Abstract

Species of the genus Bartonella are involved in an increasing variety of human diseases. In addition to the 14 currently recognized species, several Bartonella strains have been recovered from a wide range of wild and domestic mammals in Europe and America. Such a high diversity of geographic distributions, animal reservoirs, arthropod vectors and pathogenic properties makes clarification of our knowledge about the phylogeny of Bartonella species necessary. Phylogenetic data have been inferred mainly from 16S rDNA, 16S–23S rRNA intergenic spacer, citrate synthase and 60 kDa heat-shock protein gene sequences, which are available in GenBank. Comparison of phylogenetic organizations obtained from various genes allowed six statistically significant evolutionary clusters to be identified. Bartonella bacilliformis and Bartonella clarridgeiae appear to be divergent species. Bartonella henselae, Bartonella koehlerae and Bartonella quintana cluster together, as well as Bartonella vinsonii subsp. vinsonii and B. vinsonii subsp. berkhoffii. The fifth group includes bacteria isolated from various rodents that belong to native species from the New World and in the sixth, Bartonella tribocorum, Bartonella elizabethae and Bartonella grahamii are grouped with several strains associated with Old World indigenous rodents. The position of the other species could not be consistently determined. As some cat- or rodent-associated Bartonella appeared to cluster together, it has been suggested that these bacteria and their reservoir hosts may co-evolve. Lack of host specificity, however, seems to be frequent and may reflect the influence of vector specificity. Host or vector specificity may also explain the current geographic distribution of Bartonella species.

Introduction

Members of the genus Bartonella are Gram-negative, oxidase-negative, fastidious bacteria belonging to the α2-subclass of Proteobacteria. Common features of Bartonella include transmission by an arthropod vector and survival within mammalian reservoir hosts [1]. During recent years, an increasing number of Bartonella species have been isolated and characterized, and the genus, extended by unification with the genera Rochalimaea and Grahamella, currently consists of 14 recognized species [2,3]. To date, seven of the 14 species have been implicated in human disease, and Bartonella are considered as emerging pathogens [4]. Bartonella bacilliformis is the causative agent of bartonellosis (Carrion's disease), a biphasic disease endemic to Andean valleys. Bartonella quintana and Bartonella henselae, etiologic agents of trench fever and cat-scratch disease respectively, have also been involved in endocarditis and in bacillary angiomatosis in immuno-compromised patients [4]. Bartonella clarridgeiae is suspected to be an additional agent of cat-scratch disease [5]. Bartonella elizabethae and Bartonella vinsonii subsp. berkhoffii have been shown to cause endocarditis [6,7], while B. vinsonii subsp. arupensis was first isolated from a febrile patient with valvular disease in the USA [8]. Bartonella grahamii has been implicated in cases of neuroretinitis [9]. The other species, B. vinsonii subsp. vinsonii, Bartonella taylorii, Bartonella doshiae, Bartonella talpae, Bartonella peromysci, Bartonella alsatica, Bartonella tribocorum, Bartonella koehlerae and Bartonella birtlesii were isolated from the blood of non-human mammals but have not yet been associated with human disease [3,10,,12]. Additionally, an increasing number of Bartonella strains have been recovered from a wide range of mammals, including rodents, cervids and cattle, in Europe and America [1,13,14]. Although partial, the genetic characterization of these isolates suggests that some of them may represent new Bartonella species [15,,17].

As the diversity of the genus Bartonella, including bacterial species and strains, animal hosts and vectors, geographic distribution, and pathogenic properties, appears to be far more extended and complex than initially thought, phylogenetic studies are urgently needed to clarify the taxonomic organization within the genus and the evolutionary relationships between Bartonella species. By identifying lineages of bacteria that share a common evolutionary lineage as well as common epidemiological or pathogenic characteristics, such studies may also help to recognize some of the determinants of the biology of these organisms. To date, phylogenetic data have been inferred from gene sequence comparisons, using different analytical methods. Most of the current knowledge about Bartonella taxonomy is supported by only three genes, namely the 16S rRNA-, the citrate synthase (gltA)- and the 60 kDa heat-shock protein (groEL)-encoding genes, which have been sequenced for most Bartonella species. Even with such limited data, interpretation of phylogenetic trees remains difficult: 16S rDNA-based phylogeny generally lacks statistical support at the interspecies level, and trees inferred from groEL and gltA gene sequences, although providing some statistically significant lineages, show discrepancies when different analytical methods are used [17,18]. The purpose of this brief review is to discuss the published data relating to Bartonella phylogeny and to compare the various architectures inferred from these different gene sequences in order to assess the reliability of the proposed reconstructions and to identify lineages that are consistently proposed. On the basis of such overall analyses, we discuss hypotheses that have been formulated about the natural history of the genus, especially that of co-evolution between the Bartonella and their mammalian reservoir hosts and that of divergent evolutionary process between Bartonella located in separated geographic areas [15,16].

