Professor Tsutomu Chiba, Department of Gastroenterology and Hepatology, Graduate School of Medicine, Kyoto University, Kyoto 606-8507, Japan. Email: firstname.lastname@example.org
Helicobacter pylori (H. pylori) infection plays a crucial role in the development of gastric cancer. There are two major pathways for the development of gastric cancer by H. pylori infection: the indirect action of H. pylori on gastric epithelial cells through inflammation, and the direct action of the bacteria on epithelial cells through the induction of protein modulation and gene mutation. Both pathways work together to promote gastric carcinogenesis.
Since the discovery of Helicobacter pylori (H. pylori) by Warren and Marshall in 1982,1 many epidemiological studies have revealed a strong association between H. pylori infection and gastric cancer development.2–4 As confirmation, recent retrospective and prospective studies have demonstrated that H. pylori-positive patients have a significantly higher risk of gastric cancer than H. pylori-negative patients.5,6 Careful investigations have shown more than 95% positivity for H. pylori infection in gastric cancer patients. Supporting those human data, many investigators have succeeded to develop gastric cancer in H. pylori-infected Mongolian gerbils.7–9 Thus it is evident that H. pylori is a strong causative factor for gastric cancer. However, the precise mechanism for gastric cancer development by H. pylori infection is still not completely understood.
Direct and indirect actions of H. pylori on gastric epithelial cells during gastric carcinogenesis
Similar to hepatitis C virus or hepatitis B virus-induced hepatocarcinogenesis, there are two major pathways for the development of gastric cancer by H. pylori infection: the indirect action of H. pylori on gastric epithelial cells through inflammation, and the direct action of the bacteria on epithelial cells (Fig. 1). Studies have shown the importance of gastritis in the development of gastric cancer. However, it is also established that H. pylori directly modulates epithelial cell function by bacterial agents, such as CagA.10–14 Although the relationship between the two pathways remains unclear, both pathways appear to work together to promote gastric cancer development.
Indirect action of H. pylori on gastric epithelial cells through the induction of gastritis in gastric carcinogenesis
H. pylori-positive gastric cancer is invariably associated with gastritis, and it has been shown that a patient's risk for gastric cancer is proportional to the severity of gastritis, particularly chronic atrophic gastritis of the corpus. Although the prevalence of H. pylori infection is similarly high in Asian countries, the incidence of gastric cancer in southern Asia is lower than in northern Asia, and this difference is reflected by different degrees of gastritis between the two regions.15 Indeed, in north-east Asian countries, including Japan and Korea, H. pylori infection induces inflammation not only in the antrum, but also in the corpus in association with mucosal atrophy, whereas in south Asian countries, gastritis is restricted to the antrum, and the degree of corpus gastritis is low.15 In animals, gastric cancer development by H. pylori infection is associated with severe gastritis.
Interestingly, several lines of evidence have indicated that gastric ulcers (GU) are associated with gastric cancer development, while the reverse is the case for duodenal ulcers (DU).15 In Japan, where gastric cancer is prevalent, the DU/GU ratio is less than 1.16 In contrast, the DU/GU ratio is higher in southern Asia and Western countries where the incidence of gastric cancer is low.17,18 In support of this contrasting relationship between gastric cancer and DU versus gastric cancer and GU, Uemura et al.5 reported an 8-year follow-up of Japanese patients with H. pylori infection where none of the patients with DU developed gastric cancer, but 3.4% of patients with GU developed gastric cancer. In this regard, it should be noted that patients with DU generally have antrum-predominant gastritis (type B gastritis) with little mucosal atrophy of the corpus, whereas patients with GU usually have corpus-predominant gastritis (type AB gastritis) with various degrees of mucosal atrophy. Thus the inverse relationship between gastric cancer development and the DU/GU ratio appears to reflect the different extent of corpus gastritis between DU and GU, again emphasizing the importance of corpus gastritis in the development of gastric cancer.
