INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

EARLY POSTNATAL DEVELOPMENT of the heart is characterized by increasing contractile performance that reflects substantial changes in myocardial Ca²⁺ handling. During the neonatal period, when the sarcoplasmic reticulum (SR) is poorly developed structurally and functionally, cardiac contraction and relaxation are more dependent on Ca²⁺ fluxes across the sarcolemma (SL) than in adults in most mammalian species (21). There are two major systems responsible for reducing cytosolic Ca²⁺ from a high systolic level to a low resting level during cardiac relaxation: the SL Na⁺/Ca²⁺ exchanger (NCX) and sarcoendoplasmic reticulum Ca²⁺-ATPase (SERCA2). Whereas the activity of NCX is high after birth and progressively decreases during maturation, the rate of SERCA2 transport increases markedly during ontogeny (5, 36, 37). In fact, SR Ca²⁺ handling predominates in cellular Ca²⁺ regulation of excitation-contraction coupling in adults (25). For rabbit and rat myocardium, it has been shown recently (5, 36) that this developmental shift from the SL- to the SR-mediated Ca²⁺ transport is based, at least in part, on reciprocal changes in the expression of NCX and SERCA2. In addition, alterations in membrane lipids (22, 32) as well as levels and states of phosphorylation of the SERCA2 modulatory protein phospholamban (23, 33) could be further contributing factors.

Although common regulatory mechanisms controlling the balance between the long-term activities of the two Ca²⁺ transport proteins in the developing heart are presently unknown, it is noteworthy that the expression of SERCA2 (1, 29) and NCX (4) can be up- and downregulated, respectively, by thyroid hormones. An altered thyroid status, especially hypothyroidism, appears to have more serious consequences during the period of rapid cardiac growth and differentiation than in adults. In the rat, thyroid hormone blood levels increase postnatally and reach peak values in the third postnatal week (38). Thyroid hormones are essential for the normal postnatal redistribution of L-type Ca²⁺ channels from the surface nonjunctional SL into specialized areas of the SL and T tubules associated with junctional structures of the SR that are enriched with ryanodine-sensitive Ca²⁺ release channels (40). Neonatal hypothyroid rats exhibited decreased SR function together with increased dependency on trans-SL Ca²⁺ fluxes as indicated by lower inhibition of contraction by ryanodine and higher sensitivity of the heart to verapamil compared with euthyroid controls; hyperthyroidism was associated with opposite changes (14).

The present study was designed to examine the mechanisms by which thyroid hormones affect the function of NCX and SERCA2 in the developing rat ventricular myocardium. For this purpose, steady-state mRNA and protein levels together with the Ca²⁺ transporting activities of the two systems were measured in euthyroid, hypothyroid, and hyperthyroid 21-day-old animals. Furthermore, the question was addressed as to whether the amount and state of phosphorylation of phospholamban were altered by postnatal hypo- and hyperthyroidism. Our main finding was that physiological thyroid hormone levels appear to be essential for normal reciprocal changes in the expression and function of cardiac NCX and SERCA2 during postnatal development. Moreover, from the comparison of SR Ca²⁺ uptake rates and phospholamban phosphorylation levels, it appears that thyroid hormone-dependent changes in phospholamban content and its state of phosphorylation at the protein kinase A-specific site are major determinants of altered SR Ca²⁺-pump activity in the developing hypo- and hyperthyroid heart.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animal model. Litters of newborn Wistar rats were evened to eight pups per dam and maintained throughout the experiment. We made the pups hypothyroid with the administration of 0.05% 6-n-propyl-2-thiouracil (PTU) in drinking water given to nursing mothers from days 2 to 21 postpartum. Hyperthyroidism was induced by daily subcutaneous injections of 3,3',5-triiodo-L-thyronine (T₃; 10 µg/100 g body wt) during the same period of time. Euthyroid control rats received no treatment. In another group of animals, the PTU-induced hypothyroidism was simultaneously treated with 2.5 µg T₃/100 g body wt. This lower dose of T₃ was selected because 10 µg T₃/100 g body wt given daily caused high mortality in PTU-treated animals.

