Rodents as potential reservoirs for Borrelia spp. in northern Chile

danigonz@udec.cl Abstract Small mammals play an essential role in the transmission and maintenance cycles of Borrelia spirochetes. In Chile, recent studies have characterized novel Borrelia genotypes in ticks collected from small mammals, a fact that suggests these vertebrates are hosts for spirochetes from this genus. Considering this evidence, the goal of this study was to determine the presence of Borrelia DNA in small mammals inhabiting northern Chile. In winter of 2018, 58 small mammals were captured in five localities. Blood samples were collected from rodents and DNA was extracted to determine the presence of Borrelia DNA by PCR targeting the flaB gene and rrs–rrlA intergenic spacer (IGS). From three individuals (5%), belonging to two rodent species of Cricetidae family ( Phyllotis xanthopygus and Oligoryzomys longicaudatus ), we retrieved three flaB and two IGS Borrelia genotypes. Phylogenetic analyses performed with both Maximum Likelihood and Bayesian inferences showed that our sequences grouped with homologous genotypes from the relapsing fever and Lyme borreliosis groups. Our findings suggest that P. xanthopygus and O. longicaudatus rodents may play a role as reservoirs for borrelial spirochetes in A10 sp. BLASTn comparisons revealed that flaB sequences Borrelia sp. A10 and Borrelia sp. A44


Introduction
In Chile, studies on bacterial infections in small mammals have been performed mostly in the central and southern regions of the country (Müller et al., 2018;Llanos-Soto & González-Acuña, 2019). In the northern region of Chile, research regarding bacterial infection on mammals consists in three surveys focusing on Escherichia and Salmonella of marine vertebrates (Otaria flavescens) (Salinas et al., 2010;Sturm et al., 2011;Toro et al., 2015). Data on vector-borne bacterial pathogens transmitted by mites or ticks are absent for this region of the country.
Rodentia is among the most diverse mammal order in Chile, with 69 species distributed along the country (MMA, 2018). This group of vertebrates plays an important role in the maintenance and propagation of tick-borne pathogens (bacterial, protozoan and viral) in urban and natural environments (Llanos-Soto & González-Acuña, 2019). Globally, rodents act as hosts for tick populations and serve as reservoirs for zoonotic pathogenic agents, such as Borrelia species (Cutler, 2015;Cutler et al., 2017).
Despite their importance for the maintenance of Borrelia infections elsewhere, the role of rodents as sylvatic reservoirs for this genus of spirochetes has not been properly addressed in Chile, since most studies have focused on vectors rather than potential vertebrate reservoirs. For instance, only one valid genospecies, Borrelia chilensis, has been identified in Ixodes stilesi ticks collected from the environment, and from long-tailed pygmy rice rats (Oligoryzomys longicaudatus) (Ivanova et al., 2014). In this case, although ticks were infected with B. chilensis, this does not necessarily mean that O. longicaudatus were carrying the spirochetes, since positive nymphs could have acquired the bacterium through a previous blood meal (Guttman et al., 1996). A similar scenario encompasses the recent findings of novel Borrelia genotypes in Ixodes sigelos s.l. group, and an Ornithodoros sp. closely related with Ornithodoros atacamensis in northern Chile (Muñoz-Leal et al., 2019a, b), for which their associated hosts, i. e. rodents of genus Phyllotis, were not assessed for Borrelia infection (Muñoz-Leal et al., 2019a). In this context, to elucidate the identity of vertebrate reservoirs for Borrelia is still a crucial step to understand transmission cycles of these bacteria in Chilean ecosystems. In this study, we aimed to assess the role of rodents and marsupials from northern Chile as potential reservoirs for Borrelia spp. through molecular analyses performed in blood obtained from these mammals.

Study area
This study surveyed rodents in five localities belonging to hyper-arid hydrographic regions from northern Chile ( Figure 1) during July (Austral winter) of 2018. Hyper-arid hydrographic region in Chile is characterized by having an annual precipitation and potential evapotranspiration ratio <0.05; the annual precipitation does not exceeds 100 mm, presenting an annual water deficit higher than 1200 mm; and dryness prevails throughout the year with a short peak of humidity that lasts one month (MMA, 2018).

