SPIROPYRROLIZIDINE AND PIPERAZINE DERIVATIVES: SYNTHESIS AND EVALUATION ON KRAS EXPRESSION LEVELS IN HUMAN COLON CANCER CELLS by Victor P. Liu B.Sc., University of Northern British Columbia, 2016 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN BIOCHEMISTRY UNIVERSITY OF NORTHERN BRITISH COLUMBIA April 2019 © Victor P. Liu, 2019 Abstract KRAS-driven cancers are notoriously difficult to treat due to poor pharmacodynamics of downstream inhibitors and resistance to anti-EGFR drugs. IMP-1 is a post-transcriptional regulator of KRAS mRNA. As a novel therapeutic approach, the targeting of the IMP-1-KRAS mRNA complex with a spiropyrrolizidine derivative (UNBC152), was studied. LC-MS analysis of UNBC152 indicated presence of impurities. The purpose of this study was to synthesize UNBC152 and determine the responsible bioactive molecule within the impurities. LC-MS and TLC suggested the presence of a bioactive [3+3] cycloaddition side product (SPOPP) in UNBC152. SPOPP suppressed KRAS expression in human colorectal cancer cells. Fluorescence polarization determined that SPOPP did not impact the IMP-1-KRAS mRNA interaction. SPOPP induced G2/M cell cycle arrest as shown by flow cytometry. MTT assay confirmed the SPOPP-induced growth inhibition in SW480 (IC50 = 4.17 μM) and HT29 (IC50 = 6.76 μM). These findings represent a first reporting on the bioactivity of SPOPP. I Table of Contents Abstract I Table of Contents II List of Tables VIII List of Figures IX List of Abbreviations XV Acknowledgements XVIII References XIX Chapter 1 - Introduction 1 1.1. Oncogenes 1 1.2. RNA Regulation 2 1.3. IMP-1: Structure and Function 4 1.4. IMP-1: Role in Development and Oncogenesis 7 1.5. Inhibitors of IMP-1 9 1.6. Oncogenic KRAS 11 1.7. Strategies for Targeting KRAS-Driven Cancers 14 II 1.7.1. Inhibition by siRNA 14 1.7.2. Post-Translational inhibition by FTI and GGTI 17 1.7.3. Inhibiting Sub-cellular Trafficking by Targeting PDE6δ 18 1.7.4. KRAS Activation Lowering by SOS1 Inhibition 19 1.7.5. Mutant-specific Inhibition of KRAS G12C 20 1.7.6. Targeting Effectors in Related Pathways 22 1.8. Interaction Between KRAS and IMP-1 23 1.9. Principle of Fluorescence Polarization 25 2.0. Research Goals 27 Chapter 2 – Synthesis, Purification and Characterization of Spiropyrrolizidine and Piperazine Derivatives 29 2.1. Methodology – Synthesis, Purification and Characterization of Spiropyrrolizidine and Piperazine Derivatives 30 2.1.1. General Information 30 2.1.2. Synthesis of (2E)-2-(2-bromobenzylidine)-1-indanone (3) 31 2.1.3. Synthesis of UNBC152 (7) 32 2.1.4. Spiro[2′,3]-bis(acenaphthene-1′-one)perhydrodipyrrolo- [1,2-a:1,2-d]pyrazine (SPOPP; 8) and NP6A 32 III 2.2. Results and Discussion 34 2.2.1. Synthesis of (2E)-2-(2-bromobenzylidine)-1-indanone (3) 34 2.2.2. Synthesis of UNBC152 (7) 36 2.2.3. Synthesis of Spiro[2′,3]-bis(acenaphthene-1′-one)perhydrodipyrrolo[1,2a:1,2-d]-pyrazine (8) 2.2.4. HPLC-MS Analysis of Synthesized Products. 40 44 2.2.4.1. Comparison of UNBC152 (7) and VLA9 (7) 44 2.2.4.2. Temperature Dependency of the Synthesis of SPOPP (8) 49 2.2.5. Isolation of SPOPP (8) and NP6A 53 2.2.6. Purification of Other Peaks in the SPOPP (8) Mixture 57 2.2.6.1. Initial Purification and Analysis 57 2.2.6.2. F1 Sample Production as a Means for Comparison 62 2.2.6.3. HPLC Optimization of Fraction 1 Mixtures 65 2.2.7. Structural Characterization 2.2.7.1. (2E)-2-(2-bromobenzylidine)-1-indanone (3) 69 69 2.2.7.1.1. Analysis 69 2.2.7.1.2. NMR Spectrum 69 IV 2.2.7.2. SPOPP (8) 70 2.2.7.2.1. Analysis 70 2.2.7.2.2. NMR Spectrum 73 2.2.7.3. NP6A 78 2.2.7.3.1. Analysis 78 2.2.7.3.2. NMR Spectrum 78 2.2.8. Summary 83 Chapter 3 – Biochemical Characterization of Spiropyrrolizidine and Piperazine Derivatives 86 3.1 Methodology – Fluorescence Polarization Assay to Assess Effect on IMP-1-KRAS mRNA Interaction 3.1.1. Generation of Recombinant IMP-1 86 86 3.1.1.1 Protein Generation 86 3.1.1.2 Protein Purification 86 3.1.1.3 Protein Refolding and Quantification 87 3.1.2. Fluorescence Polarization Assay 87 3.2 Methodology – Analysis of Spiropyrrolizidine and Piperazine Derivatives on Human Colorectal Cancer Cells 88 V 3.2.1. General Information for Cell Preparation 88 3.2.2. Total Protein Harvesting and Western blot Detection of KRAS 89 3.2.3. Total RNA Harvesting and Quantitative PCR of KRAS Steady-state Levels 90 3.2.4. MTT Assay Quantification of Cell Viability 92 3.2.5. Flow cytometry 93 3.2.5.1. Apoptosis and Necrosis Detection 93 3.2.5.2. Analysis of Cell Cycle Population 95 3.3. Results and Discussion 95 3.3.1. Fluorescence Enhancers and Quenchers Found in Fluorescence Polarization Assay 3.3.2. Effect of Inhibitors on KRAS Expression in CRC Cells 95 101 3.3.2.1. Activity in SW480 101 3.3.2.2. Activity on HT29 101 3.3.3. Cell Viability Analysis of Compound-treated CRC Cells 107 3.3.3.1. Proliferation in SW480 107 3.3.3.2. Proliferation in HT29 109 3.3.4. Effect of SPOPP (8) on KRAS mRNA Levels in HT29 Cells 112 VI 3.3.5. Investigating the Anti-proliferative Mechanism of SPOPP (8) in SW480 113 3.3.5.1. Necrosis versus Apoptosis 113 3.3.5.2. Cell Cycle Distribution 115 3.3.6. Summary 118 Chapter 4 – General Discussion 119 4.1 Project Overview 119 4.2 Synthesis of Spiropyrrolidine Derivatives and the Focus on SPOPP (8) 120 4.3 HPLC-MS Analyses of SPOPP (8) and UNBC152-3 (7) 121 4.4 Purification and Identification of the Fused Piperazines, SPOPP (8) and NP6A 124 4.5 Finding Inhibitors of the IMP-1-KRAS Interaction by Fluorescence Polarization Assay 125 4.6 Reduction of KRAS Expression in CRC 127 4.7 SPOPP (8) Induces Cellular Death through Necrosis 128 4.8 G2/M Inhibition by SPOPP (8) 130 4.9 Concluding Remarks 130 VII List of Tables Table 2.1.1. Crude batches of 8 synthesized at varying times and temperatures. 33 Table 2.1.2. HPLC purification procedure for SPOPP (8) and NP6A. 34 Table 2.2.4.1. Time and temperature dependent yield of 8 and NP6A according to area on VWD chromatogram. Table. 2.2.5.1. Purification of VLB133 by FCC normal phase. 52 54 Table 2.2.6.1. Comparison of notable ions found by LRESI-MS analysis of F1 samples of 8 as compared to the collaborator’s UNBC152-3 (7). 64 Table 2.2.7.1. Chemical shifts of protons and carbons for SPOPP (8). 72 Table 3.2.1. Summary of antibodies used in Western blot analyses. 90 Table 3.2.2. Primer sequences for qPCR analysis of spiro-treated CRC cells. 91 Table 3.3.1. Summary of cell cycle population percentages for SPOPP-treated SW480. 118 VIII List of Figures Figure 1.3.1. Schematic of wild-type coding region determinant binding protein (WT IMP-1) displaying domains and associated nucleotide positions 4 Figure 1.3.2. Secondary structure differences of KH domain types. 5 Figure 1.3.3. Crystal structure of truncated IMP-1. 6 Figure 1.5.1. Structure of an IMP-1 inhibitor 6896009. 10 Figure 1.6.1. Downstream effectors of KRAS signaling. 12 Figure 1.7.1. Chemical modification of the furanose rings in nucleosides for second generation antisense oligonucleotides. 16 Figure 1.7.2. Inhibitors of KRAS prenylation. 17 Figure 1.7.3. Inhibitors of the PDE6δ-KRAS interaction. 18 Figure 1.7.4. DCAI: An inhibitor of the SOS1-KRAS interaction. 19 Figure 1.7.5. Allosteric inhibitors of KRAS G12C. 21 Figure 1.8.1. UNBC152 inhibitor of the IMP-1-KRAS mRNA interaction. 24 Figure 1.8.2. Hypothesis for the UNBC 152 inhibition of IMP-1-KRAS interaction. 25 Figure 1.9.1. Principle of fluorescence polarization assay. 26 Figure 2.1.1. Reaction scheme for the synthesis of (2E)-2-(2-bromobenzylidine)-1Indanone (3). 31 IX Figure 2.1.2. Synthesis of spiropyrrolizidine derivative, UNBC152 (7). 32 Figure 2.1.3. Synthesis of the dispiropiperazine derivatives SPOPP (8), and 9. 32 Figure 2.2.1.1. Mechanism for the base-catalyzed aldol condensation between 1Indanone (1) and 2-bromobenzaldehyde (2). 35 Figure 2.2.1.2. TLC reaction monitoring for the synthesis of (2E)-2 -(Bromobenzylidene)-1-indanone (3). 36 Figure 2.2.2.1. Formation of the azomethine ylide (6) from the reaction between Acenaphthenequinone (4) and ʟ-proline (5). 37 Figure 2.2.2.2. Mechanism for the 1,3-dipolar cycloaddition formation of UNBC152 (7). 38 Figure 2.2.2.3. Reaction monitoring for the synthesis of UNBC152 (7). 39 Figure 2.2.2.4. Comparison of various batches of UNBC152 (7). 39 Figure 2.2.3.1. Proposed Mechanism of SPOPP (8) by S. Haddad et al. 2015. 41 Figure 2.2.3.2. Reaction monitoring for the synthesis of Spiro[2′,3]-bis(acenaphthene-1′ -one)perhydrodipyrrolo- [1,2-a:1,2-d]-pyrazine (8). 42 Figure 2.2.3.3. Comparison between UNBC 152 (7) and the two-component synthesis VLA25 (8/9). 43 Figure 2.2.4.1. HPLC spectral analysis and comparison of UNBC152-3 (7) and VLA9 (7). Figure 2.2.4.2. Low resolution ESI-MS comparison of UNBC152-3 (7) and VLA9 (7). 45 46 Figure 2.2.4.3. Ion analysis of the dominant peaks in the UNBC152-3 (7) sample by low resolution ESI-MS. 47 X Figure 2.2.4.4. Ion analysis of the dominant peaks in the VLA9 (7) sample by low resolution ESI-MS. Figure 2.2.4.5. HPLC-MS analysis of crude VLA37 (8/9). Figure 2.2.4.6. Ion analysis of the dominant peaks in the VLA37 (8/9) sample by low resolution ESI-MS. 48 50 51 Figure 2.2.4.7. HPLC spectral analysis on the effect of temperature variance for the synthesis of 8. 52 Figure 2.2.5.1. Purification schematic for the isolation of 8 and NP6A. 53 Figure 2.2.5.2. HPLC spectral analysis of NP6A samples. 55 Figure 2.2.5.3. Ion analysis of the purified NP6A by low resolution ESI-MS. 55 Figure 2.2.5.4. HPLC-LRESI-MS analysis of purified SPOPP (8). 56 Figure 2.2.5.5. Ion analysis of the purified SPOPP (8) by low resolution ESI-MS 57 Figure 2.2.6.1. Purification schematic to produce fraction 1 (F1) components of crude 58 8/9. Figure 2.2.6.2. Reverse phased (C18) fractionation and TLC analysis of VLA37 (8/9) 59 Figure 2.2.6.3. HPLC-LRESI-MS analysis of the VLA37-1. 60 Figure 2.2.6.4. Ion analysis of the dominant peak in VLA37-1 fraction by LRESI-MS. 60 Figure. 2.2.6.5. HPLC-MS analysis of NP4. 61 XI Figure 2.2.6.6. Ion analysis of NP4 from LRESI-MS. 62 Figure 2.2.6.7. HPLC trace comparison of various fraction 1 samples from C18 FCC purification of crude batches of 8. 63 Figure 2.2.6.8. HPLC optimization of the purification of VLB188-1. 66 Figure 2.2.6.9. HPLC trace of fractionated VLB188-1. 68 Figure 2.2.7.1. 1H-NMR spectra of (2E)-2-(2-bromobenzylidene)-1-indanone (3). 70 Figure 2.2.7.2. Numbered positions for the structure of SPOPP (8). 71 Figure 2.2.7.3. 1H-NMR spectra of SPOPP (8). 74 Figure 2.2.7.4. 13C-NMR spectra of SPOPP (8). 75 Figure 2.2.7.5. COSY spectra of SPOPP (8). 76 Figure 2.2.7.6. HSQC of SPOPP (8). 77 Figure 2.2.7.7. 1H-NMR spectra of NP6A. 79 Figure 2.2.7.8. 1H-NMR spectra of 9. 80 Figure 2.2.7.9. 13C-NMR spectra of NP6A. 81 Figure 2.2.7.10. 13C-NMR spectra of 9. 82 Figure 2.2.7.11. COSY of NP6A. 83 Figure 3.2.1. Scheme for the reduction of MTT to the formazan form. 93 XII Figure 3.2.2. Structure of 7-Aminoactinomycin D. 94 Figure 3.2.3. Structure of propidium iodide. 95 Figure 3.3.1. Comparison of crude spiropyrrolizidine and dispiropiperazine compounds on the IMP-1-KRAS mRNA interaction. 97 Figure 3.3.2. Comparison of purified SPOPP (8) and NP6A versus crude batches on the IMP-1-KRAS mRNA interaction. 98 Figure 3.3.3. Effect of semi-purified F1 components on the IMP-1-KRAS mRNA interaction. 100 Figure 3.3.4. Effect of purified SPOPP (8) and NP6A on KRAS expression in SW480 CRC. 102 Figure 3.3.5. Effect of semi-purified F1 samples on KRAS expression in SW480 CRC. 103 Figure 3.3.6. Effect of SPOPP (8) and reagents on KRAS expression in HT29 CRC. 104 Figure 3.3.7. Effect of SPOPP (8) and NP6A on KRAS expression in HT29 CRC. 105 Figure 3.3.8. Effect of SPOPP (8) and NP6A on KRAS expression in HT29 CRC. 106 Figure 3.3.9. Aggregated western blot data for KRAS protein expression in compound-treated HT29 106 Figure 3.3.10. MTT assay analysis on SPOPP-treated SW480. 108 Figure 3.3.11. MTT assay analysis on F1 -treated SW480. 109 Figure 3.3.12. MTT assay analysis of SPOPP (8) and semi-purified F1 on HT29. 110 XIII Figure 3.3.13. MTT assay analysis of SPOPP (8) and NP6A on HT29. 111 Figure 3.3.14. Analysis of KRAS mRNA steady-state levels in SPOPP-treated HT29 by quantitative PCR. 112 Figure 3.3.15. Distinguishing mechanism of cellular death on SPOPP-treated SW480 cells using flow cytometry. Figure 3.3.16. Cell cycle analysis of SPOPP-treated SW480 cells using flow cytometry. 114 117 XIV List of Abbreviations 4HBD – four-helical bundle domain 7-AAD – 7-aminoactinomycin-D A2BP1 – ataxin 2-binding protein 1 ACN – acetonitrile AcQ – acenaphthenequinone Ago2 – Argonaut 2 AON – anti-sense oligonucleotides C – cysteine cDNA – complementary DNA Colorectal cancer – CRC COSY – homonuclear correlation spectroscopy D – aspartate DC – dendritic cells EGFR – epidermal growth factor receptor EtOAc – Ethyl acetate FBDD – fragment-based drug design/discovery FCC – flash column chromatography FP – fluorescence polarization FTI – farnesyl transferase inhibitor G – glycine GAP – GTPase- activating protein GAPDH – glyceraldehyde 3-phosphate dehydrogenase GEF – guanine exchange factor GGTI – geranylgeranyl transferase inhibitor XV HMGB1 – High mobility group 1 protein hnRNP – heterogeneous nuclear ribonucleoprotein HPLC – high performance liquid chromatography HRP – horseradish peroxidase HVR – hyper variable region IMP-1 – Insulin-like growth factor 2 mRNA-binding protein 1 IPTG – isopropyl β-D-1-thiogalactopyranoside KH – K Homology KRAS – Kirsten Rat Sarcoma Viral Oncogene KRAS-FL – fluorescein-labeled KRAS LNA – locked nucleic acid LRESI-MS – low resolution electrospray ionization mass spectrophotometry MAPK – mitogen activated protein kinase MDR-1 – Multi-Drug Resistance 1 miRNA – micro RNA MLKL – mixed lineage kinase domain-like protein MTT – 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide Ni-NTA – nickel nitrilotriacetic acid NMR – nuclear magnetic resonance NP – normal phase PB2 – positively charged residues PDE6δ – δ-subunit of cGMP phosphodiesterase type 6 PE – Phycoerythrin PI – propidium iodide PI3K – phophoinositide 3-kinase XVI PS – phosphatidylserine Q – Glutamine qPCR – quantitative polymerase chain reaction RBP – RNA binding protein Rf – retention factor RIP1 – protein serine/threonine kinase receptor interacting protein 1 RISC – RNA induced silencing complex RNAi – RNA interference molecules ROS – reactive oxygen species RRM – RNA recognition motifs RSV – Rous Sarcoma Virus S-IIP – switch 2 pocket (of KRAS) siRNA – small interfering RNA SOS1 – Son of Sevenless SPOPP – Spiro[2′,3]-bis(acenaphthene-1′-one)perhydrodipyrrolo- [1,2-a:1,2-d]-pyrazine TLC – thin layer chromatography UTR – untranslated region V – valine VEGFR – vascular-endothelial growth factor receptor VICKZ – Vg1, RBP/Vera, IMP-1,2,3, CRD-BP, KOC and ZBP-1 VLA – Victor Liu Book A VLB – Victor Liu Book B v-src – Rous Sarcoma Virus viral gene VWD – variable wavelength detector βTrCP1 – Beta-Transducin Repeat Containing E3 Ubiquitin Protein Ligase XVII Acknowledgements First and foremost, I would like to thank my supervisor, Dr. Chow Lee for giving me the opportunity to work in an excellent cancer research lab. Your support and guidance throughout the past few years have helped me grow substantially as a researcher. I would never forget the time spent in this lab, as it also helped me grow personally. To Dr. Tina Bott, my co-supervisor, I would like to thank you for your patience. You have always been a great mentor for organic chemistry, from undergrad till now. Your help has been tremendous, especially during the initial syntheses. To Dr. Maggie Li, thank you for joining me on the front lines in the assay screens to help solve the UNBC152 puzzle. You always had wise words during my quandaries. I would like to acknowledge Dr. Kerry Reimer for passing on excellent theoretical knowledge for HPLC purifications. I would like to acknowledge Dr. Guy Plourde for always inspiring me to think about the mechanistic nature of reactions. I would like to thank Dr. Hossein Kazemian for being a gracious host of the Northern Analytical Lab. Thank you to Charles Bradshaw for keeping the HPLC running. To Dr. Liz Dunn and Dr. Kaila Fadock, thank you for the technical knowledge on NMR. To Sebastian and Corbin, thank you for all the great times in and out of the lab; I am fortunate to have met the both of you. To my other close friends, Cale, Laura, Levon, Jonah, Aaron and Joshua, thank you for always being around, even when I seemingly drop of the face of the Earth to do research. Last but not least, I would like to thank my parents. To my mother Kit Pang, thank you for always believing in me, especially when I was meandering through different educational paths. To my late father, David Liu, thank you for supporting me throughout the years. I miss you more than you would ever know. XVIII Chapter 1 Introduction 1.1. Oncogenes The discovery of oncogenes began in the early 1900s with a researcher named Peyton Rous.1,2 During his studies on the domestic fowl, he found that a tumor extract, when injected into a healthy chicken, will initiate tumorigenesis.1 The extract was found to contain a virus and was later named the Rous Sarcoma Virus (RSV). Discovery of the RSV viral gene (v-src) was accomplished with a combination of oligonucleotide finger printing and sucrose gradient centrifugation.3 The discovery led to the idea of possible non-viral genes which can promote tumorigenesis. Mike Bishop and Harold Varmus created a v-src complementary DNA (cDNA) probe through reverse transcription to find the endogenous avian src gene.4 Hybridization experiments using v-src cDNA and avian DNA led to the discovery of the sought after src gene; this led to the term proto-oncogene. From the discovery, research for additional proto-oncogenes in cancer proceeded.5 Conversion of proto-oncogenes to oncogenes can occur from one of four main events: point mutation, upregulation, chromosome translocation or gene amplification.6 Mutations include single nucleotide replacement, nucleotide deletion(s), frameshift mutation, and nucleotide insertion. Following ribosomal translation of the mutated gene, the newly synthesized protein has altered structural conformation. Structural changes range from complete mis-folding of the protein to slight steric hindrance of an active site. The end result is either altered activity or altered post transcriptional regulation. In both cases, cellular homeostasis is imbalanced, leading to deleterious 1 effects. Upregulation results in the increased production of protein; this often occurs from malfunctioning transcriptional regulators or an increase in messenger RNA stability. Chromosomal translocation often leads to the creation of fusion genes. Fusion gene products alter the cellular interactions of their respective counterparts.6 Gene amplification results in multiple copies of the oncogene. Identification of oncogenes are important since the products are often involved in regulation of the cell cycle or apoptosis.6 Several well-known oncogenes include the following: Myc, epidermal growth factor receptor (EGFR), phosphoinositide 3-kinase (PI3K), and Ras. The method in which a proto-oncogene is activated will determine its cellular effect. Myc is a transcriptional regulator and is located on position 24 on the long arm of the chromosome 8 (8q24) for humans. Translocation of Myc to gene positions 14q, 22q and 2p causes the gene to be positioned near an enhancer leading to Burkitt’s lymphoma. In EGFR mutations, the ligand binding domain is often deleted which leads to a perpetual on state (constitutive signaling). EGFR and related proteins are found in breast and lung cancer. PI3K is a tyrosine kinase and activates Akt in a signal transduction pathway involved with cell growth and survival.6 Ras is a transmembrane protein which, upon activation, signals the canonical mitogen activated protein kinase (MAPK) and PI3K pathways, leading to growth and differentiation of cells.7 More information on Ras will be discussed in section 1.6. Given the diversity of oncogenes and the interplay with various signaling pathways, it is difficult to determine a singular target for anticancer therapy. 1.2. RNA Regulation Precise spatiotemporal regulation of RNA is the main factor which allows a cell to rapidly and efficiently adjust to cellular demands.8 Steady state levels of RNA fluctuate depending on cell 2 type, cell cycle and potential exposure to negative stimuli, such as carcinogens and toxins. Other factors influencing steady state mRNA levels are pre-mRNA splicing, transcription rates, RNA degradation and interaction with microRNAs and RNA binding proteins.8 Expression of RNA can be modulated by RNA interference (RNAi) molecules.9 A subset of RNAi, microRNA (miRNA), are produced from double stranded RNA after processing by the RNase III enzymes Dicer and Drosha. The miRNA is a short fragment of approximately 22 nucleotides in length and associates with mRNA through Watson-Crick base pairing.10 If the miRNA is perfectly matched, the mRNA is degraded by the RNaseH enzyme named Argonaute which is located in the RNA-induced silencing complex (RISC).9 Alternatively, mismatched siRNA causes translational inhibition of the mRNA. RNA binding proteins (RBP) are one of the major post-transcriptional regulators of RNA11. RBPs aid in capping, translation, localization, turnover, splicing and various other processes relating to steady state levels of RNA. RBPs often contain RNA recognition motifs (RRM) and heterogeneous nuclear ribonucleoprotein (hnRNP) K Homology (KH) domains. The domains serve to bind RNA based on a sequence specific motif. Currently there are 424 known RBPs with varying functionalities but interestingly, some of the prerequisite motifs for binding are not unique to each protein variant; this indicates that a singular RBP regulates more than one RNA species. Any perturbation in RBP levels can impact RNA regulation.11 A specific example would be the transcriptomic analysis performed by Voineagu et al. on autistic brains.12 Downregulation of the RBP, ataxin 2-binding protein 1 (A2BP1), led to 212 significant alternative splicing events. Numerous implicated exons from the alternative splicing were found to be involved in synaptic function of neuronal cells; thus there is a significant correlation between A2BP1 expression levels and regulation of synaptic events.12 In prostate cancer, the RBP named Sam68 is upregulated and 3 is correlated with neoplastic effects.13 Downregulation of Sam68 through RNA interference reduced the proliferation of the LNCaP prostate cancer cell line.13 RBP expression can vastly affect the post-transcriptional regulation of RNA targets. Any dysregulation of RBP, such as a small expression fold change, may greatly impact RNA steady-state levels leading to a diseased state. 1.3. IMP-1: Structure and Function The Insulin-like Growth Factor 2 mRNA-binding Protein 1 (IMP-1) was initially discovered through a UV cross-linking experiment.14 Through this process, the 75 kDa protein was chemically bound to the coding region determinant of the c-Myc transcript. At the time of discovery, IMP-1 was suggested to protect the transcript from endoribonuclease attack.14 IMP-1is a member of a larger family of highly conserved RNA binding proteins called VICKZ (Vg1, RBP/Vera, IMP1,2,3, CRD-BP, KOC and ZBP-1).15 According to NCBI blast, the murine IMP-1 protein has a 99% amino acid sequence identity to the human variant. As seen in Figure 1.3.1, IMP-1 contains two N-terminal RRMs and four KH domains which arrange in the following di-domains: RRM1RRM2, KH1-KH2 and KH3-KH4.16 The discrete di-domains are separated from each other through a long variable region.17 Within the di-domain pair, a shorter and more conserved spacer exists.17 Figure 1.3.1. Schematic of wild-type coding region determinant binding protein (WT IMP1) displaying domains and associated nucleotide positions.16 4 RNA binding in IMP-1 occurs through the four KH domains.18 KH domains exist in two forms. Eukaryotes have the type I fold while prokaryotes have the type II fold. As seen in Figure 1.3.2, type I and type II folds differ by the placement of β’ and α’ leading to differences in secondary structure. The KH domains in IMP-1 contain the type I fold. Binding of RNA occurs through the region between the left side (α1, GXXG, α2) and right side (β2 and a variable loop).18 Contrary to traditional RRM domains, IMP-1 does not require the RRM1 or RRM2 to bind RNA.17 Figure 1.3.2. Secondary structure differences of KH domain types. (A) Eukaryotic KH domain, (B) Prokaryotic KH domain.18 Further elucidation of the binding requirements was done in a mutational experiment by Barnes et al. in 2015.16 Since the GXXG motif was implicated in binding RNA, the first glycine (G) of the motif was mutated by site-directed mutagenesis into an aspartate (D) in order to abrogate binding. The single mutants, KH1, KH2, KH3 and KH4 were created. For example, the KH1 mutant has the GXXG loop mutated to DXXG for the KH1 domain while the KH2, KH3 and KH4 domains remain in its wildtype form. Dual domain mutants were also created such as KH1-2 and KH3-4. It was found that the single point mutants KH1, KH2, KH3 and KH4 had varying ability 5 to bind to both CD44 and c-Myc RNA. Interestingly the KH4 mutant had a higher affinity towards c-Myc RNA than CD44. Mutations in any two domains, however, abrogated binding completely. The KH3-4 di-mutant was the only exception to the finding; the dual mutation still resulted in binding. Given the variation in binding ability of the mutants towards the target RNA, it suggests that different domains are required for sequence specificity. Additionally, the data suggests that at least two domains are required for strong binding.16 Structural elucidation of IMP-1 has been historically challenging due to the flexible variable regions between the di-domain pairs. The only crystal structure solution of a VICKZ protein was a truncation of IMP-1 as solved by Chao, et al. in 2010 to a resolution of 2.75Å.19 Figure 1.3.3. Crystal structure of truncated IMP-1. (A) KH3 and KH4 dimer. (B) Hydrophobic residues involved for dimerization.19 Hydrophobic residues located on β1 and α3 of both KH3 and KH4 form electrostatic interactions leading to sequestering and the formation of the pseudo-dimer (Figure. 