Physiological role and regulation of iodothyronine deiodinases: a 2011 update.

T4 is a prohormone secreted by the thyroid. T4 has a long half life in circulation and it is tightly regulated to remain constant in a variety of circumstances. However, the availability of iodothyronine selenodeiodinases allow both the initiation or the cessation of thyroid hormone action and can result in surprisingly acute changes in the intracellular concentration of the active hormone T3, in a tissue- specific and chronologically-determined fashion, in spite of the constant circulating levels of the prohormone. This fine-tuning of thyroid hormone signaling is becoming widely appreciated in the context of situations where the rapid modifications in intracellular T3 concentrations are necessary for developmental changes or tissue repair. Given the increasing availability of genetic models of deiodinase deficiency, new insights into the role of these important enzymes are being recognized. In this review, we have incorporated new information regarding the special role played by these enzymes into our current knowledge of thyroid physiology, emphasizing the clinical significance of these new insights.

Type 2 iodothyronine deiodinase is highly expressed in human thyroid.

Type 2 iodothyronine deiodinase (D2) is a recently cloned selenodeiodinase thought to provide intracellular 3,5,3' triiodothyronine (T3) to a restricted group of tissues. We report here the presence of D2 mRNA in human thyroid at levels 50-150-fold higher than in placenta. Surprisingly, while type 1 deiodinase (D1) is known to be present in human thyroid, D2 has not been evaluated previously. D2 mRNA was especially high in thyroids from Graves' patients and in follicular adenomas. Stimulated thyroids had higher D2 to D1 mRNA ratios than normal or multinodular glands suggesting differential regulation of D1 and D2 expression. Microsomes from normal, Graves', and TSH-stimulated thyroids contained low Km D2 activity resistant to propylthiouracil (1 mM) or to inactivation by N-bromoacetyl T3, agents which block or inactivate D1. At 2 nM thyroxine (T4), 100 times the physiological-free T4 levels, 60-80% of T4 to T3 conversion in stimulated, but only 27% of that in normal thyroids, is catalyzed by D2. We conclude that intrathyroidal T4 to T3 conversion by D2 may contribute significantly to the relative increase in thyroidal T3 production in patients with Graves' disease, toxic adenomas, and, perhaps, iodine deficiency.

The type 2 iodothyronine deiodinase is essential for adaptive thermogenesis in brown adipose tissue.

Type 2 iodothyronine deiodinase (D2) is a selenoenzyme, the product of the recently cloned cAMP-dependent Dio2 gene, which increases 10- to 50-fold during cold stress only in brown adipose tissue (BAT). Here we report that despite a normal plasma 3,5,3'-triiodothyronine (T3) concentration, cold-exposed mice with targeted disruption of the Dio2 gene (Dio2(-/-)) become hypothermic due to impaired BAT thermogenesis and survive by compensatory shivering with consequent acute weight loss. This occurs despite normal basal mitochondrial uncoupling protein 1 (UCP1) concentration. In Dio2(-/-) brown adipocytes, the acute norepinephrine-, CL316,243-, or forskolin-induced increases in lipolysis, UCP1 mRNA, and O(2) consumption are all reduced due to impaired cAMP generation. These hypothyroid-like abnormalities are completely reversed by a single injection of T3 14 hours earlier. Recent studies suggest that UCP1 is primarily dependent on thyroid hormone receptor beta (TR beta) while the normal sympathetic response of brown adipocytes requires TR alpha. Intracellularly generated T3 may be required to saturate the TR alpha, which has an approximately fourfold lower T3-binding affinity than does TR beta. Thus, D2 is an essential component in the thyroid-sympathetic synergism required for thermal homeostasis in small mammals.

Type 3 lodothyronine deiodinase: cloning, in vitro expression, and functional analysis of the placental selenoenzyme.

