Protection against CD95-Induced Apoptosis by Chlamydial Infection at a Mitochondrial Step

Abstract

Chlamydiae are obligate intracellular bacteria that infect human epithelial and myeloid cells. Previous work has established that chlamydiae are able to protect a cell against apoptosis induced by certain experimentally applied stimuli. Here we provide an analysis of this protective activity against the signal transduction during CD95-induced apoptosis. In HeLa cells overexpressing CD95, infection with Chlamydia trachomatis inhibited the appearance of apoptotic morphology, effector caspase activity, the activation of caspase-9 and -3, and the release of cytochrome c from mitochondria. However, caspase-8-processing and activity (measured as cleavage of Bid) were unaffected by the chlamydial infection. Similarly, infection with the species C. pneumoniae did not prevent the activation of caspase-8 but inhibited the appearance of effector caspase activity upon signaling through CD95. Furthermore, infection with C. trachomatis was able to inhibit CD95-induced apoptosis in Jurkat lymphoid cells, where a mitochondrial contribution is required, but not in SKW6.4 lymphoid cells, where caspase-8 directly activates caspase-3. Taken together, these data show that chlamydial infection can protect cells against CD95-induced apoptosis but only where a mitochondrial signaling step is necessary for apoptotic signal transduction.

The genus chlamydia contains three species pathogenic to humans: Chlamydia trachomatis, C. pneumoniae, and C. psittaci. C. trachomatis is the agent responsible for a variety of diseases, especially trachoma (an eye infection prevalent in tropical countries) and a common form of sexually transmitted diseases (8). C. pneumoniae (the classification in a new genus Chlamydophila has been proposed) (4) causes a range of respiratory infections, more commonly upper airway infections but also pneumonia (7). Chlamydiae are obligate intracellular pathogens, i.e., they depend on a host cell to replicate. Upon infection, chlamydiae have been found to interfere with cellular functions in various ways, and one such way is the modulation of the host cell's apoptosis machinery.

Cell death by apoptosis is a frequent event in the human body, and evidence is accumulating that apoptosis plays an important role in the defense against infectious microorganisms. Cell death by apoptosis results from the activation of a specialized signal transduction pathway. On a molecular level, the release of cytochrome c from the mitochondria appears to be a critical signaling event in most cases of apoptosis. The release of cytochrome c is largely controlled by members of proteins of the Bcl-2 family, which can act to promote or to inhibit this release. Free cytosolic cytochrome c initiates the formation of a signaling complex (the so-called apoptosome) that encompasses the molecules Apaf-1 and the proteases caspase-9 and caspase-3. In the formation of this complex, caspase-9 is activated, which in turn activates caspase-3. Active caspase-3 then cleaves cellular substrates to bring about the morphological changes of apoptosis such as nuclear condensation.

An alternative activation of the apoptotic pathway can occur through stimulation of a death receptor. Death receptors are plasma membrane receptors that can, upon binding of the specific ligand (or stimulating antibodies) directly activate the apoptotic pathway. A well-studied death receptor is CD95 (fas/APO-1). Upon stimulation, CD95 recruits a “death-inducing signaling complex” (DISC) to the membrane. The principal signaling components in the DISC are the adapter protein FADD/MORT1 and procaspase-8. During DISC formation, procaspase-8 is activated. A distinction has been proposed to classify cells according to the events that occur following this step. In some cells active caspase-8 is sufficient to activate caspase-3 directly (“type I cells”). In some cells, however, CD95-mediated caspase-8 activation is not enough to activate caspase-3, and in these cells a mitochondrial amplification is necessary: the cleavage of the proapoptotic Bcl-2 family member Bid by caspase-8 causes the release of mitochondrial cytochrome c, thus feeding into the central apoptosis pathway (“type II cells”) (28).

Their dependency on the intact host cell suggested that chlamydiae might interfere with the apoptotic apparatus, and several recent studies have found that this is indeed the case. Both pro- and antiapoptotic activities have been detected: C. trachomatis and C. pneumoniae have been reported to inhibit externally induced apoptosis (5, 6, 26). C. psittaci has been described to induce apoptosis in epithelial cells and macrophages (22) and to inhibit apoptosis against external proapoptotic stimuli (3).

