Vitamin E hydroquinone is an endogenous regulator of ferroptosis via redox control of 15-lipoxygenase

RESEARCH ARTICLE

Vitamin E hydroquinone is an endogenous

regulator of ferroptosis via redox control of

15-lipoxygenase

Andrew Hinman^‡☯, Charles R. Holst^‡☯, Joey C. Latham, Joel J. Bruegger, Gozde Ulas,

¨

Kevin P. McCusker, Akiko Amagata, Dana Davis, Kevin G. Hoff, Amanda H. Kahn-Kirby,

Virna Kim, Yuko Kosaka, Edgar Lee, Stephanie A. Malone, Janet J. Mei, Steve

James Richards, Veronica Rivera, Guy Miller, Jeffrey K. Trimmer, William D. Shrader*

a 1111111111

BioElectron Technology Corporation, Inc., Mountain View, California, United States of America

☯ These authors contributed equally to this work.

‡ AH and CH are joint senior authors on this work.

* bshrader@bioelectron.com

Abstract

Ferroptosis is a form of programmed cell death associated with inflammation, neurodegen-

eration, and ischemia. Vitamin E (alpha-tocopherol) has been reported to prevent ferropto-

sis, but the mechanism by which this occurs is controversial. To elucidate the biochemical

mechanism of vitamin E activity, we systematically investigated the effects of its major vita-

mers and metabolites on lipid oxidation and ferroptosis in a striatal cell model. We found that

a specific endogenous metabolite of vitamin E, alpha-tocopherol hydroquinone, was a dra-

matically more potent inhibitor of ferroptosis than its parent compound, and inhibits 15-lipox-

ygenase via reduction of the enzyme’s non-heme iron from its active Fe³⁺state to an

inactive Fe²⁺state. Furthermore, a non-metabolizable isosteric analog of vitamin E which

retains antioxidant activity neither inhibited 15-lipoxygenase nor prevented ferroptosis.

These results call into question the prevailing model that vitamin E acts predominantly as a

non-specific lipophilic antioxidant. We propose that, similar to the other lipophilic vitamins A,

D and K, vitamin E is instead a pro-vitamin, with its quinone/hydroquinone metabolites

responsible for its anti-ferroptotic cytoprotective activity.

OPEN ACCESS

Citation: Hinman A, Holst CR, Latham JC,

Bruegger JJ, Ulas G, McCusker KP, et al. (2018)

Vitamin E hydroquinone is an endogenous

regulator of ferroptosis via redox control of 15-

lipoxygenase. PLoS ONE 13(8): e0201369. https://

doi.org/10.1371/journal.pone.0201369

Editor: Rui Kang, University of Pittsburgh, UNITED

STATES

Received: May 15, 2018

Accepted: July 14, 2018

Published: August 15, 2018

access article distributed under the terms of the

Creative Commons Attribution License, which

permits unrestricted use, distribution, and

reproduction in any medium, provided the original

author and source are credited.

Data Availability Statement: All relevant data are

within the paper and its Supporting Information

files.

Introduction

Ferroptosis is a form of iron-dependent, lipid oxidation-mediated programmed cell death

implicated in a wide array of disease pathologies including inflammation, ischemia-reperfu-

sion injury, and neurodegeneration [1–5]. Vitamin E modulates ferroptosis via regulation of

the oxidoreductase enzyme 15-lipoxygenase (15-LO) through an unknown mechanism [6–8].

The ascribed clinical benefits of vitamin E include regulation of oxidative stress, inflammation,

and cell death, each of which results from ferroptosis [1]. To elucidate the biochemical mecha-

nism of vitamin E activity, we systematically investigated its major vitamers and metabolites,

their effect on 15-LO, and the regulation of ferroptosis. We report that an endogenous

Funding: This work was funded by BioElectron

Technology Corporation, Inc. The funding entities

had a role in study design, data collection and

analysis, decision to publish, or preparation of the

manuscript.

Competing interests: The authors have read the

journal’s policy and the authors of this manuscript

have the following competing interests: All authors

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

1 / 22

Redox regulation of ferroptosis

are employees of, and/or hold equity in BioElectron

Technology Corporation, Inc. This does not alter

the authors’ adherence to PLOS ONE policies on

sharing data and materials.

metabolite of vitamin E, alpha-tocopherol hydroquinone, is a potent inhibitor of ferroptosis

through its reduction of the non-heme iron in 15-lipoxygenase from its active Fe³⁺state to its

inactive Fe²⁺state. We propose that the anti-ferroptotic effect of vitamin E results from a four-

step biochemical mechanism: 1) oxidative hydrolysis of alpha-tocopherol to alpha-tocopherol

quinone; 2) reduction of alpha-tocopherol quinone to alpha-tocopherol hydroquinone; 3)

inhibition of 15-LO via reduction of its non-heme Fe³⁺center to the inactive Fe²⁺state by

alpha-tocopherol hydroquinone; and 4) inhibition of the ferroptosis cascade by blocking for-

mation of lipid peroxidation products. These results support a reinterpretation of the mecha-

nism of vitamin E, which has previously been attributed to a non-specific antioxidant activity.

As with the other fat-soluble vitamins (A, D, and K), our data suggest that vitamin E is a pro-

vitamin that is converted to an active form, vitamin E hydroquinone, which functions at a spe-

cific set of targets. These findings link vitamin E’s purported clinical effects to an anti-ferropto-

tic mechanism-of-action.

Materials and methods

General method for the chemical conversion of a vitamin E chroman to a

vitamin E quinone

To a round-bottomed flask, the chroman (1 eq, approximately 0.5 g) was dissolved in a 2:1 ace-

tonitrile:water mixture (15 mL). This solution was cooled to 0˚C. A solution of ceric ammo-

nium nitrate (2.1 eq) in water (5 mL) was added dropwise. The reaction mixture was stirred at

0˚C for 2 h. Upon reaction completion, the mixture was neutralized with sat. aq. NaHCO₃and

extracted 3x with ethyl acetate. The organic phase was dried over Na₂SO₄and concentrated in

vacuo. The residue was purified by silica gel chromatography to obtain pure product.

General method for the chemical conversion of a vitamin E quinone to a

vitamin E hydroquinone

To a 20 mL septum-capped vial, the vitamin E quinone (1 eq, approximately 0.5 g) was dis-

solved in tetrahydrofuran (THF) (15 mL). Then the vessel was purged with H₂. Lindlar catalyst

(for the α-, γ- and δ-tocotrienol quinone) or 10% Pd/C (for the α-, γ- and δ-tocopherol qui-

none) (0.1 eq by weight) was added and the vessel was purged again with H₂. The resulting sus-

pension was stirred at room temperature overnight under a H₂atmosphere. Upon reaction

completion, celite was added. The resulting slurry was filtered through a fritted funnel. The fil-

ter cake was washed successively with THF. The filtrate was concentrated in vacuo to obtain

pure product.

Cells and cell culture

Mouse striatal Q7 cells (STHdh^Q7/Q7; immortalized by SV large T antigen) were obtained from

Coriell. Cells were cultured in high-glucose (25 mM) DMEM (Gibco) medium supplemented

with fetal bovine serum (10% v/v; Sigma), penicillin (100 U/mL; Gibco), and streptomycin

(100 μg/mL; Gibco). For routine maintenance, Q7 culture medium also contained Geneticin

Sulfate (G418; Santa Cruz Biotechnology). Cells were routinely passaged every 3–4 days, main-

taining sub-confluent cell densities.

Cellular survival assay. Ferroptotic cell death was induced in Q7 cells using RSL3, a

GPX4 inhibitor previously described [9]. All four stereoisomers of RSL3 were synthesized and

purified using published methods. All subsequent cell treatment experiments utilized the sin-

gle stereoisomer, (1S,3R)-RSL3 that most potently (IC₅₀~200 nM) inhibited Q7 cell viability.

To assess compound rescue activity and potency from ferroptotic cell death, Q7 cells were

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

2 / 22

Redox regulation of ferroptosis

seeded in clear bottom, black wall tissue culture-treated polystyrene plates (Corning) using

either an electronic multichannel pipette or a Multidrop™ Combi Reagent Dispenser (Thermo-

Fisher Scientific). Cells were incubated for 5 hours at 33˚C in an atmosphere with 95% humid-

ity and 5% CO₂to allow attachment of the cells to the culture plate. Test compounds were

administered to the desired final concentrations using a D300e Digital Dispenser (Tecan), fol-

lowed within 15 minutes by RSL3 (2 μM final concentration). DMSO diluent was back-filled

to a final concentration of 0.3% (v/v). Cell plates were incubated at 33˚C (5% CO₂, 95%

humidity) for 18 hours. After a 15-minute equilibration to room temperature, cell number/via-

bility was assessed using the CellTiter-Glo¹2.0 Cell Titer Glo 2.0 reagent (a luminescence-

based assay to quantitate ATP; Promega) which was added using the Multidrop™ Combi

Reagent Dispenser (Thermo Fisher Scientific). After 15 minutes of incubation at room temper-

ature, the luminescence (100 ms integration time) per well was determined using the BioTek

Synergy plate reader. Data was imported into and analyzed using Dotmatics Studies software

suite. EC₅₀values for each compound were estimated using standard four-parameter curve fit-

ting algorithms. pEC₅₀is defined as negative-log₁₀(EC₅₀) in M units. RSL3 treatment typically

reduced cell number/viability by >80%. The assay performance was assessed by Z-prime cal-

culations on each assay plate, with observed Z-prime values of >0.7. We also note that maxi-

mally efficacious concentrations of the vitamin E compounds restored RSL3-treated cell

viability to levels comparable to cells that had not been treated with RSL3, indicating complete

rescue of cell viability by these compounds.

