INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ACTIVATION of the sympathetic nervous system leads to the secretion of catecholamines, which stimulate -adrenergic receptors in the heart. Cardiac -adrenergic stimulation activates membrane-bound adenylyl cyclase and leads to the formation of cAMP and activation of protein kinase A (PKA). There are several intracellular proteins that can be phosphorylated by PKA, but the dominant ones quantitatively are phospholamban (PLB) (20, 32, 34) and troponin I (TnI) (19, 33). PLB is a key regulator of the sarcoplasmic reticulum (SR) Ca-ATPase in cardiac cells. Dephosphorylated PLB is closely associated with the SR Ca-ATPase and acts as an inhibitor of the SR Ca pump (13, 15, 30, 31). On phosphorylation by PKA, the inhibitory effect is relieved (20, 34). This allows greater Ca transport at a given intracellular Ca concentration ([Ca]_i), and hence, the heart relaxes faster (lusitropic effect). Phosphorylation of TnI, which interacts with troponin C (TnC), decreases the affinity of TnC for Ca (33). This reduction in Ca affinity is mediated by an increase in the off rate of Ca from the myofilaments and could thereby accelerate cardiac relaxation (28).

Because both PLB and TnI are phosphorylated during -adrenergic stimulation of the myocytes, it is difficult to isolate the individual roles of PLB and TnI on the lusitropic effect of -adrenergic activation in the heart. Indirect studies in ventricular muscle have suggested that the acceleration of relaxation with isoproterenol (Iso) could be attributed mainly to the SR Ca-pump stimulation rather than decreased myofilament Ca sensitivity (9, 26). Indeed, after withdrawal of -adrenergic activation in the intact heart, inotropic and lusitropic states recover with a time course that more closely matches that of the dephosphorylation of PLB than of TnI (35). Recent studies in skinned cardiac muscle fibers using photolysis of caged Ca chelator diazo-2 have been equivocal (14, 38). Whereas Zhang et al. (38) concluded that cardiac TnI phosphorylation increases the rate of muscle relaxation, Johns et al. (14) concluded that TnI phosphorylation was not involved in the cardiac lusitropic effect.

Luo et al. (23) ablated the PLB gene in the mouse and found that basal cardiac contractility and relaxation rates were enhanced. Furthermore, both inotropic and lusitropic effects of -adrenergic activation were attenuated in the PLB knock-out (PLB-KO) mouse, including results from isolated ventricular myocytes (8, 17, 22, 24, 27, 29, 37). These results are clearly consistent with a major role of PLB in regulating relaxation (as well as contractility). However, differences in contraction amplitude and system nonlinearities (7) in previous studies have prevented true quantitative comparison of the relative effects of PLB and TnI phosphorylation on relaxation. In this study, we overcome this limitation to provide clear quantitative data concerning the respective roles of TnI and PLB phosphorylation on the PKA-mediated lusitropic effect.

In the present study, we compare relaxation for contractions of matched amplitude in the presence and absence of Iso in both ventricular myocytes and muscles from both wild-type (WT) and PLB-KO mice (23). This allows direct quantitative comparison of the lusitropic effect of Iso on TnI alone (present in the PLB-KO) versus TnI plus PLB (present in the WT). Our results indicate that phosphorylation of TnI has no effect on the rate of myocyte or muscle relaxation during unloaded contractions (relengthening). On the other hand, when isometric force is developed, TnI phosphorylation produces a modest acceleration of relaxation that is dependent on the load. Thus the lusitropic effect of -adrenergic agonists mediated via TnI is extremely load dependent.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cardiac myocyte and muscle preparation. Isolation of ventricular myocytes from WT and PLB-KO mice was carried out as previously described (22). Briefly, hearts were excised from adult mice (species-matched WT and PLB-KO, 30-45 g) that were anesthetized with pentobarbital sodium (70 mg/kg ip). Hearts were mounted on a Langendorff perfusion apparatus and perfused with nominally Ca-free Tyrode solution for 6 min at 37°C. Perfusion was then switched to the same solution containing 0.8 mg/ml collagenase (type B, Boehringer Mannheim, Indianapolis, IN) and 0.03 mg/ml pronase (Boehringer Mannheim), with perfusion continuing until the heart became flaccid (~7-12 min). The ventricular tissue was then dispersed and filtered. The cell suspension was rinsed several times, and [Ca] was gradually increased to 1 mM. Before experimental use, the myocytes were plated onto Plexiglas superfusion chambers, with the glass bottoms of the chambers treated with laminin (GIBCO, Grand Island, NY) to increase cell adhesion.

