Confocal imaging of CICR events from isolated and immobilized SR vesicles.

The Ca2+ concentration inside the sarcoplasmic reticulum ([Ca2+]SR) is a difficult parameter to measure in ventricular cardiac myocytes. Interference from Ca2+-sensitive dye loading into cellular compartments other than the SR interferes with free Ca2+ measurement. In addition, the composition of the cytosol surrounding the SR in intact cells cannot be easily controlled. We have developed a method to measure localized [Ca2+]SR in immobilized membrane vesicles during rapid solution switches. Ca2+ uptake and release in rat SR membrane vesicles was monitored using confocal microscopy. Vesicles were immobilized on a coverslip using an agarose matrix. Perfusion with a Ca2+-containing solution supplemented with ATP initiated SR Ca2+ uptake, causing a rise in intravesicular fluorescence in vesicles containing the low-affinity Ca2+ indicator fluo-5N. Perfusion with caffeine caused SR Ca2+ release and a decrease in intravesicular flourescence. Although caffeine-dependent release was readily visible with extravesicular Ca2+-green, Ca2+ which leaked from the SR was detected only indirectly as eventless release. We conclude that SR Ca2+ uptake and release can be selectively measured in functional SR vesicles using a confocal microscope. Caffeine-dependent release is directly measurable though SR Ca2+ leak can only be inferred as subresolution events, presumably because channels in separate vesicles were not close enough to result in concerted Ca2+-induced Ca2+ release.

Allosteric regulation of Na/Ca exchange current by cytosolic Ca in intact cardiac myocytes.

The cardiac sarcolemmal Na-Ca exchanger (NCX) is allosterically regulated by [Ca](i) such that when [Ca](i) is low, NCX current (I(NCX)) deactivates. In this study, we used membrane potential (E(m)) and I(NCX) to control Ca entry into and Ca efflux from intact cardiac myocytes to investigate whether this allosteric regulation (Ca activation) occurs with [Ca](i) in the physiological range. In the absence of Ca activation, the electrochemical effect of increasing [Ca](i) would be to increase inward I(NCX) (Ca efflux) and to decrease outward I(NCX). On the other hand, Ca activation would increase I(NCX) in both directions. Thus, we attributed [Ca](i)-dependent increases in outward I(NCX) to allosteric regulation. Ca activation of I(NCX) was observed in ferret myocytes but not in wild-type mouse myocytes, suggesting that Ca regulation of NCX may be species dependent. We also studied transgenic mouse myocytes overexpressing either normal canine NCX or this same canine NCX lacking Ca regulation (Delta680-685). Animals with the normal canine NCX transgene showed Ca activation, whereas animals with the mutant transgene did not, confirming the role of this region in the process. In native ferret cells and in mice with expressed canine NCX, allosteric regulation by Ca occurs under physiological conditions (K(mCaAct) = 125 +/- 16 nM SEM approximately resting [Ca](i)). This, along with the observation that no delay was observed between measured [Ca](i) and activation of I(NCX) under our conditions, suggests that beat to beat changes in NCX function can occur in vivo. These changes in the I(NCX) activation state may influence SR Ca load and resting [Ca](i), helping to fine tune Ca influx and efflux from cells under both normal and pathophysiological conditions. Our failure to observe Ca activation in mouse myocytes may be due to either the extent of Ca regulation or to a difference in K(mCaAct) from other species. Model predictions for Ca activation, on which our estimates of K(mCaAct) are based, confirm that Ca activation strongly influences outward I(NCX), explaining why it increases rather than declines with increasing [Ca](i).

Phospholamban decreases the energetic efficiency of the sarcoplasmic reticulum Ca pump.

