Targeting the STING pathway in tumor-associated macrophages regulates innate immune sensing of gastric cancer cells

Abstract

Rationale: STING is a critical player in the innate and adaptive immune system, sensing cytosolic DNA to activate the expression of interferon genes and regulate T lymphocytes, which drives immunogenic responses to cancer cells. Tumor-associated macrophages (TAMs), abundantly present in the tumor microenvironment, play a key role in cancer development. Gastric cancer is one of the most common and leading causes in cancer-related death worldwide. However, studies on the function and regulation of STING in TAMs and their roles in gastric cancer progression are still limited. Methods: We analyzed STING and CD68 expression of 200 pairs of gastric cancer and adjacent normal tissues by immunohistochemistry to identify the prognostic values of STING, as well as the correlations between STING and CD68 in gastric cancer. The characteristics of STING-altered macrophages, as well as their effects on cancer cell apoptosis and T cell differentiation were examined by flow cytometry. Cytokines secreted by STING-altered macrophages were identified by the Human Inflammation Array3 kit. Concentrations of soluble IL24 and IFN-β were measured by ELISA. In vivo models, including spontaneous gastric cancer in p53^+/- mice and cell line-based xenografts, were established, and clinical benefits of STING-altered macrophages were examined. Results: Our study identifies STING as a prognostic factor for gastric cancer, and for the first time demonstrated that knocking-down STING and STING activation by 2'3'-c-GAMP both promote TAMs polarizing into pro-inflammatory subtype and induce apoptosis of gastric cancer cells, mechanistically through IL6R-JAK-IL24 pathway. Conclusions : This study evaluated effects of targeting STING in TAMs in anti-gastric-cancer therapies. Moreover, we unveil a novel function of STING to activate the IL6R-JAK-IL24 pathway in macrophages.

Keywords: Gastric carcinoma; IL-24; IL-6R; STING; apoptosis.

Figure 1

High STING expression is enriched in macrophages, predicting poor survival of gastric cancer patients. (A) Left panel, immune-score of STING expression in normal mucosa and gastric tumors; right panel, representative pictures of STING IHC staining in adjacent normal mucosa and tumor tissue of a GC patient. ***, p < 0.001. (B) Kaplan-Meier analysis showing overall survival of GC patients with high vs. low STING expression. The STING intensity that can best separate the low and high groups is used as the cut-off. (C) Upper panel, immunoblot analysis showing STING expression in indicated cells. Vinculin was used as a loading control. Lower panel, the ratio of STING/Vinculin was quantified, and statistical significance was analyzed by comparing to the monocytes. Data are presented as the mean±SD (n=3).*, p < 0.05; **, p < 0.01; ***, p < 0.001. (D) Left panel, immune-score of CD68 expression in normal mucosa and gastric tumors; right panel, representative pictures of CD68 IHC staining in adjacent normal mucosa and tumor tissue of a GC patient. ***, p < 0.001. (E) Correlation analysis showing expression of STING and CD68 in the 200 pairs of adjacent normal mucosa and gastric cancer samples as in (A) and (D). (F) Immunofluorescent staining of STING (red) and CD68 (green) of a human GC sample. DAPI (blue) stained for nuclei. Scale bar represents 100 µm.

Figure 2

Knocking-down STING and STING activation promote both PBMC-DMs and BM-DMs differentiating into pro-inflammatory subtype. (A) Upper panel, flow cytometric analysis of TAMs for pro-inflammatory (CD45⁺CD11b⁺CD80⁺) and anti-inflammatory (CD45⁺CD11b⁺CD163⁺) subtypes; Lower panel, quantification of pro-inflammatory, anti-inflammatory macrophages and the corresponding ratios (n=8); *, p < 0.05; **, p < 0.01. (B, C) Representative flow cytometric analysis of pro-inflammatory (CD11b⁺/CD80⁺ in human (B) and mouse(C)) and anti-inflammatory macrophages (CD11b⁺/CD163⁺ in human (B) and CD11b⁺/CD206⁺ in mouse(C)) in human PBMC-DMs from two healthy donors (B) and mouse BM-DMs (C) treated as indicated. (D, E) Quantifications of pro-inflammatory and anti-inflammatory macrophages in human PBMC-DMs (D) and mouse BM-DMs (E) as in (B) and (C), respectively. ***, p < 0.001; **, p < 0.01; *, p < 0.05. (F, G) RT-PCR analysis of the macrophage subtype markers in human PBMC-DMs and mouse BM-DMs treated as indicated; Data in D, E, F and G are presented as the mean±SD (n=3).***, p < 0.001; **, p < 0.01; *, p < 0.05. Control stands for the average of undistinguishable controls of scrambled sequence (SC; for shSTING), empty vector (EV; for STING overexpression), and PBS treatment (for 2'3'-c-GAMP treatment).

