In a series of experiments, cultured myotubes were exposed to passive stretch or pharmacological agents that block contractile activation. Under these experimental conditions, the formation of Z lines and A bands (morphological structures, resulting from the specific structural alignment of sarcomeric proteins, necessary for contraction) was assessed by immunofluorescence. The addition of an antagonist of the voltage-gated Na+ channels [tetrodotoxin (TTX)] for 2 days in developing rat myotube cultures led to a nearly total absence of Z lines and A bands. When contractile activation was allowed to resume for 2 days, the Z lines and A bands reappeared in a significant way. The appearance of Z lines or A bands could not be inhibited nor facilitated by the application of a uniaxial passive stretch. Electrical stimulation of the cultures increased sarcomere assembly significantly. Antagonists of L-type Ca2+ channels (verapamil, nifedipine) combined with electrical stimulation led to the absence of Z lines and A bands to the same degree as the TTX treatment. Western blot analysis did not show a major change in the amount of sarcomeric α-actinin nor a shift in myosin heavy chain phenotype as a result of a 2-day passive stretch or TTX treatment. Results of experiments suggest that temporal Ca2+ transients play an important factor in the assembly and maintenance of sarcomeric structures during muscle fiber development.
- Z lines
- A bands
muscle fibers change their myosin heavy chain (MHC) phenotype in response to a number of external perturbations, such as increased or decreased usage (5), electrical stimulation (37), denervation (44), anabolic steroid combined with exercise (36), or thyroid hormone (30). Although in vivo experiments can explore long-term changes, they make it difficult to dissect the individual contributions of mechanical loading or neural recruitment on muscle fibers. A cell culture approach allows the manipulation of one variable at a time, and the use of a developmental model may assist in the elucidation of mechanisms, which can then be applied in adult muscle fibers.
Contractile activation has an effect on a number of molecules, for example, acetylcholinesterase (45), L-type Ca2+ channels (18), and glucose transporter 4 (25), and it sets off a series of second messenger cascades (43). The role of mechanical factors in cellular development is not well established, but it is receiving more attention (27, 47). The independent contribution of mechanical factors and neural factors in myofibrillogenesis remains to be established. In cultured myosatellite cells, long-term electrical stimulation induces a slower contractile phenotype (53) and neural input is a requirement for specific stages of myogenesis (15). The molecular mechanisms responsible for the phenotypic changes as well as the mechanisms for the subsequent sarcomere assembly remain to be resolved. Sarcomere assembly is likely to entail several mechanisms, such as that illustrated by the organization of α-actinin in the formation of premyofibrils (13) or by the self-assembly of myosin into thick filaments before they are inserted into sarcomeres (3,26).
To study the separate contributions of mechanical stretch and contractile activation on sarcomere assembly, I used a mammalian cell culture model coupled with the controlled application of passive stretch or electrical stimulation and evaluated myofibrillogenesis. My findings suggest that contractile activation, and not mechanical stretch, is important for myofibrillogenesis. Moreover, low-frequency electrical stimulation and its subsequent increased intracellular flux of Na+ and especially Ca+ transients are required for the assembly and maintenance of the Z lines and A bands.
MATERIALS AND METHODS
Muscle cells were isolated from the hindlimbs of neonatal Sprague-Dawley rats. For each individual experiment, the muscles from the hindlimbs of one litter (10 pups) were dissected, freed from connective tissue, minced, pooled in a 60-mm dish, and treated with DNase (10 μg/dish; Sigma Chemical, St. Louis, MO) and collagenase (40 mg/dish; Life Technologies, Gaithersburg, MD) in Wyles solution (137 mM NaCl, 5 mM KCl, 21 mM HEPES, 0.7 mM Na2HPO4, 100 mM glucose, and BSA at 0.1 mg/ml) for 2 h at 37°C. After gentle trituration, the cells were resuspended at a density of 106/ml in sterile DMEM (Life Technologies) containing 10% fetal bovine serum (Life Technologies). To establish the cultures, 0.8 × 106 cells were plated on sterile glass coverslips (VWR, Bridgeport, NJ) or 0.5 × 106 cells were plated on Matrigel (0.15 mg/mm2); on average, each preparation yielded 36 coverslips. The next day, the cultures were supplemented with DMEM containing 10% fetal bovine serum. On day 4, the medium was replaced with DMEM containing 10% fetal bovine serum and cytosine arabinoside (Sigma Chemical) at a final concentration of 20 μM. On day 7, the cultures received DMEM containing 3% horse serum and cytosine arabinoside. The latter was changed every 3 days.