Genotypic studies and phylogenetic trees

To date, comparisons of sequences from seven genes or genetic loci have been used for phylogenetic analyses of Bartonella species. These are the 16S rRNA-encoding gene, the 16S/23S intergenic spacer region (ITS), and five protein-encoding genes: the riboflavin synthetase (ribC), the cell division protein (ftsZ), the citrate synthetase (gltA), the 60 kDa heat-shock protein (groEL), which are involved in basic cellular functions, and more recently the 17 kDa antigen, an outer membrane protein eliciting a strong immune response in infected hosts [17,,,20]. Comparison of 16S rRNA gene sequences is considered as being generally useful for establishing evolutionary relationships, because of its high information content, conservative nature and universal distribution [21]. Interest has focused on the protein-encoding genes because they appeared to be conserved at the species level but highly variable between different species, and also because sequences can be analyzed at the protein level as well as at the genomic level, for better reliability [17,18,22]. The use of hypervariable ITS sequence comparison has been controversial due to doubts concerning the reliability of such sequences as indicators of molecular evolution [22,23]. Nevertheless, the ability of a specific sequence to produce meaningful phylogenetic data can only be appreciated after comparison and reconstruction inferable, by assessing the statistical support of the resulting branching order. In practical terms, only four gene data sets are currently usable for phylogenetic studies: these are 16S rDNA, ITS, gltA, and groEL. Use of other genes is hampered by the lack of Bartonella sequences [19,20].

After gene sequence alignment, phylogenetic trees are constructed using three different analytical approaches: the parsimony method and two distance methods, neighbor-joining and maximum likelihood [24]. All 16S rDNA, ITS, gltA, and groEL sequences currently available in GenBank, including those from validated Bartonella species and those from partially characterized strains, have been used to infer the trees. The four parsimony trees are presented in Fig. 1. To investigate the stability of these trees, bootstrap values were obtained from 100 randomly generated sample trees. Bootstrap values are considered to be an approximation of a statistical confidence interval. Values >90% are considered to be statistically significant as they reflect a risk of <10% of accepting a wrong branching order [25,26].

1

Comparison of parsimony trees derived from 16S rRNA gene, ITS, groEL, and gltA for Bartonella species and strains. The gltA analysis was based on a 259 bp alignment of the gene, the ITS, groEL and 16S rDNA analysis on the complete sequences of the respective genes. The support for each branch, as determined from 100 bootstrap samples, is indicated by the value at the node. This analysis provided tree topology only, and the lengths of both vertical lines and horizontal lines are not significant. Boxes contain clusters that are consistently found with at least two genes. Cluster A: B. bacilliformis. Cluster B: B. clarridgeiae. Cluster C: B. quintana, B. henselae, and B. koehlerae. Cluster D: Nine strains associated with New World native rodents (Sigmodon sp., Oryzomys sp., Phylottis sp.) with one Rattus sp.-associated isolate from the USA. Cluster E: B. vinsonii subsp. vinsonii and B. vinsonii subsp. berkhoffii. Cluster F: B. elizabethae, B. grahamii, and B. tribocorum with seven strains associated with Old World native rodents (Rattus sp., Mus sp.).

1

Comparison of parsimony trees derived from 16S rRNA gene, ITS, groEL, and gltA for Bartonella species and strains. The gltA analysis was based on a 259 bp alignment of the gene, the ITS, groEL and 16S rDNA analysis on the complete sequences of the respective genes. The support for each branch, as determined from 100 bootstrap samples, is indicated by the value at the node. This analysis provided tree topology only, and the lengths of both vertical lines and horizontal lines are not significant. Boxes contain clusters that are consistently found with at least two genes. Cluster A: B. bacilliformis. Cluster B: B. clarridgeiae. Cluster C: B. quintana, B. henselae, and B. koehlerae. Cluster D: Nine strains associated with New World native rodents (Sigmodon sp., Oryzomys sp., Phylottis sp.) with one Rattus sp.-associated isolate from the USA. Cluster E: B. vinsonii subsp. vinsonii and B. vinsonii subsp. berkhoffii. Cluster F: B. elizabethae, B. grahamii, and B. tribocorum with seven strains associated with Old World native rodents (Rattus sp., Mus sp.).

