calsequestrin rescues function an strutture of ca . .. in skeletal muscles

doi: 10.1152/ajpcell.00119.2011302:C575-C586, 2012. First published 2 November 2011;Am J Physiol Cell Physiol

Volpe, Feliciano Protasi and Alessandra NoriMirta Tomasi, Marta Canato, Cecilia Paolini, Marco Dainese, Carlo Reggiani, Pompeomicecalcium release units in skeletal muscles of CASQ1-null Calsequestrin (CASQ1) rescues function and structure of

You might find this additional info useful...

54 articles, 29 of which you can access for free at: This article citeshttp://ajpcell.physiology.org/content/302/3/C575.full#ref-list-1

including high resolution figures, can be found at: Updated information and serviceshttp://ajpcell.physiology.org/content/302/3/C575.full

found at: can beAmerican Journal of Physiology - Cell Physiology about Additional material and information

http://www.the-aps.org/publications/ajpcell

This information is current as of March 15, 2013.

1522-1563. Visit our website at http://www.the-aps.org/. Rockville Pike, Bethesda MD 20814-3991. Copyright © 2012 the American Physiological Society. ESSN:molecular physiology. It is published 24 times a year (twice monthly) by the American Physiological Society, 9650

is dedicated to innovative approaches to the study of cell andAmerican Journal of Physiology - Cell Physiology

at Biblioteca P

olo Di C

oppito on March 15, 2013

http://ajpcell.physiology.org/D

ownloaded from

http://ajpcell.physiology.org/

Calsequestrin (CASQ1) rescues function and structure of calcium release unitsin skeletal muscles of CASQ1-null mice

Mirta Tomasi,1 Marta Canato,1 Cecilia Paolini,2 Marco Dainese,2 Carlo Reggiani,1 Pompeo Volpe,1

Feliciano Protasi,2 and Alessandra Nori11Departments of Experimental Biomedical Sciences and of Anatomy and Physiology, University of Padova, Padova, Italy;and 2CeSI-Center for Research on Ageing and DNI-Department of Neuroscience and Imaging, University Gabrieled’Annunzio, Chieti, Italy

Submitted 18 April 2011; accepted in final form 28 October 2011

Tomasi M, Canato M, Paolini C, Dainese M, Reggiani C, VolpeP, Protasi F, Nori A. Calsequestrin (CASQ1) rescues function andstructure of calcium release units in skeletal muscles of CASQ1-nullmice. Am J Physiol Cell Physiol 302: C575–C586, 2012. First pub-lished November 2, 2011; doi:10.1152/ajpcell.00119.2011.—Ampli-tude of Ca2� transients, ultrastructure of Ca2� release units, andmolecular composition of sarcoplasmic reticulum (SR) are altered infast-twitch skeletal muscles of calsequestrin-1 (CASQ1)-null mice. Todetermine whether such changes are directly caused by CASQ1ablation or are instead the result of adaptive mechanisms, here weassessed ability of CASQ1 in rescuing the null phenotype. In vivoreintroduction of CASQ1 was carried out by cDNA electro transfer inflexor digitorum brevis muscle of the mouse. Exogenous CASQ1 wasfound to be correctly targeted to the junctional SR (jSR), as judged byimmunofluorescence and confocal microscopy; terminal cisternae(TC) lumen was filled with electron dense material and its width wassignificantly increased, as judged by electron microscopy; peak am-plitude of Ca2� transients was significantly increased compared withnull muscle fibers transfected only with green fluorescent protein(control); and finally, transfected fibers were able to sustain cytosolicCa2� concentration during prolonged tetanic stimulation. Only theexpression of TC proteins, such as calsequestrin 2, sarcalumenin, andtriadin, was not rescued as judged by Western blot. Thus our resultssupport the view that CASQ1 plays a key role in both Ca2� homeo-stasis and TC structure.

calcium homeostasis; excitation-contraction coupling; sarcoplasmicreticulum

EXCITATION-CONTRACTION (E-C) coupling is the mechanism thatlinks transverse (T)-tubule depolarization to Ca2� release fromthe sarcoplasmic reticulum (SR). Transient increase of cyto-plasmic Ca2� concentration ([Ca2�]c) triggers, in turn, musclecontraction. Subsequent [Ca2�]c decrease brings about skeletalmuscle relaxation via multiple mechanisms: Ca2� reuptake andstorage in the SR by Ca2� ATPase (SERCA) and calsequestrin(CASQ), respectively; Ca2� binding by cytoplasmic proteins,such as parvalbumin; and, locally, Ca2� uptake by mitochon-dria (18, 46). Recently, a mechanism involved in Ca2� storerefilling (store-operated calcium entry) has been described notonly in nonexcitable cells but also in skeletal muscle (22)where it might play a role in specific conditions entailingprolonged stimulations and, thus, leading to Ca2� store deple-tion (23).

CASQ is a high-capacity/low-affinity Ca2� binding proteinof the SR lumen, restricted to junctional SR (jSR; Ref. 14),

which forms Ca2�-dependent oligomers/polymers (37, 38) thatincrease the Ca2� binding capacity up to 80 mol Ca2�/molCASQ (14). There are two genes encoding two CASQ iso-forms (1): the cardiac (CASQ2) isoform is expressed in heart,slow-twitch muscle fibers and in fast-twitch muscle fibers onlyduring early stages of postnatal development, whereas theskeletal (CASQ1) isoform is expressed in both fast- and slow-twitch muscles at increasing levels from birth to adulthood.Conceivably, the functions of CASQ2 are different from thoseof CASQ1, as inferred from in vitro studies (2, 40) on ryano-dine receptor 1 (RYR1) regulation by recombinant and purifiedCASQ.

CASQ1, on the luminal face of jSR, interacts with severalSR proteins. A large body of in vivo and in vitro data existsconcerning the interaction of CASQ1 with triadin (TR) andjunctin (JCT; for a recent review, see Ref. 13), and it has beenproposed that the molecular ratio between CASQ1 and TR/JCTis relevant for both SR biogenesis and structure of calciumrelease units (CRUs; Refs. 3, 28, 50). In the assembly of jSR,constitutive proteins appear to have different turnover, i.e., TRis much less mobile than JCT and RYR, so that TR is strictlyassociated to the junctional quaternary macromolecular com-plex whereas JCT establishes transient interactions (7). On theaccount of its Ca2� binding properties, multiple roles havebeen envisioned for CASQ1 in Ca2� handling: Ca2� bufferwithin the SR and Ca2� sensor for RYR1 in E-C coupling (2),regulator in store-operated calcium entry (48, 53), molecularplayer in SR biogenesis, and fiber type specification (15).

The role of CASQ1 in adaptive mechanisms to differentstimuli is not well known (44). Interestingly, structural altera-tions of CRUs in fast-twitch muscle fibers have been observedin some pathological conditions characterized by CASQ1 over-expression in denervation and reinnervation models (12, 25,49, 54) as well as in knockout murine models of ER/SRresident proteins, e.g., pan-TR-null mice (47, 34) or hexose-6-phosphate dehydrogenase (H6PD) null mice (24), where down-regulation of CASQ1 and of other structural proteins is de-scribed. Whether CASQ1 has a direct or indirect role inmaintaining CRUs geometry and determining CRUs plasticityremains likewise to be established in skeletal muscle.

Notwithstanding the important functions ascribed to CASQ,genetic ablation of either CASQ1, CASQ2, or both (35, 20, 45,5, 36) was not birth lethal in the mouse. In CASQ1-null mice,contractile responses of muscle fibers were not significantlyaltered, but there was profound ultrastructural remodeling ofCRUs, reduced content of stored Ca2�, reduced amplitude ofCa2� transients (35), and inability to support sustained con-tractile activity (5). Additional adaptive features, i.e., increased

Address for reprint requests and other correspondence: A. Nori, Dept. ofExperimental Biomedical Sciences, Univ. of Padova, Viale G. Colombo 3Padova, I-35131, Italy (e-mail: [email protected]).

Am J Physiol Cell Physiol 302: C575–C586, 2012.First published November 2, 2011; doi:10.1152/ajpcell.00119.2011.

0363-6143/12 Copyright © 2012 the American Physiological Societyhttp://www.ajpcell.org C575

at Biblioteca P

olo Di C



ownloaded from


numbers of mitochondria and density of RYR1 were alsoreported (35). In addition, CASQ1-null mice (mostly males)displayed a high risk of spontaneous sudden death and under-went lethal episodes of a syndrome resembling both humanmalignant hyperthermia and environmental heat stroke whenexposed to heat or halogenated anesthetics (8, 39).

