Integrating Machine Learning and Structure-Guided Discovery of a Novel Type II SYK Inhibitor for Treating Triple-Negative Breast Cancer

Spleen tyrosine kinase (SYK) is a cytoplasmic nonreceptor kinase involved in immune signaling and homologous recombination (HR) repair of DNA double-strand breaks, and its aberrant activation promotes therapeutic resistance in triple-negative breast cancer (TNBC). We developed a machine learning (ML)-corrected virtual screening strategy that reduces false positives in docking-based screening and identified a benzimidazole lead (L-5). SAR-guided optimization yielded a potent type II SYK inhibitor, 5i (IC 50 = 16 nM), which stabilizes the DFG-out inactive conformation via hinge, αC-helix, and DFG interactions. 5i shows strong antiproliferative, pro-apoptotic, and antimigratory effects in TNBC cells and suppresses tumor growth in an MDA-MB-231 xenograft model without overt toxicity. Mechanistically, 5i increases DNA damage by inhibiting SYK-mediated CtIP phosphorylation and shows synergy with the PARP inhibitor Olaparib. These findings establish 5i as a promising therapeutic candidate and demonstrate the potential of SYK inhibition as a strategy to overcome HR-mediated resistance in TNBC.