Dendritic organization of sensory input to cortical neurons in vivo

Mammalian cortical neurons compute sensory information that arrives through numerous synaptic inputs located on their dendrites. These inputs are essential for various computational functions of neurons. However, it has long been a major challenge to identify sensory stimulation-activated individual synaptic inputs in cortical neurons in vivo . I have implemented and applied the combined method of two-photon calcium imaging and whole-cell patch-clamp recording to study layer 2/3 cortical neurons in mouse visual cortex and identified orientation-specific dendritic input sites for the first time. When the neuron was slightly hyperpolarized just below action potential firing threshold, multiple dendritic calcium hotspots could be observed upon visual stimulation of a given orientation. These hotspots were spatially restricted to small dendritic domains and sensitive to NMDA receptor antagonist, thus representing sensory evoked synaptic input sites. An important finding was that input sites with different orientation preference were intermingled throughout the dendritic tree. Furthermore, all neurons received multiple inputs of different orientation preferences, regardless whether the action potential output had a specific preference or not. Overall, I have established the combined method of in vivo whole-cell recording and two-photon calcium imaging and implemented efficient data analysis algorithms for studying sensory evoked dendritic inputs in mammalian cortical neurons in vivo . Articles involving these major findings have been published or accepted by peer-reviewed journals. As a result, this dissertation is composed of the accumulation of related articles together with brief summary and specification of personal contributions.