Restricted Distribution of mRNAs Encoding a Sarcoplasmic Reticulum or Transverse Tubule Protein in Skeletal Myofibers

Abstract

Calsequestrin (CSQ) and dihydropyridine receptor (DHPR) are muscle cell proteins that are directed into the endoplasmic reticulum (ER) during translation. The former is subsequently found in the sarcoplasmic reticulum (SR) and the latter in the transverse tubule membrane. To elucidate the potential role of mRNA targeting within muscle cells, we have analyzed the localization of CSQ and DHPR proteins and mRNAs in primary cultured rat myotubes, in skeletal muscle cryosections, and in isolated flexor digitorum brevis muscle fibers. In the myotube stage of differentiation, the mRNAs distributed throughout the cell, mimicking the distribution of the endogenous ER marker proteins. In the adult skeletal myofibers, however, both CSQ and DHPRα1 transcripts located perinuclearly and in cross-striations flanking Z lines beneath the sarcolemma, a distribution pattern that sharply contrasted the interfibrillar distribution of typical ER proteins. Interestingly, all nuclei of the myofibers were transcriptionally active. In summary, the mRNAs encoding either a resident SR protein or a transverse tubule protein were located beneath the sarcolemma, implying that translocation of the respective proteins to the lumen of ER takes place at this location.

Keywords

in situ hybridization skeletal muscle sarcoplasmic reticulum

Myogenic development involves cell-cell fusion and a differentiation process via multinucleated myotubes to innervated myofibers. The myofibers are filled with the contractile apparatus comprising myofibrils that are enwrapped by the sarcoplasmic reticulum (SR) and transverse tubules that are both organized in a cross-striated fashion. The SR membrane system contains typical endoplasmic reticulum (ER)-marker proteins that occupy both core and cortex regions of the fibers (Volpe et al. 1992; Kaisto and Metsikkö 2003). It has, however, remained obscure whether the entire ER/SR system is capable of cotranslational translocation. A key question in this regard is whether the core regions contain rough ER and, accordingly, mRNAs that encode proteins that are translocated to the ER. Transcript targeting to specific ER subdomains was observed when immature myotubes were investigated (Ralston et al. 1997). In the adult myofibers, mRNAs encoding membrane proteins of the neuromuscular junction were clearly restricted beneath the junctional region (Merlie and Sanes 1985; Fontaine and Changeux 1989; Jasmin et al. 1993). Only a few transcript species encoding extrajunctional ER-translocated proteins have been localized. Accordingly, mRNAs encoding dystroglycan (Mitsui et al. 1997) or sodium channel (Awad et al. 2001) were concentrated beneath the sarcolemma, typically following the localization of their protein products.

In this study, we have analyzed the localization of mRNA species encoding proteins that are translocated to the ER and then sorted into organelles located throughout the fibers. For this purpose, we selected transcripts that encode calsequestrin (CSQ), a peripheral protein of the SR lumen, and dihydropyridine receptor (DHPR) α1 subunit that is transported from the ER to the transverse tubule membrane. CSQ is a high-capacity calcium-binding protein that provides a reserve for calcium within the SR. DHPR is a dihydropyridine-sensitive l-type calcium channel that functions as a voltage sensor of the transverse tubule triad region (MacLennan 2000). In situ hybridization studies showed that in the myotube stage, there was no compartmentalization of the transcripts, and their distribution mimicked that of the conventional ER markers. In the adult myofibers, the distribution surprisingly did not match the distribution of the ER, the SR, or the transverse tubule markers, but the mRNAs were restricted beneath the sarcolemma. Interestingly, all the nuclei were active. Because the inner regions of the fibers seemed to be lacking transcripts, it is reasonable to conclude that the respective proteins were synthesized beneath the sarcolemma.

Materials and Methods

Cell and Tissue Sample Preparation

Myofibers were isolated by collagenase digestion from rat flexor digitorum brevis muscle as previously described by Rahkila et al. (1996). The fibers were fixed with 3% paraformaldehyde (PFA) in 10 mM phosphate-buffered saline (PBS) and applied on 40-μm cell strainers (Becton Dickinson Labware; Franklin Lakes, NJ), and in situ protocol was performed on the freely floating fibers. Alternatively, the isolated fibers were plated on Matrigel-coated (Becton-Dickinson) dishes and cultivated for at least 18 hr before fixation with 3% PFA for the in situ protocol. Primary myotubes were obtained from satellite cells. Accordingly, myofiber cultures were maintained for 1 week, during which the satellite cells proliferated and differentiated into myotubes as described (Rahkila et al. 1998). The cultures were fixed for 15 min with 3% PFA and processed for the in situ procedure. For whole-muscle cryosections, muscle pieces were excised and frozen by immersion in 2-methylbutane cooled by liquid nitrogen. The tissue samples were cryosectioned and placed on Super Frost Plus Plus slides (Fisher; Pittsburgh, PA) and immediately fixed with 3% PFA. Alternatively, rat muscle pieces were fixed, dehydrated, and mounted in paraffin and sectioned. The sections were then rehydrated and subjected to in situ hybridization.

