Ca2+ release, which is necessary for muscle contraction, occurs at the j-SR (junctional domain of the sarcoplasmic reticulum). It requires the assembly of a large multiprotein complex containing the RyR (ryanodine receptor) and additional proteins, including triadin and calsequestrin. The signals which drive these proteins to the j-SR and how they assemble to form this multiprotein complex are poorly understood. To address aspects of these questions we studied the localization, dynamic properties and molecular interactions of triadin. We identified three regions, named TR1 (targeting region 1), TR2 and TR3, that contribute to the localization of triadin at the j-SR. FRAP experiments showed that triadin is stably associated with the j-SR and that this association is mediated by TR3. Protein pull-down experiments indicated that TR3 contains binding sites for calsequestrin-1 and that triadin clustering can be enhanced by binding to calsequestrin-1. These findings were confirmed by FRET experiments. Interestingly, the stable association of triadin to the j-SR was significantly decreased in myotubes from calsequestrin-1 knockout mice. Taken together, these results identify three regions in triadin that mediate targeting to the j-SR and reveal a role for calsequestrin-1 in promoting the stable association of triadin to the multiprotein complex associated with RyR.

INTRODUCTION

In skeletal muscle fibres, the T-tubules (transverse-tubules) and the SR (sarcoplasmic reticulum) membranes are closely associated to form highly specialized intracellular junctions called triads [1,2]. These junctions allow an efficient cross-talk between voltage-dependent L-type Ca2+ channels, the DHPRs (dihydropyridine receptors) located on the T-tubule membrane, and the intracellular Ca2+ release channels, RyRs (ryanodine receptors), localized on the j-SR (junctional domain of the SR) [3]. Functional coupling between voltage-activated DHPRs and RyRs results in the regulated release of Ca2+ from the SR and the initiation of muscle contraction, a mechanism known as excitation–contraction coupling [4]. Several proteins, including triadin and calsequestrin, are known to assemble into a multimolecular complex with RyRs at the j-SR [510] as demonstrated by protein co-immunoprecipitation and other biochemical techniques [5,811]. This is also supported by immunofluorescence microscopy analysis, which shows that j-SR proteins initially co-localize into discrete clusters, whereas later they are progressively ordered at triads, which in mammals are localized in the SR region surrounding the A-I band junction of the sarcomere [12,13].

The mobility of membrane proteins can be influenced by several factors, including the establishment of protein interactions, which alter their ability to move freely in the lipid bilayer [14]. Generally speaking, the distribution of membrane proteins to specific compartments requires one or more mechanisms, including targeting, which assists proteins to reach a specific membrane compartment, and trapping, which favours the retention of the proteins at a given site [1518]. The mechanisms whereby single j-SR proteins are incorporated into this multiprotein complex are, however, still not clear. Results from available knockout mice indicate that the absence of a single j-SR protein does not alter cluster formation or the localization of other proteins in the j-SR complex, thus suggesting that different or multiple mechanisms/interactions may be involved in targeting j-SR proteins to the multiprotein complex [1926]. In previous studies, we used the FRAP technique to study the dynamic properties of proteins of the j-SR, like RyR and triadin [27,28]. Mobility of membrane proteins can be evaluated by such experiments, which can yield information on the diffusion coefficient and on the fraction of the protein of interest that is free to diffuse within the lipid bilayer, also referred to as the mobile fraction [14,29]. These two parameters are used to define the mobility properties of a protein. The recovery after photobleaching of the fluorescent signal associated with GFP-tagged j-SR proteins was very efficient in undifferentiated myocytes. Such dynamic behaviour indicates that these proteins were not significantly limited in their mobility. Following myotube differentiation, j-SR proteins formed discrete clusters and the mobile fraction of the GFP-tagged proteins decreased, suggesting the establishment of stable protein–protein interactions within the assembling multimolecular complex [27,28].

To understand further the mechanisms involved in the targeting of a membrane protein to the j-SR, we focused our studies on triadin. Four isoforms of triadin have been identified; their molecular masses are 95, 51, 49 and 32 kDa, where the 95 kDa and 32 kDa isoforms are the predominant isoforms of skeletal and cardiac muscle respectively [3032]. All triadin isoforms contain a short N-terminal cytoplasmic domain, a single transmembrane region and a C-terminal region, located in the lumen of the SR. The length of the C-terminal region determines the size of the different isoforms [3133]. Triadins bind to both RyR and calsequestrin and thus have been proposed to serve as a linker between the two proteins [5,34,35]. In skeletal muscle, the deletion of the triadin gene results in a reduction in the amplitude of Ca2+ transients, an increase in myoplasmic resting free Ca2+ and a reduction in muscle strength [24,26,36,37]. In addition, triadin knockout is also associated with the down-regulation of RyR, calsequestrin, junctin, junctophilin-1 and junctophilin-2 protein levels and by the appearance of longitudinally oriented triads [24,26].

In the present paper, we report the results of studies on protein dynamics using FRAP and protein–protein interaction experiments aimed at identifying amino acid regions in triadin that regulate the targeting and retention of this protein to the j-SR. Collectively, the results obtained indicate that different regions in triadin are required for targeting and retention to the j-SR and that the interaction with calsequestrin-1 is a key element in promoting the stable association of triadin with the multiprotein complex linked with the RyR Ca2+ release channel.

EXPERIMENTAL

Microsome preparation

Microsomes were prepared from skeletal muscle of CD1 mouse or from HEK (human embryonic kidney)-293 cells transfected with the constructs of interest [38]. The use of animal experiments has been approved by the responsible Italian authorities (Project number J-19/12/2011). Tissue samples or cultured cells were homogenized in ice-cold buffer A (0.32 M sucrose, 5 mM Hepes, pH 7.4, and 0.1 mM PMSF) using a potter for cells or a homogenizer for tissues. Homogenates were centrifuged at 7000 g for 5 min at 4°C. The supernatant was centrifuged at 100000 g for 1 h at 4°C. The microsomes were resuspended in buffer A and stored at −80°C. The protein concentration of the microsomal fraction was quantified using the Bradford protein assay kit (Bio-Rad Laboratories).

Microsome solubilization and GST pull-down

Microsomes prepared from mouse skeletal tissue were solubilized according to the method described previously [39] with some minor modifications. Briefly, microsomes were solubilized at a protein concentration of 1 mg/ml in a buffer containing 2% Triton X-100, 1 M NaCl, 1 mM DTT, 20 mM Tris/HCl, pH 7.4, and a protease inhibitor mixture [leupeptin, aprotinin, antipain, chymostatin and pepstatin A at a final concentration of 2 μg/ml each (Sigma)] for 1 h at 4°C. Microsomes prepared from HEK-293 cells were solubilized at a protein concentration of 1 mg/ml for 3 h at 4°C in a buffer containing 10 mM Tris/HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 10% glycerol, 1 mM Na3VO4, 1 mM PMSF and a protease inhibitor mixture. Solubilized proteins were obtained by centrifugation (Beckman Ti90 rotor) at 145000 g for 45 min at 4°C. GST pull-down was performed as described previously [40] with some minor modifications. A total of 250 μg of solubilized microsomal proteins were incubated with 25 μg of GST fusion proteins in a buffer containing 10 mM Tris/HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 10% glycerol, 1 mM Na3VO4, 1 mM PMSF and a protease inhibitor mixture for 2 h at 4°C. After incubation, the GST fusion protein complexes were washed three times with 20 mM Tris/HCl, pH 7.4, 150 mM NaCl and 0.2% Triton X-100. The bound proteins were eluted by boiling in SDS/PAGE sample buffer and subjected to SDS/PAGE.

Production and purification of GST fusion proteins

GST fusion proteins were induced in BL21 cells with 1 mM IPTG for 3 h at 30°C. Cells were harvested and centrifuged at 3345 g for 10 min at 4°C. The pellet was resuspended in an ice-cold buffer containing PBS, 1% Triton X-100 and 20 mM EDTA and lysed by sonication on ice. The soluble fraction was obtained by centrifugation at 16100 g for 15 min at 4°C. The fusion proteins were immobilized by incubating 1 ml of the soluble fraction with 100 μl of glutathione–Sepharose 4B beads (GE Healthcare) for 10 min and washed three times with 1 ml of a buffer containing PBS and 1% Triton X-100. The beads were finally resuspended with an equal volume of PBS.

