Surveys consistently report that people fear total blindness more than any other disability, and currently the major cause of untreatable blindness is retinal disease. The retina, a part of the brain that extends into the eye during development, initiates vision by first detecting light with the rod and cone photoreceptors. Four classes of retinal neurons then begin the analysis of visual images. Defects in the optical media that transmit and focus light rays onto the retina (lens and cornea) can usually be dealt with surgically, although such treatments are not available in some parts of the world, resulting in as many as 20 to 30 million legally blind individuals worldwide. Untreatable retinal disease potentially causes legal or total blindness in more than 11 million people in the United States alone, but progress in treatments raises the possibility of restoring vision in several types of retinal blindness (1).

Retinal neurons comprise bipolar and horizontal cells, which are second-order neurons that receive signals from the photoreceptors in the outer retina. Third-order amacrine and retinal ganglion cells are activated in the inner retina by bipolar cells. Axons from the ganglion cells form the optic nerve and carry the visual message to the rest of the brain (see the figure). The cells most susceptible to blinding retinal disease are the photoreceptors and ganglion cells. Whereas progress has been made in combating blindness caused by photoreceptor degeneration, little can be done currently to address ganglion cell loss, such as occurs in glaucoma.

The approach that has been most successful in restoring photoreceptor loss that results in complete blindness is the use of retinal prosthetic devices, with two now approved for clinical use (2). These devices electrically stimulate either bipolar or ganglion cells. They require goggles that have a camera that converts visual stimuli into electrical stimuli that activate the device, which in turn stimulates the retinal cells. Several hundred of these devices have been implanted in blind or virtually blind individuals, 70 to 80% of whom report improvement in quality of life. For those who are completely blind, the ability to experience again at least some visual function is viewed as a miracle.

There are substantial limitations to the devices, however. The best visual acuity attained so far is poor (20/500) and visual field size is limited, but many improvements, mainly technical, are being developed and tested, including the potential use of electronic low-vision devices to increase visual field size and acuity (3). Retinal prostheses are not useful for patients who are blind because of loss of ganglion cells and/or the optic nerve, but prostheses that bypass the retina and stimulate more central visual structures, including the lateral geniculate nucleus (the intermediary between retina and cortex) and visual cortex, are being developed and tested in humans (4). There remain considerable technical issues, but preliminary data indicate that such devices are feasible.

A second approach to treat photoreceptor degeneration and potential blindness, now in the clinic, is gene therapy (5). This involves injecting a viral construct into the eye that contains a normal gene to replace an abnormal one. Success so far has been limited to the treatment of Leber congenital amaurosis (LCA) type 2, a rare form of retinitis pigmentosa in which the gene whose product is required to form the correct isomer of vitamin A aldehyde, the chromophore of the visual pigments, is mutated. Little of the correct isomer is made in LCA patients, resulting in substantial loss of photoreceptor light sensitivity. This is reversed when viral constructs encoding the normal gene are injected deep into the eye between the photoreceptors and pigment epithelium.

Two factors make this approach feasible in LCA: The genetic defect is monogenic, and many of the photoreceptors in the patients remain alive, although compromised. Thus, how broadly feasible gene therapy will be for treating the enormous range of inherited retinal diseases now known to exist (300) remains to be seen. But at least a dozen other gene therapy trials on monogenic inherited eye diseases similar to LCA are under way (6). Other methods to manipulate genes are now available, including CRISPR-mediated editing of retinal genes. So far, the experiments have been mainly on isolated cells or retinas, but these powerful techniques are likely to have eventual clinical applications.

A variation on the use of gene therapy techniques is optogenetics, in which light-sensitive molecules are introduced into non-photosensitive retinal cells. This approach holds much promise for restoring vision to totally blind individuals whose photoreceptors have been lost. Using viruses to insert genes encoding light-sensitive molecules into bipolar and ganglion cells, as well as surviving photoreceptor cells that are no longer photosensitive, has been accomplished in animals and shown to restore some vision (7). Again, technical issues remain: The cells made light-sensitive require bright light stimuli, and the light-sensitive cells do not adapt. That is, whereas photoreceptors normally allow vision over as much as 10 log units of light intensity, the cells made light-sensitive respond only to a range of 2 to 3 log units. Various methods to overcome these limitations are now being developed, and at least one clinical trial is under way. Experiments to make cortical neurons sensitive to light or other stimuli that better penetrate the skullmagnetic fields or ultrasound, for exampleare also being developed and tested in animals.

Other promising approaches to restore vision are being explored. In cold-blooded vertebrates, retinal cells (in fish) and even the entire retina (in amphibians) can regenerate endogenously after damage. Regeneration of retinal cells in zebrafish is now quite well understood (8). The regenerated neurons come from the major glial cell in the retina, the Mller cell. After retinal damage, Mller cells reenter the cell cycle and divide asymmetrically to self-renew and produce a progenitor cell that proliferates to produce a pool of cells capable of differentiating into new retinal cells that repair the retina.