Probes

Approximately 0.2 kb-cDNA fragments of the CSQ and di-hydropyridine-sensitive l-type calcium channel (DHPR) α1 were cloned by RT-PCR using total RNA from rat skeletal muscle. Primers CSQ-forward (GAA GGA TCC TCA AAG ACC CAC CCT AC) and CSQ-reverse (AGT CGT CTG GGT GAA TTC ACA AGA TG) were chosen for CSQ2 cDNA amplification from gene bank sequence U33287 (bases 855-1064) and EcoRI and BamHI cloning sites were added to the primers. Similarly, primers DHPRα1-forward (GCA ATG AAT TCT CCA AGA TGA CTG) and DHPR-reverse (TCC TAG GCC TTG TAC AGT AGC TGT) were designed from the gene bank sequence L04684 to clone DHPRα1 cDNA fragment (bases 730-928). The restriction sites for EcoRI and NcoI were added to the DHPRα1 primer sequences. Both PCR products were cloned into pGEM-T vector (Promega Corp.; Madison, WI) according to standard procedure. The plasmid DNAs were isolated and sequenced to certify correct insert sequences and orientations. The digoxigenin (DIG)-labeled RNA antisense and sense probes were synthesized from the linearized CSQ and DHPRα1 cDNA clones according to the instructions of the DIG-RNA labeling kit (Roche Diagnostics GmbH; Penzberg, Germany) using SP6- or T7-polymerase transcription. The concentrations of the DIG-labeled probes were estimated by the spot test according to the protocol from Roche Diagnostics.

RNA In Situ Hybridization

In situ hybridizations were performed based on protocol described by Ylä-Outinen et al. (2002) with slight modifications. Briefly, PFA-fixed samples were washed in 0.1 M phosphate buffer, permeabilized by treatment with 0.5% Triton X-100 in PBS, treated with proteinase K (10 μg/ml), washed with cold water, incubated in 0.2 M HCl containing 0.15 M NaCl, and washed in saline sodium citrate buffer (SSC). The myotube samples were not treated with proteinase K. The samples were acetylated with 0.25% acetic anhydride and incubated overnight at 45C in the hybridization solution containing DIG-labeled RNA probe (300 ng/ml), 50% formamide, 0.6 M NaCl, 2 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1.0 mg/ml BSA, 0.02% polyvinylpyrrolidine, 0.02% Ficoll, 10 mM dithiothreitol, 0.2 mg/ml ssDNA, and 10% dextran sulfate. Posthybridization treatment included stringent SSC washes at 50C and a wash with RNase A for 30 min at 37C.

The DIG-labeled RNA-RNA hybrids were detected by immunofluorescence. Briefly, samples were blocked by incubation with 6% BSA-PBS solution and then incubated with mouse anti-DIG antibody (1:100 dilution in 6% BSA-PBS; Roche). After washes in PBS, the AlexaFluor 488-conjugated IgG (Molecular Probes; Leiden, the Netherlands) was used for detection. The samples were counterstained with the nuclear Hoechst Dye 33,258 according to the standard protocol. The mounted samples were examined with a Zeiss LSM 510 (Carl Zeiss Inc.; Jena, Germany) confocal laser scanning microscope. Confocal section thickness was 0.5 μm. Intensity profiles across the scanned images were displayed by the LSM 510 examiner program. In situ hybridizations with the sense probes were used as negative controls in all experiments.

RNA Isolations and Northern Analysis

Total RNAs from adult rat muscle tissues were isolated with Trizol reagent (Gibco Life Technologies; Karlsruhe, Germany) according to the manufacturer's instructions. RNA samples (10 μg) were loaded on the denaturating agarose gel, electrophoresed, and transferred to a nylon membrane (Roche Diagnostics) by capillary transfer and fixed with UV-crosslinker (Stratagene GmbH; Heidelberg, Germany). DHPRα1 antisense probe hybridization and washes were performed at 68C. CSQ antisense probe was hybridized at non-stringent temperatures (48C) and washed at 52C to detect the transcripts of the CSQ1 and CSQ2 isoforms. The CSQ and DHPRα1 DIG-labeled antisense riboprobes (100 ng/ml) were used for hybridization with DIG EASY hyb hybridization solution (Roche Diagnostics). After washes, alkaline phosphatase-conjugated sheep anti-DIG antibody and CSPD chemiluminescent reagent (Roche Diagnostics) were used for detection of the hybrids according to the manufacturer's instructions.

