Attenuation of Helicobacter pylori CagA·SHP-2 Signaling by Interaction between CagA and C-terminal Src Kinase*

Helicobacter pylori (H. pylori) is a causative agent of gastric diseases ranging from gastritis to cancer. The CagA protein is the product of thecagA gene carried among virulent H. pyloristrains and is associated with severe disease outcomes, most notably gastric carcinoma. CagA is injected from the attached H. pylori into gastric epithelial cells and undergoes tyrosine phosphorylation. The phosphorylated CagA binds and activates SHP-2 phosphatase and thereby induces a growth factor-like morphological change termed the “hummingbird phenotype.” In this work, we demonstrate that CagA is also capable of interacting with C-terminal Src kinase (Csk). As is the case with SHP-2, Csk selectively binds tyrosine-phosphorylated CagA via its SH2 domain. Upon complex formation, CagA stimulates Csk, which in turn inactivates the Src family of protein-tyrosine kinases. Because Src family kinases are responsible for CagA phosphorylation, an essential prerequisite of CagA·SHP-2 complex formation and subsequent induction of the hummingbird phenotype, our results indicate that CagA-Csk interaction down-regulates CagA·SHP-2 signaling by both competitively inhibiting CagA·SHP-2 complex formation and reducing levels of CagA phosphorylation. We further demonstrate that CagA·SHP-2 signaling eventually induces apoptosis in AGS cells. Our results thus indicate that CagA-Csk interaction prevents excess cell damage caused by deregulated activation of SHP-2. Attenuation of CagA activity by Csk may enable cagA-positive H. pylori to persistently infect the human stomach for decades while avoiding excess CagA toxicity to the host.

Helicobacter pylori is a micro-aerophilic spiral-shaped bacterium (1) and is estimated to infect about half of the world's population. It colonizes the human stomach and persists for several decades, causing chronic gastritis and peptic ulcer diseases (2). Recent epidemiological studies have further indicated that H. pylori infection is associated with the develop-ment of gastric adenocarcinoma and gastric mucosa-associated lymphoid tissue lymphoma (3)(4)(5)(6).
The H. pylori cagA gene encodes the 120-to 145-kDa CagA protein, and it is a marker for the presence of the cag pathogenicity island (7)(8)(9). More recently, the biological heterogeneity of H. pylori has been recognized, and H. pylori strains harboring the cagA gene have been suspected to have a specific responsibility in promoting the atrophic-metaplastic mucosal lesions that represent the most recognized pathway in multistep intestinal-type gastric carcinogenesis. Thus they are considered to be more virulent than cagA-negative strains (5,10,11). Consistently, molecular epidemiological studies have suggested that cagA-positive H. pylori infection significantly increases the risk of gastric carcinoma (5,12,13).
In vitro infection of gastric epithelial cells with cagA-positive, but not cagA-negative, H. pylori induces a unique morphological change termed the hummingbird phenotype, which is characterized by strong elongation of the cell (14 -16). During the infection, the CagA protein, which is produced within H. pylori, is translocated from the bacteria into the attached epithelial cells via the bacterial type IV injection apparatus. The translocated CagA protein then localizes at the inner surface of the plasma membrane and undergoes tyrosine phosphorylation by the Src family of protein-tyrosine kinases such as c-Src, Lyn, Fyn, and Yes (17)(18)(19)(20)(21)(22)(23). In vivo tyrosine phosphorylation sites of CagA are characterized by Glu-Pro-Ile-Tyr-Ala (EPIYA) motifs, which vary in number among different H. pylori strains (24,25).
We as well as others have demonstrated that CagA is the essential and sufficient H. pylori factor in the induction of the hummingbird phenotype (15,16,22). Tyrosine phosphorylation of CagA is a prerequisite for the morphological changes. We have further shown that CagA binds and stimulates an SH2 1 domain-containing protein-tyrosine phosphatase, SHP-2, in a tyrosine phosphorylation-dependent manner (16). Thus, CagA induces the hummingbird phenotype by recruiting and activating SHP-2 at the plasma membrane. Considering the critical roles of SHP-2 in transmission of mitogenic signals as well as in regulation of cell motility (26 -28), perturbation of SHP2 by CagA is thought to be substantially involved in pathological processes that are associated with cagA-positive H. pylori infection.
