Sex-Differential Gene Expression in Human Brain: Implications for Autism Spectrum Disorders

Friday, May 16, 2014: 3:30 PM
Meeting Room A703 - A704 (Marriott Marquis Atlanta)
D. M. Werling1, N. N. Parikshak1,2 and D. H. Geschwind3, (1)Interdepartmental PhD Program in Neuroscience, Brain Research Institute, University of California, Los Angeles, Los Angeles, CA, (2)Program in Neurogenetics, Department of Neurology, David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA, (3)Program in Neurogenetics, Department of Neurology, David Geffen School of Medicine, UCLA, Los Angeles, CA
Background: Autism spectrum disorders (ASDs) show a consistently male-biased prevalence, and while genetic factors are known to influence ASD risk, the mechanism(s) by which genetic variants interact with sex-differential biology to exacerbate risk in males and attenuate risk in females is incompletely understood. Gene expression in the brain provides an intermediate phenotype between causal risk factors and the sex-biased presentation of neuropsychiatric disorders, and so the analysis of sex-differential expression patterns may elucidate key pathways influencing sex-biased risk.

Objectives: We aimed to identify and characterize sex-differential gene expression patterns in neurotypical human brain in order to evaluate whether the male bias in ASD prevalence may be driven by differential expression (DE) of ASD risk genes in male and female brain, or whether sex-differential factors act in downstream or interacting pathways to modulate the effects of genetic variants.

Methods: RNA-sequencing data from the BrainSpan project from post-mortem human cerebral cortex at two developmental stages, prenatal (12-22 post-conception weeks; n=14, 7 females) and adult (19-40 years; n=8, 4 females), was processed and analyzed for sex-differential expression using LIMMA. Genes DE by sex (sex-DE) with a minimum fold difference of 1.2 and p-value<0.005 were then compared with sets of ASD-implicated genes, genes differentially expressed in ASD brain tissue, neural cell type markers, and other functionally annotated gene sets from the literature.

Results: We identify over 800 sex-DE genes from both sex chromosomes and autosomes, with a greater number of genes expressed more highly in male than in female cortex. We observe no enrichment of ASD-implicated risk genes among sex-DE genes from either the prenatal or the adult stage. In contrast, among genes up-regulated in male adult cortex, we do find significant enrichment of an up-regulated, ASD-associated co-expression module (M16) identified in post-mortem ASD brain (Voineagu et al., 2011); this module is enriched in astrocyte and microglia markers. Furthermore, sex-DE genes that are also members of the M16 module show parallel up-regulation, with higher expression levels in males versus females and in ASD cases versus controls. Beyond the M16 module, we find that genes that are significantly DE by sex and DE in ASD show concordant directionality and function: genes up-regulated in neurotypical males and ASD mark astrocyte and other glial function, and genes up-regulated in neurotypical females and down-regulated in ASD mark neuronal function.

Conclusions: Our findings suggest that sex-differential ASD risk may not be driven by sex-differential expression of genes carrying currently identified risk variants, but may instead be related to downstream modulation of the effects of genetic risk variants by neuron-glial interactions.