Serological detection and risk factors for equine piroplasmosis in the semiarid region of Pernambuco, Northeastern Brazil.

Equine piroplasmosis, an economically important disease in horses, has so far not been reported in Pernambuco state, Brazil. This study aimed to evaluate the seroprevalence of anti-Babesia caballi and anti-Theileria equi antibodies based on the detection of these agents in equine blood and in ticks on horses in the municipality of Petrolina, Pernambuco, northeastern Brazil. Blood samples were drawn from 393 horses and sera were examined by ELISA. The presence of tick infestations was evaluated, and 101 ticks were subjected to DNA amplification for the detection of Babesia spp. by polymerase chain reaction (PCR). No parasites were detected in the blood smears. Anti-B. caballi and anti-T. equi antibodies were found in 27.2% (107/393) and 34.8% (137/393) horses, respectively. Infestation by Dermacentor nitens was detected in 4.3% (17/393) of the horses. There was no DNA amplification of the agents in ticks. The risk factors for the presence of anti-T. equi antibodies (P < 0.05) were: purebred (P < 0.001), animals older than 156 months (P = 0.014), and the presence of ticks (P = 0.001). No risk factors for B. caballi were identified. This study confirmed the circulation of agents of equine piroplasmosis in the municipality of Petrolina, state of Pernambuco, Brazil.


Introduction
Equine piroplasmosis is an important tick-borne disease caused by the intraerythrocytic protozoan parasites Theileria equi and/or Babesia caballi, which affect wild and domesticated equids (WISE et al., 2014). In Brazil, T. equi and B. caballi are transmitted mainly by Rhipicephalus microplus and Dermacentor nitens, respectively. Amblyomma cajennense has also been associated with the transmission of T. equi (KERBER et al., 2009). These agents can also be transmitted transplacentally (SANTOS et al., 2008), as well as iatrogenically through blood transfusions and the use of contaminated needles, nasogastric tubes and endoscopes (WISE et al., 2014).
The disease is considered endemic in tropical, subtropical and temperate regions (BRÜNING, 1996;KAPPMEYER et al., 2012;AHARONSON-RAZ et al., 2014). It is on the list of notifiable diseases, so infected animals are subject to international travel restrictions (OIE, 2018a), causing problems to the horse industry (VIEIRA et al., 2018). Clinical signs of the disease, which may range from asymptomatic to acute, are characterized by fever, jaundice, intravascular hemolysis based on free hemoglobin in the blood, anemia, hemoglobinuria, edema, bilirubinuria, azotemia, acute renal failure, depression, hepatomegaly, splenomegaly, and lymphadenopathy (DE WAAL, 1992;GUIMARÃES et al., 1998;MEHLHORN & SCHEIN 1998;WISE et al., 2013;ADAM et al., 2017). However, obvious signs of the disease are often absent common, making it difficult for regions to remain free of the agent. Animals infected with T. equi can manifest the disease after stress, strenuous exercise, immunosuppression and the administration of steroids (WISE et al., 2014).

Study area and sampling
This study was carried out from December 2011 to May 2012 in rural and urban areas in the municipality of Petrolina, situated in the semiarid region of the São Francisco Valley, which covers an area of 4,558.537 km 2 in the state of Pernambuco. For this study, 393 horses were randomly selected, regardless of sex, breed or age. Samples to be evaluated were determined considering an estimated prevalence of 50%, a confidence interval of 95% and an absolute precision of 7%.
Blood samples were drawn by jugular venipuncture into vacutainer tubes with and without anticoagulant. Tubes without anticoagulant were kept at room temperature until clot retraction occurred. After retraction, the blood samples were centrifuged (3,000g, 10 min) to obtain serum, which was then stored at -20°C. The ticks parasitizing the horses were removed with anatomical tweezers, placed in tubes containing 70% ethanol solution and identified as described by Aragão & Fonseca (1961

Blood smears
Immediately after drawing blood without anticoagulant, blood smears were prepared and stained using fast staining kits (Fast Panoptic, RENYLAB  ). The blood smears were examined under a light microscope equipped with a 10x eyepiece and an immersion objective lens (1000x) for hemoprotozoan examination.

