Jana N. Radin,1 Christian Gonzalez-Rivera,2 Susan E. Ivie,1Mark S. McClain,1 and Timothy L. Cover1,2,3*
Department of Medicine, Vanderbilt University School of Medicine, Nashville, Tennessee 372321; Department ofMicrobiology and Immunology, Vanderbilt University School of Medicine, Nashville, Tennessee 372322; and
Veterans Affairs Tennessee Valley Healthcare System, Nashville, Tennessee 372123
Received 23 December 2010/Returned for modification 27 January 2011/Accepted 30 March 2011
Helicobacter pylori is a Gram-negative bacterium that colonizes the human stomach and contributes tothe development of peptic ulcer disease and gastric cancer. The secreted pore-forming toxin VacA is oneof the major virulence factors of H. pylori. In the current study, we show that AZ-521 human gastricepithelial cells are highly susceptible to VacA-induced cell death. Wild-type VacA causes death of thesecells, whereas mutant VacA proteins defective in membrane channel formation do not. Incubation ofAZ-521 cells with wild-type VacA results in cell swelling, poly(ADP-ribose) polymerase (PARP) activation,decreased intracellular ATP concentration, and lactate dehydrogenase (LDH) release. VacA-induceddeath of these cells is a caspase-independent process that results in cellular release of histone-bindingprotein high mobility group box 1 (HMGB1), a proinflammatory protein. These features are consistentwith the occurrence of cell death through a programmed necrosis pathway and suggest that VacA can beincluded among the growing number of bacterial pore-forming toxins that induce cell death throughprogrammed necrosis. We propose that VacA augments H. pylori-induced mucosal inflammation in thehuman stomach by causing programmed necrosis of gastric epithelial cells and subsequent release ofproinflammatory proteins and may thereby contribute to the pathogenesis of gastric cancer and pepticulceration.
Helicobacter pylori is a Gram-negative bacterium that colo-nizes about half of the world’s population. H. pylori coloniza-tion of the human stomach is consistently associated with gas-tric mucosal inflammation and is a risk factor for thedevelopment of peptic ulcer disease and distal gastric adeno-carcinoma (14, 63). One of the major virulence factors of H.pylori is the vacuolating toxin VacA (13, 19, 26). VacA isexpressed as a 140-kDa protoxin and undergoes proteolyticprocessing to yield an 88-kDa secreted toxin (13). VacA issecreted through an autotransporter (type V) pathway as asoluble protein into the extracellular space, and a proportionalso remains attached to the bacterial cell surface (13). Thesecreted 88-kDa VacA protein forms anion-selective mem-brane channels in planar lipid bilayers (18, 50, 64), and conse-quently, VacA is classified as a pore-forming toxin. Multiplereceptors for VacA have been identified, including sphingomy-elin, receptor protein-tyrosine phosphatase alpha � (RPTP�),and RPTP� on the surface of gastric epithelial cells and �2integrin on the surface of T cells (28, 35, 53, 62, 73, 74). Uponinternalization by cells, VacA localizes to endosomal compart-ments (31) as well as to mitochondria (3, 7, 21, 27, 30, 70).
VacA causes a wide array of alterations in target cells, in-cluding cell vacuolation, depolarization of the plasma mem-brane potential, permeabilization of epithelial monolayers, de-
tachment of epithelial cells from the basement membrane,disruption of endosomal and lysosomal function, autophagy,interference with antigen presentation, and inhibition of T-cellactivation and proliferation (13, 19, 26, 66). In addition, VacAcan induce death of gastric epithelial cells. Thus far, moststudies of VacA-induced cell death have been conducted usingAGS or MKN28 gastric epithelial cell lines, as well as HeLacells (5, 11, 16, 30, 44, 54, 70, 71). VacA-induced death of thesecells is preceded by activation of Bax and Bak, induction ofmitochondrial damage, reduction of the mitochondrial mem-brane potential, and cytochrome c release (30, 39, 70, 71, 75)and is accompanied by DNA fragmentation (16). On the basisof these observations, VacA-induced cell death has been clas-sified as an apoptotic process (5, 11, 16, 44, 54).
Among several available gastric epithelial cell lines, theAZ-521 cell line is one of the most highly susceptible toVacA-induced cell death. AZ-521 cells have been used inprevious studies for the identification of several VacA re-ceptors and for studies of cellular alterations caused byVacA (20, 28, 53, 73–75). In the current study, we undertookan in-depth study of the process by which VacA causesdeath of these cells. We show that VacA-induced death ofAZ-521 gastric epithelial cells occurs by a process consistentwith programmed necrosis, resulting in extracellular releaseof cellular constituents. This leads to the hypothesis thatVacA-induced programmed necrosis and the resulting re-lease of proinflammatory cellular components augment H.pylori-induced gastric mucosal inflammation and therebycontribute to the pathogenesis of gastric cancer and pepticulceration.
* Corresponding author. Mailing address: Division of InfectiousDiseases, A2200 Medical Center North, Vanderbilt University Schoolof Medicine, Nashville, TN 37232-2605. Phone: (615) 322-2035. Fax:(615) 343-6160. E-mail: [email protected].
Bacterial strains and culture conditions. H. pylori wild-type (WT) strain 60190(ATCC 49503) and isogenic mutants expressing VacA-G14A (50) or VacA �6-27(68), mutant toxins that are defective in membrane channel formation, weregrown on Trypticase soy agar plates containing 5% sheep blood at 37°C inambient air containing 5% CO2. H. pylori liquid cultures were grown in brucellabroth supplemented with 5% fetal bovine serum (FBS; Atlanta Biologicals) or0.5% activated charcoal. WT VacA and VacA �6-27 were purified in oligomericforms from H. pylori culture supernatants, as described previously (15). Beforeaddition to cells, purified VacA was dialyzed into phosphate-buffered saline(PBS) and was then acid activated by the slow addition of 200 mM HCl until apH of 3.0 was reached (15). For experiments using H. pylori broth culturesupernatant (derived from cultures in brucella broth containing FBS), superna-tants were concentrated 30-fold by ultrafiltration with a 30-kDa-cutoff mem-brane. The relative concentrations of VacA in broth culture supernatants fromWT and mutant H. pylori strains were determined by Western blot analysis usingrabbit anti-VacA antiserum (serum no. 958), and the concentrations of VacA inindividual preparations were then normalized. Clostridium perfringens epsilon-toxin was expressed in Escherichia coli, purified, and trypsin activated as de-scribed previously (49, 56).
