Introduction

Cryptosporidiosis is a zoonotic protozoan disease of worldwide distribution, affecting a wide range of vertebrate hosts. There are 16 valid species of Cryptosporidium so far, namely, Cryptosporidium andersoni, Cryptosporidium baileyi, Cryptosporidium bovis, Cryptosporidium canis, Cryptosporidium felis, Cryptosporidium galli, Cryptosporidium hominis, Cryptosporidium meleagridis, Cryptosporidium molnari, Cryptosporidium muris, Cryptosporidium parvum, Cryptosporidium varanii, Cryptosporidium scophthalmi, Cryptosporidium serpentis, Cryptosporidium suis, and Cryptosporidium wrairi. In addition, over 40 genotypes with no species names have been described (Xiao et al. 2004).

C. andersoni was separated from C. muris and was established as a new species based on molecular data and a transmission study (Lindsay et al. 2000; Xiao et al. 2004). To data, there are many reports of C. andersoni in cattle (Andersoni 1991; Enemark et al. 2002; Fayer et al. 2006; Koyama et al. 2005; Kvac et al. 2004; Kvac and Vitovec 2003, 2006; Masuno et al. 2006; Matsubayashi et al. 2004, 2005; Robinson et al. 2006; Satoh et al. 2003; Xiao et al. 2000). Cross-transmission study showed C. andersoni isolates from different geographic region had different infection characteristics. C. muris-like oocysts from cattle in the USA were not infectious to mice or even cattle (Andersoni 1991). A Danish C. andersoni isolate was not transmissible to mice, but oocysts were detected in the feces of one experimentally infected calf (Enemark et al. 2002). C. andersoni oocysts from cattle in the Czech Republic were not transmissible to neonatal or adult BALB/C mice, SCID mice, common voles, bank voles, common field mice, desert gerbils, guinea pigs, rabbits, lambs, or goats; only cattle, Mongolian gerbils, and southern multimammate mice became infected (Koudela et al. 1998; Kvac et al. 2007). Contrary to other findings, C. andersoni isolates in Japan were found to be infective to SCID mice (Matsubayashi et al. 2004; Koyama et al. 2005; Satoh et al. 2003).

In the present study, we have characterized two C. andersoni isolates from a bactrian camel (Camelus bactrianus) in China by sequence analyses of four genes and by experimental infections of laboratory animals. Results of the study suggest that the Chinese isolates were biologically more similar to European and American C. andersoni isolates.

Materials and methods

C. andersoni isolates

Two fecal samples were collected from a bactrian camel (3-year-old) in the Wild Animals Rescue Centre of Henan Province in China in November 2005 and April 2006. Both samples were positive for Cryptosporidium by bright filed microscopy under 400× and 1,000× after oocysts were concentrated by the Sheather’s sugar flotation technique. Oocysts were purified from the samples by discontinuous sucrose gradient centrifugation as described previously (Arrowood and Sterling 1987; Arrowood and Donaldson 1996). The lengths and widths (to the nearest micrometer) of 50 oocysts (each isolate) were measured and the shape index (the ratio of length to width) of each oocyst was calculated. Oocysts were stored in 2.5% potassium dichromate solution at 4°C before they were used in experimental infections within 2 weeks.

DNA extraction

Genomic DNA was isolated from the purified oocysts using Mag Extractor-Genome kit (Toyobo, Osaka, Japan), based on chaotropic extraction followed by absorption onto silica-coated magnetic beads. Briefly, 50 μl of oocyst suspension was resuspended in 750 μl of lysis buffer. After five cycles of freeze–thaw (−80°C for 5 min and 37°C for 5 min), 40 μl of silica-coated magnetic beads were added to the oocyst lysate, and the tube was vortexed vigorously for 10 min. The magnetic beads were then separated from the suspension using a magnet and washed twice in 900 μl of washing buffer and once in 900 μl of 70% ethanol. Afterwards, the magnetic beads were resuspended in 25–100 μl of double-distilled water, and the tube were vortexed vigorously for 10 min. The bead suspension was then centrifuged at 2,000×g for 2 min, and the supernatant was collected. The supernatant containing DNA was kept at −20°C before it was used in polymerase chain reaction (PCR) analysis.

