Superinhibition of Sarcoplasmic Reticulum Function by Phospholamban Induces Cardiac Contractile Failure*

To determine whether selective impairment of cardiac sarcoplasmic reticulum (SR) Ca2+ transport may drive the progressive functional deterioration leading to heart failure, transgenic mice, overexpressing a phospholamban Val49 → Gly mutant (2-fold), which is a superinhibitor of SR Ca2+-ATPase affinity for Ca2+, were generated, and their cardiac phenotype was examined longitudinally. At 3 months of age, the increased EC50 level of SR Ca2+ uptake for Ca2+ (0.67 ± 0.09 μm) resulted in significantly higher depression of cardiomyocyte rates of shortening (57%), relengthening (31%), and prolongation of the Ca2+ signal decay time (165%) than overexpression (2-fold) of wild type phospholamban (68%, 64%, and 125%, respectively), compared with controls (100%). Echocardiography also revealed significantly depressed function and impaired β-adrenergic responses in mutant hearts. The depressed contractile parameters were associated with left ventricular remodeling, recapitulation of fetal gene expression, and hypertrophy, which progressed to dilated cardiomyopathy with interstitial tissue fibrosis and death by 6 months in males. Females also had ventricular hypertrophy at 3 months but exhibited normal systolic function up to 12 months of age. These results suggest a causal relationship between defective SR Ca2+ cycling and cardiac remodeling leading to heart failure, with a gender-dependent influence on the time course of these alterations.

To determine whether selective impairment of cardiac sarcoplasmic reticulum (SR) Ca 2؉ transport may drive the progressive functional deterioration leading to heart failure, transgenic mice, overexpressing a phospholamban Val 49 3 Gly mutant (2-fold), which is a superinhibitor of SR Ca 2؉ -ATPase affinity for Ca 2؉ , were generated, and their cardiac phenotype was examined longitudinally. At 3 months of age, the increased EC 50 level of SR Ca 2؉ uptake for Ca 2؉ (0.67 ؎ 0.09 M) resulted in significantly higher depression of cardiomyocyte rates of shortening (57%), relengthening (31%), and prolongation of the Ca 2؉ signal decay time (165%) than overexpression (2-fold) of wild type phospholamban (68%, 64%, and 125%, respectively), compared with controls (100%). Echocardiography also revealed significantly depressed function and impaired ␤-adrenergic responses in mutant hearts. The depressed contractile parameters were associated with left ventricular remodeling, recapitulation of fetal gene expression, and hypertrophy, which progressed to dilated cardiomyopathy with interstitial tissue fibrosis and death by 6 months in males. Females also had ventricular hypertrophy at 3 months but exhibited normal systolic function up to 12 months of age. These results suggest a causal relationship between defective SR Ca 2؉ cycling and cardiac remodeling leading to heart failure, with a gender-dependent influence on the time course of these alterations.
Cardiac hypertrophy and failure are highly complex disorders that arise as a result of a combination of mechanical, hemodynamic, hormonal, and pathological stimuli (1). In response to these effectors, the heart undergoes an adaptive response of compensatory hypertrophy (2) followed by decompensated heart failure that is characterized by defects in Ca 2ϩ handling during excitation-contraction coupling. Studies of end-stage-failing hearts have shown that the disturbed calcium homeostasis is associated with alterations in the expression levels or the activity of key Ca 2ϩ -handling proteins, leading to abnormal excitation contraction coupling and diastolic as well as systolic dysfunction (3,4). Specifically, alterations in SR 1 Ca 2ϩ -ATPase (SERCA2a) activity, the major Ca 2ϩ transport protein in SR, have been implicated as important determinants in the deteriorated function of the failing heart (5)(6)(7). The activity of SERCA2a is regulated by phospholamban (PLB), a 52-amino acid, muscle-specific SR phosphoprotein (8 -10). Dephosphorylated PLB inhibits the Ca 2ϩ affinity of SERCA2a, whereas the phosphorylated form of PLB dissociates from SERCA2a leading to increases in Ca 2ϩ uptake rates and accelerated ventricular relaxation (11)(12)(13).