Cell culture. AZ-521 human gastric adenocarcinoma cells (Culture Collectionof Health Science Resource Bank, Japan Health Science Foundation) weregrown in minimal essential medium supplemented with 10% FBS (Atlanta Bio-logicals) and 1 mM nonessential amino acids. Initial experiments showed thatacid-activated preparations of purified VacA induced significantly greater cellvacuolation and cell death than did nonactivated VacA preparations, and bothVacA-induced cell vacuolation and cell death were potentiated by supplemen-tation of the tissue culture medium with 5 mM ammonium chloride (data notshown). Therefore, in all subsequent experiments the tissue culture medium wassupplemented with 5 mM ammonium chloride, and all preparations of purifiedVacA were acid activated prior to addition to cells (15, 16). Addition of 5 mMammonium chloride did not alter the pH of the tissue culture medium. VacAconcentrations of 10 �g/ml correspond to 114 nM. Madin-Darby canine kidney(MDCK) epithelial cells were grown in Leibovitz L-15 medium supplementedwith 10% FBS.
Cell viability assay. For cell viability assays, AZ-521 or AGS cells were seededat 4 � 104 cells/well into 96-well plates and incubated overnight. Cells were thenincubated with purified WT VacA or purified VacA �6-27 or with serial dilutionsof H. pylori supernatant containing WT VacA or VacA-G14A. Cell viability wasassessed using the CellTiterAQueous One Solution cell proliferation assay (Pro-mega) according to the manufacturer’s instructions. As a control, cells weretreated with 1 �M staurosporine (Cell Signaling), an agent known to causeapoptosis, for various time intervals, and cell viability was assessed as describedabove.
Analysis of HMGB1 release by confocal microscopy and Western blotting. Foranalysis of histone-binding protein high mobility group box 1 (HMGB1) byconfocal microscopy, AZ-521 and MDCK cells were cultured on plastic chamberslides (5 � 104 cells/well in an 8-well chamber slide; LabTek) overnight. AZ-521cells were then incubated with purified WT VacA or VacA �6-27 (40 �g/ml) for12 h. MDCK cells were treated with 30 nM epsilon-toxin for 45 min. Cells werewashed three times with PBS, fixed with 4% formaldehyde for 10 min at roomtemperature, and washed again with PBS. Cells were permeabilized with 0.25%Triton X-100 (in PBS) for 5 min at room temperature, followed by 3 washes withPBS, and then incubated with Image-iT signal enhancer (Invitrogen) for 30 minat room temperature. Cells were stained for HMGB1 using an anti-HMGB1antibody (1:400 in 1% bovine serum albumin [BSA]–PBS; Abcam), followed byincubation with a secondary anti-goat Alexa Fluor 488-conjugated antibody (1:200 in PBS containing 1% BSA). The nuclei were stained by incubating the cellswith 7-amino-actinomycin D (7-AAD; BD Biosciences). Samples were mountedwith Fluor-Gel (Electron Microscopy Sciences) and examined at a �65 magni-fication using an LSM 510 inverted confocal microscope (Carl Zeiss). Noncon-focal differential interference contrast (DIC) images were collected simultane-ously with the confocal images.
For analysis of HMGB1 by Western blotting, AZ-521 cells were seeded at 7 �105 cells/well into 6-well plates and incubated overnight. Following incubationwith WT VacA or VacA �6-27, tissue culture supernatants were collected,proteins were precipitated from the supernatant using trichloroacetic acid(TCA), and cells were lysed in CelLyticM cell lysis reagent (Sigma). Sampleswere standardized on the basis of protein concentrations, which were determinedusing a bicinchoninic acid (BCA) assay kit (Pierce). HMGB1 in whole-cell lysatesor preparations of TCA-precipitated supernatant proteins was detected by West-ern blot analysis using an anti-HMGB1 antibody (1:1,000; Abcam), followed by
a horseradish peroxidase-conjugated secondary antibody (1:10,000; Promega).Proteins were visualized by incubation with a chemiluminescent substrate solu-tion (Pierce) and exposure to X-ray film.
LDH release assay. Cells (AZ-521, AGS, or MDCK) were seeded at 4 � 104
cells/well into 96-well plates. Following incubation with WT VacA, mutant VacAproteins, 1 �M staurosporine, or 30 nM epsilon-toxin, lactate dehydrogenase(LDH) release was measured using a CytoTox 96 nonradioactive cytotoxicityassay (Promega) according to the manufacturer’s instructions. Maximum LDHrelease was determined following treatment of cells with lysis buffer.
PARP activation assay. AZ-521 cells were seeded at 7 � 105 cells/well into6-well plates and incubated overnight. Cells were then incubated for 8 h with H.pylori supernatant containing WT VacA or VacA-G14A or 1 �M staurosporine.Poly(ADP-ribose) polymerase (PARP) activity was measured using an HT Uni-versal colorimetric PARP assay kit (Trevigen) according to the manufacturer’sinstructions.
Analysis of PARP cleavage by Western blotting. Cells (AZ-521 or MDCK)were seeded at 7 � 105 cells/well into 6-well plates and incubated overnight.AZ-521 cells were then incubated with H. pylori supernatant containing WTVacA, or they were treated with 1 �M staurosporine. MDCK cells were treatedwith 30 nM epsilon-toxin. Cells were lysed in CelLyticM cell lysis reagent(Sigma). Cleavage of PARP was assessed by Western blot analysis using ananti-PARP antibody (1:1,000; Clontech), followed by a horseradish peroxidase-conjugated secondary antibody (1:10,000; Promega). Proteins were visualized byincubation with a chemiluminescent substrate solution (Pierce) and exposure toX-ray film.
Measurement of intracellular ATP levels. AZ-521 cells (7 � 105 cells/well in6-well plates) were incubated with H. pylori supernatant containing WT VacA orVacA-G14A, or they were treated with 1 �M staurosporine. MDCK cells (7 �105 cells/well in 6-well plates) were treated with 30 nM epsilon-toxin. Cells wereharvested by trypsinization and resuspended in 200 �l of PBS. Cell suspensionswere mixed with 200 �l of a 10% TCA–4 mM EDTA solution and incubated onice for 10 min. The extracts were centrifuged at 12,000 rpm for 10 min at 4°C, andthe supernatant was collected. ATP was measured using an Enliten ATP assaysystem bioluminescence detection kit (Promega) according to the manufacturer’sinstructions. Samples were standardized on the basis of protein concentrationusing a BCA assay kit (Pierce), and the ATP concentrations of samples werecompared to ATP concentrations detected in extracts from untreated controlcells. Fluorescence measurements and luminescence were measured using aBioTek FLx800 plate reader.
Caspase inhibition. AZ-521 cells were seeded at 4 � 104 cells/well into 96-wellplates and were incubated overnight. Cells were then preincubated with 100 �MQ-VD-OPh (a general caspase inhibitor; R & D Systems) for 3 h, followed byincubation with serial dilutions of H. pylori supernatant containing WT VacA orVacA-G14A for 24 h or incubation with staurosporine for the same length oftime. Cell viability was then assessed using the One CellTiterAQueous Solutioncell proliferation assay (Promega) according to the manufacturer’s instructions.