Although extensive work has been done in CASQ1-nullmice, it still remains to be determined to which extent thecomplex CASQ1-null phenotype is either directly caused byCASQ1 ablation or by compensatory mechanisms that maskCASQ1 functions. To this aim, the present study assesseswhether acute reintroduction of CASQ1 can rescue somefeatures of the CASQ1-null phenotype.

MATERIALS AND METHODS

CASQ1-Null and Wild-Type Mice

Experiments were carried out in 4-mo-old CASQ1-null and wild-type (WT) C57 BL/6 female mice. CASQ1-null mice were obtained aspreviously described (35). All experimental protocols have beenapproved by Institutional Ethical Committees of the Universities ofPadova and of Chieti. Mice were kept in accredited animal facilitiesand killed by anaesthetic euthanasia.

Plasmids and cDNAs

Enhanced green fluorescent protein (eGFP) the bicistronic mam-malian expression vector, pCMS-eGFP (BD Biosciences, ClontechLaboratories, San Jose): the first box is a multi-cloning site (MCS)downstream of the immediate early promoter of cytomegalovirus(PCMVIE) that allows expression of the gene of interest; the secondbox contains the SV40 strong promoter upstream of the eGFP suitableas a transfection marker. eGFP, a red-shifted variant of WTGFP, iscompatible with fura-2 excitation and emission spectra and was usedas indicator of transfection in Ca2� transient recordings. For CASQ1-eGFP, the murine CASQ1 cDNA was cloned in the MCS of pCMS-eGFP; the resulting vector, pCMS-eGFP-CASQ1, was kindly pro-vided by Dr. E. Rios (Rush University, Chicago, IL). Amplificationand electroporation grade purification of plasmidic DNA were per-formed using the Maxiprep purification kit (Qiagen Sciences, Ger-mantown, MD) according to the manufacturer’s protocols.

In Vivo Electroporation

Flexor digitorum brevis (FDB) was chosen as the experimentalmodel because there are established and reliable protocols for both invivo electroporation of the muscle and enzymatic dissociation ofsingle muscle fibers (41). Electroporation of posterior pad muscleswas performed according to DiFranco et al. (11) with modifications:mice were deeply anesthetized using 4% isoflurane in O2 with anapproved gas anesthetic machine. Anesthesia was maintained using arodent face mask; injection of 5 �l of ialuronidase was followed 1 hlater by injection of 10 �l (10 �g) of plasmidic DNA for each foot.After 10 min, electroporation was performed with an Electro SquarePorator ECM 830 (20 pulses of 100 V/cm, 20 ms). Mice were killed7 days after electroporation. Transfected muscles were referred to asWT � eGFP (WT FDB muscle transfected with pCMS-eGFP, mock),CASQ1-null � eGFP (CASQ1-null FDB muscle transfected withpCMS-eGFP, control), and CASQ1-null � CASQ1 (CASQ1-nullFDB muscle transfected with pCMS-eGFP-CASQ1).

Preparation of Total Homogenates and of Single Fibers Pools

Total homogenates from FDB muscles. Seven days after transfec-tion, FDB muscles from control and CASQ1-null mice were dis-sected, homogenized for 3 min in 3% SDS (wt/vol), 1 mM EGTA,boiled for 5 min, and centrifuged at 900 g for 15 min. Protein

concentration of supernatants, referred to as total homogenates, wasdetermined according to (26).

Single fibers pools. After dissociation of FDB muscles, 10 or 20fibers were collected under the dissection microscope (at �60) fromindividual FDB muscles and stored in 20 �l of Laemmli’s solubili-zation buffer (21). Silver staining of total homogenate from WT andCASQ1-null muscles was performed according to Morrissey (31).

Western Blot Analysis of Total Homogenates and of Single FibersPools

Each sample. For each sample, 15–50 �g of protein were loaded on7.5–10% SDS-polyacrylamide gels depending on the protein beinginvestigated: 50 �g for dihydropyridine receptor (DHPR), SERCA1,TRs, and calreticulin; 25 �g for CASQ2; 15 �g for BiP; 40 �g forJCT; and 30 �g for sarcalumenin and GP 53. The following dilutionsand primary antibodies were used: monoclonal anti SERCA1 (1:5,000), polyclonal anti both CASQ1/CASQ2 (1:1,000), monoclonalanti CASQ1 (1:1,000), monoclonal anti-TR (1:1,000), polyclonal antiBiP (1:500), polyclonal anti calreticulin (1:1,000), and monoclonalanti-�1-subunit of DHPR1 (1:500) were all from Affinity Bioreagents(Affinity Bioreagents, Golden, CO); and monoclonal anti-sarcalu-menin and 53 kDa glycoprotein (1:200) and polyclonal anti JCT(1:200) were from Santa Cruz (Santa Cruz Biotechnology, SantaCruz, CA). Secondary anti-mouse or anti-rabbit (1:5,000) alkalinephosphatese-conjugated antibodies were from Sigma (Milan, Italy).

Single fiber pools. For pools of single fibers, SDS-PAGE wascarried out on 10% polyacrylamide gels and Western blot was per-formed using monoclonal antibodies against CASQ1 (dilution 1:800),TRs (dilution 1:800), and fast myosin heavy chain (1:1,000; Sigma).Secondary anti-mouse or anti-rabbit (1:5,000) alkaline phosphatase-conjugated antibodies were from Sigma (Milan, Italy). The immuno-decorated bands were visualized by a ready-to-use, precipitatingsubstrate system for alkaline phosphatase (BCIP/NBT liquid substratesystem; Sigma).

Densitometric Analysis and Image Processing

Images were obtained with an HP Scanjet scanner and AdobePhotoshop CS2 version 9.0. Densitometry was performed with ScionImage Software without modification of the images (raw images) toquantify protein band intensities. Normalization of the signal wasobtained relative to that of actin stained by Ponceau red the immu-nological signal referable to TR of each pool of fibers was divided bythe signal obtained by actin stained by Ponceau red for the very samepool of fibers (internal normalization; see Fig. 5C). This ratio wasobtained for each pool and t-test was performed between two groups:control (fibers transfected with pCMS-eGFP) and CASQ1 expressing(fibers transfected with pCMS-eGFP-CASQ1). In the case of CASQ2,to correlate densitometric signals to protein content, calibration curveswere generated as follows: increasing amounts of total homogenate ofWT FDB (ranging between 10 and 30 �g) were loaded on the samegel and processed at the same time. A calibration curve was obtainedfor each blot and extrapolation was made to the relative proteinquantity (see also Ref. 30). For display purposes, blots of JCT (seeFig. 1) and of TR 95 and 51 (see Fig. 5) were modified by AdobePhotoshop linear tools for brightness and contrast.

Immunofluorescence Microscopy

Single fibers from FDB were obtained 7 days after electroporation,plated on laminin coated coverslips for 48 h, and then fixed with 4%(wt/vol) formalin Sigma. Samples were permeabilized with 0.1%(vol/vol) Triton X-100 and 15% (vol/vol) goat serum in PBS for 1 hand incubated overnight with primary antibodies at 4°C withoutTriton X-100. Both monoclonal anti-CASQ1 antibodies and poly-clonal anti-CASQ1/CASQ2 antibodies were diluted 1:500 in PBS.After three 20-min washes, samples were incubated for 2 h at room

C576 RESCUE OF FUNCTION/STRUCTURE IN CASQ1-NULL FIBERS

AJP-Cell Physiol • doi:10.1152/ajpcell.00119.2011 • www.ajpcell.org

at Biblioteca P

olo Di C



ownloaded from


temperature with secondary rhodamine goat anti-mouse and/or goatanti-rabbit antibodies (1:500; Jackson ImmunoResearch Laboratories,Lexington, KY). Specimens were observed in either a fluorescencemicroscope (Leica DMLB; Leica Microsystem, Mannheim, Germany)or a confocal microscope Leica TCS-SP2.

Preparation of Samples for Electron Microscopy and Data Analysis

FDB muscles were dissected, fixed, and prepared for electronmicroscopy (EM) as described in (34). Ultrathin sections were cutin a Leica Ultracut R microtome (Leica Microsystem) using a

Diatome diamond knife (Diatome CH-2501; Biel, Switzerland).After being stained in 4% uranyl acetate and lead citrate, sectionswere examined in a FP 505 Morgagni Series 268D electronmicroscope (FEI, Brno, Czech Republic) at 60 kV equipped with aMegaview III digital camera and AnalySIS software (OlympusSoft Imaging Solutions).

