, and thus increasing affinity for binding T he indole pyridine core especially derivatives containing the tri fluoromethyl group, appears to be a promising new pharma coph ore for future development as selective small molecule Mcl 1 inhibitor s The form and content of this abstract are approved. I recommend its publication. Approved: Lisa Julia n

PAGE 6

! #" ACK NOWLE D GEMENTS I would like to thank Dr. Lin 's lab, and in particular Cece Johnson Sasso and Lesley Tyk for their diligence in finding the important data for the molecular modeling of these compounds. Dr. Lin's lab has been instrumental to the success of t his project Of course I would like to profoundly thank Dr. Lisa Julian for her mentorship. She has enabled me to progress throughout this endeavor with a smile, even with all the many twists and turns the research inevitably led us through She is bar no ne the best mentor I could ever have hoped for. And finally I would like to thank my family, especially my wife Cathy. She has always supported me throughout this entire project, and surely I would not have made it to this end without that. I thank her for her love and support and certainly feel as if this is our achievement.

PAGE 7

! #"" TABLE OF CONTENTS CHAPTER I. INTRODUCTION 1 I ntrinsic and e xtrinsic apoptosis p athways 2 The Bcl 2 family 3 Anti apoptotic Mcl 1: a challenging yet intrigui ng target 6 A survey of current Mcl 1 inhibitors 9 DNA intercalating agents 9 Peptides and peptide mimetics 1 1 Prodigiosin derivatives 13 Gossypol and gossypol derivatives 1 4 Sulfonamides 15 Pyridine based rhodani ne derivatives 1 7 Other key small molecule inhibitors 1 8 University of Colorado at Denver, Julian research group focus 1 9 II. THE HETEROCYCLE DERIVATIVES 23 Molecular docking method 2 3 The pyrrole derivative: 2 (6 ((3,5 dimethyl 1H pyrr ol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) 2 5 Molecular docking of the pyrrole derivative of indole pyridine core ( 18 ) 2 5 Synthesis of the indole pyridine c ore 28 Protection of the N proton of the indole ring 2 8 Formation of N Boc indole 2 boronic acid ( 22 ) 29 Suzuki Miyaura cross coupling to form N Boc protected indole pyridine core ( 19 ) 29 Acid catalyzed coupling of dimethyl pyrrole ( 27 ) to the indole pyridine core aldehyde ( 19 ) 32 Reduction of the olefin linker of the indole pyridine pyrrole ( 28 ) 33 Deprotection of the indole nitrogen on the pyridine N Boc indole pyrrole ( 29 ) 3 4

PAGE 8

! #""" The thiophene derivative: 2 (6 (thiophen 2 ylmethyl)pyridin 2 yl) 1H indole ( 30 ) 35 Molecular docking of thiophene derivative ( 30 ) 3 5 Thiophene coupling to indole pyridine carboxaldehyde ( 19 ) 38 Reduction of the biaryl al cohol ( 32 ) 38 Binding affinity testing of the thiophene derivative to Mcl 1 3 9 Experimental s ection 42 General experi mental details 42 N ( tert butoxycarbonyl) indole ( 21) ! ! 43 N ( tert butoxycarbonyl) indole 2 boronic acid ( 22 ) 44 tert butyl 2 (6 formylpyridin 2 yl) 1H indole 1 carboxylate ( 19 ) 45 tert butyl 2 (6 ((3,5 dimethyl 2H pyrrol 2 ylidene)methyl) pyridin 2 yl) 1H indole 1 carboxylate ( 28 ) 46 tert butyl 2 (6 ((3,5 dimethyl 2H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole 1 carboxylate ( 29 ) 47 2 (6 ((3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) 48 tert butyl 2 (6 (hydroxy(thiophen 2 yl)methyl)pyridin 2 yl) 1H indole 1 carboxylate ( 32 ) 49 2 (6 (thiophen 2 ylmethyl)pyridin 2 yl) 1H indole ( 30 ) 50 III. METHANOL, VINYL, AND ETHYL DERIVATIVES 51 The methanol derivative: (6 (1 H indol 2 yl)pyridin 2 y l)methanol ( 33 ) 51 Molecular docking of methanol derivative ( 33 ) 51 Reduction of indole pyridine aldehyde ( 34 ) 54 Deprotection of indole nitrogen 54 Binding affinity testing of (6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 33 ) with Mcl 1 5 5 The vinyl derivative: 2 (6 vinylpyridin 2 yl) 1 H indole ( 35 ) 55 Molecular docking of vinyl indole pyridine derivative ( 35 ) 55 Unstabilized Wittig reaction of indole pyridine carboxaldehyde ( 19 ) with a methyl Wittig reagent 56

PAGE 9

! "$ Deprotection o f indole nitrogen 58 Binding affinity testing of 2 (6 vinylpyridin 2 yl) 1 H indole ( 35 ) with Mcl 1 5 9 The ethyl derivative: 2 (6 ethylpyridin 2 yl) 1 H indole ( 37 ) 5 9 Molecular docking of the ethyl derivative ( 37 ) 5 9 Reduction of unpro tected vinyl indole pyridine ( 35) 61 Binding a ffinity testing of 2 (6 ethylpyridin 2 yl) 1 H indole ( 37 ) with Mcl 1 61 Experimental s ection 62 tert butyl 2 (6 (hydroxymethyl)pyridin 2 yl) 1 H indole 1 carboxylate ( 34 ) 6 2 ( 6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 33 ) ! 6 3 tert butyl 2 (6 vinylpyridin 2 yl) 1 H indole 1 carboxylate ( 36 ) 6 4 2 (6 vinylpyridin 2 yl) 1 H indole ( 35 ) 6 5 2 (6 ethylpyridin 2 yl) 1 H indole ( 37 ) 6 6 IV (E/Z) 1 PROPENYL AND PROPYL DERIVATIV ES 67 The trans 1 propenyl derivative: ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 38 ) 67 Molecular docking of the trans 1 propenyl derivative ( 3 8 ) 67 Unstabilized Wittig reaction of indole pyridine carboxaldehyde ( 19 ) with ethyl Witti g reagent 70 Deprotection of indole nitrogen 70 Binding affinity testing of ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 38 ) with Mcl 1 71 The cis 1 propenyl derivative: ( Z ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 41 ) 71 Molecular docking of cis 1 propenyl indole pyridine derivative ( 41 ) 71 Deprotection of indole nitrogen 71 Binding affinity testing of ( Z ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 41 ) for Mcl 1 72 The propyl derivative: 2 (6 prop ylpyridin 2 yl) 1 H indole ( 42 ) 72 Molecular docking of the propyl derivative ( 42) 7 2

PAGE 10

! $ Deprotection of indole nitrogen 75 Reduction of unprotected ( E / Z) 1 propenyl indole pyridine ( 38 & 41 ) 7 5 Binding a ffinity testing of 2 (6 propylpyridin 2 yl) 1 H indole ( 42 ) with Mcl 1 7 6 Experimental s ection 77 (E/Z) tert butyl 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1H indole 1 carboxylate ( 39 & 40 ) 77 ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 38 ) 78 (Z) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1H indole ( 41 ) 79 2 (6 propylpyridin 2 yl) 1 H indole ( 42 ) 80 V. CONCLUSION 81 Outlook 85 VI. SUPPLEMENTAL INFORMATION 90 REFERENCES 121

PAGE 11

! $" LIST OF FIGURES FIGU RE 1 1. Intrinsic and Extrinsic Pathways of Apoptosis 1 1 2. The Bcl 2 family 4 1 3. Conformational shifts of Bcl X L and Mcl 1, respectively, after binding of BIM BH3 5 1 4. Secondary structure of Mcl 1 bound to Bax BH3 7 1 5. X ray structure o f Mcl 1 (PDB ID 3MK8) showing the binding pockets 1 6. P1 P4 8 1 7. Molecular structure of ABT 737 (1) 8 1 8. Molecular structure of S1 (2) a proposed Mcl 1 inhibitor 11 1 9. X ray crystallography of a stabilized alpha helix of Bcl 2 domains SAHB D ) 12 1 10. Prodigiosin (3) molecular structure 13 1 11. Obatoclax (4) molecular structure 14 1 12. Gossypol (5) molecular structure 14 1 13. The sulfonamide functional group 15 1 14. Molecular Structure of ABT 263 (6) 16 1 15. Takeda sulfonamide ( 7 ) derivative and AbbVie sulfonamide ( 8 ) derivative 17 1 16. Molecular Structure of BH3I 1 ( 9 ) and one of the developed derivative s ( 10 ) 17 1 17. A 7 hydroxyquinoline derivative Mcl 1 inhibitor 18 1 18. Vanderbilt study on a novel Mcl 1 Inhibitor 19 1 19. Mol ecular docking of obatoclax ( 4 ) to Mcl 1 20 2 1. Obatoclax ( 4 ) molecular structure 24 2 2. Obatoclax's ( 4 ) best configuration with Mcl 1 (PDB # 2PQK ) using molecular docking 25 2 3. Molecular structure of 2 (6 ((3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) 26 2 4. Molecular docking of pyrrole derivative 18 to Mcl 1 27 2 5. Catalyt ic cycle of the Suzuki Miyaura cross coupling r eaction 30 2 6 Molecular s tructure of thiophene derivative ( 30 ) 35 2 7. Molecular docking of thiophene derivative 30 to Mcl 1 37

PAGE 12

! $"" 2 8. Illustration of the Surface Plasmon Resonance (SPR) assay to determine ligand protein binding affinity 40 3 1. Molecular structure of (6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 33 ) 52 3 2 Molecular docking of methanol derivative 33 to Mcl 1 53 3 3 Molecular s tructure of vinyl indole pyridine derivative ( 35 ) 55 3 4. General schematic for a Wittig reaction 56 3 5 Formation of an ylide from the Wittig reagent 57 3 6 Mol ecular structure of ethyl indole pyridine derivative 59 3 7. Molecular docking of ethyl derivative 37 to Mcl 1 60 4 1 Molecular structure of trans 1 propenyl indole pyridine derivative ( 38 ) 68 4 2 Molecular docking of trans 1 propenyl derivative 38 to Mcl 1 69 4 3. Molecular structure of cis 1 propenyl indole pyridine derivative ( 41 ) 71 4 4 Molecular Structure of propyl indole pyridine derivative ( 42 ). 73 4 5 Molecular docking of propyl derivative 42 to Mcl 1 7 4 5 1. Molecular docking o f SR3 3893 84 5 2 Surface Plasmon Resonance (SPR) data of SR3 3893 with Mcl 1 85 5 3 Synthetic targets for a Mcl 1 inhibitor to establish a structure activity relationship (SAR). 86 5 4 Proposed new synthetic target ( 43 ) based on initia l SAR results 8 6 5 5 Molecular docking of compound ( 43 ) with Mcl 1 8 8

PAGE 13

! $""" LIST OF SCHEMES S CHEME 1 1. Structure based design scheme for the development of a novel Mcl 1 inhibitor 22 2 1. Conceptual retro synthesis for Mcl 1 inhibitor derivati ves 28 2 2. Protection of the indole N proton 29 2 3. Boronic acid ( 22 ) synthesis 2 9 2 4. Suzuki Miyaura cross coupling reaction in an obatoclax synthesis by Da•ri et al 31 2 5. Suzuki Miyaura cross coupling reaction, yielding the indole pyr idine core ( 19 ) 32 2 6. Dimethylpyrrole ( 27 ) coupling to the indole pyridine core ( 19 ) 33 2 7. Reduction of olefin linker of the indole pyridine dimethylpyrrole compound ( 28 ) 34 2 8. TFA deprotection to yield the final product, 2 (6 (( 3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) 35 2 9. Nucleophilic addition of thienyl lithium ( 31 ) to an aldehyde ( 19 ) 38 2 10. Triethyl silane reduction of biaryl alcohol ( 32) 39 3 1. NaBH 4 reduction of indole pyridine aldehyde 54 3 2. Boc deprotection of N Boc protected indole pyridine methanol derivative ( 34 ) 55 3 3. Methyl Wittig reaction with indole pyridine carboxaldehyde ( 19 ) 58 3 4. Deprotection of the Boc protected indole pyridine vinyl derivative ( 36 ) 58 3 5. Palladium on carb on reduction of vinyl indole pyridine compound ( 35 ) 61 4 1. Ethyl Wittig reaction with indole pyridine aldehyde ( 19 ) 70 4 2. Deprotection of the Boc protected indole pyridine trans 1 propenyl derivative ( 39 ) 71 4 3. Deprotection of the Boc protected indo le pyridine cis 1 propenyl derivative ( 4 0 ) 72 4 4. Deprotection of a mixture of trans ( 39 ) & cis ( 40 ) stereoisomers of the Boc protected 1 propenyl indole pyridine derivative 75 4 5. Palladium on carbon reduction of olefin of the unprotected ( E / Z) 1 propenyl indole pyridine analogues ( 38 & 41 ) 75

PAGE 14

! $"# LIST OF TABLES T ABLE 5 1. Obatoclax (4) and the d erivatives synthesized by the Julian Lab 81 5 2. Derivatives synthesized by Scripps Florida 83

PAGE 15

! $# LIST OF ABBREVIATIONS AIDS Acquired Immu ne Deficiency Syndrome Bak Bcl 2 homologous antagonist killer Bax Bcl 2 associated X Bcl 2 B Cell Lymphoma 2 Bcl X L B cell lymphoma extra large BH Bcl 2 Homology Boc tert butyloxycarbonyl DCM Dichloromethane DISC Death Inducing Signaling Complex E SP Electrostatic potential charge FADD Fas Associated protein with Death Domain FTIR Fourier transform infrared spectroscopy HIV Human Immunodeficiency Virus HPLC High performance liquid chromatography HSA Human serum albumin k a Association rate k d Dissociation rate K d Ligand protein binding constant LCMS Liquid chromatography mass spectroscopy LDA Lithium diisopropylamine LRMS Low resolution mass spectrometry Mcl 1 Myeloid Cell Leukemia sequence 1 NA Natural atomic charge NMR Nuclear Magnetic Resonance

PAGE 16

! $#" Noxa Phorbol 12 myristate 13 acetate induced protein 1 PDB Protein data bank PEST proline / glutamic acid / serine / threonine sequence PUMA p53 upregulated modulator of apoptosis RMSD Root mean squared deviation SPhos 2 dicyclohexylphosph ino 2,6 dimethoxybiphenyl SPR Surface plasmon resonance TFA Trifllouroacetic acid TLC Thin layer chromatography TM transmembrane sequence XIAP X linked Inhibitor of Apoptosis

PAGE 17

1 C HAPTER I INTRODUCTION A current growing area of research for the treatment of cancer centers on the anti apoptotic proteins in the B Cell Lymphoma 2 class (Bcl 2) family. If a healthy cell undergoes a stress that cannot be recovered, such as DNA damage o r UV radiation, many times it has no recourse but to undergo apoptosis (regulated cell death) This normal cellular process ensures that damaged DNA is not propagated which could ultimately cause detrimental effects such as cancer T here are two cellular signaling pathways that can lead to apoptosis : the intrinsic and extrinsic pathways (Figure 1 1 ) The Bcl 2 family of proteins are involved in the intrinsic pathway of apoptosis as both pro apoptotic and anti apoptotic regulators. Figure 1 1 Intrinsic a nd Extrinsic Pathways of Apoptosis 1

PAGE 18

2 Intrinsic and e xtrinsic a poptosis p athway s Th e intrinsic pathway is triggered when intracellular stress signals occur such as ones caused by DNA damage or growth factor withdrawal T hese signals can trigger tumor suppre ssor p53 protein to upregulate the expression of several pro apoptotic proteins, including p53 upregulated modulator of apoptosis ( PUMA ) and Phorbol 12 myristate 13 acetate induced protein 1 ( Noxa ) which are both Bcl 2 family members T hese two apoptosis a ctivators sequester anti apoptotic Bcl 2 family members, such as Bcl 2, B cell lymphoma extra large (Bcl X L ) and Myeloid Cell Leukemia sequence 1 (Mcl 1), thereby preventing them from binding to the pro apoptotic proteins Bcl 2 associated X ( Bax ) and Bcl 2 homologous antagonist killer ( Bak) Bax and Bak, also Bcl 2 family members, are in this manner allowed to oligomerize at the mitochondria, which allows outer mitochondrial membrane permeabil ity via the mitochondrial voltage dependent anion channel 2 or by forming an oligomeric pore 3 This permeability of the outer mitochondrial membrane releases many downstream pro apoptotic proteins, including cytochrome c, which eventually leads to cellular apoptosis via the ca sp ase cascade 4 In contrast, the extrinsic pathway for apoptosis is initiated by extra cellular ligands that bind to transmembrane receptors called "death" receptors. Once such a binding of a ligand to the death receptor (FasR) occurs, it leads to a trimeri z ation which then also allows Fas Associat ed protein with Death Domain (FADD) to bind 5 Once this conformational change occurs procaspase 8 and 10 are sequestered to FADD and their binding creates the complex know as the Death Inducing Signaling Complex (DISC) 6 Once in close proximity to each ot her within the DISC these procapsases are cleaved into their active forms, caspases 8 and 10, which then activate the caspase cascade leading to apoptosis. Other extrinsic pathways also occur with slightly different mechanisms but all have in common the extracellular receptor mediated initiation.