Caspase-3 activation. AZ-521 cells were seeded at 7 � 105 cells/well into6-well plates and incubated overnight. Cells were then incubated with either WTVacA or 1 �M staurosporine for 4.5 h, and caspase-3 activation was measuredusing a caspase-3 colorimetric assay kit (Genscript) according to the manufac-turer’s instructions.
Select agent. Plasmid DNA capable of expressing Clostridium perfringens ep-silon-protoxin (or epsilon-toxin) is considered a select agent by the U.S. Depart-ment of Health and Human Services.
RESULTS
VacA causes death of AZ-521 gastric epithelial cells. Mostprevious research on VacA-induced cell death was done withAGS or MKN28 human gastric epithelial cell lines or HeLacells (16, 30, 44, 54, 70). Another human gastric epithelial cellline, AZ-521, has also been used extensively for studies ofVacA, and several putative VacA receptors have been identi-fied using this cell line (20, 28, 53, 73, 74). Previous studieshave not investigated whether there are substantial differencesin the susceptibility of different gastric cell lines to VacA-induced cell death. Therefore, we compared AZ-521 and AGScells. Cells were incubated with serial dilutions of H. pyloribroth culture supernatant containing WT VacA or an inactivemutant protein (VacA-G14A) (Fig. 1A) (50). Alternatively,
cells were incubated with purified WT VacA or a purifiedinactive mutant protein (VacA �6-27) (Fig. 1B) (68). Cellviability was quantified by measuring cellular metabolic activ-ity. WT VacA induced cell vacuolation within several hoursafter addition of the toxin, whereas the mutant toxins did notcause cell vacuolation (data not shown). Within 24 h, WTVacA induced death of AZ-521 cells, whereas the mutant
toxins did not (Fig. 1A and B). WT VacA did not cause deathof AGS cells after 24 h (Fig. 1C) and, consistent with previousresults (11, 16, 44), caused death of AGS cells only after longerincubation periods (data not shown). These initial experimentsshowed that, in comparison to AGS cells, AZ-521 cells aremuch more susceptible to VacA-induced cell death. We con-ducted further experiments to investigate the process by whichVacA causes death of AZ-521 cells.
Incubation of AZ-521 cells with VacA leads to cell swelling.To determine whether VacA-induced death of AZ-521 cellsoccurred by an apoptotic or a programmed necrosis pathway,we used several different experimental approaches. The firstexperiment examined cell size after incubation with VacA.Necrotic cells typically exhibit cell and organelle swelling (33),whereas apoptotic cells typically exhibit a reduction in cell size(29). To determine whether incubation of cells with VacAresults in cell swelling or cell shrinkage, we incubated AZ-521cells with WT VacA (10 �g/ml) or medium alone, and cell sizewas quantified by flow cytometric measurements of forwardscatter (FSC-A). Treatment of cells with WT VacA resulted ina 1.6-fold increase in mean cell size compared to that of un-treated cells (Fig. 2; P � 0.0001, Student’s t test). The observedincrease in cell size is atypical for an apoptotic process andsuggested that VacA might cause death of AZ-521 cellsthrough a necrotic pathway.
Cell death induced by VacA results in LDH release. Inaddition to cell swelling, necrotic cells typically undergo organ-elle damage and plasma membrane rupture, leading to therelease of intracellular components such as LDH (33, 52). Incontrast, the cell membranes of apoptotic cells remain intactuntil late stages, when the cells undergo secondary necrosis(29, 59). We therefore assessed whether incubation of AZ-521cells with VacA resulted in LDH release. Cells were incubatedwith either WT VacA or VacA �6-27 for various time intervals,and LDH release was analyzed at each time point. WT VacAinduced the release of LDH beginning about 10 h after theaddition of toxin, whereas VacA �6-27 did not induce LDHrelease (Fig. 3A). By 15 to 20 h after addition of WT VacA,
FIG. 1. Effects of VacA on viability of AZ-521 cells. (A) AZ-521cells were incubated with serial dilutions of H. pylori broth culturesupernatant containing WT VacA or VacA-G14A. Concentrations ofWT VacA and VacA-G14A were normalized by Western blot analysis,as described in Materials and Methods. After 24 h, cell viability wasassessed using the CellTiterAQueous One Solution cell proliferationassay. (B) AZ-521 cells were incubated with the indicated concentra-tions of purified WT VacA or VacA �6-27, and cell viability wasassessed after 24 h. Error bars represent means � standard deviationsbased on triplicate determinations from a single representative exper-iment. The experiment was performed three times with similar results.(C) AGS cells were incubated with serial dilutions of H. pylori brothculture supernatant containing WT VacA or VacA-G14A. Cell viabil-ity was assessed after 24 h. Error bars represent means � standarddeviations based on combined results of two independent experiments,each performed in triplicate.
FIG. 2. Incubation of AZ-521 cells with VacA leads to cell swelling.AZ-521 cells were incubated with WT VacA (10 �g/ml) or mediumalone for 6 h, and cell size was quantified by flow cytometric measure-ments of forward scatter (FSC-A). Treatment of cells with WT VacAresulted in a 1.6-fold increase in mean cell size compared to that ofuntreated cells (mean forward scatter fluorescence of 117,687 � 10,133compared to 71,943 � 4,607) (P � 0.0001, Student’s t test). Meanfluorescence values reflect results of three independent experiments,and the images depict representative results. SSC-A, side scatter.
VOL. 79, 2011 H. PYLORI VacA AND PROGRAMMED NECROSIS 2537
maximum LDH release (corresponding to 100% cell killing)was detected.
For comparison, we analyzed the effects of staurosporine, anagent that is known to induce apoptosis. On the basis of theresults of a cell viability assay, staurosporine-treated cells un-derwent cell death beginning several hours after addition ofthe drug, but the cells did not release any substantial amountsof LDH until much later time points (Fig. 3B). As a control foran agent expected to induce necrosis, we analyzed effects ofanother bacterial pore-forming toxin, C. perfringens epsilon-toxin, on a susceptible cell line (MDCK cells). Incubation ofMDCK cells with 30 nM epsilon-toxin resulted in near-maximal LDH release within 6 h after addition of the toxin(Fig. 3C).
To evaluate whether the observed necrosis of AZ-521 cells
in response to VacA was due to primary necrosis or secondarynecrosis (i.e., subsequent to apoptosis), we assessed cell viabil-ity as well as LDH release at early time points. We incubatedAZ-521 cells with WT VacA and VacA-G14A for 4, 6, and 24 hand quantified both LDH release and cell viability. These ex-periments showed that there was relatively little loss of cellviability at early time points (�6 h), and at these early timepoints, the level of LDH release was comparable to the pro-portion of dead cells (Fig. 4). When they are combined withthe results shown in Fig. 3, these data suggest that the observedcell death is due to primary necrosis rather than secondarynecrosis.