1.3.3B). Of note, the 6 GXXG motif between α1 and α2 of KH3 and KH4 lie in opposition to each other (Figure. 1.3.3A). Due to the orientation of both GXXG motifs, Chao, et al. hypothesized that the RNA substrate wraps around KH3 and KH4 in a 180° fashion. The minimum nucleotide requirement for specificity was also explored during their in vitro analysis of the KH3 and KH4 domains in ZBP1. Using the electrophoretic mobility shift assay, it was shown that the KH3 and KH4 domains displayed similar high affinity to the first 28 nucleotides of the zipcode as compared to the fulllength protein. A deletion of 2 to 4 nucleotides in the middle of the zipcode did not change the protein’s affinity. However, a deletion of 6 nucleotides led to a magnitude lower in affinity and a deletion of 8 nucleotides abrogated binding completely.19 The finding suggests the importance of nucleotide length for high affinity bipartite recognition by the KH3 and KH4 domains of VICKZ proteins. A minimum of 17 nucleotides are required to be energetically favorable for wrapping around the two domains. In reality, the in vivo affinity of VICKZ proteins to their substrates are vastly different due to substrates’ length being in the thousands of nucleotides. Binding affinity would then be enhanced by the presence of all 4 KH domains since the KH1 and KH2 domains acts as extra binding surfaces for electrostatic interactions with RNA. The RNA will then hypothetically loop around both di-domain pairs (KH3 and KH4; KH2 and KH1). The crystallization of the full protein is warranted for the elucidation of all binding surfaces and RNA orientation. An attempt to crystallize IMP-1, which consisted of KH1 to KH4 domains, was done by Mackedenski but unfortunately, the crystal twinned and could not be diffracted.20 1.4. IMP-1: Role in Development and Oncogenesis IMP-1 is the first oncofetal RBP that was reported.21 Leeds et. al. monitored IMP-1 expression in rat liver tissue during the fetal and neonatal stage and found expression. However, the protein was not found to be expressed during the adult stage. To further clarify the role in 7 development, IMP-1 expression was also monitored in liver tissue following partial liver resection. Interestingly, IMP-1 expression was absent in the liver tissue. Analysis of the hepatoma cell lines, HepG2, HuH7 and Hep3B resulted in redetection of IMP-1; therefore IMP-1 is only found during early development and in cancer.21 To further solidify IMP-1’s role in development, IMP-1 deficient mice were developed to determine potential phenotypic changes.22 The knockout mice exhibited dwarfism and the vast majority of the organs such as the brain, spleen, small intestine and liver exhibited hypoplasia. In terms of life-span, only 50% of mice lived 3 days postnatal. The rationale for organ hypoplasia was due to the alteration in mRNA transcript levels. In intestines, 24 transcripts changed more than 2-fold; changes included alteration related to extracellular matrix transcripts, gene encoding carbonic anhydrase III and various others. The liver also experienced two fold changes in 20 different transcripts.22 The global RNA level changes were related to the function of IMP-1. Depending on when IMP-1 is expressed, it can either be beneficial or detrimental. Since IMP-1 is a RBP, it can bind to a wide array of mRNA transcripts during development and oncogenesis. One of the main functions of IMP-1 is the protection from endoribonucleases.14 The role of IMP-1 shielding transcripts from endoribonucleolytic degradation was confirmed in 2006 by Sparanese and Lee.23 IMP-1 protected the c-Myc and MDR-1 transcripts from degradation by the endoribonuclease, APE1 in a concentration dependent manner.23,24 Protection from micro RNA mediated degradation was also reported.25 IMP-1 prevented miR-183 from targeting βTrCP-1 for degradation by argonaute-2 (Ago2), which is the catalytic subunit located in the RNA-induced silencing complex.25 Although, IMP-1 is necessary for embryogenesis, its ability to bind important oncogenes is detrimental during oncogenesis. IMP-1 is able to bind and protect βTrCP1, c-Myc, CD44, MDR-1 and KRAS.23,25–27 βTrCP1 is involved in the signaling pathway for ubiquitination 8 and proteasomal degradation of key substrates related to the inhibition of nuclear factor κB and βcatenin.25 MDR-1 codes for the multi-drug resistance protein that allows for efflux of chemotherapeutic agents.28 c-Myc codes for the important transcription factor c-Myc which when expressed aids in the transcription of genes for growth and development pathways.29 CD44 codes for a cell-surface antigen which are involved in processes such as cellular adhesion, lymphocyte activation, angiogenesis and various other roles relating to the extracellular matrix.30 Important oncogene turnover rate is lowered through IMP-1 binding and thus will exacerbate the cancer condition. Prevalence of IMP-1 in varying cancers has been widely reported. The protein is found in 58.5% of breast cancers, 81% of colorectal cancers and is found in basal cell carcinoma.31–33 The expression of IMP-1 also has a negative correlation with survival rates as found after examining patients with colorectal cancer; this indicates the protein is a good predictor of survival rate.34 Nonsmall cell lung carcinomas (25%) and high grade brain cancers such as glioblasomas multiform (61%) also has expression of IMP-1.35 1.5. Inhibitors of IMP-1 Initial work on IMP-1 inhibitors focused on RNAi inhibition of IMP-1. In a study on osteosarcoma, miR-150 was found to be downregulated.36 Consequently, overexpression of miR150 in osteosarcoma cell lines was found to downregulate IMP-1 protein levels through transcript base-pairing at the 3321-3328 bp region of the 3’ UTR. Downregulation of IMP-1 led to decreased cell invasion and migration.36 MicroRNA miR-506 was found to target IMP-1 in glioblastoma cells and similarly suppressed cell proliferation and migration.37 Specific oligonucleotides were developed to inhibit the IMP-1-CD44 mRNA interaction and was found to reduced steady state levels of CD44 in vivo.26 Antibiotics such as neomycin and streptomycin were also tested in the 9 study. Although the antibiotics inhibited the binding interaction, the suggested rationale was due to non-specific nucleotide interactions.26 Currently there is only one reported small molecule inhibitor of IMP-1 that was deemed effective. Mahapatra, et al. (2013) analyzed 17600 small molecules using a high-throughput fluorescence anisotropy assay for inhibitors of the IMP-1-c-Myc mRNA interaction.38 Compounds were tested at a concentration of 5 μM and only molecules with more than 25% binding inhibition were sent for cellular analysis. IMP-1 was set at 10 nM while the c-Myc probe was set at 1 nM; this represented approximately 90% of maximal binding. Compounds which increased or decreased total fluorescence by 30% or greater were not included. The molecule labeled 6896009 inhibited the interaction by 72% ± 3.6% and inhibited cell proliferation of IGROV-1 (ovarian cancer cell line) by approximately 50% as according to MTT assay. The structure of the molecule contains nitrogen heterocycles (Figure. 1.5.1).38 Br N N H N S N N O Figure 1.5.1. Structure of an IMP-1 inhibitor 6896009.38 The vast majority of the other compounds either proved ineffective at inhibiting the interaction or were cytotoxic.38 The cytotoxicity was based on a counter-screen which employed IMP-1 negative PC-3 prostate cancer cells. The report of the novel inhibitor is monumental. Specific targeting of RBPs is a difficult task due to the lack of knowledge on the relationship 10 between sequence specificity for RNAs and integral structural motifs required for binding. Although, this molecule was able to inhibit the IMP-1-c-Myc mRNA interaction both in vitro and in vivo, little is known about the specific interaction between the protein and compound.38 Currently there are no reports on the second generation variants of the lead molecule. Structural activity relationships would be necessary to progress forward in the development of IMP-1 inhibitors. Knowledge on the molecule binding site with IMP-1 is also warranted. 1.6. Oncogenic KRAS The Kirsten rat sarcoma viral oncogene (KRAS) is part of the RAS family of GTPase proteins that includes N-Ras and H-Ras.39 RAS proteins are 188 amino acids in length and shares similarity in the first 120 amino acids.40 KRAS, however, exists as two different splice variants named KRAS4A and KRAS4B.41 The KRAS4A splice variant is due to the additional exon 4A that is alternatively spliced in as compared to KRAS4B. Exon 4 is a highly variable region (HVR) for the RAS family and allows for differential post translational modification. In the case of KRAS4A, the HVR contains a site for palmitoylation at cysteine 180 whereas KRAS4B does not undergo this modification. A cluster of positively charged residues (PB2) on KRAS4A HVR from lysine 182 to lysine 185 allows for association with the plasma membrane through electrostatic interactions, even in the absence of palmitoylation. KRAS4B similarly relies on positively charged residues to interact with the inner plasma membrane. Tsai et al. removed the palmitoylation site and mutated the PB2 amino acids to neutral glutamine residues to determine localization changes of KRAS4A.41 Absence of both palmitoylation and PB2 for KRAS4A instead changed cellular localization exclusively to the endomembrane and cytosol. To add further complexity, KRAS4B interacts with the δ-subunit of cGMP phosphodiesterase type 6 (PDE6δ) for cytosolic transport to the plasma membrane whereas KRAS4A does not.41 11 A distinguishing factor of KRAS is the involvement in embryogenesis.42 KRAS knockout mice exhibited weakened ventricular walls and increased apoptosis of neural cells in the spinal cord.42 N-Ras knockout mice, however, did not exhibit deleterious effects during embryo development and were indistinguishable from wildtype mice post-natally.43 KRAS is suggested to be crucial for embryonic development.42,43 The importance of KRAS is due to its position relative to its downstream effectors. Once localized on the plasma membrane, KRAS affects several major pathways (Figure. 1.6.1).44 In the inactive form, KRAS is normally bound to guanosine diphosphate (GDP). Association with a guanine exchange factor (GEF) such as Son of Sevenless 1 (SOS1) accelerates the exchange of GDP to GTP, thus activating KRAS for signal transduction. The two major pathways activated by KRAS are the canonical mitogen activated protein kinase (MAPK) and phosphoinositide 3-kinase (PI3K); these pathways are intimately involved in cell cycle regulation. Figure 1.6.1. Downstream effectors of KRAS signaling.44 12 The MAPK pathway is responsible for cell cycle progression and transcription while PI3K is responsible for processes such as cell growth, survival and migration (Figure. 1.6.1). PI3K activation also leads to inhibition of the tumor suppressor p53. Subsequent inactivation of KRAS occurs with GTP exchange that is enabled by GTPase-activating proteins (GAP) such as Neurofibromin 1 (NF1).44 A single point mutation in KRAS results in deregulation of downstream effectors and is one of the main causes of KRAS driven cancers.7 Mutations commonly occur at glycine 12 (G12) for 85% of cases and can be mutated to aspartate, valine or cysteine. G12 mutations to arginine, serine or alanine exist but are much rarer. Once mutated, steric hindrance at the active site occurs and prevents hydrolysis of GTP from RAS leading to a constitutive signaling state. Prevention of hydrolysis is due to the catalytic glutamine (Q61) shifting orientation in G12 substitutions.45 Q61 normally coordinates with a water molecule to initiate hydrolysis of the γ-phosphate in GTP. Mutations at G13 (14% of cancers) and Q61 (2% of cancers) has also been reported.46 Since KRAS controls MAPK and PI3K pathways, the constitutive signaling leads to uncontrolled cell growth and progression resulting in a cancerous state.7 KRAS driven cancers have been widely reported to this date. Malfunction in KRAS is found in 27.3% of lung adenocarcinoma, 98% of pancreatic ductal adenocarcinoma and 42.4% of colorectal cancer.47–49 The type of mutation also dictates phenotypic outcomes. In colorectal cancer, patients who harbor the valine (V) mutation at the glycine 12 position (G12V) had an increased rate of mortality by 30%.50 G12 mutations in murine NIH/3T3 fibroblasts resulted in lowered apoptosis, increased cell density and spontaneous anchorage independent growth; those 13 characteristics are hallmarks of cancer.51 The G13 mutation, however, did not have as pronounced phenotypic effects as the G12 mutation and had reduced cellular proliferation which in this case, is better for patient survivability.51 Proliferation of colorectal cancer cells was due to a KRAS G12D mutation and was mediated through activation of MEK in the MAPK pathway.52 Not all KRAS driven cancers are due to mutation. Genetic analysis of primary endometrial carcinoma lesions revealed that 3% of total samples (414) contained KRAS gene amplification.53 Patients who had amplification of KRAS was reported to have a lower 5 year survival rate of 46% compared to the 87% for cases which are unamplified. Compared to the primary cancer, metastatic lesion samples contained a higher percentage (18%) of gene amplification. The findings led to a new prognostic marker for metastatic endometrial carcinoma and inspired research into other targeted therapies.53 Due to the prevalence in cancers and the mutant-dependent phenotypic variance, KRAS has been a widely studied protein for potential targeted therapies.54 1.7. Strategies for Targeting KRAS-Driven Cancers 1.7.1. Inhibition by siRNA Like IMP-1, effective targeted therapies towards KRAS in the pre-clinical setting began with analysis of RNA interference. Two types of RNA were analyzed: small interfering RNA (siRNA) and micro RNA (miRNA). The difference between the two lies in origin; siRNA is synthetically produced, and miRNA is found endogenously. An example of an effective miRNA is miR-143; overexpression of endogenous miR-143 in colorectal cancer cells reduced cell proliferation through targeting the 3’ UTR of the KRAS mRNA.55 In oral squamous cell carcinoma, miR-181a also targeted a region of the KRAS 3’UTR and reduced both cell proliferation and anchorage-dependent growth.56 Due to the success of the miRNA inhibition, 14 some researchers took the approach of siRNA therapies in the form of synthetic antisense oligonucleotides (AON). In 2017, Ross et al. designed a third generation AON called AZD4785 (3ʹ-TTTCTGAGGATTATCG-5ʹ) that targets the 3’ UTR of KRAS and initiates RNAse Hmediated degradation of the transcript.57 AZD4785 was designed with a phosphorothioate backbone and modified furanose rings containing constrained ethyl groups at the ends of the AON; 3 nucleotides on each end were modified.58 Specifically, the 2ʹ OH of the furanose was modified to a 2ʹ,4ʹ-constrained 2ʹO-ethyl bridge causing a C3ʹ-endo pucker of the ring (Figure. 1.7.1). Conformationally locked nucleic acids (LNA) allows for improved hybridization with RNA as found previously with AON thermal melt (Tm) analysis. The Tm increased by 1.5°C with each LNA incorporation which indicated strengthened intermolecular association between the AON and the RNA. LNA design allowed for shorter sequences for AONs, from the traditional 22 nucleotides (nts) to 14 nts and protection from 3ʹ-exonucleases. However, LNA still suffered from poor pharmacodynamics in mouse models due to hepatotoxicity. Modification of the R group from H to (S)-cEt solved the hepatotoxicity issue while retaining similar EC50 to LNA (Figure. 1.7.1).58 Due to the specificity, AZD4785 was able to down regulate KRAS without affecting N-Ras or HRas in the A431 endometrial carcinoma cell line.57 Tumor size reduction was also seen upon treatment in the lung cancer patient derived xenograft (PDX) model, LXFA 983.57 No adverse events were found with AON treatment on murine and cynomolgus monkey models.57 Although the results were promising, the pharmacokinetics and pharmacodynamics of AZD4785 have yet to be tested in humans. 15 Figure 1.7.1. Chemical modification of the furanose rings in nucleosides for second generation antisense oligonucleotides.58 As reviewed by Juliano et al. siRNA therapies encounter unfavorable barriers in terms of pharmacokinetics.59 The main barrier would be delivery to specific tissues. Although the phosphorothioate backbone enhances plasma stability, it causes non-specific interactions with proteins and in some cases, leads to heightened toxicity. The siRNAs have been found to preferentially bioaccumulate in the liver and kidneys. Increased dosing of siRNAs to compensate for accumulation merely led to an increase in renal excretion. A specialized macrophage in the liver, named Kupffer cells, also uptake foreign siRNA through phagocytosis and shuttles the phagosome to the lysosome for enzymatic degradation. Of the siRNA that reach the target tissue, it is still subject to the rate-limiting steps of crossing the plasma membrane and subsequent release from endosomes into either the cytoplasm or the nucleus.59 Due to the challenges of siRNA treatments, KRAS researchers also focused on other druggable aspects in regards to posttranslational modification, sub-cellular trafficking, protein activation and related effector pathways. 16 1.7.2. Post-translational Inhibition by FTI and GGTI Inhibition of KRAS post-translational modification was one of the approaches attempted. Farnesyl transferase inhibitors (FTI) were once considered a method for inhibiting the RAS family of proteins (Figure. 1.7.2.).60,61 FTIs function by inhibiting the farnesyl transferase protein that targets RAS for prenylation at the CAAX motif. Once inhibited, the RAS protein is unable to anchor to the plasma membrane and remains inactive. FTI GGTI Figure 1.7.2. Inhibitors of KRAS prenylation.61 Unfortunately, most of the studies focused on inhibiting the related H-Ras protein family member. Through further studies, researchers found that the KRAS4B variant and N-Ras were able to be post-translationally modified by the geranylgeranyl transferase 1 enzyme. With the addition of the geranylgeranyl group, both KRAS4B and N-Ras variants were able to anchor normally to the plasma membrane thus rendering FTIs ineffective.60 Combination therapy with FTI and geranylgeranyl transferase inhibitors (GGTI) were attempted in order to specifically target KRAS.61 Although the treatment proved effective in vivo, the dosage required for KRAS inhibition in a mouse model was severely toxic.61 17 1.7.3. Inhibiting Sub-cellular Trafficking by Targeting PDE6δ Failure in targeting prenylation led to a focus on the inhibition of the interaction between PDE6δ and KRAS. The success of disruption would lead to sub-cellular localization of KRAS to the endomembrane leading to inactivation. In 2018, Chen et al. used a fragment-based drug design (FBDD) to create improved inhibitors of PDE6δ.62 Prior attempts at inhibition resulted in the design of the benzimidazole inhibitor named Deltarasin (Figure. 1.7.3A). A) C) B) Figure 1.7.3. Inhibitors of the PDE6δ-KRAS interaction. (A) Deltarasin, (B) quinazolinone-imidazole inhibitor 11b, (C) Molecular docking of 11b to PDE6δ.62 Although the drug had nanomolar affinity for PDE6δ, other non-specific interactions resulted in cytotoxicity. The FBDD approach revealed two active sites within PDE6δ. High affinity fragments were linked together as the basis of the newer lead compound. Structural activity relationship analysis led to the discovery of the quinazolinone-imidazole inhibitor, 11b (Figure. 18 1.7.3B). Through molecular docking, the new compound was found to hydrogen bond with arginine 61 and glutamine 88 (Figure. 1.7.3C). Treatment with 11b in Capan-1 pancreatic adenocarcinoma cells led to downregulation of the KRAS downstream effectors, Erk and Akt. Further proof of KRAS inactivation was seen in the enhanced sub-cellular localization of the protein to the endomembrane and concomitant apoptosis at 25 μM concentration.62 The PDE6δ inhibitor presents an alternative method for targeting KRAS driven cancers, however, more optimization work would be required to lower the dosage before attempting murine models. Targeting GEFs which activate KRAS for signaling is the next logical step if disruption of subcellular localization is deemed ineffective. 1.7.4. KRAS Activation Lowering by SOS1 Inhibition As mentioned previously, SOS1 predominantly catalyzes the exchange of GDP to GTP for KRAS activation. Thus, small molecule intervention to prevent SOS1 association with KRAS was explored. Using FBDD, Maurer et al. screened for novel compounds which interact with KRAS.63 Through heteronuclear single quantum coherence spectroscopy (HSQC) analysis of GDP-bound KRAS, a compound called DCAI was found to shift the residues V8, L56, D57 T74, and G75, suggesting a physical association with KRAS (Figure. 1.7.4). Figure 1.7.4. DCAI: An inhibitor of the SOS1-KRAS interaction.63 19 The residue site was found to be adjacent to the SOS1-KRAS interaction surface. DCAI was found to block SOS1-mediated nucleotide exchange in an in vitro nucleotide release assay. KRAS inactivation was detected in DCAI-treated HEK293T kidney cells as indicated by reduction of cRAF recruitment to the plasma membrane.63 Unfortunately, there were no further tests in cancer cell lines. DCAI merely represented the first step in designing a mature drug for the inhibition of SOS1 activation of KRAS. As part of FBDD, additional fragments which can bind adjacent surfaces of KRAS would need to be discovered. Through linking high affinity fragments, a more mature drug can be designed. 1.7.5. Mutant-specific Inhibition of KRAS G12C Direct targeting of KRAS is notoriously difficult due to the protein’s picomolar affinity towards GTP/GDP.64 Despite the difficulties, a novel allosteric site within the KRAS G12C mutant was reported in a landmark journal article.65 Ostrem et al. screened fragments for their ability to target the nucleophilic thiol of the cysteine 12 (C12) residue. C12 was chosen due to its location in the nucleotide binding pocket between switch I and switch II (Figure. 1.7.5C). Switch I and switch II were known for their conformational shift upon GTP/GDP binding. The fragment screen resulted with compound 6H05 being the most suitable hit (Figure. 1.7.5A). Co-crystallization of 6H05 with KRAS revealed the new allosteric site that was not previously found with other crystal structures (Figure. 1.7.5C). The residues glycine 60 and glutamate 99 form hydrogen bonds with 6H05 causing reordering of switch II residues and creating the switch II pocket (S-IIP). Further structure-activity relationship analysis of 6H05 revealed the new structure 12 that utilizes acrylamide as the electrophile (Figure. 1.7.5B). The rationale for the structural shift was to create a molecule that bound irreversibly to the mutant form which was a characteristic that was not seen with the disulphide compounds. Binding of the molecule causes KRAS G12C mutant to 20 preferentially bind GDP leading to the inactive state. In addition, the molecule impairs Raf association with KRAS which further downregulates the MAPK pathway.65 Further modifications of the G12C inhibitor by Ostrem et al. was performed by a joint project between Array BioPharma and Mirati Therapeutics in 2018.66 The authors kept the fragment containing the acrylamide linker and piperazine ring but made modifications to optimize non-covalent interactions in the S-IIP (Figure. 1.7.5D). A) C) O S N H N Cl S S H N O B) D) O Cl OH I N H OH O N N N N N O N N O N Figure 1.7.5. Allosteric inhibitors of KRAS G12C. A) Molecule 6H05, B) Molecule 12, C) Co-crystal of KRAS with 6H05 (cyan) and GDP, D) Array BioPharma & Mirati Therapeutics modification.65,66 21 Addition of the naphthalene ring allowed the compound to occupy a hydrophobic space proximate to methionine 72 in the S-IIP. The naphthalic hydroxyl group also allowed hydrogen bonding to aspartate 69, thus creating more specificity.66 The ability to target the G12C mutant is a monumental breakthrough and will pave the way for clinical trials for a sub-sector of the population. However, there are currently no other specific compounds for the other allelic forms of KRAS since the acrylamide linker only allows targeting of the cysteine residue. 1.7.6. Targeting Effectors in Related Pathways Current efforts for treating KRAS driven cancers relies on the vulnerabilities of targets in related effector pathways but the approach suffers from some disadvantages. The epidermal growth factor receptor (EGFR) is part of the signaling cascade leading to SOS activation of KRAS.67 EGFR inhibition have been studied in the past and led to the development of FDA approved drugs such as Erlotinib.68 Although patients respond well initially to the treatment, there is an inevitable acquired resistance from a secondary mutation. The mechanism for resistance lies in the mutation at exon 20 of the EGFR gene causing an amino acid substitution of threonine to methionine at position 790 (T790M).68 Inhibition of downstream effectors of KRAS have also been a popular approach. Raf-1 (alternate name of c-RAF) in the MAPK pathway is commonly inhibited by the potent drug, Sorafenib.69 The success of the drug lies in its ability to also inhibit important targets such as vascular-endothelial growth factor receptor (VEGFR) and the Raf homologue, BRAF.69 Resistance to Sorafenib have been reported.70 Cells which are resistant have been found to be more invasive and less reliant on anchorage-dependent growth.70 Mek inhibition with the allosteric inhibitors such as Trametinib, is another potential approach.71 The disadvantage of inhibiting Mek is the low therapeutic index. As reported in a phase II study, treatment with Trametinib at 22 meaningful therapeutic dosages can cause serious retinopathy in some patients; however, the findings are an improvement over predecessor drugs.71 The RAS field has grown substantially over the past few decades since its discovery as an oncogene and great effort at producing drugs for RAS-driven cancers have been documented. However, the direct targeting of all mutant forms of KRAS remains a monumental challenge. The investment into alternative approaches remains important. A field of the research is focused on the post-transcriptional regulation of KRAS. If one can down-regulate translation of KRAS in the implicated cancers, then the hypothesized result would be a subsequent decrease in proliferation. 