Association of Helicobacter pylori vacuolating toxin (VacA) with lipid rafts

A variety of extracellular ligands and pathogens interact with raft domains in the plasma membrane of eukaryotic cells. In this study, we examined the role of lipid rafts and raft-associated glycosylphosphatidylinositol (GPI)-anchored proteins in the process by which Helicobacter pylori vacuolating toxin (VacA) intoxicates cells. We first investigated whether GPI-anchored proteins are required for VacA toxicity by analyzing wild-type Chinese hamster ovary (CHO) cells and CHO-LA1 mutant cells that are defective in production of GPI-anchored proteins. Whereas wild-type and mutant cells differed markedly in susceptibility to aerolysin (a bacterial toxin that binds to GPI-anchored proteins), they were equally susceptible to VacA. We next determined whether VacA physically associates with lipid rafts. CHO or HeLa cells were incubated with VacA, and Triton-insoluble membranes then were separated by sucrose density gradient centrifugation. Immunoblot analysis revealed that a substantial proportion of cell-associated toxin was associated with detergent-resistant membranes (DRMs). DRM association required acid activation of the purified toxin prior to contact with cells, and acid activation also was required for VacA cytotoxicity. Treatment of cells with methyl-beta-cyclodextrin (a cholesterol-depleting agent) did not inhibit VacA-induced depolarization of the plasma membrane, but interfered with the internalization or intracellular localization of VacA and inhibited the capacity of the toxin to induce cell vacuolation. Treatment of cells with nystatin also inhibited VacA-induced cell vacuolation. These data indicate that VacA associates with lipid raft microdomains in the absence of GPI-anchored proteins and suggest that association of the toxin with lipid rafts is important for VacA cytotoxicity.     

Enhanced internalization of ricin in nigericin-pretreated Chinese hamster ovary cells.

Biochemical and electron microscopic autoradiographic studies with [125I] ricin have revealed that nigericin-pretreated Chinese hamster ovary cells are more efficient than untreated cells in the internalization of the toxin into the cells. These results suggest that the enhanced rate of internalization of ricin in nigericin-pretreated cells may account for the enhancement of cytotoxicity of ricin in Chinese hamster ovary cells by nigericin.     

Receptor-mediated endocytosis of a ricin-colloidal gold conjugate in vero cells. Intracellular routing to vacuolar and tubulo-vesicular portions of the endosomal system

We have prepared a conjugate (Ri-Au) of the toxic plant protein ricin and colloidal gold (particle size 5 nm) and used it for internalization studies in monolayer cultures of Vero cells. The Ri-Au conjugate was very stable, with only little release of ricin ([125I]Ri) from the gold particles within a pH range of 4.5-8.0. Within 2 h at 37 degrees C, only very little intracellular degradation of the ricin preparation ([125I]Ri-Au) occurred. The cells bound the same proportion of native ricin ([125I]Ri) and Ri-Au from the medium, and the kinetics of toxicity (decrease in cellular incorporation of [3H]leucine) of [125I]Ri and [125I]Ri-Au were also comparable. At 4 degrees C, the cell-surface binding of Ri-Au was continuous and distinct, as revealed by electron microscopy. This binding was specific, since almost no Ri-Au surface binding occurred at 4 degrees C in the presence of 0.1 M lactose or 1 mg/ml native (unlabelled) ricin. Within the first 30 min of warming prelabelled cells to 37 degrees C, the amount of surface-associated Ri-Au decreased considerably (from 150 to 60 gold particles per micron cell surface in 40 nm sections). Coated pits and vesicles were involved in the internalization of Ri-Au, and within 5-30 min at 37 degrees C Ri-Au had been delivered to vacuolar and tubulo-vesicular portions of the endosomal system, and later also to lysosomes. Analysis of very thin (ca 20 nm) serial sections revealed that most of the tubulo-vesicular elements were separate structures not connected to the membrane of the vacuolar portion. Data here presented indicate that our ricin conjugate, like many "physiological' ligands and viruses, is internalized by receptor-mediated endocytosis via the coated pit-endosomal pathway.     

