Detection and molecular analysis of Toxoplasma gondii and Neospora caninum from dogs with neurological disorders

Introduction: Toxoplasma gondii and Neospora caninum are related Apicomplexa parasites responsible for systemic diseases in many species of animals, including dogs. Methods: This study aimed to determine the occurrence of T. gondii and N. caninum infections in 50 dogs with neurological signs that were admitted to the Veterinary Hospital of Universidade Estadual Paulista, City of Botucatu, Brazil. All animals were screened for antibodies using an immunofluorescent antibody test for both parasites. Tissues of positive animals were bioassayed in mice (T. gondii) and gerbils (N. caninum), and DNA was analyzed using the polymerase chain reaction (PCR). Positive samples for T. gondii by PCR were typed using restriction fragment length polymorphism-PCR for 11 markers: SAG1, SAG2 (5′-3′-SAG2 and alt.SAG2), SAG3, Btub, GRA6, L358, c22-8, c29-6, PK1 and Apico, and CS3 marker for virulence analysis. Results: Specific antibodies were detected in 11/50 (22%; 95% confidence interval (CI95%), 12.8-35.3%) animals for T. gondii and 7/50 (14%; CI95%, 7.02-26.3%) for N. caninum. In the bioassay and PCR, 7/11 (63.6%; CI95%, 34.9-84.8%) samples were positive for T. gondii and 3/7 (42.9%; CI95%I, 15.7-75.5%) samples were positive for N. caninum. Three different genotypes were identified, but only 1 was unique. Conclusions: These data confirm the presence of T. gondii and N. caninum in dogs from Brazil, indicating the importance of this host as a sentinel of T. gondii for human beings, and the genotypic variation of this parasite in Brazil.

Neospora caninum and Toxoplasma gondii are related coccidian intracellular protozoan parasites that infect many warm-blooded vertebrates causing systemic disease in many species of animals, including production animals and dogs 1,2 .Clinical canine toxoplasmosis rarely results from a primary infection; instead, most dogs that die of toxoplasmosis have a distemper virus infection or other immunosuppressive conditions [2][3][4] .Toxoplasmosis is recognized as an opportunistic disease in dogs, which is characterized by neuromuscular, respiratory, and gastrointestinal signs or by generalized infection, in addition to its most common neurological impairments, e.g., ataxia, behavioral changes, circling, seizures, paralysis, paraplegia, twitching, and tremors 5,6 .T. gondii comprises different clonal lineages that may influence the progression and severity of the disease in animals and humans 7 .
Dogs are the definitive hosts of N. caninum and play a pivotal role in its transmission to other animals, including cattle.Neosporosis has polymorphic clinical signs in dogs.In general, the clinical findings in dogs are similar to those of toxoplasmosis, but neurologic deficits and muscular abnormalities predominate.It also causes myocardial, pulmonary, and dermal disease in dogs 8 .
The simi lar it y bet ween s y mptomat ic toxoplasmosis and neosporosis in dogs reinforces the importance of the differential diagnosis of these diseases.Therefore, this study aimed to determine the occurrence of toxoplasmosis and neosporosis in naturally infected dogs with neurological symptoms and the genotypes circulating in this species.

METHODS
This study was conducted in the ambulatory attendance clinic of the Infectious Disease Section, School of Veterinary Medicine and Animal Science (FMVZ), São Paulo State University (UNESP), São Paulo, Brazil.

Studied animals
Fifty dogs admitted to the Veterinary Hospital from September 2009 to July 2010 presenting with neurological symptoms, such as ataxia, seizures, behavioral changes, paralysis and paraplegia of limbs, and tremors, were studied.

Serology
Dog serum samples were screened for T. gondii and N. caninum antibodies.Serum samples were diluted 4-fold in phosphate-buffered saline (0.01 mol/L, pH 7.2) from a 1:16 dilution and promptly tested using an immunofluorescent antibody test (IFAT) for T. gondii immunoglobin G (IgG) antibodies, as described by Camargo 9 , and for N. caninum IgG antibodies under a fluorescence microscope (Zeiss SG250), as described by Dubey et al. 3 , using 1:25 as a cut-off point.Both tests were performed using a commercially available mouse fluorescin isothiocynale (FITC)-labeled IgG Fc antibody conjugate (Bethyl Laboratories Inc., USA).Positive sera controls for toxoplasmosis were obtained by the chronic infection of mice inoculated with the RH strain by the subcutaneous (s.c.) route; while for neosporosis, they were obtained by the chronic infection of gerbils inoculated with the NC-1 strain by the s.c.route.Negative control sera for toxoplasmosis and neosporosis were obtained from mice and gerbils, respectively, inoculated with sterile saline solution by the s.c.route.
For the serological test, tachyzoites of N. caninum (NC-1 strain) and T. gondii (RH strain) were used as antigens.Dog serum samples showing complete (non-polar or bipolar) parasite fluorescence at dilutions ≥1:16 or ≥1:25 for toxoplasmosis and neosporosis, respectively, were classified as positive.

Bioassay in the experimental models
Brain samples from all seropositive animals were bioassayed by s.c.injection with 1 mL of each tissue sample in 4 mice for the isolation of T. gondii and 4 gerbil for the isolation of N. caninum, on the basis of its specific serology 10 .Peritoneal fluid from all of the mice or gerbils that died was examined for T. gondii and N. caninum tachyzoites.From the surviving animals, blood was collected by retroorbital sinus puncture on day 30 post-inoculation.The animals were considered positive when tachyzoites or tissue cysts were found in the brain or when showing positive serology in IFAT 3 .

