Can Equids Be a Reservoir of Leishmania braziliensis in Endemic Areas?

In this study, we detected Leishmania (Viannia) braziliensis infection in equids living in endemic regions of cutaneous leishmaniasis. To determine the role of these animals in the Leishmania cycle, we used two approaches: serological and molecular methods. Antibodies to the parasite were assayed using the Enzyme Linked Immunosorbent Assay (ELISA). Blood samples were collected and tested by polymerase chain reaction (PCR), and the positive products were sequenced. The results showed that 11.0% (25/227) of the equids were seropositive for Leishmania sp, and 16.3% (37/227) were PCR positive. Antibodies were detected in 20 horses, 3 donkeys, and 2 mules, and the parasite DNA was detected in 30 horses, 5 donkeys, and 2 mules. Sequencing the amplified DNA revealed 100% similarity with sequences for Viannia complex, corroborating the results of PCR for L. braziliensis. Our results show that equids are infected with L. braziliensis, which could be food sources for phlebotomines in the peridomiciliary environment and consequently play a role in the cutaneous leishmaniasis cycle.

The parasite can infect humans in the tropics and subtropics worldwide. The life cycle of Leishmania basically involves wild animals and phlebotomines. Human outbreaks have been detected only in extractive activities, deforestations, construction of highways, petroleum exploration, and during scientific exploration [5,6,7,8].
During the last 60 years, in some regions of Brazil, particularly in the southeast and south, there have been extensive changes in the environment and increases in agricultural and pastoral activities. This resulted in the virtual disappearance of cutaneous leishmaniasis (CL) from the end of the 1950s to the 1970s. However, in the 1980s, an increased incidence of CL was observed not only in endemic areas but in new areas, and outbreaks were also reported in regions where leishmaniasis was considered eradicated [9,10,11,12,13,14].
Cutaneous leishmaniasis is strongly associated with environmental factors, and its anthropogenic modifications would be considered as modifying actions of risk factors for transmission of the infection. To identify the key and more easily modified factors that prevent/promote the transmission of CL, reservoirs should be considered as essential factors in any investigation to establish the risk factors for transmission of the infection. The reservoir has been well established for numerous Leishmania species. However, even though L. braziliensis is found throughout the neotropics, its epidemiology is poorly understood. This is because when an infected animal is diagnosed, the strain generally is not isolated and identified, so questions remain regarding the primary reservoir and other potential reservoirs of L. braziliensis [5,14,15]. Domestic animals, such as dogs and equines, probably play a role in the transmission of L. braziliensis since they are closely related with humans and are also present in agricultural and pastoral areas that are generated by deforestation.
Infection of L. braziliensis in a horse was first reported in 1927 in an animal from Argentina and in 1959 in a donkey from Brazil [16]. Currently, Leishmania infection in horses in the New World has been reported in Brazil, Venezuela, Puerto Rico, and the United States [9,10,13,15,17,18].
In the New World, the leishmaniases may be mutating from exoanthropic zoonoses to anthroponoses [6], or euzoonoses, with domestic animals playing a fundamental role in the maintenance of the parasite. To answer this question, our group investigated L. braziliensis infection in equids from an endemic rural area of southern Brazil using serological and molecular methods as tools in order to determine if the infection is widely distributed in this animal species in endemic areas.

Ethics statement
All experiments conducted in the present study comply with current Brazilian laws. All equids owners provided consent to have their animals sampled. None of the animals sampled had clinical signs of leishmaniasis, and all of them had good general health. The equids were submitted only for sample collection, and none of them were subjected to any treatment or invasive procedure. The present study and its animal protocol were approved by CNPq (National Council for Scientific and Technological Development), process number: 301906/2008-4.

