Cardiac-specific Overexpression of Mouse Cardiac Calsequestrin Is Associated with Depressed Cardiovascular Function and Hypertrophy in Transgenic Mice*

Calsequestrin is a high capacity Ca2+-binding protein in the sarcoplasmic reticulum (SR) lumen. To elucidate the functional role of calsequestrin in vivo, transgenic mice were generated that overexpressed mouse cardiac calsequestrin in the heart. Overexpression (20-fold) of calsequestrin was associated with cardiac hypertrophy and induction of a fetal gene expression program. Isolated transgenic cardiomyocytes exhibited diminished shortening fraction (46%), shortening rate (60%), and relengthening rate (60%). The Ca2+ transient amplitude was also depressed (45%), although the SR Ca2+storage capacity was augmented, as suggested by caffeine application studies. These alterations were associated with a decrease in L-type Ca2+ current density and prolongation of this channel’s inactivation kinetics without changes in Na+-Ca2+ exchanger current density. Furthermore, there were increases in protein levels of SR Ca2+-ATPase, phospholamban, and calreticulin and decreases in FKBP12, without alterations in ryanodine receptor, junctin, and triadin levels in transgenic hearts. Left ventricular function analysis in Langendorff perfused hearts and closed-chest anesthetized mice also indicated depressed rates of contraction and relaxation of transgenic hearts. These findings suggest that calsequestrin overexpression is associated with increases in SR Ca2+ capacity, but decreases in Ca2+-induced SR Ca2+ release, leading to depressed contractility in the mammalian heart.

In cardiac excitation-contraction coupling, the sarcoplasmic reticulum (SR) 1 plays an essential role in the regulation of the cytosolic free Ca 2ϩ concentration. There are three major functions of the SR: (a) Ca 2ϩ uptake from the cytosol into the SR lumen, resulting in muscle relaxation; (b) Ca 2ϩ storage in the SR lumen; and (c) Ca 2ϩ release from the SR into the cytosol, resulting in muscle contraction. The main SR proteins responsible for these functions are the Ca 2ϩ transport ATPase, the Ca 2ϩ storage protein calsequestrin (1), and the Ca 2ϩ release channel or ryanodine receptor, respectively. Phospholamban is another SR protein that plays a crucial role in the modulation of myocardial contractility and relaxation (2).
Recent studies with isolated human myocardial preparations suggested that the impaired Ca 2ϩ handling of the SR may be an important subcellular mechanism contributing to the depressed contractility in heart failure (3)(4)(5). However, it is still controversial whether the expression levels of SR Ca 2ϩ -handling proteins are altered in failing human hearts. Some groups have reported alterations in mRNA levels and/or protein expression levels of SR Ca 2ϩ -ATPase, phospholamban, and the SR Ca 2ϩ release channel in human heart failure (3,6,7). However, other groups observed no significant changes in the expression levels of these proteins in hearts with end-stage heart failure (8,9). Interestingly, in all studies, the calsequestrin expression levels appeared to be unaltered. These findings suggest that calsequestrin expression is under specific and rigid regulation in cardiac muscle. However, the physiological significance of the apparent tight control of calsequestrin expression is not known.
There are currently two different calsequestrin genes, encoding for the cardiac and fast-skeletal muscle products. The cardiac isoform of calsequestrin is highly conserved among species (1), and it is the only isoform expressed during cardiac development (10,11). Several laboratories (for review, see Ref. 12) have implicated calsequestrin, which binds Ca 2ϩ with high capacity and low affinity, as a key player in the regulation of SR Ca 2ϩ release. Actually, calsequestrin appears to be physically connected to the ryanodine receptor (13), and formation of a stable complex among calsequestrin, the ryanodine receptor, junctin, and triadin at the junctional SR has been proposed to be important for the operation of Ca 2ϩ release during cardiac muscle excitation-contraction coupling (14). However, it is not currently clear how calsequestrin buffers luminal Ca 2ϩ in vivo, what is the buffering capacity of this protein in intact cells, and what are the direct or indirect effects of altered calsequestrin levels on SR Ca 2ϩ load, ryanodine receptor gating, and myocardial contractility. Furthermore, the role of calsequestrin in the decreased amplitude of the intracellular Ca 2ϩ transient in failing hearts, which may reflect a reduction in the amount of releasable Ca 2ϩ from the SR (5,15,16), is not clear. Thus, this study was designed to elucidate the functional role of calsequestrin in cardiac physiology and/or pathophysiology by altering the expression levels of this protein in vivo using transgenesis. Cardiac-specific overexpression of calsequestrin resulted in increased SR Ca 2ϩ storage capacity, but this SR Ca 2ϩ was not available for release during excitation-contraction coupling, leading to a depressed amplitude of the Ca 2ϩ transient in cardiomyocytes and depressed contractile parameters assessed in isolated myocytes, perfused hearts, and intact animals. This cardiac phenotype was accompanied by alterations in the expression levels of several key proteins and induction of a fetal gene program leading to cardiac hypertrophy.

