ABCC10-mediated cGAMP efflux drives cancer cell radiotherapy resistance

Abstract

Although radiotherapy (RT) is used in more than 50% of cancer patients, the intrinsic radioresistance of cancer cells, characterized by metabolic adaptation, significantly limits its clinical efficacy. However, the mechanisms underlying RT resistance (RTR) remain incompletely understood. In this study, we used high-throughput metabolic CRISPR library screening and identified ABCC10 as a novel molecular contributor to RTR. Functional assays, including vesicle transport, molecular docking, and an enzyme-linked immunosorbent assay, confirmed that the R545 site of ABCC10 binds to and effluxes 2'3'-cyclic GMP-AMP (cGAMP) in an ATP-dependent manner. Mechanistically, RNA transcriptomics, along with overexpression and silencing experiments, demonstrated that ABCC10-mediated export of cGAMP suppresses the STING-TBK1-IRF3 signaling pathway. This efflux reduces RT-induced intercellular accumulation of reactive oxygen species and DNA damage. In vivo, a combination of RT and nilotinib, a potential ABCC10 inhibitor, synergistically inhibited tumor growth. In summary, we identified ABCC10 as a novel exporter of cGAMP in RTR. Our results highlight its potential role as a biomarker for predicting RT response and as a therapeutic target for overcoming RTR.

Fig. 1. Human metabolic CRISPR-Cas9 screen to identify novel candidates for radiotherapy resistance.

a Schematic of a metabolic in vitro CRISPR screen for identifying regulators of RT. b Scatter plot showing the top 10 negative regulatory genes in two rounds of the CRISPR-Cas9 screen using MAGeCK analysis. c Indication of the combined analysis of the two rounds of the functional screen. d Gene ontology (GO) analysis of molecular function (MF) for the intersection of 99 negative regulatory genes. e Survival analysis for BRCA patients was conducted separately within the high and low expression groups, comparing patients treated with RT and those who did not receive RT. f Immunoblot analysis of the protein expression levels of ABCC10 in parallel Patu8988T cells and radioresistant Patu8988T cells.

Fig. 2. ABCC10 confers radioresistance to neoplastic cells.

a Immunoblot analysis of the protein expression levels of ABCC10 in Patu8988T and Calu-1 cells treated with radiotherapy (RT) (8 Gy) at the indicated time points. b Cell viability in Patu8988T and Calu-1 ABCC10-knockout cells treated with RT (8 Gy). Representative images (c) and quantification (d) of clonogenic survival analysis of Patu8988T and Calu-1 ABCC10-knockout cells treated with the indicated dose of ionizing radiation. e Cell viability in WT cells, ABCC10-knockout cells and ABCC10-knockout cells with re-expression of ABCC10 treated with RT (8 Gy). Representative images (f) and quantification (g) of clonogenic survival analysis of WT cells, ABCC10-knockout cells and ABCC10-knockout cells with re-expression of ABCC10 treated with the indicated dose of ionizing radiation. h Cell viability in Patu8988T and Calu-1 ABCC10-overexpressing cells treated with RT (8 Gy). Representative images (i) and quantification (j) of clonogenic survival analysis of Patu8988T and Calu-1 ABCC10-overexpressing cells treated with the indicated dose of ionizing radiation. k–n NXG mice were transplanted subcutaneously with doxycycline (DOX)-inducible ABCC10-knockdown Patu8988T cells and treated as indicated. A diagram of tumor growth delay experiments performed in vivo and the ionizing radiation fractionated treatment protocol is shown in (k); tumor volumes were calculated (l); tumor images were acquired as shown in (m); tumor weights (n) were measured. Experiments were repeated three times, and data are expressed as mean ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Fig. 3. ABCC10 alleviates post-radiotherapy ROS-dependent DNA damage.

a Immunoblot analysis of the protein expression levels of γ-H2A.X in Patu8988T and Calu-1 ABCC10-knockout cells at the indicated times after RT (8 Gy). b Immunoblot analysis of the protein expression levels of γ-H2A.X in Patu8988T and Calu-1 ABCC10-overexpressing cells at the indicated times after RT. c Protein levels of γ-H2A.X in tumor tissues were detected. d Quantification of γ-H2A.X protein levels in in tumor tissues. e Immunoblot analysis of the protein expression levels of ABCC10 and γ-H2A.X in WT cells, ABCC10-knockout cells and ABCC10-knockout cells with re-expression of ABCC10 treated with RT (8 Gy) after 6 h. f Quantification of flow cytometry-based analysis of ROS levels in Patu8988T WT and ABCC10 KO cells at 24 h after RT. g Quantification of flow cytometry-based analysis of ROS levels in Patu8988T vector and ABCC10-overexpressing cells at 24 h after RT. h Quantification of flow cytometry-based analysis of ROS levels in WT and ABCC10 KO1 Patu8988T cells pretreated with or without NAC (1 mM) at 24 h after RT. i Immunoblot analysis of the protein expression levels of γ-H2A.X in WT and ABCC10 KO1 Patu8988T cells pretreated with or without NAC (1 mM) at 6 h after RT. j Cell viability of Patu8988T WT and ABCC10-knockout cells pretreated with or without NAC (1 mM) at 72 h after RT. Experiments were repeated three times, and the data are expressed as mean ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Fig. 4. ABCC10 loss further activates the STING pathway after radiation.

