OUP user menu

Enhanced phosphorylation of phospholamban and downregulation of sarco/endoplasmic reticulum Ca2+ ATPase type 2 (SERCA 2) in cardiac sarcoplasmic reticulum from rabbits with heart failure

Susan Currie, Godfrey L. Smith
DOI: http://dx.doi.org/10.1016/S0008-6363(98)00241-7 135-146 First published online: 1 January 1999

Abstract

Objectives: To assess the phosphorylation of myocardial phospholamban (PLB) and quantify protein levels of PLB and sarco/endoplasmic reticulum Ca2+ ATPase type 2 (SERCA 2) in a rabbit model of heart failure. Furthermore, to correlate these parameters with the rate of Ca2+ uptake into sarcoplasmic reticulum (SR) vesicles. Methods: Heart failure in the rabbit was indicated by the pronounced ventricular contractile dysfunction accompanied by post-mortem evidence of lung and liver congestion 8 weeks after a coronary artery ligation procedure. Phosphorylation of PLB was measured by reduced mobility of the phosphorylated forms on Tris–glycine gels. Phosphoserine and phosphothreonine-specific antibodies against PLB were used to determine the phosphorylated residues. Immunoblotting combined with densitometry was used to assess PLB and SERCA 2 levels. Finally, oxalate-supported Ca2+ uptake into SR vesicles was studied using the fluorescent indicator Fura-2. Results: The phosphorylation state of PLB was significantly higher in myocardium isolated from left ventricles of heart failure rabbits (8.3±0.42 P-PLB) when compared with sham-operated animals (4.0±1.7 P-PLB). The kinase activity associated with SR vesicles isolated from animals with heart failure was a factor of 1.58±0.21-times higher than sham hearts, as assessed by the initial rate of phosphorylation of PLB. This higher kinase activity observed in heart failure was not completely abolished by inhibitors of either A-kinase, C-kinase or Ca2+/calmodulin-dependent protein kinase (CaM-kinase). Abundance of SERCA in heart failure myocardial homogenates was significantly less than sham values (0.68±0.11 vs. 1.74±0.27) as was PLB (0.41±0.08 vs. 0.69±0.13), similar reductions were seen in vesicle preparations. The rate constant of Ca2+ uptake into the isolated SR vesicles was lower in preparations from heart failure myocardium than from sham myocardium (2.50±0.23 ms vs. 4.43±0.3 ms). Conclusions: The higher level of phosphorylation of PLB observed in the left ventricle of rabbits with heart failure is associated with a higher intrinsic kinase activity of the SR. However, the abundance of both of SERCA 2 and PLB proteins are lower in heart failure. The net effect of these changes appears to be a reduced rate of Ca2+ uptake by the SR in heart failure.

Keywords
  • Heart failure
  • Rabbit
  • Phospholamban
  • SERCA 2
  • Phosphorylation
  • Ca2+ uptake
  • Sarcoplasmic reticulum
Abbreviations
  • SR, sarcoplasmic reticulum
  • PLB, phospholamban
  • SERCA 2, sarco/endoplasmic reticulum Ca2+ ATPase type 2
  • PKA, cAMP-dependent protein kinase
  • PKC, protein kinase C
  • CaM-kinase, Ca2+/calmodulin-dependent protein kinase
  • SDS, sodium dodecyl sulphate
  • PAGE, polyacrylamide gel electrophoresis
  • BSA, bovine serum albumin
  • EGTA, ethylene glycol bis(β-aminoethyl ether)tetraacetic acid
  • EDTA, ethylenediaminetetraacetic acid
  • PP1, phosphatase type 1

Time for primary review 34 days.

1 Introduction

The clinical syndrome of heart failure is characterised by ventricular contractile dysfunction. The syndrome can have a variety of pathological origins, yet the resulting contractile dysfunction is thought to be due to changes in function of the myocardium associated with compensatory hypertrophy. The mechanisms underlying the systolic and diastolic dysfunction in heart failure are unclear, but the prolongation and/or reduction in the amplitude of the intracellular Ca2+ transient observed in human and animal models of heart failure is thought to be a critical factor [8, 12, 13]. The altered time course and amplitude of the intracellular Ca2+ transient may be caused by an altered expression of several genes encoding proteins involved in the regulation of intracellular [Ca2+] [15]. In particular, changes in the level or function of sarcoplasmic reticulum (SR) proteins involved in Ca2+ uptake, i.e., sarco/endoplasmic reticulum Ca2+ ATPase type 2 (SERCA 2) and its regulatory protein phospholamban (PLB), may contribute to the pathophysiology of the disease [3]. There have been numerous studies examining the levels of mRNA and protein for SERCA 2 and PLB in various animal models and human heart failure. In the majority of animal models, severe heart failure is associated with reduced RNA and protein levels of SERCA 2 and PLB [3]. These models commonly involve pressure induced hypertrophy (caused by aortic banding or systemic hypertension) leading to heart failure; only a few studies have used myocardial infarction to induce heart failure [1]. In studies of human myocardium where heart failure is often attributable to a mixed aetiology, a consensus concerning SR protein levels has yet to be reached [16, 26, 34].

