Identification of SLiMs: Mapping and characterizing motif-based protein interactions

Sammanfattning: During the last twenty years it has become evident that about 35-40% of amino acids in the proteome are in regions that have evolved to remain unstructured. These intrinsically disordered regions contain short linear motifs (SLiMs), which serve as docking sites for protein-protein interactions. SLiMs often mediate low-to-medium affinity interactions that are transient in their nature. The characteristics of SLiM-based interactions make them difficult to be captured using conventional approaches like affinity-purification coupled to mass spectrometry or yeast-two-hybrid. We therefore used and developed a dedicated method for large-scale screening of SLiM-based interactions termed proteomic peptide phage display (ProP-PD).Using ProP-PD, We identified large sets of ligands, for the binding pocket of shank1 PDZ domain, containing C-terminal or internal binding motifs and established the consensus motifs to be xTxL/F-COOH and xTxFx respectively. We further validated interactions using biophysical affinity determinations and pulldown experiments. Using X-ray crystallization, we uncovered that shank1 PDZ binds to internal xTxFx motifs using a binding mode similar to that for C-terminal peptides.Adding a level of complexity, we explored interactions of the multiple binding pocket containing FERM domains from four closely related proteins: ezrin, radixin, moesin and merlin. We found hundreds of FERM ligands, which contained binding motifs of at least four different classes. By combining docking simulations with experiments, we established ligands binding to different pockets, and uncovered a complex interplay between distinct pockets.We further developed an optimized version of a phage library that displays intrinsically disordered regions of the human proteome. We benchmarked the library using a set of protein domains and reported better recovery of known SLiM-based interactions. Furthermore, we highlighted the functional aspects of identified SLiMs, in the case of nuclear localization signals, found for binding to importin-subunit alpha-3. Finally, we validated predicted binding of SLiMs in the Sars-CoV-2 host receptor ACE2, which illustrates the importance of fundamental knowledge for SLiMs and their binding partners.This work, taken together, contributes with method development for expansion of motifs based interactomes and provide insights into the plastic yet selective nature of peptide binding proteins.