While 16S rRNA-based analyses have played a major role in establishing the phylogenetic position of the genus Bartonella among Proteobacteria and also in defining the limits of this genus [2,3], they are considered insensitive for intragenus phylogeny, because levels of 16S rDNA sequence similarity are very high (97.5–99.9%) between Bartonella species [5,10,,12] and inferences are difficult to make with confidence from sequences that exhibit more than 97% similarity [21]. The branching orders of the trees derived from comparison of 16S rDNA genes have only moderate significance, as no branch, other than that containing B. bacilliformis, is supported by more than 90% of the bootstrap samples. Comparison of tree topology inferred using parsimony and distance methods has also revealed marked differences [17]. To date, only one study has assessed ITS-based phylogeny [27]. Intra-Bartonella phylogenies derived from gltA and groEL data have some statistical significance and resulted in the same topology with all used methods. The bootstrap values of the gltA trees were higher when using distance methods than when using parsimony, whereas GroEL trees showed greater support when using parsimony analysis [17,18]. Additionally, trees based on groEL and 17 kDa antigen amino acid sequences have been generated by using parsimony methods (the PROTPARS algorithm of PHYLIP) to overcome the potential bias resulting from synonymous base substitutions, but these received little statistical support [18,21,28].

Current phylogeny

Comparisons of phylogenetic information resulting from 16S rDNA-, ITS-, gltA- and groEL-based genotypic studies have resulted in six clusters being consistently identified (i.e. clusters which were statistically supported to at least two different genes, with all analytical methods) (Fig. 1). The unique, relatively deep-rooted divergence of B. bacilliformis (group A) is supported with 100% bootstrap values by 16S rDNA, ITS, gltA and groEL analyses, as well as by its particular fatty acid, protein and antigenic composition, growth requirements and colony morphology [4,17,18,29]. Levels of similarity for amino acid sequences of ftsZ proteins, as well as for nucleotide sequences of ribC genes, are low between B. bacilliformis and other Bartonella species, supporting its divergent position [19,20]. According to ITS and groEL analyses, B. clarridgeiae may also be a divergent species (group B). A close evolutionary relationship between B. clarridgeiae and cat-associated strains from the Netherlands and the USA is suggested by 16S rDNA-based phylogeny. As B. clarridgeiae and B. bacilliformis are two flagellated species, they have been suspected to share close relationships, but statistical support is lacking for such organization [15,18,27]. The third group (C), which contains B. quintana and B. henselae, is found with bootstrap values >90% in trees derived from gltA and groEL sequences. The high degree of conservation between ftsZ homologs, ribC proteins and 17 kDa antigens from B. henselae and B. quintana also indicates that these two species are closely related [19,20,28]. Comparison of 16S rDNA and gltA sequences suggests that B. koehlerae may cluster with this group, and more closely with B. henselae.

Group D, determined by gltA-based phylogeny only, included taxons associated with mice and rats caught in Peru and in the USA, respectively [1,15,,17]. This group can be divided into two subgroups, one including the mouse-associated bacteria and the other including the rat-associated species (Fig. 1). A strong association between B. vinsonii subsp. vinsonii and B. vinsonii subsp. berkhoffii is observed with ITS, groEL and gltA sequence studies (group E), but not with those of the 16S rDNA gene [30]. The position of this B. vinsonii cluster relative to other Bartonella species remains unclear. The sixth group (F) is obtained with 16S rRNA, ITS, gltA and groEL gene sequence comparisons. Like group D, it contains rodent-associated organisms. B. elizabethae, B. grahamii, B. tribocorum cluster into this clade, together with uncharacterized strains found in mice and rats from the USA, Peru and Europe [1,16,17]. Within group F, B. elizabethae clusters with isolates from Peruvian rats in a separate sub-clade supported by gltA analysis. The taxonomic position of other species and strains, mainly B. taylorii, B. doshiae and B. alsatica, could not be consistently determined on the basis of genetic sequence data available to date.

Host specificity and phylogeny

One of the main interests of phylogenetic studies is the possibility of eventually identifying lineages of bacteria that share common epidemiological or biological characteristics, as well as shared taxonomic origins. It can provide clues to help understand the natural history of Bartonella species, their relationships with mammalian hosts and arthropod vectors, and their geographical repartition. As Bartonella parasites a wide range of mammals, the existence of a specific association between one Bartonella species and a single mammalian host has been suggested. The notion that specific hosts are vulnerable to infection by only one species of bacterium, reflecting a co-evolution phenomenon, is supported by the fact that, to be a competent reservoir, an animal must be susceptible to a parasite in a way that allows development of a chronic infection [13,15,17]. Table 1 summarizes the identified vertebrate hosts, arthropod vectors and geographic distribution of the Bartonella species belonging to the major phylogenetic clusters that have been identified.