Intracellular Ca2� Measurements

Single FDB fibers were isolated by a modified collagenase/proteasemethod as described previously (10). There was no difference in fibersyield between transfected and nontransfected muscles. Forty-eighthours after dissociation, fibers were incubated with 5 �M fura-2acetoxymethyl ester (fura-2 AM; Invitrogen, San Giuliano Milanese,Italy) in 25 mM NaCl, 5 mM KCl, 1 mM MgSO4, 1 mM KH2PO4, 5.5mM glucose, 1 mM CaCl2, 20 mM HEPES, and 1% BSA pH7.4(incubation buffer) for 30 min at 37°C. After fura-2 loading, fiberswere washed twice for 10 min in the incubation buffer withoutBSA. A minimum of 30 min was allowed for fura-2 deesterificationbefore the fibers were imaged. Intracellular Ca2� transients weremeasured at 25°C using a dual-beam excitation fluorescence pho-tometry setup (IonOptix, Milton, MA). Single muscle fibers werestimulated with two distinct frequency protocols and routinely fivefibers were examined in each Petri dish: 1) after 10 min ofsteady-state pacing at 0.5 Hz, 10 Ca2� transients were recordedfrom each fiber; and 2) a train of stimulation (60 Hz for 2 s) wasdelivered with a recovery time of 5 min between registrations.Ca2� transients were analyzed using IonWizard software designedby IonOptix. Peak amplitude was calculated subtracting basalfluorescence values from peak values.

Statistical Analysis

Data were expressed as means � SE. Student’s t-test was used forcomparisons between CASQ1-null and WT data. One-way ANOVAtest was used when three groups were compared using Prism Graph-Pad. Statistical significance was set at P � 0.05.

RESULTS

Expression of SR and T-Tubule Markers in FDB Muscles

Quantitative Western blot analysis of total homogenates ofFDB muscles from CASQ1- null and WT mice was performedto assess the expression of proteins involved in Ca2� handlingand SR structure. In Fig. 1A, controls for gel loading andelectrotransfer are shown; the comparison between 3 �g ofcrude homogenates from WT and CASQ1-null muscles did notreveal any appreciable difference in protein pattern. The mainband at 42 kDa represents actin, as identified by anti-actinantibody, and was used as the loading control in Fig. 1B.

Fig. 1. Western blot analysis of flexor digitorum brevis (FDB) muscles fromcalsequestrin-1 (CASQ1)-null and wild-type (WT) mice. A, left lanes: silverstaining of crude homogenates from WT and CASQ1-null FDB muscles andno appreciable differences in protein pattern. Arrow indicates position ofCASQ in WT crude homogenates. A, right lanes: protein pattern of homoge-nates from WT electroporated FDB muscles stained with Ponceau red (PR):note that the major band at 45 kDa corresponds to actin as demonstrated byimmunodetection with anti-actin antibody. Ponceau red stained actin was usedas the loading control in B. B: representative Western blots of FDB homoge-nates from WT and CASQ1-null mice. C: densitometric analysis of Westernblots; data are expressed as percentage of the WT signal � SE, and WT signalwas set at 100%. Samples were obtained from n � 3 WT and n � 3CASQ1-null muscles. In samples from CASQ1-null mice, CASQ2, triadin(TR) 95, sarcoplasmic-endoplasmic reticulum calcium ATPase 1 (SERCA1);and sarcalumenin (Srl) were significantly decreased compared with WT (*P �0.001, unpaired two-tailed t-test). DHPR, dihydropyridine receptor; JCT,junctin; BiP, immunoglobulin binding protein; CRL, calreticulin.

C577RESCUE OF FUNCTION/STRUCTURE IN CASQ1-NULL FIBERS


at Biblioteca P

olo Di C



ownloaded from


Representative immunoblots and densitometric analysis areshown in Fig. 1, B and C, respectively. CASQ2, TR 95,sarcalumenin, and SERCA1 were significantly reduced in FDBfrom CASQ1-null muscles (P � 0.05), whereas no significantchanges were detected for DHPR, TR 51, JCT, glycoprotein53, calreticulin, and BiP. Based on calibration curves (seeMATERIALS AND METHODS for details), mean values of the CASQ2signals in arbitrary units, 16.82 � 1.52 and 77.55 � 4.93 inCASQ1-null and WT FDB, respectively, were converted toprotein equivalents (WT � 22.42, CASQ1-null � 12.42), i.e.,CASQ2 was reduced by 50% in FDB muscles from CASQ1-null mice. Thus, SR proteins endowed with Ca2� bindingproperties and potential substitutes for CASQ1 were foundeither not changed, e.g., calreticulin, or decreased, e.g.,CASQ2 and sarcalumenin.

Structural features of jSR-T-Tubule Junctions and CRUs inFDB Muscle Fibers

Figure 2 shows the ultrastructural differences between CRUsof WT and CASQ1-null fibers from FDB muscles. Changesaffecting CRUs of CASQ1-null FDB muscles were qualita-tively identical to those previously described in EDL muscles(35), the main difference being the reduction in size of the SR

terminal cisternae. The average width of SR terminal cisternaewas calculated in triads and in multiple CRUs (only present inCASQ1-null fibers) by measuring the distance between theopposing membranes of TCs, as marked by dashed lines in Fig.2, B, E, and F, and in Fig. 3. Ablation of CASQ1 resulted indramatic shrinkage of the SR lumen: 25.0 � 4.5 vs. 62.4 �10.5 nm in CASQ1-null and WT fibers, respectively (Fig. 3).Measurements of TC width were variable. However, distribu-tion graphs (Fig. 3, bottom) showed that only 15 out of 716TCs presented a lumen width �40 nm in WT fibers, whereasonly 18 out of 487 measurements were �40 nm in CASQ1-nullfibers. Additional morphological changes were detected inCRUs of CASQ1-null fibers, as follows: 1) Several CRUsdisplayed multiple rows of RYRs (see arrows in Fig. 2F), inCASQ1-null fibers, whereas junctions between T-tubules andjSR (or couplons) were usually in the form of triads (arrows inFig. 2A) and usually contained two rows of RYRs (arrows inFig. 2, B and C), in WT fibers. 2) Several CASQ1-null fiberscontained multilayered junctions (larger arrows in Fig. 2 D).Multilayered CRUs (Fig. 2F) were also found in EDL muscles(35), although the percentage of fibers presenting multilayeredjunctions was significantly lower in FDB compared with EDLmuscles (35 vs. 70%). 3) Several CRUs in CASQ1-null fibers

Fig. 2. Morphology of calcium release units (CRUs) in FDBmuscles from WT and CASQ1-null mice. A–C: in WTsamples, CRUs were 1) positioned on both sides of the Zline (A, arrows), 2) usually formed by 3 elements (B and C,triads), and 3) transversely oriented. Ryanodine receptors(RYRs) formed 2 rows along the junctional face of terminalcisternae (TC; B and C, arrows), which presented a widelumen (dashed lines in B). D–F: in CASQ1-null samples,CRUs were abnormal (D, arrows) because 1) often longi-tudinally oriented (D, empty arrow), 2) frequently com-posed of multiple layers (D and F, large arrows), and3) contained multiple rows of RYRs decorating the junc-tional membrane of TC (F, arrows), which presented areduced lumen (distance between dashed lines in E and F;cfr. Ref. 35). Bars: A and D � 0.2 �m; B, C, E, and F � 0.1�m. SR, sarcoplasmic reticulum.



at Biblioteca P

olo Di C



ownloaded from


were longitudinally oriented (empty arrow in Fig. 2D), afeature which is rare in adult WT fibers.

