INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

VASCULAR ENDOTHELIAL GROWTH FACTOR (VEGF), a heparin-binding endothelial cell-specific mitogen, is an important mediator of hypoxia-induced angiogenesis. VEGF increases vascular permeability and endothelial cell proliferation and migration, promotes angiogenesis in vivo, and has been demonstrated to play a critical role during embryonic development, wound healing, and tumor angiogenesis (11, 12). Hypoxia and exercise have both been shown to increase VEGF mRNA in the skeletal muscle of humans (18, 20, 28) and animals (7, 13). Similarly, electrically stimulated skeletal muscle contractions have also been shown to increase VEGF gene expression (19, 30). In addition, hypoxia has also been shown to increase the gene expression of two VEGF receptors, flt-1 and flk-1/KDR (17, 34), and, just recently, both receptors have also been found to increase after exercise (15, 16, 27). Taken together, these findings implicate VEGF, and perhaps its receptors, flt-1 and flk-1, as important angiogenic factors regulating skeletal muscle angiogenesis, especially in response to physical activity.

Endurance exercise training, whether performed at altitude (in hypoxia) or at sea level (normoxia), is well known to increase aerobic capacity of skeletal muscle by increasing oxidative enzyme activity, number of mitochondria, and muscle capillarity (6). However, still under considerable debate is whether exercise in a hypoxic environment enhances the aerobic and/or metabolic adaptations to training (1, 3, 5, 9). In a recent report, acute hypoxia combined with exercise augmented the gene response of VEGF, and to a lesser degree transforming growth factor-₁ (TGF-₁), in the gastrocnemius muscle of intact rats (7). These findings are compatible with the hypothesis that hypoxic training (at a given exercise intensity) potentates skeletal muscle angiogenesis. Nevertheless, not all evidence points to enhanced angiogenesis in hypoxia. Banchero (4), in 1987, concluded that simple normothermic hypoxia was not sufficient to increase skeletal muscle capillarity even when the muscle is active. Moreover, we have recently showed that chronic hypoxia (8 wk at 12% O₂) ultimately downregulates VEGF mRNA levels in the gastrocnemius muscle of untrained, i.e., sedentary, rats (27). Furthermore, in the brain, VEGF mRNA and protein levels are also reported to return to baseline values after prolonged exposure to hypoxia (22). Thus, despite the observation the acute hypoxia coupled with exercise augments skeletal muscle gene expression of VEGF and TGF-₁, it is unclear whether training in chronic hypoxia would actually enhance muscle angiogenic gene expression and thereby the capillary response itself.

Accordingly, we sought to evaluate gene expression of VEGF, the VEGF receptors flt-1 and flk-1, and TGF-₁ after endurance training performed under chronic hypoxic (12% O₂) conditions. Although basic fibroblast growth factor (bFGF) has been found not to respond acutely to exercise, we sought to determine whether exercise training might also increase bFGF mRNA levels, since chronic nerve stimulation has been shown to increase skeletal muscle protein levels of the fibroblast growth factors (26). Our training program was designed to isolate the effects of hypoxia by imposing the same absolute training intensity in normoxia (room air) and chronic hypoxia. The rationale was based on the previous finding that exercise-induced VEGF mRNA levels are dependent on the intensity of exercise and that acute hypoxic exercise enhances VEGF and TGF-₁ mRNA levels (7). We hypothesized that 1) training at the same absolute exercise level will produce a greater increase in skeletal muscle capillarity in animals trained in chronic hypoxia compared with normoxia, and thus 2) skeletal muscle angiogenic gene expression of VEGF, its receptors flt-1 and flk-1, TGF-₁, and perhaps bFGF would be greater in animals trained in chronic hypoxia compared with those trained in normoxia.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study was approved by the University of California, San Diego, Animal Subjects Committee and the Loma Linda University Animal Research Committee. Twenty-four female Wistar rats were subjected to an 8-wk training program in either room air (RA-T; n = 12) or chronic hypoxia [CH-T; inspired O₂ fraction (FI_O2) = 0.12; n = 12]. The mean age was 73 ± 5 (SD) days and mean body mass was 218 ± 7 (SD) g at the start of training. To assess the effects of training, we compared the present data with those from 24 untrained (i.e., sedentary) rats also exposed to room air (RA-S; n = 12) or chronic hypoxia (CH-S; n = 12), which we have recently reported (27). The mean age and body mass before the 8-wk exposure period for the untrained animals were 68 ± 2 (SD) days and 191 ± 12 (SD) g, respectively.

