|Home | About | Journals | Submit | Contact Us | Français|
Microbial pathogens contribute to the development of more than 1 million cases of cancer per year. Gastric adenocarcinoma is the second leading cause of cancer-related death in the world, and gastritis induced by Helicobacter pylori is the strongest known risk factor for this malignancy. H. pylori colonizes the stomach for years, not days or weeks, as is usually the case for bacterial pathogens, and it always induces inflammation; however, only a fraction of colonized individuals ever develop disease. Identification of mechanisms through which H. pylori co-opts host defenses to facilitate its own persistence will not only improve diagnostic and therapeutic modalities, but may also provide insights into other diseases that arise within the context of long-term pathogen-initiated inflammatory states, such as chronic viral hepatitis and hepatocellular carcinoma.
A strong link has been established between chronic H. pylori colonization and a diverse spectrum of diseases, including gastric and duodenal ulceration, gastric adenocarcinoma, mucosa-associated lymphoid tumor (MALToma), and non-Hodgkin's lymphoma of the stomach. However, the cognate niche inhabited by H. pylori, the human stomach, poses numerous barriers that normally prevent sustained microbial colonization including gastric peristalsis, an acidic pH, and inflammation-generated oxidative stress. Further, H. pylori must nimbly escape elimination by the host immune response to effectively persist within the stomach. In addition to constituents required for colonization (Table 1), H. pylori harbors several virulence factors that can influence disease, but pathologic outcomes ultimately depend on a dynamic interplay between bacteria and host over prolonged periods of coexistence.
One of the initial obstacles encountered by H. pylori is the harsh, acidic environment of the stomach, which is frequently at or below pH 2. To circumvent this deterrent, H. pylori produces copious amounts of urease, which catalyzes the hydrolysis of urea into ammonia and carbonic acid. Studies in C57BL/6 and nude mice have demonstrated that urease activity is unequivocally required for successful colonization.1–3 In macaques, infection with H. pylori strains producing low levels of urease activity leads to an expansion of bacterial subpopulations in which urease production is enriched, further emphasizing the importance of urease activity for successful colonization.3
H. pylori can also manipulate the intragastric pH by stimulating the production of IL-1β, a pro-inflammatory cytokine with potent acid-suppressive properties. In H. pylori-infected Mongolian gerbils, gastric mucosal IL-1β levels increase 6–12 weeks post-challenge, which is accompanied by a reciprocal decrease in gastric acid output.4 Administration of recombinant IL-1 receptor antagonist normalizes acid outputs, implicating IL-1β as a pivotal modulator of acid secretion within inflamed gastric mucosa.4 Informative polymorphisms within genes regulating IL-1β production in humans have allowed case-control studies to be performed that relate genotypes to H. pylori-associated diseases. El-Omar et al. were the first to demonstrate that IL-1β polymorphisms associated with increased expression of IL-1β heighten the risk for atrophic gastritis and gastric adenocarcinoma.5–9 In addition to IL-1β, high-expression alleles of TNFα, another pro-inflammatory cytokine with acid-suppressive properties, also increase the risk for gastric cancer.6,8 Importantly, these relationships are only present among H. pylori-colonized persons and not uninfected individuals, emphasizing the significance of host-environment interactions and inflammation in the progression to gastric cancer.