Type 3 iodothyronine deiodinase (D3) catalyzes the conversion of T4 and T3 to inactive metabolites. It is highly expressed in placenta and thus can regulate circulating fetal thyroid hormone concentrations throughout gestation. We have cloned and expressed a 2.1-kb human placental D3 cDNA which encodes a 32-kD protein with a Km of 1.2 nM for 5 deiodination of T3 and 340 nM for 5' deiodination of reverse T3. The reaction requires DTT and is not inhibited by 6n-propylthiouracil. We quantitated transiently expressed D3 by specifically labeling the protein with bromoacetyl [125I]T3. The Kcat/Km ratio for 5 deiodination of T3 was over 1,000-fold that for 5' deiodination of reverse T3. Human D3 is a selenoenzyme as evidenced by (a) the presence of an in frame UGA codon at position 144, (b) the synthesis of a 32-kD 75Se-labeled protein in D3 cDNA transfected cells, and (c) the presence of a selenocysteine insertion sequence element in the 3' untranslated region of the mRNA which is required for its expression. The D3 selenocysteine insertion sequence element is more potent than that in the type 1 deiodinase or glutathione peroxidase gene, suggesting a high priority for selenocysteine incorporation into this enzyme. The conservation of this enzyme from Xenopus laevis tadpoles to humans implies an essential role for regulation of thyroid hormone inactivation during embryological development.

A novel retinoid X receptor-independent thyroid hormone response element is present in the human type 1 deiodinase gene.

We identified two thyroid hormone response elements (TREs) in the 2.5-kb, 5'-flanking region of the human gene encoding type 1 iodothyronine deiodinase (hdio1), an enzyme which catalyses the activation of thyroxine to 3,5,3'-triiodothyronine (T3). Both TREs contribute equally to T3 induction of the homologous promoter in transient expression assays. The proximal TRE (TRE1), which is located at bp -100, has an unusual structure, a direct repeat of the octamer YYRGGTCA hexamer that is spaced by 10 bp. The pyrimidines in the -2 position relative to the core hexamer are both essential to function. In vitro binding studies of TRE1 showed no heterodimer formation with retinoid X receptor (RXR) beta or JEG nuclear extracts (containing RXR alpha) and bacterially expressed chicken T3 receptor alpha 1 (TR alpha) can occupy both half-sites although the 3' half-site is dominant. T3 causes dissociation of TR alpha from the 5' half-site but increases binding to the 3' half-site. Binding of a second TR to TRE1 is minimally cooperative; however, no cooperativity was noted for a functional mutant in which the half-sites are separated by 15 bp, implying that TRs bind as independent monomers. Nonetheless, T3 still causes TR dissociation from the DR+15, indicating that dissociation occurs independently of TR-TR contact and that rebinding of a T3-TR complex to the 3' half-site occurs because of its slightly higher affinity. A distal TRE (TRE2) is found at bp -700 and is a direct repeat of a PuGGTCA hexamer spaced by 4 bp. It has typical TR homodimer and TR-RXR heterodimer binding properties. The TRE1 of hdio1 is the first example of a naturally occurring TRE consisting of two relatively independent octamer sequences which do not require the RXR family of proteins for function.

Selective inhibition of selenocysteine tRNA maturation and selenoprotein synthesis in transgenic mice expressing isopentenyladenosine-deficient selenocysteine tRNA.

Selenocysteine (Sec) tRNA (tRNA([Ser]Sec)) serves as both the site of Sec biosynthesis and the adapter molecule for donation of this amino acid to protein. The consequences on selenoprotein biosynthesis of overexpressing either the wild type or a mutant tRNA([Ser]Sec) lacking the modified base, isopentenyladenosine, in its anticodon loop were examined by introducing multiple copies of the corresponding tRNA([Ser]Sec) genes into the mouse genome. Overexpression of wild-type tRNA([Ser]Sec) did not affect selenoprotein synthesis. In contrast, the levels of numerous selenoproteins decreased in mice expressing isopentenyladenosine-deficient (i(6)A(-)) tRNA([Ser]Sec) in a protein- and tissue-specific manner. Cytosolic glutathione peroxidase and mitochondrial thioredoxin reductase 3 were the most and least affected selenoproteins, while selenoprotein expression was most and least affected in the liver and testes, respectively. The defect in selenoprotein expression occurred at translation, since selenoprotein mRNA levels were largely unaffected. Analysis of the tRNA([Ser]Sec) population showed that expression of i(6)A(-) tRNA([Ser]Sec) altered the distribution of the two major isoforms, whereby the maturation of tRNA([Ser]Sec) by methylation of the nucleoside in the wobble position was repressed. The data suggest that the levels of i(6)A(-) tRNA([Ser]Sec) and wild-type tRNA([Ser]Sec) are regulated independently and that the amount of wild-type tRNA([Ser]Sec) is determined, at least in part, by a feedback mechanism governed by the level of the tRNA([Ser]Sec) population. This study marks the first example of transgenic mice engineered to contain functional tRNA transgenes and suggests that i(6)A(-) tRNA([Ser]Sec) transgenic mice will be useful in assessing the biological roles of selenoproteins.

