Clinical stage generative artificial intelligence (AI)-driven biotechnology company InSilico Medicine (“InSilico”), today announced that the Journal of Medicinal Chemistry, an academic journal focusing on critical studies about molecular structure and biological activity, has published the company’s discovery of a novel PHD enzyme inhibitor for the treatment of anemia. The academic breakthrough is powered by Chemistry42, its proprietary generative chemistry platform consisting of more than 40 selected generative models. As suggested in previous studies, the inhibition of prolyl hydroxylase domain enzymes (PHD) influences fundamental biological processes, including red blood cell production by regulating the Nobel prize-winning HIF-α pathway, thus indicating potential for the treatment of anemia in chronic kidney disease (CKD).
Guided by a structure-based drug discovery (SBDD) strategy, scientists gathered structure information on the PHD target and known molecules, and generated series of molecule. Utilizing built-in filters covering drug-likeness, pharmacophore clues and more, the AI-generated candidates were ranked and prioritized before a hit compound was produced for further optimization. Afterward, several rounds of synthesis test optimization yielded lead compound 15, which demonstrated a favorable in vitro/in vivo ADMET profile, a clean safety profile and promising pharmacokinetic properties in multiple species. Moreover, the compound required relatively simple synthesis steps and was proven to alleviate anemia in a rat disease model. Given that more than 10% of the global population suffers from CKD, this new molecule could be a meaningful for further investigations.
But it’s not the only sector where AI is being investigated by these scientists: oncology can also benefit from whatever help is available. Across the world, breast and gynecological cancers pose serious threats to women’s health, fertility, and overall quality of life. In order to identify potential targets for new therapeutics, the research team leverage Insilico’s proprietary AI-driven target identification platform, PandaOmics, to analyze data of five forms of gynecological cancers, including breast cancer particularly triple-negative breast cancer. Remarkably, MYT1 consistently ranked at the forefront across all diseases in terms of relevance. MYT1 is a protein kinase rarely expressed in most normal tissues but highly expressed in most cancers. It has been reported that MYT1 inhibition and CCNE1 gene amplification, a condition known as synthetic lethality, play crucial functions in cell cycle regulation,
This indicates MYT1 inhibition is a promising therapeutic strategy for the treatment of cancers with genome instability. Using the usual SBDD strategies and applying rigorous filters for similarity and selectivity, Insilico researchers designed an array of compounds targeting MYT1 from scratch. Amongst these novel compounds, one series emerged as hit compounds. Scientists then conducted X-ray crystal structure analysis of the complex and found significant impact on the activity of subtle chemical structure modifications. This knowledge provided guidance for further optimization, leading to the discovery of the hit molecule, Compound 21. This presents good MYT1 activity and excellent selectivity over Wee1 kinase and the other kinase panel, that reduces the potential risk for off-target effects and might translate to a safer profile.
In preclinical studies, the drug shows potent in vivo antitumor effect and a promising profile in pharmacokinetic and pharmacodynamics.
- Edited by Dr. Gianfrancesco Cormaci, PhD, specialist in Clinical Biochemistry.
Scientific references
Xu J, Ding X, Fu Y et al. J Med Chem 2024 Jan 8; in press.
Wang S, Wang C et al. J Med Chem 2023; 67(1):420-432.
Zuk A, Si Z et al. J Pharmacol Exp Ther. 2022; 383(1):11-24.
Chen H et al. J Clin Pharm Ther. 2021 Aug; 46(4):999-1009.