Cryptosporidium species generally lack distinguishing morphological traits, and consequently, molecular methods are commonly used for parasite identification. Various methods for Cryptosporidium identification have been proposed, each with their advantages and disadvantages. In this study, we show that capillary electrophoresis coupled with single-strand conformation polymorphism (CE-SSCP) is a rapid, simple and cost-effective method for the identification of Cryptosporidium species and genotypes. Species could be readily differentiated based on the SSCP mobility of amplified 18S rRNA gene molecules. Clones that differed by single-nucleotide polymorphisms could be distinguished on CE-SSCP mobility. Profiles of species known to have heterogenic copies of 18S rRNA gene contained multiple peaks. Cloning and sequencing of Cryptosporidium parvum, Cryptosporidium hominis, Cryptosporidium fayeri and Cryptosporidium possum genotype 18S rRNA gene amplicons confirmed that these multiple peaks represented type A and type B 18S rRNA gene copies. CE-SSCP provides a reliable and sensitive analysis for epidemiological studies, environmental detection and diversity screening.
18S rRNA gene
Cryptosporidium is a genus of apicomplexan protozoan parasites that has been identified in more than 150 vertebrate hosts (Fayer et al., 2000). There are 21 currently recognized species of Cryptosporidium with several host groups susceptible to more than one species: Cryptosporidium serpentis and Cryptosporidium varanii (syn. Cryptosporidium saurophilum) in reptiles; Cryptosporidium molnari and Cryptosporidium scophthalmi in fish; Cryptosporidium fragile in frogs; Cryptosporidium baileyi and Cryptosporidium galli in birds; Cryptosporidium meleagridis in birds and humans; Cryptosporidium fayeri and Cryptosporidium macropodum in marsupials; Cryptosporidium suis in pigs; Cryptosporidium muris and Cryptosporidium wrairi in rodents; Cryptosporidium bovis, Cryptosporidium ryanae and Cryptosporidium andersoni in cattle; Cryptosporidium xiaoi in sheep; Cryptosporidium felis in cats; Cryptosporidium canis in dogs; Cryptosporidium hominis in humans; and Cryptosporidium parvum in humans and ruminants (Fayer et al., 2000, 2001, 2005; Alvarez-Pellitero & Sitja-Bobadilla, 2002; Ryan et al., 2003a–c, 2008; Jirku et al., 2008; O'Brien et al., 2008; Power & Ryan, 2008; Fayer & Santin, 2009). Molecular methods have shown that the genus is more diverse than previously thought, with >40 cryptic species identified using molecular markers.
The identification of Cryptosporidium species using morphological characters is problematic. The small size of Cryptosporidium oocysts makes examination of the internal structures difficult (Fayer et al., 2000), and the similarities in oocyst size of many Cryptosporidium species prevent ready identification (Fall et al., 2003). To overcome these limitations, Cryptosporidium identification and differentiation is commonly achieved using molecular approaches. Cryptosporidium species have been differentiated using sequence analysis of a variety of loci. The more commonly used loci include 18S ribosomal DNA (18S rRNA gene) (Morgan et al., 1997, 1998; Xiao et al., 1999b), heat shock protein 70 (Sulaiman et al., 1999) and actin (Sulaiman et al., 2000). However, the high costs of DNA sequencing have led to the development of more rapid and inexpensive gel-based electrophoretic methods for species differentiation. Both restriction fragment length polymorphism (RFLP) (Spano et al., 1997; Morgan et al., 1999; Patel et al., 1999) and single-stranded conformation polymorphism (SSCP) have been used to identify the genetic variation in 20 Cryptosporidium species (Jex et al., 2007a) and for investigating the intraspecies variation in C. parvum and C. hominis (Gasser et al., 2004; Jex et al., 2007b).
Capillary electrophoresis coupled to RFLP (terminal RFLP) and SSCP (CE-SSCP) have proven to be more reliable and sensitive than analysis by conventional gel electrophoresis. In this study, we investigated the ability of CE-SSCP on the 18S rRNA gene to discriminate between species and genotypes of Cryptosporidium both within host groups and between host groups.
