A Helicobacter pylori Vacuolating Toxin Mutant That Fails To Oligomerize Has a Dominant Negative Phenotype
http://www.100md.com
感染与免疫杂志 2006年第3期
IRIS, Chiron Srl, Via Fiorentina 1, 53100 Siena, Italy
Venetian Institute of Molecular Medicine, Via Orus 2, 35129 Padova, Italy
Department of Evolutionary Biology, University of Siena, Via Aldo Moro 2, 53100 Siena, Italy
Dipartimento di Scienze Chimiche, Universita di Padova, via Marzolo 1, 35131 Padova, Italy
Dipartimento di Biologia, Universita di Padova, Via G. Colombo 3, 35121 Padova, Italy
ABSTRACT
Most Helicobacter pylori strains secrete a toxin (VacA) that causes massive vacuolization of target cells and which is a major virulence factor of H. pylori. The VacA amino-terminal region is required for the induction of vacuolization. The aim of the present study was a deeper understanding of the critical role of the N-terminal regions that are protected from proteolysis when VacA interacts with artificial membranes. Using a counterselection system, we constructed an H. pylori strain, SPM 326-49-57, that produces a mutant toxin with a deletion of eight amino acids in one of these protected regions. VacA 49-57 was correctly secreted by H. pylori but failed to oligomerize and did not have any detectable vacuolating cytotoxic activity. However, the mutant toxin was internalized normally and stained the perinuclear region of HeLa cells. Moreover, the mutant toxin exhibited a dominant negative effect, completely inhibiting the vacuolating activity of wild-type VacA. This loss of activity was correlated with the disappearance of oligomers in electron microscopy. These findings indicate that the deletion in VacA 49-57 disrupts the intermolecular interactions required for the oligomerization of the toxin.
INTRODUCTION
Helicobacter pylori is a gram-negative, microaerophilic, and spiral-shaped bacterium that colonizes the gastric mucosa of half of the world's population (26). This bacterium is associated with gastritis and peptic ulcers (4) and, for long-term chronic infections, the development of gastric carcinoma and mucosa-associated lymphoid tissue lymphoma (34).
Most H. pylori strains produce and secrete a toxin (VacA) which, together with two other virulence factors, CagA (8) and BabA (21), is significantly associated with strains that are isolated from patients with more severe disease (35). The most extensively characterized activity of VacA is its capacity to induce vacuolation in mammalian cells (9). These vacuoles contain markers for late endosomes and lysosomes and have an internal acidic pH (29, 32). Vacuolation is not the only activity of VacA, as it also increases the permeability of polarized epithelial cell monolayers to various ions and small uncharged molecules (33). After internalization by cells, VacA induces the release of cytochrome c into the cytosol, leading to apoptosis (11, 17, 50, 51). VacA also interferes with the immune system by altering the process of antigen presentation (30), and recently, it was shown to inhibit T-cell proliferation by two different mechanisms (6, 18, 43). Moreover, when administered intragastrically to mice, it causes gastric epithelial erosion (16, 45).
The vacA gene codes for a protoxin of 140 kDa, which is processed during export to the surface to yield a mature secreted toxin of 88 kDa (9, 45). Each molecule of VacA has two domains, p33 and p55, that are connected by a hydrophilic loop that can be cleaved after release from the bacterium. After cleavage, the two fragments remain noncovalently associated (45), indicating that the two subunits interact in such a way as to maintain the integrity of the 88-kDa structure. The p55 domain is responsible for host cell binding (37) and host tropism (31), whereas the p33 domain, together with the N-terminal 192 amino acids of p55, is sufficient to cause vacuolation when expressed in the cytoplasm of host cells (14).
Mature 88-kDa VacA monomers assemble into water-soluble, high-molecular weight-oligomers containing 6 to 7 or 12 to 14 copies of the mature toxin polypeptide (1, 10, 23). The purified oligomeric form of the toxin is biologically inactive, but after exposure to acidic pH, it disassembles into monomeric components (10); this permits insertion of the toxin into lipid membranes to form anion-conductive channels (44, 46). The VacA channel formation induces an osmotic imbalance of intracellular acidic compartments, leading to cytoplasmic vacuolation (46).
A previous study has reported that a mutant toxin of VacA, lacking the hydrophobic amino terminus region, has a dominant negative phenotype (48). This mutant, which lacks cytotoxic activity and completely inhibits the vacuolating activity of the wild-type toxin, is indistinguishable from wild-type VacA in its secretion, assembly into oligomeric structures, and uptake by HeLa cells. It was demonstrated later that this mutant lacks three tandem GXXXG motifs that are necessary for the dimerization of VacA within membranes and consequent anion channel formation (27).
In this study, we describe another dominant negative mutant, VacA 49-57, which has a deletion of eight amino acids in a region known to be protected from proteolysis following VacA interaction with artificial membranes (49). This mutant also inhibited the activity of wild-type VacA and failed to form oligomeric structures.
MATERIALS AND METHODS
Bacterial strains, yeast strains, and culture conditions. H. pylori CCUG 17874 was used as the source of VacA. H. pylori SPM 326, which encodes an s1m1-type vacA (2, 25), was the parental strain used for the construction of the mutant.
Colonies of H. pylori grown on blood agar plates (Columbia agar with 5% horse blood) were inoculated into Brucella broth containing 0.2% -cyclodextrin and were cultured for 2 days at 170 rpm in microaerophilic conditions. Yeast two-hybrid experiments were performed with the Saccharomyces cerevisiae EGY48 (MAT ura3 trp1 his3 6LexA-operator-LEU2) strain transformed with the plasmid pSH18-34 containing the lacZ reporter gene. Yeast strains were grown in synthetic minimal medium (SD) supplemented with the required amino acids at 30°C as previously described (19).
Construction of H. pylori SPM 326-49-57. First, a recipient strain was created by introduction of the kan-sacB cassette into the vacA gene of H. pylori SPM 326. The kan-sacB module, amplified from the plasmid pEnKSF (7), was inserted into the NcoI/EcoNI site of pBlueScriptKS p33, a plasmid containing a substantial portion of the vacA SPM 326 gene to create pKan/sacB (Fig. 1). The NcoI site, which is absent in the vacA gene, was added using the QuikChange site-directed mutagenesis kit (Stratagene) by changing the two nucleotides immediately before the sequence coding the translational start site. A kanamycin-resistant, sucrose-sensitive clone, in which the expression of the VacA molecule was inactivated, was selected. This strain, H. pylori SPM 326KO2, was then transformed by pBlueScriptKS p33-49-57, and a sucrose-resistant, kanamycin-sensitive transformant, which was obtained by homologous recombination between the two vacA fragments, was selected as summarized in Fig. 1 (7). The strain was analyzed by nucleotide sequencing to verify that the desired mutation was present.
The plasmid pBlueScriptKS p33-49-57 was obtained from pBlueScriptKS p33 by the introduction of an in-frame vacA deletion mutation using the QuikChange site-directed mutagenesis kit (Stratagene). Briefly, oligonucleotide primers, each being complementary to opposite strands of the vector and containing the deletion, were extended by PCR using Pfu DNA polymerase. Following thermal cycling, template DNA was eliminated by DpnI digestion and the PCR products were used to transform Escherichia coli XL1-Blue.
Purification of VacA and VacA 49-57. VacA from H. pylori strain CCUG 17874 was purified from the broth culture supernatant as described previously (24). For VacA 49-57, the biomass from a 1 liter culture of H. pylori SPM 326-49-57 was removed by centrifugation at 11,000 x g for 20 min. The supernatant was passed through a 0.2-μm filter and then applied at a flow rate of 1 ml/min to a column (5 by 3.5 cm, XK column; Amersham Pharmacia Biotech) containing Matrex Cellufine Sulfate (Amicon). The proteins were washed extensively with 100 mM NaCl-20 mM phosphate buffer (pH 6.5). VacA 49-57 was eluted from the column with 350 mM NaCl-20 mM phosphate buffer (pH 6.5). The mutant toxin was found to be highly pure at this stage; this was in contrast to the wild-type toxin which required a further step of purification by gel filtration in order to achieve the same purity. The eluate was brought to 50% saturation with ammonium sulfate and centrifuged at 20,000 x g for 30 min. The pelleted proteins were resuspended in phosphate-buffered saline (PBS) and dialyzed extensively against the same buffer. Purified toxins were stored at 4°C, and their concentrations were determined by using a microbicinchoninic acid assay (Pierce).
Cell vacuolation assay. HeLa cells were seeded at 7.5 x 104 ml–1 in 96-well plates in Dulbecco modified Eagle's medium containing 10% fetal calf serum and 2 mM glutamine at 37°C in 5% CO2 16 h before the assay. Before addition to cells, purified toxin preparations were acid-activated by adjusting to pH 2.0 for 5 min at room temperature and then neutralized as described previously (12). The extent of vacuolation was determined quantitatively by measuring the uptake of neutral red dye after the incubation of the toxin with cells at 37°C for 9 h in Dulbecco modified Eagle's medium containing 2% fetal calf serum and 15 mM ammonium chloride (9).
Deep-etch electron microscopy. VacA and VacA 49-57 molecules were prepared for microscopy by a procedure of absorption to mica, followed by quick-freezing and deep-etching (20). The samples were processed as described previously (23). For the experiment showing the effect of VacA 49-57 on VacA, an equimolar concentration of each toxin, acid activated when indicated, was mixed before the treatment.
Glycerol gradient centrifugation analysis. For analysis, 14-ml 10 to 30% glycerol gradients were prepared in either 60 mM Tris (pH 7.5) or 100 mM glycine (pH 3.0) containing 100 mM NaCl. Samples of 200 μl (dialyzed ammonium sulfate-precipitated proteins from broth culture supernatant diluted to an optical density at 600 nm [OD600] of 1) were layered on the gradients and centrifuged at 39,000 rpm for 10 h at 4°C in a SW40 Ti rotor (Beckman Instruments) as described previously (10). For the experiment showing the effect of VacA 49-57 on VacA, 100 μl of each concentrated supernatant, acid activated when indicated, was mixed before centrifugation. For the control, 100 μl of SPM 326 acid-activated concentrated supernatant was mixed with 100 μl of SPM 326KO2. Gradients were fractionated from the top by using a piston gradient fractionator (Biocomp). Proteins from the fractions were precipitated with 10% trichloroacetic acid and 0.02% sodium deoxycholate. The presence of VacA and VacA 49-57 was detected by Western blotting with rabbit anti-VacA serum (45).
Immunofluorescence and confocal microscopy. For indirect immunofluorescence analysis, HeLa cells were grown on chamber slides (Nunc). After 4 h of incubation with 5 μg/ml of purified VacA 49-57 and acid-activated purified VacA, the cells were fixed with 3.7% paraformaldehyde in PBS containing 5 mM Ca2+ and Mg2+ for 15 min. Aldehyde groups were quenched with 0.2 M glycine for 5 min. Plasma membrane-localized toxin was stained with a mouse monoclonal antibody specific for the toxin used at 2 μg/ml (36) followed by Texas Red goat anti-mouse antibody. For cytosolic staining, the previously surface-stained cells were then permeabilized with 0.2% Triton X-100 for 10 min at room temperature (22). PBS containing 5% fetal calf serum was used as a blocking solution. Labeling of the cytoplasmic toxin was obtained by using the same monoclonal antibody and fluorescein isothiocyanate (FITC) goat anti-mouse was used as secondary antibody. Slides were mounted with a SlowFade light antifade kit (Molecular Probes). Confocal images were obtained by using an Ultraview microscope (PerkinElmer).