Interplay between termination and translation machinery in eukaryotic selenoprotein synthesis.

Termination of translation in eukaryotes is catalyzed by eRF1, the stop codon recognition factor, and eRF3, an eRF1 and ribosome-dependent GTPase. In selenoprotein mRNAs, UGA codons, which typically specify termination, serve an alternate function as sense codons. Selenocysteine incorporation involves a unique tRNA with an anticodon complementary to UGA, a unique elongation factor specific for this tRNA, and cis-acting secondary structures in selenoprotein mRNAs, termed SECIS elements. To gain insight into the interplay between the selenocysteine insertion and termination machinery, we investigated the effects of overexpressing eRF1 and eRF3, and of altering UGA codon context, on the efficiency of selenoprotein synthesis in a transient transfection system. Overexpression of eRF1 does not increase termination at naturally occurring selenocysteine codons. Surprisingly, selenocysteine incorporation is enhanced. Overexpression of eRF3 did not affect incorporation efficiency. Coexpression of both factors reproduced the effects with eRF1 alone. Finally, we show that the nucleotide context immediately upstream and downstream of the UGA codon significantly affects termination to incorporation ratios and the response to eRF overexpression. Implications for the mechanisms of selenocysteine incorporation and termination are discussed.

Multihormonal regulation of the human, rat, and bovine growth hormone promoters: differential effects of 3',5'-cyclic adenosine monophosphate, thyroid hormone, and glucocorticoids.

We have analyzed the effects of a variety of hormones on activity of the rat GH (rGH), human GH, (hGH), and bovine GH (bGH) promoters. After transient transfection of rat pituitary tumor cells, all three promoters are induced by addition of 8-bromo-cAMP. Sequences required for the cAMP responsiveness of the hGH and rGH promoter lie within 183 base pairs of the mRNA start site. Although the rGH promoter is thyroid hormone (T3) responsive in this system, a construct containing 2.7 kilobases of the hGH promoter 5'-flanking sequences is not. Since we also found that the bGH promoter is T3 responsive in these cells, the hGH results are not likely to be due to a species specific factor required for induction in rat pituitary cells. The hGH promoter is weakly induced by dexamethasone whereas the rGH promoter does not respond to glucocorticoids. The hGH and rGH promoters are not responsive to TRH. These results illustrate the potential heterogeneity in hormonal responses of the same gene in different species.

Dominant negative inhibition by mutant thyroid hormone receptors is thyroid hormone response element and receptor isoform specific.

The heterogeneity of tissue-specific manifestations of generalized resistance to thyroid hormone (GRTH) could result from differential interactions between the mutant thyroid hormone (T3) receptor-beta (TR beta) on T3 response elements (TREs) in different T3-responsive genes. To explore this hypothesis, the mutant TR beta associated with kindred A, P448H; a TR beta mutant, P448L; and a comparable TR alpha mutant (P398H) were tested for intrinsic function and for inhibition of wild-type TR alpha- and -beta-induced expression from four structurally distinct TREs, the rGH ABC*, the rGH palindrome (PAL), the rat malic enzyme (ME), and the chicken lysozyme silencer F2 (F2). The relative function of the mutants was similarly reduced on the four TREs studied and was T3 concentration dependent. The TR alpha mutant retained the intrinsically greater potency characteristic of this isoform, but remained impaired with respect to wild-type TR alpha even at 500 nM T3. In general, dominant negative inhibition of wild-type TR alpha and -beta function was dependent upon the T3 concentration, as expected from the decreased affinity for ligand conferred by this mutation. A T3 concentration sufficient to relieve the inhibition of wild-type TR function on the ABC*, PAL, and ME TREs (50 nM) had no effect on inhibition of the F2 TRE by the mutant TRs. Receptor isoform preferential inhibition was observed on the ABC*, PAL, and ME TREs by the mutant TRs. Thus, both TRE structure and the isoform of endogenously active receptor could determine the degree of inhibition of a specific gene in GRTH individuals. Further, the lack of dominant negative potentials does not explain the absence of TR alpha mutations in GRTH kindreds.

Structural and functional differences in the dio1 gene in mice with inherited type 1 deiodinase deficiency.