Cultures were established as above on silicon membranes (Flexercell 2000; Flexercell, McKeesport, PA) coated with Matrigel. The cell culture wells containing the silicon membranes were placed in a computerized, vacuum-operated valve system (Flexercell 2000). A uniaxial passive stretch was applied for 48 h between days 7 and 9 of culture. The passive stretch regimen was adapted from the work presented by other investigators (49). In my experiments, the average passive stretch of the membrane was 5%; the stretch was applied as a train lasting 60 s (1 s on, 2 s off, 20 counts/min), with a rest period of 180 s between trains.
Electrical field stimulation was directly applied to the culture as described by Wehrle et al. (53). Coverslips were first transferred in 3% horse serum in DMEM in a 100-mm sterile petri dish, and single bipolar pulses were applied for 48 h starting onday 7 postplating. A stimulator (S-48; Grass, Warwick, MA) connected to a polarity changer, to eliminate electrolysis, was used with a field pulse with a duration of 2 ms, a frequency of 0.1 Hz, and an amplitude of 400 mV/mm. The two stainless steel electrodes were totally submerged (0.4 cm2) in the medium. These parameters generated an average current in the dish of 1.2–1.4 mA.
Tetrodotoxin (TTX; Sigma Chemical), an inhibitor of voltage-gated Na+ channels, was used at 3 μM (46). Inhibitors of L-type Ca2+ channels (nifedipine and verapamil; RBI, Natick, MA) were used at 10 μM (21). Stretch-activated ion channels were inhibited by using gadolinium (Sigma Chemical) at 10 μM (41), and 2,3-butadione 2-monoxime (BDM) was used to inhibit of myofibrillar ATPase (24). To chelate extracellular or intracellular Ca2+, EGTA (Sigma Chemical) or 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid-acetoxymethyl ester (BAPTA-AM; Calbiochem, San Diego, CA) was used, respectively, at different concentrations. The effect on Z-line and A-band morphology was studied in a dose-response design; in the experiments, control cultures were treated with vehicle only. At these concentrations, toxic effects, such as membrane blebbing or cell detachment, were not observed. Cell viability, determined by trypan blue exclusion, was not affected. The number of myotubes per area was obtained by counting the number of myotubes per area (1.357 mm2). Because myotubes may encompass several optical fields, the stage was moved at least 5 mm to obtain a random sample of different myotubes; only five random images per culture dish were evaluated.
Ca2+ fluorescence imaging.
The effect of TTX concentration on the Ca2+ transient in the myotube preparation was assessed in fluo 4-loaded myotubes grown on glass for 7 days, similar to studies using disassociated myofibers (33, 46). Myotube cultures were loaded for 15–30 min with 10 μM fluo 4-AM (Molecular Probes, Eugene, OR) in Krebs-Ringer solution (in mM: 135 NaCl, 4.0 KCl, 1.8 CaCl2, 1.0 MgCl2, 10.0 glucose, and 10.0 HEPES, pH 7.4) . After the loading protocol, the cells were allowed to rest in Krebs-Ringer without the fluorescent dye. Myotube cultures were then incubated in either control Ringer or Ringer with TTX (1 nM to 10 μM). Myotubes were selected randomly and monitored for functional Ca2+release at 37°C in a custom-built air-jacketed chamber on an inverted microscope (Olympus IX-70 ×60–1.4 numerical aperture oil or ×60–1.3 numerical aperture water objective) coupled to a Bio-Rad MRC-600 laser scanning confocal system (488-nm excitation) used inxy-mode (1-s acquisition time; 2 ms/line, 768 pixels/line). Representative mature myotubes were selected by phase-contrast microscopy; on average, 10 myotubes per 35-mm dish were analyzed, and data were collected from the central part of the myotube. After a baseline collection (3–4 images), a voltage test pulse (2 ms at 400 mV/mm) was delivered ∼100 ms into the subsequent image generating a brief fluorescent transient. Pretest pulse baseline images were summed and averaged to generate an average fluorescence image (F0), which was then subtracted from each of the subsequent images in the series to create a change in fluorescence images (F − F0). The mean peak F − F0 (>10% of peak F − F0) during the test pulse was determined by manually selecting pixel regions of the Ca2+ transient. The average pixel intensity of the region of interest was quantified by image analysis software (IDL 5.0, Boulder, CO).
