Structural alterations in the amygdala and impaired social incentive learning in a mouse model of a genetic variant associated with neurodevelopmental disorders

Abstract

Copy number variants (CNVs) are robustly associated with psychiatric disorders and their dimensions and changes in brain structures and behavior. However, as CNVs contain many genes, the precise gene-phenotype relationship remains unclear. Although various volumetric alterations in the brains of 22q11.2 CNV carriers have been identified in humans and mouse models, it is unknown how the genes in the 22q11.2 region individually contribute to structural alterations and associated mental illnesses and their dimensions. Our previous studies have identified Tbx1, a T-box family transcription factor encoded in 22q11.2 CNV, as a driver gene for social interaction and communication, spatial and working memory, and cognitive flexibility. However, it remains unclear how TBX1 impacts the volumes of various brain regions and their functionally linked behavioral dimensions. In this study, we used volumetric magnetic resonance imaging analysis to comprehensively evaluate brain region volumes in congenic Tbx1 heterozygous mice. Our data show that the volumes of anterior and posterior portions of the amygdaloid complex and its surrounding cortical regions were reduced in Tbx1 heterozygous mice. Moreover, we examined the behavioral consequences of an altered volume of the amygdala. Tbx1 heterozygous mice were impaired for their ability to detect the incentive value of a social partner in a task that depends on the amygdala. Our findings identify the structural basis for a specific social dimension associated with loss-of-function variants of TBX1 and 22q11.2 CNV.

Keywords: 22q11.2; 22q11.2 CNV; CNV; Tbx1; amygdala; autism; schizophrenia; social incentive motivation; volumetric MRI.

Figure 1.

(A) Cohen’s D values of the % volume differences between wild-type (WT) and Tbx1 heterozygous (HT) mice. The % volume of each region relative to the total volume was used to compute Cohen’s D values (HT–WT). In HT mice, volume increases are shown as positive values, and volume decreases are shown as negative values (see Table S1 for all values). WT, n=11; HT, n=7. CI, confidence interval. (B) The mean % ± standard error of the mean of the amygdala relative to the whole brain volume (mm³) is smaller in heterozygous (HT) mice than in wild-type (WT) mice (t(16)=2.451, p=0.026). WT, n=11; HT, n=7. *p<0.05. (C) Voxel-based analysis of two-group differences between WT and HT mice. Permutation-based nonparametric testing with 500 random permutations was performed (small volume correction within the amygdala region of interest, p<0.05). A total of 409 voxels (0.21 mm³) survived this testing.

Figure 2.

(A) Cohen’s D values of the percent volume differences between wild-type (WT) and Tbx1 heterozygous (HT) mice. The percent volume of each cortical segment relative to the total volume was used to compute Cohen’s D values (HT–WT). In HT mice, volume increases are shown as positive values, and volume decreases are shown as negative values (see Table S2 for all values). Among 72 cortical regions examined (see Table S2), only those with Cohen’s D values greater than 1.0 and smaller than −1.0 are listed. The statistically significant differences between the selected 10 regions between WT and HT mice survived the correction for multiple comparisons with a false-discovery rate of 5%. WT, n=11; HT, n=7. CI, confidence interval. (B) A 3D visualization of cortical segments with Cohen’s D values greater than 1.0 and smaller than −1.0. A clustering of volume reductions in and around the amygdala is noticeable (see also Fig. S4 for a movie version).

Figure 3.

(A) Calretinin-positive neuropil in the amygdalopiriform transition area (red arrows) in wild-type (WT) and Tbx1 heterozygous (HT) mice. The calretinin-positive neuropil was delineated within the anteroposterior extent of the region (Bregma from −2.46 mm to −3.64 mm from bregma), WT, n=4; HT, n=4. (B) The size of the calretinin-enriched neuropil in the amygdalopiriform transition zone was smaller in HT than in WT mice (t(6)=3.489, *p=0.0130).

Figure 4.

A) Experimental design. Each mouse was given a 30-min preconditioning session, two 24-hr conditioning sessions, and a 30-min postconditioning session. Each mouse test subject was caged with a littermate cage partner of the same or different genotype from weaning to testing. During conditioning, each test subject was conditioned with its own cage partner in one compartment and no partner in the other. Mice from a separate group were placed in both compartments without their cage partner. On the postconditioning day, each mouse was evaluated for their preference for or avoidance of each compartment without the cage partner. The average ± standard error of the mean time shift was calculated from the pre- to postconditioning sessions of wild-type (WT) and Tbx1 heterozygous (HT) mice, conditioned with (B) or without (C) a home-cage partner. Wilcoxon nonparametric tests were applied to examine a shift in time from preconditioning session to postconditioning session. *p<0.05, **p<0.01.