In this study, we investigated infections with Bartonella
spp. in water buffaloes from Thailand and cattle from five countries across the world. We provided the first evidence that water buffaloes are also infected with Bartonella
spp., with a prevalence of 7%. In cattle, bartonella prevalence varied widely between the studied countries and between different regions in the same country. In Japan, Kenya, and the Mestia district of Georgia, the cattle were free from the infection; in contrast, infection prevalence reached as high as 90% in the Marneuli district of Georgia, and a moderate prevalence was observed in other areas. It is of interest to note that the prevalence of infection in cattle varied widely across districts in Georgia, from apparent absence to nearly 100%. The observed variation in bartonella prevalence in cattle between different countries/regions can be the result of multiple factors. Since bartonellae are mainly vector-transmitted [11
], we speculate that the distribution and abundance of specific arthropods play a major role in this matter. It may be that a heavy level of infestation is required for transmission. Cattle ticks (Rhipicephalus microplus
) and biting flies (Diptera spp.) are implicated as potential vectors that may transmit bartonella between cattle [31
]. Interestingly, low ectoparasite infestation on cattle in Japan was noticed (Maruyama, personal communication). Such a fact could provide a plausible explanation for the absence of bartonella infection observed in cattle from Japan. In addition to ectoparasites, environmental factors, such as geographic characters, landscape, etc., may also have influences on bartonella prevalence. Of the three districts in Georgia, Mestia district (with no bartonella infection in cattle) is located in a mountainous area in the northwest of the country, while the other two districts (with high bartonella infection in cattle) are located in lower-lying areas in the northeast and southeast.
The cultures obtained from the study belonged to three Bartonella
spp. based on gltA
sequences. The majority of strains (172/182), identified as B. bovis
, were the most common in cattle from all investigated places. This is concordant with previously reported observations. The other two species, B. chomelii
and B. schoenbuchensis
, are not common, although both have been previously reported in cattle [17
]. Interestingly, both of these species were only found in cattle from Georgia. This may suggest that the bartonella community associated with cattle in Georgia is more diverse comparing to other places. All nine B. chomelii
isolates were obtained from cattle in Dusheti District located in northeastern Georgia. These observations may be associated with the composition of the local ectoparasite community. Future studies of cattle ectoparasites should test hypotheses about whether any particular arthropod species act as vectors for bartonella transmission between cattle. Finally, all isolates obtained from water buffaloes in Thailand also belong to B. bovis
, indicating that water buffaloes can also serve as hosts for B. bovis
Although B. bovis is known to be widely distributed in bovines, no study has demonstrated how strains from different geographic areas and hosts vary. As a potential pathogen for domestic animals, information about genetic diversity of B. bovis is important. In this study, we developed an MLST scheme for B. bovis based on nine loci to characterize B. bovis strains among geographically diverse populations. Resolving into 22 STs among the 28 strains from different geographical regions, our MLST data demonstrated B. bovis strains can be genetically different. Although some STs can be quite close, the same ST was never found in strains from different regions.
Multi-locus sequencing typing has been applied in a number of studies previously, in which the MLST was based on the comparison of multiple housekeeping gene sequences [28
]. In this study, we included ITS. Although non-functional, sequence comparison of the ITS region is widely used in taxonomy and molecular phylogeny for its ease of amplification and high degree of variation, even between closely-related species. Our data showed either insertion or deletion has occurred in some of the tested strains, which likely are associated with host species. Such results suggested that ITS might be more powerful in differentiating genetic diversities among strain populations, compared to some regularly-used housekeeping genes.
Identification of three close but distinct lineages among the STs suggests a clonal population structure for the species. All STs of cattle-originated strains fall into either lineage I or II, while STs of water buffalo-originated strains exclusively fall in lineage III, showing the specific host relationship. Meanwhile, we suspect that lineage I and lineage II each is associated with a particular lineage of Bos taurus. By general understanding, cattle from US, Europe, and Georgia belong to the ‘taurine’ lineage (former B. taurus subsp. taurus); while those from Thailand belong to the ‘zebu’ lineage (former B. taurus subsp. indicus); and in Guatemala, cattle were of mixed breeds. Based on this information, we hypothesize that lineage I is associated with cattle of ‘taurine’ lineage and lineage II is associated with cattle of ‘zebu’ lineage. Further studies are required to confirm this hypothesis.
MLST analysis also showed B. bovis strains retain geographical particularity. This observation can be explained by the association of B. bovis with specific breed of cattle as well. For example, all cattle-originated strains from Thailand belong to lineage II; while all strains from Georgian cattle together with strains from the US and France, belong to lineage I. Contrastingly, strains from Guatemalan cattle fall into both lineages I and II. Mixed breed of cattle in Guatemala likely is the cause of crossing lineages.