Radiotherapy and Immunotherapy Promote Tumoral Lipid Oxidation and Ferroptosis via Synergistic Repression of SLC7A11

Abstract

A challenge in oncology is to rationally and effectively integrate immunotherapy with traditional modalities, including radiotherapy. Here, we demonstrate that radiotherapy induces tumor-cell ferroptosis. Ferroptosis agonists augment and ferroptosis antagonists limit radiotherapy efficacy in tumor models. Immunotherapy sensitizes tumors to radiotherapy by promoting tumor-cell ferroptosis. Mechanistically, IFNγ derived from immunotherapy-activated CD8⁺ T cells and radiotherapy-activated ATM independently, yet synergistically, suppresses SLC7A11, a unit of the glutamate-cystine antiporter xc^-, resulting in reduced cystine uptake, enhanced tumor lipid oxidation and ferroptosis, and improved tumor control. Thus, ferroptosis is an unappreciated mechanism and focus for the development of effective combinatorial cancer therapy. SIGNIFICANCE: This article describes ferroptosis as a previously unappreciated mechanism of action for radiotherapy. Further, it shows that ferroptosis is a novel point of synergy between immunotherapy and radiotherapy. Finally, it nominates SLC7A11, a critical regulator of ferroptosis, as a mechanistic determinant of synergy between radiotherapy and immunotherapy.This article is highlighted in the In This Issue feature, p. 1631.

Figure 1.. Radiotherapy induces tumor cell ferroptosis

A, ID8 tumor growth following radiotherapy (8 Gy, single fraction, arrow) and/or liproxstatin-1 treatment (50 mg/kg, administered daily for 5 days, bar) in vivo. n=10 per group, ns P>0.05, ****P<0.0001, two-way ANOVA. B, B16F10 tumor lipid ROS following radiotherapy (8 Gy, single fraction) and/or liproxstatin-1 treatment (50 mg/kg, administered daily for 5 days) in vivo. DMSO, n=17; RT, n=14; Liproxstatin-1, n=10; RT+liproxstatin-1, n=12. ns P>0.05, ****P<0.001, two-way ANOVA. C, D, RSL-3 resistant B16F10 (C) tumor growth in vivo and (D) tumor lipid ROS following radiotherapy (8 Gy, single fraction, arrow). n=14 per group, (C), ****p<0.0001, two-way ANOVA; (D) ns P>0.05, **P<0.01, two-way ANOVA. E, Erastin resistant ID8 tumor growth following radiotherapy (8 Gy, single fraction, arrow) in vivo. n=10 per group. ns P>0.05, *P<0.05, two-way ANOVA. F, Clonogenic survival of HT1080 cells treated with DMSO, ferrostatin-1 (5 μM), liproxstatin-1 (5 μM), trolox (200 μM), DFO (2 μM) as indicated then irradiated (6 Gy) in vitro. Representative biological triplicate shown, mean±SD. ****p<0.0001, one-way ANOVA. G, HT1080 lipid ROS in vitro following radiotherapy (8 Gy) in vitro. Representative biological triplicates shown, mean±SD. ****P<0.0001, unpaired Student's t-test. H, Clonogenic survival of HT1080 cells following treatment with DMSO, RSL-3 (0.05 μM), erastin (2 μM), atorvastatin (5 μM), or sulfasalazine (5 μM) as indicated and radiotherapy (4 Gy) in vitro. Representative biological triplicate shown, mean±SD. *P<0.05, **P<0.01, ***P<0.001, one-way ANOVA. I, Clonogenic survival of HT1080 at indicated radiotherapy dose following cyst(e)inase (10 μM) treatment in vitro. Representative biological triplicates shown, mean±SD. **P<0.005, unpaired Student's t-test. J, B16F10 cell death following sulfasalazine (SAS, 5 μM) or cyst(e)inase (10 μM) and irradiation (20 Gy) in vitro. ****P<0.0001, two-way ANOVA. K, L, B16F10 tumor growth (K) and tumor lipid ROS (L) following radiotherapy (8 Gy, one fraction, arrow) and/or cyst(e)inase (87.5 mg/kg, arrowhead) in vivo. n=10 per group. ns P>0.05, ****P<0.0001, two-way ANOVA. M, ACSL4 knockout and parental B16F10 tumor growth following radiotherapy (10 Gy, single fraction, arrow) in vivo. n=10 per group. ****P<0.0001, two-way ANOVA. N, ACSL3 knockout and parental B16F10 tumor growth following radiotherapy (8 Gy, single fraction, arrow) in vivo. n=10 per group. ***P<0.001, two-way ANOVA. O, Relative lipid ROS in B16F10 cells treated with SAS (5 μM), RSL-3 (0.05 μM) or cyst(e)inase (10 μM) and radiotherapy (2 Gy) in vitro. Representative biological triplicate shown, mean±SD. ***P<0.001, two-way ANOVA. P, Tumor lipid ROS in B16F10 tumors following indicated radiotherapy dose fractionation in vivo. n=8-10 per group. **P<0.005, ***P<0.001, ****P<0.0001, one-way ANOVA. Data are representative of at least two independent experiments (A-P).

