Functional alterations in adult rat myocytes after overexpression of phospholamban with use of adenovirus

KERRY DAVIA1, ROGER J. HAJJAR2, CESARE M. N. TERRACCIANO1, NATASHA SINGH KENT1, HARDEEP K. RANU1, PETER O'GARA1, ANTHONY ROSENZWEIG2 and SIAN E. HARDING1

1 Cardiac Medicine, National Heart and Lung Institute, Imperial College School of Medicine, London SW3 6LY, United Kingdom
2 Cardiovascular Research Center, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts 02115


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Davia, Kerry, Roger J. Hajjar, Cesare M. N. Terracciano, Natasha Singh Kent, Hardeep K. Ranu, Peter O'Gara, Anthony Rosenzweig, and Sian E. Harding. Functional alterations in adult rat myocytes after overexpression of phospholamban with use of adenovirus. Physiol. Genomics 1: 41–50, 1999.—An increased phospholamban (PLB)-to-sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA) ratio has been suggested to contribute to the slowing of relaxation in failing human ventricle. We have used an adenoviral vector carrying the sequence for PLB to increase this ratio in isolated adult rat ventricular myocytes, and we have examined the functional consequences. With use of adenoviral vectors, the PLB content of adult rat myocytes was increased 2.73-fold, with SERCA2a levels unchanged. Maximum contraction amplitude of PLB-overexpressing myocytes was decreased to 6.9 ± 0.3% shortening compared with 11.2 ± 0.8% for 24-h controls (Con; P < 0.001, 5 preparations, 103 myocytes). Maximum rates of shortening and relengthening were also significantly decreased. Ca2+ transient amplitudes were slightly depressed, and time to 50% decay of the transients was significantly increased: 237 ± 18 (n = 14 myocytes) and 432 ± 32 ms in Con and PLB (n = 15) myocytes, respectively (P < 0.001). The amount of Ca2+ in the sarcoplasmic reticulum stores was reduced by 21% (P < 0.05). Relaxation was significantly slower in PLB than in Con myocytes when the Na+/Ca2+ exchanger was blocked but not when sarcoplasmic reticulum Ca2+ uptake was inhibited. Adenovirus infection with Ad.RSV.PLB was therefore able to produce functional changes in adult cardiac myocytes within 24 h, consistent with overexpression of PLB and similar to those seen in failing human heart.

gene transfer; cardiomyocyte; sarco(endo)plasmic reticulum calcium-adenosinetriphosphatase; relaxation; calcium transient


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
PROLONGED RELAXATION is an important characteristic of human heart failure and is thought to result from an alteration in intracellular Ca2+ handling. Under normal physiological conditions, there are two main mechanisms for Ca2+ removal from the cytosol, the sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA2a) and the Na+/Ca2+ exchanger, with smaller contributions from the sarcolemmal Ca2+-ATPase and the mitochondrial Ca2+ uniporter. In mammalian myocardium the sarcoplasmic reticulum (SR) plays a major role in the regulation of intracellular Ca2+ during contraction and relaxation and is responsible for most of the Ca2+ removal on a beat-to-beat basis. A number of studies have identified a decrease in the activity of SERCA2a in failing human heart (1, 15, 20, 24, 30), and SERCA2a mRNA levels are often low (20, 24, 30), although there is disagreement about the reduction in protein levels (20, 24, 26, 27, 30). Complete inhibition of SERCA2a by thapsigargin can mimic in myocytes from nonfailing heart the main effects of failure, such as prolonged relaxation and an impaired contractile response to increasing stimulation frequency (8).

The activity of SERCA2a is modulated by phospholamban (PLB). In its unphosphorylated state, PLB interacts with SERCA2a and inhibits Ca2+ uptake. Phosphorylation of PLB by cAMP-dependent protein kinase or Ca2+-calmodulin-dependent protein kinases removes this inhibition, and the SR is able to sequester Ca2+ more rapidly. Decreased activity of SERCA2a could therefore be due to 1) reduced protein levels of SERCA2a, 2) increased PLB, or 3) decreased PLB phosphorylation. Most studies show maintained PLB in failing human myocardium (6, 13, 22, 26, 30), and an increase in the PLB-to-SERCA2a ratio has been observed (24). Low basal cAMP levels have been shown in human heart (6, 12, 36) and have been linked to a decrease in PLB phosphorylation (3), although a change in PLB phosphorylation state has not been detected in other studies (6, 25).

Much useful information about functional effects of changes in the PLB-to-SERCA2a ratio has been obtained from the study of transgenic mice with PLB overexpression or after PLB knockout (7, 37). However, there remains the possibility that compensatory changes not associated with the PLB alterations have modified the contractile response during the lifetime of these mice (14). Another method is to use vectors to overexpress PLB acutely in myocyte cultures, and we previously showed that increasing PLB levels depresses contraction and relaxation parameters in neonatal rat cells (17). These effects can be overcome by normalizing the PLB-to-SERCA2a ratio through simultaneous overexpression of SERCA2a (17). However, although neonatal cultures have the advantage of stability over 1–5 days, their disadvantage is that the relative contributions of the various relaxation mechanisms are different from those of adult myocytes. Studies have shown that uptake of Ca2+ into the SR via SERCA contributes more to relaxation and the Na+/Ca2+ exchanger contributes less in adult than in neonatal myocytes (2, 23, 32, 35). Heterogeneity of neonatal myocytes has been observed, with some cells costaining for SERCA2a and PLB and others containing SERCA2a but not PLB (34). It is therefore possible that overexpressed PLB more easily overwhelms the undeveloped SR in neonatal myocytes and that adult cells would be relatively resistant. Additionally, it is more difficult to perform sequential concentration-response curves or detailed experiments on relaxation mechanisms with these immature cells.

The purpose of the present study was to overexpress PLB in adult rat myocytes by using an adenoviral vector and to investigate in detail the effects on contractility and Ca2+ handling. Functional effects were determined from changes in contraction amplitude and duration, Ca2+ transient amplitude and duration, and rate of Ca2+ decline after release of SR Ca2+ stores. The effects of PLB overexpression were compared with those of thapsigargin in concentrations producing maximal or submaximal inhibition of SERCA2a function. The ability of ß-adrenoceptor stimulation by isoproterenol to overcome the inhibition of SERCA2a by the excess PLB was examined. Changes consistent with submaximal SERCA inhibition were shown, although, unexpectedly, the effects were more pronounced on the Ca2+ transient than on contraction. The surprising observation that PLB infection caused an apparent increase in relaxation through non-SR-, non-Na+/Ca2+ exchanger-dependent mechanisms was also made.


    METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

Construction of E1-deleted recombinant adenoviral vectors.
Construction of the adenoviral vectors Ad.RSV.ß-gal and Ad.RSV.PLB has been described in detail previously (16, 17). For experiments where positive identification of an individual infected cell was essential, an adenovirus carrying expression cassettes for PLB and green fluorescent protein (EGFP) was constructed. The backbone was derived from pBHG11 by direct ligation of the EGFP expression cassette into a unique Pac I site in E3. The PLB cDNA was subcloned into the expression cassette in the adenoviral shuttle vector pAd.RSV. The PLB-pAd.RSV shuttle vector and E3.EGFP.pBHG11 viral vector were then cotransfected into 293 cells to generate Ad.E3-GFP.E1.PLB. The virus Ad.E3-GFP.E1.PLB is a bifunctional recombinant adenovirus that generates two separate mRNA transcripts, one coding for green fluorescent protein (GFP) and one for PLB, which are not linked. The recombinant viruses were prepared as high-titer stocks by propagation in 293 cells, as previously described (17). The titer of stocks used for these experiments was 108 plaque-forming units (pfu)/well for Ad.RSV.PLB, Ad.RSV.ß-gal, or Ad.E3-GFP.E1.PLB.

Isolation and culture of rat myocytes.
Adult rat myocytes were isolated using a low-Ca2+ solution (in mmol/l: 120 NaCl, 5.4 KCl, 5 MgSO4, 5 pyruvate, 20 glucose, 20 taurine, 10 HEPES, and 5 nitrilotriacetic acid, preoxygenated with 100% O2) and collagenase and protease enzymes, as previously described (19). The myocytes were washed three times by gentle spinning in Dulbecco's solution (Sigma Chemical) containing 10,000 U/ml penicillin and 10 mg/ml streptomycin solution (Pen/Strep, GIBCO). The myocytes were plated in a 12-well culture dish in medium 199, with addition of 0.2% (wt/vol) BSA, 100 mmol/l ascorbate, 5 mmol/l creatine, 5 mmol/l taurine, 2 mmol/l carnitine, 0.1 µmol/l insulin, and Pen/Strep (11).

Adenovirus infection.
For maximum efficiency of myocyte infection, 108 pfu of adenovirus expressing the required protein were initially added to each well containing 2 x 104 myocytes (104 rods) in 0.5 ml of medium for the first 2 h; the medium was then increased to 2 ml/well for the remaining 24 h. The cells were maintained at 37°C in a humidified atmosphere of 5% CO2. Myocytes were then removed by gentle pipetting and washed three times using low-Ca2+ solution plus 200 µmol/l Ca2+ and without nitrilotriacetic acid. Control cells and Ad.RSV.ß-gal-infected cells were stained using X-gal to give an indication of the percentage of infected myocytes. Ad.E3-GFP.E1.PLB-infected cells were identified using fluorescence microscopy, with excitation wavelength 485 nm and emission at 520 nm.

Myocyte contraction experiments.
Myocytes were placed in a bath on an inverted microscope stage, as previously described (19), in Krebs-Henseleit (K-H) solution (in mmol/l: 119 NaCl, 1 CaCl2, 4.7 KCl, 0.94 MgSO4, 1.2 KH2PO4, 25 NaHCO3, and 11.5 glucose) that was bubbled with 95% O2-5% CO2 to pH 7.4. Experiments were carried out at 32°C with field stimulation at 0.5 Hz, and contraction was monitored by a video edge-detection device with spatial resolution of 1 in 256 or 512 and time resolution of 10 or 20 ms. After 15 min of equilibration in K-H solution with 1 mmol/l Ca2+, 5–15 myocytes were selected, as far as possible randomly, for 3- to 5-min recordings of contraction amplitude (percent cell shortening), time to peak contraction (TTP), time to 50% relaxation (R50), and time to 90% relaxation (R90). Myocytes were subjected to increasing concentrations of Ca2+ (2, 4, 6, and 8 mmol/l) until a maximum was reached (no further increase in amplitude or development of spontaneous contractions). At this point, contraction amplitude and duration were measured for 3–5 min in a further 5–15 myocytes contracting in the high-Ca2+ solution. Importantly, the high Ca2+ concentration for the Ad.RSV.ß-gal- or Ad.RSV.PLB-treated myocytes was matched to that for the control on that day. After return to 1 mmol/l Ca2+ K-H solution, a myocyte was selected and superfused with increasing concentrations of isoproterenol in half-log units from a starting concentration of 1 nmol/l. When a maximum had been reached, defined by the same criteria as for the Ca2+ curve, five further cells in the bath were again examined for amplitude and duration. In a separate aliquot, a myocyte contracting in a maximally activating Ca2+ concentration was exposed to 3 µmol/l thapsigargin for 10 min. Amplitude and duration were measured on up to five other myocytes in the same bath after thapsigargin treatment.

Contraction experiments on myocytes from control and Ad.RSV.ß-gal- and Ad.RSV.PLB-treated cells were done as far as possible concurrently for the same preparation to obviate the chance of time-dependent differences.

Measurements of cytoplasmic Ca2+.
Intracellular Ca2+ was monitored using the Ca2+-sensitive, single-excitation, dual-emission fluorescent dye indo 1, as described previously (33). Cells were incubated for 20 min with 10 µM indo 1-AM (Molecular Probes). The supernatant was then removed and replaced with fresh DMEM (GIBCO). Background fluorescence subtraction was routinely carried out using a cell-free region as zero fluorescence. Indo 1 fluorescence was calibrated as described by Terracciano and MacLeod (33).

