Color is the starting point of how we perceive the world. From the blazing sunflowers in Van Gogh’s paintings to the shifting light and shadow in Monet’s water lilies, our experience of beauty and understanding of nature depend to a great extent on our distinctive trichromatic vision.Yet most mammals, including domestic cats and dogs, are actually dichromats, and the world they see is far less colorful than ours. So why can humans perceive red, green, and blue? When and how did this extraordinaryability evolve? The answer resides in the three cone visual pigments within the human retina.Now, a collaborative team led by Professor Zhang Jin from the School of Basic Medical Sciences and the Second Affiliated Hospital of Nanchang University, in partnership with Dr. Patrick Scheerer's group at Charité, Universitätsmedizin Berlin and other international researchers,has unveiled this mystery. Their findings appear as the cover article in Science, titled “Cryo-electron microscopy structures of human cone visual pigments,” with Nanchang University as the primary affiliation. This landmark studyreveals the molecular mechanism of human trichromatic vision and provides a structural basis for future therapies targeting color blindness and cone-related color-vision deficiencies.

I. From Tetrachromacy to Dichromacy: The “Downgrading” of Mammalian Vision

Three hundred million years ago, the vertebrate ancestor possessed tetrachromatic vision. Four cone visual pigments—LWS for red, SWS1 for violet or ultraviolet, SWS2 for blue, and RH2 for green—worked together. This "full-color system" persists today in birds and reptiles.However, the rise of mammals took a different turn. During the Late Triassic, the age of dinosaurs, early mammals adopted a nocturnal lifestyle, and in doing so, they first lost RH2, the green-sensitive pigment. Later, during the evolution of marsupials and eutherian, or placental, mammals, they also lost SWS2, the blue-sensitive pigment. Mammalian vision ultimately regressed to a dichromatic system containing only LWS, sensitive to red or long wavelengths, and SWS1, sensitive to ultraviolet or short wavelengths.This is the fundamental reason why dogs, cats, and most other mammals are red–green color-blind.

II. The Primate “Comeback”: The Birth of Trichromatic Vision

After the extinction of the dinosaurs approximately 65 million years ago, primate ancestors returned to a daytime lifestyle and began feeding on fruit and young leaves. The survival advantage of distinguishing ripe red fruit from unripe green fruit created strong selective pressure.Trichromatic vision provided primates with a major evolutionary advantage. The most widely accepted explanation is the “foraging hypothesis”: individuals with trichromatic vision could detect brightly colored ripe fruit more efficiently among dense green foliage.Studies have shown that trichromatic individuals are better than dichromatic individuals at detecting red or orange food. Long-term observations of rhesus macaques have found that trichromatic females locate and consume fruit significantly faster.In fact, the evolution of trichromatic vision was closely linked to the evolution of fruit coloration. Over long periods of time, the two shaped each other through coevolution. Fruit-bearing plants evolved brightly colored skins to attract primates, which helped disperse their seeds, while primates evolved more sensitive color vision to obtain nutrient-rich food. This process of reciprocal selection is an excellent illustration of Darwin’s theory of natural selection.

New World monkeys, found in Central and South America, diverged from Old World monkeys and apes—the lineage to which humans belong—approximately 35 million years ago.In New World monkeys, no gene duplication occurred on the X chromosome. Instead, differences in color vision arise from multiple versions, or alleles, of the same LWS gene locus. Only females carrying two different versions can possess trichromatic vision, whereas males are always dichromatic.Interestingly, these different allelic versions are highly similar to the separate L- and M-opsin genes produced by duplication in Old World monkeys. This indicates that spectrally distinct versions of the LWS gene already existed before the two groups of monkeys diverged.

The ancestor of Old World monkeys and apes subsequently duplicated the gene, converting versions that had previously been distributed among different individuals into two separate, fixed genes. This produced stable trichromatic vision and represents the evolutionary origin of human trichromacy.

III. Why Is Red–Green Color Blindness More Common in Men?

The genes encoding the L- and M-opsins are both located on the X chromosome. They are tightly linked and share extremely high sequence homology.Men carry only one X chromosome. Therefore, a mutation or abnormal recombination involving the L- or M-opsin gene on that chromosome directly causes red–green color blindness.Women carry two X chromosomes. When one chromosome contains a defective gene, the normal gene on the other chromosome can compensate for the impaired light-sensing function.As a result, the prevalence of red–green color blindness is approximately 7% in men but only about 0.5% in women. This is why red–green color blindness is the most common color-vision deficiency in humans.

IV. The Brain’s “Color Palette”: How Do Three Cone Pigments Produce Millions of Colors?

Humans do not perceive color because the eyes directly “see” color. Instead, the nervous system converts the physical properties of light into neural signals.Light of different wavelengths reflected by objects enters the eye and is detected by the three types of cone cells in the retina. The brain then acts like a painter, producing a particular color according to the relative activation of the three cone types. An activation ratio of 1:0:0 produces red, 1:1:0 produces yellow, and 1:1:1 produces white.

