Reexamining the Role of Dendrites in Neuronal Function

The textbook model of a neuron is one where dendrites merely serve as recipients of excitatory or inhibitory synaptic inputs, integrated at the soma to generate action potentials. However, dendrites generate local spikes that could be critical nonlinear decision points, and recent data suggest that excitatory inputs may only have an effect in the neuronal output if they are activated together in clusters, triggering dendritic spikes. We want to re-examine the role of dendrites in neuronal function in vivo and help usher in, a new “dendrocentric” paradigm of how neurons work, as opposed to the current “somacentric” one. To bring this change, we will assemble a “Dendrite Consortium” of complementary laboratories, expert in genetically-encoded voltage indicators, volumetric two-photon imaging, holographic optogenetics and optochemistry, dendritic patching, EM connectomics, superresolution synaptic mapping and computational models. Wwe will use holographic optogenetics to activate dendritic spines in arbitrary patters (as if one were “playing the piano”), while imaging voltage in 3D with new GEVIs in dendrites from a subtype of L2/3 pyramidal neuron from mouse visual cortex in vivo during sensory stimulation and spontaneous activity, combining this with direct patch recordings from dendrites. These experiments will characterize the functional regimes that generate dendritic spikes in vivo and elucidate their biophysical mechanisms and overall impact on somatic spiking. We will then use connectomics and expansion microscopy to reconstruct complete morphologies and synapse compositions of these dendrites, including imaged ones. These combined functional and structural data will be used to build a rigorous computational model of the neurons, where the functional and computational roles of dendritic spikes will be explored numerically and systematically. Through a collective, integrated effort, we will elucidate the computational logic and functional roles of dendrites and help usher in a new working model of a neuron. We will also generate for the field open-access morphological, functional and computational datasets of complete dendritic trees from one pyramidal neuron subtype, activated by sensory stimulation or intracortical activity. These datasets, including “Rosetta” ones of the same exact neuron, could enable answers to outstanding basic questions on dendritic function in normal and pathological states, as dendrites are central functional elements in all brain circuits and are affected in many neurological and mental disorders.