Development of T-tubular vacuoles in eccentrically damaged mouse muscle fibres

Abstract

Single fibres were dissected from mouse flexor digitorum brevis muscles and subjected to a protocol of eccentric stretches consisting of ten tetani each with a 40 % stretch. Ten minutes later the fibres showed a reduced force, a shift in the peak of the force–length relation and a steepening of the force-frequency relation. Addition of the fluorescent dye sulforhodamine B to the extracellular space enabled the T-tubular system to be visualized. In unstimulated fibres and fibres subjected to 10 isometric tetani, the T-tubules were clearly delineated. Sulforhodamine B diffused out of the T-tubules with a half-time of 18 ± 1 s. Following the eccentric protocol, vacuoles connected to the T-tubules were detected in six out of seven fibres. Sulforhodamine B diffused out of the vacuoles of eccentrically damaged fibres extremely slowly with a half-time of 6.3 ± 2.4 min and diffused out of the T-tubules with a half-time of 39 ± 4 s. Vacuole production was eliminated by application of 1 mm ouabain to the muscle during the eccentric protocol. On removal of the ouabain, vacuoles appeared over a period of 1 h and were more numerous and more widely distributed than in the absence of ouabain. We propose that T-tubules are liable to rupture during eccentric contraction probably because of the relative movement associated with the inhomogeneity of sarcomere lengths. Such rupture raises intracellular sodium and when the sodium is pumped from the cell by the sodium pump, the volume load of Na⁺ and water exceeds the capacity of the T-tubules and causes vacuole production. The damage to the T-tubules may underlie a number of the functional changes that occur in eccentrically damaged muscle fibres.

Muscles that are stretched during contraction (eccentric contractions) are susceptible to damage (Hough, 1902), particularly when the eccentric exercise is prolonged and unaccustomed. For example, when walking down a mountain, the quadriceps group of muscles is used to deaccelerate the body and undergo eccentric contraction. Characteristically the muscles involved are weak but pain-free immediately after the exercise, but in the subsequent days the muscle becomes stiff, tender and may become even weaker. This latter collection of symptoms is called delayed onset muscle soreness (DOMS). Full recovery may require several weeks and there is a pronounced training effect; thus if the same exercise is repeated the symptoms are substantially reduced (Balnave & Thompson, 1993).

The mechanisms involved in this damage are complex and multifactorial (for recent reviews see Morgan & Allen, 1999; Proske & Morgan, 2001; Warren et al. 2001). Early ultrastructural studies established that there are characteristic changes in the sarcomere patterns including localized regions of overstretched sarcomeres and irregular and distorted Z lines (Fridén et al. 1981). Morgan (1990) proposed a mechanism to explain the sarcomere inhomogeneities on the basis of the mechanical properties of eccentrically contracting muscle. He pointed out that when the force on a muscle exceeded about 1.6 P_o (P_o is the maximum isometric force), the stretching velocity increased uncontrollably (Katz, 1939). Furthermore, sarcomeres on the descending limb of the force–length curve are intrinsically unstable (Huxley, 1980); this arises when a sarcomere on the descending limb becomes longer than its neighbour; it is then weaker than the neighbour and therefore inclined to stretch further. As a consequence during stretch on the descending limb of the force–length curve, individual half-sarcomeres can stretch suddenly and dramatically (the popping sarcomere hypothesis) and much of the increase in muscle length can occur in a small number of abnormally stretched sarcomeres (Brown & Hill, 1991; Talbot & Morgan, 1996). Most of these overstretched sarcomeres reinterdigitate when the muscle relaxes (Talbot & Morgan, 1996), but during repeated eccentric contractions it would be expected that the regions of overstretched sarcomeres can extend laterally and longitudinally through the muscle leading to the characteristic structural changes (Allen, 2001).

It is less clear how these early structural changes lead to the reductions of force and to the inflammation and even cell death that are apparent after several days (McCully & Faulkner, 1985). Sarcomeres are connected in series in a muscle so that damage to some sarcomeres would not necessarily affect the maximum force. Instead one might predict that they would act as increased compliance and cause the force–length curve to shift to longer lengths as has been observed (Katz, 1939; Talbot & Morgan, 1998). However even when muscles are stretched to the peak of the new force–length relation, force is often reduced. Warren et al. (1993) suggested that damage to excitation-contraction (E-C) coupling occurred in eccentrically damaged mammalian muscle and this was subsequently confirmed by direct measurements of intracellular calcium (Balnave & Allen, 1995; Ingalls et al. 1998). It would be expected that T-tubules and sarcoplasmic reticulum (SR) would be susceptible to damage in the overstretched regions and this may be the basis for the E-C coupling damage. It is already known that T-tubular vacuoles can occur in a range of muscle interventions involving osmotic loads (Krolenko et al. 1998; Khan et al. 2000,Lännergren et al. 2000). Of particular interest is a report that gross structural damage to a muscle by cutting it in half leads to rapid and profuse development of vacuoles associated with the T-tubules near the site of damage (Casademont et al. 1988). Very recently it has been shown in electron microscopy (EM) studies that the T-tubules are abnormal in eccentrically stretched muscles (Takekura et al. 2001). However, there are no previous reports of vacuole development in association with eccentric muscle damage.

