Cell death, BAX activation, and HMGB1 release during infection with Chlamydia

Microbes and Infection 6 (2004) 1145–1155 www.elsevier.com/locate/micinf

Original article

Cell death, BAX activation, and HMGB1 release during infection with Chlamydia Thomas Jungas a, Philippe Verbeke a, Toni Darville b, David M. Ojcius a,c,* a

b

Institut Jacques Monod, Université Paris 7, 75251 Paris cedex 5, France Department of Microbiology and Immunology, University of Arkansas for Medical Sciences, Little Rock, AR 72205, USA c School of Natural Sciences, University of California, Merced, CA 95344, USA Received 22 July 2004; accepted 29 July 2004 Available online 17 September 2004

Abstract Infection by a number of Chlamydia species leads to resistance of the host cell to apoptosis, followed by induction of host-cell death. In a population of infected cells that displays protection against staurosporine-induced apoptosis among the adherent cells, we find that cells that had been recovered from the supernatant share characteristics of both apoptosis and necrosis, as assayed by the propidium iodide (PI)–annexin V double-labeling technique. Cell death was observed in both an epithelial cell line and primary fibroblasts, although the primary cells had a higher propensity to die through apoptosis than the immortalized cell line. Staurosporine-mediated activation of the pro-apoptotic BCL2 family member, BAX, was inhibited in the epithelial cell line infected for 32 h with the lymphogranuloma venereum (LGV/L2) but not the murine pneumonitis (MoPn) strain of C. trachomatis, but inhibition of staurosporine-mediated BAX activation disappeared after 48 h of infection with the LGV/L2 strain. Conversely, infection with MoPn (C. muridarum) but not LGV/L2 led to BAX activation after 72 h, as previously reported for shorter (48 h) infection with the guinea pig inclusion conjunctivitis (GPIC) serovar of C. psittaci (C. caviae). These results suggest that the ability to inhibit staurosporine-mediated BAX activation or to activate BAX due to the infection itself may vary as a function of the chlamydial strain. Interestingly, both the epithelial cells and the fibroblasts also released high mobility group box 1 protein (HMGB1) during infection, although much less HMGB1 was released from fibroblasts, consistent with the higher level of apoptosis observed in the primary cells. HMGB1 is released preferentially by necrotic or permeabilized viable cells, but not apoptotic cells. In the extracellular space, HMGB1 promotes inflammation through interaction with specific cell-surface receptors. Higher levels of HMGB1 were also measured in the genital-tract secretions of mice infected vaginally with C. trachomatis, compared to uninfected controls. These results suggest that cells infected with Chlamydia release intracellular factors that may contribute to the inflammatory response observed in vivo. © 2004 Elsevier SAS. All rights reserved. Keywords: Chlamydia; Apoptosis; Necrosis; Inflammation

1. Introduction A number of recent reports have shown that Chlamydia infection modulates both survival and death of the host cell. Thus, cells infected with different species and strains of Chlamydia are resistant to death due to treatment with external inducers of apoptosis [1–5]. Resistance to apoptosis may favor persistence of infection [5], a potential factor contributing to long-term complications of Chlamydia infection [6–8]. At the same time, chlamydiae must be able to exit from infected cells at the end of their developmental cycle in order to begin a new round of infection. Numerous studies since * Corresponding author. Tel.: +1-209-724-2948. E-mail address: [email protected] (D.M. Ojcius). 1286-4579/$ - see front matter © 2004 Elsevier SAS. All rights reserved. doi:10.1016/j.micinf.2004.07.004

the 1970s have shown that Chlamydia infection is lytic for the host cell [9–16]. Both apoptotic and necrotic mechanisms are probably at play during host-cell lysis [17,18], with the two types of cell death having very different consequences for the efficiency of infection and the host immune response. Apoptotic cell death may participate in propagation of chlamydial infection, while necrotic cell death may contribute to the inflammatory response [18]. The role played by apoptosis in propagation is suggested by studies on C. trachomatis infection in cells and in mice infected vaginally. Thus, the pro-apoptotic protein, BAX, is activated during Chlamydia infection with the guinea pig inclusion conjunctivitis (GPIC) serovar of C. psittaci [19], and the yield of infection with the mouse pneumonitis (MoPn) strain of C. trachomatis is higher after several cycles

1146

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

of infection in wild-type cells than in Bax-deficient primary fibroblasts [18], suggesting that BAX-dependent cell death may help bacteria to exit from infected cells and begin a new cycle of infection. Similarly, infection with the MoPn strain disappears more quickly from Bax-deficient mice than from wild-type mice [18]. Unexpectedly, Bax-deficient mice demonstrate a higher level of inflammation and increased genital-tract pathology after infection, compared to wild-type mice [18], suggesting that when chlamydial-induced apoptosis is blocked, chlamydial growth in infected cells causes lysis or necrosis of the cell rather than apoptosis. However, even in wild-type primary cells, some infection-related necrosis inevitably occurs [18]. Cellular necrosis allows for release of the cell’s internal contents, which does not happen during apoptosis. The cellular contents act as “danger signals” and induce greater amounts of inflammation [20]. Currently-known danger signals include ATP, glycosylated proteins with exposed mannose residues, uric acid, and the chromosomal protein high mobility group box 1 protein (HMGB1, also known as HMG1 or amphoterin) [20–23], whose extracellular form can be detected by specific receptors on monocytes, macrophages or dendritic cells. Release of any or all of these molecules from infected cells should contribute to the inflammation observed during Chlamydia infection in humans and experimental animal models [24,25]. Cells dying during infection with Chlamydia display some properties of apoptosis, such as DNA fragmentation and nuclear condensation [15,26]. However, cell death during infection does not involve activation of caspases [19], which are required for some but not all pathways of apoptosis [27–30]. As alternative mechanisms of cell death have also been described [30,31], we have further characterized cell death during Chlamydia infection using flow cytometry and propidium iodide (PI)–annexin double-labeling, which measures phosphatidylserine (PS) exposure on apoptotic cells and loss of plasma membrane integrity in predominantly necrotic cells. We find that the type of cell death measured depends in part on the host-cell type used, but that traditional features associated with both apoptosis and necrosis are detected in dying cells. As infection with Chlamydia inhibits apoptosis induced by external inducers [2], we also evaluated the ability of staurosporine to activate BAX during Chlamydia infection, and the ability of the MoPn and lymphogranuloma venereum (LGV/L2) strains of Chlamydia to activate BAX at later time points. We began to address the biological consequences of Chlamydia-induced cell death by measuring release of an intracellular protein during infection. For this study, we have focused on HMGB1, which associates loosely with chromatin in intact cells, where it regulates gene transcription. However, HMGB1 is also released into the extracellular medium when plasma membrane integrity is lost in permeabilized or dying cells [22,32]. Interestingly, apoptotic cells retain HMGB1 more effectively than necrotic cells or viable cells that had been permeabilized with detergent [32]. In the extra-

