16p11.2 deletion mice exhibit compromised fronto-temporal connectivity, GABAergic dysfunction, and enhanced attentional ability

Abstract

Autism spectrum disorders are more common in males, and have a substantial genetic component. Chromosomal 16p11.2 deletions in particular carry strong genetic risk for autism, yet their neurobiological impact is poorly characterised, particularly at the integrated systems level. Here we show that mice reproducing this deletion (16p11.2 DEL mice) have reduced GABAergic interneuron gene expression (decreased parvalbumin mRNA in orbitofrontal cortex, and male-specific decreases in Gad67 mRNA in parietal and insular cortex and medial septum). Metabolic activity was increased in medial septum, and in its efferent targets: mammillary body and (males only) subiculum. Functional connectivity was altered between orbitofrontal, insular and auditory cortex, and between septum and hippocampus/subiculum. Consistent with this circuit dysfunction, 16p11.2 DEL mice showed reduced prepulse inhibition, but enhanced performance in the continuous performance test of attentional ability. Level 1 autistic individuals show similarly heightened performance in the equivalent human test, also associated with parietal, insular-orbitofrontal and septo-subicular dysfunction. The data implicate cortical and septal GABAergic dysfunction, and resulting connectivity changes, as the cause of pre-attentional and attentional changes in autism.

Fig. 1. Overt alterations in brain region metabolism and altered inter-regional functional connectivity in 16p11.2 DEL mice.

a 16p11.2 DEL mice have decreased activity in the piriform (PirC, p = 0.008) and auditory (AudC, p = 0.014) cortex with increased metabolic activity in the medial septum (MS, p = 0.001) and mammillary body (MB, p = 0.014). The impact of 16p11.2 deletion on metabolism in these regions was not significantly influenced by sex, so the data are pooled for both sexes. Increased functional activity in the hippocampal subiculum (Sub) was found only in males (sex x genotype interaction: p = 0.03, ANOVA). *P < 0.05 and ***P < 0.001 difference from WT (ANOVA, main effect of genotype). ^#P < 0.05 difference from same sex WT (Tukey’s HSD). Box plots show median and interquartile range with “Tukey” whiskers. b Heatmap showing the altered inter-regional connectivity of regions with altered metabolism in 16p11.2 DEL mice, determined using PLSR. Red denotes inter-regional connectivity present in 16p11.2 DEL but not WT mice (VIP 95% CI > 1.0 in 16p11.2 DEL and <1.0 in WT). Light red denotes z > 1.96 and dark red z > 2.58 difference between genotypes. Blue denotes lost inter-regional connectivity in 16p11.2 DEL mice (VIP 95% CI < 1.0 in 16p11.2 DEL and >1.0 in WT). Light blue denotes z > 1.96 and dark blue denotes z > 2.58 difference between genotyes. Data for Sub show connectivity changes selectively found in 16p11.2 DEL male mice that are not found in 16p11.2 DEL female mice. n = 12 (6 male) mice/genotype. Prefrontal cortex: aPrL Anterior Prelimbic Cortex, FRA Frontal Association Area, DLO Dorsolateral Orbital Cortex, MO Medial Orbital Cortex, LO Lateral Orbital Cortex, mPrL Medial Prelimbic Cortex, IL Infralimbic Cortex, Cg1 Cingulate Cortex; Cortex: Ins Insular Cortex, Piri Piriform Cortex, RSC Retrosplenial Cortex, S1 Somatosensory Cortex; Thalamus: AM Anteromedial Thalamus, TRN Thalamic Reticular Nucleus, aRT Anterior Reticular Thalamus, AV Anteroventral Thalamus, MD Mediodorsal Thalamus, VL Ventrolateral Thalamus, VM Ventromedial Thalamus; Mesolimbic: AcbC Nucleus Accumbens Core, AcbSh Nucleus Accumbens Shell, VTA ventral Tegmental Area; Basal Ganglia: DLST Dorsolateral Striatum, VMST Ventromedial Striatum, GP Globus Pallidus, SNC Substantia Nigra pars Compacta, SNR Substantia Nigra pars Reticulata; Amygdala: BLA Basolateral Amygdala, CeA Central Amygdala, MeA Medial Amygdala; Auditory: IC Inferior Colliculus, MG Medial Geniculate, DB Septum/Diagonal Band of Broca, LS Lateral Septum, MS Medial Septum, HDB Horizontal Limb of DB, VDB Vertical Limb of DB; Hippocampus: DSub Dorsal Subiculum, DH CA1 Dorsal Cornu Ammonis 1, DH CA2 Dorsal Cornu Ammonis 2, DH Mol Dorsal Molecular Layer, DH DG Dorsal Dentate Gyrus, VH CA1 Ventral Cornu Ammonis 1, VH CA3 Ventral Cornu Ammonis 3, VH Mol Ventral Molecular Layer, VH DG Ventral Dentate Gyrus; Multimodal: Hab Habenula, MB Mammillary Body. Regional abbreviations are also shown in the supplemental material (Table S3).

