Helicobacter pylori strains that possess the cag pathogenicity island induce more severe gastritis and augment the risk of developing peptic ulcer disease and distal gastric cancer. A specific mechanism by whichcag + strains may enhance gastritis is strain-selective regulation of interleukin (IL)-8 production. On contact with gastric epithelial cells, H. pylori activates multiple signal transduction cascades that regulate IL-8 secretion, including nuclear factor-κB and mitogen-activated protein kinases, and these events are dependent on genes within the cagisland. An independent effect of cag-mediated cellular contact is translocation and phosphorylation of H. pyloriproteins within the host epithelial cell. The redundancy of intracellular signaling cascades activated by H. pylori and the divergent epithelial cell responses induced by components of thecag island may contribute to the ability of this organism to persist for decades within the gastric niche.
- nuclear factor-κB
- mitogen-activated protein kinases
the gastrointestinal tract represents an important barrier between human hosts and microbial populations. One potential consequence of host-microbial interactions is the development of mucosal inflammation. Regardless of the initiating event, if allowed to become persistent, inflammatory states may lead to clinically apparent disease. A paradigm for such chronic host-microbial relationships is carriage of Helicobacter pylori, Gram-negative bacteria that colonize the stomachs of humans and primates. H. pylori colonization induces chronic gastritis in essentially all hosts, a process that increases the risk of developing peptic ulceration, distal gastric adenocarcinoma, and gastric mucosal lymphoproliferative disease (Fig.1). However, only a small percentage of persons carrying H. pylori develop clinical sequelae; enhanced risk may be related to differences in expression of specific bacterial products, to variations in the host inflammatory response to the bacteria, or to specific interactions between host and microbe. The recent demonstration that a subset of H. pylori strains are associated with a reduced risk of developing esophageal adenocarcinoma (28) further underscores the importance of identifying mechanisms that predispose colonized individuals toward disease.
Certain H. pylori components such as urease and porins are ubiquitous and necessary for colonization and survival, and variation in clinical outcome is unlikely to develop as a result of these highly conserved traits. Genes that are heterogeneously represented among H. pylori strains, however, may encode candidate virulence factors that influence the pathological course. One specific phenotype shown to differ among H. pylori isolates is production of an immunodominant protein encoded by the genecagA (3), which is present in ∼60% of US strains. Persons colonized with cagA + strains are at increased risk of developing more severe gastritis, peptic ulcer disease, and distal gastric cancer compared with persons harboringcagA − strains. cagA is a component of and a marker for a 40-kb region of chromosomal DNA acquired by horizontal transfer called the cag pathogenicity island, which is inserted at the 3′-terminus of the glutamate racemase gene (3). Several cag island genes possess homology to components of a type IV secretion system that, in other prokaryotic species, functions as a conduit for export of multimeric proteins and nucleoproteins across both the inner and outer bacterial membrane. The H. pylori cag island is required for both translocation of bacterial proteins into host cells (2, 20,26) and induction of proinflammatory cytokine release (3). Thus cagA + strains are disproportionately represented among persons who develop serious sequelae of H. pylori infection, and genes within thecag island are necessary for induction of epithelial cell responses relevant to pathogenesis.
The mucosal inflammatory infiltrate that develops in response toH. pylori consists of neutrophils, lymphocytes, plasma cells, and macrophages. The presence of both acute and chronic inflammatory components within colonized mucosa suggests that soluble mediators capable of attracting cells derived from varying lineages may be key regulators in the development of gastritis. Interleukin (IL)-8 is a potent neutrophil- and lymphocyte-activating chemotactic cytokine (chemokine), and gastrointestinal epithelial cells secrete biologically activated IL-8 in response to infection with pathogenic bacteria (7). Chemokines produced by activated enterocytes bind to the extracellular matrix, thereby establishing a chemotactic gradient that directs inflammatory cell migration toward the epithelial cell surface. IL-8 expression is enhanced within H. pylori-colonized mucosa (6, 21), and increased IL-8 protein is primarily localized to gastric epithelial cells (6). Carriage of cagA + strains further augments mucosal IL-8 expression, and such increases are directly related to the more severe inflammatory response induced by these strains (21). In vitro H. pyloristimulates IL-8 expression in gastric epithelial cells, and these events require an active interplay between live bacteria and host cells (25). Similar to findings within gastric tissue,cagA + strains induce significantly higher levels of IL-8 in vitro than cagA − strains (4). Thus a paradigm for H. pylori-induced gastric inflammation is that contact between bacteria and epithelial cells stimulates IL-8 secretion that then regulates neutrophilic and monocytic infiltration into the gastric mucosa. Because cytokine production and severity of gastric inflammation are enhanced bycagA + strains, this themes article will focus on strain-specific activation of molecular signaling events that regulate IL-8 expression as a mechanistic model for H. pylori-induced gastritis.