cellular medium, HMGB1 binds with high affinity to receptor for advanced glycation end products (RAGE), which leads to secretion of proinflammatory cytokines by monocytes [33,34]. According to a recent study, HMGB1 also interacts with the Toll-like receptors (TLR), TLR2 and TLR4, which leads to activation of the nuclear transcription factor, NF-jB, in cells exposed to HMGB1 [35]. Thus, release of this nuclear protein from damaged cells during Chlamydia infection should result in an enhanced inflammatory response. 2. Materials and methods 2.1. Cells and bacterial strains HeLa 229 cells (American Type Culture Collection, ATCC; Manassas, VA) were maintained in a humidified incubator at 37 °C with 5% CO2 in Dulbecco modified Eagle medium Glutamax-1 (Life Technologies, Inc., Rockville, MD) supplemented with 10% heat-inactivated fetal calf serum and 25 µg/ml gentamicin. Murine embryonic fibroblasts (MEF) were obtained from Dr. Stanley Korsmeyer (Harvard Medical School) [18] and cultivated in RPMI-1640 supplemented with 10% heat-inactivated calf serum, 2 mM L-glutamine, and 25 µg/ml gentamicin. The LGV/L2 strain of C. trachomatis was from the ATCC; while the MoPn strain of C. trachomatis, also known as C. muridarum [36,37] was obtained from Roger Rank (University of Arkansas) and was grown and prepared as previously described [38]. The quantity of bacterial inclusion-forming units (IFU) was determined by incubating the chlamydiae with HeLa cells for 1 day and revealing the presence of bacteria by immunofluorescence with anti-Chlamydia genus mAb (1:250 dilution; Argene, France), followed by a fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgG + IgM (1:200 dilution; Argene). Other reagents were previously described [18,39]. 2.2. Infection of cells For infection experiments, HeLa cells growing at 70% confluence on tissue culture flasks (Costar) were infected with the indicated multiplicity of infection (m.o.i.) of the LGV/L2 or MoPn strain of C. trachomatis for 1 h at room temperature. The medium was then removed and replaced by fresh medium to eliminate non-invasive bacteria that had remained in the supernatant. The plates were incubated for various times at 37 °C in a CO2 incubator [18,39]. Cells were collected for flow cytometric analysis or were fixed for fluorescence microcopy, as described below. 2.3. Flow cytometric analysis of cell death Cell death was identified by flow cytometry using two complementary labeling techniques that identify different features of cell death. Early apoptotic changes were identified using FITCconjugated annexin V-fluos (Roche Applied Science, India-

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

1147

napolis, IN), which binds to PS exposed on the outer leaflet of apoptotic cell membranes [40]. PI (Sigma) was used for the discrimination of necrotic cells from the annexin V-positively stained cell cluster. Briefly, HeLa cells were grown on 12-well plates (Falcon, BD Biosciences, San Jose, CA) at a density of 1–2 × 105 cells per well, incubated with various treatments as described above, and dissociated using 0.05% trypsin/0.53 mM EDTA (Gibco BRL, Gaithersburg, MD). The HeLa cell samples were collected, washed in cold phosphate-buffered saline (PBS), and resuspended in 100 µl annexin V-fluos binding solution containing 10 µl annexin V-fluos labeling reagent per 1000 µl Hepes buffer (10 mM Hepes/NaOH, pH 7.4, 140 mM NaCl, 5 mM CaCl2) and 1 µg/ml PI. After 15 min incubation in the dark at room temperature, 400 µl incubation buffer was added to each sample and the cells were analyzed by flow cytometry using 488 nm excitation and a 515 nm band pass filter for fluorescein detection and a 585 nm filter for PI detection. Cells incubated in the binding buffer with only annexin V or PI separately served as controls.

0.1% Triton X-100 in 0.1% sodium citrate solution for 2 min on ice, washed with PBS, and incubated with anti-Chlamydia genus mAb (1:250 dilution; Argene) in PBS containing 0.1% BSA for 30 min at room temperature. Cells were then washed in PBS and incubated with goat anti-mouse IgG + IgM Texas-red-conjugated antibody (1:200 dilution) in PBS containing 0.1% BSA and 10 µM Hoechst 33258 dye (Sigma) for an additional 30 min at room temperature. The coverslips were washed with PBS, mounted onto slides using Dako mounting medium (DAKO), and viewed with either a Zeiss Axiovert 200M fluorescence microscope (Carl Zeiss, Jena, Germany) attached to a cooled charge-coupled device camera, or a LSM 510 Zeiss confocal microscope. Images were obtained with Metamorph software and analyzed with Adobe Photoshop software; Chlamydia inclusions were identified by fluorescence staining. For each experimental condition, cells from 10 random fields were counted and divided by the total number of cells to calculate the percentage of apoptotic cells among cells that contained Chlamydia inclusions or were not infected.

2.4. Flow cytometric analysis of BAX activation in infected and uninfected cells

2.6. Immunoblot and flow cytometry analysis of HMGB1 release

HeLa cells at 70% confluence were infected with C. trachomatis MoPn or C. trachomatis LGV/L2 at an m.o.i. of 1. For induction of apoptosis with an external inducer, cells were treated with a broad-spectrum protein kinase inhibitor, staurosporine (1 µM) (Sigma), overnight after the indicated period of infection with or without the bacteria. All treatments were performed in culture media at 37 °C in the CO2 incubator. Both adherent cells and cells in suspension were collected as described above, washed, and used directly for incubation with antibodies, without the fixation step. Cells were then incubated with polyclonal anti-BAX antibody (BAX N-20; Santa Cruz Biotechnology, Santa Cruz, CA) at 1:70 dilution in PBS containing 1% BSA and 0.1% saponin for 30 min on ice. After washing twice in PBS/BSA/saponin, cells were incubated with PE-conjugated goat anti-rabbit secondary antibody at 1:100 dilution in PBS/BSA/saponin for 30 min on ice. The cells were washed again in PBS/BSA/saponin and then PBS before analyzing by flow cytometry as above. Cells incubated with secondary antibody alone served as controls.

