SGM Journals
MICROBIAL PATHOGENICITY

Type III secretion system 1 of Vibrio parahaemolyticus induces oncosis in both epithelial and monocytic cell lines

Xiaohui Zhou, Michael E. Konkel, Douglas R. Call

  • 1Department of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA, USA
  • 2School of Molecular Biosciences, Washington State University, Pullman, WA, USA
  • Correspondence
    Douglas R. Call
    drcall{at}wsu.edu

    • Microbiology 2009; 155(3):837–851 · https://doi.org/10.1099/mic.0.024919-0

    View at publisher PubMed

    Abstract

    The Vibrio parahaemolyticus type III secretion system 1 (T3SS1) induces cytotoxicity in mammalian epithelial cells. We characterized the cell death phenotype in both epithelial (HeLa) and monocytic (U937) cell lines following infection with V. parahaemolyticus. Using a combination of the wild-type strain and gene knockouts, we confirmed that V. parahaemolyticus strain NY-4 was able to induce cell death in both cell lines via a T3SS1-dependent mechanism. Bacterial contact, but not internalization, was required for T3SS1-induced cytotoxicity. The mechanism of cell death involves formation of a pore structure on the surface of infected HeLa and U937 cells, as demonstrated by cellular swelling, uptake of cell membrane-impermeable dye and protection of cytotoxicity by osmoprotectant (PEG3350). Western blot analysis showed that poly ADP ribose polymerase (PARP) was not cleaved and remained in its full-length active form. This result was evident for seven different V. parahaemolyticus strains. V. parahaemolyticus-induced cytotoxicity was not inhibited by addition of the pan-caspase inhibitor carbobenzoxy-valyl-alanyl-aspartyl-[O-methyl]-fluoromethylketone (Z-VAD-FMK) or the caspase-1 inhibitor N-acetyl-tyrosyl-valyl-alanyl-aspartyl-aldehyde (Ac-YVAD-CHO); thus, caspases were not involved in T3SS1-induced cytotoxicity. DNA fragmentation was not evident following infection and autophagic vacuoles were not observed after monodansylcadaverine staining. We conclude that T3SS1 of V. parahaemolyticus strain NY-4 induces a host cell death primarily via oncosis rather than apoptosis, pyroptosis or autophagy.

    • Two supplementary tables listing strains of Vibrio parahaemolyticus and PCR primers used in this study are available with the online version of this paper.

    Edited by: D. L. Gally

    INTRODUCTION

    Vibrio parahaemolyticus is a Gram-negative food-borne pathogen that is commonly associated with consumption of raw or undercooked seafood (Joseph et al., 1982). Symptoms of infection include diarrhoea, nausea, vomiting, headache, fever and chills. In addition to typical gastrointestinal infections, approximately 5 % of V. parahaemolyticus infections advance to septicaemia (Hlady & Klontz, 1996) and these infections may be fatal, especially in immunocompromised patients or those with a preexisting medical condition such as liver disease or diabetes (Yeung & Boor, 2004).

    Thermostable direct haemolysin (TDH) is considered the primary virulence factor in V. parahaemolyticus. In addition to its ability to form pores in red blood cells, TDH causes increased short circuit current, increased Cl− secretion in human colonic epithelial cells, and enhanced Ca2+ entry from the extracellular medium (Takahashi et al., 2001). An early study of a tdh deletion mutant demonstrated significantly reduced ability to cause fluid accumulation in ileal loops of a rabbit model (Nishibuchi et al., 1992). In contrast, Park et al. (2004) showed that a tdh deletion mutant has the ability to cause fluid accumulation in the ligated intestine of rabbits. Furthermore, both TDH-positive and -negative strains of V. parahaemolyticus can disrupt the epithelial tight junction, a precursor to dissemination of bacteria into the circulation (Lynch et al., 2005). These studies indicate that there are factors in addition to TDH that contribute to the pathogenesis of V. parahaemolyticus.

    During the early stage of infection, bacterial-induced epithelial cell death may contribute to the translocation of bacteria from the apical side to the basal side of the intestinal epithelium. Thereafter, some bacteria have adapted a pathogenic mechanism whereby they can highjack the resident macrophages in the Peyer's patches, allowing dissemination of bacteria via the circulatory system. For example, Listeria monocytogenes and Legionella pneumophila reside intracellularly in the macrophage, thereby avoiding an antimicrobial environment and promoting their dissemination throughout the host without being exposed to antibodies (Duclos & Desjardins, 2000; Pizarro-Cerda & Cossart, 2006; Swanson & Fernandez-Moreira, 2002). In contrast, Salmonella and Shigella can survive within macrophages as well as kill macrophages, depending on the physiological state of the macrophage (Navarre & Zychlinsky, 2000). Finally, Yersinia is also known to induce macrophage death (Hersh et al., 1999; Mills et al., 1997; Palmer et al., 1999).

    Apoptosis, pyroptosis, autophagy and oncosis are four distinct pathways used by bacteria to trigger host cell death (Fink & Cookson, 2005). The pathogenic outcomes of these four pathways are distinct, whereby cell death by an apoptotic or autophagy pathway occurs without significant tissue disruption and inflammation, while host cell death by oncosis and pyroptosis leads to secretion of inflammatory cytokines, such as interleukin (IL)-1β and IL-8, and subsequent recruitment of activated neutrophils to the site of infection (Navarre & Zychlinsky, 2000). Interestingly, the same species of bacteria may use more than one mechanism to induce the death of epithelial cells and macrophages. For example, Salmonella can induce apoptosis of macrophages in a type III secretion system (T3SS)-dependent manner (Chen et al., 1996). Salmonella can also cause pyroptosis of macrophages, which is characterized by the activation of caspase-1 (Brennan & Cookson, 2000). Salmonella can also induce autophagy of macrophages that features degradation of cellular components, vacuolization and slight chromatin condensation. Finally, T3SS (SPI) of Salmonella can form pores in the host cell membrane, leading to oncosis of infected host macrophages characterized by cellular swelling, increased membrane permeability, and depletion of cellular energy (Fink & Cookson, 2006).

    Recently, V. parahaemolyticus has been shown to induce cell death in HeLa cells by apoptosis, as evidenced by DNA fragmentation, and this apoptosis is dependent on T3SS1 (Ono et al., 2006). Nevertheless, the clinical manifestation of vibriosis is consistent with an inflammatory response, including neutrophil recruitment to the site of intestinal infection (Qadri et al., 2003); this inflammatory response is more consistent with oncosis or pyroptosis than apoptosis or autophagy. Furthermore, Burdette et al. (2008) have recently reported that V. parahaemolyticus induces autophagy in cell culture, but blocking this pathway does not block cell death. Thus, it is very likely that alternative pathways are primarily responsible for cell death.

    In this report we further characterized V. parahaemolyticus-induced cell death in both epithelial (HeLa) and monocytic (U937) cells, and evaluated invasiveness in these cell types. We confirmed that V. parahaemolyticus causes cytotoxicity to both HeLa and U937 cells in a T3SS1-dependent manner, but that cell death occurred by oncosis, as demonstrated by morphological and biochemical analysis; this process does not require intracellular V. parahaemolyticus, although the process appears to be contact-dependent.

    METHODS

    Bacterial strains and growth conditions.

    Wild-type V. parahaemolyticus strains used are listed in Supplementary Table S1. Escherichia coli S-17 was used in the gene deletion experiments. V. parahaemolyticus was cultured in Luria–Bertani (LB) medium supplemented with 2.5 % sodium chloride. E. coli was cultured in LB medium.

    Cell lines.

    HeLa cells (CCL-2) and U937 cells (CRL-1593.2) were purchased from the American Type Culture Collection. HeLa cells were maintained as monolayer in Dulbecco's modified Eagle's medium (DMEM; Invitrogen) supplemented with 10 % fetal calf serum (FCS) at 37 °C. HeLa cells were passaged every 1–2 days and were washed once before inoculation. U937 cells were maintained with RPMI 1640 (ATCC 30-2001) supplemented with 10 % FCS at 37 °C in a humidified 5 % (v/v) CO2 incubator and were passaged every 2–3 days. Cells were resuspended in RPMI 1640 culture medium supplemented with phorbol 12-myristate 13-acetate (PMA; Sigma) at a final concentration of 10 ng ml−1 to induce differentiation (Tsuchiya et al., 1982). Cells were dispensed into 12-well tissue-culture plates at a density of approximately 1×106 cells per well and incubated for 24 h in a 37 °C humidified 5 % (v/v) CO2 incubator. After incubation, PMA-containing medium was aspirated and adherent (differentiated) cells were washed once with culture medium before inoculation.

