Science1 publisher2 min readPublished
A blind mouse's visual cortex responds to light after an injection of carbon nitride particles
An Aarhus-led team injected 300-nanometre semiconductor particles into mice with advanced retinitis pigmentosa and recorded light-driven activity in the visual cortex, a route to restoring light sensitivity that does not depend on the patient's mutation.
The Scientist · Science desk

What happened
- The particles are hollow graphitic carbon nitride spheres roughly 300 nanometres across, built with an architecture borrowed from plant chloroplasts to harvest visible light.
- The multi-institutional study, led by Associate Professor Menglin Chen at Aarhus University, was published in Nature Biomedical Engineering.
- Injected into the eyes of mice with advanced retinitis pigmentosa, the particles settled on the retinal surface and came to rest close to the retinal ganglion cells.
- When the researchers illuminated those eyes, they recorded clear evoked electrical activity in the animals' visual cortex.
- The same particles stimulated ganglion cells in isolated retinal tissue taken from pigs, which the team used as a cross-species check.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- capability One formulation would apply to a degenerated retina whatever mutation caused the degeneration, because the particles work on cells the disease leaves behind.
- decision A group weighing this against a gene therapy limited to known mutations, or against optogenetics with its viral delivery step, is comparing scope claims; no acuity figures have been published to settle the efficacy comparison.
- exposure Delivery by needle moves the risk from a surgical one to a materials one, and that material sits for months on the retina the treatment is meant to serve.
Graphitic carbon nitride is a semiconductor, and under visible light it does chemistry. The Aarhus group describes the illuminated particles as inducing local physical and chemical reactions in their immediate environment, and those reactions generate the signals that stimulate the living cells beside them [8]. Retinal ganglion cells are the conduits that relay visual information to the brain [10], so a signal produced at that layer carries through to the cortex.
The design aims there because of how the disease runs: in retinitis pigmentosa the light-capturing photoreceptors deteriorate and die, while the deeper retinal neurons often stay healthy and functional [9]. The engineering problem is handing light to cells that were never built to detect it. "Instead of genetically modifying these cells, we use the nanoparticles to create a new connection between light and the nerve cells," said Menglin Chen, the Aarhus associate professor who led the work [14].
The mice also changed their behavior when the light came on [6]. That is a detection result. Acuity, contrast and how fine a pattern the particles can write are separate measurements, and the Aarhus paper does not report how many animals were used, the light intensity needed to fire the cells, or how long the particles keep working after injection [18].
Mutation-independence here follows from where the particles act, on the surviving native neurons of a degenerated retina, whatever mutation degenerated it [17]. The material lets that interface be delivered by injection [1].
"That brings us closer to our long-term goal of developing a new type of retinal prosthesis," he said [16]. Electronic implants, the incumbent in that category, require invasive ocular surgery and bulky hardware [11]. A suspension delivered through a needle avoids both from the start. Nobody has yet shown that the particles stay safe and functional on a living retina over months.
"Once the photoreceptors are lost, the options for restoring light sensitivity are still very limited, and each approach in development carries its own constraint," said Henri Leinonen, a co-author and retina specialist [15].
What to watch
- The Nature Biomedical Engineering methods: animal numbers, the irradiance needed to evoke cortical responses, and how many days a single injection keeps working.
- Histology after chronic illumination, and whether the particles clear from the eye or accumulate on the retinal surface.
- A replication of the cortical recordings by a group outside the collaboration, in a second model of photoreceptor degeneration.