Helicobacter pylori in health and disease

H. pylori as a member of the normal human microbiota

From birth to death, humans are in contact with microbes, either transiently or persistently. Virtually every mucosal and cutaneous surface in the human body is colonized by persistent residential microbes ^1–5 (Figure 1). In most niches of the human body, including the oral cavity, esophagus, colon and skin, many bacterial species are present and no single species predominates. The distribution of the microbes is not accidental; each niche is colonized by microbes that are either conserved among most humans or host-specific. It has been presumed that the conserved microbiota have specific adaptations that permit persistence at particular locales.

What is known about bacterial colonization of the human stomach? Most studies of this topic have focused on Helicobacter pylori. Several points can be summarized:

• Natural colonization by H. pylori is restricted to humans and possibly several other primates (although the latter is not certain).

• The stomach is the major habitat of H. pylori. There may be extension of the H. pylori habitat into the proximal duodenum or distal esophagus, usually in the presence of gastric metaplasia in those sites ^{6, 7}. H. pylori also has been found overlying ectopic gastric epithelium in Meckel’s diverticulum, but this is an uncommon circumstance ⁸. H. pylori genetic sequences have been identified in oral and colonic contents, but it is not clear whether these organisms are transient or residential.

• H. pylori gastric colonization is acquired early in life (almost always before the age of 10 years), and in the absence of antibiotic therapy, generally persists for life ^{9, 10}.

• When present, H. pylori usually is the numerically dominant gastric microorganism. Studies of the bacterial flora of the human stomach, based on PCR amplification of 16S rRNA sequences, show that H. pylori represents a high proportion (70%–95%) of the clones identified ^{11, 12}. The human stomach is occasionally colonized by “Candidatus Helicobacter heilmannii” ¹³, which is closely related to H. pylori, but such colonization is relatively uncommon. Colonization of the human stomach by a single dominant species is similar to the bacterial colonization pattern sometimes observed in the human vagina ¹⁴. However, in the vagina, the dominant organism may be one of several Lactobacillus species, whereas in the stomach, only a single species (H. pylori) is typically present. Thus, H. pylori can be considered as the dominant microbiota of the human stomach.

H. pylori in human populations

H. pylori is present in human populations throughout the world. Phylogeographic studies indicate that humans have been colonized by H. pylori for at least 58,000 years, since before the most recent (but pre-historic) out-of-Africa migration ¹⁵. As humans traveled around the world populating new geographic regions, they carried their ancestral H. pylori with them ¹⁶. Based on the presence of gastric Helicobacter species (but not H. pylori) in other mammals (reviewed in ¹⁷), it is possible that gastric helicobacters are ancestral in mammals, and we may have carried the ancestors of present-day H. pylori before we evolved into humans. Unlike most other residential microbiota of which we are aware, H. pylori is becoming less common in human populations with socioeconomic development; this clearly has been happening over the course of the 20^th century in Western countries ^{18, 19}. Since humans are the only natural host for H. pylori, the decreasing prevalence can be attributed to diminished transmission among humans and perhaps a decreased duration of gastric colonization. Contributing factors include improved sanitation, smaller family sizes, and the frequent use of antibiotics during childhood. Thus, H. pylori is a major human residential organism that is becoming increasingly less common.

Tropism of H. pylori for the human stomach

H. pylori is highly adapted to colonize the human stomach, whereas most other bacteria cannot persistently colonize this niche. The major factors that limit bacterial colonization of the human stomach are: (i) acidity, (ii) peristalsis, (iii) nutrient availability, (iv) host innate and adaptive immunity, and (v) competing microbes. Specific features of H. pylori allow it to resist each of these stresses (Table 1). H. pylori resists acid by hydrolyzing urea to yield ammonia, and by regulating gene expression to respond to changes in pH ^20–22. H. pylori expresses multiple paralogous outer membrane proteins, many of which are phase-variable; several of these appear to bind to receptors on the surface of gastric epithelial cells and could diminish the rate of bacterial wash-out due to peristalsis ^{23, 24}. H. pylori has numerous mechanisms to obtain nutrients, including the induction of tissue inflammation and the presence of systems that facilitate transport and uptake of nutrients. Innate and adapative host immune responses are limited by several secreted H. pylori proteins (discussed in the next section) and multiple systems counteract the actions of reactive oxygen and nitrogen species ^{25, 26}. H. pylori produces anti-bacterial peptides ²⁷ that might reduce competition from other microbes.

