Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Distinction between Borrelia and Borreliella is more robustly supported by molecular and phenotypic characteristics than all other neighbouring prokaryotic genera: Response to Margos' et al. "The genus Borrelia reloaded" (PLoS ONE 13(12): e0208432)

  • Radhey S. Gupta

    Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Writing – original draft, Writing – review & editing

    gupta@mcmaster.ca

    Affiliation Department of Biochemistry and Biomedical Sciences, McMaster University, Hamilton, Ontario, Canada

Distinction between Borrelia and Borreliella is more robustly supported by molecular and phenotypic characteristics than all other neighbouring prokaryotic genera: Response to Margos' et al. "The genus Borrelia reloaded" (PLoS ONE 13(12): e0208432)

  • Radhey S. Gupta
PLOS
x

Abstract

In a recent publication in PLOS ONE, Gabriele Margos and colleagues have questioned the division of the genus Borrelia into two genera on the basis that the differences in percentage of conserved proteins (POCP) between these two groups is >50%, which an earlier study has suggested as the threshold for differentiating prokaryotic genera. However, the POCP threshold is a poorly characterized and rarely used criterion for establishing distinction among prokaryotic genera. Detailed evaluation of the intergeneric POCP values for 37 genera from 3 different families (viz. Enterobacteriaceae- 24 genera, Morganellaceae-8 genera and Cystobacteraceae-5 genera) presented here shows that the POCP values for all genera within each of these families exceeded >58%. Thus, the suggested POCP threshold is not a useful criterion for delimitation of genus boundary and the objection by Margos et al. on this ground is invalid. Additionally, Margos et al. have questioned the specificities of ~15–20% of the conserved signature indels (CSIs) described in our work. However, as shown here, this concern is due to misunderstanding of the results and the CSIs in question are still highly-specific characteristics of the members of these genera and they provide important information regarding the evolutionary relationships of two new reptiles-echidna-related species viz. Borrelia turcica and Candidatus Borrelia tachyglossi to other Borrelia species. Results presented here show that both these species are deeper-branching members of the genus Borrelia and their placement within this genus is strongly supported by phylogenetic analyses and multiple uniquely shared CSIs with the other Borrelia species. Based on the large body of evidence derived from phylogenetic, genomic, molecular, phenotypic and clinical features, it is contended that the characteristics clearly distinguishing the Borrelia and Borreliella genera are far more numerous and of different kinds than those discerning most (all) other neighbouring genera of prokaryotes. Thus, the placement of these two groups of microorganisms into distinct genera, Borrelia and Borreliella, which clearly recognizes the differences among them, is highly appropriate and it should lead to a better understanding of the clinical, molecular and biological differences between these two important groups of microbes.

Introduction

The family Borreliaceae includes species that are causative agents of Lyme disease (LD) and others that are causative agents of tick- and louse-borne relapsing fever (RF) [15]. Our earlier comprehensive phylogenomics and comparative studies on protein sequences from Borreliaceae genomes provided compelling evidence for the existence of two genetically distinct groups of organisms within this family [6,7]. Of these two groups, one group included all species that are causative agents of the clinically distinctive disorder known as RF, whereas the second group encompassed causative agents of LD along with some other closely related species [6]. The existence of these two groups was supported by different independent lines of evidence which included: (i) Distinct branching of the LD and RF groups of species in the 16S rRNA trees and multiple genome scale trees based on protein sequences [68]; (ii) Clear distinction of the LD and RF groups of species based on pairwise comparison of either the average nucleotide identity (ANI) or the average amino acid identity (AAI) of different genes/proteins from the Borreliaceae genomes [6,7]; (iii) Identification of >70 highly-specific molecular signatures consisting of conserved signature insertions/deletions (indels) (CSIs) in protein sequences and conserved signature proteins (CSPs) that are exclusively shared by different members of either the LD or the RF group of species [6,7]; and (iv) Several phenotypic characteristics known from earlier work including the distinct pathogenicity profiles of the two groups of organisms and differences in arthropod vectors used by them [1,2,4,5]. Based on the robust evidence provided by all of these analyses, we have previously proposed a division of the family Borreliaceae (and the genus Borrelia) into two main genera, Borrelia and Borreliella [6]. In this proposal, all of the species that are part of the RF group were retained within the genus Borrelia, whereas all species related to the LD group were placed into a new genus called Borreliella [6]. This latter group of species is widely referred to by the name "Borrelia burgdorferi sensu lato", recognizing their distinctness from the RF group of species [1,2,5,9].

Recently, Margos et al. [10,11] have analyzed the genome sequences from two new Borreliaceae species, viz. Borrelia turcica and Candidatus Borrelia tachyglossi, which are associated with reptiles and echidna. In their publication, Margos et al. [10] acknowledge that the LD and RF groups of species “have different clinical, biological, and epidemiological characteristics, and phylogenetic data is concordant with this, demonstrating that these two groups are genetically similar yet distinct and form independent monophyletic sister clades that once shared a common ancestor”. Additionally, they state that the proposal by Adeolu and Gupta [6] to divide the genus Borrelia into two genera “was largely based on the identification of conserved signature insertions/deletions (indels) (CSIs) and conserved signature proteins (CSPs) that are differentially present in the LD or RF Borrelia genogroup, as well as average nucleotide identity (ANI) values calculated between whole genomes of 18 Borrelia species including eight LD species and ten RF species … it is uncontested that these differences exist between LD and RF Borrelia”.

Margos et al. [10] have questioned the division of genus Borrelia into two genera on three accounts. Their main argument for questioning the division is based on the consideration that the differences in the ANI or AAI values between these two groups, shown in our work as part of the evidence indicating that these two groups of species differ from each other [6,7], are not suitable means for differentiation of prokaryotic genera. Instead, they assert that a method proposed by Qin et al. [12] based on percentage of conserved proteins (POCP) between genomes from different species is a more reliable means for the determination of a genus level boundary. In addition to this main objection, the authors also criticize our work on two other grounds: (i) that the methodology used in our work only identifies CSIs and CSPs that are exclusive to only one Borrelia genogroup and it precludes the detection of those characteristics that are shared non-exclusively between both genogroups, and (ii) that upon inclusion of sequence information for the two new Borrelia species, about 17–20% of the previously reported 53 CSIs are unable to differentiate between the LD and the RF groups of species. I discuss below our responses to all of these criticisms and specifically the problem of using or relying on the suggested POCP threshold as a criterion for the delimitation of prokaryotic genera, which is the main basis of Margos et al.’s [10] resistance to our division of the genus Borrelia into two genera.

