Skip to main content
NCBI home page
Endocrinology logoLink to Endocrinology
. 2010 Oct 6;151(12):5961–5970. doi: 10.1210/en.2010-0553

Inhibition of the Type 2 Iodothyronine Deiodinase Underlies the Elevated Plasma TSH Associated with Amiodarone Treatment

Matthew L Rosene 1, Gábor Wittmann 1, Rafael Arrojo e Drigo 1, Praful S Singru 1, Ronald M Lechan 1, Antonio C Bianco 1
PMCID: PMC2999495  PMID: 20926587

Abstract

The widely prescribed cardiac antiarrhythmic drug amiodarone (AMIO) and its main metabolite, desethylamiodarone (DEA), have multiple side effects on thyroid economy, including an elevation in serum TSH levels. To study the AMIO effect on TSH, mice with targeted disruption of the type 2 deiodinase gene (D2KO) were treated with 80 mg/kg AMIO for 4 wk. Only wild-type (WT) mice controls developed the expected approximate twofold rise in plasma TSH, illustrating a critical role for D2 in this mechanism. A disruption in the D2 pathway caused by AMIO could interfere with the transduction of the T4 signal, generating less T3 and softening the TSH feedback mechanism. When added directly to sonicates of HEK-293 cells transiently expressing D2, both AMIO and DEA behaved as noncompetitive inhibitors of D2 [IC(50) of >100 μm and ∼5 μm, respectively]. Accordingly, D2 activity was significantly decreased in the median eminence and anterior pituitary sonicates of AMIO-treated mice. However, the underlying effect on TSH is likely to be at the pituitary gland given that in AMIO-treated mice the paraventricular TRH mRNA levels (which are negatively regulated by D2-generated T3) were decreased. In contrast, AMIO and DEA both exhibited dose-dependent inhibition of D2 activity and elevation of TSH secretion in intact TαT1 cells, a pituitary thyrotroph cell line used to model the TSH feedback mechanism. In conclusion, AMIO and DEA are noncompetitive inhibitors of D2, with DEA being much more potent, and this inhibition at the level of the pituitary gland contributes to the rise in TSH seen in patients taking AMIO.

The widely prescribed cardiac antiarrhythmic drug amiodarone disrupts the TSH feedback mechanism by noncompetitive inhibition of the type 2 deiodinase-mediated thyroid hormone activation in the thyrotrophs.

Amiodarone (AMIO) is a widely prescribed antiarrthymic drug used in the treatment of cardiac arrhythmias, including ventricular tachycardia, ventricular fibrillation, paroxysmal supraventricular tachycardia, atrial fibrillation, and flutter (1). However, as a side effect, patients placed on AMIO exhibit important alterations in plasma TSH, thyroxine (T4), and 3,5,3′-triiodothyronine (T3) concentrations through mechanisms that have yet to be fully understood. The high content of iodine in AMIO is one important factor influencing the function of the thyroid gland. However, this is insufficient to explain the spectrum of thyroid test abnormalities seen in patients started on this drug. It is unclear whether AMIO alone, or its main metabolite desethylamiodarone (DEA) plays the more dominant role in changing thyroid economy. DEA is produced via metabolism of AMIO by the cytochrome P4503A (CYP3A), and both compounds exhibit long half-lives, between 40–58 d and 36–61 d, respectively (2,3). These long half-lives stem from drug accumulation in various tissues and organs, including adipose tissue, liver, lungs, and to a lesser extent, kidneys, heart, skeletal muscle, thyroid, and brain (4).

The chemical structures of AMIO and DEA are very similar to T3, and some of its effects have been attributed to inhibition of thyroid hormone transport across the plasma membrane (5), and/or direct binding to the thyroid hormone receptors, TRα and TRβ (6,7), and possibly even TR-dependent gene transcription (8). However, it is widely accepted that the effects of AMIO on thyroid hormone plasma concentrations are at least in part due to interference with the iodothyronine deiodinases, which metabolize thyroid hormones (9). These are enzymes that can activate (outer ring deiodination, ORD) or inactivate (inner ring deiodination, IRD) thyroid hormone via sequential removal of iodine moieties (10). For example, T4 is activated to T3 via ORD; T4 and T3 are both inactivated by IRD to rT3 and T2, respectively. The type 1 deiodinase (D1) catalyzes both ORD and IRD and, in humans, it plays a secondary role in determining plasma T3 levels. The type 2 deiodinase (D2) catalyzes ORD and is thought to be the major source of plasma T3 in humans and also plays a critical role as a source of intracellular T3 in a number of cell types. Lastly, type 3 deiodinase (D3) is restricted to IRD and terminates thyroid hormone action in the brain, placenta and fetal tissues.

Multiple studies indicate that D1 activity is decreased in homogenates of liver, heart, and kidney of animals treated with AMIO, in a dose-dependent fashion, but details are missing regarding the mechanism by which this is achieved (4,11,12,13,14,15,16,17,18). Similar findings have been observed in cells, for example in hepatocytes exposed to AMIO (11). It has been suggested that AMIO and/or DEA inhibit D1 directly via a competitive mechanism (19), which is in agreement with the observation that D1 mRNA levels are not affected by AMIO treatment (20). Much less is known about the effects of AMIO on D2 activity. Given the fundamental role played by D2 in determining the plasma levels of TSH (21) and TSH releasing hormone (TRH) (22) secretion, as well as plasma T3 (23), it is conceivable that a substantial component of the AMIO effects on thyroid economy are related to an inhibition of this enzyme. In a previous study, AMIO inhibited D2 activity in human skin cells by about 33% (24) but when added to pituitary sonicates at high concentrations the decrease in D2 activity was only 13% (25). Decreased ORD in pituitary sonicates of AMIO-treated rats has been reported, however that study did not specifically measure D2 activity (26).

