Autophagy inhibition radiosensitizes in vitro, yet reduces radioresponses in vivo due to deficient immunogenic signalling

Abstract

Clinical oncology heavily relies on the use of radiotherapy, which often leads to merely transient responses that are followed by local or distant relapse. The molecular mechanisms explaining radioresistance are largely elusive. Here, we identified a dual role of autophagy in the response of cancer cells to ionizing radiation. On one hand, we observed that the depletion of essential autophagy-relevant gene products, such as ATG5 and Beclin 1, increased the sensitivity of human or mouse cancer cell lines to irradiation, both in vitro (where autophagy inhibition increased radiation-induced cell death and decreased clonogenic survival) and in vivo, after transplantation of the cell lines into immunodeficient mice (where autophagy inhibition potentiated the tumour growth-inhibitory effect of radiotherapy). On the other hand, when tumour proficient or deficient for autophagy were implanted in immunocompetent mice, it turned out that defective autophagy reduced the efficacy of radiotherapy. Indeed, radiotherapy elicited an anti-cancer immune response that was dependent on autophagy-induced ATP release from stressed or dying tumour cells and was characterized by dense lymphocyte infiltration of the tumour bed. Intratumoural injection of an ecto-ATPase inhibitor restored the immune infiltration of autophagy-deficient tumours post radiotherapy and improved the growth-inhibitory effect of ionizing irradiation. Altogether, our results reveal that beyond its cytoprotective function, autophagy confers immunogenic properties to tumours, hence amplifying the efficacy of radiotherapy in an immunocompetent context. This has far-reaching implications for the development of pharmacological radiosensitizers.

Figures

Figure 1

Knockdown of ATG5 or BECN 1 alters autophagy after irradiation and enhances radiation sensitivity of A549, H460 and CT26 cells. (a) Decreased ATG5 and BECN1 expression in A549 and H460 cells by RNA interference was observed by immunoblot. A representative immunoblot of three independent experiments is shown. (b–d) Cell lines as indicated. Radiation sensitivity of ATG5- and BECN1-depleted cells was assessed by clonogenic assay. Cells were transfected with a control shRNA or with an shRNA-targeting ATG5 or BECN1, then cells were left untreated or irradiated by 2 or 4 Gy. Cellular colonies were coloured and counted after 12–14 days and survival fraction (SF) was calculated and represented as described before. Quantitative data are reported as means±S.E.M. (n=3 triplicate samples, Student's t-test, ***P<0.001, as compared with untreated cells)

Figure 2

IR-induced cell death is increased in ATG5-depleted cells. (a) ATG5 depletion triggers mitochondrial membrane depolarization of A549 cells after ionizing radiation. A549 cells transfected by scrambled siRNA (SCR) or by specific shRNA for ATG5 (ATG5KD) were irradiated by 4 Gy and then subjected one day later to dual staining with the vital propidium iodide (PI) dye and with the mitochondrial transmembrane potential-sensitive dye 3,3 dihexyloxacarbocyanine iodide DiOC6(3) dye. Cell death was quantified by cytofluorometric analysis. (b) Histogram reporting the experiment shown in (a). Quantitative data are shown as means±S.E.M. (n=3 triplicate samples, Student's t-test, ***P<0.001 comparing ATG5KD (4 Gy) with ATG5KD (0 Gy) and SCR (4 Gy) with SCR (0 Gy); ###P<0.001 comparing ATG5KD (4 Gy) with SCR (4 Gy))

Figure 3

Autophagy delays radiation-induced growth in vivo. (a–c) Effects of ATG5 or BECN1 knockdowns on the growth of irradiated A549 or H460 tumours. Five nude BALB/c mice were subcutaneously injected with ATG5- or BECN1-depleted A549 cells (a and b) or H460 cells (c). Once xenografted tumours became palpable, mice were randomly assigned to a control group or to a group that received localized tumour irradiation with a single dose 8 Gy. Then, tumour volumes were monitored every 2–4 days. Quantitative data are reported as means±S.E.M. (Student's t-test, **P<0.01, ***P<0.001, as compared with the control group)

Figure 4

Autophagy inhibition decreases irradiation-induced ATP release and impairs anti-tumoural immune response. (a–c) Role of ATG5 on the growth of irradiated murine CT26 tumours implanted on immunodeficient mice. Nude BALB/c mice were subcutaneously injected with SCR or ATG5-depleted CT26 cells, then irradiated or left untreated with (c) or without an inhibitor of ecto-ATPase ARL67156 (ARL) (a). Tumour volumes were monitored every 3 or 4 days by direct measurement, as previously described. (b) Same experiment as in (a) in an immunocompetent host. Wild-type BALB/c mice were subcutaneously injected with ATG5-depleted CT26 cells, left untreated or irradiated, then tumour volume was monitored. Quantitative data presented are reported as means±S.E.M. (Student's t-test, *P<0.05, **P<0.01). (d and e) Tumour growth of irradiated tumours is inhibited by ATP release via autophagy. To determine the impact of irradiation and autophagy on ATP release, ATG5depleted CT26 (ATG5KD) or control (SCR) cells were irradiated with 4 Gy and incubated with or without an inhibitor of ecto-ATPase ARL67156 (ARL). ATP release was determined in vitro by using enzymatic methods at indicated times (e). Quantitative data are reported as means±S.E.M. (n=3 triplicate samples, Student's t-test, *P<0.05 comparing ATG5KD (4 Gy+ARL) with ATG5KD (Co) at 24 h; #P<0.05 comparing ATG5KD (4 Gy+ARL) with ATG5KD (4 Gy) at 24 h). To assess the impact of extracellular ATP release by irradiation on tumour growth, Balb/c mice were injected with scramble or Atg5-depleted CT26 tumour cells, then xenografted tumours were randomly assigned to treatment groups that received localized tumour irradiation with one single 8 Gy irradiation, with ARL (d). Tumour volumes were determined as indicated above and quantitative data are reported as means±S.E.M. (Student's t-test, ***P<0.001)

Figure 5

ARL67156 restores the lymphocyte infiltration in Atg5-depleted CT26 tumour after irradiation. (a) Representative pictures of CT26-recovered tumours. Scale bar, 10 μm. Arrows point lymphocytes. Quantitative data are reported in (b). Samples were compared using one-tailed Student's t-test. Error bars indicate S.E.M. **P<0.01, comparing SCR (IR) with SCR (Co) and SCR (IR+ARL) with SCR (Co); ##P<0.01, comparing SCR (IR+ARL) with SCR (IR); **P<0.01 comparing ATG5KD (IR) with ATG5KD (Co) and ATG5KD (IR+ARL) with ATG5KD (Co); ##P<0.01 comparing ATG5KD (IR+ARL) with ATG5KD (IR) in (b)