Cytosolic Ca2� Transients Induced by Electrical Stimulationin CASQ1-null FDB Muscle Fibers

The ratiometric probe fura-2 was used to determine the free[Ca2�]c. At rest, [Ca2�]c levels were similar in WT andCASQ1-null muscle fibers (data not shown), whereas thetransient induced by a single pulse showed a lower amplitudein muscle fibers lacking CASQ1: mean peak amplitude valueswere 0.6934 � 0.0653 and 0.5129 � 0.0342, respectively (see

also Ref. 35). Moreover CASQ1-null muscle fibers were un-able to sustain high and steady levels of cytosolic [Ca2�]c

during prolonged tetanic stimulations (60 Hz for 2 sec)whereas WT fibers could, as shown in Fig. 4A. This behavioris similar to that observed in CASQ2/CASQ1-null FDB musclefibers (5). The decline of [Ca2�]c was quantified measuring thearea below each tracing and measuring the peak value at thefirst pulse (t0) and at the last pulse of the train (t2). As reportedin Fig. 4B, the area was significantly smaller in CASQ1-nullmuscle fibers, compared with WT. In CASQ1-null fibers, theamplitude of the last response (t2) was significantly lower than

Fig. 3. Measurements of TC width. Data are means �SD. Column I and II: width of SR terminal cisternaeprofile (A–D, dashed lines) was measured in triadsand in multiple CRUs (only present in CASQ1-nullfibers); ablation of CASQ1 resulted in significantshrinkage of SR terminal cisternae and distributiongraphs (bottom) clearly indicate a higher frequencyof cisternae with a reduced lumen width (see leftshift). Column III: data were obtained from fibersclassified as transfected whereas not transfected andovertransfected fibers were not taken into account(see RESULTS for more details). Upon reexpressionof CASQ1, junctional SR morphology (C) and lu-men width were restored. Distribution graphs (bot-tom) confirm that SR cisternae with an enlargedlumen were much more frequent than in CASQ1-null fibers. Column IV: in control fibers transfectedonly with pCMS-enhanced green fluorescent protein(eGFP), morphology of SR lumen remained virtu-ally identical to that of CASQ1-null fibers (compareD and B), further proving that changes shown in Cand column III were indeed the result of CASQ1reexpression. Measurements and distribution graphfully support this interpretation. *Data from Ref. 35.†P � 0.001, significantly different from CASQ1-null group. #P � 0.1, not significantly differentfrom WT group. ‡P � 0.1, not significantly differ-ent from CASQ1-null. §P � 0.001, significantlydifferent from WT and CASQ1-null � CASQ1.Scale bar � 0.1 �m.

Fig. 4. Cytosolic Ca2� concentrations during 60-Hzstimulation in muscle fibers of FDB from CASQ1-null and WT mice. A: representative tracings offura-2 fluorescence obtained with high frequencystimulation (60 Hz) trains (2-s duration). B andC: quantitative analysis of Ca2� transients (n � 19for WT fibers; n � 26 for CASQ1-null fibers). B:area (mean values � SE) measured under the fura-2tracings is highly significantly reduced in CASQ1-null (***P � 0.001) compared with WT.C: amplitude (mean values � SE) of initial (t0) andfinal (t2) peaks at 60-Hz stimulation. Note thatinitial and final peaks of WT fibers were not differ-ent, whereas initial and final peaks of CASQ1-nullfibers were significantly different (***P � 0.001paired t-test). Both t0 and t2 were lower in CASQ1-null fibres compared with WT; however, only thedifference of t2 was statistically significant (P � 0.05).Mean values of area (�SE) for WT � 1.834 � 0.338and CASQ1-null � 0.525 � 0.049; initial peak (t0) forWT � 0.879 � 0.092 and CASQ1-null � 0.705 �0.042; and final peak (t2) for WT � 0.827 � 0.101,and CASQ1-null � 0.229 � 0.025.



at Biblioteca P

olo Di C



ownloaded from


t0, whereas in WT fibers was not (Fig. 4C). Moreover, the t2value was significantly smaller in CASQ1-null muscle fibers,compared with WT, whereas no significant difference wasdetectable between t0 values of CASQ1-null and WT fibers.

Reconstitution of the CASQ1 Complement: CorrectLocalization of CASQ1 Despite Persisting Low Levels of TR

Acute reexpression of CASQ1 in FDB muscles of CASQ1-null mice was achieved by cDNA electrotransfer. Seven daysafter electroporation with either pCMS-eGFP-CASQ1 orpCMS-eGFP, FDB muscles were enzymatically dissociatedand single fibers were plated. All fibers reexpressing CASQ1were readily detected under the dissection microscope becauseof the green fluorescence, and ranged between 30 and 80% ofthe total plated fibers, i.e., the efficiency of transfection, dis-sociation, and plating was variable. The correlation betweeneGFP expression and CASQ1 signal was, by and large, onlyqualitative: in fact, the eGFP expression did not show signif-icant positive correlation with the CASQ1 amount in totalhomogenates. The CASQ1 content was measured in greensingle fibers and in total homogenates by densitometry ofWestern blot, estimated to be variable in single fibers anddetermined to be, on average, 61% of that of WT in totalhomogenates (not shown), i.e., there was a sizeable althoughincomplete reconstitution of the total CASQ1 complement.

To determine whether reexpression of CASQ1 could rescueexpression of TR, we used Western blot analysis in pools of 20single green fibers, dissociated from FDB muscles expressingeither eGFP alone (control) or both eGFP and CASQ1. Rep-resentative Western blots of such pools obtained by eithercontrol or CASQ1-transfected muscles are shown in Fig. 5, Aand B. Transfected CASQ1-null fibers expressed CASQ1 (A)in variable amounts compared with WT fibers, whereas fiberstransfected with empty vector did not express CASQ1. In Fig.5, A and B, variability of myosin and actin among the poolswas visible. This indicates that the dimension of the fibers ineach pool was not constant and that normalization of the TRsignal was necessary to compare its expression between controland CASQ1-expressing pools. The TR signal was expressed aspercentage of the actin signal of the same pool that wasconsidered 100%. Three pools of CASQ1-null � eGFP andthree pools of CASQ1-null � CASQ1 were compared with thet-test. The mean values and SE were plotted in the bar graph ofFig. 5C. Densitometric analysis (Fig. 5C) indicated that ex-pression of TR 95 and TR 51 was not significantly changed,despite the presence of large amounts of exogenous CASQ1.Moreover, changes in expression of other SR proteins wereevaluated in total FDB homogenates because of lower concen-tration of relevant proteins. Western blot with antibodies spe-cific for either CASQ2, BiP, calreticulin, sarcalumenin, glico-protein 53, or JCT showed that the expression of none of theseproteins was significantly altered by acute expression ofCASQ1 (data not shown).

The correct localization of exogenous CASQ1 was assessedby immunofluorescence and confocal microscopy. Followingelectroporation with pCMS-eGFP-CASQ1, dissociation, andplating, single fibers were analyzed by immunofluoresencewith anti-CASQ1 antibodies. Confocal analysis showed thatexpression of eGFP was variable. Almost all fibers expressingeGFP were also immunolabeled with anti-CASQ1 antibodies,

whereas all fibers negative for eGFP were also negative foranti-CASQ1 antibodies. In eGFP-positive fibers (Fig. 6, E andF), the well-known, characteristic double rows of anti-CASQ-positive foci at the A-I interface, distributed in the whole fiber(Fig. 6F, inset) that identify exogenous CASQ1, were detect-able. Since the immunostaining pattern was identical to that ofWT fibers (Fig. 6B), these results provide evidence that exog-enous CASQ1 was correctly targeted to TC. As a negativecontrol, such a regular immunostaining pattern was absent inCASQ1-null fibers transfected with eGFP alone (Fig. 6D).

Content and Width of TC are restored by Expression ofExogenous CASQ1 in FDB Muscle Fibers

Qualitative and quantitative EM analysis of CRUs in FDBfibers after electroporation with either pCMS-eGFP-CASQ1 orpCMS-eGFP is reported in Fig. 7 and Fig. 3. Seven days afterelectroporation, eGFP fluorescence was detectable in all mus-

Fig. 5. In vivo transfection of FDB muscles: detection of CASQ1 and TR inpools of single fibers. A and B: Western blots carried out on pools of 10 and20 green fibers obtained from WT and CASQ1-null muscles transfected witheither eGFP (control) or eGFP � CASQ1. Each lane represents an individualpool of single fibers (sf) loaded on the gel. A: Ponceau red staining of actinshows protein loading control, the same blot was developed with anti-CASQ1antibodies. B: Western blots carried out on pools of 20 green fibers obtainedfrom CASQ1-null muscles transfected with either eGFP (control) or eGFP �CASQ1. Each lane represents an individual pool of 20 green single fibers.Western blots stained with Ponceau red (actin) or developed with anti-myosinheavy chain antibody (myosin) were used for normalization as described in theMATERIALS AND METHODS. C: comparison between normalized signals obtainedwith anti-TR antibodies was not significant (ns). Apparent molecular mass ofTR isoforms, i.e., 95 and 51 kDa, was determined with molecular massstandards that are omitted.



at Biblioteca P

olo Di C



ownloaded from


cles after dissection but disappeared at the end of the fixationprocedure for EM and was, thus, not useful to select transfectedfibers in EM experiments. Moreover, expression of CASQ1following electroporation with pCMS-eGFP-CASQ1 was vari-able, as shown above (Fig. 5, A and B) and Western blotanalysis could not be performed in fibers fixed for EM. Fol-lowing electroporation with pCMS-eGFP-CASQ1, the moststriking change was the appearance of clusters of fibers inwhich the size of the TC lumen was increased. In FDBelectroporated with pCMS-eGFP-CASQ1, three groups of fi-bers were identified based on morphological appearance of TC:group 1: fibers displaying enlarged TCs (see Fig. 7B, A-C ofFig. 3); group 2: fibers with flat TCs, virtually identical to thosein CASQ1-null muscles (see Fig. 7C); and group 3: fibers inwhich TCs were overswollen (see Fig. 7, D and E). In threedifferent FDB muscles electroporated with pCMS-eGFP-