Cellular lipid oxidation assay

Ferroptosis was induced by RSL3 (2 μM final concentration) in Q7 cells as described above.

The protective activity of co-treated compounds was assessed also as described above. The rate

of cellular lipid oxidation after GPX4 inhibition was assessed by monitoring the time-depen-

dent changes in green fluorescence of cells pre-labeled with BODIPY™ 581/591 C11 dye (5 μM,

30-minute labeling period; Thermo Fisher/Invitrogen) using the IncuCyte S3 live-cell imaging

apparatus (Sartorius) using a 10x objective. Cells (three images per field; at least three replicate

wells per condition) were imaged once hourly for up to 18 hours after compound and RSL3

addition. Quantification of cell images was performed using the IncuCyte Zoom software

using the Basic Analyzer tool which utilized an algorithm that calculated the total green-posi-

tive area per well where “green-positive” was defined as 1.45 times the local background-sub-

tracted fluorescent signal. After a 2-3-hour lag period post-RSL3 treatment, the cellular green

fluorescence signal began increasing. The rate of change in total green-positive area per well

from 3–9 hours post-compound treatment was calculated, then expressed relative to that

observed in the RSL3-only control treatment group, which was defined as 100%.

In vitro intra-cellular quantification of alpha-tocopherol (αT), alpha-

tocopherol quinone (αTQ) and alpha-tocopherol hydroquinone (αTHQ)

concentrations

Q7 cells were cultured in standard culture medium for 18 hours prior to compound addition.

After the specified treatment time, the culture medium was removed and adherent cells were

washed with PBS. The wash was removed and adherent cells were harvested by adding 1 mL of

100 mg/mL succinic anhydride in 95:5 (acetonitrile:triethylamine) followed by immediate cell

scraping. After cell scraping, samples were transferred to a 1.5 mL Eppendorf tube, vortexed

for 15 seconds, and kept on wet ice until all samples were harvested. Samples were vortexed

again for 15 seconds and centrifuged in a Thermo Sorvall Legend XFR centrifuge at 12,000

rpm, 4˚C, for 3 minutes.

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

3 / 22

Redox regulation of ferroptosis

Following centrifugation, samples were protein-precipitated by addition of 3 volumes of

acetonitrile. Following precipitation, samples were vortexed for 30 seconds and centrifuged in

a Thermo Sorvall Legend XFR centrifuge at 4000 rpm, 4˚C, for 10 minutes. Subsequently, 1

volume of supernatant was transferred to a 384-well injection plate and diluted with one vol-

ume of mobile phase A and pipette mixed. Internal standard was added to the 384-well plate

and pipette mixed. The injection plate was vortexed and centrifuged for a final time at 4000

rpm, 4˚C, for 10 minutes and subsequently placed in a chilled Shimadzu SIL-30AC autosam-

pler at 4˚C for LC-MS/MS analysis.

Samples were analyzed using a developed method on a Shimadzu 30AD using micro-flow

with a Shimadzu SIL-30AC autosampler, and coupled to an AB Sciex 6500+ QTRAP mass

spectrometer. The MS/MS instrument was operated in positive ESI mode. Electrospray condi-

tions for the μLC-ESI-MS/MS method were optimized and were as follows: Ion spray voltage

was set to 5500V, temperature of 250˚C, curtain gas of 10, CAD gas of 10, and ion source gas 1

and 2 of 15 and 60 psi respectively. Analyzer parameters were optimized for each compound

using a combination of manual tuning and compound optimization. The following MRM

transitions were used to quantify concentrations of αT, αTQ and αTHQ. αT: 430.257!165.1;

αT-succinate: 531.368!265.3; αTQ: 429.253!165.1; αTHQ: 431.253!165.1; αTHQ-mono

succinate: 572.208!554.1; αTHQ-bis succinate: 666.458!531.4. For concentrations reported:

αT = [αT] + [αT-succinate]; αTQ = αTQ; αTHQ = [αTHQ] + [αTHQ-mono succinate] +

[αTHQ-bis succinate]. Note: αTHQ and αTHQ-mono succinate were not observed in this

experiment.

A Phenomenex Kinetex C₁₈(2.1 x 50 mm, 1.3 μm) column using a reverse phase gradient

mobile phase method was used for chromatographic separation. The composition of mobile

phase A was 0.1% formic acid in water. The composition of mobile phase B was 3:1 (0.1% for-

mic acid in (acetonitrile: isopropyl alcohol)). Mobile Phase B was ramped from 30% to 98%

over 7 minutes. The total run time was 8 minutes. Data were analyzed using Sciex Analyst

Chromatography software, version 1.6.2 and Sciex MultiQuant software. The standard curve

equation (y = mx +b) is generated from the calibration standards with weighted regression.

Our recent PCT Publication No. WO 2018/081644 also describes the hydroquinone succinate

capping methodology in more detail. Intracellular concentrations were estimated assuming

spherical cells with an average diameter of 18 μm (estimated using Coulter impedance mea-

surement), with input of 800,000 cells per 1 mL extraction volume.

d₄-α-Tocopherol (d₄-αT) in vivo biodistribution and quinone conversion

study

This study was approved by the MPI Research IACUC Committee. Animal treatment was in

compliance with the U.S. Department of Agriculture’s (USDA) Animal Welfare Act (9 CFR

Parts 1, 2 and 3) and performed in an AAALAC accredited facility (File #000243) following

The Guide for the Care and Use of Laboratory Animals, Institute of Laboratory Animal

Resources, National Academy Press, Washington, D.C. The facility maintains an Animal Wel-

fare Assurance statement with the National Institutes of Health Office of Laboratory Animal

Welfare.

Fed male CD¹Sprague-Dawley rats (aged ~6 weeks, average animal weight 183 g; n = 4)

were dosed once daily for four days by oral gavage to achieve 100 mg/kg dose of d₄-α-tocoph-

erol formulated in sesame oil. Four hours after final dose, 0.5 mL of whole blood was collected

into Lithium Heparin tubes via the sublingual vein. Blood was centrifuged within 15 minutes

of collection and clear supernatant was transferred to 0.2 mL cryotubes. Plasma samples were

rapidly frozen and then stored at -80˚C until analysis. Euthanasia was performed by carbon

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

4 / 22

Redox regulation of ferroptosis

dioxide inhalation, after which the tissues were collected and weighed, then chilled PBS was

added at a ratio of 3 mL per gram of tissue for homogenization. The homogenate was frozen

on dry ice and stored at -80˚C until further analysis.

Quantitation of d₄-α-tocopherol and d₄-alpha-tocopheryl quinone in plasma and tissues

was performed using an Eksigent microLC 200 coupled to a Sciex 6500 Qtrap. Briefly, samples

were thawed at 4˚C then vortexed. Both plasma and tissue homogenates were protein-precipi-

tated by the addition of ice cold acetonitrile in a 3:1 ratio (ACN:sample). Following precipita-

tion, samples were vortexed for 30 seconds and centrifuged at 4000 rpm for 10 minutes at 4˚C.

Fifty microliters of supernatant was added to a PCR injection plate containing 50 μL of a 0.1%

Formic Acid solution then sealed for mass spectrometry analysis.

d₄-α-Tocopherol and d₄-α-tocopheryl quinone were baseline resolved on a Waters

BEH-C₁₈column (1.0 x 50 mm, 1.7 μm). The composition of mobile phase A was 0.1% formic

acid in water. The composition of mobile phase B was 0.1% formic acid in 3:1 acetonitrile: iso-

propyl alcohol. Gradient elution was employed, with mobile phase B being ramped from 20%

at 0 to minutes, to 98% at 2 minutes and held for 2.5 minutes. The total run time was 5 min-

utes. Electrospray conditions for the μLC-ESI-MS/MS method were optimized for a flow rate

of 65 μL/min and were as follows: Ion spray voltage was set to 5500 V, temperature of 150˚C,

curtain gas of 20, CAD gas of 10, and ion source gas 1 and 2 of 20 and 60 psi, respectively. Q1

and Q3 transitions for each analyte and voltage settings (e.g. collision energy, declustering

potential) were optimized for each compound using a combination of manual tuning and

compound optimization.

Concentrations of d₄-α-tocopherol and d₄-α-tocopheryl quinone in tissue homogenates

were determined with μg analyte / g input tissue units, then converted to μM units using the

appropriate analyte molecular weights and assuming the following tissue densities (liver 1.06

g/mL, heart 1.045 g/mL, small intestine 1.05 g/mL–refer to Appendix A of Basics of Biomedical

Ultrasound for Engineers, H. Azhari, 2010). Quinone conversion percentages were calculated

using the formula Q / (Q + C) ^Ã100 where Q = quinone concentration (in μM) and C = parent

chroman concentration (in μM).

ORAC assay

ORAC assay kits were purchased from Cell Biolabs. Assays were performed per the manufac-

turer’s instructions for lipophilic samples with minor modifications. DMSO stock solutions of

compounds were diluted in an anaerobic glovebox to final concentrations of 25 and 12.5 μM

in triplicate. A standard curve of Trolox¹between 0–100 μM in triplicate per concentration

was also generated on each assay plate. Diluted compounds and standards were removed from

the glovebox and reactions initiated by addition of aerobic fluorescein reagent and radical ini-

tiator. Fluorescein fluorescence was monitored at excitation wavelength of 480 nm and emis-

sion wavelength of 520 nm every minute for one hour. AUC analysis was performed as

described in the kit protocol. The “Trolox equivalents” at each concentration of test compound

were calculated and the average Trolox equivalents at the two compound concentrations are

presented. The errors were determined by propagating errors from the SEMs of the triplicate

determinations to the averages across the two triplicate determinations.

