Molecular detection and genetic characterization of small rodents associated Bartonella species in Zhongtiao Mountain, China

The prevalence and molecular characteristics of Bartonella infections in small rodents in the Zhongtiao Mountain, China have been explored. In this study, the liver, spleen and kidney tissues of captured rodents were used for Bartonella spp. detection and identification by combination of real-time PCR of transfer-mRNA (ssrA) gene and traditional PCR and sequencing of citrate synthase (gltA) gene. It was shown that 49.52% of the rodents (52/105) were positive for Bartonella spp.. The infection rate in different gender (χ2 = 0.079, P = 0.778) and tissues (χ2 = 0.233, P = 0.890) of small rodents did not have statistical difference, but that in different small rodents (Fisher’s exact test, P < 0.001) and habitats (χ2 = 5.483, P = 0.019) had statistical difference. And, the sequencing data suggests that Bartonella sequences (n = 31) were identified into three species, including 14 of B. grahamii, 3 of B. queenslandensis and 14 of unknown Bartonella species. Phylogenetic analysis showed that B. grahamii sequences were clustered with the isolates from South Korea and China, and B. queenslandensis sequences were mainly closely related to the isolates from China and Thailand. The genetic diversity analysis showed that B. grahamii and B. queenslandensis sequences exhibited noticeable intraspecies diversity. Taken together our data demonstrates the high prevalence and genetic diversity of Bartonella infections in small rodents in the Zhongtiao Mountain, especially a potential novel Bartonella specie was detected, which could benefit the prevention and control of rodent-Bartonella species in this area.


Introduction
Bartonella species are an emerging family of vector-borne, facultative, gram-negative, haemotrophic bacteria, known to infect mammalian erythrocytes and endothelial cells and might cause human Bartonellosis [1]. There are currently over 40 species of Bartonella identified, and 14 of them are known to be zoonotic [2]. They  cats, dogs, rodents, bats, and so on [3]. Humans can be infected by close contact with rodents, cats and dogs, including typical and common diseases, such as cat scratch disease [4] and trench fever [5], and atypical diseases, such as neuroretinitis, arthritis, endocarditis, myocarditis, osteomyelitis, bacteremia, etc [6][7][8].
The Zhongtiao Mountain, is one of the major mountains, in the southern Shanxi Province of China, with an average elevation of 1200-2300 m, spread across the three cities of Linfen, Yuncheng and Jincheng. Here, we selected Pingquan Village within the Yangcheng County of Jincheng City as the rodents trapping site, as it is located at the east of Zhongtiao Mountain, and 30 km from the Manghe National Nature Reserve. Pingquan Village, as the Red Tourist Spot, with the development of tourism, the direct or indirect contact between humans and rodents increased, and the risk of transmission of rodents-associated Bartonella infection also increased. However, investigations of Bartonella species in small rodents in this area have not been reported. Given this, this study was designed to explore the prevalence and genetic diversity of Bartonella species in small rodents in the parts of Zhongtiao Mountain, in an effort to provide the necessary scientific information to help support the creation of a plan for the control and prevention of Bartonella infection in humans in this area.

Rodents collection
Small rodents were captured using snap traps from Pingquan Village (35.36˚N, 112.32˚E) within Yangcheng County of Shanxi Province during May 2021. The trapped rodents were identified by morphology and DNA barcoding via the Cytochrome C oxidase subunit I (COI) gene. The liver, spleen and kidney tissues were then harvested under sterile conditions from each rodent after euthanasia, and stored at -80˚C until use.

Phylogenetic analysis
The sequences generated in this study were submitted to the GenBank (accession numbers: MZ672181-MZ672211). The nucleotide sequence homology was blasted against reported Bartonella species sequences in the GenBank using the BLAST program available from the National Center for Biotechnology Information website (http://blast.ncbi.nlm.nih.gov/Blast. cgi). Besides, the gltA sequences of B. grahamii and B. queenslandensis in GenBank released before July 2021 were collected for traceability analysis of these two Bartonella species in our study. For the sequences isolated of the same strain from the same host in the same laboratory at the same time, we randomly selected a sequence as the reference. Phylogenetic tree was created using the maximum-likelihood method with MEGA version 7.0, and bootstrap values were calculated with 1000 replicates [34,35]. Brucella abortus was used as the outgroup.

Genetic diversity analysis
The polymorphism of nucleotide sequences, including the number of polymorphic sites (S), the number of haplotypes (H), the nucleotide diversity (π), the average number of nucleotide differences (k) and the haplotype diversity (Hd), were analyzed using DNASP 5.10 software. We used a sliding window interval of 25 bp to determine which segment of the target gene sequence had the highest nucleotide diversity (π) by analyzing 100 bp at a time across the length of the gene.

Statistical analysis
The positive rates of Bartonella spp. in different genders, tissues, and habitats of small rodents were analyzed using the chi-square test. The positive rates of Bartonella in different rodents were analyzed using Fisher's exact probability method. All data were analyzed using SPSS 22.0 (SPSS, Inc., Chicago, IL, USA). P < 0.05 was considered statistically significant.