Although it is known that the majority of the CagA proteins in host cells form complexes with SHP-2, we found in this work that CagA is also capable of binding C-terminal Src kinase (Csk) (29 -33). Upon binding with Csk, CagA stimulates its kinase activity and thereby inactivates Src family kinases. Given that Src family kinases are responsible for CagA phosphorylation (22,23) and that this CagA phosphorylation is an essential prerequisite for CagA⅐SHP-2 complex formation and subsequent induction of the hummingbird phenotype (16), our findings indicate that CagA-Csk interaction down-regulates CagA⅐SHP-2 signaling that perturbs cellular functions.
Construction of DNA-Mammalian expression vectors for HA-tagged wild-type CagA (WT CagA-HA) isolated from NCTC11637 H. pylori and its phosphorylation-resistant mutant (PR CagA-HA) have been described previously (16). EPIYA mutants of WT CagA-HA, abCCC and ABccc, were generated from HA-tagged wild-type CagA by substituting tyrosine residues that constitute EPIYA motifs with alanine residues using a Chameleon site-directed mutagenesis kit (Stratagene), and the DNA fragments encoding these mutants were cloned into a pSP65SR␣ mammalian expression vector. Rat Csk cDNA was C-terminal Mycepitope-tagged and cloned into pSP65SR␣ (Csk-Myc). To disable the SH2 domain of Csk, serine 109 of Csk was substituted with the cysteine residue (CskS109C-Myc) by site-directed mutagenesis. A kinase-dead mutant of human c-Src was made by replacing ATP-binding lysine 298 to alanine (Src⌬K). A kinase-dead and autophosphorylation-defective double mutant of human c-Src was made by replacing lysine 298 and tyrosine 419 with alanine and phenylalanine, respectively (Src⌬K⌬Y). These cDNAs were cloned into pSP65SR␣. Myc-tagged SHP-2 and its constitutively active mutant, myr-SHP-2⌬SH2, have been described previously (16).
Cell Culture and Transfection-AGS human gastric epithelial cells and monkey COS-7 cells were, respectively, cultured in RPMI 1640 medium and Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum. Expression vectors were transiently transfected into the cells by using LipofectAMINE 2000 reagent (Invitrogen) as previously described (16). Cells were harvested at 36 h after transfection. The morphology of the AGS cells was examined at 17 h post-transfection.
Immunoprecipitation and Immunoblotting-Cells were lysed in lysis buffer as described previously (16). Cell lysates were treated with the appropriate antibody, and then immune complexes were trapped on protein A-or protein G-Sepharose beads. Total cell lysates and immunoprecipitates were subjected to SDS-PAGE. Proteins transferred to polyvinylidene difluoride membranes (Millipore) were soaked in solutions of primary antibodies and then visualized using Western blot chemiluminescence reagent (PerkinElmer Life Sciences). Quantitation of chemiluminescence on the immunoblotted membrane was performed by using a luminescence image analyzer (LAS1000, FUJIFILM).
In Vitro Kinase Assay-Csk kinase activity was measured with the use of poly(Glu,Tyr) 4:1 (Sigma) as a substrate (34). Cells were lysed in lysis buffer mB (50 mM HEPES-NaOH, pH 8.0, 100 mM NaCl, 5 mM EDTA, and 1% Brij-35) containing 2 mM Na 3 VO 4 , 2 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 10 g/ml trypsin inhibitor, and 10 g/ml aprotinin. Cell lysates were treated with anti-Myc monoclonal antibody, and the immune complex was then trapped on protein G-Sepharose beads. After wash for three times with lysis buffer mB and two times with wash buffer (50 mM HEPES-NaOH, pH 8.0, 100 mM NaCl, 3 mM MnCl 2 , and 0.1 mM Na 3 VO 4 ), immunoprecipitates were incubated in 25 l of kinase assay buffer (50 mM HEPES-NaOH, pH 8.0, 100 mM NaCl, 10 mM MgCl 2 , and 0.1 mM Na 3 VO 4 ) containing 20 M ATP and 2 g of poly(Glu,Tyr) 4:1 at 30°C for indicated time. The reaction was terminated by the addition of SDS gel loading buffer. The reaction mixtures were then subjected to SDS-PAGE followed by immunoblotting. The amount of phosphorylated tyrosine residues in poly-(Glu,Tyr) 4:1 were quantified using the luminescence image analyzer.
