De novo synaptogenesis induced by GABA in the developing mouse cortex

Abstract

Dendrites of cortical pyramidal neurons contain intermingled excitatory and inhibitory synapses. We studied the local mechanisms that regulate the formation and distribution of synapses. We found that local γ-aminobutyric acid (GABA) release on dendrites of mouse cortical layer 2/3 pyramidal neurons could induce gephyrin puncta and dendritic spine formation via GABA type A receptor activation and voltage-gated calcium channels during early postnatal development. Furthermore, the newly formed inhibitory and excitatory synaptic structures rapidly gained functions. Bidirectional manipulation of GABA release from somatostatin-positive interneurons increased and decreased the number of gephyrin puncta and dendritic spines, respectively. These results highlight a noncanonical function of GABA as a local synaptogenic element shaping the early establishment of neuronal circuitry in mouse cortex.

Fig. 1. GABA induces de novo formation of gephyrin puncta and dendritic spines during early development

(A) Images of newly formed gephyrin puncta (green arrowheads) and dendritic spines (pink arrowheads) in vitro and in vivo. (B and C) Success rate of de novo gephyrin and spine formation by GABA and glutamate HFU in vitro [(B); GABA, n = 29 trials, 16 cells; glutamate, n = 24 trials, 12 cells; GABA in younger, n = 17 trials, 7 cells; GABA in old, n = 36 trials, 16 cells] and in vivo [(C); GABA, n = 61 trials, 55 cells, 5 mice; mock stimulation, n = 52 trials, 45 cells, 5 mice; spontaneous, n = 71 trials, 52 cells, 6 mice]. (D) Summary plots of distance-dependent de novo gephyrin and spine formation. (E) Time course for size changes of new gephyrin puncta (n = 18) and spines (n = 24). (F) Time-lapse images of GABA HFU–induced gephyrin puncta and spines. (G) Stability of newly formed gephyrin puncta (12 of 15) and dendritic spines (15 of 24). **P < 0.01; n.s., not significant. Error bars in (E) denote SEM.

Fig. 2. Molecular mechanisms of GABA-induced de novo gephyrin puncta and dendritic spine formation

(A) Current-voltage curves measured by perforated patch-clamp recordings. (B) Summary graph of E_Cl at P6 to P7 (−45.9 ± 2.7 mV, n = 5), P11 to P14 (−57.2 ± 2.8 mV, n = 7), and P18 to P24 (−66.3 ± 2.1 mV, n = 6). (C) Image of a dendrite coexpressing GCaMP6s, tdTomato, and Teal-gephyrin. Dashed line indicates line-scan path. Green and red fluorescences were measured by line scanning before and after GABA uncaging. (D) Averaged traces and summary ΔG/R [(G/R)_peak/(G/R)_baseline, where G (green) is GCaMP6s fluorescence and R (red) is tdTomato fluorescence] in young (EP6 to EP8) and old (EP16 to EP18) neurons (with GCaMP6s, n = 13 cells; without GCaMP6s, n = 8; GABAzine, n = 9; mock stimulation, n = 9; old neurons, n = 8). (E and F) Success rate and stability of de novo gephyrin puncta and spine formation (no drug, n = 29 trials, 16 cells; 10 μM GABAzine, n = 30 trials, 17 cells; 3 μM CGP55845, n = 21 trials, 12 cells; 10 μM nifedipine, n = 30 trials, 14 cells; 10 μM mibefradil, n = 26 trials, 13 cells; 10 μM bumetanide, n = 21 trials, 9 cells; GABA LFU, n = 18 trials, 8 cells, 1 Hz; 0 mM Ca²⁺, n = 26 trials, 14 cells). *P < 0.05, **P < 0.01; error bars represent SEM.

Fig. 3. Rapid accumulation of functional receptors at newly formed gephyrin puncta and dendritic spines

(A) Time-lapse images of a dendrite after GABA HFU. Green and red arrowheads indicate target and control spots, respectively. (B) uIPSC traces (average of 8 to 10 trials at 0.1 Hz, +10 mV) measured by whole-cell voltage-clamp recordings. (C and D) Time courses of the changes in gephyrin fluorescence (C) and uIPSC amplitudes (D) at targets (n = 10 regions) and neighbors (n = 10, 10 cells). (E) uIPSCs from a target at different time points. Time course of gephyrin fluorescence and uIPSC amplitude changes at targets. (F) Scatterplot between gephyrin expression levels and uIPSC amplitudes (n = 10 cells; up to 40 min). (G) Time-lapse images of a dendrite after GABA HFU (blue cross). Red arrowheads indicate new spines. uEPSCs were evoked by glutamate uncaging (red cross, 8 to 10 trials at 0.1 Hz, −65 mV) from a new spine measured by whole-cell voltage-clamp recordings. (H) AMPAR-mediated uEPSCs from newly formed dendritic spines (n = 12 spines, 5 cells). *P < 0.05, **P < 0.01; error bars represent SEM.

Fig. 4. GABA release from SOM interneurons is sufficient and necessary for inhibitory and excitatory synaptogenesis during early development

(A) Experimental schematic and images of newly formed gephyrin cluster (green arrowhead) and spines (red arrowheads). (B) Summary data of photoactivation-induced de novo synaptogenesis (young, n = 23 dendrites, 13 cells; GABAzine, n = 22 dendrites, 13 cells; old at EP16 to EP18, n = 19 dendrites, 11 cells; spontaneous, n = 13 dendrites, 11 cells). (C) Axons of SOM and de novo gephyrin punctum (green arrowhead) and dendritic spine (red arrowhead) imaged at 1045 nm. (D) Proximity between an axon of SOM and a new gephyrin cluster (<2 μm: 28 of 33 new puncta, 17 cells) or spine (<2 μm: 35 of 39 new spines). (E) Experimental schematic and images of dendrites in noninfected and TeTxLC-infected slices at EP12. (F and G) Quantitative analysis of gephyrin puncta density at EP11 and EP12 (control, n = 13 cells; TeTxLC, n = 15 cells) and spine density at EP7 (control, n = 8 cells; TeTxLC, n = 10 cells). (H and I) Traces of mIPSCs (H) and mEPSCs (I) measured by whole-cell patch-clamp recordings. Quantitative analyses of mIPSCs (control, n = 11 cells; TeTxLC, n = 11 cells) and mEPSCs (control, n = 7 cells; TeTxLC, n = 11 cells) are shown. *P < 0.05, **P < 0.01; error bars represent SEM.