摘要

Human ADAMs (a disintegrin and metalloproteinases) have been established as an attractive therapeutic target of inflammatory disorders such as inflammatory bowel disease (IBD). The ADAM metallopeptidase domain 17 (ADAM17 or TACE) and its close relative ADAM10 are two of the most important ADAM members that share high conservation in sequence, structure and function, but exhibit subtle difference in regulation of downstream cell signaling events. Here, we described a systematic protocol that combined computational modeling and experimental assay to discover novel peptide hydroxamate derivatives as potent and selective inhibitors for ADAM17 over ADAM10. In the procedure, a virtual combinatorial library of peptide hydroxamate compounds was generated by exploiting intermolecular interactions involved in crystal and modeled structures. The library was examined in detail to identify few promising candidates with both high affinity to ADAM17 and low affinity to ADAM10, which were then tested in vitro with enzyme inhibition assay. Consequently, two peptide hydroxamates Hxm-Phe-Ser-Asn and Hxm-Phe-Arg-Gln were found to exhibit potent inhibition against ADAM17 (K-i = 92 and 47 nM, respectively) and strong selectivity for ADAM17 over ADAM10 (similar to 7-fold and similar to 5-fold, S=0.86 and 0.71, respectively). The structural basis and energetic property of ADAM17 and ADAM10 interactions with the designed inhibitors were also investigated systematically. It is found that the exquisite network of nonbonded interactions involving the side chains of peptide hydroxamates is primarily responsible for inhibitor selectivity, while the coordination interactions and hydrogen bonds formed by the hydroxamate moiety and backbone of peptide hydroxamates confer high affinity to inhibitor binding.