Antibodies against sarcomeric α-actinin (clone EA-53) were obtained from Sigma Chemical. Antibodies against the following rat MHCs were used: fast MHC (clone MY 32, which cross-reacts with neonatal MHC, Sigma), types IIa and IIb MHC (clones SC-71 and BF-F3, respectively; American Type Culture Collection, Manassas, VA), and developmental/neonatal and slow MHC (clones RNMy2/9D2 and WB-MHCs, respectively; Novacastra, Newcastle-upon-Tyne, UK).
Cultures that were maintained under control conditions or were experimentally manipulated by mechanical stretch, electrical stimulation, or addition of drugs were fixed for 10 min with 2% paraformaldehyde in PBS (pH 7.2). After permeabilization with 0.5% Triton X-100 in PBS, the samples were incubated with BSA-PBS (1 mg/ml) for 30 min. The primary antibodies were applied at a concentration of 5 μg/ml for 1 h. The samples were washed three times in BSA-PBS and incubated for 1 h with a fluorescein-conjugated secondary antibody (goat anti-mouse IgG; Jackson Immunochemicals, West Grove, PA) diluted at 1:100. After three washes in PBS-BSA, coverslips were mounted in 90% glycerol and 10% 1 M Tris · HCl, pH 8.0, supplemented with 1 mg/mlp-phenylenediamine and observed through a PlanNeofluar ×63–1.40 numerical aperture oil-immersion objective on a Zeiss IM 35 microscope (Oberkochen, Germany). In double-labeling experiments, A bands were visualized with monoclonal antibodies against MHC (clone MY-32), then rabbit anti-mouse Fab fragments (Jackson Immunochemicals), and then tetramethylrhodamine-conjugated goat anti-rabbit IgG (Jackson Immunochemicals). The Z lines were visualized with the monoclonal antibody (clone EA-53), which was then labeled by a fluorescein-conjugated goat anti-mouse IgG. Digital images were obtained with a Zeiss 410 confocal laser-scanning microscope (Carl Zeiss) without further processing. The slides were randomized, and the integrity and presence of Z lines (Fig.1 E) or A bands (Fig.1 F) were evaluated in eight random fields in each slide. Approximately 100 myotubes were scored per slide. Myotubes labeled with anti-α-actinin continuous Z lines (e.g., Fig. 1 E) were counted as positive cells. Partial Z lines and interrupted Z lines were counted as negative (e.g., Figs. 1 B and 5 E). Similarly, myotubes labeled for MHC that did not show spatially segregated A bands (e.g., Fig. 1 D) were counted as not having A bands in fully mature myofibrils. The data were presented as a percentage of total cells containing Z lines or A bands. ANOVA and appropriate post hoc tests (Tukey's, Scheffé's) were used to analyze the data. The data are expressed as mean ± SE. Differences between data sets were considered to be statistically significant at P < 0.05.
Myotube cultures were rinsed three times in PBS medium and harvested at 4°C by scraping in SDS-PAGE sample buffer. After samples were boiled, protein concentrations were determined using amido black (23). Samples were then subjected to SDS-PAGE (29) and transferred to nitrocellulose electrophoretically. Immunolabeling was performed as described above with specific primary antibodies and then goat secondary antibodies conjugated to alkaline phosphatase. The bands were visualized using chemiluminescence (Tropix, Bedford, MA) and radiographic film (Biomax ML; Kodak, Rochester, NY).
