Introduction

Tick-borne diseases have been increasingly studied in Brazil, but there are still many unexplored places, especially in the Amazon region. Canine babesiosis is a tick-borne disease of domestic and wild canids characterized by fever, depression, and anaemia (^{KUTTLER, 1988}). Previous parasitological and serological studies carried out in Brazil have shown that canine babesiosis due to Babesia canis is distributed among different states with rates of seropositivity ranging from 1.9 to 66.9% in Minas Gerais (^{RIBEIRO et al., 1990}; ^{RODRIGUES et al., 2002}; ^{BASTOS et al., 2004}; ^{SOARES et al., 2006}), 35.7% in Paraná (^{TRAPP et al., 2006}), 5.2% in Rio de Janeiro (^{O'DWYER et al., 2001}), and 10.3% in São Paulo (^{DELL'PORTO et al., 1993}). In addition, the disease was also reported in the state of Mato Grosso, where it was molecularly confirmed asB. canis vogeli (^{SPOLIDORIO et al., 2011}).

Canine monocytic ehrlichiosis, caused by Ehrlichia canis, is the most important tick-borne disease of dogs in Brazil. Currently,Ehrlichia canis is the only Ehrlichia species that has been isolated in cell culture from vertebrates in South America. A preliminary investigation for Ehrlichia species in the northern and southeastern regions of Brazil failed to detect Ehrlichia DNA inAmblyomma ticks, humans, dogs, capybaras, and febrile human blood samples (^{LABRUNA et al., 2007a}). In contrast, ehrlichial DNA compatible with Ehrlichia chaffeensis,Ehrlichia ewingii, or an agent closely related toEhrlichia ruminantium were recently reported in animal blood samples from southeastern Brazil (^{MACHADO et al., 2006}; ^{OLIVEIRA et al., 2009}; ^{WIDMER et al., 2011}).

Most of the published studies on tick-borne diseases in the Brazilian Amazon region have focused on rickettsiosis, mainly in western Amazon, state of Rondônia (^{LABRUNA et al., 2004}, ^2007b). In the aforementioned region, someRickettsia species were described for the first time in Brazil. At the same time, there is no information on rickettsioses from the eastern part of the Amazon.

In this study, we evaluated seroprevalence to Babesia canis vogeli, Ehrlichia canis, andRickettsia spp. in dogs from rural and urban areas within the state of Pará, eastern Amazon, Brazil.

Materials and Methods

During 2008-2009, a total of 129 dogs of different breeds and ages were sampled from an urban area and from different farms in rural areas. Those samples were also used for a serological study to investigate the prevalence ofNeospora caninum, Toxoplasma gondii, andLeishmania infantum (formerly chagasi), as previously described (^{VALADAS et al., 2010}). Of all the dogs, 77 (59.7%) were males and 52 (40.3%) were females; 57 samples (44.2%) were collected from urban stray dogs from the municipality of Santarém, and 72 (55.8%) were from dogs from 39 rural properties in 20 different municipalities.Figure 1 shows the municipalities where the respective numbers of dogs were sampled. The rural properties were selected from a prevalence study of other parasitic and infection agents in cattle (^{MINERVINO et al., 2008}; ^{CHIEBAO, 2010}). Blood samples were collected from the jugular or brachial vein of the dogs, and sera was obtained by centrifugation. Samples were stored at −20 °C until tested.

Figure 1. Map of Pará state with the municipalities where canine serum samples were collected (number of dogs). 1 Água Azul do Norte (one); 2 Altamira (six); 3 Belterra (one); 4 Canaã dos Carajás (one); 5Conceição do Araguaia (11); 6 Goianésia do Pará (two);7 Itupiranga (six); 8 Jacundá (eight); 9 Marabá (four); 10 Novo Repartimento (two); 11 Ourilândia do Norte (two);12 Pau D'Arco (two); 13 Redenção (four); 14 Santa Maria das Barreiras (two);15 Santana do Araguaia (six); 16Santarém (three rural, 57 urban); 17 São Domingos do Araguaia (two); 18 São Felix do Xingu (two);19 São João do Araguaia (1); 20Tucumã (six). 