Materials and methods
Parasite DNA and 18S rRNA gene amplification
Genomic DNA from 28 Cryptosporidium isolates representing 15 species and genotypes were used in this study (Table 1). All samples had been identified previously to the species level by sequence analysis of multiple loci. DNA was extracted using the Qiagen stool kit or prepGEM™ (Zygem Corporation Ltd, Hamilton, New Zealand) (Ferrari et al., 2007). Amplification and sequencing of an ∼300-bp fragment of the 18S rRNA gene was performed using a previously described nested PCR protocol (Ryan et al., 2003a–c), with minor modifications. Primary reactions consisted of 20 pM of the following primers: 18S CF2 5′-GACATATCATTCAAGTTTCTGACC-3′ and 18S CR2 5′-CTGAAGGAGTAAGGAACAACC-3′, 1 × PCR buffer, 20 mM DMSO, 200 uM dNTPs, 1 U Accutaq (Sigma) and 2 μL of DNA template. Cycling conditions comprised 94 °C for 2 min, 58 °C for 1 min and 68 °C for 2 min, followed by 35 cycles, each consisting of 94 °C for 40 s, 58 °C for 30 s and 68 °C for 30 s and a final extension step of 68 °C for 7 min.
Isolates tested and the apparent mobilities of the 18S rRNA gene amplicon from various Cryptosporidium species determined using CE-SSCP against the internal size standard LIZ5001
No. of isolates
Farm animals and humans
Dogs and humans
Birds and humans
Avian genotype I
Avian genotype II
Rodents and humans
Mouse genotype I
Mouse genotype II
*Cloning identified this peak to be a third copy of the 18S rRNA and not an ambiguity in the trace.
Secondary reactions were performed using 1 μL of a 1/20 dilution of primary PCR product as a template and the primers 18SIF 5′-AGTGACAAGAAATAACAATACAGG-3′ and 18SIR 5′-CCTGCTTTAAGCACTCTAATTTTC-3′. For fluorescence detection of SSCP products, primer 18SIF was labeled at the 5′- end with 6-FAM (Proligo, Australia). The secondary reactions were performed in a total volume of 50 μL with reaction constituents and cycling conditions identical to those used for primary reactions. PCR products were purified using the Qiagen spin column PCR purification kit (Qiagen, Hilden, Germany) and DNA concentrations were determined using a Biophotometer (Eppendorf, Australia).
For CE-SSCP analysis, 1 μL of PCR product containing ∼1 ng of DNA was combined with 9.9 μL HiDi formamide (Applied Biosystems, Foster City, CA) and 0.1 μL of the internal lane standard LIZ500 (Applied Biosystems). Samples were denatured at 99 °C for 10 min and then snap chilled on ice for 10 min. Samples were run on an ABI 3130xl capillary electrophoresis analyzer and separated using 6% or 7% Conformation Analysis Polymer prepared as per the manufacturer's instructions using supplied buffer (Applied Biosystems). Three run temperatures of 20, 25 and 30 °C were tested to determine the optimal temperature for species differentiation. Samples were injected for 15 s at 1.6 kV and run for 50 min. Analysis was performed using genemapper v 4.0 software (Applied Biosystems).
Assessment of 18S rRNA gene diversity
CE-SSCP analysis of amplified 18S rRNA gene generated multiple peaks for five Cryptosporidium species. To determine whether these peaks represented distinct sequences types, C. parvum, C. hominis, C. fayeri and C. sp. possum genotypes were cloned using the TA TOPO vector cloning system (Invitrogen, CA). For cloning, amplifications of the 18S rRNA gene using the primers described above were performed with RedHot Taq polymerase (Abgene, Surrey, UK) to facilitate TA cloning. PCR inserts from positive transformants were amplified using the CE-SSCP 18S rRNA gene protocol as above and their mobilities were determined using CE-SSCP. Clone inserts that differed in SSCP mobility were selected for sequencing, which was performed using the ABI 3130xl Genetic analyzer (Applied Biosystems) with the BigDye™ terminator kit (Applied Biosystems). The sequences of clones were subjected to blastn searches and aligned using clustalw (Thompson et al., 1994). The nucleotide sequences for the clones generated in this study were submitted to GenBank under the accession numbers FJ218151-FJ218162.
The average mobility and SE of the mean of the mobilities of six isolates of both C. parvum and C. hominis were determined within a single run and across three runs. Microsoft excel was used to generate the average mobility, peak separation and SE of the means.
