The Intermediate Region of Helicobacter pylori VacA Is a Determinantof Toxin Potency in a Jurkat T Cell Assay

Christian González-Rivera,a Holly M. Scott Algood,b Jana N. Radin,b Mark S. McClain,b and Timothy L. Covera,b,c

Department of Pathology, Microbiology and Immunologya and Department of Medicine,b Vanderbilt University School of Medicine, Nashville, Tennessee, USA, andVeterans Affairs Tennessee Valley Healthcare System, Nashville, Tennessee, USAc

Colonization of the human stomach with Helicobacter pylori is a risk factor for peptic ulceration, noncardia gastric adenocarci-noma, and gastric lymphoma. The secreted VacA toxin is an important H. pylori virulence factor that causes multiple alterationsin gastric epithelial cells and T cells. Several families of vacA alleles have been described, and H. pylori strains containing certainvacA types (s1, i1, and m1) are associated with an increased risk of gastric disease, compared to strains containing other vacAtypes (s2, i2, and m2). Thus far, there has been relatively little study of the role of the VacA intermediate region (i-region) intoxin activity. In this study, we compared the ability of i1 and i2 forms of VacA to cause functional alterations in Jurkat cells. Todo this, we manipulated the chromosomal vacA gene in two H. pylori strains to introduce alterations in the region encoding theVacA i-region. We did not detect any differences in the capacity of i1 and i2 forms of VacA to cause vacuolation of RK13 cells. Incomparison to i1 forms of VacA, i2 forms of VacA had a diminished capacity to inhibit the activation of nuclear factor of acti-vated T cells (NFAT) and suppress interleukin-2 (IL-2) production. Correspondingly, i2 forms of VacA bound to Jurkat cells lessavidly than did i1 forms of VacA. These results indicate that the VacA i-region is an important determinant of VacA effects onhuman T cell function.

Helicobacter pylori is a Gram-negative microaerophilic bacte-rium that persistently colonizes the human stomach (3, 10).

H. pylori infection elicits a gastric mucosal inflammatory responseand is associated with an increased risk of peptic ulcer disease,gastric adenocarcinoma, and gastric lymphoma (2, 52, 63). One ofthe important virulence factors produced by H. pylori is a secretedpore-forming toxin known as VacA (9, 17, 23, 38).

The vacA gene encodes a 140-kDa protein, which undergoesproteolytic processing to yield an amino-terminal signal se-quence, an 88-kDa secreted toxin, and a carboxyl-terminal �-bar-rel domain (15, 22, 55, 66). The 88-kDa toxin (passenger domain)is secreted by a type V or autotransporter mechanism (15, 16, 22,55). Two domains of the 88-kDa secreted toxin have been identi-fied and are designated p33 and p55 (50, 66, 68, 75). Amino acidsequences within both the p55 domain (51, 72) and the p33 do-main (31, 68) contribute to the cell-binding capacity of VacA. Thecrystal structure of the p55 domain has been determined and pre-dominantly consists of a right-handed parallel �-helix (27).

The secreted 88-kDa toxin can assemble into large water-solu-ble flower-shaped oligomeric complexes (12, 21, 42). Upon expo-sure to acid or alkaline pH, the oligomers dissociate into mono-meric 88-kDa components (12, 48). In comparison to intact VacAoligomers, which have relatively little effect on human cells,oligomers exposed to acid or alkaline pH conditions are highlyactive on human cells (18, 47). A current model proposes thatVacA monomers interact with the plasma membrane and subse-quently oligomerize, which allows the formation of VacA pores incell membranes (9). VacA causes a wide range of alterations inhuman gastric cells (9), including the formation of large cytoplas-mic vacuoles (11, 40), permeabilization of the plasma membrane(65), reduction of mitochondrial transmembrane potential (19,24, 26, 74), mitochondrial cytochrome c release (19, 24, 26, 74),mitochondrial fragmentation (35), activation of mitogen-acti-vated protein kinases (49), induction of autophagy (67), and celldeath (13, 26, 35, 53). Most of these effects (but not all) are

dependent on membrane channel formation by VacA (30, 34,46, 71).

VacA also has effects on cells of the immune system and hasbeen classified as an immunomodulatory toxin (7, 29, 64). VacAinteracts with �2 integrin on the surface of human T cells (57) andis then internalized through a clathrin-independent pathway (58).Once inside T cells, VacA inhibits the activation and nuclear trans-location of nuclear factor of activated T cells (NFAT) (29, 58). Asa consequence, VacA inhibits the expression and secretion of in-terleukin-2 (IL-2) (29). Effects of VacA on IL-2 production havebeen studied most extensively in Jurkat cells (29, 57). In additionto its effects on IL-2 production by Jurkat cells, VacA inhibits theactivation-induced proliferation of primary human T cells and Bcells (57, 58, 64, 69).

The vacA alleles of H. pylori strains from unrelated humansexhibit a high level of genetic diversity, and several vacA types havebeen recognized based on sequence diversity in specific regions(Fig. 1) (4, 5, 28). Until recently, most studies focused on diversityat the 5= end (s-region) or within the middle region (m-region) ofvacA (4). Two main families of s-region and m-region sequenceshave been recognized (designated types s1 and s2, m1 and m2) (4,5, 28). H. pylori strains containing type s1 or m1 vacA alleles areassociated with a higher incidence of gastric disease than arestrains containing type s2 or m2 vacA alleles (4, 70). Type s1 VacAproteins cause numerous cellular alterations in vitro, whereas type

Received 16 January 2012 Returned for modification 16 February 2012Accepted 4 May 2012

Published ahead of print 14 May 2012

Editor: B. A. McCormick

Address correspondence to Timothy L. Cover, timothy.l.cover@vanderbilt.edu.

Copyright © 2012, American Society for Microbiology. All Rights Reserved.

doi:10.1128/IAI.00052-12

2578 iai.asm.org Infection and Immunity p. 2578–2588 August 2012 Volume 80 Number 8

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

s2 VacA proteins lack detectable activity in most in vitro assays (4,25, 39, 43). In comparison to type m2 VacA proteins, type m1VacA proteins cause vacuolation in a wider range of cells; this hasbeen attributed to differences in the cell-binding properties of m1and m2 VacA proteins (36, 51, 61, 73). Differences in the activitiesof type s1 and m1 forms of VacA compared to s2 and m2 forms ofVacA have been detected not only in studies of epithelial cells butalso in studies of T cells (1, 29, 57).

A third polymorphic region, known as the intermediate region(i-region), was recently identified within the p33 domain of vacA(54). Similar to the s- and m-regions, two families of i-regionsequences have been recognized, and these are designated type i1and i2 (54). One study reported that type i1 VacA proteins causedvacuolation of HeLa and RK13 cells (derived from human cervixand rabbit kidney, respectively), whereas type i2 VacA proteinscaused vacuolation of RK13 cells but not HeLa cells (54). There-fore, it was concluded that the i-region is a determinant of VacAcell-type specificity. Within the i-region, there are three main clus-ters of amino acid diversity, known as polymorphic clusters A, B,and C (Fig. 1). Sequence variation within clusters B and C wasreported to be responsible for the observed variation in cell-typespecificity (54). Importantly, H. pylori strains containing i1 vacAalleles have been associated with a higher incidence of gastric dis-ease, in comparison to H. pylori strains containing i2 vacA alleles(6, 8, 20, 33, 37, 54, 60).

Although multiple studies reported that the vacA i-region type

is a marker of disease outcome, thus far there have been very fewstudies comparing the activities of type i1 and type i2 VacA pro-teins. In the current study, we tested the hypothesis that type i1and i2 VacA proteins differ in the capacity to cause functionalalterations in human T cells. We report that, in comparison totype i1 VacA proteins, type i2 VacA proteins have a reduced ca-pacity to inhibit the function of Jurkat cells. In addition, we dem-onstrate that the difference in activities of i1 and i2 VacA proteinsis attributable, at least in part, to differences in the binding of theseVacA proteins to T cells.

