botulinum and C. cellulolyticum), which does not cluster selleckchem with other members of the Clostridium genus (Figure 1). That genus displays a high degree of variation and re-classification of some of the members of this genus is in progress (see for example [27]). That two members of the Clostridia are even placed outside the Firmicutes phylum is an indication of 16S rRNA gene sequence heterogeneity within this class. Figure 1 Phylogenetic neighbor-joining tree based on 16S rRNA genes extracted from 145 genomes (24 Negativicutes and 121 prokaryotic genomes representing 26 phyla). Bootstrap values of 50 and higher are indicated. With the exception of the Negativicutes, branches … Next, all protein-coding genes of the analyzed genomes were compared and a composition vector tree (CVtree) was produced, based on amino acid sequences (Figure 2).

The topology of the resulting tree is generally in accordance with the 16S rRNA tree shown in the previous figure. As indicated by the collapsed branches, the CVtree grouped most genomes according to their known taxonomic phyla, although not all Spirochaetes cluster together. In contrast to the 16S rRNA tree, in this protein tree all the Firmicutes cluster together, and are distinct from other phyla. The Negativicutes genomes, nested within the Firmicutes, again have the Acidaminococcaceae placed within the Veillonellaceae, while all Veillonella spp. are found in one cluster. All Clostridia, this time divided into two collapsed branches, are positioned as the closest relatives to Negativicutes.

It is of interest that among the closest relatives to Firmicutes, based on this analysis, are the Fusobacteria and the Elusimicrobia; these are atypical diderm bacteria that produce GSK-3 lipopolysaccharides [28]. However, the spirochete, Brachyspira murdochii, does not possess two membranes, but is nevertheless grouped with atypical diderms. On the other hand while the Synergistetes are atypical diderm bacteria, they are placed elsewhere in the tree (Figure 2). Figure 2 Phylogenetic tree based on composition vector analysis (CVtree) of all protein coding genes (amino acid sequences) derived from the analyzed genomes. Note that the branch lengths in this plot are artificial. The coloring is the same as in Figure 1 and … A third analysis was based on a subset of proteins found conserved amongst all analyzed genomes. These conserved proteins were selected based on a protein BLAST (a cutoff of 50% identity and 50% coverage of the query length was used) and single linkage clustering. The analysis identified 29 genes that are shared among all 145 genomes [Table 2]. A consensus tree was constructed based on these 29 conserved proteins (Figure 3).