Fas/CD95-Mediated Apoptosis of Type II Cells Is Blocked by Toxoplasma gondii Primarily via Interference with the Mitochondrial Amplification Loop

Abstract

The intracellular protozoan Toxoplasma gondii induces persistent infections in various hosts and is an important opportunistic pathogen of humans with immature or deficient immune responses. The ability to survive intracellularly largely depends on the blocking of different proapoptotic signaling cascades of its host cell. Fas/CD95 triggers an apoptotic cascade that is crucial for immunity and the outcome of infectious diseases. We have determined the mechanism by which T. gondii counteracts death receptor-mediated cell death in type II cells that transduce Fas/CD95 ligation via caspase 8-mediated activation of the mitochondrial amplification loop. The results showed that infection with T. gondii significantly reduced Fas/CD95-triggered apoptosis in HeLa cells by inhibiting the activities of initiator caspases 8 and 9 and effector caspase 3/7. Parasitic infection dose dependently diminished cleavage of caspase 8, the BH3-only protein Bid, and the downstream caspases 9 and 3. Importantly, interference with Fas/CD95-triggered caspase 8 and caspase 3/7 activities after parasitic infection was largely dependent on the presence of caspase 9. Within the mitochondrial amplification loop, T. gondii significantly inhibited the Fas/CD95-triggered release of cytochrome c into the host cell cytosol. These results indicate that T. gondii inhibits Fas/CD95-mediated apoptosis in type II cells primarily by decreasing the apoptogenic function of mitochondria.

The apicomplexan Toxoplasma gondii is one of the most common intracellular parasites of humans and animals and an important opportunistic pathogen of immunocompromised hosts. After infection, T. gondii renders host cells resistant to apoptosis induced by various proapoptotic stimuli (3, 21, 31). This is assumed to be a crucial adaptation of the parasite to its lifestyle that allows sustained intracellular survival and establishing of long-term persistence within its hosts, including up to 30% of humans worldwide. The significance of being able to block host cell apoptosis is underscored by the undertaking of multiple antiapoptotic activities by T. gondii. This includes the inhibition of cytochrome c release from host cell mitochondria (4, 11), upregulation of antiapoptotic proteins of the Bcl-2 and IAP families (11, 25), interference with proapoptotic Bax and Bad (4), modulation of cell death-regulating kinase activities (4, 15), inhibition of caspase 8 (34), and possibly modulation of cytochrome c-induced caspase 3/7 activation (14).

Fas/CD95 belongs to the tumor necrosis factor (TNF) receptor superfamily and is the prototype of death receptors that initiate an apoptotic cascade. Its triggering plays critical roles in the regulation of the immune system, immunity to microorganisms, and the pathogenesis of infectious diseases (8, 16), including toxoplasmosis (10, 13, 20, 26, 32). Ligation of Fas/CD95 recruits and activates the initiator caspase 8 (2, 24). In so-called type I cells, the amount of active caspase 8 suffices to directly activate downstream effector caspases 3, 6, and 7, which then execute the apoptotic program (29). In contrast, in type II cells, small amounts of caspase 8 do not suffice to directly activate caspase 3 (30). Instead, the proapoptotic signal has to be amplified via cleavage of the BH3-only protein Bid, Bax/Bak-assisted release of cytochrome c from the mitochondria, and activation of caspase 9 and subsequently caspase 3 (17, 29-30). Indeed, the concept of two different signaling pathways following Fas/CD95 ligation has recently been confirmed by genetic analyses (28) and may relate in vivo to different cell types or different sensitivities to Fas/CD95 ligation (1).

We have recently shown that, in type I cells, T. gondii inhibits Fas/CD95-triggered apoptosis by inducing noncanonical processing and degradation of caspase 8 (34). However, due to the profound differences in the regulation of death receptor-mediated signaling in type I and type II cells, the parasite mechanisms to block apoptosis in these cells may also differ considerably. We have therefore pinpointed here the cellular target of T. gondii in cells in which Fas/CD95-ligation has to be amplified via the mitochondrial amplification loop in order to induce host cell death. Our results identify the mitochondrial amplification loop as the regulatory step at which Fas/CD95-triggered apoptosis in type II cells is strongly inhibited after infection. This indicates that T. gondii has evolved at least two mechanisms to block Fas/CD95-triggered apoptosis, thereby ensuring the viability of its host cell under different physiological conditions.

