Toward the Miracle of Retinal Reanimation

“Any sufficiently advanced technology is indistinguishable from magic.”

Arthur C. Clarke, “Hazards of Prophecy: The Failure of Imagination,” 1962

The restoration of vision to the blind is—quite literally—miraculous. The Catholic church long has recognized vision restoration as a miracle worthy of canonization, and descriptions of miraculous recovery of vision are found in nearly every major religion. As ophthalmologists, we perhaps have become a bit inured to this miracle; nearly every practitioner has had the privilege of restoring vision to individuals with cataracts, corneal disease, retinal detachment, and a host of other maladies. However, for those with hereditary or acquired retinal degeneration, the notion of restoring vision—reanimating the retina—in which the rods and cones have died still seems the stuff of miracles or at least science fiction.

Yet 4 approaches—optoelectronics, cell replacement, gene therapy, and photopharmacology—have seen remarkable progress in the past decade. These techniques exploit the persistence of retinotopically mapped, functioning retinal ganglion cells after degeneration of the rods and cones. By conferring light sensitivity to these cells through exogenous means, researchers have succeeded with each in restoring some visual function, at least in animal models of retinal degeneration. I wish to give a brief update for the practicing ophthalmologist of the progress toward this objective that has occurred in the past decade, as well as detailing some of the remaining challenges for achieving this noble goal.

Optoelectronics

Optoelectronics is the most advanced approach of the 4, because it has achieved proof of principle in humans and Food and Drug Administration approval. In this approach, a microelectronic device is implanted epiretinally or subretinally and directly stimulates the retinal ganglion cells, using local field potentials. The image is fed into the device by an exogenous camera. Exemplified by the ARGUS II implant (Second Sight Medical Products, Inc., Sylmar, CA), this approach has been shown to restore the ability to distinguish high-contrast letters in patients with advanced degeneration, as well as to improve mobility and independence.¹ Although the ARGUS II has only 60 stimulating electrodes (limiting resolution), newer-generation implants in development have at least a 10-fold increase. However, increasing electrode density to match the approximately 1.3 million retinal ganglion cells of the intact retina is problematic because the current necessary to depolarize cells extracellularly generates substantial heat. Present-day technologies also are challenged to provide power, which currently requires a hard-wired connection between the intraocular device and an extra-ocular power source. Additionally, implantation typically requires a challenging vitreoretinal surgery. Nonetheless, this approach has established the feasibility of retinal reanimation and its potential for vision restoration.

Cell Replacement Therapy

The biochemical code that directs stem cells to become ocular tissues largely has been delineated.² In 2011, Eiraku et al³ published the remarkable results of growing an entire primordial eyecup from a stem cell, establishing the potential of this approach. Retinal pigment epithelium is one of the more readily differentiated cell types, and nearly a dozen clinical trials are now underway to implant stem cell-derived retinal pigment epithelium in the subretinal space to treat dry age-related macular degeneration and other retinal degenerative diseases, such as Stargardt’s disease.⁴ Nascent photoreceptors and neural retina also can be differentiated, and Chao et al⁵ recently demonstrated survival and even ganglion cell projection to the brain in stem cell-derived retinal precursors implanted subretinally in primates. A recent remarkable study by Sanges et al⁶ highlights an alternative approach. This group transplanted hematopoietic stem and progenitor cells into degenerating mouse retinas and observed these cells to fuse with Müller glia cells and subsequently to reprogram into photoreceptors within the retina. (A similar, Müller cell-dependent mechanism is thought to underlie the ability of fish retinas to regenerate naturally.⁷) Although in animal models, restoration of some visual function (as measured by electroretinography) has been shown, demonstration of the correctness of synapse formation, long-term persistence of cells, and absence of immune responses for allograft transplants remains to be established. Additionally, some early reports of generation of photoreceptors in the outer retina now seem to be the result of transfer of proteins between cells, rather than persistence and differentiation of photoreceptors.⁸ Cell replacement therapy shows great promise; the ongoing trials with stem cell-derived retinal pigment epithelium cell replacement will provide valuable data as to the viability of this approach. However, significant additional basic science studies will be needed before photoreceptor-restorative approaches can be attempted in humans.