Radical-trapping antioxidant (RTA) activity assays

Liposomes containing 2 mol% of αT, αTCC, αTQ or αTHQ in DOPC (18:1 1,2-dioleoyl-sn-

glycero-3-phosphocholine) were prepared by thin-film hydration followed by extrusion to

obtain large unilamellar vesicles at 120 nm in diameter with a polydispersity index (PDI) < 0.1

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

5 / 22

Redox regulation of ferroptosis

in DPBS buffer (pH 7.5). PDI measurements were determined using a Malvern Zetasizer

Nano-ZSP.

RTA assays were prepared by diluting liposomes (to 2.5 mM [lipid]) in DPBS buffer and incu-

bating at room temperature with 500 nM BODIPY™ 581/591 C11 dye for 15 minutes. A fresh

stock of 20 mM AMVN (2,2’-azobis(4-methoxy-2,4-dimethylvaleronitrile), Wako Chemicals

USA, Inc.) was prepared in MeOH and added to the liposome mix to a final concentration of 2

mM immediately before data collection. The decrease in the reduced dye’s fluorescence emission

was measured every 30 seconds at 37˚C, using excitation/emission of 575/600 nm with a 590nm

cutoff filter, on a Molecular Devices Spectramax M2. The data plots were normalized to 100% fluo-

rescence using the initial time points. The AUC analysis was performed using GraphPad Prism 7.

Enzyme assay materials

Rabbit 15-lipoxygenase (rLOX15, part number BML-PL015, lot number 3-P9009a) was

obtained from Enzo Life Sciences, Inc. Sodium cholate (part number C6445, lot number

094K0209) was obtained from Sigma-Aldrich. Dihydrorhodamine (part number 10055, lot

number 12D0215) was obtained from Biotium. Arachidonic acid (part number 90010, lot

number 0509246–4) and soybean 15-lipoxygenase (SLO, part number 60700, lot number

0500889–1) were obtained from Cayman Chemical. The 96 half-well black, flat bottom assay

plates (part number 3993, lot number 19316045) were obtained from Corning, Inc.

Lipoxygenase enzyme inhibition assay

Enzyme assays (100 μL final volume) were conducted in Corning 96 half-well black, flat bot-

tom assay plates and contained final concentrations of 50 μM arachidonic acid as a substrate,

1:10 (v:v) cholate mix (2% (w:v) sodium cholate and 2% (v:v) DMSO), 40 μM dihydrorhoda-

mine 123 and 3 U/mL rLOX15 in buffer A (100 mM Tris-HCl, pH 7.5). Stock solutions of test

compounds were prepared in DMSO and then diluted to final assay concentrations in cholate

mix such that final cholate and DMSO concentrations were maintained at 0.2%. Reactions

were initiated by addition of 50 μL of enzyme mix (80 μM dihydrorhodamine 123 and 6 U/mL

rLOX15 in buffer A) to wells containing 50 μL substrate and cholate mix in Buffer A and linear

rates were assessed via fluorescence, using excitation/emission of 500/536 nm, on a Molecular

Devices Spectramax M2 every 10 seconds for 5 minutes at room temperature.

EPR sample preparation

Lipoxygenase: Soybean 15-lipoxygenase (SLO) was buffer exchanged into 50 mM borate

buffer, pH 9.0 containing 10% (v/v) glycerol. The resulting protein solution was concentrated

to approximately 100 μM, based upon the supplier’s activity per mL and specific activity per

mg. The active-site iron oxidation state was further evaluated by EPR under three treatment

conditions: 1) No treatment, 2) oxidation with 3-fold molar excess of lipid hydroperoxide

(final concentration of 300 μM) from an ethanol stock, 3) oxidation with 3-fold molar excess

of lipid hydroperoxide followed by 10-fold molar excess of αTHQ or αT3HQ (final concentra-

tion of 1 mM) from a DMSO stock. All samples contained the same final concentrations of

ethanol and DMSO as sample 3. The EPR spectra of buffer alone were also evaluated to allow

for baseline subtraction.

Electron Paramagnetic Resonance (EPR) spectroscopy

Lipoxygenase. CW-EPR spectra were recorded at 150 K on a Bruker ELEXYS E540 spec-

trometer fitted with an X-band microwave bridge, a Super High Sensitivity Resonator (SHQ),

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

6 / 22

Redox regulation of ferroptosis

and a Nitrogen VT apparatus. The spectra were recorded at 9.43 GHz with 1.002 mW micro-

wave power, 10 G modulation amplitude, 100 kHz modulation frequency, 25 scans.

αTCC. CW-EPR spectra were recorded at room temperature in ambient air on a Bruker

EMXnano spectrometer fitted with an X-band microwave bridge. All spectra were recorded at

9.61 GHz with 10 mW microwave power and 5 G modulation amplitude, 100 kHz modulation

frequency, 25 scans.

Targeted arachidonate metabolite release cellular assay

Ferroptosis was initiated by RSL3 (2 μM final concentration) in Q7 cells as described above.

The protective activity of co-treated compounds was assessed also as described above. Four to

five hours after RSL3 treatment, the conditioned medium was collected into 1.2-mL cluster

tubes containing ice-cold acetonitrile at a ratio of 1:2. Tubes were snap frozen on dry ice and

stored at -80˚C.

Prior to analysis, samples were thawed on wet ice and then vortexed briefly. Samples were

centrifuged at 4000 rpm for 15min at 4˚C. Supernatant (40 μL) was transferred to a 96-well

injection plate and 10 μL of both internal standard and HPLC grade water were added. A Shi-

madzu Nexera MP system coupled to a Sciex 6500 Qtrap and configured for micro-flow was

used to analyze all samples. Samples were maintained at 4˚C in the Nexera MP autosampler

for the duration of the analytical run.

A Thermo Gold-C₈column (1.9 μm, 1.0 x 100 mm) operated at 45˚C was used for chro-

matographic separation. The total run time of the liquid chromatography method was 10 min.

Mobile phase A consisted of water with 0.1% formic acid. Mobile phase B was composed of

0.1% formic acid in 3:1 acetonitrile:isopropyl alcohol. Gradient elution was employed, with

mobile phase B being ramped from 5% at 0min, to 50% at 2 min, and finally 98% at 7.5 min.

Electrospray conditions for the μLC-ESI-MS/MS method were optimized for a flow rate of

125 μL/min and were as follows: Ion spray voltage was set to -4500 V, temperature of 250˚C,

curtain gas of 20, CAD gas of 10, and ion source gas 1 and 2 of 20 and 40 psi, respectively. Ana-

lyzer parameters were optimized for each compound using a combination of manual tuning

and compound optimization. The following MRM transitions were used to quantify concen-

trations of 15-HETE, 12-HETE and Arachidonic Acid (AA). 15-HETE: 319!219; 12-HETE:

319!179; AA: 303!259.

Statistical methods

All statistical analyses were performed using GraphPad Prism v.7.03 for Windows. If all groups

showed apparently normal distributions (by Shapiro-Wilk test), then one-way ANOVA tests

were performed, with Sidak’s correction for multiple comparisons applied when appropriate.

If even one group did not meet normality assumptions, then a non-parametric Kruskal-Wallis

test was used, with Dunn’s multiple comparisons test applied when appropriate. When only

comparing two groups, then an unpaired t-test with Welch’s correction was used. Statistical

significance throughout was defined as p<0.05.

Results

Previously, we reported that quinone metabolites of vitamin E have potent rescue activity in a

glutathione (GSH) depletion cell survival assay which exhibited features of ferroptotic cell

death [10–14]. To test the hypothesis that the quinone and/or hydroquinone metabolites of

alpha-tocopherol (αT) are responsible for its previously reported anti-ferroptotic activity

[6,15–19], we evaluated the activity of alpha-tocopherol quinone (αTQ) and hydroquinone

(αTHQ) in a ferroptotic cell death assay using immortalized mouse striatal cells (STHdh^Q7/Q7

,

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

7 / 22

Redox regulation of ferroptosis

RSL3 Ferroptosis

Cell Assay

15-Lipoxygenase

Enzyme Assay

ORAC Antioxidant Radical Trapping

Activity Assay

Assay

Average

EC₅₀(μM)

% activity

at 100 μM

SD

(μM)

Trolox

Equiv.

