Key questions in neuroscience are: how are complex neural circuits assembled in young animals and how do they process information in adults?
The retina may be the first part of the mammalian brain for which satisfactory answers to these questions will be obtained. The retina is about as complex as any other part of the brain, but it has several features that facilitate analysis: it is accessible, compact, and structurally regular, and we already know a lot about what it does.
Visual information is passed from retinal photoreceptors to interneurons to retinal ganglion cells (RGCs) and then on to the rest of the brain. Each RGC responds to a visual feature –for example motion in a particular direction– based on which of the interneuronal types synapse on it. In one set of studies, we used single cell transcriptomic methods to generate an atlas of all retinal cell types in mice, identifying 46 types of RGCs and some 80 types of interneurons. We extended these studies to generate atlases of 20 vertebrate species, allowing us to trace the evolution of these cell types over hundreds of millions of years. To understand how these circuits form, we marked retinal cell types transgenically in mice, mapped their connections, sought recognition molecules that mediate their connectivity, used genetic methods to manipulate these molecules, and assessed the structural and functional consequences of removing or swapping them.