This earlier study has indicated that the main inhibitory quality generated by chlamydiae in infected cells prevents the release of cytochrome c from mitochondria. In the study reported here we analyzed the potential inhibition of a CD95 death signal in infected cells with two issues in mind: first, this pathway has been very well studied; the events that lead to cytochrome c release are at least partly understood and can be studied closely. The distinction of type I and type II cells further provides a tool to analyze and map an inhibitory potential. Second, one important mechanism in the immune system's attack on infected cells is the deployment of cytotoxic T lymphocytes (CTL), and one effector mechanism of CTL is the CD95-mediated induction of apoptosis in infected cells. The sensitivity of an infected cell to a CD95 death signal may therefore bear on the efficiency of the antichlamydial immune response.

MATERIALS AND METHODS

Cell lines and chlamydial organisms.

The human B-lymphocyte cell line SKW6.4, the human T-lymphocyte cell line Jurkat, and the human laryngeal carcinoma cell line Hep2 were obtained from the American Type Culture Collection (ATCC). The human cervical adenocarcinoma cell line HeLa-fas (HeLa cells stably transfected to express human CD95 (30) was kindly provided by H. Wajant, University Stuttgart, Stuttgart, Germany. Cells were cultured at 37°C-5% CO₂ in Dulbecco modified Eagle medium for HeLa-fas and Hep2 cells and in RPMI 1640 for SKW6.4 and Jurkat cells, each supplemented with 10% heat-inactivated fetal calf serum (FCS) and 2 mM l-glutamine. For the propagation of bacteria, cycloheximide (1 μg/ml; Sigma) was added, and culture was done in culture medium without further addition of FCS. No antibiotics were included. The mycoplasma-free strains C. pneumoniae strain CM-1 (VR-1360) and C. trachomatis strain L2 were both obtained from ATCC.

For the propagation of chlamydiae, Hep2 cells were infected in cell culture plates at an MOI of 1 to 3 inclusion forming units (IFU). At 48 h of C. trachomatis infection, and 72 h of C. pneumoniae infection, bacterial organisms were released from the cells by homogenization and purified on a density gradient as described previously (11, 13). Infectious titers were determined by a serial dilution of preparations in HeLa-fas, SKW6.4 and Jurkat cells followed by intracellular staining for chlamydial inclusions with a fluorescence-labeled anti-Chlamydia-lipopolysaccharide (LPS) antibody (Progen, Heidelberg, Germany). Harvests were checked for mycoplasmal contamination by PCR, and purified elementary bodies were frozen in aliquots at −70°C for up to 2 months and thawed immediately before infection.

Infection of cells and induction of apoptosis.

HeLa-fas cells (2.5 × 10⁵ cells/well in six-well plates), SKW6.4 cells, or Jurkat cells (10⁶ cells/well in 12-well plates) were infected with C. pneumoniae or C. trachomatis. The infectious dose used was in the range of 1 to 3 IFU per cell. In none of the experiments where apoptosis was investigated was cycloheximide added. Infection was controlled morphologically, and >95% of cells were found to be infected by using this protocol. Cells in medium without FCS were infected by the addition of C. trachomatis or C. pneumoniae, followed by centrifugation for 45 min at 800 × g at 35°C (centrifugation only for C. pneumoniae); 10% FCS was added after 3 h. Mock-infected cells were subjected to the same procedure in the absence of chlamydiae. For UV inactivation, chlamydial suspensions were exposed for 10 min to UV light in a transilluminator box (Stratagene) as described previously (23). Apoptosis was induced by addition of staurosporine (1 μM; Sigma) or anti-CD95 MAb (CH11, 100 ng/ml; Upstate Biotechnology).

Assay for nuclear apoptosis.

Cells were infected with chlamydiae or mock infected. At 24 h (C. trachomatis) or 48 h (C. pneumoniae) after infection, three replicates each were treated with staurosporine or anti-CD95 for 5 h. Cells were then stained with 20 μM Hoechst 33258 (Sigma) for 30 min at 37°C and washed with phosphate-buffered saline (PBS), and nuclear morphology was assessed under a fluorescence microscope. At least 300 nuclei per sample were counted.

Assay for caspase activity.