SEM

AUC

R₃

R₁

R₂

IC₅₀(μM)

N

SE

Abbreviation (Compound)

(Rel. Control)

Vitamin E Vitamers

R₁

O

R₃

HO

R₂

αT (α-tocopherol)

αT3 (α-tocotrienol)

γT (γ-tocopherol)

γT3 (γ-tocotrienol)

δT (δ-tocopherol)

δT3 (δ-tocotrienol)

α-CEHC

CH₃CH₃Phytyl

CH₃CH₃Triene

5.161

0.167

3.217

0.180

5.670

0.517

8.509

2.844

0.040

0.509

0.011

3.471

0.148

0.624

9

13

3

>100

78%

64%

34%

33%

45%

46%

24%

0.073

0.008

5.123

0.009

CH₃

H

Phytyl

Triene

Phytyl

Triene

-

H

4

3

CH₃CH₃CEHC

Vitamin E quinone metabolites

O

OH

R₃

R₁

R₂

O

αTQ (α-tocopherol quinone)

αT3Q (α-tocotrienol quinone)

γTQ (γ-tocopherol quinone)

γT3Q (γ-tocotrienol quinone)

δTQ (δ-tocopherol quinone)

CH₃CH₃Phytyl

CH₃CH₃Triene

0.233

0.043

0.101

0.052

0.236

0.095

0.015

0.028

0.002

0.061

13

53

6

>100

-

75%

62%

note¹

0.044

0.005

1.217

0.001

CH₃

H

Phytyl

Triene

Phytyl

Triene

3

6

H

0.163

1.248

0.011

0.144

3

4

-

note¹

82%

δT3Q (δ-tocotrienol quinone)

α-CHEC quinone (lactone)

CH₃CH₃CEHC

Vitamin E hydroquinone metabolites

OH

R₃

R₁

R₂

OH

αTHQ (α-tocopherol hydroquinone)

αT3HQ (α-tocotrienol hydroquinone)

γTHQ (γ-tocopherol hydroquinone)

γT3HQ (γ-tocotrienol hydroquinone)

δTHQ (δ-tocopherol hydroquinone)

δT3HQ (δ-tocotrienol hydroquinone)

CH₃CH₃Phytyl

CH₃CH₃Triene

0.090

0.043

0.064

0.036

0.084

0.146

0.032

0.009

0.013

0.006

0.019

0.039

9

3

4

2.45

1.99

2.19

1.12

2.09

1.37

15%

9%

18%

7%

17%

14%

0.106

0.011

1.855

0.002

CH₃

H

Phytyl

Triene

Phytyl

Triene

H

Metabolically stable -tocopherol analog

HO

-

αTCC (α-tocopherol carbochroman)

>30

6

>100

77%

0.075

0.008

4.706

0.014

Phytyl =

R₃=

Triene =

CEHC =

CO₂H

¹Unable to determine values due to non-enzyme-mediated assay signal from these compounds

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

8 / 22

Redox regulation of ferroptosis

Fig 1. Ferroptosis cell rescue activity, 15-lipoxygenase enzyme assay data, ORAC and RTA values for 18 vitamin E vitamers and metabolites. Summary of

cellular and enzyme assay potencies. Q7 ferroptosis cell rescue, 15-lipoxygenase (15-LO) enzyme inhibition, oxygen radical absorbance capacity (ORAC) values and

the radical-trapping antioxidant (RTA) activity assays were measured for αT, its selected metabolites, and the non-metabolizable αTCC.

https://doi.org/10.1371/journal.pone.0201369.g001

hereafter referred to as Q7) [20]. These specific compounds were part of a larger screen which

included 6 vitamin E vitamers and 14 known vitamin E metabolites (Fig 1) [21–23].]

A

α

T

100

75

50

25

0

TQ

THQ

TCC

0.01

0.1

1

10

[compound] (μM)

B

100

75

50

25

0

α

T

TQ

THQ

TCC

0.01

0.1

1

10

[compound] (μM)

Fig 2. The anti-ferroptotic activity of vitamin E and its metabolites. (A) Ferroptosis was induced in Q7 cells by RSL3 treatment

(2 μM 18 h), as measured by CellTiter-Glo¹2.0 ATP assay. The dose-dependent cytoprotective activity of vitamin E (αT), its

metabolites (αTQ, αTHQ), or the synthetic carbochroman (αTCC) was assessed by co-treatment (in 24-point titration) with RSL3.

Curves are representative of ꢀ9 independent experiments, of which 7 evaluated αTCC to a top concentration of 30 μM. (B)

Ferroptosis was initiated in BODIPY^TM581/591 C11-prelabeled Q7 cells by RSL3 (2 μM) treatment, with cellular lipid oxidation

assessed by the rate of change in green cellular fluorescence using time-lapse microscopy. Rescue compounds were co-treated (in

6-point titration) with RSL3. Mean SEM shown, (n = 6 to 9 technical replicates for each condition. Results are representative of 2

independent experiments.

https://doi.org/10.1371/journal.pone.0201369.g002

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

9 / 22

Redox regulation of ferroptosis

A

αT-d₄dosed (100 mg/kg PO, 4h)

αT-d₄(parent) quantified

αTQ-d₄(metabolite) quantified

% αTQ

mean SD (n=4)

B

C

α

THQ

TQ

T

Basal

α

T

α

TQ

Basal

α

T

α

TQ

Basal

α

T

α

TQ

Basal

α

T

αTQ

Dosed Compound:

10 μM

24 h

10 μM

24 h

10 μM

24 h

10 μM

24 h

Fig 3. In vivo conversion of α-tocopherol to its quinone metabolite and quantification of intracellular concentrations of vitamin E and its metabolites in cells in

vitro. (A) To assess the in vivo conversion of αT to its quinone metabolite, stable deuterium-labelled α-tocopherol (d₄-αT) was dosed orally to Sprague-Dawley male

rats. Four hours after the final dose, plasma and tissues were collected for bioanalytical quantification of d₄-αT and d₄-αTQ by LC-MS/MS. While <1% of deuterium-

labeled αT was detected as the quinone in the plasma fraction, varying levels of quinone conversion were observed in the tissues assessed, ranging from ~2% (liver) to

~50% (small intestine). (B) Vitamin E (αT) and its metabolites, vitamin E quinone (αTQ) and vitamin E hydroquinone (αTHQ), were detected simultaneously in Q7

striatal cells under basal growth conditions or supplementation with αT or αTQ (10 μM, 24 h). Results displayed are mean SD, n = 6; results from 1 experiment

representative of 3 similar experiments. (C) Stacked bar graphs showing the mean proportions of αT, αTQ, and αTHQ quantified simultaneously under basal, αT- or

αTQ-supplemented conditions using the succinate capping methodology described in Methods.

https://doi.org/10.1371/journal.pone.0201369.g003

As ferroptotic cell death has been shown to arise from genetic knockdown or chemical inhi-

bition of phospholipid hydroperoxide glutathione peroxidase (GPX4) [6,9,24–26], we utilized

(1S,3R)-RSL3, a well-described irreversible inhibitor of GPX4, to induce ferroptosis in Q7

cells. RSL3 treatment of Q7 cells resulted in dose- and time-dependent loss of cell viability.

Typically, at RSL3 concentrations ꢀ1 μM, 80–95% loss of cell viability was observed 18 h after

RSL3 treatment, with onset of cell death observed ~5 h after RSL3 addition. Consistent with

published reports, αT prevented RSL3-induced cell death with an EC₅₀of 5.2 2.8 μM

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

10 / 22

Redox regulation of ferroptosis

A

αT

αTCC

αTQ

αTHQ

B

Control

α

T

TCC

THQ

TQ

C

Control

αT

αTCC

αTQ

αTHQ

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

11 / 22

Redox regulation of ferroptosis

Fig 4. Oxygen radical absorbance capacity (ORAC) and radical-trapping antioxidant (RTA) activity assays with

αT, αTCC, αTQ, αTHQ in comparison with their ferroptosis rescue potency. (A) ORAC activity of each compound

compared to Trolox. Mean SEM (N = 6) displayed. (B) Overlay of kinetics of RTA activity of each compound

(50 μM) prepared as 2 mol% in DOPC liposomes, detected through quenching of BODIPY™ 581/591 C11 (0.5 μM)

emission at 600 nm. Mean SD (N = 3) displayed. (C) Comparison of RTA BODIPY quenching area-under-the-curve

(AUC) of each compound (relative to the Control group, defined as 1) and pIC₅₀values for inhibition of RSL3-induced

cellular BODIPY oxidation. Mean + SEM displayed; N = 3 for RTA assay, N = 6 for cell BODIPY assay.

https://doi.org/10.1371/journal.pone.0201369.g004

(mean SD, N = 9; Fig 2). αTQ and αTHQ also prevented ferroptotic cell death, but were

markedly more potent (22- and 57-fold, respectively) than αT itself (αTQ EC₅₀0.23

0.10 μM, N = 13; αTHQ EC₅₀0.09 0.03 μM, N = 9, Fig 2A). αT, αTQ, and αTHQ also pre-

vented RSL3-induced lipid oxidation dose-dependently, following the same trend as their

cytoprotective potency (as assessed by the prevention of BODIPY™ 581/591 C11 oxidation by

time-lapse quantitative video microscopy, Fig 2B).

The improved anti-ferroptosis potency of the αTQ and αTHQ metabolites compared to

the parent αT led us to assess the prevalence of conversion in biological systems. To demon-

strate in vivo the oxidative chroman ring opening of αT to αTQ, we synthesized stable-labelled

d₄-αT and dosed fed male rats once daily for four days by oral gavage at 100 mg/kg (S1

Method). The tissues and plasma were isolated four hours after the final dose and analyzed by

LC-MS/MS to determine the concentrations of both d₄-αT and d₄-αTQ. The conversion and

distribution of d₄-αT and d₄-αTQ was tissue- and compartment-specific, ranging from a ratio

of 1:1 (d₄-αT:d₄-αTQ) in the small intestine, reaching concentrations of approximately

200 μM, to plasma where the (d₄-αT:d₄-αTQ) ratio was greater than 100:1. Due to the rapid

and spontaneous re-oxidation of the hydroquinone (αTHQ) to the quinone (αTQ) [27], we

did not attempt to quantify the tissue αTHQ exposure in this study. In summary, these results

directly demonstrate the in vivo conversion of αT to αTQ and are consistent with the findings

of others (Fig 3A)[28–36].

