Date of Award
Doctor of Philosophy (PhD)
Alzheimer’s disease (AD) is a neurodegenerative disorder featuring progressive cognitive and functional deficits. Pathologically, AD is characterized by tau and amyloid β protein deposition in the brain. As the sixth leading cause of death in the U.S., the disease course usually last from 7 to 10 years on average before the consequential death. In 2019 there are estimated 5.8 million Americans living with AD affecting 16 million family members. At certain stage of the disease course, patients with inability of maintaining their daily functioning highly depend on caregivers, primarily family caregivers, that incur estimated 18.4 billion unpaid hours of cares, which is equivalent to 232 billion dollars. These huge economic burdens and inevitable emotional distress on the family and the society would also increase as the number of AD affected population could triple by 2050.
Altered cellular composition is associated with AD progression and decline in cognition, such as neuronal loss and astrocytosis, which is a key feature in neurodegeneration but has often been overlooked in transcriptome research. To explore the cellular composition changes in AD, I developed a deconvolution pipeline for bulk RNA-Seq to account for cell type specific effects in brain tissues. I found that neuronal and astrocyte relative proportions differ between healthy and diseased brains and also among AD cases that carry specific genetic risk variants. Brain carriers of pathogenic mutations in APP, PSEN1, or PSEN2 presented lower neuron and higher astrocyte relative proportions compared to sporadic AD. Similarly, the APOE ε4 allele also showed decreased neuronal and increased astrocyte relative proportions compared to AD non-carriers. In contrast, carriers of variants in TREM2 risk showed a lower degree of neuronal loss compared to matched AD cases in multiple independent studies. These findings suggest that genetic risk factors associated with AD etiology have a specific effect on the cellular composition of AD brains. The digital deconvolution approach provides an enhanced understanding of the fundamental molecular mechanisms underlying neurodegeneration, enabling the analysis of large bulk RNA-sequencing studies for cell composition. It also suggests that correcting for the cellular structure when performing transcriptomic analysis will lead to novel insights of AD.
With deconvolution methods to delineate cell population changes in disease condition, it would help interpret transcriptomics results and reveal transcriptional changes in a cell type specific manner. One application demonstrated in this dissertation work is to use cell type proportion as quantitative trait to identify genetic factors associated with cellular composition changes. I performed cell type QTL analysis and identified a common pathway associated with neuronal protection underlying aging brains in the presence or absence of neurodegenerative disease symptoms. A protective variant of TMEM106B, which was previously identified with a protective effect in FTD, was identified to be associated with neuronal proportion in aging brains, suggesting a common pathway underlying neuronal protection and cognitive reservation in elderly. This extended analysis yield from deconvolution results demonstrated one promising direction of using deconvolution followed by cell type QTL analysis in identifying new genes or pathways underlying neurodegenerative or aging brains.
To understand the complexity of the brain under disease condition, network analysis as a large-scale system-level approach provides unbiased and data-driven view to identify gene-gene interactions altered by disease status. Using network analysis, I replicated and reconfirmed the co-expression pattern between MS4A gene cluster and TREM2 in sporadic AD, from which further evidence was inferred from Bayesian network analysis to show that MS4A4A might be a potential regulator of TREM2 that is validated by in-vitro experiments. In Autosomal Dominant AD (ADAD) cohort, disrupted and acquired genes were identified from PSEN1 mutation carriers. Among these genes, previously identified AD risk genes and pathways were revealed along with novel findings. These results demonstrated the great potential of applying network approach in identifying disease associated genes and the interactions among them.
To conclude the dissertation work from methodological, empirical, and theoretical levels, deconvolution pipeline for bulk RNA-Seq, cell type QTL analysis, and network analysis approaches were applied to understand transcriptome changes underlying disease etiology. From which previous AD related findings were replicated that validated the methods, and novel genes and pathways were identified as potential new therapeutic targets. Based on prior knowledge and empirical evidence observed from this dissertation work, a model is proposed to explain how genetic factors are assembled as a highly interconnected interactome network to affect proteinopathy observed in neurodegenerative disorders, that cause cellular composition changes in the brain, which ultimately leads to cognitive and functional deficits observed in AD patients.
Chair and Committee
Sharlee Climer, Joseph Corbo, Christina Gurnett, Oscar Harari,
Li, Zeran, "Integrative Analysis to Investigate Complex Interaction in Alzheimer’s Disease" (2019). Arts & Sciences Electronic Theses and Dissertations. 1810.