Deiodination Of rT3 Research Articles

Regioselective Deiodination of Thyroxine by Iodothyronine Deiodinase Mimics: An Unusual Mechanistic Pathway Involving Cooperative Chalcogen and Halogen Bonding

Iodothyronine deiodinases (IDs) are mammalian selenoenzymes that catalyze the conversion of thyroxine (T4) to 3,5,3'-triiodothyronine (T3) and 3,3',5'-triiodothyronine (rT3) by the outer- and inner-ring deiodination pathways, respectively. These enzymes also catalyze further deiodination of T3 and rT3 to produce a variety of di- and monoiodo derivatives. In this paper, the deiodinase activity of a series of peri-substituted naphthalenes having different amino groups is described. These compounds remove iodine selectively from the inner-ring of T4 and T3 to produce rT3 and 3,3'-diiodothyronine (3,3'-T2), respectively. The naphthyl-based compounds having two selenols in the peri-positions exhibit much higher deiodinase activity than those having two thiols or a thiol-selenol pair. Mechanistic investigations reveal that the formation of a halogen bond between the iodine and chalcogen (S or Se) and the peri-interaction between two chalcogen atoms (chalcogen bond) are important for the deiodination reactions. Although the formation of a halogen bond leads to elongation of the C-I bond, the chalcogen bond facilitates the transfer of more electron density to the C-I σ* orbitals, leading to a complete cleavage of the C-I bond. The higher activity of amino-substituted selenium compounds can be ascribed to the deprotonation of thiol/selenol moiety by the amino group, which not only increases the strength of halogen bond but also facilitates the chalcogen-chalcogen interactions.

Kinetics and thiol requirements of iodothyronine 5′-deiodination are tissue-specific in common carp (Cyprinus carpio L.)

Iodothyronine deiodinases determine the biological activity of thyroid hormones. Despite the homology of the catalytic sites of mammalian and teleostean deiodinases, in-vitro requirements for the putative thiol co-substrate dithiothreitol (DTT) vary considerably between vertebrate species. To further our insights in the interactions between the deiodinase protein and its substrates: thyroid hormone and DTT, we measured enzymatic iodothyronine 5'-deiodination, Dio1 and Dio2 mRNA expression, and Dio1 affinity probe binding in liver and kidney preparations from a freshwater teleost, the common carp (Cyprinus carpio L.). Deiodination rates, using reverse T3 (rT3, 3,3',5'-triiodothyronine) as the substrate, were analysed as a function of the iodothyronine and DTT concentrations. In kidney rT3 5'-deiodinase activity measured at rT3 concentrations up to 10 μM and in the absence of DTT does not saturate appreciably. In the presence of 1mM DTT, renal rT3 deiodination rates are 20-fold lower. In contrast, rT3 5'-deiodination in liver is potently stimulated by 1mM DTT. The marked biochemical differences between 5'-deiodination in liver and kidney are not associated with the expression of either Dio1 or Dio2 mRNA since both organs express both deiodinase types. In liver and kidney, DTT stimulates the incorporation of N-bromoacetylated affinity labels in proteins with estimated molecular masses of 57 and 55, and 31 and 28 kDa, respectively. Although primary structures are highly homologous, the biochemistry of carp deiodinases differs markedly from their mammalian counterparts.

Thyroid hormone transport in developing brain

To discuss the recent advances on thyroid hormone transport in the brain. A special attention is paid to the X-linked thyroid hormone cell transport (THCT) defect (also known as the Allan-Herndon-Dudley syndrome), caused by mutations of the specific thyroid hormone transporter MCT8 gene. MCT8 is involved in thyroid hormone transport in the brain. MRI of patients with THCT defect showed myelination delays, probably related to impaired thyroid hormone action on oligodendrocytes. MCT8 is also expressed in the thyroid and has an important role in thyroid hormone secretion. The altered circulating concentrations of thyroid hormone in the patients are partly because of impaired secretion and altered peripheral metabolism. Increased deiodinase activity is important in the pathophysiology of the syndrome. High D1 activity in liver and kidney increases T4 and rT3 deiodination, and contributes to the increased serum T3. High D2 activity in the brain contributes to compensate the deficient T3 transport by increasing local T3 production. Patients with suspected X-linked leukoencephalopathy should be screened for MCT8 gene mutations. Research on the brain pathophysiology of the THCT defect should focus on the specific role of Mct8 on oligodendrocytes and myelination.

A Nonselenoprotein from Amphioxus Deiodinates Triac But Not T3: Is Triac the Primordial Bioactive Thyroid Hormone?

