Persistent Occurrence of Cryptosporidium hominis and Giardia duodenalis Subtypes in a Welfare Institute

Few data are available on the transmission dynamics of intestinal protozoa in children in welfare institutes. In this study, fecal specimens were collected from 396 children in a welfare institute in Shanghai, China during December 2011 (207 specimens), June 2012 (78 specimens), and September 2013 (111 specimens), and examined for Cryptosporidium spp., Giardia duodenalis, and Enterocytozoon bieneusi by PCR analysis of the small subunit rRNA, triosephosphate isomerase, and internal transcribed spacer genes, respectively. The Cryptosporidium hominis and G. duodenalis assemblage A identified were further subtyped by multilocus sequence typing. Altogether, Cryptosporidium was detected in 39 (9.8%) children, with infection rates of 11.6% (24/207), 9.0% (7/78), and 7.2% (8/111) in December 2011, June 2012, and September 2013, respectively. Infection rates were higher in children of 0–12 months (20.4% compared to 0–7.3% in other age groups, P = 0.0001) and those with diarrhea (17.9% compared to 7.7% in those with no diarrhea, P = 0.006). In contrast, G. duodenalis was detected in 161/396 (40.7%), with infection rates of 48.3% (100/207), 35.9% (28/78), and 29.7% (33/111) in December 2011, June 2012, and September 2013, respectively. There were no significant gender- or diarrhea-associated differences, but the G. duodenalis infection rate in children of 13–24 months (50%) was significantly higher than in the age groups of 0–12 months and > 48 months (29.8–36.5%, P = 0.021). Co-infection of Cryptosporidium and G. duodenalis was seen in 19 (4.8%) children, but no E. bieneusi infection was detected in this study. All Cryptosporidium-positive specimens belonged to the subtype IaA14R4 of C. hominis, while all G. duodenalis-positive specimens belonged to sub-assemblage AII. Both were the same subtypes in a previous outbreak of cryptosporidiosis and giardiasis in a hospital ward hosting children from the welfare institute. Results of the study indicate that there was a persistent occurrence of limited C. hominis and G. duodenalis subtypes in the small enclosed community, with differences in age distribution and association with diarrhea occurrence between cryptosporidiosis and giardiasis.


INTRODUCTION
Cryptosporidium spp., Giardia duodenalis, and Enterocytozoon bieneusi are common enteric pathogens in humans and animals, causing diarrhea in children, the elderly and AIDS patients. Humans become infected by these parasites through the fecaloral route via direct contact with infected persons (anthroponotic route) or animals (zoonotic route) and ingestion of contaminated food or water (foodborne or waterborne route) (Xiao, 2010;Feng and Xiao, 2011;Santin and Fayer, 2011). Children are particularly vulnerable to infections, especially those who live in enclosed, crowded and insanitary environment such as daycare centers. In addition to diarrhea, cryptosporidiosis, giardiasis, and microsporidiosis in this population could cause malnutrition and cognitive impairments (Dillingham et al., 2002;Oliveira-Arbex et al., 2016).
Over 30 Cryptosporidium species and more than 40 genotypes have been identified to date, but Cryptosporidium hominis and C. parvum are responsible for > 90% of cryptosporidiosis cases in humans (Feng et al., 2018). Similarly, among the eight described assemblages (A-H) of G. duodenalis genotypes based on genetic characterizations, assemblages A and B are the causes of almost all cases of human giardiasis (Feng and Xiao, 2011). Multiple groups of genotypes with various host ranges are also present within E. bieneusi, with humans being mostly infected with Group 1 genotypes (Thellier and Breton, 2008). The identification of Cryptosporidium spp., G. duodenalis, and E. bieneusi at the species/genotype and subtype levels requires the use of molecular diagnostic tools. As different species and subtypes have different host ranges, the use of molecular diagnostic tools in epidemiological investigations has significantly improved our understanding of the transmission of Cryptosporidium spp., G. duodenalis and E. bieneusi in general populations (Thellier and Breton, 2008;Xiao and Feng, 2017).
Microscopy studies in Egypt, Peru, Thailand, Turkey, and some Eastern European countries have indicated a common occurrence of cryptosporidiosis, giardiasis and microsporidiosis in children in welfare institutes (Janoff et al., 1990;Makhlouf et al., 1994;Leelayoova et al., 2005;Borekci and Uzel, 2009;Bailey et al., 2013;Boontanom et al., 2014;Strkolcova et al., 2016;Kasprzak et al., 2017). However, the persistence of enteric pathogens in welfare institutes has not been examined longitudinally or using molecular diagnostic tools. In addition, there are no data on the occurrence of cryptosporidiosis, giardiasis and microsporidiosis in welfare institutes in China. Therefore, molecular tools were used in the present study to assess the occurrence, persistence and genetic characteristics of Cryptosporidium spp., G. duodenalis, and E. bieneusi in a welfare institute in Shanghai, China.