Figure. 2:. CD8⁺ T cells promote radiotherapy induced ferroptosis via IFN_γ

A, B16F10 tumor growth following irradiation (8 Gy, single fraction, arrow) and/or CD8⁺ T cells depletion (arrowhead) in vivo. n=20 per group, ****p<0.0001, two-way ANOVA. B, Clonogenic survival of B16F10 cells treated with of naïve or activated (OT-I) CD8⁺ T cell supernatant following RT (4 Gy) in vitro. Representative biological triplicate shown, mean±SD. *P<0.05, one-way ANOVA. C, Relative lipid ROS of B16F10 cells treated with naïve or activated (OT-I) CD8⁺ T cell supernatant following RT (4 Gy) with or without IFN_γ signaling blockade in vitro. Representative biological triplicate shown, mean±SD. ***P<0.001, ****P<0.0001, two-way ANOVA. D, Clonogenic survival of ID8 cells treated with IFN_γ (10 ng/ml) and/or RT (4 Gy) in vitro. Representative biological triplicate shown, mean±SD. ****p<0.0001, two-way ANOVA. E, ID8 cell relative lipid ROS following treatment of and/or RT (4 Gy) in vitro. ****p<0.0001, two-way ANOVA. F, Cell death in B16F10 following treatment with IFN_γ (10 ng/ml) and/or RT (20 Gy) in vitro. Representative biological triplicate shown, mean±SD. ****p<0.0001, two-way ANOVA. G, Clonogenic survival of HT1080 following treatment with IFN_γ (10 ng/ml), liproxstatin-1 (5 μM), and/or RT (4 Gy) in vitro. Representative biological triplicate shown, mean±SD. ****P<0.0001, two-way ANOVA. H, I, HT1080 tumor growth (H) and 4-HNE quantification (I) following treatment with RT (8 Gy, one fraction, arrow) and/or IFN_γ (65 μg/mouse, intraperitoneal, arrowhead) in vivo. n=10 per group. ***P<0.001, ***P<0.0001, two-way ANOVA. J, HT1080 tumor growth following treatment with liproxstatin-1 (intraperitoneal, 50 mg/kg, bar) and/or IFN_γ (65 μg/mouse, arrowhead) and radiotherapy (6 Gy, arrow) in vivo. DMSO, n=10; Liproxstatin-1, n=11; IFN_γ+RT, n=15; IFN_γ+RT+liproxstatin-1, n=14, ***P<0.001, ****P<0.0001, two-way ANOVA. Data are representative of at least two independent experiments (A-J).

Figure. 3:. Radiotherapy and IFN_γ synergistically suppress system SLC7A11

A, Normalized heatmap of HT1080 RNA-seq following treatment with IFN_γ (10 ng/ml), and/or RT (20 Gy) in vitro. B, qPCR for SLC7A11 in HT1080 cells treated with IFN_γ (10 ng/ml) and/or RT (10 Gy) in vitro. **P<0.01, ***P<0.001, ****P<0.0001, two-way ANOVA. C, Immunoblot for SLC7A11 in HT1080 cells treated with IFN_γ (10 ng/ml) and/or RT (10 Gy) in vitro. D, Representative SLC7A11 immunohistochemistry in HT1080 tumors treated as in Fig. 2H in vivo. E, ¹⁴C-cystine uptake in HT1080 cells treated with IFN_γ (10 ng/ml) and/or RT (10 Gy) in vitro. **P<0.01, ***P<0.001, ****P<0.0001, two-way ANOVA. F, Glutathione quantification in HT1080 cells treated with IFN_γ (10 ng/ml) and/or RT (16 Gy) in vitro. ****P<0.0001, two-way ANOVA. G, Immunoblot in HT1080 cells irradiated with RT (10 Gy) and/or KU60019 in vitro. H, HT1080 lipid ROS following RT (4 Gy) and/or KU60019 (1 μM) in vitro. Representative biological triplicate shown, mean±SD. ns P>0.05, *P<0.05, two-way ANOVA. I, Immunoblot for indicated in HT1080 cells treated with 10 Gy (RT) and/or siRNA targeting ATM (si-ATM) in vitro. J, Immunohistochemical quantification of SLC7A11 in HT1080 tumors of the indicated genotypes following treatment with RT (8 Gy, single fraction, arrow) in vivo. ns P>0.05, ****P<0.0001, two-way ANOVA. K, Immunoblot for indicated protein in HT1080 cells treated with IFN_γ (10 ng/ml) and/or siRNA targeting STAT1 (si-STAT1) in vitro. L, Immunohistochemical quantification of SLC7A11 in HT1080 tumors, treated with or without IFN_γ (65 μg/mouse, intraperitoneal). ns P>0.05, ****P<0.0001, two-way ANOVA. M, Clonogenic survival of HT1080 cells at indicated RT dose following shRNA targeting SLC7A11 (shSLC7A11) and liproxstatin-1 treatment in vitro. Representative biological triplicate shown, mean±SD. ***P<0.001, one-way ANOVA. N, O, SLC7A11 knockout B16F10 cells tumor growth (N) and tumor lipid ROS level (O) following irradiation (8 Gy, single fraction, arrow) in vivo. Parental, n=15; Parental+RT, n=19; SLC7A11 KO, n=17; SLC7A11 KO+RT, n=18. (N) ***P<0.001, two-way ANOVA; (O) **P<0.01, ****P<0.0001, two-way ANOVA. P, Clonogenic survival of wild type and SLC7A11 overexpressing HT1080 cells following 8 Gy and IFN_γ (10 ng/ml) in vitro. **P<0.001, unpaired Student's t-test. Q, SLC7A11 knockout with/without SLC7A11 overexpression B16F10 cells tumor growth following irradiation (8 Gy, single fraction, arrow) in vivo. n=10; *P<0.05, ****P<0.0001, two-way ANOVA Data are representative of at least two independent experiments (A-Q).

