The Distribution and Diversity of Bartonella Species in Rodents and Their Ectoparasites across Thailand

Our study highlights the surveillance of Bartonella species among rodents and their associated ectoparasites (ticks, fleas, lice, and mites) in several regions across Thailand. A total of 619 rodents and 554 pooled ectoparasites (287 mite pools, 62 flea pools, 35 louse pools, and 170 tick pools) were collected from 8 provinces within 4 regions of Thailand. Bandicota indica (279), Rattus rattus (163), and R. exulans (96) were the most prevalent species of rats collected in this study. Real-time PCR assay targeting Bartonella-specific ssrA gene was used for screening and each positive sample was confirmed by PCR using nuoG gene. The prevalence of Bartonella DNA in rodent (around 17%) was recorded in all regions. The highest prevalence of Bartonella species was found in B. savilei and R. rattus with the rate of 35.7% (5/14) and 32.5% (53/163), respectively. High prevalence of Bartonella-positive rodent was also found in B. indica (15.1%, 42/279), and R. norvegicus (12.5%, 5/40). In contrast, the prevalence of Bartonella species in ectoparasites collected from the rats varied significantly according to types of ectoparasites. A high prevalence of Bartonella DNA was found in louse pools (Polyplax spp. and Hoplopleura spp., 57.1%) and flea pools (Xenopsylla cheopis, 25.8%), while a low prevalence was found in pools of mites (Leptotrombidium spp. and Ascoschoengastia spp., 1.7%) and ticks (Haemaphysalis spp., 3.5%). Prevalence of Bartonella DNA in ectoparasites collected from Bartonella-positive rodents (19.4%) was significantly higher comparing to ectoparasites from Bartonella-negative rodents (8.7%). The phylogenetic analysis of 41 gltA sequences of 16 Bartonella isolates from rodent blood and 25 Bartonella-positive ectoparasites revealed a wide range of diversity among Bartonella species with a majority of sequences (61.0%) belonging to Bartonella elizabethae complex (11 rodents, 1 mite pool, and 5 louse pools), while the remaining sequences were identical to B. phoceensis (17.1%, 1 mite pool, 5 louse pools, and 1 tick pool), B. coopersplainensis (19.5%, 5 rodents, 1 louse pool, and 2 tick pools), and one previously unidentified Bartonella species (2.4%, 1 louse pool).


Introduction
Bartonella bacteria are new emerging pathogens causing diseases in humans and animals [1,2]. The members of genus Bartonella are rod-shaped gram negative facultative intracellular bacteria that are fastidious and slow growing at aerobic conditions. They infect human and other mammalian hosts via infected-vectors such as fleas, ticks, and lice or the bite/scratch of an infectedanimal [3][4][5]. Moreover, the infected arthropods could transmit Bartonella bacteria to human and other mammalian hosts via feces through superficial scratches in skin [6]; for example, B. henselae and B. quintana were transmitted to hosts via contaminated feces of infected cat fleas (Ctenocephalides felis) and human body lice (Pediculus humanus), respectively [7]. Pathogenesis involves the invasion of host's erythrocytes, endothelial cells, and dendritic cells which play an important role in the first line immune response to fight against pathogens [8,9]. As a result of the immune system failure, a bacteremia persistent infection might occur [8,10].
Bartonella genus comprises over 30 species and subspecies [11]. At least thirteen known or suspected species are thought to contribute to blood-borne infections in human [12]. Moreover, several studies suggested the role of Bartonella species as a potential causative agent for cases of unknown febrile illness as well as endocarditis in patients in Thailand [13]. The diversity of Bartonella species in several countries in Southeast Asia (Lao PDR, Cambodia, and Thailand) has been reported and the findings revealed that Bartonella species in rodents are much more diverse than in other animals, except bats. The species found in rodents included B. elizabethae, B. coopersplainsensis, B. phoceensis, B. queenslandensis, B. rattimassiliensis, B. tribocorum and three genotypes presumably representing new Bartonella species [14].