The type 1 deiodinase (D1) provides the major portion of the circulating T3 in vertebrates. In C3H and certain other inbred mice, liver and kidney D1 activity is 5- to 10-fold lower than in the common phenotype, C57. The lower D1 levels are paralleled by a decreased normal-sized dio1 mRNA and hyperthyroxinemia. Low activity cosegregates with a restriction fragment length variant (RFLV) in both inbred and recombinant strains, indicating it is due to differences in the dio1 gene. The exonic structure and the deduced amino acid sequences are identical for both strains and highly homologous to that of the rat. The RFLV is due to an approximately 150-base pair expansion of repetitive sequences in the second intron of the C3H gene, but this segment does not differentially affect the transient expression of a human GH gene. The promoter and 5'-flanking regions of the C3H and C57 dio1 genes are very similar and are GC rich without TATA or CCAAT boxes. However, functional assays of 1.5-kilobase 5'-flanking dio1-CAT constructs showed 2- to 3-fold higher activity of the C57-CAT constructs. Deletion mutants showed that sequences between -705 and -162 were the cause of this. In this region, the only major difference between the two genes is a 21-base pair insert containing five CTG repeats in the C3H promoter. This difference also cosegregates with low D1 activity and the intron RFLV in four other mouse strains. The correlation of the CTG repeat insert with both in vitro and in vivo expression and the absence of other significant sequence differences in the 5'-flanking region argue that this is the major explanation for the impaired expression of the dio1 gene and the resulting hyperthyroxinemia of the C3H mouse.

Selective proteolysis of human type 2 deiodinase: a novel ubiquitin-proteasomal mediated mechanism for regulation of hormone activation.

We investigated the mechanism by which T4 regulates its activation to T3 by the type 2 iodothyronine deiodinase (D2). D2 is a short- lived (t1/2 50 min), 31-kDa endoplasmic reticulum (ER) integral membrane selenoenzyme that generates intracellular T3. Inhibition of the ubiquitin (Ub) activating enzyme, E1, or MG132, a proteasome blocker, inhibits both the basal and substrate-induced acceleration of D2 degradation. Using a catalytically active transiently expressed FLAG-tagged-NH2-D2, we found rapid synthesis of high molecular mass (100-300 kDa) Ub-D2 conjugates that are catalytically inactive. Ub-D2 increases when cells are exposed to D2 substrate or MG132 and disappears rapidly after E1 inactivation. Fusion of FLAG epitope to the COOH terminus of D2 prolongs its half-life approximately 2.5-fold and increases the levels of active and, especially, Ub-D2. This indicates that COOH-terminal modification interferes with proteasomal uptake of Ub-D2 that can then be deubiquitinated. Interestingly, the type 1 deiodinase, a related selenoenzyme that also converts T4 to T3 but with a half-life of >12 h, is inactivated but not ubiquitinated or degraded after exposure to substrate. Thus, ubiquitination of the ER-resident enzyme D2 constitutes a specific posttranslational mechanism for T4 regulation of its own activation in the central nervous system and pituitary tissues in which D2-catalyzed T4 to T3 conversion is the major source of intracellular T3.

The human, but not rat, dio2 gene is stimulated by thyroid transcription factor-1 (TTF-1).

Types 1 and 2 iodothyronine deiodinases (D1 and D2) catalyze the production of T(3) from T(4). D2 mRNA is abundant in the human thyroid but very low in adult rat thyroid, whereas D1 activity is high in both. To understand the molecular regulation of these genes in thyroid cells, the effect of thyroid transcription factor 1 (TTF-1) and the paired domain-containing protein 8 (Pax-8) on the transcriptional activity of the deiodinase promoters were studied. Both the approximately 6.5-kb hdio2 sequence and its most 3' 633 bp were activated 10-fold by transiently expressed TTF-1 in COS-7 cells, but the hdio1 was unaffected. Surprisingly, the response of the rdio2 gene to TTF-1 was only 3-fold despite the 73% identity with the proximal 633-bp region of hdio2 including complete conservation of a functional cAMP response element at -90. Neither human nor rat dio2 nor human dio1 was induced by Pax-8. The binding affinity of four putative TTF-1 binding sites in hdio2 were compared by a semiquantitative gel retardation assay using in vitro expressed TTF-1 homeodomain protein. Only two sites, D and C1 (both of which are absent in rdio2), had significant affinity. Functional analyses showed that both sites are required for the full response to TTF-1. These results can explain the differential expression of dio2 in thyroid and potentially other tissues in humans and rats.

