A phylogenomic approach to microbial evolution

Abstract

To study the origin and evolution of biochemical pathways in microorganisms, we have developed methods and software for automatic, large-scale reconstructions of phylogenetic relationships. We define the complete set of phylogenetic trees derived from the proteome of an organism as the phylome and introduce the term phylogenetic connection as a concept that describes the relative relationships between taxa in a tree. A query system has been incorporated into the system so as to allow searches for defined categories of trees within the phylome. As a complement, we have developed the pyphy system for visualising the results of complex queries on phylogenetic connections, genomic locations and functional assignments in a graphical format. Our phylogenomics approach, which links phylogenetic information to the flow of biochemical pathways within and among microbial species, has been used to examine more than 8000 phylogenetic trees from seven microbial genomes. The results have revealed a rich web of phylogenetic connections. However, the separation of Bacteria and Archaea into two separate domains remains robust.

Figure 1

The phylome map of R.prowazekii (25). The outer circles represent the best hits obtained in searches against public databases using the program BLAST (12). The inner circles represent the phylogenetic connections inferred from phylogenetic reconstructions using the neighbour joining algorithm (20) and the program PAUP* (19). The lengths of the lines in the inner circle are proportional to the number of taxa in the phylogenetic trees. The key to the colour coding system is shown in Figure 2. The thick arrows in the R.prowazekii phylome map show the location of the aminoacyl-tRNA synthetases.

Figure 2

Schematic representation of the functional distribution of phylogenetic connections derived from six microbial phylomes. In, informational genes; Op, operational genes. The following gene category abbreviations are used: A, amino acid synthesis; B, biosynthesis of cofactors; C, cell envelope proteins; E, energy metabolism; I, intermediary metabolism; L, fatty acid and phospholipid biosynthesis; N, nucleotide biosynthesis; O, other; P, cell processes; R, replication; S, transcription; T, translation; X, transport; Y, tRNA synthesis; Z, regulation.

Figure 3

Phylogenetic analyses of oligopeptide ABC transporters in T.maritima (TM), P.abyssi (PAB) and Pyrococcus horikoshii (PH). Schematic picture showing the co-location (A) and phylogenetic relationships based on the individual (B) and combined (C) protein sequences of the oligopeptide ABC transporters. Neighbour joining (NJ) and maximum parsimony (MP) methods gave similar topologies. Values at nodes indicate the percentage of 1000 neighbour joining bootstraps. Values <70% are not shown. Arrows indicate sites of putative horizontal gene transfer events.

Figure 3

Phylogenetic analyses of oligopeptide ABC transporters in T.maritima (TM), P.abyssi (PAB) and Pyrococcus horikoshii (PH). Schematic picture showing the co-location (A) and phylogenetic relationships based on the individual (B) and combined (C) protein sequences of the oligopeptide ABC transporters. Neighbour joining (NJ) and maximum parsimony (MP) methods gave similar topologies. Values at nodes indicate the percentage of 1000 neighbour joining bootstraps. Values <70% are not shown. Arrows indicate sites of putative horizontal gene transfer events.

Figure 3

Phylogenetic analyses of oligopeptide ABC transporters in T.maritima (TM), P.abyssi (PAB) and Pyrococcus horikoshii (PH). Schematic picture showing the co-location (A) and phylogenetic relationships based on the individual (B) and combined (C) protein sequences of the oligopeptide ABC transporters. Neighbour joining (NJ) and maximum parsimony (MP) methods gave similar topologies. Values at nodes indicate the percentage of 1000 neighbour joining bootstraps. Values <70% are not shown. Arrows indicate sites of putative horizontal gene transfer events.