Quantitative organization of GABAergic synapses in the molecular layer of the mouse cerebellar cortex

Abstract

In the cerebellar cortex, interneurons of the molecular layer (stellate and basket cells) provide GABAergic input to Purkinje cells, as well as to each other and possibly to other interneurons. GABAergic inhibition in the molecular layer has mainly been investigated at the interneuron to Purkinje cell synapse. In this study, we used complementary subtractive strategies to quantitatively assess the ratio of GABAergic synapses on Purkinje cell dendrites versus those on interneurons. We generated a mouse model in which the GABAA receptor alpha1 subunit (GABAARalpha1) was selectively removed from Purkinje cells using the Cre/loxP system. Deletion of the alpha1 subunit resulted in a complete loss of GABAAR aggregates from Purkinje cells, allowing us to determine the density of GABAAR clusters in interneurons. In a complementary approach, we determined the density of GABA synapses impinging on Purkinje cells using alpha-dystroglycan as a specific marker of inhibitory postsynaptic sites. Combining these inverse approaches, we found that synapses received by interneurons represent approximately 40% of all GABAergic synapses in the molecular layer. Notably, this proportion was stable during postnatal development, indicating synchronized synaptogenesis. Based on the pure quantity of GABAergic synapses onto interneurons, we propose that mutual inhibition must play an important, yet largely neglected, computational role in the cerebellar cortex.

Figure 1. Loss of GABA_ARs from Purkinje cells of PC-Δα1 mice.

(A1,A2) Confocal images showing the distribution of GABA_ARα1 (green) and GABA_ARγ2 (red) in the cerebellar cortex of control (A1) and PC-Δα1 mice (A2). Note that in both conditions most GABA_ARγ2-positive clusters colocalize with the α1 subunit. A few clusters labelled for GABA_ARγ2 but not for GABA_ARα1 likely represent synapses containing α3-GABA_ARs. No surface labelling is visible in Purkinje cells (Pc) of PC-Δα1 mice, whereas MLIs (asterisks) are recognized by extrasynaptic labelling of the α1 subunit (A2). Double-labelled clusters (yellow) identify postsynaptic GABA_AR aggregates on the cell body and the dendrites of MLIs (A2). (B1) Clustered distribution of GABA_ARα1 (green) in control Purkinje cells labelled for Car8 (red). (B2,B3) GABA_AR clusters (B2: GABA_ARα1; B3: GABA_ARγ2) are not visible in Purkinje cells of PC-Δα1 mice. Scale bars: 15 µm.

Figure 2. Synaptic organization in the cerebellum of PC-Δα1 mice.

(A1,A2) Double labelling for GABA_ARγ2 (green) and NL2 (red) in control (A1) and PC-Δα1 mice (A2). NL2 colocalizes extensively with the γ2 subunit and also clusters at postsynaptic sites lacking GABA_ARs in Purkinje cells of PC-Δα1 mice. (B,C) Electron micrographs of the ML of a PC-Δα1 mouse showing that GABA-immunopositive axon terminals (Ax) make both conventional, symmetric synapses (B, arrowheads) with Purkinje cell dendrites (Pc) and heterologous synapses (C) with spines (Sp). (D1,D2) Similar distribution of α3-GABA_ARs in the ML of control (D1) and PC-Δα1 mice (D2). PCL, Purkinje cell layer. Scale bars: A = 15 µm. B,C = 200 nm. D = 20 µm.

Figure 3. Ablation of GABA_ARs is protracted during postnatal development and is asynchronous among different Purkinje cells.

Double labelling for GABA_ARα1 (red) and NL2 (green) in the cerebellar cortex of P7 (A), P16 (B) and P23 (C) PC-Δα1 mice. All P7 Purkinje cells express postsynaptic GABA_AR clusters, colocalized with NL2 (A). Purkinje cells labelled for NL2 but not GABA_ARs (asterisks) are visible at P16 (B). By P23, practically all Purkinje cells are immunonegative for GABA_ARs (C). Scale bar: 20 µm.

Figure 4. Dystroglycan is present at GABAergic synapses in Purkinje cells but not in cerebellar interneurons.

(A1–A3) Double labelling for α-dystroglycan (green) and GABA_ARγ2 (red) in the cerebellar cortex of a WT mouse. Labelling for α-dystroglycan outlines the cell bodies and major dendrites of Purkinje cells and colocalizes precisely with GABA_ARγ2. Labelling for the γ2 subunit is weaker at perisomatic synapses, scarcely visible in these low magnification images. (B1,B2) Triple labelling for α-dystroglycan (green), GABA_ARγ2 (red) and NL2 (blue) in the cerebellar cortex of a PC-Δα1 mouse. Dystroglycan colocalizes with NL2 exclusively at silent synapses that lack GABA_ARs (B2, cyan). NL2 associates with GABA_ARs at interneuron-interneuron synapses, where immunolabelling for α-dystroglycan is not visible (B2, magenta). (C,D) High-magnification images of the ML showing that in WT mice a subset of GABAergic synapses contain α-dystroglycan (yellow clusters), whereas in PC-Δα1 mice α-dystroglycan never colocalizes with GABA_ARγ2-positive clusters. Scale bars: A,B = 30 µm. C,D = 10 µm.

Figure 5. Early synaptic localization of α-dystroglycan and postnatal development of GABAergic synapses in the ML.

(A) Double labelling for α-dystroglycan and GABA_ARγ2 in the cerebellar cortex of a P7 wild-type mouse showing that α-dystroglycan clusters at developing GABAergic synapses onto Purkinje cells (Pc). (B,C) Representative images showing that α-dystroglycan associates with a subset of GABA_ARγ2-positive synapses in the ML of P10 (B) and P21 (C) mice. (D) Synapses were quantified in the iML by counting clusters immunopositive for GABA_ARγ2 (black bars) and α-dystroglycan (grey bars). Percentage values express the ratio of α-dystroglycan-positive clusters over the entire population of GABA synapses labelled for GABA_ARγ2 (n = 3 mice for each developmental stage). Note that the values vary little during postnatal development. Scale bars: A = 15 µm. B,C = 6 µm.