INTRODUCTION
In the life-cycle of Cryptosporidium parasites, the infection initiates with the ingestion of oocysts and the release of infectious sporozoites into the gastro-intestinal (GI) tract. Consistent with the dissemination of oocysts in the environment, and with their characteristic longevity (Yang et al. Reference Yang, Healey and Du1996; Widmer et al. Reference Widmer, Orbacz and Tzipori1999; Jenkins et al. Reference Jenkins, Trout, Higgins, Dorsch, Veal and Fayer2003), oocysts have evolved to maximize sporozoite survival and protect the sporozoites from the moment they are released from the intestinal epithelial cell, until they are ingested by a new host and reach the preferred organ of the GI tract. The precisely timed release of the 4 sporozoites contained in each oocyst is essential for parasite survival, as sporozoites are short lived and their mobility is limited. Depending on the species, sporozoites are released in the stomach or the small intestine. A small number of species, including C. muris and C. andersoni, multiply in the pits of the stomach and abomasum (Tyzzer, Reference Tyzzer1907; Lindsay et al. Reference Lindsay, Upton, Owens, Morgan, Mead and Blagburn2000). In contrast, the majority of Cryptosporidium species, including the most important agents of human cryptosporidiosis C. parvum and C. hominis, infect primarily the small intestine, with occasional involvement of the bile duct and colon.
Whichever the evolutionary mechanism which has led to the partition of the genus in a gastric and intestinal clade, the mechanism controlling oocyst excystation must be under strong selection to maximize the probability of infection. Since the infectious dose can be fewer than 10 oocysts (Okhuysen et al. Reference Okhuysen, Chappell, Crabb, Sterling and Dupont1999), the precise release of sporozoites in the right organ of the GI tract is essential to parasite survival. Oocysts from gastric and intestinal species can be expected to have adapted to different requirements; oocysts from gastric species to promptly release the sporozoites when exposed to body temperature and to stomach acidity, whereas the trigger for intestinal species can be expected to be more complex, as the oocyst must transit through the stomach and reach the duodenum before excysting. Premature release of sporozoites adapted to the condition of the small intestine would cause the sporozoite to be inactivated in the stomach's acid environment.
The conditions which induce excystation of C. parvum oocysts have been extensively investigated (Fayer and Leek, Reference Fayer and Leek1984; Woodmansee, Reference Woodmansee1987; Campbell et al. Reference Campbell, Robertson and Smith1992; Robertson et al. Reference Robertson, Campbell and Smith1993; Kato et al. Reference Kato, Jenkins, Ghiorse and Bowman2001). The research described in these publications was primarily driven by the need to optimize in vitro excystation to generate sporozoites for research, and to determine oocyst viability. Our motivation to study this process in different Cryptosporidium species was to improve our understanding of the evolution of different organ tropisms in this genus, and the adaptation of pathogens to the host environment in general. Although not significant human or veterinary pathogens, C. muris and C. andersoni are among the few species which have adapted to the extreme environment of the stomach, and the study of these parasites is expected to uncover unique mechanisms of adaptation. This study was undertaken to test the assumption that C. muris and C. andersoni oocysts have adapted to excyst in response to elevated temperature and acid, whereas C. parvum and C. hominis oocysts do not respond to acidity.
MATERIALS AND METHODS
Parasites
Table 1 lists the Cryptosporidium isolates used in this study. Three C. parvum, 1 C. hominis, 2 C. andersoni and 1 C. muris isolate were used. MD/TU114 is an isolate obtained by co-infecting a mouse with 2 C. parvum isolates MD and TU114. P2 is a recombinant line derived from such a mixed infection which differs genetically from MD and TU114 (Tanriverdi et al. Reference Tanriverdi, Blain, Deng, Ferdig and Widmer2007). TU502 is a C. hominis reference isolate, which was used for sequencing the C. hominis genome (Xu et al. Reference Xu, Widmer, Wang, Ozaki, Alves, Serrano, Puiu, Manque, Akiyoshi, Mackey, Pearson, Dear, Bankier, Peterson, Abrahamsen, Kapur, Tzipori and Buck2004) and has also been characterized with respect to host specificity (Akiyoshi et al. Reference Akiyoshi, Feng, Buckholt, Widmer and Tzipori2002). TU502 has been maintained by serial propagation in gnotobiotic piglets for over 5 years. C. andersoni isolate JS6 was isolated from a fecal sample collected in a pen which housed bulls and heifers aged approximately 6 months and 2 months, respectively, on a dairy farm in Waianae, Hawaii. C. andersoni VUZV-A was isolated from a 6-month-old Gascogne bull kept at the Institute of Animal Science, Prague, Czech Republic. VUZV-N was obtained from an experimentally infected 1-day-old Holstein calf kept at the same location. Oocysts of C. muris isolate RN66 (Taylor et al. Reference Taylor, Marshall, Green and Catchpole1999) were originally purchased from Waterborne Inc., New Orleans, Louisiana. The isolate has been maintained at TCSVM in Nu/Nu mice for over 6 years.