Thyroid hormone (TH) is important for metamorphosis in many species, including the cephalochordate Branchiostoma floridae, a marine invertebrate (amphioxus) living in warmer coastal areas. Branchiostoma expresses a TH receptor, which is activated by 3,3',5-triiodothyroacetic acid (TA(3)) but not by T(3). The Branchiostoma genome also contains multiple genes coding for proteins homologous to iodothyronine deiodinases in vertebrates, selenoproteins catalyzing the activation or inactivation of TH. Three Branchiostoma deiodinases have been cloned: two have a catalytic Sec, and one, bfDy, has a Cys residue. We have studied the catalytic properties of bfDy in transfected COS1 cells by HPLC analysis of reactions with (125)I-labeled substrates and dithiothreitol as cofactor. We could not detect deiodination of T(4), T(3), or rT(3) by bfDy but observed rapid and selective inner ring deiodination (inactivation) of TA(3) and 3,3',5,5'-tetraiodothyroacetic acid (TA(4)). Deiodination of TA(3) by bfDy was optimal at 25 C and 10 mm dithiothreitol. bfDy was extremely labile at 37 C, showing a half-life of less than 2 min, in contrast with a half-life of more than 60 min at 25 C. Deiodination of labeled TA(3) was inhibited dose dependently by unlabeled TA(3)≈TA(4)>T(4)≈T(3). Michaelis-Menten analysis yielded Michaelis-Menten constant values of 6.8 and 68 nm and maximum velocity values of 1.4 and 5.4 pmol/min·mg protein for TA(3) and TA(4), respectively. bfDy was not inhibited by propylthiouracil and iodoacetate and only weakly by goldthioglucose and iopanoic acid. In conclusion, we demonstrate rapid inactivation of TA(3) and TA(4) but not of T(3) and T(4) by the first reported natural nonselenodeiodinase. Our findings support the hypothesis that TA(3) is a primordial bioactive TH.

Halogenated Phenolic Contaminants Inhibit the In Vitro Activity of the Thyroid-Regulating Deiodinases in Human Liver

Halogenated contaminants, particularly brominated flame retardants, disrupt circulating levels of thyroid hormones (THs), potentially affecting growth and development. Disruption may be mediated by impacts on deiodinase (DI) activity, which regulate the levels of active hormones available to bind to nuclear receptors. The goal of this study was to develop a mass spectrometry-based method for measuring the activity of DIs in human liver microsomes and to examine the effect of halogenated phenolic contaminants on DI activity. Thyroxine (T4) and reverse triiodothyronine (rT3) deiodination kinetics were measured by incubating pooled human liver microsomes with T4 or rT3 and monitoring the production of T3, rT3, 3,3'-diiodothyronine, and 3-monoiodothyronine by liquid chromatography tandem mass spectrometry. Using this method, we examined the effects of several halogenated contaminants, including 2,2',4,4',5-pentabromodiphenyl ether (BDE 99), several hydroxylated polybrominated diphenyl ethers (OH-BDEs), tribromophenol, tetrabromobisphenol A, and triclosan, on DI activity. The Michaelis constants (K(M)) of rT3 and T4 deiodination were determined to be 3.2 ± 0.7 and 17.3 ± 2.3μM. The V(max) was 160 ± 5.8 and 2.8 ± 0.10 pmol/min.mg protein, respectively. All studied contaminants inhibited DI activity in a dose-response manner, with the exception of BDE 99 and two OH-BDEs. 5'-Hydroxy 2,2',4,4',5-pentabromodiphenyl ether was found to be the most potent inhibitor of DI activity, and phenolic structures containing iodine were generally more potent inhibitors of DI activity relative to brominated, chlorinated, and fluorinated analogues. This study suggests that some halogenated phenolics, including current use compounds such as plastic monomers, flame retardants, and their metabolites, may disrupt TH homeostasis through the inhibition of DI activity in vivo.

Molecular basis for the substrate selectivity of cat type I iodothyronine deiodinase.

The type I iodothyronine deiodinase (D1) catalyzes the activation of T4 to T3 as well as the degradation of T3 (rT3) and sulfated iodothyronines. A comparison of the catalytic activities of D1 in liver microsomal preparations from several species revealed a remarkable difference between cat D1 on one hand and rat/human D1 on the other hand. The Michaelis constant (Km) of cat D1 for rT3 (11 microm) is 30-fold higher than that of rat and human D1 (0.2-0.5 microm). Deiodination of rT3 by cat D1 is facilitated by sulfation [maximal velocity (Vmax)/Km rT3 = 3 and Vmax/Km rT3S = 81]. To understand the molecular basis for the difference in substrate interaction the cat D1 cDNA was cloned, and the deduced amino acid sequence was compared with rat/human D1 protein. In the region between amino acid residues 40 and 70 of cat D1, various differences with rat/human D1 are concentrated. By site-directed mutagenesis of cat D1 it was found that a combination of mutations was necessary to improve the deiodination of rT3 by cat D1 enzyme. For efficient rT3 deiodination, a Phe at position 65 and the insertion of the Thr-Gly-Met-Thr-Arg48-52 sequence as well as the amino acids Gly and Glu at position 45-46 are essential. Either of these changes alone resulted in only a limited improvement of rT3 deiodination. At the same time the combination of the described mutations did not affect the already quite efficient outer ring deiodination of rT3S nor the inner ring deiodination of T3S, whereas each of the described changes alone did affect rT3S deiodination. Our findings suggest great flexibility of the active site in D1 that adapts to its various substrates. The active site of wild-type cat D1 is less flexible than the active site of rat/human D1 and favors sulfated iodothyronines.