    Secretion of Vp1656 by V. parahaemolyticus.

    Protein secretion was determined as described previously (Zhou et al., 2008). Briefly, to collect extracellular proteins, an overnight culture was diluted (1 : 100) in LB medium supplemented with 2.5 % NaCl and grown for 2 h at 37 °C with shaking. Expression of ExsA was induced by the addition of IPTG to a final concentration of 1 mM. Bacterial culture (15 ml) was centrifuged at ∼3000 g for 15 min, and the supernatant for each sample was passed through a 0.22 μm pore-size syringe filter to exclude the remaining bacteria within the supernatant. Proteins in the supernatant were precipitated by adding TCA (10 %, v/v). To collect the total proteins for detection of Vp1656, pellets from each sample (15 ml) were resuspended in PBS (pH 7.4, 1 ml) and sonicated for 1 min to lyse the bacteria and shear the chromosomal DNA. Supernatant and pellet samples were loaded onto a 12 % SDS-PAGE gel and Western blot analysis was performed with polyclonal anti-Vp1656 antibody (Zhou et al., 2008).

    Infection and lactate dehydrogenase (LDH) assay.

    V. parahaemolyticus was harvested from overnight culture and pelleted by centrifugation at 10 000 g at room temperature. The bacterial pellets were resuspended in DMEM containing 1 % (v/v) FCS. Bacterial suspensions (10 μl) were added to each well, containing approximately 106 host cells, to achieve an m.o.i. of 100 c.f.u. per cell. Plates were centrifuged at 1000 g to synchronize bacteria–host cell contact. In experiments using caspase inhibitors, cells were incubated with the pan-caspase inhibitor carbobenzoxy-valyl-alanyl-aspartyl-[O-methyl]-fluoromethylketone (Z-VAD-FMK; 100 μM; R&D Systems) or the caspase-1 inhibitor N-acetyl-tyrosyl-valyl-alanyl-aspartyl-aldehyde (Ac-YVAD-CHO; 20 μM) (CalBiochem) for 1 h prior to inoculation or treatment with staurosporine (STS; Sigma; 2 μM). To inhibit bacterial internalization, cells were treated with cytochalasin D (CalBiochem) (1 μg ml−1) 1 h prior to inoculation. For the LDH release assay, the medium was free of antibiotics before inoculation. Supernatant was collected at different time points after infection and LDH activity was measured with a Cytotoxicity Detection kit (Promega) according to the manufacturer's instructions. Maximum release was achieved by lysis of cells with the lysis buffer provided in the kit. Spontaneous LDH release in the supernatant of uninfected cells was also measured as described above. Percentage cytotoxicity was calculated using the formula:

    Percentage cytotoxicity=(test LDH release−spontaneous release)/(maximal release−spontaneous release)

    Transmission electron microscopy.

    For transmission electron microscopy, V. parahaemolyticus strains and U937 cells were co-incubated at 37 °C for 3 h within 12-well cell-culture plates. Infected U937 cells were scraped from the well, transferred to 1.5 ml microfuge tubes, and fixed overnight at 4 °C in 2 % glutaraldehyde and 2 % paraformaldehyde. Samples were then rinsed in cacodylate buffer, incubated overnight at 4 °C in 1 % osmium tetroxide, and incubated for 1 h in 1 % tannic acid at room temperature. The samples were dehydrated, embedded in Spurr's solution, sectioned to 70–100 nm thin sections, and post-stained using 1 % uranyl acetate and lead citrate. Sections were examined with a transmission electron microscope at the Washington State University Electron Microscopy Center. All the reagents were provided by the Electron Microscopy Center.

    Fluorescence microscopy.

    Pore formation was detected by fluorescence microscopy as described elsewhere (Kirby et al., 1998). Briefly, HeLa and U937 cell monolayers were grown on coverslips before inoculation with strains of V. parahaemolyticus and after 3 h of infection coverslips were covered with PBS containing 25 μg ml−1 ethidium bromide and 5 μg ml−1 acridine orange (AO). Coverslips were observed using FITC (543 nm) and rhodamine (488 nm) filters with a Zeiss confocal microscope. Autophagic vacuoles were detected by monodansylcadaverine (MDC; Sigma) labelling and fluorescence microscopy as described elsewhere (Munafo & Colombo, 2001). Briefly, HeLa cells seeded on the coverslips in a six-well plate overnight to achieve a monolayer were infected with V. parahaemolyticus strains for 3 h. The infected HeLa cells were rinsed twice with PBS and incubated with 0.05 mM MDC in PBS at 37 °C for 10 min. After incubation, cells were washed four times with PBS and coverslips were analysed using FITC (543 nm) filters with a Zeiss confocal microscope. For positive and negative controls, HeLa cells were treated with rapamycin (2 μg ml−1; Fisher) for 24 h or left untreated, and observed after MDC labelling.

    Deletion mutants.

    Deletion mutants were generated by homologous recombination. A chromosomal deletion in the vcrD (vp1662) gene was constructed by allelic exchange using a suicide vector, pDM4, carrying DNA fragments corresponding to vcrD flanking regions (Milton et al., 1996). PCR-amplified DNA fragments used for constructing an in-frame deletion mutation of vcrD were generated by overlap PCR. Two DNA fragments were amplified by PCR with V. parahaemolyticus strain NY-4 chromosomal DNA as a template with primer pairs vcrD-1F and vcrD-1R and vcrD-2F and vcrD-2R, respectively (Supplementary Table S2). Primers vcrD-1R and vcrD-2F were complementary to each other. These two fragments were gel-purified and used as templates in a second PCR with primers vcrD-1F and vcrD-2R, resulting in the construction of a fragment with a deletion of 1682 bp in the vcrD gene. The fragment containing the deletion was purified and digested with XhoI and XbaI, and then cloned into a suicide vector, pDM4, that had been pretreated with the same enzymes. This vector was then transferred to NY-4 by conjugation and both ampicillin- and chloramphenicol-resistant transconjugants were selected. Ampicillin (100 μg ml−1) was used to select against E. coli and chloramphenicol (5 μg ml−1) was used to select transconjugants. To complete the allelic exchange, the integrated suicide plasmid was induced to excise from the chromosome by growth on LB medium containing sucrose (5 %). Chloramphenicol-sensitive clones were then obtained and screened by PCR with primers vcrD-up and vcrD-down, which lie adjacent to the flanking regions cloned into the plasmid (Supplementary Table S2). One clone with a vcrD gene deletion hereafter designated NY-4v1 was selected. Construction of an escV deletion mutant was performed in the same manner by using the following primers: escV1F, escV1R, escV2F, escV2R, escV-up and escV-down (Supplementary Table S2). One clone, designated NY-4e1, was selected. VcrD/escV deletion mutants were generated by constructing an escV deletion mutant in strain NY-4v1 to produce strain NY-4ve1.

    Detection of poly ADP ribose polymerase (PARP) cleavage by immunoblotting.

    Infected, uninfected, STS (2 μM) or t-butylhydroperoxide (TBH; 200 μM)-treated HeLa cells and U937 cells were harvested and lysed in radioimmunoprecipitation assay (RIPA) buffer, consisting of 25 mM Tris/HCl, pH 7.6, 150 mM NaCl, 1 % Triton X-100, 1 % sodium deoxycholate and 0.1 % SDS. Samples (100 μl) were dissolved in 100 μl SDS sample buffer containing 2.3 % (w/v) SDS, 10 % (w/v) glycerol, 5 % (v/v) 2-mercaptoethanol and 10 μg bromophenol blue ml−1 in 125 mM Tris/HCl, pH 6.8, and heated at 100 °C for 5 min and loaded on a 10 % (w/v) polyacrylamide gel. After electrophoresis, proteins were electrophoretically transferred to a nitrocellulose membrane for 12 h. The membrane was blocked with 5 % skimmed milk in Tris-buffered saline (20 mM Tris, 137 mM NaCl, pH 7.6) containing 0.1 % Tween 20 and probed with monoclonal anti-PARP antibody C-20 (Invitrogen) for 1 h at room temperature. The secondary antibody was an anti-mouse immunoglobulin conjugated to horseradish peroxidase. The blots were developed by using the Western Blot kit (Bio-Rad) according to the manufacturer's instructions.