HeLa-fas, SKW6.4, or Jurkat cells were infected with chlamydiae, and caspase-3 activity was measured as described previously (6). Briefly, cells were treated with staurosporine or with anti-CD95 for the indicated times. The cells were then lysed in NP-40 lysisbuffer (150 mM NaCl, 1% Ipegal CA-630, 50 mM Tris [pH 8.0]) for 15 min on ice. Cell lysates were cleared by centrifugation for 5 min at 15,000 × g at 4°C. Triplicates of 10-μl aliquots of the supernatant were added to 90 μl of caspase-3-recognition sequence (DEVD) assay buffer (50 mM NaCl, 2 mM MgCl₂, 40 mM β-glycerophosphate, 5 mM EGTA, 0,1% CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}, 100 μg of bovine serum albumin/ml, 10 mM HEPES [pH 7.0]) containing 10 μM (final concentration) DEVD-AMC fluorimetric substrate. Reactions were incubated for 1 h, free AMC was measured, and values are presented as arbitrary relative fluorescence units (mean ± the standard error of the mean of the triplicate reactions described above).

Detection of active caspase-3 by flow cytometry.

HeLa-fas cells were infected with C. trachomatis or left uninfected and treated with anti-CD95 or staurosporine. Cells were harvested, fixed in 0.5% paraformaldehyde for 30 min, permeabilized with 1% saponin, and incubated with anti-active caspase-3 (Pharmingen) and fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit (Dianova) as a secondary antibody. Flow cytometry was performed in a FACScalibur (Becton Dickinson), and at least 10⁶ cells per sample were recorded.

Western blot analysis.

Infected or mock-infected HeLa-fas cells were treated with anti-CD95 or staurosporine, harvested at the times indicated, washed, and lysed in NP-40 lysis buffer. Lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to nitrocellulose membranes, and probed with antibodies specific for human caspase-3 (Becton Dickinson), active caspase-9, caspase-8 (Cell Signaling Technology), Bid (R&D Systems), and β-actin (Sigma). Proteins were visualized by using peroxidase-conjugated secondary antibodies and a chemiluminescence detection system (Perkin-Elmer Lifescience, Boston, Mass.).

Intracellular staining for cytochrome c.

HeLa-fas cells were mock infected or infected with C. trachomatis as described above. At 24 h after infection, replicate wells were treated with staurosporine or anti-CD95 for 5 h and fixed with 2% formalin. For costaining of cytochrome c and mitochondria, cells were stained with anti-cytochrome c MAb (Becton Dickinson)-Cy3-labeled anti-mouse antiserum (Jackson), followed by staining with Mito Tracker Green FM (Molecular Probes). For costaining of cytochrome c and chlamydiae, cells were stained for cytochrome c-FITC-labeled anti-mouse antiserum (Dianova) as described above, followed by the addition of Alexa Fluor 546-labeled mouse anti-chlamydial-LPS antibody (Progen). Pictures were obtained with a Zeiss laser scanning microscope.

Statistical analysis.

Continuous variables are expressed via the mean and the standard deviation (SD). Kruskal-Wallis test was performed for overall comparison of more than two groups. In case of significant differences bivariate post hoc tests were performed by using the method of Schefe to adjust for multiple testing.

RESULTS

Inhibition of CD95-induced nuclear apoptosis by infection with C. trachomatis.

The various intracellular events of apoptotic signaling were investigated in cells that had been infected with chlamydiae. Since epithelial cells are an important host for chlamydial infection, HeLa epitheloid cells were used for most experiments. HeLa cells are, due to a low expression level of CD95, not susceptible to apoptosis induced by CD95 signaling (at least in the absence of enhancing substances). For the present study we therefore used cells stably transfected with the death receptor CD95/Fas/APO-1 (HeLa-fas cells) that readily undergo apoptosis in response to CD95 signaling. This cell line was extensively characterized in its response to the apoptosis-inducing stimuli anti-CD95 and staurosporine. The known steps in the apoptotic signal transduction were indistinguishable between maternal HeLa and HeLa-fas cells in response to staurosporine. The apoptotic response to CD95 was very similar to the one in Jurkat lymphoid cells as also reported for other HeLa lines overexpressing CD95 (30; data not shown).

HeLa-fas cells were infected with C. trachomatis for 24 h, and apoptosis was induced by stimulation with anti-CD95. In most experiments staurosporine was used as a control. We and others have shown before that staurosporine-induced apoptosis is potently inhibited by infection with chlamydiae. Cells treated identically but without chlamydiae were included in parallel in all experiments (mock infected).

