ATM inhibition enhance immunotherapy by activating STING signaling and augmenting MHC Class I

Abstract

Accumulating evidence supports the concept that DNA damage response targeted therapies can improve antitumor immune response by increasing the immunogenicity of tumor cells and improving the tumor immune microenvironment. Ataxia telangiectasia mutated (ATM) is a core component of the DNA repair system. Although the ATM gene has a significant mutation rate in many human cancers, including colorectal, prostate, lung, and breast, it remains understudied compared with other DDR-involved molecules such as PARP and ATR. Here, we found that either gene knockout or drug intervention, ATM inhibition activated the cGAS/STING pathway and augmented MHC class I in CRC cells, and these effects could be amplified by radiation. Furthermore, we found that MHC class I upregulation induced by ATM inhibition is dependent on the activation of the NFκB/IRF1/NLRC5 pathway and independent of STING. Animal experiments have shown increasing infiltration and cytotoxic function of T cells and better survival in ATM-deficient tumors. This work indicated that ATM nonsense mutation predicted the clinical benefits of radiotherapy combined with immune checkpoint blockade for patients with CRC. It also provides a molecular mechanism rationale for ATM-targeted agents for patients with CRC.

Fig. 1. ATM inhibition activates cGAS/STING pathway.

A Gene Set Enrichment Analysis plot for CYTOSOLIC DNA SENSING PATHWAY. B, C Immunoblots showing STING, p-STING, TBK1, p-TBK1 and β-actin expression level in vector control and ATM-KO CT26 cells and HCT116 cells. D qRT-PCR was used to measure the mRNA level of CCL5, CXCL10 in vector control and ATM-KO CT26 cells and HCT116 cells. E, F Immunoblots showing STING, p-STING, TBK1, p-TBK1 and β-actin expression level in CT26 cells and HCT116 cells treated with increasing dose of KU60019 for 24 h. G qRT-PCR was used to measure the mRNA level of CCL5, CXCL10 in CT26 cells and HCT116 cells treated with increasing dose of KU60019 for 24 h. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.

Fig. 2. Radiation enhances cGAS/STING pathway activation induced by ATM inhibition.

A Immunoblots of γH2AX and cGAS in vector control and ATM-KO CT26 cells treated with increasing doses of radiation for 24 h. B, C Immunoblots of γH2AX in vector control and ATM-KO CT26 cells and its quantitative data analysis at different times after radiation. D–H Vector control and ATM-KO CT26 cells were irradiated with 6 Gy for 24 h. Fluorescence microscopy was used to assess the dsDNA and cGAS (D). Scale bar: 50 μm. The percentage of micronuclei dsDNA positive cells relative to the total number of cells in three randomly selected fields (E). Immunoblots showing STING, p-STING, TBK1, p-TBK1, ATM, and β-actin expression levels (F and G). qRT-PCR was used to measure the mRNA level of CCL5, CXCL10, and IFNβ (H). I–K CT26 cells were treated with KU60019 (3 μmol/L, 24 h) or radiation (6 Gy, 24 h). Immunoblots showing STING, p-STING, TBK1, p-TBK1, and β-actin expression levels (I and J). qRT-PCR was used to measure the mRNA level of CCL5, CXCL10, and IFNβ (K). L, M ELISA was used to detect the level of CXCL10 (L) and IFNβ (M) in shFluc and shATM HCT116 after different treatments for 24 h. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 3. ATM inhibition enhances MHC-I, IRF1 and NLRC5 expression.

A, B FPKM quantification of antigen processing and presentation-related genes (A), IRF1, and NLRC5 (B). C, D Immunoblots showing b2M and GAPDH expression levels in vector control, ATM-KO, and KU60019 (3 μM, 24 h) treated CT26 cells and HCT116 cells. E–J Vector control, ATM-KO, KU60019 (3 μM, 24 h) treated CT26 cells, and HCT116 cells were immunostained with the anti-MHC-I antibody and analyzed with flow cytometry (E–H). qRT-PCR was used to measure the mRNA level of MHC-I, IRF1, and NLRC5 (I and J). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 4. IRF1/NLRC5 is required for ATM inhibition-induced expression of MHC-I.

