SERCA2a activity correlates with the force-frequency relationship in human myocardium

¹ Laboratory of Muscle Research and Molecular Cardiology, Klinik III für Innere Medizin; ² Department of Cardiac and Thoracic Surgery; and ³ Department of Anatomy, Universität zu Köln, 50924 Cologne, Germany

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present investigation addresses whether protein expression and function of sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA2a) and phospholamban (PLB) correlate in failing and nonfailing human myocardium. SERCA2a activity and protein expression, PLB phosphorylation, and the force-frequency relationship (FFR) have been determined in right atrium (RA) and left ventricle (LV) from nonfailing (NF, n = 12) and terminally failing [dilated cardiomyopathy (DCM), n = 12] human hearts. Only in LV of DCM hearts was SERCA2a activity significantly decreased [maximal turnover rate (V_max) = 196 ± 11 and 396 ± 30 nmol · mg¹ · min¹ in LV and RA, respectively], whereas protein expression of SERCA2a in the different chambers was unchanged in NF (3.9 ± 0.3 and 3.2 ± 0.4 densitometric units in LV and RA, respectively) and DCM hearts (4.8 ± 0.8 and 3.4 ± 0.1 densitometric units in LV and RA, respectively). Phosphorylation of PLB was higher in LV than in RA in NF (Ser¹⁶: 180.5 ± 19.0 vs. 56.8 ± 6.0 densitometric units; Thr¹⁷: 174.6 ± 11.2 vs. 37.4 ± 8.9 densitometric units) and DCM hearts (Ser¹⁶: 132.0 ± 5.4 vs. 22.4 ± 3.5 densitometric units; Thr¹⁷: 131.2 ± 10.9 vs. 9.2 ± 2.4 densitometric units). SERCA2a function, but not protein expression, correlated well with the functional parameters of the FFR in DCM and NF human hearts. Regulation of SERCA2a function depends on the phosphorylation of PLB at Ser¹⁶ and Thr¹⁷. However, direct SERCA2a regulation might also be affected by an unknown mechanism.

heart failure; phospholamban phosphorylation; sarco(endo) plasmic reticulum calcium-adenosine 5'-triphosphatase

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE NEGATIVE FORCE-FREQUENCY relationship (FFR) in human heart failure has been suggested to reflect the disturbances of intracellular Ca²⁺ handling in terminally failing human myocardium (6, 19, 26, 28). A prolonged Ca²⁺ transient and elevated diastolic intracellular Ca²⁺ have been described in failing human myocardium (2, 7). These alterations are in accordance with a decreased Ca²⁺ uptake (6, 14, 27) and reduced sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA2a) activity in the myocardium from failing hearts (27). The structural basis for this reduced SERCA2a activity is not yet resolved. Downregulation of protein expression of SERCA2a and its regulatory protein phospholamban in myocardium from patients with terminal heart failure seems not to be the case (15, 18, 27), although a small reduction of protein expression was also reported (9, 17). SERCA2a activity is under inhibitory control of phospholamban (10, 12), and phosphorylation of phospholamban relieves the inhibitory action on SERCA2a activity (29). Therefore, lower phosphorylation could be the key problem for decreased SERCA2a activity in human heart failure. Recently, Schwinger et al. (23) showed blunted protein kinase A (PKA)-dependent phosphorylation at the Ser¹⁶ site of phospholamban in failing human cardiac preparations, which contributes to the decreased Ca²⁺ sensitivity of SERCA2a. Although alterations of Ca²⁺ handling in the failing myocardium are not homogenous in different chambers of the heart, investigations are widely restricted to the left ventricle (LV). Few reports on Ca²⁺ homeostasis and FFR in the atria from failing hearts are available. Besides the negative FFR in the LV of terminally failing hearts, a positive FFR in atria has been reported (25).

Thus we investigated the function and protein expression of SERCA2a as well as the phosphorylation status of phospholamban in ventricular and atrial myocardium to analyze the regulation of protein expression and function in human heart failure in atria and ventricles. Moreover, we addressed whether SERCA2a activity, protein expression of SERCA2a and phospholamban, or phosphorylation status correlates with FFR in normal and diseased human myocardium in ventricular and atrial myocardium.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Myocardium from terminally failing human hearts (LV myocardium, n = 12) was obtained from patients with terminal heart failure (New York Heart Association stage IV) during cardiac transplantation (6 men and 6 women, 45 ± 14 yr old, 28 ± 5% ejection fraction) due to idiopathic dilated cardiomyopathy (DCM). The preoperative diagnosis was DCM in all patients. The hemodynamic data of the patients with heart failure were assessed before transplantation. For DCM patients, right atrial (RA) mean pressure was 9 ± 6 mmHg, right ventricular systolic, diastolic, and mean pressures were 50 ± 12, 8 ± 4, and 13 ± 5 mmHg, respectively, mean pulmonary arterial pressure was 35 ± 7 mmHg, mean pulmonary artery wedge pressure was 23 ± 10 mmHg, and LV end-diastolic pressure was 20 ± 4 mmHg. End-diastolic volume in the LV was 314 ± 19 ml, and cardiac output was 2.4 ± 1.1 l/min. Echocardiographic data showed increased diameters of the RA (49 ± 5 mm) and LV (69 ± 13 mm). Patients were on regular medication with vasodilators (nitrates and angiotensin-converting enzyme inhibitors), cardiac glycosides, and diuretics. None of the patients had received a Ca²⁺ antagonist within 7 days of surgery or -adrenoceptor agonists within 48 h of surgery. Patients gave written informed consent before the operation. As controls, we used nonfailing (NF) hearts from brain-dead organ donors (n = 12, 6 men and 6 women, 43 ± 11 yr old) where transplantation of the heart was not possible because of suspected chest trauma after resuscitation under emergency conditions, positive cytomegalovirus titers, or technical reasons. Absence of heart disease was verified by normal electrocardiogram, absence of cardiovascular medication, and normal pump performance by echocardiography, and the consultant cardiologist recommended the heart acceptable for transplantation. Echocardiographic data showed normal diameters of the RA (39 ± 3 mm) and LV (48 ± 7 mm). Hemodynamic data were available in most cases of the NF organ donors: mean RA pressure was 7 ± 2 mmHg, right ventricular systolic, diastolic, and mean pressures were 26 ± 8, 4 ± 2, and 8 ± 2 mmHg, respectively, mean pulmonary arterial pressure was 17 ± 4 mmHg, mean pulmonary artery wedge pressure was 14 ± 3 mmHg, and cardiac output was 5.9 ± 1.0 l/min.

