Transition state mimics as inhibitors of methyltransferase enzymes

Abstract

The work described in this thesis focuses on the development of inhibitors against two distinct methyltransferase enzymes; the protein arginine N-methyltransferases (PRMTs) and nicotinamide N-methyltransferase (NNMT). PRMTs are enzymes that methylate arginine side-chains in proteins, including histones, and by doing so regulate the transcription of genes. The dysregulation of PRMTs is involved in a wide variety of diseases, including many forms of cancer. NNMT methylates nicotinamide (vitamin b3) and other pyridines and was initially thought to be involved in detoxification and metabolic pathways only. However, recent publications show its involvement in cancer, obesity and Parkinson’s disease. Initially, small molecule inhibitors of PRMTs were designed to contain structural elements of the two substrates involved in the enzymatic reaction. By covalently linking parts of the methyl donor S-adenosyl-L-methionine (AdoMet) to a guanidine moiety, as present in the arginine side-chain, bisubstrate inhibitors were prepared. These bisubstrate compounds were designed to simultaneously occupy both the methyl donor and arginine-substrate binding sites and in doing so mimic the transition state of the enzymatic reaction. The inhibition results against three pronounced PRMTs revealed potent inhibition with surprising selectivity for PRMT4. Crystal structures of the most potent inhibitors bound with PRMT2 and PRMT4 were obtained, confirming the hypothesized binding of the compounds in the PRMT active site. Follow-up research was done focussing on transition state mimics of PRMT4 synthesized with different PRMT4-specific peptide fragments. Using a newly developed solid phase synthesis procedure, the adenosine part of AdoMet was covalently linked to the side chain of the target arginine in peptide fragments of PRMT4-specific protein substrates. Inhibition studies showed these compounds to be potent inhibitors of PRMT4 with high selectivity over the most prevalent PRMT1. In addition, the transition state mimics were found to form stable complexes with the enzyme, thereby facilitating co-crystallization. Co-crystal structures of several transition state mimics with PRMT4 were obtained. Importantly, this approach has the potential to provide new structural insights into substrate binding by PRMTs. While recently it has become clear that NNMT is involved in a variety of diseases, very few inhibitors have been described to date. In order to evaluate inhibitors against NNMT, a suitable screening method for the analysis of this enzymatic reaction first had to be developed. Using the properties of the highly polar enzymatic products, we developed an Ultra-High Performance, Hydrophilic Interaction Liquid Chromatography method coupled to Mass Spectrometry (UHP-HILIC-MS). The suitability of this method was demonstrated by probing the substrate tolerance of NNMT in a quantitative manner. In this study, the NNMT-mediated methylation of a range of pyridine-containing substrates was tested. For compounds identified as NNMT substrates, kinetic parameters were determined providing insight into the functional group tolerance of the enzyme active site. In addition, the method was found to be applicable for the characterization of NNMT inhibitors. By linking fragments of the AdoMet cofactor and nicotinamide substrate, a small library of bisubstrate compounds were prepared, designed to mimic the transition state of the NNMT-mediated methylation reaction. The inhibitory activity of these compounds was determined and the results revealed a new trivalent inhibitor with activity similar to that of the commonly used, general methyltransferase inhibitors sinefungin and S-adenosyl-L-homocysteine (AdoHcy). Docking studies suggest that the compound can bind the three main binding pockets known to contribute to the NNMT active site.