Western blot analysis

Protein samples were separated by SDS/PAGE (5 or 10% gel), as described previously [38]. Filters were incubated with the following primary antibodies: rabbit anti-triadin antibody (kindly provided by Professor Isabelle Marty, Grenoble Institut des Neurosciences, France); mouse anti-calsequestrin-1 antibody, clone MA3-913 (Thermo Scientific); rabbit serum anti-RyR1 antibody (produced in our laboratory); mouse anti-SERCA-1 [SR/ER (endoplasmic reticulum) Ca2+-ATPase 1] antibody (Developmental Studies Hybridoma Bank); or mouse anti-c-Myc antibody (Clontech/Zymed Laboratories), which were diluted in blocking buffer overnight at 4°C with agitation. Filters were washed three times with washing buffer (0.5% non-fat dried skimmed milk powder, 50 mM Tris/HCl, pH 7.4, 150 mM NaCl and 0.2% Tween 20) for 10 min each and incubated with horseradish peroxidase-conjugated secondary antibody and detected using the ECL system (ECL Western Blot Detection Reagents, GE Healthcare).

Generation of expression vectors

Human triadin cDNA was cloned into the pEGFP-N2 vector (Clontech) using the EcoRI–BamH1 sites. Triadin mutants were generated by PCR using specific primers to amplify the regions of interest. The amplified sequences were cloned in pEGFP-N1, -N2 or -N3 vectors, according to the translation reading frame and sequenced using an ABI Prism 7900 apparatus (Applied Biosystems). For experiments that required the expression of a Myc-tagged triadin, the cDNAs coding for either full-length triadin or triadin deletion mutants were cloned into the pcDNA3.1 vector (Invitrogen). For FRET analysis the triadin cDNA was cloned into the pEYFP or the pECFP vectors (Clontech). Calsequestrin-1 cDNA was amplified from total RNA extracted from human skeletal muscle tissue using specific primers. The cDNA was inserted into the pEYFP or the pECFP vectors for FRET analysis or into the pcDNA3.1 vector, for expression of a Myc-tagged calsequestrin-1. For GST pull-down experiments, triadin proteins were generated by PCR using specific primers to amplify the regions of interest. The amplified sequences were cloned into the pGEX4T1 vector (GE Healthcare). RyR1 expression vectors were generated as described previously [41].

Primary cultures of rat and mouse skeletal muscle cells, culture of cell lines and DNA transfection

Myoblasts were obtained from the hind limb muscles of 2-day-old rats (Sprague–Dawley, Harlan Laboratories), CD1 mice, calsequestrin-1 or RyR1 knockout mice. The cell suspension was plated on to 0.025% laminin-coated LabTek chambers (Nalge Nunc International) or 0.025% laminin-coated glass coverslips. Cells were grown at 37°C at 5% CO2. After 2 days, cells were transfected with the Lipofectamine™-Plus method (Invitrogen) using the DNA plasmids as described above (i.e. all triadin–GFP expression vectors, triadin and calsequestrin pEYFP and pECFP expression vectors), following the manufacturer's instructions. Myoblasts were induced to differentiate with αMEM [α-minimum essential medium (Sigma)] containing 2 mM L-glutamine (Lonza), 100 μg/ml streptomycin (Lonza), 100 units/ml penicillin (Lonza), 1 mM sodium pyruvate (Lonza), 1 mM dexamethasone and 50 mM cortisol, supplemented with 10% heat-inactivated FBS (Sigma–Aldrich) and 5% heat-inactivated horse serum (Biochrom).

NIH 3T3 cells were cultured in DMEM (Dulbecco's modified Eagle's medium), supplemented with 10% heat-inactivated FBS (Lonza), 2 mM L-glutamine, 100 μg/ml streptomycin (Bio-Whittaker), 100 units/ml penicillin and 1 mM sodium pyruvate. NIH 3T3 cells were transfected with the Lipofectamine™-Plus method with GFP expression vectors for FRAP analysis or with YFP or CFP expression vectors for FRET analysis, as described in the Results section.

HEK-293 cells were cultured in αMEM containing 2 mM L-glutamine, 100 μg/ml streptomycin (Lonza), 100 units/ml penicillin and 1 mM sodium pyruvate, supplemented with 10% heat-inactivated FBS (Lonza). Cells were transfected with the Lipofectamine™-Plus method with triadin, calsequestrin and RyR1 expression vectors. Microsomal proteins from transfected HEK-293 cells were prepared and used for GST pull-down experiments.

For localization studies, immunofluorescence staining was performed as described previously [41]. A mouse monoclonal antibody recognizing RyRs was used to detect the j-SR regions in differentiated myotubes (clone 34C, Thermo Fisher Scientific). Alternatively, a rabbit polyclonal antibody against RyR1, produced in our laboratory was used to counterstain myotubes expressing triadin–Myc [41]. A monoclonal antibody against α-actinin (clone EA-53; Sigma–Aldrich) was used to identify the Z-disks in differentiated myotubes. A monoclonal anti-Myc antibody (Clontech/Zymed Laboratories) was used to detect triadin–Myc fusion proteins. Cy3 (indocarbocyanine) or Cy2 (carbocyanine)-conjugated anti-mouse or anti-rabbit secondary antibodies (Jackson Laboratories) were used for immunofluorescence detection.

FRAP

FRAP experiments were performed on 12-day differentiated rat or mouse myocytes or on NIH 3T3 cells expressing the GFP-fusion proteins of interest, using a confocal laser-scanning microscope (ZEISS LSM 510, Carl Zeiss). Cells were imaged in buffered medium, containing 140 mM NaCl, 5 mM KCl, 10 mM glucose, 1 mM MgCl2, 0.1 mM CaCl2, 20 mM Hepes and 0.4 mM EGTA at 37°C. A 63×1.4 NA (numerical aperture) Plan-Apochromat oil immersion objective was used and cells were imaged with a pinhole aperture of 4.96 airy units, corresponding to a confocal section thickness of 3.5 μm. GFP fluorescence before bleaching and its recovery after bleaching was measured with the 488 nm line of argon laser at low laser power (0.5%). After the acquisition of ten pre-bleach images, a 50% photobleaching was performed using the argon laser lines 458, 477 and 488 nm. The photobleached area was 1.8 μm in diameter. Recovery was measured by time-lapse imaging at 50–300 ms intervals over a period of 1–10 min, until the fluorescence level reached a plateau. Throughout the experiment, fluorescence intensities were acquired for the bleached region (Ifrap), for the whole cell (Iwhole) and for the background (Ibase). Data analyses were performed using the IgorPro software (WaveMetrix). Data were collected from at least three independent experiments. Statistical analysis was performed using either an unpaired Student's t test or a Kruskal–Wallis test followed by a Dunn post-test.

FRET

FRET analyses were performed on NIH 3T3 cells fixed 2 days after transfection or on primary skeletal muscle cells fixed 4, 8 and 12 days after induction of differentiation. Fixation was obtained with 3% paraformaldehyde and 2% sucrose in PBS for 7 min. Paraformaldehyde was then removed and cells were washed with 0.2% BSA in PBS solution at room temperature (22°C) and mounted with Mowiol. FRET was measured using a confocal laser-scanning microscope (ZEISS LSM 510). CFP or YFP images were collected with a 63×1.4 NA Plan-Apochromat oil immersion objective lens with a pinhole aperture of 8 airy units, corresponding to a confocal section thickness <5.5 μm. FRET was measured with the 488 nm line of argon laser set on 458 nm to excite CFP and to 514 nm to excite YFP. FRET was determined with the acceptor photobleaching protocol: cells were scanned in the YFP channels in a defined ROI (region of interest) for four iterations using a 514 nm argon laser line. ROIs were corrected for background fluorescence and for bleaching during acquisition. FRETe (FRET efficiency) was then estimated using the following equation: FRETe=(Fpb−Fd)/Fpb, where Fpb and Fd are the donor fluorescence intensity before and after acceptor photobleaching respectively. Data were collected from at least three independent experiments. A statistical analysis was performed using an unpaired Student's t test.