Ethics Statement
Fecal specimens in this study were collected as part of the public health investigation of a hospital-associated outbreak of cryptosporidiosis and giardiasis in inpatients from one welfare institute in Shanghai, China (Feng et al., 2012;Wang et al., 2013). During that investigation, children in one hospital ward hosting inpatients from the welfare institute had high occurrence of two subtypes of C. hominis and one subtype of G. duodenalis assemblage A over a 14-month period. Informed consent and permission for the collection of fecal specimens from diapers were obtained from the welfare institute manager and custodians of the children. The attending physician of the welfare institute submitted the fecal specimens to investigators of this study as part of a request to investigate the cryptosporidiosis and giardiasis outbreak. The research personnel had no physical access to the welfare institute and no direct contact with children in the welfare institute. Information provided by caregivers at the sampling included only age and gender of the children sampled and the presence or absence of diarrhea. As work in this report is part of the routine clinical diagnosis of pathogens for diarrhea in a health care facility as well as a public health investigation of an outbreak, the research protocol was considered an exempted human subject research. It was reviewed and approved by the Ethics Committee of East China University of Science and Technology.

Specimen Collection
A total of 396 fresh fecal specimens were collected in December 2011 (207 specimens), June 2012 (78 specimens), and September 2013 (111 specimens) from children in the welfare institute in Shanghai, China. These children were 1-125 months in age, including 203 boys, 184 girls, and 9 with missing data. Among the specimens collected, 84 were from children with diarrhea. These children were mostly abandoned because of serious illnesses, including neurological/mental disability, growth retardation, cardiovascular diseases, and other conditions. Specimens were stored at 4 • C in 2.5% potassium dichromate until detection of pathogens using molecular diagnostic tools.

DNA Extraction
Each fecal specimen was washed twice with distilled water and centrifuged at 2,000 × g for 10 min to remove potassium dichromate. DNA was extracted from 200 µl of the washed fecal material using the FastDNA SPIN Kit for Soil (MP Biomedicals, Santa Ana, CA, United States), and stored at −80 • C until being analyzed by PCR.
To detect G. duodenalis, a ∼530-bp fragment of the triosephosphate isomerase (tpi) gene was amplified by nested PCR (Feng et al., 2008). To identity multi-locus genotypes (MLGs) of G. duodenalis, a ∼511-bp fragment of the β-giardin (bg) gene (Caccio et al., 2008) and a ∼530-bp fragment of the glutamate dehydrogenase (gdh) gene (Abe et al., 2003) were amplified for all tpi-positive specimens, following the standard MLG terminology for G. duodenalis assemblage A (Feng and Xiao, 2011). For the detection of E. bieneusi, a nested PCR targeting a ∼530-bp fragment of the rRNA unit containing the entire 392-bp internal transcribed spacer (its) was used (Sulaiman et al., 2003). In all PCR analyses, each specimen was analyzed twice using 1 µl of extracted DNA per PCR. Non-acetylated bovine serum albumin (Sigma-Aldrich, St. Louis, MO, United States) was used in the primary PCR at the concentration of 400 ng/µl to neutralize residual PCR inhibitors within the extracted DNA.

Sequence Analysis
All positive secondary PCR products in this study were sequenced bi-directionally using the Big Dye Terminator v3.1 Cycle Sequencing Kits (Applied Biosystems, Foster City, CA, United States) on an ABI 3130 Genetic Analyzer (Applied Biosystems). Sequences were read and assembled using ChromasPro 1.5 1 and compared with reference sequences in NCBI database using ClustalX 2 to determine the genotypes and subtypes of pathogens. The novel hsp70 gene sequence generated in this study was submitted to GenBank under the accession number MH329305.

Data Analysis
The Chi-square test was used to compare differences in infection rates among samplings, age groups, genders and diarrhea status. Differences were considered significant at the level of P < 0.05. The statistical analysis was performed using the SPSS Statistics V21.0 for Windows (IBM Corp., New York, NY, United States). As children were sampled anonymously, no attempts were made to differentiate infection episodes among samplings.

