T9 - Combining Autism And Intellectual Disability Exome Data Implicates Disruption Of Neocortical Development In Both Disorders

Abstract

Background Autism spectrum disorder (ASD) and intellectual disability (ID) are known to be highly co-morbid, but whether their genetic risk converges on a common neurodevelopmental process is unknown. Integrative approaches using genetic evidence and brain co-expression data have proven useful in pinpointing specific developmental epochs and, within them, discrete neural circuits where risk for ASD maps. These analyses have shown that the strongest signal occurred in the neocortex, specifically in the prefrontal cortex and primary motor-somatosensory cortex (PFC), during mid-fetal development. However, it is unknown how ASD and ID risk intersect at this critical spatio-temporal window for neurodevelopment. These findings motivate a genetic analysis that contrasts and combines discoveries and leverage genetic evidence to shed light on the neurobiology underlying shared risk. Methods We used a statistical method called TADA (Transmission And De novo Association) to identify likely risk genes in ASD, ID and both disorders in 4,216 ASD and 1,479 ID families, as well as 869 ASD cases and 2,829 ancestry-matched controls. We integrated the genetic scores derived from the TADA analyses in ASD with the gene co-expression data from the BrainSpan mid-fetal PFC. The resulting DAWN network was evaluated to identify tightly integrated, functionally-related gene communities. We also conducted an analysis of single-cell RNA-sequencing data from single cells laser-microdissected from human neocortex specimens at 12 and 13 weeks post-conception (apical progenitors in the ventricular zone, subventricular basal progenitors, and neurons in the cortical plate). Results Using TADA, we defined genes associated with risk for ASD and ID (referred to as tASD, tID to denote TADA-implicated genes). Within the DAWN network built on the TADA scores in ASD and gene co-expression in the mid-fetal PFC, we identified two functionally related gene communities: the ‘chromatin modification’ and the ‘transcription factor’ community. Both communities were significantly enriched for tID genes, with the ‘chromatin modification’ community having the strongest enrichment. Notably, many of the transcription factors connected in the community belong to tightly regulated cascades that control the specification of laminar fate identity in progenitors and neurons. Protein-protein interaction analyses corroborated these conclusions, with two linked modules related to synaptic transmission, and another two related to chromatin modification. Within this critical spatio-temporal nexus for risk, apical (ap) and basal progenitors (bp), and neurons (n) are all strongly expressing for tASD genes, with a steady increase in the enrichment from types ap to bp to n. A milder and more symmetrically distributed enrichment was observed for tID genes. Discussion Our analyses of gene expression of mid-fetal neocortical cells and tissue-level co-expression implicate developmental disruption of cortical projection neurons in the etiology of both ASD and ID. Our analyses suggest that mutations in ASD risk genes prominently affect processes relevant to postmitotic neurons, while ID risk might extend to processes preceding acquisition of neuronal differentiation. Within these developmental frames, chromatin remodeling and transcriptional networks were reaffirmed as prominent pathways in both ASD and ID.