PAGE 19

3 Interestingly, p53 also has a role in the extrinsic pathway by up regulating the death receptors i n the cell and thus increasing the likelihood for apoptosis to be activated. Regulation of these two pathways is c omplex and yet critical for normal human function. Dysfunction by either overexpression or suppression of the regulators of these pathways usually leads to life threatening diseases. Suppression or deactivation of anti apoptotic regulators and/or overexpre ssion of pro apoptotic regulators can lead to neurodegenerative and hematologic diseases, as well as tissue damage. In fact, deactivation of Bcl 2 (an anti apoptotic protein) by a specific Human Immunodeficiency Virus (HIV) viral protein has been implicate d as a key player in the progression of Acquired Immune Deficiency Syndrome (AIDS) 7 In contrast, overexpression of anti apoptotic proteins and/or suppression of pro apoptotic proteins can lead to cancer, autoimmune diseases, and inflammatory diseases. As an example, human non small cell lung cancer NCI H460 cells overexpress the protein X linked Inhibitor of Apoptosis (XIAP), which allows the immortality of that cell line 8 Because of their major role in human disease, apoptotic regulators have long been a focus for pharmacologic research. The Bcl 2 family The Bcl 2 family of proteins is a group of proteins involved in the apoptotic pathways that share certain regions call ed Bcl 2 Homology (BH) regions (that in some cases make up the binding domains). The y can be further classified as subsets of anti apoptotic (green in Figure 1 2) or pro apoptotic (red in Figure 1 2) proteins. All of the anti apoptotic members of this family, along with pro apoptotic Bak and Bax, share all four BH regions as well as a tra nsmembrane sequence (grouping a in Figure 1 2). However, some member s share only one of the four of these BH regions, specifically the

PAGE 20

4 BH3 domain and the transmembrane sequence and are named as BH3 only proteins (grouping b in Figure 1 2 ). Figure 1 2 The Bcl 2 family Members in green represent those with anti apoptosis activity. Members in red represent those with pro apoptosis activity. Grouping a includes members with all four BH domains. Grouping b includes BH3 only members. TM = t ransmembrane sequ ence. Secondary structure alpha helix locations are indicated above the sequence as well. 9 The secondary structures of all of the multi BH domain family members are very similar. All share a common eight helix bundle that creates a groove in which BH3 dom ain peptides can bind. Once this binding occurs, a conformational shift does occur, however to different degrees for each family member. For instance, a large shift occurs with Bcl X L when it binds to the B im BH3 ligand. However, when Mcl 1 binds to the sa me ligand a much smaller shift occurs (Figure 1 3). This has an impact of potential ligands binding, as some ligands require an opening of the hydrophobic pocket to occur before

PAGE 21

5 they can bind. Thus, it might be challenging for some ligands to bind with an y specificity for Mcl 1 compared with Bcl X L Therefore certain specific three dimensional configurations of these proteins may also bind to specific BH3 ligands, allowing for extreme specificity. Figure 1 3. Conformational shifts of Bcl X L and Mcl 1 re spectively after binding of BIM BH3 10 ApoBcl X L (yellow); Bcl X L bound (green) to BIM BH3 (purple); ApoMcl 1 (yellow); Mcl 1 ( blue) bound to BIM BH3 (purple) The anti apoptosis Bcl 2 family members function as regulators of the activation and oligomeriza tion of the BH3 only molecules and therefore direct activation of apoptosis at the mitochondrial level This subset shares certain regions named Bcl 2 homology (BH) domains that are involved in the binding of BH3 only proteins, which effectively renders th e BH3 only protein's pro apoptotic function inactive. This leads to cell survival even when cellular stress signals and/or death signals have been activated. By creating BH3 mimetic compounds that selectively bind to the BH 3 domain of the Bcl 2 family, apo ptosis could be activated in cells that were previously immortalized via Bcl X L Mcl 1

PAGE 22

6 uncontrolled expression of the Bcl 2 family proteins. Over expression of these proteins has been linked to cancer, neurodegeneration and autoimmunity 11 Anti apoptotic Mcl 1 : a c halle nging yet i ntriguing t arget Myeloid Cell Leukemia sequence 1 (Mcl 1), a member of the anti apoptosis subset of the Bcl 2 family, is of particular interest due to its high degree of regulation and high degree of amplification in many human cancers, includi ng multiple myeloma, hepatocellular carcinomas, colon carcinomas and chronic myeloid leukemia 12 Studies have indicated that it has a key role in the intrinsic pathway Mcl 1 is 1) up regulated by several growth factors and cytokines 13 2) crucial for the ear ly survival of many human cell lineages, 3) involved as a truncated form in normal mitochondrial function. Furthermore, Mcl 1 has also been linked to chemotherapeutic resistance and relapse rates, especially with drugs serving as other Bcl 2 member inhibit ors. Because of the unique functions in uncontrolled cellular growth, Mcl 1 has been stud ied extensively to determine it s structure and modes of regulation to better understand its function. However, even with this focus, selective, small molecule inhibito rs for Mcl 1 have yet to make it to in vivo trials and only moderate potencies in vitro have been achieved. The structural differences between Mcl 1 and other anti apoptosis family members may provide insight into not only why it is such an important targ et, but also why Mcl 1 inhibitors have been unsuccessful to date. Mcl 1 (Figure 1 4 ) is larger (350 residues) than most Bcl 2 family proteins, such as Bcl 2 (239 residues ) and Bcl X L (233 residues ) It has a longer N terminal region than on other Bc l 2 fam ily members that contains crucial potential regulation sites including two PEST (proline / glutamic acid / serine / threonine containing) sequences that have a significant role in protein degradation by possibly acting as a peptide signal for proteasomes In addition, several

PAGE 23

7 potential phosphorylation sites which may be used to regulate its function to a higher degree than for any of the other family members have been found in this region It has also been found to have a very short half life ( less than o ne hour ) compared to the Bcl 2 protein ( 20 hours 14 ) which also indicat es its highly regulated function. The secondary structure has now been revealed by nuclear magnetic resonance (NMR) and crystallography studies, and the binding grooves have been charac terized in recent years. These findings have enabled an elucidation into the differences between Mcl 1 and other family members in terms of binding of BH3 ligands and other small molecules Figure 1 4. Secondary structure of Mcl 1 (blue) bound to Bax B H3 (green) 15 S how s the characteristic eight helix bundle. Due to the aforementioned structural differences, Mcl 1 has distinctive pockets in the binding grooves (Figure 1 5) that behave differently than in other family members. For instance, the P2 pocket on Mcl 1 has a much higher degree of plasticity, which in turn expands upon ligand binding to form a large hydrophobic cavity. However, the P4

PAGE 24

8 pocket of Mcl 1 is not as defined and is more exposed to solvent, leading to a shallower and more hydroph ilic tar get for potential inhibitors. For this reason, it has been hypothesized that finding inhibitors selective for the P2 pocket would allow for more selectivity and therefore better potenc y 16 To illustrate the challenges inherent in finding a Mcl 1 inhibitor t hat binds selectively, ABT 737 ( 1 Figure 1 6 ) was developed by Abbott as a Bcl X L inhibitor from a fragment screening and was found to selectively bind to both the P2 and P4 pockets of that protein, thus allowing for sub nanomolar potencies. Likely due to the shallow and solvent exposed binding groove of the P4 pocket specific Mcl 1 inhibitors binding to this region has proven difficult to achieve Therefore, Mcl 1 inhibitor candidates will likely initially have lower in vitro potencies than for other fam ily members due to the shorter half life and because effective binding surface interactions will b e harder to optimize These structural obstacles have certainly directly impact ed the ability for researchers to find a selective inhibitor. Figure 1 5 X ray structure of Mcl 1 (PDB ID 3MK8) showing the binding pockets P1 P4 17

PAGE 25

9 (1) Figure 1 6 Molecular structure of ABT 737 (1) The resea rch groups focused on finding selective Mcl 1 inhibitor s have had limited success thus far Such an inhibitor, or BH3 mim etic, would function to allow a cell that is overexposing Mcl 1 and thus essentially immortalized, to reestablish its ability to undergo apoptosis. Th is function would hopefully lead to tumor growth inhibition in vivo and has already been shown to restore chemosensitivity in chemoresistant cells. Indeed, studies have already shown that certain cancer cell lines are dependent on Mcl 1 to proliferate, and its ablation results in cell death. Therefore, BH3 mimetic drugs are now being explored in order to bind excess cellular Mcl 1 in cells overexpressing this important anti apoptotic protein. Due to the high degree of expression in cancer cells, this potential drug could have far reaching effects in cancer therapeutics. A s mentioned earlier, to date very littl e in vivo data has been obtained and clinical trials have yet to be established for any selective inhibitor of Mcl 1. Yet, several research groups have been successful in vitro with finding a Mcl 1 selective inhibitor. A s urvey of c urrent Mcl 1 i nhibitors DNA i ntercalating a gents DNA intercalat ing agents have been useful in cancer research for many years, including ones that are used in current treatments for Hodgkin's Lymphoma 18 The se N N N H O S NH S N NO 2 Cl O O

PAGE 26

10 compounds function to stop replication of the cancer cells DNA and e ven cause DNA breaks thereby heralding the initiation of apoptosis Normally planar and rigid (Figure 1 6) DNA intercalators insert themselves in between the base pairs of the DNA strand which effectively creates mismatched base pairs and interferes wit h DNA synthesis machinery This process can also cause DNA repair mechanisms to malfunction or turn off and thus many DNA intercalating agents are carcinogenic as well. Develop ment of S1 ( 2 ) was intended to be a DNA intercalating agent, but after later studies it was shown to have no such action. However, this molecule did show strong anti tumor activity. After several iterations of characterizing the molecular structure of S1, Song et al 19 were able to determine the cur rent accepted structure ( 2 ) It w as originally thought of as a pan Bcl 2 family inhibitor, as it had a dissociation (or binding) constant ( K d ) of 58 n M affinity for Mcl 1 and a K d of 310 n M affinity for Bcl 2, but further studies indicated that it actually acted as an upregulator for BH3 only Noxa 20 which selectively binds Mcl 1 and also contributes to its degradation Other anti cancer cell acti vity involving Endoplasmic Reticulum (ER) stress that initiates a separate pathway for cell apoptosis also ha s been implicated, possibly making th is compound not a tru e BH3 mimetic. Other S1 derivatives have been synthesized in order to improve S1 activity, with some even increasing affinity to Mcl 1 six fold, but no in vivo studies are reported As illustrated with S1 ( 2 ), a compound with significa nt biological activity, these pathways are complex and intertwined enough that what may be thought of as a Mcl 1 inhibitor actually might have a cytotoxic action in a different manner.

PAGE 27

11 ( 2 ) Figure 1 7 M olecular structure of S1 ( 2 ) a proposed Mcl 1 inhibi tor Peptides and p eptide m imetics Mimicking the BH3 only members selective affinities to the anti apoptotic members of the Bcl 2 family has also been employed in pharmacologic research. For instance, because of the selectivity of Noxa for Mcl 1 a design ed Noxa BH3 mutant peptide was created and was found to bind to Mcl 1, Bcl X L A1 and Bcl w. This method has been employed usually to characterize the BH3 binding grooves, but short peptides have also been created to act as drug agents themselves Unfort unately short peptides are challenging to develop as drugs because their binding affinities are typically low most likely due to the lack of secondary structure. In addition, peptides rarely progress to clinical trials because of their poor drug propertie s: typically short half lives and poor oral absorption. To solve the secondary structure problem, a method called "stapling" was developed that replaces non essential amino acids in the peptide with ones that are non natural and are able to connect to each other via a metathesis reaction. This essentially locks the peptide into a helical arrangement, giving the strand the secondary structure it needs to interact with the BH3 binding groove fully (Figure 1 7) 21 Unlike the parent peptides, these "stapled" str ands have shown the ability to penetrate the cell membrane and have longer half lives, probably due to the O CN CN N S

PAGE 28

12 locked in secondary structure shielding them of degradation by potential proteases (amide bond cleavage) Figure 1 8 X ray crystallography of a st abilized alpha helix of Bcl 2 domains (SAHB D ). B ased on the BH3 binding domain of Mcl 1 itself, binding to Mcl 1 ( PDB 3MK8) Also illustrates the "staple" applied to the peptide (in light blue). This stapling was applied to Noxa BH3 only peptide strand kn own to bind selectively to Mcl 1. Muppidi et al. were able to obtain nanomolar affinities for Mcl 1 (K d = 4.9 nM) while also maintain ing a high selectivity (200 fold more selective to Mcl 1 than to Bcl X L ) 22 In addition, cell membrane permeability was more effective than non stapled peptides and proteolytic stability was vastly increased over native Noxa. To date no known in vivo studies have been reported.

PAGE 29

13 Pro digiosin d erivatives After a screening of natural products was conducted to find a potent Mcl 1 inhibitor, a group of molecules called p rodigiosin es were identified as potential compounds with activity Prodigiosin ( 3 ) is a secondary metabolite of Serratia marcescens that is responsible for the red pigment seen when observing the bacterium Its de fining structural characteristic is a system of three pyrrole rings linked together. This key structural element was found to be involved in the binding to Mcl 1 in further studies, which prompted development of new analogues with this moiety ( 3 ) Figure 1 9 P rodigiosin ( 3 ) molecular structure One such compound was found to be a pan Bcl 2 family inhibitor and subsequently was synthesized by Gen X This compound originally was named o batoclax ( 4 ) now named GeminX has an indole ring attached to two linke d pyrrole rings Activity of o batoclax ( 4 ) for Bcl 2 family members was found to be in the low micromolar range ( dissociation constant of the Mcl 1 inhibitor complex [ Ki ] of 0.5 7 M), however this relatively poor affinity has been also attributed to the compounds poor solubility. Phase I/II c linical trials were initiated in 2008, but as of January 2015, the study has been terminated due to neurotoxic effects as well as slow accrual of data and inadequate drug supply 23 N H N H N O

PAGE 30

14 ( 4 ) Figure 1 10 Obatoclax ( 4 ) molecular structure Gossypol and g ossypol d erivatives Gossypol ( 5 ) is a natural product derived from the cotton plant that has been long used in therapeutics. This compound had alread y shown effective pharmacokinetic properties with low toxicity. Furthermore, recent s tudies have shown that this compound binds effectively to Bcl 2 (K i = 320 nM), Bcl X L (K i = 480 nM) and Mcl 1 (K i = 180 nM). Ascenta Therapeutics has therefore developed t his compound and named it AT 101, which has entered into multiple clinical trials with some encouraging results However, as we see from the binding affinities, this molecule is a pan inhibit or and not selective for Mcl 1. ( 5 ) Figure 1 1 1 Gossypol ( 5 ) mo lecular structure Numerous derivatives of gossypol have been developed as well, with varying degrees of success. Most bind t o Bcl 2, Bcl X L and Mcl 1 in the submicromolar range and do have moderate cytotoxic activity. Yet just like Gossypol itself no ne o f the N H N O HN HO HO O H OH HO O H OH OH

PAGE 31

15 reported derivatives of Gossypol have been selective for Mcl 1 n or ha ve any of them been tested in vivo Sulfonamides One of the most intriguing categories of current Bcl 2 family inhibitors includes the sulfonamides Antibacterial agents contain ing the sulfonamide functional group ( Figure 1 1 2 ) have long been used and other uses have been reported since the 1930's. Drugs that contain this functional group sometimes do elicit allergic responses, and solubility issues are a concern as well. However they have long been in use and have been the subject of many patents over the years. Figure 1 1 2 The sulfonamide functional group The most prominent sulfonamide used as a Bcl 2 family inhibitor is ABT 737 ( 1 ) This compound was developed by first usi ng an NMR based screening of a chemical library to find small molecules that bind to the hydrophobic BH3 groove of Bcl X L Two fragments were selected (K d = 0.3 mM and K d = 4.3 mM ) that interacted with three specific residues known to be involved in BH3 bi nding of Bak with Bcl X L By connecting these fragments with a linking chain that includ ed a sulfonamide functional group to allow for the correct angle into a specific hydrophobic pocket much higher affinity was achieved (K i = 36 nM). Challenges arose wi th this compound that it was much less effective in human serum upon further testing. However, a fter human serum albumin (HSA) binding was limited using structure based design optimization and then c oncurrent synthesis, the compound known as ABT 737 ( 1 ) w as found. This analogue R 1 S N R 3 R 2 O O

PAGE 32

16 showed two to three fold better affinity (Ki 1 nM) towards Bcl 2, Bcl X L and Bcl w that had been reported at the time (2005) ABT 737 ( 1 ) also showed effective cytotoxicity in numerous cell lines and animal models. The drawbacks of ABT 737 ( 1 ) are that it has a very large molecular weight too many functional groups, many rotatable bonds, and is in general not very drug like These complications have made it orally unavailable. However, ABT 263 ( 6 ) has also been developed to ci rcumvent some of these problems without reducing potency, and has been involved in multiple successful clinical trials. Not surprisingly, due to the lack of affinity of ABT 263 ( 6 or ABT 737 1 ) for Mcl 1, ineffectiveness of this drug in certain cancers ha s been linked to Mcl 1 mediated resistance. Therefore, a combination of ABT 263 ( 6 ) and a novel specific Mcl 1 inhibitor could in theory drastically change the landscape of cancer therapy. (6) Figure 1 1 3 Molecular Structure of ABT 263 (6) To this end, at least three research groups have reported compounds with specific motifs (notably the sulfonamide motif) common to ABT 737 ( 1 ) merged with other fragments know n for Mcl 1 affinity. These studies ha ve yielded several promising pharmacophores with affinit ies to Mcl 1 of IC 50 = 88 nM (Takeda Pharmaceutical Company 7 ) and IC 50 = 30 nM (AbbVie 8 ). No in vivo data has been reported as of yet. N N N H O S NH S N SO 2 CF 3 Cl O O O

