Myoblast fusion requires cytosolic calcium elevation but not activation of voltage-dependent calcium channels

Cell Calcium (1996) 19(5), 365-374 0 Pearson Professional Ltci 1996

Research

Myoblast fusion requires cytosolic calcium elevation but not activation of voltage-dependent calcium ‘channels B. Constantin, Laboratory

of General

C. Cognard, G. Raymond Physiology,

URA CNRS

1669,

University

of Poitiers,

Poitiers,

France

Many studies of in vitro skeletal myogenesis have demonstrated that fusion of myoblasts into multinucleated myotubes is regulated by calcium-dependent processes. Calcium ions appear to be necessary at the outer face of the membrane, and an additional internal calcium increase seems required to promote fusion of aligned myoblasts. It has been proposed that a calcium influx could take place prior to fusion and that this may be mediated by voltage-dependent calcium channels. Previously, we showed that two types of voltage-dependent calcium currents were expressed in multinucleated myotubes but not in rat myoblasts growing in primary culture before the withdrawal of the growth medium. We also showed that the previous formation of multinucleated synticia was not a prerequisite of developmental appearance of calcium currents, suggesting that the two events were time-correlated but not sequentially dependent. These features led us to investigate changes in internal calcium activity and the possible appearance of voltage-dependent calcium influx pathways just after the promotion of fusion by the change of culture medium. The results confirm that a rise in cytosolic calcium activity occurs slightly before fusion in confluent myoblasts and remained in newly formed myotubes. Reducing this elevation by internal calcium buffering lowered myoblast fusion and, reciprocally, blocking cell fusion prevented calcium increase. Treatment with the organic calcium channel blockers nifedipine (5 PM) and PN 200-l 10 (1 PM) did not alter cytosolic calcium changes nor cell fusion, and voltage-dependent calcium currents were never observed by the perforated patch-clamp technique in aligned fusion-competent myoblasts. Other voltage-operated mechanisms of calcium rise were not detected since depolarization with hyperpotassium solutions failed to elicit increases in intracellular calcium. On the contrary, acetylcholine was able to promote extracellular calcium-dependent calcium transients. Our results confirm the requirement of an increase in resting calcium during fusion, but do not support the hypothesis of an influx through voltage-dependent channels or other voltage-operated pathways. The elevation of internal calcium activity may result from other mechanisms, such as a cholinergic action for example. Summary

INTRODUCTION

For several decades, myogenic cell culture have been an informative model reproducing part of the sequential Received Revised Accepted

13 October 30 January 1 February

Correspondence URA CNRS France

1869,

1995 1996 1996

to: Dr 8. Constantin, 40 av. du Recteur

Laboratoire de Physiologie GBn&ale. Pineau, F-86022 Poitiers Cedex,

events involved in the early development of skeletal muscle [1,2]. Briefly, cycling precursors proliferate, withdraw from the cell cycle and align along their long axis. Under control of endogenous and exogenous factors, myoblasts are able to undergo cytodifferentiation by expressing muscle-specific proteins and functions and fusing to form multinucleated syncitia. Many factors have been implicated in fusion events [3]. The early study of Shainberg et al. [4] highlighted extracellular calcium requirements for the formation of myotubes and Przybylski et al. [5] recently demonstrated that myoblast fusion is dependent on the continuous presence of calcium on the cell 365

366

B Constantin, C Cognard, G Raymond

surface. Calcium ions may be involved in many processes mediating fusion, such as cell surface interactions via Ncadherins [6,7] or phosphorylation of surface proteins [8]. In addition, David et al. [9] proposed that a net calcium influx into myoblasts is obligatory for synticia formation. More recently, the determination of exchangeable free calcium concentration in mass culture also suggested the increase of cytosolic free calcium concentration ([Ca*+]J prior to fusion [5] and the authors concluded that a calcium influx was necessary, but not sufficient, to promote myotube formation. m view of these findings, it is important to ascertain whether a calcium elevation occurring prior to fusion can be observed by means of calcium measurements performed upon individual cells. If so, which pathways could mediate calcium movements in myoblasts committed to fuse? Entwistle et al. [lo] suggested that fusion-competent myoblasts have a high resting membrane potential and that depolarization initiated by ion channels operated by prostaglandins and acetylcholine triggered a rise in calcium permeability through voltage-dependent calcium channels (VDCC) thus promoting myoblast fusion. This hypothesis, mainly supported by D600 and KC1 effects on embryonic chick myoblast fusion, needs to assume that functional VDCC are present in the membrane of fusioncompetent myoblasts. However, neither dihydropyridine receptors nor voltage-dependent calcium currents were found in myoblasts developing in primary culture until they were maintained in growth medium [ 11,121. Nevertheless, the recent demonstration that myoblast fusion is not a prerequisite for the appearance of voltage-dependent calcium currents [ 131, shows that both events are not sequentially linked and suggests the possibility of a concomitant or slightly earlier appearance of calcium currents during triggering of fusion. Indeed, results obtained with murine line cells BC3Hl and C2 suggest that the appearance of functional VDCC in mononucleated cells can be observed only after serum withdrawal but not in the presence of mitogens [ 14,151. However, it is not certain that BC3HI and C2 behavior can be extrapolated directly to the behavior of normal skeletal muscle cells. Whether or not committed myoblasts express functional VDCC after serum withdrawal and prior to fusion remains an open question. Since, our cell culture model permits us to experimentally trigger myoblast fusion by growth medium withdrawal, the possible appearance of voltage-dependent calcium currents was investigated during the critical period of the first formation of myotubes (within 20 h after growth medium withdrawal). During the same period, changes in cytosolic calcium activity were followed at the individual cell level by means of ratiometric cytofluorimetry, and the effects of a permeant calcium buffer (BAPT’AM) and of organic L-type VDCC blockers Cell Calcium

(1996)

1 S(S), 365-374

(nifedipine and PN 200-l 10) were tested on temporal changes in [Ca2+li and on multinucleated myotube formation. MATERIALS