Isolation of RNA. Total cellular RNA was isolated from separated right or left ventricles according to the single-step protocol of Chomczynski and Sacchi (6) by using TRIzol reagent (GIBCO-BRL, Grand Island, NY), a monophasic solution of phenol and guanidine isothiocyanate. Typically, between 50 and 100 µg total RNA were obtained from 100 mg tissue. RNA samples were stored at 80°C in diethyl pyrocarbonate-treated water. The integrity of the RNA was checked by agarose gel electrophoresis using ethidium bromide staining. RNA concentration was evaluated in triplicate by absorbance at 260 nm.

Northern blot analysis. Fifteen micrograms of total RNA were denaturated in 50% formamide, 17.5% formaldehyde, and 1× MOPS buffer (20 mM MOPS, 5 mM sodium acetate, and 1 mM Na₂-EDTA) at 60°C for 15 min. The samples were then separated for 6 h at 65 V in a 6% formaldehyde/1% agarose gel and transferred to a nylon membrane (Amersham Buchler, Braunschweig, Germany) by overnight capillary blotting. RNA fixation was done by ultraviolet irradiation. Complete transfer of RNA to the membrane was controlled by ethidium bromide staining of the gel.

Membrane preparation. Membrane fragments were isolated at 4°C from powdered ventricular tissue of single hearts essentially as described previously (37). Briefly, homogenization was performed three times for 10 s each time at 21,000 revolutions/min in 10 vol of 0.75 mM KCl, 1 mM EDTA, 0.2 mM dithiothreitol (DTT), 15 mM NaF, 0.1 mM phenylmethylsulfonyl fluoride (PMSF), and 5 mM histidine-HCl (pH 7.4) using a Brinkmann Polytron PT 3000 (Kinematica, Littau/Lucerne, Switzerland) equipped with a PT-DA3007/2 piston. The homogenate was pelleted at 150,000 g using an Optima TL ultracentrifuge (Beckman Instruments, Munich, Germany), rehomogenized, and sedimented as before. After an additional wash with hypotonic buffer containing 0.2 mM DTT, 0.1 mM PMSF, 15 mM NaF, and 10 mM histidine-HCl (pH 7.4), the 150,000-g membrane pellet was resuspended in 250 mM sucrose and 10 mM histidine-HCl (pH 7.4) to a final protein concentration of 4-8 mg/ml. Samples were shock frozen and stored in liquid nitrogen until use. For phosphorylation experiments, samples of the final membrane suspension and the original homogenate were diluted threefold in phosphoprotein protection buffer containing (in mM) 250 sucrose, 10 histidine (pH 7.4), 150 KH₂PO₄, 50 NaF, and 30 EDTA and were stored at 80°C. Another sample of the final membrane suspension was diluted in 50 mM sodium bicarbonate buffer (pH 9.4) to a final protein concentration of 25 µg/ml and was stored at 20°C until used for immunochemical quantitation by ELISA. Membrane fragments appeared mainly as closed vesicles as revealed by transmission electron microscopic analysis of thin layers of the final 150,000-g pellet. Ultracentrifugation at 150,000-g at all preparation steps ensured complete recovery of SL and SR membrane fragments from the original tissue homogenate. This allows quantitative evaluation of membrane proteins and activities per milligram of membrane protein as well as per unit of tissue mass (37). Enrichment in microsomal markers was mainly due to the removal of soluble and contractile proteins at high ionic strength. For the reported experiments, the purification of the SR marker phospholamban was 3.1 ± 0.4-, 3.4 ± 0.5-, and 3.6 ± 0.5-fold in membrane preparations from cardiac homogenates of euthyroid control, hypothyroid, and hyperthyroid animals, respectively. Crude cardiac membranes isolated according to this protocol comprised ~30% of total ventricular protein in all experimental groups.