1.8. Interaction Between KRAS and IMP-1 In 2011, Mongroo et al. confirmed the interaction between IMP-1 and KRAS from analysis of human colorectal cancer cells.27 In an effort to characterize the IMP-1-c-Myc interaction, siRNA against IMP-1 was transfected into SW480 colorectal cells. Western blot analysis showed that loss of IMP-1 did not change protein levels of c-Myc but lowered KRAS protein levels by 60%. Cell proliferation, growth and survival were also reduced with the siRNA treatment. Since IMP-1 regulates many oncogenes, a UV cross-linking experiment was performed to determine possible association with the KRAS transcript. After cross-linking, IMP-1 was immune-precipitated and was found to be in complex with both the coding region and 3’ UTR of KRAS.27 In 2018, Mackedenski et al. performed a more specific analysis of IMP-1 interaction with the KRAS transcript.72 Specificity for KRAS was further refined to the 1-185 nts segment of the coding region after systematic analysis of transcript fragments using the electrophoretic mobility shift assay technique. GXXG mutational experiments were also performed to determine domain requirements for binding the 1-185 nts fragment. It was found that all four KH domains are 23 required for high affinity towards the substrate. Immunoprecipitation experiments involving expressed mutant proteins in HeLa (ovarian cancer) cells verified the result (unpublished).20 A fluorescence polarization method was developed in an effort to conveniently study IMP1-RNA interaction and to screen for inhibitors of the IMP-1-KRAS mRNA interaction.73 Ms. Chuyi Wang, a prior lab member, screened a small library of 217 molecules of which a subset were dispiropyrrolizidine derivatives. Fluoresence anisotropy assay analyses revealed a positive hit called UNBC152, which inhibited the IMP-1-KRAS mRNA interaction in vitro (Figure. 1.8.1). Figure 1.8.1. UNBC152 inhibitor of the IMP-1-KRAS mRNA interaction.73 Subsequent testing of UNBC152 in two colorectal cancer cell lines confirmed knockdown of KRAS protein levels.73 The current model for the interaction of the drug is illustrated in Figure 1.8.2. In the presence of the inhibitor (UNBC152), IMP-1 is prevented from interacting with the KRAS transcript. KRAS RNA is then readily exposed and is vulnerable to decay by miRNA or endoribonuclease. 24 Figure 1.8.2. Hypothesis for the UNBC 152 inhibition of IMP-1-KRAS interaction. Interaction with either miRNA or siRNA will lead to translational inhibition and a subsequent decrease in KRAS protein levels. With the down-regulation of protein, proliferation of cells would decrease. The alternative pathway leading to endoribonucleic degradation of the KRAS after IMP-1 inhibition have yet to be seen. 1.9 Principle of Fluorescence Polarization As reviewed by Lea, fluorescence polarization (FP) analyzes molecular interactions through the usage of linearly polarized light.74 The linear light causes excitation in a fluorophore and emission is read by either a parallel or perpendicular emission polarizer.74 For the analysis of the IMP-1-KRAS mRNA interaction, the activity was monitored indirectly by excitation of the fluorescein-labeled KRAS substrate (KRAS-FL). The degree of polarization of KRAS-FL was dependent on both Brownian motion and ability to bind to IMP-1.74 The principle of FP is based on Perrin’s 1926 equation as follows: 𝜇𝜇 = (3𝜂𝜂𝜂𝜂)/(𝑅𝑅𝑅𝑅). The molecular relaxation time (µ) is proportional to the viscosity of the environment (𝜂𝜂), and the molecular volume (V). Molecular relaxation is inversely proportional to the gas constant (R) and temperature (T). Assuming stable atmospheric pressure, temperature and consistent viscosity, the rotation of the fluorophore is 25 dependent solely on the molecular volume.74 As the control, unbound KRAS-FL was tested in absence of IMP-1 protein. Since the molecular volume of the probe was low, the fluorophore tumbled rapidly in solution leading to scattering of the plane-polarized light (Figure. 1.9.1A.). A B Figure 1.9.1. Principle of fluorescence polarization assay. (A) Scattering of polarized light by unbound fluorophore, (B) Maintenance of polarized light by bound fluorophore. Image adapted from www.hi-techsci.com. 26 Upon reception by polarized filters, there was only a marginal difference between perpendicular and parallel light which lead to low FP units. When KRAS-FL bound to IMP-1, the molecular volume vastly increased thus the tumbling of the complex slowed down. The slower tumbling of the complex resulted in a significant increase of fluorescence polarization when compared to the random motion of unbound KRAS-FL (Figure. 1.9.1B.). 2.0 Research Goals This project is a continuation of the research on the UNBC152 compound, a molecule that has shown potential to inhibit the interaction between IMP-1 and the mRNA substrate, KRAS. Specifically, this project was focused on the confirmation of the active component of UNBC152, which, after chemical analysis, was discovered to be a complex mixture of components. Due to the nature of high-throughput library creation by parallel synthesis, individual compounds within a screening library, including UNBC152, are not necessarily confirmed to be pure. Therefore, a further refinement of the bioactive mixture is often necessary for substrate characterization and potential enhancement of bioactivity. To further add to this complexity in this particular case, the collaborator, Dr. Yeong Keng Yoon from Universiti Sains Malaysia, synthesized three batches of the UNBC152 compound, yet all 3 batches exhibited marked differences on ability to disrupt the IMP-1-KRAS RNA interaction as assessed by fluorescence polarization method. The first goal of my project was to synthesize UNBC152 in-house using the collaborator’s protocol and compare the composition of the synthesized batch to the original batch showing high activity towards the inhibition of the interaction between IMP-1 and KRAS mRNA. The second goal was to devise a purification strategy for the consistent isolation of any molecules of interest 27 within the UNBC152 mixture and the structural elucidation of these molecules. The third goal was to validate the performance of the purified products against the original high-activity batch of UNBC152, and to determine which pure species contributed to the original cellular activity in colorectal cancer cell (CRC) lines. 28 Chapter 2 Synthesis, Purification and Characterization of Spiropyrrolizidine and Piperazine Derivatives The main goal of this study was to find an inhibitor of the IMP-1-KRAS mRNA interaction. It began with a library screen conducted by Ms. Chuyi Wang of 217 small molecules generated by our collaborator, Dr. Yeong Keng Yoon from Universiti Sains Malaysia. The library was screened using the fluorescence polarization (FP) method (Section 1.9). Of the 217 samples screened, two spiropyrrolizidine derivatives were found to lower the percent bound of IMP-1 to KRAS mRNA: UNBC152 and UNBC143. UNBC152 was the most effective in lowering the fluorescence polarization signal; thus, suggesting that the percentage bound of IMP-1 to the fluorescein-labeled KRAS probe (KRAS-FL) was also lowered. LC-MS analysis by Dr. Yeong’s group, however, revealed that UNBC152 contained a complex mixture of compounds, most of which were unrelated to the proposed product (Figure 1.8.1). To add further complexity, the collaborator also synthesized three batches of UNBC152, and each batch had varying activity in the FP assay. The third batch of UNBC152 (UNBC152-3) was the most potent. This chapter examines the synthesis and evaluation of the molecules present in UNBC152 and the comparison of various synthetic batches with UNBC152-3 by thin layer chromatography (TLC), high performance liquid chromatography (HPLC) and low-resolution electrospray ionization mass spectrophotometry (LRESI-MS). The rationale for switching focus to the synthesis and purification of a dispiropiperazine derivative (SPOPP) will be discussed as well. Purification of the compounds was performed using flash chromatography as the initial step and HPLC was 29 used to refine the semi-purified product in order to meet biological assay requirements (≥95 % purity). Furthermore, purification of several side products formed during the synthesis of SPOPP was conducted to determine if there were other molecules with potential bioactivity. The experiments of this chapter were guided by results from biological analyses such as FP assay, Western blot, and the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Specific details regarding the bioassays conducted will be discussed in Chapter 3. Structural characterization using NMR was performed for validation of the correct product. 2.1 Methodology – Synthesis, Purification and Characterization of Spiropyrrolizidine and Piperazine Derivatives 2.1.1. General information Synthesis materials (acenaphthenequinone, 1-indanone, 2-bromobenzaldehyde and proline) were purchased from Sigma Aldrich and used without further purification. Solvents such as methanol, acetonitrile, hexanes and ethyl acetate were HPLC grade quality, purchased from either Sigma Aldrich or Fischer Scientific and used without further purification. Thin layer chromatography was performed on aluminum plates pre-coated with 0.2 or 1 mm silica gel with fluorescent indicator (UV254) or 0.2 mm reverse phase (C18, 23% carbon load). Flash chromatography columns were packed with 230-400 mesh silica gel (Silicycle) or reversed phase C18 (23% carbon load, 230-400 mesh; Silicycle). HPLC was performed using an Agilent 1200 Infinity series; the fraction collector module was the Agilent 1260 Infinity II series. Low resolution mass spectrophotometry (electrospray ionization) was performed using an Agilent 6120 LC-MS module. NMR structural characterization utilized either a Bruker Fourier 300 or Bruker Avance 600 system. 30 2.1.2. Synthesis of (2E)-2-(2-bromobenzylidine)-1-indanone (3)75 O O O NaOH + H EtOH, rt Br 1 2 3 Br Figure 2.1.1. Reaction scheme for the synthesis of (2E)-2-(2-bromobenzylidine)-1indanone (3). An equimolar ratio of 1-indanone (1, 19.5 mmol) and 2-bromobenzaldehyde (2, 19.5 mmol) was dissolved in 195 mL of ethanol (Figure. 2.1.1). To the mixture, 100 mL of 1M NaOH (aq) was added slowly while stirring, producing a white precipitate. The reaction stirred at room temperature for 4 h and reaction progress was monitored by TLC (1 ethyl acetate: 2 hexanes). Upon completion of the reaction, the solution was neutralized to pH 7 using 6 M HCl. The precipitate was removed by vacuum filtration and washed sufficiently with H2O. After air drying, the precipitate was recrystallized with 100% methanol to yield the (2E)-2-(2-bromobenzylidine)1-indanone product (3, 3.701g, 63%). 31 2.1.3. Synthesis of UNBC152 (7)76 Br O O O O + + OH O 3 4 O reflux NH Br N MeOH H 5 7 Figure 2.1.2. Synthesis of spiropyrrolizidine derivative, UNBC152 (7). An equimolar mixture of 3 (299 mg, 1 mmol), 4 (182 mg, 1 mmol) and 5 (115 mg, 1 mmol) was added to 10 mL of methanol (Figure. 2.1.2). The mixture was heated to reflux for 6 h and progress was monitored by TLC (1 ethyl acetate: 2 hexanes). After cooling, the product was isolated by vacuum filtration to yield the crude product as a yellow-colored solid (384 mg). 2.1.4. Synthesis of spiro[2′,3]-bis(acenaphthene-1′-one)perhydrodipyrrolo- [1,2-a:1,2-d]pyrazine (SPOPP; 8) and NP6A77 O O H + NH 4 O OH O 5 MeOH N N + reflux O O 8 N N O 9 Figure 2.1.3. Synthesis of dispiropiperazine derivatives SPOPP (8) and 9. 32 An equimolar mixture of 4 (1.822 g, 10.0 mmol) and 5 (1.151 g, 10.0 mmol) was heated in 10 mL of 100 % methanol. Reaction temperature was set at either 25, 35, 45 or 55 °C and reaction time was set between 1.5 and 24 h (Table 2.1.1.). Reaction progress was monitored using TLC (1 ethyl acetate: 2 hexanes). Typically, an orange precipitate was formed during the first hour. This precipitate changed to an orange-brown color after ~1 h reflux time and to a dark brown color after completion of the reaction. Table 2.1.1. Crude batches of 8 synthesized at varying times and temperatures. Name Time (h) Temperature (°C) VLA21 3 64 VLA25 3 64 VLA37 3 35 VLA159 3 35 VLA161 24 25 VLA175 3 45 VLB133 3 55 VLB135 3 35 VLB187 1.5 35 The solvent was removed under reduced pressure leaving a dark brown powder (1.772 g). Crude product containing both 8 and NP6A (250 mg) was mixed with 2 g of normal phase silica in 200 mL of 100 % methanol. Solvent was removed under reduced pressure and the silica mixture was dry-loaded into the column (10.5 cm height x 3.1 cm diameter). Purification by flash chromatography, using normal-phase silica (50 g) and a 9:1 mixture of hexanes and ethyl acetate at a flow rate of 12 mL/min resulted in the isolation of SPOPP (8) and NP6A. Two fractions, a yellow-colored band eluting at 320-405 mL and an orange-colored band eluting at 530-830 mL, were isolated. The purification of crude material was conducted 3 times and 686 mg of crude material was processed. The semi-purified orange product was dried under reduced pressure to 33 produce 280 mg of semi-purified SPOPP (8). SPOPP (8) was subsequently purified by HPLC using a Phenomenex phenyl-hexyl column (4.6 x 250 mm, 5 micron) and the purification procedure listed in Table 2.1.2. For the orange product (retention time of 4.08 min), a total of 12.5 mg was injected into the HPLC and 5 mg (40 %) of purified SPOPP (8) was retrieved. Table 2.1.2. HPLC purification procedure for SPOPP (8) and NP6A. Time H2O Acetonitrile (min) (%) (%) 0 20 80 6 8 92 8 20 80 Flow rate (mL/min) 2 2 2 NP6A, which was eluted as the yellow-colored band between 320-405 mL from the same purification step as SPOPP (8), produced 120 mg of semi-purified, yellow-colored NP6A after solvent was removed under reduced pressure. NP6A was also HPLC purified using the phenylhexyl column and the purification scheme listed above in Table 2.1.2. For the yellow product (retention time of 6.02 min), a total of 120 mg was injected into the HPLC and 21 mg (17.5 %) of purified NP6A was retrieved. 2.2 Results and Discussion 2.2.1. Synthesis of (2E)-2-(2-bromobenzylidine)-1-indanone (3) The synthesis of UNBC152 (7) began with the preparation of the starting reagent (2E)-2(bromobenzylidene)-1-indanone (3). The reaction proceeds via a base-catalyzed aldol condensation of 1-indanone (1) with 2-bromobenzaldehyde (2, Figure. 2.2.1.1). Deprotonation of 1-indanone (1), followed by nucleophilic attack on the carbonyl carbon of 2-bromobenzaldehyde (2) and proton transfer, led to the formation of the aldol product. Subsequently, dehydration yielded the condensation product 3. 34 OH O O H H O O H H Br 1 O H OH O H O OH O 2 H O O H H Br OH Br Br O 3 Br Figure 2.2.1.1. Mechanism for the base-catalyzed aldol condensation between 1-indanone (1) and 2-bromobenzaldehyde (2). The reaction was monitored using TLC. As seen in Figure 2.2.1.2, the 1-indanone (1) can be monitored using the vanillin stain whereas the 2-bromobenzaldehyde (2) was better monitored using UV light (short wave). When stained and heated, the 1-indanone (1) changes to a deep red color. 35 Figure 2.2.1.2. TLC reaction monitoring for the synthesis of (2E)-2-(Bromobenzylidene)1-indanone (3). Normal phase TLC used (1 ethyl acetate: 2 hexanes); Visualized with vanillin staining. Tracking the depletion of 1-indanone (1) as a marker for reaction completion was then relatively easy. Upon the addition of the 1M NaOH, the reaction began immediately. Reaction progress was monitored from 0 h and presence of 1-indanone (1, Rf = 0.53) was seen in reaction lane C. After 2 h, the reaction was monitored again and a new spot (Rf = 0.6) appeared which had a migration distance between 2-bromobenzaldehyde (2, Rf = 0.73) and 1-indanone (1). The reaction lane C did not contain presence of 1-indanone (1) as indicated by the lack of a red spot at Rf = 0.6. The reaction proceeded for another hour before the product was filtered. In order to ensure purity of the compound and to save time, the white precipitate was recrystallized in 100% methanol. After recrystallization, thin ivory-colored crystals remained. From the initial 2.575 g of 1-indanone (1) and 3.60 g of 2-bromobenzaldehyde (2) used for the reaction, 3.701g (63%) was retrieved in crystalline form. 2.2.2. Synthesis of UNBC152 (7)76 The formation of UNBC152 (7) was the result of a 1,3-dipolar cycloaddition reaction between an azomethine ylide (6) and (2E)-2-(2-bromobenzylidine)-1-indanone (3), which is a 36 popular synthetic route towards the regio- and stereoselective formation of 5-membered ring heterocycles.77 The synthesis began with a reaction between ʟ-proline (5), acenaphthenequinone (4, Figure. 2.2.2.1.). Acenaphthenequinone (4), in presence of weak acid, is protonated and subsequently undergoes nucleophilic attack by the nitrogen from the ʟ-proline (5). Proton transfer and dehydration generated zwitterionic iminium ion (I) which cyclizes to the lactone intermediate (II). Subsequent decarboxylation provides the azomethine ylide for the next part of the reaction. H O O O O O OH O H O OH O H N HN O H H HN 5 4 O H O O O O H O O N O N H O OH H N O I II -CO2 H H N O 6' O N Azomethine Ylide 6 Figure 2.2.2.1. Formation of the azomethine ylide (6) from the reaction between acenaphthenequinone (4) and ʟ-proline (5). 37 The formation of the azomethine ylide is key to the success of the 1,3-dipolar cycloaddition. Once generated, it reacts with (2E)-2-(2-bromobenzylidene)-1-indanone (3), as shown in Figure. 2.2.2.2., to form UNBC152 (7) in a concerted fashion. Br Br H O N O O N O 3 6 7 Figure 2.2.2.2. Mechanism for the 1,3-dipolar cycloaddition formation of UNBC152 (7). The reaction began with the reflux of equimolar portion of 3, 4, and 5. Reaction progress was monitored by normal phase TLC with the solvent system of 1 ethyl acetate: 2 hexanes (Figure. 2.2.2.3). At 0 h, differences in spots on the TLC can be seen in the reaction lane D. There is a long streak below reactant 3 (Rf = 0.63) that persisted past reactant 4 (Rf = 0.45). Disappearance of reactant 4 was seen at the 2 h and 20-min mark. The reaction was carried out for another 20 min past the 6 h reaction time as previously recommended by Dr. Yeong Keng Yoon. 38 Figure 2.2.2.3. Reaction monitoring for the synthesis of UNBC152 (7). Normal phase TLC used (1 ethyl acetate: 2 hexanes); Visualized with short wave UV. The rationale for extending the time was to ensure most of the reactants have been consumed in the reaction. However, even beyond the 6 h mark, some of reactant 3 remained as indicated by the reddish-colored spot at Rf = 0.63. Since the lab was interested in investigating the rationale for the third batch of UNBC152’s (UNBC152-3, 7) biological activity, the sample was compared to the newly synthesized batch (VLA9, 7) by TLC as an initial low-resolution analysis (Figure. 2.2.2.4.). Figure 2.2.2.4. Comparison of various batches of UNBC152 (7). Normal phase TLC used (1 ethyl acetate: 2 hexanes); Visualized by short wave UV. 39 UNBC152-3 (7) did contain some of reactant 3 but three other spots at Rf of 0.55, 0.29 and 0.0 were also present. VLA9 (7) shared similar spots as compared to the 1st batch of UNBC152 (7, UNBC152-1). When both samples were compared to UNBC152-3 (7), however, both lacked the streak found below Rf of 0.27. Since UNBC152-3 (7) had the most biological activity based on the highest reduction of fluorescence anisotropy (FA) units in FA analysis of the IMP-1-KRAS mRNA interaction, it was hypothesized that the contents within the streak were responsible for the in vitro activity. To test this hypothesis, crude VLA9 (7) was sent forth for initial biological testing. It should be noted that although crude material was not optimal for biological testing, the original batches of UNBC152 (7) were also tested as a crude material; this idea offered at least a first basis for comparison. 2.2.3. Synthesis of Spiro[2′,3]-bis(acenaphthene-1′-one)perhydrodipyrrolo- [1,2-a:1,2-d]pyrazine (8) As reported by Haddad et al. in 2015, the 1,3-dipolar cycloaddition which utilized ʟproline and acenaphthenequinone as the azomethine ylide can have a side reaction.77 The ylides (6) dimerize together to form the fused dispiropiperazines 8 and 9 (Figure. 2.2.3.1). As proposed by Haddad, the ylide forms in the transoid or cisoid orientation leading to the two dominant products of 8 and 9 with the other isomers 8ʹ and 9ʹ not formed due to steric hinderance. 40 Figure 2.2.3.1. Proposed Mechanism of SPOPP (8) by S. Haddad et al. 2015.77 Since this side reaction can occur, we wanted to synthesize 8 and 9 as controls for our bioassay. Equimolar parts of ʟ-proline (5) and acenaphthenequinone (4) were dissolved in 100% methanol and the reaction was monitored by TLC over ~3-h (Figure. 2.2.3.2). A TLC sample taken at the start of the reaction (0 h) displayed three additional spots to the starting material at the Rf of 0.0, 0.57 and 0.64. 41 Figure 2.2.3.2. Reaction monitoring for the synthesis of Spiro[2′,3]-bis(acenaphthene-1′one)perhydrodipyrrolo- [1,2-a:1,2-d]-pyrazine (8). Normal phase TLC used (1 ethyl acetate: 2 hexanes); Visualized with short wave UV. At the beginning of this investigation, it was quickly determined that the UNBC152-3 (7) reaction was complex and can produce multiple compounds that could be responsible for the previously observed bioactivity. Initially, the reaction produced an orange-colored precipitate. Beyond the 1 h mark, the precipitate began changing to a brown color. The presence of additional TLC spots from 0.0 to 0.25 Rf were first seen in reaction lane C after ~1 h, which suggested the creation of newer compounds not observed with VLA9 (7). After ~3 h, the TLC results did not appear to change and thus the reaction was stopped. By naked-eye, a yellow spot (Rf = 0.1) and an orange spot (Rf = 0) were observed, which was consistent with UNBC152-3 (1 ethyl acetate: 2 hexanes; Figure. 2.2.3.2.). Haddad reported that compounds 8 and 9 were orange and yellow respectively, which suggested that this reaction had generated these molecules as well. The crude material (VLA25, 8/9) was dried under reduced pressure and then compared to UNBC152-3 (7) by TLC (Figure. 2.2.3.3.). 42 Figure 2.2.3.3. Comparison between UNBC152 (7) and the two-component synthesis VLA25 (8/9). Normal phase TLC used (1 ethyl acetate: 2 hexanes); Visualized with short wave UV. From analysis, the UNBC152-3 (7) and VLA25 (8/9) both contained the tailing at the early retention factor region from 0.0 to 0.27. Since UNBC152-3 (7) was the most bioactive and VLA25 (8/9) also contained early Rf compounds, it was hypothesized that the piperazines 8 and 9 were responsible for the biological activity. More information on the bioassays will be given in chapter 3. In brief, FP analysis of VLA9 (7) on the IMP-1-KRAS determined that the mixture was not able to abrogate the interaction. The result was shocking considering the UNBC152-3 (7) was able to lower FP units in a dose-dependent manner. However, the VLA25 (8/9) sample was able to lower FP units, suggesting the inhibition of the IMP-1-KRAS mRNA interaction. Additionally, the VLA25 (8/9) sample was able to inhibit the interaction better than the original UNBC152-3 (7) sample created by our collaborator. The project focus was then switched from the purification of the 3-component synthesis of the 1,3-dipolar cycloaddition (isolation of 7 from UNBC-152) to the isolation and purification of 8 and 9. The purified compounds would then act as controls for the biological testing and would be compared against UNBC152-3 (7). 43 2.2.4. HPLC-MS Analysis of Synthesized Products. 2.2.4.1. Comparison of UNBC152 (7) and VLA9 (7) In order to confirm the assumptions resulted from the HPLC and LRESI-MS analyses of the crude reaction mixtures described above, the development of an isolation procedure for the compounds of interest was undertaken. In order to obtain higher resolution information on the components of UNBC152-3 (7) and VLA9 (7), the crude products were separated by means of high-performance liquid chromatography (HPLC). Material was purified using the Phenomenex phenyl-hexyl 4.6 x 250 mm analytical column using a gradient solvent system (80:20 acetonitrile to H2O initial; increase to 100% acetonitrile in 10 min). C18 was the initial matrix used, however, it was found that the phenyl-hexyl modification provided not only the hydrophobicity required for separation of spiropyrrolizidine compounds, but it also reduced the peak tailing (data not shown). Additionally, the phenyl group is theorized to have potential for greater π-π interaction with the naphthalene rings of 7 and provides alternate selectivity for stereoisomers. The gradient system also reduced peak width (data not shown). During the separation of UNBC152 (7) by HPLC, three dominant peaks were observed using a variable wavelength detector (VWD) set at 265 nm (Figure. 2.2.4.1). The peaks were found at 5.627, 6.945 and 7.996 min. VLA9 (7) also contained multiple peaks (5.335, 6.606, 7.640, and 9.108 min) using the same experimental conditions, however, the two reaction mixtures did not share many similarities in retention time. 44 Figure 2.2.4.1. HPLC VWD analysis and comparison of UNBC152-3 (7) and VLA9 (7). Absorbance was monitored at using a VWD at 265 nm. Analytes were chromatographed using a phenyl-hexyl 4.6 x 250 mm column with a gradient solvent system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). Electrospray ionization mass spectrophotometry was utilized to determine the lowresolution mass of the peaks identified by the HPLC separation (Figure. 