INTRODUCTION
Protozoa infection of the phylum Apicomplexa including Plasmodium spp., Toxoplasma gondii, Eimeria spp. and Cryptosporidium spp. occurs worldwide in humans and animals (Black and Boothroyd, Reference Black and Boothroyd2000; Xiao et al. Reference Xiao, Fayer, Ryan and Upton2004; Daugschies and Najdrowski, Reference Daugschies and Najdrowski2005; Ashley et al. Reference Ashley, McGready, Proux and Nosten2006; Morris and Gasser, Reference Morris and Gasser2006). These parasites, which are obligate intracellular parasites, have complex life-cycles involving transmission, and invade, grow and replicate within host cells. Invasive zoites of apicomplexa parasites possess unique organelles called the apical complex which play an important role in its motility and/or invasion into host cells. This apical complex is generally composed of rhoptries, micronemes and dense granules, which are secretory organelles (Sam-Yellowe, Reference Sam-Yellowe1996), and apical or polar rings, conoid and subpellicular microtubules, which provide a cytoskeletal framework during host cell invasion (Lindsay et al. Reference Lindsay, Blagburn and Toivio-Kinnucan1991, Reference Lindsay, Speer, Toivio-Kinnucan, Dubey and Blagburn1993; Hu et al. Reference Hu, Roos and Murray2002). It has been thought that microtubules composed of polymers of α and β tubulins form these cytoskeletal frames of zoites, such as conoid as well as subpellicular microtubules.
C. parvum is an intracellular protozoa belonging to the phylum Apicomplexa, and is a common cause of waterborne gastroenteritis throughout the world in animals and humans (O'Donoghue, Reference O'Donoghue1995; Chen et al. Reference Chen, Keithly, Paya and LaRusso2002). In immunocompetent patients, C. parvum develop a self-limited watery diarrhoea lasting 1–2 weeks; however, the disease can be life-threatening in immunodeficient patients, such as those with acquired immunodeficiency syndrome (AIDS) (O'Donoghue, Reference O'Donoghue1995; Chen et al. Reference Chen, Keithly, Paya and LaRusso2002). Despite the magnitude and severity of C. parvum infections, there are no effective pharmaceutical therapies or immunotherapy, including approved vaccines (O'Donoghue, Reference O'Donoghue1995; Perryman et al. Reference Perryman, Kapil, Jones and Hunt1999; Sagodira et al. Reference Sagodira, Buzoni-Gatel, Iochmann, Naciri and Bout1999; Chen et al. Reference Chen, Keithly, Paya and LaRusso2002).
Like other apicomplexan parasites, the invasive form (sporozoite or merozoite) of C. parvum is thought to possess characteristic apical organelles. Recent morphological studies using electron microscopy have revealed that C. parvum sporozoites have a single rhoptry, multiple micronemes and dense granules around the apical region (Tetley et al. Reference Tetley, Brown, McDonald and Coombs1998); however, compared with these secretory organelles, the cytoskeletal structure of C. parvum zoites has not been fully analysed. So far, genes encoding tubulins of C. parvum have been characterized genetically (Bonafonte et al. Reference Bonafonte, Garmon and Mead1999; Hu et al. Reference Hu, Roos and Murray2002; 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); furthermore, anti-microtubule drugs which can disrupt the microtubule network within sporozoites inhibit C. parvum infection of host cells (Wiest et al. Reference Wiest, Dong, Johnson, Tzipori, Boeklheide and Flanigan1994). These findings, in previous studies, proved that cytoskeletal organelles, especially microtubules, were present on C. parvum zoites and related with the host cell invasion of C. parvum; however, these morphological frame structures of C. parvum zoites remain unclear. Thus, in the present study, we morphologically examined C. parvum sporozoites using electron or immunoelectron microscopy to clarify the ultrastructural frames of invasive stages.
MATERIALS AND METHODS
Parasites
C. parvum oocysts, strain HNJ-1, were originally obtained from the faeces of a patient in Japan (Abe et al. Reference Abe, Kimata and Iseki2002). These oocysts were passaged in severe combined immunodeficient (SCID) mice, purified by the sugar flotation method, and stored at 4°C in phosphate-buffered saline (PBS, pH 7·2) until 1 month before use, or were freeze-dried for Western blot analysis. C. parvum sporozoites were prepared as previously reported except for using RPMI medium 1640 (Invitrogen) (Matsubayashi et al. Reference Matsubayashi, Kimata, Iseki, Lillehoj, Matsuda, Nakanishi, Tani, Sasai and Baba2005).