Structure-activity relationships for thyroid hormone deiodination by mammalian type I iodothyronine deiodinases.

The bioactivity of thyroid hormone is determined to a large extent by the monodeiodination of the prohormone T4 by the hepatic selenoenzyme type I iodothyronine deiodinase (IDI), i.e. by outer ring deiodination (ORD) to the active hormone T3' or by inner ring deiodination (IRD) to the inactive metabolite rT3. IDI also catalyzes the IRD of T3 and the ORD of rT3' both to T2, as well as the deiodination of different iodothyronine sulfates, e.g. IRD of T3S and ORD of T2S. Previous studies have indicated important differences in catalytic specificity between dog IDI (dID1) and human ID1 (hID1), in particular with respect to the ORD of rT3. This study was done to investigate the relationship between structure and catalytic function of this enzyme by comparing the deiodination of T4, T3, rT3, T3S, and T2S by native dID1 and hID1 in liver microsomes as well as by recombinant wild-type, chimeric and mutated d/hID1 enzymes expressed in HEK293 cells. With both native and recombinant wild-type enzymes, the substrate specificity was T3S > T2S approximately rT3 approximately T4 > T3 for dID1, and rT3 > > T2S approximately T3S > T4 approximately T3 for hID1. Whereas ORD of T4 and of T4, T3, and T3S showed relatively little variation between the different d/hID1 constructs, large differences were found for the ORD of rT3 and T2S. Both reactions were favored by the presence of the amino acids G, E and, in particular, F, present in hID1 at positions 45, 46, and 65, instead of the dID1 residues N, G, and L, respectively. However, although ORD of rT3 was not affected by the presence (hID1) or absence (dID1) of the TGMTR(48-52) sequence, the ORD of T2S was markedly inhibited by the presence of this sequence. Therefore, we have identified structural elements in ID1 that have substrate-specific impacts on deiodination. Our results suggest the specific interaction of the mono-substituted inner ring of the substrates rT3 and T2S but not the disubstituted inner ring of T3, T3S, or T4, with the aromatic ring of F65 in Id1, perhaps by pi-pi interactions.

Phenolic and tyrosyl ring iodothyronine deiodination and thyroid hormone concentrations in the human central nervous system.

In the present study we investigated the biochemical properties of in vitro phenolic (5'D) and tyrosyl (5D) ring deiodination and the tissue concentrations of T4, T3, and rT3 in adult human central nervous system (CNS) tissue. All samples were obtained from nontumoral tissue at autopsy (n = 6) or neurosurgical operation (n = 5). Both phenolic and tyrosyl ring deiodinase activities were demonstrable in all samples obtained intraoperatively, whereas only tyrosyl ring deiodination was evident in the tissues obtained postmortem. The phenolic ring deiodination pathway corresponded to the type II 5'-deiodinase isoenzyme with regard to its high affinity for T4 and rT3 (Km = 2.2 and 2.4 nmol/L, respectively), its insensitivity to 6-propyl-n-2-thiouracil (PTU), and the sequential reaction mechanism. No PTU-sensitive 5'-deiodination of rT3 was demonstrable. Tyrosyl ring deiodination of both T4 and T3 showed typical type III 5D kinetics (Ka, 6.5 nmol/L for T4 and 3.4 nmol/L for T3) and was PTU insensitive. Nanomolar concentrations of tissue T4, T3, and rT3 were detected in samples obtained both intraoperatively and postmortem. They were very similar to the absolute values of the apparent Km for T4, T3, and rT3 in the phenolic and tyrosyl ring deiodination pathways. In conclusion, we have demonstrated the coexistence of both phenolic and tyrosyl ring deiodinase activities in the human CNS. Their kinetic characteristics, substrate specificity, and reaction mechanisms are very similar to the corresponding type II 5'- and type III 5-iodothyronine deiodinase activities in rat brain. In contrast to the findings in the rat CNS, no PTU-sensitive phenolic ring deiodinase (i.e. type I 5'D) activity was found in the human CNS. This may explain the relatively high tissue concentrations of rT3.

