Background

Cryptosporidium is a socioeconomically significant genus of parasitic protists that infect humans and other vertebrates worldwide. Species within this genus are transmitted via the faecal-oral route, often through direct contact with infected people or animals, food and/or water [28,29,30], and this species has been recorded mainly in immuno-compromised persons and in children [11, 31,33], with a potential to lead to chronic cryptosporidiosis [34]. The source of C. meleagridis infection in humans was unclear in most published reports, and the assumption has been that birds can be significant reservoirs of this protist, although anthroponotic transmission might also occur [33]. PCR was conducted in 50 μl containing 50 mM KCl and 10 mM Tris-HCl (pH 8.4; Promega, Madison, USA), 3.0 mM of MgCl2, 200 μM of each deoxynucleotide triphosphate, 50 pmol of each primer and 1 U of MangoTaq DNA polymerase (Bioline, London, UK). Known test-positive, test-negative and no-template (including ‘carry-over’) controls were included in each step of each set of PCRs. PCR products were resolved on 1.5% agarose gels, stained with ethidium bromide prior to sequencing. Then, aliquots (5 μl) of individual amplicons (undigested) were treated with the enzyme Exo I and a thermosensitive alkaline phosphatase (FastAP, Thermo Fisher, Carlsbad, USA), according to the manufacturer's instructions, and subjected to automated sequencing (BigDye Terminator v.3.1 chemistry, Applied Biosystems, Foster City, USA) employing the same primers (separately) as used in PCR.

Sequences were aligned using the program MAFFT [36], and alignments manually adjusted employing the program Mesquite v.3.10 [37]. Sequences were then compared with sequence data available via GenBank (NCBI) using BLASTn (Additional file 1: Table S1). Phylogenetic analysis of pSSU or pgp60 sequence data (including selected reference sequences; Additional file 1: Table S1) was conducted by Bayesian inference (BI) using Monte Carlo Markov Chain (MCMC) analysis in MrBayes v.3.2.6 [38]. The likelihood parameters set for BI analysis of pgp60 data were based on the Akaike Information Criteria test [39] in jModeltest v.2.1.7. The number of substitutions (Nst) was set at 6, with an invariant gamma-distribution. Posterior probability (pp) values were calculated by running 2,000,000 generations with four simultaneous tree-building chains. Trees were saved every 100th generation. At the end of each run, the standard deviation of split frequencies was < 0.01, and the potential scale reduction factor approached one. A 50% majority rule consensus tree for each analysis was constructed based on the final 75% of trees generated by BI. Analyses were run three times to ensure convergence and insensitivity to priors. The outgroups used in the phylogenetic analyses of pSSU and pgp60 sequence data sets were C. molnari (GenBank: HM243547) and C. meleagridis subtype IIId (GenBank: DQ067570.1), respectively. The Chi-square test was performed using SPSS Statistics 24 software (IBM, New York, USA).

Results

All 471 individual faecal samples from chickens were analysed molecularly for the presence of Cryptosporidium species, genotypes and subtypes. The pSSU amplicons were generated from 48 of the 471 DNA samples, equating to an overall prevalence of Cryptosporidium of 10.2% (Table 1), with prevalence values ranging from 5.0% to 18.3% on eight of 14 farms from the five of the six geographical regions (Table 1). Cryptosporidium was detected exclusively in layer chickens, but not in broilers (Table 1).

Cryptosporidium was detected in both age groups (Table 2), and young chickens (≤ 4 months) tended to have a higher infection rate (15.1%) than the chickens of > 4 months (11.7%). Cryptosporidium baileyi and C. meleagridis were detected in birds of both age groups. The prevalence of C. baileyi seemed higher in older chickens (> 4 months), although there was no statistical difference between age groups (χ2 = 0.75, df = 1, P = 0.387). In contrast, the prevalence of C. meleagridis was higher in younger chickens (≤ 4 months) (χ2 = 7.8, df = 1, P = 0.005) (Table 2).

Table 2 Age groups of chickens in which Cryptosporidium species were detected using PCR-based tools (cf. Table 1)
Full size table

The identification of Cryptosporidium species and genotypes was achieved through the sequencing of pSSU amplicons (n = 48). This analysis revealed C. baileyi in 68.8% and C. meleagridis in 31.2% of the 48 samples; no mixed-species infections were detected. Nine distinct pSSU sequences that represented all 48 samples and both Cryptosporidium species were deposited under GenBank accession numbers MG969393-MG969401, and the relationships of these sequences with selected reference sequences from GenBank (Additional file 1: Table S1) were established through a phylogenetic analysis (Fig. 2). Specifically, eight of these sequences (GenBank: MG969393-MG969400) clustered with known pSSU reference sequences for C. baileyi, and the other one (GenBank: MG969401) clustered with a representative sequence for C. meleagridis, on a branch with ‘zoonotic’ species including C. felis, C. hominis and C. parvum (see Fig. 2).

