INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

NUMEROUS STUDIES HAVE shown that chronic low-frequency stimulation (CLFS) transforms fast-twitch, fast-fatigable muscles into slow-twitch, fatigue-resistant muscles (for review, see Ref. 22). In rabbit extensor digitorum longus (EDL) and tibialis anterior (TA) muscles, this process encompasses sequential fiber-type transitions in the order of type IId/x  type IIa  type I. These fiber-type transitions correspond to changes in myosin heavy chain (MHC) isoform composition in the order of MHC IId/x  MHC IIa  MHC I (18). In addition, CLFS elicits profound alterations in the metabolic profile of the muscle fibers, such as severalfold increases in mitochondrial volume and enzyme activities of terminal substrate oxidation, concomitant with reduced glycogenolytic and glycolytic enzyme activities (22). CLFS has also been shown to have strong effects on capillarization, leading to pronounced increases in capillary density (CD) and enhanced perfusion and oxygen supply (3, 4, 9, 14-17, 20). Together, these changes may contribute to the enhanced aerobic-oxidative capacity and fatigue resistance of chronically stimulated muscles.

EDL and TA muscles of the rabbit are fast-twitch muscles with a mixed fiber population. As judged from MHC analyses, the two muscles predominantly contain type IId/x and a smaller fraction of type IIa fibers (1). These two fiber types differ with regard to their myofibrillar protein isoform patterns and their metabolic profiles (21). In view of these differences, the question arises whether CLFS equally affects both fiber populations and, generally, whether changes in capillarization and mitochondrial enzyme activity follow similar or different time courses. It is not easy to answer this question on the basis of the existing literature because, due to different experimental conditions, studies from different laboratories are difficult to compare. With the exception of an investigation from our laboratory, previous morphometric studies on capillarization and quantitative measurements of mitochondrial enzyme activities have not been conducted on the same samples in the same experimental series. The results of our study indicated that stimulation-induced increases in capillarization may precede increases in mitochondrial enzyme activities (24).

The purpose of the present study was to investigate in more detail temporal changes in capillarization and oxidative enzyme activity in TA muscles of rabbit exposed to CLFS for different durations. The animals were stimulated between 1 and 50 days. To relate changes in capillarization to specific fiber types, analyses were performed on fibers histochemically classified according to their myofibrillar actomyosin ATPase (mATPase) activity. Because muscles stimulated for longer than 8 days exhibit pronounced histochemical signs of fiber-type transitions, the analysis of changes in capillarization with regard to fiber types was restricted to the early phase of CLFS. With the use of a newly developed software for image analysis, changes in capillarization were assessed on the basis of several parameters. In addition to the commonly investigated parameters such as CD, capillary-to-fiber ratio (C/F), and number of capillaries around fibers (CAF), we examined intercapillary distance (ICD) and diffusion distance (DD), the capillary arrangement around fibers (CA), and a newly defined capillary supplying factor (SUPF). The study of these parameters appeared to be meaningful because muscle fibers are not a uniform population with regard to size and shape, as well as with regard to the CA. To correlate morphometrically assessed alterations in capillarization with the expression of the vascular endothelial growth factor (VEGF), thought to control adaptive changes of the capillary bed, we studied changes in tissue levels of VEGF mRNA by reverse transcription-polymerase chain reaction (RT-PCR). Finally, citrate synthase (CS) activity, a mitochondrial marker of terminal substrate oxidation and aerobic-oxidative potential, was biochemically determined in the same samples used for morphometry. Because quantitative microphotometric measurements of succinate dehydrogenase (SDH) activity in single muscle fibers had previously been performed on the same muscles investigated in the present study, these results were included, since they provided additional information on the time-dependent changes in aerobic-oxidative metabolic potential.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals and muscles. Adult male White New Zealand rabbits (initial body wt 3,170 ± 290 g and 4,540 ± 130 g after 50 days of CLFS) were used. CLFS (10 Hz, 12 h/day, 1 h on-1 h off) of the left TA muscle was performed via electrodes implanted laterally to the peroneal nerve (26). After different periods of stimulation (0, 2, 4, 6, 8, 10, 14, 21, and 50 days), the animals (3 for each time point) were killed and TA muscles from both the stimulated and contralateral hindlimb were excised and weighed and thin longitudinal fiber bundles were dissected from the superficial layer of the midbelly region. The fiber bundles were divided transversely, with half for biochemical studies and the other half for morphometric analysis. The bundles were quickly frozen in a slightly stretched position in melting isopentane (159°C). Cross sections (9 µm thick) of composite blocks from stimulated and the corresponding contralateral muscles were cut on a microtome in a cryostat at 25°C. For the study of changes in VEGF expression, animals stimulated for up to 14 days were used. To assess the early changes, additional rabbits stimulated for 1 and 3 days (3 for each time point) were included in the study.