RESULTS
Occurrence of Cryptosporidium spp.
Among the 78 and 111 specimens examined at the second and third sampling, 7 (9.0%) and 8 (7.2%) were positive for Cryptosporidium spp., respectively. The difference in Cryptosporidium infection rates was not significant among the three samplings (P = 0.438). As observed at the initial sampling, the age group of 0-12 months had higher infection rates at the second (16.7% or 7/42) and third (18.4% or 7/38) samplings than other age groups (0-2.9%). Likewise, children with diarrhea had a higher infection rate than those without diarrhea at the second (17.6% compared with 6.6%; P = 0.171) and third (14.3% compared with 5.6%; P = 0.173) sampling (Table 2).
None of the 396 fecal specimens analyzed in the study were positive for E. bieneusi.

Cryptosporidium Species and Subtypes
All Cryptosporidium-positive secondary PCR products of the ssu rRNA gene in this study were successfully analyzed by RFLP, showing the presence of only C. hominis in the 39 PCRpositive specimens. Among them, 26 were successfully subtyped by PCR and sequence analysis of the gp60 gene, leading to the identification of the only subtype IaA14R4 (GenBank accession number KC734574) ( Table 1). Among the 26 gp60-positive specimens, 17, 21, 22, 22, 23, and 23 specimens were successfully subtyped by PCR and sequence analysis of the hsp70, cp47, cp56, dz-hrgp, msc6-7, and rpgr genes, respectively. Sequences obtained in the study were identical to each other at each genetic locus (GenBank accession numbers KC734581 for msc6-7, KC734584 for rpgr, KC734585 for dz-hrgp, KC734587 for hsp70, KC734590 for cp47, and KC734593 for cp56), except for the hsp70 locus, which had one single nucleotide substitution of T to C (nucleotide position at 268) in one specimen (GenBank accession number MH329305).

Giardia duodenalis Genotype and Subtypes
Among the 161 G. duodenalis-positive specimens, tpi products from 153 specimens were successfully sequenced, showing the presence of subtype A2 (GenBank accession number U57897) at all three samplings (Table 1). Among the 161 PCR-positive specimens at the tpi locus, 138 were positive in the bg-based PCR, and 128 were positive in the gdh-based PCR. All these PCR products were successfully sequenced, showing the presence of subtypes A2 (in 93 specimens), A3 (in 38 specimens), and A8 (in 7 specimens) at the bg locus (GenBank accession numbers AY072723, FJ971415, and JX898208, respectively) and A2 (in 128 specimens) at the gdh locus (GenBank accession number AY178737). At the bg locus, subtype A2 was found in 47, 21, and 25 specimens at the first, second, and third sampling, respectively; subtype A3 was detected in 34 and 4 specimens at the first and third sampling, respectively; whereas subtype A8 was only found in 7 specimens at the initial sampling. Altogether, sequence data were available at all three loci for 120 G. duodenalis-positive specimens, generating three MLGs of G. duodenalis in this study. All the three MLGs belonged to the sub-assemblage AII of assemblage A, with 83 specimens being positive for AII-1, 33 for AII-8 and 4 for AII-13 (Table 1).

