Discovery of novel ULK1 inhibitors by virtual screening, synthesis, in vitro assay and molecular dynamics simulations

The serine/threonine kinase UNC-51-like autophagy-activating kinase 1 (ULK1) has emerged as a promising target for cancer treatment. This study utilized an integrative approach combining computational and experimental methods to discover novel ULK1 inhibitors. Through molecular fingerprint similarity and shape-based screenings of the ChEMBL database, along with fragment growth modifications of DCC-3116, three datasets comprising a total of 1.5 million compounds were generated. A multi-level molecular docking workflow shortlisted 1520 compounds, from which 9 promising candidates (SX1-SX3, D1-D5, and CL130) were identified based on their strong binding modes, favorable binding free energies, and desirable ADMET properties. These candidate compounds were synthesized and demonstrated nanomolar inhibitory activity in the ULK1 ADP-Glo kinase assay. Notably, D1-D3 exhibited significantly higher inhibitory potency compared to the reference SBI-0206965 (IC₅₀ = 38.19 nM), with IC₅₀ values of 14.91 nM, 0.74 nM, and 1.06 nM, respectively. The anti-proliferation assay of D1 was conducted on HeLa cells, yielding an IC₅₀ value of 0.83 μM. Molecular dynamics simulations further confirmed the stability of these compounds within the ULK1 binding pocket, while principal component analysis and dynamic cross-correlation matrix analyses revealed distinct conformational and binding behaviors. Binding free energy calculations indicated that the interactions between the candidate compounds and ULK1 were either stronger or comparable to those of SBI-0206965. Overall, these results suggest that SX1-SX3, and D1-D5 were promising ULK1 inhibitors, providing a solid foundation for further development as potential anticancer drugs.