Effects of fibroblast growth factor on type I 5'-deiodinase in FRTL-5 rat thyroid cells.

In the present study, we used FRTL-5 cells to study the effects of fibroblast growth factors (FGFs) on 5'-deiodinase (5'D) activity and messenger RNA (mRNA) levels. In FRTL-5 cells deprived of TSH for 7 days, type I 5'-deiodinase (5'D-I) activity decreased to low, but detectable levels. Incubation of cells with acidic and basic FGFs significantly decreased 5'D-I activity below the basal levels. After 7 days of TSH deprivation, the addition of TSH (100 microU/ml) to the medium for 3 days resulted in an increase in 5'D-I activity. This TSH-induced increase in 5'D-I activity was blocked by the FGFs in a dose-dependent manner. Kinetic analysis revealed that both acidic and basic FGFs decreased the maximum velocity of 5'D-I activity in the presence or absence of TSH, without any significant effect on the Km of enzyme binding. HPLC analysis of the products of the 5'D-I assay revealed that there was no sequential deiodination of rT3. Incubation of FRTL-5 cells with acidic or basic FGF did not affect basal cAMP concentrations, nor did they block the TSH-induced rise in cAMP. However, acidic and basic FGFs inhibited forskolin- and (Bu)2cAMP-induced increases in 5'D-I activity. Incubation of FRTL-5 cells with TSH, (Bu)2cAMP, and forskolin increased 5'D-I mRNA levels. Incubation of FRTL-5 cells with acidic and basic FGFs decreased steady state 5'D-I mRNA levels and blocked the TSH-, forskolin-, and (Bu)2cAMP-induced increases in 5'D-I mRNA. In conclusion, we have demonstrated that FGFs inhibit 5'D-I activity and mRNA levels in FRTL-5 cells in the presence or absence of TSH. The inhibitory effect of FGFs on 5'D-I in FRTL-5 cells is mediated through either a cAMP-independent pathway or pathways distal to the generation of cAMP. The present data together with the identification of FGF in the thyroid gland suggest that FGF may play a physiological role in the regulation of thyroid hormone secretion.

Identification of critical amino acids for 3,5,3'-triiodothyronine deiodination by human type 1 deiodinase based on comparative functional-structural analyses of the human, dog, and rat enzymes.

The selenoenzyme, type 1 iodothyronine deiodinase (type 1 DI), catalyzes the activation of thyroxine (T4) to 3,5,3'-triiodothyronine (T3) but 3,3',5'-triiodothyronine (rT3) is the preferred substrate for the human enzyme. Since the dog type 1 DI has a significantly lower affinity for rT3, we cloned the dog type 1 DI to identify amino acids critical for rT3 binding. The Km of the transiently expressed dog enzyme for rT3 5'-deiodination is 25-fold higher than that of the human enzyme. However, the Ki of T4 for rT3 deiodination by dog type 1 DI is only 3-fold higher than that for the human, suggesting that the differences between the two proteins affect binding of rT3 more than that of T4. Comparative competition studies in which rT3 or T4 is used to block covalent bromoacetyl T3 binding to the two proteins support this. Mutational studies showed that the critical differences between the dog (D) and human (H) enzymes are Asn (D) 45 Gly (H), Gly (D) 46 Glu (H), and Leu (D) 60:Phe (H) 65. Substitution of the human residues for those of the dog at these positions causes the predicted changes in the Km (rT3) and vice versa. A Phe65 to Leu mutation alone in the human enzyme increases the Km (rT3) 10-fold. We speculate that Phe65 is especially important for rT3 binding due to an interaction between the tyrosyl ring of rT3 and the aromatic ring of Phe65.

The postnatal serum 3,5,3'-triiodothyronine (T3) surge in the rat is largely independent of extrathyroidal 5'-deiodination of thyroxine to T3.