Histochemical analysis. Serial cross sections of TA muscles were mounted on glass coverslips, air dried, and subsequently stained for mATPase as previously described (8). Capillaries were identified by two methods. Method 1 was the same as used for mATPase, but the preincubation step was performed at pH 3.80. Method 2 was based on a modification of the immunohistochemical staining for fibronectin, originally described by Erzen and Maravic (5). A mouse anti-human monoclonal antibody (Boehringer Mannheim) was used in the primary immunoreaction, and a peroxidase-coupled anti-mouse immunoglobulin Fab fragment from sheep (Boehringer Mannheim) was used for visualization. Cryosections were air dried and fixed for 10 min in acetone at 20°C. In accordance with manufacturer's instructions, sections were then incubated for 20 min in 5% fetal calf serum in phosphate-buffered saline (PBS) for inhibition of nonspecific reaction. Subsequently, the sections were covered with the primary antibody (1:4 diluted in PBS, containing 1% bovine serum albumin) and incubated for 60 min in a humid chamber at room temperature. After sections were washed in PBS, they were incubated for 60 min at room temperature with the secondary antibody (1:60 in PBS, containing 1% bovine serum albumin). After an extensive washing in PBS, bound peroxidase was visualized by incubation in a solution containing 0.05% 3,3'-diaminobenzidine (Sigma), 50 mM tris(hydroxymethyl)aminomethane (Tris) · HCl (pH 7.3), 0.08% H₂O₂, and 0.06% NiCl₂. After a washing in PBS, the sections were dehydrated and embedded in Entellan (Merck). Control sections were treated in the same way, except the incubation with the primary antibody was omitted.

Morphometry. Measurements were taken with ×25 magnifying objective. Evaluation of morphometric data was performed by image analysis using a Leitz Orthoplan (Wetzlar, Germany) microscope with a scanning stage and an attached video camera (Bosch model TYK 91D, Vidicon tube; linearity  95%). A personal computer (Intel 586, 100 MHz, 1 GB hard disk, 16 MB RAM, 250 MB streamer) with a frame grabber (Vision EZ, Data Translation) was used for analysis. The resolution was 768 × 512 pixels with 8 bits/pixel, corresponding to 256 gray levels. Digitized pictures were displayed in pseudocolors on the computer monitor, and single fiber areas were marked by surrounding their circumference with a computer mouse. Up to six different fiber types could be marked on the computer screen by selecting different color codes. Similarly, the computer mouse was used for marking the capillaries. Capillaries in the periphery of the field of vision, i.e., marginal to peripheral fibers, were delineated and marked in a specific color. Video pictures were independently displayed on a monochrome monitor to compare original and digitized pictures on the computer monitor. Digitized pictures, additional information, and computed results (fiber masks, fiber-type labels, x and y coordinates, point of gravity, capillary marks, comments, and so forth) were stored on the hard disk of the computer.

Morphometric characteristics of muscle fibers. In addition to fiber number and percentage of individual mATPase-defined fiber types, we evaluated the following fiber geometries (Fig. 1): 1) cross-sectional area (CSA), 2) maximum and minimum diameters (DIA_max and DIA_min, respectively), and 3) fiber shape factor. The computer program was designed to automatically evaluate these characteristics from the marked capillaries and fiber areas.

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Schematic representation of morphometric parameters related to fiber cross-sectional area (CSA) and capillarization in cross-sectioned skeletal muscle. CA, capillary arrangement; CAF, capillary around fiber; DIAmax, DIAmin, maximum and minimum diameters, respectively; DD, diffusion distance; ICD, intercapillary distance; PG, point of gravity; SUPF, supplying factor. A-H represent areas of fiber subdivided into 4 segments of equal length, which partition the total fiber area into 8 sectors.