Western blot analysis and ELISA. For Western blotting, 100 µg of membrane protein was solubilized for 30 min at room temperature in 1.5% SDS, 62.5 mM Tris · HCl (pH 6.8), 7.5% glycerol, 3.8% mercaptoethanol, 0.0005% pyronin, and 0.04% bromphenol blue, and proteins were separated by electrophoresis in SDS-polyacrylamide gels (total monomer concentration 7.6%, cross-linking monomer concentration 2.67%). The standard Laemmli protocol (17) was modified by including 4 M urea in the gels. Proteins were transferred to nitrocellulose membrane BA 85 (Schleicher and Schuell, Dassel, Germany) for 1.1 h at a constant voltage of 100 V at 4°C using a Minitransblot cell (Bio-Rad Laboratories, Richmond, CA) as described earlier (35). Membranes were blocked for 1.5 h at room temperature in Tris-buffered saline containing Tween 20 [TBS-T; 10 mM Tris · HCl (pH 7.4), 150 mM NaCl, and 0.05% Tween 20] with 3% ovalbumin. Three subsequent washes of 10 min each in TBS-T were performed before the addition of the primary antibodies. For detecting the SR Ca²⁺-ATPase, the membranes (3 µg protein per lane) were first incubated with rabbit anti-SERCA2 antiserum (dilution 1:5,000; kind gift of Dr. Dillmann, San Diego, CA) for 1.5 h at room temperature in detergent-free TBS [50 mM Tris · HCl (pH 7.4) and 120 mM NaCl] containing 1% BSA and 0.04% NaN₃. They were then washed two times for 10 min in TBS-T with 3% ovalbumin and, finally, two times for 5 min in TBS. The nitrocellulose membranes were incubated for 1 h with goat anti-rabbit IgG-horseradish peroxidase conjugate (1:5,000 and in TBS with 0.1% ovalbumin) and washed twice for 10 min in TBS-T and TBS, respectively. For detecting immunoreactive NCX, 30 µg solubilized protein per lane was electrophoresed and a 1:1,000 diluted rabbit anti-canine NCX antiserum (Swant, Bellizona, Switzerland) was used. Immunoreactive SERCA2 and NCX were visualized using an enhanced chemoluminescence analysis kit (ECL, Amersham, Little Chalfont, UK) and Kodak X-OMAT AR X-ray film. Exposure time was 1 min. The optical density of the immunoreactive bands was quantitated after densitometric scanning using a PDI scanner (model DNA 35; PDI, Huntington Station, NY). Distinct amounts of protein (1-10 µg/lane for SERCA2 and 5-50 µg/lane for NCX) were used to check the linearity between the amount of applied protein and the intensity of the optical signal. The value of that parameter was considered to reflect the relative amounts of SERCA2 and NCX. Detection of phospholamban by Western blotting analysis was performed as described elsewhere (11).

Ca²⁺ transport measurements. Initial rates of SR oxalate-supported Ca²⁺ uptake were estimated in crude membrane vesicles at 0.21 µM free Ca²⁺ by using a standard procedure (36). The reaction medium contained 40 mM imidazole (pH 7.0), 100 mM KCl, 5 mM MgCl₂, 5 mM Tris-ATP, 6 mM phosphocreatine, 10 mM K-oxalate, 0.2 mM EGTA, 10 mM NaN₃, 0.1 mM ⁴⁵CaCl₂ (1.9 × 10¹¹ Bq/mol), and 10-20 µg membrane protein per 0.25 ml. To determine the maximum rate of Ca²⁺ uptake (V_max) and the free Ca²⁺ concentration at half-maximal Ca²⁺ uptake rate (EC₅₀), the total Ca²⁺ concentration was varied in the assay to yield 10 different free Ca²⁺ concentrations between 0.04 and 2.4 µM. The free Ca²⁺ concentration was calculated using Fabiato's computer program as described elsewhere (35). After 2 min of preincubation at 37°C, the measurement was started by addition of membranes. At 0.5-min intervals, 0.05-ml samples were filtered through 0.45-µm Millipore filters with the use of a vacuum pump. Filters were then washed twice with 3 ml ice-cold solution containing 100 mM KCl, 2 mM EGTA, and 40 mM imidazole (pH 7.0). Radioactivity associated with the dry filters was determined by liquid scintillation counting. The rate of uptake was calculated by linear regression of data points and expressed as nanomoles of Ca²⁺ per milligram of membrane protein per minute. Enzfitter, a nonlinear regression data analysis program by R. J. Leatherbarrow (BIOSOFT, Cambridge, UK) was used to determine EC₅₀ and V_max values.

