Short-term response to cold promotes a small but significant rise in serum T3 in euthyroid rabbits, where the heart is an important target of T3 action. In this work, we measured changes in sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA2a) and phospholamban (PLB) in hearts of hypo- and hyperthyroid rabbits and compared them with modifications induced by short- and long-term cold exposure. Short-term cold exposure promotes a small increase in T3 and, similar to hyperthyroidism, induces an increase of heart SERCA2a expression. The total PLB content does not change in hyperthyroidism, but short-term cold exposure promotes a significant decrease in total PLB and an increase in the ratio between phosphorylated and total PLB. The temperature of a given tissue depends on the balance between the heat provided by blood circulation and the rate of heat production by the tissue. In an attempt to evaluate the heat contribution of cardiac tissue, we measured mitochondrial respiration in permeabilized cardiac muscle and heat produced by cardiac sarcoplasmic reticulum (SR) during Ca2+ transport. We observed that there was an increase in oxygen consumption and heat production during Ca2+ transport by cardiac SR in both hyperthyroidism and short-term cold exposure. In contrast, both the mitochondrial respiration rate and heat derived from Ca2+ transport were decreased in hypothyroid rabbits. The heart changes in oxygen consumption, SERCA2a-PLB ratio, and Ca2+-ATPase activity detected during short-term cold exposure were abolished after cold adaptation. We hypothesize that the transient rise in serum T3 contributes to the short-term response to cold exposure.
- calcium adenosinetriphosphatase
- mitochondrial respiration
thyroid hormone significantly affects the cardiovascular system, producing profound changes in cardiac contractility, cardiac output, blood pressure, and myocardial oxygen consumption (9, 30). The mechanisms underlying these changes involve direct gene transcriptional regulation induced by thyroid hormone, which, in turn, affects the excitation-contraction coupling by modifying both the cytosolic Ca2+ cycling (33, 50, 55) and myofilament expression (25, 36). Furthermore, thyroid hormone alters energetic metabolism in the myocardium (48) and interacts with the sympathetic nervous system (SNS) by increasing adrenergic receptor function and/or density (4, 52).
The sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA) plays a central role in the coupling of contraction and relaxation in the myocardium. This enzyme transports Ca2+ inside the SR lumen by using ATP hydrolysis as an energy source (19, 26, 28). In addition to being important for muscle contraction, SERCA is one of the heat sources contributing to the thermogenesis of animals without brown adipose tissue (BAT) (17, 29). Three distinct genes encode SERCA isoforms, but the physiological significance of this isoform diversity is not clear. Cardiac muscle expresses only SERCA2a protein, white muscles express SERCA1, and red muscles express both SERCA1 and SERCA2a. The SERCA2b and SERCA3 genes are expressed in nonmuscular tissues such as blood platelets and lymphoid tissue (39–41).
The thyroid hormone 3,5,3′-triiodo l-thyronine (T3) regulates both the expression and function of SERCA proteins; their effects vary depending on the muscle type (1, 2, 51, 54). In the heart, SERCA2a expression is modulated by T3 levels; this effect is accompanied by alterations in phospholamban (PLB) content, a protein that inhibits SERCA activity through direct protein-protein interaction (1, 43, 45, 50). Recently, it has been demonstrated that SERCA1 is able to hydrolyze part of the ATP consumed through an alternative catalytic route that is uncoupled from Ca2+ transport. In this route, the enzyme converts the energy that should be used for Ca2+ transport (work) into heat (21). This uncoupled activity is positively regulated by thyroid hormone in skeletal muscle (2, 49) and is responsible for changes in muscular thermogenesis (20). In contrast to the role of T3 in skeletal muscle, little is known about the role of thyroid hormone in cardiac muscle thermogenesis.
