Cell Biology
Ferroptosis
Iron-dependent regulated cell death — lipid peroxidation, GPX4 failure, system xc- collapse
Ferroptosis is a form of iron-dependent regulated cell death in which the polyunsaturated fatty acids of cellular membranes undergo runaway peroxidation until the membrane physically ruptures. Its central guardian is the selenoenzyme GPX4 (glutathione peroxidase 4), which uses glutathione to detoxify phospholipid hydroperoxides; glutathione, in turn, depends on cysteine imported through system xc- (the SLC7A11/SLC3A2 antiporter). When that defense collapses, ferrous iron drives Fenton chemistry and a self-propagating lipid radical chain. The death is caspase-independent and morphologically distinct from apoptosis. The term was coined in a 2012 Cell paper by Scott Dixon and Brent Stockwell, after the small molecule erastin was found to kill RAS-mutant cancer cells without triggering apoptosis — and it is now a leading target in cancer therapy, ischemia-reperfusion injury, and neurodegeneration.
- NamedDixon & Stockwell, 2012 (Cell)
- Master defenseGPX4 selenoenzyme
- Cysteine gatesystem xc- (SLC7A11/SLC3A2)
- Catalystlabile Fe²⁺ / Fenton chemistry
- Caspasesnone — caspase-independent
- Inhibitorsferrostatin-1, deferoxamine
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
Why ferroptosis matters
- A genuinely new way to die. For decades, cell death was sorted into apoptosis (regulated, silent) versus necrosis (accidental, messy). Ferroptosis broke that binary: it is regulated and genetically tractable, yet uses no caspases, no BAX/BAK pore, and no dedicated executioner protein — the killing is done by iron-catalyzed lipid chemistry. That reframing, formalized in 2012, opened an entirely new druggable axis of cell death.
- Cancer's Achilles' heel. Therapy-resistant, mesenchymal, and dedifferentiated tumor states carry a heavy load of oxidizable polyunsaturated lipids and become addicted to GPX4 and system xc- to survive it. Drug-tolerant "persister" cells that ride out targeted therapy are strikingly ferroptosis-sensitive, making ferroptosis induction a way to clear the very cells that seed relapse.
- Ischemia-reperfusion injury. When blood flow returns to an oxygen-starved, iron-loaded tissue after a stroke or heart attack, the reoxygenation burst drives massive lipid peroxidation. Ferrostatins, liproxstatins, and iron chelators shrink infarct size in animal models of stroke, myocardial infarction, and acute kidney injury — ferroptosis inhibition is a candidate organ-protective therapy.
- Neurodegeneration. Neurons are PUFA-rich, accumulate iron with age, and run low on antioxidant reserves. Iron deposition in the substantia nigra in Parkinson's disease, glutathione depletion, and lipid-peroxidation adducts like 4-hydroxynonenal all point to a ferroptotic component in Parkinson's, Alzheimer's, and Huntington's disease.
- Immunity and tumor surveillance. CD8+ T cells activated during immunotherapy secrete interferon-gamma, which downregulates SLC7A11 in tumor cells and pushes them toward ferroptosis. This wires ferroptosis directly into the anti-tumor immune response and offers a rationale for combining ferroptosis inducers with checkpoint blockade.
- A conserved vulnerability of lipid membranes. Because ferroptosis is ultimately membrane chemistry, it is broadly conserved and matters wherever cells store polyunsaturated lipids — from plant heat-stress death to kidney tubule injury. It is the price a cell pays for building flexible, fluid membranes out of oxidizable fatty acids.
Common misconceptions
- "Ferroptosis is just oxidative-stress necrosis." It looks chaotic, but it is exquisitely regulated. Specific genes and enzymes set the threshold: ACSL4 and LPCAT3 load PUFAs into membranes, ALOX and POR seed peroxides, and GPX4, FSP1, and system xc- restrain the death. Knock out or drug any one node and you shift the whole cell's ferroptosis sensitivity in a predictable direction.