Infected and uninfected HeLa cells were grown in six-well plates (Falcon), washed in ice cold PBS, and lysed in a buffer containing 50 mM Tris–HCl pH 7.2, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, and protease inhibitor cocktail (10 mg/ml) (Sigma). For some samples, HeLa cells were permeabilized with Triton X-100 in order to release all unbound HMGB1 from the cytosol, before lysing cells for immunoblot analysis. The amount of protein in each sample was determined by BCA protein assay (Pierce Biotechnology, Rockford, IL), and equal amounts of protein were subjected to 16% SDS-PAGE gel electrophoresis. Proteins were blotted onto nitrocellulose membrane, blocked in 5% dried milk solution diluted in PBS for 1 h at room temperature with slow agitation. The membrane was then incubated with primary anti-HMG1 antibody (Pharmingen, San Diego, CA) at 1:100 dilution in PBS containing 0.2% Tween-20 for 1 h RT, washed three times for 10 min with PBS containing 0.2% Tween-20 and 500 mM NaCl, treated with HRP conjugated secondary antibody at 1:5000 (Biosource International) for 1 h in 5% dried milk with 0.2% Tween-20, and washed again three times for 10 min. Results were visualized by an ECL detection kit (Amersham Biosciences, Piscataway, NJ). HMGB1 release from cells was also measured by flow cytometry. HeLa cells grown on 12-well plates were treated, infected and collected as described above. The cells were divided into two samples before fixing with 4% PFA: one of the samples was subjected to a permeabilization step using 0.1% Triton X-100 in PBS citrate for 10 min on ice in order to release all unbound HMGB1 from the cytosol before fixation. Permeabilized cells were then washed and incubated with polyclonal anti-HMGB1 antibody at 1:100 dilution 1 h on ice. After washing once in PBS, cells were incubated with

2.5. Detection of cell death by fluorescence microscopy HeLa cells at 70% confluence were infected with an m.o.i. of 0.5–1 of C. trachomatis MoPn or LGV/L2 in 12-well plates (Falcon). For induction of apoptosis studies, cells were treated with 1 µM staurosporine for 6 h after 42 h infection with or without the bacteria. All treatments were performed in culture media at 37 °C in the CO2 incubator. Adherent cells were assayed for all apoptosis measurements by microscopy. Coverslips for each time point were fixed in 4% paraformaldehyde in PBS at room temperature, permeabilized with

1148

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

phycoerythrin-conjugated goat anti-rabbit secondary antibody at 1:200 dilution in PBS/0.1% BSA for 1 h on ice. The cells were washed again in PBS/BSA/saponin and then PBS, before analyzing by flow cytometry as above. Cells incubated with secondary antibody alone served as controls. 2.7. Animal infections and measurement of HMGB1 in genital-tract secretions C57BL/6 (H-2b) mice were purchased from Jackson Laboratories, Bar Harbor, ME. Mice were given food and water ad libitum in an environmentally controlled room with a cycle of 12 h of light and 12 h of darkness. Female mice 7 weeks of age were infected by placing 30 µl of 250 mM sucrose–10 mM sodium phosphate–5 mM L-glutamic acid (SPG) containing 107 IFU of MoPn into the vaginal vault. Infection was performed with the mice under sodium pentobarbital anesthesia. The mice received 2.5 mg of progesterone (Depo-Provera, Upjohn, Kalamazoo, MI) subcutaneously 7 days before vaginal infection to synchronize all mice in a state of anestrus. The course of infection was followed by swabbing the vaginal vault and ectocervix with a Calgiswab (Spectrum Medical Industries, Los Angeles, CA) at various times following infection and by enumerating IFUs by isolation on McCoy cell monolayers [25]. Genital-tract secretions were collected from mice on multiple days prior to and throughout the course of infection and analyzed by Western blot for HMGB1 protein. At intervals before and after infection, an aseptic surgical sponge (2 by 5 mm) (DeRoyal ear wicks; Powell, IN) was inserted into the vagina of an anesthetized animal and retrieved 30 min later. The sponges were held at –70 °C until the day of the assay. Each sponge was placed in a Spin-X microcentrifuge tube (Fisher Scientific) containing a 0.2 µM cellulose acetate filter and incubated in 300 µl of sterile PBS with 0.5% BSA and 0.05% Tween20 for 1 h on ice and then centrifuged for 5 min. Spin-X filters were first preblocked with 0.5 ml of sterile PBS with 2% BSA and 0.05% Tween-20 for 30 min at 25 °C, centrifuged and washed twice with 0.05 ml of sterile PBS. The amount of protein in each sample was determined by BCA protein assay (Pierce Biotechnology), and equal amounts of protein were subjected to 16% SDS-PAGE gel electrophoresis and Western blotting was performed as above for cell lysates. 2.8. Statistics Data are presented as the mean ± standard deviation (S.D.) of “n” experiments. The statistical difference was determined using the paired Student’s t-test. A P value of less than 0.05 was considered statistically significant. 3. Results 3.1. Cell-surface exposure of phosphatidylserine and loss of plasma membrane integrity during infection Previous studies from this laboratory have shown that infection of HeLa (epithelial) cells with the GPIC serovar of

C. psittaci at an m.o.i. of 1.0 or the MoPn serovar of C. trachomatis at an m.o.i. of 6.6 led to fragmentation of DNA, as assayed by gel electrophoresis, and condensation of host-cell nuclei, as assayed by cytofluorimetry using detergentpermeabilized cells labeled with PI [26,41]. For infection with either strain, there was no significant nuclear condensation before about 2 days post-infection. In order to assay cell death with a complementary assay, we have now infected HeLa cells with either the MoPn or LGV/L2 serovar of C. trachomatis at an m.o.i. of 1.0, collected both adherent cells and cells in suspension, and measured cell death by double-labeling cells with annexin V and PI. At the lower concentration of PI used in this assay than in the previous assay with detergent-permeabilized cells, PI labels only cells that have lost their plasma membrane integrity, while annexin V binds to PS exposed on the surface of apoptotic and/or necrotic cells [40]. As previously observed, very little cell death was measured within the first 2 days of infection, while approximately 40% were dying after a 56 h infection with MoPn. A lower but reproducible level of cell death was observed after a 56 h infection with LGV/L2 (Fig. 1). Almost all of the cells were also labeled only with PI, suggesting a necrotic-type of death. A heterogeneous family of clostridial cytotoxin homologs encoded in the genomes of C. trachomatis and C. psittaci that may be responsible for the necrotic features has in fact been described [17], and overexpression of BAX has also been shown to lead to an apoptotic-like death accompanied by loss of plasma membrane permeability [31]. It should be noted that little host-cell death among infected cells was observed in an immunofluorescnce study that limited itself to analysis of infected cells that were still adherent [42]. Using the microscopy criteria, the level of host-cell death for some chlamydial strains was lower after 72 h of infection than at 48 h of infection, consistent with the possibility that dying cells were being lost in the supernatant. 3.2. Protection against apoptosis in adherent cells infected with C. trachomatis Several studies have demonstrated that cells infected with a number of different strains of C. trachomatis are resistant to apoptosis triggered by external inducers such as staurosporine and TNFa [1,2,5,42]. During both acute and persistent infection, resistance to externally-induced apoptosis is due to inhibition of cytochrome c release from mitochondria, and consequently inhibition of caspase-3 activation [2,5]. Infected cells are not resistant to apoptosis in cells that do not require mitochondria for caspase-3 activation [1]. To further investigate the mechanism of protection, we have infected HeLa cells with the LGV/L2 serovar, and measured nuclear condensation by immunofluorescence microscopy of adherent cells that had been treated with staurosporine. While staurosporine could induce nuclear condensation in uninfected cells, infected cells rarely if ever displayed

