Helicobacter pylori represses proton pump expression and inhibits acid secretion in human gastric mucosa

Abstract

Background and aims: Helicobacter pylori infection of gastric mucosa causes gastritis and transient hypochlorhydria, which may provoke emergence of a mucosal precancer phenotype; H pylori strains containing a cag pathogenicity island (PAI) augment cancer risk. Acid secretion is mediated by the catalytic alpha subunit of parietal cell H,K-ATPase (HKalpha). In AGS gastric epithelial cells, H pylori induces nuclear factor-kappaB (NF-kappaB) binding to and repression of transfected HKalpha promoter activity. This study sought to identify bacterial genes involved in HKalpha repression and to assess their impact on acid secretion.

Methods and results: AGS cells transfected with an HKalpha promoter construct or human gastric body biopsies were infected with wild-type (wt) or isogenic mutant (IM) H pylori strains. AGS cell HKalpha promoter activity, and biopsy HKalpha mRNA, protein and H(+) secretory activity were measured by luminometry, reverse transcription-PCR, immunoblotting and extracellular acidification, respectively. Wt H pylori and DeltavacA, DeltaureA, Deltaslt and DeltaflaA IM strains repressed HKalpha promoter activity by approximately 50%, a DeltacagA IM strain repressed HKalpha by approximately 33%, and DeltacagE, DeltacagM and DeltacagL IM strains elicited no HKalpha repression. Wt H pylori-infected biopsies had markedly reduced HKalpha mRNA and protein compared with IM strain infections or mock-infected controls. Histamine-stimulated, SCH28080-sensitive biopsy acid secretion was significantly inhibited by wt but not by DeltacagL IM H pylori infection compared with vehicle-only controls.

Conclusions: It is concluded that H pylori cag PAI gene products CagE, CagM, CagL and, possibly, CagA are mechanistically involved in repression of HKalpha transcription. Further, acute H pylori infection of human gastric mucosa downregulates parietal cell H,K-ATPase expression, significantly inhibiting acid secretion.

Figure 1

The H. pylori cag pathogenicity island (PAI) represses H,K-ATPase α subunit (HKα) promoter activity in AGS cells. Cells were transiently transfected with the HKα2179 promoter—Luc reporter construct and infected (multiplicity of infection 25, 6 h) with wild-type (wt) H pylori strains 7.13, 8823, 7.13 ΔcagPAI or 8823 ΔcagPAI. HKα2179 promoter activity was measured as normalised relative light units (RLU) of luciferase activity. Data represent the mean ±SD from three independent experiments (***p<0.001).

Figure 2

The cag pathogenicity island (PAI)-specific genes cagM, cagE, cagL and cagA are responsible for H,K-ATPase α subunit (HKα) promoter repression in AGS cells. AGS cells were transfected with the HKα2179 promoter—Luc reporter construct and infected (multiplicity of infection 25, 6 h) with wild-type H pylori or one of eight H pylori isogenic mutant strains (ΔcagM, ΔcagE, ΔcagA, ΔcagL, Δslt, ΔvacA, ΔflaA and ΔureA). HKα2179 promoter activity was expressed in normalised relative light units (RLU) of luciferase activity. Data represent the mean ±SD from three independent experiments (***p<0.001; *p<0.05).

Figure 3

H pylori represses H,K-ATPase α subunit (HKα) levels in human gastric biopsies. Biopsies were infected (1—2×10⁵ bacteria/mg wet weight biopsy, 24 h) with wild-type H pylori strain 60190. (A) RNA was extracted from biopsies and the HKα mRNA content was measured by reverse transcription—PCR. Open bars represent mock-infected biopsies, and shaded bars represent patient-matched infected biopsies (means, SD, n=3). (B, C) Gastric biopsies before (B) and after (C) 24 h infection with wild-type H pylori (H&E stain, ×40 magnification; arrows indicate parietal cells). (D) Biopsies were lysed and HKα protein content was assessed by immunoblotting using antibody HK 12.18 against HKα and β-actin antibody as gel loading control (representative gel of three individual patient replicates).

Figure 4

cag pathogenicity island (PAI)-specific genes are necessary for H,K-ATPase α subunit (HKα) repression in human gastric mucosa. Gastric biopsies were infected ex vivo (1—2×10⁵ bacteria/mg wet weight biopsy, 24 h) with (A) wild-type (wt) H pylori strain 7.13 or 7.13 ΔcagPAI isogenic mutant (IM) strain; (B) wt H pylori strain P12 or P12 ΔcagL IM strain; (C) wt H pylori strain 7.13 or 7.13 ΔcagE IM strain; and (D) wt H pylori strain 7.13 (open bar), 7.13 ΔcagA, 7.13 ΔcagM or 7.13 Δslt (shaded bars; all biopsies from the same patient). RNA was extracted from biopsies and the HKα mRNA content was measured by reverse transcription—PCR (means, SD, n=3; ***p<0.001, **p<0.01, *p<0.1).

Figure 5

cag pathogenicity island (PAI) genes markedly downregulate H,K-ATPase α subunit (HKα) expression in human gastric mucosa. Gastric biopsies were infected (1—2×10⁵ bacteria/mg wet weight biopsy, 24 h) with wild-type (wt) H pylori strains 7.13, 60190 or P12 or the corresponding isogenic mutant strains 7.13 ΔcagPAI, 60190 ΔcagA, 7.13 ΔcagM, 7.13 ΔcagE, P12 ΔcagL or 60190 ΔvacA. Biopsies were lysed and HKα protein content was assessed by immunoblotting using antibody HK 12.18 against HKα and β-actin antibody as gel loading control (representative gels of three individual patient replicates).

Figure 6

Gastric acid secretion by human gastric mucosal biopsies is inhibited by wild-type H pylori strains. Gastric biopsies were infected with wild-type H pylori strain 60190, P12 or P12 ΔcagL (1—2×10⁵ bacteria/mg wet weight biopsy, 15 h) or Brucella broth alone, and then incubated with 150 μM 5-(N-ethyl-n-isopropyl) amiloride (EIPA) with or without 50 μM SCH28080 for 30 min. The pH of the medium bathing the biopsies was measured continuously for 3 min at 7 min intervals. After 21 min, histamine (or vehicle) was added to some biopsies to a 1 mM final concentration (arrows). (A) Changes in extracellular pH of three representative biopsies are shown as a function of time. Open squares, biopsy treated with vehicle alone; open triangles, biopsy treated with histamine; and open circles, biopsy infected with wt H pylori strain 60190 and treated with histamine. (B) The slopes of the initial rates of change of extracellular pH of medium bathing the biopsies as shown in A were transformed for buffer capacity to yield the biopsy proton production rate (PPR; pmol H⁺/min). Open squares, biopsies treated with vehicle alone; open triangles, biopsies treated with histamine; and open circles, biopsies infected with wild-type H pylori strain 60190 and treated with histamine. (C) Open triangles, biopsy treated with histamine in the absence of SCH28080; open diamonds, biopsy treated with 50 μM SCH28080 for 30 min and then treated with histamine. (D) Open triangles, biopsy treated with histamine; open diamonds, biopsy infected with wild-type H pylori strain P12; asterisks, biopsy infected with the H pylori P12 ΔcagL isogenic mutant. Data points in (B) and (C) are the mean PPR ±SD, n=3 biopsies; data points in (D) are measurements of single biopsies.