When cells were treated with anti-CD95 for 5 h about half of the mock-infected cells showed nuclear morphological changes typical of apoptosis (condensation of DNA, nuclear fragmentation; Fig. 1, arrows show apoptotic cells). Treatment with staurosporine for 5 h induced apoptosis in ca. 90% of the cells (Table 1). In chlamydia-infected cells, the percentage of apoptotic nuclei was much lower than in uninfected cells both when cells were treated with anti-CD95 and when treated with staurosporine. Infection with C. trachomatis therefore reduces the sensitivity of HeLa-fas cells to anti-CD95-induced apoptosis. Infection with C. trachomatis in the absence of further treatment caused apoptosis in a small percentage of cells (Table 1). This is consistent with earlier observations of apoptosis induction by this organism (24, 25) and suggests that the outcome of the infection is determined by the balance of apoptosis-inducing and -inhibiting mechanisms. During the normal course of infection, however, a pronounced resistance to apoptosis induction can be seen.

FIG. 1.

Infection with C. trachomatis inhibits CD95-induced nuclear apoptosis in HeLa-fas cells. HeLa-fas cells were either left uninfected or infected with C. trachomatis at about 2 IFU per cell. At 24 h after infection, anti-CD95 MAb (100 ng/ml) was added to some aliquot cultures. After 5 h, cells were stained with Hoechst dye, and photos of representative areas were taken under a fluorescence microscope. Similar results were obtained in three similar experiments.

TABLE 1.

Quantitation of C. trachomatis-mediated protection against CD95 signaling in HeLa-fas cells^a

Treatment	Mean % cells (SD) with apoptotic nuclear morphology
	Mock infected	C. trachomatis infected
Untreated	1.9 (0.43)	8.4 (1.09)
Anti-CD95	58.7 (3.75)	11.6 (4.05)
Staurosporine	91.5 (3.48)	16.1 (5.13)

^aHeLa-fas cells were either mock infected or infected with C. trachomatis at about 2 IFU per cell. At 24 h after infection, anti-CD95 (100 ng/ml) or staurosporine (1 μM) was added. After 5 h, cells were stained with Hoechst dye, and the nuclear morphology was scored as normal or apoptotic. The percentages of cells with apoptotic nuclear morphology was counted (300 nuclei per sample) and are given as the means of three independent experiments. Inhibition of apoptosis was statistically significant (P < 0.001 for both anti-CD95 and staurosporine in mock-infected and chlamydia-infected cells [Kruskal-Wallis test]).

Infection with C. trachomatis inhibits CD95-induced apoptosis upstream of caspase activation.

Nuclear apoptosis is a result of caspase-3 activity (or other effector caspases, such as caspase-7). Caspase-3 itself is activated either directly by active caspase-8 or upon recruitment into a signaling complex (the apoptosome) containing cytochrome c, Apaf-1, caspase-9, and caspase-3. The activity of caspase-3 that is generated during the induction of apoptosis can be measured with a fluorogenic substrate that contains a DEVD. C. trachomatis-infected and mock-infected HeLa-fas cells were treated with anti-CD95 or with staurosporine for 4 h, and DEVD-specific proteolytic activity was measured. In mock-infected cells a high caspase-3-like activity was detected upon induction of apoptosis with both stimuli. As reported earlier, infection strongly reduced the detectable DEVD-cleaving activity upon staurosporine treatment. The activity generated by anti-CD95 treatment was also clearly reduced (Fig. 2A). As a control, UV-inactivated chlamydia were included. UV-treated bacteria induced only minimal if any protection against CD95-induced effector caspase activation (Fig. 2B).

FIG. 2.

Infection with C. trachomatis blocks CD95-induced apoptosis upstream of caspase activation. (A) HeLa-fas cells were mock infected or infected with C. trachomatis at about 2 IFU/cell. At 48 h after infection, cells were either treated with anti-CD95 (100 ng/ml) or staurosporine (1 μM) for 4 h or left untreated as indicated. Cells were then lysed, and DEVD-cleaving activity was measured in cell extracts. Each bar represents one well of a six-well plate, with the SD as indicated. Inhibition of caspase activity was statistically significant (P < 0.001 for both anti-CD95 and staurosporine [Kruskal-Wallis test]). The data are representative of five experiments. (B) HeLa-fas cells were either mock infected or infected with C. trachomatis or UV-treated C. trachomatis at about 2 IFU/cell. After 48 h of infection, cells were treated with with anti-CD95 (100 ng/ml) for 4 h or left untreated. Cells were lysed and caspase-3-like activity was measured as DEVD-cleaving activity. Each bar represents one well of a six-well plate, with the SD as indicated. (C) HeLa-fas cells were either mock infected or infected with C. trachomatis at about 2 IFU per cell. At 24 h postinfection anti-CD95 (100 ng/ml) or staurosporine (1 μM) was added. After the indicated time intervals cells were extracted, and extracts were analyzed by Western blotting for active caspase-9 (top; only the active enzyme is detected irrespective of the amount of total caspase-9) and β-actin as a loading control. In a similar experiment caspase-3-processing was measured by Western blotting (bottom; the inactive procaspase is 32 kDa [closed arrow], the active caspase-3 is 17 kDa [open arrow]). (D) HeLa-fas cells were infected with either C. trachomatis (2 IFU/cell) or mock infected, and anti-CD95 (100 ng/ml) or staurosporine (1 μM) was added after 24 h. For flow cytometry, cells were collected and stained for active caspase-3 (the antibody used recognizes only the cleaved caspase-3 but not inactive procaspase-3). Normal line, chlamydia-infected cells; bold line, mock-infected cells.