Bartonella transmission occurs mainly via horizontal transmission when arthropod vectors acquire Bartonella bacteria during the feeding of infected host and later they become infected and the infected vectors then transfer the bacteria to another host [5,15]. Interestingly, some studies suggested that vertical and transstadial transmissions of Bartonella species in Ixodes ticks [16], deer ked [17], and transplacental in rodent populations [18].
High prevalence of Bartonella DNA and genotype diversity have been detected in arthropod vectors around the world. For example, ticks collected from dogs and donkeys in Peru were found to carry several Bartonella species, such as B. rochalimae, B. quintana and B. elizabethae [19]. In Taiwan, B. tribocorum, B. elizabethae, B. queenslandensis, B. rochalimae-like bacteria, B. phoceensis, and B. rattimassiliensis were detected in fleas and louse pools [20]. Several studies in Thailand have reported the detection of B. henselae, B. clarridgeiae, and B. koehlerae from cats and flea pools collected from the Thai-Myanmar border [21] and in the Bangkok area [22]. Moreover, novel species such as B. tamiae was recently isolated from whole blood of febrile patients from Thailand [23] and DNA belonging to this species was also detected from the pools of ticks and mites collected from rats in Thailand [15].
Though a number of papers on Bartonella in rodents from Thailand have been published, the comparative analysis of bartonellae between rodent hosts and ectoparasites has not been done. Our aim was to investigate the prevalence and diversity of Bartonella species in rodents and their ectoparasites, and to estimate the importance of this host-vector relationship for the transmission of Bartonella species in natural habitats of Thailand. Our results indicated a significant difference between bacterial communities recognized in mammals and arthropods.

Study sites and samples processing
The study sites were located in different regions of Thailand. Rodents and their associated ectoparasites (ticks, fleas, mites, and lice) were collected from eight provinces within four regions of Thailand during the period of December 2012 to November 2013. The regions included the Northern region (Chiang Rai and Phayao provinces), the Southern region (Chumphon and Surat Thani provinces), the Eastern region (Rayong and Trat provinces) and the Northeastern region (Loei and Nong Bua Lam Phu provinces) ( Table 1). This study was carried out on private lands and the owners of the lands gave permission to conduct the study on their sites and the field studies did not involve endangered or protected species. Rodents were captured by live traps baited with bananas or dried fish. Rodents were collected from orchards, cultivated rice-fields, grassland areas, edges of dense forest, stream margins, and around houses. Traps were set for 3-5 nights and were checked early in the morning. Then, rodents were removed from the traps and later identified to species [24]. Captured rodents were killed by carbon dioxide and processed on the same day and at the site of capture. Blood and serum samples and rodent tissue samples (liver, spleen, kidney and lung) were collected and stored on dry ice, and transported to the AFRIMS laboratory. Rodent's ears were cut and stored in 70% ethanol for mite collections and the other ectoparasites (ticks, fleas, and lice) were collected from individual rodents by combing and stored in 70% ethanol for transportation to the laboratory. Mites in their larval stage (chigger) were collected from rodent's ears by paintbrush under the stereomicroscope and pooled by host. Three to five mites were selected from each pool and mounted on glass slides for morphological identification to genera and species if possible using taxonomic key [25]. Ectoparasites of each type (fleas, ticks, and lice) were identified morphologically [26,27] and pooled by host, type, stage, and gender in 1.5 ml microcentrifuge tube. Pools of ectoparasites were subjected to DNA extraction procedures as described below. Louse species identification was performed following the previously published protocol [28]. Details of the ectoparasites collected from rodents in this study are provided in Table A in S1 File. Genomic DNA extraction