Response to Comment on "Emergence of Novel Color Vision in Mice Engineered to Express a Human Cone Photopigment"

Makous suggests that the novel color vision documented in knock-in mice neither requires visual system plasticity nor implies the emergence of a new dimension of sensory experience. We explain why we disagree.

¹ Neuroscience Research Institute and Department of Psychology, University of California, Santa Barbara, CA 93106, USA.
² Departments of Neuroscience, Ophthalmology, Molecular Biology and Genetics, and the Howard Hughes Medical Institute, Johns Hopkins Medical School, Baltimore, MD 21205, USA.

^* To whom correspondence should be addressed. E-mail: jacobs{at}psych.ucsb.edu

The experiments reported in (1) did not address the neural mechanisms that allow knock-in mice to make color discriminations, but two possible arrangements were noted: the formation of conventional center/surround ganglion cell receptive fields based on differences in middle-wavelength (M) and long-wavelength (L) cone inputs to the two regions, or a subsequent extraction of color information based on a comparison of variation among ganglion cells in their total M versus L weightings. Makous (2) presumes the second of these possibilities and further claims that this arrangement yields a percept such that homogenous monochromatic test fields appear "blotchy" to the L/M heterozygous mouse, thus appearing identical to what might be produced in a mouse having only a single type of L or M cone as it views a monochromatic field containing spatial variations in luminance. Because the percepts produced by these two different stimulus circumstances are identical, Makous concludes there is no new sensory experience in the L/M heterozygotes.

The presence of a dimension of visual experience can be inferred from tests of visual discrimination. Thus, color vision is formally defined as the capacity to distinguish two spectrally distinct lights regardless of their radiances (3). Color appearance, an attribute related to the experience associated with this capacity, has long been studied in humans. Although there is an understandable temptation to use the human context to infer the color experiences of other species, that step is unjustified and, for the most part, the experience of color in nonhuman species remains mysterious. This is the case for the L/M heterozygous mice, so any speculations about the nature of the percepts supporting their color discriminations are just that—speculations. Even so, it is notable that the blotch pattern claimed as a "probable" experience in the knock-in mouse will be caused not by variations in differential receptor activation across the mosaic, but rather by variations in the M and L cone input weightings to ganglion cells. The latter is unchanging as gaze shifts around the homogenous field, and one might expect the resulting blotchy percept (if, indeed, there is one) to be similar to that of retinally stabilized or entoptic images—initially perceived as shifting in space with eye position and then quickly fading (4). In contrast, the blotches associated with luminance variations in the monochromatic field would be continuously visible and should not shift in perceived location as the eye moves. Thus, the sensory experiences seem very unlikely to be identical in the two cases as is required by this interpretation.

More generally, we think it useful to remember that color vision derives from a comparison of the signals from different cone types and that each cone resides at a distinct spatial location. Thus, all color vision systems must deconvolute two independent attributes of the visual stimulus: (i) local variation in luminance across the retina and (ii) local variation in spectral composition. How one views this deconvolution process depends in part on the retinal area over which it operates. At one extreme, one might imagine an animal whose retina had a complete nasal/temporal separation of M and L cones and, like Makous, be tempted to consider chromatic discrimination as a spatial differencing operation. The primate fovea, with its midget cell system, represents the opposite extreme, with M and L cones finely intermingled. However, in both cases, color vision only emerges from a comparison of signals from different cone types; thus, both cases require that the nervous system correctly decipher the meaning of excitation at different retinal locations.

Relative to the examples just described, the retinas of L/M heterozygous mice, and perhaps the L/M system of the primate peripheral retina, can be thought of as intermediate cases. In the mouse retina, the spatial graininess of the L/M mosaic is about 50 µm, just a little finer than the graininess of mouse retinal ganglion cell receptive fields (5). To succeed at the color vision tests, the L/M heterozygous mice learned to recognize a spatially fine-grained pattern of neural activity reflecting the capture of photons in spectrally distinct cone types, and to extract a chromatic feature from it, which is, we believe, both evidence of the emergence of a new dimension of visual experience and an indication of the plasticity of the mammalian visual system. How these mice accomplish this feat is a separate and intriguing question that may ultimately be resolved by analyzing the response properties of single cells in the visual system of mice with both L and M cones and comparing them to those of mice with only a single type of L or M cone.

References and Notes

The editors suggest the following Related Resources on Science sites:

In Science Magazine

TECHNICAL COMMENTS
Comment on "Emergence of Novel Color Vision in Mice Engineered to Express a Human Cone Photopigment"

Walter Makous (12 October 2007)
Science 318 (5848), 196b. [DOI: 10.1126/science.1146084]
 |  Abstract »  |  Full Text »  |  PDF »

REPORTS
Emergence of Novel Color Vision in Mice Engineered to Express a Human Cone Photopigment

Gerald H. Jacobs, Gary A. Williams, Hugh Cahill, and Jeremy Nathans (23 March 2007)
Science 315 (5819), 1723. [DOI: 10.1126/science.1138838]
 |  Abstract »  |  Full Text »  |  PDF »  |  Supporting Online Material »

ADVERTISEMENT