Muscle Strip Preparation

After explantation, hearts were kept in cardioplegic Brettschneider solution (in mmol/l: 15 NaCl, 10 KCl, 4 MgCl₂, 180 histidine hydrochloride, 2 tryptophan, 30 mannitol, and 1 potassium dihydrogen oxoglutarate) at 4°C during transportation between the cardiac surgery operation room and the research laboratory. On arrival in the laboratory, hearts were dissected in chilled Brettschneider solution, and part of the tissue was immediately shock frozen in liquid nitrogen and stored at 80°C until further use. Muscle strips were placed in 75-ml organ baths containing gassed (95% O₂-5% CO₂) Tyrode solution (in mmol/l: 119.8 NaCl, 5.4 KCl, 1.05 MgCl₂, 1.8 CaCl₂, 22.6 NaHCO₃, 0.42 NaH₂PO₄, 5.05 glucose, 0.28 ascorbic acid, and 0.05 Na₂-EDTA) at 37°C. Stimulation was applied with two platinum electrodes (with variable frequency, 5-ms pulse duration, 5 V) with a Grass S88 stimulator. After 45 min of equilibration, the isometric force of contraction was measured with a force transducer (W. Fleck and FMI) and recorded with a Gould recorder. The experimental protocol was carried out according to the method described previously in detail (26).

Tissue Preparation

For Western blot and SERCA2a activity measurement, homogenization and crude membrane preparation were performed as previously described (20). Briefly, 0.5-1 g of myocardium from the free LV wall or the RA was powdered in liquid nitrogen and slowly thawed on ice in three volumes of chilled preparation buffer (in mmol/l: 300 sucrose, 1 phenylmethylsulfonyl fluoride, 20 PIPES, 10 EDTA, and 50 NaH₂PO₄, pH 7.4), as previously described (20). Care was taken to dissect myocardial tissue from connective tissue, vessels, and epicardium. Samples were minced with three 30-s strokes with an Ultraturrax (Janke und Kunkel, IKA-Werke, Staufen, Germany) after homogenization with a glass-Teflon potter (Braun Melsungen, Melsungen, Germany) at constant temperature of 4°C. The resulting homogenate was further diluted with the same volumes of a storage buffer containing (in mmol/l) 400 sucrose, 5 HEPES, 5 Tris, 10 EDTA, and 50 NaH₂PO₄, pH 7.2, frozen in liquid nitrogen, and stored at 80°C until use in Western blot experiments. After homogenization, crude membranes (U3) were purified for SERCA2a activity assay with one 8,000-g centrifugation step (model JA 20, Beckman Instruments, München, Germany) for 20 min. Supernatants were filtered through four layers of gauze and centrifuged for 60 min at 100,000 g (35,000 rpm; Sorvall A 641, DuPont Nemours, Bad Nauheim, Germany), as previously described (20). Finally, crude membranes were resuspended in the storage buffer as described above, frozen in liquid nitrogen, and stored at 80°C until use. Protein content was measured according to Bradford's assay (3).

SERCA2a Activity

SERCA2a activity was determined with an enzyme-coupled assay according to the method of Chu et al. (4). The reaction was catalyzed by SERCA2a, in combination with pyruvate kinase and lactate dehydrogenase in the presence of Ca²⁺ ionophore A-23187 (1 µmol/l). The oxidation of NADH was monitored continuously by the decrease in absorbance at 340 nm with a spectrophotometer (model DU 640, Beckman). Measurement of reduction of light absorption by NADH at 340 nm was considered equivalent to the hydrolysis of ATP by SERCA2a activity. Membranes were incubated in buffer containing 21 mmol/l MOPS, 4.9 mmol/l NaN₃, 0.06 mmol/l EGTA, 100 mmol/l KCl, 3 mmol/l MgCl₂, 0.2 mmol/l NADH, 1 mmol/l phosphoenolpyruvate, 8.4 units of pyruvate kinase, and 12 units of lactate dehydrogenase. The reaction was started with ATP (1 mmol/l) and was performed at 37°C. SERCA2a activity was calculated from the change of absorption at 340 nm divided by the extinction coefficient of NADH (6.22 cm² · µmol¹ · mg protein¹ · min¹). Basal activity was measured in Ca²⁺-free buffer in the presence of EGTA (4 mmol/l) and then in ascending Ca²⁺ concentrations. SERCA2a maximal turnover rate (V_max) was calculated from nonlinear regression analysis of the single Ca²⁺-SERCA2a activity curves with Prism software.

Western Blot Analysis

For the detection of proteins, we performed immunoblots in homogenates according to Towbin et al. (30) with modifications as previously described (20). Preparations were thawed on ice, suspended in buffer (0.5 mmol/l Tris · HCl, 10% glycerol, 2% SDS, 5% 2-mercaptoethanol, and 0.05% bromphenol blue) without boiling of membrane preparations. Protein separation was performed in a discontinuous SDS-polyarcylamide gel with 4% stacking gel and 12% separation gel (SE 600, Hoefer Scientific Instruments, San Francisco, CA) under constant current of 60 and 100 mA. Protein transfer to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories, Hercules, CA) was achieved with 80-mA constant current at 4°C overnight with a Hoefer Transphor unit. Transfer efficiency was verified by total protein staining of the gels with Coomassie blue. Membranes were blocked in Tris-buffered saline containing 5% milk. Antibody binding on the PVDF membrane with specific antibodies was performed overnight at 4°C. After thorough washing procedures, the membrane was exposed to the secondary antibody, a peroxidase-conjugated anti-mouse IgG or anti-rabbit monoclonal antibody, for 2 h at room temperature. For detection of antibody binding, an enhanced chemiluminescence assay (ECL kit, Amersham Life Science, Buckinghamshire, UK) was used. PVDF membrane was exposed to an X-ray film (Amersham) according to the signal. Quantification of protein expression was achieved after films were scanned into a personal computer, and densitometric volume of bands was analyzed with a commercially available computer program (IQant).

