An Artemisinin Derivative ART1 Induces Ferroptosis by Targeting the HSD17B4 Protein Essential for Lipid Metabolism and Directly Inducing Lipid Peroxidation

Open AccessCCS ChemistryRESEARCH ARTICLE1 Jan 2022An Artemisinin Derivative ART1 Induces Ferroptosis by Targeting the HSD17B4 Protein Essential for Lipid Metabolism and Directly Inducing Lipid Peroxidation Jingjing Xie†, Guangya Zhu†, Ming Gao†, Jie Xi, Ge Chen, Xinxin Ma, Yu Yan, Zhiyuan Wang, Ze-Jun Xu, Hui-Jun Chen, Hong-Dong Hao, Yaoyang Zhang, Zhu-Jun Yao and Jidong Zhu Jingjing Xie† Interdisciplinary Research Center on Biology and Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 201203 University of the Chinese Academy of Sciences, Beijing 100049 †Co-first authors.Google Scholar More articles by this author , Guangya Zhu† Interdisciplinary Research Center on Biology and Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 201203 University of the Chinese Academy of Sciences, Beijing 100049 †Co-first authors.Google Scholar More articles by this author , Ming Gao† State Key Laboratory of Coordination Chemistry, Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 †Co-first authors.Google Scholar More articles by this author , Jie Xi State Key Laboratory of Coordination Chemistry, Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 Google Scholar More articles by this author , Ge Chen School of Life Sciences, Westlake University, Hangzhou 310012 Google Scholar More articles by this author , Xinxin Ma Interdisciplinary Research Center on Biology and Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 201203 Google Scholar More articles by this author , Yu Yan Department of Translational Research in Psychiatry, Max Planck Institute of Psychiatry, Munich 80804 Google Scholar More articles by this author , Zhiyuan Wang Interdisciplinary Research Center on Biology and Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 201203 University of the Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author , Ze-Jun Xu State Key Laboratory of Bioorganic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032 Google Scholar More articles by this author , Hui-Jun Chen State Key Laboratory of Bioorganic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032 Google Scholar More articles by this author , Hong-Dong Hao State Key Laboratory of Bioorganic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032 Google Scholar More articles by this author , Yaoyang Zhang Interdisciplinary Research Center on Biology and Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 201203 Google Scholar More articles by this author , Zhu-Jun Yao *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Coordination Chemistry, Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, Jiangsu 210023 Google Scholar More articles by this author and Jidong Zhu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Interdisciplinary Research Center on Biology and Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 201203 Center for Excellence in Molecular Synthesis, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202000691 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Artemisinin and its derivatives, commonly known as antimalarial drugs, have gradually come to be regarded as potential antitumor agents, although their cytotoxic efficacy and mechanisms of action remain to be settled. Herein, we report that an artemisinin analog, ART1, can potently induce ferroptosis in a subset of cancer cell lines. Structure–activity relationship (SAR) analysis reveals that both the endoperoxide moiety and the artemisinin skeleton are required for the antitumor activity of ART1. Aided with ART1-based small-molecule tools, chemical proteomic analysis identified the HSD17B4 protein as a direct target of ART1. HSD17B4 resides in peroxisomes and is an essential enzyme in the catabolism of very-long-chain fatty acids. Our results demonstrate that ART1 initiates ferroptosis through selective oxidation of the fatty acids in peroxisomes by hijacking the HSD17B4 protein without disturbing its enzymatic function, providing a promising mechanism to develop therapeutics for cancer treatment. Download figure Download PowerPoint Introduction Cell death is a critical biological process in mammalian development and homeostasis. Some kinds of cell death caused by accidental insults cannot be reversed, but regulated cell death mediated by specific intracellular mechanisms can be modulated with pharmacological and genetic tools. Apoptosis was the first identified form of