In vitro studies have used protein markers to distinguish between myogenic cells isolated from fast and slow skeletal muscles. The protein markers provide some support for the hypothesis that satellite cells from fast and slow muscles are different, but the data are equivocal. To test this hypothesis directly, three-dimensional skeletal muscle constructs were engineered from myogenic cells isolated from fast tibialis anterior (TA) and slow soleus (SOL) muscles of rats and functionality was tested. Time to peak twitch tension (TPT) and half relaxation time (RT1/2) were ∼30% slower in constructs from the SOL. The slower contraction and relaxation times for the SOL constructs resulted in left shift of the force-frequency curve compared with those from the TA. Western blot analysis showed a 60% greater quantity of fast myosin heavy chain in the TA constructs. 14 days of chronic low-frequency electrical stimulation resulted in a 15% slower TPT and a 14% slower RT1/2, but no change in absolute force production in the TA constructs. In SOL constructs, slow electrical stimulation resulted in an 80% increase in absolute force production with no change in TPT or RT1/2. The addition of cyclosporine A did not prevent the increase in force in SOL constructs after chronic low-frequency electrical stimulation, suggesting that calcineurin is not responsible for the increase in force. We conclude that myogenic cells associated with a slow muscle are imprinted to produce muscle that contracts and relaxes slowly and that calcineurin activity cannot explain the response to a slow pattern of electrical stimulation.
- tissue engineering
- electrical stimulation
- engineered muscle
skeletal muscle is a highly plastic tissue that changes its phenotype in response to changes in loading and recruitment (18, 19). One model that has been used to experimentally alter muscle phenotype is chronic low-frequency electrical stimulation (CLFS). A tonic stimulus that mimics the impulse pattern of a slow motor neuron slows the speed of both contraction and relaxation in fast-twitch muscles (36). The importance of electrical activity on the phenotype of skeletal muscle was reinforced when Salmons and Sréter (35) showed that the slow-to-fast muscle transition after cross re-innervation of the soleus with the perineal nerve could be reversed by CLFS. This experiment showed conclusively that the role of the nerve in determining muscle phenotype was largely dependent on activation patterns that could be emulated by electrical stimulation, and not chemical signals derived from the nerve. Since then, the CLFS-induced transition has been shown to result in coordinated changes in the isoforms of the contractile (5), regulatory (22, 34, 37), and Ca2+-sequestering proteins (20, 32), as well as inducing a concomitant mitochondrial biogenesis (38).
Although muscle is highly plastic, it is generally accepted that the myoblasts within a muscle are patterned to reflect the phenotype of that muscle (33). This assumption has been made on the basis of the expression of protein markers that are thought to represent the phenotype of muscle, such as myosin heavy chain (MHC), myosin light chain (MLC), and myogenic regulatory factors (see Table 1). However, depending on the protein analyzed, the culture conditions, and the organism studied, this difference has not always been observed (3, 4, 12–14, 30).
Regardless of whether there are differences in the expression of a single protein, the rate of contraction and relaxation of muscle is the result of numerous protein systems, Ca2+ release, regulatory proteins (MLC, troponins, and tropomyosin), MHC, and Ca2+ sequestering, all working together to produce a faster muscle. Therefore, analyzing a single protein or even representative proteins could produce erroneous conclusions. We have recently reported a model system that permits the functional analysis of skeletal muscle engineered from isolated cells (24). Using this model, we tested the hypothesis that three-dimensional (3-D)-engineered muscles generated from myogenic cells isolated from a slow muscle (soleus) would contract and relax more slowly than those generated from a fast muscle [tibialis anterior (TA)] and that electrical stimulation would modulate the contractility of both engineered constructs similarly.
The data presented here show that functional muscles engineered from myogenic cells of the slow soleus muscle contract and relax 30% slower than similar tissues engineered from the fast TA muscle. Two weeks of chronic low-frequency electrical stimulation resulted in a slowing of time-to-peak tension (TPT) and half relaxation time (RT1/2) without altering force production in TA constructs, whereas in the SOL constructs there was no change in the rate of contraction and relaxation but force increased twofold. Fast-patterned electrical activity had no effect on any of the constructs. To evaluate the potential role of calcineurin in the increase in SOL construct force production, constructs were stimulated in the presence or absence of cyclosporine A (CsA). CsA treatment resulted in a 28% faster TPT without altering the RT1/2 of the SOL constructs. Treatment with CsA alone increased force production, and when combined with CLFS, augmented the increase in force production induced by stimulation. These data suggest that, in rats, myogenic cells from a slow muscle are imprinted not only to produce muscle that contracts and relaxes more slowly but also responds to chronic low-frequency electrical stimulation differently than myogenic cells from fast muscle.
