The role of the cytoskeleton in the motility of coccidian sporozoites

Following ingestion of oocysts, sporozoites of Eimeria species are released within the lumen of the gut of their host, from where they invade the endothelial lining of the gut wall and subsequently undergo schizogony. Schizogonie development is highly site-specific for individual species. This site specificity may result from either the directed migration of the released sporozoites (Marquardt, 1973) or a sitelimited potential for development of the intracellular parasite.

Despite the potential importance of motility in the coccidian life cycle the cellular basis for locomotion has been the subject of only indirect and fragmentary studies. Jensen & Edgar (1976) demonstrated that various antiphagocytic agents will inhibit the invasion of professional macrophages by sporozoites of Eimeria magna, and concluded that a microfilamentous system could be active during invasion. They also speculated that the subpellicular microtubules do not function in locomotion, unless they are formed of ‘aggregates of contractile elements sensitive to the inhibitory action of cytochalasin B’. Dubremetz & Ferreira (1978) correlated the motility of Eimeria sporozoites with the capping of the surface-membrane marker, cationized ferritin, and found that both were inhibited by cytochalasin B and low temperatures. Vanderberg (1974) showed that the malarial sporozoite has a limited repertoire of motility, that the sporozoites can attach to substrates, and that they leave a trail behind them, which he speculates to have originated from the rhoptries that may act as a propulsive mechanism similar to a ‘camphor-boat’.

In contrast to the fragmentary studies on locomotion, the ultrastructure of the coccidian and malarial parasites has been extensively examined. The motile stages show a remarkably conservative organization (Roberts & Hammond, 1970; Jensen & Edgar, 1978; Porchet-Hennere & Vivier, 1971 ; Sinden, 1978). Structures of potential significance in locomotion are the extensive microtubular cytoskeleton, and the tri-membranous pellicle with its distinctive inner pellicular membranes bearing linear arrays of intramembranous particles (Dubremetz & Torpier, 1978; Porchet & Torpier, 1977; Dubremetz & Ferriera, 1978; D’haese, Mehlhorn & Peters, 1977).

This paper analyses the motility of sporozoites of Eimeria acervulina and E. tenella and reports the determination of the structural basis of each component of motility. A model for the cellular mechanisms of motility is proposed and its implications in studies on cell invasion are discussed.

Oocysts of E. acervulina were kindly supplied by Dr J. Spelman (May & Baker, Ongar, Essex) and those of E. tenella by Dr R. Williams (Burroughs-Wellcome, Berkhamsted, Berkshire). These were stored in 2% potassium dichromate solution until required, whereupon they were excysted by treatment with hypochlorite, shaking with grade 7 Ballotini beads, and incubating in trypsin and bile in Hanks’ buffered salts solution (HBSS) at 41 °C (Davis, 1973). A cleaned suspension of sporozoites was obtained by sieving the incubated preparation through a 10-μm pore bolt cloth filter.

Preliminary studies revealed a slight decline in sporozoite motility if oocysts were stored in potassium dichromate solution at 4 °C for more than 2 weeks. All studies have therefore been confined to sporozoites obtained from oocysts stored for less than this period. Motility was analysed by videotape recording of closed circuit television (CCTV) images from a Wild phase-contrast microscope with a controlled environment chamber maintained at 41 °C.

The effects of various concentrations of drugs on Sporozoite motility were recorded as described above, and ultrastructural changes were examined by electron microscopy (see below). Microtubule-assembly inhibitors included exposure to low temperature 0—4 °C and the compounds colchiçine, vinblastine sulphate, griseofulvin (at concentrations of 0—200 μg/ml) and nocodazole (at 0–20 μg/ml). Incubation periods varied from 5 min to 6 h, the cells being incubated at either 4 °C or 41 °C. Microfilament inhibitors used included cytochalasin B (Lin, Lin & Flanagan, 1978) and magnesium salt solutions at high concentrations (5–15 mM-MgCl₂) (McGee-Russell & Allen, 1971; Weihing, 1976). Following solution in dimethyl sulphoxide (DMSO), cytochalasin B solutions (1–10μg/ml) were adjusted to contain 0·1% (v/v) DMSO in HBSS. Experiments in this case were controlled by sporozoites suspended in 0·1% (v/v) DMSO. No difference was observed between these controls and sporozoites in HBSS alone.