There is also considerable evidence of surface membrane damage following eccentric contractions. For instance, some studies have shown reductions in the membrane potential immediately after eccentric damage (McBride et al. 2000), although others have not (Warren et al. 1993). We have recently shown that pH regulation is abnormal immediately after eccentric damage; since transporters in the surface membrane, such as the Na⁺-H⁺ exchanger and the lactate transporter, dominate pH regulation, this suggests some loss of membrane function (Yeung et al. 2002). Substantial increases in the plasma protein levels of soluble muscle proteins such as creatine kinase may be observed several days after eccentric damage (Jones et al. 1986), although this may be more indicative of the start of cell breakdown than early membrane damage.

In this study, T-tubule morphology and function were compared between fibres which had not contracted or had undergone only isometric contractions, and fibres which had undergone eccentric contractions. The morphology of the T-tubules was examined using confocal microscopy and an extracellular fluorescent dye, sulforhodamine B. In addition the accessibility of the T-tubules from the extracellular space was evaluated from the rate of washout of sulforhodamine B from the T-tubules. Finally, the role of the Na⁺,K⁺-ATPase in the development of T-tubule vacuoles was evaluated by performing eccentric contractions in the presence of ouabain.

METHODS

Single fibre dissection and mounting

Adult, male mice were killed by rapid neck disarticulation and single muscle fibres were dissected from the flexor brevis muscles as previously described (Lännergren & Westerblad, 1987). These procedures were approved by the Animal Ethical Committee of the University of Sydney. The isolated fibres were mounted between an Akers AE801 force transducer (SensoNor, Horten, Norway) and the arm of a motor (Model 300H, Cambridge Technology, Cambridge, MA, USA). The motor allowed known length changes to be imposed on the muscle fibre. The fibres were stimulated with platinum-plate electrodes using pulses of 0.5 ms duration at an intensity of ∼1.2 × threshold. All tetanic contractions were 400 ms in duration and the standard stimulation frequency was 100 Hz.

Solutions

The dissection was performed in a solution of the following composition (mm): 136.5 NaCl, 5 KCl, 1.8 CaCl₂, 0.5 MgCl₂, 0.4 NaH₂PO₄ and 11.9 NaHCO₃ (pH 8.0). During the experiment, the fibres were superfused at 2.2 ml min⁻¹ in the following standard solution (mm): 121 NaCl, 5 KCl, 1.8 CaCl₂, 0.5 MgCl₂, 0.4 NaH₂PO₄, 24 NaHCO₃ and 5.5 glucose, with 0.2 % (v/v) fetal calf serum (Gibco). The solution was bubbled with 95 % O₂-5 % CO₂ (pH 7.4). In some experiments, 1 mm ouabain (Sigma-Aldrich) was added to inhibit the activity of Na⁺,K⁺-ATPase. Ouabain was dissolved in dimethyl sulfoxide (DMSO) to give a concentration of 100 mm. This stock solution was stored at a temperature of −20 °C, protected from light, and diluted into the standard solution immediately before use. All experiments were performed at room temperature (22–24 °C).

Eccentric contraction protocol

The experimental protocol used was similar to those described previously (Balnave & Allen, 1995; Yeung et al. 2002). The basic protocol (eccentric series) involved superimposing a stretch on each of a series of 10 tetani. It is important to distinguish between the force deficit caused by fatigue and that associated with eccentric contractions (Morgan & Allen, 1999) and for this reason an isometric (control) series of tetani was also performed. We also wished to establish that the force-frequency behaviour and the force–length properties of the fibre were unaffected by the control series, but modified by the eccentric series. Thus the full experimental protocol was as follows. (i) The force–length relation was established using tetani separated by 1 min and changing length in 40 μm steps. The fibre was adjusted to L_o (the length giving maximum force). (ii) Tetani at 10, 20, 30, 50, 70 and 100 Hz were produced with 1 min of rest between each tetanus (force-frequency relation). (iii) Ten isometric tetani at 4 s intervals were given (the control series). (iv) After 10 min, the force–length relation was redetermined. (v) The force-frequency relation was redetermined. (vi) Ten eccentric tetani were given at 4 s intervals. The length change was either 25 % or 40 % of L_o. The stretch was imposed over 100 ms starting after 200 ms of the tetanus. The fibre was shortened back to the L_o over 100 ms starting 200 ms after the end of the tetanus. (vii) After 10 min the force–length relation was re-established and the fibre length kept at the original L_o. (viii) The force-frequency relation was redetermined. (ix) The fibre was then stretched to the new optimum length (post eccentric L_o). (x) The force-frequency relation was redetermined. (xi) Finally, the fibre was removed from the motor and positioned on the coverslip of a new chamber used for confocal microscopy.

