Zombie Cell Nanoparticles: Restored Vision in AMD

The story of zombie cells—those that refuse to die on time and poison the surrounding tissue—is one of the most exciting stories in the anti-aging world of the last decade. In 2015, a team from the Mayo Clinic led by Zhu and colleagues first showed (in *Aging Cell*) that they could be selectively eliminated with the drug combination dasatinib + quercetin (D+Q). Three years later, in 2018, a team led by Xu and colleagues (in *Nature Medicine*) showed that eliminating senescent cells in old mice improved physical function and extended healthspan. Since then, fisetin, navitoclax, and dozens of other senolytic molecules have entered research. But they all share a common problem: when administered systemically, through the blood, they indiscriminately damage senescent cells throughout the body.

On March 18, 2026, a Korean study was published in the journal Nature Communications offering a completely different approach. A team led by Prof. Yoo Ja-hyung from the Department of Chemistry at UNIST (Ulsan National Institute of Science and Technology) and Prof. Chung Hye-won from the Department of Ophthalmology at Konkuk University Hospital developed nanoparticles that identify zombie cells in the retina, deliver a senolytic drug to them, and trigger controlled cell death within them. In mouse models, the treatment led to a significant reduction in the number of senescent cells in the retina, restoration of retinal function (as measured by electroretinography), and re-thickening of the photoreceptor layer. This is an early but compelling demonstration of a principle: senolytics delivered in a targeted manner to an organ, rather than through the bloodstream.

Anyone who has followed the senolytics field in recent years knows why this is interesting. Systemic senolytics have a glass ceiling of side effects, and the next logical step is a move toward organ-targeted treatments. The Korean study is an early proof-of-concept for this, in mice only. It is important to emphasize right now: this is a preclinical study, years away from human use, and there is currently no approved eye-targeted senolytic treatment.

What is Age-Related Macular Degeneration (AMD)?

AMD is the number one cause of vision loss in adults over 60 in the Western world. In the US alone, more than 11 million people suffer from it. In Israel, it is also one of the leading causes of significant visual impairment in old age.

• Macula: A small area in the center of the retina, responsible for sharp, central vision.
• RPE (Retinal Pigment Epithelium) cells: A layer of cells that maintains the photoreceptors. They are the 'maintenance crew' of the retina.
• Two main forms: Dry AMD (most cases, gradual deterioration), Wet AMD (growth of pathological blood vessels, rapid and aggressive).
• Symptoms: Central blurring, distortion of straight lines, difficulty reading and recognizing faces.
• Existing treatment: Monthly injections of anti-VEGF (Eylea, Lucentis) into the eye, mainly for the wet form, and they only slow, not cure.

For the dry form, which constitutes the majority of cases, there is currently no cure. AREDS2 supplements (zinc, copper, lutein, zeaxanthin) slow deterioration to some extent. In the original AREDS study, the formula reduced the risk of progression to advanced AMD by about 25% in people at moderate-high risk. More recent analyses of AREDS2 showed a slowing of about 55% in the rate of geographic atrophy expansion toward the center, though not in all subgroups, and there are also newer, conflicting data.

The deterioration of dry AMD is slow but continuous. Patients start with mild blurring when reading, progress to difficulty recognizing faces, and end with central functional blindness. Patients describe the experience as a 'black hole in the center of the image': peripheral vision is preserved, but everything they look at directly disappears. Driving, reading, watching TV, and recognizing family members up close become difficult.

The impact on quality of life is enormous. Progressive vision loss in old age is associated with a significant decline in quality of life, an increased risk of depression, and loss of independence. This is why any progress toward a treatment that stops or reverses the process generates great interest.

The Connection to Zombie Cells: The Mechanism

RPE cells divide very little throughout life. They are exposed to strong light, high oxygen, and byproducts from the photoreceptors they 'clean up'. All of these cause chronic oxidative stress and accumulation of DNA damage. With age, an increasing percentage of RPE cells enter a state of *senescence*, cellular aging, but do not die.

In this state, they become 'zombies': alive, but secreting a toxic cocktail of inflammatory cytokines (SASP), tissue-degrading enzymes, and abnormal growth factors. They poison the healthy cells around them, promote chronic inflammation, and accelerate the deterioration of the entire retina.

The question that has hovered over the field for years: If we eliminate the zombie cells in the retina, will we stop or improve AMD? Attempts with systemic senolytics have shown their safety limitations, hence the need for an organ-targeted approach.

The major problem with a systemic approach: even a healthy body needs some cells that might be marked as 'senescent', and a treatment that spreads throughout the body increases the risk of harming useful cell populations. An eye-targeted approach tries to solve this by leaving the rest of the body out of the equation. This is precisely the challenge the Korean team tried to tackle: how to identify and mark only the senescent cells, and only in the retina.

How Does the Nanoparticle Identify a Zombie Cell?

