IP₃ receptor function and localization in myotubes: an unexplored Ca²⁺ signaling pathway in skeletal muscle

In multiple types of mammalian cells the level of intracellular Ca²⁺ controls an enormous number of functions, including division, secretion, motility and contractility. In skeletal muscle, when one thinks of Ca²⁺ signals, one thinks of the exquisite control of excitation-contraction (E-C) coupling involving dihydropyridine receptors (DHPRs) and ryanodine receptors (RyRs), the Ca²⁺ release channels of the sarcoplasmic reticulum (SR). If skeletal muscle uses Ca²⁺ as a second messenger to control functions other than those of contraction, it is likely that both the kinetics of the signals and the subcellular location, as well as the nature of the Ca²⁺ release channels, will differ from those linked to E-C coupling. In many cells, intracellular increases in Ca²⁺ are mediated by the inositol 1,4,5-trisphosphate (IP₃) cascade (Berridge, 1993); the activation of various receptors in different cells induces the hydrolysis of membrane phosphatidylinositol 4,5-bisphosphate to generate IP₃ and diacylglycerol, which are activators of intracellular Ca²⁺ channels and protein kinase C, respectively.

The skeletal muscle fiber possesses the basic molecular machinery for a functioning IP₃ messenger system, including phosphatidylinositol-4-kinase, phosphatidylinositol-4 phosphate-5-kinase, phospholipase C, inositolphosphate phosphatases, IP₃ kinase and G-proteins (Carrasco and Figueroa, 1995; Carrasco et al., 1988; Carrasco et al., 1993; Carrasco et al., 1994; Carrasco et al., 1997; Hidalgo et al., 1986; Sanchez et al., 1991; Scherer et al., 1987). Some of these enzyme activities, particularly that of phosphatidylinositol-4 phosphate-5-kinase, have been localized to the transverse tubule (T-tubule) system (Carrasco et al., 1988). In addition, IP₃-sensitive Ca²⁺ channels have been found in the SR (Suarez-Isla et al., 1988; Volpe et al., 1986). Thus, depolarization of the T-tubule system could lead to release of IP₃ and an ensuing release of internal Ca²⁺ in addition to the Ca²⁺ released through the RyRs. Indeed, in skeletal muscle cells, Ca²⁺ release from internal stores may follow more than one set of kinetics and may have more than one function. Calcium waves, unrelated to Ca²⁺ spikes involved in E-C coupling, have been reported in chick and rodent myotubes (Flucher and Andrews, 1993; Powell et al., 1996). We have described the complex Ca²⁺ release response, induced by elevated K⁺, involving two components with different kinetics (Jaimovich and Rojas, 1994; Jaimovich et al., 2000; Estrada et al., 2000; Estrada et al., 2001). These components included a fast Ca²⁺ transient, associated with E-C coupling, and a slower transient with as yet poorly defined causes and unknown functions. Also, depolarization of cultured muscle leads to an increase in IP₃ mass (Jaimovich et al., 2000). It is interesting to note that slow Ca²⁺ signals have a distinct nuclear component (Jaimovich et al., 2000; Estrada et al., 2000; Estrada et al., 2001) and appear to be associated with IP₃. The presence of IP₃ receptors (IP₃Rs) has been shown both biochemically (Liberona et al., 1997; Liberona et al., 1998) and immunocytologically (Jaimovich et al., 2000; Estrada et al., 2001) in both normal and diseased cultured muscle, further suggesting a role for IP₃ signals. Finally, biochemical localization of IP₃Rs to isolated nuclei of developing muscle (Liberona et al., 1998; Jaimovich et al., 2000) as well as the presence of Ca²⁺ signals in cells lacking RyRs (Estrada et al., 2001) further suggest a nuclear involvement in the IP₃ cascade. To explore the role of IP₃Rs in the propagation of such Ca²⁺ transients and, perhaps, in the modulation of skeletal muscle development, we have investigated the expression and subcellular localization of these receptors in cultured muscle. We have further explored the link between IP₃Rs and slow Ca²⁺ signals, and searched for targets of these signals in muscle cells.

Cells for primary mouse skeletal muscle were prepared as described (Flucher et al., 1991). Briefly, hindlimb muscle from newborn mice was minced, dissociated in 0.125% trypsin (Life Technologies) in Ca²⁺-Mg²⁺-free balanced saline solution, centrifuged and resuspended in complete medium [10% horse serum, 10% fetal bovine serum (FBS), 2% chick embryo extract in Dulbecco’s modified Eagle’s (DME) medium (Life Technologies)]. Cells were plated at 1×10⁴ cells/35 mm primaria culture dish (Falcon).