CASQ1, 44 out 130 fibers (34% on average) displayed en-larged TC size (group 1) and were classified as transfectedfibers. Such fibers always contained many (even if not all)triads, CRUs with dilated SR (such as that in Fig. 7B, dashedlines), which appeared filled with an electron dense matrix andwere virtually identical to those of WT fibers (Fig. 2, B and C).Fibers containing no triads (or multilayered CRUs) with dilatedTC (group 2) were classified as not transfected, i.e., theirCRUs (Fig. 7C) were identical to those of CASQ1-null fibers(compare with Fig. 2, E and F). Finally, very few fibers (�5%)with overswollen TC were classified as overtransfected (group3). In three control FDB muscles electroporated with pCMS-eGFP, only fibers containing flat TCs (Figs. 3D and 7F),virtually identical to those in CASQ1-null muscles, werefound. Therefore, electroporation with pCMS-eGFP-CASQ1restored the TC content in a group of fibers classified as

Fig. 6. Localization of exogenous CASQ1 in single FDBfibers from CASQ1-null mice after in vivo transfection.Three fibers representative of the three groups described inMATERIALS AND METHODS are shown. A, B, C, and D: WT (Aand B) and a CASQ1-null fiber (C and D) electroporatedwith pCMS-eGFP immunostained with primary anti-CASQ1 antibody and secondary rhodamin conjugated anti-body. A and C: fibers imaged with laser beam settingssuitable for eGFP (green) fluorescence. B and D: rhodamine(red) fluorescence referable to endogenous CASQ1 (noteabsence of the signal in D). E and F: CASQ1-null fibertransfected with pCMS-eGFP-CASQ1 and immunostainedas before, the image was obtained with the same lasersettings as B and F. Double-row, cross striation visible in Band F was taken as indication of correct targeting of exog-enous CASQ1 to CRUs. Note that each row represents thetriads aligned at the A-I interface, and in some area 4 dotsbelonging to the same sarcomere are resolved (arrows ininset in F). Bar � 12.5 �m. F, inset: area at highermagnification (�4).



at Biblioteca P

olo Di C



ownloaded from


transfected. The width of TC lumen was measured in trans-fected fibers and in control fibers electroporated with pCMS-eGFP (Fig. 3). All CRUs (both triads and multiple junctions)were measured: the average width of TC in transfected fiberswas 66.3 � 10.2 nm (see Fig. 3), a value significantly different(P � 0.001) from that of CASQ1-null fibers (25.0 � 4.5 nm),but very similar to that of WT fibers (62.4 � 10.5 nm).Distribution graphs (Fig. 3, bottom) show that in transfectedfibers 285 out of 302 TC displayed a lumen width �40 nm (seedashed lines). Measurements of TC width and relative distri-bution graphs of control fibers (Fig. 3, column IV) show thatthe average TC width was 27.5 � 5.9 nm, with 330 out of 368TC being �40 nm, i.e., virtually identical to those of CASQ1-null muscles. Very few TCs present a lumen width of 40 nmfurther confirming that electroporation per se and expression ofeGFP alone did not increase TC size. Finally, transfection withpCMS-eGFP-CASQ1 yielded very few fibers with overswollenTCs (Fig. 7, D and E). These fibers classified as overtrans-fected are comparable to cardiac myocytes overexpressingCASQ2, which causes large swelling of SR terminal cisternae(19). Junctional SR of swollen TCs is entirely filled with anelectron dense matrix (see asterisks in Fig. 7, D and E), whosestructure is referable to CASQ network (cfr. 19). Not trans-fected and overtransfected fibers were not included in mea-surements of TC width, as reported in Fig. 3, column III.

Longitudinally oriented and multilayered CRUs are notcommonly found in adult WT skeletal muscle fibers but arefrequently found both in CASQ1-null fibers (see Fig. 2) and intransfected fibers. Quantitative analysis, although, showed thatfrequency of multiple junctions was decreased: the number ofmultiple junctions/100 �m2 dropped from 5.51 � 1.03 inCASQ1-null fibers to 1.68 � 0.6 in transfected fibers.

Cytosolic [Ca2�] and Amplitude of Electrically Induced Ca2�

Transients Are Restored by Expression of Exogenous CASQ1

To assess whether CASQ1 reexpression was able to restoreSR function, fura-2 signals were recorded in three groups ofsingle fibers: 1) WT fibers transfected with eGFP (mock),2) CASQ1-null fibers transfected with eGFP (control), and 3)CASQ1-null fibers transfected with CASQ1 � eGFP. Theexperimental protocol included mock-transfected WT fibers toavoid artifacts of transfection per se on intracellular Ca2�

kinetics. Figure 8A shows that the resting basal level of fura-2,indicative of basal Ca2� concentration was virtually identicalat 25°C in fibers obtained from CASQ1-null mice transfectedwith either CASQ1 or eGFP, and in WT fibers. In contrast, thepeak amplitude of the Ca2� transient induced by a singleelectrical pulse showed a significant increase upon CASQ1reexpression (Fig. 8A). Mean values of fluorescence ratio(amplitude) were as follows: WT � eGFP: 0.635 � 0.043,CASQ1-null � eGFP: 0.466 � 0.024, CASQ1-null � CASQ1:0.566 � 0.031. Moreover, the peak mean value of CASQ1-expressing fibers was not significantly different (P � 0.36512)from that of mock-transfected WT fibers, further suggestingthat exogenous CASQ1 was able to rescue peak amplitude tovalues comparable to those of WT fibers.

Comparison of the Ca2� transients induced by high frequency(60 Hz) stimulation trains showed that, upon transfection ofCASQ1, the decline of fura-2 signals during the stimulation,typically observed in CASQ1-null fibers (see Fig. 4), was signif-icantly reduced (Fig. 8, B and C). CASQ1-transfected musclefibers of CASQ1-null mice were able to sustain high [Ca2�]c

during tetanus, whereas eGFP-transfected fibers (control) werenot (Fig. 8B). The global cytoplasmic Ca2� was also significantly

Fig. 7. Ultrastructural appearance of TC in FDBfibers upon expression of exogenous CASQ1.A–C: in CASQ1-null FDB muscles electropo-rated with pCMS-eGFP-CASQ1, the averagewidth of TC (dashed lines in B) of transfectedfibers was increased (see Fig. 3 for more de-tails), and the TC lumen was filled by an elec-tron dense matrix (arrow in B). TC presenting awider lumen were in the form of either triads ormultilayer junctions. Not all fibers presentedenlarged TC (C) and were classified as not-transfected fibers on the account of the observa-tion that their CRUs were identical to those ofCASQ1-null fibers (compare with Fig. 2F).D: fiber in which TC are extremely dilated (*):these fibers were classified as overtransfected.Enlargement in E shows how these TC are indeedcompletely filled with electron dense material.F: lack of rescue in CASQ1-null FDB fibers elec-troporated with pCMS-eGFP. Bars: A and F � 0.2�m; B, C, and E � 0.1 �m; D � 0.5 �m.



at Biblioteca P

olo Di C



ownloaded from


increased upon CASQ1 reexpression, as judged by the areaincrease under the tracings. Responses to the first (t0) and last (t2)stimuli of the train were also increased upon CASQ1 reexpres-sion, compared with control fibers (Fig. 8C).

DISCUSSION

The aim of the present study was to assess the rescue of thecomplex CASQ1-null phenotype attempted by acute in vivoreintroduction of exogenous CASQ1. Figure 9 summarizes themain results of the present study. Assessment of the rescuedphenotype, 7 days after conspicuous expression of CASQ1 (upto 61% of the WT complement), was carried out at themolecular, functional, and ultrastructural levels. The functional

rescue, as judged by Ca2� release and storage properties, wasvirtually complete, the ultrastructural changes were substan-tially, even if not completely, reversed, whereas the expressionof SR proteins, including CASQ1 molecular partners such asTR and intraluminal Ca2� binding proteins, was not signifi-cantly modified.

