Introduction
Bartonella are gram-negative bacteria classified within the alpha-Proteobacteria. Over the last decade the number of identified Bartonella spp. has increased rapidly and the known diversity of the Bartonellaceae family continues to expand, as recently demonstrated in various mammals 1,2, including humans 3. The geographical distribution of Bartonella species depends on their hosts and vectors. Mammalian species, such as cats, dogs, rodents and ruminants, are the main bartonellae reservoirs. Our objective was to detect and identify, by molecular methods, the presence of Bartonella species in wild rodents from the Lao PDR (Laos), a country where epidemiological and clinical studies of zoonoses are scarce.
Methods
Mammal samples
Rodents were trapped using cage and break-back traps and purchased from local markets in June and December 2006 in four Lao provinces. All specimens were photographed, vouchered (specimens will be registered in the Australian National Wildlife Collection, CSIRO Sustainable Ecosystems, Canberra, Australia) and identified subsequently (by KA) 4. Their livers and spleens were stored in ethanol prior to laboratory analysis for the detection of Bartonella.
Molecular methods
Total genomic DNA was extracted from livers and spleens with a QIAamp Tissue kit (QIAGEN, Hilden, Germany), as described by the manufacturer. Samples were handled under sterile conditions to avoid the risk of cross-contamination.
Samples were screened for the presence of Bartonella species DNA by the use of a real-time (RT) quantitative PCR targeting the ITS gene. Positive samples at screening were further studied by PCR amplification and DNA sequencing of the genes encoding the 16S–23S rDNA intergenic spacer (ITS), citrate synthase (gltA), β-subunit of the RNA polymerase (rpoB) using previously described primers and conditions 5. Similarity rates between all the identified Bartonella species were determined for each gene to assess the taxonomic position of Bartonella isolates at the species levels. Phylogenetic relationships among the studied bartonellae were inferred from sequence alignments of each gene and from concatenated gene sequences using the maximum parsimony and neighbour-joining methods within the MEGA version 3.1 software package (Megasoftware, Tempe, AZ, USA).
Results
We examined 568 tissue samples (311 livers and 257 spleens) from 310 rodents and one tree shrew (see Table 1). One hundred and twenty-eight tissue samples (22.5%), corresponding to 79 rodents (25.5%), were positive for Bartonella by RT-PCR. Positive and negative controls showed expected results in all tests. Using standard ITS PCR and sequencing we identified the presence of five Bartonella species, including B. phoceensis, B. elizabethae, B. tribocorum, and two new Bartonella species that we named Lao/Nh1 and Lao/Nh2, in three different regions of Laos (Table 1). Lao/Nh1 shared less than 96.0% sequence similarity for a 327-bp gltA fragment of the validated Bartonella spp. and less than 95.4% for a 825-bp rpoB fragment. Lao/Nh2 shared less than 95.4% for a 825-bp rpoB fragment of the validated Bartonella spp. and a 327-bp gltA fragment with 97.8% similarity with B. tribocorum.
Table 1.
Species of rodents and Bartonella species detected at different collection sites in Lao PDR
| Location, % rodents with Bartonellae | Rodent species | No. of spleen/liver | Bartonella | % all rodents with Bartonella |
|---|---|---|---|---|
| Vientiane City, 15.5% | Rattus rattus | 79/77 | B. phoceensis (1) and B. eizabethae (2), B. tribocorum (2) and Lao/Nh2 (3) | 10.1% |
| R. exulans | 23/23 | B. elizabethae (1), B. phoceensis (1) and B. tribocorum (5) | 30.4% | |
| Mus cervicolor | 6/6 | Lao/Nh2 (1) | 16.7% | |
| Mus caroli | 2/2 | B. phoceensis (1) | 50% | |
| Luang Prabang, 11.2% | R. rattus | 141/139 | B. elizabethae (4), B. phocensis (8), Lao/Nh2 (8), Lao/Nh1 (1) | 9.1% |
| Cannomys badius | 2 | Lao/Nh2 (1) | 50% | |
| B. indica | 2 | 0 | 0 | |
| M. caroli | 1 | 0 | 0 | |
| Champasak, 12.4% | M. cervicolor | 64/65 | 0 | 0 |
| R. rattus | 1/1 | 0 | 0 | |
