Prevalence and molecular characterization of Cryptosporidium spp. and Giardia duodenalis in deer in Henan and Jilin, China

Little is known about the prevalence and zoonotic potential of Cryptosporidium spp. and Giardia duodenalis in deer in China. In this study, 662 fecal samples were collected from 11 farms in Henan and Jilin Provinces between July 2013 and August 2014, and were screened for the presence of Cryptosporidium and G. duodenalis with genotyping and subtyping methods. Cryptosporidium spp. and G. duodenalis were detected in 6.80% (45/662) and 1.21% (5/662) of samples, respectively. Six Cryptosporidium species/genotypes were identified based on the small subunit ribosomal ribonucleic acid (SSU rRNA) gene: C. parvum (n = 11); C. andersoni (n = 5); C. ubiquitum (n = 3); C. muris (n = 1); C. suis-like (n = 1); and Cryptosporidium deer genotype (n = 24). When five of the 11 C. parvum isolates were subtyped by sequencing the 60 kDa glycoprotein (gp60) gene, zoonotic subtypes IIaA15G2R2 (n = 4) and IIdA19G1 (n = 1) were found. According to a subtype analysis, three C. ubiquitum isolates belonged to XIIa subtype 2. In contrast, only assemblage E was detected in the five Giardia-positive samples with small subunit ribosomal ribonucleic acid (SSU rRNA) gene sequencing. To our knowledge, this is the first study to report C. andersoni, as well as C. parvum zoonotic subtypes IIaA15G2R2 and IIdA19G1 in cervids. These data, though limited, suggest that cervids may be a source of zoonotic Cryptosporidium and Giardia. Cervids in the present study are likely to be of low zoonotic potential to humans, and more molecular epidemiological studies are required to clarify the prevalence and public health significance of Cryptosporidium and G. duodenalis in cervids throughout China.


Background
Cryptosporidium and Giardia are two common protozoan parasites responsible for diarrhea in a broad range of vertebrate hosts, including humans, and domestic and wild animals worldwide. Transmission of both pathogens is by the fecal-oral route with both zoonotic and anthroponotic transmission cycles [1,2]. The host plays an important role in the clinical impact of Cryptosporidium and G. duodenalis infections and the expression of disease. Drug treatments for these infections are inadequate, and do not provide a reliable strategy for their control [3].
As an important center of mammalian evolution and dispersal, China possesses an abundance of deer species [30]. Deer and their products are of high economic value and deer farming has become an important component of China's animal breeding industry. Sika deer, red deer, sambar deer, white-lipped deer, reindeer, Eld's deer and Père David's deer are the major species farmed in China. There are approximately 550,000 domesticated sika deer, most of which are distributed in northwestern China [31]. Velvet antlers, important in traditional Chinese medicine, are one of the main products derived from sika deer. Père David's deer is an endemic species in China, but has become extinct in the wild [32].
Some epidemiological surveys exist concerning Cryptosporidium and Giardia in cervids around the world, but little is known about the prevalence and molecular characteristics of Cryptosporidium and Giardia in cervids in China. Only one study was conducted in Zhengzhou, where 124 fecal specimens were examined and two C. ubiquitum isolates were identified in sika deer [14]. In the present study, deer-derived Cryptosporidium and Giardia isolates were genetically characterized to better understand the distribution and zoonotic potential of the two pathogens in cervids in Henan and Jilin provinces.

