On October 5, the Nobel Assembly at the Karolinska Institute announced that the 2026 Nobel Prize in Physiology or Medicine will be awarded to Karl Deisseroth, Peter Hegemann and Georg Nagel. The prize recognizes "discoveries concerning light-gated ion channels and optogenetics," honoring work that laid the foundation for optogenetics, a technique for controlling neural activity with light.
What the official announcement recognized was the ability to make only a specific type of neuron respond to light and to alter its activity at a chosen moment. Neuroscience gained a way not only to observe the relationship between brain activity and behavior, but also to test directly what happens when particular cells are stimulated.
Since then, optogenetics has expanded from basic research on the mechanisms of memory to efforts to restore visual function. Looking separately at what each study actually confirmed, and under what conditions, gives a more accurate picture of what the technique has made possible.
Karl Deisseroth is affiliated with Stanford University and the Howard Hughes Medical Institute in the United States, Peter Hegemann with Humboldt University of Berlin, and Georg Nagel with the University of Würzburg in Germany. The prize honors the discovery and development of a technique that neuroscience has already used for many years.
A Protein From Algae Became a Switch for Neurons
The starting point was the way the single-celled green alga Chlamydomonas swims in response to light. Peter Hegemann and colleagues were investigating why an electric current arises in cells that receive light. Channelrhodopsin, identified through collaboration with Georg Nagel and others, combined the ability to sense light and the ability to pass ions in a single protein.
Following the 2002 report on channelrhodopsin-1, Georg Nagel and colleagues examined the properties of channelrhodopsin-2 in detail in a peer-reviewed paper in 2003. In the original paper published in PNAS, they had frog oocytes and mammalian cultured cells produce the protein, and measured as electric current the passage of cations through the cell membrane when light was applied.
There was no need to transfer the alga's entire complex intracellular signaling machinery. Introducing a single protein was enough for it to function as a light-gated ion channel.
When blue light strikes channelrhodopsin-2, the channel opens, cations flow into the cell, and the membrane potential changes. In neurons, this change can trigger action potentials, the brief electrical signals neurons use.
Because the path from receiving light to changing membrane potential is direct, activity can be controlled on the millisecond timescale at which neurons exchange information.
The protein alone cannot sense light, however. It also requires a light-absorbing molecule called retinal. In the frog oocyte experiments, retinal was added externally, but in the later experiments with mammalian cultured neurons, the amount already present in the cells was sufficient.
The fact that no large set of additional signaling molecules or accessory substances had to be introduced was a major advantage for applying the protein to neuroscience.
In 2005, Edward Boyden, Feng Zhang and colleagues in Karl Deisseroth's laboratory introduced the channelrhodopsin-2 gene into cultured neurons and showed that neuronal firing could be controlled with millisecond precision using light. The paper in Nature Neuroscience also lists Georg Nagel and Ernst Bamberg as co-authors.
Research clarifying how a light-sensing molecule works and research using it as a tool to manipulate neurons were joined here.
The history of optogenetics, however, does not rest on these three laureates alone. The Nobel Prize scientific background also describes earlier work by Gero Miesenböck and colleagues in 2002, who combined several molecules to make neurons respond to light, as well as demonstrations using channelrhodopsin by other research groups.
The prize can be seen as recognizing, within a field developed by many researchers, the discovery of channels that open directly in response to light and the work that turned them into a tool for neuroscience.
Choosing Which Cells to Activate, and When
A major strength of optogenetics is that it combines rapid control by light with cell selection through genetic methods.
Even within a single brain region, neurons with different roles are intermingled. From the results of stimulating a whole region alone, it is difficult to isolate which group of cells was involved in a particular behavior.
With optogenetics, only cells of a particular type, or cells belonging to a particular neural circuit, can be made to produce light-sensitive proteins. By varying the timing and duration of the light, researchers can compare, for example, activating cells just before a behavior with activating them during it.
Genetics selects which cells are manipulated, and light decides when. Being able to combine the two is what characterizes optogenetics.
Of course, methods for investigating the causal relationship between neural activity and behavior existed before optogenetics. Electrical stimulation allows fast manipulation, and studies that damage part of the brain, along with pharmacological interventions, have also been used to probe how the brain works.
It would therefore not be accurate to say that causal relationships could not be investigated before optogenetics.
What optogenetics greatly improved was achieving, at the same time, precision in selecting only the targeted cell population and the high temporal precision needed to match neural activity.
Applying a technique that works in a culture dish to the brain of a living animal also required a way to deliver light into the brain.
In a 2007 study, viruses were used to introduce the target gene into neurons, and light from an optical fiber placed in the brain stimulated the motor cortex of rodents while whisker movements were measured.
Because gene delivery and placement of an optical fiber in the brain are required, this method as a whole cannot be called "noninvasive." It is not a technique that can manipulate arbitrary neurons simply by shining light from outside on an untreated brain.