regulated cell death, but recently other nonapoptotic programmed cell deaths, including necroptosis, pyroptosis, and ferroptosis, have been discovered.1–3 Ferroptosis is an iron-dependent mode of cell death in which lipid peroxides accumulate to lethal levels.4–6 Both depletion of reduced glutathione (GSH) levels and inactivation of the lipid-peroxide-reducing enzyme (GPX4) can lead to ferroptotic cell death.7–10 Meanwhile, oxidative stress and excess iron are associated with cancer progression.11 Studies have shown that metastasis-prone mesenchymal cancer cells, usually resistant to targeted and chemotherapies are highly sensitive to ferroptosis.12,13 Consequently, activation of ferroptosis may be a legitimate approach to cancer treatment.8,10,14–16 Artemisinin [Qinghaosu (QHS)], a sesquiterpene lactone containing an unusual peroxide bridge, was isolated from the plant Artemisia annua, and, together with some of its derivatives, has been recognized as a potent drug for the treatment of malaria.17,18 It is widely accepted that the iron-mediated cleavage of artemisinin’s endoperoxide bridge and further release of highly reactive radicals are responsible for the rapid death of the malaria parasite.19,20 Recent studies have shown that artemisinins possess antitumor activity in various cancer cell lines derived from leukemia, liver cancer, brain glioma, breast cancer, and colon cancer.21–25 Although their cytotoxicity against cancer cells is often weak, artemisinins exhibit minimal cytotoxic effects on normal cells and are able to overcome drug resistance.26,27 The reported cytotoxic effects exerted by artemisinins are complex, including inhibition of cell proliferation, induction of cell apoptosis, arrest of the cell cycle, reduction of angiogenesis, inhibition of cell invasion and metastasis, induction of oxidative damage reactions, and regulation of tumor microenvironment.23–25,27–30 Previous studies also attempted to identify the protein target(s) of artemisinin and decipher the molecular mechanisms by which artemisinins kill cancer cells, but to date, the exact antitumor-related targets of artemisinins have not been clearly identified.11,29,31–34 It suggests that artemisinins may target a broad spectrum of protein substrates in mammalian cells, and the specificity of artemisinins treating cancers with mechanism-based patient stratification has become an issue. HSD17B4 protein, also called multifunctional enzyme type 2 (MFE-2) and D-bifunctional protein (DBP), is found in peroxisomes where it plays a central role in peroxisomal β-oxidation.35 It interacts with most, if not all, peroxisomal β-oxidation substrates and catalyzes the second and third steps of peroxisomal β-oxidation of fatty acids and fatty acid derivatives.36,37 It is a 79 kDa homodimeric enzyme, and each monomer has three functional units: a (3R)-hydroxyacyl-CoA dehydrogenase unit, a 2-enoyl-CoA hydratase 2 unit, and a sterol carrier protein 2-like (SCP-2L) unit.38–40 HSD17B4 protein is responsible for the breakdown of very-long-chain fatty acids (VLCFA) such as C26∶0, long-chain polyunsaturated fatty acids (PUFA), including C24∶6, branched-chain fatty acids, and bile acid intermediates.35 Although the function of HSD17B4 in the metabolism of fatty acid has been well studied, its role in ferroptosis and its linkage with the anticancer action of artemisinin and its derivatives have not been explored. In this study, a preliminary screening of our laboratory-made small library of artemisinin derivatives and analogs showed that ART1, an artemisinin derivative bearing a naphthalene unit [(3R,5aS,6R,9R,10R,12R,12aR)-3,6,9-trimethyl-10-(naphthalen-2-yloxy)decahydro-12H-3,12-epoxy[1,2]-dioxepino-[4,3-i]isochromene, Figure 1a] exhibits potent and selective anti-proliferative activity against cancer cell lines over normal cells. Mechanistic studies indicate that ART1 inhibits cancer cell viability through induction of noncanonical ferroptosis without affecting its intracellular GSH level and GPX4 activity. Further chemical proteomic studies revealed that the HSD17B4 protein, which is involved in the β-oxidation of VLCFA, is the direct target of ART1. Our finding that artemisinin analog ART1 targets HSD17B4 protein in peroxisome may represent a novel selective antitumor mechanism through induction of ferroptotic cell death. Figure 1 | ART1 inhibits cancer growth. (a) Structures of ART1, ART2, ART3, and artemisinins (QHS and DHA). (b) Effects of ART1, ART2, ART3, and artemisinins on viability of H1299 cells. Data are