A survey of Cryptosporidium prevalence among birds in two zoos in China

Background Cryptosporidiosis is an important zoonotic protozoan disease worldwide, but few studies on this disease have been performed in wild birds; thus, our knowledge of this disease is insufficient, even in zoo birds. Animals in zoos are possible zoonotic disease reservoirs, potentially resulting in zoonotic agent spillover to humans; accordingly, our understanding of such phenomena should be improved. Methods A total of 263 fresh fecal samples from 43 avian species were randomly collected from the Beijing Zoo and Harbin North Forest Zoo and screened for the prevalence of Cryptosporidium by 18S rRNA gene sequencing. Cryptosporidium species were distinguished based on the combined results of phylogenetic tree and genetic distance analyses conducted with the inclusion of seven avian Cryptosporidium species and 13 avian Cryptosporidium genotypes. The genetic diversity of Cryptosporidium parvum among different hosts, including humans, cattle, dogs, and birds, and the genetic diversity of avian C. parvum among avian hosts in China, Iraq and Brazil were determined based on C. parvum 18S rRNA haplotypes. Results The results of PCR targeting the 18S rRNA gene revealed that 1.9% (5/263) of the samples were Cryptosporidium-positive. Four of the five Cryptosporidium-positive samples originated from white cranes (Grus leucogeranus), and one originated from a flamingo (Phoenicopteridae). Avian C. parvum isolates, including the isolates examined in the present study, showed gene flow with other isolates from different types of hosts, including humans, cattle and dogs, indicating that zoo birds potentially pose zoonotic and pathogenic risks to humans and animals. Additionally, gene flow between avian C. parvum isolates from China and Brazil was detected. Conclusions To the best of our knowledge, our results demonstrate C. parvum infection in a flamingo (Phoenicopteridae) and white cranes (Grus leucogeranus) for the first time. The results of our study provide an important reference for understanding the host range, biological characteristics, and molecular epidemiology of C. parvum.

The small intestine and cecum will develop lesions when avian are infected with C. parvum (Nakamura & Meireles, 2015). An outbreak of C. parvum infection among stone curlews was reported in Dubai; the symptoms included sudden-onset diarrhea, lethargy and reduced feed intake (Zylan et al., 2008). However, detailed descriptions of wild birds infected with C. parvum are lacking.
Species identification is a crucial step in preventing and curing Cryptosporidium infections, but it is difficult to distinguish species based on only morphology, as they share similar morphological characteristics (Ryan, Fayer & Xiao, 2014). Molecular methods targeting markers in the 18S rRNA gene are useful, practical tools for Cryptosporidium species identification, with high detection limits (Ryan et al., 2003).
Zoos are potential environmental sources of zoonotic diseases because they serve as temporary homes for wild birds. Tourists visiting a zoo may come into close contact with certain wild birds, which is a practical pathway of C. parvum spillover from birds to humans. Therefore, this pathway should be studied to increase our understanding of these spillover events. The objectives of the present study were (1) to investigate Cryptosporidium species and their prevalence among zoo birds in China and (2) to evaluate the potential risk of transmission of Cryptosporidium from zoo birds to other animals or humans.

Fecal samples
We commissioned the zoo staff to help us collect samples, all staff wore disposable gloves and placed bird droppings into sterile centrifuge tubes. All samples were transported to the laboratory on dry ice. In the laboratory, the samples were stored in a sterile centrifuge tube containing a 5% potassium dichromate solution in a refrigerator at 4 • C until analysis. A total of 263 fresh fecal samples from birds belonging to 16 families, 30 genera, and 43 species were collected from March 2019 to September 2019 at the Beijing Zoo and the Harbin North Forest Zoo (Table S1). Among the 263 fecal samples, 66 were from the Harbin Zoo, and 197 were from the Beijing Zoo. Parasite eggs were separated in a saturated solution of sucrose used as floating medium and stored at −20 • C until DNA extraction.

