INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IT IS WELL KNOWN that the additional functional demands imposed on skeletal muscles by training or electrical stimulation regimes can induce adaptations in their size and phenotype (41) and capillary supply (1, 29). Similarly, removal of synergistic muscles leading to compensatory overload of the remaining agonist muscles (20, 36) or imposition of passive stretch by fixation of muscles in a lengthened position (24, 27) both result in an increased muscle mass. Increased capillary supply has been reported after compensatory overload of 5 wk in chicken muscles (27) or 11 wk in mice (36). In rat plantaris, overloaded by synergist denervation, capillary density did not change, but the local capillary-to-fiber ratio (C/F) was increased for all fiber types, demonstrating capillary proliferation (16). In addition, an increased appearance of capillaries in extensor digitorum longus (EDL) muscles of the rat overloaded by synergist removal for up to 22 wk was noted but not quantified (20).

The mechanism for this capillary growth in muscles subjected to compensatory overload is not known. It has been postulated that mechanical factors such as higher luminal shear stress and capillary wall tension, associated with a sustained increase in blood flow, represent an important stimulus for capillary growth (28). This has been demonstrated in chronically stimulated muscles where limitation of functional hyperemia by arterial ligation prevented capillary growth (31). Armstrong et al. (3) showed that there was no difference in blood flow in plantaris or soleus muscles hypertrophied as a result of removal of both heads of gastrocnemius, and acute stretch of muscles is known to decrease blood flow (22, 45, 49). It appears, therefore, that blood flow would not change in compensatory overloaded muscles in such a way as to act as a stimulus for capillary growth and that angiogenesis would have to be initiated by some other means.

Under many conditions, especially during development and under pathological circumstances, angiogenesis is stimulated by growth factors (9). Basic fibroblast growth factor (FGF-2) is a potent mitogen for endothelial cells in vitro and has been localized to the extracellular matrix and fiber endomysium in skeletal muscles of mice (17). FGF-2 activity was enhanced in rat plantaris muscle hypertrophied by removal of synergistic muscles for 7-10 days (50), and it is possible that its release from the extracellular matrix by overload-induced stretch could contribute to capillary growth. In addition, endothelial-cell-stimulating angiogenic factor (ESAF), a low-molecular endothelial cell mitogen (40), has previously been shown to be linked with capillary growth in chronically stimulated muscles (6). The purpose of this study was, therefore, to investigate in muscles exposed to compensatory overload the relationship between capillary proliferation, blood flow, and the presence of growth factors FGF-2 and ESAF. In addition, it was reported that the fatigability of overloaded rat EDL muscles was reduced (20), and we therefore also examined the functional consequences of angiogenesis by determining the time course of changes in muscle performance in relation to capillary supply and blood flow in overloaded muscles. Preliminary data have been reported in abstract form (19).

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals and surgical procedures. Two groups of male Wistar rats weighing initially around 260 g (n = 20) and 175 g (n = 9) underwent unilateral extirpation of the tibialis anterior muscle as described previously (19, 20), randomized with respect to side, under inhalation anesthesia by 2% halothane (Fluothane, ICI) in oxygen. Animals were treated postoperatively with analgesics and antibiotics (0.1 ml sc temgesic and engemycin), in accordance with the Animals (Scientific Procedures) Act of 1986. They were provided with regular chow and water ad libitum. The two groups were taken into final experiments 2 and 8 wk after operation, when body weights were 347 ± 7 and 447 ± 10 g, respectively. Two unoperated groups of rats weighing ~350 g (n = 4) and 450 g (n = 5) were also taken to provide body-weight-matched controls.

Measurement of blood flow, muscle tension, and fatigue. Animals were anesthetized with pentobarbital sodium, initially 50 mg/kg ip (Sagatal, May and Baker), supplemented as necessary via an intravenous cannula. They were fixed supine on a board by moulded retaining bars, with both hind paws held in clamps. Tendons from EDL muscles of both legs were isolated and attached to force transducers (Dynamometer UF1) for measurement of tension developed in response to bilateral stimulation of the peroneal nerves. Isometric twitches were produced at optimal muscle length by stimulation at 4-Hz with 0.3-ms square-wave pulses using supramaximal voltage. Fatigue was assessed as fatigue index (FI) during 5 min of the stimulation by the ratio of final to peak developed tension. Muscle blood flow was determined at rest and at the end of 5-min contractions by radiolabeled microspheres (¹¹³Sn, ⁴⁶Sc; DuPont NEN, Mechelen, Belgium) delivered into the left ventricle through a catheter introduced via the right carotid artery. Reference flow was obtained by withdrawal from one brachial artery by using a precision withdrawal pump (Braun, Melsungen, Germany), and mean arterial pressure was measured by using a catheter in the other brachial artery connected to a pressure transducer (Bell & Howell, Wembley, UK).