Confocal microscopy

To visualize the morphology of the T-tubular system, the fibre was examined with an inverted confocal microscope (Leica TCS SL, Heidelberg, Germany), using a × 63, NA 1.2 water immersion objective lens. After the contraction protocol, the muscle fibre was transferred into a perfusion chamber. The two tendons of the fibre were fixed to the coverslip with silicone grease. The perfusion chamber was narrow so that the washout of extracellular space after a solution change was reasonably fast (see Fig. 4A). T-tubular morphology was examined by adding 0.5 mm sulforhodamine B (Molecular Probes, OR, USA) to the perfusate which diffused into the T-system (Endo, 1966; Lännergren et al. 2000). Sulforhodamine B fluorescence was excited using the 543 nm line of a helium-neon laser operated at 50 % maximum power and emitted light was collected between 570 and 610 nm. The resolution of the microscope in this configuration was estimated using 170 nm diameter green fluorescent beads (PS-Speck, Molecular Probes). The half-maximal width of the resulting fluorescence signal was 0.35 μm in the horizontal direction and 0.9 μm in the vertical direction.

Figure 4. Fluorescence intensity during dye washout.

A, measurements from large areas of the fibre. ▵, eccentric series (n = 7); ○, isometric series (n = 4); •, background fluorescence tracing of the extracellular dye (n = 11). Error bars represent s.e.m. values. B, the fluorescence intensity of vacuoles in eccentrically damaged fibres (•) (n = 5), T-tubules from eccentrically damaged fibres without vacuoles and T-tubules from 25 % eccentric stretches (▴) (n = 4) and T-tubules from isometric-only series (○) (n = 4).

Confocal microscopy was performed on four unstimulated fibres and on three fibres that underwent only the control series of isometric tetani. Seven fibres underwent the full protocol described above. The entire length of the muscle fibre and multiple depths were examined for morphological alterations and/or presence of vacuolation, but because of dead fibres and connective tissue the ends of the fibre were harder to visualize. Images are presented in two formats. In some figures, e.g. Fig. 2, sulforhodamine B was present in the extracellular space and produced a very large fluorescent signal that caused the detector to saturate (indicated by a blue colour). Even in a confocal microscope some fractions of the fluorescence above and below the fibre contributed to fibre fluorescence and this reduced the resolution of fluorescent structures within the fibre. For this reason we also show figures, e.g. Fig. 5, collected 1 min after extracellular sulforhodamine B had been washed out of the extracellular space. Images were scanned at a resolution of 512 × 512 pixels.

Figure 2. Confocal images of the muscle fibres in the presence of extracellular sulforhodamine B.

A, confocal image of a fibre following isometric contraction showing well-aligned fluorescent patterns of the T-tubules. In all the confocal images shown, increasing fluorescence intensity is indicated by the following colour order: black, dark red, yellow, white. Saturation of the detector occurs with dye outside the fibre and is indicated by the blue colour. B and C, fibre that had been stretched by 40 % L_o showing vacuolation of the T-system. B shows an example of the ovoid-shaped vacuoles appearing longitudinally over the fibre and C shows roughly spherical vacuoles located randomly throughout the fibre. Scale bar, 10 μm.

Figure 5. Confocal images of muscle fibres in ouabain experiments.

A, fluorescence images following isometric (control) series. Aa, 20 min after isometric contractions, ouabain present throughout. Ab, after ouabain had been washed out for 60 min. No vacuoles were observed at any period. B, fluorescent images following the eccentric series. Ba, 20 min after the eccentric series, ouabain present throughout, no vacuoles observed. Bb, after 40 min washout, hints of vacuoles began to appear. Bc and d correspond to 60 and 80 min after the removal of ouabain. Extensive vacuoles appear over this period. Scale bars, 10 μm. All images in this figure were collected 1 min after washout of sulforhodamine B, which was present for 10–20 min before washout.

In addition to studying the morphological alterations of the T-tubules, a second objective of our study was to examine the function of the T-tubules following eccentric contractions. Water-soluble dyes, such as sulforhodamine B (MW 559), diffuse in and out of the T-tubular system with a time course in the range of 4–10 s (Endo, 1966). It is thus possible to examine the accessibility of the T-tubules by measuring the rate of exit of the fluorescent dye. In order to achieve this, fibres were superfused with the standard solution containing sulforhodamine B for 15–20 min before observations. The dye-containing solution was then removed and washed out with standard solution for at least 20 min. Confocal images of the fibres were obtained every 10 s for the first minute, every minute for the next 4 min and then at 5 min intervals over 20 min. The rate of loss of sulforhodamine B from the T-system was estimated from the decline of fluorescence, which was analysed using NIH Image (Scion Corporation, MD, USA). The average fluorescence intensity was determined in a window of roughly half the diameter of the fibre placed over a representative region of the fibre. Before exposure to sulforhodamine B the fibres exhibited no significant fluorescence. Contributions to the fluorescence from dye outside the fibre profile were judged to be small because the intensity of vacuoles changed little as the extracellular dye was washed away (compare vacuoles in Fig. 3Ba and b). This procedure was repeated for several regions of each fibre to ensure that fluorescence intensity measurements were representative. When vacuoles were present a much smaller box comparable to the size of a few vacuoles was used and placed over the same group of vacuoles in each of a series of images.

