The diverse genetic genotypes of Bartonella species circulating in rodents from Inner Mongolia, Northern China

Bartonella are generally recognized as zoonotic pathogens of mammals, including many rodent species. However, data on the genetic diversity of Bartonella in some regions are still absent in China. In this study, we collected rodent samples (Meriones unguiculatus, Spermophilus dauricus, Eolagurus luteus, and Cricetulus barabensis) from Inner Mongolia located in Northern China. The Bartonella were detected and identified by sequencing the gltA, ftsZ, ITS, and groEL genes in them. An overall 47.27% (52/110) positive rate was observed. This may be the first report that M. unguiculatus and E. luteus harbor Bartonella. Phylogenetic and genetic analysis on gltA, ftsZ, ITS, and groEL genes indicated that the strains were divided into seven distinct clades, suggesting the diverse genetic genotypes of Bartonella species in this area. Of those, Clade 5 meets the criteria for identification as a novel species based on gene sequence dissimilarity to known Bartonella species and herein we name it “Candidatus Bartonella mongolica”.


Introduction
The genus Bartonella belonging to family Bartonellaceae, order Rhizobiales (Hyphomicrobiales) represents a group of Gram-negative, facultative intracellular bacteria with worldwide distribution. To date, this genus is composed of at least 53 species, 7 subspecies, and some unnamed Bartonella spp. according to the National Center for Biotechnology Information database [1]. With the expanded sampling and improved detecting methods, novel species from all over the world are still being discovered. Many Bartonella species have been recognized as etiologic agents of emerging infectious diseases for both humans and domestic animals. Until 2021, at least 20 Bartonella species have been reported to infect humans [1,2], with the syndromes varying from mild and unspecific clinical signs to life-threatening symptoms [3]. First identified in 1905 by Alberto Leonardo Barton Thompson, B. bacilliformis transmitted by sandflies is proved to be the agent of Carrion's disease with high lethality in the absence of adequate treatment [4]. As the most common Bartonella species infecting humans, B. quintana is the agent of trench fever mainly representing endocarditis and chronic bacteremia, while B. henselae causes cat scratch disease representing endocarditis, bacillary angiomatosis, and peliosis hepatis [5,6]. In the past decades, more Bartonella species have been described to infect humans, including B. clarridgeiae, B. elizabethae, B. grahamii, B. koehlerae, B. kosoyi, B. melophagi, B. rochalimae, B. tamiae, and B. washoensis, mainly manifesting with symptoms of fever, fatigue, and endocarditis [7][8][9][10][11][12][13][14][15]. Of those, B. elizabethae, B. grahamii, B. kosoyi, and B. washoensis are all hosted by rodents.
Epidemiological studies have demonstrated that Bartonella are mainly transmitted by hematophagous arthropods such as fleas and lice by their bites [16]. Bartonella species have been detected in multiple mammalian hosts including carnivores, rodents, bats, deer, and even marine animals [17][18][19]. In the past decades, a remarkable diversity of Bartonella has been reported in rodents, which were considered one of the major reservoir hosts. However, increasing research efforts in bats and other taxa might reveal similarly high diversity [18][19][20][21]. Bartonella infects the erythrocytes and endothelial cells of these animals, and then establishes persistent infections producing bacteremia that may last for months [22]. In some rodents, vertical transmission of Bartonella was also observed [22]. These characteristics strongly suggest that rodents play an important role in preserving and transmitting Bartonella in nature.
In recent years, lots of studies have been performed on molecular investigations of Bartonella in rodents from China. Up to 2021, multiple Bartonella species have been detected in various provinces including Heilongjiang, Shaanxi, Fujian, Xinjiang, Henan, Zhejiang, Yunnan, Guangdong, Hainan, and Qinghai Provinces, China [23][24][25][26][27][28]. However, some of the studies are preliminary and only one single short gene fragment was analyzed. Furthermore, in the Inner Mongolia Autonomous Region located in Northern China, a region of plague epidemic with an area of 1183, 000 km 2 , almost no studies on Bartonella have been performed except for one study reporting Meriones unguiculatus rats harboring Bartonella tribocorum and B. grahamii in this area [29]. To improve our knowledge on the diversity and epidemiology of Bartonella in China, we collected rodent samples in this region and studied the Bartonella bacteria in them.

Ethics declaration
This study protocol was approved by the Ethics Committee of the National Institute for Communicable Disease Control and Prevention, the Chinese CDC (Approval No. 2021-011), and the General Center for Disease Control and Prevention of the Inner Mongolia Autonomous Region. All the rodents were treated in a humane manner in accordance with the "Rules for Implementation of Laboratory Animal Medicine" from the National Health Commission, China. The studies in this manuscript adhere to the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines for the reporting of animal experiments.