represented as mean ± SEM (n = 3). (c) Effects of ART1, ART2, ART3, and artemisinins on viability of A549 cells. Data are represented as mean ± SEM (n = 3). (d) IC50 values of ART1 and DHA against a panel of PDTOs derived from NSCLCs patients. Data are represented as mean ± SEM (n = 2). (e) Representative images showing the anticancer efficacy of ART1 and DHA against PDTOs. Scale bar = 100 μm. (f) Effects of ART1, ART2, and ART3 on the viability of leukemia MV4;11 cells. Data are represented as mean ± SEM (n = 3). (g) Effects of ART1 on the viability of several normal cells. Data are represented as mean ± SEM (n = 3). Download figure Download PowerPoint Experimental Methods Cell lines Cells were maintained according to the guidance from American Type Culture Collection. Human HEK293T (female) and HEK293FT (female) cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS), 100 units/mL penicillin and 100 mg/mL streptomycin. Human NCI-H1838 (female), NCI-H1975 (female), NCI-H1650 (male), NCI-H1299 (male), A549 (male), and MV4;11 (male) cells were cultured in RPMI1640 supplemented with 10% (v/v) FBS, 100 units/mL penicillin and 100 mg/mL streptomycin. All cells were cultured at 37 °C in a humidified atmosphere of 95% air and 5% CO2. Cells obtained from the cell bank of Shanghai Institute of Biochemistry and Cell Biology (Shanghai, China) and Cobioer Company (Nanjing, Jiangsu Province, China) were authenticated by short tandem repeat (STR) analysis and mycoplasma detection. Other cells were verified through periodic morphology checks and mycoplasma detection. Patient-derived tumor organoid Surgery tissues were obtained from patients with advanced non-small cell lung carcinoma (NSCLC) after ethical approval by the Ethics Committee of Hebei People’s Hospital (NCT03453307; Shijiazhuang, Hebei, China) and informed consent from all participants or next of kin. The tissues were washed with cold phosphate-buffered saline (PBS) containing antibiotics and chopped into approximately 5 mm pieces with surgical scissors. Tissues were further washed with 10 mL advanced DMEM/F12 (Thermo Fisher Scientific, Waltham, MA) containing 1× Glutamax, 10 mM N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES), and antibiotics and digested in 10 mL advanced DMEM/F12 containing 2% fetal calf serum (FCS) and 2 mg/mL collagenase (Sigma, C9407) on an orbital shaker at 37 °C for 1–2 h. The pellet was resuspended in 10 mL advanced DMEM/F12 containing 2% FCS and centrifuged again at 400g. Dissociated cells were collected in advanced DMEM/F12 (Thermo Fisher Scientific), suspended in growth factor reduced (GFR) matrigel (Corning Inc., Corning, NY), and seeded. The matrigel was then solidified and overlaid with 500 μL of complete human organoid medium, which was subsequently refreshed every 2 days. Patient-derived tumor organoids (PDTOs) were cultured in advanced DMEM/F12, supplemented with 1× B27 additive and 1× N2 additive (Thermo Fisher Scientific), 0.01% bovine serum albumin (BSA), 2 mM l-glutamine, 100 units/mL penicillin-streptomycin, and containing the following additives: epidermal growth factor (EGF), noggin, R-spondin 1, gastrin, FGF-10 (FGF = fibroblast growth factor), FGF-basic, Wnt-3A, prostaglandin E2, Y-27632, nicotinamide, A83-01, A83-01, SB202190, and hepatocyte growth factor (HGF) (Pepro-Tech, London, UK). Passaging of patient derived organoids (PDOs) was performed using TrypLE (Thermo Fisher Scientific). PDOs were biobanked in FBS (Thermo Fisher Scientific), containing 10% dimethyl sulfoxide (DMSO) (Sigma-Aldrich, St. Louis, MI). Experiments were performed with two biological replicates. ART16 pull down from cell lysates H1299 and MV4;11 cells incubated with 10 μM ART16 or biotin for 2 h were harvested and lysed in lysis buffer (50 mM Tris–HCl 7.5, 150 mM NaCl, 0.4% NP40, 1.5 mM MgCl2, 1 mM dithiothreitol (DTT), 5% glycerol, 1% proteinase cocktail). Then transferred supernatants were incubated with 30 μL streptavidin-agarose (20359; Thermo Fisher Scientific) for 40 min at room temperature. The beads were washed four times with lysis buffer. The precipitates were analyzed by silver staining or western blot with anti-HSD17B4 antibody. Experiments were performed three independent times. ART16 pull down with purified protein Purified HSD17B4 protein (final, 1 μM) was pretreated with increasing concentrations of ART1 for 1 h followed by treatment with 5 μM ART16 for 2 h and was further incubated with Dynabeads MyOne (65002; Thermo Fisher Scientific) for 1.5 h at 4 °C. The beads were washed