Computational and experimental screening identifies 6-chloro-2-hydroxy-1-naphthaldehyde derivatives as disruptors of the HEG1–KRIT1 protein–protein interaction

The HEG1–KRIT1 protein–protein interaction is a central regulatory interface in cerebral cavernous malformation signaling and represents a potential target for small-molecule modulation. In this study, a focused library of 1,121 compounds containing a 2-hydroxy-1-naphthaldehyde scaffold was screened to identify candidate HEG1–KRIT1 disruptors. Compounds were prioritized through drug-likeness filtering, PAINS/reactivity assessment, diversity and purchasability evaluation, AutoDock Vina docking, short molecular dynamics refinement, and MM/PBSA rescoring. Three candidates, HN1, HN2, and HN3, were selected for 1000 ns molecular dynamics simulations, free-energy analyses, in silico ADMET profiling, and experimental evaluation using a bead-based flow-cytometry HEG1–KRIT1 competition assay. Docking and simulation analyses suggested that all three candidates could occupy the HEG1-recognition region of the KRIT1 FERM domain and maintain ligand–protein contacts during the simulations. Among the tested compounds, HN1 yielded a modest positive ΔΔG PPI estimate of + 2.66 kcal/mol, whereas HKi2, HN2, and HN3 yielded negative estimates; these values were interpreted descriptively as approximate indicators of ligand-dependent changes in the HEG1–KRIT1 interface. Experimental testing showed concentration-dependent inhibition of EGFP–KRIT1 FERM binding to HEG1-coated beads. Under the applied assay conditions, HN1 produced the numerically lowest IC₅₀ estimate at 3.51 ± 0.14 µM, compared with 3.90 ± 0.13 µM for HKi2, while HN2 and HN3 produced IC₅₀ estimates of 24.12 ± 1.59 and 31.63 ± 1.99 µM, respectively. These numerical differences do not establish superiority or equivalence among the compounds. Collectively, the computational and experimental findings support HN1 as a preliminary hit that warrants further target-engagement, mechanistic, and cellular validation.