SUMMARY
Ferroptosis is a form of regulated cell death that can be induced by inhibition of the cystine-glutamate antiporter, system xc−. Among the existing system xc− inhibitors, imidazole-ketone-erastin (IKE) is a potent, metabolically stable inhibitor of system xc− and inducer of ferroptosis potentially suitable for in vivo applications. We investigated the pharmacokinetics and pharmacodynamics for IKE in a diffuse large B cell lymphoma (DLBCL) xenograft model and demonstrated that IKE exerted an antitumor effect by inhibiting system xc−, leading to glutathione depletion, lipid peroxidation, and the induction of ferroptosis biomarkers both in vitro and in vivo. Using untargeted lipidomics and quantitative PCR (qPCR), we identified distinct features of lipid metabolism in IKE-induced ferroptosis. In addition, biodegradable polyethylene glycol-poly(lactic-co-glycolic acid) nanoparticles (PEG-PLGA NPs) were employed to aid in IKE delivery and exhibited reduced toxicity compared to free IKE in a DBLCL xenograft model.
eToc (In brief)
Zhang et al. identified imidazole ketone erastin (IKE) as a potent and metabolically stable inhibitor of system xc− and inducer of ferroptosis. The authors show that IKE exerts an anti-tumor effect in a mouse xenograft model and exhibits distinct lipid metabolism effects in vitro and in vivo.
Graphical Abstract
INTRODUCTION
Ferroptosis is a type of regulated cell death driven by iron-dependent lipid peroxidation that can be inhibited by iron chelators and lipophilic antioxidants(Dixon et al., 2012, Stockwell et al., 2017). The ferroptosis-suppressing cystine (Cys2)/glutamate (Glu) antiporter system xc− is required in some contexts to provide cells with cysteine(Stockwell et al., 2017); inhibition of system xc− can induce ferroptosis through cysteine deprivation(Dixon et al., 2012), subsequent glutathione (GSH) depletion, and ultimately inactivation of glutathione peroxidase 4 (GPX4) (Yang et al., 2014, Dixon et al., 2014). However, existing system xc− inhibitors, including sulfasalazine, glutamate, sorafenib, and erastin, are not suitable for in vivo evaluation due to the lack of potency, selectivity, and/or metabolic stability. IKE is an erastin analog with nanomolar potency, high metabolic stability and intermediate water solubility, thus potentially representing a suitable candidate for in vivo evaluation of the impact of ferroptosis driven through system xc− inhibition in mouse models of cancer(Larraufie et al., 2015).
DLBCL is an aggressive malignancy of B-lineage lymphocytes, accounting for 30–40% of non-Hodgkin’s lymphoma. DLBCL cell lines are particularly sensitive to system xc− inhibition(Skouta et al., 2014, Yang et al., 2014), due to their inability to use the transsulfuration pathway to synthesize cysteine from methionine (Gout et al., 2001). DLBCL is clinically heterogeneous, with 60% of patients curable with combination therapy, and the remainder succumbing to the disease(Chapuy et al., 2018). Thus, therapies with distinct mechanisms of action may be beneficial for refractory patients, preventing disease relapse, and improving therapeutic outcomes. Here, we identified ferroptosis-sensitive and resistant DLBCL cell lines, and established a subcutaneous DLBCL xenograft model to study the role of ferroptosis in treating DLBCL. We found that IKE exhibits antitumor activity in this DLBCL xenograft model and an associated increase of lipid peroxidation in tumors. These findings suggest that IKE can be effective as a potential therapeutic regimen for DLBCL.
While small molecule system xc− inhibitors are promising agents for inducing ferroptosis in cancers that are highly dependent on cystine import(Narang et al., 2007, Guan et al., 2009, Lu et al., 2018), there are a number of potential concerns in translating such agents for therapeutic benefit. First, many compounds, including some system xc− inhibitors, do not accumulate at sufficiently high concentrations in target tumor tissues, resulting in a minimal pharmacodynamic effect. Second, while system xc− inhibitors are expected to be largely tolerable based on the observation that Slc7a11 knockout mice have few phenotypes(Sato et al., 2005, McCullagh and Featherstone, 2014), it is possible that systemic inhibition of system xc− would cause toxicity in normal tissues, or that off-target effects of system xc− inhibitors might cause undesirable toxicities. For these reasons, we hypothesized that improving the formulation and delivery of IKE to tumors might enhance its therapeutic index and translational potential. For example, nano-carriers could be used to increase tumor targeting through the enhanced permeability and retention (EPR) effect(Yue et al., 2013, Blanco et al., 2015). Indeed, the PEG-PLGA NP carrier we employed caused increased IKE accumulation in tumors and inhibited tumor growth in mice with less toxicity compared to free IKE. These IKE PEG-PLGA NPs were formulated using a scalable microfluidic NanoAssemblr® Platform(Gdowski et al., 2018), with mean size of ~80 nm, polydispersity index of 0.2, surface charge of −17 mV and loading efficiency of 24% by weight. This formulation could provide a potential therapeutic option and research tool for in vivo studies of DLBCL and other ferroptosis-sensitive cancers.
Several studies have shown the important roles of lipids in ferroptosis(Skouta et al., 2014, Yang et al., 2016, Doll et al., 2017, Kagan et al., 2017, Magtanong et al., 2016). The significant depletion of polyunsaturated fatty acids (PUFAs) and PUFA-containing phospholipids has been reported in erastin-treated and erastin-analog-treated HT1080 cells(Skouta et al., 2014, Yang et al., 2016). Phosphatidylethanolamines (PEs) containing arachidonyl and adrenoyl have been suggested as preferred substrates in RSL3-induced ferroptosis(Kagan et al., 2017, Doll et al., 2017). Moreover, some lipid metabolism genes, including acyl-CoA synthetase long-chain family member 4 (ACSL4) and lysophosphatidylcholine acyltransferase 3 (LPCAT3), have been reported as regulators of ferroptosis(Doll et al., 2017, Kagan et al., 2017). In this study, we applied untargeted mass spectrometry (MS) in conjunction with ultra-high performance liquid chromatography (UPLC) to profile lipid changes in DLBCL cells and a DLBCL xenograft mouse model in response to IKE treatment. We observed significant changes in 62 lipid species in vitro and 37 lipid species in vivo upon IKE treatment, including phospholipids, triglycerides (TAG), diglycerides (DAG), monoglycerides (MAG) and free fatty acids (FA). In addition, a series of genes encoding lipid metabolism enzymes, including the lipid de novo biosynthetic enzymes acetyl-CoA carboxylase 1 (ACC1) and elongation of very long chain fatty acids protein 7 (ELOVL7), lipid-remodeling enzymes adipose triglyceride lipase (ATGL), secretory phospholipase A2f (sPLA2F), lysophosphatidylethanolamine acyltransferase 1 (LPEAT1), and LPCAT4, and lipid peroxidation enzymes lipoxygenases 12 and 15 (ALOX12 and ALOX15), were upregulated upon IKE treatment. Co-treatment with ferroptosis inhibitors ferrostatin-1 (fer1) or β-mercaptoethanol (β-Me) in IKE-treated cells prevented the upregulation of these genes, indicating that upregulation comes from system xc− inhibition and downstream lipid peroxidation. Collectively, these results suggest the involvement of lipids and lipid metabolism genes in IKE-induced ferroptosis and the protective effect of ferroptosis inhibitors in preventing IKE-induced lipid changes.
RESULTS
IKE potently reduces DLBCL cell number
The ferroptosis inducer and system xc− inhibitor erastin is a useful small molecule for in vitro applications, but it is metabolically labile and has low water solubility and potency, precluding its use in vivo. The small molecule IKE is an erastin analog incorporating a carbonyl (Fig. 1A), that can potentially form a reversible covalent interaction with proteins, resulting in >100x potency improvement comparing to erastin in some cell lines(Larraufie et al., 2015). Substitution of an ethoxy moiety with isopropoxy resulted in improved metabolic stability, and the imidazole moiety in IKE helped increase water solubility and stability of the ketone, making IKE soluble under acidic conditions.
Figure 1.
IKE is a potent ferroptosis inducer. (A) The small molecule IKE with isopropoxy, ketone, and imidazole substitutions is more potent than erastin. (B) DLBCL cell line sensitivity was measured by incubating cells with a two-fold series dilution of IKE (starting from 100 μM) for 24 hours followed by a Cell Titor-Glo luminescent cell viability test. LY8, DOHH2, SUDHL5, SUDHL2, SUDHL6, SUDHL10, and SUDHL16 were classified as sensitive cell lines (red). HBL1, SUDHL9, WSU-NHL, SUDHL8, RIVA, and KARPAS422 were classified as moderately resistant cell lines (black). SUDHL7, U937, A3/KAW, LY7, and U2932 were classified as resistant cell lines (blue). (C) GSH level was measured with a fluorometric-green probe in SUDHL6 cells treated with different concentration of IKE with or without 10 μM β-ME for 24 hours. (D) Lipid ROS detected with C11-BODIPY fluorescence was measured in SUDHL6 cells treated with different concentration of IKE with or without 10 μM fer-1 co-treatment. (E) RT-qPCR performed in SUDHL6 cells treated with 500 nM IKE for different lengths of time detected SLC7A11, CHAC1, and PTGS2 mRNA induction compared with DMSO-treated samples. (F) Co-treatment of 10 μM fer-1 with 500 nM IKE for 6 hours prevented PTGS2 upregulation, but not SLC7A11 and CHAC1 upregulation, while 10 μM β-ME co-treatment prevented the upregulation of all three mRNAs. Data are plotted as the mean ± s.d., n=2 (B) and (C), or n=3 (E) and (F) biological replicates. Three independent experiments were performed with similar results (B) to (F). See also Figure S1.
We sought to evaluate DLBCL cell line sensitivity to ferroptosis and to establish a lymphoma xenograft model using a ferroptosis-sensitive DLBCL cell line. A panel of 18 DLBCL cell lines representing distinct DLBCL subtypes, including germinal center B cell-like (GCB), activated B cell-like (ABC), and unclassified subgroups, were evaluated. The 18 DLBCL cell lines showed differential sensitivity to IKE inhibition, with cell lines exhibiting IC50 < 100 nM classified as sensitive cell lines, those with IC50 > 10 μM classified as resistant cell lines, and those with IC50 values between 100 nM to 10 μM classified having intermediate resistance (Fig. 1B). We further tested the degree of IKE-induced lethality upon co-treatment with the ferroptosis inhibitor fer-1, a radical-trapping antioxidant that inhibits lethal lipid peroxidation during ferroptosis (Skouta et al., 2014, Zilka et al., 2017). Co-treatment with fer-1 rescued cell death induced by IKE in DLBCL cell lines, indicating that IKE-induced lethality in these cell lines resulted from lipid peroxidation and ferroptosis (Fig. S1A). Among sensitive DLBCL cell lines, we selected SUDHL6 for generating a subcutaneous xenograft model in 6-week-old male NCG mice.
IKE pharmacodynamic (PD) study in vitro
We aimed to investigate whether IKE specifically inhibited system xc− and induced ferroptosis in DLBCL cells. Previous studies found that IKE inhibited glutamate release, and the IKE parental analog erastin inhibited cystine uptake. Thus, we tested the cellular level of reduced glutathione (GSH), which requires cysteine for its biosynthesis, as a readout of IKE potency. A fluorometric method revealed dose-dependent GSH depletion by IKE (Fig. 1C); this effect was reversed by co-treatment with 10 μM β-ME, which reduces cystine to cysteine, allowing its import into cells through systems A, ASC, and L, thus circumventing inhibition of system xc−. The IC50 of GSH depletion by IKE was 34 nM (Fig. S1B) in SUDHL-6 cells, while, sulfasalazine’s IC50 for GSH depletion is in the millimolar range(Narang et al., 2007, Lo et al., 2010).
We next sought to evaluate whether IKE treatment causes lipid peroxidation, a marker of ferroptosis, in DLBCL cells. Analysis of lipid ROS by flow cytometry using the lipid peroxidation probe C11-BODIPY revealed a dose-dependent increase in lipid ROS upon IKE treatment in SUDHL-6 cells (Fig. 1D). Co-treatment with 10 μM fer-1 inhibited this signal, as expected (Fig. 1D). We reasoned that lipo-oxidative stress markers, including lipid-peroxide-product-derived protein modifications might also be used as ferroptosis biomarkers. Malondialdehyde (MDA) is a naturally occurring reactive carbonyl compound that is derived from lipid peroxidation of PUFAs. Immunofluorescence staining with an anti-dihydropyridine-MDA-lysine-adduct antibody (mAb 1F83)(Yamada et al., 2001), revealed that dihydropyridine-MDA-lysine-adduct abundance was increased upon IKE treatment (Fig. S1C and S1D).
