H. pylori VacA

1

Interactions between p-33 and p-55 domains of the Helicobacter pylori Vacuolating

Cytotoxin (VacA)

Victor J. Torres1, Mark S. McClain2 and Timothy L. Cover1,2,3

From the Department of Microbiology and Immunology1, and Department of Medicine2

Vanderbilt University School of Medicine, Nashville, Tennessee 37232-2605, and

Department of Veterans Affairs Medical Center3, Nashville, Tennessee 37212

Corresponding author. Mailing address: Division of Infectious Diseases, A3310 MCN,

Vanderbilt University School of Medicine, Nashville, TN 37232. Phone: (615) 322-2035.

Fax: (615) 343-6160. E-mail: timothy.L.cover@vanderbilt.edu.

Running title: H. pylori VacA

Copyright 2003 by The American Society for Biochemistry and Molecular Biology, Inc.

JBC Papers in Press. Published on October 30, 2003 as Manuscript M310159200 by guest on M

arch 3, 2018http://w

ww

.jbc.org/D

ownloaded from

H. pylori VacA

2

Summary

The VacA toxin secreted by Helicobacter pylori is considered to be an important

virulence factor in the pathogenesis of peptic ulcer disease and gastric cancer. VacA

monomers self-assemble into water-soluble oligomeric structures and can form anion-

selective membrane channels. The goal of this study was to characterize VacA-VacA

interactions that may mediate assembly of VacA monomers into higher order structures.

We investigated potential interactions between two domains of VacA (termed p-33 and p-

55) by using a yeast two-hybrid system. p-33 / p-55 interactions were detected in this

system, whereas p-33 / p-33 and p-55 / p-55 interactions were not detected. Several p-33

proteins containing internal deletion mutations were unable to interact with wild-type p-

55 in the yeast two-hybrid system. Introduction of these same deletion mutations into the

H. pylori vacA gene resulted in secretion of mutant VacA proteins that failed to assemble

into large oligomeric structures and that lacked vacuolating toxic activity for HeLa cells.

Additional mapping studies in the yeast two-hybrid system indicated that only the N-

terminal portion of the p-55 domain is required for p-33 / p-55 interactions. To further

characterize p-33 / p55 interactions, we engineered an H. pylori strain that produced a

VacA toxin containing an enterokinase cleavage site located between the p-33 and p-55

domains. Enterokinase treatment resulted in complete proteolysis of VacA into p-33 and

p-55 domains, which remained physically associated within oligomeric structures and

retained vacuolating cytotoxin activity. These results provide evidence that interactions

between p-33 and p-55 domains play an important role in VacA assembly into oligomeric

structures.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

3

Introduction

Helicobacter pylori is a gram-negative, spiral-shaped, microaerophilic bacterium

that colonizes the gastric mucosa of more than 50% of the human population (1).

Colonization of the gastric mucosa by H. pylori results in mucosal inflammation and is

recognized as a major risk factor for the development of peptic ulcer disease, distal

gastric adenocarcinoma and gastric lymphoma (2-4).

Most virulent H. pylori strains secrete a cytotoxin known as VacA into the

extracellular space (5-8). Several lines of evidence indicate that VacA contributes to the

capacity of H. pylori to colonize the stomach and that this toxin is an important virulence

factor in the pathogenesis of peptic ulceration and gastric cancer (6-12). The most

extensively characterized activity of VacA is its capacity to induce vacuolation in

mammalian cells (5,13). The membranes of these VacA-induced vacuoles contain

markers for late endosomal and lysosomal compartments (14). The precise mechanism

of VacA-induced vacuole formation is not completely understood but is thought to

involve binding of the toxin to the plasma membrane, internalization of the toxin, and

action of the toxin in an intracellular site (6-8). One model proposes that VacA forms

anion-selective channels in the membrane of late endocytic compartments (6-8,15,16). In

addition to inducing the formation of intracellular vacuoles, VacA causes multiple other

effects on target cells, including depolarization of the membrane potential (16,17),

permeabilization of epithelial monolayers (18), apoptosis (19-22), detachment of

epithelial cells from the basement membrane (12), interference with the process of

antigen presentation (23), and inhibition of T lymphocyte activation (24).

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

4

The vacA gene is translated as a 139-140 kDa protoxin and shows no striking

similarity to any other known bacterial toxin (5,11,25,26). Upon expression, this

protoxin undergoes amino- and carboxyl-terminal processing to yield a mature 88 kDa

secreted VacA toxin (27). The mature secreted 88 kDa VacA toxin can undergo

spontaneous proteolytic degradation into fragments that are about 33 kDa and 55 kDa in

mass (11,27-29). This degradation has been observed especially when preparations of the

purified toxin are stored for prolonged periods (11,28). The site of proteolytic cleavage is

predicted to be located within a hydrophilic surface-exposed loop (11,27). It has been

presumed that the two fragments (termed p-33 and p-55) represent two domains or

subunits of VacA (11,30,31). It should be noted that the nomenclature for describing

these fragments has not been uniform; several previous studies have described 37 kDa

and 58 kDa fragments (11,30,32-36). In the current study, we designate the two

fragments as p-33 and p-55, based on results on mass spectrometry analysis (27).

Previous studies have identified specific functions that are attributable to either

the p-33 or the p-55 domain. Several lines of evidence indicate that amino acid

sequences within the p-55 domain (located at the C-terminus of the mature secreted

toxin) mediate binding of VacA to host cells (28,31,37,38). VacA binding to cells is

inhibited by antiserum reactive with the p-55 fragment, but not inhibited by antiserum to

the p-33 fragment (28). Also, a truncated form of VacA containing mainly the p-55

VacA fragment was reported to bind to target cells in a manner similar to the full-length

VacA (31). Amino acid sequences located within a hydrophobic region near the N-

terminus (within the p-33 domain) are required for membrane channel formation by

VacA (39,40). When expressed in transiently transfected cells, neither the p-33 nor the

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

5

p-55 domain alone is sufficient for vacuolating toxin activity (32,34,41). In transiently

transfected cells, expression of the entire p-33 domain as well as about 109 amino acids

from the N-terminus of the p-55 domain is required for vacuolating toxin activity (32).

Secreted VacA proteins assemble into large water-soluble flower-shaped

structures comprised predominantly of 12 to 14 VacA monomers (29,30,42). Assembly

of VacA into oligomeric structures is presumably required for membrane channel

formation. This hypothesis is supported by electron microscopy studies in which VacA

associated with membranes appears as oligomeric structures (15,42), as well as

electrophysiologic studies in which the kinetics of membrane channel formation by VacA

have been investigated (43). Thus far, the process by which VacA assembles into higher

order structures has not been studied in any detail. Therefore, the goal of this study was

to characterize the interactions that may mediate assembly of VacA into oligomeric

structures. We describe here several lines of evidence indicating that interactions occur

between the p-33 and p-55 domains of VacA, and we identify candidate regions within

these two domains that are required for these interactions. In addition, we propose that

interactions between p-33 and p-55 domains are essential for the assembly of VacA into

oligomeric structures and for VacA vacuolating cytotoxin activity.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

6

Experimental Procedures

Bacterial and yeast strains, media, and growth condition. H. pylori strain

60190 (ATCC 49503) was grown on Trypticase soy agar plates containing 5% sheep

blood at 37 oC in ambient air containing 5% CO2. Liquid cultures were grown in sulfite-

free brucella broth supplemented with either 5% fetal bovine serum or 0.5% activated

charcoal. Escherichia coli XL1-Blue (Stratagene) was used for plasmid propagation and

was grown on Luria-Bertani (LB) broth or LB agar at 37 oC. Yeast-two hybrid

experiments were performed with Saccharomyces cerevisiae strain YRG-2 (MAT ura3-

52 his3-200 ade2-101 lys2-801 trp1-901 leu2-3,112 gal4-542 gal80-538 LYS2::UASGAL1

-TATAGAL1-HIS3 URA3::UASGAL4 17mers(x3)-TATACYC1-lacZ) (Stratagene). Yeast strains

were grown in rich medium (yeast extract, peptone, dextrose [YPAD]) or in synthetic

defined (SD) minimal medium (supplemented with required amino acids and glucose) at

30 oC as described in the GAL4 Two-Hybrid Phagemid manual (Stratagene).