Transmission electron microscopy (TEM)
The C. parvum sporozoites prepared as described above were fixed with 2% glutaraldehyde (GA, Electron Microscopy Sciences) and 1% tannic acid (Wako Pure Chemical Industries Ltd) in 0·1 m phosphate buffer (PB, pH 7·4) at room temperature for 2 h, additionally fixed with 2% GA at 4°C for 1 week, and washed in 0·1 m PB. They were post-fixed with 2% OsO4 in 0·1 m PB at 4°C for 2–3 h, dehydrated in an ethanol series and embedded in Epon resin. Ultra-thin sections were stained with 2% uranyl acetate (Cerac Inc.) in distilled water for 15 min, lead staining solution (Hanaichi et al. Reference Hanaichi, Sato, Iwamoto, Malavasi-Yamashiro, Hoshino and Mizuno1986) for 5 min, and examined with an electron microscope (JEM-1200, JEOL).
Negative staining
Extraction by Triton X-100 was performed to reveal the detailed organization of the cytoskeleton (Nichols and Chiappino, Reference Nichols and Chiappino1987). Briefly, 106C. parvum sporozoites, excysted as described above, were suspended in 0%, 0·1% or 0·5% Triton X-100 (Sigma-Aldrich Co.) in RPMI medium 1640 (Invitrogen) for 5–10 min, fixed in 0·1% GA (TAAB Laboratories Equipment Ltd) and 2% paraformaldehyde (PFA, Merck Ltd) in 0·1 m PB at 4°C for 4–6 days, and washed in ultra-pure water (UPW). Five μl of parasite suspension, resuspended in 400 μl of UPW, were dropped onto carbon-stabilized 200-mesh nickel grids (Nissin EM Co. Ltd) covered with thin collodion film (Nissin EM Co. Ltd), and allowed to settle for 20 min in a humid chamber. Excess fluid was removed and the grids were air-dried. The parasites on the grids were stained with 2% phosphotungstic acid (PTA, pH 7·5, TAAB Laboratories Equipment Ltd) or 1% uranyl acetate (TAAB Laboratories Equipment Ltd) for 30 sec and examined with an electron microscope (H-7500, Hitachi).
Indirect immunofluorescence assay (IFA)
For IFA with C. parvum sporozoites, slides were prepared as previously reported (Matsubayashi et al. Reference Matsubayashi, Kimata, Iseki, Lillehoj, Matsuda, Nakanishi, Tani, Sasai and Baba2005), fixed using methanol, acetone or 0·05% GA and 2% PFA for 10 min or non-fixed, and stored at −80°C until use. Parasites fixed with 0·05% GA and 2% PFA were blocked with 0·1 m glycine (Nacalai Tesque) in PBS before immunostaining. After washing and drying, prepared slides were treated with 0%, 0·1% or 1% Triton X-100 in PBS for 10 min and washed 3 times for 5 min in PBS. After blocking with 1% bovine serum albumin (BSA, Sigma-Aldrich Co.) in PBS for 10 min, the slides were incubated with 50 μl of primary antibody, mouse anti-α tubulin (monoclonal, against the C-terminal region of α tubulin isoform derived from sea urchin sperm axonemes; T5168, Sigma-Aldrich Co.) at 1:1000 dilution or mouse anti-β tubulin (monoclonal, against purified rat brain tubulin; T4026, Sigma-Aldrich Co.) at 1:500 dilution in 1% BSA in PBS for 40 min. As a control, purified normal mouse IgG (Sigma-Aldrich Co.) at 1:1000 dilution was used. The slides were washed 3 times with PBS containing 0·05% Tween 20 (PBST) and incubated with 50 μl of fluorescein isothiocyanate (FITC)-conjugated rabbit anti-mouse IgG at 1:100 dilution (Sigma-Aldrich Co.) for 30 min. After immunostaining, the slides were mounted with 50% glycerol (Nacalai Tesque) in PBS for examinations using an epi-fluorescent microscope (Nikon). All steps were performed at room temperature.
