Testing the excitation/inhibition imbalance hypothesis in a mouse model of the autism spectrum disorder: in vivo neurospectroscopy and molecular evidence for regional phenotypes

Fourteen Nf1 ^+/− mice were used in the experiments; these mice were obtained from animals backcrossed to Taconic C57Bl/6 mice at least 10 times and bred once with 129T2/SvEmsJ before experiments. This breeding procedure was put in place to keep the genetic background of the Nf1 ^+/− mice constant across experiments and studies [17, 18]. Seven- to eight-month-old male and female mice were used. All experiments used littermates as controls (n = 13) and were carried and analyzed with the experimenters blinded to genotype and treatment. Animals were group housed (2–4) on a 12-h light/dark cycle in vivarium at UCLA, CAU, and CNC.IBILI. The experiments were carried out in accordance with the European Union Council Directive (2010/63/EU), the National Regulations, and the Internal Review Board of the University of Coimbra. Experimental protocols at the University of California were approved by the Chancellor’s Animal Research Committee of the University of California, Los Angeles, in accordance with NIH guidelines. All included animals were healthy (discomfort score 0), and all efforts were made to minimize the number of animals used and their suffering.

Acquisitions were performed on a 9.4T MR pre-clinical scanner (Bruker Biospec, Billerica MA) equipped with a standard Bruker crosscoil setup using a volume coil for excitation (86/112 mm of inner/outer diameter) and quadrature surface coil for detection (Bruker Biospin, Ettlingen, Germany). Mice were anesthetized by inhalation of 1.5% isoflurane with 100% O₂ and placed on temperature-controlled animal beds. A tooth bar and head restrain were used to reduce motion artifacts. Physiological parameters were continuously monitored and body temperature was maintained at 37 °C using a mouse heating cover (Bruker Biospin, Ettlingen, Germany). To assure accurate voxel positioning, we first obtained T2-weighted images in axial, coronal, and sagittal planes using a rapid multi-slice acquisition refocusing echoes (RARE) sequence directions with the following parameters: repetition time (TR) = 2500 ms, echo time (TE) = 33 ms, matrix size 256 × 256, field of view 20 × 20 mm², 22 slices, slice thickness = 0.5 mm, RARE factor = 8, and 1 average. First- and second-order shims were adjusted using the fieldmap-based MAPSHIM routine, leading to water linewidths between 14 and 20 Hz. MRS voxels were positioned in three locations: striatum (1.8 × 2.0 × 2.2 mm), hippocampus (1.9 × 1.3 × 1.9 mm), and pre-frontal cortex (2.0 × 1.2 × 2.0 mm). Spectra were acquired using a point-resolved spectroscopy (PRESS) sequence with outer volume suppression (OVS) and VAPOR water suppression [19, 20]. The following parameters were used: TR = 2500 ms, TE = 16,225 ms, number of averages = 720, 3 flip angles = 90°, 142°, 142°, bandwidth = 5000 Hz, number of acquired points = 2048 yielding a spectral resolution of 1.22 Hz/pt. Total acquisition time was 30 min. Before each spectrum, we acquired an unsuppressed water spectrum at the same voxel location (TE = 16,225 ms, TR = 2500 ms, 16 averages, scanning time = 40 s). Data analysis of ¹H-MRS spectra was performed using linear combination modeling LCModel (Stephen Provencher Inc., Toronto, Canada) [21]. Metabolite quantification was performed applying the internal water reference method. Concentrations in millmole units were calculated for metabolites, and results are presented in institutional units (i.u.). Only metabolites with Cramér–Rao bounds < 20% were considered for statistical analysis.

Mice were sacrificed following the ¹H-MRS procedure. Brains from half of the animals (7 Nf1 ^+/− and 7 WT mice) were removed, and the hippocampi, prefrontal cortex, and striata were dissected on ice, to isolate the synaptossomal fraction using Syn-PER Synaptic Protein Extraction Reagent (Thermo Scientific, Pierce Biotechnology, Rockford, USA) according to the manufacturer’s instructions. The protein fraction was then quantified using the BCA method and stored at − 80 °C until further use. The proteins were separated on SDS-polyacrylamide gel electrophoresis and transferred onto polyvinylidine difluoride membrane (PVDF; Milipore, Madrid, Spain). After blocked, membranes were incubated with anti-GABA(A)R α1 (1:1000; ref. AGA-001 from Alomone Labs, Jerusalem, Israel), washed and then incubated with alkaline phosphatase-conjugated secondary antibodies (1:20,000; ref. RPN4301 from Amersham, GE Healthcare Life Science, USA) and visualized using ECF reagent (Amersham) on Typhoon FLA 9000 (GE Healthcare Bio-Science AB, Uppsala, Sweden). Immunoblots were reprobed with β-actin antibody (1:10,000; ref. A5441 from Sigma-Aldrich) to ensure equal sample loading, and densitometric analyses were performed using the NIH ImageJ 1.44 analysis software.

Following imaging analysis, the other half of animals (7 Nf1 ^+/− and 6 WT mice) were sacrificed and their brains were removed, frozen on dry ice, and stored at − 80 °C. Coronal sections (14 μm) were cut on a cryostat (Thermo Shandon Inc.) from the posterior to anterior brain. Before starting the immunohistochemistry procedure, the brain slices were fixed in 4% paraformaldehyde (PFA). Then, immunolabeling was performed for the GABA(A)α1 receptor subunit. Briefly, brain sections were rinsed, blocked, and incubated with anti-GABA(A)α1 (1:2000; ref. AGA-001 from Alomone Labs). Brain slices were then washed and incubated with Alexa Fluor 488 anti-rabbit secondary antibody (1:200; ref. ab150077 from Invitrogen, Inchinnan, Business Park, UK). Sections were again washed and stained with Hoechst 33342 (5 μg/mL; ref. 94403 from Sigma-Aldrich), mounted in Dako fluorescence medium (Dako North America, Carpinteria, USA), and images were recorded using LSM 710 Meta Confocal microscope (Carl Zeiss, Oberkochen, Germany). The quantification of fluorescence intensities of GABA(A)α1 receptor labeling was performed using the NIH Fiji ImageJ 2.0.0 analysis software.

Given the demonstration that NF1-associated learning deficits are rescued by a GABA(A) antagonist [9], and our recent study in humans suggesting postsynaptic GABA changes [15], it was important to verify the impact of NF1 depletion on GABA(A) R levels. We found that NF1 is associated with a profound pattern of modification in GABA(A) R binding that seems to be region specific. Hippocampal receptor content was significantly increased in the synaptosomal compartment (p = 0.0270) (Fig. 2a, b). On the contrary, synaptossomes of Nf1 ^+/− prefrontal cortex showed reduction in GABA(A) R α1 protein levels (p = 0.0012) as compared to WT (Fig. 2a, b). No changes were detected in the striatum (Fig. 2a, b). These results were corroborated by immunohistochemistry studies, as shown in Fig. 3a. Our results showed again a significant, with a quite large effect size, increase of immunoreactivity of GABA(A) R α1 in Nf1 ^+/− hippocampal sub-regions, dentate gyrus (DG), and cornu ammonis 1 (CA1) and 3 (CA3; p = 0.0002), while cortical (p = 0.0286) brain slices exhibited immunoreactivity reduction (Fig. 3a). Furthermore, in the striatum, we observed formation of clusters surrounded by hypodense regions and reduction of GABA(A) R-expressing cells suggesting redistribution of these molecules to these hiperdense clusters (p = 0.0286; Fig. 3a). These results were confirmed by immunoreactivity quantification in all analyzed brain regions (Fig. 3b).