UC Irvine

Alzheimer’s disease is characterized by plaques and tangles that lead to neurodegeneration and dementia. Clinical trials for AD drugs have a high failure rate and could benefit from better mouse models of late onset AD. Changes in gene expression, alternative splicing and chromatin profiles have been described as indicators of the pathology. The focus of this thesis is to characterize already available models of AD using single-cell and long-read transcriptomics. Chapter 2 is a time course of neurodegeneration in the 3xTg-AD mouse, which is the only mouse AD model that has plaques and tangles, similar to human AD. We use bulk RNA-seq in the hippocampus of 3xTg mice to identify distinct gene modules associated with microglia and oligodendrocytes that increase with aging and pathology. We further investigate the changes in cell populations using single-nucleus RNA-seq of the hippocampus and cortex of 3xTg and 5xFAD mice to detect major changes in astrocytes and oligodendrocytes groups. We recover a common path of astrocyte activation with the 5xFAD mouse and find that 3xTg derived astrocytes seem to be at an earlier stage of activation. In order to investigate the activation of microglia in 3xTG, we also generated a single-cell RNA-seq dataset of microglial cells and found multiple subtypes, including a set of microglia with distinct transcription factor expression profile that is associated with an early increase in Csf1 expression before the full onset of DAM gene expression. Finally, scATAC-seq reveals a set of chromatin accessible areas shared across multiple activation states found in the scRNA-seq that matches glial activation processes. Overall, differences between the main glial groups point to a slower activation process in the 3xTg model when compared to the 5xFAD. Our study contributes to the identification of progressive transcriptional changes of glial cells in a model that has plaques and tangles.Single-cell microfluidic systems are optimized for smaller cell types than most cells in the brain, which are also difficult to dissociate. The Split-seq barcode strategy without any microfluidics and fixation steps before cell labeling allows for multiplexed cells and nuclei to be sequenced at the same time. We use Split-seq in Chapter 3 to sequence the transcriptome of 24,270 nuclei as well as single-cell microglia from the cortex and hippocampus of one 24mo female 3xTg-AD mouse. Comparison of Split-seq cell clusters against clusters from our existing time course study of 3xTg-AD (Chapter 2), we recover all of the main cell types and detect genes that were problematic, such as Gfap in astrocytes. However, nuclei from derived microglia lack the major identifiers of DAM, which were detectable at low levels in single-cells. Sub-clustering of Astrocytes recovers 11 distinct clusters including an activation cluster that overlaps not only with previously identified markers such as Gfap but also novel markers such as Thy1 expression. The Split-seq protocol show promise for scaling up future single-cell transcriptomics studies of AD. AD has been extensively characterized using short-read sequencing. However, most studies focus on gene expression changes and rarely analyze isoform changes. Full-length, high-throughput mRNA sequencing using long-read technologies is the best way to explore transcript isoform diversity, as regular short-reads do not provide enough information about the connectivity between distant exons. We explore in Chapter 4 the transcriptome of the mouse C57BL6/J and 5xFAD cortex and hippocampus at 8 months of age. We recover >90% of genes previously associated with the 5xFAD genotype. We further detect 244 and 471 isoform switches in cortex and hippocampus respectively. We also found 194 genes with TSS switches and 714 for TES switches relevant for the 5xFAD genotype. Genes presenting isoform changes include genes such as Csf2ra, Csf1 and Lamp2. Long-read transcriptome analysis of mouse models of disease can provide additional insights into how isoform switches can alter gene activity during disease progression.