We then compared the ability of VacA to cause LDH re-lease from AZ-521 cells with its effect on AGS cells. Similar towhat we observed in the cell viability assays (Fig. 1), WT VacAcaused substantial LDH release from AZ-521 cells but minimalLDH release from AGS cells after 24 h of incubation (Fig. 5Aand B). After 48 h of incubation, VacA caused LDH releasefrom AGS cells corresponding to about 30 to 40% of maximallysis (Fig. 5D). These experiments confirmed that AZ-521 cellsare much more susceptible to VacA-induced cell death than
FIG. 3. Cell death induced by VacA results in LDH release.(A) AZ-521 cells were incubated with purified WT VacA or VacA�6-27 (20 �g/ml; 228 nM), and LDH release was quantified at varioustime points. (B) AZ-521 cells were treated with 1 �M staurosporine.LDH release and cell viability (based on a metabolic assay) wereassessed at various time points. (C) MDCK cells were treated with C.perfringens epsilon-toxin (30 nM), and LDH release was quantified atvarious time points. Error bars represent means � standard deviationsbased on combined results of three independent experiments, eachperformed in triplicate. LDH release is expressed as a percentage ofcontrol, which is the maximal amount of LDH that can be released bythe cells.
FIG. 4. Incubation of AZ-521 cells with VacA results in primarynecrosis. AZ-521 cells were incubated with H. pylori broth culturesupernatants containing normalized concentrations of WT VacA orVacA-G14A. (A) LDH release was quantified 4, 6, and 24 h afteraddition of VacA. (B) Cell viability was assessed 4, 6, and 24 h afteraddition of VacA. Error bars represent means � standard deviationsbased on combined results of three independent experiments, eachperformed in six replicates. LDH release is expressed as a percentageof control, which is the maximal amount of LDH that can be releasedby the cells.
AGS cells and revealed that VacA induces LDH release, acharacteristic of necrotic cell death, in both cell lines.
Cell death induced by VacA results in PARP activation.Poly(ADP-ribose) polymerase is a nuclear enzyme that cata-lyzes the transfer of ADP-ribose moieties from NAD� to itselfand other acceptor proteins in response to DNA damage (36,52). PARP initially functions to alter chromatin structure andfacilitate DNA repair following low levels of DNA damage.However, following severe damage, sustained PARP activityleads to depletion of NAD� and ATP, resulting in irreversibleenergy failure and eventually necrotic cell death (8). PARPactivation is commonly observed in necrotic cells but not inapoptotic cells. PARP is inactivated by caspase cleavage inapoptosis, an ATP-dependent process (46, 61), to preventPARP-induced ATP depletion.
To determine whether VacA treatment induced PARP ac-tivation, we treated AZ-521 cells with WT VacA or VacA-G14A for 8 h and then analyzed activation of PARP. Incuba-tion of AZ-521 cells with WT VacA resulted in increasedPARP activity, while incubation with VacA-G14A or stauro-sporine did not (Fig. 6A). We also examined PARP cleavageby Western blot analysis. As expected, PARP cleavage oc-curred in staurosporine-treated cells, but PARP cleavage wasnot detected in VacA-treated cells or in MDCK cells treatedwith epsilon-toxin (Fig. 6B). Since WT VacA did not inducePARP cleavage, the responses to WT VacA and VacA-G14Awere indistinguishable (Fig. 6B). Activation of PARP and ab-sence of detectable PARP cleavage in response to VacA orepsilon-toxin are consistent with necrotic cell death.
Cell death induced by VacA leads to reduction of intracel-lular ATP levels. Sustained PARP activity can lead to deple-
tion of NAD� and ATP, and a reduction in intracellular ATPlevels is considered to be a hallmark of necrosis (23). In con-trast, apoptosis is an ATP-dependent process and intracellularATP levels typically remain relatively unchanged in apoptoticcells until the late stages of apoptosis (34). We therefore as-sessed whether incubation of AZ-521 cells with VacA resultedin decreased intracellular ATP levels. Cells treated with WTVacA showed progressively decreasing ATP levels over time(Fig. 7A), whereas this phenomenon was not observed whencells were treated with VacA-G14A. Similarly, when MDCKcells were treated with 30 nM epsilon-toxin, a toxin that isknown to induce necrosis, intracellular ATP was rapidly de-pleted (Fig. 7B). Cells treated with staurosporine exhibitedonly a small reduction in ATP levels (Fig. 7A). The observeddepletion of intracellular ATP induced by VacA is consistentwith induction of cell death by a necrotic pathway.
VacA-induced cell death does not depend on caspase acti-vation. Since caspases are important mediators of apoptosis(42, 67), we next investigated whether inhibition of caspaseactivation would prevent VacA-induced cell death. AZ-521cells were preincubated with the general caspase inhibitor Q-VD-OPh, which can prevent apoptosis by the three majorapoptotic pathways (caspase 9/3, caspase 8/10, and caspase 12)(9), and we then treated the cells with either WT VacA orstaurosporine. While inhibition of caspase activation blockedapoptosis induced by staurosporine (Fig. 8A), it had no effecton VacA-induced cell death (Fig. 8B).
To further analyze the role that caspases play in VacA-induced cell death, we assessed whether incubation of cellswith VacA resulted in activation of caspase-3. While incuba-tion of AZ-521 cells with staurosporine resulted in activationof caspase-3, incubation of cells with VacA had no effect on
FIG. 5. Cell death induced by VacA results in LDH release in bothAZ-521 and AGS cells. AZ-521 and AGS cells were incubated withserial dilutions of H. pylori broth culture supernatants containing nor-malized concentrations of WT VacA or VacA-G14A. LDH release wasquantified at 24 h (A and B) and 48 h (C and D) after addition ofVacA. Error bars represent means � standard deviations based oncombined results of two independent experiments, each performed intriplicate. LDH release is expressed as a percentage of control, whichis the maximal amount of LDH that can be released by the cells.
FIG. 6. Cell death induced by VacA results in PARP activation.(A) AZ-521 cells were treated for 8 h with H. pylori broth culturesupernatants containing normalized concentrations of WT VacA orVacA-G14A or with 1 �M staurosporine. The background level ofPARP activity detected in untreated cells was subtracted from thevalues determined for treated cells. Error bars represent means �standard deviations from combined results of two independent exper-iments, each performed in triplicate. (B) Western blot analysis ofPARP in AZ-521 cells treated for 8 h with H. pylori supernatantcontaining WT VacA or VacA-G14A or with 1 �M staurosporine (ST)and MDCK cells treated with C. perfringens epsilon-toxin (30 nM) for30 min.
VOL. 79, 2011 H. PYLORI VacA AND PROGRAMMED NECROSIS 2539
caspase-3 (Fig. 8C). These results provide evidence that VacA-induced cell death occurs through a caspase-independent pro-cess.
