The transgenic expression of highly inhibitory monomeric forms of phospholamban in mouse heart impairs cardiac contractility.

Transgenic mice were generated with cardiac-specific overexpression of the monomeric, dominant-acting, superinhibitory L37A and I40A mutant forms of phospholamban (PLN), and their phenotypes were compared with wild-type (wt) mice or 2-fold overexpressors of wt PLN (wtOE). The level of PLN monomer in cardiac microsomes was increased 11-13-fold, and the apparent affinity of the sarco(endo)plasmic reticulum Ca(2+)-ATPase for Ca(2+) was decreased from pCa 6.22 in wt or 6.12 in wtOE to 5.81 in L37A and 5.72 in I40A. Basal physiological parameters, measured in isolated myocytes, indicated a significant reduction in the rates of shortening (+dL/dt) and relengthening (-dL/dt). Hemodynamic measurements indicated that peak systolic pressure was unaffected but that pressure changes (+dP/dt and -dP/dt) were lowered significantly in both mutant lines, and relaxation time (tau) was also lengthened significantly. Echocardiography for both mutants showed depressed systolic function and an increase in left ventricular mass of over 1.4-fold. Significant decreases in left ventricular shortening fraction and velocity of circumferential shortening and increases in ejection time were corrected by isoproterenol. The use of antibodies specific against Ser(16)- and Thr(17)-PLN peptides showed that phosphorylation of both pentameric and monomeric PLN were increased between 1.2- and 2.4-fold in both the L37A and I40A lines but not in the wtOE line. These observations show that overexpression of superinhibitory mutant forms of PLN causes depression of contractile parameters with induction of cardiac hypertrophy, as assessed with echocardiography.

The contraction of cardiac muscle depends, in part, on the release of Ca 2ϩ from the lumen of the sarcoplasmic reticulum; relaxation depends on removal of Ca 2ϩ from the myoplasm by the combined activity of Ca 2ϩ pumps in the sarcoplasmic reticulum (SERCA 1 pumps), the plasma membrane pumps, and by Na ϩ /Ca 2ϩ exchange at the plasma membrane (1,2). The activity of SERCA2a, the cardiac/slow-twitch isoform of the Ca 2ϩ pump, is regulated by the integral membrane protein phospholamban (PLN) (3). Dephosphorylated PLN inhibits the activity of SERCA2a at low Ca 2ϩ concentrations by lowering its apparent affinity for Ca 2ϩ . Inhibition is overcome by phosphorylation of PLN by either protein kinase A or Ca 2ϩ -calmodulin kinase (3,4).
That PLN regulates left ventricular basal contractile parameters and their responses to ␤-agonists has been demonstrated through the analysis of the function of myocytes infused with an antibody against PLN (5), of PLN-null mice (6,7), and of mice overexpressing wild-type (wt) PLN (8). Hemodynamic tests proved that the ablation of PLN is associated with significant increases in intraventricular systolic pressure, in the rate of change of both positive (ϩdP/dt) and negative (ϪdP/dt) hemodynamic pressure, and in , a measure of the rate of cardiac muscle relaxation. These elevated parameters could be stimulated only minimally with the ␤-adrenergic agonist, isoproterenol (9). By contrast, the 2-fold overexpression of wt PLN in transgenic mice resulted in an inhibition of Ca 2ϩ transport by the sarcoplasmic reticulum, decreased Ca 2ϩ kinetics, contractile parameters in transgenic ventricular myocytes, and depressed basal left ventricular systolic function in vivo (8).
Wild-type PLN exists in pentameric and monomeric forms (10), but mutations in one face of the PLN transmembrane helix, including Leu 37 to Ala (L37A) and Ile 40 to Ala (I40A), lead to monomer formation (11,12). These mutations also enhance the inhibitory function of PLN, leading to the proposal that monomeric PLN is the inhibitory species (13,14). This view has been supported by measurement of the monomer concentration in membranes (15). If the monomer is the inhibitory species, the superinhibitory function of monomeric mutant forms of PLN can be explained, at least in part, by mass action, since the normal PLN pentamer to monomer ratio of about 10:1 is reduced to less than 1:10 (13). PLN mutants N27A and N30A are also superinhibitors, even though they are not monomeric (16). We have proposed that these mutant PLN pentamers dissociate normally and inhibit as monomers, but the monomer has enhanced affinity for SERCA2a (13,16). Anomalous mutants have also been described. The C41A mutant is about 75% monomeric but does not gain inhibitory function at 25°C (13). Furthermore, the C41F mutant is largely monomeric at 37°C (17) but loses partial inhibitory function in vitro (18) and in vivo (19), when compared with controls.
