Optic Nerve Regeneration: How will we get there?

Abstract

Restoration of vision in patients blinded by advanced optic neuropathies requires technologies that can either 1) salvage damaged and prevent further degeneration of retinal ganglion cells (RGCs), or 2) replace lost RGCs. In this review article, we discuss the different barriers to strategies for optic nerve regeneration and provide an update regarding what progress that has been made to overcome them. We also provide an update on current stem cell-based therapies for optic nerve regeneration. As neuro-regenerative and cell-transplantation based strategies for optic nerve regeneration continue to be refined, researchers and clinicians can work together to determine who will be a good candidate for such therapies.

A. Introduction

Restoration of vision in patients blinded by advanced optic neuropathies requires technologies that are neuro-protective approaches (prevent further degeneration of retinal ganglion cells [RGCs]), neuro-regenerative approaches (salvage damaged RGCs), or cell-replacement approaches (replace lost RGCs). Neuro-protective and neuro-regenerative approaches are more naturally suited for chronic, progressive optic neuropathies such as glaucoma, hereditary, toxic, and metabolic optic neuropathies than for acute optic neuropathies such as traumatic, ischemic, demyelinating, and inflammatory conditions. Although there may be a critical window in which patients with acute optic neuropathies could benefit from neuro-protective therapies, they are more likely to require cell replacement therapies.

There are a number of barriers that must be overcome before cell replacement strategies can be translated to the clinical setting (Fig. 1). First and foremost, transplantation approaches require a source of healthy RGCs to transplant. Second, these cells must be integrated into the native retina so that they may interface with amacrine and bipolar cells. Integrated cells must then sprout an axon that will grow into the nerve fiber layer of the retina, towards the optic nerve head, and then through the lamina cribrosa into the optic nerve towards various synaptic partners in the diencephalon, without taking detours. Finally, once reaching the optic tectum, the axons of these RGCs must generate new, retinotopic appropriate synaptic connections. In this article, we will discuss these different challenges and provide an update regarding the progress that has been made to overcome them. We will finish with a discussion of current stem cell based optic nerve regeneration methods that are being evaluated in clinical trials.

Figure 1: Schematic of proposed steps and obstacles for developing stem cell based therapies for optic nerve regeneration.

Cell-replacement based strategies for optic nerve regeneration need a source of healthy retinal ganglion cells (RGCs) for transplantation. 1) Fibroblasts isolated from patients can be used to develop inducible pluripotent stem cells (iPSCs) which could then be 2) transdifferentiated into RGCs via either 3D retinal organoids or 2D planar cultures. 3) RGCs would then be purified and injected into patient eyes. 4) Currently, transplanted RGCs have poor rates of integration into the retina. 5) Of the RGCs that successfully integrated, limited axonogenesis out of the eye into the optic nerve or past glial scar has been observed. 6) Regenerating axons have also been observed to overshoot their synaptic target. Upon reaching the lateral geniculate nucleus ⁷⁷, new synaptic connections would need to be made.

B. Limited Endogenous Regeneration in the Adult Mammalian CNS

In mammals, axon regeneration after injury is the rule for the peripheral nervous system (PNS) but the exception when it comes to the central nervous system (CNS). The limited regenerative capacity of the CNS was thought to be mostly cell intrinsic until experiments demonstrated that CNS neurites readily grow into peripheral nerve grafts. ^1,2 The notion that environmental cues in the CNS limit regeneration, rather that factors solely intrinsic to the neurons, was further supported by experiments demonstrating blockage of sciatic nerve regeneration with optic nerve grafts. ³ What these landmark studies also suggested is that extrinsic environmental factors can override intrinsic limitations to regeneration in the CNS. These studies fueled investigations to identify growth permissive molecules within the PNS and growth inhibitory molecules within the CNS, with the hope that manipulating their relative levels would unlock the regenerative capacity of the CNS. Decades of work have culminated in the identification of intrinsic regulators of neuronal growth (e.g., cAMP, mTOR/PTEN, APC-Cdh1, KLF4) ^4,5 and extrinsic factors such as myelin byproducts (e.g., myelin-associated glycoprotein, Nogo, and oligodendrocyte myelin glycoprotein) and glial scar as playing important roles in limiting regeneration in the CNS. For example, myelin byproducts are shed from injured axons and are cleared by Schwann cells and blood-borne macrophages in the PNS. ^6–10 The absence of Schwann cells in the CNS is thought to account, in part, for the relatively poor axon regeneration seen in the CNS.