A fragment of the 18S rRNA gene was amplified using genomic DNA from 10 recognized Cryptosporidium species and five cryptic species (Table 1). For all samples, PCR generated clear products ranging from 289 to 296 bp when analyzed using agarose electrophoresis (data not shown). Optimal CE-SSCP conditions, in terms of the separation and sharpness of individual peaks, enabled the selection of standard conditions of 25 °C, 7% conformation polymer and capillary loading of 0.1–1 ng of sample for subsequent CE-SSCP runs.
Analysis of 18S rRNA gene amplicons from the Cryptosporidium samples using CE-SSCP resulted in defined peaks with mobilities ranging from 300 to 345 compared with the Liz500 internal standards (Table 2). There was some variation in sample mobility between runs, of between 2 and 10 U. Although the absolute mobility values differed slightly from run to run, the relative difference in the mobilities between different samples was consistent for each species, and for multiple peaks where these occurred within a single sample. For example, the major peaks of C. parvum and C. hominis consistently migrated 6-bp apart in any run (Table 2). Despite between-run variation, apparent mobilities were consistent within and across runs for multiple isolates of C. parvum and C. hominis (Table 2). To control for run-to-run variation, C. parvum and C. hominis were used as reference control isolates in all CE-SSCP runs. The relative mobilities of CE-SSCP peaks from test samples were then calibrated to the apparent mobility of major peaks of C. parvum and C. hominis. These were set at 317 and 323 U, respectively.
Variation in CE-SSCP migration of six isolates of Cryptosporidium parvum and Cryptosporidium hominis within a single run and between three runs
Number of isolates
The mobility of the major peaks allowed Cryptosporidium species from within host groups to be discriminated. For example, the three species of most concern to humans, C. parvum, C. hominis and C. meleagridis, had major peaks at 317, 323 and 318, respectively (Table 2). The three species/genotypes from marsupials, C. fayeri, C. macropodum and the C. sp. possum genotype, could also be differentiated by the mobility of major peaks (Table 1). However, there was only a single unit difference in the mobilities between C. fayeri and C. macropodum from marsupials, and C. parvum and C. meleagridis from humans. The presence of two peaks provided an additional means of differentiation, making it possible to separate these species (Table 1). The three species/genotypes found in rodents, C. muris and mouse genotypes I and II had peaks of 307, 326 and 322, respectively, and could be differentiated readily by CE-SSCP (Table 1).
Some species, specific to hosts from different vertebrate orders, could not be differentiated, such as C. macropodum and C. canis, which both had apparent mobilities of 312. Three additional species, C. muris, C. andersoni and the C. sp. possum genotype, had major peak mobilities of 307. The C. sp. possum genotype consistently exhibited a secondary peak, with an apparent mobility of 342, enabling differentiation from the two species with similar mobilities, C. muris and C. andersoni, but the latter two species could not be differentiated. The mobilities of C. muris and C. andersoni were also very similar to the single peak of C. serpentis, with a mobility of 306. For birds, C. baileyi, C. meleagridis and avian II could be differentiated by the mobility of primary peaks. However, the mobility of the primary peaks for C. baileyi and avian genotype I differed only by a single unit, but the presence of a secondary peak enables differentiation. Nucleotide sequence alignments for the partial 18S rRNA gene region of species and genotypes in this study showed that variability ranged from as few as 5 bp (C. hominis and C. parvum, and C.muris and C. andersoni) up to 46 bp between C. andersoni and C. parvum (Fig. 3).
Alignment of the variable regions of 18S rRNA gene clones of C.hominis, C. parvum, Cryptosporidium fayeri and Cryptosporidium possum genotype against 18S rRNA gene type A and type B reference strains of Cryptosporidium parvum, Cryptosporidium fayeri, Cryptosporidium possum genotype type A and type B and Cryptosporidium hominis polyT region. Sequences of clones confirmed that the mobility of multiple peaks detected by CE-SSCP corresponded to heterogeneity within the 18S rRNA gene. The positions of the variable regions of bp 639 and 689 were determined against C. parvum reference sequences GenBank accession no. AF108864.