Right ventricular papillary muscles were also dissected from WT and PLB-KO mice and mounted in a small volume (0.1 ml) tissue bath (5). One end was attached to a fixed post and the other to a hook connected to a stiff transducer (AE875, Aksjeselskapet Micro-Elektronikk, Horten, Norway) held by a micromanipulator. The muscle tapered from the ventricular wall toward the attachment to the valve such that the two-dimensional shape approached a triangle with a base width of ~0.6 mm and depth of 0.1-0.25 mm at the base and length of 2-3 mm. Force was normalized per cross-sectional area at the base using the approximation area = 0.75 × width × depth. Platinum wires in the chamber wall, parallel to the muscle length, were used to field stimulate the muscle at a normal frequency of 0.5 Hz.

Measurement of shortening and [Ca]_i. Myocyte shortening was measured as previously described (2). Cells were under field stimulation (square waves) at 0.5 Hz, and experiments were conducted at room temperature (23°C). Cells were transilluminated by a red light source (to avoid interference with indo 1 epifluorescence measurement), and shortening was measured using a video-edge detection system (Crescent Electronics, Sandy, UT). This same system was used to monitor muscle shortening after the muscle was slackened to the point of zero developed force, by adjustment of the manipulator connected to the force transducer.

[Ca]_i measurements were conducted simultaneously with contraction measurement in some cells. To obtain [Ca]_i measurements while allowing myocytes to control their own intracellular environment (including [Na]) and action potential, cells were loaded with indo 1 by incubation with the acetoxymethyl ester (AM) form of the dye (indo 1-AM, 10 µM; Molecular Probes, Eugene, OR) for 20 min at room temperature. After loading was completed, the cells were superfused with normal Tyrode solution for at least 30 min to wash out excess indicator and allow deesterification (2). The excitation and emission system were as described (22). Briefly, excitation wavelength was 355 ± 5 nm, and the emission wavelength was selected as 405 ± 20 and 485 ± 20 nm. The average background fluorescence recorded at both wavelengths from cells not loaded with indo 1 was subtracted before the fluorescence ratio was calculated. Cell fluorescence signals were digitized at 120-133 Hz (filtered at 60 Hz) and stored on a computer. Details of the protocols are described in RESULTS.

Phosphorylation of isolated ventricular myocyte proteins. Isolated mouse ventricular myocytes were labeled with ³²P, and protein phosphorylation was determined using a modified version of the protocol described by Wolska et al. (37). Myocytes from a single heart were suspended in 1 ml of the normal Tyrode solution containing 1 mM Ca and incubated with 0.15 mCi [³²P]orthophosphate for 30 min at room temperature. Aliquots (100 µl) of the myocyte suspension were then mixed with 1.0 µM Iso. The reaction was stopped after 30, 60, 90, 120, or 180 s of treatment with Iso by adding SDS-stop solution containing 1 mM dithiothreitol, 30 mM Tris · HCl, 3 mM EDTA, 6% SDS, 15% glycerol, and a trace of bromophenol blue, pH 7.8. For a phosphorylation control, a myocyte aliquot was also incubated with 5 mM dibutyryl cAMP for 20 min, previously shown to produce maximal phosphorylation (21). All samples were boiled in microcentrifuge tubes for 5 min to dissociate the oligomeric state of PLB to its monomeric form before the samples were subjected to SDS-PAGE (using a 4-20% polyacrylamide gradient gel). Aliquots of samples containing 50 µg of myocyte protein, as determined by the Bio-Rad protein assay, were applied to each well. The SDS-PAGE gels were placed in the Storage Phosphor Screen Cassettes (Molecular Dynamics) for overnight exposure. Phosphorimage and densitometric data were obtained using a phosphorimager and the ImageQuant software from Molecular Dynamics. This method does not distinguish which amino acids are phosphorylated but provides simultaneous quantitative data for both PLB and TnI phosphorylation.

Solutions and statistical analaysis. Unless otherwise stated, reagents used were of analytic grade and supplied by Sigma (St. Louis, MO). The normal Tyrode (NT) solution contained (in mmol/l): 140 NaCl, 4 KCl, 1 MgCl₂, 1 or 2 CaCl₂, 10 glucose, and 5 HEPES, with the pH adjusted to 7.4 with NaOH at 23°C.

