Binocular Integration in the Mouse Lateral Geniculate Nuclei

A key task for the visual system is to combine spatially overlapping representations of the environment, viewed by either eye, into a coherent image. In cats and primates, this is accomplished in the cortex 1Casagrande V.A. Boyd J.D. The neural architecture of binocular vision.Eye (Lond.). 1996; 10: 153-160Crossref PubMed Scopus (18) Google Scholar, with retinal outputs maintained as separate monocular maps en route through the lateral geniculate nucleus (LGN). While this arrangement is also believed to apply to rodents 2Coleman J.E. Law K. Bear M.F. Anatomical origins of ocular dominance in mouse primary visual cortex.Neuroscience. 2009; 161: 561-571Crossref PubMed Scopus (54) Google Scholar, 3Sterratt D.C. Lyngholm D. Willshaw D.J. Thompson I.D. Standard anatomical and visual space for the mouse retina: computational reconstruction and transformation of flattened retinae with the Retistruct package.PLoS Comput. Biol. 2013; 9: e1002921Crossref PubMed Scopus (53) Google Scholar, this has not been functionally confirmed. Accordingly, here we used multielectrode recordings to survey eye-specific visual responses across the mouse LGN. Surprisingly, while we find that regions of space visible to both eyes do indeed form part of a monocular representation of the contralateral visual field, we find no evidence for a corresponding ipsilateral representation. Instead, we find many cells that can be driven via either eye. These inputs combine to enhance the detection of weak stimuli, forming a binocular representation of frontal visual space. This extensive thalamic integration marks a fundamental distinction in mechanisms of binocular processing between mice and other mammals.