QM/MM-based Rational Design of an Orthogonal X∙∙∙O∙∙∙H Interaction System at the Complex Interface of EGFR Kinase Domain with RALT Third Hotspot Peptide

Human epidermal growth factor receptor (EGFR) has been established as a primary druggable target for the treatment of patients with diverse cancers. The receptor-associated late transducer (RALT) protein is a negative regulator of EGFR signaling by using its Seg1 segment to bind at the allosteric dimerization interface of EGFR kinase domain to disrupt the dimerization and then inactivate the kinase. Previous studies found that the Seg1 adopts three separated hotspots to interact with the EGFR dimerization interface, in which the third hotspot (TH) is located at the core region of the interface but its derived TH peptide (356PKYVS360) and Tyr358Phe mutant (356PKPVS360) cannot bind effectively to the interface in an independent manner (KD > 1000 μM). In this study, we demonstrate that the TH peptide can be converted from nonbinder to a moderate binder of EGFR by just adding an orthogonal nonbonded interaction system (X∙∙∙O∙∙∙H) between a halogen bond and a hydrogen bond involving peptide Phe358 residue and EGFR Val948/Trp951 residues. Both the hybrid quantum mechanics/molecular mechanics (QM/MM) and electron-correlation quantum chemistry calculations are utilized to rigorously characterize and rationally design the X∙∙∙O∙∙∙H system, which is then optimized with different halogen atoms and at different substituting positions. It is revealed that there is a synergistic effect between the halogen bond (X∙∙∙O) and hydrogen bond (H∙∙∙O) of the orthogonal interaction system; formation of the halogen bond can enhance the interaction strength of the hydrogen bond. In silico energetic analysis and in vitro fluorescence assay reach a consistence that Br-substitution at the m-position of peptide Phe358 phenyl moiety is the best choice that can render strong interaction for the X∙∙∙O∙∙∙H system (ΔEX•••O = -3.48 kcal/mol for X∙∙∙O and ΔEH•••O = -2.57 kcal/mol for H∙∙∙O), which also makes the peptide ‘bindable’ to EGFR kinase domain (KD = 247 μM), while F/Cl/I-substitution at the same position can only improve the peptide affinity moderately or modestly (KD = 835/582/690 μM). In contrast, the Br-substitution at the o- and p-positions of peptide Phe358 phenyl moiety cannot define effective X∙∙∙O∙∙∙H interaction and thus does not confer additional affinity to the EGFR–peptide complex (KD > 1000 μM).