Cells for primary rat skeletal muscle culture were obtained by collagenase treatment of neo-natal rat hindlimb muscle, as previously described (Jaimovich and Rojas, 1994). Well-differentiated, contracting myotubes were collected between 7 and 10 days of culture for RT-PCR studies and for IP₃R isotype experiments, and from 6 to 7 days for Ca²⁺ signal and signal-transduction studies. C2C12 cells were maintained in DME:F12 (1:1) supplemented with 10% FBS and 0.5% chick embryo extract. To induce differentiation, the medium was changed to DME:F12 supplemented with 1% FBS and 2% horse serum.

The reagents and kits for RT and PCR procedures were obtained from Life Technologies, Inc. unless specified. Total RNA was prepared by TRIzol extraction. Five μg of RNA were used as a template for first-strand cDNA synthesis in a 20 μl-reaction with SuperScript II reverse transcriptase and random hexamers or oligo dT primers. PCR was performed using 10% of this reaction product as template either with a primer pair that amplified all three cDNA isoforms (co-amplification), or with primer pairs specific for rat IP₃R types 1, 2 and 3 (IP₃R-1, -2 and -3), (separate amplification) both amplifying products from a region flanking the alternative splicing S II sequence of rat IP₃R-1.

The primers common to the three receptors were: forward 5′ GAA GAG AAA CTG TGC ATT 3′ and reverse 5′ CAG GAT TGG CTG CAT GAT 3′ corresponding to bases 5330-5347 and 6222-6239 of rat IP₃R-1, respectively (Mignery et al., 1990). The forward nucleotide sequence aligns with bases 5244-5261 from type 2 and with bases 5109-5126 from type 3. The reverse primer aligns with bases 5998-6015 from type 2 and with bases 5742-5759 from IP₃R-3. These primers amplified products of 800, 700 and 650 bp.

To amplify separately the three isoforms, the forward sequence 5′ GAA GAG AAA CTG TGC ATT 3′ was kept. The reverse primers were: 5′ GGC AAT GGT CCA CTA TCA 3′ (5606-5623 type 1); 5′ CTG GAC AGG ATG CCG ATA 3′ (5437-5454 type 2) and 5′ TCC GCC ATC GCA GCC ACC 3′ (5289-5306 type 3). The amplified products for types 1, 2 and 3 IP₃R are 180, 214 and 200 bp, respectively. PCR conditions were: initial denaturation at 94°C for 5 minutes, followed by 30 cycles of denaturation at 94°C for 1 minute, annealing at 55°C for 1 minute and extension at 72°C for 2 minutes, with a final extension of 10 minutes.

All PCR products were purified using Wizard PCR Preps kit (Promega, Madison, WI) after electrophoresis on 2% low melting point agarose gel, and quantified by visual comparison with Low DNA Mass Ladder. After PCR amplification with Big Dye Terminator kit (Applied Biosystems, Foster City, CA), purification by precipitation with isopropanol and resuspension in formamide, the nucleotide sequencing was performed by means of a genetic analyzer (model 310, Applied Biosystems, Foster City, CA). The sequences of the PCR products obtained both by separate amplification or by co-amplification corresponded to the expected IP₃R isoforms.

Total cell lysates were obtained from the primary cultures. Cells were washed with PBS and solubilized at 4°C in lysis buffer containing 50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 0.2 mM 4-(2-aminoethyl)benzene sufonyl fluoride (AEBSF), 1 mM benzamidine, 10 μg/ml leupeptin, 1 μg/ml aprotinin and 1 μM pepstatin. After incubation on ice for 20 minutes, cells were scraped from the dishes, sonicated for 1 minute and left on ice for 20-30 minutes. Cellular debris was removed by microcentrifugation at 17,000 g for 20 minutes. The supernatant (total cell lysate) contains the nuclear membrane because of the treatment with detergent.

Nucleus enriched fractions were obtained from the cell line cultures. Cells were frozen in liquid nitrogen for 2 hours, thawed and harvested in a minimal amount of a solution containing: 20 mM Tris-HCl pH 7.5, 1 mM MgCl₂, 1 mM DTT, 1 mM PMSF, 0.1 mM leupeptin and 0.001 mM pepstatin A. The suspension was centrifuged at 1000 g for 10 minutes. To obtain a more purified nuclear fraction, the pellet was resuspended in the same buffer plus 1.2 M sucrose, and centrifuged at 100,000 g for 20 minutes.