Colony Suppression Assay-AGS cells (1.5 ϫ 10 6 ) were transfected with 1.5 g of the puromycin-resistant gene (pBabe-puro) and 28.5 g of expression plasmid by using LipofectAMINE 2000 reagent. At 16 h after the transfection, cells were collected, split into twelve 100-mm plates, and selected in RPMI 1640 medium/10% fetal bovine serum containing 0.4 g/ml puromycin for 10 days. After the drug selection, the cells were stained with May-Giemsa solution. Densities of colonies were quantitated by using the LAS1000.
Flow Cytometric Analysis-AGS cells (1.5 ϫ 10 6 ) were transfected with expression vector or control empty vector. At 36 h after the transfection, cells were collected, treated with Cy3-conjugated Annexin V (Bio Vision), and subjected to flow cytometric analysis using FACSCalibur (BD Biosciences).

RESULTS
Physical Interaction between CagA and Csk-The cagA gene was isolated from the genome of H. pylori standard strain NCTC11637 and, after addition of a C-terminal hemagglutinin (HA) tag, was cloned into the mammalian expression vector pSP65SR␣ as previously described (16). The cagA gene product, CagA, undergoes tyrosine phosphorylation when expressed in mammalian cells. Notably, NCTC11637-derived CagA protein possesses five glutamic acid-proline-isoleucine-tyrosine-alanine (EPIYA) motifs that are potential tyrosine phosphorylation sites in gastric epithelial cells (Fig. 1A). Complex formation between CagA and SHP-2 is reconstituted in COS-7 cells by co-transfection of CagA and SHP-2 expression vectors (16). Using this reconstitution system, we screened additional SH2containing cellular proteins that are capable of physically interacting with CagA and found that C-terminal Src kinase (Csk) bound wild-type CagA when co-expressed in COS-7 cells (Fig. 1B, lanes 1-4). In contrast, phosphorylation-resistant CagA, in which all of the five tyrosine-phosphorylatable EPIYA motifs were replaced with non-phosphorylatable EPIAA, was incapable of binding Csk. This indicates that the complex formation is entirely dependent on tyrosine phosphorylation of CagA. Csk is known to inactivate c-Src protein-tyrosine kinase by phosphorylation tyrosine 530 (Tyr-530) in the C-terminal tail of c-Src (Tyr-527 in avian c-Src). Csk is considered to phosphorylate and inactivate other Src family kinases as well. Using the co-expression system in COS-7, we also overexpressed the p85 subunit of phosphatidylinositol 3-kinase, c-Src, Gab1, Gab2, Shc, Grb2, Crk-II, or Sos, together with CagA. However, we were not able to detect any specific interaction between CagA and these proteins in COS-7 cells. Thus, CagA specifically binds SHP-2 and Csk. 2 To determine which EPIYA motifs were involved in Csk binding in COS-7 cells, we employed two EPIYA mutants of CagA, abCCC and ABccc (Fig. 1A). In the abCCC mutant, the former two EPIYA motifs were converted into phosphorylationresistant EPIAA sequences. Similarly, the latter three EPIYA motifs, which were generated by duplication of an EPIYAcontaining 34-amino acid sequence three times, were replaced by EPIAA to make the ABccc mutant. Co-expression studies in COS-7 cells revealed that both abCCC and ABccc mutants were capable of binding Csk, indicating that multiple EPIYA motifs can independently bind Csk once they are tyrosine-phosphorylated (Fig. 1B, lanes 5-8).
To confirm the CagA-Csk interaction in gastric epithelial cells, CagA was transiently expressed in the AGS human gastric carcinoma cell line. From the cell lysates prepared, Csk was immunoprecipitated using anti-Csk, and the anti-Csk immunoprecipitates were immunoblotted with anti-HA, which specifically detects the HA-tagged CagA. As expected, endogenous Csk was capable of specifically binding wild-type, but not the phosphorylation-resistant, CagA in AGS cells (Fig. 1C). Hence, CagA formed a complex with endogenous Csk in gastric epithelial cells, and the complex formation was strictly dependent on tyrosine phosphorylation of CagA.

Involvement of the Csk SH2 Domain in CagA⅐Csk Complex
Formation-The above-described observations suggested that CagA-Csk interaction involved the phosphotyrosine-containing EPIYA motifs of CagA. Like SHP-2, Csk possesses an SH2 domain, which specifically binds a phosphotyrosine-containing peptide (35)(36)(37)(38). To determine the role of the Csk SH2 domain in CagA binding, we generated a Csk mutant, CskS109C, in which serine 109 was replaced by cysteine to destroy the structural integrity of the SH2 domain as previously reported (36). When expressed in COS-7 cells, CskS109C was incapable of binding wild-type CagA (Fig. 2). Accordingly, we concluded that CagA⅐Csk complex formation is mediated by the interaction between tyrosine-phosphorylated EPIYA motifs of CagA and the SH2 domain of Csk.
