The present study investigated the effect of an acute exercise bout on the mRNA response of vascular endothelial growth factor (VEGF) splice variants in untrained and trained human skeletal muscle. Seven habitually active young men performed one-legged knee-extensor exercise training at an intensity corresponding to ∼70% of the maximal workload in an incremental test five times/week for 4 wk. Biopsies were obtained from the vastus lateralis muscle of the trained and untrained leg 40 h after the last training session. The subjects then performed 3 h of two-legged knee-extensor exercise, and biopsies were obtained from both legs after 0, 2, 6, and 24 h of recovery. Real-time PCR was used to examine the expression of VEGF mRNA containing exon 1 and 2 (all VEGF isoforms), exon 6 or exon 7, and VEGF165 mRNA. Acute exercise induced an increase (P < 0.05) in total VEGF mRNA levels as well as VEGF165 and VEGF splice variants containing exon 7 at 0, 2, and 6 h of recovery. The increase in VEGF mRNA was higher in the untrained than in the trained leg (P < 0.05). The results suggest that in human skeletal muscle, acute exercise increases total VEGF mRNA, an increase that appears to be explained mainly by an increase in VEGF165 mRNA. Furthermore, 4 wk of training attenuated the exercise-induced response in skeletal muscle VEGF165 mRNA.
- vascular endothelial growth factor165
- transient transcriptional activation
vascular endothelial growth factor (VEGF) is a potent angiogenic factor, inducing proliferation of endothelial cells (18) and the formation of a vascular network (33). VEGF has been proposed to be of importance for angiogenesis in skeletal muscle (2, 4) and has been shown to be upregulated by a single bout of dynamic exercise (13, 28).
The human gene for VEGF resides on chromosome 6p21.3 (31), and the coding region spans ∼14 kb and contains eight exons (15, 30). Alternative splicing of a single pre-mRNA generates at least five homodimeric isoforms. The monomers consist of 121, 145, 165, 189, and 206 amino acids (VEGF121, VEGF145, VEGF165, VEGF189, and VEGF206), respectively (20). The VEGF isoforms differ by the presence or absence of sequences encoded by exons 6 and 7, whereas the amino acids encoded by exons 1–5 and 8 are conserved in all isoforms (30). Exon 6 and 7 encode two distinct heparin-binding domains, which influence solubility and receptor binding. VEGF165 has a moderate affinity for heparin, owing to the presence of 15 basic amino acids within the 44 residues encoded by exon 7 (9), and VEGF165 is accordingly also moderately diffusible. VEGF121 lacks the heparin-binding domains encoded by exon 6 and 7 and is thus highly diffusible. Isoforms containing the domain encoded by exon 6 (VEGF145, VEGF189, and VEGF206) and in addition exon 7 (VEGF189 and VEGF206) are bound tightly to heparin-containing proteoglycans on the cell surface and in the extracellular matrix (8, 27). In most tissues, VEGF165 and to some extent VEGF121 are found to be the most angiogenic isoforms (15, 19).
The physiological importance of the different isoforms remains unclear. The heparin-binding affinity and binding of VEGF to its receptors are likely to be important events defining the characteristics and thus the physiological roles of the various isoforms (21). Thus distinguishing between expression of splice variants with exon 6 or 7 provides a unique opportunity to examine regulation of the synthesis of distinct isoforms with different biological properties. As different isoforms of VEGF possess different properties, it might be expected that acute exercise and exercise training would regulate the expression of the different isoforms in different ways. To our knowledge only three studies, one in humans (34), one in rats (16), and one in mice (22), have examined the presence of VEGF splice variants in skeletal muscle. All three studies found that VEGF165 (VEGF164 in rat/mouse) is the most abundant and VEGF121 (VEGF120) the least. However, no previous human studies have examined the effect of exercise and training on the expression of the VEGF isoforms. As VEGF165 has been proposed to be one of the most important isoforms for angiogenesis, we hypothesized that exercise would mainly affect the expression of this isoform.
Studies in human skeletal muscle have shown that acute exercise increases VEGF mRNA levels (13, 28). When studying the effect of training on mRNA expression, it is therefore important to know the time course of changes in VEGF mRNA levels after acute exercise so that the results reflect training status rather than the last exercise bout. Recently studies in human skeletal muscle (10, 14) have revealed that the time course of the total VEGF mRNA response to acute exercise is different from what has been found in rats with VEGF mRNA levels responding faster and returning faster to baseline levels in rats (4). The time course of response for the various isoforms of VEGF after an acute exercise bout has not previously been assessed.