MATERIALS AND METHODS
The SDS-PAGE gels were from Cambrex Bioscience (Rockland, ME). The horseradish peroxidase-conjugated secondary antibodies and WestDura chemiluminescent reagents were purchased from Pierce (Rockford, IL). Polydimethylsiloxane (PDMS) (Sylgard; type 184 silicone elastomer) was purchased from Dow Chemical (Midland, MI). Antibodies raised against eukaryotic initiation factor 2 were from cell signaling (Danvers, MA). The total (MF 20) and the type II (F59) MHC antibodies were developed by Dr. D. A. Fischman and F. E. Stockdale, respectively, and obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the National Institute of Child Health and Human Development and maintained by the University of Iowa's Department of Biological Sciences. All other reagents were from Sigma (St. Louis, MO).
Isolation of myogenic cells from muscle.
Myoblasts were isolated as described previously (24). Briefly, hindlimb muscles (soleus and TA) from Sprague-Dawley rats were dissected and washed with PBS three to four times to remove hair and other debris. The muscles were cut into small pieces and cleaned of excess connective tissues and tendons. Dissociation of muscles was carried out in 10 ml of digestion buffer (F-12 medium containing 0.1% collagenase and 0.05% dispase) per 100 mg tissue. The digested tissues were filtered through a 100-μm cell strainer (Becton Dickinson, Franklin Lakes, NJ) and centrifuged at 1,500 g for 6 min. The supernatant was discarded, and the cell pellet was resuspended in growth medium (10% heat-inactivated FBS and 5 ng/ml FGF-2 in F-12 or F-12K medium with 100 U/100 mg/ml penicillin-streptomycin and 2.5 μg/ml fungizone). After being preplated overnight, the myoblast-enriched supernatant was transferred to a new plate and expanded for 5–6 days.
Formation of 3-D engineered skeletal muscle.
3-D muscles were engineered as described previously (24). Briefly, two 1.0-gauge sutures were secured to a Sylgard-coated 35-mm plate. To form the fibrin gel, the plates were coated with 700 μl of growth media containing 5 units of thrombin and 4 mg of fibrinogen. One hundred thousand cells were added to the top of the gel and the plates were placed in a standard incubator (37°C and 5% CO2). After 14 days in culture, the 3-D muscle constructs had completely formed and were ready for stimulation.
Determination of engineered muscle contractility.
All contractile properties were measured 28 days after the cells were plated. The temperature of the engineered muscles was maintained at 37 ± 1°C during the measurement of contractile properties using a heated aluminum platform. For measurements of contractility, one of the sutures was freed from the PDMS substrate and a force transducer was attached to its minutien pin with the use of canning wax. The custom-built force transducer has a resolution range of 1 to 2,000 μN and a sampling rate of 1 kS/s (10), allowing the determination of TPT and RT1/2. Constructs were stimulated three times with either a single twitch pulse of 4 ms at 15 V or a 1-s tetanus at 5, 10, 20, 40, 60, 80, 100, and 150 Hz 1.2-ms pulses at 15 V and the average force recorded over the three trials was determined. Baseline force was measured as the average baseline passive force preceding the onset of stimulation.
Determination of MHC protein levels.
Four weeks after being plated, SOL and TA constructs were frozen, sonicated in 100 μl of 1× Laemmli sample buffer, and boiled for 10 min to denature and free intracellular proteins. Protein concentration was determined using the noninterfering protein assay (Upstate, Waltham, MA) and equal aliquots of protein were separated by SDS-PAGE, transferred to nitrocellulose, and blocked in 5% milk. Blots were exposed to primary antibodies (eEF2 at 1:1,000; MF20 and F59 at 1:10 of unpurified hybridoma supernatant) overnight at 4°C and the bound antibody was detected chemiluminescently with the use of WestDura chemiluminescent reagent and a Syngene Chemigenius 2 gel documentation system.