Circular dichroism (CD) spectroscopy. Spectra for VacA (70 μg/ml) and VacA 49-57 (112 μg/ml) in PBS at pH 7.4, PBS at pH 3, and PBS after reneutralization were acquired with a Jasco J-715 spectropolarimeter at room temperature in a quartz cell with an optical path length of 0.1 cm. The spectra were recorded in the 195-to-260-nm wavelength range, using a 2-nm bandwidth and a 2-s time constant at a scan speed of 50 nm/min. All spectra are reported in terms of mean residue molar ellipticity []R degree cm2 dmol–1. Data processing was carried out using a J-700 software package. The signal-to-noise ratio was improved by the accumulation of at least 10 scans. CDNN 2.1 software was used to analyze the secondary structure content of the proteins (5).
Yeast two-hybrid system. Sequences encoding the wild-type p33, p33 49-57, and p55 VacA domains were PCR amplified from pBlueScriptKS p33, pBlueScriptKS p33-49-57, and pBlueScriptKS p55, respectively. PCR products were cloned into plasmids encoding the LexA DNA binding domain (pBD) and/or the transcription activation domain B42 (pAD) (15, 19). Primers were designed in order to amplify the region of vacA coding for amino acids 1 to 319 (p33 domain) and 320 to 837 (p55 domain) of the mature toxin from H. pylori SPM 326. Before transformation, the plasmids were analyzed by automated nucleotide sequencing.
The EGY48/pSH18-34 yeast strain was cotransformed with 2 μg of bait and prey containing plasmids in all possible combinations by using the lithium acetate method (41) and cultured at 30°C on SD without Trp, without His, and with Leu. In order to test which interactions between VacA domains could be detected in this system, and which combination was able to activate the transcription of the LEU2 and the lacZ reporter genes, several independent clones of each cotransformant were seeded on a selective plate. Three isolates of each cotransformant were then chosen for the -galactosidase interaction assay.
To quantify the level of interaction between the different domains, selected cotransformants were grown in 4 ml SD containing 2% galactose, 1% raffinose, and 50 μg/ml leucine at 30°C for 24 h until the culture reached an OD640 of 4. For each strain, a volume of culture containing 107 cells and corresponding to an OD640 of 2 was centrifuged at 3,000 x g, 5 min and washed once with 1 ml of Z buffer (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1.8 mM MgSO4, 0.27% -mercaptoethanol, pH 7) (42) and the pellet was resuspended in 150 μl of Z buffer. Fifty-microliter aliquots of this cell suspension were permeabilized by using Z buffer containing 70% CHCl3 and 0.02% sodium dodecyl sulfate and incubated with 700 μl of o-nitrophenyl--D-galactopyranoside (2 mg/ml) at 30°C. The reaction was stopped at selected time intervals by the addition of 500 μl of 1 M K2CO3, and the reaction time was recorded. The OD420 values of clarified reaction supernatants were measured. The -galactosidase activity was calculated using the Miller equation (19). The values presented here are the averages of three independent cotransformants for each combination of bait and prey that were assessed in three different experiments.
RESULTS
VacA 49-57 is devoid of vacuolating activity. A study of the protection of VacA from protease digestion, following its membrane adsorption, suggested that its N-terminal region is involved in VacA membrane interaction (49). On the other hand, the cell vacuolating activity of VacA is due to its capability of forming trans-membrane anion-selective channels (44, 46). Therefore, we decided to investigate the role of the N terminus of VacA. On the basis of a preliminary in silico analysis, we selected the region of amino acids 49 to 57 as a putative membrane interactive segment and we used a counterselection marker (7, 38) to introduce an in-frame deletion of eight amino acids into the chromosomal vacA gene of H. pylori SPM326 (Fig. 1). Western blot analysis of soluble extracts and supernatants showed that the mutant strain, H. pylori SPM 326-49-57, produces and secretes a toxin of the expected molecular mass for the mature polypeptide (data not shown). However, the VacA 49-57 protein behaved differently in affinity chromatography from the wild-type protein. VacA 49-57 eluted from a Matrex Cellufine Sulfate column with 350 mM NaCl, whereas wild-type VacA needs a higher salt concentration (24). This difference of affinity for the Matrex Cellufine Sulfate resin implies that the mutant toxin has different properties relative to those of wild-type VacA.
The amino acids that were deleted are located in the p33 domain, which is known to be essential for vacuolating activity (Fig. 1). Purified VacA 49-57 was tested for activity in a neutral red uptake assay using HeLa cells. As shown in Fig. 1C, the mutant toxin did not induce vacuoles.
VacA 49-57 enters epithelial cells. VacA binds epithelial cells and is then internalized into the cell cytosol, where it displays its activity (13). We investigated whether the lack of vacuolation that was observed was due to the incapacity of VacA 49-57 to be internalized by HeLa cells. After 4 h of incubation, cells treated with either the wild-type or the mutant toxin were fixed and immunostained with a monoclonal antibody, C1G9, which is specific for VacA and Texas Red-labeled anti-mouse immunoglobulin antibodies. The cells were then permeabilized and restained with the same monoclonal antibody and a FITC-labeled secondary antibody. Intracellular VacA 49-57 was observed as green-labeled spots in the perinuclear region of cells which was comparable to the staining that was obtained in cells incubated with wild-type VacA (Fig. 2). Hence, the mutant toxin internalized into HeLa cells in a manner similar to that of the wild-type protein.
VacA 49-57 fails to form the oligomeric structure. Purified VacA forms water-soluble oligomeric structures which can be observed by transmission electron microscopy after a quick-freeze, deep-etch preparation (23). Since VacA 49-57 has a lower affinity for the Matrex Cellufine Sulfate resin, we checked the integrity of the mutant toxin. Transmission electron microscope visualization of purified wild-type VacA revealed the expected structures with either six- or sevenfold radial symmetry. These structures were not observed in preparations of VacA 49-57 (Fig. 3). After treatment at low pH, VacA oligomers disassemble into monomers and the two structures can be separated by velocity sedimentation on glycerol density gradients (10). As expected, wild-type VacA from strain SPM 326 sedimented at the bottom of the gradient, reflecting the high-molecular-weight oligomeric structure that is visible in the electron micrographs (Fig. 3). At pH 3.0, VacA was found at the top of the gradient. VacA 49-57 was found at the top of the gradient at both neutral and acid pH values. This demonstrates that VacA 49-57 is defective in oligomerization at a neutral pH and that the structure observed by transmission electron microscopy is likely to be a monomer. A very small quantity of VacA 49-57 was found at the bottom of the gradient at a neutral pH. This may imply that the mutant toxin does form oligomers, although it does so very inefficiently. However, no oligomers were ever observed in electron micrographs of the mutant toxin.
The deletion of amino acids 49 to 57 affects the secondary structure content of VacA and its sensitivity to low pH. We used far-UV circular dichroism spectroscopy to determine whether the eight-amino-acid deletion 49-57 affects the integrity of VacA. As shown in Fig. 4A, the spectrum of VacA at pH 7.4 displays a negative band at around 220 nm and has a substantial amount of -like secondary structure (44%) as previously described (12). VacA 49-57 has a lower -sheet structure content (33%) compared to that of VacA, as revealed by the presence of two negative bands in the CD spectrum, one that was intense at 206 nm and one that was weaker at approximately 220 nm (Fig. 4A).
Upon acidification (pH 3), the negative band of the VacA spectrum slightly shifted to a lower wavelength, suggesting a small change in the secondary or quaternary structure content, but returned to the same profile after neutralization (Fig. 4B). For the mutant toxin, the signal was markedly weaker after acid exposure and showed a significant change in the profile. In fact, VacA 49-57 exhibited an irreversible conformation change (Fig. 4C). Hence, the eight-amino-acid deletion that was introduced in the p33 domain decreased the -sheet structure content but also rendered the toxin sensitive to low pH.
VacA 49-57 is a dominant negative mutant. The vacuolating activity of the native toxin can be abolished by introducing mutated VacA into the oligomer (48). We investigated whether mixing wild-type VacA with VacA 49-57 can interfere with the formation of vacuoles. HeLa cells were incubated with a fixed concentration of acid-activated wild-type VacA, to which increasing concentrations of VacA 49-57 were added prior to incubation with the cells. VacA 49-57 clearly affects the cytotoxic activity of wild-type VacA in a dose-dependent manner (Fig. 5). When the two toxins were present in equimolar concentrations, VacA 49-57 completely inhibited the activity of the native protein. Furthermore, significant inhibition was achieved with a 7:1 molar ratio of wild-type to mutant toxin. Hence, in this experimental condition, VacA 49-57 dominantly inhibited the vacuolating activity of wild-type VacA. Heat-inactivated VacA 49-57 (10 min at 95°C) was also tested at an equimolar concentration, but it did not present any inhibitory effect (data not shown). Non-acid-activated VacA 49-57 completely inhibited the vacuolating activity of wild-type VacA, whereas acid-activated VacA 49-57 reduced the cytotoxic activity by only 50% at the same relative concentration. This is likely due to the denaturing effect of low pH on the mutant toxin observed in the CD spectra.
VacA 49-57 inhibits wild-type VacA oligomer reformation. Disassembly of VacA oligomers by treatment at a low pH is reversible (10). After neutralization, the VacA monomers reassemble into oligomers. We sought to determine whether VacA 49-57 could interfere with the reassembly of wild-type VacA oligomers. A solution with the same concentrations of VacA 49-57 and wild-type VacA was examined by transmission electron microscopy after a quick-freeze, deep-etch preparation. The electron micrograph revealed two structures, the oligomer characteristic of native VacA and a smaller structure, corresponding to the VacA 49-57 monomer (Fig. 6A), also observed in Fig. 3. Therefore, the mutant toxin had no effect on the structure of the wild-type toxin oligomers. When the wild-type VacA oligomers were disrupted at low pH and then neutralized in the presence of the mutant toxin, the wild-type toxin failed to reassemble (Fig. 6B). In order to prove that the absence of oligomers was due to VacA 49-57 and not to acid treatment or to contaminants in the VacA 49-57 preparation, purified VacA treated at low pH and then neutralized in the presence of PBS was also submitted to a quick-freeze, deep-etch preparation. As shown in Fig. 6C, this procedure yielded flower-shaped structures similar to those of the VacA oligomers already described (10). Furthermore, glycerol gradient sedimentation experiments on concentrated culture supernatant (Fig. 6D to F) confirmed that VacA 49-57 is responsible for the failure of wild-type VacA to reassemble since supernatant of SPM 326-49-57 blocked the reassembly of wild-type VacA, whereas supernatant from a strain lacking the vacA gene (SPM 326KO2) did not (Fig. 6F). In conclusion, our results indicate that the loss of vacuolating cytotoxic activity of wild-type VacA in the presence of VacA 49-57 is correlated with the inability of the toxin to reassemble into oligomeric structures.