For these experiments, 6-7 day rat myotubes were cultured for 24-36 hours in serum-free medium. Cells were washed with Ca²⁺-Mg²⁺-free PBS and maintained in Krebs-Ringer under resting conditions for 30 minutes (in mM: 20 HEPES-Tris, pH 7.4, 118 NaCl, 4.7 KCl, 3 CaCl₂, 1.2 MgCl₂ and 10 glucose). In Ca²⁺-free experiments, no Ca²⁺ and 0.5 mM EGTA were added to the incubation media in both resting and depolarization conditions. Depolarization was induced by changing to a medium containing 84 mM KCl; Na⁺ concentration was decreased proportionally to maintain osmolarity. We have studied extracellular signal-regulated kinases 1 and 2 (ERKs 1/2) and cAMP response element-binding protein (CREB) phosphorylation after depolarization. Rat myotubes were exposed to the elevated K⁺ concentration either for several minutes and then harvested (protocol 1), or the cultures were exposed to the depolarizing medium for 30 seconds to 1 minute, re-fed with resting condition medium, and harvested at different times (protocol 2). When the IP₃R inhibitor 2-APB (2-aminoethoxydiphenyl borate; Aldrich, Milwaukee) was used, the cells were pretreated in the presence of control medium or freshly prepared inhibitor solution (50 μM) for 30 minutes and then depolarized, as above, in the absence or presence of 2-APB. After appropriate time intervals, cells were solubilized at 4°C in 0.1 ml of lysis buffer containing 50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 5 mM Na₃VO₄, 20 mM NaF, 0.2 mM AEBSF, 1 mM benzamidine, 10 μg/ml leupeptin, 1 μg/ml aprotinin and 1 μM pepstatin. After incubation on ice for 20 minutes, cells were scraped from the dishes, sonicated for 1 minute and left on ice for 30 minutes. Nuclear and cellular debris were removed by microcentrifugation at 17,000 g for 20 minutes. The protein concentration of the supernatants was determined using BSA as the standard.

To identify the isotypes of IP₃Rs, proteins (from total lysates of primary cultures or from isolated nuclei of cell lines) were separated by 7% SDS-PAGE and transferred to PVDF membranes (Millipore). After blocking, membranes were incubated with 1:1000 dilutions of polyclonal anti-IP₃R-1 (PA1-901, Affinity Bioreagents, Inc.) or monoclonal anti-IP₃R-3 antibodies (I31220, Transduction Laboratories). After incubation with HRP-conjugated anti-rabbit secondaries (dilution 1:2000) (Pierce) or with HRP-conjugated anti-mouse secondaries (dilution 1:5000) (Sigma), membranes were developed by enhanced chemiluminescence (NEN Life Sciences).

For western blot analysis of P-ERKs or P-CREB, proteins were resolved in 10% SDS-PAGE and transferred to PVDF or nitrocellulose membranes. Primary antibody incubations at dilutions of 1:1000 (P-CREB, 9191S, New England Biolabs) and 1:2000 (P-ERKs, 9101S, New England Biolabs), were carried out at 4°C overnight. After incubation with HRP-conjugated secondary antibodies, membranes were developed by enhanced chemiluminescence. To correct for loading, membranes were stripped and blotted with antibody for total ERKs (dilution 1:2000, New England Biolabs), or total CREB (dilution 1:750, Upstate Biotechnology, Inc.). A densitometric analysis of the bands was performed with Scion Image from NIH.

Cultures were processed for immunocytochemistry essentially as previously reported (Powell et al., 1996). The primary monoclonal antibodies were: anti-α₁ DHPR (1:100) (Chemicon), anti-α-actinin (1:100) (Sigma) anti-skeletal myosin (fast) (1:400) (Sigma), anti-Ca²⁺ATPase (CaF2-5D2, fast twitch SR) (1:2) (Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA). Affinity Bioreagents supplied the anti-IP₃R-1 epitope-affinity purified polyclonal antibody (PA1-901) (1:25, 54 μg/ml), the IP₃R-1 peptide epitope and the pre-immune serum. The polyclonal antibodies to IP₃R-1 were raised against the COOH-terminal domain of the human IP₃R-1 peptide as determined from the published nucleotide sequence, N1829KKKDDEVDRDAPSR-KKAKE1848; IP₃R-1, found in brain, is specifically localized in Purkinje cells of the cerebellum. G. Mignery provided a polyclonal anti-IP₃R-1 antibody raised against the C-terminal residues of the strictly conserved region (Galvan et al., 1999). Positive staining was obtained with 1:25-1:100 dilutions of a 1.0 mg/ml epitope-affinity purified serum, and with use of 2% paraformaldehyde fixation and incubation procedures (Galvan et al., 1999).

Culture dishes were mounted in Vectashield (Vector Laboratories, Inc., Burlingame, CA) for viewing with an IM Zeiss phase contrast microscope equipped with epifluorescence optics or a Leica TCS NT Scanning Laser Confocal Microscope. To obtain confocal images, TRITC and FITC emissions were either collected simultaneously and composite images were created automatically by the software, or the photomultiplier tubes were operated separately and digital overlays were created using Adobe Photoshop™. Controls and scanner adjustments insured that FITC signals did not contaminate the TRITC data. The ‘glow over under’ function of the TCS system was used to minimize the possibility that electronic signal amplification would lead to biases in the data. The images herein were manipulated in Adobe Photoshop™ to improve clarity; no data were added or deleted.