Histology for estimation of capillary supply and muscle fiber size. After anesthetic overdose, EDLs from both legs were excised and weighed, and slices from the midbelly were frozen in isopentane precooled in liquid nitrogen. Ten-micrometer cryostat sections were stained for alkaline phosphatase to visualize the location of capillaries by using an indoxyl tetrazolium method (52). Capillary density (mm²) and C/F were estimated from counts of vessels and fibers in two fields covering 0.25 mm² of each muscle cross-sectional area. Mean fiber size was derived from counts of fiber numbers within a coherent square-counting frame at a magnification of ×250 (18).

Measurement of muscle sarcomere length. To determine the effect of compensatory overload on the remaining muscles after removal of tibialis anterior, sarcomere length was measured in extensor hallucis proprius (EHP) muscle, a dorsiflexor synergistic with EDL. In eight animals from each group (2 and 8 wk postsurgery), after the EDL muscles had been excised for analysis, EHP was fixed in situ at resting length by superfusion with phosphate-buffered 2.5% glutaraldehyde (350 mosmol, pH 7.4). Blocks 2 mm³ were then cut, further fixed overnight by immersion in fresh fixative, postfixed in 1% osmium tetroxide, and embedded in Araldite resin. Semithin sections 0.5-µm thick were obtained of longitudinal fiber orientation, stained with toluidine blue, and examined at ×500. An average of 20 sarcomere lengths, where they appeared in register, were counted per muscle.

Extraction of RNA and analysis of mRNA for FGF-2. In a subsample of four rats taken 2 wk after operation, muscles were quickly reweighed after removal of the portion for histology and snap-frozen in liquid nitrogen for estimation of mRNA for FGF-2. Total cellular RNA was extracted from experimental and contralateral EDL muscles by using the RNAzol B method (Biogenesis, Poole, UK) and examined by Northern hybridization (39) and the ribonuclease protection assay (RPAII, Ambion, Austin, TX). A ³²P-labeled FGF-2 cDNA probe (0.5 kb of the coding region of rat FGF-2 cDNA, clone RobFGF103; gift of Dr. A. Baird, The Whittier Institute, La Jolla, CA) was used to detect specific hybridization (43). Briefly, samples were hybridized overnight at 45°C and then digested with 40 µg/ml RNase A (Sigma Chemical) and 500 U/ml of RNase T1 (GIBCO) for 30 min at 30°C. Proteinase K (100 µg, Boehringer Mannheim) in 10% SDS was added to the samples, and the solution was incubated at 37°C for an additional 20 min. The samples were denatured at 95°C after a phenol/chloroform extraction and ethanol precipitation, and 20 µg of each were loaded immediately onto a 4% polyacrylamide/8 M urea gel. Internal controls were added containing 20 µg tRNA. Controls were treated identically to the samples, with the exception of the undigested control, which remained untreated with RNase A or T1. Multiple bands, observed in addition to the expected 0.5-kb protected fragment, were due to different polyadenylation sites within the RNA, and slight smearing, due to the inevitable delay in RNAzol access during homogenization of muscle tissue, was accounted for by comparison of sample and digested control lanes.

Immunolocalization of FGF-2 peptide. In five rats taken 2 wk after operation, immunohistochemistry was carried out on 8-µm cryostat sections taken adjacent to those used for estimation of capillary supply. Polyclonal antibodies to FGF-2 (anti-bovine FGF-2 124 or anti-human recombinant FGF-2, both the gift of Dr. Andrew Baird) were applied after fixation, removal of endogenous peroxidases, and blocking of nonspecific binding with normal goat serum (1.5%). Sections were incubated with primary antibodies or nonimmune serum (control) at 4°C overnight, rinsed with PBS, incubated with biotinylated anti-rabbit IgG (Amersham, Aylesbury, UK), rinsed again, and treated with ABC reagents (Vector, Peterborough, UK). Immunoreactivity was visualized by using diaminobenzidine, and sections were counterstained with hematoxylin or toluidine blue.

