Expression of the Ryanodine Receptor Type 3 Calcium Release Channel during Development and Differentiation of Mammalian Skeletal Muscle Cells*

In vertebrate skeletal muscles, the type 1 isoform of ryanodine receptor (RyR1) is essential in triggering contraction by releasing Ca2+ from the sarcoplasmic reticulum in response to plasma membrane depolarisation. Recently, the presence of another RyR isoform, RyR3, has been detected in mammalian skeletal muscle cells, raising the question of the eventual relevance of RyR3 for muscle cell physiology. The expression of RyR3 was investigated during differentiation of skeletal muscle cells. Using antibodies able to distinguish the different RyR isoforms and Western blot analysis, the RyR3 protein was detected in the microsomal fractions of differentiated skeletal muscle cells but not of undifferentiated cells. Accordingly, blocking muscle differentiation by the addition of either transforming growth factor-β or basic fibroblast growth factor prevented the expression of the RyR3 protein. In differentiated skeletal muscle cells, RyR3 was expressed independent of cell fusion and myotube formation. The expression of RyR3 was also investigated during development of the diaphragm muscle. The RyR3 content in the diaphragm muscle increased between the late stage of fetal development and the first postnatal days. However, at variance with RyR1, which reached maximum levels of expression 2–3 weeks after birth, the expression of RyR3 was found to be higher in the neonatal phase of the diaphragm muscle development (2–15 days after birth) than in the same muscle from adult mice. The differential content of RyR3 in adult skeletal muscles was found not to be mediated by neurotrophic factors or electrical activity. These findings indicate that RyR3 is preferentially expressed in differentiated skeletal muscle cells. In addition, during skeletal muscle development, its expression is regulated differently from that of RyR1.

Ryanodine receptors, together with the inositol trisphosphate (InsP 3 ) receptors, 1 form a superfamily of intracellular channels that mediate calcium release from intracellular calcium stores into the cytosol (1)(2)(3)(4)(5). Three isoforms of RyRs, encoded by three different genes, have been identified. RyR1 and RyR2 are predominantly expressed in skeletal and cardiac muscles (6 -9), respectively, and at lower levels in other tissues (10). A third isoform named RyR3 has been found to be present at low levels in almost all tissues (11)(12)(13), but so far no convincing evidence for a preferential association with a specific cell type or function has been obtained.
Most of our knowledge on RyR1 and RyR2 isoforms of the calcium release channel comes from studies in vertebrate striated muscles, where they play an important role for the release of calcium from the sarcoplasmic reticulum to activate contraction (14 -15). Transduction of the action potential signal from the cell surface to the interior of muscle cells, often referred to as excitation-contraction coupling, involves the dihydropyridine receptors (DHPRs) on the plasma membrane and RyRs on the sarcoplasmic reticulum (16 -17). Despite some similarities, the mechanism of excitation-contraction coupling differs in skeletal and cardiac striated muscles. In cardiac fibers, a calcium influx from the extracellular environment through the cardiac-specific DHPR activates the RyR2 isoform, probably via a calcium-induced calcium release mechanism (18 -20). In adult skeletal muscle, an influx of extracellular calcium through the skeletal DHPR is apparently not required, so that it has been proposed that the skeletal muscle DHPR detects the action potential and activates the RyR1 located on the terminal cisternae of the sarcoplasmic reticulum, probably by a direct physical coupling (21)(22). Accordingly, in vertebrate skeletal muscles, DHPRs and RyRs are colocalized in clusters soon after their synthesis and form the junctional domains observed in the space between transverse tubules (T tubules) and the terminal cisternae of the sarcoplasmic reticulum (23). Although junctional domains start to organize during the fetal stages of development, it is only after birth that this system undergoes a final reorganization (24). In mammals, within the period of time required after birth for the development and organization of a mature system of junctional domains, some quantitative and qualitative changes also occur in the biochemical composition of this system. In a few days after birth, the expression of DHPR genes is strongly increased, attaining maximal levels before the gene encoding the RyR1 isoform, which requires about 3-4 weeks to reach the high levels that are maintained throughout adult life (25). During the postnatal period of skeletal muscle development, the transient expression of the cardiac isoform of DHPR has also been observed (26 -27). The immature state of the excitation-contraction coupling apparatus at birth results in the apparent requirement of extracellular calcium in order for electrical stimulation to trigger contraction (28 -30). In mammalian skeletal muscles, the organization of a mature excitation-contraction coupling apparatus is established within 4 weeks of birth (24).