Phenotype of CASQ1-Null muscle Fibers of FDB:Functional, Ultrastructural, and Molecular Characteristics

In agreement with previously published observations onCASQ1-null fibers, average amplitude of the cytosolic freeCa2� induced by low frequency stimulation was reduced andbasal resting cytosolic Ca2� was unchanged. Interestingly,

Fig. 8. Depolarization-induced Ca2� transients in single FDB muscle fibers after CASQ1 reintroduction. A: low frequency stimulation. Mean values (�SE) offluorescence ratio obtained from fura-2 signals in single transfected fibers: WT � eGFP fibers (n � 15), CASQ1-null � eGFP fibers (n � 41), and CASQ1-null �CASQ1 fibers (n � 28). Basal fluorescence was not different among the three groups. Peak amplitude values of CASQ1 reexpressing muscle fibers were higherthan those of CASQ1-null control-transfected muscle fibers, and comparable to those of WT mock-transfected muscle fibers (P � 0.001 by ANOVA). B andC: high frequency stimulation. B: representative tracings of Ca2� transients belonging to the 3 groups: WT � eGFP fibers (n � 15), CASQ1-null � eGFP fibers(n � 37), and CASQ1-null � CASQ1 fibers (n � 27). C: quantitative analysis of fura-2 tracings with respect to area and amplitude calculated as described inFig. 4. Area mean values for CASQ1-null � eGFP fibers were smaller than those for WT � eGFP and CASQ1-null � CASQ1 fibers. Only data obtained withCASQ1-null control transfected muscle fibers were significantly different from all other data [area ANOVA, P � 0.001; amplitude t(0), P � 0.05; and amplitudet(2), P � 0.001]. Green fibers displaying extremely high eGFP fluorescence were not used in functional studies, since overexpression of CASQ1 might beassociated to overswollen TC, as shown in Fig 7D, which were not considered even in measurement of TC width (see Fig. 3). Statistical analysis was performedusing a one-way ANOVA test: bars identified with the same symbol exhibit significant differences. Mean values of the initial peak (t0) were as follows:WT � eGFP: 0.677 � 0.050; CASQ1-null � eGFP: 0.594 � 0.033; and CASQ1-null � CASQ1: 0.620 � 0.042; and mean values of the final peak (t2) wereas follows: WT � eGFP: 0.558 � 0.055; CASQ1-null � eGFP: 0.323 � 0.002; and CASQ1-null � CASQ1: 0.491 � 0054.



at Biblioteca P

olo Di C



ownloaded from


CASQ1-null FDB fibers were unable to sustain high level ofcytosolic Ca2� during a tetanus (Fig. 4) in keeping withanalogous results obtained in fibers lacking both CASQ1 and 2(5, 36). This represents a likely indication that, in CASQ1-nullFDB fibers, the amount of Ca2� releasable is reduced inrelation to a decreased content of SR and that contribution ofCASQ2 is not relevant. The functional impact on Ca2� ho-meostasis might derive from either CASQ1 absence and/oradaptive molecular and structural alterations.

Structurally, FDB fibers of CASQ1-null mice exhibitedmodifications intermediate between those previously observedin soleus and EDL (35). For example, multilayered junctions,mostly found in fast glycolytic fibers, which are abundant inEDL (70%) but scarce in soleus, were present in FDB but inlower percentage (35%). Since FDB has a fiber type compo-sition intermediate between EDL and soleus (4, 9, 17), anintermediate phenotype can be expected in FDB fibers. Thereasons for the different impact of CASQ1 ablation in relationto fiber type composition are not known; plausible explanatoryhypotheses contemplate the complete lack of CASQ2 in thesefibers; different content of cytoplasmic Ca2� binding proteinssuch as parvalbumin, different molecular ratios of CASQ1,SERCA, and RYR1; and specific kinetics of Ca2� fluxes inslow-twitch vs. fast-twitch fibers (32, 42).

In total homogenates from FDB, expression of some SRproteins (CASQ2, TR, sarcalumenin) was severely reduced inCASQ1-null mice, whereas expression of JCT was not signif-icantly increased (Fig. 1). Franzini-Armstrong (13) recentlyreviewed the issue of the jSR ultrastructure in muscle fiberseither overexpressing or lacking CASQ, TR, and JCT andsuggested that the overall SR architecture is influenced by themolecular ratios of these three SR proteins. Our results on FDBof CASQ1-null mice support such view and add a few inter-esting clues: expression of TR 95 is directly correlated withthat of CASQ1, whereas JCT expression is independent ofCASQ1 and could correlate with that of RYR1; therefore, TRand CASQ1 have a critical role in maintaining the integrity ofTC, whereas JCT has minor structural effects also in keepingwith heart models of TR-null mice and junctin-null mice (47,6, 52).

CASQ1 Controls Morphology of SR Terminal Cisternae

Acute expression of CASQ1 in CASQ1-null FDB was at-tained by cDNA electrotransfer of pCMS-eGFP-CASQ1: tim-ing of expression (7 days) appears adequate for reversion ofmany, even if not all, ultrastructural and functional changes(7). Expression of exogenous CASQ1 in CASQ1-null fibersresults in correct protein targeting to jSR, in proximity ofCRUs. This is proved by two independent sets of data: immu-nofluorescence showing correct positioning of CASQ1-posi-tive foci (Fig. 6) and EM revealing accumulation of an electrondense matrix in the lumen of TC, whose size increases signif-icantly (Figs. 3 and 7). Since control fibers electroporated withpCMS-eGFP did not show any sign of morphology rescue, theplausible interpretation is that exogenous expression ofCASQ1 restored electron dense content of TC and enlarged TClumen and that the overall process is specifically CASQ1dependent. Since the TC width of transfected fibers was similarto that of WT fibers, our interpretation is that exogenousexpression of CASQ1 restored TC width. Intracellular target-ing and concentration of CASQ1 to jSR, enlargement of TClumen, and sizeable reduction of multilayered junctions oc-curred without rescue of TR content and without major alter-ations in SR protein composition (Fig. 5). These findings implythat morphology of TC is finely controlled by the amount ofCASQ1, supported also by the fact that overexpression ofCASQ1 results in overswelling of TC (Fig. 7, D and E), aneffect identical to that described in cardiomyocytes overex-pressing CASQ2 (19). Moreover, these findings indirectlysupport the oligomerization model for CASQ1 segregation toTC, regardless of TR, one of the so-called anchoring proteins(33, 29).

The quantitative analysis of density of abnormal, multilay-ered junctions indicates that, upon expression of CASQ1, therewas a trend for regression and suggests that CASQ1 also playsa direct role in plasticity of the entire sarcotubular system and,therefore, could be involved in pathological conditions, such asdenervation, and in knock-out murine models of ER/SR resi-dent proteins (49, 34). Fast formation and regression of mul-tiple CRUs have been observed after denervation and reinner-

Fig. 9. Cartoon summarizing the rescue of Ca2� transients and TC structure in CRUs of CASQ1-null mice upon expression of exogenous CASQ1. A andB: functional and structural changes caused by ablation of CASQ1 in FDB muscles (see also Ref. 35): compared with WT, CRUs lacking CASQ1 present anarrower TC profile (pink) are often longitudinally oriented, and formed by multiple elements. CASQ1-null fibers also show reduced Ca2� transients (orange)and are unable to sustain prolonged Ca2� release during a high frequency stimulation. C: reexpression of exogenous CASQ1 in CASQ1-null FDB fibers resultedin 1) rescue of the TC width and electron dense content; 2) decreased frequency of multilayered junctions compared with CASQ1-null FDB fibers; and 3)significant rescue of Ca2� transients amplitude and ability to sustain prolonged Ca2� release during a tetanic stimulation. TT, transverse-tubule.



at Biblioteca P

olo Di C



ownloaded from


vation in rats, respectively (49). Since restoration of normalCRUs was incomplete, it would be interesting to assesswhether longer times of expression were needed to restorecompletely the morphology of CRUs. Moreover, it is notknown whether there is a direct correlation between the num-ber of remaining abnormal triads and the amount of reex-pressed CASQ1.

Longitudinal CRUs are still present after CASQ1 reintro-duction. In this respect, it is known that, during normal,postnatal muscle development, TC orientation is completedmuch later than the attainment of the adult CASQ1 comple-ment (16, 51). Moreover, in TR-null mice, where oblique orlongitudinally oriented triads are described, JCT is also down-regulated: thus, TR and JCT are possible candidates for thecontrol of CRUs orientation. Among JCT and the different TRisoforms, our data suggest that TR 95, decreased in CASQ1-null mice (Fig. 1) and not rescued upon CASQ1 reintroduction(Fig. 5), might play such a role. Interestingly, TR 51 is notdownregulated in CASQ1-null muscles, in agreement with adifferent role of TR 51 previously suggested yet still elusive(27, 43).

