Proteochemometrics analysis of substrate interactions with dengue virus NS3 proteases

Abstract

The prime side specificity of dengue protease substrates was investigated by use of proteochemometrics, a technology for drug target interaction analysis. A set of 48 internally quenched peptides were designed using statistical molecular design (SMD) and assayed with proteases of four subtypes of dengue virus (DEN-1-4) for Michaelis (Km) and cleavage rate constants (kcat). The data were subjected to proteochemometrics modeling, concomitantly modeling all peptides on all the four dengue proteases, which yielded highly predictive models for both activities. Detailed analysis of the models then showed that considerably differing physico-chemical properties of amino acids contribute independently to the Kmand kcatactivities. For kcat, only P1′ and P2′ prime side residues were important, while for Kmall four prime side residues, P1′-P4′, were important. The models could be used to identify amino acids for each P′ substrate position that are favorable for, respectively, high substrate affinity and cleavage rate. © 2008 Elsevier Ltd. All rights reserved.