Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Clin Microbiol Infect. Author manuscript; available in PMC 2010 April 1.
Published in final edited form as:
PMCID: PMC2837774

CagA C-terminal variations in Helicobacter pylori strains from Colombian patients with gastric precancerous lesions


The C-terminus of the Helicobacter pylori CagA protein is polymorphic, bearing different EPIYA sequences (EPIYA-A, B, C or D), and one or more CagA multimerization (CM) motifs. The number of EPIYA-C motifs is associated with precancerous lesions and gastric cancer (GC). The relationship between EPIYA, CM motifs and gastric lesions was examined in H. pylori-infected Colombian patients from areas of high and low risk for GC. Genomic DNA was extracted from H. pylori strains cultured from gastric biopsies from 80 adults with dyspeptic symptoms. Sixty-seven (83.8%) of 80 strains were cagA positive. The 3′ region of cagA was sequenced, and EPIYA and CM motifs were identified. CagA proteins contained one (64.2%), two (34.3%) or three EPIYA-C motifs (1.5%), all with Western type CagA-specific sequences. Strains with one EPIYA-C motif were associated with less severe gastric lesions (non-atrophic and multifocal atrophic gastritis), whereas strains with multiple EPIYA-C motifs were associated with more severe lesions (intestinal metaplasia and dysplasia) (p <0.001). In 54 strains, the CM motifs were identical to those common in Western strains. Thirteen strains from the low-risk area contained two different CM motifs: one of Western type located within the EPIYA-C segment and another following the EPIYA-C segment and resembling the CM motif found in East Asian strains. These strains induced significantly shorter projections in AGS cells and an attenuated reduction in levels of CagA upon immunodepletion of SHP-2 than strains possessing Western/Western motifs. This novel finding may partially explain the difference in GC incidence in these populations.

Keywords: CagA, cagA 3′region, CagA multimerization motif, EPIYA, ESS, H. pylori, WSS


Helicobacter pylori is a spiral, Gram-negative microaerophilic pathogenic bacterium that chronically colonizes the gastric epithelium of approximately one-half of the world’s population. In most infected subjects, this bacterium survives for decades without clinically apparent consequences for the host, especially in affluent populations with low risk for gastric cancer (GC). However, in a fraction of infected subjects, gastric inflammation induced by H. pylori can progress to peptic ulceration, mucosa-associated lymphoid tissue lymphoma, or gastric adenocarcinoma [1,2]. An H. pylori virulence factor that is associated with more severe disease is the cagA gene [1].

cagA encodes a 120- to 145-kDa CagA protein, and is a marker for the cag pathogenicity island [35]. Many molecular epidemiological studies have shown that cagA-positive H. pylori strains are associated with an increased risk of GC [1,6,7]. cagA-positive H. pylori strains are associated with the progression of inflammatory changes in the gastric mucosa that may lead to GC: severe non-atrophic gastritis, atrophic gastritis, intestinal metaplasia, and dysplasia [8]. Following bacterial attachment and injection of CagA into the epithelial cell by the type IV secretion system [9,10], the protein localizes to the inner surface of the plasma membrane and under-goes tyrosine phosphorylation at Y-972, Y-899 and Y-912, which is mediated by Ab1 and Src family kinases [1114]. Phosphorylation of CagA at Y-972 has a function in actinbased cytoskeletal rearrangements [11]. Once phosphorylated, CagA binds to a cytoplasmic Src Homology 2 (SH2) domain of Src Homology 2 phosphatase (SHP-2). Because CagA–SHP-2 complexes disrupt signal transduction pathways of the cell, the complexes may be involved in the development of atrophic gastritis and the transition from atrophy to intestinal metaplasia [15].

CagA proteins isolated from different H. pylori strains exhibit sequence polymorphisms and duplications, especially in their C-terminal regions containing EPIYA (glutamic acid-proline-isoleucine-tyrosine-alanine) motifs [4]. Based upon sequences surrounding those motifs, the segments are designated as EPIYA-A, EPIYA-B, EPIYA-C or EPIYA-D [16]. EPIYA-A and EPIYA-B motifs are present in the CagA proteins of almost all isolates of H. pylori of Western and East Asian origin. The EPIYA-C motif, typically present in one to three copies, is characteristic of CagA proteins from strains isolated in European countries, the Americas and Australia. The EPIYA-C segments also contain a sequence designated as Western CagA-specific sequence (WSS) [17]. CagA proteins carried by most Western H. pylori isolates have a single EPIYA-C segment, and thus are classified as ABC types [16,18]. However, some CagA proteins contain two or three EPIYA-C motifs, and are termed ABCC or ABCCC types. CagA proteins from strains isolated in East Asian countries do not possess a WSS, but instead possess an East Asian CagA-specific sequence (ESS) located within an EPIYA-D segment [1618]. The EPIYA-C and EPIYA-D motifs contain the major tyrosine phosphorylation sites of CagA. Among Western CagA species, the number of EPIYA-C sites is directly related to levels of CagA tyrosine phosphorylation, SHP-2 binding activity, and cytoskeletal alterations known as the ‘hummingbird phenotype’ [16,19]. This finding suggests that Western CagA proteins possessing a greater number of EPIYA-C motifs are more virulent and thus more carcinogenic than those having a smaller number of EPIYA-C motifs [13]. A recent study of South African subjects revealed that five of six Western-type H. pylori strains from gastric carcinoma patients possessed multiple EPIYA-C sites, compared with only one of 19 strains from non-cancer patients [20]. Yamaoka et al. found a significant association between the EPIYA-C motifs in triplicate and gastric atrophy and intestinal metaplasia among H. pylori isolates from Colombian patients [18].

