Restoring Sight in Blind Mice

Reduced vision or blindness caused by retinal degeneration is generally due to the loss of photoreceptor cells in the retina. However, the retina is usually still connected to the visual cortex of the brain by other types of remaining neurons. It is therefore possible to improve vision by replacing the photoreceptor cells with a prosthesis that can convert light into electrical signals and stimulating the functional neurons of the retina.

However, existing prosthetic treatments are limited by the need of an external power supply, insufficient retinal stimulation or lack of long-term efficacy due to issues with the biocompatibility and reliability of the devices. Furthermore, the surgical implantation of such prosthetics is complex, making the technique less desirable if it is only a short-term solution.

In new work, a team of researchers in China has developed a retinal nanoprosthesis based on tellurium nanowire networks (TeNWNs) that can reportedly process both visible and infrared light, restoring vision loss and enhancing natural vision in the near-infrared (Science, doi: 10.1126/science.adu2987). The scientists say they were also able to implant the prosthesis into the subretinal space of mice and macaque monkeys using a method that is both safe and simple.

Device development

The researchers, led by Shuiyuan Wang, Fudan University, created TeNWNs by running the tellurium (Te), a light-sensitive element used as a semiconductor, through a chemical vapor deposition process. This causes the single-helical chain atomic structure of Te to form an interconnected network, with the typical wire thickness reaching about 150 nm. One benefit of this woven architecture, according to the scientists, is that it can be easily implanted into the subretinal space.

The team tested two different models of the device—both consisting of two semi-infinite electrodes and a Te channel with vacancy and substitution defects—to see how internal defects and external interfaces affected the generated photocurrent. One model used heavily doped Te electrodes, forming a homogeneous interface with the nanowires, while the second model used bulk gold (Au) electrodes, which create a heterogeneous interface. The Te‒Au interface was meant to serve as a representation of the heterogenous contacts that occur between TeNWNs and retinal cells when the device is implanted into the retina.

The test with the first model, which had no defects, showed weak photocurrent density (1 mA cm⁻²). The second model, which had defects and the heterogenous interface, exhibited a stronger photocurrent (the maximum density reaching 288 mA cm⁻²) that spanned the visible light and near-infrared range. The researchers believe this is due to the second model’s narrow bandgap and the strong optical absorption of Te.

In additional testing focused on photocurrent densities, Wang and colleagues found that densities increased linearly with laser intensities across the visible, first and second near-infrared ranges. They also say that the technology’s ability to convert photonic energy into electrical current distinguishes it from existing retinal prosthetics, which need an external power source.

Stimulating sight

They tested retinal ganglion cell activity in retinas from nonblind mice and from the blind mice with the attached prosthesis and found that the blind retinas showed activity in response to both visible and near-infrared illumination.

The team assessed photocurrent stimulation of the TeNWNs in three stages. First, they projected shapes and digits directly onto the TeNWNs using a 635-nm laser at 42.5 mW mm⁻² to see if the device could exhibit the patterns in the photocurrent distribution. Next, they repeated the test with various laser intensities and pulse durations while the TeNWN was in a saline solution. Finally, the researchers implanted the TeNWNs into retinas from blind mice. They tested retinal ganglion cell activity in retinas from nonblind mice and from the blind mice with the attached prosthesis and found that the blind retinas showed activity in response to both visible and near-infrared illumination. The nonblind mouse retinas, in contrast, showed ganglion cell activity only in response to visible light.

Once photocurrent stimulation was confirmed, the scientists needed to confirm if the information generated by the photocurrents—especially from the near-infrared—was making it to the brain. They implanted their technology into blind mice and performed electrophysiological experiments, which showed a similar neuronal response in the visual cortex to that of nonblind mice.

These findings were corroborated by behavioral studies in which the blind mice with the implant and nonblind mice were trained to lick water in response to a light being turned off. The now-not-so-blind mice had similar test results to the nonblind mice. In behavioral tests with near-infrared light, however, nonblind mice failed to show a response, while the mice with the TeNWNs implant did indeed react. Interestingly, the mice with the retinal prosthesis showed improvement in tests with both visible and near-infrared light over the course of seven days.

 To validate the biocompatibility, safety and feasibility of these retinal nanoprosthetics, the team implanted the TeNWNs into the retinas of macaque monkeys, which serve as a nonhuman primate model. They checked and tested the implants over the course of 112 days from the procedure and found no abnormalities or bleeding around the device and that the vasculature of the retina had come to overlay the device. As for efficacy, the implant did not impede the visual response from visible light, and near-infrared light elicited a response 19 days after implantation, which increased over the next 90 days.

“This successful animal study showcases the potential of this prosthesis to restore visible vision and expand augmented infrared perception for blind humans, and offer a safer, more effective and wider-spectrum solution than existing technologies,” said the team.