Liver uptake of thyroxine (T4) is mediated by transporters and is rate limiting for hepatic 3,3′,5-triiodothyronine (T3) production. We investigated whether hepatic mRNA for T4transporters is regulated by thyroid state using Xenopus laevis oocytes as an expression system. Because X. laevis oocytes show high endogenous uptake of T4, T4 sulfamate (T4NS) was used as an alternative ligand for the hepatic T4 transporters. Oocytes were injected with 23 ng liver mRNA from euthyroid, hypothyroid, or hyperthyroid rats, and after 3–4 days uptake was determined by incubation of injected and uninjected oocytes for 1 h at 25°C or for 4 h at 18°C with 10 nM [125I]T4NS. Expression of type I deiodinase (D1), which is regulated by thyroid state, was studied in the oocytes as an internal control. Uptake of T4NS showed similar approximately fourfold increases after injection of liver mRNA from euthyroid, hypothyroid, or hyperthyroid rats. A similar lack of effect of thyroid state was observed using reverse T3 as ligand. In contrast, D1 activity induced by liver mRNA from hyperthyroid and hypothyroid rats in the oocytes was 2.4-fold higher and 2.7-fold lower, respectively, compared with euthyroid rats. Studies have shown that uptake of iodothyronines in rat liver is mediated in part by several organic anion transporters, such as the Na+/taurocholate-cotransporting polypeptide (rNTCP) and the Na-independent organic anion-transporting polypeptide (rOATP1). Therefore, the effects of thyroid state on rNTCP, rOATP1, and D1 mRNA levels in rat liver were also determined. Northern analysis showed no differences in rNTCP or rOATP1 mRNA levels between hyperthyroid and hypothyroid rats, whereas D1 mRNA levels varied widely as expected. These results suggest little effect of thyroid state on the levels of mRNA coding for T4 transporters in rat liver, including rNTCP and rOATP1. However, they do not exclude regulation of hepatic T4 transporters by thyroid hormone at the translational and posttranslational level.
- thyroid state
- plasma membrane
nonthyroidal illness (NTI) is any condition, not originating in the thyroid, resulting in a decreased peripheral 3,3′,5-triiodothyronine (T3) production (12). The serum thyroxine (T4) and thyrotropin (TSH) levels are normal or reduced, whereas serum 3,3′,5′-triiodothyronine (reverse T3) is usually elevated. Also surgery, fasting, and administration of certain drugs may cause this “low T3syndrome” (for reviews, see Refs. 12 and 13). Because serum T3 is one of the major factors that determine energy consumption and protein turnover of the body, low serum T3during NTI and fasting is seen by many authors as an adaptive mechanism to save energy and protein, i.e., organ function (13). If this is true, it is important for the body to maintain a lowered serum T3 until recovery from illness.
The main secretory product of the thyroid gland, T4, has little biological activity, whereas it is assumed that the type I deiodinase (D1) in the liver is a major site for the production of the active hormone T3 as well as for the degradation of the metabolite rT3. Therefore, the low serum T3 and the high serum rT3 during NTI and fasting have been ascribed at least in part to decreased hepatic D1 activity (23). During the last 20 yr, evidence has been collected that specific plasma membrane transporters mediate uptake of T4 and T3 in the different tissues. Most publications point to the Na+ gradient, temperature, and energy dependence of the transport mechanisms and their inhibition by thyroid hormone analogs (11, 13, 21-23, 27). Because the hepatic T3 production depends on the intracellular T4 concentration, regulation of T4 transport is potentially important in the determination of plasma T3production in addition to changes in D1 activity during pathophysiological circumstances (7, 13, 21-23).
