Discovery of KAT6A degrader enabled by structural proteomics and AI

Lysine acetyltransferase 6 (KAT6A/B) is a MYST family member of histone acetyl transferases (HATs), with elevated activity across different cancer types. Due to its driving role in activation of hormonal signaling, KAT6 is a promising drug target specifically in estrogen receptor positive (ER+) breast cancer. Indeed, efficacy has been demonstrated in clinical trials investigating KAT6 inhibition in ER+ breast cancer. Proteolysis-targeted chimeras (PROTACs) are heterobifunctional molecules inducing the proximity between a target protein and an E3 ligase (e.g. CRBN), leading to target ubiquitination and subsequent degradation. KAT6 PROTAC can target both its enzymatic and scaffold functionalities, potentially leading to improved efficacy. In addition, selectivity towards KAT6A could increase drug tolerability and reduce adverse events.In the current work, we present a novel KAT6A-selective PROTAC, developed using the AIMS™ proteomics and AI platform. Using structural mass spectrometry (MS) data and AI modeling, we generated confident structures of the KAT6-E3-PROTAC ternary complex and identified PROTAC-dependent conformations, enabling accurate modeling of the complex for further optimization. The resulting PAI-PROTAC demonstrates efficient KAT6A degradation and growth inhibition in ER+ breast cancer cell lines, with high selectivity over KAT6B and over other KAT enzymes, as well as no degradation activity in CRBN neosubstrates. Altogether, these results illustrate how proteomics-aware-AI enables rational PROTAC design, leading to discovery of a potent and selective KAT6A degrader.