We tested the hypothesis that increased Sarcoplasmic reticulum (SR) Ca content ([Ca](SRT)) in phospholamban knockout mice (PLB-KO) is because of increased SR Ca pump efficiency defined by the steady-state SR [Ca] gradient. The time course of thapsigargin-sensitive ATP-dependent (45)Ca influx into and efflux out of cardiac SR vesicles from PLB-KO and wild-type (WT) mice was measured at 100 nm free [Ca]. We found that PLB decreased the initial SR Ca uptake rate (0.13 versus 0.31 nmol/mg/s) and decreased steady-state (45)Ca content (0.9 versus 4.1 nmol/mg protein). Furthermore, at similar total SR [Ca], the pump-mediated Ca efflux rate was higher in WT (0.065 versus 0.037 nmol/mg/s). The pump-independent leak rate constant (k(leak)) was also measured at 100 nm free [Ca]. The results indicate that k(leak) was < 1% of pump-mediated backflux and was not different among nonpentameric mutant PLB (PLB-C41F), WT pentameric PLB (same expression level), and PLB-KO. Therefore differences in passive SR Ca leak cannot be the cause of the higher thapsigargin-sensitive Ca efflux from the WT membranes. We conclude that the decreased total SR [Ca] in WT mice is caused by decreased SR Ca influx rate, an increased Ca-pump backflux, and unaltered leak. Based upon both thermodynamic and kinetic analysis, we conclude that PLB decreases the energetic efficiency of the SR Ca pump.

Regulatory role of phospholamban in the efficiency of cardiac sarcoplasmic reticulum Ca2+ transport.

Phospholamban is an inhibitor of the sarcoplasmic reticulum Ca(2+) transport apparent affinity for Ca(2+) in cardiac muscle. This inhibitory effect of phospholamban can be relieved through its phosphorylation or ablation. To better characterize the regulatory mechanism of phospholamban, we examined the initial rates of Ca(2+)-uptake and Ca(2+)-ATPase activity under identical conditions, using sarcoplasmic reticulum-enriched preparations from phospholamban-deficient and wild-type hearts. The apparent coupling ratio, calculated by dividing the initial rates of Ca(2+) transport by ATP hydrolysis, appeared to increase with increasing [Ca(2+)] in wild-type hearts. However, in the phospholamban-deficient hearts, this ratio was constant, and it was similar to the value obtained at high [Ca(2+)] in wild-type hearts. Phosphorylation of phospholamban by the catalytic subunit of protein kinase A in wild-type sarcoplasmic reticulum also resulted in a constant value of the apparent ratio of Ca(2+) transported per ATP hydrolyzed, which was similar to that present in phospholamban-deficient hearts. Thus, the inhibitory effects of dephosphorylated phospholamban involve decreases in the apparent affinity of sarcoplasmic reticulum Ca(2+) transport for Ca(2+) and the efficiency of this transport system at low [Ca(2+)], both leading to prolonged relaxation in myocytes.

Reverse mode of the sarcoplasmic reticulum calcium pump and load-dependent cytosolic calcium decline in voltage-clamped cardiac ventricular myocytes.

We have characterized [Ca](i) decline in voltage-clamped rabbit ventricular myocytes with progressive increases in sarcoplasmic reticulum (SR) calcium load. "Backflux" through the SR calcium pump is a critical feature which allows realistically small values for SR calcium leak fluxes to be used. Total cytosolic calcium was calculated from the latter part of [Ca](i) decline using rate constants for cellular calcium buffers. Intra-SR calcium buffering characteristics were also deduced. We found that the net SR calcium pump flux and rate of [Ca](i) decline decreased as the SR free [Ca] rose, with pump parameters held constant. We have therefore characterized for the first time in intact myocytes both forward and reverse SR calcium pump kinetics as well as intra-SR calcium buffering and SR calcium leak. We conclude that the reverse flux through the SR calcium pump is an important factor in comprehensive understanding of dynamic SR calcium fluxes.

Potentiation of fractional sarcoplasmic reticulum calcium release by total and free intra-sarcoplasmic reticulum calcium concentration.