The studies of Luo et al. (6) have shown that PLN ablation, which results in chronic relief of the inhibitory properties of PLN, alters cardiac function but does not lead to cardiomyopathy. As a corollary, it was of interest to determine whether chronic high inhibition of SERCA2a by superinhibitory, monomeric forms of PLN might affect cardiac function and possibly lead to cardiomyopathy. Accordingly, we created transgenic mice in which L37A and I40A are overexpressed. Overexpression results in a dramatic shift to lower affinity in the range over which Ca 2ϩ is regulated, compared with overexpression of similar amounts of wild-type PLN (13). Hemodynamic and echocardiographic measurements and studies of myocytes support the view that the transgenic overexpression of the L37A and I40A superinhibitory forms of PLN impairs cardiac contractility.

EXPERIMENTAL PROCEDURES
Construction of PLN L37A and I40A Transgenic Mice-The mutations L37A and I40A were introduced into rabbit PLN cDNA as described previously (13). The construct used previously was modified by the addition of an XhoI restriction endonuclease site at the 5Ј end and by modification of the region to a consensus translation initiation sequence (20) for efficient mRNA translation. The 5Ј nucleotide sequence, used as a forward primer in PCR amplification of the clone, was 5Ј-CTG TTA CTC GAG GCC ACC ATG GAG AAA GTT C-3Ј. A HindIII restriction endonuclease site was introduced into the 3Ј end of the clone using the reverse PCR primer 5Ј-CTG TCA AGA AGC TTA GTC AGA GAA GCA TGA CGA TG-3Ј).
The amplified DNA fragments were ligated into the SalI and HindIII sites of plasmid pC126 (21) downstream of the 5.5-kb mouse cardiac ␣-MHC promoter and upstream of the poly(A) signal sequence from the human growth hormone. The presence of the mutations and the sequences of the expression constructs were confirmed by restriction mapping and nucleotide sequencing. The complete 6.3-kb expression constructs were excised from the plasmid vector by NotI restriction endonuclease digestion, gel-purified, and extracted from the gel using a Qiaquick DNA purification kit (Qiagen). DNA samples were then introduced by microinjection into the pronuclei of one-cell-stage C57BL/6 ϫ CBA murine embryos (in the laboratory of Dr. F. Jirik, University of British Columbia) to generate transgenic mice.
Mice carrying the mutated PLN transgene were identified by the presence in DNA extracted from tail tips of a 6.243-kb BamHI restriction fragment that included most of the ␣-MHC promoter sequence, the 159-bp PLN coding sequence (including the stop codon), and the 3Ј end of the human growth hormone cDNA sequence. A 1.108-kb DNA fragment, cut from the middle of the ␣-MHC promoter sequence between restriction sites SphI and EcoRI, was used as a probe for non-radioactive genomic Southern and slot-blot analysis. The probe was labeled (22) with dioxygenin-11-dUTP following the recommendations of the manufacturer (Roche Molecular Biochemicals) and hybridized to DNA on positively charged membranes, as described by Hauge and Goodman (23), except that 1% bovine serum albumin was replaced by 1% blocking reagent (Roche Molecular Biochemicals). Alkaline phosphatase activity was detected on membranes with CDP-Star as described by the manufacturer (Roche Molecular Biochemicals).
Since the endogenous ␣-MHC gene yielded a 6035-bp BamHI fragment in which 5439 bp were from the promoter region and thus were identical to the transgene, this fragment hybridized to the same probe in wild-type animals. The copy number of the construct was estimated by comparison of densitometric scans of lanes in Southern blots of DNA samples from transgenic mice with those from wt mice in which the copy number of the endogenous ␣-MHC gene was assumed to be 2.
A 500-bp PCR product that covered the junction between the ␣-MHC promoter and PLN-encoding sequences was generated using the forward primer (MHC, 5Ј-CAG CCT CTG CTA CTC CTC TTC CTG CCT GTT C-3Ј) and the reverse primer (PLN, 5Ј-GAC GTG CTT GTT GAG GCA TTT CAA TGG TG AGG-3Ј) and used for the routine identification of L37A, I40A, and wtOE transgene-positive animals.
