Phyloproteomic and functional analyses do not support a split in the genus Borrelia (phylum Spirochaetes)

Background: The evolutionary history of a species is frequently derived from molecular sequences, and the resulting phylogenetic trees do not include explicit functional information. Here, we aimed to assess the functional relationships among bacteria in the Spirochaetes phylum, based on the biological processes of 42,489 proteins in reference proteomes of 34 Spirochaetes species. We tested the hypothesis that the species in the genus Borrelia might be sufficiently different to warrant splitting them into two separate genera. Results: A detrended canonical analysis demonstrated that the presence/absence of biological processes among selected bacteria contained a strong phylogenetic signal, which did not separate species of Borrelia. We examined the ten biological processes in which most proteins were involved consistently. This analysis demonstrated that species in Borrelia were more similar to each other than to free-life species (Sediminispirochaeta, Spirochaeta, Sphaerochaeta) or to pathogenic species without vectors (Leptospira, Treponema, Brachyspira), which are highly divergent. A dendrogram based on the presence/absence of proteins in the reference proteomes demonstrated that distances between species of the same genus among free-life or pathogenic non-vector species were higher than the distances between the 19 species (27 strains) of Borrelia. A phyloproteomic network supported the close functional association between species of Borrelia. In the proteome of 27 strains of Borrelia, only a few proteins had evolved separately, in the relapsing fever and Lyme borreliosis groups. The most prominent Borrelia proteins and processes were a subset of those also found in free-living and non-vectored pathogenic species. In addition, the functional innovation (i.e., unique biological processes or proteins) of Borrelia was very low, compared to other genera of Spirochaetes. Conclusions: We found only marginal functional differences among Borrelia species. Phyloproteomic networks that included all pairwise combinations between species, proteins, and processes were more effective than other methods for evaluating the evolutionary relationships among taxa. With the limitations of data availability, our results did not support a split of the arthropod-transmitted spirochaetes into the proposed genera, Borrelia and Borreliella.