Statistical Analysis

Values are means ± SE of 12 DCM and 12 NF hearts. SERCA2a activity (in nmol ATP hydrolyzed · mg protein¹ · min¹) is given as the mean of 12 NF and 12 DCM hearts. Densitometric units were normalized to protein concentration applied per slot and expressed as densitometric units per microgram of protein. Protein expression of SERCA2a and phospholamban was also normalized to calsequestrin expression as an internal protein standard and expressed as SERCA2a-to-phospholamban or phospholamban-to-calsequestirn ratio. For optimal frequency and maximal tension development in the FFR, the respective frequency or maximal tension as change in maximal developed tension (mN) was determined by manual analysis of the FFR curves. Correlation between optimal stimulation frequency and maximal force development was determined by linear regression analysis with Prism software. Differences between groups were compared with a two-way ANOVA, and P < 0.05 was considered significant.

Materials

Salts were high grade and were purchased from Merck (Darmstadt, Germany); PIPES, MOPS (99.5%), phenylmethylsulfonyl fluoride, and Ca²⁺ ionophore A-23187 were obtained from Sigma Chemical (St. Louis, MO) and 30% acrylamide/bisacrylamide from Bio-Rad. Primary antibodies against SERCA2a were mouse monoclonal IgG1 antibodies against canine SERCA2a protein (ABR, Affinity Bioreagents, Golden, CO). Antibodies against phospholamban were mouse monoclonal IgG antibodies against canine phospholamban (Upstate Biotechnology, Lake Placid, NY). Polyclonal rabbit antibodies against canine calsequestrin were acquired from SWant (Bellizona, Switzerland). Phosphoantibodies against Ser¹⁶ and Thr¹⁷ phospholamban (PhosphoProtein Research, West Yorkshire, UK) are polyclonal mouse anti-canine IgG antibodies, which selectively bind to the phosphorylated form only and specifically distinguish the amino acids Ser¹⁶ and Thr¹⁷, respectively (5). Secondary antibodies were peroxidase-conjugated monoclonal sheep anti-mouse IgG and peroxidase-conjugated goat anti-rabbit IgG (Sigma Immunochemicals, St. Louis, MO).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The total protein yield for crude membrane preparations was 1.59 ± 0.23 and 1.57 ± 0.27 mg/g tissue wet wt in NF and DCM hearts, respectively. There was no apparent difference in the protein yield from NF and DCM hearts. The protein yield for homogenates was also similar in NF and DCM preparations (126.3 ± 7.3 and 121.5 ± 11.7 mg/g tissue for NF and DCM, respectively).

Comparison Between RA and LV

SERCA2a activity. In NF control hearts, no difference in SERCA2a V_max was found between preparations from LV and RA (Fig. 1; 253 ± 10 and 286 ± 19 nmol ATP · mg protein¹ · min¹ for LV and RA, respectively). In DCM hearts, SERCA2a V_max was significantly (P < 0.0001) increased in preparations from the RA in comparison to the LV (196 ± 11 and 396 ± 30 nmol ATP · mg protein¹ · min¹ for LV and RA, respectively; Fig. 1).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA2a) activity in human myocardium as assessed by an enzyme-coupled photometric NADH assay. Maximal turnover rate (Vmax) of SERCA2a is demonstrated in left ventricular (LV) and right atrial (RA) myocardium from nonfailing (NF) controls and patients with severe heart failure (New York Heart Association stage IV) due to idiopathic dilated cardiomyopathy (DCM). Values are means ± SE of 12 experiments. * Significantly different (P < 0.05) from NF; # significantly different (P < 0.05) from LV.

Immunoblots. The chemiluminescence signal for the binding of specific antibodies to SERCA2a and phospholamban, depending on the protein application, was tested from 5 to 100 µg of protein per slot. Over a wide range of protein applied, the densitometric signal was linear, with a regression coefficient between 0.92 and 0.98, as previously shown (20).

Ten micrograms of protein were applied in the assay to study protein expression of SERCA2a and phospholamban in NF and DCM hearts to compare RA and LV protein expression. SERCA2a expression normalized to protein was similar in LV and RA of NF (19 ± 2 and 21 ± 3 densitometric units in LV and RA, respectively) and DCM myocardium (20 ± 2 and 20 ± 2 densitometric units in LV and RA, respectively; Fig. 2). Phospholamban expression normalized to protein was significantly higher in LV than in RA of NF hearts (16 ± 2 vs. 12 ± 3 densitometric units) and slightly lower in LV than in RA from DCM tissue (18 ± 3 vs. 23 ± 5 densitometric units; Fig. 2).

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Representative Western blots for SERCA2a, calsequestrin, and phospholamban protein expression in human myocardial tissue. Myocardial preparations (10 µg) of NF controls from LV and RA (NFLV and NFRA) and from DCM patients undergoing cardiac transplantation (DCMLV and DCMRA) are shown.

When protein expression of SERCA2a was normalized to calsequestrin protein as an internal standard for overall protein expression, SERCA2a expression was also unchanged in the LV compared with the RA in NF (3.9 ± 0.3 and 3.2 ± 0.4 densitometric units in LV and RA, respectively) and DCM hearts (4.8 ± 0.8 and 3.4 ± 0.1 densitometric units in LV and RA, respectively). In contrast, in NF hearts, phospholamban expression normalized to calsequestrin was significantly higher in LV than in RA (3.0 ± 0.5 vs. 1.7 ± 0.3 densitometric units), whereas in DCM tissue, no statistical difference was apparent (4.5 ± 1.3 and 3.7 ± 0.6 densitometric units in LV and RA, respectively).