2.2.4.2.). After separation by HPLC, a portion of the analytes were diverted to the mass spectrophotometer solvent stream which contained 0.1% formic acid. In brief, the formic acid acts as a positively charged vehicle to shuttle the analytes.78 Analytes exit the capillary tube inside the MS chamber and forms a Taylor cone. Eventually the coulombic repulsion forces exceed the surface tension of the cone and causes electrospray into the desolvation zone. A nebulizing gas (nitrogen) dries and shrinks the droplets, causing a build-up of positive charge at the perimeter of the droplet resulting in coulombic explosion of the droplets; this leads to the release of protonated analytes for entry into the quadrupole. The MS quadrupole separates analytes based on their mass to charge ratio (m/z) and then a detector detects the mass. Mass peak data was observed at 2.992, 5.778 and 7.112 min for 45 UNBC 152-3 (7). It should be noted that the delay in time for peaks observed in the ESI in comparison to the HPLC is due to a physical limitation of the system. The analyte is detected first in the variable wavelength detector (VWD) before traveling into the LRESI-MS for ionization and detection. An addition peak was observed at 8.224 min and tailed beyond 11.866 min. VLA9 (7) had three dominant mass peaks, two of which were broad. Peaks were found at 4.154, 7.197, 8.107 and 9.859 min. Figure 2.2.4.2. Low resolution ESI-MS comparison of UNBC152-3 (7) and VLA9 (7). Ions from 100 to 1000 m/z were analyzed. Analytes were chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient solvent system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). The first peak for UNBC152-3 (7) at 2.992 was a DMSO solvent peak as indicated by a [M2+H]+ of 157.1 m/z (Figure. 2.2.4.3). The peak at 5.778 min had a [M+H]+ peak of 300.0 which is the mass of the protonated form of 3 (299.16 g/mol); this indicated that not all of the reactants were consumed in the production of UNBC152-3 (7). Peaks at 7.112 and 8.224 min had a [M+H]+ 46 of 535 m/z. The 535 m/z correlates to a protonated form of 7, which normally has a molecular weight of 534.44 g/mol. The data suggested the presence of two stereoisomers of 7 were produced in the UNBC152-3 batch. Additionally, a m/z of 471.1 was detected at 7.818 min in the ESI scans; this suggested the potential presence of a fused dispiropiperazine such as 8 or 9 (Figure. 2.2.4.3). Figure 2.2.4.3. Ion analysis of the dominant peaks in the UNBC152-3 (7) sample by low resolution ESI-MS. Analytes were chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient solvent system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). 47 Although the abundance of the 471.1 m/z peak was at low levels, it was still detectable. The finding was congruent with what was reported in the Haddad article; that the 3 component 1,3-dipolar cycloaddition reaction may produce the fused dispiropiperazine side products 8 and 9.77 In comparison, VLA9 (7) contained trace amounts of the 300.0 m/z peak which also indicated an incomplete consumption of starting reactants (Figure. 2.2.4.4). Closer analysis of the peaks at 7.197, 8.107 and 9.859 min indicated presence of compound 7 (535 m/z). Figure 2.2.4.4. Ion analysis of the dominant peaks in VLA9 (7) by low resolution ESI-MS. Analytes were chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient solvent system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). 48 As there were three peaks found in VLA9 (7) which ionized to a mass of 535 m/z, it suggests the formation of three different isomers of 7. This finding was also congruent with Haddad’s article which stated that different time and temperature modifications to the 1,3-dipolar cycloaddition may produce upwards of three stereoisomers of 7.77 2.2.4.2. Temperature Dependency of the Synthesis of SPOPP (8) To determine if temperature could influence the formation of the biologically-active component of the SPOPP (8) reaction mixture, multiple batches were synthesized (Table. 2.1.1.). The crude reaction mixtures, which contained the unpurified versions of 8 (SPOPP) and 9 were analysed using HPLC and low resolution LRESI-MS. A VWD 265 nm scan of VLA37 (8/9), a batch synthesized at 35°C temperature, indicated several notable peaks at 2.792, 3.418, 3.686, 7.615, 8.522, 9.531 and 10.286 min (Figure. 2.2.4.5A.). Interestingly, the peak at 3.686 min was also found in the UNBC152-3 (7) sample (Figure. 2.2.4.1.). Four peaks, starting from 7.615 min, were all successfully ionized for LRESI-MS after protonation using 0.1% formic acid (Figure. 2.2.4.5B.). Closer analysis of the LRESI-MS spectrum indicated that all 4 peaks had a m/z 471.1 which matched the [M+H]+ form of 8/9 (Figure. 2.2.4.6.). The peaks were all well resolved by the phenyl-hexyl column using the gradient solvent system which suggested that the peaks were different isomers of each other. 49 A B Figure 2.2.4.5. HPLC-MS analysis of crude VLA37 (8/9). (A) HPLC scan at VWD 265 nm, B) low resolution ESI-MS trace; scan from 100 to 1000 m/z. Analytes were chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient solvent system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). This finding was not congruent with the report by Haddad, who indicated that only two fused dispiropiperazines will form from the dimerization reaction.77 Although it is true that there would be steric hinderance between the naphthalene ring and the pyrrolidine, making the formation of 8ʹ and 9ʹ unfavorable, this does not mean the regioisomers are impossible to form in small quantity under different reaction conditions (Figure. 2.2.3.1).77 50 Figure 2.2.4.6. Ion analysis of the dominant peaks in the VLA37 (8/9) sample by low resolution ESI-MS. Analytes were chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient solvent system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). Given that time and temperature variations affected the ratio of regio- and stereoisomers of 7 produced in the 1,3-dipolar cycloaddition, further syntheses of 8 were also conducted to probe similar potential effects. The reaction time (3 h) was not changed during this evaluation, but the reaction temperature was varied. Peaks relating to products with the correct m/z were seen at 7.3 min and beyond with the phenyl-hexyl separation (Figure. 2.2.4.7.). With the increase in temperature from 35 °C to 55 °C, early retention time peaks were reduced in magnitude relative to the major peak at 7.3 min. From the result, it was found that higher temperatures caused the formation of more compounds with m/z of 471.1. The peak at 7.3 min remained the dominant one 51 at 55 °C and was hypothesized to be 8, as that was the major product reported in the Haddad article.77 Figure 2.2.4.7. HPLC spectral analysis on the effect of temperature variance for the synthesis of 8. Absorbance was monitored at 265 nm. Analytes were chromatographed using a phenyl-hexyl 4.6 x 250 mm column with a gradient solvent system 80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). Table 2.2.4.1. Time and temperature dependent yield of 8 and NP6A according to area on VWD chromatogram. Name Time (h) Temperature % Product (8) % Product (°C) (NP6A) VLA161 24 25 22 6 VLB135 3 35 30 15 VLA175 3 45 38 9 VLB133 3 55 56 11 52 2.2.5. Isolation of SPOPP (8) and NP6A In order to confirm the assumptions made by the HPLC and LRESI-MS analysis of the crude reaction mixtures described above, the development of an isolation procedure for the compounds of interested was undertaken. The purification of several batches of the reaction mixture (VLB133; 8/9) began by flash chromatography as the initial step (Figure. 2.2.5.1.). Crude material was dry-loaded into the column due to solubility issues. Figure 2.2.5.1. Purification schematic for the isolation of 8 and NP6A. 53 Three column purifications were performed and a total of 686 mg of crude material was processed. Since the 1 ethyl acetate: 2 hexanes system was able to separate a yellow and an orange compound on TLC, the system was then adapted for column purification (Figure. 2.2.3.2.). The hexanes amount was changed to 90% to increase the retention time of the products and to ensure better separation between the orange and yellow compound, believed to be 8 and 9. NP6A (yellow solid, 120 mg) and SPOPP (8, orange solid, 280 mg) were retrieved from the column and dried under reduced pressure (Table. 2.2.5.1.). Table. 2.2.5.1. Purification of VLB133 by FCC normal phase. 686 mg of crude material was invested and was purified using 50g silica. Solvent system used was 1 ethyl acetate: 9 hexanes. Name NP6A NP6B (SPOPP;8) Elution Volume (mL) 320-405 560-830 Milligrams 120 280 Percent Yield % 17.5 40.7 The low percent yield after this step was not surprising considering there were multiple other peaks in the HPLC chromatogram of VLB133 (8/9, Figure. 2.2.4.7.). HPLC analysis determined that NP6A was 87% pure according to peak area (Figure. 2.2.5.2A.). 54 A B Figure 2.2.5.2. HPLC spectral analysis of NP6A. (A) Post FCC purification. (B) Post HPLC purification Absorbance was measured at 265 nm. Analyte was chromatographed with a phenylhexyl 4.6 x 250 mm column using the gradient solvent system (80:20 acetonitrile: H2O to 92:8 acetonitrile: H2O in 6 min; flow rate of 2 mL/ min). After purification by HPLC the semi-purified sample was increased to 97% purity (Figure. 2.2.5.2B.). LRESI-MS determined the purified compound to have a [M+H]+ mass of 471 (Figure. 2.2.5.3.). A sample was sent for high-resolution NMR analysis (600 MHz) to determine the identity of the molecule. Figure 2.2.5.3. Ion analysis of the purified NP6A by low resolution ESI-MS. Analyte was chromatographed with a phenyl-hexyl 4.6 x 250 mm column using the gradient solvent system (80:20 acetonitrile: H2O to 92:8 acetonitrile: H2O in 6 min; flow rate of 2 mL/ min). 55 Initially, the purification of NP6B was performed directly on the HPLC from the crude material, VLB133 (8/9, Figure. 2.2.5.1.). Crude material was solubilized in 80 % acetonitrile and syringe filtered with a 0.2 μm filter. Material was subsequently injected (100 μL per round) into the HPLC. After a total of 345 HPLC purification rounds, 28.7 mg of purified material was obtained. Secondary HPLC analysis of the combined fractions suggested 99 % purity (Figure. 2.2.5.4A.). A LRESI-MS scan from 100 to 1000 m/z only displayed a singular dominant peak, which further supported the conclusion that the sample was of high purity (Figure. 2.2.5.4B.). A B Figure 2.2.5.4. HPLC-LRESI-MS analysis of purified SPOPP (8). (A) Absorbance measured at 265 nm. (B) Mass was detected from 100 to 1000 m/z. Analyte was chromatographed with a phenyl-hexyl 4.6 x 250 mm column using the gradient solvent system (80:20 acetonitrile: H2O to 100% acetonitrile in 6 min; flow rate of 2 mL/ min). 56 Analysis of the mass scan at 4.001 min showed the ions of 471.2, 472.2, 473.2 and 236.1 (Figure. 2.2.5.5.). The m/z of 471.2 correspond to the [M+H]+ ion while the 236.1 corresponds to the [M+2H]+ ion. The compound was sent for structural analysis to confirm its identity. Since the purity was greater than 95%, it was also assessed for biological activity. SW480 colorectal cancer cells treated with SPOPP (8) revealed favorable bioactivity as determined by Western blot and, MTT assay analyses. More details will be discussed in Chapter 3. Figure 2.2.5.5. Ion analysis of the purified SPOPP (8) by low resolution ESI-MS. 2.2.6. Purification of Other Peaks in the SPOPP (8) Mixture 2.2.6.1. Initial Purification and Analysis The HPLC chromatograms (VWD 265 nm) of the crude synthetic batches of 8 (VLA175, VLB133 & VLB135), showed that there were other compounds present that did not correspond to 8 or isomers of 8 (Figure. 2.2.4.7.). The original bioactive sample UNBC152-3 (7) also contained additional peaks that did not correspond to these expected products either (Figure. 2.2.4.1.). One peak of interest, which had a retention time of 3.7 min on the phenyl-hexyl column separation, was found to be common to both UNBC152-3 (7) and for all synthetic batches of 8 (Figure. 2.2.4.1. and 2.2.4.7.) and we decided to investigate these compounds further. 57 To this end, crude VLA37 (8/9) was fractionally separated using a C18 matrix and fractions were collected based on the color of the band on the column. The crude was separated using the same flash column as previously described for the separation of 8 and 9. C18 silica was packed into the column using 100% isopropanol and the column was equilibrated with 5 column volumes of the solvent system, 8 H2O: 2 methanol: 4 acetonitrile: 0.25 acetic acid. Following equilibration, crude material (200 mg) was dry-loaded with 2 g of C18 silica. Fraction 1 (F1) was eluted in the first 30 mL and the solution was placed in a round bottom flask (Figure. 2.2.6.1). Figure 2.2.6.1. Purification schematic to produce fraction 1 (F1) components of crude 8/9. The solvent was removed under reduced pressure to produce a reddish-brown powder (21 mg, 10%) (Figure. 2.2.6.2A.). Subsequent fractions were also collected in blocks of 30 mL portions. Naming of the fractions follows the convention of VLBX-Y, where X is numbered after the page number the experiment was recorded on and Y is numbered after the fraction number. On TLC, F1 was not identified as a pure compound, however, it did have a dominant spot at a Rf of 0.23 (Figure. 2.2.6.2B.). 58 A B Figure 2.2.6.2. Reverse phased (C18) fractionation and TLC analysis of VLA37 (8/9). (A) Approximately 50 mL fractions. (B) TLC of fractions. FCC solvent system used was 4 acetonitrile: 2 methanol: 8 H2O: 0.25 acetic acid; TLC solvent system for analysis was 7 ethyl acetate: 2 methanol: 1 H2O: 0.25 acetic acid. Visualization was done with long wave UV. HPLC analysis of VLA37-1 using VWD at 264 nm revealed that the sample had 2 major peaks at 2.370 and 2.588 min (Figure. 2.2.6.3A). The peaks, however, were not well ionized in the LRESI-MS, and only one mass peak seen was at 2.850 min (Figure. 2.2.6.3B.). This peak had a m/z of 157.1, which was the DMSO solvent peak commonly found as the [M2+H]+ ion (Figure. 2.2.6.4.). The sample was originally dissolved in DMSO for biological testing, so this finding was not unexpected. VLA37-1 demonstrated pronounced biological activity in terms of abrogating the IMP-1-KRAS mRNA interaction (Chapter 3, section 3.3.1.). The fraction also displayed activity in lowering KRAS protein expression for SW480 human colorectal cancer cells. The sample was consumed quickly from biological testing and further F1 samples were produced to see if the bioactivity was repeatable and to generate more material (VLA159, VLB135 & VLB187) for subsequent purifications. 59 A B Figure 2.2.6.3. HPLC-LRESI-MS analysis of VLA37-1. A) VWD 265 nm, B) MS abundance scan from 100 to 1000 m/z. The analyte was chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). Figure 2.2.6.4. Ion analysis of the dominant peak in VLA37-1 fraction by LRESI-MS. The analyte was chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). A subsequent yellow-colored fraction (F2) contained a dominant TLC spot at a Rf of 0.9 (Figure. 2.2.6.2B). Fractions containing the dominant yellow compound were pooled together and further purified using preparative TLC (9 hexanes: 1 ethyl acetate). The band containing the yellow 60 compound was scraped from the TLC, extracted with 100% acetonitrile, filtered, and dried under reduced pressure. The newly dried product was named NP4. NP4 contained two dominant peaks: one at 2.485 min and one at 3.7 min (Figure. 2.2.6.5A.). The only peak which ionized in this sample was also the DMSO solvent found at 2.850 min (Figure. 2.2.6.5B.). A B Figure. 2.2.6.5. HPLC-MS analysis of NP4. A) VWD 265 nm, B) MS abundance scan from 100 to 1000 m/z. The analyte was chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). At approximately, 3.9 min of the LRESI-MS spectra, an ion of 250.1 and 251.1 was found (Figure. 2.2.6.6.). Since NP4 was found at 3.7 min, and a similar peak was observed at the same retention time in the original analysis of UNBC152-3 (7), it was hypothesized that this compound could represent a common side product in both reactions. The identity of the compound is currently unknown, however, FP assay determined that NP4 significantly lowered IMP-1 percent bound to 61 KRAS-FL probe but caused total fluorescence to increase by 289 % (section 3.3.1.) NP4 also displayed modest bioactivity in SW480. The complications of NP4 presence will be discussed later. Figure 2.2.6.6. Ion analysis of NP4 from LRESI-MS. Analyte was chromatographed with a phenyl-hexyl 4.6 x 250 mm column using the gradient solvent system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/ min). 2.2.6.2. F1 Sample Production as a Means for Comparison Various other batches of F1 were also obtained from the temperature dependent experiments mentioned previously (section 2.2.4.) Several were also synthesized at 35 °C to determine reaction reproducibility (Table 2.2.6.1.). The C18 purification of crude samples was repeated to see possible differences in peak profile data on both the HPLC and LRESI-MS ions. Various F1 samples were produced by FCC C18 purification as previously mention and analyzed by both HPLC and LRESI-MS (Table 2.2.6.1.). The F1 samples, VLB188-1 and VLB189-1 were purified from crude batches, VLB187 (8/9) and VLA161 (8/9) respectively. VLB187 (8/9) was synthesized at 35 °C for 1.5 h and VLA161 (8/9) was synthesized at 25 °C for 24 h. Similar peaks were seen from the HPLC traces of VLB165-1, VLB159-1, VLB188-1 and VLB189-1 (Figure. 2.2.6.7). The peak at approximately 3.7 min, which corresponded to NP4, can be seen for all four 62 samples, regardless of time or temperature variation in the prior synthesis step. The only difference between the synthesis variants was the relative intensity for the peak at 2.940 min; this peak was the largest relative to the others for the 24 h synthesis at room temperature (Figure. 2.2.6.7D). A) B) C) D) Figure 2.2.6.7. HPLC trace comparison of various fraction 1 samples from C18 FCC purification of crude batches of 8. A) VLB165-1, B) VLB159-1, C) VLB188-1, 1.5H, D) VLB189-1 RT, 24H. Crude batches were synthesized at 35°C for 3 h unless otherwise stated. The analyte was chromatographed with a phenyl-hexyl 4.6 x 250 mm column using a gradient system (80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate of 1 mL/min). The presence of noteworthy ions was recorded and the results were compared across all main samples (Table 2.2.6.1.). The [M+H]+ m/z ion of 183 was tracked due to the initial presence in Dr. Yeong Keng Yoon’s LC-MS analysis of the UNBC152 (7) batches. The 199 and 200 m/z ions were hypothesized to be 1,8-naphthalic anhydride (198.174 g/mol) which is the oxidized version of acenaphthenequinone (4); this ion was found in both samples of the acenaphthenequinone (4) reagent purchased from Sigma-Aldrich and Alfa Aesar. Presence of 199 63 m/z in the F1 fractions would then suggest remaining unreacted material. The 250 m/z ion was found in almost all the F1 fractions, as well as both UNBC152-3 and UNBC142; this indicated that NP4 was present. UNBC142 was another library compound in the 1,3-dipolar cycloaddition series. The synthesis of UNBC142 involved similar reagents to those found for UNBC152 (see section 2.1.3.) but the dipolarophile was different. It originally was used by Ms. Chuyi Wang as a negative control for the cellular bioassays. Table 2.2.6.1. Comparison of notable ions found by LRESI-MS analysis of F1 samples of 8 as compared to the collaborator’s UNBC152-3 (7) Presence of ion is indicated by the letter Y while not detectable is indicated by N. Numbers before Y indicate more than 1 occurrence of the specified ion. All F1 components were kept in semi-crude form when testing for biological activity with respect to the IMP-1-KRAS mRNA interaction; this was strategically done to determine if the sum of all the side products were able to interrupt the protein-RNA interaction. Although testing crude is not optimal, the original UNBC152-3 (7) was found in its crude form. The semi-crude F1 fraction would serve as a basis for comparison against the original UNBC152-3 (7). F1 was found to inhibit the IMP-1-KRAS mRNA interaction as found by FP assay (Chapter 3, Figure. 3.3.3.). 64 Additionally, various F1 samples lowered cellular proliferation of SW480 CRC as determined by MTT assay (Figure. 3.3.10.). Given this result, purification of the F1 samples were investigated to determine if it was possible to purify further and perhaps increase bioactivity. 2.2.6.3. HPLC Optimization of Fraction 1 Mixtures F1 samples were purified further on the phenyl-hexyl column to determine if the most abundant peak was contributing to the observed biological activity (Figure. 2.2.6.8.). Batch VLB188-1 was used to optimize purification procedures. The solvent system was modified in order to better separate and retain the fast-eluting peaks observed when using the 80-92 % acetonitrile in H2O gradient (0 to 6 min) system. Initially, the solvent system was reduced to 50:50 acetonitrile: H2O, increasing to 70:30 acetonitrile: H2O over a period of 6 min. By reducing the percentage of acetonitrile, the eluting compounds were expected to better interact with the hydrophobic phenyl-hexyl matrix of the column. Additionally, 0.2% formic acid (aq) was used as an additive to neutralize anionic molecules or protonate basic molecule, with the goal of potentially changing column selectivity. If by protonation, some compound became more neutral, the retention time would then increase and would cause better separation. The change in solvent system led to better retention of compounds (Figure. 2.2.6.8B.). 65 A B C Figure 2.2.6.8. HPLC optimization of the purification of VLB188-1. (A) 80:20 acetonitrile: H2O to 100% acetonitrile in 10 min; flow rate 1 mL/min, (B) 50:50 acetonitrile: H2O to 70:30 acetonitrile: H2O in 6 min; 0.2% FA; flow rate 0.7mL/min. (C) Acenaphthenequinone (50:50 acetonitrile: H2O to 70:30 acetonitrile: H2O in 6 min; 0.2% FA; flow rate 0.7mL/min.). Analytes were chromatographed with a phenyl-hexyl 4.6 x 250 mm column. A large peak was seen at 2.936 min with several smaller peaks at 4.242, 5.844 and 8.095 min. Fraction-collection of the peaks at 2.936 and 4.242 min became the focus for this part due to their abundance. The other peaks were not collected due to the lower abundance. As another purification check, the retention time of acenaphthenequinone (AcQ) was analyzed under the same 66 solvent system. Presence of AcQ would be problematic in the bioassays due to its previously reported bioactivity in A549 lung cancer cells.79 AcQ was found to be retained at 8.676 min in the new solvent system conditions and was safely excluded from the purification process (Figure. 2.2.6.8C.). The flow rate at 0.7 mL per min was increased to 1.0 mL per min to increase acquisition rate without decreasing peak separation (data not shown). As seen in Figure 2.2.6.9., the purification process for VLB188-1 was only moderately successful. Fraction 1 (VLB188-1.1) of the newly collected sample contained one dominant peak at 1.933 min and was a total peak area of 82 % (Figure. 2.2.6.9A.). The peak did not have a gaussian distribution which may indicate presence of other over-lapping peaks. Smaller peaks can be seen at 2.917 min and beyond but the total area of the peaks only totaled to 18 %. Fraction 2 (VLB188-1.2) and 3 (VLB188-1.3) have more peaks compared to VLB188-1.1 (Figure. 2.2.6.9.). 67 A B C Figure 2.2.6.9. HPLC trace of fractionated VLB188-1. A) Fraction 1, B) Fraction 2, C) Fraction 3. Analytes were chromatographed using a phenyl-hexyl 4.6 x 250 mm column using a gradient solvent system (50:50 acetonitrile: H2O to 70:30 acetonitrile: H2O in 6 min; 0.2% FA; flow rate of 1.0 mL/min). Collection of the regions encompassing the peaks originally found at 2.936 and 4.242 min, caused enhancement of previously unseen peaks and further refinement of the fractions was halted due to a shift in experimental priorities. The stability of these compounds was also questionable since there was a chance that these were merely intermediates in producing the fused piperazines 68 8 and 9. Regardless, these samples were also tested for possible biological activity. MTT assay determined that the more purified fractions such as VLB188-1.1 had reduced activity as compared to the semi-purified F1 form (Figure. 3.3.11.). In light of this, the pursuit in the further refinement of F1 was lowered in experimental priority. 2.2.7. Structural Characterization 2.2.7.1. (2E)-2-(2-bromobenzylidine)-1-indanone (3) 2.2.7.1.1. Analysis The 1H-NMR spectra of 3 matches the 1H-NMR spectra reported previously by Martinez et al. (Figure. 2.2.7.1.).80 The peak at 3.99 ppm corresponded to the CH2 proton located on the indanone ring. Eight aromatic protons can be found from 7.30 to 7.72 ppm. The vinylic proton was was found at 8.01 ppm and is overlapping a signal from an Ar-H proton. The peaks at 3.52, 1.73, 1.29 and 0.90 ppm all integrated to less than 1, which indicated the presence of minor contaminants. The peak at 3.52 ppm corresponded to methanol. The two peaks at 1.29 and 0.90 ppm corresponded to hexanes. The peak at 1.73 ppm is unknown. Since the spectra of 3 matches the 1H-NMR spectra as reported previously by Martinez et al., no further analysis was pursued in this case. 2.2.7.1.2. NMR Spectrum Ivory colored thin needles; Rf = 0.58 (1 Ethyl acetate: 2 hexanes); Molecular formula of C16H11OBr; LRESI-MS [M+H]+ m/z of 300; 1H-NMR (300 MHz, CDCl3) δ = 3.99 (s, 2H, CH2), 7.30 (m, 1H, Ar–H), 7.43–7.72 (m, 6H, Ar–H), 8.00 (d, 1H, J = 7.6 Hz, Ar–H), 8.01 (s, 1H, HC=C). 69 Figure 2.2.7.1. 1H-NMR spectra of (2E)-2-(2-bromobenzylidene)-1-indanone (3). Sample was dissolved in CDCl3 and processed at 300 MHz. 2.2.7.2. SPOPP (8) 2.2.7.2.1. Analysis The SPOPP (8) structure (Figure. 