Flexible Coupling of Synaptic and Intrinsic Plasticity in a Cerebellar Circuit

It is increasingly recognized that learning and memory depend, not only on changes in synaptic strength, but also on experience-dependent modifications of intrinsic neuronal excitability. However, how these two forms of plasticity interact within a neural circuit to shape behavior remains unclear. Here, we investigated the coordination of the synaptic and intrinsic plasticity in the cerebellar flocculus, which supports oculomotor learning. Using optogenetics and ex-vivo electrophysiology, we examined plasticity mechanisms underlying different learned modifications of the vestibulo-ocular reflex (VOR) in male and female mice. Optogenetic tetantization of granule cells mimicked and occluded VOR-decrease learning, suggesting a role for synaptic and/or intrinsic LTP in this form of learning. Ex-vivo recordings further revealed that both VOR-decrease learning and VOR habituation were associated with synaptic LTP at parallel fiber-Purkinje cell synapses. However, whereas VOR-decrease learning did not alter Purkinje cell intrinsic excitability, VOR habituation induced long-term depression of intrinsic excitability (LTD-IE). Paralleling the measured intrinsic plasticity, parallel fiber-elicited spiking was unchanged after VOR-decrease learning, and decreased after habituation, as it did after VOR-increase learning. The results demonstrate that synaptic and intrinsic plasticity can be flexibly recruited in different combinations to support different modifications of a given behavior by learning. Consequently, learning-related changes in synaptically-driven spiking cannot reliably be predicted from the changes in synaptically-driven currents alone. Rather, intrinsic excitability can play a dominant role in determining whether there are changes in synaptically-driven spiking after learning. Significance statement Research on the neural mechanisms of learning has focused on synaptic plasticity, yet there is also emerging evidence for a role of changes in the intrinsic excitability of neurons in learning. We analyzed how such intrinsic plasticity interacts with synaptic plasticity to shape learning. Results from experiments combining optogenetic, behavioral, and ex vivo electrophysiology approaches demonstrate that synaptic and intrinsic plasticity are recruited in different combinations in the cerebellar flocculus to support different forms of oculomotor learning. Moreover, intrinsic plasticity seems to gate the ability of synaptic modifications to alter the spiking output of the postsynaptic neurons. The flexible recruitment of synaptic and intrinsic plasticity may enhance the adaptive capacity and computational versatility of neural networks.