Dynamic O-GlcNAcylation coordinates ferritinophagy and mitophagy to activate ferroptosis

Ferroptosis is a regulated form of necrotic cell death characterized by iron-dependent lipid peroxidation leading to membrane permeabilization. Small molecule compounds such as Erastin, which inhibits cystine import, and RSL3, which directly inactivates the phospholipid peroxidase GPX4, collapse cellular redox homeostasis, elevate lipid peroxidation, and trigger ferroptosis, linking glutathione metabolism to defense against lipid peroxidation. Numerous other processes—including amino acid and polyunsaturated fatty acid metabolism, phospholipids, NADPH, and coenzyme Q10—were found to suppress or drive ferroptosis, placing ferroptotic control at the intersection of amino acid, lipid, and iron metabolism. Ferroptosis contributes to diverse pathophysiological conditions, including neurodegenerative and cardiovascular diseases and cancers, raising opportunities to exploit or defend ferroptotic vulnerabilities. Redox-active iron initiates and amplifies free radical-mediated lipid peroxidation during ferroptosis, and chelation of intracellular labile ferrous iron (e.g., with deferoxamine) inhibits hydroxyl radical formation (Fenton reaction) and ferroptotic death. Iron homeostasis is tightly coordinated: transferrin-bound iron is internalized by transferrin receptors, released in endosomes, and stored by ferritin heteropolymers of heavy and light chains. Ferritin is delivered to lysosomes via its cargo receptor NCOA4 and degraded to release stored iron (ferrihydrite), contributing to ferroptosis. Loss of FTH increases labile iron and accelerates ferroptosis, whereas loss of NCOA4 blocks this process. However, the precise activation mechanism of ferritinophagy-dependent ferroptosis that diverts iron homeostasis under stress remains unclear. O-GlcNAcylation is a reversible post-translational modification adding a single O-linked N-acetylglucosamine to serine/threonine residues on cytoplasmic, nuclear, and mitochondrial proteins, controlled by O-GlcNAc transferase (OGT) and O-GlcNAcase (OGA). It serves as a nutrient sensor of glucose flux through the hexosamine biosynthetic pathway (HBP), with UDP-GlcNAc as the donor substrate. O-GlcNAcylation is highly dynamic and transiently responds to diverse environmental and physiological stresses, acting as a rheostat to tune cellular pathways. It has context-dependent roles, being acutely protective yet contributing to chronic pathologies (e.g., diabetes). Disruption of O-GlcNAc homeostasis is implicated in diabetes, cancer, and neurodegeneration. Whether and how O-GlcNAcylation responds to pro-ferroptotic stress was unclear. Here, the study demonstrates that O-GlcNAcylation senses ferroptotic stress and coordinates both ferritinophagy and mitophagy to drive ferroptosis, uncovering a link between metabolic stress, iron homeostasis, mitochondria, and ferroptosis.

- Cell models: U2OS cells primarily; additional cell lines include HUVEC, HT1080, and HT29. - Ferroptosis induction: RSL3 (commonly 10 μM) applied for indicated times (e.g., 0.5–8 h in time course; 6 h and 24 h endpoints; longer-term 48–72 h for RNAi experiments). Other ferroptosis inducers tested include ML210, iPSP1, and Erastin. Genetic inhibition of GPX4 by RNAi was also used. - O-GlcNAc modulation: OSMI-1 (OGT inhibitor; 1 μM, 12 h pre-incubation) to inhibit O-GlcNAcylation; TMG (Thiamet-G, OGA inhibitor; 10 μM, 12 h pre-incubation) to elevate O-GlcNAcylation. - Immunoblotting: Time-course analysis of global protein O-GlcNAcylation after RSL3 treatment using anti–O-GlcNAc antibodies; β-actin as loading control. - Flow cytometry: Relative O-GlcNAc intensity and lipid ROS assessed. Lipid ROS quantified by BODIPY 581/591 C11 staining with flow cytometry (Ex/Em 488/515 nm for oxidized dye). U2OS and HT29 cells were pre-incubated with OSMI-1 or TMG for 12 h and co-treated with RSL3 (10 μM) for 12 h for ROS assays. - Cell death/viability: Trypan blue staining to quantify dead cells after 12 h OSMI-1 or TMG treatment with or without 6 h RSL3. Cell viability measured by CCK8 assay (OD 450 nm) after OSMI-1 or TMG pre-treatment and 24 h RSL3 exposure; n = 10. - Immunofluorescence microscopy: Cells treated with DMSO or OSMI-1 and then RSL3; stained with antibodies against TR1 and GM130; images acquired with 10 μm scale bars. - Additional reagents/controls: Liproxstatin-1 was used to inhibit lipid peroxidation and assess its impact on O-GlcNAcylation dynamics under RSL3. Expression of GFPT1 (rate-limiting HBP enzyme) at mRNA and protein levels was measured during early RSL3 response. - Replicates and statistics: Experiments repeated at least three times; statistical significance reported (e.g., p < 0.01, p < 0.0001).