A prerequisite for CagA–SHP-2 interaction is CagA multimerization, which is mediated by a conserved sequence of 16 amino acids (FPLXRXXXVXDLSKVG) recently identified and designated as the CagA multimerization (CM) motif [21]. Eleven of the 16 amino acids in the CM motif are well conserved between Western and East Asian CagA species. Western CagA strains possess multiple CM motifs, located within each EPIYA-C segment, plus one CM motif distal to the last EPIYA-C, whereas East Asian CagA strains possess a single CM motif, located distal to the EPIYA-D segment. The type and number of CM motifs may influence the potential of individual CagA proteins to multimerize in host cells, and this may affect the ability of CagA to disturb host cell function via SHP-2 deregulation [21].

Previously we examined H. pylori strains isolated from subjects living in two areas of Southwestern Colombia, which differ markedly in GC incidence. Residents of Andean regions such as Tuquerres are of mixed Native American and Spanish ancestry, and are mainly farmers. This population has a high incidence of GC. In contrast, residents of the Pacific coast are of mixed African and Spanish ancestry, and have a low risk of GC; their economy is based primarily on fishing. We found that infection with H. pylori occurs at an early age in both populations [22], but that the Andean residents have a higher prevalence of more virulent (cagA+, vacA s1m1) strains [23,24]. We asked whether these two populations might also harbour H. pylori strains that differed in cagA 3′ end sequences encoding EPIYA or CM motifs and whether those differences would be associated with the severity of the gastric precancerous lesions.

Materials and Methods


Male subjects, 39–60 years old, with dyspeptic symptoms, who underwent upper gastrointestinal tract endoscopy, were recruited in 2006 from the two areas with contrasting GC risks in the State of Nariño, Colombia. According to our previous observations in these populations, we expected men of this age group to have high prevalence of H. pylori infection and of preneoplastic lesions. Exclusion criteria were previous gastrectomy, serious chronic diseases, or ingestion of H2-receptor antagonists, proton pump inhibitors, or antimicrobials in the 4 weeks prior to the endoscopic procedure. Upper gastrointestinal tract endoscopy was performed by an experienced gastroenterologist. Gastric mucosa biopsy samples were obtained from the antrum, incisura angularis, and corpus, and embedded in paraffin for histology. One additional antral biopsy for H. pylori culture was immediately frozen in glycerol and thioglycolate. The samples were shipped on dry ice to Vanderbilt University (Nashville, TN, USA) for analysis. Eighty subjects, from whom H. pylori could be isolated, were included in this study. Forty-one individuals (51.3%) were residents of Tuquerres, in the Andes Mountains, where the GC incidence is high, and 39 (48.7%) were residents of Tumaco, on the Pacific coast, where the GC incidence is low. Persons performing molecular analyses and histopathological assessments of biopsies were unaware of the diagnoses and residency of the patients. Ethics committees of the two local hospitals and of the Universidad del Valle in Cali, Colombia, approved the protocol for this study. All participants provided informed consent.

Isolation and culture of H. pylori strains

Antral biopsy specimens were homogenized under sterile conditions in 100 μL of sterile phosphate-buffered saline (PBS, pH 7.4) using a homogenizer (Kimble-Kontes, Vineland, NJ, USA). The homogenate was plated onto selective Trypticase soy agar (TSA) with 5% sheep blood containing vancomycin (20 mg/L), bacitracin (200 mg/L), nalidixic acid (10 mg/L) and amphotericin B (2 mg/L) (Sigma, St Louis, MO, USA), and a 1:10 dilution was plated on a TSA plate (BBL; LABSCO, Nashville, TN, USA) with no antibiotic. Agar plates were incubated under microaerobic conditions (6% O2, 6% CO2 and 88% N2; Campy Pak Plus envelope, BBL) at 37°C from 4 to 8 days until small grey translucent colonies appeared. Gram stains and assays for oxidase and urease were performed. Colony morphology was consistent with the characteristic shape of H. pylori colonies. Swabs and single colonies were used for DNA extraction.

Helicobacter pylori DNA extraction

DNA was extracted from bacterial pellets using Gentra Puregene reagents, according to the manufacturer’s instructions (Qiagen, Valencia, CA, USA). DNA from H. pylori reference strains 26695 (ATCC 700392), J99 (ATCC 700824) and Tx30a (ATCC 51932) were prepared for use as positive and negative controls for the presence of cagA. Quantification of DNA from all strains was performed spectrophotometrically.