with PBST (phosphate buffer saline, 0.05% Tween-20). The precipitates were analyzed by western blotting with anti-HSD17B4 antibody. Experiments were performed three independent times. Assessment of lipid peroxidation with c11-BODIPY H1299 cells were seeded in a six-well plate prior to the experiment so that the cell density reached 70–80% confluency at the day of ART1 treatment. Cells were treated with DMSO or 5 μM ART1 with or without deferoxamine (DFO; 100 μM) and incubated for 6 h at 37 °C. Then cells were harvested by trypsinization, washed, and resuspended in PBS containing 1 μM c11-BODIPY (4,4-difluoro-3a,4a-diaza-s-indacene; Thermo Fisher Scientific) and incubated for 30 min at 37 °C. Fluorescence intensity was measured on the FL1 channel with gating to rule out dead cells; 10,000 cells were analyzed per condition. Experiments were performed three independent times. Intracellular GSH/GSSG measurement H1299 cells were treated with vehicle (DMSO) or ART1 (10 μM) for 7 and 24 h: one set of cells was used for total GSH measurement, and the other set of cells was used for oxidized glutathione (GSSG) measurement. Treatments were then removed and replaced with corresponding lysis reagent provided in the GSH/GSSG-Glo Assay (Promega V6611). The intracellular GSH/GSSG levels were examined per the kit’s assay protocol. Experiments were performed with three biological replicates. GPX4 activity assay using tert-butyl hydroperoxide H1299 cells with 80% confluency in a 10 cm dish were treated with vehicle (DMSO), ART1(10 μM) for 12 h, or RSL3 (1 μM) for 2 h. Cells were harvested by trypsinization, washed twice in PBS, and resuspended in lysis buffer (25 mM Tris–HCl, pH 8.0, 150 mM NaCl, 0.3% Triton X-100, 0.1 mM DFO). Cells were further sonicated on ice for 20 min and centrifuged at 14,000g for 15 min. The Bradford method was used to measure the protein concentration in cell lysates. Enzymatic reactions were performed in 96-well microplate (100 μL volume). In lysis buffer containing reduced nicotinamide adenine dinucleotide phosphate (NADPH) (final 0.25 mM), GSH reductase (final 0.5 U/mL, Sigma G3664), 1.40 μg/μL cell lysates were added and mixed. The enzymatic reaction was initiated by adding 1 μL of 30 mM tert-butyl hydroperoxide (t-BuOOH) (Sigma 458139) to the reaction mixture. The depletion of NADPH was examined kinetically by reading OD340 at 10 s intervals over 30 min. Experiments were performed with two biological replicates. Biolayer interferometry The binding between ART16 and HSD17B4 protein was examined using biolayer interferometry (BLI) Octet RED96 (ForteB́io Inc., Menlo Park, CA). Streptavidin biosensors (ForteB́io Inc.) were presoaked for 30 min in assay buffer (25 mM Tris–HCl pH 8.0, 150 mM NaCl, 2 mM DTT, 1% BSA, and 0.02% Tween-20), then coated in an assay buffer containing 5 μM ART-16 for 5 min. Sensors were quenched in assay buffer containing 50 μM biocytin for 1 min, followed by a 2 min wash with assay buffer. Indicated concentrations of HSD17B4 protein (4.50, 1.93, 0.82, 0.35, 0.15, and 0.06 μM) were diluted in assay buffer and flowed through biosensors coated with ART16 or biocytin for 5 min. A 10 min dissociation followed. Data were analyzed using OctetRED analysis software. Experiments were performed two independent times. HSD17B4 protein enzymatic assay The enzymatic activity of HSD17B4 protein was examined accordingly to the literature procedure.36d-3-Hydroxyacyl CoA dehydrogenase assay consists of the following: the 5× assay buffer contains 125 mM Tris–HCl, pH 8.0; 5 mM NAD+; 250 mM KCl; 0.05% Triton X-100, and 0.25% BSA. Purified HSD17B4 protein (1.25 ng/μL) was diluted in assay buffer, coincubated with indicated concentrations of ART1 for 30 min, then 125 μM substrate dl-β-hydroxybutyryl-CoA lithium salt (Sigma H0261) was added, and immediately, the excitation and emission at 340 and 460, respectively, for NADH was measured with a kinetic process to quantify the dehydrogenase reaction. The hydratase assay consists of the following: the 2× assay buffer contains 0.64 M Tris–HCl, pH 7.4; 11.8 mM ethylenediaminetetraacetic acid (EDTA), and 0.012% BSA. Purified HSD17B4 protein (2.50 ng/μL) was diluted in assay buffer, coincubated with indicated concentrations of ART1 for 30 min, then 1 mM substrate Crotonyl CoA (Sigma 28007) was added, and immediately OD280 was measured for the remaining substrate with a kinetic process to quantify the dehydrogenase reaction. Experiments were performed with two biological replicates. Measurement of lipid peroxidation in