a A volcanic map was used to analyze differentially expressed transcriptome genes of ABCC10 KO1 cells treated with or without radiotherapy (RT) for 6 h. b Heatmap of STING pathway-related genes in ABCC10 KO1 cells compared to WT cells. c Gene Ontology (GO) molecular function enrichment analysis of upregulated genes from ABCC10 KO1 Patu8988T cells compared to WT cells treated with RT. d Western blot analysis of ABCC10, pTBK1, TBK1, pIRF3, and IRF3 protein levels in ABCC10 WT and KO Patu8988T cells. e Western blot analysis of FLAG, pTBK1, TBK1, pIRF3, IRF3, pSTING, and STING protein levels in vector and ABCC10-overexpressing Patu8988T STING-overexpression (OE) cells. f Quantification of flow cytometry-based analysis of reactive oxygen species (ROS) levels in Patu8988T WT and ABCC10 KO cells pretreated with H151 24 h after RT. g Western blot analysis of γ-H2A.X protein levels in Patu8988T and BxPC3 WT and ABCC10 KO cells pretreated with H151 (1 μM) 6 h after RT. h Quantification of clonogenic survival analysis of Patu8988T WT and ABCC10 KO cells pretreated with H151 (1 μM) after RT. Experiments were repeated at least three times, and the data are expressed as mean ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Fig. 5. ABCC10-mediated cGAMP export limits activation of the STING pathway.

a DNA was detected using PicoGreen dye in WT and ABCC10 KO1 cells. The nucleus was stained with DAPI dye. Fluorescence intensity of PicoGreen was calculated using ImageJ from three different areas, measuring the signal from PicoGreen in non-nuclear regions. b Cytoplasmic DNA from vector and ABCC10-overexpressing Patu8988T cell lysates was separated on an agarose gel. c Vector and ABCC10-overexpressing Patu8988T cells were treated with radiation (8 Gy) and measured for cGAMP in cell lysates and supernatants using an ELISA kit 8 h later. d WT and ABCC10-knockout Patu8988T cells were treated with radiation (8 Gy) and then measured for cGAMP in cell lysates and supernatants using an ELISA kit 8 h later. e 20-min vesicle transport assays using 293 T cell-derived vesicles expressing human ABCC10 or control vesicles with cGAMP (5 μM) in the presence of ATP or AMP. f Western blot analysis of vector and ABCC10-overexpressing cells treated with 2′3′-cGAMP after the indicated time points. g Western blot analysis of FLAG, ABCC10, pTBK1, TBK1, pIRF3, IRF3, pSTING, and STING expression in vector and ABCC10-overexpressing Patu8988T STING-OE cells. h qPCR analysis of IFNB1 mRNA expression in vector and ABCC10-overexpressing Patu8988T cells transfected with ctDNA at the indicated time points. i Western blot analysis of ABCC10 WT and KO Patu8988T cells transfected with ctDNA at the indicated time points. j qPCR analysis of IFNB1 mRNA expression in WT and ABCC10 KO1 Patu8988T cells transfected with ctDNA at the indicated time points. Experiments were repeated three times, and the data are expressed as mean ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Fig. 6. ABCC10 binds to cGAMP through the R545 site.

a Docking scores × (-1) of ABCC10 substrates. b 2′3′-cGAMP docked with the ABCC10-binding pocket. c Patu8988T cells transfected with empty vector, wild-type (WT), R545A, or R899A mutant ABCC10-overexpression lentivirus were harvested after stimulation with ctDNA for 4 h. The 2′3′-cGAMP level in cell lysates and supernatants were measured using an ELISA kit. d 20-min vesicle transport assays using 293 T cell-derived vesicles expressing human WT ABCC10, R545A mutant, R899A mutant or control vesicles with cGAMP in the presence of ATP. e Western blot analysis of Patu8988T cells transduced with WT, R545A, R899A mutant ABCC10, or control (empty vector) stimulated with ctDNA for 4 h. f Cell viability of Patu8988T cells transfected with empty vector, WT, R545A, or R899A mutant ABCC10-overexpression lentivirus was assessed 72 h after RT. g Quantification of clonogenic survival analysis of vector, WT, R545A, or R899A mutant ABCC10-overexpressing Patu8988T cells treated with RT. Experiments were repeated three times, and the data are expressed as mean ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Fig. 7. The ABCC10 inhibitor nilotinib enhances the therapeutic efficacy of radiotherapy both in vitro and in vivo.

a Docking scores × (-1) of ABCC10 inhibitors. b Nilotinib docked with the ABCC10-binding pocket. c 20-min vesicle transport assays using 293 T cell-derived vesicles expressing human WT ABCC10 and R545A mutant with cGAMP and nilotinib in the presence of ATP. d ELISA was performed to detect intracellular and extracellular cGAMP content in Patu8988T cells treated with nilotinib stimulated with ctDNA for 4 h. e Cell viability was assessed using CCK-8 in Patu8988T and KPC mouse cells treated with nilotinib at 72 h after RT. f Western blot analysis of γ-H2A.X in Patu8988T and KPC mouse cells pretreated with nilotinib at 6 h after RT. g Western blot analysis of pTBK1, TBK1, pIRF3, and IRF3 protein levels in Patu8988T and KPC cells pretreated with nilotinib at 24 h after RT. h Schematic diagram of in vivo tumor growth and fractionated treatment protocol with radiation. Tumor growth curves (i) and tumor weights (j) for tumors generated from KPC mouse cells implanted subcutaneously in C57BL/6 mice with the indicated treatments. k Representative photographs of isolated tumor tissues following the indicated treatments. l Immunohistochemical analysis of pSTING, ISG15, and γ-H2A.X protein levels in harvested tumor tissues. Experiments were repeated three times, and the data are expressed as mean ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). Data were analyzed using Student’s t-test or two-way analysis of variance. m A schematic model showing the role of ABCC10 in radioresistance. ABCC10 blocks the activation of the STING pathway by causing cGAMP efflux after RT. Moreover, nilotinib can overcome radioresistance by inhibiting ABCC10 activity, thereby activating the STING pathway and inducing DNA damage.