PCR analysis

The primers used to amplify the 3′ region of the cagA gene were CAG1 (sense: 5′-ACC CTA GTC GGT AAT GGG TTA-3′) and CAG2 (anti-sense: 5′-GTA ATT GTC TAG TTT CGC-3′) [17], resulting in generation of fragments varying from 591 to 856 bp. PCR mixtures consisted of 10 mM Tris–HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 300 μM of each dNTP, 0.3 μmol of each primer, and 0.1 U of Taq DNA Polymerase (Sigma) in a final volume of 35 μL. Three microlitres of DNA extracted from the culture of the single colony was added to each reaction mixture as template. The PCR program for cagA amplification was as follows: 35 cycles consisting of 1 min at 95°C, 1 min at 50°C, and 1 min at 72°C, plus a final 7-min extension step at 72°C. PCR products (10 μL of each sample) were separated by electrophoresis through 2% agarose gels for 2 h at 60 V. To test for false-negative results due to variation in sequences hybridizing with primers, we employed a second, independent set of primers to determine the presence of cagA (cagAF: 5′-TTG ACC AAC AAC CAC AAA CCG AAG-3′; and cagAR: 5′-CTT CCC TTA ATT GCG AGA TTC C-3′) [25]. These primers amplify a 183-bp fragment in the 5′ end of the cagA gene. Positive and negative reagent control reactions, including a no-template control, were included in each amplification.

DNA sequence analysis

PCR products were purified using Microcon YM-100 Centrifugal Concentrators (Millipore Corp., Bedford, MA, USA). Sequencing was performed using BigDye terminator chemistry with an ABI 3730xl DNA Analyzer. Sequencing primers were the same as those used for PCR amplification. All cagA sequences were compared with sequences of reference WSS and ESS strains: H. pylori NCTC 11637 (GenBank AF202973), H. pylori 26695 (GenBank AE000511), H. pylori J99 (GenBank AE001439), H. pylori GC401 (F32) (GenBank AF202972), and H. pylori strain N364 (kindly provided to us by T. Ando from Nagoya, Japan). Sequence construction and alignment, as well as translation were performed using GENEIOUS software, version 2.5.3. (Biomatters, Auckland, New Zealand). The individuals performing all the genotyping and sequencing reactions were blinded to the location of residence of the subjects whose samples they were analysing.


Histopathological diagnoses for each participant were independently assessed by two experienced pathologists (P.C. and M.B.P.) in H&E-stained sections from antrum, incisura angularis and corpus, according to the updated Sydney System classification [26], and to the Padova international classification for dysplasia [27]. The categories were non-atrophic gastritis (NAG), multifocal atrophic gastritis without intestinal metaplasia (MAG), intestinal metaplasia (IM), and dysplasia. NAG or MAG were considered less severe lesions, and IM or dysplasia were considered more severe lesions. Cases with discordant diagnoses were reviewed in a double-headed microscope and a consensus was reached. The presence of H. pylori was assessed histologically by using a modified Steiner silver stain.

Induction of cell morphological changes

AGS cells were distributed into six-well plates at a density of cells of the GC line AGS 5 × 105 cells/well and allowed to adhere at 37°C in a 5% CO2 air-humidified atmosphere overnight.

Infection with 13 H. pylori strains (seven containing Western/Western and six containing Western/East Asian CM motifs) was performed at a concentration 100:1 (bacteria-to-cell ratio) followed by 24 h of incubation. Uninfected AGS cells as well as cells infected with cagA-positive (VT-2) and cagA-negative (Tx30a) H. pylori strains were used as controls. Cells were examined for the hummingbird phenotype by microscopy by two independent readers (L.S. and M.B.P.). The lengths of cytoskeletal projections were assessed on 100 consecutive hummingbird cells in randomly selected fields. We defined a hummingbird cell as a cell having a cellular needle-like protrusion >5 μm in length.

SHP-2 immunoprecipitation/immunodepletion and Western blot

AGS cells were co-cultured for 6 h with ten H. pylori strains (four containing Western/Western and six containing Western/East Asian CM motifs) at a 100:1 bacteria-to-cell ratio. H. pylori pellets (controls) and AGS cells at the end of co-culture were lysed in RIPA lysis buffer (50 mM Tris–HCl, pH 7.2/150 mM NaCl/1% Triton X-100/0.1% SDS). In order to determine whether there were differences in complex formation with SHP-2 and translocated CagA from different H. pylori strains, in preliminary experiments we performed co-immunoprecipitation assays with antibodies against SHP-2 and against CagA. However, sequential immunoprecipitation –Western blot analysis using those two antibodies in either order produced signal artefacts. Therefore, as an alternative method to assess SHP-2 and CagA complex formation, we performed immunodepletion assays using anti-SHP-2 antibodies in which we immunodepleted SHP-2 from the lysates and assessed whether there was concomitant depletion of the translocated CagA from the lysates. The lysates of AGS cells co-cultured with H. pylori were immunodepleted of SHP-2 by addition of SHP-2 antibody (Cell Signaling Technology, Inc., Danvers, MA, USA), and protein A/G plus agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The pre- and post-SHP-2-immunodepleted lysates were resolved by SDS–PAGE and electrotransferred to nitrocellulose membranes. The membranes were then blotted with mouse anti-CagA monoclonal antibodies (1:3000; Austral Biologicals, San Ramon, CA, USA) and goat IgG-HRP anti-mouse antibodies (Santa Cruz Biotechnology). To determine the extent of removal of CagA–SHP-2 complexes by the anti-SHP-2 antibody, we then performed densitometric analysis of the CagA bands with and without immunodepletion by the anti-SHP-2 antibody. Blots were developed using ECL detection reagents (Perkin Elmer LAS, Inc., Boston, MA, USA). Densitometric analysis of protein band density was performed on the X-ray films using IMAGEJ software. The relative extents of immunodepletion were calculated by dividing the pre-immunodepletion CagA band density by the post-immunodepletion CagA band, and the corrected band densities were calculated by dividing the pre-immunodepletion CagA band densities by the band density of a CagA standard loaded for each PAGE gel.

