Molecular and functional heterogeneity of cerebellar granule cell terminals expands temporal coding in molecular layer interneurons

Abstract

In the cerebellum, molecular layer interneurons (MLIs) play an essential role in motor behavior by exerting precise temporal control of Purkinje cells, the sole output of the cerebellar cortex. The recruitment of MLIs is tightly controlled by the release of glutamate from granule cells (GCs) during high-frequency activities. Here we study how single MLIs are recruited by their distinct unitary GC inputs during burst of GC stimulations. Stimulation of individual GC-MLI synapses revealed four classes of connections segregated by their profile of short-term plasticity. Each class of connection differentially drives MLI recruitment. Molecular and ultrastructuralanalyses revealed that GC-MLI synaptic diversity is underlain by heterogeneous expression of synapsin II at individual GC terminals. In synapsin II knock-out mice, the number of classes is reduced to profiles associated with slow MLI recruitment.Our study reveals that molecular diversity across GC terminals enables diversity in temporal coding by MLIs and thereby influences the processing of sensory information by cerebellar networks.

Origins, Development, and Compartmentation of the Granule Cells of the Cerebellum

“…Thus, a superficial GL patch will project to the more distal PC dendritic arbors while parallel fibers from a deeper patch will terminate on dendritic branches more proximal to the PC somata. It is not clear whether proximal parallel fiber inputs are more influential on PC firing (Gundappa-Sulur et al, 1999 ; Dorgans et al, 2018 ). Either way, we believe that molecular differences imply functional differences, and focusing mossy fiber inputs to specific PC dendritic subdomains provides an additional subtlety that may serve to enhance the computational capacity of the mossy fiber pathway.…”

Models of Purkinje cell dendritic tree selection during early cerebellar development

“…Such inhomogeneity in granule cells can be represented by, for example, 2 different types of molecular diffusions in the model to facilitate the zoning. Also, we can set different types of granule cell objects and assign different preferences in making synapses with distal or proximal regions of Purkinje cells’ dendritic trees as observed [ 53 , 61 ]. It is technically also possible to include granule cell proliferation, but in the absence of tissue expansion in the model this did not seem useful.…”

Granule cells affect dendritic tree selection in purkinje cells during cerebellar development

“…Such inhomogeneity in granule cells can be represented by, for example, 2 different types of molecular diffusions in the model to facilitate the zoning. Also, we can set different types of granule cell objects and assign different preferences in making synapses with distal or proximal regions of Purkinje cells ‘dendritic trees as observed (28, 29). It is technically also possible to include granule cell proliferation, but in the absence of tissue expansion in the model this did not seem useful.…”