Increased Sociability in Mice Lacking Intergenic Dlx Enhancers

Abstract

The Dlx homeodomain transcription factors play important roles in the differentiation and migration of GABAergic interneuron precursors. The mouse and human genomes each have six Dlx genes organized into three convergently transcribed bigene clusters (Dlx1/2, Dlx3/4, and Dlx5/6) with cis-regulatory elements (CREs) located in the intergenic region of each cluster. Amongst these, the I56i and I12b enhancers from the Dlx1/2 and Dlx5/6 locus, respectively, are active in the developing forebrain. I56i is also a binding site for GTF2I, a transcription factor whose function is associated with increased sociability and Williams-Beuren syndrome. In determining the regulatory roles of these CREs on forebrain development, we have generated mutant mouse-lines where Dlx forebrain intergenic enhancers have been deleted (I56i^(-/-), I12b^(-/-)). Loss of Dlx intergenic enhancers impairs expression of Dlx genes as well as some of their downstream targets or associated genes including Gad2 and Evf2. The loss of the I56i enhancer resulted in a transient decrease in GABA⁺ cells in the developing forebrain. The intergenic enhancer mutants also demonstrate increased sociability and learning deficits in a fear conditioning test. Characterizing mice with mutated Dlx intergenic enhancers will help us to further enhance our understanding of the role of these Dlx genes in forebrain development.

Keywords: Dlx; GABA; Williams–Beuren syndrome; autism; development; gene regulation.

FIGURE 1

Targeting strategy for the generation of mutant mice with deletions of forebrain Dlx enhancers. (A) Strategy for the deletion of enhancer I56i. (B) Strategy for the deletion of enhancer I12b. Homologous recombination in embryonic stem cells was completed using a BAC targeting construct wherein the enhancer was replaced with a floxed neomycin cassette. Recombinant clones were transferred into a blastocyst to generate chimeric embryos. Mice positive for the recombinant Dlx loci were crossed with CRE mice to remove the neomycin cassette. Black arrows, transcriptional orientation of each gene; black triangles, lox p locations. A, AgeI; H, HindIII; S, SalI; E, EcoRI; EV, EcoRV; K, KpnI; X, XbaI.

FIGURE 2

Dlx and Dlx target gene expression are reduced in the developing forebrain of ΔI56i mice. (A–F,A′–F′) In situ hybridization on the ventral telencephalon at E11.5 for (A,A′) Dlx1, (B,B′) Dlx2, (C,C′) Dlx5, (D,D′) Dlx6, (E,E′) Gad2, and (F,F′) Evf2. (A–F) WT mice and (A′–F′) homozygous ΔI56i mice. Scale bar = 50 μm. (G–I) qRT-PCR of ventral telencephalon isolated from WT and homozygous I56i^(–/–) mice. (G) E11.5, (H) E14.5, and (I) P0. Bars are WT (black), homozygous I56i^(–/–) (white). Data presented as mean for the n values as indicated except for Gad1 and Gad2 at P0 for which n = 3. Error bars represent SEM. Data analyzed using a two tailed t-test (**p < 0.01, ***p < 0.001). SVZ, subventricular zone; VZ, ventricular zone.

FIGURE 3

Dlx expression is reduced in the developing forebrain of mice homozygous for a variant I56i enhancer. (A–C,A′–C′) In situ hybridization on forebrain for Dlx5 at (A,A′) E11.5, (B,B′) E13.5 rostral slice, and (C,C′) E13.5 caudal slice. (A–C) WT mice and (A′–C′) homozygous vI56i mice. Scale bar = 1 mm (A,A′) and 1 mm (B,C,B′,C′). LGE, lateral ganglionic eminence; MGE, median ganglionic eminence; CGE, caudal ganglionic eminence; Di, diencephalon. (D–F) qRT-PCR of ventral telencephalon isolated from WT and homozygous vI56i mice. (D) E11.5, (E) E13.5, and (F) P0. WT (black), homozygous vI56i (white). Data presented as mean for the n values listed in brackets. Error bars represent SEM. Data analyzed using a two tailed t-test (*p < 0.05).

FIGURE 4

I12b enhancer deletion and combined I12b/I56i mutations impairs Dlx and Gad2 expression levels. qRT-PCR of ventral telencephalon isolated from WT mice and mice (A) I12b^(–/–), (B) homozygous for both the I12b and I56i deletions, and (C) homozygous for both I12b^(–/–) and vI56i. Bars are WT (black), mutant (white). Data presented as a mean for the n values listed in brackets. Error bars represent SEM. Data analyzed using a two tailed t-test (*p < 0.05, **p < 0.01, ***p < 0.001).

FIGURE 5

Mutations in the I56i enhancer decreases number of GABA⁺ cells at midgestation. (A–D,A′–D′) Immunostaining on E13.5 forebrain tissue with an anti GAD65 antibody. (A,C) WT, (B) homozygous I56i^(–/–), and (D) homozygous vI56i. Scale bar (A–D) = 100 μM and (A′–D′) = 50 μM. (E) schematic drawing of the mouse P35 forebrain, with a box indicating the position of the somatosensory cortex imaged. (F,G,F′,G′) Immunostaining for calbindin using an anti-calbindin antibody on P35 somatosensory cortex. (F) WT and (G) homozygous I56i^(–/–). (H,I,H′,I′) Immunostaining for parvalbumin using an anti-parvalbumin antibody on P35 somatosensory cortex. (H) WT and (I) homozygous I56i^(–/–). Scale bar (F–I) = 50 μm and (F′–I′) = 25 μm. LGE, lateral ganglionic eminence; CB, calbindin; PV, parvalbumin.

FIGURE 6

Mice with deletions of Dlx enhancers show an increased propensity for sociability and memory/learning defects. (A) Social interaction test performed on WT (black), homozygous I56i^(–/–) (white) and homozygous I56i^(–/–) – I12b^(–/–) (stripes). (B) Social interaction test preformed on WT (black) and homozygous vI56i (white). (C) Fear conditioning test preformed on WT (black) homozygous I56i^(–/–) (white) and homozygous I56i^(–/–) – I12b^(–/–) (stripes). Numbers of mice listed in the brackets of each graph. Data shown as mean, error bars represent SEM. Data was analyzed using one-way ANOVA with Turkey’s multiple comparison test (*p < 0.05, ***p < 0.001).