Fig. 2. Altered inter-regional connectivity underlies the altered hub region connectivity seen in the functional brain networks of 16p11.2 DEL mice.

Chord diagrams showing lost (dark blue) and gained (red) inter-regional connectivity for the a Frontal Association Cortex (FRA), b Dorsal Hippocampus Dentate Gyrus (DH DG) c Insular Cortex (Ins) d Ventromedial Thalamus (VM) in e Vertical Limb of the Diagonal Band of Broca (VDB) and f Horizontal Limb of the Diagonal Band of Broca (HDB) in 16p11.2 DEL mice. Dark blue denotes significantly lost inter-regional connectivity (VIP 95% CI < 1.0 in 16p11.2 DEL and >1.0 in WT; z <−1.96 in DELs v WT) and red denotes significantly gained connectivity (VIP 95% CI > 1.0 16p11.2 DEL and <1.0 WT; z > 1.96 in DELs v WT) in 16p11.2 DEL mice. n = 12 (6 male) mice/genotype. Prefrontal cortex: aPrL Anterior Prelimbic Cortex, FRA Frontal Association Area, DLO Dorsolateral Orbital Cortex, MO Medial Orbital Cortex, LO Lateral Orbital Cortex, mPrL Medial Prelimbic Cortex, IL Infralimbic Cortex, Cg1 Cingulate Cortex; Cortex: Ins Insular Cortex, Piri Piriform Cortex, RSC Retrosplenial Cortex, S1 Somatosensory Cortex; Thalamus: AM Anteromedial Thalamus, TRN Thalamic Reticular Nucleus, aRT Anterior Reticular Thalamus, AV Anteroventral Thalamus, MD Mediodorsal Thalamus, VL Ventrolateral Thalamus, VM Ventromedial Thalamus; Mesolimbic: AcbC Nucleus Accumbens Core, AcbSh Nucleus Accumbens Shell, VTA ventral Tegmental Area; Basal Ganglia: DLST Dorsolateral Striatum, VMST Ventromedial Striatum, GP Globus Pallidus, SNC Substantia Nigra pars Compacta, SNR Substantia Nigra pars Reticulata; Amygdala: BLA Basolateral Amygdala, CeA Central Amygdala, MeA Medial Amygdala; Auditory: IC Inferior Colliculus, MG Medial Geniculate, DB Septum/Diagonal Band of Broca, LS Lateral Septum, MS Medial Septum, HDB Horizontal Limb of DB, VDB Vertical Limb of DB; Hippocampus: DSub Dorsal Subiculum, DH CA1 Dorsal Cornu Ammonis 1, DH CA2 Dorsal Cornu Ammonis 2, DH Mol Dorsal Molecular Layer, DH DG Dorsal Dentate Gyrus, VH CA1 Ventral Cornu Ammonis 1, VH CA3 Ventral Cornu Ammonis 3, VH Mol Ventral Molecular Layer, VH DG Ventral Dentate Gyrus; Multimodal: Hab Habenula, MB Mammillary Body. Brain region abbreviations are also shown in the supplemental material (Supplementary Table 3). Neural systems are colour coded and indicated in the Figure legend.

Fig. 3. 16p11.2 DEL mice have decreased parvalbumin (Pvalb) and increased calbindin (Calb1) mRNA expression in the prefrontal (PFC) and orbital (OFC) cortex.

Representative autoradiographs for a Pvalb and c Calb1 mRNA expression. b Pvalb expression is decreased in the prelimbic cortex (PrL), infralimbic cortex (IL) and medial orbital (MO) cortex, but not the lateral orbital (LO) cortex of 16p11.2 DEL mice. d Calb1 expression is increased in the medial prefrontal (PrL and IL), but not the orbital (MO and LO) regions in 16p11.2 DEL mice. e location of regions analysed: PrL (orange), IL (purple), MO (blue) and LO (green). Scale bar represents 750 μm. Box plots show median and interquartile range with “Tukey” whiskers. WT = wild-type mice; 16p11.2 DEL = 16p11.2 deletion mice. Data are pooled for sex, as no significant sex x genotype interactions were found. Data were analysed by ANOVA; F and p values are provided in Supplementary Table 1. *P < 0.05, ***P < 0.001, ANOVA main effect of genotype. n = 12 (8 male) mice/genotype.

Fig. 4. Male, but not female, 16p11.2 DEL mice have decreased GABAergic gene expression in the septum and insular cortex.