Molecular analyses
DNA was extracted directly from the samples by using the Stool DNA Kit (Omega Bio-Tek, China Technical Support Center) following the manufacturer's instructions. Nucleic acids were eluted in 50 µl of elution buffer to increase the quantity of DNA recovered. The 18S rRNA gene sequence was amplified by nested PCR (Koehler et al., 2018). The primer sets 18SCiF2: 5 -GACATATCATTCAAGTTTCTGACC-3 and 18SCiR2: 5 -CTGAAGGAGTAAGGAACAACC-3 were selected for the primary reaction, and 18SCiF1: 5 -CCTATCAGCTTTAGACGGTAGG-3 and 18SCiR1: 5 -TCTAAGAATTTCACCTCTGACTG-3 were employed for the secondary reaction. The reaction mixture consisted of 11 µl of double distilled water, 2 µl of each primer (Comate Biosciences Co., Ltd. Changchun, China), 25 µl of Taq DNA polymerase (TaKaRa, Dalian, China), and 10 µl of template DNA, with a total reaction volume of 50 µl. The PCR program consisted of an initial heating step at 94 • C for 5 min, followed by 45 cycles of 94 • C for 30 s, 55 • C for 30 s, and 72 • C for 30 s (for primary and nested PCR) and a final extension step of 7 min at 72 • C. Each PCR product was identified in a 0.8% agarose gel to verify its size. The PCR products were sent to Comate Biosciences Co., Ltd. (Changchun, China), for sequencing.
The haplotypes of the 18S rRNA genes of 7 avian Cryptosporidium species and C. parvum from 4 different host classes were analyzed by using DnaSP 5.0 software. A phylogenetic tree of the Cryptosporidium spp. haplotypes was constructed with the same parameters mentioned above. The genetic distances between different Cryptosporidium species were calculated with the MEGA 7.0 program. Arlequin 3.0 was used to calculate the pairwise fixation index (Fst) values of C. parvum between different host classes and different countries using the 18S rRNA gene. The migration number of each generation (Nm) was determined according to the following formula: Fst = 1/ (4Nm + 1). The corresponding heatmap was generated with the TBtools program. We used statistical and descriptive analysis in infection rate calculation.
We found 5, 3, 1, 6, 1, 3 and 8 unique 18S rRNA gene haplotypes in avian C. parvum, C. meleagridis, C. avium, C. baileyi, C. muris, C. andersoni and C. galli, respectively. Figure 2 shows that the Cryptosporidium-positive samples (G3, G4, G5, G6 and P1) were included in the clade of C. parvum, which was parallel to the clade of C. meleagridis. C. parvum Hap1 and C. parvum Hap 5 were defined in the present study. Hap1 was shared between avian C. parvum from white crane and flamingo in the present study and has been reported previously in other species. However, Hap5 was found in only white cranes (Table S2).
The analysis of the genetic distances between the 31 haplotypes of the 18S rRNA genes in the 7 avian Cryptosporidium species showed that the intraspecific genetic distances of C. parvum were within 0.007, but the interspecific genetic distances between C. parvum and the other six Cryptosporidium species were all greater than 0.007 (Fig. 2), which together with the results of the phylogenetic tree analysis, verified that the Cryptosporidium species obtained in the present study was C. parvum. We identified 7, 5, 9 and 3 haplotypes in C. parvum from hosts in 4 different classes (cattle, dogs, humans and birds), and Hap1 and 5 were present in all four groups (Table 1). Hap5 was shared between avian C. parvum in white cranes and flamingo in the present study; however, Hap4 was found in only white cranes (Table S3).
The Fst and Nm values based on the 18S rRNA genes between the bird and dog groups were 0.18475 and 1.10318, between the bird and cattle groups were 0.22421 and 0.86503, Figure 1 Phylogenetic tree for Cryptosporidium spp. Maximum Likelihood tree based on 18S rRNA gene sequences of seven Cryptosporidium species (C. meleagridis, C. parvum, C. galli, C. baileyi, C. avium, C. muris and C. andersoni) and 13 Cryptosporidium genotypes (avian genotypes I-III, and VI-IX; goose genotypes I-II; duck genotype; finch genotype; ostrich genotype; and C.xiaoi-like genotype).
Full-size DOI: 10.7717/peerj.12825/ fig-1 and between the dog and cattle groups were 0.58897 and 0.17447, respectively. The values among C. parvum from birds, dogs and cattle indicate gene flow between the C. parvum isolates from birds and those from dogs and cattle. The values between the cattle and human groups were 0.05063 and 4.68778, and between the bird and human groups were 0.02137 and 11.4486, respectively, and it seems that birds, like cattle, may transmit C. parvum to humans (Table 2). We calculated the Fst and Nm values of C. parvum isolated from three countries: Brazil, Iraq and China (Table S4). The Fst and Nm values between Brazil and China were 0.01530 and 16.08987, between China and Iraq were 0.45714 and 0.29688, and between Brazil and Iraq were 0.24490 and 0.77082, respectively (Table 3). Based on this data, there has been  gene flow between C. parvum from Brazil and China; however, based on geography, there should theoretically be no gene flow between the two countries.