Immunolabeling

Monoclonal antibodies against CSQ, DHPRα1 subunit, and protein disulfide isomerase (PDI) were from Affinity Bio-Reagents, Inc. (Golden, CO). Anti-tubulin antibody (clone DM1A) and polyclonal anti-desmin antibodies were obtained from Sigma (St Louis, MO). Routine immunohistochemical protocols were used. Briefly, PFA-fixed samples were permeabilized with 0.5% Triton X-100, and the nonspecific binding was blocked with 1% BSA-PBS solution. The primary antibodies were applied on the samples for 1 hr at 37C. AlexaFluor-488 or −568-conjugated secondary antibodies were used for detection. The samples were mounted and examined with a Zeiss LSM510 confocal laser-scanning microscope.

Drug Treatment, Metabolic Labeling, and SDS-PAGE

Myofibers were cultured overnight before incubation for 1 or 2 hr in the presence of nocodazole (10 μM), cytochalasin D (20 μM), latrunculin A (5 μM), puromycin (200-800 μg/ ml), or cycloheximide (40 μg/ml). After treatment with the drugs, the myofibers were subjected to in situ hybridization. Metabolic labeling for the puromycin- or cycloheximide-treated samples was performed by incubation for 1 hr in the presence of [³⁵S]methionine (5 μCi/ml). SDS-PAGE was performed on 10% polyacrylamide gels as described by Laemmli (1970) using the Bio-Rad Mini Protean 3 cell (Bio-Rad Laboratories; Hercules, CA). The electrophoresis was performed at 50 mA current for 4 hr. The gel was dried, and radioactivity incorporated into proteins was detected with Fuji Bioimager (BAS 1800 II, Fuji Fujifilm; Sendai, Japan). Images were processed with Adobe Photoshop6.

Results

Northern Analysis

Northern analysis for CSQ and DHPRα1 mRNAs was performed to verify the specificity of the antisense probes. The DHPRα1 probe showed a band of ∼6 kb in leg muscle tissue, corresponding to the size of the transcript reported by Ray et al. (1995). Analysis of canine and rabbit CSQ2 isoform has shown that it is present in both slow- and fast-twitch fibers, whereas CSQ1 isoform is present in fast-twitch fibers only (Scott et al. 1988; Fliegel et al. 1989). The CSQ antisense probe was homologous to the CSQ2 isoform, but due to the extensive sequence homology between the isoforms, it likely recognizes the CSQ1 mRNA present in rat muscle (Volpe et al. 1994). Therefore, we analyzed the rat slow-twitch soleus and the fast-twitch extensor digitorum longus (EDL) muscle separately under the low-stringency conditions. Total RNA of the soleus muscle exhibited a 2.6-kb band corresponding to the CSQ2 mRNA, and the EDL muscle exhibited the 1.9-kb transcript corresponding to the CSQ1 mRNA (Park et al. 1998). This means that the antisense CSQ probe recognized both isoforms. The results are shown in Figure 1.

Figure 1

Northern analysis of the antisense probes. The DHPRα1 antisense probe hybridized to ~6.0-kb transcript in total leg muscle RNA (lane 1). The CSQ probe recognized a 2.6-kb transcript in the soleus (lane 2), and a 1.9-kb transcript in the EDL muscle (lane 3). Equal amount of soleus and EDL RNA were loaded onto the gel, as verified by ethidium bromide staining of rRNAs.

In Primary Cultured Myotubes, the mRNAs Encoding CSQ or DHPRα1 Distribute throughout the Cell

Myotubes are multinucleated muscle cells that are not terminally differentiated. We first analyzed whether the translocation to the ER was compartmentalized in myotubes. For this purpose, we localized the mRNA encoding CSQ that is a resident protein of the SR lumen and has been shown to be cotranslationally translocated (Reithmeier et al. 1980). We also localized the mRNA encoding the DHPRα1 subunit that is a multi-spanning polypeptide purported to travel through the Golgi elements and to be targeted to the transverse tubule of the triads (Neuhuber et al. 1998). Primary cultured myotubes were investigated because immunofluorescence staining with specific antibodies showed that these cells contain both the CSQ and the DHPR proteins. Analysis of the transcript distribution with antisense riboprobes indicated a reticular structure throughout the cytoplasm, but no compartmentalization was seen (Figure 2). The staining pattern mimicked that of respective proteins or endogenous ER proteins such as PDI (Figure 2, bottom panel). The staining was specific, inasmuch as sense probes showed very weak or no staining. These findings are compatible with the idea that in the myotubes, the mRNA species of interest show no compartmentalization to ER domains.