Figure 3

STING knock-out and activation in THP1-derived macrophages promote pro-inflammatory subtype differentiation. (A) Left panel, immunoblot analysis of STING expression in human THP1-derived macrophages treated as indicated. α-Tubulin was used as a loading control; right panel, the ratio of STING/α-Tubulin was quantified; ***, p < 0.001; *, p < 0.05. (B) Luciferase activity driven by the IFN-β promoter in THP1-derived macrophages treated as indicated. ***, p < 0.001. (C) Flow cytometric analysis of surface marker expression for pro-inflammatory macrophages (CD11b⁺CD80⁺) in human THP1-derived macrophages treated as indicated. *, p < 0.05. (D) RT-PCR analysis of the macrophage-subtype markers in stable THP1-derived macrophages treated as indicated. ***, p < 0.01; p < 0.01; *, p < 0.05. Data are presented as the mean±SD (n=3). Control stands for the average of undistinguishable controls of scrambled sequence (SC; for shSTING), and PBS treatment (for 2'3'-c-GAMP treatment), empty vector (EV) was used as a control of STING overexpression.

Figure 4

Macrophages with knocking-down STING or STING activation have apoptoic effects on gastric cancer cells. (A) Colony formation assay of human HGC-27 GC cells cocultured with human PBMC-DMs treated as indicated. (B) Left panel, representative flow cytometric plots of apoptosis markers (Annexin V⁺) in HGC-27 cells cocultured with human PBMC-DMs treated as indicated; right panel, quantification of percentage of Annexin V⁺ apoptotic cells. ***, p < 0.001; **, p < 0.01; *, p < 0.05. (C) Colony formation assay of mouse MFC GC cells cocultured with mouse BM-DMs treated as indicated. (D) Left panel, representative flow cytometric plots of apoptosis markers (Annexin V⁺) in MFC cells cocultured with mouse BM-DMs treated as indicated; right panel, quantification of percentage of Annexin V⁺ apoptotic cells. **, p < 0.01; *, p < 0.05. (A-D) Control stands for the average of undistinguishable controls of scrambled sequence (for shSTING), empty vector (for STING overexpression), and PBS treatment (for 2'3'-c-GAMP treatment). (E) Colony formation assay of human HGC-27 cells cocultured with human THP1-derived macrophages treated as indicated. (F) Left panel, representative flow cytometric plots of apoptosis markers (Annexin V⁺) in HGC-27 cells cocultured with human THP1-derived macrophages treated as indicated; right penal, quantification of percentage of Annexin V⁺ apoptotic cells. **, p < 0.01; *, p < 0.05. Data in B, D, F are presented as the mean±SD (n=3). (E, F) Control stands for the average of undistinguishable controls of scrambled sequence (SC; for shSTING), and PBS treatment (for 2'3'-c-GAMP treatment), empty vector (EV) was used as a control of STING overexpression.