Estimation of ESAF levels. After removal of the portion for histology, contralateral and experimental muscles from six animals operated 2 wk earlier were snap-frozen in liquid nitrogen for assay of ESAF levels, according to a modification of the method described earlier (10). The tissue was homogenized in 10 ml (50 mM NH₄HCO₃, pH 7.9, 2 M MgCl₂) and centrifuged at 2,800 rpm for 30 min; 500 µl were taken for protein assay, the pellet was discarded, and the volume made back to 10 ml with the homogenizing solution. This extract was centrifuged through a Filtron Macrosep column (3K cutoff) overnight at 4°C. The diafiltrate was rotary evaporated to reduce the volume to 5 ml. Chromatography was performed by using a C₁₈ column octodecyl silane bonded-to-silicate Bond-Elut column (Varian, Cambridge, UK), which was activated with 5 ml of 100% methanol and washed with 5 ml of distilled water before the sample was applied to the column. To remove crude impurities, the column was washed with 10 ml distilled water and then with 3 ml of 20% methanol. ESAF was eluted with 3 ml of 100% methanol. This final extract was mixed with 10 ml distilled water and freeze-dried. ESAF was quantified by its ability to activate latent collagenase (47), and results were expressed as units of ESAF, where one unit is the percentage of activation of the enzyme per hour per milligram of protein in the supernatant. Two determinations were performed on each extract.

Statistical treatment. Treatment effects were determined by ANOVA (factorial), with intergroup comparisons assessed by post hoc tests (Fisher paired least significant difference), and significance was set at  = 0.05. Data are presented throughout as means ± SE.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Table 1 shows body and muscle weight data for operated animals. Unoperated animals of ~350 and 450 g (21, 25) were used as controls for 2- and 8-wk overload, respectively. Those animals operated 8 wk before and their contralateral muscle weights were significantly larger than those from animals operated 2 wk before, following the normal ontogenetic growth curve. Consequently, muscle-to-body weight ratios for contralateral muscles were not different between the 2- and 8-wk groups (Table 1), and both were similar to values for EDL muscles in the unoperated animals of matched body weights, 0.049 ± 0.001 (n = 12) and 0.045 ± 0.001 (n = 10), respectively. In the experimental limbs, EDL underwent significant hypertrophy to reach 119 ± 2 and 151 ± 4% of the weight of contralateral muscles at 2 and 8 wk, with muscle-to-body weight ratios increasing to a similar extent (Table 1). Mean muscle fiber size was similar in contralateral and hypertrophied muscles from the two groups, being only 6.6 ± 0.4 and 6.6 ± 0.6% larger, respectively, with no histological evidence of tissue edema. Stretch of the overloaded muscles was evidenced by a 20 ± 2% increase in sarcomere length in EHP after 2 wk, from 2.41 ± 0.07 µm in contralateral muscles to 2.89 ± 0.01 µm (P < 0.01). After 8 wk, there was no difference between sarcomere lengths from contralateral (2.71 ± 0.10 µm) and experimental muscles (2.83 ± 0.03 µm).

C/F in normal rat EDL muscles was found to be 1.44 ± 0.06 in both control unoperated groups of animals, irrespective of body weight (6, 21). The value in muscles contralateral to those overloaded for 2 wk was slightly but not significantly higher than this, whereas in the overloaded muscles themselves C/F was significantly increased by 21 ± 6%, compared with contralateral muscles (Table 1), and by 33%, compared with weight-matched controls. After 8-wk overload, C/F was 42 ± 14% (P < 0.05) greater than in contralateral muscles, but these muscles had a C/F ratio that was itself 17% greater than in normal control muscles. Whereas this would indicate that modest adaptive changes in capillarity have occurred in the contralateral muscles after 8 wk, the true increase in C/F in response to 8-wk overload relative to control muscles would be much greater, ~60%.