In Adult Myofibers, the CSQ and DHPRα1 mRNAs Are Restricted beneath the Sarcolemma

To determine the distribution of the mRNAs encoding CSQ or DHPRα1 subunit in mature skeletal muscle fibers, we performed in situ hybridization studies on rat muscle cryosections. Because our CSQ probe was against the mRNA encoding the slow-twitch CSQ2 isoform, we first analyzed soleus muscle that contains mainly slow-twitch fibers expressing the CSQ isoform in question (Fliegel et al. 1989). In transversely cut sections, the in situ hybridization revealed an intense subsarcolemmal staining (Figure 3) that was also detected in longitudinally cut sections. In the inner regions of the fibers, the intensity was at the background level seen in the sections incubated with the sense probe. All the nuclei in the sections showed relatively strong perinuclear rings. The sections of the EDL muscle that contain only the fast-twitch CSQ1 isoform also exhibited intense subsarcolemmal and perinuclear staining that was not seen with the sense probe. Similar results were obtained when other rat leg muscles were analyzed.

Figure 2

Transcripts encoding CSQ or DHPRα1 are evenly distributed within primary myotubes. RNA-RNA in situ hybridization was performed using digoxigenin-labeled probes that were visualized by immunofluorescence staining. Antisense or sense CSQ and DHPRα1 probes were used as indicated. The bottom panel shows immunofluorescence staining for CSQ, DHPRα1, and for PDI that is an endogenous ER-marker protein. Confocal sections are shown. Bars = 10 μm.

We next subjected sections of various muscles, including soleus and EDL, to in situ hybridization using the DHPRα1 probe. All muscle types analyzed showed a similar subsarcolemmal and perinuclear staining pattern that could not be detected when the sense probe was used. An example of the soleus muscle is shown in Figure 3. When we compared the staining intensities of the inner regions of the fibers in the sections, we could not find any marked difference between antisense and sense probes (Figure 3, intensity profiles). In situ hybridization on paraffin sections also showed subsarcolemmal staining with the CSQ and DHPRα1 probes (not shown). It is notable that the distribution patterns of these transcripts differed essentially from the patterns displayed by the corresponding proteins. Importantly, the transcript distribution also differed from the ER distribution detected by classical ER protein markers. Figure 3 (bottom panel) shows examples of the distribution of CSQ, DHPRα1, and PDI proteins, indicating that the proteins distributed equally in the inner as well as the subsarcolemmal regions.

Figure 3

CSQ and DHPRα1 transcripts locate to subsarcolemmal regions of myofibers, whereas the respective proteins or an ER marker protein show an even distribution throughout the fiber cross-section. In situ hybridization using antisense or sense CSQ probes was performed on EDL or soleus muscle sections and DHPRα1 antisense or sense probes on soleus sections, as indicated. Intensity profile analyses are shown below each panel, indicating that in the inner regions of the fibers, the antisense and sense probe intensity levels were practically the same. The path of the intensity profile line is indicated in each panel. The bottom panel shows immunofluorescence staining of CSQ and DHPRα1 proteins, and PDI that is an endogenous ER protein. Bars = 10 μm.

Figure 4

CSQ and DHPRα1 transcripts in myofibers show subsarcolemmal cross-striations that are in register with the I bands. Myofibers were isolated, fixed, and processed for the in situ hybridization labeling. Antisense or sense CSQ and DHPRα1 probes were used as indicated. The right panels show double labeling using antisense probe together with desmin protein immunofluorescense staining. Green color indicates the CSQ or DHPRα1 transcripts, red color indicates desmin, and yellow indicates colocalization. The top panel shows confocal sections from the middle of the fiber, while middle and bottom panels show confocal sections beneath the sarcolemma. Bars = 10 μm.

To obtain an overall view of the distribution of the CSQ and DHPRα1 mRNA in the myofibers, we performed in situ hybridization on isolated myofibers. The procedure was done either immediately after the isolation process or after cultivation on Matrigel substratum. Confocal sectioning indicated a strictly subsarcolemmal distribution in both cases and with both probes. Remarkably, the labeling pattern comprised cross-striations that often appeared as double strands (Figure 4). A prominent perinuclear staining was present around every nucleus. The cross-striated staining distributed uniformly over the entire myofiber length with the exception that the intensity often decreased with the distance from the nuclei (not shown). High-stringency hybridization conditions essentially weakened the labeling in all fibers, but the staining patterns remained unaltered. Double staining with antibodies against desmin indicated that the cross-striated labeling was over the I-band areas (Figure 4, right panels).