Bartonella tracing
We then went on to explore the likely origin of the Bartonella species in this area, using the gltA sequences of B. grahamii and B. queenslandensis placed in GenBank before July, 2021 as reference in the traceability analysis. It was shown that B. grahamii was mainly clustered into two clusters, indicating that these sequences are likely to have the different origins, with 11 of the B. grahamii sequences, including nine from A. agrarius, one from A. draco and one from A. peninsulae, shown to be most closely related to B. grahamii from A. agrarius collected in South Korea, and three B. grahamii sequences obtained from N. confucianus were most closely related to B. grahamii isolated from T. triton in China. The two B. queenslandensis sequences (NC37SXYC and NC104SXYC) obtained from N. confucianus were most closely related to B. queenslandensis from N. confucianus in China and N. fulvescens in Thailand, and one B.

PLOS ONE
queenslandensis sequence (NC59SXYC) creating a unique cluster that did not correspond with any of the existing B. queenslandensis reference strains (Fig 3). This indicated that Bartonella infection demonstrates some specificity for specific rodent species, and that there is a high degree of the genetic diversity in the Bartonella species prevalent in this area.  Table 4 and Fig 4).

Discussion
Bartonella species can cause a latent infection in their reservoirs, and this may lead to a high prevalence of Bartonella in the wild rodents [36]. Previous studies have shown that Bartonella species are widely distributed at a high frequency in wild rodents around the world. For example, the positive rate of Bartonella was 90.4% in O. torridus in the United States, 69% in Apodemus in Japan, 17% in M. minutus in Russia, and 78% in the rodents in Thailand, and so on [37,38]. This indicates that the prevalence of Bartonella varies greatly in different rodents from different geographical regions.
In the present study, the prevalence and molecular characteristics of Bartonella species in small rodents in the Zhongtiao Mountain were observed. To our knowledge, this is the first report of the investigation in this area. Our results indicted that the infection rate of Bartonella species was 49.52% in small rodents, was higher than that in some areas of China, such as Zhejiang (31.4%) [28], Fujian (16.19%) [39], Yunnan (39.2%) [40], and Qinghai (18.99%, 38.61%) [41,42], and was lower than that in Heilongjiang (57.7%) [43]. Currently, the most efficient and convenient way to detect Bartonella infection is by PCR. In this study, the real-time PCR targeting ssrA gene was used for Bartonella detection. The ssrA gene has the same species identification function as the traditional gltA gene, indicating that the amplification of ssrA fragment can be used as a fast and accurate method for the detection and classification of Bartonella [32]. Here we used a combination of liver, spleen and kidney tissues for Bartonella detection, with multi-tissue detection likely to capture a more accurate positive rate and produce as many sequences as possible for evaluation. We detected Bartonella species in six of the eight small rodent species identified in this study, and determined that their infection rates were significantly different. In addition, the infection rate was significantly different between different habitats, but not between genders, which was similar to the previous study [42].
DNA sequence homology and phylogenetic analyses of gltA gene identified three Bartonella species in this cohort of animals, including B. grahamii, B. queenslandensis and unknown Bartonella species. B. grahamii was detected in A. agrarius, N. confucianus, A. draco and A. peninsulae, which infections are primarily associated with neuroretinitis and cat scratch disease (CSD) [18,19], suggesting B. grahamii can infect a variety of rodents, and has the ability to cause human disease in this area. In addition, B. queenslandensis was only detected in N. confucianus, and unknown Bartonella was only detected in A. agrarius. Unexpectedly, no Bartonella sequences was obtained from E. inez in this study, might be associated with the low quantity of  bacteria in E. inez in this area, which still needs further investigation. These results also indicated that there is some host specificity for Bartonella in rodents, and that A. agrarius may be more susceptible to multiple species of Bartonella, which is similar to the conclusions of several other studies [43][44][45][46]. In addition, multiple Bartonella species can be detected in the same rodent species (e.g., A. agrarius), and one Bartonella specie (e.g., B. grahamii) can infect a variety of rodents, indicating that this pathogen has a strong adaptability in rodents. Traceability analysis showed that B. grahamii was mainly clustered into two clusters, one showing close association with the isolates from South Korean A. agrarius and the other with strains from Chinese T. triton samples. Our B. queenslandensis strains were most closely related to the isolates from N. confucianus in China and N. fulvescens in Thailand. In addition, we had one B. queenslandensis sequence that did not cluster with the reported B. queenslandensis reference strains, indicating that it might be a novel genotype. In addition, genetic diversity analysis demonstrated that B. grahamii and B. queenslandensis exhibited noticeable intraspecies genetic diversity, which was similar to the results of a previous study [47].
In conclusion, we identified three species of Bartonella, B. grahamii, B. queenslandensis and unknown Bartonella in four species of rodents, A. agrarius, N. confucianus, A. draco and A. peninsulae from the Zhongtiao Mountains in China. These results also showed that B. grahamii and one potential novel Bartonella species were dominant in this region and that B. grahamii and B. queenslandensis had the high genetic diversity in this area. And the biological characteristics of the potential novel Bartonella species need to be further investigated. Our study provided a better understanding of the prevalence and molecular characteristics of Bartonella species in small rodents in the Zhongtiao Mountain, which could benefit prevention and control of rodent-Bartonella species in this area.