Prevalence and Multilocus Genotyping Analysis of Cryptosporidium and Giardia Isolates from Dogs in Chiang Mai, Thailand

The occurrence and zoonotic potential of Cryptosporidium spp. and Giardia duodenalis isolated from dogs in Chiang Mai, Thailand were determined. Fecal samples were collected from 109 dogs between July and August 2008. Cryptosporidium spp. infection was determined by immunofluorescent assay (IFA), PCR assays that amplify Cryptosporidium heat-shock protein 70 kDa (hsp70), and two PCR assays that amplify a small subunit-ribosomal RNA (SSU-rRNA). Giardia duodenalis infection was identified using zinc sulfate centrifugal flotation, IFA, and four PCR assays that amplify the Giardia glutamate dehydrogenase (gdh), beta-giardin (bg), and generic and dog-specific assays of triosephosphate isomerase (tpi) genes. Overall prevalence of Cryptosporidium spp. and G. duodenalis was 31.2% and 45.9%, respectively. Sequence analysis of 22 Cryptosporidium-positive samples and 21 Giardia-positive samples revealed the presence of C. canis in 15, and C. parvum in 7, G. duodenalis Assemblage C in 8, D in 11, and mixed of C and D in 2 dogs. Dogs in Chiang Mai were commonly exposed to Cryptosporidium spp. and G. duodenalis. Cryptosporidium parvum can be isolated from the feces of dogs, and all G. duodenalis assemblages were dog-specific. Dogs could be a reservoir for a zoonotic Cryptosporidium infection in humans, but further studies will be required to determine the clinical and zoonotic importance.


Introduction
Cryptosporidium spp. and Giardia duodenalis are common intestinal protists that can infect humans and animals worldwide [1]. The clinical signs of cryptosporidiosis and giardiasis in dogs vary from sub-clinical to severe diarrhea [2,3].
At least 27 species of Cryptosporidium spp. and eight assemblages (A-H) of G. duodenalis have been described [4,5]. Although dogs are commonly infected with species-specific C. canis and G. duodenalis (Assemblages C and D), the occurrence of zoonotic C. parvum and G. duodenalis (Assemblages A and B) in dogs have raised concern that these animals may serve as a potential reservoir for human transmission [6].
In Thailand, studies regarding cryptosporidiosis and giardiasis in dogs and their zoonotic potential are limited. In one study, C. canis was identified in 2 of 95 temple dogs in central Thailand using PCR that amplify an 830-bp fragment of small subunit-ribosomal RNA (SSU-rRNA) gene [7]. The prevalence of G. duodenalis infection in temple dogs in the Bangkok area varied from 7.9-56.8% depending on the test used [8,9]. The majority of G. duodenalis isolates recovered in these samples were Assemblages A and D. It has been noted that similar genotypes (Assemblage A) were recovered from dogs and humans in the same monastery. In another study in a shelter in Nakornnayok province, the prevalence of Giardia infection in shelter dogs using a formalin-ether concentrating technique was 2.8% [10]. To our knowledge, there has been no previous research concerning Cryptosporidium spp. and G. duodenalis infections and their zoonotic potential in dogs in this area. Since these protist infections are a potential public health concern, determining the prevalence and genotypes of these organisms in dogs living in close proximity to humans and other animals is a priority. Therefore, the aims of this study were to estimate the prevalence of Cryptosporidium spp. and G. duodenalis infections in dogs in Chiang Mai, Thailand, and to characterize the organism isolates using molecular techniques in order to determine the potential for zoonotic transmission.

Study Location
Chiang Mai is the second largest province of Thailand. It is located in the northern part of the country at geographic coordinates 18 • 47 N and 98 • 59 E. The city of Chiang Mai maintains its deep roots of traditional community culture in a hybrid landscape of rural and urban city development, and includes agricultural, industrial, and tourism areas. Chiang Mai also represents a tropical environment, which exists in many parts of the world.

Sample Collection
Between July and August 2008, 109 canine fecal samples were obtained from animals visiting the Small Animal Hospital of the Faculty of Veterinary Medicine, Chiang Mai University (n = 36), private clinics (n = 9), a shelter (n = 15), or breeders (n = 49) in Chiang Mai province, Thailand. The samples were collected on a volunteer basis regardless of the health status of the animals. Demographic information (age, sex, and housing types) was recorded. Fecal consistency was determined using the Nestle Purina Fecal Scoring System for Dogs and Cats (Nestle-Purina Pet Food Co, St. Louis, MO, USA). Fecal scores of 1-3 were considered as normal, with 4-7 classified as diarrheic.