Patients and experimental animals that have been placed on AMIO tend to have a transient elevation in serum T4 and TSH concentrations that lasts for several months (27,28,29). This is puzzling given that T4 acts in a negative feedback loop, via D2 activity in the pituitary and hypothalamus, to decrease TSH and TRH production, respectively. This phenotype resembles that of the D2 knockout mouse (D2KO) (30) and it is thus conceivable that by inhibiting D2 in the hypothalamic median eminence tanycytes and/or pituitary thyrotropes, AMIO and/or DEA decrease local thyroid hormone signaling, which leads to increased TRH/TSH secretion. Here we show that D2 is critical in the AMIO-induced increase in plasma TSH by virtue of being a noncompetitive inhibitor of D2. The elevation in plasma TSH in AMIO-treated animals depends on a functional D2 and, most likely, stems from D2 inhibition at the level of the pituitary gland.

Materials and Methods

Materials

Unless otherwise specified all materials were obtained from Sigma-Aldrich (St. Louis, MO). DEA was a kind gift of Dr. H. C. van Beeren (Academisch Medisch Centrum, Amsterdam, The Netherlands).

Animals

All studies were performed according to a protocol approved by the Animal Care and Use Committee of University of Miami Miller School of Medicine. D2KO mice were backcrossed for 10 generations into C57BL/6J mice from our colony and age matched with wild-type C57BL/6J mice purchased from Jackson Laboratories (Bar Harbor, ME). All mice were maintained on normal maintenance chow and housed under a 12-h light, 12-h dark cycle at 22 C.

Animal experiments

Wild-type and D2KO mice were ip injected with either vehicle (VH) or 80 μg/g of AMIO for 25 d with 12 mice in each group. A new stock of 1000 × AMIO was prepared every 7 d in 95% ethanol and diluted fresh everyday in sterile PBS. At the end of 25 d, the mice where anesthetized with 200 μl of a ketamine-xylezene mixture (16.7 mg/ml ketamine and 3.3 mg/ml xylezene). Blood was obtained via cardiac puncture with 1cc insulin syringes (BD, San Jose, CA) and placed into EDTA coated Microtainer Tubes (BD, San Jose, CA). Six mice from each group were perfused for in situ hybridizations. Perfusions were carried out after removing some blood by cardiac puncture and the passing through heart of about 27 ml DEPc treated PBS (Invitrogen, Carlsbad, CA) with 15000 U/liter heparin followed by about 45 ml of 4% PFA (Electron Microscopy Sciences, Hatfield, PA) prepared in DEPc treated PBS. After perfusions, the brains were removed and placed in 4% PFA for 24 h and then stored in 25% sucrose solution. The remaining mice were anesthetized, blood obtained via cardiac puncture and then brain and pituitary tissues were harvested and snap frozen in liquid nitrogen for deiodinase assays.

In situ hybridizations and analysis

In situ hybridization histochemistry was performed on every fourth section through the paraventricular nucleus (PVN) using a 741 base single stranded [35S]UTP labeled cRNA probe for mouse TRH (31) following methods previously described (32). In vitro transcription was performed using SP6/T7 systems (Promega Corporation, Madison, WI) and [35S] α-UTP (1250 Ci/mmol; PerkinElmer, Waltham, MA). The hybridization was performed under plastic coverslips in a buffer containing 50% formamide, a twofold concentration of standard sodium citrate (2XSSC), 10% dextran sulfate, 0.5% sodium dodecyl sulfate, 250 μg/ml denatured salmon sperm DNA, and 5 × 105 cpm of radiolabeled probe for 16 h at 56 C. Slides were dipped into Kodak NTB autoradiography emulsion (Eastman Kodak, Rochester, NY) diluted 1:1 in distilled water, and the autoradiograms developed after 3 d of exposure at 4 C.

Autoradiograms were visualized under darkfield illumination using a COHU 4910 video camera (COHU, Inc., San Diego, CA). The images were captured with a color PCI frame grabber board (Scion Corporation, Frederick, MD) and analyzed with a Macintosh G4 computer using Scion Image. Background density points were removed by thresholding the image and integrated density values (density × area) of hybridized neurons on each side of the PVN measured in three consecutive sections of the midportion of the PVN where hypophysiotropic TRH neurons have been identified in the mouse (31).

Plasma hormone levels

Blood was spun down at 16,000 × g and plasma processed for measurement of T3, T4, and TSH using a Rat Thyroid Hormone Luminex kit from (Millipore, Billerica, MA) as quantified on a BioPlex (Bio-Rad, Hercules, CA). Mouse TSH standards were prepared by making dilutions of plasma from hypothyroid mice (treated for 10 d with 0.1 sodium perchlorate and 0.05% methimazole) or hyperthyroid mice (injected with 80 mg/kg T4 for 10 d). In several assays, both rat and mouse curves were parallel and separated by a factor of approximately five.

Tissue deiodinase assay

For measurement of tissue deiodinase activity (D1 and D2), tissues were powderized using the BioPulverizer (BioSpec Bartlesville, OK) and homogenized in 0.25 m sucrose, 1 mm or 10 mm DTT in PE buffer (DAB) using a Tissue Tearorizer (BioSpec Bartlesville, OK). Protein samples were sonicated for no more than 5 s and quantified with the Bradford method as described by the manufacturer (Bio-Rad, Hercules, CA). Radiolabeled substrates were purified on LH-40 column through Sephadex within 48 h of assay. D1 assays were carried out using 15 μg of liver or pituitary sonicate with a master mix that included 0.5 μm reverse Triiodothyronine (rT3), 10 mm DTT (Calbiochem, Gibbstown, NJ), and 150,000 cpm 125I rT3 (Perkin-Elmer, Boston, MA) for 1 h at 37 C. D2 assays were performed using 20 μg of pituitary or median eminence sonicate with a master mix that included 0.1 nm T4, 1 mm or 20 mm DTT, and 150,000 cpm 125I T4 (Perkin-Elmer, Boston, MA) for 3 h at 37 C. Assays were stopped as previously described with horse serum and 50% TCA (21) and free 125I counted on the 2470 Automatic γ Counter Wizard2 (Perkin-Elmer, Boston, MA). As indicated in the legend to the figures, UPLC (ACQUITY, Waters Corporation, Milford, MA) was used to validate that equimolar amounts of iodide and T3 were produced during the assay reaction (33).