PAGE 33

17 Figure 1 1 4 Takeda sulfonamide ( 7 ) derivative and AbbVie sulfonamide ( 8 ) derivative. Pyridine based r hodanine d e rivatives One research group (A*STAR) has investigated pyridine based Rhodanine derivatives of a compound named BH3I 1 ( 9 ). This small molecule is known to bind to Bcl 2 at the BH3 binding site. These derivatives ( such as 1 0 ) were reported to have binding affinities in the low micromolar range (K d 10 M), however later studies by other groups showed no biological activity even in combination with ABT 737 24 Figure 1 1 5 Molecular Structure of BH3I 1 ( 9 ) and one of the developed derivative s ( 1 0 ). Br S N O O O OH N S N O O O OH O O (9) (10) O N N O OH N H S O O NO 2 N H S N OH O H N S O O O N N (7) (8)

PAGE 34

18 Othe r k ey s mall m olecule i nhibitors Development of a 7 hydroxyquinoline derivative ( 1 1 ) by Eutropics Pharmaceuticals has yielded a promising result. This compound has shown moderate affinity for Mcl 1 (IC 50 = 310 nM) and Bcl X L (IC 50 = 40 M). In addition, th e pharmacophore has shown dose dependent cytochrome c release, and has inhibited growth of several Mcl 1 dependent cancer cell lines. In vivo studies utilizing this analogue have not been reported as of yet however tests indicate d that is physically unsta ble. ( 11 ) Figure 1 1 6 A 7 hydroxyquinoline derivative Mcl 1 inhibitor Vanderbilt University has most recently (2013) developed an exciting derivative utilizing a fragment based search (Figure 1 16) Their NMR based fragment library search yielded two di stinct classes of fragments that bound to Mcl 1 in two separate sites, but very close to each other. By merging the best fragment from each class ( 1 2 and 1 3 ) with each other, they hypothesized a better affinity for Mcl 1. Indeed, they achieved higher affin ities for Mcl 1 (K d = 0.32 M for 14 vs. K d = 131 M and K d = 60 M for the fragments 12 and 13 respectively ) with the resulting compound ( 1 4 ) and after further structure based design was performed they found an analogue ( 1 5 ) that achieved a K d of 55 nM for Mcl 1. In addition, thi s final merged compound showed 16 fold selectivity for Mcl 1 over Bcl 2 and 270 fold selectivity over Bcl X L Most important in N N N CF 3 OH

PAGE 35

19 this study was the cocrystal structures of the resultant merging of the two fragments did indeed bind to Mcl 1 in the same disti nct BH3 binding pockets on Mcl 1. Arginine 263 was labeled as an essential residue for this binding with Mcl 1 as it interacts with the carboxylic acid of the compound. Further in vivo results have yet to be reported. Figure 1 1 7 Vanderbilt study on a novel Mcl 1 Inhibitor Merging of two fragments ( 1 2 and 1 3 ) to yield ( 1 4 ) and further optimization yielded a compound ( 1 5 ) with high selectivity for Mcl 1 and high affinity (K d = 55 nM) University of Colorado at Denver Julian r esearch g roup f ocus As of January 2015, there have been several small molecule inhibitors (BH3 mimetic s ) for Mcl 1 identified however very little in vivo data has been obtained and very few had selectivity for Mcl 1 Efforts to find these inhibitors mainly has involved NMR based s creening of vast databases of molecules to observe if they have the potential f or selectively binding to Mcl 1 This is followed by performing the requisite assays to determine if there is any significant binding affinity for Mcl 1 and also biological activ ity on Mcl 1 dependent cell lines Further optimization and subsequent synthesis then follows. Finally, the molecules are tested in vitro to determine their affinity. H N O Cl Cl O OH S O Cl O OH S O OH O Cl OH O Cl Step 1 Step 2 (12) (13) (14) (15)

PAGE 36

20 Our research group's approach to finding a selective Mcl 1 inhibitor is to use structure based design created from key Mcl 1 to ligand interactions in order to determine potential synthetic targets In this way, we can design and synthesize a unique pharmacophore and then functionalize the molecule with side chains and groups that will increa se affinity for Mcl 1 In summary, the research goal is to develop a biologically active, novel small molecule that has high affinity and selectivity for Mcl 1 using structure based design. This research approach began with molecular docking simulations p rovided by Dr. Hai Lin 's research group of the University of Colorado at Denver. Initial molecular docking began with obatoclax ( 4 ) binding with Mcl 1 (Figure 1 1 8 ) to elucidate the key interactions that would give insight into future synthetic targets. Ke y interactions from this modeling show ed that the indole N proton and the central pyrrole nitrogen of obatoclax function as a key binding interface for Mcl 1. This modeling interaction showed hydrogen bonding occurred at Histidine 252 of Mcl 1. A scoring f unction ( C score ) value of 2 .0 was identified for obatoclax ( 4 ), which is an indication of how likely this interaction with Mcl 1 would take place (higher numbers on a scale of 0 to 10, suggest more likely interactions ) This scoring function also gives u s a baseline for further ranking of individual pharmac o phore candidates.

PAGE 37

21 Figure 1 1 8 Molecular docking of obatoclax ( 4 ) to Mcl 1 (gray). a) positioning of obatoclax ( 4 ) binding with Mcl 1, b) obatoclax ( 4 ) forms a hydrogen bond with His 252 (green) o f Mcl 1 in this docking simulation. a) b)

PAGE 38

22 This data le d the research team to investigate a novel analogue scaffold utilizing a similar binding motif. Initially, structure based design lead to the investigation of a pyridine ring attached to an indole ring at th e 2 position ( Scheme 1 1 ). A one carbon linker attached to the pyridine ring at position 6 allowed attachment of a dimethylpyrrole group ( 16 ) As the dimethylpyrrole was found to not be a contributor for molecular docking to Mcl 1, expl oring other derivati ves (cyclic and non cyclic) of the indole p yridine core ( 17 ) became the primary focus. Scheme 1 1 Structure based design scheme for the development of a novel Mcl 1 inhibitor. In the chapters that follow, eight novel derivatives of this indole pyridin e core are explored in terms of their molecular docking, synthesis, and biological activity. NH N HN O NH N HN NH N X Obatoclax (4) Structure-based design target (16) Indole pyridine core (17)

PAGE 39

23 CHAPTER II THE HETEROCYCLE DERIVATIVES This chapter will focus on the process by which our research group has initiated the attempt to find a selective Mcl 1 in hibitor as well as introduce two derivatives of the indole pyridine core : the pyrrole and the thiophene. The m olecular docking method used to evaluate each potential derivative will first be described. Secondly, the synthesis of the indole pyridine core wi ll be examined. Finally the first two proposed analogues involving five membered heterocycles attached to the pyridine ring will be evaluated. Molecular d ocking m ethod As mentioned previously, molecular docking began with obatoclax ( 4 ) to clarify the in teraction this inhibitor was making with Mcl 1. The human Mcl 1 structure was retrieved from the Protein Datab ank (PDB # 2PQK ) and then Tripos' Sybyl 8.0 Flexidock program was used to edit the structure and add any missing side chains or terminal residues Each ligand structure (in this case obatoclax ( 4 ) was formed into a 3D model using Wavefunction Spartan TM Once the geometries were optimized by finding the lowest energy conformations atomic charges were added in two distinct ways: electrostatic potenti al charge (ESP) or natural charge (NA). Starting geometries of the protein ligand complex were then built and rotatable bonds for both the ligand and the protein were added (if necessary). Using genetic algorithms, the global minimums for the configurati ons were then found using scans of 100 configurations per atomic charge method (ESP and NA). Groups of configurations within a four angstrom distance from each other using root mean squared deviation ( RMSD ) of the resulting geometries (termed "clustering" or "clusters") were found and the areas with the maximum clustering were recorded for future comparison. Higher amounts of configurations (out of 100)

PAGE 40

24 within a cluster indicate a higher likelihood of that specific binding mode with Mcl 1 to occur. C scores were also collected, which provides more angles of view to get a consensus on the best docked geometry. Higher amounts of geometric orientations of the molecule in the cluster relates to more probability in the binding configuration as does higher C scor es and lower relative energies. However, it should be noted that these values only indicate a slightly higher level of probability and do not constitute reasonable proof of interactions. ( 4 ) Figure 2 1 Obatoclax ( 4 ) molecular structure When this proces s was applied to obatoclax ( 4 ), a possible binding relationship was observed. As mentioned earlier, a C score of 2.0 and a clustering of 4 configurations at the best orientation were obtained. This was illustrated in Figure 1 1 8 previously, and we can see the interactions more closely in Figure 2 2 below that also shows the hydrop hobic pocket of Mcl 1 (red). A h ydrogen bond was observed in the simulation between our ligand (obatoclax) and Histidine 252 of Mcl 1 as well. Relative energies of this compounds' interaction with Mcl 1 was recorded as 77.4 kcal/mol All of these values will give us valuable insight as we analyze other compounds in relation to obatoclax ( 4 ) N H N O HN

PAGE 41

25 Figure 2 2 Obatoclax's ( 4 ) best configuration with Mcl 1 (PDB # 2PQK ) using molecular d ocking. Red indicates hydrophobicity, blue indicates hydrophilicity. The pyrrole derivative: 2 (6 ((3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) Molecular docking of the pyrrole derivative of the indole pyridine core ( 18 ) Initial mole cular docking of obatoclax led our research group to pursue a pyridine based analogue ( 24 ). This molecule would be more flexible than obatoclax ( 4 ) with no double bond between the pyridine and the dimethylpyrrole. In addition, the core interactions should still be in place with the indole N proton and the pyridine nitrogen available for hydrogen bond donating (indole) and accepting (indole and pyridine) Thus, we should expect the same Mcl 1 residue interaction with this pyrrole derivative as was made with obatoclax (His 252).

PAGE 42

26 Figure 2 3 Molecular s tructure of 2 (6 ((3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) By using the m olecular d ocking method explained previously, the results obtained were similar in nature to obatoclax ( 3 ). A C s core of 1.83 was found with a maximum clustering of 6 for this molecule Although the C score was slightly lower than for obatoclax, the higher clustering was an improvement. A n average relative energy was found to be 107.4 kcal/mol that was much lower th an obatoclax ( 77.4 kcal/mol). As can be seen in Figure 2 4a, this molecule orients slightly below obatoclax and yet it to projects a little further into the hydrophobic pocket below. It therefore could be advantageous to build hydrophobic groups off of th e pyrrole to increase affinity. Hydrogen bonds indicated an interaction with His 252 as expected, however as can be seen in Figure 2 4b, the molecule appears to be forming an internal hydrogen bond and the interaction between His 252 and the molecule is no t clearly defined. In all, this indicated a worthwhile pursuit to synthesize this compound. An account of that process follows. NH N HN (18)

PAGE 43

27 Figure 2 4 Molecular docking of pyrrole derivative ( 18 ) to Mcl 1 (gray), a) positioning of 18 binding with Mcl 1, b) 18 for ms a hydrogen bond with His 252 (green) of Mcl 1 in this docking simulation. a) b)

PAGE 44

28 Synthesis of the indole pyridine c ore In order to develop a novel, specific Mcl 1 small molecule inhibitor, our strategy was to creat e a new core that effectively mimicked the fu ndamental interactions that obatoclax ( 4 ) made with Mcl 1 but with improved affinity, increased potency and less biological side effects (neurotoxicity) After evaluation by molecular docking, it was determined that the indole hydrogen and the middle pyrro le nitrogen wer e the most crucial for hydrogen bond donating and accepting, respectively. An indole pyridine pairing was determined to be a good option after several possibilities were examined (Figure 2 3). This core structure would enable the same intera ctions with Mcl 1 to take place. However, improving interactions for the side chain could very well help the molecule orient in a better manner. Therefore, building a core with a possibility for extensions off of the 6 position of the pyridine was also an important step. This was accomplished by synthesizing a N protected indole pyridine carboxaldehyde ( 19 ). Scheme 2 1 Conceptual r etrosynthesis for Mcl 1 inhibitor derivatives. Protection of the N proton of the indole ring To ensure no further reactivi ty occurred with the N proton of the i ndole ring ( 20 ) in subsequent reactions, a tert butyloxycarbonyl ( Boc ) protection was used. This reaction, shown in scheme 2 2 is commonly used to protect amine functional groups and is easily removed with a strong ac id, such as trifluoroacetic acid (TFA). A final Boc NH N HN NH N X Structure-based design target (16) Indole pyridine core (17) N N H O Indole pyridine carboxaldehyde ( 19 ) O O

PAGE 45

29 protected indole ( 21 ) was produced with a yield of 100% and was used in subsequent reactions. Scheme 2 2 Protection of the i ndole N proton. Formation of N Boc indole 2 boronic acid ( 22 ) In preparatio n for a carbon carbon bond forming reaction, a known boronic acid ( 22 ) was created from the Boc protected indole ( 21 ). Lithium diisopropyl amide (LDA), a strong base, was prepared in order to extract the proton on the 2 position of the indole ring. This is the preferential site (most acidic) for proton extraction after the protection was performed. Once this proton is extracted, triisopropyl borate exchanges one of its three ligands (isopropoxide) for the indole ring. An aqueous workup allows the boron to e xchange all of its ligands for protons, which affords the final boronic acid product ( 22 ) in good yield (74%). Scheme 2 3 Boronic acid ( 22 ) synthesis. Suzuki Miyaura cross coupling to form N Boc protected indole pyridine core ( 19 ) One of the most powe rful carbon carbon bond forming reactions in organic synthesis, the Suzuki Miyaura cross coupling reaction, was used to obtain the final N H N O O NaH, THF 100% Boc anhydride (20) (21) N O O N B(OH) 2 O O LDA, THF then B(O i Pr) 3 0 o C, 74% (21) (22)

PAGE 46

30 product for the indole pyridine core ( 19 ) This reaction utilizes an organoboron compound and a halide as reactants, a palladium catalyst, and a base in order to form a carbon carbon single bond bridging the two reactants. This process invol ves a catalytic cycle (Figure 2 5 ) with three major sequential steps: oxidative addition, transmetalation, and reductive elimination. Figure 2 5 Catalyt ic cycle of the Suzuki Miyaura cross coupling r eaction 25 In the oxidative addition step, reduced palladium catalyst (Pd 0 ) is oxidized to Pd II as it is inserted in between the halide and carbon, breaking that bond. This forms an organ opalladium complex that undergoes further reaction with the base in the reaction to form a salt (in this case NaX) and the organopalladium base complex. This complex then undergoes transmetalation, which involves the exchange of the ligands of the organobo ron compound with the organopalladium compound. In the Suzuki Miyaura reaction, this results in a organopalladium species with the oxidized palladium flanked by

PAGE 47

31 two carbon ligands. This process also results in a transfer of the base to the boron species. T he role of the base is believed to participate both in the activation of the organoboron species and by creating the aforementioned organopalladium base complex. Finally, the reductive elimination step forms the final carbon carbon bond and releases a redu ced palladium catalyst. The resulting product is formed and the catalyst is renewed, making it available to be used in sequential cycles. In the synthesis of obatoclax ( 4 ) by Da•ri et al. 26 the Suzuki Miyaura cross coupling reaction (Scheme 2 4) was used to couple an indole 2 boronic acid ( 23 ) to a bromo 2 enamine ( 2 4 ). This reaction also involved hydrolysis of both the enamine and the N methoxy carbonyl groups to yield the final indolylpyrrole aldehyde compound ( 2 5 ). This was the crucial step as it provid ed the most difficult carbon carbon coupling in the synthesis. Scheme 2 4 Suzuki Miyaura cross coupling reaction in an obatoclax synthesis by Da•ri et al. To synthesize the indole pyridine core, the same approach was used with optimized conditions. I n the Suzuki Miyaura cross coupling reaction performed by Li et al. 27 tripotassium phosphate was used as the base, palladium (II) acetate was the catalyst, n butanol was the solvent, and a ligand (2 dicyclohexylphosphino 2,6 dimethoxybiphenyl or SPhos) was added to aid p alladium in the oxidative addition The starting materials were commercially available 6 bromo 2 pyridine carboxaldehyde ( 2 6 ) and N (tert butoxycarbonyl) indole 2 boronic acid ( 22 ) that was prepared previously. N O O B(OH) 2 + N Br OMe N K 3 PO 4 Pd(PPh 3 ) 4 dioxane, toluene H 2 O, 80-90 o C N H N H OMe H O (23) (24) (26)