Control Animals-Transgenic mice in which wt PLN is overexpressed 2-fold (wtOE) were used as a control for the overexpression of superinhibitory mutant PLN (8). The PLN L37A and PLN I40A transgenic mice, produced in the laboratory of Dr. F. Jirik, were primarily on a C57BL/6 genetic background owing to five backcrosses with C57BL/6, whereas transgenic mice overexpressing the murine wt PLN, produced in the laboratory of Dr. E. Kranias, were on an FVB/N background. Thus only first generation C57BL/6 ϫ FVB/N heterozygotes overexpressing either wild-type or superinhibitory transgenes were studied wherever a common genetic background was required for comparison of the effects of the transgene. In most experiments, first generation C57BL/6 ϫ FVB/N wt animals expressing normal amounts of wt PLN were included as controls.
Calcium Uptake Assays-Animals 12-16 weeks old were sacrificed by cervical dislocation. Hearts were excised and frozen within seconds in liquid nitrogen and stored at Ϫ70°C. The frozen hearts were powdered under liquid N 2 with a chilled mortar and pestle, and microsomal fractions were prepared as described previously (24), except that the homogenizing buffer contained 10 mM Tris-HCl, pH 7.4, 20 M CaCl 2 , 0.3 mM sucrose, 5 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, and Complete protease mixture. The apparent Ca 2ϩ affinity of SERCA2a was determined from measurement of the Ca 2ϩ dependence of Ca 2ϩ transport, as described previously (18). Protein concentrations were determined by the Bio-Rad method using bovine serum albumin as a standard.
Western Blot Analysis-Relative levels of expression of pentameric and monomeric PLN were determined by SDS-PAGE of measured amounts of cardiac microsomes at several dilutions, followed by densitometric analysis of Western blots using monoclonal antibody 1D11 against PLN (a gift from Drs. Robert Johnson and Edward McKenna, Merck). Binding of primary antibody was detected by peroxidase-conjugated secondary antibodies and ECL detection reagents (Amersham Pharmacia Biotech).
Cardiac Myocyte Studies-Ventricular myocyte isolation procedures were modified from a previously described protocol (25). Briefly, mice were anesthetized with sodium pentobarbital (6.5 mg, intraperitoneal); aortas of excised hearts were cannulated in ice-cold Ca 2ϩ -free Tyrode buffer, and the hearts were attached to a Langendorff perfusion apparatus. The composition of Tyrode was as follows: 118 mM NaCl, 5.4 mM KCl, 10 mM glucose, 10 mM HEPES, pH 7.4, 0.33 mM NaH 2 PO 4 , 2 mM MgCl 2 , and 30 mM taurine. Hearts were perfused at 2 ml/min at 37°C with oxygenated, Ca 2ϩ -free Tyrode for 3 min, followed by 25 M Ca 2ϩ / Tyrode containing a mixture (2:1) of type 2 and type 1 collagenase (Worthington) totaling 0.5 mg/ml and bovine serum albumin (1 mg/ml). When the heart became flaccid, the left ventricle was cut into 6 -8 pieces, pipette-dissociated, and filtered through 250-m nylon mesh. Cells were washed sequentially in 100 M Ca 2ϩ , 200 M Ca 2ϩ , and 1 mM Ca 2ϩ and resuspended in 1.8 mM Ca 2ϩ -Tyrode solution. Typical isolations yielded at least 30% rod-shaped cells in 1 mM Ca 2ϩ -Tyrode. Ventricular myocytes, suspended in 1.8 mM Ca 2ϩ -Tyrode, were placed in a Plexiglas chamber on the stage of an inverted epifluorescence microscope (Nikon Diaphot 200). Myocytes were stimulated at 0.5 Hz with a Grass S5 stimulator (pulse duration of 4 ms), and a video edge detector (Crescent Electronics) was used to track myocyte contractions. Signals were recorded and analyzed using a Delta Scan-1 dual-beam spectrophotofluorometer and Felix software (Photon Technologies Inc.). Myocyte length, fractional shortening, shortening rate (ϩdL/dt), and relengthening rate (ϪdL/dt) were determined for an average of nine myocytes from each heart. A single heart was considered as a single sample. The average age of the mice was 17.6 weeks.