Fluo-3 fluorescence images were obtained from mouse and rat myotubes with an inverted confocal microscope (Carl Zeiss Axiovert 135 M – LSM Microsystems). Myotubes were preloaded in the presence of fluo-3-AM, which was then de-esterified in the cytoplasm. Cells were then incubated in a ‘resting solution’ of the following composition (in mM): 145 NaCl, 5 KCl, 2.6 CaCl₂, 1 MgCl₂, 10 Na HEPES and 5.6 glucose, pH 7.4 containing 5.4 μM fluo-3-AM, for 30 minutes at 25°C. Cells attached to coverslips were mounted in a one-ml capacity perifusion chamber and placed in the microscope for fluorescence measurements after excitation with a 488 nm argon laser. The fluorescent images were collected every second and analyzed frame by frame with the data acquisition program of the equipment. Cells were exposed to high K⁺ solutions (47 mM K⁺) and depolarized by a fast (about 1 second) change of solution by perifusion. Changes in relative fluorescence intensity in particular regions of rat myotubes was measured using previously described software (Estrada et al., 2000). For IP₃ inhibition studies, myotubes were pre-incubated for either 5 minutes or 30 minutes in 50 μM 2-APB or incubated for 20 minutes in the presence of 10 μM U-73122 and depolarized in the presence of the drug.

RT-PCR and immunoblot analyses were performed on rat and mouse primary muscle cultures and on a thoroughly studied mouse muscle cell line. Our major interest in IP₃R expression is in the primary cultures; we are also employing the C2C12 cell line because of its homogeneity. With RT-PCR, using a primer pair that amplifies all three IP₃R isoforms and then separating the forms using 0.8% agarose, we found two bands both for the rat myotubes (Mr) and for the mouse cell line (C2C12 [C2]); CER, used as a control, exhibited their characteristic IP₃R-1 band (Fig. 1A, left). By changing the agarose gel concentration to 3%, three bands were obtained for primary mouse myotubes (Mm) and the C2C12 cell line (Fig. 1A, right). Under these conditions, the band of about 650 bp, corresponding to the IP₃R-3, is more marked than the two other bands in C2C12 cells; this corroborates a recent study of C2C12 cells using quantitative RT-PCR, where type 3 mRNA was shown to be the most abundant isoform (De Smedt et al., 1997). To confirm these results, PCR was performed with primer pairs specific for each of the three isoforms. For rat myotubes in primary culture, three bands were obtained (Fig. 1B). The high resolution of agarose 1000 at 3% concentration allows us to confirm the presence of the three isoforms in rat and mouse primary muscle culture and in the C2C12 cell line. At the level of protein expression, we found evidence for isoforms 1 and 3 (Fig. 1C); the type 2 isoform was not tested because we did not have an effective antibody to type 2. Cerebellar microsomes, used as a positive control for type 1, showed a band of higher molecular weight than that of the cultured cell samples when tested with anti-IP₃R-1 (Fig. 1C). This was expected as the cerebellum IP₃R-1 nascent RNA is spliced differently than IP₃R-1 in other cells (Danoff et al., 1991).

We have previously shown that we recover more IP₃ binding in the nuclear fraction than in the microsomal fraction, which contains the majority of RyR binding in these cultures (Jaimovich et al., 2000; Liberona et al., 1998). Isolated nuclei from C2C12 cells were analyzed, taking advantage of the homogeneity and purity of these preparations. Using primary cell cultures, we have analyzed total homogenates (that include the nuclear membrane, see Materials and Methods) of the rodent cells. In both whole homogenates (Fig. 1C, Mr) and in nuclear fractions (data not shown) of rat myotubes in primary culture, the presence of both types 1 and 3 was detected. The results obtained with the mouse cell line (Fig. 1C, C2) clearly indicate the presence of these two IP₃R isoforms in the nuclear fraction of cells from this species.

All immunocytochemistry was performed on cultured mouse muscle. Of the commercially available antibodies to IP₃R isotypes, types 1 and 3, only anti-IP₃R-1 bound subcellular structures. We have tested the specificity of the commercially available IP₃R-1, epitope-affinity purified antibody (at dilutions of 1:25 and 1:50 of a 1.3 mg/ml sample). It labels young myotubes in the nuclear envelope region and some internal structures within the nucleus, as well as cross striations in the cytoplasmic regions (Fig. 2, top left). The internal nuclear staining (Fig. 2, top left) is not specific for IP₃R as pre-immune serum also binds these structures; however the nuclear envelope region (but not internal nuclear) and cytoplasmic binding by the antibody is eliminated when the antibody is adsorbed with the specific peptide used to raise the antibody (data not shown) (Jaimovich et al., 2000). In differentiated, mature myotubes, the cross-striated pattern revealed by the antibody is eliminated when this antibody is adsorbed with excess of the specific peptide (Jaimovich et al., 2000). Thus, the cross-striated pattern of IP₃R-1 staining is not an artifact. In the present study this antibody bound the cytoplasm of cultured Purkinje cells isolated from 14-day mouse embryonic cerebellum (data not shown). The SR of Purkinje cells is rich in large quantities of IP₃R-1 (Katayama et al., 1996). To confirm further the specificity of the commercial antibody, we used a second polyclonal anti-IP₃R-1 (hereafter referred to as anti-IP₃R-M) provided by G. Mignery (Galvan et al., 1999) and a third polyclonal anti-IP₃R-1 from R. Wojcikiewicz (Wojcikiewicz, 1995). Positive staining in a cross-striated pattern (Fig. 2, bottom left) was obtained when cultures were fixed using Mignery’s paraformaldehyde fixation and staining protocols (Galvan et al., 1999), but not when ethanol fixation was used. In addition, although the nucleus stained with anti-IP₃R-M, the nuclear envelope did not (Fig. 2, bottom left); moreover, the commercial anti-IP₃R, which stained the nuclear envelope of methanol-fixed cultures, did not stain the nuclear envelope when paraformaldehyde fixation was used (data not shown). Wojcikiewics’ antibody, used on paraformaldehyde-fixed cells, gave positive staining in a cross-striated pattern but no nuclear envelope labeling (data not shown).

