Moleculare detection of Bartonella in ixodid ticks and plateau pika (Ochotona curzoniae) in Shiqu county, China

Background Bartonella bacteria have been associated with an increasingly wide range of human and animal diseases and were also recognized to be globally dispersed as emergent pathogens. Ticks and small rodents are known vectors of human and animal bartonellosis and play important roles in maintenance and circulation of bartonellae in nature. In China, Shiqu county is located on the eastern Qinghai-Tibetan plateau and about 26 thousands square kilometers with an average altitude of above 4,200 meters and vast area of pastureland. In present study, the occurrence of Bartonella spp. in ticks and plateau pika was �rstly investigated in Shiqu county. Results A total of 818 ticks ( Dermacentor everestianus , 79.0%, 646/818; Haemphysalis qinghaiensis , 21.0%, 172/818), were collected in 4 villages in Shiqu county. Only Bartonella melophagi was detected in tick samples with a total prevalence of 30.1% (246/818). Signi�cant difference was observed (P<0.05) between D.everestianus (17.0%) and H.qinghaiensis (79.1%).The infection rates of Bartonella spp. in ticks from Arizha, Maga, Derongma and Changxgma villages were 4.8%, 76.8%, 12.5% and 18.0%, respectively. Compared with other villages, the infection rate of Bartonella spp. in Maga was higher (P<0.01). As for plateau pika, total infection rate of Bartonella spp was 24.1%, with 20.8% (15/72), 30.9% (25/81), 13.8% (9/65) and 29.4% (20/68) in Arizha, Maga, Derongma and Changxgma, respectively. Totally, B. queenslandensis , B.grahamii and two unvalidated Bartonella species were detected. No signi�cant difference in infection rates was observed (P>0.05) between theses study sites. Conclusion


Background
The Bartonella genus, currently includes 36 named and 17 Candidatus species [1], which can be found in a wide range of mammalian and arthropod hosts and some of them are zoonoses, including: B.alsatica, B.bacilliformis, B.elizabethae, B.henselae, B.koehlerae, B.melophagi, B.quintana, B.rochalimae, B.tamiae, B.vinsonii subsp.berkhoi, B.vinsonii subsp.arupensisand B.washoensis [2][3][4][5][6][7][8].Ticks and small rodents are known vectors of human bartonellosis and play important roles in maintenance and circulation of bartonellae in nature within arthropod-mammal systems.Shiqu county is about 26 thousands square kilometers with an average altitude of above 4,200 meters and vast area of pastureland on the eastern Qinghai-Tibetan plateau, where lives a population of estimated 97 thousands with low education level and poor public health.Yak is the largest population of local livestock (about 600 thousand) and severe tick infestation is often observed.Except yak, plateau pika (Ochotona curzoniae) is the largest population of local small rodents with close interaction with local people and livestock.The signi cance of ticks (Acari: Ixodida) has long been recognized due to their ability to feed on a large range of host species and to transmit Bartonella pathogens that can infect a variety of vertebrate hosts, including humans.However, little information exists on bartonellae and their vectors in Shiqu county.The objective of this study was to provide evidence of the presence of Bartonella spp. in plateau pika and ticks and preliminary results to establish prevention and control measures for this tick-borne disease.
According to criteria (Bartonella spp.species thresholds: gltA≥96.0%and rpoB≥95.4%)proposed by La Scola, et al [13], for tick samples, only B.melophagi was detected (Table 2); For plateau pika, as shown in Table3, B.grahamii was predominantly identi ed species found in four villages with B. queenslandensis detected only in Maga and 2 unvalidated Bartonella species (Bartonella.sp*and Bartonella.sp**)found in Ariza and Changxgma, respectively.Furthermore, gltA and rpoB based phylogenetic analysis supported the classi cation of Bartonella spp.detected in the present work (Fig. 2 and Fig. 3).