1

Characteristics of the major phylogenetic clusters obtained consistently with analysis of the 16S rDNA, ITS, gltA and groEL gene sequences for Bartonella species

Cluster Gene providing significant support for cluster determination Taxons Reservoir host Arthropod vector Place of isolation Ref. 
16S rRNA, ITS, gltA B. bacilliformis human sandfly (Lutzomyia verrucarumPeru, Ecuador, Colombia [1,4
ITS, groEL B. clarridgeiae cat USA, France [5,37
groEL, gltA B. quintana human body louse (Pediculus humanisworldwide [4
  B. henselae cat cat flea (Ctenocephalides felis)? worldwide [4,33,37
  B. koehlerae cat USA [11
gltA Bartonella sp. R-PHY1 mouse (P. peruvianaPeru [17
  Bartonella sp. SH8200GA, Bartonella sp. OP6399GA, Bartonella sp. RR11755TX rats (S. hispidus, O. palustris, R. rattusUSA [15,16
ITS, groEL, gltA B. vinsonii subsp. vinsonii vole (Microtus pennsylvanicusrat flea (Nosopsyllus fasciatus), vole ear mite (Trombicula mirotiCanada [35
  B. vinsonii subsp. berkhoffii dog ticks? USA, Europe [6,30
16S rRNA, ITS, groEL, gltA B. grahamii vole (C. glareolusUK, Netherlands [3,9
  B. elizabethae rat (R. norvegicusfleas (Xenopsylla cheopis)? USA, Peru [1,7,13,16
  B. tribocorum rat (R. norvegicusFrance [10
  Bartonella sp. C5RAT rat (R. norvegicusUSA, Peru, Portugal [16,17
  Bartonella sp. MM5136CA mouse (M. musculusUSA [16
Unclassified B. doshiae vole (Microtus agrestisUK [3
  B. taylorii rat (Apodemus sp.) UK [3
  B. alsatica rabbit (Oryctolagus cuniculusFrance [12
Cluster Gene providing significant support for cluster determination Taxons Reservoir host Arthropod vector Place of isolation Ref. 
16S rRNA, ITS, gltA B. bacilliformis human sandfly (Lutzomyia verrucarumPeru, Ecuador, Colombia [1,4
ITS, groEL B. clarridgeiae cat USA, France [5,37
groEL, gltA B. quintana human body louse (Pediculus humanisworldwide [4
  B. henselae cat cat flea (Ctenocephalides felis)? worldwide [4,33,37
  B. koehlerae cat USA [11
gltA Bartonella sp. R-PHY1 mouse (P. peruvianaPeru [17
  Bartonella sp. SH8200GA, Bartonella sp. OP6399GA, Bartonella sp. RR11755TX rats (S. hispidus, O. palustris, R. rattusUSA [15,16
ITS, groEL, gltA B. vinsonii subsp. vinsonii vole (Microtus pennsylvanicusrat flea (Nosopsyllus fasciatus), vole ear mite (Trombicula mirotiCanada [35
  B. vinsonii subsp. berkhoffii dog ticks? USA, Europe [6,30
16S rRNA, ITS, groEL, gltA B. grahamii vole (C. glareolusUK, Netherlands [3,9
  B. elizabethae rat (R. norvegicusfleas (Xenopsylla cheopis)? USA, Peru [1,7,13,16
  B. tribocorum rat (R. norvegicusFrance [10
  Bartonella sp. C5RAT rat (R. norvegicusUSA, Peru, Portugal [16,17
  Bartonella sp. MM5136CA mouse (M. musculusUSA [16
Unclassified B. doshiae vole (Microtus agrestisUK [3
  B. taylorii rat (Apodemus sp.) UK [3
  B. alsatica rabbit (Oryctolagus cuniculusFrance [12

Taxonomic data supporting the hypothesis of co-evolution can be mainly found among cat- and rodent-associated Bartonella strains. Thus, the cat-associated species B. henselae and B. koehlerae appear to be closely related to each other and B. clarridgeiae forms a cluster with other cat isolates (Fig. 1). Among rodent isolates, all bacteria recovered from Phylottis mice in Peru shared close relationships [17]. Isolates from rice rats (Oryzomys palustris) also appeared to be genetically very close [15]. In a study of 325 Rattus norvegicus rats in the USA and in Portugal, all 63 Bartonella recovered from these rats clustered into a well-defined clade [16]. According to 16S rDNA and ITS analyses, B. taylorii, a species isolated from mice (Apodemus sp.), may be closely related with other Apodemus-associated strains (Fig. 1) [22]. Similarly, a large-scale sampling of multiple species of Peromyscus sp. from a variety of sites in different physiographic regions of the south-eastern USA provided several Bartonella isolates that clustered together using gltA-based analyses [15]. A study of 128 cattle from California established that all 63 cattle isolates had the same profile by using PCR/RFLP analysis of the gltA gene [14]. These results, together with recent experimental evidence among certain rodent species, strongly suggest that some host specificity may occur within the genus Bartonella, although it is far from being absolute [31].

