ABSTRACT: Aerobic exercise capacity is a strong predictor of disease and survivability but the utility of exercise intervention is largely dependent on how one’s genome interacts with an exercise-training environment. A newly developed rat model selectively bred for inherited differences in response to aerobic exercise training shows to be a useful resource to sort out the networks of genes responsible for signalling exercise-induced changes that benefit cardiac function. Animals The study consisted of four groups: 1) HRT trained rats (HRT-T), 2) LRT trained rats (LRT-T) 3) HRT untrained sedentary rats (HRT-S) and 4) LRT untrained sedentary rats (LRT-S) (n=5 in each group). Experimental protocols were approved by the Institutional Animal Research Ethics Council, Norway. Sample collection Left ventricle tissue was snap-frozen in liquid nitrogen and stored at -80oC to time of analysis. RNA isolation RNA was isolated from the left ventricle (20 mg) with the mirVana RNA isolation kit (Ambion) according to the manufacturer's instructions. RNA integrity, purity and quantity were assessed by Bioanalyzer (Agilent Technologies, Santa Clara, US) and Nanodrop (NanoDrop Technologies, Baltimore, US). The concentration of total RNA was measured by Nanodrop with ultraviolet spectrophotometry at 260/280 nm. RNA quality was assessed by electrophoresis on Bioanalyzer chips (Agilent Technologies, Santa Clara, US). Only samples with a 260/280-ratio of more than 1.8, and RNA integrity number (RIN) value above 8 was analyzed. Microarray analysis Samples were hybridized to Applied Biosystems Rat Genome Survey chips v.1.0. All samples were labelled at the same time, and hybridised over two days. The biological groups were balanced between the two hybridisation batches and the samples randomly distributed within this format. Samples were analyzed by the Applied Biosystems 1700 Array Expression System. The following threes comparisons were done: A: HRT-S vs. LRT-S B: HRT-S vs. HRT-T C: LRT-S vs. LRT-T Differentially expressed gene analysis The data files from the AB 1700 Chemiluminescent Microarray Analyzer Software were processed using J-Express Pro v.2.8. The signal intensity values were extracted per spot, and all flagged and control spots were filtered out. Before compiled into an expression profile data matrix, all arrays were quantile normalized to be comparable. Genes with at most 20% missing values were allowed in the final dataset. The signal intensities in the dataset were further log transformed, and missing values were replaced using the method LSimpute Adaptive. The search for differentially expressed genes was performed both on a single gene and gene set level. Rank Product was used to look for differentially expressed genes on a gene by gene basis, while Gene Set Enrichment Analysis (GSEA) was used to look for sets of genes sharing common characteristics that were differentially expressed between the classes examined. Gene sets were created using the Panther Biological Processes (241 gene sets) and Panther Molcular Functions (252 gene sets). This information was extracted from the Applied Biosystems Human Annotation File, dated September 30th 2006, with updated information from a Panther search carried out on February 1st 2008. Gene sets smaller than 15 and larger than 500 were excluded from the GSEA. GSEA First the genes were ranked according to Golub score, which t is the default method for GSEA. Next an enrichment score (ES) was calculated for each gene set. For instance, when calculating the ES for the gene set “T-cell mediated immunity”, GSEA starts by looking at the gene ranked on top of the genelist. If this gene is a member of T-cell mediated immunity, a positive score is added to the ES, otherwise a negative score is added. Then the next gene on the gene list is evaluated and the ES is updated. This process is repeated for every gene in the entire gene list. The green line in Figure 1 shows how the value of ES changes when evaluating the genes in the ranked list. The maximum value obtained during this “walk” is used as ES for the gene set. Also worth noting is that the positive score that is added to the enrichment score is weigthed, so that a higher score is added when a gene higher on the ranked list is found to be a member of the gene set than when a gene lower on the ranked list is marked with a hit. Therefore, a high ES means that the gene set is overrepresented towards the top of the ranked list. Gene sets with less than 10 members or more than 500 members were removed from the analysis.