PLB depresses the activity of SERCA 2 via specific interaction with the protein, thus an altered stoichiometry of PLB:SERCA 2 would change the relative activity of the Ca2+ pump [19]. The few studies that have examined the stoichiometry in detail have indicated both increases and decreases in the PLB:SERCA 2 ratio in heart failure [18, 24]. The interaction between PLB and SERCA 2, and the inhibitory effect of PLB is reduced on phosphorylation. Previous work has shown that kinase phosphorylation of PLB occurs at two sites, Ser16 (A- and C-kinase) and Thr17 [by Ca2+/calmodulin-dependent protein kinase (CaM-kinase)] [32]. Despite the importance of this regulatory mechanism, there is a lack of information concerning the phosphorylation state of PLB in heart failure. The purpose of this study was to characterise the phosphorylation state of PLB and measure PLB and SERCA 2 protein levels in a rabbit model of myocardial infarction induced heart failure. In addition, the Ca2+ pump function of the SR was assessed in heart failure and the changes correlated with the biochemical measurements.

2 Methods

2.1 Preparation of animals

A rabbit coronary artery ligation model of heart failure was used for this study. The procedures, which have been described before [29, 30], were undertaken in accordance with the Animals (Scientific Procedures) Act 1986 and conformed with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1985). Experiments were performed on New Zealand white male rabbits aged approximately 12 weeks and weighing between 2.5 and 3 kg. Animals were anaesthetised with Hypnorm and maintained with halothane and nitrous oxide/oxygen. A left thoracotomy was performed and the large circumflex branch of the left coronary artery was identified and ligated approximately midway between the left atrial appendage and the cardiac apex. This gives rise to a large homogeneous infarct in view of the minimal collateral circulation in the rabbit. Sham-operated animals underwent thoracotomy with the heart manipulated in a similar fashion to the ligated group but the artery was not tied. Animals were left for 8 weeks after the operation to allow cardiac remodelling to occur. Left ventricular function was assessed following coronary artery ligation by echocardiography, the details of which are given in previous publications [29, 30]. As indicated in Table 1, after 8 weeks of coronary artery ligation the animals showed significant haemodynamic dysfunction in terms of an increased end-diastolic dimension, increased left atrial dimension and decreased ejection fraction. Furthermore, evidence of congestion as manifest by increases in lung and liver wet weight were present on post-mortem examination. The echocardiographic and post-mortem results are similar to those reported in a more extensive haemodynamic study of this model [30]and suggest this procedure produces a significant degree of heart failure in these animals.

Prior to experiments, rabbits were given a lethal injection of pentobarbital sodium (1 ml/kg) and hearts were rapidly removed and washed several times in Ringers solution [150 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 5 mM 4-(2-hydroxyethyl(-1-piperazine-ethanesulphonic acid (HEPES), pH 7.0, 10 mM glucose].

2.2 Preparation of tissue homogenates

The isolated free wall of the left ventricle (1–2 g of tissue) was isolated and minced using fine scissors and then homogenised using an ultraturrax T8 (Labortechnik) in ice-cold homogenisation buffer [0.3 M sucrose, 10 mM imidazole, pH 7.0, 30 mM histidine, 1 mM DTT, containing 1 mM phenylmethylsulphonyl fluoride (PMSF), 1 μg/ml leupeptin, 100 μg/ml aprotinin). Homogenates were subjected to a low speed spin step to remove any large pieces of tissue and the supernatant kept on ice prior to experimental analysis. Care was taken when preparing tissue from infarcted hearts to isolate only the myocardial tissue and not include any tissue from the infarct zone.

View this table:
Table 1

Comparison between sham-operated (Sham) and heart failure (HF) rabbits

 Sham (n=19)HF (n=23)p-Value
Body weight (kg)3.6±0.13.7±0.2NS
Ejection fraction (%)72.1±1.344.7±1.5<0.001
LAD (mm)11.2±0.214.4±0.7<0.001
LVEDD (mm)17.35±0.319.7±0.4<0.001
Liver wet weight (g)82.6±2.3107.2±11.1<0.05
Lung wet weight (g)11.7±0.215.2±0.5<0.05
  • LAD=left atrial dimension, LVEDD=left ventricular end-diastolic dimension.

2.3 Isolation of SR vesicles

The preparation of SR vesicles was based on a previously reported method [7]. Briefly, the left ventricular free wall was removed and was finely minced and then homogenised twice for 30 s each time using three volumes of homogenisation buffer (see Section 2.2). The homogenate was centrifuged twice at 8000 g for 20 min at 4°C to remove large particles. The supernatant from the second spin was then centrifuged at 45 000 g and the resulting pellet was resuspended in 1 volume of ice-cold precipitation buffer (same composition as homogenisation buffer except this contains 0.6 M KCl). This was left on ice for ∼30 min with occasional agitation and then was centrifuged at 2000 g for 20 min to remove myofibrillar proteins. The supernatant from this spin was then centrifuged a second time at 45 000 g for 90 min and the resulting pellet could either be stored in precipitation buffer at −70°C or prepared in assay buffer for immediate use.

2.4 PLB phosphorylation studies

Phosphorylation of SR components was performed in the presence of 2 mM ATP and 1 μM cAMP-dependent protein kinase (PKA; Boehringer Mannheim). In studies examining intrinsic kinase activity, no PKA was added. Reactions were performed at 30°C and were initiated by the addition of 50 μg (total protein) of SR preparation in assay buffer containing protease and phosphatase inhibitors (50 mM Tris, pH 7.4, 5 mM MgCl2, 0.1 mM EGTA, 1 mM PMSF, 1 μg/ml leupeptin, 100 μg/ml aprotinin, 1 μM microcystein, 50 nM calyculin A, 0.2 mM Na3VO4). Reactions were allowed to proceed for the appropriate times and were terminated by the addition of 4×sample buffer [6% sodium dodecyl sulphate (SDS), 30% glycerol, 235 mM Tris, pH 6.8, 0.005% bromophenol blue, 8 mM β-mercaptoethanol]. Where appropriate, kinase inhibitors were added to the reaction tubes. The PKA inhibitory peptide PKI (Phosphoprotein Research) was used at a final concentration of 500 nM, the protein kinase C (PKC) inhibitory peptide [19–36](Calbiochem) was used at a final concentration of 5 μM and the CaM-kinase inhibitor KN-62 (Calbiochem) was used at a final concentration of 1 μM.