Our aim was to measure the influence of sarcoplasmic reticulum (SR) calcium content ([Ca](SRT)) and free SR [Ca] ([Ca](SR)) on the fraction of SR calcium released during voltage clamp steps in isolated rabbit ventricular myocytes. [Ca](SRT), as measured by caffeine application, was progressively increased by conditioning pulses. Sodium was absent in both the intracellular and in the extracellular solutions to block sodium/calcium exchange. Total cytosolic calcium flux during the transient was inferred from I(Ca), [Ca](SRT), [Ca](i), and cellular buffering characteristics. Fluxes via the calcium current (I(Ca)), the SR calcium pump, and passive leak from the SR were evaluated to determine SR calcium release flux (J(rel)). Excitation-contraction (EC) coupling was characterized with respect to both gain (integral J(rel)/integral I(Ca)) and fractional SR calcium release. Both parameters were virtually zero for a small, but measurable [Ca](SRT). Gain and fractional SR calcium release increased steeply and nonlinearly with both [Ca](SRT) and [Ca](SR). We conclude that potentiation of EC coupling can be correlated with both [Ca](SRT) and [Ca](SR). While fractional SR calcium release was not linearly dependent upon [Ca](SR), intra-SR calcium may play a crucial role in regulating the SR calcium release process.

Ca flux, contractility, and excitation-contraction coupling in hypertrophic rat ventricular myocytes.

Left ventricular hypertrophy (approximately 40%) was induced in rats by banding of the abdominal aorta. After 16 wk, ventricular homogenates were prepared for biochemical measurements and ventricular myocytes were isolated for functional studies. In myocytes, the effects of banding on intracellular Ca handling, contraction, and excitation-contraction (E-C) coupling were determined using indo 1 fluorescence and whole cell voltage clamp. After steady-state field or voltage-clamp stimulation to load the sarcoplasmic reticulum (SR), SR Ca content assessed by caffeine-induced Ca transients was the same in sham and banded groups. Despite this, cell shortening amplitudes were significantly depressed in the banded group, suggesting altered contractile properties. In banded rats, the SR Ca-adenosinetriphosphatase (Ca-ATPase) mRNA level was reduced, as was homogenate thapsigargin-sensitive SR Ca-ATPase, but cytosolic free Ca concentration ([Ca]i) decline attributed to SR Ca-ATPase activity in intact cells was not slowed. Banding also reduced Na/Ca exchange mRNA level but did not affect either Na-dependent sarcolemmal 45Ca transport in homogenate or the rate of [Ca]i decline in intact cells attributed to Na/Ca exchange (during caffeine-induced contractures). Banding also did not change the rate of [Ca]i decline mediated by the combined function of the mitochondrial Ca uptake and sarcolemmal Ca-ATPase in intact cells. Ca current (ICa) density and voltage dependence were the same in sham and banded groups. Ryanodine receptor mRNA, protein content, and ryanodine affinity were also unchanged in the banded group. At 1 mM extracellular Ca concentration ([Ca]o), banding did not affect E-C coupling efficacy in intact cells under voltage clamp (i.e., same contraction for given ICa and SR Ca load). However, when [Ca]o was reduced to 0.5 mM, the efficacy of E-C coupling was greatly depressed in the banded group (even though ICa and SR Ca content were matched). In summary, unloaded myocyte contraction was depressed in these hypertrophic hearts, but Ca transport was little altered, at 1 mM [Ca]o. However, reduction of [Ca]o to 0.5 mM appears to unmask a depressed fractional SR Ca release in response to a given ICa trigger and SR Ca load.

Passive Ca2+ binding in ventricular myocardium of neonatal and adult rats.

