Development of ligands for the TRIM33 bromodomain

Epigenetic information is encoded in a wide range of histone modifications, which gives rise to the “histone code” and brings an additional level of complexity to the control of gene expression. The proteins involved include “writers”, “erasers”, and “readers”, which install, remove, and recognise histone modifications, respectively. Given the involvement of epigenetics in diseases, including cancer, epigenetic regulation presents opportunities for the development of novel therapeutics. Understanding the roles of proteins involved in epigenetic regulation is, therefore, important. TRIM33 is a transcriptional regulator with E3 ubiquitin ligase activity that has been associated with leukaemia and DNA damage response pathways. It contains a tandem PHD finger-bromodomain cassette, which recognises methylated and acetylated lysine residues on histone proteins. The aim of this project was to develop ligands for the TRIM33 bromodomain, which can be used as chemical tools to further validate the role of TRIM33. This work describes the identification of hits in an initial AlphaScreen™ assay, which were then validated and used as the starting point for further development. Compound D664-0002 (Figure 1B) was identified as a promising compound and further studies focused on this. Iterations of synthesis and evaluation have enabled the exploration of structure-activity relationships (SAR) around the benzimidazolone core. A combination of biological assay techniques and computational studies have been used to guide investigation of the acetyl-lysine binding pocket. Interactions of the amine chain with glutamic acid resides E981 and E984 were identified as important for ligand design. Variation of the amine chain showed a relatively flat SAR with most ligands demonstrating low micromolar affinity, likely due to the solvent-exposed pocket. Computational studies also suggested the presence of structural water molecules in the ZA channel as a key consideration in understanding the precise binding mode and consequently, future design of a high-affinity ligand. The potential of gaining affinity via interaction with the region above the acetyl-lysine binding pocket was explored. This was achieved using an extension of the headgroup, which established the optimum vector combination between the extension and the amine chain. However, biological evaluation of compounds demonstrated that extensions larger than a methyl or ethyl group were not tolerated. Covalent analogues of the TRIM33 bromodomain ligand were explored for their ability to alkylate a mutant variant TRIM33β-E984C, as a potential strategy to enable crystallography. Analogues containing an electrophile were designed and synthesised. Alkylation experiments with TRIM33β-E984C indicated selective alkylation of the mutated cysteine residue. This result remains to be confirmed through mass spectrometry studies. The exploitation of a cysteine residue to enable crystallography will increase our understanding of the TRIM33β bromodomain pocket. In particular, understanding the ligand binding mode in relation to the ZA channel water molecules will further guide the design of a high-affinity ligand.