Cell transfection and protein preparation

HEK293 cells were grown in DMEM (Invitrogen Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS), 292 mg/liter L-glutamine (Invitrogen Carlsbad, CA), 50 μg/ml ampicillin, 15 μg/ml gentamycin, 3.7 g/liter sodium bicarbonate, and 10 nm sodium selenite and transfected with mouse wild-type D1 and D2 using the Lipofectamine2000 (Invitrogen, Carlsbad, CA) as the manufacturer describes. Cells were then harvested in PBS and centrifuged into a pellet at 16,000 × g. The supernatant was removed and cells were resuspended in 500 μl of DAB and sonicated for no more than 5 sec. Cell sonicates were mixed and protein was quantified using the Bradford method.

D1 and D2 sonicated assays

Protein from transfected and nontransfected HEK cells was prepared into a protein mix of protein and DAB for final protein concentrations of 2 μg and 10 μg for rat D1 and human D2 assays respectively. A mix of DAB and cold thyroid hormones rT3 (5–50 μm) or Thyroxine (T4) (1–10 nm) for D1 and D2 assays respectively were prepared and added to the protein samples. A mix of 150 × drug stocks in 95% EtOH and DAB was prepared for each protein sample for final concentrations of 10–200 μm AMIO and 250 nm-80 μm DEA. D1 assay master mixes were prepared for final concentrations of 10 mm DTT, 125,000 cpm 125I rT3 was finally added last for a final volume of 300 μl. D2 assay master mixes contained final concentrations of 20 mm DTT, 1 mm PTU to inhibit any D1 activity, and 125,000 cpm 125I T4. Samples were incubated at 37 C for 1 h and the reaction was stopped by adding 200 μl of Horse Serum and 100 μl of 50% TCA. The samples were then mixed in a multitube vortex two times for 1 min each and spun down at 21,000 × g for 3.5 min. Deiodinase activity was measured by counting the free 125I in the supernatants. Untransfected HEK cell sonicates were used for blanks. UPLC separation of iodothyronine products was used to verify that production of 125I and 125I-T3 is equimolar.

Cell culture

TαT1 cells, kindly provided by Dr. E.C. Ridgway, were grown in 10% FBS DMEM plus supplements. All cells were grown at 37 C with 5% CO2. All plastics were treated with a 1:30 dilution of Matrigel (BD, San Jose, CA) without growth hormone in PBS for at least 1 d before plating. When cells were confluent, experiments were carried out in DMEM media supplemented as described above but with 0.1% BSA instead of 10% FBS to make cells hypothyroid as well as to calculate the amount of free T4 in the media. As described previously (21), the free fraction of T4 in 0.1% BSA is 2.7% the total T4. Cells were treated with VH, AMIO or DEA (1 nm–20 μm).

D2 activity and TSH secretion in TαT1 cells

Confluent TαT1 cells were treated for 24 h with 0.1% BSA DMEM to make them hypothyroid and then exposed to 200,000 cpm/ml 125I-T4 in the same media containing either VH or final drug concentrations of 10 nm-10 μm AMIO or DEA. After 24 h, 300 μl of media was removed in duplicate and T4 to T3 conversion stopped with 200 μl horse serum and 100 μl 50% TCA. Radiometry was performed in the Wizard and the conversion rate calculated as previously described. The remaining hot media was then removed from the cells, they were washed with PBS, and then harvested in PBS. Cells were centrifuged into a pellet, and the supernatant was removed. Pellets were resuspended in DAB, sonicated, and protein quantified by the Bradford method. In vivo conversion was normalized to total protein per dish, and cell sonicates were then used for a D2 assay as described above without any additional drugs added. In a separate experiment, similarly prepared TαT1 cells were made hypothyroid in 0.1% BSA and then exposed to increasing concentrations of either AMIO or DEA. Thirty minutes later T4 was added to all cells at 40 pm free T4 concentration, and 24 h later media was collected for TSH determination using the Luminex kit.

Statistical analysis

Data were analyzed using PRISM software (GraphPad Software, Inc, San Diego, CA) and expressed as mean ± sem. A two-tailed Student’s t test or one-way ANOVA with Newman-Keuls Multiple Comparison test (or Dunnett’s Multiple Comparison test, where indicated) were used to compare means between groups.

Results

D2KO mice are not sensitive to AMIO-induced elevation in plasma TSH

To test the hypothesis that D2 is part of the mechanism that mediates an elevation in plasma TSH during treatment with AMIO, WT and D2KO mice were treated with either VH or AMIO (80 mg/kg BW) for 25 d. No significant changes in body weight were observed in WT or D2KO mice as a result of the treatment (ΔBody Weight: 1.1 ± 2.2 vs. 0.5 ± 1.7 g, respectively). In WT animals, this regimen only promoted mild alterations in thyroid hormone economy, with doubling of plasma TSH concentrations (P < 0.05) in the face of steady plasma T4 and T3 levels (Fig. 1). As expected, D2KO mice exhibited higher plasma TSH and T4 levels when compared with WT animals (Fig. 1). Notably, upon treatment with AMIO, the D2KO animals did not develop further increases in plasma TSH, revealing a critical role of D2 in this mechanism. Plasma T4 and T3 remained unaffected in these AMIO-treated animals (Fig. 1).

Figure 1.

Figure 1

Open in a new tab

Plasma T3, T4, and TSH levels of WT (open bars) and D2KO mice (filled bars) treated with VH or 80 mg/kg AMIO for 25 d. Data are plotted as mean ± sem with groups of 5–6. A, Plasma T4 levels. B, Plasma T3 levels. C, Plasma TSH levels. Groups of animals and treatments are indicated. *, P < 0.05 and **, P < 0.01 vs. the indicated control groups.