Studies in rodent models have provided further insight into the H. pylori constituents required for gastric colonization. Approaches such as signature-tagged mutagenesis and microarray tracking of transposon mutants have led to the identification of more than 100 bacterial genes required for gastric colonization ^28–30. The expression of several of these genes is upregulated during growth of H. pylori in the gastric environment ³¹.

H. pylori exclusively colonizes gastrointestinal sites overlying gastric mucosa ^6–8. The development of atrophic gastritis late in life (characterized by thinning of the gastric mucosa and loss of gastric acidity) diminishes or eliminates H. pylori colonization ³². Potential reasons for the specific association of H. pylori with normal gastric mucosal epithelium include a low pH requirement for metabolic processes; dependence on specific nutrients, mucins or cell-surface components that are specific features of gastric epithelium; or the inability to compete in environments where other microbes are more abundant. The reasons that H. pylori variants have not arisen that can breech the requirement for gastric epithelium are not known, but we speculate that the biologic cost of the necessary adaptations to increase the host tissue niche exceeds the benefit to the organism, in terms of transmission to new hosts, consistent with a Nash equilibrium ³³. At least in the past, when H. pylori was so successful at colonizing humans, niche expansion was not necessary, and possibly deleterious. In the future, with an increasingly narrow bottleneck for H. pylori transmission, there could be selection for variants that colonize a broader range of epithelial surfaces; such variants might be more readily transmitted to new hosts.

Biologic factors that promote the co-existence of H. pylori and humans

Most H. pylori localize within the gastric mucus layer and do not directly interact with host cells. However, some organisms adhere to gastric epithelial cells and occasionally are internalized by these cells ³⁴. Adherence of H. pylori to gastric epithelial cells stimulates numerous signaling pathways ³⁵, and many H. pylori strains secrete toxins or other effector molecules ^{36, 37}. H. pylori elicits a humoral immune response ³⁸, and tissue infiltration by mononuclear and polymorphonuclear leukocytes occurs in all humans who are persistently colonized ³⁹. The host inflammatory response to H. pylori is relatively weak in comparison to the response to many transient bacterial pathogens, but the host response to H. pylori is more substantial and complex than that which occurs in response to other intestinal luminal bacteria. Despite causing numerous alterations in the gastric environment and eliciting a host immune response, H. pylori persistently colonizes the human stomach for long time periods and usually does not have adverse effects ^{40, 41}.

What factors contribute to the stability of the H. pylori-host equilibrium? One salient factor is the localization of H. pylori within the gastric mucus layer, without any substantial invasion of host tissue ⁴². Another factor is the synthesis of H. pylori components that are highly adapted to reduce the intensity of the host immune response. H. pylori lipopolysaccharide (LPS) is characterized by modifications of the lipid A component that make it less proinflammatory than LPSs from other gram-negative bacterial species ⁴³. H. pylori flagella are poorly recognized by TLR5 (a component of the innate immune recognition system), due to modifications in the TLR5 recognition site ⁴⁴. Many H. pylori strains express LPS O antigens that are structurally related to Lewis blood group antigens found on human cells ⁴⁵. This molecular mimicry could permit H. pylori LPS to be recognized as a self antigen. Incorporation of a modified form of cholesterol into H. pylori membranes and the coating of H. pylori with host molecules such as plasminogen might represent additional types of antigenic disguise ^{46, 47}.

