Is Triac the Primordial Bioactive Thyroid Hormone?

T H Y R O I D - T R H - T S H

A Nonselenoprotein from Amphioxus Deiodinates Triac But Not T3: Is Triac the Primordial Bioactive Thyroid Hormone? Wim Klootwijk, Edith C. H. Friesema, and Theo J. Visser Department of Internal Medicine, Erasmus University Medical Center, 3000 CA Rotterdam, The Netherlands

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 (TA3) but not by T3. 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 125I-labeled substrates and dithiothreitol as cofactor. We could not detect deiodination of T4, T3, or rT3 by bfDy but observed rapid and selective inner ring deiodination (inactivation) of TA3 and 3,3⬘,5,5⬘-tetraiodothyroacetic acid (TA4). Deiodination of TA3 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 TA3 was inhibited dose dependently by unlabeled TA3⬇TA4⬎T4⬇T3. 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䡠mg protein for TA3 and TA4, respectively. bfDy was not inhibited by propylthiouracil and iodoacetate and only weakly by goldthioglucose and iopanoic acid. In conclusion, we demonstrate rapid inactivation of TA3 and TA4 but not of T3 and T4 by the first reported natural nonselenodeiodinase. Our findings support the hypothesis that TA3 is a primordial bioactive TH. (Endocrinology 152: 3259 –3267, 2011)

I

n mammals, thyroid hormone (TH) is very important for the growth and development of different tissues, in particular the brain, and for the regulation of energy metabolism in many tissues throughout life (1, 2). Most of these actions are initiated by binding of T3 to its nuclear receptors, which alters the transcription of TH-sensitive genes (1, 2). However, in humans and other animals, the main secretory product of the thyroid follicles is T4, and most T3 is produced by outer ring deiodination (ORD) of T4 in peripheral tissues (3). Conversely, inner ring deiodination (IRD) of T4 and T3 produces the receptor-inactive metabolites rT3 and 3,3⬘-diiodothyronine (T2), respectively. In mammals, three iodothyronine deiodinases (D1–D3) have been identified (3).

D1 is expressed predominantly in liver, kidney, and thyroid and has both ORD and IRD activity. It is thought to be a major site for production of plasma T3 and clearance of plasma rT3. D2 has only ORD activity and is expressed importantly in brain, pituitary, brown adipose tissue, thyroid, and skeletal muscle. It plays an important role in the local production of T3 in brain, pituitary, and brown adipose tissue, but the enzyme in thyroid and muscle may also contribute to circulating T3 levels. D3 has only IRD activity and is highly expressed in different fetal tissues, the pregnant uterus and placenta, and in adult brain and skin. It is the major enzyme for inactivation of T4 and T3 and for production of rT3 (3). Mammals have single genes for each deiodinase. Fish also have single genes coding for D1 or D2, but most fish have

ISSN Print 0013-7227 ISSN Online 1945-7170 Printed in U.S.A. Copyright © 2011 by The Endocrine Society doi: 10.1210/en.2010-1408 Received December 7, 2010. Accepted May 16, 2011. First Published Online June 7, 2011

Abbreviations: DTT, Dithiothreitol; GTG, goldthioglucose; hs, Homo sapiens; IAc, iodoacetate; IOP, iopanoic acid; IRD, inner ring deiodination; Km, Michaelis-Menten constant; ORD, outer ring deiodination; PTU, 6-propyl-2-thiouracil; T2, diiodothyronine; TA3, 3,3⬘,5triiodothyroacetic acid; TA4, 3,3⬘,5,5⬘-tetraiodothyroacetic acid; TH, thyroid hormone; TR, TH receptor; Vmax, maximum velocity.

Endocrinology, August 2011, 152(8):3259 –3267

endo.endojournals.org

3259

3260

Klootwijk et al.