PAGE 48

32 This reaction was further opti mized by using degassed 1,4 dioxane/10% v/v water as the solvent (Scheme 2 5 ). This reaction afforded a 72% yield of the tert butyl 2 (6 formylpyridin 2 yl) 1H indole 1 carboxylate ( 19 ), the target for this phase of the project. By creating the indole pyri dine core with an aldehyde at the 6 position on the thiophene ring, it provided a good functional group handle for future derivatives since numerous reactions can utilize this moiety Scheme 2 5 Suzuki Miyaura cross coupling reaction, yielding the indole pyridine core ( 19 ). Acid catalyzed coupling of dimethyl pyrrole ( 2 7 ) to the indole pyridine core aldehyde ( 19 ) To complete the final carbon carbon bond of this derivative, the same coupling reaction that was used by Da•ri et al. was employed by reactin g the indole pyridine carboxaldehyde ( 19 ) with the dimethylpyrrole ( 27 ) under acidic conditions (Scheme 2 6 ). By using catalytic amounts of HCl, protonation of the oxygen of the aldehyde carbonyl group occurs. Subsequent nucleophilic attack by the 5 positi on of the dimethyl pyrolle ring, the most reactive site on the molecule due to the resonance stability of the protonated intermediate, enables the coupling to occur. After a dehydration step, the resulting olefin is obtained. An 84% yield of tert butyl 2 ( 6 ((3,5 dimethyl 2H pyrrol 2 ylidene)methyl)pyridin 2 yl) 1H indole 1 carboxylate ( 2 8 ) was obtained in this manner. N N N Br O O K 3 PO 4 Pd(OAc) 2 SPhos, 1,4 dioxane, 10% v/v H 2 O N B(OH) 2 O O + H H O O (22) (19) (26) 72%

PAGE 49

33 Scheme 2 6 Dimethylpyrrole ( 2 7 ) coupling to the indole pyridine core ( 19 ). Reduction of the olefin linker of the indole pyridine pyrrole ( 28 ) Reduction of the double bond between the dimethylpyrrole and the pyridine proved difficult to achieve. A simple palladium on carbon reduction proved ineffective at atmospheric pressure most likely due to resonance stability of the olefin conjugate d with the pyrrole double bonds. Next a sodium triacetoxyborohydride reduction was tried, which is a gentler reductive agent than sodium borohydride. Though most typically used for reductive aminations of aldehydes and ketones, this compound has been known to reduce certain olefins. However in this case the reduction did not proceed. Finally, a sodium cyanoborohydride reduction was performed and even though some degradation occurred, the product tert butyl 2 (6 ((3,5 dimethyl 2H pyrrol 2 ylidene)methyl)pyri din 2 yl) 1H indole 1 carboxylate ( 2 9 ) was isolated with a 59% yield (Scheme 2 7 ). Interestingly, some final product of unprotected 2 (6 ((3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 24 ) was isolated during this process, as the conditions w ere acidic enough to remove the Boc protecting group as well. Even after multiple purifications, there were still some impurities present in the sample. The compound was carried forward, however, in the hope that after deprotection, purification would be e asier. N N O H N HCl, MeOH H 84% O O (19) (27) N N N O O (28)

PAGE 50

34 Scheme 2 7 Reduction of the olefin linker of the indole pyridine dimethylpyrrole compound ( 2 8 ) Deprotection of the indole nitrogen on the pyridine N Boc indole pyrrole ( 2 9 ) Final deprotection of the indole nitrogen proved facile, however challe nges with purifying the compound occurred. U ltimately it was decided that the compound is ustable. To deprotect the indole nitrogen from compound ( 2 9 ), a traditional TFA deprotection in d ichloromethane (DCM) was used (Scheme 2 8 ). The Boc ( tert butylcarbox ycarbonyl) group undergoes tert butyl carbocation cleavage in highly acidic conditions, followed by carbon dioxide release, leaving the nitrogen unprotected. From the TLC and LCMS analysis, it was observed that the product was degrading. Complete conversio n to the final product ( 18 ) was indicated via LCMS after stirring overnight. However, after numerous attempts to isolate the final product, it was determined to be unstable at room temperature likely due to oxidative decomposition processes Further effort s to pursue this derivative were halted. N N N MeOH, THF NaBH 3 CN O O N N O O HCl 59% (28) (29) N H

PAGE 51

35 Scheme 2 8 TFA deprotection to yield the final product, 2 (6 ((3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) The thiophene derivative: 2 (6 (thiophen 2 ylmethyl)pyridin 2 yl) 1H indole ( 30 ) Molecular docking of thiophene derivative ( 30 ) Challenges overcoming an intra molecular hydrogen bond in the pyrrole derivative ( 18 ) modeling led the research team to investigate other heterocycles attached to the pyridine ring. Furan and thiophene were both modeled attached via a one carbon linker to the indole pyridine core so that this intra molecular hydrogen bond would no longer form, thus freeing the derivative to bind to Mcl 1 more liberally The thiophene derivatives all consistently had higher C scores and cluster numbers on average than the furans, likely because the hydrophobic pocket was too non polar for the more polar furan moiety With this information in mind, the base thiophene derivative was pursued further ( 30 ). Figure 2 6 Molecular s tructure of thiophene derivative ( 30 ) Molecular docking data indicated the thiophene as a strong candidate for binding to Mcl 1. A C score of 3.29 with a clustering of 6 configurations (of a 100) was found at NH N HN DCM N N O O TFA (29) (18) N H (30) NH N S

PAGE 52

36 the best geometrical arrangement (Figure 2 7) which is an improvement over previous compounds Finally, a n average relative energy of 46.75 kcal/mol was determined, which was much greater than obatoclax ( 77.4 kcal/mole) or the pyrrole derivative ( 107.3 kcal/mol) indicating a less favorable inter action The thiophene derivative was shown to interact via a hydrogen bond with Histidine 224. This is a departure from the area that obatoclax interacted with Mcl 1, and a different histidine residue, but this placement actually allows for possible struct ure based design sites to enhance Mcl 1 interactions based off of the positioning of the molecule. a)

PAGE 53

37 Figure 2 7 Molecular docking of thiophene derivative ( 30 ) to Mcl 1 (gray), a) positioning of 30 binding with Mcl 1, b) 30 forms a hydrogen bond with Hi s 224 (green) of Mcl 1 in this docking simulation. As can be seen in Figure 2 7a, the thiophene analogue 30 orients itself between the P2 (above) and the P1 pocket (below). This orientation could provide a unique opportunity to extend the molecule in both directions to fill the binding pockets further. If extensions were built off of the indole ring at the 4 position, it might be possible to fill the P1 pocket. On the other hand, if groups were built off of the 5 position of the thiophene ring, the P2 pock et could be occupied as well. This could lead to an increase in binding affinity and a better drug candidate. Thus synthesis of the thiophene molecule as well as subsequent binding affinity evaluation was performed b)

PAGE 54

38 Thiophene coupling to indole pyridin e carboxaldehyde ( 19 ) A coupling to attach the thiophene heterocycle to the indole pyridine core ( 19 ) was next attempted. A reaction by Kelly et al. 28 that involved a nucleophilic addition to a furan aldehyde with phenyl lithium gave the basis for the coup ling reaction used in our synthesis. In this case, the authors achieved a 92% yield of the final product, a diaryl alcohol. Carey and Sundberg 29 also provided information on the reaction. Using the indole pyridine carboxaldehyde ( 19 ) and thienyl lithium ( 31 ) as starting materials, a final yield of 68% of the biaryl alcohol ( 32 ) product was obtained (Scheme 2 9) Scheme 2 9 Nucleophilic addition of thienyl lithium ( 31 ) to an aldehyde ( 19 ). Reduction of the b iaryl alcohol ( 3 2 ) From the biaryl alcohol ( 32 ) a triethyl silane reduction of the alcohol group provide d the N protected final product. A reaction reducing a substituted biphenyl methanol by Baguley et al. 30 was found as precedence for reducing a biaryl alcohol. This reaction also included adding TFA along with the silane, which creates an acidic environment and accelerates the reaction. However, this also deprotects the N protected indole ring of the biarylalcohol ( 3 2 ), thus the final product actually was the 2 (6 (thiophen 2 ylmethyl)pyridin 2 yl) 1 H indole target ( 30 ). The reduction potential of silanes has long been effective means to reduce carbonyl group to alcohols and alcohols to alkanes. Triethyl silane in this case was used as a hydride source in acidic media (TFA) to react with the carboca tion (or carbocation N N O S Li H O O (19) (31) N N OH O O (32) S Et 2 O, -78 o C 68%

PAGE 55

39 like complex induced by the silane) intermediate to form the alkane. These reactions in conjunction led to a final product ( 30 ) yield of 53% (Scheme 2 10 ). Scheme 2 10 Triethyl s ilane reduction of b iaryl alcohol ( 3 2 ). Binding a ffin ity t esting of the thiophene derivative to Mcl 1 A surface plasmon resonance (SPR) assay (Figure 2 8) on the thiophene derivative ( 30 ) was performed by Dr. Scott Kaufmann's group at the Mayo Clinic in order to determine the level of affinity for Mcl 1. Th is assay determines the association rate (k a ) and the dissociation rate (k d ) of a ligand protein interaction so that the binding constant (K d ) can be calculated. A single layer of the protein is bound to a metal chip by amine coupling at a high density. Th e ligand (analyte) is washed over the chip and allowed to bind to the protein for a short period of time. All the while, incident light is directed through a prism towards the chip. When the plasmon waves come into contact with a bound ligand protein compl ex, the SPR signal increases. The association rate is thus calculated from this increase. The chip is then washed over the same amount of time, and a dissociation rate is found. From this, the bindin g constant can be ca lculated using the following equation : (30) N N OH O O (32) S TFA, DCM NH N S Et 3 SiH 53% K d = k d k a

PAGE 56

40 The advantage of using this technology instead of other assays is that many require using a label to attach to the ligand. In the case of SPR, it is a label free technology. Figure 2 8 Illustration of the Surface Plasmon Resonance (SPR) assa y to d etermine ligand protein binding affinity 31 Pharmacaphore binding to target proteins will have a range of binding affinities, as was seen by the various molecule binding constants reported in the survey of current Mcl 1 inhibitors. Improvements can be made by further structure based design if initial results are promising. C oncentrations of ligand (thiophene derivative, 30 ) were tested at 50 M, 10 M and 1 M and yielded an increase in SPR signal at the 50 M level. This result indicates that this derivat ive ( 30 ) has a binding constant (K d ) of between 10 and 50 M to Mcl 1 and tests to precisely determine the K d are pending. The SPR assay was also performed Prism Gold surface of CM5 chip Dextran Conjugated Protein (Mcl 1) Ligand (Mcl 1 inhibitors)

PAGE 57

41 on positive controls obatoclax ( 4 ) and ABT 263 ( 6 ) to verify the validity of the results. Due to the se promising results, structure activity relationship (SAR) studies and further optimization of this compound should therefore be pursued.

PAGE 58

42 Experimental s ection General experimental d etails When indicated, solvents were degassed with nitrogen. In reacti ons requiring moisture or oxygen free environments, dry solvents were either purchased or they were passed through a solvent system and kept under nitrogen. All glassware was oven dried (170 o C) and under nitrogen just prior to use. All flash column chroma tography was performed with 0.5 10 M silica dioxide. Proton nuclear magnetic resonance ( 1 H NMR) experiments were performed on a Unity "Gemenf 300" spectrometer at 300 MHz or a n Agilent s pectrometer at 400 MHz. Carbon 13 nuclear magnetic resonance exp eriments were performed on the aforementioned Agilent spectrometer at 100MHz The proton signal for non deuterated solvent ( # 7.26 ppm for CHCl 3 ) was used for a reference for 1 H NMR spectra. Chemical shifts for 13 C NMR experiments are reported relative to the CDCl 3 resonance shift at # 77.2 ppm. Coupling constants are reported in Hz. Fourier Transform Infrared (FTIR) spectra were obtained via a Diamond ATR spectrometer. Mass spectroscopy was conducted by a Liquid Chromatography Mass Spectrometer (LCMS) by A gilent Technologies. Elemental Analysis was performed by Atlantic Microlabs, Inc. who uses combustion with automatic analyzers. Analytical High Performance Liquid Chromatography (HPLC) was performed by the Skaggs School of Pharmacy and Pharmaceutical Scien ces at Colorado University Anschutz Campus. Surface Plasmon Resonance (SPR) assays were performed by the Mayo Clinic at 1 M 10 M and 5 0 M concentrations using a Biocore T200 by GE healthcare SPR.

PAGE 59

43 N ( tert butoxycarbonyl) indole ( 21 ) NaH (60 % in m ineral oil, 0.440 g, 11.1 mmol,) was dissolved in dry THF (22 m L ) and cooled to 0 ¡C Indole 20 (1.01 g, 8.63 mmol) was dissolved in a separate flask with dry THF (1.7 m L ) and added dropwise to the NaH mixture. The resultant solution was stirred at 0 ¡C fo r 45 m in D i tert butyl dicarbonate (2.8 1 g, 12. 9 mmol) was added, the solution was heated to rt and then stirred 2 h D ry THF (10 m L ) was added t o rinse the sides of the flask after 1 h of reaction time Progress of the reaction was monitored by TLC (20:1 h exanes:EtOAc). Upon completion, the reaction was c ooled to 0 ¡C and water was added to quench. The organic layer was extracted with diethyl ether (3X), washed with brine (3X), dried over MgSO 4 filtered and then concentrated in vacuo to give a yellow pow der. Purification of the crude product by fla sh column chromatography (20:1 h exanes:EtOAc) afforded the Boc protected indole 21 (1.9 0 g, 100%) as a light yellow powder. C rude : 1 H NMR ( 3 00 MHz, CDCl 3 ) # 8. 15 (d, 1H), 7.65 7.55 (m, 2H), 7.4 0 7.15 (m, 2H), 6.6 0 6.5 0 (s, 1H) 2.0 0 1.6 0 (s, 9H), 1.6 0 1.4 0 (s, 9H, impurities present) N H N O O NaH, THF 100% Boc anhydride (20) (21)

PAGE 60

44 N ( tert butoxycarbonyl) indole 2 boronic acid ( 22 ) Distilled diisop rop ylamine (4.40 m L ) was added to a fla sk containin g dry THF (2.60 mL ) and cooled to 0 ¡C n B utyllithium sol ution in hexanes (23.2 mL, 60.3 mmol) was added dropwise over 0.5 h and then the lithium di isopropylamine (LDA) solution (1:10 THF:hexanes) was stir red for 30 min. N Boc indole 21 (8.95 g, 41.2 mmo l) was dissolved in dry THF (28.6 mL) in a round bottom flask. Triisopropyl borate (14.3 mL) was added and stirred at 0 ¡C for 10 min. The freshl y prepared LDA solution (25.8 mL ) was added dropwise over 50 min. Progress of the reaction was monitored by TLC (5:1 h exanes:EtOAc). More LDA solution (6.0 mL) was added after 1 h Upon completion, the reaction was quenched with 3N HCl to bring the mixture to neutral pH (~7). Extracted the organic layer with diethyl ether (3X), washed with brine (3X), dried over Mg SO 4 filtered and concentrated in vacuo to give a yellow powder. Purification of the crude product by recrystallization (20:1 hexanes: CH 2 Cl 2 ) afforded the boronic acid 22 (7.94 g, 100%) as a pale yellow powder. 1 H NMR (400 MHz, CDCl 3 ) # 8. 02 (d, J = 8 Hz, 1H), 7. 6 1 (d, 8 Hz, 1H), 7. 25 (d J = 3 Hz, 3 H unknown impurity together with 1H from indole ring ), 7. 38 7. 33 (dd, J = 8, 1 Hz, 1H) 7.29 7.23 (dd, J = 8, 1 Hz, 1H) 1. 8 0 (s, 9H) ; LRMS (ES+) m/z calcd for C 13 H 16 BNO 4 [M] + 261.12, fou nd 2 79 .2 [M + H 2 O] + N O O N B(OH) 2 O O LDA, THF then B(O i Pr) 3 0 o C, 74% (21) (22)

PAGE 61

45 tert butyl 2 (6 formylpyridin 2 yl) 1H indole 1 carboxylate ( 19 ) To a sample of indole pyridine boronic acid 22 (500 mg, 1.92 mmol) in a round bottom flask charged with a stir bar was added 6 bromo 2 pyridine carboxaldehyde 26 (3 64 mg, 1.96 mmol) and 1,4 dioxane (4.5 m L). The mixture was stirred for 15 min Palladium (II) acetate (23.5 mg, 0.10 0 mmol), SPhos (79.5 mgs, 0.19 0 mmol), K 3 PO 4 (1.14g, 5.4 0 mmol), and water (0.5 ml) were subsequently added. The r eaction mixture was then heated to 90 ¡C Progress of the reaction was monitored by TLC (10:1 h exanes:EtOAc). Upon completion ( 3 h ), the mixture was cooled to rt, filtered and concentrated in vacuo to give a brown oil. CH 2 Cl 2 was added followed by water and 3N HCl until a neutral pH was achieved The organic layer was extracted with CH 2 Cl 2 (3X), washed with water (2x) and brine (3X), dried over MgSO 4 filtered and concentrated in vacuo to give a yellow powder. Purification of the crude product via flash column chromatography (100% CH 2 Cl 2 ) afforded the indole pyridine carboxaldehyde 19 (423 mg, 71%) as a pale yellow powder. 1 H NMR (300 MHz, CDCl 3 ) # 10.15 10.09 (s, 1H), 8. 18 (d, J = 9 Hz, 1H), 7. 94 (d, J = 4.5 Hz, 2H), 7. 74 (t, J = 5 Hz 1H), 7. 61 (d, J = 9 Hz, 1H), 7.40 (t, J = 9 H z, 1H), 7.32 7.24 (m, 1H), 6.90 6.86 (s, 1H), 1.45 1.20 (s, 9H); LRMS (ES+) m/z calcd for C 13 H 15 NO 2 [M] + 322.13, found 267 .1 [M H 2 tBu] + N N N Br O O K 3 PO 4 Pd(OAc) 2 SPhos, 1,4 dioxane, 10% v/v H 2 O N B(OH) 2 O O + H H O O (22) (19) (26) 72%

