Prox1 Identified as Molecular Brake on Retinal Repair in Mammals

Researchers have uncovered a powerful molecular barrier preventing mammals from regrowing damaged retinal neurons—an ability long considered exclusive to species like zebrafish. The culprit? A homeobox transcription factor called Prox1.

Published in Nature Communications, the study led by scientists at Korea Advanced Institute of Science and Technology (KAIST) reveals that Müller glia, the retina’s resident support cells, can indeed be coaxed into regenerating neurons if Prox1 is removed from the equation. The findings not only challenge the long-held notion of mammalian retinal regenerative incapacity but also introduce a promising gene therapy candidate for retinitis pigmentosa and related degenerative eye diseases.

In zebrafish, Müller glia respond to injury by reprogramming into retinal progenitor cells (RPCs), which then proliferate and differentiate into various retinal neurons, including photoreceptors and ganglion cells. Mammals, however, show little such capacity. After injury, mammalian Müller glia enter a transient activation phase but ultimately return to a quiescent state without meaningful regeneration.

This new study identifies Prox1 as a key molecular roadblock in that process. While Prox1 is a known regulator of neural differentiation, its role in regeneration had remained murky. The KAIST team discovered that after retinal injury, mammalian Müller glia begin to accumulate Prox1 protein, not because they start producing it themselves, but because they absorb it from neighboring retinal neurons through intercellular protein transfer. This extracellular transfer is mediated by Prox1's homeodomain, a feature that allows the protein to move between cells—a trait shared by other homeodomain proteins but rarely considered in mammalian neurobiology.

To test the impact of this transfer, the researchers deployed a cleverly engineered antibody therapy: a single-chain variable fragment (scFv) targeting Prox1, delivered via an adeno-associated virus (AAV) injected into the vitreous of mouse eyes. By blocking extracellular Prox1, they not only prevented its uptake by Müller glia but also triggered a regenerative cascade. Injured Müller glia began to express progenitor markers, re-enter the cell cycle, and give rise to new photoreceptors and other retinal neurons.

What makes these results especially compelling is their application in disease models. In mice with retinitis pigmentosa (Pde6brd10/rd10 and Rp1tvrm64/tvrm64), AAV-mediated delivery of anti-Prox1 restored visual function, albeit transiently, by promoting the regeneration of rod photoreceptors. Treated mice showed increased outer nuclear layer thickness, elevated electroretinogram responses, and measurable improvements in visual acuity. In contrast, untreated or control vector–injected mice continued to lose vision.

Human relevance wasn’t left to speculation. The researchers analyzed postmortem retinas from a patient with retinitis pigmentosa and found Prox1 accumulation in Müller glia—supporting the notion that this inhibitory mechanism is conserved in humans.

Still, the treatment is not a silver bullet. While Prox1 blockade rekindled some regenerative potential, the extent of neuron regeneration in mice remained modest compared to that seen in zebrafish. This suggests that additional barriers, such as epigenetic repression or a lack of pro-regenerative signaling pathways (like Wnt or Notch modulation), may also need to be addressed.

But the implications are promising. By disrupting the passive uptake of a single inhibitory protein, mammalian Müller glia were liberated to act more like their zebrafish counterparts. It’s a therapeutic approach achieved not by adding growth factors or stem cells, but by simply removing a brake.

As gene therapy platforms evolve and promoter durability improves, this study positions anti-Prox1 delivery as a potentially viable treatment for early-stage retinal degeneration. More broadly, it reawakens the long-dormant idea that even in the adult mammalian central nervous system, the machinery for regeneration might not be lost—only silenced.

Source

Lee, Eun Jung, Museong Kim, Sooyeon Park, Ji Hyeon Shim, Hyun-Ju Cho, Jung Ah Park, Kihyun Park, et al. 2025. “Restoration of Retinal Regenerative Potential of Müller Glia by Disrupting Intercellular Prox1 Transfer.” Nature Communications 16 (1). https://doi.org/10.1038/s41467-025-58290-8.

‌