ADP-heptose attenuates Helicobacter pylori-induced dendritic cell activation

Abstract

Sophisticated immune evasion strategies enable Helicobacter pylori (H. pylori) to colonize the gastric mucosa of approximately half of the world's population. Persistent infection and the resulting chronic inflammation are a major cause of gastric cancer. To understand the intricate interplay between H. pylori and host immunity, spatial profiling was used to monitor immune cells in H. pylori infected gastric tissue. Dendritic cell (DC) and T cell phenotypes were further investigated in gastric organoid/immune cell co-cultures and mechanistic insights were acquired by proteomics of human DCs. Here, we show that ADP-heptose, a bacterial metabolite originally reported to act as a bona fide PAMP, reduces H. pylori-induced DC maturation and subsequent T cell responses. Mechanistically, we report that H. pylori uptake and subsequent DC activation by an ADP-heptose deficient H. pylori strain depends on TLR2. Moreover, ADP-heptose attenuates full-fledged activation of primary human DCs in the context of H. pylori infection by impairing type I IFN signaling. This study reveals that ADP-heptose mitigates host immunity during H. pylori infection.

Keywords: ADP-heptose; Dendritic cells; H. pylori; TLR2; type I IFN.

Figure 1.

Dendritic cells are recruited and activated upon Helicobacter pylori infection. (a) Immunohistochemical and immunofluorescence staining of FFPE sections of gastric biopsies from three H. pylori negative (Hp⁻) and three H. pylori positive (Hp⁺) gastritis patients. One representative out of 3 is shown. Five regions of interest (ROIs) per patient were harvested, selected ROIs are outlined with white lines in the lower panel. (b) Fluorescence staining of gastric biopsies using PanCK for epithelial cells (green), CD45 to identify immune cells (red) and Syto13 nuclear staining (blue). (c) Graphical depiction of GeoMx technology. CD45-positive areas were exposed to UV light to cleave DNA tags and oligos were quantified using the nCounter system. (d) CD45⁺ areas of 15 ROIs from Hp⁻ and Hp⁺ biopsies (3 patients, 5 ROIs per patient) were harvested and analyzed using the nCounter system. Fold change vs. significance of all tested markers in CD45⁺ regions comparing Hp⁺ and Hp⁻ gastritis patients are shown. (e) Pearson correlation of all tested markers in the CD45⁺ regions of Hp⁺ positive gastritis patients. Only significant correlations (p ≤ 0.05) are depicted). Red/blue annotations indicate positive/negative correlations, respectively. (f) Correlation of normalized counts of CD11c and CD4 or HLA-DR, respectively. Dots represent values of 15 individual ROIs. (g) Schematic depiction of the mucosoid/DC co-culture model. The mucosoid is cultured on a filter and infected with H. pylori, and DCs are cultivated in a drop of matrigel on the basolateral side. Created in BioRender. Neuper, T. (2022) BioRender.com/u31t534 (h,i) Surface marker expression of DCs cultured in the mucosoid/DC model (h) or in DC single culture (i) upon H. pylori infection. One representative donor of 2 (h) and 4 (i) individual donors is shown.

Figure 2.