In HeLa-fas cells the activation of caspase-3 by anti-CD95 requires the release of cytochrome c from the mitochondria and the subsequent activation of caspase-9 and -3 (as is indicated by the finding that Bcl-2 expression in HeLa cells protects against CD95-induced cell death) (18). Activation of caspase-9 was measured by Western blotting with an antibody that specifically recognizes only the processed (activated) form of caspase-9. In mock-infected cells treatment with staurosporine, as well as with anti-CD95, led to a prominent activation of caspase-9, which was detectable by using the cleavage specific antibody. Both staurosporine- and CD95-induced activation of caspase-9 were reduced in infected cells (Fig. 2C). Activation of caspase-3 was further investigated by Western blotting and flow cytometry. As found for caspase-9, caspase-3 activation by both staurosporine and CD95 signaling was reduced upon infection with C. trachomatis (Fig. 2C and D).

The release of cytochrome c from mitochondria during CD95-dependent apoptosis is reduced in cells infected with C. trachomatis.

Activation of caspase-9 is normally caused by apoptosome formation upon release of cytochrome c into the cytosol, and the inhibitory potential of chlamydial infection on cytochrome c release was next investigated. Using subcellular fractionation experiments, Fan et al. have found that infection with C. trachomatis blocked CD95-induced cytochrome c release in U937 human myeloid cells (5). We analyzed cytochrome c release during CD95 signaling in HeLa-fas cells by immunostaining and confocal microscopy. HeLa-fas cells were infected and treated with anti-CD95. Confocal microscopy revealed the typical mitochondrial pattern of cytochrome c distribution in both mock-infected and chlamydia-infected cells. Anti-CD95 treatment resulted in morphological changes and the release of cytochrome c, visible as a reduction in staining intensity (26). In infected cells, cytochrome c release was prevented, but it still occurred in uninfected cells from infected cultures (Fig. 3A). Cells were then infected and costained with a mitochondrial marker (Mito Tracker Green FM) and anti-cytochrome c antibody. In untreated cells, both markers showed colocalization in infected and in uninfected cells (Fig. 3B). Upon treatment with anti-CD95 for 5 h, cytochrome c was released into the cytosol in the majority of the mock-infected cells, whereas mitochondria were still stained by Mito Tracker Green FM (Fig. 3B). In contrast, in C. trachomatis-infected cells cytochrome c was largely still found in a mitochondrial pattern in the cells, and a partial colocalization of the two dyes was still seen (Fig. 3). Since it has been shown earlier that chlamydial infection blocks the staurosporine-induced cytochrome c release (5, 6), staurosporine treatment was included as a control; staurosporine-induced cytochrome c release was also prevented in infected cells (data not shown). Infection with chlamydiae therefore blocks the release of cytochrome c from the mitochondria into the cytosol during CD95 signaling.

FIG. 3.

Infection with C. trachomatis inhibits the CD95-dependent release of cytochrome c from mitochondria. HeLa-fas cells were plated onto glass coverslips in a 12-well plate. Cells were either mock infected or infected with C. trachomatis at about 0.5 IFU/cell (A) or 2 IFU/cell (B). At 24 h after infection, cells in some cultures were treated with anti-CD95 (100 ng/ml) for 5 h. Pictures were obtained by confocal laser scanning. (A) Cells were doubly labeled with FITC-anti-cytochrome c (green) and Alexa Fluor 546 antichlamydial antibodies (red). Arrows point to infected cells that have retained cytochrome c in their mitochondria; asterisks indicate uninfected cells where mitochondria have released cytochrome c. (B) Cells were costained with Cy3 anti-cytochrome c (red) and Mito Tracker Green FM (green). In uninfected and anti-CD95-treated cells, cytochrome c becomes almost undetectable upon its release as in panel A. In infected and anti-CD95-treated cells, partial colocalization of cytochrome c and the mitochondrial marker is retained (yellow). The pictures represent three similar experiments.