C. Critical barriers to regeneration and current cell transplantation based approaches

a. Source of RGCs

The first step in developing cell replacement-based strategies for optic nerve regeneration requires a reliable, high volume source of healthy RGCs. Given the scarcity and ethical dilemmas associated with using cadaveric retinas and human embryonic stem cells (hESCs), the scientific community was challenged with developing an alternative source. In 2006, Yamanaka followed by Thompson reported a remarkable breakthrough technology whereby adult fibroblasts could be reprogrammed into pluripotent stem cells using four transcription factors (Oct3/4, Sox2, c-Myc, and Klf4). ^11,12 This landmark development, termed inducible pluripotent stem cells (iPSCs), has turned optic nerve regeneration from science fiction to possibility. These trans-differentiated cells exhibit similar characteristics to hESCs including the ability to propagate indefinitely and the ability to differentiate into many different cell types including RGCs. Although the eye is an immune-privileged organ, the fact that these cells would have the same HLA make-up as their hosts confers another advantage over hESC-derived RGCs—iPSC derived cells are less likely to be destroyed by the immune system.

Currently, there are two different methods for generating RGCs from stem cells (hESCs and iPSCs): 1) organoid differentiation and 2) planar differentiation. Significant work is being performed to understand which of these methods is the most robust for producing RGCs that resemble the native RGC. To date, most of the published protocols to generate RGCs from stem cells have been developed with hESCs, although significant effort is being made to translate these technologies to iPSCs as well.

ai. Organoid derived-RGCs

Organoids are self-organizing, three-dimensional miniature organs developed in vitro from pluripotent stem cells. These mini organs can be used to study disease pathology and developmental biology, and for screening drugs. In 2015, Kuwahara devised a protocol to develop 3D retinal organoids from hESCs. ¹³ Using a cocktail of growth factors including bone morphogenic protein 4 (BMP4) in combination with a non-adherent tissue culture surface, hESCs can be coaxed into differentiating into self-organizing retinal cups. Remarkably, these organoids have been shown to recapitulate many aspects of retinal development in vitro, expressing the same sequence of transcription factors that characterize endogenous retinal development ^14,15 and self-organizing into layers similar to the native retina (Fig. 2). ^16,17 Mature RGCs are readily seen in culture starting day 20 of differentiation and continue to develop through day 34. Healthy organoids can be kept alive in culture for many months although RGCs start to degenerate after three months in culture. ¹⁶ RGCs can be readily purified from organoids. Photoreceptor-driven action potentials have been recorded from iPSC-derived RGCs, albeit sluggish compared to native RGCs. ¹⁸ This suggests that iPSC-derived retinal organoids develop functional, albeit rudimentary, retinal circuitry. As these technologies continue to be refined, iPSC organoids may serve as a reliable source from which large volumes of immune compatible, healthy RGCs can be purified and transplanted into patients.

Figure 2: Retinal organoids derived from human embryonic stem cell line.

A and C: Self-organizing retinal organoids derived from the H9 human embryonic stem cell line, day 25 of differentiation. Cell line courtesy of Dr. Donald Zack (Johns Hopkins University), pictures courtesy of Dr. Narine Harutyunyan (Children’s Hospital Los Angeles Stem Cell Core). B: Red fluorescent cells represent mature retinal ganglion cells (RGCs) expressing tdTomato under the control of the Brn3B promoter (a class-4 POU domain transcription factor expressed in mature RGCs). Scale bar = 100 μm.

aii. Planar derived-RGCs

An alternative to organoids, RGCs can be trans-differentiated from stem cells using traditional two-dimensional cell culture techniques. Reh devised a protocol to induce hESCs into a retinal lineage, giving rise to the major retinal cell types in vitro. ¹⁹ Mature RGCs can be identified as soon as 15 days in culture and can be reliably purified after 30 days in culture. ²⁰ hESC-derived RGCs were shown to have electrophysiological properties similar to mature, native RGCs. ²⁰ Additionally, whole-cell voltage clamp recordings demonstrate that hESC-derived RGCs are glutamate responsive. Although the hESC-derived RGCs display morphologic characteristics of native RGCs, these 2D cultures do not produce other retinal cell types and do not develop the retinal circuitry seen in organoid-based cultures.

