Neisserial porin (PorB) causes rapid calcium influx in target cells and induces apoptosis by the activation of cysteine proteases
1999; Springer Nature; Volume: 18; Issue: 2 Linguagem: Inglês
10.1093/emboj/18.2.339
ISSN1460-2075
Autores Tópico(s)Adenosine and Purinergic Signaling
ResumoArticle15 January 1999free access Neisserial porin (PorB) causes rapid calcium influx in target cells and induces apoptosis by the activation of cysteine proteases Anne Müller Anne Müller Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Dirk Günther Dirk Günther Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Frank Düx Frank Düx Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Michael Naumann Michael Naumann Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Thomas F. Meyer Corresponding Author Thomas F. Meyer Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Thomas Rudel Thomas Rudel Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Anne Müller Anne Müller Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Dirk Günther Dirk Günther Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Frank Düx Frank Düx Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Michael Naumann Michael Naumann Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Thomas F. Meyer Corresponding Author Thomas F. Meyer Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Thomas Rudel Thomas Rudel Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany Search for more papers by this author Author Information Anne Müller1,‡, Dirk Günther1,‡, Frank Düx1, Michael Naumann1, Thomas F. Meyer 1 and Thomas Rudel1 1Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustraße 2, 10117 Berlin, Germany ‡A.Müller and D.Günther contributed equally to this work *Corresponding author. E-mail: [email protected] The EMBO Journal (1999)18:339-352https://doi.org/10.1093/emboj/18.2.339 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The porin (PorB) of Neisseria gonorrhoeae is an intriguing bacterial factor owing to its ability to translocate from the outer bacterial membrane into host cell membranes where it modulates the infection process. Here we report on the induction of programmed cell death after prolonged infection of epithelial cells with pathogenic Neisseria species. The underlying mechanism we propose includes translocation of the porin, a transient increase in cytosolic Ca2+ and subsequent activation of the Ca2+ dependent protease calpain as well as proteases of the caspase family. Blocking the porin channel by ATP eliminates the Ca2+ signal and also abolishes its pro-apoptotic function. The neisserial porins share structural and functional homologies with the mitochondrial voltage-dependent anion channels (VDAC). The neisserial porin may be an analogue or precursor of the ancient permeability transition pore, the putative central regulator of apoptosis. Introduction Apoptosis is a genetically determined form of cell death, characterized by morphological and biochemical features such as nuclear and cytoplasmic condensation, blebbing of the plasma membrane, fragmentation of the nucleus and breakdown of DNA into oligonucleosomes (Kerr et al., 1972). In the final stage of apoptosis, cells display special markers on their surface leading to engulfment by phagocytic cells, a process which avoids spillage of intracellular contents and an inflammatory response. Apoptosis plays a central role during development and homeostasis of multicellular organisms. One of the key signalling pathways to apoptosis starts with ligation of death receptors of the tumour necrosis factor (TNF)/nerve growth factor receptor family (reviewed by Whyte, 1996) which leads to the activation of the main 'executioner' molecules, the caspases (Nicholson and Thornberry, 1997; Salvesen and Dixit, 1997). Caspases comprise a family of cysteine proteases with aspartic acid specificity. All members of this family are synthesized as dormant proenzymes that can be activated by removal of the regulatory prodomain and assembly into the active heteromeric protease. The mechanisms of caspase regulation are currently being actively investigated. One model suggests that cytochrome c release from the mitochondria is a prerequisite for subsequent proteolytic activation of caspase 3, a process which is regulated in turn by Bcl-2-like proteins (Golstein 1997; Kluck et al., 1997; Li et al., 1997; Reed, 1997; Bossy-Wetzel et al., 1998; reviewed in Kroemer, 1997a). These molecules have been reported to be localized to mitochondria (Hockenbery et al., 1990). The anti-apoptotic family member Bcl-xL shares structural similarities with the pore-forming domains of bacterial toxins (Muchmore et al., 1996) and, like these toxins, is able to insert into synthetic lipid bilayers and to form functional ion