Construction of VacA yeast-two hybrid vectors containing vacA fragments.

vacA sequences encoding the wild-type p-33 and p-55 VacA domains were PCR-

amplified from genomic DNA of H. pylori strain 60190 and cloned into plasmids

encoding the transcription-activation domain (pAD) and/or the DNA-binding domain

(pBD) of the GAL4 Two-Hybrid Phagemid system (Stratagene). The sequences of the

primers used for cloning the different vacA fragments into the yeast-two hybrid vectors

are described in Table 1. Primers combinations, restriction site used for cloning the PCR

products into the pAD and pBD vectors, and the amino acid numbers of the VacA

fragments are described in Table 2.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

7

To facilitate the introduction of in-frame deletion mutations into the pAD33

plasmid, vacA sequences encoding p-33 mutant domains were PCR-amplified from

genomic DNA of previously described H. pylori strains harboring the specific in-frame

vacA deletions (39), and cloned into the pAD vector (Table 2). vacA sequences encoding

p-55 fragments were PCR-amplified from genomic DNA of H. pylori strain 60190 (Table

2) and were cloned into the pBD vector. The entire vacA fragment in each p-33 and p-55

encoding plasmid was analyzed by nucleotide sequence analysis in order to verify that no

unintended mutations had been introduced.

Yeast methods. Yeast strains were co-transformed with 400 ng of individual

plasmids using the lithium acetate method (44) and cultured on SD medium

supplemented with the required amino acids and glucose at 30 °C as described in the

GAL4 Two-Hybrid Phagemid manual (Stratagene). For a positive protein-protein

interaction control, we co-transformed the yeast with pBD-WT and pAD-WT plasmids,

which encode fusion proteins consisting of amino acids 132-236 of wild-type lambda cI,

fragment C, together with either the GAL4- BD or AD, respectively. For a negative

control (i.e., two proteins that do not interact), we used the pAD-WT plasmid co-

transformed with a pBD-pLamin C plasmid, which expresses the BD of GAL4 fused to

amino acids 60-230 of human lamin C. We also used plasmids pAD-Mut and pBD-Mut,

which encode a mutated (E233K) lambda cI, fragment C, fused to either the GAL4- AD

or BD. The cI-E233K mutation interferes with the interaction between the cI monomers,

resulting in a protein-protein interaction that is weaker than that of wild-type proteins.

Plasmids expressing positive and negative interaction control proteins were obtained

from Stratagene.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

8

One colony of each co-transformant was grown in 2 ml of SD medium containing

2% (w/v) glucose and lacking Trp and Leu (SD-Leu,-Trp) at 30 °C with aeration for ~24

h. Cultures were then normalized to an OD600 of about 0.26 in a final volume of 100 l

of SD-Leu,-Trp,- is broth, and 10-fold serially diluted into SD-Leu,-Trp,- is broth in a

micro-titer plate. Diluted cultures (5 l) were seeded on duplicate SD-Leu,-Trp plates

(selection for plasmids) and SD-Leu,–Trp, is plates (selection for protein-protein

interactions) and incubated at 30 °C for 3-10 days.

To confirm the occurrence of protein–protein interactions, a -galactosidase

liquid assay was performed using the Yeast -Galactosidase Assay Kit as described by

the manufacturer (Pierce). Briefly, individual co-transformants were grown in 2 ml SD -

Leu,-Trp medium containing 2% (w/v) glucose at 30 °C with aeration for ~16-24 h until

cultures reached mid-log phase (OD600 0.4-0.5). 150 l aliquots of the cultures were then

mixed with 150 l of the working solution (equal volume of 2X -Galactosidase Assay

Buffer and Y-PERTM Reagent) and then incubated at 37 °C until solutions turned yellow

(approximately 1-4 h), at which time reactions were stopped by the addition of 175 l of

-Galactosidase Assay Stop Solution. The time required for yellow color development

was recorded for each tube. The OD410 of clarified reaction supernatants (200 l) were

measured in a 96-well plate and the -galactosidase activity was calculated by using the

Miller equation as described in the Yeast -galactosidase Assay Protocol (Pierce). The

values presented are the average of at least three independent co-transformants (means +

S.D.). Statistical significance was analyzed using Student’s t test.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

9

Introduction of DNA encoding a FLAG-tag epitope and enterokinase-site

into the vacA gene of H. pylori 60190. To modify the vacA gene so that it encoded a

VacA protein containing a FLAG-tag (DYKDADDDK) and an enterokinase cleavage site

(after the underlined K), complementary primers OP2891 and OP2892 encoding the

FLAG epitope (Table 1) were annealed and ligated into the unique StuI site of plasmid

pA178 (45). The resulting plasmid, pA178-FLAG, contained a sequence encoding the

FLAG epitope in the proper orientation inserted between amino acids 317 and 318 of

VacA. Plasmid pA178-FLAG was then used to transform H. pylori strain VM022 (39),

and transformants were selected by growth on 5.5% sucrose plates, as described

previously (39,46). Analysis of a single transformant (H. pylori VM088) by PCR and by

DNA sequence analysis confirmed that the sequence encoding the FLAG epitope and the

enterokinase cleavage site had been introduced into the desired location.

Purification of oligomeric forms of VacA. VacA was purified from broth

culture supernatants of H. pylori as described previously (29). Briefly, broth culture

supernatant proteins were concentrated by precipitation with a 50% saturated solution of

ammonium sulfate and resuspended in phosphate-buffered saline. Oligomeric VacA

(~1,000 kDa) was purified by gel-filtration chromatography using a Superose 6HR 16/50

column. To analyze the oligomeric state of the enterokinase-treated FLAG-VacA, 40 g

of cleaved toxin was applied to a Superose 6HR 16/50 column and fractions (2 ml each)

were collected. Fractions were analyzed by immunoblot analysis with an anti-VacA

serum.

Proteolysis of FLAG-VacA with enterokinase. Purified FLAG-VacA was

proteolytically cleaved with enterokinase as described by the Recombinant Enterokinase

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

10

Kit manual (Novagen). Briefly, 10 g of purified FLAG-VacA was incubated with 1 unit

of enterokinase and 5 l of 10X enterokinase cleavage buffer in a final volume of 50 l

for >16 h at room temperature. For mock treatment, FLAG-VacA was treated in the

same manner, but without enterokinase. Proteolysis of the full-length FLAG-VacA was

assessed by immunoblot analysis using an anti-VacA serum or the M2 monoclonal anti-

FLAG antibody (Sigma).

