Scientists often stick genes into organism in order to create something new. Remote-controlled flies, for example, or photographic E. coli. But by creating new kinds of life, scientists can also learn about the history of life. Give a mouse human vision, for example, and you may learn something important about how our own eyes evolved.
As mammals go, we have unusual eyes. Most mammals produce two kinds of pigments for catching light. One is sensitive to short wavelength light (at the blue end of the spectrum). The other is sensitive to a longer wavelength, in the green or red part of the spectrum. Each pigment is encoded in a single gene, and any given photoreceptor in the retina uses only one. Each photoreceptor is most likely to fire a signal to the brain in response to a single wavelength of light. But it also has a smaller probability of firing off in response to photons of close wavelengths. By comparing all the signals coming from all the photoreceptors, a mammal can perceive a wide range of colors, rather than just seeing a world made up of green and blue pixels.
We humans--along with our closest primate relatives, the apes and the monkeys of Africa and Asia--can make three pigments, not two. We carry a gene for a blue pigment (S for short, at 420 nanometer wavelength), a green pigment (M for medium, 530 nm), and a red pigment (L for long, 560 nm). This so-called trichromatic vision makes us a particularly good at making out different colors down at the yellow-to-red end of the spectrum. On the other hand, we don't see up in the ultraviolet range, unlike many vertebrates.
The most plausible explanation for this difference is that at some point in our history, our ancestors parted ways with other mammals, and acquired a third pigment gene. And many pieces of evidence suggest that what happened was this: the red and green genes are the product of an accidental duplication of an ancestral gene. For one thing, they are very similar to one another, differing by just three amino acids. For another, they sit close to one another on the X chromosome. They even appear to be controlled by the same "on-off switch," which turns on red pigments in some photoreceptors and green in others.
(Incidentally, this is why color blindness tends to appear in men. Men have only one X chromosome, while women have two. A mutation to a pigment gene on the X chromosome can knock out a pigment entirely from a man's eyes, while a woman may still be able to make good pigments off of the genes on her other X.)
The third piece of evidence for a red-green duplication is particularly intriguing. The closet living relatives to the lineage of humans, apes, and Old World Monkeys are the New World Monkeys. And they have a very odd sort of vision. They carry a blue (S) pigment like us, but they only have a single pigment gene on the X chromosome. Males can only have two pigments in their eyes. But in many species, there are different versions of the pigment gene on the X chromosome, and each gene encodes a pigment sensitive to a different wavelength. Female monkeys, with their two X chromosomes, may inherit two copies of the same X gene, and thus have two pigments as well. Or they may carry a different version on each X chromosome. In each photoreceptor, one of the X chromosomes is shut down, so that only one kind of pigment gets made. They thus end up trichromatic. Making the whole situation even stranger, many variants of the X chromosome pigment may float around a population of New World Monkeys, which means that from female to female, their vision may differ drastically. (See this pdf for a review.)
Evolutionary biologists have proposed a hypothesis for this difference by looking at the ecology of primates and their pigments. In a 2003 paper, one team of researchers proposed that ancestral monkeys ate figs and other fruit. Monkeys with three pigments might sometimes be able to pick out ripe fruit, while monkeys with two pigments might do a better job of picking out hidden fruits because they wouldn't be distracted by extra colors. The variation from monkey to monkey might benefit an entire monkey troupe. But about 30 million years ago, Old World Monkeys shifted to eating leaves, the scientists argue. They needed to be very sensitive to reddish hues in order to tell tender young leaves from inedible older ones.
The evidence from the genes of primates suggests one way in which Old World monkeys and apes got their third pigment gene. Like New World monkeys, a population of ancestral primates had a few different versions of pigment genes on the X chromosome. A mutation in a female caused the gene on one X chromosome to be pasted alongside the gene on the other X chromosome. Her descendants now inherited two genes on the same chromosome. Subsequent mutations changed the sensitivity of each pigment, to create the blue-green-red vision we now have.
While this hypothesis is consistent with the evidence, it raises many questions (like any other scientific hypothesis). Color vision, after all, is more than a matter of pigments. Once a photoreceptor snags a photon of a certain wavelength, an animal needs the right sort of circuitry in its nervous system to make sense of the sensation--to turn the signal into the perception of a color. Now imagine the first time an ancient female primate inherited two distinct X-chromosome pigments from her parents. How could her brain suddenly make sense of three separate channels of information, when her ancestors were accustomed to just two?
Over the past few years, Jeremy Nathans, a neuroscientist at Johns Hopkins, and his colleagues have working towards some answers to that question. They have figured out how to insert human pigment genes into mice. Mice, like most mammals, have two pigments. Mice have a medium-wavelength pigment sensitive to green (510 nm) and a very short wavelength pigment that's sensitive in the ultraviolet. Red is pretty much meaningless to them, in other words. Nathans's team replaced the green pigment gene in mice with the human red gene (560 nm). They then bred the mice with ordinary ones. Their offspring inherited these genes according to Mendel's rules. Males inherited either the mouse-green or human-red gene on their one X chromosome. Females could either have two copies of either gene, or they could have one of each. In other words, they resembled New World Monkeys, in which some females carry three different pigment genes.