I used immunofluorescent labeling with antibodies specific to α-actinin and MHC to assess the presence of Z lines and A bands in maturing rat myotubes. Figure 1, E and F, shows the morphological criteria used to determine whether Z lines or A bands were present. Myotubes gradually attain a sarcomeric pattern as a function of time in culture. At day 4, most of the α-actinin labeling is present at stress fibers (arrow in Fig.1 A), and, later, a striated pattern can be seen initially near the sarcolemma (arrow in Fig. 1 B). An organized MHC labeling is first seen as thin myofibrils in the cytoplasm (arrow in Fig. 1 D). Gradually, the majority of the myotubes have sarcomeric structures.
I used a qualitative method, based on the morphology as depicted in Fig. 1, E and F, to evaluate the development of myofibrils in the mammalian muscle cell culture. After cells were plated, myoblasts proliferated for several days and then fused ondays 4 and 5 to give rise to multinucleated myotubes. On day 5, some of the myotubes showed spontaneous twitching, and, consequently, the number of cells containing Z lines and A bands increased significantly (Fig. 1, H andI). This progressive development of sarcomeric elements was complete by day 8, when ∼60% of the myotubes displayed Z lines and A bands (Fig. 1, H and I). Although the Z lines appeared to develop before fully mature A bands could be visualized in several cultures, no statistical difference (P = 0.12) in the rate of development of these structures was apparent from the results. Similar to the observations of others (19), I observed that rat muscle cultures grown on Matrigel showed earlier fusion and maturation, especially in the development of Z lines. Six days after initial plating on Matrigel, 41 ± 12% of the myotubes contained Z lines compared with 24 ± 5.5% in myotubes grown on glass (P < 0.001, 3 independent experiments). By day 9, 78 ± 5% of the myotubes grown on Matrigel contained Z lines compared with 52 ± 6% for myotubes grown on glass (P < 0.001). The appearance of A bands in the myotubes did not show a substrate dependency. On day 6, 20 ± 19% of the myotubes grown on glass contained A bands and 16 ± 11% of the cells grown on Matrigel had fully mature A bands (P = 0.53). The development of mature myofibrils was dependent on the amount of spontaneous contractile activation, as TTX was able to block the development of Z lines in a dose-dependent manner (Fig.2). The 48-h pharmacological treatment in these cultures did not alter the number of myotubes per dish. In control cultures, there were 15.3 ± 5.8 myotubes/mm2; in TTX-treated cultures, there were 17.9 ± 3.4 myotubes/mm2; and, in the verapamil-treated cultures, there were 20.3 ± 6.6 myotubes/mm2. No statistical differences were noted between the number of myotubes per dish.
Having established reliable conditions for studying the development of contractile structures in myotubes in vitro, I proceeded first to study the effects of passive stretch of the myotubes grown on Matrigel. As reported for chick myotubes (49), after passive mechanical stretch, the rat myotubes aligned parallel to the longitudinal axis of the uniaxial deformation (bidirectional arrow in Fig.3, A andC). In contrast, the controls showed a more random orientation of myotubes (Fig. 3 B). By applying passive stretch, the number of myotubes containing Z lines was not significantly affected (71 ± 11%, P = 0.1) (Fig.4). The application of a 10% stretch led to the loss of the majority of the myotubes; the effect of other stretch regimens on the presence of Z lines was not further evaluated. The application of the 5% passive stretch resulted in well-developed focal adhesion-like complexes near the substrate-attached membrane (see arrow in Fig. 3 E). These complexes contained dartlike α-actinin-labeled structures associated with stress fibers (phalloidin staining is not shown), similar to structures reviewed in Ref. 7. The application of more vigorous passive stretch protocols resulted in the detachment of most of the myotubes. The data suggest that passive stretch as it was applied here does not affect Z lines. Similar to the Z lines, the presence of A bands was not influenced by the application of passive stretch (not shown).
