Genotyping Cryptosporidium andersoni in cattle in Shaanxi Province, Northwestern China.

The present study examined the prevalence and genotypes of Cryptosporidium andersoni in cattle in Shaanxi province, China. A total of 2071 fecal samples (847 from Qinchuan cattle and 1224 from dairy cattle) were examined for the presence of Cryptosporidium oocysts, and 70 samples (3.4%) were C. andersoni-positive and those positive samples were identified by PCR amplification of the small subunit ribosomal RNA (SSU rRNA) and the Cryptosporidium oocyst wall protein (COWP) genes. C. andersoni was the only species found in the examined cattle in this province. Fifty-seven C. andersoni isolates were characterized into 5 MLST subtypes using multilocus sequence typing analysis, including a new subtype in the native beef breed Qinchuan cattle. All of these C. andersoni isolates presented a clonal genetic structure. These findings provide new insights into the genetic structure of C. andersoni isolates in Shaanxi province and basic data of Cryptosporidium prevalence status, which in turn have implications for controlling cryptosporidiosis in this province.

Introduction

Cryptosporidium spp. infect a wide range of hosts including humans and cattle [1], [2]. Previous studies around the world have shown that cattle is the most common species of mammals known to be infected with Cryptosporidium spp., and is the main source of human infection with Cryptosporidium [2]–[4]. Four Cryptosporidium species, namely C. parvum, C. andersoni, C. bovis and C. ryanae, can infect cattle and cause bovine cryptosporidiosis [5]. Among them, C. andersoni has been identified to be the predominant species responsible for cattle infection in China [6]. There are subclinical signs in cattle infected with C. andersoni associating with poor weight gain and reduce of milk yield [7], [8]. C. andesoni also has been isolated from humans with diarrhoea in England [9] and paediatric patients in Malawi [10].

Traditionally, identification of Cryptosporidium spp. is based on morphologic examination, mainly using Sheather's sugar flotation technique [11] and modified acid-fast staining method [12], but this approach is not reliable for delineating Cryptosporidium species because of their morphologic similarities [13]. Alternatively, molecular tools are useful for the accurate identification of Cryptosporidium species and better understanding of population genetics of Cryptosporidium, which have important implications for studying their pathogenesis and clinical presentations [14], [15]. The small subunit ribosomal RNA (SSU rRNA) gene [16] and the Cryptosporidium oocyst wall protein (COWP) gene [17]–[19] have been used successfully as genetic markers for the identification of Cryptosporidium species and genotypes in hosts and environmental (water and food) samples. But genotyping tools have limitations in some epidemiologic investigations as a result of the low resolution power of these loci [17]. Consequently, several subtyping tools have been developed for molecular epidemiologic studies of Cryptosporidium spp. Multilocus sequence typing (MLST), based on both length polymorphism and single nucleotide polymorphism (SNP), is a high-resolution typing tool that can accurately describe genetic diversity of parasites [20]. Recently, C. muris and C. andersoni subtypes were successfully analyzed by MLST targeting microsatellite and minisatellite sequences [21], [22].

In China, Cryptosporidium infection has been reported in dairy cattle in Guangxi [23], Anhui [24], Henan [6], [25] and Qinghai provinces [26], and in beef cattle in Anhui [27], Qinghai [26], Inner Mongolia [28] and other provinces. However, these studies mainly focused on prevalence and identification of Cryptosporidium species. Qinchuan cattle, originated in Guanzhong Plain in Shaanxi province, is a famous native beef cattle breed in China, which now has been introduced to more than 20 provinces in China. However, there is no publication about Cryptosporidium infection in Qinchuan cattle except prevalence of other intestinal parasites reported by our group [29]. Here, the prevalence of C. andersoni infection in Qinchuan cattle in Shaanxi province was investigated and compared with that in dairy cattle in this province. The subtypes of C. andersoni in Qinchuan cattle and dairy cattle in this province were also characterized using MLST.

Materials and Methods

Ethics Statement

The performance of this study was strictly according to the recommendations of the Guide for the Care and Use of Laboratory Animals of the Ministry of Health, China, and our protocol was reviewed and approved by the Research Ethics Committee of Northwest A&F University. All the fecal samples were collected from animals after the permission of farm owners, with no specific permits being required by the authority for the collection of fecal samples.