In H. pylori-induced chronic gastritis mucosa, the number of CD4T cells predominates over other cell types.19 CD4T cells appear to be essential for the development of gastritis because gastritis does not develop in mice lacking CD4T cells.20 Tissue-infiltrating CD4T cells are classified as Th1 and Th2 type, which are characterized by the secretion of interferon (IFN)-γ and interleukin (IL)-4, respectively. Along with others, we have demonstrated that T cells in H. pylori-infected gastric mucosa produce a large amount of IFN-γ together with a minute amount of IL-4 on antigenic stimulation.19–22 Moreover, the development of gastritis by H. pylori infection is impaired in IFN-γ-deficient mice, whereas IL-4 deficiency enhances the development of gastritis.22 Thus Th1-type CD4T cells and their product, IFN-γ, appear to have crucial roles in the development of gastritis. To examine whether these different immune responses (Th1 and Th2) have roles in determining the extent of gastritis, we investigated those responses in GU and DU patients, which developed in the background of pangastritis (type AB gastritis) and antral-predominant gastritis (type B gastritis), respectively. T cells from GU patients produced greater amounts of IFN-γ and less IL-4 than those in DU patients. Moreover, GU patients had higher serum levels of immunoglobulin G (IgG)2, Th1-type IgG, specific to H. pylori, than DU patients.21 These data suggest that a Th1-type immune response accelerates the progression of corpus gastritis with mucosal atrophy, thus leading to the development of GU and gastric cancer. Our data showing IL-4 gene diplotypes having a significant, negative association with gastric cancer development supports such an idea.23 Recently, in addition to Th1- and Th2-type immune responses, the Th17-type immune response has been the focus of attention. Although IFN-γ appears to be essential for the development of corpus gastritis, the roles of the IL-17-type immune response in gastritis as well as gastric cancer development need to be examined.
How H. pylori induces the Th1-type response remains unknown. In this regard, recent studies revealed that Peyer's patches in the small intestine play a crucial role in the production of H. pylori-specific CD4T cells, their migration into the gastric mucosa, and eventually the development of gastritis.24,25 It was found that in the absence of Peyer's patches Helicobacter failed to develop gastritis despite of significant colonization in the gastric mucosa. Moreover, although the transfer of splenic T cells of Helicobacter-infected mice to Helicobacter-infected Rag2 knockout mice induced severe gastritis, splenic T cells from Helicobacter-infected Peyer's patch-deficient mice failed to produce gastritis in Rag2 knockout mice despite the presence of a large number of Helicobacter in the gastric mucosa. These data show the importance of Peyer's patches and also the adaptive immunity against Helicobacter induced by Peyer's patches in the development of gastritis.25 It has also been shown that those H. pylori-specific T cells are responsible for the infiltration of neutrophils in the gastric mucosa.26 Whether Peyer's patches are also important for dictating Th1-type immune response in the gastric mucosa is an interesting issue to be examined.
In addition to the infiltration of T cells and activated mononuclear cells, H. pylori-induced gastritis is characterized by the enhanced production of a variety of pro-inflammatory cytokines in the gastric mucosa, and several cytokines are suggested to play important roles in cancer development. Among them, investigators have focused on IL-1β as the factor that links gastric inflammation and gastric cancer. El-Omar et al.26 reported an association between an IL-1β gene polymorphism and an increased risk of gastric cancer due to H. pylori infection. Other studies have supported their findings: IL-1β expression is enhanced by H. pylori infection,27 and is not only an inflammatory mediator enhancing nuclear factor-κB (NF-κB) activation in both inflammatory and epithelial cells, but also a potent inducer of hepatocyte growth factor. Thus IL-1β is now considered to be an important factor that is significantly involved in the development of gastric cancer by H. pylori infection. Supporting this hypothesis, a recent study demonstrated the development of gastric cancer in IL-1β transgenic mice.28 However, there is still no definitive data demonstrating that polymorphism of the IL-1β gene influences the production of the IL-1β protein. Moreover, the data from Asian countries, including ours, do not confirm El-Omar's data, suggesting that the impor tance of IL-1β in gastric cancer development is different among different ethnic groups.23,29,30
In addition to IL-1β, the production of other cytokines such as tumor necrosis factor-α (TNF-α) IL-6, IL-7, and IL-8 is also enhanced in the gastritis mucosa,31 and both IL-1β and TNF-α enhance NF-κB activation. Recently, the important roles of NF-κB in inflammation-associated carcinogenesis have been the focus of attention. In colitic cancer models, NF-κB is a key factor for colitis-associated cancer development.32 At present, several possible roles of NF-κB are being considered in the development of gastric cancer with a background of gastritis. NF-κB in the epithelial cells may exert an anti-apoptotic action. The activation of NF-κB in immune cells may accelerate gastric inflammation by enhancing the production of various cytokines. Moreover, NF-κB activation in epithelial, inflammatory, and mesenchymal cells may enhance cyclooxygenase-2 production, increasing the risk of development of gastric cancer in the gastric mucosa. Finally, our recent finding suggested the involvement of NF-κB in H. pylori-induced mutagenesis.33