Figure 3. Confocal images showing washout of sulforhodamine B.

Aa, fibre that had been subjected to 10 isometric tetani and was being perfused with sulforhodamine B. Ab and c are after 1 and 20 min washout with standard solution. After the removal of the sulforhodamine B, the fluorescence density of the fibre reduces rapidly in the first minute. Ba, fibre that has been subjected to eccentric tetani of 40 % L_o in the presence of extracellular sulforhodamine B. Bright fluorescent vacuoles appeared longitudinally along the muscle fibre. Bb, c and d correspond to the time course of washout period of 1 min, 20 min and 2 h, respectively. Scale bars, 10 μm.

Na⁺,K⁺-ATPase inhibitor

We further examined the morphology of the T-tubules after eccentric contractions (n = 3) during exposure to 1 mm ouabain, an inhibitor of Na⁺,K⁺-ATPase. The fibre was exposed to ouabain from 15 min before the eccentric protocol and the ouabain remained present throughout the eccentric protocol and during the examination of the fibre on the confocal microscope. Subsequently ouabain was washed out using standard solution for the next 2 h and we again examined the fibre on the confocal microscope. Two fibres were treated in an identical manner except that the experimental protocol involved isometric tetani only.

Statistics

Data are quoted as means ± standard error of means (s.e.m.) with the number of experiments denoted as n. Statistical significance was determined with Student's paired t test for force recordings and optimal length. The unpaired t test was used to compare the rate of dye exit between fibres following isometric and eccentric tetani. The significance level was set at P < 0.05.

RESULTS

Effect of the eccentric series on mechanical performance

The force developed by muscle fibres following the eccentric series (40 % stretch, n = 7) showed a large reduction to 34 ± 4 % of control (P < 0.0001), measured in a 100 Hz isometric tetanus 10 min after the eccentric series and at the original L_o (Fig. 1A). Force was not significantly reduced after the isometric series (Fig. 1A). As a consequence of eccentric contractions, there was a shift in optimum length for force to longer muscle lengths which amounted to an increase of L_o by a factor of 1.24 ± 0.02. There was no significant shift in L_o after the isometric series (dashed line in Fig. 1A). This shift in L_o following eccentric contraction confirms earlier studies (Katz, 1939; Talbot & Morgan, 1998; Yeung et al. 2002). If the muscle was stretched to the post-eccentric L_o the decline in force was reduced and the developed force was 47 ± 3 % of the control conditions. Figure 1B illustrates the force-frequency relations of the fibres normalized to 100 Hz stimulation under each condition. As previously demonstrated, eccentrically damaged muscles have a steeper force-frequency relation (Jones et al. 1989; Balnave & Allen, 1995; Yeung et al. 2002). It can be seen that the pattern of force reduction was similar regardless of whether the fibre was at the original L_o or the post-eccentric L_o. As expected, the isometric series had no significant effect on the force-frequency relation (dashed line in Fig. 1B). These data establish that the muscle fibres show the previously established mechanical criteria of eccentric muscle damage and that fatigue due to the 10 isometric tetani was not a significant factor.

Figure 1. Changes in mechanical properties of muscle following eccentric contractions.

A, the length-force relations at various stages of the experimental protocol (n = 8). •, control; ▵, after the isometric series; and ○, after the eccentric series. The curves fitted to each set of data points are Gaussian distributions. B, force-frequency relations at various stages in the protocol. Each plot shows relative force normalized to 100 Hz force vs. different frequencies of stimulation (n = 8). ▵, control; •, post-isometric contractions; ○, post eccentric contractions at the original L_o; and ▾, after adjusting to new optimal length. * Statistically significant difference from control situation (P < 0.05).

Confocal images of dye-loaded muscle fibres

Under confocal microscopy, control experiments with no stimulation (n = 4) and those with isometric tetani series (n = 3) show regular, well-aligned fluorescent lines running transversely across the fibre (Fig. 2A). Only in the best resolution images (e.g. Fig. 2A and Fig. 3Aa) were we able to partly visualize two T-tubules per sarcomere as has previously been established in mammalian muscles (Lännergren et al. 2000). This inability to resolve two T-tubules per sarcomere probably results from the difficulty of dissecting the fibres sufficiently cleanly that the fibre lies very close to the coverslip (see discussion in Lännergren et al. 2000). In occasional images, longitudinal connections were seen spanning two to five transverse lines.