Earlier studies that have examined total VEGF have shown that 10 days of one-legged knee-extension exercise increases resting VEGF mRNA levels (12), whereas a training period of 8 wk provoked no changes in resting VEGF mRNA expression (29). Results from these studies suggest an adaptation to exercise that may reduce angiogenic signaling somewhere between 10 days and 8 wk of exercise training. This is supported by findings demonstrating an attenuated exercise-induced increase in total VEGF mRNA in muscle (29) in a trained compared with an untrained leg. In the present study, 4 wk of training was selected as we have observed that this training duration, albeit performed at a higher intensity, is sufficient to induce increased capillarization (17). We hypothesized that there would be no differences in the basal mRNA levels of the VEGF splice variants between the trained and untrained leg but that the acute regulation of the VEGF mRNA isoforms in response to exercise would be different between the two legs, reflecting a difference in training status.
The aim of the present study was therefore to investigate the effect of a single exercise bout as well as exercise training on the mRNA expression of the VEGF splice variants in human skeletal muscle.
MATERIALS AND METHODS
Seven healthy men between 19 and 24 yr of age with a mean height and weight of 184 ± 3 cm and 84 ± 7 kg, respectively, participated in the study. The subjects did not participate in regular physical activities. Before the study, the experimental procedure was explained to all subjects, and they were informed that they could withdraw from the experiment at any time. The subjects gave their informed, written consent before participation in the experiment. The study conformed with the guidelines laid down in the Declaration of Helsinki and was approved by the Ethical Committee of Copenhagen and Frederiksberg, Denmark and the Human Investigations Committee, Yale University.
The training consisted of one-legged knee extensor exercise, using a one-legged knee extension ergometer (1) 5 days/wk for 4 wk. The leg to be trained was randomly chosen. Before the training period, an incremental test was conducted with each leg separately to determine the maximal workload. Resistance was increased every 2 min until a kicking frequency of 60 extensions/min could no longer be maintained. The highest workload that the subjects could maintain for 2 min was set as the maximum workload (24). The first training session was performed at 70% of the maximal workload determined by the incremental test, and the session was performed until exhaustion was reached (∼1 h). As performance improved during the training period, the resistance was increased to induce exhaustion after ∼1 h of exercise.
Two tests of different intensities were performed to evaluate the effect of training on performance: test 1 was performed at an intensity similar to that performed during training, and test 2 was performed at an intensity ∼110% of the maximal workload determined by the incremental test. Both tests were expected to reflect the aerobic capacity of the muscle, and an improved performance in the high-intensity test would suggest that training also had lead to an improvement at higher work intensities. Test 1 examined time to exhaustion of the trained leg at the end of the training period at the workload that the subject could maintain for 1 h before the training period. Test 2 compared the time to exhaustion between the two legs in an all-out bout set to ∼110% of the maximal workload of the untrained leg that had been determined by the incremental test. This test was performed half way through the training period for the untrained leg and after the training period and at least 4 days before the acute exercise bout for the trained leg.
The day before the experiment, the subjects received a standardized dinner and evening snack and a light breakfast on the morning of the experimental day. About 2.5 h after breakfast, muscle biopsies were obtained from each leg from the middle portion of the vastus lateralis muscle using the percutaneous needle biopsy technique with suction (3). The acute exercise bout consisted of 3 h of two-legged knee extensor exercise. The absolute workload was the same for the two legs and corresponded for each leg to 50% of the maximal workload reached with the untrained leg in the one-legged incremental test. Force development was continuously recorded to verify that the two legs performed similar amounts of work. Biopsies were obtained from both legs 0, 2, 6, and 24 h after the end of exercise. Food intake was restricted throughout the experiment until after the 6-h biopsy after which the subjects were provided with standardized meals, including dinner and breakfast the next morning (2.5 h before the 24-h biopsy). Muscle biopsies were immediately frozen in liquid nitrogen and stored at −80°C for later analyses.
Design of primers and probe.