Enzyme-linked Immunosorbent Assay (ELISA)
Serum samples were diluted by a factor of 1:100 and subjected to ELISA for the detection of anti-B. caballi and anti-T. equi antibodies, as described by Baldani et al. (2004). This test involved the use of purified antigens of B. caballi and T. equi, alkaline phosphatase conjugated anti-horse IgG (Sigma Chemical Co., code A6063) diluted by a factor of 1:15000 in PBS-Tween-80 with 5% skim milk powder and 10 mg of pNPP (p-nitrophenyl phosphate) substrate. Optimal reactivity levels of B. caballi and T. equi antigens were 2.5 and 10 μg / mL, respectively. The reaction was detected in a microplate reader (Microplate Reader MRX TC Plus, Dynex Technology, USA) at a wavelength of 405 nm.
The immunological activity of each tested serum was calculated by determining the p-value (sample versus positive). In addition, the optical densities (OD) of sera were pooled at ELISA (NE) levels. The ELISA cut-off point was 2 1/2 times the mean OD of negative reference sera (BALDANI et al., 2004;GOLYNSKI et al., 2008;MACHADO et al., 2012).

Polymerase Chain Reaction (PCR)
Ticks were collected from 17 parasitized animals (one to five ticks per animal) and subjected to PCR. The ticks were processed individually, washed with 10 mM Tris HCl; 1mM EDTA, pH 8.0 (TE), as described by Horta et al. (2007), and subjected to DNA extraction using a commercial Wizard  Genomic DNA Purification kit (Promega, Madison, USA), following the manufacturer's instructions, to a final volume of 50 μL. Five microliters of extracted DNA were used for PCR amplification. DNase-free water was used as a negative control for DNA extractions and PCR assays. PCR amplification primers of 500 base pairs of the Babesia spp. 18S rRNA gene fragment were used, as proposed by Santos et al. (2017): BAB 143-167 [5'-CCG TGC TAA TTG TAG GGC TAA TAC A-3'] e BAB 694-667 [5'-GCT TGA AAC ACT CTA RTT TCT CAA AG-3']. PCR was carried out in a total of 50 μL aqueous solution containing 1x PCR buffer, 1.5 mM MgCl2, 0.2 mM dNTPs, 1 U of Platinum Taq DNA Polymerase (Invitrogen, Carlsbad, CA), and 0.2 mM of each primer. The amplification procedure, which was carried out in a thermocycler (Biocycler  ), involved the following steps: initial denaturation at 95°C for 5 min, followed by 35 cycles at 95°C for 30 sec, 58°C for 30 sec and 72°C for 30 sec, with a final extension at 72°C for 7 min. The amplification products were subjected to 1.5% agarose gel electrophoresis, stained with ethidium bromide, and examined under an ultraviolet light transducer.