Serum samples were submitted to indirect immunofluorescence assay (IFA) with antigens of B. canis vogeli (blood smears from splenectomized dogs that were experimentally infected in our lab) according to ^{Bicalho et al. (2004)}, using a screening dilution of 1:64. To detect antibodies against E. canis, the bacteria were cultivated in DH82 cells, as described by ^{Aguiar et al. (2007a)}, and serum samples were analyzed following the protocol by ^{Silva et al. (2010)}, but with a screening dilution of 1:80. ForRickettsia spp., IFA was run using the screening dilution of 1:64 against six Rickettsia species that occur in Brazil, namelyR. rickettsii, R. parkeri, R. amblyommii, R. rhipicephali, R. bellii, and R. felis, which were cultivated in Vero or C6/36 cells (^{LABRUNA et al., 2007b}; ^{HORTA et al., 2004}). Samples with IFA reaction at the cut-off point for each agent were considered positive and further tested in two-fold serial dilution to determine endpoint titers. Serum of aRickettsia species showing titer at least 4-fold higher than those observed for the other Rickettsia species was considered homologous to the first Rickettsia species or to a very closely related species (^{HORTA et al., 2004}, ²⁰¹⁰; ^{LABRUNA et al., 2007b}; ^{PIRANDA et al., 2008}; ^{SAITO et al., 2008}).

Possible statistical associations between gender or location (rural or urban) of dogs and the occurrence of anti- B. canis vogeli,E. canis, or any of the six Rickettsia species antibodies were analyzed by Pearson's chi-square test using Minitab statistical software (Minitab 2000). The significance adopted was 5%.

Results

The distance between the visited farms varied from 10 to 1,000 km. From the 20 municipalities visited, only three (Canaã dos Carajás, Ourilândia do Norte, and São João do Araguaia) presented negative results to all tested samples. From the 129 samples tested for B. canis vogeli, 55 (42.6%) were positive, being 34 (59.6%) from urban dogs and 21 (29.1%) from rural dogs. Antibodies againstB. canis vogeli were detected in dogs from 10 different municipalities (Conceição do Araguaia, Itupiranga, Jacundá, Marabá, Redenção, Santa Maria das Barreiras, Santana do Araguaia, Santarém, São Felix do Xingu, and Tucumã).

From the tested samples (129), 21 (16.2%) were positive to E. canis, being 9 (15.7%) from urban dogs and 12 (16.6%) from rural dogs from eight different municipalities (Água Azul do Norte, Altamira, Belterra, Conceição do Araguaia, Jacundá, Pau D'Arco, Redenção, and Santana do Araguaia). The highest anti-E. canis endpoint titer found in dogs from either rural or urban areas was 81,720.

Sera from 41 (31.7%) out of 129 canine samples were positive to at least one Rickettsia species, being 29 (24.8%) rural dogs from 13 different municipalities (Conceição do Araguaia, Itupiranga, Jacundá, Marabá, Novo Repartimento, Pau D'Arco, Redenção, Santa Maria das Barreiras, Santana do Araguaia, Santarém, São Felix do Xingu, São Domingos do Araguaia, and Tucumã) and 12 (21%) urban stray dogs from Santarém municipality. From the sixRickettsia species tested, only four were considered to be possibly present at the studied regions (R. bellii in at least eight dogs, R. rhipicephali in two dogs, R. rickettsii in one dog, and R. amblyommii in eight dogs). Canine endpoint titers to Rickettsia spp. antigens are presented in Table 1. A total of 30, 20, 15, 12, 11, and nine dogs were reactive to R. amblyommii, R. rhipicephali, R. parkeri, R. bellii,R. rickettsii, and R. felis, respectively. Endpoint titers varied from 64 to 8,192 for R. amblyommii. At least half of the 30 R. amblyommii-seroreactive dogs had endpoint titers ≥1024 for R. amblyommii. Endpoint titers for the otherRickettsia species ranged as follows: R. rhipicephali, 64-8,192; R. bellii, 128-8,192;R. parkeri, 128-2,148; R. rickettsii, 64-2,148; and R. felis, 64-512. While the median of the individual titers against both R. amblyommii and R. rhipicephali was 1,024, the median for the titers against R. bellii, R. parkeri, R. rickettsii, and R. felis were 512, 256, 128, and 256, respectively.

Regarding B. canis vogeli, female dogs presented significantly higher prevalence (53.8%) than male dogs (35.1%) (P = 0.046). In addition, B. canis vogeli seropositivity was significantly higher (P < 0.01) among urban (59.7%) than rural (29.2%) dogs. For E. canis, there was no association between the frequency of positive animals and the independent variables evaluated (P > 0.05). For Rickettsia spp., no association (P >0.05) was found relating to gender, but rural dogs presented significantly higher prevalence (40.3%) than urban animals (21.1%) (P = 0.02).

Antibodies for the three genera of tick-borne pathogens were not found simultaneously in any of the dogs. Antibodies for at least two genera were found in 16 (22.2%) dogs from the rural area, with highest association between E. canis and Rickettsia spp. (56.2%). In the urban area, there were 17 (29.8%) animals with positive results to at least two genera, whereE. canis and Rickettsia spp. (58.8%) were again the most prevalent. However, no statistically significant correlation was found in positivity for the different pathogens.