Infection with C. trachomatis does not affect activation of caspase-8 and cleavage of Bid in CD95-mediated apoptosis.

The initial events upon stimulation of the death receptor CD95 by anti-CD95 include the recruitment of the initiator caspase-8 to the CD95 receptor. During the formation of this complex caspase-8 is processed and activated. Active caspase-8 then cleaves the inactive cytosolic form of the BH3-only protein Bid to generate a truncated fragment (tBid) that translocates to mitochondria and causes the release of cytochrome c (15, 17). In C. trachomatis- and mock-infected HeLa-fas cells caspase-8 processing and Bid cleavage were compared upon anti-CD95 treatment. No difference was apparent between processing of caspase-8 in mock-infected and in C. trachomatis-infected cells (Fig. 4, upper panel). Similarly, CD95-dependent Bid cleavage was not affected by C. trachomatis infection (Fig. 4, middle panel, as assessed as the disappearance of the full-length protein). Note that the staurosporine-dependent cleavage of Bid was blocked in infected cells. In staurosporine-induced apoptosis caspase-9 (not caspase-8) is the first caspase activated (2). The consecutively activated caspase-3 (and caspase-8) that can in this case be activated by caspase-3 (29) will then cleave Bid. Since the activation of caspase-9 by staurosporine is blocked in chlamydia-infected cells (see above), Bid cleavage upon staurosporine treatment is also prevented (Fig. 4, middle panel). These results show that infection with chlamydiae does not affect activation and activity of caspase-8 by CD95 and suggest that apoptosis is blocked at a “mitochondrial” step of the signal transduction.

FIG. 4.

CD95-dependent signal transduction upstream of mitochondria is normal in C. trachomatis-infected cells. HeLa-fas cells were either mock infected or infected with C. trachomatis at about 2 IFU per cell. At 24 h postinfection anti-CD95 (100 ng/ml) or staurosporine (1 μM) was added. After the indicated time intervals, cells were extracted and extracts were analyzed by Western blotting for caspase-8 (top) and uncleaved Bid (middle; the antibody only recognizes the uncleaved form but not the truncated form of Bid). β-Actin served as a loading control (for better comparison, the same blot as in Fig. 2 probed with additional antibodies is shown here). Filled arrows indicate procaspases; open arrows indicate activated forms. This experiment was performed three times with similar results.

CD95 signaling in C. pneumoniae-infected cells.

As shown above, infection with C. trachomatis leads to a blockade of the apoptotic signaling pathway upstream of the release of cytochrome c from the mitochondria, while not affecting the death receptor pathway upstream of mitochondria. Infection with the species C. pneumoniae can also block apoptosis (6, 26). We investigated whether the inhibitory activities of the two species were the same with respect to CD95 signaling. HeLa-fas cells were either infected with C. pneumoniae or mock infected and treated with anti-CD95 for 4 h. Nuclear apoptosis upon CD95 signaling was strongly reduced in infected cells in a way similar to that for the infection with C. trachomatis (data not shown). In parallel, the detectable caspase-3-like activity during CD95-mediated apoptosis was also significantly reduced in C. pneumoniae-infected cells compared to mock-infected cells (Fig. 5A). However, no difference was seen in the processing of caspase-8 upon anti-CD95 signaling between mock-infected and C. pneumoniae-infected cells (Fig. 5B). In contrast, in staurosporine-treated cells the infection with C. pneumoniae could still block caspase-8 activation (which depends on caspase-9/-3 activity [see above]). These results are very similar to the ones obtained in C. trachomatis-infected cells. It thus appears very likely that the same antiapoptotic activity is generated upon infection with either C. trachomatis or C. pneumoniae. This activity does not affect the activation of the death receptor apoptotic pathway but is capable of blocking the events that lead to the release of cytochrome c from the cell's mitochondria.

FIG. 5.