MATERIALS AND METHODS

Cell culture and parasite infection.

Fas/CD95-transfected HeLa cells (HeLa-Fas), a caspase 9-deficient Jurkat T-cell clone (JMR), and a caspase 9-reconstituted transfectant derived thereof (F9) have been described previously (28, 35) and were cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal calf serum, 100 U of penicillin/ml, and 100 μg of streptomycin/ml. Tachyzoites of the mouse-avirulent type II strain NTE of T. gondii were propagated in L929 murine fibroblasts as host cells in RPMI 1640 medium, 1% heat-inactivated fetal calf serum, and antibiotics as described above. For infection of HeLa-Fas or Jurkat cells, parasites were isolated from L929 cocultures by differential centrifugation as described previously (34). HeLa-Fas cells were infected at parasite/host cell ratios of 10:1 or 20:1 for 24 h, which led to infection rates of 55% ± 4% and 78% ± 4%, respectively. In order to account for the lower infection efficiency, nonadherent Jurkat clones were infected at a parasite/host cell ratio of 30:1, thereby yielding infection rates of 82% ± 6%. During the final 6 h (HeLa-Fas) or the final 1 to 3 h (Jurkat clones), apoptosis was induced by using 0.25 μg of agonistic anti-Fas/CD95 antibody (clone CH11; Upstate Biotechnology, New York, NY)/ml. In some experiments, death receptor-mediated apoptosis was induced by treating Jurkat clones with 10 ng of TNF-α (Becton Dickinson, Heidelberg, Germany)/ml, along with 1 μg of cycloheximide/ml for 12 h.

Morphological detection of apoptosis and immunofluorescence staining.

The condensation of chromatin was determined as a morphological characteristic of apoptosis. For this purpose, T. gondii-infected or uninfected HeLa-Fas cells treated or not to undergo apoptosis, were fixed in 4% paraformaldehyde-0.1 M sodium cacodylate (pH 7.4) for 30 min. After having been washed in phosphate-buffered saline (PBS; pH 7.4), cells were stained with 50 ng of Hoechst 33258 (Sigma, Deisenhofen, Germany)/ml in PBS for 1 h. After being washed, cells were mounted with Mowiol (Calbiochem, Schwalbach, Germany) and examined by fluorescence microscopy. At least 500 cells from each sample were analyzed. In order to visualize apoptotic cells and intracellular parasites simultaneously, DNA strand breaks were labeled with fluorescein isothiocyanate-conjugated dUTP by using a TUNEL (terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling) assay, and parasites were labeled by immunofluorescence staining. To this end, cells were fixed as described above, quenched with 50 mM NH₄Cl in PBS for 10 min, and then permeabilized for 1 h with 0.1 mg of saponin/ml in PBS containing 1% bovine serum albumin (BSA). Parasites were sequentially immunolabeled with rabbit anti-Toxoplasma hyperimmune serum and 6.5 μg of Cy5-conjugated F(ab′)₂ fragment donkey anti-rabbit immunoglobulin G (IgG; Dianova, Hamburg, Germany)/ml diluted in PBS with saponin and BSA. After they were washed, cells were stained by the TUNEL assay as recommended by the manufacturer (Boehringer, Mannheim, Germany). Briefly, cells were permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate for 2 min on ice and then incubated with the TUNEL reaction mixture for 1 h at 37°C. Thereafter, cells were stained with 5 μg of propidium iodide/ml in PBS to visualize total cells. Coverslips were mounted with Mowiol and examined by confocal laser scanning microscopy using a Leica TCS SP2.

Fluorometric measurement of caspase activity.

