Identification of a novel set of scaffolding residues that are instrumental for the inhibitory property of Kunitz (STI) inhibitors.

For canonical serine protease inhibitors (SPIs), scaffolding spacer residue Asn or Arg religates cleaved scissile peptide bond to offer efficient inhibition. However, several designed "mini-proteins," containing the inhibitory loop and the spacer(s) with trimmed scaffold behave like substrates, indicating that scaffolding region beyond the spacer is also important in the inhibitory process. To understand the loop-scaffold compatibility, we prepared three chimeric proteins ECI(L)-WCI(S), ETI(L)-WCI(S), and STI(L)-WCI(S), where the inhibitory loop of ECI, ETI, and STI is placed on the scaffold of their homolog WCI. Results show that although ECI(L)-WCI(S) and STI(L)-WCI(S) behave like good inhibitors, ETI(L)-WCI(S) behaves like a substrate. That means a set of loop residues (SRLRSAFI), offering strong trypsin inhibition in ETI, act as a substrate when they seat on the scaffold of WCI. Crystal structure of ETI(L)-WCI(S) shows that the inhibitory loop is of noncanonical conformation. We identified three novel scaffolding residues Trp88, Arg74, and Tyr113 in ETI that act as barrier to confine the inhibitory loop to canonical conformation. Absence of this barrier in the scaffold of WCI makes the inhibitory loop flexible in ETI(L)-WCI(S) leading to a loss of canonical conformation, explaining its substrate-like behavior. Incorporation of this barrier back in ETI(L)-WCI(S) through mutations increases its inhibitory power, supporting our proposition. Our study provides structural evidence for the contribution of remote scaffolding residues in the inhibitory process of canonical SPIs. Additionally, we rationalize why the loop-scaffold swapping is not permitted even among the members of highly homologous inhibitors, which might be important in the light of inhibitor design.