Nonselenodeiodinase from Amphioxus

two genes coding for different D3 isoforms (4) (our unpublished observations). All deiodinases studied so far are homologous selenoproteins, with a Sec in the catalytic center. Replacement of this Sec residue by Cys results in a two orders of magnitude drop in catalytic activity, whereas replacement with Leu or Ala results in a completely inactive protein (3, 5–7). In all mammals, D2 contains a second Sec residue near the C terminus, which is not essential for catalytic activity (8). The function of this second Sec residue is unknown. Other pathways for the metabolism of TH involve the conjugation of the phenolic hydroxyl group with sulfate or glucuronic acid, as well as the conversion of the Ala side chain to acetic acid (9). The latter pathway results in the production of metabolites, such as 3,3⬘,5,5⬘-tetraiodothyroacetic acid (TA4) and 3,3⬘,5-triiodothyroacetic acid (TA3). A possible route for the production of iodothyroacetic acid metabolites involves decarboxylation of iodothyronines to iodothyronamines and subsequent oxidative deamination of these intermediates. Recently, interesting pharmacological properties have been reported for thyronamine and 3-iodothyronamine, but the quantitative importance and mechanism of the generation of these metabolites remain to be established (10). In animals, such as fish and amphibians, TH stimulates the metamorphosis of a larval stage into an adult animal (11). This is also the case for the cephalochordate amphioxus, an invertebrate living in warmer coastal waters, where an asymmetric larva is transformed into a symmetric adult organism. This has been studied extensively in Branchiostoma floridae, the metamorphosis of which is stimulated by TH and delayed by TH synthesis inhibitors (11–13). Recently, the genome of B. floridae has been deciphered (14, 15). Examination of this genome has indicated the presence of multiple genes coding for deiodinase-like proteins. Most of these genes code for selenoproteins, showing high sequence identity with deiodinases from vertebrates, in particular around the Sec residue (12). However, one of these genes appears to code for a nonselenoprotein, where the Sec residue is replaced by Cys. The cDNA from three of these Branchiostoma genes have been cloned. One of these, bfDy, codes for this nonselenoprotein; the two others, bfDt and bfDx, code for selenodeiodinases (Fig. 1). We have subcloned bfDy in an expression vector and studied its catalytic activity in transfected mammalian cells. This was investigated initially using 125I-labeled T4, T3, and rT3 as substrates. When this study was in progress, a paper was published by the group of Laudet, showing that the TH receptor (TR) from Branchiostoma is not stimulated by T3 but is potently activated by TA3 (13). This prompted us to also test the possible deiodination of labeled TA4 and TA3 by this Branchiostoma enzyme.


Materials and Methods Materials The amphioxus bfDy cDNA clone bfad027g07 inserted in pDONR222 was obtained courtesy of professor Yuji Kohara (National Institute of Genetics, Mishima, Japan) and subcloned in the pcDNA3.2/v5-DEST (Gateway) expression vector using LR Clonase from Invitrogen (Breda, The Netherlands). pCIneo clones containing cDNA coding for wild-type human D3 [Homo sapiens (hs)D3.pCI-neo] or its Sec144Cys mutant (hsD3Cys.pCI-neo) were obtained as previously described (7). Oligonucleotides, DMEM/F12 medium, and fetal bovine serum were obtained from Invitrogen and FuGENE 6 transfection reagent from Roche Diagnostics (Almere, The Netherlands). Unlabeled iodothyronine derivatives were obtained from Henning GmbH (Berlin, Germany), Sigma-Aldrich (Zwijndrecht, The Netherlands), and Calbiochem (Schiphol, The Netherlands). Na125I was obtained from MDS Nordion (Fleurus, Belgium). Iodothyronines were labeled with 125I using the chloramine-T method: [3⬘-125I]T3 was produced from 3,5-T2, [3⬘,5⬘-125I]rT3 from 3,3⬘-T2; [3⬘,5⬘-125I]T4 from T3; [3⬘-125I]TA3 from 3,5-TA2; and [3⬘,5⬘-125I]TA4 from TA3. The radioactive products were isolated on Sephadex LH-20 minicolumns, and the purity was confirmed by HPLC analysis as described below. 6-Propyl-2-thiouracil (PTU), iodoacetate (IAc), and goldthioglucose (GTG) were obtained from Sigma-Aldrich, and iopanoic acid (IOP) from Sterling-Winthrop (Amsterdam, The Netherlands).