Sporozoites were incubated with the following fluorescein-labelled reagents: the lectins, concanavalin A (Miles-Yeda) and those of Ricinus communis and peanut (Sigma chemicals): and the anionic marker, cationized ferritin (Sigma). Fluorescein-labelled cationized ferritin was prepared by the method of King & Preston (1977), briefly; 11·5 mg of cationized ferritin in 1 ml of 0·05 M-Tris-HCl buffer (pH 7·6) was mixed for 30 min at room temperature with 1·5 mg of fluorescein isothiocyanate (FITC) on io%(w/v) Celite. Following conjugation, excess fluorochrome was removed by chromatography on Sephadex G25. The labelled eluant was diluted 1 in 5 in HBSS for use. Preparations were examined in a Leitz Orthoplan incident fluorescence microscope. Sporozoite suspensions were incubated in the diluted fluorescein-labelled preparation or with unlabelled cationized ferritin (0·575 mg/ml) at 4 °C and 41 °C in the presence and absence of cytochalasin B (10μg/ml) or 15 mM-magnesium chloride. Sporozoite suspensions were washed 3 times in HBSS (containing cytochalasin B and magnesium ions where appropriate) and prepared for fluorescence or electron microscopy as appropriate.

Lectin distribution was examined in sporozoites both before and after fixation in 0·2% glutaraldehyde, in the latter case the parasites were incubated in 5% (w/v) bovine serum albumin to inactivate residual, aldehyde terminal groups then washed in HBSS before labelling. All sporozoites were incubated at 4 °C and 41 °C in lectin, lectin plus the appropriate saccharide inhibitor (at 100 mM) or lectin plus cytochalasin B (10 μg/ml) for 15–40 min (see Table 1), and then washed 3 times in HBSS before examination by fluorescence microscopy.

Sporozoite suspensions were negatively stained in phosphotungstic acid by the technique of Roberts & Hammond (1970). Thin sections of Araldite-embedded parasites were prepared by the technique of Tucker (1967). A technique that has successfully demonstrated paracrystalline arrays of actin in thin sections of the ciliate Nasnda (Russell, unpublished results) was used in the present study. Sporozoites were incubated in 15 mM-magnesium chloride in HBSS for 40 min at temperatures of 21 and 41 °C and then briefly washed in HBSS before being fixed in the normal manner.

Hanks’ buffered salts solution was first shown to support sporozoite motility by Millard & Long (1974). In the present study our attempts to enhance sporozoite motility by contact with cultured monolayers of Xenopus epithelial cells, mouse peritoneal macrophages, embryonic chick brain cells and Chang liver cells all failed, and attempts to prolong motility by the addition of 10 mM-glucose, fructose or mannose (Ryley, 1973) did not significantly alter their behaviour. Consequently all studies on sporozoite motility were conducted in basic HBSS.

Initial observations revealed that the repertoire of Eimeria sporozoite for movement was similar to, although much more vigorous than, that described in Plasmodium by Vanderberg (1974). Hence his terminology was adopted, i.e. movement was the change of position of one part of the cell relative to another, locomotion was the translocation of the entire cell, and motility described both movement and locomotion, Eimeria sporozoites displayed 2 types of’ movement : namely, bending and pivoting ; and only one form of locomotion, namely, gliding. The pattern of each activity during the life of a ‘typical’ sporozoite is illustrated in Fig. 1.

Bending activity was distributed continuously throughout the motile life of the sporozoites (Fig. 2), and characteristically involved the relatively slow flexion of the anterior half of the sporozoite toward the ‘inside curve’ of the body followed by a somewhat abrupt almost ‘springlike’ straightening of the body; this suggested that flexing and not straightening is the active energy-consuming process. Despite the constant proportion of time spent in bending activity, a slow decline in the intensity (i.e. number of flexions per min) was noted with increasing age of sporozoite (Fig. 2).