Cryptosporidium isolates
1 Number of replicate excystation experiments.
2 C57BL/6, immunosuppressed with dexamethasone (Yang et al. Reference Yang, Benson, Du and Healey2000).
3 Taurocholic acid.
4 CD-1-Foxn1nu nude (Charles River, Wilmington, Mass, USA).
5 Mean of 3 counts of same excystation experiment.
Oocyst purification
Different methods were used to purify the oocysts. In the first method, fecal samples were first fractionated on a 15%–25% step gradient of Histodenz (Sigma) as described (Widmer et al. Reference Widmer, Tzipori, Fichtenbaum and Griffiths1998). Fractions containing the oocysts were aspirated from the interface between the two Histodenz solutions and oocysts further purified by immunomagnetic purification (Aureon Crypto Kit, Aureon Biosystems, Vienna, Austria). To avoid interfering with subsequent excystation assays, at completion of the IMS procedure, oocysts were released from the magnetic particles by a short incubation in 1 m NaCl instead of the recommended incubation in 0·1 m HCl. As indicated in Table 1, some oocyst samples were not subjected to IMS purification. Alternatively, feces were first homogenized in water and filtered through a nylon sieve to remove large debris. The resulting suspension was precipitated by low-speed centrifugation and the concentrated sample suspended in Sheather's sucrose solution to float the oocysts. The top layer containing the oocysts was recovered, the oocysts washed once in phosphate buffered saline (PBS) and loaded onto a sucrose gradient (Arrowood and Sterling, Reference Arrowood and Sterling1987). A fraction containing the oocysts was recovered and a final purification step on a gradient of Percoll was performed as described (Kuczynska and Shelton, Reference Kuczynska and Shelton1999).
Species identification
For isolates maintained in the laboratory the species designation was known from prior analyses of the small subunit ribosomal RNA (SSU rRNA) sequence and/or the LIB13 species-specific PCR marker (Tanriverdi et al. Reference Tanriverdi, Arslan, Akiyoshi, Tzipori and Widmer2003). This is the case for MD/TU114, P2 (Tanriverdi et al. Reference Tanriverdi, Blain, Deng, Ferdig and Widmer2007) and TU502 (Xu et al. Reference Xu, Widmer, Wang, Ozaki, Alves, Serrano, Puiu, Manque, Akiyoshi, Mackey, Pearson, Dear, Bankier, Peterson, Abrahamsen, Kapur, Tzipori and Buck2004). Isolate RN66 (GenBank Accession AB089284) was confirmed as C. muris as part of an ongoing genome sequencing project (http://msc.tigr.org/c_muris/index.shtml). For JS6 a 484-basepair portion sequenced on both strands was 99% similar to position 408–890 of the C. andersoni small subunit ribosomal RNA gene (Accession AB089285) and C. muris ‘calf genotype’ (Accession AF093496). The latter designation was used for C. muris-like oocysts isolated from bovine hosts prior to C. andersoni being named a separate species (Lindsay et al. Reference Lindsay, Upton, Owens, Morgan, Mead and Blagburn2000). The JS6 partial SSU rRNA sequence was deposited in GeneBank under Accession number EF432239. Confirming the classification of JS6 as C. andersoni, sequence similarity values between the JS6 partial SSU rRNA sequence and the homologous sequence from C. muris, C. parvum and other species was less than 99%. Isolate VUZV-A and VUZV-N were identified as C. andersoni and C. parvum, respectively, based on oocyst size, oocyst morphology, and the age of the bull (6 months and <7 days, respectively) from which they were isolated. Consistent with published dimensions for C. andersoni and C. parvum oocysts (Upton and Current, Reference Upton and Current1985; Lindsay et al. Reference Lindsay, Upton, Owens, Morgan, Mead and Blagburn2000), the mean oocyst size for VUZV-A was 7·4±0·4×5·6±0·4 μm (n=16), and for VUZV-N 4·8±0·2×4·4±0·1 μm (n=25).