Values are given as means ± SE (n equals number of myocytes or muscles). Statistical comparisons were based on Student's t-test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Maximal effect of Iso on twitch amplitude. Addition of Iso increases contractions and Ca transients in both WT and PLB-KO myocytes. Figure 1A shows the effect of steady-state exposure to 1 µM Iso on cell contraction amplitude at 1 mM extracellular Ca concentration ([Ca]_o). Cell shortening increased to 196 ± 30% of control after Iso was administered in PLB-KO myocytes (n = 16, P < 0.01). In WT myocytes, contraction amplitude increased to 369 ± 87% of control due to Iso treatment (n = 7, P < 0.05). The percent increase was significantly higher in WT myocytes compared with that found in PLB-KO myocytes (P < 0.05). The results in muscles were qualitatively similar, but Iso increased developed force to only 176 ± 10% in WT and 129 ± 4% in PLB-KO muscles (Fig. 1).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Inotropic effect of isoproterenol (Iso) in phospholamban knock-out (PLB-KO) and wild-type (WT) mouse myocytes and muscles. Superfusate contained 1 mM Ca ± 1 µM Iso. Steady-state (SS) twitch amplitude was compared with that of control (Ctl). ** P < 0.05 for effect of Iso vs. Ctl. * P < 0.05 for difference between effect of Iso on WT vs. PLB-KO.

Matched contractions before and after Iso. Because contraction amplitudes have an intrinsic effect on the interpretation of relaxation kinetics (7), we compared the kinetics of cell relaxation only when contraction amplitudes were matched in the control and Iso. To accomplish this we used elevation of [Ca]_o to raise control twitch contraction amplitude to levels comparable to those achieved by Iso application.

Figure 2 shows the usual experimental protocol. Cells were initially perfused with NT containing 1 mM Ca until twitch contractions reached steady state. The solution was then switched to 2 mM Ca NT, and the cell contractions were allowed to reach a new steady state. [Ca] was then reduced back to 1 mM and the amplitude of contraction returned to the initial level. Finally, 1 µM Iso was added (in the same 1 mM Ca NT), and cell contraction amplitude gradually increased over 1-2 min to a new steady state. Contractions of matching amplitude without (Fig. 2, trace a) and with Iso (trace b) were selected for direct comparison of relaxation kinetics (see superimposed contractions in Fig 2).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Experimental protocol used to match twitch amplitude before and after Iso exposure. An isolated myocyte from a WT mouse was stimulated at 0.5 Hz, and extracellular [Ca] concentration ([Ca]o) was elevated from 1 to 2 mM. Breaks in the record occur between sets of 3 twitches and are for variable intervals, but overall equilibration time at 2 mM [Ca]o was typically 2-3 min. After this time, [Ca]o was returned to 1 mM and equilibrated before exposure to 1 µM Iso. Traces a and b, same amplitude before (Ctl) and after Iso, respectively. These twitches are superimposed at bottom. This allows assessment of effects of Iso on relaxation, uncomplicated by differences in twitch amplitude (7).

In WT myocytes, addition of Iso or elevation of [Ca] to 2 mM had similar effects on contraction amplitude, but typically the steady-state twitch in 2 mM Ca NT was compared with a twitch with Iso, which occurred slightly before the eventual maximum steady state with Iso (as in the case in Fig. 2). Not surprisingly for PLB-KO myocytes, Iso did not increase twitch amplitude nearly as much as that observed in the WT myocytes (see also Fig. 1). Thus the PLB-KO cells were always in steady state with respect to Iso exposure for the contraction comparisons. Indeed, in >50% of the PLB-KO cells, addition of 1 µM Iso (in 1 mM Ca NT) did not increase contraction amplitude as much as raising [Ca]_o from 1 to 2 mM. In these cases [Ca]_o was also increased to 2 mM in the continued presence of 1 µM Iso. This allowed selection of a twitch in the sustained presence of Iso during the phase of increasing twitch amplitude toward a final steady state with 2 mM [Ca]_o (for comparison with a control steady state twitch in 2 mM [Ca]_o without Iso). Thus we can be particularly confident that the PLB-KO cells were being compared with maximal Iso stimulation.

Phosphorylation of TnI and PLB in intact myocytes. For this protocol to be appropriate, it is important to know whether PLB and TnI really get phosphorylated as expected during the exposure to Iso. Figure 3A shows incorporation of [³²P]orthophosphate into cellular proteins in intact isolated mouse ventricular myocytes that had been treated for different times with 1 µM Iso to mimic our experimental conditions. The last lane in Fig. 3A shows the effect of exposure to 5 mM dibutyryl cAMP for 20 min as a control for maximal phosphate incorporation. Figure 3B shows pooled results for TnI and PLB phosphorylation normalized to the value for dibutyryl cAMP (taken as 100%). There was some basal phosphorylation of PLB and TnI before Iso exposure (17, 21, 37). However, exposure to 1 µM Iso caused a marked and parallel increase in phosphorylation of both TnI and PLB, reaching an apparent maximum within 2 min. This is similar to previous results in intact WT and PLB-KO hearts (17), where there was also no difference in basal TnI phosphorylation between WT and PLB-KO hearts. Moreover, the maximum level of phosphorylation achieved in 2 min of Iso in Fig. 3 was ~85% of that attained by 20 min of exposure to 5 mM dibutyryl cAMP. These results indicate that Iso treatment produced a marked increase in phosphorylation and that TnI and PLB get phosphorylated to similar extents and with similar time courses.

