The dendritic arbors of neurons are incredibly diverse in size and shape. This makes sense, because dendritic morphology is a critical determinant of the numbers and types of inputs that each neuron receives, and the inputs, in turn, lead to the incredible diversity of neuronal function that enables the complexity of brain function. Indeed, the founder of neurobiology, Ramon y Cajal, featured dendritic diversity in his pioneering studies over a century ago, and neurobiologists have delved deeply into the relationship of dendritic form to dendritic function. Much less is known, however, about how different neuronal types acquire their type-specific dendritic morphologies as they develop.
Peng and colleagues address this issue in their new paper, using the mouse retina as a model system. They focus on a set of retinal ganglion cells called ON-OFF direction-selective RGCs (ooDSGCs), which use their inputs to calculate direction of motion and then send the results of their calculation to the rest of the brain. ooDSGCs have unusual bistratified dendritic arbors in which inputs that transmit information about increased and decreased illumination levels are confined to the inner (ON) and outer (OFF) strata, respectively.
Peng et al. use a genetic method to show that a transcriptional regulator called Satb1 generates this bistratification: in its absence, ooDSGCs lose their ON arbor and become unresponsive to light-on stimuli. They then asked how Satb1 works, and found that it acts in part by promoting transcription of a gene that encodes a membrane-associated recognition molecule called Contactin 5. The Contactin, in turn, anchors the ON stratum to a cellular scaffold formed by interneurons that actually provide ON direction-selective input to the ooDSCGs. Together these results provide a model for understanding how a “hard-wired” genetic program can generate functionally important, type-specific features in a dendritic arbor.