A major limitation to using autologous iPSCs as a source of RGCs is that any underlying genetic predispositions to neuro-degeneration would still be present in the cells used to generate new RGCs. For example, Ohlemacher extracted skin fibroblasts from a patient containing a mutation in the optineuron (OPTN) gene which has been implicated in normal tension glaucoma. ²¹ He reprogrammed these cells into iPSCs and then differentiated them into RGCs. These iPSC-derived RGCs exhibited increased apoptosis relative to control RGCs but could be rescued by the addition of neuroprotective factors. Although this established an in vitro model of glaucoma, these patients are unlikely to be able to serve as autologous donors. Instead, iPSC donor banks would need to be established.

Although production of RGCs from hESCs is discussed here in general terms, there are over 30 different subtypes of RGCs. ²² These RGC subtypes include ON-OFF directionally selective ganglion cells (ON-OFF DSGCs), ON directionally sensitive ganglion cells (ON DSGCs), intrinsically photosensitive retinal ganglion cells, and alpha-RGCs (αRGCs) among others. Studies have shown that not all types of RGCs are uniformly lost in different forms of optic neuropathies. For instance, 14 days after optic nerve crush, Duan demonstrated 80% of αRGCs survived. ²³ Daniel et al. found a subtype of αRGCs (transient OFF αRGCs) to be the most vulnerable to optic nerve crush injury compared to three other RGC subtypes. ²⁴ Similar results were demonstrated in a mouse model of glaucoma where (intraocular pressure) IOP was transiently elevated. ^25,26 Interestingly, different subtypes of RGCs are not produced at equal rates in hESC cultures. Among hESC-derived RGCs, ON-OFF and ON DSGCs consist of about 17% and 30% of the total RGC population, respectively, while αRGCs were found in about 21% of the population. ²⁷ A continued understanding of which subtypes of RGCs are damaged in different types of optic neuropathies may allow for the development of more targeted therapies.

b. Integration of Transplanted RGCs

Great strides have been made to develop technologies that allow for large volume production of iPSC-derived RGCs for cell-transplantation. Unfortunately, simply injecting RGCs into the vitreous does not lead to meaningful levels of RGC integration into the ganglion cell layer (GCL) of the retina. Goldberg reported between 1–7% of RGCs integrated into the GCL after intravitreal injection with 40–60,000 RGCs. ²⁸ Interestingly, improved integration rates were noted when smaller amounts of RGCs were injected into the eye. When Goldberg co-transplanted RGCs with iPSCs into the vitreous, they saw retinal engraftment rates on the order of 20%. ²⁹ RGCs derived from healthier cultures or younger retinas were also associated with higher integration rates. ^28,30

There are a number of factors that may account for the poor rates of RGC integration: 1) structural obstacles to the GCL, such as the retinal inner limiting membrane (ILM), 2) post-injection inflammatory response, 3) extracellular matrix (ECM), and 4) immune-mediated clearance of the transplanted cells.

The ILM, which serves as a vitreo-retinal barrier, is formed by astrocytes and end-feet of Müller glial cells that are covered in a thick basement membrane comprised of ECM proteins. The finding that greater than 99% of RGCs injected into the eye lay at the vitreous-ILM interface implicated the ILM as a possible barrier to RGC integration. ³¹ This notion is supported by ex vivo cultures which showed that the amount of mesenchymal stem cell (MSC) retinal integration correlated with the amount of ILM peeled. Interestingly, ILM peeling was also correlated with reduction in glial reactivity, which represents another barrier to cell integration, as discussed below. ³² In contrast, digestion of the ILM with collagenase and plasmin did not demonstrate a significant effect on MSC proliferation, survival or migration into the retina. ³²

In addition to the ILM, intravitreal injections have been found to elicit a gliotic reaction in the retina that may inhibit RGC integration. Astrocytes and Müller glia have been shown to upregulate glial fibrillary acidic protein (GFAP) and vimentin after intravitreal injections, creating an inhospitable environment for RGC integration. ³³ When the glial cell reaction was suppressed by aminoadipic acid treatment (AAA), there was a fourfold increase in MSC migration in vivo. ³⁴ This treatment, however, may not be feasible for translation to human application because AAA impairs normal physiological function of Müller cells. On a similar vein, Goldberg showed that RGC integration rates in ex vivo retinal cultures whose optic nerves were previously axotomized were superior in the 28-day group than earlier groups. ³⁰ They proposed that more effective RGC integration can be achieved in an RGC-depleted retina after the inflammatory response has cleared. This is supported by experiments by Singhal who successfully utilized anti-inflammatory and ECM-degrading matrix agents to increase integration. ³⁵

A less understood obstacle to RGC integration is the inhospitable ECM environment that develops in the retina as it matures. ECM molecules including CD44 and neurocan, a chondroitin sulfate proteoglycan (CSPG), have been shown to increase in concentration as the retina matures. ^36,37 Reducing levels of retinal CD44 and neurocan in MRL/MpJ mice was associated with enhanced integration of photoreceptors into the mature retina and increased regeneration. ³⁸ Singhal also saw significantly improved integration of Müller stem cells into all retinal layers when chondroitinase ABC (and concomitant immunosuppression) was used to degrade CSPGs. ³⁹ Thus, it may be worthwhile to consider ECM degradation treatment as an adjuvant to future RGC transplantation approaches.