Figure 5

Nocodazole does not affect transcript distribution in isolated myofibers. Isolated myofibers cultivated on Matrigel were incubated without (control) or with nocodazole as indicated. The fibers were then processed for the in situ hybridization protocol using CSQ or DHPRα1 antisense probes, or they were subjected to immunofluorescence staining with anti-tubulin antibodies (bottom panel). Microtubules were practically disassembled with the nocodazole concentration used. Subsarcolemmal confocal sections are shown. Bars = 10 μm.

mRNA-cytoskeleton interactions are common in mononucleated cells. Therefore, we analyzed whether disrupting microtubules changed the mRNA distribution pattern in the myofibers. However, nocodazole did not change the subsarcolemmal localization or the cross-striated distribution pattern of the transcripts at concentrations that destroyed microtubules (Figure 5). Although the presence of subsarcolemmal gamma actin filaments in myofibers has been recently questioned (Nakata et al. 2001), we analyzed whether cytochalasin D or latrunculin that disassembles actin filaments altered the transcripts distribution. These drugs did not have any marked effect (not shown).

Ribosomes that synthesize membrane-associated or secreted proteins are targeted to the ER via the signal sequence of the nascent polypeptide chain. Puromycin causes release of the nascent polypeptide from the ribosome (Taneja et al. 1992) and dissociates the mRNA from the ER (Lodish and Froshauer 1977). To analyze whether the subsarcolemmal localization of the transcripts was a result of association with the ER, we incubated isolated myofibers with different concentrations of puromycin (200-800 μM) for various periods. Metabolic labeling with [³⁵S]methionine indicated that protein synthesis was blocked with the concentrations used. Puromycin blurred the cross-striated appearance to some extent at high concentration (800 μM; Figure 6), but it did not change the subsarcolemmal localization. Cycloheximide is another drug that halts protein synthesis without releasing the mRNA from the ER, but it had no effect on the localization patterns of the transcripts. The localization pattern of DHPRα1 after cycloheximide treatment is shown in Figure 6. These results indicate that association with the microtubules, actin filaments, or the ER is not crucial for the subsarcolemmal localization of the mRNAs encoding CSQ or DHPRα1.

Figure 6

Effect of puromycin or cycloheximide on the distribution of DHPRα1 transcripts. Myofibers were treated either with puromycin (800 μM) or cycloheximide (10 μM) and subjected either to in situ hybridization or to metabolic labeling with [³⁵S]methionine. Untreated myofibers were used as control. The in situ hybridization panels show subsarcolemmal confocal sections of myofibers treated as indicated, and the Bioimager image shows SDS-PAGE profiles of the [³⁵S]methionine-labeled myofibers, in the absence or presence of drugs as indicated. Bars = 10 μm.

All the Nuclei Are Active in the Myofibers

Actively transcribed mRNAs appear as two foci in a diploid nucleus, whereas transcripts of less active or inactive genes distribute diffusively throughout the nuclei (Jolly et al. 1997). The nuclear spots were used here to analyze whether there were differences in the activity between nuclei in the muscle cells. We found that in the muscle cryosections, bright intranuclear spots were constantly seen with the antisense probe but not with the sense probe. The use of isolated myofibers made it possible to analyze all the nuclei in a fiber. We found that the myofiber nuclei displayed two spots when probed for either CSQ or DHPRα1 transcripts (Figure 7). In each fiber, every nucleus displayed two spots, and their intensity within a single fiber was constant. Accordingly, the genes encoding CSQ or DHPRα1 seemed to be equally active in all the nuclei and both alleles were transcribed. In the myotube cultures, no spots in the nuclei were seen with either DHPRα1 or CSQ probes, suggesting a relatively low expression level.

Discussion

In adult myofibers, the transcripts encoding rough ER-synthesized proteins CSQ and DHPRα1 were found to be restricted beneath the sarcolemma. In contrast, several endogenous ER markers have been shown to distribute throughout the myofibers. For example, chaperoning components of the ER distributed all through the muscle cells (Volpe et al. 1992; Rahkila et al. 1997). Furthermore, the viral GFP-tagged mutant tsO45 glycoprotein that is blocked in the ER at 39C due to a folding defect also distributed in the cortex and core regions in cultured myofibers (Kaisto and Metsikkö 2003). The distribution of the transcripts analyzed here differs radically from that of the classical ER markers or the viral glycoprotein. It remains possible, though, that the inner regions of the myofibers contain small amounts of transcripts that cannot be detected.

Figure 7

All the nuclei are transcriptionally active in myofibers. In situ hybridizations using the DHPRα1 or CSQ antisense probes were performed. In a fraction of the nuclei, only one spot was visible, but confocal sectioning indicated two spots in such nuclei. Bars = 10 μm.

