Excitation/inhibition imbalance and impaired synaptic inhibition in hippocampal area CA3 of Mecp2 knockout mice

Abstract

Rett syndrome (RTT) is a neurodevelopment disorder associated with intellectual disabilities and caused by loss-of-function mutations in the gene encoding the transcriptional regulator Methyl-CpG-binding Protein-2 (MeCP2). Neuronal dysfunction and changes in cortical excitability occur in RTT individuals and Mecp2-deficient mice, including hippocampal network hyperactivity and higher frequency of spontaneous multiunit spikes in the CA3 cell body layer. Here, we describe impaired synaptic inhibition and an excitation/inhibition (E/I) imbalance in area CA3 of acute slices from symptomatic Mecp2 knockout male mice (referred to as Mecp2(-/y) ). The amplitude of TTX-resistant miniature inhibitory postsynaptic currents (mIPSC) was smaller in CA3 pyramidal neurons of Mecp2(-/y) slices than in wildtype controls, while the amplitude of miniature excitatory postsynaptic currents (mEPSC) was significantly larger in Mecp2(-/y) neurons. Consistently, quantitative confocal immunohistochemistry revealed significantly lower intensity of the alpha-1 subunit of GABAA Rs in the CA3 cell body layer of Mecp2(-/y) mice, while GluA1 puncta intensities were significantly higher in the CA3 dendritic layers of Mecp2(-/y) mice. In addition, the input/output (I/O) relationship of evoked IPSCs had a shallower slope in CA3 pyramidal neurons Mecp2(-/y) neurons. Consistent with the absence of neuronal degeneration in RTT and MeCP2-based mouse models, the density of parvalbumin- and somatostatin-expressing interneurons in area CA3 was not affected in Mecp2(-/y) mice. Furthermore, the intrinsic membrane properties of several interneuron subtypes in area CA3 were not affected by Mecp2 loss. However, mEPSCs are smaller and less frequent in CA3 fast-spiking basket cells of Mecp2(-/y) mice, suggesting an impaired glutamatergic drive in this interneuron population. These results demonstrate that a loss-of-function mutation in Mecp2 causes impaired E/I balance onto CA3 pyramidal neurons, leading to a hyperactive hippocampal network, likely contributing to limbic seizures in Mecp2(-/y) mice and RTT individuals.

Keywords: E/I balance; Hippocampus; MeCP2; Rett syndrome; intellectual disability.

Figure 1. Spontaneous quantal synaptic transmission onto CA3 pyramidal neurons

A, Representative examples of mIPSCs. B, The cumulative probability distribution of mIPSC amplitudes recorded in Mecp2^-/y CA3 pyramidal neurons was shifted towards the left compared to wildtype neurons. C, The cumulative probability distribution of mIPSC charges in Mecp2^-/y CA3 pyramidal neurons was shifted towards the left compared to wildtype neurons. D, The cumulative probability distribution of mIPSC inter-event intervals in Mecp2^-/y CA3 pyramidal neurons was shifted towards the left compared to wildtype neurons, i.e. higher mIPSC frequency. E, Representative examples of mEPSCs. F, The cumulative probability distribution of mEPSC amplitudes recorded in Mecp2^-/y CA3 pyramidal neurons was shifted towards the right compared to wildtype neurons. G, The cumulative probability distribution of mEPSC charge in Mecp2^-/y CA3 pyramidal neurons was shifted towards the right compared to wildtype neurons. H, The cumulative probability distribution of mEPSC inter-event intervals in Mecp2^-/y CA3 pyramidal neurons was shifted towards the right compared to wildtype neurons, i.e. lower mEPSC frequency. I, The ratio of the amplitudes of mEPSCs to mIPSCs in CA3 pyramidal neurons is higher in Mecp2^-/y mice compared to wildtype mice. These mEPSCs and mIPSCs where obtained from the same cells at different holding potentials. J, The ratio of the charges of mEPSCs to mIPSCs in CA3 pyramidal neurons is higher in Mecp2^-/y mice compared to wildtype mice; same cells as in I.