RESULTS

Three distinct regions contribute to targeting of triadin to the j-SR

To identify the sequences required for targeting of triadin to the j-SR, we generated a series of GFP-tagged deletion mutants of the 95 kDa triadin isoform. Plasmids expressing these proteins were transfected in differentiating primary rat myoblasts and the localization of these GFP-tagged proteins assessed using a confocal microscope. To avoid potential problems due to overexpression of exogenous proteins, cells expressing low levels of GFP fluorescence were selected. To follow muscle cell differentiation and to identify j-SR regions, the cells were immunolabelled with antibodies against endogenous RyRs. As a control, we also expressed a full-length triadin–GFP protein and followed its localization during myoblast differentiation. Confocal microscope analyses revealed that in undifferentiated myoblasts the full-length triadin–GFP protein was diffused throughout the entire cell, in agreement with the presence of a disorganized ER/SR (Figures 1A–1C). In the initial stages of differentiation, triadin–GFP began to form discrete clusters, which were dispersed throughout the myotubes. There, it co-localized with the endogenous RyRs (Figures 1D–1F). This agrees with the notion that clusters represent the initial sites of assembly of the j-SR [12,13]. In terminally differentiated myotubes, triadin–GFP, along with the endogenous RyRs, was organized in double rows that closely resembled the mature triads observed in skeletal muscle fibres (Figures 1G–1I) that are positioned at the A-I band junctions as shown by immunolabelling the Z-disks with an antibody against α-actinin (Figures 1J–1L). Cloning of the GFP protein at the N- or the C-terminus of triadin did not affect localization of the resulting fusion proteins (results not shown). Data shown hereafter thus refer to experiments performed with fusion proteins containing GFP cloned at the C-terminus of triadin (triadin–GFP).

Localization of triadin–GFP in rat myotubes

Figure 1
Localization of triadin–GFP in rat myotubes

Primary rat myoblasts were transfected with expression vectors for triadin–GFP and induced to differentiate for 4 (AC), 8 (DF) and 12 (GL) days. Cells were counterstained with monoclonal antibodies against the RyRs (B, E and H) or with antibodies against α-actinin and a Cy3-conjugated anti-mouse secondary antibody (K).

Figure 1
Localization of triadin–GFP in rat myotubes

Primary rat myoblasts were transfected with expression vectors for triadin–GFP and induced to differentiate for 4 (AC), 8 (DF) and 12 (GL) days. Cells were counterstained with monoclonal antibodies against the RyRs (B, E and H) or with antibodies against α-actinin and a Cy3-conjugated anti-mouse secondary antibody (K).

Work from several laboratories have shown that triadin contains sites of interaction with RyR1 and calsequestrin at both the N- and C-terminal domains. Sites of interaction with RyR were found in the amino acid regions 18–47 and 200–232, whereas interaction with calsequestrin was shown to occur in the amino acid region 200–224 [34,35]. A computer-aided analysis of the amino acid sequence of the 95 kDa isoform of triadin revealed a significant level of sequence similarity (42.6%) in the large intraluminal C-terminus between the regions corresponding to amino acids 232–440 and amino acids 441–729. Accordingly, we prepared six triadin–GFP mutants carrying deletions of selected regions in the N- or C-terminus. As shown in Figure 2 in fully differentiated myotubes, triadin(1)–GFP to triadin(5)–GFP mutants co-localized with endogenous RyR at the j-SR (Figures 2D–2R). By contrast, the triadin(6)–GFP mutant, which lacks most of the intraluminal triadin tail, did not co-localize with RyRs (Figures 2S–2U). These results indicated that the region between amino acids 106–214 might contain amino acid sequences required for triadin localization at the j-SR. Unexpectedly, however, a triadin–GFP fusion protein with selective deletion of amino acids 106–214 [triadin(7)–GFP] co-localized with endogenous RyR (Figures 3A–3C), indicating that in the absence of amino acids 106–214, additional sequences can support triadin targeting to the j-SR.

Localization of triadin deletion mutants in rat myotubes

Figure 2
Localization of triadin deletion mutants in rat myotubes

Primary rat myoblasts were transfected with expression vectors for triadin–GFP (AC) or triadin–GFP deletion mutants (D, G, J, M, P and S) and induced to differentiate for 12 days. Cells were counterstained with mouse monoclonal antibodies against the RyRs and a Cy3-conjugated anti-mouse secondary antibody (B, E, H, K, N, Q and T). A schematic representation of full-length human triadin and the N- and C-terminal deletion mutants are indicated. The transmembrane domain of triadin is shown as a black box. Scale bar, 2 μm.

Figure 2
Localization of triadin deletion mutants in rat myotubes

Primary rat myoblasts were transfected with expression vectors for triadin–GFP (AC) or triadin–GFP deletion mutants (D, G, J, M, P and S) and induced to differentiate for 12 days. Cells were counterstained with mouse monoclonal antibodies against the RyRs and a Cy3-conjugated anti-mouse secondary antibody (B, E, H, K, N, Q and T). A schematic representation of full-length human triadin and the N- and C-terminal deletion mutants are indicated. The transmembrane domain of triadin is shown as a black box. Scale bar, 2 μm.

Localization of triadin deletion mutants in rat myotubes

Figure 3
Localization of triadin deletion mutants in rat myotubes

Primary rat myoblasts were transfected with expression vectors for triadin–GFP deletion mutants (A, D, G, J, M, P and S) and induced to differentiate for 12 days. Cells were counterstained with mouse monoclonal antibodies against the RyRs and a Cy3-conjugated anti-mouse secondary antibody (B, E, H, K, N, Q and T). A schematic representation of the combined GFP deletion mutants is indicated. The transmembrane domain of triadin is shown as a black box. Scale bar, 2 μm.

Figure 3
Localization of triadin deletion mutants in rat myotubes

Primary rat myoblasts were transfected with expression vectors for triadin–GFP deletion mutants (A, D, G, J, M, P and S) and induced to differentiate for 12 days. Cells were counterstained with mouse monoclonal antibodies against the RyRs and a Cy3-conjugated anti-mouse secondary antibody (B, E, H, K, N, Q and T). A schematic representation of the combined GFP deletion mutants is indicated. The transmembrane domain of triadin is shown as a black box. Scale bar, 2 μm.

We next generated three double-deletion triadin–GFP mutants lacking amino acids 106–214 and either N-terminal [triadin(8)–GFP] or C-terminal [triadin(9)–GFP and triadin(10)–GFP] sequences. As shown in Figure 3, triadin(8)–GFP did not co-localize with RyRs to the j-SR (Figures 3D–3F) indicating that the combined deletion of amino acids 18–47 and amino acids 106–214 affects triadin targeting to the j-SR. By contrast, both triadin(9)–GFP and triadin(10)–GFP co-localized with RyRs. This indicated that even in the absence of amino acids 106–214, either the region between amino acids 233–440 or that between amino acids 441–729 are equally effective in targeting triadin to the j-SR (Figures 3G–3I and 3J–3L).

To complete our analysis, we prepared three additional triadin–GFP mutants to verify whether region 232–729, or either one of the regions 233–440 or 441–729, can target triadin to the j-SR in the absence of the N-terminal residues 18–47. As shown in Figures 3(M)–3(O), deletion of residues 18–47 and 233–729 [triadin(11)–GFP] abolished triadin targeting to the j-SR. By contrast, the two mutants triadin(12)–GFP and triadin(13)–GFP, carrying a deletion of residues 18–47 and of amino acids 233–440 or amino acids 441–729, were still able to localize to the j-SR (Figures 3P–3R and 3S–3U).

To exclude that the localization of the triadin deletion mutants could be affected by GFP, we expressed representative triadin deletion mutants as fusion proteins in-frame with a Myc tag epitope at the C-terminus of the triadin sequence. A full-length triadin–Myc fusion protein was also expressed as a control. No difference was observed in triadin localization when the full-length protein was expressed in-frame with either a GFP or a Myc tag (Figure 2, and Supplementary Figure S1 at http://www.biochemj.org/bj/458/bj4580407add.htm). Similarly, triadin(1)–Myc [corresponding to mutant triadin(4)–GFP] co-localized with RyRs at the j-SR in differentiated myotubes. On the other hand, triadin(2)–Myc and triadin(3)–Myc did not localize to the j-SR, but presented with a distribution pattern superimposable to that observed for their GFP-tagged counterparts [triadin(5)–GFP and triadin(7)–GFP respectively]. Taken together, these results suggest that localization of triadin mutants was not dependent on the presence of GFP at their C-terminus (Supplementary Figure S1).

In summary, results reported in Figures 2 and 3 showed that three regions in triadin may contribute to targeting to the j-SR, one in the cytoplasmic domain (amino acids 18–47) and two in the luminal tail of triadin (amino acids 106–214 and 233–729). As shown in Figures 2(D)–2(F), 2(M)–2(R) and 3(A)–3(C), the presence of at least two of these three regions is sufficient for triadin to be located at the j-SR. Conversely, mutants deleted in any two of these three regions do not localize to the j-SR (Figures 2S–2U, 3D–3F and 3M–3O). For convenience, we will therefore refer to these three regions as targeting regions TR1, TR2 and TR3. Given the sequence homology and the functional properties of the two amino acid regions 233–440 or 441–729 in TR3 we refer to them as TR3a and TR3b. Further attempts to re-define minimal sequences required for triadin-selective targeting inside TR1 and TR2 did not result in significant outcomes.