Figure. 4:. Radiotherapy and immunotherapy synergistically induce tumoral ferroptosis

A, B, B16F10 (A) tumor growth and (B) tumor lipid ROS following treatment with RT (8 Gy) and/or anti-CTLA-4 mAb in vivo. Isotype Control (IgG), n=16; RT, n=20; anti-CTLA-4 mAb (aCTLA-4), n=18; RT+a-CTLA-4 mAb (aCTLA-4+RT), n=20. (A) *P<0.05, ***P<0.001, ****P<0.0001, two-way ANOVA; (B) *P<0.05, ***P<0.001, two-way ANOVA. C, Intratumoral CD8⁺ T cell proliferation in vivo. *P<0.05, **P<0.01, two-way ANOVA. D, IFN_γ production by intratumoral CD8⁺ T cells in vivo. *P<0.05, two-way ANOVA. E, Granzyme B production by intratumoral CD8⁺ T cells in vivo. *P<0.05, ***P<0.001, two-way ANOVA. F, B16F10 tumor growth following treatment with liproxstatin-1 (50 mg/kg, intraperitoneal, bar) and/or RT (8 Gy, single dose, arrow) with anti-CTLA-4 mAb (arrowhead) in vivo. Isotype Control (IgG), n=16; RT, n=20; Liproxstatin-1, n=8; RT plus anti-CTLA-4 mAb and liproxstatin-1 (aCTLA-4+RT+liproxstatin-1), n=20. ****P<0.0001, two-way ANOVA. G, H, ID8 (G) tumor growth and (H) lipid ROS following treatment with RT (8 Gy, single dose, arrow) and/or anti-CTLA-4 mAb (arrowhead) in vivo. n=8 per group. *P<0.01, ***P<0.001, two-way ANOVA. I, J, B16F10 tumor cell growth (I) and tumor lipid ROS (J) following treatment with RT (8 Gy, single dose, arrow) and/or anti-PD-L1 mAb (200 μg/mouse, arrowhead) in vivo. Isotype Control (IgG), n=15; RT, n=16, anti-PD-L1 mAb (aPD-L1); n=17, RT and anti-PD-L1 mAb (aPD-L1+RT), n=17. (I) ****P<0.0001, two-way ANOVA; (J) *P<0.05, **P<0.01, two-way ANOVA. K, L, RSL-3 resistant B16F10 tumor growth in vivo (K) and tumoral lipid ROS (L) were treated with RT (8 Gy, one fraction, arrow) and anti-PD-L1 mAb (200 μg/mouse, arrowhead). Parental IgG, n=14; Parental RT+aPD-L1, n=16; RSL-resis IgG, n=14; RSL-resis RT+aPD-L1, n=16. (K) ns P>0.05, ****P< 0001, two-way ANOVA; (L) ns P>0.05, ****P<0.0001, two-way ANOVA. M, RSL-3 resistant or parental ovalbumin expressing B16F10 tumor growth following adoptive transfer of activated OT-I T cells in vivo (arrowhead). n=10. ****p<0.0001, two-way ANOVA. N, O, SLC7A11 knockout B16F10 cells tumor growth (N) and rechallenge inoculation (O) following irradiation (8 Gy, single fraction, arrow) and anti-PD-L1 treatment (200 μg/mouse, arrowhead) in vivo. Parental, n=10; Parental+aPD-L1+RT, n=10; SLC7A11 KO, n=10; SLC7A11 KO+aPD-L1+RT, n=25; Naïve, n=10. ****p< 0.0001, two-way ANOVA. Data are representative of at least two independent experiments (A-O).