Microsurgical Methods and in Vivo Hemodynamic Measurements-Mice were anesthetized using a mixture of ketamine (50 -100 mg/kg) and xylazine (3-6 mg/kg) by intraperitoneal injection. After achieving full anesthesia, mice were placed in a supine position. The right carotid artery was isolated and cannulated with a polyethylene tube (PE-200) stretched to an outer diameter of 150 -300 m and a 20 -30-m wall thickness. The polyethylene catheter was connected to a TXD-310 low compliance pressure transducer (MicroMed, Louisville KY) and amplified by a blood pressure analyzer (BPA model 300, MicroMed, Louisville, KY). The pressure recording system had a 28 Ϯ 2-Hz roll-off frequency measured at Ϫ3 db, which might be expected to produce some distortion of our pressure recordings. After insertion of the catheter into the carotid artery, the catheter was advanced into the aorta and then into the left ventricle to record the aortic and ventricular pressures. The parameters measured and analyzed were heart rate, aortic pressure, left ventricular (LV) systolic pressure, LV diastolic pressure, and the maximum and minimum first derivatives of the LV pressure (ϩdP/dt max and ϪdP/dt max , respectively). Both ϩdP/dt max and ϪdP/dt max give a measure of the contractility of the heart, whereas ϪdP/dt max is also strongly influenced by the rate of pressure relaxation. To complement the ϪdP/dt max estimates, the time constants for pressure relaxation were also estimated by fitting the relaxation phase of the pressure traces (from 30 to 80% of the peak systolic pressure) to a monoexponential.
Some pressure recordings (n ϭ 3 or 4 in each group) were also made using a 1.4 French Millar catheter (Millar Inc., Houston) connected to an amplifier (TCP-500, Millar Inc.), The pressure signals were filtered at 300 Hz (3 db), recorded using an A/D data acquisition board (2801A, Data Translation, Marlboro, MA) at a 1-kHz sampling rate and stored in the computer for later analysis. No significant difference in peak pressure was observed between the two recording methods. However, the maximum ϮdP/dt was about 18% lower for the fluid-filled catheters than the Millar-based measurements. Only the pressure recordings made using the fluid-filled catheters were included in the analyses reported below.
In Vivo Echocardiographic Assessment of Cardiac Function-Mmode and Doppler echocardiography were performed for non-invasive assessment of left ventricular (LV) function and dimensions, using previously described methods (26). Briefly, mice were lightly anesthetized with 2.5% avertin (0.01 ml/g) intraperitoneally and were allowed to breathe spontaneously. The chest was shaved, acoustic coupling gel was applied to the left precordium, and a warming pad was used to maintain normothermia. Mice were imaged in a shallow left lateral decubitus position using an Interspec Apogee X-200 ultrasonograph with a 9-MHz imaging and a 5-7.5-MHz pulsed-wave Doppler transducer. Studies were performed at base line and after the administration of 2.0 g/g isoproterenol intraperitoneally. Left ventricular percent fractional shortening, velocity of circumferential shortening corrected for heart rates at matched heart rates, and end diastolic wall thickness/ cavity radius ratio were calculated as described previously (8,26). LV mass was calculated using M-mode LV measurements according to American Society of Echocardiography conventions and the modified American Society of Echocardiography-cube LV mass equation (27), as described previously (8,26). The feasibility of the use of cardiac ultrasound in the in vivo assessment of LV mass in mice has been shown in a necropsy validation study (28).
Phosphorylation of PLN-Hearts were excised and frozen within seconds in liquid nitrogen and stored at Ϫ70°C. The frozen hearts were powdered under liquid N 2 with a chilled mortar and pestle, and microsomal fractions were prepared as described under "Calcium Uptake Assays," except that 50 mM sodium fluoride and 100 mM sodium phosphate, pH 7.4, were added to the homogenizing buffer to inhibit phosphatase activity. Relative levels of phosphorylation of pentameric and monomeric PLN were determined by SDS-PAGE of cardiac microsomes at several dilutions, followed by densitometric analysis of Western blots using antibodies specific against PLN-peptide-Thr(P) 17 and PLN-peptide-Ser(P) 16 (rabbit polyclonal antibodies to PLN phosphoserine peptide or PLN phosphothreonine peptide (number 010-12/010-13, Phos-phoProtein Research). Binding of primary antibody was detected by alkaline phosphatase-conjugated secondary antibodies. Alkaline phosphatase activity was detected on membranes with CDP-Star as described by the manufacturer (Roche Molecular Biochemicals).
Statistical Analysis-Data are presented as mean Ϯ S.E. The number of mice (n) used in each study are indicated. Statistical analysis was performed by t test between wild-type and transgenic mice. p Ͻ 0.05 was considered statistically significant.