Forward and reverse primers and the TaqMan probe were designed from human specific databases (Entrez-NIH and Ensembl, Sanger Institute), using computer software (Primer Express, Applied Biosystems) as given in Table 1. A VEGF probe was designed to span the boundary of exon 1 and 2 [the forward primer (FP) was located on exon 1 and the reverse primer (RP) on exon 2] and amplify all VEGF isoforms because all VEGF isoforms share exon 1 and 2. A probe, FP, and RP located within exon 7 were designed to amplify all isoforms except VEGF121 and VEGF145. A probe, FP, and RP located within exon 6 would not lead to amplification of VEGF121 or VEGF165. Furthermore, a probe spanning exon 4 and 5 and the RP spanning the boundary of exon 7 and 5 were designed to amplify isoform VEGF165. The amplified products will be named VEGFex7 (VEGF165, VEGF189, and VEGF206) and VEGFex6 (VEGF145, VEGF189, and VEGF206) to differ between isoforms encoded by different exons and total VEGF when probe and primers were located on exon 1 and 2. For each of the genes, a Blast Search revealed that sequence homology was obtained only for the target gene. The TaqMan probes were labeled with 6-carboxyfluorescein at the 5′-end and 6-carboxy-N,N,N′,N′-tetramethylrhodamine (TAMRA) at the 3′-end. Primer and probe optimization was conducted by determination of optimal primer and probe concentrations, and verification of the efficiency of the amplification was conducted. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA was amplified using a 5′-VIC- and a 3′-TAMRA-labeled predeveloped assay reagent (Applied Biosystems). Validation of the different PCR product sizes was performed by electrophoresis (ethidium bromide containing 2.5% agarose gel).
Total RNA was isolated from ∼20 mg of muscle tissue by a modified guanidinium thiocyanate-phenol-chloroform extraction method adapted from Chomczynski and Sacchi (7) as previously described (25). The final pellet was resuspended overnight (4°C) in 2 μl 0.1 mM EDTA containing nuclease-free water per milligram original tissue.
RT of total RNA was performed with the Superscript II RNase H-system using the manufacturer's recommended methods (Invitrogen, Carlsbad, CA) as previously described (25). A volume of 11 μl of RNA was reverse transcribed, and the RT product was diluted in nuclease-free water to a total volume of 150 μl.
The mRNA level of each VEGF isoform was quantified by real-time RT-PCR on the ABI Prism 7700 Sequence Detection System (Applied Biosystems, Fosters City, CA). PCR was performed in triplicate using 1 μl of diluted RT products in a total volume of 10 μl of reaction mixture containing the TaqMan probe and forward and reverse primers as determined from the prior optimization procedure, nuclease-free water, and 2× TaqMan Universal Master Mix (AmpliTaq Gold DNA polymerase, AmpErase uracil N-glycosylase (UNG), dNTPs with dUTP, ROX as passive reference; Applied Biosystems, Fosters City, CA). PCR cycles consisted of an initial step at 50°C for 2 min followed by 95°C for 10 min to activate the UNG and AmpliTaq Gold enzyme, respectively, followed by 40 cycles at 95°C for 15 s and at 60°C for 1 min. PCR amplification of the reference gene GAPDH was determined for each sample as control of sample loading and to allow for normalization between samples. Before analyses, RNA from three randomly chosen subjects was quantified using a spectrophotometer, and RT was performed on 2 μg RNA. There was no effect of acute exercise or exercise training on GAPDH levels. A standard curve was constructed from a representative cDNA human sample, and the relative expression in each sample was calculated with respect to the standard curve. All samples from each subject were run together to allow for relative comparisons.
Results are expressed as means ± SE. VEGF mRNA was normalized to GAPDH mRNA levels. Data were presented as fold change, with samples expressed relative to the corresponding untrained preexercise sample, which was set to 1. To evaluate the effect of time and exercise training, a two-way repeated-measures ANOVA was used on normally distributed data. When significant changes were found, a Student-Newman-Keuls method for multiple comparisons was used to determine where the significant changes occurred. A value of P < 0.05 was interpreted as significant.
Time to exhaustion with knee extensor exercise at the load used in the first training session increased from ∼60 min before training to >100 min after the training period (test 1). Time to exhaustion in an exercise bout set to ∼110% of the maximal workload of the untrained leg determined by the incremental test was significantly higher (P < 0.05) in the trained leg compared with untrained leg (3.6 ± 0.7 min for the untrained and 7.2 ± 1.7 min for the trained leg) (test 2).
Levels of mRNA.
For GAPDH mRNA there was no significant interaction between time and exercise training (P = 0.877). The GAPDH mRNA levels in biopsies obtained from the seven subjects before and after the acute exercise bout showed no significant dependence on time (P = 0.679) or exercise training (P = 0.353), confirming that the GAPDH mRNA content in the samples is suitable to use as endogenous control. The amount of GAPDH mRNA was close to the preexercise level, which was set to 1 at all time points, the lowest being 0.88 (untrained leg 6 h) and the highest 1.42 (untrained 24 h).
mRNA levels for total VEGF showed an interaction between time and exercise training (P = 0.048). A more marked (P = 0.045) increase in total VEGF mRNA was observed in the untrained compared with the trained leg. The main values among the different levels of time showed a significant increase (P = 0.003) at 0, 2, and 6 h after exercise with values being back at prelevel 24 h after the exercise bout. This difference was by the post hoc test found to be caused by higher total VEGF mRNA levels in the untrained leg after 0, 2, and 6 h compared with preexercise and 24 h levels (Fig. 1A).