H. pylori produces several factors that target host immune cells. For example, many H. pylori strains secrete a protein (VacA) that targets human CD4⁺ T cells, inhibiting the transcription factor NFAT and inhibiting T cell proliferation ^48–51. VacA targets not only CD4⁺ T cells, but also inhibits antigen presentation by B cells ⁵² and disrupts the normal functions of CD8⁺ T cells, macrophages and mast cells ^53–56. Two other H. pylori proteins (arginase and gamma-glutamyl transferase (GGT)) are also reported to cause alterations in T cells ^{57, 58}, and H. pylori arginase downregulates production of inducible nitric oxide synthase by macrophages ⁵⁹. In addition to the targeting of immune cells by the H. pylori proteins described above, H. pylori causes numerous additional effects on immune cells via mechanisms that have not yet been elucidated ⁴¹. By targeting host immune cells, H. pylori can potentially downregulate host responses and thereby maximize its persistence.

Heterogeneity among H. pylori

H. pylori strains isolated from unrelated individuals exhibit a high level of genetic diversity (reviewed in ^{60, 61}). Nucleotide sequences of conserved genes are 92%–99% identical among different H. pylori strains, but several H. pylori genes are more highly diverse in sequence ^62–64. In addition to variation in the sequences of individual genes among H. pylori strains, there is considerable variation in gene content. One study analyzed genomic DNA from 56 different H. pylori strains using array hybridization methods and identified 1150 genes that were present in all of the strains tested (thus representing a “core” genome) ⁶⁵. In contrast, 25% of the 1531 genes analyzed were absent from at least one of the 56 strains, indicating the extensive plasticity of the H. pylori genome.

H. pylori has evolved highly effective systems for generating diversity ^{61, 66}. Mechanisms include lack of mismatch DNA repair to maximize variation ⁶⁷, use of repetitive DNA for intragenomic recombination to change phenotypes, and natural competence for DNA uptake to facilitate acquisition of new genetic sequences. Gastric colonization with more than one distinct H. pylori strain is common; this multiplicity of infection provides substrate for acquisition of new genetic sequences and recombination events, which occur commonly ⁶⁸. Efficient systems for generating genetic diversity allow H. pylori to adapt to changing conditions within individual human stomachs and also permit bacterial adaptation to the gastric environments of new hosts ^{69, 70}. The simultaneous presence in the stomach of multiple H. pylori strains that can recombine could permit the emergence of the most flexible and robust bacterial populations; conversely, a decrease in the multiplicity of strains in the stomach, associated with socioeconomic advancement, might accelerate the loss of H. pylori ^{61, 66}.

One of the most striking differences among H. pylori strains is the presence or absence of a 40-kb region of chromosomal DNA known as the cag pathogenicity island (PAI) ⁷¹. One gene in the H. pylori cag PAI encodes an effector protein (CagA) whereas others encode proteins that assemble into a type IV secretion apparatus that translocates CagA into gastric epithelial cells ^{37, 72}. Within gastric epithelial cells, CagA is phosphorylated by host cell kinases ⁷³. Both phosphorylated CagA and non-phosphorylated CagA cause numerous alterations in gastric epithelial cells, including activation of the phosphatase SHP-2 and dephosphorylation of cellular proteins ⁷⁴, alterations in cell morphology and cell motility ^{75, 76}, alterations of tight junctions ⁷⁷, alterations in cell scattering and proliferation ⁷⁸, activation of β-catenin ⁷⁹, and perturbation of epithelial cell differentiation and polarity ^{80, 81}. In addition to the effects on gastric epithelial cells that result from actions of CagA, products of the cag PAI contribute to CagA-independent alterations in gastric epithelial cells, including stimulation of the synthesis of interleukin (IL)-8, a proinflammatory cytokine ^{82, 83}.