Fig. 2
figure 2

Relationships among Cryptosporidium taxa inferred from the phylogenetic analysis of sequences from a portion of the small subunit of the nuclear ribosomal RNA gene (pSSU) by Bayesian inference (BI). Posterior probabilities of > 0.95 are indicated at all major nodes. Bold-type indicates Cryptosporidium species or genotypes characterised from faecal DNA samples in this study. The GenBank accession number precedes the species designation; the number of samples of a particular species/genotype is indicated in parentheses. The scale-bar represents the number of substitutions per site. Cryptosporidium molnari was used as an outgroup

Full size image

Subty** was achieved through an analysis of sequence data derived from gp60 amplicons obtained (n = 13; deposited under GenBank: MG969387-MG969392). Sequence alignment and phylogenetic analysis of the sequences revealed six new subtypes of C. meleagridis in chickens. This analysis defined two subtype families (IIIb and IIIe; Fig. 3); the commonest subtype family IIIb [IIIbA22G1R1c (n = 6); IIIbA23G1R1d (n = 2) and IIIbA26G1R1b (n = 1)] was identified for nine samples, and subtype family IIIe [IIIeA17G2R1 (n = 1), IIIeA19G2R1 (n = 1) and IIIeA26G2R1 (n = 2)] for four samples. The phylogenetic analysis showed that nine sequences representing C. meleagridis subtype IIIb (GenBank: MG969390-MG969392) clustered with pgp60 sequences (GenBank: KY575457-KY575459) derived from samples from diarrhoeic children from Wuhan [23]. The pgp60 gene sequences representing C. meleagridis subtype IIIb from chickens (GenBank: MG969390-MG969392) showed high sequence similarity (92.7–100%) with those from humans (GenBank: KY575457-KY575459), being associated with subtypes IIIbA22G1R1c, IIIbA23G1R1d and IIIbA26G1R1b.

Fig. 3
figure 3

Relationships among Cryptosporidium meleagridis subtypes inferred from the phylogenetic analysis of sequences from a portion of the nuclear 60-kDa glycoprotein gene (pgp60) by Bayesian inference (BI). Posterior probabilities of > 0.95 are indicated at all major nodes. Bold-type indicates subtypes characterised from faecal DNA samples tested in this study. The GenBank accession number precedes the species designation; the number of samples of a particular species/genotype is indicated in parentheses. The scale-bar represents the number of substitutions per site. Subtype IIIhA7 was used as an outgroup

Full size image

Discussion

The present molecular investigation reports, for the first time, the presence and prevalence of Cryptosporidium in chickens on commercial farms in Hubei Province, although previous studies have recorded Cryptosporidium in chickens in Henan [40] and Zhejiang [41]. Here, both C. baileyi and C. meleagridis were identified using PCR-based tools. The overall prevalence of Cryptosporidium in chickens was ~10%, which is consistent with percentages recorded previously in China (8.9%, Henan; 9.9%, Zhejiang) [40, 41] and Syria (9.9%) [42], higher than Jordan (4.8%) [43] and Tunisia (4.5%) [44], and lower than Brazil (12.6%) [45]. Cryptosporidium infection has been recorded mainly in broiler chickens in countries including China, Algeria, Germany, Iran, Syria and Tunisia [26, 41, 42, 44, 46, 47]. Published studies indicate that young birds are more frequently infected with Cryptosporidium than adults [28, 40, 46], but more work is needed to confirm such an age-related association.

Cryptosporidium baileyi was detected in most (69%) of the 48 test-positive chicken faecal samples. Cryptosporidium baileyi was originally isolated from commercial broiler chickens [48], has been recorded in a broad range of avian hosts and is considered to be a dominant species in birds, although other taxa including C. avium, C. galli and/or C. meleagridis occur, of which only the latter species is recognised as zoonotic [4, 30, 33, 58,59,60,61,33]. Within these families, subtypes IIIbA22G1R1c, IIIbA26G1R1b, IIIeA17G2R1, IIIeA19G2R1 and IIIeA26G2R1 have been recorded in humans [33, 68, 72], and were identified here in chickens in Hubei Province. Subtype IIIbA22G1R1c was the predominant subtype in this and our previous study of children [23]. This subtype has been detected in people with travel-acquired infections. For example, subtype IIIbA22G1R1c, detected in some people from Sweden, was linked to travel to India or Thailand [21, 33]. The same subtype had been recorded in an English patient proposed to have become infected in India (sample analysed in the UK) [33, 69]. This information highlights the potential of this pathogen and various IIIb subtypes to be spread worldwide through human travel. In this context, subtype IIIbA23G1R1d was found here, for the first time, in chickens. However, subtype IIIbA23G1R1 had been recorded in quail in Brazil, and a human in Peru [71, 73], and IIIbA23G1R1a, IIIbA23G1R1b and IIIbA23G1R1c in people with histories of travel to Thailand, Indonesia and Malaysia, respectively [21, 33, 74]. We propose that subtype IIIbA23G1R1d might be transmissible from poultry to humans, but this needs to be explored in detail.

Conclusions

This is the first molecular study reporting the genetic identity and prevalence of C. baileyi and C. meleagridis in chickens in Hubei. The findings suggest that C. meleagridis subtypes IIIbA26G1R1b and IIIbA22G1R1c are cross-transmissible between chickens and humans, raising awareness about the significance of birds as potential reservoirs of zoonotic variants of Cryptosporidium. Taken together, this information emphasises the need for epidemiological studies of Cryptosporidium, particularly C. meleagridis, in humans and in birds using accurate diagnostic and analytical tools utilising multiple informative genetic loci. Such investigations should be focused on assessing the transmission patterns and dynamics of cryptosporidiosis.