Infection of the same mammal species by several Bartonella which belong to distinct phylogenetic groups has been widely reported: in species of small wild mammals caught in the UK [32], in the cotton rat population from the south-eastern USA [15] and also in deer and elk in California and Oregon [14]. Conversely, the same Bartonella species can be isolated from different mammal species that inhabit the same geographical location. Although observed on only a few occasions, the observation that some Rattus rattus-associated strains from the USA were genotypically indistinguishable from bacteria isolated from the American cotton rat, Sigmodon hispidus, also suggests that species can cross from one rodent species to another. Indeed, the ubiquitous species R. rattus frequently occurs in rural environments, where it may come into contact with indigenous New World rodents [16]. Similarly, the identity between B. elizabethae- and R. norvegicus-associated strains demonstrated, for the first time, that rodent-associated Bartonella species could act as human pathogens [1] and suggested that the predominantly urban rat R. norvegicus could be a reservoir for human infection among inner-city residents living in poor social conditions [16]. Several other animal-associated Bartonella have shown their ability to infect humans, inducing clinical manifestations which are different from those observed in natural reservoir hosts: this is the case for B. vinsonii subsp. arupensis, which may have a rodent reservoir [8], for B. vinsonii subsp. berkhoffii, initially described in dogs [6], and for the cat-associated species B. henselae and B. clarridgeiae [4,5].

To explain such lack of host specificity, it has been speculated that vector specificity may be an important factor in the natural history, parasitic cycle and distribution of Bartonella species. Thus, as several different species of fleas are naturally found as ectoparasites on small mammals, infection by different Bartonella species may originate from different flea species, with each Bartonella exhibiting species specificity to its vector. Conversely, sharing of the same flea fauna by different mammals could result in infection of various mammals by the same Bartonella species [17]. Ticks have been also suspected to play a role in transmission of Bartonella from one mammal species to another [14,33]. Collection and analysis of arthropods caught on animals and humans and in the environment is necessary to determine if they can be infected with Bartonella species [14]. All vectors of Bartonella species will need to be identified before it can be determined if co-evolution of the organisms has occurred [34].

Geographic specificity and phylogeny

Within the genus Bartonella, gltA-based analyses suggest that bacteria associated with indigenous rodents from the New World (such as cotton rats S. hispidus, rice rats O. palustris and mice Phylottis peruviana) are phylogenetically distinct from species that have been recovered from Old World rodents (such as rats R. norvegicus and R. rattus, mice Mus musculus and voles Clethrionomys glareolus) (Table 1, groups D and F, respectively) [16]. Such a phylogenetic divergence between New World and Old World Bartonella species suggests that species that were separated as a result of the continental drift may have evolved independently of one-another. Consequently, Bartonella species may have a restricted geographical distribution in as far as their animal hosts and vectors have never been exported from one continent to another. B. grahamii, B. doshiae, B. talpae and B. taylorii for the Old World, as well as B. vinsonii subsp. vinsonii and B. vinsonii subsp. arupensis for the New World, may illustrate a situation of geographical segregation resulting from the restricted distribution of their specific mammalian reservoir hosts [3,8,32,34]. The strictly limited distribution of B. bacilliformis, however, might be related to the restricted geographical area of its sandfly vector, as no animal host was found in endemic areas and humans are regarded as the most probable reservoir for this species [1,11,17].

Conversely, dissemination of infected animals or vectors resulted in the widespread extension of their associated parasites. B. quintana, B. henselae and probably B. vinsonii subsp. berkhoffii, which are associated with ubiquist hosts (human, cat and dog, respectively), harbor a widespread extension (Table 1) [4,30,35]. While originally native from the Old World, rats from the genus Rattus were introduced into the New World during the Spanish Conquest and may have carried their own Bartonella strains [1,16]. However, data currently available regarding the geographical distribution of Bartonella should be interpreted carefully, as efforts to identify these organisms to the species level are not equally active in all regions of the world. The abundant information concerning Bartonella in the USA and Europe contrasts with only very limited data derived from Asia or Africa. Further characterization of the available Bartonella isolates and efforts to identify bacteria in other animals and locations, will help in the understanding of the geographical distribution of the genus.

