HeLa cells extend and internalize pseudopodia during active invasion by Trypanosoma cruzi trypomastigotes

Trypanosoma cruzi, the protozoon that causes Chagas’ disease, is an obligatory intracellular parasite. The forms found in the mammalian host, such as blood trypomastigotes and amastigotes, which divide intracellularly, as well as the metacyclic-trypomastigote form, found in the faeces of triatomine vectors, are able to invade and to establish infection in mammalian cells (Brener, 1973; de Souza, 1984; Ley et al. 1988; Mortara,1991). Trypomastigotes and amastigotes appear to use different invasion mechanisms, possibly employing different cellular receptors. For example, amastigotes invade HeLa cells after association with surface microvilli and mobilization of actin microfilaments. Their entry is inhibited by cytochalasin D (Mortara, 1991), a microfilament-disrupting drug (Cooper, 1987). In contrast, trypomastigotes preferentially invade cells at the edges (Schenkman et al. 1988; Mortara, 1991), and can enter HeLa cells even in the presence of cytochalasin D (Schenkman et al. 1991b). Like amastigotes, trypomastigote forms are found inside a cell phagosome after invasion. Therefore, it is conceivable that the host cytoskeleton may undergo some modifications to accommodate the entry of a trypomastigote, which measures about 1 μm × 10 μm, inside such a large vacuole.

In this study we have examined the microfilament distribution of HeLa cells during trypomastigote invasion. We observed by confocal fluorescence microscopy that phalloidin stained microfilaments at the sites of invasion. In addition, the plasma membrane of HeLa cells encircles penetrating trypomastigotes, even in the presence of cytochalasin D, and parasites surrounded by these membrane processes are found in cytoplasmic vacuoles.

Slender T. cruzi trypomastigotes of the Y strain (Silva and Nussenzweig, 1953) were isolated from supernatants of infected LLC-MK2 cells (ATCC-CCL-7, Rockville, MD) as described (Schenkman et al. 1991a). Metacyclic trypomastigote forms from Y and G strains (Yoshida, 1983) were purified from aged epimastigote cultures grown at 28°C in LIT medium (Camargo, 1964), by passage through DEAE-cellulose (Sigma, St. Louis, MO) to remove epimastigotes (Yoshida, 1983). HeLa cells (ATCC-CCL 2, Instituto Adolfo Lutz, Sâo Paulo) were grown at 37°C in either Dulbecco’s modified Eagle’s (DMEM) or RPMI-1640 medium, supplemented with antibiotics (penicillin and streptomycin for DMEM, and gentamicin for RPMI) and 10% foetal bovine serum (FBS). Madin-Darby canine kidney cells, MDCK (a gift from W. Dolan, Department of Cell Biology, NYU), and Balb/3T3 fibroblasts, clone A31 (ATCC-Ccl-163), were also cultivated in DMEM medium with 10% FBS. The mammalian cells were removed from the flasks by trypsinization.

For invasion experiments, 3 × 10⁴ HeLa cells were grown at 37°C, plated onto 12 mm round glass coverslips and grown for at least 24 h. Then 1× 10⁷ parasites in 0.5 ml of medium were added to the cells and incubated for 30 min at 37°C. Metacyclic trypomastigotes (G strain) were washed with phosphate buffered saline (PBS) and pre-incubated with medium (DMEM or RPMI, with 10% FBS) for at least 2 h before centrifugation into HeLa cells, as described (Mortara, 1991). Unattached parasites were aspirated, the cells were washed three times with PBS or Hank’s solution, and the coverslips were immediately immersed in the appropriate fixative.

Cells and parasites were fixed in 3.5% formaldehyde or 4% paraformaldehyde in PBS for at least 1 h and the excess aldehyde groups quenched by a 15 min incubation with 50 mM NH4CI in PBS. The fixed coverslips were then soaked in PBS containing 0.25% gelatin and 0.05% NaN₃ (PGN), and kept at 4°C before being used. In the experiments using paraformaldehyde as fixative, when required the preparations were treated for 1 min with 0.1% Zwitergent (Calbiochem) or 0.1% Saponin (Sigma) after removal of the fixative. Labeling was performed by inverting the coverslips onto 25-pl drops of different reagents.

