Reversal of Social Recognition Deficit in Adult Mice with MECP2 Duplication via Normalization of MeCP2 in the Medial Prefrontal Cortex

Abstract

Methyl-CpG binding protein 2 (MeCP2) is a basic nuclear protein involved in the regulation of gene expression and microRNA processing. Duplication of MECP2-containing genomic segments causes MECP2 duplication syndrome, a severe neurodevelopmental disorder characterized by intellectual disability, motor dysfunction, heightened anxiety, epilepsy, autistic phenotypes, and early death. Reversal of the abnormal phenotypes in adult mice with MECP2 duplication (MECP2-TG) by normalizing the MeCP2 levels across the whole brain has been demonstrated. However, whether different brain areas or neural circuits contribute to different aspects of the behavioral deficits is still unknown. Here, we found that MECP2-TG mice showed a significant social recognition deficit, and were prone to display aversive-like behaviors, including heightened anxiety-like behaviors and a fear generalization phenotype. In addition, reduced locomotor activity was observed in MECP2-TG mice. However, appetitive behaviors and learning and memory were comparable in MECP2-TG and wild-type mice. Functional magnetic resonance imaging illustrated that the differences between MECP2-TG and wild-type mice were mainly concentrated in brain areas regulating emotion and social behaviors. We used the CRISPR-Cas9 method to restore normal MeCP2 levels in the medial prefrontal cortex (mPFC) and bed nuclei of the stria terminalis (BST) of adult MECP2-TG mice, and found that normalization of MeCP2 levels in the mPFC but not in the BST reversed the social recognition deficit. These data indicate that the mPFC is responsible for the social recognition deficit in the transgenic mice, and provide new insight into potential therapies for MECP2 duplication syndrome.

Keywords: CRISPR-Cas9; MECP2 duplication; Medial prefrontal cortex; Social recognition deficit.

Fig. 1

MECP2-TG mice display reduced locomotor activity, enhanced aversive behaviors, and a social recognition deficit. A–CMECP2-TG mice and wild-type littermates manifested no side preference (A) and normal social approach behavior (B), but MECP2-TG mice did not distinguish familiar partners from novel partners (C) as wild-type littermates did in the three-chamber test. DMECP2-TG mice travelled less in the open field than their wild-type littermates. E, F Compared to their wild-type littermates, MECP2-TG mice spent less time on the light side of the light/dark exploration box, and less time in the open arms of an elevated plus maze. G–J Percentage freezing was significantly higher in the MECP2-TG mice than in wild-type littermates during the training phase (G), the contextual fear memory testing phase (H), the cued fear memory testing phase (I), or when presented with a mouse in a new context after training (J). n = 10–14 for wild-type, n = 10–14 for MECP2-TG. Data are the mean ± SEM. ns, not significant. *P < 0.05, **P < 0.01, ***P < 0.001 (two-tailed paired t-test in A–C, two-tailed t-test in D–F and H–J, and two-way repeated-measures ANOVA in G).

Fig. 2

MECP2-TG mice display normal appetitive behaviors, novel object recognition, and spatial learning and memory. A, BMECP2-TG mice and wild-type littermates showed comparable sucrose preference (A), and cocaine conditioned place preference (B). C In the novel object recognition test, MECP2-TG mice and wild-type littermates showed no side preference during habituation, and normal preference for the novel object in the test phase. D–F No significant difference was found between MECP2-TG mice and wild-type littermates during the learning phase (D), the short-term memory test (E), and the long-term memory test (F) in the Barnes maze test. n = 10–14 for wild-type, n = 10–14 for MECP2-TG. Data represent the mean ± SEM. ***P < 0.001 (two-tailed t-test in A, B, E, and F, two-tailed paired t-test in C, and two-way repeated-measures ANOVA in D).