To facilitate molecular characterization of ferroptosis induction in DLBCL cells, we analyzed gene expression biomarkers of ferroptosis in these cell lines by RT-qPCR. Expression of genes involved in ferroptosis(Stockwell et al., 2017) were analyzed upon IKE treatment. Glutathione peroxidase 4 (GPX4), cystathionine beta-synthase (CBS), and acyl-coA synthetase long-chain family member 4 (ACSL4) were not upregulated with IKE treatment (data not shown). However, the system xc− component SLC7A11, prostaglandin-endoperoxide synthase 2 (PTGS2, which encodes cyclooxygenase-2), and ChaC glutathione-specific gamma-glutamylcyclotransferase 1 (CHAC1) expression were significantly increased in SUDHL-6 cells following IKE treatment (Fig. 1E). Compared to the upregulation of PTGS2 and CHAC1, induction of SLC7A11 expression requires a relatively longer-term IKE treatment. PTGS2 was reported to be upregulated by the lipid peroxidation product 4-hydroxy-2-nonenal (4-HNE)(Uchida, 2017) and by GPX4 depletion(Sengupta et al., 2013). Co-treatment with fer-1 inhibited PTGS2 upregulation, suggesting that PTGS2 is a functional biomarker of ferroptosis. Fer-1 co-treatment did not inhibit CHAC1 upregulation, as expected, indicating that CHAC1 is a parallel downstream marker of system xc− inhibition not affected by lipid peroxidation (Fig. 1F). In addition, co-treatment with 10 μM β-ME prevented IKE-induced upregulation of SLC7A11, PTGS2, and CHAC1, suggesting that the upregulation of these genes is downstream of cysteine starvation.
Untargeted lipidomics of IKE-treated DLBCL cells
To investigate the effects of IKE on lipid composition and metabolites, we performed untargeted mass-spectrometry-based lipidomics and gene expression analysis of related enzymes during lipid biosynthesis and oxidation. SUDHL6 cells treated with 1 μM IKE, and 500 nM IKE with or without 10 μM fer-1, 10 μM β-Me, or 10 μM deferoxamine (DFO) co-treatment, were subjected to UPLC-MS analysis. The annotations of the lipid species were determined with LipoStar software (Version 1.0.4, Molecular Discovery, UK)(Goracci et al., 2017). We observed significant (one-way ANOVA, p < 0.05) alterations in the relative abundance of 62 lipid species including lysophosphatidylcholines (LPC), phosphatidylcholines (PC), phosphatidylethanolamines (PE), and TAGs mainly containing PUFAs in IKE-treated samples, while co-treatment with β-ME reversed these effects and maintained the levels of these specific lipids to near the amounts on vehicle-treated cells (Fig. 2A and 2B). The decrease in the phospholipids PC and PE upon IKE treatment is consistent with the effects of piperazine-erastin treatment and erastin treatment on lipid composition in HT-1080 cells(Yang et al., 2016, Skouta et al., 2014). The decrease in TAG lipids upon IKE treatment indicates TAG lipids may be susceptible to oxidation during ferroptosis. Fer-1 co-treatment significantly increased TAG lipids and decreased MAG lipids, suggesting a possible protective role of TAG as a buffer against oxidative damage(Listenberger et al., 2003). With the induction of oxidants and oxidative stress, lipids can be oxidized via both iron-mediated and enzyme-mediated lipid peroxidation. While, co-treatment with DFO inhibited IKE-induced cell death in culture (Fig. S1E), it only partially eliminated lipidomic changes upon IKE treatment, possibly due to DFO inhibiting iron-mediated lipid peroxidation, but not enzyme-mediated lipid peroxidation, suggesting that only a subset of lipidomic changes are needed for inducing cell death.
Figure 2.
Heat map of significantly changed (one-way ANOVA, p < 0.05) lipid species in SUDHL6 cells treated with DMSO, 500 nM IKE, 1 μM IKE, 500 nM IKE with 10 μM fer-1 co-treatment, 500 nM IKE with10 μM β-ME co-treatment, 500 nM IKE with10 μM DFO co-treatment, measured by LC-MS. Each row represents z-score-normalized intensities of the detected lipid species in (A) negative electrospray ionization mode and (B) positive electrospray ionization mode. Each column represents an independent biological replicate. The lipid abundance is color-coded, with red indicating high signal intensity, and dark blue indicating low signal intensity.</P/>Abbreviations: PC, phosphatidylcholine; PE, phosphatidylethanolamine; PS, phosphatidylserine, LPC, lysoPC; PE P-, plasmalogen PE; TAG, triacylglycerol; DAG, diacylglycerol, MAG, monoacylglycerol. (C) Fold change in expression of ACC1, ELOVL7, ATGL, sPLA2F, LPCAT4, LPEAT1, ALOX12, and ALOX15 with 500 nM IKE treatment, 500 nM IKE with 10 μM fer-1 co-treatment, and 500 nM with 10 μM β-ME co-treatment compared with DMSO control in SUDHL6 cells. Data are plotted as the mean ± s.d., n=3 biologically independent samples. (D) Schematic view of fatty acid biosynthesis, lipid remodeling, and arachidonic acid oxidation. The genes upregulated upon IKE treatment were labeled with red. See also Figure S2.
By measuring the mRNA levels of lipid biosynthesis enzymes upon IKE treatment, we discovered that de novo lipid biosynthesis pathways, phospholipid remodeling pathways, and arachidonic acid oxidation pathways were activated (Fig. 2C). First, there was significant upregulation of ACC1, which catalyzes the rate-limiting step of fatty acids biosynthesis from acetyl-coA to malonyl-coA and ELOVL7, which catalyzes the rate-limiting reaction of long-chain fatty acid elongation, especially the elongation of C18:3 (n-3) and C18:3 (n-6)-coAs (Figure 2C and S2). Second, lipid peroxidation during ferroptosis is a deleterious process, and one way to repair such damage is to remodel lipids, by selectively cleave the oxidized PUFA tail, replacing it with non-oxidized fatty acids, and subsequently reduce lipid hydroperoxides with glutathione peroxidase(Chakraborti, 2003). The activation of the phospholipid remodeling pathway, including sPLA2, which selectively releases PUFAs at the sn-2 position of phospholipids, ATGL, which hydrolyzes TAGs, LPCAT4 and LPEAT1, which catalyze the conversion of LPC to PC, and LPE to PE, respectively, is thus important for the oxidative-damage-repair process. Third, in addition to Fenton-chemistry-mediated lipid peroxidation, enzyme-mediated lipid peroxidation was also activated in IKE-induced ferroptosis, as there was increased expression of lipoxygenases ALOX12 and ALOX15. Genes upregulated following IKE treatment were highlighted in red in the schematic lipid metabolism overview (Fig. 2D). The upregulation of these genes was partially reversed by co-treatment of 10 μM fer-1 or 10 μM β-ME (Fig. 2C), indicating that the observed changes in gene expression result from cysteine depletion and oxidative stress induced by IKE. In summary, these lipidomic changes provides us with a detailed picture of IKE-induced ferroptosis in cell culture at the level of lipid metabolism.
IKE pharmacokinetics (PK) and pharmacodynamics (PD) in vivo
To determine the suitability of IKE for in vivo studies, we first evaluated multiple dosage routes by administrating a single dose of IKE (50 mg/kg, 5% DMSO in HBSS at pH 4) using intraperitoneal (IP), intravenous (IV), and oral (PO) routes in NOD/SCID mice. Determination of IKE concentration over a period of eight hours revealed IP to be the most effective and practical means of IKE administration (Table S1). Next, IKE concentration in plasma and tumor samples was determined after a single dose of IKE (50 mg/kg, 5% DMSO in HBSS at pH 4, IP) in SUDHL6-xenograft-bearing NCG mice over a period of 24 hours. IKE reached the highest plasma concentration of 5.2 μg/mL at 1.35 h, and the highest tumor accumulation of 2.5 μg/mL at 3.30 h (Fig. 3A, Table S2).
Figure 3.
IKE-induced ferroptosis biomarker changes in a lymphoma xenograft model. (A) A pharmacokinetic study performed in SUDHL6 subcutaneously xenografted NCG mice measured IKE accumulation over time in plasma and tumor using LC-MS. (B) Analysis of GSH extracted from tumor tissue by fluorometric-green showed over 50% GSH depletion with IKE treatment starting at 4 hours. *** P<0.001, **** P<0.0001 by one-way ANOVA. Data are plotted as the mean ± s.d., n=3 individual mice (A) to (B). (C) RT-qPCR performed using RNA extracted from tumor tissue showed PTGS2, SLC7A11, and CHAC1 mRNA increase starting from 3 hours, colored as blue in the heatmap. Data are plotted as the mean, n=3 individual mice. (D) and (E) Immunofluorescence of dihydropyridine-MDA-lysine adduct in paraffin-embedded tumor samples from mice four hours after treatment with one dose of vehicle or IKE. Quantification of fluorescence intensity showed 1.6-fold increase of dihydropyridine-MDA-lysine adducts and 1.5-fold increase of 8-OH-dG in IKE-treated mouse tumor tissue relative to the vehicle-treated mouse tumor tissue (sections were cut from 3 mice in each group, five images from each section were captured on Zeiss LSM 800 63x/1.40 Oil DIC objective). *** P<0.001, **** P<0.0001 by t test. Heatmaps of dysregulated lipids in SUDHL6 subcutaneously xenografted NCG mice treated with one dosage IKE for at 0, 1, 2, 3, 4, 6, and 24 hours or with vehicle, detected by untargeted UPLC-MS analysis. Data shown in heatmaps are lipid species that were identified as being statistically significant (p < 0.05) among the groups in (F) negative electrospray ionization mode and (G) positive electrospray ionization mode. Each column represents an independent biological replicate. Each row represents z- score normalized intensity of a lipid feature. The relative abundance of each identified lipid is color coded in blue, indicating low signal intensity, and red indicating high signal intensity. Abbreviations: PC, phosphatidylcholine; PE, phosphatidylethanolamine; PI, Phosphatidylinositol; PS, phosphatidylserine, LPC, lysoPC; LPE, lysoPE; FA, free fatty acid; TAG, triacylglycerol; DAG, diacylglycerol, MAG, monoacylglycerol. See also Figure S3 and Table S2.
We collected tumor samples from each time point after a single dose of IKE (50 mg/kg, 5% DMSO in HBSS) and evaluated the abundance of ferroptosis markers. There was significant GSH depletion upon IKE dosing starting from 4 h (Fig. 3B). The GSH depletion in tumors induced by IKE was persistent at 24 h, even when there was little IKE left in the bulk tumor sample. Consistent with IKE accumulation in the tumor samples, CHAC1, SLC7A11, and PTGS2 mRNAs were upregulated starting at 3 h (3–5 fold for PTGS2 and SLC7A11, and 2-fold for CHAC1) (Fig. 3C). Immunofluorescence analysis of samples at 4 h post-treatment showed that there was increased abundance of a hydropyridine MDA-lysine adduct and 8-hydroxy-2’-deoxyguanosine (8-OH-dG), biomarkers for oxidative lesions, indicating that IKE induced lipid peroxidation and more general oxidative stress in these tumor samples (Fig. 3D to 3E).
IKE untargeted lipidomic study in vivo
We sought to investigate lipidomic changes caused by IKE treatment in vivo. We performed untargeted lipidomics on tumor tissue with a single dose of IKE at different time points. We identified significant (one-way ANOVA p < 0.05) increases in the relative abundance of free fatty acids, phospholipids, and diacylglycerols (DAG) upon IKE treatment (Fig. 3F and Fig. 3G). Differences with the cell culture experiment might stem from the different tumor microenvironment in vivo. The lipids identified were enriched in linoleic acid and arachidonic acid metabolism (Fig. S3A). The significant increase in the levels of DAGs and free fatty acids may result from ATGL-mediated-TAG hydrolysis (Fig. S3B). The increased fatty acids might in turn promote phospholipid remodeling to synthesize specific phospholipids, including PC and PE. To explore the free fatty acids effects on cells and ferroptosis, we performed a cell survival test of free fatty acids in the presence or absence of IKE. We found that PUFAs, including γ-linolenic acid, eicosapentaenoic acid, and arachidonic acid, were toxic to cells at high concentration (100 μM) (Fig. S3C). Most free fatty acids (10 μM) sensitized cells to ferroptosis, while oleic acid and palmitoleic acid (10 μM) protected against ferroptosis, which might be due to the fact these two fatty acids lack bis-allylic sites, which are the sites for lipid peroxidation initiation (Gaschler and Stockwell, 2017) (Fig. S3D).
IKE PEG-PLGA NPs have suitable properties to be applied in vivo
IKE is soluble in acidic aqueous conditions, but not to the same degree in neutral aqueous conditions (Fig. 1A). To improve delivery of the compound, we sought to use a nanoparticle formulation. We selected biocompatible and biodegradable PEG-PLGA di-block copolymer-based nanoparticles as an IKE carrier system (Fig. 4A). The PEG block was used to create a deformable hydrating layer by tight associations with water molecules, which prevents clearance by the mononuclear phagocyte system (MPS), prolonging circulation lifetime. The PLGA block was used to form a hydrophobic core to incorporate IKE, which provides sustained release by diffusion and surface and bulk erosion.
Figure 4.