Discussion
In this study, two tick species were found: H.qinghaiensis (only in Maga) and D.everestianus(in all of four sites).D.everestianus was only reported in Northwestern China and Nepal [15] with an altitude of 2600-4700m [16].Larvae and nymphs of this tick species often infest lagomorphs and rodents, while adult ticks usually utilize medium-large sized, modest and wild mammals as hosts, including hares, sheep, yaks, and horses [15,16].However, H.qinghaiensis a typical three-host tick, is only recorded in China [17][18][19][20][21], particularly prevalent in the western plateau, including the provinces of Qinghai, Gansu, Sichuan and Tibet [21].Its natural hosts include sheep, goat, yak, cattle and hare (Lepus oiostolus).It is known that all stages of the tick could develop in sheep, goat, yak and cattle [21][22][23][24][25][26][27].Comparing to ticks at high altitude, the activity of H.qinghaiensis is more frequently at low altitude.In this study, Arizha, Changxgma and Derongma villages belong to sub-frigid zone, with an altitude of 4300-4600m Maga village is located in the cold temperate zone, with an altitude of 3799m.There was a signi cant difference in altitude between Maga and the other three villages, which was probably why H.qinghaiensis was only found in Maga village.
All types of ticks were found to contain Bartonella DNA, although in varying percentages and locations.A survey of ticks from 16 states in the U.S revealed that the overall prevalence of B.henselae in Ixodes ticks was 2.5% [28].In Austria, Bartonella spp.(B.henselae, B.doshiae, and B.grahamii) were detected in 2.1% of I.ricinus with the highest rate in ticks derived from Vienna (with a 7.5% infection rate), and that adult ticks had a higher prevalence than other stages [29].Furthermore, a recent One Health perspective review on Bartonella indicated that the overall presence of Bartonella in ticks (combining evidence from multiple surveillance studies) was about 15% [30].In our results, a total prevalence of 30.1% in ticks (especially in Maga, 76.8%) indicated the severity in Shiqu county.
B. melophagi, a human bacterial pathogen, was rstly isolated from sheep blood in 2007 [31] and the same bacteria were then isolated from the blood of two female patients with pericarditis and skin lesions in the USA [32].In this study, this is the rst report of DNA of B.melophagi detected in D.everestianus and H.qinghaiensis and it was the rst molecular evidence of B.melophagi found in Shiqu county.However, currently there is no evidence supporting the ability of these ticks to transmit B. melophagi to livestock or human.To address this issue, experiments need to be performed to assess vector competency of D.everestianus and H.qinghaiensis to transmit B.melophagi in the future.
Bartonella infection has been mostly reported in Rodentia [33][34][35][36][37][38][39][40][41][42], with few cases reported in Lagomorpha.Until now, there has only one report of Bartonella infection in plateau pika with a positive rate of 18.99% [43].Totally, 15 Bartonella strains were obtained and most of them were closely related to B.taylorii and B. grahamii [43].In our results, B.grahamii, a pathogenic strain in humans, was detected in all of four villages while B.queenslandensis was detected only in Maga.Nevertheless, as B.coopersplainsensis, the zoonotic potential of B. queenslandensis has not been reported.Additionally, for two unvalidated Bartonella species (Bartonella.sp*and Bartonella.sp**)found in Ariza and Changxgma, respectively, sequences analysis showed : 1) based on gltA gene, they were clustered with B.rochalimae and B. queenslandensis, respectively; 2) based on rpoB, however, they were clustered with B. melophagi.The causes of this con icting result can be classi ed as fowllows: 1) potential presence of multiple Bartonella species in the sample although it is not common based on culturing; 2) different primer sets may have ampli cation bias towards particular species based on the annealing a nity.These complications may cause the observed Bartonella diversity to differ depending on which marker was used for ampli cation; 3) homologous recombination, a speci c form of LGT (lateral gene transfer) among Bartonella spp.This problem has been documented in several studies of Bartonella strains from cats, rodents and bats based on sequencing multiple protein-coding loci [12,[44][45][46][47][48].However, culturing, sequencing multiple loci (including 16S rRNA, ftsZ, gltA, groEL, ribC and rpoB and ITS), cloning sequences into vectors before sequencing or deep sequencing approaches can be su cient to describe a potentially novel Bartonella specie or subspecies and may differentiate these possible scenarios.
In Shiqu, plateau pika, the largest population of local small rodents, has close contact with local people and livestock and can be infested with eas and ticks, implicating them in transmission cycles of Bartonella spp.In China, Bartonella infections among humans have mainly been reported in the central plain area, such as Jiangsu, Zhejiang, Anhui, and Hubei province.No cases or suspected cases have been reported in the Qinghai-Tibetan plateau.So, the relationship of plateau pika and the transmission of Bartonella should be studied closer and more thoroughly with controlled experiments to determine the exact routes of transmission between plateau pika, the transmission between plateau pika and their vectors, as well as the transmission between plateau pika to humans and livestock.