Conclusion

Despite recent studies, the phylogeny of Bartonella remains confusing and it is likely that this may be exacerbated as new isolates of Bartonella continue to be made in large numbers from mammal populations [14,,,17,22]. Although 16S rDNA-based phylogenetic studies give little information on evolutionary processes occurring within the genus Bartonella, studies using comparisons of ITS, gltA and groEL sequences have enabled six distinct evolutionary clades to be identified within the genus. As multiple gene data sets have been shown to be useful in determining the precise molecular relationships between genera and species [18], comparisons of the sequences of the ribC, the ftsZ and the 17 kDa antigen genes might be used to improve our understanding of the evolutionary process, but this will only be possible when such data become available for all Bartonella species. This will specifically help in the identification of lateral gene transfer, which occurs more commonly in intracellular pathogens [36].

Although still partial, the phylogenetic data available for Bartonella species suggest that a phenomenon of co-evolution occurred between the bacteria and their animal reservoir hosts. This host specificity may be partially offset by concomitant vector specificity. As a consequence of the restricted geographical distribution of their hosts and vectors, geographically distant Bartonella species may have evolved separately. It would be interesting to combine genetic classifications with phenotypic data sets (such as duration of experimental bacteremia [33]) to better understand the natural history of Bartonella species, their association with specific mammal hosts, and to confirm the apparent divisions identified within the genus.

Acknowledgements

The authors thank Alia Ben Kahla for her help in the computer-assisted construction of phylogenetic trees.

References

[1]
Birtles
R.J.
Canales
J.
Ventosilla
P.
Alvarez
E.
Guerra
H.
Llanos-Cuentas
A.
Raoult
D.
Doshi
N.
Harrison
T.G.
(
1999
)
Survey of Bartonella species infecting intradomicillary animals in the Huayllacallan valley, Ancash, Peru, a region endemic for human bartonellosis
.
Am. J. Trop. Med. Hyg.
 
60
,
799
805
.
[2]
Brenner
D.J.
O'Connor
S.P.
Winkler
H.H.
Steigerwalt
A.G.
(
1993
)
Proposals to unify the genera Bartonella and Rochalimaea, with descriptions of Bartonella quintana comb. nov., and Bartonella elizabethae comb. nov., and to remove the family Bartonellaceae from the order Rickettsiales
.
Int. J. Syst. Bacteriol.
 
43
,
777
786
.
[3]
Birtles
R.J.
Harrison
T.G.
Saunders
N.A.
Molyneux
D.H.
(
1995
)
Proposals to unify the genera Grahamella and Bartonella, with descriptions of Bartonella talpae comb. nov., Bartonella peromysci comb. nov., and three new species, Bartonella grahamii sp. nov., Bartonella taylorii sp. nov., and Bartonella doshiae sp. nov
.
Int. J. Syst. Bacteriol.
 
45
,
1
8
.
[4]
Anderson
B.E.
Neuman
M.A.
(
1997
)
Bartonella spp. as emerging human pathogens
.
Clin. Microbiol. Rev.
 
10
,
203
219
.
[5]
Kordick
D.L.
Hilyard
E.J.
Hadfield
T.L.
Wilson
K.H.
Steigerwalt
A.G.
Brenner
D.J.
Breitschwerdt
E.B.
(
1997
)
Bartonella clarridgeiae, a newly recognized zoonotic pathogen causing inoculation papules, fever, and lymphadenopathy (cat scratch disease)
.
J. Clin. Microbiol.
 
35
,
1813
1818
.
[6]
Roux
V.
Eykyn
S.J.
Wyllie
S.
Raoult
D.
(
2000
)
Bartonella vinsonii subsp. berkhoffii as an agent of afebrile blood culture-negative endocarditis in a human
.
J. Clin. Microbiol.
 
38
,
1698
1700
.
[7]
Daly
J.S.
Worthington
M.G.
Brenner
D.J.
Moss
C.W.
Hollis
D.G.
Weyant
R.S.
Steigerwalt
A.G.
Weaver
R.E.
Daneshvar
M.I.
O'Connor
S.P.
(
1993
)
Rochalimaea elizabethae sp. nov. isolated from a patient with endocarditis
.
J. Clin. Microbiol.
 
31
,
872
881
.
[8]
Welch
D.F.
Carroll
K.C.
Hofmeister
E.K.
Persing
D.H.
Robinson
D.A.
Steigerwalt
A.G.
Brenner
D.J.
(
1999
)
Isolation of a new subspecies, Bartonella vinsonii subsp. arupensis, from a cattle rancher identity with isolates found in conjunction with Borrelia burgdorferi and Babesia microti among naturally infected mice
.
J. Clin. Microbiol.
 
37
,
2598
2601
.
[9]
Kerkhoff
F.T.
Bergmans
A.M.
Van der Zee
A.
Rothova
A.
(
1999
)
Demonstration of Bartonella grahamii DNA in ocular fluids of a patient with neuroretinitis
.
J. Clin. Microbiol.
 