IKE PEG-PLGA NPs < 100 nm in diameter were formulated and showed reduced toxicity in SUDHL6-subcutaneous-xenografted mice. (A) A NanoAssemblr equipped with a microfluidic mixer was used to formulate PEG-PLGA NPs (polymer structure shown). The nanoparticles were characterized with a Zetasizer Nano ZS to have a mean diameter of 80 nm, polydispersity index of 0.17, and surface charge of −17 mV. (B) IKE PEG-PLGA NP cellular activity was measured by Cell Titer-Glo luminescence cell viability assay in SUDHL6 cells with 24-hour incubation. The encapsulated IKE in PEG-PLGA NPs was measured by LC-MS. The x-axis shows IKE concentration. Data are plotted as the mean ± s.d., n=2 biologically replicates. Three independent experiments were performed with similar results. (C) Tumor volume fold change compared with day 0, measured by electronic caliper daily and calculated using the formula volume = 0.5 × length × width2, revealed a reduction of tumor growth by IKE 40 mg/kg (n=9), IKE 23 mg/kg (n=11), and IKE NP 23 mg/kg (n=13) treatment for 14 days comparing with vehicle (n=13) and NP vehicle (n=13) controls (data analyzed by two-way ANOVA). (D) Mouse weight was measured daily and indicated no weight loss upon NP and IKE NP treatments, but weight loss upon IKE 40 mg/kg and IKE 23 mg/kg treatment starting from day 9. Data are plotted as the mean ± s.d. See also Figure S4.
To formulate <100 nm nanoparticles with a reproducible method, we employed the high flow microfluidic system NanoAssemblr®(Gdowski et al., 2018, Valencia et al., 2013). Self-assembly of nanoparticles by nanoprecipitation occurs when the organic phase containing IKE and the PEG-PLGA polymer combines with the aqueous phase in a microchannel. IKE PEG-PLGA nanoparticles (NPs) with a diameter as small as 80 nm and polydispersity index (PDI) of 0.17 were formulated using this approach (Figure S4). The surface possessed a slightly negative charge with zeta potential of −17.0 mV to prevent nanoparticle aggregation. One of the problems with the use of nanoparticles as drug carriers in vivo is the low content of drug and low concentration of nanoparticles in suspension. We were able to achieve concentration factors up to 20 using centrifugal filter units without causing nanoparticle aggregation or dramatically decreasing IKE loading efficiency (Fig. S4C). As a result, IKE PEG-PLGA NPs containing IKE as high as 3 mg/mL (drug loading 3.65% by weight and encapsulation efficiency of 24%) were formulated. Compared to free IKE, IKE PEG-PLGA NPs had enhanced cellular activity in SUDHL6 cells (Fig. 4B), which might result from increased internalization of IKE NPs into cells. Co-treatment with fer-1 rescued cell death induced by IKE and IKE PEG-PLGA NPs (Fig. S4G). In addition, administration of 750 mg/kg naked PEG-PLGA NPs daily for two weeks didn’t result in any detectable weight loss or other observable signs of toxicity, suggesting that this NP system was suitable for in vivo drug delivery.
IKE inhibits tumor growth in vivo and the PEG-PLGA NP formulation enhances its therapeutic index
We investigated the efficacy of IKE in vivo in male NCG mice bearing SUDHL6 subcutaneous xenografts. Once tumor volumes reached 100 mm3, mice were randomized into five groups and treated with vehicle (5% DMSO in HBSS at pH 4), unfunctionalized PEG-PLGA NPs in water, 40 mg/kg free IKE (5% DMSO in HBSS at pH 4), 23 mg/kg free IKE (5% DMSO in HBSS at pH 4), or 23 mg/kg IKE NP (IKE PEG-PLGA NPs in water) via IP injection once daily. During the experimental period, mouse weight and tumor volume were measured daily to determine IKE’s antitumor effect and possible toxicity. Tumor growth was calculated as the fold change to original tumor volume on day 0 before the first dose (Fig. 4C). Administration of 40 mg/kg IKE, 23 mg/kg IKE, and 23 mg/kg IKE NPs caused a significant decrease in tumor growth starting from day 9 of treatment. The tumor growth inhibition effect was not significantly different between 23 mg/kg free IKE and 23 mg/kg IKE NP; however, IKE NPs showed less toxicity, as evidenced by weight loss (Fig. 4D). Compared to saline vehicle, free IKE-(5% DMSO in HBSS at pH 4)-treated mice started losing weight from day 9, which might be caused by the precipitation of IKE after administrated into the peritoneum, an environment with pH ranges of 7.5–8.0, causing damage to abdominal organs, or possible toxicity of systemic system xc− inhibition, or off-target toxicity of IKE. However, IKE-NP-treated mice had a similar weight as the saline vehicle and the NP vehicle groups; the lower toxicity of the IKE NP formulation might result from the NP’s capability to prevent the aggregation of hydrophobic drugs(Sun et al., 2014), or the NP EPR effect, which decreases the non-specific distribution and systemic toxicity associated with conventional hydrophobic drugs(Yue et al., 2013). By analyzing IKE tumor accumulation using LC-MS, we found that IKE NPs at 23 mg/kg had slightly enhanced tumor accumulation compared with free IKE at 23 mg/kg and were comparable to the free IKE 40 mg/kg treatment (Fig. S5A). Overall, the PEG-PLGA NP formulation increased IKE’s therapeutic window.
Given the fact that PTGS2 mRNA was upregulated upon IKE treatment in vivo (Fig. 3C), we hypothesized that after IKE treatment, there might be an increased level of the PTGS2 gene product (COX-2) in the tumor. Using immunofluorescence (Fig. 5A), we found that COX-2 protein expression was indeed increased in IKE-treated tumor tissue (Fig. 5D). There was no significant COX-2 protein expression increase in IKE-NP-treated tumors, which might be due to the slower release of IKE from PEG-PLGA NPs (Fig. S4F), making the free IKE concentration lower than free-IKE-treated tumors. This implies that COX-2 protein abundance is not a highly sensitive pharmacodynamic marker for low concentrations of IKE, but may more effectively report on exposure to high concentrations of IKE.
Figure 5.
IKE treatment induced lipid peroxidation in tumor tissue during the efficacy study. (A, D) Immunofluorescence of COX-2 on frozen tumor sections showed 2.0 and 2.8-fold increase on COX-2 intensity upon IKE 23 mg/kg and IKE 40 mg/kg treatment daily for 14 days. (B, E) Immunofluorescence of MDA on frozen tumor sections measured by confocal microscopy showed 1.4, 1.8, 1.8, and 2.0-fold increase in MDA intensity upon NP vehicle, IKE NP 23 mg/kg, IKE 23 mg/kg, and IKE 40 mg/kg treatment daily for 14 days. (C, F) Immunofluorescence of 8-OH-dG on frozen tumor sections measured by confocal microscopy showed 2.3, 1.8, and 2.2-fold increase in 8-OH-dG intensity upon IKE NP 23 mg/kg, IKE 23 mg/kg, and IKE 40 mg/kg treatment daily for 14 days. (Sections were cut from 5 randomly chosen mice in each group, three images from each section were captured on Zeiss LSM 800 63x/1.40 Oil DIC objective) (G) TBARS assay measuring MDA-TBA adduct fluorescence showed 1.6, 1.4, and 1.8-fold MDA increase upon IKE NP 23 mg/kg, IKE 23 mg/kg and IKE 40 mg/kg treatment. ns P>0.5, * P<0.05, ** P<0.01, *** P<0.001, **** P<0.0001 by one-way ANOVA. Data are plotted as the mean ± s.d.
We further characterized lipid peroxidation in tumor tissues during this efficacy study. Immunofluorescence analysis of tumor tissue from the five groups showed significantly increased dihydropyridine-MDA-lysine adduct levels in IKE-treated and IKE-PEG-PLGA-NP-treated groups, compared with the saline vehicle and NP vehicle groups (Figure 5B and 5E). Compared to saline vehicle, the NP vehicle group had increased levels of dihydropyridine-MDA-lysine adducts, which might due to low-level ferroptosis induction of ultrasmall (<10 nM in diameter) nanoparticles in the sample, as observed in PEG-coated silica nanoparticles previously(Kim et al., 2016). In addition, the Thiobarbituric Acid Reactive Substances (TBARS) assay, which measures fluorescence of MDA-TBA adducts formed, showed increased MDA levels in tumors in the free IKE and IKE PEG-PLGA NP groups compared to vehicle groups (Figure 5G). The TBARS assay is not as sensitive as immunofluorescence to detect small increases in MDA adducts, suggesting a possible explanation for our inability to detect this increase in the NP vehicle group. Immunofluorescence analysis of 8-OH-dG showed increased oxidative DNA damage in IKE-treated and IKE-PEG-PLGA-NP-treated groups, confirming that IKE treatment induced oxidative damage (Fig. 5C and 5F). In addition, there was no increase in caspase-3 activity in tumor tissues upon IKE treatment, suggesting that the inhibition of tumor growth was not due to apoptosis, as expected (Fig. S5B).
DISCUSSION
Recent discoveries have suggested that ferroptosis inducers can have anti-tumor efficacy(Yu et al., 2017, Gout et al., 2001, Liu et al., 2017) and may synergize with chemotherapy in some cell contexts(Sato et al., 2018, Chen et al., 2015, Yamaguchi et al., 2013, Yu et al., 2015). However, there is a lack of potent, selective, and metabolically stable tools to study ferroptosis in vivo. Study of the ferroptosis inducer IKE in an SUDHL6 xenograft model demonstrated that IKE and IKE NP reduce tumor growth by inducing ferroptosis. The IKE PEG-PLGA NP formulation exhibited a larger therapeutic window comparing with the free IKE formulation tested. Given the fact that we used a 5% DMSO in HBSS at pH 4 formulation and that IKE is less soluble at pH 7 versus pH 4, we suspect that the toxicity of high dose free IKE in this study might stem from the precipitation of IKE after injection into the peritoneal cavity.
The induction of ferroptosis theoretically has the potential to shrink tumors. IKE’s T1/2 plasma and T1/2 tumor in SUDHL6 NCG xenograft mice through IP injection were 1.83 h and 3.50 h, respectively. Compared with the pharmacokinetics of IKE, the once-per-day-dosage frequency might be too low to see a tumor shrinking effect; in other words, we might need either a longer half-life compound or formulation, or more frequent dosing to balance rapid tumor growth with induction of cell death, to cause tumor regression. The IKE nanoparticle formulation had increased accumulation in tumor tissue relative to plasma (Figure S5A), but it might not be high enough to counter rapid tumor growth in vivo; delivery of a higher dose to the tumor might be needed. Based on above considerations, combining IKE with other therapies or designing ferroptosis inducers with enhanced bioavailability and metabolic stability, and higher tumor penetration and accumulation are worth investigating to further optimize the therapeutic impact of ferroptosis inducers in diverse cancer models.
Our untargeted lipidomic study identified lipid metabolism features during ferroptosis. First, we observed decreased PEs, PCs, and TAGs upon IKE treatment, possibly resulting from the cleavage of oxidized PUFA tails in PE, PC, and TAG to prevent the oxidative damage to cells induced by IKE treatment. Co-treatment with β-ME, which allows transport of cysteine through systems A, ASC, and L, completely reversed the lipidomic changes caused by IKE treatment. Fer-1 partially prevented the phospholipid decreases in cells, but dramatically increased TAGs. The increased TAG pool might serve as a buffer against lipid-peroxidation induced by IKE(Listenberger et al., 2003). However, the iron chelator DFO didn’t reverse IKE-induced phospholipid decreases or TAG decreases, indicating that Fenton-chemistry-mediated lipid peroxidation is not likely a driver of most lipidomic changes during IKE-induced ferroptosis. Second, both the de novo and the remodeling pathways of phospholipids and TAG biosynthesis(Shindou and Shimizu, 2009) were activated following IKE treatment. Finally, lipidomic changes following induction of ferroptosis in vivo was characterized here. The data from the tumor samples revealed that DAG, MAG and phospholipids were all significantly increased upon IKE treatment, possibly resulting from the activation of the TAG hydrolysis enzyme, ATGL, and the response of tumor cells to oxidative stress.
In addition, pharmacodynamic markers developed in this study may be beneficial for future ferroptosis efficacy studies. First, the RT-qPCR experiments identified a time-dependent upregulation of numerous genes, including PTGS2, SLC7A11, and CHAC1 following IKE treatment in SUDHL6 cells and in xenograft tumor tissues derived from the same cell line. Second, immunofluorescence experiments showed increased levels of the PTGS2 gene product COX-2, 8-OH-dG, and dihydropyridine-MDA-lysine adducts in IKE-treated and IKE-PEG-PLGA-nanoparticle-treated tumors compared to vehicle, with no increased caspase-3 cleavage, a biomarker of apoptosis, in IKE-treated and IKE-PEG-PLGA-nanoparticle-treated tumor tissues. These data collectively suggest that the antitumor effects of IKE result from system-xc−inhibition-induced ferroptosis, but not apoptosis.
DLBCL cell lines have differential sensitivity to IKE-induced ferroptosis. The intrinsic factors determining DLBCL ferroptosis sensitivity remain to be explored to increase the understanding to ferroptosis and to enable selection of patients with sensitive tumors. The efficacy of IKE in additional animal cancer models would also be valuable to explore. Moreover, further studies on genes encoding lipid metabolism enzymes identified here would be useful to determine the essentiality of these genes for ferroptosis execution and the contribution of their expression to cell sensitivity to ferroptosis. In summary, we have developed a suitable small molecule and formulation for inhibiting system xc− and inducing ferroptosis in mouse tumor models and identified a series of pharmacodynamic markers of ferroptosis. These tools may be beneficial in determining whether there are specific cancer contexts in which ferroptosis induction would be therapeutically beneficial. The lipidomic study performed here may increase our understanding of the roles of lipids in ferroptotic cell death.
