If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
To whom correspondence should be addressed: Laboratory of Cell Signaling, Aging Research Center, Korea Research Institute of Bioscience and Biotechnology, 125 Gwahangno, Yusong, Daejeon 305-806, Korea. Tel.: 82-42-860-4143; Fax: 82-42-879-8596;
* This work was supported by grants from the Ministry of Education, Science and Technology (Grants 2011-0002141, 2011-0030134, and 2011-0027762) (to K. S. K.) and a grant from the Systems Biology Infrastructure Gwangju Institute of Science and Technology (to S. H. E., H. J., and D. H. K.). This article contains supplemental Movies S1 and S2 and Figs. S1 and S2. 1 Both authors contributed equally to this work.
Calsequestrin (CSQ), the major intrasarcoplasmic reticulum calcium storage protein, undergoes dynamic polymerization and depolymerization in a Ca2+-dependent manner. However, no direct evidence of CSQ depolymerization in vivo with physiological relevance has been obtained. In the present study, live cell imaging analysis facilitated characterization of the in vivo dynamics of the macromolecular CSQ structure. CSQ2 appeared as speckles in the presence of normal sarcoplasmic reticulum (SR) Ca2+ that were decondensed upon Ca2+ depletion. Moreover, CSQ2 decondensation occurred only in the stoichiometric presence of junctin (JNT). When expressed alone, CSQ2 speckles remained unchanged, even after Ca2+ depletion. FRET analysis revealed constant interactions between CSQ2 and JNT, regardless of the SR Ca2+ concentration, implying that JNT is an essential component of the CSQ scaffold. In vitro solubility assay, electron microscopy, and atomic force microscopy studies using purified recombinant proteins confirmed Ca2+ and JNT-dependent disassembly of the CSQ2 polymer. Accordingly, we conclude that reversible polymerization and depolymerization of CSQ are critical in SR Ca2+ homeostasis.
is a major Ca2+ storage reservoir in muscle. The SR Ca2+ release channel, ryanodine receptor (RyR), forms a large macromolecular complex with transmembrane, luminal, and cytoplasmic components (
). RyR recruits a low affinity (Kd = 1 mm), high capacity (60–80 mol Ca2+ mol−1) Ca2+-binding protein, calsequestrin (CSQ), to sense the environment within SR (
Complex formation between junctin, triadin, calsequestrin, and the ryanodine receptor. Proteins of the cardiac junctional sarcoplasmic reticulum membrane.
Calsequestrin binds to monomeric and complexed forms of key calcium-handling proteins in native sarcoplasmic reticulum membranes from rabbit skeletal muscle.
). The N-terminal regions of these proteins are located in the cytoplasm and the C-terminal regions in the lumen of the SR. The luminal domain of these proteins containing the KEKE repeat is proposed to directly interact with CSQ (
Complex formation between junctin, triadin, calsequestrin, and the ryanodine receptor. Proteins of the cardiac junctional sarcoplasmic reticulum membrane.
). Thus, CSQ, TRN, and JNT play an active role in relaying information on the Ca2+ load in the SR to RyR so that resting and voltage-activated Ca2+ release can be modified in response to changes in the store load through physical coupling between these proteins (
). However, the individual steps constituting this system remain to be established.
Catecholamine-induced polymorphic ventricular tachycardia (CPVT) is a familial arrhythmogenic disorder caused by the aberrant release of Ca2+ from the SR. CPVT is linked to a defective channel, either RyR2 itself or other regulatory components of the channel gating machinery. Mutations in RyR2 and CSQ2 are linked to CPVT. In total, seven CSQ2 mutations have been identified in association with CPVT (
A missense mutation in a highly conserved region of CASQ2 is associated with autosomal recessive catecholamine-induced polymorphic ventricular tachycardia in Bedouin families from Israel.
Human CSQ2, containing 391 amino acids plus a 19-residue N-terminal signal peptide, is highly acidic due to enrichment of acidic residues (60%) within the 63 carboxyl-terminal amino acids. Although human CSQ1 and CSQ2 are the products of different genes, the proteins display 65% sequence identity, are acidic, and share one glycosylation site. However, CSQ2 contains a 31-amino acid extension at its carboxyl terminus (residues 361–391) comprising 71% acidic residues, which significantly contributes to its ability to bind large amounts of Ca2+ (
), as well as a second glycosylation site. CSQ exists either as a monomer or as a wide range of polymers, depending on the Ca2+ concentration. The extended N terminus of CSQ establishes a front-to-front dimer interface through arm exchange. The C terminus of CSQ, the most negative region, is critical in forming tetrameric and higher-order linear polymers and capturing Ca2+ in the back-to-back interface (
Comparing skeletal and cardiac calsequestrin structures and their calcium binding: a proposed mechanism for coupled calcium binding and protein polymerization.