Exercise training regime. Training consisted of a 1-h run, 5 days/wk, for a total of 8 wk on a multilane rodent treadmill (model CL-4, Omnitech, Columbus, OH). In a separate study, it had previously been determined that rats could run continuously for 1 h in hypoxia (FI_O2 = 0.12) at 15 m/min on a 10° incline (7). Accordingly, this level of exercise was selected to begin the training program. The treadmill speed was gradually increased each week (keeping the incline fixed at 10°), such that the majority rats could successfully complete the 1-h exercise bout. By the end of the 8-wk period, rats in chronic hypoxia were running at a speed of 18 m/min. Rats trained in room air were subjected to the same absolute training profile (also starting at 15 m/min and increased to 18 m/min in the same fashion throughout the 8-wk training program). To motivate and keep the rats running, a shock grid and air jets were utilized at the rear of the treadmill. A minimum gas flow of 4 l/min (at the designated FI_O2) was flushed through each treadmill lane and was found to be sufficient to minimize increases in treadmill chamber temperature and CO₂ during exercise.

Environmental conditions. When not training, rats assigned to the hypoxic group were housed in a normobaric environmental chamber that electronically controlled and maintained the FI_O2 (at 0.12) and chamber temperature (22-24° C). Fans inside the chamber circulated the air across a soda lime bed, which kept CO₂ levels below 0.5% throughout the 8-wk hypoxic period. Separate O₂ and CO₂ analyzers continuously monitored and recorded O₂ and CO₂ levels, respectively. Animals and cages were visually inspected daily for animal health, normal and abnormal behavior, and food and water consumption. Rats were kept on 12:12-h day-night cycles and provided standard rat food and water. Hypoxic animals were briefly exposed to room air (10-15 min per training day) during transit to and from the treadmill apparatus.

Single acute exercise bout. On completion of the training program, rats were rested for 2 days. Thereafter, six animals were randomly selected within each group (CH-T, n = 6; RA-T, n = 6) to perform a single 1-h exercise bout at 18 m/min, on a 10° incline, in room air. The remaining six animals in each group served as trained resting controls. This strategy allowed us to evaluate the acute gene response to exercise in trained animals, as well as the basal gene expression levels after training.

Surgical procedure and muscle sampling. Immediately (<2 min) after the acute exercise bout, rats were anesthetized using pentobarbital sodium (Nembutal; 50 mg/kg ip). Once anesthetized, the entire left gastrocnemius muscle was surgically removed, weighed, and flash frozen in liquid nitrogen (occurring 20-30 min after the exercise bout). Muscles were stored at 80°C until processed for Northern blot analysis. The remaining gastrocnemius muscle (right side) was perfusion fixed using a 6.25% solution of glutaraldehyde in 0.1 M sodium cacodylate buffer (total osmolarity: 1,100 mosM; pH 7.4), using a whole body technique described by Mathieu-Costello (24). The gastrocnemius muscle was carefully excised and stored at 4°C in a glutaraldehyde solution until processed for morphometric analysis. Animals not selected to perform the final single acute exercise bout (i.e., rest group) underwent the same surgical and vascular perfusion procedure.