Protein kinase A-catalyzed phosphorylation of phospholamban. Crude membranes were incubated with [-³²P]ATP for 5 min at 30°C in the presence of C subunit of protein kinase A and in the absence of Ca²⁺ as described previously (9, 35). In this process, only those phosphorylation sites not yet phosphorylated in vivo can be filled by the in vitro phosphorylation reaction. After 2 min of preincubation, the reaction was started by addition of membranes (40 µg protein) into a 40-µl mixture containing 40 mM HEPES-Tris (pH 7.4), 10 mM MgCl₂, 1 mM EGTA, 1 mM DTT, 20 mM NaF, 2 µM C subunit, and 100 µM [-³²P]ATP (specific activity 400 dpm/pmol). The reaction was stopped by 2 ml of ice-cold 15% TCA containing 50 mM H₃PO₄ and 0.5 mM Na-ATP. After the addition of 100 µg BSA, the denaturated proteins were centrifuged (3,000 g for 10 min) and solubilized in 2% SDS, 1% mercaptoethanol, 5 mM EDTA, and 50 mM H₃PO₄-Tris (pH 6.8). Urea-SDS-PAGE, gel staining, destaining, autoradiography, and measurement of radioactivity associated with low-molecular-weight phospholamban were performed as described earlier (35). The amount of ³²P incorporated into phospholamban was expressed as picomoles ³²P per milligram of membrane protein.

Miscellaneous. Membrane protein was determined according to Lowry et al. (19) after samples were incubates with 1 M NaOH (30 min, 25°C). Ovalbumin was used as a standard. Solutions for Ca²⁺ transport measurements were made with deionized water; contaminant Ca²⁺ did not exceed 3 µM. ⁴⁵CaCl₂ was obtained from Amersham Buchler; valinomycin from Boehringer Mannheim (Mannheim, Germany); superpure NaCl, KCl, and CaCl₂ from Merck (Darmstadt, Germany); and other biochemicals from Sigma-Aldrich (Deisenhofen, Germany).

Statistical analysis. The effect of thyroid status was assessed by one-way ANOVA. When the F ratio exceeded the critical value (at P < 0.05), Dunnett's multiple comparisons test was used to compare individual experimental groups with the control group. In cases of unequal SD values, the Kruskal-Wallis nonparametric analysis was performed; at P < 0.05, Dunn's multiple comparisons test was employed for identification of significant group-to-group differences. Values are given as means ± SD. Statistical significance was assumed at P < 0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Weight parameters and crude membrane protein. In comparison with the euthyroid group, the body weight of both hypo- and hyperthyroid animals was decreased by 28 and 23%, respectively (Table 1). The total heart weight of hypothyroid rats was reduced to about one-half that of the euthyroid controls. The other heart weight parameters also decreased, except for the right-to-left ventricular weight ratio (RV/LV), which did not differ from the euthyroid value. Treatment with T₃ resulted in increased absolute heart and right ventricular weights by 37 and 55%, respectively, compared with euthyroid animals. All the relative heart weight parameters were markedly increased in hyperthyroid rats due to body growth retardation. Elevated RV/LV and unaltered absolute left ventricular weight indicate that cardiac hypertrophy induced by the excessive thyroid hormone was more pronounced in the right ventricle. Relative values similar to those in the T₃ group were obtained in animals treated simultaneously with PTU and a low dose of T₃ (Table 1).

Steady-state levels of NCX, SERCA2, and GAPDH mRNA. The same Northern blots containing total RNA from either right or left ventricles of 21-day-old rats at different thyroid states were sequentially hybridized with the NCX, SERCA2, GAPDH, and 18S rRNA probes (Fig. 1). The 1.9-kb signal of 18S rRNA indicated the amount of RNA loaded for each sample so that the relative abundance of both NCX and SERCA2 mRNAs could be assessed. For quantification of the NCX mRNA level, a predominant 7.2-kb NCX transcript was used. In addition, a minor 4-kb signal was found when the blots were probed with the NCX. With the SERCA2 cDNA, a single hybridization band was obtained at the expected position, corresponding to a mRNA size of 4.4 kb. After hybridization with the GAPDH probe, a 1.3-kb transcript was observed.