In experiments examining the phosphorylation state of PLB close to in vivo conditions, papillary muscles (70–100 mg tissue) were rapidly dissected from the left ventricle of hearts that had been removed from animals immediately following anaesthetic. The tissue was immediately placed in 4×sample buffer and minced into small pieces. The sample was then homogenised with a pellet pestle homogeniser for ∼5 min. This was then centrifuged at 8000 g to remove any remaining large pieces of tissue and samples were analysed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE).

2.5 Electrophoresis and immunoblotting

SDS-PAGE was performed as described by Laemmli [21]using the Novex system with either 6% or 14% Tris–glycine gels for SERCA 2 and PLB detection respectively. Samples, prepared in 4×sample buffer were not boiled prior to electrophoresis and hence, there was no degradation of the pentameric form of PLB, which runs at approximately 25 kilodaltons (kDa), into the monomer (∼5 kDa). Gels were subjected to electrophoresis at a constant current of 25 mA for 1.25 h in either case. Proteins were transferred to nitrocellulose membranes using the Transblot semidry transfer cell system (Bio-Rad) in transfer buffer (39 mM glycine, 1.3 mM SDS, 20% methanol, 48 mM Tris, pH 9.2). The nitrocellulose sheets were blocked in wash buffer (150 mM NaCl, 1 mM EDTA, 0.1% Triton-X-100, 10 mM Tris–HCl, pH 7.4) containing 3% bovine serum albimin (BSA; Sigma). The blots were incubated overnight at 4°C with mouse anti-PLB monoclonal antibody A1 (0.0625 μg/ml), rabbit anti-phosphopeptide PS-16 polyclonal antibody (1:10 000 dilution) or rabbit anti-phosphopeptide PT-17 polyclonal antibody (1:5000 dilution) (Phosphoprotein Research) in blotting buffer (100 mM MgCl2, 0.5% Tween 20, 1% Triton-X-100, 1% BSA, 100 mM Tris/HCl, pH 7.4) containing 5% foetal calf serum (Gibco BRL). The specificity of the phosphopeptide antibodies was checked by inclusion of either 1 μM Ser16-phosphorylated peptide (RSAIRRAS*TIEY) or 1 μM Thr17-phosphorylated peptide (RSAIRRAST*IEY) at this stage in initial experiments. Membranes were then washed three times for 5 min with wash buffer and incubated for 2 h at room temperature with goat anti-mouse IgG–horseradish peroxidase conjugate diluted 1:2000 (Transduction Laboratories) or goat anti-rabbit IgG–horseradish peroxidase conjugate diluted 1:5000 (Jackson Immuno-Research) in blotting buffer. The blots were then washed a further three times and developed using the ECL detection system (Amersham). For SERCA 2 detection, the nitrocellulose membranes were blocked in the same buffer as described above but overnight at 4°C and then probed with mouse anti-SERCA 2 ATPase monoclonal antibody (IgG1) (1:4000) (Affinity Bioreagents) for 1 h at room temperature. This was followed as above with a secondary antibody, goat anti-mouse IgG–horseradish peroxidase conjugate diluted 1:2000 (Transduction Laboratories).

2.6 Protein content and densitometry studies

Protein content was determined using the Coomassie Plus protein assay (Pierce) and BSA (0.1–1 mg/ml) as standard. Quantification of the pentameric form of PLB was achieved by scanning developed immunoblots containing known amounts of total SR protein using a Bio-Rad GS-670 Imaging Densitometer linked to an Apple McIntosh Quadra 800 microcomputer. Background was subtracted by scanning equivalent sized areas of nitrocellulose that did not contain immunoreactive protein. Measurements were performed in triplicate and the average densiometric measurement was taken over the linear range of protein loading (6.26–25 μg).

2.7 SR Ca2+ uptake studies

Cardiac SR vesicles were prepared as described above and the final pellet was resuspended in 1 ml of assay buffer (100 mM KCl, 10 mM oxalate, 5 mM MgCl2, 25 mM HEPES, pH 7.0, 0.05 mM EGTA). Freshly prepared vesicles were used for uptake studies immediately after resuspension in assay buffer. Reactions were performed in a final volume of 2 ml with 100 μg vesicle preparation, 5 mM ATP and an initial [Ca2+] of 1.5 μM as assessed by the fluorescence ratio from Fura-2 (10 μM). Reactions were initiated by addition of the required volume of vesicles and the reaction mixture was stirred constantly at 30°C for the required times. The fluorescence ratio due to excitation at 340 nm and 380 nm was measured at 30 Hz using a spinning wheel spectrophotometer (Cairn Research, Kent, UK) and the signals stored on computer for later analysis. The relationship between [Ca2+] and fluorescence ratio was established with a series of calibration experiments and analysed according to Grynkiewicz et al. [10]. Under the conditions of the assay and measuring system, Fura-2 fluorescence had the following characteristics; Rmin=0.95±0.002, Rmax=13.2±0.10, β=5.7+0.2, KD=230±11 nM. This calibration was applied to the fluorescence signal to allow the time course of the change in [Ca2+] to be analysed. Single exponential decays were fitted to the [Ca2+] records using Microcal Origin 4.1 (Microcal Software, Northampton, MA, USA). Uptake was generally monitored for ∼30 min or until the ratio had reached a steady state. In all cases, the declining phase of the [Ca2+] record was well described by a mono-exponential decrease (p<0.001). Initial experiments with a range of concentrations of vesicles from sham, stock and ligated hearts (30–300 μg/ml) indicated that Ca2+ uptake followed approximately first order kinetics. Mitochondrial contribution to Ca2+ uptake measurements was assessed by a series of experiments using the mitochondrial inhibitor sodium azide (5 mM). There appeared to be no significant mitochondrial component to the kinetics of Ca2+ uptake by the SR vesicle preparation. The nature of Ca2+ leak from the SR was assessed by monitoring the increase in [Ca2+] after addition of thapsigargin. The Ca2+ leak appeared to be linear over the [Ca2+] studied (0.1–2.0 μM).