Interaction of p33/p55 and p3349-57/p55 in a yeast two-hybrid system. The mature 88-kDa VacA monomer is sensitive to proteolytic cleavage at a hydrophilic loop connecting the p33 and p55 domains which, however, remain associated in the cleaved protein (45). Hence, p33 binds to p55 in the absence of covalent linkage. Furthermore, analysis of the three-dimensional electron microscopy structure of VacA and a truncated molecule lacking the p33 domain led to a model which suggested that the oligomeric structure is formed by the interaction of the p33 moiety with the p55 moiety of the adjacent monomer (37). Therefore, p33 and p55 have two sites of interaction, one which maintains the integrity of the monomer and the other that is involved in intermolecular interactions in the oligomer. Evidence for the interaction of p33 with p55 comes from experiments using the yeast two-hybrid system (47). However, these experiments could not distinguish between p33/p55 interactions involved in maintaining the structure of the mature monomeric polypeptide (45) and the interactions between p33 of one monomer with p55 in the adjacent monomer in the oligomeric structure (37). Since the 49-57 deletion in the p33 domain appears to affect the interaction between the VacA domains that are necessary for the formation of oligomers, we decided to use this system to test whether the interactions between the p33 and p55 domain were compromised by the deletion. The vacA sequence encoding p33 was cloned into a plasmid containing the DNA binding domain of LexA (pBD), whereas p55 was cloned into a plasmid containing the B42 transcription activation domain (pAD) (19). The EGY48 yeast strain carrying the lacZ reporter plasmid (pSH18-34) was transformed with plasmids pBD-p33 and pAD-p55, and cotransformants were selected on specific medium. Three clones were chosen at random for further analysis. Interaction between wild-type p33 and p55 in this system was confirmed by measuring the activation of the lacZ reporter gene. The level of interaction determined by using the -galactosidase liquid assay on total yeast extracts is reported in Fig. 7. This combination was used as a positive control for further investigation of p33 49-57/p55 interactions. As shown, the 49-57 deletion in the p33 domain caused a 50% decrease in the interaction with p55 (Fig. 7). This result is consistent with a loss of the p33/p55 interactions involved in oligomerization without interfering with the interactions involved in maintaining the monomeric structure after cleavage of the hydrophilic loop between p33 and p55. None of the LexA fusion constructs (pBD-p33 and pBD-p33 49-57) were able to activate the expression of the reporter gene in the absence of the pAD-p55 hybrid. Similarly, extracts from the yeast strain that were transformed with only the p55 moiety fused to the activating domain showed very low -galactosidase activity in absence of the other hybrid.
DISCUSSION
Purified VacA assembles in high-molecular-weight oligomers which have flower-like structures when observed by quick-freeze, deep-etch electron microscopy. Mutational analysis of p33 has shown that the region spanning amino acids 28 to 196 is important for interaction between two adjacent monomers and formation of the global structure (47, 48). In this study, we show that the deletion of amino acids 49 to 57 of p33 disrupts this interaction.
The mutated VacA protein described in this study (VacA 49-57) was produced and secreted by H. pylori in a manner similar to that of the wild-type toxin, but it lacked vacuolating cytotoxic activity. This absence of activity correlated with its lack of oligomerization. In agreement with previous work (37, 48), this finding indicates that VacA oligomerization is essential for vacuole formation, although it cannot be ruled out that the mutant toxin is defective in some other way as, for example, in the interaction with cytoplasmic target cell proteins. It is noteworthy that this is the smallest deletion which has been found to be capable of blocking the oligomerization of VacA. However, since the deletion results in a slightly altered CD spectrum, we cannot conclude that this region is directly involved in oligomerization.
A previous study showed that the p55 molecule was able to interact with the surface of target cells but was not internalized (37). This suggests that either the p33 subunit or the ability to form an oligomeric structure is required for cell entry. Here we show that the VacA 49-57 monomer has the capacity to enter HeLa cells and localize in the cytoplasm in a manner similar to that observed in VacA-infected cells (22), demonstrating that oligomerization is not necessary for internalization. This mechanism differs from that described for the C2 toxin of Clostridium botulinum, a related channel-forming molecule which needs to form oligomers for cellular uptake (3). We propose that a region of p33, one that is different from the region deleted in this study and still unknown, must be involved in VacA internalization.
A remarkable property of VacA 49-57 is its capacity to inhibit the cytotoxic activity of the wild-type toxin. It is unlikely that this is due to competition for the binding to the VacA receptor. The p55 domain has been shown to be responsible for binding to the host cell and can fold independently of p33 (37), therefore VacA 49-57 is expected to bind the VacA receptor with the same affinity as the wild-type toxin. When the two proteins were mixed in equimolar concentrations, a complete inhibition of cytotoxic activity was observed. Thus the inhibition is unlikely to be due to competition for the receptor, particularly since HeLa cells exhibit a high level of nonspecific binding for radiolabeled VacA. Only a small reduction in the activity was observed in the presence of a 100-fold excess of unlabeled VacA (28, 39).
VacA 49-57 was able to abolish the cytotoxic activity of wild-type VacA in a manner similar to that of VacA 6-27, another dominant negative mutant of H. pylori-vacuolating toxin (48). However, VacA 6-27 presented a dominant negative phenotype only when it was acid activated, while VacA 49-57 was able to completely inhibit wild-type VacA activity even without acid activation. The inhibition of activity was also significant at a molar ratio of 7:1 for wild-type to mutant toxin, suggesting that one mutant molecule per oligomeric complex is sufficient to cause substantial inhibition of vacuolating activity. VacA 6-7 forms an oligomer and, after exposure to low pH, monomers of VacA 6-7 can interact with monomers of VacA to form dysfunctional mixed oligomers (48). In contrast, VacA 49-57 does not oligomerize and does not need to be acid activated to interact with wild-type VacA monomers. In fact, acid-activated VacA 49-57 inhibited only 50% of the wild-type VacA. Accordingly, the far-UV CD data indicate that the lower inhibition observed with acid-activated VacA 49-57 results from structural damage to the mutant toxin after exposure to acidic pH.
In a recent study, using a FLAG-VacA toxin, which can be cleaved into the p33 and p55 domains, it has been demonstrated that the two fragments remain physically associated after proteolytic cleavage and were still able to form a VacA oligomeric structure (47). This suggests that there are two types of interactions between the two domains of VacA: intramolecular interactions between the p33 and p55 domains of an individual VacA monomer and intermolecular interactions between p33 and p55 of different VacA molecules, which are necessary for the formation of oligomeric structures. As shown by the yeast two-hybrid system, the deletion of eight amino acids in the p33 domain reduced the level of interaction with the p55 domain by 50%, consistent with the loss of the latter but not the former p33/p55 interaction. We propose that p33 49-57 lacks the capacity to interact with the p55 domain of another monomer. Thus, after acid activation, the p33 domain of monomers of the wild-type toxin may interact with the p55 domain of VacA 49-57; however, the reduced ability of the p33 domain of VacA 49-57 to interact with the wild-type toxin would prevent the complete formation of the active oligomer in the membrane. This would explain why inhibition could be achieved at even low mutant-to-wild-type toxin ratios.
Over the last 10 years, many studies have shown that due to the formation of a VacA anion-selective channel, the toxin causes multiple effects on target cells in vitro (6, 9, 11, 17, 18, 30, 43, 50, 51) but the in vivo function of VacA and its relevance for H. pylori initial colonization of the stomach remains unclear (40). Since VacA 49-57 is produced and correctly secreted by H. pylori and has the capacity to enter epithelial cells without inducing vacuolation, it could be a useful tool to understanding the role of VacA in H. pylori infection.
ACKNOWLEDGMENTS
C. Genisset was supported by Marie Curie Industry Host Fellowship QLK4 1999 50407.
We thank R. Janulczyk and C. Montecucco for critically reading the manuscript. We are grateful to D. Bison and E. Pasqualetto for technical assistance with the confocal microscope and CD spectroscopy respectively. We thank G. Corsi for the artwork and S. Pasquini, L. Fini, and S. Magi for the medium preparation.
REFERENCES
1. Adrian, M., T. L. Cover, J. Dubochet, and J. E. Heuser. 2002. Multiple oligomeric states of the Helicobacter pylori vacuolating toxin demonstrated by cryo-electron microscopy. J. Mol. Biol. 318:121-133.
2. Atherton, J. C., P. Cao, R. M. Peek, Jr., M. K. Tummuru, M. J. Blaser, and T. L. Cover. 1995. Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori. Association of specific vacA types with cytotoxin production and peptic ulceration. J. Biol. Chem. 270:17771-17777.
3. Barth, H., D. Blocker, J. Behlke, W. Bergsma-Schutter, A. Brisson, R. Benz, and K. Aktories. 2000. Cellular uptake of Clostridium botulinum C2 toxin requires oligomerization and acidification. J. Biol. Chem. 275:18704-18711.
4. Blaser, M. J. 1990. Helicobacter pylori and the pathogenesis of gastroduodenal inflammation. J. Infect. Dis. 161:626-633.
5. Bohm, G., R. Muhr, and R. Jaenicke. 1992. Quantitative analysis of protein far UV circular dichroism spectra by neural networks. Protein Eng. 5:191-195.
6. Boncristiano, M., S. R. Paccani, S. Barone, C. Ulivieri, L. Patrussi, D. Ilver, A. Amedei, M. M. D'Elios, J. L. Telford, and C. T. Baldari. 2003. The Helicobacter pylori vacuolating toxin inhibits T cell activation by two independent mechanisms. J. Exp. Med. 198:1887-1897.
7. Copass, M., G. Grandi, and R. Rappuoli. 1997. Introduction of unmarked mutations in the Helicobacter pylori vacA gene with a sucrose sensitivity marker. Infect. Immun. 65:1949-1952.
8. Covacci, A., S. Censini, M. Bugnoli, R. Petracca, D. Burroni, G. Macchia, A. Massone, E. Papini, Z. Xiang, N. Figura, and R. Rappuoli. 1993. Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proc. Natl. Acad. Sci. USA 90:5791-5795.
9. Cover, T. L., and M. J. Blaser. 1992. Purification and characterization of the vacuolating toxin from Helicobacter pylori. J. Biol. Chem. 267:10670-10675.
10. Cover, T. L., P. I. Hanson, and J. E. Heuser. 1997. Acid-induced dissociation of VacA, the Helicobacter pylori vacuolating cytotoxin, reveals its pattern of assembly. J. Cell Biol. 138:759-769.
11. Cover, T. L., U. S. Krishna, D. A. Israel, and R. M. Peek, Jr. 2003. Induction of gastric epithelial cell apoptosis by Helicobacter pylori vacuolating cytotoxin. Cancer Res. 63:951-957.
12. de Bernard, M., E. Papini, V. de Filippis, E. Gottardi, J. Telford, R. Manetti, A. Fontana, R. Rappuoli, and C. Montecucco. 1995. Low pH activates the vacuolating toxin of Helicobacter pylori, which becomes acid and pepsin resistant. J. Biol. Chem. 270:23937-23940.
13. de Bernard, M., B. Arico, E. Papini, R. Rizzuto, G. Grandi, R. Rappuoli, and C. Montecucco. 1997. Helicobacter pylori toxin VacA induces vacuole formation by acting in the cell cytosol. Mol. Microbiol. 26:665-674.