) demonstrated that knockdown of CSQ2, but not CSQ1, led to reduced Ca2+ storage and release in C2C12 skeletal muscle myotubes. They also found a significant reduction in both the sarco/endoplasmic reticulum Ca2+ ATPase (SERCA) and the RyR1 proteins in the CSQ2 knockdown myotubes. Thus, we chose to evaluate CSQ2 instead of CSQ1 for the experiments with C2C12 cells.
Although in vitro studies have reported reversible CSQ polymerization and depolymerization, no direct in vivo evidence has been obtained to date. Earlier electron microscopy analyses have revealed a fibrous array assumed to be Ca2+-bound CSQ in the SR at the junctional membrane (
). Cross-linking studies have also demonstrated that the majority of CSQ in the SR forms CSQ-CSQ complexes, supporting the physiological relevance of these polymers (
). Here, we transiently overexpressed a fluorescent protein-tagged CSQ2, along with JNT, and examined the in vivo dynamics of the macromolecular structure of the CSQ complex using live cell imaging. Our data provide in vivo evidence that the occurrence of CSQ2 depolymerization is synchronous with Ca2+ depletion and that this conformational change requires coexpression of JNT.
EXPERIMENTAL PROCEDURES
Cell Culture and Transfection
C2C12, a murine skeletal muscle cell line, was grown in Dulbecco's modified Eagle's medium (DMEM; Invitrogen). Media were supplemented with heat-inactivated 10% fetal bovine serum (Invitrogen), non-essential amino acids, 50 units/ml penicillin, and 50 units/ml streptomycin (Invitrogen). Cells were grown at 37 °C and 5% CO2 and transfected with Lipofectamine reagent (Invitrogen). The amounts of expression plasmids used are specified in the figure legends, and equivalent amounts of plasmid DNA were maintained with a relevant empty vector.
Plasmids and Lentiviral Infection
Human CSQ2 cDNA with a C-terminal yellow fluorescent protein (YFP) fusion was inserted into pLenti M1.4-MCMV, and that containing a C-terminal hemagglutinin (HA) tag was inserted into pcDNA3 vectors for the expression of CSQ2-YFP fusion and CSQ2-HA proteins in mammalian cells, respectively. Mouse JNT cDNA with a C-terminal cyan fluorescent protein (CFP) sequence was inserted into pLenti M1.4-MCMV for the expression of JNT-CFP fusion in mammalian cells. VSV-G-pseudotyped lentivirus was produced in HEK293T cells cotransfected with pLenti M1.4 construct, pLp1, pLp2, and pVSV-G. Culture medium was replaced with serum-free DMEM at 12 h after transfection. Supernatant fractions containing lentivirus particles were collected 48 h later, filtered through a 0.45-μm filter unit, and used to infect C2C12 cells at 60% confluence in the presence of 4 μg/ml of Polybrene or stored at −70 °C.
Immunoprecipitation and Immunoblotting
Cells were harvested, washed with PBS, and lysed in lysis buffer (20 mm HEPES, pH 7.2, 50 mm NaCl, 10% glycerol, 0.5% Triton X-100, 100 μm phenylmethanesulfonyl fluoride, 1 μg/ml each of leupeptin and aprotinin). Lysates were centrifuged and incubated with anti-GFP antibody preincubated with protein A/G-Sepharose (GE Healthcare) or anti-HA-agarose beads (Sigma) for 12 h at 4 °C. Immobilized proteins were collected by centrifugation, washed three times with lysis buffer, and solubilized by boiling for 3 min in SDS-polyacrylamide gel electrophoresis sample buffer. Proteins were separated by electrophoresis on 10–12% SDS-polyacrylamide gels and transferred to nitrocellulose membranes. After blocking with 5% skimmed milk in Tris-buffered saline (20 mm Tris-HCl, pH 7.4, 150 mm NaCl) containing 0.05% Tween 20, membranes were probed with anti-GFP (Santa Cruz Biotechnology), anti-CSQ (Affinity BioReagents), and anti-HA (Sigma) antibodies. Blots were washed with Tris-buffered saline containing Tween 20, incubated with peroxidase-conjugated anti-rabbit IgG (Sigma) antibody, and developed using the chemiluminescence system (Pierce).