Construction and Identification of Transgenic Animals-To
create the PLN L37A founder mice, 290 embryos were injected; 142 embryos were implanted, and 10 pups were born. One of two positive pups died early, and the other founded the line used in this study. For PLN I40A mice, 286 embryos were injected, 159 embryos were implanted, and 20 pups were born. Of these, 5 were positive, but 3 did not transmit or showed poor expression levels; 1 line could not be propagated, and the other founded the line used in this study. This resulted in one trans-genic line for each of the PLN mutant constructs that could be propagated for further analysis. However, the L37A and I40A lines have similar phenotypes in vitro and in vivo, so that creation of two comparable lines from two comparable mutations makes insertional mutagenesis and inter-line variability of phenotypes unlikely. The founders of both transgenic lines and their progeny did not appear to differ from their littermates in either behavior or reproductive ability.
In the Southern blots shown in Fig. 1B, the presence of a heavily stained 6243-bp BamHI fragment representing most of the ␣-MHC promoter, the PLN coding sequence, and part of the human growth hormone sequence in the transgene vector was used to identify transgenic mice. In control lanes, a weakly stained 6035-bp BamHI fragment, which overlaps the mobility of the 6243-bp transgene fragment, represents the endogenous murine ␣-MHC gene, and the level of staining is equivalent to two copies of the gene. The level of hybridization of the probe to the endogenous ␣-MHC gene fragment in wt mice compared with the level of hybridization of the probe to the transgenic ␣-MHC gene fragment revealed that the L37A and I40A transgenes were integrated into the mouse genome at copy numbers of about 20 and 15, respectively.
Overexpression of Monomeric PLN-The expression of monomeric mutant PLN in cardiac microsomes from transgenic mice was assessed by Western blotting with the PLN-specific monoclonal antibody, 1D11. In previous experiments on the transient expression of L37A and I40A in HEK-293 cells, L37A was 94% monomeric, whereas I40A was 100% monomeric (13). Thus any increase in PLN monomer content in L37A and I40A transgenic mice could be attributed to the expression of these mutant transgenes. Data presented in Fig. 2 and Table I show that content of pentameric PLN was constant in microsomal fractions from wt, L37A, and I40A transgenic lines, whereas the content of pentameric PLN was increased about 2-fold in the wtOE line, as described previously (8). The content of monomeric PLN was also constant for wt and wtOE lines at about 5.5-5.7% of total PLN. In the L37A line, however, the distribution between monomer and pentamer was increased to 40.5% and in I40A to 46.2% of total PLN (Table I) Ca 2ϩ Uptake by Sarcoplasmic Reticulum-Ca 2ϩ dependence of Ca 2ϩ uptake in cardiac microsomes from transgenic and wt mice was used to assess the effect of overexpression of L37A and I40A on SERCA2a function (Table II). Apparent Ca 2ϩ affinity, expressed as K Ca in pCa units, was lowered from pCa 6.22 in wild-type and 6.12 in mice overexpressing wild-type PLN to pCa 5.81 in L37A and pCa 5.72 in I40A. No change was observed in V max . The K Ca values for SERCA2a activity in transgenic and wild-type mouse heart preparations were strikingly similar to the values obtained in co-expression experiments in HEK-293 cells (13), demonstrating that L37A and I40A mutant PLN, expressed in transgenic mice, retain the ability for superinhibitory interaction with SERCA2a. Thus these transgenic animals provide an excellent model for analysis of the physiological role of highly inhibitory PLN.
Characterization of Isolated Cardiomyocytes-Isolated left ventricular myocytes from L37A and I40A mice exhibited a 2-fold depression of mechanical function compared with wt control myocytes (Table III). Fractional shortening of I40A and L37A myocytes was decreased (p Ͻ 0.05) by 46 and 33%, whereas rates of contraction (ϩdL/dt) of I40A and L37A myocytes were depressed by 53 (p Ͻ 0.05) and 40% (p Ͻ 0.05), respectively. Finally, myocyte relengthening rates (ϪdL/dt) were depressed by 63 (p Ͻ 0.05) and 47% (p Ͻ 0.05) in I40A and L37A myocytes, respectively. These results support in vitro data and suggest that I40A mutant PLN is a more potent inhibitor of SERCA2a than L37A. By contrast, the mechanical function of myocytes from wtOE mice expressing 2-fold elevated levels of pentameric wt PLN showed no functional depression compared with wt cells. Since one of our previously published studies (8) showed significant decreases in mechanical parameters of myocytes overexpressing wt PLN in the FVB/N genetic background (at 0.25 Hz), we reassessed the functional parameters of myocytes from this transgenic model. We observed significant decreases in fractional shortening (18%), ϩdL/dt (31%), and ϪdL/dt (36%) in myocytes (at 0.5 Hz) overexpressing wt PLN in the FVB/N genetic background, consistent with our previous report (9). However, the background of the mice used in the current study reflects an F1 cross between FVB/N and C57BL/6. This suggests that the lack of functional depression in wtOE myocytes may be a function of the altered genetic background.
