Maternal loss of Ube3a produces an excitatory/inhibitory imbalance through neuron type-specific synaptic defects

Abstract

Angelman syndrome (AS) is a neurodevelopmental disorder caused by loss of the maternally inherited allele of UBE3A. AS model mice, which carry a maternal Ube3a null mutation (Ube3a(m-/p+)), recapitulate major features of AS in humans, including enhanced seizure susceptibility. Excitatory neurotransmission onto neocortical pyramidal neurons is diminished in Ube3a(m-/p+) mice, seemingly at odds with enhanced seizure susceptibility. We show here that inhibitory drive onto neocortical pyramidal neurons is more severely decreased in Ube3a(m-/p+) mice. This inhibitory deficit follows the loss of excitatory inputs and appears to arise from defective presynaptic vesicle cycling in multiple interneuron populations. In contrast, excitatory and inhibitory synaptic inputs onto inhibitory interneurons are largely normal. Our results indicate that there are neuron type-specific synaptic deficits in Ube3a(m-/p+) mice despite the presence of Ube3a in all neurons. These deficits result in excitatory/inhibitory imbalance at cellular and circuit levels and may contribute to seizure susceptibility in AS.

Figure 1. Inhibitory synaptic deficits arising through development in Ube3a^m−/p+ mice are not due to decreased density of inhibitory interneurons

(A) Photomicrographs of immunocytochemical markers for inhibitory interneurons in visual cortex of wildtype (WT) and Ube3a-deficient (m−/p+) mice. (B) Quantification of parvalbumin, calretinin, and somatostatin positive neurons in L2/3 of mouse primary visual cortex at ~P80 (n=4 mice/genotype). (C) Sample mIPSC recordings from ~P25 (P21-P28) and ~P80 (P70-P90) WT and Ube3a^m−/p+ mice (scale bar = 30 pA, 400 ms). (D) Average mIPSC amplitude was similar between WT and Ube3a^m−/p+ mice at ~P25 (WT n=10 cells; m−/p+ n=8 cells) and ~P80 (n=13 cells/genotype) in L2/3 pyramidal cells. (E) Average mIPSC frequency for ~P25 (WT n=10 cells; m−/p+ n=8 cells) and ~P80 (n=13 cells/genotype) (see also Table S1). (F) Illustration of stimulation (L4) and recording (L2/3 pyramidal neuron) configuration in primary visual cortex, and sample recordings of evoked IPSCs at stimulation intensities of 0, 10, 30, and 100 μA (scale bar = 500 pA, 30 ms). (G) Evoked IPSCs at ~P25 (WT n=26 cells; m−/p+ n=29 cells) and (H) ~P80 (WT n=20 cells; m−/p+ n=14 cells). (I) Sample recordings of evoked IPSCs in paired-pulse experiments (scale bar = 100pA, 20 ms) and quantification of paired-pulse ratio in mice aged ~P25 (WT n=12 cells; m−/p+ n=11 cells) and ~P80 (WT n=13 cells; m−/p+ n=8 cells.) (see also Figure S1). All data are represented as the mean ± SEM; **p<0.01, ***p<0.001.

Figure 2. Ube3a loss leads to neuron type-specific defects in inhibitory neurotransmission

(A₁) Illustration and photomicrograph of a filled L2/3 pyramidal neuron and sample recordings from ~P80 WT (upper) and Ube3a^m−/p+ (lower) mice (scale bar = 40 pA, 300 ms.) (A₂) Average mIPSC amplitude (n=11 cells/genotype, p=0.16) and (A₃) frequency (n=11 cells/genotype) from L2/3 pyramidal neurons. (B₁) Illustration and fill of a L2/3 pyramidal neuron and sample recordings from ~P80 WT (upper) and Ube3a^m−/p+ (lower) mice (scale bar 15 pA, 300 ms). (B₂) Average mEPSC amplitude (WT n=6 cells; m−/p+ n=9 cells) and (B₃) frequency (WT n=6 cells; m−/p+ n=9 cells) from L2/3 pyramidal neurons. (C₁) Illustration and fill of a L2/3 FS inhibitory interneuron and sample recordings from ~P80 WT (upper) and Ube3a^m−/p+ (lower) mice (scale bar 40 pA, 300 ms). (C₂) Average mIPSC amplitude (WT n=13 cells; m−/p+ n=10 cells) and (C₃) frequency (WT n=13 cells; m−/p+ n=10 cells; p=0.57) from L2/3 FS inhibitory interneurons. (D₁) Illustration and fill of a L2/3 FS inhibitory interneuron and sample recordings from ~P80 WT (upper) and Ube3a^m−/p+ (lower) mice (scale bar 10 pA, 300 ms). (D₂) Average mEPSC amplitude (WT n=13 cells; m−/p+ n=12 cells) and (D₃) frequency (WT n=13 cells; m−/p+ n=12 cells) from L2/3 FS inhibitory interneurons (see also Figure S2 and Table S2). Bar graphs represent the mean ± SEM; p*<0.05, p***<0.001.