TR3, but not TR1 or TR2, contributes to the regulation of triadin mobility at the j-SR

Previous results on the dynamic properties of triadin revealed that, following localization at the j-SR, the triadin mobile fraction was significantly reduced, indicating that the protein became stably connected within the multiprotein complex associated with the RyR Ca2+ release channel [27]. In order to investigate whether triadin mobility is dependent on TR1, TR2 and/or TR3, three triadin–GFP deletion mutants lacking either one of these three regions were expressed in differentiating myotubes and their mobile fraction analysed by FRAP experiments (Figure 4). Cells expressing low levels of GFP fluorescence were selected to avoid potential problems due to overexpression. The mobile fraction of single deletion mutants was evaluated in at least three independent experiments. The mobile fraction of triadin mutants with a TR1 or TR2 deletion (ΔTR1–GFP or ΔTR2–GFP mutants) was not significantly different (19.33±9.24%, n=6 and 23.92±7.68%, n=13 respectively), from that of full-length triadin–GFP (26.1±9.11%, n=22). By contrast, the triadin mutant with a TR3 deletion (ΔTR3–GFP) displayed a significantly higher mobile fraction (69.84±15.36%, n=27). Since we observed previously that TR3a and TR3b are equally able to support the localization of triadin to the j-SR, we next measured the mobile fraction of triadin proteins deleted in either TR3a or TR3b. As shown in Figure 4, triadin mutants containing either one of the two regions displayed a mobile fraction comparable with that of full-length triadin (ΔTR3a–GFP, 34.75±10.98%, n=16; and ΔTR3b–GFP, 26.27±11.48%, n=18), indicating that both TR3a and TR3b regions are sufficient to mediate a stable association of triadin to the j-SR. Triadin mutants containing only TR3a or TR3b in the absence of either TR1 or TR2, showed mobile fraction values which were intermediate between those of the full-length triadin and those observed in triadin mutants deleted of the entire TR3 (Figure 4).

FRAP analysis on 12-day differentiated myotubes expressing triadin–GFP and triadin deletion mutants

Figure 4
FRAP analysis on 12-day differentiated myotubes expressing triadin–GFP and triadin deletion mutants

FRAP analysis was performed on 12-day differentiated myotubes expressing either triadin–GFP or triadin–GFP deletion mutants. Data are expressed as percentage of mobile fraction±S.D. The n values were as follows: triadin–GFP (n=22), ΔTR1–GFP (n=6), ΔTR2–GFP (n=13), ΔTR3–GFP (n=27), ΔTR3a–GFP (n=16), ΔTR3b–GFP (n=18), triadin ΔTR1+ΔTR3b–GFP (n=15), triadin ΔTR2+ΔTR3b–GFP (n=17), triadin ΔTR1+ΔTR3a–GFP (n=10) and triadin ΔTR2+ΔTR3a–GFP (n=12). The asterisks indicate statistical significance compared with the mobile fraction of triadin–GFP, as evaluated by Kruskal–Wallis and Dunn statistical test analysis (P<0.05).

Figure 4
FRAP analysis on 12-day differentiated myotubes expressing triadin–GFP and triadin deletion mutants

FRAP analysis was performed on 12-day differentiated myotubes expressing either triadin–GFP or triadin–GFP deletion mutants. Data are expressed as percentage of mobile fraction±S.D. The n values were as follows: triadin–GFP (n=22), ΔTR1–GFP (n=6), ΔTR2–GFP (n=13), ΔTR3–GFP (n=27), ΔTR3a–GFP (n=16), ΔTR3b–GFP (n=18), triadin ΔTR1+ΔTR3b–GFP (n=15), triadin ΔTR2+ΔTR3b–GFP (n=17), triadin ΔTR1+ΔTR3a–GFP (n=10) and triadin ΔTR2+ΔTR3a–GFP (n=12). The asterisks indicate statistical significance compared with the mobile fraction of triadin–GFP, as evaluated by Kruskal–Wallis and Dunn statistical test analysis (P<0.05).

To identify specific domains in TR3 that could mediate an association with the j-SR we performed a sequence analysis of the TR3 region. This revealed that TR3 contains a coiled-coil domain (amino acids 308–335), two cysteine residues at positions 270 and 671 and an N-glycosylation consensus site at residue 647 [42]. To investigate whether these domains can affect the mobile fraction of triadin, we generated mutant ∆335–729–GFP (containing the coiled-coil domain), C270S–GFP, C671S–GFP and the double mutant C270S/C671S–GFP (where cysteine residues were mutated to serine) and mutant N647S–GFP (where the N-glycosylation site was mutated). Transfection of cells with these mutants indicated that they were all able to localize to the j-SR (results not shown). Furthermore, analysis of the mobile fraction of all these mutants upon expression in differentiated myotubes showed that none of these sites can independently contribute to the association of triadin to the j-SR (Table 1).

Table 1
FRAP analysis on primary rat myotubes

FRAP analysis was performed on 12-day differentiated myotubes expressing triadin–GFP mutants. Data are expressed as mean value of percentage of mobile fraction±S.D. ***P<0.05 compared with triadin–GFP and analysed by Kruskal–Wallis and Dunn statistical tests.

Protein expressed Mobile fraction (%) 
Triadin–GFP 26.41±9.11 (n=22) 
Triadin Δ335–729–GFP 46.9±16.97 (n=18)*** 
Triadin C270S–GFP 30.11±17.2 (n=18) 
Triadin C671S–GFP 24.37±16.3 (n=18) 
Triadin C270S/C671S–GFP 31.54±13.2 (n=26) 
Triadin N647S–GFP 25.8±8.87 (n=5) 
Protein expressed Mobile fraction (%) 
Triadin–GFP 26.41±9.11 (n=22) 
Triadin Δ335–729–GFP 46.9±16.97 (n=18)*** 
Triadin C270S–GFP 30.11±17.2 (n=18) 
Triadin C671S–GFP 24.37±16.3 (n=18) 
Triadin C270S/C671S–GFP 31.54±13.2 (n=26) 
Triadin N647S–GFP 25.8±8.87 (n=5) 

Multiple sites of interaction with calsequestrin-1 in TR2 and TR3 promote triadin–triadin interactions in vitro

Since the mobile fraction of a protein can be affected by the establishment of protein–protein interactions, we investigated whether TR1, TR2 or TR3 might contain sites of interaction with proteins of the j-SR. In vitro pull-down experiments were performed using GST fusion proteins covering either TR1 (GST–TR1), TR2 (GST–TR2), TR3a (GST–TR3a) or TR3b (GST–TR3b). A GST fusion protein containing an extended TR2 (amino acids 68–264, GST–TR2*) was also prepared to cover the boundary region between TR2 and TR3. Experiments performed with solubilized microsomal proteins from mouse skeletal muscle tissue showed that GST–TR2, GST–TR3a and GST–TR3b, but not GST–TR1, were all able to pull-down RyR1, calsequestrin 1 and triadin. No interaction was observed with SERCA, a protein of the longitudinal SR (Figure 5A). To distinguish between direct or indirect protein interactions, Myc-tagged triadin, RyR1 or calsequestrin-1 were expressed in the HEK-293 cells and pull-down experiments were performed with proteins solubilized from transfected cells (Figure 5B). No interaction was observed between all GST–triadin fusion proteins and RyR1. GST–TR2* was the only GST fusion protein able to pull-down triadin expressed in transfected cells. By contrast, GST–TR2, GST–TR2*, GST–TR3a and GST–TR3b were all able to bind calsequestrin-1 expressed in transfected HEK-293 cells. To verify whether the triadin–triadin interaction observed in pull-down experiments with solubilized microsomal proteins from skeletal muscle tissue could be mediated by calsequestrin-1, we co-expressed triadin and calsequestrin-1 in HEK-293 cells. Under these experimental conditions, all GST proteins covering the triadin intraluminal domain were able to interact with the Myc-tagged triadin co-expressed with calsequestrin-1 (Figure 5C), suggesting that calsequestrin-1 acts as a scaffold in promoting triadin clustering.

