Genome sequence-based criteria for demarcation and definition of species in the genus Rickettsia

Abstract

Over recent years, genomic information has increasingly been used for prokaryotic species definition and classification. Genome sequence-based alternatives to the gold standard DNA–DNA hybridization (DDH) relatedness have been developed, notably average nucleotide identity (ANI), which is one of the most useful measurements for species delineation in the genomic era. However, the strictly intracellar lifestyle, the few measurable phenotypic properties and the low level of genetic heterogeneity made the current standard genomic criteria for bacterial species definition inapplicable to Rickettsia species. We evaluated a range of whole genome sequence (WGS)-based taxonomic parameters to develop guidelines for the classification of Rickettsia isolates at genus and species levels. By comparing the degree of similarity of 74 WGSs from 31 Rickettsia species and 61 WGSs from members of three closely related genera also belonging to the order Rickettsiales ( Orientia , 11 genomes; Ehrlichia , 22 genomes; and Anaplasma , 28 genomes) using digital DDH (dDDh) and ANI by orthology (OrthoANI) parameters, we demonstrated that WGS-based taxonomic information, which is easy to obtain and use, can serve for reliable classification of isolates within the Rickettsia genus and species. To be classified as a member of the genus Rickettsia , a bacterial isolate should exhibit OrthoANI values with any Rickettsia species with a validly published name of ≥83.63 %. To be classified as a new Rickettsia species, an isolate should not exhibit more than any of the following degrees of genomic relatedness levels with the most closely related species: >92.30 and >99.19 % for the dDDH and OrthoANI values, respectively. When applied to four rickettsial isolates of uncertain status, the above-described thresholds enabled their classification as new species in one case. Thus, we propose WGS-based guidelines to efficiently delineate Rickettsia species, with OrthoANI and dDDH being the most accurate for classification at the genus and species levels, respectively.