Structural analysis of chemically probed pre-mRNA structure using RADIS

ABSTRACT: Pre-mRNA secondary structure can modulate the regulatory function of intronic sequences by masking or exposing splice-site signals and altering the accessibility of other features. Yet experimentally grounded structural models for human pre-mRNAs have been largely restricted to experimentally grounded structural models for human pre-mRNAs have been largely restricted to splice-site-proximal regions, leaving the deep intronic regions, where ~70% of intronic disease-causing variants reside, almost entirely unmapped. We developed RADIS (Reactivity Analysis of Deep and Intergenic RNA Structure), a strategy that yields strand-resolved chemical probing reactivity profiles across entire intron-rich loci by creating tiled arrays of RNAs that comprehensively span long genomic distances. This approach removes the abundance and locus-specificity constraints that otherwise limit pre-mRNA probing approaches. RADIS recovers known E. coli ribosomal RNA architectures and yields high-correlation in-cell dimethyl sulfate (DMS)-MaP reactivities at representative splice sites (r = 0.87–0.90). RADIS-constrained folding of 109 5′ and 88 3′ splice sites uncovers an inverse relationship between base-pairing across the spliceosome footprint and splice-site strength, and partitions 81 branchpoints into three structural classes. Across 233 full-length Alu elements, sense Alu RNAs are more structured than antisense elements, and both partition into strand- and lineage-dependent structural classes. RADIS provides a generalizable route to experimentally grounded interpretation at intron-rich disease-associated and intergenic loci, complementing sequence-based variant-effect predictors.

ORGANISM(S): Homo sapiens

PROVIDER: GSE341315 | GEO | 2026/09/13

REPOSITORIES: GEO