Specific disruption of thalamic inputs to the auditory cortex in schizophrenia models

Abstract

Auditory hallucinations in schizophrenia are alleviated by antipsychotic agents that inhibit D2 dopamine receptors (Drd2s). The defective neural circuits and mechanisms of their sensitivity to antipsychotics are unknown. We identified a specific disruption of synaptic transmission at thalamocortical glutamatergic projections in the auditory cortex in murine models of schizophrenia-associated 22q11 deletion syndrome (22q11DS). This deficit is caused by an aberrant elevation of Drd2 in the thalamus, which renders 22q11DS thalamocortical projections sensitive to antipsychotics and causes a deficient acoustic startle response similar to that observed in schizophrenic patients. Haploinsufficiency of the microRNA-processing gene Dgcr8 is responsible for the Drd2 elevation and hypersensitivity of auditory thalamocortical projections to antipsychotics. This suggests that Dgcr8-microRNA-Drd2-dependent thalamocortical disruption is a pathogenic event underlying schizophrenia-associated psychosis.

Fig. 1. Specific deficit of TC synaptic transmission to the ACx of 22q11DS mice

(A) Map of 22q11DS orthologs in Df(16)1/+ mice. (B) Voltage-clamp recordings in thalamorecipient L3/4 pyramidal neurons. Arrow: electrical stimulation (top). TC slices containing MGv, hippocampus (Hipp), and ACx (bottom). (C–E) Input–output relationships between stimulation intensity and EPSCs at TC (MGv-L3/4) (C), L3/4-L3/4 (D), and L1-L3/4 (E) projections. (F, G) CT EPSC and CA3-CA1 fEPSP recordings (left) and input–output relationship (right). (H) L3/4 pyramidal neuron filled with Fluo-5F and Alexa 594 (left) to visualize dendritic spines (right). Red line represents the line scan. (I) Calcium transients in a dendritic spine in response to a single TC stimulation (arrows) repeated 10 times (0.1 Hz). (J) TC input location on dendritic trees of L3/4 pyramidal neurons (48 [WT]/38 [Df(16)1/+] spines from 17 [WT]/11 [Df(16)1/+] neurons) (left); (0;0), soma coordinates. Average distances from soma to TC inputs (right). (K, L) Calcium transient amplitudes (K) and probabilities (L) in response to 10 single TC stimulations. Numbers of spines tested are shown inside columns. *P < 0.05 (C, L).

Fig. 2. The 22q11DS microdeletion renders TC projections abnormally sensitive to antipsychotics due to increased Drd2s in the MGv

(A) TC EPSCs (top) and access resistance (Ra) (bottom) during repeated TC stimulations (0.05 Hz, 600 µA) before and after haloperidol (1 µM). (B, C) Mean EPSCs normalized to initial baseline at TC (P < 0.001, 16/21 neurons) (B) and L3/4-L3/4 CC projections (P = 0.75, 15/14 neurons) (C) before and after haloperidol. (D) Normalized TC EPSCs before and after L-741,626 (20 nM) and haloperidol (5/12 neurons). Insets: representative EPSC traces. (E) Average mRNA levels of Drd2 normalized to Gapdh in different brain regions. (F) Drd2 in situ hybridization. (G, H) Average Drd2 and DRD2 protein levels normalized to β-actin in mice (G) and MGv from postmortem tissues of healthy human controls and patients with SCZ (H). Numbers of mice or human samples are shown inside columns or parentheses. *P < 0.05 (E, G, H).

Fig. 3. Dgcr8 haploinsufficiency causes abnormal TC sensitivity to antipsychotics and acoustic startle deficit in 22q11DS mice

(A) Mouse strains carrying hemizygous multigene or single-gene deletions within the 22q11DS region. Light gray boxes: mouse strains used. (B) Average haloperidol effect on TC EPSCs (EPSC_haloperidol) in 13 mutant mouse lines and WT littermates normalized to baseline (before haloperidol). EPSC slopes were measured 30 min after drug application. Baselines were established by adjusting stimulation intensities to elicit peak EPSC amplitudes of 100−200 pA. (C) Average maximal acoustic startle response at 120 dB SPL. (D) Mean PPI of acoustic startle before and after haloperidol. (E) Mean Drd2 mRNA levels (normalized to Gapdh) in MGv (left) and cortex (right). (F) Low- (top) and high-magnification (bottom) images of TC slice from Dgcr8^fl/+ mice injected with AAV-hSynapsin-Cre-GFP (left) or ACSF (Sham, right). (G) Mean TC EPSCs before and after haloperidol in Dgcr8^fl/+ mice injected in vivo with AAV-hSynapsin-Cre-GFP or sham (P < 0.05, 8/6 neurons). Insets show representative TC EPSCs. Dotted line: baseline. Numbers of neurons (B) or mice (C, E) shown inside columns or parentheses.*P < 0.05.

Fig. 4. Aberrant Drd2 elevation in MGv neurons is necessary and sufficient to render TC projections sensitive to antipsychotics

(A, B) Drd2 (A, top) and control (B, top) siRNA viruses. In vivo MGv viral injection (A, bottom, left). Mean TC EPSCs before and after haloperidol in mice injected with Drd2 siRNA (A, bottom, right) (P = 0.67, 9/11 neurons) or control siRNA viruses (B, bottom) (P = 0.016, 7/9 neurons). (C) Composition of the Drd2-OE virus (woodchuck hepatitis posttranscriptional regulatory element [WPRE], polyA signal [pA]) (top) and TC slice showing GFP in MGv neurons (bottom). (D) Representative TC EPSCs (left) and mean normalized TC EPSCs (right) before and after haloperidol from WT mice injected with Drd2-OE or sham (P < 0.05, 7/9 neurons). (E, F) Acoustic startle response (E) at 120 dB SPL and PPI (F) in WT mice injected with Drd2-OE or sham (10 mice each) into MGv, *P < 0.05. Dotted lines: baseline; insets: representative TC EPSCs.