Another obstacle that must be confronted within the gastric niche is peristalsis. H. pylori invokes several mechanisms to overcome this impediment, including efficient motility, chemotaxis, and the ability to specifically adhere to gastric epithelium. H. pylori possesses between 3–5 polar flagella, which are comprised primarily of FlaA and FlaB subunits. Deletion of flaA results in flagellar truncation and decreased motility in vitro, while in vivo, FlaA as well as other proteins necessary for flagellar assembly are essential for colonization of rodent and piglet models of infection.10–12 Unlike FlaA of Salmonella or other Gram-negative pathogens that colonize mucosal surfaces, H. pylori FlaA is a non-inflammatory molecule, which may contribute to long-term persistence.13,14 In contrast to flaA− deficient strains, H. pylori mutants lacking expression of the flagellar motor protein MotB, which is also required for motility, are capable of acutely colonizing mice; however, the 50% infectious dose (ID50) is >4 log-fold higher than that observed for wild-type H. pylori and mutants are unable to be recovered >48 hours after co-challenge with wild-type strains.15,16
In addition to flagella-mediated motility, chemotaxis plays a role in the ability of H. pylori to establish residence in the stomach. To sense external stimuli, H. pylori expresses at least four chemotaxis receptor proteins: TlpA, TlpB, TlpC and TlpD. TlpA is a receptor for arginine and sodium bicarbonate while TlpB is required for pH taxis, but the specific activators for TlpC and TlpD are unknown.17,18 In contrast to FlaA or FlaB, loss of Tlp proteins does not alter levels of colonization in wild-type infected animals; however, mutations in tlpB significantly reduce levels of inflammation in infected C57BL/6 mice or gerbils.19,20
H. pylori response regulators are also important in chemotaxis. Mutants defective in the receptor-kinase adaptor CheW, the histidine kinase CheA, or the response regulator CheY are non-chemotactic, but are able to infect mice at low levels.21 CheW appears to be important for initial steps in colonization, as the ID50 is 1–2 log-fold higher than that observed for wild-type strains, although at 6 months post-challenge, colonization levels are similar between mutant and wild-type isolates.21 Further, cheY and cheA mutants induce lower levels of inflammation in mice and, similar to tlpA and tlpB mutants, are not found closely associated with gastric epithelial cells as frequently in vivo when compared to the topography of wild-type H. pylori.20 Collectively, these data suggest a critical role for motility and chemotaxis for initial as well as long-term colonization of the stomach.
Though most H. pylori are free-living within the semipermeable mucous gel layer, approximately 20% of the bacterial population binds to gastric epithelial cells which is thought to be important for colonization. Further, the bacteria that bind epithelial cells are found in close apposition to epithelial tight junctions.22,23 Tight junctions are dynamic structures located at the most apical region of cell-cell contact points and play critical roles in maintenance of barrier function, cell polarity, and intercellular adhesion. H. pylori not only adhere to epithelial cells in close proximity to tight junctions22,23 but also alter the localization patterns of component proteins that form junctional complexes, such as occludin, claudins, junctional adhesion molecules (JAMs), and zonula occludens-1 (ZO-1).23–30 Occludin has been implicated in the regulation of barrier function and is directly linked to the actin cytoskeleton via its C-terminus and indirectly through its interactions with ZO-1.26,31–35 Expression of occludin at the level of the tight junction is disrupted by H. pylori in cultured canine duodenal epithelial cell monolayers, and our laboratory has recently shown that H. pylori induces barrier dysfunction in gastric epithelial cells via urease-mediated dysregulation of occludin.25,26
H. pylori expresses numerous proteins that function as epithelial cell adhesins. One of the most well-studied is the outer membrane protein (OMP) BabA, which is encoded by babA2 and binds Lewisb (Leb) blood-group antigens expressed on gastric epithelial cells (Fig. 1). Though not all H. pylori express BabA, BabA+ strains are associated with an increased risk for gastric cancer.36 Expression of BabA is regulated by slipped strand mispairing during replication as well as intragenomic recombination, which can result in aberrant localization of babA2 into the locus occupied by the closely related gene babB.37,38 Evidence of altered regulation of babA expression in vivo has been demonstrated through H. pylori infection of macaques, which revealed that babA2 could be replaced by babB, leading to an inability of recovered isolates to bind Leb in vitro.37
OipA is another H. pylori adhesin that can be regulated by slipped-strand mispairing in a manner similar to BabA.39 Expression of OipA in cell co-culture systems leads to higher levels of production of the pro-inflammatory cytokine IL-8 by infected gastric epithelial cells, while in vivo, OipA expression is associated with more severe inflammation in mice as well as cancer in Mongolian gerbils.39,40 SabA is an H. pylori OMP that binds the sialylated glycan Lex (Fig. 1). In situ studies using tissue from Rhesus monkeys demonstrated that an H. pylori mutant lacking BabA bound gastric epithelium in a pattern that mirrored host sialyl-Lex expression. Further, BabA deficient H. pylori pretreated with sialyl-Lex could bind tissue from Leb-, sialyl-Lex expressing mice but with less than 10% efficiency compared with untreated bacteria.41 Thus, H. pylori OMPs that function as adhesins also appear to be critical regulators of efficient colonization of the host.