Discussion and Conclusions

To the best of our knowledge, Babesia species in dogs from the Amazon region has never been reported. Due to this lack of information about canine babesiosis in the Brazilian Amazon region, the occurrence of positive dogs to B. canis vogeli (42.6%) can only be compared to other Brazilian studies, in which the prevalence of antibodies against this parasite in sera from dogs varied from 35.7% in Paraná state (^{TRAPP et al., 2006}) to 66.9% in Minas Gerais state (^{RIBEIRO et al., 1990}). We can infer from our results that the higher seroprevalence to B. canis vogeli observed among urban dogs is probably due to the predominating tick species recently found in the urban area of the municipality of Santarém, namely, R. sanguineus (^{SERRA-FREIRE, 2010}), which is the only known vector of B. canis vogeli (^{DANTAS-TORRES, 2008}).

Ehrlichia was first reported in Brazil in 1973, in dogs from Minas Gerais state, southeastern Brazil (^{COSTA et al., 1973}). From the 26 Brazilian states, only one remains with no report onEhrlichia species, as recently reviewed or reported (^{SPOLIDORIO et al., 2010}; ^{AZEVEDO et al., 2011}; ^{VIEIRA et al., 2011}). Considering the Legal Amazon region, the only reports onEhrlichia species are from Rondônia state (western Amazon), where 36% of dogs were seroreactive to E. canis by IFA (^{AGUIAR et al., 2007b}; ^{VIEIRA et al., 2011}). Also in Rondônia state, ^{Labruna et al. (2007a)} found four out of five dogs infected with E. canis by molecular methods, which was the first molecular report of E. canis in the Amazon region. We have sampled a total of 129 domestic dogs from 20 different municipalities in the state of Pará and found 21 (16.2%) positive results to E. canis by IFA. Our results showed lower occurrence when compared to previous studies in dogs from western Amazon.

Overall, serum endpoint titers to Rickettsia spp. indicate that seropositive rural dogs had been predominately infected by R. amblyommii or by a closely related genotype, whereas seropositive urban dogs had been predominately infected by R. bellii or closely related genotypes (Table 1). Even though we have not recorded ectoparasitism of dogs, it has been reported that R. amblyommii infects exclusively ticks of the genusAmblyomma. In fact, in the Brazilian Amazonian region,R. amblyommii was detected in Amblyomma cajennense, Amblyomma coelebs, Amblyomma longirostre, and Amblyomma geayi (^{LABRUNA et al. 2004}, ²⁰¹¹). The former two ticks are known to parasitize dogs within the rural areas of the Amazon region (^{LABRUNA et al., 2005}). Therefore, they might be related to the serological status of the dogs of the present study. Urban dogs reacted predominately to the non-Spotted Fever group agent R. bellii, an observation that should be further evaluated. It is worth mentioning that this rickettsia and closely related genotypes have been reported infecting species of nearly all tick genera of the New World (^{PHILIP et al., 1983}, ^{LABRUNA et al., 2011}), as well as a broad range of insects and diverse organisms, including amoeba (^{WEINERT et al., 2009}).

In contrast to southern and southeastern Brazil, rickettsioses have never been reported in humans from northern Brazil (the Amazon region) (^{Labruna et al., 2011}). Indeed, our results indicate that there is circulation of a Spotted Fever group agent closely related toR. amblyommii in the rural area of the study region. To date,R. amblyommii is still considered of unknown pathogenicity. However, it has been proposed that some of the rickettsiosis cases reported as Rocky Mountain spotted fever (RMSF) (presumably caused by R. rickettsii) in the USA may have been caused by R. amblyommii (^{APPERSON et al., 2008}). This assumption relied on serological results, which demonstrated a four-fold increase in endpoint titers to R. amblyommii, but not to R. rickettsii, in acute and convalescent sera samples taken from clinical cases compatible with RMSF (^{APPERSON et al., 2008}).

In summary, this first study on tick-borne agents in dogs from the Brazilian eastern Amazon indicates that these dogs are exposed to several vector-borne infectious agents. These include Babesia organisms, mostly in the urban area, where B. canis vogeli is possibly being transmitted by R. sanguineus ticks; and Spotted Fever groupRickettsia organisms, mostly in the rural area, whereR. amblyommii is possibly being transmitted byAmblyomma ticks. In addition, dogs from both rural and urban areas are similarly exposed to Ehrlichia organisms. While it is well known that E. canis is transmitted by R. sanguineus ticks in the urban areas of the Amazon region, it is possible that other Ehrlichia species are transmitted by native tick species in the rural areas, resulting in similar seroprevalence values between urban and rural dogs. Further studies on tick species and isolation of tick-borne agents from ticks and vertebrate hosts in the Amazon region are needed to better elucidate the epidemiology of tick-borne diseases in the region.