Assessment of the SR Ca2+ content and relative contribution of Ca2+ regulatory mechanisms to relaxation.
After stimulation was stopped, cells were superfused with an Na+-free, Ca2+-free (0 Na-0 Ca) solution (Na+ was substituted with Li+) for 2 s, then with 0 Na-0 Ca solution containing 10 mM caffeine for another 2 s. The amplitude of the increase in indo 1 fluorescence was used as an indication of the SR Ca2+ content. To study the relative contributions of Ca2+ regulatory mechanisms to relaxation, different solutions were then used. Relaxation was performed in caffeine only to study the contribution of non-SR-dependent mechanisms, in 0 Na-0 Ca solution to study non-Na+/Ca2+-exchange dependent relaxation, or in 0 Na-0 Ca + caffeine to measure the relative contribution to relaxation by non-SR, non-Na+/Ca2+ exchange mechanisms (mitochondrial and sarcolemmal Ca2+-ATPase) (4). Monoexponential curves were fitted on the recovery phase of the Ca2+ traces, and analysis was performed as described by Terracciano and MacLeod (33).

Statistical analysis.
Results for a number of myocytes for a given preparation were pooled, so that n values for statistics refer to preparations, except where virus coexpressing PLB and GFP was used, where n refers to myocytes. Error bars indicate 1 SE. Differences between freshly isolated myocytes, 24-h cultured myocytes not exposed to virus (control), and myocytes exposed to Ad.RSV.PLB or Ad.RSV.ß-gal were analyzed by one-way ANOVA with Fisher's test for pairwise comparison of means. Concentration-response curves were compared using repeated-measures ANOVA.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Initial experiments were performed using Ad.RSV. PLB and Ad.RSV.ß-gal, and large numbers of cells were studied from each preparation to minimize possible influences of the small proportion of uninfected myocytes. Subsequent development of the dual-expression Ad.E3-GFP.E1.PLB virus allowed positive identification of infected myocytes. This was used to confirm key points and to perform experiments such as the determination of SR Ca2+ loading, where it is not possible to process large numbers of myocytes from a given preparation.


Reporter gene expression and effects on contractility in adult myocytes.
Expression of ß-galactosidase was consistent in myocytes infected with Ad.RSV.ß-gal and cultured for 24 h, averaging 85 ± 2% (n = 12). Expression was confined to viable rod-shaped myocytes, with the nonviable rounded cells rarely staining positive. It was observed that increasing the concentration of virus from 107 to 108 pfu/well could increase the intensity of ß-galactosidase staining and spread color outside the nucleus, without increasing the percentage of rod-shaped myocytes that stained positive. It is likely that the 15% of myocytes that do not express protein are in some way compromised and that increasing viral titers simply leads to multiple infections of the susceptible cells.

The effects on contraction of 24 h of culture after infection with reporter virus Ad.RSV.ß-gal (or in one case virus coding for GFP) are shown in Table 1 compared with freshly isolated myocytes (day 0) and cells cultured for 24 h without virus (day 1 control). Contraction amplitude in 1 mmol/l Ca2+ dropped significantly between day 0 and day 1, but there was no further drop with Ad.RSV.ß-gal treatment. Maximum contraction amplitudes in high Ca2+, isoproterenol, or after thapsigargin treatment did not differ between the three groups. Culture for 24 h was associated with increases in TTP or R90 in high or low Ca2+, but again Ad.RSV.ß-gal treatment did not exacerbate this. In several cases the slowing of the beat was less pronounced in the infected cells, raising the possibility that viral infection could accelerate contraction and relaxation. Overall, the conclusion is that infection with Ad.RSV.ß-gal did not produce deleterious changes in contractility in addition to those resulting from 24 h of culture.


View this table:
[in this window]
[in a new window]
 
Table 1. Effect of 24 h of culture with and without Ad.RSV.ß-gal infection on contraction of adult myocytes

 
PLB expression.
Overexpression of PLB protein 24 h after infection was shown by Western blotting in seven of nine preparations tested; only preparations demonstrating increased PLB protein were used for comparison of contraction data. Those that did not show an increase did not have the characteristic changes described below. The average increase in the level of PLB protein expression in the preparations the contractile characteristics of which are analyzed below was 2.73 ± 0.23-fold (n = 4) for Ad.RSV.PLB and 2.31 ± 0.06-fold (n = 3) for Ad.E3-GFP.E1.PLB compared with freshly isolated myocytes. In the same preparations, Ad.RSV.ß-gal infection did not alter PLB levels (0.97 ± 0.13-fold compared with fresh, n = 3). SERCA2a levels were not changed 24 h after PLB overexpression (0.98 ± 0.06-fold compared with fresh, n = 7).

Myocyte contraction in low and high Ca2+ after infection with Ad.RSV.PLB.
The number of myocytes with visible contractions in response to electrical stimulation (1-ms pulse width, 50 V) in 1 mmol/l Ca2+, with use of a x16 objective, was compared between groups. Thirty to 40 cells were counted for each condition in each preparation. At this Ca2+ concentration, contraction was visible in significantly fewer myocytes (66.4 ± 13%) in the Ad.RSV.PLB-infected cells than in freshly isolated myocytes (day 0, 98.2 ± 0.75%) or 24-h cultured untreated myocytes (day 1, 90.2 ± 5.9%, P < 0.05 for both). The number of contracting myocytes from the Ad.RSV.ß-gal-infected group showed a slight but not significant drop (80.2 ± 6.9%). When the Ca2+ concentration was raised to >=4 mmol/l, few myocytes were noncontractile in any group. It is therefore likely that the lack of visible contraction in 1 mol/l Ca2+ simply represents an undetectably small contraction amplitude.

In low Ca2+ (1 mmol/l) there was little difference in contraction amplitude between myocytes from day 0 and day 1 control or Ad.RSV.ß-gal- or Ad.RSV.PLB-treated groups (Fig. 1A). However, in high Ca2+ (4–8 mmol/l, Fig. 1B), myocytes infected with Ad.RSV.PLB had a significantly depressed contraction amplitude. It is possible that the lack of effect at 1 mmol/l Ca2+ is due to the higher proportion of cells with undetectably small contraction amplitudes that were not included in the calculation.