Cell culture and vacuolating assays. HeLa cells were grown in minimal

essential medium (modified Eagle’s medium containing Earle’s salts; MEM)

supplemented with 10% fetal bovine serum. For vacuolating assays, HeLa cells were

seeded (1.5 x 104) into 96-well plates 24 h prior to each experiment. Protein

concentrations of the purified VacA toxins were determined by using a Micro-BCA assay

(Pierce). Serial dilutions of purified VacA toxins (standardized by protein concentration)

were incubated with cells in serum-free medium containing 10 mM ammonium chloride

as described previously (5). Purified VacA preparations (VacA, FLAG-VacA and

cleaved FLAG-VacA) were acid-activated by adjusting them to pH 3 by the addition of

250 mM hydrochloric acid before they were added to cell culture wells (47,48).After 24 h

incubation of VacA toxins with the cell monolayers, cells were examined by inverted

light microscopy. Toxins that induced vacuoles in more than 50% of the cells were

scored positive for vacuolating cytotoxic activity. Vacuolating activity was also

quantified by a neutral red uptake assay as described previously (49). Neutral red uptake

data are presented as OD540 values (means + S.D.). Statistical significance was analyzed

using Student’s t test.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

11

Results

Detection of p-33 / p-55 interactions in a yeast-two hybrid system. Previous

studies have reported that 88 kDa VacA monomers commonly undergo spontaneous

degradation, yielding 33 and 55 kDa fragments (Figure 1) (11,27,28). These two

fragments have been considered to represent two domains or subunits of VacA

(11,20,30,31). To investigate potential interactions among these two VacA fragments,

we investigated whether they were able to interact in a GAL4 transcription activator

yeast-two hybrid system (GAL4 Two-Hybrid Phagemid; Stratagene) (50). This

particular yeast-two hybrid system was selected based on the low expression of the bait

and prey fusion proteins, which in turn, is thought to reduce the number of false positive

interactions (51). In this system, the yeast GAL4 transcription activator has been divided

into two separate functional domains (52): (i) the transcription-activation domain (AD)

present on plasmid pAD-GAL4-2.1 (pAD), which encodes for the LEU2 gene as a

selectable marker, and (ii) the DNA-binding domain (BD) present on the plasmid pBD-

GAL4-Cam (pBD), which encodes for the TRP1 gene as a selectable marker. If two

fusion proteins interact in this system, they will bring in close proximity the GAL4

transcription-activation domain and the GAL4 DNA-binding domain to the GAL4/GAL1

promoter, which in turn, will initiate the transcription of the HIS3 and lacZ reporter

genes. Protein-protein interactions are then detected by the ability of co-transformed

yeast cells to grow in selective medium lacking Leu, Trp, and His (SD-Leu,-Trp,-His)

and by production of -galactosidase activity.

vacA sequences encoding p-33 and p-55 fragments were cloned into plasmids

encoding the transcription-activation domain (pAD) and the DNA-binding domain

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

12

(pBD), generating plasmids pAD33, pBD33, pAD55 and pBD55 (Table 2). In order to

test whether interactions among VacA fragments could be detected in this system, we co-

transformed all four possible combinations of the p-33 and p-55 yeast-two hybrid

plasmids into the yeast reporter strain (Figure 2A). Serial dilutions of co-transformed

yeast were then tested for their ability to grow on SD medium lacking Leu and Trp, as

well as SD medium lacking Leu, Trp, and His. All of the co-transformed yeast grew on

the former media, which confirmed that the transformation had been successful and

provided an indication of the relative number of co-transformed yeast cells present in

each dilution. Yeast co-transformed with plasmids encoding p-33 and p-55 fusion

proteins grew in SD medium lacking Leu, Trp, and His, indicating that there was an

interaction between these fusion proteins (Figure 2B). In contrast, yeast co-transformed

with plasmids encoding p-33 / p-33 or p-55 / p-55 fusion proteins did not grow on SD

medium lacking Leu, Trp and His (Figure 2B). The p-33 / p-55 interaction observed

within co-transformed yeast was further confirmed by the analysis of -galactosidase

activity, which reflects the activation of the lacZ reporter gene (Figure 2C). The p-33 / p-

55 interaction was independent of the protein fused (AD or BD) to the p-33 or p-55

fragment, since both combinations of tested plasmids (pAD33 / pBD55 and pAD55 /

pBD33) were able to activate the HIS3 and lacZ reporters (Figure 2B and 2C).

Furthermore, none of the fusion-constructs were able to activate the reporter genes when

transformed alone into the yeast reporter strain (data not shown), indicating that the

interaction detected in the yeast-two hybrid system is due to a specific interaction

between p-33 and p-55 VacA fragments.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

13

p-33 / p-55 interactions of VacA mutant fragments. Previously, we have

reported the construction of H. pylori strains containing in-frame deletion mutations in

the vacA gene (39). Results from this previous study indicated that a mutant VacA

protein with a deletion of amino acids 6-27, VacA-( 6-27), could assemble into water-

soluble oligomeric structures, and suggested that VacA proteins containing 28-108,

56-83, 85-127, or 112-196 deletion mutations were defective in the capacity for

oligomer assembly (39) (Table 3). We hypothesized that the latter mutations might

disrupt interactions between p-33 and p-55 domains, and that such a defect might account

for failure of these VacA mutant proteins to form oligomeric structures. In order to

directly test this hypothesis, we introduced these same deletion mutations (i.e., 6-27,

28-108, 56-83, 85-127 and 112-196) into the pAD33 plasmid, as described in the

Experimental Procedures (Table 2). The pAD33 plasmid was chosen for mutagenesis

based on the observation that higher reporter activity was detected when wild-type p-33

was fused to the transcription-activation domain and wild-type p-55 fused to the DNA-

binding domain, compared to the opposite orientation (Figure 2C). Mutant pAD33

plasmids were co-transformed into yeast along with the wild-type pBD55 plasmid and the

activation of the reporter genes was analyzed as described above. Wild-type pAD33 and

pAD 6-27p-33 were each able to activate both the HIS3 and lacZ reporter genes when

co-transformed with wild-type pBD55 (Figure 3), whereas the other pAD33 mutants

failed to activate the reporter genes when co-transformed with wild-type pBD55. The p-

55 / 6-27p-33 interaction was specific, since yeast cells co-transformed with pBD33 and

pAD 6-27p-33 were not able to activate the reporter genes (data not shown). Thus,

mutations that disrupted oligomerization of VacA secreted by H. pylori (39) also

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

14

disrupted p-33 / p-55 interactions in the yeast-two hybrid system (Table 3). Interestingly,

yeast co-transformed with plasmids expressing 6-27p-33 and p-55 consistently

produced higher levels of -galactosidase activity than did yeast expressing wild-type p-

33 and p-55 (Figure 3B).

To further map putative amino acid sequences involved in the p-33 / p-55

interaction, we generated different p-55 fragments (i.e., encoding amino acids: 313-700,

313-550, 313-478, 550-821, and 479-821) and cloned them into the pBD vector as

described in Experimental Procedures (Table 2). These new fusion proteins were tested

for their ability to interact with wild-type p-33 as described above (Figure 4A). Yeast co-

transformed with the pAD33 plasmid and plasmids expressing p-55 fragments containing

amino acids 313-700, 313-550, or 313-478 were able to grow on SD-Leu,-Trp,-His,

whereas yeast co-transformed with the pAD33 plasmid and plasmids expressing p-55

fragments containing amino acids 479-821 or 550-821 were not able to grow (Figure 4B).