Imaging
Imaging was carried out using an inverted confocal microscope (LSM 510 META and LIVE 5, Carl Zeiss) with a 40× objective. To monitor cytosolic Ca2+ levels, cells were loaded with 2 μm fluo-4-acetoxymethylester (Fluo-4-AM) (Molecular Probes) in physiological salt solution (150 mm NaCl, 4 mm KCl, 2 mm CaCl2, 1 mm MgCl2, 5 mm glucose, 5 mm HEPES) for 30 min at 37 °C. Fluo-4 was excited with a 488-nm laser line and fluorescence was acquired at wavelengths of 505–530 nm. Cells were treated with thapsigargin (Calbiochem), 2,5-di-(tert-butyl)-1,4-benzohydroquinone (Sigma), cyclopiazonic acid (Sigma), or isoproterenol (Sigma) for intracellular Ca2+ perturbation. Fluorescence resonance energy transfer (FRET) was measured with the acceptor photobleaching method using CFP and YFP fusion proteins as donor and acceptor, respectively. This method estimates the release of donor quenching after acceptor photobleaching, thus providing a measurement of close colocalization between the fusion proteins. CFP and YFP were excited with the 405- and 514-nm lines of an argon laser, respectively. Emitted fluorescence was collected at 420–480 nm for CFP and 530–600 nm for YFP. Repeated scans with unattenuated 514 nm illumination were used to photobleach YFP. FRET efficiency was estimated using the following equation: (FCFPPostbleach − FCFPPrebleach)/FCFPPostbleach × 100%.
Recombinant CSQ2 and JNT Protein Purification
Human cardiac CSQ2 (NP_001223) (amino acids 20–399) and mouse JNT (AAG40812) without the transmembrane domain (amino acids 68–207) were cloned into pET-28b vector (Novagen) with an N-terminal His6 tag for affinity purification. For amplification of recombinant CSQ2 and JNT sequences, the following synthetic oligonucleotide primers were used: forward, 5′-GTA GAA CGC AGA ACT CTC ATA TGA AGA GAA CTC AC-3′; and reverse, 5′-CCC CCC CTC GAG GTC GAC CTC GAGCTA TTC ATC ATC ATC GTC ATC-3′ for CSQ2; forward, 5′-GTG GCT GTC CAT ATG TTT GAC TTG GTC GAT TAT GAA G-3′, and reverse, 5′-GCC CTT GCT CAC CTC GAGTCA GTT CTT TCT CTT C-3′ for JNT. The forward primer introduced a restriction enzyme site for NdeI (underlined), whereas the reverse primer introduced sites for XhoI (underlined) and the stop codon (bold). PCR was performed using Pfu DNA polymerase (SolGent) with the following cycling profile: initial denaturation at 95 °C for 2 min, 30 cycles of denaturation at 95 °C for 1 min, annealing at 50 °C for 1 min, elongation at 72 °C for 1 min, and final extension at 72 °C for 10 min. PCR products digested with NdeI and XhoI were inserted into the vector using the corresponding restriction sites. The resulting constructs were verified by DNA sequencing. Prior to the cloning procedure, the first mutation was generated within CSQ2 cDNA to eliminate an undesirable NdeI site by converting CATATG to CTTATG, using the QuikChange II technique (Stratagene). Cells were grown at 37 °C to an optical density (A600) of ∼0.6, and protein was induced with 0.2 mm isopropyl β-d-thiogalactopyranoside for 3 h. Next, cells were suspended in 20 mm Tris-Cl (pH 8.0), 500 mm NaCl, and 0.5% Triton X-100 and lysed with sonication. Proteins were subjected to affinity purification with Ni2+ affinity chromatography, and the His6 tag was removed by incubation with thrombin for 12 h at 4 °C. Proteins were subsequently purified using a gel filtration column (HiLoad 16/60, Amersham Biosciences) in final buffer (20 mm HEPES, 150 mm NaCl, 2 mm DTT, pH 7.5) with the FPLC system (Amersham Biosciences).
Atomic Force Microscopy (AFM)
Purified CSQ2 alone and the CSQ2-JNT mixture (10 μm each) were dialyzed against polymerization-inducing buffer (5 mm CaCl2, 150 mm NaCl, 2 mm DTT, 20 mm HEPES, pH 7.5) for 5 h at room temperature. An aliquot of each sample was deposited on film, dried under vacuum, and washed with distilled water. AFM measurements were performed with a Dimension 3100 scanning probe microscope (Digital Instruments Veeco) in tapping mode. Silicon nitride cantilevers were used as square pyramidal tips with a scan rate of 1 Hz for a 500 × 500-nm2 scale image. The resonance frequency was 320 kHz, and the force constant was 20 newtons/m. Raw AFM data were processed using a flattening/low pass fitting process.
Electron Microscopy (EM)
Purified CSQ2 and CSQ2-JNT mixtures (10 μm each) were incubated in polymerization-inducing buffer with or without 5 mm Ca2+, and the supernatant fractions were isolated by centrifugation. Aliquots (5 μl) of the reactions were applied to glow-discharged carbon-coated grids and negatively stained with 2% (w/v) uranyl acetate solution. Micrographs were recorded with a CCD camera (2k, Gatan) using a Tecnai F20 field emission gun electron microscope operated at 200 kV (FEI).