MATERIALS AND METHODSBacterial strains and culture conditions. Bacterial strains and the plas-mids used in this study are listed in Table 1. The wild-type (WT) H. pylori60190 strain (ATCC 49503) and strain X47 (generously provided byDouglas Berg) were grown on Trypticase soy agar plates containing 5%sheep blood at 37°C in ambient air containing 5% CO2. H. pylori mutantstrains were grown on brucella agar plates containing 10% fetal bovineserum, supplemented with metronidazole (3.75 �g/ml) or chloramphen-icol (5 �g/ml) when indicated. H. pylori liquid cultures were grown inbrucella broth supplemented with either activated charcoal or 5% fetalbovine serum (FBS) (11).

Preparation of H. pylori broth culture supernatants and normaliza-tion of VacA concentrations. For experiments using H. pylori broth cul-ture supernatant (derived from bacteria cultured in brucella broth con-taining FBS), supernatants were concentrated 50-fold by ultrafiltrationwith a 30-kDa-cutoff membrane (Millipore). The relative concentrations

FIG 1 Amino acid sequences of the VacA i-region in different H. pylori strains. The secreted p88 VacA protein (comprising 821 amino acids in H. pylori strain60190) contains two domains, designated p33 and p55. Amino acid sequence variation in VacA proteins from different H. pylori strains is especially prominentin three regions, which are designated s-, i-, and m-regions. The i-region and a portion of the s-region are localized within the p33 domain, and the m-region islocalized within the p55 domain. The lower portion of this figure shows the sequences of VacA i-regions from strain 60190 (GenBank no. U05676), Tx30a(GenBank no. U29401), and X47. The figure illustrates the sequence of the secreted VacA protein of H. pylori strain 60190 (amino acids G120 to L236) andcorresponding regions of VacA in two other wild-type strains (Tx30a and X47). Clusters of amino acid polymorphisms defined as clusters A, B, and C areindicated. Asterisks (*) indicate amino acids that are identical in all three sequences. The sequence of VacA from strain 60190 is designated a prototype for thei1 region, and the sequence of VacA from strain Tx30a is designated a prototype for the i2 region. VacA from strain X47 contains a type i2 sequence in cluster C,and cluster B is chimeric.

H. pylori VacA Intermediate Region

August 2012 Volume 80 Number 8 iai.asm.org 2579

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

of VacA in broth culture supernatant preparations from WT and mutantH. pylori strains were determined by Western blot analysis using anti-VacA antiserum no. 958 (prepared by immunization of a rabbit with VacAoligomers purified from H. pylori broth culture supernatant) (56). Thisanti-VacA antiserum reacted equally well with i1 and i2 VacA proteins inan antigen detection enzyme-linked immunosorbent assay (ELISA) (datanot shown). When necessary, the concentrations of VacA in individualpreparations were normalized by diluting samples with the appropriatevolumes of concentrated brucella broth containing FBS.

Purification of VacA from H. pylori broth culture supernatants. Forexperiments using purified VacA, VacA oligomers were purified from H.pylori culture supernatants as described previously (12). Prior to addingpurified VacA to eukaryotic cells, the oligomeric VacA preparations wereacid activated by the slow addition of 200 mM HCl until a pH of 3.0 wasreached.

Mutagenesis of vacA. To generate unmarked H. pylori mutant strains,we used a negative selection method (41, 59). As a first step, metronida-zole-resistant forms of strains 60190 and X47, designated 60190 �rdxAand X47 �rdxA, were generated by deletion of the rdxA gene. PCR analysisconfirmed that the rdxA locus was deleted from the mutant strains. As anext step, cloned vacA sequences were disrupted by insertion of a cat-rdxAcassette. This cassette confers resistance to chloramphenicol mediated bythe chloramphenicol acetyltransferase (cat) gene from Campylobacter coli,and susceptibility to metronidazole is mediated by an intact rdxA gene(HP0954) from H. pylori 26695 (41). For mutagenesis of vacA in H. pyloristrain 60190, the cat-rdxA cassette (described above) was ligated into anStuI site in plasmid pA178, which contains a vacA DNA fragment from H.pylori 60190 (43). The resulting plasmid (pCGR1), which is unable toreplicate in H. pylori, was used to transform the H. pylori 60190 �rdxAstrain, and single colonies resistant to chloramphenicol (5 �g/ml) butsensitive to metronidazole (3.75 �g/ml) were selected. For mutagenesis ofvacA in H. pylori strain X47, a DNA fragment encoding VacA amino acids

4 to 727 was PCR amplified from this strain, and the PCR product wascloned into pGEMT-Easy (Promega). The resulting plasmid was digestedwith EcoRV, and the cat-rdxA cassette was ligated into this restriction site.The resulting plasmid (pCGR2) was transformed into the H. pylori X47�rdxA strain, and single colonies resistant to chloramphenicol but sensi-tive to metronidazole were selected. Immunoblot analysis revealed theloss of VacA expression in these mutants, and insertion of the cat-rdxAcassette into the vacA gene was confirmed by PCR amplification and nu-cleotide sequence analysis of PCR products. To introduce alterations intothe i-region of the chromosomal vacA gene in H. pylori strains, we con-structed various plasmids using an inverse PCR approach with the 5=phosphorylated primers listed in Table 2. The vacA sequence from eachplasmid was sequenced to ensure that unintentional mutations were notintroduced. These plasmids were used to transform H. pylori strains con-taining the cat-rdxA cassette, and transformants resistant to metronida-zole were selected. The presence of the desired mutations was confirmedby PCR and nucleotide sequence analysis of PCR products.

Cell culture. RK13 cells were obtained from the American Type Cul-ture Collection (ATCC CCL-37) and were cultured in minimal essential

TABLE 1 H. pylori strains and plasmids

Strain/plasmid Relevant characteristics Reference

Strains60190 Wild type (ATCC 49503); vacA s1/i1/m1 4060190 �rdxA Same as 60190 except HP0954 (rdxA) gene deleted; metronidazole resistant This study60190 vacA::cat rdxA Same as 60190 �rdxA except cat cassette and rdxA inserted in vacA; chloramphenicol resistant

and metronidazole sensitive; expression of VacA is disruptedThis study

60190 i2B Same as 60190 �rdxA except vacA cluster B changed to i2 This study60190 i2C Same as 60190 �rdxA except vacA cluster C changed to i2 This study60190 i2BC Same as 60190 �rdxA except vacA clusters B and C changed to i2 This study60190 i1/i2C Same as 60190 �rdxA except vacA cluster C has 4 amino acids changed to i2 This studyX47 Wild type; vacA s1/m2, chimeric i-region 32X47 �rdxA Same as X47 except HP0954 (rdxA) gene deleted; metronidazole resistant This studyX47 vacA::cat rdxA Same as X47 �rdxA except cat cassette and rdxA inserted in vacA; chloramphenicol resistant

and metronidazole sensitive; expression of VacA is disruptedThis study

X47 i1C Same as X47 �rdxA except vacA cluster C changed to i1 This study

PlasmidspMM672 Allows deletion of rdxA in H. pylori strains 41pCGR1 Contains cat-rdxA cassette in StuI site; derived from pA178 plasmid This studypCGR2 Contains cat-rdxA cassette in EcoRV site from X47 vacA This studypCGR3 60910 cluster B changed from i1 to i2 by inverse PCR using primers B1F and B1R This studypCGR4 60190 cluster C changed from i1 to i2 by inverse PCR using primers C1F and C1R This studypCGR5 60190 clusters B and C changed from i1 to i2 by inverse PCR using primers C1F and C1R and

pCGR3 as the templateThis study

pCGR6 A portion of 60190 cluster C changed from i1 to i2 by inverse PCR using primers C2F and C2R This studypCGR7 X47 cluster C changed from i2 to i1 by inverse PCR using primers C3F and C3R This studyp55 Expresses VacA p55 27p33 Expresses VacA p33 31p33 i2 Expresses p33 i2 This study

TABLE 2 PCR primers used for mutagenesis of the vacA i-region

Primer Sequence (5=¡3=)B1F ATTACAAGCCGTGAAAATGCTGAAATTTCTCTTTATGB1R TTTTTCTGAACTTTTCAAAGTCAAAACCGTAGAGCC1F TATATGGTAAGGTGTGGATGGGCCGTTTGCC1R GATCAACGCTCTGATTTGAGCTTGAAACCAAATTGAGCGT

AGCGCCATCC2F AACCAAAGCGTTAAATTAAATGGCAATGTGC2R GCTGTTTGACACCAAATTGAGCGTAGCGCCAC3F TTAAATGGCAATGTGTGGATGGGCCGTTTGCAATAC3R TTTAACGCTGTTTGAAGCCAAATTGAGCGTGGCGCCATCATA

González-Rivera et al.