14. de Bernard, M., D. Burroni, E. Papini, R. Rappuoli, J. Telford, and C. Montecucco. 1998. Identification of the Helicobacter pylori VacA toxin domain active in the cell cytosol. Infect. Immun. 66:6014-6016.
15. Di Giaimo, R., M. Riccio, S. Santi, C. Galeotti, D. C. Ambrosetti, and M. Melli. 2002. New insights into the molecular basis of progressive myoclonus epilepsy: a multiprotein complex with cystatin B. Hum. Mol. Genet. 11:2941-2950.
16. Fujikawa, A., D. Shirasaka, S. Yamamoto, H. Ota, K. Yahiro, M. Fukada, T. Shintani, A. Wada, N. Aoyama, T. Hirayama, H. Fukamachi, and M. Noda. 2003. Mice deficient in protein tyrosine phosphatase receptor type Z are resistant to gastric ulcer induction by VacA of Helicobacter pylori. Nat. Genet. 33:375-381.
17. Galmiche, A., J. Rassow, A. Doye, S. Cagnol, J. C. Chambard, S. Contamin, V. de Thillot, I. Just, V. Ricci, E. Solcia, E. Van Obberghen, and P. Boquet. 2000. The N-terminal 34 kDa fragment of Helicobacter pylori vacuolating cytotoxin targets mitochondria and induces cytochrome c release. EMBO J. 19:6361-6370.
18. Gebert, B., W. Fischer, E. Weiss, R. Hoffmann, and R. Haas. 2003. Helicobacter pylori vacuolating cytotoxin inhibits T lymphocyte activation. Science 301:1099-1102.
19. Golemis, E., and I. Serebriiski. 1998. Two-hybrid system/interaction trap, p. 69-69.40. In D. L. Spector, R. D. Goldman, and L. A. Leinwand (ed.), Cells: a laboratory manual, vol. 1. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
20. Heuser, J. E. 1983. Procedure for freeze-drying molecules adsorbed to mica flakes. J. Mol. Biol. 169:155-195.
21. Ilver, D., A. Arnqvist, J. Ogren, I. M. Frick, D. Kersulyte, E. T. Incecik, D. E. Berg, A. Covacci, L. Engstrand, and T. Boren. 1998. Helicobacter pylori adhesin binding fucosylated histo-blood group antigens revealed by retagging. Science 279:373-377.
22. Ilver, D., S. Barone, D. Mercati, P. Lupetti, and J. L. Telford. 2004. Helicobacter pylori toxin VacA is transferred to host cells via a novel contact-dependent mechanism. Cell. Microbiol. 6:167-174.
23. Lupetti, P., J. E. Heuser, R. Manetti, P. Massari, S. Lanzavecchia, P. L. Bellon, R. Dallai, R. Rappuoli, and J. L. Telford. 1996. Oligomeric and subunit structure of the Helicobacter pylori vacuolating cytotoxin. J. Cell Biol. 133:801-807.
24. Manetti, R., P. Massari, D. Burroni, M. de Bernard, A. Marchini, R. Olivieri, E. Papini, C. Montecucco, R. Rappuoli, and J. L. Telford. 1995. Helicobacter pylori cytotoxin: importance of native conformation for induction of neutralizing antibodies. Infect. Immun. 63:4476-4480.
25. Marchetti, M., B. Arico, D. Burroni, N. Figura, R. Rappuoli, and P. Ghiara. 1995. Development of a mouse model of Helicobacter pylori infection that mimics human disease. Science 267:1655-1658.
26. Marshall, B. J., and J. R. Warren. 1984. Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet i:1311-1315.
27. McClain, M. S., H. Iwamoto, P. Cao, A. D. Vinion-Dubiel, Y. Li, G. Szabo, Z. Shao, and T. L. Cover. 2003. Essential role of a GXXXG motif for membrane channel formation by Helicobacter pylori vacuolating toxin. J. Biol. Chem. 278:12101-12108.
28. McClain, M. S., W. Schraw, V. Ricci, P. Boquet, and T. L. Cover. 2000. Acid activation of Helicobacter pylori vacuolating cytotoxin (VacA) results in toxin internalization by eukaryotic cells. Mol. Microbiol. 37:433-442.
29. Molinari, M., C. Galli, N. Norais, J. L. Telford, R. Rappuoli, J. P. Luzio, and C. Montecucco. 1997. Vacuoles induced by Helicobacter pylori toxin contain both late endosomal and lysosomal markers. J. Biol. Chem. 272:25339-25344.
30. Molinari, M., M. Salio, C. Galli, N. Norais, R. Rappuoli, A. Lanzavecchia, and C. Montecucco. 1998. Selective inhibition of Ii-dependent antigen presentation by Helicobacter pylori toxin VacA. J. Exp. Med. 187:135-140.
31. Pagliaccia, C., M. de Bernard, P. Lupetti, X. Ji, D. Burroni, T. L. Cover, E. Papini, R. Rappuoli, J. L. Telford, and J. M. Reyrat. 1998. The m2 form of the Helicobacter pylori cytotoxin has cell type-specific vacuolating activity. Proc. Natl. Acad. Sci. USA 95:10212-10217.
32. Papini, E., B. Satin, C. Bucci, M. de Bernard, J. L. Telford, R. Manetti, R. Rappuoli, M. Zerial, and C. Montecucco. 1997. The small GTP binding protein rab7 is essential for cellular vacuolation induced by Helicobacter pylori cytotoxin. EMBO J. 16:15-24.
33. Papini, E., B. Satin, N. Norais, M. de Bernard, J. L. Telford, R. Rappuoli, and C. Montecucco. 1998. Selective increase of the permeability of polarized epithelial cell monolayers by Helicobacter pylori vacuolating toxin. J. Clin. Investig. 102:813-820.
34. Parsonnet, J., S. Hansen, L. Rodriguez, A. B. Gelb, R. A. Warnke, E. Jellum, N. Orentreich, J. H. Vogelman, and G. D. Friedman. 1994. Helicobacter pylori infection and gastric lymphoma. N. Engl. J. Med. 330:1267-1271.
35. Prinz, C., M. Schoniger, R. Rad, I. Becker, E. Keiditsch, S. Wagenpfeil, M. Classen, T. Rosch, W. Schepp, and M. Gerhard. 2001. Key importance of the Helicobacter pylori adherence factor blood group antigen binding adhesin during chronic gastric inflammation. Cancer Res. 61:1903-1909.
36. Reyrat, J. M., M. Charrel, C. Pagliaccia, D. Burroni, P. Lupetti, M. de Bernard, X. Ji, N. Norais, E. Papini, R. Dallai, R. Rappuoli, and J. L. Telford. 1998. Characterisation of a monoclonal antibody and its use to purify the cytotoxin of Helicobacter pylori. FEMS Microbiol. Lett. 165:79-84.
37. Reyrat, J. M., S. Lanzavecchia, P. Lupetti, M. de Bernard, C. Pagliaccia, V. Pelicic, M. Charrel, C. Ulivieri, N. Norais, X. Ji, V. Cabiaux, E. Papini, R. Rappuoli, and J. L. Telford. 1999. 3D imaging of the 58 kDa cell binding subunit of the Helicobacter pylori cytotoxin. J. Mol. Biol. 290:459-470.
38. Reyrat, J. M., V. Pelicic, B. Gicquel, and R. Rappuoli. 1998. Counterselectable markers: untapped tools for bacterial genetics and pathogenesis. Infect. Immun. 66:4011-4017.
39. Ricci, V., A. Galmiche, A. Doye, V. Necchi, E. Solcia, and P. Boquet. 2000. High cell sensitivity to Helicobacter pylori VacA toxin depends on a GPI-anchored protein and is not blocked by inhibition of the clathrin-mediated pathway of endocytosis. Mol. Biol. Cell. 11:3897-3909.
40. Salama, N. R., G. Otto, L. Tompkins, and S. Falkow. 2001. Vacuolating cytotoxin of Helicobacter pylori plays a role during colonization in a mouse model of infection. Infect. Immun. 69:730-736.
41. Schiestl, R. H., and R. D. Gietz. 1989. High efficiency transformation of intact yeast cells using single stranded nucleic acids as a carrier. Curr. Genet. 16:339-346.
42. Sledziewski, A. Z., A. Bell, C. Yip, K. Kelsay, F. J. Grant, and V. L. MacKay. 1990. Superimposition of temperature regulation on yeast promoters. Methods Enzymol. 185:351-366.
43. Sundrud, M. S., V. J. Torres, D. Unutmaz, and T. L. Cover. 2004. Inhibition of primary human T cell proliferation by Helicobacter pylori vacuolating toxin (VacA) is independent of VacA effects on IL-2 secretion. Proc. Natl. Acad. Sci. USA 101:7727-7732.
44. Szabo, I., S. Brutsche, F. Tombola, M. Moschioni, B. Satin, J. L. Telford, R. Rappuoli, C. Montecucco, E. Papini, and M. Zoratti. 1999. Formation of anion-selective channels in the cell plasma membrane by the toxin VacA of Helicobacter pylori is required for its biological activity. EMBO J. 18:5517-5527.
45. Telford, J. L., P. Ghiara, M. Dell'Orco, M. Comanducci, D. Burroni, M. Bugnoli, M. F. Tecce, S. Censini, A. Covacci, Z. Xiang, E. Papini, C. Montecucco, L. Parente, and R. Rappuoli. 1994. Gene structure of the Helicobacter pylori cytotoxin and evidence of its key role in gastric disease. J. Exp. Med. 179:1653-1658.
46. Tombola, F., C. Carlesso, I. Szabo, M. de Bernard, J. M. Reyrat, J. L. Telford, R. Rappuoli, C. Montecucco, E. Papini, and M. Zoratti. 1999. Helicobacter pylori vacuolating toxin forms anion-selective channels in planar lipid bilayers: possible implications for the mechanism of cellular vacuolation. Biophys. J. 76:1401-1409.
47. Torres, V. J., M. S. McClain, and T. L. Cover. 2004. Interactions between p-33 and p-55 domains of the Helicobacter pylori vacuolating cytotoxin (VacA). J. Biol. Chem. 279:2324-2331.
48. Vinion-Dubiel, A. D., M. S. McClain, D. M. Czajkowsky, H. Iwamoto, D. Ye, P. Cao, W. Schraw, G. Szabo, S. R. Blanke, Z. Shao, and T. L. Cover. 1999. A dominant negative mutant of Helicobacter pylori vacuolating toxin (VacA) inhibits VacA-induced cell vacuolation. J. Biol. Chem. 274:37736-37742.
49. Wang, X., R. Wattiez, C. Pagliaccia, J. L. Telford, J. Ruysschaert, and V. Cabiaux. 2000 Membrane topology of VacA cytotoxin from H. pylori. FEBS Lett. 481:96-100.
50. Willhite, D. C., and S. R. Blanke. 2004. Helicobacter pylori vacuolating cytotoxin enters cells, localizes to the mitochondria, and induces mitochondrial membrane permeability changes correlated to toxin channel activity. Cell Microbiol. 6:143-154.
