Marine bacteria are comprehensively existed in the ocean environment and produce large amounts of different secondary metabolites. However, traditional screening for bioactive natural compounds is time-consuming and has high probability to rediscover known compounds. In my thesis, a chemical fingerprint database of marine bacteria based on their secondary metabolite profile was presented. A software system as well as an online interface for data processing and management were developed and provided freely for the community to use. Till now, 1430 bacterial strains had been fermented and their metabolite profiles were imported to this database.

Several applications of this metabolite database were discussed, which included correlation of marine bacterial taxonomy and secondary met...[ Read more ]

Marine bacteria are comprehensively existed in the ocean environment and produce large amounts of different secondary metabolites. However, traditional screening for bioactive natural compounds is time-consuming and has high probability to rediscover known compounds. In my thesis, a chemical fingerprint database of marine bacteria based on their secondary metabolite profile was presented. A software system as well as an online interface for data processing and management were developed and provided freely for the community to use. Till now, 1430 bacterial strains had been fermented and their metabolite profiles were imported to this database.

Several applications of this metabolite database were discussed, which included correlation of marine bacterial taxonomy and secondary metabolite profiles, discovery of "species-specific" secondary metabolites, and avoidance of rediscovering known active compounds. The results showed that, classification based on the secondary metabolite profile was not in complete agreement with 16s rRNA. However, some "species-specific" metabolites may present and serve as a phenotypic trait for bacterial classification. Moreover, the screening strategy for marine natural products could be substantially improved. A good case was the discovery of 14 new thalassospiramides from Thalassospira and Tistrella species by using a reported strain – Thalassospira sp. CNJ-328 to search against the whole metabolite database. The bioassays showed that all these novel cyclopeptides possessed nanomolar level inhibitory activity against human calpain 1.

A highly unusual biosynthetic mechanism was elucidated by the observed molecular diversity and two putative homologous gene clusters. In order to discover more thalassospiramides and investigate the evolutionary relationship between the two originally identified gene clusters, an extended study of 76 Thalassospira, 29 Tistrella and some other additional Rhodospirillaceae was undertaken. Totally 21 new molecules were identified and their distribution in different strains suggested horizontal gene transfer from Thalassospira to other families. To prove this hypothesis, a comparative genome analysis of 55 Rhodospirillaceae was performed and totally 7 gene cluster patterns that were likely responsible for thalassospiramide biosynthesis were identified. The comparison of bacterial function genes and essential genes indicated that gene cluster 1 might be conserved in the Thalassospira xiamenensis and be the ancestral cluster that evolved to other patterns via horizontal gene transfer and gene loss events. Gene cluster 7, the smallest one, still produced large amounts of thalassospiramide molecules, demonstrating the potential benefits to the bioengineering community of elucidating and mimicking biological shortcuts in nature.

Thalassospiramides have no traditional "warhead" but possess nanomolar-level inhibitory activity against human calpain 1. To investigate their activity mechanism, a combination of chemical modification, mass spectrometric techniques, site-directed mutagenesis, and molecular modeling were performed. The structure-activity relationship data supported the hypothesis that the rigid 12-membered ring containing an α, β-unsaturated carbonyl moiety is the pharmacologically active functional group, in contrast to classic electrophilic "warheads" in known calpain inhibitors. The good inhibitory activity was achieved by the covalent binding of thalassospiramide’ s α, β-unsaturated carbonyl moiety to the thiol group of calpain’s catalytic cys115 residue by a Michael 1,4-addition reaction. The preliminary selectivity and toxicity results indicate thalassospiramides have the potential to be a promising drug leads.

The impact of culture media and physical factor on bacterial metabolite production was also discussed in my thesis. Ten different culture conditions were designed by the single factor substitute of carbon source, nitrogen source, temperature, pH and salinity, and 40 microbes were selected to examine grow status and metabolite production. The results showed that different culture media and physical factors affected microbial growth and metabolite production. Especially, the correlation of microbe metabolite profile and culture condition suggests the comparison of different metabolite profiles has to be performed under the same culture condition. For the specific metabolites - thalassospiramides and didemnins, the examination of the production pattern under different culture conditions showed that the production of two molecules has no response to the change of culture conditions, indicates biosynthesis of these secondary metabolites may be important for bacterial survival.