Caspase activities were determined fluorometrically by measuring the cleavage of caspase-specific substrates essentially as described previously (34). Briefly, HeLa-Fas cells or Jurkat clones (10⁶ cells/sample) were isolated by trypsinization or resuspension, respectively. After they were washed in PBS, the cells were extracted with 50 μl of 1% Nonidet P40, 150 mM NaCl, 50 mM Tris-HCl (pH 8.0), and complete protease inhibitor cocktail (EDTA-free; Roche, Mannheim, Germany) for 15 min on ice. After centrifugation for 5 min at 14,000 × g and 4°C, 10-μl portions of the supernatants were mixed in triplicate with 90 μl of either 10 μM Ac-DEVD-amino-4-methyl-coumarin (AMC; caspase 3/7-specific; Bachem, Weil am Rhein, Germany), 50 μM Ac-IETD-AMC (caspase 8-specific; Alexis, Grünberg, Germany), or 50 μM Ac-LEHD-AMC (caspase 9-specific; Bachem) in 0.1 mg of BSA/ml, 0.1% CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}, 10 mM HEPES, 50 mM NaCl, 40 mM β-glycerophosphate, 2 mM MgCl₂, and 5 mM EGTA (pH 7.0). The kinetics of substrate cleavage were recorded over a period of 60 min at 37°C at excitation and emission wavelengths of 380 and 460 nm, respectively, using a Victor V Multilabel counter (Perkin-Elmer, Boston, MA). The increase in substrate cleavage with time was calculated as a measure of caspase activity.

Western blot analyses.

Toxoplasma-infected HeLa-Fas cells and Jurkat clones or uninfected controls were collected by trypsinization or resuspension, respectively. For total cellular lysates, the cells were lysed for 1 h in 1% Triton X-100, 150 mM NaCl, 50 mM Tris-HCl (pH 8.0), 50 mM NaF, 5 mM sodium pyrophosphate, 1 mM phenylmethylsulfonyl fluoride, 1 mM EDTA, 1 mM sodium orthovanadate, and 5 μg each of leupeptin, aprotinin, and pepstatin/ml. In order to determine the subcellular distribution of cytochrome c, digitonin-soluble and -insoluble fractions were prepared as described previously (11). Briefly, HeLa-Fas cells were resuspended in PBS and mixed with an equal volume of 150 μg of digitonin/ml in 0.5 M sucrose. After 30 s, heavy organelles, including mitochondria, were pelleted at 14,000 × g for 1 min, and the supernatants were saved as cytosol-containing digitonin-soluble fractions. The digitonin-insoluble pellets were then extracted as described above. After centrifugation, soluble proteins were separated by standard sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose by semidry blotting. Equal loading of each lane was confirmed by staining with Ponceau S. Prior to the immunodetection of specific proteins, nonspecific binding sites had been blocked by using 5% dry skimmed milk-0.2% Tween 20-0.02% NaN₃ in PBS (pH 7.4). Thereafter, membranes were probed overnight at 4°C with mouse monoclonal anti-caspase 8 (clone 1C12, recognizing an epitope within the p18 fragment; 1:1,000 [Cell Signaling Technology, Beverly, MA]), rabbit polyclonal anti-caspase 9 (directed against the p35 fragment; 1:250 [Santa Cruz Biotechnology, Heidelberg, Germany]), goat polyclonal anti-caspase 3 (1:1,000 [R&D Systems, Wiesbaden-Nordenstadt, Germany]), rabbit polyclonal anti-Bid (1:250 [BD Pharmingen, Heidelberg, Germany]), mouse monoclonal anti-actin (clone C4; 1:10,000 [kindly provided by J. Lessard, Cincinnati, OH]), 2 μg of mouse anti-cytochrome c (clone 7H8.2C12 [BD Pharmingen])/ml, or 2 μg of anti-cytochrome c oxidase (COX) subunit IV (clone 10G8-D12-C12 [Molecular Probes, Leiden, The Netherlands])/ml diluted in 5% dry skimmed milk-0.05% Tween 20 in PBS (pH 7.4). After three washes with 0.05% Tween 20 in PBS (pH 7.4), immune complexes were labeled for 90 min with horseradish peroxidase-conjugated anti-mouse, anti-rabbit, or anti-goat IgG (1:10,000 [Dianova, Hamburg, Germany]). After extensive washes, bound antibodies were visualized by ECL chemiluminescence (GE Healthcare, Freiburg, Germany).

Subcellular distribution of cytochrome c.

Cytochrome c was morphologically detected by immunofluorescence staining and confocal microscopy. For this purpose, infected and uninfected HeLa-Fas cells were fixed with 4% paraformaldehyde, quenched, and permeabilized with saponin as described above. The cells were then incubated for 1 h with 10 μg of anti-cytochrome c (clone 6H2.B4; BD Pharmingen)/ml, along with rabbit anti-Toxoplasma hyperimmune serum diluted in PBS containing saponin and BSA. After being washed, the cells were incubated for 1 h with 12 μg of Cy2-conjugated F(ab′)₂ fragment anti-rabbit IgG/ml, along with 1.6 μg of Cy3-conjugated F(ab′)₂ fragment anti-mouse IgG (Dianova)/ml. Coverslips were mounted with Mowiol and examined by using confocal laser scanning microscopy.