37
,
4034
4038
.
[10]
Heller
R.
Riegel
P.
Hansmann
Y.
Delacour
G.
Bermond
D.
Dehio
C.
Lamarque
F.
Monteil
H.
Chomel
B.
Piémont
Y.
(
1998
)
Bartonella tribocorum sp. nov., a new Bartonella species isolated from the blood of wild rats
.
Int. J. Syst. Bacteriol.
 
48
,
1333
1339
.
[11]
Droz
S.
Chi
B.
Horn
E.
Steigerwalt
A.G.
Whitney
A.M.
Brenner
D.J.
(
1999
)
Bartonella koehlerae sp. nov., isolated from cats
.
J. Clin. Microbiol.
 
37
,
1117
1122
.
[12]
Heller
R.
Kubina
M.
Mariet
P.
Riegel
P.
Delacour
G.
Dehio
C.
Lamarque
F.
Kasten
R.
Boulouis
H.J.
Monteil
H.
Chomel
B.
Piémont
Y.
(
1999
)
Bartonella alsatica sp. nov., a new Bartonella species isolated from the blood of wild rabbits
.
Int. J. Syst. Bacteriol.
 
49
,
283
288
.
[13]
Breitschwerdt
E.B.
Kordick
D.L.
(
2000
)
Bartonella infection in animals: carriership, reservoir potential, pathogenicity, and zoonotic potential for human infection
.
Clin. Microbiol. Rev.
 
13
,
428
438
.
[14]
Chang
C.
Chomel
B.B.
Kasten
R.W.
Heller
R.
Kocan
K.M.
Ueno
H.
Yamamoto
K.
Bleich
V.C.
Pierce
B.M.
Gonzales
B.J.
Swift
P.K.
Boyce
W.M.
Jang
S.S.
Boulouis
H.J.
Piémont
Y.
(
2000
)
Bartonella spp. isolated from wild and domestic ruminants in North America
.
Emerg. Infect. Dis.
 
6
,
306
311
.
[15]
Kosoy
M.Y.
Regnery
R.L.
Tzianabos
T.
Marston
E.L.
Jones
D.C.
Green
D.
Maupin
G.O.
Olson
J.G.
Childs
J.E.
(
1997
)
Distribution, diversity, and host specificity of Bartonella in rodents from the southeastern United States
.
Am. J. Trop. Med. Hyg.
 
57
,
578
588
.
[16]
Ellis
B.A.
Regnery
R.L.
Beati
L.
Bacellar
F.
Rood
M.
Glass
G.G.
Marston
E.
Ksiazek
T.G.
Jones
D.
Childs
J.E.
(
1999
)
Rats of the genus Rattus are reservoir hosts for pathogenic Bartonella species: an old world origin for a new world disease
.
J. Infect. Dis.
 
180
,
220
224
.
[17]
Birtles
R.J.
Raoult
D.
(
1996
)
Comparison of partial citrate-synthase gene (gltA) sequences for phylogenetic analysis of Bartonella species
.
Int. J. Syst. Bacteriol.
 
46
,
891
897
.
[18]
Marston
E.L.
Sumner
J.W.
Regnery
R.L.
(
1999
)
Evaluation of intraspecies genetic variation within the 60 kDa heat-shock protein gene (groEL) of Bartonella species
.
Int. J. Syst. Bacteriol.
 
49
,
1015
1023
.
[19]
Bereswill
S.
Hinkelmann
S.
Kist
M.
Sander
A.
(
1999
)
Molecular analysis of riboflavin synthesis genes in Bartonella henselae and use of the ribC gene for differentiation of Bartonella species by PCR
.
J. Clin. Microbiol.
 
37
,
3159
3166
.
[20]
Kelly
T.M.
Padmalayam
I.
Baumstarck
B.R.
(
1998
)
Use of the cell division protein ftsZ as a means of differentiating among Bartonella species
.
Clin. Diagn. Lab. Immunol.
 
5
,
766
772
.
[21]
Olsen
G.J.
Woese
C.R.
(
1993
)
Ribosomal RNA: a key to phylogeny
.
FASEB J.
 
7
,
113
123
.
[22]
Birtles
R.J.
Hazel
S.
Bown
K.
Raoult
D.
Begon
M.
Bennett
M.
(
2000
)
Subtyping of uncultured bartonellae using sequence comparison of 16S/23S rRNA intergenic spacer regions amplified directly from infected blood
.
Mol. Cell. Probab.
 
