Molecular Survey of Bacterial Zoonotic Agents in Bats from the Country of Georgia (Caucasus)

Bats are important reservoirs for many zoonotic pathogens. However, no surveys of bacterial pathogens in bats have been performed in the Caucasus region. To understand the occurrence and distribution of bacterial infections in these mammals, 218 bats belonging to eight species collected from four regions of Georgia were examined for Bartonella, Brucella, Leptospira, and Yersinia using molecular approaches. Bartonella DNA was detected in 77 (35%) bats from all eight species and was distributed in all four regions. The prevalence ranged 6–50% per bat species. The Bartonella DNA represented 25 unique genetic variants that clustered into 21 lineages. Brucella DNA was detected in two Miniopterus schreibersii bats and in two Myotis blythii bats, all of which were from Imereti (west-central region). Leptospira DNA was detected in 25 (13%) bats that included four M. schreibersii bats and 21 M. blythii bats collected from two regions. The Leptospira sequences represented five genetic variants with one of them being closely related to the zoonotic pathogen L. interrogans (98.6% genetic identity). No Yersinia DNA was detected in the bats. Mixed infections were observed in several cases. One M. blythii bat and one M. schreibersii bat were co-infected with Bartonella, Brucella, and Leptospira; one M. blythii bat and one M. schreibersii bat were co-infected with Bartonella and Brucella; 15 M. blythii bats and three M. schreibersii bats were co-infected with Bartonella and Leptospira. Our results suggest that bats in Georgia are exposed to multiple bacterial infections. Further studies are needed to evaluate pathogenicity of these agents to bats and their zoonotic potential.

Introduction Bats (Chiroptera) represent one of the most successfully evolved mammalian groups on Earth for their unique characteristics, such as a long lifespan, the capability to fly long distances during foraging and particularly during seasonal migrations, the ability to inhabit a multitude of diverse ecological niches, and the colonial habitation. The role of bats in epidemiology of zoonotic diseases is very important as they frequently live in close proximity to humans and serve as reservoirs to different pathogens that include viruses, bacteria, fungi and parasites [1,2]. Many previous and ongoing research activities predominantly focused on viral agents in bats [3,4], but little is known about a presence of bacterial pathogens [5].
The bacterial genera Bartonella, Brucella, Leptospira, and Yersinia each consist of multiple species, some of which are zoonotic pathogens causing diseases in domestic or companion animals and in humans. Bartonella infections have been reported from a variety of animals occurring over a broad geographic distribution. Around 30 species have been described within the genus and the number is still increasing [6]. Controversy has been raised in several studies regarding host specificity of Bartonella [7][8][9]. Importantly, bats in the Northern Hemisphere have been implicated as a reservoir of B. mayotimonensis that was described from a human case of endocarditis in the USA [10,11], although the mechanism of transmission between bats and humans remains unresolved.
Brucellosis is an important zoonotic disease caused by bacteria of the genus Brucella. Domestic animals such as cattle, goats, sheep, pigs, camel, buffalo and dogs serve as reservoir hosts. Humans can be infected after contacting infectious animals or drinking raw milk [12]. Knowledge of Brucella ecology in wildlife is limited although several species were described in rodents, foxes, and marine mammals [13][14][15][16][17]. Except for an old report of anti-Brucella agglutinins in vampire bats (Desmodus rotundus) in Brazil [18], no other studies have reported Brucella infection in bats.
Leptospirosis is a bacterial zoonosis caused by L. interrogans and other pathogenic spirochetes of the genus Leptospira. Animals and humans acquire the infection through contact with water or soil contaminated with the urine of infected animals or by direct contact with these animals [19,20]. Leptospira spp. are distributed worldwide in rats and many other mammalian species [21]. Recently, leptospiral infections have been identified in bats from several countries [22][23][24].
Among infections caused by Yersinia species, plague (Y. pestis) causes the most notorious disease, infecting many mammalian species along with humans. Yersiniosis, occurring as an enteric disease in humans, is caused by Y. pseudotuberculosis and Y. enterocolitica, both of which have a broad distribution [25]. These bacteria have been frequently isolated from a variety of wild and domestic animals [26,27], but compared to Y. pestis their association with wildlife is not as well studied. Recently, bats have been reported to be infected with Yersinia [28,29].
The country of Georgia is located between the Greater Caucasus and Lesser Caucasus mountain ridges at the intersection of Europe and Asia. There are 109 mammalian species and many associated zoonotic agents in this region [30,31]. Recent studies conducted in this country have demonstrated the presence of diverse Bartonella species in wild rodents [32] and suggested the role of rat-associated Bartonella as a causative agent for a human illness [33]. Brucellosis is endemic in the area, with B. abortus and B. melitensis actively circulating in livestock and affecting local residents [34]. Leptospirosis is also broadly distributed in the country with increasing morbidity in recent years [35].
Bats are abundant in Georgia with at least 29 species identified [36]. However, information on bacterial infectious agents in bats from this region was absent. Understanding the prevalence and distribution of zoonotic pathogens in local bats would be significant from the veterinary and public health perspectives. In this study, we evaluated the presence and distribution of Bartonella, Brucella, Leptospira, and Yersinia in bats collected from eight colonies located in four regions of Georgia.