Sample collection and DNA extraction
In an ongoing plague surveillance project in 2021, rodents were live-trapped using fried food as baits in the desert steppe in three locations ( Traps were checked for rodents each day and cages containing rodents were transported to the lab. After morphological identification [30], the rodents were anesthetized using pentobarbital sodium (40 mg/kg) to minimize their suffering and then sacrificed by cervical dislocation. The leftover liver tissues were then collected and stored at -20˚C until DNA extraction. After being washed three times with Phosphate Buffered Saline (PBS) and ground, liver samples (25 mg) were then subjected to DNA extraction. The DNeasy Blood &Tissue kit (QIAGEN, Germany) was used to extract the DNA according to the manufacturer´s instructions. The eluted DNA samples were stored in a -80˚C refrigerator until PCR detection of Bartonella.

Molecular detection of Bartonella bacteria by amplifying gltA gene
As previously indicated, the primers forward BhCS871.p and reverse BhCS1137.n [31] were used for the detection of Bartonella amplifying a conserved region of the citrate synthase gene (gltA). PCR amplification was performed using a conventional Sensoquest PCR System LabCycler (Germany). Positive and negative controls were set in each PCR run. The PCR parameters are as follows: denaturing at 95˚C for three minutes, 40 cycles of 95˚C for one minute, 56˚C for one minute, and 72˚C for one minute. PCR amplification was completed after a final incubation at 72˚C for 10 minutes. The PCR products were analyzed by electrophoresis on agarose gel and then observed under UV light. All the PCR products of the expected size (approximately 380 bp) were sent for DNA sequencing. The obtained gltA sequences, which were commonly used for the taxonomic identification of Bartonella species [32], were then analyzed by BLASTn (https://blast.ncbi.nlm.nih.gov/Blast.cgi? PROGRAM = blastn&PAGE_TYPE = BlastSearch&LINK_LOC = blasthome) to preliminarily determine their genetic similarity to known species. The DNA of Bartonella sp. we previously detected in Guizhou Province was used as the positive control [33] and ddH 2 O was used as the negative control.

PCR amplification and genetic analysis of ftsZ, groEL, and ITS genes
Based on genetic analysis of the obtained gltA sequences, 23 representative Bartonella strains were selected for further characterization. The cell division protein (ftsZ) gene was amplified using the BaftsZF and BaftsZR primers as shown [34]. The primers used for 16S-23S intergenic spacer region (ITS) gene amplification were as shown [35]. The PCR conditions were as described previously. For amplification of the 60 kDa chaperonin protein (groEL) gene, heminested primers were designed in this study (as shown in S1 Table) according to the conserved regions of all available Bartonella sequences downloaded from the GenBank Database, amplifying approximately 415 bp region. All the PCR products of the expected size were subjected to DNA sequencing and then analyzed by BLASTn algorithm.
All the sequences recovered in this study have been submitted to the GenBank Database and the Accession Numbers have been assigned (Shown in S2 Table).

Phylogenetic analysis
The gltA, ftsZ, groEL, and ITS gene sequences, as well as those from known Bartonella strains deposited in the GenBank database, were manually aligned using ClustalW method within MEGA7.0 and trimmed using SeqMan software (DNASTAR Inc., Madison, USA) to remove poor quality sequences [36]. A phylogeny was constructed based on the nucleotide sequences using the maximum-likelihood (ML) method implemented in PhyML software [37]. The bestfit phylogenetic model was determined by the substitution model test. Bootstrap analysis was carried out with 100 re-samplings. All the trees were mid-point rooted. Accession numbers and rodent hosts of the sequences in this study were shown in the trees.  (Table 1). Of those, 23 representative strains whose gltA sequences represent genetically distinct genotypes were selected based on their nucleotide sequences for further investigations.