In the last few years, we and others have shown that hepatic uptake of iodothyronines is mediated, at least in part, by the Na+-taurocholate-cotransporting polypeptide (NTCP) and Na+-independent organic anion-transporting polypeptides (OATPs; see Refs. 9-17). The uptake of iodothyronines, as well as their sulfamate and sulfate derivates, is stimulated in Xenopus laevis oocytes after injection with cRNA for rat (r) NTCP, human (h) NTCP, rOATP1, rOATP2, rOATP3, rOATP4, rOATP-E, hOATP-A, hOATP-C, hOATP-E, and hOATP8 (1, 5, 15, 16, 18,19, 24, 26, 31). However, studies also suggest the existence of a major Na+-dependent transporter in addition to the multispecific organic anion transporters (16). If regulation of thyroid hormone transport is important during pathophysiological conditions, it is likely that this transport is not regulated by multispecific transporters but at the level of as yet unidentified, specific thyroid hormone transporters. For this reason, we investigated the expression of thyroid hormone transporters inX. laevis oocytes after nanoinjection of rat liver mRNA (10). By use of this technique, it is possible to obtain an indication of the total mRNA levels of all thyroid hormone transporters contributing to the uptake of thyroid hormone without knowing their exact molecular structure (3, 40).
For the expression of thyroid hormone transporters, X. laevis oocytes have a major drawback in that they show high endogenous uptake of iodothyronines, which makes it difficult to quantify stimulation of uptake by injected mRNA (10). With the use of T4 sulfamate (T4NS) as a ligand, this high background is strongly reduced, since this analog is hardly taken up by native oocytes (17). Inhibition of T4NS uptake by T4 and stimulation of uptake of both ligands by the same rat liver 1.7- to 2.2-kb mRNA size fraction strongly suggests the involvement of a common transporter (10,17). Previous findings have also indicated that uptake of T4 and rT3 by rat hepatocytes is mediated by the same transporter (29). Therefore, the influence of thyroid state on the mRNA levels of thyroid hormone transporters in rat liver was evaluated in X. laevis oocytes with the use of T4NS and rT3 as the ligands. In addition, the levels of rNTCP and rOATP1 mRNA were estimated by Northern and dot-blot analysis in livers from euthyroid, hypothyroid, and hyperthyroid rats. Because D1 expression is known to be under positive control of T3 (3, 29), the D1 mRNA levels in these livers were also determined in oocytes and by Northern analysis as a positive control.
MATERIALS AND METHODS
[3′,5′-125I]T4, [3′,5′-125I]rT3, [32P]UTP, and [32P]dATP were obtained from Amersham Pharmacia (Uppsala, Sweden); T4, methimazole (MMI), dithiothreitol (DTT), BSA, and 3-aminobenzoic acid ethyl ester were from Sigma (St. Louis, MO); rT3 was from Henning (Berlin, Germany); Sephadex LH-20 was from Amersham Pharmacia; and collagenase B was from Roche (Mannheim, Germany). [125I]T4NS was prepared as previously described (35). [125I]T4NS, [125I]T4, and [125I]rT3 were purified immediately before use by Sephadex LH-20 chromatography (35). All other chemicals used in this study were of reagent grade.
Adult X. laevis females (2–3 yr old) were obtained from Hubrecht (Utrecht, The Netherlands). They were maintained in a water-filled tank with three dark sides at a temperature of 18–22°C. A 12:12-h light-dark cycle was maintained to reduce seasonal variations in oocyte quality. The frogs were fed two times a week, and water was changed 2 h after feeding.
Twelve male Wistar rats (weighing 350–500 g) were divided into three groups of four rats. All rats were provided with food and drinking water ad libitum. Rats were made hyperthyroid by daily intraperitoneal injections of 10 μg T4/100 g body wt. Hypothyroidism was induced by adding 0.05% (wt/vol) MMI to the drinking water. Euthyroid control rats received plain drinking water. After 4 wk, rats were anesthetized with ether. Blood was collected by cardiac puncture, and serum was stored at −20°C until analysis of T4, T3, and TSH. Livers were isolated, frozen immediately in liquid nitrogen, and stored at −80°C until isolation of RNA. All procedures were approved by the Animal Welfare Committee of the Erasmus University Rotterdam.