Materials and methods

Construction of the phylogenetic tree based on the core genome proteins and the calculation of percentage of conserved proteins (POCP) between different genomes was carried out using an internally developed software pipeline described in earlier work [1315]. Information regarding genome sequences for different species from the families Enterobacteriaceae [15], Morganellaceae [14] and Cystobacteraceae, for which the POCP values were calculated is provided in S1 and S2 Tables. Briefly, using the CD-HIT program [16], proteins sharing a minimum of 50% sequence identity and sequence length were identified in different genomes. Based on this information, the POCP between different pairs of genomes was calculated as described by Qin et al. [12]. Multiple sequence alignment (MSA) of the proteins which were found in at least 80% of the input genomes (a total of 703 protein families) were created using Clustal Omega [17]. For phylogenetic analysis, the sequence alignments were trimmed using TrimAl [18] before their concatenation into a single file. The combined sequence for the 703 core genome proteins, which after trimming consisted of 248452 aligned amino acids, was utilized for phylogenetic analysis. A maximum likelihood (ML) tree based on this sequence alignment was constructed and optimized in RAxML 8 as described in our earlier work [1315].

The 16S rRNA gene sequences for different Borreliaceae species were downloaded from All-Species Living Tree Project [19] and aligned using ClustalX2. The tree was constructed using the Maximum-likelihood (ML) method in MEGA6 [20]. Updating of the sequence information and group specificity of different CSIs and CSPs was carried out by performing BLASTp searches on the sequences of the indicated proteins. Formatting of the sequence alignment files was carried out using SIG_CREATE and SIG_STYLE programs described in our work [21]. It should be mentioned that based on different lines of evidence, the following Borrelia species (viz. B. bissettii, B. lanei, B. mayonii and B. yangtzensis) consistently group with the LD group. Unlike the other LD group of species, which are now transferred to the genus Borreliella [6], the proposal to reclassify these four species to the genus Borreliella has not yet been made. However, in the interim, to avoid any confusion due to the grouping of these Borrelia species within other Borreliella species, the genus name of these species is abbreviated as “Bor.” in the manuscript and different Figs.

Results and discussion

The inadequacy of using a 50% POCP threshold for genus level boundaries

Prokaryotic systematics involves assemblage of organisms into groups of different ranks from most inclusive to least inclusive (e.g. Phylum, Class, Order, Family, Genus and Species) on the basis of their observed similarities and differences and phylogenetic/evolutionary relationships [2225]. Species are the basic unit of any biological classification scheme. For prokaryotic organisms, although a formal definition of “the species” is lacking, for practical purposes, it is now generally accepted that strains showing >70% similarity in DNA-DNA hybridization values, or >98.65% sequence similarity in 16S rRNA, or those exhibiting >95% similarity in ANI values provide comparable means for delimiting a prokaryotic species and for identification of new species [23,2533]. In contrast to these accepted criteria for species delimitation, there are no commonly accepted or used criteria for identification of genus or higher level taxa [34]. A genus is commonly defined as “a monophyletic grouping of species with many characters in common”[22,35]. Further, there is a general consensus that the division into higher taxonomic ranks including genus level taxon should reflect phylogenetic relationships.

While there are no accepted criteria for genus level boundaries, some authors have suggested that the 16S rRNA similarity values between 94.5% and 86.6% [26] or the POCP values <50% [12] can be used as thresholds for differentiation among genera. However, these suggestions are based on studies using a limited number of prokaryotic taxa and the general utilities of these methods (or suggested thresholds) for delimitation of prokaryotic genera remains to be properly evaluated. Let us now specifically consider the utility of using the 50% POCP threshold value as a genus level boundary, which Margos et al. [10] contend provides a more suitable method for demarcation of prokaryotic genera. The study by Qin et al. [12], which suggested the use of POCP values for genus level delimitation was based on a limited number of prokaryotic taxa and the inferences based on it suffer from a number of drawbacks: (i) Interspecies POCP comparison in this study was carried out for only 17 genera. Of these, several genera such as Bacillus, Lactobacillus and Clostridium are highly polyphyletic and only a selected group of closely related species were chosen from them to represent the entire genera [12]. Due to arbitrary delimitation of these genera to a small group of selected species, the closest relatives of these genera, which are other species from the same genera, were not considered in either the interspecies or intergeneric POCP comparisons. (ii) Intergeneric POCP comparisons were carried out with only 1 arbitrarily chosen species from these 17 genera to only single species from other genera, families and orders of bacteria [12]. As many of these latter comparisons were made for species that are part of different families or orders of bacteria, the POCP values obtained for them do not reflect intergeneric differences, but rather family or order level differences. The latter values are expected to be lower than intergeneric differences and the results from such comparisons should not have been included in the comparison of intergeneric POCP values as they artificially lower the observed intergeneric values. (iii) Several genera used for interspecies comparison viz. Thermotogae, Clostridium, Mycobacterium, for which the POCP values were indicated to be higher than 50%, have since been divided into multiple genera [13,3638] indicating that the POCP threshold is not a useful or required criterion for genus level separation.

To further evaluate the usefulness of POCP values for genus level separation/boundary, we have independently determined interspecies and intergeneric POCP values for a number of families each containing multiple genera. Three well-studied families that we have examined in this regard include the family Enterobacteriaceae containing 24 genera [15], the family Morganellaceae containing 8 genera [14], and the family Cystobacteraceae containing 5 genera [39]. For all of these families, pairwise interspecies and intergeneric POCP values were determined for all species for which genome sequences were available. From the pairwise POCP matrix, average POCP values were determined for different species within each genus (interspecies POCP values) and for different genera within each of these three families (intergeneric POCP values). The results of these comparisons for the family Enterobacteriaceae are presented in the pairwise POCP matrix in Fig 1.

thumbnail
Fig 1. A comparison matrix showing the averages of the percentage of conserved proteins (POCP) within and between different genera of the family Enterobacteriaceae.

POCP was determined for all genome sequenced species from the family Enterobacteriaceae detailed in our earlier work [26]. The values along the diagonal shows the average POCP values for different species within a given genus (i.e. interspecies values), whereas all other values represent average intergeneric POCP values for different genera within this family. The blank cells indicate that only a single species was available for these genera and hence their interspecies values could not be calculated.

https://doi.org/10.1371/journal.pone.0221397.g001

As seen from the matrix in Fig 1, the intergeneric POCP values for all 24 genera that are part of this family range from a low of 59.4% to a high of 82.0% and they are all higher than 50%. Similarly, the intergeneric POCP values for the 8 genera that are part of the family Morganellaceae range from 58.7% to 84.5% (S3 Table), and for the family Cystobacteraceae, they range from 61.9%– 82.4% (S3 Table). Thus, if a POCP cut-off value of <50% was to be used for genus level boundary, then all of the different genera present within each of these three families would be part of a single genus. These results demonstrate that the usefulness of the 50% POCP threshold value for determination of genus level boundaries is very limited, if any.