Ethics statement
The work was performed in compliance with the protocol approved by the CDC Institutional Animal Care and Use Committee (protocol #2096FRAMULX-A3). Permits for the list of bat species sought and the number of animals per colony available for sampling were obtained from the Ministry of Environmental and Natural Resources Protection of Georgia.

Study sites and bat tissues collection
In June 2012, bats were captured manually or by using nets from eight colonies (found in caves, building attics, or monasteries) within four sites located in four regions of Georgia: one colony in Martvili (42˚N, 42˚E, Samegrelo-Zemo Svaneti region in western Georgia); three colonies in Tskaltubo (42˚N, 42˚E, Imereti region in west-central Georgia); one colony in Gardabani district (41˚N, 45˚E, Kvemo Kartli region in southern Georgia); and three colonies in David Gareja (41˚N, 45˚E, Kakheti region in eastern Georgia). Martvili and Tskaltubo, Gardabani and David Gareja, are neighboring sites, respectively (Fig 1). Captured bats were delivered to the processing site in individual cotton bags. The bats were sexed, weighed, and identified to species based on external morphological characteristics. Bats were anesthetized using ketamine (0.05-0.1 mg/g body mass) and exsanguinated by cardiac puncture. Tissue samples (including spleen, kidney, intestine, and others) were collected from the bats. Samples were stored on dry ice in the field, then transferred to a -80˚C freezer in the laboratory of Georgian NCDC before shipping to the US CDC's laboratory in Fort Collins, Colorado for bacterial testing.

