Stray dogs in Nepal have high prevalence of vector-borne pathogens: a molecular survey

Background Population of stray dogs is significant in large cities of Nepal, such as Kathmandu. Most of stray dogs suffer a lack of basic health care. Considering the clinical relevance, the broad distribution and the lack of information of canine vector borne diseases (CVBD) in Nepal, the aim of this study was to evaluate the prevalence of different vector-borne pathogens (VBP) in stray dogs living in the metropolitan area of Kathmandu, and to assess different traits as possible risk factors. Methods A total of 70 canine blood samples from stray dogs attended at the Kathmandu Animal Treatment Centre during August 2017 were collected on filter paper (Flinders Technology Associates (FTA) cards). Data regarding signalment, clinical signs and epidemiological characteristics were recorded for each animal. Real-time polymerase chain reaction assays were performed for Leishmania spp., Ehrlichia spp./Anaplasma spp., Babesia spp./Theileria spp. and Hepatozoon canis. Results The overall prevalence detected was 31.43% for Hepatozoon canis, 31.43% for Anaplasma platys, 27.14% for Ehrlichia canis, 18.57% for Leishmania donovani species complex, 12.86% for isolates corresponding to Theileria spp., 12.86% for Babesia vogeli and 2.86% for B. gibsoni. A total of 81.43% of the dogs were positive to at least one of the VBP tested. Co-infections were detected in 41.43% of the dogs. Dogs positive to any of the VBP tested, and particularly to E. canis, were older than those that were negative. Conclusions To our knowledge, this is the first molecular detection of VBP in stray dogs from Kathmandu, Nepal. The high prevalence of VBP detected highlights the need to implement a surveillance programme and control strategies for these CVBD in the population of stray dogs in this area.

and appropiate for the presence of different vectors of relevance for the dog population such as sand flies and hard ticks.
The aim of this study was to evaluate the prevalence of different VBP in stray dogs living in the metropolitan area of Kathmandu, Nepal, and to assess different traits as possible risk factors. Taking into account the clinical relevance, the broad distribution and the lack of information in dogs from Nepal, the pathogens evaluated in this study were Leishmania spp. Ehrlichia spp./Anaplasma spp., Babesia spp./Theileria spp. and Hepatozoon canis.

Methods
Canine blood samples from stray dogs attending the Kathmandu Animal Treatment Centre during August 2017 were collected for further molecular analyses. Different data regarding signalment, clinical signs, and epidemiological characteristics were recorded for each animal. We aimed to perform real-time polymerase chain reaction assays for Leishmania spp., Ehrlichia spp./Anaplasma spp., Babesia spp./Theileria spp. and Hepatozoon canis.

Animals and samples, and data collection
A total of 70 blood samples from stray dogs attended at the Kathmandu Animal Treatment Centre (KAT Centre) during August 2017 were collected from the core of the metropolitan area of Kathmandu, Nepal. Stray dogs were opportunistically sampled when they were brought into the clinic for medical treatment or for neutering procedures as a part of animal birth control (ABC) programmes. Dogs with and without clinical signs were included in this study. Dog blood samples obtained by venipuncture were spotted on Whatman FTA ® (Flinders Technology Associates) classic cards (Whatman International Ltd, Maidstone, UK) for further molecular analyses. To avoid cross-contamination, these samples were stored in separate plastic bags that included storage desiccant packets to ensure that FTA cards remained dry during transport and storage. Samples were kept at room temperature during the entire campaign and were later stored at − 20 °C until DNA extraction. Data regarding signalment, clinical signs and presence of ectoparasites were recorded for each animal.
DNA extraction and quality assessment DNA was extracted using SpeedTools Tissue DNA extraction commercial kit (Biotools, B&M Laboratories, S.A., Madrid, Spain) according to the manufacturer's instructions. All DNA samples were stored at − 20 °C until use. The extraction yield (quality and quantity of the extracted DNA) was assessed by means of spectrophotometry (Nanodrop TM ; Thermo Fisher Scientific, Waltham, USA). The eukaryotic 18S ribosomal ribonucleic acid (18S rRNA) (Thermo Fisher Scientific) was used as an internal reference of canine genomic DNA to ensure proper extraction (presence/absence of DNA inhibition factors).