    Osmoprotection assay.

    HeLa cells and U937 cells were infected with strains of V. parahaemolyticus or treated with TBH in the presence of 30 mM of various osmoprotectants. Osmoprotectants included sucrose (diameter=0.9 nm), PEG1450 (diameter=4 nm), and PEG3350 (diameter=3.8 nm). All these chemicals were purchased from Sigma. LDH release was measured after 3 h of infection and experiments were performed at least three times.

    Gentamicin protection assay.

    Invasion of HeLa or U937 cells was assessed by a gentamicin protection assay. Overnight V. parahaemolyticus culture (10 μl; ∼108 bacterial cells) was resuspended in DMEM and added to each well containing HeLa or U937 monolayers (∼106 cells per well). V. parahaemolyticus-infected monolayers were incubated at 37 °C for 1 h, after which the DMEM was aspirated and wells were gently washed once with PBS buffer. DMEM containing 200 μg gentamicin ml−1 was added to each well for 1 h. Medium was aspirated and cells were washed with PBS two times, and then cell monolayers were lysed by forcibly pulling the culture through a 27 gauge needle 10 times. Intracellular c.f.u. of V. parahaemolyticus was determined by serial plating on LB agar with 3 % NaCl and expressed as total c.f.u. minus extracellular c.f.u.

    Statistical analysis.

    Student's t test was used to compare groups in osmoprotectant- and caspase inhibitor-treatment assays, and P<0.05 was considered statistically significant. Pearson's correlation coefficient (r) was used to examine relationships between intracellular c.f.u. and LDH release.

    RESULTS

    T3SS1 of V. parahaemolyticus induces rapid death of both HeLa and U937 cells

    Earlier studies have shown that T3SS1 induces cytotoxicity in epithelial cells after 4 h infection and that this process involves an apoptotic pathway (Ono et al., 2006; Park et al., 2004). In the present study, we examined T3SS1 of V. parahaemolyticus-induced cell death for both HeLa and U937 cells. We generated T3SS1 mutant (NY-4v1), T3SS2 mutant (NY-4e1) and T3SS1–T3SS2 double mutant (NY-4ve1) strains and determined the secretion of Vp1656 (a known T3SS1 secretion protein) by the wild-type and each mutant strain. Previous work had shown that T3SS1 genes are expressed when the T3SS1 positive regulator (ExsA) is overexpressed in V. parahaemolyticus (Zhou et al., 2008). To confirm that our knockout strains specifically blocked the T3SS1 secretion pathway, we transformed the wild-type and each mutant strain of V. parahaemolyticus with an exsA plasmid and determined whether they secreted a known T3SS1 secretion protein, Vp1656. Vp1656 was expressed in all strains when they were transformed with an exsA plasmid, pexsA (Fig. 1⇓, upper panel). However, Vp1656 was secreted only by wild-type and T3SS2 mutant strains, and was not secreted by the T3SS1 mutant and T3SS1–T3SS2 double mutant (Fig. 1⇓, lower panel). These results confirmed that T3SS-1-dependent secretion of Vp1656 was eliminated in the T3SS1 mutant and T3SS1–T3SS2 double mutant strains, while wild-type and T3SS2 mutant strains were competent for the secretion of Vp1656.

    Figure image not available in archive
    Fig. 1.

    Vp1656 is secreted by wild-type (NY-4) and T3SS2 mutant (NY-4e1) strains, but not by T3SS1 mutant (NY-4v1) and T3SS1–T3SS2 double mutant (NY-4ve1) V. parahaemolyticus strains. V. parahaemolyticus NY-4 : pexsA, NY-4v1 : pexsA, NY-4e1 : pexsA and NY-4ve1 : pexsA strains were grown in LB medium supplemented with 2.5 % NaCl for 2 h. Pellet (upper panel) and supernatant (lower panel) samples were loaded onto the SDS-PAGE gel and detected with polyclonal anti-Vp1656 antibody. ‘N’ in the upper panel indicates a non-specific protein (Zhou et al., 2008).

    We then infected HeLa and U937 cells with wild-type, T3SS1 mutant (NY-4v1), T3SS2 mutant (NY-4e1) and T3SS1–T3SS2 double mutant (NY-4ve1) strains. After 3 h of infection, the wild-type and T3SS2 mutant strains caused ∼95 and 58 % of HeLa cell death, respectively, as measured by an LDH-release assay (Fig. 2a⇓, white bars). In contrast, both the T3SS1 mutant and the T3SS1–T3SS2 double mutant strains killed <20 % of HeLa cells (Fig. 2a⇓). HeLa cells infected with wild-type and T3SS2 mutant strains exhibited rounding and appeared to be swelling (Fig. 2b⇓, d). In contrast, HeLa cells infected with the T3SS1 mutant or the T3SS1–T3SS2 double mutant strains maintained a uniform monolayer and normal cell appearance compared with uninfected cells (Fig. 2c⇓, e and f). In a similar manner, ∼80 and 70 % of U937 cells infected with wild-type and T3SS2 mutant strains were killed within 3 h, while T3SS1 mutant and T3SS1–T3SS2 double mutant strains induced 20 and 10 % cell death, respectively (Fig. 2a⇓, grey bars). Taken together, these results confirmed that V. parahaemolyticus induced cytotoxicity to both epithelial and monocytic cells in a T3SS1-dependent manner.

    Figure image not available in archive
    Fig. 2.

    V. parahaemolyticus induces rapid death of both HeLa and U937 cells in a T3SS1-dependent manner. (a) HeLa cells (white bars) and U937 cells (grey bars) were killed by the wild-type (NY-4) and T3SS2 mutant (NY-4e1) strains, but much lower cytotoxicity was observed for cells infected with the T3SS1 mutant (NY-4v1) and T3SS1–T3SS2 double mutant (NY-4ve1) strains (3 h infection). The results represent the mean±sd of three biological replicates. Morphological observations under a light microscope for (b) HeLa cells infected with the wild-type, (c) T3SS1 mutant, (d) T3SS2 mutant and (e) T3SS1–T3SS2 double mutant strains, and (f) uninfected HeLa cells. Bar, 100 μm; ×100 magnification.

    T3SS1-induced cytotoxicity is invasion-independent and cell contact-dependent

    Salmonella-induced cytotoxicity in macrophages is dependent on its invasive phenotype (Chen et al., 1996; Hersh et al., 1999); thus, we hypothesized that the cytotoxic differences between wild-type and T3SS1 mutant strains are due to different invasive ability. To test this hypothesis, we performed a gentamicin protection assay to measure the invasion rate of both wild-type and T3SS mutant strains in HeLa and U937 cells. The c.f.u. was similar for intracellular wild-type and T3SS mutant strains recovered from HeLa cells and U937 cells (Fig. 3a⇓). These results indicate that different levels of cytotoxicity caused by wild-type and T3SS mutant strains are not due to a differential invasive phenotype. To explore this further, we also measured the invasiveness and cytotoxicity for an additional 12 strains of V. parahaemolyticus and one strain of Salmonella typhimurium. Salmonella invasion was ∼10–100 times greater than that of V. parahaemolyticus. Strains NY-2 and B-4 showed the highest level of invasiveness; nevertheless, while strain NY-2 induced 80 % cell death for HeLa cells, strain B-4 induced only 20 % cell death 4 h post-infection (Fig. 3b⇓, c), indicating that the high cell death induction of strain B-4 was not related to its invasiveness. In contrast, strains NY-10, NY-11 and NY-12 induced high levels of cytotoxicity with low levels of invasiveness (Fig. 3b⇓, c), suggesting that the high cytotoxicity of these strains was not due to their invasive phenotypes. Overall, there was no correlation between recovery of intracellular c.f.u. and LDH release (r=−0.005).

    Figure image not available in archive
    Fig. 3.