Opsin Gene Therapy

The discovery and characterization of channelopsins opened the era of optogenetics. Channelopsins are opsin-family proteins that, rather than stimulating a G-protein to signal, directly open or close cell membrane ion channels. Chlamydomonas channelrhodopsin—isolated from green algae—can depolarize mammalian ganglion cells directly when expressed by gene transfer, causing them to fire in response to light. Delivery of these proteins is accomplished through virally mediated gene therapy, a technology currently in numerous clinical trials for conditions including Leber congenital amaurosis, X-linked juvenile retinoschisis, achromotopsia, and choroideremia, among others. At least partial restoration of retinal function has been demonstrated using this approach in multiple mouse models of degeneration.⁹ Perhaps most promising is the virally-mediated gene transfer of rhodopsin (and, presumably soon, cone opsin) to bipolar cells. Achieved by 2 laboratories in 2015 in mice,^10,11 this approach offers recapitulation of some of the cell signaling pathways in the retina with a photopigment capable of high-yield quantum efficiency and marked signal amplification. (That these pigments work in cells they were not intended to be expressed in is yet another minor miracle!) Gene therapy-mediated opsin expression also has been used in nonhuman primates to correct the most common genetic vision defect of all, red-green color blindness.¹² Challenges for gene therapy include optimization of the viral vectors for efficient infection of the appropriate cell types (particularly after intravitreal rather than subretinal injection), the permanence of treatment (i.e., it may be difficult to retreat individuals and upgrades to new constructs may not be possible), maintaining long-term expression of proteins, and limiting inflammation after treatment. Nonetheless, this is a most promising avenue for vision restoration and has achieved the most impressive results in animals to date.

Photopharmacology

Photopharmacology is the newcomer to the group. Organic chemists have modified existing drugs with a molecular light switch (an azobenzene moiety) that creates a light-dependent steric confirmation.¹³ Existing drugs thus may become light-or dark-activated. When linked to a potassium channel blocker (which will lead to depolarization and firing of retinal ganglion cells), this allows cell firing to be driven by light, thus achieving the same result as optoelectronic or opsin gene therapy approaches. Variants include the 1 component (i.e., photoswitched drug alone) or 2 component (gene therapy with a modified channel or receptor that reacts only to photoswitched compound) methods. These approaches have restored light responsiveness and some visual function to mice and rats blind from outer retinal degeneration.¹⁴ Newer-generation compounds feature long half-lives (on the orders of weeks)¹⁵ and activation of upstream cellular components such as bipolar cells.¹⁶ Advantages of this approach include the need only for small-molecule treatment (in the 1-component system), the reversibility or upgradability of compounds (whereby the optimal agent and its dosing could be titrated for the individual), and a more familiar small-molecule Food and Drug Administration approval path. Challenges include maintaining continuous delivery and open questions on long-term toxicity.

Common Challenges

Perhaps the largest challenge facing all these technologies is our collective lack of understanding of the neural code. The retina performs substantial local processing of the nascent image, extracting information about motion, direction, color, contrast, and intensity. The question of how these attributes are encoded in the firing of approximately 1.3 million retinal ganglion cells remains fundamentally unsolved. Initial attempts to replicate the natural encoding of vision in mice have shown promising results,¹⁷ but until the human code is cracked, attempts at retinal reanimation likely will not restore normal vision, and instead will create usable information for the cerebral cortex that can be interpreted to allow function. Such has been the case with the cochlear implant—the pioneer of sensory restoration—with which recipients report the auditory experience as being distinct from native hearing, but sufficient for functional hearing restoration. Other challenges common to these approaches include the central scotoma problem (methods restoring ganglion cell function will create a central scotoma because of the slight displacement of retinal ganglion cells from the fovea), the color-encoding challenge (there is no way at present to determine which ganglion cells encode which color channels), and the retinal rewiring problem (whereby substantial remodeling of intraretinal connections occurs in outer retinal degenerative disease,¹⁸ a problem that is particularly challenging for cell replacement technologies). None of these hurdles are insurmountable, but substantial research will be needed to address each. Finally, there is the issue of cost. The ARGUS II implant costs more than $100 000 (exclusive of surgical implantation costs). Gene therapies in the past have been priced even higher; Glybera (alipogene tiparvovec, [uniQure, Amsterdam, Netherlands], a therapy for lipoprotein lipase deficiency) was the first approved AAV gene therapeutic (approved in Europe in 2012), at a cost of $1 million per treatment (!). The value of vision restoration to the patient will be enormous, as will the value to society. However, the market forces that determine its pricing (at least in the United States) are not yet clear. Cost could prove a barrier to advancement or adoption of promising technologies.

Although the optoelectronic approach to vision restoration has an almost 25-year history, all the remaining approaches are less than a decade old. The progress in this domain has been remarkable, and it is likely that several of these approaches will be in clinical trials before the next decade dawns. These approaches offer great hope to our patients with currently untreatable retinal disease and eventually will bring treatments to one of the few areas of ophthalmology that have no treatment options at present.