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

1149

Fig. 1. Cell death during infection of HeLa cells with the MoPn or LGV/L2 serovars of C. trachomatis, measured by cytofluorimetry with annexin–PI double-labeled cells. (a) Representative dot plot of uninfected HeLa cells. (b) Dot plot of cells that had been infected with the MoPn serovar for 56 h. The annexin-positive population corresponds to apoptotic cells, the PI-positive population corresponds to necrotic cells, and the double-positive positions corresponds to necrotic and/or late apoptotic cells. (c) Cell death as a function of time during infection with the MoPn serovar. White bars, viable cells; gray bars, apoptotic cells; black bars, necrotic cells; striped bars, necrotic and/or late apoptotic cells. (d) Cell death as a function of time during infection with the LGV/L2 serovar. Results are given as averages and S.D.

Fig. 2. Protection of infected cells against staurosporine-mediated apoptosis (a, upper left-hand panel). Uninfected cells labeled with the nuclear stain, Hoechst (a, upper right-hand panel). Uninfected cells treated with staurosporine, showing condensed (apoptotic) nuclei stained with Hoechst (apoptotic nuclei indicated with arrowheads) (a, lower left-hand panel). Cells that had been infected with the LGV/L2 serovar were treated with staurosporine. Condensed nuclei were found mainly in cells that did not have a Chlamydia inclusion (inclusions indicated with stars) (a, lower right-hand panel). Cells that had been infected with LGV/L2 but had not been treated with staurosporine, showing a lower cell density than in the uninfected controls (a, upper right-hand panel). (b) Immunofluorescence quantification of cells with segmenting (apoptotic) nuclei, measured by Hoechst staining. There was little nuclear segmentation in the absence of staurosporine treatment (black bar), compared to uninfected cells that were treated with staurosporine (gray bar). In a population of adherent cells that contained both infected and uninfected cells, staurosporine caused a lower level of nuclear segmentation (white bar). In the same cell population, there were essentially no apoptotic nuclei among cells that contained a Chlamydia inclusion (striped bar).

nuclear condensation (Fig. 2a). However, we were not able to demonstrate protection against staurosporine-mediated apoptosis in cells infected with the MoPn serovar (data not

shown). In another study, the MoPn serovar was shown to be less efficient than LGV/L2 in protecting infected cells against staurosporine-induced apoptosis [42]. We thus re-

1150

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

stricted ourselves to the LGV/L2 serovar for subsequent experiments on protection. Resistance to apoptosis could also be observed in cell populations that contained both infected and uninfected cells. As previously reported [2,42], uninfected cells in the total population were still sensitive to staurosporinemediated apoptosis, while cells containing Chlamydia inclusions were resistant to apoptosis (Fig. 2b). Nonetheless, we consistently observed that the density of adherent cells remaining after longer periods of infection was lower than in uninfected controls (Fig. 2a), suggesting that, despite the resistance to staurosporine-mediated apoptosis, infected cells were also detaching from the microscope coverslips. 3.3. BAX activation during infection with two strains of C. trachomatis We have previously reported that a 48-h infection with the GPIC serovar leads to activation of BAX in HeLa cells [19]. Given the role that Bax-deficiency plays during vaginal infection of mice with the MoPn serovar [18], we verified whether MoPn activates BAX in infected HeLa cells. MoPn also activated BAX (Fig. 3), to a lower extent than GPIC but in line with the role played by BAX during MoPn infection in vivo. Nonetheless, we failed to observe BAX activation due to infection with LGV/L2 for up to 72 h (not shown). Since HeLa cells infected with LGV/L2 are resistant to apoptosis induced by staurosporine, we investigated whether

Fig. 4. Inhibition of staurosporine-mediated BAX activation in cells infected with the LGV/L2 serovar. (a) Treatment of uninfected cells with staurosporine leads to a large increase in BAX activation (gray line), but there is no BAX activation in HeLa cells that had been infected with LGV/L2 for 32 h (black line). (b) Staurosporine treatment leads to a large increase in BAX activation (gray line), and the same level of staurosporine-induced BAX activation is found in HeLa cells that had been infected with LGV/L2 for 56 h (black line). The dotted gray line with internal vertical lines represents the fluorescence of uninfected cells in the absence of first (anti-BAX) antibody.

BAX was still activated during the time period when protection against apoptosis was observed. Measuring BAX activation by cytofluorimetry with an antibody that recognizes an epitope of BAX that becomes exposed during activation, a large level of staurosporine-mediated activation was observed in uninfected cells (Fig. 4). Conversely, there was a complete inhibition of staurosporine-dependent BAX activation in cells that had been infected with LGV/L2 for 32 h. But significantly, inhibition of staurosporine-mediated BAX activation disappeared after 56 h of infection (Fig. 4). 3.4. HMGB1 release from epithelial cells infected with C. trachomatis

Fig. 3. Activation of BAX during infection with the MoPn serovar. (a) There is no BAX activation in HeLa cells infected for 32 h (black line), compared to uninfected controls (gray line). (b) There is significant BAX activation in HeLa cells infected for 72 h (black line), compared to uninfected controls (gray line). The dotted gray line with internal vertical lines represents the fluorescence of uninfected cells in the absence of first (anti-BAX) antibody.