View larger version (46K): [in this window] [in a new window]   Fig. 3.   Phosphorylation of cardiac troponin I (TnI) and PLB in isolated WT mouse ventricular myocytes. A: representative autoradiograph showing time course of 32P incorporation into TnI and PLB after treatment with 1 µM Iso or 5 mM dibutyryl cAMP (db-cAMP, 20 min) in presence of 1 mM Ca and 0.15 mCi [32P]orthophosphate. B: phosphorylation of TnI and PLB relative to levels obtained in same preparation by 5 mM dibutyryl cAMP, used as a control. Values are means ± SE from 4 preparations. LC2, myosin light chain.

Relaxation of twitches in isolated myocytes. Figure 4, A and B, shows superimposed cell shortening traces in control and Iso from an individual PLB-KO and WT myocyte, respectively. Time to peak and overall duration of contraction were much shorter in PLB-KO than WT myocytes. The time constant () for control cell relaxation during steady-state twitch was also significantly faster in the PLB-KO mouse ( = 65 ± 4 ms in WT vs. 28.7 ± 2 ms in PLB-KO; n = 5 and 8, respectively; P < 0.001). This is consistent with previous reports (22, 37).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Myocyte twitches in PLB-KO and WT myocytes of matched amplitudes in presence and absence of Iso. Superimposed twitches in control and after Iso exposure in PLB-KO (A) and WT myocyte (B). Pooled data from experiments like those in Figs. 2 and 3, A and B, for exponential time constants () of relaxation (C) and amplitude of contraction, relative to Ctl (D) (n = 8 and 5 for PLB-KO and WT, respectively). Control twitch amplitude in PLB-KO and WT myocytes were 15.5 ± 2.7 and 12.2 ± 3.3% of resting cell length (RCL), respectively. These values served as normalization points for amplitude data in D. D shows amplitudes with respect to control. This does not indicate population variance but does show how well matched amplitudes were before and after Iso. ** P < 0.05 for effect of Iso.

Contraction amplitudes in Fig. 4, A and B, were well matched before and after Iso treatment in both PLB-KO and WT myocytes. In the PLB-KO myocyte, cell relaxation followed almost identical kinetics ( was 23 and 24 ms, before and after Iso, respectively). On the other hand, the WT myocyte relaxed much faster after Iso treatment, with  of 34 ms in Iso versus 72 ms in control. Figure 4C shows pooled data from experiments like those in Figs. 2 and 4, A and B. In PLB-KO myocytes there was no change in the rate of cell relaxation ( = 28.7 ± 2.2 ms in control vs. 28.9 ± 3.3 ms in Iso, n = 8). In contrast in WT myocytes, the rate of relaxation was much faster after Iso treatment ( = 65 ± 4 ms in control vs. 46 ± 5 ms in Iso, n = 5; P < 0.01). These kinetic comparisons were based on contractions of matched amplitude, as shown in Fig 4D. These results suggest that for matched contraction amplitudes, 1 µM Iso does not accelerate myocyte relaxation in the absence of PLB.

With 1 µM Iso, the relaxation  in the WT (46 ± 5) was still almost 60% longer than the relaxation in the PLB-KO with or without Iso treatment (29 ms). One possible explanation is that even with extremely strong -adrenergic activation, PLB still partially inhibits Ca transport by the SR Ca-ATPase (16, 27). Other adaptive changes in the PLB-KO cannot be ruled out.

Indo 1-loaded cells and Ca transients. In some cells, intracellular Ca transients were also measured. However, the data in Figs. 2 and 4 were from cells that were not loaded with intracellular Ca indicators. This was done to avoid the potential effects of added intracellular Ca buffer on relaxation kinetics. Indeed, in indo 1-loaded cells the control  of relaxation was longer than for non-indo 1-loaded cells for both WT (115 ± 27 vs. 65 ± 4 ms) and PLB-KO cells (46 ± 9 vs. 29 ± 2 ms). The higher variance for indo 1-loaded cells may also indicate cell-to-cell variability in the extent of indo 1 loading. Twitch amplitude was also reduced in indo 1-loaded cells (in PLB-KO from 15.6 ± 2.7 to 11.0 ± 2.7% of resting cell length and in WT from 12.2 ± 3.3 to 7.8 ± 0.7% of resting cell length).