View larger version (27K):
[in this window]
[in a new window]
 
Fig. 1. Contraction amplitude (left) and time to peak contraction + time to 50% relaxation (TTP + R50; right) in myocytes from 5 preparations at 1 mM Ca2+ (A) and maximally activating Ca2+ (4–8 mM, B). Myocytes were studied directly after isolation (day 0) or 24 h later (day 1) after incubation without virus (Con) or with Ad.RSV.ß-gal (LacZ) or Ad.RSV.PLB (PLB). Total number of myocytes studied was 128 for 1 mM Ca2+ and 178 for maximum Ca2+. * Significantly different from day 0, day 1 Con, and LacZ (P < 0.001).

 
Beat duration (TTP + R50) in high or low Ca2+ was increased markedly in some preparations but not at all in others, so that the difference was not significant when all preparations were considered (Fig. 1). Separate analysis of relaxation only showed a similar pattern, with R50 at maximum Ca2+ of 176 ± 41 ms in Ad.RSV.PLB-treated myocytes compared with 99 ± 17 ms in 24-h controls (not significant). Analysis is performed for n preparations, not n myocytes. In preparations where beat duration was increased, TTP, R50, and R90 were affected. This is illustrated in Fig. 2, which is taken from a preparation where amplitude and duration were significantly altered. There was no apparent relation between the increase in beat duration and the extent of decrease in amplitude or the degree of PLB overexpression.



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 2. Sample trace of myocytes from a preparation where contraction amplitude and beat duration were affected by phospholamban (PLB) overexpression. Contraction was at maximum Ca2+. GFP, green fluorescent protein.

 
Maximum rates of shortening and relengthening were significantly decreased in Ad.RSV.PLB- but not Ad.RSV.ß-gal-treated myocytes (shortening: 175 ± 35, 146 ± 18, and 80 ± 8 µm/s for Con, ß-gal, and PLB, respectively, P < 0.05 vs. Con; lengthening: 146 ± 26, 127 ± 18, and 63 ± 9 µm/s for Con, ß-gal, and PLB, respectively, P < 0.01 vs. Con). However, in unloaded cardiac myocytes a decrease in amplitude will contribute to this effect.

To confirm that PLB infection could produce decreases in contraction amplitude without increases in beat duration, Ad.E3-GFP.E1.PLB was used and virus-treated cells positive for GFP fluorescence were compared with virus-treated cells not expressing GFP and with myocytes not exposed to virus. At 108 pfu Ad.E3-GFP.E1.PLB/well, 60 ± 6% of myocytes showed strong green fluorescence after 24 h (n = 8 preparations). GFP-expressing myocytes without slowed relaxation were observed, and this was true for experiments at room temperature [R50: 0.31 ± 0.01 ms in control cells (n = 13 myocytes), 0.35 ± 0.04 ms in infected GFP-positive cells (n = 12), and 0.35 ± 0.03 ms in infected GFP-negative cells (n = 9), not significant] and at 35°C (see Ca2+ transients in myocytes infected with Ad.E3-GFP.E1.PLB). Extending the culture period to 48 h increased the GFP-positive myocytes from 66 to 80% in one preparation but did not increase relaxation times significantly (data not shown).

To discover whether partial inhibition of SERCA2a could produce conditions where amplitude was affected but beat duration was preserved, freshly isolated rat myocytes were exposed to increasing thapsigargin concentrations (Fig. 3). The decrease in amplitude preceded the major change in TTP and R50, demonstrating that compensatory mechanisms could initially preserve the speed of contraction and relaxation during submaximal SERCA2a inhibition.



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 3. Effect of increasing thapsigargin concentration ([thapsigargin]) on contraction amplitude (A) and beat duration (TTP + R50, B) in same myocytes (n = 4). Cells were contracting in presence of thapsigargin for 10 min before measurements were taken.

 
Ca2+ transients in myocytes infected with Ad.E3-GFP.E1.PLB.
Myocytes positively identified as infected with Ad.E3-GFP.E1.PLB by fluorescence microscopy had Ca2+ transients with significantly slowed rate of decline compared with untreated myocytes cultured for the same time (24 h; Figs. 4 and 5). The amplitude of the Ca2+ transient was smaller, although not significantly so [untreated cells: 0.058 ± 0.007 (n = 14); GFP-positive cells: 0.044 ± 0.007 (n = 15)], and diastolic Ca2+ levels were not increased [untreated cells: 0.29 ± 0.01 (n = 14); GFP-positive cells: 0.27 ± 0.02 (n = 15)].



View larger version (13K):
[in this window]
[in a new window]
 
Fig. 4. Time to 50% relaxation of cell length compared with time to 50% decline of Ca2+ transient in uninfected myocytes at 24 h (Con, n = 14 myocytes, 4 preparations) and in Ad.E3-GFP.E1.PLB-treated cells (PLB-GFP, n = 15 myocytes). *** P < 0.001 compared with Con.

 


View larger version (22K):
[in this window]
[in a new window]
 
Fig. 5. Sample traces of Ca2+ transient (A) and contraction (B) for myocytes cultured for 24 h without virus (Con) and with Ad.E3-GFP.E1.PLB (PLB-GFP). Amplitudes are normalized to show differences in relaxation times. [Ca2+], Ca2+ concentration.

 
Direct comparison of the duration of cell relengthening with decline of the Ca2+ transient (Fig. 4) after PLB overexpression showed that slowing of the transient was more pronounced (Fig. 5). It was also clear that cell relengthening occurred more rapidly than the restoration of diastolic Ca2+ levels. When paired data were examined for individual myocytes, time to 50% reduction in Ca2+ was longer than time to 50% relengthening by 51 ± 28 ms for untreated myocytes (n = 8, not significant) and 216 ± 47 ms for GFP-positive cells (n = 7, P < 0.01). Time to peak effect was slightly (22 ms) but not significantly longer for the Ca2+ transient than for the cell length change; this did not differ between untreated and GFP-positive myocytes. These experiments were performed at 35°C, where a dissociation of Ca2+ transient times from cell length changes has been observed previously (29).

Relaxation after SR Ca2+ release in myocytes infected with Ad.E3-GFP.E1.PLB.
Myocytes identified as infected with Ad.E3-GFP.E1.PLB were subject to rapid caffeine spritzes to release the SR Ca2+ load. The amount of Ca2+ released, as estimated from the resulting Ca2+ transient, was decreased by 21% in GFP-positive myocytes compared with untreated cells cultured for the same time (24 h; Fig. 6A). This indicates a reduction in SR Ca2+ load in PLB-overexpressing myocytes.