channels (Minn et al., 1997). Another protease often activated during apoptosis is the Ca2+-dependent cysteine protease calpain (Sarin et al., 1993; Squier et al., 1994; Martin and Green, 1995; Nath et al., 1996). One of its substrates is reported to be α–fodrin, a cytoskeletal protein that may be involved in the regulation of apoptotic membrane changes such as phosphatidylserine exposure, membrane blebbing and cellular fragmentation into apoptotic bodies, as well as detachment from the surrounding tissue. The induction of apoptosis in host cells has been described for a variety of bacterial pathogens (Zychlinsky and Sansonetti, 1997). The facultative intracellular pathogens Salmonella (L.M.Chen et al., 1996; Monack et al., 1996) and Shigella induce apoptosis in macrophages (Zychlinsky et al., 1992, 1994; Y.J.Chen et al., 1996b). Another actively investigated bacterial pathogen is Listeria monocytogenes, which induces apoptosis in hepatocytes as well as dendritic cells (Guzman et al., 1996; Rogers et al., 1996). Epithelial pathogen-induced apoptosis has been reported for Helicobacter pylori (Chen et al., 1997) after infection of human gastric epithelial cell lines. The facultative, intracellular, human-specific pathogen Neisseria gonorrhoeae (Ngo) is the etiological agent of the sexually transmitted disease gonorrhea. During the course of infection the pathogen penetrates the mucosa and causes a massive inflammatory response in the subepithelial tissue. Ngo can enter epithelial cells as well as professional phagocytes. Several factors play a role in the infection, most notably the pili, which mediate primary adherence (McGee et al., 1983; Rudel et al., 1992), the Opa-proteins which mediate adhesion and invasion (Makino et al., 1991), and the PorB porin. Porins are found in the outer membrane of Gram-negative bacteria and also in mitochondria. In bacteria they function as homotrimers and generate pores allowing the passage of solutes. PorB of pathogenic Neisseriae has the unusual feature of translocating from the outer membrane of the bacteria into artificial membranes as well as into target cell membranes (Lynch et al., 1984; Blake and Gotschlich, 1987; Weel and van Putten, 1991). The insertion process leads to the formation of a functional channel which, strikingly, is regulated by the eukaryotic host cell (Rudel et al., 1996). Similar to mitochondrial porins, PorB interacts with purine nucleoside triphosphates which down-regulate the pore size and cause a shift in voltage dependence and ion selectivity. The remarkable parallels between PorB and mitochondrial pores led to the hypothesis that early in evolution, intracellular pathogens may have used porin-like channels to induce death of the host when the cytoplasmic level of ATP/GTP was low (Frade and Michaelidis, 1997). In this study we provide evidence that gonococci cause apoptosis in epithelial and phagocytic cells lines in vitro. An intriguing novel mechanism is proposed: translocated neisserial porin induces apoptosis by causing a rapid calcium influx, followed by the activation of the calcium-dependent cysteine protease calpain and the central apoptosis-executing molecules, the caspases. Results Infection of the HeLa epithelial cell line with Ngo strains induces apoptosis Since gonorrhoea is often associated with a strong inflammatory reaction we investigated the potential cytotoxic effects of Ngo during in vitro infections of HeLa cells, a well established infection model for gonococci. Infection of HeLa cells with invasive or piliated Ngo strains at a multiplicity of infection (m.o.i.) of 1 for 15 h resulted in apoptosis detectable by morphological as well as biochemical criteria (Figure 1A–D). The chromosomal DNA of infected cells was fragmented extensively into oligonucleosomes, clearly visible as a characteristic DNA ladder as soon as 12 h after infection (Figure 1A). Neither non-cytotoxic commensal Neisseria strains nor an invasive Escherichia coli strain expressing the invasin of Yersinia pseudotuberculosis induced fragmentation of DNA (data not shown). HeLa cells treated with TNFα in combination with cycloheximide to induce apoptosis exhibited a similar kinetic of chromosomal fragmentation as the infected cells, whereas actinomycin D showed a delayed response. Next, a method that displays fragmented DNA (Figure 1B) was employed, allowing quantification of the cell population. HeLa cells were infected with a piliated (P+), adherent Ngo strain as well as its non-piliated (P−), non-adherent derivative and a commensal Neisseria sicca strain. Cells showing decreased fluorescence after propidium iodide (PI) staining (marked by a gate in Figure 1B) compared to those in G1 or G2 phase of the cell cycle are depicted to the left of the diagram. This