Cell death induced by VacA results in the release ofHMGB1. The results obtained so far suggested that VacA-induced cell death occurs primarily by a necrotic pathwayrather than an apoptotic pathway. Previous studies have re-ported that Clostridium septicum alpha-toxin- and Staphylococ-cus aureus alpha-toxin-induced necrosis of cells results in therelease of HMGB1 (a proinflammatory protein) into the cyto-plasm and then into the culture supernatant (24, 38). Releaseof HMGB1 is a hallmark of necrotic cells but not apoptoticcells (60). To determine whether VacA-induced cell deathresulted in release of HMGB1, we incubated AZ-521 cells withpurified WT VacA or VacA �6-27 and assessed localization ofHMGB1 by confocal microscopy (Fig. 9A). In untreated cellsand cells that were treated with VacA �6-27, HMGB1 re-mained in the nucleus. In contrast, in cells treated with WTVacA, HMGB1 localized in the cytoplasm and not the nucleus.Immunoblot analysis confirmed that, in response to treatmentwith WT VacA, there was a progressive reduction of cellularHMGB1 accompanied by release of HMGB1 into the culturesupernatant (Fig. 9B). As expected, HMGB1 colocalized withthe nucleus of untreated MDCK cells, but as was observed withVacA-treated cells, HMGB1 was released into the cytoplasmof epsilon-toxin-treated cells (Fig. 9C) and was subsequently
released into the culture supernatant (data not shown). Theseresults indicate that VacA induces cell death by a necroticpathway, resulting in the release of HMGB1.
DISCUSSION
Bacteria and bacterial protein toxins are capable of causingcell death by multiple pathways, including apoptosis, necrosisor lysis, programmed necrosis, pyroptosis, and autophagic celldeath (45, 58). In contrast to the cellular necrosis that occurs inresponse to nonspecific cellular insults (such as treatment withdetergent, high concentrations of peroxide, or freeze-thaw cy-
FIG. 7. Cell death induced by VacA leads to reduction of intracel-lular ATP levels. (A) AZ-521 cells were treated for the indicated timeswith H. pylori broth culture supernatants containing normalized con-centrations of WT VacA or VacA-G14A, or cells were treated with 1�M staurosporine. Cellular ATP concentrations were measured atvarious time points and are reported as a percentage of the ATPconcentrations of untreated control cells. (B) MDCK cells weretreated for the indicated times with C. perfringens epsilon-toxin (30nM), and cellular ATP concentrations were determined. Error barsrepresent means � standard deviations from combined results of threeindependent experiments, each performed in triplicate. Statistical sig-nificance was assessed using analysis of variance and Dunnett’s posthoc test. *, P 0.05 compared to time zero.
FIG. 8. VacA-induced cell death does not depend on caspase ac-tivation. AZ-521 cells were pretreated with the general caspase inhib-itor Q-VD-OPh (100 �M) or dimethyl sulfoxide (DMSO) alone for 3 hand then treated with 2 �M staurosporine for 24 h (A) or incubated for24 h with serial dilutions of H. pylori broth culture supernatant con-taining WT VacA (B). The inhibitor remained present throughout theexperiment. Cell viability was assessed using the CellTiterAQueous OneSolution cell proliferation assay. Error bars represent means � stan-dard deviations of measurements from combined results of two inde-pendent experiments, each performed in triplicate. (C) AZ-521 cellswere incubated for 4.5 h with H. pylori broth culture supernatantcontaining WT VacA or with 1 �M staurosporine, and caspase-3 ac-tivation was determined using a caspase-3 colorimetric assay kit.OD415, optical density at 415 nm.
cles), programmed necrosis (also known as pyronecrosis, on-cosis, necroptosis, or PARP-induced cell death) is initiatedthrough activation of specific signaling pathways (29, 33).These include PARP-1 hyperactivation (77), tumor necrosisfactor alpha receptor signaling (10), and RIP1 kinase signaling(12). Programmed necrosis can be induced by extracellularsignals or can be the result of intracellular perturbations (78).
Apoptotic cells and necrotic cells share several commonfeatures but can be differentiated by a variety of morphologicaland biochemical criteria. Typical morphological features ofcells undergoing apoptosis include cell rounding, cell shrink-age, and plasma membrane blebbing (29, 33). Other featuresof apoptotic cells include caspase activation, PARP cleavage,retention of intracellular ATP levels, and absence of plasmamembrane rupture (29). Cell swelling, vacuolation of the cy-toplasm, PARP activation, ATP depletion, early plasma mem-brane rupture, and the release of proinflammatory proteins arecharacteristics of cells undergoing programmed necrosis, butthese features are not characteristics of apoptotic cells (40, 47).Mitochondrial alterations and DNA damage can occur in bothcells undergoing apoptosis and necrotic cells (6, 32, 33, 40,43, 72).
In the current study, we observed that AZ-521 gastric epi-thelial cells were much more susceptible to VacA-induced celldeath than were AGS gastric epithelial cells. Therefore, weinvestigated the process by which VacA causes death of AZ-521 cells. As controls, we analyzed the effects of staurosporine(an agent known to cause apoptosis) and C. perfringens epsilon-toxin (a pore-forming toxin that causes necrosis). Our resultsshowed that the effects of VacA were different from those ofstaurosporine and resembled the effects of epsilon-toxin.Therefore, we conclude that VacA induces death of AZ-521cells by a programmed necrosis pathway.
Previous studies have reported that VacA is capable of caus-ing death of AGS gastric epithelial cells and several other celllines (5, 7, 16, 39, 44). On the basis of the observations thatVacA-induced cell death is preceded by the activation of Baxand Bak (75), that VacA can localize to mitochondria (7, 21,27, 30, 70), and that incubation of cells with VacA results inreduction of the mitochondrial transmembrane potential andrelease of cytochrome c (30, 39, 70), VacA-induced cell deathof these cell lines was classified as an apoptotic process (5, 16,44, 54, 71). However, these phenomena can occur in cellsundergoing either apoptotic or nonapoptotic cell death (33, 40,41). Several findings in a previous study (70) suggested thatthere might be differences between the process by which VacAinduces death of HeLa cells compared to the process by whichknown apoptosis-inducing agents cause cell death. (i) Inhibi-tion of cellular caspases had no effect on VacA-induced reduc-tion of the mitochondrial transmembrane potential or cyto-chrome c release (70). (ii) VacA-induced reduction of themitochondrial transmembrane potential occurred prior to cy-tochrome c release (70), whereas the apoptosis-inducing agentactinomycin D induced reduction in the mitochondrial trans-membrane potential only after the release of cytochrome c(69). (iii) A lower VacA concentration was needed to inducethe reduction of the mitochondrial transmembrane potentialthan was needed to induce cytochrome c release (70). In thecurrent study, we did not conduct detailed studies of the pro-cess by which VacA causes death of AGS or HeLa cells. How-ever, we observed that VacA-induced death of AGS cells isassociated with LDH release. This suggests that, similar toVacA-induced death of AZ-521 cells, VacA-induced death ofAGS cells may occur at least in part through programmednecrosis.
