Skip to main content

Computer-aided Design and synthesis of inhibitors of EED-EZH2 interaction as a novel approach for anticancer therapy

Final Report Summary - HKMTIS (Computer-aided Design and synthesis of inhibitors of EED-EZH2 interaction as a novel approach for anticancer therapy)

The broad objective of this project was to discover small molecule inhibitors of histone lysine methyltransferases (HKMTs), which should provide potential therapeutics for cancer treatment. Our specific goal was to design inhibitors of an important HKMT, enhancer-of-zeste 2 (EZH2), which is one of the key components of PRC2 complex. EZH2 contains catalytic site in its SET domain which methylates lysine 27 of histone 3 (H2K27), a repressive epigenetic mark, and there is a large body of evidence suggesting EZH2 is an exciting and potentially important cancer target.

Our original strategy was to design inhibitors of protein-protein interactions between EZH2 and EED. The latter is another key component of PRC2 and its interaction with EZH2 was reported to be essential for the overall enzymatic activity of the PRC2 complex. However, a report was published by another research group at the beginning of this project where they had used an analogous strategy to that which we had proposed. We therefore changed our focus and revisited our goals, instead choosing to focus on direct enzymatic inhibitors rather than protein-protein interaction inhibitors.

We employed the 2,4-diaminoquinazoline class of compounds, such as BIX01294, which were reported to be substrate-competitive inhibitors G9a (EHMT2), another HKMT. In view of a high proposed similarity between the substrate binding pocket of G9a and EZH2, we hypothesized that 2,4-diaminoquinazoline scaffold could be exploited for the discovery of (the first) substrate-competitive inhibitors of EZH2. Hence, our objective was to to synthesize a library of 2,4-diaminoquinazolines, supported by computational design where appropriate, and to evaluate the compounds in enzymatic and cell based assays. Pleasingly, our efforts resulted in the identification of EZH2 inhibitors with IC50 values up to 1 µM, as judged by the enzymatic assay. In accordance with the design rationale, and unlike other known co-factor site binding EZH2 inhibitors, our molecules were found to be substrate-competitive, thus adding novelty to this project. In cell based assays the active molecules were also found to increase the m-RNA levels of genes known to be silenced by EZH2; as may be expected from inhibition of an HKMT which installs a repressive epigenetic mark. Finally, one of the tested analogues exhibited activity in a MDA-MB- 231 (breast cancer) xenograft mouse model.

HKMTs are conserved and important in other species. Building on prior work in the group, we also pursued the 2,4-diaminoquinazoline scaffold to target the HKMTs of the malarial parasite Plasmodium. In Plasmodium, epigenetic gene regulation mediated by histone-modifying enzymes has been shown to play a key role at various stages of its life-cycle, including the control of virulence genes involved in immune evasion. So far Plasmodium HKMTs have not been isolated and thus, the molecules were evaluated for parasite-killing activity initially. Interestingly, several analogues based on the 2,4-diaminoquinazoline scaffold were found to possess potent and rapid parasite-killing activity against erythrocytic and transmission stages of parasite life-cycle. Additionally, the most efficacious analogues were shown to reduce the amount of methylated histones in the treated parasites, giving evidence for direct inhibition of Plasmodium HKMTs by the synthesised compounds. Most importantly, these analogues were also found to be effective against multiple malaria causing species and artemisinin-resistant field isolates. Thus, our medicinal chemistry and pharmacology work has helped to disclose detailed SAR of 2,4-diaminoquinazolines antimalarial activity and positioned this scaffold as an efficient antimalarial class targeting malarial HKMTs.

Finally, we also wanted to discover novel non-diaminoquinazoline scaffold that can be exploited to target human HKMTs, G9a and EZH2. To realize this goal, we employed a virtual screening (VS) approach, which is a computational substitute of physical high throughput screening and widely used in drug discovery. Thus, a field-based pharmacophore (Cresset) derived from the G9a-bound co-crystallized conformation of a 2,4-diaminoquinazoline analogue was used to screen a selection of the ZINC database. The top ranked hits belonging to diverse chemotypes were handpicked for in vitro assays against G9a and Plasmodium HKMTs. Interestingly, the analogues based on the newly identified scaffold were found to exhibit nano-molar potency against both G9a and the malarial parasite, thus, validating the computational hypothesis. Further study of these compounds is underway.

In summary, the current project has resulted in new ‘lead’ molecules as human and malarial HKMT inhibitors. In long term, these molecules have potential of yielding new drugs for the treatment of cancer and/or malaria and may have immense positive impact on socio-economics of European Union. Cancer and malaria are responsible for a large number of deaths across the globe. Despite the availability of other drugs against these diseases, there is need for more efficacious, safe and cheaper therapies. Based on the current data generated from this project, we have secured additional funding for the further characterization and development of this class of molecules.