    No direct relationship was observed between cytotoxicity and invasiveness. (a) Average c.f.u. of internalized bacteria (NY-4, NY-4v1, NY4e1 and NY-4ve1) recovered after co-culture with HeLa (white bars) and U937 (grey bars) cells for 1 h. (b) Average c.f.u. of internalized bacteria for 12 strains of V. parahaemolyticus and S. typhimurium LT2 recovered after co-culture with HeLa cells for 1 h. (c) LDH release by HeLa cells infected with 12 additional wild-type V. parahaemolyticus strains for 4 h. (d) Average c.f.u. of internalized bacteria (NY-4, NY-4v1, NY4e1 and NY-4ve1) recovered after co-culture with HeLa cells treated (white bars) or not treated (grey bars) with cytochalasin D. (e) LDH release by HeLa cells treated (grey bars) or untreated (white bars) with cytochalasin D prior to infection with wild-type V. parahaemolyticus. The results represent the mean±sd of three replicates.

    To further test if intracellular bacteria are necessary for cytotoxicity, we treated HeLa cells for 1 h with cytochalasin D, which inhibits actin polymerization and therefore bacterial invasion of mammalian cells, and then infected the cells with wild-type V. parahaemolyticus. As reported elsewhere (Akeda et al., 1997), bacterial c.f.u. recovered from HeLa cells treated with cytochalasin D was ∼1–2 % of that recovered from HeLa cells that were not treated with cytochalasin D (Fig. 3d⇑), indicating that cytochalasin D significantly inhibited bacterial internalization by HeLa cells. Nevertheless, cell killing was not inhibited by treatment with cytochalasin D (Fig. 3e⇑), confirming that bacterial invasiveness was not responsible for the cell killing by T3SS1. In addition, cytochalasin D treatment without infection did not cause significant LDH release (Fig. 3e⇑), indicating that cytochalasin D did not promote LDH release during these experiments.

    To determine whether cell contact is required for the T3SS1-induced cytotoxicity, we collected culture media from HeLa cells 4 h post-infection (m.o.i. 100 : 1) and centrifuged it to remove cells. The supernatant was filter-sterilized (0.22 μm pore-size) and gentamicin (10 μg ml−1) was added before adding the supernatant back (neat) to fresh HeLa cell monolayers. Supernatant from HeLa cells infected with wild-type, T3SS1 mutant, T3SS2 mutant and T3SS1–T3SS2 double mutant strains caused 5.6±0.7, 4.8±1.5, 5.5±1.5 and 6.5±1.5 % cell death, respectively. We concluded from this experiment that cell contact is required for V. parahaemolyticus to induce host cell death.

    Pore forming activity of T3SS1

    Salmonella uses a T3SS-dependent mechanism to form pores in the host cell membrane, thereby leading to osmotic lysis of host cells (Fink & Cookson, 2006). Based on the morphology (cell rounding and swelling) of infected HeLa cells as observed by light microscopy (Fig. 2b⇑, d), we hypothesized that T3SS1 of V. parahaemolyticus can form a pore structure on the host membrane, resulting in osmotic lysis. Infected cells were stained with a membrane-impermeable dye (ethidium bromide; EtBr) and a membrane-permeable dye (AO). When AO penetrates the cell, it exhibits green fluorescence, whereas the combination of AO and EtBr results in yellow fluorescence, thereby confirming membrane permeability. After 3 h of infection with wild-type and T3SS2 mutant strains, both HeLa cells and U937 cells appeared yellow under a fluorescence microscope, indicating that both EtBr and AO entered cells (Fig. 4a⇓, c, f and h). In contrast, uninfected cells and cells infected with the T3SS1 mutant and T3SS1–T3SS2 double mutant strains showed a green colour, indicating that only AO entered cells (Fig. 4b⇓, d, e, g, i and j). These results confirmed that the T3SS1 is involved in pore formation in the cell membrane, leading to the uptake of membrane-impermeable dye (EtBr) and osmotic lysis.

    Figure image not available in archive
    Fig. 4.

    Pore formation on the membranes of HeLa and U937 cells by T3SS1 of V. parahaemolyticus. HeLa cells (a–e) and U937 macrophages (f–j) adherent to glass coverslips were infected with wild-type strain (a, f), T3SS1 mutant strain (b, g), T3SS2 mutant strain (c, h) and T3SS1-T3SS2 double mutant strain (d, i) at an m.o.i. of 100 : 1 for 3 h. (e) and (j) show uninfected HeLa and U937 cells, respectively. Cells were stained with the membrane-permeable dye AO (green) and the membrane-non-permeable dye ethidium bromide (red). Adherent cells were visualized by confocal fluorescence microscopy. Bars, 50 μm.

    To estimate the size of the pores produced by T3SS1 of V. parahaemolyticus, we infected HeLa and U937 cells with the NY-4 strain in the presence of osmoprotectants of different molecular mass. The premise of this assay is to introduce increasingly larger osmoprotectants until one is encountered that provides protection by blocking membrane pores. LDH release was significantly reduced when HeLa and U937 cells were infected with NY-4 in the presence of PEG3350, while PEG1450 and sucrose had no protective effect (Fig. 5⇓). PEG1450 and PEG3350 have diameters of 2.4 and 3.8 nm, respectively, indicating that the pore-size produced by T3SS1 of V. parahaemolyticus is between 2.4 nm and 3.8 nm. LDH release of control U937 cells induced by TBH, an organic hydroperoxide that induces oxidative stress, resulting in ATP depletion and cell death (Nieminen et al., 1995), was not prevented by any of the osmoprotectants (Fig. 5⇓). Collectively, these data indicate that V. parahaemolyticus is causing host cell death by producing pores in HeLa and U937 cells via the T3SS1 pathway.

    Figure image not available in archive
    Fig. 5.

    Protection against HeLa and U937 cell death by osmoprotectants (PEG3350). LDH release for HeLa cells and U937 macrophages infected with the wild-type strain of V. parahaemolyticus or treated with TBH in the presence of osmoprotectants. The results represent the mean±sd of three biological replicates.

    Morphological and biochemical characterization of T3SS1-induced cell death

    After 3 h of infection, intracellular bacteria were evident within the U937 cells infected with wild-type and T3SS mutant strains (white arrows in Fig. 6f–i⇓), while uninfected U937 cells harboured no intracellular bacteria (Fig. 6j⇓). Intracellular bacteria were not distinguishable in the dying HeLa cells (Fig. 6a–e⇓), probably due to the lower invasion rate for HeLa cells compared with U937 cells (Fig. 3a⇑). One interesting finding was that U937 and HeLa cells infected with wild-type and T3SS2 mutant strains displayed severe cytoplasmic vacuolation (black arrows in Fig. 6a⇓, c, f and h) while cells infected with the T3SS1 or T3SS1–T3SS2 double mutant strain have no obvious changes (Fig. 6b⇓, d, g and i) compared with uninfected cells (Fig. 6e⇓, j). The membrane integrity of the cells infected with wild-type and T3SS2 mutant strains was disrupted, while there was no evidence of similar changes in the cells infected with T3SS1 and T3SS1–T3SS2 double mutant strains. Changes typical of apoptosis, such as chromatin condensation, membrane blebbing and apoptotic bodies, were not observed in U937 and HeLa cells infected with either wild-type or mutant strains of V. parahaemolyticus.

    Figure image not available in archive
    Fig. 6.

    V. parahaemolyticus-infected HeLa (a–e) and U937 (f–j) cells showed vacuolation and damage of membranes in a T3SS1-dependent manner. Transmission electron micrographs of (a, f) wild-type-infected, (b, g) T3SS1 mutant-infected, (c, h) T3SS2 mutant-infected, (d, i) T3SS1–T3SS2 double mutant-infected, and (e, j) uninfected U937 and HeLa cells. White arrows indicate putative intracellular bacteria and black arrows indicate vacuoles in the cytoplasm of both HeLa and U937 cells. Bars: (c, d, e, i, j), 5 μm; (a, b, f, g, h), 2 μm.