SIGNIFICANCE
Emerging evidence has shown the potential antitumor effects of using system xc− inhibitors and ferroptosis inducers as single reagents or in combination with chemotherapy in cell culture and some types of xenografts. It is thus important to identify biomarkers and understand the molecular mechanisms of ferroptosis inducers in both tumor cells and cancer models. Here we identify and establish the pharmacokinetic and pharmacodynamic properties of the ferroptosis inducer IKE in a xenograft model, which will be useful for future study of the therapeutic impact of ferroptosis in cancers and other conditions. The work presented here also identified a set of genes that are upregulated by IKE treatment in cell culture and demonstrated that the upregulation of these genes is due to system xc− inhibition, as illuminated following co-treatment with ferroptosis inhibitors. The integration of small molecule tools, untargeted lipidomics, and RT-qPCR provides an efficient means to explore changes in lipid composition and lipid metabolic mechanisms during regulated cell death. Moreover, nanoparticle carriers provide an effective way to improve drug efficacy and reduce systemic toxicity because of targeted localization in tumors. Despite the various nanocarrier systems available and numerous advantages of nanoparticle therapeutics, there are challenges to apply nanocarriers in vivo. One such challenge is to manufacture uniform nanoparticles with high loading capacity at large scale. We utilized a scalable microfluidic platform, which allows for the precise engineering of PEG-PLGA nanoparticles. Given the rapid optimization process and the commercially availability of PEG-PLGA polymers, the nanoparticle system we used here may be useful to other researchers exploring in vivo delivery of ferroptosis inducers.
STAR Methods
CONTACT FOR REAGENTS AND RESOURCES SHARING
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact Brent R. Stockwell ([email protected]).
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Cell lines and cell culture information is listed below. The SUDHL-5 (female), SUDHL-6 (male), SUDHL-16 (male), and HT-1080 (male) cell lines were obtained from ATCC. The DOHH-2 (male) cell line was obtained from DSMZ. The HBL-1 (male), U2932 (female), SUDHL-7 (male), SUDHL-9 (male), A4/FUK (female), WSU-NHL (female), Ly18 (OCI-Ly18) (male), Karpas422 (female), SUDHL-1 (male), SUDHL-2 (female), SUDHL-8 (male), SUDHL-10 (male), A3/KAW (female), RIVA (female), Ly9 (OCI-Ly9) (male), U937 (male), and Ly7 (OCI-Ly7) (male) cell lines were provided by Dr. Owen A. O’Connor (Columbia University) and the Columbia Genome Center. HT-1080 cells were grown in DMEM with glutamine and sodium pyruvate (Corning 10–013) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Life Technologies, 10437036) 1% non-essential amino acids (Thermo Fisher Scientific, 11140076) and 1% penicillin-streptomycin mix (pen-strep) (Thermo Fisher Scientific, 15140148). SUDHL-1, SUDHL-2, SUDHL-6, SUDHL-8, SUDHL-10, SUDHL-16, A3/KAW, RIVA, Ly8 and U937 cells were cultured in RPMI-1640 (ATCC 30–2001) with 10% FBS and 1% pen-strep. DOHH-2, HBL-1, U2932, SUDHL-7, SUDHL-9, A4/FUK, WSU-NHL, and Ly8 were cultured in RPMI-1640 with 10% heat-inactivated FBS and 1% pen-strep. SUDHL-5 and Karpas422 were cultured in RPMI-1640 with 20% FBS and 1% pen-strep. Ly7 was cultured in IMDM (Thermo Fisher Scientific, 12440053) with 15% heat-inactivated FBS and 1% pen-strep. All cells were maintained in a humidified environment at 37 °C and 5% CO 2 in an incubator.
The animal models used in the paper is listed below. The animal studies reported in this manuscript adhere to the ARRIVE guidelines. All animal study protocols were approved by the Columbia University Institutional Animal Care and Use Committee (IACUC). NODPrkdcem26Cd52Il2rgem26Cd22/NjuCrl (NCG) mice (Charles River, strain code 572) (male, 5-weeks of age) and NOD.CB17-Prkdcscid/J (NOD SCID) mice (The Jackson Laboratory, stock number 001303) (male, 5-weeks of age)) were acclimated after shipping for >3 days before beginning experiments. Mice were fed a standard diet (PicoLab 5053) and maintained with no more than 5 mice per cage.
METHODS DETAILS
Animal Studies
Pharmacokinetic analysis in mice with three different administration routes
NOD/SCID mice (12-weeks of age and ~28 g weight) were weighed before injection and divided into groups of 3 mice per cage. IKE was dissolved in 5% DMSO/95% Hank’s Balanced Salt Solution (HBSS) (Thermo Fisher Scientific, 14025076), pH 4, to create a 5 mg/mL solution. 5% DMSO/95% HBSS at pH 4 solution (Vehicle 1) without IKE was used as vehicle. The solution was sterilized using a 0.22 μm Steriflip filter unit (Thomas Scientific 1189Q46). Mice were dosed using three different routes, IP and PO with 50 mg/kg IKE, and IV with 17 mg/kg IKE. Samples were collected at 0, 1, 3, 4, and 8 h from three mice per time point. Additionally, three mice per group were used as controls by administration with equivalent amount of vehicle 1 by IP, PO, and IV, and samples were collected at 8 h. At the appropriate time, mice were sacrificed by CO2 asphyxiation for 3 min and ~0.5 mL of blood was collected via cardiac puncture. Blood was immediately put into K3 EDTA micro tube (SARSTEDT 41.1504.105) and placed on ice. Samples were centrifuged for 10 min at 2,100 × g at 4°C, then plasma was transferred to a clean tube. Plasma samples were flash frozen in liquid nitrogen and stored at −80°C. IKE was extracted from plasma by adding 900 μL acetonitrile to 100 μL plasma. Samples were mixed for at least 5 min by rotating at room temperature and were sonicated prior to concentration for 10 min at 4,000 × g and 4°C. The supernatant was removed and dried on a GeneVac evaporator overnight on an HPLC setting. After drying, the samples were re-suspended in 100 μL of methanol and analyzed on the liquid chromatography mass spectrometry (LC-MS), with each sample analyzed twice. Quality control standard samples were prepared by dissolving IKE in 100 μL water and extraction with the same procedures to ensure that the extraction was efficient. LC-MS analysis was performed on a platform comprising a Thermo Scientific Dionex Ultimate 3000RS controlled by Chromeleon (Dionex) and a Bruker amazon SL ESI ion-trap mass spectrometer.
Chromatographic separation was performed at 20°C on Agilent Eclipse Plus C18 column (2.1 × 50 mm, 3.5 μm) at 20°C over a 12 minute gradient elution. Mobil e phase A consisted of water with 0.1% acetic acid v/v and mobile phase B was methanol with 0.1% acetic acid v/v. After injection, the gradient was held at 80% mobile phase A for 1 min. The gradient was then ramped in a linear fashion to 80% mobile phase over 0.5 min. Over the next 3.5 min, the gradient was ramped in linear fashion to 100% mobile phase B and held at 100% mobile phase B for 3.25 min. The gradient was then ramped in a linear fashion to 80% mobile phase over 0.5 min and held there for the duration of the run. The flow rate was set to 400 μL/min and injection volumes were 5 μL. The retention time for IKE in this gradient was 2.8 min.
Mass spectrometry analysis was performed on a Bruker Amazon SL (Billerca, MA) in positive ESI mode. Trap Control was used to control the ESI settings with the inlet capillary held at −4500 V and the end plate offset at −500 V. Nitrogen was used as the desolvation gas. Hystar v3.2 was used to integrate the UHPLC and MS applications, and data analysis was performed with the Compass DataAnalysis software. The base peak chromatogram at m/z 655.2 with a width of ± 0.1 was integrated and peak area quantified by standard curve.
Pharmacokinetics (PK) of IKE were assessed using Prism fitted with lognormal of one phase exponential decay. The PK parameters are summarized in Table S1.
Pharmacokinetic and pharmacodynamic analysis in NCG mice bearing SUDHL6 xenografts
IKE was dissolved in 5% DMSO/95% HBSS at pH 4 to create a 5 mg/mL solution or 3 mg/mL solution. 5% DMSO/95% HBSS at pH 4 was used as vehicle 1. IKE PEG-PLGA nanoparticles and unfunctionalized PEG-PLGA nanoparticles (without IKE) (vehicle 2) prepared with a NanoAssemblr were dialyzed with deionized water overnight, and the water was changed at least twice. Dialyzed IKE-PEG-PLGA nanoparticles and unfunctionalized PEG-PLGA nanoparticles were concentrated by Amicon Ultra-15 Centrifugal Filter Units to create a solution with 80 mg/mL PEG-PLGA nanoparticles. All above solutions were sterilized by filtering through a 0.2 μm syringe filter.
NCG mice 6 weeks of age were injected with 10 million SUDHL-6 cells subcutaneously. Visible tumors appeared after 2 weeks. Tumor size was measured by electronic caliper every 2 days and calculated using the formula: 0.5 × length × width2. After another 2 weeks, mice were randomly separated into group of 3 mice per cage with roughly tumor size (1,200 mm3). Mice were dosed at 50 mg/kg IKE using IP at one time and samples were collected at 0, 1, 2, 3, 4, 6, and 24 h with three mice per time point. Mice were dosed with DMSO vehicle by IP and samples were collected at 24 h. Mice were euthanized using a CO2 gas chamber before xenograft dissection and cardio puncture. Plasma samples were collected and analyzed as described above. Tumors were dissected and divided into 6 segments, frozen on dry ice, and stored at −80°C. Before IKE extraction, tumor tissu e samples were thawed at room temperature and weighed. A 2.5-fold ratio of volume of phosphate-buffered saline (PBS) (mL/g) was added to the sample and homogenized using Bead Ruptor 4 at speed 4 for 30 sec. The homogenization step was repeated until the samples were homogenized well (final weight/volume was 0.4 g/mL). 100 μL of homogenized tissue (equal to 40 mg) was added to a new microfuge tube, 900 μL of acetonitrile was added. The samples were mixed for at least 5 minutes by rotating on a shaker and were sonicated for at least 30 sec prior to centrifugation for 10 min at 4,000 × g and 4°C. The supernatant was removed and dried on the GeneVac overnight on the HPLC setting. After drying, the samples were resuspended in 100 μL of methanol and analyzed on the LC-MS with each sample analyzed twice. Technical replicates were averaged. The concentration of IKE in the sample was determined by comparison against a standard curve of IKE in the range of 25 to 2,500 ng/mL. Pharmacodynamic analysis of ferroptosis biomarkers is described in the RT-qPCR, glutathione, immunofluorescence, and lipidomics sections below. Pharmacokinetics parameters of IKE was assessed using Prism fitted with Lognormal of one phase exponential decay. The PK parameters are summarized in Table S2.
IKE efficacy study
IKE was dissolved in 5% DMSO/95% HBSS at pH 4 to create a 4 mg/mL solution. 5% DMSO/95% HBSS at pH 4 was used as vehicle 1. IKE PEG-PLGA nanoparticles and unfunctionalized PEGPLGA nanoparticles (without IKE loading) (vehicle 2) prepared with a NanoAssemblr were dialyzed with deionized water overnight; the water was changed at least twice. Dialyzed IKE-PEG-PLGA nanoparticles and unfunctionalized PEG-PLGA nanoparticles were concentrated by Amicon Ultra-15 Centrifugal Filter Units to create a solution with 80 mg/mL PEG-PLGA nanoparticles. All above solutions were sterilized by filter through a 0.2 μm syringe filter.
NCG mice, 6-weeks old, were injected with 10 million SUDHL0–6 cells subcutaneously. The mice were treated after the tumor size reached 100 mm3. Mice were separated randomly into treatment groups and dosed with 23 mg/kg IKE, 40 mg/kg IKE, 23 mg/kg IKE PEG-PLGA nanoparticles (equal to 700 mg/kg PEG-PLGA), vehicle 1 (based on volume), and vehicle 2 (700 mg/kg PEG-PLGA) once daily by IP for 14 days. Tumor volume was measured daily with electronic caliper and calculated using the formula: 0.5 × length × width2. 3 h after the final dosage, mice were euthanized with CO2 and tumor tissue was dissected, weighed, divided into 4–6 segments, frozen, and stored at −80°C. Tumor vo lume change was analyzed in Prism using one-way ANOVA. IKE accumulation in tumor tissues were analyzed using the same method in the above sections. Biomarker characterization in tumor tissues were described in the RT-qPCR, immunofluorescence, TBARS, and lipidomics sections below.
DLBCL Lines Sensitivity Measurement
DLBCL cells were plated at 10,000 cells per well in white 384-well plates (32 μL per well) in technical duplicates and incubated overnight. The cells were then treated with 8 μL medium containing a two-fold dilution series of vehicle (DMSO), IKE (starting from 100 μM) with or without Fer-1 (starting from 200 μM). After 24 h incubation, 40 μL of 50% CellTiter-Glo (Promega) 50% cell culture medium was added to each well and incubated at room temperature with shaking for 15 min. Luminescence was measured using a Victor X5 plate reader (PerkinElmer). All cell viability data were normalized to the DMSO vehicle condition. Experiments were performed three independent times with different passages for each cell line. From these data, dose-response curves and IC50 values were computed using Prism 7.0 (GraphPad).
