Date of Award

2018

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Drug Discovery and Biomedical Sciences

College

College of Graduate Studies

First Advisor

Patrick M. Woster

Second Advisor

Ramsay Camp

Third Advisor

Danielle Townsend

Fourth Advisor

Campbell McInnes

Fifth Advisor

Craig C. Beeson

Abstract

The polyamine oxidase family of enzymes has been extensively studied and plays a critical role in cell development. These enzymes catalyze the oxidation of many different polyamines, and catalyze transformations on DNA-associated histones in order to alter the epigenetic code to modulate gene expression. Polyamines are known to be essential in all forms of living organisms at multiple sites, and they have been identified in animals, plants and fungi. Polyamine oxidases (PAOs), including enzymes such as lysine-specific demethylase 1 (LSD1, KDM1A), spermine oxidase (SMOX) and acetylpolyamine oxidase (APAO) oxidize a variety of polyamines and other amine-containing substrates. Each one of these polyamine oxidases serves a unique purpose in cell cycle regulation, acting on various pathways, and inhibition of these oxidases affects the cell in a variety of different ways. Further, polyamine metabolism is highly regulated, and dysregulation of this pathway has been a prominent feature in various human cancers. The histone demethylase KDM1A was the first to be discovered, and there are now more than 30 known histone demethylases. The enzyme encoded by the gene KDM1A is a flavin-dependent amine oxidase that demethylates mono and dimethylated lysine in histone tails, specifically on histone 3, lysine 4 (H3K4) and histone 3, lysine 9 (H3K9). It has been shown that KDM1A is up regulated and directly linked to multiple diseases, resulting in decreased methylation at the chromatin marks H3K4 and H3K9. Both of these chromatin marks have been identified as novel targets for pharmacological intervention. Because methylated H3K4 is known to be an activating chromatin mark, we sought to promote re-expression of aberrantly silenced tumor suppressor genes by discovering novel, potent, small molecule inhibitors to attenuate the cellular activity of KDM1A, thus increasing levels of monomethyl H3K4 (H3K4me) and dimethyl H3K4 (H3K4me2). However, through these studies, we discovered a unique interplay between modulating these histones marks and regulation of the polyamine back conversion pathway via the homologous enzyme spermine oxidase (SMOX). We employed powerful computational chemistry approaches to simulate ligand binding to both KDM1A and SMOX, and verified these in silico results using a fluorescence-based medium throughput assay, and were able to identify a novel series of compounds that serve as dual inhibitors by modulating KDM1A and down regulating SMOX function simultaneously. More than 100 inhibitors were synthesized and evaluated, and many exhibited low nanomolar IC50 values against KD1MA, as well as low micromolar IC50 values against SMOX. These compounds exhibited >100-fold selectivity when compared to the flavin-dependent amine oxidases monoamine oxidase A (MAOA) and B (MAOB), as well against the peroxisomal polyamine oxidase acetylpolyamine oxidase APAO. Selected inhibitors were a shown to significantly increase H3K4 methylation in multiple cancer cell lines in which LSD1 is known to be overexpressed. These compounds proved to be more “drug like” and to produce reversible inhibition, and can be synthesized via new, versatile and facile synthetic routes. KDM1A is a flavin adenine dinucleotide amine oxidase that has been shown to be over expressed in a multitude of tumors, recently making all demethylases an attractive, emerging target. KDM1A is one of many lysine demethylases (KDMS) in the family such as KDM1B, KDM2A/B, KDM3A/B, KDM4-F, KDM5A-D all the way through KDM8. Each of these demethylase proteins possess specific target sites for lysines to be demethylated. Many of these demethylases in the family contain the transcriptional repressor domain, jumanji (jmj), which actively partake in chromatin regulation and development.1 KDM1A specifically demtheylates mono and di methylated lysines on Histone 3 and 9. The demethylation of these specific markers lead to the condensation of chromatin, repressing the transcription of multiple antitumor gene regions such as p53, DMNT1,p21 WAF-1, GATA-1/2 etc. KDM1A also has multiple binding partners with which it complexes with in order to perform Its various functions. Inhibition of KDM1A function results in tumor suppressor gene transcription leading to suppression of cancer. The KDM1A methylation pathway has been well studied, as many groups are working on synthesizing inhibitors which specifically target its’ function. SMOX is a member of the ubiquitous flavin-dependent amine oxidase family, and has significant sequence homology to KDM1A and other amine oxidases. Like KDM1A and other related oxidases, SMOX requires a flavin-containing cofactor that is crucial for the catalytic activity of the enzyme. SMOX is the major catabolic enzyme in the mammalian polyamine metabolic pathway, and converts the natural polyamines spermine and spermidine to the next lower polyamine spermidine and putrescine, respectively, by oxidative deamination at an aminopropyl end group. The by-products of the SMOX reaction are 3-acetamidopropanal and hydrogen peroxide. The tetraamine spermine is highly conserved in the polyamine pathway, but it has been shown that spermine alone is not sufficient for survival in mammalian organisms; in fact, the transformation of spermine into spermidine is vital. Without this conversion, spermine levels can accumulate and cause growth irregularities, but more importantly if levels of spermine are too high, metabolic dysregulation ensues, and could lead to cancer initiation and progression. Spermidine and spermine participate in a broad range of cellular functions including cell cycle modulation, scavenging reactive oxygen species (ROS), stabilization of DNA structure, protein synthesis and control of gene expression. We have demonstrated that our second generation of derivatives are up to 50% more potent against SMOX when compared to the previously reported inhibitor MDL72527. Using purified proteins and endogenous substrates, the spermine/spermidine conversion ratio can be measured via LCMS analysis. Importantly, SMOX has been shown to be overexpressed in many diseases including pancreatic cancers and is an attractive target for therapeutic intervention. Because the specificity of our compounds for the target enzymes KDM1A and SMOX is a critical factor in avoiding off-target effects, we have evaluated active compounds against related flavin-dependent enzymes in the amine oxidase family. Activity against the related amine oxidases monoamine oxidase A and B (MAO-A and MAO-B must be ruled out, since inhibition of these enzymes could cause off-target effects that would preclude their use in the clinic. In addition, the peroxisomal polyamine oxidase acetylpolyamine oxidase (APAO), which oxidizes acetylated polyamines produces by spermidine/spermine-N1-acetyltransferase (SSAT) must be ruled out. The inhibitors we have produced to date show a significant selectivity toward the target enzymes when compared to MAO-A and B, and also to APAO. Since pancreatic tumor cell lines exhibit increased levels of LSD1 as well as SMOX, these enzymes represent logical targets for chemotherapeutic intervention, and some of the inhibitors we have developed to date produce antitumor responses similar to the gemcitabine, the current first-line therapy for pancreatic cancer. Selected analogues produce cytotoxicity at low micromolar concentrations against the immortalized pancreatic adenocarcinoma cell lines Miapaca1, Panc-1 and BXPC-3, but are relatively non-toxic to non-cancerous mixed intestinal cells. Our current lead analogs demonstrate favorable half lives in vitro, possess promising bioavailability profiles and appear to have acceptable therapeutic indices. Importantly, selected compounds in the 3,5-diamino-1,2,4-triazole class rival the current best-in-class reversible KDM1A inhibitors, and are the most potent SMOX inhibitors described to date. Thus, these analogues can be considered as dual KDM1A/SMOX inhibitors with potential for the treatment of pancreatic cancer.

Rights

All rights reserved. Copyright is held by the author.

Download

Share

COinS