Specimen Collection and Examination

Between November 2010 and May 2012, a total of 2071 fecal samples from intensively reared pre-weaned calves (3 weeks −<3 months old), post-weaned calves (3–11 months old), heifers (1–2 years old), and adult cattle (>2 years old) were obtained directly from the rectum of each animal using sterile disposable gloves, then placed in clean plastic bags labeled with the animal’s breed, age, and geographical origin (Table 1) in Shaanxi province, China. These samples were then examined by Sheather's sugar flotation technique and microscopy at 400×magnification [11]. Cryptosporidium-like samples were kept in 2.5% potassium dichromate solution before DNA extraction.

Table 1

Factors associated with prevalence of Cryptosporidium andersoni genotypes in cattle in Shaanxi province, Northwestern China.

Factors		Qinchuan cattle						Dairy cattle
		Sample size	No. positive (%)	No. subtypes*				Sample size	No. positive (%)	No. subtypes*
				A	B	C	D			A	E
Collection site	Yangling district	561	33 (5.88)	20	3	1	1	445	11 (2.47)	1	6
	Tongchuan city	84	0	0	0	0	0	188	8 (4.26)	0	8
	Mei county	138	0	0	0	0	0	78	2 (2.56)	0	2
	Qian county	34	5 (14.71)	5	0	0	0	143	6 (4.20)	0	6
	Dali county	0	0	0	0	0	0	86	0	0	0
	Shenmu county	0	0	0	0	0	0	14	0	0	0
	Xi 'an city	30	0	0	0	0	0	270	5 (1.85)	0	4
	Total	847	38 (4.49)	25	3	1	1	1224	32 (2.61)	1	26
Age group	<3 month	14	3 (21.43)	3	0	0	0	119	0	0	0
	3–11 months	67	17 (25.37)	11	1	1	0	250	11 (4.40)	0	10
	1–2 year	447	14 (3.13)	9	2	0	1	224	8 (3.57)	0	7
	>2 year	319	4 (1.25)	2	0	0	0	631	13 (2.06)	1	9
	Total	847	38 (4.49)	25	3	1	1	1224	32 (2.61)	1	26

Note: *A, B, C, D, E respresent subtypes A4, A4, A4, A1; A2, A4, A4, A1; A2, A4,A2, A1; A4, A4, A2, A1; A1, A4, A4, A1, respectively.

DNA Isolation and PCR Amplification

The positive samples were washed extensively in distilled water to remove the potassium dichromate solution. DNA was extracted from each microscopically positive sample using the E.Z.N.A.® Stool DNA Kit (OMEGA) according to the manufacturer’s instructions, and stored at −20°C until further processed. Cryptosporidium species and genotypes were determined by nested PCR amplification of the small subunit ribosomal RNA (SSU rRNA) gene [16] and by single PCR amplification of the Cryptosporidium oocyst wall protein (COWP) gene [30]. Subtyping was achieved by amplifying the minisatellite targets (four loci, namely CM-MS1 coding for hypothetical protein; CM-MS2 coding for 90 kDa heat shock protein; CM-MS3 coding for hypothetical protein; CM-MS16 coding for leucine rich repeat family protein) according to Feng et al [21] and Wang et al [22]. DNA of Cryptosporidium andersoni was used as the positive control for each target gene-based PCR analyses, and samples without DNA and host (cattle) DNA were included in each amplification run to exclude contamination. The primers used in PCR analysis of all gene targets, annealing temperatures, and sizes of the expected PCR products are listed in Table 2. Amplification products were examined by 1.5% agarose gel electrophoresis and stained with ethidium bromide.

Table 2

Primers used in the study, annealing temperatures used in the PCR and expected sizes of the PCR products.