Statistical analysis.

Results are expressed as means ± the standard error of the mean (SEM) of at least three independent experiments unless otherwise indicated. Significant differences between mean values were identified by using the Student t test. In order to avoid false-positive results due to multiple comparisons, P values were multiplied with the number of comparisons (Bonferroni correction). Corrected P values of <0.05 were considered significant.

RESULTS

T. gondii inhibits apoptosis in Fas-triggered type II cells.

The capacity of T. gondii to inhibit Fas-induced apoptosis was evaluated in HeLa cells stably transfected with a plasmid encoding human Fas/CD95 (35). HeLa cells have been shown to behave as type II cells since overexpression of Bcl-2 prevents CD95-mediated apoptosis (23, 36). Treatment of these cells with agonistic anti-Fas/CD95 induced apoptosis in the large majority of cells; however, prior infection with T. gondii significantly blocked Fas/CD95-triggered cell death, as was expected from previous results (Fig. 1A and B; P = 0.008; Student t test). Importantly, simultaneous detection of DNA strand breaks and T. gondii indicated that HeLa-Fas cells harboring intracellular parasites were largely protected from undergoing Fas/CD95-triggered apoptosis (Fig. 1A, arrows), whereas parasite-negative cells were not. After Hoechst staining and quantification of cells with condensed chromatin, however, we also recognized that T. gondii did not completely block Fas/CD95-mediated apoptosis since 78% ± 4% infected cells (data not shown) corresponded to a 60% ± 4% reduction of cells with condensed chromatin (Fig. 1B). In order to confirm that apoptosis induced via Fas/CD95 in HeLa-Fas cells largely depended on the mitochondrial amplification loop, they were treated with a caspase 9-specific inhibitor. The results showed that Z-LEHD-FMK, but not its vehicle completely abrogated apoptosis, as revealed by the strong inhibition of both caspase 9 and effector caspase 3/7 activities (Fig. 1C). Together, these results indicated that infection with T. gondii renders type II cells resistant to Fas/CD95-induced apoptosis.

FIG. 1.

Inhibition of Fas/CD95-mediated apoptosis in type II cells by T. gondii. (A and B) HeLa-Fas cells were infected with T. gondii at a parasite/host cell ratio of 20:1 for 24 h or were left uninfected. Apoptosis was induced during the last 6 h of infection with an agonistic anti-Fas/CD95 antibody where indicated. (A) After fixation, DNA strand breaks and T. gondii were detected by TUNEL assay (green fluorescence) and immunofluorescence staining (blue fluorescence), respectively. Total cells were visualized by propidium iodide (red fluorescence). (B) Alternatively, cells were stained with Hoechst 33258, and the percentage of cells with condensed chromatin was determined by fluorescence microscopy. Bars represent means ± the SEM from three independent experiments. Significant differences between percentages of apoptotic cells for infected and noninfected cells are marked (**, P < 0.01; Student t test). (C) Uninfected HeLa-Fas cells were treated with anti-Fas/CD95 in the presence or absence of the caspase 9 inhibitor LEHD-FMK or its vehicle dimethyl sulfoxide. After lysis of the cells, cleavage of the caspase3/7-specific substrate DEVD-AMC (▪) or the caspase 9-specific substrate LEHD-AMC (▒) was measured fluorimetrically. The data represent mean cleavage activities ± the SEM from two independent experiments.

Caspase cleavage and activities are broadly inhibited after infection.

