Abstract

SummaryColor vision, originating with opponent processing of spectrally distinct photoreceptor signals, plays important roles in animal behavior.1, 2, 3, 4 Surprisingly, however, comparatively little is understood about color processing in the brain, including in widely used laboratory mammals such as mice. The retinal gradient in S- and M-cone opsin (co-)expression has traditionally been considered an impediment to mouse color vision.5, 6, 7, 8 However, recent data indicate that mice exhibit robust chromatic discrimination within the central-upper visual field.9 Retinal color opponency has been reported to emerge from superimposing inhibitory surround receptive fields on the cone opsin expression gradient, and by introducing opponent rod signals in retinal regions with sparse M-cone opsin expression.10, 11, 12, 13 The relative importance of these proposed mechanisms in determining the properties of neurons at higher visual processing stages remains unknown. We address these questions using multielectrode recordings from the lateral geniculate nucleus (LGN) in mice with altered M-cone spectral sensitivity (Opn1mwR) and multispectral stimuli that allow selective modulation of signaling by individual opsin classes. Remarkably, we find many (∼25%) LGN cells are color opponent, that such cells are localized to a distinct medial LGN zone and that their properties cannot simply be explained by the proposed retinal opponent mechanisms. Opponent responses in LGN can be driven solely by cones, independent of cone-opsin expression gradients and rod input, with many cells exhibiting spatially congruent antagonistic receptive fields. Our data therefore suggest previously unidentified mechanisms may support extensive and sophisticated color processing in the mouse LGN.

Highlights

  • We set out to unambiguously determine the extent to which cone-opponent mechanisms account for color vision in mice by performing multielectrode recordings spanning the visual thalamus in mice (n = 27) in which the native M-cone opsin was replaced with the human L-cone opsin[20] (Opn1mwR; Figure S1A)

  • We examined the prevalence of cells exhibiting nonopponent and opponent responses, revealing a pronounced functional subdivision where opponent neurons strongly localized to medial dLGN and IGL and non-opponent neurons dominated the surrounding lateral geniculate nucleus (LGN) regions (Figure 1J)

  • We modeled the likelihood that the observed LGN color opponency could be an emergent property of asymmetries in cone contributions to center and surround receptive fields (RFs) components (‘‘random wiring’’; Figures S3F–S3I; STAR Methods)

Read more

Summary

Introduction

We employed previously validated[21] multispectral stimuli to selectively modulate effective light intensity for L- and/or Scone opsin, against a background that recreated a wild-type mouse’s experience of natural daylight (Figures S1B and S1C). Using these stimuli, cone-opponent responses are readily identifiable via opposite responses to selective modulation of either cone opsin (‘‘LOnly’’ and ‘‘SOnly’’ stimuli) and via stronger responses to antiphasic (‘‘L-S,’’ producing a substantial change in color) rather than in synchronous changes in excitation of both cone opsins (‘‘L+S,’’ which changes ‘‘illuminance’’ without altering color)

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call