Real-time PCR assays
Real-time PCR assays were used to target selected species of VBP, including Babesia/Theileria species, Ehrlichia/ Anaplasma species, Hepatozoon canis and Leishmania donovani complex species as previously described [6,7].
Real-time PCR reactions were conducted in a 20 µl reaction mixture containing PowerUp SYBR Green master mix (Thermo Fisher Scientific, Carlsbad, CA, USA), specific primers (Table 1) and 4 µl of 1/5 diluted DNA. The thermal cycling profile was 50 °C for 2 min and 95 °C for 2 min, followed by 40 cycles at 95 °C for 3 s and 60 °C for 30 s and a dissociation curve added at the end of the run to asses PCR-specificity. Commercial DNA and water were used with each amplification run as a positive and negative PCR controls, respectively. Positive-PCR amplicons were directly sequenced to characterize pathogens at the species level using the same primers. Subsequent sequencing was performed in all the positive samples, using the Big Dye ® Terminator version 3.1 Cycle Sequencing kit (Thermo Fisher Scientific, Carlsbad, CA, USA) following the manufacturer's instructions. Sequences obtained were compared with those deposited on GenBank using the Basic Local Alignment Search Tool (BLAST).

Statistical analysis
The results were statistically analysed using the software SAS, version 9.4 (SAS Institute, Cary, NC, USA). Statistical associations between VBP and epidemiological data recorded for each stray dog were analysed using Chi-square test or Fisherʼs exact test, where appropriate, and Studentʼs t-test to evaluate mean age with positive or negative effects for different vector-borne pathogens. In order to exclude possible confounding factors, complementary logistic regression analysis with backward elimination was performed with those variables that showed a statistical association. The level of statistical significance was established at P < 0.05.
Ehrlichia canis was the only pathogen displaying a statistical association with dog-derived traits. When looking to E. canis infection specifically, it was observed that dogs with a positive PCR result were also older (5.16 ± 3.39 years-old) than negative dogs (3.27 ± 2.59 yearsold) (Studentʼs t-test, t = 2.49, P = 0.0153). On the other hand, neutered status was also associated with PCR-positive result for E. canis (χ 2 = 5.07, df = 1, P = 0.024). However, complementary logistic regression analysis with backward elimination was performed including those variables in a model and demonstrated that these associations were not statistically significant (neutered status (P = 0.09; OR: 2.68; 95% CI: 0.85-8.43) and age (P = 0.067; OR: 1.2; 95% CI: 0.99-1.47).