EXPERIMENTAL PROCEDURES
The handling and maintenance of the animals in this study were approved by the ethics committee of the University of Cincinnati. 8 -13week-old mice of either sex were used for the following studies.
Isolation of cDNA Encoding Mouse Cardiac Calsequestrin-The mouse cardiac 5Ј-stretch plus cDNA library (ML5002a; CLONTECH) was screened using a 32 P-labeled polymerase chain reaction product by a pUC/M13 forward primer (24-mer; Promega), a reverse primer corresponding to 60 nucleotides in the rabbit cardiac calsequestrin cDNA sequence (17,18), and the rabbit cardiac calsequestrin cDNA as a template. The cDNA was excised from the bacteriophage DNA with NotI and subcloned into the pBluescript SK(Ϫ) vector (Stratagene), and DNA sequence analysis was performed by an automated DNA sequencer (Model 373A, Perkin-Elmer Applied Biosystems).
Generation of Transgenic Mice-A 2180-bp fragment encompassing 60 bp of 5Ј-untranslated region, the entire calsequestrin coding region, and 863 bp of 3Ј-untranslated region was generated using polymerase chain reaction methodology (flanked with SalI and EcoRV sites). This fragment was linked to the 3Ј-end of the mouse ␣-myosin heavy chain promoter and to the blunted 5Ј-end of the human growth hormone polyadenylation signal sequence (19) in pBluescript SK(ϩ). The entire 8.3-kilobase NotI fragment was excised from the plasmid sequence and used for microinjection of fertilized mouse eggs (FVB/N). To identify transgenic mice, polymerase chain reaction analysis of tail genomic DNA was carried out using a forward primer corresponding to the 5Ј-end of the mouse ␣-myosin heavy chain gene sequence (5Ј-CACAT-AGAAGCCTAGCCCACAC-3Ј) and a reverse primer corresponding to the 5Ј-end of the mouse cardiac calsequestrin cDNA sequence (5Ј-TTC-TTCTCAGAAAGGCTGACC-3Ј).
Dot-blot Analysis-Dot-blot analysis of total RNA from cardiac ventricles was performed as described previously (21).
Quantitative Immunoblotting-Mouse hearts were homogenized at 4°C in a buffer containing 10 mM imidazole (pH 7.0), 300 mM sucrose, 1 mM dithiothreitol, 1 mM sodium metabisulfite, 0.3 mM phenylmethylsulfonyl fluoride, 5 g/ml leupeptin, 2 mM EDTA, 10 g/ml soybean trypsin inhibitor type II-S (Sigma), and 7 g/ml pepstatin A. After solubilization of the homogenates, SDS-polyacrylamide gel electrophoresis and immunoblotting were performed as described previously (18). For the quantitation of phospholamban, the solubilized samples were boiled for 5 min to fully dissociate the pentameric form of phospholamban into monomers. Binding of the primary antibody was detected by peroxidase-conjugated secondary antibodies and ECL (Amersham Pharmacia Biotech). Protein concentration was determined by the Bradford method (Bio-Rad) with bovine serum albumin as a standard.
Isolated Cardiomyocytes-Left ventricular tissue was enzymatically digested to isolate individual myocytes, and measurements of myocyte mechanics and Ca 2ϩ transients were performed at a pacing rate of 0.25 Hz as described previously (22). Viability of myocytes from wild-type and transgenic mice after the isolation was ϳ80 and 40%, respectively. Rod-shaped viable cells were selected for the experiments. To monitor intracellular free Ca 2ϩ transients, the cells were loaded with 7 M Fura-2/AM, and fluorescent signals were measured (22). Subsequently, the cells were perfused for 1 min with 0 Na ϩ /0 Ca 2ϩ buffer containing 132 mM LiCl, 4.8 mM KCl, 1.2 mM MgCl 2 , 5 mM glucose, 10 mM HEPES, and 10 mM EGTA (pH 7.3). Then, Ca 2ϩ transients induced by caffeine (10 mM) were obtained in the absence of Na ϩ and Ca 2ϩ to determine the SR Ca 2ϩ capacity in myocytes (23).