In the rat, selenium deficiency causes a near-complete loss of the selenoenzyme type I 5'-deiodinase (5'D-I), resulting in a marked decrease in hepatic T4 to T3 conversion. In adult rats, serum T4 concentrations are consistently increased, whereas serum T3 and rT3 concentrations are unaffected or slightly decreased and increased, respectively. In rat fetuses near term, serum T4 and rT3 concentrations are not affected by selenium deficiency. We have now studied the effect of selenium deficiency on thyroid function in the neonatal rat. Weanling female rats were fed either a selenium-supplemented or a selenium-deficient diet for 4 weeks before mating and then throughout gestation and lactation. Neonatal rats were killed at 7, 14, 21, and 28 days. Selenium deficiency was confirmed by a more than 89% decrease in liver 5'D-I activity in mothers and pups. Selenium deficiency resulted in significant increases in serum T4 concentrations in 3- and 4-week-old pups. In contrast, selenium deficiency led to a striking increase in serum rT3 concentrations. The normal postnatal serum T3 surge was not affected by selenium deficiency at any age. In 2- and 4-week-old selenium-deficient pups obtained from a second litter from the same mothers, liver 5'D-I activity was markedly decreased, but thyroid 5'D-I activity was not affected. The increased serum rT3 and, less so, T4 concentrations observed in selenium-deficient pups were associated with a significant decrease in brain 5'D-II activity in 14- and 28-day-old pups and in brown adipose tissue 5'D-II activity in 14-day-old pups. In conclusion, the present study demonstrates that the increase in serum T4 concentrations consistently observed in selenium-deficient adult rats occurs only after the second week of life. The normal physiological postnatal 12-fold increase in serum T3 concentrations observed in selenium-deficient pups despite the marked decreases in liver 5'D-I and brain and brown adipose tissue 5'D-II activities suggests that T4 to T3 conversion by peripheral tissues may not be a major source of T3 in the neonate. In contrast, the thyroid gland, whose 5'D-I activity is not affected by selenium deficiency, is probably the principal source of circulating T3 in the neonate. Finally, the early and marked increase in serum rT3 concentrations observed in selenium-deficient pups suggests that liver 5'D-I is important in rT3 deiodination.

Inhibition of thyroxine transport into cultured rat hepatocytes by serum of nonuremic critically ill patients: effects of bilirubin and nonesterified fatty acids.

We investigated bilirubin and oleic acid as causes of low plasma T3 in nonuremic critically ill patients with gross changes in serum thyroid hormone levels (T4, < or = 60; T3, < or = 1.1; rT3, > or = 0.45 nmol/L) and elevated bilirubin concentrations (> or = 33 mumol/L). Iodide production from [125I]T4 was inhibited by 42% when rat hepatocytes in primary cultures were incubated with 10% serum from these patients. The mean serum concentration of albumin was reduced by 41%, while the concentrations of bilirubin and nonesterified fatty acids (NEFA) were increased by 2022% and 115%, respectively, in the patients. The molar ratios of bilirubin/albumin and NEFA/albumin in the patients were 0.42 and 3.18, respectively. Addition of oleic acid (50-400 mumol/L) and bilirubin (3-130 mumol/L) to 10% normal human serum (albumin, 70 mumol/L; NEFA, 54 mumol/L; bilirubin, 1.1 mumol/L) progressively inhibited the production of iodide by rat hepatocytes. The decreased iodide production was presumed to be caused by inhibition of T4 transport into hepatocytes. The deiodination of rT3 by rat liver microsomes was unaltered by free bilirubin and free oleic acid concentrations up to 0.1 mumol/L. These free concentrations are at least 1 order of magnitude higher than that attained in nonthyroidal illness. The inhibition of iodide production by the sera of critically ill patients (n = 12) was significantly correlated with the molar ratios of bilirubin/albumin (r = 0.72; P < 0.01) and NEFA/albumin (r = 0.58; P < 0.05). Extensive dialysis or treatment of the sera with charcoal did not completely remove the inhibitory activity on iodide production. Serum concentrations of indoxyl sulfate, 3-carboxy-4-methyl-5-propyl-2-furan propanoic acid, and hippuric acid in the critically ill patients (other known T4 transport inhibitors into hepatocytes) were similar to those in the normal subjects. This study together with the well known effects of carbohydrate on T3 neogenesis suggest that elevated bilirubin and NEFA and the low albumin level in non-uremic critical illness may be at least partly responsible for the T4 transport inhibition in T3-producing tissues (e.g. the liver) and, thus, the low plasma T3 levels in these critically ill patients. The question of whether inhibitors of T4 transport into the hepatocytes are also present in other patients with nonthyroidal illness who show only mild changes in thyroid hormone levels and have low concentrations of bilirubin and NEFA remains to be determined.