Infection with C. pneumoniae inhibits the generation of effector caspase activity but not CD95-induced activation of caspase-8 in HeLa-fas cells. (A) HeLa-fas cells were mock infected (□) or infected with C. pneumoniae at about 2 IFU/cell (▪). At 72 h after infection, cells were either treated with anti-CD95 (100 ng/ml) or staurosporine (1 μM) for 4 h or left untreated as indicated. Cells were then lysed, and DEVD-cleaving activity was measured in cell extracts. Each bar represents one well of a six-well plate, with the SD as indicated. Inhibition of caspase activity was statistically significant (P < 0.001 for both anti-CD95 and staurosporine in mock-infected and chlamydia-infected cells [Kruskal-Wallis test]). The data are representative of three experiments. (B) HeLa-fas cells were either mock infected or infected with C. pneumoniae at about 2 IFU per cell. At 48 h postinfection anti-CD95 (100 ng/ml) or staurosporine (1 μM) were added. After the indicated time periods cells were extracted, and extracts were analyzed by Western blotting for caspase-8. Filled arrows indicate procaspases; open arrows indicate activated forms. This experiment was performed three times with similar results.

Chlamydial infection fails to inhibit apoptosis in type I cells.

Dependent on the cellular context, CD95 signaling can cause different signal transduction events. In some cells (type II cells), caspase-8 activation by anti-CD95 is not sufficient for immediate activation of downstream caspases but requires the Bid-mediated release of cytochrome c (such as HeLa cells, above; a well-studied example is the human T-cell line Jurkat). In other cells (type I; for instance, the B-cell line SKW6.4) such a mitochondrial involvement is not necessary, and caspase-8 directly activates caspase-3; therefore, a factor that blocks only the mitochondrial path to apoptosis will inhibit the activation of caspase-3 and apoptosis in type II cells but not in type I cells as convincingly shown with Bcl-2 (that acts to inhibit mitochondrial cytochrome c release) (28). Relevance of a protection against a CD95-signal in vivo will depend on the protection afforded in either cell type. We therefore investigated CD95-dependent apoptosis induction in chlamydia-infected SKW6.4 cells. Since the above results have been obtained by using a type II epitheloid cell line and SKW6.4 are lympoid cells, we further examined a lymphoid type II cell line, Jurkat (28). SKW6.4 cells were infected with C. trachomatis at a multiplicity of infection of 5 and treated with anti-CD95. As expected, caspase-8 processing was not affected by chlamydial infection (data not shown). Significantly, chlamydial infection did not block CD95-induced apoptosis or caspase-activation in SKW6.4 cells, although staurosporine-induced apoptosis was inhibited (Fig. 6A and C). In Jurkat cells, the results obtained were similar to the ones in HeLa cells, i.e., infection with C. trachomatis afforded protection against both staurosporine and CD95-induced apoptosis (Fig. 6B and D). A second type I cell line, MCF7 breast carcinoma stably expressing CD95 (12), was also investigated. Since these cells have no caspase-3 protein, they were transfected with an expression vector for human caspase-3, infected with C. trachomatis, and effector caspase activity was measured. As in SKW6.4 cells, infection afforded protection to MCF7 cells against staurosporine but not CD95-induced caspase activation (data not shown). The antiapoptotic activity (or activities) generated during infection with chlamydiae does thus not affect the activation of the death receptor apoptotic pathway but is capable of blocking the events that lead to the release of cytochrome c from the cell's mitochondria. The protection against death receptor signaling afforded by chlamydiae is therefore available to cells in which a mitochondrial contribution is required for the induction of apoptosis but does not extend to cells in which caspase-8 directly activates caspase-3.

FIG. 6.

Infection with C. trachomatis protects Jurkat cells (type II), but not SKW6.4 cells (type I), against CD95-mediated apoptosis. SKW6.4 cells (A and C) or Jurkat cells (B and D) were mock infected (□) or infected with C. trachomatis (▪) at about 5 IFU/cell (by morphology, >95% of cells were found to be infected by using this protocol in both cell lines). At 24 h after infection, cells were either treated with anti-CD95 (100 ng/ml) or staurosporine (1 μM) for 5 h or left untreated. (A and B). Nuclear apoptosis. After 4 h, cells were stained with Hoechst dye, and nuclear morphology was scored as normal or apoptotic as described above. (C and D). Effector caspase activity. Cells were lysed, and caspase activity was measured as the DEVD-cleaving activity in cell extracts. Each bar represents 1 well of a 12-well plate, with the SD as indicated. For Jurkat cells inhibition of caspase activity was statistically significant (P < 0.001 for both anti-CD95 and staurosporine [Kruskal-Wallis test]). For SKW6.4 cells, no reduction was seen in CD95-induced caspase activity but staurosporine-mediated caspase activation was significantly inhibited by chlamydial infection (P < 0.001 [Kruskal-Wallis test]). The data are representative of three experiments.