CS activity. Frozen muscle tissue was pulverized under liquid N₂ in a small steel mortar and homogenized (1:20, wt/vol) in a cold 100 mM sodium-potassium phosphate buffer (pH 7.2) containing 2 mM EDTA. The homogenate was sonicated three times for 30 s under intense cooling, stirred for 15 min on ice, and centrifuged for 10 min in a refrigerated Eppendorf centrifuge. The supernatant fraction was decanted and used for determining CS activity in a coupled optical test (24).

Preparation of total RNA. Frozen muscle was pulverized under liquid N₂ and homogenized (1:20, wt/vol) in cold TriStar reagent (AGS, Heidelberg, Germany). Total RNA was isolated according to the producer's instructions using three modifications. 1) Proteins and unsoluble material were removed after homogenization by 10-min centrifugation at 12,000 g (+4°C). 2) Phase separation was performed using 1-bromo-3-chloropropane (Fluka, Buchs, Switzerland). 3) For RNA precipitation, 0.25 ml isopropanol and 0.25 ml of a solution containing 1.2 M sodium citrate and 0.8 M NaCl were used. Pellets were resuspended in 40 µl of diethyl pyrocarbonate (DEPC)-treated H₂O, and RNA concentration was assessed spectrophotometrically. Each preparation was diluted to 1 µg RNA/µl in DEPC-treated H₂O.

Oligonucleotide primers. Oligonucleotide primers specific to a sequence common to VEGF1 (GenBank accession no. S38083), VEGF2 (GenBank accession no. S38100), and VEGF3 (GenBank accession no. S37052) from mouse (positions 184-409) were derived and compared for sequence similarity with other species. Sense primer and antisense primers were derived from sequences identical in mouse, rat, guinea pig, pig, bovine, macaque, and human. The sense primer was GTG GAC ATC TTC CAG GAG TA; the antisense primer was TCT TTG GTC TGC ATT CAC A. For -skeletal actin (GenBank accession no. J00692), the sense primer was (23) CGC GAC ATC AAA GAG AAG CT; the antisense primer was GGG CGA TGA TCT TGA TCT TC.

cDNA synthesis, PCR, and product detection. cDNA synthesis was performed with 2 µl of RNA stock solution in 10 µl volume using the following assay mixture: 10 units avian myeloblastosis virus reverse transcriptase (Pharmacia, Uppsala, Sweden), 12.5 units RNAguard (Pharmacia), 0.625 mM each deoxynucleoside triphosphate, 1 µM specific antisense primer, 50 mM Tris · HCl (pH 8.3), 8 mM MgCl₂, and 30 mM KCl. Incubation lasted for 60 min at 50°C. PCR was performed on several dilutions of the RT assay. One microliter of template was transferred into 24 µl of the PCR incubation mixture for VEGF consisting of 10 mM Tris · HCl (pH 9.0), 1.5 mM MgCl₂, 50 mM KCl, 0.2 µM antisense and sense primers, 0.25 mM dNTPs, and 0.75 units Taq polymerase (Pharmacia). The assay mixture for -skeletal actin PCR differed in the amount of MgCl₂ (2 mM) and a lower pH of 8.3. The annealing temperature was 57°C for VEGF and 59°C for -skeletal actin. The number of cycles was 28 for VEGF and 22 for actin. Product detection [226 nucleotides for VEGF, 367 for -skeletal actin (23)] was performed after electrophoresis on a 6% polyacrylamide gel by ethidium bromide staining. The signals were photographically documented and evaluated by densitometry. At least two measurements were performed on each sample (animal and time point). Changes in mRNA amounts were evaluated by comparison with the amounts of products amplified from cDNA dilutions of control muscles. Dilutions of 1:10, 1:20, 1:40, and 1:100 were used throughout. To correct for changes in total RNA, the results were normalized to actin mRNA levels in the same preparations.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   cDNA sequence of the amplified reverse transcription-polymerase chain reaction (RT-PCR) product specific to rabbit vascular endothelial growth factor (VEGF) mRNA (GenBank accession no. AF022179). Sequences from which sense and antisense primers were derived are printed in bold.