In 2003, Nathans's group reported that mice could not only inherit all three genes, but they could also produce all three pigment in their retinas. They even found that the retinas became sensitive to UV, green, and red light. The next challenge was to move into the brain--to see if the mice could actually make sense of all three colors.
Today in the journal Science the researchers report that the answer is yes. They carried out behavioral tests on the mice, showing them three colored panels. In some cases one of the panels was a different color, and in others they were the same. They gave the mice a reward of soy milk if they could correctly recognize when the colors were different. Not surprisingly, ordinary mice only got the right answer a third of the time, even after thousands of trials. In other words, they could only guess. But female mice with all three pigments got the right answer four times out of five.
What's particularly intriguing about these results is that we--and other primates--have a special system of neurons called midget cells dedicated to distinguishing between red and green on the way to the brain. Mice and other mammals lack midget cells. And yet it's obvious that they didn't need a special set of neurons to take in this extra dimension of color. Their brains simply organized themselves in such a way that they could perceive it.
If you didn't know this, you might wonder if trichromatic vision is too complex to have evolved--that adding an extra color was impossible because you need a complicated system made of many parts in order to see if at all. But that's not true. When a single mutation occurs, altering a single gene, its effects can ripple out across an animal's body. Simply moving a pigment gene from one chromosome to another appears to have triggered a drastic rewiring of the brain, taking advantage of the brain's ability to adjust itself. Natural selection could have fine-tuned the connections from the eye to brain in later generations, producing midget cells and other adaptations, in order to improve color perception. What seems complex now may not be so complex in the first place.
- Log in to post comments
Woot! Great piece there. This story is especially fun because you can set up a pretend scenario in your head about what the mice are trying to communicate to each other about all the different colors they're seeing. And what if a couple mice escaped and begun a brand new race of mice that suddenly can better navigate the human world? Mwa ha ha.
This made my day. I now have a new favorite example of gene duplication!
As I understand it, other vertebrates use four pigment genes.
What would happen if the missing two pigment genes were grafted into a mouse? Would we get a mouse that could see in shades of four colours?
Nice post - Sean Carroll also discusses dichromatic and trichromatic vision in mammals (incl. the split between Old and New World primates) in his latest book "The Making of the Fittest".
Very interesting piece. I've always had a fascination with tetrachromacy in females and how their brains deal with the extra forth color input. Have you done any research on this phenomena?
re: tetrochromacy...
i once knew a woman who was extraordinarily sensitive to purple. she wore it every day but always denied that it was 'purple': it was 'lavender' or 'fuscia' or 'magenta' silly!
i got to wondering later whether she had a mutant(?) allele of the red allele on one X chromosome.
the question this begs is: could we someday see a human seeing in the infra-red or ultraviolet colour spectrums like some animals?
I believe you can find evidence for the fact that under special circumstances it is possible for humans to sense near UV.
I have had one such experience. At least, I have no other way to interpret the sensation of a extra band in a rainbow...
I wonder if this can be used to cure color-blindness in humans.
Also, maybe sometime soon we can have pets that see in color as well. This is great news.
The story's missing just a little something. Is there any speculation about how the mice's brains could possibly just adapt themselves to see a brand new color? Like could we splice lizard genes into humans and have them grow up able to see ultraviolet? Makes no sense. And that's the crucial piece of the story for explaining why color vision isn't irreducibly complex. We can see it isn't, somehow, but it would be nice to know why. To make sure it wasn't just an experimental error, for one thing.
Carl, you write marvelous material.
Being color blind, I can only wonder about what fellow humans see. The chart on Carl's previous post had to be interpreted for me, for example. I remember quipping to my interpreter, "It should have been in shades of gray". When it comes to color, I feel like some of those New World monkeys.
As Sean Carrol's "The Making of the Fittest" clearly demonstrates, trichromatic vision has evolved -- and been lost -- many times in evolutionary history.
Fish and birds have trichomatic vision -- some birds can see ultraviolet light, as well -- but the ability to see three distinct wavelengths of light was lost in mammals, who are mostly nocturnal.
Because nocturnal animals need to see in low light conditions, color vision came to be less important for survival, and natural selection did not kill off mammals who lost three-color vision through mutation of the opsin gene for seeing light in the red part of the spectrum.
Color vision evolved again in the apes and old world monkeys. It also evolved separately and later in one monkey species found in the Western Hemisphere -- the Howler.
This raises an interesting question about the neural pathways for processing trichomatic vision. Were those pathways preserved from an earlier period of evolution before mammals evolved, or did mutation likewise render them inoperative, like the opsin gene for seeing red, requiring them to evolve a second time?
Do tetrachromatic females need computer monitors with four color guns? Or is three enough to produce a big enough color space?
Ultra sensitive to purple? The human eye generally isn't that sensitive to blue...
Infrared or ultraviolet human vision might be made by inserting appropriate genes from animals that have those abilities. I can hear the ethics debates already. And people will probably want to start wearing clothing that isn't transparent in those frequencies.
