Presence and molecular characterization of Cryptosporidium and Giardia in recreational lake water in Tianjin, China: a preliminary study

Little is known about the occurrence of Cryptosporidium and Giardia in recreational water in China. A total of 52 samples were collected from recreational lakes in Tianjin during a high-occurrence season (June–October) for the waterborne cryptosporidiosis and giardiasis, and the occurrence and genotypes of Cryptosporidium and Giardia were investigated. The results showed that 82.7% (43) and 98.1% (51) of samples were positive for Cryptosporidium oocyst and Giardia cysts, respectively. The mean concentration of parasites was 3.65 oocysts/10 L and 12.58 cysts/10 L, respectively. Molecular characterization revealed that the presence of Cryptosporidium parvum, C. andersoni, C. hominis, C. meleagridis, C. fragile, C. ubiquitum, and Giardia lamblia assemblage A, B and D. The protozoan contamination in the studied lakes may originate from animal feces on ground, which was washed into the lake by stormwater runoff. Nevertheless, there is a potential risk of infection during recreational activities in the lake because the dominant detected protozoan genotypes are common human pathogens. Moreover, microbial indicators analysis does not adequately indicate the protozoan contamination in recreational water. The information from this study will be valuable for future protozoan source tracking, and any further control interventions against Cryptosporidium and/or Giardia infection associated with recreational water.

Prevalence of Cryptosporidium and Giardia in water samples. Of the 52 tested recreational lake water samples, 43 (82.7%) were positive for Cryptosporidium, with 92.3% (12/13), 84.6% (11/13), 84.6% (11/13), and 69.2% (9/13) in samples from sites 1#, 2#, 3#, and 4#, respectively. Giardia cysts were detected in all samples (98.1%, 51/52) except for the one collected on October 5, 2015 from site 4#. The counts of parasites ranged from 0 to 15 oocysts with a mean of 3.65 oocysts, and 0 to 43 cysts with an average of 12.58 cysts per 10 liters (Table 1). The positive rate for both parasites in samples collected from sampling site 4# was lower than those from other sites. Nevertheless, the difference of concentration of Cryptosporidium oocyst or Giardia cyst among all of the sampling sites was not significant (P > 0.05).
Variations of the concentrations of Cryptosporidium oocyst and Giardia cyst in different times are shown in Fig. 2. Except the peak in late July, the average counts of (oo)cysts in water samples showed a general downtrend during the survey.
Genotyping of protozoa. Nested PCR amplifications were performed for Cryptosporidium and Giardia, respectively, on all the water samples and the expected size of amplicons were produced in 25% (13 out of 52) water samples. DNA sequencing of PCR products confirmed the presence of C. parvum, C. andersoni, C. fragile, C. ubiquitum and Giardia assemblage D in samples collected from site 1#; C. parvum, C. andersoni, C. meleagridis, C. hominis and Giardia assemblage A from site 2#; C. andersoni and Giardia assemblage A from site 3#; and Giardia assemblage B from site 4#. Cryptosporidium parvum, C. andersoni, and Giardia assemblage A were the most commonly detected protozoan species/assemblage (Table 2).
Correlation between protozoa and basic water quality parameters. The results of mean, median and range of the total coliforms, turbidity and pH of lake water are presented in Table 3. As expected, all of the samples were positive for total coliforms, with concentrations ranging between 2 × 10 3 and 6.5 × 10 5 most probable numbers (MPN) per liter. Nonparametric Spearman's correlation two-tailed test showed there was a significant positive correlation between Cryptosporidium and Giardia as well as between them and turbidity (P < 0.01). Neither of the parasites displayed significant correlations with the total coliforms, which are commonly used to indicate microbial contamination (Table 4).

Discussion
Currently, the monitoring of Cryptosporidium oocyst and Giardia cyst in water is largely performed by using the United States Environmental Protection Agency (USEPA) Method 1623, which involved filtration, isolation   of (oo)cysts by immunomagnetic separation, immunofluorescence assay 28 . However, the high cost of Method 1623 has restricted its usage in most developing countries. A recent review found that only 5% of publications on Cryptosporidium and Giardia detection in water from Central/South America and Africa adopted the Method 1623 28 . Similar testing methods from the United Kingdom and other countries have been involved as alternatives 19 . In this study, we employed a concentration and purification method adopted by Japan 29 . As shown in Table S1, this method permitted mean recoveries of 41.25% (relative standard deviation (RSD), 21.54%) for Cryptosporidium and 38.32% (RSD, 26.09%) for Giardia, which could meet the acceptable levels of the Method 1623 (38-100% (RSD < 37%) for Cryptosporidium, 27-100% (RSD < 39%) for Giardia) 30 .
Oocysts of Cryptosporidium and cysts of Giardia have been reported in recreational water in many previous studies. In Spain, the concentration of Cryptosporidium in recreational rivers varied from 10 to 600 oocysts/10 L and Giardia ranged between 10 and 1600 cysts/10 L 31 . Each of them was much higher than that reported in the present study (3.46-3.85 oocysts/10 L and 11.15-14.54 cysts/10 L, respectively), while slightly lower protozoan contamination (1-4 Cryptosporidium oocysts and 1-8 Giardia cysts per 10 liters) was reported in some recreational lake waters from the Netherlands 32 . No Cryptosporidium oocyst was found in samples from a Malaysia recreational lake but a density of 1.7-11 cysts per 10 liters for Giardia was detected 33 . So the contamination of Cryptosporidium oocysts and Giardia cysts in the recreational lakes herein is within the range those reported elsewhere.

