GPX4: Its Critical Role in Cell Health and Disease

GPX4, or glutathione peroxidase 4, is the only enzyme in the human body that can directly neutralize oxidized fats embedded in cell membranes, and its loss triggers a distinctive form of cell death called ferroptosis. That makes it one of the most consequential single proteins in mammalian biology. Mice engineered to lack GPX4 die in the womb by midgestation, and adult mice stripped of it die within two weeks. From brain neurons to sperm cells to tumor-fighting immune cells, GPX4 acts as a last line of defense against a chain reaction that, left unchecked, tears membranes apart. Its involvement in cancer, neurodegeneration, heart disease, and male infertility has turned it into one of the most actively studied drug targets of the past decade.

What GPX4 Actually Does

Cell membranes are built from fatty molecules called phospholipids. When iron-driven chemistry or other oxidative stress hits those fats, they become lipid hydroperoxides, essentially damaged fats that can propagate further damage in a chain reaction. Most antioxidant enzymes cannot touch these hydroperoxides once they are already lodged in a membrane. GPX4 is the exception. It reduces those bulky, membrane-bound hydroperoxides directly back into harmless alcohols, stopping the chain reaction before it spreads.

Structurally, GPX4 is unusual among human glutathione peroxidases. Crystal structure data show it works as a single-unit protein rather than forming multi-unit complexes the way other family members do. Its active site contains selenocysteine, a rare amino acid that incorporates the trace element selenium, and x-ray data suggest the catalytic cycle involves three distinct chemical states of that selenium atom rather than the two states long assumed for other peroxidases.¹ That reliance on selenium has practical consequences: when selenium levels in the body drop, GPX4 production falters, a point we will return to later.

Three Versions for Three Compartments

The GPX4 gene produces three protein isoforms that end up in different parts of the cell. Cytoplasmic GPX4 is the workhorse for general membrane protection and the primary brake on ferroptosis. Mitochondrial GPX4 carries a targeting signal that routes it into mitochondria, where it guards the inner mitochondrial membrane against oxidative damage and helps maintain energy production. Nuclear GPX4 protects the lipids and genetic material inside the nucleus, preserving chromatin structure and genetic integrity.²

The nuclear isoform has a surprising extra job: it plays a role in sperm nuclear maturation. That dual life, where one protein moonlights between antioxidant defense and structural biology, hints at how deeply embedded GPX4 is in basic cellular operations.

How GPX4 Prevents Ferroptosis

Ferroptosis is a form of regulated cell death driven specifically by the accumulation of lipid hydroperoxides. Unlike apoptosis, the orderly self-destruction most people have heard of, ferroptosis does not depend on the usual executioner enzymes. Instead, it is essentially death by membrane corrosion, driven by iron-catalyzed lipid damage that overwhelms the cell’s repair capacity. GPX4 is the central gatekeeper: ferroptosis occurs when GPX4 activity is lost or overwhelmed.³

GPX4 does not work alone. It depends on an upstream supply chain. A membrane transporter called System Xc⁻ imports cystine into the cell, which gets converted into cysteine, a building block of glutathione (GSH). GSH is the fuel GPX4 burns to do its work: each catalytic cycle uses GSH to reduce a lipid hydroperoxide. Block any step in this System Xc⁻/GSH/GPX4 pathway and you cut off ferroptosis defense.⁴ That is why some ferroptosis-inducing drugs target the transporter upstream rather than GPX4 itself.

What Happens When GPX4 Disappears

The clearest evidence for GPX4’s importance comes from knockout experiments in mice. Complete deletion is lethal before birth: embryos lacking GPX4 die around day 7.5 of gestation and fail to form normal tissue structures.⁵ That alone tells you the protein is not optional.

What about adults? Researchers generated mice whose GPX4 gene could be switched off after they had already matured. Adult mice that lost roughly 75 to 85 percent of their GPX4 across major organs lost weight and died within two weeks. Their mitochondria showed clear damage, their livers produced less energy, and neurons in the hippocampus region of the brain began dying off, accompanied by an increase in astrocyte activity, a classic sign of brain injury.⁶ These findings confirmed that GPX4 is not just critical for embryonic development but remains essential throughout adult life.

Intriguingly, feeding high-dose vitamin E to mice with tissue-specific GPX4 deletions partially rescued damage in several organs, including the testes, liver, and heart.⁷ Vitamin E is a lipid-soluble antioxidant that can intercept some of the same chain reactions GPX4 handles, so this makes mechanistic sense. It also suggests that in situations where GPX4 is compromised rather than absent, boosting other antioxidant pathways could offer partial protection.

GPX4 and Male Fertility

GPX4 has an unusual double life in sperm cells. During spermatogenesis, it works as an enzyme, detoxifying lipid hydroperoxides. But as sperm mature, GPX4 undergoes a transformation: it cross-links with structural proteins in the mitochondrial capsule of the sperm midpiece, essentially becoming a building material. It oxidizes protein thiol groups and then becomes physically embedded in the capsule structure.⁸ This moonlighting function makes GPX4 the major selenoprotein in sperm, involved in hydroperoxide detoxification, mitochondrial capsule assembly, and chromatin compaction.⁹

The practical upshot is that anything compromising GPX4 activity, whether a genetic variant, selenium deficiency, or environmental toxin, could impair sperm function and contribute to male infertility. Researchers have begun investigating genetic variations in the GPX4 gene as potential contributors to unexplained infertility in men.