Margos et al. [10] have also presented a comparison of the POCP values for the Borrelia and Borreliella genera along with some other genera within the phylum Spirochaeta. However, of the four other genera for which the POCP comparisons were made, Brachyspira and Leptospira are part of two separate orders viz. Brachyspirales and Leptospirales within the phylum Spirochaeta [40,41]. Based on the 16S rRNA sequence similarity comparisons, Yarza et al. [26] have previously noted that the species from these two orders, which are very distantly related to each other as well as other orders within the phylum Spirochaeta, should in fact be assigned class level ranks within the phylum. Thus, a comparison of the POCP values for these two genera with the other genera is misleading as they provide an indication of the order or class level differences and not intergeneric differences. The other two genera included in the comparison are Treponema and Spirochaeta. Although both of these genera are part of the family Spirochaetaceae [40,41], in phylogenetic trees, members of these genera form different clades indicating extensive divergence (unpublished results) [19,41,42]. Based on the results shown by Margos et al. (S1 Table of their publication) [10], the interspecies POCP values for members of these two genera are mostly in the range of 20–40% with an average POCP value of 33.7% for the Treponema species and 35.5% for the Spirochaeta species. Based on the 50% POCP threshold value for genus level boundaries, the species from both Treponema and Spirochaeta genera should each be divided into multiple genera. These results again point to the inadequacy of using the suggested POCP threshold value as a reliable means for the genus level boundaries.

Although a specific POCP value is not very useful for establishing a genus level boundary, a comparison matrix based on POCP, similar to the matrices based on ANI or AAI values, can still provide an overall indication of the genomic similarity and differences between two closely related groups of species. In the POCP matrix presented by Margos et al. [10], while the species from the genus Borreliella (LD group) exhibited a high degree of similarity to each other, the species from the Borrelia (RF) group exhibited considerable variability and this group was not clearly differentiated. However, the POCP matrix constructed by Margos et al. [10] was based on genome sequences that included genes present on both the linear chromosomes as well as different plasmids. The distribution of plasmids is highly variable in different Borreliaceae species/strains unlike the conservation of linear chromosome structure and chromosomal genes, [5,4346] and inclusion of plasmid sequences will introduce considerable variability in genome sequence or POCP comparison. Thus, in order to reliably compare the POCP values among different species, such comparisons should be based only on the chromosomal genes not including the plasmid genes. A POCP matrix for the Borreliaceae species based on genes present on chromosomal sequences is presented in Fig 2. As seen, this matrix clearly distinguishes the Borreliaceae species into two groups corresponding to the Borrelia and Borreliella genera. Based on this matrix, the average POCP for species from the genera Borrelia and Borreliella are 93.4% and 94.7%, respectively, whereas the average POCP value between these two groups is only 82.2%. Thus, a comparison of the POCP values based on chromosomal genes actually supports the genetic distinction between the Borrelia and Borreliella genera.

thumbnail
Fig 2. A pair-wise comparison matrix based on percentage of conserved proteins (POCP) in chromosomal genes from different genome sequenced Borreliaceae species.

The matrix was constructed using an internally developed pipeline [13,14]. Genome pairs sharing higher POCP are shaded more darkly (red). Based on their POCP values, species belonging to the family Borreliaceae form two main groups, with one group containing all of the LD and related species (or Borreliella), and the other encompassing RF group of species together with the reptile-and echidna- associated species B. turcica and Candidatus Borrelia tachyglossi (genus Borrelia).

https://doi.org/10.1371/journal.pone.0221397.g002

Specificity of the molecular signatures for the Borrelia and Borreliella genera

In the Margos et al. [10] paper, concerns were also raised regarding our methodology for identifying CSIs and CSPs, which they assert only considered those molecular signatures which were exclusively found in one Borrelia genogroup and precluded detection of such characteristics that are shared non-exclusively between both genogroups. However, in our original work, in addition to the CSIs and CSPs that are specific for the two main groups (viz. Borrelia and Borreliella), we also reported 31 CSIs and 82 CSPs that are specifically found in all Borreliaceae species [6,41]. This information was also provided and emphasized in our rebuttal response [7] to an earlier criticism of our work by these authors [47]. By non-exclusive, however, if Margos et al. [10] mean that the CSIs or CSPs are commonly shared by only some members from each of the two main clades of Borreliaceae species, then in our work we have not come across significant number of such characteristics showing any specific pattern. However, isolated characteristics of this kind can result from lateral gene transfers and they are not useful for understanding evolutionary relationships or for taxonomic purposes [32,48].

Margos et al. [10] also state that between 17–20% of the CSIs identified by us are not specific for Borrelia or Borreliella genera and do not differentiate between these two groups. However, subsequent to our earlier work describing the specificities of the CSIs for two Borreliaceae genera [6], genome sequences have become available for two new Borrelia isolates viz. B. turcica and Candidatus Borrelia tachyglossi [10,11], and they were included by Margos et al. [10] in their analyses. Of these two species/strains, B. turcica is associated with reptiles whereas Candidatus Borrelia tachyglossi was isolated from an echidna (Tachyglossus aculeatus) species [11]. In phylogenetic trees based on 16S rRNA sequences as well as multiple genome-scale phylogenetic trees and trees based on individual protein sequences (Fig 3), these two species form deeper branching lineages of the Borrelia (RF) clade [10,11]. Although in a number of trees, particularly those based on large datasets of protein sequences (Fig 3A) [10], these two species form a clade, such an association is often not seen in trees based on sequences for many individual proteins (see Fig 3B) or in the tree based on 16S rRNA gene sequences (Fig 3C). However, we will refer to B. turcica and Candidatus Borrelia tachyglossi as the Reptiles-related (RR) group/clade in this work.

thumbnail
Fig 3. Phylogenetic trees showing the branching of Borreliaceae species.

(A) A maximum-likelihood (ML) tree based on concatenated sequences of 703 core proteins found in the genomes of Borreliaceae species; (B) A tree based on sequence alignment for the RNA polymerase β’- subunit (RpoC protein). (C) A ML tree for Borreliaceae species based on 16S rRNA gene sequences.

https://doi.org/10.1371/journal.pone.0221397.g003

The inclusion of these two new species in the dataset, depending upon their branching position, is expected to alter the specificity of some of the identified signatures. In our earlier rebuttal response to Margos et al. [7,47], we had clearly outlined the different scenarios of how the inclusion of sequence information for the RR group of species, depending upon their branching positon within the family Borreliaceae, will affect the group-specificity of some of the identified CSIs. It was stated that if “the RR species/strains branch either within the RF group or as an outgroup of this clade, then such a group of species is expected to contain either some or all of the signatures for the RF clade, but generally none for the LD group”[7]. This is exactly what is observed upon the inclusion of sequence information for B. turcica and Candidatus Borrelia tachyglossi sequences. Thus, the questions raised by Margos et al. [10], regarding the specificities of some of the CSIs indicate that they are misinterpreting the results for the species distribution of the indicated CSIs.

