However, PLX3397 considering that the rate of evolution is faster within the Erm clade than in the corresponding clusters of bacterial KsgA (Figs 1 and 2) and that the short sequences (234 amino acid positions) do not provide sufficient signal for deep-level phylogeny reconstruction, the tree

cannot be rooted unequivocally and such an apparent paralogy is considered to be an artifact caused by long-branch attraction or the lack of a phylogenetic signal. To examine the congruence of detailed branching orders of Erm and KsgA subtrees, trees were constructed separately with all the Erm methylases detected in the databases and the corresponding KsgA proteins that exist in the same or closely related species that harbor erm genes (Figs 3 and 4). Figure

3 shows that the clustering of KsgA proteins is in good accordance with Silmitasertib research buy the typical taxonomy. In the case of Erm, the bifurcation of the main clade into two branches of the Actinobacteria and Firmicutes is consistent with that of the KsgA proteins. However, phylogenetic anomalies generated by horizontal gene transfer and duplication are recognized within the Erm clades (Fig. 4). In Fig. 4, the obvious horizontal transfers of the erm genes between phylogenetically distant bacteria are expressed as shaded boxes, and the occurrences of gene duplication are shown as shaded without boxes. It is noticeable that most of the incidents of horizontal gene transfer occurred within the clade of the Firmicutes, whereas all of the gene duplications were detected in the clade of the Actinobacteria. The comparison of the G+C content of the erm gene with that of host chromosomal DNA also supports the frequent occurrence of erm horizontal gene transfer in pathogenic bacteria, and many of these genes are related by self-transferable plasmids or transposons (Table 1). We performed a comprehensive phylogenetic analysis with extensive Erm and KsgA/Dim1 sequences found in three domains of life. The phylogenetic tree provides some insights into the origins of the erm genes with KsgA/Dim1 sequences as an appropriate outgroup. The early branching

of the Actinobacteria and Firmicutes asserts that the origin of the current erm genes in pathogenic bacteria cannot be explained by recent horizontal gene transfer from antibiotic producers. As for the origin of the present-day erm genes, those found in pathogenic bacteria may have originated from antibiotic-resistance Selleckchem Ibrutinib determinants of drug-producing bacteria (Arthur et al., 1987). If this belief is true, the main Erm clade of the Actinobacteria should be the precursor of the monophyletic tree of the Erm methylases. However, the phylogenetic tree did not place the origin of the erm genes at the ancestral node of the Actinobacteria, implying that an antibiotic producer might not provide antibiotic-resistance genes in pathogens. However, considering the frequent occurrences of horizontal erm gene transfer (Brisson-Noel et al., 1988; Berryman and Rood, 1995; Gupta et al.