High throughput gene expression measurement platforms such as microarrays have enabled system-wide analyses of biological systems. In this thesis work, microarray technology was used to discover how a key blood pressure regulatory nucleus in the brain responds to acute hypertension. It is believed that one class of essential hypertension is caused by imbalances between inhibition and excitation of sympathetic vasomotor neurons (cells in the brain responsible for heart rate modulation and peripheral resistance) favoring sympathetic discharge. The nucleus tractus solitarius (NTS) has been identified as a key player in integrating blood pressure sensory input and controlling heart rate, stroke volume, and peripheral vascular resistance as to regulate blood pressure through a process called the baroreflex. Many membrane receptors have been implicated in NTS baroreflex modulation through one-at-a-time pharmacological studies of specific receptor agonists and antagonists. However, the detailed intracellular molecular mechanisms of baroreflex modulation adaptation leading to essential hypertension are largely unknown. Microarrays can help identify into the molecular mechanisms by simultaneously measuring the aggregate system-wide transcriptomic response of all baroreflex implicated receptors to a hypertensive stimulus. Also, activated receptors can be inferred from the gene expression because the receptors are often endocytosed requiring the cell to replace them. Microarrays were used to study the NTS transcriptomic response to an acute hypertensive stimulus as a first step in uncovering the intracellular molecular mechanisms of long term baroreflex control.
Microarrays have been successfully used in cancer and cell cycle studies where gene expression changes are very large (10 fold). However, gene expression changes in neurons under normal physiologically relevant perturbations (such as acute hypertension) are typically less than 2 fold. The first aspect of this dissertation deals with reengineering microarray experiments to gain better measurement sensitivity. Another related aspect identifies how many replicates are needed to robustly determine small gene expression changes (30%-50% changes). Finally, quantitative real time polymerase chain reaction (qRT-PCR), a microarray result validation platform, was reengineered to improve sensitivity to a similar range (20% changes were detectable).
Ethanol adaptation was used as a case study to determine the number of animal replicates to achieve 85% sensitivity with 20% false discovery rate, tissue source, and number of time points, as well as the applicability of various statistical methods and enrichment analysis methods.
After the appropriate analysis methods and experimental design parameters were derived, acute hypertension gene expression data were collected. The hypertension data was analyzed in the context of previous literature. There is a large existing literature that describes the pharmacological effects of various receptor agonists and antagonists on baroreflex modulation associated with either bradycardic (decreasing heart rate) or tachycardic (increasing heart rate) effects. Other literature as linked these receptors to various Na+, K+, and Ca2+ channel types. Finally a third body of literature has used protein activity blockers to link signaling pathways to receptor activity. These three bodies of literature were curated and assembled into two models: one detailed model describing well-known links between baroreflex implicated receptors and their downstream pathways and channels, and one abstract model linking receptors and channels to canonical pathways. The former model contains more components than the latter but only describes receptor types with detailed downstream signaling. The latter model is less detailed, but gives a broader overview of the components and is skewed towards receptors that have known connections to downstream pathways but the components and interactions in these pathways are not fully described. These models were annotated with the gene expression data. Pathways that showed coordinated regulation in various receptor and pathway components were identified. These data and analyses provide physiological context to the existing pharmacological literature.
This dissertation improves both the neuronal cardiovascular control field, and biological studies in general. The specific pathway components with coordinate regulation described in this dissertation are the focus of followup experiments in an ongoing 3 year NIH grant. Those pathway components will be inhibited with and without hypertensive stimulus to infer their individual roles in the hypertension regulation mechanisms. Further, the new curated models provide a framework for analysis of NTS related gene, protein, channel, and receptor data. The matrix of receptor, channel, and pathway interactions in the NTS is about 50% complete, outlining clear experiments for future researchers. The reengineered microarray and qRT-PCR experimental designs provide the entire field of high-throughput biology the tools to measure gene expression with unprecedented accuracy (typically gene expression changes over 2-fold are investigated). The microarray noise simulator can be used to determine the number of replicates needed to obtain specific false positive and false negative rates under any expression signal range (although this dissertation focuses on changes between 20% and 50%). Finally, the model curation, annotation, and interpretation methods can be extended to other biological systems.