vitro Lipid peroxides were examined per the modified absorbance of products from the reaction with thiobarbituric acid (TBARS).14,41 Thiobarbituric acid (TBA) solution 0.67% (w/v) was prepared in 10% trichloroacetic acid (TCA) with heating to 65 °C to dissolve the TBA solid. Arachidonic acid (AA; 500 μM) was mixed with 25 μM ART1 without 5 μM FeSO4 at 37 °C for 30 min in buffer containing H2O with 0.02% Triton-X100. Equal volumes of 0.67% (w/v) TBA in 10% TCA were added to the reaction mixture and vortexed and incubated at 95 °C for 15 min. The MDA-TBA pink adduct formed by the reaction of malondialdehyde (MDA) and TBA was measured colorimetrically at 532 nm. Experiments were performed with three biological replicates. Chemical synthesis Unless otherwise noted, all reactions were carried out under a nitrogen atmosphere with dry solvents under anhydrous conditions. Tetrahydrofuran (THF) was distilled immediately from sodium benzophenone ketyl prior to use. Methylene chloride (CH2Cl2) was distilled immediately before use from calcium hydride. External bath temperatures were used to record all reaction temperatures. All other solvents were processed per the reference Purification of Laboratory Chemicals (Seventh Edition) (detailed information could be found in Supporting Information Pages S5–S14). Silica gel (300–400 mesh) and petroleum ether, ethyl acetate and acetone were used for product purification by flash column chromatography. Analytical thin-layer chromatography (TLC) was performed with glass TLC plates. Visualization was accomplished with UV light, phosphomolybdic acid staining, and subsequent heating. 1H and 13C NMR spectra were recorded on either 400 or 500 MHz instruments (Bruker AM-400 spectrometer). IR spectra were recorded on a Fourier Transform infrared spectrometer and listed in cm−1. High-resolution mass spectrometry (HRMS) analyses were determined on a Agilent 6540 quadrupole time-of-flight mass spectrometer(Q-TOF MS). Optical rotations were measured with a polarimeter with a sodium lamp. Detailed synthesis processes are listed in the Supporting Information. Results and Discussion Cytotoxicity screen identifies ART1 as a potential antitumor agent To identify the lead compound derived from artemisinin (QHS) for cancer treatment, a series of endoperoxide derivatives that maintain the artemisinin scaffold but with different aryl substituents in the C-10 position (ART1, ART2, and ART3) were prepared and screened (Figure 1a, see Supporting Information for detailed information). We compared the antitumor activities of these compounds together with artemisinin (QHS) and its derivative dihydroartemisinin (DHA) in two lung cancer cell lines: H1299 and A549. We found that consistent with previous studies,11,26,42 QHS and DHA exhibit weak cytotoxic activity in the lung cancer cells with IC50 values ranging from 1 μM to more than 100 μM (Figures 1b and 1c). Among these derivatives, ART1, in which a naphthalene ring attaches to the artemisinin skeleton, exhibited the most potent cytotoxicity against the lung cancer cells (Figures 1b and 1c and Table 1). To determine whether ART1 could inhibit primary cancer cells, we further examined its anti-proliferative efficacy in PDTOs models. Experiments showed that ART1 can suppress the growth of PDTOs derived from NSCLC patients <10-fold more effectively than DHA (Figures 1d and 1e). We further compared the cytotoxicity of ART1, ART2, and ART3 in leukemia MV4;11 cell lines. Again, ART1 was proven to be the most potent compound to inhibit the proliferation of leukemia cells (Figure 1f and Table 1). Modification of the artemisinin skeleton resulted in varied antitumor activity suggesting that, apart from the endoperoxide bridge motif, the artemisinin scaffold also plays a critical role mediating the cancer-cell killing. To prove this, we further screened a library of organic peroxides to compare their antitumor activities ( Supporting Information Figure S1).43,44 Interestingly, although all the compounds contained the peroxide bridge motif, they barely affected the cell proliferation of MV4;11 cells, while ART1 strongly inhibited the cell proliferation. This strongly suggested that the artemisinin skeleton is indispensable for the cytotoxicity of ART1. More importantly, ART1 displays great selectivity, exhibiting very weak anti-proliferative activity against normal cells (Figure 1g). All the above results indicate that ART1 is a promising potential lead of therapeutics for cancer treatment. Table 1 | IC50 Values of ART1, ART2 