Insulin receptors internalize by a rapid, saturable pathway requiring receptor autophosphorylation and an intact juxtamembrane region

The effect of receptor occupancy on insulin receptor endocytosis was examined in CHO cells expressing normal human insulin receptors (CHO/IR), autophosphorylation- and internalization-deficient receptors (CHO/IRA1018), and receptors which undergo autophosphorylation but lack a sequence required for internalization (CHO/IR delta 960). The rate of [125I]insulin internalization in CHO/IR cells at 37 degrees C was rapid at physiological concentrations, but decreased markedly in the presence of increasing unlabeled insulin (ED50 = 1-3 nM insulin, or 75,000 occupied receptors/cell). In contrast, [125I]insulin internalization by CHO/IRA1018 and CHO/IR delta 960 cells was slow and was not inhibited by unlabeled insulin. At saturating insulin concentrations, the rate of internalization by wild-type and mutant receptors was similar. Moreover, depletion of intracellular potassium, which has been shown to disrupt coated pit formation, inhibited the rapid internalization of [125I]insulin at physiological insulin concentrations by CHO/IR cells, but had little or no effect on [125I]insulin uptake by CHO/IR delta 960 and CHO/IRA1018 cells or wild-type cells at high insulin concentrations. These data suggest that the insulin-stimulated entry of the insulin receptor into a rapid, coated pit-mediated internalization pathway is saturable and requires receptor autophosphorylation and an intact juxtamembrane region. Furthermore, CHO cells also contain a constitutive nonsaturable pathway which does not require receptor autophosphorylation or an intact juxtamembrane region; this second pathway is unaffected by depletion of intracellular potassium, and therefore may be independent of coated pits. Our data suggest that the ligand-stimulated internalization of the insulin receptor may require specific saturable interactions between the receptor and components of the endocytic system.     

Internalization-sequestration and degradation of cholecystokinin (CCK) in tumoral rat pancreatic AR 4-2 J cells

Cholecystokinin (CCK) receptors were investigated in the tumoral acinar cell line AR 4-2 J derived from rat pancreas, after preincubation with 20 nM dexamethasone. At steady state binding at 37 degrees C (i.e., after a 5 min incubation), less than 10% of the radioactivity of [125I]BH-CCK-9 (3-(4-hydroxy-[125I]iodophenyl)propionyl (Thr34, Nle37) CCK(31-39)) could be washed away from intact cells with an ice-cold acidic medium, suggesting high and rapid internalization-sequestration of tracer. By contrast, more than 85% of the tracer dissociated rapidly after a similar acid wash from cell membranes prelabelled at steady state. In intact AR 4-2 J cells, internalization required neither energy nor the cytoskeleton framework. Tracer internalization was reversed partly but rapidly at 37 degrees C but slowly at 4 degrees C. In addition, two degradation pathways of the tracer were demonstrated, one intracellular and one extracellular. Intracellular degradation occurred at 37 degrees C but not at 20 degrees C and resulted in progressive intracellular accumulation of [125I]BH-Arg that corresponded, after 1 h at 37 degrees C, to 35% of the radioactivity specifically bound. This phenomenon was not inhibited by serine proteinase inhibitors and modestly only by monensin and chloroquine. Besides, tracer degradation at the external cell surface was still observable at 20 degrees C and yielded a peptide (probably [125I]BH-Arg-Asp-Tyr(SO3H)-Thr-Gly). This degradation pathway was partly inhibited by bacitracin and phosphoramidon while thiorphan, an inhibitor of endopeptidase EC 3.4.24.11, was without effect.     

