Helicobacter: Inflammation, immunology, and vaccines

Abstract

Helicobacter pylori infection induces a chronic gastric inflammation which can lead to gastric ulcers and cancer. The mucosal immune response to H. pylori is first initiated by the activation of gastric epithelial cells that respond to numerous bacterial factors, such as the cytotoxin-associated gene A or the lipopolysaccharide intermediate heptose-1,7-bisphosphate. The response of these cells is orchestrated by different receptors including the intracellular nucleotide-binding oligomerization domain-containing protein 1 or the extracellular epidermal growth factor receptor. This nonspecific response leads to recruitment and activation of various myeloid (macrophages and dendritic cells) and T cells (T helper-17 and mucosal-associated invariant T cells), which magnify and maintain inflammation. In this review, we summarize the major advances made in the past year regarding the induction, the regulation, and the role of the innate and adaptive immune responses to H. pylori infection. We also recapitulate efforts that have been made to develop efficient vaccine strategies.

1 |. INFLAMMATION

1.1 |. Helicobacter pylori factors that regulate the activation of epithelial cells

Helicobacter pylori is essentially an extracellular bacterium that generally does not invade the gastric mucosa of the host. Gastric epithelial cells (GECs) are a primary target for H. pylori infection and actively contribute to the development of an acute and chronic inflammation. Natural immune response strongly depends on host recognition of invariant structures, namely pathogen-associated molecular patterns (PAMPs), via signaling through pattern recognition receptors such as toll-like receptors (TLRs). But H. pylori also regulates the host innate responses by injecting one of its main virulence factors, the cytotoxin-associated gene A (CagA), directly into epithelial cells using a type 4 secretion system (T4SS).

Thus, Li et al investigated the effect of H. pylori on GEC auto-phagy. Using immunohistochemistry, they demonstrated that the SQSTM1/p62 (sequestosome1) protein, which accumulates when autophagy is dysfunctional, accumulates in the gastric epithelium of patients infected with cagA⁺/vacAs1m2 H. pylori strains compared to uninfected individuals or to patients infected with cagA^–/vacAs1m2 isolates.¹ Then, in vitro investigations demonstrated that the CagA dampens the formation of autophagy by stimulating the c-Met/Akt signaling pathway.¹ Furthermore, because the inhibition of autophagy by CagA was associated with an increase of proinflamma-tory cytokine release,¹ Li et al concluded that this mechanism is an unrecognized cellular event that may sustain gastric inflammation and consequently lead to gastric carcinogenesis.

Members of the Pellino proteins (PELI1, PELI2, PELI3) have been shown to sustain TLR signaling by functioning as scaffold proteins² and by inducing polyubiquitination of interleukin (IL)-1 receptor-associated kinase 1 (IRAK1).³ TLR2 is involved in the recognition of H. pylori lipopolysaccharide (LPS) in GECs⁴ and leads to the activation of numerous signaling events, to nuclear factor-kappa B (NF-κB) activation, and to the transcription of the genes encoding chemokines such as C-X-C motif chemokine ligand 8 (CXCL8, also known as IL-8), which plays a major function in the development of gastritis and gastric cancer.⁵ Using TLR2-expressing HEK293T cells and H. pylori LPS stimulation, Smith et al demonstrated that PELI1 and PELI3 overexpression enhances NF-κB activation and CXCL8 gene expression;⁶ in contrast, the innate response was dampened by PELI3.⁶ Moreover, PELI1 mRNA expression was stimulated in the human GEC line MKN45 in response to H. pylori LPS and live bacteria.⁶ This emphasizes that the balance between PELI1/PELI2 and PELI3 in H. pylori-infected patients might have effects on the development of gastritis. In the same way, in the last year, three groups reported the critical effect of an LPS intermediate metabolite, namely heptose-1,7-bisphosphate (HBP), on the innate response of GECs.^7–9 Using direct delivery of HBP in host cells and mutants defective in the genes required for synthesis of HBP or for the T4SS, they all found that HBP (a) is delivered intracellularly by a functional T4SS, (b) is sensed by tumor necrosis factor (TNF) receptor-associated factor-interacting protein with a forkhead-associated domain (TIFA), and (c) signals in GECs to stimulate an NF-κB-dependent innate immune response,^7–9 independently of nucleotide-binding oligomerization domain-containing protein 1 (NOD1).⁸ Further, Zimmermann et al proposed that TIFA, TNF receptor-associated factor 2 (TRAF2), and tumor growth factor-beta-activated kinase 1 binding protein 2 (TAB 2) form large protein complexes called TIFAsomes.⁹ These discoveries have revived the lost interest in the role of H. pylori LPS on the induction of the immune response and the increased complexity of the interaction between the T4SS and GECs. Now, it is important to understand whether the H. pylori HBP/TIFA signaling pathway also has a biological significance in myeloid cells.