As myotubes begin to organize into sarcomeres (soon after fusion) in culture, IP₃Rs are simultaneously found in a diffuse and a cross-striated pattern in the cytoplasm (Fig. 2, top left). Myotubes show cross striations of both IP₃Rs and Ca²⁺ATPase (Fig. 2, top middle). In fact, the cross-striated regions appear to be colocalized, as seen by the yellow cross striations resulting from superimposition of the fluorescein image of IP₃R with the rhodamine image of Ca²⁺ATPase (Fig. 2, top right). However, in preparations fixed in methanol and stained with anti-IP₃R-1 (Affinity Bioreagents) the nuclear envelope region is positive for IP₃R-1 but negative for Ca²⁺ATPase (Fig. 2, top middle). We believe the SR may be continuous with the nuclear envelope (see below) and the nuclear envelope region may possess IP₃Rs but not Ca²⁺ATPase (or at least not the same type of Ca²⁺ATPase). IP₃R cross striations in differentiated myotubes also localize to the region positive for α-actinin (Fig. 2, bottom middle; Fig. 3A-C), that is, the Z-line (Fig. 5). Because of the apparent overlapping of IP₃R staining with labels for both Ca²⁺ATPase and α-actinin, we believe the IP₃Rs are most probably localized to the same region of the SR as Ca²⁺ATPase, the area of the ‘I-band SR’ near the Z-line. If so, we would expect the staining of IP₃Rs to alternate with myosin staining in the A-band. Such is the case as can be seen in Fig. 3D,E,F. In that illustration (Fig. 3F), orange (myosin) and yellow-green (IP₃R), alternate. The colors are not pure red and green as there is a slight background staining in both single laser images.

Finally, we compared localization of IP₃Rs with a specific triad marker, the α₁ subunit of the DHPR (Fig. 3G). In this fully striated myotube in culture, the proximity of the DHPRs to the IP₃R region of the SR appears as orange-yellow dots in confocal fluorescein/rhodamine composite images (Fig. 3G). In such mature myotubes, packed with cross-striated IP₃R-staining SR, several of the DHPR dots overlay (arrows in Fig. 3G) or are at the edge of (arrowheads in Fig. 3G) the IP₃R-SR staining, i.e. at the A-I junction (see also Fig. 5 for an explanation of composite images). At this resolution in the confocal microscope, both arrangements of IP₃R and α₁ staining suggest that α₁ is found in the region of the terminal cisternae where the more extensive staining of IP₃Rs is also found (Fig. 4, top; Fig. 5). In the mature myotube in culture we find areas of DHPR presence that lack IP₃R, and vice versa (Fig. 3G). We interpret this to mean that not all SR at this developmental stage shows IP₃R-SR staining, and there may be IP₃R not yet associated with α₁-bearing T-tubules.

In clearly stained, very mature myotubes, the IP₃Rs are localized in a double cross-striated pattern (Fig. 4). As illustrated here, this double pattern could be explained by higher concentrations of receptors in the regions of the terminal cisternae of the I-band SR or, alternatively, a larger expanse of SR surface in terminal cisternae. The image also suggests continuity in IP₃R-SR staining and the staining of IP₃Rs in the nuclear envelope region. In another myotube, in which the optical section is scanned at a better resolution (Fig. 4, bottom), the continuity is quite convincing. We have never seen such continuity in any other protein in our mouse myotubes (Fig. 2, top).