Electrophysiology-Whole-cell L-type Ca 2ϩ currents were recorded by applying depolarizing pulses every 10 s from a holding potential of Ϫ50 mV. Cell membrane capacitance was measured using voltage ramps of 0.8 V/s from a holding potential of Ϫ50 mV (24). Na ϩ -Ca 2ϩ exchanger currents were recorded by the method of Kimura et al. (25).
Langendorff Perfusion-Contractile parameters of isolated hearts were determined at 37°C as described previously (26).
In Vivo Left Ventricular Function-Left ventricular contractile parameters were assessed in closed-chest anesthetized mice using a 1.4-French scale Millar catheter (27). The hearts were paced at 300 beats/ min, using a 1-French scale bipolar pacing wire advanced into the right atrium through the jugular vein.
Materials-Generous gifts of materials included rabbit cardiac calsequestrin cDNA from Dr. M. Periasamy (University of Cincinnati), mouse ␣-myosin heavy chain promoter from Dr. J. Robbins (Children's Hospital Medical Center, Cincinnati, OH), and rabbit polyclonal antijunctin affinity-purified antibody (28)  Statistics-Data are presented as mean Ϯ S.E. Comparisons between groups were evaluated using Student's t test, with significance imparted at the p Ͻ 0.05 level.

RESULTS
cDNA Encoding Mouse Cardiac Calsequestrin-An adult mouse cardiac cDNA library was screened using the 5Ј-end of the rabbit cardiac calsequestrin cDNA as a probe (17). Twenty positive colonies were obtained from ϳ20,000 plaques. The longest cDNA clone encoding mouse cardiac calsequestrin was 2275 bp long and included a single open reading frame of 1245 bp. It also contained 129 and 901 bp of 5Ј-and 3Ј-untranslated regions, respectively, but there was no typical polyadenylation signal ( Fig. 1). 2 The deduced amino acid sequence of mouse cardiac calsequestrin indicated a protein composed of 396 amino acid residues and an amino acid homology of 90% to either canine or rabbit cardiac calsequestrin (17,29). There was a cleavable N-terminal signal sequence (19 residues) ( Fig.  1) and a highly acidic C terminus, but mouse cardiac calsequestrin had a few more acidic amino acids compared with canine (5 more) and rabbit (6 more) cardiac calsequestrin (1,20,32).
Cardiac Hypertrophy-The increases in calsequestrin protein expression levels were associated with increases in heart/ body weight ratios, which were 7.8 Ϯ 0.2 mg/g for line 398 (n ϭ 13), 6.7 Ϯ 0.3 mg/g for line 412 (n ϭ 10), and 6.1 Ϯ 0.1 mg/g for line 418 (n ϭ 41) compared with 4.7 Ϯ 0.1 mg/g for the wild types (n ϭ 38). Examination of gross cardiac morphology revealed mild myocyte hypertrophy in the left ventricular free walls and intraventricular septa of transgenic hearts. No other changes were observed in the ventricles and valves of transgenic mouse hearts at low magnification. At higher magnification, however, there was an irregular vacuolization of the left atrial myocytes in lines 398 and 412 (data not shown). Thus, further characterization studies of the transgenic phenotype were carried out using mice from line 418. The increases in heart/body weight ratios were associated with increased expression of fetal genes, particularly the ␤-myosin heavy chain isoform (␤-MHC), atrial natriuretic factor, and ␣-skeletal actin, which have been previously reported to increase in physiologically relevant experimental models of cardiac hypertrophy (30 -32). The relative increases in ␤-MHC, atrial natriuretic factor, and ␣-skeletal actin mRNAs in transgenic mice were 4-, 5-, and 11-fold over the wild types, respectively (Fig. 3).