In order to evaluate the amount of CagA protein that had been phosphorylated by the AGS cell kinases (evidence of translocation), after stripping the blots, we blotted with mouse anti-phosphotyrosine PY-99 antibodies (1:300; Santa Cruz Biotechnology) and goat IgG-HRP anti-mouse antibodies (Santa Cruz Biotechnology). Detection of bound antibodies was performed as above for blotting with the anti-CagA antibody.


χ2, Fisher’s exact and Student’s t-tests were used as appropriate to assess the statistical significance of the differences in age, H. pylori status and variants of CagA in the risk area and for the histological diagnosis. Statistical analyses for hummingbird phenotype assessment were performed using linear fixed effects models. Data were analysed using STATA 9.2 (Stata Corp., College Station, TX, USA).



Among the 80 participants, 34 (42.5%) were classified as having NAG, 14 (17.5%) as having MAG, 29 (36.3%) as IM, and two cases (2.5%) were diagnosed as having dysplasia (displaying characteristics of the category ‘indefinite for dysplasia’ [27]). A subject with gastritis (1.2%), in which atrophy was impossible to assess, was excluded from the majority of the analyses. The distribution of cases according to histopathological diagnoses according to area is shown in Table 1. NAG was more frequent in men from the low-risk area (55.3%), whereas IM was most common in those from the high-risk area (48.8%). Overall differences in distribution of histopathological diagnoses according to area were significant, revealing a prevalence of more advanced lesions in the high-risk area for GC (p 0.034). H. pylori infection was highly prevalent (>90%) in subjects from both areas. Histological examination of the silver-stained sections of gastric biopsy specimens showed the presence of typical spiral-shaped H. pylori in 76 (95%) of the 80 cases and in 63 of 67 biopsy specimens from patients infected with cagA-positive strains (94%).

Status of Helicobacter pylori infection and histopathological diagnoses in Colombian subjects

Detection of 3′ variable region of the cagA gene

Amplification of the 3′ end of cagA from 80 DNA samples from cultured H. pylori strains generated fragments ranging from 591 to 856 bp in 67 subjects, indicating cagA-positive infection in those patients (83.8%; Table 1). All sequences from those fragments aligned well with the cagA 3′ ends of strains 26695 and NCTC 11637. Amplification of DNA from strains from the other 13 subjects did not generate cagA 3′ end fragments, and those subjects were considered cagA negative and excluded from further molecular examinations. The results of amplification of the smaller cagA fragment (183 bp) in all 80 samples agreed completely with the results from amplification of the 3′ end of the gene. The prevalence of cagA-positive strains was slightly higher in patients from the high-risk area compared with those from the low-risk region; however, this difference was not statistically significant (p 0.688; Table 1).

Characterization of the C-terminal variable region by EPIYA motif

Analysis of DNA and of predicted protein sequences revealed that in 43 (64.2%) of 67 strains, the fragment designated as ‘short’ (591–654 bp) encoded three EPIYA motifs in an ABC arrangement (Table 2 and Fig. 1). The shortest fragment (591 bp) encoded two EPIYAs in an arrangement of BC type, missing entirely the EPIYA-A segment. All 43 strains which produced a short fragment included a single EPIYA-C motif, as shown in Table 2.

FIG. 1
EPIYA-repeat polymorphisms in the C-terminal region of the CagA protein and their frequency in the Colombian Helicobacter pylori strains included in this analysis.
Number of EPIYA-C motifs according to histopathological diagnoses by area of risk for gastric cancer

In the other 24 strains, the C-terminal region was longer than 654 bp and encoded two or three EPIYA-C motifs. The shortest fragment of 693 bp was missing sequences encoding the WSS motif in the first EPIYA-C segment. In all other Colombian strains, each EPIYA-C segment contained a WSS. Twenty-three of 24 long fragments encoded two EPIYA-C motifs in an ABCC, ABCBC (856 bp) or ABBCC (772 bp) pattern (Fig. 1; Table 2). The ABCBC fragment included a sequence encoding an insert of 28 amino acids characteristic of a B segment. The pattern of 772 bp was designated ABBCC as it encoded a peptide with an insert of 19 amino acids between EPIYA-A and EPIYA-B. One of 24 long fragments included a sequence encoding five EPIYA motifs assembled in an ABCCC pattern (850 bp; Fig. 1; Table 2) possessing three EPIYA-C motifs. Patterns of the cagA 3′ end fragments in H. pylori strains and their frequency are shown in Fig. 1. Nucleotide and predicted protein sequences of all strains were deposited in GenBank, accession numbers FJ542221–FJ542279.