Hemodynamic Properties-Hemodynamic characteristics of L37A and I40A mice were examined at 8 -12 weeks of age. The results are summarized in Table IV. There were no significant differences in mean aortic pressures and peak systolic left ventricular pressures for the L37A and I40A mice compared with age-matched littermate control mice. Consistent with Ca 2ϩ uptake measurements and cell shortening studies described above, the magnitude of both left ventricular ϩdP/dt max and ϪdP/dt max was significantly depressed in L37A and I40A transgenic mice compared with controls. Neither heart rates nor end diastolic pressures were significantly different between the groups, suggesting that the differences in dP/dt max between the groups were not due to changes in preload or heart rate. These differences in hemodynamic parameters between the groups were not the result of the limited frequency response of our pressure recording system, since signal filtering will tend to reduce (not enhance) the observed differences between L37A/ I40A and wt control mice. As expected from the differences in ϪdP/dt max between the groups, the relaxation times () of the pressure traces were significantly prolonged in the I37A and I40A mice compared with either wt or wtOE littermate controls.
As with our studies of single myocytes and our echocardiographic measurements described below, measurements carried out with wtOE mice were indistinguishable from control mice. This was unexpected, since 2-fold overexpression of wild-type PLN was found to be inhibitory when compared with wild-type controls in two earlier studies (8,19). However, our findings may be a consequence of the genetic background of all of the animals in our study (C57BL/6/FVB/N), which was different from the genetic background in the earlier studies (FVB/N).
Echocardiographic Properties-Echocardiographic studies revealed that, when compared with wt or wtOE mice, L37A and I40A mice displayed significant reductions in left ventricular fractional shortening and velocities of circumferential shortening (Table V). The relative magnitude of these differences in the extent and velocity of shortening between the groups was remarkably similar to the reductions in contractility as measured by ϩdP/dt max and ϪdP/dt max . Ejection times were also prolonged in the L37A and I40A mice, relative to wt and wtOE mice, but this difference did not reach significance for I37A mice. The reductions in indices of cardiac contractility were accompanied by increases in left ventricular mass to body weight ratios and left ventricular dimensions in both L37A and I40A mice, when compared with wtOE mice. A 1.47-fold increase in LV mass for the L37A mice and a 1.44-fold increase in LV mass for the I40A mice suggest that compensatory hypertrophy occurred in these two lines of mice. Reduced contractility in L37A and I40A mice, compared with wtOE mice, could be completely overcome by treatment with isoproterenol (Table  VI).
PLN Phosphorylation in Transgenic Animals-The lack of mortality in most of the L37A and I40A transgenic mice suggests that compensatory mechanism(s) may be activated to counteract the superinhibitory effects of mutant PLN. The most likely compensatory change would result in a sustained increase in PLN phosphorylation, leading to dissociation of PLN from the inhibited PLN⅐SERCA2a complex. To test this hypothesis, the content of phospho-PLN in pooled hearts from three animals was determined by its reaction with anti-Ser(P) and anti-Thr(P) antibodies. Increased phosphorylation was observed in both pentameric and monomeric PLN in all three of the transgenic samples (Fig. 3). Changes could also be deduced from the alterations in mobility that are characteristic of phosphorylation following immunoblotting with antibody 1D11 (Fig. 2).
The key question, however, was whether the increase in phosphorylation was greater than the increase in PLN expression. The most straightforward measurement of the increase in phosphorylation levels was through measurement of the phos-  Table I. Arrows indicate pentameric (p) and monomeric (m) PLN forms. Multiple bands in monomeric PLN are likely due to different levels of phosphorylation. phorylation of pentameric PLN, since the content of pentameric PLN in L37A or I40A was similar to wt ( Fig. 2 and Table I), and pentameric PLN was overexpressed by 2-fold in wtOE. By using the same microsomal preparations as those used in Fig.  2, permitting a correction for the relative amount of pentameric protein expressed, the phosphorylation of pentameric Ser 16 was found to be increased 1.86-fold in L37A and 2.39-fold in I40A but to be reduced to 0.74-fold in wtOE, when compared with wt Ser 16 phosphorylation. Thr 17 phosphorylation was found to be increased 1.24-fold in L37A and 1.63-fold in I40A but to be reduced to 0.58-fold in wtOE, when compared with wt Thr 17 phosphorylation. Data obtained by analysis of the phosphoryl content of the monomeric forms gave similar results but were less accurate because of the large multiplication factors required. These results demonstrate that compensatory mechanisms for enhancement of the phosphorylation of both Ser 16 and Thr 17 are activated in the L37A and I40A lines but not in the wtOE line.