I used TTX, an inhibitor of voltage-gated Na+ channels, to block the spontaneous contractile activation of the myotubes in developing cultures from days 7 to 9 postplating. A dose response for TTX was established (Fig. 2), and, in all the experiments described herein, we used 3 μM. The lack of contractile activation resulted in myotubes lacking Z lines, and the sarcomeric α-actinin labeling was found in stresslike fibers (arrow in Fig.3 H) and in a punctate pattern throughout the sarcoplasm (arrowhead in Fig. 3 H). The application of passive stretch did not reverse the absence of Z lines. In the TTX-treated cells, under both stretched and nonstretched conditions, the number of myotubes with Z lines decreased to 2 ± 3% and 3 ± 3%, respectively (Fig. 4). Similar changes occurred in the organization of MHC in that the TTX treatment resulted in the nearly complete absence of recognizable A bands (data not shown). An experiment in which spontaneously twitching myotubes were treated with gadolinium (10 μM, 48 h) showed no significant alteration in the number of A-band-containing myotubes (64 ± 7%, n = 3). However, a 24-h treatment of 8-day-old cultures with BDM (5 μM), which inhibits myosin-actin interaction, led to a decrease in the number of Z-line-containing myotubes compared with control cultures (39.6 ± 23 and 78.5 ± 7.7%, respectively;n = 3, P = 0.05, power = 0.50), and a similar effect was noticed in the number of A-band-containing myotubes (39.6 ± 23 and 76 ± 1.5%, respectively;n = 3, P = 0.025, power = 0.71). Passively stretching the TTX-treated myotubes had no effect on either Z lines or A bands. These results suggest that contractile activation and perhaps the internal mechanical strains are a major factor in controlling myofibrillogenesis.
The effect of TTX was at least partially reversible. In these experiments, the removal of TTX on day 9 after a 4-day treatment and the resultant reallowance of spontaneous contractile activation for a 2-day recovery period resulted in the appearance of Z lines. Under these culture conditions, 39.3 ± 7.5% (see Fig.6 A) of the myotubes contained Z lines and 35 ± 6.5% of the cells contained A bands (see Fig. 6 B). The removal of TTX (recovery) resulted in Z lines and A bands having a morphology similar to that of the control cultures (Fig.5, C and D). In 11-day control cultures, the number of cells containing Z lines was 57.3 ± 4.3% (Fig. 6 A) and the number containing A bands was 60.3 ± 4.4% (Fig.6 B), which was similar to the data acquired previously when the myotubes were grown on glass (Fig. 1).
To test whether contractile activation is required for the maintenance of Z lines or A bands, TTX was added at different time points during development. Myotubes were treated with TTX during sarcomeric assembly (days 5–11) or when the assembly process was already accomplished for most of the myotubes (days 9–11). As mentioned above, the addition of TTX from days 5 to11 resulted in a loss of sarcomeric structures. The number of cells containing Z lines was 0.75 ± 0.1% (Fig.6 A), and the number of cells having recognizable A bands was 1.1 ± 0.5% (Fig. 6 B). To determine whether the Z lines and A bands were susceptible to disassembly, TTX was added at a later stage (days 9–11), at which time most of the myotubes had assembled these structures. The TTX treatment resulted in the nearly complete loss of assembled Z lines and A bands (Fig. 5,E and F). Under these conditions, 0.83 ± 0.5% of the cells contained Z lines (Fig. 6 A) and 2.3 ± 1 of the cells contained A bands (Fig. 6 B). These experiments suggest that electrical activity of the sarcolemma and contractile activation, by either a direct or an indirect mechanism, are required for the development and maintenance of normal sarcomeric structures.
