Cross-Tissue Gene Networks Distinguish Normal from Abnormal Brain Development in ASD Toddlers
Objectives: To identify functional genomic abnormalities underlying neural development and risk signatures in ASD.
Methods: A general population screening approach was used to allow prospective, unbiased recruitment and study of ASD and control (typically developing and mildly language delayed) infants and toddlers from community pediatric clinics. Whole-genome leukocyte expression and MRI-based neuroanatomic measures were analyzed in a sample of 142 males ages 1-4 years. Co-expression and connectivity patterns analyses were performed to identify gene modules and hub-genes associated with variations in neuroanatomic measures. Protein-protein interaction analysis revealed hub-genes physical interactions.
Results: Two modules, cell cycle and protein folding networks, were strongly correlated in control toddlers with brain size, cortical surface area, and cerebral gray and white matter. Genes in these modules were found highly expressed in the human brain during first and second trimesters and with reduced expression after birth (BrainSpan). The same two modules were only weakly correlated with neuroanatomic measures in ASD. In control toddlers, modulation of expression in hub-genes with cell cycle functions accounted for smaller to larger brains. Such hub-genes were lost in ASD, and the new set had minimal impact on regulating neuroanatomy. In ASD toddlers with enlarged brains, none of the top ten normal hub-genes were related to brain size measures, and in ASD toddlers with reduced brains, a different set of cell cycle genes impacted size. Unlike controls, ASD toddlers displayed significant correlations with an abnormal array of different gene networks including cell adhesion, immune/inflammation, and translation. A genomic signature enriched in immune/inflammation and translation genes displayed 77% to 82% classification accuracy.
Conclusions: The functional genomic pathology underlying variation in early brain development and size in ASD involves disruption of cell cycle and protein folding gene networks, which govern neuron number and synapse formation. Genomic pathology in ASD was present across all developmental brain sizes from small to abnormally large. Importantly, genomic pathology underlying small brain size in ASD differed significantly from the one underlying brain enlargement. Cell cycle abnormalities included changes in expression, connectivity and network membership patterns of hub-genes. ASD genomic pathology also involved the abnormal relationship of cell adhesion, immune/inflammation and translation networks to brain development. Previously, we reported cell cycle gene networks are disrupted in prefrontal cortex in postmortem ASD children. Thus, we hypothesize that prenatal disruption of key developmental gene networks in ASD may lead to known defects of abnormal neuron number, brain and body growth, and synaptic development and function. A candidate gene expression signature of risk for autism in infants and toddlers was identified. Knowledge of genomic pathology and relationships to heterogeneity of early brain growth abnormality in ASD will illuminate the various brain bases of different subtypes of ASD and facilitate discovery of genomic early diagnostic leading to earlier treatment as well as development of targeted biotherapeutics. Brain size and genomic defects are important autism endophenotypes.