The N-terminal 34 kDa fragment of Helicobacter pylori vacuolating cytotoxin targets mitochondria and induces cytochrome c release

The pathogenic bacterium Helicobacter pylori produces the cytotoxin VacA, which is implicated in the genesis of gastric epithelial lesions. By transfect ing HEp-2 cells with DNAs encoding either the N-terminal (p34) or the C-terminal (p58) fragment of VacA, p34 was found localized specifically to mitochondria, whereas p58 was cytosolic. Incubated in vitro with purified mitochondria, VacA and p34 but not p58 translocated into the mitochondria. Microinjection of DNAs encoding VacA-GFP and p34-GFP, but not GFP-VacA or GFP-p34, induced cell death by apoptosis. Transient transfection of HeLa cells with p34-GFP or VacA-GFP induced the release of cytochrome c from mitochondria and activated the executioner caspase 3, as determined by the cleavage of poly(ADP-ribose) polymerase (PARP). PARP cleavage was antagonized specifically by co-transfection of DNA encoding Bcl-2, known to block mitochondria-dependent apoptotic signals. The relevance of these observations to the in vivo mechanism of VacA action was supported by the fact that purified activated VacA applied externally to cells induced cytochrome c release into the cytosol.     

Leupeptin and ammonium chloride inhibit intrinsic factor mediated transcytosis of [57Co]cobalamin across polarized renal epithelial cells.

The [125I] intrinsic factor (IF) mediated transcytosis of [57Co]Cyanocobalamin (Cbl) by polarized opossum kidney cells was inhibited (greater than 80%) by preincubation of the cells with lysosomotropic agents leupeptin or ammonium chloride. Inhibition of Cbl transcytosis resulted in the intracellular accumulation of both [125I]IF (48 kDa) and [57Co]Cbl. Intracellular degradation of [125I]IF occurred during normal cellular transcytosis of [57Co]Cbl and in one h following internalization the major intracellular degradation products of IF were two polypeptides of Mr 29 kDa and 19 kDa. The size of the major degradation product of IF in the basolateral media was 10 kDa. Based on these results, we suggest that IF is internalized by the renal epithelial cells and is degraded by leupeptin-sensitive acid proteases during Cbl transcytosis.     

Inhibition of coated pit formation in Hep2 cells blocks the cytotoxicity of diphtheria toxin but not that of ricin toxin

It has been recently shown (Larkin, J. M., M. S. Brown, J. L. Goldstein, and R. G. W. Anderson, 1983, Cell, 33:273-285) that after a hypotonic shock followed by incubation in a K+-free medium, human fibroblasts arrest their coated pit formation and therefore arrest receptor-mediated endocytosis of low density lipoprotein. We have used this technique to study the endocytosis of transferrin, diphtheria toxin, and ricin toxin by three cell lines (Vero, Wi38/SV40, and Hep2 cells). Only Hep2 cells totally arrested internalization of [125I]transferrin, a ligand transported by coated pits and coated vesicles, after intracellular K+ depletion. Immunofluorescence studies using anti-clathrin antibodies showed that clathrin associated with the plasma membrane disappeared in Hep2 cells when the level of intracellular K+ was low. In the absence of functional coated pits, diphtheria toxin was unable to intoxicate Hep2 cells but the activity of ricin toxin was unaffected by this treatment. By measuring the rate of internalization of [125I]ricin toxin by Hep2 cells, with and without functional coated pits, we have shown that this labeled ligand was transported in both cases inside the cells. Hep2 cells with active coated pits internalized twice as much [125I]ricin toxin as Hep2 cells without coated pits. Entry of ricin toxin inside the cells was a slow process (8% of the bound toxin per 10 min at 37 degrees C) when compared to transferrin internalization (50% of the bound transferrin per 10 min at 37 degrees C). Using the indirect immunofluorescence technique on permeabilized cells, we have shown that Hep2 cells depleted in intracellular K+ accumulated ricin toxin in compartments that were predominantly localized around the cell nucleus. Our study indicates that in addition to the pathway of coated pits and coated vesicles used by diphtheria toxin and transferrin, another system of endocytosis for receptor-bound molecules takes place at the level of the cell membrane and is used by ricin toxin to enter the cytosol.     

Internalization of ricin in Chinese hamster ovary cells.

Internalization of ricin into Chinese hamster ovary cells has been investigated. Combined treatment with galactose and pronase at 0 degrees C resulted in a complete release of surface-bound [125I]ricin into the media. Galactose-pronase-resistant cell-bound [125I]ricin represents internalized ricin molecules inside the cells. The internalization process is time, temperature, and concentration dependent. The pH optimum of internalization of ricin is about pH 7. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis has revealed that intact ricin molecules are internalized. Neither reduction nor proteolytic processing of ricin is required for the entry of ricin into Chinese hamster ovary cells.     

