Cryptosporidium in asymptomatic children in Southern Xinjiang, China and the potential of zoonotic transmission

Background: Cryptosporidium is a primary cause of diarrhea in children globally. However, there is limited information on the prevalence and genetic characterization of Cryptosporidium in children in Xinjiang, China. This study aimed to assess the genetic characteristics and epidemiological status of Cryptosporidium in asymptomatic children in Southern Xinjiang, China. Methods: A total of 609 fecal samples were collected from kindergartners aged 2-6 y from 11 counties of Southern Xinjiang, China. We used nested PCR amplication of partial SSU rDNA gene to screen the samples for Cryptosporidium spp. The isolates containing C. parvum and C. hominis were further subtyped by a 60-kDa glycoprotein (gp60). We used MEGA7 to construct a phylogenetic tree to study the genetic relationship between the gp60 subtypes of these two species via the Maximum Likelihood method based on the Tamura-Nei model. Results: Only 1.3% (8/609) of asymptomatic children were conrmed as infected with Cryptosporidium with 2.0% (6/299) infection rate in boys and 0.6% (2/310) infection rate in girls. Three Cryptosporidium species were identied including C. felis (37.5%; 3/8), C. hominis (37.5%; 3/8) and C. parvum (25.0%; 2/8). Three C. hominis subtypes (IbA9G3, IdA14 and IfA12G1) and two C. parvum subtypes (IIdA14G1 and IIdA15G1) were also found. Conclusions: This study was the rst to identify the presence of cryptosporidium in asymptomatic children in Southern Xinjiang, China. The presence of zoonotic C. parvum subtypes IIdA14G1 and IIdA15G1 indicates possible crossspecies transmission of Cryptosporidium between children and animals.


Background
Cryptosporidium spp. are obligate intracellular protozoan parasites, which can be found in all classes of vertebrates [1]. Cryptosporidium infects the gastric and intestinal epithelial cells leading to self-limiting diarrhea or asymptomatic symptoms in healthy people. However, infection can be fatal amongst immunocompromised individuals [2]. Cryptosporidium infection leads to stunted growth and weight loss, among other features. Cryptosporidiosis accounts for approximately 30.0%-50.0% global deaths per-year in children < 5 y of age and is the second leading cause of diarrhea-related fatalities in children after rotavirus [3]. Cryptosporidium can be transmitted via various routes, including direct contact with infected individuals/animals, or uptake of contaminated food/water [4].
Cryptosporidiosis is a complex genus in which a total of more than 40 species and over 40 genotypes (yet to be assigned a species) have been formally described, all with high genetic diversity [5,6]. Molecular studies have demonstrated that human cases of cryptosporidiosis are caused by minimum 20 different species. The dominant subtypes include C. hominis and C. parvum followed by Cryptosporidium meleagridis, C. ubiquitum and C. viatorum. Meanwhile, C. canis, C. felis, C. cuniculus, and C. andersoni, in addition to three Cryptosporidium genotypes (horse, chipmunk I and skunk) cause disease in humans and are zoonotic [7]. Various gp60-targeted subtyping tools have been designed to assess the importance of zoonotic transmission, using comparative analyses of human and animal samples [7,8].
In China, recent studies have focused on Cryptosporidium in children in Central and Southern China but molecular detection methods have been used in no more than three reported studies [9]. Four species of Cryptosporidium, including C. meleagridis, C. hominis, C. canis, and C. felis have been identi ed in children from China. Amongst them, C. hominis is the predominant species and four subtype families including Ia, Ib, Id and Ig, are present within the C. hominis isolates [9]. Nevertheless, the species and subtypes of Cryptosporidium in children in China remain poorly de ned, particularly in speci c provinces or regions such as in Xinjiang Uygur Autonomous Region (XUAR), where a number of zoonotic species of Cryptosporidium are distributed in local animals. We therefore propose that the scale of infection by Cryptosporidium in children has been underestimated in China. This population poses a signi cant threat if the infections continue to be ignored.
In this study, we investigated the epidemiology of Cryptosporidium infection in children aged 1-6 years in XUAR, North-West China, and tracked possible sources of infection and assessed the crossspecies spread of Cryptosporidium spp. between children and animals using genotyping and subtyping tools. Such knowledge can provide insight into control and prevention strategies against infections for this pathogen.