As in our study of rat myotubes (Jaimovich et al., 2000), confocal imaging allowed us to dissect one fast and one or more slow intracellular fluorescence rises when mouse myotubes were rapidly perifused with saline containing high K⁺. Fig. 6 represents Ca²⁺ imaging of fluo-3 fluorescence following high K⁺-induced depolarization of a mouse myotube. The depolarization was followed by the fast wave associated with E-C coupling and then the slow waves (see below). In this representative experiment, a sequence of images of fluo-3 fluorescence acquired every second is displayed. The sequence shows the fast increase (Fig. 6, 1-5 seconds) in fluorescence that immediately followed an increase in K⁺ concentration. Seconds after the fast signal faded, a focus of fluorescence appeared in the right extreme of the myotube (Fig. 6, 8-15 seconds). Fluorescence then propagated slowly to the left (Fig. 6, 9-16 seconds) and two components can be distinguished in this signal: a more rapid, more diffuse component of low fluorescence (green) and a group of localized, high fluorescence (red) spots, which always follow the rapid, diffuse component. The localized high fluorescence normally correlates with the positions of cell nuclei.

Rat myotubes display similar signals and relative fluorescence; the signal in a region of a cell can be clearly separated into two components (fast and slow) as a function of time (Fig. 7, top). The slow component showed a more complex kinetics as discussed above. When cells were pre-incubated and tested in the presence of either 50 μM of the IP₃R inhibitor 2-APB (Fig. 7, bottom, •, n=7) or 10 μM of the phospholipase C inhibitor U-73122 (□, n=8), the fast signal remained, sometimes slightly diminished, but the slow signal was completely abolished. The same effect was observed in mouse cells (data not shown, n=9). These experiments suggest that the slow Ca²⁺ signal is dependent on IP₃Rs. The fact that the fast Ca²⁺ transient was unaffected suggests that Ca²⁺ stores were not altered by these treatments. This was confirmed by experiments using 10 mM caffeine (data not shown), which elicited a large Ca²⁺ transient in 2-APB-treated cells.

To confirm that the signals described have long-term effects on the cells, we studied ERKs and CREB phosphorylation. A transient increase in both ERKs 1 and 2 phosphorylation was obtained with both depolarization protocols used (30-60 seconds or continuous exposure to high K⁺, see Materials and Methods). Fig. 8A illustrates results with protocol 1 (continuous exposure to high K⁺); a representative western blot is presented in the top panel. The center panel shows the mean, normalized effect for three to six independent experiments. For both protocols, an increase in ERKs 1 and 2 phosphorylation was seen in 22 out of 24 experiments, the net increase after 5 minutes being 3.3±1.2-fold (data not shown). The stimulation of phosphorylation, expressed as fold-induction over the controls, was similar for ERK 1 and ERK 2 under both conditions (Fig. 8A). In cells incubated in the presence of 50 μM 2-APB, the increase in ERK 1 phosphorylation induced by high K⁺ was clearly inhibited (Fig. 8A, bottom); mean inhibition was 63±7% (n=12). ERK 2 was also significantly inhibited, but to a lesser extent (32±12%).

An increase in CREB phosphorylation was already detected at 1 minute following depolarization (Fig. 8B), the shortest time examined. A maximum at 5 minutes with an increase over basal levels of threefold was evident (Fig. 8B, middle). The results shown in Fig. 8B correspond to experiments with myotubes continuously exposed to high K⁺ for the times indicated. The stimulation of phosphorylation was also found in experiments using protocol 2, where the myotubes were exposed to high K⁺ for only 30 seconds and incubated under resting conditions (data not shown). CREB phosphorylation was also inhibited in cells previously incubated with 50 μM 2-APB (see representative blot in Fig. 8B, bottom). Mean inhibition reached 55±4% of paired controls for four independent experiments. Nevertheless, basal values of P-CREB before depolarization were consistently higher in 2-APB-treated cells and, when compared with initial values, phosphorylation was completely blocked.

We have used several different cell types and found similar results for all of the rodent muscle cells. As cell lines of mouse origin are readily available for a number of studies (Estrada et al., 2001), it was important to analyze Ca²⁺ signals and the subcellular distribution of IP₃Rs in primary mouse cultures. Rat cultures were also used both for comparison and to understand the functional meaning of previously reported signals.

At the mRNA and protein levels, the presence of three IP₃R isoforms in different proportions in a number of rodent cell lines has been previously demonstrated (De Smedt et al., 1997; Newton et al., 1994; Wojcikiewicz, 1995). In particular, the study of mRNA expression performed by De Smedt et al. (De Smedt et al., 1997) has shown that in three mouse skeletal cell lines, including C2C12 cells, IP₃R-3 is clearly the most abundant transcript. The proportion of types 1 and 2 vary in these cell lines, but there was a tendency for type 2 to increase and for type 3 to decrease as differentiation proceeded. The results presented here build on our IP₃ nuclear binding results (Jaimovich et al., 2000; Liberona et al., 1998), i.e. we now know that at least two isoforms are present in the nuclear region. Nuclear localization of specific isoforms of IP₃R in skeletal muscle has not been described in the literature. Our own previous biochemical localization studies of IP₃R (IP₃ binding studies) indicated that the majority of the IP₃Rs are found in the nuclear fraction (Jaimovich et al., 2000; Liberona et al., 1998) in both cell lines and primary cultures. To confirm and expand these findings we have looked at the expression (mRNA and protein, Fig. 1) and the localization of the IP₃R. Only two of the isoforms, 1 and 3, could be studied by immunoblot procedures (Fig. 1C) due to the fact that type 1 and 3 were the only commercial antibodies available. In the immunofluorescence localization studies we could identify only type 1 as the commercial anti-type 3 did not bind the cultures.