The most striking feature after 40 % stretch was the presence of vacuoles in six of the seven fibres examined (Fig. 2B and C). These vacuoles were present at our earliest observation, which was 10–20 min after the eccentric contractions. Vacuolation did not appear throughout the whole preparation, but was confined to a focal region of the fibre, so it is possible that they were present in the fibre in which we failed to identify vacuoles, but in a very restricted region. In the seven fibres, we counted the number of vacuoles in a 50 μm length of the fibre and a single plane of focus in the region where vacuoles were most prevalent. The number of vacuoles varied between 32 and 0 (11 ± 4). Because the vacuoles were filled with dye from the extracellular space, they must have had access to the extracellular space presumably through the T-system. The vacuoles varied in size and shape with some ovoid in cross-section running parallel to the length of the muscle fibre (Fig. 2B). In other experiments, the vacuoles were larger (up to 2 μm) and roughly spherical in shape (Fig. 2C). Sulforhodamine B was often applied several times and the vacuoles persisted for more than 2 h. They showed no sign of disappearing in contrast to vacuoles after fatiguing stimulation, which had largely resolved after 1 h (Lännergren et al. 2000). In the eccentrically damaged fibres, the T-tubules formed a wavy line across the fibre, the edges of the T-tubules were less clearly defined, and in some regions increased numbers of longitudinal connections were apparent. These features were also present in the one fibre in which no vacuoles could be located. There was no obvious correlation between of the presence or number of vacuoles and the magnitude of the force deficit; for instance the fibre in which no vacuoles were located had a roughly similar decline in force. In an attempt to examine whether vacuoles could develop during the eccentric stretches, but then seal over (Fraser et al. 1998), we performed the eccentric protocol in two fibres in the presence of sulforhodamine B. In these experiments the vacuoles appeared similar to other fibres; furthermore the dye was gradually lost from all vacuoles when extracellular dye was removed. Thus we found no evidence of sealed vacuoles that developed during the eccentric contraction. Milder stretches of 25 % of L_o did not lead to the production of vacuoles though force was reduced to 65 ± 10 % of control (n = 2).

Rate of fluorescence washout from muscle fibres

The time course of sulforhodamine B washout gives an estimate of the accessibility of the T-system to the extracellular space (Endo, 1966). Figure 3 illustrates representative images obtained during such experiments. Figure 3Aa is a control image in the presence of sulforhodamine B in a fibre that had the isometric series of contractions only. Figure 3Ab and c show images after 1 and 20 min washout. Note that the major part of the dye had left the T-tubules at 1 min, but faint T-tubules of roughly similar intensity are present in the images both at 1 min and 20 min. Figure 3Ba–d shows an equivalent series from an eccentrically damaged fibre. Note the prominent vacuoles in the image in the presence of sulforhodamine B (Fig. 3Ba) and that the T-tubules are less clear than in the equivalent control image. After a 1 min washout, the intensity of dye in the vacuoles was little changed, but most dye had left the T-tubules. After 20 min and even 2 h, vacuoles still contained some dye and faintly stained T-tubules were still detectable.

These issues are explored more quantitatively in the graphs shown in Fig. 4. In Fig. 4A the filled circles show the washout of the extracellular space, which was reduced to ∼5 % at 1 min. The open circles show the rate of decline of fluorescence in T-tubules in fibres following isometric tetani, which was roughly exponential with a half-time of 18 ± 1 s. As expected from the work of Endo (1966), the main decline lags behind the extracellular space by 10–20 s only, but there is an irreducible tail of fluorescence, which is about 20 % of the maximum and shows only a minor decline over 20 min. We are uncertain of the origin of this fluorescence, but the most likely possibility is that of a small fraction the dye, possibly caused by a small degree of lipid solubility, remains bound to the membranes of the T-system and dissociates very slowly. Following eccentric contractions, the T-system (vacuoles + T-tubules) (n = 7) showed significant slowing of washout rate when compared with those fibres after isometric contractions (P < 0.01). Note that in these data, which include both T-tubules and vacuoles, there seems to be two phases in the decline, with a rapid phase over the first minute and then a slower phase over at least 20 min.

To determine if the efflux rate of the dye differs from various structures, we first examined the decline of fluorescence in vacuoles (Fig. 4B, filled circles, n = 5). It is clear that the efflux rate is very slow and seems to follow a roughly exponential time course with a half-time of 6.3 ± 2.4 min. We also examined the efflux from T-tubules of fibres that had contracted eccentrically, but did not show vacuoles in the field of view (Fig. 4B, filled triangles, n = 4). These four fibres included the one eccentrically damaged fibre that did not show vacuoles, two fibres from experiments with 25 % stretch that also failed to exhibit vacuoles, and one fibre that showed vacuoles, but the measurements were made from an area that did not exhibit vacuoles. None of these data were significantly different from each other and they have been combined in Fig. 4B. The dye leaves the eccentrically damaged T-tubules much faster than vacuoles and the decline is roughly exponential with a half-time of 39 ± 4 s. Finally, in Fig. 4B the open circles show the decline of fluorescence from T-tubules of isometrically contracted fibres, which is clearly faster (same data as Fig. 4A).