The role of PLB in the regulation of basal contractility has been elucidated through the development of genetically engineered mouse models. Phospholamban ablation resulted in significant increases in cardiac contractile parameters, whereas overexpressing PLB was associated with depressed cardiac function (14,15). Actually, an inverse relationship between the ratio of PLB/SERCA2a and myocardial contractility was observed (16), suggesting that any process that alters the relative levels of these proteins and/or the interaction between them may result in altered myocardial contractility. However, it is not currently known whether a primary defect in SR Ca 2ϩ uptake, elicited by altered interactions between PLB and SERCA2a, may result in depressed cellular Ca 2ϩ homeostasis and drive the progressive deterioration of function leading to heart failure.
Early studies evaluating the interaction between PLB and SERCA2a were directed toward the cytoplasmic domain (amino acid residues 1-30), which includes the phosphorylation sites in PLB (17,18). More recently, the transmembrane interaction sites (amino acid residues 31-52) were shown to mediate the regulatory effects of PLB on SERCA2a affinity, with some mutations yielding increased inhibition, whereas others abolish the PLB inhibitory effects on SERCA2a in vitro (19 -21). Extension of these studies to in vivo models indicated that overexpression of either L37A or I40A mutant PLB resulted in depressed cardiac myocyte Ca 2ϩ kinetics and mechanical parameters associated with hypertrophy (22). Further support for the functional significance of the PLB transmembrane domain in mediating its regulatory effects on SERCA2a was provided by a recent study, in which a Val 49 3 Ala (V49A) mutant PLB was expressed in vivo, using recombinant adenoviruses. The V49A mutant acted dominantly to increase contractility in cardiac ventricular cells in the absence of catecholamines (23). This study pointed out the importance of the PLB carboxyl terminus in mediating its interaction with SERCA2a and proposed that interfering with this interaction may provide a novel therapeutic approach for the prevention of dilated cardiomyopathy. To further elucidate the functional significance of the PLB Val 49 residue in vivo, we generated a mutant, Val 49 3 Gly (V49G), which acts as a potent inhibitor of the Ca 2ϩ affinity of SERCA2a. Cardiac-specific overexpression of the V49G mutant PLB resulted in superinhibition of cardiac contractility and cardiac remodeling, which progressed to dilated cardiomyopathy and early mortality. Thus, enhanced inhibition of SERCA2a by PLB may serve as a prime candidate for driving the onset and progression of heart failure.

EXPERIMENTAL PROCEDURES
In Vitro Co-expression of SERCA2a and Mutant Phospholamban-Rabbit PLB wild type or V49G mutant and SERCA2a cDNAs were co-transfected into the human embryonic kidney cell line (HEK-293 cells), and microsomal Ca 2ϩ transport activity was assayed according to a previous study (24).
Creation of Mutant Phospholamban Mice-The site-specific mutation of V49G (GTG to GGG) was introduced into PLB cDNA by polymerase chain reaction (24). The entire expression construct was composed of the cardiac-specific ␣-myosin heavy-chain promoter (␣-MHC P , 5.5 kb, a gift from Dr. J. Robbins, Children's Hospital, Cincinnati, OH), the PLB coding region with the V49G mutation (0.65 kb), and the SV40 polyadenylation signal. Microinjection and identification of transgenic mice were performed (24).
Biochemical Assays-Quantitative immunoblotting of cardiac homogenates was used to determine the levels of PLB and SR calciumhandling proteins (25). Cardiac gene expression was assayed by dot blot analysis of total RNA (10 g) using 32 P-labeled oligonucleotide probes (25). Oxalate-supported Ca 2ϩ uptake in cardiac homogenates was measured by a modified Millipore filtration technique (15).
Functional Measurements-Mouse left ventricular (LV) cardiomyocytes were isolated, and cardiomyocyte mechanical properties and Ca 2ϩ transients were examined (24). Whole-cell L-type Ca 2ϩ channel currents were recorded, using patch-clamp techniques (26). Briefly, Ca 2ϩ channel currents were elicited by 300-ms depolarizing pulses from a holding potential of Ϫ50 mV, applied every 10 s. Cell capacitance was measured using voltage ramps of 0.8 V/s from a holding potential of Ϫ50 mV.