As our double-labeling studies show IP₃Rs overlapping α-actinin staining in the region of the Z-line, colocalizing with Ca²⁺ATPase that is found in the I-band of the SR and, finally, alternating with myosin staining in the A-band region, we conclude that the IP₃Rs are located in the I-band region of the SR. Localization to the SR might be expected as IP₃Rs are found in the smooth endoplasmic reticulum of most mammalian cells. Perhaps the IP₃Rs and Ca²⁺ATPase are preferentially found in the I-band SR because there is more membrane here compared to the A-band SR.

Ca²⁺ATPase has been found in a perinuclear region in early chick myotubes in culture (Kaprielian and Fambrough, 1987), only to disappear from this region in more mature myofibers. By contrast, in these mouse cultures only the IP₃R staining of the SR membrane continues into the nuclear envelope region (Fig. 4).

In adult mammalian muscle, DHPRs of the triads should be found lined up near the edge of terminal cisternae of the SR at the A-I interface, very similar to what we find in the mature myotube in culture (Fig. 3G). The DHPR clusters, at the triads, are found on the margins of the IP₃R-staining I-band of the SR. We have not yet been able to obtain clean pictures of myotubes doubly labeled for IP₃Rs and the RyR; however, we have previously shown that the RyR and the α1 subunit of the DHPR colocalize precisely in this culture system at the light microscope level (Flucher et al., 1993), as would be expected from the known structure of the triad. This one-to-one relationship holds true even at the earliest expression of these receptors (Flucher and Franzini-Armstrong, 1996). Thus, we believe that the relationship of IP₃R localization to RyR will prove similar to that of IP₃R to α1.

Although we find no staining by anti-IP₃R in the nuclear envelope region in control experiments, using a paraformaldehyde fixation, we believe this may be due to penetration problems of the nuclear envelope. Our previous biochemical binding studies (Jaimovich et al., 2000; Liberona et al., 1998) show a significant IP₃ binding to muscle nuclei and we believe this is substantiated by the major immunocytochemical results.

Our previous papers (Estrada et al., 2000; Jaimovich et al., 2000; Jaimovich and Rojas, 1994) have shown the presence of two separate Ca²⁺ release systems in cultured muscle cells. In addition to the known fast Ca²⁺ transients associated with RyR, we have found a slow Ca²⁺ signal, associated with increases of IP₃ concentration and with the presence of IP₃Rs.

Several Ca²⁺-sensitive dyes show different binding affinities depending on intracellular compartment; fluo-3 has been shown to be the best indicator among those dyes that have been calibrated in situ (Thomas et al., 2000) but no quantitative measurements can be made. Increased fluorescence of this dye thus represents relative changes in intracellular [Ca²⁺]. The diffuse, low fluorescence, cytosolic component we see always associated with the localized nuclear Ca²⁺ rise is not associated with muscle cell contraction (Estrada et al., 2000; Jaimovich et al., 2000), so the concentration change must be too small to trigger contraction. The role for IP₃Rs in the I band region could be to modulate cytosolic Ca²⁺ concentrations within the levels, subcellular regions, and time scale required to activate nuclear Ca²⁺ release. An important finding was the inhibition of the slow Ca²⁺ signals in cells pre-incubated with 2-APB or U-73122 (Fig. 7). 2-APB has been shown to be an inhibitor of IP₃-mediated Ca²⁺ signals in various cell systems (Ascher-Landsberg et al., 1999; Gysembergh et al., 1999; Maruyama et al., 1997), although its specificity has been a matter for discussion (Gregory et al., 2001; Missiaen et al., 2001; Broad et al., 2001). The presence of fast Ca²⁺ signals together with lack of slow Ca²⁺ waves in the presence of 2-APB clearly suggests that these signals are indeed mediated by IP₃Rs. An effect of 2-APB on either Ca²⁺ pumps (Missiaen et al., 2001) or on store-operated Ca²⁺ channels (Gregory et al., 2001) is unlikely as internal Ca²⁺ stores appear unaltered as judged by both the presence of the fast Ca²⁺ signal and the fact that the response to caffeine was not inhibited. However, the signals we are looking at are independent of extracellular Ca²⁺ given that experiments performed in the absence of extracellular Ca²⁺ give the same results (data not shown) (Jaimovich et al., 2000; Estrada et al., 2001). Additional evidence of the role of IP₃ comes from the current experiments using U-73122; this PLC inhibitor has essentially the same effect as 2-APB. As the PLC inhibitor eliminates the slow wave, it is likely that depolarization causes IP₃ increase via the activation of PLC (Smith et al., 1990). The mechanism responsible for such activation should be a matter for further study. As IP₃R channels need both IP₃ and Ca²⁺ for activation, the function of IP₃Rs located in the I-band SR region could be to propagate a local Ca²⁺ wave that would not cause contraction but would assure enough Ca²⁺ available in the nuclear region to activate further local Ca²⁺ release for an, as yet, hypothetical regulatory role in gene expression, as proposed for neurons (Hardingham et al., 1999; Hardingham et al., 2001; Hu et al., 1999).