For cells in which Ca transients were measured together with relaxation, the protocol was still the same, and twitches with Iso were matched to control by contraction amplitude (101 ± 1% for WT and 101 ± 2% for PLB-KO). As in the non-indo 1-loaded cells (Fig. 4B),  of relaxation was unaltered by Iso in the PLB-KO myocytes (46 ± 9 vs. 45 ± 9) but was strongly accelerated in the WT myocytes (115 ± 27 vs. 54 ± 13; P < 0.05).

Figure 5 shows Ca transients for contractions where the amplitude of contraction was matched. The time course of [Ca]_i decline was similar before and after Iso treatment in PLB-KO myocytes (Fig. 5A). In WT myocytes, Iso again accelerated [Ca]_i decline (Fig. 5B). Average results are shown in Fig. 5, C and D. As for relaxation, [Ca]_i decline was greatly accelerated by Iso in the WT but not significantly altered in the PLB-KO.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Myocyte Ca transients from cells in which twitch amplitudes were matched in PLB-KO and WT myocytes in presence and absence of Iso. Superimposed intracellular Ca concentration ([Ca]i) records in control and after Iso exposure in PLB-KO (A) and WT myocyte (B). F405/F485, fluorescence ratio between 405 and 485 nm. Pooled data from these experiments for exponential  of [Ca]i decline (C) and amplitude of Ca transient ([Ca]i), relative to Ctl (D) (n = 7 and 4 for PLB-KO and WT, respectively). Control twitch amplitude in PLB-KO and WT myocytes were 0.06 ± 0.02 and 0.03 ± 0.01 ratio units, respectively. These values served as normalization points for amplitude data in D. ** P < 0.05 for effect of Iso.

Figure 5D shows that the amplitudes of Ca transients were similar, where contraction amplitudes were matched (±Iso). Because PKA-dependent phosphorylation of TnI is known to decrease myofilament Ca sensitivity, it was expected that higher Ca transient peaks might be required for comparable contractions after Iso. Whereas there was a slight trend toward higher [Ca]_i for matched contraction amplitude (especially in the WT), the difference did not reach statistical significance.

Relaxation of isometric force in multicellular preparations. The above-mentioned experiments were carried out in isolated myocytes that were shortening without appreciable external load. The influence of TnI phosphorylation on relaxation may be altered when the myocytes generate substantial force. Thus parallel experiments were performed to measure force development and relaxation in papillary muscles from WT and PLB-KO mice.

Figure 6 shows results from muscle force measurements exactly analogous to those in Fig. 4 with myocyte shortening. Again, Iso dramatically accelerated relaxation of force in the WT by ~50% (from  = 80.5 ± 10.3 to 41.5 ± 8.4 ms). The effect of Iso on the  of force relaxation in the PLB-KO was modest (17.5%) but significant (from 38.5 ± 2.5 to 31.8 ± 2.0 ms). This contrasts with results with unloaded myocyte shortening (Fig. 4C) and indicates that there may be a load dependence to the lusitropic effect of TnI phosphorylation.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Muscle force development in PLB-KO and WT with matched amplitudes in presence and absence of Iso. Superimposed twitches in control and after Iso exposure in PLB-KO (A) and WT myocyte (B). Pooled data for exponential  of relaxation (C) and amplitude of contraction, relative to Ctl with 2 mM [Ca]o (D). Data are shown for both Ctl with [Ca]o of 1 mM (1Ca) and 2 mM (2Ca). Actual matching of contraction amplitudes (±Iso) was extremely close, as in Fig. 4D (101.6 ± 1.1% for PLB-KO and 100.8 ± 0.8 for WT; n = 8 and 5 for PLB-KO and WT, respectively). ** P < 0.05 for effect of Iso.

Figure 6, C and D, also includes control data in 1 mM [Ca]_o (i.e., when twitch amplitudes were not matched with Iso twitches). It can be seen that relaxation is a bit slower for the smaller amplitude control twitches for both PLB-KO and WT, as expected for Ca transients (7). Even without our method to match twitch amplitude for  comparisons, the effect of Iso on both WT and PLB-KO relaxation is apparent.