from rodent tissue and ectoparasites Genomic DNA was extracted from rodent livers using the Wizard1 Genomic DNA purification kit (Promega, Madison, WI) according to the manufacturer's instructions with some modifications. Briefly, the liver tissue was cut into pieces of approximately 3 millimeters in diameter and added to 600 μl of Nuclei Lysis Solution (Promega, Madison, WI). The mixture was homogenized with beads using a TissueLyser II machine (Qiagen, Hilden, Germany) at 25 Hz for 5 min twice. Subsequently, the mixture was incubated with 20 μl of Proteinase K solution (20 mg/ml) at 55°C for 1 hr, and then with 3 μl of RNase A (10 mg/ml) at 37°C for 15 min. Then 200 μl of protein precipitation solution (Promega, Madison, WI) was added and mixed vigorously by vortex. The mixture was kept on ice for 5 min. Insoluble materials were removed by centrifugation at 13,000 rpm for 4 min and the supernatant was transferred to a new tube. DNA was precipitated by adding 600 μl of isopropanol and then centrifuged at 13,000 rpm for 1 min. DNA pellet was washed using 70% ethanol and dried by SpeedVac™ concentrator (Thermo Scientific, Waltham, MA). Two hundred microliters of EB buffer (10 mM Tris Cl, pH 8.5) were used to resuspend dried DNA and stored at -20°C until further analysis. DNA extraction from the ectoparasites was performed according to the tissue extraction protocol from QIAamp 1 DNA Mini Kit (Qiagen, Hilden, Germany) with some modification. Briefly, pools of ticks, fleas and lice were puncture in the presence of liquid nitrogen in a 1.5 ml microcentrifuge tube. Mites were punctured with a fine needle under microscopy. Next, ninety microliters of ATL lysis buffer were added to each sample and mixed thoroughly. Then, ten microliters of Proteinase K solution (20 mg/ml) was added and incubated at 56°C for 3 hr. One hundred microliters of AL buffer was added to the samples and mixed by pulse-vortexing for 15 sec then incubated at 70°C for 10 min. After that, 100 μl of absolute ethanol was added and the mixture was mixed by pulse-vortexing for 15 sec. Finally, the mixture was transferred to a QIAamp spin column and DNA was eluted in 50 l AE buffer. DNA solution was stored at -20°C until further analysis.
Bartonella detection in rodent tissues and ectoparasites DNA extracts obtained from rodent tissues and ectoparasites were screened for the presence of Bartonella species using real-time PCR assay (qPCR) with TaqMan probe. A genus-specific assay targeting a transfer-mRNA gene (ssrA) of Bartonella species was used in this study following previously published protocol [29]. The primer pair, ssrA-F/ ssrA-R, and ssrA-probe were used to amplify 301 bp fragment of ssrA gene. The qPCR reaction (25 μl) consisted of 12.5 μl Platinum1 Quantitative PCR SuperMix-UDG (Invitrogen, Grand Island, NY), 0.5 μM of each primer, and 0.1 μM of ssrA-Probe and 2 μl of DNA template or nuclease free water as non-template control. The qPCR conditions were performed as follows: UDG incubation at 50°C for 2 min and then initial denaturation at 95°C for 2 min followed by 45 cycles of 95°C for 15 sec and 60°C for 30 sec using the Chromo4™ Real-Time Detector (Bio-Rad, Hercules, CA). Every sample with a positive signal from the screening assay was subjected to confirmatory test. A different target gene of NADH Dehydrogenase Gamma Subunit gene (nuoG) of Bartonella species was used to confirm the positivity by conventional PCR assay. A 346 bp fragment of nuoG gene was amplified with nuoG-F and nuoG-R primer pair according to previously published protocol [30]. The PCR reaction mixture (25 μl) consisted of 1X of PCR buffer, 0.2 μM of each primer, 0.2 mM of dNTP, 1.25 U of Taq DNA Polymerase (Invitrogen, Grand Island, NY) and 5 μl of DNA template or nuclease free water as non-template control. PCR amplification was carried out using the Veriti1 96-well Thermal Cycler (Applied Biosystems, Foster City, CA) with the initial denaturation at 94°C for 3 min followed by 45 cycles of denaturation at 94°C for 45 sec, annealing at 55°C for 1 min and extension at 72°C for 1 min 30 sec, then incubated at 72°C for 10 min for the final extension step. The amplification product of 346 bp was observed with 1.5% agarose gel electrophoresis under the UV visualization.