Flow Cytometry Assay
0.20 million SUDHL-6 cells were seeded in six-well plates and treated with DMSO, IKE, or fer-1 at specific concentration. The final cell density was 0.05 million cells/mL. After 24 h, cells were harvested by centrifuging at 300 × g for 5 min. Cells were resuspended in 500 μL HBSS containing 2 μM C11-BODIPY (BODIPY 581/591 C11) (Thermo Fisher Scientific, D3861) and incubated at 37 °C for 15 min. Cells were pelleted and resuspended in HBSS. Fluorescence intensity was measured on the FL1 channel with gating to record live cells only (gate constructed from DMSO treatment group). A minimum of 10,000 cells were analyzed per condition.
Reduced Glutathione Measurement
2.4 million cells were incubated with DMSO, IKE, or β-mercaptoethanol (β-ME) at a density of 0.2 million cells/mL for 24 h. Cells were collected and washed with cold PBS once. Cell number was counted by Vi-Cell. Cells were resuspended in ice-cold RIPA buffer (Thermo Fisher Scientific, 89900) with 100 μL/one million cells. Samples were centrifuged for 15 min at 4°C at 17,000 × g. The resulting supernatant was deproteinized using a trichloroacetic acid and sodium bicarbonate solution and kept on ice. The sample was diluted with assay buffer provided in GSH/GSSG Ratio Detection Assay Kit (Abcam, ab13881) ten-fold. Reduced glutathione (GSH) levels were determined using Fluorometric-Green provided in the kit following the manufacture’s protocol. 384-well (Corning) low volume black flat bottom polystyrene non-treated microplates 1230F99 were used.
To perform reduced glutathione measurement in tumor tissue, 10 mg of tumor tissue was mixed with 400 μL RIPA buffer, homogenized at speed 5 for 30 sec using a Bead Ruptor 4 (OMNI International). The sample was centrifuged at 17,000 × g for 5 min, then the above deprotenization and measurement steps performed.
TBARS Assay
10–25 mg tumor tissue was placed into a 1.5 mL Eppendorf microfuge tube, and the sample was placed on dry ice until use. 20 μL/mg ice-cold RIPA buffer with cocktail protease inhibitor (Roche) was added. The tumor tissue was homogenized at speed 5 for 30 sec using Bead Ruptor 4. The sample was centrifuged at 4°C at 1,60 0 × g for 10 min. The sample was kept on ice and the supernatant used for analysis. Tissue homogenates did not need to be diluted before assaying. TBARS was measured using a TBARS Assay Kit (Cayman 700870), following the manufacture’s protocol.
qPCR
To perform qPCR analysis in cell culture, DLBCL cells were treated with 0–10 μM IKE, fer-1, or β-ME for the indicated times. RNA was extracted using the Qiashredder and Qiagen RNeasy Mini kits (Qiagen) according to the manufacture’s protocol. 2 μg RNA from each sample was reversed transcribed to cDNA using a High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, 4368814). Quantitative PCR reactions were performed using Power SYBR Green PCR Master Mix (Applied Biosystems) with triplicate measurement on ViiA 7 Real-Time PCR instrument (Thermo Fischer). HPRT1 was used as an internal reference. Differences in mRNA levels compared with HPRT1 were computed between vehicle and experimental groups using the ΔΔCt method. The primers used in the study was listed below.
| Genes | Foward | Reverse |
|---|---|---|
| HPRT1 | 5’-GCCCTGGCGTCGTGATTAGTG-3’ | 5’-GCCTCCCATCTCCTTCATCAC-3’ |
| CHAC1 | 5’- GAACCCTGGTTACCTGGGC-3’ | 5’- CGCAGCAAGTATTCAAGGTTGT-3’ |
| SLC7A11 | 5’-GGTGGTGTGTTTGCTGTC-3’ | 5’-GCTGGTAGAGGAGTGTGC-3’ |
| PTGS2 | 5’-ATATGTTCTCCTGCCTACTGGAA-3’ | 5’-GCCCTTCACGTTATTGCAGATG-3’ |
| ACSL4 | 5’-CCCCGCTATCT CTCAGACAC-3’ | 5’-GGTGCTCCAACTCTGCCAG-3’ |
| FASN | 5’-CTTCCGAGATTCCATCCTACGC-3’ | 5’-TGGCAGTCAGGCTCACAAACG-3’ |
| ELOVL1 | 5’-TCCCTCTACATTGTCTATGAGTTCC-3’ | 5’-TTCAGTTGGCCTTGACCTTGGCAATACC-3’ |
| ELOVL7 | 5’-AGGATCCATGGCCTTCAGTGATCTTACATCGAGG-3’ | 5’-AACCACCTGCAGCAAATTTGACTCC-3’ |
| ALOX12 | 5’-AGAAAAGTTGACTAGTCCAGTGTGGTGAA-3’ | 5’-AAAAGCTGTGCTAAACCAATTCCGAACAGATTCTCA-3’ |
| ALOX15 | 5’-GGAGCCTTCCTAACCTACAGC-3’ | 5’-CTCACGATTCCTTCCACATACC-3’ |
| LPCAT1 | 5’-CACAACCAAGTGGAAATCGAG-3’ | 5’-GCACGTTGCTGGCATACA-3’ |
| LPCAT3 | 5’-ATCACTGCCGTCCTCACTAC-3’ | 5’-AGTCAACAGCCAAACCAATC-3’ |
| LPCAT4 | 5’-GGTGGGAGAGAATGCCACTT-3’ | 5’-ATGCAAGGGATGATGGCTGT-3’ |
| LPEAT1 | 5’-CTGAAATGTGTGTGCTATGAGCG-3’ | 5’-TGGAAGAGAGGAAGTGGTGTCTG-3’ |
| ACC1 | 5’-GAGGGCTAGGTCTTTCTGGAAG-3’ | 5’-CCACAGTGAAATCTCGTTGAGA-3’ |
| sPLA2F | 5’-TGACGACAGGAAAGGAAGCCGCAC-3’ | 5’-AGGGAAGAGGGGACTCAGCAACGAG-3’ |
| PLA2G6C | 5’-CTGGAACCTGTGTTGGACCT-3’ | 5’-CGGTGATATCTGTGGTCACG-3’ |
| ATGL | 5’-CAACGCCACTCACATCTACGG-3’ | 5’-GGACACCTCAATAATGTTGGCAC-3’ |
To perform qPCR analysis in mouse tumor tissue, ~20 mg of tumor tissue was prepared and kept on dry ice. The frozen tissue was homogenized with Bead Ruptor 4 at speed 4 for 30 sec. The homogenization was repeated once if there remained unhomogenized tissue. Total mRNA was prepared from the cleared homogenate using Ribopure Kit (Thermo Fisher Scientific, am1924). RNA was reverse transcribed to cDNA then used in quantitative PCR experiment following the same steps used in cell qPCR experiment.
Immunofluorescence Study and Quantification
Immunofluorescence study and quantification on cells
SUDHL-6 cells were treated with IKE. The cells were harvested by centrifugation and washed with PBS once. The cells were resuspended in PBS, fixed by adding equal volume of 4% paraformaldehyde (PFA), and incubated at room temperature for 15 min. The cells were washed with PBS/0.1% Tween 20 (PBST) twice, resuspended in 10% goat serum (ThermoFisher 50197Z) for 1 h. The cells were incubated with mouse mAb 1F83, which specifically recognizes the malondialdehyde (MDA)-lysine adduct 4-methyl-1,4-dihydropyridine3,5-dicarbaldehyde (MDHDC)(Yamada et al., 2001) (1:100 dilution) overnight at 4°C. The cells were washed with PBS three times by centrifugation. The cells were incubated with goat anti-mouse IgG H&L (Alexa Fluor 647) (Abcam ab150115) (1:1000) at room temperature for 1 h. The cells were washed with PBST twice by centrifugation, then resuspended in PBS in 24-well plate with poly-lysine-(Sigma Aldrich P4832)-coated cover slips and centrifuged at 1,000 × g for 10 min. ProLong Diamond antifade mountant with DAPI (ThermoFisher P36962) was added to stain the nucleus. All images were captured on a Zeiss LSM 800 confocal microscope at Plan-Apochromat 63x/1.40 Oil DIC objective with constant laser intensity for all samples. When applicable, line-scan analysis was performed on representative confocal microscopy images using Zeiss LSM software to qualitatively visualize fluorescence overlap. The intensity above threshold of the fluorescent signal of the bound antibodies was analyzed using NIH ImageJ software. Data were expressed as fold change compared with the vehicle.
Immunofluorescence study and quantification on frozen tissue sections
Tumor tissues were fixed in 4% paraformaldehyde (PFA) for 24h at 4°C followed by washing with PBS three times. The tissues were perfused in 30% sucrose for 24h at 4°C for cryoprotection. The samples were embedded in OCT cryostat sectioning medium, and then moved directly into a cryostat. After equilibration of temperature, frozen tumor tissues were cut into 5 μm thick sections. Tissue sections were mounted on to poly-L-lysine coated slides by placing the cold sections onto warm slides. Slides were stored at −80°C until staining. For staining, slides were warmed to room temperature followed by washing with PBS twice. A hydrophobic barrier pen was used to draw a circle on each slide. The slides were permeabilized with PBS/0.4% Triton X-100 twice before non-specific-binding blocking by incubating the sections with 10% goat serum (ThermoFisher 50197Z) for 30 minutes at room temperature. The sections were separately incubated with mouse anti-MDA mAb 1F83 (1:1000 dilution), anti-cyclooxygenase 2 (COX 2, AKA PTGS2) antibody (Abcam, ab15191, 1:200 dilution), or anti-8-OH-dG (DNA/RNA damage) antibody (Abcam, ab62623, 1:200 dilution) overnight at 4°C in humidified chambers. Sections were washed with PBST for twice before incubating with goat anti-mouse IgG H&L (Alexa Fluor 647) (Abcam, ab150115, 1:1000 dilution) or goat anti-rabbit IgG H&L highly cross-absorbed secondary antibody (Alexa Fluor488, Thermo Fisher Scientific, A-11034, 1:1000) at room temperature for 1 h. Slides were then washed twice with PBST. ProLong Diamond antifade mountant with DAPI (ThermoFisher P36962) was added onto slides, which were then covered with the coverslips, sealed by clear fingernail polish and observed under confocal microscopy. All images were captured on a Zeiss LSM 800 confocal microscope at Plan-Apochromat 63x/1.40 Oil DIC objective with constant laser intensity for all analyzed samples. The intensity above threshold of the fluorescent signal of the bound antibodies was analyzed using NIH ImageJ software. Data were expressed as fold change comparing with the vehicle.
Immunofluorescence study and quantification on paraffin-embedded tissue sections
Tumor tissue was fixed in 4% paraformaldehyde (PFA) for 24 h at 4°C followed by washing three times with PBS. The samples were fixed in paraffin. Six series of 5 μM sections were obtained with a sliding microtome. The serial sections were then mounted on gelatin-coated slide. The paraffin-embedded tissue sections were deparaffinized with xylene three times, 5 min each, followed by rehydrating in 100%, 90%, 70%, and 50% ethanol, two washes 5 min each, then rinsed with distilled water. Antigen retrieval was performed in Tris-EDTA buffer, pH 9.0, 95–100°C for 10 min. Then sections were rinsed in PBST, 2 min each. Then the above steps were followed for staining and quantification.
Western Blot
10–25 mg tumor tissue was placed into a 1.5 mL Eppendorf microfuge tube, and the sample was placed on dry ice until use. 20 μL/mg ice cold RIPA buffer with cocktail protease inhibitor (Roche) was added. The tumor tissue was homogenized at speed 5 for 30 sec using Bead Ruptor 4. The sample was centrifuged at 4°C at 1,600 × g for 10 min. The sample was kept on ice and the supernatant used for analysis. Caspase-3 antibody which could recognize both full-length caspase-3 and cleaved caspae-3 was used. α-tubulin antibody was used as the reference.
Nanoparticle Formulation and Characterization
Formulation
(Poly(ethylene glycol) methyl ether-block-poly(lactide-co-glycolide) (PEG-PLGA) with PEG average Mn 5,000, PLGA Mn 15,000, and lactide:glycolide 50:50 was purchased from SigmaAldrich (900948). PEG-PLGA nanoparticles were prepared using a microfluidic mixer with the NanoAssemblr Benchtop instrument (Precision NanoSystems Inc., Vancouver, BC) by mixing oil and aqueous phases. PEG-PLGA was dissolved in 25% acetone/75% DMSO at concentration of 10 mg/mL. Milli-Q water was used as aqueous phase. For synthesis of PEG-PLGA nanoparticles, two phases were injected through two inlets of the microfluidic mixer with a speed of 8 mL/min and a volume ratio of 1:1. Nanoparticles were collected in the sample collection tube by discarding an initial and final waste volume of 0.25 and 0.05 mL. Samples were dialyzed in at least 500 × deionized water overnight using 10 kD cut-off dialysis bag (ThermoFisher, SnakeSkin Dialysis Tubulin, 68100) and the water was changed twice in between to remove organic solvent. Then the purified nanoparticles were concentrated by centrifuging using Amicon Ultra-15 Centrifugal Filter Units (Millipore, UF9050) at 2,168 × g to the desired concentration. For encapsulation of IKE, IKE was dissolved in a solution of PEG-PLGA in 25% acetone and 75% DMSO. The final concentration of PEG-PLGA was 10 mg/mL and IKE was 1.5 mg/mL. The process was the same with PEG-PLGA nanoparticles preparation. Nanoparticle size and potential were measured using a Malvern Zetasizer Nano ZS.