Normalization of SERCA2a to phospholamban resulted in lower ratios in LV than in RA from NF hearts (1.3 ± 0.2 vs. 1.8 ± 0.2). In DCM tissue, no apparent difference in the SERCA2a-to-phospholamban ratio between LV and RA was found (1.35 ± 0.30 and 1.11 ± 0.06 in LV and RA, respectively).

There was marked reduction of the phosphorylation at Ser¹⁶ of phospholamban in RA compared with LV in the NF and the DCM groups (Fig. 3). Phosphorylation at Thr¹⁷ of phospholamban was also lower in RA than in LV of NF (174.6 ± 11.2 vs. 37.4 ± 8.9) and DCM hearts (131.2 ± 10.9 vs. 9.2 ± 2.4; Fig. 3).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Phosphorylation status of regulatory protein phospholamban at specific amino acid Ser16 and Thr17 in human myocardium as assessed by Western blot and site-specific phosphoantibodies. Phosphorylation status in LV and RA from NF controls and patients with DCM was investigated; 10 µg of protein from NF and DCM samples were applied per slot. Values are means ± SE of 12 NF and 12 DCM samples. * Significantly different (P < 0.05) from NF; # significantly different (P < 0.05) from LV.

The ratio of phosphorylated phospholamban at Ser¹⁶ was significantly higher in LV than in RA of the NF (18.0 ± 3.2 vs. 6.1 ± 0.7) and DCM groups (14.8 ± 2.8 vs. 3.9 ± 0.4). The same is also true for Thr¹⁷ phosphorylation. A higher fraction of phosphorylated phospholamban in LV than in RA was found at Thr¹⁷ in the NF (10.0 ± 2.3 vs. 3.4 ± 0.9) and DCM groups (8.3 ± 1.6 vs. 0.6 ± 0.2).

FFR. In papillary muscle strip preparations from LV of NF control hearts, a positive FFR from 0.5 to 3.0 Hz frequency, with increase of force generation from 1.4 ± 0.2 to 2.5 ± 0.3 mN, was found, in accordance with previous reports. In NF control RA, a positive FFR was evident, with a transient increase of the developed force of contraction from 0.5 to 1.5 Hz from 2.5 ± 0.4 to 3.7 ± 0.4 mN. At 3.0 Hz, a small decline to 2.7 ± 0.4 was apparent in NF RA (Figs. 4 and 5). The time course of the contraction changed in NF LV with increasing stimulation frequencies. The contraction (+T) and relaxation (T) velocities increased gradually over the range of stimulation (+T: 11.6 ± 1.0 and 34.8 ± 3.0 mN/s at 0.5 and 3.0 Hz, respectively; T: 8.5 ± 1.2 and 24.5 ± 3.1 mN/s at 0.5 and 3.0 Hz, respectively). In RA from NF hearts, the velocity of contraction and relaxation increased to a peak value at 2.0 Hz and slightly declined at 3.0 Hz (+T: 32.1 ± 7.5 and 55.6 ± 7.4 mN at 0.5 and 2.0 Hz, respectively; T: 21.1 ± 5.1 and 36.0 ± 5.3 mN at 0.5 and 2.0 Hz, respectively). When the contraction and relaxation parameters were compared in both chambers of the NF group, FFR gradually increased with increasing stimulation parameters. However, contraction and relaxation were significantly faster in RA than in LV.

View larger version (78K): [in this window] [in a new window]   Fig. 4.   Representative original twitch recordings from human myocardium after increase of stimulation frequency from 0.5 to 3 Hz. Slow and fast recordings are demonstrated for RA from explanted hearts from NF organ donors (RA NF) and explanted hearts from patients undergoing cardiac transplantation due to terminal heart failure (RA DCM) and from LV from explanted hearts of NF organ donors (LV NF) and from patients with terminal heart failure due to DCM undergoing cardiac transplantation (LV DCM). Time course and force generation are indicated by horizontal and vertical bars, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Force recordings from isolated muscle strips from RA and LV from explanted human hearts. Values are means ± SE from 12 nonfailing organ donors where cardiac transplantation could not be performed (A) and 12 patients with terminal heart failure undergoing cardiac transplantation (B).

In contrast, LV muscle preparations from DCM hearts developed a negative FFR from 0.5 to 3.0 Hz frequency. The force of contraction declined from 1.5 ± 0.2 mN at 0.5 Hz to 0.5 ± 0.1 mN at 3.0 Hz. In isolated muscle strips from the RA from DCM hearts, the force of contraction increased steadily from 1.2 ± 0.2 mN at 0.5 Hz over the whole range of frequencies to 2.5 ± 0.5 mN at 3.0 Hz (Figs. 4 and 5). The contraction and relaxation velocities in LV myocardium from DCM hearts did not increase with increasing stimulation frequencies (+T: 10.8 ± 1.7 and 11.0 ± 2.5 mN/s at 0.5 and 3.0 Hz, respectively) or even decreased (T: 8.2 ± 2.7 and 5.9 ± 1.8 mN/s at 0.5 and 3.0 Hz, respectively). In contrast, RA contraction parameters in DCM hearts increased gradually over the range of stimulation frequencies (+T: 21.3 ± 0.8 and 57.3 ± 7.5 mN/s at 0.5 and 3.0 Hz, respectively; T: 10.6 ± 0.4 and 41.6 ± 5.2 mN/s at 0.5 and 3.0 Hz, respectively). When contraction and relaxation velocities in the DCM group were compared, RA contracted significantly faster than LV in DCM hearts, and contraction and relaxation gradually increased in RA and decreased in LV from DCM hearts.

Comparison Between NF and DCM Hearts

SERCA2a activity. SERCA2a V_max in preparations from LV was significantly (P < 0.001) reduced in DCM tissue preparations compared with NF control tissue (253 ± 10 vs. 196 ± 11 nmol ATP · mg protein¹ · min¹). In contrast, V_max of SERCA2a was significantly increased (P < 0.05) in RA from DCM preparations compared with RA from NF controls (286 ± 19 vs. 396 ± 30 nmol ATP · mg protein¹ · min¹; Fig. 1).