Bartonella culture from rodent blood Selected rodents, which were Bartonella-positive by molecular assays, were subjected to Bartonella culture following previously published protocol [31]. At the field sample site, whole blood was collected from each rodent by cardiac puncture and preserved in EDTA. Rodent whole blood was kept in -70°C until use. Briefly, whole blood was retrieved from the -70°C freezer and thawed at 4°C. Then, homogeneous whole blood was diluted 1:4 in 1X Dulbecco's Phosphate Buffered Saline (GIBCO, Grand Island, NY) containing 5-10% Fungizone (GIBCO, Grand Island, NY). Diluted blood sample (0.1 ml) was pipetted onto Brain Heart Infusion agar plates containing 5% rabbit blood (BBL, Becton Dickinson Microbiology Systems, Cockeysville, MD). Four to five agar plates were kept in a plastic bag and incubated at 37°C, 5% CO 2 for up to 4 weeks. The agar plates were monitored once a week after the initial inoculation and once per three days after sub-culturing. Sub-culturing was continued until a pure culture was obtained. Bartonella-like colonies were recognized by colony morphology and then harvested into 10% sterile glycerol and kept in -70°C freezer for further confirmation and characterization. A portion of each Bartonella-like colony was subjected to DNA extraction using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) following manufacturer's instruction and confirmed to be Bartonella species by citrate synthase gene (gltA) sequence identity.

Citrate synthase gene (gltA) amplification
Amplification of Bartonella gltA gene was done using published primers, BhCS.781p and BhCS.1137n [32]. The PCR reaction (50 μL) consisted of 1X of PCR buffer, 2.5 mM of MgCl 2 , 0.2 μM of each primer, 0.2 mM of dNTP, 2.5 U of AmpliTaq Gold1 DNA Polymerase (Applied Biosystems, Foster City, CA) and 2.5 μl of DNA template or nuclease free water as non-template control. The amplification conditions were as follows: the initial denaturation at 95°C for 3 min followed by 35 cycles of denaturation at 95°C for 1 min, annealing at 56°C for 1 min and extension at 72°C for 1 min, then the final extension step at 72°C for 10 min. The amplification product (379 bp) was observed on 1.5% agarose gel electrophoresis.

DNA sequencing
Amplification products of ssrA or gltA genes were purified using QIAquick PCR Purification Kit (QIAGEN Inc., Valencia, CA) following the manufacturer's instruction and sent for sequencing at AITbiotech Pte. Ltd. (Singapore)

Statistical analyses
The difference in the prevalence of Bartonella DNA in ectoparasites collected from Bartonellapositive and Bartonella-negative rodents was confirmed by Chi-Square Test and the critical range (P < 0.005) was used. The statistical calculations were performed with IBM1 SPSS1 Statisic (version 22) software (Chicago, IL).

Sequence and Phylogenetic analysis
Sequence data were assembled using the Sequencher 5.1 software (Gene Code Corporation, Ann Arbor, MI) and the consensus sequences were used for analyses. Sequences were aligned and constructed a similarity matrix with the reference sequences of Bartonella species using Muscle algorithm implemented in MEGA 6.0 software [33]. Maximum likelihood (ML) trees based on Kimura's 2-parameter model (K2+G+I) were constructed using molecular evolutionary genetics analysis (MEGA) 6.0 software [33] and bootstrap analyses with 1,000 resamplings performed to test the robustness of the branching.