Reactive oxygen species such as superoxide and reactive nitrogen intermediates exert damaging effects on proteins, membranes, and DNA, which can have dire consequences for colonizing microbes.42 During the course of colonization, H. pylori induces a robust inflammatory response that leads to production of molecules that can potentially induce oxidative and nitrosative damage. However, H. pylori has developed strategies to overcome these obstacles, including expression of a superoxide dismutase (SodB). SodB catalyzes the conversion of superoxide to H2O2, a substrate that can rapidly be inactivated by peroxide or catalase, and SodB-deficient H. pylori colonize mice at a rate of only 4% compare to colonization rates seen in wild-type strains.43
In addition to SodB, H. pylori expresses three peroxiredoxins: alkyl hydroperoxide reductase (AhpC), thiolperoxidase (Tpx), and bacterioferritin comigratory protein (BCP). Disruption of genes encoding any of these proteins results in a dramatic decrease in the ability to colonize mice.44–46 H. pylori also expresses catalase (KatA) which facilitates the conversion of H2O2 into oxygen and water. Mutants lacking this enzyme, or the catalase-associated protein KapA, exhibit an increased sensitivity to H2O2.47 Although the short-term colonization efficiency of catalase mutants is essentially the same as wild-type strains, persistent colonization is only 20–50% of wild-type when assessed 24 weeks post-challenge, suggesting a role for these elements in facilitating chronic colonization within an inflammatory milieu.47,48 Another H. pylori gene, mdaB, encodes a NADPH quinone reductase that also has a role in resisting H2O2-mediated damage and MdaB-deficient H. pylori colonize mice at significantly lower frequency and density when compared to wild-type strains.49
H. pylori NapA was originally reported to function in neutrophil recruitment; however, NapA mutants have also been shown to exhibit decreased resistance to oxidative stress and colonize mice with lower efficiency when compared to wild-type strains.50,51 Additionally, H. pylori mutants deficient in SodB, AhpC or KatA express increased levels of NapA, suggesting that NapA may play a compensatory role in oxidative stress resistance under these conditions.46,52 Finally, H. pylori RocF, which converts arginine to urea and ornithine, has also been shown to be important for evading clearance mechanisms that utilize nitric oxide as well as providing shelter from acid-mediated damage. RocF mutants are 1,000 times more sensitive to acid than wild-type H. pylori, which may be due to conversion of arginine-generated urea to ammonia by urease.53,54 RocF can also compete with macrophages for the iNOS substrate arginine, resulting in decreased host iNOS production thereby permitting clearance by nitric oxide defense mechanisms.55,56
The most well-studied defense mechanism mounted by human hosts against microbial infection is the immune response, but as might be predicted for a chronic pathogen, H. pylori possesses numerous mechanisms to overcome this obstacle and establish chronic colonization. Gram-negative bacteria such as H. pylori possess lipopolysaccharide (LPS) as a component of the bacterial cell wall, which typically elicits a robust inflammatory response. However, the LPS of H. pylori is relatively anergic, exhibiting >3 log-fold less endotoxin activity when compared with LPS from other Gram-negative bacteria.57–59 In vivo studies indicate that H. pylori LPS may also have a role in modulating the immune response as mice immunized with bacterial sonicate containing LPS elicited a strong Th1 immune response while LPS-depleted sonicates induced a Th2 response.58,60
Another host constituent activated by H. pylori that may dampen the immune response is protease-activated receptor (PAR)-1, a G-protein coupled receptor expressed on many cell types including gastric epithelial cells. A recent study indicates that PAR-1 is protective against H. pylori induced inflammation and that it acts to suppress proinflammatory cytokines including MIP-2 in mice.61