Left ventricular M-mode and Doppler measurements (27) were performed at baseline and after the administration of isoproterenol (2.0 g/g intraperitoneally) in mice, which were anesthetized with 2.5% avertin (0.01 ml/g).
Histology-Hearts from mutant PLB and control mice were excised, photographed, and subjected to histological analysis. Briefly, hearts were collected, fixed overnight in 10% formalin, buffered with phosphate-buffered saline for up to 24 h, dehydrated in 70% ethanol, and transferred to xylene and then into paraffin. Paraffin-embedded hearts were sectioned at 4 m and stained with Masson's trichrome.
Statistical Analysis-Data are presented as mean Ϯ S.E. The number n indicates the number of mice unless otherwise stated. Statistical analysis was performed by one-way or two-way analysis of variance, followed by Student-Newman-Keuls' test for multiple comparisons. Values of p Ͻ 0.05 were considered to be statistically significant.

In Vitro Co-expression of Mutant Phospholamban and
SERCA2a-We identified a mutation (V49G) in the transmembrane domain of PLB that did not alter the monomer-to-pentamer ratio, but resulted in superinhibition of SERCA2a's Ca 2ϩ affinity in vitro. Co-expression of wild type PLB with SERCA2a in HEK-293 cells was associated with a significant shift in the Ca 2ϩ dependence (EC 50 ) of Ca 2ϩ uptake activity from 0.25 Ϯ 0.05 M to 0.55 Ϯ 0.05 M (n ϭ 9, p Ͻ 0.05). However, when the PLB V49G mutant was co-expressed with SERCA2a, there was a further significant (p Ͻ 0.05) shift in the EC 50 value (1.29 Ϯ 0.14 M, n ϭ 9), indicating a greater inhibitory effect compared with wild type PLB.
Characterization of Transgenic Mice Expressing Mutant Phospholamban-To assess the physiological significance of the superinhibitor mutant PLB in vivo, transgenic mice with cardiac-specific expression of mutant PLB (PLB-MT) were generated. Four founder (lines 1, 2, 3, and 4) mice, harboring the mutant PLB transgene were identified. The expression of a transgenic transcript migrating at ϳ1.0 kb was only detected in the hearts of transgenics, using Northern blot analysis of total RNA (data not shown) (14).
Western blot analysis of cardiac homogenates from 3-monthold male or female transgenic and wild type mice revealed that the PLB protein levels were increased by 2-fold in lines 1, 3, and 4 and by 1.8-fold in line 2, compared with wild types. Assessment of the SERCA protein levels in transgenic mouse hearts showed no significant changes (Fig. 1, A and B). Lines 3 and 4 were then propagated for further characterization along with transgenic mice overexpressing similar (2-fold) levels of wild type PLB (14). To avoid gender differences, initial characterization studies were carried out using male mice at 3 months of age.
Sarcoplasmic Reticulum Ca 2ϩ Uptake Assays-The initial rates of SR Ca 2ϩ uptake, assessed over a wide range of [Ca 2ϩ ] (Fig. 1C), indicated that overexpression of mutant PLB (line 3) resulted in a significant increase in the EC 50 value for Ca 2ϩ (0.67 Ϯ 0.01 M), compared with wild type PLB overexpression (0.48 Ϯ 0.04 M) and control hearts (0.30 Ϯ 0.02 M). Similar results were obtained with line 4. These data suggest that the mutant PLB interacts with, and inhibits, SERCA2a to a greater extent than wild type PLB.