Determination of Cryptosporidium and Giardia Infections
Cryptosporidium spp. infection was determined using immunofluorescent assay (IFA) and PCR techniques. Giardia duodenalis infection was determined using zinc sulfate centrifugal flotation, immunofluorescent assay, and PCR techniques.

Zinc Sulfate Centrifugal Flotation and Microscopy
Fecal consistency was determined upon the receipt of the sample, and all fecal samples were stored in closed plastic containers at 4 • C. Microscopic examination of feces after the performance of a conventional zinc sulfate centrifugal flotation was used to determine intestinal parasitic infection within 5 days of collection, and the remaining fecal samples were stored at −20 • C until being shipped to Colorado State University for IFA and molecular analysis. All fecal samples were shipped to the USA on dry ice and stored at −20 • C until processed.

Fecal Concentration and Immunofluorescent Assay
Prior to IFA and DNA extraction, all fecal samples were concentrated using sucrose gradient centrifugation technique as previously described [11,12]. The IFA slides were processed according to the manufacturer's instructions (Merifluor ® Cryptosporidium/Giardia IFA kit, Meridian Diagnostic Corporation, Cincinnati, OH, USA). The remaining concentrated fecal material was stored at −20 • C until DNA extraction was performed.
2.3.3. Molecular Detection of Cryptosporidium spp. and Giardia duodenalis Infection Three hundred microliters of each fecal concentrate were subjected to DNA extraction following an established protocol [13]. Three PCR assays for Cryptosporidium identification were performed. PCR assays amplify a 325-bp fragment of the heat-shock protein (hsp70), a~290-bp fragment (one-step PCR), and an~830-bp fragment (nested PCR) of the small subunit-ribosomal RNA (SSU-rRNA) genes were utilized to detect the presence of Cryptosporidium spp. [14][15][16]. For Giardia molecular identification, four nested PCR assays targeting a 432-bp fragment of glutamate dehydrogenase (gdh), a 510-bp fragment of beta-giardin (bg), and a 511-bp fragment of triose phosphate isomerase genes using generic primers (tpigen) and dog-specific primers (tpiD) were performed as previously described [17][18][19][20]. All PCR assays had several modifications from original publication. PCR mix consisted of 1× HotStarTaq Master Mix (Qiagen, Valencia, CA, USA), 10 pmol of each primer, and 1 µL of template DNA in a final volume of 25 µL for each targeting gene. PCR positive and negative controls were included in every PCR reaction. The Giardia positive control was obtained from a dog sample that tested positive for G. duodenalis by all four Giardia PCR assays, and was subsequently sequenced. The Cryptosporidium positive control was obtained from a C. parvum-positive cow. The negative control contained the PCR reagents but no DNA. In addition, in nested PCR assays, the negative controls from primary PCRs were included in the secondary PCR assays to evaluate the possibility of contamination.

DNA Sequencing and Genotyping Analysis
The PCR products were evaluated by nucleotide sequencing using a commercially available service (Proteomics and Metabolomics Facility, Colorado State University). The obtained sequences were compared with nucleotide sequences from the nucleotide database from the GenBank by BLAST analysis (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Data Analysis
A sample was considered positive for Cryptosporidium if the sample was positive by either IFA or any of the Cryptosporidium PCR assays, and considered positive for Giardia if the sample was positive by either zinc sulfate fecal flotation, IFA, or any of the Giardia PCRs. Overall prevalence and 95% confidence intervals (CI) were calculated [21]. Associations between G. duodenalis or Cryptosporidium spp. infections and age (less than one year or one year or more), sex, diarrhea status (yes or no), and housing type (household or breeding kennel/shelter) were assessed using Fisher's exact test [21]. Odds ratios and 95% CI were estimated using univariate logistic regression analysis to measure the strength of association of each independent variable including age, sex, diarrhea status, housing type, and the presence of co-infection (having both Cryptosporidium and Giardia). A multivariate logistic regression model against either Cryptosporidium spp. or G. duodenalis infection in dogs was constructed using a backward stepwise elimination procedure [22]. Variables found to be associated with Cryptosporidium spp. or G. duodenalis infection in the univariate logistic regression (p < 0.25) were included in the multivariable logistic regression analysis. Variables were retained in the model based on the likelihood ratio χ 2 statistic, at p ≤ 0.05. All statistical analyses were performed using the Stata statistical software release 10.1 (Stata Corp., College Station, TX, USA).

Detection of Cryptosporidium spp. and Giardia duodenalis Isolates
A single fecal sample was collected from 109 dogs. The characteristics of the samples are shown in Table 1 Table 2). In addition, in dogs, single infections with Cryptosporidium spp. or G. duodenalis were 14.7% (16/109) and 29.4% (32/109), respectively. Co-infection of G. duodenalis and Cryptosporidium spp. was shown in 16.5% (18/109) of the samples.

