Abnormal thyroid function is usually associated with altered cardiac function. Mutations in the thyroid hormone (TH)-binding region of the TH β-receptor (TRβ) that eliminate its TH-binding ability lead to the thyroid hormone resistance syndrome (RTH) in humans, which is characterized by high blood TH levels, goiter, hyperactivity, and tachycardia. Mice with “knock-in” mutations in the TH α-receptor (TRα) or TRβ that remove their TH-binding ability have been developed, and those with the mutated TRβ (TRβPV/PV) appear to provide a model for RTH. These two types of mutants show different effects on cerebral energy metabolism, e.g., negligible change in glucose utilization (CMRGlc) in TRβPV/PV mice and markedly reduced CMRGlc, like that found in cretinous rats, in the mice (TRαPV/+) with the knock-in mutation of the TRα gene. Studies in knockout mice have indicated that the TRα may also influence heart rate. Because mutations in both receptor genes appear to affect some parameters of cardiac function and because cardiac functional activity and energy metabolism are linked, we measured heart glucose utilization (HMRGlc) in both the TRβPV/PV and TRαPV/+ mutants. Compared with values in normal wild-type mice, HMRGlc was reduced (−77 to −95%) in TRαPV/+ mutants and increased (87 to 340%) in TRβPV/PV mutants, the degree depending on the region of the heart. Thus the TRαPV/+ and TRβPV/PV mutations lead, respectively, to opposite effects on energy metabolism in the heart that are consistent with the bradycardia seen in hypothyroidism and the tachycardia associated with hyperthyroidism and RTH.
- thyroid hormone resistance syndrome
- glucose metabolism
neonatal hypothyroidism impairs maturation in most tissues with major effects on brain, bone, and body growth. The heart is also a target of thyroid hormone (TH) actions during both early development and adulthood. Heart rate and cardiac output are decreased in hypothyroidism and elevated in hyperthyroidism (17). A number of molecular processes that have an impact on heart function are under TH control, e.g., various ion channels, genes coding for proteins involved in heart muscle contraction, and genes coding for a number of enzymes and proteins involved in a variety of regulatory processes (10, 17). Most of the effects of TH at the molecular level are mediated by TH nuclear receptors (TR) belonging to the v-erb A family. These ligand-dependent transcription factors are encoded by two genes, TRα and TRβ (23, 25). Several nuclear TR isoforms have been identified, TRα1, TRβ1, TRβ2, and TRβ3 (27), all of which are activated by TH, and also a splicing variant, TRα2, that is unable to bind TH (28). There are some other effects of TH, particularly in the heart, that occur too rapidly to be mediated at the gene level and are considered to be nongenomic effects (3).
Studies in mice with knockout mutations in the two major TH receptors, TRα and TRβ, have shown that mice lacking TRα1 have a slower heart rate, prolonged ventricular repolarization, lower body temperature, and mild hypothyroidism (26). Deletion of both TRα1 and TRβ results in a phenotype similar to that obtained with deletion of TRα1 alone, indicating that it is the TRα1 receptor that has the more important role in regulating heart rate (13).
A variety of mutations in the TH-binding region of TRβ have been described. These mutations lead to the syndrome of thyroid hormone resistance (RTH) (1, 2, 22). RTH is characterized by high levels of free TH in the serum, failure to suppress pituitary thyroid-stimulating hormone (TSH) secretion, and refractoriness to various TH actions in the peripheral tissues. RTH patients also exhibit goiter, hyperactivity, and tachycardia and may also manifest, in addition to the tachycardia, other cardiac effects characteristic of thyrotoxicosis, such as atrial fibrillation and heart failure. Complete deletion of either TH receptor has rarely been found in humans, but it also produces the RTH phenotype similar to that observed with TRβ mutations.
The present studies of cardiac glucose utilization were carried out in mice with targeted “knock-in” (PV) mutations of the α- or β-receptor (15, 16) in which cerebral glucose utilization was also simultaneously measured. The findings in the brain have already been reported (11). The PV-mutated β-receptor was derived from a patient with severe RTH who exhibited elevated blood TH and normal TSH levels, goiter, and tachycardia. The PV mutation is located in exon 10 and leads to complete loss of TH-binding activity. The phenotype of the TRβPV/PV knock-in mice is consistent with that seen in human patients with RTH (1, 2, 22). They exhibit severe dysfunction of the thyroid-pituitary axis, impaired hearing, retarded growth, delayed bone maturation, and abnormal patterns in the expression of TH-targeted genes, but generally normal cerebral energy metabolism. Previous studies in knockout mouse mutants have indicated that TRα may also have effects that influence heart rate (7, 13, 20, 26). We therefore also measured cardiac glucose utilization in knock-in TRα mutants in which the mutation in the TH-binding region leads to a receptor protein that, like that seen in the TRβPV mutants, is unable to bind TH (15, 16). Because homozygous mutation of the TRαPV mutant is lethal, we were limited in our studies to heterozygous TRαPV/+ mutant mice, which survive but exhibit increased mortality, infertility, dwarfism, and markedly reduced cerebral glucose utilization (11, 16).