Statistical analyses. A paired t-test was used to determine if differences existed between methods 1 and 2 for capillary staining in serial sections. An unpaired t-test was applied to evaluate differences in fiber types, morphometric data, and VEGF mRNA levels between stimulated and control muscles for animals stimulated for different time periods. Time-dependent changes in capillary characteristics and CS activity were analyzed using a general linear model. When a significant F ratio was obtained, the Duncan's multiple range test (SAS Statistical Software Package, SAS Institute) was used to determine differences between time points. The acceptable level of significance was set at P < 0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Changes in fiber-type distribution. As judged from mATPase histochemistry, low-frequency stimulation induced changes in fiber-type distribution as early as 4 days after stimulation onset. In 4-day-stimulated muscle, the fraction of type IId fibers was reduced by ~12%, but this decrease was compensated for by a corresponding increase in type IIda fibers (Table 1). Whereas no further changes were observed in the distribution of type IId and type IIda fibers at 6 days, a major shift from type IId to type IIda fibers occurred between 6 and 8 days. Type IId fibers could no longer be delineated in 8-day-stimulated muscles. By this time, the fiber composition was characterized by a shift from type IIda to type IIa (Table 1).

Changes in fiber CSA and shape. Early changes in fiber-type distribution were accompanied by alterations in CSA. In agreement with previous results on rabbit adductor magnus and EDL muscles (8), type IId fibers in normal TA muscle were found to be larger than type IIa fibers (Table 2). This pattern was markedly altered in 4-day-stimulated muscles in which type IIda and type IIa fibers displayed a significantly enlarged CSA (P < 0.01). On the basis of previous observations (10, 19), this increase most likely resulted from a transient edema. Compared with the 4-day values, the CSA of type IIda and IIa fibers were smaller again in 6-day-stimulated muscles. Compared with 2-day-stimulated muscles, the mean CSA of type IId fibers was unaltered at 4 days, although it was elevated relative to the contralateral muscles. In 6-day-stimulated muscles, the CSA of type IId fibers was also markedly reduced (P < 0.001), and this reduction continued with the growing populations of type IIda and type IIa fibers. Changes in DIA_min displayed similar trends but were less pronounced than the changes in CSA (Table 2).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Time-dependent changes in muscle weight () and in CSA () of chronically stimulated rabbit tibialis anterior (TA) muscle. Inset: weight changes of stimulated muscles were referred to the weight of contralateral muscles. * Significant differences between time points (P < 0.05).

Histochemical identification of capillaries. Two approaches were chosen for histochemically identifying capillaries in muscle cross sections, mATPase staining after acid preincubation and immunohistochemical staining of fibronectin. To compare the reliability of the two methods, mATPase and fibronectin stainings were evaluated on serial sections from all contralateral and stimulated muscles. As documented by the results from 12,696 fibers and 34,251 capillaries, both methods yielded identical results in the muscles under study (Table 4).

Changes in CD, C/F, ICD, and MCA. CLFS induced pronounced increases in capillarization. This was reflected by steep elevations in CD and C/F, as well as by decreases in ICD and MCA (Figs. 4 and 5). An almost immediate effect of CLFS was observed in 2-day-stimulated muscles. Although increases in CD reached significance only by 6 days (Fig. 4), the decreases in ICD and MCA were significant 2 days after the onset of stimulation (Figs. 4 and 5). These changes continued, reaching their maxima in muscles stimulated for 2 wk. Longer stimulation periods did not lead to further changes.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Time course of changes in intercapillary distance () and capillary density () in low-frequency-stimulated rabbit TA muscle. * Significant differences between time points (P < 0.05).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Time course of changes in mean capillary area () and capillary-to-fiber ratio () in low-frequency-stimulated rabbit TA muscle. * Significant differences between time points (P < 0.05).

Fiber-type-specific changes in capillarization. To investigate fiber-type-specific responses to CLFS, changes in capillarization were evaluated with regard to specific fiber types. Fiber-type-specific patterns of capillarization were obvious in unstimulated muscles. Thus, as judged from Table 5, type IId fibers had fewer capillaries and longer DDs than IIda and IIa fibers. The difference between fiber types became smaller in stimulated muscles. First, increases in CAF and SUPF, as well as decreases in DD, were noted in type IId fibers after 2 days. These changes progressed in 6-day- and 8-day-stimulated muscles, although type IId fibers could no longer be delineated from type IIda fibers at 8 days. As seen in Table 5, type IIda fibers exhibited first changes in these three parameters after 6 days. These early changes seemed to be less pronounced in type IIa fibers, with significant increases in SUPF only after 8 days.