so-called hypo-diploid part of the population is constituted of apoptotic cells with degraded DNA which incorporate less PI than cells with intact DNA. Infection of HeLa cells with adherent Ngo or treatment with TNFα/cycloheximide resulted in an increase in the hypo-diploid population from 15 to 36% and 27%, respectively. Non-adherent (P−, Opa−) gonococci and a N.sicca strain did not show this effect (21 and 16%, respectively). Apoptotic cells could also clearly be identified by their distinct morphology (Figure 1C). The observed nuclear changes such as chromatin condensation and nuclear fragmentation (data not shown) as well as membrane blebbing, rounding up and detachment from the culture plate are typical apoptotic features. Only cells infected with the strongly binding P+, Opa− strain N138 (Figure 1C, panel b) displayed these characteristics. In contrast, cells infected with its P−, Opa− variant (panel c) or the commensal N.sicca strain (panel d) were indistinguishable from control cells (panel a). Figure 1.Prolonged infection of HeLa cells with Ngo induces apoptosis. (A) Fragmentation of cellular DNA into oligonucleosomes after stimulation of HeLa cells with actinomycin D or TNFα, or infection with gonococcal strain N242 (P−, Opa+) for the indicated time points. DNA from 2×105 cells was extracted, separated on a 1.8% agarose gel and visualized by ethidium bromide staining. Lane 1, treatment with 1 μg/ml actinomycin D; lane 2, untreated control; lane 3, treatment with 10 ng/ml TNFα and 20 μg/ml cycloheximide; lane 4, infection with strain N242 at an m.o.i. of 10. (B) FACS analysis of the DNA content of infected cells. Cells were either infected with the Ngo strains of the indicated phenotypes at an m.o.i. of 1, or treated with TNFα and cycloheximide or left untreated. Cells were harvested, permeabilized and stained with PI. 10 000 cells per sample were analysed by FACS and cell counts were plotted against PI fluorescence. The percentage of cells of the hypo-diploid population is indicated in every diagram. (C) Morphology of infected HeLa cells. Cells were infected at an m.o.i. of 1 with strain N138 (P+, Opa−) (panel b), a P− variant of N138 (panel c) or N.sicca, selected for adherence, (panel d) for 15 h or left untreated (panel a). (D) FACS analysis of the annexin V binding properties of infected cell populations. Cells were infected for 15 h with the indicated strains, harvested and stained with annexin V-FITC. After counterstaining with PI which labels the permeable, necrotic cells, 10 000 cells per sample were analysed by flow cytometry. Only the PI-negative cells were included in the histogram. The percentage of apoptotic cells showing strong annexin V binding is 3% in the control, 27% in the population infected with strain N138 (P+, Opa−), 6% in the one infected with its P− variant and 22% in the population infected with strain N242 (P−, Opa+). Download figure Download PowerPoint In order to quantify the apoptotic population with an alternative method, the phosphatidylserine exposure was measured by annexin V (FITC) binding and PI counterstaining to distinguish the necrotic cells (Figure 1D). After 15 h, 22% of the cells infected with the P−, Opa+ strain N242, and 27% of the cells infected with the P+, Opa− strain, N138, stained annexin V-positive, while the non-infected control population showed only 3% spontaneous apoptosis. Infection with a P−, Opa− variant of the same strain, which was unable to adhere to or invade epithelial cells, had hardly any effect (6%). In contrast to the adherent Ngo strains, a commensal N.sicca Opa+ strain selected for adherence to HeLa cells was unable to induce apoptosis (not shown). The necrotic population remained unchanged in all experiments (∼5%). After 60 h of infection with the adherent Ngo strains, nearly all the cells were dead but, since the majority of the cells had also incorporated PI at this stage, it was impossible to distinguish apoptosis from secondary necrosis. Adherent Ngo strains also induced apoptosis in other epithelial cells, for instance the 293 cell line (data not shown). Porin induces apoptosis in epithelial and monocytic cell lines Since both adherent and invasive strains induced apoptosis to a similar extent in HeLa cells, and adherence alone is not sufficient, we addressed the question whether bacterial factors other than the Opa invasins or pili are responsible for the observed cytotoxic effect. Therefore, purified neisserial factors were tested for their apoptosis-inducing capacity on epithelial cells. The secreted neisserial IgA protease was ineffective in this respect (data not shown). In contrast, incubation of cultured epithelial or monocyte-like cells with purified PorB porins, e.g. PorBIA of Ngo strain N242 (Figure 2A–E) or