AND METHODS

Cell culture

Primary cultures of mammalian skeletal muscle cells were initiated from satellite cells obtained by trypsinisation of muscle pieces from hind limbs of I-3-day-old neonatal rats. Dissociation and culture techniques have been reported in detail elsewhere [ 121. For 2 days following plating, cells were maintained in growth medium (300 l.tM [Ca*+],), consisting of HAM F12 (Gibco BRL, Life Technologies, Eragny, France) with 10% heat-inactivated horse serum (Gibco BRL), 10% fetal calf serum (Boehringer Marmheim, Meylan, France), and 1% antibiotics. Myoblasts underwent myogenesis in differentiation medium (1.8 mM [Ca2+]J, containing DMEM (Gibco BRL) supplemented with 5% heat-inactivated horse serum. After 48 h of culture, this control medium (DMEM + serum) was used to promote the formation of myotubes which occurs within 15- 18 h. The fusion-promoting conditions were provided by the presence of a higher calcium concentration (1.8 mM) and horse serum. This medium exchange was used as time zero for the culture (time after induction of fusion). When stated, this medium was supplemented with the membrane-permeant form of BAPTA (BAPTA/AM), with nifedipine or PN 200-l 10. Incubation with BAPTA/AM was performed for 2 h for loading the cells with the calcium chelator, and BAPTA/AM was removed from the culture medium to prevent side effects due to long term exposure. As previously described [13], two different media were also used to prevent cell fusion: (i) a low-calcium medium (50 PM CaCl, supplied by 5% horse serum + ‘calcium-free DMEM’ prepared by Gibco BRL); and (ii) a control medium supplemented with 1.25 pg/ml cytochalasin B (Sigma). Evaluation of myoblast fusion

After 20 h (68 h after plating), cultures were fixed with 4% paraformaldehyde in PBS for 20 min, washed 3 times for 10 min with PBS, stained with Giemsa blue for 25 min and then washed 3 times in PBS. Ten different fields per Petri dish (around 1000 different cells) were randomly observed under an inverted microscope (objective 40x). Fused cells were determined by direct microscopic examination, only by considering cells with clear cytoplasmic continuity and at least 3 nuclei in the cytoplasm. Myoblasts were morphologically designated as aligned mononucleated cells with marked bipolar, spindle-shaped morphology. Flattened and extended mononucleated 0 Pearson

Professional

Ltd 1996

Calcium influx and myoblast fusion

cells were classified as fibroblasts and discarded. The fusion index was estimated as the total number of nuclei observed in multinucleated cells divided by the total number of nuclei. Resting cytosolic free calcium concentration intracellular calcium transients

and

Cells were loaded with the calcium fluorescent probe, its membrane permeant form Indo- 1, through @do-l/AM, Sigma). Details of the technique have been reported elsewhere [ 161. Intracellular free Ca2+ concentrations were measured at room temperature by means of a ratiometric fluorescence method using a laser cytometer ACAS 570 (Meridian Instruments, Okemos, MI, USA). Indo-l fluoresced in response to excitation by laser light (beam width = 0.6 lun) in the LJV wavelengths (351 and 364 nm rays). Calcium activity was estimated as the ratio of the 405/485 nm emissions of both bound and Ca2+free forms of the dye, respectively. Variations of the calcium activity with time were measured at a fixed point in the cell, to obtain good time resolution, whereas the resting [Caz+], was evaluated from laser scanning images. Since the laser beam is fixed, two dimensional specimen scanning was performed by the use of a two-axis, microstepper-motor drive system and a precision X-Y microscope stage. Fluorescence images measured simultaneously by single frame scanning at 405 m-n and at 485 nm were ratioed pixel by pixel. The ratio was calculated in the entire cell as the average of each pixel’s ratioed value measured in a hand-delimited polygonal area. The cytosolic free calcium activities were calculated from the Grynkiewicz equation [ 17] : [Caz+], = Kd x B x [(R - R,,)/(R,,

- R)]

where R is the ratio of the fluorescence signals (F,,,/F,,J, the Kd of Indo-l was assumed to be fixed at the in vitro calculated value 250 nM [ 171 and the Rti and R,, values were determined in vivo by a calibration method described elsewhere [ 161. For rapid changes in intracellular calcium activity under stimulation (100 mM KCI, 10 PM acetylcholine), the fluorescence ratio was measured at a single point, which gave a better time resolution. Recording

of transmembrane

currents

Ionic transmembrane currents were recorded at room temperature (ZO-22°C) by means of the whole-cell configuration of the patch-clamp technique [18] or the derived perforated patch method [ 191 to avoid ‘wash out’ effects in small myoblasts. In this latter case, pipettes (2-5 MSZ) were backfilled by sonicated amphotericin-B (150 pg/ml) diluted in pipette solution after filling the pipette tip with an amphotericin-B free internal solution. 0 Pearson

Professional

Ltd 7996

367

Pipettes were then connected to the head stage of a patch-clamp amplifier (RK300, Bio-Logic, &ix, France), driven by a PC-AT compatible microcomputer equipped with a Labmaster A/D conversion board (Scientific Solutions, Solon, MA, USA). Membrane voltage clamping, data acquisition and analysis were performed by means of a software package (PClamp, Axon Instruments, Foster City, USA), and data graphics through the Fig.P software (Fig.P Corp., Biosoft, Cambridge, UK). Data were corrected for leakage currents from linear extrapolation of membrane current magnitudes (assumed to be ohmic) elicited by small depolarisations. Experimental

