Projection-dependent heterogeneity of cerebellar granule cell calcium responses

Abstract

Cerebellar granule cells (GCs) relay mossy fiber (MF) inputs to Purkinje cell dendrites via their axons, the parallel fibers (PFs), which are individually located at a given sublayer of the molecular layer (ML). Although a certain degree of heterogeneity among GCs has been recently reported, variability of GC responses to MF inputs has never been associated with their most notable structural variability, location of their projecting PFs in the ML. Here, we utilize an adeno-associated virus (AAV)-mediated labeling technique that enables us to categorize GCs according to the location of their PFs, and compare the Ca2+ responses to MF stimulations between three groups of GCs, consisting of either GCs having PFs at the deep (D-GCs), middle (M-GCs), or superficial (S-GCs) sublayer. Our structural analysis revealed that there was no correlation between position of GC soma in the GC layer and location of its PF in the ML, confirming that our AAV-mediated labeling was important to test the projection-dependent variability of the Ca2+ responses in GCs. We then found that the Ca2+ responses of D-GCs differed from those of M-GCs. Pharmacological experiments implied that the different Ca2+ responses were mainly attributable to varied distributions of GABAA receptors (GABAARs) at the synaptic and extrasynaptic regions of GC dendrites. In addition to GABAAR distributions, amounts of extrasynaptic NMDA receptors appear to be also varied, because Ca2+ responses were different between D-GCs and M-GCs when glutamate spillover was enhanced. Whereas the Ca2+ responses of S-GCs were mostly equivalent to those of D-GCs and M-GCs, the blockade of GABA uptake resulted in larger Ca2+ responses in S-GCs compared with D-GCs and M-GCs, implying existence of mechanisms leading to more excitability in S-GCs with increased GABA release. Thus, this study reveals MF stimulation-mediated non-uniform Ca2+ responses in the cerebellar GCs associated with the location of their PFs in the ML, and raises a possibility that combination of inherent functional variability of GCs and their specific axonal projection contributes to the information processing through the GCs.

Developmental timing-dependent organization of synaptic connections between mossy fibers and granule cells in the cerebellum

“…Another study found that neighboring GC somas in the GCL labeled with calcium indicator gave rise to sparsely distributed PFs 10 . Our previous results obtained by using the AAV-GABRα6 were consistent with these studies, in terms that there is no association in the anterior-posterior axis between positions of GC somas in the GCL and PF locations in the ML 12 . In addition, our present study demonstrated random distributions of GC somas also in the dorsal-ventral or lateral axis, regardless of their bundled PFs.…”

“…A recent study suggested the differences in GC electrophysiological properties according to their PF locations, based on the correlation between PF locations and GC soma locations that was suggested using sparse GC labeling 8 . However, other studies 9 – 11 , including our previous study 12 , demonstrated that there is no apparent correlation between GC soma locations and their PF locations along the anterior-posterior axis. Thus, such association between synaptic organizations from MFs to GCs in the GCL and sublayers of PF projections remained under debate.…”

“…We have previously demonstrated that stereotaxic injection of AAV-GABRα6 at a proper timing during postnatal development is capable of triggering selective molecular expression in a group of GCs, whose PFs are bundled together at approximately one third of the ML 12 , 15 . We confirmed the expression patterns by injecting AAV-GABRα6-GFP into lobule IV/V of the cerebellum at different postnatal times (P7–P13) and observing GFP expression at P21–P23 (Supplementary Fig.…”

“…Evident heterogeneity in overall GCs has been reported regarding molecular expression 34 , morphology 35 and electrophysiological properties 8 , 36 – 38 . We previously succeeded to combine calcium imaging using indicator dyes with AAV-GABRα6-mediated labeling of GCs, and found non-uniform calcium responses in GCs associated with the location of their PFs 12 . Combining the analyses to detect the heterogeneous properties with our labeling technique using AAV-GABRα6 would have the potential to correlate these properties of GCs with their PF locations.…”

“…In the present study, we took advantage of the characteristic features of an adeno-associated viral (AAV) vector with the minimum region of GABAA receptor α6 subunit (GABRα6) promoter that we previously developed 14 . Using stereotaxic injection of this vector, AAV-GABRα6, expressing fluorescent proteins during postnatal development, we were able to distinctly label a specific group of GCs that have PFs located at a certain sublayer of the ML 12 , 15 . Our imaging analyses using this labeling and hypothetical network model demonstrated the overall organization of synaptic connectivity from MFs to GCs based on their PF locations in the ML.…”