In this study, passive Ca2+ binding was determined in ventricular homogenates (VH) from neonatal (4-6 days) and adult rats, as well as in digitonin-permeabilized adult ventricular myocytes. Ca2+ binding sites, both endogenous and exogenous (Indo-1 and BAPTA) were titrated. Sarcoplasmic reticulum and mitochondrial Ca2+ uptake were blocked by thapsigargin and Ru360, respectively. Free [Ca2+] ([Ca2+]F) was measured with Indo-1 and bound Ca2+ ([Ca2+]B) was the difference between [Ca2+]F and total Ca2+. Apparent Ca2+ dissociation constants (Kd) for BAPTA and Indo-1 were increased by 10-20 mg VH protein/ml (from 0.35 to 0.92 microM for Indo-1 and from 0.20 to 0.76 microM for BAPTA) and also by ruthenium red in the case of Indo-1. Titration with successive CaCl2 additions (2.5-10 nmoles) yielded delta[Ca2+]B/delta[Ca2+]F for the sum of [Ca2+]B at all three classes of binding sites. From this function, the apparent number of endogenous sites (Ben) and their Kd (Ken) were determined. Similar Ken values were obtained in neonatal and adult VH, as well as in adult myocytes (0.68 +/- 0.14 microM, 0.69 +/- 0.13 microM and 0.53 +/- 0.10 microM, respectively). However, Ben was significantly higher in adult myocytes than in adult VH (1.73 +/- 0.35 versus 0.70 +/- 0.12 nmol/mg protein, P < 0.01), which correspond to approximately 300 and 213 mumol/l cytosol. This indicates that binding sites are more concentrated in myocytes than in other ventricular components and that Ben determined in VH underestimates cellular Ben by 29%. Although Ben values in nmol/mg protein were similar in adult and neonatal VH (0.69 +/- 0.12), protein content was much higher in adult ventricle (125 +/- 7 versus 80 +/- 1 mg protein/g wet weight, P < 0.01). Expressing Ben per unit cell volume (accounting for fractional mitochondrial volume, and 29% dilution in homogenate), the passive Ca2+ binding capacity at high-affinity sites is approximately 300 and 176 mmol/l cytosol in adult and neonatal rat ventricular myocytes, respectively. Additional estimates suggest that passive Ca2+ buffering capacity in rat ventricle increases markedly during the first two weeks of life and that adult levels are attained by the end of the first month.

Cyclic stretch down-regulates calcium transporter gene expression in neonatal rat ventricular myocytes.

Abnormal intracellular Ca2+ handling in hypertrophied and failing hearts is partly due to changes in Ca2+ transporter gene expression, but the mechanisms responsible for these alterations remain largely unknown. We previously showed that intrinsic mechanical load (i.e. spontaneous contractile activity) induced myocyte hypertrophy, and down-regulated SR Ca2+ ATPase (SERCA2) gene expression in cultured neonatal rat ventricular myocytes (NRVM). In the present study, we examined whether extrinsic mechanical load (i.e. cyclic stretch) also induced NRVM hypertrophy, and led to down-regulation of SERCA2 and other Ca2+ transporter genes which have been associated with cardiac hypertrophy and failure in vivo. NRVM were maintained in serum-free culture medium under control conditions, or subjected to cyclic mechanical deformation (1.0 Hz, 20% maximal strain, 48 h). Under these conditions, cyclic stretch induced NRVM hypertrophy, as evidenced by significant increases in total protein/DNA ratio, myosin heavy chain (MHC) content, and atrial natriuretic factor (ANF) secretion. Cyclic stretch also induced the MHC isoenzyme "switch" which is characteristic of hemodynamic overload of the rat heart in vivo. Cyclic stretch significantly down-regulated SERCA2 and ryanodine receptor (RyR) mRNA and protein levels, while simultaneously increasing ANF mRNA. In contrast, Na+-Ca2+ exchanger and phospholamban mRNA levels were unaffected. Load-dependent SERCA2 and RyR down-regulation was independent of Ca2+ influx via voltage-gated, L-type Ca2+ channels, as cyclic stretch down-regulated SERCA2 and RyR mRNA levels in both control and verapamil-treated NRVM. These results indicate that extrinsic mechanical load (in the absence of other exogenous stimuli) induces NRVM hypertrophy and causes down-regulation of Ca2+ transporter gene expression. This in vitro model system should prove useful to dissect the intracellular signaling pathways responsible for transducing this phenotype during cardiac hypertrophy and heart failure in vivo.