Thyroid Hormone Deiodination in Brain, Liver, Gill, Heart and Muscle of Atlantic Salmon (Salmo salar) during Photoperiodically-Induced Parr-Smolt Transformation. II. Outer- and Inner-Ring 3,5,3′-Triiodo-l-Thyronine and 3,3′,5′-Triiodo-l-Thyronine (Reverse T3) Deiodination

Outer-ring deiodinase (ORD) and inner-ring deiodinase (IRD) pathways for 3,5,3′-triiodo-l-thyronine (T3) and 3,3′,5′-T3 (reverse T3, rT3) were examined in microsomal fractions of liver, heart, gill, brain, and skeletal muscle of 20-month-old Atlantic salmon induced to undergo parr-smolt transformation (PST) in late February and March by imposing a 16-hr photoperiod. All tissues showed negligible T3ORD activity. T3IRD activity was detected in both the liver (Km = 0.65 nM; Vmax = 15.5 pmol T3 deiodinated · hr-1 · mg microsomal protein-1) and brain of smolts, but not in gill, heart, or skeletal muscle. rT3ORD was detected in liver, brain, and muscle, and at very low levels in gill and heart. rT3IRD activity occurred to some extent in all tissues except brain. T3IRD activity changed in brain during PST, and was low in brain and liver of post-smolts examined in late October. We conclude that (i) deiodination of T3 proceeds exclusively through an IRD pathway, which may permit regulation of T3 degradation independently of the ORD pathway responsible for T3 formation; (ii) deiodination of rT3 proceeds mainly through an ORD pathway but rT3IRD activity does occur in some tissues; and (iii) the altered brain T3IRD activity during PST suggests regulation of T3 turnover in the brain at this time.

A furan fatty acid and indoxyl sulfate are the putative inhibitors of thyroxine hepatocyte transport in uremia.

We studied the effects of 3-carboxy-4-methyl-5-propyl-2-furan-propanoic acid (CMPF), indoxyl sulfate, and hippuric acid on iodide production by rat hepatocytes in primary cultures. We questioned whether these substances could explain the alteration of serum thyroid hormone parameters observed in renal failure. Iodide production from [125I]T4 by rat hepatocytes was significantly inhibited in the presence of serum from uremic patients. Serum concentrations of CMPF, indoxyl sulfate, and hippuric acid were markedly elevated in uremic patients. The minimum concentration that inhibited iodide production, when expressed as a molar ratio of the inhibitor to BSA, was 0.13 for CMPF, 0.53 for indoxyl sulfate, and 1.33 for hippuric acid. This molar ratio was lower than the corresponding mean molar ratio in uremic sera for CMPF (0.38) and indoxyl sulfate (0.63), while it was higher than that found for hippuric acid (0.85). The inhibition was reproduced when the inhibitors were added to normal human serum. The decreased iodide production was not due to the inhibition of deiodinase activity. The deiodination of rT3 by rat liver microsomes was unaffected by these inhibitors. Charcoal adsorption of uremic serum normalized the iodide production by hepatocytes. This normalization coincided with almost complete removal of CMPF and indoxyl sulfate, with a concomitant reduction of the free T4 fraction. Dialysis of uremic serum only partially restored iodide production. Even though indoxyl sulfate and hippuric acid were no longer detectable, a high concentration of CMPF remained in the serum. The serum free T4 fraction remained elevated in uremic patients after dialysis. Our studies indicate that CMPF and indoxyl sulfate in concentrations normally present in the serum of uremic patients inhibit cellular transport and subsequent deiodination of T4. These substances may account for the low total T3 level in uremic patients.

Molecular cloning of the selenocysteine-containing enzyme type I iodothyronine deiodinase

It has proved impossible to purify the three types of iodothyronine deiodinase to homogeneity by use of standard physiochemical techniques. However, using expression cloning with Xenopus oocytes, we were successful in isolating a cDNA coding for the rat type I iodothyronine deiodinase. An extremely puzzling aspect of this cDNA was an in-frame thymine-guanine-adenine (TGA) codon at position 126. By mutagenesis studies we showed that this coded for selenocysteine, consistent with the previously described requirement for selenium for hepatic and renal thyronine deiodination. The cloned enzyme can be expressed transiently in JEG, a human cell line, or COS, a monkey cell line, by gene-transfer techniques, and the expressed protein has the appropriate biochemical characteristics of type I iodothyronine deiodinase. These include a newly identified extreme sensitivity to competitive inhibition of reverse-3,5,3'-triiodothyronine (rT3) deiodination by gold. The identification of this elusive protein should prove helpful in understanding the process of thyroid hormone activation.

Cloning and in vitro expression of the human selenoprotein, type I iodothyronine deiodinase.

