Autophagy promotes ferroptosis by degradation of ferritin

ABSTRACT

Macroautophagy/autophagy is an evolutionarily conserved degradation pathway that maintains homeostasis. Ferroptosis, a novel form of regulated cell death, is characterized by a production of reactive oxygen species from accumulated iron and lipid peroxidation. However, the relationship between autophagy and ferroptosis at the genetic level remains unclear. Here, we demonstrated that autophagy contributes to ferroptosis by degradation of ferritin in fibroblasts and cancer cells. Knockout or knockdown of Atg5 (autophagy-related 5) and Atg7 limited erastin-induced ferroptosis with decreased intracellular ferrous iron levels, and lipid peroxidation. Remarkably, NCOA4 (nuclear receptor coactivator 4) was a selective cargo receptor for the selective autophagic turnover of ferritin (namely ferritinophagy) in ferroptosis. Consistently, genetic inhibition of NCOA4 inhibited ferritin degradation and suppressed ferroptosis. In contrast, overexpression of NCOA4 increased ferritin degradation and promoted ferroptosis. These findings provide novel insight into the interplay between autophagy and regulated cell death.

Autophagy is an evolutionarily conserved degradation pathway that maintains homeostasis.¹ In many cases, induction of autophagy promotes cell survival in response to environmental stressors such as nutritional starvation and energy depletion.² In some cases, excessive and impaired autophagy may contribute to cell death.³ Thus, defining the context-dependent function of autophagy in the regulation of cell death is important.

Ferroptosis, a novel form of regulated cell death, is characterized by a production of reactive oxygen species from accumulated iron and lipid peroxidation.⁴ According to the original study from Dr. Brent R. Stockwell in 2012, ferroptosis induced by erastin in cancer cells lacks the morphological and biochemical characteristics of apoptosis, necrosis, and autophagy.⁵ Potent autophagy inhibitors (e.g., bafilomycin A₁, 3-methyladenine, and chloroquine [CQ]) cannot prevent erastin-induced ferroptosis in certain cancer cells (e.g., HT-1080 and BJeLR).⁵ However, whether genetic inhibition of the autophagic pathway regulates the process and function of ferroptosis is unclear.

Autophagy-related (Atg) genes play a central role in the mediation of autophagy.⁶ Among them, Atg5 and Atg7 are critical for the formation of the autophagosome. We first investigated whether knockout of Atg5 or Atg7 in immortalized mouse embryonic fibroblasts (MEFs) regulates ferroptosis. Compared with wild-type (WT) MEFs, knockout of Atg5 or Atg7 inhibited erastin-induced cell death in atg5^−/−or atg7^−/−MEFs (Fig. 1A). Staurosporine (STS) is a classical inducer of apoptosis.⁷ Knockout of Atg5 or Atg7 enhanced STS-induced cell death (Fig. 1A). Moreover, knockdown of ATG5 and ATG7 by shRNA in human pancreatic cancer cell lines (PANC1 and PANC2.03) and the human fibrosarcoma cell line HT-1080 (Fig. 1B) also inhibited erastin-induced, but increased STS-induced, growth inhibition (Fig. 1C). As expected, both intracellular ferrous iron (Fe²⁺) levels and end products of lipid peroxidation (e.g., malondialdehyde [MDA]) (Fig. 1D) were significantly decreased in ATG5- or ATG7-deficient cells (MEFs and HT-1080) following treatment with erastin, suggesting that ATG5- and ATG7-mediated autophagy contributes to ferroptosis. In contrast, erastin-induced cell death, Fe²⁺, and MDA production did not significantly change after treatment with CQ (Fig. 1E), which is similar to the results of Dr. Stockwell's study.⁵ CQ and bafilomycin A₁ are not specific inhibitors to autophagic processes and can inhibit lysosomal processes, as well as endocytic pathways.⁸ Thus, the interpretation of data from studies using autophagy inhibitors should be combined with genetic approaches to more specifically inhibit the autophagy pathway.⁸

Figure 1.

Autophagy promotes erastin-induced ferroptosis. (A) The indicated immortalized MEFs were treated with erastin and STS for 24 h and cell viability was assayed using a Cell Counting Kit-8 (Sigma, 96992) (n = 3; *, p < 0.05 versus WT group). (B-C) Knockdown of ATG5 and ATG7 by specific shRNA (human ATG5_shRNA sequence: CCGGCCTGAACAGAATCATCCTTAACTCGAGTTAAGGATGATTCTGTTCAGGTTTTTTG; human ATG7_shRNA sequence: CCGGGCCTGCTGAGGAGCTCTCCATCTCGAGATGGAGAGCTCCTCAGCAGGCTTTTT) inhibited erastin-induced growth inhibition in the indicated human cancer cell lines (n = 3; *, p < 0.05 vs. control shRNA group). (D) Indicated MEFs or HT-1080 cells were treated with erastin for 24 h. Fe²⁺ and MDA levels were assayed using commercial kits (Abcam, ab83366 and ab118970) (n = 3; *, p < 0.05 versus WT group). (E) MEFs and PANC1 cells were treated with erastin “Era” in the presence or absence of CQ for 24 h. Fe²⁺ and MDA levels were assayed using commercial kits (Abcam, ab83366 and ab118970) (n = 3). (F) Western blot analysis monitoring protein expression in MEFs or PANC1 cells following treatment with erastin for 24 h. (G) The indicated MEFs were transfected with a GFP-LC3 plasmid for 24 h and then treated with erastin for 24 h. The GFP-LC3 puncta were assayed using image analysis. (H-I) The effects of knockdown of NCOA4 by shRNA (human NCOA4_shRNA(1) sequence: CCGGTCAGCAGCTCTACTCGTTATTCTCGAGAATAACGAGTAGAGCTGCTGATTTTTG; human NCOA4_shRNA(2) sequence: CCGGTGAACAGGTGGACCTTATTTACTCGAGTAAATAAGGTCCACCTGTTCATTTTTG), or overexpression of NCOA4 by transfection of human NCOA4 cDNA (OriGene Technologies, RC226676) on cell viability, Fe²⁺, MDA, and GSH levels in the indicated cells following treatment with erastin for 24 h (n = 3; *, p < 0.05 vs. control shRNA group or control cDNA group). (J) Schematic of the mechanism by which autophagy promotes ferroptosis.