In order to determine underlying mechanisms of the antiapoptotic effect of T. gondii in type II cells, the activities of caspases 8, 9, and 3/7 were determined by measuring the cleavage of specific substrates fluorometrically. Treatment of uninfected HeLa-Fas cells with anti-Fas/CD95 antibody strongly activated the initiator caspases 8 and 9, as well as the effector caspase 3/7 (Fig. 2). In contrast, the activities of all caspases were significantly decreased after infection with T. gondii at a parasite/host cell ratio of 10:1 (P < 0.05; Student t test). Infection of HeLa-Fas cells with increasing numbers of parasites further decreased the substrate cleavage to background levels (P < 0.01 for caspases 9 and 3/7; Fig. 2). Since caspase activities rely on the proteolytic cleavage of inactive zymogens, we next determined the appearance of caspase subunits after Fas/CD95 ligation by immunoblotting. The results showed that after treatment of uninfected HeLa-Fas cells with anti-Fas/CD95, caspases 8, 9, and 3 were cleaved into intermediate cleavage products and/or active caspase subunits (Fig. 3). As expected, the BH3-only protein Bid was also cleaved into proapoptotic tBid, thus confirming the activation of the mitochondrial amplification loop in anti-Fas/CD95-treated HeLa-Fas cells (Fig. 3). Importantly, infection with T. gondii prior to Fas/CD95 activation dose dependently decreased the cleavage products of caspases 8, 9, and 3, as well as tBid (Fig. 3). In addition, we also recognized a decrease of the full-length protein levels, particularly of Bid but also caspase 9 and 3, after Fas/CD95-triggering that occurred independently of the cleavage into active subunits (Fig. 3). In contrast to caspase 8, however, such a decrease of full-length Bid and caspases 9 and 3 was not further accelerated in the presence of T. gondii, thus arguing against a parasite-induced process. Together, these results indicated that infection with T. gondii conferred broad defects in the activation of the death-receptor signaling and the mitochondrial amplification loop after Fas/CD95 ligation.

FIG. 2.

Infection of HeLa-Fas cells with T. gondii inhibits Fas/CD95-triggered initiator and effector caspase activities. Cells were infected with T. gondii at different parasite/host cell ratios or were left uninfected. After 18 h, cells were induced to undergo apoptosis with agonistic anti-Fas/CD95 for an additional 6 h. Cells were then lysed, and protein extracts were tested in triplicates for the activities of caspase 8, 9, and 3/7 by measuring the cleavage of specific substrates fluorometrically. The data represent mean cleavage activities ± the SEM from three independent experiments. Significant differences between caspase activities of infected and uninfected cells are marked (*, P < 0.05; **, P < 0.01 [Student t test]).

FIG. 3.

Expression and Fas/CD95-triggered cleavage of caspases and the BH3-only protein Bid after infection with T. gondii. HeLa-Fas cells were infected with T. gondii at different parasite/host cell ratios for 24 h or were left uninfected. During the final 6 h of infection, cells were induced to undergo apoptosis with agonistic anti-Fas/CD95. Protein extracts and a control extract of Fas/CD95-treated Jurkat cells (Ctr) were separated by SDS-PAGE and immunoblotted with antibodies recognizing full-length proteins or cleavage products of caspases 8, 9, and 3, as well as Bid. Immunolabeling of actin was performed to ensure equal loading of each lane. The results are representative of three independent experiments.

T. gondii intersects death receptor signaling in type II cells primarily during mitochondrial amplification.

Fas/CD95-mediated signaling in type II cells does not only rely on the apoptogenic function of mitochondria but may also be amplified by a positive feedback activation of the initiator caspase 8 downstream of the mitochondria (17, 29). Therefore, we next determined at which step of the signaling cascade T. gondii interferes with death receptor-mediated apoptosis. In order to address this question, we used caspase 9-deficient Jurkat cells and a Jurkat clone in which caspase 9 was stably complemented (28). Jurkat cells, like HeLa cells, behave as type II cells because Bcl-2 overexpression abolishes CD95-mediated apoptosis (29, 30). Immunoblotting of complete cellular lysates confirmed the complete absence of caspase 9 in JMR and its complementation in F9 transfectants (Fig. 4B). Until 2 h after ligation of Fas/CD95, caspase 8 and caspase 3/7 were strongly activated in F9 but not in JMR cells, thereby confirming that death receptor-mediated apoptosis in these cells largely depended on the mitochondrial amplification loop (Fig. 4A). Control immunoblot analysis excluded the possibility that the defective caspase activities in JMR cells were due to impaired or absent caspase 8 expression (Fig. 4B). Importantly, concomitant infection of the cells with T. gondii abolished caspase 8 and 3/7 activities in F9 cells (Fig. 4A; P < 0.05; Student t test), whereas no differences were observed between infected and uninfected JMR cells. Caspase activities further increased until 3 h posttreatment in uninfected F9 cells but were again significantly reduced in parasite-infected cells (P < 0.01). Only at that time point were caspase 8 and caspase 3/7 slightly activated in caspase 9-deficient JMR cells (Fig. 4A). Interestingly, the activity of caspase 8 at 3 h was significantly lower in infected than in uninfected JMR cells (P < 0.05), whereas that of caspase 3/7 was not diminished by T. gondii. These results clearly established that T. gondii blocked Fas/CD95-triggered apoptosis in type II cells primarily within the mitochondrial amplification loop, whereas parasite-mediated interference with caspase 8 contributed only to a minor extent.