PAGE 62

46 tert butyl 2 (6 ((3,5 dimethyl 2H pyrrol 2 ylidene)methyl)pyridin 2 yl) 1H indole 1 carboxylate ( 2 8 ) To a sample of pyridine N Boc indole carboxaldehyde 19 (290 mg, 0.90 0 mmol) was added MeOH (4.5 m L ) and 2,4 dimethylpyrrole 27 (120 L, 1.2 0 mmol). The solution was stirred at rt for 10 min and then HCl in MeOH (1.0 mL, 1. 3 mmol) was added dropwise. The reaction was monitored by TLC and LCMS and showed completion after 45 min. The reaction was quenched with sat NaHCO 3 A small amount of CH 2 Cl 2 was then added The organic layer was extracted with CH 2 Cl 2 (3X), washed with water (2X) and brine (2X), dried over MgSO 4 filtered and concentrated in vacuo to give a reddish brown oil. Purification of the crude product via flash column chromato graphy (2% MeOH in CH 2 Cl 2 ) afforded the indole pyridine dimethylpyrrole olefin 2 8 (302 mg, 84%) as a light yellow powder. LRMS (ES+) m/z calcd for C 25 H 25 N 3 O 2 [M] + 399.48, found 400 .1 N N O H N HCl, MeOH H 84% O O (19) (27) N N N O O (28)

PAGE 63

47 tert butyl 2 (6 ((3,5 dimethyl 2H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole 1 carboxylate ( 2 9 ) To a sample of indole pyridine dimethylpyrrole olefin 2 8 (284 mg, 0.71 0 mmol) was added THF (3.6 ml) and MeOH (0.25 mL). NaCNBH 3 (134 mg, 2.13 mmol) was then added and the mixture was heated to 65 ¡C. The reaction was monitored by TLC and LCMS and showed to be near completion after 24 hrs. A small amount of CH 2 Cl 2 was then added. The organic layer was extracted with CH 2 Cl 2 (3X), washed with sat. NaHCO 3 (2X), water (2X) and brine (2X ) dried over MgSO 4 filtered and concentrated in vacuo Purification of the crude product via flash column chromatography (2% MeOH in CH 2 Cl 2 ) afforded the reduced indole pyridine dimethylpyrrole 2 9 (168 mg, 59%) as a light yellow powder. LRMS (ES+) m/z calcd for C 25 H 27 N 3 O 2 [M] + 401.50, found 402 .2 N N N MeOH, THF NaBH 3 CN O O N N O O HCl 59% (28) (29) N H

PAGE 64

48 2 (6 ((3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) To a sample of indole pyridine dimethylpyrrole 29 (297 mg, 0.74 0 mmol) was added CH 2 Cl 2 (1.0 mL) and TFA (1.0 mL). Additional aliquots of TFA (2 X 1.0 mL) were added over the first h of the reaction. The reaction was monitored by TLC and LCMS and sho wed near completion after 24 h The r eaction was quenched with sat NaHCO 3 until pH was neutral. A small amount of CH 2 Cl 2 was then added. The organic layer was extracted with CH 2 Cl 2 (3X), washed with water (2X) and brine (2X), dried over MgSO 4 filtered and concentrated in vacuo Purification of the crude product via flash column chromatography and recrystallization was ineffective. Further pursuit of this compound was halted due to chemica l instability. 1 H NMR (300 MHz; CDCl 3 ): # 9.86 ( br s, 1H), 7.81 ( br s, 1H), 7.68 7.60 (m, 3 H), 7.42 (d, J = 8.0 Hz, 1H), 7.27 7.22 (m, 1H), 7.16 7.11 (m, 1H), 7.04 (d, J = 1.3 Hz, 1H), 6.99 (dd, J = 6.3, 2.2 Hz, 1H), 5.71 (d, J = 2.4 Hz, 1H), 4.09 (s, 2H) 2.16 (s, 3H), 2.10 (s, 3H ) NH N HN DCM N N O O TFA (29) (18) N H

PAGE 65

49 tert butyl 2 (6 (hydroxy(thiophen 2 yl)methyl)pyridin 2 yl) 1H indole 1 carboxylate ( 32 ) To a portion of indole pyridine carboxaldehyde 19 (195 mg, 0.61 0 mmol) was added dry diethylether (3.0 mL). This solution was allowe d to stir for 10 min and then cooled to 78 ¡C. Thienyl lithium 31 (350 L, 3.22 mmol) was added dropwise. Reaction was allowed to return to rt. The reaction was monitored by TLC (10:1 MeOH: CH 2 Cl 2 ) and LCMS. A fter 1 h the reaction was not progressing O ne aliquot of thienyl lithium 31 (350 L, 3.22 mmol) was added dropwise. T LC and LCMS indicated reaction completion after 45 min The reaction mixture was concentrated in vacuo and then CH 2 Cl 2 and water was added. The organic layer was extracted with CH 2 Cl 2 (3X), washed with water (2X) and brine (2X), dried over MgSO 4 filtered and concentrated in vacuo Purification of the crude product via flash column chromatography (2% MeOH in CH 2 Cl 2 ) afforded the biaryl alcohol 32 (172 mg, 68%) as a light yellow powder. 1 H NMR (300 MHz; CDCl 3 ): # 8.19 (d, J = 8.2 Hz, 1H), 7.75 (t, J = 7.8 Hz, 1H), 7.61 (d, J = 7.7 Hz, 1H), 7.47 (d, J = 7.6 Hz, 1H), 7.39 (td, J = 7.8, 1.1 Hz, 1H), 7.31 7.21 (m, 4H), 7.09 (d, J = 3.2 Hz, 1H), 6.98 (dd, J = 5.0, 3.5 Hz, 1H), 6.83 (s, 1H), 6.07 (s, 1H), 1.38 (s, 9H) ; LRMS (ES+) m/z calcd for C 23 H 22 N 2 O 3 S [M] + 406.14, found 407 .1 N N O S Li H O O (19) (31) N N OH O O (32) S Et 2 O, -78 o C 68%

PAGE 66

50 2 (6 (thiophen 2 ylmethyl)pyridin 2 yl) 1H indole ( 30 ) To a small vial charged with a stir bar was added the biaryl alcohol 32 (51 mgs, 0.12 mmol) and CH 2 C l 2 (0.9 mL). The solution wa s allowed to stir at rt for 10 min and then TFA (0.1 0 mL, 1.3 m mol) was added dropwise. T riethyl silane (0.1 0 m L 0.63 mmol) was added and then the mixture was stir red under nitrogen overnight. The reaction was monitored by TLC (100% CH 2 Cl 2 ) which indicat ed reaction completion after 24 h The reaction was q uenched with sat. NaHCO 3 to neutral pH and a small amount of CH 2 Cl 2 was added. The organic layer was extracted with CH 2 Cl 2 (3X), washed with sat. NaHCO 3 (1X) and brine (2X), dried over MgSO 4 filtered an d concentrated in vacuo Purification of the crude product via flash column chromatography (2% MeOH in CH 2 Cl 2 ) afforded the final thiophene derivative 30 (19 mg, 53%) as a light yellow powder. Further synthesis for complete characterization is required. 1 H NMR (300 MHz; CDCl 3 ): # 9.47 (s, 1 H), 7.68 7.58 (m, 3 H), 7.48 (d, J = 8.2 Hz, 1H), 7.34 (dd, J = 8.1, 1.7 Hz, 1 H), 7.27 7.17 (m, 2 H), 7.13 7.07 (m, 1 H), 7.04 6.91 (m, 1 H), 6.74 (dt, J = 3.5, 1.3 Hz, 1H), 6.59 (s, 1H), 4.40 (s, 2H) ; FT IR (KBr, cm 1 ) 3401, 3058, 2926, 1616, 1590, 1568, 1455, 1437, 1361, 1338, 1306, 1228, 1149, 1118, 1085, 1039, 1012, 992. Anal. Calcd for C 14 H 12 N 2 O: C, 74.45; H, 4.86; N, 9.65 Found: C, 71.02; H, 6.62; N, 8.6. (30) N N OH O O (32) S TFA, DCM NH N S Et 3 SiH 53%

PAGE 67

51 CHAPTER III METHANOL, VINYL, AND ETHYL DERIVATIVES From evaluat ion of the derivative m olecular docking, and structure based design, our research group determined that alkyl and alkenyl chains off of the 6 position on the pyridine ring might create viable fragments with good affinity for Mcl 1. Molecular docking experi ments showed that the heterocycle off of the 6 position of the pyridine ring was not contributing to the binding affinity of the derivatives This led to the decision that a replacement of the heterocycle was needed with: 1) hydrophobic chains to fill the hydrophobic pockets, 2) a hydrophilic alcohol group to test if it would increase affinity By adding these exte nsions off of the pyridine ring, better interactions with Mcl 1 should be possible The resulting fragment derivatives were classified into one, two and three carbon chains and were investigated by molecular docking studies prior to and during synthesis. If moderate affinity was not created by one fragment, building a structure with two or more might be a way to increase binding affinity if differe nt hydrogen bond interactions were observed. C reating two hydrogen bond interactions from the merged compound would afford greater affinity for Mcl 1. The methanol derivative: (6 (1 H indol 2 yl)pyridin 2 yl)methanol (3 3 ) Molecular docking of methanol der ivative ( 33 ) To investigate whether a hydrogen bond donor (or acceptor) off the side chain of the pyridine would be important to the affinity for the core, a methanol derivative off of the 6 position of the pyridine ring ( 3 3 ) was evaluated by molecular do cking studies.

PAGE 68

52 Figure 3 1 Molecular s tructure of (6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 3 3 ) The results of the docking simulations were improved from those of obatoclax ( 4 ) by a significant amount. A C score of 3.5 and a cluster maximum of 8 were obtained from the best molecular configuration of 33 (Figure 3 2a). Both data points were clearly higher than those from obatoclax ( 4 ), indicating a possible improvement on binding affinity. The average r elative energy for this compound 's simulated binding was found to be 70.7 kcal/mol, which is comparable to the obatoclax docking ( 77. 4 kcal/mol) The docking studies also showed a unique binding of the methanol derivative to Mcl 1. While at the top of the BH3 binding groove between the P2 and P3 pockets ( Figure 3 2b), the molecule oriented itself in such a way as to create two hydrogen bonds at a sparagine (Asn) 239 and a spartic acid (Asp) 242. The a spartic acid (aspartate) residues at a physiological pH of 7 are hydrogen bond acceptors, whereas a sparagine can be either a hydrogen bond donator or an acceptor. From the top view shown in Figure 3 2b, the Asn residue (orange) clearly forms an interaction (hydrogen bond) with the alcohol group (hydrogen bond donator) of the methanol derivative. Also shown in Fig ure 3 2b is the interaction between Asp residue (blue), a hydrogen bond acceptor with the indole nitrogen proton. (33) NH N OH

PAGE 69

53 Figure 3 2 Molecular docking of methanol derivative ( 3 3 ) to Mcl 1 (gray), a) positioning of 3 3 binding with Mcl 1, b) 3 3 forms a hydrogen bond with Asn 239 (orange) and with Asp 242 (blue) of Mcl 1 in this docking simulation. a) b)

PAGE 70

54 From this information, and the fact that the synthesis would be facile, the research group deemed it important to synthesize the methanol derivative. A possibility of forming two hydrogen bonds with this fragment could be highly useful for future derivatives. Reduction of indole pyridine aldehyde ( 3 4 ) Reduction of the aldehyde functional group of the indole pyridine core molecule ( 19 ) was accomplished by a simple sod ium borohydride (NaBH 4 ) reduction. Sodium borohydride is known to be an effective reducer of aldehydes and ketones to yield alcohols. Just like triethyl silane, this compound functions as a hydride source that adds to the carbonyl group An aqueous workup will protonate the resulting alkoxide and the alcohol is formed. This reaction was performed successfully to form the Boc protected primary alcohol ( 34 ) with a 93% yield. Scheme 3 1 NaBH 4 reduction of indole pyridine aldehyde. Deprotection of i ndole n i trogen Deprotection of the indole nitrogen on the primary alcohol ( 34 ) was accomplished as previously described with TFA in DCM. This reaction (77%) completed the alcohol derivative ( 3 3 ) of the indole pyridine core. N N O H O O (19) N N O O (34) OH NaBH 4 MeOH 93%

PAGE 71

55 Scheme 3 2 Boc deprotection of N B oc protected indole pyridine methanol derivative ( 3 4 ). Binding a ffinity t esting of (6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 3 3 ) with Mcl 1 Affinity testing using SPR analysis revealed no substantial affinity (at or under 5 0 M ) for Mcl 1. The vinyl derivative: 2 (6 vinylpyridin 2 yl) 1 H indole (3 5 ) Molecular d ocking of vinyl indole pyridine derivative ( 3 5 ) Molecular docking experiments were not performed on the vinyl derivative of the indole pyridine core ( 3 5 ). Figure 3 3 Molecular s tructure of vi nyl indole pyridine derivative ( 3 5 ) (33) TFA, DCM NH N OH 77% N N O O (34) OH (35) NH N

PAGE 72

56 Unstabilized Wittig r eaction of indole p yridine carboxaldehyde ( 19 ) with a m ethyl Wittig r eagent. The Wittig reaction is a powerful reaction for synthetic organic chemists. In fact, Georg Wittig was awarded the Nobel Prize in Chemistry in 1979 for discovering the method. Any carbon carbon bond forming reaction is highly important to organic chemists, and this reaction falls into that category. An olefin is added to a molecule with this reaction, enabling further exten sions of a carbon chain or adding branching and even adding new functional groups to the parent compound. In general, this feat is accomplished by reacting a phosphonium ylide to an aldehyde or ketone (Figure 3 4). Figure 3 4 General schematic for a Wi ttig reaction. Formation of the ylide is the very important first step in the reaction (Figure 3 5), which involves adding a strong base 32 to a Wittig reagent. This Wittig reagent is deprotonated by a strong base and the zwitterionic ylide is formed. The ca rbanion o n the alpha carbon of the ylide is unstable and can be made more or less unstable by the R groups attached to it (R 3 and R 4 ). If for instance both R groups are alkyl groups (or hydrogens), the ylide will be more unstable due to the mildly electron donating capabilities of alkyl groups. This is what is known as an unstabilized Wittig reaction. However, if the R group is an electron withdrawing group, like a carbonyl, the ylide will be more stabilized and less reactive. This is because of the electro n withdrawing R 1 R 2 O + R 3 R 4 PPh 3 R 1 R 2 R 3 R 4 + Ph 3 P=O X Base Aldehyde or Ketone Wittig Reagent Olen Product triphenylphosphonium oxide

PAGE 73

57 capabilities of the carbonyl carbon that is attached directly to the anionic alpha carbon. Thus the electrons of the carbanion are slightly delocalized, thereby stabilizing the ylide. Figure 3 5 Formation of an ylide from the Wittig reage nt. The degree of stabilization of the ylide has significant implications on the conditions of these reactions. For instance, with stabilized ylide reactions, side reactions are less prone to occur and the ylides are less sensitive to moisture or oxygen. H owever, if the Wittig reagent is unstabilized, the ylides will be more reactive and much more sensitive to water and oxygen. The ylide, once formed, will initiate nucleophilic attack on the carbonyl carbon of the aldehyde or ketone. Through multiple inter mediates, including a four membered ring called an oxaphosphatane, the products are an olefin and triphenylphosphonium oxide. The main driving force for the reaction energetically is the very stable bonds that form between oxygen and phosphorous in the tri phenylphosphonium oxide. This is also the reason the reactions, especially the unstabilized Wittig, must stay free of water or oxygen. By employing an unstabilized methyl Wittig reaction with the indole pyridine core ( 19 ), it was possible to synthesize tw o separate analogues: the vinyl ( 3 5 ) and ethyl ( 3 7 ) derivatives. Methyltriphenylphosphonium bromide was employed as the Wittig reagent and the strong base potassium bis(trimethylsilyl)amide (KHMDS) was used to create the R 3 R 4 PPh 3 X Wittig Reagent : Base R 3 R 4 PPh 3 Ylide

PAGE 74

58 ylide. KHMDS is a bulky, non nucleo philic base, which is important in these reactions so that the base itself does not evoke nucleophillic attack on the aldehyde or ketone. By using this non nucleophilic base, side reactions are thereby minimized. Precedence for this reaction was found by K omanduri et al. 33 who employed this reaction with a substituted quinolone carboxaldehyde. In this manner, the Boc protected vinyl derivative ( 3 6 ) was achieved (63% yield). Scheme 3 3 Methyl Wittig r eaction with indole pyridine carboxaldehyde ( 19 ) Depro tection of indole nitrogen Deprotection of the indole nitrogen on the Boc protected vinyl derivative ( 3 6 ) was accomplished as previously performed with TFA in DCM. This reaction completed the vinyl derivative ( 3 5 ) of the indole pyridine core (83%). Schem e 3 4 D eprotection of the Boc protected indole pyridine vinyl derivative ( 36 ) N N O H O O (19) N N O O (36) Ph 3 PCH 3 Br 63% KHMDS, THF (35) TFA, DCM NH N 82% N N O O (36)