In muscles contralateral to overloaded ones, blood flow at rest was not significantly different 2 or 8 wk after operation (14.6 ± 4.0 and 7.5 ± 1.6 ml · min¹ · 100 g¹, respectively) from that in control unoperated muscles (8.3 ± 1.3 ml · min¹ · 100 g¹). Muscles hypertrophied by overload for 2 wk also had similar resting flows (10.0 ± 2.1 ml · min¹ · 100 g¹). After 8-wk overload, however, resting blood flow was 15.8 ± 3.0 ml · min¹ · 100 g¹, almost twice that of control muscles (7.4 ± 1.0 ml · min¹ · 100 g¹; see Ref. 25) (P < 0.05). Blood flow was also measured during indirect stimulation at 4 Hz, which is the frequency producing maximum blood flow in contracting muscles in the rat (25). Figure 1 shows that, in muscles overloaded for 2 and 8 wk, maximum blood flows (132.1 ± 37.4 and 177.0 ± 22.2 ml · min¹ · 100 g¹, respectively) were no different from those in normal muscles of unoperated weight-matched controls (155.6 ± 12.2 and 149.8 ± 10.3 ml · min¹ · 100 g¹, respectively; see Refs. 21, 25). There was thus no indication that any alteration in blood flow would be likely to initiate the increase in capillarity upon overload. Maximum flows in contralateral muscles were also similar after 8 wk (145.4 ± 24.9 ml · min¹ · 100 g¹) but showed some reduction after only 2 wk (89.0 ± 17.9 ml · min¹ · 100 g¹, P < 0.01 vs. weight-matched control).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Maximal extensor digitorum longus (EDL) muscle blood flow (MBF) measured by radiolabeled microspheres at end of 5-min isometric contractions at 4 Hz. Data from control unoperated rats (open bars) that were weight-matched with groups in which EDL was overloaded by stretch following extirpation of a synergist muscle for 2 and 8 wk (stippled bars).

Specific peak tension developed during contractions by muscles overloaded for 2 wk and by their contralateral muscles was similar (264 ± 13 and 266 ± 20 g/g, respectively; not significant) as was FI (0.85 ± 0.04 and 0.80 ± 0.03). However, after 8 wk, specific peak tension was reduced (183 ± 11 g/g, P < 0.01) and FI increased (0.92 ± 0.02, P < 0.01), compared with contralateral muscle values of 301 ± 12 g/g and 0.77 ± 0.03 for specific peak tension and FI, respectively (Fig. 2), and control values of 271 ± 13 g/g and 0.64 (25), respectively.

View larger version (39K): [in this window] [in a new window]   Fig. 2.   A: peak specific twitch tension per gram muscle weight, measured during 5-min isometric contractions at 4 Hz in 2- and 8-wk overloaded EDL muscles (stippled bars) and their respective contralateral control muscles (open bars). B: fatigue index during 5-min isometric twitch contractions at 4 Hz expressed as final/peak tension × 100. * P < 0.01 vs. contralateral.

With respect to the presence of growth factors, mRNA for FGF-2 was demonstrated in almost all contralateral and 2-wk overloaded EDL muscles, but there was no consistent relationship between the intensity of expression and capillary supply (Fig. 3). Similarly, FGF-2 peptide immunoreactivity was observed in all muscles examined, localized to nerves, larger blood vessels, and around the fiber basement membrane (Fig. 4). However, there was no discernible difference in the distribution or intensity of staining reaction between overloaded and contralateral muscles that could be related to the increase in C/F. In contrast, activity of the endothelial mitogen ESAF was significantly higher in overloaded EDL, in comparison with the contralateral muscles (6.22 ± 0.59 vs. 4.05 ± 0.25 units, respectively, P < 0.01).

View larger version (82K): [in this window] [in a new window]   Fig. 3.   Gel showing ribonuclease protection assay for basic fibroblast growth factor (bFGF-2). A, undigested probe; B, digested probe. 1, tibialis anterior muscle that was extirpated to produce overload on EDL; 2, EDL muscle after 2-wk overload; 3, contralateral control EDL; 4, contralateral tibialis anterior. Loading of each lane was equal, with 20-µg RNA. Bars = 10 µm.

View larger version (134K): [in this window] [in a new window]   Fig. 4.   Cryostat sections of EDL muscles after immunostaining for FGF-2 peptide. Top: contralateral (left) and 2-wk overloaded (right) EDL showing immunolocalization of FGF-2 in nerves (*), blood vessels (large arrows) and extracellular matrix (small arrows). Middle: overloaded EDL after incubation with FGF-2 antibody (left) and parallel section incubated with nonimmune serum (right) to show specificity of reaction;  and * identify the same fiber in both sections. Bottom: overloaded EDL stained for alkaline phosphatase to depict capillaries (left) and parallel section showing localization of FGF-2 immunostaining (right). Bars = 10 µm.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

