The research team, based at the Korea Institute of Radiological and Medical Sciences, set out to examine how metformin and phenformin influence ferroptosis triggered by cysteine deprivation, a metabolic stress that strips cancer cells of a raw material they desperately need. Cysteine is the precursor for glutathione, the cell’s master antioxidant, and also supplies sulfur for iron-sulfur cluster biogenesis. When cysteine runs out, glutathione synthesis collapses, the antioxidant enzyme glutathione peroxidase 4 loses support, and membranes become vulnerable to runaway lipid peroxidation, the biochemical signature of ferroptosis.

Working with H1299 human non-small cell lung cancer cells and HEYA8 ovarian cancer cells, the researchers cultured the cells in custom-made cysteine-free medium and then added metformin or phenformin at various concentrations. The results were unambiguous. Both drugs dose-dependently rescued cell viability that would otherwise have plummeted, and propidium iodide staining confirmed that the proportion of dead cells dropped sharply. Neither drug alone killed more than about ten percent of cells at the concentrations tested, yet in the cysteine-starved environment they acted as powerful survival shields.

To confirm that the cell death being blocked was genuinely ferroptotic, the team deployed a panel of death-pathway inhibitors. Ferrostatin-1 and liproxstatin-1, canonical ferroptosis inhibitors, strongly restored viability, as did deferoxamine, an iron chelator, consistent with iron-driven lipid peroxidation as the dominant death mechanism. The apoptosis inhibitor z-VAD-fmk had no effect. Intriguingly, necrostatin-1, a necroptosis inhibitor, also rescued viability and reduced lipid peroxidation, suggesting that cysteine deprivation kills cells through a hybrid program in which ferroptosis and necroptotic signaling overlap, with lipid peroxidation serving as a common mediator.

Flow cytometry with the fluorescent lipid peroxidation probe C11-BODIPY 581/591 revealed that metformin and phenformin dramatically suppressed the peroxidation wave triggered by cysteine withdrawal, and the mitochondrial probe MitoPeDPP showed that the protection extended to the mitochondrial membranes themselves. Because reactive oxygen species drive lipid peroxidation, the researchers next measured ROS with DCFH-DA and mitochondrial superoxide with MitoSOX. Cysteine deprivation sent both soaring; both drugs pulled the levels back down, even though neither drug altered ROS on its own.

Here is where the story becomes mechanistically fascinating. The canonical view of ferroptosis centers on the glutathione-glutathione peroxidase 4 axis: no cysteine means no glutathione, and no glutathione means an undefended membrane. Yet metformin did not restore glutathione levels, did not change the ratio of oxidized to reduced glutathione, and did not rescue glutathione peroxidase 4 protein expression. Its protective effect was entirely independent of the antioxidant system that ferroptosis researchers usually target. Something else was going on, and the trail led to the mitochondria.

Using FerroOrange and Mito-FerroGreen probes, the team found that cysteine deprivation caused ferrous iron to pile up both in the cell overall and specifically inside mitochondria, and that metformin and phenformin blunted both accumulations. This matters because cysteine starvation disrupts iron-sulfur cluster biogenesis, destabilizing cellular iron homeostasis and raising the pool of labile ferrous iron that catalyzes lipid-damaging radical reactions. Notably, rotenone, a direct inhibitor of mitochondrial Complex I, and CCCP, which dissipates the mitochondrial membrane potential, similarly reduced mitochondrial iron accumulation, implicating mitochondrial bioenergetics in the iron dysregulation.

The mitochondrial connection deepened when the researchers measured membrane potential with TMRE staining and oxygen consumption with an extracellular oxygen probe. Within just two hours of cysteine deprivation, before any substantial cell loss, mitochondria became hyperpolarized and oxygen consumption spiked, exactly the alterations previously shown to be required for this form of ferroptosis. Metformin and phenformin, both established Complex I inhibitors, suppressed the hyperpolarization and curbed the excess oxygen consumption. Ferrostatin-1 also attenuated the hyperpolarization and mitochondrial ROS, reinforcing that these mitochondrial changes are not bystanders but active participants in the death program.

The authors are careful about what these findings do and do not mean clinically. The concentrations of metformin that blocked ferroptosis in this study, in the millimolar range, far exceed the roughly 10 to 40 micromolar plasma levels achieved in patients, and micromolar metformin showed no detectable protection. The results should therefore be read primarily as mechanistic evidence that tuning mitochondrial bioenergetic activity can suppress cysteine deprivation-induced ferroptosis, not as a reason to expect metformin alone to shield tumors at therapeutic doses. They also note that other lung cancer cell lines, including A549, H460, H23, and H820, were far less sensitive to cysteine deprivation in the first place, underscoring how cell-type-specific these responses are. Adding another layer of complexity, prior studies have reported that metformin induces or enhances ferroptosis in lung, breast, and liver cancer models through pathways such as NRF2/HO-1 suppression and SLC7A11 destabilization, and has even shown anti-ferroptotic protection in diabetic osteoporosis and pancreatic beta-cell injury, making metformin’s role in ferroptosis regulation unmistakably context-dependent.

The broader implication is a caution flag for the ferroptosis field. As pharmaceutical interest surges in therapies that starve tumors of cysteine or block the SLC7A11/xCT cystine importer, the new data suggest that biguanides, and perhaps other mitochondrial modulators, could quietly undermine such strategies. The study also hints at unexplored territory: cyclosporine A, which blocks the mitochondrial permeability transition pore, protected cells without preventing hyperpolarization, raising the possibility that transient, low-conductance pore opening contributes to ferroptotic progression in ways that standard membrane potential measurements miss. For now, the message is clear: if you want to kill cancer cells by starving them of cysteine, you had better make sure their mitochondria, and any diabetes medication in the mix, are not working against you.

Subject of Research: Inhibition of cysteine deprivation-induced ferroptosis by the biguanide drugs metformin and phenformin through suppression of mitochondrial dysfunction

Article Title: Metformin Inhibits Ferroptotic Cell Death by Suppressing Cysteine Deprivation‐Induced Mitochondrial Dysfunction

Article References: Jang, S.-K., Kim, S., Kim, G., Hong, J., Park, I.-C., & Jin, H.-O. (2026). Metformin Inhibits Ferroptotic Cell Death by Suppressing Cysteine Deprivation‐Induced Mitochondrial Dysfunction. Pharmacology Research & Perspectives, 14(5), Article e70330. https://doi.org/10.1002/prp2.70330

Image Credits: AI Generated

DOI: 10.1002/prp2.70330

Keywords: metformin, phenformin, ferroptosis, cysteine deprivation, mitochondria, lipid peroxidation, reactive oxygen species, iron accumulation, cancer, biguanides, mitochondrial complex I, cell death