2580 iai.asm.org Infection and Immunity

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

medium supplemented with 10% FBS and 1 mM nonessential aminoacids. Jurkat T lymphocytes (clone E6-1, ATCC TIB-152) were cultured inRPMI 1640 medium containing 2 mM L-glutamine, 1.5 g/liter sodiumbicarbonate, 4.5 g/liter glucose, 10 mM HEPES, 1.0 mM sodium pyruvate,and 10% FBS. Jurkat lymphocytes containing stable luciferase reporterswere cultured as described above, except that the medium was supple-mented with 1 �M puromycin.

Neutral red uptake assay. To quantify VacA-induced cell vacuolation,RK13 cells were seeded at a density of 2 � 104 cells/well into 96-well platesfor 24 h prior to the experiment. Serial dilutions of concentrated H. pyloriculture supernatants containing different forms of VacA were added toserum-free tissue culture medium (supplemented with 10 mM ammo-nium chloride) overlying cells and incubated overnight at 37°C. VacA-induced cell vacuolation was detected by inverted light microscopy andquantified by a neutral red uptake assay, a well-established method that isbased on rapid uptake of neutral red into VacA-induced cell vacuoles (11,14). Background levels of neutral red uptake by untreated cells were sub-tracted to yield net neutral red uptake values.

Analysis of IL-2 production by Jurkat cells. Jurkat T cells were platedin 96-well plates at a density of 1 �105 cells/well, and H. pylori brothculture supernatant preparations or purified VacA proteins were added tocells for 30 min at 37°C. The cells were then stimulated with phorbolmyristate acetate (PMA) (50 ng/ml; Sigma) and ionomycin (500 ng/ml;Sigma) and maintained in RPMI 1640 medium containing 10% FBS for24 h. Cells were pelleted, and levels of IL-2 in the supernatants were quan-tified by ELISA, according to the manufacturer’s protocol (R&D Systems;human IL-2 immunoassay). To ensure that IL-2 production was not al-tered by T-cell apoptosis, we monitored the viability of Jurkat cells in eachexperiment by using trypan blue staining and did not detect any signifi-cant effect of VacA on viability of the cells (data not shown), a result thatis consistent with previous publications (29, 64).

Expression and purification of recombinant VacA proteins. Recom-binant p33 and p55 proteins, derived from H. pylori strain 60190, wereexpressed in Escherichia coli and purified as described previously (27, 31).In addition, we modified the plasmid encoding the i1 p33 protein derivedfrom H. pylori strain 60190, so that it expressed an i2 form of p33. To dothis, we first changed the sequence in the vacA i-region polymorphic clus-ter B from type 1 to 2 by inverse PCR, using the WT p33 plasmid as atemplate and primers B1F and B1R (Table 2). The resulting plasmid,containing a type 2 cluster B and type 1 cluster C, was then used as atemplate to change the amino acid sequence of cluster C to type 2, usingprimers C1F and C1R (Table 2). The modifications in the i-region wereconfirmed by nucleotide sequence analysis. VacA p33 and p55 were ex-pressed by culturing Escherichia coli BL21(DE3) in Terrific broth (Fisher)supplemented with 25 �g/ml kanamycin (TB-KAN) at 37°C overnightwith shaking. Cultures were diluted 1:100 in TB-KAN and grown at 37°Cuntil they reached an absorbance (A600) of 0.6. Cultures were inducedwith a final isopropyl �-D-thiogalactopyranoside (IPTG) concentration of0.5 mM and incubated at 25°C for 16 to 18 h (p55 proteins) or at 37°C for3 h (p33 proteins). VacA p55 was purified under native conditions bynickel affinity, ion exchange, and gel filtration chromatography (27).VacA p33 proteins were purified under denaturing conditions from in-clusion bodies by using Ni-affinity resin (Novagen). The purified dena-tured VacA p33 proteins were then refolded by dialysis and were purifiedfurther by gel filtration chromatography (31).

Flow cytometric analysis of VacA binding to cells. Purified p55 waslabeled with Alexa 488 (Molecular Probes) according to the manufactur-er’s instructions (31). Jurkat cells (1 � 105 cells per condition) weretreated with Alexa 488-labeled p55 alone (10 �g/ml) or with a mixture ofAlexa 488-labeled p55 plus either purified refolded p33 i1 or p33 i2 pro-teins (each at 5 �g/ml) at 4°C for 1 h. Cells were then washed three timesin cold phosphate-buffered saline (PBS) containing 0.5% bovine serumalbumin (BSA) and fixed in 2% paraformaldehyde. The cells were col-lected using a flow cytometer (LSR II system; BD, San Alta, CA) andanalyzed using BD Diva (1). Immunofluorescent microscopy experi-

ments indicated that the VacA proteins were not internalized at 4°C (datanot shown).

Immunoblot analysis of VacA binding to cells. Jurkat cells (1 � 106

cells per condition) were cultured in serum-free medium for 8 h and thenincubated with preparations of H. pylori broth culture supernatants at 4°Cfor 1 h. Cells were washed three times with cold PBS, pelleted, and heatedat 100°C for 5 min in sodium dodecyl sulfate loading buffer. Samples wereelectrophoresed on a 4 to 20% gradient precast acrylamide gel (Bio-Rad)and transferred onto nitrocellulose membranes. Membranes were immu-noblotted with rabbit anti-VacA serum (serum number 958, diluted1:10,000) or anti-GAPDH serum (abcam, diluted 1:1,000), followed byhorseradish peroxidase-conjugated secondary antibodies (Promega, di-luted 1:10,000). Immune complexes were revealed by using an enhancedchemiluminescence system (ECL Western Blotting Analysis System; GEHealthcare).

Generation of a Jurkat cell line with a stable NFAT luciferase re-porter. Jurkat lymphocytes were transduced with replication-deficientlentiviral particles encoding an NFAT reporter or a negative-control re-porter (Cignal Lenti NFAT reporter assay and Cignal Lenti reporter neg-ative control; Qiagen), according to the manufacturer’s protocol. Briefly,Jurkat cells (1 � 104 cells per condition) were infected with lentiviralparticles carrying the desired reporter at a multiplicity of infection (MOI)of 50 viral particles per cell. After 3 days, the cell culture medium waschanged and supplemented with 1 �M puromycin. After 3 additionaldays, surviving clones were used for further experiments.

Luciferase assay. Jurkat cells carrying a stable luciferase reporter(NFAT or negative control) were cultured (1 � 105 cells per condition)and treated with viable H. pylori strains (MOI of 50 bacterial cells perJurkat cell) or H. pylori broth culture supernatant preparations for 1 h at37°C. Cells were then stimulated with PMA (50 ng/ml) and ionomycin(500 ng/ml; Sigma) for 6 h. Luciferase activity was measured using theluciferase assay system with reporter lysis buffer (Promega) according tothe manufacturer’s protocol. Luciferase activity is expressed as relativevalues (luciferase activity of cells containing NFAT reporter divided byluciferase activity of cells containing the negative-control reporter), andthe values for control cells (stimulated with PMA-ionomycin, withoutVacA treatment) are assigned a relative value of 1 (or 100%).

VacA binding to integrin. VacA binding to �2 integrin was evaluatedby ELISA. The wells of microtiter plates (Immunolon IB) were coatedwith 50 �l of recombinant �M�2 integrin or �V�3 integrin derived fromhuman CHO cells (R&D system) at 4°C for 24 h. Unbound protein wasthen removed and the wells were blocked with PBS containing 5% BSA at4°C for 48 h. After blocking, serial dilutions of H. pylori culture superna-tants containing equivalent concentrations of different forms of VacAwere added to the wells at 25°C for 1 h. Wells were then washed three timeswith PBS-0.05% Tween 20, and bound VacA was detected by incubatingthe wells with anti-VacA rabbit serum (diluted 1:1,000/serum no. 958),followed by incubating with horseradish peroxidase-conjugated second-ary antibody (diluted 1:1,000; Promega), each at 25°C for 1 h. Rabbitserum and secondary antibody were diluted in PBS containing 3% BSA.ELISA was developed using Strep Ultra TMB-ELISA (Thermo Scientific).