However, L- and M-opsins share as much as 96% sequence identity. Why, then, is one sensitive to red light and the other to green light? S-opsin differs substantially from both; how does it detect blue light? And why do cone cells respond much faster than rod cells?Although the structure of rhodopsin, the visual pigment of rod cells, was determined as early as 2000—and was the first G protein-coupled receptor, or GPCR, structure ever resolved—high-resolution three-dimensional structures of cone visual pigments remained unavailable for many years.

Using cryo-electron microscopy, the present study determined, for the first time, high-resolution active-state structures of the three human cone visual pigments: red-sensitive LWS opsin, green-sensitive MWS opsin, and blue-sensitive SWS opsin. These structures reveal the molecular mechanisms underlying the brain’s “color palette.”

Discovery 1: How Do Three Cone Pigments Produce Millions of Colors?

Why do the three cone visual pigments prefer different colors? The secret lies in the microenvironment surrounding retinal.The protonated Schiff base of retinal must be stabilized by a negatively charged counterion. This mechanism resembles adjusting the tension of a musical instrument’s strings: it determines which wavelength of light the pigment can “play.”

L- and M-opsins: These pigments use a distinctive counterion mechanism. In addition to retaining E129 as the primary counterion, they contain an L/M-specific E102 residue in transmembrane helix 2, which functions as a secondary counterion. In the inactive state, H197 in extracellular loop 2 serves as a chloride-binding site that mediates a spectral red shift.

S-opsin: Four polar serine residues—S87, S183, S289, and S292—surround the Schiff base to form a “serine ring.” This creates an electrostatic environment that shifts the pigment’s sensitivity toward blue light.

Together, these differences provide the structural basis for the spectral specificity of the three cone visual pigments.

Discovery 2: The Secret of Rapid Activation and Deactivation—Why Do Cone Cells Respond Faster Than Rod Cells?

Under physiological conditions, the active state of rhodopsin, known as Meta II, has a lifetime of approximately 100 milliseconds, whereas the active states of cone visual pigments last only 3–10 milliseconds.This ability to reset rapidly is essential for the adaptation of cone cells to photopic, or daylight, vision.Structural comparisons showed that all three cone visual pigments lack the E122–H211–W126 hydrogen-bond triad that stabilizes the active Meta II state of rhodopsin.S-opsin also possesses a distinctive “quick-release” architecture. An S-opsin-specific C84–C296 disulfide bond connects transmembrane helices 2 and 7, and the conserved aspartate residue D2.50 is absent. Together, these features create a larger, continuous water-filled cavity extending from the Schiff base toward the intracellular side.The reduced stability of the active-state environment shortens the lifetime of activated cone visual pigments and accelerates their decay.

Discovery 3: A Structural Map of Color-Blindness Mutations—From a “Spectral Tuning Knob” to a “Folding Switch”

Structural comparisons and quantum mechanics/molecular mechanics, or QM/MM, calculations identified residue 285 as a key “spectral tuning knob.”L-opsin contains threonine at this position, T285, whereas M-opsin contains alanine, A285. When A285 in M-opsin was mutated to threonine, the absorption maximum shifted approximately 12 nanometers toward longer wavelengths. This is comparable to making a fine adjustment to a knob on a color palette and thereby changing the hue of the resulting color.The researchers also mapped known color-blindness mutations onto the three-dimensional structures. They found that these mutations generally cause color-vision deficiencies by disrupting protein folding or membrane trafficking, rather than by directly altering spectral properties.In other words, the problem is more like a blocked tube of paint than a defect in the color of the paint itself. In vitro expression experiments showed that most mutant proteins either failed to express or exhibited markedly reduced absorption, confirming this conclusion.

V. Significance and Outlook

This study systematically addresses the fundamental scientific question of how humans perceive color. It explains the molecular mechanism of spectral tuning from the perspective of the retinal microenvironment, reveals the structural basis of the rapid kinetics of cone visual pigments, and provides an important structural blueprint for understanding the molecular origins of visual disorders such as color blindness.These findings not only provide a precise mutation map for the development of gene therapies for color blindness, but also offer natural templates for designing spectrally tunable optogenetic tools for biomimetic visual brain–computer interfaces. In addition, they open new avenues for developing next-generation artificial-intelligence vision models with biologically realistic color-perception capabilities.

Peng Qi and Cheng Xinyu, doctoral students at the School of Basic Medical Sciences of Nanchang University; Professor Li Jian of Gannan Medical University; and Jiang Haihai of the School of Basic Medical Sciences of Nanchang University are co-first authors of the paper.Professor Zhang Jin of Nanchang University and Professor Patrick Scheerer of Charité – Universitätsmedizin Berlin are the corresponding authors.

Original article:

https://www.science.org/doi/10.1126/science.adz8141