Skip to main page content
U.S. flag
See More

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jan;13(1):81-90.
doi: 10.1038/nchembio.2238. Epub 2016 Nov 14.

Oxidized arachidonic and adrenic PEs navigate cells to ferroptosis

Valerian E Kagan  1   2   3   4 , Gaowei Mao  1   5 , Feng Qu  1 , Jose Pedro Friedmann Angeli  6 , Sebastian Doll  6 , Claudette St Croix  7 , Haider Hussain Dar  1 , Bing Liu  8 , Vladimir A Tyurin  1 , Vladimir B Ritov  1 , Alexandr A Kapralov  1 , Andrew A Amoscato  1 , Jianfei Jiang  1 , Tamil Anthonymuthu  5 , Dariush Mohammadyani  1 , Qin Yang  5 , Bettina Proneth  6 , Judith Klein-Seetharaman  9 , Simon Watkins  7 , Ivet Bahar  8 , Joel Greenberger  4 , Rama K Mallampalli  10 , Brent R Stockwell  11   12 , Yulia Y Tyurina  1 , Marcus Conrad  6 , Hülya Bayır  1   5
Affiliations

Oxidized arachidonic and adrenic PEs navigate cells to ferroptosis

Valerian E Kagan et al. Nat Chem Biol. 2017 Jan.

Abstract

Enigmatic lipid peroxidation products have been claimed as the proximate executioners of ferroptosis-a specialized death program triggered by insufficiency of glutathione peroxidase 4 (GPX4). Using quantitative redox lipidomics, reverse genetics, bioinformatics and systems biology, we discovered that ferroptosis involves a highly organized oxygenation center, wherein oxidation in endoplasmic-reticulum-associated compartments occurs on only one class of phospholipids (phosphatidylethanolamines (PEs)) and is specific toward two fatty acyls-arachidonoyl (AA) and adrenoyl (AdA). Suppression of AA or AdA esterification into PE by genetic or pharmacological inhibition of acyl-CoA synthase 4 (ACSL4) acts as a specific antiferroptotic rescue pathway. Lipoxygenase (LOX) generates doubly and triply-oxygenated (15-hydroperoxy)-diacylated PE species, which act as death signals, and tocopherols and tocotrienols (vitamin E) suppress LOX and protect against ferroptosis, suggesting a homeostatic physiological role for vitamin E. This oxidative PE death pathway may also represent a target for drug discovery.

PubMed Disclaimer

Conflict of interest statement

Competing financial interests

The authors declare no competing financial interests.

