pubmed.ncbi.nlm.nih.gov

Quantifying homologous replacement of loci between haloarchaeal species - PubMed

Quantifying homologous replacement of loci between haloarchaeal species

David Williams et al. Genome Biol Evol. 2012.

Abstract

In vitro studies of the haloarchaeal genus Haloferax have demonstrated their ability to frequently exchange DNA between species, whereas rates of homologous recombination estimated from natural populations in the genus Halorubrum are high enough to maintain random association of alleles between five loci. To quantify the effects of gene transfer and recombination of commonly held (relaxed core) genes during the evolution of the class Halobacteria (haloarchaea), we reconstructed the history of 21 genomes representing all major groups. Using a novel algorithm and a concatenated ribosomal protein phylogeny as a reference, we created a directed horizontal genetic transfer (HGT) network of contemporary and ancestral genomes. Gene order analysis revealed that 90% of testable HGTs were by direct homologous replacement, rather than nonhomologous integration followed by a loss. Network analysis revealed an inverse log-linear relationship between HGT frequency and ribosomal protein evolutionary distance that is maintained across the deepest divergences in Halobacteria. We use this mathematical relationship to estimate the total transfers and amino acid substitutions delivered by HGTs in each genome, providing a measure of chimerism. For the relaxed core genes of each genome, we conservatively estimate that 11-20% of their evolution occurred in other haloarchaea. Our findings are unexpected, because the transfer and homologous recombination of relaxed core genes between members of the class Halobacteria disrupts the coevolution of genes; however, the generation of new combinations of divergent but functionally related genes may lead to adaptive phenotypes not available through cumulative mutations and recombination within a single population.

PubMed Disclaimer

Figures

F<sc>ig</sc>. 1.—
Fig. 1.—

(A) ML phylogenetic reconstruction from 59 concatenated ribosomal protein sequences from 21 haloarchaea with edge lengths scaled to substitutions per site. Two sets of nanohaloarchaeal and one mesophilic methanogen from Methanomicrobia were used as an outgroup. Protein homologs inferred as causing compositional heterogeneity were excluded, and the deepest bipartitions were collapsed due to inconsistency among nonparametric bootstrapped replicates and evidence of LBAA. (B) Bayesian sampled phylogeny inferred from the same data set with edge lengths scaled to a relaxed molecular clock. As an example, the edges marked d1–4 in (A) and the regions labeled "t" in (B) indicate the genetic distance between and the duration of coexistence respectively of the ancestral lineages of Halalkalicoccus and of Haloarcula and Halomicrobium used in HGT frequency versus genetic distance modeling. All pairwise, coexisting, nonsister edges were included.

F<sc>ig</sc>. 2.—
Fig. 2.—

A diagram indicating a horizontal transfer of a protein coding ORF inferred by embedded quartet decomposition and tree reconciliation with an adjacent ORF inferred to have been horizontally transferred in the same event. The three horizontal lines represent regions of chromosomes from Halobacterium salinarum R1 (top, putative donor of transferred genetic material), Haloarcula californiae ATCC 33799 (middle, putative recipient), and Halorhabdus utahensis DSM 12940 (bottom, a reference genome). Units are megabases (Mb). Horizontal arrows represent 3′–5′ strand direction and range of protein coding regions. Shared colors indicate most recent homology except for gray, which indicates no local homology. The vertical red arrow indicates which homologs were inferred by embedded quartet decomposition and tree reconciliation to have been transferred between the ancestor of Halobacterium salinarum R1 and Halorhabdus utahensis DSM 12940 and the direction. The reference genome was selected for being more closely related to the putative recipient than donor according to the ribosomal protein phylogeny, plotted to the left side, and to have not been inferred to have been affected by HGT for the gene analyzed with embedded quartets. ML estimates of evolutionary distances measured in substitutions per site are indicated between homologous protein coding regions with the shorter distance indicated by a color.

F<sc>ig</sc>. 3.—
Fig. 3.—

(A) Directed hierarchical network of extant and ancestral genomes showing HGTs of cobalamin (vitamin B12) biosynthesis genes. The nodes in the network are defined by the ribosomal protein phylogeny of nine haloarchaea, a subset of the full analysis of 21 genomes. The edges of the network are the vertical arrows indicating inferred HGTs by HR and are colored by homology. The network is arranged to also show the chromosomal regions coding for the proteins involved in cobalamin (vitamin B12) biosynthesis. The phylogeny edge length units are arbitrary time units; the chromosome map scale units are megabases (Mb). Horizontal arrows depict ORFs in 3′–5′ strand direction and are scaled to base pairs. Shared colors indicate most recent homology except for gray, which indicates no local homology. (B) Directed hierarchical network depicting the same HGTs as in figure 1(A) with arrows also colored by most recent homology. Gray lines indicate all other inferred HGTs, the directions of which are not indicated to maintain clarity. The green numbers indicate the total HGT events to or from that ancestral population.

F<sc>ig</sc>. 4.—
Fig. 4.—

Scatter plot of frequency of HGT events between two lineages versus the distance across the ribosome phylogeny in substitutions per site. Black: HGT between terminal edges and red: HGT between internal edges. Frequency was calculated as the total inferred HGT events between two edges on the ribosome phylogeny per overlapping edge length when scaled to a relaxed molecular clock. The dashed line is a linear model fitted after natural log transformation of the frequencies between terminal edges, R2 = 0.78, P < 6 × 10−15. Terminal edges are considered more reliable than internal edges because tips are at present day. Fitted HGT frequencies between all edges, R2 = 0.72, P < 1 × 10−16. An inverse log-linear relationship for recombination rates versus genome similarity has also been observed in experimental studies and inferred from sequence data of closely related bacterial genera.

Similar articles

Cited by

References

    1. Abascal F, Zardoya R, Posada D. ProtTest: selection of best-fit models of protein evolution. Bioinformatics. 2005;21:2104–2105. - PubMed
    1. Aivaliotis M, et al. Large-scale identification of N-terminal peptides in the halophilic archaea Halobacterium salinarum and Natronomonas pharaonis. J Proteome Res. 2007;6:2195–2204. - PubMed
    1. Allers T, Mevarech M. Archaeal genetics—the third way. Nat Rev Genet. 2005;6:58–73. - PubMed
    1. Amorós-Moya D, Bedhomme S, Hermann M, Bravo I. Evolution in regulatory regions rapidly compensates the cost of nonoptimal codon usage. Mol Biol Evol. 2010;27:2141–2151. - PubMed
    1. Andam C, Harlow T, Papke RT, Gogarten JP. Ancient origin of the divergent forms of leucyl-tRNA synthetases in Halobacteriales. BMC Evol Biol. 2012;12:85. - PMC - PubMed

Publication types

MeSH terms

Substances