|Home | About | Journals | Submit | Contact Us | Français|
Analysis of the complete genome of Thermococcus sp. strain AM4, which was the first lithotrophic Thermococcales isolate described and the first archaeal isolate to exhibit a capacity for hydrogenogenic carboxydotrophy, reveals a proximity with Thermococcus gammatolerans, corresponding to close but distinct species that differ significantly in their lithotrophic capacities.
Thermococcus sp. strain AM4 was isolated from 1 of 13 enrichments of coccoid cells obtained from hydrothermal venting structures (East Pacific Rise; 13°N, 2,600-m depth) anaerobically on CO at 80°C and growing with H2 and CO2 production (9). Carboxydotrophic growth of strain AM4 required the presence of 50 mg/liter yeast extract. In the absence of CO, AM4 could grow on peptone or yeast extract with elemental sulfur as the electron acceptor.
Within the marine microbe sequencing project (www.moore.org/microgenome), genomic libraries for Thermococcus sp. AM4 were constructed and sequenced by the Sanger method to an 8-fold level of coverage. Sequence reads were assembled with the J. Craig Venter Institute (JCVI) Consed and Manatee packages to 17 contigs.
The contigs were connected by PCR. Open reading frames were predicted with Glimmer (http://www.cbcb.umd.edu/software/glimmer) and RAST (1). The annotation was manually cured using the BLAST and the NCBI nr databases.
The Thermococcus sp. AM4 genome consists of a circular chromosome of 2,086,428 bp without extrachromosomal elements. It contains 2,235 protein-coding genes, 46 tRNA genes, and two copies of 5S and one copy of 16S-23S rRNA genes.
Strain AM4 is most closely related to Thermococcus gammatolerans EJ3 (3, 13). The in silico hybridization yielded an average nucleotide identity (ANI) of shared protein-coding genes of 87% (about 80% of genes shared). This ANI value is lower than the 95 to 96% value shown to correspond to the 70% DNA-DNA hybridization level accepted to delimit microbial species (2, 7, 11). Thus, AM4 and T. gammatolerans represent phylogenetically close but distinct species.
Genomic comparisons have revealed an organotrophic potential comparable in the two thermococci (13). Both genomes encode a CO dehydrogenase (CODH) which allows lithoheterotrophic growth on CO plus sulfur with H2S formation (D. A. Kozhevnikova and T. G. Sokolova, unpublished data). However, AM4 possesses an extra CODH clustered in the genome with an energy-converting hydrogenase. This gene cluster, also conserved in Thermococcus onnurineus and Thermococcus barophilus (5, 6, 10, 12), enables these species to grow on CO following the reaction CO + H2O → H2 + CO2 (5, 10), while T. gammatolerans is unable to produce hydrogen from CO and H2O (T. G. Sokolova, unpublished data). Noticeably, T. gammatolerans and AM4 possess a formate hydrogenlyase gene cluster, but only T. gammatolerans possesses the formate transporter gene that allows this species to grow via hydrogenogenic oxidation of exogenous formate (4).
Like all sequenced genomes of the Thermococcales, the AM4 genome encodes a type III RubisCO but lacks a phosphoribulokinase to provide RubisCO its substrate in the Calvin cycle. Given the extensive requirements of carboxydotrophic thermococci for amino acids (5, 13), possession of an autotrophic CO2 fixation pathway would be paradoxical. Furthermore, the hypothetical CO2 fixation cycle that involves RubisCO and can proceed without phosphoribulokinase (5, 8) cannot be closed in thermococci, including carboxydotrophic ones, due to the lack of transaldolase genes. Nor is acetyl coenzyme A synthase, a key enzyme of the Wood-Ljungdahl pathway, encoded by any of the Thermococcales genomes. Thus, carboxydotrophy, while providing energy to diverse thermococci, cannot be considered to contribute to the primary production of hydrothermal vent communities.
The final annotated genome of Thermococcus sp. AM4 is available in GenBank under accession number CP002952.
We thank the Gordon and Betty Moore Foundation for sequencing support at the J. Craig Venter Institute. Thanks are also due to Justin Johnson at the JCVI.
This work was supported in part by the Agence Nationale de la Recherche (ANR-10-BLAN-Living Deep) to P.O., the Russian Foundation for Basic Research (project numbers 11-04-01723-a and 09-04-01787-a), federal targeted program for 2009-2013 “Scientific and Pedagogical Personnel of Innovative Russia” (GK number P646), and the program of the Presidium of the Russian Academy of Sciences “Molecular and Cell Biology.”