抄録
Marine microorganisms are now recognized as a promising source for the development of new pharmaceuticals. We have been collecting "new" bacteria from the marine environment, mainly from marine invertebrates. ("New" means that the homology of the 16S rDNA sequence of the collected bacterium is less than 98% by a BLAST search.) In the course of screening for antitumor substances from "new" bacteria, we found the cyclic peptide compound, mechercharmycin A (1), and its linear congener, mechercharmycin B (2), from the new marine-derived Firmicutes, Thermoactinomyces sp. YM3-251. We also found urukthapelmycin A (3) from Thermoactinomyces sp. YM11-542, which was closely related to the mechercharmycins-producing strain. Mechercharmycin A (1) is a cyclic peptide bearing four oxazoles, a thiazole and the peptide moiety. The connection of the peptide moiety was easily determined as an order of dehydroalanine, valine and isoleucine by the analyses of 2D NMR data. The remaining structure, which is presumed to have contained four consecutive oxazoles and a thiazole, could not be fully determined by the NMR studies, because the lack of any correlation signals between the oxazole rings by HMBC analyses. Consequently, the structure of 1 was determined by an X-ray crystallographic analysis. The absolute configuration of 1 was determined from X-ray anomalous dispersion of the S atom. Mechercharmycin B (2) is a linear congener of 1 which has four oxazoles and the same peptide moiety as that of 1. The connections of the oxazole-2-carboxylic acid methyl ester and of the oxazole-4-carboxylic acid amide could not determined by the NMR studies, but instead by a fragment analysis of the LC-MS/MS data, enabling the complete structure of 2 to be determined. The absolute stereochemistry of 2 was determined by the Marfey's method and by a chiral HPLC analysis. The structure of urukthapelmycin A (3) was determined from an analysis of the NMR data and by a comparison of the spectral data with 1. The absolute stereochemistry of 3 was determined by the same procedure as that used for 2. The antitumor activities of 1 and 3 were relatively strong, the respective IC_<50> values for A549 (human lung cancer) cells being 4.0×10^<-8>M and 1.2×10^<-8>M, under our assay conditions, while 2 did not exhibit any inhibitory activity toward these cells even at 4×10^<-6>M. The cyclic structure of 1 and 3 must have been essential for its strong antitumor activity. Further investigations of the antitumor potential of 1 and 3 are in progress.
