Nobel Prize in Physiology or Medicine 2026
Channelrhodopsin: a protein that turns a quantum event into a nerve impulse
One photon twists one molecule in a few hundred femtoseconds. The protein around it decides the colour, steers the twist and converts it into a current that fires a neuron.
The 2026 Nobel Prize in Physiology or Medicine went to Karl Deisseroth, Peter Hegemann and Georg Nagel for light-gated ion channels and optogenetics.
Twelve orders of magnitude in time, one protein
Channelrhodopsin-2 comes from the eyespot of the green alga Chlamydomonas reinhardtii. Inside it sits retinal, the same chromophore your eyes use, bound to a lysine as a protonated Schiff base. What makes it a quantum device is the hand-off between three very different physics regimes.
- ~2 fsAbsorption
A 470 nm photon lifts a π electron across retinal's HOMO–LUMO gap. The nuclei have no time to move.
- ~300 fsTwist through the funnel
On the excited surface the C13=C14 bond loses its double-bond character, twists, and drops through a conical intersection to 13-cis.
- ~1 msPore opens
The strained chromophore pushes the helices apart. Cations flow in and depolarise the neuron toward its firing threshold.
The protein sets the colour
Retinal's π electrons are delocalised along its conjugated chain, so to a first approximation they behave like particles in a one-dimensional box of length L. Filling levels two electrons at a time, the lowest excitation has energy
ΔE = (N + 1) h² / 8 mₑL², λ = 8 mₑ c L² / (N + 1) h
The free-electron model puts retinal near 600 nm. In methanol the same chromophore absorbs near 440 nm, in channelrhodopsin-2 near 470 nm, and in bacteriorhodopsin near 570 nm. That spread, the opsin shift, is the protein at work: the negative counter-ions near the Schiff-base nitrogen and the polar residues lining the pocket stabilise the ground and excited states by different amounts.
For protein engineers this is the first design lever. Red-shifted channelrhodopsins reach deeper into tissue, and they are made by moving charges around an unchanged chromophore.
Retinal has 6 (five C=C and the C=N). The model ignores bond alternation, end effects and the protein, which is why it lands redder than 470 nm.
Through the funnel: why the twist is so efficient
Near the crossing, two electronic states mix through a coupling V. The adiabatic surfaces are the eigenvalues of a 2 × 2 Hamiltonian,
E± = (E₁ + E₂)/2 ± √[ ((E₁ − E₂)/2)² + V² ]
and they touch only where E₁ = E₂ and V = 0 at once. With two internal coordinates that point is the tip of a double cone: a conical intersection. Passing a crossing at sweep rate α, the Landau–Zener formula gives the probability of jumping between adiabatic surfaces,
P = exp(−2πV² / ħα)
At a conical intersection V vanishes, P approaches 1, and the wavepacket goes straight from S1 to S0 in a single pass. The protein's job is to place the funnel where the twist leads and to bias which valley the molecule falls into.
Reduced units with ħ = 1. Shrink V toward zero to turn the avoided crossing into a conical intersection.
After the twist: a proton relay
Isomerization is only the first move. In the following microseconds to milliseconds, the Schiff-base proton leaves the nitrogen for a nearby carboxylate counter-ion and travels along a chain of residues and bound water. The deprotonated intermediate absorbs in the near ultraviolet, around 390 nm, before the channel reaches its conducting state.
Proton transfer along hydrogen-bonded networks is where quantum effects in proteins are argued about most: zero-point energy, isotope effects and possible tunnelling contributions. Channelrhodopsin offers a well-characterised system to test those ideas, with a light switch that starts every event at a known time.
Rhodopsin: from photosensing to neuroscience
Rhodopsins are a large family of proteins built on the same quantum switch: retinal, bound to a lysine, isomerised by light through a conical intersection. Animal visual rhodopsin twists 11-cis retinal to all-trans in about 200 fs, with a quantum yield near 0.65, and then signals through a G-protein cascade. Microbial rhodopsins such as bacteriorhodopsin and channelrhodopsin run the reverse switch, all-trans to 13-cis, and act directly as pumps or channels. The physics is shared; the protein scaffold decides what the twist is used for.
Channelrhodopsin came from a question about behaviour. Peter Hegemann wanted to know how the single-celled alga Chlamydomonas reinhardtii swims toward light. In the early 2000s he and Georg Nagel found a protein on the cell surface that opens an ion channel when illuminated with blue light. No second messengers are involved: the protein that absorbs the photon is the protein that conducts the current. They also found that any cell expressing it became light-sensitive.
Karl Deisseroth put the channelrhodopsin gene into rat neurons and triggered nerve signals with blue light, published in 2005. Two years later the same switch worked in the brains of living mice. Optogenetics lets researchers activate chosen neurons with millisecond timing and test causes, not just correlations, in circuits governing memory, feeling and behaviour. Clinical groups are now testing it to restore some sight in people with retinal degeneration.
“Optogenetics provides opportunities for mapping the brain in a way that we could once only dream of.” — Per Svenningsson, Chair of the Nobel Committee for Physiology or Medicine
The 2026 Nobel Prize recognises a tool whose first step is a quantum event: one photon, one isomerization. It is a concrete example of why it pays to understand proteins as quantum devices.
Why this belongs on Quantum Proteins
Channelrhodopsin is the cleanest established case of the platform's central idea: a protein that shapes a quantum event and turns it into classical biological function. Ultrafast spectroscopy has measured the quantum step directly, so it anchors the more exploratory questions we pursue elsewhere, from tunnelling in enzymes to the energy landscapes of folding. See the open problems and the theory behind them.
Learn the full physics
The companion course on courseshub.world derives every step here from first principles: the particle-in-a-box colour, the two-state model of the conical intersection, and the membrane as an RC circuit, with Python simulations and worked exercises.
Key papers
- Nagel G. et al. Channelrhodopsin-1: a light-gated proton channel in green algae. Science 296 (2002).
- Nagel G. et al. Channelrhodopsin-2, a directly light-gated cation-selective membrane channel. PNAS 100 (2003).
- Boyden E. S., Zhang F., Bamberg E., Nagel G., Deisseroth K. Millisecond-timescale, genetically targeted optical control of neural activity. Nature Neuroscience 8 (2005).
- Nobel Prize in Physiology or Medicine 2026, press release.
The 3D model and energy curves are schematic teaching models, not quantum-chemical calculations.