In an intriguing manner, in this last year, it has been reported that H. pylori also induces a rapid noncanonical NF-κB signaling in human HeLa cells and in the GEC line NCI-N87, characterized by an increase in the expression of the mitogen-activated protein kinase kinase kinase 14 (MAP3K14, also named NIK) and by phosphorylation/activation of p100.¹⁰ The authors then questioned the cellular mechanism leading to the activation of this pathway. The phosphorylation of p100 was not observed when a virB7 mutant strain, which lacks a functional T4SS, was used to stimulate GECs; in contrast, a cagA-deficient strain had the same effect as the parental strain,¹⁰ demonstrating that the T4SS, but not CagA, is required for this signaling event. The authors also demonstrated that MAP3K14 induction and p100 activation are dependent on the lymphotoxin beta receptor that signals via TRAF2.¹⁰ At last, they found that MAP3K14 protein expression is also increased in gastric biopsies from H. pylori-infected patients compared to uninfected individuals,¹⁰ providing a further rationale that NIK could represent a marker of H. pylori infection and may have a potential effect in H. pylori-mediated diseases. In the light of the finding that HBP,^7–9 peptidoglycan,¹¹ and DNA¹² can be transferred into GECs by a T4SS, it would be of interest to understand how H. pylori signals in cells activate the noncanonical NF-κB pathway.

Helicobacter pylori, which is an auxotroph for cholesterol, extracts the lipid from host membranes to incorporate it into its own outer membrane using the enzyme cholesterol-α-glucosyltransferase (Cgt).¹³ Cholesterol glucosylation and extraction from host cells result in lipid raft destruction and/or alteration of the membrane architecture, which has been linked to immune evasion and bacterial persistence.¹³ Phosphorylation/activation of signal transducer and activator of transcription 1 (STAT1) is stimulated in the GEC lines MKN45 and AGS, as well as in primary human GEC, treated with interferon-gamma (IFN-γ), but not in these cells preinfected for 24 h with H. pylori.¹⁴ This phenomenon was not observed with an H. pylori mutant strain lacking cgt. Furthermore, H. pylori coated with water-soluble cholesterol before infection did not block the cellular response to IFN-γ.¹⁴ At last, the reduction of cholesterol levels by H. pylori in GEC blocks IFN-γ signaling by preventing assembly of IFN-γ receptor subunits via lipid raft destruction.¹⁴ In an important way, IL-6 and IL-22 signaling in GEC was also perturbed by H. pylori via a mechanism involving a Cgt-dependent cholesterol depletion. In mice infected for 3 days, the H. pylori PMSS1 Δcgt was not able to colonize mice as the wild-type strain did but the expression of the host IFN-γ downstream response gene Irf1 was significantly higher in PMSS1 Δcgt-infected mice compared to animals infected with the parental strain.¹⁴ Therefore, H. pylori Cgt reduces cholesterol levels in infected GEC, thereby blocking T-cell signaling, allowing the bacteria to escape the host inflammatory response.