Indirect Immunofluorescence Microscopy-To determine the subcellular localization of overexpressed calsequestrin in cardiomyocytes, immunostaining and confocal microscopy were performed. Sections of cardiac tissue from transgenic animals exhibited similar striated staining patterns as the wild types (Fig. 4). The somewhat discontinuous, transversely striated staining patterns observed in sections from wild-type ventricle (Fig. 4B) were consistent with previously described distributions of calsequestrin (33). For this study, wild-type and transgenic sections were stained in parallel with the same antibody solutions. Images from the wild-type sections were first collected, and the transgenic cardiac tissue was examined using the same parameters. The increased staining in the calsequestrin-overexpressing tissue was consistent with the quantitative immunoblotting data (Fig. 2). Further examination of the transgenic sections revealed that the staining pattern of transverse striations was predominant (Fig. 4A), although in a few areas, this pattern was somewhat less organized. In control experiments, using fluorescein-conjugated anti-rabbit IgG without prior incubation with the anti-calsequestrin antibody, there was no staining observed in either wild-type or transgenic cardiac tissue (data not shown).
Contractile Parameters in Isolated Cardiomyocytes-To assess the effects of calsequestrin overexpression on cardiac function, the contractile parameters and intracellular Ca 2ϩ transients of Ca 2ϩ -tolerant left ventricular cardiomyocytes from transgenic and age-matched wild-type mice were measured in parallel. Table I indicates that the extent of cardiomyocyte shortening and the maximal rates of shortening (ϩdL/dt) and relengthening (ϪdL/dt) were depressed by 46, 60, and 60%, respectively, in transgenic cardiomyocytes compared with wildtype cardiomyocytes. The decreases in contractile parameters reflected a depression (45%) in the amplitude of the systolic Ca 2ϩ transient ( Fig. 5A and Table I). However, the time for 80% decay of the Ca 2ϩ signal (t 80 ) and the base-line levels of free cytosolic Ca 2ϩ , approximated by the 340/380 nm ratio, were not statistically different between transgenic and wildtype cardiomyocytes. To examine whether overexpression of calsequestrin altered the SR Ca 2ϩ storage capacity, caffeine (10 mM) was added to the media in the absence of extracellular Na ϩ and Ca 2ϩ in order to open the SR Ca 2ϩ release channels and to allow the release of Ca 2ϩ from the SR into the cytosol. Addition of caffeine caused a transient increase in the Ca 2ϩ signals (Fig. 5B). However, the peak amplitude of the caffeine-induced Ca 2ϩ signal in transgenic cardiomyocytes was significantly higher than that in wild-type cardiomyocytes (340/380 nm signal: 5.97 Ϯ 0.27 in transgenic (n ϭ 10) and 1.09 Ϯ 0.17 in wild-type FIG. 2. Quantitative immunoblotting for cardiac calsequestrin. Total cardiac homogenate protein was electrophoresed on an 8% SDS-polyacrylamide gel, electroblotted onto a nitrocellulose membrane, and probed with a calsequestrin-specific antibody followed by a peroxidase-conjugated secondary antibody. The same control homogenate (C) was used to generate a regression line for quantitation of expression levels in wild-type (WT) and transgenic (TG) mouse hearts. In transgenic mouse lines 398, 412, and 418, cardiac calsequestrin levels were increased by 43 Ϯ 5-fold (n ϭ 4), 56 Ϯ 2-fold (n ϭ 4), and 20 Ϯ 2-fold (n ϭ 8) compared with the wild types, respectively.

FIG. 4. Indirect immunofluorescence localization of calsequestrin in calsequestrin-overexpressing transgenic and wild-type mouse ventricular myocytes.
In longitudinal sections from both transgenic (A; line 418) and wild-type (B) ventricles, calsequestrinspecific staining was predominantly observed as transverse striations with a periodicity of ϳ2 m. In the sections from transgenic hearts, there were also areas in which the striated pattern was not as regular (see, for example, the area between the asterisks), indicating some disruption of the normal distribution of terminal cisternae at the level of the Z-disc. Calsequestrin was also apparently concentrated in the perinuclear regions (see, for example, the thick arrows), presumably in the Golgi apparatus. This pattern was also observed in the wild-type sections (see, for example, the thin arrow), but it was not as apparent as in transgenic animals. Bar ϭ 13.4 m.