To assess directly the cellular exposure of αT and its metabolites, αTQ and αTHQ, in our

cellular ferroptosis model system, we established an LC-MS/MS method to quantitate all three

analytes simultaneously in cell extracts. Because αTHQ is rapidly oxidized back to αTQ upon

exposure to atmospheric oxygen, [27] we quantified its cellular concentration by trapping it as

the stable bis-succinate ester [37]. We observed that Q7 striatal cells cultured under basal

growth conditions had detectable αT, αTQ, and αTHQ (αT:αTQ:αTHQ proportions 88:3:8;

Fig 3B and 3C), likely due to the αT (quantified at 2.3 μM) in the culture medium containing

10% fetal bovine serum. After αT supplementation (10 μM, 24 h) in the Q7 cell medium, the

cellular αT, αTQ, and αTHQ concentrations were each elevated (Fig 3B), yet maintained a

similar proportional distribution (αT:αTQ:αTHQ, 87:4:9; Fig 3C) to those observed under

basal growth conditions. These observations suggest that a relationship exists between the con-

centration of αT and the corresponding concentrations of αTQ and αTHQ, likely mediated

by αT to αTQ oxidative conversion and subsequent reduction to αTHQ through a cellular

activity such as the widely-expressed NAD(P)H:quinone oxidoreductase 1 (NQO1) [38,39].

When Q7 cells were supplemented with αTQ (10 μM, 24 h), the cellular exposure of αT was

unaffected (Fig 3B, left panel), demonstrating that the dosed αTQ did not biochemically

“back-convert” to αT in these cells. By contrast, supplementation with αTQ led to a dramatic

increase in cellular αTQ and αTHQ exposure (αTQ, >300-fold increase, to over 250 μM;

αTHQ, >600-fold increase, to over 900 μM; Fig 3B). αTQ supplementation of Q7 cells

resulted in intracellular αT:αTQ:αTHQ exposure ratios of 1:22:76 (Fig 3C), a marked alter-

ation from the baseline proportions. These results demonstrate that the dosed quinone (αTQ)

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

12 / 22

Redox regulation of ferroptosis

A

*

40000

30000

20000

10000

0

DMSO

RSL3

only

ns

B

C

ns

**

*

****

0.004

0.002

0.000

0.0003

0.0002

0.0001

LLOQ

0.0000

DMSO

αTQ

only

RSL3

only

RSL3 +

αTQ

DMSO

αTQ

only

RSL3

only

RSL3 +

αTQ

D

E

ns

0.02

0.01

0.00

ns

0.002

0.001

0.000

DMSO

αTQ

only

RSL3

only

RSL3 +

αTQ

DMSO

αTQ

only

RSL3

only

RSL3 +

αTQ

F

G

ns

***

2.5

2.0

1.5

1.0

0.5

0.0

0.25

0.20

0.15

0.10

0.05

0.00

DMSO

αTQ

only

RSL3

only

RSL3 +

αTQ

DMSO

αTQ

only

RSL3

only

RSL3 +

αTQ

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

13 / 22

Redox regulation of ferroptosis

Fig 5. Lipid changes quantitated in RSL3-treated Q7 cells or their culture medium. (A) RSL3-Q7 treated cells have elevated cellular levels

of a lyso-PE-palmitoyl species (^Ãp<0.05, Unpaired t-test with Welch’s correction). (B, C) RSL3-treated Q7 cells have elevated cellular and

conditioned media levels of 15-HETE, an effect prevented by αTQ (5 μM) co-treatment (^Ãp<0.05, ^ÃÃp<0.01, ^ÃÃÃp<0.001, Shapiro-Welk

normality test followed by ANOVA with Sidak’s multiple comparisons test or Kruskal-Wallis test with Dunn’s multiple comparisons test,

respectively). (D, E) 12-HETE levels were not altered by RSL3 or αTQ treatment (ns, p>0.05, Shapiro-Welk normality test followed by

Kruskal-Wallis test with multiple comparisons test. (F, G) Conditioned medium of RSL3-treated Q7 cells has elevated AA, but cellular levels

are not significantly changed (^ÃÃÃp<0.001, Shapiro-Welk normality test followed by ANOVA with Sidak’s multiple comparisons test).

https://doi.org/10.1371/journal.pone.0201369.g005

is reduced to the hydroquinone (αTHQ) form readily in cells; indeed, the αTHQ exposures

exceed those of αTQ by 3- to 4-fold in all cases tested.

To test whether metabolic conversion of the αT chroman ring to the corresponding αTQ

quinone is necessary for anti-ferroptotic activity, we designed and synthesized a novel non-

metabolizable analog of αT we refer to as αT carbochroman (αTCC, Fig 1). αTCC is a conser-

vative, one-atom, oxygen-to-methylene substituted analog of αT which retains the same phe-

nolic group in αT that is widely regarded as the moiety responsible for vitamin E’s radical

scavenging activity [40] (see S2 Method. Procedure for the synthesis of αTCC). We confirmed

that, similar to αT, αTCC forms a radical through hydrogen atom donation, and this radical is

delocalized over the carbochroman ring system (S1 Fig). Furthermore, when assessed in a

solution-based oxygen radical absorbance capacity (ORAC) assay, as expected, αTCC had

comparable antioxidant activity to αT (0.075 0.008 vs. 0.073 0.008 Trolox equivalence

units, respectively; mean SEM, n = 6 each; p>0.05 by two-tailed t-test; Figs 1 and 4A) [41].

Despite the similarity in antioxidant activity and radical formation, however, the anti-ferropto-

tic cellular activity of αTCC was markedly decreased compared to αT, with an EC₅₀> 30 μM

(Fig 2A) and only 42 21% (mean SD; N = 7 independent experiments) rescue at 30 μM, the

highest concentration tested due to solubility limitations. We confirmed cellular exposure of

αTCC to be >50 μM (6 technical replicates in each of 4 independent experiments, S3 Method),

indicating that the weak rescue activity was not due to poor cellular uptake.

Ferroptosis occurs via oxidation of the lipid membrane, the primary location of the highly

lipophilic vitamin E species. We next sought to determine whether the rank order of antifer-

roptotic potency of αT, αTQ, αTHQ, and αTCC could be predicted by antioxidant activity in

a more biologically relevant lipid environment. We therefore tested the radical-trapping activ-

ity (RTA) of αT, αTCC, αTQ, and αTHQ in liposome vesicles as measured by an increased

lag period prior to quenching of BODIPY™ 581/591 C11 fluorescence (Fig 4A) [8,42]. Under

these conditions we found that the time to quench BODIPY™ 581/591 C11 fluorescence in

DOPC liposomes showed a rank order of αT> αTCC>> αTHQ> αTQ (Fig 4B). This result

is further highlighted by an area-under-the-curve (AUC) comparison of each compound

determined with respect to the liposome only control, which shows that both αT and αTCC

have similar activity and are between 2.5–4 fold greater than αTQ and αTHQ (Fig 4B). These

data are in stark contrast to the concentrations required to prevent RSL3-induced cellular

BODIPY™ 581/591 C11 oxidation (Fig 4B, compare open vs. closed bars), and argues that radi-

cal trapping antioxidant activity alone is not sufficient to explain the increased potency of

αTQ and αTHQ (Fig 4C). Our results suggest that the radical scavenging activity of αT is not

sufficient to prevent ferroptosis, consistent with the hypothesis that the quinone/hydroqui-

none metabolites are largely responsible for the anti-ferroptotic effect attributed to αT. Fur-

thermore, while αTHQ exhibits weak but quantifiable ORAC and RTA activity, these activities

alone cannot explain the anti-ferroptotic potency of αTHQ (Fig 4B and 4C).

It has been reported that ferroptosis requires lipoxygenase-mediated production of peroxi-

dated polyunsaturated fatty acids [6,26]. Therefore, we studied the ability of the αT metabolite,

αTQ, to modulate the generation of endogenous oxidized lipids in Q7 cells [38]. During

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

14 / 22

Redox regulation of ferroptosis

A

0.006

0.004

0.002

α

T

TQ

LLOQ

0.000

0.001

0.01

0.1

1

10

[compound] (μM)

B

100

50

0

α

T

TQ

THQ

TCC

0.001

0.01

0.1

1

10

100

[compound] (μM)

SLO-Fe³⁺

g = 6.1

[Fe³⁺] = 100%

SLO untreated

SLO + Lipid-OOH

SLO + Lipid-OOH + αTHQ

SLO + Lipid-OOH + αTHQ3

C

Free Fe³⁺

g = 4.3

50%

28%

600

800

1000

1200

1400

1600

1800

2000

Magnetic Field (G)

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

15 / 22

Redox regulation of ferroptosis

Fig 6. Inhibitory activity of 15-LO by vitamin E and its metabolites in an isolated enzyme assay and in Q7 cells.

(A) αTQ dose-dependently prevents RSL3-induced cellular release of 15-HETE (IC₅₀~0.6 μM). Mean SD shown

(n = 3), representative of 2 independent experiments. (B) αTHQ inhibited rabbit 15-lipoxygenase (15-LO) enzyme

activity (IC₅₀2.45 μM) utilizing arachidonic acid as substrate, whereas αT, αTQ, and αTCC did not (up to 100 μM).