IL-6 and IL-11 are also increased in the H. pylori-induced gastritis mucosa. Recently, Giraud et al. demonstrated the development of gastric cancer by the constitutive activation of signal transducers and activators of transcription-3 (STAT3) in mice, suggesting the importance of IL-6 or IL-11 signaling in the development of gastric cancer.34,35 STAT3 activation accelerate nuclear localization of β-catenin.36 We recently demonstrated that H. pylori infection enhances the expression of RegIα, a potent growth factor for the gastric mucosa in gastric epithelial cells.37 Interestingly, the expression of RegIα in gastric mucosal cells is not stimulated directly by H. pylori, but by IL-6 in a STAT3-dependent manner.38 Thus IL-6 appears to accelerate epithelial cell growth not only directly, but also indirectly by stimulating RegIα production through STAT3 activation in epithelial cells. We also showed that in addition to IL-6, RegIα production is stimulated by IFN-γ, a representative Th1 cytokine.39 These cytokines appear to be coordinately involved in gastric cancer development.
Direct action of H. pylori on gastric epithelial cells
In addition to the indirect pathway for the development of gastric cancer by H. pylori infection through the induction of inflammation, data have suggested that H. pylori promotes gastric carcinogenesis by acting directly on gastric epithelial cells (Fig. 2). In vitro studies using human gastric epithelial cells have demonstrated that H. pylori acts directly on gastric epithelial cells to modulate various cellular functions. For instance, H. pylori infection to gastric epithelial cells or the introduction of various H. pylori genes into gastric epithelial cells modulates cellular growth, apoptosis, or cell migration, and induces the hummingbird phenomenon characterized by an extremely elongated cell shape. H. pylori or products of H. pylori may act on cellular membranes, and some of the effects may be mediated by cell surface receptors, such as Toll-like receptors (TLR). However, several investigators have shown that TLR are not important in epithelial cell responses to H. pylori.40,41 Recent studies have demonstrated that a bacterial type IV secretion apparatus, encoded by the H. pylori cag pathogenicity island (cagPAI), plays an essential role in mediating many of the direct actions of H. pylori on epithelial cells by delivering bacterial agents into the cells. It was found that the CagA protein exerts its action on epithelial cells by entering the cells through this structure.12,42,43
Roles of CagA in gastric carcinogenesis
H. pylori has been subdivided into cagA-positive and cagA-negative strains. The cagA-positive strains are considered to be more potent in inducing mucosal damage and developing atrophic gastritis than cagA-negative strains.44 Clinical studies have shown a strong association between infection with cagA-positive H. pylori and gastric cancer development.45,46 Moreover, gastric cancer patients are almost invariably infected with cagA-positive strains. Accordingly, many investigators have been interested in CagA in cancer development.
As described, CagA is delivered into gastric epithelial cells via the bacterial type IV secretion apparatus, where it undergoes tyrosine phosphorylation by Src family kinases or the Abl kinase at the EPIYA motifs that are present in the C-terminal region.12,13,43,47 Subsequently, the delivered CagA binds to and activates Src homology 2-containing protein tyrosine phosphatase (SHP2),13,43 and the CagA-deregulated SHP2 exerts various functions, such as the activation of the extracellular regulated kinase and mitogen-activated protein kinase cascade,13 the modulation of the focal adhesion kinase with resulting induction of the hummingbird phenomenon,14,48 and the inhibition of Src family kinases by activating the C-terminal Src kinase, generating a feedback regulation loop for the tyrosine phosphorylation cascade. CagA also binds to Crk inducing cell scattering.49 In contrast to such a pathway involving the tyrosine phosphorylation of CagA, it has also been recently shown that CagA deregulates Grb2 and c-Met in a phosphorylation-independent manner.50,51 Moreover, CagA has been shown to activate the nuclear factor of activated T cells by stimulating calcineurin, regardless of its phosphorylation status.52 More recently, CagA has been shown to impair cell–cell adhesion independently of CagA tyrosine phosphorylation by disrupting tight junctions and causing a loss of cell polarity by inhibiting the PAR1/MARK polarity-regulating kinase.53 CagA was also shown to destabilize the E-cadherin–β-catenin complex with the resulting activation of β-catenin signal.54 All of these actions of CagA may contribute to gastric cancer development by providing epithelial cells with a suitable environment for neoplastic transformation.