These data also show that the steady-state T-tubule-fluorescence is greater in eccentrically stretched fibres than in control fibres. Possible explanations will be considered in the Discussion.

Effects of Na⁺,K⁺-ATPase inhibitor on the formation of vacuoles

Casademont et al. (1988) showed that gross damage to muscle fibres caused the production of vacuoles which could be inhibited with 1 mm ouabain. We therefore sought to determine whether the production of vacuoles following eccentric damage was also inhibited by ouabain. In three fibres the tetanic contraction and the force- frequency relation were not significantly affected by the addition of 1 mm ouabain and the decline of force following eccentric contractions in the presence of ouabain was to 28 ± 8 % of the control (not significantly different from the eccentric protocol in the absence of ouabain). In these three fibres exposed to ouabain throughout the eccentric protocol, no vacuoles could be detected for a period of 30 min after the eccentric protocol (Fig. 5Ba). This number of vacuoles is significantly smaller than in the seven fibres studied in the absence of ouabain (Mann-Whitney rank sum test, P < 0.05). The ouabain was then washed out and after 45–60 min small vacuoles began to appear (Fig. 5Bb and c). The extent of vacuolation progressively increased over the next 60 min (Fig. 5Bd). Bright fluorescent vacuoles appeared in large numbers and mostly formed along the longitudinal axis in all three fibres observed. In one experiment, ouabain was reapplied in an attempt to examine whether the vacuoles were reversible. However in this fibre vacuoles persisted for another 60 min of observation, suggesting that once formed, the vacuoles were stable and did not require a continuous influx of Na⁺ and accompanying H₂O for their maintenance. Experiments on two fibres subjected to isometric contractions did not show any vacuoles either in the presence of ouabain (Fig. 5Aa) or after the removal of ouabain (Fig. 5Ab)).

DISCUSSION

In this study we used an established model of eccentric damage to single mouse muscle fibres (Balnave & Allen, 1995; Yeung et al. 2002). We used a stretch of 40 % L_o, which is within the physiological range (Brooks et al. 1995). The fibres remained excitable and produced force, but showed three characteristic signs of eccentric damage: (i) reduced force, (ii) a shift in the peak of the force–length relation (Wood et al. 1993; Talbot & Morgan, 1998; Yeung et al. 2002) and (iii) reduced force at low frequencies of stimulation compared to high frequencies (Jones et al. 1989; Balnave & Allen, 1995; Yeung et al. 2002). The design of the study established that these changes were a consequence of the eccentric component of the contraction and not simply the fatigue that might result from repeated isometric tetani. The main novel findings of this study were that eccentric muscle damage caused the formation of vacuoles connected to the T-tubules and that the diffusion of molecules into and out of both the T-system and the vacuoles was slowed. We also showed that the formation of vacuoles was dependent on the activity of the sodium pump. These new results provide a more functional understanding of the damage to T-tubules and extend the structural findings of a recent EM study (Takekura et al. 2001).

The study by Takekura et al. (2001) involved rats that performed a series of downhill runs over 90 min. The animals were killed at intervals up to 10 days after the termination of exercise. Identified fast or slow fibres were studied with high voltage EM and a staining technique that identified T-tubules or SR. Four changes in the T-system were identified: (i) increase in the longitudinal elements of the T-tubules, (ii) changes in the organization of triads, (iii) the development of calveolar clusters and (iv) the appearance of multiple connections between two and three T-tubules and three and four terminal cisternae (pentads and heptads). The increase in longitudinal T-tubules was greatest at 3 days and, although we used objective methods to search for them in our fibres, we could not detect increased longitudinal T-tubules in our time frame (0.5–2 h). The calveolar clusters are interesting and were apparent immediately after exercise but only in slow fibres. Nevertheless it seems possible that some vacuoles and caveolar clusters might be similar or identical structures.

Mechanism of production of vacuoles in skeletal muscle

Vacuoles have previously been described in a range of situations including glycerol removal (Krolenko et al. 1998; Khan et al. 2000), muscle fatigue (Lännergren et al. 2000) and gross membrane damage (Casademont et al. 1988). An extensive review of the mechanism of production and the significance of skeletal muscle vacuoles has recently been published (Krolenko & Lucy, 2001). However vacuoles have not previously been described as part of eccentric damage, although with the benefit of hindsight, structures that might be vacuoles are visible in the study by Warrren et al. (1995) (see their Fig. 6).

Figure 6. Hypothesis for mechanism and consequences of T-tubular rupture in eccentric muscle damage.

A, diagram illustrating sarcomere inhomogeneities in an eccentrically damaged fibre. Sarcomeres 1 and 4 are of normal length, sarcomere 2 is shorter than average while sarcomere 3 is longer. T-tubules in mammalian muscle are at the overlap of thick and thin filaments and are assumed to be subject to shearing stress (indicated by dashed lines). Where the shear stress is greatest (between sarcomeres 2 and 3), T-tubules are assumed to rupture. B, illustration of some hypotheses of the consequences of eccentric damage following rupture of T-tubules; AP, action potential. See text for further discussion.