The type I 5' iodothyronine deiodinase (5' DI) catalyzes the deiodination of T4 to the biologically active hormone T3 and accounts for a significant fraction of its production. We have recently cloned the complementary DNA (cDNA) for the rat 5' DI, which contains the rare amino acid selenocysteine, and used this to screen human liver and kidney cDNA libraries to identify a human 5' DI cDNA clone. From these, we constructed a cDNA encoding a functional 5' DI. The 2222 base pair human 5' DI cDNA is approximately 200 nucleotides shorter than the 2.4-kilobase hybridizing band in Northern blots of human liver, kidney, and thyroid, because of missing 5' untranslated sequence and the poly A tail. The deduced amino acid sequence codes for a protein of 28.7 kilodaltons assuming the UGA codon at position 382 encodes selenocysteine, and is highly homologous (88% similarity) to the rat. We transiently expressed the 5' DI in COS-7 cells to establish that it encodes a functional enzyme and to study its kinetics. These show saturable deiodination of rT3 (Ka 0.52 +/- 0.04 mumol/L and Vmax 63.2 +/- 16.4 pmol min-1 mg-1). T4 and gold thioglucose are competitive inhibitors of rT3 deiodination. 6-n-Propylthiouracil (PTU) is an uncompetitive inhibitor (with rT3) and competitive inhibitor (with dithiothreitol) of rT3 deiodination. 6-n-Propylthiouracil inhibits T4 to T3 conversion. Labeling of COS-7 cells transiently transfected with the human 5' DI cDNA with bromoacetyl-125I-T3 demonstrates a 28-kilodalton protein. This indicates that in the human, as well as in the rat messenger RNA, the UGA encodes selenocysteine and translation terminates at the UAA codon at nucleotides 754 to 756. Reverse T3 and gold thioglucose (100 nmol/L) block bromoacetyl-125I-T3 labeling of the transiently expressed human and rat 5' DI proteins. These results demonstrate that the human 5' DI is a selenoprotein, analogous to the rat enzyme. Given the previously demonstrated critical role of the selenium atom in catalyzing deiodination by this protein, we conclude that this trace element is essential for normal thyroid hormone action in man.

Deiodination of Iodothyronine Sulfamates by Type I Iodothyronine Deiodinase of Rat Liver*

The substrate behavior of synthetic N-sulfonated iodothyronines (iodothyronine sulfamates, TiNS) for the type I deiodinase was compared with that of the naturally occurring 4'-O-sulfonated iodothyronines (iodothyronine sulfates, TiS), which have been shown to be deiodinated 40-200 times more efficiently than the native iodothyronines. Deiodination was studied in incubations of rat liver microsomes with unlabeled or 3' (5')-125I-labeled T4NS, rT3NS, T3NS, and 3,3'-T2NS at 37 C and pH 7.2 in the presence of 5 mM dithiothreitol. Reaction products were analyzed by RIA or Sephadex LH-20 and HPLC. Kinetic studies were performed under initial reaction rate conditions to determine the apparent Michaelis Menten (Km) constants and maximum velocity values. In contrast to T4S, which is converted only by inner ring deiodination (IRD), T4NS underwent both IRD and outer ring deiodination (ORD), similar to T4, but more rapidly. At 10 nM T4NS substrate, T3NS was the major product observed, while no rT3NS accumulated due to its rapid conversion to 3,3'-T2NS. At least one third of the 3,3'-T2NS was converted by IRD, unlike 3,3'-T2 which is a pure ORD substrate. The type I deiodination efficiencies of T4NS IRD and ORD were 17-fold higher than with T4, mainly due to approximately 32-fold lower apparent Km values. Deiodination of rT3, the preferred type I substrate, was not improved by sulfamation. T3NS and 3,3'-T2NS were deiodinated 4-10 times more efficiently than T3 and 3,3'-T2, respectively, due to 2- to 4-fold decreases in apparent Km values with a concomitant doubling of maximum velocity values. N-Sulfonation stimulates type I deiodination to a similar extent as other side-chain modifications that eliminate the positive charge of the nitrogen (e.g. iodothyroacetic acids). However, the effects are less dramatic than those induced by 4'-sulfation with respect to both efficiency and specificity of the catalytic process.

Selenocysteine confers the biochemical properties characteristic of the type I iodothyronine deiodinase

The conversion of thyroxine to 3,5,3'-triiodothyronine (T3) is the first step in thyroid hormone action, and the Type I iodothyronine deiodinase supplies most of this extrathyroidal T3 in the rat. We found that the cDNA coding for this enzyme contains an in-frame UGA encoding the rare amino acid selenocysteine. Using site-directed mutagenesis, we have converted selenocysteine to cysteine and expressed the wild-type and cysteine mutant enzymes in JEG-3 cells by transient transfection. The kinetic properties of the transiently expressed wild-type enzyme are nearly identical to those reported for rat liver Type I deiodinase. Substitution of sulfur for selenium causes a 10-fold increase in the Km of the enzyme for the favored substrate 3,3',5'-triiodothyronine (rT3), a 100-fold decrease in the sensitivity of rT3 deiodination to competitive inhibition by gold and a 300-fold increase in the apparent Ki for uncompetitive inhibition by 6-n-propylthiouracil. These results demonstrate that selenium is responsible for the biochemical properties which characterize Type I iodothyronine monodeiodination.