Internalization and degradation of cholera toxin by cultured cells: relationship to toxin action

Using anticholeragen antibodies and 125I-protein A, we developed a specific and quantitative assay for measuring choleragen on the surfaces of cultured cells. When neuroblastoma cells containing bound toxin were incubated at 37 degrees C, surface toxin disappeared with a half-life of approximately 2 h and a significant loss was detected by 10 min. When cells were incubated with 125I-choleragen in order to measure toxin degradation, cell-associated radioactivity disappeared with time and a corresponding amount of TCA-soluble label appeared in the culture medium with a half-life of 4-6 h. No degradation was detected until 45 min. Although there was a lag of 15 min before bound choleragen activated adenylate cyclase, the enzyme became maximally activated between 45 and 60 min. Similar results were obtained with Friend erythroleukemia cells. Internalization, degradation, and activation all were blocked when the cells were maintained at 4 degrees C. At 22 degrees C, internalization and activation occurred, albeit at a slower rate, whereas degradation was effectively inhibited. These results indicated that choleragen does not have to be degraded by intact cells in order for it to activate adenylate cyclase. Some internalization of the toxin, however, appears to precede the activation process.     

Internalization and recycling of CD4 transfected into HeLa and NIH3T3 cells.

The internalization of CD4, a T cell differentiation antigen and the receptor for the human immunodeficiency viruses (HIV-1 and -2), has been examined in HeLa and murine 3T3 cells transfected with CD4 cDNA. Fab' fragments of the anti-CD4 monoclonal antibody Leu3a were generated by pepsin digestion and used as a specific monovalent, non-crosslinking ligand for CD4. These Fab' fragments were shown to bind to CD4 on the transfected cells with an affinity similar to that of HIV gp120, and inhibited HIV infection of lymphocytic cells. The Fab' fragments were radioiodinated and used in an acid-stripping endocytosis assay to demonstrate that the CD4 expressed on transfected HeLa and NIH3T3 cells was internalized. Approximately 1.5-2% of the total cell-bound [125I]Fab' fragments were internalized per minute. Furthermore, the internalized [125I]Fab' fragments could be shown to recycle to the cell surface. After 30-60 min a steady state was reached between internalization and recycling, with approximately 30-40% of the total cellular CD4 pool residing inside the cell. Similar results were obtained in studies with the intact divalent radiolabelled Leu3a antibody. These data demonstrate that CD4 expressed on transfected non-lymphoid cells is constitutively endocytosed and recycled.     

Internalization and degradation of receptor bound C-reactive protein by U-937 cells: induction of H2O2 production and tumoricidal activity

The presence of a membrane receptor for C-reactive protein (CRP-R) on the human monocytic cell line U-937 was the basis for determining the metabolic fate of the receptor-bound ligand and the functional response of the cells to CRP. Internalized [125I]CRP was measured by removing cell surface-bound [125I]CRP with pronase. Warming cells to 37 degrees C resulted in the internalization of approx. 50% of the receptor-bound [125I]CRP or receptor-bound [125I]CRP-PC-KLH complexes. U-937 cells degraded about 25% of the internalized [125I]CRP into TCA-soluble radiolabeled products. The lysosomotrophic agents (chloroquine, NH4Cl) greatly decreased the extent of CRP degradation without altering binding or internalization. In addition, a pH less than 4.0 resulted in dissociation of receptor-bound [125I]CRP. Treatment of U-937 cell with monensin, a carboxylic ionophore which prevents receptor recycling, resulted in accumulation of internalized [125I]CRP. Therefore, it appears that the CRP-R complex is internalized into an endosomal compartment where the CRP is uncoupled from its receptor and subsequently degraded. CRP initiated the differentiation of the U-937 cells so that they acquired the ability to produce H2O2 and also display in vitro tumoricidal activity. The results support the concept that internalization and degradation of CRP leads to the activation of monocytes during inflammation.     