The former group of co-transformed yeast grew slowly on SD-Leu,-Trp,-His plates and

expressed a relatively low level of -galactosidase activity (Figure 4B and 4C). This

suggests that the interactions between the 313-700, 313-550 and 313-478 p-55 fragments

and p-33 are weaker than the interaction between full-length wild-type p-55 and p-33.

When we co-transformed yeast with a mutated version of the positive control plasmids

(Mutant + control described in Experimental Procedures) known to encode proteins that

interact with relatively low affinity (53), a similar low level of -galactosidase activity

was detected (Figure 4C). The data obtained with these mutant control plasmids supports

our conclusion that the p-33 interacts weakly with 313-700, 313-550, and 313-478 p-55

fragments.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

15

Proteolytic cleavage of H. pylori VacA into 33 kDa and 55 kDa fragments.

We sought to confirm the finding that p-33 interacts with p-55 using VacA purified from

H. pylori. Numerous previous studies have demonstrated that VacA proteins produced

by H. pylori can degrade into 33 kDa and 55 kDa fragments (11,27,28). However, it

remains unclear whether proteolytic cleavage of the 88 kDa VacA protein into 33 kDa

and 55 kDa fragments alters VacA cytotoxic activity or alters its oligomeric structure.

One reason why this issue has not yet been addressed in any detail is that it has been

difficult to experimentally induce the desired cleavage. VacA preparations commonly

undergo partial proteolytic degradation during storage (Figure 1), but complete

proteolysis of the 88 kDa VacA protein into smaller fragments occurs uncommonly.

Furthermore, spontaneous proteolytic degradation of VacA often involves cleavage at

multiple sites (27).

In order to test experimentally whether cleavage of VacA into 33 and 55 kDa

fragments alters its activity, we sought to develop a system in which a protease could be

used to induce proteolytic cleavage at a specific site located between the 33 and 55 kDa

domains. To accomplish this, we constructed a modified H. pylori strain in which an

oligonucleotide encoding the FLAG-tag epitope and an enterokinase cleavage site was

inserted into the vacA allele of H. pylori 60190 as described in Experimental Procedures

(Figure 5A). This enterokinase cleavage site (Figure 5B) provided a mechanism by

which the 88 kDa VacA protein could be specifically cleaved into fragments similar to

the spontaneously arising 33 kDa and 55 kDa VacA fragments. Immunoblotting studies

with an anti-VacA serum indicated that an H. pylori strain (VM088) containing this

engineered vacA allele expressed and secreted VacA at levels similar to wild-type H.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

16

pylori strain 60190 (data not shown). The FLAG-VacA toxin was purified from H. pylori

VM088 culture supernatant as large oligomeric structures which eluted from a gel

filtration column in the same fractions as wild-type VacA. Purified FLAG-VacA

underwent spontaneous partial degradation into two forms of the p-33 and p-55 VacA

fragments (Figure 6B, lane 1), which is consistent with previous observations suggesting

that proteolysis can occur at multiple sites (27). As expected, enterokinase treatment

resulted in the complete cleavage of the full-length FLAG-VacA into p-55 and FLAG-

tagged p-33 (FLAGp-33) fragments as seen by immunoblot analysis using an anti-VacA

serum or an anti-FLAG monoclonal antibody (Figure 6). In contrast, enterokinase

treatment of the wild-type 60190 VacA toxin (lacking the FLAG-tag epitope) did not

induce any detectable cleavage (data not shown).

Vacuolating toxic activity of intact FLAG-VacA and proteolytically cleaved

FLAG-VacA. We next compared the cytotoxic activities of intact FLAG-VacA and

FLAG-VacA that had been proteolytically cleaved into p-55 and FLAGp-33 fragments.

Purified oligomeric FLAG-VacA was either treated with enterokinase or mock treated as

described in Experimental Procedures. Proteins were then acid activated (a procedure

used to enhance vacuolating activity) and added to the neutral-pH medium overlying

HeLa cells (47,48). The vacuolating activities of intact FLAG-VacA and proteolytically

cleaved FLAG-VacA were indistinguishable as determined by microscopic examination

(data not shown) and analysis by a neutral-red uptake assay (Figure 7). The vacuolating

activities of full-length FLAG-VacA and proteolytically cleaved FLAG-VacA were not

detectably different when tested in a range of toxin concentrations from 2.5 to 10 g/ml

(Figure 7 and data not shown).

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

17

VacA purified from wild-type H. pylori strain 60190 exhibits very little

vacuolating toxic activity unless it is acid- or alkaline-activated prior to contact with cells

(47,48,54). We hypothesized that proteolytically cleaved VacA might exhibit

vacuolating toxic activity even in the absence of acid-activation. However, neither intact

nor proteolytically cleaved FLAG-VacA exhibited detectable vacuolating activity in the

absence of acid activation, as determined by microscopic examination (data not shown)

and analysis by a neutral-red uptake assay (Figure 7).

Oligomeric state of proteolytically cleaved FLAG-VacA. The secreted FLAG-

VacA assembles into large water-soluble oligomeric structures (Figure 8A), in a manner

similar to wild-type VacA (5,29). To determine whether p-33 and p-55 fragments of

VacA remained physically associated following proteolytic cleavage of FLAG-VacA,

enterokinase-treated FLAG-VacA was fractionated by gel filtration chromatography, and

its elution pattern monitored by immunoblot analysis with an anti-VacA serum.

Proteolytically cleaved FLAG-VacA and intact FLAG-VacA were found to elute in the

same high molecular mass fractions (Figure 8), but not in any of the lower molecular

mass fractions (data not shown), indicating that VacA fragments (i.e. FLAGp-33 and p-

55) remain associated within an oligomeric structure (Figure 8).

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

18

Discussion

Over the past decade, multiple lines of investigation have provided evidence for

the occurrence of homotypic interactions among VacA proteins (5,11,31,34,35,39,45).

The earliest evidence suggesting oligomerization of VacA proteins was the observation

that purified VacA from H. pylori broth culture supernatant assembles into structures of

~1,000 kDa in mass, corresponding to a buoyant density of about 22 S (5,29).

Ultrastructural studies have shown that these VacA complexes have a flower-like shape,

and that they contain anywhere from 6 to 14 subunits (29,30,42). These flower-shaped

structures disassemble into VacA monomers after exposure to acid or alkaline pH, and

then reassemble into oligomeric structures after neutralization (29,54). Further evidence

for the assembly of VacA into oligomeric structures has resulted from experiments in

which two different VacA toxins have been shown to form mixed-oligomeric structures

(34,39,45). VacA-VacA interactions have also been detected by transiently transfecting

cells with different fluorescently tagged VacA-expressing plasmids and using Fluorescent

Resonance Energy Transfer (FRET) methodology (35).