51. Willhite, D. C., T. L. Cover, and S. R. Blanke. 2003. Cellular vacuolation and mitochondrial cytochrome c release are independent outcomes of Helicobacter pylori vacuolating cytotoxin activity that are each dependent on membrane channel formation. J. Biol. Chem. 278:48204-48209.(Christophe Genisset, Cesi)
Venetian Institute of Molecular Medicine, Via Orus 2, 35129 Padova, Italy
Department of Evolutionary Biology, University of Siena, Via Aldo Moro 2, 53100 Siena, Italy
Dipartimento di Scienze Chimiche, Universita di Padova, via Marzolo 1, 35131 Padova, Italy
Dipartimento di Biologia, Universita di Padova, Via G. Colombo 3, 35121 Padova, Italy
ABSTRACT
Most Helicobacter pylori strains secrete a toxin (VacA) that causes massive vacuolization of target cells and which is a major virulence factor of H. pylori. The VacA amino-terminal region is required for the induction of vacuolization. The aim of the present study was a deeper understanding of the critical role of the N-terminal regions that are protected from proteolysis when VacA interacts with artificial membranes. Using a counterselection system, we constructed an H. pylori strain, SPM 326-49-57, that produces a mutant toxin with a deletion of eight amino acids in one of these protected regions. VacA 49-57 was correctly secreted by H. pylori but failed to oligomerize and did not have any detectable vacuolating cytotoxic activity. However, the mutant toxin was internalized normally and stained the perinuclear region of HeLa cells. Moreover, the mutant toxin exhibited a dominant negative effect, completely inhibiting the vacuolating activity of wild-type VacA. This loss of activity was correlated with the disappearance of oligomers in electron microscopy. These findings indicate that the deletion in VacA 49-57 disrupts the intermolecular interactions required for the oligomerization of the toxin.
INTRODUCTION
Helicobacter pylori is a gram-negative, microaerophilic, and spiral-shaped bacterium that colonizes the gastric mucosa of half of the world's population (26). This bacterium is associated with gastritis and peptic ulcers (4) and, for long-term chronic infections, the development of gastric carcinoma and mucosa-associated lymphoid tissue lymphoma (34).
Most H. pylori strains produce and secrete a toxin (VacA) which, together with two other virulence factors, CagA (8) and BabA (21), is significantly associated with strains that are isolated from patients with more severe disease (35). The most extensively characterized activity of VacA is its capacity to induce vacuolation in mammalian cells (9). These vacuoles contain markers for late endosomes and lysosomes and have an internal acidic pH (29, 32). Vacuolation is not the only activity of VacA, as it also increases the permeability of polarized epithelial cell monolayers to various ions and small uncharged molecules (33). After internalization by cells, VacA induces the release of cytochrome c into the cytosol, leading to apoptosis (11, 17, 50, 51). VacA also interferes with the immune system by altering the process of antigen presentation (30), and recently, it was shown to inhibit T-cell proliferation by two different mechanisms (6, 18, 43). Moreover, when administered intragastrically to mice, it causes gastric epithelial erosion (16, 45).
The vacA gene codes for a protoxin of 140 kDa, which is processed during export to the surface to yield a mature secreted toxin of 88 kDa (9, 45). Each molecule of VacA has two domains, p33 and p55, that are connected by a hydrophilic loop that can be cleaved after release from the bacterium. After cleavage, the two fragments remain noncovalently associated (45), indicating that the two subunits interact in such a way as to maintain the integrity of the 88-kDa structure. The p55 domain is responsible for host cell binding (37) and host tropism (31), whereas the p33 domain, together with the N-terminal 192 amino acids of p55, is sufficient to cause vacuolation when expressed in the cytoplasm of host cells (14).
Mature 88-kDa VacA monomers assemble into water-soluble, high-molecular weight-oligomers containing 6 to 7 or 12 to 14 copies of the mature toxin polypeptide (1, 10, 23). The purified oligomeric form of the toxin is biologically inactive, but after exposure to acidic pH, it disassembles into monomeric components (10); this permits insertion of the toxin into lipid membranes to form anion-conductive channels (44, 46). The VacA channel formation induces an osmotic imbalance of intracellular acidic compartments, leading to cytoplasmic vacuolation (46).
A previous study has reported that a mutant toxin of VacA, lacking the hydrophobic amino terminus region, has a dominant negative phenotype (48). This mutant, which lacks cytotoxic activity and completely inhibits the vacuolating activity of the wild-type toxin, is indistinguishable from wild-type VacA in its secretion, assembly into oligomeric structures, and uptake by HeLa cells. It was demonstrated later that this mutant lacks three tandem GXXXG motifs that are necessary for the dimerization of VacA within membranes and consequent anion channel formation (27).
In this study, we describe another dominant negative mutant, VacA 49-57, which has a deletion of eight amino acids in a region known to be protected from proteolysis following VacA interaction with artificial membranes (49). This mutant also inhibited the activity of wild-type VacA and failed to form oligomeric structures.
MATERIALS AND METHODS
Bacterial strains, yeast strains, and culture conditions. H. pylori CCUG 17874 was used as the source of VacA. H. pylori SPM 326, which encodes an s1m1-type vacA (2, 25), was the parental strain used for the construction of the mutant.
Colonies of H. pylori grown on blood agar plates (Columbia agar with 5% horse blood) were inoculated into Brucella broth containing 0.2% -cyclodextrin and were cultured for 2 days at 170 rpm in microaerophilic conditions. Yeast two-hybrid experiments were performed with the Saccharomyces cerevisiae EGY48 (MAT ura3 trp1 his3 6LexA-operator-LEU2) strain transformed with the plasmid pSH18-34 containing the lacZ reporter gene. Yeast strains were grown in synthetic minimal medium (SD) supplemented with the required amino acids at 30°C as previously described (19).
Construction of H. pylori SPM 326-49-57. First, a recipient strain was created by introduction of the kan-sacB cassette into the vacA gene of H. pylori SPM 326. The kan-sacB module, amplified from the plasmid pEnKSF (7), was inserted into the NcoI/EcoNI site of pBlueScriptKS p33, a plasmid containing a substantial portion of the vacA SPM 326 gene to create pKan/sacB (Fig. 1). The NcoI site, which is absent in the vacA gene, was added using the QuikChange site-directed mutagenesis kit (Stratagene) by changing the two nucleotides immediately before the sequence coding the translational start site. A kanamycin-resistant, sucrose-sensitive clone, in which the expression of the VacA molecule was inactivated, was selected. This strain, H. pylori SPM 326KO2, was then transformed by pBlueScriptKS p33-49-57, and a sucrose-resistant, kanamycin-sensitive transformant, which was obtained by homologous recombination between the two vacA fragments, was selected as summarized in Fig. 1 (7). The strain was analyzed by nucleotide sequencing to verify that the desired mutation was present.
The plasmid pBlueScriptKS p33-49-57 was obtained from pBlueScriptKS p33 by the introduction of an in-frame vacA deletion mutation using the QuikChange site-directed mutagenesis kit (Stratagene). Briefly, oligonucleotide primers, each being complementary to opposite strands of the vector and containing the deletion, were extended by PCR using Pfu DNA polymerase. Following thermal cycling, template DNA was eliminated by DpnI digestion and the PCR products were used to transform Escherichia coli XL1-Blue.
Purification of VacA and VacA 49-57. VacA from H. pylori strain CCUG 17874 was purified from the broth culture supernatant as described previously (24). For VacA 49-57, the biomass from a 1 liter culture of H. pylori SPM 326-49-57 was removed by centrifugation at 11,000 x g for 20 min. The supernatant was passed through a 0.2-μm filter and then applied at a flow rate of 1 ml/min to a column (5 by 3.5 cm, XK column; Amersham Pharmacia Biotech) containing Matrex Cellufine Sulfate (Amicon). The proteins were washed extensively with 100 mM NaCl-20 mM phosphate buffer (pH 6.5). VacA 49-57 was eluted from the column with 350 mM NaCl-20 mM phosphate buffer (pH 6.5). The mutant toxin was found to be highly pure at this stage; this was in contrast to the wild-type toxin which required a further step of purification by gel filtration in order to achieve the same purity. The eluate was brought to 50% saturation with ammonium sulfate and centrifuged at 20,000 x g for 30 min. The pelleted proteins were resuspended in phosphate-buffered saline (PBS) and dialyzed extensively against the same buffer. Purified toxins were stored at 4°C, and their concentrations were determined by using a microbicinchoninic acid assay (Pierce).
Cell vacuolation assay. HeLa cells were seeded at 7.5 x 104 ml–1 in 96-well plates in Dulbecco modified Eagle's medium containing 10% fetal calf serum and 2 mM glutamine at 37°C in 5% CO2 16 h before the assay. Before addition to cells, purified toxin preparations were acid-activated by adjusting to pH 2.0 for 5 min at room temperature and then neutralized as described previously (12). The extent of vacuolation was determined quantitatively by measuring the uptake of neutral red dye after the incubation of the toxin with cells at 37°C for 9 h in Dulbecco modified Eagle's medium containing 2% fetal calf serum and 15 mM ammonium chloride (9).
Deep-etch electron microscopy. VacA and VacA 49-57 molecules were prepared for microscopy by a procedure of absorption to mica, followed by quick-freezing and deep-etching (20). The samples were processed as described previously (23). For the experiment showing the effect of VacA 49-57 on VacA, an equimolar concentration of each toxin, acid activated when indicated, was mixed before the treatment.
Glycerol gradient centrifugation analysis. For analysis, 14-ml 10 to 30% glycerol gradients were prepared in either 60 mM Tris (pH 7.5) or 100 mM glycine (pH 3.0) containing 100 mM NaCl. Samples of 200 μl (dialyzed ammonium sulfate-precipitated proteins from broth culture supernatant diluted to an optical density at 600 nm [OD600] of 1) were layered on the gradients and centrifuged at 39,000 rpm for 10 h at 4°C in a SW40 Ti rotor (Beckman Instruments) as described previously (10). For the experiment showing the effect of VacA 49-57 on VacA, 100 μl of each concentrated supernatant, acid activated when indicated, was mixed before centrifugation. For the control, 100 μl of SPM 326 acid-activated concentrated supernatant was mixed with 100 μl of SPM 326KO2. Gradients were fractionated from the top by using a piston gradient fractionator (Biocomp). Proteins from the fractions were precipitated with 10% trichloroacetic acid and 0.02% sodium deoxycholate. The presence of VacA and VacA 49-57 was detected by Western blotting with rabbit anti-VacA serum (45).
Immunofluorescence and confocal microscopy. For indirect immunofluorescence analysis, HeLa cells were grown on chamber slides (Nunc). After 4 h of incubation with 5 μg/ml of purified VacA 49-57 and acid-activated purified VacA, the cells were fixed with 3.7% paraformaldehyde in PBS containing 5 mM Ca2+ and Mg2+ for 15 min. Aldehyde groups were quenched with 0.2 M glycine for 5 min. Plasma membrane-localized toxin was stained with a mouse monoclonal antibody specific for the toxin used at 2 μg/ml (36) followed by Texas Red goat anti-mouse antibody. For cytosolic staining, the previously surface-stained cells were then permeabilized with 0.2% Triton X-100 for 10 min at room temperature (22). PBS containing 5% fetal calf serum was used as a blocking solution. Labeling of the cytoplasmic toxin was obtained by using the same monoclonal antibody and fluorescein isothiocyanate (FITC) goat anti-mouse was used as secondary antibody. Slides were mounted with a SlowFade light antifade kit (Molecular Probes). Confocal images were obtained by using an Ultraview microscope (PerkinElmer).