Load dependence of lusitropic effect of Iso. To explore this issue further, muscle and sarcomere length were adjusted to vary developed force, from maximal to zero developed force (in which case unloaded muscle shortening was measured). Figure 7A shows that in PLB-KO muscle, relaxation during unloaded shortening was not altered by Iso, as seen for myocyte shortening (Fig. 4C). However, with higher developed force (Fig. 7B and C) there is a progressive increase in the lusitropic effect of Iso in the PLB-KO. This can be appreciated in Fig. 7D where the ratio of  with Iso versus control (_Iso/_Ctl) is shown as a function of developed force.

View larger version (23K): [in this window] [in a new window]   Fig. 7.   PLB-KO muscles show load-dependent, Iso-induced acceleration of relaxation. A: superimposed unloaded muscle shortening traces before and after Iso exposure (Ctl trace has been normalized to allow more direct comparison of time course). B: developed force for same muscle at different muscle lengths in absence and presence of Iso. Broken-lined curves, in presence of Iso. Resting force has been subtracted to allow comparison of active force traces for twitches where developed force was equal before and after Iso (i.e., muscle length and resting force were necessarily different). C:  of relaxation are indicated for muscle in A and B. D: ratio of  values from C (Iso/ctl) and regression slope was significantly different from zero.

Figure 8 shows analogous results in WT muscles. In the WT mouse the control relaxation is much slower than that observed in the PLB-KO mouse. In addition, the acceleration of relaxation induced by Iso is comparable at all levels of developed force, including unloaded shortening. The load-independent effect of Iso in the WT mouse is emphasized by _Iso/_Ctl in Fig. 8D. This is consistent with PLB phosphorylation being dominant in the WT, obscuring a relatively small effect of TnI phosphorylation on relaxation.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   WT muscles do not show load-dependent, Iso-induced acceleration of relaxation. A: superimposed unloaded muscle shortening traces before and after Iso exposure (Ctl trace has been normalized to allow direct comparison of time course). B: developed force for same muscle at different muscle lengths in absence and presence of Iso. Broken-lined curves, in presence of Iso. Resting force has been subtracted to allow comparison of active force traces for twitches where developed force was equal before and after Iso (i.e., muscle length and resting force were necessarily different). C:  of relaxation are indicated for muscle in A and B. D: Iso/ctl using values at nearly matched force levels.

In these experiments a different strategy was used to match twitch contraction amplitudes before and after Iso (altering muscle and sarcomere length vs. changing [Ca]_o as in Figs. 2, 4-6). This has the advantage that Ca transients are not expected to change appreciably with sarcomere length. Thus similar results here serve as a control to rule out potential complicating effects of altered [Ca]_o (such as activation of [Ca]_i-dependent kinases and phosphatases). When force is varied by changing sarcomere length, there was also no apparent acceleration of relaxation as a function of load (Figs. 7D and 8D). This contrasts with the effects when amplitude was varied by changing [Ca]_o or frequency (Fig. 6C, Refs. 7 and 22).

Figure 9A shows pooled data for _Iso/_Ctl in WT and PLB-KO muscles from experiments like Figs. 7 and 8. The slope of the regression line is significantly different from zero only in the PLB-KO case. The data in Fig. 9A may create the false impression that Iso causes acceleration of relaxation in PLB-KO that reaches the same extent as in WT. However, the control relaxation is very different in the PLB-KO versus the WT such that the fast [Ca]_i decline in PLB-KO optimizes conditions for detection of an effect of TnI phosphorylation on relaxation.

View larger version (20K): [in this window] [in a new window]   Fig. 9.   TnI phosphorylation contributes to Iso-induced acceleration of relaxation only when force is generated. A: pooled results of force dependence of Iso-induced acceleration of relaxation based on experiments as in Figs. 7 and 8 (n = 4 for PLB-KO and n = 3 for WT). Regression line slope was significantly different from zero only for PLB-KO. B: percentage of Iso-induced acceleration of relaxation attributed to PLB and TnI phosphorylation for unloaded shortening and during isometric contractions (see text for details).