CASQ1 Controls Ca2� Homeostasis

The recovery of Ca2� storage and release properties afterCASQ1 reintroduction was virtually complete (Fig. 8). In FDBfrom CASQ1-null muscles, detailed investigations of the SRprotein profile (Fig. 1) revealed marked decrease of sarcalumenin,SERCA1, CASQ2, and TR 95, all proteins reported to modulateSR refilling. Consequently, the inability to store Ca2� within theSR might be due to the combination of reduced intra-luminalCa2� buffer capacity (absence of CASQ1 and deficiency ofsarcalumenin), slowed Ca2� uptake rate (deficiency of SERCA1),and decreased Ca2� uptake capacity (rapid inactivation ofSERCA1 by high intra-luminal Ca2� concentration). However,since SERCA1, sarcalumenin, CASQ2 and TR 95 content did notincrease significantly in CASQ1-transfected muscles, the recoveryof Ca2� storage recorded in our experiments (Fig. 8) was directlydetermined only by accumulation of exogenous CASQ1 in TC,which increases SR Ca2� capacity and facilitates Ca2� storerefilling via SERCA1. These results support the view that CASQ1alone is sufficient to provide the amount of calcium required for aprolonged tetanic contraction.

Conclusions

This study defines some of the functional and structural rolesof CASQ1 in vivo. Reconstitution of the CASQ1 complementhas been successfully accomplished for the first time in FDBmuscle fibers of CASQ1-null mice. Restoration of the Ca2�

storage capacity allows CASQ1-transfected muscle fibers toeffectively cope with either single electrical pulses or pro-longed high frequency train of pulses, preventing SR depletion.This is unambiguous evidence that the ability to store Ca2� isdirectly dependent on CASQ1 and that proper coordination ofCa2� reuptake and Ca2� release also requires CASQ1. Inaddition, our data also confirm the crucial role of CASQ1 indetermining shape and size of TC in skeletal muscle fibers.

ACKNOWLEDGMENTS

We thank E. Rios for providing the pCMS-eGFP-CASQ1 and for sharingthe enthusiasm for the experimental model represented by the in vivo electro-poration of skeletal muscles.

GRANTS

This study was supported by Research Grant GGP08153 from the ItalianTelethon Foundation to F. Protasi, C. Reggiani, and A. Nori and MIUR-PRINto P. Volpe and F. Protasi.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: M.T., M.C., C.P., C.R., P.V., F.P., and A.N. concep-tion and design of research; M.T., M.C., C.P., M.D., F.P., and A.N. performedexperiments; M.T., M.C., C.P., C.R., F.P., and A.N. analyzed data; M.T., M.C.,C.P., C.R., P.V., F.P., and A.N. interpreted results of experiments; M.T., M.C.,C.P., and A.N. prepared Figs.; M.T., C.R., P.V., F.P., and A.N. draftedmanuscript; M.T., C.R., P.V., F.P., and A.N. edited and revised manuscript;M.T., M.C., C.P., M.D., C.R., P.V., F.P., and A.N. approved final version ofmanuscript.

REFERENCES

1. Arai M, Otsu K, MacLennan DH, Periasamy M. Regulation of sarco-plasmic reticulum gene expression during cardiac and skeletal muscledevelopment. Am J Physiol Cell Physiol 262: C614–C620, 1992.

2. Beard NA, Sakowska MM, Dulhunty AF, Laver DR. Calsequestrin isan inhibitor of skeletal muscle ryanodine receptor calcium release chan-nels. Biophys J 82: 310–320, 2002.

3. Beard NA, Wei L, Dulhunty AF. Ca(2�) signaling in striated muscle:the elusive roles of triadin, junctin, and calsequestrin. Eur Biophys J 39:27–36, 2009.

4. Calderon JC, Bolanos P, Caputo C. Myosin heavy chain isoformcomposition and Ca(2�) transients in fibres from enzymatically dissoci-ated murine soleus and extensor digitorum longus muscles. J Physiol 588:267–279, 2010.

5. Canato M, Scorzeto M, Giacomello M, Protasi F, Reggiani C, StienenGJ. Massive alterations of sarcoplasmic reticulum free calcium in skeletalmuscle fibers lacking calsequestrin revealed by a genetically encodedprobe. Proc Natl Acad Sci USA 107: 22326–22331, 2010.

6. Chopra N, Yang T, Asghari P, Moore ED, Huke S, Akin B, CattolicaRA, Perez CF, Hlaing T, Knollmann-Ritschel BE, Jones LR, PessahIN, Allen PD, Franzini-Armstrong C, Knollmann BC. Ablation oftriadin causes loss of cardiac Ca2� release units, impaired excitation-contraction coupling, and cardiac arrhythmias. Proc Natl Acad Sci USA106: 7636–7641, 2009.

7. Cusimano V, Pampinella F, Giacomello E, Sorrentino V. Assembly anddynamics of proteins of the longitudinal and junctional sarcoplasmicreticulum in skeletal muscle cells. Proc Natl Acad Sci USA 106: 4695–4700, 2009.

8. Dainese M, Quarta M, Lyfenko AD, Paolini C, Canato M, Reggiani C,Dirksen RT, Protasi F. Anesthetic- and heat-induced sudden death incalsequestrin-1-knockout mice. FASEB J 23: 1710–1720, 2009.

9. Danieli-Betto D, Esposito A, Germinario E, Sandona D, Martinello T,Jakubiec-Puka A, Biral D, Betto R. Deficiency of alpha-sarcoglycandifferently affects fast- and slow-twitch skeletal muscles. Am J PhysiolRegul Integr Comp Physiol 289: R1328–R1337, 2005.

10. Defranchi E, Bonaccurso E, Tedesco M, Canato M, Pavan E, RaiteriR, Reggiani C. Imaging and elasticity measurements of the sarcolemma offully differentiated skeletal muscle fibres. Microsc Res Tech 67: 27–35,2005.

11. DiFranco M, Neco P, Capote J, Meera P, Vergara JL. Quantitativeevaluation of mammalian skeletal muscle as a heterologous protein ex-pression system. Protein Expr Purif 47: 281–288, 2006.

12. Donoghue P, Ribaric S, Moran B, Cebasek V, Erzen I, Ohlendieck K.Early effects of denervation on Ca(2�)-handling proteins in skeletalmuscle. Int J Mol Med 13: 767–772, 2004.

13. Franzini-Armstrong C. Architecture and regulation of the Ca2� deliverysystem in muscle cells. Appl Physiol Nutr Metab 34: 323–327, 2009.

14. Franzini-Armstrong C, Kenney LJ, Varriano-Marston E. The struc-ture of calsequestrin in triads of vertebrate skeletal muscle: a deep-etchstudy. J Cell Biol 105: 49–56, 1987.

15. Froemming GR, Murray BE, Harmon S, Pette D, Ohlendieck K.Comparative analysis of the isoform expression pattern of Ca(2�)-regu-latory membrane proteins in fast-twitch, slow-twitch, cardiac, neonatal



at Biblioteca P

olo Di C



ownloaded from


and chronic low-frequency stimulated muscle fibers. Biochim BiophysActa 1466: 151–168, 2000.

16. Froemming GR, Ohlendieck K. Oligomerisation of Ca2�-regulatorymembrane components involved in the excitation-contraction-relaxationcycle during postnatal development of rabbit skeletal muscle. BiochimBiophys Acta 1387: 226–238, 1998.

17. Gonzalez E, Messi ML, Zheng Z, Delbono O. Insulin-like growthfactor-1 prevents age-related decrease in specific force and intracellularCa2� in single intact muscle fibres from transgenic mice. J Physiol 552:833–844, 2003.

18. Hajnoczky G, Csordas G, Madesh M, Pacher P. The machinery of localCa2� signalling between sarco-endoplasmic reticulum and mitochondria. JPhysiol 529: 69–81, 2000.

19. Jones LR, Suzuki YJ, Wang W, Kobayashi YM, Ramesh V, Franzini-Armstrong C, Cleemann L, Morad M. Regulation of Ca2� signaling intransgenic mouse cardiac myocytes overexpressing calsequestrin. J ClinInvest 101: 1385–1393, 1998.

20. Knollmann BC, Chopra N, Hlaing T, Akin B, Yang T, Ettensohn K,Knollmann BE, Horton KD, Weissman NJ, Holinstat I, Zhang W,Roden DM, Jones LR, Franzini-Armstrong C, Pfeifer K. Casq2 dele-tion causes sarcoplasmic reticulum volume increase, premature Ca2�

release, and catecholaminergic polymorphic ventricular tachycardia. JClin Invest 116: 2510–2520, 2006.

21. Laemmli UK. Cleavage of structural proteins during the assembly of thehead of bacteriophage T4. Nature 227: 680–685, 1970.

