Protein-protein interactions (PPIs) form the basis of cellular regulation with virtually every protein involved in mutually exclusive binding relationships. With each protein involved in modulating a given pathway, the nodal, interfacial protein therefore represents the regulatory node controlling pathway crosstalk. Interestingly, therapeutically relevant interfacial point mutants are in several cases involved in causing a gain of function by abrogating or weakening a given PPI. We believe that by modulating these competitive regulatory networks using small molecules we can identify novel therapeutics. To this end, we have developed a computational database that identifies mutually exclusive interactions by mining all protein interactions in the PDB, along with matching all known protein domains with domains in the PDB. This computational database provides structural rationale to interactions recovered by BioGRID but lacking structural information. In addition, we identified several novel interactions absent from BioGRID or literature-wide aggregation. As a proof of principle we explored mothers against decapentaplegic (SMADs) proteins where receptor SMADs (R-SMADs) compete for the nodal SMAD4 protein to control functional outputs related to TGFβ and BMP signaling. By overexpressing, knocking-down, and modulating these systems using small molecules, we demonstrate how structurally induced competition leads to meaningful remodulation of competing functional pathways as measured by luciferase-reporter, immunofluorescence and viability assays. Finally, we define the rationale for supplementing our established assays with single molecule tracking (SMT) to characterize weak and transient interactions that if stabilized or inhibited could provide a unique lever toward modulating a pathway of interest in a therapeutically advantageous manner.
Driouchi et al. (Sun,) studied this question.