PAGE 75

59 Binding a ffinity t esting of 2 (6 vinylpyridin 2 yl) 1 H indole ( 3 5 ) with Mcl 1 Affinity testing of the vinyl derivative ( 3 5 ) using SPR analysis revealed a binding affinit y for Mcl 1. A binding constant ( K d ) of between 10 and 50 M for Mcl 1 was found, indicating that this derivative could be pursued for future fragment based design. The ethyl derivative: 2 (6 ethylpyridin 2 yl) 1 H indole (3 7 ) Molecular docking of the ethyl derivative ( 3 7 ) Molecular docking si mulations of the ethyl indole pyridine derivative ( 3 7 ) gave analogous scoring results to the methanol derivative. The C score obtained from the best configuration was 3.4 compared to 3.5 for the methanol derivative. In addition, the clustering of molecules in that configuration amounted to 7, as opposed to eight for the methanol derivative. The average relative energy of the docking was found to be 42.7 kcal/mol, the highest of all of the derivatives tested. However, the hydrogen bonds formed in the simula tion were quite different, and therefore the location of the compound when it was bound to Mcl 1 was vastly different. Figure 3 6 Molecular s tructure of e thyl indole pyridine derivative H istidine 224 was found to interact with this molecule via a hydrog en bond (Figure 3 7). Orientation of the analogue is situated right in between the P2 and the P1 pocket. This could be advantageous to extend chains in either direction of the molecule in order to fill both pockets with alkyl or alkenyl chains. For instanc e, if the P2 pocket was the destination, building a chain off of the 8 position of the ring could yield a nice (37) NH N

PAGE 76

60 possibility to enhance affinity. Likewise, if the 4 position of the indole ring was utilized, the P1 pocket could be filled. This data, and the p romising C score and clustering values, led our group to pursue the synthesis of this ethyl derivative ( 3 7 ). Figure 3 7 Molecular docking of ethyl derivative ( 3 7 ) to Mcl 1 (gray), a) positioning of 3 7 binding with Mcl 1, b) 3 7 forms a hydrogen bond wi th His 224 (green) of Mcl 1 in this docking simulation. a) b)

PAGE 77

61 Reduction of unprotected vinyl indole pyridine ( 3 5 ) Reduction of the olefin off of the 6 th position of the vinyl indole pyridine molecule ( 3 5 ) was accomplished by a simple palladium on carbon reducti on in methanol (Scheme 3 5). An overnight reaction was required for completion of the reduction. However, the olefin did undergo reduction to produce the final ethyl derivative of the indole pyridine core ( 3 7 ) with a n 87% yield. Scheme 3 5 Palladium on c arbon reduction of vinyl indole pyridine compound ( 3 5 ). Binding a ffinity testing of 2 (6 ethylpyridin 2 yl) 1 H indole ( 3 7 ) with Mcl 1 Affinity testing of the ethyl derivative ( 3 7 ) using SPR analysis revealed no substantial affinity (at or under 5 0 M ) f or Mcl 1. (35) NH N Pd on C (10% w/w) 87% MeOH (37) NH N

PAGE 78

62 Experimental s ection tert butyl 2 (6 (hydroxymethyl)pyridin 2 yl) 1 H indole 1 carboxylate ( 3 4 ) To a sample of pyridine N Boc indole carboxaldehyde 19 (220 mg, 0.68 0 mmol) was added MeOH (2.0 mL). All 19 did not go into solution, so THF (1 m L) was added and the resulting solution was cooled to 0 ¡ C. NaBH 4 (32 mg, 0.85 mmol) was then added and the reaction was stirred at 0 ¡ C for 45 min. The reaction was monitored by TLC (100% CH 2 C l 2 ) Upon completion, t he reaction was quenched with water. The organic layer was extracted with EtOAc (3X), washed with brine (2X), dried over MgSO 4 filtered and concentrated in vacuo to give a yellow powder. Purification of the crude product via flash column chromatography (2% MeOH in CH 2 C l 2 ) afforded the Boc prote cted methanol derivative 34 (205 mg, 93%) as a light yellow powder. 1 H NMR (300 MHz; CDCl 3 ): # 8.19 (d, J = 8.3 Hz, 1H), 7.77 (t, J = 7.7 Hz, 1H), 7.60 (dt, J = 7.7, 0.6 Hz, 1H), 7.45 (d, J = 7.7 Hz, 1H), 7.38 (td, J = 7.7, 0.9 Hz, 1H), 7.29 7.24 (m, 1H), 7.22 (d, J = 7.8 Hz, 1H), 6.80 (s, 1H), 4.82 (s, 2H), 3.89 ( br s, 1H), 1.36 (s, 9H) N N O H O O (19) N N O O (34) OH NaBH 4 MeOH 93%

PAGE 79

63 (6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 3 3 ) To a sample of the N Boc protected methanol derivative 3 4 (200 mg, 0.62 0 mmol) was added CH 2 Cl 2 (1.0 mL) and TFA (1.0 mL). The reaction was monitored by TLC and LCMS which indicated the reaction was complet e after 2 h. The r eaction was quenched with sat. NaHCO 3 until the pH was neutral. A small amount of CH 2 Cl 2 was then added. The organic layer was extracted with CH 2 Cl 2 (3X), washed with sat. NaHCO 3 ( 1X ) and brine (2X ), dried over MgSO 4 filtered and concent rated in vacuo Purification of the crude product, a yellow powder, via flash column chromatography (2% MeOH in CH 2 Cl 2 ) afforded the final methanol derivative 3 3 (107 mg, 77%) as a light yellow powder. 1 H NMR (300 MHz; CDCl 3 ): # 9.74 (s, 1H), 7.73 7.65 (m 3H), 7.43 (d, J = 8.2 Hz, 1H), 7.26 7.21 (m, 1H), 7.13 (ddd, J = 8.5, 6.5, 1.0 Hz, 2H), 7.05 (d, J = 1.6 Hz, 1H), 4.84 (s, 2H), 3.76 ( br s, 1H) ; 13 C NMR (100 MHz, CDCl 3 ) # 158.6, 149.6, 137.7, 136.8, 136.3, 129.2, 123.6, 121.4, 120.5, 118.9, 118.8, 111.6 101.5, 64.5. FT IR (KBr, cm 1 ) 3279, 3051, 2930, 2856, 1591, 1570, 1544, 1466, 1454, 1432, 1385, 1339, 1306, 1264, 1229, 1188, 1142, 1090, 1050, 1004, 994. Anal. Calcd for C 14 H 12 N 2 O: C, 74 98 ; H, 5. 3 9; N, 12. 49 Found: C, 74 56 ; H, 5. 56 ; N, 12. 25 (33) TFA, DCM NH N OH 77% N N O O (34) OH

PAGE 80

64 ter t butyl 2 (6 vinylpyridin 2 yl) 1 H indole 1 carboxylate ( 3 6 ) Methyltriphenylphosphonium bromide (335 mg, 0.94 0 mmol) was added to a dry round bottom flask charged with a stir bar under nitrogen. Dry THF (1 mL) was added and the solution was brought to 0 ¡ C KHMDS (1.9 mL, 0.95 mmol) was added and the ylide was allowed to form at 0 ¡ C for 15 min. Separately, a solution of N Boc indole carboxaldehyde 19 (252 mg, 0.78 0 mmol) in dry THF (0.5 mL) was stirred at rt until homogenous. The solution of N Boc indole c arboxaldehyde 19 in THF was added dropwise to the ylide mixture The empty round bottom flask was rinsed with dry THF (2X 0.5 mL) and the resulting mixture was added to the ylide solution. The reaction was monitored by TLC (100% CH 2 Cl 2 ), which indicated th e reaction was complete after 2.5 h The reaction was quenched with sat. NH 4 Cl (aq) The organic layer was extracted with diethyl ether (3X), washed with sat. NH 4 Cl (aq.) (1X), water ( 1X ), and brine (3X) dried over MgSO 4 filtered and concentrated in vacu o to give a yellow oil. Purification of the crude product via flash column chromatography (100% CH 2 Cl 2 ) afforded the Boc protected vinyl derivative 3 6 (157 mg, 63%) as a white powder. 1 H NMR (300 MHz; CDCl 3 ): # 8.22 (dd, J = 8.4, 0.7 Hz, 1H), 7.71 (t, J = 7.8 Hz, 1H), 7.59 (dd, J = 7.8, 0.5 Hz, 1H), 7.41 7.31 (m, 3H), 7.26 (td, J = 7.5, 1.0 Hz, 1H), 6.87 (dd, J = 17.5, 10.8 Hz, 1H), 6.78 (d, J = 0.5 Hz, 1H), 6.27 (dd, J = 17.5, 1.3 Hz, 1H), 5.52 (dd, J = 10 .8, 1.3 Hz, 1H), 1.31 (s, 9H) N N O H O O (19) N N O O (36) Ph 3 PCH 3 Br 63% KHMDS, THF

PAGE 81

65 2 (6 vinylpyridin 2 yl) 1 H indole ( 3 5 ) To a sample of the N Boc protected vinyl derivative 3 6 (162 mg, 0.5 06 mmol) was added CH 2 Cl 2 (1.0 mL) and TFA (1.0 mL). The reaction was monitored by TLC and LCMS, which indicated the reaction was complete after 2 h. The reaction was quenched with sat. NaHCO 3 until the pH was neutral. A small amount of CH 2 Cl 2 was then added. The organic layer was extracted with CH 2 Cl 2 (3X), washed with sat. NaHCO 3 (X1) and brine (3X ), dried over MgSO 4 filtered and concentrated in vacuo Purification of the crude product, a yellow powder, via flash column chromatography (100% CH 2 C l 2 ) afforded the final vinyl derivative 3 5 (91 .0 mg, 82%) as a white powder. 1 H NMR (300 MHz; CDCl 3 ): # 9.60 (s, 1H), 7.67 ( m 3H), 7.44 (d, J = 8.2 Hz, 1H), 7.26 7.20 (m, 2H), 7.12 (td, J = 7.5, 1.0 Hz, 1H), 7.03 (s, 1H), 6.89 (t, J = 12.7 Hz, 1H), 6.35 (dd, J = 17.4, 1.3 Hz, 1H), 5.56 (d, J = 10.8 Hz, 1H), 13 C NMR (100 MHz, CDCl 3 ) # 155.2, 150.1, 137.4, 137.0, 136.7, 129.4, 1 23.4, 121.4, 120.3, 119.9, 11 8 9 118.7, 111.6, 100. 9 FT IR (KBr, cm 1 ) 34 31 305 1 3018, 158 2 156 2 145 4 141 8 139 9 1361, 1335, 1305, 1229, 11 99, 1181 11 59 1113, 1091 10 70, 1040, 1006 989. Anal. Calcd for C 15 H 12 N 2 : C, 81. 79 ; H, 5 49 ; N, 12. 72 Found : C, 8 1 69 ; H, 5 69 ; N, 1 2 59 (35) TFA, DCM NH N 82% N N O O (36)

PAGE 82

66 2 (6 ethylpyridin 2 yl) 1 H indole ( 3 7 ) P alladium (10% w/w) on c arbon (15.4 mg, 30% w/w) was added to a dry round bottom flask equipped with a stir bar under nitrogen. The vinyl derivative 35 (50.4 mg, 0.2 29 mmol) and MeO H (2 mL) were added to the f lask and then it was fitted with a three way stopcock. A hydrogen balloon was attached at one end of the stopcock and a vacuum line was fixed to the other end The flask was evacuated and subsequently filled with hydrogen gas (5 X) while stirring. The reaction was allowed to stir a t rt overnight under hydrogen. The mixture was then f iltered over wet Celite 500 and concentrated in vacuo TLC (1:1 hexanes: CH 2 Cl 2 ) was used to deter mine if the product was present Purification of the crude product via flash column chromatography (1:5 hexanes: CH 2 Cl 2 ) afforded the final ethyl derivative 3 7 (44 .0 mg, 87%) as a white powder. 1 H NMR (300 MHz; CDCl 3 ): # 9.67 (s, 1H), 7.66 7.61 (m, 3H), 7.43 (d, J = 8.2 Hz, 1H), 7.22 (td, J = 7.6, 1.2 Hz, 1 H), 7.11 (ddd, J = 7.9, 7.0, 1.0 Hz, 1H), 7.05 (dd, J = 5.1, 3.5 Hz, 1H), 7.01 (d, J = 1.2 Hz, 1H), 2.89 (q, J = 7.6 Hz, 2H), 1.38 (t, J = 7.6 Hz, 3H) ; 13 C NMR (100 MHz, CDCl 3 ) # 163.2, 149.7, 137.2, 136.6, 129.4, 123.2, 121.3, 120.6, 120.2, 117.3, 111.6, 100.4, 31.5, 30.0, 14.1. FT IR (KBr, cm 1 ) 3420, 3053, 2965, 2927, 2870, 1662, 1587, 1564, 1455, 1419, 1392, 1335, 1305, 1229, 1184, 1160, 1146, 1113, 1 097 10 59 98 9 Anal. Calcd for C 15 H 14 N 2 : C, 8 1 .0 5 ; H, 6. 35 ; N, 12.60 Found: C, 80. 78 ; H, 6. 64 ; N, 11. 76 (35) NH N Pd on C (10% w/w) 87% MeOH (37) NH N

PAGE 83

67 CHAPTER IV (E/Z) 1 PROPENYL AND PROPYL DERIVATIVES Alkyl and alkenyl extensions off of the 6 position of the indole pyridine core were next examined to determine if the hydrophobic pocket could be filled with longer chains. These hydrophobic interacti on s (van der Waals) between the deep hydrophobic P2 pocket of Mcl 1 and the alkenyl or alkyl group of the derivative could enable better binding affinity. In addition, the rigidity of the 1 propenyl structure due to the olefin could provide a better geomet ric fit into the P2 pocket, thus increasing affinity. The trans 1 propenyl derivative: ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole (3 8 ) Molecular docking of the trans 1 propenyl derivative ( 3 8 ) The trans 1 propenyl derivative ( 3 8 ) was docked using t he molecular simulations as described previously. Interestingly, the C score (3.5) was the same as the methanol derivative ( 3 3 ). Both of these values along with the C score of the propyl derivative ( 42 ) to be discussed shortly had the highest C scores se en in these derivative classes. Also the trans 1 propenyl derivative had 20 configurations in the best cluster (out of 100) that was by far the best cluster amount of the tested derivatives This indicates an increased reliability that this molecule orien ts itself in a specific manner with respect to Mcl 1 The relative energy was found at 46.1 kcal/mol, which is on the high end of the derivatives explored here. The hydrogen bond that this analogue forms with Mcl 1 was determined to be with Threonine 266 which is a unique interaction with respect to the other derivatives

PAGE 84

68 Figure 4 1 Molecular structure of trans 1 propenyl indole pyridine derivative ( 3 8 ) As seen in the molecular docking in F igure 4 2, the molecule orients itself as almost a bridge f rom the P1 pocket to the P2 pocket of Mcl 1. This could provide valuable opportunities to fill both pockets with chains originating from the 6 or 7 position on the indole ring for the P1 pocket or from extensions of the trans 1 propenyl carbon chain to fi ll the P2 pocket. Interaction with t hreonine (Thr) 266 could possibly involve a hydrogen bond donor or acceptor, however in this case the oxygen of the alcohol group on Thr 266 appears as if it is accepting a hydrogen bond from the indole nitrogen proton ( Figure 4 2b) The pyridine is in close proximity, but the indole nitrogen proton is certainly closer. T his interaction is intriguing as the orientation of the molecule gives derivitization of the molecule with affinity boosting functional groups a real pos sibility. The propenyl group might be contributing to this binding due to the filling of a small hydrophobic pocket at the end of the chain. (38) NH N

PAGE 85

69 Figure 4 2 Molecular docking of trans 1 propenyl derivative ( 3 8 ) to Mcl 1 (gray), a) positioning of 3 8 binding with Mcl 1, b) 3 8 forms a hydrogen bond with Thr 266 (yellow) of Mcl 1 in this docking simulation. a) b)

PAGE 86

70 Unstabilized Wittig reaction of indole pyridine carboxaldehyde ( 19 ) with an ethyl Wittig reagent The Wittig reaction was again employed in order to form the 3 carbon derivatives off of the 6 position of the pyridine in the indole pyridine core ( 19 ). As described earlier, the unstabilized Wittig reaction was used but in this case an ethyltriphosphonium bromide Wittig reagent was used to extend the chain on e more carbon. The reaction generated a mixture of trans ( 39 ) and cis ( 40 ) stereoisomers of N Boc protected 1 propenyl derivatives. The unstabilized Wittig reaction normally favor s the formation of the cis stereoisomer over the trans stereoisomer but the mixture of stereoisomer products (68% yield) was near 1:1 (E:Z). This may be due to the steric bulk of the tert butoxycarbonyl (Boc) group protecting the indole nitrogen, thus preventing the syn addition of the Wittig reagent from several directions. Final ly, t he t wo stereoisomers were separated, but not completely, using flash column chromatography. Scheme 4 1 Ethyl Wittig reaction with indole pyridine aldehyde ( 19 ) Deprotection of indole nitrogen Deprotection of the indole nitrogen of the Boc protec ted trans 1 propenyl analogue ( 3 9 ) was accomplished as previously described with TFA in DCM This reaction completed the trans 1 propenyl derivative ( 3 8 ) of the indole thiophene core with a 90% yield. N N O H O O (19) N N O O (39) Ph 3 PCH 2 CH 3 Br 68% KHMDS, THF N N O O (40) +

