Senior Author

Colour vision is the result of a remarkably complicated process that requires a host of photoreceptor cells to be arrayed so that they can catch, and then process, light of different wavelengths. How these receptors are arranged, and what role each type plays in processing the small components of an image into a colourful whole, is a mystery.

On page 174 of this issue, a team of developmental biologists explores how the random — or stochastic — distribution of photoreceptors in fruitflies' eyes can give some insight into human sight. They found that, although photoreceptor distribution is random and complex, the expression of one gene plays a crucial role in how they are arranged. Nature spoke to team leader Claude Desplan, a biologist at New York University, to find out how flies — and humans — reconstruct colour.

Why look at flies to understand human vision?

Most mammal animal models don't have colour vision because they are nocturnal. It is not absolutely obvious that insects are good models for human vision, but insects have a pretty well-developed colour vision; they can compare the vector of light polarization for navigation as well as different-colour wavelengths. There are some differences, though. Flies have three eye units, each of which processes different kinds of visual information.

How hard was the experimental work?

We know the system extremely well. The eye is a system that can be manipulated. You let the body of the fly be wild type and you make the genetic mutation almost at the single-cell level. Now that we know that one gene — one transcription factor — plays a role in the random distribution of receptors we can change its timing, level or pattern of expression.

What is the role of randomness in colour vision?

In flies, the colour receptors are stochastically distributed. This is true in humans, too, but with a very fixed ratio of one kind of colour receptor to another. Interestingly, in zebrafish, receptors are amazingly well organized in an array — which seems to make more sense. I don't fully understand why zebrafish receptors are more organized than the more random fly or human retinas.