Most H. pylori strains secrete a protein known as VacA via an autotransporter mechanism (reviewed in ⁸⁴). The VacA protein was originally identified based on its capacity to cause vacuolation in cultured human epithelial cells ³⁶, but multiple other activities of this protein have subsequently been identified ⁸⁴. All H. pylori strains contain a vacA gene, but there is marked variation among strains in vacA nucleotide sequences. vacA alleles have been classified into separate families based on diversity at several loci (designed s, i, m); variations in sequence are associated with variations in VacA functional activity in cell culture assays ^{62, 85–87}. Active forms of VacA cause detectable alterations in gastric epithelial cells and immune cells whereas inactive forms of VacA (predominantly type s2) lack activity in most in vitro cell culture assays^{62, 88, 89}. Effects of active VacA on gastric epithelial cells include alterations of late endocytic compartments ⁹⁰, increased plasma membrane permeability ⁹¹, increased mitochondrial membrane permeability ^{92, 93} and apoptosis ⁹⁴ (reviewed in ⁸⁴). Most VacA-induced alterations are attributable to insertion of VacA into cell membranes, oligomerization, and formation of anion-selective channels ^{91, 95–97}. Since the inactive s2 form of VacA is well-conserved, it is likely to have a functional role that it not yet understood.

Individual H. pylori strains differ considerably in the expression and binding properties of outer membrane proteins (OMP) that function as adhesins. In particular, there are differences among strains in the expression and binding properties of BabA (an OMP that binds the fucosylated Lewis b receptor on gastric epithelial cells) and SabA (an OMP that binds to sialyl Lewis X receptors) ^{23, 24}. These differences among strains in adhesin expression result in strain-specific variations in binding of H. pylori to gastric epithelial cells. There also are differences among strains in the expression of the outer membrane OipA (HopH) due to phase variation, which could result in strain-specific variations in H. pylori-induced signaling in gastric epithelial cells ⁹⁸.

H. pylori strains can be broadly categorized into 2 groups: strains that express multiple factors that interact with host tissue (including proteins encoded by the cag PAI, active forms of VacA, and outer membrane proteins such as BabA) and strains that lack these factors ^{62, 99, 100}. Strains with intermediate properties have been identified, although less frequently than expected than if the distribution of H. pylori virulence factors were completely random. Recent studies indicate that CagA and active forms of VacA have reciprocal or antagonistic actions; consequently, there may be selection for strains that encode both of these factors or strains that lack both factors ^101–103.

H. pylori strains that express multiple ‘interaction factors’ (CagA⁺, s1-VacA⁺, BabA⁺ strains) are predicted to be highly interactive with the host, whereas strains that lack these factors would be relatively non-interactive (Figure 2). Concordant with these predictions, CagA⁺, s1-VacA⁺, BabA⁺ strains are associated with increased gastric mucosal inflammatory cell infiltration and increased gastric epithelial injury, compared to strains that do not express these factors ^{99, 104}. In addition, the colonization density of CagA⁺, s1-VacA⁺, BabA⁺ strains is typically higher than that of strains that do not express these factors ¹⁰⁵.

H. pylori strains expressing multiple interaction factors and strains that lack these factors might occupy different niches in the gastric environment or each could have selective advantages at different times during prolonged colonization. Currently, people in developing countries are predominantly colonized by cagA⁺ strains, whereas those in many developed countries are colonized by an almost equal proportion of cagA⁺ and cagA⁻ strains ^{100, 106}. This suggests that there is an accelerated loss of cagA⁺ strains from some populations ¹⁸. cagA⁺ strains induce the production of beta-defensin 2 and other antimicrobial effectors to a greater extent than cagAstrains ¹⁰⁷, which might render cagA⁺ strains more susceptible to eradication from the host. In addition, cagA⁺ strains seem to be more efficiently eradicated by antibiotics than are cagAstrains ¹⁰⁸.

H. pylori and gastroduodenal disease

Although H. pylori typically colonizes the human stomach for many decades without adverse consequences, the presence of H. pylori is associated with an increased risk for several diseases, including peptic ulcers, non-cardia gastric adenocarinoma, and gastric MALT lymphoma (reviewed in ^{109, 110}). What factors account for the development of these diseases in subsets of people who harbor H. pylori?