Although the eye is an immune-privileged organ, it is possible that immune-mediated clearance of injected RGCs contributes to poor integration rates. Bull found immunosuppression to improve rates of transplanted cell survival. ⁴⁰

Upon successful migration into the GCL, another challenge to RGC integration involves the formation of appropriate synaptic connections with retinal amacrine and bipolar interneurons. Remarkably, Goldberg noted that RGCs that migrated into the GCL displayed normal RGC-like arboritic structures. ²⁸ Furthermore, using synaptic marker staining and light stimulation, they demonstrated the formation of novel and functional synapses between donor RGCs and host retina. These connections, however, were found to be weaker than native RGCs as increased stimulation was needed to elicit an action potential in transplanted RGCs.

Given experiences with photoreceptor transplantation, one concern regarding experiments reporting RGC integration is whether true RGC integration even occurs. Transplantation studies with GFP-labeled photoreceptors have found that rather than integrating, transplanted photoreceptors were transferring cytoplasmic material (i.e. green fluorescent tracing material) to endogenous photoreceptors, giving the false appearance of cellular integration. Although the mechanism is unknown, it is thought that photoreceptors transfer cytoplasmic material through exosome vesicular transport. ^41,42 This phenomenon of cytoplasmic transfer, however, is thought to be unique to photoreceptors who routinely shed their outer segments and does not to apply to RGCs. This argument is bolstered by reports from Goldberg who observed 1) morphological differences between the axon terminals of GFP labeled donor cells and mature host RGC termini and 2) only noted single nuclei in transplanted RGCs. ²⁸ Future studies involving interspecies transplantation (i.e. human iPSC-derived RGCs into rat retina) will be able to use markers of donor genetic material to demonstrate true integration.

c. Axon growth

As detailed above, when RGCs were injected into the vitreous, few integrated into the retina. Of the RGCs that successfully integrated into the GCL, many sprouted axons that grew towards the optic nerve head but few grew past the lamina cribrosa (~10%). ²⁸ Whether poor axonogenesis is a result of absence of external guidance cues or presence of inhibitory molecules is under active investigation. During development, surface proteins including ephrin-A and their respective ligands, EphA, are expressed in gradients within the superior colliculus and RGC population, respectively. ⁴³ These molecules have been implicated in directing axonogenesis during optic nerve development, helping establish the topographic map between the retina and superior colliculus. Experiments aimed at exploiting these pathways to recapitulate development and thereby direct optic nerve regeneration, however, have been met with limited success; for example, ephrin-A2 expression in the superior colliculus was found to have an inhibitory effect on RGC axon regeneration. ⁴⁴ A similar effect was reported with semaphorin-3A. ⁴⁵ Failure to direct optic nerve regeneration may stem from the fact that these pathways work in not only spatial but also temporal gradients, which are difficult to recapitulate in vivo.

Other approaches to promote axonogenesis include providing RGCs with neurotrophic support such as with nerve growth factor and brain-derived neurotrophic factor. ⁴⁶ The rationale for these approaches stem from the finding that neurons depend on retrograde transport of endosomes containing neurotrophic factors for survival. ⁴⁷ By providing neurotrophic support, RGCs can be kept alive longer, increasing the opportunity for axon regeneration. The problem however is that increased neuronal survival did not equate to new axon formation ^48,49. More targeted approaches aimed at resetting cell intrinsic programs to pro-axonogenic states have been met with greater success. Activation of growth-promoting molecules such as mechanistic target of rapamycin (mTOR) and Janus kinase/signal transducers and activators of transcription (JAK/STAT) have been associated with long distance RGC axon regeneration after optic nerve crush injury. ^50–52 There are a number of caveats, however, associated with these approaches. First, these therapies appear to benefit only a subset of RGCs: activation of mTOR or overexpression of onocomudulin and insulin-like growth factor 1 selectively promoted regeneration of alpha-RGCs which comprise only 6% of the total RGC population. ²³ Second, the efficacy of mTOR and JAK/STAT in directing optic nerve regeneration have only been demonstrated when their levels were upregulated before or concurrently with axon injury ^50,53, which may not be clinically practical. Moreover, given that RGC axon degeneration after crush injury occurs over the course of weeks ⁵⁴, it is difficult to determine whether these experiments promoted de novo axon growth or instead prevented axon degeneration. Third, axons have been observed to grow towards aberrant targets (e.g. into the contralateral optic nerve or back on itself) ^50,55,56, implying a need for not just cues that promote but also ones that direct growth.