2.2.7.2) contains an axis of symmetry, which is demonstrated in the 1H-NMR spectra. The heterocyclic protons adjacent to N on C14/25 and C12/28 were the most uniquely identified for this structure. Due to the stereocenters at the piperazine rings, the pyrrolidine ring residues shared different integration values for the CH2 groups. Due to the proximity to the aromatic ring system, H-12/28 (2 x CH for each) are slightly more de-shielded and found at 2.15 and 2.6 ppm. H-14/26 (2 x CH) were the most de-shielded due to proximity to the nitrogen atom. H-15 (CH2) and H-30 (CH2) are proximal to H-14 and H-26 70 which caused a slight shift to 1.97 and 1.7 ppm. H-11 (CH2) and H-29 (CH2) were the least deshielded and were found at 1.78 and 1.61 ppm. Figure 2.2.7.2. Numbered positions for the structure of SPOPP (8). The 13C-NMR displayed a peak at 207.3 ppm which indicated the presence of the carbonyl carbons of C-35 and C-33. From 119.4 to 141.4 ppm are the signals for the aromatic carbons found for the naphthalene rings. Since these rings is equivalent, only 10 peaks were visible for the 20 total carbons. The distinguishing signals in this case were the aliphatic signals located at 72.3, 60.5, 47.6, 27.9 and 21.3 ppm. Due to the axis of symmetry, the spectra only displayed 5 peaks out of the possible 10 carbon signals. 71 Table 2.2.7.1. Chemical shifts of protons and carbons for SPOPP (8). 1 13 Position H Shifts (ppm) C Shifts (ppm) 1, 21 7.18 127.98 2, 20 7.38 124.8 3, 19 130.16 4, 18 141.4 5, 33 136.16 6, 16 7.46 123.28 7, 25 7.67 130.77 8, 24 7.43 127.68 9, 23 7.65 119.4 10, 22 134.09 11, 29 1.61, 1.78 21.3 12, 28 2.15, 2.6 47.6 14, 26 3.89 60.5 15, 30 1.7, 1.97 27.9 31, 32 72.3 35, 33 207.3 C-31 and C-32 were located at 72.3 ppm due to their position in relation to both the nitrogen atom and the carbonyl. C-14 and C-26 were found at 60.5 ppm, which is congruent with the de-shielding by the nitrogen atom found in the 1H-NMR. C-12 and C-28 were found at 47.6 ppm due to proximity to the nitrogen atom. C-15 and C-30 were located at 27.9 ppm. C-11 and C-29 were the most shielded and was found at 21.3 ppm. Homonuclear correlation spectroscopy (COSY) was performed to determine proton-proton coupling by neighboring protons (Figure. 2.2.7.5.). Correlations were observed between protons at 3.89 ppm (H14/26) peak, those at 1.97 ppm and 1.7 ppm (H15/30. The protons at 2.6 ppm (H12/28) correlated to 1.78 ppm (H11/29) and 2.15 ppm (H12/28). The peak at 2.15 ppm (H12/28) also correlated to those at 1.78 ppm (H11/29). The protons at 1.97 ppm (H15/30) correlated to those at 1.7 ppm. The protons at 1.78 ppm correlates to those at 1.61 ppm. Although this spectrum 72 identifies both geminal and vicinal coupling interactions, the absolute stereochemistry of the molecule cannot be explicitly determined. Given that there are two pyrrolidine rings for the dispiropiperazine isomers, the maximum proton signals for the aliphatic hydrogens would be seven if the protons were all shielded differently. There are seven visible signals between 1.69 to 3.98 ppm which indicated some symmetry. However, a truly symmetrical molecule would have only four proton signals between 1.69 to 3.98 ppm; the integration values for the protons would also be twice the value due to overlapping signals. In this case, most of the proton signals have an integration value of approximately 2. Haddad et al also indicated that the more symmetrical isomers, 8ʹ and 9ʹ, would have these overlapping signals for the aliphatic region (Section 2.2.3.).77 Haddad et al. required x-ray diffraction to solve the structure of 8 and thus the technique might be necessary to solve this structure.77 Regardless, the 1H-NMR spectrum of SPOPP matches what is reported in the literature as compound 8. Heteronuclear single quantum correlation experiment (HSQC) determined correlation between the aliphatic protons from the 1H-NMR spectra and the 13C-NMR spectra for SPOPP (8, Figure. 2.2.7.6.). The H14/26 protons at 3.89 ppm were found correlated with the carbon signal at 60.5 ppm. Due to a degree of symmetry, the three carbon signals at 47.6, 27.9 and 21.3 ppm each correlated to two proton signals. The carbon signal at 21.3 ppm correlates with H11/29. The carbon signal at 27.9 ppm correlates with H15/30. The carbon signal at 47.6 ppm correlates with H12/28. 2.2.7.2.2. NMR Spectrum Orange powder; Rf = 0.33 (1 Ethyl acetate: 9 hexanes, ran twice); LRESI-MS [M+H]+ m/z of 471 1H-NMR (600 MHz, CDCl3) δ = 7.17−7.68 (m, 12H), 3.89 (m, 2H), 2.59−2.62 (m, 2H), 2.13−2.17 (m, 2H), 1.96−2.01 (m, 2H), 1.69−1.82 ppm (m, 6H); 13C-NMR (125 MHz, CDCl3) δ 73 = 141.42, 136.86, 134.07, 130.76, 130.18, 127.93, 127.68, 124.81, 123.30, 119.35, 72.30, 60.54, 47.63, 27.89, 21.31 ppm. Figure 2.2.7.3. 1H-NMR spectra of SPOPP (8). Sample was dissolved in CDCl3 and processed at 600 MHz. 74 Figure 2.2.7.4. 13C-NMR spectra of SPOPP (8). Sample was dissolved in CDCl3. 75 Figure 2.2.7.5. COSY spectra of SPOPP (8). Sample was dissolved in CDCl3. 76 Figure 2.2.7.6. HSQC of SPOPP (8). Sample was dissolved in CDCl3. 77 2.2.7.3. NP6A 2.2.7.3.1. Analysis The spectra of NP6A is different from what was reported in the literature for compound 9 (Figure. 2.2.7.7.). Initially, we hypothesized that NP6A was compound 9, based on the LRESI-MS data and the report by Haddad.77 Upon NMR analysis of NP6A, however, several discrepancies were identified. The aromatic signals of NP6A were much more symmetrical than those observed for 9 (Figure. 2.2.7.8.). Three clusters of peaks, each integrating to 4, can be seen for NP6A, which is not the case for compound 9 (Figure. 2.2.7.7.). The peak observed at 4.0 ppm for 9 is not present in NP6A, however, NP6A was observed to have a slightly more downfield peak at 4.88 ppm (Figure. 2.2.7.7.). The peak at 7.26 ppm was due to the residual solvent signal of CDCl3 while the peak at 0.0 ppm was the TMS standard. Ten signals could be found within the aromatic region for 13 C-NMR which still indicated the symmetrical naphthalene rings of the fused piperazine (Figure. 2.2.7.9.). The difference was the aliphatic region, which only contained 4 carbon peaks as compared to the 13C-NMR for 9 which contained 5 (Figure. 2.2.7.10.). Unfortunately, the COSY spectrum of NP6A did not give enough information to correctly identify which regioisomer it was (Figure. 2.2.7.11.). Additional spectral analysis might not correctly lead to the solution of the NP6A structure. According to Haddad, x-ray diffraction was utilized to differentiate between 8 and 9. Solving NP6A might require the same technique. 2.2.7.3.2. NMR Spectrum Yellow powder, Rf = 0.53 (1 ethyl acetate: 9 hexanes, resolved twice on the same TLC); LRESI-MS [M+H]+ m/z of 471. 1H-NMR (300 MHz, CDCl3) δ = 8.14-8.16 (m, 4H), 7.76-7.93 (m, 4H), 7.74-7.76 (m, 4H), 4.88 (m, 2H), 2.45-2.55 (m, 4H), 1.37-1.53 (m, 7H), 0.56-0.58 (m, 78 2H). 13C-NMR (125 MHz, CDCl3) δ = 208.89, 141.95, 140.84, 132.22, 131.74, 130.29, 128.70, 128.04, 124.45, 121.61, 120.58, 60.67, 60.60, 46.12, 24.80. Figure 2.2.7.7. 1H-NMR spectra of NP6A. Sample dissolved in CDCl3 with 0.3% TMS and processed at 300 MHz. 79 Figure 2.2.7.8. 1H-NMR spectra of 9. Sample dissolved in CDCl3 and processed at 300 MHz..77 80 Figure 2.2.7.9. 13C-NMR spectra of NP6A. Sample dissolved in CDCl3 with 0.3% TMS. 81 Figure 2.2.7.10. 13C-NMR spectra of 9. Sample dissolved in CDCl3 with 0.3% TMS.77 82 Figure 2.2.7.11. COSY of NP6A. Sample dissolved in CDCl3 with 0.3% TMS. 2.2.8. Summary The original goal of replicating UNBC152-3 (7) changed drastically with the discovery of the side product 8. The combination of TLC, HPLC and LRESI-MS suggested that the original UNBC152-3 contained 8. Additionally, FP analysis led us to focus on crude 8 due to its enhanced 83 activity at reducing the percentage of IMP-1 binding to KRAS-FL. Purifications of the crude dispiropiperazine mixture containing 8 resulted in the retrieval of both 8 and a structurally unknown isomer called NP6A. The result was surprising since NP6A did bear the same yellow coloration of 9 as described in the Haddad et al. article.77 Additional experiments are required to elucidate the structure of NP6A; perhaps X-ray diffraction or nuclear overhauser effect spectroscopy (NOESY) could be used. If the results are positive, NP6A would represent a new, unreported dispiropiperazine derivative that can be generated from a one-pot synthesis. The purification of other side products generated by the synthesis 8 did not result in generating a more bioactive molecule. These side products could not be purified further by HPLC; this could be a result of the side products being unstable. Regardless, NP6A and 8 were isolated and a detailed analysis of their bioactivity is discussed in Chapter 3. 84 Chapter 3 Biochemical Characterization of Spiropyrrolizidine and Piperazine Derivatives This chapter focuses on the biochemical analyses of the purified and semi-purified synthetic compounds described in Chapter 2. The bioassays performed followed the same techniques that were previously used by Ms. Chuyi Wang in the analysis of UNBC152-3 (7). Initial test with fluorescence polarization (FP) assay was aimed at determining which of the compounds in UNBC152-3 (7) were responsible for the reduction in FP. The crude mixture of dispiropiperazines 8 and NP6A was initially used as a control to compare alongside UNBC152-3 (7) in the FP assay. Western blot analyses were performed to determine the effects of 8, NP6A, 3, 4, and 5 on the repression of KRAS protein expression in SW480 and HT29 human colorectal cancer cells. Preliminary cellular analyses with Fraction 1 (F1) components was also explored. The rationale for focusing on compound 8 (SPOPP) is discussed herein. An MTT assay was conducted on all relevant compounds and mixtures to determine their effect on cellular proliferation. The rationale for this choice was that there can be a discrepancy between protein repression and decrease in cellular proliferation. A quantitative polymerase chain reaction was performed to determine steady-state mRNA levels of KRAS; this was to determine the potential mechanism for the repression in KRAS protein expression. Lastly, investigations into the biochemical mechanism of the anti-proliferative activity of SPOPP (8) were conducted using flow cytometry analyses. 85 3.1 Methodology – Fluorescence Polarization Assay to Assess the Effect on IMP-1-KRAS mRNA Interaction 3.1.1. Generation of Recombinant IMP-1 3.1.1.1 Protein Generation For the fluorescence polarization assay, two truncated IMP-1 variants were generated. One truncation removed the N-terminal RRM1 and RRM2 motifs leaving the presence of KH1 to KH4 domains (KH1to4) and the other truncation removed the RRM1, RRM2, KH1 and KH2 domains, leaving only KH3 and KH4 (KH3to4) present. The gene for the truncations was previously cloned into the pET41c plasmid, sequenced by Macrogen and verified using FinchTV and Serial Cloner. Plasmids were transformed into the calcium competent E. coli BL21 strain by heat shocking at 42 °C. Transformed BL21 was plated on Luria-Bertani (LB) broth plates containing kanamycin (25 μg/mL) and grown overnight. After overnight growth, approximately 100 colonies were selected to grow as a starter culture in 100 mL of LB-kanamycin broth for 3 h at 37 °C. The culture was then transferred to a Fernbach flask containing 900 mL LB-kanamycin broth for further growth. Once the OD600 reached 0.5, IPTG (1 M) was added to induce protein production. The culture was then incubated for another 4 h. Cells were then pelleted at 3000 x g for 15 min and the supernatant was removed. Cell pellets were then transferred to the -80 °C freezer for storage. 3.1.1.2 Protein Purification The cell pellets were thawed on ice and then resuspended with 12 mL Buffer B (8 M urea, 100 mM NaH2PO4, 10 mM Tris-Cl; pH 8) for lysis. The slurry was transferred to a 50 mL conical tube and placed on a waver for 1 h. After incubation, the sample was pelleted at 13200 rpm for 30 min and the supernatant was removed. The supernatant was then syringe filtered (0.45 μM) to a new 50 mL conical tube and 2 mL of Ni-NTA resin was added. The mixture was incubated for 1 86 h for binding. Buffer C (pH 6.5), D (pH 6.3), E (pH 5.9) and F (pH 4.5) were generated from Buffer B in preparation for column purification. The lysate-bounded Ni-NTA resin was added to a BIO-RAD Poly-Prep® column (0.8 x 4 cm). Impurities were washed with 5 mL Buffer B, 5 mL Buffer C and 5 mL Buffer D. Protein was then collected in 1 mL fractions upon addition of 12 mL Buffer E and 12 mL Buffer F. The purity of the fractions was analyzed using a 10% SDS-PAGE gel and visualized using Coomassie blue. 3.1.1.3 Protein Refolding and Quantification Approximately 1 mL of the purest protein fractions were dialyzed using the Thermo Scientific Slide-A-Lyzer™ dialysis cassette (3500 MWCO) in 250 mL Buffer A (200 mM NaCl, 20 mM Tris-HCl, 1 mM GSH, 0.1 mM GSSG, 10 % v/v glycerol, 2 M urea, 0.01 % v/v Triton X100; pH 7.4) at 4 °C for 24 h. After 24 h, the protein was then dialyzed in 250 mL Buffer A2 (200 mM NaCl, 20 mM Tris-HCl, 10 % v/v glycerol, 0.01 % v/v Triton X-100; pH 7.4) at 4 °C for 2 h. The protein was transferred to a new container with 1000 mL Buffer A2. The newly refolded protein was quantified using the Thermo Scientific BCA assay kit. In brief, 200 μL of reagent (1 B: 50 A) was mixed with 10 μL protein in a Sarstedt 96-Well Tissue Culture plate and incubated at 37 °C for 30 min before being read at 562 nm with the BioTek® Synergy 2 plate reader. 3.1.2. Fluorescence Polarization Assay The KRAS probe used for fluorescence polarization analysis was purchased from Thermo Scientific. The sequence was derived from the coding region located between 139 to 185 nts (5’AUGGAGAAACCUGUCUCUUGGAUAUUCUCGACACAGCAGGUCAU-6-FL-3’). Additionally, the probe was fluorescein labeled at the 3’ end. Binding Buffer (50 mM Tris-Cl pH 7.4, 12.5 mM EDTA pH 8, 25 % glycerol, 0.01 % Triton-X and ddH2O) was used in every reaction. 87 Components were added to a Thermo Scientific MicroFluor® 384-Well microplate in the proportion of 8 protein, 4 Binding Buffer, 3 KRAS probe and 4 inhibitor in either a 19 μL or 100 μL reaction. Proteins were diluted to a final concentration of either 300 nM or 60 nM which represents ~80% bound and KRAS was diluted to either 10 or 2 nM depending on the solubility of the compounds that were tested. Compounds were tested from 0 to 100 μM. After addition to the microplate, the reaction was incubated at 37 °C for 30 min before being read on the BioTek Synergy 2 plate reader. Percentage bound was calculated based on the ratio of fluorescence anisotropy (treatment/DMSO control). 3.2 Methodology – Analysis of Spiropyrrolizidine and Piperazine Derivatives on Human Colorectal Cancer Cells 3.2.1. General Information for Cell Preparation The human colorectal cancer cell lines, HT29 and SW480, were maintained in T75 flasks at 37 °C under 5 % CO2. The growth media used was 20 mL of Minimum Essential Medium Eagle (EMEM) from Lonza. In preparation for plating, cells were grown to 80 % confluency. The T75 was washed with 7 mL Dulbecco’s Phosphate Buffered Saline (DPBS). Cells were then trypsinized with 2 mL Trypsin-EDTA (0.25 %) at 37°C for 5 min. Trypsinization was stopped with the addition of 18 mL EMEM (with 10 % Fetal Bovine Serum). A cell sample (20 μL) was stained with Trypan Blue and manually counted with the Bright-Line hemocytometer. Cells were then plated in the 6-well plate at the concentration of 30 x 104 cells per well (SW480) or 60 x 104 cells per well (HT29) and incubated overnight at 37 °C. Following incubation, compounds were added to each well to a final concentration of 10 or 20 μM with 2 % DMSO being the control. Plates were incubated for another 48 h before being harvested for further analysis. 88 3.2.2. Total Protein Harvesting and Western blot Detection of KRAS Following the 48-h compound treatment, each well was washed twice with 1 mL DPBS. To each well, 150 μL of lysis buffer (150 mM NaCl, 50 mM Tris-Cl, 1 % Triton-X 100 and 0.5 % SDS) was added and lysed cells were transferred into 1.5 mL Eppendorf tubes (300 μL per tube). Acetone (4 x volume; 1200 μL) was added to each tube and tubes were transferred to the -20 °C freezer for overnight precipitation of protein. Following precipitation, tubes were centrifuged at 10,000 x g for 10 min at 4 °C. Acetone was removed and the pellets were left to dry for approximately 15 min. Pellets in each tube were then reconstituted with 20 μL of MilliQ H2O. Protein was then quantified using the BCA kit as before. Harvested protein (12 uL, 30 μg per sample) was boiled with 3 uL 5x SDS loading buffer (0.25 % Bromophenol blue, 0.5 M DTT, 50 % glycerol, 10 % SDS and 5 % β-mercaptoethanol). Boiled samples were loaded into a 13.3 % SDS-PAGE and was ran for 50 min at 200 V. The SDSPAGE was then transferred to a GE Healthcare Amersham™ Protran™ 0.45 μm nitrocellulose membrane. The transfer was performed at 4 °C at 100V for 1 h. Following the transfer, the membrane was blocked with 5 % (w/v) skim milk in TBST (0.05 M Tris-Cl, 0.15 M NaCl, 0.1 % Tween-20, pH 7.4) and was left on a waver overnight at 4 °C. The blocked membrane was washed 3 times (10 min each) with TBST and incubated with anti-K-Ras mouse mAb (1:500) from SigmaAldrich for 2 h. The blot was then washed 3 times as before and incubated with goat anti-mouse IGg secondary antibody (1:4000). The blot was washed 3 more times. Visualization was achieved by addition of SuperSignal™ West Fempto Maximum Sensitivity solution (1 mL) from Thermo Scientific. The image was acquired using the ProteinSimple FluorChem Q image and protein bands were quantified using the AlphaView Q software. Luminescence analysis was performed on the house keeping gene (HKG; thioredoxin and GAPDH) and KRAS bands. Background noise was 89 also subtracted from the target bands using the software. A ratio of luminescence for KRAS: HKG was calculated for each treatment including the DMSO control to determine expression values. The expression value for a specific treatment was divided by the expression value for the DMSO control to produce an expression ratio. This process was repeated for the other proteins of interest using the antibodies found in Table 3.2.1. Table 3.2.1. Summary of antibodies used in Western blot analyses. Name Expected Species Dilution Company Band Size Origin (kDa) Thioredoxin 12 Rabbit 1:2000 Abcam KRAS 21 Mouse 1:500 Sigma-Aldrich GAPDH 37 Mouse 1:20000 Sigma-Aldrich Anti-Rabbit HRPGoat 1:2000 Promega linked Anti-Mouse HRPGoat 1:4000 Promega linked Catalogue Number Ab26320 WH0003845M1 G8795 W401B W402B 3.2.3. Total RNA Harvesting and Quantitative PCR of KRAS Steady-state Levels Total RNA was extracted using the Thermo Scientific mirVana™ miRNA isolation kit using the manufacturer’s protocol. Treated cells were harvested after a 48-h treatment window. Cells were washed with 2 mL DPBS and 200 μL of Lysis buffer were added per well. Lysate (~600 μL) was pooled together into a clean 1.5 mL Eppendorf tube and was vortexed gently. To the lysate, 1/10 x volume of Homogenate additive was added. The tube was mixed by inversion and left on ice for 10 min. Afterwards, 1 x volume of Acid-phenol: Chloroform solution was added. The Eppendorf was vortexed for 45 sec and centrifuged for 5 min at 10,000 x g (room temperature). The upper phase was transferred to a new Eppendorf and the volume was recorded. To the upper phase, 1.25 x volume of 100 % ethanol was added. The mixture was loaded onto the filter cartridge and was centrifuged at 10,000 x g for 15 sec. The flow-through was discarded and 700 μL of Wash 90 Solution #1 was added. The cartridge was centrifuged at 10,000 x g for 15 sec and the flow-through was discarded. The cartridge was then centrifuged after addition of 500 μL of Wash Solution #2/3 and this process was repeated once more. Excess solvent was removed with a final centrifugation step (1 min, 10,000 x g) before the sample was eluted with 100 μL of 95°C nuclease-free H2O. RNA was quantified using the Thermo Scientific NanoDrop spectrophotometer. RNA was cleaned using the Ambion DNA: free™ kit using the manufacturer’s protocol. RNA (1 μg, 8 μL) was added to 1 μL of 10 x DNase I buffer and 1 μL DNase I. The sample was incubated at 37 °C for 30 min. After incubation, 2 μL of Inactivating reagent was added and the sample was incubated at room temperature for 2 min. The sample was centrifuged at 10,000 x g for 1.5 min at room temperature. The supernatant was removed and the pellet was discarded. Following clean-up, cDNA was synthesized using BIO-RAD’s iScript™ cDNA synthesis kit. Each sample (1 μg in 10 μL) was added to 4 μL of 5 x Master Mix, 5 μL of nuclease-free H2O and 1 μL of reverse transcriptase in a new Eppendorf tube. The Master Mix contains a universal primer. The cDNA was synthesized using a thermocycler under the following conditions: 25°C for 5 min, 42 °C for 30 min and finally 85 °C for 5 min. Following incubation, samples were prepared for quantitative PCR using Quantabio’s PerfeCTa SYBR Green Fast Mix kit. Each sample was prepared as the following mixtures: 5.3 μL nuclease-free H2O, 7.5 μL SYBR green master mix, 0.6 μL forward primer (10 μM), 0.6 μL reverse primer (10 μM) and 1 μL cDNA. The primer sequences can be found in Table 3.2.2. 91 Table 3.2.2. Primer sequences for qPCR analysis of spiro-treated CRC cells. Name Annealing Forward Sequence Reverse Sequence Temperature (°C) KRAS 58 5’-CGA ATA TGA TCC AAC 5’- ATG TAC TGG TCC CTC AAT AGA G-3’ ATT-3’ GAPDH 60 5’-GTC TTC ACC ACC ATG 5’-AGT TGT CAT GGA TGA GAG AAG-3’ CCT TGG-3’ The samples were added to a 96-well qPCR plate (Axygen, California, USA). Control wells which did not contain primers were also added. Technical triplicates were performed for each reaction. The thermocycler used was the BIO-RAD iCycler and analysis was performed using the BIO-RAD iQ5 software. The thermocycler protocol starts with 3 min at 95 °C followed by 40 cycles of 10 s at 95 °C and 30 s at 58 °C (KRAS) or 60 °C (GAPDH). A thermal melt curve was performed following data collection. The cycle threshold (Ct) values of each technical replicate was recorded and averaged. The averaged Ct for the target gene was subtracted from the Ct of the HKG to produce the ∆Ct value. The ∆Ct was then normalized with the DMSO control to produce the ∆∆Ct value. Lastly, the expression ratio was calculated by normalizing the treatment ∆∆Ct value with the DMSO ∆∆Ct value. 3.2.4. MTT Assay Quantification of Cell Viability The 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay analyses were performed by Dr. Maggie Li. Cells were grown to 80 % confluency in T75 flasks, trypsinized and counted as indicated before. SW480 and HT29 cells were plated in Sarstedt 96-well cell culture plates at the concentration of 1500 cells/well and 2000 cells/well respectively and incubated overnight. Cells were then treated with compounds from 0 to 40 μM concentration (100 μL per well) and incubated for 48 h. Afterwards, 50 μL of MTT dye (1 mg/mL) was added to each well and the plate was incubated for 3 h at 37 °C. Live cells containing active mitochondrial reductase 92 causes reduction of the yellow-colored tetrazolium MTT dye to the purple-colored formazan product (Figure. 3.2.1.). N N S N N Br N Mitochondrial Reductase N 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide MTT S N H N N N (E,Z)-5-(4,5-dimethylthiazol-2-yl)-1,3-diphenylformazan Formazan Figure 3.2.1. Scheme for the reduction of MTT to the formazan form. The supernatant was removed and 100 μL of 100% DMSO was added to each well causing solubilization of formazan crystals. The plate was then read at 570 nm with the BioTek® Synergy 2 plate reader. 3.2.5. Flow Cytometry 3.2.5.1. Apoptosis and Necrosis Detection The BD Pharmingen™ PE Annexin V Apoptosis Detection kit was used for staining of cells for apoptosis and necrosis detection. After treatment of cells at 24 and 44 h, cells were washed with 1 mL DPBS and trypsinized with 200 uL trypsin-EDTA (0.25 %). The digestion was stopped with the addition of 1 mL EMEM (10% FBS). Cells were transferred to a 1.5 mL Eppendorf tube and centrifuged at 400 x g for 5 min (RT). The supernatant was removed and the pellet was washed with 1 mL cold DPBS. Cells were counted manually as before. Cells were then resuspended in Binding Buffer (0.01M Hepes, 0.14M NaCl and 2.5 mM CaCl2; pH 7.4) at the 93 concentration of 1 x 106 cells/mL. Approximately 1 x 105 cells (100 μL) was transferred into a 5 mL culture tube and stained with 5 μL of 7-aminoactinomycin-D (7-AAD) and 5 μL of AnnexinV-PE (Figure. 3.2.2.). Annexin V-PE is a conjugation of the proteins, Annexin V and Phycoerythrin (PE). PE functions as the chromophore (structure not shown). Cells were incubated in the dark for 15 min at room temperature. Binding Buffer (400 μL) was added and cells were filtered before being separated and detected by the BD FACSMelody™ cell sorter. Unstained and single stain controls (7-AAD or Annexin V-PE) were also produced. Single stain for Annexin VPE follows the same protocol. The 7-AAD control has an extra freeze-thaw step. Approximately 100 μL of cells were placed at -80°C for approximately 5 min. The cells were then thawed at 37 °C and stained with 7-AAD using the previously described protocol. Data was processed using the FlowJo software package. O O N Me N NMe O N O O O O HN O NMe O HN H N NMe O O NH O O N NH2 O O O H 2N Figure 3.2.2. Structure of 7-Aminoactinomycin-D. 94 3.2.5.2. Analysis of Cell Cycle Population Cells (5 x 105 cells/mL) in DPBS were acquired from the previous preparation described above and is instead, processed using the BD Cycletest™ Plus DNA Reagent Kit. The cells were pelleted at 400 x g for 5 min. The supernatant was removed and 250 μL of Solution A containing trypsin was added. The sample was incubated for 10 min at room temperature. Solution B (200 μL), the trypsin inhibitor, was then added and the sample was incubated for another 10 min. Cold Solution C (200 μL), which contains propidium iodide (PI), was added (Figure. 3.2.3.). The tube was mixed and incubated in the dark on ice for 10 min. Cells were then detected with the BD FACSMelody™ and data was processed with FlowJo as before. NH2 CH3 H 2N N N I I CH3 CH3 Figure 3.2.3. Structure of propidium iodide. 