2.8 Statistical analysis

The data shown are mean±SEM. For comparison between groups, a Student's t-test was used. A value of p<0.05 was considered significant.

3 Results

3.1 Phosphorylation state of PLB in freshly isolated tissue

In an effort to assess the in vivo phosphorylation profile of PLB, gel electrophoresis was performed on crude tissue homogenates prepared from papillary muscles rapidly removed from the left ventricle of terminally anaesthetised animals. As described in Section 2, the tissue was immediately homogenised in SDS sample buffer thereby immobilising all enzyme activity and the samples subjected to SDS-PAGE. Nitrocellulose membranes were probed with the monoclonal A1 antibody to determine the phosphorylation state of PLB. Fig. 1 shows the typical phosphorylation profile of myocardial PLB found in age-matched control, sham-operated and heart failure animals. PLB from the myocardium of heart failure rabbits had a higher phosphorylation state than either control or sham-operated animals with significant phosphorylation above P5 detectable. In contrast, phosphorylation states above P5 were not detected in control or sham hearts. On average the maximum phosphorylation state of PLB isolated from failing rabbit hearts was 8.3±0.42 compared with 4.0±1.7 and 3.2±0.8 in control and sham groups respectively (see Table 2).

3.2 Intrinsic kinase activity of isolated SR vesicles

To study the pattern of phosphorylation of PLB in more detail, isolated SR vesicles were used to examine the time course of phosphorylation. Vesicles were incubated in a nominally Ca2+ free solution (3 mM EGTA) containing 5 mM ATP and phosphatase inhibitors for increasing time periods. The incubations were terminated with SDS sample buffer and subsequent immunoblotting was used to examine the time-dependent changes in phosphorylation of PLB caused by a Ca2+-independent intrinsic kinase activity associated with SR preparations. The nature of the kinase is unknown, however Ca2+-insensitive forms of both PKA [22]and CaM-dependent kinase [20]have previously been identified. Endogenous type I phosphatase (PP1) activity has also been reported [23], hence the inclusion of phosphatase inhibitors in the assay. Fig. 2A shows the results from a typical time course experiment demonstrating intrinsic kinase activity using isolated SR vesicle preparations in the absence of any added kinase. The relative intensities of staining of PLB in the two experimental groups does not reflect the relative abundance of these proteins. This is studied in detail in Section 3.4. A time-dependent increase in the levels of phosphorylated PLB was observed as indicated by the gradual mobility shift. Phosphorylation of the whole pentamer (P1–P5 forms present) is evident after a 1 min incubation period in the sham preparation. This contrasts with SR vesicle preparations prepared from heart failure animals where equivalent levels of phosphorylation are achieved earlier (∼30 s). This difference between SR vesicle preparations isolated from sham and failing hearts is further shown in Fig. 2B. Here the average maximum phosphorylation state detected at each time point is shown graphically. These results indicate that comparable amounts of SR vesicle preparations isolated from animals with heart failure have significantly higher intrinsic kinase activities. An estimate of the relative difference in kinase rate can be made by measuring the increase in phosphorylation state over the first minute of incubation. Using this method, SR vesicle preparations isolated from heart failure tissue have a kinase activity that is approximately 150% of the sham preparations (see Table 2).

Fig. 1

Examples of immunoblots showing the differential separation of the phosphorylated and unphosphorylated forms of PLB at different protein loads prepared from freshly isolated myocardium from control, sham-operated and heart failure myocardium. The phosphorylation state P1–P10 is shown as a calibration bar on the right hand side of the diagram.

View this table:
Table 2

Comparison of results between age-matched controls (control), sham-operated (Sham) and heart failure (HF) rabbits