One important thing to note, however, is that many of the data on the direct actions of CagA were obtained from in vitro studies by transfecting the cagA gene to epithelial cells, but not by H. pylori infection. Thus the possibility remains that some of the known, direct actions of CagA on epithelial cells are not physiological, and thus have been overestimated. Accordingly, the roles of CagA should be examined by H. pylori infection in vitro.
NF-κB activation by H. pylori
Another well-known direct action of H. pylori on gastric epithelial cells is the activation of NF-κB.55In vitro studies have shown that only H. pylori strains with functional cagPAI induce the activation of NF-κB. Importantly, although CagA translocation and its phosphorylation are required for the appearance of the hummingbird phenomenon, these events are dispensable for H. pylori induction of NF-κB activation in epithelial cells, suggesting the presence of unknown H. pylori factors that are translocated into epithelial cells via type IV secretion apparatus and then activate NF-κB.40,56 However, subsequent studies with various mutated cagPAI genes failed to identify a candidate molecule within the cagPAI gene.56 Viala et al.57 suggested H. pylori-derived peptidoglycan as a candidate molecule. They demonstrated that H. pylori-derived peptidoglycan is delivered into epithelial cells through type IV secretion apparatus, which is recognized by NOD1, an intracellular pathogen-recognition molecule with specificity for Gram-negative peptidoglycan that eventually activates NF-κB. Hirata et al.,58 however, showed that cagPAI-dependent NF-κB activation in epithelial cells by H. pylori involves Myd88, a molecule known to be located downstream from TLR. The data suggest another H. pylori molecule that activates NF-κB.
Thus it is evident that the direct action of H. pylori on epithelial cells involves both CagA-dependent and CagA-independent mechanisms, and that NF-κB may be a molecule representing the CagA-independent pathway (Fig. 2). In this regard, almost all the cagPAI-positive clinical isolates of H. pylori have cagA. However, it should be noted that the functional significance of cagPAI and cagA is different, and it is now clear that some of the actions of cagPAI-positive H. pylori are not attributed to CagA. Thus the roles of CagA and the type IV secretion apparatus composed of the gene products encoded by cagPAI in gastric carcinogenesis should be separately discussed.
Induction of gene mutations by H. pylori
Cancer development is characterized by the accumulation of gene modulations, including mutations. Various gene mutations, such as those in TP53, are present in gastric cancers; the TP53 gene mutation is present in nearly 50% of gastric cancers.59 Previous clinical studies have shown that mutations of TP53 are already present in the H. pylori-infected chronic gastritis mucosa.60,61 Moreover, animal experiments using Big Blue mice revealed a high mutation frequency in the H. pylori-infected gastric mucosa.62 Thus H. pylori infection appears to induce gene mutations in gastric mucosal cells. It is possible that H. pylori-induced gastritis may enhance gene mutations by producing mutagens, such as radical oxygen species; however, whether H. pylori directly affects epithelial cells to induce gene mutations is unknown. Recently, we demonstrated that H. pylori directly induces gene mutation by enhancing the expression of activation-induced cytidine deaminase (AID) in gastric mucosal cells.34 AID, a member of the cytidine deaminase family that acts as an editor of DNA and RNA, is essential for somatic hypermutation and class-switch recombination of immunoglobulin genes in B lymphocytes.63 Importantly, AID is expressed exclusively in B cells under physiological conditions. We first found that the constitutive expression of AID in transgenic mice resulted in gastric cancer development with high mutation frequencies, suggesting that AID played a role in gastric carcinogenesis by inducing gene mutations. Surprisingly, we also found that AID was ectopically expressed in the H. pylori-induced gastritis mucosa as well as gastric cancer tissues in humans, and the eradication of H. pylori could reduce its expression. Moreover, in vitro studies using human gastric cancer cells revealed that H. pylori infection elicited the ectopic expression of AID in association with a high mutation frequency of the TP53 gene. AID expression in B cells is stimulated with CD40 ligation by T cells via a NF-κB-dependent mechanism.64 Therefore, we also examined the NF-κB dependency of the H. pylori-induced AID expression, and found that the H. pylori-induced AID expression in gastric epithelial cells is also mediated by NF-κB activation. AID expression in gastric epithelial cells is elicited only by cagPAI-positive H. pylori that can activate NF-κB. Finally, we showed that H. pylori-induced TP53 mutation in vitro could be inhibited by blocking AID using AID siRNA. Taken together, these data strongly suggested that H. pylori directly induces gene mutations in epithelial cells by enhancing the AID expression through NF-κB activation (Fig. 2). Since NF-κB is activated by cytokines, such as IL-1β and TNF-α, it is possible that the AID expression is also induced indirectly by these cytokines in the gastritis mucosa. Of note, the most prevalent base substitution observed in human gastric cancer is a cytidine to thymidine transition,65 and AID theoretically induces a cytidine to thymidine transition. Interestingly, although the frequency of cytidine to thymidine transitions was the highest in our in vitro experiments, there were also other mutations. The reason for this unexpected observation remains to be clarified.