For VEGF165 mRNA there was no interaction between time and exercise training (P = 0.129). A higher (P = 0.004) level of VEGF165 mRNA was observed immediately after exercise but with no difference from rest at 2, 6, and 24 h after exercise (P > 0.05). No differences were found between the untrained and trained leg (P = 0.121) (Fig. 1B).
No interaction between legs and time was observed (P = 0.102) for VEGFex7 mRNA (VEGF165, VEGF189, and VEGF206). There was a main effect of time (P = 0.001) with higher mRNA levels at 0 and 2 h after than before exercise but similar (P > 0.05) to rest at 6 and 24 h after exercise. Training elicited no effect (P = 0.127) on VEGFex7 mRNA levels (Fig. 1C).
The interaction effect for VEGFex6 mRNA (VEGF145, VEGF189, and VEGF206) was P = 0.84. VEGFex6 level was not altered (P = 0.33) by exercise and was not different (P = 0.85) in the untrained and trained leg (Fig. 1D).
The main finding of the present study was that acute exercise induced an increase in VEGF165 and VEGF mRNA encoded by exon 7 (VEGF165, VEGF189, and VEGF206), whereas no increase was observed in VEGF mRNA encoded by exon 6 (VEGF145, VEGF189, and VEGF206). After a 4-wk training period, the exercise-induced increase in total VEGF mRNA was higher in the untrained leg compared with the trained leg.
Expression of splice variants containing exon 7 (VEGF165, VEGF189, and VEGF206) was found to increase at 0 and 2 h after exercise, and the levels had returned to those preexercise after 24 h. In contrast, splice variants containing exon 6 were found not to be altered by exercise. This observation suggests that exercise not only increases the level of VEGF mRNA but also regulates splicing of VEGF mRNA into different isoforms of VEGF. However, the present results do not support a role of training in regulating the expression pattern of splice variants because no differences were observed between the untrained and trained leg in either the expression of VEGFex7 (VEGF165, VEGF189, and VEGF206), VEGFex6 (VEGF145, VEGF189, and VEGF206) or VEGF165. The existence of multiple VEGF isoforms raises the possibility that individual VEGF isoforms might mediate different aspects of vascular growth. This concept is supported by the fact that individual VEGF isoforms exhibit different binding affinities toward heparin and heparan sulfate and, hence, have different extracellular localizations and accessibility to their receptors (23, 26). The present observation that VEGF165 mRNA as well as the splice variants containing exon 7 (VEGF165, VEGF189, and VEGF206) were increased by exercise whereas those containing exon 6 remained unaltered suggests that mainly VEGF165 is upregulated in response to exercise. This is in agreement with our hypothesis and fits with the observation that VEGF165 is the most abundant isoform in resting human skeletal muscle (34). Why it is an advantage to specifically upregulate VEGF165 cannot be concluded from the present study but could be due to different binding properties of the isoforms and hence regulation of the release to the interstitium and availability and stability of VEGF (23) or, plausibly, due to the affinity of VEGF to its receptors (21). Several studies have shown that the more diffusible isoforms (VEGF121, VEGF145, and VEGF165) induce proliferation of endothelial cells and in vivo angiogenesis (19, 23, 27), whereas VEGF189, which binds tightly to heparin sulfate, may be involved in vessel maintenance and specialization. Consistent with this hypothesis are the findings in mice that the highest levels of VEGF188 mRNA (corresponding to isoform 189 in humans) were detected in organs initially vascularized by means of vasculogenesis (e.g., lung, heart, and liver), whereas organs vascularized primarily by means of angiogenesis (e.g., skeletal muscle, brain, eye, and kidney) had higher levels of VEGF164 mRNA (isoform 165 in humans) (22). Thus the specific upregulation of VEGF165 observed in the present study is in accordance with the general belief that exercise-induced capillary growth occurs by angiogenesis.