It has been reported that a cytosolic enzyme, lactate dehydrogenase (LDH), is released from cells infected with Chlamydia [43], making it more than likely that smaller molecules, such as ATP, and molecules of a similar size as LDH, such as HMGB1, may also be released. To verify whether HMGB1 is in fact released, HeLa cells were infected for 24, 48 or 72 h with MoPn, and the amount of HMGB1 still associated with the host cells was compared by Western blot analysis, using an antibody that recognizes HMGB1. As seen

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

1151

Fig. 5. HMGB1 release from HeLa cells infected with Chlamydia. (a) Western blot showing a decrease in the quantity of cell-associated HMGB1 in cells that had been infected with the MoPn or LGV/L2 serovars for the indicated time (in hours). Western blots were probed with a polyclonal anti-HMGB1 antibody and identified as a 29 kDa band; 0, control uninfected cells; T, uninfected cells that had been permeabilized with Triton before being transferred to the lysis buffer. (b) Plots showing cell-associated HMGB1 in cells analyzed by cytofluorimetry. A decrease in the fluorescence signal corresponds to loss of cell-associated HMGB1. HeLa cells were infected with the MoPn (top panel) or LGV/L2 (bottom panel) serovars for 72 h. Gray line, uninfected cells; dotted line, infected cells; and black line, cells that were pre-permeabilized with Triton to release most intracellular HMGB1. (c) Time course of HMGB1 release from HeLa cells infected with the MoPn (white bar) or LGV/L2 (black bar) serovars.

in Fig. 5a, epithelial cells infected with MoPn or LGV/L2 lose a significant amount of HMGB1 into the supernatant, implying that HMGB1 is secreted or, more likely, released from dying cells. In order to quantify the amount of HMGB1 that was still associated with host cells, infected and uninfected cells were first fixed and then permeabilized with detergent, and incubated with the anti-HMGB1 antibody and secondary fluorescently-labeled antibodies (Fig. 5b). As a control for complete release of HMGB1, uninfected cells were first permeabilized with detergent before the fixation step. By cytofluorimetry, a significant percentage of HeLa cells infected with MoPn or LGV/L2 had lost their HMGB1 (Fig. 5c), in qualitative agreement with the percentage of cells that had lost their plasma membrane integrity, measured by PI staining (Fig. 1). These results suggest that a significant percentage of HeLa cells are necrotic and/or late apoptotic. 3.5. Cell death and HMGB1 release during infection of primary fibroblasts A higher level of apoptosis and lower level of necrosis was previously observed in primary fibroblasts infected with MoPn for 48 h [18], as measured by the PI–annexin V double-labeling assay. We, therefore, infected primary fibroblasts with MoPn or LGV/L2 as a function or time, and measured cell death by double-staining with PI and annexin V. Significant levels of cell death were measured within 48 h of infection with either chlamydial strain, but for both MoPn and LGV/L2, most of the dying cells were stained only with annexin V (Fig. 6a, b), implying apoptotic cell death. These results suggest that the type of cell death features measured

during infection with Chlamydia may depend strongly on the type of host cell that is infected. Consistent with the predominantly apoptotic nature of cell death, very little HMGB1 was released from primary fibroblasts infected with either strain of Chlamydia (Fig. 6c). 3.6. HMGB1 secretion in the murine genital tract during infection with C. trachomatis Genital-tract swabs of female mice that had been uninfected (day –2 and –1) or infected for 2, 4 or 7 days were eluted and analyzed by Western blotting, using the polyclonal antibody against HMGB1. There was essentially no HMGB1 release in uninfected mice and a low level of release in mice after 2 days of infection (Fig. 7). A large level of release was observed after 4 days of infection, which subsided by 7 days of infection. However, as the level of HMGB1 release coincides with the intensity of the inflammatory response (not shown; and [18,25]), it is not possible with these data to determine whether HMGB1 release is due to secretion of inflammatory cytokines or whether HMGB1 release contributes to the inflammatory response.

4. Discussion The present experiments demonstrate that an epithelial cell line (HeLa) infected with the LGV/L2 or MoPn strains of C. trachomatis dies through a process that displays features of necrosis, as shown by lack of annexin V labeling, indicative of little PS exposure on infected cells, but loss of plasma membrane integrity and release of HMGB1. This is in con-

1152

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

Fig. 6. Cell death and HMGB1 release from primary fibroblasts infected with Chlamydia. (a) Cell death as a function of time during infection with the MoPn serovar. White bars, viable cells; gray bars, apoptotic cells; black bars, necrotic cells; striped bars, necrotic and/or late apoptotic cells. Cell death was measured by cytofluorimetry with annexin–PI double-labeled cells. (b) Cell death as a function of time during infection with the LGV/L2 serovar. Results are given as averages and S.D. (c) Western blot showing a slight decrease or no decrease in the quantity of cell-associated HMGB1 in fibroblasts that had been infected with the MoPn or LGV/L2 serovars, respectively, for the indicated time (in hours).

Fig. 7. HMGB1 release into the genital-tract secretions of female mice after chlamydial infection. Western blot analysis of HMGB1 genita-tract secretions from pre- (day –2, –1) and post- (day 2, 4, 7) vaginal inoculation of mice with the MoPn serovar. Secretions were probed with antiHMGB1 antibody by Western blot. Representative results from two groups of mice infected on separate dates.

trast with strong PS exposure found on HeLa cells infected with C. caviae (the GPIC strain of C. psittaci), as assayed by cytofluorimetry of annexin V-labeled cells [44]. Similar results were obtained with C. caviae-infected cells when sodium binding to PS was measured noninvasively by nuclear magnetic resonance (NMR) [44], suggesting that annexin V-labeling of total cell populations is a reliable measure of the extent of cell death. Nonetheless, a higher level of apoptosis was measured when primary fibroblasts were infected with C. trachomatis. Hence, the balance between apoptotic and necrotic cells may be affected by both the host-cell type and the chlamydial strain. In this respect, it is worth noting that the main cell type used for most cell death studies until now, the HeLa epithelial cell line, was originally derived from a cervical carcinoma and expresses constitutively the E6 and E7 oncogenes from the human papilloma virus [45]. Among other activities, the E6 and E7 oncogene products regulate the cell cycle, stimulate telomerase activity, and inhibit apoptosis [45–47]. Although lung fibroblasts are not physiological targets for Chlamydia infection, these results suggest that primary epithelial cells from the genital tract or from the respiratory tract should be used in future studies to

determine the extent of cell death or resistance to death during infection with C. trachomatis or C. pneumoniae. For both HeLa cells and primary fibroblasts, HMGB1 was released into the extracellular medium following infection with C. trachomatis, but much less HMGB1 was released from fibroblasts, consistent with the observation that these cells die preferentially through an apoptotic-type mechanism. Until recently, the HMGB1 protein has been studied mainly in the context of its association with nucleosomes, where it facilitates the assembly of DNA binding proteins to specific binding sites [48]. HMGB1 is expressed in all nucleated cells and is only 10 times less abundant than core histones. However, unlike histones, HMGB1 is not stably associated with chromatin but is also present in the extracellular space, where it plays an important role in inflammation and tumor metastasis, due to its ability to bind with high affinity to the receptor RAGE [33,49]. Significantly, HMGB1 binds to chromatin in apoptotic cells with a higher affinity than to chromatin in necrotic or viable cells [32]. HMGB1 can also be secreted from a limited number of professional inflammatory cells such as monocytes and macrophages stimulated by LPS, TNFa or IL-1b [33,34]. Along similar lines, injection of HMGB1 into mice causes toxic shock; and conversely, injection of anti-HMGB1 antibodies confers significant protection against the lethality of intratracheal or i.p. endotoxin [33,50]. Finally, septic patients show high levels of HMGB1 in serum, which correlate with the severity of infection [22]. These results suggest that HMGB1 may be secreted from sites of inflammation during infection or released passively from infected cells that are undergoing cell death. However, release of HMGB1 from an infected cell in vitro has not been described until now. We find that HMGB1 is released from HeLa cells with a time course similar to cell death, measured by PI–annexin V

