Humans tend to be a self-obsessed lot. We have problems imagining how other creatures see the world. So, when Craig Hawryshyn and Jim Bowmaker announced independently in the early 1980s that fish see ultraviolet (UV) wavelengths,`the observation was met with some surprise' says Hawryshyn. However, 30 years on the tables have turned. Biologists have accepted that some vertebrates see not only UV but polarization too. But why have these creatures invested in visual systems that we lack? Hawryshyn explains that there is a strong selective pressure for UV and polarized light vision. While colour can be affected by environmental and atmospheric conditions, UV polarization remains stable beneath water or in shady conditions. Having established that UV and polarized light vision are standard for some fish, Hawryshyn is keen to understand the complex neural networks that deliver polarized light information from the retina to the central nervous system, starting with the retina ().
Teaming up with Samuel Ramsden, Leslie Anderson and Martina Mussi,Hawryshyn began recording the electric signals generated in the retina and optic nerve as the researchers shone flashes of polarized light into a trout's eye. Changing the angle of the polarized light, Ramsden recorded the retina and optic nerve's electric activity, expecting to see similar activities in both tissues.
But when Ramsden appeared in Hawryshyn's office with both electric traces,Hawryshyn was in for a shock. They were different. The electric activity in the optic nerve showed two peaks of activity, one when the light was vertically polarized at 0°/180° and the second when the flash was horizontally polarized at 90°/270°, but two extra polarization sensitivity peaks had appeared at 45° and 135° in the retina's trace. Hawryshyn already knew that the fish's longwave cones responded to horizontal polarization and the UV cones responded to vertical polarization, but which cells were causing the 45° and 135° sensitivities, and how were they doing it?
Hawryshyn and Ramsden suspected that the extra peaks may be generated by feedback so they reduced the fish's sensitivity to horizontal polarization by exposing the retina to longwave light and checked to see how the fish responded to 45° and 135° polarized light. The 45° and 135°sensitivity peaks shifted. And when they reduced the retina's vertical polarization sensitivity with UV light, the peaks shifted again, but in the opposite direction. There was feedback between the cones, but which cells were generating the feedback?
Calling up his long time collaborator Maarten Kamermans in Amsterdam,Hawryshyn described his unexpected results. Kamermans suggested repeating the electrical recordings while varying the polarization, but this time switching off the negative feedback on cones with cobalt chloride to see if the horizontal cells were the source of the mysterious peaks. Ramsden and Hawryshyn were amazed when the 45° and 135° peaks vanished. The horizontal cells were responsible for the intermediate peaks.
But why is horizontal cell feedback on cones activated so strongly when the polarization is 45°? Hawryshyn explains that at planes of polarization close to 45°, the cones that respond to horizontal and vertical polarization are stimulated almost equally, generating strong negative feedback and electric activity in the horizontal cells.
Having found that some of the earliest cells in the visual neural network are involved in processing polarized light sensory information, Hawryshyn is keen to find out more about the retina's role in polarized light vision.