Ethics Statement
Rodents trapping were carried out in the different locations of the provinces according to the institutional animal collection protocol entitled "Field Sampling of Small Mammal (Orders: Erinaceomorpha; Soricomorpha; Scandentia; Macroscelidea and Rodentia) Populations to Support Zoonotic Diseases Surveillance and Ectoparasite Collection" (PN# 12-06) reviewed and approved by the USAMC-AFRIMS Institutional Animal Care and Use Committee (IACUC). All sampling procedures and experimental manipulations were reviewed and approved as part of obtaining the animal collection protocol (PN# 12-06). Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals, NRC Publication, 2011 edition.

Results
Prevalence rate of Bartonella species among rodent and their associated ectoparasites in Thailand  Table B in S1 File)], with evident decline in ticks (3.5%, 6/170 pools of female ticks, Table C in S1 File) and mites (1.7%, 5/287 pools). A high Bartonella prevalence was found in rats collected , fleas were identified as Xenopsylla cheopis, and ticks were identified as Haemaphysalis spp. Mites collected from rats were more diverse than other types of ectoparasites. Three to five mites were selected from each pool and morphologically identified to genus (subgenus), and species if possible. They were all identified as trombiculid mites (Trombiculidae family, Trombiculinae subfamily). The most predominant genera were as follows: Gahrliepia (39.9%), Leptotrombidium (34.3%), Ascoschoengastia (14.6%), Blankaartia (5.4%), Schoengastia (4.0%), Helenicula (1.5%) and Lorillatum (0.3%). Two major genera collected from rats were further identified to subgenus. The results showed that within the genus Gahrliepia, subgenus Walchia was the most prevalent followed by Schoengastiella, and then Gahrliepia. Almost all mites in Leptotrombidium genus belonged to subgenus Leptotrombidium.  Among the tested ectoparasites, a high prevalence of Bartonella DNA was detected in lice (57.1%, 20/35) and fleas (25.8%, 16/62). High prevalence of Bartonella DNA was detected in 17/28 louse pools (60.7%) collected from R. rattus and 13/45 flea pools collected from R. exulans. Mites and ticks were mostly collected from B. indica and R. rattus and only 1.7% (5/287) of mite pools and 3.5% (6/170) of tick pools were positive for Bartonella DNA. Among Bartonella-positive trombiculid mites, 3 pools were Leptotrombidium genus, 1 pool was Ascoschoengastia genus, and the last pool did not have slide for morphological identification.

Identification of Bartonella species in rodent hosts and their associated ectoparasites based on gltA sequence variations
A total of 16 Bartonella isolates were successfully cultured from 26 individual rodent blood samples. Bartonella species detected from rats and ectoparasites based on sequences and phylogenetic analyses of 318 bp gltA amplicon are presented in Table 5    the reference Bartonella gltA sequences of Bartonella species detected in this study are summarized in Table 6. Maximum-likelihood (ML) tree (Fig 1) shows the relationship between sequences generated from Bartonella-positive samples clustered into 8 different cladograms as described below. The majority of identified sequences (25/41 sequences) fell within B. elizabethae species complex. Within this B. elizabethae complex group (25 sequences), 15 B. rattimassiliensis were detected from 9 rats, 5 louse pools, and 1 mite pool sharing 96.5-100% identity with strain 16115 (AY515125) and strain THNA5-R09 (JX158360). Eight sequences of B. tribocorum were detected from 2 rats and 6 flea pools sharing 99.3-100% identity with strain IBS506 (AJ005494) and strain THSKR-020 (JX158363). One sequence of B. queenslandensis was detected from a flea pool sharing 99.6% identity with strain AUST/NH5 (EU111799). One sequence of undescribed Bartonella species within B. elizabethae complex was detected from a flea pool (#F1644) with a percent identity ranging from 96.8 to 97.1% to B. tribocorum strain IBS506 (AJ005494) and strain THSKR-020 (JX158363). The rest of the detected Bartonella sequences (16 sequences) were identified into eight B. coopersplainsensis which were detected from 5 rats, 2 tick pools, and 1 louse pool sharing 99.0-100% identity with strain AUST/NH20 (EU111803). Seven B. phoceensis were detected solely from ectoparasites: 5 in louse pools, 1 in a mite pool, and 1 in a tick pool sharing 98.4-100% identity with B. phoceensis strain 16120 (AY515126) and B. phoceensis strain BR07 (GU056197). Interestingly, one sequence obtained from Bartonella-positive louse pool (#L1026) was distantly related to the rest of the Bartonella species (66.3-68.2% identity) detected in this study, however, it was similar to some new strains of Bartonella species strain BCF03 (GU056190) and strain BR10 (GU056200) obtained from Taiwan with 99.3% identity ( Table 6). The distribution of Bartonella species found in the North and northeast seems to be a region-specific distribution. Four samples of B. tribocorum were found in the Northeast and 8 samples of B. coopersplainsensis were detected in the North, even though one B. rattimassiliensis was also detected in the North (Fig 1 and Table 5). In contrast, Bartonella species found in samples collected from the East and the South are quite diverse consisting of six different species; B. rattimassiliensis, B. queenslandensis, B. tribocorum, Bartonella species within B. elizabethae species complex, and a presumably new strain of Bartonella species.