Gene	Primer	Sequence (5'–3')	Annealing temperature (°C)	Fragment length (bp)	References
SSU rRNA	F1	CCCATTTCCTTCGAAACAGGA	55	830	[14]
	R1	TTCTAGAGCTAATACATGCG
	F2	AAGGAGTAAGGAACAACCTCCA	58
	R2	GGAAGGGTTGTATTATTAGATAAAG
COWP	F	TGTAGCGTTTTCTCCACCTGATAAA	56	450	[27]
	R	GTTGTGTTGATGCGGTGTTC
CM-MS1	F1	ACCATCTAGAGATAACGAGCGA	55	550	[19]
	R1	GAATCAGAAGATGAGCGACAA
	F2	CGTGATAGTGGGTATGAATTGGACA	55
	R2	CGACTGCGATACTCACGTCCT
CM-MS2	F1	TTGCAACTGTACCTAAATTAGTA	55	457	[19]
	R1	GTGAGACTTCTGGGGTCCTGA
	F2	TCATGACGCGTCATACCAACA	52
	R2	ACTTAGACAGTTCTATGCTGA
CM-MS3	F1	AACCAAGTGAATCACGAACTT	55	536	[19]
	R1	TCAAGTACAGCAGTCTATTGCTT
	F2	GCAATATCTTCGACGATCCCA	55
	R2	ATGGGAATAATTCTTCATCATCAA
CM-MS16	F1	GAAGAGGTCGAAGTTAAGCTA	50	597	[19]
	R1	GACAATCATCTAAATCGTGTT
	F2	AAGTTTCATCTAGGTACACTAAGA	55
	R2	CACTACCTAATCTCGTGTACTT

Statistical Analysis and Sequence Analysis

The prevalence of Cryptosporidium oocysts and the factors of breed, collection origins and ages were evaluated using Regression Analysis in Statistic Package for Social Science (SPSS) for Windows with 95% confidence intervals (CI). Probability levels (P) of <0.05 were regarded as statistically significant.

The positive PCR products were sent to Shanghai Sangon Biological Engineering Biotechnology Company for sequencing using ABI 377 automated DNA sequencer (BigDye Terminator Chemistry) to identify the species/genotype and subtype. The Cryptosporidium nucleotide sequences obtained were aligned with reference sequences from the GenBank™ database using the BLAST (http://www.ncbi.nlm.nih.gov) and computer program Clustal X 1.83 [31]. Phylogenetic analysis based on SSU rRNA and COWP gene sequences were conducted to identify species of Cryptosporidium isolates in the present study. Neighbor-joining (NJ) method was carried out using Phylip 3.64 [32] with the Kimura two-parameter model. The consensus tree was obtained after bootstrap analysis, with 1000 replications. NJ analysis based on minisatellite sequences was also used to study the relationships of C. andersoni isolates with other Cryptosporidium by retrieving the reported sequences available in the GenBank™ by Feng et al. [21] and Wang et al. [22]. The C. andersoni subtypes were named according to the repeat characteristics of minisatellite repeats in four genetic loci by Feng et al. [21] (Table 3). The genetic diversity of C. andersoni was analyzed using DnaSP version 5.10.01 (http://www.ub.edu/dansp/). Linkage disequilibrium was tested using LIAN version 3.5 (http://adenine.biz.fh-weihenstephan.de/cgi-bin/lian/lian.cgi.pl) by a parametric method for four microsatellite/minisatellite loci. STRUCTURE version 2.3.3 was used to describe the population substructure of C. andersoni in allelic variation by K-means partitional clustering and the admixture model.

Table 3

The repeat characteristics of minisatellite repeats at four genetic loci.

Locus	Targeted repeat
CM-MS1	(TAAAGGGCGAGA)3 and (GAACGAGATAGG)12–17
CM-MS2	(CCATACCTC)10–11
CM-MS3	(TGTTGGTGTTGCTGT)2 and (TGCTGCAGCTGC)2–3
CM-MS16	(CTTCTTCAT)12,14

Nucleotide Sequence Accession Number

Sequences generated in the present study have been deposited in the GenBank database under accession numbers KC580754-KC580823 (for 18S rRNA), KC580824-KC580893 (for COWP), KC580894-KC580958 (for MS1), KC580959-KC581022 (for MS2), KC581023-KC581090 (for MS3) and KC581091-KC581158 (for MS16).

