Cheminformatics-based drug design approach for identification of inhibitors targeting the characteristic residues of MMP-13 hemopexin domain.

Background: MMP-13, a zinc dependent protease which catalyses the cleavage of type II collagen, is expressed in osteoarthritis (OA) and rheumatoid arthritis (RA) patients, but not in normal adult tissues. Therefore, the protease has been intensively studied as a target for the inhibition of progression of OA and RA. Recent reports suggest that selective inhibition of MMP-13 may be achieved by targeting the hemopexin (Hpx) domain of the protease, which is critical for substrate specificity. In this study, we applied a cheminformatics-based drug design approach for the identification and characterization of inhibitors targeting the amino acid residues characteristic to Hpx domain of MMP-13; these inhibitors may potentially be employed in the treatment of OA and RA.

Methodology/principal Findings: Sequence-based mutual information analysis revealed five characteristic (completely conserved and unique), putative functional residues of the Hpx domain of MMP-13 (these residues hereafter are referred to as HCR-13(pf)). Binding of a ligand to as many of the HCR-13(pf) is postulated to result in an increased selective inhibition of the Hpx domain of MMP-13. Through the in silico structure-based high-throughput virtual screening (HTVS) method of Glide, against a large public library of 16908 molecules from Maybridge, PubChem and Binding, we identified 25 ligands that interact with at least one of the HCR-13(pf). Assessment of cross-reactivity of the 25 ligands with MMP-1 and MMP-8, members of the collagenase family as MMP-13, returned seven lead molecules that did not bind to any one of the putative functional residues of Hpx domain of MMP-1 and any of the catalytic active site residues of MMP-1 and -8, suggesting that the ligands are not likely to interact with the functional or catalytic residues of other MMPs. Further, in silico analysis of physicochemical and pharmacokinetic parameters based on Lipinski's rule of five and ADMET (absorption, distribution, metabolism, excretion and toxicity) respectively, suggested potential utility of the compounds as drug leads.

Conclusions/significance: We have identified seven distinct drug-like molecules binding to the HCR-13(pf) of MMP-13 with no observable cross-reactivity to MMP-1 and MMP-8. These molecules are potential selective inhibitors of MMP-13 that can be experimentally validated and their backbone structural scaffold could serve as building blocks in designing drug-like molecules for OA, RA and other inflammatory disorders. The systematic cheminformatics-based drug design approach applied herein can be used for rational search of other public/commercial combinatorial libraries for more potent molecules, capable of selectively inhibiting the collagenolytic activity of MMP-13.