LETTER

Complementary studies of exercised-induced angiogenic growth factors in human skeletal muscle

With regard to the science contained within the two papers, it is clear that there are many similarities but also some differences. Both studies focus on the angiogenic response of human skeletal muscle to exercise in terms of VEGF and basic fibroblast growth factor (bFGF), with the Gustafsson paper (1) including hypoxia-inducible factor (HIF) subunits. Both studies utilized the single leg knee-extensor exercise model but with apparently different motives. Gustafsson et al. (1) chose this model for its easy combination with femoral venous blood samples (PO₂ and lactate) directly from the exercising muscle, whereas we wanted an isolated small muscle mass to allow myoglobin saturation measurements, rest and exercised samples to be taken on the same subject on the same day, and allow a repeated measures experimental design. Both studies collected muscle samples following the exercise bout: Gustafsson et al. (1) at 30 min and our study at an average of 60 min. Although both studies were interested in the effect of varying oxygen availability, the two studies achieved this in two very different fashions: Gustafsson et al. utilized external pressure that resulted in ischemia, whereas we used alterations in inspired oxygen concentration and no restriction to blood flow, thus making the two studies complementary. The assessment of this treatment at the tissue level was also very different between the two studies, with Gustafsson et al. using indirect measures of cellular oxygenation state (venous oxygen saturation, PO₂, venous lactate), whereas we directly measured intracellular myoglobin saturation to determine intracellular PO₂. At this point it should be recognized that there may be significant differences in the physiology associated with ischemia and hypoxia (3, 4), but oxygen delivery may play a fundamental role in both treatments (2).

The scientific conclusions of both of these studies is that VEGF mRNA is upregulated in human muscle in response to exercise, whereas bFGF appears not be affected. This VEGF upregulation is closely correlated to the change in both HIF subunits. However, the level of VEGF mRNA is not obviously related to alterations in oxygen availability during exercise, but measurements of intracellular PO₂ suggest a possible "threshold" of intracellular PO₂ (easily achieved in normoxia or with no flow restriction), beyond which no greater angiogenic stimulus is produced. Even the correlation found between femoral venous lactate production, as reported by Gustafsson et al. (1), should be interpreted with caution as a link to increasing intracellular hypoxia because we have previously documented that these two variables are unrelated (5).

In conclusion, the failure to cite Gustafsson et al. (1) in our paper (6) is regretted but was the consequence of the complicated review history of our paper, which was originally submitted before Gustafsson and colleagues' work was published. We would like to suggest that since the two papers arrived at similar conclusions using quite different protocols, they should be viewed as providing complementary information on angiogenic growth factor gene regulation as the result of exercise.