Characteristics of L-type Ca 2ϩ Channel and Na ϩ -Ca 2ϩ Exchanger Currents-The altered Ca 2ϩ handling in transgenic myocytes prompted us to examine the characteristics of the L-type Ca 2ϩ channel current (I Ca ) in these cells. The membrane capacitance was significantly larger in cells isolated from transgenic (143.8 Ϯ 4.1 pF, n ϭ 72) compared with wild-type (112.5 Ϯ 3.8 pF, n ϭ 37) myocytes, in agreement with the hypertrophic phenotype of calsequestrin-overexpressing mice. Fig. 6A shows typical current-voltage relationships for wildtype and transgenic cardiomyocytes. When peak I Ca was normalized to cell capacitance, the current-voltage relationships were similar between the two groups, but the current density of peak I Ca was significantly reduced in transgenic myocytes. The average values were 5.9 Ϯ 0.3 pA/pF (n ϭ 72) for transgenic myocytes and 9.8 Ϯ 0.4 pA/pF (n ϭ 37) for wild-type myocytes. Furthermore, the current decay during depolarization was significantly slower in transgenic myocytes compared with wildtype myocytes. The time to half-decay (t1 ⁄2 ) at ϩ10 mV, where I Ca reaches a maximal value, was 18.7 Ϯ 1.2 ms in wild-type myocytes (n ϭ 37) and 43.5 Ϯ 2.0 ms in transgenic myocytes (n ϭ 52).
We also examined the magnitude of the Na ϩ -Ca 2ϩ exchanger current by replacing external Na ϩ with Li ϩ to suppress the exchanger activity (Fig. 6B). When the peak outward current magnitude was normalized relative to cell capacitance, the average values of current densities were 0.94 Ϯ 0.10 pA/pF (n ϭ 18) and 0.82 Ϯ 0.10 pA/pF (n ϭ 14) for wild-type and transgenic cells, respectively. This suggests that the functional activity of the Na ϩ -Ca 2ϩ exchanger was not significantly altered in transgenic myocytes.
Expression Levels of Ca 2ϩ -handling Proteins-To determine whether calsequestrin overexpression was associated with any alterations in the expression levels of other SR proteins, the levels of the major Ca 2ϩ -handling proteins were assessed using quantitative immunoblotting. The SR Ca 2ϩ -ATPase and phospholamban protein levels were increased by 57 and 74%, respectively, in calsequestrin-overexpressing mouse hearts (Table II). It is interesting to note that the relative ratio of phospholamban to SR Ca 2ϩ -ATPase was similar between FIG. 5. Representative recordings of the Ca 2؉ transients in Fura-2-loaded cardiomyocytes from wild-type and calsequestrin-overexpressing transgenic mice. A, shown are Ca 2ϩ transients in ventricular myocytes at a pacing rate of 0.25 Hz. Cardiomyocytes from transgenic animals had a reduced Ca 2ϩ transient amplitude. B, subsequent superfusion with caffeine (10 mM, in 0 Na ϩ /0 Ca 2ϩ buffer) inclusion evoked increases in the intracellular Ca 2ϩ concentration in both quiescent wild-type (WT) and transgenic (TG) cardiomyocytes, but the effects were more pronounced in transgenic cells.

FIG. 6. I Ca and Na ؉ -Ca 2؉ exchanger current in wild-type and calsequestrin-overexpressing transgenic ventricular myocytes.
A: upper panels, typical whole-cell L-type Ca 2ϩ currents recorded in wild-type (WT) and transgenic (TG) cells. Shown are traces of currents that were recorded from a holding potential of Ϫ50 mV to the indicated test potentials. Lower panels, the peak current-voltage relationships obtained from these cells. The extracellular Ca 2ϩ concentration was 2 mM. B: Na ϩ -Ca 2ϩ exchanger currents measured from wild-type and transgenic cells. The Na ϩ -Ca 2ϩ exchanger current was estimated as the outward current when the external solution was changed from 150 mM Na ϩ to 150 mM Li ϩ . The traces show continuous recording at a holding potential of Ϫ40 mV. Extracellular and intracellular Ca 2ϩ concentrations were adjusted to 1 and 67 nM, respectively. transgenic and wild-type hearts. Calreticulin, another high capacity Ca 2ϩ -binding protein in the lumina of the sarcoplasmic and endoplasmic reticula, was also increased by 57%, similar to recent findings in an experimental model of pressureoverload cardiac hypertrophy (34). However, the protein expression levels of the ryanodine receptor, junctin, and an ϳ95-kDa triadin isoform were not significantly different in transgenic hearts compared with wild-type hearts. Smaller isoforms of triadin (35) were not detected by the monoclonal antibody MA3-927. Examination of the protein levels of FKBP12 (the 12-kDa FK506-binding protein), which has been shown to regulate the gating properties of cardiac ryanodine receptors (36,37), revealed a 24% decrease in transgenic hearts compared with wild-type hearts. FKBP12.6, an isoform of FKBP12, was barely detected by the polyclonal antibody PA1-026 in mouse hearts.