View larger version (10K):
[in this window]
[in a new window]
 
Fig. 6. A: sarcoplasmic reticulum Ca2+ content, determined from Ca2+ transient after a caffeine spritz, in uninfected myocytes at 24 h (Con, n = 14 myocytes, 4 preparations) compared with Ad.E3-GFP.E1.PLB-treated cells (PLB-GFP, n = 12 myocytes). * P < 0.05 compared with Con. B: relaxation from caffeine spritz in 0 Na-0 Ca medium in uninfected myocytes at 24 h (Con, n = 12 myocytes) compared with Ad.E3-GFP.E1.PLB-treated cells (PLB-GFP, n = 12 myocytes). * P < 0.02 compared with Con.

 
Reuptake of Ca2+ after the spritz was done in one of four conditions: 1) normal buffer, 2) the continued presence of caffeine to inhibit SR Ca2+ accumulation, 3) medium without Na+ or Ca2+ (0 Na-0 Ca) to inhibit the Na+/Ca2+ exchanger, or 4) caffeine plus 0 Na-0 Ca to inhibit main mechanisms of Ca2+ removal. As would be predicted, removal of Ca2+ was significantly slowed in Ad.E3-GFP.E1.PLB-treated myocytes compared with control when relaxation was in the presence of 0 Na-0 Ca solutions (Fig. 6B) but was not significantly different in the presence of caffeine (Table 2). The SR-mediated decline was therefore calculated to be 1,682 ± 184 (n = 11) for control and 3,035 ± 352 (n = 12) for Ad.E3-GFP.E1.PLB-treated myocytes (P < 0.005). Surprisingly, however, relaxation in the presence of 0 Na-0 Ca and caffeine was significantly faster for Ad.E3-GFP.E1.PLB-infected myocytes (Table 2), indicating that a further Ca2+ removal mechanism has assumed a greater importance in the infected myocytes.


View this table:
[in this window]
[in a new window]
 
Table 2. Rate of decline in Ca2+ transient after caffeine spritz

 
Effect of isoproterenol in Ad.RSV.PLB-infected myocytes.
Isoproterenol treatment should increase the phosphorylation of PLB and so reverse effects of overexpression. Contraction amplitude and beat duration of myocytes in maximally stimulating concentrations of isoproterenol (in 1 mmol/l Ca2+) are shown in Fig. 7. Contraction amplitude was not significantly different between groups in the presence of maximally effective concentrations of isoproterenol. Some variability between preparations was apparent, with isoproterenol able to overcome the effects of Ad.RSV.PLB treatment in some cases but not in others. Although numbers are too small to allow correlations to be performed, there was a trend for isoproterenol to be more effective in preparations where only amplitude was depressed and less so when amplitude and duration had been altered by Ad.RSV.PLB. Figure 8 shows the effect of isoproterenol in two preparations with the different patterns of response.



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 7. Contraction amplitude (A) and TTP + R50 (B) in myocytes from 5 preparations in maximally activating isoproterenol (30–100 nM). Myocytes were studied directly after isolation (day 0) or 24 h later (day 1) after incubation without virus (Con) or with Ad.RSV.ß-gal (LacZ) or Ad.RSV. PLB (PLB). Total number of myocytes studied was 98. There were no significant differences between groups.

 


View larger version (20K):
[in this window]
[in a new window]
 
Fig. 8. Severe (prep 1) and mild (prep 2) overexpression of PLB: reversibility by isoproterenol (Iso). Contraction amplitude (percent shortening, A and C) and TTP + R50 (B and D) are shown in myocytes from 2 preparations with different levels of PLB overexpression. Number of myocytes is indicated within bars. In prep 1, amplitude was significantly decreased (* P < 0.05) and TTP + R50 increased (*** P < 0.001) after PLB expression. Isoproterenol did not reverse effects of overexpression. Difference between control (day 1 uninfected) and PLB remained significant during isoproterenol exposure. In prep 2, amplitude was significantly decreased (* P < 0.05) but TTP + R50 was not increased. Isoproterenol reversed depression of contraction amplitude (# P < 0.01 between Ca and Iso in PLB myocytes) and significantly accelerated TTP + R50 in PLB myocytes (## P < 0.005, PLB vs. Con in Iso).

 
Effect of thapsigargin in Ad.RSV.PLB-infected myocytes.
Because thapsigargin inhibits SERCA2a, differences between control and PLB-overexpressing myocytes should be abolished after thapsigargin treatment. Contracting myocytes were exposed to a maximally inhibiting concentration of thapsigargin for 10 min, by which time a stable depression of amplitude had been obtained. There was no significant difference in contraction amplitude or beat duration between control and Ad.RSV.PLB-treated myocytes after exposure to thapsigargin (Fig. 9). Thapsigargin appeared to slow contraction and relaxation most markedly in the day 1 control group. However, the low amplitudes after thapsigargin treatment made measurements difficult to obtain, with a proportion of cells ceasing to contract altogether. Data gathered in the presence of thapsigargin should, therefore, not be overinterpreted.



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 9. Contraction amplitude and TTP + R50 in myocytes from 5 preparations after 10 min in maximum Ca2+ plus 3 µM thapsigargin. Myocytes were studied directly after isolation (day 0) or 24 h later (day 1) after incubation without virus (Con) or with Ad.RSV. ß-gal (LacZ) or Ad.RSV.PLB (PLB). Total number of myocytes studied was 88. There were no significant differences between groups.