Figure 9B shows the approximate fraction of  acceleration, which might be attributed to PLB phosphorylation versus TnI phosphorylation. For this simplified analysis we used  data from 1) control WT and PLB-KO + Iso as the upper and lower limits for , 2) the  shift by Iso in PLB-KO as the TnI component, and 3) the  shift in PLB-KO versus WT as the maximum PLB component. We infer that without mechanical load, 100% of the lusitropic effect of Iso in normal mouse heart is due to PLB phosphorylation (for both myocyte and muscle). Similarly, under near-maximal isometric load the TnI phosphorylation contributes 14%, whereas PLB contributes 86%. These values were from the experiments where [Ca]_o was altered to match force before and after Iso (Fig. 4). Similar values were obtained when alteration of sarcomere length was used (Figs. 7 and 8).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We found that in the absence of PLB, cardiac relaxation at low mechanical load is not accelerated by -adrenergic receptor activation. However, with increasing developed force in PLB-KO, Iso-dependent TnI phosphorylation can accelerate relaxation. The key novel aspects that allow quantitative analysis of this issue here are the use of myocytes and muscles from PLB-KO mice and comparing twitches of the same amplitude. Whereas Na/Ca exchange can contribute to relaxation, the contribution in the mouse is very small and not regulated by PKA (22).

PLB and TnI phosphorylation may both alter relaxation. It has been known for some time that PLB phosphorylation accelerates Ca transport by the cardiac SR Ca-ATPase and that this participates in accelerated relaxation in response to -adrenergic activation (18, 34). Indeed, the extremely rapid relaxation in the PLB-KO mouse at the level of the intact animal, isolated heart, and myocyte emphasizes the important natural inhibitory effect of PLB on the SR Ca-ATPase and relaxation (8, 12, 17, 22, 23, 29, 37). Furthermore, there is much less inotropic or lusitropic effect of Iso in the PLB-KO mouse compared with the WT mouse (22, 23, 27, 37). Thus there are compelling data indicating a major contribution of PLB phosphorylation to the lusitropic effect of -adrenergic agonists.

There are also compelling data to indicate that -adrenergic agonists lead to phosphorylation of TnI and that this decreases Ca affinity of the myofilaments via an increase in dissociation rate of Ca from TnC (28, 33, 38). Indeed, Zhang et al. (38) argued that PKA-dependent phosphorylation of TnI also contributes to the lusitropic effect of -adrenergic agonists, based on flash photolysis of the caged Ca chelator diazo-2 in porcine skinned fibers (relaxation half-time decreased from 110 to 70 ms). On the other hand, in similar experiments in skinned guinea pig ventricular trabeculae, Johns et al. (14) found no effect of PKA phosphorylation on the relaxation  (90 ms).

There have been few studies that have compared the influence of these two effects of PKA on relaxation in intact ventricular myocytes. McIvor et al. (26) used acetylcholine (which increased myofilament Ca sensitivity) to offset the effect of TnI phosphorylation by PKA but found that muscle relaxation was still accelerated. This clever strategy made a strong case for the dominance of PLB versus TnI phosphorylation in the lusitropic effect of Iso, but the use of acetylcholine limited the interpretation in quantitative terms. Talosi et al. (35) showed that withdrawal of Iso in the intact heart led to the recovery of the inotropic and lusitropic state comparable in time to the dephosphorylation of PLB rather than TnI (which occurred more slowly). A key limitation has always been that the TnI and PLB effects coexist in the cell, and if the TnI effect is somewhat less than the PLB effect, it may be more difficult to detect in the normal cellular environment (with a larger PLB effect).

Lusitropic effect of Iso in WT and PLB-KO myocytes. In the present study, we took advantage of the PLB-KO mouse to examine the impact of TnI phosphorylation in the complete absence of PLB. To extend previous work with the PLB-KO mouse above, we also took great care to compare twitches of the same amplitude before and after Iso treatment. This eliminated complications that may accrue from nonlinearities in the numerous processes involved when one uses normalization to compare twitches without Iso to the larger twitches produced by Iso exposure.

The mouse ventricular action potential lacks a plateau phase and is very short (like in the rat), such that it is essentially over before relaxation begins. This avoids potential complications due to the effects of action potential duration on relaxation. These concerns may be more relevant for larger mammals, which have a long action potential plateau phase and where [Ca]_i decline is considerably slower than in the mouse. The mouse (like the rat) also has a relatively weak Na/Ca exchange system, which is the main Ca transport system that competes with the SR Ca-ATPase (3, 22). This system is therefore also unlikely to confound interpretation of our experiments.

The very rapid [Ca]_i decline and relaxation in the PLB-KO mouse myocytes is perhaps the fastest natural cellular [Ca]_i decline among cardiac species normally studied. This should really enhance the ability to detect a rate-limiting dissociation of Ca from TnC if it is physiologically relevant. It was thus initially surprising that even in the PLB-KO mouse, Iso and phosphorylation of TnI had no effect on myocyte relaxation (when contraction amplitude was matched). This suggests that even during these very rapid Ca transients (with [Ca]_i decline much faster than normal), Ca transport rather than Ca dissociation from TnC is the rate-limiting step in the relaxation of isolated ventricular myocytes. The same observations were made in muscle preparations when they were not mechanically loaded. An important implication from this is that transport of Ca out of the cytosol is extremely critical in both the development of diastolic dysfunction and the strategies aimed at improving cardiac relaxation.