- Dynamic O-GlcNAcylation during ferroptosis: RSL3 triggered a robust, transient increase in global O-GlcNAcylation that peaked at ~2 h and declined thereafter; lipid peroxidation (BODIPY C11) likewise peaked at ~2 h. - Generality across inducers and models: Similar biphasic O-GlcNAc responses observed with other ferroptosis inducers (ML210, iPSP1, Erastin) and upon genetic GPX4 knockdown (O-GlcNAc accumulation at 48 h followed by decline at 72 h), and in multiple cell lines (HUVEC, HT1080, U2OS, HT29). - Dependence on lipid peroxidation: Liproxstatin-1 blunted the RSL3-induced rise in O-GlcNAcylation, linking lipid peroxidation to O-GlcNAc dynamics. Early RSL3 treatment reduced UDP-GlcNAc, and GFPT1 mRNA/protein increased as compensation. - Modulation alters ferroptotic sensitivity: Pharmacologic inhibition of O-GlcNAcylation with OSMI-1 increased RSL3-induced cell death and lipid ROS; elevation of O-GlcNAcylation with TMG decreased death and ROS (assessed by Trypan blue, CCK8 OD450, BODIPY C11 flow cytometry). Reported statistics included n = 10 for viability assays and significance thresholds of p < 0.01 and p < 0.0001. - Iron handling and mitochondria: Inhibition of O-GlcNAcylation promoted ferritinophagy, driving labile iron accumulation toward mitochondria; it also caused mitochondrial fragmentation and enhanced mitophagy, providing additional labile iron and sensitizing cells to ferroptosis. - Mechanism: De–O-GlcNAcylation of ferritin heavy chain (FTH) at Ser179 promoted its interaction with NCOA4, the ferritinophagy receptor, increasing ferritin turnover and labile iron release to activate ferroptosis. - Overall: Protein O-GlcNAcylation acts as a metabolic sensor that coordinates ferritinophagy and mitophagy to regulate iron availability and ferroptotic execution.

The study addresses how cells interpret pro-ferroptotic stress to activate ferroptosis by identifying O-GlcNAcylation as a dynamic metabolic rheostat linking lipid peroxidation to iron mobilization and mitochondrial quality control. The early, transient rise in O-GlcNAcylation following ferroptotic insults appears to be a compensatory response to oxidative stress and UDP-GlcNAc depletion, involving upregulation of GFPT1. Functionally, tuning O-GlcNAcylation alters ferroptotic susceptibility: lowering O-GlcNAcylation enhances ferritinophagy and mitophagy, increases labile iron delivery to mitochondria, elevates lipid ROS, and accelerates cell death, whereas elevating O-GlcNAcylation is protective. Mechanistically, de–O-GlcNAcylation of FTH at S179 facilitates NCOA4 binding to drive ferritin turnover. These findings integrate nutrient sensing with iron metabolism and mitochondrial dynamics to explain decision-making in ferroptosis. The work suggests actionable nodes—OGT/OGA activity, HBP flux (GFPT1), and the FTH S179–NCOA4 interaction—for therapeutic intervention in diseases where ferroptosis contributes (e.g., neurodegeneration, cardiovascular disease, cancer).

This work reveals that dynamic O-GlcNAcylation coordinates ferritinophagy and mitophagy to control labile iron availability and lipid peroxidation, thereby regulating ferroptotic cell death. A biphasic O-GlcNAc response to ferroptotic stimuli is observed, and reducing O-GlcNAcylation enhances ferritinophagy/mitophagy and sensitizes cells to ferroptosis via de–O-GlcNAcylation of FTH S179 that promotes NCOA4 interaction. These insights uncover an unrecognized link between metabolic stress signaling and iron homeostasis in ferroptosis, offering potential targets for modulating ferroptosis in disease. Future studies should delineate the broader proteome of O-GlcNAc-regulated ferroptotic effectors, test in vivo disease models, and assess therapeutic strategies to fine-tune O-GlcNAc cycling or the FTH–NCOA4 axis.