Sample collection
Rodents and marsupials were captured using Sherman-like traps. Eighty traps remained active during two consecutive nights (10 hours per night) in each locality, and were placed along four parallel lines distanced approximately 100 m from each other, with 20 traps per line (spaced 10 m between each other). Animal handling was performed according to protocols used in field and laboratory studies on rodents (Herbreteau et al., 2011). Fifty microliters of blood were collected from each captured rodent through puncture of the caudal ventral vein and stored in sterile tubes with 96% ethanol (Sigma-Aldrich  ). All rodents were identified to the species level using a taxonomic guide (Iriarte, 2008), and after blood collection, were released in the same place of capture. DNA extraction and gene amplification DNA was extracted from blood using the DNeasy Blood & Tissue Kit (QUIAGEN, GERMANY). Forty microliters of Buffer AE (10 mM Tris-Cl; 0.5 mM EDTA, pH 9.0) were used to suspend the final DNA yield. To test successful extractions, and rule out the presence of PCR inhibitors, DNA quantity (concentration) and quality (purity and integrity) of DNA was assessed by A 260 / A 280 absorbance (A) in each sample using an Epoch™ Microplate Spectrophotometer. Samples with an A 260 / A 280 DNA ratio ranging between 1.6-2.0 were considered pure, and suitable for PCR amplification (Khare et al., 2014). Additionally, samples were tested through conventional PCR targeting endogenous gapdh (glyceraldehyde-3phosphate dehydrogenase) gene as an internal control following Birkenheuer et al. (2003).
BI was performed with two independent tests of 10 7 generations, running four MCMC chains, sampling trees every 1000 generations, and discarding the first 25% as burn-in. MCMC chain correlation was confirmed with Tracer v1.7.1 (Rambaut et al., 2018). Statistical support of internal nodes was evaluated employing Bayesian posterior probabilities (BPPs) and considering values ≥0.70 as strong support . The ML analysis was carried using rapid hill-climbing and stochastic disturbance methods, evaluating the robustness of the inferred tree with 1000 pseudo-replicates of ultrafast bootstrapping. We used the criteria of Minh et al. (2013) to evaluate the ultrafast bootstrap: values <70% were considered non-significant statistical support; values between 70-94% as moderately significant; and values ≥ 95% as highly significant.

Positive animals and PCR
A total of 58 small mammals belonging to 12 species in the families Cricetidae, Muridae and Didelphidae were captured ( Table 2). Although DNA purity obtained after measurements of A 260 / A 280 absorbance ratio was optimal in 56/58 of the samples (97%), five samples (including the two with low A 260 / A 280 absorbance ratio) were negative after GAPDH gene PCR and excluded from further analyses. Three out of 53 rodents (5%) were positive for Borrelia flaB screening, and two of these samples were positive for IGS (4%) ( Table 2).
Successfully extracted samples were screened for Borrelia flagellin gene (flaB), and rrs-rrlA intergenic spacer (IGS) through nested PCR protocols, using primers listed in Table 1. Reactions were performed into a final volume of 25 μL containing 12.5 μL of Dream Taq Green PCR Master Mix (Thermo Scientific, USA), 1 μL of each primer (10 pmol), 2 μL of DNA for conventional PCR, 1 μL of the product for nested rounds, and ultra-pure water to complete the final volume of the mix. Amplicons were verified in 1.5% agarose gels stained with SYBR Safe (Life Technologies/Thermo Fisher Scientific, Carlsbad, CA), and visualized through UV light. Positive samples were purified and sequenced in both directions at AUSTRAL-omics (Valdivia, Chile).  Two different IGS sequences were obtained from the same samples of P. xantophygus positive for flaB gene. Sequences were 96.49% identical between them and 94.36% identical (202/214 bp, 41% query cover, 3 gaps, 8e-85 E-value) to Borrelia sp. TM (DQ000283; referred as "cf. Borrelia crocidurae" amplified from ticks). IGS sequences from Borrelia sp. A10 and A44 were deposited in Genbank under accession numbers MN598782 and MN598783, respectively.

Phylogenetic analyses
Overall, BI and ML phylogenetic trees for flaB and IGS depicted similar and well-supported logic topologies, grouping Borrelia spp. into LB and RF groups. In particular, phylogenetic analyses for flaB gene positioned our sequences within a clade with Borrelia sp. 95325 (HM583797) having statistically significant support (BPP=1 in Figure 2A, and Bootstrap=100 in Figure 2B). For the BI analysis, the clade composed by our sequences and Borrelia sp. 95325 formed a clade with high statistical support (BPP=0.98 in Figure 2A) with Borrelia latyschewii (JF708952), Borrelia microti (JF708951), Borrelia duttonii (NC011229), Borrelia recurrentis (CP000993), Borrelia crocidurae (CP004267), Borrelia hispanica (MF432465), and Borrelia persica (NZAYOT01000225). However, the internal relations of this clade were undefined. On the other hand, with high support (Bootstrap=97 in Figure 2B), the ML analysis supported Borrelia sp. A10 and A44 as independent branches into a monophyletic group with Borrelia sp. 953225 (HM583797). On the other hand, and with high statistical support (BPP= 0.84 in Figure 2A, and Bootstrap= 90 in Figure 2B), Borrelia sp. A53 clustered with Borrelia genotypes characterized from ticks belonging to the I. sigelos group (MH187987, MH178397), with B. chilensis VA1 (CP009910), and Borrelia sp. ISIG1 (KX417768) obtained from I. cf. neuquenensis, and I. sigelos from Argentina as sister groups (Figure 2).
With high statistical support (BBP= 1 in Figure 3A, and Bootstrap= 96 in Figure 3B), phylogenetic analyses for IGS sequences showed that Borrelia sp. A10 and Borrelia sp. A44 form an independent clade related to Borrelia spp. belonging to the relapsing fever group (Figure 3).