DISCUSSION
Inhibitory interactions between PLN and SERCA2a establish the range of myoplasmic Ca 2ϩ concentrations over which elevated Ca 2ϩ can shift SERCA2a Ca 2ϩ transport from base line to full activity. When PLN is phosphorylated, these inhibitory interactions are abolished, and the Ca 2ϩ activation range is shifted, temporarily, toward higher apparent Ca 2ϩ affinity. Following the ablation of PLN, the Ca 2ϩ activation range is shifted chronically toward higher apparent Ca 2ϩ affinity, and apparent Ca 2ϩ affinity is unaffected by ␤-adrenergic stimulation. Under these conditions, an enhanced rate and extent of Ca 2ϩ uptake into the sarcoplasmic reticulum can be correlated with enhanced cardiac contractility and with a decrease in relaxation time. These sustained changes in the performance of the heart do not induce cardiomyopathy, and the PLN-null animals enjoy a normal life span (29).
In earlier studies (13,16), we showed that several dominantacting mutant forms of PLN, including L37A, I40A, and N27A, gain inhibitory function by inducing dramatic decreases in the apparent affinity of SERCA2a for Ca 2ϩ . We predicted that this shift in Ca 2ϩ affinity might compromise cardiac function, leading to cardiomyopathy (13). In order to test this hypothesis, we have created animal models in which the dominant, highly inhibitory L37A and I40A PLN mutant proteins are overexpressed. We obtained one transgenic line for each PLN mutant, but since the L37A and I40A mutants have similar phenotypes in vitro and in vivo, creation of two comparable lines from two comparable mutations would rule out insertional mutagenesis and inter-line variability of phenotypes and, in this way, would be comparable to creation of more than one line from a single mutant. The L37A mice showed a 1.54-fold increase in overall PLN expression and an 11-fold increase in monomeric PLN expression; the I40A mice showed a 1.71-fold increase in overall PLN expression, and a 13-fold overexpression of monomeric PLN in cardiac microsomes. Measurement of the Ca 2ϩ dependence of Ca 2ϩ uptake in these microsomes recreated with high fidelity the observations made earlier in HEK-293 cells coexpressing SERCA2a and PLN (13). Thus we were successful in transferring the experimental system from heterologous cell culture to intact animals where physiological correlations could be measured. Chu et al. (19) investigated the effect of transgenic overexpression of the C41F mutant of PLN. This mutant is anomalous in that it was largely monomeric in SDS gels (17) but did not gain inhibitory function in in vitro tests (18). In line with the in vitro observations, the transgenic overexpression of the C41F mutant in mice did not lead to superinhibition of Ca 2ϩ -ATPase function in isolated cardiac microsomes (19). The level of inhibition was slightly greater than that observed in wild-type mice but less than that observed in mice expressing a 2-fold excess of wild-type PLN. Analysis of cardiomyocyte mechanics and Ca 2ϩ kinetics indicated that the inhibitory effects of C41F mutant PLN overexpression were also less pronounced than those of 2-fold overexpression of wild-type PLN. Thus, the C41F mutant mouse has not proven to be a suitable model for the investigation of the effects of overexpression of highly inhibitory monomeric forms of PLN such as the mutants L37A and I40A.