Figure 2. Inhibitory and excitatory postsynaptic puncta in area CA3

A top, Representative confocal microscopy images of dual immunolabeling of GABA_Aα1 and the neuronal nuclear marker NeuN in area CA3 of wildtype (left) and Mecp2^-/y mice (right). Region inside white box is enlarged at the bottom, and shows postsynaptic GABA_AR clusters around CA3 pyramidal neurons. bottom, Quantification of the pixel intensity and numerical density of GABA_Aα1 immunopositive puncta in the different layers of area CA3 (S.O. stratum oriens; S.P. stratum pyramidale; S.L. stratum lucidum; S.R. stratum radiatum). B top, Representative confocal microscopy images of immunolabeling of GluA1 in area CA3 of wildtype (left) and Mecp2^-/y mice (right). Region inside white box is enlarged at the bottom, and shows postsynaptic AMPAR clusters in the dendritic regions of CA3. Scale bars are 50μm (top) and 10μm (insets). bottom, Quantification of the pixel intensity and numerical density of GluA1 immunopositive puncta in the different layers of area CA3; asterisks denote p<005.

Figure 3. Evoked synaptic inhibition onto CA3 pyramidal neurons

A, Electrode placement and representative examples of IPSCs evoked in CA3 pyramidal neurons clamped at 0mV by stimulation of local interneurons within CA3 stratum lucidum in the presence of APV/CNQX. B, Input-Output (I-O) relationship between the amplitude of evoked IPSCs and stimulation intensity. The fitted slope (m) of the linear portion of the I-O curve was significantly smaller in Mecp2^-/y slices than in wildtype slices.

Figure 4. Interneuron cell density in area CA3

A left, Representative example of a wildtype brain section including the hippocampus, which was immunostained with anti-GAD1 antibodies to illustrate the region used for quantitative analyses of interneuron cell density (confocal microscopy). Lines enclose the region analyzed for cell density: the wedge of dorsal hippocampus from the tips of the dentate gyrus laterally to the point of inflection of Cornu Ammonis. Arrowheads point to GAD1 immunopositive cells. right, number of GAD1 positive cells per mm² of area CA3. B left, Representative examples of PV and GAD1 double immunostaining in area CA3 of wildtype (top) and Mecp2^-/y (bottom) mice. Arrowheads point to PV/GAD1 double-positive cells. right, number of PV/GAD1 double-positive positive cells per mm² of area CA3. C left, Representative examples of SST and GAD1 immunostaining in area CA3 of wildtype (top) and Mecp2^-/y (bottom) mice. Arrowheads point to SST/GAD1 double-positive cells. right, number of SST/GAD1 double-positive positive cells per mm² of area CA3.

Figure 5. Intrinsic and synaptic properties of CA3 fast-spiking basket interneurons

A. Representative subthreshold voltage responses in CA3 fast-spiking basket cells of wildtype (left) and Mecp2^-/y mice (right). Current injections from −100 to 60pA in 20pA increments. B. I-O relationship between current intensity and subthreshold voltage responses. No significant differences were observed between wildtype (n=24 cells/14 mice) and Mecp2^-/y interneurons (n=29 cells/20 mice). C,D. The amplitudes of the fAHP (C) and sAHP (D) were similar between two genotypes. E. Representative firing responses of fast-spiking basket cells from wildtype (left) and Mecp2^-/y mice (right) evoked by 80pA (top) and 200pA (bottom) current injections. F. I-O relationship between current intensity and the number of action potentials. Two-way repeated ANOVA analysis shows significant differences between two genotypes (p=0.003); post hoc comparisons indicate tendency to significance for fewer spikes in Mecp2^-/y mice for the low current injection groups (60pA, p=0.068; 80pA, p=0.079; 100pA, p=0.078). G. Representative example of a wildtype fast-spiking basket cell filled with biocytin, stained with Alexa-488-conjugated streptavidin, and imaged by confocal microscopy; scale bar=100μm. H. Representative examples of spontaneous firing recorded in CA3 fast-spiking basket cells. I. Cumulative probability distribution of spontaneous firing in Mecp2 knockout interneurons was shifted towards the right (wildtype n=14 cells/8 mice, Mecp2^-/y n=23 cells/11 mice; K-S test p<0.001). J. Representative examples of mEPSCs in CA3 fast-spiking basket cells of wildtype (top) and Mecp2^-/y mice (bottom). K,L. Cumulative probability distribution of mEPSC amplitude (K) in Mecp2 knockout interneurons was shifted towards the left (wildtype n=7 cells/6 mice, Mecp2^-/y n=13 cells/8 mice; K-S test p<0.001), and that of mEPSC inter-event intervals (L) towards the right (K-S test p<0.001).