Discovery of a Novel and Potent LCK Inhibitor for Leukemia Treatment via Deep Learning and Molecular Docking.

The lymphocyte-specific protein tyrosine kinase (LCK) plays a crucial role in both T-cell development and activation. Dysregulation of LCK signaling has been demonstrated to drive the oncogenesis of T-cell acute lymphoblastic leukemia (T-ALL), thus providing a therapeutic target for leukemia treatment. In this study, we introduced a sophisticated virtual screening strategy combined with biological evaluations to discover potent LCK inhibitors. Our initial approach involved utilizing the PLANET algorithm to assess and contrast various scoring methodologies suitable for LCK inhibitor screening. After effectively evaluating PLANET, we progressed to devise a virtual screening workflow that synergistically combines the strengths of PLANET with the capabilities of Schrödinger's suite. This integrative strategy led to the efficient identification of four potential LCK inhibitors. Among them, compound 1232030-35-1 stood out as the most promising candidate with an IC of 0.43 nM. Further bioassays revealed that 1232030-35-1 exhibited robust antiproliferative effects on T-ALL cells, which was attributed to its ability to suppress the phosphorylations of key molecules in the LCK signaling pathway. More importantly, 1232030-35-1 treatment demonstrated profound antileukemia efficacy in a human T-ALL xenograft model. In addition, complementary molecular dynamics simulations provided deeper insight into the binding kinetics between 1232030-35-1 and LCK, highlighting the formation of a hydrogen bond with Met319. Collectively, our study established a robust and effective screening strategy that integrates AI-driven and conventional methodologies for the identification of LCK inhibitors, positioning 1232030-35-1 as a highly promising and novel drug-like candidate for potential applications in treating T-ALL.