Introduction

Vision loss from optic nerve injury or glaucoma happens because retinal ganglion cells (RGCs) fail to regrow their axons. In adult mammals, the intrinsic growth program of RGCs is normally shut off, so damaged nerves do not heal on their own (pmc.ncbi.nlm.nih.gov). Recent mouse studies show that gene therapy can reactivate these growth pathways. For example, deleting the PTEN gene (a brake on cell growth) in adult RGCs turns on the mTOR growth pathway and leads to strong axon regrowth (pmc.ncbi.nlm.nih.gov). In this article we review how manipulating PTEN/mTOR, KLF-family genes, and Sox11 can stimulate RGC axon regeneration, what this has achieved in mice, the safety issues (like cancer risk), how genes are delivered (AAV viral vectors, intravitreal or suprachoroidal injection), and what steps are needed to move from acute injury models to chronic glaucoma treatment.

Intrinsic Growth Pathways in RGCs

PTEN/mTOR Pathway

Under normal conditions, adult RGCs keep the mTOR pathway largely off, which limits their ability to grow new axons (pmc.ncbi.nlm.nih.gov). PTEN is a gene that inhibits mTOR. Scientists found that removing PTEN in adult mouse RGCs unleashes mTOR signaling and allows axon regrowth (pmc.ncbi.nlm.nih.gov). In one landmark study, conditional knockout of PTEN in adult mice led to robust optic nerve regeneration (pmc.ncbi.nlm.nih.gov). About 8–10% of the surviving RGCs extended axons more than 0.5 mm past the injury, with some axons growing beyond 3 mm and even reaching the optic chiasm by 4 weeks after injury (pmc.ncbi.nlm.nih.gov). Knocking out another brake on mTOR, the TSC1 gene, also induced axon regrowth (pmc.ncbi.nlm.nih.gov).

Deleting PTEN not only spurred regrowth but also improved RGC survival (about 45% survival vs ~20% in controls) (pmc.ncbi.nlm.nih.gov). However, there is a safety concern: PTEN is a tumor suppressor. Long-term PTEN loss can promote uncontrolled cell growth. Indeed, a major regeneration study noted that permanently deleting PTEN would be clinically unacceptable because of cancer risk (pmc.ncbi.nlm.nih.gov). To address this, researchers suggest using controllable gene therapy (for example, AAV-delivered shRNA under a switchable promoter) so PTEN activity can be turned off during regrowth and then back on (pmc.ncbi.nlm.nih.gov). In short, PTEN/mTOR is a powerful internal growth switch, but it must be carefully controlled.

KLF Family and Sox11

Researchers have also targeted transcription factors that control axon growth. The Krüppel-like factors (KLFs) are a family of such genes. A key finding is that KLF4 acts as a brake on axon growth: RGCs lacking KLF4 grow better than normal (pmc.ncbi.nlm.nih.gov). In mice engineered so RGCs have no KLF4, these neurons extended much longer neurites in culture and, after optic nerve crush, many more axons grew out. For example, two weeks after injury, KLF4-knockout mice had significantly more regenerating fibers beyond 1 mm from the crush site than wild-type mice (pmc.ncbi.nlm.nih.gov). Other KLFs have varied roles: some (like KLF6 and KLF7) promote growth, while others (like KLF9) suppress it (pmc.ncbi.nlm.nih.gov). Thus, rebalancing KLF expression can lift some of the developmental “brakes” on RGC growth (pmc.ncbi.nlm.nih.gov) (pmc.ncbi.nlm.nih.gov).

Another transcription factor is Sox11, important in development. Overexpressing Sox11 in adult RGCs (using AAV gene delivery) was also found to boost regeneration. In one study, RGCs with extra Sox11 showed a marked increase in axon regrowth after injury (pmc.ncbi.nlm.nih.gov). However, Sox11 has mixed effects: it promotes regrowth in certain RGC types but can kill others. Notably, Sox11 overexpression killed nearly all the so-called “alpha” RGCs (a subtype of RGC) that usually respond well to PTEN-based treatments (pmc.ncbi.nlm.nih.gov). In other words, Sox11 reprograms some RGCs into a growth-competent state, but it also harms others (pmc.ncbi.nlm.nih.gov). Scientists conclude that different RGC subtypes require different regrowth strategies.