DNA extraction and PCR detection
A small piece (~10 mg) of spleen, kidney, and intestine of each bat were homogenized separately using a Bullet Blender1 Gold homogenizer (Next Advance, Averill Park, NY) following the protocols provided by the manufacturer. The homogenates were then transferred to a QIAxtractor (Qiagen, Valencia, CA) platform for DNA extraction using the tissue protocol, and the DNA was used as the template for downstream analyses. The kidney DNA was tested for Bartonella and Leptospira; the spleen DNA was tested for Brucella and Yersinia; and the intestine DNA was tested for Yersinia only. Molecular detection was performed using a conventional PCR assay carried out in a C1000 Touch Thermal Cycler (Bio-Rad, Hercules, CA) and/or real-time PCR assay carried out in a CFX96 Real-Time System (Bio-Rad, Hercules, CA). The relevant genes targeted were 16S -23S internal transcribed spacer (ITS), insertion sequence (IS711), 32-kDa lipoprotein (lipL32) gene, and peptidoglycan-associated lipoprotein (pal) gene for Bartonella, Brucella, Leptospira, and Yersinia, respectively. The reactions for ITS were run under conventional PCR settings; the reactions for IS711, pal and lipL32were run under the real-time multiplex PCR settings, following protocols published elsewhere [37][38][39]. Ct value < 36 with an amplification curve is recorded as positive. Positive samples for Brucella DNA (IS711) and Leptospira DNA (lipL32) identified by real-time PCR were further tested by conventional PCR targeting a 223 base pair-fragment in 31 kDa gene (bcsp31) for Brucella [40] and a 423 base pair-fragment in lipL32 for Leptospira using primers [41] different from the real-time PCR. All primers and probes used in this study are listed in Table 1. The pal primers and probes for detection of Yersinia species were developed for this study based on a wholegenome scan (M. Diaz, unpublished data). For any conventional PCR, the PCR products were analyzed for the presence of amplicons of the expected size by electrophoresis on 1.5% agarose gels containing GelGreen stain (Biotium, Hayward, CA). Positive and negative controls were included in each PCR assay to evaluate the presence of appropriately sized amplicons and to rule out potential contamination, respectively. Leptospira infection rate was compared between study sites and between bat species using chi-square tests.
Sequencing and phylogenetic analysis for Bartonella species and Leptospira species Samples positive for Bartonella (ITS) and Leptospira (lipL32) were further identified by sequencing analyses. The PCR amplicons were purified using a QIAquick PCR Purification Kit (Qiagen, Valencia, CA) according to manufacturer's instructions, and then sequenced in both directions using ABI 3130 Genetic Analyzer (Applied Biosystems, Foster City, CA). Forward and reverse sequences were assembled using the SeqMan Pro program in Lasergene v.12 (DNASTAR, Madison, WI). Assembled sequences obtained from all samples in the present study were compared between themselves and with reference sequences available in GenBank after alignment using the Clustal algorithm in the MegAlign program in Lasergene. Using the neighbor-joining method, cladogram (showing the cladistics relationship rather than phylogenetic relationship) was generated for Bartonella among the ITS sequences. Sequences were assigned to clades visually based on monophyletic clusters. Phylogenetic tree was constructed for Leptospira, Branch support was estimated using 1000 bootstrap replicates. Newly identified sequence variants were submitted to GenBank.
The number of bat species varied by site, with three, four, five, and three in Martvili, Tskaltubo, Gardabani, and David Gareja, respectively. All M. schreibersii bats (n = 27) were collected in Tskaltubo. M. blythii bats were mainly captured in Tskaltubo and David Gareja (Table 3).

Sequencing analysis
Sequencing analyses were performed on ITS sequences of Bartonella and lipL32 sequences of Leptospira, both of which were amplified from bat kidney DNA. The ITS sequences of Bartonella exhibited considerable heterogeneity. The 77 Bartonellapositive DNA represented 25 genetic variants (a variant is defined when at least one nucleotide difference is observed between compared sequences). These variants were all unique and submitted to GenBank (accession numbers KX420713-KX420737). Only a few variants were relatively close (96.3% similarity) between themselves, while the majority were largely distant from each other and clustered into 21 clades (a clade is a cluster of sequences following neighbor-joining analysis) (Fig 2). Except for E. serotinus and P. pygmaeus in which Bartonella was detected in one individual for each species, each of the other bat species was associated with 4 to 9 Bartonella clades. For example, M. blythii was associated with nine lineages (I-IV, XV, and XVII-XXI); M. schreibersii was associated with six lineages (III-VI, XVI, and XVIII); R. euryale was associated with eight lineages (VII-XIII and XVII). Most Bartonella lineages were specific to certain bat genus/species. In particular, lineages I-VI and lineages VII-XIII were associated with Myotis spp. and Rhinolophus spp., respectively; while lineages XV-XXI could be associated with multiple bat genera (Fig 2).
The 25 LipL32 sequences of Leptospira detected in the bats represented five variants with sequence distances of 3.7% -7.3% (Fig 3). All variants were novel and assigned GenBank accession numbers KX420708 -KX420712. Of the 25 sequences, the four sequences recovered from M. schreibersii were identical (variant KX420712); while the other 21 sequences from M. blythii were of four variants, each of which was detected in 11, 4, 2, and 4 individuals, respectively. These variants were close to some Leptospira species that are known zoonotic pathogens. Specifically, variant KX420710 that was detected in 11 M. blythii was closest to L. interrogans with genetic identity of 98.6%; while a few other variants were relatively close to L. borgpetersenii with genetic identity of 96% -97%.