- "Any iron overload causes ferroptosis." It is specifically the labile (chelatable) ferrous iron pool that matters, not iron locked in ferritin or heme. Ferritinophagy — NCOA4-mediated autophagic release of iron from ferritin — sensitizes cells, while ferritin overexpression or the exporter ferroportin protects them. The location and redox state of iron is everything.
- "GPX4 and glutathione are the only defense." A parallel, glutathione-independent system exists. FSP1 (ferroptosis suppressor protein 1, formerly AIFM2) sits at the membrane and regenerates reduced ubiquinol (CoQ10-H2), a lipophilic radical-trapping antioxidant that halts the peroxidation chain. GCH1/BH4 and DHODH provide further parallel shields. That is why some GPX4-low cells still resist ferroptosis.
- "Ferroptosis is always bad and should always be blocked." In cancer it is a therapeutic asset, not a liability — the goal there is to trigger ferroptosis, not prevent it. Whether you want to induce or inhibit depends entirely on context: induce in tumors, inhibit in stroke, heart attack, and neurodegeneration.
- "Mitochondria are irrelevant, since this is a membrane death." Mitochondria are a hallmark of ferroptotic morphology — they shrink, their cristae condense, and outer-membrane density increases. Mitochondrial metabolism (TCA-cycle activity, membrane potential, and lipid supply) actively contributes to cysteine-deprivation ferroptosis, and DHODH on the inner mitochondrial membrane is a defense node.
- "Ferroptosis was invented in 2012 out of nowhere." The chemistry was seen much earlier under other names — the "oxytosis" of glutamate-treated neurons and glutathione-depletion death studied in the 1980s–90s were the same process. The 2012 work unified those scattered observations under one mechanism and one name.
How ferroptosis works
Ferroptosis is a race between two lipid-membrane processes: the loading of oxidizable fuel and its detoxification. On the fuel side, the enzyme ACSL4 (acyl-CoA synthetase long-chain family member 4) selectively activates polyunsaturated fatty acids — chiefly arachidonic and adrenic acid — and LPCAT3 esterifies them into membrane phospholipids, especially phosphatidylethanolamines. These PUFA tails are the substrate for peroxidation: their bis-allylic hydrogens are easy to abstract, and once a radical forms it propagates down the chain. Lipoxygenases (notably ALOX15) and cytochrome-P450 reductase (POR) seed the first phospholipid hydroperoxides, and the labile ferrous iron pool amplifies them through Fenton chemistry, generating lipid-alkoxyl and hydroxyl radicals that abstract more hydrogens and spawn more peroxides in a self-sustaining chain reaction.
On the defense side stands GPX4 (glutathione peroxidase 4), the only mammalian enzyme that can reduce phospholipid hydroperoxides already embedded in the membrane back to harmless lipid alcohols. GPX4 is a selenoprotein — its catalytic selenocysteine resists over-oxidation — and it consumes two molecules of reduced glutathione per turnover. Glutathione (a cysteine-glycine-glutamate tripeptide) is synthesized from cysteine, and the rate-limiting cysteine supply arrives as cystine through system xc-, the SLC7A11/SLC3A2 antiporter that trades one intracellular glutamate for one extracellular cystine. So the defensive axis runs cystine → cysteine → glutathione → GPX4 → detoxified lipid peroxides. Break any link — inhibit system xc- (erastin), deplete glutathione, or inactivate GPX4 (RSL3, ML162) — and peroxides accumulate.