Statistical analysis
The animals' owners answered an epidemiological questionnaire aimed at identifying possible risk factors associated with the presence of anti-B. caballi and anti-T. equi antibodies. The variables analyzed were age, breed, sex, presence of irrigated area, presence of ticks, degree of infestation, mucosal evaluation, use of tick repellents and uses of the animals.
The risk factor analysis was conducted in two steps: in univariate analysis, each independent variable was crossed with a dependent variable, and the ones showing a p value of ≤ 0.20 by the chi-square test or Fisher's exact test were selected for multivariate analysis, using multiple logistic regression. The level of significance adopted in the multiple analysis was 5%. All the analyses were performed using SPSS 20.0 for Windows software.
The presence of anti-B. caballi antibodies was found in 27.2% of the horses, anti-T. equi antibodies in 34.9% and both agents in 10.7%. In the univariate analysis, the variables that showed a significant association (P < 0.20) with the presence of anti-T. equi antibodies were purebred horses, animals older than 156 months, the presence of irrigated areas on farms, the presence of ticks, the use of tick repellent, and uses of the animals (Table 1). However, when these independent variables were subjected to multivariate analysis, the breed -category of purebred (OD = 8.91, P < 0.001), age -senile animals (OD = 3.54, P = 0.014) and presence of ticks (OD = 7.13, P = 0.001) were considered risk factors for T. equi (Table 2). In the univariate analysis of antibodies of both agents found in the same animal, breed and the presence of irrigated areas on farms showed a significant association (P < 0.20) (Table 3). In the multivariate analysis, the breed -category of purebred (OD = 2.61, P = 0.012) was considered a risk factor.
No risk factors associated with the presence of anti-B. caballi antibodies were identified, and no intra-erythrocyte form of T. equi or B. caballi was observed in the blood smears examined under the microscope.
The OD values of cutoff points, negative control sera and positive control sera tested for B. caballi and T. equi were 0.351 and 0.269, 0.122-0.145 and 0.096-0.162, and 0.948-1.175 and 0.891-1.531, respectively. The average OD values of samples reactive to the presence of anti-B. caballi and anti-T. equi antibodies were 0.501 and 0.615, respectively.
One to five adult ticks were collected from each of 17 horses that were parasitized at the time of evaluation, making a total of 101 ticks, all identified as Dermacentor nitens. This tick species transmits only Babesia, so the ticks were subjected only to PCR to test for this agent. However, DNA amplification of the agents in the ticks subjected to PCR did not occur.