Mechanical Properties and Ca 2ϩ Transients in Isolated Cardiomyocytes-To elucidate whether the decreased affinity of the SERCA2a for Ca 2ϩ following mutant PLB overexpression was associated with alterations in cardiac function, the contractile parameters of LV cardiomyocytes were assessed. Fig. 2 indicates that the extent of cardiomyocyte shortening and the maximal rates of shortening (ϩdL/dt: 57%) and relengthening (ϪdL/dt: 31%) were depressed in transgenic (line 3) cardiomyocytes overexpressing mutant PLB, compared with wild type PLB overexpression (ϩdL/dt: 68%, ϪdL/dt: 64%) cardiomyocytes (wild types: 100%). Furthermore, resting cell length was significantly increased (132%) in mutant, compared with wild type PLB overexpression or wild type cardiomyocytes (100%). The changes observed in contractile parameters suggested that alterations would be found in Ca 2ϩ kinetics in transgenic myocytes. Thus, cardiomyocytes were loaded with 4 M Fura-2-AM to permit detection of changes in free Ca 2ϩ during stimulation. Peak amplitude was not significantly different in mutant relative to wild type PLB overexpression and wild type control myocytes (Fig. 2). The time to 80% decay of the Ca 2ϩ signal (T 80 ) was more prolonged (165%) in cardiomyocytes overexpressing the mutant PLB than in cardiomyocytes overexpressing wild type PLB (125%), compared with controls (100%).
To clarify whether the superinhibitory effects could be relieved by ␤-agonists, cardiomyocytes were subjected to maximal stimulation (100 nM) by isoproterenol. Administration of isoproterenol resulted in complete reversal of the inhibitory effects in cardiomyocytes overexpressing wild type PLB. However, in transgenic cardiomyocytes overexpressing mutant PLB, the maximally stimulated T 80 of the Ca 2ϩ transient parameter remained depressed, compared with maximally stimulated values in controls.
When peak I Ca amplitude, normalized relative to cell capacitance (pA/pF) as a function of voltage (I Ca Ϫ V relationship), was determined, the current density was significantly in-creased in mutant PLB myocytes (13.7 Ϯ 0.6 pA/pF, n ϭ 39) compared with wild type mocytes (8.7 Ϯ 0.3 pA/pF, n ϭ 64). Myocytes overexpressing wild type PLB exhibited no significant differences in L-type Ca 2ϩ channel current compared with wild types (data not shown).
In Vivo Assessment of Cardiac Function-To determine whether the depressed function, observed in cardiomyocytes from mutant PLB mice, correlated with similar depression in LV systolic function in vivo, M-mode and Doppler echocardiography was performed in the three groups under basal (Fig. 4) and isoproterenol stimulation conditions. End diastolic and end systolic dimensions were significantly higher, whereas heart rate, fractional shortening, and velocity of circumferential fiber shortening were significantly depressed in mutant hearts (Table I). The depressed heart rate was probably due to the decreased Ca 2ϩ release from the SR in pacemaker cells (28). Furthermore, LV-to-body mass (Table I), LV anterior wall thickness, and posterior wall thickness were increased in mutant compared with wild type PLB overexpression or wild type control mice (data not shown). The relative wall thickness (h/r) was also higher in mutant hearts (Table I), indicating the presence of concentric hypertrophy.
Isoproterenol administration enhanced ejection time, heart rate, fractional shortening, and velocity of circumferential fiber shortening, and the percent increase, relative to baseline values, was similar between all groups. However, with the exception of ejection time, the maximally stimulated parameters in mutant PLB mice were attenuated, which is in agreement with the findings in cardiomyocytes. Echocardiography was also performed in line 4 under basal and isoproterenol stimulation conditions. The results indicated similar alterations (data not shown) to those observed with line 3.
Cardiac Hypertrophy-Because mutant PLB cardiomyocytes exhibited increased cell capacitance and M-mode echocardiography indicated the presence of concentric hypertrophy in mutant hearts (Table I), gravimetric analysis of heart and body weights was pursued in 3-month-old males. Overexpression of the mutant PLB was associated with increases in heart-to-body weight ratios, which were 6.30 Ϯ 0.13 (n ϭ 12), compared with 3.88 Ϯ 0.09 (n ϭ 12) for wild type PLB overexpression, and 3.95 Ϯ 0.18 (n ϭ 12) for wild type control males. The lung-tobody weight and liver-to-body weight ratios were also significantly higher in mutant (5.44 Ϯ 0.18 and 51.70 Ϯ 0.09 mg/g, respectively) compared with wild type PLB overexpression (4.38 Ϯ 0.20 and 37.98 Ϯ 0.10 mg/g, respectively) and wild type controls (4.60 Ϯ 0.18 and 39.5 Ϯ 0.11 mg/g, respectively). Line 4 also showed increased heart-to-body weight ratios (6.89 Ϯ 0.07 mg/g, n ϭ 12).