Electrical stimulation of engineered muscles.
SOL and TA constructs were electrically stimulated for 14 days starting 2 wk after being seeded using a protocol that mimicked either fast or slow motor neuron activity. The slow protocol delivered 5 pulses at 20 Hz every 4 s, whereas the fast protocol delivered 5 pulses at 100 Hz every 100 s. Both protocols used the same voltage (5 V) and pulse width (1.5 ms).
Control and chronically stimulated SOL constructs were treated with 500 ng/ml CsA for 2 wk. At the end of the 2-wk incubation period, the constructs were stimulated as described above.
Data are presented as means ± SE for 4 to 6 engineered muscle constructs per group. Differences in mean values were compared within groups and significant differences were determined by ANOVA with post hoc Tukey-Kramer's honestly significant difference test. The level of significance was set at P < 0.05.
Isolation of myoblasts from fast and slow muscle.
Myoblasts isolated from both the TA and SOL muscles grew well in culture. Digestion of the soleus muscle tended to provide more myoblasts per milligram of muscle than the TA muscle as has been reported previously (16). After expansion, equal numbers of myoblasts isolated from the SOL or TA were seeded on the fibrin gels.
Contractility of the engineered muscles.
To determine whether the cells isolated from the soleus muscle were imprinted to be slower than those of the TA, the rate of contraction and relaxation of the engineered tissues was determined. The TPT and RT1/2 of the SOL constructs were 30% and 31% slower than the TA constructs, respectively (SOL construct TPT = 52 ± 0.3 RT1/2 = 45.3 ± 2.3; TA construct TPT = 39.8 ± 0.5 ms RT1/2 = 35.3 ± 1.7; Fig. 1A). As a result of the slower twitch kinetics, the force frequency curve of the SOL constructs was shifted to the left compared with the TA constructs (Fig. 1B). The SOL constructs had a significantly higher percentage of maximal force production from 20 to 60 Hz.
MHC isoform expression.
To assess whether the rate of contraction reflected differences in the contractile proteins expressed by the constructs, the level of fast MHC was determined by Western blot analysis (Fig. 2). With the use of an antibody to detect all forms of the fast MHC, the TA constructs had 57% more fast MHC than the SOL constructs (TA = 580 ± 13.0, SOL = 370 ± 35.9).
Stimulation alters the function of engineered muscle.
Both SOL and TA constructs were stimulated with either a slow or a fast paradigm of electrical activity (see materials and methods) to determine the effect on contractile function. Two weeks of slow electrical activity resulted in a 14.5% longer TPT and a 13.5% longer RT1/2 in the TA constructs (Fig. 3B). This increase in TPT and RT1/2 was not accompanied by a change in the amount of force produced by the constructs (Fig. 4B). In contrast, the TPT and RT1/2 of the SOL constructs was unaffected by CLFS (Fig. 3A) but this paradigm of stimulation increased the force produced by the SOL constructs by 80.4% (Fig. 4A). The increase in force production after CLFS was not the result of an increase in MHC in the stimulated SOL constructs. Total MHC, as determined by Western blot analysis, showed a trend toward decreasing (49.2% and 53.6%, respectively, of control) in two separate experiments. The change in MHC did not reflect a change in the number of cells in the constructs because the amount of elongation factor 2 was unchanged after stimulation. The fast paradigm of electrical activity had no effect on the force production or the contractile dynamics of either the SOL or TA constructs (Figs. 3 and 4).
Role of calcineurin in the increase in SOL construct force after CLFS.
Because calcineurin may play a role both in the control of muscle fiber type and muscle hypertrophy, we sought to determine whether the increase in force production in the SOL constructs after CLFS was the result of the activation of calcineurin in the SOL constructs. Control and chronically stimulated SOL constructs were treated with 500 ng/ml CsA for 2 wk, in an effort to prevent calcineurin activation, and the effect on contractile dynamics and force production were determined. Two weeks of CsA treatment alone increased force production 230% (CTL = 203 ± 31.4 μN; CsA = 471 ± 71.0 μN; Fig. 5). As before, CLFS increased force production (326 ± 201.0 μN). The combination of CsA and CLFS resulted in an additive effect on force production to a level 339% of control (690 ± 85.1 μN). Unlike electrical stimulation, CsA treatment increased total MHC within the SOL constructs (Fig. 6). This increase in MHC reflects not only an increase in the total amount of MHC but also a 3.4-fold increase in the percentage of the fast isoform. Concurrent treatment with CsA and CLFS had the greatest increase in total MHC (9.5-fold). However, the increase in the proportion of fast MHC was prevented by CLFS.
