Mild expression differences of MECP2 influencing aggressive social behavior

Abstract

The X-chromosomal MECP2/Mecp2 gene encodes methyl-CpG-binding protein 2, a transcriptional activator and repressor regulating many other genes. We discovered in male FVB/N mice that mild (~50%) transgenic overexpression of Mecp2 enhances aggression. Surprisingly, when the same transgene was expressed in C57BL/6N mice, transgenics showed reduced aggression and social interaction. This suggests that Mecp2 modulates aggressive social behavior. To test this hypothesis in humans, we performed a phenotype-based genetic association study (PGAS) in >1000 schizophrenic individuals. We found MECP2 SNPs rs2239464 (G/A) and rs2734647 (C/T; 3'UTR) associated with aggression, with the G and C carriers, respectively, being more aggressive. This finding was replicated in an independent schizophrenia cohort. Allele-specific MECP2 mRNA expression differs in peripheral blood mononuclear cells by ~50% (rs2734647: C > T). Notably, the brain-expressed, species-conserved miR-511 binds to MECP2 3'UTR only in T carriers, thereby suppressing gene expression. To conclude, subtle MECP2/Mecp2 expression alterations impact aggression. While the mouse data provides evidence of an interaction between genetic background and mild Mecp2 overexpression, the human data convey means by which genetic variation affects MECP2 expression and behavior.

Figure 1

Spontaneous home cage activity of FVB/N and C57BL/6N mice of both genders is modulated by mild Mecp2 overexpression. Results for male and female mice of both strains are presented. With the exception of male FVB/N mice, locomotion, immobility and climbing reveal similar TG effects across genders and genetic backgrounds. N = 10–28; mean ± s.e.m. given.

Figure 2

Territorial aggressive behavior in male mice is influenced by Mecp2 overexpression and genetic background. A Latency to attack in the resident-intruder test is significantly reduced in male FVB/N TG mice but increased in C57BL/6N. Note the different cut-off for the 2 strains. B Sociability testing in the tripartite chamber reveals a highly significant preference of male mice, independent of the genetic background, for a stranger mouse as compared to an object (empty cage). Data information: N = 10–24; mean ± s.e.m. given.

Figure 3

Male TG C57BL/6N mice show reduced territorial social interaction as well as inferior social competence. Upper 2 rows: Frequency and duration of determinants of territorial social interaction, that is, follow/chase behavior, nose/snout sniff and anogenital sniff, are consistently reduced in TG carriers. Lower row: TG mice are inferior in nest building, social dominance behavior and ultrasound vocalization. N = 10–23; mean ± s.e.m. given.

Figure 4

Pentylenetetrazole-induced seizure propensity is increased upon mild Mecp2 overexpression independent of strain and gender, whereas the startle response is augmented in females only. Upper row: Higher seizure scores are found in TG carriers across gender and genetic backgrounds. N = 6–14; Lower row: Significant increase in the startle response is observed only in female mice of both genetic backgrounds. N = 11–28; mean ± s.e.m. given.

Figure 5

Basic genetics of MECP2: Gene structure, Hardy-Weinberg statistics, linkage disequilibrium, and case-control analysis in schizophrenic (GRAS) patients and healthy individuals. A Schematic overview of MECP2 isoforms e1 and e2, including SNP positions. Digits depict exon numbers, solid black lines exon usage for the respective isoform. Dashed lines denote SNP positions. Black fillings in boxes denote coding sequence, isoform-specific in exons 1 and 2. B Test for deviation from Hardy-Weinberg-Equilibrium (HWE) in females only due to X-chromosomal location of MECP2. C Case–control association analysis - separate for both genders - reveals similar distribution of SNPs in patients and controls. D Linkage disequilibrium for all included GRAS patients and controls.

Figure 6

SNP rs2734647 in the 3'UTR of MECP2: Search for mechanistic insight. A Human miRNAs predicted to bind to the MECP2 3′UTR in an rs2734647 allele-specific manner. The bases corresponding to the SNP position are black-shadowed. Numbers left and right of the nucleotide sequence refer to its base-pair position within the miRNA sequence. B, C Luciferase assay results showing relative luciferase activity in HEK293 and N2a cells after co-transfection of candidate miRNAs with phRL-TK rs2734647C, or rs2734647T, respectively; mean ± s.e.m.;N = 7 (refers to biological replicates) for all conditions. Statistical significance was calculated relative to the non-transfection control (100%). D, E Relative expression of hsa-miR-511 and of MECP2 isoform 2 (MECP2 e2) or both isoforms (MECP2 e1&2) in aggression/impulsivity – relevant brain areas: FC=frontal cortex, PFC, prefrontal cortex; TC, temporal cortex; OC, occipital cortex; HC, hippocampus; AM, amygdala; as well as in placenta (N = 1) as control tissue. Numbers of individual brains included in the analysis are given in brackets; mean ± s.e.m. F Relative expression of MECP2 e2 or MECP2 e1&2 in peripheral blood mononuclear cells (PBMC) of male patients dependent on rs2734647 genotype; N numbers of individuals in brackets; mean ± s.e.m. G Alignment of human and mouse MECP2 3′UTRs around rs2734647 SNP position (black-shadowed) and human and mouse miR-511, illustrating a perfect species-specific seed match; hsa-miR-511 perfectly matches the human MECP2 3′UTR in case of rs2734647 T. Additional mismatches are gray-shadowed.