Chronic exposure to low temperature is associated with a high incidence of hypertension and related cardiovascular diseases, including cardiac hypertrophy (24, 59). Acute cold exposure leads to 1) increased heart rate, 2) a rise in both systolic and diastolic blood pressures, and 3) enhancement of cardiovascular responsiveness to β-adrenergic stimulation (5, 24). Previous studies have shown that cold exposure activates the SNS (47) and increases circulating thyroid hormones (46). There have been several reports on the effect of cold exposure in rats (13, 53); however, as far as we know, the effect of cold exposure on the rabbit heart has not been studied. Recently, it has been proposed that the Ca2+ cycling in the human heart is more similar to rabbits than to rats (6, 58). One of the main differences between rats and rabbits is the presence of BAT in rats. It is well documented that thermoregulation in rats depends significantly on BAT function and is thyroid hormone dependent (7, 8, 12). Hypothyroid rats are not able to survive a sudden cold exposure. Recently, we observed that, unlike rats, hypothyroid rabbits are able to survive a sudden cold exposure and, therefore, are able to adapt to low temperature (3). In this report, we investigate the rate of heat production by cardiac Ca2+-ATPase in hypo- and hyperthyroid rabbits and compare them with modifications induced by short- and long-term cold exposure. In addition, we show the changes in SERCA2a and PLB expression and mitochondrial respiration in permeabilized cardiac muscle.
MATERIALS AND METHODS
Hyperthyroidism, hypothyroidism, and cold acclimation.
Adult male New Zealand White rabbits were divided into different groups: euthyroid, hypothyroid, and hyperthyroid rabbits kept at room temperature (25–28°C) and euthyroid animals exposed to 4°C for either 72 or 240 h. The hypothyroid state was induced by oral administration of propylthiouracil (0.08% wt/vol) in drinking water for 21 days (3). Hyperthyroidism was induced by subcutaneous injection of T4 (100 μg/kg body wt) for 10 days (1, 2). Blood was obtained at the day of death by cardiac puncture, and serum T4 and T3 levels were determined by specific radioimmunoassays using 125I as a tracer (2). During all of the experiments, the animals were treated in accordance with the published rules for animal laboratorial care; our protocol was approved by the Institutional Committee for the Use of Animals for Research.
Gel electrophoresis and Western blotting.
Protein samples were resolved on polyacrylamide gels: 7.5% for SERCA2a protein, 10% for serine-16 (Ser16) phosphorylated PLB, and 13% for total PLB, according to Laemmli (37). Electrotransfer of proteins from the gel to polyvinylidene difluoride (PVDF) membranes was performed for 20 min at 250 mA per gel in 25 mM Tris, 192 mM glycine, and 10% methanol using a Mini Trans-Blot cell from Bio-Rad. Membranes were blocked with 3% nonfat dry milk in Tris-buffered saline with 0.1% Tween 20 for 1 h at room temperature. Membranes were then washed and incubated for 1 h with anti-SERCA2a, anti-PLB, and anti-phospho(Ser16)-PLB monoclonal antibodies at room temperature. The membranes were washed and incubated with anti-mouse or anti-rabbit secondary antibody. The immunoblots were revealed using an ECL detection kit from Amersham-Pharmacia Biotech (Little Chalfont, UK). Monoclonal anti-SERCA2 antibody (clone IID8) was obtained from Affinity Bioreagents (Sao Paulo, Brazil), and monoclonal anti-PLB and polyclonal anti-phospho(Ser16)-PLB antibodies were obtained from Upstate Biotechnology (Sao Paulo, Brazil). Anti-PLB appears as a double band when the samples are not boiled before being loaded on the gel, and this is attributed to PLB oligomerization (18, 45, 56).
Sarcoplasmic reticulum vesicles.
Vesicles derived from the longitudinal sarcoplasmic reticulum (SRV) of rabbit cardiac muscle were prepared as previously described (22) and stored at −70°C.
This parameter was measured by the filtration method (15). For 45Ca uptake, trace amounts of 45Ca were included in the assay medium. The reaction was arrested by filtering samples of the assay medium through Millipore filters. After filtration, the filters were washed five times with 5 ml of 3 mM La(NO3)3, and the radioactivity remaining on the filters was counted using a liquid scintillation counter.