To go over their objections, let us consider the results for different CSIs that were reported previously and how they have been affected upon the inclusion of sequence information for B. turcica and Candidatus Borrelia tachyglossi. As noted earlier, 31 identified CSIs were specific for the family Borreliaceae (Table 2 in Ref. [6]). These CSIs, as expected, are also present in protein homologs from B. turcica and Candidatus Borrelia tachyglossi (results not shown). The remaining CSIs, which distinguished the two main groups within the family Borreliaceae were/are of two kinds. Of these, the first category of 15 CSIs are in proteins whose homologs besides the family Borreliaceae are also found in other bacteria (i.e. outgroup species) (Fig 4A and 4B). Based on the presence or absence of these CSIs in the outgroup species, one can infer whether these CSIs represent an insert(s) or deletion(s) and at what specific stage in the evolution of Borreliaceae family the genetic changes responsible for these CSIs have occurred [6,7,32,41]. Of these 15 CSIs, based on the available information, 7 CSIs were indicated to be specific for the LD group, whereas in the remaining 8, the genetic changes leading to the CSIs occurred in the lineage leading to the RF group of species. Upon inclusion of sequence information for B. turcica and Candidatus Borrelia tachyglossi, which form deeper branching lineages of the RF group, no changes were observed in the specificities of any of the CSIs specific for the LD group and the homologs of the two new species lacked these CSIs (see Table 1).

thumbnail
Fig 4. A summary diagram showing the species specificities of different CSIs reported in our earlier work [6].

The CSIs described in our earlier work were of two kinds. Panels (A) and (B) present the results for CSIs, where sequence information for outgroup species was available, whereas panels (C) and (D) show results for CSIs which are found in proteins that are limited to the Borreliaceae species (i.e. no homologs in any outgroup species). Panels (A) and (C) show the results as reported earlier [6], whereas panels (B) and (D) show how the observed specificities of the CSIs have been affected upon inclusion of sequences for B. turcica and Candidatus Borrelia tachyglossi. The asterisks (*) marks the CSIs whose specificities have been questioned by Margos et al. [10]. As shown here and as discussed in the text, these CSIs remain specific for the RF group (genus Borrelia) in addition to providing important information regarding the branching or phylogenetic placement of B. turcica and Candidatus Borrelia tachyglossi within the genus Borrelia and family Borreliaceae.

https://doi.org/10.1371/journal.pone.0221397.g004

thumbnail
Table 1. Conserved signature indels (CSIs) found in widely distributed proteins that are specific for the members of the Lyme disease Borrelia (i.e. genus Borreliella).

https://doi.org/10.1371/journal.pone.0221397.t001

However, the CSIs which were previously indicated to be specific for the RF clade showed two patterns. Of these, 4 CSIs are commonly shared by all members of the RF group as well as B. turcica and Candidatus Borrelia tachyglossi (RR group), whereas the remaining 4 CSIs were only found in the RF group of species and not found in the two deeper branching RR group of species. In Fig 5, an example of CSIs showing the two types of patterns are presented. Information regarding the species specificities of all other CSIs for this group is presented in Table 2. The species distribution pattern of the CSIs for this group is exactly as we had predicted previously and the observed results, independent of the phylogenetic trees, strongly support the following inferences: (i) RR group of species, i.e. B. turcica and Candidatus Borrelia tachyglossi, are specifically associated with the RF group (i.e. genus Borrelia) as indicated by the 4 CSIs they uniquely share with the other RF group of species (Table 2A; Figs 4 and 5); (ii) B. turcica and Candidatus Borrelia tachyglossi are earlier branching members of the genus Borrelia and the genetic changes in the 4 CSIs that are absent in these two species have occurred in a common ancestor of the other Borrelia species, after the divergence of these two species (Table 2B; Figs 4 and 5). Thus, the species distribution patterns of the CSIs, upon inclusion of sequence information for B. turcica and Candidatus Borrelia tachyglossi, rather than showing any lack of specificity, provide important information clarifying and strongly supporting the observed evolutionary relationship of these species to the other Borreliaceae species (Fig 4). The CSIs whose specificities are questioned by Margos et al. [10] are marked by an asterisk (*) in Fig 4.

thumbnail
Fig 5. Partial sequence alignments of two CSIs in proteins with outgroup species that were previously reported as specific for the RF clade.

Panel (A) shows a 6 aa insert in a hypothetical protein BDU327 (BB_0326) that is specifically found in all members of the genus Borrelia including B. turcica and Candidatus Borrelia tachyglossi. (B) This panel shows a 1 aa insert in the L-lactate permease protein, which is only shared by all RF clade species but is absent in the B. turcica and Candidatus Borrelia tachyglossi homologs, which are deeper branching members of the genus Borrelia (see Figs 3 and 4). Dashes (-) in all alignments shows sequence identity with the amino acids on the top line.

https://doi.org/10.1371/journal.pone.0221397.g005

thumbnail
Table 2. Conserved signature indels in proteins that are specific for either all members of the Genus Borrelia or those lacking in the deeper branching Borrelia turcica and Candidatus Borrelia tachyglossi.

https://doi.org/10.1371/journal.pone.0221397.t002

The remaining 38 CSIs, which constitute the second category, are present in proteins that are found only in different Borreliaceae species [6]. Although these CSIs differentiate members of the LD and the RF group of species, due to the absence of these proteins in outgroup species, it is difficult to determine whether the genetic changes giving rise of these CSIs represent insertion(s) in the LD (RF) group, or deletion(s) in the RF (LD) group (see Fig 4C). Thus, Margos et al. [10] are misinterpreting the results for these CSIs, when they indicate that a specific CSI of this kind is an insert or a deletion in the RF or the LD group of species. Nonetheless, with the inclusion of sequences for B. turcica and Candidatus Borrelia tachyglossi, which are deeper branching species associated with the RF group, depending upon where the genetic changes responsible for these CSIs have occurred, the species distribution pattern of some of these CSIs will be altered. The presence and absence of the indels in all 38 CSIs from this category and their correct interpretation is provided in Table 3.

thumbnail
Table 3. Conserved signature indels in Borreliaceae-specific proteins distinguishing Borrelia and Borreliella and showing deeper branching of the RR group of species within the genus Borrelia.