To evaluate the MHC isoforms and quantitative changes, I used Western blot analysis. In myotube cultures grown on Matrigel, the 48-h treatment with passive stretch or TTX did not result in major differences. I used several antibodies to determine the type of myosin present in these cultures. Proteins in homogenates from different experimental groups were separated by SDS-PAGE, transferred to nitrocellulose membranes, and labeled with antibodies specific for sarcomeric α-actinin, developmental MHC, type IIa MHC, type IIb MHC, and slow MHC. The results indicate that these cultures contain mostly developmental (neonatal) MHC and some type IIa (Fig.7). Separation of the MHC in high-glycerol-containing SDS gels, according to the work presented by Caiozzo et al. (8), pointed to the same conclusions: the predominant type of MHC in these cultures is the neonatal form, with small amounts of types IIa and IIx present (not shown). Filters stained with antibodies against type IIb MHC (clone BF-F3) and slow MHC did not result in a visible signal. We compared the blots from four independent experiments and could not detect any major changes in signal intensity (representative blot depicted in Fig. 7).
Because spontaneous activity (contractile activation) promoted myofibrillogenesis and blocking inhibited myofibrillogenesis, I reasoned that exogenous electrical stimulation might accelerate the development of Z lines and A bands. I therefore exposed the myotube cultures to electrical stimulation for 2 days at a frequency of 0.1 Hz. In addition, inhibition with the use of antagonists of the L-type Ca2+ channels resulted in a decrease in the amount of myotubes containing Z lines (Fig. 8). As expected, the electrical stimulation protocol resulted in an increased number of cells with Z lines (Fig.9 A) and A bands (Fig.9 B) on day 9. In three independent electrical stimulation experiments, the number of Z-line-containing myotubes increased from 50 ± 4% in controls to 73 ± 5% (P = 0.05), and the number of A-band-containing myotubes increased from 50 ± 9 to 71 ± 3% (P = 0.02) (Fig. 9). The addition of TTX in combination with the electrical stimulation led to the nearly complete absence of Z lines (0.6 ± 0.7%, P < 0.001) and, to a lesser extent, A bands (10.5 ± 5.5%, P < 0.001) (data not shown). Thus low-frequency electrical stimulation increased the number of mature myotubes in the dish, perhaps by synchronizing the population of developing myotubes.
I used pharmacological agents to explore L-type Ca2+channels and their role in the appearance of sarcomeric structures. Blocking the influx of Ca2+ had an effect similar to that of blocking the membrane depolarization by TTX in both the spontaneous and the electrically stimulated cultures. Verapamil and nifedipine (both at 10 μM), L-type Ca2+ channel blockers, led to the absence of the Z lines (Fig. 9 A) and A bands (Fig.9 B). In spontaneously twitching cultures, the addition of verapamil reduced the number of Z-line-containing cells to 8.5 ± 3.6% (P < 0.001); the addition of nifedipine had a similar effect (6 ± 3.3%, P < 0.001). The number of A bands in spontaneously contracting myotube cultures was reduced to 3.6 ± 1.9% (P < 0.001) when treated with verapamil and to 8.8 ± 6.5% (P < 0.001) when treated with nifedipine. Electrical stimulation combined with the Ca2+ channel blockers led to similar results. Although the data suggested a lesser effect on the sarcomeric structures by verapamil than by nifedipine, the data were not statistically different from those of the other pharmacologically treated groups. The application of both verapamil and electrical stimulation in the cultures yielded 19 ± 5% Z-line-containing myotubes (Fig.9 A, P < 0.001) and 10.5 ± 4.6% A-band-containing myotubes (Fig. 9 B, P < 0.001). The treatments with nifedipine and electrical stimulation led to 2.7 ± 2.5% Z-line-containing myotubes (P < 0.001) and 1 ± 1.7% A-band-containing myotubes (P < 0.001). The data indicate the role of Ca2+ in sarcomeric assembly.
