Study area, rodent capture, and blood collection

During 2019 and 2021, rodents were captured in Baja California (32°7′13.59″N, 115°15′4.61″W) and Chihuahua (30°39′8.50″N, 108°31′23.76″W). In both states, agricultural activities are practiced, such as intensive and extensive cattle raising and agriculture, which includes different crop types such as onions, wheat, cotton, and chilli.

Rodent trapping was conducted over two seasons by placing grids of 7×7 Sherman traps (7.6 cm × 8.9cm × 22.9 cm; H.B. Sherman Traps, Inc., Tallahassee, FL). Each grid was spaced at least ≥500m apart. Traps were baited with a mixture of oats and vanilla extract and opened for three consecutive nights. In total, 12 and 18 grids were placed in Baja California and Chihuahua, respectively. The quadrats were distributed in surrounding areas and far from human settlements; most of them were in a mosaic landscape dominated by mesquite shrublands, grassland, and croplands vegetation; and others were placed in oak forest areas.

Once the rodents were captured, they were weighed and sexed, and morphometric measurements were taken for identification using a mammal identification guide [Reference Reid23]. Subsequently, blood samples were obtained from the retro-orbital plexus, placed in cryovials with EDTA, and stored at −70°C. After handling, each rodent was released at the capture site.

Molecular analysis

DNA extraction from rodent blood samples was performed using the DNeasy Blood and Tissue kit (Qiagen®), following the supplier’s recommendations. Once the extraction was performed, polymerase chain reaction (PCR) was continued by amplifying 767 bp of the gltA gene using the following primers: CS443f: 5′ GCTATGTCTGCATTCTCTCTCTCTCTCTATCA 71 3′ and CS1210r: 5′ GATCYTCAATCATTTCTTTCCA 3′ [Reference Billeter24]. Amplification consisted of a 25 μl final volume mix containing 12.5 μl Top Taq® Master Mix, 5 μl nuclease-free water, 2.5 μl CoralLoad buffer, 2.5 μl DNA, and 1.25 μl (10 μM) of each primer using the following parameters: initial denaturation (2 min 94°C), followed by 45 cycles at 94°C for 30 s, 48°C for 1 min, and 72°C for 1 min, and a final extension of 72°C for 7 min. The PCR products were visualized for amplicons of the expected size by electrophoresis in a 1% agarose gel with ethidium bromide staining. Some amplicons were purified from an agarose gel (1.5%) with the EZ-10 Spin Column DNA Gel Extraction kit (Bio Basic). Subsequently, purification of the remaining PCR products, and sequencing which was performed in both directions, was carried out in Korea by Macrogen. The phylogenetic relationship of our sequences and the reference sequences obtained from GenBank were aligned using MUSCLE in the MEGA 11 program [Reference Tamura, Stecher and Kumar25]. Using the same program, the phylogenetic tree was reconstructed by maximum likelihood by Tamura 3-parameter method and a bootstrap calculation with 1000 replicates. The visualization and modification of the phylogenetic tree style were carried out at Fig Tree version 1.4.4. The novel sequences were submitted to GenBank (PQ655038-PQ655044).

Bartonella prevalence

A total of 586 rodents belonging to four families (Cricetidae, Heteromyidae, Muridae, and Sciuridae), 12 genera, and 28 species were captured (Table 1). PCR was performed on 408 rodents (Chihuahua=285 and Baja California= 123) obtaining an overall prevalence of Bartonella spp. of 39.71%. The highest prevalence was found in the state of Chihuahua with 42.80%, followed by Baja California with a prevalence of 32.52% (Table 1). Previously, a study conducted in 2014 in the same area in Chihuahua reported a higher prevalence in rodents (50.01%) [Reference Rubio12], and later in 2016, a prevalence of 40.4% was found in wild rodent fleas [Reference Fernández-González9]. To our knowledge, our study is the first to report Bartonella DNA in wild rodents in Baja California; this provides new information in this state, and together with previous studies conducted in Chihuahua and Sonora with rodents and/or their fleas, we confirm the circulation of this bacterium in the three north-western states of Mexico (Baja California, Sonora, and Chihuahua) that border the United States [Reference Fernández-González9, Reference Rubio12, Reference Zapata-Valdés22].