RESULTS
SR Ca2+ Depletion Induces Decondensation of CSQ2 Speckles
To investigate the in vivo dynamics of CSQ2 oligomerization in response to perturbation of the Ca2+ concentration, CSQ2 was ectopically expressed with a fluorescent protein tag in C2C12 myoblasts, and the real-time fluorescence image was examined with confocal microscopy. As CSQ interacts with JNT or TRN (
Complex formation between junctin, triadin, calsequestrin, and the ryanodine receptor. Proteins of the cardiac junctional sarcoplasmic reticulum membrane.
) and undergoes assembly into the RyR complex, JNT was cotransfected with CSQ. CSQ appeared as speckles in myoblasts, consistent with a previous finding (
Overexpression of calsequestrin in L6 myoblasts: formation of endoplasmic reticulum subdomains and their evolution into discrete vacuoles where aggregates of the protein are specifically accumulated.
). Ca2+ perturbation was performed by treating cells with thapsigargin, which inhibits SERCA, resulting in SR Ca2+ depletion within a few minutes and a corresponding transient Ca2+ increase in the cytoplasm (
). Upon coexpression with JNT, CSQ speckles became diffuse in live cells after thapsigargin treatment (Fig. 1A, lower panel; supplemental Movies S1 and S2). Sequential decondensation progression was additionally evident in the high-resolution images of cells fixed at intervals after thapsigargin treatment (Fig. 1B). Speckle shapes began to change at 5 min after thapsigargin treatment and became fully diffuse in 30 min (Fig. 1B). However, upon expression of CSQ only, speckles remained unchanged following thapsigargin treatment (Fig. 1A, upper panel). This result suggests that JNT assists in the decondensation of CSQ speckles that occurs upon Ca2+ depletion. We examined SR Ca2+ release by monitoring fluorescence of Fluo-4 placed in the cytoplasm after thapsigargin treatment. Cells coexpressing CSQ and JNT showed similar patterns of Ca2+ transients upon thapsigargin treatment as untransfected control cells (supplemental Fig. S1), implying Ca2+ release from SR to the cytoplasm. We speculated that Ca2+ release from CSQ induced by thapsigargin promotes changes in the speckle shape. To evaluate the state of CSQ2, we performed native gel electrophoresis of cell lysates expressing CSQ2 (supplemental Fig. S2). As expected, CSQ2 was present in the form of a variety of polymers in a normal condition and was also found in the depolymerized form, which revealed a high content of monomeric CSQ2 following thapsigargin treatment. This finding is consistent with previous in vitro experiments showing that lower environmental Ca2+ concentrations trigger release of Ca2+ from the CSQ polymer, with concomitant disruption of its polymeric structure (
), also affect decondensation of CSQ speckles (Fig. 2, B and C). However, caffeine, an RyR channel opener, did not alter the CSQ speckle shape. We suggest that the transient Ca2+ decrease caused by caffeine treatment is not sufficient (in regard to concentration or duration) to induce major conformational changes in the CSQ polymer but that the irreversible SR Ca2+ depletion as a result of treatment with SERCA inhibitors is enough to do so. As JNT is necessary for this process, we examined the stoichiometry of molar interactions for CSQ speckle decondensation. The molar ratio of JNT to CSQ was important in the thapsigargin response. Specifically, when the JNT:CSQ molar ratio approached 1 (Fig. 3), speckle decondensation progressed clearly. However, at significantly lower concentrations of JNT than CSQ (for example, 1:3), speckle decondensation was marginal (Fig. 3). This result is in keeping with previous in vitro experiments showing that loss of interactions with JNT (or TRN) induces compact folding of CSQ at high Ca2+ concentrations (
Complex formation between junctin, triadin, calsequestrin, and the ryanodine receptor. Proteins of the cardiac junctional sarcoplasmic reticulum membrane.
). At higher JNT levels than CSQ (for example, 3:1), inefficient speckle decondensation was observed (Fig. 3). Accordingly, we suggest that stoichiometric interactions of CSQ and JNT are necessary for Ca2+-dependent CSQ depolymerization.
FIGURE 1JNT promotes decondensation of CSQ2 speckles upon SR Ca2+ depletion.A, time-lapse images of CSQ2 decondensation in the presence of JNT. C2C12 cells expressing YFP-tagged CSQ2 (CSQ2-YFP) with or without CFP-tagged JNT (JNT-CFP) were treated with 5 μm thapsigargin (TG), and YFP fluorescence was traced in live cells under live confocal microscopy with incubation for 30 min. B, high-resolution images of CSQ2 speckle decondensation in cells coexpressing CSQ and JNT. Upper panels represent magnified photographs of the boxes in the lower panels. Cells cultured on coverslips were treated with thapsigargin, fixed at 5-min intervals, and observed under a confocal microscope. C, C2C12 cells were transfected with JNT-CFP and CSQ2-YFP or GFP control vector. At 40 h after transfection, cell lysates were prepared and analyzed by immunoblotting (I/B) with anti-GFP antibodies. n.s. denotes nonspecific.