| R. exulans | 56/56 | B. phoceensis (2), B. elizabethae (1), B. tribocorum (1) and Lao/Nh2 (6) | 17.9% | |
| Petinomys phayrei | 2/2 | 0 | 0 | |
| Luang Nam Tha, 19.2% | R. rattus | 160/159 | B. phoceensis (11), B. elizabathae (2), B. tribocorum (3), Lao/Nh1 (5), Lao/Nh2 (11) | 20.1% |
| R. argentiventer | 1/1 | 0 | 0 | |
| Rhizomys sumatrensis | 2/3 | 0 | 0 | |
| Callosciurus erythraeus | 2/2 | 0 | 0 | |
| Dremomys rufigenis | 1/1 | 0 | 0 | |
| B. savilei | 1/1 | 0 | 0 | |
| R. exulans | 1/1 | 0 | 0 |
Discussion
Comparison of DNA sequences has been the most commonly used approach for Bartonella species identification as the members of the genus Bartonella are fastidious bacteria that possess few phenotypic markers that are useful for species delineation. We used the gltA, the rpoB and the ITS genes to determine the taxonomic status of Bartonella strains from Laos as these genes have good discriminating power 5. According to current molecular criteria 5, Lao/Nh1 and Lao/Nh2 could be defined as new species. Using concatenation of the sequences obtained, Lao/Nh1 clustered with B. birtlesii and B. taylorii whereas Lao/Nh2 clustered with B. tribocorum. Finally, Lao/Nh1 was detected only in Rattus rattus collected from Luangnatha Province and Luang Prabang. It is yet unknown whether these Bartonella species strains from Laos are human pathogens and this should be investigated.
To the best of our knowledge, this is the first study of the prevalence of Bartonella species in rodents in Lao PDR. We found a high prevalence of Bartonella spp. (25.5%), including B. elizabethae, B. tribocorum and B. phoceensis. B. elizabethae has been already isolated in small mammals from Bangladesh 2 and B. tribocorum in rodent fleas from China 6. Interestingly, B. tamiae, a newly recognised pathogen isolated from three human patients from Thailand 3, was not detected in Lao rodents. Concerning B. phoceensis, it is its first molecular detection in this part of the world.
Bartonella spp. were identified in rodents derived from both agrarian and urban environments. Rodents are abundant in both contexts in Laos, hence a large proportion of the human population lives in close contact with rodents. Accordingly, it is possible that some unknown diseases in Laos may be caused by Bartonella spp. Further investigations are warranted in order to isolate these new Bartonella spp. and to determine if they can cause any clinical manifestations.
Acknowledgments
We thank the technicians of molecular biology for their technical support.
Genbank accession numbers for the new isolates are: Lao/Nh1 glta EU755060, Lao/Nh1 rpoB EU714973, Lao/Nh1 ITS EU714976, Lao/Nh2 glta EU755061, Lao/Nh2 rpoB EU714974, Lao/Nh2 ITS EU714977.
References
- 1.Fournier PE, Taylor C, Rolain JM, Barrassi L, Smith G, Raoult D. Bartonella australis sp. nov. from kangaroos, Australia. Emerg Infect Dis. 2007;13(12):1961–1962. doi: 10.3201/eid1312.060559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bai Y, Montgomery SP, Sheff KW, et al. Bartonella strains in small mammals from Dhaka, Bangladesh, related to Bartonella in America and Europe. Am J Trop Med Hyg. 2007;77(3):567–570. [PubMed] [Google Scholar]
- 3.Kosoy M, Morway C, Sheff KW, et al. Bartonella tamiae sp. nov., a newly recognized pathogen isolated from three human patients from Thailand. J Clin Microbiol. 2008;46(2):772–775. doi: 10.1128/JCM.02120-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Aplin KP, Brown PR, Jacob J, Krebs C, Singleton GR. Field Methods for Rodent Studies in Asia and the Pacific. ACIAR Monograph No 100. Canberra: Australian Centre for International Agricultural Research; 2004. p. pp397. [Google Scholar]
- 5.La Scola B, Zeaiter Z, Khamis A, Raoult D. Gene-sequence-based criteria for species definition in bacteriology: the Bartonella paradigm. Trends Microbiol. 2003;11(7):318–321. doi: 10.1016/s0966-842x(03)00143-4. [DOI] [PubMed] [Google Scholar]
- 6.Li DM, Liu QY, Yu DZ, Zhang JZ, Gong ZD, Song XP. Phylogenetic analysis of Bartonella detected in rodent fleas in Yunnan, China. J Wildl Dis. 2007;43(4):609–617. doi: 10.7589/0090-3558-43.4.609. [DOI] [PubMed] [Google Scholar]