When defense loses the race, phospholipid hydroperoxides pile up faster than they can be reduced. The peroxidation front sweeps across the plasma membrane, degrading the lipid packing until pores form and the membrane loses integrity — the cell swells slightly and ruptures, spilling oxidized lipids and damage-associated molecular patterns that can inflame surrounding tissue. A parallel glutathione-independent shield can buy time: FSP1 (AIFM2) regenerates ubiquinol at the membrane to trap radicals, and GCH1-derived tetrahydrobiopterin (BH4) and mitochondrial DHODH provide further antioxidant capacity. Pharmacologically, the death is stopped cold at two chemical choke points — lipophilic radical-trapping antioxidants such as ferrostatin-1 and liproxstatin-1 quench the propagating radicals, and iron chelators such as deferoxamine remove the catalyst. Caspase inhibitors, by contrast, do nothing.
Ferroptosis vs apoptosis vs necroptosis vs pyroptosis
| Feature | Ferroptosis | Apoptosis | Necroptosis | Pyroptosis |
|---|---|---|---|---|
| Core trigger | Iron-dependent lipid peroxidation | BH3-only / death receptors | RIPK1/RIPK3 on caspase-8 block | Inflammasomes (NLRP3, AIM2) |
| Executioner | None — lipid radical chemistry | Caspases-3/7/8/9 | MLKL pore | Caspase-1, gasdermin D |
| Key defense | GPX4, system xc-, FSP1 | BCL-2, BCL-xL, MCL-1 | Caspase-8, FLIP | — |
| Membrane fate | Peroxidation-driven rupture | Intact, blebs outward | MLKL pores, swells, lyses | Gasdermin pores, lyses |
| Nucleus / DNA | Largely normal, no laddering | Condensed, 180-bp laddering | Random degradation | Random degradation |
| Mitochondria | Shrunken, condensed cristae | Cyt c release, MOMP | Often intact early | Often intact early |
| Inflammation | Pro-inflammatory (oxidized lipids, DAMPs) | Silent (anti-inflammatory) | High (DAMPs) | Very high (IL-1β, IL-18) |
| Blocked by | Ferrostatin-1, deferoxamine, vit E | zVAD-fmk, BCL-2 | Necrostatin-1 (RIPK1) | MCC950 (NLRP3) |
Ferroptosis inducers vs inhibitors
| Agent | Class | Molecular target | Net effect |
|---|---|---|---|
| Erastin | Inducer (class 1) | system xc- (SLC7A11) | Cysteine starvation → GSH loss |
| Sulfasalazine / sorafenib | Inducer (class 1) | system xc- (SLC7A11) | Cysteine starvation → GSH loss |
| RSL3 / ML162 | Inducer (class 2) | GPX4 selenocysteine (covalent) | Peroxides accumulate directly |
| FIN56 | Inducer (class 3) | GPX4 degradation + CoQ10 depletion | Loss of two defense arms |
| Ferrostatin-1 / liproxstatin-1 | Inhibitor | Propagating lipid radicals (RTA) | Chain-breaking antioxidant |
| Deferoxamine | Inhibitor | Labile ferrous iron | Removes the Fenton catalyst |
| Vitamin E (α-tocopherol) | Inhibitor | Lipid peroxyl radicals | Endogenous chain-breaking RTA |
| Rosiglitazone / thiazolidinediones | Inhibitor | ACSL4 | Fewer PUFAs esterified into membranes |
Famous experiments
- Erastin discovery (Dolma, Stockwell 2003). A synthetic-lethal screen for compounds that kill engineered RAS-mutant cells identified erastin, which killed without the nuclear morphology of apoptosis and could not be rescued by caspase inhibitors. It was the first hint of a distinct, oxidative, iron-linked death mode.
- Naming ferroptosis (Dixon, Stockwell 2012, Cell). Scott Dixon and colleagues showed erastin-induced death is iron-dependent, caspase-independent, rescued by iron chelators and the newly characterized antioxidant ferrostatin-1, and traced erastin's target to system xc-. They coined the term "ferroptosis" — iron (ferrum) plus death.