Number of EPIYA motifs and histopathology

We next investigated the relationship between the number of EPIYA-C motifs or CagA types and histopathological diagnoses in both high- and low-risk areas. H. pylori strains with only one EPIYA-C motif (ABC pattern) in the CagA proteins were more frequently associated with NAG and MAG; strains with two and three EPIYA-C motifs were more prevalent in subjects with IM and dysplasia, considering either the total number of cases (p <0.001), or segregated by low- (p 0.007) or high- (p 0.026) risk regions separately (Table 2). The significance for the high-risk area was stronger when two and three EPIYA-C motifs were combined (p 0.012, Table 2). The number of H. pylori strains with one EPIYA-C motif (ABC pattern), as well as strains with two and three EPIYA-C motifs in their CagA protein, was not significantly different in low- and high-risk areas (p 0.700; Table 3). No EPIYA-D motifs were detected in our strains. One-half of the Colombian strains carried CagA with a modified EPIYAB motif, with the threonine substituted for alanine as EPIYT instead of EPIYA. However, we found no association between EPIYA/EPIYT sequences and risk area (p 0.218) or disease outcome (data not shown).

Structural differences of the CagA protein of Helicobacter pylori strains from two geographic areas varying in risk for gastric cancer

Characterization of the C-terminal variable region by CM motif

In fifty-four of the 67 cagA-positive strains, the CM motifs located within EPIYA-C segments and immediately distal to the last repeat of the EPIYA-C segment had the peptide sequence FPLKRHDKVDDLSKVG, typically found in H. pylori strains from Western countries. Twenty-nine (69%) of 42 strains with three EPIYA motifs (ABC type, Table 3) contained two identical Western CM motifs (Fig. 2a), and the other 13 strains contained two different CM motifs (Fig. 2b; GenBank accession numbers EU250990–EU251001). The CM motif located within the EPIYA-C segment had the peptide sequence found in strains from Western countries; the CM motif located downstream of the EPIYA-C segment varied at positions 4, 6, 7, 8 and 10, in the pattern FPLRRSAA VEDLSKVG (two cases) and FPLRRSAKVEDLSKVG or FPLRRSAKVDDLSKVG (11 cases). The CM motifs following the EPIYA-C segment of the first two strains perfectly matched the CM motifs from 87 strains isolated from patients from East Asian countries (NCBI; Fig. 2c) and CM motifs from reference strain F32 and Japanese strain N364 (Fig. 3). The CM motifs located after the EPIYA-C of the other 11 strains exactly matched two clones of an H. pylori strain from Mexico and seven H. pylori strains from Japan. Because of the predominance of these CM motifs in strains from Japan, China and Korea, we designated these CM motifs as East Asian.

FIG. 2
CagA multimerization (CM) motifs in Colombian and Japanese Helicobacter pylori strains: (a) Colombian H. pylori strain Co5054 possesses two identical Western CM motifs: the first one is located in the N-terminal region of the EPIYA-C segment and the second ...
FIG. 3
Amino acid sequence characteristics of CagA C-terminal regions from six strains containing three EPIYA motifs. The reference strain F32 and the strain N364 isolated from a Japanese patient contain an East Asian CagA-specific sequence (ESS) (underlined, ...

Twelve of 13 strains bearing both Western and East Asian CM motifs were isolated from subjects from the low-risk area (p <0.001 for sequence and area; Table 3) and nine of those 12 (69.2%) were isolated from patients with less severe gastric lesions (NAG and MAG). Four strains bearing two different CM motifs were found in patients with IM (30.8%). As the CM motif distribution differed in low- and high-risk areas, we next determined if CM motifs were associated with functional differences.

Hummingbird phenotype

We aimed to assess whether the type of CM motif in our clinical ABC-type H. pylori strains was associated with the extent of hummingbird change. We found that H. pylori strains possessing Western/Western CM motifs were associated with a greater mean length (21.8 μm; 95% CI 20.7–22.8) of hummingbird elongations compared with strains possessing Western/East Asian CM motif (18.3 μm; 95% CI 17.2–19.4, p <0.001).