The increases in heart-to-body weight ratios were associated with increased expression of a fetal cardiac gene program, particularly the ␤-myosin heavy chain isoform (8.25-fold), ␣-skeletal actin (8.5-fold), and ventricular expression of atrial natriuretic factor (ANF) (11-fold). The relative increases in ␤-myosin heavy chain and ␣-skeletal actin protein levels were also assessed, and they were 7.35-and 8-fold over wild types, respectively, when normalized to calsequestrin protein levels (data not shown).
To further assess the hypertrophic alterations in the mutant PLB transgenic male hearts, histologic analysis was performed at 3 months of age. Every mutant PLB mouse analyzed showed a dramatic increase in heart size relative to wild type controls (Fig. 5A). Furthermore, histologic analysis revealed myocardial fibrosis in hearts of mutant PLB transgenics (Fig. 5B).
Effect of Aging on Cardiac Function-Male mutant PLB mice died by 6 -7 months of age from congestive heart failure, which is supported by the increases in lung-to-body weight ratios (6.71 Ϯ 0.4 mg/g at 6 months; n ϭ 8). To evaluate LV systolic function at that time, M-mode and Doppler echocardiography were performed. The fractional shortening and velocity of circumferential fiber shortening were significantly depressed in mutant PLB (22.61 Ϯ 0.80% and 4.02 Ϯ 0.31 circ/s, respectively; n ϭ 8) at 6 months of age compared with 3-month-old males (Table I). Furthermore, the end diastolic dimension, end systolic dimension, LV mass, and ejection time were significantly higher at 6 months (4.92 Ϯ 0.33 mm, 3.82 Ϯ 0.30 mm, 112.3 Ϯ 14.1 mg, and 89.43 Ϯ 4.05 ms, respectively; n ϭ 8) compared with 3-month values. The wall thickness-to-cavity radius ratio (0.29 Ϯ 0.03; n ϭ 8) was also further decreased. In addition, gravimetric analysis revealed further increases in heart and liver-to-body weight ratios (7.38 Ϯ 0.45 and 62.01 Ϯ 0.23, respectively, n ϭ 8). However, no significant differences were observed in wild type PLB overexpression or wild type control mice upon aging (data not shown). Thus, the mutant PLB mice exhibited LV concentric hypertrophy and contractile dysfunction at 3 months, which was followed by progressive LV

FIG. 3. Properties of L-type Ca 2؉ current (I Ca ) in PLB-MT and WT ventricular myocytes. Representative whole-cell L-type Ca 2ϩ currents recorded in (A) wild type (WT) and (B) mutant PLB (PLB-MT)
cells. Currents were elicited from a holding potential of Ϫ50 mV to the indicated test potentials. C, peak current voltage relationships obtained from these cells. I Ca was measured and normalized to the cell capacitance to obtain current densities (pA/pF). Values are mean Ϯ S.E. (n ϭ 34 -64).

FIG. 4. Representative M-mode echocardiography of 3-monthold transgenic and wild type mice. Left ventricular M-mode and
Doppler measurements were performed at baseline and after the administration of isoproterenol (2.0 g/g intraperitoneally) in mice, which were anesthetized with 2.5% avertin (0.01 ml/g). In vivo assessment of cardiac function revealed increases in end diastolic dimension (EDD), end systolic dimension (ESD), and slower heart rate in transgenic mice overexpressing mutant PLB (PLB-MT), compared with wild type PLB overexpression (PLB-WT) and wild type (WT) controls. dilation, deterioration of LV systolic function, and development of overt congestive heart failure at 6 months.