The O-antigen of H. pylori LPS contains various human Lewis antigens including Lex, Ley, Lea and Leb, and inactivation of Lex and Ley-encoding genes prevents H. pylori from colonizing mice.62–65 Approximately 85% of H. pylori clinical isolates express Lex and Ley and although both Lex and Ley can be detected on individual strains, one antigen usually predominates.66–68 Several studies have demonstrated that H. pylori Lewis antigens can undergo phase variation in vitro68–71 and in vivo studies using Rhesus monkeys (which like humans, express Lex and Ley on gastric epithelial cells), demonstrated that the Lewis expression pattern of colonizing bacteria varies in response to the expression pattern of their cognate hosts.72,73 A recent study using transgenic mice expressing Leb also demonstrated that H. pylori recovered from these animals, but not wild type mice, expressed Leb.74 This suggests that Lewis antigens facilitate molecular mimicry of the host and allow H. pylori to escape host immune defenses by preventing the formation of antibodies against shared bacterial and host epitopes.
H. pylori has recently been shown to decrease production of specific heat shock proteins (HSPs) in vitro and within colonized gastric mucosa.75 Since HSPs can modulate both innate and adaptive immune responses, inhibition of HSP production may represent an additional mechanism of immune evasion that promotes long-term colonization. Finally, a subpopulation of H. pylori has been shown to maintain residence within gastric epithelial cells as well as within the lamina propria of gastric mucosa, which may represent sanctuary sites that protect against immune clearance.76–78
Although most persons colonized with H. pylori never develop clinically apparent disease, colonization can lead to the development of peptic ulcer disease and gastric cancer. Numerous studies have revealed that there is substantial genetic variability between H. pylori isolates, and indeed, isolates harvested from a single person are genetically unique.79–81 Despite this tremendous degree of diversity, several factors associated with disease have been identified.
CagA is an immunodominant protein encoded by cagA, the terminal gene within the cag pathogenicity island (PAI), and is one of the most intensely studied virulence factors expressed by H. pylori. The presence of the cag PAI varies among H. pylori populations and is found in approximately 60–70% of strains in the United States.82–84 H. pylori cag PAI+ strains are associated with an increased risk for atrophic gastritis, peptic ulcer disease, and distal gastric cancer compared to strains that lack this genetic locus.85–92
The cag PAI encodes a type IV secretion system which functions to translocate bacterial components such as cagA into host epithelial cells. CagL is expressed, located at the tip of the pilus, and binds α5β1-integrins on target cells in an RGD dependent manner. This interaction activates FAK (focal adhesion kinase) and Src and has a role in H. pylori-induced IL-8 secretion by host cells.93 Recently, CagA, CagI and CagY have also been demonstrated to bind β1-integrin inducing conformational changes permitting CagA translocation.94 Intracellular CagA can be sequentially phosphorylated by Src and Abl kinases, and phosphorylated CagA activates the eukaryotic phosphatase SHP-2 leading to dephosphorylation of host cell proteins and morphological changes including cellular elongation and cell spreading (Fig. 2).95–98 Phosphorylation sites have been identified as EPIYA motifs which are located at the carboxy-terminus of CagA and up to 5 motif copies may be present within a single CagA protein. Four variants of the EPIYA motif have been described: EPIYA-A, -B, -C and -D.96,99,100 EPIYA-A and -B motifs are found in strains throughout the world; in contrast, EPIYA-C motifs segregate with CagA proteins expressed by western strains while EPIYA-D motifs predominate in CagA proteins from strains of East Asian origin. In vitro studies have indicated that the number of C-type phosphorylation motifs is associated with the intensity of CagA phosphorylation, cellular morphological changes, and induction of proinflammatory cytokines.101–103 Further, in some studies, a greater number of EPIYA sites is linked to gastric atrophy, intestinal metaplasia and gastric cancer, indicating a role for these sites in pathogenesis.99,102,104,105 The sequence surrounding EPIYA-D motifs matches the consensus sequence for SHP-2 binding perfectly and as a result, strains expressing an East Asian CagA exhibit greater SHP-2 binding leading to increased morphological changes compared with western strains.106 However, unphosphorylated CagA moieties also play important roles in pathogenesis by altering cell signaling, including inducing aberrant activation of β-catenin, disrupting apical-junctional complexes, activating STAT3, and inducing a loss of cell polarity (Fig. 2).95–98,107,108