FIG. 4.

T. gondii inhibits death receptor-initiated apoptosis in type II cells via interference with a step that relies on caspase 9. Caspase 9-deficient Jurkat cells (JMR) and a derivative clone thereof complemented with caspase 9 (F9) were infected with T. gondii for 24 h (open circles and bars) or were left uninfected (closed circles and bars). Cells were treated with agonistic anti-Fas/CD95 during the final 1 to 3 h of infection (A) or else were treated with TNF-α and/or cycloheximide (CHX) during the last 12 h (C). Protein extracts were prepared and tested in triplicates for the activities of caspase 8 or 3/7 by measuring the cleavage of specific substrates fluorometrically. The data represent mean cleavage activities ± the SEM from three independent experiments. In order to visualize the effects of T. gondii in JMR cells more clearly, bar graphs with a reduced y-axis scale displaying caspase activities at 0 and 3 h of anti-Fas/CD95 treatment have been partially included (insets). Significant differences between caspase activities of infected and uninfected cells are marked (*, P < 0.05; **, P < 0.01 [Student t test]). (B) Extracts from uninfected JMR and F9 cells were separated by SDS-PAGE and immunoblotted with antibodies recognizing caspase 9, caspase 8, or actin.

In order to test whether T. gondii generally intersects with death receptor signaling in type II cells within the amplifying mitochondrial pathway, we measured caspase activities in infected and noninfected JMR and F9 cells after treatment with TNF-α in the presence of cycloheximide. Cycloheximide was added in order to prevent NF-κB-mediated upregulation of antiapoptotic proteins that would counterbalance the proapoptotic function of TNF-α through TNF receptor I-mediated signaling. After 12 h of treatment with TNF-α-cycloheximide, both caspase 8 and caspase 3/7 activities were strongly increased in F9 cells; however, prior infection with T. gondii completely abolished TNF-α-mediated signaling (P < 0.05 and P < 0.01, respectively; Fig. 4C). In contrast, TNF-α/cycloheximide did not induce substantial caspase activities in JMR cells, and only caspase 3/7 activity was slightly reduced after parasitic infection (P < 0.05). Thus, death receptor-mediated apoptosis in type II cells appears to be generally inhibited by T. gondii at the level of the mitochondrial amplification loop.

Different mechanisms may enable T. gondii to inhibit the mitochondrial apoptotic pathway; some of these mechanisms operate upstream of the mitochondria, whereas others act downstream (4, 11, 14, 25). In order to distinguish between these two possibilities during Fas/CD95-mediated apoptosis in type II cells, the release of cytochrome c from the mitochondria into the cytosol was determined. Immunoblot analyses of subcellular fractions revealed that the amount of cytochrome c strongly increased in the cytosol of HeLa-Fas cells after Fas/CD95 activation and concomitantly decreased in the mitochondrion-containing fraction (Fig. 5A). An additional band at approximately 60 kDa indicated the formation of SDS-resistant cytochrome c multimers, as also shown previously (11). Importantly, the Fas/CD95-triggered redistribution of cytochrome c was considerably lower in T. gondii-infected cells than in uninfected controls. Staining of the mitochondrial marker protein COX excluded the possibility that the detection of cytochrome c in the cytosolic fraction after induction of apoptosis was due to a contamination with mitochondria. In addition, reprobing with an anti-actin antibody revealed that the decrease of cytosolic cytochrome c levels was not due to an unequal protein loading (Fig. 5A). Immunocytochemical analysis confirmed that treatment of uninfected HeLa-Fas cells with agonistic anti-Fas/CD95 led to the release of cytochrome c from mitochondria into the cytosol, as revealed by the homogeneous distribution in the majority of cells (Fig. 5B). When cells had been previously infected with T. gondii, however, cytochrome c was mostly associated with vesicular structures, thus indicating a mitochondrial distribution. Quantitative analyses showed that the percentages of cells with a homogeneous distribution of cytochrome c was indeed significantly decreased after parasitic infection (P < 0.05; Student t test). Together, these results suggested that, within the mitochondrial amplification loop, T. gondii considerably inhibits Fas/CD95-triggered apoptosis in type II cells via decreased redistribution of cytochrome c.

