Modulation of extracellular matrix genes reflects the magnitude of physiological adaptation to aerobic exercise training in humans

BACKGROUND: Regular exercise reduces cardiovascular and metabolic disease partly through improved aerobic fitness. The determinants of exercise-induced gains in aerobic fitness in humans are not known. We have demonstrated that over 500 genes are activated in response to endurance-exercise training, including modulation of muscle extracellular matrix (ECM) genes. Real-time quantitative PCR, which is essential for the characterization of lower abundance genes, was used to examine 15 ECM genes potentially relevant for endurance-exercise adaptation. Twenty-four sedentary male subjects undertook six weeks of high-intensity aerobic cycle training with muscle biopsies being obtained both before and 24 h after training. Subjects were ranked based on improvement in aerobic fitness, and two cohorts were formed (n = 8 per group): the high-responder group (HRG; peak rate of oxygen consumption increased by +0.71 ± 0.1 L min(-1); p < 0.0001) while the low-responder group (LRG; peak rate of oxygen consumption did not change, +0.17 ± 0.1 L min(-1), ns). ECM genes profiled included the angiopoietin 1 and related genes (angiopoietin 2, tyrosine kinase with immunoglobulin-like and EGF-like domains 1 (TIE1) and 2 (TIE2), vascular endothelial growth factor (VEGF) and related receptors (VEGF receptor 1, VEGF receptor 2 and neuropilin-1), thrombospondin-4, α2-macroglobulin and transforming growth factor β2. RESULTS: neuropilin-1 (800%; p < 0.001) and VEGF receptor 2 (300%; p < 0.01) transcript abundance increased only in the HRG, whereas levels of VEGF receptor 1 mRNA actually declined in the LRG (p < 0.05). TIE1 and TIE2 mRNA levels were unaltered in the LRG, whereas transcription levels of both genes were increased by 2.5-fold in the HRG (p < 0.01). Levels of thrombospondin-4 (900%; p < 0.001) and α2-macroglobulin (300%, p < 0.05) mRNA increased substantially in the HRG. In contrast, the amount of transforming growth factor β2 transcript increased only in the HRG (330%; p < 0.01), whereas it remained unchanged in the LRG (-80%). CONCLUSION: We demonstrate for the first time that aerobic training activates angiopoietin 1 and TIE2 genes in human muscle, but only when aerobic capacity adapts to exercise-training. The fourfold-greater increase in aerobic fitness and markedly differing gene expression profile in the HRG indicates that these ECM genes may be critical for physiological adaptation to exercise in humans. In addition, we show that, without careful demonstration of physiological adaptation, conclusions derived from gene expression profiling of human skeletal muscle following exercise may be of limited value. We propose that future studies should (a) investigate the mechanisms that underlie the apparent link between physiological adaptation and gene expression and (b) use the genes profiled in this paper as candidates for population genetic studies.