Genotyping of Cryptosporidium spp. and Giardia duodenalis Isolates
Eleven sequences from Cryptosporidium hsp70, eleven sequences from Cryptosporidium one step SSU-rRNA, and eight sequences from Cryptosporidium nested SSU-rRNA PCR positive samples were available for genotyping analysis. Using BLAST analyses, 15 dog isolates were typed as C. canis and seven were typed as C. parvum (Table 3). C. parvum n/a n/a TH08Dog46 C. canis C. canis C. canis TH08Dog54 C. parvum n/a n/a TH08Dog55 C. canis C. canis C. canis TH08Dog58 C. canis n/a n/a TH08Dog61 n/a n/a C. canis TH08Dog68 n/a n/a C. canis TH08Dog69 n/a C. canis n/a TH08Dog71 n/a C. canis n/a TH08Dog76 C. parvum. n/a n/a TH08Dog86 C. parvum. n/a n/a TH08Dog87 C. parvum. n/a n/a TH08Dog92 C. canis n/a n/a TH08Dog96 C. canis n/a n/a TH08Dog101 C. parvum. n/a n/a TH08Dog102 n/a C. canis n/a TH08Dog107 n/a C. canis C. canis n/a = not available.
Twenty-one sequences from gdh, 18 sequences from bg, 8 from generic tpi, and 15 dog-specific tpi PCR positive samples were available for analysis. Eight dog isolates were typed as G. duodenalis Assemblage C, 12 were typed as D, and one C or D depending on target genes (Table 4). Table 4. Giardia genotypes determined by nucleotide sequence analyses of glutamate dehydrogenase (gdh), β-giardin (bg), and triose phosphate isomerase (tpi) PCR products from dog samples in Chiang Mai, Thailand.

ID gdh bg tpigen a tpid b
TH08Dog5 D D n/a D TH08Dog15 D D n/a n/a TH08Dog17 D n/a n/a n/a TH08Dog43 D D n/a D TH08Dog45 D ash D n/a D TH08Dog73 C C C C TH08Dog93 D n/a n/a n/a TH08Dog96 D n/a n/a n/a TH08Dog100 C C n/a n/a TH08Dog101 C C C C TH08Dog103 C C n/a n/a TH08Dog107 D ash C ash C ash D TH08Dog108 C ash C C C ash a tpi with generic primers; b tpi with dog specific primers; n/a = not available; ash = allelic sequence heterogeneity.

Statistical Analysis
Using χ 2 or Fisher's exact tests, age was significantly associated with the prevalence of both Cryptosporidium spp. and G. duodenalis (Table 2). Other variables were not associated with infection.