Although many lines of evidence indicate that VacA can form oligomeric

structures, thus far there has been relatively little investigation of the specific VacA-

VacA interactions that are required for assembly into higher order structures. In this

study we investigated the role of p-33 and p-55 interactions in oligomer formation by

VacA. These two VacA fragments have been considered to represent two domains or

subunits of VacA. In the present work, we demonstrated that p-33 / p-55 interactions can

be detected using the yeast-two hybrid system. In contrast, p-33 / p-33 and p-55 / p-55

interactions were not detected in this system. The lack of detectable p-33 / p-33 and p-55

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

19

/ p-55 interactions was not due to the lack of expression of these fusion proteins in the

yeast, since all four fusion proteins were able to interact when tested for p-33 / p-55

interactions. Importantly, the failure to detect p-33 / p-33 or p-55 / p-55 interactions in

the yeast-two hybrid system does not exclude the possibility that such interactions might

occur in other environments. A previous study analyzed the properties of a modified p-

55 fragment secreted by H. pylori, and reported that this protein could form homodimers

but not large oligomeric structures (31). However, the VacA protein analyzed in that

study contained at least 27 residues of the p-33 domain in addition to the p-55 fragment

(31). Another previous study provided evidence that an amino-terminal hydrophobic

portion of the p-33 fragment could undergo transmembrane protein homodimerization

within E.coli membranes (40,55). The inability to detect p-33 / p-33 interactions with the

yeast-two hybrid system is consistent with a model in which p-33 / p-33 interactions

occur only within membranes (40,55).

In an effort to map putative interacting regions within p-33 and p-55, we used the

yeast-two hybrid system to analyze a series of mutant proteins containing internal

deletions, as well as truncated VacA proteins. The yeast-two hybrid data suggest that

amino acids 28-196 within the p-33 domain and 313-478 within the p-55 domain

contribute to the p-33 / p-55 interaction. Only one of the p-33 deletion mutants ( 6-27p-

33) was able to interact with wild-type p-55. VacA amino acids 6-27 comprise a

hydrophobic region of VacA that inserts into lipid membranes and is required for VacA

cytotoxic and membrane anion-channel activity (39,40,55). Interestingly, when co-

expressed with p-55, the 6-27p-33 mutant fragment induced higher reporter activity in

the yeast-two hybrid assay than did wild-type p-33. A possible explanation for this result

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

20

could be that the 6-27p-33 mutant fragment is able to translocate to the yeast nuclei

more efficiently than can the wild-type p-33, since previously, it has been reported that

hydrophobic sequences decrease nuclear protein translocation (56). Another possibility

could be that the 6-27p-33 mutant fragment interacts more strongly with the wild-p55

fragment than does wild-type p-33. This latter hypothesis is consistent with data from a

cell transfection system, in which cells co-transfected with plasmids expressing wild-type

full-length VacA and VacA-( 6-26) produced a stronger FRET signal than did cells co-

transfected with plamids expressing two wild-type VacA proteins (35).

By constructing a FLAG-VacA toxin, which can be efficiently cleaved into p-33

and p-55 domains, we were able to study the p-33 / p-55 interactions that occur within the

VacA oligomer produced by H. pylori. Our results (Figure 8) clearly indicate that p-33

and p-55 fragments remain physically associated after proteolytic cleavage of VacA. In

agreement with this interpretation, when we incubated proteolytically cleaved FLAG-

VacA with an ANTI-FLAG M2 Affinity Gel (Sigma), both the p-55 and FLAGp-33

fragments were captured (data not shown). Neither FLAGp-33 nor p-55 fragments could

be eluted from the affinity gel under pH conditions ranging from 7.5 to 4.5, nor with 1%

CHAPS (3-[(3-Cholamidopropyl)-dimethylammonio]-1-propanesulfonate), 10% Triton

X-100 or 2 M urea, suggesting that there is a stable physical interaction between the p-33

and p-55 VacA domains (data not shown). Furthermore, our results demonstrate that the

vacuolating activity of intact FLAG-VacA and fully proteolytically cleaved FLAG-VacA

is indistinguishable (Figure 7). A likely explanation for intact activity of proteolytically

cleaved VacA is that the fragments remain associated and do not undergo any extensive

adverse conformational changes (unfolding) following proteolytic cleavage.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

21

Two lines of evidence presented in the current study suggest that the p-33 / p-55

interactions play an important role in VacA oligomer assembly. First, the intact

oligomeric structure of proteolytically-cleaved VacA indicates that p-33 / p-55

interactions occur within VacA oligomers. Second, the properties of mutant p-33

fragments in the yeast-two hybrid system correlate perfectly with the capacity of secreted

H. pylori mutant proteins to form oligomeric structures. This suggests that mutations

disrupting the p-33 / p-55 interaction also disrupt the formation of large VacA oligomeric

structures. It is possible that there may be two different types of interactions: (i)

intramolecular interactions between the p-33 and p-55 domains of an individual VacA

monomer that might be important for proper folding of the VacA monomers within the

oligomeric structures and (ii) intermolecular interaction between p-33 and p-55 domains

of different 88 kDa molecules to form the larger oligomeric structures. Currently, we are

not able to differentiate between these two interactions.

VacA toxins that fail to oligomerize consistently lack vacuolating cytotoxic

activity, suggesting that oligomerization is important for VacA activity (39). This

hypothesis is supported by the observation that neither p-33 nor p-55 are able to exhibit

vacuolating cytotoxic activity in a transient transfection assay when expressed

individually within cells (32,34,41). In contrast, when both domains are co-expressed

within cells, they interact and exhibit vacuolating cytotoxic activity (32,34). Thus, the

data presented in the current study, as well as data obtained using other systems, indicate

that p-33 / p-55 interactions are essential for VacA assembly into oligomeric structures

and also for vacuolating cytotoxic activity.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

22

Oligomerization is an important feature of many proteins, in particular pore-

forming proteins (57,58). Examples of these proteins include perforin, Bcl-2 family

members, as well as several bacterial toxins. Several bacterial pore-forming toxins,

including protective antigen (PA) of Bacillus anthracis, -hemolysin of Staphylococcus

aureus, streptolysin O of Streptococcus pyogenes, -toxin of Clostridium septicum, El

Tor cytolysin of Vibrio cholera, and aerolysin of Aeromonas hydrophila, assemble in

target cell membranes forming transmembrane channels (59-65). These proteins are

interesting because, like VacA, they are synthesized as water-soluble molecules but then

insert into membranes. Further studies of the processes by which these toxins assemble

into oligomeric structures may ultimately lead to the development of therapeutic

inhibitors that block the oligomerization and the activity of these toxins in vivo.

Acknowledgments-We thank Ping Cao, Beverly Hosse, Arlene D. Vinion-Dubiel,

Leslie Morrison, and Mark Hansberger for technical assistance and helpful discussions.

DNA oligonucleotides were synthesized by the Vanderbilt University DNA Chemistry

Core Facility, and DNA sequence analysis was performed by the Vanderbilt University

DNA Sequencing Laboratory.

This work was supported by NIH grants AI39657 and DK53623 and by the

Medical Research Department of the Department of Veterans Affairs (T. Cover). V.

Torres was supported by a structural biology predoctoral fellowship from the Vanderbilt

University Center for Structural Biology. The Vanderbilt University DNA Sequencing

Laboratory is supported by the Vanderbilt-Ingram Cancer Center.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

23

References

1. Warren, J. R., and Marshall, B. J. (1983) Lancet 1, 1273-1275

2. Cover, T. L., Berg, D. E., Blaser, M. J., and Mobley, H. L. T. (2001) in Principles

of bacterial pathogenesis (Groisman, E. A., ed), pp. 510-558, Academic Press,

San Diego

3. Dunn, B. E., Cohen, H., and Blaser, M. J. (1997) Clin Microbiol Rev 10, 720-741

4. Suerbaum, S., and Michetti, P. (2002) N Engl J Med 347, 1175-1186

5. Cover, T. L., and Blaser, M. J. (1992) J Biol Chem 267, 10570-10575

6. Montecucco, C., Papini, E., de Bernard, M., Telford, J. L., and Rappuoli, R.

(1999) in The comprehensive sourcebook of bacterial protein toxins (Alouf, J. E.,

and Freer, J. H., eds), pp. 264-286, Academic Press, London ; San Diego, Calif.