To ascertain whether the TTX treatment during the electrical stimulation protocol led to a lack of Ca2+ fluxes, an experiment was undertaken to visualize the intracellular amount of Ca2+ in these developing myotubes (Fig.10). The data show that, under the TTX concentrations and electrical stimulation conditions used in the experiments, Ca2+ flow was significantly inhibited. In an additional set of experiments, a Ca2+ transient was visualized at a TTX concentration of 10 μM by increasing the voltage in the stimulation protocol to 800 mV/mm (not shown). To determine whether the internal Ca2+ pools contributed to sarcomeric assembly, I used a pharmacological approach. The addition of EGTA to the medium led to a dose-dependent decrease in the number of Z-line-containing myotubes (Fig.11 A), whereas the chelation of internal Ca2+ by using BAPTA-AM at different concentrations did not affect Z line formation (Fig. 11 B). To reverse the effect of reduced external Ca2+ entry by verapamil, the release of internal Ca2+ stores was induced by using nanomolar amounts of ryanodine (Fig. 11 B); such an approach did not lead to an increase in Z-line-containing myotubes. Immunofluorescent labeling with antibodies against the ryanodine receptor indicated a punctate, suggesting an unorganized internal membrane system (data not shown).
In vivo models of muscle development and muscle plasticity may not be able to determine completely the independent contributions of the mechanical environment (i.e., the stresses and strains generated within the tissue) or of the electrical activity (dictated by the neural recruitment of the motoneurons). The independent effects of passive stretch (an external strain) and contractile activation (inducing internal strains) on muscle development may be different. In an effort to study these effects, I used a mammalian cell culture model of myofibrillogenesis. By combining pharmacological agents with either passive stretch or electrical stimulation, I was able to assess the effects of these treatments on the development of sarcomeric structures. A passive, external stretch, as was applied here, had no apparent effect. In contrast, electrical activity, by either spontaneous contractile activation or induced electrical stimulation, is not only required for the assembly of Z lines and A bands but is also necessary for the maintenance of these structures.
Role of passive stretch in myofibrillogenesis.
The in vitro application of passive stretch is used to evaluate the synthesis of prostaglandins in fibroblasts (1), and the deformation of the substrate in this application can be modeled mathematically (20). The application of a passive stretch is used in chick cultures to assess α-actin gene transcription (9) or to evaluate morphological changes (49). I applied passive stretch to the cultured rat myotubes by using the approach based on the work of Vandenburgh and Karlisch (49). Using this method, I could (by adding pharmacological agents such as antagonists for voltage-gated Na+ or Ca2+ channels) inhibit the spontaneous contractile activation, an approach that was used earlier (4,46). I could thereby independently test the effects of passive stretch or contractile activation on the development of sarcomeres in mammalian skeletal muscle cells. Using morphological criteria similar to those used by others (40), I could not show that a cyclic passive stretch regimen (averaging 5%) facilitated or inhibited the formation of Z lines or A bands. In vitro, passive stretch may actually have a series of catabolic effects in muscle, such as a decreased level of transcription (9), induction of membrane damage (12), and the generation of phospholipase-derived products (11, 50, 51). It is possible that passive stretch may have a role during the myoblast stage rather than at the myotube stage, similar to what is shown in alveolar cells (42) and cardiac fibroblasts (34). In cultured, mononuclear cells, passive stretch gives rise to increased cell proliferation (42) and activation of cell proliferation-related kinases, such as extracellular signal-related kinase and c-Jun NH2-terminal kinase (34).
Role of contractile activation and electrical stimulation in sarcomeric formation.
By using pharmacological agents, I could inhibit the ion flow during spontaneous contractile activation or electrical stimulation. Although adult skeletal muscle responds to TTX at nanomolar levels, in denervated and developing muscle a different voltage gated Na+ channel (SkM2) is expressed that is sensitive to TTX in the micromolar range (55). In culture, both genes are expressed (55), necessitating the use of TTX in the micromolar range (Fig. 2). Blocking the extracellular Na+movement with or without a passive stretch regimen led to the absence of Z lines and A bands, suggesting that membrane depolarization is more important for sarcomeric assembly than is passive stretch. Direct electrical stimulation of the cultures combined with the pharmacological manipulations pointed more immediately to this conclusion. Inhibition of either the voltage-dependent Na+channels or the L-type Ca2+ channels in the mammalian myotube culture resulted in the failure to assemble and maintain Z lines and A bands.