ATP hydrolysis was estimated by measuring the Pi released using a colorimetric method (23). The reaction was stopped with trichloroacetic acid (final concentration 5% wt/vol). Two different ATPase activities can be distinguished in SRV (19, 26, 28). The magnesium-dependent activity (Mg2+-ATPase) requires only Mg2+ for its activation and is measured in the presence of 10 mM EGTA to remove contaminating Ca2+ in the medium. The Ca2+-dependent ATPase activity (Ca2+-ATPase), which is responsible for Ca2+ transport, was determined by subtracting the Mg2+-ATPase from the activity measured in the presence of both Mg2+ and Ca2+. The SERCA ATPase activities that are described in the data correspond to the Ca2+-ATPase. Unlike skeletal muscle, vesicles derived from heart sarcoplasmic reticulum contain high Mg2+-ATPase and low Ca2+-ATPase activities. Taffet and Tate (57) reported that the ATP Km for the Mg2+-ATPase is 6–10 times higher than that for the Ca2+-ATPase (500 and 50 μM ATP, respectively). We took advantage of this Km difference to improve the measurement of the Ca2+-ATPase by using an ATP-regenerating system so that the ATP concentration in the assay medium would be low enough to reduce the proportion of ATPase activity that corresponds to Mg2+-ATPase. The assay medium contained 50 μM ATP, 2 mM phosphoenolpyruvate, and 10 U/ml pyruvate kinase. There was substantial Mg2+-ATPase activity, and only 37% of the total activity was Ca2+-ATPase when using 1 mM ATP (Fig. 1A). With the ATP-regenerating system, there was a sharp decrease of the Mg2+-ATPase activity such that ∼73% of the total activity corresponded to Ca2+-ATPase (Fig. 1B).
Heat of reaction.
The heat of reaction was measured using an OMEGA isothermal titration calorimeter from Microcal (Northampton, MA). The calorimeter cell (1.5 ml) was filled with reaction medium, and the reference cell was filled with Milli-Q water. After an equilibration at 35°C, the reaction was started by injecting vesicles into the sample cell, and the heat change was recorded for 30 min. The volume of vesicle suspension injected into the reaction cell varied between 0.03 and 0.05 ml. The heat change that was measured during the initial 2 min after vesicle injection was discarded to avoid artifacts, such as heat derived from the dilution of the vesicle suspension and binding of ions to the Ca2+-ATPase. The duration of these events is less than 1 min. Calorimetric enthalpy (ΔHcal) was calculated by dividing the amount of heat released by the amount of ATP hydrolyzed during the same time (21). The units used were moles for substrate hydrolyzed and kilocalories for heat released. Negative values indicate that the reaction is exothermic. In a typical experiment, the assay medium was divided into three samples, which were used for the simultaneous measurement of Ca2+ uptake, ATP hydrolysis, and heat release. These different measurements were started simultaneously with the addition of vesicles to a final concentration of 50 μg protein/ml.
Oxygen consumption and citrate synthase activity.
For oxygen consumption measurements, a small piece (1–2 mg) of the left ventricular muscle was placed into a petri dish on ice with 1 ml of relaxing solution containing 10 mM Ca2+-EGTA buffer, 0.1 μM free calcium, 20 mM imidazole, 50 mM K+-2-(4-morpholino)ethanesulfonic acid, 0.5 mM dithiothreitol, 6.56 mM MgCl2, 5.77 mM ATP, and 15 mM phosphocreatine, pH 7.1. Individual fiber bundles were separated with a sharp forceps. The fiber bundles were permeabilized for 30 min in 3 ml of ice-cold relaxing solution containing 50 μg/ml saponin. The fibers were then washed two times for 10 min with a medium (MitoMed2; Oroboros, Innsbruck, Austria) containing 0.5 mM Na2EDTA, 5 mM MgCl2·6H2O, 10 mM KH2PO4, 110 mM mannitol, 60 mM KCl, and 60 mM Tris. The muscle bundles were then immediately transferred into a respirometer (Oxygraph-2k; Oroboros) containing an air-saturated respiration medium at 25°C. We chose 25°C because at this temperature the amount of oxygen that is available in the assay medium is greater than at 37°C. The respiration medium (MiR05; Oroboros) contained 110 mM sucrose, 60 mM potassium lactobionate, 0.5 mM EGTA, 3 mM MgCl2·6H2O, 20 mM taurine, 10 mM KH2PO4, 20 mM HEPES, and 2 mg/ml bovine serum albumin, pH 7.1. The maximal respiratory rate was measured in a mixture containing 10 mM succinate, 5 mM pyruvate plus malate, and 2 mM ADP. The Oxygraph-2k is a two-chamber titration-injection respirometer with a limit of oxygen flux detection of 1 pmol·s−1·ml−1 (10). Citrate synthase activity was determined spectrophotometrically, as described previously (27).