Tools were also added for suppressing neural activity, not just enhancing it. In 2007, a study was reported that suppressed neural activity using halorhodopsin, which transports chloride ions when it receives light. Later, channels that pass anions were also used.
Switching a single molecule, channelrhodopsin-2, does not allow both turning neural activity on and off at will.
A Memory Experiment Showed That Stimulation Can Elicit a Response
A 2012 memory study makes it easy to see what optical intervention in neural activity can reveal.
Xu Liu, Steve Ramirez and colleagues in Susumu Tonegawa's laboratory had neurons in the hippocampal dentate gyrus of mice that were active during fear learning produce channelrhodopsin-2. When the same group of cells was later reactivated with light in an environment different from where the learning took place, the mice showed more freezing behavior, in which they stop moving.
This peer-reviewed study was published in Nature.
In Figure 3 of the paper, the main experimental group of 12 mice was compared with a group of 12 trained without electric shocks and a group of 12 that did not express the light-responsive channel.
Averaged over five days of testing, freezing increased in the main experimental group while light was on, whereas no similar increase was seen in these control groups. The control experiments thus confirmed that the result is hard to explain as a mere reaction to the light itself.
What this result directly showed is sufficiency: reactivating the cell population that was active during learning can trigger the behavior corresponding to that memory.
To confirm necessity, that is, whether stopping the activity of that cell population makes it impossible to express the memory, a separate experiment would be needed.
From the result that stimulation produced a response alone, one cannot conclude that the cell group is the only place where the memory is stored.
What this study measured was also the behavior of mice. It was not an experiment that read subjective memories experienced by humans or freely wrote arbitrary memories.
There is also a gap between being able to elicit memory-related behavior by activating an artificially selected group of cells and explaining the entire process of natural memory recall.
Controlling neuronal firing in cultured cells, eliciting a fear response in mice, and restoring part of a patient's visual function are demonstrations at different stages.
Organizing representative original papers by subject, what was manipulated and what was actually measured shows the differences.
| Study / year | Subject | What was manipulated | What was measured |
|---|---|---|---|
| Georg Nagel et al., 2003 | Frog oocytes and mammalian cultured cells | Cells made to produce channelrhodopsin-2 and exposed to light | Current flowing through a light-gated cation channel |
| Edward Boyden et al., 2005 | Cultured mammalian neurons | Channelrhodopsin-2 expressed and light applied | Control of neuronal firing at millisecond timescales |
| Xu Liu et al., 2012 | Mice that underwent fear learning | Dentate gyrus cells active during learning restimulated with light | Increase in freezing behavior during illumination |
| José-Alain Sahel et al., 2021 | One patient with retinitis pigmentosa | Gene delivery to the retina combined with light-stimulating goggles | Detection and localization of objects, touching them and counting them |
This table summarizes the scope that each paper actually demonstrated. It does not compare effects obtained in different subjects on the same scale.
Moving from being able to make neurons fire, to changing animal behavior, to compensating for lost function in humans requires separate verification at each stage.
Restoring Vision Required Gene Delivery and Special Goggles
In humans, a case in which part of the visual function was restored in one patient with retinitis pigmentosa was reported in Nature Medicine in 2021. It is a peer-reviewed case report by José-Alain Sahel, Botond Roska and colleagues.
In a retina that had lost its photoreceptors, the method made the remaining retinal ganglion cells sensitive to light.
The research team injected an AAV vector carrying the gene into the eye, causing retinal ganglion cells to produce a light-sensitive protein called ChrimsonR.
They also used special camera-equipped goggles to capture the surrounding scene and project amber pulses of light onto the retina corresponding to areas where brightness changed. The retinal ganglion cells, now able to respond to light, receive that stimulation and send information to the brain.
This differs from a treatment that regenerates the lost photoreceptors themselves.
After repeated training, the patient, using the treated eye together with the special goggles, became able to locate and touch objects and count them.
According to an announcement from the IOB, which took part in the research, the patient succeeded in 36 of 39 trials in a task of finding and touching a notebook on a white table.
However, this figure is the result of one patient repeating the same kind of task, and does not indicate a treatment success rate in a patient population.
According to the abstract of the original paper, before gene delivery the patient could not visually detect objects even with the goggles, and after gene delivery could not detect them without the goggles.
When the patient was looking at objects, electroencephalogram measurements also confirmed activity in the visual cortex associated with object recognition.
In other words, what was confirmed is that part of visual function was restored under conditions combining gene delivery with a special device. Normal vision was not fully restored.
Turning a technique for studying brain circuits into a way of supporting patients' daily lives requires more than making cells respond to light.
It works only when technology for delivering genes to the target cells, a device that converts images into the necessary light stimulation, and training to make use of the new visual information are combined.
Future clinical studies will need to confirm whether similar improvements can be reproduced in other patients and whether the effect lasts over the long term.
If safety can be verified, the burden of the device and training reduced, and the ability to find objects in the laboratory extended into vision that is useful in daily life, optogenetics, which began with the way algae sense light, may develop further as a technology for compensating for lost sensory function.