Figures

Figure 1
Figure 1. Oxygenation of esterified (but not free) AA contributes to RSL3-induced ferroptosis in WT and Acsl4 KO Pfa1 cells
(a) Live cell fluorescence imaging of lipid hydroperoxides in WT and Acsl4 KO cells treated with RSL3 (100 nM, 6 hrs, scale bar 5 μm). (b) LiperFluo fluorescence intensity following RSL3 treatment in WT and Acsl4 KO cells. Insert: Fluorescence time course after RSL3 treatment (100 nM) in WT and Acsl4 KO cells with a time control. The data were from a minimum of 10 stage positions. For the statistical analysis, each stage position counted as one data entry. *P < 0.05 vs. WT+RSL3 cells (t-test). (c) WT and Acsl4 KO cells were treated with RSL3 (100 nM) for 2, 4, and 6 hrs before cell death analysis. Data are mean ± s.d., n=3. *P < 0.05 vs. ACSL4 KO+RSL3 (t-test). (d) Fluorescence responses from Mito-FAP (red, upper panel) and ER-FAP (red, lower panel) vs. LiperFluo (green, both panels, scale bars 5 μm). (e) ACSL inhibitor, Triacsin C, suppresses ferroptosis in Pfa1 cells. AA (2.5 μM,16 hrs); RSL3 (100 nM, 6 hrs); Triacsin C (2.5 μM, 6 hrs). Data are mean ± s.d., n=3. *P < 0.05 vs. control (t-test). (f) Contents of AA-CoA and AdA-CoA in WT and Acsl4 KO cells. Data are mean ± s.d., n=3. *P < 0.05 vs. WT cells (t-test). (g) Distribution of free and esterified PUFA-OOH in WT and Acsl4 KO Pfa1 cells treated with RSL3 (100 nM, 6 hrs).
Figure 2
Figure 2. Effects of exogenous AA on RSL3-triggered ferroptosis
(a) WT and Acsl4 KO cells were pretreated with AA (2.5 μM, 16 hrs) followed by RSL3 (100 nM, 6 hrs). Data are mean ± s.d., n=3; *,#,$ indicate P < 0.05 vs. control, RSL3 (or Acsl4 KO+RSL3), RSL3+AA, respectively (t-test). (b) LC-MS based heat maps for oxygenated esterified AA (C20:4) and AdA (C22:4) (normalized to corresponding WT group, in black), showing their relative changes after different treatments. (c) Schema of metabolic pathways for C20:4, C22:4 and their oxygenated products. (d) Levels of oxidized AA and AdA esterified into phospholipids in WT and Acsl4 KO Pfa1 cells. Data are mean ± s.e.m., n=3. *P < 0.05 vs. WT cells (one-tailed t-test). (e) Lpcat3 KD decreased RSL3-induced ferroptosis in MLE cells. Data are mean ± s.d., n=3. *P < 0.05 vs. WT cells (t-test). Insert: LPCAT3 decrease was confirmed by western blotting (48, 72 and 96 hrs) after Cre addition (for the full blot, see Supplementary Fig. 18a).
Figure 3
Figure 3. Screening of phospholipids and their oxidation products identifies ferroptosis death signals
(a) Representative normal phase LC-MS/MS chromatogram and mass spectra for six major classes of phospholipids in Pfa1 cells. (b) Scatter plot of RSL3 induced changes in the levels of oxygenated phospholipids (log2 (fold-change), X-axis) vs. significance (−log10 (p-Value), Y-axis, by t-test) of live (cell death <15%, number of replicate data points is 26) and ferroptotic cells (cell death >15%, number of replicate data points is 18). No oxygenated CLs were found (hence no oxygenated CL dots are present). (c) Re-categorized (based on the fatty acyls in sn-2 position) scatter plot of RSL3 induced changes in oxygenated phospholipids. Note that AA (20:4)- and AdA (22:4)-containing phospholipid species were most responsive. (d) Levels of di- and tri-oxygenated PE-(C18:0/C20:4) and PE-(C18:0/C22:4) in live (cell death <15%) and ferroptotic Pfa1 cells (cell death >15%).
Figure 4
Figure 4. Oxygenated PE species are identified in ferroptotic GPX4 deficient cells and kidney of GPX4-deficient mice
(a) Accumulation of oxygenated PE-(C18:0/C20:4) in WT and Gpx4 KO cells. Data are mean ± s.d., n=3. *P < 0.05 vs. WT Pfa1 cells (t-test). (b) Accumulation of oxygenated PE-(C18:0/C22:4) in WT and Gpx4 KO cells. Data are mean ± s.d., n=3. *P < 0.05 vs. WT Pfa1 cells (t-test). (c) LC-MS based heat maps for oxygenated esterified C20:4 and C22:4 from Gpx4 WT and KO Pfa1 cells. (d, e) Elevated levels of di- and tri-oxygenated species of PE-(C18:0/C20:4) and PE-(C18:0/C22:4) in kidney of GPX4-deficient mice. Data are mean ± s.d., n=3. *P < 0.05 vs. WT (t-test). (f) Ferroptosis inhibitor, liproxstatin-1, decreases the accumulation of di- and tri-oxygenated PE species in kidney of GPX4-deficient mice below the levels detected in WT animals. Data are mean ± s.d., n=3. *P < 0.05 vs. mice without liprostatin-1 (t-test).
Figure 5
Figure 5. Labeling with d8-AA unravels pathways leading to oxygenated di-acylated PE ferroptotic signals
(a, b) Accumulation of non-deuterated and deuterated (insert) oxygenated PE-(18:0/20:4) (a) and PE-(18:0/22:4) (b) in WT and Acsl4 KO cells treated with d8-AA (16 hrs) and after that exposed to RSL3 (100 nM, 6 hrs). Data are mean ± s.d., n=3. *P < 0.05 vs. WT Pfa1 cells (t-test). (c) Quantitative assessment of PE molecular species (C18:0/C20:4 and C18:0/d8-C20:4) (left panel) and (C18:0/22:4 and C18:0/d8-C22:4) (right panel) in WT and Acsl4 KO Pfa1 cells treated with d8-AA. Data are mean ± s.d., n=3. *P < 0.05 vs. WT (t-test). (d) 3D-representation of mass spectra of deuterated-PE from WT and Acsl4 KO cells treated with d8-AA. (e) Typical MS/MS spectrum illustrating fragmentation of non-oxygenated PE-(C18:0/C20:4) species as well as mono-, di-and tri-oxygenated species formed in RSL3 treated Pfa1 cells.
Figure 6
Figure 6. 15-LOX phospholipid oxidation product, 15-Hydroperoxy-SAPE, triggers ferroptosis in WT and Acsl4 KO Pfa1 cells
(a) Effects of AA-OOH (2.5 μM), AA-OOH-CoA (2.5 μM), and (b) SAPE-OOH on RSL3 (100 nM, 6 hrs) triggered ferroptosis in WT or Acsl4 KO cells. Without RSL3, SAPE-OOH (0.9 μM) did not induce ferroptosis (b, insert). LDH release levels: 8.1% (WT), 6.2% (WT+SAPE-OOH), 11% (Acsl4 KO), 10.6% (Acsl4 KO+SAPE-OOH). Data are mean±s.d., n=3; *,# indicate P < 0.05 vs. WT, Acsl4 KO, respectively (t-test). (c) LC-MS identification of 15-LOX induced SAPE oxidation products in cell lysates. Left panels: MS2 and MS3 (inserts) fragmentation of di- (m/z 798.5300) and tri-oxygenated (m/z 814.5262) SAPE. Right panels: formulas of oxygenated SAPE and detected fragments (red). Identified products were: 15-OOH-AA-PE, 15-OOH-8-OH-AA-PE, 15-OOH-9-OH-AA-PE, and 15-OOH-12-OH-AA-PE. Characteristic fragments formed: m/z 351 and m/z 333 - carboxylate anions of AA-OOH and AA-OOH minus H2O, respectively; m/z 113 - OOH at 15th carbon of AA; 176 and 155, 165 and 139, 203 - OH at 8th, 9th and 12th carbons of AA. (d) Inhibitors of LOX (but not COX or P450) protect WT cells against RSL3 (50 nM, 18 hrs) induced ferroptosis. Inhibitors: 12/15LOX-ML351 (10 μM), 15LOX-PD146176 (0.5 μM), COX-piroxicam (20 μM), P450-MSPPOH (10 μM), 5LOX-zileuton (10 μM), 12LOX-NCTT-956 (20 μM). Data are mean±s.d., n=3. *,$ indicate P < 0.05 vs. control, RSL3, respectively (t-test). (e) Etomoxir (β-oxidation inhibitor) enhances RSL3 (100 nM, 6 hrs) induced ferroptosis in WT cells. Data are mean±s.d., n=3. *P < 0.05 vs. RSL3 without etomoxir (t-test).
See this image and copyright information in PMC