Differential capacity of wild type promoter elements for binding and trans-activation by retinoic acid and thyroid hormone receptors.

Retinoic acid receptor (RAR) and thyroid hormone receptor (T3R) are structurally similar and can bind as homodimers or T3R-RAR heterodimers to a single synthetic DNA response element. The interaction of these two types of receptors with wild type elements, however, has not been systematically investigated. Promoter elements from genes regulated by retinoic acid (RA) or thyroid hormone (T3) were tested for response to T3 and RA in transient transfections in both JEG and COS cells. The elements were classified as primarily responsive to RA or to T3 or responsive to both ligands. Binding of highly purified RAR alpha and T3R alpha to the various elements was assessed using the gel shift assay. Those elements predominantly responsive to one ligand showed preferential binding to the appropriate receptor. A series of point mutations were introduced into the rat GH T3 response element to further define sequence requirements for response to both RA and T3. Down-mutations in any of the three hexamers (previously demonstrated to be required for full response to T3 and full binding of T3R) also decreased RA induction and RAR binding. However, only one of two sets of up-mutations for T3 response also increased RA induction, demonstrating differences in hexamer preference between RAR and T3R. Variation in spacing of the three hexamers did not influence RA vs. T3 induction or RAR vs. T3R binding according to the predictions of a simple hexamer spacing model. There was a strong correlation between the extent of T3R dimer binding and strength of T3 induction for a subset of elements studied in JEG cells (r = 0.97, P < 0.01) and a weaker but significant correlation in COS cells (r = 0.65, P < 0.05)). In contrast, RAR dimer binding by the wild type elements did not quantitatively correlate with RA induction in either JEG (r = 0.13, P > 0.05) or COS cells (r = 0.21, P > 0.05). These results suggests that RAR interacts with a heterodimer partner(s) which influences binding site specificity, whereas T3R heterodimer partner(s) is less likely to alter binding site recognition. The observed difference in COS and JEG cells as well as the weak T3R binding-function relationship of the malic enzyme element, however, suggest that the influence of T3R heterodimer partner(s) on binding site specificity is likely to vary with cell type and the specific element tested.

Mutations of the rat growth hormone promoter which increase and decrease response to thyroid hormone define a consensus thyroid hormone response element.

We have previously identified sequences required for thyroid hormone (T3) induction of the rat GH (rGH) promoter, which lie in a region from -188 to -164 upstream of the mRNA start site. Within this region, Domains A, -189 to -184 and B, -179 to -174, are imperfect direct repeats, and domain C, -172 to -167, is a divergent inverted copy that matches the A domain at 4/6 positions. A series of synthetic mutant versions of this sequence were inserted upstream of a truncated rGH promoter, or as a replacement for wild-type sequences in a synthetic 237 base pair rGH promoter or upstream of the heterologous thymidine kinase promoter. Mutations changing the B domain to a perfect copy of the A domain significantly increased T3 induction (21.3-fold) relative to the wild type (3.6-fold). A single point mutation making the C domain a better match to the A domain also increased T3 induction to 16.2-fold. Combining this up-mutation with any of three down-mutations in the A, B, or C domains strongly decreased response, showing that all three domains contribute to the amplified T3 response. Binding affinity of the various mutant oligonucleotides was assessed using in vitro translated receptor and affinity paralleled the functional responses for most binding site mutations. Requirements for in vitro binding were, however, less rigorous than those for functional T3 induction. Based on these results, we propose a consensus T3 receptor binding half-site, AGGT(C/A)A, at least two copies of which are required for a T3 response.(ABSTRACT TRUNCATED AT 250 WORDS)

Capacity for cooperative binding of thyroid hormone (T3) receptor dimers defines wild type T3 response elements.

