GPX4 gene therapy helps injured optic nerves regenerate and shields retinal cells in mice

Glaucoma is the leading cause of irreversible blindness worldwide, and current treatment focuses on lowering eye pressure, which does not stop neurons from dying in many patients. Therapies that directly protect retinal ganglion cells (RGCs) or help their axons regrow in the optic nerve are a major unmet need. Earlier studies linked lipid peroxidation and a form of cell death called ferroptosis to optic nerve disease, but the role of GPX4, the main enzyme that reduces phospholipid hydroperoxides, had not been tested as a gene therapy.

The team examined GPX4 in two mouse models: traumatic optic nerve crush (ONC) and the silicone oil-induced ocular hypertension (SOHU) glaucoma model. Using immunostaining, RGC single-cell RNA-seq data, Retro-Seq, and RiboIP-seq, they found GPX4 was increased, not decreased, in surviving RGCs, in regenerating RGCs, and in glaucomatous RGCs. They then used AAV vectors with an RGC-specific promoter to overexpress mitochondrial or cytosolic GPX4 (with SECIS elements for selenocysteine incorporation). They compared this with liproxstatin-1 (a lipid radical-trapping antioxidant) and deferiprone (an iron chelator and ferroptosis inhibitor).

Both GPX4 forms drove significant optic nerve regeneration after crush injury, and liproxstatin-1 did the same. Crush raised lipid peroxidation markers (MDA and 4-HNE), and GPX4 lowered them. GPX4 and liproxstatin-1 also increased RGC survival. In glaucomatous mice, GPX4 overexpression did not change eye pressure but increased thickness of the retinal ganglion cell complex on OCT, preserved pattern ERG responses and optokinetic visual acuity, and increased survival of RGC cell bodies and optic nerve axons. Deferiprone reduced retinal iron and modestly improved survival, but it did not promote regeneration after crush and did not meaningfully improve visual function in glaucoma.

The authors conclude that lipid peroxidation itself, rather than iron-dependent ferroptosis, limits axon regeneration and drives glaucomatous neurodegeneration. They suggest oxidized membrane lipids may stiffen membranes and disrupt lipid rafts and signaling needed for axon regrowth, though this mechanism was not tested here. GPX4 gene delivery or lipid-peroxidation inhibitors could become neuroprotective strategies independent of eye-pressure lowering.

Caveats are significant: all findings come from mice, treatment was given around the time of injury, and long-term safety of sustained GPX4 expression is unknown. Mechanisms remain speculative, and the full methods were only in supplemental documents.