solutions

The culture medium was exchanged before each experiment by a saline bath solution allowing for sodium or calcium current recordings. The presence of sodium current constituted the test for the functionality of the cell and stabilisation of the rising rate and the maximum amplitude of sodium currents testified to the good perforation of the membrane under the pipette tip. The solution for sodium current contained (in mM): 100 tetraethylammonium chloride (TEA-Cl), 30 NaCI, 5 CaCl,, 0.8 MgCl,, 10 HEPES, 5.6 glucose, pH 74 adjusted with TEA-OH. This bath solution was replaced by a sodium-depleted solution allowing the suppression of sodium currents and detection of the two types of calcium currents (T and Ltypes). This solution contained (in n&I): 135 TEA-Cl; 5 CaCI,, 0.8 MgCI,, 10 HEPES, 5.6 glucose, pH 74 adjusted with TEA-OH. When stated, CaCI, was replaced by 2.5 mM BaCl,. For ratiometric cytometry, 135 mM TEA-Cl was replaced by 133 mM NaCl + 2 mM KC1 and the saline solution contained 2.5 mM Cad,. For the calciumfree condition, a calcium-free solution was perfused and the saline solution was supplemented with 5 mM EGTA. Patch pipettes were filled with a saline solution containing (in mM): 145 CsCl, 1 MgCl,, 0.005 CaCI,, 1 EGTA, pH 72 adjusted with CsOH. Calcium-free solution, KC1 and acetylcholine were applied by means of a microsuperfusion device, at concentrations indicated in the text. RESULTS Measurements

of resting cytosolic

calcium activity

Firstly, the evolution of cytosolic resting calcium activity during the early phases of development was followed in cells from the same culture (Fig. lA), thereby distinguishing 4 steps in myoblast development, that is 20 h, 48 h, 68 h and 76 h after plating. The resting [Ca2+li was higher in early myoblasts at 20 h and 48 h after plating (134 f 5 nM and 123 * 12 nM, respectively), compared to differentiated myotubes (around 100 nM [16]), but in agreement Cell Calcium

(1996)

19(5),

365-374

368

B Constantin, C Cognard, G Raymond

HAM

DMEM

160 1

DMEM CB

DMEM BAPTAIAM

ns

100

60

HAM

DMEM NIF Fig. 1 Average values of resting calcium activity measured in whole cells by means of laser scanning microscopy, and evaluated in different culture conditions and at different development stages. Cells cultured in the growth medium (HAM Fi 2, 300 pM CaCI,) are represented by white bargraphs whereas cells cultured in the fusion-promoting medium (DMEM, 1.8 mM CaCI,) are represented by graduated shading ones. (A) Age-dependent evolution of cytosolic resting calcium activity during early phases of in vitro myogenesis. The age after plating is indicated in hours under each bargraph. Mb, mononucleated myoblasts; Mt, multinucleated myotubes. (B) Effects of fusion blockers and calcium buffers on the cytosolic calcium change observed in myoblasts at 68 h after 20 h in fusion-promoting medium (DMEM). Black bargraphs represent the group of cells treated with 2.5 f.rM cytochalasin B (CB), or 15 pM of the cell permeant form of BAPTA. (C) Organic VDCC blockers are ineffective in preventing the elevated calcium level observed at 68 h after 20 h in the fusion-promoting medium (DMEM). Black bargraphs represent groups of cells treated with 1 pM PN 200-110 (PN) or 5 uM nifedipine (NIF). The gray bargraph corresponds to the average calcium level from 68 h myoblasts in growth medium (HAM) obtained from values reported in A and B. Bargraphs show the mean + SEM from several different cells. ns, not significantly different (Students f-test) vs the unlabelled class: P 2 0.05; *significantly different, P c 0.05; **significantly different, P < 0.01; n = number of tested cells. Cell Calcium

(1996)

19(5),

DMEM

365-374

DMEM PN

with the elevated global value reported for myoblasts in a previous paper (13 1 + 9 nM [ 161). Prolonged exposure to the growth medium prevented myoblast fusion and the mean resting [Caz+], significantly decreased with time (from 13 1 to 90 + 4 nM at 68 h) to reach a mean value similar to that measured in differentiated myotubes. When the growth medium (HAM in Fig. 1A) was exchanged at 48 h with the fusion-promoting medium containing a higher calcium concentration and lower horse serum concentration (DMEM in Fig. lA), the first myotubes appeared within a few hours, and 20 h later (68 h) newly-formed myotubes were numerous, although aligned fusion-competent myoblasts were still present. During this intense period of fusion, the resting [Ca’+], was significantly higher both in interacting myoblasts and newly-formed myotubes (125 + 5 nM and 150 f 4 nM, respectively), compared to myoblasts at the same time maintained in growth medium (HAM). When the growth medium was exchanged at 68 h for the fusionpromoting medium, resting [Ca’+], was significantly increased (128 + 2 nM) in myoblasts after 8 h of exposure (Mb 76 h). This observation is consistent with the idea that [Caz+li increases slightly before the fusion process, and shows that newly formed myotubes have not yet down-regulated their basal [Caz+],. During the following period of cytodifferentiation and maturation of myotubes, [Ca2+li progressively decreased with time to values around 100 nM (1 day, 146 + 10 nM; 2-3 days, 119 + 6 nM. 4-6 days, 108 * 4 nM [16]). Since cytochalasin B (CB) is an actin-disrupting agent well known to block cell fusion without preventing cytodifferentiation, [ 13,202 1],48 h myoblasts were exposed to 2.5 )LM CB supplemented with the differentiation-promoting medium (DMEM + low serum). 20 hours later (68 h), the basal [Ca2+li of fusion-arrested myoblasts (Fig. lB, DMEM CB) was not significantly elevated (83 * 7 nM) compared to 68 h myoblasts (80 + 8 nM) maintained in the growth medium (Fig lB, HAM). From this and the previous observation (Fig. lA), a tight relation between the regulation of the resting [Caz+], and the fusion process is strongly suggested: stimulating fusion of confluent myoblasts (Fig. 1A) was accompanied by an increase in the basal [Ca2+], while the blockade of fusion was correlated with a prevention of the rise in resting [Ca2+], (Fig. 1B). The elevation of cytosolic [Ca2+], could not be due to passive mechanisms following the elevation of extracellular calcium in culture medium (from 300 ).tM to 1.8 mM), because all the measurements of cytosolic [Ca2+li were not done in the culture medium. For all types of cell, measurements were preceded by 1 h of incubation in a saline solution containing 2.5 mM CaCl, (for loading the calcium probe) and were performed in the same saline solution. Thus, if cytosolic [Ca2+], were to be passively correlated to extracellular 0 Pearson