Contractile Parameters in Langendorff Perfused Hearts-To determine whether the alterations in contractile parameters observed at the cellular level reflected similar alterations at the organ level, isolated hearts from transgenic mice and wild-type littermates were perfused in parallel using the Langendorff mode. The spontaneous heart rates were similar, but the maximal rates of both pressure development (ϩdP/dt) and decline (ϪdP/dt) were significantly depressed in transgenic hearts compared with wild-type hearts (Table III) Contractile Parameters in Closed-chest Mice-An important question is whether the depressed function at the cellular and organ levels would translate into depressed function in vivo or whether compensatory mechanisms in the intact animal may mask the effects of cardiac calsequestrin overexpression. Thus, the effects of calsequestrin overexpression on ventricular function were assessed in intact mice using a Millar catheter that was inserted into the left ventricle via the carotid artery. At matched heart rates (300 beats/min), both ϩdP/dt and ϪdP/dt were significantly depressed in transgenic hearts compared with wild-type hearts (3406 Ϯ 241 versus 5104 Ϯ 375 mm Hg/s for ϩdP/dt (p Ͻ 0.05) and 2656 Ϯ 224 versus 4948 Ϯ 595 mm Hg/s for ϪdP/dt (p Ͻ 0.05; n ϭ 6) for transgenic versus wildtype). DISCUSSION This study demonstrated that overexpression of calsequestrin in the heart resulted in increases in SR Ca 2ϩ storage capacity, but this pool of Ca 2ϩ was not accessible for release, leading to attenuation of the Ca 2ϩ transient and contractile parameters. This is the first evidence indicating that alterations in calsequestrin expression levels are associated with alterations in basal cardiac contractile parameters assessed at the cellular, organ, and intact animal levels.
Cardiac-specific overexpression of calsequestrin was achieved using the ␣-myosin heavy chain promoter (38) and cDNA encoding mouse cardiac calsequestrin to assure complete homology between the overexpressed and endogenous proteins. Calsequestrin overexpression (20-fold) resulted in mild left ventricular hypertrophy without substantial morphological abnormalities. Immunofluorescence labeling of calsequestrin in transgenic ventricular tissue was generally observed as transverse striations, indicating that the overexpressed protein was predominantly localized to terminal cisternae at the Z-lines, similar to endogenous calsequestrin. Thus, the SR lumen, unlike the SR membrane (39,40), appears capable of accommodating relatively high levels of calsequestrin expression without significant architectural alterations in cell morphology. However, these findings differ from recent observations in a transgenic mouse model overexpressing lower levels (10-fold) of dog cardiac calsequestrin in the heart (41). Introduction of the heterologous protein also resulted in cardiac hypertrophy, but there was elimination of clear striations in ventricular myocytes, suggesting disruption of normal ultrastructure.