Earlier studies have located mRNAs encoding sarcolemmal proteins such as dystroglycan (Mitsui et al. 1997) or sodium channel (Awad et al. 2001) preferentially beneath the sarcolemma. An explanation for this could be that the large muscle cells contain ER sub-compartments, and transcripts encoding exported proteins are sorted to ER domains beneath the sarcolemma. CSQ, on the contrary, is not exported from the ER to the sarcolemma but is targeted to the SR that occupies the inner regions of the fibers. Therefore, it is surprising that its transcripts were restricted beneath the sarcolemma as well. The transcripts of DHPRα1 were also restricted beneath the sarcolemma, although the protein product distributed throughout the fiber cross-section. These findings, together with previous in situ hybridization studies, suggest that in skeletal muscle fibers, transcripts encoding proteins to be translocated to the ER are restricted beneath the sarcolemma. If so, the rough ER would be restricted to subsarcolemmal regions as well. This is compatible with the notion that the SR membranes occupying the interfibrillar spaces are smooth-surfaced (Saito et al. 1984), implying that they do not contain ribosomes and do not participate in protein synthesis. One explanation for why ER-specific proteins that are not directly linked to calcium storage are present in the entire SR is that they are needed for chaperoning processes in the muscle cells that are regularly subjected to stress conditions. It is notable that the smooth ER of hepatocytes also contains PDI (Akagi et al. 1988) and the folding-defective tsO45 viral glycoprotein or BiP has free access to the smooth-ER compartment in UT-1 cells (Bergmann and Fusco 1990).

We found that in the primary myotubes, the mRNAs encoding ER-translocated proteins were distributed throughout the multinucleated cells. This resembles the situation in C2 myotubes, where transferrin receptor mRNA was concentrated in the core region, although its distribution did not fully match the distribution of the ER (Ralston et al. 1997). In the primary myotubes, the mRNAs seem to also occupy the inner regions, in contrast to the situation in adult myofibers. Accordingly, mRNA compartmentalization mechanisms seem to appear during the terminal differentiation process. Studies supporting this include the finding that the transcripts of a cytoplasmic nitric oxide synthase shifted from their intracellular location to the subsarcolemmal regions upon terminal differentiation (Luck et al. 1998). Similarly, vimentin and vinculin transcripts were diffuse during the first week of development but were detected in costameres as soon as costameres developed (Cripe et al. 1993; Morris and Fulton 1994). It should be noted that the inner regions of myofibers do not exclude all mRNA species, because myosin light-chain mRNA has been found evenly distributed within the cross-section of the fibers (Billeter et al. 1992). Other studies have reported preferentially peripheral localization to myosin mRNA (Dix and Eisenberg 1991; Jostarndt et al. 1994), but this seems to depend on the type of chain the transcript encodes (Shoemaker et al. 1999). Myoglobin transcripts have been located at the A bands of interfibrillar spaces (Mitsui et al. 1994), in total contrast to the location of the CSQ or DHPRα1 transcripts.

Because the mRNA encoding CSQ was not detected in the inner regions of the myofibers, the protein product has to move relatively long distances from the subsarcolemmal areas to the core of the cell. Providing that the synthesis compartment has a common lumen with the SR, diffusion is one possibility, but experimental evidence favors the view that membrane vesicles carry CSQ from the ER to the terminal cisternae of the SR (Thomas et al. 1989; Nori et al. 2004). In contrast, DHPR is transported through the Golgi elements to transverse tubules, and in myofibers a majority of the Golgi elements are located beneath the sarcolemma (Lu et al. 2001). Furthermore, microtubules that act as tracks for vesicular trafficking in mononucleated cells are dense in the subsarcolemmal region of myofibers (Boudriau et al. 1993). These findings, together with the idea that transcripts of ER-translocated proteins reside beneath the sarcolemma, suggest that membrane protein synthesis, as well as exocytic trafficking, is restricted beneath the sarcolemma in skeletal myofibers. We could not elucidate a crucial role for microtubules or actin filaments, or association with the ER in keeping the transcripts beneath the sarcolemma. Therefore, the positioning mechanisms of mRNAs in myofibers remain an open question.

Earlier work has shown that gene products in myotubes do not reach far from the nucleus of origin (Pavlath et al. 1989; Ralston and Hall 1989). Regarding adult myofibers, it has been shown that gene products located at the neuromuscular junction are solely encoded by the nuclei in the junctional region (Merlie and Sanes 1985). It has remained obscure, however, whether the extrajunctional nuclei are equal or if there are quiescent nuclei present. Our present results show that the genes encoding CSQ or DHPRα1 were active in all the nuclei, and it seems that there were no marked differences in activity between the nuclei.