References

    1. Allocati N, Masulli M, Di Ilio C, De Laurenzi V. Die for the community: an overview of programmed cell death in bacteria. Cell Death Dis. 2015;6:e1609. - PMC - PubMed
    1. Byrne JM, et al. Redox cycling of Fe(II) and Fe(III) in magnetite by Fe-metabolizing bacteria. Science. 2015;347:1473–6. - PubMed
    1. Dixon SJ, Stockwell BR. The role of iron and reactive oxygen species in cell death. Nat Chem Biol. 2014;10:9–17. - PubMed
    1. Dixon SJ, et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell. 2012;149:1060–72. - PMC - PubMed
    1. Yang WS, Stockwell BR. Ferroptosis: Death by Lipid Peroxidation. Trends Cell Biol. 2016;26:165–76. - PMC - PubMed

Reference for Online Methods

    1. Telmer CA, et al. Rapid, specific, no-wash, far-red fluorogen activation in subcellular compartments by targeted fluorogen activating proteins. ACS Chem Biol. 2015;10:1239–46. - PMC - PubMed
    1. Szent-Gyorgyi C, et al. Fluorogen-activating single-chain antibodies for imaging cell surface proteins. Nat Biotechnol. 2008;26:235–40. - PubMed
    1. Tardi PG, Mukherjee JJ, Choy PC. The quantitation of long-chain acyl-CoA in mammalian tissue. Lipids. 1992;27:65–7. - PubMed
    1. Minkler PE, Kerner J, Ingalls ST, Hoppel CL. Novel isolation procedure for short-, medium-, and long-chain acyl-coenzyme A esters from tissue. Anal Biochem. 2008;376:275–6. - PMC - PubMed
    1. Sun D, Cree MG, Wolfe RR. Quantification of the concentration and 13C tracer enrichment of long-chain fatty acyl-coenzyme A in muscle by liquid chromatography/mass spectrometry. Anal Biochem. 2006;349:87–95. - PubMed