Activation of Csk by Complex Formation with CagA-To elucidate the biological consequences of the CagA-Csk interaction, we decided to examine the effect of CagA binding on Csk kinase activity. To do so, Myc-tagged Csk was ectopically expressed in COS-7 cells in the absence or presence of CagA. Cell lysates were prepared and Csk was immunoprecipitated from the lysates with anti-Myc (Fig. 3A). The immunopurified Csk was then subjected to an in vitro kinase assay using poly(Glu, Tyr) 4:1 as a substrate (34). As shown in Fig. 3 (B and C), Csk kinase activity was strongly potentiated when Csk formed a complex with CagA. Based on the observations, we concluded that CagA stimulates Csk kinase activity through the physical complex formation. Next, to determine whether CagA indeed activates Csk in cells, we generated a human c-Src derivative, Src⌬K⌬Y, in which lysine 298, an ATP-binding site, and tyrosine 419, an autophosphorylation site (39,40), were respectively replaced by arginine and phenylalanine residues. The resulting Src⌬K⌬Y is enzymatically inactive and is not autophosphorylated at Tyr-419 because of the mutation, yet it undergoes tyrosine phosphorylation at Tyr-530 by Csk. The c-Src mutant was co-expressed together with CagA in AGS cells, and its phosphorylation levels were examined by anti-phosphotyrosine immunoblotting. As shown in Fig. 4A, Src⌬K⌬Y became more tyrosine-phosphorylated when CagA was co-expressed in cells. Because Src⌬K⌬Y possesses only one tyrosine phosphorylation site, Tyr-530, the result indicates CagA stimulated Csk and the activated Csk phosphorylated Tyr-530 of c-Src.
Inactivation of c-Src by CagA-Once Tyr-530 is tyrosinephosphorylated by Csk, c-Src forms a "closed" inactive structure. In contrast, c-Src changes to an "open" active state by dephosphorylation of Tyr-530 and autophosphorylation at Tyr-419 (29 -31, 39, 40). The above-described observations indicate that CagA-Csk interaction activates Csk and thereby inhibits c-Src kinase activity through inhibitory phosphorylation at Tyr-530. To investigate this, we examined c-Src kinase activity using anti-phospho-Src and anti-nonphospho-Src antibodies, which detect active c-Src that is phosphorylated at Tyr-419, and inactive c-Src, which is non-phosphorylated at Tyr-419, respectively. CagA was co-expressed in AGS cells with the kinase-dead c-Src, Src⌬K, in which Lys-298 was substituted with arginine, and levels of phosphorylation on Tyr-419 were examined using these antibodies. Under conditions in which comparable amounts of Src⌬K were expressed, CagA significantly increased the levels of inactive c-Src that was specifically detected by anti-nonphospho-Src, whereas the levels of active c-Src, which were detected by anti-Phospho-Src, were reciprocally reduced (Fig. 4B). The observation indicated that CagA inactivated c-Src by activating Csk upon physical com- plex formation. As the result, the levels of autophosphorylation at Tyr-419 in Src⌬K were reduced. Notably, overall tyrosine phosphorylation levels of Src⌬K, which possesses both Tyr-419 and Tyr-530, slightly increased in the presence of CagA as determined by anti-phosphotyrosine immunoblotting. This was most probably due to elevated levels of Src⌬K phosphorylated at Tyr-530 by Csk. Because Csk is considered to be the universal negative regulator for the Src family kinases (29 -31, 39, 40), it should collectively phosphorylate and inactivate the Src family members expressed in gastric epithelial cells upon complex formation with CagA. Notably, Src⌬K underwent autophosphorylation at Tyr-419, yet it is catalytically inactive. Hence, autophosphorylation of c-Src is an intermolecular, rather than intramolecular, process caused by c-Src dimerization as described previously (41). quires activation of SHP-2, which is induced by CagA-SHP-2 interaction (16). Because tyrosine phosphorylation of CagA by Src family kinases is an essential prerequisite for CagA⅐SHP-2 complex formation, inhibition of Src family kinases by Csk was suspected to reduce phosphorylation levels of CagA and thus to inhibit the hummingbird phenotype induction by CagA.

