Orchestration of Aberrant Epithelial Signaling by Helicobacter pylori CagA

Although colonization of humans by pathogenic bacteria is common, disease follows in only a fraction of infected persons. Helicobacter pylori is a Gram-negative bacterial species that colonizes human gastric epithelium for decades, in part because of its ability to limit the bactericidal effects of proinflammatory molecules (1), to vary its antigenic repertoire of surface-exposed proteins (2), and to suppress actively the adaptive immune response (3-5). Microbial persistence implies a physiological linkage between host and microbe in which signals of the colonizing organism affect signals of the host, and because of its coevolution with humans, H. pylori can send and receive signals from gastric epithelium. However, there are biological costs to the long-term relationship between H. pylori and humans, in that chronic colonization confers an increased risk of distal gastric cancer (6-9). Despite the designation of H. pylori as a class I carcinogen for gastric cancer, only a fraction of colonized persons ever develop neoplasia; enhanced cancer risk is related to H. pylori strain differences, inflammatory responses governed by host genetic diversity, specific interactions between host and microbial determinants, or some combination of these factors. These observations underscore the importance of identifying signaling pathways that regulate biological interactions of these organisms with their hosts--pathways that promote gastric carcinogenesis. Results generated by such studies would permit physicians to focus H. pylori diagnostic and eradication strategies on identified high-risk populations to optimize prevention of subsequent neoplastic events.

H. pylori strains from different individuals are extremely diverse, because of point mutations and gene insertions and deletions (10-12); indeed, genetically unique derivatives of a single strain are present simultaneously within an individual human host (13). Although this extraordinary diversity has hindered characterization of the bacterial factors associated with diseases such as gastric cancer, genetic loci have been identified for which particular variants are associated with different risks of disease.

The cag pathogenicity island is a 40-kb locus that is present in ~60% of H. pylori strains isolated in the United States (10, 11). Although all H. pylori strains induce gastritis, cag⁺ strains augment the risk for severe inflammation, peptic ulcer disease, and distal gastric cancer compared with that after infection with cag^- strains (14-18). The cag island encodes a type IV secretion system (the class of bacterial secretion system with the capacity to deliver macromolecules intracellularly), and peptidoglycan is delivered by the cag secretion system into host cells, where it is recognized by intracellular Nod1, which subsequently activates nuclear factor κB (NF-κB) (19). Another substrate of this secretion system is the product of the terminal gene in the island, CagA, which is translocated into epithelial cells after bacterial attachment, where it undergoes tyrosine phosphorylation by members of the Src family of kinases (20-24) (Fig. 1). Phospho-CagA subsequently binds to and activates a eukaryotic phosphatase (SHP-2, for Src homology-2 domain-containing protein-tyrosine phosphatase), which thereby induces a cellular elongation (hummingbird) phenotype (23) (Fig. 1), and this process of CagA-dependent SHP-2 activation mirrors events within inflamed mucosa as CagA becomes phosphorylated and binds SHP-2 within gastric epithelial cells in vivo (25). SHP-2 contains two tandem SH2 domains, and binding of phosphorylated proteins, such as phospho-CagA, to either or both of these sites stimulates its phosphatase activity (24). SHP-2 is an integral component in various intracellular pathways that convey signals transduced by growth factors and cytokines regulating proliferation, morphogenesis, and motility (24). The cytoskeletal derangements induced by CagA-dependent activation of SHP-2 are similar to morphogenetic alterations induced by growth factors. The ability of CagA to activate SHP-2 aberrantly has provided insights into mechanisms through which H. pylori cag⁺ strains may initiate carcinogenesis, because altered SHP-2 signaling induced by manipulation of the interleukin-6 (IL-6) coreceptor gp130 culminates in the development of gastric adenocarcinoma in genetically engineered mice (26).

• Download high-res image
• Open in new tab
• Download Powerpoint

Although cag⁺ strains are disproportionately represented among infected hosts who develop gastric cancer, and genes within the cag island are necessary for induction of pathogenic epithelial responses, most persons colonized by these strains remain asymptomatic. This paradox has fueled efforts to define more clearly the mechanisms that may underpin differences in the biological activity of CagA proteins present in different H. pylori strains. Among H. pylori strains harvested from individuals residing in Western countries where the prevalence of gastric cancer is low, the number of CagA tyrosine phosphorylation motifs (TPMs) can vary substantially, and SHP-2 binding affinity and induction of cellular elongation are potentiated by an increasing number of these motifs (27) (Fig. 1). Moreover, H. pylori strains with an increased number of Western-type CagA TPMs are more closely associated with gastric cancer (28, 29). In contrast, in regions where gastric cancer rates are high, such as East Asia, CagA TPMs are distinct from Western-type TPMs, and sequences flanking the major East Asian-type TPM perfectly match the consensus SH2 binding site for SHP-2 (27). As would be predicted, binding of SHP-2 and morphogenetic activity are induced more potently by CagA proteins containing East Asian-type compared with Western-type TPMs (Fig. 1), which may explain, in part, the strikingly different rates of gastric cancer in these regions.

