Environmental genomics of "Haloquadratum walsbyi" in a saltern crystallizer indicates a large pool of accessory genes in an otherwise coherent species

Abstract

Background: Mature saturated brine (crystallizers) communities are largely dominated (> 80% of cells) by the square halophilic archaeon "Haloquadratum walsbyi". The recent cultivation of the strain HBSQ001 and the sequencing of its genome allows comparison with the metagenome of this taxonomically simplified environment. Similar studies carried out in other extreme environments have revealed very little diversity in gene content among the cell lineages present.

Results: The metagenome of the microbial community of a crystallizer pond has been analyzed by end sequencing a 2000 clone fosmid library and comparing the sequences obtained with the genome sequence of "Haloquadratum walsbyi". The genome of the sequenced strain was retrieved nearly complete within this environmental DNA library. However, many ORF's that could be ascribed to the "Haloquadratum" metapopulation by common genome characteristics or scaffolding to the strain genome were not present in the specific sequenced isolate. Particularly, three regions of the sequenced genome were associated with multiple rearrangements and the presence of different genes from the metapopulation. Many transposition and phage related genes were found within this pool which, together with the associated atypical GC content in these areas, supports lateral gene transfer mediated by these elements as the most probable genetic cause of this variability. Additionally, these sequences were highly enriched in putative regulatory and signal transduction functions.

Conclusion: These results point to a large pan-genome (total gene repertoire of the genus/species) even in this highly specialized extremophile and at a single geographic location. The extensive gene repertoire is what might be expected of a population that exploits a diverse nutrient pool, resulting from the degradation of biomass produced at lower salinities.

Figure 1

Environmental fosmids 02B08 and 07B05 and comparison with corresponding region on the "Haloquadratum walsbyi" HBSQ001 chromosome. The table underneath displays identity and similarities of each gene from the environmental clones compared to the genome sequence. In the overlapping areas two values are shown corresponding the gene in 02B08 and 07B05 respectively. When only one value is shown both similarities differed only by decimals.

Figure 2

Histogram distribution of DNA sequence identity to HBSQ001 among fosmid-end sequences.

Figure 3

Scatter plot of DNA sequence identity to HBSQ001 versus GC-content. Each dot corresponds to one fosmid-end sequence.

Figure 4

GC% distribution of genes from selected halophilic genomes, of fosmid-end sequences and of crystallizer DNA. GC values of individual genomes were obtained from public databases except for Salinibacter ruber and "Haloquadratum walsbyi" that were obtained from their respective genome databases [42, 43]. The GC plot from the crystallizer DNA determined by reassociation kinetics (differential absorbance) is from reference [12].

Figure 5

Distribution of the fosmid-end sequences along the HBSQ001 chromosome showing coverage and discontinuities. Position for each fosmid-end is obtained from the best blast hit between the fosmid-end and the HBSQ001 chromosome. 'GC-HBSQ001' consists of the number of guanine (G) and cytosine (C) nucleotides out of every 10 nucleotide on the HBSQ001 chromosome (averaged over 10 Kb). At any corresponding nucleotide position on the HBSQ001 genome (x-axis): 'coverage' consists in the number (averaged over 10 Kb) of fosmids from the library where both ends have more than 80% DNA identity to HBSQ001 and for which distance between corresponding BLAST hit positions on the HBSQ001 genome is smaller than 60 Kb; discontinuities consist of the number (averaged over 60 Kb) of 1) fosmid end sequences with more than 80% DNA identity to HBSQ001 but for which the other end gives no hit to HBSQ001 and 2) fosmids for which distance between corresponding BLAST hit positions on the HBSQ001 chromosome is greater than 60 Kb. "Choppy" regions are highlited by a grey background.

Figure 6

Trinucleotide analysis of fosmid end sequences. The distance between any two dots (sequences) represent relative proximity between their sequence signatures (multidimensional scaling was performed from 64-dimensions to 2-dimensions). The genome of HBSQ001, cut in chunks of 818 bp (the average size of fosmid-end sequences) is represented by dots tagged "HBSQ001, 818 bp". Dots with the "H. walsbyi" + "Haloquadratum", "high-GC" and "accessory pool" tags represent fosmid ends sequences from corresponding categories as described in the text.

Figure 7

Functional category distribution of putative genes found on fosmid-end sequences and comparison with "H. walsbyi" HBSQ001 genes. "H. walsbyi", "Haloquadratum" and "Accessory pool" are all putative genes found in the corresponding sequence categories described in the text.