After the Axon Grows: Three Checkpoints Standing Between Regeneration and Actually Seeing Again
Growing an optic nerve axon and restoring sight are not the same achievement. Recent research across three separate subfields — myelination, cellular metabolism, and retinotopic mapping — maps out what still stands between the two, and these findings rarely get discussed together.
The optic nerve carries roughly 1.2 million fibers from the eye to the brain, and in humans it does not repair itself — once those fibers are cut or degenerated, by injury, glaucoma, or other disease, they stay that way. The last decade of research has changed part of that picture: labs have found real ways to make these fibers grow again in animal models, something long thought biologically impossible in mammals. But a growing axon and a working eye are not the same achievement, and most public attention — including a companion piece I wrote earlier — stops at the growth question. This piece looks at three checkpoints, drawn from recent literature, that still stand between a regenerated axon and restored sight, and that rarely get discussed together.
Checkpoint One: A Grown Axon That Can't Carry a Signal
Even a successfully regenerated axon is not automatically useful. Signal conduction in the mature visual pathway depends on myelin sheaths structured for saltatory conduction — and a regenerating axon leaves behind a supportive developmental environment for a chronically disrupted adult tract. A 2025 review from the University of Portsmouth and University of Surrey, part of a growing body of work on "functional restoration" rather than mere axon extension, frames this directly: activating a neuron's growth program is only the first of several distinct barriers standing between injury and restored vision, and remyelination of newly grown fibers is one of them.
Consistent with this, a separate 2025 review in Neurotherapeutics on CNS repair mechanisms notes that remyelination efficiency itself varies by location and declines with age, as oligodendrocyte precursor cells become exhausted and less able to differentiate over time — meaning a regenerated axon in an older visual system faces a double disadvantage.
One of the more advanced pharmacological candidates here is clemastine fumarate, an over-the-counter antihistamine repurposed as an M1 muscarinic receptor antagonist that promotes oligodendrocyte precursor differentiation. It has already shown a measurable reduction in visual-evoked-potential latency — a direct marker of remyelination — in people with MS-related optic neuropathy, and a 2025 mini-review from Jichi Medical University surveys its effects across several animal models of white matter injury, including internal capsule and stroke-related demyelination.
Sources: Beard, El-Bushra, Kader, Sun, Au, "From Regeneration Failure to Functional Restoration," CNS Neuroscience & Therapeutics, 2026; Gluck, Gerstein, Kaunzner, "Repair mechanisms of the central nervous system: From axon sprouting," Neurotherapeutics, 2025; Yamazaki, Ohno, "The potential of repurposing clemastine to promote remyelination," Frontiers in Cellular Neuroscience, 2025; MS Trust (UK), clemastine overview.
Checkpoint Two: Regeneration Runs on Sugar, Not Just Mitochondria
The intuitive assumption is that a growing axon simply needs more mitochondrial energy. A September 2025 study in iScience, using RNA sequencing on adult zebrafish retinal ganglion cells after optic nerve crush, found something closer to the opposite. During the phase of successful axon regrowth, genes for oxidative phosphorylation were actually downregulated. In their place, the cells sharply upregulated glycolysis and the pentose phosphate pathway, paired with a strong induction of the thioredoxin antioxidant system — a coordinated shift that supplies both fast energy and the NADPH needed for biosynthesis, while neutralizing the oxidative stress that damaged mitochondria would otherwise generate.
What makes this finding more than a zebrafish curiosity is that the same signature turned up in the mammalian PTEN/SOCS3 co-deletion model — the same genetic manipulation behind one of the best-established mouse regeneration results. The researchers also validated the mechanism functionally: blocking glycolysis or the thioredoxin system in retinal explants impaired axon regrowth. In other words, this isn't a side effect of regeneration — inhibiting it stops regeneration.
Sources: "Successful axonal regeneration is associated with intraneuronal metabolic reprogramming and thioredoxin-dependent detoxification," iScience, 2025; Au, Raza, Kumar, Asthana et al., "A small molecule M1 promotes optic nerve regeneration," PNAS, 2022.
Checkpoint Three: Reaching the Target Isn't the Same as Wiring Correctly
The final barrier isn't distance — it's precision. Vision depends on strict retinotopic mapping: specific regions of the retina must connect to precise, corresponding coordinates within the superior colliculus and lateral geniculate nucleus. A 2020 review from USC's Department of Ophthalmology lays out why this matters practically, not just theoretically — during development, guidance cues like ephrin-A and its receptor EphA are expressed in gradients across the retina and superior colliculus that establish this topographic map, and in the adult injured system, those same gradients are altered or absent. Even when axons do reach their general target region, the review notes, they tend to arrive with disorganized, imprecise connectivity rather than the clean map vision requires — meaning "the axon got there" and "the axon reconnected correctly" are two different, independently unsolved problems.
Where the Real Gaps Are
Read together, these three checkpoints suggest specific, testable questions that don't appear to have been addressed head-on in the published literature:
- Does remyelination need to be synchronized with axon growth, rather than applied after it? Every clemastine and anti-LINGO-1 trial tests remyelination of an axon already fully formed and simply stripped of its sheath. None appear to test whether a still-elongating regenerated axon needs pro-myelinating support delivered progressively, segment by segment, as it extends — timed to the growth front itself, not to the injury date.
- Is the zebrafish metabolic shift necessary for regeneration, or just accompanying it? Blocking glycolysis and the thioredoxin system stops zebrafish axons from regrowing — strong evidence the pathway is required. Untested is the reverse: whether inducing that same shift in an otherwise non-regenerating mammalian RGC, without genetic manipulation like PTEN deletion, is enough on its own to unlock regrowth.
- Does the axon's own metabolic state affect whether it gets remyelinated at all? Oligodendrocyte wrapping is itself metabolically demanding, requiring substantial lipid synthesis. Whether a regenerating axon running on glycolysis rather than oxidative phosphorylation sends a different signal to nearby oligodendrocyte precursor cells than a mature axon does sits exactly at the seam between checkpoints one and two — and neither body of research currently speaks to the other.
- Is there a minimum threshold of retinotopic precision below which regeneration provides no usable vision at all? Regenerated axons often reach their general target but wire imprecisely. Whether that imprecision degrades vision gradually, or whether there's a sharp threshold below which the brain cannot extract a coherent image at all, changes what "success" should even mean in a regeneration therapy.
None of these are claims this piece can settle — they're the questions that only become visible once checkpoint research, usually published in isolation within separate subfields, is read side by side.
This piece is a research synthesis and opinion essay, not a peer-reviewed publication, and is intended to prompt discussion among specialists in axon regeneration, remyelination biology, and neuronal metabolism.