H. PYLORI-INDUCED IL-8 EXPRESSION IS MEDIATED BY ACTIVATION OF NUCLEAR FACTOR-κB
The human IL-8 gene contains several binding sites within its promoter region (Fig. 2). A nuclear factor (NF)-κB binding motif is located at nucleotides (nt) −80 to −70, and a NF-IL-6 site lies immediately adjacent to this motif (nt −94 to −81). In addition to these loci, a binding site for c-fos and c-jun, which together comprise the transcription factor AP-1, is present at nt −126 to −120 (Fig. 2).
NF-κB is a transcription factor sequestered in the cytoplasm, whose activation and regulation are tightly controlled by a class of inhibitory proteins termed IκBs (IκBα, IκBβ, and IκBε). Through noncovalent association, IκB proteins mask the nuclear localization signals of NF-κB, thereby preventing movement of NF-κB to the nucleus. On stimulation with signaling molecules such as tumor necrosis factor (TNF)-α, phosphorylation of IκBα and IκBβ leads to the ubiquitination and 26S proteosome-mediated degradation of phospho-IκBα, thereby liberating NF-κB to enter the nucleus, where it regulates transcription of a variety of genes including those involved in inflammation (Fig. 3). Two cytokine-inducible kinases, IκB kinase-α (IKK-α) and -β (IKK-β), have recently been characterized, and these enzymes phosphorylate IκBα in response to proinflammatory cytokines (16). An upstream mediator of these events is NF-κB-inducing kinase (NIK), a novel member of the mitogen-activated protein kinase/extracellular signal-regulated kinase kinase (MEKK) family, which has been shown to activate both IKK-α and IKK-β through interaction with adaptor proteins associated with receptors for TNF-α and IL-1 (15). The adaptor proteins TRAF2 and TRAF6 belong to the TNF receptor-associated factor (TRAF) family and act as effectors for activated TNF-α and IL-1 receptors, respectively (Fig. 3) (14).
Stimulation of NF-κB does not require protein synthesis, thereby allowing efficient activation of target genes such as IL-8. This system is particularly utilized in immune, inflammatory, and acute phase responses, where rapid activation of defense genes after exposure to pathogens is critical for survival of an organism. Several studies demonstrated that contact between H. pylori and gastric epithelial cells results in brisk activation of NF-κB that is followed by increased IL-8 mRNA and protein expression (11,24). The ability of H. pylori to activate NF-κB in vitro has been corroborated in vivo because activated NF-κB is present within gastric epithelial cells of infected but not uninfected patients (11), and this site of localization mirrors the pattern of increased IL-8 protein levels within colonized mucosa. Recently, Maeda et al. (14) identified upstream mediators that regulate H. pylori-induced NF-κB-dependent IL-8 production. Using in vitro transfections, these investigators demonstrated that H. pylori-mediated NF-κB activation is inhibited by IKK-α and IKK-β kinase-deficient mutants, indicating that IκBα is degraded via activation of IKK-α and IKK-β duringH. pylori coculture (14). Kinase-deficient NIK as well as dominant-negative mutants of the upstream adaptor proteins TRAF2 or TRAF6 also inhibit the ability of H. pylori to activate NF-κB (14). These results have established a hierarchical series of events culminating in NF-κB activation in which H. pylori activates NIK via TRAF2 and TRAF6, which, in turn, phosphorylates and activates IKK-α and IKK-β. The activated IKKs then phosphorylate IκBα, leading to its proteosome-mediated degradation, with subsequent release and nuclear translocation of NF-κB and induction of IL-8 (Fig. 3).
REGULATION OF H. PYLORI-INDUCED IL-8 PRODUCTION BY MITOGEN-ACTIVATED PROTEIN KINASE CASCADES
Although the studies described have identified intracellular intermediaries through which H. pylori activates NF-κB and induces IL-8, recent investigations implicated mitogen-activated protein kinases (MAPK) as additional mediators of H. pylori-dependent NF-κB activation and IL-8 expression. MAPK cascades are signal transduction networks that target transcription factors and thus participate in a diverse array of cellular functions including cytokine production. MAPK cascades are organized in three-kinase tiers consisting of a MAPK, a MAPK kinase (MKK), and a MKK kinase (MKKK). Transmission of signals occurs by sequential phosphorylation and activation of the components specific to a respective cascade. In mammalian systems, five MAPK modules have been identified and characterized to date; these include extracellular signal-regulated kinase 1 and 2 (ERK 1/2), p38, and c-Jun NH2-terminal kinase (JNK) (Fig.4). In addition to regulating NF-κB, MAPK can activate other transcription factors, such as AP-1, that regulate cytokine gene expression. Because MAPK can activate both NF-κB and AP-1, and the IL-8 gene promoter contains motifs for both of these DNA binding proteins (Fig. 2), a fundamental extension of previous studies focused on intracellular signaling has been to determine whether MAPK activation is also required for H. pylori-mediated IL-8 production.