RNA isolation and Northern blot analysis. Total cellular RNA was isolated from the whole left medial gastrocnemius muscle by the method of Chomczynski and Sacchi (8). RNA (10 µg) preparations were quantitated by absorbance at 260 nm, and intactness was assessed by ethidium bromide staining of RNA that was separated by electrophoresis using a 6.6% formaldehyde-1% agarose gel. Fractionated RNA was transferred to a Zeta probe membrane (Bio-Rad, Hercules, CA) and cross-linked to the nylon membrane by ultraviolet irradiation and stored at 4°C. Blots were probed with oligolabeled [-³²P]dCTP cDNA probes, which had specific activities of 1 × 10⁹ disintegrations · min¹ · µg DNA¹ (10). The rat VEGF probe is a 0.9-kb cDNA PstI/SmaI insert cloned into pBluescript II KS⁺ vector. The band corresponding to VEGF₁₆₅ isoform (~4.7 kDa) was analyzed. The 1.2-kb rat KDR/flk-1 cDNA EcoRI insert of pUC18 and the 0.6-kb rat flt-1 cDNA EcoRI/HindIII insert of pUC119 were kindly provided by Dr. Masabumi Shibuya (36). The rat TGF-₁ cDNA probe is a 0.985-kb HindIII/XbaI insert cloned into pBluescript II KS⁺ vector. The bFGF probe is a 1-kb XhoI fragment of human bFGF cDNA. Prehybridization and hybridizations were performed in 50% formamide, 10× SSC (20× SSC is 0.3 M sodium chloride and 0.3 M sodium citrate), 5× Denhardt's solution (100× Denhardt's is 2% Ficoll and 2% polyvinylpyrrolidone), 50 mM sodium phosphate (pH 6.5), 1% SDS, and 250 µg/ml sonicated salmon sperm DNA at 42°C. Blots were washed with 2× SSC and 0.1% SDS at room temperature and 0.1× SSC and 0.1% SDS at 65°C for VEGF mRNA, at 60°C for bFGF mRNA, and at 50°C for TGF-₁, flk-1, and flt-1 mRNAs. Blots were exposed to X-OMAT AR-5 X-ray film (Eastman Kodak, New Haven, CT) using a Cronex Lighting Plus screen at 80°C. Autoradiographs were quantitated via a computer densitometry software package (Gel-Pro Analyzer, Media Cybernetics, Silver Spring, MD) and normalized to ribosomal 18S RNA levels.

Morphological processing and analysis. Using anatomic landmarks, a small section of muscle was taken from the deep midbelly portion (composed of mixed fiber-type distribution) within the medial gastrocnemius muscle. Muscle samples were then cut into small longitudinal strips, dehydrated, and embedded into resin blocks. From each animal, four blocks were cut into four (1-µm-thick) transverse sections using a LKB Ultrotome III and stained with 0.1% aqueous toluidine blue solution as previously described (24). Using light microscopy, measurement of fiber cross-sectional area (FCSA), capillary density (CD; capillary number/FCSA), and the number of capillaries around a fiber (N_CAF) were obtained. The ratio of the number of capillaries to the number of fiber (C/F) was calculated by multiplying CD and FCSA.

Data analysis. All values are expressed as means ± SE. Two-way ANOVA and Student's t-tests were used to determine significance. Significance was accepted at the 0.05 confidence level.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Mean body mass before beginning the training program was not significantly different between the room air and chronic hypoxia groups (Table 1). At first glance, it might appear that room air training (RA-T) increased body mass compared with sedentary control (RA-S), but when the initial body mass is taken into account it is evident that the RA-T animals gained less weight compared with RA-S. This was not true for animals that trained in chronic hypoxia, because both CH-S and CH-T groups gained the same amount of weight. However, it was found that both groups (i.e., sedentary and trained) subjected to chronic hypoxia gained less weight than the untrained (i.e., sedentary) room air animals.

Morphological analyses revealed that the N_CAF and the C/F in the gastrocnemius muscle of CH-T animals increased by 15 and 16% (P < 0.05), respectively, when compared with CH-S animals (Table 2). Room air training (which was performed at the same absolute workload as in hypoxia) did not increase N_CAF or C/F. Because of a significantly higher FCSA in the RA-T group, a concomitant, but nonsignificant, decrease in CD was also observed. Neither FCSA nor CD was found to be significantly different between CH-S and CH-T groups.