Cardiac energy metabolism is normally closely correlated with cardiac function and work. Glucose, when it is adequately available in the blood, is one of the major substrates for that metabolism, and its utilization may then correlate with the total energy metabolism of the heart and cardiac work. We used the 2-[14C]deoxyglucose (2-[14C]DG) method, which was originally developed for determination of local rates of glucose utilization in the brain (24) and subsequently adapted for use in the heart (18), to measure local rates of cardiac glucose utilization (heart metabolic rate; HMRGlc) in both the TRβPV/PV and the TRαPV/+ mutant mice. Local rates were determined in several regions of the heart, because previous studies had shown regional heterogeneity in local rates of HMRGlc in euthyroid rats, e.g., the rate in the right ventricle being ∼50% lower than that in the left ventricle (18).
MATERIALS AND METHODS
Generation and characterization of mutant mice.
The mice with the targeted TRβ and TRα gene mutations were generated by knock-in mutations in the carboxy-terminal 14-amino acid sequence that contains the TH-binding site. These mutants produce defective TRαPV and TRβPV proteins that cannot bind TH (15, 16).
All procedures performed on animals were in strict accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals and were approved by the local Animal Care and Use Committee. Local rates of HMRGlc were determined in 4-wk-old wild-type (n = 7) and TRαPV/+ (n = 4) or TRβPV/PV (n = 5) mutant mice. The determinations were made in paired-sibling male wild-type and mutant mice, but the litters varied between studies. The animals were maintained on a 12:12-h light-dark cycle, with humidity and temperature controlled at normal levels, and allowed food and water ad libitum until studied. In preparation for the determination of HMRGlc, they were anesthetized with halothane (5% for induction, 1.0–1.5% for maintenance) in 70% N2O-30% O2, and polyethylene catheters (PE10; Clay-Adams, Parsippany, NJ) were inserted into the left femoral artery and vein. The skin incision was then sutured and treated with 5% lidocaine, and a loose-fitting plaster cast was fitted to the lower torso and pelvis and taped to a plastic box so as to allow movement but no locomotion. At least 3 h were allowed for recovery from the anesthesia and surgery before initiation of the procedure for measuring HMRGlc. Body temperature was maintained throughout the procedure by warming with a heating lamp and hand warmer.
Mean arterial blood pressure was measured with a Digi-Med Blood Pressure Analyzer (Micro-Med, Louisville, KY). Hematocrit was determined in arterial blood samples centrifuged in a Microfuge B (Beckman Instruments, Fullerton, CA). Arterial plasma glucose concentrations were measured in a Beckman Glucose Analyzer 2 (Beckman Instruments).
Determination of local rates of HMRGlc.
HMRGlc was determined by the quantitative autoradiographic 2-DG method (24) as adapted for use in the heart (18). The determination was initiated by an intravenous pulse of 2-deoxy-d-[1-14C]glucose (120 μCi/kg, specific activity 50–55 μCi/μmol; DuPont NEN, Boston, MA) followed by timed arterial blood sampling during the subsequent 45 min of the experimental period. Blood sample volumes were kept to a minimum to avoid excessive blood loss in such small animals. The blood samples were immediately centrifuged to separate the plasma from the red cells, and plasma glucose and 2-[14C]DG concentrations were measured in the Beckman Glucose Analyzer 2 and by liquid scintillation counting with external standardization (TRI-CARB model no. 2250CA; Packard Instruments, Downers Grove, IL), respectively. At ∼45 min after the pulse, the mice were injected with a lethal dose of pentobarbital, and the hearts were rapidly removed and frozen in isopentane maintained at −40 to −50°C with dry ice. The frozen hearts were sealed in plastic containers and kept in the freezer (−70°C) until they were cut into 20-μm-thick sections in a cryostat maintained at −22°C. The sections were thaw-mounted on glass coverslips, immediately dried on a hot plate at ∼60°C, and autoradiographed together with calibrated [14C]methylmethacrylate standards on Kodak EMC-1 X-ray film (Kodak, Rochester, NY). The autoradiograms were digitized in a Howtek MultiRAD 850 scanner (Howtek, Hudson, NH) and displayed on a computer monitor. Local tissue 14C concentrations were determined from the optical densities of the representations of the tissues in the autoradiographs and a calibration curve relating optical density to 14C concentration derived from the optical densities and the concentrations of the calibrated standards (24). Local HMRGlc was computed from the local tissue 14C concentrations and the time courses of the arterial plasma glucose and 2-[14C]DG concentrations according to the operational equation of the method and the program developed by G. Mies (Max Planck Institut für Neurologische Forschung, Köln, Germany) for use with the National Institutes of Health (NIH) image program (W. Rasband, National Institute of Mental Health, Bethesda, MD). Kinetic constants required by the equation were those previously determined in the rat (18, 24).
