STING-Dependent Cytosolic DNA Sensing Promotes Radiation-Induced Type I Interferon-Dependent Antitumor Immunity in Immunogenic Tumors

Abstract

Ionizing radiation-mediated tumor regression depends on type I interferon (IFN) and the adaptive immune response, but several pathways control I IFN induction. Here, we demonstrate that adaptor protein STING, but not MyD88, is required for type I IFN-dependent antitumor effects of radiation. In dendritic cells (DCs), STING was required for IFN-? induction in response to irradiated-tumor cells. The cytosolic DNA sensor cyclic GMP-AMP (cGAMP) synthase (cGAS) mediated sensing of irradiated-tumor cells in DCs. Moreover, STING was essential for radiation-induced adaptive immune responses, which relied on type I IFN signaling on DCs. Exogenous IFN-? treatment rescued the cross-priming by cGAS or STING-deficient DCs. Accordingly, activation of STING by a second messenger cGAMP administration enhanced antitumor immunity induced by radiation. Thus radiation-mediated antitumor immunity in immunogenic tumors requires a functional cytosolic DNA-sensing pathway and suggests that cGAMP treatment might provide a new strategy to improve radiotherapy.

Copyright © 2014 Elsevier Inc. All rights reserved.

Figures

Figure 1. STING signaling is required for the antitumor effect of radiation

MC38 tumors established in WT mice and KO mice were treated locally with one dose of 20Gy ionizing radiation (IR) or untreated. (A) 500μg anti-IFNAR1 was administered intratumorally in WT mice on day 0 and 2 after radiation. Tumor growth was monitored after radiation. (B) Tumor growth in WT and Myd88−/− mice after radiation. (C) Tumor growth in WT and Trif−/− mice after radiation. (D) 200μg anti-HMGB1 was administered i.p. in WT mice with tumors on day 0 and 3 after radiation. Tumor growth was monitored after radiation. (E) Tumor growth in WT and Camp−/− mice after radiation. (F) Tumor growth in WT and Tmem173−/− mice after radiation. STING-deficient mice are represented by Tmem173−/−, whereas CRAMP-deficient mice are represented by Camp−/−. Representative data are shown from three (A-F) experiments conducted with 5 (A-D) or 6 to 8 (E-F) mice per group. Data are represented as mean ± SEM. *P < 0.05, **P < 0.01 and ns No significant difference (Student's t test). See also Figure S1.

Figure 2. STING signaling is essential for IFN-β induction by radiation

(A-B) Tumors were excised on day 3 after radiation and homogenized in PBS with protease inhibitor. After homogenization, Triton X-100 was added to obtain lysates. ELISA assay was performed to measure IFN-β (A) and CXCL10 (B). (C) 72 hours after radiation, single cell suspensions from tumors in WT and Tmem173−/−mice were stained with 7-AAD and conjugated antibodies against CD45, CD11c and CD11b, and then sorted into different cell populations by flow cytometry. Ifn-β mRNA in different cell subsets was quantified by real-time PCR assay. STING-deficient mice are represented by Tmem173−/−. Representative data are shown from three experiments conducted with 4 mice per group. Data are represented as mean ± SEM. *P<0.05, **P < 0.01 and ***P < 0.001 (Student's t test). See also Figure S2.

Figure 3. STING-IRF3 axis in DCs is activated by irradiated-tumor cells

(A-C) BMDCs were cultured with 40Gy-pretreated MC38-SIY cells or non-irradiated-MC38-SIY cells. Subsequently purified CD11c+ cells were co-cultured with isolated CD8+ T cells from naive 2C mice for three days and analyzed by ELISPOT assays. (A) BMDCs from WT or Tmem173−/− mice were used for co-culture with irradiated or non-irradiated MC38-SIY cells. DC cross-priming activity was analyzed by ELISPOT assays. (B) BMDCs from WT or Irf3−/− mice were used for co-culture with irradiated or nonirradiated MC38-SIY cells. DC cross-priming activity was analyzed by ELISPOT assays. (C) WT and Tmem173−/− BMDCs were cultured with 40Gy-pretreated MC38-SIYhi cells. 10ng/ml IFN-β was added into the co-culture of Tmem173−/− BMDC and irradiated-MC38-SIY cells. 100μg/ml DMXAA was added to isolated Tmem173−/− CD11c+ cells for additional three hours incubation. DC cross-priming activity was analyzed by ELISPOT assays. (D) WT and Tmem173−/− BMDCs were co-cultured with 40Gy-pretreated MC38-SIY cells. The purified CD11c+ cells were incubated for additional two days and the supernatants were collected to measure IFN-β by ELISA assay. STING-deficient mice are represented by Tmem173−/−. Representative data are shown from three (A-D) experiments. Data are represented as mean ± SEM. *P<0.05, **P < 0.01, ***P < 0.001 and ns No significant difference (Student's t test). See also Figure S2 and S3.

