T. gondii isolates in North America and Europe present a highly clonal population structure mainly consisted of 4 lineages, namely type I, II, III, and 12 [6-9], which have genetic and biological differences from that in South America [10-12]. Several studies have documented the genotype of T. gondii isolates from different animals in China [13-15], but information about genotyping of T. gondii isolates from dogs in China is limited. Therefore, the current study was conducted to determine the prevalence of T. gondii infection in dogs from Zhanjiang, southern China, and to study the genotypes of the T. gondii isolates.

A total of 364 blood samples and 432 liver tissue samples were collected from dogs in Zhanjiang city, southern China, between December 2012 and January 2013. Dogs were randomly selected from the dog farms. These dogs are farmed for meat and are mainly consumed by the local people. One blood sample or 1 liver tissue sample was collected from a dog. However, the blood samples did not correspond to the tissue samples by the number due to the constraint of field condition. When the dog was slaughtered, tissue samples were collected, and blood samples were drawn from jugular vein into a sterile, plain centrifuge tube. Then, the tissue samples were saved in the microtubes at 4˚C. The blood samples were left to clot at room temperature for 6 hr and centrifuged at 3,000 rpm for 10 min. The separated sera were stored at -20˚C until needed for ELISA. This study was approved by the Animal Ethics Committee of Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences (permit code, LVRIAEC2012-007). Dogs were handled in strict accordance with the Animal Ethics Procedures and Guidelines of the People's Republic of China.

IgG antibodies to T. gondii were determined using a commercially available ELISA kit (Haitai, Zhuhai, Guangdong Province, China) according to the manufacturer’s recommendations. A serum sample was considered positive when the value was 1.1 times higher than the mean value of positive control, negative control, and blank control. Genomic DNA was extracted from these liver tissues using TIANamp Genomic DNA kit (TianGen™, Beijing, China) according to manufacturer’s recommendations and previous descriptions [15]. Then, a semi-nested PCR based on B1 gene was employed to detect the T. gondii DNA following the previously described method [16]. B1 gene positive DNA samples were submitted to nested PCR amplification of 3'- and 5'- SAG2 [17,18], followed by digestion with restriction enzymes MboI and Hha1, respectively. The products were resolved in 2.5-3.0% agarose gel to display restriction fragment length polymorphisms (RFLP) using a gel document system (UVP GelDoc-It™ Imaging System, Cambridge, UK). Six T. gondii strains, namely GT, PTG, CTG, MAS, TgCatCa1, and TgCatBr5, were used as references.

Of the 364 serum samples of dogs in Zhanjiang city, southern China, 189 (51.9%) reacted positively. A previous study reported a seroprevalence of 70.9% in 175 stray dogs housed in shelters at Umuarama city, Brazil [19]. Alvarado-Esquivel et al. [20] reported a high seroprevalence of 67.3% in dogs in Veracruz, Mexico. The overall seroprevalence (51.9%) of T. gondii in dogs in Zhanjiang city was lower than that of Umuarama and Veracruz, but much higher than that observed in other parts of China, such as Guangzhou [21], Lanzhou [22], Kunming [23], and Shenyang [24]. High prevalence of T. gondii infection in dogs in Zhanjiang city reported here indicates a high environmental contamination with oocysts. Cats are the only definitive hosts of T. gondii, playing a significant role in the transmission of this parasite [1]. During the present investigation, we observed that cats were present on the farm and had free access to the dogs and their feed. This may represent a major risk factor for T. gondii infection in the dog farm.

T. gondii DNA was detected in 37 of the 432 (8.6%) liver tissue samples by using semi-nested PCR targeting the B1 gene. The target fragment was about 130 bp in length (Fig. 1). Four of these positive PCR products were randomly selected and sent to sequencing. Sequence comparison and analysis revealed 100% homology with the published T. gondii B1 gene sequence (GenBank accession no. AF179871). Genotyping of positive DNA samples was performed by employing PCRRFLP technique. Due to low DNA concentration, only 8 of 37 positive DNA samples gave the PCR-RFLP data on 3'- and 5'-SAG2, and they were identified as type I (Fig. 2). The results of genotyping of these isolates and 6 references are summarized in Table 1.

ELISA is among the most commonly used methods for investigation of IgG antibody. IgG antibodies usually appear within 1-2 weeks of acquisition with T. gondii infection, peak within 1-2 months, decline at various rates, and usually persist for life. Because of its high sensitivity and specificity, low cost, and ease of practice, ELISA is widely used for diagnosis of T. gondii infection. In the present study, seroprevalence of IgG antibodies (51.9%) does not keep in agreement with the prevalence of T. gondii DNA (8.6%). Actually, IgG antibody based seroprevalence mainly reflects that exposure of dogs to T. gondii infection in the investigated geographic area may be very common, while DNA detection reveals the presence of viable T. gondii in dogs and that they may be mostly acute infections. New detection methods such as IgG antibody avidity test and more studies are needed to explore the T. gondii infection details in dogs in Zhanjiang city. Anyway, what we could confirm in this study was that the dog farm is seriously contaminated with T. gondii oocysts. Urgent measures should be taken to prevent infection from spreading.

Limited data about genotyping of T. gondii isolates from dogs in China is available. A previous study reported the genotype of T. gondii isolates from dogs in Henan province and considered it as a type I variant [25]. The present result shared the same type at the (3'+5') SAG2 loci with T. gondii isolates from dogs in Henan [25]. This may suggest that T. gondii isolates from dogs in Zhanjiang city may belong to type I or a type I variant. However, further studies of sampling more dog samples from wider geographical locations are needed to draw a valid conclusion.

The present survey showed that T. gondii prevalence in dogs in Zhanjiang city, southern China is high. The dog meat is consumed in this region by the local people, and T. gondii is considered as an important food-borne parasite. Thus, dogs can serve as a transport host for T. gondii to humans. Therefore, it is essential to implement integrated measures to prevent and control T. gondii infection in dogs. Moreover, it is urgent to improve the eating habit of the local people and implement T. gondii-inspection during dog slaughtering and processing.