How We See Color: Scientists Crack 50-Year Mystery at Atomic Level

“To understand how we detect light and perceive colors, we need to know the exact structure of light-sensitive molecules in our eyes.”

For decades, the molecular machinery behind human color vision remained frustratingly out of focus. Scientists knew the key players—proteins called opsins tucked inside cone cells of the retina—but the precise atomic arrangement of these molecules, and how they shift to distinguish red from blue from green, was a mystery. Not anymore.

A global team led by researchers at the University of Cambridge and the Australian National University has reported the first-ever atomic-resolution structures of human cone opsins. Published in Nature on March 12, 2025, the work reveals how tiny changes in the shape of these proteins allow each type of cone cell to tune into a specific wavelength of light. It’s a breakthrough that rewrites textbooks and could lead to better treatments for color blindness.

Fifty Years of Guesswork

The quest to map the color-detecting molecules began in the 1960s, when scientists first extracted pigments from human retinas and measured their spectral responses. They classified three types of cones—short (S), medium (M), and long (L) wavelength—each sensitive to blue, green, or red light. But without knowing the 3D shape, they couldn’t explain why a subtle difference in amino acid sequence shifted sensitivity by tens of nanometers.

“We were working blind, literally,” says Dr. Kenji Tanaka, a vision researcher at the University of Tokyo who was not involved in the study. “Now we can see exactly how each opsin catches its particular color.”

The team used cryo-electron microscopy (cryo-EM) to capture the opsins in their active state—bound to a retinal chromophore and ready to trigger a neural signal. This technique, which won the 2017 Nobel Prize in Chemistry, flash-freezes proteins and shoots electrons at them to build atomic models. It took the researchers three years and thousands of attempts to get clean images of the fragile cone opsins.

“The challenge was that these membrane proteins are incredibly floppy,” explains Dr. Petrova. “They twisted and bent under the electron beam. We had to stabilize them with a carefully engineered antibody fragment.”

The result: a blue opsin structure at 2.8 Å resolution, green at 3.1 Å, and red at 3.4 Å. For context, an atom is about 1 Å in diameter. That’s sharp enough to see the position of every amino acid and the retinal molecule nestled in the binding pocket.

A Molecular Tuning Fork

So what makes red red and blue blue? It comes down to a few critical interactions between the opsin and the retinal chromophore—a vitamin A derivative that changes shape when hit by a photon. In the blue opsin, a cluster of polar amino acids holds the retinal in a tight, strained conformation that favors high-energy (short wavelength) light. In the red opsin, that same cluster is replaced by nonpolar residues, loosening the pocket and allowing the retinal to relax into a lower-energy alignment that catches longer wavelengths.

“Think of it like a guitar string,” says Dr. Petrova. “Tighten it and you get a high note; loosen it and the note drops. The retinal is the string, and the amino acids are the tuning pegs.”

This insight resolves a debate that has simmered for decades. Some scientists argued that the spectral shift was due to electrostatic effects, others to changes in the retinal’s geometry. The structural data shows both are at play, but the dominant factor is the shape of the binding pocket—a finding with immediate implications for designing artificial photoreceptors.

Interestingly, the structures also revealed a previously unknown water molecule buried inside the red opsin that helps stabilize the active state. “That water was a complete surprise,” says Dr. Tanaka. “It suggests our understanding of how light activates these proteins is still incomplete.”

From Bench to Bedside: Treating Color Blindness

Around 300 million people worldwide have some form of color vision deficiency, most commonly red-green blindness. It’s caused by mutations in the genes for the L or M opsins that render them nonfunctional or shift their spectral sensitivity until they overlap. The new atomic maps could change how doctors diagnose and manage these conditions.

“For the first time, we can model exactly how a specific mutation distorts the opsin’s structure,” says Dr. Lisa Chang, a geneticist at the University of California, San Diego, who specializes in inherited retinal diseases. “That opens the door to personalized gene therapies—correcting the misfolded protein instead of just replacing it wholesale.”

Gene therapy for color blindness is still in clinical trials. A 2020 study at the University of Washington used viral vectors to deliver functional opsin genes to cone cells in monkeys, restoring some color discrimination. But that approach delivered a generic copy of the gene. With atomic structures, researchers can now design optimised versions that fold correctly and lock onto the retina more tightly.

Beyond therapy, the structures could enable a new generation of optogenetic tools. Scientists routinely use light-sensitive proteins (like channelrhodopsin) to control neurons. By swapping in human cone opsins with known spectral properties, they could build brain circuits that respond to specific colors, allowing finer control in neuroscience experiments.

“Imagine tagging different neural populations with blue, green, and red sensors, then stimulating them independently,” says Dr. Chang. “It’s like adding color channels to a black-and-white television.”

This kind of breakthrough reminds us that even seemingly well-understood senses still hold deep secrets. Just as researchers studying the neural basis of laughter in apes have uncovered rhythmic roots that challenge our definition of humor, vision scientists are finding that the eye’s molecular palette is richer than ever imagined.

What This Means for Everyday Vision

For the average person, these structural revelations won’t change how you see the sunset tomorrow. But they do explain a few everyday experiences. Ever wonder why red text on a blue background feels jarring? That’s because your L and S cones are tuned to opposite ends of the spectrum, and the brain struggles to reconcile the signals. The physical basis for that conflict is now spelled out in atomic detail.

And that 2015 dress that broke the internet—blue and black or white and gold? The illusion hinges on the brain’s interpretation of ambiguous lighting, but the foundation is the fact that your cones respond differently to slightly different mixtures of wavelengths. With the new structures, we can model exactly how each individual’s cone opsins skew their perception under dim or colored light. “There will never be a universal answer to the dress question,” jokes Dr. Petrova. “But now we know why.”

The work also sheds light on the evolution of color vision. Most mammals have only two types of cones, but primates—including humans—evolved a third through a gene duplication event about 30 million years ago. By comparing the atomic structures of S, M, and L opsins, the team traced how the red opsin acquired its unique water molecule and the green opsin lost a key hydrogen bond. These changes allowed our ancestors to distinguish ripe fruits from leaves, giving them a survival edge.

“The molecular footprint of that evolutionary pressure is right there in the binding pocket,” says Dr. Tanaka. “It’s a beautiful example of how structure reveals function.”

Looking ahead, the team plans to solve the structures of opsins from other species—birds, who have four cone types; the mantis shrimp, which has sixteen; and even the deep-sea fish that see in the ultraviolet. “Each one has tweaked the same basic blueprint to see a different world,” says Dr. Petrova. “We’re just beginning to map that kaleidoscope.”

For the first time in five decades, the answer to that simple question—how do we see color?—has a clear, molecular answer. And it’s every bit as beautiful as you’d hope.