Gender-dependent Survival Rate and Cardiac Function-A Kaplan-Meier analysis demonstrated that male mutant PLB mice died between 6 and 7 months of age, whereas female mutant PLB mice died between 15 and 18 months of age (Fig.  6A). There were no deaths in wild type male or female controls during this period. These findings suggested a gender-dependent progression in cardiac dysfunction. Thus, LV function and its time course of deterioration were further examined in female mutant PLB mice. M-mode and Doppler echocardiography indicated that end diastolic dimension, end systolic dimension, fractional shortening, velocity of circumferential fiber shortening, heart rate, and ejection time were not significantly different between female mutant and wild type PLB overexpression mice at 3 months (Table II). However, the LV mass and wall thickness-to-cavity radius ratio were higher in mutant than wild type PLB overexpression females, indicating a concentric hypertrophic response. Gravimetric analysis also revealed increases in heart-to-body weight ratios (6.18 Ϯ 0.55, n ϭ 6), compared with wild type PLB overexpression (3.90 Ϯ 0.07, n ϭ 6) and wild type control (3.79 Ϯ 0.07, n ϭ 6) females. The lung-to-body weight and liver-to-body weight ratios were   n ϭ 32, A). Echocardiographic assessment of LV-to-body mass ratio (B), end diastolic dimension (EDD, C), fractional shortening (FS, D), and velocity of circumferential fiber shortening (Vcf c , E) at 3 and 6 months of age in male (n ϭ 14 at 3 and 8 at 6 months of age) and female (n ϭ 3 at 3, 12 at 6, and 4 at 12 months of age) mutant PLB (PLM-MT) mice. The same group of mice was used for sequential echocardiographic measurements. *, p Ͻ 0.05 versus same age; #, p Ͻ 0.05 versus same sex. also significantly higher in mutant (7.12 Ϯ 0.67 and 58.00 Ϯ 0.67 mg/g, respectively) compared with wild type PLB overexpression (5.11 Ϯ 0.25 and 49.00 Ϯ 0.14 mg/g, respectively) and wild type controls (5.43 Ϯ 0.14 and 50.00 Ϯ 0.05 mg/g, respectively).
Isoproterenol enhanced fractional shortening and velocity of circumferential fiber shortening, in mutant and wild type PLB overexpression mice, and the percentage increases, relative to baseline values, were similar between both groups. Further analysis of fractional shortening, velocity of circumferential fiber shortening, and ejection time at 6 months did not show any significant alterations in mutant females compared with wild type PLB overexpression females, indicating that the LV systolic function was preserved. However, end diastolic dimension was significantly increased in mutant females at 6 months compared with wild type PLB overexpression mice, indicating the onset of LV dilation in these hearts. Upon aging to 12 months, the mutant females exhibited no further alterations in end diastolic dimension and LV-to-body mass. Interestingly, fractional shortening was decreased compared with 3-monthold mutant and age-matched wild type PLB overexpression female controls. Between 12 and 15 months of age, the LV function deteriorated rapidly and progressed to cardiac failure. These findings suggested that mutation of V49G in PLB resulted in preserved contractile function and concentric LV hypertrophy in females at 3 months of age, which consequently progressed to dilation and contractile dysfunction during the aging process. Gravimetric analysis also revealed increased heart-to-body weight ratios at 6 and 12 months of age (7.28 Ϯ 0.45 and 7.89 Ϯ 0.52, respectively), but these increases were not significantly different between 6 and 12 months.