Professional

Ltd 1996

Calcium influx and myoblast

[Ca*+], the [Caz+li should be the same in all cells. It is clear that the differences in [CaZ+],are due to differences in calcium handling and that muscle cells strongly regulate the resting level of cytosolic calcium, whatever the external calcium concentration: 68 h-old CB-treated cells in the presence of 1.8 mM CaCl, maintained the [Ca2+], at a low level, as done by 68 h-old myoblasts in the presence of 300 @I external CaCl,. Moreover, [Ca2+li is progressively decreased after the fusion period in myotubes differentiating in the presence of 1.8 mM CaCI, (150 nM at 20 h after fusion stimulation, in Fig. lA, to 108 nM at 4-6 days, in [ 161). Since myoblasts seems to actively elevate the basal [Ca2+li during the fusion process, it was therefore interesting to examine the consequence of buffering [Caz+], with a cell permeant calcium chelator to test its effects on the fusion process. First of all, cells were loaded with the membrane permeant form of BAPTA and the effects on the resting [Ca”+], were tested after 20 h of exposure to the fusion-promoting medium. This allowed the buffering of the [Ca*+], without affecting the [Ca2+10, which is known to regulate myoblast fusion, partly by mechanisms acting at the external face of the sarcolemma [5]. Figure 1B shows that previous incubation with 15 @l BAPTA/AM (DMEM BAP’WAM) significantly reduced [Ca2+], (111 f 7 r&I) in respect to the [Ca2+li (136 + 9 nM) measured in 68 h fusion-competent myoblasts (Fig. lB, DMEM), demonstrating that changes in [Ca2+li are attenuated due to the presence of BAPTA in the cytoplasm. Further attempts to prevent more drastically the increase in resting [Caz+], by augmenting the concentration of BAPTA!AM appeared unsuccessful because of cytotoxic effects. In Figure 1C a pharmacological approach was used to tentatively elucidate the potential involvement of DHPsensitive VDCC in leading to the high resting [Ca2+11 increase in fusion-competent myoblasts. The organic VDCC blockers, PN 200-l 10 and nifedipine were used at concentrations (1 @I and 5 l.t.M, respectively) which have been demonstrated to fully suppress L-type calcium currents in cultured rat muscle cells 122,231. These drugs failed to significantly prevent the increase in the resting [Caz+li, since the mean resting [Ca2+], values measured at 68 h (143 f 8 nM and 132 + 6 nM for PN 200-l 10 and nifedipine, respectively) were not significantly different from values (130 f 6 nM) obtained with control fusioncompetent myoblasts (Fig. lC, DMEM). The action of internal calcium buffering and organic VDCC blockers on the fusion process was then explored by evaluating the fusion index after 20 h in fusion-promoting medium (68 h), using the two following control conditions as reference: (i) myogenic cells grown in fusion-promoting medium as positive control (Fig. 2, CTRL); and (ii) muscle cells grown in a low calcium medium (Fig. 2A, LCA) as a negative reference, the blocking d Pearson

Professional

Lfd 1996

fusion

369

CTRL

LCA

CTRL

BAPTAIAM 15pY

CTRL

NIF

CTRL

PN

Fig. 2 Determination of the fusion index at 68 h performed in different culture conditions. Graduated shading bargraphs: cultures in control fusion-promoting medium (CTRL). Black bargraphs: treated cultures. (A) Blockage of fusion with a ‘low-calcium medium’ (LCA) containing 50 pM CaCI,. (B) Reduction of the fusion index after 2 h incubation with 15 uM of the calcium buffer BAPTA in its cell permeant form, and 20 h in the fusion-promoting medium. (C,D) No changes in the fusion index after incubation for 20 h in the fusion-promoting medium supplemented with 5 uM nifedipine (NIF), or 1 PM PN 200-110 (PN). Bargraphs represent mean values in percentage f SEM obtained from several different cultures: n = number of cultures.

effect of low extracellular calcium being well documented [4]. At 68 h, the fusion process was drastically and almost totally blocked by culture in LCA medium (0.7 f 0.06% against 28 + go/o), and was considerably more efficient than the presence of BAPTA/AM (18 f 4% instead of 25 f 5%) in the differentiation medium (Fig. 2C). This latter effect matched well with the limited effect on resting [Ca”+], observed in the same culture conditions (Fig. 1B). On the other hand, the presence of 1 @I PN 200-l 10 or 5 @vI nifedipine (Fig. 2C,D) did not sign& cantly modify the fusion index at 68 h (respectively, 35 + 8% against 34 & 5% and 30 + 8% against 3 1 * 60/o),which is in agreement with their inability to prevent a high resting [Ca2+], (Fig. 1B). The obvious correlation between resting [Caz+li changes and formation of myotubes is in Cell Calcium

(1996)

19(5).