PAGE 87

71 Scheme 4 2 D eprotection of the Boc protected indole p yridine trans 1 propenyl derivative ( 3 9 ). Binding a ffinity testing of ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 3 8 ) with Mcl 1 Affinity testing of the trans 1 propylene derivative ( 3 8 ) using SPR analysis revealed no substantial affinity (at o r under 5 0 M ) for Mcl 1. The cis 1 propenyl derivative: ( Z ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 41 ) Molecular docking of cis 1 propenyl indole pyridine derivative ( 41 ). Molecular docking of the cis 1 propenyl derivative of the indole pyridine core (Figu re 4 3) was not performed. Figure 4 3 Molecular structure of cis 1 propenyl indole pyridine derivative ( 41 ) Deprotection of indole nitrogen Deprotection of the indole nitrogen of the Boc protected cis 1 propenyl analogue ( 40 ) was accomplished as pre viously described with TFA in DCM. This reaction (38) TFA, DCM NH N 90% N N O O (39) (41) NH N

PAGE 88

72 completed the cis 1 propenyl derivative of the indole pyridine core ( 41 ). The product was obtained as a white powder with a yield of 73%. Scheme 4 3 Deprotection of the Boc protected indole pyridine cis 1 propenyl derivative ( 40 ). Binding a ffinity testing of ( Z ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 41 ) with Mcl 1 Affinity testing of the cis 1 propenyl derivative ( 41 ) using SPR analysis revealed no substantial affinity (at or under 10 M ) for M cl 1. Further testing at an increased dosage (50 M ) is pending. The propyl derivative: 2 (6 propylpyridin 2 yl) 1 H indole (4 2 ) Molecular d ocking of the propyl derivative ( 42 ) The propyl analogue ( 4 2 ) of the indole pyridine core showed almost identical m olecular docking scoring values as the methanol derivative ( 3 3 ). A C score of 3.5 and cluster of 8 molecules was determined at the lowest energy configuration. The average relative energy was found to be 84.4 kcal/mol as well. These values are analogous t o the methanol derivative. However, the propyl derivative forms a hydrogen bond with Arginine (Arg) 263 instead of Asp 242 or Asn 239 as the methanol derivative does. (41) TFA, DCM NH N 73% N N O O (40)

PAGE 89

73 Interestingly, t his residue (Arg 263) coincides with what the Vanderbilt research team id entified as the key residue in the binding of their derivative to Mcl 1 Figure 4 4 Molecular Structure of propyl indole pyridine derivative ( 4 2 ). As can be seen in Figure 4 5, the propyl derivative ( 42 ) orients itself above the P2 pocket of Mcl 1. Th e deep hydrophobic regions of this pocket again allow for possible alkyl or non polar extensions into the region to increase affinity. The Arg 263 residue as a hydrogen bond donor is forming a hydrogen bond with the pyridine nitrogen in this docking How ever, another N proton off of Arg 263 is also in close proximity to the indole nitrogen of the compound indicating another possible contributing interaction. The propyl chain seems to be curling away slightly into the solvent exposed area, although the al pha carbon of the chain may be contributing some van der Waals interactions with a hydrophobic residue nearby (42) NH N

PAGE 90

74 Figure 4 5 Molecular docking of propyl derivative ( 4 2 ) to Mcl 1 (gray), a) positioning of 4 2 binding with Mcl 1, b) 4 2 forms a hydrogen bond with Arg 263 (cyan) of Mcl 1 in this docking simulation. a) b)

PAGE 91

75 Deprotection of i ndole n itrogen Deprotection of the indole nitrogen was accomplished as previously described with TFA in DCM (Scheme 4 4). This reaction was performed on a mixture of unseparated pr oducts from the previous syntheses ( 3 9 and 40 ). The final product was a mixture of trans ( 38 ) and cis ( 41 ) stereoisomers of unprotected 1 propenyl derivatives with a yield of 97%. Scheme 4 4 Deprotection of a mixture of trans ( 39 ) & cis ( 40 ) stereoisomer s of the Boc protected 1 propenyl indole pyridine derivative. Reduction of unprotected ( E / Z) 1 propenyl indole pyridine ( 3 8 & 41 ) Reduction of the 1 propenyl olefin of the indole pyridine core molecule s ( 38 & 41 ) was accomplished by using a p alladium on c arbon reduction in methanol (Scheme 3 5) as described previously. The product was the final propyl derivative ( 4 2 ) of the indole pyridine core with a yield of 86%. Scheme 4 5 Palladium on c arbon reduction of olefin of the unprotected ( E / Z) 1 propenyl indole pyridine analogue s ( 3 8 & 41 ) N N O O (39) N N O O (40) + TFA, DCM 97% NH N (38) NH N (41) + Pd on C (10% w/w) 86% MeOH (42) NH N NH N (38) NH N (41) +

PAGE 92

76 Binding a ffinity t esting of 2 (6 propylpyridin 2 yl) 1 H indole ( 4 2 ) with Mcl 1 Affinity testing of the propyl derivative ( 4 2 ) using SPR analysis revealed no substantial affinity (at or under 5 0 M ) for Mcl 1.

PAGE 93

77 Exp erimental s ection ( E/Z ) tert butyl 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole 1 carboxylate ( 3 9 40 ) Ethyltriphenylphosphonium bromide (692 mg, 1.86 mmol) was added to a dry round bottom flask equipped with a stir bar under nitrogen. Dry THF (2.1 0 mL) was then added and the solution was cooled to 0 ¡ C. KHMDS (3.7 0 mL, 1.85 mmol) was added and the ylide was allowed to form at 0 ¡ C for 15 min. Separately, a solution of N Boc indole carboxaldehyde 19 (500 mg, 1.55 mmol) in dry THF (2.0 mL) was stirred at rt until homogenous. The solution of N Boc indole carboxaldehyde 19 in THF was added dropwise to the ylide solution The reaction was monitored by TLC (100% CH 2 Cl 2 ), which indicated the reaction was complete after 5 h. The reaction was quenched with sat. N H 4 Cl (aq.) The organic layer was extracted with diethyl ether (3X), washed with sat. NH 4 Cl (aq.) (1X), water ( 1X ), and brine (3X) dried over MgSO 4 filtered and concentrated in vacuo to give a yellow oil. Purification of the crude product via flash colum n chromatography (1:5 hexanes: CH 2 Cl 2 ) afforded the mixture of Boc protected (E/Z) 1 propenyl derivatives 39 & 40 (157 mg, 63% total) as a white powder, as well as individual separated stereoisomer fractions: trans 39 ( approximately 147 mgs, 0.440 mmol, 57% ) and cis 40 ( approximately 183 mgs, 0.548 mmol, 71%) trans : 1 H NMR (300 MHz; CDCl 3 ): # 8.21 (dd, J = 8.3, 0.7 Hz, 1H), 7.68 (dt, J = 13.8, 7.2 Hz, 1H), 7.58 (dd, J = 7.7, 0.5 Hz, 1H), 7.38 7.31 (m, 2H), 7.27 7.18 (m, 1 H), 6.88 6.79 (m, 1H), 6.76 (s, 1H), 6.55 (d, J = 15.8 Hz, 1H), 2.13 (dd, J = 7.3, 1.8 Hz, 1H ), 1.93 (dd, J = 6.7, 1.6 Hz, 3H), 1.31 (s, 9H). cis : 1 H NMR (300 MHz; CDCl 3 ): # 8.21 (d, J = 8.1 Hz, 1H), 7.70 (t, J = 7.8 N N O H O O (19) N N O O (39) Ph 3 PCH 2 CH 3 Br 68% KHMDS, THF N N O O (40) +

PAGE 94

78 Hz, 1H), 7.59 (dd, J = 7.7, 0.5 Hz, 1H), 7.39 7.33 (m, 2H), 7.28 7.21 (m, 1 H), 6.77 (s, 1H), 6.52 (dd, J = 11.7, 1.3 Hz, 1H), 6.04 (dq, J = 11.8, 7.3 Hz, 1H) 2.13 (dd, J = 7.3, 1.8 Hz, 3H), 1.93 (dd, J = 6.7, 1.6 Hz, 1H ), 1.32 (s, 9H). ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 3 8 ) To a purified sample of the N Boc protected trans 1 propenyl derivative 3 9 (103 mg, 0.3 08 mmol) was added CH 2 Cl 2 (1.5 m L) and TFA (1.0 mL) and the mixture was allowed to stir at rt. The reaction was monitored by TLC which indicated the reaction was complete after 2 h. The reaction was quenched with sat. NaHCO 3 until the pH was neutral. A small amount of CH 2 Cl 2 was then ad ded. The organic layer was extracted with CH 2 Cl 2 (3X), washed with sat. NaHCO 3 (X1) and brine (3X ), dried over MgSO 4 filtered and concentrated in vacuo Purification of the crude product, a yellow powder, by flash column chromatography (1:1 hexanes: CH 2 Cl 2 ) afforded the final trans 1 propenyl derivative 3 8 (65.3 mg, 91 %) as a white powder. 1 H NMR (300 MHz; CDCl 3 ): # 9.57 ( br s, 1H), 7.66 7.63 (m, 3H), 7.44 (d, J = 8.0 Hz, 1H), 7.26 7.20 (m, 1H), 7.11 ( m 2H), 7.02 (s, 1H), 6.97 6.85 (m, 1H), 6.58 (d, J = 14.9 Hz, 1H), 1.99 (dd, J = 6.7, 1.6 Hz, 3H) ; 13 C NMR (100 MHz, CDCl 3 ) # 155.6, 149.8, 137.25, 137.18, 136.6, 13 1.34, 131.25, 129.4, 123.2, 121.4, 120.3, 119.5, 118.0, 111.6, 100.7, 18.6. FT IR (KBr, cm 1 ) 3429, 3048, 2912,1662, 1584, 1561, 1454, 1417, 1391, 1335, 1305, 1229, 1195, 1191, 1158, 1113, 1100, 1007, 988. Anal. Calcd for C 16 H 14 N 2 : C, 82.02; H, 6.02; N, 11. 96 Found: C, 81.50; H, 6.06; N, 11.81. (38) TFA, DCM NH N 90% N N O O (39)

PAGE 95

79 ( Z ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 41 ) To a purified sample of the N Boc protected cis 1 propenyl derivative 40 (139 mg, 0.4 16 mmol) was added CH 2 Cl 2 (1.5 mL) and TFA (1.0 mL) and the solution was a llowed to stir at rt. The reaction was monitored by TLC and LCMS, which indicated the reaction was complete after 3 h. The reaction was quenched with sat. NaHCO 3 until the pH was neutral. A small amount of CH 2 C l 2 was then added. The organic layer was extra cted with CH 2 Cl 2 (3X), washed with sat. NaHCO 3 (X1) and brine (3X ), dried over MgSO 4 filtered and concentrated in vacuo Purification of the crude product, a yellow solid, via flash column chromatography (1:1 hexanes: CH 2 Cl 2 ) afforded the final cis 1 prop enyl derivative 41 (70.6 mg, 73 %) as a white powder. 1 H NMR (300 MHz, CDCl 3 ) # 9.70 9.30 ( br s, 1H), 7.74 7.58 (m, 3H), 7.48 7.40 (d, J = 7.5 Hz, 1H), 7.30 7.20 (m, 1H), 7.18 7.08 (m, 2H), 7.06 7.00 (s, 1H), 6.62 6.48 (d, J = 10.5 Hz, 1H), 6.17 6.02 (m, 1H), 2.28 2.17 (dd, J = 7.5, 2.4, 3H); 13 C NMR ( 100 MHz, CDCl 3 ) # 15 6 3 149.8, 13 6 9 136.6, 1 31 7 12 9 6 12 9 4 12 3 3 12 2 6 1 21 4 1 20.3 119 .5, 117.5 111 6, 100.7, 15.8 FT IR (KBr, cm 1 ) 3 428 30 20 29 90 1662, 1583, 1561, 1454, 1417, 1392, 1335, 1305, 1 229, 1195, 1181, 1158, 1113, 1100, 1007, 988 Anal. Calcd for C 16 H 1 4 N 2 : C, 8 2 0 2; H, 6. 0 2; N, 11. 96 Found: C, 8 0 04 ; H, 6. 14 ; N, 11. 34 (41) TFA, DCM NH N 73% N N O O (40)

PAGE 96

80 2 (6 propylpyridin 2 yl) 1 H indole ( 4 2 ) A sample of unseparated and unprotected (E/Z) 1 propenyl analogues 3 8 & 4 1 (60 mg, 0.26 mmol) was placed in a dry round bottom flask with palladium (10% w/w) on carbon (18 mgs, 30% w/w) under nitrogen. MeOH (2 mL) was added t o the flask and it was then fitted with a three way stopcock. A hydrogen balloon was attached to one end o f the stopcock and a vacuum line was attached to the other end. The flask was evacuated and subsequently filled with hydrogen gas while stirring (5X). The reaction was allowed to stir at rt overnight under hydrogen. M ore MeOH was added and then the mixtur e was filtered over wet Celite 500 and concentrated in vacuo TLC (1:1 hexanes: CH 2 Cl 2 ) was performed to determine if the product was present Purification of the crude product, a yellow solid, via flash column chromatography (1:1 hexanes: CH 2 Cl 2 ) afforded t he final propyl derivative 42 (52 mg, 86%) as a white powder. 1 H NMR (300 MHz; CDCl 3 ): # 9.60 ( br s, 1H), 7.66 7.63 (m, 3H), 7.44 (d, J = 8.2 Hz, 1H), 7.22 (td, J = 7.6, 1.0 Hz, 1H), 7.11 (ddd, J = 7.9, 7.0, 1.0 Hz, 1H), 7.03 (t, J = 5.4 Hz, 2H), 2.83 (t, J = 7.4 Hz, 2H), 1.84 (dq, J = 15.0, 7.5 Hz, 2H), 1.02 (t, J = 7.4 Hz, 3H). 13 C NMR (300 MHz, CDCl 3 ) # 152.0, 149.8, 137.3, 137.0, 136.6, 129.5, 123.2, 121.31, 121.28, 120.2, 117.3, 111.6, 40.5, 23.2, 14.2 FT IR (KBr, cm 1 ) 3223, 3055, 2957, 2928, 2869, 1593 1569, 1437, 1429, 1390, 1339, 1305, 1253, 1230, 1188, 1159, 1145, 11 2, 1082, 1005, 993, 796, 750. Anal. Calcd for C 16 H 16 N 2 : C, 81.32; H, 6.82; N, 11.85 Fou nd: C, 81.31; H, 6.60; N, 11.83 Pd on C (10% w/w) 86% MeOH (42) NH N NH N (38) NH N (41) +

PAGE 97

81 CHAPTER V CONCLUSION In order to find a selective Mcl 1 inhibitor a structure based design approach utilizing molecular modeling and subsequent synthesis was employed. Of the eight novel derivatives pursued (Table 5 1) only seven could be tested for Mcl 1 affinity as the pyrrole derivative ( 18 ) was unstable at room t emperature. From those seven analogues containing the pyridine indole core, two ( 30 and 3 5 ) had affinity at concentrations between 10 and 50 M Specific binding constants (K d ) for these two compounds are currently being determined. These two compounds are considered "fragment like", due to their relatively small size and the lack of affinity boosting functional groups. Therefore a more potent inhibitor of Mcl 1 could be constructed using two methods: 1) fragment based design by fusing two fragments or ele ments of those fragments into one compound or 2) by performing a structure activity relationship (SAR) study, to determine which key functional group s boost affinity to Mcl 1 and adding these moieties to an already promising scaffold Our research group wi ll use the second approach and will pursue a SAR study to find a novel and selective Mcl 1 inhibitor. Table 5 1 Obatoclax ( 4 ) and the d erivatives synthesized by the Julian Lab. MD = molecular docking nd = not determined Structure MD C score MD c lusteri ng MD H bond SPR K d ( "M) 2.0 4 His 252 0.01 1.83 6 His 252 nd* NH N O HN NH N HN 18 4

PAGE 98

82 3.29 6 His 224 10 50 3.5 8 Asn 239 Asp 242 No affinity (under 50 M) nd* nd* nd* 10 50 3.4 7 His 224 No affinity (under 50 M) 3.5 20 Thr 266 No affinity (under 50 M) nd* nd* nd* No affinity (under 50 M) 3.5 8 Arg 263 No affinity (under 50 M) Th e Julian research group has had the opportunity to partner with Scripps Florida to create more complex analogues of the pyridine indole core Scripps Florida had previously constru cted the M311 compound ( 11, Table 5 2 ), however the compound was physically unstable. B y merging components of M311 ( 11 ) with the indole pyridine core NH N S 30 NH N OH 33 NH N 35 NH N 37 NH N 38 NH N 41 NH N 42

