Gene therapy gives full colour vision to colour-blind monkeys

People with red-green colour blindness find it difficult to tell red hues from green ones because of a fault in a single gene. Their inheritance robs them of one of the three types of colour-sensitive cone cells that give us colour vision. With modern technology, scientists might be able to insert a working copy of the gene into the eye of a colour-blind person, restoring full colour vision.

 You might think that the brain and eye would need substantial rewiring to make use of the new hardware, but Katherine Mancuso from the University of Washington thinks otherwise. She has used gene therapy to give full colour vision to adult squirrel monkeys that had been red-green colour-blind since birth, opening up a world of formerly invisible reds and oranges, right in front of their eyes.

Her success proved that adding a third cone cell into a retina with just two isn't as much of a technical challenge as it initially seems - the new cells slot in with all the ease of plug-and-play hardware.  The experiments suggest that early mammals could have evolved three-colour vision simply by developing a third type of cone cell, with little in the way of extra genetic control or neural wiring. If anything, the third cone probably exploited the circuitry that was already in place to process the signals from one of its siblings.


The human eye has three types of cone cells - the S cones that are sensitive to short violet-ish wavelengths, M-cones that are sensitive to medium greenish-blue wavelengths, and L-cones that are sensitive to longer reddish wavelengths. Each type of cone cell contains a different light-sensitive pigment - an opsin - and each of these is produced by a single gene. Red-green colour-blindness is what happens when the genes for the M-opsin or the L-opsin are flawed.

In squirrel monkeys, females see a more colourful world than males. While they have the same three opsins that humans do, males lack the gene for L-opsin and can't see red.  Mancuso changed that by loading the human L-opsin gene into a virus and injecting it into the monkeys' retinas. As a result, around 15-30% of each animal's M-cones were also producing L-opsins. It's a trick that other researchers have used to give mice the ability to see red and it clearly worked for monkeys too.

The process took a lot of tweaking and failed attempts, but once Mancuso perfected its, the results were superb.  She tested the monkeys with three collections of dots, two of which were grey and one of which was coloured. Their job was simple - touch the coloured set and earn a rewarding sip of juice.

i-b230674322f518ee3f5220cd5f8f620a-Daltonchooses.jpgFemale monkeys can see red, green, blue and yellow dots, but with only two types of cones, male monkeys can only make out blue and yellow ones. But after their viral injections, they too could pick out all four hues and they could even discern blue-green colours against red-violet backgrounds. Over the course of 5 months, a new palette of reds and oranges were open to them.  

Their new abilities were still there after two years and Mancuso plans to continue watching the monkeys to see if they have any long-term side effects. For the same treatment to be used in humans, such long-term safety data will be essential, as will years of further research.

However, there's no denying that Mancuso's success is exciting and promising, particularly because many aspects of our sight are set during "critical periods" of development during childhood. If our brains don't experience a particular type of signal within these time windows, they don't develop the neural circuits needed to process them. As an example, David Hubel and Torsten Weisel's seminal experiments with kittens showed that stopping them from using one eye during early life prevented them from developing normal binocular vision later on.

But Mancuso's experiments show that this isn't true for colour vision, which means that gene therapy could potentially 'fix' red-green colour-blindness in adults, without being restricted to the very young. While some sight disorders will be impossible to 'solve', Mancuso predicts "that other [technologies], like gene therapy for red-green colour blindness, will provide vision where there was previously blindness."

As a sidenote, it irritates me that one of the press releases (and presumably numerous reports) say that the scientists "cured" colour-blindness in monkeys. They did no such thing. Male squirrel monkeys only have two types of cones naturally. It'd be like saying that we've cured a human of bipedality by grafting two extra legs onto them.

More on eyes:

Reference: Nature doi:10.1038/nature08401


More like this

Evolution mostly involves small, gradual changes, and for good reason - we might expect that large changes to an animal's genetic code, and therefore to its body plan, simply wouldn't work. It would be like shoving an extra cog into a finely-tuned machine and expecting it to fit in - the more…
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…
What has always attracted me to developmental biology is the ability to see the unfolding of pattern—simplicity becomes complexity in a process made up of small steps, comprehensible physical and chemical interactions that build a series of states leading to a mostly robust conclusion. It's a bit…
We are all familiar with the idea that there are strikingly different kinds of eyes in animals: insects have compound eyes with multiple facets, while we vertebrates have simple lens eyes. It seems like a simple evolutionary distinction, with arthropods exhibiting one pattern and vertebrates…

Wow...that is amazing work.