365-374

370

13Constantin, C Cognard, G Raymond

30Nal5Ca

100pA

1 -

OmV

100 ms

B

ONa/5Ca -40 mV OmV

Fig. 3 Expression of sodium currents but not calcium currents in myoblasts cultured for 20 h in fusion-promoted medium. Examples of whole-cell ionic currents (patch-clamp recordings) elicited by square depolarizing pulses from a holding potential of -90 mV to different values indicated near the traces. (A) Currents were recorded from a 66 h aligned myoblast in a control condition containing 30 mM NaCl and 5 mM CaC!,. (B) The sodium-containing bath solution was changed to a sodium free medium containing 5 mM CaCI,. (C) Calcium ions present in the sodium-free medium were replaced by barium ions (2.5 mM).

agreement with the involvement of a cytosolic calcium signal for the fusion of myoblasts. However, these results do not support the involvement of calcium entry through DHP-sensitive VDCC in either the resting [Ca*+], changes, or in the fusion process. Although we had previously reported that, after 48 h in growth medium, myoblasts did not present any voltage-dependent calcium currents [ 121, voltage-dependent L-type calcium currents were detected soon after fusion in newly-formed myotubes containing 2-9 nuclei [ 121. Therefore, the question of a rapid expression of functional VDCC concomitant to the triggering of fusion is yet to be answered. Thus, voltage-dependent calcium currents were searched for in aligned fusion-competent myoblasts during the first hours of exposure to the fusion-promoting medium. The classic whole-cell configuration of the patch-clamp technique was used in some Cell Calcium

(1996)

19(5),

365-374

cases for comparisons with currents obtained in previous work [ 121. However, because of the small dimensions of most of the spindle-shaped myoblasts, the rupture of the patch membrane under the pipette tip rapidly resulted in total wash out of the cytoplasm by the intrapipette solution. To maintain more physiological conditions and to keep the intracellular regulation systems intact, current recordings were performed by means of the perforated patch-clamp technique using amphotericin B. The presence of the sodium currents previously demonstrated in myoblasts [12,24-261 was used to check the proper access to the cell interior. After several minutes, the rising rate and the maximum amplitude of sodium currents became stable, attesting to a proper recording of the whole-cell currents. From Figure 3A, which shows the inward sodium current obtained at -40 mV (at the peak of current amplitude) and at 0 mV (at a membrane potential where calcium currents are expected to be maximum), it is clear that no signs of slow and long-lasting inward currents were apparent. When sodium ions were withdrawn from the ionic bath solution (0 Na/5 Ca), rapid inward currents disappeared at all potentials (from -80 mV to 60 mv), attesting to the nature of these currents being sodium (Fig. 3B). Although this saline bath solution was designed to reveal inward calcium currents from other ionic conductances (absence of both external and internal Na+ preventing any sodium current recording, and K+ currents blocked by external tetraethylammonium and internal Cs+), no transitory or long-lasting calcium currents were detectable in all fusion-competent myoblasts (18 cells tested, 7 in whole-cell and 11 in perforated patch). Even when Ca2+ ions were changed for Ba2+, (5 cells tested) to enhance currents through L-type VDCC, a slow inward current was not detected (Fig. 1C). The ability to trigger Ca2+ influx via voltage-dependent mechanisms in fusion-competent myoblasts was tested by a second approach using the ratiometric cytofluorescence method with the calcium indicator Indo-1, and a hyperpotassium solution to abruptly depolarize the cell membrane. As previously reported for 100 mM KC1 pulses applied to 48 h myoblasts cultured in growth medium [ 161, the application of 100 mM KC1 failed to induce any detectable calcium elevation in 68 h fusioncompetent myoblasts cultured 20 h in the fusion-promoting medium. The absence of calcium transients in response to hyperpotassium solution can be interpreted as the absence of voltage-dependent mechanisms involved in calcium mobilization. The absence of calcium currents suggests the absence of functional VDCC. On the contrary, 52% (n = 23) of the 68 h fusion-competent myoblasts responded to acetylcholine (Ach) application by a slow calcium transient (Fig. 4A) as for 48 h myoblasts. However, responses obtained in 68 h fusioncompetent myoblasts were more rapid and larger than in 0 Pearson

Professional

L td 1996

Calcium influx and myoblast fusion

371

10 pM ACH I

A

0 Ca/EGTA

B [F$l

n 50 1

20 s

Fig. 4 Examples of extraceltular calcium-dependent calcium transient induced by acetylcholine in a 68 h-old aligned myoblast cultured for 20 h in fusion-promoting medium. (A) Application of 10 PM acetylcholine (black bar) induced a transitory increase in cytosolic free calcium concentration recorded by a ratiometric fluorescence method. (B) Perfusion of a free calcium solution (white bar) shorten the acetylcholineinduced calcium transient.

early 48h myoblasts, and were similar in rise-time (around 90 nM/s) and increment amplitude (around 370 nM) to those observed in l-day-old newly-formed myotubes [ 161. Since these cells did not significantly respond to hyperpotassium depolarization, these Ach-induced calcium transients likely reflected calcium entry through voltage-independent pathways, like the acetylcholineoperated nicotinic channels. The idea of a calcium influx is supported by the observation that calcium increase induced by Ach perfusion could be drastically reduced by application of an external free-calcium solution supplemented with 5 mM EGTA (Fig. 4B). DISCUSSION

The first features reported in this study provide evidence for a progressive decrease of the resting [Caz+], when cultured myoblasts are left in growth medium, going from a high level at 24-48 h to a lowered one in older confluent myoblasts (68 h). This decrease of [Ca’+], could reflect the progressive control of free calcium concentration as a result of myoblast commitment into the cytodifferentiation program. It is known that myoblast in culture first actively replicate 11,271and finally withdraw from the cell cycle 128,291 regardless of the presence of mitogenic agents [30,31]. Just prior to fusion (at 65 h of culture), rat myoblasts have been shown 1321 to express in the cytoplasm, calsequestrin and the Ca2+ pump/Mg2+-dependent ATPase, two sarcoplasmic reticulum proteins involved in Q Pearson

