Dying Eye Cells Can Pull Back from the Brink. Here’s the Machinery That Saves Them

The cell death program is almost ostentatiously thorough. Once a cell commits to apoptosis, the internal machinery fires in sequence: the mitochondria fragment, membrane proteins migrate to the outer surface to flag the cell for disposal, enzymes called caspases begin dismantling structural proteins, the membrane blisters and blebs. Textbooks have described this as a one-way trip. You don’t come back from caspase activation. You don’t come back from PARP cleavage. And you certainly don’t come back from full membrane blebbing, which is essentially the cellular equivalent of packing your desk and heading for the door. Except, as a team at the University of Michigan has just demonstrated in photoreceptor cells, sometimes you do.

The finding matters enormously for a specific reason. Most cells in the body, when damaged, can regenerate. Photoreceptors cannot. These are the light-sensing cells of the retina, roughly 120 million of them per eye, responsible for translating photons into neural signals. When they die, they stay dead, and diseases that kill them, including retinitis pigmentosa, age-related macular degeneration, and retinal detachment, cause irreversible blindness as a result. There are no replacements. Whatever you lose, you lose permanently.

Which is precisely what makes the Michigan results so consequential. In a study published in Cell Death & Disease, David Zacks and his team exposed murine cone photoreceptor cells to two different stressors: the apoptosis-inducing compound staurosporine, and hypoxic conditions (roughly 1% oxygen) that mimic what photoreceptors experience during retinal detachment when the retina separates from its blood supply. Both treatments produced the expected apoptotic response. Cells rounded up, their membranes blebbed, caspase-3 and PARP were cleaved, phosphatidylserine flipped to the outer membrane. Standard death sequence. Then the team removed the stressor.

Recovery began within about 2.5 hours. Within 24 hours it was essentially complete. Cells that had been displaying the full suite of late-apoptotic features returned to normal morphology and resumed function. The flow cytometry numbers are rather striking: immediately after staurosporine treatment, only 14% of cells were classifying as healthy by Annexin V/propidium iodide staining. After 24 hours of recovery, that figure was 51.7%.

“These results were exciting because even if we can’t cure the underlying disease, we can try to activate those survival pathways and keep cells alive,” said Zacks, a professor of ophthalmology and visual sciences at the University of Michigan and member of the Caswell Diabetes Institute.

The phenomenon has a name, though it’s not yet widely known outside cell biology circles. Anastasis, from the Greek for “rising to live,” describes the ability of cells to reverse the apoptotic cascade after the initiating signal is removed. It was first characterised in HeLa cells a little over a decade ago by Helen Tang and colleagues, who showed that even cells displaying membrane blebbing and caspase activation could recover if the chemical trigger was taken away. The Michigan study is the first to show anastasis occurring in photoreceptor cells specifically, which is the distinction that gives the finding clinical weight. HeLa cells proliferate; photoreceptors don’t. Recovery here isn’t just biologically interesting. It’s potentially the difference between sight and its loss.

The key to that recovery, the Michigan team found, lies with the mitochondria. During apoptosis, mitochondria become dysfunctional, their reactive oxygen species production climbs, their ATP output drops. The chain of signals this produces is, in part, what drives the death program forward. The question is: what happens to them when the stressor is removed? The answer is that they recover too, but not passively. The cells activate a process called mitophagy, the selective autophagy of damaged mitochondria, essentially packaging the most damaged organelles and degrading them, then rebuilding mitochondrial capacity from healthier stock. ATP levels recovered after stressor removal. Mitochondrial reactive oxygen species fell back to baseline. Markers of mitophagy, including Pink1, Parkin, Fundc1, and LC3B, all spiked during the recovery phase.

“It’s like having a corroding battery in the cell that is leaking toxins,” Zacks said. “Mitophagy gets rid of those bad batteries.”

The team then tested whether mitophagy was merely correlating with recovery or actually driving it. Using a selective mitophagy inducer (MF-094, which boosts Pink1/Parkin signalling), they found that enhanced mitophagy further reduced apoptosis during recovery. Using an inhibitor (Mdivi-1, which blocks mitochondrial fission and therefore mitophagy), recovery was blocked entirely. Cells that should have pulled back from the brink didn’t. This suggests mitophagy isn’t just part of the recovery landscape; it’s a load-bearing component of it.

The in vitro results were then validated in a mouse model. A standard experimental model of retinal detachment keeps the retina permanently separated from the underlying retinal pigment epithelium, so no reattachment and no recovery is possible. The Michigan team developed a variation using a more dilute injection of sodium hyaluronate that produces a detachment resolving naturally within three days, mimicking the situation after successful surgical repair of retinal detachment in human patients. The reattached retinas showed significantly fewer dying photoreceptors (by TUNEL staining), less caspase-3 activity, and better preserved photoreceptor morphology than permanently detached ones. Recovery from apoptosis, it seems, occurs in vivo too, at least when the insult is sufficiently transient.

There are caveats worth keeping in mind. The current work demonstrates association rather than direct causality between mitophagy and recovery, and the paper is forthright about this: genetic approaches will be needed to establish the causal chain. The in vitro model also lacks single-cell tracking of apoptotic markers, which would be required to formally claim that individual cells have undergone anastasis rather than that a surviving subpopulation is masking aggregate recovery. And of course, these are mouse cells, not human photoreceptors in an aging eye under conditions of AMD or retinitis pigmentosa, diseases that are vastly more complex than a 72-hour hypoxia protocol.

But there’s a coherent clinical implication taking shape here. Clinical studies of retinal detachment have long shown that surgical reattachment within roughly a week of onset produces substantially better visual recovery than later repair, suggesting photoreceptors have some capacity for recovery if the insult isn’t too prolonged. Until now, the molecular events underlying that window were unknown. The Michigan study offers a candidate explanation: anastasis, driven by mitophagy, operating within a time window that closes as mitochondrial damage accumulates beyond recovery. The practical consequence, if the mechanism holds in human tissue, is that pharmacological enhancement of mitophagy could extend that window, keeping photoreceptors viable for longer after detachment, reducing permanent visual loss in patients who can’t reach surgery quickly.

Which is probably not the last word on what this window can be made to do. There are other retinal diseases in which photoreceptors die slowly under chronic stress, AMD and retinitis pigmentosa among them, and the question of whether milder, sustained apoptotic activation might also be reversible if the stress could be reduced or interrupted remains open. Whether mitophagy enhancement could be therapeutic in those contexts, or whether the mechanism only applies to acute, removable stressors, is the kind of question a field tends to spend the next decade on.

DOI / Source: https://doi.org/10.1038/s41419-026-08436-3

Frequently Asked Questions