Sample Collection
Between August 2017 to January 2019, 609 fecal samples were collected from kindergarteners aged 2-6 y from 11 counties of Southern Xinjiang, China ( Table 1). The parents/guardians who provided consent on behalf of their children were trained with relevant guidelines by the staff and were provided a labeled plastic fecal collector marked with the date of collection and patient identity information (age and sex).
None of the kids experienced diarrhea during the sampling period. After collection, samples were stored at 4°C.

DNA Extraction and PCR Ampli cation
Approximately 200 mg of each fecal sample was extract the genomic DNA used the E.Z.N.A® stool DNA kit, and stored the extracted DNA at -20°C before PCR analysis.
We used nested PCR ampli cation of an approximately 830 bp partial SSU rDNA gene fragment to screen the samples for the presence of Cryptosporidium spp. We used primers described previously in a study by Xiao et al. [10]. Additionally, we used nested PCR to amplify an 850 bp gene fragment of a 60-kDa glycoprotein (gp60) to subtype Cryptosporidium positive isolates. This study was done using primers described previously in a study by Alves et al. [11]. Also, dH 2 O was included as a negative control and DNA from chicken-derived C. bailey was used as the positive control.

Sequence and Phylogenetic Analyses
We used GENEWIZ to bidirectionally sequence positive secondary PCR products. Sequencing of PCR products was performed to con rm sequence accuracy during DNA preparation. We used DNASTAR Lasergene EditSeq v7.1.0 (http://www.dnastar.com/) to edit the derived sequences and used Clustal X v2.1 (http://www.clustal.org/) to align them with reference sequences downloaded from GenBank.
A phylogenetic tree was constructed using MEGA7 using the Maximum Likelihood method based on the Tamura-Nei model to evaluate the genetic relationship between the gp60 subtypes of C. hominis and C. parvum.

Results
Prevalence of Cryptosporidium Genetic characterizations at the SSU rRNA gene Three Cryptosporidium species were identi ed, including C. felis (37.5%; 3/8), C. hominis (37.5%; 3/8), and C. parvum (25.0%; 2/8). The three, three and two SSU rDNA sequences of C. felis, C. hominis and C. parvum were identical, respectively and showed 100% similarity with the sequences in Genbank: AF159113 for C. felis, MK990042 for C. hominis and MK982463 for C. parvum. Amongst them, C. hominis was identi ed at three collection sites (Payzawat, Poskam and Yopurga), C. felis in Yopurga and Yecheng, and C. parvum only in Kuqa. C. felis was only identi ed in boys, whilst C. hominis and C. parvum were both identi ed in both boys and girls.
Subtyping of C. hominis and C. parvum of the gp60 gene All six C. hominis and C. parvum isolates were successfully ampli ed at the gp60 gene. Sequence analysis of three C. hominis isolates showed three subtypes, termed IbA9G3 (1), IdA14 (1) and IfA12G1 (1). The sequences of two C. parvum isolates identi ed in the study showed 100% similarity to those of IIdA14G1 and IIdA15G1, respectively.