Pivoting was only seen when sporozoites were attached to the substratum by their posterior pole. Sporozoites fixed to the slide always describe an anticlockwise rotation of the anterior pole (N.B. reversed to clockwise by microscope optics). The same unidirectional rotation was also noted during locomotion (see below). Pivoting, unlike bending, was a very spasmodic activity, and was commonly encountered in the first 8 min of the sporozoites life (at 41 °C) but more rarely with increasing age of sporozoite (Fig. 1). The intensity of activity (i.e. number of turns/min) also declined significantly with time (Fig. 3), and subjective observation suggested that the rate of turning, although highly variable throughout the life of the sporozoite, also declined slightly with increasing age.

Locomotion, like pivoting, was a spasmodic activity (Fig. 1), and these two activities were frequently contiguous. The frequency of locomotion decreased with increasing sporozoite age. The distance moved declined noticeably during the sporozoite lifetime. This reflected a decrease in the speed of sporozoite movement and a shortening duration of active periods with time (Fig. 4). The initial speed of locomotion was 4–8 μm/s. Locomotion was achieved by a gliding motion of the rotating, helically coiled sporozoite. Single-frame analysis of recorded movement revealed a distinct ‘unit sequence’ of cell postures (Fig. 5). From reconstructions of the movement it was apparent that the anterior tip of the body did not connect with the substratum and that a helical footprint was described along the posterior J of the body surface while the sporozoite was in contact with the substratum (Fig. 6). The sporozoite ceased movement only at the end of a unit sequence of movement, but could perform several units consecutively. The total distance moved per unit sequence was 7–10 μm (N.B. body length is 12–13 μm).

None of the 4 microtubule inhibitors used in this study (colchicine, vinblastine sulphate, griseofulvin or nocodazole) at the range of concentrations stated above had any detectable effect on sporozoite motility. Neither did treatment at 0–4 °C for 15–360 min prior to examinatión at 41 °C.

Negatively stained and thin-sectioned sporozoites, treated by any of the above methods, were examined in the electron microscope to determine whether any morphological changes in the microtubular cytoskeleton could be detected. These studies showed that the microtubule length and number (Fig. 7), the subpellicular localization and the microtubular diameter (Fig. 8), and. the protofilament and molecular architecture (Fig. 9) were, all identical to those Seen in normal untreated sporozoites. Thus, whilst it was not directly proven that the drugs used successfully penetrated the sporozoites, the use of prolonged incubating times and physical treatments (low-temperature shock), and the success and rapidity of action of other inhibitors (see below), all suggested that the microtubular skeleton was exposed to these inhibitors but was exceptionally stable, i.e. that the microtubules were not in a dynamic polymerizing-depolymerizing state.

The effects of cytochalasin B on the 3 components of motility are illustrated in Figs. 2–4. From these results it was clear that pivoting and locomotory components of motility were significantly depressed by concentrations above 2·5 μg/ml, whereas bending was relatively insensitive to the compound. The effects observed followed the addition of the drug to the suspending medium almost instantaneously and, conversely, motility recommenced within seconds of the removal of the compound.

Treatment of sporozoites with 15 mM-magnesium ions caused the cessation of all motility. This and the studies with cytochalasin B suggest most strongly that all aspects of motility were microfilament-based, but that pivoting and locomotion were determined by one and the same set of microfilaments, whereas bending was controlled by a different complex, the latter system being less accessible or less sensitive to cytochalasin B.

Despite the indications that a microfilamentous system is involved in motility, electron microscopic studies of untreated sporozoites failed to reveal any significant concentrations of microfilaments (Fig. 8). Further, treatment with magnesium ions failed to reveal any localized deposits of actin-sized filaments as demonstrated in the Gymnostome ciliate Nassula (Russell, unpublished results).

The apparent adhesion of sporozoites to the substratum during locomotion, and the dramatic inhibition produced by treatment with cytochalasin B both suggested that motility could be mediated by an actin-based capping-like activity. The presence of specific adhesion molecules on the sporozoite, e.g. fibronectin and other identifiable membrane coat components, were therefore studied using FITC-labelled R. communis lectin; other lectins binding to glycoproteins normally associated with cell adhesion, and the anionic site marker, cationized ferritin. None of the lectins used could be found to bind in detectable quantities on the sporozoite membrane in either fixed or unfixed preparations despite treatment with trypsin, which often reveals otherwise unmarked sites (Sethi, Rahmun Pelster & Brandis, 1977). Oocysts remaining in the sporozoite preparation, however, reacted vigorously with all the lectins used (Table 1).