DISCUSSION

In this study we demonstrate that the infection with chlamydiae has the potential to inhibit CD95-induced apoptosis in some but not all cells. In infected cells, the mitochondrial contribution to the signal transduction of apoptosis appears to be prevented and, accordingly, cells in which such a contribution is necessary for apoptosis, are protected.

Although there is still the possibility that chlamydiae may induce apoptosis later in the infection cycle (22) by a process involving the activation of Bax (25), the ability of chlamydiae principally to inhibit apoptosis has been demonstrated several times. This antiapoptotic ability is seen in various cell types, such as epithelial cells, monocytes, and macrophages (1, 5, 6, 26). The protective capacity has been found to extend to a number of different stimuli: infection inhibited apoptosis induced by staurosporine, tumor necrosis factor alpha, etoposide, granzyme B/perforin, and UV light (5, 6). In most of these cases of apoptosis induction, cytochrome c release from the mitochondria appears to be an important step in signal transduction. However, the steps leading to cytochrome c release are not completely understood but are likely to involve the activation of BH3-only proteins and the proteins Bax and/or Bak (21). The CD95 death receptor pathway is, on the other hand, well characterized, and the individual steps upstream of mitochondria are known in some detail. Formation of the DISC allows for activation of caspase-8 which then cleaves and thereby activates the BH3-only protein Bid. After activation, Bid can activate monomeric Bax and Bak. Bax/Bak translocate from the cytosol to the mitochondrial membrane to form mixed clusters containing thousands of molecules, and Bid-mediated activation of Bax is sufficient to cause the release of cytochrome c (14, 20). Infection by chlamydiae did not affect caspase-8 activation or -activity (measured as the cleavage of Bid) but prevented the release of cytochrome c. The most likely explanation is therefore that chlamydiae block the Bid-induced activation of Bax (the protein expression levels of Bax/Bak do not change during chlamydial infection [unpublished results]).

Since the apoptosis-inducing stimuli that are inhibited by chlamydial infection are also blocked by cellular Bcl-2, it would be easiest to argue that chlamydial infection induces the expression of a Bcl-2-like protein. However, our searches of the genome of C. pneumoniae failed to find such a protein, and the published results of analyses of the cellular transcription in infected cells likewise yielded no indication that a cellular Bcl-2-like protein may be strongly upregulated (10, 31). Another possibility, although not supported by any experimental evidence and perhaps unlikely, is that chlamydiae generate different antiapoptotic activities that counter different stimuli. A further possibility that should be taken into account is the following: we have described earlier that cytosol prepared from cells infected with C. pneumoniae is refractory to cytochrome c, i.e., externally added cytochrome c fails to activate caspases in these preparations (6). Although cytochrome c release may be the first step in activating caspases, a feedback loop is likely to provide amplification by caspase-3-driven mitochondrial desintegration. Although there are some other early reports, a sophisticated study is the recent investigation of granzyme B action (19) that indicates that such amplification is critical for bringing about the apoptotic changes. If, therefore, chlamydiae blocked the caspase activation by the initial release of a small quantity of cytochrome c, the lack of feedback amplification may account for a slower release of remaining cytochrome c, thereby explaining the results we describe here.

We can still only speculate what the precise role is that chlamydial antiapoptotic activities play for the infection (for a discussion, see reference 9). One possible function is a protection against the host cell's (likely) propensity to undergo apoptosis upon infection. Another possibility is the defense against an immune attack. T cells play a role in the clearance of infection with C. pneumoniae in mice (27), and one of the mechanisms by which cytotoxic T cells kill infected target cells is by triggering CD95-dependent apoptosis (16). Whether target cells of chlamydial infection in vivo will be type I or type II (with respect to CD95-induced apoptosis) is unknown. Although the available data indicate that ocular, genital, and respiratory infections by chlamydia are effectively cleared by the body's defense systems, it remains a distinct possibility that the blockade of apoptosis that is afforded by chlamydial infection contributes in a small number of individuals to a resilience to the immune system's efforts to clear the infection.