In addition to stimulating SHP-2 activity, phospho-CagA also activates C-terminal Src kinase, which inhibits the activity of Src, leading to a reciprocal decrease in the intracellular level of phosphorylated CagA (30, 31). Although this negative feedback loop likely contributes to the long-term equilibrium between H. pylori and its host, unphosphorylated CagA can also exert effects within the host cell that influence pathogenesis. For example, translocation, but not phosphorylation, of CagA leads to disruption of apical junctional complexes in polarized epithelial cells and a loss of cellular polarity, alterations that play a role in carcinogenesis (32). Unmodified CagA can bind to growth factor receptor-binding protein 2 (Grb2), which results in activation of the Ras mitogen-activated or extracellular signal-regulated protein kinase kinase (MEK)-extracellular signal-regulated kinase (ERK) mitogen-activated protein kinase (MAPK) pathway (33), and ERK activation regulates pathogenic epithelial responses, including secretion of the proinflammatory cytokine IL-8 and of MMP-7, an epithelial cell-derived matrix metalloproteinase with tumor-initiating properties (34, 35). CagA binds the c-Met receptor (a receptor tyrosine kinase implicated in tumor progression and metastasis) in a phosphorylation-independent manner, which stimulates a motogenic response in gastric epithelial cells termed the cell-scattering phenotype that is also dependent on ERK signaling (36). These results suggest that CagA may function as a pivotal docking molecule that coopts different host effectors to deregulate multiple signaling pathways; however, the relationships between CagA and intracellular signaling pathways have not been completely defined. SHP-2, which clearly binds phosphorylated CagA, can function in parallel to Ras to activate ERK, and indeed, SHP-2 is required for maximal and sustained ERK activation in other cell systems (24). These observations raise the question as to whether unmodified or phosphorylated CagA is the primary inducer of ERK-mediated cellular derangements that are related to disease outcome.

In a recent article that addresses this question, Higashi et al. developed gastric epithelial cells that stably expressed either wild-type or phosphorylation-resistant CagA under the control of an inducible promoter in order to define more precisely the morphologic and molecular responses that depend on tyrosine phosphorylation of CagA (37). Induction of wild-type, but not phosphorylation-resistant, CagA reproducibly induced hummingbird and cell-scattering phenotypes identical to changes induced by transient expression of wild-type CagA (37). These results were confirmed more rigorously by treating CagA-expressing cells with a chemical inhibitor of Src (the kinase responsible for CagA phosphorylation) or transiently transfecting gastric cells with a panel of phosphorylation-resistant CagA constructs; under each of these conditions, morphogenetic responses were abolished (37). The role of SHP-2 in this process was established by using small interfering RNA to silence expression of SHP-2 in CagA-expressing cells, which correspondingly inhibited cellular elongation and migration (37).

Having shown that Src-dependent phosphorylation of CagA and concomitant SHP-2 activation were required for epithelial cytoskeletal derangements, the authors then investigated the role of ERK activation. CagA-expressing cells pretreated with an inhibitor of ERK, but not inhibitors of p38, c-jun N-terminal kinase (JNK), or phosphoinositide 3-kinase (PI3K), failed to develop a hummingbird phenotype, which specifically implicated ERK in phospho-CagA-SHP-2-dependent alterations of cellular morphology (37). Expression of wild-type CagA and SHP-2 was also shown to prolong ERK activation. Finally, to define the role of Ras in this cascade, constitutively active and dominant-negative Ras constructs were transfected into CagA-expressing cells, and although mutant Ras proteins successfully interfered with Raf1 phosphorylation, neither construct had any effect on CagA induction of the hummingbird phenotype (37). Collectively, these findings indicate that SHP-2 is involved in Ras-independent modification of ERK, which, in turn, is necessary for the morphogenetic activity of CagA (Fig. 1).

Detailed mechanistic studies such as this from Higashi et al. are extremely valuable, primarily because of the importance of H. pylori as a human pathogen. About 650,000 people worldwide will die this year alone from gastric cancer, and 5-year survival rates in the United States are 15%. Establishment of H. pylori as a risk factor for gastric cancer permits an approach to identify people at increased risk; however, infection with this organism is extremely common and most colonized people never develop cancer. Thus, techniques to identify high-risk subpopulations must use other biological markers. Because both strain genotypes and induced host responses influence the risk for carcinogenesis by differentially affecting epithelial cell physiology, molecular delineation of intracellular pathways activated by H. pylori-epithelial cell interactions will not only improve our understanding of H. pylori-induced gastric carcinogenesis, but will also facilitate identification of potential therapeutic targets for prevention and more effective treatment of this disease.