Samples
In total, 662 samples were collected between July 2013 and August 2014 from 11 farms in Henan and Jilin provinces, from 16 red deer (Cervus elaphus), 47 Père David's deer (Elaphurus davidianus) and 599 sika deer (Cervus nippon) ( Table 1). Red deer and sika deer were in shed-feeding and housed in separate breeding houses according to different deer species and age groups. These 615 animals had a wide age distribution, ranging from 1 month to 15 years. However, the Père David's deer from one forest farm were so agile, solitary and secretive, that it was difficult to determine their precise ages. Approximately 50 g of fresh feces was collected from each deer immediately after its defecation onto the ground, using a sterile disposal latex glove, and was then placed individually into a disposable plastic bag. No obvious clinical signs were observed in these deer, except for one case of diarrhea. The specimens were transported to the laboratory in an insulated container containing cold packs. Upon arrival, a portion of each specimen was examined by microscopy to detect Cryptosporidium oocysts and Giardia cysts using Sheather's sugar flotation technique and Lugol's iodine stain method, respectively. Wet smears were examined using a bright-field microscope with 100× and 400× magnification. All of the fecal specimens were stored in 2.5% potassium dichromate solution at 4°C prior to DNA extraction.

DNA extraction
The fecal specimens were washed three times in distilled water and centrifuged at 3000× g for 10 min to remove the potassium dichromate. DNA was extracted from 200 mg of each fecal specimen using the E.Z.N.A. Stool DNA Kit (Omega Biotek Inc., Norcross, GA, USA), according to the manufacturer's instructions. The extracted DNA was stored at -20°C.

Genotyping and subtyping
Cryptosporidium species were identified by nested PCR amplification and sequencing of an~830 bp fragment of the small subunit ribosomal ribonucleic acid (SSU rRNA) gene, as described previously [2]. Cryptosporidium parvum and C. ubiquitum were subtyped by a sequence analysis of the 60 kDa glycoprotein (gp60) gene [33,34]. The assemblages of G. duodenalis were determined by sequencing the small subunit ribosomal ribonucleic acid (SSU rRNA), β-giardin (bg), glutamate dehydrogenase (gdh), and triosephosphate isomerase (tpi) genes [35,36]. Replicate analyses were done at each locus using both positive and negative controls.

Sequence analysis
All PCR amplicons were sequenced on an ABI Prism™ 3730 XL DNA Analyzer (Applied Biosystems, Foster, CA, USA), using the BigDye Terminator v.3.1 Cycle Sequencing Kit (Applied Biosystems). The sequencing accuracy was confirmed by two-directional sequencing. The sequences were identified by their alignment with reference sequences downloaded from GenBank using the MEGA 6.0 software [37]. Representative nucleotide sequences generated in this study have been deposited in the GenBank database under the accession numbers KX259127-KX259145 and MG921620-MG921622.

Statistical analysis
The statistical analysis was performed with the SPSS 22. 0 software. Chi-square test and 95% confidence intervals (CIs) were used to compare the Cryptosporidium prevalence rates among different locations and age groups, and differences were considered significant at P < 0.05.
When the 11 C. parvum isolates sequences were compared with each other, five variants were detected. Variant 1 was identical to sequence AB513881 derived from calves in Egypt [38], while variants 2-5 had one or two nucleotide substitutions at six sites in the SSU rRNA sequence; these variants have only been found in this study. Two variants of C. andersoni were identified, and variant 1 and variant 2 were identical to sequences KF826313 and KF826314, respectively, isolated from Chinese outpatients with diarrhea [39].

Cryptosporidium subtypes
All 11 C. parvum isolates were subtyped by sequence analysis of the gp60 gene. However, only five of them produced the expected PCR products, which were identified as two subtypes: IIaA15G2R2 (n = 4, GenBank: KX259142) and IIdA19G1 (n = 1, GenBank: KX259141). IIaA15G2R2 was identical to strains isolated from cattle in the USA (DQ630517), and humans in Canada (DQ192501) and the USA (JX575583), while IIdA19G1 was identical to strains isolated from goats in China (KM199738), cattle in China (HQ009809) and Egypt (JX237824), and humans in China (JQ796092 and JF691561) and Sweden (KU852713). The C. ubiquitum isolates were subtyped to family XIIa and belonged to subtype 2 (GenBank: KX259143-KX259144), sharing 100% homology with a strain isolated from goats in China (KM199742) and a human-derived isolate from Turkey (JX412919). To the best of our knowledge, this is the first time IIaA15G2R2 and IIdA19G1 have been detected in deer.