RESULTSManipulation of the vacA i-region. The VacA proteins secretedby different WT H. pylori strains vary markedly in amino acidsequences, and there are also differences among strains in thelevels of VacA secretion (4, 25). To facilitate analysis of the VacAi-region, we used an approach in which we manipulated the chro-mosome of reference H. pylori strains (60190 and X47) in a man-ner so that we altered the region of vacA encoding the i-region andmaintained all other regions of vacA without changes. For initialstudies, we altered vacA in strain 60190 (which contains type s1/i1/m1 vacA) so that two clusters of polymorphisms in the i-region(cluster B and cluster C) were changed from an i1 form to an i2

H. pylori VacA Intermediate Region

August 2012 Volume 80 Number 8 iai.asm.org 2581

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

form, as described in Materials and Methods. The modified strainwas designated 60190 i2BC (Table 1 and Fig. 2A). Immunoblotanalysis indicated that the modified H. pylori strain expressed andsecreted VacA in a manner similar to the WT strain (data notshown). The WT strain and modified H. pylori strain were grownin broth cultures, and bacterial supernatants were concentratedand normalized so that they contained equivalent concentrationof VacA, as described in Materials and Methods. To investigatewhether there were any detectable differences in the ability ofthese proteins to cause alterations in gastric epithelial cells, serialdilutions of the supernatant preparations were added to RK13cells. Consistent with results of a previous study (54), we did notdetect any difference in the ability of the i1 and i2 forms of VacA tocause vacuolation of RK13 cells (Fig. 2B). This suggests that ma-nipulation of the VacA i-region by this approach does not result inmisfolding of the protein.

Effects of type i1 and i2 VacA on IL-2 production by Jurkatcells. Previous studies have shown that type i1 forms of VacA cansuppress IL-2 secretion from Jurkat cells (1, 29, 57). To determinewhether type i1 and i2 forms of VacA differ in this activity, wecompared the ability of the WT i1 form of VacA and the i2BC formdescribed above to suppress IL-2 secretion by Jurkat cells. We alsomanipulated H. pylori strain 60190 so that individual polymor-phic regions within the vacA i-region (cluster B or cluster C) werechanged to type i2 (Fig. 3A). These modified strains are designated60190 i2B and 60190 i2C. Cluster A was not manipulated, sincesequence variation at this site has not been linked to disease out-come (54). Immunoblot analysis indicated that each of the mod-ified H. pylori strains expressed and secreted VacA, similar to theWT strain (data not shown). The WT and modified H. pyloristrains were grown in broth cultures, and supernatant prepara-tions containing equivalent concentrations of VacA were pre-pared, as described in Materials and Methods. Jurkat cells werepretreated with broth culture supernatant preparations from theWT and modified strains and were then stimulated with PMA andionomycin. IL-2 production by Jurkat cells was quantified byELISA, as described in Materials and Methods. In comparison to

supernatant from a vacA-null mutant strain, supernatant contain-ing WT i1 VacA suppressed IL-2 secretion, as expected (Fig. 3B).Supernatants containing i2 forms of VacA (i2B, i2C, or i2BC) alsosuppressed IL-2 secretion, but in comparison to i1 VacA, the i2forms of VacA had a significantly reduced capacity to suppressIL-2 secretion (Fig. 3B).

We next investigated the role of the VacA i-region in the con-text of H. pylori strain X47, which contains an s1/m2 type of vacA.The vacA gene in this strain contains a type i2 sequence in clusterC of the i-region, and cluster B is chimeric (Fig. 1). A previousstudy (54) reported that polymorphisms in cluster C accountedfor differences in the activity of i1 and i2 forms of VacA on epi-thelial cells. Therefore, we investigated whether changing clusterC of vacA in this strain from type i2 to type i1 would result in anincreased capacity of VacA to suppress IL-2 secretion from Jurkatcells. To do this, we manipulated H. pylori strain X47 as describedin Materials and Methods such that amino acids in cluster C of thevacA i-region were changed from an i2 to an i1 form, resulting ina strain designated X47 i1C (Fig. 3C). The WT and modified H.pylori strains were grown in broth cultures, and supernatant prep-arations containing equivalent concentrations of VacA were pre-pared, as described in Materials and Methods. Jurkat cells werepretreated with the H. pylori culture supernatant preparations andwere then stimulated with PMA and ionomycin. In comparison toVacA produced by the H. pylori X47 WT strain (X47 i2), VacAcontaining an i1 form of cluster C (X47 i1C) had an increasedinhibitory effect on IL-2 secretion by Jurkat cells (Fig. 3D).

Effects of purified VacA proteins on IL-2 production by Jur-kat cells. To further analyze the activities of type i1 and i2 VacAproteins, we purified i1 and i2 VacA proteins from broth culturesupernatant of either WT H. pylori 60190 (expressing i1 VacA) orthe 60190 i2BC strain (expressing an i2 form of VacA) (Table 1)and then tested the effects of these proteins on Jurkat cells. Thepurified VacA i1 protein suppressed IL-2 secretion from Jurkatcells, whereas the purified VacA i2 protein had relatively littleeffect (Fig. 4A). To corroborate the conclusion that i1 and i2 pro-teins differed in activity, we generated an additional modified

FIG 2 VacA-induced vacuolation of RK13 cells. (A) Amino acid sequence of the VacA i-region in WT H. pylori strain 60190 (type i1) and a strain expressing amodified VacA protein in which clusters B and C were changed from type i1 to type i2 (60190 i2BC). (B) Broth culture supernatants derived from the WT strain(type i1) or strain 60190 i2BC were concentrated and normalized so that they contained equivalent VacA concentrations, as described in Materials and Methods.Serial dilutions of VacA-containing preparations were then added to RK13 cells. Vacuolating activity was measured by a neutral red uptake assay. Relative VacAconcentrations are indicated. Results represent the means � standard deviations (SD) from triplicate samples.

González-Rivera et al.

2582 iai.asm.org Infection and Immunity

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

form of VacA and analyzed the activity of this purified protein.Specifically, we mutated the 5= end of vacA cluster C in H. pyloristrain 60190 so that it contained amino acids corresponding to i2sequences. Cluster C in this modified form of VacA, designatedi1/i2C VacA, contained an A-to-V substitution and an SNQ inser-

tion (VSNSNQSVKLNGN; for comparison, the type i1 sequenceis ASNSVKLNGN and the type i2 sequence is VSSSNQSVDLYGK[Fig. 1]). We then purified the WT i1 protein and the VacA i1/i2Cprotein (containing i2 amino acids in the 5= region of cluster C)from H. pylori supernatants and tested these proteins for their

FIG 3 Role of the VacA i-region in inhibition of IL-2 secretion by Jurkat cells. (A) Amino acid sequence of the VacA i-region in WT H. pylori strain 60190 (typei1) and modified strains. Modified strains were constructed so that 60190 i2B contains an i2 sequence in polymorphic cluster B and an i1 sequence in cluster C,60190 i2C contains an i2 sequence in cluster C and an i1 sequence in cluster B, and 60190 i2BC contains i2 sequences in both clusters B and C. (B) H. pylori strainswere cultured in broth, and preparations of culture supernatants were standardized so that they contained equivalent concentrations of VacA, as described inMaterials and Methods. Jurkat cells were pretreated with 1:20 or 1:50 dilutions of culture supernatant preparations, each containing the indicated VacA protein,and then stimulated with PMA-ionomycin. After 24 h, the cells were pelleted and the IL-2 content of supernatants was analyzed by ELISA. (C) Amino acidsequence of the VacA i-region in WT strain X47 and a modified strain. WT strain X47 contains an i2 sequence in cluster C, and the X47 i1C strain contains ani1 sequence in cluster C. (D) Jurkat cells were pretreated with 1:20 or 1:50 dilutions of culture supernatant preparations, each containing the indicated VacAprotein, and cells were then stimulated with PMA-ionomycin. After 24 h, the cells were pelleted, and the IL-2 content of supernatants was analyzed by ELISA.Results represent the means � standard deviations of triplicate samples of a single experiment. Similar results were obtained in two additional experiments. *, Pvalue of �0.05 compared to WT i1 VacA (A) or WT VacA from strain X47 (B) at a 1:20 dilution; ***, P value of �0.05 at a 1:50 dilution (analysis of variance[ANOVA] followed by Dunnett’s post hoc test for panel B; Student t test for panel D). Levels of IL-2 secretion are expressed as relative values (levels of IL-2 secretedby cells treated with WT VacA or modified VacA proteins, divided by levels of IL-2 secreted by cells treated with supernatant from the VacA-null mutant strain).Values for cells treated with supernatant from the VacA-null mutant strain are assigned a relative value of 1 (or 100%).