PorBIB of Ngo MS11 (not shown), induced apoptosis very efficiently. Cells treated with 7 μg/ml of porin for 15 h showed the DNA fragmentation typical of apoptotic cells, as shown here for U937 cells [Figure 2A, compare lane 4 with lane 2 (control)]. A similar apoptotic effect was achieved by treatment with 10 ng/ml TNFα in combination with 20 μg/ml cycloheximide (Figure 2A, lane 1). Porin at 3.5 μg/ml does not show any effect (Figure 2A, lane 3). Porin-treated cells were assessed further regarding their PI binding capacity (Figure 2B). A dose-dependent increase in the hypo-diploid population was observed after treatment with 4, 7 and 10 μg/ml porin compared with the untreated control. The dose dependence was not linear (15, 21 and 65% apoptotic cells, respectively), indicating that a certain threshold per cell must be exceeded before cells undergo apoptosis. These results were confirmed by microscopy (Figure 2C). The morphology of porin-treated HeLa cells, such as cell shrinkage and membrane blebbing (Figure 2C, phase-contrast, left panel) as well as nuclear fragmentation and chromatin condensation (nuclear staining, right panel) were indistinguishable from apoptotic cells previously observed in infected cultures (Figure 2C, c and d compared with controls in a and b). Since the porin was purified from Ngo in its native trimeric form, it contained lipopolysaccharide (LPS) as well as 0.025% of the detergent LDAO. However, the buffer as well as LPS isolated from gonococcal strain N242 and solubilized in porin purification buffer had no effect on the morphology of the cells (not shown). Figure 2.Treatment of HeLa cells with purified porin induces apoptosis. (A) Fragmentation of cellular DNA into oligonucleosomes after treatment of U937 cells with TNFα and cycloheximide (lane 1) or 3, 5 and 7 μg/ml of porin isolated from strain N242 for 15 h (lanes 3 and 4). An untreated control is shown in lane 2. DNA from 2×105 cells was extracted and separated on a 1.8% agarose gel and visualized by ethidium bromide staining. (B) FACS analysis of the DNA content of porin-treated cells. Cells were incubated with indicated amounts of porin for 15 h, harvested, permeabilized and stained with PI. 10 000 cells per sample were analysed. PI fluorescence was plotted against cell number. The percentage of cells in the hypo-diploid population is indicated in the diagrams. (C) Morphology of porin-treated HeLa cells. Cells were seeded onto Nunc chamber slides, treated with 7 μg/ml porin for 15 h (c–h) or left untreated (a and b), fixed and stained with the DNA binding dye Hoechst 333420. Some wells were treated with 50 μM zVAD-fmk (e and f) or 50 μM calpain I inhibitor LLnL (g and h) for 30 min prior to application of porin. Both phase contrast and fluorescence pictures were taken of the same section. (D) FACS analysis of the annexin V binding properties of a porin-treated population. Cells were incubated with 7 μg/ml porin for 15 h, harvested and stained with annexin V and PI to mark the apoptotic and the permeable, necrotic cells, respectively. 10 000 cells were analysed per sample and annexin V binding was plotted against cell number in a histogram of the PI-negative cells. A control of untreated cells as well as an LPS control is included. The bold line represents the porin treated cells (55% apoptosis), the thin line the untreated control (5% apoptosis) and the dotted thin line the LPS control (5% apoptosis). (E) FACS analysis of the annexin V binding properties of cells treated with recombinant His-tagged porin purified from E.coli. Cells were incubated with 7 μg/ml porin for 15 h and analysed as described. The dotted thin line represents the untreated control (6% apoptosis), the bold line represents the porin-treated cells (38% apoptosis). A Coomassie Blue-stained gel of purified porins from Ngo strain N242 (lanes 2 and 3) and recombinant E.coli expressing His-tagged porin (lane 4) is shown as inset. Visible are the monomeric form around 34 kDa and some bands representing oligomeric forms of porin (lane 2). Due to the insertion of six histidine residues, the recombinant porin exhibits a slightly higher molecular weight. Samples where either incubated at 37°C (lane 2) or boiled (lane 3 and 4) prior to electrophoresis. Molecular weight marker (kDa) was applied in lane 1. Download figure Download PowerPoint A large percentage of the porin-treated cells bound annexin V (Figure 2D). The annexin V-positive cells were not permeable for PI, indicating that the mechanism of cell death induced by porin is clearly apoptotic and not necrotic. Buffer (data not shown) as well as LPS (Figure 2D) had no effects on the annexin V binding capacity of the cells, even at concentrations 10 times higher than those found in the porin preparations (100 