Type of experimental measurementExperimental preparationControlShamHFp-Value (Sham vs. HF)
PLB phosphorylation stateHomogenate(n=4)(n=5)(n=6) 
  Ejection fraction (%) 72.1±1.344.7±1.5<0.001
  Maximum P-PLB 4.0±1.73.2±0.88.3±0.42<0.001
PLB phosphorylation rateSR vesicles(n=7)(n=5) 
  Ejection fraction (%) 73.5±3.145.7±6.0<0.001
  P incorporated/min (1st min) 1.0±0.221.58±0.21<0.05
SERCA 2 proteinHomogenate(n=4)(n=4)(n=4) 
  Ejection fraction (%) 74.1±3.946.6±6.1<0.001
  Density (/μg protein) 1.51±0.231.74±0.270.68±0.11<0.05
PLB proteinHomogenate(n=4)(n=4)(n=4) 
  Ejection fraction (%) 72.3±2.942.4±6.8<0.001
  Density (/μg protein) 0.83±0.270.69±0.130.41±0.08<0.05
Ca2+ uptakeSR vesicles(n=5)(n=5)(n=7) 
  Ejection fraction (%) 77.2±1.842.3±2.0<0.001
  Rate constant (ms−1/100 μg) 4.90±0.24.43±0.32.50±0.23<0.001
  • Five groups of measurements are summarised under separate headings with the associated ejection fractions for each group. PLB phosphorylation state is expressed as the average maximum phosphorylation state observed at 20 μg total protein. PLB phosphorylation rate is a measure of the increase in phosphorylation state of PLB over the first minute of incubation with ATP. SERCA 2 and PLB abundance measurements were made using an imaging densitometer. Immunoblots were analysed in triplicate. Ca2+ uptake rates are expressed as rate constants of the exponential decline of extravesicular [Ca2+].

3.3 Properties of the intrinsic SR kinase

An insight into the nature of the intrinsic SR kinase was provided by examining the sites on PLB phosphorylated by kinase activity. The monoclonal A1 antibody staining of PLB is not phosphorylation site-specific and provides no information regarding phosphorylation of specific residues. Therefore the phosphorylation site-specific antibodies against the phosphorylated Ser16 site on PLB (PS-16) and the phosphorylated Thr17 site (PT-17) were used. Fig. 3 shows an example of sequential antibody staining of the same time course of PLB phosphorylation from sham-operated and heart failure rabbit SR vesicles. Samples that had been transferred to nitrocellulose membranes were first probed with the A1 antibody. After washing off the antibody, the same membrane was reprobed with PS-16, and then washed and the process repeated for the PT-17 antibody. Fig. 3A shows the series of blots typical of PLB identified in SR vesicles prepared from sham-operated hearts. The A1 antibody immunoblot (top) indicates that over 30 min PLB was progressively phosphorylated, while after 60 min incubation some dephosphorylation was noted. PS-16 antibody staining (Fig. 6A, middle) revealed an increase in Ser16 phosphorylation over 30 min, reaching a maximum after approximately 10 min. The Thr17 phosphorylation was absent in some samples, but when present as it is in Fig. 3A (bottom), it appeared to occur over a longer time course showing no evident peak after 60 min. These results suggest that over the first 30 min of incubation Ser16 phosphorylation is dominant, while Thr17 phosphorylation, when present, occurs later. These observations agree with previous reports on the chronology of PLB phosphorylation [17].

Fig. 2

Panel A, immunoblots illustrating the time course of the progressive increase in phosphorylation state of PLB after addition of ATP to SR vesicle preparations prepared from sham-operated and heart failure animals. Panel B, mean maximum phosphorylation state of PLB in the two experimental groups (±SEM) at the time intervals shown. Closed symbols, sham-operated; open symbols, heart failure.

As shown in Fig. 3B, SR preparations from heart failure animals showed phosphorylation of both Ser16 and Thr17 which was evident almost immediately following incubation with ATP. Extra time points were included to highlight the temporal changes in the site of phosphorylation. There is an initial peak in Ser16 phosphorylation at ∼2–5 min, followed by a peak in Thr17 phosphorylation at ∼20 min. These results suggest that the higher phosphorylation rates observed in SR vesicles isolated from failing myocardium is caused by increased rates of phosphorylation at both Ser16 and Thr17 sites with no evidence of one site being phosphorylated predominately more than another. This suggests that intrinsic kinase activity associated with the SR has both A-kinase and CaM-kinase characteristics and that both forms of activity are upregulated in heart failure.

In an effort to characterise the enzyme activity responsible for the intrinsic activity of the SR vesicle preparations, the vesicles were incubated in a series of specific peptide inhibitors and the progress of PLB phosphorylation monitored. Fig. 4A shows that in sham-operated SR vesicle preparations, inhibition of either PKA (with PKI) or PKC (with PKC inhibitor peptide) substantially reduces the intrinsic kinase activity. However, the intrinsic kinase activity was insensitive to treatment with the CaM-kinase inhibitor KN-62.

In contrast to the results from sham hearts, the higher intrinsic kinase activity of SR vesicle preparations prepared from failing hearts was insensitive to treatment with PKI while PKC inhibitory peptide only partially inhibited PLB phosphorylation (Fig. 4B). This suggests that in failing hearts a kinase other than PKA may act along with PKC to phosphorylate PLB in heart failure. Recently, a Ca2+-independent form of the CaM-kinase (δ-CaM-kinase) has been reported in cardiac SR [4]and this kinase may contribute to the higher kinase activity observed in failing hearts. However, as shown in Fig. 4B, treatment with the selective CaM-kinase inhibitor, KN-62, had little effect on the phosphorylation associated with failure preparations. Therefore, the identity of this intrinsic kinase(s), is still uncertain and currently under further investigation.