Immunoblots. In a comparison of SERCA2a protein expression in NF and DCM preparations, no significant differences were found in LV (19 ± 2 and 20 ± 2 densitometric units in NF and DCM, respectively) or RA (21 ± 3 and 20 ± 2 densitometric units in NF and DCM, respectively; Fig. 2). When phospholamban protein expression was compared in NF and DCM hearts, protein expression was unchanged in LV myocardium (16 ± 2 and 18 ± 3 densitometric units in NF and DCM, respectively), but phospholamban expression was significantly increased in DCM RA compared with NF RA (12 ± 3 vs. 23 ± 5 densitometric units). Calsequestrin protein expression was unchanged in NF compared with DCM preparations in LV and RA myocardium (5 ± 1 and 7 ± 1 densitometric units in NF LV and RA, respectively, and 4 ± 1 and 6 ± 1 densitometric units in DCM LV and RA, respectively). When immunoblots were normalized to calsequestrin as internal standard, comparison of NF and DCM preparations showed unchanged SERCA2a expression normalized to calsequestrin in LV (3.9 ± 0.3 and 4.8 ± 0.8 densitometric units in NF and DCM, respectively) and RA (3.2 ± 0.4 and 3.4 ± 0.1 densitometric units in NF and DCM, respectively). Comparison of NF and DCM hearts showed unchanged phospholamban expression normalized to calsequestrin in LV (3.0 ± 0.5 and 4.5 ± 1.3 densitometric units in NF and DCM, respectively) and RA (1.7 ± 0.3 and 3.7 ± 0.6 densitometric units in NF and DCM, respectively). Comparison of the NF and DCM groups showed that the differences in SERCA2a-to-phospholamban ratios did not reach significant levels in RA or LV.

Phosphorylation of phospholamban at Ser¹⁶ was lower in DCM than in NF hearts in LV (180.5 ± 19.0 vs. 132.0 ± 5.4) and RA (56.8 ± 6.0 vs. 22.4 ± 3.5; Fig. 3). Phosphorylation at Thr¹⁷ of phospholamban was also lower in both RA and LV in DCM than in NF hearts (174.6 ± 11.2 and 131.2 ± 10.9 in LV in NF and DCM, respectively, and 37.4 ± 8.9 and 9.2 ± 2.4 in RA in NF and DCM, respectively; Fig. 3). Moreover, the ratio of phosphorylated phospholamban at Ser¹⁶ was significantly lower in the DCM group in the LV (18.0 ± 3.2 and 14.8 ± 3.9 in NF and DCM, respectively) and RA (6.2 ± 0.7 and 3.9 ± 0.4 in NF and DCM, respectively).

FFR. Comparison of the FFR in NF RA (2.5 ± 0.4 and 3.7 ± 0.4 mN at 0.5 and 1.5 Hz, respectively) and DCM RA (1.2 ± 0.2 and 2.5 ± 0.5 mN at 0.5 and 3.0 Hz, respectively) showed increased force of contraction with increasing stimulation frequencies in both groups. In contrast, comparison of LV in NF and DCM hearts showed a positive FFR in NF LV (1.4 ± 0.2 and 2.5 ± 0.3 mN at 0.5 and 3.0 Hz, respectively) and a negative FFR in DCM LV (1.5 ± 0.2 and 0.5 ± 0.1 mN at 0.5 and 3.0 Hz, respectively; Figs. 4 and 5). Comparison of the contraction cycle velocities in NF RA vs. DCM RA showed no statistical differences in contraction (55.6 ± 18.1 and 57.0 ± 7.5 mN/s in NF and DCM, respectively) or relaxation (36.0 ± 13.1 and 41.6 ± 5.2 mN/s in NF and DCM, respectively). The cycle data in the LV showed a marked decrease in contraction (34.8 ± 3.0 and 11.0 ± 2.5 mN/s in NF and DCM, respectively) and relaxation (24.5 ± 3.1 and 8.2 ± 2.7 mN/s in NF and DCM, respectively) in NF vs. DCM preparations.

The optimal stimulation frequency for the force development was significantly higher in DCM than in NF control atria (1.5 ± 0.8 and 3.0 ± 0.4 Hz in NF and DCM RA, respectively). In LV the optimal stimulation frequency for developed tension was significantly higher in NF than in DCM muscle strips (2.5 ± 0.7 vs. 0.7 ± 0.9 Hz). Moreover, maximal tension was unchanged in NF RA compared with DCM RA (0.97 ± 0.28 and 1.3 ± 0.02 mN in NF and DCM RA, respectively). In contrast, in DCM LV, maximal tension was significantly reduced compared with NF LV (1.26 ± 0.25 vs. 0.02 ± 0.01 mN). Good correlation was found between FFR parameters and SERCA2a activity. When the NF and DCM groups are taken together, the correlation between optimal stimulation frequency and SERCA2a V_max was 0.47 in LV (P = 0.0004) and 0.62 in RA (P < 0.0001). The correlation between maximal tension development and SERCA2a V_max was 0.61 in LV (P < 0.0001) and 0.38 in RA (P = 0.0021) from the NF and DCM groups taken together (Fig. 6). Moreover, good correlation between SERCA2a V_max and contraction and relaxation parameters was found (r² = 0.28, P = 0.007 for +T; r² = 0.19, P = 0.03 for T).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Correlation between contractile parameters of force-frequency relation and SERCA2a Vmax in human myocardium. Left: RA myocardium; right: LV myocardium. Top: correlation between SERCA2a Vmax and optimal stimulation frequency in force-frequency relation that elicits maximal tension development in multicellular human myocardial preparations. Correlation coefficient (r2) = 0.62 (P < 0.0001) in RA and 0.47 (P = 0.0004) in LV myocardium. Bottom: correlation between SERCA2a Vmax and maximal force generation in force-frequency relation in multicellular human myocardial preparations; r2 = 0.38 (P = 0.0021) in RA, and r2 = 0.61 (P < 0.0001) in LV.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

SERCA2a Activity and FFR

Comparison of RA and LV. A novel finding of the present study is that the FFR in the LV and RA of failing and nonfailing myocardium is tightly correlated to altered SERCA2a activity in the different chambers of the human heart. The FFR in nonfailing hearts was positive in RA and LV; however, in failing hearts the negative response of the force generation to increasing heart rates in the LV stands in contrast to a positive FFR in RA. Altered FFR parameters correlated with SERCA2a activity, which was increased solely in RA compared with LV of failing hearts but not in RA compared with LV of nonfailing hearts. The original finding of the present study is that different activity of SERCA2a in different heart chambers of the failing heart can be explained, at least partly, by a decrease of the phosphorylation status of the regulatory protein phospholamban.