H. pylori VacA Intermediate Region

August 2012 Volume 80 Number 8 iai.asm.org 2583

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

ability to inhibit IL-2 production by Jurkat cells. In comparison tothe WT i1 VacA protein, the i1/i2C VacA protein was less potent inits ability to inhibit IL-2 production (Fig. 4B). We also attemptedto purify WT and modified VacA proteins expressed by H. pyloristrain X47, but this was not feasible due to a failure of this strain togrow in medium free of FBS (which is essential for purification ofVacA).

As another approach, we tested the activity of purified recom-binant VacA i1 and i2 proteins. We have previously shown that amixture of i1 p33 plus p55 VacA domains reconstitutes toxin ac-tivity in assays using HeLa cells (31, 68). Therefore, we expressedboth i1 and i2 forms of p33 as described in Materials and Methods,mixed either purified VacA p33 i1 or p33 i2 proteins with purifiedp55 (1:1 mass ratio), and tested the effects of these preparations onJurkat cells. In comparison to a mixture of p33 i1 plus purifiedp55, a mixture of p33 i2 plus purified p55 had a significantly re-duced capacity to suppress IL-2 secretion from Jurkat cells(Fig. 4C). Taken together, these results indicate that the i-region isan important determinant of the capacity of VacA to inhibit IL-2secretion.

Analysis of VacA effects on NFAT activation. Previous studieshave shown that the effect of VacA on IL-2 secretion by Jurkat cellsis dependent on inhibition of NFAT (29). We therefore investi-gated whether the composition of the VacA i-region influences theability of VacA to inhibit NFAT activation. We first transducedJurkat cells with replication-deficient lentiviral particles that carryan NFAT luciferase reporter or a negative-control reporter andselected for puromycin-resistant cells that contain the reporters,as described in Materials and Methods. We then cocultured thecells with viable H. pylori (60190 WT strain, 60190 vacA-null mu-tant strain, or 60190 i2BC) (Fig. 5A). After 1 h of incubation, cellswere stimulated with PMA-ionomycin for an additional 6 h, andluciferase was measured as described in Materials and Methods. Incomparison to the WT H. pylori strain (expressing type i1 VacA),which inhibited NFAT activation, H. pylori expressing type i2VacA (60190 i2BC) had an impaired ability to inhibit NFAT acti-vation (Fig. 5A). Similar results were obtained when analyzing H.pylori broth culture supernatant preparations (Fig. 5B). As shownin Fig. 5B, supernatant from the vacA-null mutant strain causedsome detectable inhibition of NFAT activation, which might beattributable to actions of other factors besides VacA on NFATactivation or nonspecific effects of the preparation on the lucifer-ase assay. In summary, the results obtained in these studies ofNFAT activation were concordant with results obtained in theIL-2 assays.

Analysis of VacA binding to Jurkat cells. To investigate a pos-sible mechanism for the observed differences in activities of i1 andi2 VacA proteins, we analyzed the binding properties of VacAproteins containing type i1 or type i2 i-regions. Broth culture su-pernatant preparations from H. pylori 60190 strains expressingeither i1 VacA or i2 VacA (60190 i2BC) proteins, as well as asupernatant preparation from a vacA-null mutant strain, wereincubated with Jurkat cells for 1 h at 4°C. Cells were then washedand immunoblotted with an anti-VacA antibody to detect VacAbinding. As shown in Fig. 6A and B (top), immunoblot analysis ofthe supernatants in the absence of Jurkat cells indicated that thelevels of VacA were similar in the normalized preparations fromWT and modified strains. In comparison to i2 VacA, i1 VacAbound more avidly to the cells (Fig. 6A, bottom). Additionally, wetested the binding of VacA proteins produced by H. pylori strain

FIG 4 Effects of purified VacA proteins on IL-2 secretion by Jurkat cells. (A)Jurkat cells were pretreated with purified p88 VacA proteins secreted by eitherWT strain 60190 (expressing type i1 VacA) or a modified strain expressingi2BC VacA, which contains i2 sequences in polymorphic clusters B and C. (B)Jurkat cells were pretreated with purified H. pylori VacA proteins secreted byeither WT strain 60190 (type i1) or a modified strain expressing an i1/i2Cprotein (as described in Results) at the indicated protein concentrations. Cellswere then stimulated with PMA-ionomycin, and after 24 h the cells were pel-leted and the IL-2 content of supernatants was analyzed by ELISA. (C) Recom-binant purified p33 proteins containing either i1 or i2 (clusters BC) amino acidsequences were mixed with purified p55. The indicated protein concentrationsfor the VacA p33-p55 mixture (1:1 mass ratio) correspond to the total proteinconcentration. Jurkat cells were pretreated with the VacA preparations andwere then stimulated with PMA-ionomycin. After 24 h, IL-2 production wasquantified by ELISA. Results represent the means � standard deviations oftriplicate samples of a single experiment. Similar results were obtained in twoadditional experiments. *, P value of �0.05 as determined by Student’s t test.Levels of IL-2 secretion are expressed as relative values (levels of IL-2 secretedby cells treated with WT VacA or modified VacA proteins, divided by levels ofIL-2 secreted by cells treated with buffer alone). Values for cells treated withbuffer are assigned a relative value of 1 (or 100%).

González-Rivera et al.

2584 iai.asm.org Infection and Immunity

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

X47 (X47 WT VacA [i2] and X47 i1C). Consistent with the resultsobtained when analyzing VacA proteins produced by strain60190, the WT VacA i2 protein from strain X47 exhibited de-creased avidity of binding compared to the i1C VacA protein (Fig.6B, bottom).

As another approach for analyzing VacA binding, we quanti-fied VacA binding to Jurkat cells using a flow cytometry-basedassay. For these experiments, we used recombinantly expressedp33 and p55 VacA domains. We have previously shown that p33can facilitate the binding of purified p55 to HeLa cells (31, 68).

Therefore, in the current experiments, we labeled the recombi-nant VacA p55 protein with Alexa 488, as described previously(31), and mixed the labeled p55 protein with either unlabeled p33i1 or unlabeled p33 i2 protein (1:1 mass ratio). These proteinmixtures were incubated with Jurkat cells for 1 h at 4°C, and cellswere then washed and analyzed by flow cytometry. When com-bined with unlabeled p33 i1, the labeled p55 protein bound moreavidly to Jurkat cells than when the labeled p55 protein alone wasadded to Jurkat cells (Fig. 6C, D). The labeled p55 protein boundsignificantly less avidly when mixed with p33 i2 than when mixedwith p33 i1 (Fig. 6C, D). Representative histograms are presentedin Fig. 5C, and quantification of levels of VacA binding is shown inFig. 5D. Collectively, these experiments indicate that, compared totype i2 forms of VacA, type i1 forms of VacA bind more avidly toJurkat cells.

Binding of type i1 and i2 VacA to �2 integrin. Previous stud-ies have shown that �2 integrin is a receptor for VacA in T cells(57). We hypothesized that the observed difference in binding ofi1 and i2 VacA proteins to Jurkat cells might be due to differencesin the binding of these proteins to �2 integrin. To test this hypoth-esis, we performed an ELISA-based binding assay as described inMaterials and Methods. Both VacA i1 and VacA i2 proteins boundto �M�2 integrin in a dose-dependent manner, and no significantdifferences in binding avidity were detected (Fig. 7). As expected,both forms of VacA bound less avidly to a control protein (�V�3integrin) than to �M�2 integrin. Taken together, these resultssuggest that the observed difference in binding of i1 and i2 VacAproteins to Jurkat cells is not attributable to differences in VacAbinding to the �2 integrin receptor.