    PARP is a 116 kDa enzyme that is cleaved into an 89 kDa inactive form to conserve the energy required for apoptosis (Affar et al., 2001); thus, detection of the cleaved form of PARP is a hallmark of apoptotic cell death. In contrast, for the other three mechanisms of cell death (autophagy, oncosis and pyroptosis), PARP is maintained as a full-length protein and the active form depletes its substrate NAD and ATP, leading to cell death. As shown in Fig. 7(a, b)⇓ (lanes 1–4), PARP remained in its active uncleaved state in both wild-type and mutant infected HeLa cells (Fig. 7a⇓) and U937 cells (Fig. 7b⇓). In control cells treated with STS, a protein kinase inhibitor that induces apoptosis, PARP was clearly cleaved into an 89 kDa inactive form (Fig. 7a⇓, b, lanes 5). These results indicated that the apoptotic pathway leading to PARP cleavage does not function during strain NY-4-induced cell death for either U937 or HeLa cells, and therefore apoptosis does not occur when either HeLa or U937 cells are infected with strain NY-4. To determine whether this conclusion is applicable to other strains of V. parahaemolyticus, we infected HeLa cells with six different strains that have been shown to be cytotoxic to HeLa cells (Fig. 3c⇑) and then measured the cleavage of PARP. PARP remained in its active, uncleaved state regardless of the infecting strain (Fig. 7c⇓), indicating that the apoptotic pathway is not induced by any of the seven strains tested in this experiment.

    Figure image not available in archive
    Fig. 7.

    PARP was not cleaved in V. parahaemolyticus-infected HeLa and U937 cells. (a) HeLa and (b) U937 cells were infected with wild-type (NY-4; lane 1), T3SS1 mutant (lane 2), T3SS2 mutant (lane 3) and T3SS1–T3SS2 double mutant (lane 4) strains of V. parahaemolyticus. HeLa and U937 cells were treated with STS (lane 5), TBH (lane 6) or remained untreated (lane 7). (c) HeLa cells were infected with strain NY-1 (lane 3), NY-2 (lane 4), NY-10 (lane 5), NY-11 (lane 6), NY-12 (lane 7) or B-3 (lane 8). Lanes 1 and 2 are samples from HeLa cells treated with STS and untreated, respectively. Cells were lysed and separated by SDS-PAGE and immunoblotted with monoclonal anti-PARP antibody specific for both 116 kDa PARP and 89 kDa cleaved PARP.

    V. parahaemolyticus-induced host cell death is caspase-independent

    Because PARP is cleaved by activated caspase-3 and caspase-7 (Affar et al., 2001; Germain et al., 1999; Lazebnik et al., 1994; Tewari et al., 1995) and our results showed that PARP is not cleaved during cell death induced by V. parahaemolyticus (Fig. 7⇑), this indicated that caspase-3 and caspase-7 are not activated during the process of cell death. To further confirm that caspase is not involved in the V. parahaemolyticus-induced cell death, we treated HeLa cells with a pan-caspase inhibitor Z-VAD-FMK for 1 h prior to infection and measured LDH release; others have shown that at a concentration of 100 μM, Z-VAD-FMK completely inhibits Fas-mediated apoptosis (Kato et al., 2000). Results showed that the pan-caspase inhibitor significantly protected STS-treated cells from death (Fig. 8a⇓), indicating that Z-VAD-FMK inhibited activation of caspases in STS-induced apoptotic cells. The pan-caspase inhibitor did not prevent TBH-treated cells from dying (Fig. 8a⇓), indicating that, as expected, caspases are not involved in TBH-induced cell death. The pan-caspase inhibitor did not prevent cell death of HeLa cells infected with wild-type and escV mutant strains. Thus, caspases are not activated and are not involved in V. parahaemolyticus-induced cytotoxicity. DNA fragmentation analysis showed that neither wild-type nor mutant strains of V. parahaemolyticus induced DNA fragmentation in HeLa cells, while the apoptosis inducer STS induced significant DNA fragmentation (Fig. 8b⇓). These results confirmed that apoptosis is not involved in the host cell death caused by T3SS1 of V. parahaemolyticus.

    Figure image not available in archive
    Fig. 8.

    (a) Effect of the pan-caspase inhibitor Z-VAD-FMK on wild-type and T3SS2 mutant strain-induced cytotoxicity. HeLa cells were preincubated for 1 h with 100 μM Z-VAD-FMK and then infected with the wild-type and T3SS2 mutant strains prior to the measurement of LDH release. STS- and TBH-treated cells were included as positive and negative controls, respectively. The results represent the mean±sd of three biological replicates. (b) DNA fragmentation analysis showing that V. parahaemolyticus does not induce apoptosis. HeLa cells were infected with the wild-type and each mutant strain or treated with STS for 3 h, and genomic DNA was isolated before electrophoresis on 1.5 % agarose gels.

    To verify that pyroptosis does not play a role in V. parahaemolyticus-induced cell death, we treated U937 cells with caspase-1 inhibitor 1 h prior to infection and measured LDH release. Salmonella-induced cell death was significantly inhibited by treatment with caspase-1 inhibitor (Fig. 9⇓), consistent with an earlier report that Salmonella induces caspase-1-dependent pyroptosis (Fink & Cookson, 2006). Cell death induced by wild-type and T3SS2 mutant strains of V. parahaemolyticus was not prevented (Fig. 9⇓), indicating that caspase-1 is not involved in V. parahaemolyticus-induced cytotoxicity.

    Figure image not available in archive
    Fig. 9.

    Effect of caspase-1 inhibitor Ac-YVAD-CHO on cytotoxicity induced by wild-type and T3SS2 mutant strains of V. parahaemolyticus and Salmonella. U937 cells were preincubated for 1 h with 20 μM Ac-YVAD-CHO and then infected with the wild-type or T3SS2 mutant strain, or S. typhimurium prior to the measurement of LDH release. LDH release was determined 8 or 4 h after U937 cells were infected with Salmonella or V. parahaemolyticus strains, respectively. The results represent the mean±sd of three biological replicates.

    V. parahaemolyticus does not induce autophagy in HeLa cells

    Wild-type and T3SS2 mutant strains induced vacuoles in both HeLa and U937 cells (Fig. 7⇑). To determine whether these vacuoles were autophagic vacuoles, infected cells were labelled with an autofluorescent drug, MDC, a selective marker for autophagic vacuoles (Biederbick et al., 1995). MDC-labelled vesicles were not evident in HeLa cells infected with wild-type (Fig. 10a⇓), T3SS1 mutant (Fig. 10b⇓), T3SS2 mutant (Fig. 10c⇓) and T3SS1–T3SS2 double mutant (Fig. 10d⇓) strains. As a negative control, uninfected HeLa cells did not show any MDC-labelled vesicles (Fig. 10e⇓). As a positive control, HeLa cells were treated with rapamycin, a known autophagy inducer (Iwai-Kanai et al., 2008), for 24 h and then were stained with MDC. As shown in Fig. 10(f)⇓, the distinct dot-structures with punctuate MDC fluorescence appeared in the cytoplasm or in the perinuclear regions (white arrows). These results indicated that V. parahaemolyticus does not induce autophagy in HeLa cells.

    Figure image not available in archive
    Fig. 10.

    Autophagy is not evident in HeLa cells infected with the wild-type or each mutant strain of V. parahaemolyticus. HeLa cells were infected with (a) wild-type, (b) T3SS1 mutant, (c) T3SS2 mutant and (d) T3SS1–T3SS2 double mutant strains of V. parahaemolyticus for 3 h, and then incubated with MDC for 10 min at 37 °C. (e) Untreated HeLa cells or (f) HeLa cells treated with rapamycin (2 μg ml−1) for 24 h were included as negative and positive controls, respectively. White arrows indicate MDC-labelled vacuoles in rapamycin-treated HeLa cells.

    DISCUSSION

    The genome of V. parahaemolyticus harbours two distinct T3SSs of which T3SS1 is similar to the Ysc secretion system in Yersinia, while T3SS2 is similar to Inv–Mxi–Spa secretion system in Salmonella and Shigella (Denecker et al., 2002). The Ysc secretion system is typically associated with cytotoxicity, while the Inv–Mxi–Spa secretion system usually contributes to host cell invasion (Denecker et al., 2002). Earlier studies have shown that one fully sequenced clinical strain of V. parahaemolyticus (RIMD2210633) can infect HeLa cells, which in turn display apoptotic changes including DNA fragmentation and the presence of phosphatidylsenine in the outer membrane leaflet of the plasma membrane. These apoptotic changes are dependent on T3SS1 (Ono et al., 2006; Park et al., 2004).