In both chambers of the nonfailing hearts, a positive FFR was documented, in accordance with previous studies (8, 25, 26). It is generally accepted that the ability of the heart to increase force generation with increasing heart rates is tightly coupled to the removal of Ca²⁺ from the cytoplasm of the myocyte into the SR by SERCA2a (7, 9, 27). The positive FFR was in accordance with comparable SERCA2a activity in RA and LV from nonfailing hearts. However, the maximal increase in force generation was achieved at lower stimulation frequencies in RA than in LV of nonfailing hearts. This difference in the FFR in nonfailing atrial vs. ventricular myocardium might not be explained completely by differences in SERCA2a function. Besides removal of Ca²⁺ from the cytoplasm into the SR, transsarcolemmal Ca²⁺ transport via the Na⁺/Ca²⁺ exchanger might play a role. A prominent role of the Na⁺/Ca²⁺ exchanger in the action potential in human atrial myocardium has been reported in an electrophysiological study (1). Additionally, Ca²⁺ influx into the cell and properties of the contractile system might be of importance for the FFR of the heart. Differences in Ca²⁺ current of the L-type Ca²⁺ channels between the atrial and the ventricular myocardium have been reported (16) and might also contribute to the differences in atrial FFR.

Although hemodynamic differences exist between RA and LV in patients with terminal heart failure, we have evidence of failure from the echocardiographic and hemodynamic data of increased chamber diameters and increased pressures in RA and LV of the failing heart. Nevertheless, besides a negative FFR in LV myocardium from failing hearts, the RA increased force generation with an increase of the stimulation frequency, in accordance with previous findings (8, 25, 26). The FFR parameters were in good agreement with increased SERCA2a function in RA compared with LV from failing hearts. The underlying cause for the increase of SERCA2a function could not be resolved in the present study. Changes in SERCA2a or phospholamban protein expression could not help explain the change in SERCA2a regulation. Phosphorylation of phospholamban relieves the inhibitory action of phospholamban on SERCA2a activity (29). However, phosphorylation of the regulatory protein phospholamban could not explain the altered SERCA2a function, inasmuch as phospholamban phosphorylation was markedly lower in RA than in LV from failing hearts. Possibly direct phosphorylation of SERCA2a (31, 34) might play a role in increased SERCA2a activity and positive FFR in atria from failing hearts. Besides changes in the Ca²⁺ homeostasis of the failing heart, Ca²⁺ sensitivity of myofibrils might also influence the FFR of LV and RA. Decreased Ca²⁺ sensitivity of the contractile apparatus in atria vs. ventricles in failing hearts as a result of different myosin isozyme expression has been reported (32).

Comparison of failing and nonfailing myocardium. It is well documented that muscle strips of LV from failing myocardium develop a negative FFR (11, 19, 24, 26). The present data underline the importance of SERCA2a activity for the contractile parameters of the FFR not only in LV, but also in RA, from failing hearts. A decrease of SERCA2a activity was found in LV from failing hearts in comparison to nonfailing controls in the present study. This again is in accordance with previous findings concerning the decreased Ca²⁺ uptake (9, 14, 27) and reduced function of SERCA2a in failing human ventricles (27). Functional alterations of SERCA2a in failing LV can be explained by different phosphorylation status of phospholamban, leading to an increased inhibition of SERCA2a in failing myocardium (11, 18). Recently, Schwinger et al. (23) showed lower PKA-dependent phosphorylation at Ser¹⁶ of phospholamban in failing human cardiac preparations, which contributes to the decreased Ca²⁺ sensitivity of SERCA2a. The importance of phosphorylation for the regulation of SERCA2a was underlined in the present study. A novel finding is lower phosphorylation at the PKA-dependent Ser¹⁶ site and the calmodulin kinase-dependent Thr¹⁷ site of phospholamban in failing LV myocardium, which explains the reduced SERCA2a in failing LV compared with nonfailing LV.

The importance of SERCA2a activity for FFR could be demonstrated by good correlations between SERCA2a V_max and FFR parameters such as tension development, optimal stimulation frequency, and cycle parameters such as contraction and relaxation velocity. In recent work from our laboratory, it could be demonstrated that, even in nonfailing tissue, force frequency depends on SERCA2a activity (28). Inhibition of SERCA2a by cyclopiazonic acid in nonfailing myocardium from LV papillary muscle strips mimics the negative FFR found in terminally failing human myocardium.

The FFR in RA tissue from failing hearts was positive and, compared with RA from nonfailing hearts, shifted to higher frequencies in accordance with increased SERCA2a activity. Thus the ability of the myocardium to increase the force of contraction at higher heart rates seems to be very closely linked to SERCA2a activity in atria and ventricles, even in diseased myocardium. A previous report emphasizes the importance of the SERCA2a-to-phospholamban ratio for myocardial contractility (13); however, the present data do not support this hypothesis, inasmuch as the highest SERCA2a-to-phospholamban ratio was found in atria from nonfailing hearts, whereas the highest SERCA2a activity was found in atria from failing hearts.