Taken together, CSQ and DHPRα1 transcripts neither followed the distribution of classical ER markers nor had the respective target organelles. Our findings here locate ER-translocated mRNAs to perinuclear regions and repeating rib-like structures that are in register with I bands beneath the sarcolemma. The results imply that the synthesis machinery of the respective proteins was similarly organized. Furthermore, there was no marked division of labor among the nuclei to transcribe the mRNAs of interest.

Footnotes

Acknowledgements

We thank Pirkko Peronius and Marja Paloniemi for expert assistance. This study was supported by the Academy of Finland (grant no. 73054).

References

Akagi

Yamamoto

Yoshimori

Masaki

Ogawa

Tashiro

(1988) Distribution of protein disulfide isomerase in rat hepatocytes. J Histochem Cytochem 36:1533–1542

Awad

Lightowlers

Young

Chrzanowska-Lightowlers

Lomo

Slater

(2001) Sodium channel mRNAs at the neuromuscular junction: distinct patterns of accumulation and effect of muscle activity. J Neurosci 21:8456–8463

Bergmann

Fusco

(1990) The G protein of vesicular stomatitis virus has free access into and egress from the smooth endoplasmic reticulum in UT-1 cells. J Cell Biol 110:625–635

Billeter

Messerli

Wey

Puntschart

Jostarndt

Eppenber

Parriard

(1992) Fast myosin light chain expression in chicken muscle studied by in situ hybridization. J Histochem Cytochem 40:1547–1557

Boudriau

Vincent

Cote

Rogers

(1993) Cytoskeletal structure of skeletal muscle: identification of an intricate exosarcomeric microtubule lattice in slow- and fast-twitch muscle fibers. J Histochem Cytochem 41:1013–1021

Cripe

Mollis

Fulton

(1993) Vimentin mRNA location changes during muscle development. Proc Natl Acad Sci USA 90:2724–2728

Dix

Eisenberg

(1991) Distribution of myosin mRNA during development and regeneration of skeletal muscle fibers. Dev Biol 143:422–426

Fliegel

Leberer

Green

McLennan

(1989) The fast-twitch muscle calsequestrin isoform predominates in rabbit slow-twitch soleus muscle. FEBS Lett 242:297–300

Fontaine

Changeux

J-P

(1989) Localization of nicotinic acetylcholine receptor alpha-subunit transcripts during myogenesis and motor endplate development in the chick. J Cell Biol 108:1025–1037

10.

Jasmin

Lee

Rotundo

(1993) Compartmentalization of acetylcholinesterase mRNA and enzyme at the vertebrate neuromuscular junction. Neuron 11:467–477

11.

Jolly

Morimoto

Robert-Nicoud

Vourc'h

(1997) HSF1 transcription factor concentrates in nuclear foci during heat shock: relationship with transcription sites. J Cell Sci 110:2935–2941

12.

Jostarndt

Puntschart

Hoppeler

Billeter

(1994) The use of 33P-labelled riboprobes for in situ hybridizations: localization of myosin alkali light-chain mRNAs in adult human skeletal muscle. Histochem J 26:32–40

13.

Kaisto

Metsikkö

(2003) Distribution of the endoplasmic reticulum and its relationship with the sarcoplasmic reticulum in skeletal myofibers. Exp Cell Res 289:47–57

14.

Laemmli

(1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685

15.

Lodish

Froshauer

(1977) Binding of viral glycoprotein mRNA to endoplasmic reticulum membranes is disrupted by puromycin. J Cell Biol 74:358–364

16.

Joseph

Bugnard

Zaal

Ralston

(2001) Golgi complex reorganization during muscle differentiation: visualization in living cells and mechanism. Mol Biol Cell 12:795–808

17.

Luck

Oberbäumer

Blottner

(1998) In situ identification of neuronal nitric oxide synthase (NOS-1) mRNA in mouse and rat skeletal muscle. Neurosci Lett 246:77–80

18.

MacLennan

(2000) Ca²⁺ signalling and muscle disease. Eur J Biochem 267:5291–5297

19.

Merlie

Sanes

(1985) Concentration of acetylcholine receptor mRNA at synaptic regions of adult muscle fibers. Nature 317:66–68

20.

Mitsui

Kawai

Naruo

Saito

(1994) Ultrastructural localization of myoglobin mRNA in human skeletal muscle. Histochemistry 101:99–104

21.

Mitsui

Kawai

Shono

Kawajiri

Kunishige

Saito

(1997) Preferential subsarcolemmal localization of dystrophin and beta-dystroglycan mRNA in human skeletal muscles. J Neuropathol Exp Neurol 56:94–101

22.

Morris

Fulton

(1994) Rearrangement of mRNAs for costamere proteins during costamere development in cultured skeletal muscle from chicken. J Cell Sci 107:377–386

23.