As previously described, H. pylori is frequently found in close association with epithelial cell tight junctions. At junctional complexes, translocated CagA associates with the tight-junction proteins ZO-1 and junctional adhesion molecule (JAM), leading to disruption of tight junctions and alterations of cell polarity.23,108–111 Intracellular CagA also binds to E-cadherin at adherens junctions.110 Disruption of cellular organization has now been shown to be due to aberrant inactivation of the kinase Par1, a regulator of cell polarity.108,111,112 Binding of Par1 by CagA also results in perturbation of microtubule formation and myosin II activation.112
Another bacterial molecule that is translocated by the cag T4SS into host epithelial cells is peptidoglycan, which interacts with the host pattern recognition molecule NOD1. NOD1 functions as an intracellular sensor for peptidoglycan components that originate from Gram-negative bacteria. The interaction of H. pylori peptidoglycan with NOD1 leads to increased NFκB activation, resulting in increased production of inflammatory molecules such as IL-8.113 Translocated peptidoglycan also leads to enhanced PI3K-AKT signaling, which decreases H. pylori-induced apoptosis and increases cell migration, phenotypes that may set the stage for carcinogenesis.114 A recent study has now indicated that H. pylori peptidoglycan can also be delivered intracellularly via bacterial outer membrane vesicles.115
In addition to effects induced by translocated CagA and peptidoglycan, the cag PAI-encoded T4SS per se can also exert effects on host cells. The cag PAI protein CagL interacts directly with host cell integrins, resulting in local membrane ruffling and clustering of integrins into focal adhesion complexes.93 These events lead to FAK and Src kinase activation and CagA phosphorylation, resulting in morphological and cell signaling alterations that are similar to growth factor stimulation (Fig. 2).93
Another bacterial factor that is clearly associated with H. pylori-induced disease is the vacuolating cytotoxin VacA. VacA was first identified to harbor toxigenic properties by the demonstration that H. pylori VacA-containing culture supernatants produced vacuolation in eukaryotic cells in vitro.116 VacA is an approximately 140 kDa protein which undergoes proteolytic cleavage to yield an 88 kDa product that can be further cleaved to 33 kDa and 55 kDa fragments that can ultimately assemble into large oligomeric complexes.117–121 Unlike cagA, the gene encoding VacA (vacA) is found in all H. pylori strains, but exists as allelic variants that exert differences in epithelial cell responses. Polymorphic sites exist within the signal (s) region, the mid (m) region and the more recently identified intermediate (i) region122–124 and vacA s1/m1/i1 variants are more strongly associated with gastric adenocarcinoma.122
Oligomeric VacA complexes can insert into cellular lipid bilayers, forming anion-selective membrane channels.125–128 The vacuolating activity of VacA is dependent on formation of these channels in conjunction with weak bases such as ammonium chloride.129–131 After membrane insertion, VacA is internalized by a pinocytic-like process, leading to the formation of anion-selective channels in endosomal membranes.124,126,132 Vacuolation relies on numerous host factors previously implicated in endosome formation such as V-ATPase, Rab7, Rac1, dynamin and PIKfyve.133–135 VacA also induces alterations in late endocytic compartments which leads to inhibition of EGF degradation, clustering of endocytic compartments, and inhibition of antigen presentation.131,136,137
