Lab mice's gene upgrade gives them a full-color world

Providing a kaleidoscopic upgrade to creatures that are largely color-blind, scientists have endowed mice with a human gene that allows the rodents to see the world in full Technicolor splendor.

The advance, which relied on imaginative tests to confirm that the mice can perceive all the hues that people see, helps resolve a long-standing debate about how color vision arose in human ancestors tens of millions of years ago. That seminal event brought a host of practical advantages, such as the ability to spot ripe fruit, and unveiled a host of new aesthetic pleasures - autumn foliage, magenta sunsets and the blush of a potential mate, among them.

The work also points to the possibility of curing some of the millions of color-blind Americans - and even enhancing the vision of healthy people, allowing them to experience a richer palette than is possible with standard-issue eyes. "It opens up huge doors to understanding how color vision evolved and where it can go," said Brian Verrelli, an evolutionary geneticist who studies color vision at Arizona State University and was not involved in the work, published Friday in the journal Science.

Mice, like most mammals, have limited color perception, equivalent to that of people with red-green color blindness. Their eyes have just two kinds of color detectors, or "cone" cells, each sensitive to a different part of the spectrum. Unable to differentiate between reds and greens, they see the world as a blend of blues and yellows, with gray overlays added by black- and white-registering "rod" cells. By contrast, most people - along with Old World primates and South and Central American female monkeys - have three kinds of cones. That gives birth to the vibrant world of reds and a vast repertoire of related colors.

Scientists studying the evolution of color vision had already identified the DNA mutation that gave rise to the third kind of cone cell. But they have argued over whether that mutation immediately conferred a new breadth of color perception or whether generations had to pass until changes in the brain's neural circuitry could take advantage of the novel inputs. "This experiment says it had an advantage immediately," said Jeremy Nathans, of the Johns Hopkins Medical School in Baltimore, who led the new study with Gerald Jacobs of the University of California at Santa Barbara. "And if you think about it, that's a very good way to build a brain, so that even a small evolutionary tweak can immediately give you an advantage."

Nathans, Jacobs and their colleagues snipped the red-detecting gene from human retinal cells and inserted it into mouse embryos. The resulting mice had the usual two kinds of mouse cone cells and also the human one - everything required for "trichromatic" vision. Tests showed that all the light-detecting cells were working. But were the mice really seeing red? To find out, the team placed each engineered mouse in a box containing three small illuminated screens. Working at first with changing combinations of black or white screens, they trained the mice to touch the screen that was different. Those giving correct answers were rewarded with a drop of soy milk. Then the team tested the rodents' ability to discriminate among various colors, including thousands of trials focusing on the crucial green-red distinction. To make sure the animals were not simply noting a difference in brightness, they randomly made the red or green screens brighter.

Of five mice tested, three proved they could make use of their new cones, choosing correctly about 80 percent of the time. Only 33 percent would be expected by chance. Why two of the mice failed remains unclear. "Maybe they were just too dumb. Maybe they weren't motivated. One just doesn't know why, when an animal fails," Nathans said. That three could immediately use the new information, however, proves that the brain is flexible enough to detail some of its neurons to handling new inputs, he said. It is impossible to know how the enhanced mice are experiencing their spectrally enhanced world, Jacobs said - whether the ordinary seems psychedelic, and whether their quality of life is better. "I don't know what you experience when you see colors," Jacobs said, "and I know even less about what the mouse is experiencing." Indeed, some said, the fact that mice never acquired - or at least never retained - trichromatic vision suggests it may not be of much evolutionary value to them, perhaps because they are nocturnal.

By contrast, biologists theorize that the sudden ability to see red was a windfall for primates. Pink skin may signal good health in a prospective mate, a red face can warn of a rival's anger, and a lack of red can indicate fear - a useful cue in a touch-and-go situation. Jay Neitz, a vision scientist at the Medical College of Wisconsin in Milwaukee, said the work confirms his unpublished research in which he has cured red-green color blindness in monkeys by injecting into their eyes the missing color-receptor gene. If further studies prove the approach is safe, he said, it could be used in people to correct color blindness and perhaps to add gene variants that would allow a finer parsing of the spectrum. "You'd think that the color world of a tetrachromat would be very rich compared to ours," Neitz said. In fact, researchers have identified some people who already seem to have four different color receptors. Some animals have even more, while some mammals that eventually took to the oceans have lost one of their three, apparently having found them unnecessary in their uniformly bluish environs.