1.2 |. Regulation of inflammation by cellular receptors

Intracellular and extracellular receptors of PAMPS or immune effectors display different effects on H. pylori pathogenesis. NOD1 is expressed in GECs, senses peptidoglycan, and modulates innate immune response.¹² Tran et al investigated whether NOD1 can also dampen the inflammatory response by inducing the cytokine IL-33. Indeed, the mouse GEC line GSM06 infected with H. pylori releases the anti-inflammatory cytokine IL-33 by a NOD1-dependent pathway; this was confirmed in primary GECs from wild type (WT) and Nod1-deficient mice.¹⁵ Similar to that, the human AGS cells transfected with shRNA specific for NOD1 or with a NOD1 knockout using CRISPR/Cas9 technology failed to produce IL-33 in response to H. pylori.¹⁵ As expected, the NOD1-dependent production of IL-33 was orchestrated by a functional T4SS.¹⁵ At 8 weeks postinfection, WT mice, but not Nod1^−/− mice, exhibited a significant increase in Il33 mRNA expression and IL-33 concentration in the gastric mucosa;¹⁵ this suggests that NOD1-driven IL-33 may protect against H. pylori-mediated inflammatory responses thus favoring the persistence of the bacterium in the stomach.

Of all the hormones produced in the stomach, gastrin plays a major function in cellular processes involved in carcinogenesis such as proliferation, apoptosis, angiogenesis, motility, and tumor cell invasion. Hence, transgenic mice overexpressing gastrin under the transcriptional control of the insulin promoter (INS-GAS mice) rapidly develop gastric malignancies after H. pylori infection.¹⁶ Gunawardhana et al found that the induction of the gene encoding gastrin by H. pylori is abolished in the presence of AG1478, an inhibitor of epidermal growth factor receptor (EGFR) in AGS cells.¹⁷ Using siRNAs to specifically knockdown various signaling molecules, these authors demonstrated that H. pylori first induces the release of heparin-binding epidermal growth factor-like growth factor (HB-EGF), which in turn signals via Raf-1 proto-oncogene, serine/threonine kinase (RAF1), mitogen-activated protein kinase kinase 1 (MAP2K1, previously known as MEK1), and mitogen-activated protein kinase 1 (MAPK1, also called ERK2) to stimulate the transcription of the gene encoding gastrin.¹⁷ In the same way, it has been already reported that there is increased EGFR phosphorylation in the H. pylori-induced human gastric inflammation and cancer development.¹⁸ This year, the Wilson group also showed that treatment of H. pylori-infected INS-GAS mice or Mongolian gerbils with the EGFR inhibitor gefitinib resulted in (a) reduced activation of MAPK1/3 and activator protein 1 in GECs, (b) inhibition of chemokine synthesis by the gastric mucosa, (c) a decrease in myeloperoxidase-positive inflammatory cells, and (d) a marked reduction of gastric dysplasia and carcinoma.¹⁹ Of particular interest, similar data were obtained in mice with epithelial cell-specific deletion of Egfr (Egfr^Δepi) and key findings were recapitulated using culture of primary GEC monolayers of WT and Egfr^Δepi mice. Altogether, these data prove that the recruitment of myeloperoxidase-positive inflammatory cells to the infected gastric mucosa is orchestrated by epithelial EGFR. The same group had already demonstrated that EGFR signaling in myeloid cells is also involved in H. pylori-mediated mucosal immune response.²⁰ In this context, the inhibition of EGFR may represent an interesting chemopreventive strategy to alter the development of a persistent mucosal immune response, which may lead to carcinogenesis.