The Korean team started by identifying a unique biological address. Using comparative transcriptomics analysis of RPE cells from mice, they identified that a protein called Bst2 (also known as CD317 or tetherin) is highly expressed on the surface of senescent RPE cells, and barely on young, healthy cells. Bst2 serves here only as an 'anchor' for identification, not as an active factor in aging itself.

Based on this address, the nanoparticle platform, named B-Z-PON in the study, was built. It consists of mesoporous organosilica nanoparticles about 150 nm in diameter, onto which a *recombinant Fc-binding domain* is attached, and to this domain are attached anti-Bst2 antibodies. The antibodies give the particle the ability to selectively adhere to senescent cells displaying Bst2.

The advantage of this architecture is its modularity: the Fc-binding domain acts as a 'plug' into which different antibodies can be 'connected', making the platform tunable for different targets. In this study, the connection was an anti-Bst2 antibody.

The porous core of the nanoparticle is loaded with the drug navitoclax (ABT-263), a BCL-2/BCL-xL inhibitor that triggers apoptosis in senescent cells. This is a single drug, not a 'dual payload'. The clever point is the release mechanism: the nanoparticle is built with disulfide bridges that break down in a reducing, glutathione-rich (GSH) environment, as exists inside senescent cells. Thus, the drug is released mainly inside the correct cell. Control particles without GSH sensitivity remained intact and did not release the drug in the same way.

In short: the antibody directs the particle to the senescent cell (Bst2), and the core chemistry (GSH-dependent degradation) ensures the release of ABT-263 mainly inside it. A healthy cell, displaying little Bst2, simply does not bind with the same strength.

Why Injection into the Eye, Not Drops?

The first question for most readers: why do the nanoparticles need to be injected? Why not give them as eye drops? The answer is the eye's barriers, anatomical structures that protect the retina from foreign substances, similar to the blood-brain barrier. A large molecule like a drug-loaded nanoparticle does not easily cross them from the outside.

In the study, the nanoparticles were administered via intravitreal injection, placing them close to the RPE layer intended for treatment. This way, they reach the site directly, instead of spreading throughout the body.

The Current Evidence: What the Study Actually Showed

Two Mouse Models, in C57BL/6J Mice

The researchers tested the treatment in two complementary models, both in C57BL/6J mice:

• Doxorubicin-induced aging model: Subretinal injection of doxorubicin created local cellular senescence in the retina within days, allowing the treatment to be tested on 'aging' tissue in a controlled manner.
• Naturally aged mice: Mice aged 24 months (equivalent to old age in humans), which had naturally accumulated senescent RPE cells.

Treatment Regimen

In the old mice, the treatment was given as three intravitreal injections at intervals of about three weeks, and the analysis was performed several weeks after the last injection. So, not a single 'magic' injection, but a series of several injections.

The Results, as Reported

It is important to clarify: the study reported the results mainly qualitatively and with statistically significant comparisons against a control group, not as precise miracle percentages. What was reported:

• Significant reduction in the number of senescent cells in the retina (measured by p16-positive cells and SA-β-gal staining area), with selective targeting of senescent cells only and preservation of healthy cells.
• Restoration of retinal function: A significant increase in electrical responses to light in electroretinography (ERG), including a, b, and c waves, a sign that the photoreceptors and RPE layer had returned to better function.
• Re-thickening of the photoreceptor layer (ONL, outer nuclear layer), meaning structural evidence of tissue recovery, not just a halt in deterioration.

The plausible interpretation: after the senescent cells are eliminated, the remaining healthy cells manage to return to normal activity and better maintain the neighboring photoreceptors. This is a nice demonstration that the tissue environment, not just the individual cell, is important. But again, this is in mice only.

What About Other Eye Diseases?

The concept of nanoparticles targeting senescent cells is not theoretically unique to AMD. If the approach proves itself, one could imagine additional applications where the accumulation of senescent cells contributes to deterioration:

• Glaucoma: Senescent retinal ganglion cells contribute to optic nerve deterioration. Current glaucoma treatment mainly focuses on lowering eye pressure.
• Diabetic Retinopathy: Chronic inflammation from senescent RPE accelerates damage. An eye-targeted approach is particularly interesting for diabetic patients, where systemic drugs could disrupt metabolic balance.
• Aging of other ocular tissues: As unique markers for senescent cells in different tissues are identified, nanoparticles could in principle be adapted for each.

But this is still a vision, not data. The broader vision is a modular platform that can be 'loaded' with a different antibody for each target, but this too needs to be tested separately for each tissue and disease. There is no sweeping promise here, only a research direction.

Should We Start Expecting This Treatment?

The excitement is legitimate, but there are important caveats to be aware of.

The Gap Between Mouse and Human

Results in preclinical models, even when impressive, do not translate directly to humans. A large proportion of treatments that show excellent results in mice fail in human trials. A human eye differs from a mouse eye in size, anatomy, and the nature of AMD.

