Comparative genomics study of polyhydroxyalkanoates (PHA) and ectoine relevant genes from Halomonas sp. TD01 revealed extensive horizontal gene transfer events and co-evolutionary relationships

Abstract

Background: Halophilic bacteria have shown their significance in industrial production of polyhydroxyalkanoates (PHA) and are gaining more attention for genetic engineering modification. Yet, little information on the genomics and PHA related genes from halophilic bacteria have been disclosed so far.

Results: The draft genome of moderately halophilic bacterium, Halomonas sp. TD01, a strain of great potential for industrial production of short-chain-length polyhydroxyalkanoates (PHA), was analyzed through computational methods to reveal the osmoregulation mechanism and the evolutionary relationship of the enzymes relevant to PHA and ectoine syntheses. Genes involved in the metabolism of PHA and osmolytes were annotated and studied in silico. Although PHA synthase, depolymerase, regulator/repressor and phasin were all involved in PHA metabolic pathways, they demonstrated different horizontal gene transfer (HGT) events between the genomes of different strains. In contrast, co-occurrence of ectoine genes in the same genome was more frequently observed, and ectoine genes were more likely under coincidental horizontal gene transfer than PHA related genes. In addition, the adjacent organization of the homologues of PHA synthase phaC1 and PHA granule binding protein phaP was conserved in the strain TD01, which was also observed in some halophiles and non-halophiles exclusively from γ-proteobacteria. In contrast to haloarchaea, the proteome of Halomonas sp. TD01 did not show obvious inclination towards acidity relative to non-halophilic Escherichia coli MG1655, which signified that Halomonas sp. TD01 preferred the accumulation of organic osmolytes to ions in order to balance the intracellular osmotic pressure with the environment.

Conclusions: The accessibility of genome information would facilitate research on the genetic engineering of halophilic bacteria including Halomonas sp. TD01.

Figure 1

Phylogenetic tree of putative PhaC from Halomonas sp. TD01 with reported PHA synthases. Halophiles were highlighted in bold. The trees were constructed using the neighbor-joining algorithm with MEGA (version 5.03) software. The GenBank accession numbers were given after the microorganism names. The numbers besides the nodes indicated the bootstrap values based on 500 replications. Bar 0.1 substitutions per site were indicated on the graph.

Figure 2

Multiple alignment of putative PhaC from Halomonas sp. TD01 with reported PHA synthases. Black shading indicated the conserved catalytic triad residues, and gray shading indicated lipase-like box residues. The sequences were listed in the same order of the phylogenetic tree in Figure 1.

Figure 3

The organization of phaP and phaC1. (A) The phasin gene (phaP) was located on the upstream of phaC1 with a space of 92 bp. (B) The predicted promoter structure of phaC1. Bold capital letters indicated key base pairs. TSS, transcription start site; TLS, translation start site of phaC1; ST, stop codon of phaP; -10, Pribnow box; -35, -35 element. Arrow indicated the transcription from TSS.

Figure 4

Phylogenetic tree based on the 16S rDNA sequences of strains with homologues of PHA and ectoine relevant enzymes. The trees were constructed using the neighbor-joining algorithm with MEGA (version 5.03) software. The numbers besides the nodes indicated the bootstrap values based on 500 replications. Bar 0.05 substitutions per site were indicated on the graph. The GenBank accession numbers of 16S rDNA sequences of the strains with the homologues of PHA and ectoine relevant enzymes were listed in Additional file 1, Table S1.

Figure 5

Evolutionary analysis based on the 16S rDNA, PHA and ectoine relevant enzymes sequences. The evolutionary distances of 16s rDNA, PHA (A) and ectoine (B) enzymes between Halomonas sp. TD01 and other strains were calculated from the multiple alignments with ClustalW [33]. For each strain, the evolutionary distance of 16s rDNA was plotted on the X axis, and evolutionary distances of PHA (A) and ectoine (B) relevant enzymes were plotted on the Y axis. The same enzyme of different species was linked, while broken lines indicated the genes missing from the genomes of the corresponding strains (Additional file 1, Table S1).

Figure 6

The correlation of evolutionary distance for any two of the 16S rDNA, PHA and ectoine relevant sequences. The heatmap (A) and cluster dendrogram (B) based on their evolutionary correlation coefficient. The self-distance was set at zero. Then, the heatmap was plotted by using the function heatmap.2 in package gplots of R programming language. The color intensity in the heatmap corresponded to the distance. The cluster dendrogram was plotted by the hierarchical clustering method in R programming language.