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
This study confirms adenovirus infection as a rapid and efficient means of introducing protein-coding sequences into adult cardiac myocytes and shows that functional effects can be demonstrated 24 h after infection. Overexpression of PLB had many of the effects that would be predicted, consistent with partial inhibition of SERCA2a. The Ca2+ transient decline was significantly slowed, and SR Ca2+ load was decreased. Removal of Ca2+ after a caffeine spritz to release SR Ca2+ was slowed in the presence of 0 Na-0 Ca (to inhibit the Na+/Ca2+ exchanger), and the calculated contribution of SR uptake to Ca2+ decline was approximately halved. Contraction amplitude (percent cell shortening) was reproducibly depressed, with slow contraction and relaxation seen in some preparations but not in others. Quantitatively, the effect on contraction amplitude was equivalent to a submaximal concentration of the SERCA2a inhibitor thapsigargin. After treatment with maximally effective concentrations of thapsigargin, the difference between control and Ad.RSV.PLB-treated myocytes was abolished, consistent with an effect on SR Ca2+ uptake. Isoproterenol treatment, which is predicted to reverse PLB inhibition by stimulation of phosphorylation, increased contraction amplitude in the PLB-overexpressing myocytes to a level not significantly different from that of control.

Slowing of the Ca2+ transient decline was more consistently observed than slowing of cell relengthening after PLB overexpression. For individual myocytes, a paired comparison showed that the time to 50% decrease in Ca2+ was longer than the time to 50% relaxation of cell length. This was particularly pronounced for the myocytes overexpressing PLB. This dissociation of the transient from mechanical relaxation was observed previously when myocytes were used at more physiological temperatures; e.g., rabbit myocytes at 35°C had an R50 for twitch relaxation of 50 ms compared with 117 ms for the Ca2+ transient (29). It is likely that there is a threshold for contractile activation by Ca2+ and that part of the fall in Ca2+ occurs below this threshold, so that the cell is already fully relaxed. This may be a mechanism that protects against ectopic contractions that might result from small fluctuations in diastolic Ca2+. One implication of the observation in the present context is that the functional consequences of manipulation of Ca2+ handling should be deduced from studies that include cell shortening data rather than Ca2+ transients alone.

PLB overexpression has been achieved previously in transgenic mice (7), in neonatal rat myocytes with use of adenoviral vectors (17), and in rat heart in vivo (18). Quantitatively, the prolongation of Ca2+ transient relaxation was similar between the present studies in adult myocytes, where R50 was increased to 183% of control, and studies in neonates, in which a similar viral construct was used (206% of control) (17). For transgenic mice with a twofold overexpression of PLB, the R50 for the Ca2+ transient was similarly increased to 206% of the control value (7). The amplitude of unloaded myocyte shortening in these mice was decreased by 45%, which is comparable to our finding of a 38% decrease in the Ad.RSV.PLB-treated myocytes. Rates of shortening and relengthening in the PLB-overexpressing transgenic mice were decreased by 48 and 50%, respectively, which again are close to our values of 54 and 57%. However, a decrease in amplitude in unloaded myocytes will itself reduce rates of shortening and relengthening, even when there is no change in beat duration. In the in vivo rat hearts infected with Ad.RSV.PLB, a 2.8-fold increase in PLB accompanied a peak left ventricular pressure reduction of 37% after 2 days as well as a doubling of the time constant of relaxation (18). There is, therefore, good quantitative agreement between these studies concerning the functional effects of PLB overexpression.

One surprising observation is that blockade of the two main mechanisms of relaxation, SERCA and the Na+/Ca2+ exchanger, revealed a difference between uninfected and PLB-overexpressing myocytes. Under these conditions, where only minor Ca2+ removal mechanisms are active, the PLB-overexpressing cells relaxed twice as fast as controls (Table 2). The increased importance of a non-SR-, non-Na+/Ca2+ exchange-dependent relaxation in PLB-overexpressing myocytes may account for the lack of effect on diastolic Ca2+ levels, in contrast to the results for neonatal myocytes (17). However, complete inhibition of SERCA by thapsigargin strongly depressed contraction and relaxation in control and PLB-overexpressing myocytes and eliminated the difference between the two. This suggests that the effects on Ca2+ removal brought about by this non-SR-, non-Na+/Ca2+ exchange-dependent mechanism are only detectable when the other mechanisms are inhibited but may be too slow to play a role during a twitch.

Could changes in the PLB-to-SERCA ratio account for the poor contraction and relaxation of failing human heart, as has been suggested (21)? Qualitatively and quantitatively, the alterations in the PLB-overexpressing rat myocytes are similar to those in failing human ventricle. R50 in myocytes from failing human left ventricle was increased from 160 to 250 ms compared with nonfailing human ventricle (10), and contraction amplitude (or force of contraction in muscle strips) was approximately halved (5, 9, 28, 31). This compares with an increase in average R50 from 99 to 176 ms, a doubling in duration of Ca2+ transient decline, and a 38% reduction in maximum contraction amplitude for PLB-overexpressing rat myocytes. However, the change in the PLB-to-SERCA ratio was greater for the rat myocytes in the present study than has been reported for failing human heart. The greatest change in the PLB-to-SERCA ratio demonstrated for human ventricle was 28% (24), whereas the PLB-to-SERCA ratio in infected rat myocytes in the present study was increased by 130–170%. If the diluting effect of nonexpressing rat cells (uninfected rods plus rounded myocytes) is taken into account, this difference might rise even further. Rat myocardium is, if anything, more sensitive to SERCA2a inhibition than human myocardium (cf. Figs. 3 and 9 with thapsigargin on human myocytes in Ref. 8). It therefore seems unlikely that the 28% increase in the PLB-to-SERCA ratio in failing human heart would have been sufficient per se to produce the changes in contractility observed. One caveat for this comparison is the possibility that overexpressed PLB in the rat may not have been targeted to the membrane of the SR as efficiently as that produced naturally by the cell.

In summary, overexpression of PLB protein with use of adenoviral vectors can produce within 24 h functional changes in adult rat myocytes consistent with submaximal SERCA2a inhibition. Slowing of relaxation and Ca2+ removal and depression of maximum contraction amplitude were observed in PLB-overexpressing myocytes and were similar to changes seen in failing human heart. However, the change in the PLB-to-SERCA ratio producing these effects was greater in the rat than has been reported for human ventricle.


    ACKNOWLEDGMENTS
 
This work was supported in part by National Institutes of Health Grants HL-50361 and HL-57623 (R. J. Hajjar) and HL-54202, HL-59521, HL-61557, and AI-40970 (A. Rosenzweig), British Heart Foundation Grants PG/97064 (H. K. Ranu) and PG/98043 (K. Davia), and Pfizer (K. Davia).