Immunoprecipitation and Western blots

In Western blots with the anti-phosphotyrosine antibody, phosphorylated CagA was detected in the same lysates as CagA, suggesting internalization of CagA into the cytosol of AGS cells (data not shown). We did not find any differences in the frequency of translocation of the CagA from H. pylori strains containing two different (one East-Asian and one Western) CM motifs or two identical (two Western) CM motifs: three out of six and two out of four CagAs were translocated, respectively (Fig. 4, Table 4). Densitometric analysis of the injected CagA demonstrated that the amounts of CagA injected into AGS cells by strains with two identical and two different CM motifs were comparable (average corrected band density 3.9 and 4.6, respectively). Comparing strains containing two different (one East Asian and one Western) vs. two identical (two Western) CM motifs, we found a quantitative difference in the amount of CagA that was removed by the anti-SHP-2 antibody and therefore complexed to SHP-2. There was an average 2.67-fold decrease in CagA upon immunodepletion of SHP-2 in the two Western CM motif samples, compared with an average 1.67-fold decrease in CagA in the strains having the East Asian/Western CM motifs. This suggests that there is a stronger affinity between CagA and SHP-2 from the Western/Western CM motif strains compared with the East-Asian/Western CM motif strains.

FIG. 4
Translocated CagA co-immunoprecipitates with SHP-2. AGS cells were co-cultured with Helicobacter pylori strains 1–10 for 6 h. Cell lysates were then immunodepleted of SHP-2 with an anti-SHP-2 antibody. Lysates with and without immunodepletion ...
CagA internalization in AGS cells by Helicobacter pylori strains possessing different CM motif type


In this study, we examined CagA C-terminal variations in H. pylori strains from Colombian patients from two areas with different GC risks. We found that the C-terminal region varied considerably in size, from 591 to 856 bp, and in number and structure of EPIYA and CM motifs. Encoded in their cagA 3′ end, Colombian H. pylori strains had one to three EPIYA-C motifs in an ABC, ABCC, ABCBC, ABBCC or ABCCC CagA pattern. We also found both Western and East Asian CM motifs in Colombian H. pylori strains containing WSS in their C-terminal CagA, and these variations were associated with functional differences. Recent studies have revealed much about the functional significance of variations in the C-terminal portions of the CagA proteins from clinical H. pylori isolates [13,28]. In co-culture experiments, the number of EPIYA-C motifs was directly correlated with levels of tyrosine phosphorylation and SHP-2-binding activity among Western CagA varieties [16,19,20]. Variations in the number of tyrosine phosphorylation sites in the C-terminal of CagA were associated with dephosphorylation and inactivation of focal adhesion kinase, induction of the hummingbird phenotype, and deregulation of the SHP-2 oncoprotein [13,29,30]. The CM motif, which is variably duplicated among distinct Western CagA species, may also disturb host cell function via SHP-2 deregulation [21]. CagA variations, particularly in the C-terminal region at the SHP-2 binding site, could affect the potential of different strains of H. pylori to promote gastric carcinogenesis [28]. We tested whether the number of EPIYA-C motifs or the type of CM motifs was associated with differences in GC risk in two Colombian areas.

We examined gastric mucosal lesions spanning multiple stages of the gastric carcinogenesis cascade [31] and their relationship to the number of EPIYA-C motifs. Our finding of significant associations between two and three EPIYA-C motifs in the CagA protein with more severe lesions is consistent with identification of that variation as a virulence marker in strains isolated from South Africa, Japan, Korea, and the central region of Colombia [17,18,20,32]. The proportion of strains bearing two (34.3%) and three (1.5%) EPIYAC sites appeared to be higher in our population of strains than previously reported in Mexican H. pylori strains (30%; [33]) or in Peruvian ones (11.5%; [34]). In the Peruvian study, the majority of variants contained one, rarely two, but not three EPIYA-C motifs. Of 33 H. pylori strains isolated in Costa Rica (GeneBank AF289432–289464; [35]), 21 contained one EPIYA-C, and 12 contained two EPIYA-C motifs. No strains contained three or more EPIYA-C motifs. By contrast, in a study of isolates from Colombian subjects, Yamaoka et al. [18] found that 9.5% of the H. pylori strains contained three or more EPIYA-C motifs, but this difference could be due to the selection of patients with different diagnoses, as almost half of their patients had GC, whereas the others had chronic gastritis and duodenal ulcer disease. No H. pylori strains in this study contained an EPIYA D motif despite their isolation from a Native American population with ancestry in East Asia.

In our previous studies of the well-established virulence markers of H. pylori (presence or absence of cagA, vacA s and m regions), we found a higher proportion of more virulent strains (cagA+, vacA s1/m1) in the Colombian region of high risk for GC [23,24]. In the present study, the distribution of strains with one, two and three EPIYA-C motifs was not significantly different in the two areas, suggesting the CagA type is not the cause of different GC incidence rates in those two areas. Other cagA 3′ end variations or other bacterial virulence factors, other genetic or environmental factors may be responsible for the high incidence of GC in the Andean regions of Colombia.