In this study, we found that the distribution of Bartonella species among rodents supports a host-specific pattern. Thus, five B. coopersplainsensis were solely detected from B. indica. Almost all B. rattimassiliensis (8/9) were detected from R. rattus, while two B. tribocorum were detected from R. norvegicus only. However, no specific pattern has been observed among ectoparasites. For example, two different Bartonella species (B. rattimassiliensis and B. phoceensis) were detected in mite pools (Table 5)   Prevalence and distribution of Bartonella species detected in ectoparasites collected from Bartonella-positive and Bartonella-negative rats The difference of Bartonella DNA prevalence in ectoparasites collected from Bartonella-positive and Bartonella-negative rats was investigated and presented in Table 7. Among Bartonellapositive rats, 20/103 (19.4%) ectoparasite pools were positive for Bartonella species. In contrast, only 27/309 (8.7%) ectoparasite pools were positive from Bartonella-negative rats. Of these Bartonella-positive ectoparasites, louse pools possessed the highest prevalence of 65% (13/20) and 46.7% (7/15) for positive and negative rats, respectively (Table 7). In general, Bartonella DNA prevalence in ectoparasites collected from positive rats (19.4%) were higher significantly (Chi-Square Tests, P = 0.003) comparing to ectoparasites from negative rats (8.7%).

Discussion
Our study highlights the surveillance of Bartonella species among rodents and their associated ectoparasites (ticks, fleas, lice, and mites) in several regions across Thailand. The data demonstrated the high prevalence of Bartonella DNA in rats and their associated ectoparasites, especially in lice and fleas, as well as the finding of a diverse range of Bartonella species circulated among them. Previous studies have shown the high prevalence of Bartonella species among rats captured from Chiang Rai province (8.7%) [34] and from several regions across the country (41.5%) [35]. Although the prevalence in the latter study by Bai et [36].