With respect to the biochemical composition of the excitation-contraction coupling apparatus, it has been shown that in most avian, amphibian, and fish skeletal muscles, two isoforms of RyRs named ␣ and ␤, which correspond to mammalian RyR1 and RyR3 (31)(32)(33)(34)(35), are expressed. Recent evidence has indicated that also in mammalian skeletal muscles, in addition to RyR1, the RyR3 isoform is present (10, 36 -37). In mammalian skeletal muscles, RyR3 is expressed at lower levels than RyR1; however, whereas RyR1 expression appears to be rather homogeneous, it is intriguing to note that the levels of RyR3 vary among different skeletal muscles, with maximal expression in the diaphragm and intermediate levels in the soleus. In other skeletal muscles, RyR3 levels are either extremely low or below detection limits (36). To further extend our knowledge of RyR3 in mammalian skeletal muscles, we analyzed whether the expression of RyR3 was associated with the differentiated state of cultured myocytes and studied its expression during the development of the diaphragm muscle. RyR3 content was also analyzed in denervated adult muscles to verify the possible influence of neurotrophic factors and electrical activity on the differential expression of the RyR3 isoform among different skeletal muscles.
Microsomal Proteins Preparation-Mouse diaphragm muscles were isolated from mice of 18 days of fetal development at days 2, 5, and 15 after birth and from adult mice (Ͼ2 months old). Bovine tissues were used to prepare the microsomal fractions of hippocampus, skeletal, and cardiac muscles used as controls. Microsomes were prepared as described previously (10,36). Briefly, tissue samples or cultured cells were homogenized in ice-cold buffer A (320 mM sucrose, 5 mM Na-Hepes, pH 7.4, and 0.1 mM phenylmethylsulfonyl fluoride) using a Teflon potter for cells or a Dounce homogenizer for tissues. Homogenates were centrifuged at 7000 ϫ g for 5 min at 4°C. The supernatant obtained was centrifuged at 100,000 ϫ g for 1 h at 4°C. The microsomes were resuspended in buffer A and stored at Ϫ80°C. Protein concentration of the microsomal fraction was quantified using the Bradford protein assay kit (Bio-Rad).
Western Blot Analysis-Microsomal proteins were separated by SDSpolyacrylamide gel electrophoresis as described (10). Proteins were then transferred to a nitrocellulose membrane (Schleicher & Schuell, Inc.) using a transfer buffer containing 192 mM glycine, 25 mM Tris, 0.01% SDS, and 10% methanol for 5 h at 350 mA at 4°C. Filters were blocked for 3 h in a buffer containing 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 0.2% Tween 20, 5% no-fat milk and incubated overnight at room temperature with specific antibodies. Polyclonal rabbit antisera were developed against purified glutathione S-transferase fusion proteins corresponding to the region of low homology situated between the transmembrane domains 4 and 5 (divergent region 1) of the RyR1, RyR2, and RyR3 proteins as described previously (10). These antibodies are able to distinguish the three RyRs and have been shown not to cross-react with each other (10,32,36). A monoclonal antibody against Ca 2ϩ -ATPase (YIF4) was kindly provided by Dr J. M. East. The antibodies against the RyRs were used at 1:3000 dilution, whereas the cell culture supernatant containing the YIF4 monoclonal antibody was used undiluted. Antigen detection was performed using the alkaline phosphatase detection method.
Calcium Measurements-Undifferentiated and differentiated C 2 C 12 cells were detached from plates with a trypsin/EDTA mix and washed with fresh cell culture medium by centrifugation at low speed for 5 min.