To assess gender differences in LV remodeling, hypertrophy, and function in response to the mutant PLB superinhibitory effect, the LV and body mass, end diastolic dimension, and LV systolic function in male and female mutant PLB mice were compared at 3 and 6 months of age. Female mutant mice had significantly lower body weights (23 Ϯ 0.74 and 27 Ϯ 0.44 g, at 3 and 6 months, respectively) compared with male mutant (30 Ϯ 1.0 and 34 Ϯ 0.7 g, at 3 and 6 months, respectively) mice. However, body weights were similar between female mutant mice and their age-matched wild type controls, and this was also observed for male mice. At 3 months, both males and females developed the same degree of hypertrophy (Fig. 6B), measured by percent increases in their LV-to-body mass over their male or female wild type controls, which was also supported by similar gravimetric measurements described above. The fractional shortening and velocity of circumferential fiber shortening declined in males but not in females, which exhib-ited preserved LV systolic function at 3 and 6 months of age (Fig. 6, D and E), indicating differences in their adaptive hypertrophic responses (concentric versus eccentric). Upon aging to 6 months, the end diastolic dimension was increased in mutant PLB male mice, indicating decompensation with cavity dilation and transition to heart failure (Fig. 6C). Mutant PLB females also exhibited significant increases in end diastolic dimension, compared with their 3-month-old counterpart, but their LV function was preserved at 6 months. DISCUSSION This is the first study to demonstrate that a mutation in PLB, which is associated with superinhibition of SERCA2a and contractile parameters, may lead to dilated cardiomyopathy, overt heart failure, and early mortality, which are modified by gender. Previous studies in animal models or human with end stage heart failure showed that the increases in diastolic calcium and impaired relaxation were linked to inhibition of SR Ca 2ϩ transport (3,29). However, the depressed SR function was suggested to be secondary to insults arising from extracellular effectors and may act as a modifier for the progression of cardiac deterioration (30). Thus, it is not currently clear whether inhibition of SR Ca 2ϩ transport is sufficient to cause heart failure or it is a contributing factor to cardiomyocyte dysfunction in the context of pre-existing heart disease. To better address the role of depressed SR Ca 2ϩ transport activity in the onset and progression of heart failure, we generated a transgenic model harboring the V49G mutation in PLB, which we identified as a superinhibitor of SERCA2a Ca 2ϩ affinity in vitro, and studied the cardiac phenotype over the life span of the mouse. Overexpression of this mutant PLB was associated with significant depression in SR Ca 2ϩ uptake rates. The inhibited SR function resulted in impaired cardiac contractile parameters in isolated cardiomyocytes and intact mice at 3 months of age. In cardiomyocytes, the attenuated mechanical parameters reflected a significant prolongation in the rate of the Ca 2ϩ signal decay (T 80 ) without alteration in the peak amplitude of the Ca 2ϩ transient. The lack of peak amplitude alteration may reflect increased Ca 2ϩ influx through the L-type Ca 2ϩ channels, whose density was increased in transgenic myocytes, indicating an important compensatory response in an attempt to normalize systolic Ca 2ϩ levels in these cells. The V49G mutant myocytes also exhibited significantly slower I Ca inactivation, despite increased Ca 2ϩ influx through the L-type Ca 2ϩ channels, suggesting that the SR Ca 2ϩ release was diminished (26,31). Importantly, similar phenotypic alterations were exhibited in a second line (line 4), which expressed the same levels of mutant PLB, indicating that the inhibitory effect  was due to transgene expression and not any possible insertional mutation.
The mechanisms underlying the superinhibitory effects of V49G mutant PLB in vivo are not clear. However, this study together with a previous one, utilizing a V49A mutant (23), emphasize that the Val 49 residue in PLB, which is highly conserved among species, is important for the PLB/SERCA2a interaction and changes in this region have profound effects on SR function and cardiac contractility. Thus, the V49G mutation, employed in this study, may influence the three-dimensional structure of PLB and enhance its interaction with SERCA2a. Importantly, the normal ratio (10:1) of pentamers to monomers was not altered by this mutation in PLB, further supporting that the superinhibitory effect of the mutant may be due to its conformational change. An enhanced association between PLB and SERCA2a would be consistent with the inability of isoproterenol to fully relieve the V49G mutant PLB superinhibitory effects, similar to previous findings with the N27A PLB mutant (24). In contrast, the reduced contractility in L37A and I40A PLB mutant mice could be completely over come by isoproterenol stimulation (22).