GST pull-down experiments on the microsomal fraction of mouse skeletal muscles and of HEK-293 cells

Figure 5
GST pull-down experiments on the microsomal fraction of mouse skeletal muscles and of HEK-293 cells

(A) The microsomal fraction (250 μg) from mouse skeletal muscles were used in GST pull-down experiments using GST–triadin fusion proteins containing TR1, TR2, TR2*, TR3a and TR3b. Proteins were separated by SDS/PAGE, transferred to membranes and detected by specific antibodies. Solubilized microsomes (30 μg) were loaded in the input lane. (B) The microsomal fraction (250 μg) of HEK-293 cells expressing either RyR1, triadin–Myc or calsequestrin-1–GFP, were used in GST pull-down experiments using GST–triadin fusion proteins containing TR1, TR2, TR2*, TR3a and TR3b. Proteins were separated by SDS/PAGE, transferred to membranes and detected by specific antibodies against RyR, calsequestrin-1 and the Myc tag. (C) The microsomal fraction (250 μg) of HEK-293 cells co-transfected with triadin–Myc and calsequestrin-1–GFP were used in GST pull-down experiments using GST–triadin fusion proteins containing TR1, TR2, TR2*, TR3a and TR3b. Proteins were separated by SDS/PAGE, transferred to membranes and detected by an anti-Myc antibody to analyse the Myc-tagged triadin protein.

Figure 5
GST pull-down experiments on the microsomal fraction of mouse skeletal muscles and of HEK-293 cells

(A) The microsomal fraction (250 μg) from mouse skeletal muscles were used in GST pull-down experiments using GST–triadin fusion proteins containing TR1, TR2, TR2*, TR3a and TR3b. Proteins were separated by SDS/PAGE, transferred to membranes and detected by specific antibodies. Solubilized microsomes (30 μg) were loaded in the input lane. (B) The microsomal fraction (250 μg) of HEK-293 cells expressing either RyR1, triadin–Myc or calsequestrin-1–GFP, were used in GST pull-down experiments using GST–triadin fusion proteins containing TR1, TR2, TR2*, TR3a and TR3b. Proteins were separated by SDS/PAGE, transferred to membranes and detected by specific antibodies against RyR, calsequestrin-1 and the Myc tag. (C) The microsomal fraction (250 μg) of HEK-293 cells co-transfected with triadin–Myc and calsequestrin-1–GFP were used in GST pull-down experiments using GST–triadin fusion proteins containing TR1, TR2, TR2*, TR3a and TR3b. Proteins were separated by SDS/PAGE, transferred to membranes and detected by an anti-Myc antibody to analyse the Myc-tagged triadin protein.

Calsequestrin-1 is required for triadin–triadin interactions in vivo

To validate further the possible role of calsequestrin-1 in mediating triadin–triadin interactions, acceptor photobleaching FRET experiments were performed. Both triadin and calsequestrin-1 proteins tagged with CFP or YFP were expressed in primary myotubes and in vivo interactions between the two proteins assessed on the basis of the calculated FRETe. The CFP or YFP fluorescent proteins were cloned at either the N- or C-terminus of the proteins and FRETe was accordingly calculated (Table 2). In myotubes co-expressing triadin proteins fused with CFP or YFP, a FRET signal was observed indicating that triadin–triadin interactions occur in differentiated myotubes. Similarly, a FRET signal was observed when calsequestrin-1–CFP and triadin–YFP or YFP–calsequestrin-1 and triadin–CFP were co-expressed in differentiating myotubes, confirming further the occurrence of an interaction between triadin and calsequestrin-1 in muscle cells.

Table 2
FRET analysis on primary rat myotubes or NIH 3T3 cells

FRET analysis on primary rat myotubes or NIH 3T3 cells. FRET analysis was performed on either differentiating myotubes or NIH 3T3 cells expressing YFP- or CFP-tagged triadin (TRD) or calsequestrin-1 (CSQ-1). Data are expressed as percentage of FRET efficiency±S.D.

Protein expressed 12-day myotubes NIH 3T3 cells 
eCFP/eYFP control 27.83±11.95 (n=29) 19.45±4.59 (n=31) 
CFP–TRD/YFP–TRD 36.30±12.82 (n=10) 0 (n=13) 
TRD–CFP/TRD–YFP 15±4.5 (n=11) 0 (n=32) 
CSQ-1/CFP–TRD/YFP–TRD – 11.68±5.46 (n=10) 
CSQ-1/TRD–CFP/TRD–YFP – 12±3.62 (n=19) 
CSQ–CFP/TRD–YFP 19.8±6.5 (n=10) 0 (n=19) 
YFP–CSQ/TRD–CFP 18±6 (n=10) 0 (n=21) 
Protein expressed 12-day myotubes NIH 3T3 cells 
eCFP/eYFP control 27.83±11.95 (n=29) 19.45±4.59 (n=31) 
CFP–TRD/YFP–TRD 36.30±12.82 (n=10) 0 (n=13) 
TRD–CFP/TRD–YFP 15±4.5 (n=11) 0 (n=32) 
CSQ-1/CFP–TRD/YFP–TRD – 11.68±5.46 (n=10) 
CSQ-1/TRD–CFP/TRD–YFP – 12±3.62 (n=19) 
CSQ–CFP/TRD–YFP 19.8±6.5 (n=10) 0 (n=19) 
YFP–CSQ/TRD–CFP 18±6 (n=10) 0 (n=21) 

To verify whether triadin–triadin interactions can occur in the absence of other muscle proteins, either the CFP–triadin and YFP–triadin pair or the triadin–CFP and triadin–YFP pair were transfected in NIH 3T3 cells. In contrast with the results obtained in myotubes, no FRET signal was observed when either set of pairs of triadin fusion proteins were expressed in NIH 3T3 cells (Table 2). However, when calsequestrin-1 was co-transfected in NIH 3T3 cells along with the CFP–triadin and YFP–triadin pair or with the triadin–CFP and triadin–YFP pair, a clear FRET signal was observed (Table 2). These results support those obtained from pull-down experiments in vitro, indicating that calsequestrin-1 can actually promote triadin–triadin interaction in vivo as well.

Binding to calsequestrin-1 decreases triadin mobility in the non-muscle NIH 3T3 cell line

Results obtained by GST pull-down experiments and FRET analysis showed that calsequestrin-1 can bind directly to the intraluminal region of triadin and promote triadin–triadin interactions. We therefore wondered whether binding to calsequestrin-1 might also directly affect triadin mobility. Accordingly, we measured the mobile fraction of triadin–GFP in the presence or in the absence of calsequestrin-1 in NIH 3T3 cells. FRAP analysis performed in NIH 3T3 cells showed that the mobile fraction of triadin–GFP has an average value of 88.06±8.14% (n=53); however, co-expression of triadin–GFP with calsequestrin-1 resulted in a significant reduction in the mobile fraction of triadin–GFP to 56.55±20.93%, n=68 (P<0.05) (Figures 6A and 6B). In agreement with the proposed role of TR3 in regulating the mobility of triadin in muscle cells via interactions with calsequestrin-1, no reduction was observed in the mobile fraction of triadin–GFP proteins deleted in the luminal domain (Δ215–729 or Δ106–729) when co-expressed with calsequestrin-1 in NIH 3T3 cells (Figures 6C and 6D). Analogous experiments based on co-expression of RyR1 and triadin–GFP in NIH 3T3 cells did not result in a significant modification of triadin mobility (mobile fraction 84.53±10.37%, n=40) (Figure 6A).

FRAP analysis on NIH 3T3 cells expressing triadin–GFP, triadin–GFP mutants, triadin–GFP plus calsequestrin-1 or triadin–GFP plus RyR1

Figure 6
FRAP analysis on NIH 3T3 cells expressing triadin–GFP, triadin–GFP mutants, triadin–GFP plus calsequestrin-1 or triadin–GFP plus RyR1

(A) FRAP analysis was performed on NIH 3T3 cells transfected with expression vectors for either triadin–GFP, or co-transfected with expression vectors for triadin–GFP and calsequestrin-1 or triadin–GFP and RyR1. Data are expressed as mean value of percentage of mobile fraction±S.D. The n values were as follows: triadin–GFP (n=53), triadin–GFP plus calsequestrin-1 (n=68) and triadin–GFP + RyR1 (n=40). The asterisks indicate statistical significance compared with the mobile fraction of triadin–GFP, as evaluated by Kruskal–Wallis and Dunn statistical test analysis (P<0.05). (B) FRAP analysis was performed on NIH 3T3 cells transfected with either triadin–GFP (n=53) or triadin–GFP plus calsequestrin-1 (n=68). (C) Triadin Δ215–729–GFP (n=27) or triadin Δ215–729–GFP plus calsequestrin-1 (n=28) or (D) triadin Δ106–729 (n=12) or triadin Δ106–729 plus calsequestrin-1 (n=18) expression vectors. Data are expressed as percentage of mobile fraction and each symbol in the diagrams indicates the mobile fraction value measured per single cell.