Because force can alter Ca dissociation from the myofilaments (1), it was also important to examine this issue during relatively isometric contractions. Whereas the effect of Iso on  of force relaxation is much smaller in the PLB-KO than in WT muscle (Fig. 6), the effect is highly significant. Our results do not determine which proteins are responsible for the load dependence of relaxation effected by PKA. Moreover, this lusitropic effect becomes negligible as the developed force is reduced (Fig. 7). In contrast, the lusitropic effect of Iso in WT muscle was independent of load and likely to be almost completely due to PLB phosphorylation (Fig. 8). It is possible that with the slower Ca transients in WT mice with PLB present, Ca dissociation from the myofilaments is not rate limiting, but [Ca]_i decline is (even with Iso).

A simplified comparison of relaxation  values allows the first quantitative estimates of the relative contributions of PLB and TnI phosphorylation to the Iso-induced acceleration of relaxation in the WT mouse (Fig. 9B). The values of 14-18% for TnI and 82-86% for PLB are certainly consistent with previous work, which suggested that PLB was more important for this effect (22, 24, 26, 35, 37). It should be reemphasized that the [Ca]_i decline in PLB-KO mouse myocytes is much faster than in larger mammals and even in comparison to the WT mouse. Because a very fast [Ca]_i decline creates a situation where dissociation of Ca from TnC is more likely to be rate limiting, 14-18% may be an upper limit for the contribution of TnI phosphorylation to the kinetics of cardiac relaxation in the intact myocyte. This would be especially true for myocytes in which the  of relaxation is much slower than the 20-30 ms observed in the PLB-KO mouse. Indeed, there was no significant force-dependent acceleration of relaxation in the presence of Iso, even in the WT mouse (Fig. 8). This may have been expected if TnI phosphorylation was a significant contributor based on the characteristic effect of load in the PLB-KO with Iso (Fig. 7). We conclude that TnI phosphorylation and acceleration of Ca dissociation are probably extremely minor contributors to the normal lusitropic effect of -adrenergic activation in heart. Rather, the rate of Ca transport and [Ca]_i decline are likely the main rate-limiting steps during relaxation.

Inotropic effect of Iso in WT and PLB-KO myocytes. The two most important primary factors in mediating the positive inotropic effects of PKA activation are generally accepted to be phosphorylation of L-type Ca channels and PLB (6). Of course there could be other PKA targets that contribute to the inotropic state, such as ryanodine receptor (36) or C protein (11, but see also Ref. 10). TnI phosphorylation would by itself be negatively inotropic such that increases in Ca transient would have to more than offset this effect to allow increased contraction. For simplicity, we will focus on PLB and Ca current effects and assume that enhanced Ca current is the primary cause of the Iso-induced doubling of twitch amplitude in the PLB-KO.

Because Ca influx is only a very small fraction of activator Ca in rat and mouse myocytes (6, 22), the two- to threefold increase in Ca current typical with Iso (25) cannot be directly responsible for increased myofilament activation. A more likely explanation is the following. With more Ca entry per twitch, even an unaltered SR Ca pump (in PLB-KO) will result in higher SR Ca content and thus a greater pool of SR Ca will be available for release. Furthermore, the higher Ca current with Iso will provide a larger trigger of Ca-induced Ca release, resulting in increased fractional SR Ca release. In addition, even a modest increase in SR Ca load may increase the fraction of SR Ca released for a given Ca current trigger (4). There may also be increased Ca sensitivity of SR Ca release due to PKA-dependent phosphorylation of the ryanodine receptor (36). Thus there are multiple synergistic explanations for the increase of twitch amplitude even without PLB present. Clearly, these can overcome the negative effect of reduced myofilament Ca sensitivity induced by TnI phosphorylation. If we then add the stimulatory effect of PKA phosphorylation of PLB on the SR Ca-ATPase in the WT mouse, we would expect a much greater increase in SR Ca content available for release and consequently also greater fractional release. This could easily explain the twofold greater inotropic effect of Iso in the WT versus the PLB-KO mouse.

In conclusion, the effect of -adrenergic activation on relaxation is mediated entirely by PLB phosphorylation in the absence of external mechanical load. However, during isometric contractions TnI phosphorylation may contribute up to 14-18% of the acceleration of relaxation of developed force in response to Iso in WT mouse.