Discussion
To our knowledge, this study addressed the first molecular survey of VBP in stray dogs from the metropolitan area of Kathmandu, Nepal, and the first report of H. canis infection in dogs from this region. A high prevalence of VBP was detected in this study using real-time PCR technique. The results presented here support the scarce findings of previous studies that detected Leishmania spp. in dogs in some regions of Nepal [8]. Babesia sp., Ehrlichia sp. and Anaplasma sp. have also been recently detected by microscopy in blood smears from hyperthermic owned dogs in Kathmandu Valley [9,10]. Previous studies in the region have found an overall prevalence of haemoparasites between 10-17.14% in dogs considered under risk based on clinical signs like fever or hyperthermic dogs, or those carrying tickinfestation [9]. However, our study shows a higher prevalence of selected VBP (81.43%). This could be explained because previous studies were performed on owned dogs (that usually receive the adequate veterinarian care). Most of the stray dogs involved in our study were highly exposed to tick and sand fly bites. In addition, differences in the techniques employed for the detection of these pathogens could explain, at least partially, the differences in the prevalences detected. It has been previously established that real time-PCR is considerably more sensitive than visualization of the VBP in a blood smear, especially in those cases of low burden of VBP [11][12][13][14].
Hepatozoon canis and A. platys were the most frequently detected agents. Hepatozoon canis has been recently detected in dogs from Northeast India with a prevalence of 38% [15], slightly higher than the prevalence described here (31.43%). These apicomplexan protozoans are transmitted by the ingestion of an infected tick [16]. The presence of ticks on dogs is sometimes used in practice as a predictor of some CVBD, especially in the lack of specific diagnostic tests. However, this study shows the lack of significant association between H. canis infection and detection of ticks in dogs during the physical exam (P = 0.34).
Sequencing confirmed that 31.43% of the dogs were positive to A. platys, and 27.14% to E. canis. Both rickettsial agents are transmitted by the same vector, Rhipicephalus sanguineus (sensu lato). As previously described for H. canis, tick infestation was detected in a high percentage of dogs infected by A. platys or E. canis (63.63% and 68.42%, respectively). Previous studies showed high prevalence of ectoparasite infestation in dogs from this area [2], and a recent review did show that a wide variety of tick species are parasites of dogs in Nepal [17].
Infections with Anaplasma spp. (1.34%) and Ehrlichia spp. (10.66%) have been recently detected in blood smears from hyperthermic dogs in Kathmandu [10]. The dogs infected with E. canis were older than the negative dogs in our study, probably due to a higher exposure to the vectors and their pathogens. Ehrlichia canis infection was more frequently detected in male dogs, although statistical test was non-significant (P = 0.08). This could be related to behavioural characteristics, as previously suggested [18].
Nepal is recognized as endemic for L. donovani [19]. However, the role of the dog as a reservoir in the cycle of this anthroponotic visceral leishmaniasis (AVL) (also known as kala-azar) and transmitted by Phlebotomus argentipes (anthropophilic vector) remains unknown. Some studies in India have hypothesized that dogs may constitute a reservoir for this species of Leishmania [20]. A previous study performed in the Terai region (southern Nepal) has shown that P. argentipes and P. papatasi with predilection to feed on humans or cattle, also feed on dog blood [8]. Phlebotomus papatasi has not been incriminated as a vector for L. donovani [8,21]. However, this sand fly is a reported vector for other Leishmania spp. (L. major and L. infantum) in the Old World, suggesting that Nepal could become endemic for zoonotic leishmaniosis [22]. Our study shows a prevalence of 18.57% of Leishmania donovani complex in the tested stray dogs. There was no statistically significant correlation between Leishmania spp. detection and any signalment data, clinical findings, and environmental data. It was not possible to establish if these animals were suffering a clinical presentation of leishmaniosis or were just infected and carrying the pathogen at a low burden. The lack of compatible clinical signs and the low number of copies of parasite DNA detected, could support the second hypothesis. Thus, having confirmed the presence of Leishmania spp. by molecular tests, it would be interesting in the future to establish the role of the dog in the human leishmaniosis in Kathmandu.
Canine babesiosis has been recently reported in dogs from Kathmandu [9] and B. gibsoni constituted the most detected vector-borne agent found in dogs from Northeast India (43% of the dogs positive to B. gibsoni and 3% positive to B. vogeli) [15]. In contrast, in our study, prevalence was higher for B. vogeli (12.86%), and lower for B. gibsoni (2.86%). Babesia gibsoni has been widely described in Asia [15] and is transmitted by Haemaphysalis longicornis [23]. Other species of the genus Haemaphysalis have been previously detected in Nepal [17]. On the other hand, the main vectors described for B. vogeli are ticks from different genera (Dermacentor spp., Ixodes spp., and R. sanguineus (s.l.)) also found in dogs from Nepal [17]. The authors of the present study are cautious with the DNA sequences of Theileria spp. detected in dogs.
An unexpected finding from this study was the high rate of co-infections, detected in 29 out of 70 dogs (41.43%). In agreement with this finding, another recent study in India has shown a high prevalence of dogs presenting co-infection by two or more agents (44/130, 34%) [15]. The most common co-infection detected in our study was H. canis + E. canis. Canine hepatozoonosis and ehrlichiosis are both tick-borne diseases transmitted by R. sanguineus (s.l.). Some interactions between the pathogens and host cells in this co-infection have been previously suggested [24]. The high ratio of co-infections could be explained by the interactions of some of these pathogens when infecting dogs, as previously described [25].
There was no effect of BCS or sex on the prevalence of these VBP. The animals included in the study were stray dogs with difficulties to feed properly and most of them presented a low BCS. This could be related to the lack of statistical differences regardless of whether they present clinical signs. A previous study in Nepal detected that most of the dogs (80%; n = 47) had an adequate BCS. However, ecto-and endoparasites were also detected in most of them (83%; n = 49) [2].The lack of association between the presence of infection and the sex of the dogs in this study is consistent with other studies evaluating VBP in dogs from India [15,26] and from Kathmandu Valley, Nepal [9,10].
Clinical findings detected in the dogs included in the study were not directly related with the classical clinical picture of CVBD, and consisted mainly on wounds, myiasis, bone fractures, fungal skin lesions, and limb loss, among others. The lack of clinical signs in asymptomatic carriers, the co-infection with multiple VBP in the same animal or even the poor health condition in non-infected dogs constitute one of the barriers in the clinical diagnosis approach for these CVBD. In agreement with our work, a previous study performed in Chitwan District (central Nepal), also showed that there was a considerable number of apparently healthy dogs carrying haemoprotozoans and ectoparasites (49/59; 83%) [2].
Future studies should focus on the evaluation of possible risk factors, as well as the presence of competent arthropod vectors to understand biological cycles of the parasites, the role of the dog as a possible reservoir in zoonosis transmission, and the control of these VBP affecting dog populations and, potentially, the human population.

Conclusions
To our knowledge, this study constitutes the first molecular report of VBP in stray dogs from Kathmandu, Nepal. Our results highlight the importance of several CVBD that should not be underestimated in the metropolitan area of Kathmandu. The prevention and treatment of these CVBD must be taken under consideration and implemented in the veterinary control in concurrence with population awareness programmes, animal birth control or rabies vaccination programmes. Further on this, considering the high detected prevalence of VBP and following the WSAVA vaccination recommendations, a check-up including ecto-and endoparasite control should be performed prior to vaccination in apparently healthy dogs that could be potentially infected. Further studies are needed to understand the role of the dogs and arthropod vectors in the transmission of these and other VBP in the metropolitan area of Kathmandu.