A number of transcription factors and other factors identified as being involved in retinal regeneration in zebrafish have been shown to stimulate some Mller cell proliferation and neuronal regeneration in mice. Regenerated bipolar and amacrine cells, as well as rod photoreceptors, have so far been identified in mouse retinas, and these cells are responsive to light stimuli (9, 10). Further, cells postsynaptic to the regenerated neurons are activated by light stimuli, indicating that the regenerated neurons have been incorporated into the retinal neural circuitry. So far, the regenerative capacity of mammalian Mller cells is limited, but directed differentiation of specific types of neurons with a mix of factors appears to be a possibility. Regrowth of ganglion cell axons after the optic nerve is disrupted is also under active investigation, and although the number of axons regrowing is low (10%), those that do regrow establish synaptic connections with their correct targets (11). Therefore, endogenous regeneration is still far from clinical testing, but substantial progress has occurred.

The retina lines the back of the eye and consists of rod and cone photoreceptors, as well as four types of neuron: second-order bipolar and horizontal cells and third-order retinal ganglion cells (RGCs) and amacrine cells. Mller glial cells fill the spaces between the neurons. The pigment epithelium, critical for photoreceptor function, underlies the retina. Photoreceptors and RGCs are most susceptible to blinding retinal disease. Progress in combating photoreceptor degeneration has been made, but there are few strategies to address RGC loss.

A long-studied area of research is transplantation of retinal cells, particularly photoreceptors, into diseased retinas. In experiments with mice, transplanted postmitotic rod photoreceptor precursor cells derived from embryonic retinas or from stem cells appeared to integrate into diseased retinas in reasonable numbers and to be functional. A surprising and unexpected complication in the interpretation of these experiments was recently discovered. Rather than integrating into diseased retinas, the donor cells appear to pass material (RNA or protein) into remaining host photoreceptor cells, rejuvenating them, and these appear to be most of the functional cells (12). The current evidence suggests that only a small proportion of the donor cells integrate, but progress in overcoming this setback is being made.

More success has been reported with stem cells induced to become pigment epithelial (PE) cells, which provide essential support for photoreceptors. A number of blinding retinal diseases relate to the degeneration of the PE cells, and replacement using such cellsin a suspension or on a scaffoldis being actively pursued. PE cells do not need to integrate synaptically with retinal cells; they simply need to contact the photoreceptor cells. This is achieved when PE cells are placed between the retina and the back of the eye. Early clinical trials suggest that the transplants are safe, but retinal detachment, a serious complication, can occur and efficacy has yet to be shown (13).

The finding that donor photoreceptor cells can help diseased host retinal cells to recover function suggests that certain substances can provide neuroprotection. Indeed, a substantial number of such neuroprotective molecules have been shown to affect retinal disease progression, especially degeneration of photoreceptor cells. No one factor has been shown to be effective generally, but two have received much attention. One, ciliary neurotrophic factor (CNTF), promotes photoreceptor survival in light-induced photoreceptor degeneration and in several other models of retinal degeneration (14). Some evidence suggests that CNTF acts primarily on Mller cells, but how it works, and on what cells, is still unclear. The other factor, rod-derived cone viability (RDCV) factor, has received less research attention, but with recent industrial support, it is now being advanced to the clinic. Current evidence indicates that RCDV factor protects cones after rod degeneration.

Two of the most common retinal diseases in developed countriesage-related macular degeneration (AMD), the leading cause of legal blindness (visual acuity of less than 20/200), and glaucoma, the leading cause of total blindnessare not monogenic diseases, and so genetic treatments for them are not obvious. Attempts to understand the etiology of these diseases are under way, but currently their underlying causes are still unclear. A difficulty presented by AMD is that no animal model is readily available, because it is a disease of the fovea, which mediates high-acuity vision. Except for primates, other mammals do not possess this small critical retinal area. Whereas large primates are not feasible for extensive cellular or molecular studies, small primates such as marmosets that have a fovea are potential models but have not been used much to date.

Other approaches for restoring vision have been suggested and have even yielded some progress. From both normal humans and those with an inherited retinal disease, skin biopsy cells can be induced to form tiny retinal eyecups called organoids (15). Containing all retinal cell types, these structures could be a source of retinal cells for studying retinal disease development and possible therapies, as well as for cell transplantation. A fovea has not been observed in any organoid so far, but this is not beyond the realm of possibility. Another suggested approach is to surgically transplant whole eyes into blind individuals. This appears feasible, but whether there is sufficient optic nerve regrowth remains an open question.

Read the original:
Restoring vision to the blind - Science Magazine