To find out whether the AMIO effect on plasma TSH is mediated at the pituitary level and/or the hypothalamus, TRH mRNA levels were measured by in situ hybridization analysis in the PVN of the AMIO-treated mice. In the AMIO-treated WT animals, there was a significant reduction (∼55%; P < 0.05) in TRH mRNA levels (Fig. 2), possibly as a result of AMIO acting as a partial TR agonist (6,7) or interference with membrane transport of thyroid hormone. These findings indicate that changes at the hypothalamus are unlikely to underlie the elevated plasma TSH. Remarkably, PVN TRH mRNA levels were not elevated in the D2KO mice (Fig. 2), indicating that the elevated serum TSH levels found in these animals are due to a specific lack of D2 in the pituitary gland and not mediated by an elevation in TRH expression. In the AMIO-treated D2KO animals, there was a tendency for TRH mRNA levels to be reduced but the difference did not reach statistical significance (Fig. 2).

Figure 2.

Figure 2

Open in a new tab

In situ hybridization of TRH mRNA in the PVN of AMIO-treated WT and D2KO mice as indicated in the legend to Figure 1. A, Coronal section of a VH-treated WT mouse brain showing the medial basal hypothalamus and the III ventricle. B, Same as in A, except that animals were treated with AMIO. C and D, VH- and AMIO-treated D2KO mice, respectively. E, Diagram representing computerized image analysis of the in situ hybridization autoradiograms (six animals per group); values are the mean ± sem and n = 5–6; *, P < 0.01 vs. the indicated control group.

AMIO is a noncompetitive inhibitor of D2

The hypothesis that AMIO triggers some of its effects on thyroid economy via inhibition of D2 activity was further tested using sonicates of HEK-293 cells transiently expressing D2. These cell sonicates were processed for deiodinase assay in the presence of AMIO or its main derivative, DEA. Assays were performed in the presence of increasing concentrations of either drug (10 nm–100 μm) and substrate (T4; 1–10 nm), while the DTT concentration was kept constant at 20 mm. Data reduction and analysis using the Lineweaver-Burk plotting indicate that AMIO functions as a weak noncompetitive inhibitor of D2 (Ki >100 μm) (Fig. 3A), with the caveat that inhibition did not even reach 50% at the maximum AMIO concentration used (Fig. 3B). DEA is a much more potent D2 inhibitor, also functioning as a noncompetitive inhibitor with an apparent Ki of about 3.1 μm (Fig. 3C); the IC (50) value for DEA is approximately 7 μm (Fig. 3D). For comparison, typical serum concentration of these drugs is about 3 μm for AMIO and 2 μm for DEA (3,34).

Figure 3.

Figure 3

Open in a new tab

In vitro inhibition of D2 by AMIO or DEA. Both drugs were added to sonicates of HEK cells transiently expression human D2 enzyme. A, AMIO is a weak noncompetitive inhibitor of D2 as seen in a Lineweaver-Burk plot. B, AMIO-induced D2 inhibition. C, Lineweaver-Burk plot of DEA-induced D2 inhibition, also noncompetitive in nature. D, The IC (50) of DEA on D2 activity is calculated to ∼3 μm DEA; values are the mean ± sem and n = 5–6. UPLC was used to validate that equimolar amounts of iodide and T3 were produced during the assay reaction.

As a comparison, similar studies were performed using sonicates of HEK-293 cells transiently expressing D1 (Fig. 4). These studies indicate a kinetic inhibitory pattern compatible with mild noncompetitive inhibition by AMIO (Fig. 4A; Ki>100 μm), calculated based on assays in which AMIO did not inhibit enzyme activity to the 50% level (Fig. 4B). In contrast, DEA markedly inhibited D1 (Ki ∼1.6 μm) (Fig. 4C), with an IC (50) value for DEA at about 10 μm (Fig. 4D).

Figure 4.

Figure 4

Open in a new tab

In vitro inhibition of D1 by AMIO or DEA. Both drugs were added to sonicates of HEK cells transiently expressing human D1 enzyme. A, B, and C are as in the legend to Figure 3. The IC (50) of DEA on D1 activity is calculated to ∼10 μm DEA (D); values are the mean ± sem and n = 5–6. UPLC was used to validate that equimolar amounts of iodide and T3 were produced during the assay reaction.

We next assayed tissues of the AMIO-treated mice for deiodinase activities. This included measurement of D2 in the pituitary and median eminence and D1 activity in the liver and pituitary gland. Notably, no deiodinase inhibition was detected in these tissue samples (Table 1). The experiment was repeated and included different time points (1–4 weeks), with similar results (Table 1). To explore the possibility that this was due to the weak nature of the AMIO inhibition on deiodinase activity that is dissipated once tissue sonicates are diluted in assay buffer and processed in vitro with massive concentration of the cofactor (i.e., 10–20 mm DTT), we repeated these experiments (administration of AMIO for 3 weeks) and processed all tissues with only 1 mm DTT. Although deiodinase velocity was substantially less across all groups, it is now clear that D1 and D2 activities are significantly decreased in the tissues of AMIO-treated mice (Table 1). The decrease in liver D1 activity was 20–30%, whereas in the pituitary and median eminence the decrease in D2 activity was 40–50% (Table 1).

Table 1.