    We describe herein another cytopathic effect that differs from apoptosis and that is induced by T3SS1 of strain NY-4 of V. parahaemolyticus in both HeLa and U937 cell lines. The cytopathic effect is manifested by LDH release, cellular swelling, pore formation on the cell membrane, and cytoplasmic vacuolation. In addition, infected cells retained PARP in its active, uncleaved state and DNA fragmentation did not occur. These results indicated that the NY-4 strain of V. parahaemolyticus does not induce epithelial and monocytic cell death via a mechanism of apoptosis. In addition, a panel of six other strains that are cytotoxic to both HeLa cells and U937 cells did not cause PARP cleavage (Fig. 7c⇑), indicating that these strains also did not cause apoptosis. It is unclear why our results contradicted earlier results, although this may reflect different phenotypes associated with different strains, thereby illustrating phenotypic and genotypic diversity attributable to V. parahaemolyticus. For example, it is possible that the wild-type strain (NY-4) and the other six strains used in the present study caused rapid cell death (within 4 h) by unrecognized effector protein(s) of T3SS1 that are absent in the strain used in previous studies (RIMD2210633; unavailable for the present study) and that this rapid cell death prevented detection of apoptosis in our study. Burdette et al. (2008) also reported that host cell death induced by V. parahaemolyticus is independent of apoptosis.

    T3SSs are needle-like structures expressed on the surface of Gram-negative bacteria and upon contact with eukaryotic cells (Cornelis & Van Gijsegem, 2000; Cornelis, 2002, 2006; Hueck, 1998; Mota & Cornelis, 2005). These structures become inserted into the cell membrane to create pores in the membrane surface (Neyt & Cornelis, 1999); these pores assist the translocation of effector proteins from bacteria into host cells (Neyt & Cornelis, 1999). In some cases, pore formation results in the entry of small ions and water from the extracellular milieu into the cytosol of host cells leading to cellular oncotic lysis (Dacheux et al., 2001; Fink & Cookson, 2006; Sun et al., 2005). For example, T3SSs of Pseudomonas, Salmonella and Burkholderia induce oncosis of macrophages that can be prevented by addition of osmoprotectants such as PEG with a molecular diameter larger than that of the pores. Our results showed that wild-type V. parahaemolyticus and T3SS2 mutant strains, not T3SS1 and T3SS1–T3SS2 double mutant strains, induced host cell LDH release, cell rounding (Fig. 2a⇑, b, d), uptake of membrane-impermeable dye (Fig. 4a⇑, c, f, h) and disruption of the integrity of cell membrane (Fig. 6a⇑, c, f, h). These findings are consistent with cell death induced by oncosis; this phenotype was abrogated by addition of PEG3350, but not by osmoprotectants with a molecular diameter less than that of PEG3350, suggesting that the pore sizes on the host membrane are between 2.4 and 3.8 nm. This is larger than pores that are created by Pseudomonas (1.3–2.8 nm), Salmonella (1.1–2.4 nm) and Burkholderia (1.4–2.4 nm) (Dacheux et al., 2001; Fink & Cookson, 2006; Sun et al., 2005).

    We can exclude the possibility that V. parahaemolyticus induces an apoptotic cell death pathway, because PARP is maintained as a full-length protein (Fig. 7⇑), caspase is not involved in the process of cell death (Fig. 8a⇑) and chromosomal DNA fragmentation does not occur (Fig. 8b⇑). The hallmark of pyroptosis is the activation of caspase-1, and inhibition of caspase-1 reduces the host cell death (Fink & Cookson, 2005). Our results showed that inhibition of caspase-1 did not prevent host cell death induced by T3SS1 of V. parahaemolyticus (Fig. 9⇑), thus excluding the possibility of pyroptotic cell death.

    One interesting finding is that U937 and HeLa cells infected with wild-type and T3SS2 mutant strains displayed severe cytoplasmic vacuolation. The intracellular vacuolation is assumed to represent an early pathophysiological event leading to eventual cell death. This activity has been associated with a few toxins; the most extensively studied cytotoxin that induces vacuolation is VacA in Helicobacter. Cellular vacuolation induced by VacA is the result of alteration of intracellular vesicle trafficking leading to the merger of the late endosome and lysosome (Reyrat et al., 1999; Wada et al., 2004). ShlA, a secreted haemolysin of Serratia marcescens, causes vacuoles in several epithelial cell lines that eventually lead to cell lysis (Hertle et al., 1999; Hertle, 2000). ShlA-induced vacuoles, however, do not appear to be derived from late endosomes. Alpha toxin of Xenorhabdus triggers endoplasmic reticulum vacuolation by increasing monovalent cation permeability of the membrane, resulting in cell death by colloid-osmotic lysis (Ribeiro et al., 2003). An early in vivo study showed that V. parahaemolyticus-infected mice displayed congestion, oedema and vacuolation in mucosal and submucosal tissues (Hoashi et al., 1990). Our in vitro observations in cell culture systems corroborate this earlier finding and, in addition, we found that T3SS1 is involved in vacuole formation.

    It is unclear whether the vacuoles observed in the present study were induced by a specific mechanism or were a general consequence of host cell death in response to the V. parahaemolyticus infection. Autophagy is associated with vacuolization, although these vacuoles are usually double-membrane vesicles with sequestrated cytoplasmic materials (Fink & Cookson, 2005). We have no evidence to show that vacuoles induced by T3SS1 of V. parahaemolyticus have double membranes or contain any cytoplasmic materials (Fig. 6⇑). Furthermore, there were apparently no MDC-labelled autophagic vacuoles within the cytoplasm of HeLa cells infected with the wild-type and each mutant strain of V. parahaemolyticus compared with HeLa cells treated with rapamycin, a known autophagy inducer (Fig. 10⇑). Thus, we conclude that if autophagy is occurring, it is not likely to play a major role in host cell death.

    In contrast to our findings, Burdette et al. (2008) showed that V. parahaemolyticus is able to induce autophagy in a T3SS1-dependent manner. The authors observed (by electron microscopy) autophagic vesicles within the HeLa cells that were infected with the POR3 strain of V. parahaemolyticus (Burdette et al., 2008). Importantly, however, inhibition of autophagy by addition of wortmannin to the cells made no significant impact on host cell death. Thus, while autophagy may occur, the primary cause of host cell death is independent of this pathway. In support of this conclusion, Burdette et al. (2008) hypothesized that other signalling pathways are activated by T3SS effectors that in turn result in cell rounding and lysis. Collectively, our results indicate that oncosis is this mechanism, at least with respect to interactions with both U937 and HeLa cells.

    Oncosis caused by V. parahaemolyticus may lead to the release of pro-inflammatory cytokines that attract activated neutrophils, thereby leading to inflammatory diarrhoea. This is similar to the mechanism by which Salmonella and Shigella cause enteritis by inducing oncosis of macrophages, thereby leading to the release of pro-inflammatory cytokines such as IL-1β (Brennan & Cookson, 2000; Fernandez-Prada et al., 1997; Nonaka et al., 2003). Interestingly, both Salmonella and Shigella must be located intracellularly to induce the death of host cells, which is evident because cytochalasin D inhibits the internalization of bacteria and efficiently prevents cytotoxicity. Our results indicated that intracellular bacteria are not required for V. parahaemolyticus-induced cytotoxicity because cytochalasin D treatment does not prevent cell death, even though invasiveness is significantly reduced (Fig. 3d⇑, e). In addition, there was no significant difference in the invasiveness of wild-type and mutant strains, indicating that invasiveness is not responsible for the differing cytotoxicity of wild-type and mutant strains of V. parahaemolyticus. In this context, V. parahaemolyticus-induced cytotoxicity more closely resembles Yersinia-induced cytotoxcity, which occurs from an extracellular location (Mills et al., 1997). Our results also indicate that cell contact is necessary for V. parahaemolyticus to induce oncosis. It is likely that T3SS1 is activated in a cell-contact-dependent manner and that an effector protein(s) responsible for initiating oncosis is translocated directly into host cells without being released to the extracellular medium. Further work is needed to identify the effector protein(s) involved in this process and to confirm that such proteins are translocated into the host cell.