Thyroid hormone response elements (T3REs) have been identified in a variety of promoters including those directing expression of rat GH (rGH), alpha-myosin heavy chain (rMHC), and malic enzyme (rME). A detailed biochemical and genetic analysis of the rGH element has shown that it consists of three hexamers related to the consensus [(A/G)GGT(C/A)A]. We have extended this analysis to the rMHC and rME elements. Binding of highly purified thyroid hormone receptor (T3R) to T3REs was determined using the gel shift assay, and thyroid hormone (T3) induction was measured in transient tranfections. We show that the wild type version of each of the three elements binds T3R dimers cooperatively. Mutational analysis of the rMHC and rME elements identified domains important for binding T3R dimers and allowed a direct determination of the relationship between T3R binding and function. In each element two hexamers are required for dimer binding, and mutations that interfere with dimer formation significantly reduce T3 induction. Similar to the rGH element, the rMHC T3RE contains three hexameric domains arranged as a direct repeat followed by an inverted copy, although the third domain is weaker than in rGH. All three are required for full function and T3R binding. The rME T3RE is a two-hexamer direct repeat T3RE, which also binds T3R monomer and dimer. Across a series of mutant elements, there was a strong correlation between dimer binding in vitro and function in vivo for rMHC (r = 0.99, P less than 0.01) and rME (r = 0.67, P less than 0.05) T3REs. Our results demonstrate a similar pattern of T3R dimer binding to a diverse array of hexameric sequences and arrangements in three wild type T3REs. Addition of nuclear protein enhanced T3R binding but did not alter the specificity of binding to wild type or mutant elements. Binding of purified T3R to T3REs was highly correlated with function, both with and without the addition of nuclear protein. T3R dimer formation is the common feature which defines the capacity of these elements to confer T3 induction.

Effects of varying the position of thyroid hormone response elements within the rat growth hormone promoter: implications for positive and negative regulation by 3,5,3'-triiodothyronine.

The thyroid hormone response element (T3RE) of the rat GH (rGH) promoter is located at -188 to -165 relative to the mRNA start site (TSS). Similar sites have been identified in other genes regulated by T3. We have investigated some of these T3REs in positions within the rGH promoter to assess the relative influences of DNA-binding site and position on positive and negative regulation by T3. Synthetic oligonucleotides were used with sequences from the rGH T3RE and proposed negative T3REs (nT3RE) from the rat and human alpha-subunit and rat beta TSH genes. The nT3REs were placed in the background of the wild-type rGH promoter in two positions, at -55 and down-stream of the TSS, with up- and down-mutations of the rGH T3RE. Rat GH T3RE elements were placed 700 basepairs up-stream of a basal rGH promoter and some also at the -55 and TSS positions. Constructions were tested in a transient transfection assay in rat pituitary tumor cells. Two copies of the rGHPAL (palindromic T3RE) placed 700 basepairs up-stream of the rGH promoter conferred 10-fold T3 induction. In the -55 position, the rGHPAL increased T3 induction compared to that in controls, whereas a fragment from the rat and human alpha-subunit gene in the same position reduced induction. Negative T3REs from rat beta TSH and human alpha-subunit reduced T3 induction 50% when placed at the TSS position of a rGH promoter containing an up-mutant T3RE. The T3REPAL placed at the same site increased T3 induction.(ABSTRACT TRUNCATED AT 250 WORDS)

The role of selenocysteine 133 in catalysis by the human type 2 iodothyronine deiodinase.

Human type 2 iodothyronine deiodinase (hD2) catalyzes the activation of T4 to T3. D2, like types 1 and 3 deiodinases, contains selenocysteine (Sec) in the highly conserved active center at position 133. To evaluate the contribution of Sec133 to the catalytic properties of hD2, we generated mutants in which cysteine (Cys) or alanine (Ala) replaced Sec133. The Km (T4) of Cys133D2 was 2.1 microM, strikingly higher than that of native D2 (1.4 nM). In contrast, the relative turnover number was 10-fold lower for Cys133D2, illustrating the greater potency of Se than S in supporting catalysis. The AlaD2 mutant was inactive. Studies in intact cells transiently expressing the native or Cys133D2 enzyme exhibited saturation kinetics expected from the Km as measured under in vitro conditions, indicating rapid equilibration of extracellular and intracellular T4. Blockade of the NTCP, OATP1-3, and LST-1 transporters with 10 mM sodium taurocholate did not alter the deiodination rate of T4 by Cys133D2 transiently expressed in intact cells, suggesting that intracellular transport of T4 is not rate limiting. These results illustrate that selenium plays a critical role in deiodination catalyzed by hD2.

Pituitary cells respond to thyroid hormone by discrete, gene-specific pathways.