Lännergren et al. (2000) showed that vacuoles could develop during the recovery from repeated isometric tetani in amphibian muscle fibres. Thus it is important to establish that the vacuoles we observed in the present experiments were a consequence of the eccentric damage and not caused by some aspect of the repeated tetani. Firstly, like Lännergren et al. (2000), we did not observe vacuoles in mouse fibres after repeated isometric tetani. Secondly, vacuoles were localized following eccentric damage, whereas they were widespread in fatigued amphibian fibres. Thirdly, vacuoles in association with fatigue disappeared over the course of an hour or so, whereas those associated with eccentric damage seemed quite stable over several hours. Thus the vacuoles described in the present study have different characteristics to those observed after repeated contractions. It seems that the vacuoles we observed are specific to eccentric damage.

In all situations where vacuoles have been described, the muscle is subject to an abnormal osmotic load (Krolenko & Lucy, 2001). For instance when glycerol is removed from the extracellular space after a period of exposure, the muscle swells as water enters to equilibrate intracellular and extracellular osmolarity. Subsequently the fibre slowly returns to its normal volume as glycerol and accompanying water leave the cell. Vacuoles (and detubulation) occur during this latter phase and it is thought that the volume load of water and glycerol leaving the cell via the T-tubules exceeds their capacity causing them to swell and form vacuoles. Similarly, during fatigue it is postulated that the load of lactic acid and accompanying water leaving the cell in the recovery period causes vacuole development (Lännergren et al. 2000). It is less clear whether the volume load simply exceeds the normal flow of the T-system or whether the osmotic forces also lead to compression of some T-tubules and reduce their fluid transport capability.

What might cause an osmotic load in association with eccentric muscle damage? Key observations were made by Casademont et al. (1988) who showed that when a muscle was cut in half, very large numbers of vacuoles developed that were attached to the T-tubules close to the cut region. They showed that vacuole production could be limited by a range of procedures including removing Na⁺ from the extracellular fluid and addition of a sodium pump inhibitor. They argued that following surface membrane damage, Na⁺ entered the myoplasm from the extracellular space. This stimulated the Na⁺ pump which would then pump the excess Na⁺ into the extracellular space of the T-tubules. The volume flow of H₂O which accompanies the Na⁺ was proposed to exceed the flow capability of the T-tubules and to cause the vacuole production. We propose that in eccentric damage, T-tubules suffer shearing damage when adjacent myofilaments show disparate degrees of stretch (Fig. 6A). Once disruption of a T-tubule occurs there will be leakage of extracellular contents into the myoplasm and vice versa. One important consequence of this process will be an increase in [Na⁺]_i and [Ca²⁺]_i close to the end of the disrupted T-tubule. Presumably disrupted T-tubules either reconnect or seal over and remain disconnected; however, we have no information on the time course or frequency of these two processes. We propose that this localized increase in [Na⁺]_i and [Ca²⁺]_i initiates two consequences of the initial T-tubular disruption as indicated in Fig. 6B.

The main evidence in favour of our proposal is extensive earlier work by Casademont et al. (1988) in a situation of indisputable membrane damage and our demonstration that ouabain can completely prevent the development of vacuoles. It might be argued that the failure to observe vacuoles in ouabain is simply due to chance, since not all eccentrically stretched fibres exhibited vacuoles, but this explanation is eliminated by the subsequent observation of extensive vacuole development as the ouabain was washed out of the preparation. The increased frequency of vacuole development after ouabain is consistent with an increase in intracellular sodium when the Na⁺ pump is inhibited. The slow time course of the appearance of vacuoles as ouabain is washed off (which is slower than the appearance of vacuoles after eccentric contractions) is consistent with the slow washout of ouabain from muscles (Nielsen & Clausen, 1996). The gradual increase in background fibre fluorescence as ouabain is washed off (compare Fig. 5Bb to d) probably represents a slow increase in the number of vacuoles and in the number of blocked T-tubules so that sulforhodamine leaves the fibre more slowly with each subsequent washout.

Our hypothesis for the formation of vacuoles requires the presence of Na⁺ pumps in the T-tubules. Early studies showed that Na⁺ pumps are present at a reduced density in the T-tubules (Venosa & Horowicz, 1981). More recent studies have confirmed the original observation and, in addition, demonstrate that the distribution of Na⁺ pump isoforms is different in the T-tubule compared to the surface membrane (Williams et al. 2001).

Slowed diffusion in and out of the T-system

It is established that molecules enter and leave the T-system with a time course that is only moderately greater than that expected for free diffusion (Hodgkin & Horowicz, 1960). In keeping with this, sulforhodamine B enters and leaves the T-system with a half-time of between 5 and 20 s depending on the diameter of the fibre (Endo, 1996); we confirmed this finding in normal fibres which had a half-time for removal of sulforhodamine B of 18 ± 1 s. In contrast, removal of sulforhodamine B from the vacuoles is greatly slowed with a half-time of around 6 min. It is clear from these studies that a substantial barrier to diffusion is located at the point of connection of the vacuoles with the T-system.