Role of calcineurin in altering TPT and RT1/2.
Determination of TPT and RT1/2 showed that both the CsA alone and the CsA with CLFS groups showed a 25% decrease in TPT (CTL = 52.7 ± 1.8 ms; CsA = 39.6 ± 1.2 ms; CsA + CLFS = 39.3 ± 0.6 ms; Fig. 7), whereas there was no change in TPT with CLFS alone (53.6 ± 2.7 ms). In contrast to TPT, RT1/2 was unchanged for any of the groups studied (CTL = 45.9 ± 1.8 ms; CsA = 41.8 ± 3.6 ms; CLFS = 48.5 ± 6.7 ms; CsA + CLFS = 39.9 ± 2.4 ms).
This is the first demonstration that myogenic cells isolated from a slow muscle are functionally distinct from those isolated from a fast muscle, contracting and relaxing 30% slower. The differences between myogenic cells are much greater than simply differential expression of MHC, MLC, or myogenic regulatory factors, as previously described in the rodent (6, 7, 12, 26, 33, 39). Not only are the rates of contraction and relaxation different, but the engineered muscles also show a differential response to electrical stimulation. This indicates that the plasticity of rat muscle is dependent on a long-term reprogramming of the composite cells. The first essential steps that are required for this reprogramming remain to be identified.
Before discussing the significance of the findings, we must first discuss one issue with the techniques. The variability in the basal measurements from experiment to experiment reflects the fact that each experiment is generated from a separate isolated muscle cell population. As a result, there may be differences in the total number of cells in the constructs and in the ratio of myoblasts to fibroblasts from experiment to experiment. For this reason, all of the experiments have been repeated at least three times and the percentage of change among the trials is equivalent (compare Fig. 4A slow stimulation with Fig. 5 slow stimulation; each shows an 87% increase in force).
As was expected from the studies by Salmons and Vrbova (36) and Wehrle et al. (39), electrical stimulation altered the contractility of the muscle constructs. Unexpectedly, the changes in contractility depended on the muscle used to generate the constructs. CLFS slowed the rate of both contraction and relaxation in the TA constructs without affecting force production. In contrast, neither the TPT nor the RT1/2 was affected in the SOL constructs. Instead, CLFS of the SOL constructs resulted in a 61–80% increase in force production without an increase in total MHC. The shift in dynamics in the TA constructs is likely due to the shift in MHC and sarco(endo)plasmic reticulum Ca2+ ATPase to a slower isoform, as has been reported both in vivo (21) and in vitro (39). The increase in force production in the SOL constructs is more difficult to explain and could not have been observed in the 2-D culture systems that have previously been used. The increase in force in the stimulated SOL constructs may reflect the fact that, to slow myogenic cells, chronic low-frequency stimulation promotes the reorganization of sarcomeres within the SOL constructs resulting in greater force production with the same or fewer myosin molecules or that CLFS results in an alteration in the matrix produced by the SOL muscle cells allowing for better transmission of the force to the anchors. We initially thought that hypertrophy was occurring within the SOL constructs and that this was due to calcineurin, which has been hypothesized to be involved in both the development of slow fiber phenotype and muscle growth (9, 31). However, treatment with CsA, which should inhibit calcineurin, did not prevent the increase in force production in the stimulated SOL constructs.