It is of interest that cardiac-specific overexpression of the PLB V49G mutant induced cardiac hypertrophy in an attempt to normalize cardiac function (32,33), which was severely depressed in mutant hearts. However, the molecular mechanisms that link depressed SR Ca 2ϩ cycling to reprogramming of gene expression in the PLB V49G hearts are not clear. Hypertrophy was associated with induction or re-expression of a fetal gene program, including ANF, ␤-myosin heavy chain, and ␣-skeletal actin in these hearts (34). The increases in the protein expression levels of the slow myosin heavy chain (␤-MHC) isoform (35) may also contribute to the observed attenuation of contraction and relaxation rates in the mutant PLB hearts. Interestingly, the hypertrophic remodeling, which provided an initial compensatory phase, progressed to cardiac failure and premature death in this model expressing the V49G PLB superinhibitor. Previous studies on the N27A (24) and L37A and I40A (22) PLB superinhibitors also showed severely depressed SR and cardiac function in vivo, which were associated with hypertrophy by 3 months of age. However, the hypertrophic phenotype of the N27A, L37A, and I40A mutant mice did not progress to overt heart failure, as observed with the V49G mutant. The apparent differences between these transgenic models may be due to the expression levels of the various PLB mutants in vivo, the age of the mice studied, and/or the time course of transition between the compensatory hypertrophic response and heart failure. In addition, there may be differences in activated intrinsic hypertrophic signaling pathways, triggered by SR dysfunction, which lead to different courses of myocyte maladaptation among these transgenic mice. However, the present findings and those in other models (22)(23)(24)37) suggest that multiple pathways, including disturbed SR Ca 2ϩ homeostasis, may operate in concert to induce a hypertrophic response leading to heart failure.
Several studies have indicated that cardiovascular mortality is higher in males than in females, which frequently exhibit preserved cardiac performance, and suggested a gender-dependent influence on the onset and progression of heart failure (36). In the present study, we also observed an early mortality in males (6 months) compared with females (15 months) in our model with genetically induced SR Ca 2ϩ -handling defects. The striking survival benefit in females was not due to differences in their levels of PLB, compared with age-matched males. Therefore, we investigated gender-specific differences in cardiac function and in the progression to heart failure. Males and females were examined during compensatory hypertrophy and after the appearance of symptoms of heart failure, using echocardiography. LV-to-body mass was about 50% increased by 3 months of age, and cardiac function was significantly depressed in male PLB mutant mice. The remodeling process accelerated the progression to heart failure, characterized by significant chamber dilation and a decrease in the ratio of wall thickness to chamber diameter (h/r) by 6 months of age. However, the progression to failure was delayed in females (37), indicating that, despite a similar degree of LV hypertrophy between males and females, there were significant gender differences in the LV-adaptive response to pathological hypertrophy and depressed function, similar to recent studies in the TNF-␣ overexpression model and in a rat model of pressure overload hypertrophy (38). The etiology of these differences is unclear but may be related to: (a) a reduced adaptive hypertrophic reserve in males (39); (b) the lower mitochondria respiratory and lysosomal enzyme activity in females (40); (c) a higher percentage of the V1 myosin isozyme, which is up-regulated by estrogen in females (41); and (d) the intrinsic gender-specific differences in cardiac muscle physiology and biochemistry (42). Furthermore, estrogen signaling through the adult myocyte estrogen receptor may contribute to gender differences in gene expression in pathological hypertrophy (43). Interestingly, in humans with congenital aortic stenosis and cardiac hypertrophy, there appears to be an overcompensation of myocardial contraction early on (44), which is similar to that described in mutant PLB females in this study.
In summary, our findings point to a primary defect in SR function as an inducer of a phenotype that resembles human dilated cardiomyopathy and ventricular remodeling. The enhanced inhibition of SERCA2a by the V49G mutant PLB, associated with depressed myocyte calcium homeostasis, resulted in a remodeling process, which involved interaction between all components of the myocardium leading to overt heart failure. Furthermore, gender significantly influenced the evolution of the early responses to LV remodeling, including the transition to heart failure in the mutant PLB model. Future studies on the cellular and molecular mechanisms underlying these responses, using gene expression profiling and proteomics, may unveil specific molecules or pathways by which depressed SR function influences the onset of hypertrophy and the transition to dilated cardiomyopathy in this model.