Data are means ± SE. Analysis of differences between groups was performed with Student's t-test (P < 0.05 was considered statistically significant).
Hyperthyroidism and hypothyroidism.
Hyperthyroidism produced a significant reduction in body mass, corroborating a previous report (2) (Table 1). Serum levels of T4 and T3 were significantly elevated in hyperthyroid rabbits and decreased in hypothyroid rabbits (Table 2). Compared with the euthyroid rabbits, the normalized heart weight (heart weight/body weight) was significantly higher in hyperthyroid rabbits (Table 2). This result is due to both a decrease in body weight and an increase in heart weight (Table 1), indicating that the hyperthyroid rabbits developed cardiac hypertrophy. In hypothyroid rabbits, heart weight decreased but body weight did not change. Therefore, unlike hyperthyroid rabbits, hypothyroid animals experienced a decrease in normalized heart weight mainly due to a reduction in the cardiac weight (Tables 1 and 2).
We found no change of SERCA2a protein in hypothyroid rabbit hearts (Fig. 2A). However, we observed a 1.44-fold increase of total PLB protein (Fig. 2B), in agreement with previous reports (11, 32, 44). A decrease in the SERCA2a-to-PLB ratio resulted, mainly because of the increase in PLB. The SERCA2a-to-PLB ratio is one of the parameters used to estimate the calcium cycling kinetics in cardiac myocytes (31, 35). In hypothyroid rabbits, the fraction of phosphorylated (Ser16) PLB was smaller than that measured in euthyroid rabbits (Fig. 2C).
In accordance with earlier reports (1, 33, 43, 45, 50), we found a 2.1-fold increase in SERCA2a expression associated with hyperthyroidism (Fig. 2A). In addition, the total PLB protein expression remained unchanged (Fig. 2B), and there was a 1.3-fold increase in phosphorylated (Ser16) PLB (Fig. 2C). Unlike hypothyroidism, the SERCA2a-to-PLB ratio was higher, due to the enhancement in SERCA2a levels. PLB Ser16 phosphorylation is catalyzed by a cAMP-dependent protein kinase (PKA) and is known to relieve the inhibition of SERCA2a caused by nonphosphorylated PLB. Phosphorylation of PLB at Ser16 is the primary mechanism by which β-adrenergic agonists exert positive inotropic actions on the heart in both rats and rabbits (14, 16, 35).
To correlate the changes noted in Fig. 2 in hypo- and hyperthyroid rabbits with SERCA function, we isolated vesicles derived from rabbit cardiac muscle and measured both Ca2+ uptake and Ca2+-dependent ATPase activity. In accordance with previous report (55), Ca2+ uptake and Ca2+-ATPase activity were enhanced during hyperthyroidism and decreased during hypothyroidism (Fig. 3).
There was no change in body weight, heart weight, or the normalized heart weight ratio in euthyroid rabbits exposed to the cold for either 72 or 240 h (Tables 1 and 2). Compared with rabbits kept at room temperature, the rabbits given a short-term cold exposure presented a small but significant increase in T3 serum levels. Serum T3 returned to normal values when the cold acclimation time was prolonged to 240 h (Table 1). These data suggest that short-term response to cold rapidly stimulates thyroid gland secretion or the peripheral metabolism of T4, increasing serum T3 levels. Thyroid function returns to normal after cold adaptation.
Although the increase (∼20%) in serum T3 levels (Table 2) was small, the short-term response to cold promoted changes in SERCA2 expression and PLB phosphorylation that resemble those noted in hyperthyroid rabbits that have a higher level of serum T3 (Fig. 2, A, C, D, and F). The difference between the two groups of rabbits was a decrease in total PLB (compare Fig. 2, B and E), and the enhancement of the phosphorylated PLB in cold-exposed rabbits was more pronounced than that detected in hyperthyroid rabbits (compare Fig. 2, C and F). SERCA2 is inhibited by nonphosphorylated PLB. Thus, although in the cold-exposed rabbit the enhancement of SERCA2 was smaller than that detected in hyperthyroid rabbits, the degree of inhibition by PLB was smaller, and the overall effect was that Ca2+ accumulation, ATPase activity, and rate of heat production were similar in the two conditions (Fig. 4, A and B). The serum T3 level and the rates of Ca2+ uptake, Ca2+-ATPase activity, and heat released returned to normal values after adaptation to low temperature (Fig. 4, A and B).