Fig. 3

Significant ReHo value differences between MECP2-TG and wild-type mice in the whole brain resting-state fMRI. A Regions manifesting significant ReHo value differences between MECP2-TG and wild-type mice (red, MECP2-TG > wild-type; blue, wild-type > MECP2-TG; color bar represents significance of difference). B Quantification of ReHo values in specific brain areas [n = 10 for wild-type, n = 14 for MECP2-TG; data represent the mean ± SEM; *P < 0.05, **P < 0.01 (two-tailed t-test)]. mPFC, medial prefrontal cortex; AI, agranular insular area; ACB, nucleus accumbens; CP, caudoputamen; PIR, piriform area; BST, bed nuclei of the stria terminalis; ACA, anterior cingulate area; TH, thalamus; HY, hypothalamus; HIP, hippocampal region; Amy, amygdala; RSP, retrosplenial area; VIS, visual areas; PAG, periaqueductal gray.

Fig. 4

Strategy for specific knockout of the human MECP2 copy in MECP2 duplication neurons and brain areas of interest in MECP2 duplication mice. A AAV-SpCas9, AAV-Guide, and AAV-GFP expression vectors. B SpCas9 target location of the human MECP2 locus (blue, targeted genomic locus; yellow, PAM (protospacer adjacent motif) sequence; red, mismatched bases between mouse Mecp2 and the human MECP2 gene in the targeted locus). C Left, protocol for and analysis of on-target gene editing in vitro. The efficiency of indel mutations in the MECP2 locus was ~17.8%, but no indel mutations were detected in the Mecp2 locus in infected cultured neurons. Right, representative mutation patterns detected by sequencing of the MECP2 locus (red arrowhead, Cas9 cutting site). D Representative images and analysis of MeCP2 expression after CRISPR-Cas9-mediated human MECP2 knockout in the mPFC in vivo (scale bars, 500 μm; n = 9 brain slices from 3 mice for each genotype). E Western blots and analysis of MeCP2 protein expression 1.5 months after AAV injection [n = 6/group; data represent the mean ± SEM; ***P < 0.001 (two-tailed t-test in D, and one-way ANOVA followed by Tukey’s honest significant difference post-hoc correction for multiple comparisons in E)].

Fig. 5

Restoration of normal MeCP2 levels in the mPFC reverses the social recognition deficit in adult MECP2-TG mice. A Schematic of the experimental design. B–D Normalizing MeCP2 levels in the mPFC reverses impaired social recognition behavior (D) in the three-chamber test. No side preference (B) and normal social approach (C) were found in the test. E–J Normalizing MeCP2 levels in the mPFC of MECP2-TG mice did not reverse the decreased locomotor activity in the open field test (E), the heightened anxiety-like behaviors in light/dark exploration (F) and elevated plus maze (G) tests, the percentage freezing when tested on contextual fear memory (H) and cued fear memory (I), and when presented with a mouse in a new context (J). n = 15 each for Wildtype–CTL, MECP2-TG–CTL, and MECP2-TG–normalized; data represent the mean ± SEM; ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.001 (two-tailed paired t-test in B–D, and one-way ANOVA followed by Tukey’s honest significant difference post-hoc correction for multiple comparisons in E–J).

Fig. 6

Normalization of MeCP2 levels in the BST does not reverse the behavioral defects in MECP2-TG mice. A Schematic of the experimental design. B Representative image of the injection site in the BST (scale bar, 500 μm). C–K Normalizing MeCP2 levels in the BST of MECP2-TG mice did not reverse the social recognition deficit in the three-chamber test (C–E), the decreased locomotor activity in the open field test (F), the heightened anxiety-like behaviors in the light/dark exploration (G) and elevated plus maze tests (H), and the fear generalization phenotype in the fear conditioning test (I–K). n = 14 for wild-type–CTL, n = 14 for MECP2-TG–CTL, n = 13 for MECP2-TG–normalized. Data represent the mean ± SEM; ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.001 (two-tailed paired t-test in C–E, one-way ANOVA followed by Tukey’s honest significant difference post-hoc correction for multiple comparisons in G–K, and one-way ANOVA followed by Fisher’s least significant difference post-hoc test in F).