Neurodegeneration and Motor Neuron Disease

Neurons are especially vulnerable to ferroptosis. They are rich in the polyunsaturated fatty acids that serve as substrates for lipid peroxidation, and they have limited capacity to regenerate. When researchers conditionally deleted GPX4 from neurons in adult mice, the animals developed rapid-onset paralysis. Examination of their spinal cords revealed dramatic motor neuron degeneration, with hallmarks of ferroptosis rather than apoptosis: no activation of the apoptotic executioner enzyme caspase-3, no DNA fragmentation typical of apoptosis, but elevated inflammatory markers and activation of specific signaling pathways associated with ferroptotic death.¹⁰ Notably, the cortex was relatively spared while spinal motor neurons bore the brunt of the damage, suggesting different neuron types have different vulnerability thresholds.

Growing evidence links GPX4 dysfunction to human neurodegenerative conditions. Dysregulation of GPX4-mediated antioxidant defense has been implicated in Alzheimer’s disease, Parkinson’s disease, and other conditions where oxidative stress and iron imbalance play a role.¹¹ This is still primarily preclinical work, but the connection between iron accumulation in aging brains, lipid peroxidation, and GPX4 decline is attracting serious therapeutic interest.

Cancer and the Ferroptosis Vulnerability

If GPX4 keeps normal cells alive, cancer cells often depend on it even more. Tumor cells frequently face higher oxidative stress due to their chaotic metabolism and rapid proliferation. Some cancer cells adopt a therapy-resistant state, sometimes called a mesenchymal or drug-tolerant persister state, in which they become particularly reliant on GPX4 to survive. Researchers have found that these resistant cancer cells are selectively vulnerable to GPX4 inhibition.¹² In other words, the very trait that makes them hard to kill with conventional chemotherapy exposes them to ferroptosis-based attack.

This insight has driven a rush to develop GPX4 inhibitors. The best-known laboratory tools, RSL3 and ML162, were long assumed to work by binding directly to GPX4’s active-site selenocysteine. But crystal structure studies revealed a surprise: both compounds actually bind to a different site on the protein, cysteine 66, triggering inhibition and degradation of GPX4 through an allosteric mechanism.¹³ Newer inhibitor designs have taken a different approach, with at least one compound shown to bind the active-site selenocysteine directly, opening up multiple strategies for shutting GPX4 down in tumor cells.¹⁴

Another strategy avoids direct binding altogether. The compound FIN56 triggers GPX4 degradation through an autophagy-dependent mechanism, essentially tricking the cell’s recycling machinery into digesting its own ferroptosis shield. Researchers showed that blocking autophagy prevented FIN56 from destroying GPX4, confirming autophagy as the degradation route.¹⁵ These indirect approaches could be valuable because they sidestep the challenge of designing a drug that binds a small, somewhat featureless active site.

The Immune System Needs GPX4 Too

Here is a complication for the cancer-therapy story: immune cells also need GPX4 to function. Both CD4⁺ and CD8⁺ T cells lacking GPX4 failed to expand in response to infection and could not protect mice against viral or parasitic challenge. High-dose vitamin E supplementation rescued this immune defect, confirming that the failure was driven by lipid peroxidation rather than some other consequence of losing the enzyme.¹⁶

This creates a therapeutic tension. Inhibiting GPX4 in tumors could kill therapy-resistant cancer cells, but it might simultaneously cripple the immune cells needed to fight that same cancer. One creative workaround involves an oncolytic virus engineered to deliver GPX4 directly to the tumor microenvironment. In a pancreatic cancer model, a vaccinia virus carrying the GPX4 gene replicated inside tumor cells and promoted GPX4 expression in nearby T cells. The result was a reshaped immune environment with less T cell exhaustion and improved responses to immunotherapy.¹⁷ The strategy essentially arms immune cells while the virus kills tumor cells, threading the needle between pro-ferroptosis and anti-ferroptosis approaches.

Cardiovascular Disease and Ischemia-Reperfusion Injury

Atherosclerosis, the buildup of fatty plaques in artery walls, is fundamentally a disease of oxidized lipids. Mice bred to develop atherosclerosis showed significantly smaller lesions when they overexpressed GPX4, with fewer advanced plaque features like fibrous caps and dead-cell zones. The protective effect tracked closely with reduced levels of a lipid peroxidation marker in the aorta.¹⁸ GPX4 also reduced the downstream effects of oxidized lipids on blood vessel cells, including the inflammatory adhesion of immune cells to vessel walls that drives plaque growth.