IKE quantification in PEG-PLGA nanoparticles
IKE concentration was determined in PEG-PLGA nanoparticles using Liquid Chromatography- Mass Spectrometry (LC -MS). IKE was extracted from nanoparticles by mixing 100 μL of samples with 900 μL of acetonitrile. After mixing for 1 h, the mixture was sonicated for 5 min and then centrifuged at 4000 rpm for 10 min. The supernatant was transferred to a new vial. The organic solvent was removed by Genevac evaporation. The residues were re-dissolved in 100 μL MeOH in a LC-MS vial. Calibration standards and quality control samples were prepared spanning a range of 25 ng/ml to 1250 ng/ml IKE in MeOH. Peak integration and data analysis were performed. Using the standard curve, IKE concentration in the sample was determined.
Measurement of IKE PEG-PLGA cellular activity
DLBCL cells or HT-1080 cells were plated at 2,000 cells per well in white 384-well plates (32 μL per well) in technical duplicates and incubated overnight. The cells were then treated with 8 μL medium containing a two-fold dilution series of vehicle 1 (DMSO), IKE, fer-1, vehicle 2 (unfunctional PEG-PLGA nanoparticles), or IKE PEG-PLGA nanoparticles. The final concentration of IKE in PEG-PLGA nanoparticles was measured by LC-MS as described in the sections below. After 24 h incubation with compounds, 40 μL of 50% CellTiter-Glo (Promega) 50% cell culture medium was added to each well and incubated at room temperature with shaking for 15 min. Luminescence was measured using a Victor X5 plate reader (PerkinElmer). All cell viability data were normalized to the DMSO vehicle condition. Experiments were performed in three independent times with different passages for each cell line. From these data, dose-response curves and IC50 values were computed using Prism 7.0 (GraphPad).
Kinetics of IKE release from PEG-PLGA nanoparticles
200 μL newly generated IKE PEG-PLGA nanoparticles were mixed with 800 μL PBS solution. They were incubated on a shaker at 37°C. Samples we re collected followed by centrifugation at 10,000 × g for 10 min. The obtained supernatant was evaporated using a GeneVac. The obtained residues were resuspended in 100 μL methanol and analyzed with LC-MS with the methods described above. Peak integration and data analysis were performed to quantify released IKE concentration in the samples.
Solubility Measurements
Solubility measurement samples were prepared in a 96-well Corning Costar plate with two at 1% and 5% DMSO v/v with a total sample volume of 200 μL in MilliQ water. A BMG Labtech NEPHELOstar nephelometer was used to measure turbidity of the samples with the following parameters per the instrument manufacturer: Gain = 80, Laser Focus= 2.50 mm, and Laser Intensity = 90%. Samples were shaken in the nephelometer by orbital shaking for 3s prior to the turbidity measurement. Sample values were blanked in the instrument software using the corresponding water and DMSO condition values. Turbidity 4000 NTU Calibration Standard-Formazin (Millipore-Sigma) was used as a positive control for all turbidity measurements. The turbidity of each sample was measured within 20 minutes of initial preparation. After the initial measurement, the sample plates were covered with aluminum sealing film and allowed to sit overnight at room temperature prior to re-measuring.
IKE Synthesis
IKE was synthesized using methods previously reported(Larraufie et al., 2015) except that the reduction step of nitrobenzene to aniline was modified, which is described here. To a solution of substituted nitrobenzene (1 equiv.) in MeOH (0.1 M), Pd/C (0.2 equiv.) were added. The reaction mixture was gas exchanged with hydrogen, then stirred at room temperature overnight under 1 atm hydrogen gas. Upon completion, the mixture was filtered with celite and concentrated under vacuo. The resulting mixture was purified by combiflash 0–10% methanol/dichloromethane to afford the corresponding aniline. The characterization of the final product IKE was:
1 H NMR (400 MHz, Dimethyl sulfoxide-d6) δ 8.31 (ddd, J = 8.0, 1.5, 0.6 Hz, 1H), 8.07 (dd, J = 8.7, 2.3 Hz, 1H), 8.01 (d, J = 2.2 Hz, 1H), 7.83 (ddd, J = 8.4, 7.0, 1.5 Hz, 1H), 7.80 – 7.73 (m, 1H), 7.69 (s, 1H), 7.54 (ddd, J = 8.2, 7.0, 1.3 Hz, 1H), 7.29 – 7.21 (m, 2H), 7.25 – 7.13 (m, 2H), 6.98 (s, 1H), 6.93 – 6.83 (m, 2H), 5.48 (d, J = 17.6 Hz, 1H), 5.36 (d, J = 17.6 Hz, 1H), 5.32 (s, 2H), 4.78 – 4.68 (m, 1H), 4.63 (s, 2H), 3.62 (s, 1H), 3.44 (d, J = 16.3 Hz, 4H), 3.29 (d, J = 14.0 Hz, 1H), 3.22 (d, J = 14.0 Hz, 1H), 2.50–2.13 (m, 4H), 1.28 (dd, J = 16.2, 6.0 Hz, 6H). MS (m/z): [M]+ calculated for C35H35ClN6O5, 655.14; found 655.24. HPLC detected purity: 99%
Mass Spectrometry-based Untargeted Lipidomics
Sample preparation in vitro study
Lipids were extracted from each sample as described below. 5 million cells treated with DMSO, 1 μM IKE, 500 nM IKE with or without 10 μM fer-1 and 10 μM β-Me for 24 hours were homogenized in 250 μL cold methanol containing 0.1% butylated hydroxyl toluene (BHT) with micro tip sonicator. Homogenized samples were transferred to fresh glass tubes containing 850 μL of cold methyl-tert-butyl ether (MTBE) and vortex-mixed for 30 sec. To enhance extraction efficiency of lipids, the samples were incubated overnight at 4 °C on the shaker. On the next day, 200 μL of cold water was added to each sample, and incubated for 20 min on ice before centrifugation at 3,000 rpm for 20 min at 4 °C. The organic layer was collected followed by drying under a gentle stream of nitrogen gas on ice and stored at −80 °C until LC-MS analysis. The protein pellet was used to measure protein concentration using Bio-Rad protein assay. The samples were re-constituted in a solution containing IPA/ ACN /water (4:3:1, v/v/v) containing mixture of internal standard for further MS analysis. A quality control (QC) sample was prepared by combining 40 μL of each sample to assess the reproducibility of the features through the runs.
Ultra-high performance liquid chromatography analysis
Chromatographic separation of extracted lipids was carried out at 55 °C on Acquity UPLC HSS T3 column (2.1× 150 mm, 1.8 μm) over a 17-min gradient elution. Mobile phase A consisted of ACN/water (60:40, v/v) and mobile phase B was IPA/ACN/water (85:10:5, v/v/v) both containing 10 mM ammonium acetate and 0.1% acetic acid. After injection, the gradient was held at 60% mobile phase A for 1.5 min. For the next 12 min, the gradient was ramped in a linear fashion to 100% B and held at this composition for 3 min. The eluent composition returned to the initial condition in 1 min, and the column was re-equilibrated for an additional 1 min before the next injection was conducted. The flow rate was set to 400 μL/min and Injection volumes were 5 μL using the flow-through needle mode in both positive and negative ionization modes. The QC sample was injected between the samples and at the end of the run to monitor the performance and the stability of the MS platform. This QC sample was also injected at least 5 times at the beginning of the UPLC/MS run, in order to condition the column.
Mass spectrometry analysis
The Synapt G2 mass spectrometer (Waters, Manchester, U.K.) was operated in both positive and negative electrospray ionization (ESI) modes. For positive mode, a capillary voltage and sampling cone voltage of 3 kV and 32 V were used. The source and desolvation temperature were kept at 120 °C and 500 °C, respectively. Nitro gen was used as desolvation gas with a flow rate of 900 L/hr. For negative mode, a capillary voltage of −2 kV and a cone voltage of 30 V were used. The source temperature was 120 °C, and d esolvation gas flow was set to 900 L/hr. Dependent on the ionization mode the protonated molecular ion of leucine encephalin ([M+H]+, m/z 556.2771) or the deprotonated molecular ion ([M-H]−, m/z 554.2615) was used as a lock mass for mass accuracy and reproducibility. Leucine enkephalin was introduced to the lock mass at a concentration of 2 ng/μL (50% ACN containing 0.1% formic acid), and a flow rate of 10 μL/min. The data was collected in duplicates in the centroid data independent (MSE) mode over the mass range m/z 50 to 1600 Da with an acquisition time of 0.1 seconds per scan.
The QC samples were also acquired in enhanced data independent ion mobility (IMS-MSE) in both positive and negative modes for enhancing the structural assignment of lipid species. The ESI source settings were the same as described above. The traveling wave velocity was set to 650 m/s and wave height was 40 V. The helium gas flow in the helium cell region of the ion-mobility spectrometry (IMS) cell was set to 180 mL/min to reduce the internal energy of the ions and minimize fragmentation. Nitrogen as the drift gas was held at a flow rate of 90 mL/min in the IMS cell. The low collision energy was set to 4 eV, and high collision energy was ramping from 25 to 65 eV in the transfer region of the T-Wave device to induce fragmentation of mobility-separated precursor ions.
Data pre-processing and statistical analysis
All raw data files were converted to netCDF format using DataBridge tool implemented in MassLynx software (Waters, version 4.1). Then, they were subjected to peak-picking, retention time alignment, and grouping using XCMS package (version 3.2.0) in R (version 3.5.1) environment. For the peak picking, the CentWave algorithm was used with the peak width window of 2–25 s. For peak grouping, bandwidth and m/z-width of 2 s and 0.01 Da were used, respectively. After retention time alignment and filling missing peaks, an output data frame was generated containing the list of time-aligned detected features (m/z and retention time) and the relative signal intensity (area of the chromatographic peak) in each sample. Technical variations such as noise were assessed and removed from extracted features’ list based on the ratios of average relative signal intensities of the blanks to QC samples (blank/QC >1.5). Also, peaks with variations larger than 30% in QCs were eliminated. Multivariate and univariate analyses were performed using (version 4.0) and in R (version 3.5.1) environment. Group differences were calculated using VIP scores of PLS-DA model and one-way ANOVA (p < 0.05) and false discovery rate of 5% to control for multiple comparisons.
Structural assignment of identified lipids
Identification and structural characterization of significant lipid features were confirmed with LipoStar(Goracci et al., 2017) (Version 1.0.4, Molecular Discovery, UK). Lipidomix standard (Avanti Polar Lipids, INC., Alabaster, AL, USA) and quality control samples were analyzed in LipoStar with the recommended data processing parameters(Goracci et al., 2017) except MS/MS signal filtering threshold was set to 20 for both positive and negative ionization mode. The precursor ion (MS) and fragment ion information obtained by data independent MS (MSE) were automatically annotated using LipoStar database library with mass tolerances of 5 ppm and 10 ppm, respectively. Annotated lipid species with the highest score and high-confidence identification (matches between experimental and theoretical MS/MS spectra) were approved, and the identified lipids with low-confidence matches were further evaluated manually using MSE data viewer (Version 1.3, Waters Corp., MA, USA). In case of co-eluting and isomeric lipid species (e.g., triacylglycerols), all the available fragments in MS/MS spectra data were reported. However, if fragments for fatty acyl compositions of lipids weren’t clear, the lipid species were annotated as sum of total number of carbons and double bonds (e.g. PC 32:2).
Untargeted lipidomics study in vivo
25 mg frozen tissue was homogenized in 15 μL/mg pre-chilled methanol containing 0.1% butylated hydroxyl-toluene (BHT) at speed 5 for 30 secs using a Bead Ruptor 4 (OMNI International). Then, 300 μL of tissue lysate were transferred to a glass vials containing 1,000 μL ice-cold methyl-tert-butyl ether (MTBE) and vortex-mixed for 30 sec. The sample are stored in the −20 °C freezer overnight to enhance lipids e xtraction. The following day, 250 μL of ice-cold methanol was added to each sample and vortex-mixed for 30 sec vigorously. Then, samples were incubated on dry ice for 20 min on the shaker followed by centrifuge at 3,000 rpm for 20 min at 4°C. Finally, 1,000 μL of upper phase containing lipids were transferred to fresh glass vials, and evaporated to dryness under the stream of N2 gas. The samples were re-constituted in a solution containing IPA/ ACN /water (4:3:1, v/v/v) containing mixture of internal standard for further MS analysis. The other steps including LC-MS analysis and data processing are the same as in vitro lipidomic analysis as mentioned above.
