Molecular investigation of the dual inhibition mechanism for targeted P53 regulator MDM2/MDMX inhibitors

Abstract

Inhibitors that competitively bind MDM2/MDMX can block the inhibition of P53 by MDM2/MDMX and restart its tumor-suppressive effect. Molecular studies targeting MDM2/MDMX inhibitors have always been a hot topic in anticancer drug design. Although numerous inhibitors have been designed previously against MDM2/MDMX, their dual inhibition efficacy has not been demonstrated, and few studies assessed the general causes affecting the dual inhibition of MDM2/MDMX by these inhibitors. Here, molecular dynamics simulations and alanine scanning combined with the interaction entropy method were employed to precisely investigate whether 16 inhibitors could dually inhibit MDM2/MDMX and the similarities and differences in the interaction modes. Thereby addressing the key residue sites affecting dual inhibition. Residues L54/M53, I61/60, M62/61, Y67/66, and V93/92 of MDM2/MDMX, which are in corresponding positions in both protein structures, provide significant conditions for these inhibitors to bind to MDM2/MDMX tightly. In addition, most of these inhibitors prefer to bind MDM2 than MDMX, and residues H96 and I99 in MDM2 are attractive targets for inhibitors, resulting in inhibitors binding to MDM2/MDMX with different affinity. These key residues should be considered in the development of dual inhibitors. For these 16 inhibitors, most have dual inhibitory potential for MDM2/MDMX based on the binding affinity of the complexes. Still, it is questionable whether they can exert excellent dual inhibition considering the assessment of the hot-spots. At least their binding affinity for MDMX is not superior to that for MDM2 due to the difference in energy of the van der Waals interactions at the key sites. Furthermore, based on the analysis of three representative inhibitors (TUZ/HRH and HRQ with different binding preferences for MDM2/MDMX), 3-chloropyridine in TUZ leads to the differential binding affinity between the inhibitor and MDM2/MDMX. It readily forms hydrophobic interactions with the surrounding residues H96 and I99. But this phenomenon does not occur in the TUZ–MDMX system, implying the critical role of residues H96/P95 and I99/L98. And the completely different binding mechanism of HRQ binding to MDM2/MDMX explains its inability to inhibit MDM2 well. Thus, we are cautious about its dual inhibitory ability. Besides, HRH is more prone to strong van der Waals interactions with MDM2 than MDMX whereas its 2-chlorofluorobenzene is detrimental to this. We hope that these findings will provide reliable molecular insights for the screening and optimization of targeting MDM2/MDMX dual inhibitors.