HOMOLOGY MODELING OF THE LACTOCOCCUS-LACTIS LEADER PEPTIDASE NISP ANDITS INTERACTION WITH THE PRECURSOR OF THE LANTIBIOTIC NISIN

A model is presented for the 3-D structure of the catalytic domain of the putative leader peptidase NisP of Lactococcus lactis, and the inte raction with its specific substrate, the precursor of the lantibiotic nisin. This homology model is based on the crystal structures of subti lisin BPN' and thermitase in complex with the inhibitor eglin. Predict ions are made of the general protein fold, inserted loops, Ca2+ bindin g sites, aromatic interactions and electrostatic interactions of NisP. Cleavage of the leader peptide from precursor nisin by NisP is the la st step in maturation of nisin. A detailed prediction of the substrate binding site attempts to explain the basis of specificity of NisP for precursor nisin. Specific acidic residues in the S1 subsite of the su bstrate binding region of NisP appear to be of particular importance f or electrostatic interaction with the P1 Arg residue of precursor nisi n after which cleavage occurs. The hydrophobic S4 subsite of NisP may also contribute to substrate binding as it does in subtilisins. Predic tions of enzyme-substrate interaction were tested by protein engineeri ng of precursor nisin and determining susceptibility of mutant precurs ors to cleavage by NisP. An unusual property of NisP predicted from th is catalytic domain model is a surface patch near the substrate bindin g region which is extremely rich in aromatic residues. It may be invol ved in binding to the cell membrane or to hydrophobic membrane protein s, or it may serve as the recognition and binding region for the modif ied, hydrophobic C-terminal segment of precursor nisin. Similar predic tions for the tertiary structure and substrate binding are made for th e highly homologous protein EpiP, the putative leader peptidase for th e lantibiotic epidermin from Staphylococcus epidermidis, but EpiP lack s the aromatic patch. Based on these models, protein engineering can b e employed not only to test the predicted enzyme-substrate interaction s, but also to design lantibiotic leader peptidases with a desired spe cificity.