Giardia assemblage
A total of five G. duodenalis isolates were successfully amplified and sequenced at the SSU rRNA (n = 5), bg (n = 3) and gdh (n = 2) loci, but amplifying failed at the tpi locus despite repeated attempts at molecular analysis using different primers. Sequence analysis showed that all the isolates belonged to G. duodenalis assemblage E. Comparison with SSU rRNA sequences available on GenBank showed 100% sequence identity with sequences of isolates previously recognized from calves (KT922263) and lambs (KT922264) in Ethiopia. Phylogenetic relationship analysis identified G. duodenalis assemblage E (Fig. 2). Two different subtypes were identified at the bg locus, which showed 100% similarity to strains isolated from sheep (GQ337972) in Norway and from calves (KT922247) in Ethiopia, respectively. At the gdh locus, the assemblage E shared 99% similarity with a yak isolate (KP334146) in China.

Discussion
Results obtained from the 662 fecal specimens by both microscopic examination and PCR concurred. PCR is a tool of high sensitivity and specificity [40], particularly for identifying morphologically indistinguishable parasites such as species of Cryptosporidium and assemblages of Giardia, and for detecting their genetic variation. In contrast, Sheather's sugar flotation technique and Lugol's iodine staining are routine diagnostics. All the PCRpositive samples found positive by microscopy, and vice versa, may be due to sufficient oocyst and cyst concentrations in fecal specimens.
The Cryptosporidium prevalence of 6.8% in this study was close to the prevalence of 7.84% (25/319, 95% CI: 4. 9-10.8%) reported in a study in Japan [9], but higher than that detected in Zhengzhou, China (1.61%, 2/124, 95% CI: 0-3.9%) [14]. It is difficult to explain the discrepancies in the prevalences of Cryptosporidium spp. among different studies because prevalences are affected Studies of Cryptosporidium spp. in cervids have been conducted in several countries and 11 Cryptosporidium species/genotypes have been detected (Table 2). In this study, Cryptosporidium deer genotype was the most frequently detected. Small numbers of the Cryptosporidium deer genotype have been detected in red and roe deer in the UK [12,15] and in white-tailed deer in the Czech Republic [10], whereas in sika deer in Japan [9] and white-tailed deer in the USA [13], Cryptosporidium deer genotype was the only genotype detected. It appears to be a host-adapted genotype, which has so far only been identified in deer.
Cryptosporidium parvum is one of the two most common Cryptosporidium species in humans [41]. We detected 11 C. parvum isolates, making C. parvum the second-largest cause of infection in cervids in this study. Cryptosporidium parvum infections were observed in red deer and roe deer in the UK [15], red deer in the Czech Republic [7], and white-tailed and black-tailed deer in the USA [4,11]. We identified zoonotic subtypes IIaA15G2R2 and IIdA19G1 based on a sequence analysis of the gp60 gene. Thus far, at least 14 C. parvum subtype families (IIa-IIi and IIk-IIo) have been found [2,42]. IIa is the predominant subtype family in animals and humans worldwide, and IId is another major zoonotic subtype family reported in Europe, Asia, Egypt and Australia [43]. In China, most C. parvum isolates belong to subtype IId, including IIdA15G1 found in rodents, cattle and yaks [44][45][46], IIdA18G1 found in yaks [46], and IIdA19G1 found in cattle, humans, goats, yaks and urban wastewater [46][47][48][49][50][51]. In contrast, only a few IIa isolates have been detected in yaks and goats [48,52], and IIc has been found in monkeys [53]. Subtype IIaA15G2R2 has previously been reported in humans, calves and water in the USA and Canada [54][55][56]. The above findings that the same gp60 gene sequences of the C. parvum isolates have been found in humans and animals, suggest that deer infected with C. parvum in the areas we investigated might pose a threat to local people and animals by shedding oocysts in their feces, thereby contaminating the environment and food sources.