Assessment of intra-SR free [Ca] and buffering in rat heart.

To measure the free intrasarcoplasmic reticulum [Ca] ([Ca]SR) in isolated rat cardiac microsomes, ventricular tissue was homogenized in the presence of the low-affinity Ca indicator furaptra. Stepwise increases in cuvette [Ca] ([Ca]c) in the presence of ATP caused progressive increases in steady-state intravesicular fluorescence ratio to a maximum (Rmax). Steady-state [Ca]SR/[Ca]c was approximately 7000. Therefore the resting [Ca]SR may approach 700 microM in the rat cardiac myocyte at [Ca]c = 100 nM. The sarcoplasmic reticulum (SR) Ca pump requires a free energy of deltaG approximately 44 kJ x mol(-1) to generate this [Ca] gradient (e.g., approximately 74% of deltaG(ATP)). Total SR 45Ca uptake was also measured in digitonin-permeabilized myocytes as a function of [Ca]c in the absence of precipitating ions. The steady-state SR Ca content at 100 nM [Ca]c was approximately 400 micromol/liter cytosolic volume. Used together, these data allowed evaluation of the in situ SR Ca-buffering properties. The SR Ca-binding site concentration was approximately 14 mM, and Kd(Ca) approximately 0.638 mM [Ca]SR.

Downregulation of sarcoplasmic reticulum Ca(2+)-ATPase during progression of left ventricular hypertrophy.

To determine whether reduced sarcoplasmic reticulum (SR) Ca(2+)-adenosinetriphosphatase (ATPase) (SERCA2) activity contributes to delayed myocardial relaxation during chronic left ventricular hypertrophy (LVH) progression, LVH was produced in rats by abdominal aortic coarctation. Systolic and diastolic functions were assessed in vivo 8 and 16 wk after surgery, and compositional alterations in LV myocardium [SERCA2 concentration, myosin heavy chain (MHC) isoenzymes, and tissue collagen] were correlated with the development of prolonged isovolumic relaxation and impaired cardiac performance over time. Myocardial relaxation was prolonged in 8-wk banded rats, despite normal isovolumic systolic function and LV end-diastolic pressure (LVEDP). No significant alterations in SERCA2 protein, beta-MHC, or fibrillar collagen levels were observed at this early time point. In contrast, LV SERCA2, beta-MHC, and fibrillar collagen concentrations were all significantly altered in 16-wk banded rats. These late compositional changes were associated with reduced cardiac performance, as manifested by a significant elevation in LVEDP (14 +/- 2 mmHg). The 34% decrease in SERCA2 protein was associated with reduced SR Ca2+ uptake and an even greater reduction (76%) in SERCA2 mRNA. SERCA2 mRNA levels were also significantly reduced to 43 +/- 10% of sham-operated rats 8 wk after banding, despite unchanged SERCA2 protein levels and normal SR Ca2+ uptake. These results argue against a significant contribution of SERCA2 downregulation to the subtle alterations in myocardial relaxation observed in compensated LVH. However, the early reduction in SERCA2 mRNA levels may serve as a molecular marker for impaired cardiac performance during the transition from compensated LVH to heart failure.

Effects of FK-506 on contraction and Ca2+ transients in rat cardiac myocytes.

