Borrelia burgdorferi: a (somewhat) clonal bacterial species

Abstract

Response from Dykhuizen and BarantonWe thank Brian Stevenson for the chance to clarify two issues. The first involves the appropriateness of defining different species in the Borrelia burgdorferi complex; the second, the likelihood of insertion of virulence factors into the Lyme Borrelia genome by horizontal transfer.Twelve bacterial taxonomists, speaking for the community as a whole, have formalized the concept of a bacterial species:‘The phylogenetic definition of a species generally would include strains with approximately 70% or greater DNA–DNA relatedness and with 5°C or less ΔTm. Both values must be considered.’ 1xReport of the ad hoc committee on reconciliation of approaches to bacterial systematics. Wayne, L.G. et al. Int. J. Syst. Bacteriol. 1987; 37: 463–464CrossrefSee all References1The same definition is used for both sexual and asexual species. All the named species in the B. burgdorferi species complex have been defined this way, and have used type specimens (original references defining each species can be found in Ref. 2xMolecular typing of Borrelia burgdorferi sensu lato: taxonomic, epidemiological, and clinical implications. Wang, G. et al. Clin. Microbiol. Rev. 1999; 12: 633–653PubMedSee all ReferencesRef. 2). As DNA–DNA hybridization is not a practical means of assigning strains to species, diagnostics, such as rrfA–rrlB intergenic region typing, ospA allele variation or the presence of particular restriction sites in the chromosome, have been developed to do this. Diagnostic criteria are used only for identification and are not part of the species definition, and later work can show that particular diagnostics are inadequate 3xTwo distinct ospA genes among Borrelia valaisiana strains. Wang, G. et al. Res. Microbiol. 2000; 151: 325–331Crossref | PubMed | Scopus (14)See all References3. In these cases, a particular strain can be reassigned to a different species. Even if a strain is assigned to two or more species depending on the criterion used, DNA–DNA hybridization with the type specimens can always be used to assign it to the ‘correct’ species.Usually, different species occupy different niches. We agree with Brian Stevenson that, until recently, little was known about the ecological differences between the species of Borrelia. However, as pointed out by Ryffel et al.4xScored antibody reactivity determined by immunoblotting shows an association between clinical manifestations and presence of Borrelia burgdorferi sensu stricto, B. garinii, B. afzelii, and B. valaisiana in humans. Ryffel, K. et al. J. Clin. Microbiol. 1999; 37: 4086–4092PubMedSee all References4, even in human pathogenesis, there are highly significant differences between species. Thus, the species distinctions are useful.We agree with Brian Stevenson that there is large-scale intragenome recombination and that in some strains large fragments can be lost from the genome. Where we disagree is that we propose that there is no large-scale intergenome recombination in Borrelia and that no new genes, particularly pathogenicity islands, are added to the genome 5xImplications of a low rate of horizontal transfer in Borrelia. Dykhuizen, D.E. and Baranton, G. Trends Microbiol. 2001; 9: 344–350Abstract | Full Text | Full Text PDF | PubMed | Scopus (68)See all References5. The test case will be the vslE region. Atypical nucleotide composition is the criterion used as evidence of incorporation of large regions of DNA from external sources 6xA bacterial genome in flux: the twelve linear and nine circular extrachromosomal DNAs in an infectious isolate of the Lyme disease spirochete Borrelia burgdorferi. Casjens, S. et al. Mol. Microbiol. 2000; 35: 490–516Crossref | PubMed | Scopus (489)See all References6. Using this criterion, only about 0.1% of the DNA in Borrelia is likely to be horizontally transferred compared with 18% in Escherichia coli 7xAmelioration of the bacterial chromosome: rates of change and exchange. Lawrence, J.G. and Ochman, H. J. Mol. Evol. 1997; 44: 383–397Crossref | PubMed | Scopus (520)See all References7, where horizontal transfer of pathogenicity islands is clearly important. Atypical nucleotide composition is a less reliable indicator of horizontal transfer than the lack of homologues in closely related species 8xCodon bias and base composition are poor indicators of horizontally transferred genes. Koski, L.B. et al. Mol. Biol. Evol. 2001; 18: 404–412Crossref | PubMedSee all References8.Given the marginal nature of the evidence using atypical DNA composition for all but the vlsE region, we feel that only this region needs to be considered carefully as evidence for transfer of pathogenicity islands. Atypical nucleotide composition can be caused either by horizontal transfer or by a different mutational spectrum. These hypotheses can be distinguished by observation of the mutational variation across strains and species. If horizontal transfer has recently added this region, then the mutational variation should be biased towards low G+C; if the mutational spectrum is different in this region, the mutational variation should reflect the 50% G+C observed in this region.