Cryptosporidium ubiquitum, previously called the Cryptosporidium cervine genotype, has been detected in sika deer in China [14], white-tailed deer in the USA [11], swamp deer in Nepal [5], roe deer in the UK [12], red deer in the Czech Republic [10] and deer in Australia [16]. In this study, three C. ubiquitum isolates were identified as zoonotic XIIa subtype 2, which has been found in domestic and wild ruminants, as well as in humans [34]. Despite its low prevalence in this study, C. ubiquitum may be a pathogen of public health concern in this area given its broad host range including rodents, carnivores, primates, domestic and wild ruminants, as well as humans in the UK, Slovenia, the USA, Canada, Spain and New Zealand [34].
Although Cryptosporidium muris is typically a parasite of mice and rats, it has a wide range of host species, including rodents, cats, marsupials (bilbies) and other mammals [57][58][59]. Cryptosporidium muris has also been identified in human cryptosporidiosis cases in many countries, including Thailand, Iran, India, Indonesia, Saudi Arabia, Kenya, Peru and France [60][61][62][63][64][65][66][67], and found in red and white-tailed deer in the Czech Republic [10]. Cryptosporidium suis naturally infects pigs worldwide, but has also been found in cattle, rodents, humans and chimpanzees [68], whereas the C. suis-like genotype has been reported in cattle and rodents, as well as in humans [12,69,70]. In this study, only one C. muris isolate and one C. suis-like isolate were identified. The Table 2 Cryptosporidium species/genotypes in cervids Species/genotype Host (Locality) Reference
At the molecular level, this study is the first to characterize G. duodenalis from cervids in China, and only assemblage E was identified. Worldwide, less than 100 G. duodenalis isolates from cervids have been analyzed, among which assemblages A, B, D and E were detected [6,13,16,[19][20][21][22][23][24][25][26][27][28][29]. Assemblage A, including sub-assemblages A-I (infecting most animals), A-II (mainly found in humans), and A-III (mainly infecting wild ruminants), have previously been reported from red deer, roe deer, fallow deer, reindeer, white-tailed deer and moose in the USA, Croatia, Spain, Poland, Italy, Norway and Australia [6,13,16,19,[21][22][23][24][25][26][27]29]. Assemblage A is most frequently found assemblage in cervids, while assemblages B and D have only been found in red and roe deer in eastern Poland and Croatia, respectively [19,28]. In Sweden assemblage E has been detected in a fallow deer [24]. Giardia duodenalis assemblage E has a wide distribution in domestic mammals, including cattle, water buffaloes, sheep, goats and pigs, and is the predominant assemblage found in these animals in the USA, Europe and Australia [1]. However, assemblage E has rarely been identified in wild hoofed animals, and thus may reflect an adaption to these animals following domestication [19]. Moreover, assemblage E has also been found in NHPs from western Uganda and China [79,80] and in humans from Egypt, Brazil and Australia, suggesting zoonotic transmission of assemblage E [81][82][83][84][85].
Although assemblage E was detected in low numbers in this study, it is still important to understand the public health risk posed by the Giardia species and assemblages infection in cervids in the region.

Conclusions
The 662 samples collected from red deer, Père David's deer, and sika deer were screened for the presence of Cryptosporidium and G. duodenalis. C. parvum, C. andersoni, C. ubiquitum, C. muris, Cryptosporidium suis-like genotype, Cryptosporidium deer genotype and G. duodenalis assemblage E were detected in this study. This is the first study to report C. andersoni, as well as C. parvum zoonotic subtypes IIaA15G2R2 and IIdA19G1 in cervids. Deer farming has become an important component of China's animal breeding industry. Farming increases the potential contact between deer and humans, and intensifies the numbers of animals, which potentially increases the numbers of shed (oo) cysts in the environment. Given that we have detected human pathogens in deer in China, further investigations into the transmission dynamics of these pathogens would be warranted.