Polymerase chain reaction
The extraction and purification of DNA from brain samples were carried out by using the Illustra Tissue and Cells GenomicPrep Mini Spin Kit (GE Healthcare Life Sciences do Brasil Ltda ® , Brazil).
Polymerase chain reaction was performed by using the primers TOX4 and TOX5 described by Homan et al. 11 , which amplify a 529-base pair (bp) fragment, AF146527 (GenBank), and repeated 200-to 300-times in the T. gondii genome.PCR for N. caninum was performed using the species-specific primer pair Np21-Np6, directed to the genomic Nc-5 region, which amplify a 328-bp fragment 12 .

Statistical analysis
Statistics associated with the results from the bioassay and PCR were calculated by adopting antibody detection as a standard screening test using the spreadsheet described by Mackinnon 18 .The associations between the epidemiological variables and the serological results were analyzed by the chi-square or Fisher's exact tests, considering α = 0.05 19 .For the analysis of the results, concordance testing was performed among the serology, bioassay, and PCR data using the McNemar test 20 .All tests were carried out using the EpiInfo TM v.3.5.1 program 21 .
Both parasitological tests (i.e., bioassay and PCR) had a 100% agreement for toxoplasmosis and neosporosis.Comparing the results obtained by IFAT and PCR for toxoplasmosis, or IFAT and the bioassay, the tests had an 82% agreement.For the comparison of the 3 diagnostic tests for neosporosis, PCR and the bioassay had an agreement of 82.4%, IFAT and the bioassay test had an agreement of 88.2%, and IFAT and PCR had an agreement of 94.1%.
No variables presented with significant differences for the IFAT results for the analysis of T. gondii and N. caninum antibodies.Significant associations were observed for both parasites in IFAT [p = 0.038; odds ratio (OR) = 6.8571 (CI95%, 1.2503-37.6068)]and PCR [p = 0.037; OR = 21.5000(CI95%, 1.5810-292.3818)].The genotyping results and the detection rates in other studies are presented in Table 1.

DISCUSSION
Neurological signs of toxoplasmosis, neosporosis, and distemper are similar, emphasizing the importance of the differential diagnosis of these diseases.Toxoplasmosis is recognized as one of the most common diseases in dogs with neurological signs 5 , and has been related to combined infections with distemper 22 .N. caninum infections have been described in animals affected with toxoplasmosis and it should be considered in the differential diagnosis or with concurrent detection in cases like the present one 23 .
In the present study, T. gondii was isolated from the brain tissue of 7 of 11 dogs with IFAT titers of ≥1:16.N. caninum was isolated from the brain tissue of 2 of 7 dogs with IFAT titers of ≥1:25.These data are in agreement with those of Cavalcante et al. 24 who identified N. caninum in dogs fed with masseter, heart, liver, and brain from infected cattle.
In the detection of T. gondii infection, PCR presented with a higher sensitivity than the conventional diagnostic methods.Two dog samples with negative results for serology and the bioassay generated positive PCR results.In this case, one can possibly suggest that these dogs had a chronic previous infection with non-detectable antibody at the 1:16 dilution, which was associated with a parasite load that was too low to be detected in the bioassay, but enough to be detected by PCR.Additionally, the sensitivity of PCR depends on whether the chosen tissue sample contains parasite DNA.
For neosporosis diagnosis, serology demonstrated a greater sensitivity; from 7 positive 2 were positive in the bioassay and 3 were positive with PCR.Additionally, serological studies using IFAT performed in different hosts have shown that there is little cross-reactivity with the coccidia of other parasites; therefore, IFAT has been considered the standard test for the diagnosis of neosporosis.PCR with the primer pair Np21/Np6 can be an efficient tool for large-scale epidemiological studies using brain tissue obtained at necropsy 12 .Despite the sensitivity and specificity of PCR, the brain tissue used may not have contained parasite DNA.This is a limitation of DNA detection, not only for N. caninum but also for other microorganisms.Hůrková & Modrý 25 detected N. caninum DNA in the brains of only 4.6% (7/152) red foxes (Vulpes vulpes).
Three genotypes were identified in the present study, TgDogBr1-3.All of them were avirulent in mice.Only one TgDogBr1 presented with a unique genotype.TgDogBr2 presented with a genotype identical to the reference strain TgCatBr1, which was previous identified by Su et al. 13 in cats from the State of Paraná, Brazil.TgDogBr3 was identical to the reference strain P89, which was previous reported by Dubey et al. 26 in a pig from Iowa State, USA.Phenotypic differences are observed between P89 and TgDogBr3.TgDogBr3 was avirulent, while P89 was classified as virulent for mice 26 , where only 1 oocyst is lethal to mice by any route of administration.TgDogBr3, as with the other genotypes, presented with the type I or III allele at the CS3 locus.Da Silva et al. 27 observed that 5 (62.5%) of 8 strains with the type II allele at the CS3 locus were avirulent in mice, while 3 (37.5%)strains with the type II allele were virulent.In other reports, 82% of strains with the type II allele at the CS3 locus were virulent in mice 16 , and 85.5% of strains with the type II allele at the CS3 locus were virulent in mice from another study in Brazil 28 .According to the classification formulated by Pena et al. 16 , TgDogBr2 was classified in the Type BrII group (intermediate virulence), while TgDogBr3 was classified in the Type BrIII group (avirulent), which correlates with the virulence observed in the present study.TgDogBr1 did not match with any of the types proposed by Pena et al. 16 .
These data confirm both types of infection in dogs, the importance of this host as a reservoir for these pathogens for humans, and the genotypic variation of this parasite in Brazil.