DISCUSSION
The rate of C. parvum positivity was 1.9% (5/263) in the present study, which was slightly lower than the C. parvum positivity rate of 8.5% (58/679) detected in wild birds from   Qinghai Lake (Jian et al., 2021), and higher than the rate of 0.1% (1/1005) detected in captive pet birds in Henan (Dong et al., 2021). Based on these data, the zoo C. parvum positivity rate was lower than that in the wild but was higher than that in captive pet birds, but more experiments are needed to support this conclusion. Currently, we also found that C. parvum could infect white crane and flamingo, and we found a unique C. parvum haplotype in white crane. The variation in Cryptosporidium species plays a key role in the clinical presentation of cryptosporidiosis in humans or animals (Santin, 2013); therefore, the Cryptosporidium species isolated in this study, C. parvum, and its particular haplotype must be studied in greater depth in the future.
In addition, C. parvum has been found to infect humans and other mammals (Ayinmode et al., 2018;Moore et al., 2016). Many of the numerous outbreaks of human cryptosporidiosis reported annually worldwide have been linked to mammals as sources of C. parvum infection. The haplotype analysis of C. parvum in four different host classes showed that there were two shared haplotypes of C. parvum among those hosts (Table 1), which indicated that avian C. parvum could spill over into humans and other mammals because of its cross-host transmission ability. The gene flow of C. parvum among the four different host classes was also in accordance with the results of share haplotypes (Table 2). Livestock, particularly cattle, are the most important reservoirs of C. parvum, potentially posing risks to humans (Mravcova et al., 2020). In our study, the Fst value between the bird group and human group was 0.02137, which is even smaller than the value between the cattle group and human group (0.05063). High gene flow suggests that birds, similar to cattle, are crucial reservoir hosts of zoonotic C. parvum pathogens, posing risks to humans.
Zoos are important environments for the protection of animals; they are important not only for enhancing the natural resources of animals but also for providing protection and conservation education. As animals in zoos are generally kept in relatively close proximity and have increased opportunities for contact with humans, such as tourists, researchers, feeders, and veterinary staff, sick birds in zoos may potentially be associated with spillover of zoonotic Cryptosporidium pathogens to humans. Therefore, more attention should be given to monitoring for Cryptosporidium infection, such as C. parvum infection, in birds in zoos due to its potential cross-host transmission ability.
Brazil, a country located in South America, is far from China; the two countries are theoretically geographically isolated, and no natural avian C. parvum gene flow should occur between these countries. However, the Fst and Nm results showed higher gene flow between avian C. parvum from Brazil and China (Table 3). The reason for this scenario may be that the C. parvum pathogen was introduced into the zoo from imported Brazilian flamingos, exchanged animals from other zoos or animals rescued from the wild. This result indicates the need for stricter regulations regarding inspection, quarantine, and isolation protocols for new animal residents.

CONCLUSIONS
The present study is the first to report that flamingos and white cranes can be infected with C. parvum. This knowledge is beneficial for the further protection of white cranes and flamingos and expands the known host range of C. parvum. The possibility of Cryptosporidium infection is considered in only a small number of wild birds, including those involved in the present study. There are additional potential wild bird hosts that have yet to be identified, and more work on this topic needs to be carried out in the future.

ADDITIONAL INFORMATION AND DECLARATIONS Funding
This work was supported by the National Natural Science Foundation of China (No. 31970501). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Grant Disclosures
The following grant information was disclosed by the authors: The National Natural Science Foundation of China: No. 31970501.