1.3 |. Beyond epithelial cells

Although the gastric epithelium is the primary site of H. pylori colonization, H. pylori can access deeper tissues and the bacterium has been found in the vicinity of endothelial cells in the gastric submucosa.²¹ By examining a previously unappreciated role of endothelial cells in eliciting a strong CXCL8 response against H. pylori infection, recent studies provide important new insights into the pathogenesis of H. pylori-associated gastric diseases. This last year, Tafreshi et al showed that, apart from gastric epithelial cells, endothelial cells may also contribute to H. pylori-induced inflammation.²² They demonstrated that primary human umbilical vein cells (HUVEC) from an immortalized HUVEC cell line (EA.Hy926) produce more CXCL8 than the human GEC line AGS when stimulated with H. pylori.²² This induction was completely inhibited when mutant strains lacking the complete cagPAI or the cagL gene, but not the cagA gene, were used to stimulate the cells; moreover, a recombinant CagL protein, but not a recombinant protein with a mutation in the arginine-glycine-aspartate (RGD) motif that is involved in the binding to the human transmembrane receptor integrin α5β1 and can activate EGFR,²³ also stimulates CXCL8 secretion by HUVECs.²² At last, CXCL8 and IL-6 production by H. pylori-infected HUVECs has been significantly reduced with the EGFR inhibitor AG1478.²² Thus, this work demonstrates that endothelial cells may play a critical role in the initiation and chronicity of the proinflammatory response in H. pylori-infected patients.

Urease displays an important function in H. pylori colonization and survival in the acidic environment of the stomach. Previous data have shown that, in a nonenzymatic way, H. pylori urease activates neutrophils to produce reactive oxygen species and also induces platelet aggregation, requiring ADP secretion modulated by the 12-lipoxygenase pathway, a signaling cascade also triggered by the physiological agonist collagen. Thus, Scopel-Guerra et al aimed to elucidate the mechanism underlying the activation of platelets exposed to an enzymatically active H. pylori urease holoenzyme and to its subunits UreA and UreB. For this purpose, they used recombinant versions of H. pylori urease and its two subunits, and they conducted platelet aggregation assays, flow cytometry and quantitative PCR to analyze how these proteins affect the platelets’ physiology. They demonstrated that the H. pylori urease, and notably its B subunit, activates platelet aggregation and that the activation of platelets by H. pylori urease appears to lead these cells to exhibit a proinflammatory phenotype.²⁴ Therefore, platelets might possibly contribute to the development and/or the persistence of H. pylori-induced inflammation.

2 |. IMMUNOLOGY

2.1 |. Activation and role of myeloid cells

Neutrophils, monocytes and macrophages, and dendritic cells are the first cells recruited in the gastric lamina propria during H. pylori infection. A remaining topic is the identification of the cellular mechanism by which myeloid cells are recruited to the gastric mucosa and are activated by H. pylori. In their recent work, Arnold et al observed that the recruitment of macrophages and CD11b⁺ dendritic cells (DCs), but not neutrophils and CD4⁺ T cells, is inhibited in mice lacking C-C Motif Chemokine Receptor 2 (CCR2), a receptor for monocyte chemoattractant protein-1.²⁵ In addition, the recruited gastric macrophages and DCs were found to express the receptor for fractalkine C-X3-C motif chemokine receptor 1 (CX3CR1), which may also play a role in myeloid cell recruitment.²⁵ Then, in an elegant study using an H. pylori strain that expresses the red fluorescent protein (RFP) and mice with a reconstituted human immune system, they showed that CX3CR1⁺ macrophages and DCs phagocytose H. pylori in the gastric mucosa;²⁵ this result is particularly important because it demonstrates that myeloid cells are in contact not only with H. pylori soluble factors, but also with live bacteria during the infection, thus providing yet another strong rationale for the study of H. pylori/myeloid cell interactions. These CX3CR1⁺/H. pylori⁺ myeloid cells expressed high levels of the inflammasome nucleotide-binding domain, leucine-rich-containing family, pyrin domain-containing-3 (NLRP3), and deletion of the gene encoding this protein led to a reduction of FOXP3⁺ Treg response, to a concomitant increased gastric T helper (Th) 1 response, and to a decreased colonization of the gastric tissue by H. pylori.²⁵ The differentiation and the recruitment of DCs were not dependent on other inflammasome-related factors such as apoptosis-associated speck-like protein containing a carboxy-terminal CARD or caspase-1. Hence, the authors concluded that H. pylori-induced NLRP3 is required to generate Treg response independently of the role of the inflammasome.²⁵ This cellular event could be considered as a novel pathway of immune escape by H. pylori. In parallel, it was reported that NLRP3 mRNA expression and IL-1β secretion is increased in the human monocytic cell line THP-1 and blood-derived monocytes/macrophages infected with H. pylori or with mutants lacking functional T4SS, CagA, the vacuolating cytotoxin A (VacA), Cgt, or flagellin,²⁶ suggesting that other factors than the typical H. pylori virulence factors are involved in the stimulation of the inflammasome in humans.