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

double-labeling. Thus, HMGB1 release correlates with cell death rather than HMGB1 secretion, and measurable quantities of HMGB1 are released independently of the predominant mode of cell death used by the host cell. Interestingly, HMGB1 was also detected in the genital secretions of mice that had been infected with C. trachomatis vaginally, suggesting that regardless of the mechanism, HMGB1 released from sites of infection or inflammation in vivo could amplify the inflammatory response. Cells dying during C. trachomatis or C. caviae infection in vivo have been detected in the lumen of the female genital tract of mice and guinea pigs [41,51]. In the absence of rapid phagocytosis by neighboring cells, it is possible that infected cells that are shed into the lumen in vivo are able to advance from apoptosis to death with features of late apoptosis or necrosis. Alternatively, HMGB1 may be released from the genital tract due to the inflammatory response, regardless of the effects of Chlamydia on host-cell death. It is safe to assume that other danger signals are also released from cells during Chlamydia infection. Several studies have shown that a cytosolic enzyme, LDH, is released from cells infected with Chlamydia (reviewed in [43]), making it more than likely that smaller molecules such as ATP also make their way to the extracellular medium. But release of danger signals is not restricted to Chlamydia infection, since macrophages infected with Mycobacterium tuberculosis also release ATP in a dose-dependent manner [52]. Some of these molecules are already known to stimulate inflammation, while others may play additional roles. In this respect, it is worthwhile noting that extracellular ATP helps to control infections through several mechanisms. Extracellular ATP binds to purinergic P2X7 receptors, which are expressed on macrophages and dendritic cells [53–55]. Engagement of the receptor leads to maturation and secretion of IL-1b [56,57], and thus contributes to initiation of the inflammatory response. However, extracellular ATP also induces death of intracellular mycobacteria and chlamydiae in infected macrophages [39,58–63]. ATP stimulation of the P2X7 receptor results in an increase in the activity of phospholipase D (PLD), an enzyme that has been previously linked to leukocyte antimicrobial mechanisms, including phagocytosis and generation of reactive oxygen species. And PLD activation appears to be directly responsible for killing of intracellular mycobacteria and chlamydiae, since PLD inhibitors decrease the level of killing of the intracellular pathogens [39,58,59]. In the case of HMGB1, its release from the nucleus leads to binding to RAGE, a receptor that belongs to the immunoglobulin superfamily and is expressed on a wide set of cells, including endothelial cells, smooth muscle cells and neurons [22]. Ligation of the receptor on monocytes leads to cytokine secretion, but in other cell types, HMGB1 binding to RAGE activates a number of different kinases in different cell types, including the MAP kinases p38 MAPK, JNK and p42/p44 MAPK, and translocation of phosphorylated ERK1/2 into the nucleus [49,64]. How these pathways are

1153

related to each other still remains to be elucidated, but ERK1/2 activation is now known to be required for optimal development of the Chlamydia vacuole [65]. It is thus tempting to propose that RAGE ligation, besides initiating an inflammatory response, could also influence the developmental cycle of Chlamydia in certain host-cell types.

Acknowledgements This work was supported by National Institutes of Health grant R01 AI054624. We thank Marie-Claude Gendron for assistance with cytofluorimetry.

References [1]

[2]

[3]

[4]

[5] [6]

[7]

[8] [9]

[10]

[11]

[12]

[13] [14]

[15]

S.F. Fischer, T. Harlander, J. Vier, G. Hacker, Protection against CD95-induced apoptosis by chlamydial infection at a mitochondrial step, Infect. Immun. 72 (2004) 1107–1115. T. Fan, H. Lu, L. Shi, G.A. McCarthy, D.M. Nance, A.H. Greenberg, et al., Inhibition of apoptosis in Chlamydia-infected cells: blockade of mitochondrial cytochrome c release and caspase activation, J. Exp. Med. 187 (1998) 487–496. K. Rajalingam, H. Al-Younes, A. Muller, T.F. Meyer, A.J. Szczepek, T. Rudel, Epithelial cells infected with Chlamydophila pneumoniae (Chlamydia pneumoniae) are resistant to apoptosis, Infect. Immun. 69 (2001) 7880–7888. R. Coutinho-Silva, J.L. Perfettini, P.M. Persechini, A. Dautry-Varsat, D.M. Ojcius, Modulation of P2Z/P2X7 receptor activity in macrophages infected with Chlamydia psittaci, Am. J. Physiol. 280 (2001) C81–C89. D. Dean, V.C. Powers, Persistent Chlamydia trachomatis infections resist apoptotic stimuli, Infect. Immun. 69 (2001) 2442–2447. W.L. Beatty, R.P. Morrison, G.I. Byrne, Persistent chlamydiae: from cell culture to a paradigm for chlamydial pathogenesis, Microbiol. Rev. 58 (1994) 686–699. D. Dean, R. Suchland, W. Stamm, Evidence for long-term cervical persistence of Chlamydia trachomatis by omp1 genotyping, J. Infect. Dis. 182 (2000) 909–916. R.P. Morrison, New insights into a persistent problem—chlamydial infections, J. Clin. Invest. 111 (2003) 1647–1649. R.R. Friis, Interaction of L cells and Chlamydia psittaci: entry of the parasite and host responses to its development, J. Bacteriol. 180 (1972) 706–721. J.W. Moulder, T.P. Hatch, G.I. Byrne, K.R. Kellogg, Immediate toxicity of high multiplicities of Chlamydia psittaci for mouse fibroblasts (L cells), Infect. Immun. 14 (1976) 277. W.J. Todd, J. Storz, Ultrastructural cytochemical evidence for the activation of lysosomes in the cytocidal effect of Chlamydia psittaci, Infect. Immun. 12 (1975) 638–646. J. Taverne, W.A. Blyth, R.C. Ballard, Interactions of TRIC agents with macrophages: effects on lysosomal enzymes of the cell, J. Hyg. (Lond.) 72 (1974) 297. G.T. Chang, J.W. Moulder, Loss of inorganic ions from host cells infected with Chlamydia psittaci, Infect. Immun. 19 (1978) 827–832. P.B. Wyrick, E.A. Brownridge, B.E. Ivins, Interaction of Chlamydia psittaci with mouse peritoneal macrophages, Infect. Immun. 19 (1978) 1061–1067. D. Gibellini, R. Panaya, F. Rumpianesi, Induction of apoptosis by Chlamydia psittaci and Chlamydia trachomatis infection in tissue culture cells, Zentralbl. Bakteriol. 288 (1998) 35–43.