DISCUSSION

VacA is one of the most important virulence factors produced byH. pylori (9, 17, 23, 38). Numerous studies have shown that H.pylori strains containing specific vacA types (such as s1 or m1) areassociated with a higher risk of gastric disease than are strainscontaining s2 or m2 vacA types (62). Correspondingly, type s1/m1forms of VacA exhibit increased cytotoxic activity in vitro com-pared to type s2/m2 forms of VacA (4, 39, 43). Recently, it wasreported that strains containing the type i1 forms of vacA are as-sociated with a higher risk of gastric disease than are strains con-taining type i2 forms of vacA (6, 8, 20, 33, 37, 54, 60). One studyreported that the i-region is a determinant of cell-type specificity(54), but thus far there has been very little study of the role of thei-region in VacA activity. In the current study, we tested the hy-pothesis that VacA i1 and i2 proteins differ in the ability to causefunctional alterations in T cells, using Jurkat cells as a model cellline.

In accordance with a previous study (54), we found that both i1and i2 forms of VacA caused vacuolation of RK13 cells. Both the i1and i2 VacA proteins inhibited IL-2 secretion and NFAT activa-tion in Jurkat cells, but the i2 VacA proteins had a reduced po-tency. Type i1 VacA proteins bound more avidly than type i2 VacAproteins to Jurkat cells, and this difference in binding probablyaccounts, at least in part, for the observed difference in activity.Previous studies have shown that binding of VacA to epithelialcells is mediated not only by the p55 domain but also by the p33domain (31, 68). The results in the current study provide addi-tional evidence that the VacA p33 domain contributes to VacAcell-binding properties.

The observed difference in the binding properties of i1 and i2

FIG 5 Effects of VacA proteins on NFAT activation. Jurkat cells stably ex-pressing an NFAT luciferase reporter or a negative-control luciferase reporterwere treated with viable H. pylori strains (WT strain 60190 [expressing i1VacA], vacA-null mutant strain, or 60190 i2BC [expressing i2BC VacA]) (A)or with H. pylori broth culture supernatant preparations derived from thesestrains (B) and then activated with PMA and ionomycin. Luciferase activitywas quantified by luminometry, as described in Materials and Methods. NFATactivity is expressed as relative values (luciferase activity of cells containingNFAT reporter divided by luciferase activity of cells containing the negative-control reporter), and the values for control cells (stimulated with PMA-iono-mycin, without VacA treatment) are assigned a relative value of 1 (or 100%).Results represent the means � standard deviations of triplicate samples of asingle experiment. Similar results were obtained in two additional experi-ments. *, P value of �0.05 as determined by using Student’s t test, comparingWT strain 60190 and a strain expressing VacA i2BC (A) or culture supernatantpreparations derived from these strains (B).

H. pylori VacA Intermediate Region

August 2012 Volume 80 Number 8 iai.asm.org 2585

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

VacA suggest that these proteins might differ in binding to a spe-cific receptor on the surface of Jurkat cells. As shown in Fig. 7, wedid not detect any significant difference in the binding of type i1and i2 VacA to �2 integrin, which is an important receptor forVacA on T cells (57). This result suggests that the i-region mightbe involved in VacA binding to alternate receptors which have notyet been characterized (57, 58). Various candidates for these alter-nate receptors include sphingomyelin, glycosylphosphatidylinosi-tol (GPI)-anchored proteins, or glycolipids (57, 58). Further stud-ies will be required to better understand the basis for thedifferential binding properties of i1 and i2 forms of VacA.

As shown in Fig. 1, type i1 and type i2 forms of VacA differ inamino acid sequences at a relatively small number of sites withinpolymorphic clusters A, B, and C, and these polymorphisms ac-count for most of the sequence variation that is observed within

the VacA p33 domain (28). Experiments in the current study in-dicate that polymorphisms in cluster C are important determi-nants of VacA activity in a Jurkat T cell assay. Prior to the currentstudy, a random mutagenesis study revealed that mutations in twoamino acids in close proximity to this region (T210A, S246L) al-tered the capacity of VacA to cause vacuolation in HeLa cells (44,45). Taken together, these studies highlight the functional impor-tance of this region of the p33 domain. At present, a crystal struc-ture is available for the p55 domain of VacA (27), but no structuraldata are available for the p33 domain. In future studies, it will beimportant to determine the structure of the p33 domain and toinvestigate the structural basis for the observed differences in ac-tivity of type i1 and i2 forms of VacA. In addition, it will be im-portant to determine whether the VacA i-region influences thepotency of other VacA activities, including a spectrum of altera-

FIG 6 Binding of VacA proteins to Jurkat cells. H. pylori strains were cultured in broth, and culture supernatant preparations were standardized so that theycontained equivalent concentrations of VacA. Jurkat cells were incubated with supernatant preparations from the indicated H. pylori strains for 1 h at 4°C. Cellswere washed and lysed, and protein samples were then analyzed by immunoblotting using an anti-VacA antibody. (A) Analysis of WT H. pylori strain 60190(expressing i1 VacA), 60190 vacA-null mutant strain, and 60190 i2BC (expressing i2BC VacA). (B) Analysis of WT H. pylori strain X47, X47 vacA-null mutantstrain, and strain X47 i1C. Top panels (labeled “60190 supernatant” and “X47 supernatant”) depict immunoblot analysis of H. pylori supernatant preparationsprior to the addition to Jurkat cells. Bottom panels (labeled “Bound VacA”) depict VacA binding to Jurkat cells. GAPDH was analyzed as a loading control. (C)Jurkat cells were treated with purified Alexa 488-labeled p55 (2.5 �g/ml) plus either purified p33 i1 or p33 i2 (2.5 �g/ml) for 1 h at 4°C. After treatment, cells werewashed and analyzed by flow cytometry. Values indicate mean fluorescence intensity (MFI), based on three independent samples, and the percent positive cells(% pos. cells), defined as the proportion of cells exhibiting detectable fluorescence in comparison to control cells. Representative histograms depicting VacAbinding are shown. (D) Graphical representation of VacA binding to Jurkat cells, based on flow cytometry analysis. These data are from an experiment performedon a separate day compared to the data in panel C. *, P value of �0.05 compared to p33 i2 and p55 Alexa 488, as determined by using Student’s t test. Resultsrepresent the means � standard deviations of triplicate samples from a single experiment.

González-Rivera et al.

2586 iai.asm.org Infection and Immunity

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

tions produced by VacA in gastric epithelial cells and several typesof immune cells (9).

It is striking that within three different regions of VacA (s-, i-,and m-regions), there is marked sequence variation among pro-teins expressed by different H. pylori strains, and analysis of eachregion indicates the existence of two main groups of VacA pro-teins categorized as type 1 (s1, i1, m1) and type 2 (s2, i2, m2) (4, 5,28, 54). In each case, the sequence variations are associated withdifferences in VacA activity toward host cells (4, 25, 36, 39, 43, 51,54, 61, 73). It may be presumed that selective forces had an im-portant role in the origin of these variations, as well as in themaintenance of the different allelic variants (28). In future studies,it will be important to determine how these different forms ofVacA each provide a selective advantage to H. pylori.

ACKNOWLEDGMENTS

We thank Borden Lacy for providing purified p55 protein, Beverly Hossefor assistance with VacA purification, and John Loh for assistance withpreparation of figures.

This work was supported by the National Institutes of Health(AI039657 and AI068009), the Molecular Microbial Pathogenesis Train-ing Program (T32 AI007281-21), the Vanderbilt University Digestive Dis-eases Research Center (DK058404), Vanderbilt Ingram Cancer Center,and the Department of Veterans Affairs.

REFERENCES1. Algood HM, Torres VJ, Unutmaz D, Cover TL. 2007. Resistance of

primary murine CD4� T cells to Helicobacter pylori vacuolating cyto-toxin. Infect. Immun. 75:334 –341.

2. Amieva MR, El-Omar EM. 2008. Host-bacterial interactions in Helico-bacter pylori infection. Gastroenterology 134:306 –323.