22. Launikonis BS, Barnes M, Stephenson DG. Identification of the cou-pling between skeletal muscle store-operated Ca2� entry and the inositoltrisphosphate receptor. Proc Natl Acad Sci USA 100: 2941–2944, 2003.

23. Launikonis BS, Murphy RM, Edwards JN. Toward the roles of store-operated Ca2� entry in skeletal muscle. Pflügers Arch 460: 813–823, 2010.

24. Lavery GG, Walker EA, Turan N, Rogoff D, Ryder JW, Shelton JM,Richardson JA, Falciani F, White PC, Stewart PM, Parker KL,McMillan DR. Deletion of hexose-6-phosphate dehydrogenase activatesthe unfolded protein response pathway and induces skeletal myopathy. JBiol Chem 283: 8453–8461, 2008.

25. Lehotsky J, Bezakova G, Kaplan P, Raeymaekers L. Distribution ofCa(2�)-modulating proteins in sarcoplasmic reticulum membranes afterdenervation. Gen Physiol Biophys 12: 339–348, 1993.

26. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measure-ment with the Folin phenol reagent. J Biol Chem 193: 265–275, 1951.

27. Marty I, Thevenon D, Scotto C, Groh S, Sainnier S, Robert M,Grunwald D, Villaz M. Cloning and characterization of a new isoform ofskeletal muscle triadin. J Biol Chem 275: 8206–8212, 2000.

28. Meissner G, Wang Y, Xu L, Eu JP. Silencing genes of sarcoplasmicreticulum proteins clarifies their roles in excitation-contraction coupling. JPhysiol 587: 3089–3090, 2009.

29. Milstein ML, Houle TD, Cala SE. Calsequestrin isoforms localize todifferent ER subcompartments: evidence for polymer and heteropolymer-dependent localization. Exp Cell Res 315: 523–534, 2009.

30. Mollica JP, Oakhill JS, Lamb GD, Murphy RM. Are genuine changesin protein expression being overlooked? Reassessing Western blotting.Anal Biochem 386: 270–275, 2009.

31. Morrissey JH. Silver stain for proteins in polyacrylamide gels: a modifiedprocedure with enhanced uniform sensitivity. Anal Biochem 117: 307–310, 1981.

32. Murphy RM, Larkins NT, Mollica JP, Beard NA, Lamb GD. Calse-questrin content and SERCA determine normal and maximal Ca2� storagelevels in sarcoplasmic reticulum of fast- and slow-twitch fibres of rat. JPhysiol 587: 443–460, 2009.

33. Nori A, Valle G, Bortoloso E, Turcato F, Volpe P. Calsequestrintargeting to sarcoplasmic reticulum of skeletal muscle fibers. Am J PhysiolCell Physiol 291: C245–C253, 2006.

34. Oddoux S, Brocard J, Schweitzer A, Szentesi P, Giannesini B, Bro-card J, Faure J, Pernet-Gallay K, Bendahan D, Lunardi J, CsernochL, Marty I. Triadin deletion induces impaired skeletal muscle function. JBiol Chem 284: 34918–34929, 2009.

35. Paolini C, Quarta M, Nori A, Boncompagni S, Canato M, Volpe P,Allen PD, Reggiani C, Protasi F. Reorganized stores and impaired

calcium handling in skeletal muscle of mice lacking calsequestrin-1. JPhysiol 583: 767–784, 2007.

36. Paolini C, Quarta M, D’Onofrio L, Reggiani C, Protasi F. Differentialeffect of calsequestrin ablation on structure and function of fast and slowskeletal muscle fibers. J Biomed Biotechnol 2011: 634075, 2011.

37. Park H, Park IY, Kim E, Youn B, Fields K, Dunker AK, Kang C.Comparing skeletal and cardiac calsequestrin structures and their calciumbinding: a proposed mechanism for coupled calcium binding and proteinpolymerization. J Biol Chem 279: 18026–18033, 2004.

38. Park H, Wu S, Dunker AK, Kang C. Polymerization of calsequestrin.Implications for Ca2� regulation. J Biol Chem 18: 16176–16182, 2003.

39. Protasi F, Paolini C, Dainese M. Calsequestrin-1: a new candidate genefor malignant hyperthermia and exertional/environmental heat stroke. JPhysiol 587: 3095–3100, 2009.

40. Qin J, Valle G, Nani A, Chen H, Ramos-Franco J, Nori A, Volpe P,Fill M. Ryanodine receptor luminal Ca2� regulation: swapping calseques-trin and channel isoforms. Biophys J 97: 1961–1970, 2009.

41. Ravenscroft G, Nowak KJ, Jackaman C, Clement S, Lyons MA,Gallagher S, Bakker AJ, Laing NG. Dissociated flexor digitorum brevismyofiber culture system–a more mature muscle culture system. Cell MotilCytoskeleton 64: 727–738, 2007.

42. Reggiani C, te Kronnie T. RyR isoforms and fibre type-specific expres-sion of proteins controlling intracellular calcium concentration in skeletalmuscles. J Muscle Res Cell Motil 27: 327–335, 2006.

43. Rezgui SS, Vassilopoulos S, Brocard J, Platel JC, Bouron A, ArnoultC, Oddoux S, Garcia L, De Waard M, Marty I. Triadin (Trisk 95)overexpression blocks excitation-contraction coupling in rat skeletal myo-tubes. J Biol Chem 280: 39302–39308, 2005.

44. Rossi AE, Dirksen RT. Sarcoplasmic reticulum: the dynamic calciumgovernor of muscle. Muscle Nerve 33: 715–731, 2006.

45. Royer L, Sztretye M, Manno C, Pouvreau S, Zhou J, Knollmann BC,Protasi F, Allen PD, Rios E. Paradoxical buffering of calcium bycalsequestrin demonstrated for the calcium store of skeletal muscle. J GenPhysiol 136: 325–338, 2010.

46. Rudolf R, Mongillo M, Magalhaes PJ, Pozzan T. In vivo monitoring ofCa(2�) uptake into mitochondria of mouse skeletal muscle during con-traction. J Cell Biol 166: 527–536, 2004.

47. Shen X, Franzini-Armstrong C, Lopez JR, Jones LR, Kobayashi YM,Wang Y, Kerrick WG, Caswell AH, Potter JD, Miller T, Allen PD,Perez CF. Triadins modulate intracellular Ca(2�) homeostasis but are notessential for excitation-contraction coupling in skeletal muscle. J BiolChem 282: 37864–37874, 2007.

48. Shin DW, Pan Z, Kim EK, Lee JM, Bhat MB, Parness J, Kim DH, MaJ. A retrograde signal from calsequestrin for the regulation of store-operated Ca2�

entry in skeletal muscle. J Biol Chem 278: 3286–3292, 2003.49. Takekura H, Tamaki H, Nishizawa T, Kasuga N. Plasticity of the

transverse tubules following denervation and subsequent reinnervation in rat slowand fast muscle fibres. J Muscle Res Cell Motil 24: 439–451, 2003.

50. Vassilopoulos S, Thevenon D, Rezgui SS, Brocard J, Chapel A,Lacampagne A, Lunardi J, Dewaard M, Marty I. Triadins are nottriad-specific proteins: two new skeletal muscle triadins possibly involvedin the architecture of sarcoplasmic reticulum. J Biol Chem 280: 28601–28609, 2005.

51. Villa A, Podini P, Nori A, Panzeri MC, Martini A, Meldolesi J, VolpeP. The endoplasmic reticulum-sarcoplasmic reticulum connection. II.Postnatal differentiation of the sarcoplasmic reticulum in skeletal musclefibers. Exp Cell Res 209: 140–148, 1993.

52. Yuan Q, Fan GC, Dong M, Altschafl B, Diwan A, Ren X, Hahn HH,Zhao W, Waggoner JR, Jones LR, Jones WK, Bers DM, Dorn GW II,Wang HS, Valdivia HH, Chu G, Kranias EG. Sarcoplasmic reticulumcalcium overloading in junctin deficiency enhances cardiac contractilitybut increases ventricular automaticity. Circulation 115: 300–309, 2007.

53. Zhao X, Min CK, Ko JK, Parness J, Kim do H, Weisleder N, Ma J.Increased store-operated Ca2� entry in skeletal muscle with reducedcalsequestrin-1 expression. Biophys J 99: 1556–1564, 2010.

54. Zorzato F, Volpe P, Damiani E, Quaglino D, Jr, Margreth A. Terminalcisternae of denervated rabbit skeletal muscle: alterations of functionalproperties of Ca2� release channels. Am J Physiol Cell Physiol 257:C504–C511, 1989.



at Biblioteca P

olo Di C



ownloaded from


calsequestrin rescues function an strutture of ca . .. in skeletal muscles

Documents