This study has established that capillary growth induced in skeletal muscle hypertrophied by long-term passive stretch during compensatory overload is independent of any increase in blood flow or changes in expression of FGF-2. It is, however, linked with higher levels of ESAF. Many previous studies have induced gross hypertrophy by bilateral extirpation of gastrocnemius or extirpation and denervation of remaining synergists. We chose a less dramatic intervention of extirpation of tibialis anterior alone as used previously (20), confirming the observation of increases in EDL muscle mass on the order of 20% after 2 wk and establishing an accompanying increase in C/F of similar magnitude. Stretch induced either by overloading chick wing muscles with weights (27) or by extirpation of agonists in mice (36) or rats (16) also increases capillary supply. However, as these changes were estimated in muscles exposed to stretch for 5 wk (16, 27), whereas muscle hypertrophy occurred much earlier (2), we investigated both shorter (2-wk) and longer (8-wk) exposure to stretch. Both muscle hypertrophy and capillary supply, expressed as C/F to take into account changes in number of capillaries and fiber size, were increased in overloaded muscles between 2 and 8 wk, suggesting that the adaptations were progressive. Compensatory hypertrophy was evidently incomplete even after 8 wk, because the increase in EDL muscle mass was still much less than the mass of the extirpated tibialis anterior muscle. Although avian models of stretch overload produce large increases in muscle fiber size and number, in rodents this appears not to be the case. Muscle mass is the product of fiber area, number, and length. Long-term passive stretch resulted in a 20% increase in sarcomere length after 2 wk, with no additional change thereafter. Because the increase in fiber size was small and there was no further increase in sarcomere length at 8 wk, the increase in muscle mass at this later stage was most likely due to addition of new sarcomeres and/or new fibers, as described previously (2, 48).

Previous work from our laboratory has suggested that mechanical factors associated with increased blood flow may act as an initial stimulus for capillary growth in skeletal muscle (29). Capillary growth can be rapidly induced, within 4-7 days, by chronic electrical stimulation (28) or by a 2-wk administration of a vasodilator, the ₁-blocker prazosin (21). Under these conditions, blood flow is increased over a prolonged period of time, and we proposed that this hyperemia, resulting in increased capillary shear stress and/or wall tension (15), is a potential stimulus for capillary growth. It has been established that blood flow is considerably decreased when stretch is applied to muscles acutely (22, 45). Using microspheres, Wisnes and Kirkebo (49) found absence of flow in the medial region of gastrocnemius on passive stretch, and Poole et al. (42) observed that acute stretch of the spinotrapezius muscle resulted in a decrease in capillary diameter and velocity of flow. From the evidence of stretch in the present study, based on the increase in sarcomere length, it is therefore unlikely that capillary growth was initially induced by changes in blood flow and that the significant increase in C/F was probably induced without a higher shear stress. This conclusion is supported by the observation, in comparison with control muscles, of similar resting and maximal blood flows in muscles overloaded for 2 wk, in which it appears that the increase in capillary supply that has already occurred has normalized any decrement in blood flow arising from the initial stretch. The higher resting blood flow found in muscles overloaded for 8 wk cannot be responsible for the initiation of capillary growth, since this had occurred earlier, but it may be a consequence of an expanded capillary bed, contributing to the continuation of capillary growth. It was noted that blood flow was somewhat reduced in contralateral muscles after 2 wk, whereas C/F ratio was increased in contralateral muscles after 8 wk. Contralateral muscles are thus clearly affected by the intervention to the opposing limb, probably because the animals use predominantly the contralateral leg for few days after the operation. For this reason, where possible, we have made comparisons with data from muscles of unoperated weight-matched controls.

Whereas mechanical factors associated with blood flow did not relate to capillary growth, it is possible that a growth factor such as FGF-2, the expression of which has been reported to increase in hypertrophied muscles (50) and which can be released in the retina by stretch (12), could be involved. The presence of FGF-2 mRNA in both contralateral and stretched muscles indicates that the growth factor is, indeed, expressed, but there was no apparent link in levels of expression with changes in C/F ratio. This is not so surprising, since expression in whole muscle extracts would not necessarily reveal changes in relation to the vasculature, because capillaries represent only a very small proportion of the whole tissue, and FGF-2 is also known to be present in the fiber sarcolemma (17). More important is the negative immunohistochemical finding in relation to capillaries. Because FGF-2 was demonstrated in the nerves and even in some larger vessels, the negative finding in capillaries cannot be due to the lack of specificity of antibodies used in these experiments. Rather, it indicates that its role in angiogenesis under relatively physiological conditions is negligible, in agreement with previous findings that FGF-2 causes proliferation only in compromised and not in normal vasculature (11). Although we were not able to demonstrate any increase in expression of FGF-2 that could contribute to capillary growth in muscles undergoing compensatory overload, it is possible that endothelial cells themselves could be rendered more susceptible to the effects of any growth factor that is present. Ingber (34) demonstrated in vitro that endothelial cells that were more spread out, by being grown on different substrata, showed more proliferation in the presence of FGF-2.