Species of Cryptosporidium (no. of samples)
Assemblage of Giardia lamblia (no. of samples)    Table 4. Correlations between the concentration of parasitic pathogens, total coliforms, turbidity and pH of the water analyzed a . a Spearman's correlation coefficient (two-tailed test), * P < 0.05, ** P < 0.01.

C. parvum C. andersoni C. hominis C. meleagridis C. ubiquitum C. fragile A B D
The occurrence of protozoa in environmental water samples is influenced by many factors, including the timing, frequency and site of sampling 34 . Sampling once in the dry season and the rainy season, respectively, was adopted in many previous studies 35,36 . However, given that summer and early autumn are the high-occurrence season for the waterborne cryptosporidiosis and giardiasis 1,2,[25][26][27] , and there was no water recreation in the park in winter, the present study therefore focused on the presence of Cryptosporidium oocyst and Giardia cyst in the hot season. Due to their small size, slow settling velocity, and resistance to environmental stress, oocyst and cyst contamination can persist in water for a long time 27 . Additionally, the lake water was disturbed by boating and the sampling intervals were relatively short. These factors resulted in a high detection rate of parasites for all samples and no significant differences in protozoan positive rate or densities at different sampling sites.
Genotyping demonstrates that a total of 6 species of Cryptosporidium and 3 assemblages of Giardia lamblia present in this study ( Table 2). Most of them have been well reported in animals across China, such as C. parvum in many kinds of animals 13 , C. meleagridis in chicken 37 , C. ubiquitum in deer and sheep 38 , C. andersoni in cattle 12,13 , and Giardia lamblia in dogs, cats and cattle 10,11 . Nevertheless, it is noteworthy that all the detected protozoan species/assemblages except C. fragile have been reported in humans 21,23,39 , and the dominant genotypes, i.e. C. parvum, C. andersoni, and Giardia assemblage A, are the common human pathogens in China 8, 16 , showing their potential public threat and requiring the attention of public health authorities.
Several studies have shown that main sources for protozoan contamination of surface water identified were the intrusion of animal feces or wastewater due to heavy rains [40][41][42] . In this study, the replenishment patterns of the lakes depend mainly on tap water and occasional heavy rainfall, so the discharge of storm water may be the major contribution of parasitic pathogens contamination. This is consistent with the variations of concentrations of Cryptosporidium oocyst and Giardia cyst. That is, the downtrend of concentration over time suggesting that no continuous protozoa was discharged into the lake, whereas the peak in late July may be attributed to the wash from fecal-polluted ground by the heavy rainfall on 19 July, 2015 (http://tj.weather.com.cn, Fig. S1). Indeed, the number of oocyst and cyst shed by infected animals was believed to be as high as 10 6 -10 8 per gram of feces 27,43 .
Considering that all detected Cryptosporidium species and Giardia assemblages have been found in animals, particularly C. fragile has only been found in amphibians, the protozoa in this study appears to be derived from animals and may be associated to the nearby zoo since there are many animals including amphibians. Thus, it is one of the important questions for future studies to determine the source of protozoa in the lakes.
Physical-chemical properties such as turbidity and bacteria such as total coliforms are commonly used as indicators for monitoring water quality. Significant correlations were observed between turbidity and parasites, and between parasites themselves in the present study. These findings are in agreement with previous studies observed in a recreational river in Taiwan, China 44 , in the Three Gorges Reservoir in China 36 , and in a reservoir in Spain 35 . Nevertheless, no parasite was related to indicator bacteria in this study. In contrast, Graczyk, et al. 45 found that bacteria count was a good indicator for the presence of Giardia and Cryptosporidium in marine recreational beach water. Meanwhile, they noted that water sample should be collected during times when bather numbers are high and tested in time. The strong resistance of Cryptosporidium and Giardia to environmental stress and long-term viability may be the main cause of different results of the above studies 46 . Previous study also found that fecal bacterial indicator was not an appropriate index to monitor the presence of Cryptosporidium or Giardia in treated recreational water 20 . Consequently, the analysis of microbial indicators does not adequately characterize contamination of protozoa in recreational water, especially if the sampling is not timely.
In conclusion, Cryptosporidium and Giardia were commonly detected in the recreational lakes, and their concentration levels are similar to those reported elsewhere. The protozoan contamination may be attributed to heavy rains that wash ground-polluted feces into the lake, and was related to the nearby zoo animals. As the dominant detected protozoan genotypes are common human pathogens, their potential threat requires the attention of public health authorities. In addition, microbial indicators analysis does not adequately indicate the protozoan contamination in recreational water. Future studies should therefore include follow-up work designed to assess the occurrence of protozoa of feces from residents nearby and the zoo animal, allowing a full evaluation of its public health.