The hypertrophic response of our cardiac-specific calsequestrin-overexpressing mice was associated with induction of reexpression of a fetal gene program in the heart, particularly increases in atrial natriuretic factor, ␣-skeletal actin, and ␤-MHC transcripts. The increases in the expression of the slow MHC isoform (␤-MHC) may be a contributing factor to the attenuation of contraction and relaxation rates in transgenic hearts. This depression in cardiac mechanics reflected a decrease in the amplitude of the Ca 2ϩ transient, although the SR Ca 2ϩ load was significantly increased as revealed by caffeine application. The apparent impairment of SR Ca 2ϩ release in transgenic hearts may be due to (a) increased SR Ca 2ϩ -buffering capacity leading to lower intraluminal free Ca 2ϩ concentrations, which is associated with lower amounts of Ca 2ϩ released (12); (b) reduced Ca 2ϩ influx through L-type Ca 2ϩ channels, due to their lower density, leading to a smaller activation of the Ca 2ϩ -induced SR Ca 2ϩ release (the reduced SR Ca 2ϩ release was also evidenced by the slow inactivation ki-  netics of L-type Ca 2ϩ currents, which are regulated by the transient increase in the Ca 2ϩ concentration of the microdomain between the L-type Ca 2ϩ channel and the ryanodine receptor (24,42)); (c) increased expression of calreticulin, which may also contribute to increased SR Ca 2ϩ buffering and thus attenuation of the SR Ca 2ϩ release mechanism; and (d) a defect in the coupling between L-type Ca 2ϩ channels and ryanodine receptors in the hypertrophic transgenic cardiomyocytes, leading to a reduction in the ability of the L-type Ca 2ϩ current to trigger SR Ca 2ϩ release. Recently, defective excitation-contraction coupling without any changes in the densities of the L-type Ca 2ϩ current and ryanodine receptors was observed in a rat model of cardiac hypertrophy, and this was proposed to contribute to decreases in Ca 2ϩ transient amplitude (43). In the calsequestrin-overexpressing hearts, there were no significant alterations in the protein levels of ryanodine receptors, but the density of the L-type Ca 2ϩ current was decreased, similar to previous observations in models of severe pressure-overload hypertrophy or cardiac infarction (44 -46). Furthermore, examination of the levels of junctin and an ϳ95-kDa triadin isoform, which have been proposed to act as calsequestrin-"anchoring proteins" and to form a functional complex with the ryanodine receptor (14,35), revealed no alterations. However, the levels of FKBP12 were significantly reduced in transgenic hearts, suggesting an important compensatory mechanism in an attempt to enhance the opening probability of the ryanodine receptors and to facilitate SR Ca 2ϩ release, which may increase the amplitude of the cytosolic Ca 2ϩ signal and contractility (39). The increased open probability of ryanodine receptors may be associated with activation of calcineurin and dephosphorylation of the transcription factor NF-AT3, leading to reprogramming of gene expression and hypertrophy (47). Furthermore, down-regulation of FKBP12, which has been postulated to interact with and to inhibit the activity of transforming growth factor-␤ signaling (48), may also contribute to development of cardiac hypertrophy through the transforming growth factor-␤ pathway. Treatment of the calsequestrin-overexpressing mice with cyclosporin (47) may allow us to distinguish between these pathways for the hypertrophic response in this model. Assessment of the protein levels of the SR Ca 2ϩ -ATPase and its regulator phospholamban indicated that calsequestrin overexpression was associated with increases compared with wildtype littermates. It is interesting to note that the increases in the levels of these proteins were similar, suggesting that the maximal velocity of the SR Ca 2ϩ -ATPase activity was increased, without any changes in the affinity of this enzyme for Ca 2ϩ (49). The increased maximal velocity of the SR Ca 2ϩ pump probably constitutes an important compensatory response in an attempt to increase the amplitude of the Ca 2ϩ transient in transgenic cardiomyocytes. Decreases in contractile parameters were also observed in intact hearts and in closed-chest animals. Reduction of left ventricular isovolumic parameters indicated that calsequestrin overexpression was uncompensated in vivo. Thus, the development of hypertrophy may provide an adaptive response to normalize contractile dysfunction in this animal model. However, induction of the fetal gene program and other compensatory responses, accompanying calsequestrin overexpression, make it rather difficult to assess the direct effects of altered calsequestrin levels on cardiac contractile parameters.
In summary, our findings indicate that calsequestrin is a regulator of the cardiac SR Ca 2ϩ storage capacity and the amount of Ca 2ϩ available for release during excitation-contraction coupling. Thus, alterations in the expression levels of this protein would reflect alterations in the levels of cytosolic Ca 2ϩ and contractile parameters. The availability of transgenic mod-els with altered expression of calsequestrin will facilitate further studies on elucidating the mechanism(s) by which calsequestrin buffers SR luminal Ca 2ϩ , the effect(s) of free luminal Ca 2ϩ on SR Ca 2ϩ release and ryanodine receptor gating properties, and the dynamic relationship between changes in SR luminal Ca 2ϩ and cytosolic Ca 2ϩ levels during the contractile cycle. Such studies will greatly improve our understanding of the regulatory mechanisms involved in SR Ca 2ϩ cycling and contractility in the mammalian myocardium.