Figure 4. cGAS is essential for DC sensing of irradiated-tumor cells

(A-C) BMDCs were cultured with 40Gy-pretreated MC38-SIY or non-irradiated-MC38-SIY cells. Subsequently purified CD11c+ cells were co-cultured with isolated CD8+ T cells from naive 2C mice for three days. (A) BMDCs from WT and Mb21d1−/− mice were used for co-culture with irradiated and nonirradiated MC38-SIY cells. DC cross-priming activity was analyzed by ELISPOT assays. (B) BMDCs were transfected with a siRNA-non-targeting control or siRNA-Mb21d1. Two days later after the transfection, the BMDCs were harvested for the co-culture with irradiated and non-irradiated MC38-SIY cells. DC cross-priming activity was analyzed by ELISPOT assays. (C) WT and Mb21d1−/− BMDCs were cultured with 40Gy-pretreated MC38-SIY cells. 10ng/ml IFN-β was added into the co-culture of Mb21d1−/− BMDC and irradiated-MC38-SIY cells. 100μg/ml DMXAA was added to isolated Mb21d1−/− CD11c+ cells for additional three hours incubation. DC cross-priming activity was analyzed by ELISPOT assays. (D) CD11c+ cells from WT or Mb21d1−/− BMDCs after co-culture with irradiated or non-irradiated MC38-SIY cells were incubated for 2 days, and then subjected to ELISA assays for IFNβ level. (E) CD11c+ cells were sorted from tumors in WT mice at 72 hour after radiation. Real-time PCR assay was performed to quantify the Mb21d1 mRNA. cGAS-deficient mice are represented by Mb21d1−/−. Representative data are shown from three (A-E) experiments. Data are represented as mean ± SEM. **P < 0.01 and ***P < 0.001 (Student's t test). See also Figure S4.

Figure 5. STING signaling is required for effective adaptive immune responses mediated by type I IFN signaling on DCs after radiation

MC38 tumors established in WT, Tmem173−/−, Cd11cCre+Ifnarf/f and Ifnarf/f mice were treated locally with one dose of 20Gy. (A) 300μg anti-CD8 mAb was administered i.p. in WT mice every three days for a total of four times starting from the day of radiation. Tumor growth was monitored. (B) Eight days after radiation, tumor draining inguinal lymph nodes (DLNs) were removed from WT and Tmem173−/− mice. Tumor antigen-specific CD8+ T cells function was measured by ELISPOT assays by co-culturing purified CD8+ cells with IFN-γ-treated MC38 tumor cells. (C) 1×1010 viral particles of Ad-null or Ad-IFN-β was administered intratumorally on day 2 after radiation. Tumor DLNs were removed on day 8 after radiation. Isolated CD8+ cells were subjected to ELISPOT assay with the presence of IFN-γ-treated MC38 tumor cells. (D) Tumor growth curves were analyzed in Cd11cCre+Ifnarf/f and Ifnarf/f mice after radiation. (E) Tumor DLNs from Cd11cCre+Ifnarf/f and Ifnarf/f mice were removed on day 8 after radiation. ELISPOT assay were performed with purified CD8+ cells and IFN-γ-treated MC38 tumor cells. STING-deficient mice are represented by Tmem173−/−. Representative data are shown from three (A-E) experiments conducted with 5-6 (A and D) or 3-4 (B-C and E) mice per group. Data are represented as mean ± SEM. **P<0.01 and ***P <0.001 (Student's t test). See also Figure S5.

Figure 6. cGAMP treatment promotes the antitumor effect of radiation in a STING-dependent manner

MC38 tumors were established in WT and Tmem173−/− mice and treated locally with one dose of 20Gy. (A-B) 10μg 2’3’-cGAMP was administered intratumorally on day 2 and 6 after radiation. Tumor volume (A) and tumor-bearing mice frequency after IR (B) in WT and Tmem173−/− mice were monitored. (C-D) 10μg 2’3’-cGAMP was administered intratumorally in WT and Tmem173−/− mice on day 2 after radiation. Tumor DLNs were removed on day 8 after radiation, and purified CD8+ T cells were subjected to ELISPOT assay with the presence of IFNγ-treated MC38 cells. (C) IFN-γ-producing CD8+ T cells were evaluated in WT mice that were treated with IR alone, cGAMP alone or the combination of IR and cGAMP. (D) IFN-γ-producing CD8+ T cells were evaluated in WT and Tmem173−/− mice that were treated with the combinaiton of IR and cGAMP. STING-deficient mice are represented by Tmem173−/−. Representative data are shown from three experiments conducted with 5-7 (A-B) or 3-4 (C-D) mice per group. Data are represented as mean ± SEM. **P<0.01 and ***P <0.001 (Student's t test in A, C and D, and log rank (Mantel-Cox) test in B).