A dominant negative mutant of Helicobacter pylori vacuolating toxin (VacA) inhibits VacA-induced cell vacuolation

Most Helicobacter pylori strains secrete a toxin (VacA) that causes structural and functional alterations in epithelial cells and is thought to play an important role in the pathogenesis of H. pylori-associated gastroduodenal diseases. The amino acid sequence, ultrastructural morphology, and cellular effects of VacA are unrelated to those of any other known bacterial protein toxin, and the VacA mechanism of action remains poorly understood. To analyze the functional role of a unique strongly hydrophobic region near the VacA amino terminus, we constructed an H. pylori strain that produced a mutant VacA protein (VacA-(Delta6-27)) in which this hydrophobic segment was deleted. VacA-(Delta6-27) was secreted by H. pylori, oligomerized properly, and formed two-dimensional lipid-bound crystals with structural features that were indistinguishable from those of wild-type VacA. However, VacA-(Delta6-27) formed ion-conductive channels in planar lipid bilayers significantly more slowly than did wild-type VacA, and the mutant channels were less anion-selective. Mixtures of wild-type VacA and VacA-(Delta6-27) formed membrane channels with properties intermediate between those formed by either isolated species. VacA-(Delta6-27) did not exhibit any detectable defects in binding or uptake by HeLa cells, but this mutant toxin failed to induce cell vacuolation. Moreover, when an equimolar mixture of purified VacA-(Delta6-27) and purified wild-type VacA were added simultaneously to HeLa cells, the mutant toxin exhibited a dominant negative effect, completely inhibiting the vacuolating activity of wild-type VacA. A dominant negative effect also was observed when HeLa cells were co-transfected with plasmids encoding wild-type and mutant toxins. We propose a model in which the dominant negative effects of VacA-(Delta6-27) result from protein-protein interactions between the mutant and wild-type VacA proteins, thereby resulting in the formation of mixed oligomers with defective functional activity.     

Involvement of the Golgi region in the intracellular trafficking of cholera toxin

The intracellular pathway following receptor-mediated endocytosis of cholera toxin was studied using brefeldin A (BFA), which inhibited protein secretion and induced dramatic morphological changes in the Golgi region. In both mouse Y1 adrenal cells and CHO cells, BFA at 1 μg/ml caused a 80–90% inhibition of the cholera toxin (CT)-elevation of intracellular cAMP. The inhibition of the cytotoxicity of CT by BFA was also observed in a rounding assay of Y1 adrenal cells. The inhibition of CT cytotoxicity by BFA was dose dependent, with the ID₅₀ value similar to the LD₅₀ of BFA in Y1 adrenal cells. Binding and internalization of [¹²⁵I]-cholera toxin in Y1 adrenal cells was not affected by BFA. Unlike the BFA-sensitive cell lines such as Y1 adrenal and CHO cells, BFA at 1 μg/ml did not inhibit the cytotoxicity of CT in PtK₁ cells, of which the Golgi structure was BFA-resistant. These results strongly suggest that a BFA-sensitive Golgi is required for the protection of CT cytotoxicity by BFA. In contrast, elevation of the intracellular cAMP by forskolin, which acts directly on the plasma membrane adenylate cyclase, was not affected by BFA. These observations indicate that the intoxication of target cells by CT requires an intact Golgi region for its intracellular trafficking and/or processing. In this respect, CT shares a common intracellular pathway with ricin, Pseudomonas toxin, and modeccin, even though their structures and modes of action are very different. © 1993 Wiley-Liss, Inc.     

Human gastric glycosphingolipids recognized by Helicobacter pylori vacuolating cytotoxin VacA

