The light-sensitive cells in our eye - the photoreceptors - are known as rods and cones. Cone cells give us our colour-vision when the light is bright enough, and rods take over to give us a monochrome view of the world when the light levels are very dim. These two cell types have long been thought to mediate all of the light responses of the vertebrate eye. But according to a report in the 2 July 1998 issue of Nature, there are light-sensitive pigments in other nerve cells in the fish retina, suggesting that photoreception is not just restricted to rods and cones.

We can identify a photoreceptor cell by the photopigment it contains. Lots of cells in the eye contain pigments, such as melanin, to absorb light at the back of the eye, and shielding pigments to separate facets of the insect eye. But photopigments are a bit different. They are made up of a protein molecule, opsin, which usually sits in the cell membrane, and a so-called ?chromophore?, the active region, made from vitamin A. When light hits the chromophore, it changes shape and is released from the opsin. Opsin molecules activated in this way in turn set off nerve signals.

There are cells with photopigments in the pineal gland of the brain, which seem to regulate daily rhythms and hormone release. But rods and cones were thought to be the only photoreceptors in the vertebrate eye.

But Bobby Soni of Imperial College in London, UK, and his colleagues, have now found a functional photopigment in the salmon retina, in nerves known as horizontal cells and amacrine cells.

Genetic codes for new types of opsin families have been found, leading naturally to the question of whether they can form functioning photopigments. The research team looked at one of the proteins known as vertebrate ancient (VA) opsin from the salmon. They found that when it was regenerated with vitamin A, it absorbed light in just the same way as the rod pigment, rhodopsin.

The researchers found none of the pigment in the fish?s rod or cone cells, but found that it was made in amacrine and horizontal cells. These cells regulate and integrate the signals from rods and cones, and were not thought to act as photoreceptors themselves. They also found the pigment in the pineal gland and in another brain region known as the habenular.

So what could these ?new? photoreceptors do? One of the jobs of horizontal and amacrine cells is to control how wide an area of the visual field other cells along the image processing pathway pay attention to. If it is not so bright, it is best to lose a little resolution and widen the pixel size to keep the cells responding well - like opening the aperture of a camera lens when the light is dim. The fact that the pigment is also found in the brain suggests that another possible function is in helping to control circadian rhythms.