Molecular Determinants of Dendritic Patterning in the Mammalian Retina

Abstract

The architecture of dendritic arbors is a critical element in neuronal connectivity and function, influencing partner choice, synaptic wiring and computational properties. Unraveling core principles in dendritic patterning is thus an outstanding objective in neurodevelopment research. Here, I report my efforts towards this goal, utilizing the morphological diversity of retinal ganglion cells (RGCs) in mammalian retina to elucidate the cellular and molecular logic behind dendritic patterning. To begin, I characterized how RGCs belonging to one type, called J-RGC, coordinate dendritic patterning across time and space to produce a stereotyped morphology. I showed that ventral asymmetry of J-RGC dendrites develops in a protracted dorsoventral wave while their lamination within the inner plexiform layer (IPL) develops rapidly in a centrifugal wave. Each ventrally-asymmetric, lamina-restricted arbor bears an asymmetric distribution of inhibitory synapses, the formation of which, surprisingly, lags behind that of excitatory synapses. Thus, distinct mechanisms operate in different dimensions of J-RGC dendritic patterning to generate the appropriate substrate for highly specific modes of synaptogenesis. Next, I present work on the intrinsic control of dendritic patterning in RGCs like J-RGCs. Using transcriptomic profiling with molecular, transgenic and spatial mosaic analyses, I discovered that the transcription factor T-box Brain 1 (Tbr1) is selectively expressed by J-RGCs and three other RGC types. All of them send dendrites to the outer half of the IPL. Using loss- and gain-of-function approaches, I demonstrated that Tbr1 is both necessary and instructive for laminar specification of RGC dendrites. Therefore, Tbr1 is a transcriptional determinant of laminar identity for RGC dendrites. Finally, I describe the roles of cell-surface molecules in Tbr1-driven, type-specific dendritic patterning. I showed that a classical cadherin, Cdh8, and a growth factor receptor, Sorcs3, are localized to J-RGC dendrites, regulated by Tbr1 and capable of affecting lamination by themselves. Thus Tbr1 acts in part via Cdh8 and Sorcs3 to specify the laminar identity of J-RGC dendrites. My work puts forth the concept that a single transcriptional determinant can drive a common dendritic lamination pattern presented by different RGC types. Each determinant can then employ multiple and type-specific mediators to establish key aspects of that lamination program.