PAGE 99

83 new trifluoromethyl containing compounds were synthesized. As an example, SR3 3893 was synthesized that has an amide group off of the 6 position of the pyridine ring, a butyl chain off of the 3 position of the indole ring and a trifluoromethyl ( CF 3 ) group off of the 5 position of the indole ring. Molecular docking and SPR data of these compounds are sh own i n Table 5 2. Table 5 2 Derivatives synthesized by Scripps Florida. nd = not determined Name Structure MD C score MD Clustering MD H bond SPR K d ( "M) M311 2.56 9 Arg 263 His 224 nd* Initial M311 Hybrid 3.3 6 Thr 266 His 224 nd* SR3 3893 3.5 20 Hi s 224 Arg 263 6.6 SR3 3891 nd* nd* nd* 1 10 SR3 3890 nd* nd* nd* 1 10 SR3 3896 nd* nd* nd* 1 10 N N OH CF 3 N H F 3 C N N N OH NH N O F 3 C N NH N F 3 C OH NH N F 3 C N NH N N Ot-Bu N O

PAGE 100

84 M olecular modeling on SR3 3893 revealed a C score of 3.5 with a clustering of 20 (Figure 5 1) which were the highest values of all the co mpounds studied. In addition, the modeling revealed a hydrogen bond ing interaction with the amide group of SR3 3893 and the Arg 263 residue of Mcl 1. Interestingly the trifluoromethyl group showed interactions with the His 224 residue of Mcl 1, effectively forming a bridge across the P1 and P2 pocket s of Mcl 1. This creates a unique opportunity to increase affinity by filling the hydrophobic pocket below with a hydrophobic alkyl group, which the butyl group off of the 3 position of the indole ring provides Other possible hydrophobic chains, such as cyclopropyl or cyclobutyl groups, may be explored to see if a particular moiety fills the hydrophobic pocket more effectively. Figure 5 1 Molecular docking of SR3 3893. Displays the trifluoromethyl group inter acting with His 224 (left) the amide group interacting with Arg 263 (right) and the butyl group filling the hydrophobic pocket SPR analysis of SR3 3893 yielded a K d to Mcl 1 of approximately 6.6 M (Figure 5.2) Hence, SR3 3893 showed moderate binding a ffinity for Mcl 1. In addition, three other compounds were measured to have below 10 M affinity (precise K d values are currently being determined) for Mcl 1 from the CU Denver Scripps Florida compounds: NH N O F 3 C N

PAGE 101

85 SR3 3891 SR3 3890 and SR3 3896 (Table 5 2 ) Three of the se four compounds contain a trifluoromethyl moiety, signaling that this functional group may be aiding in the binding to Mcl 1 corroborated by the modeling results This preliminary data provides an initial structure activi ty relationship that will b e studied in depth with future derivatives Figure 5 2 Surface Plasmon Resonance (SPR) data of SR3 3893 (red) with Mcl 1 Obatoclax (blue) was the positive control and ABT 263 (black) was the negative control for the assay. Outlook The Julian researc h group will continue to develop a potent Mcl 1 inhibitor by exploring the SAR between Mcl 1 and the pyridine indole core. Several new derivatives have been targeted for synthesis (Figure 5 3) to probe the importance of the specific functional groups withi n the Scripps synthesized analogues that showed affinity for Mcl 1. Three of these compounds ( 43 44 and 45 ) have the trifluoromethyl moiety that initial SAR indicated contributes to Mcl 1 binding. By removing the butyl group off of the 3 position of the indole ring of SR3 389 1 to yield compound 4 3 the importance of this hydrophobic chain will be explored. Similarly, the butyl group will also be removed from

PAGE 102

86 SR3 3893 and SR3 3890 to yield 44 and 45 respectively. In addition 46 will be synthesized to rev eal the importance of the trifluoromethyl group to Mcl 1 binding. Finally, other derivatives (not pictured) such as a tetrazole or a carboxylic acid containing analogue ( off of the pyridine ring ) could also be employed to strengthen the arginine residue interaction. Future SPR analysis of these four compounds ( 43 46 ) and the others mentioned will define a clear SAR between the se different moieties and Mcl 1, and thus will provide a path to synthesize a novel and specific Mcl 1 inhibitor Figure 5 3 S y nthetic targets for a Mcl 1 inhibitor to establish a structure activity relationship (SAR) Due to its facile synthesis, compound 43 (Figure 5 4 ) will be the first synthesized As mentioned previously, t his compound will help define the role of the trifluo romethyl group in relation to Mcl 1 binding and reveal the importance of the hydrophobic chain off of the 3 position of the indole ring An indole ring substituted at the 5 position with a trifluoromethyl group is commercially available and would provide a starting material for t he synthesis of the trifluoromethyl methanol derivative ( 4 3 ) Figure 5 4 Proposed new synthetic target ( 43 ) based on initial SAR results NH N F 3 C 43 OH N Br O N H B(OH) 2 + H F 3 C NH N F 3 C 43 OH NH N F 3 C N O 44 NH N F 3 C N 45 NH N OH 46

PAGE 103

87 Initial molecular modeling of this compound ( 43 ) had a cluster of 18 molecules in th e best geometric configuration and a C score of 3.15 A hydrogen bond interaction was formed with the alcohol group of 43 and His 224 of Mcl 1 In addition, the trifluoromethyl group interacts with two residues : Asn 260 and Arg 263 (Figure 5 5 ) This is an inter esting binding mode for Mcl 1 as the molecule is essentially flipped around 180 ¡ from the interactions that were forming with SR3 3893 The consistency of the binding interactions of several of the derivatives (specifically SR3 3893 and 43 ) with these thre e residues is encouraging. Thus synthesis and subsequent SPR analysis of 43 will provide valuable insight into the SAR for this particularly robust binding mode for Mcl 1

PAGE 104

88 Figure 5 5 Molecular docking of compound ( 43 ) with Mcl 1 a) P ositioning of ( 43 ) on Mcl 1. b) C lose up view. 43 forms a hydrogen bond interaction with His 224 (green) and the alcohol moiety while the trifluoromethyl group has interactions with both Arg 263 (cyan) and Asn 260 (orange). Future selective small molecule inhibitors of Mcl 1 are clearly on the horizon. The recent success of ABT263 ( 6 ) as a Bcl X L inhibitor has heightened the need for anti apoptotic Bcl 2 family member inhibitor drugs. However, results for Mcl 1 specific inhibitors have lagged behind due to many factors, including the inability to find small a) b)

PAGE 105

89 molecules that can bind to Mcl 1 effectively (low nanomolar range) and yet still have in vivo efficacy. Currently a potent Mcl 1 inhibitor is in high demand, as chemoresistance (such as resistance to ABT 263 ) is often linked to overexpression of Mcl 1. Therefore, the Julian research group aims to continue to pursue the indole pyridine core as a pharmacophore with the ultimate goal of developing a potent and selective small molecule inhibitor of Mcl 1

PAGE 106

90 CHAPTER VI SU PPLEMENTAL INFORMATION Spectroscopy of synthetic target Type of Spectra Page # N (tert butoxycarbonyl) indole ( 21 ) 1 H NMR 92 N (tert butoxycarbonyl) indole 2 boronic acid ( 22 ) 1 H NMR 93 tert butyl 2 (6 formylpyridin 2 yl) 1H indole 1 carboxylate ( 19 ) 1 H NMR 94 2 (6 ((3,5 dimethyl 1H pyrrol 2 yl)methyl)pyridin 2 yl) 1H indole ( 18 ) 1 H NMR 95 tert butyl 2 (6 (hydroxy(thiophen 2 yl)methyl)pyridin 2 yl) 1H indole 1 carboxylate ( 32 ) 1 H NMR 96 2 (6 (thiophen 2 ylmethyl)pyridin 2 yl) 1H indole ( 30 ) 1 H NMR 97 2 (6 (thiophen 2 ylmethyl)pyridin 2 yl) 1H indole ( 30 ) FTIR 98 tert butyl 2 (6 (hydroxymethyl)pyridin 2 yl) 1 H indole 1 carboxylate ( 34 ) 1 H NMR 99 (6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 33 ) 1 H NMR 100 (6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 33 ) 13 C NMR 101 (6 (1 H indol 2 yl)pyridin 2 yl)methanol ( 33 ) FTIR 102 tert butyl 2 (6 vinylpyridin 2 yl) 1 H indole 1 carboxylate ( 36 ) 1 H NMR 103 2 (6 vinylpyridin 2 yl) 1 H indole ( 35 ) 1 H NMR 104 2 (6 vinylpyridin 2 yl) 1 H indole ( 35 ) 13 C NMR 105 2 (6 vinylpyr idin 2 yl) 1 H indole ( 35 ) FTIR 106 2 (6 ethylpyridin 2 yl) 1 H indole ( 37 ) 1 H NMR 107 2 (6 ethylpyridin 2 yl) 1 H indole ( 37 ) 13 C NMR 108 2 (6 ethylpyridin 2 yl) 1 H indole ( 37 ) FTIR 109 ( E ) tert butyl 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole 1 carbox ylate ( 39 ) 1 H NMR 110 ( Z ) tert butyl 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole 1 carboxylate ( 40 ) 1 H NMR 111 ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 38 ) 1 H NMR 112 ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 38 ) 13 C NMR 113 ( E ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 38 ) FTIR 114

PAGE 107

91 ( Z ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 41 ) 1 H NMR 115 ( Z ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 41 ) 13 C NMR 116 ( Z ) 2 (6 (prop 1 en 1 yl)pyridin 2 yl) 1 H indole ( 41 ) FTIR 117 2 (6 propylpyridin 2 yl) 1 H indole ( 42 ) 1 H NMR 118 2 (6 propylpyridin 2 yl) 1 H indole ( 42 ) 13 C NMR 119 2 (6 propylpyridin 2 yl) 1 H indole ( 42 ) FTIR 120

PAGE 108

92

PAGE 109

93

PAGE 110

94

PAGE 111

95 NH N HN (18)

PAGE 112

96 N N OH O O (32) S

PAGE 113

97 (30) NH N S

PAGE 114

98 (30) NH N S

PAGE 115

99 N N O O (34) OH

PAGE 116

100 (33) NH N OH

PAGE 117

101 (33) NH N OH

PAGE 118

10 2 (33) NH N OH

PAGE 119

103 N N O O (36)

PAGE 120

104 (35) NH N

PAGE 121

105 (35) NH N

PAGE 122

106 (35) NH N

PAGE 123

107 (37) NH N

PAGE 124

108 (37) NH N

PAGE 125

109 (37) NH N

PAGE 126

110 N N O O (39)

PAGE 127

111 N N O O (40)

PAGE 128

112 (38) NH N

PAGE 129

113 (38) NH N

PAGE 130

114 (38) NH N

PAGE 131

115 (41) NH N

PAGE 132

116 (41) NH N

PAGE 133

117 (41) NH N

PAGE 134

118 3 3 2 2 1.81 1.81 0.0832 0.0832 1.65 1.65 0.942 0.942 0.954 0.954 0.399 0.399 0.802 0.802 2.42 2.42 0.731 0.731 ppm -2 -2 0 0 2 2 4 4 6 6 8 8 10 1.65 1.65 0.942 0.942 0.954 0.954 0.399 0.399 0.802 0.802 2.42 2.42 7.0 7.0 7.5 7.5 3 3 2 2 1.0 1.0 1.5 1.5 2.0 2.0 2.5 2.5 (42) NH N

PAGE 135

119 (42) NH N

PAGE 136

120 (42) NH N

PAGE 137

121 REFER ENCES 1 Ashkenazi, A. Nature Reviews Cancer 2002 2 420 430. 2 Shi, Y.; Chen, J.; Weng, C.; Chen, R.; Zheng, Y.; Chen, Q.; Tang, H. Biochemical and Biophysical Research Communications 2003 305 989 996. 3 Bu ytaert, E.; Callewaert, G.; Vandenheede, J. R.; Agostinis, P. Autophagy 2006 2 238 240. 4 Belmar, J.; Fesik, S. W. Pharmacology & Therapeutics 2015 145 76 84. 5 Boldin, M. P.; Varfolomeev, E. E.; Pancer, Z.; Mett, I. L.; Camonis, J. H.; Wallach, D. Journal of Biological Chemistry 1995 270 7795 7798. 6 Kischkel, F.; Hellbardt, S.; Behrmann, I.; Germer, M.; Pawlita, M.; Krammer, P. H.; Peter, M. E. EMBO J. 1995 14 (22) 5579 5598. 7 Alimonti, J. B.; Ball, T. B.; Fowke, K. R. Journal of General Virology 2003 84 (7) 1649 1661. 8 Yang, L.; Mashima, T.; Sato, S. Cancer Res. 2003 63 (4) 831 837. 9 Lessene, G.; Czabotar, P. E.; Colman, P. M. Nature Reviews Drug Discovery 2008 7 989 1000. 10 Lessene, G.; Czabotar, P. E.; Colman, P. M. Nature R eviews Drug Discovery 2008 7 989 1000. 11 Perciavalle, R. M.; Opferman, J. T. Trends in Cell Biology 2013 23 22 29. 12 Thomas, L. W.; Lam, C.; Clark, R. E.; White, M. R. H.; Spiller, D. G.; Moots, R. J.; Edwards, S. W. PLoS ONE 2012 7 13 Yang Yen, H F. Mcl 1: a highly regulated cell death and survival controller. Journal of Biomedical Science 2006 13 201 204. 14 Rautureau, G. J. P.; Day, C. L.; Hinds, M. G. International Journal of Molecular Sciences 2010 11 1808 1824. 15 PDB file 3PK1. Czabota r, P. E.; Lee, E. F.; Thompson, G. V.; Wardak, A. Z.; Fairlie, W. D.; Colman, P. M. Journal of Biological Chemistry 2011 286 7123 7131. 16 Sathishkumar, N.; Sathiyamoorthy, S.; Ramya, M.; Yang, D. U.; Lee, H. N.; Yang, D. C. Journal of Enzyme Inhibition and Medicinal Chemistry 2012 27 (5), 685 692. 17 Belmar, J.; Fesik, S. W. Pharmacology & Therapeutics 2015 145 76 84.

PAGE 138

122 18 Thorn, C. F.; Oshiro, C.; Marsh, S.; Hernandez Boussard, T.; McLeod, H.; Klein, T. E.; Altman, R. B. Pharmacogenetics and Genomics 2 011 21 440 446. 19 Song, T.; Chen, Q.; Li, X.; Chai, G.; Zhang, Z. Journal of Medicinal Chemistry 2013 56 9366 9367. 20 Albershardt, T. C.; Salerni, B. L.; Soderquist, R. S.; Bates, D. J. P.; Pletnev, A. A.; Kisselev, A. F.; Eastman, A. Journal of Bio logical Chemistry 2011 286 24882 24895. 21 Stewart, M. L.; Fire, E.; Keating, A. E.; Walensky, L. D. Nature Chemical Biology 2010 6 595 601. 22 Muppidi, A.; Doi, K.; Edwardraja, S.; Drake, E. J.; Gulick, A. M.; Wang, H. G.; Lin, Q. Journal of the Ameri can Chemical Society 2012 134 14734 14737. 23 Obatoclax and Bortezomib in Treating Patients With Relapsed or Refractory Multiple Myeloma Full Text View https://clinicaltrials.gov/ct2/show/NCT00719901 (accessed Mar 13, 2015 ). 25 Chemist 19, O. File:Suzu ki Coupling Full Mechanism 2.png http://commons.wikimedia.org/wiki/File:Suzuki_Coupling_Full_Mechanism_2.png#media viewer/File:Suzuki_Coupling_Full_Mechanism_2.png (accessed Mar 13, 2015 ). 26 Da•ri, K.; Yao, Y.; Faley, M.; Tripathy, S.; Rioux, E.; Billot, X .; Rabouin, D.; Gonzalez, G.; LavallŽe, J. F.; Attardo, G. Organic Process Research & Development 2007 11 1051 1054. 27 Li, G.; Lamberti, M.; Mazzeo, M.; Pappalardo, D.; Roviello, G.; Pellecchia, C. Organometallics 2012 31 1180 1188. 28 Kelly, A. R.; K errigan, M. H.; Walsh, P. J. Journal of the American Chemical Society 2008 130 4097 4104. 29 Carey, F. A. Advanced Organic Chemistry PT. B: Reactions and Synthesis ; New York$: Plenum Press, c1983 c1984.: United States, 1983 30 Baguley, T. D.; Xu, H. C.; Chatterjee, M.; Nairn, A. C.; Lombroso, P. J.; Ellman, J. A. Journal of Medicinal Chemistry 2013 56 7636 7650.. 31 Sabban, S.; Ye, H.; Helm, B. Veterinary Immunology and Immunopathology 2013 153 10 16. 32 Antonioletti, R.; Bonadies, F.; Ciammaichella, A.; Viglianti, A. Lithium hydroxide as base in the Wittig reaction. A simple method for olefin synthesis. Tetrahedron 2008 64 4644 4648.

PAGE 139

123 33 Komanduri, V.; Grant, C. D.; Krische, M. J. Journal of the American Chemical Society 2008 130 12592 12593.