Figure 1 shows that neither resting nor exercise-induced levels of VEGF mRNA levels were found to be significantly different between RA-S and RA-T groups, where acute exercise produced a 3.7- and 3.4-fold increase in VEGF mRNA levels in RA-S and RA-T rats, respectively. Although bFGF was not found to respond to acute exercise, room air training increased (P < 0.05) both resting and exercise-induced bFGF mRNA levels by 1.6- and 2.3-fold, respectively. In contrast, TGF-₁ mRNA levels increased 1.5- and 1.9-fold after acute exercise (P < 0.05) in both sedentary and trained rats, respectively. This slight increase in TGF-₁ mRNA response to exercise after training was not significantly different compared with the exercise-induced response of TGF-₁ in sedentary animals.

View larger version (55K): [in this window] [in a new window]   Fig. 1.   A: Northern blot analysis revealing the gene response of vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), and transforming growth factor-1 (TGF-1) in the rat gastrocnemius muscle after an 8-wk exercise training regime in room air, i.e., normoxia. B: quantitative Northern blot analysis normalized to 18S rRNA. Values are means ± SE. *Significant response to acute exercise bout, P < 0.05. # Significantly different compared with sedentary rest group, P < 0.05.

Similar to results shown in Fig. 1, neither resting nor exercise-induced levels of VEGF mRNA were found to be significantly altered after chronic hypoxic training (Fig. 2). However, in these animals, acute exercise increased VEGF mRNA levels only 1.8- and 2.3-fold in untrained and trained rats, respectively. In other words, compared with the VEGF mRNA response to exercise in animals kept in room air, chronic hypoxia attenuated the VEGF mRNA response to exercise in both trained and untrained animals (P < 0.05). Furthermore, chronic hypoxic training decreased resting and exercise levels of bFGF mRNA and eliminated the small, but statistically significant, exercise-induced increase in TGF-₁ mRNA observed in both room air-trained and untrained animals (Figs. 1 and 2).

View larger version (59K): [in this window] [in a new window]   Fig. 2.   A: Northern blot analysis revealing the gene response of VEGF, bFGF, and TGF-1 in the rat gastrocnemius muscle after an 8-wk exercise training regime in chronic hypoxia (inspired O2 fraction = 0.12). B: quantitative Northern blot analysis normalized to 18S rRNA. Values are means ± SE. ns, Not significant. *Significant response to acute exercise bout, P < 0.05. # Significantly different compared with sedentary rest group, P < 0.05.

Figures 3 and 4 show that muscle flt-1 and flk-1 mRNA responded divergently to a single acute exercise bout. flt-1 mRNA increased, whereas flk-1 mRNA decreased, in response to exercise. Training in room air attenuated the flt-1 mRNA response to exercise (Fig. 3), whereas chronic hypoxic training abolished the flt-1 mRNA response to exercise (Fig. 4). In contrast, the flk-1 mRNA response to exercise was not significantly altered by room air training (Fig. 3) but tended to decrease further after chronic hypoxic training.

View larger version (63K): [in this window] [in a new window]   Fig. 3.   A: Northern blot analysis revealing the gene response of the VEGF receptors, flt-1 and flk-1 mRNA, after an 8-wk exercise training regime in room air, i.e., normoxia. B: quantitative Northern blot analysis reflecting the change in flt-1 and flk-1 mRNA normalized to 18S rRNA. Values are means ± SE. *Significant response to acute exercise bout, P < 0.05. # Significantly different compared with sedentary rest group, P < 0.05.