Figure 5

STING knocking-down and activation regulate macrophage differentiation through JAK-IL24 pathway. (A) Representative pictures of IL6R in a cytokine array analysis of supernatants from PBMC-DM cultures treated as indicated, and quantified in (B); ***, p<0.001; **, p< 0.01. (C) Flow cytometric analysis of surface marker expression for pro-inflammatory macrophages (CD11b⁺/HLA-DR⁺) and anti-inflammatory macrophages (CD11b⁺/CD163⁺) in human PBMC-DMs treated as indicated. ***, p < 0.001; **, p < 0.01; *, p < 0.05. (D) RT-PCR analysis of the macrophage subtype-markers in human PBMC-DMs treated as indicated. ***, p < 0.001; **, p < 0.01; *, p < 0.05. (E) Left panel, representative flow cytometric plots of apoptosis markers (FITC-Annexin V⁺) in human PBMC-DMs treated as indicated; right panel, quantification of percentage of Annexin V⁺ apoptotic cells; **, p < 0.01; *, p < 0.05. (F) Immunoblot analysis of phospho-STAT1, total STAT1, phospho-STAT3, total STAT3, IL24 expression in PBMC-DMs from two healthy donors treated as indicated; β-Actin was used as a loading control; quantification of the immunoblot analysis of PBMC-DMs from each healthy donor was below the bands; ***, p < 0.001; **, p < 0.01; *, p < 0.05. (G) Immunofluorescent staining of IL24 (green) in human PBMC-DMs. DAPI (blue) stained for nuclei. Scale bar represents 50 µm. (H) IL24 expression measured by ELISA in supernatants of human PBMC-DMs (left) or mouse BM-DMs (right) cultures treated as indicated. ***, p < 0.001; **, p < 0.01; *, p < 0.05. (I) Left panel, representative flow cytometric plots of apoptosis markers (FITC-Annexin V⁺) in human PBMC-DMs treated as indicated; right panel, quantification of percentage of Annexin V⁺ apoptotic cells; **, p < 0.01; *, p < 0.05. Data in B,C, D, E and F are presented as the mean±SD (n=3). (A-E, H) Control stands for a representative sample of undistinguishable controls of scrambled sequence (for shSTING), empty vector (for STING overexpression), and PBS treatment (for 2'3'-c-GAMP treatment). (F, G, I) Control stands for a representative sample of undistinguishable controls of scrambled sequence (for shSTING and/or shIL24) and PBS treatment (for 2'3'-c-GAMP treatment).

Figure 6

Macrophages with Sting knocking-down or activation have killing effects on cancer cells of spontaneous gastric tumors in mice. (A) Visual examination and tumor development in stomachs of mice treated as indicated. (B) Scheme showing the time course of treatments in mice. (C) Left panel, immunoblot analysis of CD68 expression in gastric tissues from mice treated as indicated; GAPDH was used as a loading control; right panel, the ratio of CD68/GAPDH was quantified, and statistical significance was analyzed by comparing to the Clodron-treated samples; **, p < 0.01; ***, p < 0.001. Control stands for a representative sample of undistinguishable controls of scrambled sequence (for shSTING) and PBS treatment (for 2'3'-c-GAMP treatment). (D) Upper panel, visual examination and tumor development in stomachs of mice treated as indicated; lower panel, representative pictures of Ki67 and CD8 IHC staining in gastric normal and tumor tissues corresponding to the upper panel. Scale bars represent 20 µm. (E) Quantification of surface marker expression for total T cells (CD3⁺) and ratio of effector/helper T cells (CD8⁺/CD4⁺) in the blood of mice treated as indicated. *, p < 0.05. (F) Quantification of surface marker expression for total T cells (CD3⁺) and ratio of effector/helper T cells (CD8⁺/CD4⁺) in the spleens of mice treated as indicated. **, p < 0.01. *, p < 0.05. Data in B, E and F are presented as the mean±SD (n=10). (D-F) Control stands for a representative sample of undistinguishable controls of scrambled sequence (for shSTING), empty vector (for STING overexpression), and PBS treatment (for 2'3'-c-GAMP treatment).

Figure 7

Macrophages with Sting knocking-down or activation have killing effects on cancer cells of xenografted gastric tumors in mice. (A) Visual examination of isolated tumors from mice injected subcutaneously with MFC cells and treated as indicated. (B) Tumor weight in mice after indicated treatments; **, p < 0.01. (C) Tumor volume in mice with indicated treatments; *, p < 0.05. Data in B,C are presented as the mean±SD (n=6). Control stands for a representative sample of undistinguishable controls of scrambled sequence (for shSting), empty vector (for Sting overexpression), and PBS treatment (for 2'3'-c-GAMP treatment).