DISCUSSION
A common occurrence of cryptosporidiosis and giardiasis was seen in children in the welfare institute in this study. The 9.8% overall infection rate of Cryptosporidium spp. is in concordance with observations in previous studies of cryptosporidiosis in welfare institutes in Thailand (Janoff et al., 1990;Jongwutiwes et al., 1990), but significantly higher than results obtained from other studies in Egypt and Turkey (Makhlouf et al., 1994;Borekci and Uzel, 2009;Doni et al., 2013). The infection rate was similar at the first (11.6%), second (9.0%), and third (7.2%) samplings, suggesting that there was a persistent occurrence of C. hominis in this welfare institute. The 40.7% overall infection rate of G. duodenalis was significantly higher than infection rates in previous studies in welfare institutes in Thailand, Egypt, Turkey and Peru (Janoff et al., 1990;Makhlouf et al., 1994;Borekci and Uzel, 2009;Turhan et al., 2009;Bailey et al., 2013;Doni et al., 2013). There was a gradual reduction in G. duodenalis infection rates during the study period (48.3, 35.9, and 29.7% at the first, second and third sampling, respectively). The latter, however, are still significantly higher than those reported in children in the general community (Natividad et al., 2008;Wang et al., 2013;Tellevik et al., 2015;Squire and Ryan, 2017). In contrast to the frequent detection of E. bieneusi in children in developing countries (Leelayoova et al., 2005;Wang et al., 2013;Lobo et al., 2014;Yang et al., 2014), none of the specimens analyzed in the present study was positive for this pathogen.
Results of genotyping and subtyping of Cryptosporidium spp. and G. duodenalis support the persistent nature of their occurrence in the welfare institute. In the present study, we identified the occurrence of only one C. hominis subtype IaA14R4 in all Cryptosporidium-positive specimens. MLST analysis of the C. hominis strains has further supported the homogeneous nature of the pathogen, with only one nucleotide substitution observed in one specimen at the hsp70 locus. Similarly, only limited genetic heterogeneity was seen in the assemblage A of G. duodenalis. At the tpi and gdh loci, all positive specimens were identified as having the subtype A2, while some minor sequence differences were observed at the bg locus, with A2 as the dominant subtype. Thus, relatively homogeneous populations of C. hominis and G. duodenalis have been circulating among children in the welfare institute during the 22-month study period, although we cannot fully exclude the possibility of introduction of other G. duodenalis strains via new orphans. Different from the observation in the present study, several Cryptosporidium species and G. duodenalis assemblages were found in children from developing countries, including multiple subtypes of C. hominis and C. parvum and both assemblages A and B of G. duodenalis (Sanchez et al., 2017;Squire and Ryan, 2017;Naguib et al., 2018).
The IaA14R4 subtype was one of two C. hominis subtypes (the other one was IdA19) involved in a hospital-associated outbreak of cryptosporidiosis in Shanghai over a 14-month period (Feng et al., 2012). Ward A of this hospital was an inpatient ward for ill children from the welfare institute. Similarly, the G. duodenalis subtypes found in children of the welfare institute were also found in patients from this ward during the outbreak of cryptosporidiosis (Wang et al., 2013). In contrast, children in other wards of the hospital had assemblage B in addition to assemblage A (Wang et al., 2013). The detection of the same AII subtypes of G. duodenalis in the present studies supports the contribution of children from the welfare institute to the nosocomial infections of multiple enteric pathogens in Ward A. The lack of E. bieneusi detection and C. hominis IdA19 subtype in the present study suggests that introduction of pathogens from the welfare institute was not entirely responsible for the cryptosporidiosis outbreak. Previously, E. bieneusi was detected in 10.8% (8/74) of children in Ward A during the outbreak (Wang et al., 2013).
Children in the study had unique age-associated transmission patterns of C. hominis and G. duodenalis. Cryptosporidium infection rate was the highest in the age group of 0-12 months (20.4%). This is in contrast to the peak occurrence of cryptosporidiosis in children of 1-4 years in previous Chinese and African studies (Chen et al., 1992;Squire and Ryan, 2017;Ukwah et al., 2017). For example, in rural Jiangsu, China, Cryptosporidium infection rates were 3.4-3.63% in healthy children of 1-3 years, compared to 2.3% in children under 1 year (Chen et al., 1992), while in urban and suburban areas of Nigeria, the infection rates were 10.7 and 5.7% in diarrheic children of 1-4 years and under 1 year, respectively (Ukwah et al., 2017). Similarly, children of 13-24 months had the highest G. duodenalis infection rate, which appears to be different from observations in other studies in China, the Netherlands and Africa (Dib et al., 2008;Ismail et al., 2016;Pijnacker et al., 2016;Squire and Ryan, 2017). Therefore, the peak occurrence of cryptosporidiosis and giardiasis in the welfare institute is earlier in age than that seen in other studies. This is probably a reflection of complicating health conditions of children in the present study. The different age patterns between C. hominis and G. duodenalis infections could be due to some intrinsic biological differences between the two microorganisms.
The reason for the persistent occurrence of limited C. hominis and G. duodenalis subtypes in these children is unclear. Congregation of susceptible hosts, poor health, especially mental retardation in some children, and poor hand-wash of caregivers after diaper changes and before feeding could have all contributed to the fecal-oral transmission and persistent occurrence of C. hominis and G. duodenalis in this welfare institute. The role of each of the risk factors can only be assessed through thorough field investigations with extensive collection of epidemiological data.

CONCLUSION
In conclusion, data of the study have revealed a persistent occurrence of C. hominis and G. duodenalis in children in the study welfare institute, as indicated by the presence of only one subtype of each parasite over an extended period of time. Proper training of caregivers and improved hygiene should be implemented to reduce the occurrence of cryptosporidiosis and giardiasis in the study facility and the spread of these diseases to the adoption families and the general community.

AUTHOR CONTRIBUTIONS
YF and LX designed the study. YW, NL, YG, LW, and RW performed the experiments. YW, LW, YF, and LX performed the statistical analysis, interpreted the results, and developed the draft manuscript. All authors contributed to manuscript revisions and approved the final version for publication.