193 Arachnoid Cysts as Genetically-encoded Harbingers of Neurodevelopmental Pathology. (1st March 2022)

• Record Type: Journal Article
• Authors: Kundishora, Adam; Kiziltug, Emre; Allington, Garrett; Allocco, August A.; Smith, Hannah; Ocken, Jack; Moreno-De-Luca, Andres; DiLuna, Michael; Jin, Sheng C.; Kahle, Kristopher
• Is Part Of: Neurosurgery. Volume 68(2022)Supplement 1
• Page Start: 58
• Page End: 58
• Publication Date: 2022-03-01
• Subjects: Nervous system -- Surgery -- Periodicals; 617.48005
• DOI: 10.1227/NEU.0000000000001880_193
• Languages: English
• ISSNs: 0148-396X

Abstract

Abstract : INTRODUCTION: Arachnoid cysts (ACs) are the most common congenital intracranial lesions. While familial cases implicate genetic factors, their ultimate cause remains unknown. Herein, we describe a multi-omics approach to disease state characterization including traditional whole exome sequencing, integrative genomics with analytics of large bulk-RNA and single cell RNA sequencing (scRNAseq) data, and a novel phenomics technique using natural language processing and Human Phenotypic Ontological (HPO) language clustering. METHODS: Whole exome sequencing was performed on 1560 individuals (520 parent-offspring "trios"). Genes with likely pathogenic mutations were defined as those with >1 de novo variant (DNV) reaching threshold for exome-wide significance (p<2.5 x 10-6) and having high pLI score (>0.9). This high-confidence pathogenic gene set and a secondary probable pathogenic gene set (>1 DNV, regardless of exome-wide significance or pLI score) were analyzed in bulk-RNA and scRNAseq datasets to assess spatiotemporal and cell-type relationships of gene expression. Phenomic analysis was conducted by subjecting patient medical records to a natural-language-processing (NLP) algorithm. Phenotypes were catalogued with the Human Phenotype Ontology (HPO) language tree and mapped with uniform manifold approximation and projection (UMAP) clustering. RESULTS: 7 genes reached our pathogenic threshold, ADNP, ARIDB1, KDM5C, PURA, FOXP1, MAP2K1, and SCN2A. Strikingly, all but SCN2A play critical roles in epigenetic regulation of gene expression in the developing brain. Bulk-RNA datasets reveal that these genes converge in networks essential for chromatin remodeling and mRNA processing. scRNAseq datasets revealed an association with excitatory and inter-neurons known to play a role in developmental delay. HPO clustering revealed four major phenotypic clusters defined by neurodevelopmental delay (NDD), seizures, both or asymptomatic. In a subgroup of traditionally phenotyped patients, there was statistically significant enrichment of neuraxial phenotypes - including NDD - among those harboring DNVs. CONCLUSION: This novel association between ACs and multiple epigenetic regulators of gene expression provides the first insight into intracranial AC pathobiology. HPO clustering reveals clinically relevant subsets of ACs and implies that traditionally symptomatic or asymptomatic ACs may be harbingers of genetically-driven aberrant neurodevelopment.