Western blot analysis
For analyses by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting, C. parvum oocyst antigen was prepared as previously reported (Matsubayashi et al. Reference Matsubayashi, Kimata, Iseki, Lillehoj, Matsuda, Nakanishi, Tani, Sasai and Baba2005). Briefly, 20 μg of C. parvum oocysts antigens were resolved on 4% stacking/12% resolving SDS-PAGE at 200 V constant voltage (Laemmli, Reference Laemmli1970) and the separated proteins were blotted to PVDF membrane (Immobilon Transfer Membranes, Millipore). After blocking with 1% BSA in PBS, membranes of blotted individual lanes were stained for 1 h at room temperature with anti-α tubulin antibody at 1:1000 dilution, anti-β tubulin antibody at 1:100 dilution or normal mouse IgG at 1:1000 dilution as a control in 1% BSA in PBS, and washed 3 times for 5 min with PBST. Bound antibodies were detected by staining with horseradish peroxidase-labelled goat anti-mouse IgG Fab specific (Sigma-Aldrich Co.) at 1:10 000 dilution for 1 h and developed with Konica Immunostaining HRP-1000 (Konica). Molecular weights were estimated using standard molecular weight proteins after staining with 0·2% Coomassie Brilliant Blue R-250 (Wako Pure Chemical Industries Ltd).
Two-dimensional polyacrylamide gel electrophoresis (2-DE)
Two-dimensional polyacrylamide gel electrophoresis was performed using Protean IEF Cell (Bio-Rad) according to the manufacturer's protocol. Briefly, 109 freeze-dried C. parvum oocysts were dissolved in 1 ml of the Rehydration/Sample buffer (Bio-Rad), which contained 8 m urea, 2% CHAPS, 50 mm dithiothreitol (DTT), 0·2% Bio-lyte 3/10, and 0·001% bromophenol blue, sonicated in an ice bath with an Ultrasonic Processor (Taitec Co.) and centrifuged at 600 g for 5 min. The supernatant was aliquoted and cryopreserved at −80°C until use. The concentration of crude antigen was determined by Coomassie Blue Assay Kit (Pierce). For first-dimensional electrophoresis, ReadyStrip™ pH 3–10 IPG strips (7 cm; Bio-Rad) were rehydrated by loading Rehydration/Sample buffer containing 20 μg of prepared antigen for 12 h. After rehydration, focussing conditions were 20 min at 250 V to remove excess salts, 2 h at 4000 V for voltage ramping, followed by focussing for 2·5 h at 4000 V. The temperature was set at 20°C. After isoelectric focussing, strips were laid onto 4% stacking/12% gradient SDS-PAGE gels and resolved in the second dimension at 200 V constant voltage. Second-dimension gels were blotted, and the blotted membranes were immunostained as described above.
Immunoelectron microscopy (IEM)
C. parvum sporozoites treated by 0·1% Triton X-100 (Sigma-Aldrich Co.) in RPMI medium 1640 (Invitrogen) in 5–10 min were fixed in 2% PFA in 0·1 m PB at 4°C for 4–6 days, washed and resuspended in UPW. Parasites settled on grids were air-dried as described above, stained with anti-α tubulin antibody or normal mouse IgG, the same as IFA. After washing with PBST 3 times, grids were incubated with the secondary antibody, rabbit anti-mouse IgG-conjugated with 5 nm colloidal gold (EY Laboratories Inc.) or goat anti-mouse IgG conjugated with ultra small gold (Electron Microscopy Sciences) for 30 min, and then washed 3 times with PBST and UPW. Silver enhancement was carried out using GoldEnhance™ LM/Blot Formulation (Nanoprobes Inc.) for 1 min according to the manufacturer's protocol. They were washed with UPW, stained by 2% PTA and examined with an electron microscope (H-7500, Hitachi).
RESULTS
Two outer membranes, including the plasma membrane and inner membrane complex, were present throughout the entire region of C. parvum sporozoites (Fig. 1A–E). At the apical end of sporozoites, 3 apical rings (Petry and Harris, Reference Petry and Harris1999), an electron-dense collar (Fayer et al. Reference Fayer, Speer, Dubey and Fayer1997), single rhoptry, many micronemes and dense granules were observed (Fig. 1A–D). Furthermore, 2 central microtubules were observed further inside the sporozoites, at approximately 25 nm thickness, and extended from the apical rings of sporozoites (Fig. 1A, F). Between the plasma membrane and inner membrane complex, there were structures previously reported as longitudinal ridges (Uni et al. Reference Uni, Iseki, Maekawa, Moriya and Takada1987) from the apical region of sporozoites to the posterior at intervals of 30–60 nm, and were approximately 15–20 nm thickness and often observed as spiral around sporozoites in traverse sections (Fig. 1D, E, G, H).
