Ketamine Restores Thalamic-Prefrontal Cortex Functional Connectivity in a Mouse Model of Neurodevelopmental Disorder-Associated 2p16.3 Deletion

Abstract

2p16.3 deletions, involving heterozygous NEUREXIN1 (NRXN1) deletion, dramatically increase the risk of developing neurodevelopmental disorders, including autism and schizophrenia. We have little understanding of how NRXN1 heterozygosity increases the risk of developing these disorders, particularly in terms of the impact on brain and neurotransmitter system function and brain network connectivity. Thus, here we characterize cerebral metabolism and functional brain network connectivity in Nrxn1α heterozygous mice (Nrxn1α+/- mice), and assess the impact of ketamine and dextro-amphetamine on cerebral metabolism in these animals. We show that heterozygous Nrxn1α deletion alters cerebral metabolism in neural systems implicated in autism and schizophrenia including the thalamus, mesolimbic system, and select cortical regions. Nrxn1α heterozygosity also reduces the efficiency of functional brain networks, through lost thalamic "rich club" and prefrontal cortex (PFC) hub connectivity and through reduced thalamic-PFC and thalamic "rich club" regional interconnectivity. Subanesthetic ketamine administration normalizes the thalamic hypermetabolism and partially normalizes thalamic disconnectivity present in Nrxn1α+/- mice, while cerebral metabolic responses to dextro-amphetamine are unaltered. The data provide new insight into the systems-level impact of heterozygous Nrxn1α deletion and how this increases the risk of developing neurodevelopmental disorders. The data also suggest that the thalamic dysfunction induced by heterozygous Nrxn1α deletion may be NMDA receptor-dependent.

Keywords: NMDA receptor; autism; functional brain imaging; graph theory; schizophrenia.

Figure 1

Constitutive cerebral metabolism is altered in the thalamus, mesolimbic system, cortex, and amygdala in Nrxn1α^+/− mice. Nrxn1α^+/− mice show hypermetabolism in the (A–C) thalamus and (D) ventral tegmental area with a contrasting hypometabolism in (E, F) primary sensory processing cortices, (G) entorhinal cortex, and (H) central amygdala. Data shown as mean ± SEM. ⁺P < 0.05 main effect of genotype, ANOVA. ^*P < 0.05, pairwise t-test with Benjamini–Hochberg correction. VL, VM, and Re data from saline-treated animals only. VTA, SSCTX, AudC, EC, and CeA data are for the genotype effect across all treatment groups.

Figure 2

Nrxn1α heterozygosity alters functional brain network structure and inter-regional functional connectivity. (A) Average path length is significantly increased (L_p, P = 0.041) in functional AQ9 brain networks of Nrxn1α^+/− mice, while (B) mean degree (<k>, P = 0.589) and (C) the global clustering coefficient (C_p, P = 0.277) are not altered. Inter-regional functional connectivity alterations in Nrxn1α^+/− mice support reduced thalamic “rich club,” thalamic-PFC and abnormal septum/DB and raphé-PFC connectivity. (D) Heatmap showing significantly lost (blue) and abnormal/gained (red) inter-regional connectivity in Nrxn1α^+/− relative to WT mice, determined by comparison of the VIP statistic (t-test with Bonferroni correction) calculated through PLSR analysis. (E) Brain images showing the anatomical localization of altered inter-regional connectivity for the anterior prelimbic cortex (aPrL), dorsal reticular thalamus (dRT), and nucleus reuniens (Re) “seed” regions (yellow). Blue denotes functional connectivity present in WT mice (VIP 95% CI > 1.0) that is significantly lost in Nrxn1α^+/− mice (VIP 95% CI < 1.0, and P < 0.05 t-test with Bonferroni correction). 5-HT, serotonergic system; Amg, amygdala; Aud, auditory system; BG, basal ganglia; Hipp, hippocampus; Meso, mesolimbic system; PFC, prefrontal cortex; Sept/DB: septum/diagonal band of Broca. Brain images adapted from the Allen mouse brain atlas (mouse.brain-map.org/static/atlas).

Figure 3

Subanesthetic ketamine administration normalizes thalamic metabolism and restores thalamic hub and reticular thalamus–nucleus reuniens–prefrontal cortex (RT–Re–PFC) circuit connectivity in Nrxn1α^+/− mice. (A–D) Subanesthetic ketamine administration normalizes thalamic hyperactivity in Nrxn1α^+/− mice. (E) Nrxn1α^+/− mice show an enhanced cerebral metabolic response to ketamine in the nucleus accumbens core (NacC). (F, G) The impact of ketamine on PFC and hippocampal function is not altered in Nrxn1α^+/− mice. Data shown as mean ± SEM. Nrxn1α Hz = Nrxn1α^+/− mice. ^*P < 0.05, ^**P < 0.01 and ^***P < 0.001 ketamine effect within genotype and ⁺⁺P < 0.01, ⁺⁺⁺P < 0.001 genotype effect within treatment group (t-test with BH correction). ^##P < 0.01, ^###P < 0.001 ketamine effect (ANOVA) in regions where no significant genotype × treatment interaction found. (H) Summary connectivity map showing the functional connections of thalamic hub regions lost in Nrxn1α^+/− mice that are restored by ketamine administration. Black shading denotes lost connectivity in saline-treated Nrxn1α^+/− mice that is restored to the level seen in wild-type mice in ketamine-treated Nrxn1α^+/− mice. (I) Summary diagram of RT–Re–PFC circuit connectivity restored in Nrxn1α^+/− mice by subanesthetic ketamine administration. aPrL, anterior prelimbic cortex; CM, centromedial thalamus; dRT, dorsal reticular thalamus; MD, mediodorsal thalamus; Re, nucleus reuniens; VHCA2, ventral hippocampus CA2; VO, ventral orbital cortex; vRT, ventral reticular thalamus; SNC, substantia nigra pars compacta. Representative autoradiograms are shown in Supplementary Fig. 1.

Figure 4

Cerebral metabolic responses to d-amphetamine are not altered in Nrxn1α^+/− mice. Data shown as mean ± SEM of the ¹⁴C-2-DG uptake ratio. d-Amphetamine (d-Amph) administration induces (A) medial orbital cortex hypometabolism, (B–E) thalamic hypermetabolism, and (F, G) amygdala and (H, I) hippocampal hypometabolism. We found no evidence, in any brain region where d-amphetamine modified cerebral metabolism, that the response was altered in Nrxn1α^+/− mice. Data shown as mean ± SEM. ^#P < 0.05, ^##P < 0.01 and ^###P < 0.001 significant effect of d-amphetamine (ANOVA).