Ferritin is the major intracellular iron storage protein complex, which includes FTL1 (ferritin light polypeptide 1) and FTH1 (ferritin heavy polypeptide 1).⁹ Increased ferritin expression limits ferroptosis.¹⁰ Recent studies indicate that increased autophagy can degrade ferritin to increase iron levels resulting in oxidative injury by the Fenton reaction.^11,12 Remarkably, the protein level of FTH1 was significantly increased in ATG5-deficient cells (MEFs and PANC1) with or without erastin treatment (Fig. 1F), suggesting that ATG5-mediated autophagy is required for ferritin degradation. MAP1LC3/LC3 (microtubule-associated protein 1 light chain 3), a mammalian homolog of yeast Atg8, has been used as a marker to monitor autophagy.¹³ ATG5-mediated lipidation of LC3 is required for the conversion of LC3-I to LC3-II.^14,15 Knockout/knockdown of Atg5/ATG5 inhibited erastin-induced LC3-II (Fig. 1F) and GFP-LC3 puncta formation (Fig. 1G). In addition to LC3-II, expression of the autophagy substrate SQSTM1 (sequestosome 1) was decreased in response to erastin in ATG5 WT, but not ATG5-deficient cells (Fig. 1F). These findings indicate that increased autophagy flux is associated with ferroptosis in fibroblasts and cancer cells following erastin treatment.

Given that NCOA4 is a selective cargo receptor for the autophagic turnover of ferritin,^11,12,16 we next determined whether a change in NCOA4 expression regulates erastin-induced ferroptosis. The protein level of NCOA4 did not remarkably change following erastin treatment (Fig. 1F). However, knockdown of NCOA4 by specific shRNA in PANC1 or HT-1080 cells increased FTH1 expression (Fig. 1H) and limited erastin-induced death (Fig. 1I). In contrast, overexpression of NCOA4 by gene transfection in PANC1 cells suppressed FTH1 expression (Fig. 1H) and increased erastin-induced death (Fig. 1I). As expected, the levels of Fe²⁺ and MDA were decreased in NCOA4-knockdown cells, whereas the levels of Fe²⁺ and MDA were increased in NCOA4-overexpressing cells (Fig. 1I). Moreover, the levels of glutathione (GSH) were increased in NCOA4-knockdown cells and decreased in NCOA4-overexpressing cells (Fig. 1I), confirming that iron and GSH are at the crossroad of redox metabolism in ferroptosis. Thus, NCOA4-mediated ferritin degradation is involved in ferroptosis.

In summary, the present study provides important evidence that activation of the autophagy pathway promotes ferroptosis by degradation of ferritin in fibroblasts and cancer cells (Fig. 1J). Consistently, a recent study also shows that induction of autophagy contributes to erastin-induced ferroptosis in cancer cells.¹⁷ Impaired ferroptosis has been identified in various human diseases such as neurodegenerative diseases,¹⁸ ischemia/reperfusion injury,^18,19 infectious diseases,²⁰ and cancers.^5,21-25 Further functional characterization of the ATG5-ATG7-NCOA4 autophagic pathway in ferroptosis may provide insight into the treatment of diseases of iron metabolism.

Abbreviations

Atg

autophagy related

CQ

chloroquine

Fe²⁺

ferrous iron

FTH1

ferritin, heavy polypeptide 1

GSH

glutathione

MAP1LC3/LC3

microtubule-associated protein 1 light chain 3

MDA

malondialdehyde

MEFs

mouse embryonic fibroblasts

NCOA4

nuclear receptor coactivator 4

STS

staurosporine

WT

wild type

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Christine Heiner (Department of Surgery, University of Pittsburgh) for her critical reading of the manuscript.

Funding

This work was supported by the National Institutes of Health of the USA (R01GM115366 and R01CA160417 to D.T.; R01 CA181450 to H.J.Z), the National Natural Science Foundation of China (31171229 and U1132005 to X.S.), the National Natural Science Foundation of Guangdong (2016A030308 to D.T.), a Science of Guangzhou Key Project (201508020258, 201400000003-4, and 201400000004-4 to X.S.), and a Research Scholar Grant from the American Cancer Society (RSG-16-014-01-CDD to D.T.). This project partly used University of Pittsburgh Cancer Institute shared resources supported by award P30CA047904.