3.3. Results and Discussion 3.3.1. Fluorescence Enhancers and Quenchers Found in the Fluorescence Polarization Assay The fluorescence polarization assay technique was the first strategy for determining initial activity of synthesized mixtures. These mixtures were compared with the UNBC152-3 (7) with the goal of determining which compound exhibit favorable activity with reducing IMP-1 protein 95 binding with KRAS mRNA. Comparison of UNBC152-3 (7) with the crude batch of VLA9 (7, Section 2.2.2.) was performed first. As SPOPP (8) was hypothesized to be produced in the 1,3dipolar cycloaddition synthesis of UNBC152-3, the crude batches VLA25 (8/9) and VLA21 (8/9) were also included. Additional controls of ʟ-proline (5), 3 (VLA3) and acenaphthenequinone (4) were included to determine if the individual reactants had non-specific interactions with either IMP-1 protein or the KRAS RNA. Protein was set at 300 nM which represented 80 % maximal binding between IMP-1 and 10 nm KRAS RNA. Unsurprisingly, VLA3 (red), acenaphthenequinone (black) and ʟ-proline (violet) did not exhibit the ability to reduce the percent bound, even at 100 μM treatment (Figure. 3.3.1.). VLA9 (yellow) did not decrease the percent bound either. This result was surprising considering VLA9 (7) was suggested to contain all 3 regioisomers of the expected product (7). Post-synthesis of VLA9 (7), the yellow colored solid was filtered and separated from the mother liquor (supernatant). The mother liquor (M. Liq) was dried and tested in the FP analysis. Interestingly, the M. Liq was able to reduce the percent bound of IMP-1 to KRAS mRNA. UNBC152-3 was able to reduce the percent bound to 63 %. Both UNBC152-3 and VLA9 contained presence of the unreacted reagent 3. From the results, it was hypothesized that the portions of the UNBC152-3 crude mixture which could lead to an interaction with either the protein or probe would be the minor products found in the early HPLC retention time points or SPOPP (8) as seen in section 2.2.4.1 (Figure. 2.2.4.1.). 96 Figure 3.3.1. Comparison of crude spiropyrrolizidine and dispiropiperazine compounds on the IMP-1-KRAS mRNA interaction. Protein used was 300 nm KH3to4 with 10 nm KRAS-FL probe. Experiment performed in technical triplicate. VLA21 (8/9 blue) and VLA25 (8/9 purple) were found to reduce the percent bound to 55.3 % and 45.3 % respectively which suggested that components of the mixture were able to interact with either the probe or the protein to reduce FP units. Purified portions of the 2-component reaction (Section 2.2.3) were prepared and analyzed by fluorescence polarization to determine which compound(s) was responsible for the in vitro activity. Additionally, the total fluorescence of the interaction (green bars) was monitored to determine any differences between 2 % DMSO control and treatment wells (Figure. 3.3.2.). UNBC152-3 (7) and VLA9 (7) were included as positive and negative controls, respectively. For faster comparison purposes, compounds were tested at 100 μM instead of in a dose-dependent manner. UNBC152-3 (7) decreased the percent bound to 66.17 % but increased total fluorescence to 141 %. VLA9 (7) did not decrease the percent 97 bound compared to the control but had the total fluorescence increased to 109 %. The change in total fluorescence compared to control can be problematic since FP relies on fluorescence as an output. One of the common criteria for excluding compounds as potential hits is the total fluorescence change.38 In the Shapiro article on IMP-1 inhibitors, candidates that increased total fluorescence by 30 % or more were marked as fluorophore enhancers and compounds which decreased the total by 30 % or more were marked as fluorophore quenchers. Both quenchers and enhancers are typically deemed as false positives and are excluded in subsequent analyses.38 Figure 3.3.2. Comparison of purified SPOPP (8) and NP6A versus crude batches on the IMP-1-KRAS mRNA interaction. 60 nM KH1to4 was used in presence of 2 nM KRAS-FL probe. Compounds were treated at 100 µM concentration. Experiment was performed in technical triplicate. Error bars are standard deviation. 98 The FP data on UNBC152-3 (7) suggested presence of an enhancer due to the 41 % change in fluorescence. An interesting component in these reaction mixtures was NP4 which was a peak found in varying abundance in the HPLC traces of all crude batches. NP4 is the fluorescence compound seen under long wave UV in the TLC analysis of the crude mixtures in section 2.2.6.1. (Figure. 2.2.6.2.). The TLC-purified NP4 increased total fluorescence to 289 % and decreased the percent bound to 33 %; this suggested that one of the major causes of the FP unit decrease for crude batches was due to the presence of NP4 interfering with the assay. VLA159 (8/9) and VLA175 (8/9) were also found to have a greater than 30 % increase in total fluorescence. In contrast, analysis of purified SPOPP (8) and NP6A on the IMP-1-KRAS mRNA interaction revealed that the two dispiropiperazine compounds did not change percent bound or total fluorescence. This data indicated that neither SPOPP (8) nor NP6A truly inhibit the IMP-1-KRAS mRNA interaction. The F1 sample, VLA37-1, did reduce percent bound to 32 % but also decreased total fluorescence to 78 %. Since the decrease in total fluorescence is within the acceptable limits, additional F1 samples were analyzed. As compared to 2 % DMSO, all the F1 samples were able to decrease percent bound by varying amounts (Figure. 3.3.3.). 99 Figure 3.3.3. Effect of semi-purified F1 components on the IMP-1-KRAS mRNA interaction. 300 nM KH1to4 was analyzed in presence of 10 nM KRAS-FL. Compounds were treated at 100 µM. F1 components were purified from crude by FCC with C18 matrix (8 H2O: 2 methanol: 4 acetonitrile: 0.25 acetic acid). Experiment was performed in technical triplicate. Error bars are standard error. The only two samples which were found to be within acceptable fluorescence bounds were VLB165-1 and HX31AB. From the data, it was hypothesized that some side products and/or intermediates within the F1 sample may collectively interact with either the IMP-1 protein or the KRAS RNA probe to decrease binding. Selective F1 samples were tested in colorectal cancer (CRC) cells to determine bioactivity. Since there can be a disparity between in vitro assay activity and bioactivity within a cellular environment, SPOPP (8) and NP6A were also investigated. 100 3.3.2. Effect of Inhibitors on KRAS Expression in CRC Cells 3.3.2.1. Activity in SW480 SW480 CRC cells were first tested with the purified and crude forms of the spiropyrrolizidine derivatives. As a basis for comparison, treatment was kept to the 48 h time frame as previously described by Ms. Chuyi Wang.73 The compounds were initially tested at 10 μM concentration. UNBC152-3 (7) was kept as the positive control and was found to lower KRAS protein expression to 24 % when normalized against Thioredoxin (Figure. 3.3.4). VLA9 (7) was also able to reduce expression to 23 % and was the most potent at the time of testing for this particular trial. Some of the activity may be attributed to presence of 3 in both VLA9 (7) and UNBC152-3(7). The bioactivity of 3 was previously described by Dimmock et al. in the structural activity relationship (SAR) analysis of 2-arylidenebenzocycloalkones derivatives.75 Human Molt 4/C8 and CEM T-lymphocytes cell lines were treated with 3 and the IC50 was found to be 36.8 ± 1.3 and 25.4 ± 5.9 μM respectively. Purified SPOPP (8) and NP6A were also able to lower KRAS expression to 63 % and 26 % respectively. Since VLA159 (8/9) is a crude version of SPOPP (8), it was not surprising to see that it also lowered KRAS expression. The crude material, however, did contain more compounds and thus the combination was hypothesized to cause the increase in activity. As a semi-purified mixture, VLA37-1 lowered KRAS expression to 52 % but the effect was weaker than both VLA159 (8/9) and UNBC152-3 (7). 101 A B 1 0.8 0.6 0.4 SPOPP NP6A UNBC152-3 VLA159 VLA9 NP4 0 VLA37-1 0.2 DMSO Normalized Expression 1.2 Figure 3.3.4. Effect of purified SPOPP (8) and NP6A on KRAS expression in SW480 CRC. (A) Cells were incubated with 10 µM concentration compounds for 48 h. (B) Normalized expression of KRAS against Thioredoxin and relative to DMSO control. N = 1. The FP false positive, NP4, interestingly was able to reduce KRAS expression to 41 % which further complicates the analysis of the UNBC152-3 (7) puzzle. More F1 components were tested alongside NP6A to determine if it was reproducible (Figure. 3.3.5.). Compounds were treated at 20 μM to see a more pronounced reduction of KRAS expression. NP6A was found to lower KRAS expression to 68 % which was over 2-fold weaker than the previous trial. VLB1881 increased KRAS expression to 120 % while VLB189-1 lowered expression to 80 %. Treatment of AcQ (4) at 20 μM was able to reduce KRAS expression to 53 %. The sample was tested as a control due to its role as one of the reactants. A search on scientific literature indicated that 102 acenaphthenequinone has not been tested in SW480 before which made this the first reporting of its bioactivity in this cell line. VLB165-1 increased expression of KRAS to 141 %. UNBC142 was previously reported to have no activity in SW480 and was used as a negative control in this experiment.73 Due to the inconsistencies with the F1 activity, the subsequent Western blot analysis of the semi-crude material in HT29 was abandoned. The focus was also switched to analyzing SPOPP (8) since it was purified to 100 %. A B 1 VLB165-1 UNBC142 UNBC152-3 AcQ NP6A DMSO 0 VLB189-1 0.5 VLB188-1 Normalized Expression 1.5 Figure 3.3.5. Effect of semi-purified F1 samples on KRAS expression in SW480 CRC. (A) Cells were incubated with 20 µM concentration compounds for 48 h. (B) Normalized expression of KRAS against thioredoxin and relative to DMSO control. N = 1. 103 3.3.2.2. Activity on HT29 Compounds were tested on HT29 at the 20 μM concentration. Treatment with SPOPP (8) was found to lower KRAS expression to 72 % (Figure. 3.3.6.). AcQ (4) had less of an effect and reduced expression to 82 %. The activity of AcQ (4) on HT29 has not been previously reported. Treatment with 3 (VLA3) was able to reduce KRAS expression to 51 %; this result was significant due to the compound’s presence in both the VLA9 (7) and UNBC152-3 (7) samples. Although 3 was not able to reduce FP units in the fluorescence polarization assay, it was able to decrease expression of KRAS. The result suggested that 3 may have contributed to the biological activity of UNBC152-3 (7) in SW480 in both the current study and during the initial testing of this library hit by Ms. Chuyi Wang. A B 1 0.8 0.6 0.4 VLA3 AcQ 0 SPOPP 0.2 DMSO Normalized Expression 1.2 Figure 3.3.6. Effect of SPOPP (8) and reagents on KRAS expression in HT29 CRC. (A) Cells were incubated with 20 µM concentration compounds for 48 h. (B) Normalized expression of KRAS against GAPDH and relative to DMSO control. Cells were treated twice more with SPOPP (8) to produce two other biological replicates. SPOPP (8) reduced expression to 65 % which was consistent with the previous finding (Figure. 104 3.3.7.). Similarly, AcQ (4) was still able to reduce expression again to 88 % which was a very consistent reduction. NP6A, however, increased expression to 165 % which was inconsistent with the SW480 results. UNBC168 was tested to see if it can also be used as a negative control but it seemly increased KRAS protein expression. A B Normalized Expression 2 1.5 1 0.5 NP6A UNBC168 AcQ SPOPP DMSO 0 Figure 3.3.7. Effect of SPOPP (8) and NP6A on KRAS expression in HT29 CRC. (A) Cells were incubated with 20 µM concentration compounds for 48 h. (B) Normalized expression of KRAS against GAPDH and relative to DMSO control. A third biological replicate confirmed the bioactivity of SPOPP (8) and AcQ (4); KRAS was found to be reduced to 76 % and 74 % respectively (Figure. 3.3.8.). NP6A increased KRAS to 148 % which was consistent with the prior trial. The western blot data was aggregated and the reduction of KRAS protein in HT29 cells post treatment with SPOPP (8) is statistically significant with a P value of 0.0013 (Figure. 3.3.9.). 105 A B Normalized Expression 1.5 1 0.5 NP6A AcQ SPOPP DMSO 0 Figure 3.3.8. Effect of SPOPP (8) and NP6A on KRAS expression in HT29 CRC. (A) Cells were incubated with 20 µM concentration compounds for 48 h. (B) Normalized expression of KRAS against GAPDH and relative to DMSO control. Figure 3.3.9. Aggregated western blot data for KRAS protein expression in compoundtreated HT29. Cells were incubated with 20 µM concentration compounds for 48 h. Statistical analysis was performed using ANOVA. N = 3. P = 0.0013. 106 Although KRAS can be reduced by treatment with SPOPP (8) and a select few F1 samples, it does not necessarily mean there is a reduction on cellular proliferation. Compounds were analyzed by the MTT assay to validate the impact on proliferation. 3.3.3. Cell Viability Analysis of Compound-treated CRC Cells 3.3.3.1. Proliferation in SW480 The MTT assay is a quick and useful method to determine cell viability. The MTT dye is a tetrazolium salt that is reduced in the mitochondria with Mitochondrial Reductase to produce the insoluble formazan form (Figure. 3.2.1.).81 Dye formation is then proportional to the amount of live cells present. The MTT trials were performed by Dr. Maggie Li. Cells were treated with compounds ranging from 0 to 40 μM concentration and the intensity of the formazan dye was analyzed as compared to 2 % DMSO control. Treatment of SW480 cells with SPOPP (8) led to reduced cell viability with an IC50 of 4.17 μM (Figure. 3.3.10.). 107 SW480 MTT 110 100 % Cell Viability 90 80 IC50 70 60 50 40 30 1 ACQ 4.78 uM SPOPP VLA-3 HX152 UNBC168 4.17uM N/A N/A 2 N/A 3 4 5 Log Dose (nM) Figure 3.3.10. MTT assay analysis on SPOPP-treated SW480. Cells were treated using compounds ranging from 0 to 40 µM concentration. Cells were incubated for 48 h. Error bars are standard deviation. Experiment was performed in technical triplicate. N = 1. AcQ (4) also had a low IC50 of 4.78 μM. The results clearly showed that both SPOPP (8) and AcQ (4) are anti-proliferative to SW480 cells. Paradoxically, 3 did not reduce viability to the same level despite being able to reduce KRAS protein expression levels by 49 % (Figure. 3.3.6., Figure. 3.3.10.). This result suggested that KRAS expression is not necessarily correlated with cell proliferation. The results show at least one instance where a reduction of KRAS might not lead to a reduction of cell viability. HX152 (7) was the sample prepared by purification of the main peak in the crude UNBC152 (7) synthesized by Ms. Hooi Xian Lee (Section 2.2.2.). The peak was found at 6.9 min in the phenyl-hexyl analysis as previously described in Chapter 2. Interestingly, HX152 (7) did not appear to have any significant effect on SW480 cell viability; therefore, one isomer of 7 appears to be non-bioactive. F1 fractions isolated from crude batches of 8 were also analyzed. The rationale was to see if any side-products were able to collectively affect viability. Treatment 108 of the cells with VLB165-1, 159-1, 188-1 and 189-1 was able to lower viability with IC50 values close to AcQ (4) and SPOPP (8, Figure. 3.3.11.). SW480 MTT 110 100 % Cell Viability 90 80 IC50 70 ACQ 4.78 uM 165-1 5.03uM 159-1 6.29 uM 60 188-1 5.19 uM 189-1 4.06 uM 50 40 30 1 2 3 4 5 Log Dose (nM) Figure 3.3.11. MTT assay analysis on F1 -treated SW480. Cells were treated with compounds ranging from 0 to 40 µM concentration. Cells were incubated for 48 h. Error bars are standard deviation. Experiment was performed in technical triplicate. N = 1. It is possible that multiple components of the F1 fractions were able to interact with IMP1 collectively to prevent KRAS mRNA association. The unbound KRAS mRNA would then be free to interact with endonucleases or miRNA leading to reduced KRAS protein and the subsequent decrease in proliferation. Alternatively, the components of F1 may interact with multiple effectors associated with growth pathways thus leading to proliferation decrease. Either scenario is possible but would still need to be validated. To test the IMP-1-KRAS mRNA interaction decrease, immuno-precipitation of endogenous CRC IMP-1 protein will have to be performed on F1-treated cells. From the analysis, there would be a decrease in KRAS mRNA that is associated with the precipitated protein as compared to DMSO control if the case is true. 109 3.3.3.2. Proliferation in HT29 Semi-purified F1 (188-1.1) derived from VLB188-1 was tested in HT29 to determine if purification was able to increase potency (Figure. 3.3.12.). Figure 3.3.12. MTT assay analysis of SPOPP (8) and semi-purified F1 on HT29. Cells were treated with compounds ranging from 0 to 40 µM concentration. Cells were incubated for 48 h. Error bars are standard deviation. Experiment was performed in technical triplicate. N = 1. After fractionation, the 188-1.1 did not display marked improvement versus the parent F1. Initially, cell viability decreased with a maximum treatment of 2.5 μM as compared to F1. Concentrations of 188-1.1 from 5 to 40 μM shared a similar decrease in cell viability. The most potent compound tested was SPOPP (8), which had an IC50 of 6.76 μM; the potency was similar to that observed with SW480. Since SPOPP (8) was the most consistently bioactive compound, further testing was conducted and the pursuit of F1 was set aside. Several other compounds have similar activity to SPOPP (8, Figure. 3.3.13.). AcQ (4), again, was effective on HT29 and had an IC50 of 8.51 μM. UNBC168, which was used as a negative control since it was another 1,3-dipolar 110 cycloaddition product utilizing 4 and 5 as precursors but was inactive in the FP assay, did not decrease viability. Figure 3.3.13. MTT assay analysis of SPOPP (8) and NP6A on HT29. Cells were treated with compounds ranging from 0 to 40 µM concentration. Cells were incubated for 48 h. Error bars are standard deviation. Experiment was performed in technical triplicate. N = 1. It has been previously reported that resveratrol, at a concentration of 60 μM, decreases KRAS expression and cellular proliferation in HCT116 and SW480 CRC cell lines.82 In light of this information, resveratrol was used as a positive control in these experiments. Treatment with 40 μM resveratrol decreased HT29 cell viability to approximately 60 %. Interestingly, crude VLA9 (7) was able to reduce proliferation the most with an IC50 of 4.47 μM. VLA9 (7) contains both 3 and 7. Since, by part, 3 and 6 did not have activity, it was hypothesized that the combination of all the reactants and products in VLA9 (7) led to the proliferation decrease seen in HT29. NP6A, on the other hand, only had modest activity in HT29. 111 3.3.4. Effect of SPOPP (8) on KRAS mRNA Levels in HT29 Cells Steady-state levels of KRAS mRNA was analyzed in SPOPP-treated HT29 cells as an initial step to determine the mechanism whereby SPOPP (8) decreases cell proliferation. A total of four biological replicates were conducted and cells were treated with 20 μM of each compound. The results pooled from four biological replicates showed that treatment with SPOPP (8), AcQ (4) and NP6A did not significantly change the steady-state levels of KRAS mRNA in HT29 cells (Figure. 3.3.14.). Figure 3.3.14. Analysis of KRAS mRNA steady-state levels in SPOPP-treated HT29 by quantitative PCR. Results were normalized to GAPDH. Biological replicates where N = 4. P = 0.5667. Statistical analysis by ANOVA. Error bars are standard error. Since the KRAS mRNA levels did not change while suppression of KRAS protein was observed, I hypothesize that SPOPP (8) acts at the level of translational repression on KRAS. There is a possibility that SPOPP (8) can increase miRNA leading to translational repression of 112 KRAS. In the Saud et al. paper on resveratrol and CRC, tumors resected from resveratrol-treated APCCKO/KRASmut mice contained a 67% increase of microRNA 96 (miR-96).82 The miR-96 was previously found to directly target the 3ʹ-UTR of the KRAS transcript in pancreatic cancer cells.83 Microarray or RNA-seq analysis of SPOPP-treated HT29 and SW480 would have to be performed to determine global change for miRNA. Alternatively, qPCR can be performed as an inexpensive method if a subset of miRNAs, such as miR-96, were to be tested for. 3.3.5. Investigating the Anti-proliferative Mechanism of SPOPP (8) in SW480 3.3.5.1. Necrosis versus Apoptosis Flow cytometry was used as an analytical tool to determine the differences between necrotic and apoptotic cell populations. The same technique was used to determine differences in the cell cycle. In brief, the machine can sort a cell population into single cell droplets using a buffered sheath fluid as the vehicle. The droplets pass by a laser which excites fluorophore-labeled cells and the subsequent emission is detected. The machine also detects the degree of forward (FSC) or side light scatter (SSC) as a low-resolution determinant of cell size and shape. The commonly used stains to detect necrosis and apoptosis are 7-AAD and Annexin V-PE respectively. In the case of necrosis, the cell membrane becomes compromised and is permeable to large dyes. The 7-AAD can then enter the necrotic cell and intercalate with DNA.84 Annexin V-PE is used to stain for apoptotic cells.85 Normally phosphatidylserine (PS) is sequestered in the inner membrane of cells. During the early stages of apoptosis, PS flips to the outer membrane to signal phagocytes. Annexin V was proven to have high affinity to PS and thus if conjugated to a fluorophore, can act as a quick method of determining an apoptotic cell. To determine the difference between necrosis and apoptosis, cells were dual-stained with both 7-AAD and Annexin V-PE (Figure. 3.3.15.). 113 A DMSO 24 h B SPOPP 24 h C DMSO 44 h D SPOPP 44 h Figure 3.3.15. Distinguishing mechanism of cellular death on SPOPP-treated SW480 cells using flow cytometry. (A) 2% DMSO, (B) 20 µM SPOPP, (C) 2% DMSO, (D) 20 µM SPOPP. 7-Aminoactinomycin D (7-AAD) detects necrotic cells with permeated plasma membrane. Annexin V-PE (PE) detects apoptotic cells through binding with phosphatidylserine. Necrotic cells (Q1), Late Apoptosis (Q2), Early Apoptosis (Q3) and Live cells (Q4). The results were then plotted in a 2-dimensional histogram in order to visualize population differences as indicated by fluorescence intensity. As a control, 2 % DMSO-treated SW480 cells 114 were dual-stained and measured at 24 h. Most of the cellular population (91.4 %) were seen in quadrant 4 (Q4) which was set as the base-line level of fluorescence for non-specific interactions with 7-AAD and Annexin V-PE (Figure. 3.3.15A.). Necrotic cells (0.89 %) were found in quadrant 1 (Q1) as indicated by higher levels of 7-AAD staining. Cells undergoing early apoptosis (2.72 %) as indicated by higher levels of Annexin V-PE were found in quadrant 3 (Q3). Quadrant 2 (Q2) was the region which indicated high levels of both stains and contained 5.08 % of the total cell population. SPOPP (8) treatment of cells at 24 h caused the cell population to move towards a more necrotic state (Figure. 3.3.3.15B.). Q4 contained 60.9 % of the cell population which was a reduction of 30.5 % as compared to the control. Q1 contained 28.5 % of the population which was a drastic increase. The results indicated that the cellular population was moving towards a necrotic state and will ultimately end in cellular death. Activation of necrosis was further reflected by the shift in position of highest cellular density (red color) when compared to the control. Early apoptosis was not seen since Q3 only had 2.34 % of the population which was comparable to control (2.72 %). Q2 contained 10.8% of the cell population which suggested an increase in either late apoptosis or dead cells. Necrotic cells were also present for the 44 h SPOPP (8) treatment (Figure. 3.3.15D.). Q1 contained 16.9 % of total cell population which was a decrease of 11.9 % as compared to the 24 h treatment. The decrease indicated that the peak activity of SPOPP (8) for inducing necrosis in SW480 was around 24 h. 3.3.5.2. Cell Cycle Distribution Differences in cell cycle stages for a population of cells can also be measured by flow cytometry. The common dye used to stain cells is propidium iodide (PI), which serves to stoichiometrically bind DNA through intercalation.86 In the procedure, trypsin was used to cause limited digestion of the plasma membrane leaving the nuclei intact. Ribonuclease A was used to 115 remove any RNA to prevent false positives. The sample was then stained with PI, filtered to remove aggregates and analyzed using the flow cytometer. Cells are normally in the G1 phase of the cell cycle and has a ploidy of 2N. A switch to the synthesis phase (S) would cause the production of new DNA in preparation of cellular division. At the end of S phase and the beginning of the G2/M checkpoint, the cells would have twice the amount of DNA (4N ploidy). Twice the DNA would lead to twice the fluorescence from PI. Cells were monitored at 18, 24 and 44 h and 2 % DMSO was used as the control. Histograms were plotted as fluorescence intensity of PI vs cell counts. The approximation of cell cycle percentages was performed using the Watson model and was displayed as the violet-colored trace (Figure. 3.3.16.). Treatment of cells with 20 μM SPOPP (8) for 18 h led to an increase of the G2/M peak (green) to 15 % when compared to the control which was pre-dominantly at G1 as shown by the purple peak (Figure. 3.3.16A., Figure. 3.3.16D.). 116 A 18 h B 24 h C 44 h D 18 h E 24 h F 44 h Figure 3.3.16. Cell cycle analysis of SPOPP-treated SW480 cells using flow cytometry. (AC) 2 % DMSO control. (D-F) 20 µM SPOPP treatment. Cells were stained with propidium iodide (PI) which intercalates DNA. Cell cycle approximation through the Watson model is outlined in violet. The color coding is as follows: blue for G1 phase, yellow for S phase and green for G2/M phase. At the 24 h mark, treatment with SPOPP caused a further shift in cell population at the G2/M phase to 38 % (Figure. 3.3.16E.). The activity was seen to decrease at 44 h since the G2/M peak was reduced to 7.8 % (Figure. 