An important difference between apoptosis and pro-grammed necrosis is that the plasma membrane remains intactin apoptotic cells but not in necrotic cells. Apoptotic cells areengulfed in vivo by resident phagocytes, thereby preventingcells from undergoing secondary necrosis and releasing intra-cellular contents. As a consequence, apoptosis usually elicits avery limited inflammatory response (59). In contrast, necroticcells undergo early plasma membrane rupture, resulting inrelease of intracellular contents that can promote an inflam-matory response (33). One of the proinflammatory proteinsreleased by necrotic cells is HMGB1. HMGB1 is expressed byalmost all cells and is one of the most abundant proteins in thenucleus. In healthy cells, it is involved in facilitating transcrip-tion factor binding, nucleosome remodeling, and DNA repair(4). The release of HMGB1 from necrotic cells occurs by aprocess that requires PARP activation, and the release ofHMGB1 results in a proinflammatory response (57, 60, 76).
FIG. 9. Cell death induced by VacA results in release of HMGB1.(A) AZ-521 cells were treated with purified WT VacA or VacA �6-27(40 �g/ml; 456 nM) for 12 h. (B) AZ-521 cells were treated with 10�g/ml of WT VacA or VacA �6-27 for various time intervals.(C) MDCK cells were treated with C. perfringens epsilon-toxin (30 nM)for 45 min. Localization of HMGB1 was assessed by confocal micros-copy (A and C). Cell-associated HMGB1 and release of HMGB1 intothe culture supernatant were detected by Western blot analysis (B).
VOL. 79, 2011 H. PYLORI VacA AND PROGRAMMED NECROSIS 2541
When it is released from cells, HMGB1 can bind to severalreceptors on immune cells, including receptor for advancedglycation end products (RAGE), Toll-like receptor 2 (TLR2),and TLR4 (55). This results in maturation of dendritic cells,activation of T cells, and the release of proinflammatory cyto-kines by monocytes, T cells, and endothelial cells (22, 51).Additionally, HMGB1 has been shown to promote colon car-cinogenesis in an inflammation-based model (48). Our resultsindicate that VacA treatment of cells results in the release ofHMGB1, which would be capable of stimulating a proinflam-matory response.
Colonization of the stomach with H. pylori results in chronicgastric inflammation, and inflammation contributes to thepathogenesis of peptic ulcer disease and distal gastric adeno-carcinoma (14, 63). Data from numerous studies suggest thatVacA contributes to gastric mucosal inflammation and devel-opment of peptic ulceration or gastric cancer, in a manner thatis independent of the effects of the cag pathogenicity island (1,2, 25, 65). The data reported in the current study suggest thatVacA may contribute to gastric inflammation by causing pro-grammed necrosis of gastric epithelial cells and subsequentrelease of proinflammatory proteins.
In summary, these results indicate that VacA-induced celldeath of AZ-521 gastric epithelial cells occurs by a pro-grammed necrosis pathway. Several other bacterial pore-form-ing toxins, including Staphylococcus aureus alpha-toxin, Esch-erichia coli hemolysin, Clostridium septicum alpha-toxin, andClostridium perfringens epsilon-toxin, are also known to inducecell death by programmed necrosis (17, 24, 37, 38). VacA cantherefore be included among the growing number of bacterialpore-forming toxins that induce cell death by programmednecrosis and thereby trigger host inflammatory responses. Wepropose that VacA augments mucosal inflammation in thehuman stomach by causing programmed necrosis of gastricepithelial cells and subsequent release of proinflammatory pro-teins and may thereby contribute to the pathogenesis of gastriccancer and peptic ulceration.
ACKNOWLEDGMENTS
This work was funded by NIH grants AI39657, AI068009,CA116087, and AI079123, the Molecular Microbial PathogenesisTraining Program (T32 AI007281-21), and the U.S. Department ofVeterans Affairs. The VMC Flow Cytometry Shared Resource andCell Imaging Shared Resource are supported by the Vanderbilt In-gram Cancer Center (P30 CA68485) and the Vanderbilt DigestiveDisease Research Center (DK058404).
The content is solely the responsibility of the authors and does notnecessarily represent the official views of the National Institute ofAllergy and Infectious Diseases, the National Institutes of Health, orthe U.S. Department of Veterans Affairs.
We thank Beverly Hosse for assistance with VacA purification,Holly M. Algood for helpful discussions, and David K. Flaherty andBrittany Matlock for assistance with flow cytometry.
REFERENCES
1. Atherton, J. C., et al. 1995. Mosaicism in vacuolating cytotoxin alleles ofHelicobacter pylori. Association of specific vacA types with cytotoxin produc-tion and peptic ulceration. J. Biol. Chem. 270:17771–17777.
2. Atherton, J. C., R. M. Peek, Jr., K. T. Tham, T. L. Cover, and M. J.Blaser. 1997. Clinical and pathological importance of heterogeneity invacA, the vacuolating cytotoxin gene of Helicobacter pylori. Gastroenter-ology 112:92–99.
3. Blanke, S. R. 2005. Micro-managing the executioner: pathogen targeting ofmitochondria. Trends Microbiol. 13:64–71.
4. Bonaldi, T., G. Langst, R. Strohner, P. B. Becker, and M. E. Bianchi. 2002.
The DNA chaperone HMGB1 facilitates ACF/CHRAC-dependent nucleo-some sliding. EMBO J. 21:6865–6873.
5. Boquet, P., V. Ricci, A. Galmiche, and N. C. Gauthier. 2003. Gastric cellapoptosis and H. pylori: has the main function of VacA finally been identi-fied? Trends Microbiol. 11:410–413.
6. Bortner, C. D., N. B. Oldenburg, and J. A. Cidlowski. 1995. The role of DNAfragmentation in apoptosis. Trends Cell Biol. 5:21–26.
7. Calore, F., et al. 2005. Endosome-mitochondria juxtaposition during apop-tosis induced by H. pylori VacA. Cell Death Differ. 17:1707–1716.
8. Carson, D. A., S. Seto, D. B. Wasson, and C. J. Carrera. 1986. DNA strandbreaks, NAD metabolism, and programmed cell death. Exp. Cell Res. 164:273–281.
9. Caserta, T. M., A. N. Smith, A. D. Gultice, M. A. Reedy, and T. L. Brown.2003. Q-VD-OPh, a broad spectrum caspase inhibitor with potent antiapo-ptotic properties. Apoptosis 8:345–352.
10. Challa, S., and F. K. Chan. 2010. Going up in flames: necrotic cell injury andinflammatory diseases. Cell. Mol. Life Sci. 67:3241–3253.
11. Cho, S. J., et al. 2003. Induction of apoptosis and expression of apoptosisrelated genes in human epithelial carcinoma cells by Helicobacter pyloriVacA toxin. Toxicon 42:601–611.
12. Cho, Y. S., et al. 2009. Phosphorylation-driven assembly of the RIP1-RIP3complex regulates programmed necrosis and virus-induced inflammation.Cell 137:1112–1123.