This report has revealed the diversity in the CM motifs of H. pylori strains circulating in the state of Nariño, Colombia. Although the number of H. pylori strains of ABC type CagA was essentially equal in the two areas, we found different types of CM motifs in ABC type strains from each area. We found H. pylori strains bearing a CM motif described previously only in strains from East Asian countries (NCBI database). Remarkably, the East Asian pattern of the CM motif appeared in the majority of strains from subjects from the low-risk area. Identification of CM sequences associated with East Asian strains was unexpected in the low-risk region, because those populations are characterized by African, rather than Native American ancestry. In our study, the presence of the East Asian CM motif was not associated with the severity of gastric lesions. However, the majority of strains bearing a combination of one East Asian and one Western CM motif were found in patients with less severe lesions (NAG). In our in vitro experiments, immunoblotting of CagA in lysates with and without SHP-2 immunodepletion suggested that higher affinity CagA–SHP-2 complexes could be formed by strains containing two Western CM motifs compared with strains having two different CM motifs. In parallel, more intense induction of hummingbird responses were induced by strains containing two Western CM motifs in their CagA compared with those possessing one Western and one East Asian CM motif. These results suggest that the East Asian CM motif in the Colombian, Western-type CagA may be less functional in multimerization compared with the Western CM motif. Altered multimerization potential of CagA proteins in strains bearing two different CM motifs may help to explain the prevalence of less severe premalignant lesions (NAG) in the low-risk area. In vitro experiments with isogenic mutants will be necessary to establish this result unequivocally.

In conclusion, we have demonstrated that the 3′ end of the cagA gene and its CagA protein in Colombian H. pylori strains is heterogeneous with respect to size, to the number of EPIYA, WSS and CM motifs and the internal structure. H. pylori strains with two and three EPIYA-C motifs were associated with more severe gastric lesions. H. pylori strains with two different CM motifs—one Western and one East Asian—showed altered biological activity in vitro and were found predominantly in patients with less severe gastric lesions. This finding may partially explain the lower incidence of GC on the Pacific coast of Colombia.


We are very grateful to T. Ando from the University of Nagoya, Japan, for the clinical strain of Helicobacter pylori N364, containing the East Asian Specific Sequence. We thank M. Hatakeyama, Division of Molecular Oncology, Institute of Genetic Medicine, Hokkaido University, Sapporo, Japan for helpful discussions.

Transparency Declaration

This work was supported by the Health Excellence Fund of the Board of Regents of the State of Louisiana, USA (HEF 2000-05-03), by NIH Grant P01 CA 28842, by Specialized Project of Research Excellence (P50 CA95103), by NIH RO1 Grants (CA 77955, DK 58587 and DK 73902), and by the Vanderbilt University Medical Center’s Digestive Disease Research Center, supported by NIH Grant P30DK058404. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NCI or NIH.


The authors have no conflict of interest to declare.


