Autophagy mediates ER stress and inflammation in Helicobacter pylori-related gastric cancer

Abstract

Autophagy is a cellular degradation mechanism, which is triggered by the bacterium Helicobacter pylori. A single nucleotide polymorphism (SNP) in the autophagy gene ATG16L1 (rs2241880, G-allele) has been shown to dysregulate autophagy and increase intestinal endoplasmic reticulum (ER) stress. Here, we investigate the role of this SNP in H.pylori-mediated gastric carcinogenesis and its molecular pathways. ATG16L1 rs2241880 was genotyped in subjects from different ethnic cohorts (Dutch and Australian) presenting with gastric (pre)malignant lesions of various severity. Expression of GRP78 (a marker for ER stress) was assessed in gastric tissues. The effect of ATG16L1 rs2241880 on H.pylori-mediated ER stress and pro-inflammatory cytokine induction was investigated in organoids and CRISPR/Cas9 modified cell lines. Development of gastric cancer was associated with the ATG16L1 rs2241880 G-allele. Intestinal metaplastic cells in gastric tissue of patients showed increased levels of ER-stress. In vitro models showed that H.pylori increases autophagy while reducing ER stress, which appeared partly mediated by the ATG16L1 rs2241880 genotype. H.pylori-induced IL-8 production was increased while TNF-α production was decreased, in cells homozygous for the G-allele. The ATG16L1 rs2241880 G-allele is associated with progression of gastric premalignant lesions and cancer. Modulation of H.pylori-induced ER stress pathways and pro-inflammatory mediators by ATG16L1 rs2441880 may underlie this increased risk.

Keywords: ATG16L1; ER stress; Helicobacter pylori; Intestinal metaplasia; atrophic gastritis; autophagy; gastric cancer; inflammation.

Figure 1.

ER stress (GRP78) is increased in gastric IM tissues (taken by forceps biopsy from the gastric antrum). Representative pictures of antral biopsies (n = 23) immunohistochemically stained for GRP78 (brown). (a) Normal epithelium of the gastric antrum, showing no IM. (b) Representation of a biopsy including mixed normal epithelium (left) and IM (right). (c) Quantitative analyses showing that IM tissues of the gastric antrum express more GRP78 than the normal epithelium as measured by Allred score, Statistical analysis was carried out using parametric t-test (two-tailed) with errors bars representing SEM (d) Magnification of individual crypt showing strong GRP78 staining located mainly in an intestinal Paneth cell, representing an IM patch within the biopsy. IM: intestinal metaplasia

Figure 2.

ER stress (GRP78) in gastric IM epithelium is not influenced by ATG16L1 rs2241880 status. (a) Representative pictures of normal epithelium (upper panels) and intestinal metaplastic tissues (IM, lower panels), immunohistochemically stained for GRP78. Representative examples of ATG16L1 rs2241880 AA, AG and GG carriers are shown for both normal and IM tissues. (b) Quantification of Allred scores separated according to ATG16L1 rs2241880 genotype. Statistical analysis was carried out using parametric t-test (two-tailed) with errors bars representing SEM, (c) Western blot analysis of unstimulated (basal) GRP78 levels and LC3B expression in H. pylori stimulated organoids derived from intestinal metaplasia (IM) patients. (d) Quantification of Western blot analysis of GRP78 expression and LC3BII/I ratio normalized against the expression of β-actin in H. pylori stimulated organoids derived from IM patients. A* indicates AA and AG allele carriers. IM: intestinal metaplasia, AA; wild-type homozygote, AG; heterozygote, GG; mutant homozygote

Figure 3.