Experiments in which whole muscle is exposed to altered loading (5, 14, 48) convincingly showed altered gene expression, which could be the result of either a modified mechanical environment or an altered neural recruitment. To complement these experiments, additional structural studies are necessary to understand sarcomere assembly, an issue of particular importance for muscle, since its structural characteristics are related to its functional properties. My experiments suggest that, in developing myotubes, the role of contractile activation on myofibrillogenesis is more important than the influence of passive mechanical factors. Parallel experiments emphasize the importance of neural activity on the MHC gene expression by applying electrical stimulation to cultured muscle cells (16) or by using nerve-muscle cocultures (15).
How depolarization and the inward Ca2+ flow may facilitate Z-line assembly are not yet known, but these processes may preferentially affect α-actinin because this molecule is one of the first to be organized in a striated fashion in skeletal muscle. Details are not known, and a recent report documents that the actin, binding domain, the spectrin repeats, and the EF-hand motif of α-actinin and EF-hand motif of α-actinin are not required for Z-line assembly (31), suggesting that other molecules may play a role in Z-line assembly (2, 38, 54). Protein motifs, such as LIM (38) and PDZ (54) domains, may be important in the assembly of fully mature Z lines.
The interactions between the hydrophobic and hydrophilic amino acids of the myosin rod domain are thought to comprise the most important factor in the self-assembly of myosin (3). How these heterodimers assemble further into functional A bands is still not fully known, although several mechanisms have been suggested (17, 39,40). Molecules associated with myosin are thought to play a role in the assembly of A bands (52), but how this mechanism is regulated is not well described.
Synthesis and assembly of sarcomeric structures.
An important aspect of muscle development is the orchestrated interplay between synthesis, degradation, and assembly of contractile elements. I did not detect major changes in the level of MHC and α-actinin protein in two different experimental conditions, but the experiments were relatively short. Changes in muscle gene expression and altered phenotype as a result of altered loading or electrical stimulation are seen only after several days to weeks (5, 37). Increased turnover of myofibrillar proteins is shown in chick muscle culture grown for 6 or more days in high-K+-containing medium (12 mM) or in normal medium supplemented with TTX (4). Similarly, electrical stimulation of cultured muscle satellite cells induces changes at the MHC mRNA level only after several weeks (16, 35, 53).
Role of Ca2+ in myofibrillogenesis.
Altered neural activity and the altered electrical pulse pattern might lead directly to gene transcription (6), as shown for the MHC (35). The pulse pattern might lead to temporally and spatially defined increases in intracellular Ca2+ and thus gene expression. A Ca2+-dependent gene transcription is proposed in MHC fiber-type transformation (10, 28), but change in muscle phenotype as it occurs during development and in adult life would also have to entail a well-controlled protein degradation and sarcomeric assembly process. The data presented herein indicate a role for temporal, extracellular Ca2+ fluxes in sarcomere assembly and maintenance. It is interesting to note that calpain, a Ca2+-dependent protease, is able to remove Z lines from muscle fibrils without causing the degradation of α-actinin and without affecting its binding capability to actin (22,31). Treatments of muscle cell cultures with phorbol esters result first in the disassembly of Z lines and subsequently in the disassembly of A bands (32). Thus Ca2+ and related second messenger pathways might activate both selective gene transcription and sarcomeric assembly.
In conclusion, on the basis of the current work, it seems that continuous contractile activation during development and in adult life is an important factor for the further maturation of a muscle fiber and that it is a requirement for the maintenance of sarcomeric structures during muscle fiber development.
I appreciate the help of Dr. Chris Ward during the Ca2+imaging experiments, and I am grateful for the suggestions from Dr. Susan D. Kraner regarding the SkM2 Na+ channel. I also thank Dr. Martin Schneider and especially Dr. Robert Bloch for comments and helpful suggestions, Rick Meyer for technical assistance, and Dori Kelly for editorial assistance.
This work was possible through the support of the National Center for Medical Rehabilitation Research.
Address for reprint requests and other correspondence: P. G. De Deyne, PhD, MPT, Division of Orthopedics, MSTF, Rm. 400, 10 South Pine St., Baltimore, MD 21201 (E-mail:).
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 © 2000 the American Physiological Society