A DARTEL Based Analysis of Post-Natal Brain Response to Prenatal Maternal Inflammation In Early or Late Pregnancy In the Mouse

Background:  Exposure to maternal inflammation during prenatal life is thought to increase the incidence of neurodevelopmental disorders such as autism and schizophrenia. However, direct evidence for a causal link in man is difficult to establish, making animal models a useful tool.  In a well documented approach to this issue, offspring of pregnant mice exposed to the viral analogue PolyIC have been reported to have brain and behavioural phenotypes relevant to autism. However, the whole-brain extent of changes across grey and white matter tissue compartments has yet to be explored in detail. In addition, the evidence to date indicates the impact of early and late pregnancy exposure may be distinct, suggesting that the final phenotype depends upon particular windows of vulnerability in fetal life.

Objectives:  To directly test the effect of early or late prenatal inflammation on post-natal brain morphometry in the mouse model using advanced automated MRI methodology. 

Methods:  The viral analogue PolyIC (5mg/5ml/kg) or saline (control) was administered to pregnant C57/B6 mice via tail vein in early (gestation day 9) or late gestation (day 17) and 12 week old adult offspring scanned in-vivo in a 7 T small animal scanner (maximum gradient 360 mT/m; 70/16 PharmaScan, Bruker Biospoin GmbH, Germany) with 23mm quadrature RF coil. Axial images were acquired in under 1 hour/mouse [T₂ – weighted: Effective TE = 38.71ms, TR = 4614.566ms, No of Average = 6, Rare Factor =8, Acquisition Matrix = 256 × 256, FOV = 25 × 25mm, Slice thickness = 0.25mm, Scan Time = 11m4s]. The protocol was approved by the Committee on Use of Live Animals for Teaching and Research, University of Hong Kong. Final numbers available for analysis were: controls n=8, PolyI:C n=14 (PolyI:C, GD9 n=8, GD17 n=6).  Voxel-wise group differences in volumes across whole brain were analysed using a diffeomorphic framework for registering images, DARTEL with reference to a conventional mouse brain template. Statistical analysis in SPM 2 software used “Single-subject: conditions & covariates” to compare tissue volume between different groups with results thresholded at p<.001, cluster extent 30 voxels.

Results:  Early prenatal exposure to PolyIC lowered volumes in hippocampus, subiculum and striatum brain regions compared to saline exposed controls and increased volumes in mid-brain and lateral ventricles.  Mice exposed later had more extensive volume differences in cerebellum with lower volumes in hippocampus regions and lateral ventricles and greater volumes around the amygdala and mid-brain.

Conclusions:  The timing of prenatal insult determines the post-natal brain phenotype.   Broadly speaking, ventriculomegaly and lower limbic-striatal volume in the early exposed mice appears to echo more closely similar anatomical features in schizophrenia, whereas, cerebellar dysmorphology and smaller lateral ventricles in late exposed mice, have have more relevance to autism.  Thus the time of prenatal brain developmental perturbation may contribute to the phenotypic spectrum observed across related disorders of neurodevelopment.