3. Atherton JC, Blaser MJ. 2009. Coadaptation of Helicobacter pylori andhumans: ancient history, modern implications. J. Clin. Invest. 119:2475–2487.

4. Atherton JC, et al. 1995. Mosaicism in vacuolating cytotoxin alleles ofHelicobacter pylori. Association of specific vacA types with cytotoxin pro-duction and peptic ulceration. J. Biol. Chem. 270:17771–17777.

5. Atherton JC, Peek RM, Jr, Tham KT, Cover TL, Blaser MJ. 1997.

Clinical and pathological importance of heterogeneity in vacA, the vacu-olating cytotoxin gene of Helicobacter pylori. Gastroenterology 112:92–99.

6. Basso D, et al. 2008. Clinical relevance of Helicobacter pylori cagA andvacA gene polymorphisms. Gastroenterology 135:91–99.

7. Boncristiano M, et al. 2003. The Helicobacter pylori vacuolating toxininhibits T cell activation by two independent mechanisms. J. Exp. Med.198:1887–1897.

8. Chung C, et al. 2010. Diversity of VacA intermediate region amongHelicobacter pylori strains from several regions of the world. J. Clin. Mi-crobiol. 48:690 – 696.

9. Cover TL, Blanke SR. 2005. Helicobacter pylori VacA, a paradigm fortoxin multifunctionality. Nat. Rev. Microbiol. 3:320 –332.

10. Cover TL, Blaser MJ. 2009. Helicobacter pylori in health and disease.Gastroenterology 136:1863–1873.

11. Cover TL, Blaser MJ. 1992. Purification and characterization of the vac-uolating toxin from Helicobacter pylori. J. Biol. Chem. 267:10570 –10575.

12. Cover TL, Hanson PI, Heuser JE. 1997. Acid-induced dissociation ofVacA, the Helicobacter pylori vacuolating cytotoxin, reveals its pattern ofassembly. J. Cell Biol. 138:759 –769.

13. Cover TL, Krishna US, Israel DA, Peek RM, Jr. 2003. Induction ofgastric epithelial cell apoptosis by Helicobacter pylori vacuolating cyto-toxin. Cancer Res. 63:951–957.

14. Cover TL, Puryear W, Perez-Perez GI, Blaser MJ. 1991. Effect of ureaseon HeLa cell vacuolation induced by Helicobacter pylori cytotoxin. Infect.Immun. 59:1264 –1270.

15. Cover TL, Tummuru MK, Cao P, Thompson SA, Blaser MJ. 1994.Divergence of genetic sequences for the vacuolating cytotoxin amongHelicobacter pylori strains. J. Biol. Chem. 269:10566 –10573.

16. Dautin N, Bernstein HD. 2007. Protein secretion in Gram-negative bac-teria via the autotransporter pathway. Annu. Rev. Microbiol. 61:89 –112.

17. de Bernard M, Cappon A, Del Giudice G, Rappuoli R, Montecucco C.2004. The multiple cellular activities of the VacA cytotoxin of Helicobacterpylori. Int. J. Med. Microbiol. 293:589 –597.

18. de Bernard M, et al. 1995. Low pH activates the vacuolating toxin ofHelicobacter pylori, which becomes acid and pepsin resistant. J. Biol.Chem. 270:23937–23940.

19. Domanska G, et al. 2010. Helicobacter pylori VacA toxin/subunit p34:targeting of an anion channel to the inner mitochondrial membrane. PLoSPathog. 6:e1000878. doi:10.1371/journal.ppat.1000878.

20. Douraghi M, et al. 2009. Multiple gene status in Helicobacter pylori strainsand risk of gastric cancer development. Digestion 80:200 –207.

21. El-Bez C, Adrian M, Dubochet J, Cover TL. 2005. High resolutionstructural analysis of Helicobacter pylori VacA toxin oligomers by cryo-negative staining electron microscopy. J. Struct. Biol. 151:215–228.

22. Fischer W, Buhrdorf R, Gerland E, Haas R. 2001. Outer membranetargeting of passenger proteins by the vacuolating cytotoxin autotrans-porter of Helicobacter pylori. Infect. Immun. 69:6769 – 6775.

23. Fischer W, Prassl S, Haas R. 2009. Virulence mechanisms and persistencestrategies of the human gastric pathogen Helicobacter pylori. Curr. Top.Microbiol. Immunol. 337:129 –171.

24. Foo JH, et al. 2010. Both the p33 and p55 subunits of the Helicobacterpylori VacA toxin are targeted to mammalian mitochondria. J. Mol. Biol.401:792–798.

25. Forsyth MH, Atherton JC, Blaser MJ, Cover TL. 1998. Heterogeneity inlevels of vacuolating cytotoxin gene (vacA) transcription among Helico-bacter pylori strains. Infect. Immun. 66:3088 –3094.

26. Galmiche A, et al. 2000. The N-terminal 34 kDa fragment of Helicobacterpylori vacuolating cytotoxin targets mitochondria and induces cyto-chrome c release. EMBO J. 19:6361– 6370.

27. Gangwer KA, et al. 2007. Crystal structure of the Helicobacter pylorivacuolating toxin p55 domain. Proc. Natl. Acad. Sci. U. S. A. 104:16293–16298.

28. Gangwer KA, et al. 2010. Molecular evolution of the Helicobacter pylorivacuolating toxin gene vacA. J. Bacteriol. 192:6126 – 6135.

29. Gebert B, Fischer W, Weiss E, Hoffmann R, Haas R. 2003. Helicobacterpylori vacuolating cytotoxin inhibits T lymphocyte activation. Science301:1099 –1102.

30. Genisset C, et al. 2006. A Helicobacter pylori vacuolating toxin mutantthat fails to oligomerize has a dominant negative phenotype. Infect. Im-mun. 74:1786 –1794.

31. Gonzalez-Rivera C, et al. 2010. Reconstitution of Helicobacter pyloriVacA toxin from purified components. Biochemistry 49:5743–5752.

32. Handt LK, et al. 1995. Characterization of feline Helicobacter pylori

FIG 7 Binding of type i1 and i2 VacA proteins to �2 integrin. Wells of micro-titer plates were coated with �M�2 integrin or �V�3 integrin as described inMaterials and Methods. Serial dilutions of culture supernatants from H. pylori60190 strains, containing equivalent concentrations of either WT (i1) or i2BCforms of VacA, were then added and incubated for 1 h. Unbound protein wasremoved, wells were washed, and VacA binding was analyzed by ELISA, asdescribed in Materials and Methods. The background absorbance (VacA bind-ing to wells in the absence of integrin) was subtracted from all absorbancevalues. Results represent the means � standard deviations of triplicate exper-iments and are representative of three independent experiments.

H. pylori VacA Intermediate Region

August 2012 Volume 80 Number 8 iai.asm.org 2587

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

strains and associated gastritis in a colony of domestic cats. J. Clin. Micro-biol. 33:2280 –2289.

33. Hussein NR, et al. 2008. Differences in virulence markers between Heli-cobacter pylori strains from Iraq and those from Iran: potential importanceof regional differences in H. pylori-associated disease. J. Clin. Microbiol.46:1774 –1779.

34. Ivie SE, et al. 2008. Helicobacter pylori VacA subdomain required forintracellular toxin activity and assembly of functional oligomeric com-plexes. Infect. Immun. 76:2843–2851.

35. Jain P, Luo ZQ, Blanke SR. 2011. Helicobacter pylori vacuolating cyto-toxin A (VacA) engages the mitochondrial fission machinery to inducehost cell death. Proc. Natl. Acad. Sci. U. S. A. 108:16032–16037.

36. Ji X, et al. 2000. Cell specificity of Helicobacter pylori cytotoxin is deter-mined by a short region in the polymorphic midregion. Infect. Immun.68:3754 –3757.

37. Jones KR, et al. 2011. Polymorphisms in the intermediate region of VacAimpact Helicobacter pylori-induced disease development. J. Clin. Micro-biol. 49:101–110.

38. Jones KR, Whitmire JM, Merrell DS. 2010. A tale of two toxins: Helico-bacter pylori CagA and VacA modulate host pathways that impact disease.Front. Microbiol. 1:115.