Address for reprint requests and other correspondence: S. E. Harding, Cardiac Medicine, National Heart and Lung Institute, Imperial College School of Medicine, Dovehouse St., London SW3 6LY, UK (E-mail: sian.harding{at}ic.ac.uk).


    FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Arai, M., H. Matsui, and M. Periasamy. Sarcoplasmic reticulum gene expression in cardiac hypertrophy and heart failure. Circ. Res. 74: 555–563, 1994.[Medline]
  2. Balaguru, D., P. S. Haddock, J. L. Puglisi, D. M. Bers, and W. A. Coetzee. Role of sarcoplasmic reticulum in contraction and relaxation of immature rabbit ventricular myocytes. J. Mol. Cell. Cardiol. 29: 2747–2757, 1997.[Medline]
  3. Bartel, S., B. Stein, T. Eschenhagen, U. Mende, J. Neumann, W. Schmitz, E. G. Krause, P. Karczewski, and H. Scholz. Protein phosphorylation in isolated trabeculae from nonfailing and failing human hearts. Mol. Cell. Biochem. 157: 171–179, 1996.[Medline]
  4. Bassani, J. W. M., R. A. Bassani, and D. Bers. Relaxation in rabbit and rat cardiac cells: species-dependent differences in cellular mechanisms. J. Physiol. (Lond.) 476: 279–293, 1994.[Abstract]
  5. Bohm, M., K. La Rosee, U. Schmidt, C. Schulz, R. H. Schwinger, and E. Erdmann. Force-frequency relationship and inotropic stimulation in the nonfailing and failing human myocardium: implications for the medical treatment of heart failure. Clin. Investig. 70: 421–425, 1992.[Medline]
  6. Bohm, M., B. Reiger, R. H. G. Schwinger, and E. Erdmann. cAMP concentrations, cAMP dependent protein kinase activity, and phospholamban in non-failing and failing myocardium. Cardiovasc. Res. 28: 1713–1719, 1994.[Medline]
  7. Chu, G., G. W. Dorn, W. Luo, J. M. Harrer, V. J. Kadambi, R. A. Walsh, and E. G. Kranias. Monomeric phospholamban overexpression in transgenic mouse hearts. Circ. Res. 81: 485–492, 1997.[Abstract/Free Full Text]
  8. Davia, K., C. H. Davies, and S. E. Harding. Effects of inhibition of sarcoplasmic reticulum calcium uptake on contraction of myocytes from failing human ventricle. Cardiovasc. Res. 33: 88–97, 1997.[Medline]
  9. Davies, C. H., K. Davia, J. G. Bennett, J. R. Pepper, P. A. Poole-Wilson, and S. E. Harding. Reduced contraction and altered frequency response of isolated ventricular myocytes from patients with heart failure. Circulation 92: 2540–2549, 1995.[Abstract/Free Full Text]
  10. Del Monte, F., P. O'Gara, P. A. Poole-Wilson, M. H. Yacoub, and S. E. Harding. Cell geometry and contractile abnormalities of myocytes from failing human left ventricle. Cardiovasc. Res. 30: 281–290, 1995.[Medline]
  11. Ellingsen, O., A. J. Davidoff, S. K. Prasad, H. J. Berger, J. P. Springhorn, J. D. Marsh, R. A. Kelly, and T. W. Smith. Adult rat ventricular myocytes cultured in defined medium: phenotype and electromechanical function. Am. J. Physiol. 265 (Heart Circ. Physiol. 34): H747–H754, 1993.[Abstract/Free Full Text]
  12. Feldman, A. M., A. E. Cates, W. B. Veazey, R. E. Hershberger, M. R. Bristow, K. L. Baughman, W. A. Baumgartner, and C. Van Dop. Increase of the 40,000-mol wt pertussis toxin substrate (G protein) in the failing human heart. J. Clin. Invest. 82: 189–197, 1988.[Medline]
  13. Flesch, M., R. H. Schwinger, P. Schnabel, F. Schiffer, I. van Gelder, U. Bavendiek, M. Sudkamp, F. Kuhn-Regnier, and M. Bohm. Sarcoplasmic reticulum Ca2+-ATPase and phospholamban mRNA and protein levels in end-stage heart failure due to ischemic or dilated cardiomyopathy. J. Mol. Med. 74: 321–332, 1996.[Medline]
  14. Ghu, G., W. Luo, J. P. Slack, C. Tilgmann, W. E. Sweet, M. Spindler, K. W. Saupe, G. P. Boivin, C. S. Moravec, I. L. Grupp, and J. S. Ingwall. Compensatory mechanisms associated with the hyperdynamic function of phospholamban-deficient mouse hearts. Circ. Res. 79: 1064–1076, 1996.[Abstract/Free Full Text]
  15. Gwathmey, J. K., L. Copelas, R. MacKinnon, F. J. Schoen, M. D. Feldman, W. Grossman, and J. P. Morgan. Abnormal calcium handling in myocardium from patients with end-stage heart failure. Circ. Res. 61: 70–76, 1987.[Abstract]
  16. Hajjar, R. J., J. X. Kang, J. K. Gwathmey, and A. Rosenzweig. Physiological effects of adenoviral gene transfer of sarcoplasmic reticulum calcium ATPase in isolated rat myocytes. Circulation 95: 423–429, 1997.[Abstract/Free Full Text]
  17. Hajjar, R. J., U. Schmidt, J. X. Kang, T. Matsui, and A. Rosenzweig. Adenoviral gene transfer of phospholamban in isolated rat cardiomyocytes. Rescue effects by concomitant gene transfer of sarcoplasmic reticulum Ca ATPase. Circ. Res. 81: 145–153, 1997.[Abstract/Free Full Text]
  18. Hajjar, R. J., U. Schmidt, T. Matsui, J. L. Guerrero, K. Lee, J. K. Gwathmey, G. W. Dec, M. J. Semigran, and A. Rosenzweig. Modulation of ventricular function through gene transfer in vivo. Proc. Natl. Acad. Sci. USA 95: 5251–5256, 1998.[Abstract/Free Full Text]
