Single-Cell Multiomic Atlas of Human Cortical Development in Down Syndrome - Fetal Cortex Multiome Dataset

Down syndrome, caused by an extra copy of chromosome 21, is the most common genetic form of intellectual disability affecting up to 1 in 700 live births1-3. Yet, it remains unclear how the increased dosage of ~200 protein-coding genes on chromosome 21 affects brain development, particularly in the cerebral cortex—the area central to higher-level cognitive functions4-7. Here we generated a single-cell transcriptome and chromatin accessibility atlas from 30 human fetal cortical samples at mid gestation (10-20 weeks after conception), a critical period of cortical development8. We discovered an early global transcriptional network disruption, subtly altering ~600 genes involved in neural development and function, mostly in excitatory neurons accompanied by a significant reduction in RORB/FOXP1 expressing excitatory neurons. Multimodal network analyses predicted the chromosome 21 transcription factors BACH1, PKNOX1, and GABPA as key regulatory hubs controlling dozens of genes genetically linked to intellectual disability. Antisense-mediated normalization of the increased activity of these transcription factors in stem-cell-derived neural cells in vitro, which partially recapitulated molecular changes in the Down syndrome cortex, rescued expression of several of these predicted intellectual disability-associated targets. Finally, a humanized in vivo model replicates key molecular features of Down syndrome not recapitulated in neural cells in vitro. This resource defines the gene-regulatory landscape of the developing human cortex in Down syndrome, revealing early molecular and cellular signatures and candidate therapeutic targets, along with a human in vivo experimental platform for their preclinical testing. Overall design: We performed combined single-cell transcriptional and chromatin accessibility profiling (10X Genomics Multiome technology) of human brain samples from fetuses with DS and controls (CON, normal diploid karyotype). We acquired 20 CON and 19 DS foetal brain samples from surgical terminations of pregnancy spanning PCW 10 to 20. After excluding two poorly preserved samples, we performed single nucleus RNA/ATAC sequencing. After excluding 3 samples with a large fraction of non-cortical cells and stringent quality controls, we retained 248,998 high-quality cells from 15 CON and 15 DS samples for final analyses.