Professional

Ltd 7996

calcium regulation in muscle cells. It is thus probable that the progressive decrease in [Ca2+], may be related to the appearance of calcium uptake and storage functions as muscle cells undergo terminal differentiation. This study reinforces previous conclusions reported by David et al. [9] and Przybylski et al. [5], providing direct evidence for a resting [Caz+li elevation in fusion-competent myoblasts and newly-formed myotubes. As already proposed, this elevated resting [CaZ+], seems to be a prerequisite for myoblast fusion since partial buffering of internal calcium ions reduced myotube formation. Moreover, the absence of resting [Ca2+], elevation in fusion-arrested cells reinforces the idea of a correlation between the fusion process and intracellular calcium signailing. One can imagine that this prerequisite cytosolic calcium elevation is involved in mechanisms providing the signal for myoblast fusion and/or in controlling the machinery necessary for the fusion of adjacent membranes. Interestingly, Lee et al. [33] have recently proposed a possible target for the calcium signal that may activate myoblast NO synthetase to produce NO which, in turn, activates guanylate cyclase for cGMP production. These authors had previously observed a Caz+-dependent increase in cGMP concentration occurring transiently just prior to fusion [34] and this could be an important secondary messenger for fusion signalling. In addition, the increase in [Ca2+li may also activate calcium-dependent intracellular proteases which may be involved in cytoskeletal reorganisation occurring during fusion 1351. Cell Calcium

(1996)

19(5),

365-374

372

B Constantin, C Cognard, G Raymond

The present results, however, disagree with the earlier conclusions of Entwistle et al. [lo] who proposed the participation of functional VDCC in mediating a [Ca*+], increase. First of all, the culture of myoblasts in the continuous presence of organic blockers of DHP-sensitive VDCC did not affect either the myotube formation or the resting [Caz+], increase in fusion-competent myoblasts. On the contrary, it has been reported that the phenylalkylamine D600, at concentrations from 200 I&I to 2 mM [lo], 30-50 PM 191,or 20 FM [34], decreased fusion of embryonic chick myoblasts. Although this phenylalkylamine binds to a high affinity site of L-type VDCC, the large concentrations used in these studies could have affected lower affinity membrane sites including Na+,K+ channels (seereview [36]) and ACh-gated nicotinic channels 1371.Indeed, David et al. 191observed that D600 only partially blocked net calcium influx, even at 50 l&l. It is thus possible that D600 reduction of the fusion process may be related to interactions of the drug with low-affinity membrane sites other than DHP-sensitive VDCC. Entwistle et al. [lo] have also shown that lanthanum reduced the fusion process. However, this trivalent ion is not a specific blocker of VDCC and Przybylski et al. [5] have demonstrated that La3+ blocked the fusion process without blocking the [Caz+], increase, but by acting on surface membrane targets. Secondly, we did not find any DHP-sensitive L-type voltage-dependent current which could mediate D600-sensitive Caz+ influx in mononucleated myoblasts committed to fuse, even when calcium ions were replaced by Ba2+. Fusion competence and appearance of L-type calcium currents can be independent programs [ 131, but these two events seem to be tightly time coordinated after medium change, myoblast fusion occurring earlier than functional L-type VDCC appearance. Alternatively, calcium influx could be mediated by the early appearance of T-type calcium currents. However, this type of voltage-dependent calcium current was never found in fusion-competent myoblasts, and we have previously shown that it appears in myotubes later than the L-type one [12]. Thirdly, no other sources of depolarization-operated calcium rise was found in fusioncompetent myoblasts which failed to respond to the application of depolarizing hyperpotassium solution. This observation suggests the absence of functional voltagedependent mechanisms involved in calcium mobilization, like a transmembrane calcium influx through VDCC or a voltage-dependent release of calcium from internal stores. Since myoblasts seem to actively regulate calcium homeostasis by controlling plasma membrane Ca*+ transport and uptake by intracellular stores, an elevation of the basal activity of calcium in the cells will be due to changes in one or several components of these systems. One of these changes could be a modification in the entry of calcium ions through nicotinic cationic messengerCell Calcium

(1996)

19(5),

365-374

operated channels, due to changes in the concentration of the messenger or changes in the intrinsic activity of the channels. Indeed, external calcium-dependent calcium transients were obtained in response to acetylcholine, which suggested the possible influx of calcium ions through nicotinic receptor-channels. This is consistent with the results from Decker and Dani [38] who concluded from single patch-clamp experiments that the calcium influx through nicotinic acetylcholine receptorchannels is significant. Numerous studies have shown that mononucleated myoblasts express functional acetylcholine receptors prior to fusion [ 16,39-431 and they are up-regulated during cytodifferentiation (see review [44]). After withdrawal of mitogens, the responses presently obtained in 68 h adjacent myoblasts were larger and more rapid than those obtained in 48 h myoblasts cultured in growth medium [16]. This testified to the commitment of 68 h-old myoblasts and could reflect an increased expression of functional nicotinic channels involved in acquisition of fusion competence. Indeed, Entwistle and collaborators [42] have shown that acetylcholine receptors are involved in the spontaneous fusion of embryonic chick myoblasts and that these mononucleated cells specifically bound an ACh antibody. This suggests that myoblasts are able to synthesize ACh and probably provide an autocrine control of the fusion process. Moreover, preliminary observations have shown that the addition of exogenous Ach (lo6 M) potentiated the fusion process of rat myoblasts. This effect seemed to involve nicotinic receptors, since the presence of a-bungarotoxin ( 1O-8 M), blocked the stimulation of fusion by Ach. Although ACh-induced [Ca*+], increase does not stem from membrane depolarization, calcium entry through transmitter-gated cationic channels could significantly contribute to a calcium signal. This could also be amplified by ACh-induced mobilization from InsP,-sensitive stores as has been proposed for C2C12 myotubes [43]. Auxiliary mechanisms could be provided by the presence of receptors to ATP [45,46], since ATP-activated cation channels are present in the membrane of chick myoblasts and permeabilities predict that the calcium influx mediated by these receptor-gated channels could lead to significant rise in [Caz+li [47]. Provided that calcium mobilization overcomes the intracellular calcium buffering capacities, many other pathways could contribute to increase the resting [Caz+li prior to fusion, such as prostanoid-mediated activation of InsP,-sensitive calcium pools [42], mechano-sensitive cationic channels [48,49], or various voltage-independent calcium channels which are already implicated in the regulation of cell growth and differentiation (see review [50]). The present observations do not, however, rule out other voltage-dependent pathways for the regulation of the fusion process as suggested by results from Entwistle 0 Pearson

Professional

Ltd 1996

Calcium influx and myoblast fusion

et al. [42] showing a dependence of myoblast fusion on the external concentration of K+. For instance, experiments performed with ouabain have suggested an involvement of the Na+/K+ pump and the control of the resting membrane potential [51]. Finally, Widmer et al. [52] have recently demonstrated the contribution of a TEA-sensitive potassium permeability to the fusion process which could be partly mediated by a voltagegated potassium current, analogous to the delayed rectifier current. With regard to these numerous features, the increase in [Ca2+], is probably only one of many confluent processes, such as membrane hyperpolarization and increases in CAMP and cGMP levels, which contribute to the initiation of fusion.