In contrast, sporozoites were readily labelled with cationized ferritin, which under normal conditions of incubation of sporozoites (in HBSS at 41 °C) was found to cap rapidly to the posterior of the cell. The speed was such that most sporozoites, incubated with the label at 4 °C, had capped the marker before the temperature of the medium had reached 41 °C and numerous cells were found with label concentrated about the ‘permanent’ anionic site at the posterior of the cell (Fig. 10). Capping of cationized ferritin was inhibited not only by low temperature (Fig. 11) but also by treatment with cytochalasin B (Fig. 12). Removal of the cytochalasin B was immediately followed by renewed capping activity. It was noted that these responses were identical to those observed on motility.

Fluorescence studies on the distribution of cationized ferritin were confirmed by electron microscopic studies (Fig. 10c), which revealed in addition, at 4 °C, evidence for patching of the surface marker (Fig. 11c).

Despite the high motility of merozoites and sporozoites of the coccidia, so elegantly recorded in the cinematographic studies on cell invasion by early workers (Doran, John & Rinaldi, 1962), the cellular basis or indeed the ‘role’ of locomotion has not been studied in depth. Nonetheless numerous studies have made incidental observations upon the process and considerable discussion has followed (Ryley, 1980). It was therefore the prime purpose of this study to interrelate new and basic behavioural observations with morphological and molecular studies to determine the cellular basis of motility of Eimeria sporozoites.

This study has recognized 3 classes of motility in Eimeria sporozoites, two of which, pivoting and gliding locomotion, are behaviourally similar and may be achieved by the same cellular mechanism. Indeed, it is possible that pivoting merely represents an ‘aberrant’ form of locomotion. The remaining and behaviourally distinct process, bending, would seem to us to be of little physiological importance. Inhibitor studies have shown that bending, like locomotion, is sensitive to inhibitors of microfilament activity but suggested that different contractile systems may be involved in the two activities, one interacting directly with subpellicular microtubules resulting in the bending activities, the other associated with the plasmalemma, providing the drive for locomotion (see below).

We have confirmed the observations of Jensen & Edgar (1976) and Dubremetz & Ferreira (1979) that treatment with cold and with cytochalasin B inhibits sporozoite locomotion. These observations, together with the inhibitory effects of magnesium ions upon motility reported in the present study, indicate that the locomotion of Eimeria sporozoites is based upon a microfilamentous system (Lin et al. 1978; McGee-Russell & Allen, 1971).

Since the original description of sporozoite motility as being ‘serpentine’ (Grassi, 1900), it has often been assumed that sporozoites move by the propagation of waves down the body and that these waves are the result of the interaction between the subpellicular microtubules and contractile elements within the sporozoite. This assumption was first examined by Jensen & Edgar (1976), who noted that colchicine, a microtubule-depolymerizing agent, failed to inhibit the motility of Eimeria sporozoites. The authors suggested that the tubular structures were either not microtubules or not implicated in sporozoite motility. This present study has shown that 4 micro tubule inhibitors of different modes of action not only fail to inhibit sporozoite motility but also fail to have any detectable effect on the microtubule structure. The failure of these drugs and cold treatment cannot therefore be considered as evidence that sporozoite motility is not generated by the subpellicular microtubules. However, despite the shortcomings of the microtubule inhibitor studies, the behavioural observations suggest that microtubules are not responsible for locomotion because any microtubule-associated sliding mechanism, as suggested by Wong & Desser (1976) and Sinden (1978), would result in changes in body shape, i.e. waves; these Jo not occur during sporozoite locomotion. We suggest that the function of the complex skeleton of microtubules is the maintenance or adoption of a rigid body shape in the form of a part helix. This rigidity is important in the following theory of motility and cell invasion.