View larger version (82K): [in this window] [in a new window]   Fig. 1.   Autoradiographs of Northern hybridization analysis of Na+/Ca2+ exchanger (NCX), sarcoendoplasmic reticulum Ca2+ (SERCA2), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA, and 18S rRNA from the right (RV) and left ventricles (LV) of 21-day-old rats with different thyroid statuses. Con, untreated controls (euthyroid); PTU, 6-n-propyl-2-thiouracil treated (hypothyroid); T3, 3,3', 5-triiodo-L-thyronine treated (hyperthyroid); PTU + T3, combined PTU and T3 treated. Total RNA of 15 µg was subjected to size fractionation by electrophoresis in 6% formaldehyde-1% agarose gel. RNA transfer and blot hybridization are described in METHODS; 18S rRNA message estimation was used as an internal standard.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   mRNA levels of NCX (filled bars) and SERCA2 (open bars) in RV and LV of 21-day-old rats with different thyroid statuses. Con, n = 5; PTU, n = 6; T3, n = 6; and PTU + T3, n = 4. Expression levels of specific mRNA were normalized to 18S rRNA and expressed as percentages of euthyroid controls value. * P < 0.05, significantly different from euthyroid controls.

Protein levels of NCX and SERCA2. To examine whether the alterations in NCX and SERCA2 steady-state mRNA levels were accompanied by changes at the protein level, Western blot analysis of membrane preparations isolated from ventricular myocardium was performed. The use of the specific polyclonal antibody directed against the canine NCX revealed two bands: one at the expected position corresponding to 120 kDa and the other at the position of 40 kDa. Because the second band is likely to be a nonspecific signal (36), densitometric values of the major 120-kDa band (Fig. 3) were used for semiquantitative measurement of NCX protein levels. On the other hand, only one immunoreactive signal (100 kDa) was found after the incubation of membranes with anti-rat SERCA2 antibody (Fig. 3).

View larger version (62K): [in this window] [in a new window]   Fig. 3.   Effect of postnatal thyroid status on NCX, SERCA2, and phospholamban (PLB) protein expression in rat ventricular myocardium. Representative Western blots for NCX (120-kDa signals), SERCA2 (100-kDa signals), and PLB (6.5-kDa signals) are shown. Protein per lane: 30 µg (NCX) and 3 µg (SERCA2 and PLB). See also METHODS.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Protein levels of NCX (filled bars) and SERCA2 (open bars) in ventricular myocardium of 21-day-old rats with different thyroid statuses. Con, n = 6; PTU, n = 6; T3, n = 6; and PTU + T3, n = 6. Results are expressed as percentages of euthyroid control values. * P < 0.05, significantly different from euthyroid controls.

NCX and SERCA2 transport activities. SL Na⁺-dependent Ca²⁺ transport and SR oxalate-supported Ca²⁺ uptake were measured in membrane vesicles prepared from ventricular myocardium of experimental and control rats. Whereas the SR Ca²⁺ uptake was linear with time for at least 3 min, NCX-mediated Ca²⁺ transport was linear only up to 2 s (37). The activity of both systems was normalized to milligrams of crude membrane protein and expressed as a percentage of the euthyroid control level (Fig. 5). In hypothyroid hearts, the NCX activity was increased by 71%, whereas SERCA2 transport rate declined by 70%, compared with euthyroid controls. A similar decline was observed if SR Ca²⁺ uptake was assayed in the presence of the Ca²⁺-release channel inhibitor ruthenium red (data not shown). Hyperthyroid rats exhibited opposite changes: a decrease of Na⁺-dependent Ca²⁺ transport by 50% was accompanied by an elevation in the rate of SR Ca²⁺ uptake by 150%. This latter effect was even higher (by 210%) in the presence of ruthenium red (data not shown). Similar alterations were observed in animals treated simultaneously with PTU and T₃ (Fig. 5).