We have infrared goggles on the market. So, you can get an idea how this would look. I have an older web cam with a CCD that happens to be infrared sensitive. There's a filter to remove IR, but it could be removed...
Interesting post, it made me think that likely there are many polymorphisms in the human pigment genes as well. Perhaps, that lends some credence to the philosophical idea that my red is different from your red.
"Simply moving a pigment gene from one chromosome to another appears to have triggered a drastic rewiring of the brain, taking advantage of the brain's ability to adjust itself."
I assume by this you meant that the photoreceptors created by the new gene sent new info to the brain's neural network, which rewired itself to understand the info - not that the gene directly caused a rewiring.
Dawkins discusses the color vision evolution issue too in "Ancestor's Tale".
Neural plasticity is awesome.
If we developed the biological adroitness to grow eyes and transplant them into humans successfully, what might the experience of an adult human receiving eyes with nonstandard pigmentation be?
Regarding human UV vision, I seem to recall the main obstacle is the cornea. I've heard that some people who have had their corneas removed (presumably due to disease) can see UV.
For both IR and UV, there would be focusing issues doe to chromatic aberration. That is, lenses bend different colors a little differently; the depth of our eyes represents a (hopefully) happy medium among the colors we normally see. This is part of why nocturnal animals tend to have big eyes. Perhaps this wouldn't apply to the IR-sensing "pits" found in certain snakes, but I have no clue how those form an image anyway.
Caledonian: In trying to transplant eyes into an adult human, I'd think the more immediate problem would be getting the severed optic nerves to join together properly!
I'm not sure if "plasticity" is quite the right word here, but yeah, it's really impressive that the developing mouse brain "automagically" organizes itself to handle the extra channel. I find myself wondering just how many times in the murine/mammalian ancestry, have color-receptor genes been gained and lost...?
It's worth noting that there are other traits relevant to color vision. For example, while my dad and uncles had classical R/G color blindness, my Mom's side of the family runs to an unusual sensitivity to color, which affects both men and women on that side. We call it "artist's eyes" -- the whole (side of the) family (from Grandpa to my sisters and me) have dabbled in art to varying degrees.
I found a story about this on another board, and among the related links was one about tetrachromaticity. It is probably what David Harmon is talking about just a few posts above. Check it out at the Post-Gazette:
http://www.post-gazette.com/pg/06256/721190-114.stm
It's surprising to me that there is not more information in articles covering tetrachromicity and the insertion of an additional gene for color perception in mice of the actual physiology of the nerves and the electro chemical processes involved in transmitting information from the eye to the brain and processing it there.
What I have read has not been clear about the form of the signal propagated on the optic nerve to the brain, the structure of that nerve (simple or compound bundle), how signals from different sensors are transmitted (combined into a compound signal vs. transmitted on separate nerves as distinct signals?) etc.
Nor has there been any indication - in the articles on these topics I have had a chance to read - about how the brain handles incoming signals - demodulates, parses, processes, integrates the results of processing, and produces results that the brain can integrate into the rest of its activities from them.
Perhaps Carl or a commenter can point me to better information on that.
Hey, I just got this story again via Reddit.
Can somebody tell me why Scienceblog.com intentionally eliminates the Reddit user frame? What does this site's administrators have against people coming here from Reddit?
I know a man who can see into the ultraviolet. He and his twin brother were born prematurely and kept in an incubator for the first few days of their lives. Too much oxygen resulted in damage to the cornea, and he suffered the worst from it. Meaning cataract surgery, back before artificial corneas included a yellow pigment to screen out UV light.
It also meant his retina was damaged by the UV, and so he has extremely poor vision. Which is a shame, because he used to be an artist.
I have extraordinary night vision myself. At least my brothers thought so when I was a kid. Didn't matter how they moved living room furnitute around, I had no trouble negotiating safely in the dark. And this was back before light pollution got as bad as it did. On some full moon nights I can read 12 point type in certain fonts.
I can see male humans picking up the tetrachromatic trait through gene shuffling. Then the ladies evolve quatrochromacity.
Though, now that I think of it, I do recall that some trichomatic people can't distinguish the shades others can. Not because of a difference in sensors, but in a difference in how the information is procesed.
I just thought of something. There are a small number of men, born male, functionally male (with one exception obviously), who are XX. Thus it is possible a small number of men have the tetrachromacy trait.
Wow, another gleaming post to this layman's eyes, and some classy comments too.
I seem to remember Sacks says something about dichromatic-style sight being of foraging benefit in 'On the Island of the Colour-Blind'. I wonder how many 'allele-overlap' skills there are out there with communal benefits for us hairless apes?
I have read that people who could see into the UV (usually elderly men) assisted in the navigation of WWII pathfinder bombers, where the trait was useful. No citation, sorry.
Of course. I'm sorry, but this is totally obvious. Why did you even have to do an experiment to look into this? Just go check out Mouse Paint! It's right there! Sheesh. What a waste of funding.
Verrelli and Tishkoff find evidence for diversifying selection at the "red" opsin locus. One possible explanation for this is an advantage to women who carry pigments with sufficiently different spectra to be effectively tetrachromatic.