Structure-Based Design of Apogossypol-Derived Ligands Targeting the PELO–HBS1L Interface for Cancer Therapy

Background The PELO–HBS1L complex is a critical mediator of ribosome-associated quality control, responsible for resolving stalled ribosomes and maintaining translational fidelity. Recent CRISPR-based synthetic lethality screens have identified PELO and HBS1L as selective vulnerabilities in cancers with 9p21.3 deletion or microsatellite instability-high (MSI-H) phenotypes. These cancers exhibit dependency on the PELO–HBS1L complex due to coexisting destabilization of the superkiller complex (SKIc), presenting a unique opportunity for targeted therapeutic intervention. However, no small-molecule inhibitors of this complex have been reported. Methods We employed AlphaFold2-multimer modeling to generate high-confidence structural models of the PELO–HBS1L heterodimer. Structural evaluation included predicted aligned error (PAE), per-residue confidence (pLDDT), and MSA coverage. Fragment-based pocket mapping identified a candidate binding pocket, selected for docking based on accessibility and proximity to inter-chain contacts. A fragment-guided design strategy was implemented using apogossypol scaffolds bearing hydroxyamide linkers and terminal polar groups. Ligands were prepared via Open Babel and docked into the interface pocket using AutoDock Vina. Binding energy, interaction geometry, and contact distances were used to rank binding poses. Results AlphaFold2 models revealed a rigid PELO core and a flexible HBS1L tail, with PAE and pLDDT scores indicating moderate interface confidence and ligand-accessible surface features. Indole-2-carboxylic acid with guanidine substitution formed key polar contacts in the candidate binding pocket. Among apogossypol analogs, methylamine and guanidine-tailed derivatives displayed strong polar interactions at the pocket interface, including engagement with GLU215, ASN, and THR. A methylated guanidine analog further enhanced interaction density, while piperidine substitutions were poorly tolerated. Conclusion A therapeutic small molecule docked in a binding pocket at the interface of two interacting proteins can disrupt a protein-protein complex. This approach, known as protein-protein interaction (PPI) inhibition, is a well-established strategy in drug discovery. The PELO–HBS1L interface contains a druggable surface pocket that can be targeted by small molecules. Apogossypol-derived ligands with flexible, cationic extensions demonstrate promising interaction profiles and provide a foundation for further hit-to-lead optimization. These findings support development of interface-directed inhibitors to exploit synthetic lethality in PELO-dependent cancers.