Data are mean SD (n = 4, from 2 independent experiments). (C) EPR spectroscopy of soybean 15-lipoxygenase

(SLO) with αTHQ or αT3HQ. Overlaid spectra focusing on the active site Fe³⁺(g = 6.1) and the exogenous rhombic

Fe³⁺that is typically present in purified protein fractions (g = 4.3). Quantification of the Fe³⁺-SLO before treatment,

after oxidation by lipid-hydroperoxide (set to 100%), and after compound treatment.

https://doi.org/10.1371/journal.pone.0201369.g006

untargeted LC-MS/MS analysis of Q7 cells undergoing RSL3-induced ferroptosis, we observed

a significant increase in a lyso-phosphatidylethanolamine(PE)-palmitoyl species (m/z =

436.284, neg. mode), suggesting that free fatty acids might be released during this death pro-

cess (Fig 5A). Given the central role of arachidonate-containing PE lipids in ferroptosis

[2,6,26,43], we assessed whether arachidonic acid (AA) and/or 19 of its known metabolites

were released from cells into the culture medium during RSL3 treatment (S1 Table). Interest-

ingly, in three independent experiments, only two analytes were significantly altered by RSL3

treatment: 15-hydroxyeicosatetraenoic acid (15-HETE) and AA itself (p<0.05 by Kruskal-

Wallis test with Dunn’s correction for multiple comparisons, Fig 5B and 5F). None of the

hydroperoxidated arachidonyl species were detected in our experiments. We also observed

that cellular free 15-HETE but not AA was induced by RSL3 treatment (Fig 5C and 5G).

Importantly, the RSL3-dependent induction of both cellular and released 15-HETE were

inhibited by αTQ (5 μM, Fig 5B and 5C); by contrast, αTQ treatment had no effect on cellular

or released AA levels (Fig 5F and 5G). Finally, we note that neither cellular nor released

12-HETE levels were induced by RSL3 treatment, suggesting a specific role for 15-lipoxygenase

in Q7 cells undergoing ferroptosis (Fig 5D and 5E).

We sought to determine the inhibitory potency of the parent αT or its quinone metabolite

αTQ on RSL3-induced cellular 15-HETE release. When Q7 cells were cotreated with RSL3

and αT, 15-HETE production was reduced by only 19% at 10 μM with αT. In contrast, αTQ

dose-dependently returned 15-HETE concentration to background levels (IC₅₀= 0.6 μM; Fig

6A) with a potency similar to its cytoprotective activity (EC₅₀= 0.23 μM; Figs 1 and 2A). The

dose-dependent inhibition of 15-LO mediated lipid metabolites by αTQ at similar concentra-

tions to its cellular antiferroptotic potency suggests that these two activities might be mecha-

nistically linked.

To determine whether the superior anti-ferroptotic potency of αTHQ and αTQ relative to

αT and αTCC could be due to direct inhibition of 15-LO enzyme activity, the rate of AA per-

oxidation using purified rabbit 15-LO was measured over a range of compound concentra-

tions. Only αTHQ displayed potent inhibition of 15-LO (IC₅₀= 2.45 μM), while αT, αTCC,

and αTQ all failed to decrease enzyme activity by more than 20% at 100 μM, the highest con-

centration tested (Fig 6B). These data demonstrate that the 15-LO inhibitory potency of

αTHQ is in excess of 40-fold when compared with the potency of either αTQ, αT, or αTCC.

The catalytic activity of 15-Lipoxygenase requires a non-heme iron center to be in the Fe³⁺

state. We investigated whether the inhibitory activity of αTHQ resulted from the reduction of

the 15-LO Fe³⁺to its inactive Fe²⁺state using the 15-LO enzyme from soybean (SLO), and

EPR spectroscopy to measure its redox state [44,45]. Incubation of SLO with α-linoleic acid

hydroperoxide (lipid-OOH) converted all of the iron to the Fe³⁺state, which was spin quanti-

fied and set to 100% Fe³⁺(Fig 6C). Upon addition of αTHQ, a 50% decrease in the SLO-Fe³⁺

signal was observed. A lack of change in the exogenous rhombic Fe³⁺at g = 4.3 (which is not

bound to SLO) indicated that Fe³⁺reduction by αTHQ was specific to the enzyme-bound

metal.

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

16 / 22

Redox regulation of ferroptosis

Oxygen to -CH₂- substitution

Vitamine E

O

HO

α-Tocopherol (αT)

α-Tocopherol carbochroman (αTCC)

1. Conversion

No Conversion

O

HO

α-Tocopherol quinone (αTQ)

2. Reduction

NQO1

OH

HO

α-Tocopherol hydroquinone (αTHQ)

3. Inhibition

15-LO

GPX4

PUFA

PUFA-OOH

PUFA-OH

4. Ferroptosis & Inflammation

Fig 7. Proposed target and mechanism of action of vitamin E. The anti-ferroptotic effect of vitamin E results from a four-step biochemical

mechanism: 1) oxidative hydrolysis of alpha-tocopherol (αT) to alpha-tocopherol quinone (αTQ); 2) reduction of alpha-tocopherol quinone (αTQ) to

alpha-tocopherol hydroquinone (αTHQ); 3) inhibition of 15-LO via reduction of its non-heme Fe³⁺center to the inactive Fe²⁺state by alpha-

tocopherol hydroquinone (αTHQ); and 4) inhibition of the ferroptosis cascade by blocking formation of lipid peroxidation products: polyunsaturated

fatty acids (PUFA), hydroperoxidated polyunsaturated fatty acid (PUFA-OOH) and hydroxylated polyunsaturated fatty acid (PUFA-OH). The non-

metabolizable isosteric analog of vitamin E αTCC retains its antioxidant activity, but does not inhibit ferroptosis.

https://doi.org/10.1371/journal.pone.0201369.g007

To determine the key molecular features of vitamin E and its metabolites that afford the

ability to inhibit 15-LO and ferroptosis, we prepared a library of each of the naturally occur-

ring vitamers (α-, γ- and δ-tocopherols and tocotrienols) and their major metabolites (Fig 1).

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

17 / 22

Redox regulation of ferroptosis

With this library, we related structural changes in the chroman head, lipid tail, and quinone

ring functionality with 15-LO inhibition and protective anti-ferroptotic activity. Of particular

interest was determining whether the two major forms of vitamin E, tocopherols and tocotrie-

nols, function by the same mechanism-of-action. Consistent with Kagan et al., [6] we observed

that αT3 was more potent in inhibiting ferroptosis than αT (~30-fold, Fig 1), a trend that was

also observed for the γ-, and δ-forms (~18- and ~10-fold, respectively). We also found that the

quinone and hydroquinone metabolites of α-, γ- and δ-tocotrienols were more potent than

their parent chroman forms. This anti-ferroptotic activity was due to a similar mechanism of

action as the tocopherol series: αT3HQ also potently inhibited 15-LO enzymatic activity and

reduced the Fe³⁺catalytic center in SLO (Figs 1 and 6C). The potent anti-ferroptotic activity

(EC₅₀= 0.043 μM) of alpha-tocotrienol quinone αT3Q, together with a favorable pharmacoki-

netic and safety profile, position it as an attractive candidate to treat diseases involving oxida-

tive stress, inflammation, and ferroptosis[14].

Discussion

In summary, we dissected the structural and enzymatic basis for vitamin E pharmacology

based on protection from ferroptosis. To address the prevailing hypothesis that the lipid anti-

oxidant activity of vitamin E (αT) is responsible for its anti-ferroptotic activity, we used the

novel, non-metabolizable αT analog, αTCC, as a chemical biology probe and found that the

radical trapping activity of αT is insufficient to explain its anti-ferroptotic cellular activity (Fig

4C). We present evidence that the anti-ferroptotic biological activity of vitamin E results in-

stead from a direct action of its hydroquinone metabolite on the lipoxygenase active site Fe³⁺

center (Fig 7). Specifically, using the data reported herein, we propose the following model for

vitamin E mechanism-of-action: (1) oxidative chroman ring opening of αT to its quinone

metabolite, αTQ (Fig 3A); (2) in situ reduction of αTQ to its hydroquinone form (αTHQ; Fig

3B and 3C); deactivation of lipoxygenase enzyme activity (Fig 6B) by αTHQ via direct reduc-

tion of the active-site Fe³⁺to its inactive Fe²⁺state (Fig 6C); (4) thereby preventing cellular

lipid oxidation (Figs 2B, 4B, 5B, 5C and 6A) and subsequent ferroptotic cell death (Fig 2A).

In 1939, Evans and Bishop noted that both vitamin E (αT) and vitamin E quinone (αTQ)

possessed protective activity in fetal resorption studies[46,47]. For reasons that were not artic-

ulated in their initial work, vitamin E (αT chroman) was defined as the essential nutritional

factor, not the vitamin E quinone. More recently, a vitamin E effect on fetal resorption was

demonstrated in mice by action at the 15-LO/GPX4 ferroptotic pathway[7,48,49]. Our experi-

mental evidence suggests that the hallmark activity leading to the classification of alpha-

tocopherol as a vitamin may be in fact mediated by its metabolite, vitamin E hydroquinone.

Supporting information

S1 Method. The synthesis of d₄-α-tocopherol (d₄-αT) and d₄-α-tocopherol quinone (d₄-

αTQ).

(DOCX)

S2 Method. The synthesis of αTCC–alpha tocopherol carbochroman.

(DOCX)

S3 Method. Cellular quantification of alpha-tocopherol carbochroman (αTCC).

(DOCX)

S1 Fig. EPR spectrum of the radical formed upon oxidation of αTCC.

(DOCX)

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

18 / 22

Redox regulation of ferroptosis

S1 Table. The metabolites measured in culture medium during ferroptosis by LC-MS/MS.

(DOCX)

Acknowledgments

We thank Professor James E. Rothman for his insights and editorial guidance regarding this

research and manuscript. We thank Sekhar Boddupalli and Orion Jankowski for their insights

into vitamin E metabolism. We thank Viktoria Kheifets for her development and insights

regarding the fibroblast GSH depletion assay.

Author Contributions

Conceptualization: Charles R. Holst, Kevin P. McCusker, Stephanie A. Malone, Steve James

Richards, Guy Miller, William D. Shrader.