Although Bartonella DNA was detected in all types of ectoparasites (ticks, lice, fleas, and mites) collected from rats in this study, the prevalence varied substantially between ectoparasite types. The highest rate was found in lice and fleas that were in agreement with the study conducted in Taiwan [20]. Though Bartonella DNA was found in lower prevalence in ticks and mite pools in both studies, Kabeya et al. reported high prevalence in mites (82.9%) collected from rats in Thailand and all Bartonella species detected from mites were identified into B. tamiae based on their DNA sequences [15]. Although we do not believe that rat-associated lice can transmit pathogens to humans because of their high specificity to the host, for example, Polyplax and Hoplopleura lice are more likely specific to Rattus rats or to some other relative rat species [37,38], the circulation of bartonellae via different ectoparasites, including lice, can influence the diversity of a rat-associated Bartonella community. A diverse range of Bartonella species were isolated from whole blood of rats and shrews, including B. rattimassiliensis, B. grahamii, B. elizabethae, B. tribocorum, B. coopersplainensis, B. phoceensis, B. queenslandensis, and unknown genogroup [20,32,35,36]. While only three of these species, B. rattimassiliensis, B. tribocorum, B. coopersplainensis, were isolated from rats in our study. Bartonella species found in rodent-associated ectoparasites were more diverse with seven different species found, including species isolated from their rat hosts. In this study, gltA gene sequence was used to identify Bartonella species, although ssrA gene was also sequenced and analyzed. We found that both genes were effective to identify Bartonella at the genus level; however, gltA gene sequence was found to be more suitable for species identification than ssrA gene sequence since the gltA sequence has expressed a higher range of variation comparing to the ssrA. Moreover, the availability of gltA reference sequences for Bartonella species in available databases is much higher than for ssrA genes that makes it a reliable and accurate tool for Bartonella species identification. Additionally, phylogenetic trees constructed from Bartonella culture isolates using both ssrA and gltA genes created quite similar phylogenetic tree topologies as shown in S1 Fig. Kabeya et al. [15] reported the detection of B. tamiae, previously isolated from Thai febrile illness patients, in mites and tick pools collected from rats in Thailand. Two dominant mite genus most infected with B. tamiae were Leptotrombidium (66.7%) and Schoengastia (78.6%), therefore, the author suggested the role of mite as potential vector for B. tamiae transmission to human patients. In this study, Bartonella DNA was also detected from mites of Leptotrombidium and Ascoschoengastia genera, although B. tamiae was not found in our study. Our finding supports the role of trombiculid mite as a vector for Bartonella transmission. Because of the difficulty of mite species identification, the mite species in each pool was determined based on 3-5 mites selected from each pool and mounted on a glass slide as mentioned in materials and methods.
From our study, we found that Bartonella positive ectoparasites from Bartonella-positive rats was higher than in Bartonella-negative rats suggesting that this environment promotes the occurrence of horizontal transmission of Bartonella bacteria during the bite/feeding of vectors on rats. However, in order to prove infection route, transmission studies from an infected vector to naïve hosts/rats should be done in a controlled laboratory environment. Seasonal dynamics of Bartonella infection in natural populations of rats can also obscure a correlation between prevalence in rats and ectoparasites.
Bandicota and Rattus rats are the most common reservoir hosts for Bartonella infection in Southeast Asia which raises a question about the potentially important role of these rodentborne agents as sources of febrile illness in human populations in Thailand. Detection of Bartonella DNA similar to B. tamiae, isolated from three febrile patients [23], and from mites collected from rodents in Thailand [15] imply a potential connection role for transmission of the disease to humans. Data from Kosoy et al. 2010, support this potential transmission route came from isolation of Bartonella species, previously identified from rodent hosts, from Thai patients' blood and supported by analysis of history of the patients having an exposure to rats during the 2 weeks before the illness [35]. Bartonella species identified in these human cases included B. elizabethae, B. rattimassiliensis, and B. tribocorum, and the last two of these species were also detected in rodents, mites, fleas, and lice reported in this study.

Disclaimer
Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals, NRC Publication, 2011 edition. The opinions or assertions contained herein are the private views of the author, and are not to be construed as official, or as reflecting true views of the Department of the Army or the Department of Defense.
Supporting Information S1 Fig. Phylogenetic relationships between Bartonella species isolated from rodent blood and some reference species according to the ssrA and gltA phylogenies. The GenBank accession numbers are shown for each reference sequences. Trees constructed from ssrA and gltA genes were able to discriminate all samples into two major clusters (C1 and C2), although some different branching patterns of sequences in C1 group (R1012, R1033, R1135, and R1188) were noticed. (TIF) S1 File. Additional data of ectoparasites collected from rodents and Bartonella DNA prevalence in the developmental stages of tick and flea. Detail information of ectoparasites collected from rodents in this study (Table A).Bartonella DNA detection in flea pools classified by gender (Table B) and in tick pools classified by stage and gender (Table C) (DOCX)