- GPX4 as the hub (Yang & Stockwell 2014; Conrad lab). Wan Seok Yang showed that RSL3 kills by directly inhibiting GPX4, placing the selenoenzyme at the center of the pathway. Independently, Marcus Conrad's inducible Gpx4-knockout mice died of acute ferroptotic tissue degeneration — including rapid neurodegeneration — proving GPX4 is the essential in vivo guardian.
- ACSL4 sets the substrate (Doll, Conrad 2017; Kagan 2017). Genome-wide screens and lipidomics identified ACSL4 as a key determinant of ferroptosis sensitivity: it enriches membranes with the oxidizable PUFAs (arachidonoyl- and adrenoyl-phosphatidylethanolamines) that become the lethal peroxidation substrate. Losing ACSL4 makes cells ferroptosis-resistant.
- FSP1–CoQ10 parallel axis (Bersuker/Doll 2019, Nature). Two back-to-back papers showed FSP1 (AIFM2) suppresses ferroptosis independently of glutathione by regenerating ubiquinol, a membrane radical-trapping antioxidant. This explained why some GPX4-inhibited cells survive and revealed a second, druggable defense arm.
- Immunity link (Wang, Zou, Green 2019). CD8+ T-cell-derived interferon-gamma was shown to repress SLC7A11 and promote lipid peroxidation and ferroptosis in tumor cells, tying ferroptosis to the efficacy of immune-checkpoint therapy and anti-tumor immunity.
Frequently asked questions
How is ferroptosis different from apoptosis?
Ferroptosis and apoptosis are mechanistically and morphologically distinct. Apoptosis is executed by caspase proteases, requires ATP, condenses chromatin, and blebs the membrane into apoptotic bodies that are cleared silently — no membrane oxidation, no iron. Ferroptosis uses no caspases and no dedicated executioner protein at all: it is a chemistry-driven death in which ferrous iron catalyzes the peroxidation of polyunsaturated phospholipids until the membrane loses integrity. Under electron microscopy, ferroptotic cells show shrunken mitochondria with condensed, ruptured cristae and increased outer-membrane density, while the nucleus stays normal in size with no chromatin condensation — the opposite of the apoptotic picture. Caspase inhibitors like zVAD-fmk do not block ferroptosis; the lipophilic antioxidant ferrostatin-1 and the iron chelator deferoxamine do. Because the ruptured membrane spills oxidized lipids and DAMPs, ferroptosis is generally pro-inflammatory, whereas apoptosis is immunologically quiet.
What is the role of GPX4 in ferroptosis?
GPX4 (glutathione peroxidase 4) is the master repressor of ferroptosis. It is the only mammalian enzyme that can reduce phospholipid hydroperoxides — lipid peroxides already esterified within the membrane — back to non-reactive lipid alcohols, using two molecules of reduced glutathione as the electron donor. GPX4 is a selenoenzyme: its active site contains selenocysteine rather than cysteine, which makes it resistant to over-oxidation and irreplaceable. When GPX4 is inhibited (by RSL3 or ML162, which covalently target its selenocysteine) or genetically deleted, phospholipid hydroperoxides accumulate unchecked, iron-catalyzed radical chemistry propagates, and the cell dies by ferroptosis. Whole-body Gpx4 knockout in mice is embryonic lethal, and inducible or tissue-specific deletion causes acute ferroptotic degeneration — for example, neuronal Gpx4 loss produces rapid motor-neuron death and paralysis.
What is system xc- and why does erastin inhibit it?
System xc- is a plasma-membrane antiporter built from the catalytic subunit SLC7A11 (xCT) and the chaperone subunit SLC3A2 (4F2hc). It imports one cystine into the cell while exporting one glutamate, in a 1:1 exchange. Imported cystine is reduced to cysteine, the rate-limiting building block for glutathione synthesis. Glutathione in turn is the fuel GPX4 needs to neutralize lipid peroxides. Erastin — the small molecule that first defined ferroptosis — directly inhibits system xc-, starving the cell of cysteine. Glutathione levels crash, GPX4 loses its cofactor, lipid peroxides accumulate, and ferroptosis follows. This is why high extracellular glutamate (which competitively blocks cystine uptake) can be neurotoxic, and why the classic 'oxytosis' described in neurons decades earlier turned out to be the same pathway as ferroptosis.