Key Mouse Optic Nerve Crush Studies

Mouse models of optic nerve injury (optic nerve crush) have shown how these gene manipulations work in practice. A classic approach combined pathways for maximal effect. In one PNAS study, scientists applied three treatments: deleting PTEN, inducing inflammation in the eye (zymosan), and elevating cAMP. This trio triggered RGCs to regrow axons all the way through the optic nerve and into the brain’s visual centers (pmc.ncbi.nlm.nih.gov). When they examined the brains of treated mice, many regenerating fibers reached the lateral geniculate nucleus, superior colliculus, and other visual areas (pmc.ncbi.nlm.nih.gov). Importantly, this regrowth led to partial recovery of vision-related behaviors. Treated mice regained some ability to perform simple visual tasks: they could track moving patterns (an optomotor reflex) and judge depth better than injured controls (pmc.ncbi.nlm.nih.gov). (They even showed better circadian light responses [20†L33-L38], though those details can be hard to measure.) This work demonstrated that long-distance axon regeneration in adult mice can functionally reconnect parts of the visual system.

Other studies focused on individual factors. Intravitreally delivering an AAV carrying a constitutively active TrkB (a brain-derived neurotrophic factor receptor) caused even longer growth. For example, Nishijima et al. used an engineered TrkB (called F-iTrkB) delivered by AAV and saw axons regrow over 4.5 mm, with some reaching the optic chiasm (pmc.ncbi.nlm.nih.gov). Similarly, forcing a growth-promoting gene like active K-Ras (a well-known oncogene) into RGCs gave about 3 mm of regeneration (pmc.ncbi.nlm.nih.gov). Interestingly, no tumors were seen in those treated eyes, but the authors still recommend using inducible on/off gene switches for safety (pmc.ncbi.nlm.nih.gov). These and other studies confirm that turning on intrinsic growth genes can indeed drive regeneration in mouse optic nerve injury models.

Partial Visual Recovery

The mouse experiments often tracked not just anatomy but function. The optomotor reflex (mice following moving stripes) and depth perception tests are simple ways to see if vision improves. In the triple-treatment study (pmc.ncbi.nlm.nih.gov), mice showed partial restoration of these reflexes. They could once again respond to moving visuals and judge depth, whereas injured mice without treatment could not (pmc.ncbi.nlm.nih.gov). This is encouraging: it means the regrown axons made useful connections. However, the recovery was only partial. Many visual pathways (especially fine image-forming vision) remain disconnected. So far, regeneration has restored basic visual responses, but not full vision. Still, seeing any functional gain confirms the potential of these strategies.

Safety Considerations

While gene therapy for regeneration is promising, safety is a critical concern. The same growth pathways that help axons can also cause problems if left unchecked. As noted, permanently deleting PTEN is a cancer risk (pmc.ncbi.nlm.nih.gov). Likewise, chronically activating mTOR can lead to tumor growth (for instance, TSC1/2 patients get tumors). Gene therapies that push growth factors (like engineered RAS or other oncogenes) must be carefully controlled. Notably, in experimental AAV-RAS therapy no tumors were observed in mouse eyes (pmc.ncbi.nlm.nih.gov), but the authors stress using regulated (inducible) systems in case any oncogenic activity needs to be shut off (pmc.ncbi.nlm.nih.gov).

Other safety issues include cell death and immune reactions. Some interventions harm certain cells: for example, Sox11 overexpression killed many alpha-type RGCs (pmc.ncbi.nlm.nih.gov). Any therapy that kills RGCs offsets its benefit. There is also the risk of damage from injections or inflammation. Inducing inflammation (zymosan) helped regeneration in mice, but in humans it would be dangerous. Long-term effects of AAV inserts (like insertional mutagenesis) are low, but any ocular gene therapy needs careful assessment. In short, each growth-promoting gene must be balanced against potential harm: ideally delivered transiently or under tight control.

Gene Delivery Strategies

Getting genes into the right cells is a key challenge. For RGCs, adeno-associated viruses (AAVs) are the workhorse vectors. AAVs are safe, non-replicating viruses that can carry therapeutic genes into retinal cells. A common method is intravitreal injection: injecting AAV directly into the vitreous gel of the eye. AAV2 is the classic serotype for retinal transduction; it efficiently reaches RGCs when injected intravitreally (pmc.ncbi.nlm.nih.gov). In fact, one study found that intravitreal AAV2 transduced over 90% of RGCs (pmc.ncbi.nlm.nih.gov). Other capsids can be used too. For example, AAV6 given intravitreally shows very high tropism for the inner retina and RGC layer (pmc.ncbi.nlm.nih.gov). Scientists also engineer AAV2 variants (like mutations or chimeras) to cross retinal barriers even better, but those details are evolving.