In addition to inserting into plasma and endocytic membranes, VacA also interacts with mitochondrial membranes leading to a decrease in mitochondral transmembrane potential and release of cytochrome c.138–140 One study has indicated that VacA-induced cytochrome c release may be indirect and mediated instead by Bax and Bak activation.141 Transient expression of VacA in HeLa cells results in cleavage of poly-ADP-ribose polymerase (PARP), an indicator of caspase-3 activation.138 Both cytochrome c release and caspase-3 activation are pro-apoptotic events, and this mirrors the apoptotic response observed following incubation of VacA with AGS cells.142,143 Although VacA s1m1 variants induce high levels of apoptosis, VacA s2m1 forms do not, suggesting a role for strain variation in differences in gastric epithelial apoptosis that have been observed among H. pylori-infected persons.142,143
Recent studies have also suggested a role for CagA in modulating the apoptosis-inducing function of VacA.144,145 In Mongolian gerbils, CagA activates the cell survival pathway mediated by the MAPK ERK and induces expression of the anti-apoptoptic protein MCL1 in gastric pit cells.145 Further, unphosphorylated CagA antagonizes VacA-induced apoptosis at the mitochondrial membrane while phosphorylated CagA hinders VacA from reaching targeted intracellular compartments.144 Together, these data suggest a possible model in which VacA has a role in colonization of the gastric niche, but after colonization is established, its activity is modulated by translocated CagA, favoring bacterial persistence and protection against cellular apoptosis.
In addition to influencing cell turnover, VacA plays an important role in suppressing the host immune response to H. pylori. Two studies have demonstrated that VacA interrupts phagosome maturation by recruiting tryptophan-aspartate-containing coat protein (TACO or coronin 1), suggesting that VacA may have a role in protecting H. pylori from phagocytic killing.146,147 However, another study using freshly isolated human monocytes demonstrated no effect of VacA on phagosome fusion or intracellular survival.148 This discrepancy may be due to differences in cell lineages or bacterial strains that were used and additional work is needed to reconcile the results from these studies. Additionally, VacA has been shown to interfere with antigen proteolytic processing in T cells and inhibit antigen presentation through the invariant chain dependent pathway thereby interfering with adaptive immunity.136
VacA also inhibits proliferation of Jurkat T cells by interfering with IL-2 signaling, which is due to inhibition of nuclear factor of activated T-cell (NFAT) activation.149–151 Evidence suggests that VacA blocks calcium influx resulting in downregulation of the Ca+-calmodulin-dependent phosphatase calcineurin, which in turn inhibits NFAT transcription activation.149,150 VacA can also inhibit proliferation of primary human CD4+ T cells, independently of NFAT.151 In total, the anti-proliferative effects of VacA may be a mechanism by which T cells activated by H. pylori antigens can be inhibited, allowing H. pylori to escape the adaptive immune response and establish chronic infection.
Gastric adenocarcinoma is strongly associated with the presence of H. pylori, and both microbial and host factors influence the risk for carcinogenesis. Interactions between H. pylori constituents and host epithelial and immune cells play an important role in the development of gastric injury. H. pylori has evolved an array of diverse constituents to subvert obstacles presented by the gastric niche which may consequently induce host responses related to carcinogenesis within the context of long-term microbial colonization. Molecular delineation of pathways activated by such host-microbial interactions will not only improve our understanding of H. pylori-induced carcinogenesis, but may also provide mechanistic insights into other malignancies that arise within the context of inflammatory states (e.g., ulcerative colitis and colon cancer).
Previously published online: www.landesbioscience.com/journals/gutmicrobes/article/11991