18S rRNA and HSP70 genes amplification and sequencing

Primers and amplification conditions used in nested-PCR analysis of the partial 18S rRNA and HSP70 gene were previously described (Sulaiman et al. 2000; Xiao et al. 1999a). KOD-Plus amplification enzyme (Toyobo, Kita-ku, Osaka, Japan) was used for PCR amplification. The PCR products were examined by agarose gel electrophoresis and visualized after ethidium bromide staining. Direct sequencing of the PCR products was done by TaKaRa Biotechnology (Dalian, China) on an ABI PRISMTM 3730 XL DNA Analyzer (Applied Biosystems) using the Big Dye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems). Sequence accuracy was confirmed by two-directional sequencing and by sequencing a new PCR product if necessary.

Actin and COWP genes amplification, cloning, and sequencing

Primers and amplification conditions used in nested-PCR analyses of the partial actin and COWP gene were previously described (Sulaiman et al. 2002; Xiao et al. 2000). TaKaRa Ex Taq® enzyme (TaKaRa Biotechnology) was used in PCR. PCR products were purified using a Mag Extractor-PCR and Gel Clean up-kit (Toyobo). Purified PCR products were cloned into the PMD19-T vector (TaKaRa Biotechnology) as recommended by the supplier. Positive clones were screened by PCR, and four to five recombinant plasmids per product were sequenced as described above. The nucleotide sequences of the partial 18S rRNA, HSP70, actin and COWP genes of C. andersoni from the bactrian camel were deposited in the GenBank database under accession numbers DQ989570 to DQ989577.

Phylogenetic analysis

Nucleotide sequences were aligned using the ClustalX 1.81 (Thompson et al. 1997), with manual adjustment. Phylogenetic analyses were performed using the software PHYLIP version 3.64 (Felsenstein 1989). Neighbor-joining trees were constructed based on the evolutionary distances calculated using the Kimura two-parameter model. Parsimony trees were constructed using the Program Dnapars and the more thorough search. The reliability of these trees was assessed using the bootstrap analysis with 1,000 replicates, with values above 50% reported. Phylograms were drawn using the Tree View program version 1.65 (Page 1996).

Cross-transmission study

Fifty Kun-ming 18-day-old mice, six 5-week-old SCID mice (Shanghai SLAC Laboratory Animal, Shanghai, China), and two 8-day-old calves were inoculated with C. andersoni oocysts. Before the inoculation, fecal samples from all animals were examined daily for 10 days using Sheather’s sugar flotation technique. Only Cryptosporidium-negative animals were used in the experimental infection. Immunosuppression was induced in Kun-ming mice using dexamethasone at 0.06 mg per day 10 days before and 45 days after the inoculation, while immunosuppression was induced in one calf using dexamethasone at 28 mg per day 10 days before and 14 days after the inoculation. The doses used in the experimental infections are shown in Table 1.

Table 1 The experimental design for cross-transmission studies of C. andersoni oocysts
Full size table

Results

Sequence characterizations of the two C. andersoni isolates

For each of the gene targets, sequences of the two Cryptosporidium isolates were aligned with those from other C. andersoni isolates using ClustalX 1.81 (Thompson et al. 1997). There were no sequence differences between the two Cryptosporidium isolates from the camel. The sequences of the 18S rRNA gene were identical to those obtained previously from C. andersoni in other countries (AF093496 for a USA isolate and AB089285 for Japanese isolate). Likewise, the COWP sequences obtained were also identical to those previously obtained from cattle (AF266262 for a Czech isolate and AB089289 for a Japanese isolate). The HSP70 sequences obtained were identical to a sequence obtained in Japan (AB089288), however, six nucleotide differences was observed with an Australia isolate (AF221542). In addition, there were eight nucleotide differences from a USA isolate (AF382352) in the actin gene (Fig. 1a,b).