3.3.16F.). The result indicated the peak activity of SPOPP was around 24 h which was congruent with the apoptosis analysis. Unfortunately, the analysis does not differentiate between cells at the G2 check point or cells in the Mitosis (M) phase. 117 Both would have a ploidy of 4N. From the results, SPOPP is proposed to be a G2/M inhibitor which leads to necrosis of cells. The importance of G2/M inhibitors will be further discussed. Table 3.3.1. Summary of cell cycle population percentages for SPOPP-treated SW480. Treatment Time G1 Phase S Phase G2/M Phase (h) (%) (%) (%) 2 % DMSO 18 80.0 17.6 3.1 24 126.0 6.1 0.8 44 93.6 8.3 0.5 20 µM SPOPP 18 66.0 16.8 15.0 24 55.0 10.5 38.1 44 81.1 11.3 7.8 3.3.6. Summary The search for an inhibitor of the IMP-1-KRAS mRNA interaction was not successful. Purified compounds (8 and NP6A) were not able to impact the interaction. Components of F1 were able to lower IMP-1 percent bound in the FP assay but further purifications resulted in decreased potency as determined by MTT assay; thus, the pursuit of further purifications of F1 was abandoned. From FP and WB analyses, it was suggested that the previous bioactivity of UNBC152-3 (7) was due to a mixture of fluorescent false positives and the presence of previously described bioactive compounds (3 & 4). However, this is the first reporting on the bioactivity of 3 and 4 for SW480 and HT29 human colorectal cancer cells. On a positive note, 8 was found to be a potent compound for decreasing cellular proliferation in SW480 and HT29. Since 8 did not impact the steady-state levels of KRAS mRNA for HT29, it was hypothesized that 8 may upregulate miRNA(s) to produce translational repression of KRAS protein. Additionally, 8 was able to arrest SW480 cells in the G2/M phase of the cell cycle and was found to reduce cellular proliferation through necrosis. The importance of G2/M inhibition and necrosis is further discussed in Chapter 4. 118 Chapter 4 General Discussion 4.1 Project Overview IMP-1 is part of the VICKZ family of RNA binding proteins which serves in spatiotemporal regulation of mRNA during embryogenesis and is absent in adulthood.21 Reappearance of IMP-1 in colorectal cancer was detected by Ross et al. in 2001, thus earning its status as an important oncofetal protein worth interrogating.32 Furthermore, IMP-1 was found to bind several important oncogenes such as MDR-1, CD44, βTrcp-1, c-Myc and KRAS.23,26,27,87 Binding of the oncogenic transcripts further exacerbates the cancer condition due to IMP-1’s role in shielding mRNA targets from degradation by endoribonucleases.23 Of particular note is the KRAS signaling protein. Activation of KRAS through GTP binding causes signal transduction of the canonical growth pathways of MAPK and PI3K leading to growth and differentiation.7 The issue lies in the scenario of KRAS driven cancers such as colorectal cancer.50 KRAS is often mutated at the 12th amino acid position which changes the glycine to valine.52 The mutation causes KRAS to be switched on permanently leading to constitutive signaling of growth pathways and a poorer clinical outcome for patients.50 Mongroo et al. found that knockdown of IMP-1 by siRNA led to a 60 % decrease in KRAS protein levels and subsequent decrease in cell proliferation in colorectal cancer (CRC) cells.27 The study also showed that IMP-1 has high affinity to KRAS mRNA. Such important findings led to the hypothesis that inhibition of IMP-1 could lead to its dissociation from KRAS mRNA. The naked KRAS mRNA is then subjected to degradation by endoribonuclease or be translationally repressed by miRNA. A previous graduate student in Dr. Lee’s lab, Ms. Chuyi Wang, screened a small library of 217 spiropyrrolizidine derivatives using 119 the fluorescence polarization (FP) method and found 2 candidates which inhibited the IMP-1KRAS mRNA interaction: UNBC143 and UNBC152 (7).73 UNBC152 (7) was found to be the best hit from the assay. Western blot (WB) analysis of UNBC152-treated SW480 colorectal cancer cells showed suppression of KRAS protein level. However, the second batch of UNBC152 (7) made by our collaborator did not show any bioactivity. Subsequently, a third batch (UNBC152-3, 7) was synthesized which produced the most activity according to FP analysis. To determine the cause of the discrepancy between the three batches, a LC-MS scan was performed for purity analysis. The results showed that the three batches were not pure and may contain isomers of the predicted structure. The goal of this investigation was to synthesize the UNBC152 (7) molecule(s) in-house according to the collaborator’s protocol and to determine the molecule responsible for: (i) inhibition of IMP-1-KRAS RNA interaction as determined but FP analysis, and (ii) suppression of KRAS expression as determined by Western blot analysis. The focus of this investigation was on the purification of the compounds and assessing their potential biological activity. This study was not meant to focus on the organic synthesis of UNBC152 (7) and its related compounds. 4.2 Synthesis of Spiropyrrolidine Derivatives and the Focus on SPOPP (8) Synthesis of the starting reagent, (2E)-2-(2-bromobenzylidene)-1-indanone (3) was a necessary step to produce the dipolarophile required for the 1,3-dipolar cycloaddition product of UNBC152-3 (7). The cycloaddition began after in situ generation of the azomethine ylide (6) produced by the decarboxylative reaction between acenaphthenequinone (4) and ʟ-proline (5). The azomethine ylide (6) formed the 1,3-dipole and fused with 3 in a concerted motion to form the proposed cycloproduct in UNBC152 (7). The crude batch was named VLA9 (7). However, as reported by Haddad et al. in 2015, two ylide intermediates can cyclize together to form the fused piperazines 8 and 9 as side-products of the main 1,3-dipolar cycloaddition reaction.77 Originally, 120 8/9 (VLA25) was synthesized as a control to test side-by-side with UNBC152-3 (7). At the initial stages of the project, HPLC-MS was not available, and I had no access to higher detailed analyses. At that time, the only information used was the TLC analysis of both UNBC152-3 (7) and the crude batch of 8/9 (VLA25). TLC comparison between crude UNBC152-3 (7) and VLA25 (8/9) revealed a similar tailing pattern on normal phase TLC from 0 to 0.27 Rf (Figure. 2.2.3.3.). VLA9 (7) did not have this tailing pattern but contained 3 other spots in common with UNBC152-3 (7). Given this information, it was hypothesized that compound(s) that exhibited bioactivity on FP and WB analyses were present within the tailing portion of the TLC. Due to the aforementioned information and the simpler reaction in terms of reagent usage, the focus was then shifted to the synthesis and purification of 8 (SPOPP). In summary, prior to the installation of the HPLC-MS, the preliminary work was primarily guided by results from TLC analysis. 4.3 HPLC-MS Analyses of SPOPP (8) and UNBC152-3 (7) With installation of the HPLC-MS at UNBC and hence its usage in this investigation, more valuable information was gained from the various synthesized batches of both 7 and 8. UNBC1523 (7) contained 3 major peaks and 3 minor peaks whereas the batched synthesized in house (7, VLA9) contained 3 major peaks and 2 minor peaks (Figure. 2.2.4.1.). The mass spectrophotometer used functioned by the electrospray ionization principle. Ionizable functional groups were protonated using 0.1 % formic acid and was detected on the MS as having a mass one Dalton larger than original due to the additional proton. Following MS analysis, it was found that major peaks from 7.1 min and onwards corresponded to the predicted synthesized product (534 Da) as seen by the m/z ratio of 535 (Figure. 2.2.4.2.). Due to the physical delay between the variable wavelength detector (VWD) and the MS, there was an approximate 0.2 min retention time difference between the peak seen on VWD and the mass peak seen on MS. UNBC152-3 (7) contained two peaks 121 related to the cycloaddition product and one peak related to the starting reagent 3 (Figure. 2.2.4.1.). The VLA9 (7) sample contained 3 products (6.6, 7.6 and 9.1 min) and the presence of 3. The potential for the creation of three spiropyrrolizidine isomers in the 1,3-dipolar cycloaddition was previously reported by Haddad et al.77 The azomethine ylide formation between 4 and 5 can interconvert between the W-shape and the S-shape during the in situ reaction (Figure. 2.2.3.1). The dipolaraphile 3 can then interact with either the W-shape or the S-shaped form to produce 2 different regioisomers. The minimum reaction time to produce the heterocycle was reported to be 0.5 h, however, extension of the reaction time towards 24 h resulted in the creation of a third regioisomer. Creation of the third regioisomer was hypothesized to be due to a retro-1,3-dipolar cycloaddition and subsequent cycloreversion.77 Since VLA9 (7) contained three peaks of the 534 mass, it was hypothesized that main two regioisomers were formed through the main reaction. The third product would then be formed by the cycloreversion mechanism. Conversely the UNBC1523 (7) only contained two of the peaks related to the product and it was hypothesized that the cycloreversion did not occur. Bioactivity comparison on FP between UNBC152-3 (7) and VLA9 (7) was prioritized first. Structural elucidation of the three isomers was postponed. The rationale for this choice was to favor the search for the biologically active molecule using the FP screen. The goal was to determine the sole inhibitor of the IMP-1-KRAS mRNA interaction within the UNBC152-3 (7) mixture. If, as a collective, VLA9 (7) did not have activity based on FP, then there would not be any need to characterize the three regioisomers. This was especially true since the compounds were tested at concentrations of 100 μM. At that concentration, even minor peaks within the mixture would be abundant enough for testing purposes. If any component within the mixture was bioactive, it should then interact with IMP-1. The UNBC152-3 (7) mixture did contain 122 other early retention peaks from 0 to 3.7 min in the HPLC analysis (Figure. 2.2.4.1.). The peaks were hypothesized to be other side-products or intermediates of the reaction. The crude batch of SPOPP (VLA37, 8/9) was analyzed by HPLC using the same column and solvent system as the UNBC152-3 (7) analysis to search for similarly retained peaks. VLA37 (8/9) also contained early retention time peaks from 0 to 3.8 min (Figure. 2.2.4.5A.). Of particular note was the peak at 3.7 min that was shared between UNBC152-3 (7) and the crude batch of SPOPP (8). The peak was named NP4 due to the region of the TLC where it was found on. The compound was a rather distinct yellow-looking spot when viewed using a UV transilluminator at 320 nm (Figure. 2.2.6.1.). VLA37 (8/9) had four peaks located from 7.6 to 10.2 min on the VWD trace (Figure. 2.2.4.5.). All four peaks ionized to a mass of 471 m/z after protonation with 0.1% formic acid. In their paper, Haddad et al. only reported the generation of SPOPP (8) and 9 from their reaction sequence.77 Formation of the other two regioisomers, 8ʹ and 9ʹ, was deemed to be unfavorable due to steric factors between the naphthalene ring of one ylide and the pyrrolidine ring of the other ylide (Figure. 2.2.3.1.). Since the four peaks are separated by a minimum of 0.7 min by the phenyl-hexyl column as visualized by the VWD chromatogram, the retention time is far apart enough to conclude that all four peaks were different entities from each other (Figure. 2.2.4.5.). The results indicated that all four possible isomers (SPOPP, 8ʹ, 9 and 9ʹ) were generated in the VLA37 (8/9) batch. It was also found that increasing the temperature in the synthesis of SPOPP (8) from 35 °C to 55 °C favored the production of the peak at 7.3 min and reduced the relative levels of side-products (0 to 3.7 min) (Figure. 2.2.4.7.). The rationale for this experiment was to determine the optimal temperature to produce SPOPP (8) which was approximately 55 °C. To date, there have been no other reports of the time and temperature dependencies in the creation of the fused piperazines 8 & 9. 123 4.4 Purification and Identification of the Fused Piperazines, SPOPP (8) and NP6A As a comparison with UNBC152-3 (7) and VLA9 (7), SPOPP (8), as well as several other components of the reaction mixture were purified. The initial step was to purify the two fused piperazines as reported by Haddad et al. for use as controls.77 In their article, SPOPP (8) was reported as an orange-colored compound, and 9 as a yellow-colored compound. VLB133 (8/9) was used as the purification material due to the higher amount of product peaks relative to side-products (Figure. 2.2.4.7.) observed through HPLC analysis. Indeed, an orange and a yellow-colored band did appear from elution on a column packed with normal phase matrix. The yellow-colored band was named NP6A. It was hypothesized that NP6A was 9 but NMR analysis was not conducted at the time to prove the case. Unfortunately, the yield of both SPOPP (8) and NP6A was low. The yield of both compounds was further reduced upon HPLC purification. The reported benefits of cycloadditions are mild reaction conditions, ease of synthesis and highly functionalized isomers.77 However, the benefits were balanced out by the difficulty of purification and the low yields. Due to the low abundance of the two isomers that were similar to SPOPP (8) and NP6A as indicated by the peaks retained from 7 to 10 min on VWD chromatogram, they were not pursued in terms of purification (Figure. 2.2.4.7.). Further enhancement of yields for 8 and 9 might be aided by microwave-assisted synthesis as reported by Bashiardes et al. in their analysis of synthesis conditions for the [3+2] cycloaddition of azomethine ylides.88 The microwave increased yields of the major products for most cases by greater than 5 %. SPOPP was confirmed to be 8 by comparing 1 H-NMR spectra obtained to what was reported in literature.77 NP6A, however, did not match the spectra of 9 in the literature due to the shift in the peak from 4.0 ppm to 4.88 ppm and the more over-lapping peaks in the aromatic region (Figure. 2.2.7.7., Figure. 2.2.7.8.). The aliphatic region of the 13C-NMR was also missing a peak compared to 9 (Figure. 2.2.7.9., Figure. 2.2.7.10.). 124 According to the spectra, NP6A was more symmetrical than what Haddad et al. reported for 8 and 9. Higher resolution spectra of NP6A will have to be performed in order to elucidate the structure of NP6A. X-ray diffraction may also be required since Haddad et al. required the additional technique to structurally elucidate the various regioisomers reported in the paper.77 4.5 Finding Inhibitors of the IMP-1-KRAS Interaction by Fluorescence Polarization Assay At this time, an IMP-1-KRAS mRNA inhibitor does not exist in the literature. Ms. Chuyi Wang performed a preliminary screen of 217 spiropyrrolizidine derivatives on the interaction but a vast majority of the compounds were reported to be auto-fluorescent.73 NP4 was one of the autofluorescent compounds discovered during the analysis of UNBC152-3 (7) and SPOPP (8). Although it decreased the percent bound between IMP-1 and the KRAS-FL probe to 33 %, it enhanced total fluorescence to 289 % thus suggesting it as a likely false positive (Figure. 3.3.2.). NP4 was hypothesized to be present in several other compounds of the library since this particular series required the same azomethine ylide for the synthesis. UNBC152-3 (7) was proven to contain the NP4 compound; thus, its activity as a mixture was overestimated using this technique. To add further to this complexity, the F1 components of crude SPOPP (8) were also able to reduce FP units (Figure. 3.3.3.). UNBC152-3 (7) and SPOPP (8) had similar early retention time peaks in the HPLC trace. Since both compounds required 4 and 5 as starting reagents, it is hypothesized that most of the side-products were the same. Therefore, the original FP activity of UNBC152-3 (7) was attributed to the presence of the side-products rather than the main products. The conclusion was further supported by the fact that VLA9 (7), which contained all three regioisomers of UNBC152-3 (7) and the starting reagent 3 but did not affect the IMP-1-KRAS mRNA interaction (Figure. 3.3.2.). Both the purified SPOPP (8) and NP6A did not reduce the binding of IMP-1 to KRAS-FL. From the results, there was no singular molecular species found which was able to 125 affect IMP-1 binding. Bioactivity from pure compounds derived in this study was then attributed to a mechanism other than inhibition of the IMP-1-KRAS mRNA binding. The development of small molecular inhibitors of IMP-1 is a new field. Thus far, only Shapiro has discovered a molecule which can target the IMP-1-c-Myc mRNA interaction.38 However, their reporting did not include supplementary information describing the purity of the compound. Purification of the lead compound remains an integral step of the process. As found in the purification and analysis of SPOPP (8), the side products alone (F1) lead to favorable activity in the FP assay. Evidence of purity in the form of 1H-NMR, mass spectrophotometry and HPLC chromatogram should remain as part of the proof; this would give confidence to researchers that the activity was due to a singular entity and not by a collection of compounds/side-products. Finding inhibitors of IMP-1, in general, is a monumental task. IMP-1 can recognize many mRNA substrates as shown previously for the binding assays of βTrcp1, CD44, c-Myc, MDR1 and KRAS.20,23,25,26 The lack of a consensus sequence is both problematic and beneficial. From those studies, it is difficult to ascertain the true binding surface of IMP-1 that is essential for recognizing a specific RNA. Concurrently, the lack of a consensus sequence lends credence to the idea of a substrate-specific IMP-1 inhibitor. This is further supported by the fact that IMP-1 is suggested to recognize both sequence and secondary structure of RNA.20 The search for an IMP1 inhibitor can be continued for this particular study by soaking an IMP-1 crystal with the F1 components. Chao et al. were able to produce a crystal of the truncated form of IMP-1 which only contained the KH3 and KH4 domains.19 Soaking the IMP-1 crystal with F1 and usage of X-ray diffraction could determine which molecular components were able to associate with IMP-1. 126 4.6 Reduction of KRAS Expression in CRC The goal of finding an inhibitor for the IMP-1-KRAS mRNA interaction was not successful. However, analysis of SW480 and HT29 human colorectal cancer cells after treatment with the spiropyrrolizidine and piperazine derivatives still resulted in reduction of KRAS protein expression. UNBC152-3 (7) and VLA9 (7) both reduced KRAS expression in SW480 (Figure. 3.3.4.). The rationale for the biological activity was suggested to be due to the presence of reactant 3 and 4 in both samples. Interestingly, even though 3 can reduce KRAS expression, it did not affect cellular proliferation as indicated by the MTT assay (Figure. 3.3.6., Figure. 3.3.10.). Reactant 4 was able to reduce both KRAS protein expression and cellular proliferation in SW480 (Figure. 3.3.5., Figure. 3.3.10.). From the results, it was hypothesized that the biological activity of UNBC152-3 (7), as previously found by Ms. Chuyi Wang, could be overreported due to the presence of 3, 4 and perhaps trace levels of 8.73 The conclusion was further supported by the fact that HX152 (7), which is the dominant product purified from the 1,3-dipolar cycloaddition reaction of UNBC152-3 (7), did not affect cellular proliferation of SW480 (Figure. 3.3.10.). The sideproducts of UNBC152-3 (7) were also hypothesized to contribute to the overall biological activity of crude 8, since F1 was able to reduce both KRAS expression and cellular proliferation (Figure. 3.3.5., Figure. 3.3.11.). The only remaining component of UNBC152-3 (7) that was not analyzed was a pure form of the secondary regioisomer found within the mixture. Focus was switched to characterizing the bioactivity of SPOPP (8). At the time, SPOPP (8) was a purified molecule and it was unlikely that the low-abundance secondary regioisomer was able to impact the IMP-1KRAS mRNA binding. SPOPP (8) was consistently able to reduce KRAS protein expression levels in both SW480 and HT29. Cellular proliferation was also confirmed to decrease upon SPOPP (8) treatment as indicated by the MTT assay. Since SPOPP (8) does not interrupt the IMP-1-KRAS 127 mRNA interaction as indicated by the FP results and does not affect KRAS mRNA steady-state levels, a different mechanism would be occurring (Figure. 3.3.2., Figure. 3.3.14.). It is hypothesized that treatment with SPOPP (8) increased a miRNA. The miRNA would then bind to the 3’ UTR of the KRAS transcript to produce translational inhibition of KRAS protein. There are miRNA candidates that were previously found to target the 3’UTR of KRAS such as miR-143 and miR-96.55,82 The mechanism for the repression is due to the miRNA binding to the 3’UTR regions where regulatory proteins normally interact.89 Validation on the increase of a specific miRNA can be performed using qPCR. However, if the miRNA is novel, RNAseq would have to be performed to determine global RNA level changes upon SPOPP (8) treatment.90 Validation of the novel miRNA can be performed through the use of a luciferase reporter assay.55 4.7 SPOPP (8) Induces Cellular Death through Necrosis As indicated by the MTT assay, treatment of cells with SPOPP (8) reduced cellular proliferation (Figure. 3.3.10.). The IC50 was found to be 4.17 μM. To characterize the mode of reduced proliferation, flow cytometry was used to discern between apoptosis and necrosis. Dual staining with 7-AAD and Annexin V-PE determined which cells were undergoing necrosis and apoptosis respectively. During necrosis, the cell membrane is permeabilized and the 7-AAD can enter the cells to then intercalate with DNA.84 Annexin V-PE detects apoptosis through the binding of cell-surface exposed phosphatidylserine which is normally sequestered in the inner membrane for normal cells.85 Treatment of cells with SPOPP (8) for 24 h revealed a ~27 % increase in necrotic cells as indicated in Q1 by the increase in 7-AAD fluorescence as compared to the 2 % DMSO control (Figure. 3.3.15B.). The necrosis pathway was further supported by the upward movement of the cell population from Q4 to Q1 as compared to control and the lack of Annexin V-PE staining in Q3. Furthermore, maximal activity of SPOPP (8) was indicated to occur around 128 24 h which was evidenced by the 11.6 % reduction of cell population in Q1 as compared to 24 h (Figure. 3.3.15.). Although necrosis was classically known as an uncontrolled cellular process, Chan et al. found the process can be regulated.91 SPOPP (8) may act through the process of programmed necrosis (necroptosis). Necroptosis is induced through the activation of the protein serine/threonine kinase receptor interacting protein 1 (RIP1) by receptor interacting protein 3 (RIP3).92 RIP1 and RIP3 forms a complex (complex II) with Fas-associated protein with death domain (FADD) and mixed lineage kinase domain-like protein (MLKL).93,94 MLKL is then recruited to the plasma membrane due to the positively charged residues in the four-helical bundle domain (4HBD) of the N-terminal domain region.95 Oligomerization of MLKL occurs through 4HBD interactions leading to pore creation and permeabilization of the plasma membrane. The loss in osmotic pressure and cellular contents causes cell death, thus marking the end of the necroptosis pathway. Necrosis is a pro-inflammatory process due to its release of High mobility group 1 (HMGB1) protein.96 HMGB1 release into the tumor microenvironment causes recruitment and maturation of dendritic cells (DC).97 As part of the immune response, DC causes activation of CD8+ T cell lymphocytes which leads to anti-tumor immunity. Although treatment of CRC with SPOPP causes necrosis, the resultant activation of DC and recruitment of T cells would be highly beneficial. CRC have been known to evade immune surveillance with disease progression as marked by the decrease in T cells.98 However, proving the relationship between SPOPP-treated CRC and activation of CD8+ T cells would require an in vivo mouse model of CRC. Confirmation of necroptosis instead of uncontrolled necrosis can be found through dual treatment of CRC with SPOPP (8) and necrosulfonamide.94 Necrosulfonamide was found to inhibit MLKL and prevent necrosis. Alternatively, necrostatin1 could be used to inhibit RIP1 which would also result in the prevention of necrosis.92 129 4.8 G2/M Inhibition by SPOPP (8) Results from the flow cytometry analyses of cell cycle populations also rationalized the decrease in cell proliferation that was first indicated by MTT. Treatment of cells with SPOPP (8) for 24 h caused a noticeable increase in the G2/M cell population (38.1 %) as compared to 2 % DMSO control (Figure. 