Lack of ADP-heptose increases bacteria-induced Th1-associated immune responses. (a-c) Human CD1c⁺ DCs were infected with H. pylori (wt) or the ADP-heptose deficient mutant ΔrfaE (Δ) at a MOI of 5 for 16 h and cytokine (a) and chemokine (b) as well as surface marker expression (c) was analyzed by multiplex and flow cytometry, respectively (n = 8-21). (d) Extracted ion current chromatograms of ADP-heptose (m/z 618.0850) in bacterial lysates and in a commercially available standard at a concentration of 100 pg/µl. One out of two experiments is shown. (e) DCs were infected with H. pylori (Hp) or A. lwoffii (Al) at an MOI of 5 and IL-12p70 secretion and CD40 expression were analyzed 16 h post infection (n = 4). (f,g) DCs were infected with H. pylori (Hp wt), the ADP-heptose deficient mutant (Hp Δ) or A. lwoffii (Al) (MOI5). After 16 h allogenic, naive CD4⁺ T cells were added and DCs and T cells were co-cultured for another 6 days, before CD4⁺/IFNγ⁺ T cells were quantified by flow cytometry (n = 4). (h) Graphical depiction of a mucosoid/DC co-culture, and a DC/T cell co-culture. Created in BioRender. Neuper, T. (2022) BioRender.com/u31t534 (i) CXCL8 mRNA expression was analyzed in the mucosoid 40 h after infection with H. pylori (wt) (MOI 100) or the ADP-heptose deficient mutant (Δ) (n = 2). (j) IFNγ production was analyzed by flow cytometry in CD4⁺ T cells co-cultured with DCs re-isolated from the mucosoid/DC co-culture. One out of two independent donors is shown. Bars indicate mean±SD, dots represent individual donors. For statistical analysis repeated measures one-way ANOVA with a šídák’s post-hoc test was performed.

Figure 3.

ADP-heptose does not activate dendritic cells but reduces pathogen-induced DC activation. (a) Human DCs were incubated with ADP-heptose (25 µg/ml) or E. coli LPS (100 ng/ml) for 16 h and cytokine/chemokine mRNA expression was analyzed by quantitative real-time PCR (n = 6). (b) DCs were treated with Takinib (1 or 10 µM) for 20 min prior to ADP-heptose treatment (25 µg/ml) and CXCL8 secretion was analyzed after 16 h by multiplex assay (n = 2). (c,d) DCs were infected with H. pylori wildtype or the ADP-heptose deficient mutant ΔrfaE (MOI 5) as well as with 2.5 or 25 µg/ml ADP-heptose. IL-12p70 secretion (c) and CD40 expression (d) was monitored after 16 h by multiplex analysis and flow cytometry, respectively (n = 4). (e,f) DCs were infected with A. lwoffii (MOI 5) as well as 2.5 or 25 µg/ml ADP-heptose and IL-12p70 secretion (e) and CD40 expression (f) were analyzed 16 h post infection (n = 4). (g) DCs were infected with H. pylori (wt) or the ADP-heptose deficient mutant (Δ) (MOI 5) supplemented with ADP-heptose (25 µg/ml) in the absence or presence of Takinib (10 µM, grey background) (n = 2). Bars indicate mean±SD, dots represent individual donors. For statistical analysis repeated measures one-way ANOVA with a šídák’s post-hoc test was performed.

Figure 4.

TLR2 mediated Helicobacter pylori uptake is essential for potent dendritic cell activation. (a) TLR2 mRNA expression upon stimulation with H. pylori wt or mutant (Δ) (MOI 5) at indicated time points. Mean±SD of three individual donors is shown. (b) TLR2 surface expression was monitored by flow cytometry 16 h post-infection (MOI 5) (n = 4). (c,d) H. pylori strains were stained with eFluor670 proliferation dye prior to infection. One hour post infection with H. pylori wt or the ADP-heptose-deficient mutant (Δ) (MOI 5), DCs were subjected to immunofluorescence and stained for CD45, DAPI and TLR2. Internalization of bacteria as well as TLR2 localization was analyzed by confocal fluorescence microscopy. Orthogonal views of confocal z-stacks of one out of three representative donors are shown. Scale bar: 5 µm. (e) DCs were treated with a TLR2 neutralizing antibody 20 min prior to infection. After 1 h of infection with eFluor670 stained H. pylori (MOI 5), DCs were subjected to immunofluorescence and stained for CD45 and DAPI. Maximum intensity projections of confocal z-stacks of one representative out of three donors are shown. Scale bar: 5 µm. (f,g) DCs were treated with a TLR2 neutralizing antibody 20 min prior to infection with H. pylori (wt) or the ADP-heptose deficient mutant (Δ) at an MOI of 5. After 16 h CD40 expression (f) and IL-12p70 secretion (g) was monitored by flow cytometry and multiplex assay (n = 6). For statistical analysis one-way ANOVA with a šídák’s post-hoc test was performed.