Data curation: Charles R. Holst, Go¨zde Ulas, Amanda H. Kahn-Kirby, Yuko Kosaka, Veron-

ica Rivera.

Formal analysis: Charles R. Holst, Joey C. Latham, Go¨zde Ulas, Kevin G. Hoff, Amanda H.

Kahn-Kirby, William D. Shrader.

Funding acquisition: Guy Miller, Jeffrey K. Trimmer.

Investigation: Charles R. Holst, Joel J. Bruegger, Go¨zde Ulas, Akiko Amagata, Dana Davis,

Amanda H. Kahn-Kirby, Virna Kim, Yuko Kosaka, Edgar Lee, Stephanie A. Malone, Janet

J. Mei, Jeffrey K. Trimmer, William D. Shrader.

Methodology: Andrew Hinman, Charles R. Holst, Joey C. Latham, Joel J. Bruegger, Go¨zde

Ulas, Kevin P. McCusker, Akiko Amagata, Dana Davis, Kevin G. Hoff, Amanda H. Kahn-

Kirby, Virna Kim, Yuko Kosaka, Edgar Lee, Janet J. Mei, Steve James Richards, Veronica

Rivera, William D. Shrader.

Project administration: Andrew Hinman, Charles R. Holst, Steve James Richards, Jeffrey K.

Trimmer, William D. Shrader.

Resources: Kevin P. McCusker, Janet J. Mei, Guy Miller, Jeffrey K. Trimmer, William D.

Shrader.

Supervision: Andrew Hinman, Charles R. Holst, Joey C. Latham, Steve James Richards, Jeffrey

K. Trimmer, William D. Shrader.

Validation: Joey C. Latham, Joel J. Bruegger, Amanda H. Kahn-Kirby, Virna Kim, Stephanie

A. Malone, Veronica Rivera.

Visualization: Amanda H. Kahn-Kirby, William D. Shrader.

Writing – original draft: Charles R. Holst, Steve James Richards, William D. Shrader.

Writing – review & editing: Andrew Hinman, Joel J. Bruegger, Go¨zde Ulas, Kevin P. McCus-

ker, Dana Davis, Kevin G. Hoff, Amanda H. Kahn-Kirby, Stephanie A. Malone, Steve

James Richards, Guy Miller, Jeffrey K. Trimmer, William D. Shrader.

References

1. Stockwell BR, Friedmann Angeli JP, Bayir H, Bush AI, Conrad M, Dixon SJ, et al. Ferroptosis: A Regu-

lated Cell Death Nexus Linking Metabolism, Redox Biology, and Disease. Cell. 2017; 171: 273–285.

https://doi.org/10.1016/j.cell.2017.09.021 PMID: 28985560

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

19 / 22

Redox regulation of ferroptosis

2. Wenzel SE, Tyurina YY, Zhao J, St Croix CM, Dar HH, Mao G, et al. PEBP1 Wardens Ferroptosis by

Enabling Lipoxygenase Generation of Lipid Death Signals. Cell. 2017; 171: 628–641.e26. https://doi.

org/10.1016/j.cell.2017.09.044 PMID: 29053969

3. Drefs M, Thomas MN, Guba M, Angele MK, Werner J, Conrad M, et al. Modulation of Glutathione

Hemostasis by Inhibition of 12/15-Lipoxygenase Prevents ROS-Mediated Cell Death after Hepatic

Ischemia and Reperfusion. Oxid Med Cell Longev. 2017; 2017: 8325754. https://doi.org/10.1155/2017/

8325754 PMID: 28811867

´

4. Do Van B, Gouel F, Jonneaux A, Timmerman K, Gele P, Petrault M, et al. Ferroptosis, a newly charac-

terized form of cell death in Parkinson’s disease that is regulated by PKC. Neurobiol Dis. Elsevier Inc.;

2016; 94: 169–178. https://doi.org/10.1016/j.nbd.2016.05.011 PMID: 27189756

´

5. Do Van B, Gouel F, Jonneaux A, Timmerman K, Gele P, Petrault M, et al. Ferroptosis, a newly charac-

terized form of cell death in Parkinson’s disease that is regulated by PKC. Neurobiol Dis. Elsevier B.V.;

2016; 94: 169–78. https://doi.org/10.1016/j.nbd.2016.05.011 PMID: 27189756

6. Kagan VE, Mao G, Qu F, Angeli JPF, Doll S, Croix CS, et al. Oxidized arachidonic and adrenic PEs nav-

igate cells to ferroptosis. Nat Chem Biol. 2016; 13: 81–90. https://doi.org/10.1038/nchembio.2238

PMID: 27842066

7. Angeli JPF, Shah R, Pratt DA, Conrad M. Ferroptosis Inhibition: Mechanisms and Opportunities. Trends

Pharmacol Sci. Elsevier Ltd; 2017; 38: 489–498. https://doi.org/10.1016/j.tips.2017.02.005 PMID:

28363764

8. Shah R, Shchepinov MS, Pratt DA. Resolving the Role of Lipoxygenases in the Initiation and Execution

of Ferroptosis. ACS Cent Sci. 2018; acscentsci.7b00589. https://doi.org/10.1021/acscentsci.7b00589

PMID: 29632885

9. Yang WS, Stockwell BR. Synthetic lethal screening identifies compounds activating iron-dependent,

nonapoptotic cell death in oncogenic-RAS-harboring cancer cells. Chem Biol. 2008; 15: 234–45. https://

doi.org/10.1016/j.chembiol.2008.02.010 PMID: 18355723

10. Shrader WD, Amagata A, Barnes A, Hinman A, Jankowski O, Lee E, et al. Towards a modern definition

of vitamin E-evidence for a quinone hypothesis. Bioorg Med Chem Lett. Elsevier Ltd; 2012; 22: 391–5.

https://doi.org/10.1016/j.bmcl.2011.10.117 PMID: 22137789

11. Shrader WD, Amagata A, Barnes A, Enns GM, Hinman A, Jankowski O, et al. α-Tocotrienol quinone

modulates oxidative stress response and the biochemistry of aging. Bioorg Med Chem Lett. Elsevier

Ltd; 2011; 21: 3693–8. https://doi.org/10.1016/j.bmcl.2011.04.085 PMID: 21600768

12. Enns GMGM, Kinsman SLSL, Perlman SLSL, Spicer KMKM, Abdenur JEJE, Cohen BHBH, et al. Initial

experience in the treatment of inherited mitochondrial disease with EPI-743. Mol Genet Metab. Elsevier

Inc.; 2012; 105: 91–102. https://doi.org/10.1016/j.ymgme.2011.10.009 PMID: 22115768

13. Lynch DR, Willi SM, Wilson RB, Cotticelli MGG, Brigatti KW, Deutsch EC, et al. A0001 in Friedreich

ataxia: biochemical characterization and effects in a clinical trial. Mov Disord. 2012; 27: 1026–33.

https://doi.org/10.1002/mds.25058 PMID: 22744651

14. Martinelli D, Catteruccia M, Piemonte F, Pastore A, Tozzi G, Dionisi-Vici C, et al. EPI-743 reverses the

progression of the pediatric mitochondrial disease—genetically defined Leigh Syndrome. Mol Genet

Metab. Elsevier Inc.; 2012; 107: 383–8. https://doi.org/10.1016/j.ymgme.2012.09.007 PMID: 23010433

15. Khanna S, Roy S, Ryu H, Bahadduri P, Swaan PW, Ratan RR, et al. Molecular Basis of Vitamin E

Action: Tocotrienol modulates 12-lipoxygenase, a key mediator of glutamate-induced neurodegenera-

tion. J Biol Chem. 2003; 278: 43508–43515. https://doi.org/10.1074/jbc.M307075200 PMID: 12917400

¨

16. Seiler A, Schneider M, Forster H, Roth S, Wirth EK, Culmsee C, et al. Glutathione peroxidase 4 senses

and translates oxidative stress into 12/15-lipoxygenase dependent- and AIF-mediated cell death. Cell

Metab. 2008; 8: 237–48. https://doi.org/10.1016/j.cmet.2008.07.005 PMID: 18762024

17. Chen L, Hambright WS, Na R, Ran Q. Ablation of the Ferroptosis Inhibitor Glutathione Peroxidase 4 in

Neurons Results in Rapid Motor Neuron Degeneration and Paralysis. J Biol Chem. 2015; 290: 28097–

106. https://doi.org/10.1074/jbc.M115.680090 PMID: 26400084

18. Carlson BA, Tobe R, Yefremova E, Tsuji PA, Hoffmann VJ, Schweizer U, et al. Glutathione peroxidase

4 and vitamin E cooperatively prevent hepatocellular degeneration. Redox Biol. 2016; 9: 22–31. https://

doi.org/10.1016/j.redox.2016.05.003 PMID: 27262435

19. Matsushita M, Freigang S, Schneider C, Conrad M, Bornkamm GW, Kopf M. T cell lipid peroxidation

induces ferroptosis and prevents immunity to infection. J Exp Med. 2015; 212: 555–568. https://doi.org/

10.1084/jem.20140857 PMID: 25824823

20. Ruan Q, Lesort M, MacDonald ME, Johnson GVW. Striatal cells from mutant huntingtin knock-in mice

are selectively vulnerable to mitochondrial complex II inhibitor-induced cell death through a non-apopto-

tic pathway. Hum Mol Genet. 2004; 13: 669–81. https://doi.org/10.1093/hmg/ddh082 PMID: 14962977

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

20 / 22

Redox regulation of ferroptosis

21. Wu JH, Croft KD. Vitamin E metabolism. Mol Aspects Med. 2007; 28: 437–52. https://doi.org/10.1016/j.

mam.2006.12.007 PMID: 17306359

¨

22. Schmolz L, Birringer M, Lorkowski S, Wallert M. Complexity of vitamin E metabolism. World J Biol

Chem. 2016; 7: 14–43. https://doi.org/10.4331/wjbc.v7.i1.14 PMID: 26981194

23. Galli F, Polidori MC, Stahl W, Mecocci P, Kelly FJ. Vitamin E biotransformation in humans. Vitam Horm.

2007; 76: 263–80. https://doi.org/10.1016/S0083-6729(07)76009-0 PMID: 17628177

24. Angeli JPF, Proneth B, Hammond VJ, Tyurina YY, Tyurin VA, O‘Donnell VB, et al. Inactivation of the

Ferroptosis Regulator Gpx4 Triggers Acute Renal Failure in a Therapeutically Relevant Mechanism.