FTTC-phalloidin (Sigma) and FTTC-wheat germ agglutinin (FITC-WGA) (Sigma) were used at 10 μg ml⁻¹ in PGN. AntiHeLa glycoprotein (Concanavalin A-binding) antibodies raised in [A/O × LOU]FI rats were kindly provided by Gordon Koch (Koch et al. 1987). A rabbit antiserum to human transferrin receptor (HTR) was a gift from A. D. Lowe (MRC Laboratory of Molecular Biology, Cambridge, UK). Anti-T. cruzi rabbit serum was prepared by immunizing rabbits with live slender tissue culture-derived trypomasti-gotes inactivated with 8-methoxypsoralen (Andrews et al. 1985). All antibodies were diluted at least 1:100 in PGN. After several washes with PBS, bound immunoglobulins were visualized following incubation with the appropriate FTTC-labeled conjugate (Sigma), diluted 1:50 in PGN. After final washes in PBS, samples were mounted in 50% glycerol/PBS containing 1% p-phenylenediamine to reduce fading (Koch et al. 1987) and examined on a Lasersharp MRC-500 confocal fluorescence imaging system at the MRC Laboratory of Molecular Biology in Cambridge, UK. The confocal system was mounted on a Nikon epifluorescence microscope fitted with a ×63 Plan-Apo phase-contrast objective and a transmitted light extension probe. This set-up allowed the simul-taneous recording of both transmitted (phase-contrast) and confocal fluorescence images. Images were recorded on Ilford FP4 films, developed for 5 min at 23°C with Accutol (Patterson, UK) diluted 1:10 and fixed for 10 min with Kodak fixer.

Cytochalasin D (Sigma) treatment was carried out by pre-incubating the cells for 15 to 30 min in the presence of 10 μM of the drug. The parasites were added and the concentration of cytochalasin D readjusted to 10 μM when required (Schenkman et al. 1991b).

Samples were processed as for immunofluorescence with or without cytochalasin D treatment, and were fixed with 2% glutaraldehyde in 0.1 M sodium cacodylate buffer, pH 7.2, containing 0.1 M sucrose. The coverslips were kept at 4°C in fixative until processed as described previously (Schenkman et al. 1988; Mortara, 1991).

HeLa cells were plated in a 75 cm² flask and pre-treated for 30 min at 37°C with 1 ml of medium with or without 10 ×M cytochalasin D. The cells were incubated with 1 ml of medium containing 1 × 10⁸ trypomastigotes. After 20 to 30 min, the cultures were washed with Hank’s solution and fixed for at least 14 h with 2% glutaraldehyde in 0.1 M sodium cacodylate, pH 7.2, with 0.1 M sucrose. Cells and parasites were collected by scraping and pelleted by centrifugation. The pellets were post-fixed with 1% OsO₄ in 0.1 M sodium cacodylate, pH 7.2, dehydrated in an ethanol series, and embedded in Epon 812. Thin sections were mounted on Formvar/carbon-coated copper grids.

To observe peripheral microfilaments around invading trypomastigotes, HeLa cells were grown on poly-L-lysine Formvar/carbon-coated nickel grids. Trypomastigotes were added and after 30 min incubation, at 37°C, the grids were washed with Hank’s solution and fixed in the presence of nonionic detergent. We used a lysis-fixative solution described by Karlsson et al. (1984), which consisted of 0.137 M NaCl, 5 mM KC1, 1.1 mM Na-HPCL, 4 mM NaHCOa, 5.5 mM glucose, 2 mM MgC12, 2 mM EGTA, 20 mM piperazine-N,N’-bis[2-ethane sulfonic acid] (PIPES), pH 6.0, 1% Triton X-100 and 0.75% glutaraldehyde. The grids were stained with 1% aqueous uranyl acetate before air drying. All specimens were examined with a JEOL 1200 EXH electron microscope operated at 80 kV in the Centro de Microscopia Eletrônica of Escola Paulista de Medicina.

When tissue culture-derived or metacyclic trypomastigotes were incubated or centrifuged onto HeLa cells, and the preparations fixed and treated with FITC-phalloidin, fluorescent staining was observed at the sites of parasite attachment and/or invasion (Fig. 1). This staining is associated with the HeLa cell microfilaments and not with the parasites, since free parasites or parasites attached to HeLa cells pre-fixed with paraformaldehyde do not stain with phalloidin at the concentration used in this study (not shown). Phalloidin labeling on HeLa cells occurs at different magnitudes, ranging from discrete, almost punctate aggregates, through more extensive processes enveloping the trypomastigotes, which can reach the entire length of the extracellular portions of the invading parasites (Fig. 1A-C). Upon double labeling of non-permeabilized preparations, the sites of this extracellular regions stained with phalloidin cannot be detected with anti-T. cruzi IgG, which, unlike phalloidin, does not permeate paraformaldehyde-fixed cells (not shown). In some instances, a specific staining of actin microfilaments was detected around intracellular segments of parasites (Fig. 1D). Similar staining patterns were observed in 3T3-fibroblasts and in MDCK cells, but at a much lower frequency (not shown).