FK-506 binding protein (FKBP) has been reported to be closely associated with the ryanodine receptor in skeletal and cardiac muscle and to modulate sarcoplasmic reticulum (SR) Ca2+ release channel gating in isolated channels. FK-506 can inhibit the activity of FKBP, thereby reversing its effects on SR Ca2+ release. We investigated the function of FKBP during normal contractions and Ca2+ transients in intact rat ventricular myocytes loaded with fluorescent Ca2+ indicators. FK-506 significantly increased steady state twitch Ca2+ transients and contraction amplitudes even under conditions in which the SR Ca2+ load and Ca2+ current were unaltered, suggesting that FK-506 increases the fraction of SR Ca2+ released during excitation-contraction (E-C) coupling. Action potentials were somewhat prolonged, consistent with the larger Ca2+ transients causing greater inward Na(+)-Ca2+ exchange current. FK-506 did not affect SR Ca2+ uptake but modestly decreased Ca2+ extrusion via Na(+)-Ca2+ exchange in intact cells (although no effect on Na(+)-Ca2+ exchange was seen in sarcolemmal vesicles). In most cells, FK-506 caused an increase in SR Ca2+ content during steady state stimulation, as assessed by caffeine-induced contractures. This was probably due to the inhibition of Ca2+ efflux via Na(+)-Ca2+ exchange. FK-506 also accelerated the rest decay of SR Ca2+ content and increased the frequency of resting Ca2+ sparks about fourfold. The increase in frequency of these basic Ca2+ release events was not associated with changes in the amplitude or duration of the Ca2+ sparks. We conclude that FK-506 increases the fraction of SR Ca2+ released during normal twitches and enhances the rate of SR Ca2+ release during rest. FK-506 also inhibits Na(+)-Ca2+ exchange, although this effect may be indirect. These effects are consistent with an important SR-stabilizing effect of FKBP in intact rat ventricular myocytes.

Interaction of cardiac Na-Ca exchanger and exchange inhibitory peptide with membrane phospholipids.

We tested the hypothesis that the exchange inhibitory peptide (XIP) domain in the cardiac Na-Ca exchanger is a regulatory site under the control of the membrane lipid environment. We found that 125I-XIP bound to liposomes composed of phosphatidylcholine (PC) and phosphatidylserine (PS) with peak binding at 1:1 PC/PS. No binding was observed in PC liposomes. XIP and pentalysine-inhibitable bovine sarcolemmal (SL) Na-Ca exchange activity was observed in reconstituted proteoliposomes composed of 1:1 PC/PS. Proteolysis of SL membranes resulted in a twofold stimulation of Na-Ca exchange activity, but the half-maximal inhibitory concentration (IC50) for XIP (3 microM) was not significantly changed, suggesting that the XIP binding site remained intact. In contrast, the IC50 for pentalysine was decreased from 500 to 150 microM in proteolyzed membranes. These data are consistent with a model of Na-Ca exchange regulation in which the endogenous XIP domain interacts either with another region of the exchange protein to induce an inactive conformational state or with membrane lipid to produce an active conformation.

Identification of a 65 kDa endothelin receptor in bovine cardiac sarcolemmal vesicles.

Endothelin-1, an endothelial cell-derived vasoconstrictor peptide, also exerts a potent positive inotropic effect on cardiac tissue. Characterization of specific binding of endothelin-1 to bovine cardiac sarcolemmal vesicles is reported. In the presence of 1 mM CaCl2, the observed binding for 125I-endothelin-1 had a Kd of 6.2 nM with an observed Bmax of 14 pmol/mg sarcolemmal protein. In the presence of 1 mM EDTA (and no added Ca2+) Bmax was reduced to 9 pmol/mg sarcolemmal protein while the Kd remained unchanged. Binding affinity for sarafotoxin S6b was at least one order of magnitude less than for endothelin-1. 125I-Endothelin-1 covalently cross-linked to a sarcolemmal protein with an apparent molecular weight of 65 kDa. Site-directed polyclonal antibodies to a sequence located on the third extramembranal segment of a previously cloned endothelin ETA receptor from bovine lung were produced. Using Western blot analysis, the site-directed polyclonal antibody recognized a sarcolemmal protein at 65 kDa. We conclude that sarcolemmal membranes from bovine ventricular myocardium contain an endothelin binding site and that it is a protein with an apparent molecular weight of 65 kDa.