For each species with multiple peaks, the unit differences between the peaks were consistent between runs. For example the two C. parvum peaks were consistently separated by 5 U within a run, between runs, between different samples and between replicate PCRs (Table 2). The presence of two peaks in some species/genotypes was most probably caused by polymorphisms in the 18S rRNA gene multigene family. This was investigated by cloning amplicons from four species where multiple peaks were consistently detected, these being C. parvum, C. hominis, C. fayeri and C. sp. possum genotype. Clones were screened using CE-SSCP and those with apparent mobilities corresponding to one of the multiple peaks from the initial SSCP run were sequenced. Multiple alignments of cloned sequences and GenBank reference isolates showed that for C. parvum the two peaks represented type A and type B 18S rRNA gene copies. Type A clones had a mobility of 322 and type B 317. The peak height for type A 18S rRNA gene clones was approximately fourfold higher than type B (Fig. 1). Similarly, for C. fayeri, which exhibited three peaks, clones represented type A and type B, but a minor third type was also identified (Fig. 2). For C. fayeri clones, the variable region from bp 639 corresponded to type A 18S rRNA gene (mobility 313) and the region from bp 689 to type B 18S rRNA gene (mobility 317) (Fig. 2). The third peak had the lowest peak height and a mobility of 318 (Fig. 1). Similarly, the two peaks present in the Crytosporidium sp. possum genotype corresponded to type A (307) and type B (342) 18S rRNA gene copies. For C. hominis, differences in apparent mobility were related to the number of thymidine residues in the poly-T region, which ranged from 7 to 11 (Fig. 2).
CE-SSCP traces of representative Cryptosporidium species from humans and marsupials. Multiple peaks represent the polymorphism within the 18S rRNA gene and enable differentiation between species from similar hosts exhibiting similar mobilities.
Nucleotide sequence alignment of partial regions of the primary 18S rRNA gene copies for the 15 species and genotypes analyzed by CE-SSCP. Nucleotides that are identical to consensus are defined by a dot (.) and deletions by dash (-).
This study is the first to report on the application of capillary electrophoresis analyses of SSCP for the differentiation of Cryptosporidium species. Although SSCP has been used previously to differentiate Cryptosporidum species, analyses were performed using conventional nondenaturing gel electrophoresis that relied on manual scoring of band mobilities against a reference control (Jex et al., 2007a, b). In our hands, CE-SSCP provides a method for the differentiation of Cryptosporidium species both within host groups and between host groups.
The Cryptosporidium CE-SSCP electropherograms comprise a major peak that corresponds to a single strand of a fluorescently labeled PCR template. For four species, additional minor peaks were detected. Cloning and sequencing confirmed that multiple peaks corresponded to polymorphism in the 18S rRNA gene. For C. parvum and C. fayeri the peaks corresponded to type A and type B copies of the 18S rRNA gene (Le Blancq Sylvie et al., 1997; Xiao et al., 1999a, b). A third peak in C. fayeri samples appeared to be a recombinant between type A and type B 18 s rRNA gene copies; however, it is also possible that this peak was a PCR chimera of type A and type B. Similarly, the two peaks observed in the Cryptosporidium possum genotype corresponded to the 18S rRNA gene polymorphism (Hill et al., 2008). For C. hominis isolates, the two peaks observed corresponded to variations in the poly-T region of the 18S rRNA gene. The inter- and intraisolate variation for the poly-T region has been shown to range in thymidine numbers from 8 to 12 (Gibbons-Matthews & Prescott, 2003). Inter- and intraspecies variations in the poly-T region, observed in clones from five C. hominis isolates, corresponded to differences in CE-SSCP mobility. Although several Cryptosporidium species have heterogeneic copies of the 18S rRNA gene, the regions complementary to PCR primers are conserved, and hence a second fluorescent peak is present in amplicons and detectable by CE.
The three species of concern to human health, C. parvum, C. hominis and C. meleagridis, were clearly discernable by CE-SSCP, and the multiple peaks observed in C. parvum and C. hominis provided extra discriminatory power. The ability of CE-SSCP to clearly identify multiple peaks in samples indicates that it may be applicable for the detection of mixed infections. However, current PCR protocols need to be optimized for mixed species detection because preferential amplification of C. parvum has been observed in the past. Mixes of C. parvum and C. hominis DNA over a range of concentrations have shown that C. parvum is preferentially amplified (Cama et al., 2007; Waldron et al., 2009). CE-SSCP could be applied to evaluate PCR protocols for the detection of mixed infections. Additionally, CE-SSCP could be assessed for collecting quantitative data based on peak intensity comparisons of mixed species.