Deiodinase activities in tissues of AMIO-treated mice: effect of treatment time and DTT content in assay buffer

Liver D1 Pituitary D1 Pituitary D2 ME D2
Experiment 1
 7 days
 WT VH n. d. n. d. 12 ± 0.78 n. d.
 WT AMIO n. d. n. d. 11 ± 0.83 n. d.
 21 days
 WT VH n. d. n. d. 8.8 ± 0.70 n. d.
 WT AMIO n. d. n. d. 11 ± 1.2 n. d.
Experiment 2
 25 days
 WT VH 77 ± 7.8 1.3 ± 0.20 4.4 ± 0.16 0.34 ± 0.04
 WT AMIO 84 ± 8.8 1.4 ± 0.28 4.8 ± 0.14 0.36 ± 0.01
 D2KO VH 78 ± 6.8 1.2 ± 0.13 0.0 ± 0.01 n. d.
 D2KO AMIO 65. ± 3.1 1.3 ± 0.29 0.01 ± 0.0 n. d.
Experiment 3
 21 days
 WT VH 2.5 ± 0.2 und. 0.32 ± 0.01 0.06 ± 0.001
 WT AMIO 1.8 ± 0.1* und. 0.20 ± 0.01* 0.03 ± 0.001*
 D2KO VH 2.8 ± 0.2 und. und. und.
 D2KO AMIO 2.3 ± 0.3* und. und. und.
Open in a new tab

The indicated tissues (ME is hypothalamic median eminence) were harvested from mice treated with either VH or AMIO (80 mg/kg/day for the indicated times) and processed for D1 and/or D2 activity. In Experiments #1 and #2, tissues were sonicated and assayed in buffer containing 10 mm DTT, whereas 1 mm DTT was used in Experiment #3; D1 activity is reported in pmoles/min/mg protein and D2 activity in fmoles/min/mg protein; length of the treatment was as indicated (i.e. 7, 21, and 25 days; entries are mean ± sem of 3–6 animals; n. d. is not determined; und. is undetectable. 

AMIO and DEA inhibit ORD and elevate TSH secretion in intact TαT1 cells

We next took advantage of the TαT1 cell line, which has been used successfully to model the role of D2 in the TSH secretion (21,35). The advantage here is that D2 activity can be measured in live intact cells, at physiological levels of endogenous cofactor(s), and using AMIO or DEA at therapeutically relevant concentrations. Such cells were incubated with 10 nm–10 μm AMIO or DEA in the presence of 10 pm free 125I-T4. Under these conditions, ORD was measured by quantification of free 125I. Because these cells do not express D1 (21), ORD can be used as a surrogate for D2 activity. UPLC studies validated this strategy, showing that equimolar amounts of 125I and 125I-T3 are produced under these conditions (Fig. 5, A and B). Notably, exposure to AMIO significantly inhibited the D2-catalyzed fractional conversion of T4 to T3 in a dose-dependent manner to almost 50% at 10 μm (Fig. 5C). At the same time, DEA was found to be a much stronger inhibitor of the D2-catalyzed fractional conversion of T4 to T3 in these cells, with significant inhibition at 1 μm (Fig. 5D).

Figure 5.

Figure 5

Open in a new tab

Effects of AMIO or DEA on T4 to T3 conversion in intact TαT1 cells incubated in 0.1% BSA DMEM media containing about 200,000 cpm 125IT4 and increasing concentrations of AMIO or DEA (10 nm to 10 μm) for 24 h. Free 125I was measured in the media and the fractional conversion normalized based on the total cell protein. A, Representative chromatogram indicating different 125I-labeled metabolites in the cell media as resolved by UPLC at time 0 and at time 24 h (B). C, Fractional conversion of T4 to T3 was used to build the kinetic plots shown in C and D in the presence of the indicated concentrations of AMIO or DEA, respectively. E and F, Similarly treated TαT1 cells except that T4 was added at 40 pm (final free concentration in media) 30 min after addition of AMIO or DEA and media was harvested 24 h later; values are the mean ± sem and n = 4–6. *, P < 0.05 vs. control; **, P < 0.01 vs. control; ***, P < 0.005 vs. control.

To evaluate the impact of AMIO- or DEA-induced D2 inhibition on TSH secretion, TαT1 cells kept for 24 h in media containing 0.1% BSA were exposed to either one of these compounds and subsequently (30 min later) exposed to 40 pm free T4 during the next 24 h. In the absence of AMIO or DEA, addition of T4 decreased TSH secretion by about 2.5-fold (Fig. 5, E and F). However, inhibition of D2 activity by either AMIO or DEA blunted the capacity of the added T4 to inhibit TSH secretion (Fig. 5, E and F).

Discussion

Alterations in plasma thyroid hormone levels in patients receiving AMIO have been well documented over the years, but the mechanistic basis for these alterations is not completely understood. They include a simultaneous elevation in plasma T4 and TSH with normal/low plasma T3 concentration (11,12,16,17,27,29). Many groups attribute these modifications to a decrease in D1 activity, which has been documented in liver and other tissues of animals treated with AMIO or its metabolites (13,15,16,17,18). However, the D1 knockout (D1KO) mouse does not exhibit an elevated plasma TSH, whereas the D2KO mouse exhibits both an increase in plasma T4 as well as TSH (30). In both the hypothalamus and pituitary, D2 has been shown to locally activate T4 to T3 and thus mediate the negative feedback mechanism caused by plasma T4 (21,36). We therefore hypothesized that AMIO-induced inhibition of D2 in the central thyroid axis would disrupt the transduction of the T4 signal, generating less T3 in the median eminence and/or in thyrotrophs, disrupting the TSH suppression caused by plasma T4. Notably, iopanoic acid is another drug that inhibits D2 and has similar effects on serum TSH and thyroid hormone levels as AMIO (37). In the present investigation we provide experimental confirmation that inhibition of D2 is a critical feature of AMIO-induced changes in TSH levels, finding that D2 is only weakly inhibited by AMIO but much more strongly inhibited by DEA, both in a noncompetitive manner.

For these studies we developed an animal model of AMIO-induced changes in thyroid economy (80 mg/Kg for 25 d), which is very mild given that plasma levels of T4 and T3 were not statistically affected, while there was an approximate twofold elevation in plasma TSH levels (Fig. 1). At the same time, most studies on this subject have used rat models, and the only other mouse model reported required 90 mg/kg for 6 weeks to decrease serum T3 by only 23% while serum T4 and TSH were not measured (38). The mouse model was chosen because of the availability of D2KO mice that allowed us to look at the pathways affected by D2.