Discussion
The identification of wild vertebrate reservoirs implicated in the maintenance of pathogenic agents should be considered a permanent task in scientific research (Karesh et al., 2012). Rodents are important sylvatic reservoirs, as at least 217 out of 2777 known species harbor 66 zoonotic agents (Han et al., 2015). Considering this scenario, and the recent detection of Borrelia spp. in rodent-associated ticks in Chile, we aimed to assess the presence of Borrelia DNA in 12 species of small mammals (ten rodents and two marsupials) from this country. We detected DNA of Borrelia spp. belonging to the LB and RF groups in two cricetid rodents, namely P. xanthopygus and O. longicaudatus, respectively. While cricetid rodents (i. e. Peromyscus leucopus) have been previously reported as competent reservoirs for Borrelia burgdorferi sensu stricto (s.s.) in North America (Levine et al., 1985;Hofmeister et al., 1999;Bunikis et al., 2004), high prevalence for Borrelia burgdorferi s.l. have been reported in synanthropic murid rodents (Mus musculus and Rattus rattus) too (Solís-Hernández et al., 2016). In our study, synanthropic M. musculus were negative to Borrelia detection, a fact that could be attributed to the small sample of this species that was analyzed.
Apart from rodents, we assessed blood from marsupials, but with negative results. Nevertheless, Borrelia-like spirochetes have already been isolated from opossums (Marsupialia) in the United States (Hanson, 1970), so the role of Chilean marsupials as hosts for Borrelia spp. should not be discarded. Studies focusing on small mammals and enzootic cycles of borrelial spirochetes have been performed only for Northern (Levine et al., 1985;Hofmeister et al., 1999;Bunikis et al., 2004) and Central American species (Solís-Hernández et al., 2016). Whether rodents may act as reservoirs for Borrelia in South America remained unknown until the current study.  Parola et al. (2011) formed an independent clade with the sequences obtained in our study (Borrelia sp. A10 and Borrelia sp. A44), suggesting a close phylogenetic relationship. The phylogenetic resolution of the flaB gene has been useful to define lineages in the genus Borrelia (Fukunaga et al., 1996). In agreement with this fact, our BI and ML analyses indicated with high support that the detected genotypes constitute putatively new species (Figure 2). According to these results, the BI and ML phylogenies for IGS also point that the sequences of Borrelia sp. A10 and Borrelia sp. A44 constitute novel taxa, related to RF borreliae (Figure 3). To date, only one rodent parasitized by an Ornithodoros sp. has been proposed as putative reservoir for a RF Borrelia sp. in Chile (Muñoz-Leal et al., 2019b). However, the sequences detected in this study differ from Borrelia genotypes previously detected in soft ticks from this country (Muñoz-Leal et al., 2019a).
Different genotypes of flaB gene belonging to the LB group have been reported in Chile, namely Borrelia sp. Navarino (MH178398), characterized from Ixodes auritulus collected in the bird a Troglodytes musculus, and several genotypes related to B. chilensis from ticks of the I. sigelos group (MH178397, MH187987, CP009910; Ivanova et al.2014;Muñoz-Leal et al., 2019a). Even though those studies have made valuable contributions to the understanding of the diversity of Borrelia in Chile, evidence of mammal hosts acting as reservoir for these agents was previously non-existent. In this study, sequences of Borrelia flaB gene retrieved from blood of O. longicaudatus (Borrelia sp. A53) branched as independent genotypes within the LB group of borreliae ( Figure 2). Remarkably, Borrelia sp. A53 was also related to B. chilensis (CP009910), a genospecies previously reported in I. stilesi parasitizing the same rodent species in Southern Chile (Ivanova et al., 2014). Moreover, genotypes of Borrelia detected in larvae and nymphs of I. sigelos s.l. collected on P. darwini and Octodon degus (Muñoz-Leal et al., 2019a) clustered with Borrelia sp. A53 as well. This fact suggests that LB genotypes of Borrelia associated with rodents could constitute a monophyletic group related to B. chilensis. A similar hypothesis pointing a natural group of borreliae infecting rodents has been proposed through phylogenetic analyses using Borrelia genotypes detected in rodent-associated ticks in Southern Argentina (Sebastian et al., 2016).
To our knowledge, this is the first report of infection by Borrelia spp. in small mammals from Chile and South America, and the first isolation of DNA from these spirochetes in P. xanthopygus and O. longicaudatus. Our results suggest that these rodents may act as potential reservoirs for novel Borrelia genotypes in natural ecosystems of northern Chile. However, future studies are needed to further determine the competence of these rodents in maintaining Borrelia infections, and to investigate if other species of small mammals participate in the enzootic cycles of these spirochetes as well.