The statistical significance of the differences in values for the physiological variables and heart sizes between wild-type and mutant mice was evaluated by unpaired t-tests. Local rates of HMRGlc in the different regions of the heart were statistically compared by paired t-tests. Statistical significance of the differences in HMRGlc between wild-type and mutant mice was evaluated by ANOVA followed by unpaired t-tests. P values < 0.05 were considered statistically significant.
Physiological and morphological variables.
As previously observed (11), body weights of the TRαPV/+ mice were ∼41% lower and those of the TRβPV/PV mutants essentially the same as those of the wild-type mice (Table 1). Mean arterial blood pressure was significantly lower in the TRαPV/+ mice than in the wild-type and TRβPV/PV mice. Heart volume, determined from the sizes of the images of the heart sections in the autoradiograms, was ∼31% lower (P < 0.01) in TRα1PV/+ mice than in wild-type mice, a decrease somewhat less than the 41% in total body weight. Heart size in the TRβPV/PV mice did not differ in a statistically significant way from that of the wild-type mice (Fig. 1).
HMRGlc in different regions of the heart.
In the wild-type mice, HMRGlc (mean ± SE) in the lateral wall of the left ventricle (105 ± 18 μmol·100 g−1·min−1, n = 7) was markedly higher than in the right ventricular muscle (28 ± 6 μmol·100 g−1·min−1, n = 7) (P < 0.002) and apex (54 ± 15 μmol·100 g−1·min−1, n = 7) (P < 0.001), but it did not differ in a statistically significant way from the rates in the papillary muscle (123 ± 25 μmol·100 g−1·min−1, n = 7) and septum (103 ± 24 μmol·100 g−1·min−1, n = 7) (Figs. 2 and 3). Similar but lesser differences in local HMRGlc among the various regions of the heart were observed previously in normal rats (18).
HMRGlc was statistically significantly lower (−77 to −95%) in all regions of the heart in TRα1PV/+ mice than in wild-type mice (Fig. 3). In contrast, HMRGlc was markedly and statistically significantly higher (+87 to +340%) in all regions of the heart in TRβPV/PV mice than in wild-type mice.
The present studies demonstrate that the influences of TH on morphology, function, and energy metabolism of the heart are differentially mediated by the TRα and TRβ TH receptors. Mutations of the TRα and TRβ genes in their TH-binding domain that remove their ability to bind TH result in contrasting effects on cardiac size and energy metabolism. Compared with wild-type mice, the TRαPV/+ mutant mice exhibited a moderate reduction in heart size but a profound decrease in cardiac glucose utilization, whereas heart size was unchanged and HMRGlc markedly increased in the TRβPV/PV mutants. Probably because they are functionally related and can influence one another, both mean arterial blood pressure and heart size were lower by about the same degree (approximately −30%) in the TRαPV/+ mice, and both were essentially unchanged in the TRβPV/PV mice compared with their values in the wild-type mice. Body size and weight were also very much reduced in the TRαPV/+ mutants but normal in the TRβPV/PV mice. TH is known to lower systemic vascular resistance and to raise blood volume, cardiac output, and blood pressure, whereas hypothyroidism has opposite effects (17). The decreases in blood pressure, heart size, and glucose utilization in the TRαPV/+ mutant are thus consistent with the reductions in heart size and cardiac work seen in hypothyroidism. In contrast, the normal heart size but large increase in cardiac glucose utilization in the TRβPV/PV mice are more consistent with hyperthyroidism than hypothyroidism.