Discussion
The current study represents the first report of the Cryptosporidium spp. and G. duodenalis prevalence rates and genotypes/species in dogs in Chiang Mai, Thailand. The global prevalence of Cryptosporidium spp. and G. duodenalis infection in dogs varies depending on the test used, geographic location, and population tested [4,23]. In the present study, overall Cryptosporidium spp. and G. duodenalis prevalence was 31.2% and 45.9%, respectively. These high prevalences were derived by considering detection in parallel from four tests for Cryptosporidium spp. and six tests for G. duodenalis.
The prevalence of Cryptosporidium spp. is comparable to a previous report of sled dogs from Poland [24], and the prevalence of Giardia found in this study is comparable to a previous report of 56.8% in Bangkok [9] and similarly high rates in other countries such as Japan [25], Mexico [26], Brazil [27], Italy [28], and Belgium [29], where most of the studies were from breeding kennels, shelters or abandoned dogs. Nevertheless, the prevalence of these two organisms in this study may have been overestimated due to selection bias. The samples available for this study were not randomly selected, but depended on voluntary participation of the owner visiting the small animal hospital and caregivers of breeders and a shelter. Therefore, the sample may have been biased towards infected and diarrheic animals, resulting in an overestimation of the apparent prevalence.
Zoonotic species or genotypes of Cryptosporidium spp. and G. duodenalis cannot be distinguished from host-adapted organisms using morphological differentiation. Therefore, molecular characterization using PCR assay and sequence analysis is suggested due to its rapidity and specificity to differentiate the species or genotypes of these organisms. However, not all PCR assays have the same sensitivity for detecting Cryptosporidium or Giardia nucleotides in fecal samples. In the current study, hsp70 and SSU-rRNA are not in agreement. Of 22 Cryptosporidium PCR positive samples, nine were identified from hsp70 only, five from one-step SSU-rRNA only, and two from nested SSU-rRNA only, four from both one-step and nested SSU-rRNA, and two from all three PCRs. It is unclear whether hsp70 or SSU-rRNA have an advantage over each other due to the limitation on the PCR of biological or fecal samples. Of three targeting genes for Giardia detection, Giardia gdh PCR had the highest amplification rate compared to bg and tpi genes ( Table 4). This observation was similar to the study by Scorza and colleagues [12], which showed that the gdh PCR had higher amplification rate than bg or tpi. However, this observation contrasted with the studies by Covacin et al. [30] and Sprong et al. [31], which showed that gdh PCR had the least amplification rate compared to bg, tpi, and SSU-rRNA. In addition, the discrepancies of genotype determination among these three genes have also been reported. Based on this study and our experiences, PCR for the gdh gene may be suggested if the multilocus PCR assay is not affordable.
The majority of genotypes of G. duodenalis that infect dogs are host-adapted (Assemblages C or D); however, the pattern can differ geographically [4,12,32]. For example, in the Western United states, zoonotic genotypes of Giardia Assemblages A and B were highly prevalent [30], whereas in temple dogs in Bangkok, Thailand, the majority of Giardia isolates were identified as Assemblage A [9]. In the current study, all of the G. duodenalis isolates were dog-adapted assemblages (C or D). Therefore, the potential of zoonotic Giardia transmission from pet dogs in this location is possibly low.
In the current study, from 22 Cryptosporidium PCR-positive dogs, 15 specimens were identified as C. canis (68%) and 7 specimens (32%) were identified as C. parvum. The rate of C. canis detection in this study was relevant to previous studies of Cryptosporidium isolates from dogs worldwide; 41 Cryptosporidium isolates that had been previously reported in dogs, 76% of the isolates were identified as C. canis, 22% as C. parvum, and 2% as C. meleagridis [33,34]. Due to the nature of the cross-sectional study, we are not certain whether the isolated C. parvum was a pathogen circulating in the dog population, or transmitted from other animals or humans. The presence of C. parvum in the dog samples suggests that dogs could be a potential reservoir for the zoonotic transmission of Cryptosporidium spp. While C. parvum and C. hominis are significant causes of human cryptosporidiosis, the detection of C. canis in HIV patients in Thailand [35][36][37] and elsewhere [38,39], as well as the detection of C. canis in children [40] have raised concerns regarding the transmission of protozoal diseases from pets to humans even when they harbor the host-adapted pathogens. Further investigation of these parasites among humans and animals living in the same household or in close proximity are needed to confirm this relationship. However, good sanitary practices are highly recommended for all pet owners to avoid zoonotic transmission to humans.
Cryptosporidium and Giardia genotypes/species isolates from dogs in this study may not reflect the majority of genotypes/species for the dog population as a whole in Chiang Mai, Thailand, since we made our interpretation in the light of the available nucleotide sequences. The information regarding the genotype/species for 35% of Cryptosporidium-infected samples and 46% of Giardia-infected samples was unknown. Failure of PCR assays to amplify the organisms' target genes may be from the presence of a PCR inhibitor [41], or degradation of DNA material in the samples which may result from international shipment or long-term storage before PCR processing. Therefore, the failure to amplify Cryptosporidium or Giardia genes in the fecal samples using PCR did not rule out Cryptosporidium or Giardia infection. Thus, a PCR should not be used as the primary test for Cryptosporidium or Giardia clinical diagnosis for practical and cost effective reasons, but it certainly has an important role for confirmation and in research.
Young age, presence of diarrhea, feeding a home-cooked diet, presence of other enteric parasites, being an abandoned or stray dog, and having been kept in a kennel are risk factors that have been reported to be associated with Cryptosporidium and Giardia in previous studies [42][43][44][45]. Similarly, in this study, G. duodenalis infection was shown to be associated with young age, the presence of diarrhea, and coming from a breeder or a shelter. However, to help in prevention and control of these pathogens in dogs in Chiang Mai area, the important risk factors mentioned above including history of the pet's acquisition, season, and source of drinking water could be applied to this population.

Conclusions
The current information suggests that the Cryptosporidium and Giardia infections in young dogs in Chiang Mai are common. Dogs may be a reservoir for zoonotic transmission of Cryptosporidium; however, dogs may not be a primary reservoir for zoonotic transmission of G. duodenalis. Further investigation using molecular analysis of Cryptosporidium and Giardia species/genotypes isolated from animals and humans (pets and owners or shelter animals with the caregivers) may clarify the transmission cycle of these organisms between humans and animals in the same environmental setting.