FIG. 5.

Decreased cytochrome c release in T. gondii-infected type II cells after Fas/CD95 activation. HeLa-Fas cells were infected with T. gondii at a parasite/host cell ratio of 20:1 for 18 h or left uninfected and then treated or not treated with agonistic anti-Fas/CD95 for an additional 6 h. (A) Subcellular fractions were obtained by cell lysis using digitonin, separated by SDS-PAGE, and immunolabeled using antibodies recognizing cytochrome c, the mitochondrial marker protein COX, and actin. (B) Alternatively, cytochrome c (red fluorescence) and T. gondii (green fluorescence) were morphologically detected by indirect immunofluorescence staining. Representative images of both labelings were recorded by confocal laser scanning microscopy and superimposed. The data represent mean percentages ± the SEM (n = 3) of cells with a homogeneous, i.e., cytosolic distribution of cytochrome c. Significant differences are indicated.

DISCUSSION

The inhibition of host cell apoptosis is one of the hallmarks of the interaction of intracellular microorganisms with their host cells (3, 21, 31). The critical importance of apoptosis as antimicrobial effector mechanism makes its parasite-mediated inhibition a promising novel target to combat various infectious diseases that are caused by intracellular pathogens. Apoptosis-regulating therapeutic agents have indeed entered clinical testing or have even been approved for the treatment of several noninfectious disorders (9). Before such a scenario is also feasible for the treatment of infectious diseases, however, the cellular targets of the microbial interference with host cell apoptosis needs to be unraveled.

Herein, we provide clear evidence that after Fas/CD95 ligation in type II cells, T. gondii blocks the apoptotic program of the host cell primarily at the level of the mitochondrial amplification loop (Fig. 6). In contrast, parasite interference with the initiator caspase 8 was recognized only during prolonged Fas/CD95 triggering and to a minor extent. This is remarkable, since we have previously shown that the parasite abrogates Fas/CD95-triggered apoptosis in type I cells via the cleavage and degradation of caspase 8, i.e., during an early step of the apoptotic cascade (Fig. 6) (34). Parasite-mediated degradation of caspase 8 at least under proapoptotic conditions was confirmed in the present study to also occur in type II HeLa cells. It was accompanied by inhibition of Fas/CD95-triggered apoptosis and a prominent decrease in the activities and the cleavage of several caspases, including the initiator caspase 8 furthest upstream. Furthermore, the cleavage of the BH3-only protein Bid, i.e., the molecular link between the death receptor pathway and the mitochondrial amplification loop (19, 22, 38), was also prominently decreased after infection with T. gondii. Whereas these results were consistent with the view of a parasite interference with initiator caspase 8, analyses of caspase 9-deficient cells instead clearly showed the critical importance of the apoptogenic function of mitochondria in Fas/CD95-mediated activation of effector caspase 3/7 and its inhibition in T. gondii-infected type II cells. Caspase 8 in type II cells is indeed mainly activated by a step downstream from the mitochondria, i.e., a positive feedback amplification by effector caspases (Fig. 6) and can therefore be inhibited by overexpression of Bcl-2 or Bcl-x_L (17, 19, 29). We support these findings by showing a severe defect in caspase 8 activation after Fas/CD95 ligation in caspase 9-deficient cells. The prominent decrease of caspase 8 activity observed after the infection of caspase 9-expressing cells with T. gondii, therefore, was rather the consequence than the underlying mechanism of reduced caspase 9 and 3 activities in the presence of the parasite. It is worthy noting that in the absence of caspase 9, sustained Fas/CD95 or TNF receptor-triggering slightly activated caspase 8 and caspase 3/7, respectively, and that such activity was also blocked by T. gondii. This indicates that the parasite is in principle able to directly interfere with caspase 8 activation in type II cells, as previously described in type I cells (Fig. 6) (34). Since the apoptotic program in type II cells under physiological conditions, i.e., in the presence of caspase 9, however, critically depends on the apoptogenic function of mitochondria, interference with death receptor-mediated apoptosis in type II cells clearly relies on the parasite intersection with the mitochondrial amplification loop (Fig. 6).