Circular dichroism (CD) spectroscopy. Spectra for VacA (70 μg/ml) and VacA 49-57 (112 μg/ml) in PBS at pH 7.4, PBS at pH 3, and PBS after reneutralization were acquired with a Jasco J-715 spectropolarimeter at room temperature in a quartz cell with an optical path length of 0.1 cm. The spectra were recorded in the 195-to-260-nm wavelength range, using a 2-nm bandwidth and a 2-s time constant at a scan speed of 50 nm/min. All spectra are reported in terms of mean residue molar ellipticity []R degree cm2 dmol–1. Data processing was carried out using a J-700 software package. The signal-to-noise ratio was improved by the accumulation of at least 10 scans. CDNN 2.1 software was used to analyze the secondary structure content of the proteins (5).
Yeast two-hybrid system. Sequences encoding the wild-type p33, p33 49-57, and p55 VacA domains were PCR amplified from pBlueScriptKS p33, pBlueScriptKS p33-49-57, and pBlueScriptKS p55, respectively. PCR products were cloned into plasmids encoding the LexA DNA binding domain (pBD) and/or the transcription activation domain B42 (pAD) (15, 19). Primers were designed in order to amplify the region of vacA coding for amino acids 1 to 319 (p33 domain) and 320 to 837 (p55 domain) of the mature toxin from H. pylori SPM 326. Before transformation, the plasmids were analyzed by automated nucleotide sequencing.
The EGY48/pSH18-34 yeast strain was cotransformed with 2 μg of bait and prey containing plasmids in all possible combinations by using the lithium acetate method (41) and cultured at 30°C on SD without Trp, without His, and with Leu. In order to test which interactions between VacA domains could be detected in this system, and which combination was able to activate the transcription of the LEU2 and the lacZ reporter genes, several independent clones of each cotransformant were seeded on a selective plate. Three isolates of each cotransformant were then chosen for the -galactosidase interaction assay.
To quantify the level of interaction between the different domains, selected cotransformants were grown in 4 ml SD containing 2% galactose, 1% raffinose, and 50 μg/ml leucine at 30°C for 24 h until the culture reached an OD640 of 4. For each strain, a volume of culture containing 107 cells and corresponding to an OD640 of 2 was centrifuged at 3,000 x g, 5 min and washed once with 1 ml of Z buffer (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1.8 mM MgSO4, 0.27% -mercaptoethanol, pH 7) (42) and the pellet was resuspended in 150 μl of Z buffer. Fifty-microliter aliquots of this cell suspension were permeabilized by using Z buffer containing 70% CHCl3 and 0.02% sodium dodecyl sulfate and incubated with 700 μl of o-nitrophenyl--D-galactopyranoside (2 mg/ml) at 30°C. The reaction was stopped at selected time intervals by the addition of 500 μl of 1 M K2CO3, and the reaction time was recorded. The OD420 values of clarified reaction supernatants were measured. The -galactosidase activity was calculated using the Miller equation (19). The values presented here are the averages of three independent cotransformants for each combination of bait and prey that were assessed in three different experiments.
RESULTS
VacA 49-57 is devoid of vacuolating activity. A study of the protection of VacA from protease digestion, following its membrane adsorption, suggested that its N-terminal region is involved in VacA membrane interaction (49). On the other hand, the cell vacuolating activity of VacA is due to its capability of forming trans-membrane anion-selective channels (44, 46). Therefore, we decided to investigate the role of the N terminus of VacA. On the basis of a preliminary in silico analysis, we selected the region of amino acids 49 to 57 as a putative membrane interactive segment and we used a counterselection marker (7, 38) to introduce an in-frame deletion of eight amino acids into the chromosomal vacA gene of H. pylori SPM326 (Fig. 1). Western blot analysis of soluble extracts and supernatants showed that the mutant strain, H. pylori SPM 326-49-57, produces and secretes a toxin of the expected molecular mass for the mature polypeptide (data not shown). However, the VacA 49-57 protein behaved differently in affinity chromatography from the wild-type protein. VacA 49-57 eluted from a Matrex Cellufine Sulfate column with 350 mM NaCl, whereas wild-type VacA needs a higher salt concentration (24). This difference of affinity for the Matrex Cellufine Sulfate resin implies that the mutant toxin has different properties relative to those of wild-type VacA.
The amino acids that were deleted are located in the p33 domain, which is known to be essential for vacuolating activity (Fig. 1). Purified VacA 49-57 was tested for activity in a neutral red uptake assay using HeLa cells. As shown in Fig. 1C, the mutant toxin did not induce vacuoles.
VacA 49-57 enters epithelial cells. VacA binds epithelial cells and is then internalized into the cell cytosol, where it displays its activity (13). We investigated whether the lack of vacuolation that was observed was due to the incapacity of VacA 49-57 to be internalized by HeLa cells. After 4 h of incubation, cells treated with either the wild-type or the mutant toxin were fixed and immunostained with a monoclonal antibody, C1G9, which is specific for VacA and Texas Red-labeled anti-mouse immunoglobulin antibodies. The cells were then permeabilized and restained with the same monoclonal antibody and a FITC-labeled secondary antibody. Intracellular VacA 49-57 was observed as green-labeled spots in the perinuclear region of cells which was comparable to the staining that was obtained in cells incubated with wild-type VacA (Fig. 2). Hence, the mutant toxin internalized into HeLa cells in a manner similar to that of the wild-type protein.
VacA 49-57 fails to form the oligomeric structure. Purified VacA forms water-soluble oligomeric structures which can be observed by transmission electron microscopy after a quick-freeze, deep-etch preparation (23). Since VacA 49-57 has a lower affinity for the Matrex Cellufine Sulfate resin, we checked the integrity of the mutant toxin. Transmission electron microscope visualization of purified wild-type VacA revealed the expected structures with either six- or sevenfold radial symmetry. These structures were not observed in preparations of VacA 49-57 (Fig. 3). After treatment at low pH, VacA oligomers disassemble into monomers and the two structures can be separated by velocity sedimentation on glycerol density gradients (10). As expected, wild-type VacA from strain SPM 326 sedimented at the bottom of the gradient, reflecting the high-molecular-weight oligomeric structure that is visible in the electron micrographs (Fig. 3). At pH 3.0, VacA was found at the top of the gradient. VacA 49-57 was found at the top of the gradient at both neutral and acid pH values. This demonstrates that VacA 49-57 is defective in oligomerization at a neutral pH and that the structure observed by transmission electron microscopy is likely to be a monomer. A very small quantity of VacA 49-57 was found at the bottom of the gradient at a neutral pH. This may imply that the mutant toxin does form oligomers, although it does so very inefficiently. However, no oligomers were ever observed in electron micrographs of the mutant toxin.
The deletion of amino acids 49 to 57 affects the secondary structure content of VacA and its sensitivity to low pH. We used far-UV circular dichroism spectroscopy to determine whether the eight-amino-acid deletion 49-57 affects the integrity of VacA. As shown in Fig. 4A, the spectrum of VacA at pH 7.4 displays a negative band at around 220 nm and has a substantial amount of -like secondary structure (44%) as previously described (12). VacA 49-57 has a lower -sheet structure content (33%) compared to that of VacA, as revealed by the presence of two negative bands in the CD spectrum, one that was intense at 206 nm and one that was weaker at approximately 220 nm (Fig. 4A).
Upon acidification (pH 3), the negative band of the VacA spectrum slightly shifted to a lower wavelength, suggesting a small change in the secondary or quaternary structure content, but returned to the same profile after neutralization (Fig. 4B). For the mutant toxin, the signal was markedly weaker after acid exposure and showed a significant change in the profile. In fact, VacA 49-57 exhibited an irreversible conformation change (Fig. 4C). Hence, the eight-amino-acid deletion that was introduced in the p33 domain decreased the -sheet structure content but also rendered the toxin sensitive to low pH.
VacA 49-57 is a dominant negative mutant. The vacuolating activity of the native toxin can be abolished by introducing mutated VacA into the oligomer (48). We investigated whether mixing wild-type VacA with VacA 49-57 can interfere with the formation of vacuoles. HeLa cells were incubated with a fixed concentration of acid-activated wild-type VacA, to which increasing concentrations of VacA 49-57 were added prior to incubation with the cells. VacA 49-57 clearly affects the cytotoxic activity of wild-type VacA in a dose-dependent manner (Fig. 5). When the two toxins were present in equimolar concentrations, VacA 49-57 completely inhibited the activity of the native protein. Furthermore, significant inhibition was achieved with a 7:1 molar ratio of wild-type to mutant toxin. Hence, in this experimental condition, VacA 49-57 dominantly inhibited the vacuolating activity of wild-type VacA. Heat-inactivated VacA 49-57 (10 min at 95°C) was also tested at an equimolar concentration, but it did not present any inhibitory effect (data not shown). Non-acid-activated VacA 49-57 completely inhibited the vacuolating activity of wild-type VacA, whereas acid-activated VacA 49-57 reduced the cytotoxic activity by only 50% at the same relative concentration. This is likely due to the denaturing effect of low pH on the mutant toxin observed in the CD spectra.
VacA 49-57 inhibits wild-type VacA oligomer reformation. Disassembly of VacA oligomers by treatment at a low pH is reversible (10). After neutralization, the VacA monomers reassemble into oligomers. We sought to determine whether VacA 49-57 could interfere with the reassembly of wild-type VacA oligomers. A solution with the same concentrations of VacA 49-57 and wild-type VacA was examined by transmission electron microscopy after a quick-freeze, deep-etch preparation. The electron micrograph revealed two structures, the oligomer characteristic of native VacA and a smaller structure, corresponding to the VacA 49-57 monomer (Fig. 6A), also observed in Fig. 3. Therefore, the mutant toxin had no effect on the structure of the wild-type toxin oligomers. When the wild-type VacA oligomers were disrupted at low pH and then neutralized in the presence of the mutant toxin, the wild-type toxin failed to reassemble (Fig. 6B). In order to prove that the absence of oligomers was due to VacA 49-57 and not to acid treatment or to contaminants in the VacA 49-57 preparation, purified VacA treated at low pH and then neutralized in the presence of PBS was also submitted to a quick-freeze, deep-etch preparation. As shown in Fig. 6C, this procedure yielded flower-shaped structures similar to those of the VacA oligomers already described (10). Furthermore, glycerol gradient sedimentation experiments on concentrated culture supernatant (Fig. 6D to F) confirmed that VacA 49-57 is responsible for the failure of wild-type VacA to reassemble since supernatant of SPM 326-49-57 blocked the reassembly of wild-type VacA, whereas supernatant from a strain lacking the vacA gene (SPM 326KO2) did not (Fig. 6F). In conclusion, our results indicate that the loss of vacuolating cytotoxic activity of wild-type VacA in the presence of VacA 49-57 is correlated with the inability of the toxin to reassemble into oligomeric structures.
