UCLA

RNA transcripts are marked with single-nucleotide variants (SNVs), which can have genetic origins (i.e. arising from DNA), or be deposited co-/post-transcriptionally. These non-genetic SNVs often arise from RNA editing, a regulatory process in which RNA sequences are altered through substitution of nucleotides. Regardless of their origins, the presence and pattern of RNA variants can yield informative biological insights. In this work, we leverage RNA variants identified using high-throughput sequencing data to decipher gene regulatory mechanisms and disease-associated transcripts.We first demonstrate that genetic variants exhibiting allele-specific expression patterns in RNA can be utilized as an indicator for gene regulation mechanism. Gene expression is modulated jointly by transcriptional regulation and messenger RNA stability, yet the latter is often overlooked in studies on genetic variants. Leveraging metabolic labeling data (Bru/BruChase-seq) and a new computational pipeline, RNA tracker, we categorize genes as allele-specific RNA stability (asRS) or allele-specific RNA transcription events (asRT). Extensive characterization of stability-regulating variants revealed notable contribution of these genes to immune system related processes. This work highlights RNA stability as a critical, yet understudied mechanism linking genetic variation and disease.Next, we explored the usage of A-to-I editing as a marker for the age of gene transcripts. Standard transcriptomics captures steady-state abundance, obscuring the kinetics of the RNA lifecycle. We demonstrate that the progressive accumulation of endogenous adenosine-to-inosine editing events serves as a molecular timestamp to estimate RNA age. We present MEMORIA, a computational toolkit that leverages this ‘molecular clock’ to infer absolute decay and synthesis rates from single-timepoint long-read RNA-seq data, obviating the need for metabolic labeling.Finally, we leveraged a large compendium of brain-specific A-to-I editing sites as a basis for identifying double-stranded RNAs (dsRNAs). It has been hypothesized that by altering the structure of dsRNAs, RNA editing suppresses the activity of dsRNA sensors which mediate innate immune response. As such, RNA editing is highly relevant to neurodegenerative disorders such as Alzheimer’s Disease (AD), in which dysregulation of innate immune pathways contributes to disease pathogenesis. This work yields novel insights regarding the landscape surrounding immune dysregulation and RNA editing in AD.