https://doi.org/10.1371/journal.pone.0221397.t003

If the genetic change leading to the CSI occurred in a common ancestor of either the LD group or the entire RF group (inclusive of the RR group) then the CSIs will be present in one of these groups and absent in the other, similar to that reported in the earlier work. Of the 38 CSIs in this category, 29 showed this pattern and they differentiate between the members of the two Borreliaceae genera. One example of a CSI of this kind is shown in Fig 6A. However, if the genetic change in a given gene/protein occurred in a common ancestor of the RF group after the divergence of the RR group of species (viz. B. turcica and Candidatus Borrelia tachyglossi), then such a CSI will be present in the RF group of species, but absent in B. turcica and Candidatus Borrelia tachyglossi as well as the LD group of species. There were 7 CSIs, which showed this type of pattern (listed at the bottom of Table 3). One example of a CSI showing this type of pattern is shown in Fig 6B. However, as indicated in Fig 4, the genetic changes in this CSI or other CSIs of this kind should not be interpreted as showing that B. turcica and Candidatus Borrelia tachyglossi are specifically related to the LD group of species, as these CSIs, due to the occurrence of genetic changes in a common ancestor of the RF group, are only distinguishing the RF group of species from other Borreliaceae species. Further, as noted earlier and shown in Fig 3, although in phylogenetic trees based on large datasets of proteins, B. turcica and Candidatus Borrelia tachyglossi form a deeper-branching clade, the grouping together of these two species/strain is not seen in trees based on several individual protein sequences and also in 16S rRNA trees (Fig 3B and 3C and unpublished results). Due to this, in some cases the genetic change leading to the CSI can also occur in an RF-group ancestor inclusive of B. turcica (or Candidatus Borrelia tachyglossi) but after the branching of Candidatus Borrelia tachyglossi (or B. turcica). The genetic changes in two of the CSIs in Borreliaceae-specific proteins (viz. a membrane protein and DNA polymerase III subunit delta) described in our earlier work [6] appeared to have occurred at these stages of evolution. Sequence information for one of these CSIs is presented in Fig 6C. In this case, the described CSI is present in LD clade of species and Cand. Borrelia tachyglossi whereas B. turcica and the RF group of species are lacking this CSI. However, in this case, it will again be incorrect to interpret that the presence of this CSI in Cand. Borrelia tachyglossi and the LD group of species indicates that this species is specifically related to the LD group of species. A summary of the distribution pattern of different CSIs in the second category before and after the inclusion of results from B. turcica and Cand. Borrelia tachyglossi is presented in Fig 4C and 4D. The CSIs whose specificities are questioned by Margos et al. [10] are marked by asterisk (*) in Fig 4. Based on the correct interpretation of the genetic and evolutionary significance of these CSIs, as shown in Fig 4, it is clear that these CSIs are also highly specific characteristics of most members of the genus Borrelia. In addition, they are also clarifying the phylogenetic placement of the species B. turcica and Cand. Borrelia tachyglossi within this genus and the family Borreliaceae.

thumbnail
Fig 6. Partial sequence alignments of three CSIs in proteins found only in the Borreliaceae species providing differentiation among members of the genera Borrelia and Borreliella.

(A) This panel shows a 2 aa CSI in a hypothetical protein BT0110 that differentiates the members of the genera Borrelia and Borreliella. Twenty nine other CSIs also show a similar species distribution (Table 3). Due to the absence of outgroup species it is difficult to infer whether this CSI is an insert in the genus Borrelia or a deletion in the genus Borreliella. (B) A 3 aa CSI in a putative lipoprotein that is specific for the RF clade of species. Due to the absence of this CSI in the LD clade as well as in B. turcica and Candidatus Borrelia tachyglossi homologs this CSI is an insert in the RF clade of species (see Fig 4). (C) A 2 aa CSI in DNA polymerase III subunit delta, which is commonly shared by the LD clade of species and Cand. Borrelia tachyglossi, but absent in B. turcica and the RF group of species. Based on its species distribution, this CSI is inferred to be an insert in a common ancestor of the RF clade and B. turcica (see Fig 4 for additional information).

https://doi.org/10.1371/journal.pone.0221397.g006

Based on the evidence presented above, specifically, the correct interpretations of the results for the specificities of the CSIs and the inadequacy of genomic similarity (POCP threshold) as a criterion for genus level differentiation [12], it should be clear that the concerns raised by Margos et al. [10] to challenge the division of the genus Borrelia into two genera are not justified. In a recent publication, Estrada-Peña and Cabezas-Cruz [49] based on their examination of presence or absence of different biological processes in spirochetes species have inferred that members of the genus Borrelia and Borreliella are more similar to each other than other free-living (viz. Sediminispirochaeta, Spirochaeta and Sphaerochaeta) or pathogenic spirochetes such as Leptospira, Treponema and Brachyspira. However, their results are not surprising, as both Borrelia and Borreliella are part of the family Borreliaceae whose members exhibit very similar life cycle and vector(s)-host transmission characteristics [2,5,7,40]. With the exception of B. recurrentis, all other Borrelieaceae species have a tick-stage in their life cycle [2,5]. Thus, members of the genera Borrelia and Borreliella have coevolved intracellularly within their natural animal host-reservoir organisms for a long period of time. Due to this it is expected that all members of the family Borreliaceae (i.e. Borrelia and Borreliella genera) will share large number of biological processes and characteristics in common [5,47]. In our own work [6,7,41], we have described 31 CSIs and 82 CSPs which are uniquely shared by the members of these two genera. However, these shared characteristics are properties of the family and they reflect the multiple biological and phenotypic characteristics that the members of this family share in common. These shared properties and biological processes of the family Borreliaceae, which have been better studied, have likely led to the inference by Estrada-Peña and Cabezas-Cruz [49] that the members of these two genera are more closely related to each other than other spirochetes groups/genera. However, the observed similarity between these two genera, which are the shared properties of the family Borreliaceae, does not in any way minimizes or reduces the significance of large numbers of molecular, phenotypic and clinical differences that exist between the members of these two genera that are summarized in this work and which forms the basis of dividing this family into two different genera [6,7]. Estrada-Peña and Cabezas-Cruz [49] have not questioned the validity or significance of any these described characteristics and thus their resistance to splitting the family Borreliaceae is not justified.

To further clearly illustrate the differences between members of the genera Borrelia and Borreliella, in Table 4, I present a summary of some of the characteristics which distinguish members of these two genera. A number of other characteristics, which also distinguish these genera, are noted by Barbour [5] in a recent publication on the family Borreliaceae. The characteristics which distinguish members of these two genera include their different disease spectrums, multiple important differences in their epidemiology and phenotypic properties [5,7], and the clear differentiation and demarcation of these two groups based on genomic similarity and numerous molecular sequence based characteristics. Based on this evidence, it will be accurate to state that the distinction between these two groups of spirochetes is supported by more numerous and distinct types of characteristics than has been reported/observed for any two closely related groups (genera) of prokaryotes. Hence, we urge critics of this division to keep in mind the strong and incontrovertible evidence supporting the distinctness of these two groups of spirochetes.

thumbnail
Table 4. Clinical, molecular and phenotypic differences between members of the RF-RR group (genus Borrelia) and the Lyme disease group (genus Borreliella).

https://doi.org/10.1371/journal.pone.0221397.t004

As noted earlier, the species which are now part of the genus Borreliella are widely referred to by the name "Borrelia burgdorferi sensu lato", recognizing their distinctness from other Borreliaceae species, which are members of the genus Borrelia [1,2,5,9]. However, the meanings of the terms "Borrelia burgdorferi sensu lato" or “RF clade”, or which Borrelia species are part of each of these groups, or the species which fall outside of these two groups (viz. RR group of species), are not clearly understood by many scientists and others professionals working in this as well as other related fields. Hence, the substitution of these poorly understood terms with more precise and unambiguous names (Borrelia and Borreliella), which clearly differentiates the relapsing fever encompassing group of species from the different Lyme disease-causing and related microorganisms [3,4,46], should be highly beneficial to the field in terms of advancing our understanding of the molecular, biochemical and biological differences that underlie these two unique disease-causing groups of microorganisms.