FIGURE 2Effects of Ca2+ perturbation on decondensation of CSQ2 speckles. C2C12 myoblasts transfected with CSQ2-YFP and JNT-CFP were treated with SERCA inhibitors, such as thapsigargin (TG, 5 μm), cyclopiazonic acid (CPA, 10 μm), and 2,5-di-(tert-butyl)-1,4-benzohydroquinone (tBHQ, 100 μm), or a β-adrenergic receptor agonist isoproterenol (ISO, 10 μm). Cells were preincubated with 2 μm Fluo-4 AM for 30 min and washed with 1× physiological salt solution prior to drug treatment for 30 min. CSQ2 decondensation and cytoplasmic Ca2+ levels were monitored in cells coexpressing CSQ2-YFP and JNT-CFP using live confocal microscopy.
FIGURE 3The molar ratios of JNT to CSQ is important for CSQ decondensation.A, decondensation analysis of CSQ2 speckles in C2C12 cells expressing different concentrations of CSQ2-YFP and JNT-CFP. SR Ca2+ depletion was induced by thapsigargin (TG) treatment. B, expression levels of CSQ2-YFP and JNT-CFP were confirmed by immunoblotting (I/B) analysis with the anti-GFP antibody. The relative concentrations of proteins are indicated with approximate values.
To examine the interactions between exogenously expressed CSQ and JNT, C2C12 cells were cotransfected with HA-tagged CSQ2 and CFP-tagged JNT expression plasmids, and coimmunoprecipitation assays were performed. CSQ2 was detected in immunoblots of JNT-CFP immunoprecipitates with the anti-GFP antibody (Fig. 4A). In reverse order, JNT was probed in CSQ2-HA immunoprecipitates with the anti-HA antibody (Fig. 4B). Thapsigargin treatment did not influence this interaction significantly (Fig. 4B, lane 3 versus lane 4), suggesting that the levels of CSQ-JNT interactions are maintained, even after disruption of CSQ2 speckles at lower Ca2+ concentrations. To determine whether CSQ and JNT proteins colocalize in intact cells, we utilized FRET technology. CSQ2 and JNT were expressed as N-terminal fusions with YFP and CFP, respectively. In an acceptor photobleaching experiment, YFP and CFP fluorescence was monitored before and after YFP was bleached in a region of interest. If CFP and YFP are in close proximity, donor (CFP) fluorescence should increase in the region where the acceptor (YFP) is bleached. JNT-CFP was distributed throughout the cytoplasm, whereas CSQ2-YFP was localized in part of the cytoplasm as speckles (Fig. 4C, upper panels). After bleaching of the acceptor fluorophore (CSQ2-YFP), which led to almost complete removal of fluorescence intensity (Fig. 4C, lower right panel), increased fluorescence of donor molecules (JNT-CFP) was detected in the region of acceptor bleaching (Fig. 4C, lower left panel, yellow to red pseudocolor). Quantitative analysis revealed that CSQ2-YFP lost its yellow fluorescence by ∼80% after bleaching, whereas JNT-CFP gained cyan fluorescence by ∼40% (Fig. 4C). This result indicates that the two proteins are in close proximity to each other near the speckle positions. To determine whether the interactions between CSQ and JNT proteins remain after decondensation of CSQ speckles, FRET was performed after thapsigargin treatment in cells coexpressing CSQ2-YFP and JNT-CFP. Consistent with immunoprecipitation data, thapsigargin hardly influenced FRET outcome (Fig. 4D), suggestive of constant interactions between CSQ and JNT, irrespective of the Ca2+ and polymerization states.
FIGURE 4CSQ2 interacts with JNT.A, 293T cells were cotransfected with HA-tagged CSQ2 (CSQ2-HA), together with JNT-CFP or control vector. After 40 h, cells were lysed, and JNT was immunoprecipitated (IP) with an anti-GFP antibody. CSQ2 was detected by immunoblotting (I/B) with anti-HA antibody. B, 293T cells were cotransfected with CSQ2-HA and JNT-CFP or control vector. After 40 h, cells were treated with or without thapsigargin (TG) (5 μm) for 30 min prior to lysis, and CSQ2 was immunoprecipitated with an anti-HA antibody. JNT was detected by immunoblotting with an anti-GFP antibody. C, FRET analysis to detect interactions between CSQ and JNT. Representative confocal images of CFP and YFP fluorescence were acquired from myoblasts cotransfected with JNT-CFP and CSQ2-YFP fusion constructs before (upper panels) and after (lower panels) photobleaching of YFP. Images were pseudocolored for easy visualization of fluorescence intensity. The bleached regions of interest are indicated with white lines. Fluorescent intensities (●, CFP; ▴, YFP) were measured within a bleached zone at three time points before and after bleaching. Representative plots were drawn with data from the rightmost zone. D, FRET analysis was additionally performed under similar conditions after thapsigargin treatment for 30 min. n.s. denotes nonspecific.