As trypomastigotes invade HeLa cells in the presence of 10 μM cytochalasin, we examined whether the phalloidin staining could still be observed in drug-treated cells. As shown in Fig. 2, HeLa cells preincubated with 10 μM cytochalasin D lose some of their connections to the substratum, and the microfilament meshwork is disrupted. Even under these conditions, specific but less-intense staining by phalloidin was observed at the sites of trypomastigote invasion (Fig. 2).

The detection of microfilaments around invading trypomastigotes at the cell periphery, apparently at extracellular parts of invading parasites, suggests that HeLa cells extend their membrane around the invading parasite. To confirm this possibility, we studied the distribution of several HeLa surface membrane markers under the confocal microscope, in paraformaldehyde-fixed, but non-permeabilized, samples. Wheat germ agglutinin (WGA) does not react with metacyclics of the G strain and was used in these experiments. Labeling of HeLa cells with WGA showed specific staining at the extracellular region of invading parasites (Fig. 3). A polyclonal antibody raised against concan-avalin A-binding glycoproteins of HeLa cells (Koch et al. 1987) also stained the regions corresponding to parts of trypomastigotes that appear to be outside the cells (Fig. 4). In the example shown, optical sections cut through parasites positioned perpendicularly to the cells give rise to rings of fluorescence at focal planes away from that of the substratum (arrows in Fig. 4). Parasites alone did not label with the anti-HeLa glycoprotein antibodies (not shown). Similar results were also obtained with a rabbit antiserum to human transferrin receptor (not shown).

Samples of trypomastigotes incubated with, or centri fuged onto, HeLa cells were processed for scanning electron microscopy. Several images of parasites entering the cells, revealed an extension of the cell membrane, which appears to grow around the invading trypomastigotes (Fig. 5).

Ultrathin sections cuts through identical samples revealed trypomastigotes completely surrounded by a concentric double membrane process intercalated with cytoplasmic material (Fig. 6). In these sections, trypomastigote were identified by characteristic ultrastructural markers such as the flagellum, sub-pellicular microtubules and kinetoplast. This‘sleeve-like’ membrane extension or pseudopodium around trypomastigotes was similarly observed in HeLa cells that had been pre-treated with 10 μM cytochalasin D (Fig. 7). Longitudinally cut sections confirmed the above observations (Fig. 8).

Transmission electron microscopy was also performed on whole-mounts of simultaneously fixed and extracted HeLa cells infected with trypomastigotes according to the method of Karlsson et al. (1984). After staining with uranyl acetate, this procedure permits the visualization of the intact cytoskeleton of the target cell and parasites. As shown in Fig. 9, remnants of HeLa cell surface processes were visible near invading parasites. Under higher magnification these loose structures were shown to contain a meshwork of thin filaments.

Observations of invading trypomastigotes revealed that the pseudopodium surrounding the parasites was also found within cytoplasmic vacuoles containing parasites.

Fig. 10 shows a typical image where the parasite appears to be surrounded by the pseudopodium inside a cytoplasmic vacuole. These observations suggest that HeLa cells internalize the entire process. Therefore, a second pseudopodium had been formed around the membrane extension enveloping the trypomastigotes. Observation of transverse thin sections of these specimens confirmed this notion and images of an‘onion’-like double process around intracellular (Fig. HA, B) and invading (Fig. 11C, D) parasites were observed. Examination of serial thin sections confirmed that the pseudopodia enveloping the trypomastigotes are actually inside the cell cytoplasm (Fig. 12). The fact that serial thin sections cut through parasites surrounded by a pseudopodium are found in the same position inside cytoplasmic vacuoles of a single cell excludes the possibility that such images arose by sectioning of parasites invading adjacent cells (Fig. 12). These studies also revealed that the intracellular (and in some rare images, also the extracellular) double membrane structure could be found at various degrees of vésiculation (Fig. 13). Although we detected pseudopodia around invading trypomastigotes in cytochalasin D-treated HeLa cells (Fig. 7), the phenomenon of pseudopodium internalization was not detected in these samples.

In this paper we present evidence that during invasion of HeLa cells by T. cruzi trypomastigotes there is a previously undescribed reorganization of host cell membrane. The HeLa cell plasma membrane, in association with the cortical actin cytoskeleton, can envelope the entire body of invading trypomastigotes before complete parasite internalization. Further, both actin microfilaments and membrane vesicles can be found inside vacuoles containing parasites.

The specific fluorescence staining of cellular but not parasite actin around invading parasites was demonstrated by confocal fluorescence microscopy. The finding that the labeling around apparently extracellular parasites indeed corresponds to extensions of HeLa cytoplasm was confirmed by fluorescence staining of HeLa surface markers around the parasites, by the fact that anti-T. cruzi antibodies did not label the phalloidin-stained material in unpermeabilized preparations, and by direct observation of host membranes around the parasites by scanning and transmission electron microscopy. Also, the direct staining of fixed-extracted preparations shows HeLa microfilaments closely associated with the parasite.