Excystation and numerical analysis
To measure excystation ratios oocysts were suspended in 0·8% taurocholic acid (TA) in PBS, pH 7·4, or 0·1 m HCl and incubated at 37°C for 1 h. Following incubation, a portion of the oocyst suspension was loaded onto a haemocytometer and empty and intact oocysts present on the 1 mm2 gridded area enumerated at 400× magnification. Excystation ratios were calculated as the fraction of empty plus partially excysted oocysts divided by the sum of empty and full oocysts. Empty oocysts present in the unexcysted oocyst sample were subtracted. For each isolate the number of independent replicate experiments is shown in Table 1.
RESULTS
The results of the excystation assays with oocysts from 4 Cryptosporidium species exposed to 2 excystation conditions are shown in Table 1. A total of 14 independent assays were performed. Except for C. hominis, each species was tested a minimum of 3 times in both excystation conditions. Excystation of C. parvum and C. andersoni oocysts were tested independently in 2 laboratories using different isolates. Oocysts from 1 C. hominis and 1 C. muris isolate were examined. In HCl, excystation of C. parvum and C. hominis oocysts was minimal, and the highest excystation ratio obtained was 0·5%. In taurocholic acid, C. parvum and C. hominis oocyst excystation rates ranged from 14% to 34%. C. andersoni and C. muris excysted well in both conditions. Clearly, oocysts of C. muris and C. andersoni excysted better in HCl than C. parvum and C. hominis oocysts, and also excysted slightly better than C. hominis/C. parvum in taurocholic acid (t-test, P=0·02). Oocysts of the gastric species excysted equally well in both conditions (Mann-Whitney Rank Sum test, P=0·50). C. muris oocysts appeared to respond better to acid than to taurocholic acid but the means of 3 experiments were not significantly different (t-test, P=0·09).
To exclude the possibility that oocyst age may differentially affect excystation of oocysts of gastric and intestinal species, the mean oocyst age in both groups, measured from the day of excretion until the day the experiment was carried out (Table 1), was compared. The mean age for all gastric oocyst samples was 29 days (n=6, s.d. 21), whereas for intestinal species the mean age was 22 days (n=8, s.d. 21). The difference in age was statistically not significant (t-test, P=0·55), indicating that age was not a factor in the relative response to HCl and TA.
The unexpected observation that C. muris and C. andersoni oocysts excysted well in HCl and in taurocholic acid suggested that temperature may be sufficient to induce excystation of oocysts from these species. We therefore compared the ability of C. muris and C. parvum oocysts to excyst at 37°C in PBS. The excystation ratio for C. muris after a 1-h incubation at 37°C was 38·3% (s.d. 3·4) as compared to 15·3% (s.d. 2·0) for oocysts maintained on ice. For C. parvum the ratios were very low at both temperatures, 2·7% (s.d. 2·5) at 37°C, and 1·3% (s.d. 1·9) for the control oocysts kept on ice. This observation indicates that temperature is an important excystation trigger in C. muris but not in C. parvum.
DISCUSSION
The ability of Cryptosporidium species to infect different organs of the GI tract is a striking example of a pathogen's adaptation to colonizing different host environments. In this context, C. muris and C. andersoni are particularly interesting as not many pathogens have adapted to the harsh environment of the stomach. The comparison of the C. muris genome, currently being sequenced, with that of C. parvum (Abrahamsen et al. Reference Abrahamsen, Templeton, Enomoto, Abrahante, Zhu, Lancto, Deng, Liu, Widmer, Tzipori, Buck, Xu, Bankier, Dear, Konfortov, Spriggs, Iyer, Anantharaman, Aravind and Kapur2004) and C. hominis (Xu et al. Reference Xu, Widmer, Wang, Ozaki, Alves, Serrano, Puiu, Manque, Akiyoshi, Mackey, Pearson, Dear, Bankier, Peterson, Abrahamsen, Kapur, Tzipori and Buck2004) will provide a detailed insight into how these genomes have been shaped by this adaptative process. The present study indicates that the ability of oocysts to respond to different environmental conditions has also been affected by this evolutionary process. Because of the short life-span of the sporozoites, delivery to the right location in the GI tract is critical for a successful infection. It would therefore be expected that the excystation phenotype is under strong selection.