14
,
79
87
.
[23]
Roux
V.
Raoult
D.
(
1995
)
The 16S-23S rRNA intergenic spacer region of Bartonella (Rochalimaea) species is longer than usually described in other bacteria
.
Gene
 
156
,
107
111
.
[24]
Felsenstein
J.
(
1993
)
PHYLIP phylogeny inference package (version 3.5c)
.
Cladistics
 
5
,
164
166
.
[25]
Berry
V.
Gascuel
O.
(
1996
)
On the interpretation of bootstrap trees: appropriate threshold of clade selection and induced gain
.
Mol. Biol. Evol.
 
13
,
999
1011
.
[26]
Dessen
P.
Fondrat
C.
Valencien
C.
Munier
G.
(
1990
)
BISANCE: a French service for access to biomolecular sequences databases
.
CABIOS
 
6
,
355
356
.
[27]
Jensen
W.A.
Majilinde
Z.F.
Rooney
J.
Kordick
D.L.
Breitschwerdt
E.B.
(
2000
)
Rapid identification and differentiation of Bartonella species using a single-step PCR assay
.
J. Clin. Microbiol.
 
38
,
1717
1722
.
[28]
Sweger
D.
Resto-Ruiz
S.
Johnson
D.P.
Schmiederer
M.
Hawke
N.
Anderson
B.
(
2000
)
Conservation of the 17-kilodalton antigen gene within the genus Bartonella
.
Clin. Diagn. Lab. Immunol.
 
7
,
251
257
.
[29]
Birtles
R.J.
Harrison
T.G.
Fry
N.K.
Saunders
N.A.
Taylor
A.G.
(
1991
)
Taxonomic considerations of Bartonella bacilliformis based on phylogenetic and phenotypic characteristics
.
FEMS Microbiol. Lett.
 
83
,
187
192
.
[30]
Kordick
L.K.
Swaminathan
B.
Greene
C.E.
Wilson
K.H.
Whitney
A.M.
O'Connor
S.
Hollis
D.G.
Matar
G.M.
Steigerwalt
A.G.
Malcolm
G.B.
Hayes
P.S.
Hadfield
T.L.
Breitschwerdt
E.B.
Brenner
D.J.
(
1996
)
Bartonella vinsonii subsp. berkhoffii subsp. nov., isolated from dogs; Bartonella vinsonii subsp. vinsonii; and emended description of Bartonella vinsonii
.
Int. J. Syst. Bacteriol.
 
46
,
704
709
.
[31]
Kosoy
M.Y.
Saito
E.K.
Green
D.
Marston
E.L.
Jones
D.C.
Childs
J.E.
(
2000
)
Experimental evidence of host specificity of Bartonella infection in rodents
.
Comp. Immunol. Microbiol. Infect. Dis.
 
23
,
221
238
.
[32]
Birtles
R.J.
Harrison
T.G.
Molyneux
D.H.
(
1994
)
Grahamella in small woodland mammals in the U.K.: isolation, prevalence and host specificity
.
Ann. Trop. Med. Parasitol.
 
88
,
317
327
.
[33]
Lucey
D.
Dolan
M.J.
Moss
C.W.
Garcia
M.
Hollis
D.G.
Wegner
S.
Morgan
G.
Almeida
R.
Leong
D.
Greisen
K.S.
Welch
D.F.
Slater
L.N.
(
1992
)
Relapsing illness due to Rochalimaea henselae in immunocompetent hosts: implication for therapy and new epidemiological associations
.
Clin. Infect. Dis.
 
14
,
683
688
.
[34]
Hofmeister
E.K.
Kolbert
C.P.
Abdulkarim
A.S.
Magera
J.M.
Hopkins
M.K.
Uhl
J.R.
Ambyaye
A.
Telford
S.R.
III
Cockerill
F.R.
III
Persing
D.H.
(
1998
)
Cosegregation of a novel Bartonella species with Borrelia burgdorferi and Babesia microti in Peromyscus leucopus
.
J. Infect. Dis.
 
177
,
409
416
.
[35]
Baker
J.A.
(
1946
)
A rickettsial infection in Canadian voles
.
J. Exp. Med.
 
84
,
37
51
.
[36]
Doolittle
W.F.
(
1999
)
Phylogenetic classification and the universal tree
.
Science
 
284
,
2124
2128
.
[37]
Chomel
B.B.
Kasten
R.W.
Floyd-Hawkins
K.A.
Chi
B.
Yamamoto
K.
Roberts-Wilson
J.
Gurfield
A.N.
Abbott
R.C.
Pedersen
N.C.
Koehler
J.E.
(
1996
)
Experimental transmission of Bartonella henselae by the cat flea
.
J. Clin. Microbiol.
 
23
,
1952
1956
.