Oxygen consumption, citrate synthase activity, and heat production derived from Ca2+ transport.
Heat generation plays a key role in cellular metabolism and energetic balance. Recently, we reported that the rate of heat production by SRV derived from rabbit skeletal muscle is enhanced in hyperthyroidism (2). We now have shown that in cardiac muscle, both hyperthyroidism and short-term cold exposure enhanced the rate of heat produced by SRV during Ca2+ transport (Fig. 5). In contrast, hypothyroidism decreased the rate of heat production (Fig. 5A). Heat production rates were found to be altered due to changes in the amount of ATP cleaved during Ca2+ transport. Therefore, unlike red skeletal muscles (2), cardiac SRV does not experience a variation in the amount of heat produced during the hydrolysis of each ATP molecule (ΔHcal) in either hypothyroidism or hyperthyroidism or during cold exposure. The ΔHcal varied in the five conditions between −4.22 ± 0.47 and −5.18 ± 0.68, and there was not statistically significant difference.
To further evaluate heart metabolic activity in the four conditions studied, we measured the rate of mitochondrial oxygen consumption using permeabilized left ventricular muscle tissue (Fig. 6, A and B). In the hearts of hyperthyroid rabbits and rabbits acclimated at 4°C for 72 h, the rates of oxygen consumption were found to be higher than that measured in control cardiac muscle: ∼33 and ∼44%, respectively (Fig. 6, A and B). In hypothyroid hearts, there was a ∼26% decrease in mitochondrial respiration (Fig. 6A). There was no change in respiration in hearts from animals submitted to long-term cold exposure (Fig. 6B).
Citrate synthase is an enzyme of the Krebs cycle that is commonly used as a marker of mitochondrial density (27). To verify whether the alterations in oxygen consumption that are described above were correlated with changes in mitochondria biogenesis, we measured citrate synthase activity. As shown in Fig. 6, the increase in citrate synthase activity was strictly associated with the enhancement of heart mitochondrial respiration found in both hyperthyroidism and short-term cold exposure (Fig. 6, C and D). In hypothyroidism, a decrease in citrate synthase activity was detected (Fig. 6C). Therefore, as shown in Fig. 7, there is a direct relationship between mitochondrial oxygen consumption and citrate synthase activity. Long-term adaptation to cold does not promote any change in citrate synthase activity in cardiac muscle (Fig. 6D).
Various thyroid states are known to promote significant alterations in the heart, such as changes in the contractile apparatus and in the expression of sarcoplasmic reticulum proteins (25, 33, 36, 50, 55). In this study, we focused on the adaptation of the rabbit heart during cold exposure and compared it with the effects of hypo- and hyperthyroidism.
The stoichiometry of SERCA2a-PLB is an important determinant of calcium cycling kinetics in the heart (31, 35). PLB is a transmembrane protein that interacts and inhibits SERCA2a by lowering its apparent Ca2+ affinity. Inhibitory interactions are disrupted by PLB phosphorylation at two sites: Ser16 and/or Thr17. These residues are phosphorylated by cAMP-dependent or Ca2+/calmodulin-dependent protein kinases, respectively. Ser16 phosphorylation can occur independently of Thr17 in vivo and is the main target of β-agonist stimulation (14, 16, 35). In the present study, we have shown that a short-term response to cold enhances the ratio of SERCA2a-PLB, similar to the change observed in hyperthyroid animals (Fig. 2). The changes in SERCA2a-PLB are accompanied by an enhancement of the heat production rate (Fig. 5). This probably represents a compensatory mechanism for the adaptation to cold exposure. Furthermore, both animals exhibit a high level of Ser16-phosphorylated PLB (Fig. 2, C and F), promoting a significant increase in Ca2+ uptake and Ca2+-ATPase activity by vesicles containing SERCA2a (Fig. 3 and Fig. 4). These effects are associated with increased serum T3 levels (Table 2) and are probably due to the concerted action of T3 and norepinephrine. The increase in Ca2+ uptake involves not only the rate but also the final levels of Ca2+ in the steady state. Two possibilities may account for this finding. The first is an increase of Ca2+-ATPase expression. With this possibility, we should expect an increase of vesicles preparation per gram of tissue used. We did not use intact tissue or homogenate of whole muscle but rather, a constant amount of isolated vesicles. Therefore, an enhancement of expression would only be possible replacing in each vesicle proteins not involved in transport by Ca2+-ATPase. The second possibility is the rate of Ca2+ uptake measured is in fact the balance between two fluxes, the rate of Ca2+ pumping inside the vesicles and the rate of Ca2+ efflux. The enhancement of both the rate and steady state of Ca2+ accumulation may reflect a decrease of Ca2+ efflux of the vesicles prepared from rabbits exposed to a cold environment during a short period. This and other questions require a further experimentation that we intend to conduct during the future development of this project.