Discussion
The main tests used for the diagnosis of long-term equine piroplasmosis and of animals treated with antiparasitic drugs are serological tests such the indirect immunofluorescence test (IFA) and enzyme-linked immunosorbent assay (ELISA) (OIE, 2018b). A significant degree of agreement between the tests was found in a comparative study by Vieira et al. (2015). Other serological studies using these two tests also showed almost perfect agreement between them (GOLYNSKI et al., 2008;SALVAGNI et al., 2010;FARKAS et al., 2013;MAHMOUD et al., 2016). However, IFA tested an increased number of 19 horses with anti-T. equi antibodies titer in comparison with ELISA which tested only 10/19 horses with anti-T. equi antibodies titer (PIKALO et al., 2016).  Although equine piroplasmosis is widely distributed in Brazil, there is no report of its prevalence in the state of Pernambuco. Hence, this is the first study aimed at ascertaining the seroprevalence of T. equi and B. caballi in the city of Petrolina, Pernambuco, in northeastern Brazil. The ELISA assay performed in this study revealed that, among the seroreactive horses, 34.9% tested positive for T. equi and 27.2% for B. caballi. These findings are consistent with those of two other studies conducted recently in northeastern Brazil, using ELISA, the first of which reported finding 43.5% of animals positive for T. equi and 7.7% for B. caballi, while the second study found 38.1% positive for T. equi and 18.6% for B. caballi (BRAGA et al., 2017;NOGUEIRA et al., 2017). In Paraíba, also in the country's northeastern region, a study using IFA found that 59.6% of horses were seroreactive to T. equi (FERREIRA et al., 2016). In other studies conducted in Rio Grande do Sul and the Brazilian Pantanal, 18.9% and 5.5%, and 61.8% and 52.9%, were observed seropositive horses for T. equi and B. caballi, respectively, using ELISA (VIEIRA et al., 2018;CAMPOS et al., 2019). However, the agent B. caballi was detected more frequently in the states of São Paulo and Mato Grosso do Sul KERBER et al., 2009). The different prevalence rates of equine piroplasmosis in the abovementioned regions may be attributed to variations in climate, equine populations studied, diagnostic tests used, and to the dynamics of tick populations (MUJICA et al., 2011;VIEIRA et al., 2013).
The test most widely used for the diagnosis of hemoparasitic diseases used to be the blood smear test, but because its low sensitivity often leads to false negatives in horses with low parasite loads, it has been replaced by indirect and direct techniques for the detection of this parasite (GOLYNSKI et al., 2008;SGORBINI et al., 2015). The results of this study are in agreement with the findings of Munkhjargal et al. (2013), who suggest that the blood smear diagnostic technique may not be suitable for epidemiological surveys of equine piroplasmosis. Moreover, Hodgson (2002) points out that this technique presents low sensitivity in cases of low parasite loads, which leads to false negative results.
The low tick infestation rate in horses on farms in the municipality of Petrolina differs from that reported in Rio de Janeiro by Santos et al. (2011), who found a moderate to severe infestation rate, i.e., 17.09% (122/702) by D. nitens. Climate factors such as temperature, relative air humidity, altitude and high rainfall can influence the habitat of the main tick species/vectors and define areas of enzootic instability, areas that do not favor the development of ticks throughout the year (GOLYNSKI et al., 2008). Therefore, the low infestation rate found in this study may be associated with the edaphoclimatic factors of the municipality, i.e., its location in the semi-arid region of Pernambuco, whose dry climate and low rainfall are unfavorable for the development of ticks. This suggests that the municipality is located in an area where the presence of ectoparasites in equines is low, and that such ectoparasites are found only at certain times of the year or in places with higher humidity.
PCR has been used to identify and associate some tick species with agents of equine piroplasmosis (BATTSETSEG et al., , 2002IORI et al., 2010;ABEDI et al., 2014). In this study, ticks were tested only for DNA detection of Babesia spp, since only D. nitens were found in this region. However, the presence of DNA of this agent in ticks was not detected. In contrast, Nogueira et al. (2017) collected 170 ticks from 97 horses, 151 of which were identified as D. nitens, 13 as A. cajennense sensu lato and 6 as Rhipicephalus microplus. Among these ticks, 2.35% (4/170) tested positive by PCR for T. equi and 0.59% (1/170) for B. caballi.
The age of horses older than 156 months was considered a risk factor for the presence of anti-T. equi antibodies. This has also been reported by Ruegg et al. (2007), Sevinc et al. (2008) and Kouam et al. (2010), supporting the assumption that T. equi infection persists throughout life (BRÜNING, 1996). Nogueira et al. (2017) also state that the risk of T. equi infection increases as a function of age, the presence of ticks, and vaccination against other diseases.
Several of the horses evaluated here participate in sports events, which means they travel to several states and different regions. Thus, they are in contact with other animals and possibly also with the vector of T. equi, which may explain the presence of anti-T. equi antibodies in these animals. Transmission may also occur iatrogenically, i.e., through blood transfusions or contaminated needles (WISE et al., 2014). Kerber et al. (2009) observed that high infestations of horse flies were reported on all farms during the summer and spring months, which they would describe as a type of mechanical transmission. Cattle were present on some farms, but R. microplus infestations were not detected.
These factors may also be associated with purebred animals as a risk factor for T. equi. Purebred horses are more commonly used for sports and recreation activities. Some studies have been carried out solely with purebred or crossbred horses or with a specific breed predefined according to the region, since these animals are of greater economic importance in the horse market (GOLYNSKI et al., 2008;KERBER et al., 2009;GUIMARÃES et al., 2016;NUGRAHA et al., 2018). However, more specific studies are needed involving purebred and mixed breed horses in order to examine this host parasite relationship.
The identification of areas that offer a potential risk of disease transmission to horses is extremely important because it enables the implementation of adequate prophylactic measures, such as the control of vector ticks in animals and the environment, as well as the management of these animals, allied to educational actions that are essential for the prevention and control of diseases in horses.
This study confirmed, for the first time, that agents of equine piroplasmosis are circulating in the municipality of Petrolina, semi-arid region of Pernambuco, based on the indirect detection of the infection by anti-B. caballi and anti-T. equi antibodies. The municipality of Petrolina comprises an area of enzootic instability, given the low frequency of antibodies detected in its equine population, associated with the low frequency of tick infested animals. Nevertheless, additional studies in the region are needed, including studies involving purebred and mixed breed horses, in order to gain further insight into the relationship between these animals and T. equi.