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Differences in relative NCX (filled bars) and SERCA2 (open bars) activities in crude membranes isolated from 21-day-old ventricular myocardium of rats with different thyroid statuses. Sarcolemmal NCX was measured for 2 s as sodium-mediated 45Ca2+ transport, and SR ATPase-mediated Ca2+ transport was measured as oxalate-supported 45Ca2+ uptake. Values are expressed as percentages of euthyroid controls. Con, n = 7; PTU, n = 7; T3, n = 6; and PTU + T3, n = 6. See also METHODS. Absolute values of NCX and SERCA2 activities in control group were 0.362 ± 0.067 nmol Ca2+ · mg protein1 · 2 s1 and 8.26 ± 3.50 nmol Ca2+ · mg protein1 · min1, respectively. * P < 0.05, significantly different from euthyroid controls.

Protein levels and phosphorylation of phospholamban. To examine whether the observed thyroid-dependent changes in the activity of SR Ca²⁺ pump were related to alterations in the control of this system by phospholamban, immunoreactive phospholamban and protein kinase A-dependent in vitro phosphorylation of this protein were determined. Figures 2 and 6 show that immunoreactive phospholamban was increased in membranes isolated from hypothyroid hearts. In contrast, hyperthyroid animals exhibited significantly lower levels of phospholamban than euthyroid controls (Figs. 2 and 6).

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Relative levels of phospholamban measured by enzyme-linked immunosorbent assay in ventricular myocardium of 21-day-old rats with different thyroid statuses. Con, n = 6; PTU, n = 7; and T3, n = 7. Product formation of peroxidase reaction in a substrate mixture with o-phenylenediamine and H2O2 was monitored spectrophotometrically at 492 nm. A monoclonal mouse anti-phospholamban that cross-reacts with both nonphosphorylated and phosphorylated rat phospholamban was used. OD, optical density. * P < 0.05, significantly different from euthyroid controls.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Protein kinase A-catalyzed in vitro 32P incorporation into ventricular phospholamban of 21-day-old rats with different thyroid statuses. Con, n = 7; PTU, n = 7; T3, n = 6; and PTU + T3, n = 5. Phosphorylation reaction was performed for 5 min at 30°C under stringent phosphoprotein protection conditions (35) in presence of 1 mM EGTA. Heated SDS-solubilized samples (40 µg protein/lane) were subjected to urea-SDS-PAGE and autoradiography. Radioactivity associated with low-molecular-weight form of phospholamban was determined by scintillation counting of respective gel band cut out from dried gels. See also METHODS. * P < 0.05, significantly different from euthyroid controls.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Relationship between in vitro phospholamban phosphorylation catalyzed by C subunit of protein kinase A and rate of oxalate-supported SR Ca2+ uptake in ventricular myocardium of 21-day-old rats. Mean values of duplicate determinations for each analyzed membrane preparation of single hearts are shown. Line was obtained by linear regression analysis over all individual values of different experimental groups.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The main objective of this study was to investigate the effect of thyroid hormone on the expression and function of the SL NCX and SR Ca²⁺-ATPase in the immature ventricular myocardium of 21-day-old rats. Northern blot analysis was used to determine the steady-state level of mRNA encoding both Ca²⁺ transport systems. Figure 1 shows that NCX probe hybridized to an abundant major 7.2-kb transcript. In contrast to Boerth et al. (5) but in agreement with Windhager et al. (42) and Yu et al. (43), we also observed a minor 4-kb species with this probe in the rat heart. Because the ratio of 7.2- and 4-kb signals remained unchanged after either lower or higher stringency used for hybridization and washing (unpublished observations), a less abundant 4-kb transcript is likely to be a processed exchanger mRNA or an isoform. However, the proportion of both hybridization signals was not affected by hypo- or hyperthyroidism. The SERCA2 probe detected a single mRNA species of 4.4 kb; minor contamination bands found above the specific signal (Fig. 1) have already been documented by Arai et al. (1), who have also analyzed total RNA instead of poly(A)⁺ RNA. GAPDH message estimation was used as an internal standard in previous studies (1, 8, 36). In contrast to data in Ref. 1, we show that the expression of this glycolytic enzyme is significantly lowered in hypothyroid rat ventricle. This decrease was even more pronounced in atrial tissue (unpublished observations). On the other hand, the level of 18S rRNA was found to be unaltered by any of the thyroid status examined here. Therefore, this subunit was used as an internal standard to verify that a constant amount of RNA was loaded onto each lane on Northern blots.

