Comparison of tRNA and rRNA Phylogenies in Proteobacteria: Implications for the Frequency of Horizontal Gene Transfer

Abstract

The current picture of bacterial evolution is based largely on studies of 16S rRNA. However, this is just one gene. It is known that horizontal gene transfer can occur between bacterial species, although the frequency and implications of this are not fully understood. If horizontal transfer were frequent, there would be no single evolutionary tree for bacteria because each gene would follow a different tree. We carried out phylogenetic analyses of rRNA and tRNA genes from Proteobacteria (a diverse group for which many complete genome sequences are available) using RNA-specific phylogenetic methods that account for the conservation of the secondary structure. We compared trees for 16S rRNA and 23S rRNA with those derived from concatenated alignments of 29 tRNA genes that are found in all the genomes studied. The tRNA genes are scattered throughout the genomes, and would not follow the same evolutionary history if horizontal transfer were frequent. Nevertheless, the tRNA tree is consistent with the rRNA tree in most respects. Minor differences can almost all be attributed to uncertainty or unreliability of the phylogenetic method. We therefore conclude that tRNA genes give a coherent picture of the phylogeny of the organisms, and that horizontal transfer of tRNAs is too rare to obscure the signal of the organismal tree. Some tRNA genes are not present in all the genomes. We discuss possible explanations for the observed patterns of presence and absence of genes: these involve gene deletion, gene duplication, and mutations in the tRNA anticodons.