  19. Harding, S. E., G. Vescovo, M. Kirby, S. M. Jones, J. Gurden, and P. A. Poole-Wilson. Contractile responses of isolated rat and rabbit myocytes to isoproterenol and calcium. J. Mol. Cell. Cardiol. 20: 635–647, 1988.[Medline]
  20. Hasenfuss, G., H. Reinecke, R. Studer, M. Meyer, B. Pieske, J. Holtz, C. Holubarsch, H. Posival, H. Just, and H. Drexler. Relation between myocardial function and expression of sarcoplasmic reticulum Ca2+-ATPase in failing and nonfailing human myocardium. Circ. Res. 75: 434–442, 1994.[Abstract]
  21. Koss, K. L., I. L. Grupp, and E. G. Kranias. The relative phospholamban and SERCA2 ratio: a critical determinant of myocardial contractility. Basic Res. Cardiol. 92, Suppl. 1: 17–24, 1997.
  22. Linck, B., P. Boknik, T. Eschenhagen, F. U. Muller, J. Neumann, M. Nose, L. R. Jones, W. Schmitz, and H. Scholz. Messenger RNA expression and immunological quantification of phospholamban and SR-Ca2+-ATPase in failing and nonfailing human hearts. Cardiovasc. Res. 31: 625–632, 1996.[Medline]
  23. Lompre, A. M., F. Lambert, E. Lakatta, and K. Schwartz. Expression of sarcoplasmic reticulum Ca2+-ATPase and calsequestrin genes in rat heart during ontogenic development and aging. Circ. Res. 69: 1380–1388, 1991.[Abstract]
  24. Meyer, M., W. Schillinger, B. Pieske, C. Holubarsch, C. Heilmann, H. Posival, G. Kuwajima, K. Mikoshiba, H. Just, G. Hasenfuss, et al. Alterations of sarcoplasmic reticulum proteins in failing human dilated cardiomyopathy. Circulation 92: 778–784, 1995.[Abstract/Free Full Text]
  25. Movsesian, M. A., J. Colyer, J. H. Wang, and J. Krall. Phospholamban-mediated stimulation of Ca2+ uptake in sarcoplasmic reticulum from normal and failing hearts. J. Clin. Invest. 85: 1698–1702, 1990.[Medline]
  26. Movsesian, M. A., M. Karimi, K. Green, and L. R. Jones. Ca2+-transporting ATPase, phospholamban, and calsequestrin levels in nonfailing and failing human myocardium. Circulation 90: 653–657, 1994.[Abstract]
  27. Movsesian, M. A., C. Leveille, J. Krall, J. Colyer, J. H. Wang, and K. P. Campbell. Identification and characterization of proteins in sarcoplasmic reticulum from normal and failing human left ventricles. J. Mol. Cell. Cardiol. 22: 1477–1485, 1990.[Medline]
  28. Mulieri, L. A., G. Hasenfuss, B. J. Leavitt, P. D. Allen, and N. R. Alpert. Altered myocardial force-frequency relation in human heart failure. Circulation 85: 1743–1750, 1992.[Abstract]
  29. Puglisi, J. L., R. A. Bassani, J. W. Bassani, J. N. Amin, and D. M. Bers. Temperature and relative contributions of Ca transport systems in cardiac myocyte relaxation. Am. J. Physiol. 270 (Heart Circ. Physiol. 39): H1772–H1778, 1996.[Abstract/Free Full Text]
  30. Schwinger, R. H., M. Bohm, U. Schmidt, P. Karczewski, U. Bavendiek, M. Flesch, E. G. Krause, and E. Erdmann. Unchanged protein levels of SERCA II and phospholamban but reduced Ca2+ uptake and Ca2+-ATPase activity of cardiac sarcoplasmic reticulum from dilated cardiomyopathy patients compared with patients with nonfailing hearts. Circulation 92: 3220–3228, 1995.[Abstract/Free Full Text]
  31. Schwinger, R. H. G., M. Bohm, J. Muller-Ehmsen, R. Uhlmann, U. Schmidt, A. Stablein, P. Uberfuhr, E. Kreuzer, B. Reichart, H. Eissner, and E. Erdmann. Effect of inotropic stimulation on the negative force-frequency relationship in the failing human heart. Circulation 88: 2267–2276, 1993.[Abstract]
  32. Tanaka, H., and K. Shiegenobu. Effect of ryanodine on neonatal and adult rat heart: developmental increase in sarcoplasmic reticulum function. J. Mol. Cell. Cardiol. 21: 1305–1313, 1989.[Medline]
  33. Terracciano, C. M., and K. T. MacLeod. Effects of lactate on the relative contribution of Ca2+ extrusion mechanisms to relaxation in guinea-pig ventricular myocytes. J. Physiol. (Lond.) 500: 557–570, 1997.[Abstract]
  34. Vetter, R., M. Kott, W. Schulze, and H. Rupp. Influence of different culture conditions on sarcoplasmic reticular calcium transport in isolated neonatal rat cardiomyocytes. Mol. Cell. Biochem. 188: 177–185, 1998.[Medline]
  35. Vetter, R., R. Studer, H. Reinecke, F. Kolar, I. Ostadalova, and H. Drexler. Reciprocal changes in the postnatal expression of the sarcolemmal Na+/Ca2+-exchanger and SERCA2 in rat heart. J. Mol. Cell. Cardiol. 27: 1689–1701, 1995.[Medline]
  36. Von der Leyen, H., U. Mende, W. Meyer, J. Neumann, M. Nose, W. Schmitz, H. Scholz, J. Starbatty, B. Stein, H. Wenzlaff, V. Doring, P. Kalmar, and A. Haverich. Mechanism underlying the reduced positive inotropic effects of the phosphodiesterase III inhibitors pimobendan, adibendan and saterinone in failing as compared to nonfailing human cardiac muscle preparations. Naunyn Schmiedebergs Arch. Pharmacol. 344: 90–100, 1991.[Medline]
  37. Wolska, B. M., M. O. Stojanovic, W. Luo, E. G. Kranias, and R. J. Solaro. Effect of ablation of phospholamban on dynamics of cardiac myocyte contraction and intracellular Ca2+. Am. J. Physiol. 271 (Cell Physiol. 40): C391–C397, 1996.[Abstract/Free Full Text]