Searches for lead compounds intended for prevention of infective endocarditis by means of combinatorial chemical libraries.

点击翻译按钮获取中文摘要

Infective endocarditis is often caused by oral indigenous bacteria introduced into the blood stream. Abnormal or damaged heart valves including artificial valves are the most susceptible to the infection, while normal valves can be infected only by some aggressive bacteria when present in large numbers. The ability of bacteria to adhere to extracellular matrix proteins, which are exposed at the wound endocardia, is considered critical for infection. An oral viridans streptococcus, Granulicatella (Abiotrophia) adiacens, often induces endocarditis and the pathogenicity is related to its adhering activities to fibronectin (Fn). The purpose of our present study is the development of candidate compounds, which can interfere with the adherence of the microorganism. First, monoclonal antibodies against Fn were prepared and found to inhibit Fn-binding ability of the organism. On the other hand, random hexapeptide libraries were constructed by the "one peptide on one-bead" approach. A modified libraries, Gly-X1-X2-X3-X4-Gly-Tenta Gel (X = equal mixture of natural amino acids except Cys) were constructed for easier and more rapid identification. The "one peptide on one-bead" strategy is however limited to determining epitopes since quantitative evaluation is difficult. We attempted to improve the position scan strategy, in which we did not employ mixture based protocols to construct the library but each of 19 amino acid derivatives is individually coupled in divided resins. In the novel screening methodology, information obtained by use of the two types of peptide libraries supplement each other. It will not only facilitate identification of lead compounds for prevention of infective endocarditis, but also improve the efficiency in search for low molecular weight compounds that mimic interactions of higher molecular weight biological substances.