3.4 Quantification of protein levels

Tissue homogenates were prepared from the left ventricle of sham-operated animals, heart failure animals and age-matched controls. As described in Section 2.5, SDS-PAGE was used to isolate PLB and SERCA 2 which were identified using specific antibody staining. Fig. 5 shows representative immunoblots obtained with a range of protein loads for PLB and SERCA 2 and indicates that immunoreactive protein levels are reduced. Quantitation of these blots was performed by densitometric analysis, and a range of protein loads was used to ensure that the amount of protein detected was within the linear range of the measurement system. Average densitometry measurements taken from the three groups of animals are shown in Table 2 with the associated mean ejection fraction that apply to the subgroups of animals used to measure SERCA 2 or PLB abundance. In a limited set of experiments, SERCA 2 protein abundance was measured in SR vesicle preparations. On average SERCA levels in vesicles isolated from the myocardium of heart failure animals were 53±9% (n=3) of those from sham hearts. Similar measurements of PLB abundance indicated that vesicles from failing myocardium had 51±8% (n=5) of the PLB that occurred in sham hearts. Thus in both tissue homogenates and SR vesicle preparations an increased phosphorylation of PLB and decreased abundance of PLB and SERCA 2 have been observed in heart failure. The combined effect of these changes on Ca2+ uptake by the SR was studied in a separate set of experiments.

Fig. 3

Sequential immunoblots examining the time course of phosphorylation of PLB in SR vesicle preparations. As labelled on the right hand side of the diagram, the gels were sequentially probed with A1 antibody (A1), the site-specific antibodies against PLB phosphorylated at serine 16 (PS-16) and the site-specific antibodies against PLB phosphorylated at threonine 17 (PT-17). The results were obtained from SR vesicle preparations prepared from a sham-operated (Panel A) and heart failure (Panel B) animal.

Fig. 4

Sample immunoblots showing the effects of various kinase inhibitors on the progressive phosphorylation of PLB with incubation time. SR vesicles were isolated from sham-operated (panel A) and heart failure (panel B) animals. PLB is probed with A1 antibody. The kinase inhibitor used is shown below each gel, 500 nM protein kinase C inhibitory peptide (PKI); 5 μM protein kinase C (PKC) inhibitory peptide; 1 μM CaM-kinase inhibitor (KN-62). These results are representative of four other experiments.

3.5 SR vesicle Ca2+ uptake studies

The Ca2+ pump activity of SERCA 2 was assessed by measuring the rate of Ca2+ uptake into oxalate-loaded SR vesicles prepared from the left ventricles of heart failure, sham-operated and age-matched control hearts. Ca2+ uptake was initiated by adding an aliquot of vesicles to a solution containing 5 mM ATP and 1.5 μM Ca2+ (0.05 mM EGTA) at 30°C (see Section 2.7). As illustrated in Fig. 6A, addition of SR vesicles (100 μg protein) isolated from the left ventricle of sham-operated hearts caused a decrease in the extravesicular [Ca2+] with a half time (t1/2) of approximately 180 s. The time course of Ca2+ uptake into vesicles isolated from sham hearts contrasts with the records shown in Fig. 6B recorded after addition of an equivalent amount of SR vesicles (100 μg protein) isolated from the left ventricle of a failing heart. In this case the rate of decrease of [Ca2+] was slower (t1/2=330 s) and a higher steady state [Ca2+] was observed. Fig. 6C illustrates the time course of the decline of [Ca2+] on addition of three times the initial amount of SR vesicles (300 μg) from the same ligated heart. The rate of [Ca2+] uptake observed under these conditions was comparable to that detected in vesicles from sham hearts (t1/2=198 s). As illustrated in Fig. 6A–C, the decrease in [Ca2+] was well described by a mono-exponential decline between 1.5 μM and 0.1 μM. On average, SR vesicles prepared from sham-operated hearts and control hearts exhibited similar rates of Ca2+ uptake and therefore similar rate constants of exponential decline (4.9±0.2 ms−1 and 4.4±0.3 ms−1 respectively). In contrast, preparations from ligated hearts showed a markedly slower rate of uptake (2.5±0.23 ms−1). These results and the accompanying in vivo assessment of cardiac function are summarised in Table 2. The relationship between SR vesicle concentration and rate constant for sham and heart failure groups is shown in Fig. 6D. In both experimental groups, the relationship appears approximately linear up to 100 μg/ml total protein, but the rate constant values at 300 μg/ml protein are lower than expected for a linear relationship. However, at all protein concentrations, the Ca2+ uptake rate constants measured in the heart failure group were approximately 50% of that in the sham group.

Fig. 5

Panel A, immunoblots showing the protein-dependency of antibody detection of SERCA 2 in homogenates prepared from sham-operated and heart failure animals. Homogenates were subjected to SDS-PAGE (6% gel) and immunoblot probing with the anti-SERCA 2 monoclonal antibody (IgG1). Protein loading (μg) is shown above the appropriate lane. Panel B, immunoblots showing the protein-dependency of antibody detection of SERCA 2 in homogenates prepared from sham-operated and heart failure animals. Homogenates were subjected to SDS-PAGE (14% gel) and immunoblot probing with the monoclonal A1 antibody specific for PLB. Protein loading (μg) is shown above the appropriate lane.

4 Discussion

This study describes a series of measurements designed to assess the status of myocardial SERCA 2 and PLB in an infarct model of heart failure in the rabbit. While there have been numerous studies on SR function in both human heart failure and animal models of hypertrophy, little information is available concerning the phosphorylation state of cardiac PLB in heart failure.