These observations show that there is a pseudopodium extension with corresponding staining of micro-filaments around invading parasites. Formation of this‘sleeve-like’ membrane protrusion around invading trypomastigotes could be brought about by localized actin polymerization that would promote or facilitate trypomastigote internalization, as shown in the endo-cytosis of Shigella (Clerc and Sansonetti, 1987). Actin polymerization is thought to be the major mechanochemical cytoplasmic force involved in pseudopodial extension in phagocytosis by professional phagocytes (Silverstein et al. 1989), or in the movement, membrane ruffling and pseudopodial extension of non-phagocytic cells (Bray and White, 1988). The primary role of actin polymerization in all these processes has been deduced by the fact that cytochalasin D, a drug that primarily disrupts long actin microfilaments, is inhibitory.

However, we have previously shown that cytochalasin D does not inhibit T. cruzi trypomastigote invasion of HeLa cells (Schenkman et al. 1991b), and now we report that it does not prevent the formation of membrane extensions by HeLa cells. The fact that cytochalasin D had disrupted the microfilament meshwork in these cells is clearly demonstrated by the typical phalloidin staining, HeLa cell rounding, and microvilli clustering. Therefore, in the present case, membrane extension is independent of the integrity of the cortical microfilament system and the presence of microfilaments within the membrane extension may represent actin association with membrane molecules, probably triggered by the interaction of the cell surface with the parasite, as shown in many cell types (Geiger, 1989; Small, 1989). Similar independence of cytochalasin, with concomitant staining of microfilament, was observed during the attachment of T. cruzi amastigotes to HeLa cells (Mortara, 1991). In this latter case, however, parasite invasion is diminished, suggesting that amastigote invasion requires the participation of mechano-chemical forces from the target cell cytoskeleton. Another example of localized actin staining without internalization has been described during the interaction of macrophages with yeast particles (Painter et al. 1981).

The initial steps leading to the entry of T. cruzi trypomastigotes into mammalian cells involve parasite swimming driven by flagellar motion and contacts with host cell surface components, which lead to the formation of stable (possibly cooperative) associations, as described before, after studies of parasite attachment to fixed cells (Schenkman et al. 1991a). After stable contact is established, the active participation of the moving trypomastigote could provide the major mechanical force necessary for membrane extension. This would be, as previously suggested, a‘self-zippering’ process, where most of the required energy comes from T. cruzi itself. In this hypothetic model, cell microfilaments would be drawn passively into the growing projections. This rather simplistic view could explain the formation of a pseudopodium around an invading trypomastigote,’ even in the presence of cytochalasin D.

Our observations represent static observations of a highly dynamic process. The possibility cannot be excluded that pseudopodial extension around invading parasites could result from cell retraction during parasite invasion, although our scanning electron micrographs indicate that this is unlikely. Whichever the actual mechanism, these pseudopodia appear to be further internalized, since we frequently found actin microfilaments and vesicles around intracellular trypomastigotes, probably the remains of the pseudopodia, inside endocytic cytoplasmic vacuoles. This highly unusual manifestation of the endocytic process appears to evolve to the internalization of a‘self’ membrane protrusion. The participation of the cell microfilament system in this process seems to be crucial, since in cytochalasin D-treated HeLa cells no such internalization was detected.

We do not know whether the phenomenon described here is necessary for invasion. We have also examined the invasion of 3T3 and MDCK cells by trypomastigotes, and the phenomenon of pseudopodia formation around trypomastigotes seems to occur, albeit at a much lower frequency. It is possible that this is a shortlived phenomenon in those cells, or it may be especially profuse in HeLa cells. Further studies aimed at characterizing the molecular events required for the onset of this new event may provide important information for understanding the mechanism of cell invasion by T. cruzi.

We thank Gordon L. E. Koch for critical discussions and access to the confocal microscope. We also thank W. Brad Amos and John White for their help with the confocal microscope, Edith S. Robbins for the use of Scanning Electron Microscope, Rocilda P. F. Schenkman for the assistance during this work, and Victor Nussenzweig for suggestions and discussions. This work was supported by grants from the MacArthur Foundation, the UNDP/World Bank/WHO Special Program for Research and Training in Tropical Diseases and the Rockefeller Foundation, American Heart Association (Grant-in-Aid, no. 901216), Conselho Nacional de Desenvolvimento Científico e Tecnológico-CNPq (Brazil), and Fundaçào de Amparo à Pesquisa de Estado de Sâo Paulo-FAPESP.