The risks of peptic ulcer disease and non-cardia gastric adenocarcinoma are determined in part by characteristics of the H. pylori strain with which an individual is colonized. Most of the H. pylori polymorphisms associated with varying disease risk are found in genes that encode bacterial products that interact with host tissue. Numerous studies, particularly in Western countries, have shown that cag PAI-positive H. pylori strains are associated with an increased risk of peptic ulcer disease, premalignant gastric lesions and gastric cancer, compared to strains that lack the cag PAI ^{100, 111–114}. Moreover, the number of tyrosine phosphorylation (EPIYA) motifs in CagA proteins correlates with gastric cancer risk ^{115, 116}. Strains that express forms of VacA that are active in vitro (for example, s1/i1/m1) are associated with a higher risk of disease than those that express inactive forms of VacA ^{62, 85, 112, 117}. Similarly, strains that express BabA and OipA (HopH) outer membrane proteins are also associated with a higher risk of disease than strains that lack these factors ^{99, 118}. Based on data from human epidemiologic studies, it is difficult to determine which of these bacterial factors is most closely linked to adverse disease outcomes, since these interaction factors tend to cluster together in H. pylori strains ^{62, 99, 117}.

Studies involving gerbil and transgenic mouse models suggest that products of the cag PAI (including CagA) have an important role in contributing to adverse disease outcome ^119–122. Some studies in rodents suggest that products of the cag PAI and VacA enhance the ability of H. pylori contribute to colonize the stomach ^{123, 124}, but other studies have not reached the same conclusions ¹²⁵. Notably, there are limitations of various animal models in replicating the human host environment. For example, some H. pylori factors (such as those encoded by the cag PAI) are present on genetically metastable elements that are commonly deleted during colonization of mice ^{126, 127}. Furthermore, human T cells are susceptible to VacA, whereas mouse T cells are not ^{51, 128}.

Host and environmental factors also are important determinants of H. pylori-associated disease risk. For example, male gender, specific IL-1β haplotypes, and various other proinflammatory gene polymorphisms are associated with an increased risk of non-cardia adenocarcinoma ^{129, 130}. There might be synergy between bacterial and host polymorphisms in determining disease risk ¹¹². Environmental factors that may influence the risk of gastric cancer include the level of dietary salt intake, intake of fresh fruit and vegetables, and the presence of various parasitic infections ^131–133.

The use of antibiotics to eradicate H. pylori has dramatically altered the incidence and natural history of peptic ulcer disease ¹³⁴. H. pylori resistance to macrolides, fluoroquinolones, and nitroimidzaoles is gradually increasing due to the widespread use of these antibiotics in the community for multiple indications. Increasing antibiotic resistance decreases the efficacy of current triple-drug treatment regimens for H. pylori and may portend future difficulties in our ability to treat peptic ulceration with antibiotics.

Potential benefits of H. pylori

H. pylori colonization is associated with many biological costs to the host; conversely, a growing body of literature suggests that the absence of H. pylori might also be associated with an increased risk of various diseases. An absence of H. pylori could indicate that an individual was never colonized, or that the organism was present earlier in life and subsequently eradicated. The idea that H. pylori might actually confer benefits to humans has engendered considerably controversy among investigators, but we review here the current data and discuss the potential importance of health benefits that might be afforded by H. pylori. Not surprisingly, most of the potential benefits (as with the costs) come from cagA⁺ strains, which are the most interactive with their human hosts. In 1998, one of us used the term acagia to describe the absence of cagA⁺ H. pylori, a condition associated with disease risks that differ from those associated with the presence of cagA⁺ H. pylori ¹³⁵.

Conclusions

The rediscovery of gastric microbiota and the first sucessful culture of H. pylori in 1982 by Marshall and Warren opened a new chapter in human medicine ¹⁶⁵. Early work, demonstrating a relationship between H. pylori and peptic ulcer disease, changed medical practice ¹³⁴. The finding that H. pylori also increased the risk of gastric adenocarcinoma bolstered the view that H. pylori is a human pathogen. However, it is now becoming clear that the progressive disappearance of H. pylori in the 20^th and 21^st centuries, abetted by modern medical practices (including overuse of antibiotics in childhood), may have consequences. These may include an increased risk of gastroesophageal reflux disease and its sequelae, childhood asthma, and metabolic disorders. If continued studies confirm the findings reported thus far, then our medical approaches to H. pylori will need to change. These next years will be an exciting period in which the relationship between H. pylori and humans becomes more thoroughly understood.

Biographies