In contrast to the effects of CLFS, stimulating the constructs with a paradigm modeled after a fast nerve did not change contractility. This likely reflects the fact that the fast paradigm of stimulation was not appropriate for ex vivo stimulation. Previous work from our laboratory (11) has shown that the maintenance of the mass and force production of a denervated fast muscle, the extensor digitorum longus in vivo requires between 200 and 800 contractions a day. Greater than 800 contractions a day resulted in a loss of muscle mass and force production. In this study the fast protocol delivered one contraction every 100 s. This amounts to >850 contractions per day for an isolated engineered muscle construct. Although this pattern may be within the normal range for fast twitch muscle in vivo, it is likely that in culture this pattern was inappropriate. The fast protocol of electrical stimulation resulted in the loss of force within 24 h (data not shown). This is likely the result of electrically induced tissue damage. The effect of excessive electrical stimulation was also observed in slow protocols of electrical stimulation. The CLFS protocol adopted for this study (250-ms train duration every 4 s) produced only 6% active time compared with 25∼30% active time in the adult soleus (23). In preliminary experiments, doubling the train duration to 500 ms (bringing the active time to one-half that of the adult soleus) induced a decrease in force production within a couple of days. This suggests that electrical stimulation ex vivo does not precisely emulate the stimulation of muscle in vivo. Electrical stimulation is potentially destructive, both in vivo and in vitro, due to hydrolytic degradation of the fluids and tissues surrounding the electrodes, electrode breakdown, and the large field of stimulation compared with the smaller area directly involved in an intact neuromuscular junction. Overall, the greater energy flux associated with electrical stimulation means that activity patterns derived from normal nerve-muscle activity may be more than the tissue can endure via field stimulation, resulting in collateral tissue damage (11). Therefore, for both protocols used in this study the amount of electrical stimulation still needs to be optimized further in terms of frequency, pulse width, and train duration to minimize tissue damage and maximize phenotypic alterations.
The data also suggest that calcineurin has a direct effect on the rate of muscle contraction but has limited effect on the rate on relaxation. Our data are consistent with others that have shown that calcineurin directly alters the expression of MHC away from slow type I myosin (9, 29) toward the fast type II isoform (8, 25). In the data presented here, the shift toward a fast muscle is reflected not only in an increase in fast MHC levels, but for the first time both a decrease in the TPT and an increase in the force of contraction were observed in cell culture. Interestingly, there was no change in the rate of relaxation of the SOL constructs when they were incubated with CsA, despite the decrease in the TPT. This suggests that calcineurin has no direct effect on the expression of the calcium sequestering machinery and that the fast-to-slow transformation that occurs in vivo (8) requires other factors, such as thyroid or other circulating hormones. A similar effect has been seen in regenerating soleus muscle where inhibition of calcineurin in isolated fibers prevented the accumulation of type I MHC but had no effect on the expression of the slow sarco(endo)plasmic reticulum Ca2+ ATPase pump (40).
The effects of CsA on MHC expression are more difficult to explain. The increase in total MHC after treatment with CsA suggests that a target of CsA, possibly calcineurin, inhibits MHC expression. It is noteworthy that the increase in MHC in the CsA-treated constructs was primarily the result of an increase in the fast isoform of the protein. This is consistent with the in vivo work of Giger et al. (17), who showed that 9 days of CsA treatment increased the expression of both the type IIa and IIx isoforms. In contrast, Meissner and colleagues did not see any change in the mRNA of either type IIa or IIx MHC in rabbit primary bead cultures treated with CsA for 14 days. The difference between the studies likely reflects differences between the models. The engineered model described here is in agreement with the data collected in vivo and suggests that a CsA-inhibited process blocks normal expression of fast MHC isoforms.
The fact that CsA was able to drive the expression of fast MHC was reduced by slow electrical stimulation in the SOL constructs may indicate that slow electrical activity is dominant over CsA in determining MHC isoform or that the amount of CsA used was not sufficient to block calcineurin, or other downstream targets, when CLFS was added. It is interesting to note that even though the proportion of fast MHC was not increased in these constructs the TPT was faster. This likely reflects the fact that there is an increase in the total amount of fast MHC in these constructs. In small muscles like those engineered here, any increase in the amount of fast myosin may be sufficient to improve TPT.
It is clear from the data presented here that the development of a muscle and the functional determination of what constitutes a fast muscle cell is more complex than simply the expression of MHC. The identification of why SOL constructs increase force production while decreasing MHC in response to CLFS and why CsA treatment and CLFS increase total MHC more than just CsA treatment alone may give us some insight into how adult muscle fibers develop and determine skeletal muscle phenotype.
This work was supported by a grant from the United States Defense Advanced Research Projects Agency (to R. G. Dennis) and by Navy Contract N66001-02-C-8034 (to K. Baar).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society