Dietary Lipids Induce Ferroptosis in Caenorhabditiselegans and Human Cancer Cells

Abstract

Dietary lipids impact development, homeostasis, and disease, but links between specific dietary fats and cell fates are poorly understood. Ferroptosis is an iron-dependent form of nonapoptotic cell death associated with oxidized polyunsaturated phospholipids. Here, we show that dietary ingestion of the polyunsaturated fatty acid (PUFA) dihomogamma-linolenic acid (DGLA; 20:3n-6) can trigger germ-cell ferroptosis and sterility in the nematode Caenorhabditis elegans. Exogenous DGLA is also sufficient to induce ferroptosis in human cells, pinpointing this omega-6 PUFA as a conserved metabolic instigator of this lethal process. In both C. elegans and human cancer cells, ether-lipid synthesis protects against ferroptosis. These results establish C. elegans as a powerful animal model to study the induction and modulation of ferroptosis by dietary fats and indicate that endogenous ether lipids act to prevent this nonapoptotic cell fate.

Keywords: ferroptosis; germ cells; polyunsaturated fatty acids.

Figure 1.. Ferroptosis regulators modulate DGLA-induced germ cell death.

(A) Percentage (%) sterility of wild-type and ced-3 C. elegans raised on 0.1 mM or 0.13 mM DGLA and ferrostatin-1 (Fer-1). (B) Representative DAPI-stained images of wild-type C. elegans after exposure to 0.1 mM DGLA. Fertile worms may have abnormal (diminished) gonads. (C) Quantification of fertility phenotypes based on the categories in (B). (D) % sterility of worms exposed to DGLA ± Trolox (vitamin E). (E) Quantification of fertility phenotypes based on the categories in (B). (F) % sterility of wild-type or gpx-1 mutant worms after dietary exposure to DGLA (G) % sterility of wild-type or NADPH oxidase/duox (bli-3) mutant worms after dietary exposure to DGLA. (H) % sterility of wild-type or ftn-1 mutant worms exposed to dietary DGLA ± 2,2’-bipyridine (BP). (I) % sterility of wild-type worms fed the indicated combinations of oleic acid (OA) and DGLA. (J) % sterility of wild-type and fat-2 mutants raised on DGLA plates. In (A, D, and F-J), each point represents an independent experiment comprising 50 worms for each treatment. Statistical significance was determined by two-way ANOVA with a Tukey’s test for multiple comparisons (Table S1). Fatty acid composition data is reported in Table S2.

Figure 2.. DGLA induces ferroptosis in human cells.

(A) Cell death (lethal fraction) in HT-1080ⁿ cells treated ± erastin2, dihomogamma-linolenic acid (DGLA) and ± ferrostatin-1 (Fer-1, 1 μM). (B) Representative images showing HT-1080ⁿ cells treated ± DGLA (500 μM) ± Fer-1 (1 μM) at 72 h. Live cells express nuclear-localized mKate2; dead cells take up SYTOX Green. Scale bar = 50 pm. (C) Cell death ± Fer-1 (1 μM). (D) Confocal images of C11 BODIPY 581/591 (herein “C11”) in HT-1080 cells. After treatment for 8 h, cells were labeled with C11 (5 μM) and Hoechst (1 μM /mL) prior to imaging. C11^Ox: Oxidized C11; C11^Non-Ox: Non-Oxidized C11. Scale bar = 20 μm. Images are representative of two independent experiments. (E,F) Cell death in response to treatment with polyunsaturated fatty acids (all at 500 μM), or a control non- ferroptotic lethal bortezomib (200 nM), ± Fer-1 (1 μM). AA: arachidonic acid, DHA: docosahexaenoic acid, EPA: eicosapentaenoic acid. (G) Cell death over time. (H) Heat map showing the relative fold changes in fatty acid composition of various lipid classes after 6 hours of growth with DGLA (500 μM or 250 μM) versus ethanol controls. PC: phosphatidylcholine, PE: phosphatidylethanolamine, neutral: neutral lipids. Grey boxes indicate situations with fatty acid concentrations at less than 0.4%. Each data point in A,C,E,F represents an independent experiment. Data in G represent mean ± SD of three independent experiments. Data in H represent mean values from three independent experiments.

Figure 3.. Ether lipids protect cells from ferroptosis.

(A) Percentage (%) sterility of wild-type and ads-1 on various doses of DGLA and ferrostatin-1 (Fer-1). (B) Iron chelation by 2,2-bipiridine suppresses ferroptosis in both wild type and in ads-1 mutants. (C) Dietary oleic acid partially suppresses the sterility induced by DGLA in ads-1 mutants. C) Epistasis analysis of ads-1 combined with bli-3 or fat-3. In figures 3A-3D, each point represents the % of worms that are sterile in an independent experiment comprising 50 worms for each treatment. Statistical significance was determined by two-way ANOVA with a Tukey’s test for multiple comparisons (Table S1). Fatty acid composition data is reported in Table S2. (E) HT-1080^N cells were pretreated with DMSO or ZINC-69435460 (AGPSi, 500 μM) for 24 h, then co-treated as indicated. Data are mean ± SD of three independent experiments. (F) The plasmalogen species, 16:0 alkenyl, is reduced in cells treated with AGPSi (500 μM). (G) Model for dietary DGLA-induced ferroptosis.