Abstract
Genetic strategies such as linkage analysis and quantitative trait locus (QTL) mapping have identified a multitude of loci implicated in the pathogenesis of hypertension in the spontaneously hypertensive rat (SHR). While several candidate genetic regions have been identified in the SHR and its control, the Wistar-Kyoto rat (WKY), systematic follow-up of candidate identification with polymorphism discovery has not been widespread. In the current report, we develop a data-mining strategy to identify candidate genes for hypertension in the SHR, and then sequence each gene in the SHR and WKY strains. We integrate blood pressure QTL data, microarray data and data-mining methods. First, we determined the set of genes differentially expressed in SHR and WKY adrenal glands. Next, the chromosomal position of all differentially expressed genes was compared with peak marker position of all reported SHR blood pressure QTLs. We also identified the set of differentially expressed genes with the most extreme fold-change. Finally, the QTL positional candidates and the genes with extreme differential expression were proposed as candidate genes if they had biologically plausible roles in hypertensive pathology. We identified seven candidate genes that merit resequencing (catechol-O-methyltransferase [Comt], chromogranin A [Chga], dopamine beta-hydroxylase [Dbh], electron transferring flavoprotein dehydrogenase [Etfdh], endothelin receptor type B [Ednrb], neuropeptide Y [Npy] and phenylethanolamine-N-methyltransferase [Pnmt]), and then discovered polymorphism in four of these seven candidate genes. Chga is proposed as the strongest candidate for additional functional investigation. Our method for candidate gene identification is portable and can be applied to microarray data from any tissue, in any disease model with a QTL database.