Measurements of Ca²⁺ transport functions and protein levels of NCX and SERCA2 were performed in isolated membranes that were prepared with the use of an established method for quantitative recovery of cardiac membranes from ventricular homogenates (36, 37) under conditions of stringent protection of preexisting phosphoproteins. This protocol resulted in no significant intergroup differences in purification of the SR marker phospholamban. This indicates that group-to-group differences in SERCA2 expression and function cannot be due to treatment-dependent variations in microsomal enrichment in the analyzed membrane preparations. It should also be pointed out that similar intergroup differences in SR Ca²⁺ transport and phospholamban phosphorylation were observed in whole tissue homogenates. In particular, the dependence shown in Fig. 8 was also found if homogenates were analyzed. At the submicromolar free Ca²⁺ concentration of 0.21 µM used in this study, the rate of SR Ca²⁺ uptake depends also on the extent of phospholamban phosphorylation that modulates the Ca²⁺ sensitivity and velocity of the SR Ca²⁺ pump (34). That this is really a contributing factor under these assay conditions is supported by the results of kinetic analysis at varying Ca²⁺ concentrations. In fact, the Ca²⁺ sensitivity of the SR Ca²⁺ transport system was found to be highest in hyperthyroidism and lowest in hypothyroidism. Tissue handling, membrane isolation, and phosphorylation experiments were performed under conditions protecting in vivo preexisting phosphoproteins from dephosphorylation after removal of the hearts (35). The use of monoclonal anti-phospholamban antibody permitted the quantitation of relative levels of tissue phospholamban independently on the status of phosphorylation of this protein (31).

Thyroid hormone levels increase postnatally (38), and the results of the present study demonstrate that abolishing this neonatal thyroid surge in rats prevented both the normal postnatal decline in NCX activity and an increase in SR Ca²⁺-sequestering activity (36). Although the marked depression of the SERCA2 activity observed after administration of PTU is entirely consistent with previous studies in developing (14, 40) and adult rat hearts (2, 3), hypothyroidism-induced elevation of the NCX rate has not been previously reported. We found that the expression of mRNA coding for NCX was upregulated, whereas the mRNA level of SERCA2 was downregulated, in hypothyroid right and left ventricles of 3-wk-old animals. These reciprocal changes were accompanied by corresponding alterations in NCX and SERCA2 protein levels. Recently, similar results have been reported in hypothyroid rabbits of the same age; however, SERCA2 expression remained constant in this model, probably due to insufficient reduction of thyroid hormone level in the blood (4).

Our finding that the level of immunoreactive phospholamban was enhanced in hypothyroid hearts is concordant with the previous study in adult rats (13). In contrast, either no change (24) or a decrease in phospholamban expression (1) has been observed in hypothyroid rabbits. This suggests species-dependent differences in the responsiveness of the phospholamban gene to thyroid hormones. The results of the phosphorylation experiments show that in vitro ³²P incorporation into phospholamban was significantly increased in hypothyroidism. This elevation was higher than could be expected from the increase in the tissue level of phospholamban. Although we did not determine the absolute basal phosphorylation in the tissue, our data suggest that the in vivo phosphorylation of phospholamban is lower in hypothyroid hearts compared with that in euthyroid and hyperthyroid hearts, which is most probably due to a lower adrenergic responsiveness (41). Taken together, these data indicate that in developing hypothyroid ventricles, the alteration of NCX- and SERCA2-mediated Ca²⁺ transport rates is accomplished primarily by increasing (NCX) or decreasing (SERCA2) the number of functional protein molecules. In addition, higher phospholamban expression and, most likely, reduced in vivo phosphorylation of this protein result in increased levels of nonphosphorylated phospholamban. As this nonphosphorylated modulator suppresses the SR Ca²⁺ pump, a marked decrease in the SR Ca²⁺ uptake rate occurs.

