Proteases are vital for virus proliferation and have been found to play key roles in cancer metastasis. Identifying protease substrates allows determination of their function and can aid in inhibitor design. We recently reported a method of identifying substrates using cellular libraries of peptide substrates (CLiPS). Here we report improvement of this method to increase the sorting efficiency for protease substrates and high-throughput analysis of the individual substrates. We applied this method to matrix metalloprotease-1 (MMP-1) which is implicated in many diseases such as atherosclerosis and cancer. We identified an optimum substrate sequence consensus of P-(M/L/V)-(A/N/V)-(M/L/V) for residues in P3 to P1' positions. The consensus is similar to previously reported MMP-1 substrates with the most common residue of methionine in the P2 or P1' positions, but we found that optimum substrates contained one and only one methionine residue (e.g. P2 but not P1'). This demonstrates the need to investigate peptide substrates as a complete chain, since single point mutations and single position statistical analysis of substrates will not identify multi-residue interactions. Determining these higher order interactions will help understand selectivity of proteases and could aid in design of specific inhibitors.

Extended Abstract Status: Not Uploaded