Author ORCID Identifier
https://orcid.org/0000-0002-6177-9316
Date Available
12-12-2027
Year of Publication
2025
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
College
Medicine
Department/School/Program
Neuroscience
Faculty
Mark T. W. Ebbert
Faculty
Richard C. Grondin
Abstract
Late-onset Alzheimer’s disease (LOAD) is highly heritable, with genetic variation accounting for an estimated 60–80% of disease risk. LOAD risk genes average 14 annotated isoforms per gene and encode an average of five distinct proteins with potentially divergent functions. Yet RNA isoform expression in LOAD remains understudied, limiting our understanding of how these molecules contribute to disease. Prior studies relied on short-read RNA sequencing (RNA-seq), which struggles to reliably assemble and quantify isoforms, leading researchers to collapse them into a single gene measurement. In contrast, long-read sequencing spans entire RNA molecules, enabling more accurate quantification of annotated and novel isoforms and thus enabling us to better assess isoform-level mechanisms in LOAD.
We generated and analyzed the largest long-read RNA-seq cohort of human brain tissue to date, profiling 115 postmortem prefrontal cortex samples (29 LOAD males, 26 LOAD females, 30 control males, 30 control females). Our analyses uncovered >3,000 previously unannotated isoforms, including hundreds arising from medically relevant genes, and revealed widespread differential isoform expression and usage patterns. Importantly, more than half of the isoform expression quantitative trait loci (eQTL) we identified were missed by gene-level analyses, underscoring the distinct regulatory and functional signals captured only at isoform resolution. Together, these findings establish that isoform diversity is both extensive and biologically consequential in the aging human brain, and that gene-level analyses obscure much of the molecular landscape underlying LOAD. Our work provides a foundation for identifying new molecular targets and biomarkers that could ultimately enable earlier diagnosis and isoform-specific therapeutic strategies in LOAD. These data have been made easy to access and explore through a web application for wide use by the scientific community.
As part of this dissertation, we also developed RNApysoforms, a Python package that delivers interactive, fast rendering visualizations of RNA isoform structure and expression. This tool integrates isoform structure and expression into dynamic, plotly-based graphics, outperforming existing visualization approaches and making complex datasets interpretable. In addition, we performed the first systematic review and meta-analysis of bulk RNA-seq studies in LOAD brains, synthesizing evidence across datasets and identifying common patterns of differential gene expression as well as pathway-level changes in the LOAD brain. Our review highlights the need for rigor and consistency across RNA-seq studies, and establishes isoform-level resolution as an essential lens for uncovering the molecular mechanisms that drive LOAD and related neurodegenerative disorders.
Collectively, this dissertation demonstrates that isoform-level analyses are not only feasible at scale but essential for understanding the molecular basis of LOAD. By combining large-cohort long-read RNA sequencing, novel isoform discovery, isoform-level regulatory mapping, methodological tool development, and systematic evidence synthesis, this work establishes a comprehensive framework for investigating transcriptomic complexity in the human brain. These contributions expand fundamental knowledge of isoform diversity, reveal new LOAD relevant isoform signatures, and provide broadly accessible resources that will guide future studies and accelerate the development of isoform-informed biomarkers and therapeutic strategies.
Digital Object Identifier (DOI)
https://doi.org/10.13023/etd.2025.573
Funding Information
This work was supported by the National Institutes of Health [R35GM138636, R01AG068331, 5R50CA243890, T32DA035200/TL1TR001997] the BrightFocus Foundation [A2020161S], Alzheimer’s Association [2019-AARG-644082], PhRMA Foundation [RSGTMT17] and [2025 PDDS 1322707]; Ed and Ethel Moore Alzheimer’s Disease Research Program of Florida Department of Health [8AZ10 and 9AZ08]; and the Muscular Dystrophy Association.
Recommended Citation
Aguzzoli Heberle, Bernardo, "USING LONG-READ RNA SEQUENCING TO FIND POTENTIAL TARGETS FOR LATE-ONSET ALZHEIMER’S DISEASE TREATMENT AND EARLY DIAGNOSIS" (2025). University of Kentucky Doctoral Dissertations. 853.
https://uknowledge.uky.edu/gradschool_diss/853