Induction of aberrant DNA methylation by H. pylori
In addition to the induction of gene mutations, recent studies have suggested that H. pylori infection enhances aberrant DNA methylation in the gastric mucosa and that such methylation may participate in gastric carcinogenesis through silencing tumor suppressor genes. Ushijima et al. demonstrated enhanced DNA methylation in H. pylori-infected non-cancerous gastritis mucosa.66 Interestingly, they observed the same pattern of DNA methylation in non-cancerous mucosa of gastric cancer patients with and without current H. pylori infection, suggesting that the DNA methylation signature persists even after the disappearance of H. pylori.67 They also found that H. pylori infection results in DNA methylation of the genes common in Mongolian gerbils and in humans, suggesting gene specificity for aberrant DNA methylation by H. pylori infection (Dr T. Ushijima, pers. comm., 2007). It is unclear whether the aberrant DNA methylation results from a direct action of H. pylori on epithelial cells or from H. pylori-induced inflammation.
Relationship between direct and indirect actions of H. pylori on epithelial cells in gastric cancer development
At an early stage of gastritis, when a large number of H. pylori is present, H. pylori has both direct and indirect effects on epithelial cells. H. pylori accelerates gastritis not only by presenting bacterial antigens at Peyer's patches of the intestine, but also by stimulating innate immunity within the stomach. At the same time, H. pylori acts directly on gastric epithelial cells to induce gene mutations by NF-κB activation and enhance cell growth, inhibit or promote apoptosis, and modulate cell adhesion and migration through influencing various intracellular signaling cascades by bacterial agents. It should be emphasized that the direct actions should be targeted to the stem cells or the progenitor cells in the isthmus of the gastric gland because it appears difficult for the differentiated cells to be transformed to cancer cells. Thus a large number of H. pylori may be required to exert direct effects on undifferentiated cells located at the isthmus. At a late stage of gastritis, the number of H. pylori usually decreases or even disappears with the progression of atrophic gastritis in association with the development of intestinal metaplasia. At this stage, gastritis activity is decreased, but the continuous supply of H. pylori-specific T cells from the intestine may persist, even with a decreased number of H. pylori, contributing to the continuation of chronic gastritis. In contrast, with the decrease in the number of H. pylori in the gastric mucosa, it is likely that the direct influence of H. pylori on epithelial cells is reduced, and consequently, the signal transduction events in the epithelial cells evoked by H. pylori may become inactive. Thus at a late stage of H. pylori infection, the direct actions of H. pylori on epithelial cells do not appear to be important. However, once the cells, particularly stem cells or progenitor cells, acquire gene mutations by AID at an active stage of gastritis, the mutated cells persist until this stage, and eventually become malignant (Fig. 3).
Prevention of H. pylori-induced gastric cancer
As discussed, H. pylori plays crucial roles in the development of gastric cancer. Therefore, the best strategy for preventing gastric cancer development is the eradication of H. pylori. Although H. pylori eradication has been clearly shown to decrease the risk for cancer development, some patients develop gastric cancer even after H. pylori eradication.68 Since signatures for gastric cancer development, such as gene mutations, are likely to develop at an early stage of gastritis, eradication at an early stage of H. pylori infection is strongly recommended.