μg/mg porin). Interestingly, recombinant His-tagged porin, produced in E.coli and purified from inclusion bodies, induced apoptosis with similar dose-dependence and kinetics as porin purified from Ngo, as determined by annexin V (FITC) binding (Figure 2E). This demonstrates clearly that no other neisserial factors beside the porin are necessary for induction of apoptosis. Both preparations were pure as judged by SDS–PAGE analysis (Figure 2E, inset). Whereas porin from Ngo was purified in its native trimeric form (lane 2) and disassembled by boiling (lane 3), the recombinant porin was purified as monomers (lane 4). It is conceivable that trimers form immediately after insertion of porin monomers into host cell membranes. Porin also induced apoptosis in Chang conjunctiva cells, 293 cells and in the monocytic cell lines U937 and JOSK–M with a similar kinetics to HeLa cells (data not shown). Porin induces rapid calcium fluxes in epithelial and monocytic cell lines A possible mechanism by which the neisserial porin induces apoptosis might depend on its property to form ion-selective channels in the eukaryotic membranes (Rudel et al., 1996). Changes in intracellular Ca2+ have been reported to occur during the onset of apoptosis (McConkey et al., 1988; Jones et al., 1989; Martikainen et al., 1991). Therefore, we investigated the effect of porin treatment on the cytoplasmic Ca2+ levels of the monocytic cell line JOSK-M by FACS analysis. JOSK-M cells grown in medium containing 1 mM Ca2+ were loaded with the two calcium-responsive chromophores, Fluo-3 and Fura Red. The fluorescence intensity of Fluo-3 increases after Ca2+-binding, while that of Fura Red decreases. The basal fluorescence level of the indicator-loaded cells was determined for 1 min. After adding the porin, intracellular Ca2+ was monitored for another 2 min. The ionophore ionomycin was applied to a separate sample prior to the measurements in order to control the proper loading of the cells. The level of intracellular Ca2+ increased immediately upon addition of porin at concentrations of 5–10 μg/ml and peaked after 1–2 min at 2- to 3-fold (in some experiments up to 4-fold) the basal Ca2+ concentration (Figure 3A–D). Application of adequate volumes of detergent-containing buffer, as well as LPS at a concentration of 15 μg/ml, had no effect on the intracellular Ca2+ concentration (Figure 3A). Furthermore, a clear dose-dependence could be observed with increasing porin concentrations (Figure 3B). Whereas porin at 4 μg/ml had no effect on intracellular Ca2+ levels, a significant increase was recorded with 5 μg/ml which was even exceeded at 7 μg/ml. A similar dose-dependence was observed with assays quantifying apoptosis after porin treatment (Figure 2B). This supports the idea that both processes are linked. The Ca2+ measurements by FACS were performed mainly with JOSK-M cells from suspension cultures. However, primary monocytes and granulocytes isolated from fresh blood reacted very similar to JOSK-M cells in the same experiment with regard to the kinetics and dose dependence (data not shown). Adherent cell types like HeLa were not accessible to Ca2+ measurements by FACS since removing them from the cell culture plate by trypsination alone resulted in elevated intracellular Ca2+ concentrations. However, Fluo-3-loaded adherent HeLa cells monitored by confocal microscopy reacted with an increase in Fluo–3 fluorescence intensity upon addition of porin which was similar to the treatment with ionomycin (data not shown). The elevation of intracellular Ca2+ was only transient and returned to basal levels after 3–5 min. Thus, the neisserial porin evokes transient Ca2+ fluxes in both phagocytes and epithelial cells. Figure 3.Porin treatment of JOSK-M cells induces Ca2+ influx from the surrounding medium. Cells were loaded with the dyes Fluo-3 and Fura Red at 10 μg/ml final concentration. Fluorescence intensities in channels 1 (515–535 nm) and 3 (665–685 nm) were routinely measured over a period of 3 min. The baseline was determined for 1 min, then an inducer (e.g. porin or LPS) was applied and the measurement continued for another 2 min. The ratio of Fluo3/ Fura Red intensities was formed and mean values of this ratio were calculated for each time point. For presentation the mean fluorescence intensity of Fluo3/Fura Red was plotted against time. (A) Addition of 7.5 μg/ml porin-induced a Ca2+ influx (black dotted line) while the buffer control (grey dotted line) or 15 μg/ml LPS purified from neisserial strain N242 (black thin line) had no effect. (B) Porin induced Ca2+ influx is dose-dependent and blocked by EGTA. Measurements were performed in RPMI containing 1 mM Ca2+. Porin was applied at three different concentrations: 4 μg/ml (bold black line), 5 μg/ml (grey