Interaction of p33/p55 and p3349-57/p55 in a yeast two-hybrid system. The mature 88-kDa VacA monomer is sensitive to proteolytic cleavage at a hydrophilic loop connecting the p33 and p55 domains which, however, remain associated in the cleaved protein (45). Hence, p33 binds to p55 in the absence of covalent linkage. Furthermore, analysis of the three-dimensional electron microscopy structure of VacA and a truncated molecule lacking the p33 domain led to a model which suggested that the oligomeric structure is formed by the interaction of the p33 moiety with the p55 moiety of the adjacent monomer (37). Therefore, p33 and p55 have two sites of interaction, one which maintains the integrity of the monomer and the other that is involved in intermolecular interactions in the oligomer. Evidence for the interaction of p33 with p55 comes from experiments using the yeast two-hybrid system (47). However, these experiments could not distinguish between p33/p55 interactions involved in maintaining the structure of the mature monomeric polypeptide (45) and the interactions between p33 of one monomer with p55 in the adjacent monomer in the oligomeric structure (37). Since the 49-57 deletion in the p33 domain appears to affect the interaction between the VacA domains that are necessary for the formation of oligomers, we decided to use this system to test whether the interactions between the p33 and p55 domain were compromised by the deletion. The vacA sequence encoding p33 was cloned into a plasmid containing the DNA binding domain of LexA (pBD), whereas p55 was cloned into a plasmid containing the B42 transcription activation domain (pAD) (19). The EGY48 yeast strain carrying the lacZ reporter plasmid (pSH18-34) was transformed with plasmids pBD-p33 and pAD-p55, and cotransformants were selected on specific medium. Three clones were chosen at random for further analysis. Interaction between wild-type p33 and p55 in this system was confirmed by measuring the activation of the lacZ reporter gene. The level of interaction determined by using the -galactosidase liquid assay on total yeast extracts is reported in Fig. 7. This combination was used as a positive control for further investigation of p33 49-57/p55 interactions. As shown, the 49-57 deletion in the p33 domain caused a 50% decrease in the interaction with p55 (Fig. 7). This result is consistent with a loss of the p33/p55 interactions involved in oligomerization without interfering with the interactions involved in maintaining the monomeric structure after cleavage of the hydrophilic loop between p33 and p55. None of the LexA fusion constructs (pBD-p33 and pBD-p33 49-57) were able to activate the expression of the reporter gene in the absence of the pAD-p55 hybrid. Similarly, extracts from the yeast strain that were transformed with only the p55 moiety fused to the activating domain showed very low -galactosidase activity in absence of the other hybrid.
DISCUSSION
Purified VacA assembles in high-molecular-weight oligomers which have flower-like structures when observed by quick-freeze, deep-etch electron microscopy. Mutational analysis of p33 has shown that the region spanning amino acids 28 to 196 is important for interaction between two adjacent monomers and formation of the global structure (47, 48). In this study, we show that the deletion of amino acids 49 to 57 of p33 disrupts this interaction.
The mutated VacA protein described in this study (VacA 49-57) was produced and secreted by H. pylori in a manner similar to that of the wild-type toxin, but it lacked vacuolating cytotoxic activity. This absence of activity correlated with its lack of oligomerization. In agreement with previous work (37, 48), this finding indicates that VacA oligomerization is essential for vacuole formation, although it cannot be ruled out that the mutant toxin is defective in some other way as, for example, in the interaction with cytoplasmic target cell proteins. It is noteworthy that this is the smallest deletion which has been found to be capable of blocking the oligomerization of VacA. However, since the deletion results in a slightly altered CD spectrum, we cannot conclude that this region is directly involved in oligomerization.
A previous study showed that the p55 molecule was able to interact with the surface of target cells but was not internalized (37). This suggests that either the p33 subunit or the ability to form an oligomeric structure is required for cell entry. Here we show that the VacA 49-57 monomer has the capacity to enter HeLa cells and localize in the cytoplasm in a manner similar to that observed in VacA-infected cells (22), demonstrating that oligomerization is not necessary for internalization. This mechanism differs from that described for the C2 toxin of Clostridium botulinum, a related channel-forming molecule which needs to form oligomers for cellular uptake (3). We propose that a region of p33, one that is different from the region deleted in this study and still unknown, must be involved in VacA internalization.
A remarkable property of VacA 49-57 is its capacity to inhibit the cytotoxic activity of the wild-type toxin. It is unlikely that this is due to competition for the binding to the VacA receptor. The p55 domain has been shown to be responsible for binding to the host cell and can fold independently of p33 (37), therefore VacA 49-57 is expected to bind the VacA receptor with the same affinity as the wild-type toxin. When the two proteins were mixed in equimolar concentrations, a complete inhibition of cytotoxic activity was observed. Thus the inhibition is unlikely to be due to competition for the receptor, particularly since HeLa cells exhibit a high level of nonspecific binding for radiolabeled VacA. Only a small reduction in the activity was observed in the presence of a 100-fold excess of unlabeled VacA (28, 39).
VacA 49-57 was able to abolish the cytotoxic activity of wild-type VacA in a manner similar to that of VacA 6-27, another dominant negative mutant of H. pylori-vacuolating toxin (48). However, VacA 6-27 presented a dominant negative phenotype only when it was acid activated, while VacA 49-57 was able to completely inhibit wild-type VacA activity even without acid activation. The inhibition of activity was also significant at a molar ratio of 7:1 for wild-type to mutant toxin, suggesting that one mutant molecule per oligomeric complex is sufficient to cause substantial inhibition of vacuolating activity. VacA 6-7 forms an oligomer and, after exposure to low pH, monomers of VacA 6-7 can interact with monomers of VacA to form dysfunctional mixed oligomers (48). In contrast, VacA 49-57 does not oligomerize and does not need to be acid activated to interact with wild-type VacA monomers. In fact, acid-activated VacA 49-57 inhibited only 50% of the wild-type VacA. Accordingly, the far-UV CD data indicate that the lower inhibition observed with acid-activated VacA 49-57 results from structural damage to the mutant toxin after exposure to acidic pH.
In a recent study, using a FLAG-VacA toxin, which can be cleaved into the p33 and p55 domains, it has been demonstrated that the two fragments remain physically associated after proteolytic cleavage and were still able to form a VacA oligomeric structure (47). This suggests that there are two types of interactions between the two domains of VacA: intramolecular interactions between the p33 and p55 domains of an individual VacA monomer and intermolecular interactions between p33 and p55 of different VacA molecules, which are necessary for the formation of oligomeric structures. As shown by the yeast two-hybrid system, the deletion of eight amino acids in the p33 domain reduced the level of interaction with the p55 domain by 50%, consistent with the loss of the latter but not the former p33/p55 interaction. We propose that p33 49-57 lacks the capacity to interact with the p55 domain of another monomer. Thus, after acid activation, the p33 domain of monomers of the wild-type toxin may interact with the p55 domain of VacA 49-57; however, the reduced ability of the p33 domain of VacA 49-57 to interact with the wild-type toxin would prevent the complete formation of the active oligomer in the membrane. This would explain why inhibition could be achieved at even low mutant-to-wild-type toxin ratios.
Over the last 10 years, many studies have shown that due to the formation of a VacA anion-selective channel, the toxin causes multiple effects on target cells in vitro (6, 9, 11, 17, 18, 30, 43, 50, 51) but the in vivo function of VacA and its relevance for H. pylori initial colonization of the stomach remains unclear (40). Since VacA 49-57 is produced and correctly secreted by H. pylori and has the capacity to enter epithelial cells without inducing vacuolation, it could be a useful tool to understanding the role of VacA in H. pylori infection.
ACKNOWLEDGMENTS
C. Genisset was supported by Marie Curie Industry Host Fellowship QLK4 1999 50407.
We thank R. Janulczyk and C. Montecucco for critically reading the manuscript. We are grateful to D. Bison and E. Pasqualetto for technical assistance with the confocal microscope and CD spectroscopy respectively. We thank G. Corsi for the artwork and S. Pasquini, L. Fini, and S. Magi for the medium preparation.
REFERENCES
1. Adrian, M., T. L. Cover, J. Dubochet, and J. E. Heuser. 2002. Multiple oligomeric states of the Helicobacter pylori vacuolating toxin demonstrated by cryo-electron microscopy. J. Mol. Biol. 318:121-133.
2. Atherton, J. C., P. Cao, R. M. Peek, Jr., M. K. Tummuru, M. J. Blaser, and T. L. Cover. 1995. Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori. Association of specific vacA types with cytotoxin production and peptic ulceration. J. Biol. Chem. 270:17771-17777.
3. Barth, H., D. Blocker, J. Behlke, W. Bergsma-Schutter, A. Brisson, R. Benz, and K. Aktories. 2000. Cellular uptake of Clostridium botulinum C2 toxin requires oligomerization and acidification. J. Biol. Chem. 275:18704-18711.
4. Blaser, M. J. 1990. Helicobacter pylori and the pathogenesis of gastroduodenal inflammation. J. Infect. Dis. 161:626-633.
5. Bohm, G., R. Muhr, and R. Jaenicke. 1992. Quantitative analysis of protein far UV circular dichroism spectra by neural networks. Protein Eng. 5:191-195.
6. Boncristiano, M., S. R. Paccani, S. Barone, C. Ulivieri, L. Patrussi, D. Ilver, A. Amedei, M. M. D'Elios, J. L. Telford, and C. T. Baldari. 2003. The Helicobacter pylori vacuolating toxin inhibits T cell activation by two independent mechanisms. J. Exp. Med. 198:1887-1897.
7. Copass, M., G. Grandi, and R. Rappuoli. 1997. Introduction of unmarked mutations in the Helicobacter pylori vacA gene with a sucrose sensitivity marker. Infect. Immun. 65:1949-1952.
8. Covacci, A., S. Censini, M. Bugnoli, R. Petracca, D. Burroni, G. Macchia, A. Massone, E. Papini, Z. Xiang, N. Figura, and R. Rappuoli. 1993. Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proc. Natl. Acad. Sci. USA 90:5791-5795.
9. Cover, T. L., and M. J. Blaser. 1992. Purification and characterization of the vacuolating toxin from Helicobacter pylori. J. Biol. Chem. 267:10670-10675.
10. Cover, T. L., P. I. Hanson, and J. E. Heuser. 1997. Acid-induced dissociation of VacA, the Helicobacter pylori vacuolating cytotoxin, reveals its pattern of assembly. J. Cell Biol. 138:759-769.
11. Cover, T. L., U. S. Krishna, D. A. Israel, and R. M. Peek, Jr. 2003. Induction of gastric epithelial cell apoptosis by Helicobacter pylori vacuolating cytotoxin. Cancer Res. 63:951-957.
12. de Bernard, M., E. Papini, V. de Filippis, E. Gottardi, J. Telford, R. Manetti, A. Fontana, R. Rappuoli, and C. Montecucco. 1995. Low pH activates the vacuolating toxin of Helicobacter pylori, which becomes acid and pepsin resistant. J. Biol. Chem. 270:23937-23940.
13. de Bernard, M., B. Arico, E. Papini, R. Rizzuto, G. Grandi, R. Rappuoli, and C. Montecucco. 1997. Helicobacter pylori toxin VacA induces vacuole formation by acting in the cell cytosol. Mol. Microbiol. 26:665-674.
14. de Bernard, M., D. Burroni, E. Papini, R. Rappuoli, J. Telford, and C. Montecucco. 1998. Identification of the Helicobacter pylori VacA toxin domain active in the cell cytosol. Infect. Immun. 66:6014-6016.