and ART3 against H1299, A549, and MV4;11 Cancer Cellsa Cancer Cells ART1 (IC50, μM) ART2 (IC50, μM) ART3 (IC50, μM) H1299 0.09 ± 0.01 0.30 ± 0.14 1.19 ± 0.13 A549 0.44 ± 0.03 18.17 ± 6.30 29.10 ± 5.02 MV4;11 0.11 ± 0.01 1.11 ± 0.09 0.45 ± 0.20 aData are represented as mean ± SD (n = 3). ART1 induces noncanonical ferroptotic cell death Previous reports demonstrate that artemisinins cause death of cancer cells through multiple mechanisms, including apoptosis, autophagy, cell cycle arrest, and inhibition of tumor angiogenesis.23–25,27–30 We first examined whether apoptosis is induced by ART1 and found that ART1 fails to activate caspases and cause PARP cleavage ( Supporting Information Figures S2a and S2b), suggesting that cell death triggered by ART1 is independent of caspase activation and apoptosis. Immunoblotting results indicated that there was no LC3B accumulation upon ART1 treatment, suggesting that autophagy was not mediated in ART1-induced cell death either ( Supporting Information Figure S2b). To define the mechanism by which ART1 induces cancer cell death, MV4;11 cells were treated with ART1 at various concentrations alone or with a panel of cell death-suppressing compounds (Figure 2a). The lethality induced by ART1 treatment was only suppressed by the ferroptosis inhibitor ferrostatin-1 which prevents the accumulation of lipid peroxides. It was not suppressed by the apoptosis inhibitor z-VAD-FMK or the necroptosis inhibitor necrostatin-1 (Nec-1).1 This result clearly showed that ART1 treatment triggers ferroptotic cell death in cancer cells. This conclusion is further supported by the observation that the ART1-induced cell death was rescued in the presence of α-tocopherol, a lipid-soluble anti-oxidant that blocks ferroptosis4,8,45 in H1299 cells (Figure 2b). We then examined whether ART1 can cause lipid peroxidation, the definitive event of ferroptosis.6 As shown in Figure 2c, the fluorescence of the lipid peroxidation probe C11-BODIPY monitored by flow cytometry was elevated in H1299 cells treated with ART1, suggesting that the lipid peroxidation was induced by ART1. This elevation was diminished by cotreatment with the iron chelator DFO, confirming that the lipid peroxidation caused by ART1 is iron-dependent. These results indicate that the structural optimization based on artemisinin generated a highly potent compound ART1 that induces cancer cell death through ferroptosis. Figure 2 | ART1 induces cancer cell death through ferroptosis. (a) Effects of various cell death-suppressing compounds on the viability of MV4;11 cells treated with ART1. Data are represented as mean ± SEM (n = 3). (b) Effects of ferroptosis inhibitors on viability of H1299 cells treated with ART1. Data are represented as mean ± SEM (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001. (c) Ability of iron chelator DFO to prevent the production of lipid peroxidation induced by ART1. The experiment was examined by C11-BODIPY using flow cytometry. (d) GSH/GSSG levels in H1299 cells treated with 10 μM ART1. Data are represented as mean ± SEM (n = 3). (e) Effects of the ferroptosis inducers RSL3 and ART1 on GPX4 activity within GPX4-containing cell lysates. The GPX4 activity in H1299 cells was examined using t-BuOOH as a surrogate substrate by measuring the rate of NADPH oxidation coupled to t-BuOOH-reducing activity of GPX4. (f) Immunoblotting result of GPX4 protein in H1299 cells cotreated with ART1 and α-tocopherol. (g) The content of ferrous iron in H1299 cells treated with 5 μM ART1 at indicated time points. Data are represented as mean ± SD (n = 3). Download figure Download PowerPoint We next examined the mechanisms by which ART1 causes ferroptotic cell death and attempted to learn if ART1 behaves like the two canonical classes of ferroptosis inducers, one by decreasing the intracellular levels of GSH, a key cofactor for the lipid-peroxide-reducing enzyme GPX4, and the other by inhibiting GPX4 activity.4,5,8,10 We quantified the GSH in ferroptosis cells induced by ART1 using a luminescent GSH probe and found that ART1 failed to significantly alter intracellular GSH homeostasis (Figure 2d). This result is consistent with our observation that addition of GSH did not rescue the cell death induced by ART1 treatment (Figure 2a). GPX4 catalyzes the reduction of lipid peroxide to lipid alcohols, making it an important regulator in ferroptotic cell death. The GPX4 activity in H1299 cells was examined using t-BuOOH as a surrogate substrate and measuring the rate of NADPH oxidation coupled