3.3.16.). The maximal activity of SPOPP (8) was found to be at 24 h, since the G2/M population for the 44-h treatment down to 7.8 %. The percentages of the cell cycle population were estimated using the Watson model through the FlowJo software. Since PI staining was only able to measure DNA content, it cannot discern between cells at the G2 checkpoint and cells in mitosis. The ability of SPOPP (8) to halt the cell cycle combats one of the main hallmarks of cancer: resistance to anti-apoptotic signals.99 SPOPP (8) could either halt cells at the G2 phase or the mitosis phase of the cell cycle. The G2 checkpoint serves to detect aberrations in DNA before the commitment into mitosis and will halt the cell cycle through activation of checkpoint kinase 1 (Chk1).100 Piperazine derivatives have been shown to generate reactive oxygen species (ROS) within cells.101–103 SPOPP (8) can potentially generate ROS sufficient enough to cause DNA damage leading to G2/M arrest.103 Regardless of the mechanism, G2/M inhibition by SPOPP (8) could be beneficial as a therapy. G2/M arrest and resultant radiosensitivity in CRC was previously shown by Jeong et al. through treatment with metformin.104 In light of this information, SPOPP (8) could act as a radiosensitizer for CRC. 4.9 Concluding Remarks In the investigation to determine the reason for the previously reported bioactivity of UNBC152-3 (7), it was found that the mixture contained both previously reported bioactive compounds and compounds which produced a false positive in the FP assay. Furthermore, a collection of potential side-products in the F1 fraction of the crude mixture was suggested to inhibit 130 the IMP-1-KRAS mRNA interaction, but the result may be non-specific. Co-crystallization of the F1 components with the KH3to4 crystal may elucidate potential binding surfaces within the two domains. While conducting research and syntheses for the 1,3-dipolar cycloaddition of UNBC1523 (7), the molecule SPOPP (8) was synthesized as a control. Although SPOPP (8) was not the original focus, preliminary TLC data led into the direction of synthesizing and purifying SPOPP (8). SPOPP (8) does not inhibit the IMP-1-KRAS mRNA interaction, however, it clearly suppresses KRAS protein expression in human colorectal cancer cells. I hypothesize that SPOPP (8) may increase miRNA(s) leading to translational repression of KRAS protein. The common issue with cancer progression is the cellular resistance to apoptosis and the ability to evade immune surveillance. Activation of necrosis/necroptosis by SPOPP (8) may counteract both issues and present a viable therapeutic approach. Inhibition of the cell cycle at G2/M may also aid in the radiosensitization of CRC. The bioactivity of SPOPP (8) has not been reported in literature and thus the work presented herein represents novel activity of SPOPP (8) in CRC. 131 References 1. Rous, P. A Transmissible Avian Neoplasm (Sarcoma of the Common Fowl). J. Exp. Med. 12, 696–705 (1910). 2. Rous, P. A Sarcoma of the Fowl Transmissible by an Agent Separable From Tumor Cells. J. Exp. Med. 13, 397–411 (1911). 3. Wang, L.-H., Duesberg, P. H., Kawai, S. & Hanafusa, H. Location of Envelope-specific and Sarcoma-specific Oligonucleotides on RNA of Schmidt-Ruppin Rous Sarcoma Virus. Proc. Natl. Acad. Sci. 73, 447–451 (1976). 4. Stehelin, D., Guntaka, R. V, Varmus, H. E. & Bishop, J. M. Purification of DNA Complementary to Nucleotide Sequences Required for Neoplastic Transformation of Fibroblasts by Avian Sarcoma Viruses. J. Mol. Biol. 101, 349–365 (1976). 5. Martin, G. S. The Road to Src. Oncogene 23, 7910–7917 (2004). 6. Croce, C. M. Oncogenes and Cancer. N. Engl. J. Med. 358, 502–511 (2008). 7. Pylayeva-Gupta, Y., Grabocka, E. & Bar-Sagi, D. RAS Oncogenes: Weaving a Tumorigenic Web. Nat Rev Cancer 11, 761–774 (2013). 8. Ross, J. Control of Messenger RNA Stability in Higher Eukaryotes. Trends Genet. 12, 171–175 (1996). 9. Hammond, S. M. Dicing and Slicing: The Core Machinery of the RNA Interference Pathway. FEBS Lett. 579, 5822–5829 (2005). 10. Ambros, V. The Functions of Animal microRNAs. Nature 431, 350–355 (2004). 11. Ray, D., Kazan, H., Cook, K. B., Weirauch, M. T., Najafabadi, H. S., Li, X., Gueroussov, S., Albu, M., Zheng, H., Yang, A., Na, H., Irimia, M., Matzat, L. H., Dale, R. K., Smith, S. A., Yarosh, C. A., Kelly, S. M., Nabet, B., Mecenas, D., Li, W., Laishram, R. S., Qiao, M., Lipshitz, H. D., Piano, F., Corbett, A. H., Carstens, R. P., Frey, B. J., Anderson, R. A., Lynch, K. W., Penalva, Luiz, O. F., Lei, E. P., Fraser, A. G., Blencowe, B. J., Morris, Q. D. & Hughes, T. R. A Compendium of RNA-binding Motifs for Decoding Gene Regulation. Nature 499, 172–177 (2013). 12. Voineagu, I., Wang, X., Johnston, P., Lowe, J. K., Tian, Y., Horvath, S., Mill, J., Cantor, R., Blencowe, B. J. & Geschwind, D. H. Transcriptomic Analysis of Autistic Brain Reveals Convergent Molecular Pathology. Nature 474, 380–384 (2011). 13. Busà, R. Paronetto, M. P., Farini, D., Pierantozzi, E., Botti, F., Angelini, D. F., Attisani, F., Vespasiani, G. & Sette, C. The RNA-binding Protein Sam68 Contributes to Proliferation and Survival of Human Prostate Cancer Cells. Oncogene 26, 4372–82 (2007). XIX 14. Bernstein, P. L., Herrick, D. J., Prokipcak, R. D. & Ross, J. Control of c-Myc mRNA Half-life in vitro by a Protein Capable of Binding to a Coding Region Stability Determinant. Genes Dev. 6, 642–654 (1992) 15. Yisraeli, J. K. VICKZ proteins: A Multi-talented Family of Regulatory RNA-binding Proteins. Biol Cell 97, 87–96 (2005). 16. Barnes, M, van Rensburg, G., Li, W., Mehmood, K., Mackedenski, S., Chan, C., King, D. T., Miller, A. L. & Lee, C. H. Molecular Insights into the Coding Region Determinantbinding Protein-RNA Interaction through Site-directed Mutagenesis in the Heterogeneous Nuclear Ribonucleoprotein-K-homology Domains. J. Biol. Chem. 290, 625–39 (2015). 17. Git, A. & Standart, N. The KH Domains of Xenopus Vg1RBP Mediate RNA Binding and Self-association. RNA 8, 1319–1333 (2002). 18. Valverde, R., Edwards, L. & Regan, L. Structure and Function of KH Domains. FEBS J. 275, 2712–2726 (2008). 19. Chao, J. A., Patskovsky, Y., Patel, V., Levy, M., Almo, S. C. & Singer, R. H. ZBP1 Recognition of β-actin Zipcode Induces RNA Looping. Genes Dev. 24, 148–158 (2010). 20. Mackedenski, S. Molecular Characterization of the Coding Region Determinant-Binding Protein Interaction with KRAS Oncogene mRNA. (University of Northern British Columbia, 2015). 21. Leeds, P., Kren, B. T., Boylan, J. M., Betz, N. A., Steer, C. J., Gruppuso, P. A. & Ross, J. Developmental Regulation of CRD-BP, an RNA-binding Protein that Stabilizes c-Myc mRNA in vitro. Oncogene 14, 1279–1286 (1997). 22. Hansen, T. V. O., Hammer, N. A., Nielsen, J., Madsen, M., Dalbaeck, C., Wewer, U. M., Christiansen, J. & Nielsen, F. C. Dwarfism and Impaired Gut Development in Insulinlike Growth Factor II mRNA-binding Protein 1-deficient Mice. Mol. Cell. Biol. 24, 4448–64 (2004). 23. Sparanese, D. & Lee, C. H. CRD-BP Shields c-Myc and MDR-1 RNA from Endonucleolytic Attack by a Mammalian Endoribonuclease Nucleic Acids Res.. 35, 1209–1221 (2007). 24. Barnes, T., Kim, W., Mantha, A. K., Kim, S., Izumi, T., Mitra, S. & Lee, C. H. Identification of Apurinic/apyrimidinic Dndonuclease 1 (APE1) as the Endoribonuclease that Cleaves c-Myc mRNA. Nucleic Acids Res. 37, 3946–3958 (2009). 25. Elcheva, I., Goswami, S., Noubissi, F. K. & Spiegelman, V. S. CRD-BP Protects the Coding Region of BTrCP1 mRNA from miR-183-Mediated Degradation. Mol. Cell 35, 240–246 (2009). 26. King, D. T., Barnes, M., Thomsen, D. & Lee, C. H. Assessing Specific Oligonucleotides and Small Molecule Antibiotics for the Ability to Inhibit the CRD-BP-CD44 RNA Interaction. PLoS One 9, (2014). XX 27. Mongroo, P. S., Noubissi, F. K., Cuatrecasas, M., Kalabis, J., King, C. E., Johnstone, C. N., Bowser, M. J., Castells, A., Spiegelman, V. S. & Rustgi, A. K. IMP-1 Displays Cross-talk with K-Ras and Modulates Colon Cancer Cell Survival Through the Novel Proapoptotic Protein CYFIP2. Cancer Res. 71, 2172–2182 (2011). 28. Kim, R. B., Fromm, M. F., Wandel, C., Leake, B., Wood, A., Roden, D. M. & Wilkinson, G. R. The Drug Transporter P-glycoprotein Limits Oral Absorption and Brain Entry of HIV-1 Protease Inhibitors. J. Clin. Invest. 101, 289–294 (1998). 29. Chi V. Dang. MYC on the Path to Cancer. Cell 149, 22–35. (2012). 30. Sneath, R. J. & Mangham, D. C. The Normal Structure and Function of CD44 and its Role in Neoplasia. Mol. Pathol. 51, 191–200 (1998). 31. Ioannidis, P. Mahaira, L., Papadopoulou, A., Teixeira, M. R., Heim, S., Anderson, J. A., Evangelou, E., Dafni, U., Pandis, N. & Trangas, T. 8q24 Copy Number Gains and Expression of the c-Myc mRNA Stabilizing Protein CRD-BP in Primary Breast Carcinomas. Int. J. Cancer 104, 54–59 (2003). 32. Ross, J., Lemm, I. & Berberet, B. Overexpression of an mRNA-binding Protein in Human Colorectal Cancer. Oncogene 20, 6544–6550 (2001). 33. Noubissi, F. K., Kim, T., Kawahara, T. N., Aughenbaugh, W. D., Berg, E., Longley, B. J., Athar, M. & Spiegelman, V. S. Role of CRD-BP in the Growth of Human Basal Cell Carcinoma Cells. J. Invest. Dermatol. 134, 1718–1724 (2014). 34. Dimitriadis, E., Trangas, T., Milatos, S., Foukas, P. G., Gioulbasanis, I., Courtis, N., Nielsen, F. C., Pandis, N., Dafni, U., Bardi, G. & Ioannidis, P. Expression of Oncofetal RNA-binding Protein CRD-BP/IMP1 Predicts Clinical Outcome in Colon Cancer. Int. J. Cancer 121, 486–494 (2007). 35. Ioannidis, P., Kottaridi, C., Dimitriadis, E., Courtis, N., Mahaira, L., Talieri, M., Giannopoulo, A., Iliadis, K., Papaioannou, D., Nasioulas, G. & Trangas, T. Expression of the RNA-binding Protein CRD-BP in Brain and Non-small Cell Lung Tumors. Cancer Lett. 209, 245–250 (2004). 36. Qu, Y., Pan, S., Kang, M., Dong, R. & Zhao, J. MicroRNA-150 Functions as a Tumor Suppressor in Osteosarcoma by Targeting IGF2BP1. Tumor Biol. 37, 5275–5284 (2016). 37. Luo, Y. Sun, R., Zhang, J., Sun, T., Liu, X. & Yang, B. miR-506 Inhibits the Proliferation and Invasion by Targeting IGF2BP1 in Glioblastoma. Am. J. Transl. Res. 7, 2007–2014 (2015). 38. Mahapatra, L., Mao, C., Andruska, N., Zhang, C. & Shapiro, D. J. High-Throughput Fluorescence Anisotropy Screen for Inhibitors of the Oncogenic mRNA Binding Protein, IMP-1. J. Biomol. Screen. 19, 427–436 (2014). XXI 39. Vakiani, E. & Solit, D. B. KRAS and BRAF: Drug Targets and Predictive Biomarkers. J. Pathol. 223, 219–229 (2011). 40. Ellis, C. A. & Clark, G. The Importance of Being K-Ras. Cell. Signal. 12, 425–434 (2000). 41. Tsai, F. D., Lopes, M. S., Zhou, M., Court, H., Ponce, O., Fiordalisi, J. J., Gierut, J. J., Cox, A. D., Haigis, K. M. & Philips, M. R. K-Ras4A Splice Variant is Widely Expressed in Cancer and Uses a Hybrid Membrane-targeting Motif. Proc. Natl. Acad. Sci. 112, 779– 784 (2015). 42. Koera, K. Nakamura, K., Nakao, K., Miyoshi, J., Toyoshima, K., Hatta, T., Otani, H., Aiba, A. & Katsuki, M. K-ras is Essential for the Development of the Mouse Embryo. Oncogene 15, 1151–1159 (1997). 43. Umanoff, H., Edelmann, W., Pellicer, A. & Kucherlapati, R. The Murine N-Ras Gene is Not Essential for Growth and Development. Proc. Natl. Acad. Sci. 92, 1709–1713 (1995). 44. Berndt, N., Hamilton, A. D. & Sebti, S. M. Targeting Protein Prenylation for Cancer Therapy. Nat Rev Cancer. 11, 775–791 (2011). 45. Buhrman, G., Holzapfel, G., Fetics, S. & Mattos, C. Allosteric Modulation of Ras Positions Q61 for a Direct Role in Catalysis. Proc. Natl. Acad. Sci. 107, 4931–4936 (2010). 46. Hobbs, G. A., Der, C. J. & Rossman, K. L. RAS Isoforms and Mutations in Cancer at a Glance. J. Cell Sci. 129, 1287–1292 (2016). 47. Buttitta, F. Barassi, F., Fresu, G., Felicioini, L., Chella, A., Paolizzi, D., Lattanzio, G., Salvatore, S., Camplese, P. P., Rosini, S., Iarussi, T., Mucilli, F., Mezzetti, A., Sacco, R. & Marchetti, A. Mutational Analysis of the HER2 Gene in Lung Tumors from Caucasian Patients: Mutations are Mainly Present in Adenocarcinomas with Bronchioloalveolar Features. Int. J. Cancer. 119, 2586–2591 (2006). 48. Eser, S., Schnieke, A., Schneider, G. & Saur, D. Oncogenic Kras Signalling in Pancreatic Cancer. Br. J. Cancer. 111, 817–822 (2014). 49. Vaughn, C. P., ZoBell, S. D., Furtado, L. V., Baker, C. L. & Samowitz, W. S. Frequency of KRAS, BRAF, and NRAS Mutations in Colorectal Cancer. Genes. Chromosomes Cancer 50, 307–312 (2011). 50. Andreyev, H. J., Norman A. R., Cunningham, D., Oates, J., Dix, B. R., Iacopetta, B.J., Young, J., Walsh, T., Ward, R., Hawkins, N., Beranek, M., Jandik, P., Benamouzig, R., Jullian, E., Laurent-Puig, P., Olschwang, S., Muller, O., Hoffmann, I., Rabes, H. M., Zietz, C., Troungos, C., Valavanis, C., Yuen, S. T., Ho, J. W., Croke, C. T., O'Donoghue, D. P., Giaretti, W., Rapallo, A., Russo, A., Bazan, V., Tanaka, M., Omura, K., Azuma, T., Ohkusa, T., Fujimori, T., Ono, Y., Pauly, M., Faber, XXII C., Glaesener, R., de Goeij, A. F.., Arends, J. W., Andersen, S. N., Lövig, T., Breivik, J., Gaudernack, G., Clausen, O. P., De Angelis, P. D., Meling, G. I., Rognum, T. O., Smith, R., Goh, H. S., Font, A., Rosell, R., Sun, X. F., Zhang, H., Benhattar, J., Losi, L., Lee, J. Q., Wang, S. T., Clarke, P.A., Bell, S., Quirke, P., Bubb, V. J., Piris, J., Cruickshank, N. R., Morton, D., Fox, J. C., Al-Mulla, F., Lees, N., Hall, C. N., Snary, D., Wilkinson, K., Dillon, D., Costa, J., Pricolo, V. E., Finkelstein, S. D., Thebo, J. S., Senagore, A. J., Halter, S. A., Wadler, S., Malik, S., Krtolica, K. & Urosevic N. Kirsten Ras Mutations in Patients with Colorectal Cancer: The ‘RASCAL II’ Study. Br. J. Cancer. 85, 692–6 (2001). 51. Guerrero, S. Casanova, I., Farre, L., Mazo, A., Capella, G. & Mangues, R. K-Ras Codon 12 Mutation Induces Higher Level of Resistance to Apoptosis and Predisposition to Anchorage-Independent Growth Than Codon 13 Mutation or Proto-Oncogene Overexpression. Cancer Res. 60, 6750–6756 (2000). 52. Haigis, K. M., Kendall, K. R., Wang, Y., Cheung, A., Haigis, M. C., Glickman, J. N., Niwa-Kawakita, M., Sweet-Cordero, A., Sebolt-Leopold, J., Shannon, K. M., Settleman, J., Gkovannini, M. & Jacks, T. Differential Effects of Oncogenic K-Ras and N-Ras on Proliferation, Differentiation and Tumor Progression in the Colon. Nat. Genet. 40, 600– 608 (2008). 53. Birkeland, E., Wik, E., Mjø, S., Hoivik, E. A., Trovik, J., Werner, H. M. J., Kusonmano, K., Petersen, K., Raeder, M. B., Holst, F., Øyan, A. M., Kalland, K-H., Akslen, L. A., Simon, R., Krakstad, C. & Salvesen, H. B. KRAS Gene Amplification and Overexpression But Not Mutation Associates with Aggressive and Metastatic Endometrial Cancer. Br. J. Cancer. 107, 1997–2004 (2012). 54. Friday, B. B. & Adjei, A. A. K-ras as a Target for Cancer Therapy. Biochim. Biophys. Acta - Rev. Cancer 1756, 127–144 (2005). 55. Chen, X. Guo, X., Zhang, H., Xiang, Y., Chen, J., Yin, Y., Cai, X., Wang, K., Wang, G., Ba, Y., Zhu, L., Wang, J., Yang, R., Zhang, Y., Ren, Z., Zen, K., Zhang, J. & Zhang, CY. Role of miR-143 Targeting KRAS in Colorectal Tumorigenesis. Oncogene. 28474, 1385–1392 (2009). 56. Shin, K.-H., Bae, S. D., Hong, H. S., Kim, R. H., Kang, M. K. & Park, N-H. miR-181a Shows Tumor Suppressive Effect Against Oral Squamous Cell Carcinoma Cells by Downregulating K-ras. Biochem. Biophys. Res. Commun. 404, 896–902 (2011). 57. Ross, S. J. Revenko, A. S., Hanson, L. L., Ellston, R., Staniszewska, A., Whalley, N., Pandey, S. K., Revill, M., Rooney, C., Buckett, L. K., Klein, S. K., Hudson, K., Monia, B. P., Zinda, M., Blackey, D. C., Lyne, P. D. & Macleod, A. R. Targeting KRASDependent Tumors with AZD4785 , a High-Affinity Therapeutic Antisense Oligonucleotide Inhibitor of KRAS. Sci. Transl. Med. 9, 1–14 (2017). 58. Seth, P. P., Siwkowski, A., Allerson, C. R., Vasquez, G., Lee, S., Praskash, T. P., Wancewicz, E.V., Witchell, D. & Swayze, E. E. Short Antisense Oligonucleotides with XXIII Novel Analogues Show Improved Potency without Increased Toxicity in Animals. J. Med. Chem. Lett. 52, 10–13 (2009). 59. Juliano, R., Baumann, J., Kang, H. & Ming, X. Biological Barriers to Therapy with Antisense and siRNA Oligonucleotides. Mol. Pharm. 6, 686–695 (2009). 60. Baines, A. T., Xu, D. & Der, C. J. Inhibition of Ras for Cancer Treatment: The Search Continues. Future Med. Chem. 3, 1787–808 (2011). 61. Lobell, R. B., Omer, C. A., Abrams, M. T., Bhimnathwala, H. G., Brucker, M. J., Buser, C. A., Davide, J. P., deSolms, S. J., Dinsmore, C. J., Ellis-Hutchings, M. S., Kral, A. M., Liu, D., Lumma, W. C., Machotka, S. V., Rands, E., Williams, T. M., Graham, S. L. Hartman, G. D., Oliff, A. I., Heimbrook, D. C. & Kohl, N. E. Evaluation of Farnesyl: Protein Transferase and Geranylgeranyl: Protein Transferase Inhibitor Combinations in Preclinical Models. Cancer Res. 61, 8758–8768 (2001). 62. Chen, L., Zhuang, C., Lu, J., Jiang, Y. & Sheng, C. Discovery of Novel KRAS-PDE δ Inhibitors by Fragment-Based Drug Design. J. Med. Chem. 61, 2604–2610 (2018). 63. Maurer, T. Garrenton, L. S., Oh, A., Pitts, K., Anderson, D. J., Skelton, N. J., Fauber, B. P., Pan, B., Malek, S., Stoko, D., Ludlam, M. J. C., Bowman, K. K., Wu, J., Giannetti, A. M., Starovasnik, M. A., Mellman, I., Jackson, P. K., Rudolph, J., Wang, W. & Fang, G. Small-molecule Ligands Bind to a distinct Pocket in Ras and Inhibit SOS-Mediated Nucleotide Exchange Activity. Proc. Natl. Acad. Sci. 109, 5299–5304 (2012). 64. John, J., Sohmen, R., Feuerstein, J., Linke, R., Wittinghofer, A. & Goody, R. S. Kinetics of Interaction of Nucleotides with Nucleotide-Free H-ras p21. Biochemistry. 29, 6058– 6065 (1990). 65. Ostrem, J. M., Peters, U., Sos, M. L., Wells, J. A. & Shokat, K. M. K-Ras(G12C) Inhibitors Allosterically Control GTP Affinity and Effector Interactions. Nature. 503, 548–551 (2013). 66. Fell, J. B., Fischer, J. P., Baer, B. R., Ballard, J., Blake, J. F., Bouhana, K., Brandhuber, B. J., Briere, D. M., Burgess, L. E., Burkard, M. R., Chiang, H., Chicarelli, M. J., Davidson, K., Guadino, J. J., Hallin, J., Hanson, L., Hee, K., Hicken, E. J., Hinklin, R. J., Marx, M. A., Mejia, M. J., Olson, P., Savchenkov, P., Sudhakar, N., Tang, T. P., Vigers, G. P., Zeca, H. & Christensen, J. G. Discovery of Tetrahydropyridopyrimidines as Irreversible Covalent Inhibitors of KRAS-G12C with In Vivo Activity. ACS Med. Chem. Lett. 9, 1230–1234 (2018). 67. Yarden, Y. The EGFR Family and its Ligands in Human Cancer: Signalling Mechanisms and Therapeutic Opportunities. Eur. J. Cancer. 37, s3–s8 (2001). 68. Pao, W., Miller, V. A., Politi, K. A., Riely, G. J., Somwar, R., Zakowski, M. F., Kris, M. G. & Varmus, H. Acquired Resistance of Lung Adenocarcinomas to Gefitinib or Erlotinib is Associated with a Second Mutation in the EGFR Kinase Domain. PLoS Med. 2, 0225–0235 (2005). XXIV 69. Wilhelm, S. M., Carter, C., Tang, L., Wilkie, D., Mcnabola, A., Rong, H., Chen, C., Zhang, X., Vincent, P., Mchugh, M., Cao, Y., Shujath, J., Gawlak, S., Eveleigh, D., Rowley, B., Liu, L., Adnane, L.,Lynch, M., Auclair, D., Taylor, I., Gedrich, R., Voznesensky, A., Riedl, B., Post, L. E., Bollag, G. & Trail, P. A. BAY 43-9006 Exhibits Broad Spectrum Oral Antitumor Activity and Targets The RAF/MEK/ERK Pathway and Receptor Tyrosine Kinases Involved in Tumor Progression and Angiogenesis. Cancer Res. 64, 7099–7109 (2004). 70. van Malenstein, H., Dekervel, J., Verslype, C., Cutsem, E. V., Windmolders, P., Nevens, F., Pelt, J. V. Long-term Exposure to Sorafenib of Liver Cancer Cells Induces Resistance with Epithelial-To-Mesenchymal Transition, Increased Invasion and Risk of Rebound Growth. Cancer Lett. 329, 74–83 (2013). 71. Infante, J. R., Fecher, L. A., Falchook, G. S., Nallapareddy, S., Gordon, M. S., Becerra, C., Demarini, D. J., Cox, D. S., Xu, Y., Morris, S. R., Peddareddigari, V. G. R., Le, N, T., Hart, L., Bendell, J. C., Eckhardt, G., Kurzrock, R., Flaherty, K., Burris III, H. A., Messersmith, W. A. Safety, Pharmacokinetic, Pharmacodynamic, and Efficacy Data for the Oral MEK Inhibitor Trametinib: A Phase 1 Dose-Escalation Trial. Lancet Oncol. 13, 773–781 (2012). 72. Mackedenski, S., Wang, C., Li, W. M. & Lee, C. H. Characterizing the Interaction Between Insulin-Like Growth Factor 2 mRNA-Binding Protein 1 (IMP1) and KRAS Expression. Biochem. J. 475, 2749–2767 (2018). 73. Wang, C. Inhibitors of the CRD-BP-KRas Interaction. (University of Northern British Columbia, 2016). 74. Lea, W. A. & Simeonov, A. Fluorescence Polarization Assay. Expert Opin Drug Discov. 6, 17–32 (2012). 75. Dimmock, J. R., Zello, G. A., Oloo, E. O., Quail, J. W., Kraatz, H-B., Perjesi, P., Aradi, F., Takacs-Novak, K., Allen, T. M., Santos, C. L., Balzarini, J., Clercq, E. D. & Stables, J. P. Correlations Between Cytotoxicity and Topography Of Some 2Arylidenebenzocycloalkanones Determined by X-ray Crystallography. J. Med. Chem. 45, 3103–3111 (2002). 76. Wei, Ang, C., Ali, M. A., Yoon, Y. K., Choi, S. B., Osman, H., Masand, V. H. & Choon, T. S. Antimycobacterial Activity and in silico Study of Highly Functionalised Dispiropyrrolidines. Med. Chem. Res. 24, 818–828 (2015). 77. Haddad, S. Boudriga, S., Porzio, F., Soldera, A., Askri, M., Knorr, M., Rousselin, Y., Kubicki, M. M., Golz, C. & Strohmann, C. Regio- and Stereoselective Synthesis of Spiropyrrolizidines and Piperazines through Azomethine Ylide Cycloaddition Reaction. J. Org. Chem. 80, 9064–9075 (2015). XXV 78. Ho, C. S., Lam, C. W. K., Chan, M. H. M., Cheung, R. C. K., Law, L. K., Lit, L. C. W., Ng, K. F., Suen, M. W. M. & Tai, H. L. Electrospray Ionisation Mass Spectrometry: Principles and Clinical Applications. Clin. Biochem. Rev. 24, 3–12 (2003). 79. Chung, S. W., Chung, H. Y., Toriba, A., Kameda, T., Tang, N., Kizu, R. & Hayakawa, K. An Environmental Quinoid Polycyclic Aromatic Hydrocarbon, Acenaphthenequinone, Modulates Cyclooxygenase-2 Expression through Reactive Oxygen Species Generation and Nuclear Factor Kappa B Activation in A549 Cells. Toxicol. Sci. 95, 348–355 (2007). 80. Martinez, A., Fernandez, M., Estevez, J. C., Estevez, R. J. & Castedo, L. New ’2phenylnaphthalene’-mediated Synthesis of Benzo[B]naphtho[2,3-D] furan-6,11-diones and 6-Oxa-benzo[A]anthracene-5,7,12-triones: First Total Synthesis of 6-oxabenzo[A]anthracen-5-ones. Tetrahedron 61, 1353–1362 (2005). 81. Liu, Y., Peterson, D. A., Kimura, H. & Schubert, D. Diphenyltetrazolium Bromide ( MTT) Reduction. J. Neurochem. 69, 581–593 (1997). 82. Saud, S. M., Li, W., Morris, N. L., Matter, M. S., Colburn, N. H., Kim, Y. S. & Young, M. R. Resveratrol Prevents Tumorigenesis in Mouse Model of Kras Activated Sporadic Colorectal Cancer by Suppressing Oncogenic Kras Expression. Carcinogenesis. 35, 2778–2786 (2014). 83. Yu, S., Lu, Z., Liu, C., Meng, Y. & Ma, Y. miRNA-96 Suppresses KRAS and Functions as a Tumor Suppressor Gene in Pancreatic Cancer. Cancer Res. 70, 6015–6025 (2010). 84. Schmid, I., Uittenbogaart, C. H. & Giorgi, J. V. Sensitive Method for Measuring Apoptosis and Cell Surface Phenotype in Human Thymocytes by Flow Cytometry. Cytometry 15, 12–20 (1994). 85. Vermes, I., Haanen, C., Steffens-nakken, H. & Reutelingsperger, C. A Novel Assay For Apoptosis Flow Cytometric Detection of Phosphatidylserine Expression on Early Apoptotic Cells Using Fluorescein Labelled Annexin V. J. Immunol. Methods. 184, 39– 51 (1995). 86. Krishan, A. Rapid Flow Cytofluorometric Analysis of Mammalian Cell Cycle by Propidium Iodide Staining. J. Cell Biol. 66, 188–193 (1975). 87. Noubissi, F. K., Elcheva, I., Bhatia, N., Shakoori, A., Ougolkov, A., Liu, J., Minamoto, T., Ross, J., Fuchs, S. Y. & Spiegelman, V. S. CRD-BP Mediates Stabilization of βTrCP1 and c-Myc mRNA in Response to β-catenin Signalling. Nature. 441, 898–901 (2006). 88. Bashiardes, G., Safir, I., Mohamed, A. S. & Barbot, F. Microwave-assisted [3+2] Cycloadditions of Azomethine Ylides. Org. Lett. 5, 4915–4918 (2003). 89. Wilkie, G. S., Dickson, K. S. & Gray, N. K. Regulation of mRNA Translation by 5ʹ - and 3ʹ -UTR-binding factors. Trends Biochem. Sci. 28, 182–188 (2003). 90. Wang, Z., Gerstein, M. & Snyder, M. RNA-Seq : A Revolutionary Tool for Transcriptomics. Nat. Rev. Genet. 10, 57–63 (2010). XXVI 91. Chan, F. K. M., Shisler, J., Bixby, J. G., Felices, M., Zheng, L., Appel, M., Orenstein, J., Moss, B. & Lenardo, M. J. A Role for Tumor Necrosis Factor Receptor-2 and Receptorinteracting Protein in Programmed Necrosis and Antiviral Responses. J. Biol. Chem. 278, 51613–51621 (2003). 92. Cho, Y., Challa, S., Moquin, D., Genga, R., Ray, T. D., Guildford, M. & Chan, F. K-M. Phosporylation-driven Assembly of the RIP1-RIP3 Complex Regulates Programmed Necrosis and Virus-Induced Inflammation. Cell. 137, 1112–1123 (2009). 93. Micheau, O. & Tschopp, J. Induction of TNF receptor 1-mediated Apoptosis Via Two Sequential Signaling Complexes. Cell. 114, 181–190 (2003). 94. Sun, L., Wang, H., Wang, Z., He, S., Chen, S., Liao, D., Wang, L., Yan, J., Liu, W., Lei, X. & Wang, X. Mixed Lineage Kinase Domain-Like Protein Mediates Necrosis Signaling Downstream of RIP3 Kinase. Cell. 148, 213–227 (2012). 95. Dondelinger, Y., Declercq, W., Montessuit, S., Roelandt, R., Goncalves, A., Bruggeman, I., Hulpiau, P., Weber, K., Sehon, C. A., Marquis, R. W., Bertin ,J., Gough, P. J., Savvides, S., Martinou, J-C., Berrand, M. J. M. Vandenabeele, P. MLKL Compromises Plasma Membrane Integrity by Binding to Phosphatidylinositol Phosphates. Cell. Rep. 7, 971–981 (2014). 96. Scaffidi, P., Misteli, T. & Bianchi, M. E. Release of Chromatin Protein HMGB1 by Necrotic Cells Triggers Inflammation. Nature. 418, 191–195 (2002). 97. Gao, Q. Li, F., Wang, S., Shen, Z., Cheng, S., Ping, Y., Qin, G., Chen, X., Yang, L., Cao, L., Liu, C., Zhang, B., Wang, L., Sun, Y. & Zhang, Y. A Cycle Involving HMGB1, IFNγ And Dendritic Cells Plays a Putative Role in Anti-Tumor Immunity. Cell. Immunol. 1– 9 (2018). 98. Bindea, G., Mlecnik, B., Tosolini, M., Kirilovsky, A., Waldner, M., Obenauf, A. C., Angell, H., Fredriksen, T., Lafontaine, L., Berger, A., Bruneval, P., Fridman, W. H., Becker, C., Pages, F., Speicher, M. R., Trajanoski, Z. & Gallon, J. Spatiotemporal Dynamics of Intratumoral Immune Cells Reveal the Immune Landscape in Human Cancer. Immunity. 39, 782–795 (2013). 99. Hanahan, D. & Weinberg, R. A. The Hallmarks of Cancer. Cell. 100, 57–70 (2000). 100. Jackson, J. R., Gilmartin, A., Imburgia, C., Winkler, J. D., Marshall, L. A. & Roshak, A. An Indolocarbazole Inhibitor of Human Checkpoint Kinase (Chk1) Abrogates Cell Cycle Arrest Caused by DNA Damage. Cancer Res. 60, 566–572 (2000). 101. Khanam, R., Kumar, R., Hejazi, I. I., Shahabuddin, S., Meena, R., Rajamani, P., Yadav, N., Bhat, A. I. & Athar, F. New N-benzhydrylpiperazine/1,3,4-oxadiazoles Conjugates Inhibit the Proliferation, Migration, and Induce Apoptosis in HeLa Cancer Cells Via Oxidative Stress-Mediated Mitochondrial Pathway. J. Cell. Biochem. 120, 1651–1666 (2018). XXVII 102. Sampson, J. J., Donkor, I. O., Huang, T. L. & Adunyah, S. E. Novel Piperazine Induces Apoptosis in U937 Cells. Int. J. Biochem. Mol. Biol. 2, 78–88 (2011). 103. Costa, F. B., Cortez, A. P., de Ávila, R. I., de Carvalho, F. S., Andrade, W. M., da Cruz, A. F., Reis, K. B., Menegatti, R., Lião, L. M., Romeiro, L. A. S., Noël, F., Fraga, C. A. M., Barreiro, E. J., Sanz, G., Rodrigues, M. F., Vaz, B. G. & Valadares, M. C. The Novel Piperazine-Containing Compound LQFM018: Necroptosis Cell Death Mechanisms, Dopamine D4 Receptor Binding and Toxicological Assessment. Biomed. Pharmacother. 102, 481–493 (2018). 104. Jeong, Y. K., Kim, M. S., Lee, J. Y., Kim, E. H. & Ha, H. Metformin Radiosensitizes p53-deficient Colorectal Cancer Cells Through Induction of G2/M Arrest and Inhibition of DNA Repair Proteins. PLoS One 10, 1–15 (2015). XXVIII