The approach that our group is taking to overcome the problem of limited axonogensis is to use electric fields (EFs) to direct RGC axon regeneration. Often overlooked, electrical gradients co-exist with chemo-attractants and repellants and have been found to play an important role in directing tissue patterning during development ⁵⁷ and wound healing ⁵⁸. Recently, we demonstrated that RGC axons exhibit cathode-directed growth in vitro. ⁵⁹ In vivo, Borgens demonstrated that transected spinal cords of sea lampreys and guinea pigs who underwent continuous EF stimulation had more action potentials and more axons traversing across the transection site, respectively, than controls ^60,61. These approaches lost momentum, in large part, from an inability to translate well into large animal studies. Borgens applied a direct current, which is well known to be toxic above even minute levels, and others tried alternating current, which was found to be ineffectual. ⁶² The latter finding was not surprising given our work showed RGC axons making U-turns in response to 180 degree changes in electric field polarity, a condition that mimics alternating current. ⁵⁹ Since these studies, advances in electrical engineering have fueled the development of hybrid currents such as asymmetric charge-balanced waveforms that allow for both safe and effective activation neurons in the CNS. ⁶³ The efficacy of such currents on directing RGC axon regeneration in vivo are currently under active investigation in our lab and preliminary experiments show promising results. A major advantage that may be gained from taking an electrical over molecular approach is that EFs can override endogenous molecular directional cues: in vivo EF application redirected neural stem cells moving within the rostral migratory stream to novel targets ⁶⁴. This suggests that in vivo application of EFs may be able to override endogenous inhibitors of axon regeneration (e.g., myelin byproducts and glial scar). ⁴⁶

Alternative strategies for directing optic nerve regeneration include fabricated scaffolds and autologous grafts. Recently, 3D printers have been used to print micro- and nano-grooved scaffolds that can be used to direct the orientation of axonal growth. ^20,65,66 Other groups hope to exploit the regenerative capacity of the PNS by using sciatic nerve grafts to facilitate optic nerve regeneration. ⁶⁷ Ultimately, optic nerve regeneration is going to require a combinatorial approach with signals that provide not only the “drive” to grow but also directional cues that can “steer” axons towards proper targets.

d. New synapse formation

The question of whether regenerated axons will establish new, functional synapses with the diencephalon remains a significant question in the field of optic nerve regeneration. This task is further complicated by the necessity of maintaining the retinotopic map. Progress in this field has been hampered by limited long distance axonogenesis (see previous section). To circumvent this, Bei lesioned the optic tract just proximal to the superior colliculus. ⁵¹ They found that upregulation of mTOR and JAK/STAT or co-overexpression of osteopontin (OPN)/insulin-like growth factor (IGF1)/ciliary neurotrophic factor (CNTF) was associated with increased formation of functional synapses with the superior colliculus compared to controls. These mice, however, did not show improvement on visual aptitude testing. They found newly formed axons to be unmyelinated but that partial restoration of visual function could be achieved with administration of 4-aminopyridine (4-AP), a voltage-gated potassium channel blocker.

Other approaches to promote synaptogenesis exploit processes that directed synapse formation during development. In utero, before direct stimulation by light, spontaneous neural activity termed ‘retinal waves’ prime the developing visual system for eventual light stimulation. ⁶⁸ This organizational process is thought to be driven by voltage-dependent calcium influx that mediates transcriptional changes that then drives new synapse formation. After birth, visual stimulation refines established visual neural networks, which dictate retinotopy and ocular dominance. Retinal wave disruption, either pharmacologically or genetically, interferes with retinotopic mapping and eye specific segregation in the lateral geniculate nucleus and superior colliculus in mice. ⁶⁹ Building on this concept, recent data shows that high contrast visual stimulation in conjunction with mTOR activation enhanced RGC regeneration after crush injury and lead to partial recovery of visual function including the optokinetic reflex, the visual cliff test, and the looming avoidance response. ⁵³ In these experiments, Lim crushed one optic nerve and sutured the non-lesioned eye shut, forcing all light input to the visual cortex to come from the lesioned optic nerve. These mice demonstrated increased RGC regeneration, electrical activity and synaptogenesis compared to lesioned mice in which both eyelids were left open.