Study area
The equids used in the present study were from the counties of Jaboti (23u449340S, 50u049330 W), Japira (23u489460S, 50u089200 W), Pinhalão (23u479340S, 50u039210 W) and Tomazina (23u469400 S, 49u579000 W). These counties are inserted in the mid-region of northern Paraná State, the Second Paraná Plateau, and the Eastern Portal of the Paraná Basin, with altitudes from 550 to 650 m, between 22u299230 and 26u429590 south latitude and between 48u029240 and 54u379380 west longitude ( Figure 1).
The study area is defined climatically as Cfa, where C means average of the coldest month is less than 18uC; f means there is no dry season; and a indicates a hot summer. The annual rainfall is 2000 mm, distributed throughout the year, with 120 days of rain concentrated mainly in the months of November to January. Since our objective was to analyze the role of equids as reservoirs for L. braziliensis, studies were conducted in the four counties mentioned because of the incidence of human cases of cutaneous leishmaniasis occurring there, a reliable indicator for the presence of this parasite.

Reservoir survey and blood collection
Blood samples were collected from the jugular vein (10 mL) of 227 equids (181 horses, 27 donkeys, and 19 mules) from the counties mentioned above. From these samples, 5 mL were collected in BD Vacutainer tubes to collect the sera for serological assay, and 5 mL were collected in BD Vacutainer spraycoated K 2 EDTA tubes for DNA extraction of whole blood.

Enzyme-linked immunosorbent assay (ELISA)
The prevalence of anti-Leishmania sp. antibodies was evaluated using the enzyme-linked immunosorbent assay antibody (ELISA) test. Sera were submitted to two ELISA steps; in the first (ELISA I), Leishmania infantum lysate was used as antigen (1 mg/well), while in ELISA II, the protein rk 39 was used as antigen (25 ng/well). Antigens were purchased from CORIXA CORP, Seattle, WA, USA. The conjugate was anti-horse antibody (IgG-fraction) from sheep, A-6917 (SIGMA TM ), diluted to 1:10,000 in PBS buffer (0.01 M, pH 7.2), 1% of Tween 20, and 0.1% of BSA. All samples were analyzed at a 1:100 dilution in PBS, pH 7.2. The substrate used was TMB Microwell Peroxidase Substrate (Kirkegaard & Perry Laboratories). The cutoff value was obtained from the mean absorbance of 30 negative horse serum samples plus 3 standard deviations. The negative animals came from Curitiba, not an endemic area for L. braziliensis. The ELISA test was performed as proposed by the manufacturer, and the absorbance was under the cutoff value.

DNA extraction and polymerase chain reaction (PCR)
Blood samples were submitted for DNA extraction using the phenol/chloroform protocol, according to [19], with the following modifications. Samples were centrifuged for 15 min at 3500 g, the supernatant was discarded, and the sediment was washed with PBS. Next, lysozyme, 10% SDS, and RNAse were added, and the mixture was homogenized and incubated at 37 uC for 2 h. After this, proteinase K was added, homogenized, and incubated at 55 uC for 2 h. The subsequent extraction was performed by the phenol/chloroform-based protocol [20]. The final extract was airdried and resuspended in TE buffer (Tris HCl 10 mM, EDTA 1 mM, pH 8.0). This solution was stored at -80 uC.
Subsequently, blood DNA samples were assayed for the presence of Leishmania sp. by a single PCR program using the primer pair MP1L and MP3H (5 mM). Amplification reactions were performed in a total volume of 25 mL, with the primary mixture containing 2.5 mL of 106 DNA polymerase buffer (100 mM Tris-HCl, pH8.3, and 500 mM KCl); 1.5 mM MgCl 2 ; 0.8 mM of each deoxyribonucleotide triphosphate (dNTPs); 0.1% of Triton-100; 0.5 mL of each primer; 0.2 mL of Taq DNA polymerase (InvitrogenTM); and 2 mL of template genomic DNA. Reaction cycles comprised an initial denaturing step at 94 uC for 2 min, followed by 30 cycles of amplification, each cycle consisting of: 94 uC for 1 min, 58 uC for 1 min, and 72 uC for 1 min, with a final extension step at 72 uC for 10 min. Negative and positive controls were included with all PCR reactions, performed in a Hybaid GeneAmp thermocycler with a heated lid. The amplification products were electrophoresed in a 1.6% (w/v) agarose gel, stained with ethidium bromide, illuminated under UV, and photographed with a Life Technologies Gibco BRL UV transilluminator documentation system.