Table 1. Prevalence of Bartonella spp. in wild rodents from Chihuahua and Baja California

Note: Capital letters represent total rodents captured (C) and abundance of rodents considered for PCR (A)

We found Bartonella spp. in several rodent species (Table 1). In Chihuahua, we found 18 rodent species infected with Bartonella of which seven are new records (Chaetodipus intermedius, Mus musculus, Neotoma mexicana, Perognathus flavus, Peromyscus boylii, Peromyscus eremicus, and Reithrodontomys fulvescens). The species with the highest prevalence in Chihuahua were Neotoma albigula (72.22%), N. mexicana (100%), and P. boylii (100%) (Table 1). The high prevalence found in N. albigula may be common, as previous studies in the same state and in New Mexico, United States and adjacent to Chihuahua, recorded a high prevalence in this rodent species (75%–100%) [Reference Rubio12, Reference Morway19]. The prevalence we found in N. mexicana and P. boylii is higher than that reported for these species in the United States, 38.7% and 33.3%, respectively [Reference Bai26, Reference Ziedins27]. In Baja California, Bartonella DNA was found in nine rodent species, with the highest prevalence being Chaetodipus baileyi (100%), followed by D. merriami and N. albigula (each 50%). The prevalence found in D. merriami and N. albigula species has been previously reported as moderate to high [Reference Rubio12, Reference Morway19, Reference Goodrich, McKee and Kosoy28], but caution should be taken with the high prevalence we found in C. baileyi, since the only captured individual was positive. Therefore, further studies to confirm this observation is necessary.

Phylogenetic analysis

A total of 26 positive PCR products were sequenced, with both forward and reverse sequencing performed on each. Of these, only 18 consensus sequences were successfully recovered. Basic Local Assignment Search Tool (BLAST) analysis yielded similarity and query cover values for eight sequences. These sequences were subsequently deposited in GenBank. Notably, some sequences were identical and thus were deposited with the same accession number (PQ655043, PQ655044), as indicated in the supplementary material table. Seven genetic variants were determined and grouped into three clades (Figure 1). In the first clade (I), there were five variants belonging to the rodents P. boylii (3), P. leucopus (1), D. merriami (1), Peromyscus fraterculus (1), and O. leucogaster (1) (accession numbers: PQ655039-PQ655041, PQ655043, and PQ655044). These variants were associated with the B. vinsonii group that include subspecies that have been categorized as pathogenic (B. vinsonii arupensis and B. vinsonii berkhoffii) [Reference Buffet, Kosoy and Vayssier-Taussat7, Reference Kosoy and Goodrich29]. Particularly, the variants belonging to a D. merriami (PQ655041, 595 bp) captured in Chihuahua and a P. fraterculus (PQ655040, 463 bp) from Baja California were grouped with B. vinsonii arupensis (AF214557) with 98.90% and 99.14% similarity.

Figure 1. Phylogenetic relationship of Bartonella genotypes based on partial sequences of gltA gene detected in rodents captured in Baja California (BC) and Chihuahua (CH), Mexico. Each genetic variant is indicated in boldface with its accession number, capital letters show the state where the rodents were captured (CH and BC), and numbers in parentheses are the number of sequences obtained from blood. The clades of interest are represented by a rectangle of different colour and by roman numerals (I–III). The phylogenetic tree was constructed by the maximum likelihood method by Tamura 3-parameter and a bootstrap calculation with 1000 replicates.

