Search for a command to run...
Synapses are continuously formed and eliminated throughout life in the mammalian brain, and emerging evidence suggests that this structural plasticity underlies experience-dependent changes of brain functions such as learning and long-term memory formation. However, it is generally difficult to understand how the rewiring of synaptic circuitry observed in vivo eventually relates to changes in animal's behavior. This is because afferent/efferent connections and local synaptic circuitries are very complicated in most brain regions, hence it is largely unclear how sensorimotor information is conveyed, integrated, and processed through a brain region that is imaged. The cerebellar cortex provides a particularly useful model to challenge this problem because of its simple and well-defined synaptic circuitry. However, owing to the technical difficulty of chronic in vivo imaging in the cerebellum, it remains unclear how cerebellar neurons dynamically change their structures over a long period of time. Here, we showed that the commonly used method for neocortical in vivo imaging was not ideal for long-term imaging of cerebellar neurons, but simple optimization of the procedure significantly improved the success rate and the maximum time window of chronic imaging. The optimized method can be used in both neonatal and adult mice and allows time-lapse imaging of cerebellar neurons for more than 5 mo in ∼80% of animals. This method allows vital observation of dynamic cellular processes such as developmental refinement of synaptic circuitry as well as long-term changes of neuronal structures in adult cerebellum under longitudinal behavioral manipulations.