13. Cover, T. L., and S. R. Blanke. 2005. Helicobacter pylori VacA, a paradigmfor toxin multifunctionality. Nat. Rev. Microbiol. 3:320–332.
14. Cover, T. L., and M. J. Blaser. 2009. Helicobacter pylori in health and disease.Gastroenterology 136:1863–1873.
15. Cover, T. L., P. I. Hanson, and J. E. Heuser. 1997. Acid-induced dissociationof VacA, the Helicobacter pylori vacuolating cytotoxin, reveals its pattern ofassembly. J. Cell Biol. 138:759–769.
16. Cover, T. L., U. S. Krishna, D. A. Israel, and R. M. Peek, Jr. 2003. Inductionof gastric epithelial cell apoptosis by Helicobacter pylori vacuolating cyto-toxin. Cancer Res. 63:951–957.
17. Craven, R. R., et al. 2009. Staphylococcus aureus alpha-hemolysin activatesthe NLRP3-inflammasome in human and mouse monocytic cells. PLoS One4:e7446.
18. Czajkowsky, D. M., H. Iwamoto, T. L. Cover, and Z. Shao. 1999. Thevacuolating toxin from Helicobacter pylori forms hexameric pores in lipidbilayers at low pH. Proc. Natl. Acad. Sci. U. S. A. 96:2001–2006.
19. de Bernard, M., A. Cappon, G. Del Giudice, R. Rappuoli, and C. Monte-cucco. 2004. The multiple cellular activities of the VacA cytotoxin of Heli-cobacter pylori. Int. J. Med. Microbiol. 293:589–597.
20. De Guzman, B. B., et al. 2005. Cytotoxicity and recognition of receptor-likeprotein tyrosine phosphatases, RPTPalpha and RPTPbeta, by m2 VacA. CellMicrobiol. 7:1285–1293.
21. Domanska, G., et al. 2010. Helicobacter pylori VacA toxin/subunit p34: tar-geting of an anion channel to the inner mitochondrial membrane. PLoSPathog. 6:e1000878.
22. Dumitriu, I. E., et al. 2005. Release of high mobility group box 1 by dendriticcells controls T cell activation via the receptor for advanced glycation endproducts. J. Immunol. 174:7506–7515.
23. Eguchi, Y., S. Shimizu, and Y. Tsujimoto. 1997. Intracellular ATP levelsdetermine cell death fate by apoptosis or necrosis. Cancer Res. 57:1835–1840.
25. Figueiredo, C., et al. 2002. Helicobacter pylori and interleukin 1 genotyping:an opportunity to identify high-risk individuals for gastric carcinoma. J. Natl.Cancer Inst. 94:1680–1687.
26. Fischer, W., S. Prassl, and R. Haas. 2009. Virulence mechanisms and per-sistence strategies of the human gastric pathogen Helicobacter pylori. Curr.Top. Microbiol. Immunol. 337:129–171.
27. Foo, J. H., et al. 2010. Both the p33 and p55 subunits of the Helicobacterpylori VacA toxin are targeted to mammalian mitochondria. J. Mol. Biol.401:792–798.
28. Fujikawa, A., et al. 2003. Mice deficient in protein tyrosine phosphatasereceptor type Z are resistant to gastric ulcer induction by VacA of Helico-bacter pylori. Nat. Genet. 33:375–381.
29. Galluzzi, L., et al. 2009. Guidelines for the use and interpretation of assaysfor monitoring cell death in higher eukaryotes. Cell Death Differ. 16:1093–1107.
30. Galmiche, A., et al. 2000. The N-terminal 34 kDa fragment of Helicobacterpylori vacuolating cytotoxin targets mitochondria and induces cytochrome crelease. EMBO J. 19:6361–6370.
31. Gauthier, N. C., et al. 2005. Helicobacter pylori VacA cytotoxin: a probe fora clathrin-independent and Cdc42-dependent pinocytic pathway routed tolate endosomes. Mol. Biol. Cell 16:4852–4866.
32. Gold, R., et al. 1994. Differentiation between cellular apoptosis and necrosisby the combined use of in situ tailing and nick translation techniques. Lab.Invest. 71:219–225.
33. Golstein, P., and G. Kroemer. 2007. Cell death by necrosis: towards a mo-lecular definition. Trends Biochem. Sci. 32:37–43.
34. Grusch, M., et al. 2002. Maintenance of ATP favours apoptosis over necrosistriggered by benzamide riboside. Cell Death Differ. 9:169–178.
35. Gupta, V. R., et al. 2008. Sphingomyelin functions as a novel receptor forHelicobacter pylori VacA. PLoS Pathog. 4:e1000073.
36. Ha, H. C., and S. H. Snyder. 1999. Poly(ADP-ribose) polymerase is a me-diator of necrotic cell death by ATP depletion. Proc. Natl. Acad. Sci. U. S. A.96:13978–13982.
37. Jonas, D., B. Schultheis, C. Klas, P. H. Krammer, and S. Bhakdi. 1993.Cytocidal effects of Escherichia coli hemolysin on human T lymphocytes.Infect. Immun. 61:1715–1721.
38. Kennedy, C. L., D. J. Smith, D. Lyras, A. Chakravorty, and J. I. Rood. 2009.Programmed cellular necrosis mediated by the pore-forming alpha-toxinfrom Clostridium septicum. PLoS Pathog. 5:e1000516.
39. Kimura, M., et al. 1999. Vacuolating cytotoxin purified from Helicobacterpylori causes mitochondrial damage in human gastric cells. Microb. Pathog.26:45–52.
40. Kroemer, G., B. Dallaporta, and M. Resche-Rigon. 1998. The mitochondrialdeath/life regulator in apoptosis and necrosis. Annu. Rev. Physiol. 60:619–642.
41. Kroemer, G., L. Galluzzi, and C. Brenner. 2007. Mitochondrial membranepermeabilization in cell death. Physiol. Rev. 87:99–163.
42. Kroemer, G., et al. 2009. Classification of cell death: recommendations of theNomenclature Committee on Cell Death 2009. Cell Death Differ. 16:3–11.
43. Kroemer, G., N. Zamzami, and S. A. Susin. 1997. Mitochondrial control ofapoptosis. Immunol. Today 18:44–51.
44. Kuck, D., et al. 2001. Vacuolating cytotoxin of Helicobacter pylori inducesapoptosis in the human gastric epithelial cell line AGS. Infect. Immun.69:5080–5087.
45. Lamkanfi, M., and V. M. Dixit. 2010. Manipulation of host cell death path-ways during microbial infections. Cell Host Microbe 8:44–54.
46. Lazebnik, Y. A., S. H. Kaufmann, S. Desnoyers, G. G. Poirier, and W. C.Earnshaw. 1994. Cleavage of poly(ADP-ribose) polymerase by a proteinasewith properties like ICE. Nature 371:346–347.