Sequencing
Five amplified PCR products from each municipality were submitted to sequencing using the Dyenamic ET to MegaBACE (GE Healthcare Life Sciences of Brazil) to confirm the identity of L. braziliensis. The sequencing reactions were performed using 100 ng of template amplicons, 3 mL of Dyenamic ET, 1.5 mL of the primer (10 pmol/mL), and 1.5 mL of the appropriated sequencing buffer. The sequencing program comprised 25 cycles of 95 uC for 20 s, 50 uC for 15 s, and 60uC for 1 min, following which the products were precipitated and purified as described by [21]. Consensus sequences were produced using the EMBOSS GUI package software (http://bips.u-strasbg.fr/EMBOSS), the sequenced fragments were aligned with available data deposited in GenBank, and comparisons were performed using BLAST programs available on NCBI to confirm the results (http://blast. ncbi.nlm.nih.gov/Blast.cgi).

Data analysis
The data were statistically analyzed using the Fisher exact test to determine differences in prevalence between sex groups and equine species (horses x donkeys and mules). A value of P#0.05 was considered statistically significant. To compare groups (sex and species), we considered the positive results for both PCR and ELISA assays.

ELISA and PCR
Analyzing ELISA and PCR, past or current infection with L. braziliensis was detected in 10.1% (23/227) of the animals studied considering both ELISA and PCR results and in 17.2% (39/227) of the equids positive for either ELISA or PCR (Table 1). Regarding sex, considering the results for both tests, we verified that 9.1% (12/132) of male equids and 11.6% (11/95) of female equids were positive for L. braziliensis. Analyzing the PCR and ELISA data, we determined that there were no significant differences in the infection rates between species (P = 0.52) and sexes (P = 0.37).