Subsequent to our earlier work [6], a number of new species belonging to the family Borreliaceae have been described [5558]. Of these species, Borrelia bissettiae, Borrelia californiensis, Borrelia lanei, Borrelia mayonii and Borrelia yangtzensis group reliably with the members of the genus Borreliella (LD-group) in 16S rRNA trees [55] (Fig 3C), or where genome sequence information is available based on uniquely shared molecular characteristics with other members of the genus Borreliella [8,9] (see Table 1). Hence, new name combinations for these species are described below.

Description of Borreliella bissettiae comb. nov. (bis.set´ti.ae. N.L. gen. n. bissettiae, of Bissett, named after Marjorie L. Bissett, who isolated and described this spirochaete with her co-worker Warren Hill) Basonym: Borrelia bissettiae Margos et al. 2016

The strain for this species was isolated by Bissett and Hill [59] and the description of this species is as provided by Margos et al. [56] for Borrelia bissettiae

Type strain: DN127 = DSM 17990 = CIP 109136.

Description of Borreliella californiensis comb. nov. (ca.li.for.ni.en´sis. N.L. fem. adj. californiensis, belonging to California, from where the type strain was isolated)

Basonym: Borrelia californiensis Margos et al. 2016

The description of this species is the same as provided by Postic et al. [60] and Margos et al. [56] for Borrelia californiensis

Type strain: CA446 = DSM 17989 = ATCC BAA-2689.

Description of Borreliella lanei comb. nov. (la.ne′i. N.L. gen. n. lanei, in honour of Professor Robert S. Lane for his outstanding contributions to Borrelia and Ixodes research)

Basonym: Borrelia lanei Margos et al. 2017

The description of this species is the same as provided by Margos et al. [55] for Borrelia lanei

Type strain: (see also StrainInfo.net) CA28-91 = DSM 17992 = CIP 109135.

Description of Borreliella mayonii comb. nov. (ma.yo′ni.i. N.L. gen. n. mayonii, after William James Mayo and Charles Horace Mayo, founders of the Mayo Clinic).

Basonym: Borrelia mayonii Pritt et al. 2016

The description of this species is the same as provided by Pritt et al. [58] for Borrelia mayonii

Type strain: MN14-1420 = ATCC BAA-2743 = DSM 10281.

Description of Borreliella yangtzensis comb. nov. (yang.tzen′sis. N.L. fem. adj. yangtzensis, referring to the Yangtze River valley in China, where these organisms were first isolated.

Basonym: Borrelia yangtzensis Margos et al. 2015

The description of this species is the same as provided by Margos et al. [57] for Borrelia yangtzensis

Type strain: Okinawa-CW62 = DSM 24625 = JCM 17189.

Supporting information

S1 Table. Species and genome sequence information for Enterobacteriaceae species used in POCP analysis.

https://doi.org/10.1371/journal.pone.0221397.s001

(PDF)

S2 Table. Species and genome sequence information for Morganellaceae and Cystobacteraceae species used in POCP analysis.

https://doi.org/10.1371/journal.pone.0221397.s002

(PDF)

S3 Table. Averages of the intrageneric and intergeneric POCP values for species from the family Morganellaceae and Cystobacteraceae.

https://doi.org/10.1371/journal.pone.0221397.s003

(PDF)

Acknowledgments

This work was supported by research grant No. 249924 from the Natural Science and Engineering Research Council of Canada to RSG. I thank Sudip Patel for assistance in this work and Dr. Alan Barbour for helpful comments on the manuscript.

