Computational structure of the cerebellar molecular layer

Abstract

Theories regarding the computational structure of the cerebellar molecular layer have been more strongly influenced by biological data than is the case for many and perhaps even most other vertebrate brain regions. Yet, or perhaps for this reason, the computational significance of this circuitry remains a very active subject of discussion and debate, with even basic assumptions regarding the physiological organization of this circuit now being questioned. Specifically, while most current models of molecular layer circuitry assume that parallel fiber excitatory synaptic input directly drives Purkinje cells to fire, there is growing experimental evidence that this might not be the case. Recent model-based efforts to better understand the functional role of the parallel fiber system suggest that molecular layer inhibition may play a larger and more complex physiological role than previously realized. In turn, this reanalysis of the role of inhibition in cerebellar cortical circuitry suggests that the cerebellar molecular layer may implement a sensory-related contextual algorithm, rather than either a timing or associative learning function previously assumed to provide the basis for a cerebellar involvement in motor coordination.