Discussion
This study proved that equids (horses, donkeys, and mules) from southern Brazil are infected with L. braziliensis, since antibodies were detected in sera and parasite DNA were found in the blood samples. The sequencing confirmed that the parasite was Viannia complex, thus corroborating the presence of L. braziliensis. Leishmania DNA in the blood was observed in 16.3% of the animals, and 11.0% presented positivity by ELISA. A positive ELISA with negative PCR is obtained in cases when the equid has only been exposed to the parasite in the past or when the parasite is in the chronic phase of infection and not present in the blood. In addition, a positive PCR with negative serology is obtained in cases of acute infection, since we only analyzed IgG antibodies, and when the animal has some immune deficiency disease.
The presence of anti L. braziliensis antibodies is evidence that the immune system had prior contact with the protozoan, while the presence of specific DNA confirms blood circulation of the parasite. We detected L. braziliensis DNA in 23 of the 25 equids with positive sera, while 14 PCR-positive animals had no antibodies against the parasite. These data could be explained by the possibility that we detected L. braziliensis in the acute phase of infection. These results are evidence of the presence of the parasite and of the peridomiciliary transmission of this parasite in the study area (Figure 1).
Since the first report in Argentina, studies have developed the possibility of horses, donkeys, or mules being accidental hosts or reservoirs or even having a role in the cycle of Leishmania sp and CL in the peridomiciliary environment ( Table 2). The main epidemiological studies addressing the occurrence of CL in horses have been conducted in South America [3,10,15,18,22,23]. Because of the presence of horses infected with Leishmania in endemic areas for leishmaniasis in humans, these authors discussed the chance that horses could take a part in the maintenance cycle of the parasite in this transitional environment between wild and domestic habitats. In studies conducted by [10,18,22,23], the parasite was isolated and identified as Leishmania braziliensis. Most studies of leishmaniasis in horses have been conducted in South America, particularly in Brazil [3,10,13,15,17,18,25,28]. There are case reports of CL in two horses in Central America [24] without Leishmania species identification and an infected horse in North America [27], in which the identified species was L. siamensis. Case reports have also been published in Europe; in Germany [26] and in Spain [27] L. infantum was identified as the causative agent for cutaneous lesions in horses. In recent epidemiological studies conducted in Brazil, [13,25] investigated the presence of Leishmania in 132 horses. In both studies, none of these animals presented cutaneous lesions. In a serological survey performed in an area endemic for CL in humans, [25] found 22.1% of horses carried antibodies against Leishmania sp. In the study conducted by [13], of the 55 horses analyzed, 76.3% were seropositive for Leishmania sp, and 7.1% were PCR positive in blood samples. The parasite was not isolated or sequenced in either study [13,25]. In our study, 227 equids were analyzed. It is therefore the largest epidemiological survey regarding leishmaniasis and horses in a transmission environment for CL. All horses  23 the parasite was present in the blood circulation, as their blood samples were PCR positive. This finding may suggest the possibility of these infected equids not only to serve as a food source for sand flies but also to maintain the life cycle of L. braziliensis in peridomiciliary areas. This result is unprecedented in identifying blood circulation of L. braziliensis (PCR and sequencing) in seropositive horses (IgG ELISA). In this way, horses could act as infection sources for sand flies and contribute to this epidemiological context in that CL is characterized by synanthropic zoonoses. In order to prove that these animals can act as domestic reservoirs of L. braziliensis in the peridomiciliary environment is necessary to conduct infectivity tests on sand flies and perform the isolation of the parasite from samples of peripheral blood.
In rural areas of Brazil, equids are commonly used as means of transport, as pack animals in deforestation areas, and in agropastoral activities, therefore they are constantly present in endemic areas and can become a good blood source for phlebotomines [3]. This change in the habits of sand flies results from the human-induced environmental change in order to increase agricultural and pastoral areas. As a result of deforestation and human migration to these new agropastoral regions, where the population has built precarious housing on the edges of preserved woods and secondary woods, phlebotomines have adapted to the peridomiciliary environments. This altered behavior in relation to food sources might be a future adaptation of sand flies, since studies show that Lutzomyia (N.) intermedia feeds on people, rodents, birds, horses, and dogs [30]. Since this feeding habit is directly related to human activities, phlebotomines may have adapted to new ecotopes in the peridomiciliary environment, resulting in new epidemiological patterns for CL, thus creating a favorable eco-epidemiological scenario [6,20].
Studies performed in the State of Paraná have revealed high numbers of phlebotomines collected in houses and shelters near deforestation areas and in secondary woods in CL endemic regions, proving the tendency of peridomiciliary behavior of sand flies in these locations [31,32]. Our concern regarding this epidemiological context is to identify the reservoir(s) that could maintain the life cycle of L. braziliensis in the peridomiciliary environment, since in any epidemiological research, knowledge regarding the reservoirs, phlebotomines, and leishmanial parasites in any disease foci is essential for correct diagnosis and treatment and precise control strategies [14,33]. Regarding this scenario, where sand flies appear to be adapting to new ecotopes and different food sources, both close to anthropic environments, consequent pressure is being exerted on the Leishmania species, particularly on L. braziliensis. This parasite species presents a high degree of genetic variability and thus could infect a larger number of vectors and mammals, permitting the appearance of new transmission areas with different species of vectors and reservoirs [34,35,36,37,38]. Studies have shown genetic polymorphism in natural populations of different species of Leishmania sp. [34,35], which could explain the plasticity of these parasites and their ability to adapt to new ecological conditions [6].
In the New World, the identification of the primary reservoirs (wild animals x domestic animals) in the maintenance of the L. braziliensis cycle is still under investigation. It is known that CL is a zoonoses that was originally restricted to the sylvatic cycle and that human-induced environmental changes in rural and urban areas brought the parasite into the domestic cycle, increasing the incidence of human and domestic animals (dogs and horses) infected with L. braziliensis [11]. A high prevalence of domestic animals infected with L. braziliensis is usually related to epidemic outbreaks, when the same infection pressure suffered by humans is observed among dogs or horses, which show similar or lower seroprevalence. A study conducted in Paraná state, [14], discussed the role of dogs and equids in the transmission of L. braziliensis based on serology, anatomopathological tests and in vitro culture. It was not possible to isolate the parasite from intact skin, blood, lymph nodes, or internal organs of any of the equids, later diagnosed as Habronema sp. The Koch postulates are partially fulfilled and have shown that equids (horses, donkeys, and mules) are infected with L. braziliensis. Therefore, these animals can be sources of food for the sand flies outside the home, allowing maintenance of the peridomestic cycle of the parasite and thus contributing to the epidemiology of human cutaneous leishmaniasis.