References

  1. 1. Barbour AG, Hayes SF. Biology of Borrelia species. Microbiol Rev.1986; 50: 381–400. pmid:3540570
  2. 2. Wang G, Schwartz I: Genus II. Borrelia Swellengrebel 1907, 582AL. In Bergey's Manual of Systematic Bacteriology Volume 4: The Bacteroidetes, Spirochaetes, Tenericutes (Mollicutes), Acidobacteria, Fibrobacteres, Fusobacteria, Dictyoglomi, Gemmatimonadetes, Lentisphaerae, Verrucomicrobia, Chlamydiae, and Planctomycetes. Edited by Brenner DJ, Krieg NR, Garrity GM, Staley JT. New York: Springer Verlag; 2011:484–498.
  3. 3. Talagrand-Reboul E, Boyer PH, Bergstrom S, Vial L, Boulanger N. Relapsing Fevers: Neglected Tick-Borne Diseases. Front Cell Infect Microbiol.2018; 8: 98. pmid:29670860
  4. 4. Bergstrom S, Normark J. Microbiological features distinguishing Lyme disease and relapsing fever spirochetes. Wien Klin Wochenschr.2018; 130: 484–490. pmid:30074091
  5. 5. Barbour AG: Borreliaceae. In Bergey Manual of Systematics of Bacteria and Archaea. John Wiley & Sons, Inc. Bergey's Manual Trust; 2019.
  6. 6. Adeolu M, Gupta RS. A phylogenomic and molecular marker based proposal for the division of the genus Borrelia into two genera: The emended genus Borrelia containing only the members of the relapsing fever Borrelia, and the genus Borreliella gen. nov. containing the members of the Lyme disease Borrelia (Borrelia burgdorferi sensu lato complex). Antonie van Leeuwenhoek.2014; 105: 1049–1072. pmid:24744012
  7. 7. Barbour AG, Adeolu M, Gupta RS. Division of the genus Borrelia into two genera (corresponding to Lyme disease and relapsing fever groups) reflects their genetic and phenotypic distinctiveness and will lead to a better understanding of these two groups of microbes (Margos et al. (2016) There is inadequate evidence to support the division of the genus Borrelia. Int. J. Syst. Evol. Microbiol. ). Int J Syst Evol Microbiol.2017.
  8. 8. Barbour AG. Phylogeny of a relapsing fever Borrelia species transmitted by the hard tick Ixodes scapularis. Infect Genet Evol.2014; 27: 551–558. pmid:24813576
  9. 9. Margos G, Vollmer SA, Ogden NH, Fish D. Population genetics, taxonomy, phylogeny and evolution of Borrelia burgdorferi sensu lato. Infect Genet Evol.2011; 11: 1545–1563. pmid:21843658
  10. 10. Margos G, Gofton A, Wibberg D, Dangel A, Marosevic D, Loh SM, et al. The genus Borrelia reloaded. PLoS ONE.2018; 13: e0208432. pmid:30586413
  11. 11. Gofton AW, Margos G, Fingerle V, Hepner S, Loh SM, Ryan U, et al. Genome-wide analysis of Borrelia turcica and 'Candidatus Borrelia tachyglossi' shows relapsing fever-like genomes with unique genomic links to Lyme disease Borrelia. Infect Genet Evol.2018; 66: 72–81. pmid:30240834
  12. 12. Qin QL, Xie BB, Zhang XY, Chen XL, Zhou BC, Zhou J, et al. A proposed genus boundary for the prokaryotes based on genomic insights. J Bacteriol.2014; 196: 2210–2215. pmid:24706738
  13. 13. Gupta RS, Lo B, Son J. Phylogenomics and Comparative Genomic Studies Robustly Support Division of the Genus Mycobacterium into an Emended Genus Mycobacterium and Four Novel Genera. Front Microbiol.2018; 9: 67. pmid:29497402
  14. 14. Adeolu M, Alnajar S, Naushad S, Gupta RS. Genome-based phylogeny and taxonomy of the 'Enterobacteriales': proposal for Enterobacterales ord. nov. divided into the families Enterobacteriaceae, Erwiniaceae fam. nov., Pectobacteriaceae fam. nov., Yersiniaceae fam. nov., Hafniaceae fam. nov., Morganellaceae fam. nov., and Budviciaceae fam. nov. Int J Syst Evol Microbiol.2016; 66: 5575–5599. pmid:27620848
  15. 15. Alnajar S, Gupta RS. Phylogenomics and comparative genomic studies delineate six main clades within the family Enterobacteriaceae and support the reclassification of several polyphyletic members of the family. Infect Genet Evol.2017; 54: 108–127. pmid:28658607
  16. 16. Fu L, Niu B, Zhu Z, Wu S, Li W. CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics.2012; 28: 3150–3152. pmid:23060610
  17. 17. Sievers F, Wilm A, Dineen D, Gibson TJ, Karplus K, Li W, et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol.2011; 7: 539. pmid:21988835
  18. 18. Capella-Gutierrez S, Silla-Martinez JM, Gabaldon T. trimAl: a tool for automated alignment trimming in large-scale phylogenetic analyses. Bioinformatics.2009; 25: 1972–1973. pmid:19505945
  19. 19. Yilmaz P, Parfrey LW, Yarza P, Gerken J, Pruesse E, Quast C, et al. The SILVA and "All-species Living Tree Project (LTP)" taxonomic frameworks. Nucleic Acids Res.2014; 42: D643–D648. pmid:24293649
  20. 20. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol.2013; 30: 2725–2729. pmid:24132122
  21. 21. Gupta RS: Identification of Conserved Indels that are Useful for Classification and Evolutionary Studies. In Bacterial Taxonomy, Methods in Microbiology Volume 41. Edited by Goodfellow M, Sutcliffe IC, Chun J. London: Elsevier; 2014:153–182.
  22. 22. Oren A: Microbial Systematics. In Handbook of Environmental Engineering, Vol. 10: Environmental Biotechnology. Edited by Wang LK, Ivanov V, Tay J-H, Hung YT. New York: Springer Science + Business Media; 2010:81–120.
  23. 23. Oren A, Garrity GM. Then and now: a systematic review of the systematics of prokaryotes in the last 80 years 13. Antonie van Leeuwenhoek.2014; 106: 43–56. pmid:24306768
  24. 24. Parker CT, Tindall BJ, Garrity GM. International Code of Nomenclature of Prokaryotes. Int J Syst Evol Microbiol.2015.
  25. 25. Sutcliffe IC. Challenging the anthropocentric emphasis on phenotypic testing in prokaryotic species descriptions: rip it up and start again 1. Front Genet.2015; 6: 218. pmid:26136772
  26. 26. Yarza P, Yilmaz P, Pruesse E, Glockner FO, Ludwig W, Schleifer KH, et al. Uniting the classification of cultured and uncultured bacteria and archaea using 16S rRNA gene sequences 1. Nat Rev Microbiol.2014; 12: 635–645. pmid:25118885
  27. 27. Rossello-Mora R, Amann R. Past and future species definitions for Bacteria and Archaea. Syst Appl Microbiol.2015; 38: 209–216. pmid:25747618
  28. 28. Varghese NJ, Mukherjee S, Ivanova N, Konstantinidis KT, Mavrommatis K, Kyrpides NC, et al. Microbial species delineation using whole genome sequences. Nucleic Acids Res.2015; 43: 6761–6771. pmid:26150420
  29. 29. Goris J, Konstantinidis KT, Klappenbach JA, Coenye T, Vandamme P, Tiedje JM. DNA-DNA hybridization values and their relationship to whole-genome sequence similarities. Int J Syst Evol Microbiol.2007; 57: 81–91. pmid:17220447
  30. 30. Konstantinidis KT, Tiedje JM. Genomic insights that advance the species definition for prokaryotes. Proc Natl Acad Sci U S A.2005; 102: 2567–2572. pmid:15701695
  31. 31. Klenk HP, Goker M. En route to a genome-based classification of Archaea and Bacteria? 1. Syst Appl Microbiol.2010; 33: 175–182. pmid:20409658
  32. 32. Gupta RS. Impact of genomics on the understanding of microbial evolution and classification: the importance of Darwin's views on classification. FEMS Microbiol Rev.2016; 40: 520–553. pmid:27279642