Figure 3. Synaptic deficits arise from both FS and non-FS inhibitory interneurons in Ube3a^m−/p+ mice

(A) Recording configuration (upper) and representative recordings of presynaptic action potentials evoked in a FS inhibitory interneuron (middle), and resulting unitary IPSPs in a L2/3 pyramidal neuron (lower) from WT (blue) and Ube3a^m−/p+ mice (red). Average is of 12 trials. Scale bar “Pre” 20 mV, 40 ms; “Post” 4 mV, 40 ms. (B) Average short-term plasticity of inhibitory to excitatory connections (WT n=8; m−/p+ n=12) (presynaptic spikes @ 30 Hz). (C) Average unitary IPSP amplitude of 1^st IPSP of inhibitory to excitatory connections (WT n=8; m−/p+ n=12). (D) Connection probability of FS inhibitory interneurons to L2/3 pyramidal neurons (numbers of connected/total pairs written in bars); (E) inhibitory drive of FS inhibitory interneurons to L2/3 pyramidal neurons (WT n=19 pairs; m−/p+ n=30 pairs). (F) Recording configuration (upper) and representative recordings (lower) of eIPSCs during baseline (black) and 20 minutes after 500 nM ω-agatoxin-IVa perfusion (green) (scale bar = 100 pA, 20 ms). (G) A 10 minute baseline was recorded before a 10 minute perfusion of agatoxin (green bar) (WT n=10; m−/p+ n=10). (H) Input-output relationship in the agatoxin-insensitive portion of total eIPSC performed 20 minutes after cessation of agatoxin perfusion (WT n=10; m−/p+ n=10). (see also Figure S3 and Table S3). Data are represented as the mean ± SEM; *p<0.05, **p<0.01, ***p<0.001.

Figure 4. Inhibitory synapses of Ube3a^m−/p+ mice have presynaptic defects

(A₁) Electron micrographs of inhibitory synapses stained for GABA in ~P80 WT and Ube3a^m−/p+ mice (green denotes axon terminal, blue denotes soma), insets highlight CCVs (scale bar = 250 nm). Average values of (left to right) (A₂) axon terminal area, (A₃) mitochondrial area divided by terminal area, (A₄) number of synaptic vesicles per square micron, and (A₅) number of CCVs per square micron (WT n = 3; m−/p+ n = 4). (B₁) Electron micrographs of excitatory synapses in WT and Ube3a^m−/p+ mice (green denotes axon terminal, red denotes spine) (scale bar = 250 nm). Average values of (left to right) (B₂) axon terminal area, (B₃) mitochondrial area divided by terminal area, (B₄) number of synaptic vesicles per square micron, and (B₅) number of CCVs per square micron (WT n = 3; m−/p+ n = 4). (C₁) Sample recordings (left) and sample experiment (right) showing baseline (black), high frequency depletion (orange), and recovery (black) phases (scale bars; baseline = 200 pA, 20 ms; depletion = 200 pA, 200 ms; recovery = 200 pA, 20 ms). (C₂) Average depletion phase showing eIPSC amplitude normalized to baseline during 800 stimuli at 10 Hz. 10 responses are averaged for each point and a monophasic exponential was fit to the averaged responses for WT and Ube3a^m−/p+ mice (WT n=19; m−/p+ n=18). (C₃) Average recovery phase showing eIPSC amplitude normalized to baseline during 90 stimuli at 0.33 Hz. 3 responses are averaged for each point and a monophasic exponential was fit to the averaged responses for WT and Ube3a^m−/p+ mice (WT n=19; m−/p+ n=18). (D₁) Sample recordings (left) and sample experiment (right) showing baseline (black), high frequency depletion (orange), and recovery (black) phases (scale bars; baseline = 100 pA, 10 ms; depletion = 100 pA, 200 ms; recovery = 100 pA, 10 ms). (D₂) Average depletion phase showing eEPSC amplitude normalized to baseline during 800 stimuli at 10 Hz. 10 responses are averaged for each point and a monophasic exponential was fit to the averaged responses for WT and Ube3a^m−/p+ mice (WT n=10; m−/p+ n=11). (D₃) Average recovery phase showing eEPSC amplitude normalized to baseline during 90 stimuli at 0.33 Hz. 3 responses are averaged for each point and a monophasic exponential was fit to the averaged responses for WT and Ube3a^m−/p+ mice (WT n=10; m−/p+ n=11) (see also Figure S4). Data are represented as the mean ± SEM; *p<0.05, ***p<0.001.