SH3 Domain

Src homology 3 (SH3) domain, one of the most abundant protein modules found in eukaryotes, was found to specifically recognize proline-rich sequences adopting polyproline-II helices. Whereas there were 28 SH3 domains found in yeast, the human proteome is estimated to contain about 400 copies of such domains spread among a diverse array of proteins. SH3 domains contain approximately 60 residues and share a common structure fold featuring five-stranded anti-parallel beta-sheet. This domain was shown to recognize a core PXXP motif which adopts a polyproline-II helix. The SH3 domains were also found in extracellular proteins, including melanoma inhibitory activity (MIA) protein, which binds to collagen homology-2 domain of p66shc, and other EM proteins. Early analysis showed that at least 25% of human proteins contain proline-reach regions, opening a question how the hundreds of SH3 domains select their respective binding partner to execute their specific function.

Although several co-crystal structures are available for protein complexes involving SH3 domains, understanding the molecular basis of their binding specificity appeared to be important but very challenging task. This problem has been approached by several ways, including yeast proteomic scale analysis by phage display and yeast two-hybrid screening. Most of these results are still not well understood, including the ability of SH3 domains to recognize a variety of sequences that do not conform to the conventional motifs. This is a new area for me, and I will need to collect preliminary data before applying for grants on this topic. This problem will be approached similarly to what I did for studying ZF protein – DNA specificity and my experience with structure analysis of proline-rich protein motifs should also help here. Computational approaches will be applied to predict binding specificity, which will be experimentally tested on peptides by binding assays and thermodynamically characterized by isothermal titration calorimetry.