Macromolecular Structure of Polymerized CSQ2 in Vitro
Data from the present study showed that CSQ2 speckle decondensation occurs in a Ca2+- and JNT-dependent manner in skeletal muscle cells. We examined this issue in vitro using purified proteins. First, we tested the solubility of polymers generated with different concentrations of CSQ2 and JNT. Purified CSQ2 was combined with JNT at different molar ratios or left unmixed. Polymerization proceeded in buffer containing 5 mm Ca2+, as described in the previous study (
). After centrifugation, precipitates and supernatants were separated. Supernatants (Fig. 5A, Soluble I) were further prepared for negative staining electron microscopic analysis (see below). To establish whether the precipitation (polymerization) reaction is reversible, precipitates (Fig. 5A, Pellet I) were incubated in buffer without Ca2+. Resolubilized fractions (Soluble II) were separated from insoluble precipitates (Pellet II) by centrifugation, and the protein concentrations were measured (Fig. 5B). In vitro Ca2+-free conditions may mimic the in vivo conditions of SR Ca2+ depletion (FIGURE 1, FIGURE 2) that induce solubilization (decondensation) of CSQ speckles. Polymers obtained from equimolar mixtures of CSQ and JNT exhibited highest solubility with the lowest residual precipitates after Ca2+ removal (Fig. 5B, fourth bar, compare gray with dark parts) among those obtained from mixtures of various molar ratios of CSQ and JNT. The solubilization efficiency gradually decreased as the ratio diverged from equimolar concentrations. A low proportion of polymers generated from CSQ only was converted to the soluble form (<30%), whereas those from the 1:1 CSQ-JNT mixture were susceptible to solubilization (up to 90%). This result is in agreement with our in vivo data showing that CSQ-JNT stoichiometry is critical for speckle decondensation. Second, residual precipitates (Pellet II) of the CSQ polymers in the presence or absence of JNT were observed under AFM performed to analyze macromolecular structures. In the absence of JNT, significant amounts of large CSQ aggregates (heights of up to 150 nm) were observed in the AFM image (Fig. 6A). In contrast, in the presence of JNT, smaller, evenly sized aggregates appeared (Fig. 6B). This finding is compatible with data from the solubility test (Fig. 5B), indicating that JNT mediates solubilization (destabilization) of the CSQ polymer, both in vitro and in vivo.
FIGURE 5JNT promotes depolymerization of Ca2+-induced CSQ2 polymer.A, scheme of fractionation experiments for analysis of CSQ2 polymerization and depolymerization. CSQ2 (20 μm) and JNT (0–100 μm) were mixed at different molar ratios, incubated in polymerization buffer (5 mm CaCl2, 50 mm KCl, 1 mm DTT, 20 mm Tris-Cl, pH 7.4) at room temperature for 15 min, and spun down at 15,000 × g for 5 min. Precipitated polymers (Pellet I) were resolubilized in depolymerization buffer without Ca2+ (500 mm KCl, 1 mm EGTA, 20 mm Tris-Cl, pH 7.4), and the soluble II and insoluble pellet II fractions were separated after centrifugation. B, stoichiometric analysis of resolubilization of the CSQ2 polymer with JNT. Protein concentrations of the soluble II (gray) and pellet II (dark) fractions are shown.
FIGURE 6Contribution of JNT to Ca2+-induced CSQ2 polymerization. Polymers containing CSQ2 alone and both CSQ2 and JNT were examined using AFM. CSQ2 and CSQ2 plus JNT polymers were prepared by dialyzing protein solution (20 μm total protein) against polymerization-inducing buffer and washed with distilled water. Residual precipitates were observed under AFM. Images represent scans of the 5 × 5-μm area.
Third, we examined the oligomeric states of the CSQ molecule in the absence or presence of JNT using EM. Protein solutions were incubated with or without 5 mm Ca2+, and their supernatant fractions (Soluble I) were isolated by centrifugation to remove insoluble aggregates, mounted, and negatively stained for EM. Previous studies showed that CSQ exists mainly in dimeric, tetrameric, and polymeric states in the presence of 5 mm Ca2+ (
Comparing skeletal and cardiac calsequestrin structures and their calcium binding: a proposed mechanism for coupled calcium binding and protein polymerization.