Cells were then resuspended in Krebs-Ringer-Hepes medium (125 mM NaCl, 5 mM KCl, 1.2 mM KH 2 PO 4 , 1.2 mM MgSO 4 , 2 mM CaCl 2 , 6 mM glucose, and 25 mM Hepes adjusted to pH 7.4 with NaOH) at a concentration of 5 ϫ 10 6 cells/ml. Fura 2-AM (Calbiochem) in anhydrous dimethyl sulfoxide was added to a final concentration of 5 M, and the cells were incubated for 30 min at room temperature in the dark. The cells were then diluted to a final concentration of 10 6 cells/ml. Before use, 750-l aliquots of cells were washed and resuspended in Krebs-Ringer-Hepes medium with 200 M sulfinpyrazone (Sigma) and transferred to a cuvette equipped with a magnetic stirrer for whole-cell population calcium measurement. The Fura 2 fluorescence was recorded on a luminescence spectrometer LS 50 (Perkin-Elmer) with excitation at 345 nm and emission at 490 nm. Maximum and minimum fluorescence (F max and F min ) were determined by the addition of 5 mM EGTA in Tris base, 40 mM Tris, 0.1% Triton X-100 (F min ), and 3 mM CaCl 2 (F max ).
[Calcium] i was calculated from the fluorescence F max and F min using a dissociation constant for the Fura 2⅐calcium complex of 225 nM and the equation

Functional Characterization of Calcium Stores in Developing
Myotubes-In skeletal muscle cells, calcium release through RyRs is associated with the expression of the differentiated phenotype and the regulation of muscle contraction, whereas the pathway that operates through the InsP 3 receptors appears to be independent of differentiation (42)(43). The organization of calcium stores and the expression of calcium release channels has been extensively studied using in vitro cultured cell lines and in developing and adult skeletal muscles (39, 42, 44 -46). To define the calcium stores present in proliferating and differentiated muscle cells, we performed fluorometric analysis of cytoplasmic calcium concentration using Fura 2 dye and the C 2 C 12 cell line as an in vitro model system of muscle differentiation and myotube development. Cells were stimulated with either bradykinin, a specific agonist of the InsP 3 receptor family of intracellular calcium channels and with caffeine, a specific agonist of RyRs. As shown in Fig. 1, stimulation with 100 nM bradykinin induced a transient increase of intracellular calcium concentration in both differentiated and undifferentiated C 2 C 12 cells (Fig. 1, lower panels). Instead, undifferentiated cells (Fig. 1, left upper  RyR3 Expression in Mammalian Skeletal Muscle Cells-A second isoform of calcium release channel, RyR3, has been recently detected in some mammalian skeletal muscles (36). To further extend this analysis, we measured the expression of the RyR3 isoform in microsomal fractions prepared from undifferentiated and fully differentiated C 2 C 12 cells. Western blot analysis of the microsomal proteins using antibodies able to distinguish the three RyR isoforms is shown in Fig. 2. The three proteins recognized by these antisera run in SDS gels with slightly different relative mobilities, with RyR1 showing the higher molecular weight and RyR3 showing the smaller relative mobility. The specificity of the antisera used is confirmed by isoform-specific signal displayed against skeletal muscle, cardiac muscle, and hippocampal microsomal preparations used as controls (see figure legends). Both RyR1-and RyR3specific bands are detected in differentiated culture (lanes 2 and 12), whereas they cannot be detected in undifferentiated growing cells (lanes 1 and 11). RyR2 protein, as expected, was not detected in either preparation (lanes 6 and 7). Similar results were obtained with microsomal fractions prepared from differentiated and undifferentiated murine BC3H1 cells and rat L6 cells (data not shown). Although at this level we cannot distinguish the calcium released through RyR3 from the calcium released through the RyR1 channels, these results indi-cate that the presence of functional calcium release stores operated by caffeine (i.e. ryanodine receptors) parallels the regulated expression of both RyR isoforms of calcium release channels.