Previously, in the state of Chihuahua, the presence of B. vinsonii subsp. arupensis has been reported in rodents such as N. albigula, P. maniculatus, and P. leucopus; however, there was no record of this bacterium in D. merriami [Reference Rubio12]. In addition, our study adds another rodent species in Chihuahua as carrier of Bartonella (P. boylii). On the other hand, this is the first time that B. vinsonii subsp. arupensis is reported in a P. fraterculus rodent from Baja California, Mexico.

The second clade (II) consisted of a variant obtained from a D. merriami captured in Chihuahua (PQ655038, 737 bp) that had a 98.78% similarity to B. grahamii (CP001562), and other species that also parasitize rodents (B. elizabethae and B. tribocorum). A study conducted in the same state with wild rodents did not report the presence of B. grahamii [Reference Rubio12]; however, an investigation conducted on rodent fleas reported the presence of this bacterium in Meringis altipecten, Meringis arachis, and Meringis parkeri fleas collected on D. merriami [Reference Fernández-González9]. This suggests that B. grahamii was already circulating among rodents in the region, despite not having been detected in them until now. B. grahamii was recognized of medical importance from its isolation in ocular fluids of a patient, and this Bartonella species has been associated with several genera of the order Rodentia (Apodemus, Dryomys, Microtus, Mus, and Myodes) [Reference Kerkhoff30, Reference Inoue31].

The third clade (III) consisted of zoonotic Bartonella species mainly associated with carnivores (B. rochalimae and B. clarridgeiae); however, one of our variants obtained from a P. maniculatus collected in Chihuahua (PQ655042, 748 bp) had a similarity of 98.26% with Bartonella sp. (FN645480) and 97.91% with Candidatus Bartonella rudakovii (EF682090). The later was originally detected in voles from Siberia, and recently using genes such as gltA and rpoB, it has been recorded in rodents Myodes glareolus, Microtus oeconomus, and Sciurus vulgaris rodents in Switzerland, Lithuania, and Czech Republic [Reference Mardosaitė-Busaitienė32–Reference Majerová34]. In general, information on Candidatus B. rudakovii worldwide is scarce, and so far, it has not been recognized as zoonotic [Reference Cheslock and Embers35]. This is the first time that a sequence associated with this putative species of Bartonella has been obtained in Mexico, although it has not been recognized as zoonotic, it is found within the clade that integrates species such as B. clarridgeiae and B. rochalimae which compromise human health.

It is important to note that certain rodent species can adapt to environments near human settlements, where they exploit available resources for their survival. This proximity can increase the likelihood of rodent-human interactions, potentially facilitating the transmission of bacteria or other infectious agents. In our study, some of the cricetid and heteromyid species (N. albigula, N. mexicana, P. boylii, C. baileyi, D. merriami, and P. fraterculus) presented high prevalence and/or zoonotic species of Bartonella; these rodents are usually found far from human settlements, as they are distributed in landscapes composed of forests, shrublands, deserts, and grasslands, so the risk of transmission of Bartonella bacteria to humans could be low [Reference Reid23]. However, some rodent species found in our study, such as N. albigula and N. mexicana, may occasionally be found in abandoned buildings [Reference Macêdo and Mares36, Reference Cornely and Baker37]. Furthermore, it should be considered that some human activities such as agricultural and livestock production, animal trade, deforestation, travel, and tourism, among others, imply the entry of humans into areas inhabited by wild animals, which increases the probability of contact and, in turn, the risk of transmission of Bartonella or other infectious agents [Reference Esposito38].

Our study, in conjunction with existing research, indicates that Bartonella is a persistent agent in wild rodents in north-western Mexico. The identification of certain sequences that correspond to zoonotic species underscores the necessity for preventive measures to avert the dissemination and occurrence of cases in humans. Bartonella species identification was carried out with the gltA gene, which is a reliable and widely used gene; however, for further studies, we recommend the use of multiple genes to discern between Bartonella species [Reference Kosoy17, Reference La Scola39]. Additionally, it is recommended that further research be conducted on this bacterium in Baja California to gain a deeper understanding of its prevalence and diversity within the state.