Transcriptomic Profiling in Mild Cognitive Impairment and Alzheimer's Disease Using Neuroimaging Endophenotypes

Abstract

Alzheimer’s disease (AD) is a devastating neurodegenerative disease affecting more than 6 million Americans and 50 million people worldwide currently. It is an irreversible neurodegenerative disease which causes decline in memory, cognition, personality, and other functions which eventually lead to death due to complete brain failure. Recently there has been a lot of research that has focused on enabling early intervention and disease prevention in AD which could have a significant impact on this disease, be crucial for life management, assessment of risk for future generations, and assistance in end-of-life preparation. For a late-life complex multifactorial disease, such as AD, where both genetic and environmental factors are involved, integrating multiple layers of genetic, imaging, and other biomarker data is a critical step for therapeutic discovery and building predictive risk assessment tools. The multifactorial nature of AD suggests that multiple therapeutic targets need to be identified and tested together. Hence, we need a systems-level approach to build biomarker profiles which can be used for drug discovery and screening/risk assessment. The research presented in this dissertation focuses on utilizing a systems level approach to identify promising imaging genetics biomarkers that provide insight into dysregulated biological pathways in AD pathogenesis and identify critical mRNA measures that can be investigated further within the scope of novel therapeutics, as well as input variables in predictive models for AD risk, screening, and diagnosis. The overall research goal was the development of systems level, imaging genetics biomarker signatures to serve as tools for risk analysis and therapeutic discovery in AD. The specific outcomes of the analyses were characterization of patterns in gene expression at systems level using neuroimaging endophenotypes, and identification of specific driver genes and genotypic variants, which can inform predictive modeling for diagnosis, risk, and pathogenic profiling in AD.