PLOS NEGLECTED TROPICAL DISEASES
with the tree based on the gltA gene (Fig 2). Namely, all the strains belonging to certain clades in the gltA tree are still clustered together in these phylogenetic trees. To be noticed, clades 5, 6, and 7 are all located in a major clade in phylogenetic trees based on gltA, ftsZ, and ITS genes.
For the ftsZ gene, Clades 2, 3, and 4 showed the highest nucleotide homologies with B. washoensis (97.20%), B. grahamii (96.13%), and B. krasnovii (95.74%), respectively. Clades 5, 6, and 7 are all mostly related to B. taylorii, with the highest nucleotide similarities of 87.09%, 86.71%, and 87.55-87.56%, respectively (   Unexpectedly, the phylogenetic tree based on groEL sequences showed that the positions of Bartonella strains are obviously inconsistent with gltA, ftsZ, and ITS trees (Fig 3). Clade 1 (strain 18) and clade 2 (strains 16, 23, and 25) clustered together. Both of them are closely related to B. washoensis with the identities of 98.55% and 98.55-98.79%. Remarkably, strains from clade 6 (strains 42, 46, and 51) and clade 7 (strains 8, 28, 34, 41, 48, and 53) are dispersed into different lineages. For example, strains 42 and 46 belonging to Clade 6 clustered with strain 48 belonging to clade 7, while strains 28, 34, and 41 (Clade 7) are in the same lineage together with strain 51 belonging to Clade 6. This result may result from several reasons: 1. It is possible that co-infection of different Bartonella strains may present in the rodent blood and the groEL primers may have been more sensitive to a minor variant present; 2. Lack of some  (Table 2). According to the criteria for Bartonella species definition reported in the previous study [32], we suppose that clade 5 undoubtedly represents a novel species. Herein we name it "Candidatus Bartonella mongolica" based on the area where it was first detected.

Discussion
Bartonella are considered emerging zoonotic causative agents distributed worldwide. Small mammals such as rodents and bats are the primary hosts of Bartonella, and approximately two-thirds of currently reported Bartonella species are harbored by rodents. Over the past decades, an increasing number of Bartonella species has been reported worldwide. However, despite the species abundance of rodents in China (more than 200 species), the geographical distribution and genetic diversity of rodent-borne Bartonella in many areas are still pending.
In this study, a high prevalence of Bartonella spp. was observed in rodents from Inner Mongolia, with an overall positive rate of 47.27% (52/110). This rate is similar to those previously reported in rodents from Yunnan Province (43.5%) [38] and Shanxi Province (49.52%) [39]. Three of the four rodent species tested positive for Bartonella, with the positive rate varying from 35.82%-85.71%. As is currently known, this may be the first report that M. unguiculatus and E. luteus harboring Bartonella, which may expand our knowledge on the host range of Bartonella. Furthermore, the host specificity of Bartonella was also observed in this study. For example, all strains belonging to Clade 3 were detected in E. luteus, which only account for 19.09% (21/110) of the rodent samples. This is well consistent with previous reports that each Bartonella specie only infects one or a few rodent species [40]. Notably, Siziwang Banner is an epidemic area of plague. In this area, M. unguiculatus and S. dauricus are the main hosts of Yersinia pestis, the etiologic agent of plague [41]. As these rodents live in proximity to humans (mostly herdsmen) in this area, human-rodent interactions are supposed to be frequent. The high infection rate of Bartonella in these rodents makes it possible that Bartonella spp. may be transmitted to human beings in a similar pathway as plague. Bartonellosis such as cat scratch disease caused by B. henselae has been widely reported in multiple provinces of China, such as Hebei, Fujian, Anhui, Zhejiang, Guangdong, Hubei, Guizhou, and Beijing [42]. Meanwhile, human cases infected by other Bartonella species were also reported including Bartonella vinsonii subsp. berkhoffii, B. quintana, and B. coopersplainsensis [42][43][44]. Nonetheless, clinical cases of bartonellosis in Inner Mongolia have never been reported, probably due to minor symptoms (fever, malaise, etc.) and overlap with other infectious diseases. It is plausible to suppose that the risk of human infection by Bartonella is largely underestimated or ignored.
In this study, remarkable genetic diversity of Bartonella was observed, with seven distinct phylogenetic clades identified. Of those, the gltA, ftsZ, ITS, and groEL genes of Clade 5 showed the highest nucleotide identities of 92.94%, 87.09%, 88.89-89.35%, and 94.39-94.64% to reported species. According to the criteria for the definition of Bartonella species [31], these remarkable genetic divergences suggest that they represent a novel species. Herein we name it "Candidatus Bartonella mongolica". Genetic and phylogenetic analysis indicated that all the key genes of clade 3 strains are closely related to B. grahamii. Namely, this clade should represent a variant of B. grahamii. As a validated human pathogen, B. grahamii has been reported to cause cat scratch disease, with the symptoms of neuroretinitis and enlarged lymph nodes [11]. The high positive rate of B. grahamii in E. luteus and its potential human pathogenicity may suggest the risk of human infection.

Ethics approval and consent to participate
This study protocol was approved by the Ethics Committee of the National Institute for Communicable Disease Control and Prevention, the Chinese CDC (Approval No. 2021-011), and the General Center for Disease Control and Prevention of the Inner Mongolia Autonomous Region. All the rodents were treated in a humane manner in accordance with the "Rules for Implementation of Laboratory Animal Medicine" from the National Health Commission, China.

Availability of data and materials
All sequence files are available from the GenBank database (Accession Numbers are shown in S2 Table).
Supporting information S1