1. Covacci A, Telford JL, Del Giudice G, et al. Helicobacter pylori virulence and genetic geography. Science. 1999;284:1328–1333. [PubMed]
2. Cover TL, Blaser MJ. Helicobacter pylori factors associated with disease. Gastroenterology. 1999;117:257–260. [PubMed]
3. Censini S, Lange C, Xiang Z, et al. cag, a pathogenicity island of Helicobacter pylori, encodes type I-specific and disease-associated virulence factors. Proc Natl Acad Sci USA. 1996;93:14648–14653. [PubMed]
4. Covacci A, Censini S, Bugnoli M, et al. Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proc Nat Acad Sci USA. 1993;90:5791–5795. [PubMed]
5. Tummuru MKR, Cover TL, Blaser MJ. Cloning and expression of a high-molecular mass major antigen of Helicobacter pylori: evidence of linkage to cytotoxin production. Infect Immun. 1993;61:1799–1809. [PMC free article] [PubMed]
6. Blaser MJ, Perez-Perez GI, Kleanthou TL, et al. Infection with Helicobacter pylori strains possessing cagA is associated with an increased risk of developing adenocarcinoma of the stomach. Cancer Res. 1995;55:2111–2115. [PubMed]
7. Parsonnet J, Friedman GD, Orentreich N, et al. Risk for gastric cancer in people with CagA positive or CagA negative Helicobacter pylori infection. Gut. 1997;40:297–301. [PMC free article] [PubMed]
8. Correa P. Helicobacter pylori and gastric carcinogenesis. Am J Surg Pathol. 1995;19:S37–43. [PubMed]
9. Asahi M, Azuma T, Ito S, et al. Helicobacter pylori CagA protein can be tyrosine phosphorylated in gastric epithelial cells. J Exp Med. 2000;191:593–602. [PMC free article] [PubMed]
10. Backert S, Ziska E, Brinkmann V, et al. Translocation of the Helicobacter pylori CagA protein in gastric epithelial cells by a type IV secretion apparatus. Cell Microbiol. 2000;2:155–164. [PubMed]
11. Backert S, Moese S, Selbach M, et al. Phosphorylation of tyrosine 972 of the Helicobacter pylori CagA protein is essential for induction of a scattering phenotype in gastric epithelial cells. Mol Microbiol. 2001;42:631–644. [PubMed]
12. Selbach M, Moese S, Hauck CR, et al. Src is the kinase of the Helicobacter pylori CagA protein in vitro and in vivo. J Biol Chem. 2002;277:6775–6778. [PubMed]
13. Hatakeyama M, Higashi H. Helicobacter pylori CagA: a new paradigm for bacterial carcinogenesis. Cancer Sci. 2005;96:835–843. [PubMed]
14. Tammer I, Brandt S, Hartig R, et al. Activation of Abl by Helicobacter pylori: a novel kinase for CagA and crucial mediator of host cell scattering. Gastroenterology. 2007;132:1309–1319. [PubMed]
15. Yamazaki S, Yamakawa A, Ito Y, et al. The CagA protein of Helicobacter pylori is translocated into epithelial cells and binds to SHP-2 in human gastric mucosa. J Infect Dis. 2003;187:334–337. [PubMed]
16. Higashi H, Tsutsumi R, Fujita A, et al. Biological activity of the Helicobacter pylori virulence factor CagA is determined by variation in the tyrosine phosphorylation sites. Proc Nat Acad Sci USA. 2002;99:14428–14433. [PubMed]
17. Yamaoka Y, Kodama T, Kashima K, et al. Variants of the 3′ region of the cagA gene in Helicobacter pylori isolates from patients with different H. pylori-associated diseases. J Clin Microbiol. 1998;36:2258–2263. [PMC free article] [PubMed]
18. Yamaoka Y, El-Zimaity HMT, Gutierrez O, et al. Relationship between the cagA 3′ repeat region of Helicobacter pylori, gastric histology, and susceptibility to low pH. Gastroenterology. 1999;117:342–349. [PubMed]
19. Naito M, Yamazaki T, Tsutsumi R, et al. Influence of EPIYA-repeat polymorphism on the phosphorylation-dependent biological activity of Helicobacter pylori CagA. Gastroenterology. 2006;130:1181–1190. [PubMed]
20. Argent RH, Kidd MS, Owen RJ, et al. Determinants and consequences of different levels of CagA phosphorylation for clinical isolates of Helicobacter pylori. Gastroenterology. 2004;127:514–523. [PubMed]
21. Ren S, Higashi H, Lu H, et al. Structural basis and functional consequence of Helicobacter pylori CagA multimerization in cells. J Biol Chem. 2006;281:32344–32352. [PubMed]
22. Camargo MC, Yepez MC, Ceron C, et al. Age at acquisition of Helicobacter pylori infection: comparison of two areas with contrasting risk of gastric cancer. Helicobacter. 2004;9:262–270. [PubMed]
23. Bravo LE, van Doorn LJ, Realpe JL, et al. Virulence-associated genotypes of Helicobacter pylori: do they explain the African Enigma? Am J Gastroenterol. 2002;97:2839–2842. [PubMed]
24. Sicinschi LA, Correa P, Peek RM, Jr, et al. Helicobacter pylori genotyping and sequencing using paraffin-embedded biopsies from residents of Colombian areas with contrasting gastric cancer risks. Helicobacter. 2007;12:624–634. [PMC free article] [PubMed]
25. van Doorn LJ, Figueiredo C, Rossau R, et al. Typing of the Helicobacter pylori vacA gene and detection of the cagA gene by PCR and reverse hybridization. J Clin Microbiol. 1998;36:1271–1276. [PMC free article] [PubMed]
26. Dixon MF, Genta RM, Yardley JH, et al. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol. 1996;20:1161–1181. [PubMed]
27. Rugge M, Correa P, Dixon MF, et al. Gastric dysplasia—the Padova international classification. Am J Surg Pathol. 2000;24:167–176. [PubMed]
28. Hatakeyama M. Oncogenic mechanisms of the Helicobacter pylori CagA protein. Nat Rev Cancer. 2004;4:688–694. [PubMed]
29. Higashi H, Nakaya A, Tsutsumi R, et al. Helicobacter pylori CagA induces Ras-independent morphogenetic response through SHP-2 recruitment and activation. J Biol Chem. 2004;279:17205–17216. [PubMed]
30. Tsutsumi R, Higashi H, Higuchi M, et al. Attenuation of Helicobacter pylori CagA × SHP-2 signaling by interaction between CagA and C-terminal Src kinase. J Biol Chem. 2003;278:3664–3670. [PubMed]
31. Correa P, Haenszel W, Cuello C, et al. A model for gastric cancer epidemiology. Lancet. 1975;2:58–60. [PubMed]
32. Azuma T, Yamakawa A, Yamazaki S, et al. Correlation between variation of the 3′ region of the cagA gene in Helicobacter pylori and disease outcome in Japan. J Infect Dis. 2002;186:1621–1630. [PubMed]
33. Reyes-Leon A, Atherton JC, Argent RH, et al. Heterogeneity in the activity of Mexican Helicobacter pylori strains in gastric epithelial cells and its association with diversity in the cagA gene. Infect Immun. 2007;75:3445–3454. [PMC free article] [PubMed]
34. Devi SM, Ahmed I, Khan AA, et al. Genomes of Helicobacter pylori from native Peruvians suggest admixture of ancestral and modern lineages and reveal a western type cag-pathogenicity island. BMC Genomics. 2006;7:191. [PMC free article] [PubMed]
35. Occhialini A, Marais A, Urdaci M, et al. Composition and gene expression of the cag pathogenicity island in Helicobacter pylori strains isolated from gastric carcinoma and gastritis patients in Costa Rica. Infect Immun. 2001;69:1902–1908. [PMC free article] [PubMed]