Many bacterial toxins utilize cell surface glycoconjugate receptors for attachment to target cells. In the present study the potential carbohydrate binding of Helicobacter pylori vacuolating cytotoxin VacA was investigated by binding to human gastric glycosphingolipids on thin-layer chromatograms. Thereby a distinct binding of the toxin to two compounds in the non-acid glycosphingolipid fraction was detected. The VacA-binding glycosphingolipids were isolated and characterized by mass spectrometry and proton NMR as galactosylceramide (Galbeta1Cer) and galabiosylceramide (Galalpha4Galbeta1Cer). Comparison of the binding preferences of the protein to reference glycosphingolipids from other sources showed an additional recognition of glucosylceramide (Glcbeta1Cer), lactosylceramide (Galbeta4Glcbeta1Cer) and globotriaosylceramide (Galalpha4Galbeta4Glcbeta1Cer). No binding to the glycosphingolipids recognized by the VacA holotoxin was obtained with a mutant toxin with deletion of the 37 kDa fragment of VacA (P58 molecule). Collectively our data show that the VacA cytotoxin is a glycosphingolipid binding protein, where the 37 kDa moiety is required for carbohydrate recognition. The ability to bind to short chain glycosphingolipids will position the toxin close to the cell membrane, which may facilitate toxin internalization.     

Modulation of mannose receptor activity by proteolysis.

The binding, internalization and degradation of 200 pM monoiodinated human atrial natriuretic factor-(99-126) (125I-hANF) by cultured bovine aortic endothelial cells (BAECs) were studied at 37 degrees C. 125I-hANF was rapidly cleared from the extracellular medium (t1/2 approximately 10 min), whereas preincubation of the cells in the presence of 20 mM-NH4Cl or 0.2 mM-chloroquine resulted in a significant inhibition of this process. The BAECs rapidly produce three major degradation products of 125I-hANF, namely [125I]iodotyrosine 126 (125I-Y), Arg125-[125I]iodotyrosine126 (125I-RY) and Phe124-Arg125-[125I]iodotyrosine126(125I-FRY), which were detected in the extracellular medium. NH4Cl and chloroquine acted to inhibit the generation of 125I-Y and 125I-RY, but not that of 125I-FRY. Furthermore, excess unlabelled hANF (300 nM) completely blocked the rapid production of 125I-Y and 125I-RY in the first 5 min, but only partially (49%) inhibited the generation of 125I-FRY. Thus, in contrast with our previous findings with cultured smooth-muscle cells [Johnson, Arik & Foster (1989) J. Biol. Chem. 264, 11637-11642], BAECs bind, internalize and rapidly degrade 125I-hANF, resulting in the release of 125I-Y and 125I-RY into the extracellular medium. Similarly to smooth-muscle cells, the BAECs generate 125I-FRY from 125I-hANF via an extracellular proteolytic event. The rapidity of the receptor-mediated process and its sensitivity to NH4Cl and chloroquine suggest that the 125I-hANF is proteolytically processed in the endosomes of BAECs and that its receptors cycle between the cell surface and intracellular stores.     

Internalization and degradation of peptides of the bombesin family in Swiss 3T3 cells occurs without ligand-induced receptor down-regulation.

The binding of [125I]gastrin releasing peptide ([125I]GRP) to Swiss 3T3 cells at 37 degrees C increases rapidly, reaching a maximum after 30 min and decreasing afterwards. The decrease in cell-associated radioactivity at this temperature is accompanied by extensive degradation of the labelled peptide. At 4 degrees C equilibrium binding is achieved after 6 h and [125I]GRP degradation is markedly inhibited. Extraction of surface-bound ligand at low pH demonstrates that the iodinated peptide is internalized within minutes after addition to 3T3 cells at 37 degrees C. The rate of internalization is strikingly temperature-dependent and is virtually abolished at 4 degrees C. In addition, lysomotropic agents including chloroquine increase the cell-associated radioactivity in cells incubated with [125I]GRP. The binding of [125I]GRP to Swiss 3T3 cells was not affected by pretreatment for up to 24 h with either GRP or bombesin at mitogenic concentrations. Furthermore, pretreatment with GRP did not reduce the affinity labelling of a Mr 75,000-85,000 surface protein recently identified as a putative receptor for bombesin-like peptides. These results demonstrate that while peptides of the bombesin family are rapidly internalized and degraded by Swiss 3T3 cells, the cell surface receptors for these molecules are not down-regulated.     