Influence of Fasting and Refeeding on 3,3′,5′-Triiodothyronine Metabolism in Man*

To determine the influence of prolonged fasting and refeeding on rT3 metabolism in man, five euthyroid obese subjects underwent a 13-day fast, followed by a refeeding period. Each patient received an iv dose of 25 muCi [125I]rT3 during the fed control period, on days 7 and 13 of the fast, and on the fourth day after refeeding with a regular diet. Serial blood and urine samples were obtained to determine serum rT3 clearance and production rates and the urinary tracer rT3 deiodination fraction. Significant increases in serum rT3 values were noted by day 7 and remained elevated for the duration of the fast (P less than 0.01). Normalization of rT3 levels occurred after 4 days of refeeding. Both 7 and 13 days of fasting decreased rT3 clearance [132.6 +/- 8.3 L/day (P less than 0.001) and 132.2 +/- 9.5 L/day (P less than 0.001), respectively] without changing rT3 production (36.8 +/- 5.3 and 33.0 +/- 3.7 nmol/D, respectively) compared to control values (207.0 +/- 10.9 L/day and 31.8 +/- 3.8 nmol/day, respectively). Refeeding did not restore rT3 clearance (151.2 +/- 6.9 L/day; P less than 0.002), but significantly reduced blood rT3 production (18.4 +/- 3.8 nmol/day; P less than 0.003). The fractional deiodination of rT3 was significantly reduced on day 7 (42.5 +/- 4.6%; P less than 0.01) and day 13 (41.9 +/- 3.7%; P less than 0.01) of fasting compared to the control value (69.2 +/- 2.8%), while refeeding only partially restored deiodination to baseline (48.4 +/- 5.1%; P less than 0.04). The clearance of rT3 was highly dependent on the fractional deiodination rate (r = 0.83; P less than 0.001). Although rT3 production remained constant during fasting, reduced rT3 production was seen on the fourth day of refeeding. This unique observation explained the fall in serum rT3 to prefasting levels after 4 days of refeeding when rT3 clearance was still inhibited. This study, in context with previous investigations, indicates that T4 conversion to circulating T3 and rT3 in fasting is a highly complex and multifaceted process requiring further investigation to elucidate the mechanism responsible for these alterations.

Iodothyronine Deiodinase Activities in FRTL5 Cells: Predominance of Type I 5′-Deiodinase*

FRTL5 cells, a thyroid follicular cell line derived from normal rat thyroid, has been extensively used as a model system to study various aspects of the physiology of the thyroid epithelium. The capacity of these cells to metabolize iodothyronines and to generate T3 from T4 has not been previously examined. Here we studied the deiodination of T4, T3, and rT3 in homogenates of FRTL5 cells. By far, these homogenates were more potent catalyzing the 5'-deiodination (outer ring) of T4 and rT3 than the inner ring deiodination of T4 or T3. Both the production of rT3 and the degradation of newly formed T3 from T4 were very limited. Thus, when T4 was used as a substrate, T3 and iodide accumulated in a linear fashion with time, and initially the amounts of iodide and T3 were approximately equal. rT3 and 3,3'-diiodothyronine were rapidly deiodinated by these homogenates, with the 3'-deiodination of 3,3'-diiodothyronine occurring at a slower rate than the 5'-deiodination of rT3. The iodothyronine 5'-deiodinase activity corresponded to type I, as indicated by the following: the Km for T4 and T3 was in the micromolar range; rT3 was a better substrate than T4 (maximum velocity = 101 vs. 19 pmol/min.mg protein; Km = 0.83 vs. 3.1 microM, respectively); and the kinetics of inhibition by 6n-propyl-2-thiouracil were uncompetitive and substrate dependent, suggesting ping-pong kinetics. The type I 5'-deiodinase activity of FRTL5 cells was distinctly stimulated by TSH. This stimulation seems to be mediated by cAMP and requires serum as a permissive factor. In conclusion, 1) FRTL5 cells exhibit both inner and outer ring iodothyronine-deiodinating activities; 2) iodothyronine 5'-deiodination is by far more active; 3) the 5'-deiodination has been characterized as type I deiodinase based on substrate preference, enzyme kinetics, and inhibitors; 4) in all respect iodothyronine deiodination by FRTL5 cell homogenates proceeded with marked similarity to that in homogenates or microsomes of thyroid glands from several species; and 5) the FRTL5 type I deiodinase is more active than that reported in thyroid tissue and as active as that reported in liver and kidney, the prototype of type I deiodinase-containing tissues. The present studies indicate that FRTL5 cells are an excellent model system to study cellular and biochemical aspects of the regulation of this enzyme as well as its regulation by TSH and putative serum factors.