Estrogen appears to protect blood vessels in part through GPX4. In a mouse model of postmenopausal atherosclerosis, loss of estrogen accelerated plaque development by driving endothelial cell ferroptosis through a pathway that suppresses GPX4 expression. Restoring estrogen or directly blocking ferroptosis both reversed the effect and restored GPX4 levels in endothelial cells.¹⁹

When blood flow to an organ is interrupted and then restored, as happens during a heart attack or organ transplant, the returning blood paradoxically causes a burst of oxidative damage called ischemia-reperfusion injury. Ferroptosis has emerged as a key player in this damage. In myocardial ischemia-reperfusion, a signaling protein called STING was shown to promote GPX4 degradation through autophagy, worsening heart damage. When researchers delivered GPX4 to the heart using a viral vector, it counteracted this degradation and improved cardiac recovery.²⁰ Similar dynamics have been observed in the kidney, where the protein FBXW7 binds GPX4 and drives its downregulation, worsening acute kidney injury after ischemia-reperfusion events.²¹

GPX4 Is Not the Only Line of Defense

For years, GPX4 looked like the sole enzymatic brake on ferroptosis. Then in 2019, researchers discovered that a separate enzyme called FSP1 (ferroptosis suppressor protein 1) operates in parallel. FSP1 works outside the mitochondria, using coenzyme Q10 as its substrate to trap lipid radicals, entirely independent of glutathione.²² This explained a long-standing puzzle: why some cell types can survive partial GPX4 loss better than others. Cells with high FSP1 expression have a backup system; those without it are left fully dependent on GPX4.

The membrane lipid composition also matters. The enzyme ACSL4 channels polyunsaturated fatty acids into membrane phospholipids, creating more of the substrates that become dangerous when oxidized. In lung ischemia-reperfusion models, inhibiting ACSL4 reduced lipid peroxidation and boosted both glutathione and GPX4 levels.²³ Ferroptosis sensitivity, in other words, is a balance between the pro-ferroptotic lipid supply and the anti-ferroptotic enzymatic defenses. GPX4 is the most important piece on the defense side, but it is embedded in a larger network.

The Selenium Bottleneck

GPX4’s dependence on selenocysteine creates a nutritional bottleneck. The selenocysteine codon (UGA) is normally a stop signal in genetic code; incorporating selenium at that position requires specialized cellular machinery and adequate selenium availability. When selenium levels are low, ribosomes stall at the GPX4 selenocysteine codon, leading to collisions between stacked ribosomes, premature termination, and disposal of the incomplete protein fragment.²⁴ The result is a steep drop in functional GPX4 protein, leaving the cell vulnerable to ferroptosis.

This connection between dietary selenium and ferroptosis resistance is not just academic. In regions with selenium-poor soils, populations may have chronically lower GPX4 levels, a factor that could influence susceptibility to the diseases discussed throughout this article. The link also provides a rationale for selenium supplementation trials in contexts where ferroptosis-related damage is expected, though translating animal findings to clinical recommendations remains a work in progress.

Fine-Tuning GPX4 After It Is Made

Beyond the selenium supply question, the cell has multiple ways to adjust GPX4 levels after the protein is already built. GPX4 undergoes post-translational modifications including ubiquitination (tagging it for disposal), phosphorylation, and other chemical additions that alter how stable and active the protein is.²⁵ These modifications offer potential therapeutic leverage: if you could block the specific ubiquitin tag that marks GPX4 for destruction during a heart attack, for instance, you might preserve more GPX4 during the critical window of ischemia-reperfusion injury.

Environmental Toxins and GPX4 Depletion

Cadmium, a toxic heavy metal found in industrial waste, cigarette smoke, and contaminated food, causes organ damage partly by depleting GPX4. In sheep exposed to cadmium chloride, kidney damage was accompanied by drops in GPX4 protein levels and gene expression, along with the classic signs of ferroptosis: glutathione depletion, elevated lipid peroxidation, and impaired antioxidant enzyme activity. Supplementing with selenium restored GPX4 levels and reduced the damage.²⁶ A similar pattern was observed in the liver, where cadmium exposure downregulated GPX4 and selenium nanoparticles counteracted the toxicity by inhibiting ferroptosis.²⁷

These findings reinforce the selenium-GPX4 relationship from a different angle: selenium does not just help build GPX4 in the first place, it also appears to protect GPX4 function against chemical assault. For people with occupational or environmental cadmium exposure, selenium status may be a meaningful variable in how much organ damage accumulates over time.

Viruses That Exploit GPX4 Suppression

Some pathogens appear to manipulate GPX4 to their own advantage. Human cytomegalovirus (HCMV), a common virus that infects most people at some point, actively downregulates GPX4 during infection and drives up lipid peroxidation in host cells. Yet the infection does not fully trigger ferroptosis, as if the virus pushes GPX4 low enough to remodel the cell’s lipid landscape without crossing the lethal threshold. When researchers artificially pushed GPX4 even lower, virus replication suffered, suggesting the virus benefits from a specific, carefully calibrated level of GPX4 suppression.²⁸ It is an interesting example of a pathogen walking a metabolic tightrope, exploiting the host’s antioxidant defenses without destroying the cell it needs to inhabit.