1154

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155

[16] A.J. McCoy, R.C. Sandlin, A.T. Maurelli, In vitro and in vivo functional activity of Chlamydia MurA, a UDP-N-acetylglucosamine enolpyruvyl transferase involved in peptidoglycan synthesis and fosfomycin resistance, J. Bacteriol. 185 (2003) 1218–1228. [17] R.J. Belland, M.A. Scidmore, D.D. Crane, D.M. Hogan, W. Whitmire, G. McClarty, et al., Chlamydia trachomatis cytotoxicity associated with complete and partial cytotoxin genes, Proc. Natl. Acad. Sci. USA 98 (2001) 13984–13989. [18] J.L. Perfettini, D.M. Ojcius, C.W. Andrews, S.J. Korsmeyer, R.G. Rank, T. Darville, Role of proapoptotic BAX in propagation of Chlamydia muridarum (the mouse pneumonitis strain of Chlamydia trachomatis) and the host inflammatory response, J. Biol. Chem. 278 (2003) 9496–9502. [19] J.L. Perfettini, J.C. Reed, N. Israël, J.C. Martinou, A. Dautry-Varsat, D.M. Ojcius, Role of BCL-2 family members in caspase-independent apoptosis during Chlamydia infection, Infect. Immun. 70 (2002) 55–61. [20] P. Matzinger, The danger model: a renewed sense of self, Science 296 (2002) 301–305. [21] S. Gallucci, P. Matzinger, Danger signals: SOS to the immune system, Curr. Opin. Immunol. 13 (2001) 114–119. [22] S. Müller, P. Scaffidi, B. Degryse, T. Bonaldi, L. Ronfani, A. Agresti, et al., The double life of HMGB1 chromatin protein: architectural factor and extracellular signal, EMBO J. 16 (2001) 4337–4340. [23] Y. Shi, J.E. Evans, K.L. Rock, Molecular identification of a danger signal that alerts the immune system to dying cells, Nature 425 (2003) 516–521. [24] P.M. Bavoil, R.-c. Hsia, R.G. Rank, Prospects for a vaccine against Chlamydia genital disease. I. Microbiology and pathogenesis, Bull. Inst. Pasteur 94 (1996) 5–54. [25] T. Darville, J.M. O’Neill, C.W. Andrews Jr., U.M. Nagarajan, L. Stahl, D.M. Ojcius, Toll-like receptor-2, but not Toll-like receptor-4, is essential for development of oviduct pathology in chlamydial genitaltract infection, J. Immunol. 171 (2003) 6187–6197. [26] D.M. Ojcius, P. Souque, J.L. Perfettini, A. Dautry-Varsat, Apoptosis of epithelial cells and macrophages due to infection with the obligate intracellular pathogen Chlamydia psittaci, J. Immunol. 161 (1998) 4220–4226. [27] N.K. Danial, S.J. Korsmeyer, Cell death: critical control points, Cell 116 (2004) 205–219. [28] M. Jäättelä, J. Tschopp, Caspase-independent cell death in T lymphocytes, Nat. Immunol. 4 (2003) 416–423. [29] J. Lieberman, The ABCs of granule-mediated cytotoxicity: new weapons in the arsenal, Nat. Rev. Immunol. 3 (2003) 361–370. [30] M. Leist, M. Jäättelä, Four deaths and a funeral: from caspases to alternative mechanisms, Nat. Rev. Mol. Cell Biol. 2 (2001) 589–598. [31] J. Xiang, D.T. Chao, S.J. Korsmeyer, BAX-induced cell death may not require interleukin 1b-converting enzyme-like proteases, Proc. Natl. Acad. Sci. USA 93 (1996) 14559–14563. [32] P. Scaffidi, T. Misteli, M.E. Bianchi, Release of chromatin protein HMGB1 by necrotic cells triggers inflammation, Nature 418 (2002) 191–195. [33] H. Wang, O. Bloom, M. Zhang, J.M. Vishnubhakat, M. Ombrellino, J. Che, et al., HMG-1 as a late mediator of endotoxin lethality in mice, Science 285 (1999) 248–251. [34] U. Andersson, H. Wang, K. Palmblad, A.C. Aveberger, O. Bloom, H. Erlandsson-Harris, et al., High mobility group 1 protein (HMG-1) stimulates proinflammatory cytokine synthesis in human monocytes, J. Exp. Med. 192 (2000) 565–570. [35] J.S. Park, D. Svetkauskaite, Q. He, J.-Y. Kim, D. Strassheim, A. Ishizaka, et al., Involvement of Toll-like receptors 2 and 4 in cellular activation by high obility group box 1 protein, J. Biol. Chem. 279 (2004) 7370–7377.

