Molecular Phenotypes Associated with Total Cerebral Volume in Boys with Autism Spectrum Disorders

Background: We recently described molecular pathways affected by differential exon usage (DEU) / and differential alternative splicing (DAS) that were common to a number of ASD boys. In this study we investigated the divergent molecular pathways in blood associated with different total cerebral volume (TCV) phenotypes of boys with ASD. Megalencephaly is an endophenotype present in about 11-15% of subjects with ASD that has been associated with regression status. We postulated that the genetic and/or environmental factors that cause differences of TCV would affect DEU/DAS in blood and contribute to our understanding of the molecular differences associated with these ASD subgroups.

Objectives: We aimed to identify differences in predicted DAS/DEU in blood cells of 2-4 year old ASD boys with large TCV (LTCV) and normal TCV (NTCV) compared to age-matched typically developing (TD) boys.

Methods: Subjects were recruited through the Autism Phenome Project (M.I.N.D. Institute).  The study included 20 ASD boys with NTCV (3.0±0.5 years), 10 ASD boys with LTCV (3.1±0.2 years), and 20 TD boys (3.0±0.3 years). Brain MRI defined the subgroups, with the LTCV group having a mean TCV of 1.5 standard deviations greater than the average TCV of matched TD controls. Predicted DAS/DEU was assayed using whole blood on Affymetrix exon arrays. A two-level analysis defined the most reliable set of genes predicted to have DAS/DEU (Partek). First, an Alternative Splicing ANCOVA was performed on Group, with Covariates for both technical (Batch, random effect) and biological (Age, continuous variable) variation. Genes with DAS p<0.05 were considered significant. Second, an exon-level expression ANCOVA on Group, with age and batch as co-variates, was performed. Exons with p<0.005 and |Fold-Change|>1.2 on each Group comparison were considered significant. The genes common to both analyses were considered to be the most reliable because they were predicted to be display DAS and to have significant differences of exon-level expression.

Results: 764 genes are predicted to exhibit DAS/DEU in ASD_NTCV vs TD, 23 of which overlapped with genes implicated in ASD (SFARI database, 369 genes, p of overlap=0.07). The 764 genes were over-represented in Dendritic Cell Maturation, NF-kappaB, Actin Cytoskeleton, and Ephrin A (Axon Guidance). 16 of the 764 genes overlapped with the 211 genes reported by Voineagu et al(Nature, 2011) as having DAS in ASD brain (overlap p=0.01). 212 of the 764 genes pass FDR <0.1 for DAS. A different molecular signature was associated with ASD boys with LTCV when compared to TD controls with 124 genes predicted to exhibit DAS/DEU. They were over represented in 5-amidoimidazole Ribonucleotide Biosynthesis I (Nucleotide Biosynthesis), Palmitate Biosynthesis I (Fatty Acid Biosynthesis), Netrin (Axon Guidance /Nervous System Signaling), Leukocyte Extravasation (Cellular Immune Response), and Tetrahydrofolate Salvage from 5,10-methenyltetrahydrofolate (Folate Biosynthesis) signaling pathways. 14 genes overlapped the two comparisons.

Conclusions: We provide evidence for DAS/DEU in blood associated with different TCV in 2-4 year old boys with ASD. The data suggest that differences of TCV in boys with ASD are associated with specific molecular pathways and a specific pathophysiology. RNA-Seq analysis is underway to validate findings.