Recently, the Insilico team’s research entitled “An Internal Sulfur-Lone Pair Interaction Enabled the Discovery of Potential and Sub-Family Selective PKMYT1 Inhibitors” was invited for publication as a cover Feature in ChemMedChem[1].
In the high-stakes arena of oncology, targeting the serine/threonine kinase (PKMYT1) has emerged as a premier strategy for treating aggressive cancers, particularly those characterized by CCNE1 amplification. PKMYT1 is a target that has achieved clinical proof of concept in the field of synthetic lethal cancer therapy. However, because the ATP‑binding site of kinases is highly conserved across the human genome, achieving selectivity at the kinase subfamily level remains a major challenge. Current clinical leads, such as the first-runner RP-6306 (RE1), have struggled with narrow selectivity margins—often less than 10-fold—against off-target kinases like BRAF, RAF1, and SRC. These off-target hits result in dose-limiting toxicities, such as severe skin rashes, and saturated plasma concentrations at relatively low doses. To solve this puzzle, Insilico Medicine have utilized their AI-powered generative chemistry platform, Chemistry42, to design a series of novel small molecular inhibitors and extremely selective PROTACs, and published multiple peer-reviewed scientific papers[2-5].
The core innovation of this work lies in the clever exploitation of noncovalent interactions. Through a conformational restriction strategy, the molecular spatial geometry is precisely controlled while hydrogen‑bond donors that could adversely affect physicochemical properties are effectively masked. Ultimately, the researchers replaced the pyrido‑pyrrole core scaffold of RE1 with a thiazolyl‑pyrazole ring system. In this new structure, the sulfur atom on the thiazole ring interacts with the lone pair of the nitrogen on the adjacent pyrazole ring. This forces the two five-membered heteroaromatic rings to maintain a "syn-locked" coplanar conformation, which is the ideal geometry for optimal binding within the PKMYT1 active site. For medicinal chemists, this marks a shift from traditional hydrogen-bonding or rigid cyclization strategies into new territory.
Compound A4 and its active enantiomer, A4‑ent1, exhibit highly promising properties:
- High potency & selectivity: A4-ent1 exhibits high activity for PKMYT1 IC50 = 2.2 nM ) and over 100-fold selectivity against the sub-family kinase WEE1.
- Significant anti-tumor potency: A4 effectively inhibits downstream CDK1 phosphorylation and shows strong antiproliferation in various CCNE1-amplified cancer cell lines (e.g., HCC1569, Ovcar3, and MKN1), while having minimal impact on non-CCNE1-amplified lines.
- Superior drug-like properties: Compared to RE1, the physicochemical properties of A4 are significantly improved. It possesses excellent Caco-2 permeability, higher solubility in pH 7.4 buffer (217 µM vs. 45 µM), and lower in vitro clearance in liver microsome stability tests.
The discovery of compound A4 proves that "underutilized" molecular forces like sulfur–lone pair interactions can outperform conventional binding motifs. Exquisite design strategy combining AI-driven scaffold hopping with diverse conformational restrictions have not only delivered high activity and selectivity, but also masked hydrogen‑bond donors that would otherwise negatively affect permeability and solubility, thereby improving the drug‑likeness of the molecule. Subtle non-bonding interactions embody the key to rational and precise molecular design.
Since its startup, Insilico has published over 200 peer-reviewed papers. This marks the sixth publication by the company in Nature Portfolio journals since 2024. Leveraging sustained scientific breakthroughs at the intersection of biotechnology, artificial intelligence, and automation, Insilico ranked Top 100 global corporate institutions in Nature Index’s “2025 Research Leaders: global corporate institutions for biological sciences and natural sciences publications”.
Harnessing state-of-the-art AI and automation technologies, Insilico has significantly improved the efficiency of preclinical drug development, setting a benchmark for AI-driven drug R&D.While traditional early-stage drug discovery typically requires 2.5 to 4 years, Insilico has nominated 20 preclinical candidates with an average timeline—from project initiation to preclinical candidate (PCC) nomination—of just 12 to 18 months per program, with only 60 to 200 molecules synthesized and tested in each program.
About Insilico Medicine
Insilico Medicine, a leading and global AI-driven biotech company, utilizes its proprietary Pharma.AI platform and cutting-stage automated laboratory to accelerate drug discovery and advance innovations in life sciences research. By integrating AI and automation technologies and deep in-house drug discovery capabilities, Insilico is delivering innovative drug solutions for unmet needs including fibrosis, oncology, immunology, pain, and obesity and metabolic disorders. Additionally, Insilico extends the reach of Pharma.AI across diverse industries, such as advanced materials, agriculture, nutritional products and veterinary medicine.
Reference：
[1] ChemMedChem 2026, 21 (6), e202501029. https://doi.org/10.1002/cmdc.202501029.
[2] J. Med. Chem. 2024, 67 (1), 420–432. https://doi.org/10.1021/acs.jmedchem.3c01476.
[3] Eur. J. Med. Chem. 2025, 281, 117025. https://doi.org/10.1016/j.ejmech.2024.117025.
[4] Bioorg. Med. Chem. 2026, 135, 118582. https://doi.org/10.1016/j.bmc.2026.118582.
[5] Nat. Commun. 2025, 16 (1), 10759. https://doi.org/10.1038/s41467-025-65796-8.