Because VEGF189 and VEGF206 are also part of the VEGFex7 (VEGF165, VEGF189, and VEGF206) that was found to increase after exercise, these splice variants could be partly responsible for the elevation. If so, the results of a lack of change in VEGFex6 (VEGF145, VEGF189, and VEGF206) would indicate that VEGF145 decreased to a similar extent as VEGF189 and VEGF206 increased. Such an event does not appear very likely, and furthermore VEGF145 mRNA has been found in relatively few cell types and is mainly found in cells derived from reproductive organs (5, 6). Thus results from the present study imply that acute exercise has no effect on VEGF145, VEGF189, or VEGF206. Moreover, our results suggest that exercise does not enhance the VEGF121 mRNA level, because total VEGF (that includes VEGF121) measurements did not differ to a noteworthy extent from analyses of VEGFex7 (VEGF165, VEGF189 and VEGF206), which does not include VEGF121. Therefore, although VEGF121 is known to induce angiogenesis in tissues (15), the present data do not support an exercise-induced upregulation of VEGF121. It should be noted, however, that if only small amounts of VEGF121 are present a possible change would maybe not be detected because of larger amounts of other isoforms.
To study the effect of exercise on total VEGF mRNA, primers and probes amplifying splice variants containing exon 1 and 2 were used. The finding that an exercise bout increased the amount of total VEGF mRNA during recovery is in agreement with earlier studies in rat (4) and human skeletal muscle (13, 28). In several previous human studies, muscle biopsies for VEGF mRNA determination have been obtained immediately after up to 1 h after an exercise bout. These sampling times have probably been based on results obtained in rat, which have shown a maximal increase in VEGF mRNA between 0 and 30 min after exercise. In the present study it was found that VEGF mRNA content was increased immediately after 3 h of exercise and stayed elevated 2 and 6 h postexercise, which is in agreement with recent studies showing elevated VEGF mRNA at 0, 1, and 3 h (14) and 2 and 4 h (10) after an exercise bout. The return of VEGF mRNA to preexercise levels at 24 h of recovery in the present study is, moreover, in line with the study by Hiscock et al. (14). Thus these results suggest that when investigating the effect of exercise training on VEGF mRNA expression it is appropriate to obtain a biopsy 24 h after the last exercise bout. At this time point changes in expression will not be due to acute exercise, and thus a training effect can be investigated.
The effect of physical training on basal levels of VEGF mRNA remains controversial. Previous human studies have reported no changes in skeletal muscle VEGF mRNA after exercise training performed for periods between 6 wk to 3 mo (32). Similarly, the present study found that preexercise samples obtained from untrained and trained leg contained similar levels of total VEGF mRNA. Nevertheless, in heart failure patients, Gustafsson et al. (11) did observe an increase after 8 wk of knee extensor training. The observed differences in the response of VEGF mRNA expression to physical training might be attributable to the disease state and higher age of the subjects in the study by Gustafsson et al. (11). It is plausible that exercise training causes an increase in basal VEGF mRNA expression in the early stage of the training period until somewhere before 4 wk, after which time the exercise-dependent angiogenic stimuli is reduced and VEGF mRNA levels return to baseline. This possibility is supported by the observation that a period of 10 days of knee-extensor training performed at 60–70% of leg maximal O2 consumption (V̇o2 max) resulted in higher basal VEGF mRNA content in trained than in untrained leg. Moreover, in an earlier study in which the subjects trained at a higher intensity (90% of V̇o2 max), it was found that capillarization and the number of proliferating endothelial cells were increased after 4 wk of training, whereas after 7 wk capillarization was not further enhanced and the number of proliferative endothelial cells had returned to baseline levels (17).
The present study shows a higher increase in total VEGF mRNA content in the untrained compared with the trained leg in response to acute exercise performed at the same absolute workload. However, if the acute exercise bout had been performed at the same relative workload, it is possible that in the trained leg, the exercise stimulus would have been sufficiently great to elicit a greater VEGF mRNA response than observed at the same absolute workload. The present observations are in congruence with those of Richardson et al. (29) showing that the VEGF mRNA response to exercise, of one not identified VEGF isoform, was attenuated in the trained state despite the fact that exercise was performed at the same relative workload. These results suggest that training elicits a negative feedback mechanism on the expression of the total VEGF mRNA once an adequate level of capillarization has been reached. It could also be speculated that the increased capillarization in the trained muscle would reduce the effect of exercise performed at the same absolute workload and consequently reduce the signal that induces VEGF mRNA.
In conclusion, this study demonstrates that acute exercise induces a transient increase in total VEGF mRNA that mainly is explained by an increase in splice variants containing exon 7 (VEGF165, VEGF189, and VEGF206), i.e., VEGF165. Furthermore, the attenuated VEGF mRNA response in the trained state suggests that exercise training leads to adaptations in skeletal muscle microvasculature with a consequent reduction of the exercise-induced increase in VEGF mRNA.
The study was supported by grants from the Danish National Research Foundation (504-14). H. Pilegaard was in part supported by the Boje Benzon Stipendie (Denmark).
We are grateful to the subjects who participated in this study and for the technical assistance by K. Møller Kristensen.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society