Human Specific Genomic Changes in Brain Development and Disease

Abstract

The human cerebral cortex is a defining feature of our species, supporting expanded cognition and complex behaviors. How human-specific developmental programs arise, and how disruption of gene-regulatory processes leads to neurodevelopmental disease, remain elusive. Here, I addressed these questions through two complementary projects: one examining how pathogenic variation in an RNA helicase perturbs cortical development, and a second defining how a human-evolved enhancer reshapes gene regulation and neurogenesis.First, I investigated how de novo missense mutations in the RNA-binding helicase DDX3X, an established cause of DDX3X syndrome and autism spectrum disorder, disrupt neural development. While DDX3X loss of function impairs corticogenesis, mechanisms by which clinically diverse missense variants act have remained largely unknown. To model disease variants in a controlled developmental context, I established a lentiviral strategy that introduces GFP-HsDDX3X missense mutations into primary mouse neural progenitors and neurons without overexpression. I combined live-cell imaging of progenitor divisions and neuronal output with transcriptomics, proximity-labeling proteomics, and subcellular localization analyses. Across four recurrent variants spanning mild to severe clinical presentations, clinically severe mutations markedly reduce neurogenic divisions and cause profound neuronal death, whereas mild mutations have modest effects on fate and survival. Severe variants induce transcriptional remodeling enriched for DNA damage and stress-response pathways, show elevated double-strand break markers and p53 activation, and drive accumulation of cytoplasmic DNA:RNA hybrids. Severe mutations also promote stress-granule–like assemblies with distinct compositional and biophysical properties. Together, these data support a model in which severe DDX3X missense mutations derail neurodevelopment by perturbing RNA metabolism, linking DNA:RNA hybrid accumulation and cellular stress responses to neuronal loss. I then examined how human regulatory evolution can influence cortical development by modulating chromatin architecture and neurogenesis. Human Accelerated Regions (HARs) are conserved sequences enriched for human-specific substitutions and often act as neurodevelopmental enhancers, yet mechanistic links between HAR evolution, 3D genome organization, and developmental phenotypes remain limited. Using in silico screening, luciferase assays, and in utero electroporation in mouse embryos, we identified HAR1984 as a human enhancer with species-divergent activity during brain development. I generated CRISPR-edited human and chimpanzee cortical organoids in which HAR1984 orthologs were reciprocally swapped, complemented by a humanized knock-in mouse model. Across systems, human HAR1984 increases expression of ETV5 and TRA2B, expands neural progenitors, and increases neuronal output. Mechanistically, Hs-HAR1984 strengthens species-specific chromatin looping to ETV5 and TRA2B promoters, and a human-gained ETV5 binding site supports a positive feedback loop that amplifies enhancer activity. These findings show how enhancer evolution can couple sequence change to altered genome topology and neurogenic output. Collectively, these studies advance understanding of cortical development at two scales: they define mechanisms by which pathogenic missense variation in an RNA regulator compromises neurogenesis and neuronal survival, and they demonstrate how human-specific non-coding evolution reshapes chromatin architecture to bias progenitor programs and neuronal production. Together, they highlight cortical development as highly sensitive to post-transcriptional regulation and to evolutionary tuning of cis-regulatory and 3D genomic mechanisms that control when and where developmental genes are deployed.