In addition to the need to respond to specific excystation conditions, oocysts must also be able to discriminate between the GI tract of the original host in which they were formed and the newly infected host. In contrast to oocysts from other coccidia, which sporulate upon release in the environment, Cryptosporidium oocysts are thought to be excreted fully sporulated (Lindsay et al. Reference Lindsay, Upton, Owens, Morgan, Mead and Blagburn2000; Upton and Current, Reference Upton and Current1985). The observation that oocysts of gastric species readily excyst at 37°C is not consistent with the absence of a sporulation-like process, as in the absence of such a process, excystation would be expected to occur soon after oocyst formation and release from the epithelial cell, preventing the excretion of infectious oocysts. The apparent lack of specific excystation triggers other than temperature, raises the possibility of a maturation step analogous to sporulation occurring upon release of the oocysts into the environment. Microscopical observations (Lindsay et al. Reference Lindsay, Upton, Owens, Morgan, Mead and Blagburn2000; Upton and Current, Reference Upton and Current1985) indicate that such a process would not involve morphological changes which are readily apparent. Since elevated temperature by itself does not efficiently induce C. parvum oocysts to excyst, such a step would not be essential for intestinal species. While gastric species have apparently evolved methods to prevent excystation in the same host, intestinal species face the challenge of preventing premature release of the sporozoites after uptake by a new host, as excystation in the stomach would most likely lead to the immediate inactivation of the sporozoites. The observation that, regardless of conditions, C. muris and C. andersoni oocysts excyst in higher proportions than C. parvum and C. hominis oocysts is consistent with the need of oocysts from gastric species to quickly release the sporozoites following ingestion. Consistent with this view is the occasional presence of empty oocysts in preparations of C. muris oocysts. Temperature fluctuations during purification, and in particular during the various centrifugation steps, could be sufficient to initiate excystation in this species. This could also explain the presence of empty C. muris oocyst shells in samples that were not incubated at 37°C. Since excystation of C. parvum and C. hominis requires a sequence of stimuli signalling arrival into the small intestine (Campbell et al. Reference Campbell, Robertson and Smith1992), in vitro excystation conditions producing high excystation ratios are more difficult to reproduce. Among these stimuli, exposure to acid, which mimics the oocysts' travel through the stomach, has been shown to enhance excystation (Campbell et al. Reference Campbell, Robertson and Smith1992; Kato et al. Reference Kato, Jenkins, Ghiorse and Bowman2001; Robertson et al. Reference Robertson, Campbell and Smith1993).
Because Cryptosporidium oocysts cannot be produced in culture in significant numbers, and the parasites cannot be cryopreserved, comparative studies involving multiple species are challenging. The excystation experiments presented here were performed over a period of over 2 years with oocysts obtained at different times from various sources, locations, and host species. Maintaining 4 species in the laboratory to produce oocysts for simultaneous excystation assays was not feasible and, given the inclusion of 2 gastric and 2 intestinal species, was not necessary. It was, however, essential to ascertain that no bias was introduced as a result of oocyst age, host species or oocyst purification method. The fact that excystation phenotypes of gastric and intestinal species were not affected by the propagation host indicates that the oocysts' response to specific stimuli is a stable phenotype. We were also able to rule out oocyst age as an explanation for the excystation results. Although oocyst age varied over a wide range, there was no significant difference in age between gastric and intestinal group. Similarly, centrifugal fractionation on Histodenz, an iodinated gradient medium, or sucrose and Percoll, did not impact excystation in a noticeable manner. Of particular concern was also avoiding exposing the oocysts to conditions known or suspected to promote excystation, such as bleach, acid and, of course, elevated temperature (Smith et al. Reference Smith, Nichols and Grimason2005). For this reason the conventional procedure for releasing oocysts from the magnetic beads was modified by replacing HCl with a short incubation in 1 m NaCl. At no time during purification were oocysts exposed to bleach.
In spite of excystation and host cell invasion being the only phases of the Cryptosporidium life-cycle directly accessible to experimentation, our results show that our understanding of these processes remains superficial. The involvement of parasitic proteases during excystation was postulated on the basis of the inhibitory effect of certain protease inhibitors on C. parvum excystation (Forney et al. Reference Forney, Yang and Healey1996; Okhuysen et al. Reference Okhuysen, Chappell, Kettner and Sterling1996). It is unknown how excystation signals are transmitted to the sporozoite and what other genes participate in excystation and sporozoite activation. The lack of methods for efficient in vitro cultivation and for genetic manipulation of Cryptosporidium parasite is at the root of our limited ability to directly test the role of candidate genes in controlling excystation and the parasite's life-cycle in general. Analyses of the oocyst's transcriptional response to excystation conditions will shed light on the molecular mechanisms leading to excystation.
Financial support from the National Institute of Allergy and Infectious Diseases (grant AI57069) and from the Ministry of Agriculture of the Czech Republic (Project MZe-000 270 1403) is gratefully acknowledged. Our thanks to Sultan Tanrıverdi and Jodi Schumacher for oocyst samples.