). Consistent with earlier findings, we observed dimer-, tetramer-, and oligomer-like structures in EM images (Fig. 7, B and C). After magnification (Fig. 7, insets), CSQ appeared as a propeller-shaped particle (diameter, 60–80 Å) containing three to four globular domains in the presence of 5 mm Ca2+. We assume that the particle corresponds to a front-to-front dimer because dimers having three or four globular domains can be imaged, depending on the direction (angle) of attachment to the grid surface (Fig. 7B). The repetitive domain appears to be a thioredoxin-like domain (diameter ∼30 Å), previously identified in x-ray crystallography analyses (
). However, no evident structure was observed in the absence of Ca2+ (Fig. 7A). Previous intrinsic fluorescence, circular dichroism, Raman spectroscopy, NMR, and proteolytic digestion analyses performed to determine the physicochemical properties of CSQ revealed that CSQ is a mostly unfolded random coil at low ionic strength but that it folds into a compact structure with increasing ion concentrations (
). This result has been replicated in our EM experiments showing disruption of the thioredoxin-like domain in the absence of Ca2+ (Fig. 7A). Upon incubation of equal amounts of CSQ and JNT (Fig. 7C), the total number of propeller-shaped particles decreased, and the appearance of the high-order structure (chain-like) that prevailed in the CSQ-only sample also decreased (Fig. 7, B versus C). Our results consistently suggest that interactions between JNT and CSQ compete with the intermolecular interactions among CSQ, thereby facilitating disassembly or preventing the formation of CSQ polymers.
FIGURE 7Oligomeric states of the CSQ molecular in the absence or presence of JNT.B and C, negatively stained EM images of the soluble fractions of Ca2+-induced CSQ2 polymerization reaction in the absence (B) or presence of JNT (C). A, the Ca2+-free CSQ2 sample was used as a control. Insets represent magnified images.
Reversibility is a typical mechanism to maintain homeostasis in biological systems. The bidirectional reactions of CSQ polymerization and depolymerization may thus be critical in Ca2+ homeostasis. Previous studies have proposed that positive cooperative binding of Ca2+ by CSQ is concomitant with a shift in equilibrium from the soluble to the crystalline forms of CSQ (
). Thus, CSQ polymerization appears beneficial for Ca2+ deposit. Based on the crystal structures, various lengths of CSQ polymer are expected to form with increasing Ca2+ concentrations (
Comparing skeletal and cardiac calsequestrin structures and their calcium binding: a proposed mechanism for coupled calcium binding and protein polymerization.
). In addition, during FRET analysis of CSQ2 polymerization in cells cotransfected with CFP- and YFP-tagged CSQs, depletion of the endoplasmic reticulum Ca2+ stores with ionomycin resulted in a marked decrease in FRET efficiency, suggestive of CSQ depolymerization (
). Stopped-flow analysis showed that intrinsic CSQ fluorescence decreases with decreasing concentrations of Ca2+ from 1 mm to 0.1–100 μm, indicative of depolymerization (
). However, there is no direct evidence of Ca2+-dependent CSQ depolymerization in vivo with physiological relevance. Live cell imaging analysis (supplemental Movies S1 and S2 and Fig. 1) suggested JNT-dependent depolymerization of the CSQ polymer in vivo after Ca2+ depletion. CSQ polymer underwent slow conformational changes upon SR Ca2+ depletion using a SERCA inhibitor, and this was not synchronized with the twitches of normal, millisecond-long excitation-contraction coupling. We speculate that complete depolymerization does not occur during excitation-contraction coupling but that partial Ca2+ release from CSQ and a corresponding partial depolymerization do occur, which is followed by a rebound in Ca2+ and the repolymerization of CSQ. In contrast, under certain pathological conditions, virtually all of the Ca2+ becomes lost from CSQ, and this may lead to the full depolymerization of the CSQ polymer without the possibility of reconstruction. In cardiac muscle, it has been shown that the lower intraluminal SR Ca2+ concentration characteristic of heart failure results in abnormal CSQ trafficking (
Canato, M., Scorzeto, M., Giacomello, M., Protasi, F., Reggiani, C., Stienen, G. J., (2010) Massive alterations of sarcoplasmic reticulum free calcium in skeletal muscle fibers lacking calsequestrin revealed by a genetically encoded probe. Proceedings of the National Academy of Sciences of the United States of America 107, 22326–22331.
Our data strongly suggest that direct molecular interactions between CSQ and JNT are necessary for CSQ-mediated Ca2+ control in SR. JNT-null cardiomyocytes display increased frequency of Ca2+ sparks, indicative of increased SR leak, the phenotype of which is associated with arrhythmia (
). Experiments from the current study disclosed that upon overexpression of JNT, interactions between JNT and CSQ increase, inhibiting polymerization of CSQ and leading to reduced SR Ca2+ load and reduced spontaneous SR Ca2+ release. On the other hand, upon JNT ablation, the bridge between CSQ and RyR may be abolished, resulting in desynchronized SR Ca2+ release (
). We showed that the molar ratio of JNT-CSQ is critically important in Ca2+-CSQ dynamics by ectopic coexpression in C2C12 cells (Fig. 3). When the JNT level was significantly higher or lower than the CSQ level, depolymerization barely occurred (Fig. 3), possibly due to abnormal SR Ca2+ load and CSQ mislocalization, respectively. Previously, CSQ condensation was observed in L6 myoblast and HeLa cells transfected with CSQ, which remained unchanged after depletion of endoplasmic reticulum Ca2+ stores (
Overexpression of calsequestrin in L6 myoblasts: formation of endoplasmic reticulum subdomains and their evolution into discrete vacuoles where aggregates of the protein are specifically accumulated.