Fig. 1
figure 1

Sequence differences between the C. andersoni camel isolate in this study and other C. andersoni cattle isolates in the partial HSP70 (a) and actin (b) genes. The cattle reference sequences used are AF221542 for the HSP70 gene of an Australian isolate and AF382352 for the actin gene of an American isolate. Dots denote sequence identity to the bovine sequence. Dashes represent nucleotide deletions

Full size image

Phylogenetic analysis

Neighbor-joining and parsimony trees were constructed from aligned partial 18S rRNA, HSP70, COWP, and actin sequences of these two C. andersoni isolates and those downloaded from the GenBank database (Fig. 2a,b,c,d). In the 18S rRNA, HSP70 and COWP loci, the two isolates formed a cluster with C. andersoni, which was supported by bootstrap analysis. While in the actin gene, the two isolates clustered together and formed a sister clade with C. andersoni.

Fig. 2
figure 2

Phylogenetic relationship of Cryptosporidium parasites inferred by neighbor-joining analysis of the 18S rRNA (a), HSP70 (b), COWP (c), and actin genes (d) based on Kimura 2-parameter model. Bootstrap values (in percentage) above 50% from 1,000 pseudo-replicates are shown for both the neighbor-joining (the first value) and maximum parsimony analyses (the second value). Weak Node supported by method but very weak (<50%), ns node not supported by method. Scale bar indicates an evolutionary distance of 0.1 substitutions per site in the sequence

Full size image

Morphometric measurements

Oocysts of C. andersoni from the bactrian camel were ellipsoid and lacked sporocysts. The mean length and width of oocysts of the isolate 0511 were 7.0 and 5.1 μm, respectively, with a mean shape index of 1.37. The mean length and width of oocysts of the isolate 0604 was 7.2 and 5.2 μm, respectively, with a mean shape index of 1.39 (Table 2).

Table 2 Morphometric measurements (micrometers) of two C. andersoni isolates from a bactrian camel
Full size table

Cross-transmission results

All fecal samples from inoculated animals were negative for Cryptosporidium spp. as examined by the Sheather’s sugar flotation technique.

Discussion

Results of the genetic characterizations suggests that the Cryptosporidium isolates derived from a bactrian camel in China belonged to C. andersoni. In phylogenetic analyses of the 18S rRNA, HSP70 and COWP genes, the Cryptosporidium isolates formed a cluster with C. andersoni, which was supported by bootstrap analysis. At the actin locus, the present Cryptosporidium parasites clustered together and formed a sister taxon with C. andersoni. Similarity analysis of nucleotide sequences obtained showed that the camel Cryptosporidium isolates and C. andersoni were identical at the 18S rRNA and COWP loci, 100% with Japanese isolates, and 99.7% with an Australian isolate at the HSP70 locus and 99.2% at actin locus. The size of oocysts from the two isolates were also very close to the reported size for C. andersoni (mean 7.4 × 5.5 μm; range 6.0–8.1 × 5.0–6.5 μm; shape index 1.35; Lindsay et al. 2000). This is the first report of Cryptosporidium in bactrian camel in China, although Cryptosporidium spp. have been reported in bactrian camels in other countries (Morgan et al. 2000; Sulaiman et al. 2000; Xiao et al. 1999b).

To date, Cryptosporidium andersoni oocysts were not transmissible to neonatal or adult BALB/C mice, SCID mice, common voles, bank voles, common field mice, desert gerbils, guinea pigs, rabbits, lambs, or goats; only cattle, Mongolian gerbils and southern multimammate mice became infected (Andersoni 1991; Enemark et al. 2002; Koudela et al. 1998; Kvac et al. 2007). In contrast, bovine C. andersoni isolates in Japan were infectious to SCID mice, thus was named as a novel type of C. andersoni (Matsubayashi et al. 2004; Koyama et al. 2005; Satoh et al. 2003).

In this study, C. andersoni derived from a bactrian camel was not infectious to immunosuppressed or immunocompetent Kun-ming mice, SCID mice, immunosuppressed or immunocompetent calves. Such results were similar to other C. andersoni isolates from Europe and America, although there were a few nucleotide differences between the isolate in this study and the USA isolate (Fig. 1b) at the actin locus. In contrast, the results were different from the C. andersoni isolates from cattle in Japan in that the latter could infect SCID mice. Interestingly, there were no sequence differences among these isolates, Europe, America, and Japan isolates at 18S rRNA, HSP70, and COWP loci. On the other hand, the transmission study in Australia was unavailable, although some nucleotide differences were found between the isolates in this study and the Australia isolates in the HSP70 gene (Fig. 1a). The reason for the different transmission behavior is therefore unknown.