39. Letley DP, Rhead JL, Twells RJ, Dove B, Atherton JC. 2003. Determi-nants of non-toxicity in the gastric pathogen Helicobacter pylori. J. Biol.Chem. 278:26734 –26741.

40. Leunk RD, Johnson PT, David BC, Kraft WG, Morgan DR. 1988.Cytotoxic activity in broth-culture filtrates of Campylobacter pylori. J.Med. Microbiol. 26:93–99.

41. Loh JT, et al. 2011. Analysis of cagA in Helicobacter pylori strains fromColombian populations with contrasting gastric cancer risk reveals a bio-marker for disease severity. Cancer Epidemiol. Biomarkers Prev. 20:2237–2249.

42. Lupetti P, et al. 1996. Oligomeric and subunit structure of the Helicobac-ter pylori vacuolating cytotoxin. J. Cell Biol. 133:801– 807.

43. McClain MS, et al. 2001. A 12-amino-acid segment, present in type s2 butnot type s1 Helicobacter pylori VacA proteins, abolishes cytotoxin activityand alters membrane channel formation. J. Bacteriol. 183:6499 – 6508.

44. McClain MS, Cover TL. 2003. Expression of Helicobacter pylori vacuo-lating toxin in Escherichia coli. Infect. Immun. 71:2266 –2271.

45. McClain MS, et al. 2006. Random mutagenesis of Helicobacter pylori vacAto identify amino acids essential for vacuolating cytotoxic activity. Infect.Immun. 74:6188 – 6195.

46. McClain MS, et al. 2003. Essential role of a GXXXG motif for membranechannel formation by Helicobacter pylori vacuolating toxin. J. Biol. Chem.278:12101–12108.

47. McClain MS, Schraw W, Ricci V, Boquet P, Cover TL. 2000. Acidactivation of Helicobacter pylori vacuolating cytotoxin (VacA) results intoxin internalization by eukaryotic cells. Mol. Microbiol. 37:433– 442.

48. Molinari M, et al. 1998. The acid activation of Helicobacter pylori toxinVacA: structural and membrane binding studies. Biochem. Biophys. Res.Commun. 248:334 –340.

49. Nakayama M, et al. 2004. Helicobacter pylori VacA activates the p38/activating transcription factor 2-mediated signal pathway in AZ-521 cells.J. Biol. Chem. 279:7024 –7028.

50. Nguyen VQ, Caprioli RM, Cover TL. 2001. Carboxy-terminal proteo-lytic processing of Helicobacter pylori vacuolating toxin. Infect. Immun.69:543–546.

51. Pagliaccia C, et al. 1998. The m2 form of the Helicobacter pylori cytotoxinhas cell type-specific vacuolating activity. Proc. Natl. Acad. Sci. U. S. A.95:10212–10217.

52. Polk DB, Peek RM, Jr. 2010. Helicobacter pylori: gastric cancer and be-yond. Nat. Rev. Cancer 10:403– 414.

53. Radin JN, Gonzalez-Rivera C, Ivie SE, McClain MS, Cover TL. 2011.Helicobacter pylori VacA induces programmed necrosis in gastric epithe-lial cells. Infect. Immun. 79:2535–2543.

54. Rhead JL, et al. 2007. A new Helicobacter pylori vacuolating cytotoxindeterminant, the intermediate region, is associated with gastric cancer.Gastroenterology 133:926 –936.

55. Schmitt W, Haas R. 1994. Genetic analysis of the Helicobacter pylori

vacuolating cytotoxin: structural similarities with the IgA protease type ofexported protein. Mol. Microbiol. 12:307–319.

56. Schraw W, Li Y, McClain MS, van der Goot FG, Cover TL. 2002.Association of Helicobacter pylori vacuolating toxin (VacA) with lipidrafts. J. Biol. Chem. 277:34642–34650.

57. Sewald X, et al. 2008. Integrin subunit CD18 is the T-lymphocyte recep-tor for the Helicobacter pylori vacuolating cytotoxin. Cell Host Microbe3:20 –29.

58. Sewald X, Jimenez-Soto L, Haas R. 2011. PKC-dependent endocytosis ofthe Helicobacter pylori vacuolating cytotoxin in primary T lymphocytes.Cell Microbiol. 13:482– 496.

59. Shaffer CL, et al. 2011. Helicobacter pylori exploits a unique repertoire oftype IV secretion system components for pilus assembly at the bacteria-host cel l interface. PLoS Pathog. 7 :e1002237. doi:10.1371/journal.ppat.1002237.

60. Sheu SM, Hung KH, Sheu BS, Yang HB, Wu JJ. 2009. Association ofnonsynonymous substitutions in the intermediate region of the vacA geneof Helicobacter pylori with gastric diseases in Taiwan. J. Clin. Microbiol.47:249 –251.

61. Skibinski DA, Genisset C, Barone S, Telford JL. 2006. The cell-specificphenotype of the polymorphic vacA midregion is independent of the ap-pearance of the cell surface receptor protein tyrosine phosphatase beta.Infect. Immun. 74:49 –55.

62. Strobel S, et al. 1998. Identification and analysis of a new vacA genotypevariant of Helicobacter pylori in different patient groups in Germany. J.Clin. Microbiol. 36:1285–1289.

63. Suerbaum S, Michetti P. 2002. Helicobacter pylori infection. N. Engl. J.Med. 347:1175–1186.

64. Sundrud MS, Torres VJ, Unutmaz D, Cover TL. 2004. Inhibition ofprimary human T cell proliferation by Helicobacter pylori vacuolatingtoxin (VacA) is independent of VacA effects on IL-2 secretion. Proc. Natl.Acad. Sci. U. S. A. 101:7727–7732.

65. Szabo I, et al. 1999. Formation of anion-selective channels in the cellplasma membrane by the toxin VacA of Helicobacter pylori is required forits biological activity. EMBO J. 18:5517–5527.

66. Telford JL, et al. 1994. Gene structure of the Helicobacter pylori cytotoxinand evidence of its key role in gastric disease. J. Exp. Med. 179:1653–1658.

67. Terebiznik MR, et al. 2009. Effect of Helicobacter pylori’s vacuolatingcytotoxin on the autophagy pathway in gastric epithelial cells. Autophagy5:370 –379.

68. Torres VJ, Ivie SE, McClain MS, Cover TL. 2005. Functional propertiesof the p33 and p55 domains of the Helicobacter pylori vacuolating cyto-toxin. J. Biol. Chem. 280:21107–21114.

69. Torres VJ, VanCompernolle SE, Sundrud MS, Unutmaz D, Cover TL.2007. Helicobacter pylori vacuolating cytotoxin inhibits activation-induced proliferation of human T and B lymphocyte subsets. J. Immunol.179:5433–5440.

70. Van Doorn LJ, et al. 1999. Geographic distribution of vacA allelic types ofHelicobacter pylori. Gastroenterology 116:823– 830.

71. Vinion-Dubiel AD, et al. 1999. A dominant negative mutant of Helico-bacter pylori vacuolating toxin (VacA) inhibits VacA-induced cell vacuo-lation. J. Biol. Chem. 274:37736 –37742.

72. Wang HJ, Wang WC. 2000. Expression and binding analysis of GST-VacA fusions reveals that the C-terminal approximately 100-residue seg-ment of exotoxin is crucial for binding in HeLa cells. Biochem. Biophys.Res. Commun. 278:449 – 454.

73. Wang WC, Wang HJ, Kuo CH. 2001. Two distinctive cell binding pat-terns by vacuolating toxin fused with glutathione S-transferase: one high-affinity m1-specific binding and the other lower-affinity binding for vari-ant m forms. Biochemistry 40:11887–11896.

74. Willhite DC, Blanke SR. 2004. Helicobacter pylori vacuolating cytotoxinenters cells, localizes to the mitochondria, and induces mitochondrialmembrane permeability changes correlated to toxin channel activity. CellMicrobiol. 6:143–154.

75. Ye D, Blanke SR. 2002. Functional complementation reveals the impor-tance of intermolecular monomer interactions for Helicobacter pyloriVacA vacuolating activity. Mol. Microbiol. 43:1243–1253.

González-Rivera et al.

2588 iai.asm.org Infection and Immunity

on June 24, 2020 by guesthttp://iai.asm

.org/D

ownloaded from