During cold adaptation, the rise of serum T3 is transient and much smaller than that found in hyperthyroid rabbits, but the changes in SERCA occur in the same direction. Interestingly, total PLB expression levels do not change in hyperthyroidism but decreases in short-term cold exposure. Altogether, these findings may be explained by the fact that the heart is exposed to high adrenergic stimulation during short-term cold exposure, which also activates SERCA activity and may differentially regulate total PLB. Mirit et al. (42) described an interesting finding that the hearts of rats acclimated to high temperatures (34°C) for 4 wk have a lower SERCA2a-to-PLB ratio and lower levels of serum T4 and T3, where this result is the opposite of what we observed during cold exposure.
The SR Ca2+-ATPase and the phosphorylated PLB fraction that increased during short-term response to cold returned to normal values after adaptation to the cold (Fig. 2). The enhancement paralleled a small increase in serum T3, and in hyperthyroidism, the same effect is observed after a substantial increase of T3. The cardiac hypertrophy observed in hyperthyroidism (Table 2) is primarily the result of increased work imposed on the heart through enhancement of hemodynamic performance (34). Previous studies have shown that chronic cold exposure (≥4 wk) promotes a significant increase in rat heart weight, which is accompanied by elevations of systolic, diastolic, and mean blood pressures (24, 59). Our results show that in rabbits, acclimatization at 4°C for 240 h does not promoted cardiac hypertrophy (Table 2). We cannot distinguish at present whether the difference between rats and rabbits is due to species-specific differences or to the time of cold exposure needed to promote the changes noted by Fregly et al. (24) and Wang et al. (59).
In previous reports, it was shown that the amount of heat released during ATP hydrolysis does not depend solely on the rate of ATP hydrolysis but also on the amount of heat produced during the hydrolysis of each ATP molecule (ΔHcal). The energy released during ATP hydrolysis can be converted to either work or heat. Under physiological conditions, only the SERCA1 of white skeletal muscle is able to modulate the ΔHcal of ATP hydrolysis (20, 21). Cardiac muscle expresses only SERCA2a. Hence, the extra heat produced by the heart during short-term cold exposure is mainly derived from both increased of Ca2+-ATPase activity and an enhancement of mitochondrial oxidative phosphorylation. This is probably related to the maintenance of body temperature during the initial 72 h of cold exposure.
Hypothyroidism causes cardiac atrophy and leads to impaired cardiac function (38). We confirmed the cardiac atrophy (Table 2) and found, in addition, that hypothyroidism promotes a decrease in the rates of heat production, oxygen consumption, Ca2+ uptake, and Ca2+-dependent ATP hydrolysis. The ΔHcal value of ATP hydrolysis does not vary, and the decrease of heat production is related to the diminished ATP hydrolysis rate, which is probably related to the higher amount of PLB that is responsible for SERCA inhibition.
In conclusion, both hyperthyroidism and short-term response to cold enhance cardiac SERCA activity through similar mechanisms. We also have shown that in all conditions tested, unlike red muscle, cardiac muscle expressed only SERCA2a and the ΔHcal value of cardiac muscle does not vary. Therefore, in the heart, the rate of heat production by SERCA2a depends only on the rate of ATP hydrolysis.
This work was supported by grants from PRONEX-Financiadora de Estudos e Projetos, Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), and Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ). L. A. Ketzer and A. P. Arruda are recipients of fellowships from CNPq and FAPERJ, respectively.
We are grateful to Valdecir A. Suzano and Antônio Carlos Miranda for technical assistance.
- Copyright © 2009 the American Physiological Society