In contrast to FGF-2 activity, activity of low-molecular ESAF was clearly elevated in overloaded muscles, because of either increased expression or increased bioavailability. This factor has been found in tissues exhibiting capillary growth such as diabetic retina (40) as well as in skeletal muscles with intensive capillary growth elicited either by chronic electrical stimulation (6) or by prazosin treatment (7). So far, the mechanism by which ESAF could initiate capillary growth has not been elucidated. However, repetitive stretch increases tissue-type plasminogen activator production by endothelial cells in vitro (33), and plasminogen activator and plasmin are potent activators of the pro-forms of interstitial collagenase and stromelysin, which are important in enabling the newly formed capillary to proceed through the extracellular matrix. Both pro-forms of these enzymes are also activated by ESAF (38). In addition, ESAF has been shown to be both mitogenic and chemotactic for microvascular endothelial cells (40).

It is conceivable that in addition to a possible angiogenic stimulus provided by growth factors, passive stretch of a muscle may initiate capillary growth through mechanical forces imparted to endothelial cells directly via their interaction with the surrounding extracellular matrix. It is well known that endothelial cells are very responsive to mechanically induced changes in their shape (14). Stretching of endothelial cells in culture resulted in increased incorporation of [³H]thymidine (35), indicating growth, possibly triggered by reorientation of their cytoskeleton (13, 44). The latter, together with generation of inositol phosphates and diacylglycerol, has been shown to be accompanied by an increased intracellular Ca²⁺ concentration (46), which is an important signal in the stimulation of cellular proliferation (5). Furthermore, intracellular Ca²⁺ can also be increased by stretch-activated ion channels in vascular endothelial cells (37). Thus stretch of capillaries in line with the surrounding tissue may promote endothelial cell proliferation by one of these mechanisms.

The extent to which changes in capillary supply could contribute to improved fatigue resistance in compensatory overload hypertrophy was also investigated. Reported changes in muscle performance following overload include increased maximal twitch and tetanic tensions but lower specific tensions and reduced fatigability (20). Similarly, we observed that tension development by hypertrophied muscles normalized for muscle mass was significantly reduced after 8 wk. This may be related to an increase in fiber length or addition of immature sarcomeres, causing a significant reduction in tension development. In addition, decreased electromyographic activity (26), a lower myosin ATPase activity (4), and increased protein degradation (23) have all been observed in overloaded muscles. However, we did observe reduced fatigability after 8-wk overload, the origin of which may be a combination of altered muscle phenotype and adaptation of the vascular supply. Transformation toward a slow phenotype may occur in overloaded muscles, since a shift from type IIb toward type I fiber types has been reported (16, 51). On the other hand, Frischknecht and Vrbová (20) described no difference in the number of slow myosin-positive fibers and only slight qualitative differences in histochemical staining for succinate dehydrogenase (SDH) in EDL muscles of the rat overloaded by synergist removal for up to 22 wk, whereas oxidative enzyme activities in rat plantaris muscles were normal after several weeks of overload, following a transient decrease (32), or decreased in overloaded cat plantaris (8). In the present study, there was also little evidence for major differences in SDH staining (unpublished observations). Therefore, the metabolic response of muscles to overload appears variable. Little change in peak tension and fatigability was found at 2 wk, despite a 21% increase in C/F, suggesting that there may be a threshold below which capillary supply has little effect on overall resistance to fatigue. In muscles overloaded for 8 wk, endurance was improved by 20% while C/F increased by 42% relative to contralateral muscles, even though maximal blood flow was not different. It is likely that reduced diffusion distances following an increased capillary supply improve oxygen delivery, and, therefore, fatigue resistance (30).

This study has established that capillary growth induced in skeletal muscle hypertrophied by long-term passive stretch during compensatory overload is not initiated by any increase in blood flow or changes in expression of FGF-2 but that it is correlated with higher levels of ESAF. We conclude that mechanical factors acting on capillaries from the abluminal side are the primary angiogenic stimuli and that the increase in capillary supply is an important factor in the increased resistance of overloaded muscles to fatigue.