The results of this and other studies on the structure and motility of coccidian sporozoites and merozoites are consistent with the model of locomotion that, briefly stated, is as follows; the sporozoite is capable of capping a wide variety of surface ligands to its posterior pole by an organized submembranous microfilament system. The specificity and availability of the cell-surface ligands will control whether the capping reaction results in locomotion (when a substratum binding ligand is involved) or the capping and shedding of other molecules bound to the parasite surface (e.g. antibody capping in the circumsporozoite precipitation reaction).

This model proposes that invasive parasites move only when in contact with a substratum to which they are attached by components of the cell surface coat. These surface ligands may be either highly specific, e.g. for the Duffy-associated sites for malarial merozoite attachment on the erythrocyte membrane (Miller et al. 1975); ^or non-specific, e.g. those resulting in the attachment of sporozoites of Eimeria and Plasmodium to glass or plastic surfaces. The glycoprotein usually associated with cell-substratum interactions is fibronectin. This protein has been shown to be associated with the contractile proteins, actin and myosin, in mammalian fibroblasts (Badley, Woods, Smith & Rees, 1980). However, treatment with the R. communis lectin demonstrated that this glycoprotein was not exposed on the surface coat of Eimeria sporozoites. Indeed the Eimeria sporozoite cell surface, like that of Toxoplasma (Sethi et al. 1977) and Plasmodium (Schulman, Oppenheim & Vanderberg, 1980; Turner, 1980), is remarkably free of any lectin binding sites.

The surface-binding ligands are attached, either directly or indirectly through intramembranous particles (IMPs) in the plasmalemma, to a microfilamentous system. Such surface-modulating assemblies (Edelman, Wong & Yahara, 1976; Bourguignon & Singer, 1977; Rajaraman, MacSween & Fox, 1978) are responsible for the capping of the surface-binding ligands to the posterior anionic site of the cell surface and hence the progression of the cell body over the substratum. Aikawa, Cochrane, Nussenzweig & Rabbage (1979) have shown that the IMPs of the plasmalemma of the malarial sporozoite are normally distributed at random over the cell membrane, but upon the addition of antisporozoite antibody they are concentrated at the posterior pole of the parasite. This redistribution is accompanied in living specimens by the capping and shedding of the antibody in the typical circum sporozoite precipitation reaction (Vanderberg et al. 1972). In contrast Dubremetz & Ferreira (1978) have suggested that the capping of fluorescein-labelled cationized ferritin is not associated with any change in distribution of IMPs in the plasmalemma, suggesting that the antibody and cationized ferritin binding sites are attached to different components of the plasmalemma, which may be capped independently. It is therefore considered unlikely that the IMPs in the plasmalemma will be capped during locomotion.

During locomotion the sporozoite follows a unidirectional helical path, therefore the direction of capping of the surface ligands will be similarly unidirectional and helical (Figs. 5, 6). It is suggested that the helically coiled linear arrays of IMPs on the outer surface of the inner pellicular membranes (Dubremetz & Torpier, 1978; D’haese et al. 1977; Dubremetz et al. 1979) may provide the cellular basis for such direction by acting as anchorage points for the microfilament system. The organization of these linear arrays may in turn be directed by the subtending rigid sub-pellicular microtubules. It is important to note that the spiral, described by the intramembranous particles and the subpellicular microtubules, is identical to the proposed path of the surface receptors during capping (Figs. 5, 6).

The observations on coccidian motility that have led to the proposed model of locomotion and the projected properties of the model are remarkably similar to the recorded data on cell invasion by coccidian and malarial parasites (e.g. Jensen & Edgar, 1976; Miller, Aikawa, Johnson & Shiroishi, 1979; Bannister, 1977). We therefore propose that host-cell invasion by these parasites involves 3 distinct and essential cellular events. (1) The attachment of the parasite to the host cell; (2) the induction by the parasite of a parasitrophorous vacuole membrane; (3) the locomotion of the parasite into the induced vacuole. We suggest that the third component, locomotion, is the same mechanism as is used for extracellular motility even in those parasites that display little if any extracellular activity (e.g. Plasmodium merozoites).

We believe that the speed and ease with which coccidian merozoite and sporozoite locomotion may be essayed in vitro offers an accessible and relevant model for the study of this vital component of cell invasion, but is nonetheless an interesting example of a highly active cell exploiting a contractile system normally associated with slow-moving mammalian cells in culture.