Unfortunatly it means I have to stop making snide little remarks about how gene therapy will never work very well in eukaryotes. Ah well...such is science.

Screw fixing colour-blindness -- can this give me x-ray vision?

More seriously, I presume this only works because squirrel monkeys in general have the neurological wiring to deal with red-green perception, and that doing this with other colour-blind species (such as dogs) would not grant them the same abilities.

Also, it is bizarre to me that there would be a sex difference in squirrel monkeys on such a basic perceptual ability. What is the explanation for this difference (apart from "one produces L-opsin and the other doesn't")?

The way I recall it, we see red when we see light (L, responding across the spectrum) but the S (blue) and M (green) cells aren't reporting in. That's why violet looks reddish; at decreasing wavelengths, the S response falls off at a longer wavelength than the L, just as at the other end M falls off at a shorter wavelength than L.

Me, I have no objection to "cured". All diagnoses are value judgments. I would very much like to be cured of my incapacity to distinguish horizontal, vertical, and circular polarity of light.

By Nathan Myers (not verified) on 16 Sep 2009 #permalink

Tulse: Sex-linked mutations are very common. I doubt monkeys have exactly our X and Y chromosomes, but they must have some analog that is a little more stringent. Many more men than women have color blindness, because of a gene on X but not Y, so it shouldn't be too surprising to find similar variation in monkeys.

By Nathan Myers (not verified) on 16 Sep 2009 #permalink

I wonder whether it is important that the females of the species have trichromatic vision. I do not know much of developmental biology but it seems plausible that the neural circuitry would be similar in the sexes because of evolutionary constraints. (Compare nipples on human males.)

This might make it easier for the males to learn to use the new color sense they get.

Sex-linked mutations are very common.

I realize that, and I'm also familiar with the gender differences in human colour blindness. I guess what I find odd from an evolutionary perspective is why genes for a basic perceptual mechanism would be tied to the chromosomes involved in sex determination -- I would think there would be strong evolutionary pressure against such, and especially against one sex completely lacking some basic colour perception.

Dichromaticism (having just the two cone cells) is beneficial in some areas. There was a paper recently which showed that dichromatic monkeys are better at detecting contrasts, and were better at foraging in shade than trichromats (

Hate to nitpick, but the images are off. I'm red-green colorblind and those images are very obviously different to me. If they were actually indicative of what the color blind world looks like, I and those like me would see the images as the same.

Dichromaticism (having just the two cone cells) is beneficial in some areas.

Cool -- thanks for the reference, Ed.

This is awesome. So can we engineer opsins that respond to the near UV and near infrared, and use gene therapy to put them into humans, thereby creating 5-color vision? Life would be so much cooler with 5 primary colors...

By Curt Fischer (not verified) on 16 Sep 2009 #permalink


There are many different kinds of color blindness.

I, for instance, can hardly tell any color difference at all in those two photos. The grapes and tomatoes look more vibrant in the full-color picture, but the coloring in the monkey looks exactly the same.

Unfortunately, the architecture of the retina does not accommodate more than trichromatic vision.

Cool, can i have UV vision grafted on my eyes next?

The link in "It's a trick that other researchers have used to give mice the ability to see red" is broken.

Fascinating article. Right now I'm just doing my morning survey of what's been happening on the Internet while I was asleep, but will read more throroughly later.

I'm glad I'm not the only one whose immediate thought was expanding the human visible spectrum. What's our closest relative that can see infrared or ultraviolet?

By Treppenwitz (not verified) on 16 Sep 2009 #permalink

Greg, are you completely red-green colourblind, or just partially? Do you have the most common form of red-green colourblindness, or a lesser-known kind? I do not profess to be an expert, but I would think that there are a whole range of colourblindness options possible. Also my dad is colour-insensitive but on the dot tests (where you see a number if you have normal vision) he presents as red-green colourblind. We know he isn't really because we turned up the colour on the TV when they were showing the tests and he presented as having normal vision.

By Katherine (not verified) on 16 Sep 2009 #permalink

We already have things that see in UV and infrared. They're digital cameras.

Mine picks up UV and IR very well. And it's only a little Xacti C40.

I want some mantis shrimp eyes, personally.

It seems as if you could get much of the way to mantis shrimp eyes (i.e., distinguishing various polarizations) with cleverly designed contact lenses.