[36] K.D.E. Everett, R.M. Bush, A.A. Andersen, Emended description of the order Chlamydiales, proposal of Parachlamydiaceae fam. nov. and Simkaniaceae fam. nov., each containing one monotypic genus, revised taxonomy of the family Chlamydiaceae, including a new genus and five new species, and standards for the identification of organisms, Int. J. Syst. Bacteriol. 49 (1999) 415–440. [37] J. Schachter, R.S. Stephens, P. Timms, C. Kuo, P.M. Bavoil, S. Birkelund, et al., Radical changes to chlamydial taxonomy are not necessary just yet, Int. J. Syst. Evol. Microbiol. 51 (2001) 249. [38] K.H. Ramsey, W.J. Newhall, R.G. Rank, Humoral immune response to chlamydial genital infection of mice with the agent of mouse pneumonitis, Infect. Immun. 57 (1989) 2441–2446. [39] R. Coutinho-Silva, L. Stahl, M.N. Raymond, T. Jungas, P. Verbeke, G. Burnstock, T. Darville, D.M. Ojcius, Inhibition of chlamydial infectious activity due to P2X7 receptor-dependent phospholipase D activation, Immunity 19 (2003) 403–412. [40] J.-L. Perfettini, M. Gissot, P. Souque, D.M. Ojcius, Modulation of apoptosis during infection by Chlamydia, Methods Enzymol. 358 (2002) 334–344. [41] J.-L. Perfettini, T. Darville, G. Gachelin, P. Souque, M. Huerre, A. Dautry-Varsat, et al., Effect of Chlamydia trachomatis infection and subsequent TNFa secretion on apoptosis in the murine genital tract, Infect. Immun. 68 (2000) 2237–2244. [42] W. Greene, Y. Xiao, Y. Huang, G. McClarty, G. Zhong, Chlamydiainfected cells continue to undergo mitosis and resist induction of apoptosis, Infect. Immun. 72 (2004) 451–460. [43] J.W. Moulder, Interaction of chlamydiae and host cells in vitro, Microbiol. Rev. 55 (1991) 143–190. [44] M. Lupu, D. Ojcius, J.L. Perfettini, J. Patry, J.L. Dimicoli, J. Mispelter, Changes of sodium ion compartmentalization in biological systems due to pathological states. A noninvasive NMR study, Biochimie 85 (2003) 849–861. [45] R.A. DeFilippis, E.C. Goodwin, L. Wu, D. DiMaio, Endogenous human papillomavirus E6 and E7 proteins differentially regulate proliferation, senescence, and apoptosis in HeLa cervical carcinoma cells, J. Virol. 77 (2003) 1551–1563. [46] H. Zur Hausen, Papillomaviruses and cancer: from basic studies to clinical application, Nat. Rev. Cancer 2 (2002) 342–350. [47] K. Butz, T. Ristriani, A. Hengstermann, C. Denk, M. Scheffner, F. Hoppe-Seyler, siRNA targeting of the viral E6 oncogene efficiently kills human papillomavirus-positive cancer cells, Oncogene 22 (2003) 5938–5945. [48] S. Calogero, F. Grassi, A. Aguzzi, T. Voigtländer, P. Ferrier, S. Ferrari, et al., The lack of chromosomal protein HMG1 does not disrupt cell growth, but causes lethal hypoglycaemia in newborn mice, Nat. Genet. 22 (1999) 276–280. [49] A. Taguchi, D.C. Blood, G. Del Toro, A. Canet, D.C. Lee, W. Qu, et al., Blockade of RAGE-amphoterin signalling suppresses tumour growth and metastases, Nature 405 (2000) 354–360. [50] H. Yang, M. Ochani, J. Li, X. Qiang, M. Tanovic, H.E. Harris, et al., Reversing established sepsis with antagonists of endogenous highmobility group box 1, Proc. Natl. Acad. Sci. USA 101 (2004) 296– 301. [51] B.L. Soloff, R.G. Rank, A.L. Barron, Electronmicroscopic observations concerning the in vivo uptake and release of the agent of guinea pig inclusion conjunctivitis (C. psittaci) in guinea pig exocervix, J. Comp. Pathol. 95 (1985) 335–344. [52] A. Sikora, J. Liu, C. Brosnan, G. Buell, I. Chessel, B.R. Bloom, Cutting edge: purinergic signaling regulates radical-mediated bacterial killing mechanisms in macrophages through a P2X7-independent mechanism, J. Immunol. 163 (1999) 558–561. [53] R. Coutinho-Silva, P.M. Persechini, R.C. Bisaggio, J.L. Perfettini, A.C.T. Sa Neto, J.M. Kanellopoulos, et al., P2Z–P2X7 receptormediated apoptosis of dendritic cells, Am. J. Physiol. 276 (1999) C1139–C1147. [54] F. Di Virgilio, P. Chiozzi, D. Ferrari, S. Falzoni, J.M. Sanz, A. Morelli, et al., Nucleotide receptors: an emerging family of regulatory molecules in blood cells, Blood 97 (2001) 587–600.

T. Jungas et al. / Microbes and Infection 6 (2004) 1145–1155 [55] C. Mutini, S. Falzoni, D. Ferrari, P. Chiozzi, A. Morelli, O.R. Baricordi, et al., Mouse dendritic cells express the P2X7 purinergic receptor: characterization and possible participation in antigen presentation, J. Immunol. 163 (1999) 1958–1965. [56] A. MacKenzie, H.L. Wilson, E. Kiss-Toth, S.K. Dower, R.A. North, A. Surprenant, Rapid secretion of interleukin-1b by microvesicle shedding, Immunity 15 (2001) 825–834. [57] M. Solle, J. Labasi, D.G. Perregaux, E. Stam, N. Petrushova, B.H. Koller, et al., Altered cytokine production in mice lacking P2X7 receptor, J. Biol. Chem. 276 (2001) 125–132. [58] I.P. Fairbairn, C.B. Stober, D.S. Kumararatne, D.A. Lammas, ATPmediated killing of intracellular mycobacteria by macrophages is a P2X7-dependent process inducing bacterial death by phagosomelysosome fusion, J. Immunol. 167 (2001) 3300–3307. [59] D.J. Kusner, J. Adams, ATP-induced killing of virulent Mycobacterium tuberculosis within human macrophages requires phospholipase D, J. Immunol. 164 (2000) 379–388. [60] D.J. Kusner, J.A. Barton, ATP stimulates human macrophages to kill intracellular virulent Mycobacterium tuberculosis via calciumdependent phagosome-lysosome fusion, J. Immunol. 167 (2001) 3308–3315.

1155

[61] D.A. Lammas, C. Stober, C.J. Harvey, N. Kendrick, S. Panchalingam, D.S. Kumararatne, ATP-induced killing of mycobacteria by human macrophages is mediated by purinergic P2Z (P2X7) receptors, Immunity 7 (1997) 433–444. [62] C.B. Stober, D.A. Lammas, C.M. Li, D.S. Kumararatne, S.L. Lightman, C.A. McArdle, ATP-mediated killing of Mycobacterium bovis bacille Calmette-Guerin within human macrophages is calcium dependent and associated with the acidification of mycobacteriacontaining phagosomes, J. Immunol. 166 (2001) 6276–6286. [63] A. Molloy, P. Laochumroonvorapong, G. Kaplan, Apoptosis, but not necrosis, of infected monocytes is coupled with killing of intracellular bacillus Calmette-Guerin, J. Exp. Med. 180 (1994) 1499–1509. [64] B. Degryse, T. Bonaldi, P. Scaffidi, S. Müller, M. Resnati, F. Sanvito, et al., The high mobility group boxes of the nuclear protein HMG1 induce chemotaxis and cytoskeleton reorganization in rat smooth muscle cells, J. Cell Biol. 152 (2001) 1197–2006. [65] H. Su, G. McClarty, F. Dong, G.M. Hatch, Z.K. Pan, G. Zhong, Activation of Raf/MEK/ERK/cPLA2 signalling pathway is essential for chlamydial acquisition of host glycerophospholipids, J. Biol. Chem. 279 (2004) 9409–9416.