12.

13.

14.

15.

16.

ACKNOWLEDGEMENTS

The authors thank Francoise Mazin for expert technical assistance and Michael Patterson for correcting the English. This work was supported by grants from CNRS (LJRA 1869), Universite de Poitiers and Association Francaise contre les Myopathies. REFERENCES

Professional

18.

19.

1. Konigsberg LR. The differentiation of cross-striated myoflbrils in short term cell culture. &p Cell Res 1960; 21: 414-420. 2. Yaife D. Retention of differentiation potentialities during prolonged cultivation of myogenic cells. Proc Natl Acad Sci USA 1968; 61: 477-483. 3. Wakelam M.J.O. The fusion of myoblasts. Bio&m/ 1985; 228: 1-12. 4. Shainberg A., Yagil G., Yaffe D. Control of myogenesis in vitro by Ca 2 1 concentration in nutritional medium. Exp Cell Res 1969; 58: 163-167. 5. Przybylski R J., Szigeti V., Davidheiser S., Kirby A.C. Calcium regulation of skeletal myogenesis. IL Extracellular and cell surface effects. Cell Calcium 1994; 15: 132-142. 6. Knudsen K.A., Myers L., M&wee S.A., A role for the CaZ+dependent adhesion molecule, N-cadherin, in myoblast interaction during myogenesis. Exp Cell Res 1990; 188: 175-184. 7 Mege R.M., Goudou D., Diaz C. et al. N-cadherin and N-CAM in myoblast fusion: compared localisation and effect of blockade by peptides and antibodies. J Cell Sci 1992; 103: 897-906. 8. Lognonne J.L.,Wahrmann. Spontaneous myoblast fusion is mediated by cell surface Ca2+-dependent protein kinase(s). Exp CellRes1986; l&%340-356. 9. David J.D.,See W.M., Higginbotham C. Fusion of chick embryo skeletal myoblasts: role of calcium influx preceding membrane fusion. Da, Bioll98 1; 82: 297-307. 10. Entwistle A., Zalin RJ., Bevan S.,Warner A.E. The control of chick myoblast fusion by ion channels operated by prostaglandins and acetylcholine. J Cell Bioll988; 106: 1693-1702. Il. Romey G., Garcia L., Dimitriadou V., Pit-icon-Raymond M., Rieger F.,Lazdunski M. Ontogenesis and localization of CaZ+ channels in mammalian skeletal muscle in culture and role in excitation-contraction coupling. Proc Nat1 Acad Sci USA 1989; 86:2933-2937 0 Pearson

17

Ltd 1996

20. 2 1. 22.

23.

24.

25. 26. 27 28.

373

Cognard C., Constantin B., Rivet-I&tide M., Imbert N., Besse C., Raymond G. Appearance and evolution of calcium currents and contraction during the early post-fusional stages of rat skeletal muscle cells developing in primary culture. Deuelupment 1993; 117: 1153-1161. Constantin B., Cognard C., Raymond G. Myoblast fusion is not a prerequisite for the appearance of calcium current, calcium release, and contraction in rat skeletal muscle cells developing in culture. &p CefI Res 1995; 217: 497-505. Caffrey J.M., Brown A.M., Schneider M.D. Mitogens and oncogenes can block the induction of specific voltage-gated ion channels. Science 1987; 236: 570-573. Caffrey J.M., Brown A.M., Schneider M.D. CaZ+and Na+ currents in developing skeletal myoblasts are expressed in a sequential program: reversible suppression by transforming growth Beta-l, an inhibitor of the myogenic pathway. J Neurosci 1988; 9: 3443-3453. Cognard C., Constantin B., Rivet-Bastide M., Raymond, G. Intracellular calcium transients induced by different kinds of stimulus during myogenesis of rat skeletal muscle cells studied by laser cytofluorimetry with Indo-I. Cell Calcium 1993; 14: 333-348. Grynkiewicz G., Poenie M., Tsien R.Y. A new generation of Caz+ indicators with greatly improved fluorescence properties. J Biol ckw 1985; 260: 3440-3450. Hamill O.P.,Marty A., Neher E., Sakmann B., Sigworth FJ. Improved patch-clamp techniques for high-resolution current recording from cell-free membrane patches. pfliigen Arck 198 1; 391: 85-100. Horn R., Marty A. Muscarinic activation of ionic currents measured by a new whole-cell recording method. / Gen Pkysiol 1988; 92: 145-159. Holtzer H., Strahs K., Biehl J., Somlyo A.P., Ishikawa H. Thick and thin filaments in postmitotic, mononucleated myoblasts. Science 1975; 188: 943-944. Cantini M., Sartore S., Vitadello M., Schiaffino S. Development of the sarcotubular system in fusion-arrested myoblasts. Cell Biol IntRq 1979; 3: 151-156. Cognard C., Romey G., Calizzi J-P., Fosset M., Lazdunski M. Dihydropytidine-sensitive CaZ+channels in mammalian skeletal muscle cells in culture: electrophysiological properties and interactions with Ca2+ channel activator (Bay K8644) and inhibitor (PN 2.00-l 10). Proc Nat2 Acad Sci USA 1986; 83: 1518-1522. Cognard C., Rivet M., Raymond G. The blockade of excitation/contraction coupling by nifedipine in patch-clamped rat skeletal muscle cells in culture. @tigers Arch 1990; 416: 98-105. Frelin C., Vijverberg H.P.M., Romey G., Vigne P., Lazdunski M. Different functional states of tetrodotoxin sensitive and tetrodotoxin resistant Na+ channels occur during the in vitro development of rat skeletal muscle. pflgers Arch 1984; 402: 121-128. Gonoi T., Sherman S., Catteral W.A. Voltage clamp analysis of tetrodotoxin-sensitive sodium channels in rat muscle cells developing in vitro. J Neurosti 1985; 5: 2559-2564. Weiss R.E.,Horn R. Functional differences between two classes of sodium channels in developing rat skeletal muscle. Science 1986; 233: 361-364. Yaife D. Development changes preceding cell fusion during muscle differentiation in vitro. .E%rrCell Res 1971; 66: 33-48. Okazaki K., Holtzer H. Myogenesis: fusion, myosin synthesis, and the mitotic cycle. Proc Nat1 Acad Sci USA 1966; 56: 1484-1490.