Fig. 1. Transmission electron micrographs of Cryptosporidium parvum sporozoites. (A) Traverse section of the anterior region of a sporozoite. Only a single rhoptry was seen in C. parvum sporozoites. (B) Higher magnification of the sporozoite anterior region. Small arrows show apical rings and the large arrow shows the electron-dense collar. (C) Cross-section of the apical end of the sporozoite. Arrows show the electron-dense collar beneath the inner membrane complex. (D) Cross-section at middle region of the sporozoite. Longitudinal ridges are present between the plasma membrane and inner membrane complex and run as a spiral around the sporozoite (DG). (E and H) High magnifications of the boxed regions in (D) and (G), respectively. Arrows show longitudinal ridges. (F) Higher magnification of the anterior region of the sporozoite. Arrows show 2 thick central microtubules. ar, Apical rings; edc, electron-dense collar; pm, plasma membrane; imc, inner membrane complex; cm, central microtubules; lr, longitudinal ridges; r, rhoptry; m, micronemes; dg, dense granules.
In a preliminary examination of extraction by Triton X-100, only a few sporozoites greatly changed in form were observed, and many oocyst walls with 0·5% concentration of Triton X-100 (data not shown). Under this condition, sporozoites of C. parvum were thought to collapse with detergent treatment of Triton X-100; however, the organization of the cytoskeleton was clearly displayed on negatively stained C. parvum sporozoites by extraction by 0·1% Triton X-100 in 5 min as a milder condition (Fig. 2A, B). Additionally, without Triton X-100 treatment, no internal organelles of sporozoites were seen because the surface membrane of sporozoites impeded electron microscopic observations (data not shown).
Fig. 2. Electron micrographs of Cryptosporidium parvum sporozoites treated with 0·1% Triton X-100 for 5 min. Detergent-treated sporozoites are negatively stained by 2% phosphotungstic acid (A) or 1% uranyl acetate (B). Black arrows and an asterisk show apical rings and micronemes in (A), respectively. White arrows show central microtubules in (A) and (B).
In Fig. 2A, three apical rings and the electron-dense collar below the apical rings were present in the apical region of sporozoites, and micronemes were below the apical region of sporozoites. Furthermore, 2 thick central microtubules were observed inside sporozoites and extended from apical rings to the posterior region (Fig. 2A, B).
We examined the reactivity of anti-α and -β tubulin antibodies with C. parvum sporozoites extracted by 0%, 0·1% or 1% Triton X-100 in 10 min. As a result, anti-α and -β tubulin antibodies did not react to any sporozoites treated with 1% Triton X-100 (data not shown). After treatment with 0·1% Triton X-100, anti-α tubulin antibody reacted with the apical tip of sporozoites, and the organelles which extended from the apical to posterior region in sporozoites (Fig. 3A, B). Anti-β tubulin antibody reacted with only the apical region of sporozoites (Fig. 3C, D) although the reaction of anti-β tubulin antibody was very weak. Normal mouse IgG did not show any reactions with sporozoites (Fig. 3E, F). Without Triton X-100 treatment, there was less reaction of these antibodies with C. parvum sporozoites than with 0·1% Triton X-100 (data not shown). When fixed with methanol or acetone, or non-fixed, these reactions were not changed; however, strong non-specific reactions were observed on sporozoites when fixed with 0·05% GA and 2% PFA (data not shown).
Fig. 3. Immunofluorescence staining of Cryptosporidium parvum sporozoites using anti-α tubulin (A and B), anti-β tubulin (C and D) and normal mouse IgG (E and F). Photomicrographs from immunofluorescence microscopy (A, C and E) and bright-field by interference contrast microscopy (B, D and F) are shown.
Western blotting analysis showed that anti-α tubulin antibody reacted with the 50 kDa C. parvum antigens, but anti-β tubulin antibody and normal mouse IgG did not react with any bands (Fig. 4A). Further 2-DE analysis revealed that the α tubulin antigen of C. parvum was detected in a pH range of 6 to 7 (6·24) (Fig. 4B).
Fig. 4. (A) Western blot analysis probed with the following primary antibodies using 20 μg of Cryptospoidium parvum antigens. Lane 2, anti-α tubulin; lane 3, anti-β tubulin; lane 4, normal mouse IgG. Lane 1 shows molecular weight standards. (B) Immunoblot analysis of 20 μg of C. parvum antigens with anti-α tubulin antibody by 2-dimensional gel electrophoresis. The left lane shows molecular weight standards. pH values are indicated at the top.
