Identification of a potent peptide deformylase inhibitor from a rationally designed combinatorial library

Peptide deformylase catalyzes the removal of the N-terminal formyl group from nascent polypeptides during prokaryotic protein synthesis and maturation and is essential for bacterial survival. Its apparent absence from mammalian organisms makes it an attractive target for designing novel antibacterial agents. Based on the substrate specificity of peptide deformylase from Escherichia coli, a focused library of peptide thiols was synthesized on TentaGel resin using a disulfide linkage. Screening of the library against the purified deformylase was carried out in solution phase after the inhibitors were released from the resin with a reducing agent. A potent deformylase inhibitor was obtained from a 750-member library and was further optimized through rational modification into a low nanomolar inhibitor (KI = 15 nM against E. coli deformylase).

Substrate recognition through a PDZ domain in tail-specific protease

Tail-specific protease (Tsp) is a periplasmic enzyme that selectively degrades proteins bearing a nonpolar C-terminus. Its substrate specificity suggests that Tsp may contain a substrate recognition domain, which selectively binds to the nonpolar C-termini of substrate proteins, separate from its catalytic site. In this work, we show that substrate recognition of Tsp is mediated by a PDZ domain, a small protein module that promotes protein-protein interactions by binding to internal or C-terminal sequences of their partner proteins. Partial proteolysis by V8 protease at a single peptide bond immediately N-terminal to the PDZ domain resulted in two distinct and relatively stable fragments and complete loss of catalytic activity. Photoaffinity labeling with a fluorescent nonpolar peptide caused the covalent attachment of the peptide to a single site on the Tsp protein. Systematic deletion mutagenesis of Tsp localized the binding site to amino acids 206-307, a region that completely encompasses the putative PDZ domain (217-301). The isolated PDZ domain (amino acids 206-334) is capable of folding into a well-behaved structure and binds to a nonpolar peptide with a dissociation constant (K(D)) of 1.9 microM, similar to that of the intact Tsp protein. Site-directed mutagenesis of a surface residue at the peptide binding site of the PDZ domain, valine 229, to Glu or Gln resulted in an increase in the K(M) value but had no effect on the k(cat) value. The use of a separate substrate recognition domain such as a PDZ domain may be a general mechanism for achieving selective protein degradation.