Figure 6
FRAP analysis on NIH 3T3 cells expressing triadin–GFP, triadin–GFP mutants, triadin–GFP plus calsequestrin-1 or triadin–GFP plus RyR1

(A) FRAP analysis was performed on NIH 3T3 cells transfected with expression vectors for either triadin–GFP, or co-transfected with expression vectors for triadin–GFP and calsequestrin-1 or triadin–GFP and RyR1. Data are expressed as mean value of percentage of mobile fraction±S.D. The n values were as follows: triadin–GFP (n=53), triadin–GFP plus calsequestrin-1 (n=68) and triadin–GFP + RyR1 (n=40). The asterisks indicate statistical significance compared with the mobile fraction of triadin–GFP, as evaluated by Kruskal–Wallis and Dunn statistical test analysis (P<0.05). (B) FRAP analysis was performed on NIH 3T3 cells transfected with either triadin–GFP (n=53) or triadin–GFP plus calsequestrin-1 (n=68). (C) Triadin Δ215–729–GFP (n=27) or triadin Δ215–729–GFP plus calsequestrin-1 (n=28) or (D) triadin Δ106–729 (n=12) or triadin Δ106–729 plus calsequestrin-1 (n=18) expression vectors. Data are expressed as percentage of mobile fraction and each symbol in the diagrams indicates the mobile fraction value measured per single cell.

Calsequestrin-1 and RyR1 contribute to a stable association, but not to the localization of triadin at the j-SR

As reported previously, targeting of triadin to the j-SR is not affected by knockout of either RyR1 or calsequestrin-1 [22,24]. However, as shown above, we observed that calsequestrin-1 can reduce the mobility of triadin in NIH 3T3 cells. We thus wanted to verify whether triadin mobility might be affected by endogenous calsequestrin-1 or RyR1 in myotubes. To this end, we prepared myotubes from RyR1 or calsequestrin-1 knockout mice and measured the mobile fraction of triadin–GFP proteins expressed in this cellular environment. Five independent transfection experiments with calsequestrin-1 knockout mice and three independent experiments with RyR1 knockout mice were performed. In myotubes lacking RyR1, FRAP experiments revealed a moderate increase in triadin–GFP protein mobility (mobile fraction 25.83±7.3%, n=18) compared with wild-type myotubes (14.58±8.03%, n=12, P<0.05). A higher increase in triadin mobile fraction was observed in myotubes prepared from calsequestrin-1 knockout mice (39.19±20.19%, n=69, P<0.05) as shown in Figure 7, thus indicating that calsequestrin-1 may play a major role in favouring the stable association of triadin to the j-SR.

FRAP analysis on 8-day differentiated myotubes from wild-type, calsequestrin-1 or RyR1 knockout mice expressing triadin–GFP

Figure 7
FRAP analysis on 8-day differentiated myotubes from wild-type, calsequestrin-1 or RyR1 knockout mice expressing triadin–GFP

FRAP analysis was performed on 8-day differentiated myotubes from wild-type, calsequestrin-1 knockout mice or RyR1 knockout mice transfected with triadin–GFP expression vectors. Data are expressed as mean value of percentage of mobile fraction±S.D. The n values were as follows: wild-type mice (n=17), calsequestrin-1 knockout mice (n=69) and RyR1 knockout mice (n=18). The asterisks indicate statistical significance compared with the mobile fraction of triadin–GFP expressed in wild-type mice, as evaluated by Kruskal–Wallis and Dunn statistical test analysis (P<0.05).

Figure 7
FRAP analysis on 8-day differentiated myotubes from wild-type, calsequestrin-1 or RyR1 knockout mice expressing triadin–GFP

FRAP analysis was performed on 8-day differentiated myotubes from wild-type, calsequestrin-1 knockout mice or RyR1 knockout mice transfected with triadin–GFP expression vectors. Data are expressed as mean value of percentage of mobile fraction±S.D. The n values were as follows: wild-type mice (n=17), calsequestrin-1 knockout mice (n=69) and RyR1 knockout mice (n=18). The asterisks indicate statistical significance compared with the mobile fraction of triadin–GFP expressed in wild-type mice, as evaluated by Kruskal–Wallis and Dunn statistical test analysis (P<0.05).

DISCUSSION

Regulation of Ca2+ release in striated muscles requires the for-mation of junctions between the SR and T-tubules and the selective targeting of excitation–contraction coupling proteins to these sites, where they assemble to form a multimolecular complex [3]. In the j-SR, this complex is composed of a number of proteins, which include RyR1, triadin and calsequestrin-1 [5]. Earlier analysis by us of protein mobility by FRAP experiments showed that the assembly of triadin and other j-SR proteins into clusters was accompanied by a progressive decrease in protein mobility [27]. This is consistent with biochemical evidence showing that mutual protein–protein interactions occur among j-SR proteins in differentiated muscle cells [5]. In the present paper, we report that three distinct regions in triadin (TR1, TR2 and TR3) contribute to triadin localization at the j-SR and that the presence of at least two of them is sufficient for protein targeting. Indeed, combined deletion of any two of these three regions led to protein mislocalization. Whether this is actually due to deletion of specific targeting sequences contained inside these regions or to protein misfolding, which indirectly affects protein localization, remains to be defined. In addition, FRAP analysis revealed that, whereas TR1 and TR2 had no effect on triadin mobility, only TR3 was required for a triadin stable association with the j-SR multiprotein complex.

Binding sites for proteins of the j-SR have been described within the amino acid sequences corresponding to TR1, TR2 and TR3. A cytoplasmic binding site for RyR has been identified in the N-terminal region between amino acid residues 18–47 in TR1 [43]. However, in our pull-down experiments, we were unable to detect any interaction between a GST fusion protein covering this region and RyR1. Consistent with this, others [44] have reported an absence of binding between this triadin sequence and RyR1 in blot overlay experiments. Nevertheless, a weak interaction whose detection may depend on the technique used cannot be excluded.

Binding domains for RyR1 and calsequestrin-1 between amino acid residues 200–232 in TR2 have been identified by several laboratories [3435,40,44,45], whereas low- and high-affinity binding sites for RyR1 were found in TR3 [44]. Intermolecular disulfide bonds have been proposed to occur in TR3 at cysteine residues located at positions 270 and 671. This could then mediate the formation of triadin oligomers [42,46,47]. Clusters of alternating positively and negatively charged amino acids enriched in lysine and glutamic acid residues, ‘KEKE motifs’, have been observed in TR2 and TR3. It has been proposed that these motifs promote interactions among proteins associated with the RyR channels at the j-SR [46,48]. In the present paper, we report that GST fusion proteins covering TR2 or TR3 were able to pull down calsequestrin-1, RyR1 and triadin from solubilized skeletal muscle proteins. However, when pull-down experiments were performed with solubilized recombinant proteins expressed in HEK-293 cells, it was evident that triadin could bind only to a GST fusion protein covering an extended TR2 domain (amino acid region 68–264), whereas no interaction was observed with RyR1. By contrast, recombinant calsequestrin-1 expressed in HEK-293 cells was able to bind to both TR2 and TR3 and also to restore binding of triadin to TR2 and TR3. This finding extends our previous knowledge by indicating that additional interaction sites with calsequestrin-1, other than the one identified in TR2 between amino acid residues 210–224 [34,35] are present in triadin. In vitro interaction and FRET experiments performed in both muscle and NIH 3T3 cells showed that binding of calsequestrin-1 to sites present in the intraluminal tail of triadin can favour triadin–triadin protein interactions. Interestingly, EM studies identified triadin as the component of periodically located anchors connecting calsequestrin to the j-SR membrane, supporting further the occurrence of close interactions between the two proteins in muscle cells [37]. Although it is widely accepted that protein–protein interactions can also contribute to protein targeting [29,49], this mechanism cannot explain triadin localization at the j-SR since knockout mice lacking either calsequestrin-1, RyR1 or DHPR did not show any change in triadin localization compared with wild-type muscles [1923,25,50]. Indeed, the results shown in the present study confirm that triadin can localize at the j-SR independently of calsequestrin-1. Nevertheless, results from FRAP studies in myotubes prepared from calsequestrin-1 knockout mice and in NIH 3T3 cells, clearly showed that calsequestrin-1 decreased triadin mobility, suggesting that calsequestrin-1 is necessary for mediating triadin association to the stable multiprotein complex present at the j-SR.

In summary, the data described in the present study provide the first identification of specific domains that contribute to the localization of an SR protein to the j-SR domain. All three triadin domains identified (TR1, TR2 and TR3) were able to contribute to the targeting of the protein to the j-SR, provided that at least two of them were present. Only TR3 was capable of affecting the mobility of triadin, most likely through interactions with calsequestrin-1, which appears to play a major role in the establishment of a stable interaction between triadin and the multiprotein complex associated with the RyR Ca2+ release channel at the j-SR.