4.1 Evidence of myocardial contractile dysfunction and organ congestion in the heart failure group

The echocardiographic findings presented in this study revealed significant changes in cardiac function and dimensions as a result of chronic myocardial infarction secondary to coronary artery ligation (see Table 1). Ejection fraction was markedly reduced in the heart failure group. There was significant left ventricular dilatation analogous to that seen in human heart failure as a result of volume overload caused by impaired left ventricular function. Consistent with this was the increased left atrial dimension indicating raised filling pressure and possible mitral regurgitation. The significantly increased lung and liver weights indicate organ congestion suggesting a state of decompensation in the experimental group of animals. In a recent publication, a more detailed study was made of the contractile dysfunction evident in this model [29]. This showed that a lower peak systolic force and slower rate of decline of intraventricular pressure was evident in vitro and these changes correlated with a prolongation of the time course of the intracellular Ca2+ transient measured on the epicardial surface of Langendorff perfused hearts. This work suggests that the contractile abnormalities observed in this model in vivo may be related to an abnormal time course of the intracellular Ca2+ transient in a similar way to that seen in other models of hypertrophy (e.g. Refs. [13, 35]and in human heart failure [5, 12]).

4.2 Phosphorylation state of PLB in heart failure

Measurements on left ventricular tissue rapidly dissected from rabbit hearts indicated that the phosphorylation state of PLB was higher in failing myocardium (8.3±0.42 P-PLB) than that seen in sham-operated (3.2±0.8 P-PLB) or age matched controls (4.0±1.7 P-PLB). Similar results were seen when PLB from SR vesicle preparations was examined (results not shown). These initial measurements cannot distinguish between Ser16 and Thr17 phosphorylation sites, however the observation of phosphorylation states above P5 in PLB from failing hearts suggest that more than one site was phosphorylated. As indicated Section 1, few studies exist concerning the PLB phosphorylation state in heart failure, however, two recent abstracts indicate that in human and canine myocardium with chronic heart failure, the PLB phosphorylation state is lower than in PLB from normal myocardium [11, 33]. While in agreement with the present study, Boateng et al. [6]found an increased PLB phosphorylation state in hypertrophic rat myocardium. More definitive work on a human myocardium and a range of animal models is required to determine whether the phosphorylation state of PLB in heart failure is species-dependent.

Fig. 6

Fluorescence ratio records showing the change of [Ca2+] within the cuvette solution (2 ml) from a starting value of 1.5 μM, after addition of SR vesicles (Ves). Panel A: addition of SR vesicles (100 μg SR protein) prepared from a sham-operated rabbit; Panel B: SR vesicles prepared from a heart failure animal (100 μg SR protein); Panel C: an increased concentration of vesicles prepared from a heart failure animal (300 μg SR protein). The open symbols describe the calculated decline in [Ca2+] assuming a mono-exponential relationship using the following rate constants (k): Panel A, k=3.8 ms−1; Panel B, k=2.1 ms−1; Panel C, k=3.5 ms−1. Panel D, plot of the mean (±SEM) of the Ca2+ uptake rate constants vs. concentration of SR vesicle (μg/ml total protein). Filled circles, results from vesicles isolated from sham-operated group; open circles, results from vesicles isolated from heart failure group.

The cause of the higher phosphorylation state of PLB observed in this study is unknown, but may be related to the increased rate of phosphorylation of PLB observed in SR vesicle preparations isolated from failing hearts. The intrinsic kinase activity which lead to the phosphorylation of PLB in SR vesicle preparations was observed in the virtual absence of Ca2+ (3 mM EGTA). In SR vesicle preparations isolated from normal myocardium (sham-operated and age-matched control), the intrinsic kinase activity could be significantly reduced by inhibitors of A-kinase or C-kinase but not CaM-kinase. The higher intrinsic kinase activity observed in SR vesicle preparations from failing hearts was insensitive to A-kinase and CaM-kinase inhibitors and only partially inhibited by C-kinase inhibitors. Recently, a Ca2+-independent form of the CaM-kinase (δ-CaM-kinase) has been reported in cardiac SR [4], this kinase may contribute to the higher kinase activity observed in failing hearts. However, as shown in Fig. 4B, treatment with the selective CaM-kinase inhibitor, KN-62, had little effect on the intrinsic kinase activity. However, this technique will only detect large inhibitory effects on enzyme activity and further work is needed to quantify and characterise the intrinsic kinase activity associated with these SR vesicle preparations. It is tempting to associate the higher intrinsic kinase activity observed in SR vesicle preparations derived from the heart failure group with the higher phosphorylation state of PLB observed in fresh tissue homogenates and SR vesicle preparations from the same experimental group. However, this association is problematic, in particular all kinase activity measurements were performed in the presence of phosphatase inhibitors, it is conceivable that changes in kinase activities in heart failure were paralleled by changes in phosphatase activity. Without information on the relative activities of phosphatase and kinase systems, it is difficult to predict the extent to which altered kinase activity seen in isolated vesicle studies can explain the higher phosphorylation state of PLB observed in homogenates prepared from freshly dissected tissue. Another consideration is that the in vivo phosphorylation state of PLB will be affected by a range of cytosolic kinases and phosphatases some of which may be differentially influenced by heart failure. A further consideration in relating the enhanced in vivo phosphorylation state of PLB with the enhanced kinase activity of isolated SR vesicle preparations is that purified vesicle preparations represent a preferential recovery of a subfragment of the cardiac SR. The kinase activity of this preparation may not be representative of that of the complete SR, further studies are required to determine whether the activity of the intrinsic kinase is associated with distinct type of cardiac SR (i.e. superficial or terminal cisternae).

The maximum phosphorylation states observed in myocardial homogenates from the heart failure group were consistently greater than P5, suggesting that both Ser16 and Thr17 sites were routinely phosphorylated. The use of phosphorylation site-specific antibodies [9]and elaborate gel band shift procedures in this study confirmed that both Ser16 and Thr17 sites were phosphorylated in normal and heart failure myocardium. The more rapid phosphorylation of PLB observed in heart failure appeared to be caused by higher rates of phosphorylation at both sites. The agreement between the A1 mobility shift data and the PT-17- and PT-16-specific immunoblots illustrated in Fig. 3 indicates that both techniques independently confirm the higher rate of PLB phosphorylation. This is the first reported comparison between A1 mobility shifts and site-specific antibody for PLB and indicate close qualitative agreement between these two techniques.