47. Leist, M., B. Single, A. F. Castoldi, S. Kuhnle, and P. Nicotera. 1997.Intracellular adenosine triphosphate (ATP) concentration: a switch in thedecision between apoptosis and necrosis. J. Exp. Med. 185:1481–1486.
48. Maeda, S., et al. 2007. Essential roles of high-mobility group box 1 in thedevelopment of murine colitis and colitis-associated cancer. Biochem. Bio-phys. Res. Commun. 360:394–400.
49. McClain, M. S., and T. L. Cover. 2007. Functional analysis of neutralizingantibodies against Clostridium perfringens epsilon-toxin. Infect. Immun. 75:1785–1793.
50. McClain, M. S., et al. 2003. Essential role of a GXXXG motif for membranechannel formation by Helicobacter pylori vacuolating toxin. J. Biol. Chem.278:12101–12108.
51. Messmer, D., et al. 2004. High mobility group box protein 1: an endogenoussignal for dendritic cell maturation and Th1 polarization. J. Immunol. 173:307–313.
52. Moubarak, R. S., et al. 2007. Sequential activation of poly(ADP-ribose)polymerase 1, calpains, and Bax is essential in apoptosis-inducing factor-mediated programmed necrosis. Mol. Cell. Biol. 27:4844–4862.
53. Nakayama, M., et al. 2006. Clustering of Helicobacter pylori VacA in lipidrafts, mediated by its receptor, receptor-like protein tyrosine phosphatasebeta, is required for intoxication in AZ-521 cells. Infect. Immun. 74:6571–6580.
54. Oldani, A., et al. 2009. Helicobacter pylori counteracts the apoptotic action ofits VacA toxin by injecting the CagA protein into gastric epithelial cells.PLoS Pathog. 5:e1000603.
55. Park, J. S., et al. 2004. Involvement of Toll-like receptors 2 and 4 in cellularactivation by high mobility group box 1 protein. J. Biol. Chem. 279:7370–7377.
56. Pelish, T. M., and M. S. McClain. 2009. Dominant-negative inhibitors of theClostridium perfringens epsilon-toxin. J. Biol. Chem. 284:29446–29453.
57. Raucci, A., R. Palumbo, and M. E. Bianchi. 2007. HMGB1: a signal ofnecrosis. Autoimmunity 40:285–289.
58. Rudel, T., O. Kepp, and V. Kozjak-Pavlovic. 2010. Interactions betweenbacterial pathogens and mitochondrial cell death pathways. Nat. Rev. Mi-crobiol. 8:693–705.
59. Savill, J., I. Dransfield, C. Gregory, and C. Haslett. 2002. A blast from thepast: clearance of apoptotic cells regulates immune responses. Nat. Rev.Immunol. 2:965–975.
60. Scaffidi, P., T. Misteli, and M. E. Bianchi. 2002. Release of chromatinprotein HMGB1 by necrotic cells triggers inflammation. Nature 418:191–195.
61. Scovassi, A. I., and G. G. Poirier. 1999. Poly(ADP-ribosylation) and apop-tosis. Mol. Cell. Biochem. 199:125–137.
62. Sewald, X., et al. 2008. Integrin subunit CD18 is the T-lymphocyte receptorfor the Helicobacter pylori vacuolating cytotoxin. Cell Host Microbe 3:20–29.
63. Suerbaum, S., and P. Michetti. 2002. Helicobacter pylori infection. N. Engl.J. Med. 347:1175–1186.
64. Szabo, I., et al. 1999. Formation of anion-selective channels in the cellplasma membrane by the toxin VacA of Helicobacter pylori is required for itsbiological activity. EMBO J. 18:5517–5527.
65. Telford, J. L., et al. 1994. Gene structure of the Helicobacter pylori cytotoxinand evidence of its key role in gastric disease. J. Exp. Med. 179:1653–1658.
66. Terebiznik, M. R., et al. 2009. Effect of Helicobacter pylori’s vacuolatingcytotoxin on the autophagy pathway in gastric epithelial cells. Autophagy5:370–379.
67. Thornberry, N. A. 1998. Caspases: key mediators of apoptosis. Chem. Biol.5:R97–R103.
68. Vinion-Dubiel, A. D., et al. 1999. A dominant negative mutant of Helicobacterpylori vacuolating toxin (VacA) inhibits VacA-induced cell vacuolation.J. Biol. Chem. 274:37736–37742.
69. Waterhouse, N. J., et al. 2001. Cytochrome c maintains mitochondrial trans-membrane potential and ATP generation after outer mitochondrial mem-brane permeabilization during the apoptotic process. J. Cell Biol. 153:319–328.
70. Willhite, D. C., and S. R. Blanke. 2004. Helicobacter pylori vacuolating cyto-toxin enters cells, localizes to the mitochondria, and induces mitochondrialmembrane permeability changes correlated to toxin channel activity. Cell.Microbiol. 6:143–154.
71. Willhite, D. C., T. L. Cover, and S. R. Blanke. 2003. Cellular vacuolation andmitochondrial cytochrome c release are independent outcomes of Helico-bacter pylori vacuolating cytotoxin activity that are each dependent on mem-brane channel formation. J. Biol. Chem. 278:48204–48209.
72. Xiang, J., D. T. Chao, and S. J. Korsmeyer. 1996. BAX-induced cell deathmay not require interleukin 1 beta-converting enzyme-like proteases. Proc.Natl. Acad. Sci. U. S. A. 93:14559–14563.
73. Yahiro, K., et al. 1999. Activation of Helicobacter pylori VacA toxin byalkaline or acid conditions increases its binding to a 250-kDa receptor pro-tein-tyrosine phosphatase beta. J. Biol. Chem. 274:36693–36699.
74. Yahiro, K., et al. 2003. Protein-tyrosine phosphatase alpha, RPTP alpha, isa Helicobacter pylori VacA receptor. J. Biol. Chem. 278:19183–19189.
75. Yamasaki, E., et al. 2006. Helicobacter pylori vacuolating cytotoxin inducesactivation of the proapoptotic proteins Bax and Bak, leading to cytochromec release and cell death, independent of vacuolation. J. Biol. Chem. 281:11250–11259.
76. Yang, H., H. Wang, C. J. Czura, and K. J. Tracey. 2002. HMGB1 as acytokine and therapeutic target. J. Endotoxin Res. 8:469–472.
77. Yu, S. W., et al. 2002. Mediation of poly(ADP-ribose) polymerase-1-depen-dent cell death by apoptosis-inducing factor. Science 297:259–263.
78. Zong, W. X., D. Ditsworth, D. E. Bauer, Z. Q. Wang, and C. B. Thompson.2004. Alkylating DNA damage stimulates a regulated form of necrotic celldeath. Genes Dev. 18:1272–1282.
Editor: B. A. McCormick
VOL. 79, 2011 H. PYLORI VacA AND PROGRAMMED NECROSIS 2543