Another route is suprachoroidal injection, where a needle or microcannula delivers AAV between the sclera and choroid (the vascular layer). This approach spreads the vector widely under the retina. Suprachoroidal AAV8 in monkeys resulted in broad gene expression (pmc.ncbi.nlm.nih.gov). It can be done with specially designed microneedles. Suprachoroidal delivery avoids major surgery but is still invasive and can cause local inflammation. In fact, suprachoroidal AAV8 caused mild chorioretinitis (inflammation of the choroid) that required steroids, although it resolved over weeks (pmc.ncbi.nlm.nih.gov) (pmc.ncbi.nlm.nih.gov). Importantly, suprachoroidal delivery triggered weaker systemic antibody responses to the AAV capsid than intravitreal delivery (pmc.ncbi.nlm.nih.gov). This is likely because some virus escapes the eye differently. Overall, suprachoroidal injection shows promise for gene therapy to the back of the eye, but its immune effects need management.

Immunogenicity

Even though the eye is somewhat “immune-privileged,” AAV gene delivery can still provoke immune reactions. Intravitreal AAV often leaks out of the eye through drainage channels. One study in primates found that intravitreal AAV resulted in ~400–500 times more virus in the bloodstream compared to subretinal injection (pmc.ncbi.nlm.nih.gov). This caused a very strong antibody response against the AAV capsid (pmc.ncbi.nlm.nih.gov). In contrast, subretinal AAV (injected under retina) is sequestered in the eye and typically elicits almost no anti-capsid antibodies (pmc.ncbi.nlm.nih.gov). Suprachoroidal AAV lies in between: some virus stays in the eye while some reaches nearby tissues. Studies show suprachoroidal AAV causes milder anti-capsid antibody production than intravitreal (pmc.ncbi.nlm.nih.gov), but it can spur immune cells against the gene product (as seen with GFP) because it transduces cells outside the blood-retinal barrier (pmc.ncbi.nlm.nih.gov).

In addition to antibodies, T-cell responses can attack transduced cells. If the inserted gene produces a protein that the body sees as foreign (like GFP in experiments), immune cells may clear those cells. Even real human genes can sometimes trigger low-level inflammation. Clinical retinal gene trials (e.g. for RPE65) often give steroids to dampen this response. Routes that stay within the retina (subretinal, suprachoroidal) tend to be less immunogenic overall than vitreous injections (pmc.ncbi.nlm.nih.gov) (pmc.ncbi.nlm.nih.gov). Future therapies will need to balance efficient delivery with minimal immune activation, possibly using newer AAV types or immune-suppressive regimens.

Translating to Glaucoma

Glaucoma presents a different challenge than an acute nerve crush. In glaucoma, RGCs die slowly due to factors like high eye pressure, reduced blood flow, and stress. To treat glaucoma, gene therapy must work in a chronic injury setting. This means timing is important: therapies may need to be given early to protect RGCs, or periodically to retune growth signals. Fortunately, some work is starting to bridge this gap. In a recent study, researchers used AAV to deliver an always-on TrkB receptor (F-iTrkB) into the eyes of glaucoma-model mice. These mice showed both protection of RGCs and substantial axon regeneration (pmc.ncbi.nlm.nih.gov). This suggests even in glaucomatous conditions, activating growth pathways can help.

Still, moving from crush models to human glaucoma will require more steps. We need to test these gene therapies in animal glaucoma models (like induced ocular hypertension or genetic models) rather than only in crush. We also must consider aging and the diseased environment: older neurons, scar tissue, and fluctuating eye pressure. It will likely be necessary to combine gene therapy with standard glaucoma care (lowering pressure, using neurotrophic factors) and to use controlled gene systems. For example, as noted, AAV constructs could use inducible promoters so that the growth factor gene can be turned off after axons have regrown (pmc.ncbi.nlm.nih.gov) (pmc.ncbi.nlm.nih.gov). Because human glaucoma progresses slowly, a single gene injection might not be enough; repeated dosing or long-lasting vectors could be needed. In summary, translating these findings to glaucoma therapy will mean adjusting for chronic injury dynamics and ensuring treatments are safe and durable.

Conclusion

Gene therapy that modulates RGC intrinsic pathways shows exciting potential: in rodents it can make the optic nerve grow back and even restore some vision. Key strategies like PTEN/mTOR activation, KLF4 deletion, or Sox11 overexpression each give regenerative boosts through different cell programs. Mouse studies confirm axons can reinnervate the brain and improve simple visual tasks (pmc.ncbi.nlm.nih.gov). However, safety issues (oncogenic risk, cell loss, immune response) must be solved, and delivery methods refined. Progress in AAV vectors and ocular injections provides tools to target RGCs efficiently (pmc.ncbi.nlm.nih.gov) (pmc.ncbi.nlm.nih.gov). Next steps include testing in chronic glaucoma models, optimizing dosing and promoters, and combining gene therapy with glaucoma treatments. Altogether, the preclinical evidence strongly supports further development: by carefully tuning intrinsic growth pathways, we may fundamentally change the outlook for optic nerve repair.