FIG. 6.

Apoptotic signaling pathways after ligation of Fas/CD95 in type I and type II cells and the levels of interference by T. gondii. In type I cells, Fas/CD95-ligation triggers effector caspases directly by the initiator caspase 8 without the need of a mitochondrial amplification loop. Such signaling pathway is blocked by interference of T. gondii with caspase 8. Fas/CD95-mediated cell death in type II cells mainly relies on the mitochondrial amplification loop and inhibition of this apoptotic pathway by T. gondii largely occurs at a mitochondrial step.

After activation of apoptosis by actinomycin D or UV light, i.e., after triggering the intrinsic pathway of apoptosis, parasites from both the mouse-virulent RH strain and the mouse-avirulent NTE strain of T. gondii block the redistribution of cytochrome c from the host cell mitochondria into the cytosol (4, 11). Similarly, parasite-mediated inhibition of the mitochondrial apoptotic pathway after triggering of Fas/CD95 in type II cells also involved blockade of cytochrome c release as we have shown in the present study again using the NTE strain of Toxoplasma. Based on the studies mentioned above (4, 11), it can be assumed that Fas/CD95-triggered cytochrome c release is inhibited also by other parasite strains, although this needs to be confirmed.

The apoptosis-sensing function of tBid relies on the activation and insertion of proapoptotic Bcl-2 family proteins Bax and/or Bak into the outer mitochondrial membrane, which can be counteracted by antiapoptotic Bcl-2-like proteins, e.g., Bcl-2, Bcl-xL, and others (7, 19, 27, 37). The inhibition of the mitochondrial amplification loop by T. gondii after triggering Fas/CD95, therefore, likely relies on an altered balance between pro- and antiapoptotic Bcl-2 family members, as previously described after activation of the intrinsic apoptotic pathway (4, 11, 25). However, we cannot completely exclude the possibility that a defect in the cytochrome c-triggered caspase 3-activity downstream of mitochondria (4, 14) also plays a role via a feedback regulation of the apoptogenic function of mitochondria. Attempts to address this question using caspase 9-deficient type II cells were hampered by insufficient induction of cytochrome c release after Fas/CD95 triggering. It is interesting that our observation of low cytochrome c release in JMR cells independently of a Toxoplasma infection indicates that the apoptogenic function of mitochondria during Fas/CD95-triggered apoptosis in type II cells partially relies on the presence of caspase 9. This finding is in accordance with the recent finding that the activity of effector caspases 3 and 7, i.e., those downstream of caspase 9 are critical regulators of proapoptotic mitochondrial events (12, 18). Since such a positive feedback amplification loop has been shown to depend on accelerated caspase 8-mediated Bid cleavage at least during drug-induced apoptosis (33) (Fig. 6), it may also explain the delayed and deficient caspase 8 activity in JMR cells described earlier (28) and in the present study.

During toxoplasmosis in vivo, CD4⁺ and CD8⁺ T lymphocytes have been shown to be important in controlling T. gondii (reviewed in reference 5). IFN-γ is critical for the activation of effective antiparasitic effector mechanisms. In contrast, perforin-dependent cytotoxicity plays only a limited role in combating the parasite (6). Furthermore, death receptors also appear to be dispensable for the control of T. gondii infection, as revealed after infection of TNF-R or Fas/CD95 knockout mice but rather determine immunopathology during acute toxoplasmosis (10, 26, 32). These data are consistent with our finding of a decreased death receptor-mediated apoptosis in Toxoplasma-infected cells, thereby counteracting cytotoxicity as a potential antiparasitic effector mechanism and facilitating parasite survival during toxoplasmosis. Such an evasion strategy may be particularly efficient due to the fact that T. gondii can inhibit Fas/CD95-triggered apoptosis by two distinct mechanisms, i.e., interference with caspase 8 as reported previously (34) and blocking of the mitochondrial amplification loop as shown here. Which of these mechanisms plays the predominant role in the parasite-mediated blockade of cell death largely depends on the type of host cell and its regulation of Fas/CD95-triggered apoptosis. The invention of multiple mechanisms may enable the parasite to inhibit death receptor-mediated apoptosis under different physiological conditions. It may also represent an important obstacle in the development of therapeutic regimens that target the ability of T. gondii to inhibit host cell apoptosis.