Rat liver mRNA was isolated using a commercial kit (Stratagene, La Jolla, CA) according to the manufacturer's protocol. Tissue was homogenized in guanidinium isothiocyanate buffer with β-mercaptoethanol. After dilution, precipitated proteins were removed by centrifugation, and the mRNA was bound to oligo(dT) cellulose. After several wash steps, mRNA was eluted, precipitated with 0.3 M sodium acetate in ethanol, dissolved in water, and stored at −80°C. mRNA concentration was estimated by measuring absorption at 260 nm, and purity was checked by the 260- to 280-nm ratio. To check that the mRNA was not degraded, it was analyzed by agarose gel electrophoresis. Total RNA was isolated using the TRIzol reagent (Invitrogen, Breda, The Netherlands) according to the manufacturer's protocol. Quality controls were as described above for mRNA.
cDNA coding for human D1 or human type III iodothyronine deiodinase (D3) was kindly provided by Dr. P. Reed Larsen (Thyroid Division, Brigham and Women's Hospital and Harvard Medical School, Boston, MA) and subcloned in pCI-neo (Promega Benelux, Leiden, The Netherlands). Capped D3 copy RNA (cRNA) was prepared from the cDNA clone with the Ampliscribe RNA transcription kit (Epicentre Technologies, Madison, WI). For capping, the m7G[5′]ppp[5′]G cap analog was used (Epicentre Technologies).
X. Laevis Oocyte Isolation and mRNA Injection
Oocytes were isolated and prepared as described previously (10). Healthy looking stage V-VI oocytes were sorted on morphological criteria, such as size, polarization, pigmentation, and absence of follicular layer debris (10). The next day, oocytes were injected with 23 ng of mRNA from hyperthyroid, hypothyroid, or euthyroid livers and/or with 2.3 ng of D3 cRNA by means of a nanoject system (Drummond Scientific, Broomall, PA). Control oocytes were not injected, yielding the same results as water-injected oocytes (10). Oocytes were maintained for 3 or 4 days at 18°C in modified Barth's solution [in mM: 88 NaCl, 1 KCl, 0.82 MgSO4, 0.4 CaCl2, 0.33 Ca(NO3)2, 2.4 NaHCO3, and 10 HEPES, pH 7.4, containing 10 IU/ml penicillin and 10 μg/ml streptomycin] with a daily change of medium. All oocyte experiments were done double blind. In each experiment, liver mRNA was tested from one euthyroid, one hypothyroid, and one hyperthyroid animal.
Uptake assays were done as described previously (10). Eight to 10 oocytes were incubated for 1 h at 25°C or for 4 h at 18°C with 10 nM [125I]T4NS or 2.5 nM [125I]rT3 in 0.1 ml incubation buffer (in mM: 100 NaCl, 2 KCl, 1 CaCl2, 1 MgCl2, 10 HEPES, and 10 Tris, pH 7.5). After removal of the medium, the oocytes were washed four times with 2.5 ml of ice-cold incubation buffer supplemented with 0.1% BSA. Oocytes were transferred to new tubes and counted individually.
Uptake and metabolism of T4 were studied by incubation of oocytes for 3 h at 18°C with 4 nM [125I]T4. After incubation, medium was collected, and oocytes were washed as described above and homogenized in 0.2 ml of 0.1 M NaOH. The medium and the extracts were fractionated on Sephadex LH-20, resulting in the separation of [125I]iodide, conjugates, and iodothyronines, as previously described (29).
Rat liver homogenates.
Liver tissue was homogenized on ice in 5 vol of 0.1 M phosphate (pH 7.2), 2 mM EDTA, and 1 mM DTT (P100E2D1). Protein concentration was measured by Bradford analysis (4). D1 activity was determined by incubation of 25 μg/ml homogenate protein for 30 min at 37°C with 1 μM [125I]rT3 in 0.2 ml P100E2D10. The reaction was stopped by addition of 0.1 ml 5% BSA on ice. Protein-bound iodothyronines were precipitated by addition of 0.5 ml of 10% TCA. After centrifugation, 125I−was determined in the supernatant by chromatography on Sephadex LH-20, as previously described (41).
Groups of 10 oocytes were homogenized on ice in 0.25 ml P100E2D10. Subsequently, D1 activity was determined by incubation of 50 μl of homogenate for 1 h at 37°C with 50 μl of 20 nM [125I]rT3 (final concentration 10 nM) in P100E2D10. Radioiodide production was quantified as described above for D1 in liver homogenates.