We tried to stain sporozoites extracted by Triton X-100 using anti-α tubulin antibody in IEM be-cause anti-β tubulin antibody weakly reacted with C. parvum zoites in IFA. Consequently, a secondary antibody labelled with 5 nm gold particles failed to penetrate the surface of detergent-extracted sporozoites (data not shown), while ultra-small gold particles were successfully visible on the internal organelles after enhancement with silver. The silver-enhanced gold particles were observed on 3 apical rings, the electron-dense collar of the apical region and 2 central microtubules, but not other organelles (Fig. 5A, B), although there were less gold particles on the apical rings and the electron-dense collar than on the 2 central microtubules.
Fig. 5. Immunogold labelling of detergent-treated Cryptosporidium parvum sporozoites using anti-α tubulin antibody. Ultra-small gold particles were made visible by silver enhancement. Arrows show structures labelled with gold particles (A and B). (B) Higher magnification of apical region of sporozoites.
DISCUSSION
To date, there have been many reports about ultrastructural analyses of apicomplexan parasites such as Plasmodium spp., Toxoplasma gondii, or Eimeria spp. (Aikawa, Reference Aikawa1971; Dubremetz and Elsner, Reference Dubremetz and Elsner1979; Lindsay et al. Reference Lindsay, Blagburn and Toivio-Kinnucan1991; Tetley et al. Reference Tetley, Brown, McDonald and Coombs1998; Hu et al. Reference Hu, Roos and Murray2002). Zoites in invasive stages of these parasites have a collection of unique organelles, termed the apical complex, which include rhoptries, micronemes, dense granules, apical or polar rings, conoid and/or subpellicular microtubules; however, as not all apicomplexan zoites possess all these organelles in the apical region, the detailed ultrastructure of zoites of each species is different (Morrissette and Sibley, Reference Morrissette and Sibley2002). For example, conoid was not found in Plasmodium spp. and Theileria spp., and subpellicular microtubules were not found in Theileria spp. (Morrissette and Sibley, Reference Morrissette and Sibley2002). Compared with well-analysed apicomplexan parasites, the ultrastructure of Cryptosporidium zoites remains unclear. Thus, the invasive process of C. parvum, including the role of the apical complex, is also poorly understood. In the present study, our results showed that C. parvum sporozoites possessed secretory organelles, a single rhoptry (2 rhoptries in Toxoplasma gondii tachyzoites), multiple micronemes and dense granules around its apical region, and supported the findings reported previously (Tetley et al. Reference Tetley, Brown, McDonald and Coombs1998).
For cytoskeletal organelles of C. parvum sporozoites, we could observe 3 apical rings and an electron-dense collar that were present only at the apical end, and 2 central microtubules in the centre of sporozoites but not in other organelles. Therefore, we clarified that it might not possess structures such as subpellicular microtubules surrounding zoites, as in Toxoplasma gondii. In addition, many longitudinal ridges were seen in the entire region of C. parvum sporozoites between the two outer membranes. These ridges may provide support in the absence of subpellicular microtubules to organize the elongated body shape of sporozoites.
Anti-α and -β tubulin antibodies, which were generated from tubulin origins derived from sea urchin and rat brain, respectively, showed cross-reactivity with C. parvum sporozoites in IFA; however, by Western blotting analysis, only anti-α tubulin antibody recognized the approximately 50 kDa antigen of C. parvum. Although anti-β tubulin antibody recognized the 55 kDa antigen derived from vertebrates (Gozes and Barnstable, Reference Gozes and Barnstable1982), β tubulin of C. parvum might have structurally different carbohydrate modifications on the epitope from those of other origins, while the molecular weight of α tubulin antigen of C. parvum coincided with that predicted from the isolated C. parvum α tubulin gene (Bonafonte et al. Reference Bonafonte, Garmon and Mead1999) and that of α tubulin from other species (Piperno et al. Reference Piperno, LeDizet and Chang1987). Furthermore, IEM indicated that cytoskeletal organelles of C. parvum sporozoites, including apical rings, the electron-dense collar in the apical region and 2 central microtubules were formed by α tubulin. Our results suggested that α and β tubulins were phenotypically present on C. parvum sporozoites, and these cytoskeletons of C. parvum sporozoites could possess a conserved function such as motility by polymerization of tubulins, similar to tubulins of other species during the process of host cell invasion.
In conclusion, C. parvum sporozoites have a unique apical structure which is different from other apicomplexan parasites. Although the invasion process of C. parvum has not been fully elucidated, these characteristics may be related to the extra-cytoplasmic development of C. parvum, which only enter epithelial cells. Although more detailed analysis is needed to clarify these functions during attachment or invasion into host cells, our results are helpful to understand the invasive process of C. parvum sporozoites.
This work was partly supported by the Public Trust for Home Economics Research Support Fund (to M. M.).