View larger version (66K): [in this window] [in a new window]   Fig. 4.   A: Northern blot analysis revealing the gene response of the VEGF receptors, flt-1 and flk-1 mRNA, after an 8-wk exercise training regime in chronic hypoxia (inspired O2 fraction = 0.12). B: quantitative Northern blot analysis reflecting the change in flt-1 and flk-1 mRNA normalized to 18S rRNA. Values are means ± SE. *Significant response to acute exercise bout, P < 0.05. # Significantly different compared with sedentary rest group, P < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The principal finding in this study is that the same absolute training intensity in chronic hypoxia and room air leads to an increase in skeletal muscle capillarity of chronically hypoxic but not room air-maintained animals. This is consistent with the enhanced VEGF mRNA response reported by Breen et al. (7) after acute hypoxic exercise. However, contrary to our hypothesis, skeletal muscle gene expression of VEGF, its receptors flt-1 and flk-1, TGF-1, and bFGF was not elevated but rather decreased after training in chronic hypoxia compared with room air-trained animals. This could argue against the importance of these angiogenic factors in regulating skeletal muscle angiogenesis. However, it is important to note that we measured the gene expression at the end of training program, and, therefore, it is possible that changes in gene expression occurred earlier within the 8-wk training program that are not reflected in the posttraining levels. In addition, whether similar changes are observed at the protein level (via synthesis or degradation) is presently unknown and was not investigated in this study. Furthermore, we have previously shown that exposure to chronic hypoxia itself attenuates gene expression of VEGF and its receptors flt-1 and flk-1 in the skeletal muscle of similar untrained rats (27). Accordingly, the present findings appear to further support the observation that chronic hypoxia, via a mechanism that has yet to be identified, is attenuating skeletal muscle angiogenic gene expression (27) and that this still occurs after a rise in muscle capillarity induced by endurance training under chronic hypoxic conditions. Because we chose to use female Wistar rats to allow comparison with previous studies from our laboratory (15, 16, 27), we cannot exclude the possibility that hormones involved in the periodic female reproductive cycle, e.g., progesterone, may also regulate VEGF gene expression and could have played a role in the gene responses we have reported here. However, at present, we are unaware of any inconsistencies with skeletal muscle VEGF gene expression related to the female reproductive cycle or of any evidence that would suggest this under the experimental conditions we tested.

Lack of VEGF and capillarity changes in room air-trained rats. It is well established that the degree to which skeletal muscle adapts to physical activity is dependent on the duration and intensity of exercise (6) and that maximal O₂ consumption (O_2 max) and the ability to perform work are reduced at altitude (i.e., during hypoxia). Therefore, it is not surprising that training in room air at the same absolute intensity as training in hypoxia did not increase muscle capillarity, whereas hypoxic training did. Indeed, Desplanches et al. (9) have reported that humans training (i.e., cycling) in room air, at the same absolute workload as that performed in hypoxia (FI_O2 at 0.10), did not significantly increase muscle capillarity or even O_2 max, whereas hypoxic training did increase skeletal muscle capillarity and O_2 max. The lack of a capillary response observed in our room air-trained animals can most likely be attributed to exercise intensity, which was too low to elicit a change in muscle capillarity. Because the focus of the study was to determine whether changes in angiogenic gene expression as seen by Breen et al. (7) and Gavin et al. (14) ultimately correspond to a change in vascularity, it was important to use the same absolute intensity during training in room air and chronic hypoxia so as not to change more than one variable at a time. Although there was a small difference between the initial body mass of the sedentary and trained animals (Table 1), which could influence muscle fiber size and capillarization, we based our assessment and interpretation of muscle capillarity on N_CAF and C/F, which takes into account changes in fiber area. Therefore, the small difference in body mass observed is unlikely to significantly impact the interpretation of these data.