7. Atherton, J. C., Cover, T. L., Papini, E., and Telford, J. L. (2001) in Helicobacter

pylori: Physiology and Genetics (Mobley, H. L. T., Mendz, G. L., and Hazell, S.

L., eds), pp. 97-110, ASM Press, Washington, D.C.

8. Papini, E., Zoratti, M., and Cover, T. L. (2001) Toxicon 39, 1757-1767

9. Figueiredo, C., Machado, J. C., Pharoah, P., Seruca, R., Sousa, S., Carvalho, R.,

Capelinha, A. F., Quint, W., Caldas, C., van Doorn, L. J., Carneiro, F., and

Sobrinho-Simoes, M. (2002) J Natl Cancer Inst 94, 1680-1687

10. Atherton, J. C., Cao, P., Peek, R. M., Jr., Tummuru, M. K., Blaser, M. J., and

Cover, T. L. (1995) J Biol Chem 270, 17771-17777

11. Telford, J. L., Ghiara, P., Dell'Orco, M., Comanducci, M., Burroni, D., Bugnoli,

M., Tecce, M. F., Censini, S., Covacci, A., Xiang, Z., and et al. (1994) J Exp Med

179, 1653-1658

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

24

12. Fujikawa, A., Shirasaka, D., Yamamoto, S., Ota, H., Yahiro, K., Fukada, M.,

Shintani, T., Wada, A., Aoyama, N., Hirayama, T., Fukamachi, H., and Noda, M.

(2003) Nat Genet 33, 375-381

13. Leunk, R. D., P.T., J., David, B. C., Kraft, W. G., and Morgan, D. R. (1988) J.

Med. Microbiol. 26, 93-99

14. Molinari, M., Galli, C., Norais, N., Telford, J. L., Rappuoli, R., Luzio, J. P., and

Montecucco, C. (1997) J Biol Chem 272, 25339-25344

15. Czajkowsky, D. M., Iwamoto, H., Cover, T. L., and Shao, Z. (1999) Proc Natl

Acad Sci U S A 96, 2001-2006

16. Szabo, I., Brutsche, S., Tombola, F., Moschioni, M., Satin, B., Telford, J. L.,

Rappuoli, R., Montecucco, C., Papini, E., and Zoratti, M. (1999) Embo J 18,

5517-5527

17. Schraw, W., Li, Y., McClain, M. S., van der Goot, F. G., and Cover, T. L. (2002)

J Biol Chem 277, 34642-34650

18. Papini, E., Satin, B., Norais, N., de Bernard, M., Telford, J. L., Rappuoli, R., and

Montecucco, C. (1998) J Clin Invest 102, 813-820

19. Peek, R. M., Jr., Blaser, M. J., Mays, D. J., Forsyth, M. H., Cover, T. L., Song, S.

Y., Krishna, U., and Pietenpol, J. A. (1999) Cancer Res 59, 6124-6131

20. Galmiche, A., Rassow, J., Doye, A., Cagnol, S., Chambard, J. C., Contamin, S.,

de Thillot, V., Just, I., Ricci, V., Solcia, E., Van Obberghen, E., and Boquet, P.

(2000) Embo J 19, 6361-6370

21. Kuck, D., Kolmerer, B., Iking-Konert, C., Krammer, P. H., Stremmel, W., and

Rudi, J. (2001) Infect Immun 69, 5080-5087

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

25

22. Cover, T. L., Krishna, U. S., Israel, D. A., and Peek, R. M., Jr. (2003) Cancer Res

63, 951-957

23. Molinari, M., Salio, M., Galli, C., Norais, N., Rappuoli, R., Lanzavecchia, A., and

Montecucco, C. (1998) J Exp Med 187, 135-140

24. Gebert, B., Fischer, W., Weiss, E., Hoffmann, R., and Haas, R. (2003) Science

301, 1099-1102

25. Cover, T. L., Tummuru, M. K. R., Cao, P., Thompson, S. A., and Blaser, M. J.

(1994) J. Biol. Chem. 269, 10566-10573

26. Schmitt, W., and Haas, R. (1994) Mol Microbiol 12, 307-319

27. Nguyen, V. Q., Caprioli, R. M., and Cover, T. L. (2001) Infection and Immunity

69, 543-546

28. Garner, J. A., and Cover, T. L. (1996) Infect Immun 64, 4197-4203

29. Cover, T. L., Hanson, P. I., and Heuser, J. E. (1997) J Cell Biol 138, 759-769

30. Lupetti, P., Heuser, J. E., Manetti, R., Massari, P., Lanzavecchia, S., Bellon, P. L.,

Dallai, R., Rappuoli, R., and Telford, J. L. (1996) J Cell Biol 133, 801-807

31. Reyrat, J. M., Lanzavecchia, S., Lupetti, P., de Bernard, M., Pagliaccia, C.,

Pelicic, V., Charrel, M., Ulivieri, C., Norais, N., Ji, X., Cabiaux, V., Papini, E.,

Rappuoli, R., and Telford, J. L. (1999) J Mol Biol 290, 459-470

32. Ye, D., Willhite, D. C., and Blanke, S. R. (1999) J Biol Chem 274, 9277-9282

33. Ye, D., and Blanke, S. R. (2000) Infect Immun 68, 4354-4357

34. Ye, D., and Blanke, S. R. (2002) Mol Microbiol 43, 1243-1253

35. Willhite, D. C., Ye, D., and Blanke, S. R. (2002) Infect Immun 70, 3824-3832

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

26

36. de Bernard, M., Burroni, D., Papini, E., Rappuoli, R., Telford, J., and

Montecucco, C. (1998) Infect Immun 66, 6014-6016

37. Wang, H. J., Chang, P. C., Kuo, C. H., Tzeng, C. S., and Wang, W. C. (1998)

Biochem Biophys Res Commun 250, 397-402

38. Wang, H. J., and Wang, W. C. (2000) Biochem Biophys Res Commun 278, 449-

454

39. Vinion-Dubiel, A. D., McClain, M. S., Czajkowsky, D. M., Iwamoto, H., Ye, D.,

Cao, P., Schraw, W., Szabo, G., Blanke, S. R., Shao, Z., and Cover, T. L. (1999) J

Biol Chem 274, 37736-37742

40. McClain, M. S., Iwamoto, H., Cao, P., Vinion-Dubiel, A. D., Li, Y., Szabo, G.,

Shao, Z., and Cover, T. L. (2003) J Biol Chem 278, 12101-12108

41. de Bernard, M., Arico, B., Papini, E., Rizzuto, R., Grandi, G., Rappuoli, R., and

Montecucco, C. (1997) Mol Microbiol 26, 665-674

42. Adrian, M., Cover, T. L., Dubochet, J., and Heuser, J. E. (2002) J Mol Biol 318,

121-133

43. Iwamoto, H., Czajkowsky, D. M., Cover, T. L., Szabo, G., and Shao, Z. (1999)

FEBS Lett 450, 101-104

44. Woods, R. A., and Gietz, R. D. (2001) Methods Mol Biol 177, 85-97

45. McClain, M. S., Cao, P., Iwamoto, H., Vinion-Dubiel, A. D., Szabo, G., Shao, Z.,

and Cover, T. L. (2001) J Bacteriol 183, 6499-6508

46. Copass, M., Grandi, G., and Rappuoli, R. (1997) Infect Immun 65, 1949-1952

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

27

47. de Bernard, M., Papini, E., de Filippis, V., Gottardi, E., Telford, J., Manetti, R.,