The biggest gap is time: in the mouse model, aging occurred over months (or was rapidly induced with doxorubicin), whereas in humans, AMD develops over 10-20 years. The accumulation of zombie cells is much slower, and the cumulative damage is deeper. A treatment that works well in a mouse with 'fast' aging might behave differently in a human with years of accumulated damage.

Another point: mice do not have a macula in the human sense, and they rely mainly on peripheral vision. This limits the ability to directly infer from the model to human central vision.

Risks of Intraocular Injection

The treatment must be injected directly into the vitreous humor of the eye. Intravitreal injection carries a small but real risk of intraocular infection (endophthalmitis), bleeding, and increased intraocular pressure. With a series of injections over time, the cumulative risk is not negligible and must be weighed against the benefit.

What is Still Unknown

How does the nanoparticle behave in the eye over years? Does it accumulate in tissues with repeated injections? Does the eye's immune system develop a reaction to the antibody or the particle? These are questions that require further research, and subsequently, research in larger animals and humans, before a treatment can be discussed. The current study did not test monkeys, rabbits, or any other model beyond mice.

Realistic Timeline

A preclinical study in mice is an early step in a long chain. Further safety studies are needed afterward, and only then can human trials be considered, which also take years. Even in an optimistic scenario, a treatment of this type is many years away from the clinic, if it ever arrives. There is currently no approved eye-targeted senolytic treatment.

What is Happening in the Clinic in Parallel

Alongside the nanoparticle research, there are already clinical trials of senolytics for the eye. Unity Biotechnology developed UBX1325 (foselutoclax), a BCL-xL inhibitor injected into the eye targeting senescent cells in retinal blood vessels (not a nanoparticle, but a direct molecule). In the ASPIRE Phase 2b trial for diabetic macular edema, the drug did not meet its pre-specified primary endpoint (non-inferiority to aflibercept at the average of weeks 20 and 24 did not reach the pre-specified significance threshold), although it showed improvement in vision and was non-inferior at some other time points. This means that even the 'simpler' approach has not yet proven itself unequivocally in the clinic. This illustrates how long and uncertain the path from promising results to an approved treatment is.

What to Take from the Study?

1. If you have early-stage AMD, or a family history, get annual eye exams. Early detection is the most important factor in preserving vision.
2. Take AREDS2 supplements if your eye doctor recommends them. They are not a cure, but they have evidence for slowing deterioration in some cases of dry AMD.
3. Stop smoking immediately if you smoke. Smoking is one of the strongest risk factors for AMD and rapid deterioration, after age itself.
4. Protect your eyes from UV. High-quality sunglasses with UV protection reduce long-term oxidative stress on the retina.
5. Maintain a lifestyle that reduces the overall burden of senescent cells. Physical activity, quality sleep, and a balanced diet are linked to better cellular health. This is not a substitute for future treatment, but the basic layer under your control.
6. Eat fatty fish, dark leafy greens, and eggs. Omega-3 (DHA) supports retinal health, and lutein and zeaxanthin from greens and eggs accumulate in the macula and help protect against light damage. A Mediterranean diet has been linked to a lower risk of AMD.
7. Keep realistic expectations. The treatment described here is a study in mice. It marks a promising direction, not an available solution. Do not delay evidence-based treatment available today in anticipation of a future breakthrough.

The Broader Perspective

The story of senolytic nanoparticles in AMD is interesting because it illustrates a direction in the world of senolytics: a possible shift from crude systemic treatment to refined organ-targeted treatment. The first generation of senolytics (D+Q, fisetin) acted throughout the body, both where needed and where it might be harmful. A targeted approach tries to select the organ, and the cell type, and act with greater precision.

Nanotechnology is the tool enabling this attempt. A nanoparticle that can identify a specific surface marker (here Bst2), bind to it, and release a drug only under the internal conditions of the senescent cell (here, a GSH-rich environment) is an elegant idea. But there is a large gap between an elegant idea in a mouse and an approved treatment in humans, and this gap is precisely what needs to be remembered.

It is also worth remembering that nanotechnology has promised great things in the past that took a long time to reach (if at all) the clinic. Healthy caution is needed. Nevertheless, there is a solid foundation here: familiar chemistry of silica nanoparticles, an identification marker identified in data, and an existing senolytic drug (ABT-263). This is not a vision far from science, but a stepwise research advance.

And finally, the human context: if in the future we can effectively treat AMD, we will preserve not only vision but also independence, safe movement, and social connections in old age. Preserving vision is preserving quality of life. This is why research like this generates interest, even when it is still far from the clinic.

References:
Yoo et al., Bst2-targeted senotherapy restores visual function by eliminating senescent retinal cells, Nature Communications (2026)
Seoul Economic Daily - Nanoparticle Targeting Senescent Cells Restores Vision in Macular Degeneration Model