D. Current Clinical Trials

To date, few active stem cell-based clinical trials for optic nerve regeneration are listed on clinicaltrials.gov (Table 1). As emphasized above, this is because the field is still in early stages of development. This is well exemplified by the recent disastrous outcomes of patients with age-related macular degeneration who lost significant vision after intravitreal injection with adipose-derived stem cells. ⁷⁰ These trials were able to circumvent FDA regulation on the technicality that they were transplanting autologous cells. Rather than cell-replacement, current stem cell-based therapies likely work by providing neurotrophic support to host tissue.

Table 1:

Stem Cell-Based Clinical Trials for Optic Nerve Regeneration

Clinicaltrials.gov ID number Location	Treatment	Delivery Method	Condition	Status	Results/ Findings
NCT01920867 USA NCT03011541 USA NCT02638714 Jordan	Autologous Bone Marrow derived Stem Cells	Combination of retrobulbar, subtenon, intravitreal, intra-optic nerve, subretinal, and intravenous injection	Retinal Disease Macular Degeneration Hereditary Retinal Dystrophy Optic Nerve Disease Glaucoma	Enrolling by invitation	NAION: 73.6% of treated eyes gained vision with an average of 3.53 Snellen lines of vision improvement. 72 LHON: up to 35 letters of improvement by ETDRS 73

NAION: non-arteritic ischemic optic neuropathy; LHON: Leber’s Hereditary Optic Neuropathy; ETDRS: Early Treatment Diabetic Retinopathy Study

Although researchers and clinicians alike want the best for their patients who are debilitated by their vision loss, the community must proceed with caution when endorsing these trials to prevent public loss of faith in the potential of stem cell-based therapies.

E. Conclusion

In parallel to the progress being made to develop neuro-regenerative and cell transplantation-based strategies for optic nerve regeneration, researchers and clinicians together will need to determine who will be a good candidate for such therapies. Trans-synaptic retrograde and anterograde degeneration have been well described in the visual pathway. For example, patients with advanced glaucoma have documented degeneration in the lateral geniculate nucleus. ⁷¹ Patients with posterior visual pathway lesions have been shown to have correlated degeneration of their retinal nerve fiber layer, suggesting anterograde degeneration. ^72,73 Given this, the success of cell replacement-based therapies will depend on the health of the residual supporting environment. In other words, if we manage to regenerate the optic nerve in patients with advanced glaucoma, will there be neurons in the lateral geniculate nucleus for them to synapse with?

In considering the many endogenous barriers to axon regeneration in the CNS, it was interesting to learn that myelin, a lipoprotein well known for its insulating function, also robustly inhibits axon growth after injury. This finding raises the question of whether myelin serves to prevent axon growth in the non-traumatic state, assuring that neural networks, once established would not change? Prevention of axonogenesis in fully developed, healthy tissue may be important for preventing inappropriate rewiring and the formation of misguided, aberrant connections. Numerous examples of aberrant regeneration can be readily found in the PNS: reparative responses to severe damage to the oculomotor nerve such as from crush injury can result in miswiring between different muscle groups such as the levator palpebrae and inferior rectus, causing inappropriate co-contraction. Conversely, few examples of acquired CNS aberrancy have been reported in the clinical literature. ⁷⁴ Most examples of CNS synkinesis are found in congenital conditions (e.g. Joubert Disease) in which aberrant connections were likely made before myelination set in. ⁷⁵ Although myelination is associated with decreased CNS plasticity, does it in fact play a direct role in regulating the stability of neuronal networks? ⁷⁶ Is poor clearance of myelin from the CNS after injury a protective response to guard against aberrant regeneration? Given the strong anti-regenerative effect of myelin, successful CNS regeneration will likely require targeted blockade of these inhibitory signals.

Although cell replacement strategies are becoming a viable option for restoration of vision in patients blinded by advanced optic neuropathies, there are still a number of significant challenges that must be overcome before they can be safely implemented in the clinical setting. Technical advances that merge basic science with biomedical engineering to control intrinsic and extrinsic regulators of neuro-regeneration are essential to maintain the current momentum of translating RGC transplantation to the clinical arena.