Why is iron essential for ferroptosis?
Iron is the catalytic engine of ferroptosis, which is why iron chelators such as deferoxamine block the death entirely. Ferrous iron (Fe2+) drives Fenton chemistry, reacting with hydrogen peroxide and lipid hydroperoxides to generate hydroxyl and lipid-alkoxyl radicals that abstract hydrogen atoms from adjacent polyunsaturated fatty acids, initiating and propagating the peroxidation chain. Iron is also a required cofactor for the lipoxygenases (especially ALOX15) and for POR-driven electron transfer that help seed the initial lipid peroxides. Anything that raises the labile iron pool — increased transferrin-receptor uptake, ferritinophagy (NCOA4-mediated autophagic degradation of ferritin), or loss of the exporter ferroportin — sensitizes cells to ferroptosis. Anything that lowers free iron, from chelators to ferritin overexpression, protects them.
How was ferroptosis discovered?
Ferroptosis emerged from a search for compounds that selectively kill RAS-mutant cancer cells. In 2003, Brent Stockwell's lab identified erastin in a synthetic-lethality screen, and later found RSL3 (RAS-selective lethal 3). These molecules killed cells in a way that could not be blocked by caspase inhibitors, apoptosis blockers, or necroptosis inhibitors, but was fully rescued by iron chelators and lipophilic antioxidants. In a landmark 2012 Cell paper, Scott Dixon, Stockwell, and colleagues named this iron-dependent, non-apoptotic death 'ferroptosis' and traced erastin's target to system xc-. In 2014, Wan Seok Yang and Stockwell identified GPX4 as the central enzyme whose inhibition triggers the death, unifying the pathway. Marcus Conrad's independent Gpx4-knockout mouse work confirmed GPX4 as the essential in vivo guardian.
Can ferroptosis be exploited to treat cancer?
Yes — ferroptosis is one of the most active frontiers in oncology. Therapy-resistant, dedifferentiated, and mesenchymal-state cancer cells are often exquisitely dependent on GPX4 and system xc- to survive their high load of oxidizable membrane lipids, making them vulnerable to ferroptosis inducers even when they resist apoptosis-based chemotherapy. Drug-tolerant 'persister' cells that survive targeted therapy are particularly ferroptosis-sensitive. Strategies include GPX4 inhibitors, system xc- blockers, and cyst(e)inase enzymes that deplete circulating cysteine. Immunotherapy links in too: CD8+ T cells release interferon-gamma that downregulates SLC7A11 and drives tumor-cell ferroptosis, so ferroptosis induction can synergize with checkpoint blockade. The main hurdles are delivering ferroptosis inducers selectively to tumors and avoiding collateral damage to normal iron-rich tissues.
What role does ferroptosis play in neurodegeneration and stroke?
The brain is unusually vulnerable to ferroptosis: neurons are rich in polyunsaturated fatty acids, accumulate iron with age, and have relatively low antioxidant reserves. Ferroptotic features — iron accumulation, glutathione depletion, and lipid peroxidation markers such as 4-hydroxynonenal — are documented in Parkinson's disease (where iron builds up in the substantia nigra), Alzheimer's disease, Huntington's disease, and periventricular leukomalacia. In ischemia-reperfusion injury after stroke or heart attack, the sudden return of oxygen to iron-loaded tissue drives a burst of lipid peroxidation, and ferrostatins, liproxstatins, and iron chelators reduce infarct size in animal models. This has made ferroptosis inhibition an appealing neuroprotective and cardioprotective strategy, with several lipophilic radical-trapping antioxidants and GPX4 stabilizers in preclinical and early clinical development.