    It is unexpected that neither T3SS1 nor T3SS2 played a role in V. parahaemolyticus invasiveness (Fig. 3a⇑), because T3SS2 belongs to the Inv–Mxi–Spa secretion system that can translocate effector proteins into host cells and induce actin rearrangement leading to bacterial internalization (Handa et al., 2007; Hansen-Wester & Hensel, 2001; Jepson et al., 2003; Pizarro-Cerda & Cossart, 2006; Schlumberger & Hardt, 2006). A possible reason for this discrepancy is that the strain used in our study lacks the T3SS2 substrates responsible for efficient invasion; our results showed that invasiveness of NY-4 is less than that of other strains of V. parahaemolyticus (e.g. NY-2 and B-4; Fig. 3b⇑) and considerably less than that of S. typhimurium.

    Oncosis-induced inflammatory responses result in the breach of tight junctions (Lynch et al., 2005) and lead to the transport of V. parahaemolyticus from the lumen side to the basolateral side of the intestine. To survive within Peyer's patches and disseminate to internal organs, V. parahaemolyticus must be able to survive phagocytosis by resident macrophages. Our results suggest that the number of intracellular V. parahaemolyticus within U937 cells (a macrophage-like cell line) decreased significantly from 1 to 24 h (data not shown), indicating that macrophages are able to kill intracellular V. parahaemolyticus. These results indicate that V. parahaemolyticus probably survives within the Peyer's patches by inducing the death of resident macrophages before ingestion, and that this precedes systemic infection (Daniels et al., 2000).

    In summary, we showed that infection of either epithelial or macrophage-like cell lines with strain NY-4 of V. parahaemolyticus results in death that is due to the rapid induction of oncosis. The mechanism of oncosis involves contact-dependent formation of pores in the host cell membrane and induction of cytoplasmic vacuoles, which are dependent on T3SS1, but not T3SS2. Apoptosis, pyroptosis and autophagy do not appear to play a significant role in this process.

    Acknowledgments

    We acknowledge excellent technical assistance from Stacey LaFrentz and Lisa Orfe, and services provided through Washington State University Electron Microscopy Center. Dr Asim Bej (University of Alabama at Birmingham) and Dr Kathryn J. Boor (Cornell University) provided wild-type strains of V. parahaemolyticus. This project has been funded in part with Federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services, under contract number NO1-AI-30055, and by the Agricultural Animal Health Program, College of Veterinary Medicine, Washington State University.

