A novel human gastric primary cell culture system for modelling Helicobacter pylori infection in vitro

Abstract

Background and aims: Helicobacter pylori is the causative agent of gastric diseases and the main risk factor in the development of gastric adenocarcinoma. In vitro studies with this bacterial pathogen largely rely on the use of transformed cell lines as infection model. However, this approach is intrinsically artificial and especially inappropriate when it comes to investigating the mechanisms of cancerogenesis. Moreover, common cell lines are often defective in crucial signalling pathways relevant to infection and cancer. A long-lived primary cell system would be preferable in order to better approximate the human in vivo situation.

Methods: Gastric glands were isolated from healthy human stomach tissue and grown in Matrigel containing media supplemented with various growth factors, developmental regulators and apoptosis inhibitors to generate long-lasting normal epithelial cell cultures.

Results: Culture conditions were developed which support the formation and quasi-indefinite growth of three dimensional (3D) spheroids derived from various sites of the human stomach. Spheroids could be differentiated to gastric organoids after withdrawal of Wnt3A and R-spondin1 from the medium. The 3D cultures exhibit typical morphological features of human stomach tissue. Transfer of sheared spheroids into 2D culture led to the formation of dense planar cultures of polarised epithelial cells serving as a suitable in vitro model of H. pylori infection.

Conclusions: A robust and quasi-immortal 3D organoid model has been established, which is considered instrumental for future research aimed to understand the underlying mechanisms of infection, mucosal immunity and cancer of the human stomach.

Keywords: GASTRIC CANCER; HELICOBACTER PYLORI.

Figure 1

Establishment of human gastric spheroid cultures. (A) The spheroid forming efficiency of human gastric glands is highly dependent on the addition of Wnt3A/R-spondin1 and Y-27632 to the basal medium. The number of spheroids per well was counted 14 days after seeding. Representative micrographs of spheroids grown in the three different culture conditions are shown in (a)–(c). Graph represents mean+SD of at least three biologically independent replicates; *** p<0.001, Student t test. (B) Representative examples of human gastric spheroids cultured from isolated human gastric glands for between 2 and 9 days (black arrows identify the same spheroid over the course of 8 days), as well as after 9 months (m). Scale bars: 500 µm.

Figure 2

Characterisation of human gastric spheroids. (A) Confocal micrographs of human gastric spheroids (cross sections), fluorescently labelled with antibodies against the epithelial marker E-cadherin (green), the polarisation marker β-catenin (red), the gland mucous cell marker MUC6 (red), the proliferation marker Ki67 (red) or the chief cell marker pepsinogen C (green); nuclei were counterstained with Draq5 (blue). Scale bars: 100 µm; and 25 µm for bottom right panel; for technical reasons smaller spheroids were chosen here compared with the organoids shown in figure 3D. (B) Expression analysis of gastric-specific and epithelial-specific genes from early (P1) and late (P8) passage spheroids, using semi-quantitative RT-PCR. Photomicrographs are presented in reversed colours. PGC, pepsinogen C; EPCAM, epithelial cell adhesion molecule; GADPH, glyceraldehyde 3-phosphate dehydrogenase.

Figure 3

Human gastric spheroids can be differentiated into organoids. (A) Representative micrographs of different stages of a single spheroid during the differentiation process induced by withdrawal of Wnt3A and R-spondin1. (B) Global overview of differential gene expression upon withdrawal of Wnt3A and R-spondin1 for 5 days as determined by microarray analysis comparing undifferentiated corpus spheroids with differentiated corpus organoids. (C) Expression changes for a selection of gastric-specific and differentiation-specific genes highlighted in the global overview in (B) provided by microarray analysis. For genes with multiple probes, those with the largest difference were chosen. (D) Confocal micrographs of human corpus organoids (cross sections), fluorescently labelled with antibodies against the epithelial marker E-cadherin (green), the polarisation marker β-catenin (red), the pit mucous cell maker MUC5A (red) or the proliferation marker Ki67 (red); nuclei were counterstained with Draq5 (blue). Scale bars: 100 µm.

Figure 4

Region-specific molecular differences are maintained in three dimensional cultures from antrum and corpus. (A) Representative micrographs showing spheroid cultures growing with the same efficiency from isolated gastric glands from the corpus and antrum region of the human stomach. Scale bars: 500 µm. (B) Global overview of differential gene expression upon withdrawal of Wnt3A and R-spondin1 for 5 days as determined by microarray analysis comparing undifferentiated antrum spheroids with differentiated antrum organoids. (C) Expression changes for a selection of gastric-specific and differentiation-specific genes as determined by microarray analysis highlighted in the global overview in (B) provided by microarray analysis. (D) Region-specific differences in expression of the region-specific markers. Upregulated pepsinogen C (PGC) and ghrelin (GHRL) expression is elevated in corpus-derived cultures, while expression of gastrin (GAST), villin1 (VIL1) and CDX2 is elevated in antrum-derived cultures.

Figure 5

Human gastric spheroids represent an unlimited source of planar two dimensional (2D) gastric primary cells. (A) Representative photomicrographs of three dimensional (3D) human gastric spheroids from corpus cultures during conversion to a 2D monolayer of adherent primary cells. Scale bars: 500 µm (B) Global overview of differential gene expression as determined by microarray analysis comparing 3D human gastric spheroids with 2D adherent gastric primary cells. (C) Expression changes for a selection of gastric-specific and differentiation-specific genes as determined by microarray analysis highlighted in the global overview in (B) provided by microarray analysis. (D) Confocal micrographs of adherent human gastric primary cell cross-sections fluorescently labelled with antibodies against the epithelial marker E-cadherin (green), the polarisation marker β-catenin (red), the tight junction marker occludin (red) and the pit mucous cell maker MUC5A (red); nuclei were counterstained with Draq5 (blue). Scale bars: 50 µm.

Figure 6

Planar human gastric cell cultures can be successfully infected with Helicobacter pylori. (A) Representative micrographs of adherent gastric primary cells (a) non-infected, (b) infected with the human-adapted H. pylori strain P12 (MOI 100) or (c) infected with the P12 mutant P12ΔPAI, which is missing the pathogenicity island (MOI 100) for 22 h. Scale bars: 100 µm. (B) Western blotting analysis of non-infected and infected adherent gastric primary cells 22 h after infection, using antibodies against the virulence factor CagA and phospho-tyrosine 99, which detects phosphorylated CagA. β-actin was used as loading control. (C) Global overview of differential gene expression as determined by microarray analysis comparing H. pylori-infected two dimensional primary cells (MOI 100, 3 h infection) generated from corpus cultures with the non-infected control. (D) Expression changes for a selection of strongly upregulated genes as determined by microarray analysis highlighted in the global overview in (C) provided by microarray analysis (E) Gene set enrichment analysis of upregulated genes identified by microarray analysis after 3 h infection with H. pylori (MOI 100) correlated with publically available data from H. pylori-upregulated genes in infected corpus tissue sample. The analysis reveals a significant enrichment of genes upregulated in infected corpus tissue among those genes upregulated in infected 2D primary cells compared with non-infected primary cells. ES, enrichment score; NES, normalised enrichment score; FDR, false discovery rate.