Free Radic Biol Med. Elsevier; 2014; 76: S77–S78. https://doi.org/10.1016/j.freeradbiomed.2014.10.

276

25. Shimada K, Skouta R, Kaplan A, Yang WS, Hayano M, Dixon SJ, et al. Global survey of cell death

mechanisms reveals metabolic regulation of ferroptosis. Nat Chem Biol. Nature Publishing Group;

2016; 1–10. https://doi.org/10.1038/nchembio.2079 PMID: 27159577

26. Yang WS, Kim KJ, Gaschler MM, Patel M, Shchepinov MS, Stockwell BR. Peroxidation of polyunsatu-

rated fatty acids by lipoxygenases drives ferroptosis. Proc Natl Acad Sci U S A. 2016; 113: E4966–75.

https://doi.org/10.1073/pnas.1603244113 PMID: 27506793

27. Hughes PE, Tove SB. Synthesis of alpha-tocopherolquinone by the rat and its reduction by mitochon-

dria. J Biol Chem. 1980; 255: 7095–7. Available: http://www.ncbi.nlm.nih.gov/pubmed/7391070 PMID:

7391070

28. Kaneko K, Kiyose C, Ueda T, Ichikawa H, Igarashi O. Studies of the metabolism of alpha-tocopherol

stereoisomers in rats using [5-methyl-(14)C]SRR- and RRR-alpha-tocopherol. J Lipid Res. 2000; 41:

357–67. Available: http://www.ncbi.nlm.nih.gov/pubmed/10706583 PMID: 10706583

29. Csallany AS, Draper HH, Shah SN. Conversion of d-α-tocopherol-C14 to tocopheryl-p-quinone in vivo.

Arch Biochem Biophys. 1962; 98: 142–145. https://doi.org/10.1016/0003-9861(62)90159-5 PMID:

13882501

30. Plack PA, Bieri JG. Metabolic products of α-tocopherol in the livers of rats given intraperitoneal injec-

tions of [14C]-α-tocopherol. Biochim Biophys Acta—Spec Sect Lipids Relat Subj. 1964; 84: 729–738.

https://doi.org/10.1016/0926-6542(64)90031-9

31. Jain SK, Levine SN. Elevated lipid peroxidation and vitamin e-quinone levels in heart ventricles of strep-

tozotocin-treated diabetic rats. Free Radic Biol Med. 1995; 18: 337–341. https://doi.org/10.1016/0891-

5849(94)00114-Y PMID: 7744318

32. Kiyose C, Kaneko K, Muramatsu R, Ueda T, Igarashi O. Simultaneous determination of RRR- and

SRR-alpha-tocopherols and their quinones in rat plasma and tissues by using chiral high-performance

liquid chromatography. Lipids. 1999; 34: 415–22. Available: http://www.ncbi.nlm.nih.gov/pubmed/

10443975 PMID: 10443975

33. Upston JM, Witting PK, Brown AJ, Stocker R, Keaney JF. Effect of vitamin E on aortic lipid oxidation

and intimal proliferation after arterial injury in cholesterol-fed rabbits. Free Radic Biol Med. 2001; 31:

1245–53. https://doi.org/10.1016/S0891-5849(01)00721-3 PMID: 11705703

34. Terentis AC, Thomas SR, Burr JA, Liebler DC, Stocker R. Vitamin E oxidation in human atherosclerotic

lesions. Circ Res. 2002; 90: 333–9. https://doi.org/10.1161/hh0302.104454 PMID: 11861423

35. Mottier P, Gremaud E, Guy P a, Turesky RJ. Comparison of gas chromatography-mass spectrometry

and liquid chromatography-tandem mass spectrometry methods to quantify alpha-tocopherol and

alpha-tocopherolquinone levels in human plasma. Anal Biochem. 2002; 301: 128–35. https://doi.org/10.

1006/abio.2001.5486 PMID: 11811977

36. Shichiri M, Yoshida Y, Ishida N, Hagihara Y, Iwahashi H, Tamai H, et al. α-Tocopherol suppresses lipid

peroxidation and behavioral and cognitive impairments in the Ts65Dn mouse model of Down syndrome.

Free Radic Biol Med. Elsevier Inc.; 2011; 50: 1801–11. https://doi.org/10.1016/j.freeradbiomed.2011.

03.023 PMID: 21447382

37. Hinman AW. Methods of analyzing p-hydroquinone levels and ratios. WIPO—World Intellectual Prop-

erty Organization; WO 2018/081644 A1, 2018. p. 90.

38. Siegel D, Bolton EM, Burr JA, Liebler DC, Ross D. The reduction of alpha-tocopherolquinone by human

NAD(P)H: quinone oxidoreductase: the role of alpha-tocopherolhydroquinone as a cellular antioxidant.

Mol Pharmacol. 1997; 52: 300–5. Available: http://www.ncbi.nlm.nih.gov/pubmed/9271353 PMID:

9271353

39. Hayashi T, Kanetoshi A, Nakamura M, Tamura M, Shirahama H. Reduction of alpha-tocopherolquinone

to alpha-tocopherolhydroquinone in rat hepatocytes. Biochem Pharmacol. 1992; 44: 489–93. Available:

http://www.ncbi.nlm.nih.gov/pubmed/1510699 PMID: 1510699

40. Traber MG, Atkinson J. Vitamin E, antioxidant and nothing more. Free Radic Biol Med. 2007; 43: 4–15.

https://doi.org/10.1016/j.freeradbiomed.2007.03.024 PMID: 17561088

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

21 / 22

Redox regulation of ferroptosis

41. Cao G, Alessio HM, Cutler RG. Oxygen-radical absorbance capacity assay for antioxidants. Free Radic

Biol Med. 1993; 14: 303–11. https://doi.org/10.1016/0891-5849(93)90027-R PMID: 8458588

42. Shah R, Margison K, Pratt DA. The Potency of Diarylamine Radical-Trapping Antioxidants as Inhibitors

of Ferroptosis Underscores the Role of Autoxidation in the Mechanism of Cell Death. ACS Chem Biol.

2017; 12: 2538–2545. https://doi.org/10.1021/acschembio.7b00730 PMID: 28837769

43. Doll S, Proneth B, Tyurina YY, Panzilius E, Kobayashi S, Ingold I, et al. ACSL4 dictates ferroptosis sen-

sitivity by shaping cellular lipid composition. Nat Chem Biol. 2016; 13: 91–98. https://doi.org/10.1038/

nchembio.2239 PMID: 27842070

44. Kemal C, Louis-Flamberg P, Krupinski-Olsen R, Shorter AL. Reductive Inactivation of Soybean Lipoxy-

genase 1 by Catechols: A Possible Mechanism for Regulation of Lipoxygenase Activity. Biochemistry.

1987; 26: 7064–7072. https://doi.org/10.1021/bi00396a031 PMID: 3122826

45. Nelson MJ, Batt DG, Thompson JS, Wright SW. Reduction of the active-site iron by potent inhibitors of

lipoxygenases. J Biol Chem. 1991; 266: 8225–9. Available: http://www.ncbi.nlm.nih.gov/pubmed/

1850741 PMID: 1850741

46. Evans HM, Emerson OH, Emerson GA, Smith LI, Ungnade HE, Prichard WW, et al. The chemistry of

Vitamin E. XII. Specificity and relationship between chemical structure and Vitamin E activity. J Org

Chem. 1939; 04: 376–388. https://doi.org/10.1021/jo01216a003

47. Emmerson OH, Emmerson GA, Evans HM. The vitamin E activity of alpha-tocoquinone. J Biol Chem.

1939; 131: 409–412.

48. Lebold KM, Traber MG. Interactions between α-tocopherol, polyunsaturated fatty acids, and lipoxy-

genases during embryogenesis. Free Radic Biol Med. Elsevier; 2014; 66: 13–9. https://doi.org/10.1016/

j.freeradbiomed.2013.07.039 PMID: 23920314

49. Yant LJ, Ran Q, Rao L, Van Remmen H, Shibatani T, Belter JG, et al. The selenoprotein GPX4 is

essential for mouse development and protects from radiation and oxidative damage insults. Free Radic

Biol Med. 2003; 34: 496–502. https://doi.org/10.1016/S0891-5849(02)01360-6 PMID: 12566075

PLOS ONE | https://doi.org/10.1371/journal.pone.0201369 August 15, 2018

22 / 22

Article Doi

Vitamin E hydroquinone is an endogenous regulator of ferroptosis via redox control of 15-lipoxygenase

DOI: 10.1371/journal.pone.0201369

Source and publish data:

Authors:

Article abstract of DOI:10.1371/journal.pone.0201369

Full text of DOI:10.1371/journal.pone.0201369

Products guided by the article

R&D Labs maybe for 1254952-19-6

Relevant to this article

Hot Product