The basal split in the cfr phylogeny was between rlmN housekeeping methylases and their sister group formed by two major clades; one of which was comprised of S-adenosyl-methionine dependent methyltransferases from various Deltaproteobacteria and Actinomycetia, including antibiotic producer Streptomyces species; the other major clade included cfr proteins. In the latter clade, the most basal branching was from the soil alphaproteobacterium Devosia, then further basal branches contained proteins from Paenibacillus sp. and various soil-dwelling members of Thermoactinomycetaceae and Chloroflexi (Fig. 1). The clade then was split into two large sister groups, with proteins from different Bacillales in the basal branches in both groups.
One of them basally contained proteins from Paenibacillus, Brevibacillus and Cohnella, then split a branch with proteins from Ruminococcus, and a group with poorly resolved branching containing proteins from various Clostridium spp, including C. botulinum, cfrC proteins from Clostridioides difficile and a few other species (Faecalicoccus, Ruminococcus, Enterocloster, C. perfringens), and several cfrE proteins from C. difficile, Streptococcus suis, Intestinibacillus sp., Blautia sp. and Enterocloster bolteae.
The basal taxa in the other group were Bacillus cereus, B. paramycoides, Aneurinibacillus and Sporolactobacillus, then the next branch was comprised of cfr(D) proteins from a S. parasuis and a number of Enterococcus faecium isolates. A further branch contained cfr proteins from clinically relevant staphylococci (S. aureus, S. haemolyticus, S. epidermidis, S. capitis), Mammaliicoccus sciuri, Enterococcus faecium, E. faecalis as well as from various Enterobacterales (all annotated as cfr). Further branches formed a complicated structure of proteins from various Bacillales (Cytobacillus, Paenibacillus, Bacillus, Lysinibacillus, Brevibacillus), with the clade of cfr(B) intercalated containing proteins from C. difficile, Enterococcus faecalis, E. faecium, Mediterraneibacter and Ruminococcus. The most derived branch included clbA rRNA methylase proteins related to cfrs from various Bacillus spp., including B. amyloliquefaciens and B. velezensis.
Queries of cfr(B), cfr(C), cfr(D) and cfr(E) yielded very similar trees, the major differences were with the basal rlmN proteins. In case of cfr(B) these were totally absent, the root of the tree is between the S-adenosyl-methionine dependent methyltransferases and cfrs; the cfr(D) query yielded only five rlmN proteins in the basis of the tree, some of which were different from those on the cfr tree; cfr(E) resulted in a tree practically identical in topology to the tree with cfr as query.
The optrA phylogeny contained two major lineages (Fig. 2), one basally exhibited Peribacillus-derived proteins, then branches containing proteins from various Erysipelotrichaceae associated with the human microbiota. The other major lineage was split into to two sister groups, one of which contained proteins from various Lachnospiraceae associated with the human and animal microbiota, while the other had a basal branch of proteins from soil-associated Clostridium spp. and proteins from clinically important bacteria, i. e. enterococci, staphylococci, Mammaliicoccus, Lactococcus, Listeria, Campylobacter and Streptococcus suis, the latter showing up in most recognizable subgroups.
Protein sequences of poxtA were split into two sister groups with very high support values (Fig. 3). One of them contained proteins from Lachnospiraceae (Enterocloster, Hungatella and Blautia spp), while the other was comprised of a basally positioned protein from Fontibacillus solani, then a protein from Enterococcus faecium and a cluster of identical proteins from E. faecium, other enterococci (E. faecalis and E. hirae), Pediococcus spp., Staphylococcus saprophyticus, S. haemolyticus and S. aureus. The split from a protein from Protoclostridium gallicola and another from Heteroclostridium caecigallinarum was not unequivocal.