Results and Discussion

A total of 2071 fecal samples were collected from Qinchuan cattle (847) and dairy cattle (1224) between November 2010 and May 2012, and were examined for the presence of Cryptosporidium oocysts. The results are summarized in Table 1. Of these samples, the prevalence of Cryptosporidium infection in Qinchuan cattle in Shaanxi province was 4.49% (38/847), which was lower than that in Anhui province [27], Qinghai province [26] in China and in Korea [33]. The prevalence of Cryptosporidium in dairy cattle in Shaanxi province (32/1224, 2.61%) was lower than that in Guangxi [23], Anhui [24], Henan [6], [25], Qinghai [26] in China and in rural areas in Korea [33] and in Mongolia [34]. These differences may be related to the different management systems, the timing of specimen collection and the ecological conditions. The results of the SPSS analysis showed that the prevalence of Cryptosporidium in Qinchuan cattle was higher than that in dairy cattle in Shaanxi province, but with no significant differences (P>0.05). This indicates that the breed of cattle is not associated with the prevalence of Cryptosporidium infection. For Qinchuan cattle, the unique beef cattle in China, the highest Cryptosporidium prevalence (14.71%) was found in Qian county, but that of dairy cattle was observed in Tongchuan city, indicating no obvious association with the locations (P>0.05). These differences may be attributed to the number of samples examined and the timing of sample collection. An age-related difference in Cryptosporidium prevalence was observed in this study (Table 1, P<0.05). Compared with other age groups, the highest prevalence was noticed in cattle of 3–11 month old (25.37% for Qinchuan cattle, 4.40% for dairy cattle). This was different from results of previous studies that the overall prevalence of Cryptosporidium infection had negative relation with age [17], [35], [36].

All microscopically positive samples were confirmed to be infected with C. andersoni by characterizations of the SSU rRNA and COWP genes loci. A BLAST similarity search against NCBI nucleotide sequence database indicated that all the obtained sequences of the SSU rRNA and COWP genes have high identity values (>99%) with C. andersoni (GenBank accession numbers HQ009808 and AB514044 for the SSU rRNA and COWP gene fragments, respectively). Phylogenetic analysis using NJ based on the SSU rRNA and COWP gene sequences, respectively, showed that all Cryptosporidium positive samples were clustered in the clade C. andersoni (Data not shown). These results demonstrated that all cattle-derived Cryptosporidium isolates in the present study represented C. andersoni. This finding was different from that of previous studies in other geographical origins in China [24], [27], [37], India [37], United States [38]–[40] and Denmark [41], where they showed that C. parvum was the predominant species in pre-weaned calves, whereas C. bovis and C. ryanae usually infected post-weaned calves (frequently in yearlings and adult cattle). However, the results of our study are similar to that of Wang et al. [6] who concluded that C. andersoni was the predominant species in post-weaned and adult dairy cattle in China. In the present study, all Cryptosporidium isolates were identified as C. andersoni, which may be related with cattle ages in that sampled cattle were usually older than 3 months. Santín et al. [39] described that cryptosporidiosis in calves infects usually in the first 2 weeks of life. Therefore, the species and prevalence of Cryptosporidium infection in cattle before 2-week-old in this region should be investigated in further studies.

Minor intragenotypic variations sometimes can be identified by subtyping tools. At all four minisatellite loci, namely MS1, MS2, MS3 and MS16, a total of 65, 64, 68 and 68 DNA preparations were successfully amplified from each locus, respectively. For Cryptosporidium andersoni isolates, 3, 1, 2, and 1 haplotypes were identified at the MS1, MS2, MS3 and MS16 loci, respectively (Figure 1). Two prevalent subtypes were found in the present study: one named subtype A4, A4, A4, A1 herein being derived from 1 dairy cattle and 25 Qinchuan cattle, and the other subtype A1, A4, A4, A1 being unique to dairy cattle. The MLST subtype A2, A4, A4, A1, subtype A2, A4, A2, A1 and subtype A4, A4, A2, A1 were only found in Qinchuan cattle, with 3, 1, 1 isolates for each subtype, respectively (Table 1). Compared with results of Feng et al [21] and Wang et al [22], a new C. andersoni MLST subtype (A4, A4, A2, A1) in Qinchuan cattle was identified. In addition, the predominate subtype was subtype A1, A4, A4, A1 for dairy cattle in Shaanxi province, which was different from subtype A4, A4, A4, A1 in other areas of China [22]. These differences may be related to the number of examined specimens and geographic separation.

Figure 1

Phylogenetic relationship of C. andersoni subtypes based on MS1, MS2, MS3 and MS16 sequences.

The trees were reconstructed using respective site sequences of C. muris as outgroup and assessed by a neighbor-joining (NJ) analysis of the nucleotide sequences with distance calculated by the Kimura two-parameter model. CM-MS1, CM-MS2, CM-MS3, and CM-MS4 are reference sequences from the whole-genome sequencing project.