Table 1 shows the serum thyroid parameters and hepatic D1 activities in the groups of rats used in this study. The hypothyroid state of the MMI-treated rats was confirmed by the strongly decreased serum T4 and T3 levels and the strongly increased serum TSH levels. Conversely, the hyperthyroid state of the T4-treated animals was confirmed by the significantly increased T4 and T3 levels and the suppressed TSH levels. Corresponding changes in tissue thyroid state were demonstrated by the 3.2-fold decrease in hepatic D1 activity in the hypothyroid rats and the 2.5-fold increase in D1 activity in the hyperthyroid animals compared with the euthyroid controls.
To validate the X. laevis oocyte expression system to study thyroid state-dependent changes in hepatic mRNA levels for thyroid hormone transporters, we investigated the expression of D1 activity in oocytes injected with liver mRNA isolated from the different animals (Fig. 1). Uninjected oocytes showed negligible D1 activity. Vigorous D1 activity was expressed after injection of rat liver mRNA. The D1 activity induced by hypothyroid rat liver mRNA was 2.7-fold lower, whereas that induced by hyperthyroid rat liver mRNA was 2.4-fold higher than the D1 activity induced by mRNA from euthyroid rats. These changes are remarkably close to the changes in D1 activity observed in liver homogenates (Table 1), suggesting that the protein levels expressed in the oocytes reflect the hepatic mRNA levels.
Initially, studies of the expression of hepatic thyroid hormone transporters were carried out using T4 as the ligand. In agreement with previous studies (10, 15, 17), these experiments were hampered by the high endogenous uptake of T4 by uninjected oocytes. To investigate whether perhaps not all T4 uptake represented true internalization of the hormone and whether the latter could be determined more accurately, a parameter was sought reflecting intracellular T4. T4 is not metabolized by X. laevis oocytes, although rT3 is extensively sulfated (14). Therefore, we tested the metabolism of T4 in oocytes injected with cRNA coding for human D3. T4 internalized by the oocytes would thus be converted to rT3 and subsequently sulfated, which is easily detected by Sephadex LH-20 chromatography. Figure 2 shows the metabolism of T4 during incubation for 3 h at 18°C with uninjected oocytes or with oocytes injected with euthyroid rat liver mRNA, D3 cRNA alone, or both. Conjugates were hardly observed after incubation of T4 with uninjected or liver mRNA-injected oocytes. However, marked conjugation was observed in D3 cRNA-injected oocytes, showing that indeed T4 is extensively internalized by oocytes that have not been injected with liver mRNA. Coinjection of liver mRNA increased conjugation over that observed in oocytes injected with D3 cRNA alone (Fig. 2), but the difference was not large enough to be used as a parameter to assess thyroid state-dependent changes in hepatic mRNA expression for T4 transporter(s).
Previous studies in our laboratory have demonstrated that uninjected oocytes show lower uptake of rT3 than of T4 and negligible uptake of T4NS and that these analogs may be used as alternative ligands for the hepatic T4 transporter (10, 15, 17). Therefore, uptake of T4NS and rT3 was determined in uninjected oocytes and in oocytes injected with liver mRNA from euthyroid, hypothyroid, and hyperthyroid rats. The results of the incubations conducted for 1 h at 25°C are depicted in Fig. 3. T4NS uptake by native oocytes was very low and markedly stimulated after injection of liver mRNA. Uptake of T4NS by hyperthyroid rat liver mRNA-injected oocytes was insignificantly decreased, whereas uptake by hypothyroid rat liver mRNA-injected oocytes was not different from that in euthyroid controls. rT3 showed much higher endogenous uptake by uninjected oocytes, and the response to injection of rat liver mRNA was relatively much smaller compared with T4NS uptake. No differences in rT3 uptake were seen between oocytes injected with hypothyroid or hyperthyroid rat liver mRNA.
When rT3 and T4NS were incubated with uninjected and injected oocytes at 18°C, uptake was linear with incubation time for at least 6 h. Incubation for 4 h at 18°C resulted in rT3 and T4NS uptake values that were about two times as high as those observed after incubation for 1 h at 25°C. However, also under these conditions, differences in rT3 and T4NS uptake were negligible among oocytes injected with euthyroid, hypothyroid, or hyperthyroid rat liver mRNA (data not shown).