Material and Methods
Study site. The studied lakes are located in the Tianjin Waterpark (39.080° to 39.095° N, 117.159° to 117.174°E), which is the largest comprehensive park within the city boundaries of Tianjin, China. All the lakes are connected together and the surface area is about 750,000 square meters. Apart from the surface runoff during storm, the lakes depend mainly on tap water for its replenishment. It is an aquatic venue used for rowing, interactive fountain, wading, and other water recreation activities. There were hundreds of people, mainly the elderly residents nearby, swim in the early morning and evening in summer, even though swimming is banned here.
Sample collection and processing. During the period between June and October, which was the high-occurrence season for waterborne cryptosporidiosis and giardiasis 26,27,47,48 , water samples were collected from 4 sampling sties (as shown in Fig. 1) at ten-day intervals and a total of 52 samples were obtained. The sampling site 1# is close to the zoo in the south of the West lake and site 4# in the north, site 2# is located in the swimming area in the East lake and site 3# is near the cruise ship terminal. Each sample was collected using a 20-L plastic container. After collection, the samples were transported on ice to the laboratory immediately and parasitic pathogens in the samples were concentrated by a membrane filter dissolution method described previously 49 . Briefly, water samples were filtered and then the mixed cellulose ester membrane filter (diameter, 142 mm; pore size, 1 μM; Advantec MFS, Inc. Japan) was dissolved in acetone solutions, followed by centrifugation at 1,050 × g for 10 min at 4 °C. The packed pellets were resuspended in a suitable volume (2-10 mL) of distilled water. Half of the resuspension from each sample was used for protozoan enumeration and the other half for DNA extraction. Morphological examination and enumeration for protozoa. The recovered Cryptosporidium oocysts and/or Giardia cysts in the pellet were separated from debris by flotation on Percoll-sucrose gradients (specific gravity, 1.10) 49 . Each purified sample was stained with 100 μL of combined fluorescein isothiocyanate (FITC) conjugated anti-Cryptosporidium and anti-Giardia monoclonal antibodies (Waterborne, Inc., New Orleans, LA) in a humid dark chamber at room temperature for 30 min, and then stained with 50 μL of 4′,6′-diamidino-2-phenylindole (DAPI, Waterborne) solution (0.4 μg/mL) for 10 min. The prepared slides were subsequently examined microscopically at 400× magnification using fluorescence microscopy (Olympus, Japan) for the detection and enumeration of Cryptosporidium oocysts and Giardia cysts according the USEPA Method 1623 30 .
To evaluate the recovery efficiency of the method used, the initial precision and recovery efficiency was determined by spiking four 10 L of purified water samples with suspension containing enumerated oocysts and cysts 50 . Furthermore, a negative control assay, in which all the procedures were performed using deionized water instead of the water sample, was performed on each batch of samples to determine if any contamination occurred during the analysis. All negative control analyzes were negative for oocysts or cysts.
Genotyping of Cryptosporidium and Giardia. Genomic DNA was extracted from each of the sample concentrates using the FastDNA SPIN kit for soil (MP Biomedicals, France) and eluted into 50 μL of reagent-grade water according to previous study 51 . Small subunit (SSU) rRNA-based nested PCR followed by sequencing techniques were employed to determine the species/genotypes of the protozoa according to previous study 52 . Briefly, a 435-bp fragment of Cryptosporidium SSU rRNA locus and a 292-bp fragment of Giardia SSU rRNA gene were amplified by two nested PCR, respectively. The sequence of all primers and PCR conditions are specified in Table S2. All positive secondary PCR products were sent to Beijing Augct Co., Ltd. for direct sequencing with the secondary PCR primers using ABI 3730 automated DNA sequencer (BigDye Terminator Chemistry). Nucleotide sequences obtained in the study were aligned with reference SSU rRNA sequences downloaded from the Genbank using the Clustal W programs and analyzed to determine Cryptosporidium species/genotypes and Giardia assemblages using phylogenetic trees.
Microbiological analysis, turbidity and pH measurement. According to the Chinese standard examination method for drinking water-microbiological parameters, five-tube most probable numbers (MPN) procedure was used to enumerate total coliforms 53 . Briefly, after ten-fold serial dilution, 1.0 mL of each serial dilution was transferred to five tubes of lactose peptone broth containing inverted Durham tubes, and then incubated at 37 °C for 24 ± 2 h. All positive presumptive tubes that demonstrated an acidic reaction or gas production were submitted to the confirmed phase with total coliform test by using eosin methylene blue agar medium 53 . The turbidity and pH were measured on site for each sample with portable photometer. Statistical analysis. The paired-samples t-test was employed to evaluate the relationship between the concentrations of Cryptosporidium oocyst and Giardia cyst. The association between total coliform, turbidity, pH and protozoan concentrations was correlated using the nonparametric Spearman's correlation two-tailed test. Differences with P values of <0.05 were defined as being statistically significant. All statistical tests were performed using PASW Statistics 18 computer software package. Data availability. All relevant data are fully available without restriction.