15. Di Giaimo, R., M. Riccio, S. Santi, C. Galeotti, D. C. Ambrosetti, and M. Melli. 2002. New insights into the molecular basis of progressive myoclonus epilepsy: a multiprotein complex with cystatin B. Hum. Mol. Genet. 11:2941-2950.
16. Fujikawa, A., D. Shirasaka, S. Yamamoto, H. Ota, K. Yahiro, M. Fukada, T. Shintani, A. Wada, N. Aoyama, T. Hirayama, H. Fukamachi, and M. Noda. 2003. Mice deficient in protein tyrosine phosphatase receptor type Z are resistant to gastric ulcer induction by VacA of Helicobacter pylori. Nat. Genet. 33:375-381.
17. Galmiche, A., J. Rassow, A. Doye, S. Cagnol, J. C. Chambard, S. Contamin, V. de Thillot, I. Just, V. Ricci, E. Solcia, E. Van Obberghen, and P. Boquet. 2000. The N-terminal 34 kDa fragment of Helicobacter pylori vacuolating cytotoxin targets mitochondria and induces cytochrome c release. EMBO J. 19:6361-6370.
18. Gebert, B., W. Fischer, E. Weiss, R. Hoffmann, and R. Haas. 2003. Helicobacter pylori vacuolating cytotoxin inhibits T lymphocyte activation. Science 301:1099-1102.
19. Golemis, E., and I. Serebriiski. 1998. Two-hybrid system/interaction trap, p. 69-69.40. In D. L. Spector, R. D. Goldman, and L. A. Leinwand (ed.), Cells: a laboratory manual, vol. 1. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
20. Heuser, J. E. 1983. Procedure for freeze-drying molecules adsorbed to mica flakes. J. Mol. Biol. 169:155-195.
21. Ilver, D., A. Arnqvist, J. Ogren, I. M. Frick, D. Kersulyte, E. T. Incecik, D. E. Berg, A. Covacci, L. Engstrand, and T. Boren. 1998. Helicobacter pylori adhesin binding fucosylated histo-blood group antigens revealed by retagging. Science 279:373-377.
22. Ilver, D., S. Barone, D. Mercati, P. Lupetti, and J. L. Telford. 2004. Helicobacter pylori toxin VacA is transferred to host cells via a novel contact-dependent mechanism. Cell. Microbiol. 6:167-174.
23. Lupetti, P., J. E. Heuser, R. Manetti, P. Massari, S. Lanzavecchia, P. L. Bellon, R. Dallai, R. Rappuoli, and J. L. Telford. 1996. Oligomeric and subunit structure of the Helicobacter pylori vacuolating cytotoxin. J. Cell Biol. 133:801-807.
24. Manetti, R., P. Massari, D. Burroni, M. de Bernard, A. Marchini, R. Olivieri, E. Papini, C. Montecucco, R. Rappuoli, and J. L. Telford. 1995. Helicobacter pylori cytotoxin: importance of native conformation for induction of neutralizing antibodies. Infect. Immun. 63:4476-4480.
25. Marchetti, M., B. Arico, D. Burroni, N. Figura, R. Rappuoli, and P. Ghiara. 1995. Development of a mouse model of Helicobacter pylori infection that mimics human disease. Science 267:1655-1658.
26. Marshall, B. J., and J. R. Warren. 1984. Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet i:1311-1315.
27. McClain, M. S., H. Iwamoto, P. Cao, A. D. Vinion-Dubiel, Y. Li, G. Szabo, Z. Shao, and T. L. Cover. 2003. Essential role of a GXXXG motif for membrane channel formation by Helicobacter pylori vacuolating toxin. J. Biol. Chem. 278:12101-12108.
28. McClain, M. S., W. Schraw, V. Ricci, P. Boquet, and T. L. Cover. 2000. Acid activation of Helicobacter pylori vacuolating cytotoxin (VacA) results in toxin internalization by eukaryotic cells. Mol. Microbiol. 37:433-442.
29. Molinari, M., C. Galli, N. Norais, J. L. Telford, R. Rappuoli, J. P. Luzio, and C. Montecucco. 1997. Vacuoles induced by Helicobacter pylori toxin contain both late endosomal and lysosomal markers. J. Biol. Chem. 272:25339-25344.
30. Molinari, M., M. Salio, C. Galli, N. Norais, R. Rappuoli, A. Lanzavecchia, and C. Montecucco. 1998. Selective inhibition of Ii-dependent antigen presentation by Helicobacter pylori toxin VacA. J. Exp. Med. 187:135-140.
31. Pagliaccia, C., M. de Bernard, P. Lupetti, X. Ji, D. Burroni, T. L. Cover, E. Papini, R. Rappuoli, J. L. Telford, and J. M. Reyrat. 1998. The m2 form of the Helicobacter pylori cytotoxin has cell type-specific vacuolating activity. Proc. Natl. Acad. Sci. USA 95:10212-10217.
32. Papini, E., B. Satin, C. Bucci, M. de Bernard, J. L. Telford, R. Manetti, R. Rappuoli, M. Zerial, and C. Montecucco. 1997. The small GTP binding protein rab7 is essential for cellular vacuolation induced by Helicobacter pylori cytotoxin. EMBO J. 16:15-24.
33. Papini, E., B. Satin, N. Norais, M. de Bernard, J. L. Telford, R. Rappuoli, and C. Montecucco. 1998. Selective increase of the permeability of polarized epithelial cell monolayers by Helicobacter pylori vacuolating toxin. J. Clin. Investig. 102:813-820.
34. Parsonnet, J., S. Hansen, L. Rodriguez, A. B. Gelb, R. A. Warnke, E. Jellum, N. Orentreich, J. H. Vogelman, and G. D. Friedman. 1994. Helicobacter pylori infection and gastric lymphoma. N. Engl. J. Med. 330:1267-1271.
35. Prinz, C., M. Schoniger, R. Rad, I. Becker, E. Keiditsch, S. Wagenpfeil, M. Classen, T. Rosch, W. Schepp, and M. Gerhard. 2001. Key importance of the Helicobacter pylori adherence factor blood group antigen binding adhesin during chronic gastric inflammation. Cancer Res. 61:1903-1909.
36. Reyrat, J. M., M. Charrel, C. Pagliaccia, D. Burroni, P. Lupetti, M. de Bernard, X. Ji, N. Norais, E. Papini, R. Dallai, R. Rappuoli, and J. L. Telford. 1998. Characterisation of a monoclonal antibody and its use to purify the cytotoxin of Helicobacter pylori. FEMS Microbiol. Lett. 165:79-84.
37. Reyrat, J. M., S. Lanzavecchia, P. Lupetti, M. de Bernard, C. Pagliaccia, V. Pelicic, M. Charrel, C. Ulivieri, N. Norais, X. Ji, V. Cabiaux, E. Papini, R. Rappuoli, and J. L. Telford. 1999. 3D imaging of the 58 kDa cell binding subunit of the Helicobacter pylori cytotoxin. J. Mol. Biol. 290:459-470.
38. Reyrat, J. M., V. Pelicic, B. Gicquel, and R. Rappuoli. 1998. Counterselectable markers: untapped tools for bacterial genetics and pathogenesis. Infect. Immun. 66:4011-4017.
39. Ricci, V., A. Galmiche, A. Doye, V. Necchi, E. Solcia, and P. Boquet. 2000. High cell sensitivity to Helicobacter pylori VacA toxin depends on a GPI-anchored protein and is not blocked by inhibition of the clathrin-mediated pathway of endocytosis. Mol. Biol. Cell. 11:3897-3909.
40. Salama, N. R., G. Otto, L. Tompkins, and S. Falkow. 2001. Vacuolating cytotoxin of Helicobacter pylori plays a role during colonization in a mouse model of infection. Infect. Immun. 69:730-736.
41. Schiestl, R. H., and R. D. Gietz. 1989. High efficiency transformation of intact yeast cells using single stranded nucleic acids as a carrier. Curr. Genet. 16:339-346.
42. Sledziewski, A. Z., A. Bell, C. Yip, K. Kelsay, F. J. Grant, and V. L. MacKay. 1990. Superimposition of temperature regulation on yeast promoters. Methods Enzymol. 185:351-366.
43. Sundrud, M. S., V. J. Torres, D. Unutmaz, and T. L. Cover. 2004. Inhibition of primary human T cell proliferation by Helicobacter pylori vacuolating toxin (VacA) is independent of VacA effects on IL-2 secretion. Proc. Natl. Acad. Sci. USA 101:7727-7732.
44. Szabo, I., S. Brutsche, F. Tombola, M. Moschioni, B. Satin, J. L. Telford, R. Rappuoli, C. Montecucco, E. Papini, and M. Zoratti. 1999. Formation of anion-selective channels in the cell plasma membrane by the toxin VacA of Helicobacter pylori is required for its biological activity. EMBO J. 18:5517-5527.
45. Telford, J. L., P. Ghiara, M. Dell'Orco, M. Comanducci, D. Burroni, M. Bugnoli, M. F. Tecce, S. Censini, A. Covacci, Z. Xiang, E. Papini, C. Montecucco, L. Parente, and R. Rappuoli. 1994. Gene structure of the Helicobacter pylori cytotoxin and evidence of its key role in gastric disease. J. Exp. Med. 179:1653-1658.
46. Tombola, F., C. Carlesso, I. Szabo, M. de Bernard, J. M. Reyrat, J. L. Telford, R. Rappuoli, C. Montecucco, E. Papini, and M. Zoratti. 1999. Helicobacter pylori vacuolating toxin forms anion-selective channels in planar lipid bilayers: possible implications for the mechanism of cellular vacuolation. Biophys. J. 76:1401-1409.
47. Torres, V. J., M. S. McClain, and T. L. Cover. 2004. Interactions between p-33 and p-55 domains of the Helicobacter pylori vacuolating cytotoxin (VacA). J. Biol. Chem. 279:2324-2331.
48. Vinion-Dubiel, A. D., M. S. McClain, D. M. Czajkowsky, H. Iwamoto, D. Ye, P. Cao, W. Schraw, G. Szabo, S. R. Blanke, Z. Shao, and T. L. Cover. 1999. A dominant negative mutant of Helicobacter pylori vacuolating toxin (VacA) inhibits VacA-induced cell vacuolation. J. Biol. Chem. 274:37736-37742.
49. Wang, X., R. Wattiez, C. Pagliaccia, J. L. Telford, J. Ruysschaert, and V. Cabiaux. 2000 Membrane topology of VacA cytotoxin from H. pylori. FEBS Lett. 481:96-100.
50. Willhite, D. C., and S. R. Blanke. 2004. Helicobacter pylori vacuolating cytotoxin enters cells, localizes to the mitochondria, and induces mitochondrial membrane permeability changes correlated to toxin channel activity. Cell Microbiol. 6:143-154.
51. Willhite, D. C., T. L. Cover, and S. R. Blanke. 2003. Cellular vacuolation and mitochondrial cytochrome c release are independent outcomes of Helicobacter pylori vacuolating cytotoxin activity that are each dependent on membrane channel formation. J. Biol. Chem. 278:48204-48209.(Christophe Genisset, Cesi)