Inhibition of the Hummingbird Phenotype Induction by Ec
To determine whether Csk acts as a negative regulator of CagA⅐SHP-2 signaling, we examined the effect of ectopic Csk expression on the hummingbird phenotype induction by CagA. As shown in Fig. 5, Co-expression of Csk strongly inhibited the CagA-dependent morphological transformation. In contrast, the CskS109C mutant, which cannot bind CagA, exhibited much less activity to inhibit the hummingbird phenotype than the wild-type Csk did (Fig. 5, A and B). The observations indicate that, upon complex formation with CagA, Csk counteracts CagA⅐SHP-2 signaling by inactivating Src kinases and competitively inhibiting CagA⅐SHP-2 interaction. A less but significant inhibition of the CagA activity by CskS109C, which cannot bind CagA, may be due to basal kinase activity of the Csk109C mutant, which dose not require CagA binding.
Effects of CagA and SHP-2 on the Growth of AGS Cells-The above-described observations indicate that Csk, once activated by CagA, inhibits the complex formation between CagA and SHP-2, thereby down-regulating CagA-dependent activation of SHP-2. The fact that there is a negative-feedback regulation of CagA activity raises the possibility that sustained activation of SHP-2 by CagA causes adverse effects on the colonization of cagA-positive H. pylori in the stomach. To address this possibility, we examined the long term effect of CagA or SHP-2 on the growth of AGS cells. To do so, we transfected a cDNA expression vector for CagA, wild-type SHP-2, or membranetargeted, constitutively active SHP-2 (Myr-⌬N-SHP-2) (16), together with the puromycin-resistance gene. After selection of the transfected cells with puromycin, the number of drugresistant colonies was counted. As shown in Fig. 6A, expression of CagA induced strong inhibition of colony formation. Also, expression of wild-type SHP-2 resulted in a significant reduction of puromycin-resistant colonies, indicating that these proteins are growth-inhibitory. Furthermore, the constitutively active Myr-⌬N-SHP-2 exhibited stronger activity to reduce the colony number than wild-type SHP-2 did. The growth-inhibitory activity of CagA or SHP-2 may be due to cell cycle arrest or programmed cell death. To discriminate these two possibilities, we performed flow cytometry analysis using Annexin V, a sensitive method to detect apoptosis. At 36 h after transfection of CagA, wild-type SHP-2, or Myr-⌬N-SHP-2 expression vector, a significant fraction of the transfected AGS cells was positive for Annexin V. From these observations, we concluded that ectopic overexpression of CagA induces apoptosis and that this CagA-dependent apoptosis is due to deregulation of SHP-2 caused by CagA⅐SHP-2 complex formation (Fig. 6B).

DISCUSSION
CagA is translocated from H. pylori into the attached gastric epithelial cells via the type IV injection system (14,(17)(18)(19)(20)(21). In this work, we found a novel interaction between the translocated CagA and Csk in gastric epithelial cells. We previously demonstrated that CagA undergoes tyrosine phosphorylation and binds an SH2 domain-containing protein-tyrosine phosphatase, SHP-2, in a tyrosine phosphorylation-dependent manner. Upon complex formation, CagA activates SHP-2 phosphatase activity and thereby initiates a cellular morphological change termed the hummingbird phenotype (14 -16).
As is the case with SHP-2, the CagA-Csk interaction is strictly dependent on CagA tyrosine phosphorylation and involves the SH2 domain of Csk. This indicates that Csk and SHP-2 compete with each other to bind CagA. In this regard, a predominant fraction of the CagA proteins expressed in gastric epithelial cells is present in complexes with SHP-2 (16), suggesting that SHP-2 has higher affinity to bind phosphorylated CagA than does Csk and/or that SHP-2 is more abundantly expressed than is Csk in cells. Our present work thus indicates that a small but significant fraction of tyrosine-phosphorylated CagA proteins forms complexes with Csk in host cells.