A recent study has suggested that Thr17 phosphorylation was only observed in vivo under conditions when PP1 was inhibited, either by intrinsic PKA activation or addition of an inhibitor [27]. Increased Thr17 phosphorylation was observed in all the tissue preparations from the heart failure group, this may be due to a combination of decreased PP1 and increased SR kinase activities.

4.3 Abundance of SERCA 2 and PLB

As shown in Table 2, tissue homogenates from failing myocardium contained less PLB and SERCA 2 when compared to sham-operated or normal hearts. Separate groups of animals were used to measure these two proteins, yet both heart failure groups had comparable ejection fractions suggesting a similar degree of heart failure in both groups. The results concerning SERCA 2 abundance were similar to those reported by Zarain-Herzberg et al. [36]in a rat myocardial infarct model. The experimental design used in this study prevented a proper statistical analysis of the stoichiometry of PLB:SERCA 2 in this model. However, a reduction in this ratio was reported in human heart failure [24], but an increase was observed by Kiss et al. [18]studying pressure-overloaded hypertrophy in the guinea-pig. More detailed work is required to establish whether heart failure of differing origins may result in different relative changes in PLB and SERCA 2. Differences between studies could result from differences in: species; causes of heart failure or biochemical measurement techniques. This latter point has been discussed by a number of studies, in particular artefactual abundance measurements may arise when the tissue has undergone numerous sedimentation and purification steps. Under these conditions there may be preferential recovery of nonpathological SR from failing myocardium [25]. Measurements of protein abundance from tissue homogenates (as used in this study) which have not undergone purification steps has been encouraged [14]as a means of reducing artefact in the measurements. In this study, SERCA 2 and PLB abundance measurements were also performed on SR vesicle preparations. An approximate 50% reduction in protein abundance was observed in heart failure myocardium, a similar reduction to that to that seen in homogenate preparations.

4.4 SR vesicle Ca2+ uptake measurements

The Ca2+ uptake experiments (Fig. 6 and Table 2) show that preparations from ligated hearts exhibit reduced uptake rates. Numerous previous studies have examined SR Ca2+ uptake rates in failing hearts and have found similar results, e.g., Refs. [2, 18, 24, 31, 34]. In a separate study, a reduced Ca2+ uptake rate was observed in cell aggregate derived from the heart failure model used in this study [28]. Thus the reduced uptake rates observed in the purified SR preparations from failing hearts reflect the behaviour of the SR in myocardial preparations analogous to tissue homogenate. As shown in Fig. 6D, the relationship between SR vesicle concentration and rate constant deviated significantly from the linear relationship expected of a simple uptake model at protein concentrations above 100 μg/ml. The reason for this is unknown, however this may be due to nonspecific inhibitory effects of high total protein concentrations. Over the range of vesicle concentrations used in this study (30–300 μg/ml), the Ca2+ uptake rate constant measured in vesicles isolated from the heart failure group was approximately 50% of the sham-operated value. This is comparable to the reduction in abundance of SERCA 2 measured from tissue homogenates (see Table 2) and SR vesicles (results not shown). Therefore the higher phosphorylation state of PLB observed in both tissue homogenates and SR vesicles in the heart failure group was not reflected in the Ca2+ uptake measurements. However, the magnitude of the uncertainty associated with both protein abundance and the Ca2+ uptake measurements prevented a detailed comparison of the two parameters that may have revealed a component attributable to phosphorylation of PLB. Despite these reservations, it is clear that the additional degree of phosphorylation of PLB (and therefore pump activation) observed in the myocardium from the heart failure group is not sufficient to overcome the effects of decreased protein levels. It is possible that the upregulation of systems leading to increased PLB phosphorylation is a compensatory mechanism responding to the downregulation of SERCA 2. Previous suggestions for useful interventions in the treatment of heart disease have included inhibition of PLB phosphatase and stabilisation of the phosphorylated form of PLB [19]. However, in this study we have shown that, during the progression of heart failure in the rabbit, PLB is highly phosphorylated, yet due to the downregulation of SERCA 2, the net Ca2+ uptake capacity of the SR is reduced. Therefore, in this model of heart failure, there is little scope for improvement of SR function through further increases in phosphorylation of PLB. It remains to be seen whether a comparable mechanism exists at any stage of human heart failure.

In conclusion, results presented here suggest that in the rabbit coronary artery ligation model of heart failure, the remaining left ventricle possesses PLB at a higher than normal phosphorylation state. However, there is also downregulation of SERCA 2 and PLB proteins, the former effect resulting in the reduced Ca2+ uptake capacity of the SR. In addition to this, an intrinsic kinase activity associated with SR vesicles is upregulated in the heart failure group, and this enhanced activity may be responsible for the higher phosphorylation state of PLB.

Acknowledgements

This work was supported by a Medical Research Council programme grant. We thank Prof. Stuart Cobbe, Dr. Martin Hicks and technical staff at the Dept. of Cardiology, Glasgow Royal Infirmary for the preparation and characterisation of the animal model. We also wish to acknowledge Dr. John Colyer and Dr. Guido Drago, University of Leeds, for supplying, and helpful advice on, the PLB phosphorylation site-specific antibodies.

References

View Abstract