    References

    1. Affar, E. B., Germain, M., Winstall, E., Vodenicharov, M., Shah, R. G., Salvesen, G. S. & Poirier, G. G. (2001). Caspase-3-mediated processing of poly(ADP-ribose) glycohydrolase during apoptosis. J Biol Chem 276, 2935–2942.
    2. Akeda, Y., Nagayama, K., Yamamoto, K. & Honda, T. (1997). Invasive phenotype of Vibrio parahaemolyticus. J Infect Dis 176, 822–824.
    3. Bej, A. K., Patterson, D. P., Brasher, C. W., Vickery, M. C., Jones, D. D. & Kaysner, C. A. (1999). Detection of total and hemolysin-producing Vibrio parahaemolyticus in shellfish using multiplex PCR amplification of tl, tdh and trh. J Microbiol Methods 36, 215–225.
    4. Biederbick, A., Kern, H. F. & Elsasser, H. P. (1995). Monodansylcadaverine (MDC) is a specific in vivo marker for autophagic vacuoles. Eur J Cell Biol 66, 3–14.
    5. Brennan, M. A. & Cookson, B. T. (2000). Salmonella induces macrophage death by caspase-1-dependent necrosis. Mol Microbiol 38, 31–40.
    6. Burdette, D. L., Yarbrough, M. L., Orvedahl, A., Gilpin, C. J. & Orth, K. (2008). Vibrio parahaemolyticus orchestrates a multifaceted host cell infection by induction of autophagy, cell rounding, and then cell lysis. Proc Natl Acad Sci U S A 105, 12497–12502.
    7. Chen, L. M., Kaniga, K. & Galan, J. E. (1996). Salmonella spp. are cytotoxic for cultured macrophages. Mol Microbiol 21, 1101–1115.
    8. Cornelis, G. R. (2002). The Yersinia Ysc–Yop ‘type III’ weaponry. Nat Rev Mol Cell Biol 3, 742–752.
    9. Cornelis, G. R. (2006). The type III secretion injectisome. Nat Rev Microbiol 4, 811–825.
    10. Cornelis, G. R. & Van Gijsegem, F. (2000). Assembly and function of type III secretory systems. Annu Rev Microbiol 54, 735–774.
    11. Dacheux, D., Goure, J., Chabert, J., Usson, Y. & Attree, I. (2001). Pore-forming activity of type III system-secreted proteins leads to oncosis of Pseudomonas aeruginosa-infected macrophages. Mol Microbiol 40, 76–85.
    12. Daniels, N. A., MacKinnon, L., Bishop, R., Altekruse, S., Ray, B., Hammond, R. M., Thompson, S., Wilson, S., Bean, N. H. & other authors (2000). Vibrio parahaemolyticus infections in the United States, 1973–1998. J Infect Dis 181, 1661–1666.
    13. Denecker, G., Totemeyer, S., Mota, L. J., Troisfontaines, P., Lambermont, I., Youta, C., Stainier, I., Ackermann, M. & Cornelis, G. R. (2002). Effect of low- and high-virulence Yersinia enterocolitica strains on the inflammatory response of human umbilical vein endothelial cells. Infect Immun 70, 3510–3520.
    14. Duclos, S. & Desjardins, M. (2000). Subversion of a young phagosome: the survival strategies of intracellular pathogens. Cell Microbiol 2, 365–377.
    15. Fernandez-Prada, C. M., Hoover, D. L., Tall, B. D. & Venkatesan, M. M. (1997). Human monocyte-derived macrophages infected with virulent Shigella flexneri in vitro undergo a rapid cytolytic event similar to oncosis but not apoptosis. Infect Immun 65, 1486–1496.
    16. Fink, S. L. & Cookson, B. T. (2005). Apoptosis, pyroptosis, and necrosis: mechanistic description of dead and dying eukaryotic cells. Infect Immun 73, 1907–1916.
    17. Fink, S. L. & Cookson, B. T. (2006). Caspase-1-dependent pore formation during pyroptosis leads to osmotic lysis of infected host macrophages. Cell Microbiol 8, 1812–1825.
    18. Germain, M., Affar, E. B., D'Amours, D., Dixit, V. M., Salvesen, G. S. & Poirier, G. G. (1999). Cleavage of automodified poly(ADP-ribose) polymerase during apoptosis. Evidence for involvement of caspase-7. J Biol Chem 274, 28379–28384.
    19. Handa, Y., Suzuki, M., Ohya, K., Iwai, H., Ishijima, N., Koleske, A. J., Fukui, Y. & Sasakawa, C. (2007). Shigella IpgB1 promotes bacterial entry through the ELMO–Dock180 machinery. Nat Cell Biol 9, 121–128.
    20. Hansen-Wester, I. & Hensel, M. (2001). Salmonella pathogenicity islands encoding type III secretion systems. Microbes Infect 3, 549–559.
    21. Hersh, D., Monack, D. M., Smith, M. R., Ghori, N., Falkow, S. & Zychlinsky, A. (1999). The Salmonella invasin SipB induces macrophage apoptosis by binding to caspase-1. Proc Natl Acad Sci U S A 96, 2396–2401.
    22. Hertle, R. (2000). Serratia type pore forming toxins. Curr Protein Pept Sci 1, 75–89.
    23. Hertle, R., Hilger, M., Weingardt-Kocher, S. & Walev, I. (1999). Cytotoxic action of Serratia marcescens hemolysin on human epithelial cells. Infect Immun 67, 817–825.
    24. Hlady, W. G. & Klontz, K. C. (1996). The epidemiology of Vibrio infections in Florida, 1981–1993. J Infect Dis 173, 1176–1183.
    25. Hoashi, K., Ogata, K., Taniguchi, H., Yamashita, H., Tsuji, K., Mizuguchi, Y. & Ohtomo, N. (1990). Pathogenesis of Vibrio parahaemolyticus: intraperitoneal and orogastric challenge experiments in mice. Microbiol Immunol 34, 355–366.
    26. Hueck, C. J. (1998). Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiol Mol Biol Rev 62, 379–433.
    27. Iwai-Kanai, E., Yuan, H., Huang, C., Sayen, M. R., Perry-Garza, C. N., Kim, L. & Gottlieb, R. A. (2008). A method to measure cardiac autophagic flux in vivo. Autophagy 4, 322–329.
    28. Jepson, M. A., Pellegrin, S., Peto, L., Banbury, D. N., Leard, A. D., Mellor, H. & Kenny, B. (2003). Synergistic roles for the Map and Tir effector molecules in mediating uptake of enteropathogenic Escherichia coli (EPEC) into non-phagocytic cells. Cell Microbiol 5, 773–783.
    29. Joseph, S. W., Colwell, R. R. & Kaper, J. B. (1982). Vibrio parahaemolyticus and related halophilic Vibrios. Crit Rev Microbiol 10, 77–124.
    30. Kato, M., Nonaka, T., Maki, M., Kikuchi, H. & Imajoh-Ohmi, S. (2000). Caspases cleave the amino-terminal calpain inhibitory unit of calpastatin during apoptosis in human Jurkat T cells. J Biochem 127, 297–305.
    31. Kirby, J. E., Vogel, J. P., Andrews, H. L. & Isberg, R. R. (1998). Evidence for pore-forming ability by Legionella pneumophila. Mol Microbiol 27, 323–336.
    32. Lazebnik, Y. A., Kaufmann, S. H., Desnoyers, S., Poirier, G. G. & Earnshaw, W. C. (1994). Cleavage of poly(ADP-ribose) polymerase by a proteinase with properties like ICE. Nature 371, 346–347.
    33. Lynch, T., Livingstone, S., Buenaventura, E., Lutter, E., Fedwick, J., Buret, A. G., Graham, D. & DeVinney, R. (2005). Vibrio parahaemolyticus disruption of epithelial cell tight junctions occurs independently of toxin production. Infect Immun 73, 1275–1283.
    34. Mills, S. D., Boland, A., Sory, M. P., van der Smissen, P., Kerbourch, C., Finlay, B. B. & Cornelis, G. R. (1997). Yersinia enterocolitica induces apoptosis in macrophages by a process requiring functional type III secretion and translocation mechanisms and involving YopP, presumably acting as an effector protein. Proc Natl Acad Sci U S A 94, 12638–12643.
    35. Milton, D. L., O'Toole, R., Horstedt, P. & Wolf-Watz, H. (1996). Flagellin A is essential for the virulence of Vibrio anguillarum. J Bacteriol 178, 1310–1319.
    36. Mota, L. J. & Cornelis, G. R. (2005). The bacterial injection kit: type III secretion systems. Ann Med 37, 234–249.
    37. Munafo, D. B. & Colombo, M. I. (2001). A novel assay to study autophagy: regulation of autophagosome vacuole size by amino acid deprivation. J Cell Sci 114, 3619–3629.
    38. Navarre, W. W. & Zychlinsky, A. (2000). Pathogen-induced apoptosis of macrophages: a common end for different pathogenic strategies. Cell Microbiol 2, 265–273.
    39. Neyt, C. & Cornelis, G. R. (1999). Insertion of a Yop translocation pore into the macrophage plasma membrane by Yersinia enterocolitica: requirement for translocators YopB and YopD, but not LcrG. Mol Microbiol 33, 971–981.
    40. Nieminen, A. L., Saylor, A. K., Tesfai, S. A., Herman, B. & Lemasters, J. J. (1995). Contribution of the mitochondrial permeability transition to lethal injury after exposure of hepatocytes to t-butylhydroperoxide. Biochem J 307, 99–106.
    41. Nishibuchi, M., Fasano, A., Russell, R. G. & Kaper, J. B. (1992). Enterotoxigenicity of Vibrio parahaemolyticus with and without genes encoding thermostable direct hemolysin. Infect Immun 60, 3539–3545.
    42. Nonaka, T., Kuwabara, T., Mimuro, H., Kuwae, A. & Imajoh-Ohmi, S. (2003). Shigella-induced necrosis and apoptosis of U937 cells and J774 macrophages. Microbiology 149, 2513–2527.
    43. Ono, T., Park, K. S., Ueta, M., Iida, T. & Honda, T. (2006). Identification of proteins secreted via Vibrio parahaemolyticus type III secretion system 1. Infect Immun 74, 1032–1042.
    44. Palmer, L. E., Pancetti, A. R., Greenberg, S. & Bliska, J. B. (1999). YopJ of Yersinia spp. is sufficient to cause downregulation of multiple mitogen-activated protein kinases in eukaryotic cells. Infect Immun 67, 708–716.
    45. Park, K. S., Ono, T., Rokuda, M., Jang, M. H., Okada, K., Iida, T. & Honda, T. (2004). Functional characterization of two type III secretion systems of Vibrio parahaemolyticus. Infect Immun 72, 6659–6665.
    46. Pizarro-Cerda, J. & Cossart, P. (2006). Bacterial adhesion and entry into host cells. Cell 124, 715–727.
    47. Qadri, F., Alam, M. S., Nishibuchi, M., Rahman, T., Alam, N. H., Chisti, J., Kondo, S., Sugiyama, J., Bhuiyan, N. A. & other authors (2003). Adaptive and inflammatory immune responses in patients infected with strains of Vibrio parahaemolyticus. J Infect Dis 187, 1085–1096.
    48. Reyrat, J. M., Pelicic, V., Papini, E., Montecucco, C., Rappuoli, R. & Telford, J. L. (1999). Towards deciphering the Helicobacter pylori cytotoxin. Mol Microbiol 34, 197–204.
    49. Ribeiro, C., Vignes, M. & Brehelin, M. (2003). Xenorhabdus nematophila (enterobacteriacea) secretes a cation-selective calcium-independent porin which causes vacuolation of the rough endoplasmic reticulum and cell lysis. J Biol Chem 278, 3030–3039.
    50. Schlumberger, M. C. & Hardt, W. D. (2006). Salmonella type III secretion effectors: pulling the host cell's strings. Curr Opin Microbiol 9, 46–54.
    51. Sun, G. W., Lu, J., Pervaiz, S., Cao, W. P. & Gan, Y. H. (2005). Caspase-1 dependent macrophage death induced by Burkholderia pseudomallei. Cell Microbiol 7, 1447–1458.
    52. Swanson, M. S. & Fernandez-Moreira, E. (2002). A microbial strategy to multiply in macrophages: the pregnant pause. Traffic 3, 170–177.
    53. Takahashi, A., Iida, T., Naim, R., Naykaya, Y. & Honda, T. (2001). Chloride secretion induced by thermostable direct haemolysin of Vibrio parahaemolyticus depends on colonic cell maturation. J Med Microbiol 50, 870–878.
    54. Tewari, M., Quan, L. T., O'Rourke, K., Desnoyers, S., Zeng, Z., Beidler, D. R., Poirier, G. G., Salvesen, G. S. & Dixit, V. M. (1995). Yama/CPP32β, a mammalian homolog of CED-3, is a CrmA-inhibitable protease that cleaves the death substrate poly(ADP-ribose) polymerase. Cell 81, 801–809.
    55. Tsuchiya, S., Kobayashi, Y., Goto, Y., Okumura, H., Nakae, S., Konno, T. & Tada, K. (1982). Induction of maturation in cultured human monocytic leukemia cells by a phorbol diester. Cancer Res 42, 1530–1536.
    56. Wada, A., Yamasaki, E. & Hirayama, T. (2004). Helicobacter pylori vacuolating cytotoxin, VacA, is responsible for gastric ulceration. J Biochem 136, 741–746.
    57. Yeung, P. S. & Boor, K. J. (2004). Epidemiology, pathogenesis, and prevention of foodborne Vibrio parahaemolyticus infections. Foodborne Pathog Dis 1, 74–88.
    58. Zhou, X., Shah, D. H., Konkel, M. E. & Call, D. R. (2008). Type III secretion system 1 genes in Vibrio parahaemolyticus are positively regulated by ExsA and negatively regulated by ExsD. Mol Microbiol 69, 747–764.