), possibly due to imbalance in the levels of CSQ relative to JNT. We replicated the necessity for JNT in CSQ depolymerization reactions in vitro in experiments using the purified proteins (FIGURE 5, FIGURE 6, FIGURE 7). Reversible CSQ polymerization and depolymerization were dependent on the Ca2+ concentration and JNT-CSQ ratios, consistent with in vivo data (FIGURE 1, FIGURE 2, FIGURE 3).
The stoichiometry of the RyR complex has been studied in detail. FKBP12 (or FKBP12.6), calmodulin, TRN, and JNT are each known to interact with the RyR subunit in a 1:1 manner (
Complex formation between junctin, triadin, calsequestrin, and the ryanodine receptor. Proteins of the cardiac junctional sarcoplasmic reticulum membrane.
Identification of calmodulin-, Ca2+-, and ruthenium red-binding domains in the Ca2+ release channel (ryanodine receptor) of rabbit skeletal muscle sarcoplasmic reticulum.
). These results suggest a potential interdependence between these three proteins. Furthermore, CSQ function is known to depend on the stoichiometry of the association with the RyR-TRN-JNT-CSQ macromolecular complex (
). Here, we found that the stoichiometric interaction of CSQ2 with JNT was critical for the CSQ2 depolymerization reaction in vivo and have suggested that the optimal ratio of JNT-CSQ expression for this reaction is 1:1 (protein levels). However, endogenous JNT is present at lower levels than endogenous CSQ under physiological conditions (
Purification, primary structure, and immunological characterization of the 26-kDa calsequestrin binding protein (junctin) from cardiac junctional sarcoplasmic reticulum.
). A possible explanation, which may reconcile the difference between physiological conditions and our experimental conditions, is seen in Fig. 8. It is likely that some JNT molecules remain free, whereas others interact with polymerized CSQ. Thus, the final interaction of JNT and CSQ may be at a ratio less than 1:1. This suggestion is in line with our in vivo imaging data showing that ectopically expressed JNT is diffusely expressed, whereas that of CSQ is speckled (Fig. 4, C and D). Thus, the limited concentrations of endogenous JNT in vivo might be necessary and sufficient to play key roles in the reversible dynamics of Ca2+-CSQ polymerization and depolymerization, which are located adjacent to RyR.
FIGURE 8Schematic diagram of CSQ, JNT, and RyR interactions in a junctional face of SR. At increasing Ca2+ concentrations, JNT coordinates with the growing CSQ polymer in the vicinity of a RyR-localizing Ca2+ store near the release port. Upon channel opening and deployment of Ca2+ ions, CSQ tethered to JNT undergoes depolymerization. CSQ, JNT, and Ca2+ are depicted as triangles, blue tubes, and red dots.
Complex formation between junctin, triadin, calsequestrin, and the ryanodine receptor. Proteins of the cardiac junctional sarcoplasmic reticulum membrane.
Calsequestrin binds to monomeric and complexed forms of key calcium-handling proteins in native sarcoplasmic reticulum membranes from rabbit skeletal muscle.
A missense mutation in a highly conserved region of CASQ2 is associated with autosomal recessive catecholamine-induced polymorphic ventricular tachycardia in Bedouin families from Israel.
Comparing skeletal and cardiac calsequestrin structures and their calcium binding: a proposed mechanism for coupled calcium binding and protein polymerization.
Overexpression of calsequestrin in L6 myoblasts: formation of endoplasmic reticulum subdomains and their evolution into discrete vacuoles where aggregates of the protein are specifically accumulated.
Canato, M., Scorzeto, M., Giacomello, M., Protasi, F., Reggiani, C., Stienen, G. J., (2010) Massive alterations of sarcoplasmic reticulum free calcium in skeletal muscle fibers lacking calsequestrin revealed by a genetically encoded probe. Proceedings of the National Academy of Sciences of the United States of America 107, 22326–22331.
Identification of calmodulin-, Ca2+-, and ruthenium red-binding domains in the Ca2+ release channel (ryanodine receptor) of rabbit skeletal muscle sarcoplasmic reticulum.
Purification, primary structure, and immunological characterization of the 26-kDa calsequestrin binding protein (junctin) from cardiac junctional sarcoplasmic reticulum.
53264-X&id=gr1.jpg){kind=link}
53264-X&id=gr2.jpg){kind=link}
53264-X&id=gr3.jpg){kind=link}
53264-X&id=gr4.jpg){kind=link}
53264-X&id=gr5.jpg){kind=link}
53264-X&id=gr6.jpg){kind=link}
53264-X&id=gr7.jpg){kind=link}
53264-X&id=gr8.jpg){kind=link}