Abbreviations

     
  • Cy3

    indocarbocyanine

  •  
  • DHPR

    dihydropyridine receptor

  •  
  • ER

    endoplasmic reticulum

  •  
  • FRETe

    FRET efficiency

  •  
  • HEK

    human embryonic kidney

  •  
  • j-SR

    junctional domain of the SR

  •  
  • αMEM

    α-minimum essential medium

  •  
  • NA

    numerical aperture

  •  
  • ROI

    region of interest

  •  
  • RyR

    ryanodine receptor

  •  
  • SERCA

    SR/ER Ca2+-ATPase

  •  
  • SR

    sarcoplasmic reticulum

  •  
  • TR

    targeting region

  •  
  • T-tubule

    transverse-tubule

AUTHOR CONTRIBUTION

Francesca Benini and Angela Maria Scarcella performed the localization studies. Daniela Rossi, Cristina Bencini and Marina Maritati performed the FRAP and FRET experiments. Cristina Bencini, Enrico Pierantozzi and Stefania Lorenzini performed the GST pull-down experiments. Cecilia Paolini and Feliciano Protasi contributed calsequestrin knockout mice. Daniela Rossi and Vincenzo Sorrentino designed and supervised the experiments and wrote the paper.

We thank Dr Paul D. Allen (Department of Molecular Biosciences, School of Veterinary Medicine, University of California, Davis, U.S.A.) for providing RyR1 knockout mice. We thank Professor Angiolo Benedetti (Department of Molecular and Developmental Medicine, University of Siena, Italy) and Professor Fyfe Bygrave (Department of Biomedical Science and Biochemistry, Australian National University, Camberra, Australia) for their critical reading of the paper before submission and helpful discussion.

FUNDING

This work was supported by Telethon Research Grant [grant number GGP13213 (to F.P. and V.S.)] and MIUR (Ministry of Education, Universities and Research) [grant number PRIN 2009 (to V.S. and C.P.)].

References

References
1
Franzini-Armstrong
C.
Simultaneous maturation of transverse tubules and sarcoplasmic reticulum during muscle differentiation in the mouse
Dev. Biol.
1991
, vol. 
146
 (pg. 
353
-
363
)
2
Zhao
X.
Yamazaki
D.
Kakizawa
S.
Pan
Z.
Takeshima
H.
Ma
J.
Molecular architecture of Ca2+ signaling control in muscle and heart cells
Channels
2011
, vol. 
5
 (pg. 
391
-
396
)
3
Dulhunty
A. F.
Excitation–contraction coupling from the 1950s into the new millennium
Clin. Exp. Pharmacol. Physiol.
2006
, vol. 
33
 (pg. 
763
-
772
)
4
Lanner
J. Y.
Georgiou
D. K.
Joshi
A. D.
Hamilton
S. H.
Ryanodine receptors: structure, expression, molecular details, and function in calcium release
Cold Spring Harb. Perspect. Biol.
2010
, vol. 
2
 (pg. 
1
-
21
)
5
Zhang
L.
Kelley
G.
Schmeisser
G.
Kobayashi
Y. M.
Jones
L. R.
Complex formation between junctin, triadin, calsequestrin and the ryanodine receptor. Proteins of the cardiac junctional sarcoplasmic reticulum membrane
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
23389
-
23397
)
6
Treves
S.
Vukcevic
M.
Maj
M.
Thurnheer
R.
Mosca
B.
Zorzato
F.
Minor sarcoplasmic reticulum membrane components that modulate excitation–contraction coupling in striated muscles
J. Physiol.
2009
, vol. 
13
 (pg. 
3071
-
3079
)
7
Takeshima
H
Komazaki
S.
Nishi
M.
Iino
M.
Kangawa
K.
Junctophilins: a novel family of junctional membrane complex proteins
Mol. Cell
2000
, vol. 
6
 (pg. 
11
-
22
)
8
Jones
L. R.
Zhang
L.
Sanborn
K.
Jorgensen
A. O.
Kelley
J.
Purification, primary structure and immunological characterization of the 26 kD calsequestrin binding protein (junctin) from cardiac junctional sarcoplasmic reticulum
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
30787
-
30796
)
9
Terentyev
D.
Viatchenko-Karpinski
S.
Vedamoorthyrao
S.
Oduru
S.
Gyorke
I.
Williams
S. C.
Gyorke
S.
Protein–protein interactions between triadin and calsequestrin are involved in modulation of sarcoplasmic reticulum calcium release in cardiac myocytes
J. Physiol.
2007
, vol. 
583
 (pg. 
71
-
80
)
10
Golini
L.
Chouabe
C.
Berthier
C.
Cusimano
V.
Fornaio
M.
Bonvallet
R.
Formoso
L.
Giacomello
E.
Jacquemond
V.
Sorrentino
V.
Junctophilin 1 and 2 interact with the L-type Ca2+ channel dihydropyridine receptors (DHPRs) in skeletal muscle
J. Biol. Chem.
2011
, vol. 
286
 (pg. 
43717
-
43725
)
11
Wium
E.
Dulhunty
A. F.
Beard
N. A.
A skeletal muscle ryanodine receptor interaction domain in triadin
PLoS ONE
2012
, vol. 
7
 pg. 
e43817
 