dotted line) and 7 μg/ml (black dotted line). Ca2+ influx induced by 7 μg/ml is totally abolished in RPMI medium containing 2 mM EGTA (thin black dotted line). (C) Porin-induced Ca2+ measurements performed in PBS in the absence of Ca2+ (black dotted line), in the presence of 3 mM Ca2+ (grey dotted line) and 30 mM Ca2+ (black plain line). (D) The Ca2+ measurement was performed with 7 μg/ml porin (black dotted line), 7 μg/ml porin in the presence of 0.1 mM ATP (black plain line) and 0.1 mM ATP as a control (grey dotted line). Download figure Download PowerPoint Porin triggers the influx of Ca2+ from extracellular sources Cytoplasmic Ca2+ originates from two major sources: it is either released from intracellular Ca2+ stores like the endoplasmic reticulum (ER), the mitochondria or the nucleus, or enters the cells through Ca2+-specific channels of the plasma membrane from the exterior. Several strategies were used to identify the Ca2+ source in the case of porin-induced Ca2+ changes. On the one hand, measurements in medium containing 1 mM Ca2+ were performed as described above. If the extracellular Ca2+ was chelated by the addition of 2 mM EGTA, which reportedly is not cell permeable, the increase in Ca2+ was totally abolished (Figure 3B). Additionally, when the measurements were performed in Ca2+-free phosphate-buffered saline (PBS), no increase in cytoplasmic Ca2+ was recorded upon treatment of the cells with porin. However, addition of 1–30 mM Ca2+ resulted in a concentration-dependent restoring of porin-induced Ca2+ fluxes (Figure 3C). Chemicals known to selectively inhibit intracellular Ca2+ mobilization (10 μM TMB 8) or receptor-mediated Ca2+ entry from the exterior of the cell (10 μM SK&F) had no effect on the time course and the degree of Ca2+ influx (not shown). Interestingly, addition of porin preincubated with either 10 mM ATP or GTP (0.1 mM final concentration) resulted in complete inhibition of the Ca2+ influx. Also, ATP alone added at the same concentration had no effect (Figure 3D). This effect could not be observed with AMP, which reportedly binds to the porin with much lower affinity (data not shown). These findings suggest that the pore is closed after binding to ATP or GTP, as proposed previously (Rudel et al., 1996), and that the Ca2+ influx is regulated by the ATP/GTP level. Altogether these data support the idea that extracellular Ca2+ passes directly through the porin channel. That Ca2+ enters the cell via so far undefined, porin-activated Ca2+ channels appears less probable. Inhibition of porin-induced apoptosis by ATP The first striking morphological features reminiscent of apoptosis appeared as early as 30 min after application of porin to HeLa cells (Figure 4b). Cells showed extensive blebbing of the plasma membrane followed by rounding up, shrinkage and detachment from the culture vessel. After 60–90 min, blebbing gradually declined: the majority of the cells now detached (Figure 4c) and blebbing only occurred in a small population. At these early stages, the cells showed neither annexin V-binding nor DNA fragmentation. Since blocking pore channel formation with ATP completely prevented the rapid Ca2+ signal elicited by porin (Figure 3D), we asked whether ATP-treated porin still induces apoptosis. As previously demonstrated, preincubation of porin with 10 mM ATP closes the channel (Rudel et al., 1996). The above described early signs of apoptosis did not appear after incubation of porin with 10 mM ATP for 20 min at 37°C prior to application (Figure 4e and f). The resulting final concentration of 20 μM ATP alone had no effect on the cells (Figure 4d). These results provide evidence that both processes, Ca2+ influx and apoptosis, are mechanistically linked. Later biochemical features of porin-treated cells such as annexin V-binding do not reveal a clear difference with the ATP-bound form. This discrepancy between early and later characteristics can most likely be attributed to the fact that the non-covalently bound ATP readily dissociates from the porin. Figure 4.Early features of apoptosis are inhibited by pretreatment of porin with ATP. Purified porin was preincubated with 10 mM ATP for 20 min at 37°C before application to HeLa cells at a concentration of 7 μg/ml, resulting in a final ATP concentration of 20 μM. Non-fixed cells were photographed after 30 and 90 min, respectively. Cells in (a) were left untreated, cells in (b) and (c) were treated with porin for 30 and 90 min. (d) shows a control treated with 20 μM ATP for 30 min; the cells in (e) and (f) were treated with preincubated porin for 30 and 90 min. bl, cell showing extensive blebbing; de, cell detached from culture plate. Download figure Download PowerPoint Caspases and the calcium-dependent protease, cal
Referência(s)