Some species within a host group and across host groups could not be differentiated by CE-SSCP. These species tend to be closely related and differ by as few as five nucleotides, such as C. muris and C. andersoni. As the 18S rRNA gene is highly conserved, a locus that has greater variation such as actin (Sulaiman et al., 2000) may enable the differentiation of all species and strains.
Although some species had multiple peaks, consistent separation and analysis using genemapper software provides a less subjective scoring method than the visual assessment of gel electrophoresis. In contrast to the numbers of peaks detected in by CE, multiple bands, which range from three to eight, are detected when using conventional gel electrophoresis (Gasser et al., 2004; Jex et al., 2007a). Applications of SSCP for Cryptosporidium differentiation using 18S rRNA gene have not attempted to identify what the multiple bands represent, but it is likely that they are the sense and antisense strands of the type A and type B copies of the 18S rRNA gene. In CE-SSCP, only one strand is analyzed when a single fluorescent primer is used for amplifications, as performed in this study. Performing CE-SSCP with a second labeled primer would allow both sense and antisense strands to be analyzed concurrently.
Previous applications using CE have reported a run-to-run variation that has been controlled for using reference isolates (Gillings et al., 2008; Waldron et al., 2009). In this study, the absolute mobility unit for each species differed from 2 to 10 U between CE-SCCP runs, but relative mobility was consistent for all isolates within a run. The observed shifts in mobility are likely to arise from instrument factors such as variation in polymer preparation, the concentration of sample that is loaded, slight temperature fluctuations and capillary maintenance. These variables can be controlled for using a size marker and a set of reference samples with a range of mobilities that can then be used to correct the mobilities of test samples for each run.
In recent years, molecular studies of Cryptosporidium have resulted in the identification of more than 40 cryptic species/genotypes (Xiao et al., 1999a, b, 2003; Ryan et al., 2003a–c; Power et al., 2004; Zhou et al., 2004; Hill et al., 2008). Establishment of a mobility reference bank using repeated testing of described species will enable CE-SSCP prescreening and selection of variants for subsequent sequencing. At our facility, prescreening using CE-SSCP represents a threefold cost saving per sample compared with DNA sequencing. Its application to epidemiological studies will decrease the sample processing times and minimize sequencing costs. At present, genetic analyzers are expensive and the sample run time is limited by the number of samples that can be processed (commonly 16 per run). As detection and diagnostic capabilities move toward molecular-based methods, Gene analyzers must also evolve to become routine instrumentation. Tests that are simple, reliable, reproducible, sensitive and cost effective will become necessary with advancing instrumentation.
We have described a CE-based method for differentiating Cryptosporidium species from within and between host groups. Genetic variation for other parasitic species has been investigated using SSCP (Gasser & Chilton, 2001; Hutson et al., 2004; Mahnaz et al., 2006; Lin et al., 2007), suggesting that CE would also be useful for other parasites. We are currently assessing CE-SSCP for use with different Cryptosporidium loci and as a tool for assessing the biodiversity of this genus. Applications of this rapid method to detection, population genetics and identification will increase our understanding of the evolution and diversity of this important parasitic group.
Funding for this research was provided through the Macquarie University Research Fellow Scheme and an Australian Research Council Linkage grant in collaboration with NSW Health.
(2006) Leishmania major: genetic heterogeneity of Iranian isolates by single-strand conformation polymorphism and sequence analysis of ribosomal DNA internal transcribed spacer. Acta Tropica 98: 52–58.
(1999) The identification of Cryptosporidium species and Cryptosporidium parvum directly from whole faeces analysis of a multiplex PCR of the 18S rRNA gene and by PCR/RFLP of the Cryptosporidium outer wall protein (COWP) gene. Int J Parasitol 29: 1241–1247.
(1997) PCR-RFLP analysis of the Cryptosporidium oocyst wall protein (COWP) gene discriminates between C. wrairi and C. parvum, and between C. parvum isolates of human and animal origin. FEMS Microbiol Lett 150: 209–217.
(1999) Nucleotide sequence characterization of HSP70 heat-shock gene differentiates various Cryptosporidium species and genotypes of Cryptosporidium parvum. SO – Abstracts of the General Meeting of the American Society for Microbiology, 99, 1999, p. 178.
(1994) CLUSTALW: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673–4680.