Thus, our findings of an elevation in plasma TSH in this mouse model indicate that this is probably one of the most sensitive parameters affected by treatment with AMIO, compared with plasma T4 or T3. The striking implication of an inhibition of normal T4 feedback in the central thyroid axis is that AMIO inhibition of pituitary D2 causes a dissociation of TSH from T4 levels. That D2 is critically involved in this mechanism is demonstrated by the absence of AMIO effects on TSH in the D2KO mouse (Fig. 1).

Interestingly, our results indicate that the elevation in plasma TSH is triggered at the pituitary level, independent of hypothalamic changes because the PVN TRH mRNA levels are decreased by AMIO (Fig. 2). The finding of decreased TRH mRNA in the PVN of AMIO-treated mice (Fig 2) was unexpected because TRH and TSH have been shown to be similarly affected by D2-generated T3 (21,36). The reasons for the decrease in the PVN TRH mRNA levels after treatment with AMIO are not entirely clear. First, AMIO treatment is known to cause loss of body weight due to an anorectic effect, but our animals maintained their body weight probably because the doses of AMIO were small. Second, AMIO is known to function as a partial TR agonist and thus could decrease TRH expression (8). Third, AMIO could be interfering with thyroid hormone transporters in the hypothalamus and thus decrease TRH mRNA.

It is interesting however that the D2KO mice failed to have an elevated TRH mRNA at the PVN, indicating that, contrary to TSH, targeted inactivation of D2 (or AMIO-induced D2 inhibition) either does not affect or decreases TRH gene expression. These findings minimize a role for D2 in the TRH feedback but do not exclude it from the TSH mechanism given that D2-generated T3 in the median eminence is likely to reach the thyrotrophs via transport through the portal system. Most important is the fact that TSH is elevated in the AMIO-treated mice. Together with the low TRH mRNA levels, this indicates that AMIO and/or its metabolites act at the pituitary level to inhibit D2 and elevate TSH, independent of hypothalamic mechanisms.

The present data point toward the existence of previously unidentified fundamental differences between the TRH and TSH feedback mechanisms given that the net effect of D2 inhibition and/or interference with TR signaling is a decrease in TRH mRNA (Fig. 2) combined with an elevation in TSH secretion (Figs. 1 and 5). Whether this is because AMIO/DEA are partial agonists of TR and thus their effects vary from being weak agonists to antagonists, depending on how much T3 is available, remains to be investigated.

In the rat pituitary gland, ORD (combined activity of D1 and D2) is inhibited by ∼60% after treatment with AMIO (26), but the relative effects of AMIO on the deiodinases in this setting have not been reported. However, when AMIO was added to pituitary sonicates little D2 inhibition was seen (25). At the same time, D2 is inhibited by about 30% in human epidermal keratinocytes exposed to AMIO (24). In the present studies we find that AMIO indeed is a very poor inhibitor of D2 when added to cell sonicates, whereas DEA is more potent (Fig. 3, A and B). On the other hand, when live cells are exposed to AMIO (Fig. 5C), inhibition is more potent, presumably due to production of AMIO metabolites such as DEA, which are more powerful inhibitors of D2 (Fig. 3, C and D). We also addressed the kinetic mechanism of inhibition in cell sonicates and have found that both AMIO and DEA inhibit D2 in intact cells at therapeutic concentrations by a noncompetitive mechanism.

Notably, we were only able to detect D2 inhibition in AMIO-treated mouse tissue sonicates if samples were prepared with low DTT concentration (Table 1) suggesting that the effect of AMIO and/or DEA on D2 activity is weak and/or reversible upon exposure to high levels of DTT (Table 1). In fact, Safran et al. only detected an ORD inhibition in liver of AMIO-treated rats after excluding DTT from the assay (26). Here, we also took advantage of the TαT1 cell system, which has been used to model the T4-TSH feedback mechanism (21), to unequivocally demonstrate that exposure to AMIO or DEA inhibits D2 in cultured cells (Fig. 5, A–D). Remarkably, addition of increasing concentrations of either one of these molecules progressively dampened the T4-induced TSH suppression observed in these cells (Fig. 5, E and F).

D1 is also inhibited by AMIO. Much has been published on this subject, and it is well accepted that theoretical AMIO metabolites (B2, B21, and B22) inhibit D1 via a competitive inhibitory mechanism (19). In vivo this effect is potentially amplified by the drop in plasma T3 observed in AMIO-treated mice, given that T3 positively regulates D1 (39). On the other hand, our data show that DEA is a strong noncompetitive inhibitor of D1 (Fig 4) and, because B2, B21, and B22 are not found in live animals, it is likely that AMIO treatment results in noncompetitive inhibition of D1 in rodents and humans, just as it does with D2.

In conclusion, AMIO is a weak noncompetitive inhibitor of D1 and D2, whereas its metabolite, DEA, strongly inhibits both enzymes via a similar mechanism. Thus, for AMIO to exhibit significant inhibitory effects on deiodinase activity it must be added to live cells or injected in live animals, allowing for its conversion to DEA. By inhibiting D2 activity, AMIO weakens the T4-mediated feedback at the pituitary gland, thus elevating plasma TSH. TRH expression does not seem to be involved in this mechanism. It is expected that prolonged/more intense D2 inhibition with AMIO would also result in elevation of plasma T4, just as observed in the D2KO mouse.

Acknowledgments

We thank Dr. H. C. van Beeren for providing the critical DEA compound, Dr. E. Chip Ridgway for providing the Tα-T1 cells, Dr. Brian Kim for critically reviewing the manuscript, and Jessica A. Hall for insightful discussions during the course of this project.

Footnotes

This work was supported by National Institutes of Health Grants DK58538 and DK37021.

Disclosure Summary: The authors have nothing to declare.

First Published Online October 6, 2010

Abbreviations: AMIO, amiodarone; CYP3A, cytochrome P4503A; D1, type 1 deiodinase; D2, type 2 deiodinase; D2KO, D2 knockout mouse;D3, type 3 deiodinase; DEA, desethylamiodarone; FBS, fetal bovine serum; IRD, inner ring deiodination; ORD, outer ring deiodination; PVN, paraventricular nucleus; rT3, reverse triiodothyronine; TRα, thyroid hormone receptor alpha; TRβ, thyroid hormone receptor beta; TRH, TSH releasing hormone; VH, vehicle; WT, wild-type.

