Structure Dynamics Guides the Enhancement of Ligand Affinity for MdmX

Abstract

Mdm2 and MdmX are highly homologous and Mdm2 inhibitors exhibit weak affinity for MdmX. The structural mechanism underlying how Mdm2/MdmX distinguish ligand binding affinity remains elusive. Guided by NMR study, we have identified three flexible regions on N-MdmX that affected inhibitor-binding affinity. Furthermore, using disulfide staple, we found that the Phe19-binding site on MdmX was one of the most flexible regions weakening nutlin-3a binding. On the other hand, we continued to pursue high-resolution structure of Mdm2 in complex with ntutin-3a. An X-ray crystallographic structure of Mdm2/nutlin-3a at 1.10 Å identified that His72 had two conformations, a bound state and an unbound state. Molecular dynamics simulation also revealed that His72 in MdmX existed two states. Thus, our work started to search fragments that could rigidify Phe19-binding site. We designed a fusion protein model that p5315-29 was fused with MdmX through a flexible linker, i.e., p5315-29-MdmX, for fragment screening. By use of the intrinsic fluorescence signal of p5315-29 to monitor high-throughput screening, we found that the imidazoindole was a bona fide scaffold for template-based design of MdmX inhibitors. Next the p5315-29-MdmX fusion protein was modified to leave Phe19-binding site unoccupied, i.e., p5323-29-MdmX. By screening a fragment library, a small fragment, CPD-01, could bind to Phe19-binding site in an allosteric mode that enhanced the binding of p5323-29 to MdmX. Furthermore, CPD-01 was conjugated with p5323-29 to synthesize p53 peptide analogues with different linkers. These peptide analogues exhibited enhanced binding affinity for MdmX, compared with p5323-29. Then, CPD-01 was integrated into nutlin-3a to replace Phe19-binding-site-pharmacophore. The new nutlin analogue exhibited high-affinity binding MdmX and Mdm2 as well. Taken together, our work provided a promising strategy to design inhibitors of MdmX/Mdm2. This strategy should be applicable for lead optimization in drug discovery.