The Expression of RyR3 in Mammalian Skeletal Muscle Cells Is Dependent on Differentiation-TGF-␤ and bFGF are known to interfere with myogenic differentiation without inducing a significant effect on cell proliferation (39,45,47). They can be used, therefore, to verify whether RyR3 expression is related to the activation of the skeletal muscle differentiation program or is a consequence of the withdrawal from the cell cycle due to the specific culture conditions used for the induction of differentiation. Fig. 3 shows the results obtained from Western blot analysis of C 2 C 12 cells induced to differentiate in serum-free medium and in serum-free medium supplemented with TGF-␤ and bFGF for 2 and 4 days. The data reported in Fig. 3 reveal a significant reduction of the RyR1 and Ca 2ϩ -ATPase levels both at 2 and 4 days of treatment when compared with normally differentiated cells. Analysis of the RyR3 content under these conditions revealed that RyR3, undetectable in undifferentiated growing cells, is clearly expressed at day 2 of culture in serum-free medium, indicating that this channel is expressed in the early stages of differentiation and further increases at day 4 of differentiation. RyR3 signal is significantly reduced in microsomal membranes of cells cultured in serumfree medium supplemented with either TGF-␤ or bFGF, suggesting that the arrest of growth is not sufficient to induce RyR3 synthesis but that the cells have to enter the differentiation program to express this protein.

Expression of RyR3 in Mammalian Skeletal Muscle Cells Is
Not Dependent on Cell Fusion-An important step in skeletal muscle differentiation is represented by myotube formation. Soon after plating C 2 C 12 in serum-free medium, myotubes and polynucleated syncitia due to cell fusion appear. The number of myotubes in culture rises dramatically between days 4 and 5. Extracellular calcium is known to be necessary for the cell fusion process leading to myotubes development. To analyze whether RyR3 expression in differentiated muscle cells was dependent on the process of cell fusion, C 2 C 12 cells were induced to differentiate with 1.4 mM EGTA to prevent cell fusion and myotube formation. This protocol is known not to affect the execution of the differentiation program at the biochemical level (i.e. expression of muscle-specific proteins is not prevented) (40 -41). As expected, the fraction of polynucleated cells at day 4 in EGTA-supplemented medium was about 5% that observed in control differentiating cultures. However, despite a remarkable effect of EGTA on the morphological differentiation, muscle-specific proteins (see Ca 2ϩ -ATPase and RyR1 content in Fig. 4) are present at levels comparable with normally differentiated control cells, indicating that indeed the cells underwent biochemical differentiation although they remained mononucleated. RyR3 was expressed in mononucleated differentiated C 2 C 12 at levels similar to those of polynucleated differentiated cells (Fig. 4), indicating that RyR3 expression is dependent on differentiation and is induced before and independently of cell fusion, similar to what was observed for RyR1 and other skeletal markers (39).
Expression of the RyR3 Isoform in Developing Skeletal Muscle-Although most muscle-specific proteins start to be expressed during embryo development, a significant burst in the synthesis of these proteins occurs around birth. With respect to the components of the excitation-contraction coupling apparatus, it has been shown that the patterns of expression of RyR1 and DHPR ␣ 1 subunit are distinct after birth, with the DHPR reaching maximal levels of expression earlier than RyR1 (25). In this context, it was interesting to investigate the in vivo developmental pattern of expression of the RyR3 protein. The diaphragm muscle was chosen in view of its relatively high content in RyR3 protein (36). Muscle tissue was isolated from mice at day 18 of fetal development (F18), at day 2 (P2), day 5 (P5), and day 15 (P15) after birth and from adult mice (i.e. 60-day-old mice). Both RyR1 and RyR3 were detected at the earliest developmental stage analyzed, as shown in Fig. 5. A relatively constant increase in RyR1 and Ca 2ϩ -ATPase can be observed in microsomes prepared from different stages, in agreement with published data (46). With respect to the RyR3 content, a significant increase was observed in the diaphragm muscle of mice from stages F18 (Fig. 5, lane 1) to P2 (not shown), P5 (Fig. 5, lane 2), and P15 (not shown). Surprisingly, a decrease in the RyR3 content was detected in diaphragm muscles prepared from adult mice (Fig. 5, lane 3). The same results were obtained in three independent experiments, ensuring that the observed differences in the levels of RyR3 in diaphragm muscles at the different days analyzed were significant. Thus, we conclude that RyR3 is expressed at higher levels in neonatal than in adult diaphragm muscle.