Nakata

Nishita

Yorifuji

(2001) Cytoplasmic γ actin as a Z-disc protein. Biochem Biophys Res Commun 286:156–163

24.

Neuhuber

Gerster

Döring

Glossmann

Tanabe

Flucher

(1998) Association of calcium channel α_1S and β_1a subunits is required for the targeting of β_1a but not of α_1S into skeletal muscle triads. Proc Natl Acad Sci USA 95:5015–5020

25.

Nori

Bortoloso

Frasson

Valle

Volpe

(2004) Vesicle budding from endoplasmic reticulum is involved in calsequestrin routing to sarcoplasmic reticulum of skeletal muscle. Biochem J 379:502–512

26.

Park

Goo

Chung

Kim

Park

(1998) Cloning the genes encoding mouse cardiac and skeletal calse-questrins: expression patterns during embryogenesis. Gene 217:25–30

27.

Pavlath

Rich

Webster

Blau

(1989) Localization of muscle gene products in nuclear domains. Nature 337:570–573

28.

Rahkila

Alakangas

Väänänen

Metsikkö

(1996) Transport pathway, maturation, and targeting of the vesicular stomatitis virus glycoprotein in skeletal muscle fibers. J Cell Sci 109:1585–1596

29.

Rahkila

Luukela

Väänänen

Metsikkö

(1998) Differential targeting of vesicular stomatitis virus G protein and influenza virus hemagglutinin appears during myogenesis of L6 muscle cells. J Cell Biol 140:1101–1111

30.

Rahkila

Väänänen

Saraste

Metsikkö

(1997) Endoplasmic reticulum to Golgi trafficking in multinucleated skeletal muscle fibers. Exp Cell Res 234:452–464

31.

Ralston

Hall

(1989) Transfer of a protein encoded by a single nucleus to nearby nuclei in multinucleated myotubes. Science 244:1066–1069

32.

Ralston

McLaren

Horowitz

(1997) Nuclear domains in skeletal myotubes: the localization of tranferrin receptor mRNA is independent of its half-life and restricted by binding to ribosomes. Exp Cell Res 236:453–462

33.

Ray

Kyselovic

Leddy

Wigle

Jasmin

Tuana

(1995) Regulation of dihydropyridine and ryanodine receptor gene expression in skeletal muscle. Role of nerve, protein kinase C, and cAMP pathways. J Biol Chem 270:25837–25844

34.

Reithmeier

de Leon

MacLennan

(1980) Assembly of the sarcoplasmic reticulum. Cell-free synthesis of the Ca2 + + Mg2 + -adenosine triphosphatase and calsequestrin. J Biol Chem 255:11839–11846

35.

Saito

Seiler

Chu

Fleischer

(1984) Preparation and morphology of sarcoplasmic reticulum terminal cisternae from rabbit skeletal muscle. J Cell Biol 99:875–885

36.

Scott

Simmerman

Collins

Nadal-Ginard

Jones

(1988) Complete amino acid sequence of canine cardiac calsequestrin deduced by cDNA cloning. J Biol Chem 263:8958–8964

37.

Shoemaker

Ryan

Lieber

(1999) Transcript-specific mRNA trafficking based on the distribution of coexpressed myosin isoforms. Cells Tissues Organs 165:10–15

38.

Taneja

Lifshitz

Fay

Singer

(1992) Poly(A)mRNA codistribution with microfilaments: evaluation by in situ hybridization and quantitative digital imaging microscopy. J Cell Biol 119:1245–1260

39.

Thomas

Navarro

Benson

Campbell

Rotundo

Fine

(1989) Newly synthesized calsequestrin, destined for the sarcoplasmic reticulum, is contained in early/intermediate Golgi-derived clathrin-coated vesicles. J Biol Chem 264:3140–3145

40.

Volpe

Martini

Furlan

Meldolesi

(1994) Calsequestrin is a component of smooth muscles: the skeletal- and cardiac-muscle isoforms are both present, although in highly variable amounts and ratios. Biochem J 301:465–469

41.

Volpe

Villa

Podini

Martini

Nori

Panzeri

Meldolesi

(1992) The endoplasmic reticulum-sarcoplasmic reticulum connection: distribution of endoplasmic reticulum markers in the sarcoplasmic reticulum of skeletal muscle fibers. Proc Natl Acad Sci USA 89:6142–6146

42.

Ylä-Outinen

Koivunen

Nissinen

Bjorkstrand

Paloniemi

Korkiamäki

Peltonen

. (2002) NF1 tumor suppressor mRNA is targeted to the cell-cell contact zone in Ca(2 +)-induced keratinocyte differentiation. Lab Invest 82:353–361