By Nathan Myers (not verified) on 16 Sep 2009 #permalink

Ed, I wrote Katie an email asking about the ratio of cones to rods, regarding color vision vs. night vision- I've got better night vision than anyone I know but I'm red green colorblind. I've always looked at is as a trade-off. Her replies are below. Note where she corrected you.

I'm just a layman but a story about the gene therapy giving additional color vision to male squirrel monkeys caught my eye. It stated that you've been able to add color perception by adding cone cells. I'm red green color blind and have theorized that my eyes have more rods than most folks, that I've been enjoying better night vision at the cost of some of my color vision. Have you experimented any with light perception with your monkeys? Does the additional cone add some color perception at a cost? Just curious and thanks for your time, John

Hi John, Thanks for the email. What you describe about having better night vision is actually a fairly common observation made by people with color vision deficiency. It has also prompted scientists to test for this in human subjects; however, the consensus seems to be that no one has been able to conclusively demonstrate that people with colorblindness have any advantages under low lighting (scotopic) conditions. (I have attached a paper describing some of this work.)
We have not specifically tested for this in our monkeys, although it would be interesting to know. Because we only added a single gene that specifically changed the response properties of a subset of the cone photoreceptors (not the rods), and we have not done anything to change the overall number of rods vs cones, it does not seem likely that the new color vision was gained at the loss of some other aspect of vision. [By the way, some of the articles have incorrectly reported that we injected ânew cellsâ into the eye; what we actually injected was a virus vector carrying a gene â when it was taken up by the cone cells, a new visual pigment was produced within those cones.] Thanks again, Katie

Katie, So, the cells you modified, are they now dual-color perceptive? Is there a way to tell? Has perception of the other colors changed? For instance, has blue or yellow perception lessened? I'm excited to hear that there's a prospect of full color vision, especially if I don't have to trade off my night vision. Have you played any with the so-called fourth cone that some women possess? Thanks again! John

John, The subset of cones that are transfected by gene therapy co-express two pigments â green and red. Overall, this gives those cells a different spectral sensitivity (different response characteristics to colored light) compared to the original blue cones and green cones in the monkeysâ retinas. Normal color vision requires three different types of cone; whatâs interesting is that co-expressing a red pigment within some of the green cones was sufficient for the monkeyâs brains to recognize them as being a âthird cone type.â Therefore, even though we only injected a single gene, it did result in producing a third type of cone, as far as the visual system was concerned. I hope this is helpful and not too confusing.
The treatment did not have any obvious effects on their ability to see blues and yellows, although we are in the process of testing this more thoroughly.
Drs. Neitz have thought a lot about the tetrachromacy idea, and here is link to a 2006 article that describes a little about it:
You may also be interested in checking-out our website:

The paper she referred to is: "Is Color Vision Deficiency an Advantage under Scotopic Conditions?" by Matthew P. Simunovic, Benedict C. Regan, and J. D. Mollon (she sent a .pdf)


Tulse, I think they COULD get this to work in dogs. As mentioned in the post, a similar sort of procedure has been done in rats, which are normally quite colorblind. The particular method there involved breeding female rats to have a different allele for red on each x chromosome, if I remember correctly, but the method should still hold. All this implies that if you plugged in a fourth type of color signal into the brain, the mind could probably process it. Which is really cool.

John: You can experiment with better color vision just by wearing glasses with different-colored lenses. They don't need to be dramatically different. If you get two pair of sunglasses and swap lenses, you'll have two pair; wearing them on alternate days heightens the experience.

By Nathan Myers (not verified) on 18 Sep 2009 #permalink

You write "In squirrel monkeys, females see a more colourful world than males. While they have the same three opsins that humans do, males lack the gene for L-opsin and can't see red."

This is actually not correct. The L-opsin and M-opsin in squirrel monkeys are actually just alleles of the same gene, and they're located on the X chromosome. Well, actually there are three alleles, and they respond to slightly different wavelengths. The females are trichromates if they are heterozygous (and not all of them are), and the males are always dichromates - but they can have any one of the three alleles.


That should answer Tulse's question.

By HÃ¥vard (not verified) on 20 Sep 2009 #permalink

Thank you, HÃ¥vard, that makes things much clearer. It means the squirrel monkeys' color blindness is much more closely analogous to that of humans. The difference appears to be that we have more than one analog of those opsin genes on each X chromosome, so that one fully functional chromosome suffices to provide color vision, where they need a different one on each X chromosome, if they're even so fortunate as to have two Xes.

It's hard for people like "us" to look up facts like this, so your annotation is very helpful.

By Nathan Myers (not verified) on 20 Sep 2009 #permalink