References

  1. Reiffel JA, Estes 3rd NA, Waldo AL, Prystowsky EN, DiBianco R 1994 A consensus report on antiarrhythmic drug use. Clin Cardiol 17:103–116 [DOI] [PubMed] [Google Scholar]
  2. Holt DW, Tucker GT, Jackson PR, Storey GC 1983 Amiodarone pharmacokinetics. Am Heart J 106:840–847 [DOI] [PubMed] [Google Scholar]
  3. Plomp TA, van Rossum JM, Robles de Medina EO, van Lier T, Maes RA 1984 Pharmacokinetics and body distribution of amiodarone in man. Arzneimittelforschung 34:513–520 [PubMed] [Google Scholar]
  4. Martino E, Bartalena L, Bogazzi F, Braverman LE 2001 The effects of amiodarone on the thyroid. Endocr Rev 22:240–254 [DOI] [PubMed] [Google Scholar]
  5. Krenning EP, Docter R, Bernard B, Visser T, Hennemann G 1982 Decreased transport of thyroxine (T4), 3,3′,5-triiodothyronine (T3) and 3,3′,5′-triiodothyronine (rT3) into rat hepatocytes in primary culture due to a decrease of cellular ATP content and various drugs. FEBS Lett 140:229–233 [DOI] [PubMed] [Google Scholar]
  6. Bakker O, van Beeren HC, Wiersinga WM 1994 Desethylamiodarone is a noncompetitive inhibitor of the binding of thyroid hormone to the thyroid hormone beta 1-receptor protein. Endocrinology 134:1665–1670 [DOI] [PubMed] [Google Scholar]
  7. van Beeren HC, Bakker O, Wiersinga WM 1995 Desethylamiodarone is a competitive inhibitor of the binding of thyroid hormone to the thyroid hormone alpha 1-receptor protein. Mol Cell Endocrinol 112:15–19 [DOI] [PubMed] [Google Scholar]
  8. Bogazzi F, Bartalena L, Brogioni S, Burelli A, Raggi F, Ultimieri F, Cosci C, Vitale M, Fenzi G, Martino E 2001 Desethylamiodarone antagonizes the effect of thyroid hormone at the molecular level. Eur J Endocrinol 145:59–64 [DOI] [PubMed] [Google Scholar]
  9. Koenig RJ 2005 Regulation of type 1 iodothyronine deiodinase in health and disease. Thyroid 15:835–840 [DOI] [PubMed] [Google Scholar]
  10. Gereben B, Zavacki AM, Ribich S, Kim BW, Huang SA, Simonides WS, Zeold A, Bianco AC 2008 Cellular and molecular basis of deiodinase-regulated thyroid hormone signaling. Endocr Rev 29:898–938 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Aanderud S, Sundsfjord J, Aarbakke J 1984 Amiodarone inhibits the conversion of thyroxine to triiodothyronine in isolated rat hepatocytes. Endocrinology 115:1605–1608 [DOI] [PubMed] [Google Scholar]
  12. Hershman JM, Nademanee K, Sugawara M, Pekary AE, Ross R, Singh BN, DiStefano JJ, 3rd 1986 Thyroxine and triiodothyronine kinetics in cardiac patients taking amiodarone. Acta Endocrinol (Copenh) 111:193–199 [DOI] [PubMed] [Google Scholar]
  13. Gotzsche LS, Boye N, Laurberg P, Andreasen F 1989 Rat heart thyroxine 5′-deiodinase is sensitively depressed by amiodarone. J Cardiovasc Pharmacol 14:836–841 [DOI] [PubMed] [Google Scholar]
  14. Ceppi JA, Gonzalez MR, Zaninovich AA 1988 [Effect of amiodarone on the conversion of thyroxine to triiodothyronine in the myocardium of rats in vivo and in vitro]. Medicina (B Aires) 48:29–32 [PubMed] [Google Scholar]
  15. Balsam A, Ingbar SH 1978 The influence of fasting, diabetes, and several pharmacological agents on the pathways of thyroxine metabolism in rat liver. J Clin Invest 62:415–424 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Sogol PB, Hershman JM, Reed AW, Dillmann WH 1983 The effects of amiodarone on serum thyroid hormones and hepatic thyroxine 5′-monodeiodination in rats. Endocrinology 113:1464–1469 [DOI] [PubMed] [Google Scholar]
  17. Pekary AE, Hershman JM, Reed AW, Kannon R, Wang YS 1986 Amiodarone inhibits T4 to T3 conversion and alpha-glycerophosphate dehydrogenase and malic enzyme levels in rat liver. Horm Metab Res 18:114–118 [DOI] [PubMed] [Google Scholar]
  18. Ceppi JA, Zaninovich AA 1989 Effects of amiodarone on 5′-deiodination of thyroxine to tri-iodothyronine in rat myocardium. J Endocrinol 121:431–434 [DOI] [PubMed] [Google Scholar]
  19. Ha HR, Stieger B, Grassi G, Altorfer HR, Follath F 2000 Structure-effect relationships of amiodarone analogues on the inhibition of thyroxine deiodination. Eur J Clin Pharmacol 55:807–814 [DOI] [PubMed] [Google Scholar]
  20. Hudig F, Bakker O, Wiersinga WM 1994 Amiodarone-induced hypercholesterolemia is associated with a decrease in liver LDL receptor mRNA. FEBS Lett 341:86–90 [DOI] [PubMed] [Google Scholar]
  21. Christoffolete MA, Ribeiro R, Singru P, Fekete C, da Silva WS, Gordon DF, Huang SA, Crescenzi A, Harney JW, Ridgway EC, Larsen PR, Lechan RM, Bianco AC 2006 Atypical expression of type 2 iodothyronine deiodinase in thyrotrophs explains the thyroxine-mediated pituitary thyrotropin feedback mechanism. Endocrinology 147:1735–1743 [DOI] [PubMed] [Google Scholar]