Discussion
Using molecular approaches, we report the detection of multiple potential zoonotic pathogens, including Bartonella, Brucella, and Leptospira in bats from the country of Georgia. Similar to early reports from other parts of the world [8,42], Bartonella species are widely distributed in bats in Georgia with very high diversity. All genetic variants discovered in the study were novel and do not belong to any previously described Bartonella species. Further characterization is necessary to verify whether the identified DNA sequences represent novel Bartonella species. Most Bartonella variants described here show specific relationships to their bat hosts at a genus level, particularly Myotis spp. and Rhinolophus spp. Conversely, some Bartonella variants were shared among bats of different genera, which suggests frequent host-shifting or cross-species transmission potentially related to exchange of ectoparasites between bats [43,44]. We do not know whether the Bartonella variants identified in the bats are responsible for any human disease in Georgia, but a recent report that bats in northern Europe harbor a human pathogen B. mayotimonensis [11] indicates such a possibility. New information about Bartonella infection in bats in Georgia can provide additional insights towards understanding the interactions between humans, animals, and parasites.
Leptospira infection was detected in the tested bats. Interestingly, all Leptospira infectedbats were either M. blythii or M. schreibersii. No Leptospira infection was detected in bats of other species, including the M. emarginatus which is a quite common local species. Speciesspecific variations in bacterial infection rates may indicate that certain bat species are more exposed habitually (e.g. drinking the same contaminated water) or even more susceptible to Leptospira than other bat species [23,45,46]. Our observation of similar Leptospira infection rates in M. blythii in Tskaltubo and David Gareja suggests that these bats were equally exposed to the infection at different sites; while higher prevalence in M. blythii than in M. schreibersii may suggest that M. blythii is more susceptible to Leptospira. Field observations showed that M. blythii and M. schreibersii share the same roosts and M. schreibersii usually incorporate into the dense groups of M. blythii, which presumably results in close body to body contact between animals of these two species. M. blythii may transmit the infection to M. schreibersii through urinary shedding and other similar routes. The high prevalence observed suggests that these bats might play a possible role in the maintenance of Leptospira spp. in the environment [47]. On the other hand, one Leptospira variant identified in M. blythii was closely related to L. interrogans (98.6% identity), a well-known zoonotic pathogen frequently found in rats [48]. Considering the high frequency of this variant (detected in 11 of 25 or 44% infected bats), it warns that M. blythii may serve as a natural reservoir to L. interrogans and can potentially transmit the infection to humans, particularly when they roost synantropically, e.g. in monasteries. Furthermore, some of the variants were relatively close to L. borgpetersenii that also is a zoonotic pathogen.
The most intriguing finding of this work is probably the discovery of Brucella in bats. Although there was a single report of Brucella agglutinins in Desmodus rotundus bats from Brazil [18], our study represents the first detection of Brucella DNA in bats. Similar to the finding of Leptospira spp., the Brucella infections were found only in M. schreibersii and M. blythii. Interestingly, all Brucella positive bats were from the site Tskaltubo. The habitat preference and geographic origin could influence the infection prevalence in these bats. It was evident for M. schreibersii since this species was only captured in Tskaltubo; whereas absence of Brucellapositive M. blythii bats in David Gareja (where no M. schreibersii were present) may suggest that M. blythii contracted Brucella infection from M. schreibersii in Tskaltubo. Alternatively, as was the case with Leptospira, bats of both species in Tskaltubo might be exposed to Brucella independently by consuming the same contaminated source, such as water. The most studied Brucella species (B. melitensis, B. abortus, and B. suis) are so genetically similar that some researchers have argued for their combination into one species [49]. However, the epidemiological and diagnostic benefits for separating the genus based on phenotypic and ecological characteristics are more compelling. Therefore, it is important to distinguish genotypic and phenotypic differences between this newly discovered bat-associated Brucella and other Brucella species. In the current presentation, we only report the detection of Brucella DNA in the bats. Further characterization of the bacterium is in progress. Because most Brucella species are highly pathogenic, it is important to determine whether this strain has a zoonotic potential.
We did not find any Yersinia infections in this study using recently designed primers. Since the sensitivity of the utilized primers for detection of all Yersinia species in the field samples has not been sufficiently evaluated in other field studies, the negative results for this bacterial genus in Georgian bats might represent a lack of sensitivity of our assay or a need for more broadly reactive primers.
In conclusion, bats from Georgia harbor several potential bacterial pathogens. These preliminary data highlight that bats may play an important role in maintaining those agents in nature. Further studies need to be carried out to understand the importance of these agents for bats, other wildlife, veterinary and public health.