Fontana, A., Rappuoli, R., and Montecucco, C. (1995) J Biol Chem 270, 23937-

23940

48. McClain, M. S., Schraw, W., Ricci, V., Boquet, P., and Cover, T. L. (2000) Mol

Microbiol 37, 433-442

49. Cover, T. L., Puryear, W., Pérez-Pérez, G. I., and Blaser, M. J. (1991) Infect.

Immun. 59, 1264-1270

50. Fields, S., and Song, O. (1989) Nature 340, 245-246

51. James, P. (2001) Methods Mol Biol 177, 41-84

52. Ma, J., and Ptashne, M. (1987) Cell 48, 847-853

53. Gimble, F. S., and Sauer, R. T. (1989) J Mol Biol 206, 29-39

54. Yahiro, K., Niidome, T., Kimura, M., Hatakeyama, T., Aoyagi, H., Kurazono, H.,

Imagawa, K., Wada, A., Moss, J., and Hirayama, T. (1999) J Biol Chem 274,

36693-36699

55. McClain, M. S., Cao, P., and Cover, T. L. (2001) Infect Immun 69, 1181-1184.

56. Boulikas, T. (1993) Crit Rev Eukaryot Gene Expr 3, 193-227

57. Nooren, I. M., and Thornton, J. M. (2003) Embo J 22, 3486-3492

58. Gouaux, E. (1997) Curr Opin Struct Biol 7, 566-573

59. Lacy, D. B., and Collier, R. J. (2002) Curr Top Microbiol Immunol 271, 61-85

60. Bhakdi, S., Bayley, H., Valeva, A., Walev, I., Walker, B., Kehoe, M., and Palmer,

M. (1996) Arch Microbiol 165, 73-79

61. Parker, M. W., Buckley, J. T., Postma, J. P., Tucker, A. D., Leonard, K., Pattus,

F., and Tsernoglou, D. (1994) Nature 367, 292-295.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

28

62. Ramachandran, R., Heuck, A. P., Tweten, R. K., and Johnson, A. E. (2002) Nat

Struct Biol 9, 823-827

63. Moschioni, M., Tombola, F., de Bernard, M., Coelho, A., Zitzer, A., Zoratti, M.,

and Montecucco, C. (2002) Cell Microbiol 4, 397-409

64. van der Goot, F. G., Pattus, F., Wong, K. R., and Buckley, J. T. (1993)

Biochemistry 32, 2636-2642.

65. Fivaz, M., Abrami, L., Tsitrin, Y., and van der Goot, F. G. (2001) Toxicon 39,

1637-1645

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

29

Figure Legends

Figure 1. Proteolytic degradation of VacA into 33 and 55 kDa fragments.

Proteins precipitated from broth culture supernatant of H. pylori 60190 by a 50 %

saturated solution of ammonium sulfate (lanes A and B) or purified VacA (lanes C and

D) from H. pylori 60190 were electrophoresed on a 10% SDS-polyacrylamide gel,

transferred to a nitrocellulose membrane and immunobloted with polyclonal anti-VacA

serum. These preparations exhibit varying degrees of spontaneous VacA proteolysis into

p-33 and p-55 fragments.

Figure 2. Interaction of VacA fragments in the yeast-two hybrid system. (A)

Representation of the different p-33 and p-55 fusion proteins and plasmid combinations

used in the co-transformation experiments. (B) Yeast cells were co-transformed with the

plasmids depicted in panel A, or with controls plasmids described in Experimental

Procedures. Co-transformed yeast-cells (normalized based on OD) were 10-fold serially

diluted and identical inocula were plated on two types of SD agar plates. All of the co-

transformed yeast grew on SD-Leu,-Trp plates. Protein-protein interactions were

assessed by growth of the co-transformed yeast cells on SD-Leu,-Trp,-His plates. (C) -

galactosidase liquid culture assay of the co-transformed yeast-cells. Activation of the

protein-protein interaction reporters were only detected when yeast were co-transformed

with plasmids encoding wild-type p-33 together with a plasmid encoding wild-type p-55.

Results represent the mean + S.D. from triplicate determinations, each representing a

separate colony. * = P (<0.05) when compared to the negative control. + cntrl = positive

control, - cntrl = negative control.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

30

Figure 3. Interaction between wild-type p-55 and p-33 mutant VacA

fragments. (A) Co-transformed yeast-cells (normalized based on OD) were 10-fold

serially diluted and identical inocula were plated on two types of SD plates. Protein-

protein interactions were assessed by growth of the co-transformed yeast cells on SD-

Leu,-Trp, is plates -galactosidase liquid culture assay of the co-transformed

yeast-cells. Activation of the protein-protein interaction reporters were only detected

when yeast were co-transformed with plasmids encoding wild-type p-33 or 6-27p-33

together with a plasmid encoding wild-type p-55. Results represent the mean + S.D.

from triplicate colonies. * = P (<0.05) when compared to the negative control.

Figure 4. Interaction between wild-type p-33 and fragments of p-55. (A)

Representation of the different p-55 fragments used in the co-transformation experiment.

(B) Yeast-cells co-transformed with the pBD55 fragments depicted in panel A and wild-

type pAD33 (normalized based on OD) were 10-fold serially diluted and identical inocula

were plated on two types of SD agar plates. C -galactosidase liquid culture assay of

the co-transformed yeast-cells. Protein-protein interactions were assessed by growth of

the co-transformed yeast cells on SD-Leu,-Trp,-His plates and by a -galactosidase liquid

culture assay. Sample numbers in panel C are identical to those in panel B. Wild-type p-

33 was able to interact with p-55 fragments containing amino-acids 313-700, 313-550

and 313-478 but not with amino acids 550-821 or 479-821. -galactosidase liquid culture

assay results represent the mean + S.D. from triplicate colonies. * = P (<0.05) when

compared to the negative control.

Figure 5. Construction of H. pylori VM088 using a sacB-based counter-

selection approach. (A) The pMM389 plasmid contains a vacA fragment from H. pylori

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

31

strain 60190 and a sacB / kan cassette (39,46). Transformation of H. pylori 60190 with

pMM389 yielded a kanamycin-resistant transformant designated H. pylori VM022 (39).

H. pylori VM022 was transformed with pA178-FLAG plasmid, transformants selected on

sucrose-containing medium, and a strain (H. pylori VM088) with a sequence encoding a

FLAG-tag with an enterokinase site inserted into the vacA gene was thereby obtained.

(B) VacA spontaneously degrades into two fragments, termed p-33 and p-55. Open

arrows indicate the most common site at which wild-type VacA from H. pylori 60190

undergoes spontaneous cleavage and the resulting p-33 and p-55 fragments (27). Closed

arrows indicate the enterokinase cleavage site introduced with the FLAG-tag epitope and

the FLAGp-33 and p-55 fragments generated after enterokinase cleavage. The amino

acid numbering system used in this figure is based on designating the first amino acid

(alanine) of the mature secreted VacA toxin of strain 60190 as amino acid 1.