12
Flucher
B. E.
Andrews
S. B.
Daniels
M. P.
Molecular organization of transverse tubule/sarcoplasmic reticulum junctions during development of excitation–contraction coupling in skeletal muscle
Mol. Biol. Cell
1994
, vol. 
5
 (pg. 
1005
-
1118
)
13
Protasi
F.
Franzini-Armstrong
C.
Flucher
B. E.
Coordinated incorporation of skeletal muscle dihydropyridine receptors and ryanodine receptors in peripheral couplings of BC3H1 cells
J. Cell Biol.
1997
, vol. 
137
 (pg. 
859
-
870
)
14
Sprague
B. L.
McNally
J.
FRAP analysis of binding: proper and fitting
Trends Cell Biol.
2005
, vol. 
15
 (pg. 
84
-
91
)
15
Sorrentino
V.
Sarcoplasmic reticulum: structural determinants and protein dynamics
Int. J. Biochem. Cell Biol.
2011
, vol. 
48
 (pg. 
1075
-
1078
)
16
Ziolkowska
N. E.
Cristiano
R.
Walther
T. C.
Organized living: formation mechanisms and functions of plasma membrane domains in yeast
Trends Cell Biol.
2012
, vol. 
22
 (pg. 
151
-
158
)
17
Kusumi
A.
Suzuki
K. G. N.
Kasai
R. S
Ritchie
K.
Fujiwara
T. K.
Hierarchical mesoscale domain organization of the plasma membrane
Trends Biochem. Sci.
2011
, vol. 
36
 (pg. 
604
-
615
)
18
Kusumi
A
Fujiwara
T. K.
Moronea
N.
Yoshida
K. J.
Chadda
R.
Xiea
M.
Kasai
R. S.
Suzuki
K. G. N.
Membrane mechanisms for signal transduction: the coupling of the meso-scale raft domains to membrane-skeleton-induced compartments and dynamic protein complexes
Semin. Cell Dev. Biol.
2012
, vol. 
23
 (pg. 
126
-
144
)
19
Flucher
B. E.
Andrews
S. B.
Fleischer
S.
Marks
A. R.
Caswell
A.
Powell
J. A.
Triad formation: organization and function of the sarcoplasmic reticulum calcium release channel and triadin in normal and dysgenic muscle in vitro
J. Cell Biol.
1993
, vol. 
123
 (pg. 
1161
-
1174
)
20
Powell
J. A.
Petherbridge
L.
Flucher
B. E.
Formation of triads without the dihydropyridine receptor alpha subunits in cell lines from dysgenic skeletal muscle
J. Cell Biol.
1996
, vol. 
134
 (pg. 
375
-
387
)
21
Ikemoto
T.
Komazaki
S.
Takeshima
H.
Nishi
M.
Noda
T.
Iino
M.
Endo
M.
Functional and morphological features of skeletal muscle from mutant mice lacking both type 1 and type 3 ryanodine receptors
J. Physiol.
1997
, vol. 
501
 (pg. 
305
-
312
)
22
Felder
E.
Protasi
F.
Hirsch
R.
Franzini-Armstrong
C.
Allen
P. D.
Morphology and molecular composition of sarcoplasmic reticulum surface junctions in the absence of DHPR and RyR in mouse skeletal muscle
Biophys. J.
2002
, vol. 
82
 (pg. 
3144
-
3149
)
23
Knollmann
B. C.
Chopra
N.
Hlaing
T.
Akin
B.
Yang
T.
Ettensohn
K.
Knollmann
B. E. C.
Horton
K. D.
Weissman
N. J.
Holinstat
I.
, et al. 
Casq2 deletion causes sarcoplasmic reticulum volume increase, premature Ca2+ release, and catecholaminergic polymorphic ventricular tachycardia
J. Clin. Invest.
2006
, vol. 
116
 (pg. 
2510
-
2520
)
24
Shen
X.
Franzini-Armstrong
C.
Lopez
J. R.
Jones
L. R.
Kobayashi
Y. M.
Wang
Y.
Kerrick
W. G.
Caswell
A. H.
Potter
J. D.
Miller
T.
, et al. 
Triadins modulate intracellular Ca2+ homeostasis but are not essential for excitation-contraction coupling in skeletal muscle
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
37864
-
37874
)
25
Paolini
C.
Quarta
M.
Nori
A.
Boncompagni
S.
Canato
M.
Volpe
P.
Allen
P. D.
Reggiani
C.
Protasi
F.
Reorganized stores and impaired calcium handling in skeletal muscle of mice lacking calsequestrin-1
J. Physiol.
2007
, vol. 
583
 (pg. 
767
-
784
)
26
Oddoux
S.
Brocard
J.
Schweitzer
A.
Szentesi
P.
Giannesini
B.
Brocard
J.
Fauré
J.
Pernet-Gallay
K.
Bendahan
D.
Lunardi
J.
, et al. 
Triadin deletion induces impaired skeletal muscle function
J. Biol. Chem.
2009
, vol. 
284
 (pg. 
34918
-
34929
)
27
Cusimano
V.
Pampinella
F.
Giacomello
E.
Sorrentino
V.
Assembly and dynamics of proteins of the longitudinal and junctional sarcoplasmic reticulum in skeletal muscle cells
Proc. Natl. Acad. Sci. U.S.A.
2009
, vol. 
106
 (pg. 
4695
-
4700
)
28
Rossi
D.
Barone
V.
Giacomello
E.
Cusimano
V.
Sorrentino
V.
The sarcoplasmic reticulum: an organized patchwork of specialized domains
Traffic
2008
, vol. 
9
 (pg. 
1044
-
1049
)
29
Owen
D. M.
Williamson
D.
Rentero
C.
Gaus
K.
Quantitative microscopy: protein dynamics and membrane organisation
Traffic
2009
, vol. 
10
 (pg. 
962
-
971
)
30
Kobayashi
Y. M.
Jones
L. R.
Identification of triadin 1 as the predominant triadin isoform expressed in mammalian myocardium
J. Biol. Chem.
1999
, vol. 
274
 (pg. 
28660
-
28668
)
31
Thevenon
D.
Smida-Rezgui
S.
Chevessier
F.
Groh
S.
Henry-Berger
J.
Romero
N. B.
Villaz
M.
DeWaard
M.
Marty
I.
Human skeletal muscle triadin: gene organization and clonino of the major isoform, Trisk 51
Biochem. Biophys. Res. Commun.
2003
, vol. 
303
 (pg. 
669
-
675
)
32
Marty
I.
Thevenon
D.
Scotto
C.
Groh
S.
Sainnier
S.
Robert
M.
Grunwald
D.
Villaz
M.
Cloning and characterization of a new isoform of skeletal muscle triadin
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
8206
-
8212
)
33
Perez
C. F.
On the footsteps of triadin and its role in skeletal muscle
World J. Biol. Chem.
2011
, vol. 
26
 (pg. 
177
-
183
)
34
Guo
W.
Jorgensen
A. O.
Jones
L. R.
Campbell
K. P.
Biochemical characterization and molecular cloning of cardiac triadin
J. Biol. Chem.
1996
, vol. 
271
 (pg. 
458
-
465
)
35
Kobayashi
Y. M.
Alseikhan
B. A.
Jones
L. R.
Localization and characterization of the calsequestrin binding domain of triadin-1
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
17639
-
17646
)
36
Eltit
J. M.
Szpyt
J.
Li
H.
Allen
P. D.
Perez
C. F.
Reduced gain of excitation–contraction coupling in triadin-null myotubes is mediated by the disruption of FKBP12/RyR1 interaction
Cell Calcium
2011
, vol. 
49
 (pg. 
128
-
135
)
37
Boncompagni
S.
Thomas
M.
Lopez
J. R.
Allen
P. D
Yuan
Q.
Kranias
E. G.
Franzini Armstrong
C.
Perez
C. F.
Triadin/junctin double null mouse reveals a differential role for triadin and junctin in anchoring CASQ to the jSR and regulating Ca2+ homeostasis
PLoS ONE
2012
, vol. 
7
 pg. 
e39962
 
38
Conti
A.
Gorza
L.
Sorrentino
V.
Differential distribution of ryanodine receptor type 3 (RyR3) gene product in mammalian skeletal muscles
Biochem. J.
1996
, vol. 
316
 (pg. 
19
-
23
)
39
Lee
H. G.
Kang
H.
Kim
D. H.
Park
W. J.
Interaction of HRC (histidine-rich Ca2+-binding protein) and triadin in the lumen of sarcoplasmic reticulum
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
39533
-
39538
)
40
Lee
J. M.
Rho
S. H.
Shin
D. W.
Cho
C.
Park
W. J.
Eom
S. H.
Ma
J.
Kim
D. H.
Negatively charged amino acids within the intraluminal loop of ryanodine receptor are involved in the interaction with triadin
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
6994
-
7000
)
41
Rossi
D.
Simeoni
I.
Micheli
M.
Bootman
M.
Lipp
P.
Allen
P. D.
Sorrentino
V.
RyR1 and RyR3 isoforms provide distinct intracellular Ca2+ signals in HEK 293 cells
J. Cell Sci.
2002
, vol. 
115
 (pg. 
2497
-
2504
)
42
Fan
H.
Brandt
N. R.
Caswell
A. H.
Disulfide bonds, N-glycosylation and transmembrane topology of skeletal muscle triadin
Biochemistry
1995
, vol. 
34
 (pg. 
14902
-
14908
)
43
Groh
S.
Marty
I.
Ottolia
M.
Prestipino
G.
Chapel
A.
Villaz
M.
Ronjat
M.
Functional ineraction of the cytoplasmic domain of tridin with the skeletal ryanodine receptor
J. Biol. Chem.
1999
, vol. 
274
 (pg. 
12278
-
12283
)
44
Caswell
A. H.
Motoike
H. K.
Fan
H.
Brandt
N. R.
Location of ryanodine receptor binding site on skeletal muscle triadin
Biochemistry
1999
, vol. 
38
 (pg. 
90
-
97
)
45
Shin
D. W.
Ma
J.
Kim
D. H.
The Asp-rich region at the carboxy-terminus of calsequestrin binds to Ca2+ and interacts with triadin
FEBS Lett.
2000
, vol. 
486
 (pg. 
178
-
182
)
46
Knudson
C. M.
Stang
K. K.
Moomaw
C. R.
Slaughter
C. A.
Campbell
K. P.
Primary structure and topological analysis of a skeletal muscle specific junctional sarcoplasmic reticulum glycoprotein (triadin)
J. Biol. Chem.
1993
, vol. 
268
 (pg. 
12646
-
12654
)
47
Fourest-Lieuvin
A.
Rendu
J.
Osseni
A.
Pernet-Gallay
K.
Rossi
D.
Oddoux
S.
Brocard
J.
Sorrentino
V.
Marty
I.
Fauré
J.
Role of triadin in the organization of reticulum membrane at the muscle triad
J. Cell Sci.
2012
, vol. 
125
 (pg. 
3443
-
3453
)
48
Realini
C.
Rogers
S. W.
Rechsteiner
M.
KEKE motifs. Proposed roles in protein-protein association and presentation of peptides by MHC class I receptors
FEBS Lett.
1994
, vol. 
348
 (pg. 
109
-
113
)
49
Lippincott-Schwartz
J.
Snapp
E.
Kenworthy
A.
Studying protein dynamics in living cells
Nat. Rev.
2001
, vol. 
2
 (pg. 
444
-
4456
)
50
Takekura
H.
Franzini-Armstrong
C.
Correct targeting of dihydropyridine receptors and triadin in dyspedic mouse skeletal muscle in vivo
Dev. Dynam.
1999
, vol. 
214
 (pg. 
372
-
380
)

Author notes

1

These authors contributed equally to this work.

Supplementary data