Using gastric epithelial cells, Keates et al. (12) demonstrated that H. pylori rapidly induced a dose-dependent activation of ERK, p38, and JNK MAPK, and, similar to NF-κB activation, these effects were dependent on the presence of live bacteria. Preincubation with ERK- and p38-specific inhibitors completely blocked the ability of H. pylori to induce IL-8 but did not affect IκBα degradation or NF-κB activation. A recent report from the Max Planck Institute extended these observations by focusing on H. pylori-stimulated JNK activation (18). JNK is activated by MKK4 and MKK7 and targets the transcription factors c-fos and c-jun, which are components of AP-1. Upstream kinases within the JNK cascade include p21-activated kinases (PAK), which were the first kinases identified as being direct effectors for the low-molecular-weight Rho-GTPases Rac1 and Cdc42. Naumann et al. (18) have now shown thatH. pylori directly activates Rho-GTPases that subsequently stimulate PAK. PAK then activates MKK4 through an as yet unidentified intermediary MKKK, which activates JNK (18). Phosphorylated JNK then activates AP-1, possibly by activating Elk-1 (which drives expression of c-fos) and c-jun (Fig. 4).
An important question raised by these studies is whether H. pylori-induced IL-8 production is dependent on activation of NF-κB, MAPK, or both. In vitro experiments utilizing IL-8 reporter constructs revealed that H. pylori-induced IL-8 gene expression is dependent on activation of both NF-κB and AP-1 (1), and cross-talk between NF-κB and MAPK pathways has been previously demonstrated. For example, MEKK1 and NIK each can directly activate the IκB kinase signalsome, resulting in IκB phosphorylation and release of activated NF-κB (15, 16). However, although inhibition of ERK and p38 MAPK attenuates H. pylori-induced IL-8 secretion, this does not affect NF-κB activation (12), raising the possibility that synergistic interactions between AP-1 and NF-κB within the IL-8 promoter are required for maximal H. pylori-induced IL-8 production. Consistent with this hypothesis, activation of ERK by H. pylori results in Elk-1 phosphorylation and enhanced c-fos transcription (Fig. 4) (17). However, p38 is not required for H. pylori-induced activation of either Elk-1 or AP-1 (17, 18). Thus ERK (via activation of Elk-1 and c-fos) may exert regulatory effects on H. pylori-induced IL-8 production that are primarily dependent on AP-1, whereas p38 MAPK appears to influence IL-8 induction through mechanisms that are independent of AP-1 or NF-κB (Fig. 4). Collectively, these findings indicate that there is considerable redundancy in the intracellular signaling pathways activated byH. pylori that regulate IL-8 expression.
BACTERIAL COMPONENTS REQUIRED FOR H. PYLORISTRAIN-SPECIFIC ACTIVATION OF SIGNAL TRANSDUCTION PATHWAYS
The ability of H. pylori to induce epithelial cell responses related to pathogenesis is not uniform across strain populations, and, similar to associations with increased disease frequency, cagA + strains are more potent in stimulating IL-8 production than cagA − strains (4). One of the first H. pylori strain-specific constituents identified as being necessary for IL-8 production wascagE, a component of the cag pathogenicity island, and inactivation of this gene not only attenuates IL-8 expression but also decreases NF-κB activation in vitro (27). Since this report, numerous cag island genes (cagG, cagH, cagI,cagL, and cagM), but not cagA, have been demonstrated to be required for NF-κB activation (10). The identical cag island gene products responsible for NF-κB activation have now been shown to be necessary for activation of MAPK and AP-1 (12, 17, 18), confirming the pivotal role of the cag locus in activating eukaryotic signal transduction pathways that influence induction of IL-8 (Fig. 4).