Signal transduction pathways: from depolarization to gene transcription

To explore whether depolarization-induced, IP₃R-mediated increase in nuclear [Ca²⁺] could affect gene expression in our system, we looked for two intermediates in a gene expression pathway: the phosphorylation of the kinases ERKs 1/2 and the transcription factor CREB. Our results show that depolarization of muscle results in an increase in the phosphorylation of ERKs 1 and 2, and that both the slow Ca²⁺ wave and the increase in P-ERK1/2 are blocked by 2-APB. These data indicate that IP₃R activity is necessary for both the slow Ca²⁺ wave and the phosphorylation of ERKs. The activation of ERKs 1/2 as a consequence of muscle stimulation has been widely reported (Goodyear et al., 1996; Sherwood et al., 1999; Ryder et al., 2000) but only recently has a specific role been demonstrated for these mitogen activated protein (MAP) kinases in skeletal muscle gene expression (Murgia et al., 2000). In regenerating muscle, the transfection of a constitutively active Ras or a Ras mutant that selectively activates ERKs could mimic the effects of slow motor neuron activity on expression of myosin genes. Thus, the Ras-ERKs signaling pathway has a role in depolarization-induced transcription in skeletal muscle. We believe that a direct connection between Ca²⁺ signals mediated by IP₃Rs and activation of the MAP kinases is supported by the facts that: (1) depolarization as brief as 30-60 seconds can elicit ERKs phosphorylation; (2) this response is not dependent on extracellular Ca²⁺; and (3) this response is inhibited in the presence of 2-APB.

We have demonstrated that K⁺-induced depolarization of rat myotubes in primary culture causes the phosphorylation of CREB. To our knowledge, this is the first evidence of this effect in skeletal muscle. Very recently, an increase in P-CREB levels was reported to occur in K⁺-depolarized arterial smooth muscle (Cartin et al., 2000). In neurons, it is the Ca²⁺ influx, concomitant with membrane depolarization or with NMDA receptor activation, that leads to CREB phosphorylation (Hardingham and Bading, 1999). The role of Ca²⁺, enhancing the activity of Ca²⁺-dependent kinases, which can phosphorylate CREB, has been demonstrated in a number of systems (Hardingham and Bading, 1999; Shaywitz and Greenberg, 1999).

In skeletal muscle cells, two major stimuli have been identified for growth and gene expression: electrical muscle activity (involving membrane depolarization) and growth factors. Muscle activity has long been known to stimulate (Brevet et al., 1976) or inhibit (Cohen and Fischbach, 1973) the production of extrajunctional muscle-specific proteins, yet the series of molecular events linking muscle usage to the cellular expression and accumulation of contractile proteins has not been spelt out. The IP₃ signaling pathway proposed here, activated by muscle activity (usually depolarization via action potential generation) could be one mechanism for controlling activation of muscle genes. The fact that phosphorylation of compounds involved in signaling cascades leading to gene expression occurs in response to relatively short depolarization times favors the idea of a direct role of action potential trains being the primary stimuli for this process.

Why has the role of the IP₃ signaling cascade been ignored in skeletal muscle when it is so universally recognized in so many other cells? We believe it is because it is difficult to conceive of a Ca²⁺ signal other than the intensely studied signal of E-C coupling. We hope that the data presented here, in conjunction with accumulating biochemical and physiological documentation of Ca²⁺ signals and IP₃ cascade intermediates and receptors, will stimulate further investigation of this topic, which is so central to the study of skeletal muscle development, health and disease.

The monoclonal antibody CaF2-5D2 developed by D.M. Fambrough was obtained from the Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA 52242. We thank G. Mignery and R. Wojcikiewicz for their generous gifs of their own polyclonal IP₃R-1 antibodies for control studies. We are grateful to Susan Haynes and Alejandra Espinoza for some tissue culture preparations and immunostaining, and to Sara Gathers for some tissue culture preparations, immunostaining and confocal scans. We appreciate the help of Marcelo Alvarez with some of the immunoblots. This work was supported by a grant from the Muscular Dystrophy Association of America (J.A.P. and E.J.), FONDECYT 8980010 (E.J and M.A.C.) and from the Blakeslee Fund and Howard Hughes Grant for undergraduate research at Smith College (J.A.P.).