Type 1 iodothyronine deiodinase (D1) converts T4 to T3, the active thyroid hormone, by removal of the outer ring iodine. Previous studies in liver and thyroid cells have shown that T3 regulates Type 1 deiodinase (dio1) gene expression by a mechanism not requiring ongoing protein synthesis. For certain T3-regulated genes, such as rat GH, T3-induced transcription is blocked by protein synthesis inhibitors. Because the somatotrope tumor cell lines express both dio1 and GH, we compared these two positively T3-regulated genes to establish whether cycloheximide blockade of T3 effects is cell-type or gene specific. In these cells, the T3 stimulation of dio1 messenger RNA (mRNA) is not blocked by cycloheximide, whereas the T3 effect on GH mRNA synthesis is eliminated. Other differences between these two genes were also noted. Retinoic acid does not alter dio1 gene expression or the response to T3 but increases GH and synergizes with T3. Dexamethasone alone had no effect on dio1 mRNA but did enhance the effect of T3 on both dio1 and GH. These results point to distinct pathways for T3 induction of mRNA synthesis from different genes within the same cell.

Studies of the hormonal regulation of type 2 5'-iodothyronine deiodinase messenger ribonucleic acid in pituitary tumor cells using semiquantitative reverse transcription-polymerase chain reaction.

We developed a sensitive competitive RT-PCR technique for quantitating the ratio of D2 to cyclophilin messenger RNA (mRNA) and used this to study type 2 deiodinase (D2) mRNA regulation. Hyperthyroidism in rats causes a 2- to 3-fold reduction in anterior pituitary and medial basal hypothalamus (MBH). Thyroid hormone (T3) withdrawal increased the D2/cyclophilin ratio 2- to 3-fold over 48 h in both GC and GH4C1 cells. T3 additional reduced D2 gene transcription by 50% over 2 h and about 30% over the next 2 h. D2 mRNA half-life is 2 h and is not affected by T3, indicating that its effect is due to suppression of D2 gene transcription. The T3 effect did not require new protein synthesis. Longer treatment with T3 led to a maximum decrease of 70% in D2 mRNA, indicating that there is also a T3-independent transcriptional component of the D2 gene. 3,3',5'-Triiodothyronine (reverse T3) caused a slight increase D2 mRNA over 24 h but an 80-90% decrease in D2 activity, indicating that it acts posttranscriptionally. Dexamethasone, 8 Br-cAMP, and TRH also caused modest increases in D2 mRNA in pituitary tumor cells. We conclude that D2 gene transcription has both T3-dependent and T3-independent components. Thus, posttranscriptional effects of D2 substrates such as T4 will be required for complete feedback inhibition of D2 activity. The short half-life of D2 mRNA and D2 protein explains the rapid response of D2 activity to thyroid hormone administration.

Molecular biological and biochemical characterization of the human type 2 selenodeiodinase.

Type 2 deiodinase (D2) is a low K(m) iodothyronine deiodinase that catalyzes the removal of a single iodine from the phenolic ring of T4 or rT3. We sequenced and subcloned the open reading frame from a partial complementary DNA (cDNA) clone (2.1 kilobases) prepared by Genethon (Z44085) from a human infant brain cDNA library. The open reading frame encodes a putative 273-amino acid protein of 31 kDa with greater than 70% similarity to the Rana catesbeiana D2 protein. Transient expression of the cDNA produces a low K(m) (5 nM for T4; 8 nM for rT3) propylthiouracil- and gold thioglucose-resistant 5'-deiodinase in 293-HEK cells. Human D2, like human type 1 (D1) and type 3 (D3) deiodinases, is a selenoenzyme, as evidenced by 1) the presence of two in-frame UGA codons (positions 133 and 266), 2) the synthesis of a 31-kDa 75Selabeled protein in D2 cDNA-transfected cells, and 3) the requirement for a 3'-selenocysteine incorporation sequence element for its translation. Unlike D1 and D3, we were not able to covalently label overexpressed D2 with N-bromoacetyl [125I]T3 or -T4. We found that the human D2 messenger RNA is 7-8 kilobases and is expressed in brain, placenta, and, surprisingly, cardiac and skeletal muscle. Type 2 deiodinase activity was also present in human skeletal muscle. These results indicate that there are unique features of D2 that distinguish it from the two other selenodeiodinases. The expression of D2 in muscle suggests that it could play a role in peripheral, as well as intracellular, T3 production.