Spatial Expression of Long Non-Coding RNAs in Human Brains of Alzheimer's Disease

Bernard Ng,Denis R. Avey,Katia de Paiva Lopes,Masashi Fujita,Shinya Tasaki,Ricardo A. Vialle,Artemis Iatrou,Sashini De Tissera,Tien-Hao Chang,Jishu Xu,Himanshu Vyas,Nicola A. Kearns,Vilas Menon,Chris Gaiteri,Philip L. De Jager,David A. Bennett,Yanling Wang
DOI: https://doi.org/10.1101/2024.10.27.620550
2024-10-28
Abstract:Long non-coding RNAs (lncRNAs) are highly versatile in their modes of action and play critical roles in both normal physiological and disease processes. The dysregulation of lncRNA expression has been implicated in aging and many neurological disorders, including Alzheimer's disease (AD). Here, we report a spatial expression atlas of 7,634 lncRNA genes in aged human brains, covering 258,987 microdomains from 78 postmortem brain sections of 21 ROSMAP participants. We detected a greater proportion of lncRNAs expressed specifically in distinct cortical subregions than mRNAs, and most belong to antisense and lincRNA biotypes. We generated 193 gene modules from gene networks of 8 subregions and identified lncRNA-enriched gene modules implicated in multiple AD-relevant biological processes. By cross-referencing published snRNA-seq data, we detected a partially overlapping but independent pattern between sub-region and cell-type specificity of lncRNA expression. We mapped spatially differentially expressed (SDE) lncRNAs and mRNAs in AD brains and observed that SDE lncRNAs encompass more subregion-specific transcripts. Gene set enrichment analysis indicates that AD SDE lncRNAs are involved in epigenetic regulation, chromatin remodeling, RNA metabolism, synaptic signaling, and apoptosis. Particularly relevant to AD therapeutic potentials is an enrichment for HDAC target genes, including OIP5-AS1, a lncRNA extensively studied in cancer and, most recently, in AD. We then applied multivariate fine-mapping to the expression quantitative trait loci (eQTLs) proximal to the OIP5-AS1 loci and identified rs1655558 as a potential genetic driver of OIP5-AS1 expression. Using statistical modeling, we inferred that the interaction between OIP5-AS1 and HDAC proteins, especially HDAC11, is associated with tau tangles in excitatory neurons and plaque burden in microglia. Our study represents a valuable resource of lncRNA spatial expression in the aged human brain, shedding mechanistic insights into their functional roles in AD and neurodegeneration.
Neuroscience
What problem does this paper attempt to address?