MetScape analysis KEGG ID and HMDB ID
| Lpids | KEGG or HMDB ID | Full name |
|---|---|---|
| FA 14:0 | C06424 | Myristic acid |
| FA 16:1 | C08362 | Palmitoleic acid |
| FA 18:1 | C00712 | Oleic acid |
| FA 18:2 | C01595 | Linoleic acid |
| FA 18:3 | C06426 | Gamma-Linolenic acid |
| FA 20:1 | C16526 | Eicosenoic acid |
| FA 20:2 | C16525 | Eicosadienoic acid |
| FA 20:3 | C03242 | 8,11,14-Eicosatrienoic acid |
| FA 20:4 | C00219 | Arachidonic acid |
| FA 20:5, EPA | C06428 | Eicosapentaenoic acid |
| FA 22:1 | C08316 | Erucic acid |
| FA 22:4 | C16527 | Adrenic acid |
| FA 22:5 | C16513 | Docosapentaenoic acid |
| FA 22:6 | C06429 | Docosahexaenoic acid |
| FA 22:0 | C08281 | Behenic acid |
| FA 24:0 | C08320 | Tetracosanoic acid |
| LPC 18:1 | HMDB0002815 | LysoPC(18:1(11Z)) |
| LPE 18:1 | HMDB0011506 | LysoPE(18:1(9Z)/0:0) |
| PC 16:1/16:1 | HMDB0008002 | PC(16:1(9Z)/16:1(9Z)) |
| PI 18:2/20:3 | HMDB0009854 | PI(18:2(9Z,12Z)/20:3(5Z,8Z,11Z)) |
| FA 16:0 | C00249 | Palmitic acid |
| FA 17:0 | HMDB0002259 | Heptadecanoic acid |
| FA 18:0 | C01530 | Stearic acid |
| FA 24:1 | C08323 | Nervonic acid |
| DAG 34:1 | HMDB0007102 | DG(16:0/18:1(9Z)/0:0) |
| PE 16:1/18:2 | HMDB0008961 | PE(16:1(9Z)/18:2(9Z,12Z)) |
| PC 16:1/18:2 | HMDB0008006 | PC(16:1(9Z)/18:2(9Z,12Z)) |
| MAG 18:1 | HMDB0011567 | MG(18:1(9Z)/0:0/0:0) |
| MAG 16:0 | HMDB0011533 | MG(0:0/16:0/0:0) |
| DAG 34:2 | HMDB0007103 | DG(16:0/18:2(9Z,12Z)/0:0) |
| DAG 16:0/20:2 | HMDB0007109 | DG(16:0/20:2(11Z,14Z)/0:0) |
| DAG 18:1/18:2 | HMDB0007190 | DG(18:1(11Z)/18:2(9Z,12Z)/0:0) |
| LPE 18:1 | HMDB0011506 | LysoPE(18:1(9Z)/0:0) |
| TAG 16:0/20:4/20:5 | HMDB0044553 | TG(16:0/20:4(8Z,11Z,14Z,17Z)/20:5 (5Z,8Z,11Z,14Z,17Z)) |
| DAG 18:2/20:1 | HMDB0007253 | DG(18:2(9Z,12Z)/20:1(11Z)/0:0) |
| PC 16:1/16:1 | HMDB0008002 | PC(16:1(9Z)/16:1(9Z)) |
| PC 16:1/18:2 | HMDB0008006 | PC(16:1(9Z)/18:2(9Z,12Z)) |
| PE 16:0/16:1 | HMDB0008924 | PE(16:0/16:1(9Z)) |
| PE 16:0/18:1 | HMDB0008927 | PE(16:0/18:1(9Z)) |
| PS 34:3 | HMDB0012409 | PS(18:3(9Z,12Z,15Z)/16:0) |
| LPC 18:1 | HMDB0002815 | LysoPC(18:1(9Z)) |
QUANTIFICATION AND STATISTICAL ANALYSIS
T-test, one-way ANOVA, and two-way ANOVA were performed in R (version 3.5.1) environment and GraphPad Prism7 with significance and confidence level 0.05 (95% confidence interval). MetScape analysis is performed on MetScape version 3.1.
DATA AND SOFTWARE AVAILABILITY
Please follow the Key Resources Table for software downloading. Further information and requests for lipidomics data should be directed to and will be fulfilled by the Lead Contact Brent R. Stockwell ([email protected]).
KEY RESOURCES TABLE.
| REAGENT or RESOURCE | SOURCE | IDENTIFIER |
|---|---|---|
| Antibodies | ||
| Anti-dihydropyridine-MDA-lysine adduct mouse mAb 1F83 | Reference (Yamada et al., 2001) | N/A |
| Anti-8-OH-dG antibody | Abcam | ab62623 |
| Anti-xCT antibody | Abcam | ab37185 |
| Anti-cyclooxygenase 2 (COX-2) antibody | Abcam | ab15191 |
| Goat anti-mouse IgG H&L (Alexa Fluor 647) | Abcam | ab150115 |
| Goat anti-rabbit IgG H&L highly cross-adsorbed secondary antibody (Alexa Fluor488) | Thermo Fisher Scientific | A11034 |
| Donkey anti-Goat IgG (H+L) cross-adsorbed, Alexa Fluor 647, polyclonal, secondary antibody | Thermo Fisher Scientific | A21447 |
| Anti-15-F2T-isoprostane purified IgG | Oxford Biomedical Research | IS20 |
| Anti-α-tubulin antibody (DM1A) | Santa Cruz Biotechnology | sc-32293 |
| Anti-xCT/SLC7A11 (D2M7A) antibody | Cell Signaling Technology | 12691 |
| Anti-caspase-3 antibody | Cell Signaling Technology | 9662 |
| Chemicals | ||
| (Poly(ethylene glycol) methyl ether-block-poly(lactide-co-glycolide)(PEG-PLGA) | Sigma-Aldrich | 900948 |
| β-mercaptoethanol | Sigma-Aldrich | M3148 |
| Imidazole ketone erastin (IKE) | Reference(Larraufie et al., 2015) | N/A |
| Ferrostatin-1 (fer-1) | Reference(Skouta et al., 2014) | N/A |
| Critical Commercial Assays | ||
| GSH/GSSG Ratio Detection Assay Kit | Abcam | ab13881 |
| TBARS Assay Kit | Cayman | 700870 |
| BODIPY™ 581/591 C11 | Thermo Fisher Scientific | D3861 |
| RiboPure™ RNA Purification Kit | Thermo Fisher Scientific | AM1924 |
| CellTiter-Glo Luminescent Cell Viability Assay | Promega | G7573 |
| RNAeasy extraction kit | QIAGEN | 74106 |
| QIAshredder | QIAGEN | 79656 |
| High Capacity cDNA Reverse Transcription Kit | Thermo Fisher Scientific | 4368814 |
| Power SYBR Green PCR Master Mix | Thermo Fisher Scientific | 4368702 |
| Cell culture reagents and others | ||
| ProLong™ Gold anti-fade Mount with DAPI | Thermo Fisher Scientific | P36962 |
| 10% goat serum | Thermo Fisher Scientific | 50197Z |
| Donkey serum | Sigma-Alderich | D9663 |
| Poly-lysine | Sigma-Aldrich | P4832 |
| Fetal bovine serum | Thermo Fisher Scientific | 10437–028 |
| DMEM | Corning | 10–013 |
| Fetal bovine serum | Life Technologies | 10437036 |
| RPMI-1640 | ATCC | 30–2001 |
| Non-essential amino acids | Thermo Fisher Scientific | 11140076 |
| Penicillin-streptomycin mix | Thermo Fisher Scientific | 15140148 |
| IMDM | Thermo Fisher Scientific | 12440053 |
| HBSS | Thermo Fisher Scientific | 14025076 |
| Steriflip filter unit | Thomas Scientific | 1189Q46 |
| Amicon Ultra-15 centrifugal filter units | Millipore | UF9050 |
| K3 EDTA micro tube | SARSTEDT | 41.1504.105 |
| RIPA buffer | Thomas Scientific | 89900 |
| Software and Algorithms | ||
| ChemDraw Ultra, Version 14.0 | Perkin Elmer | http://www.perkinelmer.com/category/chemdraw |
| Prism, Version 7.0 | GraphPad Software | https://www.graphpad.com/scientific-software/prism/ |
| LipoStar, Version 1.0.4 | Molecular Discovery | https://www.moldiscovery.com/software/lipostar/ |
| MetaboAnalyst, Version 4.0 | MetaboAnalyst | https://www.metaboanalyst.ca/faces/docs/About.xhtml |
| MassLynx, Version 4.1 | Waters | http://www.waters.com/waters/en_US/MassLynx-Mass-Spectrometry-Software-/nav.htm?cid=513164&locale=en_US |
| XCMS package, Version 3.2.0 | Bioconductor | http://packages.renjin.org/package/org.renjin.bioconductor/xcms |
Supplementary Material
Highlights.
IKE is a potent, selective, and metabolically stable system xc− inhibitor
IKE inhibited tumor growth in a diffuse large B cell lymphoma mouse model
Untargeted lipidomic profiling revealed lipid metabolism features of ferroptosis
PEG-PLGA nanoparticles have reduced toxicity in a mouse xenograft model
ACKNOWLEDGEMENTS
We thank Dr. Brandon Fowler of Columbia University for assistance with mass spectrometry of IKE measurement and Hanqing Xiong of Columbia University for assistance with immunofluorescence and pharmacokinetics analysis. This research was supported by the National Cancer Institute (R35CA209896 and P01CA087497 to B.R.S).
Footnotes
DECLARATION OF INTERESTS
BRS holds equity in and consults with Inzen Therapeutics. BRS is an inventor on a patent related to this work. BRS and YZ are inventors on a patent application related to this work.
SUPPLEMENTAL INFORMATION
Supplemental Information includes five figures, two tables, and methods file.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- BLANCO E, SHEN H & FERRARI M 2015. Principles of nanoparticle design for overcoming biological barriers to drug delivery. Nature biotechnology, 33, 941. [DOI] [PMC free article] [PubMed] [Google Scholar]
- CHAKRABORTI S 2003. Phospholipase A2 isoforms: a perspective. Cellular signalling, 15, 637–665. [DOI] [PubMed] [Google Scholar]
- CHAPUY B, STEWART C, DUNFORD AJ, KIM J, KAMBUROV A, REDD RA, LAWRENCE MS, ROEMER MGM, LI AJ, ZIEPERT M, STAIGER AM, WALA JA, DUCAR MD, LESHCHINER I, RHEINBAY E, TAYLOR-WEINER A, COUGHLIN CA, HESS JM, PEDAMALLU CS, LIVITZ D, ROSEBROCK D, ROSENBERG M, TRACY AA, HORN H, VAN HUMMELEN P, FELDMAN AL, LINK BK, NOVAK AJ, CERHAN JR, HABERMANN TM, SIEBERT R, ROSENWALD A, THORNER AR, MEYERSON ML, GOLUB TR, BEROUKHIM R, WULF GG, OTT G, RODIG SJ, MONTI S, NEUBERG DS, LOEFFLER M, PFREUNDSCHUH M, TRUMPER L, GETZ G & SHIPP MA 2018. Molecular subtypes of diffuse large B cell lymphoma are associated with distinct pathogenic mechanisms and outcomes. Nat Med, 24, 679–690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- CHEN L, LI X, LIU L, YU B, XUE Y & LIU Y 2015. Erastin sensitizes glioblastoma cells to temozolomide by restraining xCT and cystathionine-γ-lyase function. Oncology reports, 33, 1465–1474. [DOI] [PubMed] [Google Scholar]
- DIXON SJ, LEMBERG KM, LAMPRECHT MR, SKOUTA R, ZAITSEV EM, GLEASON CE, PATEL DN, BAUER AJ, CANTLEY AM & YANG WS 2012. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell, 149, 1060–1072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- DIXON SJ, PATEL DN, WELSCH M, SKOUTA R, LEE ED, HAYANO M, THOMAS AG, GLEASON CE, TATONETTI NP & SLUSHER BS 2014. Pharmacological inhibition of cystine–glutamate exchange induces endoplasmic reticulum stress and ferroptosis. Elife, 3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- DOLL S, PRONETH B, TYURINA YY, PANZILIUS E, KOBAYASHI S, INGOLD I, IRMLER M, BECKERS J, AICHLER M, WALCH A, PROKISCH H, TRUMBACH D, MAO G, QU F, BAYIR H, FULLEKRUG J, SCHEEL CH, WURST W, SCHICK JA, KAGAN VE, ANGELI JP & CONRAD M 2017. ACSL4 dictates ferroptosis sensitivity by shaping cellular lipid composition. Nat Chem Biol, 13, 91–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- GASCHLER MM & STOCKWELL BR 2017. Lipid peroxidation in cell death. Biochem Biophys Res Commun, 482, 419–425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- GDOWSKI A, JOHNSON K, SHAH S, GRYCZYNSKI I, VISHWANATHA J & RANJAN A 2018. Optimization and scale up of microfluidic nanolipomer production method for preclinical and potential clinical trials. Journal of Nanobiotechnology, 16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- GORACCI L, TORTORELLA S, TIBERI P, PELLEGRINO RM, DI VEROLI A, VALERI A & CRUCIANI G 2017. Lipostar, a comprehensive platform-neutral cheminformatics tool for lipidomics. Analytical chemistry, 89, 6257–6264. [DOI] [PubMed] [Google Scholar]
- GOUT P, BUCKLEY A, SIMMS C & BRUCHOVSKY N 2001. Sulfasalazine, a potent suppressor of lymphoma growth by inhibition of the x c− cystine transporter: a new action for an old drug. Leukemia, 15, 1633. [DOI] [PubMed] [Google Scholar]
- GUAN J, LO M, DOCKERY P, MAHON S, KARP CM, BUCKLEY AR, LAM S, GOUT PW & WANG Y-Z 2009. The x c− cystine/glutamate antiporter as a potential therapeutic target for small-cell lung cancer: use of sulfasalazine. Cancer chemotherapy and pharmacology, 64, 463. [DOI] [PubMed] [Google Scholar]
