Structural basis for the binding of proline-rich peptides to SH3 domains.

A common RXL motif was found in proline-rich ligands that were selected from a biased combinatorial peptide library on the basis of their ability to bind specifically to the SH3 domains from phosphatidylinositol 3-kinase (PI3K) or c-Src. The solution structure of the PI3K SH3 domain complexed to one of these ligands, RKLPPRPSK (RLP1), was determined. Structure-based mutations were introduced into the PI3K SH3 domain and the RLP1 ligand, and the influence of these mutations on binding was evaluated. We conclude that SH3 domains recognize proline-rich motifs possessing the left-handed type II polyproline (PPII) helix conformation. Two proline residues directly contact the receptor. Other prolines in the ligands appear to function as a molecular scaffold, promoting the formation of the PPII helix. Three nonproline residues consisting of combinations of arginine and leucine interact extensively with the SH3 domain and appear to confer ligand specificity.