In contrast to the reduction of phospholamban phosphorylation in LV myocardium in patients with DCM, the increase in SERCA2a in RA from patients with DCM cannot be explained by the phosphorylation of phospholamban. However, direct regulation of SERCA2a activity, independent of phospholamban, could also be present. Whether phosphorylation of SERCA2a is direct is under debate and has not been resolved (21, 22, 31, 34). Thus, in RA myocardium, not only the phosphorylation status or the protein expression of phospholamban, but a direct modification of SERCA2a function, might explain differences in SERCA2a regulation. In addition to altered SERCA2a function, different Ca²⁺ sensitivity of the contractile proteins due to different phosphorylation of troponin I additionally affects the FFR in failing human myocardium (33).

Conclusion

SERCA2a function is a major determinant for the FFR in human myocardium. Marked differences in SERCA2a activity and FFR are present in LV and RA from terminally failing human hearts. Changes in SERCA2a in LV from failing hearts can be at least partly explained by lower phosphorylation of phospholamban. Changes in SERCA2a function in atria from failing human hearts are not due to changes in phospholamban protein expression or phosphorylation and, hence, may be indicative of an altered direct SERCA2a regulation.

Limitations of the Study

The alteration in FFR in LV myocardium is a well-documented phenomenon. It is widely accepted that alterations of Ca²⁺ homeostasis of the failing myocardium play a key role in the negative FFR in LV from terminally failing human hearts. However, data on FFR in different chambers of the failing heart, particularly in RA of hearts from patients with DCM, are sparse. The present study confirms previous findings of FFR and SERCA2a in LV of hearts from patients with DCM but gives new insights into the activity and regulation of SERCA2a and phospholamban in RA from terminally failing human hearts. This study gives new evidence that phosphorylation of phospholamban plays an important role in SERCA2a regulation but also raises questions concerning different modulatory aspects of SERCA2a function in human atrial myocardium.

	ACKNOWLEDGEMENTS

The authors are very grateful to Prof. E. R. DeVivie (Dept. of Cardiac Surgery, Universität zu Köln) and Prof. B. Reichart (Dept. of Cardiac Surgery, Universität München) and their colleagues for providing the human cardiac tissue. The authors thank Andrea Herber and Tatjana Schewior for excellent technical assistance.

	FOOTNOTES

The experimental work was supported by the Deutsche Forschungsgemeinschaft (to R. H. G. Schwinger).

This work contains part of the doctoral thesis of H. Reuter and B. Bölck.

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Address for reprint requests and other correspondence: R. H. G. Schwinger, Klinik III für Innere Medizin, Laboratory of Muscle Research and Molecular Cardiology, Universität Köln, Joseph-Stelzmann-Str. 9, 50924 Cologne, Germany (E-mail: robert.schwinger{at}medizin.uni-koeln.de).

Received 2 July 1999; accepted in final form 6 December 1999.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

1.   Benardeau, A, Hatem SN, Rücker-Martin C, Le Grand B, Mace L, Dervanian P, Mercadier J-J, and Coraboeuf E. Contribution of Na⁺/Ca²⁺ exchange to action potential of human atrial myocytes. Am J Physiol Heart Circ Physiol 271: H1151-H1161, 1996[Abstract/Free Full Text].

2.   Beuckelmann, DJ, Nabauer M, and Erdmann E. Intracellular calcium handling in isolated ventricular myocytes from patients with terminal heart failure. Circulation 85: 1046-1055, 1992[Abstract/Free Full Text].

3.   Bradford, MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254, 1976[ISI][Medline].

4.   Chu, A, Dixon MC, Saito A, Seiler S, and Fleischer S. Isolation of sarcoplasmic reticulum fractions referable to longitudinal tubules and junctional terminal cisternae from rabbit skeletal muscle. Methods Enzymol 189: 36-46, 1988.

5.   Drago, GA, and Colyer J. Discrimination between two sites of phosphorylation on adjacent amino acids by phosphorylation site-specific antibodies to phospholamban. J Biol Chem 269: 25073-25077, 1994[Abstract/Free Full Text].

6.   Feldman, MD, Gwathmey JK, Phillips P, Schoen F, and Morgan JP. Reversal of the force-frequency relationship in working myocardium from patients with end-stage heart failure. J Appl Cardiol 3: 273-283, 1988[ISI].

7.   Gwathmey, JK, Copelas L, MacKinnon R, Schoen FJ, Feldman MD, Grossman W, and Morgan JP. Abnormal intracellular calcium handling in myocardium from patients with end-stage heart failure. Circ Res 61: 70-76, 1987[Abstract/Free Full Text].

8.   Gwathmey, JK, Slawsky MT, Hajjar RJ, Briggs GM, and Morgan JP. Role of intracellular calcium handling in force-interval relationships of human ventricular myocardium. J Clin Invest 85: 1599-1613, 1990.

9.   Hasenfuss, G, Reinecke H, Studer R, Meyer M, Pieske B, Holtz J, Holubarsch C, Posival H, Just H, and Drexler H. Relation between myocardial function and expression of sarcoplasmic reticulum Ca²⁺-ATPase in failing and nonfailing human myocardium. Circ Res 75: 434-442, 1994[Abstract/Free Full Text].

10.   James, P, Inui M, Tada M, Chiesi M, and Carafoli E. Nature and site of phospholamban regulation of the Ca²⁺ pump of sarcoplasmic reticulum. Nature 342: 90-92, 1989[Medline].

11.   Kim, HW, Steenaart NA, Ferguson DG, and Kranias EG. Functional reconstitution of the cardiac sarcoplasmic reticulum Ca²⁺-ATPase with phospholamban in phospholipid vesicles. J Biol Chem 265: 1702-1709, 1990[Abstract/Free Full Text].

12.   Kimura, Y, Kurzydlowski K, Tada M, and MacLennan DH. Phospholamban regulates the Ca²⁺-ATPase through intramembrane interactions. J Biol Chem 271: 21726-21731, 1996[Abstract/Free Full Text].

13.   Koss, KL, Grupp IL, and Kranias EG. The relative phospholamban and SERCA2 ratio: a critical determinant of myocardial contractility. Basic Res Cardiol 92 Suppl1: 17-24, 1997.

14.   Limas, CJ, Olivari MT, Goldenberg IF, Levine TB, Benditt DG, and Simon A. Calcium uptake by cardiac sarcoplasmic reticulum in human dilated cardiomyopathy. Cardiovasc Res 21: 601-605, 1987[ISI][Medline].

