Classic alterations in skeletal muscle phenotype following physical training include improved fatigue resistance, enhanced aerobic capacity and greater insulin sensitivity [9, 23, 25, 35, 36]. The significance of an individual's ability to adapt to exercise training may ultimately influence multiple risk factors important for long term cardiovascular health . In the present study we demonstrate for the first time that ANG genes are substantially modulated in humans following a six-week aerobic training program. Overall, many ECM-gene transcripts were only modulated in subjects that demonstrated a concurrent improvement in aerobic capacity. Our data suggest that activation of ECM genes may help determine the cardiovascular adaptation to aerobic exercise in humans. The present findings also indicate that, when carrying out expression studies of gene transcripts in humans, prior to any interpretation of muscle gene expression responses, attention must be afforded to the presence of physiological adaptation.
Modulation of genes that regulate extracellular matrix remodeling
We still have an incomplete understanding of the endogenous processes that regulate physiological adaptation to aerobic exercise. Vascular growth factors not only regulate tissue blood vessel density, but also enhances the expression of proteins that regulate oxygen levels in tissue . Hence, it is naïve to think about the role of such growth factors only in terms of regulating tissue capillary density. Altered expression of HIF responsive genes (e.g. VEGF or VEGFR1) typically reflects the posttranslational stabilization of HIF1α protein , and consistent with this dogma, HIF mRNA was not significantly upregulated. VEGFR2, considered the major mediator of VEGF-A-related angiogenesis, was significantly upregulated in the HRG only. Furthermore, NP-1, a facilitator of VEGF165 action at VEGFR2, was markedly elevated in the HRG and unchanged in the LRG (Fig. 1). Upregulation of the VEGFR2 coreceptor transcript provides some evidence for greater VEGF activity in the HRG, as enhanced NP-1 gene expression can be mediated by VEGF signaling [39, 40]. The significance of the downregulation of VEGFR1 in the LRG is unclear, but plausibly reflects an (unsuccessful) compensation for the general lack of VEGFR2/NP-1 gene activation in the LRG, as studies indicate VEGFR1 may oppose VEGF signaling via VEGFR2 in some situations .
The present study represents the first characterization of expression levels of the human ANG gene family in response to endurance-exercise training. Recently, the Terjung laboratory examined the impact of exercise on the angiopoietin system in rodents . Alterations in TIE2, ANG1 and ANG2 transcript expression were profiled 2 h post exercise in various muscle groups taken from Sprague-Dawley rats after 1 to 24 days of intense aerobic running . In rodents, activation of TIE2 and ANG2 was, broadly speaking, rather similar across each muscle tissue, especially when one considers that recruitment patterns would differ between the various muscle groups studied. The ANG1 response, however, differed both across muscle groups and when compared with our human data. In contrast with the upregulation of ANG1 seen in humans in the HRG (Fig. 2), in the rat ANG1 expression was slightly downregulated in oxidative muscle groups, while 'fast' gastrocnemius demonstrated only a modest increase in ANG1 expression . It has recently been established that either pre-  or concurrent administration  of ANG1 synergizes with VEGF to promote hindlimb angiogenesis. Thus it makes sense that effective aerobic training might result in stable increases in ANG1 expression (as we have found). Therefore, differences between our human data and the rodent study by Lloyd et al  perhaps reflect differing muscle sampling times. Future human studies should ideally use multiple time points post exercise to clarify these issues. However, care should be taken to verify that the subjects studied are able to demonstrate a measurable aerobic training response otherwise such results may be unreliable.
It has been hypothesized that changes in the ratio between levels of ANG1 and ANG2 is of physiological importance [27, 42] by contributing to the stabilization of the primary endothelial structures . While ANG1's ability to antagonize ANG2 at the TIE2 receptor may be cell-type specific [42, 44, 45] and has yet to be proven to be important in vivo, it is interesting to note that ANG1 and ANG2 appear to have identical binding affinities for the TIE2 receptor, whereas a threefold molar excess of ANG2 is required to antagonize ANG1 activity . If the gene expression patterns observed in the present study translate to greater Tie2 receptor signaling, then our data supports the idea that ANG1 cooperates with other growth factors in vivo to regulate and promote functional angiogenesis. This hypothesis clearly requires further investigation. ANG1 and TGFB signaling facilitate the maturation of VEGF-stimulated collateral vessel growth in adult tissue [29, 30, 46]. Notably, TGFB2 and TGFBR2 were substantially upregulated only in the HRG. This observation again suggests that the ECM-related gene response in the HRG may have allowed for greater tissue remodeling, which would have contributed to the fourfold greater increase in aerobic capacity.
To further examine the idea that the pattern of transcript expression in the HRG contributed to the enhanced aerobic adaptation, we profiled a second set of genes chosen from a gene-array study . For example, upregulation of potent growth factors should be accompanied by upregulation of endogenous regulators, so that physiological control could be maintained. A2M is an ECM protein known to bind and regulate growth factor activity [47, 48]. A2M was substantially upregulated in the HRG only (Fig. 3) suggesting that there was more active growth factor signaling within the ECM of the HRG. Thrombospondins also regulate ECM growth factors, including TGFB activity. In addition, a loss of function polymorphism in THBS4 is strongly associated with premature coronary artery disease . In the present study a more substantial increase in THBS4 mRNA expression was noted in the HRG (Fig. 3). As dysfunction of THBS4 and lack of cardiorespiratory fitness are both risk factors for cardiovascular disease, it is plausible that THBS4 plays a role in the cardioprotective effects of exercise. Further analysis is required to establish if such a relationship exists within a larger human population.