Energy metabolism in the heart is related to its level of physical work. In brain, glucose is normally an obligatory and almost the sole substrate for energy metabolism. The heart, however, is more facultative in its choice of substrates and can use, in addition to glucose, other substrates, e.g., lactate, fatty acids, and so forth, more or less in proportion to their relative levels in the blood. To ensure, therefore, that blood glucose levels were at least maintained at normal or higher levels, the mice were allowed full access to food and water right up to the time of the experiments. This resulted in blood glucose concentrations in the wild-type and mutant mice that were in the high normal range and, furthermore, not statistically significantly different from one another. Blood levels and rates of utilization of other potential substrates were not, however, measured in the present studies, and their role in the altered HMRGlc cannot, therefore, be excluded. There is, however, another reason to believe that the reduced HMRGlc observed in the TRαPV/+ mice is due to a reduction in total cardiac energy metabolism and not merely a switch from glucose utilization to that of other potential substrates. Total body weight was disproportionately decreased more than heart size (i.e., −41 and −31%, respectively), which is consistent with reduced demand for cardiac output, cardiac work, and, therefore, energy demand per unit weight of heart tissue.
The striking reduction in HMRGlc in the TRαPV/+ mutants was distributed throughout all regions of the heart, e.g., those that perform high (e.g., left ventricles and papillary muscles) and low (e.g., right ventricles) levels of physical work. Reduced HMRGlc in the TRαPV/+ mutants appears to be consistent with the alterations in cardiac function, e.g., bradycardia and reduced cardiac output, found in hypothyroidism (17). TRα-knockout mice have also been reported to have an ∼20% decrease in heart rate (8, 13, 26), but further work is clearly required to compare the effects of the knockout and knock-in mutations on cardiac metabolism. There is an important difference between knockout and knock-in mutants to be considered. In the TRα-knockout mice, there is no α-receptor protein produced, whereas in the knock-in TRα mutant, a receptor protein lacking the ability to bind TH is produced. It is conceivable that the decreased cardiac glucose metabolism in the TRα mutant might result from a negative effect of this abnormal receptor.
In contrast to the decreases found in the TRαPV/+ mutants, HMRGlc was markedly elevated in the TRβPV/PV mice, despite the lack of any overt evidence of altered cardiac function. There were no changes in arterial blood pressure, hematocrit, or heart size that might have indicated conditions that could alter cardiac work. In any case, the magnitudes of the reductions in HMRGlc in the TRαPV/+ mutants and the increases in the TRβPV/PV mutants are probably too great to be explained entirely by alterations in cardiac work, and intrinsic alterations in energy metabolism and/or energy demand must be considered.
Large amounts of energy are utilized in the heart by a number of ATP-consuming processes regulated by TH-dependent gene expression. Among these are 1) the ATPase activity associated with the actin-myosin complex (i.e., the fetal and adult myosin heavy chains) (10, 19, 20, 21); 2) the SERCA2 ATPase (9), which consumes ATP in the transport of Ca2+ from the sarcoplasmic reticulum to the cytosol, where it stimulates heart muscle contraction; and 3) the Na+-K+-ATPase (7), which consumes large amounts of ATP to restore membrane potentials after depolarization, as well as a number of other enzymes (17). The role that the TRα receptor may play in the mediation of the actions of TH on the expression of these enzymes remains, however, to be determined.
Increased blood levels of TH might explain the increased HMRGlc found in the TRβPV/PV mutants. TRβPV/PV protein is unable to bind and, therefore, to mediate the actions of TH in all tissues, including the pituitary (15, 29), where TH normally has an inhibitory effect on the release of TSH. Because of the lack of this inhibition, circulating levels of TSH and TH are consequently elevated in the TRβPV/PV mutant. TRα is the dominant TH receptor in the heart, e.g., threefold higher than the TRβ receptor (6), and is functionally normal in the TRβPV/PV mice. It may therefore be mediating the effects of the increased circulating TH to produce a thyrotoxic response in the heart, manifested by the increased heart rate previously observed in TRβ-deficient knockout mutant mice (13) and the increased HMRGlc observed in the TRβPV/PV mice in the present study.
The effects of the TRβPV/PV mutation on glucose utilization in the heart are also consistent with several features of the RTH mutation in human patients, i.e., high blood levels of THs but normal levels of TSH that are rather high for the elevated blood TH levels, goiter, and tachycardia (22). Heart rate, stroke volume, cardiac output, and systolic aortic flow velocity are all considerably greater in RTH than in euthyroid and hypothyroid human subjects (14). Some other parameters of heart function are less affected in RTH than in hyperthyroid patients, suggesting an incomplete thyroid hormone response in RTH (14). It remains to be determined whether cardiac glucose utilization is also increased in humans with RTH.
We thank Daniel Glen for assistance and critical contributions to the computerized analyses of the autoradiograms and the preparation of the illustrations, particularly the quantitative color-coded autoradiograms of local rates of glucose utilization in various regions of the heart.
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. Section 1734 solely to indicate this fact.
- Copyright © 2004 by American Physiological Society