Conclusion
At present, only D.everestianus and H.qinghaiensis were found in Shiqu county with high infection of Bartonella spp. in theses ticks and plateau pika and further research need to be conducted to determine the risk of Bartonella infections to humans and livestock.

Biological sampling
Ticks were carefully removed from blanket and stored in 70% ethanol at 4℃.The specimens were morphologically identi ed using the guidelines for tick identi cation [9].Then, molecular identi cation of tick species was performed targeting the mitochondrial 16S rRNA gene [10].Plateau pikas were captured using mouse snap traps.After capturing the animals, spleens were collected under sterile conditions, and stored in liquid nitrogen until use.The bodies of pika were deeply buried to avoid being eaten by dogs, cats and other wild carnivores.
DNA extraction, PCR and sequencing Ticks were sectioned longitudinally and one half per each tick was used for DNA extraction; For all of spleen samples, an average of 30 mg of tissue was used.Total DNA of all samples were extracted using the TIANamp Genomic DNA Kit (TIANGEN Biotech Co., Ltd, Beijing, China; Cat No: DP304) for tick molecular identi cation and characterization of Bartonella spp.All samples were submitted to previously described PCR assays targeting gltA (379 bp) [11].All gltA-positive samples were further analyzed with PCR targeting rpoB (379 bp) [12].In this study, all primers were listed in Table 1.PCR ampli cations were conducted in a 25μl reaction mixture consisting of 1μl of genomic DNA (2-3 ng), 1μl of each primer (10μM), 12.5μl of PCR Supermix (Transgen Co., Ltd, Beijing, China; Cat No: AS111-11) and 9.5μl of nuclease-free water.Each PCR reaction included a positive control (DNA of B. henselae, preserved in lab) and a negative control (nuclease-free water).Observed bands were puri ed using the QIAquick Gel Extraction Kit and sent for sequencing (Sangon Biotech (Shanghai) Co., Ltd).Obtained sequences were analyzed by using the Bioedit v.7.0.2 and submitted for nucleotide BLAST search through the NCBI database.Sequences with ≥95% quality cover and identity were considered to be positive for Bartonella spp.and compared with validated Bartonella species in GenBank/EMBL/DDBJ using the Clustal X program (http://www.clustal.org/clustal2/).Clones that share ≥96.0% and ≥95.4% similarity in gltA and rpoB sequences with the validated species, respectively, can be considered as the same species [13].

Phylogenetic analysis and statistics
For phylogenetic analysis, Neighbor-Joining phylogenetic trees were constructed based on Bartonella gltA and rpoB sequences using the Kimura two-parameter model with partial gap deletion and a cutoff of 95% site coverage, respectively.The evolutionary distance was calculated and bootstrap analysis with 1,000 iterations was carried out with the MEGA6 [14].SPSS19.0 (One-way ANOVA) was applied to compare the difference in prevalence of Bartonella spp. between different sampling locations, plateau pika and tick species.A p-value of ≤ 0.05 was considered signi cant.

Figure 1 The
Figure 1