Stimulation-induced changes in VEGF mRNA content. Stimulation led to pronounced increases in the tissue level of mRNA specific to VEGF (Fig. 6). Its increase followed a bimodal time course with a first elevation (2.5-fold) in 1-day-stimulated muscles and a second rise (9-fold) by days 6-8. Thereafter, VEGF mRNA started to decay. Compared with its initial level, it was still 2.9-fold elevated in 14-day-stimulated muscles.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Time course of changes in VEGF mRNA. Tissue levels of VEGF mRNA and -skeletal actin mRNA were determined in total RNA preparations by RT-PCR. To correct for stimulation-induced changes in total RNA content, VEGF mRNA data were referred to -skeletal actin mRNA levels and set equal to one for unstimulated muscles at zero time point. , Contralateral; , stimulated. * Significant differences between stimulated and control muscles (P < 0.05).

Stimulation-induced increases in CS activity. To correlate changes in capillarization with stimulation-induced increases in the aerobic-oxidative potential of energy metabolism, CS activity was measured in homogenates of the same muscles used for histochemistry and morphometric analysis. Low-frequency stimulation for up to 50 days induced a fourfold increase in CS activity. First, significant increases were observed in muscles exposed to low-frequency stimulation for periods longer than 8 days (Fig. 7). A steep rise in activity occurred between 8 and 14 days. Thereafter, it continued to increase at a lower rate. However, as indicated from the slope of the curve, the maximum value was not yet attained by 50 days. Obviously, the rise in total CS activity occurred much later than the increases in capillarization. This observation is confirmed by the increase in SDH activity previously investigated by quantitative microphotometry in large numbers of single fibers from the same muscles used in the present study (29). The increase in SDH activity, therefore, has been included in Fig. 7. Except for an initial decrease in CS, the time-dependent increases of the two enzyme activities were nearly identical. Most likely, the initial decrease in CS activity resulted from a transitory edema of the stimulated muscle. This decrease was not observed by microphotometric determination of SDH activity because measurements were performed on single fibers and results were computed per unit fiber area.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Time course of changes in citrate synthase activity () and succinate dehydrogenase (SDH) activity () of low-frequency-stimulated rabbit TA muscle. Values for SDH are from a previous study (29) in which enzyme activity was assessed microphotometrically on the same muscles analyzed in the present study. * Significant differences between time points (P < 0.05).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The present study focuses on the temporal patterns of adaptive changes in fiber morphology, capillarization, and aerobic-oxidative potential of energy metabolism in response to sustained contractile activity by CLFS. As shown in previous studies, CLFS leads to marked decreases in muscle fiber CSA, as well as to enhanced capillarization and elevations in mitochondrial enzyme activities of terminal substrate oxidation (for review, see Ref. 22). These changes are thought to contribute to an enhanced resistance to fatigue, a conspicuous property of chronically stimulated muscle. Our interest to elucidate in more detail the time course of these changes originates from a previous study in which we showed that stimulation-induced elevations in CS activity and improvements in resistance to fatigue did not occur in a strictly coordinate manner, especially during the first days when improvements in resistance to fatigue preceded the rise in CS activity (27). Maximum increase in resistance to fatigue was observed after 14 days, whereas CS continued to rise with longer stimulation periods. This discrepancy suggested that additional factors, such as early increases in perfusion and capillarization, might contribute to enhanced fatigue resistance. Increases in capillarization had previously been observed in 2-day- (24) and 4-day-stimulated rabbit TA muscles (3).

The present results confirm and extend these findings. However, increases in capillarization and mitochondrial enzyme activities seem to follow different time courses. This discrepancy is most conspicuous during the first days of CLFS but is evident also with prolonged stimulation. Whereas significant changes in ICD, C/F, and MCA first occur as soon as after 2 days and progress with ongoing stimulation, increases in enzyme activities of terminal substrate oxidation become evident only several days later. Significant elevations in CS and SDH activities are recorded after stimulation periods exceeding 8 days. However, as previously shown by studies on key metabolites of energy metabolism, 8-day-stimulated TA muscle exhibits properties suggesting that its energy supply is no longer "glycolytic" but is increasingly based on oxidative phosphorylation (7). Because this change occurs before substantial increases in mitochondrial enzyme activities, we suggest that an elevated capacity for oxygen supply by increased capillarization is a major factor contributing to this qualitative switch of energy metabolism.