Several caveats to this model require discussion. AlthoughcagA + strains stimulate higher levels of IL-8 than cagA − strains, there is substantial heterogeneity among cagA + isolates in the ability to induce IL-8 in vitro, and a small proportion of wild-typecagA − strains can stimulate IL-8 production (25). Furthermore, certain isogenic mutant strains (i.e.,cagE −, cagL −) can induce a limited IL-8 response compared with uninfected control cells (3). These observations suggest that undetected differences in the genetic composition of the cag island or genetic loci exogenous to the island may also contribute to the ability of H. pylori to induce IL-8. Censini et al. (3) reported that considerable differences exist in cag island gene content among strains that contain the cagA gene (and hence are classified as cagA +). Yamaoka and colleagues (29) provided further insights into these relationships by identifying a H. pylori outer membrane protein (oipA) that is synergistic with cagE in inducing IL-8 in vitro, supporting the contention that additional as yet unidentified strain-specific genes may also be required for induction of IL-8. Identification of such genes using molecular fingerprinting techniques such as restriction fragment length polymorphisms or random arbitrarily primed PCR is limited by the extensive genetic diversity that exists among H. pyloristrains. It is likely, however, that additional H. pyloricomponents required for stimulation of IL-8 will be identified in the near future with the development and refinement of whole genome microarrays (22), which will allow comprehensive genomic comparisons to be made between isolates that differ in their ability to induce proinflammatory cytokines.
H. PYLORI CAG-MEDIATED EPITHELIAL CELL RESPONSES THAT ARE INDEPENDENT OF INFLAMMATION
One hypothesis generated by the preceding data is that thecag type IV secretion system translocates a bacterial factor into the epithelial cell that activates NF-κB and/or MAPK with subsequent induction of IL-8. Three studies have now demonstrated that tyrosine phosphorylation of CagA occurs within the host epithelial cell after H. pylori-epithelial cell contact (2, 20,26). However, disruption of cagA does not affect NF-κB or MAPK activation or IL-8 release (12, 24, 27), indicating either that an independent bacterial factor is injected into the host cell by the cag island or that perturbation of the eukaryotic membrane per se by the type IV secretion system affects IL-8 expression (Fig. 4). What then might be the biological importance of CagA translocation and phosphorylation? The significance of phosphorylated bacterial proteins within host cells is a phenomenon that is becoming increasingly recognized. For example, the translocated intimin receptor (Tir) of enteropathogenic Escherichia coliis injected into and tyrosine phosphorylated within epithelial cells and serves as a receptor for intimin adhesion (13). Recent data indicate that phosphorylated CagA induces cytoskeletal changes including cell elongation, cell spreading, and production of filapodia and lamellipodia (23). Because CagA tyrosine phosphorylation triggers host cell morphological changes, pathways that control organization of the actin cytoskeleton may represent targets of intracellular CagA modification. Phosphorylated CagA may recapitulate intracellular events induced by Shigella flexneri IcsA and bind directly to neural Wiskott-Aldrich syndrome protein (N-WASP) with subsequent binding to the ARP2/3 actin nucleator (8), thus stimulating actin polymerization and pedestal formation (Fig. 4). Although this is unconfirmed, one may speculate that phosphorylation of CagA within the host cell confers a survival advantage that allowsH. pylori to persist for prolonged periods of time within the hostile gastric environment. An example of this mechanism is utilization of forced actin polymerization that is mediated by tyrosine phosphorylation of the vaccinia viral protein A36R, which is required for propagation of virions from cell to cell (9).
Considerable efforts have focused on delineating the precise mechanisms by which H. pylori may induce gastric inflammation. However, another avenue for future research is to consider that H. pylori may also possess means to downregulate the host inflammatory response, a requirement seemingly inherent for an organism that persists for the lifetime of its host. Crabtree et al. (5) found that inactivation of acag island gene (cag10) results in a paradoxical increase in IL-8 secretion compared with levels induced by wild-typeH. pylori. Recent studies in Salmonella typhimurium, which contains a type III secretion system, have identified dramatic differences between pathogenic and nonpathogenic strains in the ability to regulate NF-κB-dependent IL-8 production. Certain nonpathogenic Salmonella attenuate IL-8 secretion induced by pathogenic bacteria or by inflammatory cytokines by inhibiting the ubiquitination of IκBα, a novel mechanism for dampening the inflammatory response (19). It must be recognized, however, that contact between epithelial cells and H. pylori is likely to induce levels of host-bacteria adaptation that are not found during cellular interactions with acute pathogens that possess similar secretion systems. The question of whether H. pylori strains that are not associated with disease (i.e.,cagA − strains) similarly inhibit signal transduction pathways involved in generation of inflammatory cytokines is a fertile area of interest.
The author thanks Dawn A. Israel for constructive review and insightful critique of this manuscript.
Address for reprint requests and other correspondence: R. M. Peek, Jr., Div. of Gastroenterology, Vanderbilt Univ. School of Medicine, C-2104 Medical Center North, Nashville, TN 37232-2279 (E-mail:).
- Copyright © 2001 the American Physiological Society