Speaker
Description
Genomic studies have identified an ever-expanding set of de novo gene mutations in Epilepsy Risk Genes (ERGs) conferring high risk for neurodevelopmental epilepsy. ERGs are known to converge on synaptic and ion channel pathways in neurons, but we currently lack a systematic annotation of their functional genomic features.
To address this, we collated twelve diverse multiomic data sets spanning diverse spatiotemporal scales of the human brain. Ten were whole-neurogenomic transcriptomic studies, capturing expression patterns across brain regions, embryonic to adult development, cell types, and subcellular structures; one capturing reported direct protein-protein interactions between proteins coded for by the genes; one with metrics of intolerance to loss-of-function mutations. We included 18,450 protein-coding genes which were present in at least 6 of the 12 datasets. 557 (3%) are identified as ERGs because single-nucleotide variants have been linked to epilepsy risk.
Comparison of ERGs with null gene sets recovered their relative intolerance to loss of function mutations and their enrichment in gene sets marking inhibitory neurons, but also revealed several other novel associations. For example, regional expression is highest in the frontal and temporal cortices, especially in neonatal samples. At a cellular level alongside GABA-ergic interneurons, expression was high in intratelencephalic excitatory projection neurons in cortical layers 2 to 4. Subcellularly, ERG proteins mostly localised to the cellular membrane and mitochondria, consistent with previous protein function studies. At the scale of protein-protein interactions, ERGs have significantly higher degree and are more commonly connected with other ERGs, particularly SCN1A, SCN2A, CACNA1A, but with notable neuro-related non-ERGs (e.g. CREB1, CAMK2A, TH).
This work provides an integrative view of the genomic features that distinguish ERGs - pointing towards a multiscale signature threading through cortical regions, layers, cell-types compartments and biological pathways, informing mechanistic models and setting the stage for deeper predictive modelling.
| Author(s) | Jack Highton* , Mathilde Ripart, Sophie Adler, Armin Raznahan, Konrad Wagstyl |
|---|---|
| Affiliation(s) | "School of Biomedical Engineering & Imaging Sciences, King's College London, London, United Kingdom., School of Biomedical Engineering & Imaging Sciences, King's College London, London, United Kingdom., UCL Great Ormond Street Institute of Child Health, London, United Kingdom., Section on Developmental Neurogenomics, National Institute of Mental Health, National Institutes of Health, Bethesda, MD, United States., School of Biomedical Engineering & Imaging Sciences, King's College London, London, United Kingdom.," |