The CagA-Csk interaction described in this work is thought to play an important role in the regulation of CagA⅐SHP-2 signaling. Csk has been well characterized as a negative regulator of Src family kinases; it phosphorylates the C-terminal tail of c-Src and probably those of other Src family members as well, resulting in an intramolecular interaction between this phosphorylated tyrosine residue and the c-Src SH2 domain that renders c-Src inactive (29 -31, 39, 40). Src family kinases expressed in gastric epithelial cells (such as c-Src, Fyn, Lyn, and Yes) are responsible for CagA phosphorylation, and this CagA phosphorylation is an essential prerequisite for CagA⅐SHP-2 complex formation as well as subsequent induction of the hummingbird phenotype (16,22,23). Our findings therefore indicate the presence of a molecular circuitry that regulates the biological activity of CagA within certain ranges of intensity by controlling tyrosine-phosphorylation levels of CagA. In this regulation (Fig. 7), the translocated CagA protein first undergoes tyrosine phosphorylation by Src family kinases, which are somehow activated in gastric epithelial cells. A predominant fraction of tyrosine-phosphorylated CagA proteins binds and deregulates SHP-2, generating signals that give rise to the hummingbird phenotype. Simultaneously, a small but significant fraction of tyrosine-phosphorylated CagA proteins binds and stimulates Csk, which in turn phosphorylates and inactivates Src family kinases. Reduced Src family kinase activity then down-regulates levels of CagA phosphorylation, followed by diminished CagA⅐SHP-2 complex formation. Accordingly, Csk functions as a negative regulator of CagA⅐SHP-2 signaling. Activated Csk also down-regulates CagA⅐Csk complex formation, eventually diminishing Csk kinase activity as well. Such a transient nature of Csk activation may provoke oscillation in the intensity of CagA⅐SHP-2 signaling.
The biological relevance of down-regulation of CagA⅐SHP-2 signaling by CagA⅐Csk has been shown by the observation in the study that sustained activation of SHP-2 by CagA triggers apoptosis in gastric epithelial cells. Given the physiological role of SHP-2 in transmission of mitogenic signals between receptor tyrosine kinase and Ras, the CagA⅐SHP-2 complex is most likely to stimulate Ras as well as other signaling molecules. Notably, forced expression of activated Ras induces apoptosis due to signaling imbalance provoked by the deregulated Ras activity (42,43). Deregulated stimulation of SHP-2 by CagA may also cause an inappropriate Ras activation, which eventually triggers programmed cell death. Massive loss of gastric epithelial cells by CagA-dependent apoptosis, which might provoke severe gastro-duodenal ulcers, is obviously disadvantageous for H. pylori in maintaining long term colonization in the stomach.
Attenuation of CagA⅐SHP-2 signaling by Csk may prevent excess CagA toxicity and enable cagA-positive H. pylori to persistently infect the human stomach for decades yet, at the same time, provoke chronic gastric damage by occasional but transient deregulation of SHP-2 upon CagA injection. Our findings with AGS cells are supported by the observation that cagA-positive H. pylori infection is associated with profound changes in the pattern of epithelial cell turnover in gastric glands. Results of both in vitro and in vivo studies on H. pylori A, AGS cells were co-transfected with pBabe-puro and indicated expression vector or control empty vector. Following transfection, cells were selected with puromycin and, on day 10, were fixed and stained. Densities of colonies in each dishes were quantitated by using a fluorescence image analyzer LAS1000 (FUJIFILM), and relative values are indicated. Bars represent 2ϫ S.D. of the three independent experiments. B, at 36 h after transfection, cells were harvested, incubated with Cy-3-labeled Annexin V, and subjected to flow cytometric analysis. infection indicate that the cagA-positive strain is associated with an increase in apoptosis (44 -47). Furthermore, apoptosis was found to be increased in the stomach in patients with cagA-positive, but not cagA-negative, H. pylori strains (48), although the results of some studies have not supported the notion of a pro-apoptotic role of cagA-positive H. pylori (49,50). Again, the conflicting results may be due to methods used to determine cagA status. Host genetic differences may also play a role in determining cellular response to H. pylori.
A moderate but continuous induction of apoptosis by CagA may underlie the elevated epithelial cell turnover associated with cagA-positive H. pylori infection (44 -50). Importantly, extra rounds of DNA replication in gastric cells would increase the chance of genetic mutations. In particular, mutations in genes, such as p53, that evade apoptosis may change the host cell response to CagA⅐SHP-2 signaling from apoptosis to deregulated proliferation, a cellular situation in which further mutations in oncogenes and tumor suppressor genes may progressively accumulate. Indeed, in vivo studies have indicated that H. pylori infection is associated with acquired genetic instability and induction of p53 point mutations in patients with chronic gastritis (51,52). Accordingly, cellular response to CagA, either apoptosis or proliferation, may depend on the p53 status of host cells.