Our findings on the effects of sustained contractile activity suggest that elevations in the capacity of mitochondrial end oxidation follow, but do not precede, enhanced capillarization. In other words, the differences between the temporal patterns of improved capacities for local oxygen supply and oxygen consumption clearly point to sequential inductions, i.e., improved oxygen supply first and mitochondrial induction second.

C/F has been reported to be tightly coupled to muscle mitochondrial volume density in 28-day-stimulated fast-twitch muscle of rat (20). However, as shown in the present study, increases in capillarization and mitochondrial volume density must not necessarily occur in parallel to attain a stable ratio in long-term-stimulated muscle. Interestingly, CS and cytochrome oxidase activities reach maximum elevations in rat TA muscle only by 30 days of CLFS (11, 27). In rabbit muscle, the possibility must also be considered that damage and regeneration of some fibers, passing through the myotube stage (19), may transiently perturb the otherwise tight relationship between the aerobic-oxidative potential of muscle fibers and the size of their capillary bed.

The continuing rise of mitochondrial enzyme activities after capillarization has attained a maximum cannot be explained by the present data but possibly relates to a qualitative switch of energy supply. For example, due to continuing elevations in enzyme activities involved in activation and transport and oxidation of fatty acids (24, 25), the latter may be utilized as major fuel in long-term-stimulated muscles. It is noteworthy that, also, the activity levels of CS (present study) and SDH (previous study, Ref. 29) have not reached their maximum by 14 days but continue to rise with stimulation up to 50 days (Fig. 6). In view of these increases, the plateau of the CD, MCA, and ICD values by 14 days supports the notion that maximum capillarization has been attained by this time. This interpretation is in agreement with data on blood flow in low-frequency-stimulated rabbit muscles. Thus blood flow in contracting ankle flexors reaches maximum values by day 14 (13).

The improvement in local oxygen supply relates to two complementary processes, angiogenesis and decrease in fiber CSA. In the present study, both seem to be more or less complete after 2-3 wk. As indicated by the changes in absolute and relative weight (Fig. 3) and transitory increases in CSA (Table 2), the muscle develops an edema during the first days of stimulation, which seems to obscure the initial increase in capillarization, as reflected by decreases in ICD and MCA (Figs. 4 and 5).

Local increases in blood pressure and endothelial injury have been suggested to be important factors initiating the formation of new capillaries (12). In low-frequency-stimulated TA muscles of rat, capillary growth has been related to an increased expression of an endothelial cell-stimulating angiogenic factor (4). An independent study on low-frequency-stimulated EDL and TA muscles of rat demonstrated by Northern blot analyses that the mRNA specific to VEGF was elevated during 3 wk. VEGF mRNA peaked in 4-day-stimulated muscles (6-fold elevation) and declined thereafter (9). An even higher increase (9-fold) is shown in the present study for rabbit muscle. Interestingly, the increase in VEGF mRNA follows a bimodal time course. Although our results refer to mRNA data, VEGF protein(s) might also rise in a bimodal manner, following the changes in mRNA levels. This suggestion is deduced from the bimodal time course of changes in the C/F ratio with increases after 2 and 10 days (Fig. 5).

Three different splice variants have been suggested for VEGF in mouse and four in the human (6). Therefore, the possibility exists that the bimodal increase in VEGF mRNA in rabbit muscle relates to differential expression of splice variants. The latter variants are not distinguished by the sequence under study because this sequence seems to code for a region common to all splice variants. A bimodal temporal pattern could also be explained by assuming that an initial increase of VEGF results from mRNA stabilization, whereas the second and major rise results from enhanced transcription.

In summary, we show that CLFS leads to increases in capillarization as soon as 2 days after the onset of stimulation. Changes in the CAF, SUPF, and DD occur in a fiber-type-specific manner, affecting type IId fibers before types IIda and IIa. Increases in VEGF mRNA precede the changes of the capillary bed. As mirrored by the changes in CD, C/F, ICD, and MCA, enhanced capillarization plateaus by 2-3 wk of CLFS, whereas mitochondrial CS and SDH activities continue to increase in the same muscles. These different time courses indicate that improvements in oxygen supply precede the adaptive increases in aerobic-oxidative potential of energy metabolism.