  22. Fekete C, Lechan RM 2007 Negative feedback regulation of hypophysiotropic thyrotropin-releasing hormone (TRH) synthesizing neurons: role of neuronal afferents and type 2 deiodinase. Front Neuroendocrinol 28:97–114 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Maia AL, Kim BW, Huang SA, Harney JW, Larsen PR 2005 Type 2 iodothyronine deiodinase is the major source of plasma T3 in euthyroid humans. J Clin Invest 115:2524–2533 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kaplan MM, Pan CY, Gordon PR, Lee JK, Gilchrest BA 1988 Human epidermal keratinocytes in culture convert thyroxine to 3,5,3′-triiodothyronine by type II iodothyronine deiodination: a novel endocrine function of the skin. J Clin Endocrinol Metab 66:815–822 [DOI] [PubMed] [Google Scholar]
  25. Kaplan MM, Breitbart R 1984 Conversion of thyroxine to triiodothyronine in the anterior pituitary gland and the influence of this process on thyroid status. Horm Metab Res Suppl 14:79–85 [PubMed] [Google Scholar]
  26. Safran M, Fang SL, Bambini G, Pinchera A, Martino E, Braverman LE 1986 Effects of amiodarone and desethylamiodarone on pituitary deiodinase activity and thyrotropin secretion in the rat. Am J Med Sci 292:136–141 [DOI] [PubMed] [Google Scholar]
  27. Burger A, Dinichert D, Nicod P, Jenny M, Lemarchand-Beraud T, Vallotton MB 1976 Effect of amiodarone on serum triiodothyronine, reverse triiodothyronine, thyroxin, and thyrotropin. A drug influencing peripheral metabolism of thyroid hormones. J Clin Invest 58:255–259 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Melmed S, Nademanee K, Reed AW, Hendrickson JA, Singh BN, Hershman JM 1981 Hyperthyroxinemia with bradycardia and normal thyrotropin secretion after chronic amiodarone administration. J Clin Endocrinol Metab 53:997–1001 [DOI] [PubMed] [Google Scholar]
  29. Nademanee K, Singh BN, Hendrickson JA, Reed AW, Melmed S, Hershman J 1982 Pharmacokinetic significance of serum reverse T3 levels during amiodarone treatment: a potential method for monitoring chronic drug therapy. Circulation 66:202–211 [DOI] [PubMed] [Google Scholar]
  30. Schneider MJ, Fiering SN, Pallud SE, Parlow AF, St Germain DL, Galton VA 2001 Targeted disruption of the type 2 selenodeiodinase gene (DIO2) results in a phenotype of pituitary resistance to T4. Mol Endocrinol 15:2137–2148 [DOI] [PubMed] [Google Scholar]
  31. Kádár A, Sánchez E, Wittmann G, Singru P, Füzesi T, Marsili A, Larsen P, Liposits Z, Lechan R, Fekete C 2010 Distribution of hypophysiotropic thyrotropin-releasing hormone (TRH)-synthesizing neurons in the hypothalamic paraventricular nucleus of the mouse. J Comp Neurol [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Fekete C, Kelly J, Mihaly E, Sarkar S, Rand WM, Legradi G, Emerson CH, Lechan RM 2001 Neuropeptide Y has a central inhibitory action on the hypothalamic-pituitary-thyroid axis. Endocrinology 142:2606–2613 [DOI] [PubMed] [Google Scholar]
  33. Freitas BC, Gereben B, Castillo M, Kallo I, Zeold A, Egri P, Liposits Z, Zavacki AM, Maciel RM, Jo S, Singru P, Sanchez E, Lechan RM, Bianco AC. Paracrine signaling by glial cell-derived triiodothyronine activates neuronal gene expression in the rodent brain and human cells. J Clin Invest 120:2206–2217 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Falik R, Flores BT, Shaw L, Gibson GA, Josephson ME, Marchlinski FE 1987 Relationship of steady-state serum concentrations of amiodarone and desethylamiodarone to therapeutic efficacy and adverse effects. Am J Med 82:1102–1108 [DOI] [PubMed] [Google Scholar]
  35. Yusta B, Alarid ET, Gordon DF, Ridgway EC, Mellon PL 1998 The thyrotropin beta-subunit gene is repressed by thyroid hormone in a novel thyrotrope cell line, mouse T alphaT1 cells. Endocrinology 139:4476–4482 [DOI] [PubMed] [Google Scholar]
  36. Tu HM, Kim SW, Salvatore D, Bartha T, Legradi G, Larsen PR, Lechan RM 1997 Regional distribution of type 2 thyroxine deiodinase messenger ribonucleic acid in rat hypothalamus and pituitary and its regulation by thyroid hormone. Endocrinology 138:3359–3368 [DOI] [PubMed] [Google Scholar]
  37. Suzuki H, Kadena N, Takeuchi K, Nakagawa S 1979 Effects of three-day oral cholecystography on serum iodothyronines and TSH concentrations: comparison of the effects among some cholecystographic agents and the effects of iopanoic acid on the pituitary-thyroid axis. Acta Endocrinol (Copenh) 92:477–488 [DOI] [PubMed] [Google Scholar]
  38. Le Bouter S, El Harchi A, Marionneau C, Bellocq C, Chambellan A, van Veen T, Boixel C, Gavillet B, Abriel H, Le Quang K, Chevalier JC, Lande G, Leger JJ, Charpentier F, Escande D, Demolombe S 2004 Long-term amiodarone administration remodels expression of ion channel transcripts in the mouse heart. Circulation 110:3028–3035 [DOI] [PubMed] [Google Scholar]
  39. Berry MJ, Kates AL, Larsen PR 1990 Thyroid hormone regulates type I deiodinase messenger RNA in rat liver. Mol Endocrinol 4:743–748 [DOI] [PubMed] [Google Scholar]

Articles from Endocrinology are provided here courtesy of The Endocrine Society

RESOURCES