15.   Linck, B, Boknik P, Eschenhagen T, Müller FU, Neumann J, Nose M, Jones LR, Schmitz W, and Scholz H. Messenger RNA expression and immunological quantification of phospholamban and SR Ca²⁺-ATPase in failing and nonfailing human hearts. Cardiovasc Res 31: 625-632, 1996[ISI][Medline].

16.   Mewes, T, and Ravens U. L-type calcium currents of human myocytes from ventricle of non-failing and failing hearts from atrium. J Mol Cell Cardiol 26: 1307-1320, 1994[ISI][Medline].

17.   Meyer, M, Schillinger W, Pieske B, Holubarsch C, Heilmann C, Posival H, Kuwajima G, Mikoshiba K, Just H, Hasenfuss G, Alterations of sarcoplasmic reticulum proteins in failing human dilated cardiomyopathy. Circulation 92: 778-784, 1995[Abstract/Free Full Text].

18.   Movsesian, MA, Karimi M, Green K, and Jones LR. Ca²⁺-transporting ATPase, phospholamban, and calsequestrin levels in nonfailing and failing human myocardium. Circulation 90: 653-657, 1994[Abstract/Free Full Text].

19.   Mulieri, LA, Hasenfuss G, Leavitt B, Allen PD, and Alpert NR. Altered myocardial force-frequency relation in human heart failure. Circulation 85: 1743-1750, 1992[Abstract/Free Full Text].

20.   Münch, G, Bölck B, Hoischen S, Brixius K, Bloch W, Reuter H, and Schwinger RHG Unchanged protein expression of sarcoplasmic reticulum Ca²⁺-ATPase, phospholamban, and calsequestrin in terminally failing human myocardium. J Mol Med 76: 434-441, 1998[ISI][Medline].

21.   Odermatt, A, Kurzydlowski K, and MacLennan DH. The V_max of the Ca²⁺-ATPase of cardiac sarcoplasmic reticulum (SERCA2a) is not altered by Ca²⁺/calmodulin-dependent phosphorylation or by interaction with phospholamban. J Biol Chem 271: 14206-14213, 1996[Abstract/Free Full Text].

22.   Reddy, LG, Jones LR, Pace RC, and Stokes DL. Purified, reconstituted cardiac Ca²⁺-ATPase is regulated by phospholamban but not by direct phosphorylation with Ca²⁺/calmodulin-dependent protein kinase. J Biol Chem 271: 14964-14970, 1996[Abstract/Free Full Text].

23.   Schwinger, RH, Münch G, Bölck B, Karczewski P, Krause EG, and Erdmann E. Reduced Ca²⁺ sensitivity of SERCA2a in failing human myocardium due to reduced serine-16 phospholamban phosphorylation. J Mol Cell Cardiol 31: 479-491, 1999[ISI][Medline].

24.   Schwinger, RHG, Böhm M, and Erdmann E. Inotropic and lusitropic dysfunction in myocardium from patients with dilated cardiomyopathy. Am Heart J 123: 116-128, 1992[ISI][Medline].

25.   Schwinger, RHG, Böhm M, Koch A, Uhlmann R, Überfuhr P, Kreuzer E, Reichart B, and Erdmann E. Force-frequency relation in human atrial and ventricular myocardium. Mol Cell Biochem 119: 73-78, 1993[ISI][Medline].

26.   Schwinger, RHG, Böhm M, Müller J, Ehmsen Uhlmann R, Schmidt U, Stäblein A, Überfuhr P, Kreuzer E, Reichart B, Eissner HJ, and Erdmann E. Effect of inotropic stimulation on the negative force-frequency relationship in the failing human heart. Circulation 88: 2267-2276, 1993[Abstract/Free Full Text].

27.   Schwinger, RHG, Böhm M, Schmidt U, Karczewski P, Bavendiek U, Flesch M, Krause EG, and Erdmann E. 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].

28.   Schwinger, RHG, Böhm M, Uhlmann R, Schmid U, Überfuhr P, Kreuzer E, Reichart B, and Erdmann E. Magnesium restores the altered force-frequency relationship in failing human myocardium. Am Heart J 126: 1018-1021, 1993[ISI][Medline].

29.   Tada, M, Kirchberger MA, and Katz AM. Phosphorylation of a 22,000-dalton component of the cardiac sarcoplasmic reticulum by adenosine 3',5'-monophosphate-dependent protein kinase. J Biol Chem 250: 2640-2647, 1975[Abstract/Free Full Text].

30.   Towbin, H, Staehelin T, and Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76: 4350-4354, 1979[Abstract/Free Full Text].

31.   Toyofuku, T, Curotto Kurzydlowski K, Narayanan N, and MacLennan DH. Identification of Ser³⁸ as the site in cardiac sarcoplasmic reticulum Ca²⁺-ATPase that is phosphorylated by Ca²⁺/calmodulin-dependent protein kinase. J Biol Chem 269: 26492-26496, 1994[Abstract/Free Full Text].

32.   Wankerl, M, Böhm M, Morano I, Ruegg JC, Eichhorn M, and Erdmann E. Calcium sensitivity and myosin light chain pattern of atrial and ventricular skinned cardiac fibers from patients with various kinds of cardiac disease. J Mol Cell Cardiol 22: 1425-1438, 1990[ISI][Medline].

33.   Wolff, MR, Buck SH, Stoker SW, Greaser ML, and Mentzer RM. Myofibrillar calcium sensitivity of isometric tension is increased in human dilated cardiomyopathies: role of altered -adrenergically mediated protein phosphorylation. J Clin Invest 98: 167-176, 1996[ISI][Medline].

34.   Xu, A, Hawkins C, and Narayanan N. Phosphorylation and activation of the Ca²⁺-pumping ATPase of cardiac sarcoplasmic reticulum by Ca²⁺/calmodulin-dependent protein kinase. J Biol Chem 268: 8394-8397, 1993[Abstract/Free Full Text].