Measurements of myocyte function provide functional correlates of changes in Ca 2ϩ handling by L37A and I40A overexpression. As expected from the Ca 2ϩ uptake studies, the extent of shortening and the rates of shortening and relengthening for wt or wtOE myocytes, which were not different from each transgenic lines Equal aliquots of microsomal fractions from three pooled hearts from each mouse line (wt; L37A: I40A; wt OE) were separated by SDS-PAGE, followed by densitometric analysis of Western blots using monoclonal antibody 1D11 against PLN (see Fig. 2). Binding of primary antibody was detected by peroxidase-conjugated secondary antibodies and ECL detection reagents. The content of pentameric PLN in each line was normalized to pentameric wt (1.0). The content of monomeric PLN is presented relative to pentamer for each line and as a percent of total PLN in that line. The relative amount of pentameric Ser 16 and Thr 17 phosphorylation in each line (normalized to 1.0 for pentameric wt) was determined by densitometric analysis of Western blots using antibodies specific against PLN-peptide-Ser(P) 16 or PLN-peptide-Thr(P) 17 (rabbit polyclonal antibodies to PLN phosphoserine peptide or PLN phosphothreonine peptide). Binding of primary antibody was detected by alkaline phosphataseconjugated secondary antibodies with CDP-Star, as described by the manufacturer. other, were far greater (11%) than either L37A (7.7%) or I40A (6.2%) myocytes. Consistent with the myocyte studies, both ϩdP/dt max and ϪdP/dt max were found to be depressed in hemodynamic measurements carried out with the L37A and I40A mice, establishing that contractility is impaired in these mice. Associated with ϩdP/dt max changes, there were also reductions in the fractional ventricular shortening, ejection time, and velocity of circumferential shortening. Modulation of these ventricular parameters are all expected to result from reduced [Ca 2ϩ ] i transient amplitude as a result of reduced sarcoplasmic reticulum Ca 2ϩ uptake and release following impairment of SERCA2a function. Diminished contractility coincided with a nearly 50% increase in LV heart weight to body weight ratio, a significant increase in LV mass, and an increased chamber size in both L37A and I40A mice (Table IV). These morphological changes are anticipated since cardiac hypertrophy and remodeling is a common compensatory response of the myocardium to alterations in cardiac function or increased hemodynamic loads (30 -32). In our hemodynamic recordings, significant slowing of the rate of pressure relaxation was also observed, as reflected in the time constants for pressure relaxation () and ϪdP/dt max . These changes corresponded with differences in myocyte   FIG. 3. Immunoblotting of pentameric and monomeric forms of PLN with antibodies specific against PLN-peptide-Ser(P) (left) and PLN-peptide-Thr(P) (right). An aliquot of the same cardiac microsomes (5 g of total protein) obtained from the pooled hearts of three L37A, I40A, and wtOE transgenic mice that were used in the experiment described in Fig. 2 was subjected to SDS-PAGE on 15% acrylamide prior to immunoblotting. Further details are provided in the legend to Table I.  ϪdL/dt and with rates of Ca 2ϩ uptake in sarcoplasmic reticulum vesicles and are proposed to result from a reduced rate of Ca 2ϩ reuptake following inhibition of SERCA2a. Although this might be expected to produce elevated diastolic pressures in L37A and I40A mice, this was not observed. Reduced contractility in L37A and I40A mice, compared with wtOE mice, was completely overcome by treatment with isoproterenol (Table V). These results indicate that phosphorylation of PLN was not maximal in these animals and that considerable potential for ␤-adrenergic stimulation remained in the L37A and I40A mice. We used antibodies specific against Ser 16 -and Thr 17 -PLN peptides to show that phosphorylation of both pentameric and monomeric PLN was increased in both the L37A and I40A lines but was decreased in the wtOE line ( Fig.  3 and Table I). The increases in Thr 17 phosphorylation in L37A and I40A might be attributed to activation of Ca 2ϩ calmodulin kinase, but the increases in Ser 16 phosphorylation for L37A and I40A would be due to protein kinase A, implying adrenergic activation. These findings suggest that the inhibitory effects of overexpression of L37A and I40A on Ca 2ϩ cycling might be reduced in the transgenic mice by compensatory phosphorylation of PLN. Further studies of compensatory mechanisms will be useful in assessing cross-talk between Ca 2ϩ handling proteins of the sarcoplasmic reticulum and other regulatory proteins in maintaining contractile parameters under basal and isoproterenol-stimulated conditions in the hearts of L37A and I40A mice.
An unexpected finding was that wtOE transgenic animals did not differ in any functional properties from wt animals when the genetic background was an F1 cross between FVB/N and C57BL/6. In earlier studies in which comparisons were made using inbred animals with an FVB/N background, the 2-fold overexpression of wt PLN was found to be inhibitory (8,19). This suggests that the lack of functional depression in wtOE myocytes may be a function of the altered genetic background; the effect of recessive traits is maximized in inbred lines and minimized in hybrid animals.
In summary, we have found that overexpression of two monomeric, superinhibitory forms of PLN results in significant depression of contractility in mouse hearts. Our findings are particularly relevant, since they are the corollary of the recent findings demonstrating that ablation of PLN can rescue the cardiomyopathic phenotype caused by LIM protein disruption (33). Both of these studies highlight the critical role of SERCA2a function in normal and myopathic hearts.