Gad1 (Gad67) mRNA expression is decreased in the a medial septum, b lateral septum and c insular cortex in male 16p11.2 DEL mice. Sst mRNA expression is not altered in the f insular cortex of 16p11.2 DEL mice. Pvalb expression g is increased in female 16p11.2 DEL mice in the medial septum, but h unaltered in the insular cortex of 16p11.2 DEL mice. Representative autoradiographs for d Gad1 and i Pvalb expression in male mice. e Location of regions analysed: LS (orange), MS (blue) and Insula (red). Scale bar represents 1000 μm. Box plots show median and interquartile range with “Tukey” whiskers. WT F = wild-type female mice; WT M = wild-type male mice; 16p11 DEL F = 16p11.2 deletion female mice; 16p11.2 DEL M = 16p11.2 deletion male mice. Data were analysed by ANOVA. F and p values are provided in Supplementary Table 1. ^#P < 0.05, ^##P < 0.01, Fisher’s post hoc test, n = 12 (8 male) mice/genotype.

Fig. 5. Male, but not female, 16p11.2 DEL mice show reduced GABAergic gene expression in the reticular thalamic nucleus (TRN) and parietal cortex (ParC).

Levels of Gad1 mRNA in the a TRN and b ParC are reduced in 16p11.2 DEL male mice. Pvalb mRNA expression is unaltered in the d TRN, but increased in the e ParC of 16p11.2 mice of both sexes. Sst mRNA expression is decreased in the g TRN and unaltered in the h ParC of 16p11.2 DEL male mice. Representative autoradiographs for c Gad1, f Pvalb and j Sst mRNA expression in male mice. i Location of regions analysed: TRN (green) and ParC (purple). Scale bar represents 1000 μm. Box plots show median and interquartile range with “Tukey” whiskers. WT F = wild-type female mice; WT M = wild-type male mice; 16p11 DEL F = 16p11.2 deletion female mice; 16p11.2 DEL M = 16p11.2 deletion male mice. Data were analysed by ANOVA. F and p values are provided in Supplementary Table 1. ***P < 0.001 main effect of genotype (ANOVA). ^#P < 0.05, ^##P < 0.01, Tukey’s post-hoc test vs same sex WT group. n = 12 (8 male) mice/genotype. Arrows in c also show ParC location of Gad1 expression.

Fig. 6. 16p11.2 DEL mice show decreased PPI.

a The startle reflex is unaltered but b PPI is significantly reduced in 16p11.2 DEL mice. a Data shown as mean ± s.e.m. b Box plots show median and interquartile range with “Tukey” whiskers. WT = wild-type mice, 16p11.2 DEL = 16p11.2 deletion mice. Data are pooled for sex as no significant sex x genotype interaction was found (ANOVA). PPI was assessed at 3 different prepulse sound intensities –4, 8 and 16 dB above background. n = 16 (8 male) mice/genotype. **P < 0.01, ANOVA main effect of genotype.

Fig. 7. 16p11.2 DEL mice have enhanced performance in the rCPT, when fully trained (stage 6), mainly due to a higher hit rate (HR) in 16p11.2 DEL males.

a Hit Rate (HR), b False Alarm Rate (FAR), c Sensitivity Index (SI), d Perceptual Sensitivity (d’) and e Responsivity Index (RI) were monitored in 16p11.2 DEL mice and WT controls. Box plots show median and interquartile range with “Tukey” whiskers. n = 12 (6 male) mice/genotype. WT F = wild-type female mice, WT M = wild-type male mice, 16p11.2 DEL F = 16p11.2 deletion female mice, 16p11.2 DEL M = 16p11.2 deletion male mice. p values for overall genotype effects are shown (ANOVA). ^##p < 0.01, ^###p < 0.001 Tukey’s post hoc test compared to same sex WT group. **p < 0.01, ***p < 0.001 significant main sex effect.

Fig. 8. 16p11.2 DEL mice have enhanced performance in the rCPT under conditions of increased attentional load (variable stimulus duration).

a Hit rate (HR), b False Alarm Rate (FAR), c Sensitivity Index (SI), d Perceptual Sensitivity (d’) and e Responsivity Index (RI) were monitored in 16p11.2 DEL mice and WT controls under 3 different stimulus durations—1 s (as in the standard conditions at stage 6), Box plots show median and interquartile range with “Tukey” whiskers. 0.6 s and 0.2 s. n = 12 (6 male) mice/genotype. WT = wild-type, 16p11.2 DEL = 16p11.2 deletion mice. Data are pooled for sex. p values for overall genotype effects are as shown; *p < 0.05, **p < 0.01 (a) or p = 0.02 (d), *** p < 0.001, ANOVA; ^###p < 0.001, Tukey’s post hoc test vs same stimulus WT group.