Expression of RyR1 and RyR3 in Adult Skeletal Muscles Is Not Dependent on Electrical Activity/Neurotrophic Factors-
Expression of many skeletal muscle genes and proteins has been shown to be dependent on various factors including nerveinduced electrical activity and neurotrophic factors (48 -49). In light of the observed differences in the RyR3 isoform content among mammalian skeletal muscles, we analyzed last whether the expression of RyR3 was regulated by innervation. This was analyzed in rat soleus and extensor digitorum longus muscles after removal of a section of the sciatic nerve and in the right hemidiaphragm muscle after unilateral removal of a section of the phrenic nerve, using an intercostal approach. Microsomes were isolated from the indicated muscles 10 days after nerve recision and from the corresponding contralateral control muscles. Proteins were size-fractionated on SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose membranes, and decorated with antibodies able to discriminate the RyR1, RyR2, and RyR3 isoforms. Results reported in Fig. 6 indicate that the levels of RyR1 and RyR3 in denervated muscles (lanes 2, 4, and 6, respectively) were essentially unchanged with respect to controls (lanes 1, 3, and 5, respectively), thus indicating that expression of RyR3 in soleus and diaphragm as well as its absence in the extensor digitorum longus muscle is independent of muscle innervation. No expression of RyR2 in denervated muscles was detected.

DISCUSSION
The process of skeletal muscle development is a complex mechanism that implies a fine regulation of the expression of contractile and sarcoplasmic proteins. This has been extensively studied using both primary myoblasts and cell lines able to differentiate and form myotubes in vitro (24). Studies on the structural and functional organization of intracellular calcium stores indicated that the activity of these organelles depends on a complex level of organization, including the regulated expression of the genes encoding proteins that transport calcium ions in the luminal side of these vesicles, calcium binding proteins, and calcium release channels (1). Previous studies on the composition and function of intracellular calcium stores during in vitro differentiation of skeletal muscle cells revealed that, whereas InsP 3 -sensitive stores can be present in undifferentiated and in differentiated cells, caffeine-sensitive stores can be detected only in differentiated cells (42)(43). These observations suggest that the InsP 3 receptor plays a role in regulating intracellular calcium homeostasis with respect to housekeeping activities, whereas expression and function of the skeletal muscle isoform, RyR1, have been shown to be mainly related to the functions of the differentiated muscle cells.
In this report we have analyzed the expression of the RyR3 isoform using cultured myogenic cell lines and during the development of skeletal diaphragm muscle. Western blot analysis of microsomal fractions prepared both from undifferentiated and differentiated cells indicated that two isoforms of calcium  2, 4, and 6, respectively) and from the corresponding contralateral control muscles (lanes 1, 3, and 5, respectively). 10 g of proteins were tested for RyR1, whereas 50 g were tested for RyR2 and RyR3. release channel, RyR1 and RyR3, were present in differentiated muscle cells but not in growing, undifferentiated cells. This was confirmed in experiments where RyR3 expression was completely abolished when differentiating cells were cultured in the presence of growth factors such as bFGF and TGF-␤, which inhibit myogenic differentiation. This indicates that as reported for RyR1 and other muscle proteins (39,45,47), RyR3 synthesis is not dependent on cell cycle withdrawal but is rather related to the activation of the myogenic differentiation program. In addition, RyR3 expression was found to occur before and independent from the appearance of morphologically differentiated myotubes, as stated by experiments where cell fusion and myotube formation was prevented by removal of extracellular calcium with EGTA. In conclusion, all of the experiments performed with differentiating myoblasts indicate that expression of the RyR3 isoform in muscle cells shows a trend similar to that of most of the known muscle-specific proteins.