Discussion
Almost all countries in the world except Antarctica has been recorded to have a high incidence of cryptosporidiosis in humans [12].  [9].
Other than age, sensitivity and speci city of the detection techniques, host health, and living criteria might also affect disease prevalence [14]. The estimated prevalence was high in those from low-income countries, individuals with gastrointestinal symptoms and residents not living in urban areas [12]. In developing countries, the prevalence of Cryptosporidium in children under ve years with diarrhea was 27.4% in India (ELISA) [15], 27.8% and 32.0% in Ghana and Guatemala (microscopic analysis), respectively [16,17], and 25.0% and 10.4% in Uganda and Tanzania (PCR), respectively [18,19]. In China, signi cantly higher prevalence of Cryptosporidium was observed in rural population (1.8%-12.9%) than in urban population (0-3.7%) [9]. In the USA, cryptosporidiosis was mainly observed in children between 1-9 y of age; its high incidence was associated with recreational water use and communal swimming venues, resulting in peak infection in the summer [20].
There was a limited of information on the prevalence of asymptomatic infection. Cryptosporidium was detected in 20 (7.2%) of 276 asymptomatic aboriginal children living in villages in Malaysia [21]. In Jeddah, South Africa, almost 4.7% Cryptosporidium-positive cases were asymptomatic compared with 32.0% cases with diarrhea from pediatric clinics [22]. This study is the rst study to explore the infection of Cryptosporidium in asymptomatic children in China with low infection rates (1.3%; 8/609).
Three Cryptosporidium species were identi ed, including C. felis (37.5%; 3/8), C. hominis (37.5%; 3/8) and C. parvum (25.0%; 2/8). C. felis oocysts were rst identi ed in cat feces and were therefore considered a host-adapted species [23]. Human C. felis infections have shown more recent prevalence in developing countries including China, where it was found to cause Cryptosporidium infection in minimum eight human cases [7]. In addition to cats and humans, C. felis has also been found in other animals including non-human primates, calves, horses, and foxes, suggestive of a possible risk of zoonotic transmission [24]. C. hominis is a pathogenic species commonly found in humans and natural infections have been reported in nonhuman primates, cattle, dugong, marsupials, and goats [25]. The major Cryptosporidium species in Chinese population is C. hominis, which has been shown to cause approximately 48.3% (127/263) of human cases [9]. Recent studies have identi ed C. hominis as the dominant Cryptosporidium species in other species in China, such as horses and donkeys [26][27][28]. This supports the theory that these animals have been infected by human feces. C. parvum has been shown to have the highest zoonotic potential among the Cryptosporidium species. In China, 16.7% (44/263) of human cases are caused by C. parvum [9]. In this study, we identi ed C. parvum in yaks, sheep, goats, golden takins, horses, cattle, and donkeys, indicating zoonotic potential.
Subtyping tools based on the analysis of the gp60 gene have been developed for human-pathogenic Cryptosporidium spp. to track potential sources of infection [7]. In this study, three subtype families (Ib, Id, and If) were identi ed for C. hominis, composed of IdA14, IfA12G1 and IbA9G3. Elevated frequency of all three subtype families (Ib, Id, and If) were identi ed in rhesus monkeys in Guizhou, where humans and animals are known to closely interact [29]. This study is the rst work that is performed on humans with the discovery that those subtypes can spread between humans and monkeys.
There is a limited availability of data on the subtypes of C. parvum in humans in China compared to C. hominis. Only two subtypes (IIa and IId) have been identi ed and only IIdA19G1 was found [30]. In this study, subtyping successfully identi ed two C. parvum isolates of the IId family; IIdA15G1 and IIdA14G1. These were the dominant groups in animals in China indicating that these animals may represent an important source of zoonotic Cryptosporidium in China [7].

Conclusions
This study was the rst to identify the presence of asymptomatic cryptosporidium infections in children in Southern Xinjiang China. Three species of Cryptospridium including C. felis, C. hominis and C. parvum were found. Meanwhile, three C. hominis subtypes (IdA14, IfA12G1 and IbA9G3) and two C. parvum subtypes (IIdA14G1 and IIdA15G1) were identi ed. The presence of common zoonotic C. parvum subtypes highlights the possible cross-species transmission of Cryptosporidium between children and animals.
Abbreviations gp60: 60-kDa glycoprotein; SSU: small subunit Declarations Figure 1 Molecular Phylogenetic analysis. The Tamura-Nei model-based Maximum Likelihood method was used to determine the evolutionary history. Here is the tree with the highest log likelihood with branches showing the % of trees with clusters of relevant taxa. The NJ and BioNJ algorithms were used to obtain the initial tree(s) for heuristic searches using the MCL method. We selected the topology with a superior log likelihood value. The tree has been drawn to scale, and the number of substitutions per-site indicate