- KAGAN VE, MAO G, QU F, ANGELI JPF, DOLL S, ST CROIX C, DAR HH, LIU B, TYURIN VA & RITOV VB 2017. Oxidized arachidonic and adrenic PEs navigate cells to ferroptosis. Nature chemical biology, 13, 81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- KIM SE, ZHANG L, MA K, RIEGMAN M, CHEN F, INGOLD I, CONRAD M, TURKER MZ, GAO M, JIANG X, MONETTE S, PAULIAH M, GONEN M, ZANZONICO P, QUINN T, WIESNER U, BRADBURY MS & OVERHOLTZER M 2016. Ultrasmall nanoparticles induce ferroptosis in nutrient-deprived cancer cells and suppress tumour growth. Nat Nanotechnol, 11, 977–985. [DOI] [PMC free article] [PubMed] [Google Scholar]
- LARRAUFIE M-H, YANG WS, JIANG E, THOMAS AG, SLUSHER BS & STOCKWELL BR 2015. Incorporation of metabolically stable ketones into a small molecule probe to increase potency and water solubility. Bioorganic & medicinal chemistry letters, 25, 4787–4792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- LISTENBERGER LL, HAN XL, LEWIS SE, CASES S, FARESE RV, ORY DS & SCHAFFER JE 2003. Triglyceride accumulation protects against fatty acid-induced lipotoxicity. Proceedings of the National Academy of Sciences of the United States of America, 100, 3077–3082. [DOI] [PMC free article] [PubMed] [Google Scholar]
- LIU DS, DUONG CP, HAUPT S, MONTGOMERY KG, HOUSE CM, AZAR WJ, PEARSON HB, FISHER OM, READ M, GUERRA GR, HAUPT Y, CULLINANE C, WIMAN KG, ABRAHMSEN L, PHILLIPS WA & CLEMONS NJ 2017. Inhibiting the system xC(−)/glutathione axis selectively targets cancers with mutant-p53 accumulation. Nat Commun, 8, 14844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- LO M, LING V, LOW C, WANG Y & GOUT P 2010. Potential use of the anti-inflammatory drug, sulfasalazine, for targeted therapy of pancreatic cancer. Current oncology, 17, 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- LU B, CHEN XB, YING MD, HE QJ, CAO J & YANG B 2018. The role of ferroptosis in cancer development and treatment response. Frontiers in pharmacology, 8, 992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- MAGTANONG L, KO PJ & DIXON SJ 2016. Emerging roles for lipids in non-apoptotic cell death. Cell Death and Differentiation, 23, 1099–1109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- MCCULLAGH EA & FEATHERSTONE DE 2014. Behavioral characterization of system xc-mutant mice. Behav Brain Res, 265, 1–11. [DOI] [PubMed] [Google Scholar]
- NARANG VS, PAULETTI GM, GOUT PW, BUCKLEY DJ & BUCKLEY AR 2007. Sulfasalazine-induced reduction of glutathione levels in breast cancer cells: enhancement of growth-inhibitory activity of Doxorubicin. Chemotherapy, 53, 210–7. [DOI] [PubMed] [Google Scholar]
- SATO H, SHIIYA A, KIMATA M, MAEBARA K, TAMBA M, SAKAKURA Y, MAKINO N, SUGIYAMA F, YAGAMI K, MORIGUCHI T, TAKAHASHI S & BANNAI S 2005. Redox imbalance in cystine/glutamate transporter-deficient mice. J Biol Chem, 280, 37423–9. [DOI] [PubMed] [Google Scholar]
- SATO M, KUSUMI R, HAMASHIMA S, KOBAYASHI S, SASAKI S, KOMIYAMA Y, IZUMIKAWA T, CONRAD M, BANNAI S & SATO H 2018. The ferroptosis inducer erastin irreversibly inhibits system xc− and synergizes with cisplatin to increase cisplatin’s cytotoxicity in cancer cells. Scientific reports, 8, 968. [DOI] [PMC free article] [PubMed] [Google Scholar]
- SENGUPTA A, LICHTI UF, CARLSON BA, CATAISSON C, RYSCAVAGE AO, MIKULEC C, CONRAD M, FISCHER SM, HATFIELD DL & YUSPA SH 2013. Targeted disruption of glutathione peroxidase 4 in mouse skin epithelial cells impairs postnatal hair follicle morphogenesis that is partially rescued through inhibition of COX-2. Journal of Investigative Dermatology, 133, 1731–1741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- SHINDOU H & SHIMIZU T 2009. Acyl-CoA: lysophospholipid acyltransferases. Journal of biological chemistry, 284, 1–5. [DOI] [PubMed] [Google Scholar]
- SKOUTA R, DIXON SJ, WANG J, DUNN DE, ORMAN M, SHIMADA K, ROSENBERG PA, LO DC, WEINBERG JM & LINKERMANN A 2014. Ferrostatins inhibit oxidative lipid damage and cell death in diverse disease models. Journal of the American Chemical Society, 136, 4551–4556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- STOCKWELL BR, ANGELI JPF, BAYIR H, BUSH AI, CONRAD M, DIXON SJ, FULDA S, GASCÓN S, HATZIOS SK & KAGAN VE 2017. Ferroptosis: a regulated cell death nexus linking metabolism, redox biology, and disease. Cell, 171, 273–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- SUN T, ZHANG YS, PANG B, HYUN DC, YANG M & XIA Y 2014. Engineered nanoparticles for drug delivery in cancer therapy. Angewandte Chemie International Edition, 53, 12320–12364. [DOI] [PubMed] [Google Scholar]
- UCHIDA K 2017. HNE as an inducer of COX-2. Free Radical Biology and Medicine, 111, 169–172. [DOI] [PubMed] [Google Scholar]
- VALENCIA PM, PRIDGEN EM, RHEE M, LANGER R, FAROKHZAD OC & KARNIK R 2013. Microfluidic platform for combinatorial synthesis and optimization of targeted nanoparticles for cancer therapy. ACS Nano, 7, 10671–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- YAMADA S, KUMAZAWA S, ISHII T, NAKAYAMA T, ITAKURA K, SHIBATA N, KOBAYASHI M, SAKAI K, OSAWA T & UCHIDA K 2001. Immunochemical detection of a lipofuscin-like fluorophore derived from malondialdehyde and lysine. Journal of Lipid Research, 42, 1187–1196. [PubMed] [Google Scholar]
- YAMAGUCHI H, HSU JL, CHEN C-T, WANG Y-N, HSU M-C, CHANG S-S, DU Y, KO H-W, HERBST R & HUNG M-C 2013. Caspase-Independent Cell Death Is Involved in the Negative Effect of EGF Receptor Inhibitors on Cisplatin in Non–Small Cell Lung Cancer Cells. Clinical cancer research. [DOI] [PMC free article] [PubMed] [Google Scholar]
- YANG WS, KIM KJ, GASCHLER MM, PATEL M, SHCHEPINOV MS & STOCKWELL BR 2016. Peroxidation of polyunsaturated fatty acids by lipoxygenases drives ferroptosis. Proceedings of the National Academy of Sciences, 113, E4966–E4975. [DOI] [PMC free article] [PubMed] [Google Scholar]
- YANG WS, SRIRAMARATNAM R, WELSCH ME, SHIMADA K, SKOUTA R, VISWANATHAN VS, CHEAH JH, CLEMONS PA, SHAMJI AF & CLISH CB 2014. Regulation of ferroptotic cancer cell death by GPX4. Cell, 156, 317–331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- YU HT, GUO PY, XIE XZ, WANG Y & CHEN G 2017. Ferroptosis, a new form of cell death, and its relationships with tumourous diseases. Journal of Cellular and Molecular Medicine, 21, 648–657. [DOI] [PMC free article] [PubMed] [Google Scholar]
- YU Y, XIE Y, CAO L, YANG L, YANG M, LOTZE MT, ZEH HJ, KANG R & TANG D 2015. The ferroptosis inducer erastin enhances sensitivity of acute myeloid leukemia cells to chemotherapeutic agents. Molecular & cellular oncology, 2, e1054549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- YUE Z, YOU Z, YANG Q, LV P, YUE H, WANG B, NI D, SU Z, WEI W & MA G 2013. Molecular structure matters: PEG-b-PLA nanoparticles with hydrophilicity and deformability demonstrate their advantages for high-performance delivery of anti-cancer drugs. Journal of Materials Chemistry B, 1, 3239–3247. [DOI] [PubMed] [Google Scholar]
- ZILKA O, SHAH R, LI B, FRIEDMANN ANGELI JP, GRIESSER M, CONRAD M & PRATT DA 2017. On the Mechanism of Cytoprotection by Ferrostatin-1 and Liproxstatin-1 and the Role of Lipid Peroxidation in Ferroptotic Cell Death. ACS Cent Sci, 3, 232–243. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Please follow the Key Resources Table for software downloading. Further information and requests for lipidomics data should be directed to and will be fulfilled by the Lead Contact Brent R. Stockwell ([email protected]).
KEY RESOURCES TABLE.
| REAGENT or RESOURCE | SOURCE | IDENTIFIER |
|---|---|---|
| Antibodies | ||
| Anti-dihydropyridine-MDA-lysine adduct mouse mAb 1F83 | Reference (Yamada et al., 2001) | N/A |
| Anti-8-OH-dG antibody | Abcam | ab62623 |
| Anti-xCT antibody | Abcam | ab37185 |
| Anti-cyclooxygenase 2 (COX-2) antibody | Abcam | ab15191 |
| Goat anti-mouse IgG H&L (Alexa Fluor 647) | Abcam | ab150115 |
| Goat anti-rabbit IgG H&L highly cross-adsorbed secondary antibody (Alexa Fluor488) | Thermo Fisher Scientific | A11034 |
| Donkey anti-Goat IgG (H+L) cross-adsorbed, Alexa Fluor 647, polyclonal, secondary antibody | Thermo Fisher Scientific | A21447 |
| Anti-15-F2T-isoprostane purified IgG | Oxford Biomedical Research | IS20 |
| Anti-α-tubulin antibody (DM1A) | Santa Cruz Biotechnology | sc-32293 |
| Anti-xCT/SLC7A11 (D2M7A) antibody | Cell Signaling Technology | 12691 |
| Anti-caspase-3 antibody | Cell Signaling Technology | 9662 |
| Chemicals | ||
| (Poly(ethylene glycol) methyl ether-block-poly(lactide-co-glycolide)(PEG-PLGA) | Sigma-Aldrich | 900948 |
| β-mercaptoethanol | Sigma-Aldrich | M3148 |
| Imidazole ketone erastin (IKE) | Reference(Larraufie et al., 2015) | N/A |
| Ferrostatin-1 (fer-1) | Reference(Skouta et al., 2014) | N/A |
| Critical Commercial Assays | ||
| GSH/GSSG Ratio Detection Assay Kit | Abcam | ab13881 |
| TBARS Assay Kit | Cayman | 700870 |
| BODIPY™ 581/591 C11 | Thermo Fisher Scientific | D3861 |
| RiboPure™ RNA Purification Kit | Thermo Fisher Scientific | AM1924 |
| CellTiter-Glo Luminescent Cell Viability Assay | Promega | G7573 |
| RNAeasy extraction kit | QIAGEN | 74106 |
| QIAshredder | QIAGEN | 79656 |
| High Capacity cDNA Reverse Transcription Kit | Thermo Fisher Scientific | 4368814 |
| Power SYBR Green PCR Master Mix | Thermo Fisher Scientific | 4368702 |
| Cell culture reagents and others | ||
| ProLong™ Gold anti-fade Mount with DAPI | Thermo Fisher Scientific | P36962 |
| 10% goat serum | Thermo Fisher Scientific | 50197Z |
| Donkey serum | Sigma-Alderich | D9663 |
| Poly-lysine | Sigma-Aldrich | P4832 |
| Fetal bovine serum | Thermo Fisher Scientific | 10437–028 |
| DMEM | Corning | 10–013 |
| Fetal bovine serum | Life Technologies | 10437036 |
| RPMI-1640 | ATCC | 30–2001 |
| Non-essential amino acids | Thermo Fisher Scientific | 11140076 |
| Penicillin-streptomycin mix | Thermo Fisher Scientific | 15140148 |
| IMDM | Thermo Fisher Scientific | 12440053 |
| HBSS | Thermo Fisher Scientific | 14025076 |
| Steriflip filter unit | Thomas Scientific | 1189Q46 |
| Amicon Ultra-15 centrifugal filter units | Millipore | UF9050 |
| K3 EDTA micro tube | SARSTEDT | 41.1504.105 |
| RIPA buffer | Thomas Scientific | 89900 |
| Software and Algorithms | ||
| ChemDraw Ultra, Version 14.0 | Perkin Elmer | http://www.perkinelmer.com/category/chemdraw |
| Prism, Version 7.0 | GraphPad Software | https://www.graphpad.com/scientific-software/prism/ |
| LipoStar, Version 1.0.4 | Molecular Discovery | https://www.moldiscovery.com/software/lipostar/ |
| MetaboAnalyst, Version 4.0 | MetaboAnalyst | https://www.metaboanalyst.ca/faces/docs/About.xhtml |
| MassLynx, Version 4.1 | Waters | http://www.waters.com/waters/en_US/MassLynx-Mass-Spectrometry-Software-/nav.htm?cid=513164&locale=en_US |
| XCMS package, Version 3.2.0 | Bioconductor | http://packages.renjin.org/package/org.renjin.bioconductor/xcms |