Chronic hypoxic training and VEGF gene expression. Contrary to our hypothesis, exercise-induced VEGF mRNA levels were not augmented but attenuated after chronic hypoxic training compared with similar room air-trained animals (Fig. 5). Because the final acute exercise bout (at the end of the training) for both room air- and hypoxic-trained animals was performed in room air (to appropriately compare their gene responses), it is possible that the attenuated VEGF response to exercise could be explained by the greater number of capillaries (i.e., C/F) in the hypoxic-trained animals (Table 2), which may have enhanced O₂ availability and perhaps resulted in a greater intracellular, i.e., muscle, PO₂ level. Similarly, a greater increase in myoglobin concentration associated with hypoxic training (21, 35) might also have resulted in higher muscle PO₂ levels, thus attenuating VEGF gene response to exercise. These hypotheses, however, seem unlikely given that untrained animals exposed to chronic hypoxia, which did not increase skeletal muscle capillarity (Table 2), also demonstrated a similar decrease in VEGF gene expression (Fig. 5). Because long-term exposure to hypoxia is well known to stimulate changes, such as increased alveolar ventilation and increased red blood cell concentration (Hb concentration), both of which can increase the amount of O₂ delivered to the tissues, these mechanisms may also play a role in attenuating the gene expression of VEGF and its receptors in the rat gastrocnemius muscle. However, given that cells (human hepatoma 3B and rat cardiomyocytes) grown in culture, which have also been preconditioned to hypoxia, demonstrate a blunted VEGF gene response to subsequent hypoxic exposures (23), it seems unlikely that the systemic mechanisms involving acclimatization to altitude or chronic hypoxia could explain, or at least fully explain, the attenuated gene response of VEGF and its receptors we observed. Indeed, Levy (23) provides convincing evidence that the blunted VEGF gene response to hypoxia stems from failure to activate VEGF transcriptionally, presumably via the transcriptional binding protein hypoxia inducible factor-1 (HIF-1). Coupled with the observation that chronic hypoxia decreases HIF-1 binding activity in rat neurons (31, 33), it is possible that alterations in HIF-1 ability to activate VEGF may be responsible for the decreases in skeletal muscle gene expression we observed in VEGF, and perhaps for the VEGF receptors as well.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Summary of VEGF mRNA at rest and after a single acute exercise bout in the gastrocnemius muscle of trained and untrained rats (sedentary, i.e., untrained, data were previously published, see Ref. 27) exposed to 8-wk of normoxia (room air; A) or chronic hypoxia (B). Data in A and B are normalized to untrained room air rest group, which was set to equal 1.0. Values are means ± SE.

VEGF receptor responses. In untrained rats, we have previously reported that flt-1 and flk-1 respond divergently to a single acute exercise bout (27). In this study, we report a similar finding in trained rats (Figs. 3 and 4). The divergent response of flt-1 and flk-1 to exercise is consistent with the observation that these receptors are differentially regulated in response to hypoxia (17, 32). On the basis of these data (Figs. 3 and 4), it may be tempting to speculated that flk-1 is not important in skeletal muscle angiogenesis or at least in response to exercise. However, it should be noted that, in our study, muscles were removed less than 1 h after completion of the exercise bout (~20-30 min), a time point within the expected maximal gene responses for VEGF, TGF-₁, and bFGF (7). During the course of our study, Gavin et al. (16) reported that flk-1 mRNA also increases with exercise but that this response occurs ~16 h after the exercise bout. Moreover, in separate report, flt-1 demonstrated a biphasic response to exercise, increasing at 1 and 16-24 h postexercise (15). Therefore, in the case of flk-1, and perhaps also for flt-1, it appears that the postexercise time frame was not sufficient to assess the peak gene response to exercise under our experimental design. Undoubtedly, further studies looking at the gene responses at these later time points will be necessary to fully appreciate the skeletal muscle gene responses of these VEGF receptors.

Gene expression of bFGF and TGF-₁ in response to training. Thus far, the roles of bFGF and TGF-₁ in exercise-induced skeletal muscle angiogenesis remain largely undefined. Whereas bFGF mRNA has been shown not to respond to acute exercise, we found that room air training produced a small, but significant, rise in skeletal muscle bFGF mRNA levels (Fig. 1). This is consistent with the observation that chronic electrical stimulation (>3 wk) has been shown to increase skeletal muscle bFGF protein levels (26). Therefore, although not found to respond to acute exercise, bFGF may play a role in the long-term response to exercise (e.g., training). This may be particularly important because bFGF has been shown to act synergistically with VEGF (2) and therefore may also play an indirect role in promoting exercise-induced skeletal muscle angiogenesis. However, in view of the fact that bFGF mRNA levels were found to increase only after room air training (where an increase in skeletal muscle capillarity was not observed) and decreased after hypoxic training (where an increase in skeletal muscle capillarity was observed, Table 2), it is unclear what role, if any, bFGF may play in exercise-induced angiogenesis.