Autism spectrum disorder (ASD) has been often associated to reduced or aberrant functional brain connectivity as measured with resting state functional Magnetic Resonance imaging (rsfMRI) [1]. However, great heterogeneity exists in the expression of these alterations, and little is known on the pathophysiological and genetic etiology of these deficits. Mouse models harbouring mutations with robust ASD-association can be used to probe causal relationships between genetic variations and aberrant connectivity. By tightly controlling motion and physiological parameters we recently demonstrated the possibility of using rsfMRI to reliably map large-scale connectivity networks in the mouse brain [2, 3]. The approach led to the first identification of possible mouse homologues of the human salience and default mode (DMN) networks in this species [2, 4].

Human chromosome 16p11.2 microdeletion, a genetic trait associated with mild intellectual disability, is the most common gene copy number variation in autism, accounting for approximately 0.5–1% of all ASD cases. A mouse line harbouring a chromosomal deletion syntenic to human 16p11.2 has been recently generated [5]. 16p11.2^+/- mice exhibit cognitive impairments and mild social deficits [5]. However, the macroscale substrates of the cognitive and social impairments associated with 16p11.2 microdeletion remain elusive. We here used mouse rsfMRI to probe whether 16p11.2 microdeletion results in macroscale functional connectivity changes. Our approach permits to begin to establish causal relations between connectivity impairments and specific ASD-related genetic etiologies.

All experiments were carried out in accordance with Italian regulations governing animal welfare and protection. rsfMRI timeseries were acquired on N=11 16p11.2^+/- (heterozygous) adult male mice and N=12 control (16p11.2^+/+) littermates as previously described (6). Animals were imaged under shallow halothane anesthesia (0.7%) and artificial ventilation. All experiments were performed on a 7.0 Tesla MRI scanner, using a single-shot EPI sequence with TR/TE 1200/15 ms, flip angle 30°, matrix 100 × 100, field of view 2 × 2 cm2, 24 coronal slices, slice thickness 0.50 mm and a total rsfMRI acquisition time of 6 min. Image preprocessing: images were motion corrected, spatially normalized, smoothed band-pass filtered and regression of motion traces and the mean ventricular signal were applied. Seed based correlation maps were computed using AFNI based on anatomical regions chosen a priori from an anatomical atlas as previously described [6]Seed-correlation mapping in the retrosplenial cortex (Rs), a key hub of the mouse DMN, revealed reduced pre-frontal connectivity in 16p11.2^+/- mice compared to control littermates. The effect was not seed-dependent and was apparent along the whole retrosplenial-cingulate midline of the DMN (Fig. 1). We next probed insular-cingulate rsfMRI connectivity, a network that recapitulates anatomical features of the human salience network [2] (Fig. 2). 16p11.2^+/- mice showed reduced connectivity between insular and anterior cingulate areas. Thalamo-prefrontal connectivity has also been reported to be compromised in ASD [7]. We thus probed connectivity of ventral (infralimbic) PFC regions, as these contain robust reciprocal projection with thalamic areas, and observed foci of prefrontal-thalamic connectivity in 16p11.2+/- (Fig. 3). To investigate the presence of alerations in axonal connectivity we also employed rabies virus retrograde tracing of infralimbic regions [3] (Fig. 4) to which thalamic afferents project. The tracing result revealed the presence of dense afferent neuronal clusters in mid thalamic regions of both genotypes, suggesting that thalamo-cortical dysconnectivity observed does not trivially reflect absence of projecting fibers. Homotopic inter-hemispheric connectivity appeared to be broadly preserved in all the cortical and subcortical regions assessed, including nuclei of the basal ganglia (p>0.10, all ROIs). No genotype-dependent differences in anesthesia sensitivity were detected as seen with a) halothane minimal alveolar concentration recordings (p=0.18), b) mean arterial blood pressure mapping 17 (p=0.32), c) amplitude of cortical BOLD signal fluctuations (p=0.31), thus arguing against a confounding effect of anesthesia on the inter-strain connectivity differences.We show that 16p11.2^+/- mice exhibit long-range functional hypo-connectivity affecting prefrontal integrative hubs of the “default-mode” and “salience” networks. The observed hypo-connectivity recapitulates hallmark neuroimaging findings in ASD and serves as a possible macroscale substrate for the cognitive deficits produced by 16p11.2 microdeletion. Our approach permits to begin to establish causal relations between connectivity impairments and specific ASD-related genetic etiologies and pave the way to the investigation of the neurobiological underpinnings of highly heterogeneous clinical connectivity findings.