MicroRNAs (miRNAs) are short noncoding RNAs (20–22 nt) that inhibit mRNA transduction and induce their degradation. Pagliari et al were able to identify the miRNAs induced by H. pylori in human macrophages differentiated from blood-derived monocytes using colony stimulating factor 1 (CSF1 or MCSF). Using miRNA microarray, the authors observed 55 downregulated and 46 upregulated miRNAs after 24 h of infection.²⁷ Using a sophisticated comparative mapping of the miRNAs inhibited by H. pylori and of the mRNAs induced during the infection, they found that the downregulation of miR-4270 was correlated with a marked overexpression of the activating receptor CD300E, which is involved in the inducible expression of IL-1β and IL-6, and in the inhibition of MHC-II-dependent pathway of antigen presentation in human macrophages.²⁷ Thus, these findings highlight a novel mechanism by which H. pylori elicits persistent inflammation and escapes T-cell-dependent specific immune response.

2.2 |. The adaptive immune responses

CD4⁺ Th cells are central to many aspects of immune response. The proinflammatory response characterized by the Th1 and Th17 subtypes contributes to protection against H. pylori via cytokine secretion and activation of effector cells, but also supports chronic inflammation and injury that can ultimately lead to development of gastric cancer.²⁸ In the last year, clinical investigations emphasized that Th1 and Th17 responses are elicited during H. pylori infection. Adamsson et al reported that the level of expression of the genes encoding the prototype Th17 cytokine IL-17A and the Th1 lymphokine IFN-γ is enhanced in the corpus and antrum of Swedish patients with H. pylori infection compared to uninfected individuals;²⁹ more importantly they confirmed by immunofluorescence that the number of IL-17A⁺ and IFN-γ⁺ cells is also increased in H. pylori-infected patients.²⁹ Similar to that, a significant increase in IL17A mRNA transcripts was observed in the gastric tissue of H. pylori-infected patients from Iran³⁰ and France³¹ compared to uninfected ones. Arachchi et al also reported that H. pylori infection is associated with elevated IL-17 serum concentrations in a human cohort from Sri-Lanka.³² As expected, the severity of the gastritis was positively correlated with the production of IL-17.³² However, two studies showed that the number of infiltrating Th17 cells is higher in H. pylori-infected patients with peptic ulcers compared to individuals with H. pylori gastritis.^29,30 To converse, Cremniter et al described that the same level of IL17A mRNA is found in patients with gastritis, ulcer, or precancerous and cancerous lesions.³¹

An alternative picture to the Th1/Th17 synergy and antagonism emerged from a study by D’Souza et al These authors assessed the induction and the role of mucosal-associated invariant T (MAIT) cells during H. pylori infection; these cells expressed the MHC class I-related gene protein (MR1), were activated by riboflavin-producing bacteria, secreted IFN-γ and IL-17, and displayed a protective role in bacterial infections.^33,34 They found that MAIT cells were enhanced in the gastric mucosa of patients or mice infected with H. pylori.³⁵ Transgenic Vα19iCα^−/−Mr1^+/+ mice, which express only the TCRα chain (Vα19i) harbored by MAIT cells,³⁶ exhibited high levels of MAIT cells in the stomach and showed increased gastritis compared to Vα19iCα^−/−Mr1^−/− mice upon H. pylori infection.³⁵ Similar results were obtained in animals with a boosted MAIT cell levels using preinfection with Salmonella enterica serovar Typhimurium or pretreatment with Pam2Cys plus 5-(2-oxopropylideneamino)-6-d-ribit ylaminouracil. In an important way, these last animals infected with H. pylori exhibited not only more gastric T cells, but also an increase in neutrophils, macrophages, eosinophils, and dendritic cells. These data emphasize that MAIT cells should be taken into consideration to elucidate of the immunopathology of H. pylori infection; however, the level of bacterial colonization in mice with increased MAIT cells was not provided in this article, so we are not able to estimate the effect of MAIT cells on the H. pylori burden.³⁶