Of special interest is our observation that, in eccentrically damaged fibres that did not exhibit vacuoles, the rate of diffusion of sulforhodamine from the preparation was significantly slowed to 39 ± 4 s. This is important because it establishes that there is some change in T-tubular properties in eccentrically damaged fibres even in the absence of vacuoles. Furthermore it suggests that at least some of the barrier to diffusion lies distributed in the T-system. One possibility is that the T-system is compressed by the osmotic changes reducing both volume flow and diffusion. Another possibility is that the rupture of T-tubules and subsequent sealing over make the diffusion pathway more restricted and/or more tortuous.

Another interesting observation is the presence of fluorescence in the T-tubules of normal fibres after 20 min and the increase in this intensity in eccentrically damaged fibres. We argued earlier that the fluorescence in normal fibres could represent binding of the dye to the membrane. The increased fluorescence following eccentric damage could arise (i) because the surface area of T-system has increased, (ii) because the diffusion pathway is more tortuous, particularly if many T-tubules near the surface membrane were disrupted, (iii) because some T-tubules have formed sealed sections from which the dye cannot escape or (iv) because some dye has entered the myoplasm and is trapped there. Some of these possibilities could be distinguished if higher resolution images of T-tubular staining were available or EM studies of the T-tubules at the appropriate time were made.

Consequences of T-tubular damage for eccentric muscle damage

The observation of vacuoles is the most compelling evidence of T-tubular damage, but do vacuoles necessarily affect muscle function? In muscle fatigue, vacuoles are a prominent feature of the recovery period in amphibian fibres, but they correlate poorly with the degree of functional recovery (Lännergren et al. 2000). Similarly in our experiments all the features of eccentric damage were present in 1/7 fibres subjected to 40 % stretch and 2/2 fibres subjected to 25 % stretch and yet no vacuoles could be found. In addition, the distribution of vacuoles was often quite localized while the sarcomere inhomogeneities and ionic changes appeared to be much more widely distributed (Balnave et al. 1997). Furthermore, the degree of reduction of force was similar in the experiments in the presence of ouabain despite the fact that no vacuoles were detected. These features lead us to the view that the fundamental T-tubular defect is not the presence of vacuoles, but a more minor degree of damage to the T-tubules that is not detectable with our present images. However we do have evidence for this abnormality based on the slower diffusion of sulforhodamine B from the T-tubules of eccentrically damaged fibres that did not exhibit vacuoles. Thus we propose that shearing damage to T-tubules at multiple sites in the muscle, triggered by the sarcomere inhomogeneities, is the primary T-tubular defect (Fig. 6A). Vacuoles presumably develop in sites where this primary damage is so great that the osmotic load of pumping out the consequent rise in Na⁺ is sufficiently great to produce vacuoles (central column of Fig. 6B).

Figure 6B illustrates how we propose that the shearing damage to T-tubules contributes to other aspects of eccentric damage. The left hand pathway illustrates some of the consequences of Ca²⁺ influx from the extracellular space. A rise in resting [Ca²⁺]_i has been noted in several studies (Balnave & Allen, 1995; Ingalls et al. 1998) though focal rises, which would be predicted by our hypothesis, have not been detected (Balnave et al. 1997), perhaps because they are very transient. Several studies have shown that a rise of the time-averaged [Ca²⁺]_i in a muscle can inhibit SR Ca²⁺ release (Lamb et al. 1995; Bruton et al. 1996; Chin & Allen, 1996) so the rise in resting [Ca²⁺]_i could conceivably cause the reduced Ca²⁺ transients that have been observed in eccentrically damaged muscles (Balnave & Allen, 1995; Ingalls et al. 1998). It is also widely thought that rises in [Ca²⁺]_i may contribute to activation of proteases and phospholipases and to the ensuing cell damage and inflammation (Belcastro et al. 1998).

The right hand pathway illustrates some of the possible consequences of the ruptured or compressed T-tubules. It is possible that inward conduction of the action potential would be affected and this is another possible mechanism for the reduced SR Ca²⁺ release, which is characteristic of eccentrically damaged mammalian muscle (references above). This component of damage could be the cause of the slowed diffusion of sulforhodamine B, and if removal of protons is also slowed, it may explain the impaired pH regulation that we have recently described (Yeung et al. 2002).

Conclusion

These observations of damage and functional changes in the T-tubular system offer new insights into some of the early changes in eccentric damage. The formation of vacuoles, which can be inhibited by ouabain, is strong evidence of intracellular Na⁺ loading, presumably through damaged T-tubules. The ionic changes secondary to T-tubular damage and the reduced exchange of ions, metabolites and fluid across the T-tubular network are capable of explaining a range of phenomena that occur in eccentrically damaged muscle.