A–C Vector control and ATM-KO CT26 cells were transfected with siRNA-Ctrl or siRNA-IRF1 for 48 h. Surface expression of H2Kd/H2Dd was determined by flow cytometry (A and B). qRT-PCR was used to measure the mRNA level of H2Kd/H2Dd, IRF1 and NLRC5 (C). D–F Vector control and ATM-KO CT26 cells were transfected with siRNA-Ctrl or siRNA-NLRC5 for 48 h. Surface expression of H2Kd/H2Dd was determined by flow cytometry (D and E). qRT-PCR was used to measure the mRNA level of H2Kd/H2Dd, IRF1 and NLRC5 (F). G–J Vector control and ATM-KO CT26 cells were transfected with shRNA-Ctrl or shRNA-STING. Surface expression of H2Kd/H2Dd was determined by flow cytometry (G and H). STING expression was determined by Immunoblots (I). IFNβ and CXCL10 expression were determined by qRT-PCR (J). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 5. NF-κB transcriptional pathway is involved in the upregulation of MHC-I, IRF1, and NLRC5 caused by ATM inhibition.

A–C Vector control and ATM-KO CT26 cells were treated with MG132 (8 μM, 1 h). Surface expression of H2Kd/H2Dd was determined by flow cytometry (A and B). qRT-PCR was used to measure the mRNA level of H2Kd/H2Dd, IRF1 and NLRC5 (C). D Gene set enrichment analysis plot for NF-KAPPA B SIGNALING PATHWAY. E, F Representative immunoblot analysis of cytoplasmic and nuclear relocalization of NF-κB family members in Vector control and ATM-KO CT26 cells after different treatments. G–I Vector control and ATM-KO CT26 cells were transfected with siRNA-Ctrl, siRNA-RELA, or siRNA-RELB for 48 h. Surface expression of H2Kd/H2Dd was determined by flow cytometry (G). qRT-PCR was used to measure the mRNA level of H2Kd/H2Dd, IRF1 and NLRC5 (H and I). J Vector control and ATM-KO CT26 cells were irradiated with 6 Gy for 24 h. Fluorescence microscopy was used to assess P65 and P52. Scale bar: 20 μm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 6. Radiation and anti-PD-L1 enhance therapeutic efficacy in ATM-deficient tumor model.

A Schema of the mice experimental protocol. B, C The tumor volume curve of BALB/c mice inoculated with approximately 1 × 10⁶ vector control (B) or ATM-KO CT26 cells (C). n = 6. D, E Kaplan–Meier survival curves for vector control (D) or ATM-KO CT26 tumor (E). n = 6. F, G Representative IHC images of CD8 staining (F) and statistical analysis (G) of vector control and ATM-KO CT26 tumors. Scale bar: 50 μm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 7. Radiation and anti-PD-L1 enhances lymphocyte function in ATM-deficient tumor model.

A–D Representative flow cytometry profiles of KI67 in CD3+ T cells (A), KI67 in CD3+ CD8+ T cells (B) from TIL of vector control CT26 tumors and their quantitative analysis (C and D). n = 4. E Quantitative analysis of the MFI of CD86 in CD45+ CD11c+ MHC-II+ cells from tumor-draining lymph node of vector control CT26 tumors. n = 4. F–K Representative flow cytometry profiles of KI67 in CD3+ T cells (F), KI67 in CD3+ CD8+ T cells (J), and perforin in CD3+ CD8+ T cells (H) from TIL of ATM-KO CT26 tumors and their quantitative analysis (I, J and K). n = 4. L Quantitative analysis of the MFI of CD86 in CD45 + CD11c+ MHC-II+ cells from tumor-draining lymph node of ATM-KO CT26 tumors. n = 4. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 8. ATM inhibitor is synergistic with radiation and anti-PD-L1 in the syngeneic mouse CRC model.

A Schema of the mice experimental protocol. B Body weight curve of BALB/c mice inoculated with ~1 × 10⁶ CT26 cells and divided into different groups. n = 6. C, D The tumor volume curve (C) and Kaplan–Meier survival curves (D) of tumor-bearing BALB/c mice. n = 6. E, F Representative IHC images of CD8 staining (E) and statistical analysis (F) of CT26 tumors. Scale: 50 μm. G–L Representative flow cytometry profiles of KI67 in CD3+ T cells (G), KI67 in CD3+ CD8+ T cells (H) and perforin in CD3+ CD8+ T cells (I) from TIL of CT26 tumors and their quantitative analysis (J–L). n = 4. M Quantitative analysis of the MFI of CD86 in CD45 +CD11c+ MHC-II+ cells from tumor-draining lymph node of CT26 tumors. n = 4. *P < 0.05, ** P < 0.01, *** P < 0.001, ****P < 0.0001.

Fig. 9

Schematic illustration of ATM inhibition affects STING signaling and MHC Class I expression in tumor cells.