29. Bischoff R., Holtzer H. Mitosis and the processes of Cell Calcium

(1996)

19(5),

365-374

374

30. 3 1.

32.

33. 34. 35.

B Constantin,

C Cognard,

G Raymond

differentiation of myogenic cells in vitro. J CeU Viol 1969; 41: 188-200. Emerson C., Beckner S. Activation of myosin synthesis in fusing and mononucleated myoblasts. / Mol Bioll975; 93: 43 l-447. Lin Z., Lu M-H., Schultheiss T. et al. Sequential appearance of muscle-specific proteins in myoblasts as a function of time after cell division: evidence for a conserved myoblast differentiation program in skeletal muscle. Cell Motil Cytos~elefon 1994; 29: 1-19. Jorgensen A.O., Kalnins, VI., Zubrzycka E., MacLerman D.H. Assembly of the sarcoplasmic reticulum. Localization by immunofluorescence of sarcoplasmic reticulum proteins in differentiating rat skeletal muscle cell cultures. J Cell Bioll977; 74: 287-298. Lee K.H., Back M.Y., Moon K.Y. et al. Nitric oxide as a messenger molecule for myoblast fusion. / Biol Chem 1994; 269: 14371-14374. Choi S.W.,Back M.Y., Rang M-S. Involvement of cyclic cGMP in the fusion of chick embryonic myoblasts in culture. Fxp Cell Res 1992; 199: 129-133. Kwak K.B., Chung S.S.,Kim O-M., Kang M-S., Ha D.B., Chung C.H. Increase in the level of m-calpain correlates with elevated cleavage of fdamin during myogenic differentiation of embryonic muscle cells. Biochim Biopkys A& 1993; 1175: 243-249.

36. McDonald T.F.,Pelzer S., Trautwein W., Pelzer D.J.Regulation and modulation of calcium channels in cardiac, skeletal, and smooth muscle cells. Physiol Rev 1994; 74: 365-502 37. Miledi R., Parker L. Blocking of acetylcholine induced channels by extracellular or intracellular application of D600. Proc R Sot Land 1980; 211: 143-150. 38. Decker E.R., Dani J.A. Calcium permeability of the nicotinic acetylcholine receptor: the single-channel calcium influx is significant. JNeurosti 1990; 10: 3413-3420. 39. Vogel Z., Sytkowski A.J.,Nirenberg M.W. Acetylcholine receptors of muscle grown in vitro. Proc Null Acud Sci USA 1972; 69: 3180-3184.

Cell Calcium (1996) 19(5), 365-374

40. Patrick J., Heinemann SF., Lindstrom J., Schubert D. Appearance of acetylcholine receptors during differentiation of a myogenic cell line. Pm Natl Acad Sci USA 1972; 69: 2762-2766. 4 1. Smilovitz H., Fishbach G.D. Acetylcholine receptors on chick mononucleate muscle precursor cells. Dew Bioll978; 66: 539-549. 42. Entwide A., Zalin RJ., Warner A.E., Bevan S. A role for acetylcholine receptors in the fusion of chick myoblasts. / Cell Biol 1988; 106: 1703-1712. 43. Grassi F.,Futile S., Eusebi F. Ca2+ signalling pathways activated by acetylcholine in mouse C2C12 myotubes. pfriigers Arck 1994; 428: 340-345. 44. Fambrough D.M. Control of acetylcholine receptors in skeletal muscle. Pkysiol Rev 1979; 59: 165-227. 45. Kolb H-A., Wakelam J.O. Transmitter-like action of ATP on patched membranes of cultured myoblasts and myotubes. Nature 1983; 303: 645-648. 46. Hume RI., Honig M.G. Excitatory action of ATP on embryonic chick muscle. J Neurosci 1986; 6: 681-690. 4% Thomas S.A. Hume R.I. Permeation of both cations and anions through a single class of ATP-activated ion channels in developing chick skeletal muscle. J Gen Physioll990; 95: 569-590. 48. Franc0 A., Lansman J.B. Stretch-sensitive channels in developing muscle cells from a mouse cell line. JPkysioll990; 427: 361-380. 49. France-Obreg6n A., Lansman J.B. Mechanosensitive ion channels in skeletal muscle from normal and dystrophic mice. / PhysioZl994; 481: 299-309. 50. Felder CC., Singer-Lahat D., Mathes C. Voltage-independent calcium channels. Regulation by receptors and intracellular calcium stores. Biockem Pkumaco~ 1994; 48: 1997-2004. 5 1. Pauw P.G., Hermann GJ. Ouabain is a reversible inhibitor of myogenic fusion. In Vitro CeU Dev Bioll994; 30A: 9- 1 I. 52. Widmer H., Hamann M., Barofflo A., Bijlenga P., Bader C.R. Expression of a voltage-dependent potassium current precedes fusion of human muscle satellite cells (myoblasts). J Cell Pkysiol 1995; 162: 52-63.

0 Pearson ‘Professional LM 7996