Essential role of a GXXXG motif for membrane channel formation by Helicobacter pylori vacuolating toxin

Helicobacter pylori secretes a toxin, VacA, that can form anion-selective membrane channels. Within a unique amino-terminal hydrophobic region of VacA, there are three tandem GXXXG motifs (defined by glycines at positions 14, 18, 22, and 26), which are characteristic of transmembrane dimerization sequences. The goals of the current study were to investigate whether these GXXXG motifs are required for membrane channel formation and cytotoxicity and to clarify the role of membrane channel formation in the biological activity of VacA. Six different alanine substitution mutations (P9A, G13A, G14A, G18A, G22A, and G26A) were introduced into the unique hydrophobic region located near the amino terminus of VacA. The effects of these mutations were first analyzed using the TOXCAT system, which permits the study of transmembrane oligomerization of proteins in a natural membrane environment. None of the mutations altered the capacity of ToxR-VacA-maltose-binding protein fusion proteins to insert into a membrane, but G14A and G18A mutations markedly diminished the capacity of the fusion proteins to oligomerize. We then introduced the six alanine substitution mutations into the vacA chromosomal gene of H. pylori and analyzed the properties of purified mutant VacA proteins. VacA-G13A, VacA-G22A, and VacA-G26A induced vacuolation of HeLa cells, whereas VacA-P9A, VacA-G14A, and VacA-G18A did not. Subsequent experiments examined the capacity of each mutant toxin to form membrane channels. In a planar lipid bilayer assay, VacA proteins containing G13A, G22A, and G26A mutations formed anion-selective membrane channels, whereas VacA proteins containing P9A, G14A, and G18A mutations did not. Similarly, VacA-G13A, VacA-G22A, and VacA-G26A induced depolarization of HeLa cells, whereas VacA-P9A, VacA-G14A, and VacA-G18A did not. These data indicate that an intact proline residue and an intact G(14)XXXG(18) motif within the amino-terminal hydrophobic region of VacA are essential for membrane channel formation, and they also provide strong evidence that membrane channel formation is essential for VacA cytotoxicity.     

Helicobacter pylori vacuolating toxin forms anion-selective channels in planar lipid bilayers: possible implications for the mechanism of cellular vacuolation

The Helicobacter pylori VacA toxin plays a major role in the gastric pathologies associated with this bacterium. When added to cultured cells, VacA induces vacuolation, an effect potentiated by preexposure of the toxin to low pH. Its mechanism of action is unknown. We report here that VacA forms anion-selective, voltage-dependent pores in artificial membranes. Channel formation was greatly potentiated by acidic conditions or by pretreatment of VacA at low pH. No requirement for particular lipid(s) was identified. Selectivity studies showed that anion selectivity was maintained over the pH range 4.8-12, with the following permeability sequence: Cl- approximately HCO3- > pyruvate > gluconate > K+ approximately Li+ approximately Ba2+ > NH4+. Membrane permeabilization was due to the incorporation of channels with a voltage-dependent conductance in the 10-30 pS range (2 M KCl), displaying a voltage-independent high open probability. Deletion of the NH2 terminus domain (p37) or chemical modification of VacA by diethylpyrocarbonate inhibited both channel activity and vacuolation of HeLa cells without affecting toxin internalization by the cells. Collectively, these observations strongly suggest that VacA channel formation is needed to induce cellular vacuolation, possibly by inducing an osmotic imbalance of intracellular acidic compartments.     

Binding of epidermal growth factor by human colon carcinoma cell (Caco-2) monolayers

The objective of this study was to investigate whether Caco-2 cells bind and internalize epidermal growth factor (EGF). [125I]EGF was presented to the apical (AP) or basolateral (BL) side of Caco-2 monolayers, grown on microporous membranes, at different times in culture. At day 10, [125I]EGF binding (at 37 degrees C) to the BL membrane was 2-3 times greater than binding to the AP membrane. Of that [125I]EGF bound to the AP membrane 76% was internalized within 3 h while internalization from the BL membrane was 90%. At lower temperatures membrane-bound [125I]EGF increased while internalization decreased. At day 16, AP and BL binding decreased and then remained constant through day 25. [125I]EGF was bound to the BL membrane of 10 days old monolayers with a Kd of 0.67 nM. There was a single binding site whose numbers in the BL membrane was about 5500/cell.