To determine the presence of the clonal or epidemic genetic structure for C. andersoni from cattle in Shaanxi province, the linkage disequilibrium (LD) analysis were performed according to Wang et al. [22] (Table 4). Samples amplified successfully at every locus were included in LD analysis. The standardized index of association (I) was above zero and the pairwise variance (V) was greater than the 95% confidence limitation (L), indicating the presence of LD and the clonal genetic structure of C. andersoni in this province. These results were consistent with C. andersoni isolates from other geographical origins in China [22].

Table 4

Analysis of linkage disequilibrium in C. andersoni from cattle.

Area	No. completely typed	I SA	VD	L	P value
China	99	0.1737	0.7261	0.5256	4.23×10−33
Shaanxi province	57	0.0314	0.3431	0.3251	2.55×10−09

I =  standardized index of association, V =  the pairwise variance, L = 95% critical value.

The C. andersoni samples in China including isolates from Shaanxi province and sequences available in GenBank™ were used in cluster analysis. All the samples formed three clusters (Figure 2). Samples in most provinces except Heilongjiang, Shanxi and Jilin provinces were dispersed in different clusters. All the samples in Shaanxi province were positioned in three clusters, with Qinchuan cattle in cluster 2 and 3, and dairy cattle in cluster 1 and 3. These results were slightly different from finding of Wang et al. [22] in that only two clusters were identified in China, possibly due to the number of sampled specimens and geographical origins. The MLST subtype A1, A4, A4, A1 (n = 32) and A4, A4, A4, A1 (n = 50), the two most common subtypes in cattle in China, located into cluster 1 and cluster 3, respectively. The new MLST subtype A4, A4, A2, A1 was identified within Cluster 2.

Figure 2

Population structure inferred by Bayesian clustering using multiocus information.

A thin vertical line represents each individual, and the geographic regions lie in the bottom. HLJ = Heilongjiang province; SHX = Shanxi province; SX = Shaanxi province; HE = Henan province; JL = Jilin province; SC = Sichuan province; GX = Guangxi province; B = beef cattle; D = dairy cattle; Q = Qinchuan cattle.

Subtyping tools have proven useful for understanding the biologic characteristics of Cryptosporidium spp., identifying virulence and clinical presentations among different subtypes, and epidemiological investigation of Cryptosporidium species [2], [21], [42]. The 60 kDa glycoprotein (GP 60) gene was the initial subtyping target widely used for studying C. parvum and C. hominis [41], [42]. However, single locus may be deviated and some genotypes may be missed because of selection of locus and its length. The multilocus typing (MLT) and MLST have been used to genotype C. parvum and C. hominis [20], [43], [44]. The MLT method may identify the subtype of multiple loci, but sometimes can miss the single nucleotide polymorphisms (SNP), unable to detect some subtypes. Compared with MLT, the MLST tool, which is dependent principally on the genetic heterogeneity by DNA sequencing of the amplified PCR products, can directly and accurately analyze the genotypes and subtypes of Cryptosporidium spp. [20]. C. andersoni was the preponderant Cryptosporidium species in cattle in China [6], resulting in the pathological lesions of the gastric glands and the gastric mucosa, poor weight gain and falling of milk yield [7], [8], [45], [46]. In 2011, a MLST technique for subtyping C. andersoni was developed by Feng et al. [21], and used later by Wang et al. [22]. This method can identify successfully the subtypes of C. andersoni using the microsatellite and minisatellite markers.

In conclusion, using MLST, the present study identified 5 MLST subtypes among 57 C. andersoni-positive specimens from dairy cattle and Qinchuan cattle in Shaanxi province, including a new subtype A4, A4, A2, A1 in the native beef breed Qinchuan cattle. The subtypes A1, A4, A4, A1 and A4, A4, A4, A1 were the prevalent subtypes in dairy cattle and Qinchuan cattle, respectively. C. andersoni in cattle in Shaanxi province presented a clonal genetic structure. These findings showed new insights into the genetic structure of C. andersoni isolates in Shaanxi province, Northwestern China. The accurate genotyping of C. andersoni isolates provided valuable basic data for developing strategies to control C. andersoni infection in cattle and evaluate risk of Cryptosporidium infection to humans. These findings should have implications for controlling cattle cryptosporidiosis in this province as well as in China.