Deep-sea soft sediments from trench systems and depths in the northwestern Pacific Ocean ranging from less than 300 to 10,897 m in depth have been analyzed for three target genera of actinomycetes: Micromonospora, Rhodococcus, and Streptomyces. Only culturable strains, recovered at atmospheric pressure on selective isolation media, have been examined to date. Maximum recoveries of culturable bacteria were greater that 10(7)/ml wet g sediment, but actinomycetes comprised a small proportion of this population (usually less than 1%). The target actinomycetes were isolated at all depths except from the Mariana Trench sediments. Actinomycete colonies were defined initially on the basis of colony morphologies, and preliminary identification then was made by chemotaxonomic tests. Pyrolysis mass spectrometry (PyMS) of deep-sea mycolic acid-containing actinomycetes gave excellent correspondence with numerical (phenetic) taxonomic analyses and subsequently was adopted as a rapid procedure for assessing taxonomic diversity. PyMS analysis enabled several clusters of deep-sea rhodococci to be distinguished that are quite distinct from all type strains. 16S rRNA gene sequence analysis has revealed that several of these marine rhodococci have sequences that are very similar to certain terrestrial species of Rhodococcus and to Dietzia. There is evidence for the intrusion of terrestrial runoff into these deep trench systems, and the inconsistency of the phenotypic and molecular taxonomies may reflect recent speciatiion events in actinomycetes under the high-pressure conditions of the deep sea. The results of DNA-DNA pairing experiments point to the novelty of Rhodococcus strains recovered from hadal depths in the Izu Bonin Trench. Biotransformation studies of deep-sea bacteria have focused on nitrile compounds. Nitrile-metabolizing bacteria, closely related to rhodococci, have been isolated that grow well at low temperature, high salt concentrations, and high pressures, suggesting that they are of marine origin or have adapted to the deep-sea environment.

A large number of mycolate actinomycetes have been recovered from deep-sea sediments in the NW Pacific Ocean using selective isolation methods. The isolates were putatively assigned to the genus Rhodococcus on the basis of colony characteristics and mycolic acid profiles. The diversity among these isolates and their relationship to type strains of Rhodococcus and other mycolate taxa were assessed by Curie point pyrolysis mass spectrometry (PyMS). Three major (A, C, D) and two minor (B, E) groups were defined by PyMS. Cluster A was a large group of isolates recovered from sediment in the Izu Bonin Trench (2679 m); Cluster C comprised isolates from both the Izu Bonin Trench (6390 and 6499 m) and from the Japan Trench (4418, 6048 and 6455 m). These Cluster C isolates showed close similarity to Dietzia maris and this was subsequently confirmed using molecular methods. Cluster D contained isolates recovered from a sediment taken from a depth of 1168 m in Sagami Bay and were identified as members of the terrestrial species Rhodococcus luteus. Clusters B and E had close affinities with members of the genera Gordonia and Mycobacterium. The presence of Thermoactinomyces in certain of the deep-sea sediments studied was indicative of the movement of terrestrial material into the ocean depths. 16S ribosomal RNA gene sequence analyses produced excellent definition of most genera of the mycolata, and indicated that the among the deep sea isolates (1) were novel species of Corynebacterium, Gordonia and Mycobacterium, and (2) a Sea of Japan isolate the phylogenetic depth of which suggests the possibility of a new genus. Polyphasic taxonomic analysis revealed considerable diversity among the deep sea rhodococci and evidence for recently diverged species or DNA groups.

A continual need in natural product discovery is dereplication, that is the ability to exclude previously tested microorganisms from screening programmes. Whole-cell fingerprinting techniques offer an ideal solution to this problem because of their rapidity and reproducibility, dependence on small samples, and automation. One such technique, Curie-point pyrolysis mass spectrometry (PyMS), has been deployed for the characterisation of a unique collection of actinomycetes recovered from Pacific Ocean sediments approximately 2000 to 6500 m below sea level. This paper addresses the question: to what extent are pyrogroups, defined on the basis of PyMS fingerprinting, related to classifications derived from more conventional microbial systematics? A collection of 44 randomly chosen deep-sea rhodococci were coded and subjected to a double-blind PyMS and numerical taxonomic (NT) analysis; the latter sorted the strains into clusters (taxospecies) using large sets of equally weighted phenotypic data. At the end of the experiment the codes were disclosed and the NT classification shown to generate 6 homogeneous clusters corresponding to different deep-sea sites. The matching of these clusters with the resulting pyrogroups was very high with an overall congruence of nearly 98%. Thus, PyMS characterisation is directly ascribable to the phenotypic variation being sought for biotechnology screens. Moreover, the exquisite discriminatory power of PyMS readily revealed infraspecific diversity in these industrially important bacteria.

No abstract