Figure 5.

ADP-heptose attenuates dendritic cell activation by suppressing type I IFN signaling. (a) Heatmap showing z-scores of normalized intensities for differentially expressed proteins in DCs upon infection with H. pylori (wt) or the ADP-heptose deficient mutant (Δ) (MOI 5). Red/blue annotation for the direction of change indicates up/down regulation of differentially expressed proteins present in the comparison of Δ vs. wt, respectively (n = 3). (b) Volcano plot displaying proteins in the comparison of H. pylori mutant infected DCs (Δ) vs. H. pylori wildtype infected DCs (wt). The dashed horizontal line indicates the p_adj cutoff < 0.05. Black dots represent single proteins and red large dots indicate proteins of interest (n = 3). (c) Bar graph with the normalized enrichment score, showing pathways enriched in DCs infected with the ADP-heptose deficient H. pylori mutant (Δ) (positive NES score) or enriched in H. pylori wt (negative NES score) (n = 3). (d) DCs were infected with H. pylori (wt) or the ADP-heptose deficient mutant (Δ) (MOI 5) and type I IFN signaling was monitored. Using Western Blot STAT1 and STAT2 phosphorylation was analyzed, as well as total protein levels of the ISGF3 components (STAT1, STAT2, IRF9) after 16 h. One out of five representative donors is shown. (e) IFNA2 and ISG15 mRNA expression was monitored by qPCR upon infection (MOI 5) at the indicated time points. Mean±SD of three individual donors is shown. (f) Secretion of IFNα and CXCL10 was monitored after 16 h of infection with the indicated strains (n = 14). (g) DCs were infected with H. pylori wildtype (wt) or the ADP-heptose deficient mutant ΔrfaE (MOI 5) and stimulated with 2.5 or 25 µg/ml ADP-heptose. Cytokine secretion was analyzed by multiplex assay (n = 3). (h) H. pylori (MOI 5) -induced type I IFN signaling (STAT1, STAT2, IRF9) was monitored upon addition of ADP-heptose (25 µg/ml) by Western Blot analysis. One representative out of three donors is shown. (i) DCs were infected with H. pylori (wt) and supplemented with IFNα and IFNβ (10 µg/ml each) or infected with the ADP-heptose deficient mutant (Δ) and IL-12p70 was analyzed after 16 h (n = 4). (j) Histograms of HLA-ABC surface expression after 16 h of infection with H. pylori wt or ΔrfaE (MOI 5) analyzed by flow cytometry. One representative out of three donors is shown. (k) Heatmap displaying log2FC of wt and ΔrfaE infected DCs (MOI 5) compared to untreated DCs. Red annotation indicates upregulation of the indicated proteins. (n = 3). (l) Peripheral blood mononuclear cells (PBMCs) were infected with H. pylori (wt) or the ADP-heptose deficient mutant (Δ) at a MOI of 0.1 and granzyme B (GZMB) secretion was monitored by ELISA after 6 days (n = 7). (m) DCs were infected with the indicated H. pylori strains (MOI 5) for 16 h, before allogenic pan T cells were added. After 6 days of co-culture granzyme B (GZMB) production was monitored using flow cytometry (n = 8). (n) Correlation of normalized counts of CD8 and granzyme B (GZMB) in H. pylori negative (Hp⁻, left panel) and H. pylori positive (Hp⁺, right panel) gastritis samples. Dots represent values of individual ROIs. Bars indicate mean±SD, dots represent individual donors. For statistical analysis Student’s T-test or repeated measures one-way ANOVA with a šídák’s post-hoc test was performed.