Helicobacter pylori reduces ER stress levels in IM tissues. (a-d) GRP78 (a, c) and LC3II expression (b, d) determined by Western blot analysis in GES-1 normal gastric epithelial cells (a, b) and SK-GT-2 gastric cancer cell line (c, d). Representative example as well as quantification of 4 individual experiments is shown. GRP78 expression and LC3B II/I ratio relative to B-actin and normalized against unstimulated sample, are shown. Statistical analysis was carried out using Mann-Whitney test (two-tailed) with errors bars representing SEM (e) Representative images of GRP78 staining of antral biopsies from a patient with active infection with H. pylori (E.I), the same patient 1 year after eradication of H. pylori (HP) (E.II), and an antral biopsy from a second patient who has not encountered H. pylori (E.III). (f) Bar graph of quantified Allred scores of the average of biopsies stained for GRP78 during infection (n = 24), after eradication (n = 23) and in patients who have never been infected (n = 23). IM: intestinal metaplasia

Figure 4.

Helicobacter pylori-induced ER stress reduction may be correlated to ATG16L1 rs2241880 status. (a) Immunohistochemical staining for GRP78 on IM organoids from the same patient (AA) and same passage number, stimulated with tunicamycin for 16 hours. (b) Western blot analysis for GRP78 of IM organoid samples (n = 3, GT: AA) stimulated with increasing concentrations of tunicamycin for 16 hours. (c) Western blot analysis for GRP78 and LC3B of IM organoid samples from patients with differing ATG16L1 rs2241880 genotypes. (d) Bar graph of quantified Western blot analysis for GRP78 and LC3B II/I ratio relative to β-actin and normalized against unstimulated samples. Statistical analysis was carried out using Mann-Whitney test (two-tailed) with errors bars representing SEM. A* denotes AA/AG genotypes. IM: intestinal metaplasia, AA; wild-type homozygote, AG; heterozygote, GG; mutant homozygote

Figure 5.

ATG16L1 rs2241880 status does not affect IL-8 production in immune cells but modulates IL-8 and TNF-α production in AGS cells upon infection with Helicobacter pylori. (a) Expression of IL8 mRNA in H. pylori strain GC026-challenged edited and non-edited AGS cells. (b) Concentration of IL-8 in H. pylori strain GC026-challenged edited and non-edited AGS cells. (c) Isolated PBMCs derived from healthy Dutch volunteers (AA n = 4; AG n = 6; GG n = 4) release substantial amounts of IL-8 after 2 hours of stimulation with H. pylori, but this is not affected by donor ATG16L1 rs2241880 status. (d) Expression of IL18 mRNA in H. pylori strain GC026-challenged edited and non-edited AGS cells. (e) Expression of TNFA mRNA in H. pylori strain GC026-challenged edited and non-edited AGS cells. (f) Concentration of TNF-α in H. pylori strain GC026-challenged edited and non-edited AGS cells. Experiments with the AGS cells were conducted in triplicates. Statistical analysis was carried out using parametric t-test (two-tailed) with errors bars representing SEM. AA; wild-type homozygote, AG; heterozygote, GG; mutant homozygote

Figure 6.

Graphical representation of ATG16L1 rs2241880 effect on gastric carcinogenesis. (a) According to Correa’s cascade, gastric carcinogenesis starts by infection with H. pylori. While within the general population carriers of the ATG16L1 rs2241880 G-allele are believed to be more prone to H. pylori infection, A-allele carriers are also infected. (b) In these A-allele carriers, H. pylori-evokes increased autophagy and TNF-α production, and less IL-8 production and endoplasmic reticulum (ER) stress, as compared to G-allele carriers, progressing only to mild intestinal metaplasia (IM). Enhanced IL-8 production and ER stress contribute to enhanced severity of gastric lesions in G-allele carriers. (c) While H. pylori is a leading cause in the carcinogenic process, this bacterium is generally not seen in established IM and gastric cancer. Thus, progression of IM to cancer may not rely solely on H. pylori-stimulated effects at these stages of the Correa sequence. As IM and cancer in themselves are prone to development of ER stress due to deregulated cellular processes, increased ER stress may no longer be dependent on ATG16L1 rs2241880. However, H. pylori-independent epithelial IL-8 production is enhanced in ATG16L1 rs2241880 G-allele carriers, which may further drive the carcinogenic process. AA; wild-type homozygote, AG; heterozygote, GG; mutant homozygote