3 |. VACCINES

During the past year, studies sought to identify H. pylori antigens and epitopes that may lead to the development of efficient vaccine protocols were performed. By identifying the immunodominant antigens of 30 different groups of H. pylori proteins, based on their molecular weight, Sun et al determined that inosine 5′-monophosphate dehydrogenase, type II citrate synthase, and urease subunit beta are three protective antigens inducing dominant Th1 and Th17 responses.³⁷ Moreover, immunization of mice with recombinant H. pylori FliD protein was shown to elicit strong Th1 and Th17 immune responses and a significant reduction in H. pylori bacterial load by more than a 2-log order compared to mice treated with the adjuvant only.³⁸ At last, Li et al identified that only the epitopes UreB_373–385 and UreB_317–329 are effective in stimulating a strong T-cell response and are protective against H. pylori challenge in mice.³⁹

More particularly, Guo et al developed multivalent epitope-based vaccines, called CWAE vaccines. They cloned in Escherichia coli two plasmids containing CD4⁺ Th cell and B cell epitopes in antigenic H. pylori proteins predicted by T Cell Epitope Prediction Tools and B Cell Epitope Prediction Tools. The first harbored (a) two molecular adjuvants (mucosal adjuvant cholera toxin B subunit and Th1-type adjuvant), (b) tandem copies of the selected B and Th cell epitopes (UreA_27–53, UreA_183–203, HpaA_132–141, and heat shock protein 60_189–203), and (c) epitope-rich regions UreB_158–251 containing four known Th or B cell epitopes and UreB_321–385 containing three Th or B cell epitopes.⁴⁰ The second was composed of (a) the mucosal adjuvant cholera toxin B subunit, (b) Th and B-cell epitopes (UreA_27–90, UreA_183–225, UreB_327–385, membrane-associated lipoprotein Lpp20_58–97, HpaA_52–194, and CagL_21–139), and (c) an Lpp20 mimotope.⁴¹ These proteins were administered per os together with aluminum hydroxide adjuvant or polysaccharide adjuvant in gerbils infected with H. pylori SS1, 6 weeks after infection; H. pylori burden was assessed 5 weeks after vaccination. Stomach colonization was reduced by 3–4-log orders in animals treated with both CWAE vaccines compared to infected gerbils and by 1-log versus gerbils vaccinated with either a univalent epitope-based vaccine against H. pylori urease⁴⁰ or a urease vaccine.⁴¹ Antibody response and gastritis were also improved with the CWAE vaccine.⁴⁰ Thus, this multivalent epitope-based vaccine may be a promising candidate to control H. pylori infection.

As an alternative to classical vaccines, two groups engineered recombinant Lactococcus lactis strains to overexpress H. pylori antigens: (a) the complete H. pylori adhesin A (HpaA),⁴² a lipoprotein located in the cellular outer membrane and flagellum sheath, and (b) the neutrophil-activating protein A subunit (NapA),⁴³ a virulence factor that recruits and stimulates neutrophils and protects H. pylori against oxidative damage.⁴⁴ Animals inoculated once a week for four weeks with the recombinant L. lactis strain developed anti-HpaA and anti-NapA IgG, and fecal IgA.^42,43 The immunization with the NapA-expressing L. lactis strain also initiated a Th1 and Th17 response by splenic cells under an antigenic recall assay using H. pylori lysates, a partial reduction in H. pylori colonization,⁴³ but no significant change in the level of gastritis. It would now be of interest to assess the role of the HpaA-expressing strain on H. pylori colonization and clinical outcomes.