  33. 33. Sutcliffe IC, Trujillo ME, Whitman WB, Goodfellow M. A call to action for the International Committee on Systematics of Prokaryotes. Trends Microbiol.2013; 21: 51–52. pmid:23238020
  34. 34. Sangal V, Goodfellow M, Jones AL, Schwalbe EC, Blom J, Hoskisson PA, et al. Next-generation systematics: An innovative approach to resolve the structure of complex prokaryotic taxa. Sci Rep.2016; 6: 38392. pmid:27924912
  35. 35. Oren A: Concepts about Phylogeny of Microorganisms—an Historical Overview. In Molecular Phylogeny of Microorganisms. Edited by Oren A, Papke RT. Norfolk: Caister Academic Press; 2010:1–21.
  36. 36. Bhandari V, Gupta RS. Molecular signatures for the phylum (class) Thermotogae and a proposal for its division into three orders (Thermotogales, Kosmotogales ord. nov. and Petrotogales ord. nov.) containing four families (Thermotogaceae, Fervidobacteriaceae fam. nov., Kosmotogaceae fam. nov. and Petrotogaceae fam. nov.) and a new genus Pseudothermotoga gen. nov. with five new combinations. Antonie van Leeuwenhoek.2014; 105: 143–168. pmid:24166034
  37. 37. Yutin N, Galperin MY. A genomic update on clostridial phylogeny: Gram-negative spore formers and other misplaced clostridia. Environ Microbiol.2013; 15: 2631–2641. pmid:23834245
  38. 38. Lawson PA, Citron DM, Tyrrell KL, Finegold SM. Reclassification of Clostridium difficile as Clostridioides difficile (Hall and O'Toole 1935) Prevot 1938. Anaerobe.2016; 40: 95–99. pmid:27370902
  39. 39. Awal RP, Garcia R, Gemperlein K, Wink J, Kunwar B, Parajuli N, et al. Vitiosangium cumulatum gen. nov., sp. nov. and Vitiosangium subalbum sp. nov., soil myxobacteria, and emended descriptions of the genera Archangium and Angiococcus, and of the family Cystobacteraceae. Int J Syst Evol Microbiol.2017; 67: 1422–1430. pmid:28141508
  40. 40. Paster BJ: "Phylum XV. Spirochaetes Garrity and Holt (2001)". In Bergey's Manual of Systematic Bacteriology Volume 4: The Bacteroidetes, Spirochaetes, Tenericutes (Mollicutes), Acidobacteria, Fibrobacteres, Fusobacteria, Dictyoglomi, Gemmatimonadetes, Lentisphaerae, Verrucomicrobia, Chlamydiae, and Planctomycetes. Edited by Brenner DJ, Krieg NR, Garrity GM, Staley JT. New York: Springer Verlag; 2010.
  41. 41. Gupta RS, Mahmood S, Adeolu M. A phylogenomic and molecular signature based approach for characterization of the phylum Spirochaetes and its major clades: proposal for a taxonomic revision of the phylum. Front Microbiol.2013; 4: 217. pmid:23908650
  42. 42. Norris SJ, Paster BJ, Smibert RM: Genus IV. Treponema Schaudinn 1905, 1728AL (Spironema Vuillemin 1905, 1568; Microspironema Stiles and Pfender 1905, 936). In Bergey's Manual of Systematic Bacteriology Volume 4: The Bacteroidetes, Spirochaetes, Tenericutes (Mollicutes), Acidobacteria, Fibrobacteres, Fusobacteria, Dictyoglomi, Gemmatimonadetes, Lentisphaerae, Verrucomicrobia, Chlamydiae, and Planctomycetes. Edited by Brenner DJ, Krieg NR, Garrity GM, Staley JT. New York: Springer Verlag; 2010:501–531.
  43. 43. Miller SC, Porcella SF, Raffel SJ, Schwan TG, Barbour AG. Large linear plasmids of Borrelia species that cause relapsing fever. J Bacteriol.2013; 195: 3629–3639. pmid:23749977
  44. 44. Casjens SR, Mongodin EF, Qiu WG, Luft BJ, Schutzer SE, Gilcrease EB, et al. Genome stability of Lyme disease spirochetes: comparative genomics of Borrelia burgdorferi plasmids. PLoS ONE.2012; 7: e33280. pmid:22432010
  45. 45. Schuler W, Bunikis I, Weber-Lehman J, Comstedt P, Kutschan-Bunikis S, Stanek G, et al. Complete genome sequence of Borrelia afzelii K78 and comparative genome analysis. PLoS ONE.2015; 10: e0120548. pmid:25798594
  46. 46. Casjens SR, Di L, Akther S, Mongodin EF, Luft BJ, Schutzer SE, et al. Primordial origin and diversification of plasmids in Lyme disease agent bacteria. BMC Genomics.2018; 19: 218. pmid:29580205
  47. 47. Margos G, Marosevic D, Cutler S, Derdakova M, Diuk-Wasser M, Emler S, et al. There is inadequate evidence to support the division of the genus Borrelia. Int J Syst Evol Microbiol.2017; 67: 1081–1084. pmid:27930271
  48. 48. Gao B, Gupta RS. Microbial systematics in the post-genomics era. Antonie van Leeuwenhoek.2012; 101: 45–54. pmid:22048742
  49. 49. Estrada-Pena A, Cabezas-Cruz A. Phyloproteomic and functional analyses do not support a split in the genus Borrelia (phylum Spirochaetes). BMC Evol Biol.2019; 19: 54. pmid:30760200
  50. 50. Barbour AG: Relapsing fever. In Harrison's Principle of Internal Medicine. Edited by Kasper DL, Longo DL, Jameson JL, Loscalzo . New York: McGraw Hill; 2015:1146–1149.
  51. 51. Barbour AG, Schwan TG: Borrelia, Swellengrebel 1907, 582AL. In Bergey Manual of Systematics of Bacteria and Archaea, Online @ 2015 Bergey's Manual Trust. John Wiley & Sons, Inc. Bergey's Manual Trust; 2019:1–22.
  52. 52. Barbour AG, Weigang Q: Borreliella, Adeolu and Gupta 2014, 1014VP. In Bergey Manual of Systematics of Bacteria and Archaea, Online @ 2015 Bergey's Manual Trust. John Wiley & Sons, Inc. Bergey's Manual Trust; 2019:1–23.
  53. 53. Wang Z, Wu M. A phylum-level bacterial phylogenetic marker database. Mol Biol Evol.2013; 30: 1258–1262. pmid:23519313
  54. 54. Barbour AG: Specificity of Borrelia-tick vector relationships. In SGM Symposium 63: Microbe-Vector Interactions in Vector-Borne Disease. Edited by Gillispie S, Smith G, Osbourn A. Cambridge: Cambridge University Press; 2004:75–90.
  55. 55. Margos G, Fedorova N, Kleinjan JE, Hartberger C, Schwan TG, Sing A, et al. Borrelia lanei sp. nov. extends the diversity of Borrelia species in California. Int J Syst Evol Microbiol.2017; 67: 3872–3876. pmid:28884668
  56. 56. Margos G, Lane RS, Fedorova N, Koloczek J, Piesman J, Hojgaard A, et al. Borrelia bissettiae sp. nov. and Borrelia californiensis sp. nov. prevail in diverse enzootic transmission cycles. Int J Syst Evol Microbiol.2016; 66: 1447–1452. pmid:26813420
  57. 57. Margos G, Chu CY, Takano A, Jiang BG, Liu W, Kurtenbach K, et al. Borrelia yangtzensis sp. nov., a rodent-associated species in Asia, is related to Borrelia valaisiana. Int J Syst Evol Microbiol.2015; 65: 3836–3840. pmid:28875924
  58. 58. Pritt BS, Mead PS, Johnson DK, Neitzel DF, Respicio-Kingry LB, Davis JP, et al. Identification of a novel pathogenic Borrelia species causing Lyme borreliosis with unusually high spirochaetaemia: a descriptive study. Lancet Infect Dis.2016; 16: 556–564. pmid:26856777
  59. 59. Bissett ML, Hill W. Characterization of Borrelia burgdorferi strains isolated from Ixodes pacificus ticks in California. J Clin Microbiol.1987; 25: 2296–2301. pmid:3323225
  60. 60. Postic D, Garnier M, Baranton G. Multilocus sequence analysis of atypical Borrelia burgdorferi sensu lato isolates—description of Borrelia californiensis sp. nov., and genomospecies 1 and 2. Int J Med Microbiol.2007; 297: 263–271. pmid:17374507