In nonmammalian skeletal muscles, the expression of equivalent levels of two isoforms of calcium release channels, namely ␣ and ␤, is a predominant feature of most species. The two channels display distinct calcium release properties (50 -52), suggesting that RyR1 and RyR3 isoforms may properly fit with the functional model based on morphological studies of triads at the electron microscope, where one out of every two RyRs is not opposed to DHPRs (53)(54). How the uncoupled RyRs are activated is yet unclear, although their activation could be mediated by the isoforms coupled to the DHPRs (14). Up to now, there is no direct evidence for a role of the RyR3 isoform in mammalian skeletal muscle physiology and excitation-contraction coupling. Lack of response to electrical stimulation has been reported in the skeletal muscle of mice and chickens that carry mutations which abolish the expression of the RyR1/␣-RyR isoform, indicating that RyR3 cannot substitute for RyR1 function in this process. Caffeine stimulation does, however, cause calcium release in the RyR1-deficient mouse myocytes in support of a potential role of RyR3 in skeletal muscle intracellular calcium regulation, probably through a calcium-induced calcium release mechanism (37,(55)(56). The recent studies of knockout mice that do not express a functional RyR3 protein indicate that these mice can develop normal skeletal muscle tissue in the absence of the RyR3 isoform (57). 2 However, the presence of subtle qualitative modifications in the excitationcontraction coupling of muscles from these mice cannot yet be completely ruled out and requires further studies. 2 Expression of most muscle-specific proteins is strongly increased after birth probably as a consequence of the activation of muscle activity and reaches maximal levels a few weeks after birth (24 -25). However, the kinetics of expression of skeletal muscle genes around birth are not synchronous. Analysis of RyR3 expression during mouse skeletal muscle development revealed that RyR3 was already expressed during fetal development (F18) similar to the RyR1 isoform and that expression of both isoforms increased at birth. However, after birth, a distinct pattern of expression was observed for the two receptors. Whereas the RyR1 content continued to increase in the period from P2-P15 to adult life, the RyR3 content was found to be higher in the days following birth (P2-P15) than in the diaphragm muscle of adult mice, indicating that in this muscle RyR3 is preferentially expressed during the neonatal phase of development. This finding indicates that the relative ratio between RyR1 and RyR3 isoforms of calcium release channels in mammalian skeletal muscles differs not only among different muscles but also in the same muscle during postnatal development.
It is interesting to note that for a period of a few weeks after birth the excitation-contraction coupling apparatus of the skeletal muscle appears to be functionally immature compared with that of adult skeletal muscle. In fact, during this period skeletal muscle contraction appears to depend on external calcium (28 -30). Another biochemical marker of the differences in the composition of the excitation-contraction coupling apparatus during the neonatal period of muscle development is represented by the temporary expression of the cardiac DHPR channel, which is otherwise never detected in skeletal muscles with the exception of regenerating muscles (26 -27, 58 -59). Therefore, although RyR3 is apparently not essential for the development of skeletal muscles, RyR3 is expressed in differentiated skeletal muscle cells, and its expression is developmentally regulated. The coexpression of the RyR1 and RyR3 calcium release channels in skeletal muscles may result in a diversification of the excitation-contraction coupling apparatus to obtain a fine regulation of muscle contraction. In nonmammalian vertebrates, differential usage of either one or two isoforms of RyRs seems to reflect the contractile properties of specific muscles (60). Thus it is not unreasonable that the different content in RyR3 observed in mammalian skeletal muscles or in at least a specific subset of them may fulfill a similar task (36). The preferential expression of RyR3 in neonatal diaphragm muscle suggests a possible role of the RyR3 isoform in excitation-contraction coupling of immature diaphragm muscle and may help to unravel the role of RyR3 in this muscle. Further work is clearly required to address this specific aspect.