Studies have shown that uptake of iodothyronines by the rat liver is mediated in part by the multispecific transporters rNTCP and rOATP1 (15, 16). Therefore, in addition to the oocyte experiments, the effects of thyroid state on the hepatic mRNA levels coding for rNTCP and rOATP1 as well as the D1 mRNA levels were measured by Northern and dot-blot analysis (data not shown). We found no influence of thyroid state on the NTCP or OATP1 mRNA levels, although a slight downregulation of rOATP1 during hyperthyroidism could not be excluded. However, D1 mRNA levels were strongly decreased in hypothyroid rats and strongly increased in hyperthyroid rats, as expected.
Uptake of T4 by the liver is mediated by transporters and is rate limiting for subsequent metabolism (9, 21,23). Both in humans and in rats, the liver plays a major role in plasma T3 production (13). Previous experiments in our laboratory have shown that liver uptake of T3 is decreased in hyperthyroid rats (7). The aim of this study was to investigate whether thyroid state influences the mRNA levels for hepatic T4 transporter(s) by use of theX. laevis oocyte expression system.
Initially, we attempted to address this question by studying the uptake of T4 by oocytes injected with liver mRNA from euthyroid, hypothyroid, or hyperthyroid rats. These studies were hampered by the high endogenous uptake of T4 by oocytes, which makes it difficult to assess liver mRNA-induced T4 uptake (10). Our results show that this high endogenous uptake of T4 by oocytes indeed represents the internalization of the hormone. Incubation of T4 with D3 cRNA-injected oocytes resulted in the marked generation of conjugates because of the conversion of T4 to rT3 and the subsequent sulfation of this metabolite. Extensive sulfation of rT3, but not of T4, has been demonstrated in uninjected oocytes (14). Compared with total T4 uptake, conjugate formation by D3 cRNA-injected oocytes may be a more sensitive parameter reflecting liver mRNA-induced T4 transport. Indeed, coinjection of liver mRNA resulted in an increase in conjugate formation during incubation of the oocytes with T4, but the difference with oocytes injected with D3 cRNA alone was too small to be useful.
Uptake of rT3 by native X. laevis oocytes is significantly lower than that of T4 (15). Because there is evidence suggesting that hepatic uptake of T4 and rT3 is mediated by the same transporter in rats (29), we decided to test rT3 as an alternative ligand in our experiments. Injection of oocytes with euthyroid rat liver mRNA resulted in a slight increase in rT3 uptake. However, the signal was too small to draw firm conclusions from the differences in rT3 uptake by oocytes injected with liver mRNA from hypothyroid or hyperthyroid animals.
Compared with T4, uptake of T4NS by uninjected oocytes is negligible. The stimulation of T4NS and T4 uptake in oocytes by the same 1.8- to 2.2-kb rat liver mRNA size fraction as well as the inhibition of the induced T4NS uptake by T4 suggest that T4NS and T4 are ligands for the same transporter (10,17). This led us to use T4NS as an alternative ligand to study expression of the rat hepatic T4transporter in X. laevis oocytes. Uptake of T4NS was strongly stimulated after injection of oocytes with rat liver mRNA, but again negligible differences were seen in T4NS uptake among oocytes injected with mRNA from euthyroid, hypothyroid, or hyperthyroid rats. This lack of effect of thyroid state on hepatic mRNA levels coding for T4NS transporter(s) was not because of an insufficient degree of the hypothyroid or hyperthyroid state of the rats or to other experimental conditions, since large thyroid state-dependent differences in D1 expression were observed in the oocytes that perfectly matched the D1 activities in the rat liver homogenates. Although the duration of thyroid hormone treatment and of thyroid hormone withdrawal was sufficient to induce differences in D1 mRNA levels, it may not have been optimal to alter levels of mRNA coding for thyroid hormone transporters.