Hyperthyroidism was associated with opposite changes in the activities of the two Ca²⁺ transport systems examined. The elevated rate of SR Ca²⁺ uptake (at submicromolar free Ca²⁺) in isolated membranes that we observed after T₃ treatment confirms earlier data obtained with the same model in whole tissue homogenates (14). Similar results have been also described in thyroid hormone-treated adult rat hearts (18) and neonatal rat cardiomyocytes (12). In contrast, Wibo et al. did not observe an anticipated increase in the thapsigargin-sensitive SERCA2 activity in cardiac homogenates in 21-day-old hyperthyroid rats (40). Our finding that NCX activity decreases in hyperthyroid animals has not been previously reported.

In contrast to functional data, both mRNA and protein levels of ventricular NCX and SERCA2 remained unaffected by T₃ treatment. That the SERCA2 protein content is not significantly elevated in hyperthyroid versus euthyroid hearts is supported by the V_max values in Table 2, which do not differ significantly between both groups. This suggests that the normal postnatal level of thyroid hormones in the blood, which is known to reach peak values during the third postnatal week (38), might be sufficient for maximal stimulation of SERCA2 expression in the rat heart. In line with this view, T₃ administration accelerated the normal postnatal increase in densities of ryanodine receptors and -adrenoceptors, and the redistribution of L-type Ca²⁺ channels from nonjunctional to junctional domains of the SL and T tubules in 7- and 14-day-old rats (39) but not in 21-day-old rats (40). It remains to be determined whether excessive T₃ could stimulate the expression of SERCA2 at earlier developmental stages. Adult hyperthyroid animals exhibited higher expression of cardiac SERCA2 (1, 24, 29) and lower expression of cardiac NCX (4) than age-matched euthyroid controls. This may be related to the fact that the level of thyroid hormones in the blood of adult rats is only about one-half of that in 3-wk-old animals (38).

However, regulatory mechanisms other than altered expression of the respective genes seem to be responsible for elevated SR Ca²⁺ uptake and reduced NCX rate in 21-day-old hyperthyroid rats. Our data together with previous reports (1, 12, 24) indicate that the expression of cardiac phospholamban is significantly depressed in hyperthyroidism. Moreover, in vitro ³²P incorporation into this protein was also markedly decreased, suggesting that the ratio of phosphorylated to nonphosphorylated phospholamban was increased by T₃ treatment. We assume that the reduction in inhibitory nonphosphorylated phospholamban, due to its decreased expression and enhanced in vivo phosphorylation, rather than enhanced SERCA2 expression accounts primarily for elevated SR Ca²⁺ transport activity in postnatal hyperthyroidism. The decrease in NCX rate might be secondary to a significant increase in Na⁺-K⁺-ATPase activity, which was observed in hyperthyroid rats of the same age (40). Previously reported data on reciprocal regulation of cardiac Na⁺-K⁺-ATPase and NCX during altered thyroid status (20) support this view. Moreover, it should not be dismissed that modifications of cardiac membrane lipids such as cholesterol, phosphatidylethanolamine, and phosphatidylcholine by hypo- and hyperthyroidism (15, 32) could also contribute to the observed changes in NCX activity (16, 26).

In conclusion, thyroid hormones play an important role in the reciprocal control of NCX (decreasing activity) and SERCA2 (increasing activity) during postnatal development of the rat heart. Hypothyroidism prevented these changes, particularly those due to altered expression of the genes encoding the respective transport proteins. In addition, increased expression and reduced protein kinase A-catalyzed in vivo phosphorylation of phospholamban contribute to suppression of the cardiac SR Ca²⁺-transport activity in this state. In contrast, a moderate decline in expression of phospholamban and a strong enhancement of in vivo phosphorylation of this regulatory protein are most likely the major contributing mechanisms for an elevated SR Ca²⁺-transport activity in postnatal hyperthyroidism, rather than an alteration in SERCA2 expression. In this state, the decline of NCX activity appears to be due to a mechanism that does not involve altered expression of the NCX gene. The antithetical regulation of the two competing Ca²⁺ transport systems by thyroid hormones could be an important mechanism determining the postnatal development of myocardial Ca²⁺ handling.