Figure 6. Enterokinase treatment of full-length FLAG-VacA results in

production of 33 and 55 kDa VacA fragments. Purified VacA from H. pylori VM088

(10 g) was either treated with enterokinase (1 Unit) or mock treated as described in

Experimental Procedures. Toxins (100 ng) were then electrophoresed on a 10% SDS-

polyacrylamide gel, transferred to a nitrocellulose membrane and reacted with anti-FLAG

M2 monoclonal antibody (Sigma) (A). The blot was then stripped and reacted with

polyclonal anti-VacA serum (B). Enterokinase treatment induced complete cleavage of

full-length FLAG-VacA into ~33 and ~55 kDa fragments. In preparations of FLAG-

VacA not treated with enterokinase, the FLAG-tag epitope is detected predominantly in

full-length 88 kDa VacA and in the p-55 fragment (Figure 6A lane 1). Only the ~33 kDa

fragment retained the FLAG-tag epitope after enterokinase treatment (Figure 6A lane 2).

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

32

Figure 7. Vacuolating cytotoxic activity of intact and proteolytically cleaved

FLAG-VacA. Identical aliquots of purified FLAG-VacA were either treated with

enterokinase as described in Experimental Procedures or left untreated. Toxins were then

either acid activated (pH 3) or left untreated and added to tissue culture medium

overlying HeLa cells. The final VacA concentration of all samples was 5 g/ml.

Vacuolating activity was quantified using a neutral red uptake assay. Enterokinase-

treated FLAG-VacA and untreated FLAG-VacA did not differ significantly in

vacuolating cytotoxin activity. Furthermore, both enterokinase-treated and untreated

FLAG-VacA required acid activation in order to induce vacuolating activity. Results

represent the mean + S.D. from triplicate samples and are expressed as a percent of

neutral red uptake induced by full-length acid-activated FLAG-VacA.

Figure 8. Oligomeric state of full-length FLAG-VacA and cleaved FLAG-

VacA. Concentrated broth culture supernatant from Helicobacter pylori strain VM088

(A) and 40 g of purified FLAG-VacA treated with enterokinase as described in

Experimental Procedures (B), were applied to a Superose 6 HR FPLC column. Forty

fractions of 2 ml each were collected, with fraction 1 corresponding to the void volume.

Samples (40 l) were then electrophoresed on a 10% acrylamide gel, transferred to a

nitrocellulose membrane and immunoblotted with anti-VacA serum. Full-length VacA as

well as enterokinase-cleaved VacA eluted in the same fractions with an elution peak at

fraction 13, corresponding to a molecular mass of about 1,000 kDa (5,29). VacA was not

detected in any of the lower molecular mass fractions (data not shown). p-55 fragments

are detected more prominently than p-33 fragments in panel B due to the relatively

weaker reactivity of the anti-VacA serum against the p-33 fragment. When the

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

33

membrane shown in panel B was stripped and immunoblotted with polyclonal anti-FLAG

antibody, the p-33 band was detected in fractions 8-16 (data not shown).

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

34

Table 1 Oligonucleotides Used in This Study

Primer Name Nucleotide Sequence AND510a 5’-TATAGCCCGGGCCGCCTTTTTTACAACCGTGATAND511 5’-CCCCTCGACTTATTTAGCACCACTTTGAGAAG AND512 5’-CCCCGGATCCGCCTTTTTT ACAACCGTGAT AND515a 5’-CCCCGTCGACTTAAGCGTAGCTAGCGAAACGCGAND516 5’-CCCCGGATCCAACGAC AAACAAGAGAGCAG AND6098 5’-CCCCGAATTCAACGACAAACAAGAGAGCAG OP4175 5’-CCCCGGATCCGCCTTTTTTACAACCCTTGG OP4176 5’-CCCCGTCGACTTACGTATCAATACCTTTAAAAT BAR1556 5’-CCCCGAATTCGGTAATGGTGGTTTCAACAC BAR1557 5’-CCCCGAATTCACTAGGTCAATCTTTTCTGG BAR1558 5’-CCCCGTCGACTTAGCCAGTTTCCAAACGCACG BAR1559 5’-CCCCGTCGACTTAGATTTTCGCTTTCAATAAAACA OP2891 5’-CCGACTACAAGGATGACGA CGACAAAG OP2892 5’-CTTTGTCGTCGTCATCCTTGTAGTCGG

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

35

Table 2 Yeast-two hybrid vectors containing vacA fragments

Y2H

Plasmidsa

Forward

Primersb

Reverse

Primersb

Restriction

Enzymesc

VacA

Amino acidsd

pBD33 AND510a AND511 SrfI / SalI 1-312 pAD33 AND512 AND511 BamHI / SalI 1-312 pBD55 AND6098 AND515a EcoRI / SalI 313-821 pAD55 AND516 AND515a BamHI / SalI 313-821 pAD 6-27 OP4175 AND511 BamHI / SalI 1-312 ( 6-27)pAD 28-108 AND512 AND511 BamHI / SalI 1-312 ( 28-108)pAD 56-83 AND512 AND511 BamHI / SalI 1-312 ( 56-83)pAD 85-127 AND512 AND511 BamHI / SalI 1-312 ( 85-127)pAD 112-196 AND512 AND511 BamHI / SalI 1-312 ( 112-196)pBD313-700 AND6098 BAR1559 EcoRI / SalI 313-700 pBD313-550 AND6098 BAR1558 EcoRI / SalI 313-550 pBD313-478 AND6098 OP4176 EcoRI / SalI 313-478 pBD550-821 BAR1557 AND515a EcoRI / SalI 550-821 pBD479-821 BAR1556 AND515a EcoRI / SalI 479-821

a Name of the yeast-two hybrid (Y2H) plasmids expressing VacA fragments.

b Oligonucleotides used to PCR-amplify the different vacA sequences. Oligonucleotide

sequences are described in Table 1. c Restriction sites used to clone the PCR products into the pAD and pBD yeast-two

hybrid plamids. d The VacA amino acid numbering system used in this table is based on designating the

first amino acid (alanine) of the mature secreted VacA toxin of strain 60190 (25) as amino acid 1. indicates amino acids that are deleted in the mutant VacA fragments.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

H. pylori VacA

36

Table 3 Characterization of mutant VacA toxins

p-33a

p-55b

Y2H

Interactionc

Oligomer

Formationd

Cytotoxic

Activitye

WT WT + + + 6-27 WT + + - 28-108 WT - - - 56-83 WT - - - 85-127 WT - - - 112-196 WT - - -

a The p-33 VacA domain contains amino acids 1-312.

b The p-55 VacA domain contains amino acids 313-821.

c Interaction between p-33 and p-55 VacA fragments was determined

using the yeast-two hybrid system (Y2H) as described in the text. d Oligomer formation of full-length VacA toxins secreted by H. pylori

was analyzed by determining whether these proteins eluted as large oligomeric structures (~1,000 kDa) from a Superose 6 HR 16/50 FPLC column (5,39).

eCytotoxic vacuolating activity of VacA toxins for HeLa cells was

analyzed as described in Experimental Procedures.

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Victor J. Torres, Mark S. McClain and Timothy L. Covercytotoxin (VacA)

Interactions between p-33 and p-55 domains of the helicobacter pylori vacuolating

published online October 30, 2003J. Biol. Chem. 

  10.1074/jbc.M310159200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on March 3, 2018

http://ww

w.jbc.org/

Dow

nloaded from