Evidence has been obtained that hepatic uptake of thyroid hormones is mediated in part by the organic anion transporters rNTCP and rOATP1 (15). However, the same studies also suggest the existence of a major, still unidentified Na+-dependent transporter in addition to the multispecific organic anion transporters (15,21). Quantitation of the hepatic rNTCP and rOATP1 levels by Northern and dot-blot analysis showed, in comparison with hepatic D1 levels, no obvious difference in expression of these transporters in the livers of euthyroid, hypothyroid, and hyperthyroid rats. These data are in apparent agreement with the results obtained in the oocyte experiments. This lack of effect of thyroid state on the expression of rOATP1 mRNA in rat liver contrasts with the suppression of the promoter function of hOATP-A in HepG2 cells by 1 μM T3(30). Whether this discrepancy is because of species differences in the regulation of the expression of these transporters or differences in experimental conditions remains to be investigated.
Previous observations with the use of isolated perfused rat liver indicate that hepatic uptake of T3 is decreased in hyperthyroid animals (7), whereas the present findings suggest no effect of hyperthyroidism on hepatic T4 uptake. A different regulation of hepatic uptake of T4 and T3 is consistent with the existence of separate transporters for T4 and T3 (22,29). Although rNTCP and rOATP1 may also mediate hepatic uptake of T3, rOATP1 preferentially transports T4 and, in particular, rT3 (15), suggesting that hepatic T3 is largely mediated by an as yet unidentified transporter.
The lack of regulation of hepatic T4 transport by thyroid state at the transcriptional level does not exclude regulation by translational or posttranslational mechanisms or by other factors. Previous studies suggested that factors other than protein synthesis are involved in the reduced uptake of T4 and T3by livers of starved rats (8, 25). A decrease in intracellular ATP levels, caused by the administration of fructose, results in a decreased hepatic T4 uptake in both humans and rats (6). A decreased cellular ATP content may also downregulate T4 transport during hyperthyroidism (28).
Hepatic uptake of thyroid hormone may also be under the control of cAMP, which has been shown to stimulate uptake of taurocholate in hepatocytes by stimulating translocation of rNTCP from the endosomes to the plasma membranes (37). Hepatic cAMP levels decrease during hypothyroidism and increase during hyperthyroidism (2,36). This would imply decreased rNTCP activity during hypothyroidism and increased activity during hyperthyroidism.
Besides intracellular factors such as ATP and cAMP, hepatic uptake of thyroid hormone is also regulated by certain substances in the serum. When rat hepatocytes are incubated with serum of uremic or nonuremic critically ill patients, or with 3-carboxy-4-methyl-5-propyl-2-furanpropanoic acid, indoxyl sulfate, bilirubin, or nonesterified fatty acids (NEFAs) at concentrations similar to their free concentrations in the serum of NTI patients, iodide production from T4 is reduced, which has been ascribed to a decrease in transport of T4 into the hepatocytes (32, 34). It has also been found that, during starvation, serum NEFAs are increased to concentrations that inhibit T4 transport into hepatocytes (33). The increased serum NEFA concentrations in hyperthyroid subjects may thus decrease hepatic T4 uptake, whereas the decreased serum NEFA levels in hypothyroid subjects may be associated with an increased T4 uptake (20, 38).
In conclusion, we found no evidence for an influence of thyroid state on the mRNA expression for T4 transporters in rat liver with the use of X. laevis oocytes as an expression system and T4NS and rT3 as ligands. Consistent data were obtained by Northern analysis of the mRNA levels of rNTCP and rOATP1, multispecific transporters that have been demonstrated to be involved in hepatic T4 uptake. If regulation of thyroid hormone transport is important during pathophysiological conditions, it is likely that this transport is not regulated at the level of those multispecific transporters but at the level of still unidentified, specific thyroid hormone transporters. The possible regulation of these specific transporters at the translational or posttranslational level should be explored in future studies, when these transporters are characterized.
We thank E. P. C. M. Moerings for help with the oocyte experiments and Dr. J. P. Sanders for help with Northern blotting. We are also grateful to Dr. P. Reed Larsen for the generous gift of D1 and D3 cDNA and to Drs. P. J. Meier, B. Hagenbuch, and B. Stieger for the generous gift of rNTCP and rOATP1.
Address for reprint requests and other correspondence: T. J. Visser, Dept. of Internal Medicine, Rm. Bd234, Erasmus Univ. Medical Center, PO Box 1738, 3000 DR Rotterdam, The Netherlands (E-mail:).
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.
August 20, 2002;10.1152/ajpendo.00214.2002
- Copyright © 2002 the American Physiological Society