UC San Diego

Complex phenotypes are distinguished by the interplay of multiple interacting molecules and pathways. Effective study of these phenotypes requires a comprehensive approach utilizing unbiased, genome-scale datasets, including transcriptomics, proteomics and genome sequencing. These datasets can be computationally analyzed to identify known pathways and processes likely to contribute to the phenotype and can be integrated to produce models and make testable predictions. The latter analyses are aided by incorporation of proteome-wide protein-protein interaction data, which can be represented as a network and which may permit the identification of novel functional modules that contribute to the pathogenesis or physiology of the complex phenotype. I used a systems approach to study two complex phenotypes in animal models of the human conditions. In Chapter 2, I looked at rhesus macaques suffering from SIV encephalopathy (SIVE), a model for human HIV-Associated Dementia (HAD). Previous work using a human microarray had identified significant upregulation of inflammatory molecules in monkeys suffering from SIVE but little significant gene downregulation. I integrated gene expression data obtained using a newly available rhesus- specific microarray with a large human protein-protein interaction network I constructed from multiple sources, and then applied a module-finding algorithm to identify modules that discriminated between control and SIVE monkeys. I identified EGR1, which plays a role in memory and learning, as a candidate gene and further work led to a model linking infection-associated downregulation of EGR1 to the cognition deficits seen in HAD. In Chapters 3, 4, and 5, I investigated hypoxia tolerance in Drosophila melanogaster that had been adapted to 4% O2 over generations of selection at progressively lower oxygen tension. Hypoxia contributes to the morbidity and mortality of several important human diseases, including myocardial infarction and stroke, and plays a role in the chemo- and radio-resistance of solid tumors. Understanding mechanisms of hypoxia tolerance may help design new therapeutic approaches to these diseases. In Chapter 3, I analyzed the genome sequences of control and hypoxia- tolerant flies, and in Chapter 4, I analyzed gene expression data. Polymorphisms and gene expression changes identified in the hypoxia-tolerant flies both pointed to involvement of the Wnt signaling pathways in acquisition of hypoxia tolerance. In Chapter 5, I confirmed Wnt signaling involvement through experimental studies of Wnt pathway gene overexpression and knockdown