Bioinformatics Partial cDNA sequences for bfDy (accession codes BW699364 and BW718012), bfDt (FE585018 and FE585017), bfDx (BW885150 and BW942918), and genomic sequences were obtained from GenBank. The full insert sequence of the bfDy.pcDNA3.2 clone was determined on an automated ABI 3100 capillary sequencer, using the Big Dye Terminator Cycle Sequencing method (Applied Biosystems, Nieuwerkerk a/d IJssel, The Netherlands). Reference amino acid sequences were also obtained from GenBank for human (Homo sapiens) D1 (NP_000783), D2 (NP_000784), and D3 (NP_001353); mouse (Mus musculus) D1 (NP_031886), D2 (NP_034180), and D3 (NP_742117); chicken (Gallus gallus) D1 (NP_001091083), D2 (NP_989445), and D3 (NP_001116120); frog (Xenopus laevis) D1 (NP_001089136), D2 (AAK40121), D3a (NP_001081332), and D3b (NP_001087559; zebrafish (danio rerio) D1 (NP_001007284), D2 (NP_997954), D3a (NP_001171406), and D3b (EH432430, EH458785, and other EST sequences). Amino acid sequences were aligned using the ClustalW2 program available on the European Bioinformatics website (http:// www.ebi.ac.uk/Tools/clustalw2/index.html). The amino acid alignment was edited using the freeware GeneDoc program (http://www.nrbsc.org/gfx/genedoc/). Molecular phylogenetic analysis was done with the maximum likelihood method based on the Poisson correction model with bootstrap using MEGA5 freeware (http://megasoftware.net/) (16). The bfDy amino acid sequence was analyzed for the presence of possible transmembrane domains using different prediction programs available on the ExPASy proteomics server (http://expasy.org/).



3261

FIG. 1. A, Nucleotide and amino acid sequences of bfDy. The catalytic Cys residue is indicated in red. B, Phylogenetic tree of amphioxus and vertebrate deiodinases based on complete amino acid sequences and constructed using the maximum likelihood method with 500 bootstrap replications. The percentage of replicate trees in which the associated taxa clustered together is indicated next to the branches. C, Alignment of the core amino acid sequences of bfDy, bfDx, bfDt, and representative vertebrate deiodinase sequences (mm, Mouse; gg, chicken; xl, frog; and dr, zebrafish). The catalytic Cys/Sec residue is indicated by the arrow.

Expression of bfDy and analysis of deiodinase activity COS1 cells were cultured at 37 C under 5% CO2 in air in six-well culture dishes with DMEM/F12 medium plus 9% heatinactivated fetal bovine serum and 100 nM sodium selenite. Cells were transfected with 1 ␮g empty pcDNA3 vector or bfDy.p-

cDNA3.2 using 3 ␮l FuGENE 6. After 2 d, cells were lyzed in ice-cold 0.1 M phosphate (pH 7.2), 2 mM EDTA (PE buffer) plus 1 mM dithiothreitol (DTT), briefly sonicated, and stored in aliquots at ⫺80 C. Protein concentrations were determined with the Bradford method using the Bio-Rad (Veenendaal, The Netherlands) protein assay reagent and BSA as a standard.

3262

Klootwijk et al.


Deiodinase activity was determined by incubation of transfected cell lysates (25–500 ␮g protein/ml) usually for 60 min at 25 C with 1 nM 125I-labeled substrate in 0.1 ml PE buffer with 10 mM DTT. The incubation was stopped by addition of 0.1 ml ice-cold ethanol, and the mixture was kept on ice for 30 min. After centrifugation, 0.1 ml of the supernatant was mixed with 0.1 ml 0.02 M ammonium acetate (pH 4), and 0.1 ml of the mixture was applied to a 250 ⫻ 4.6 mm Symmetry C18 column connected to an Alliance HPLC system (Waters, Etten-Leur, The Netherlands) and eluted using a 20-min gradient of 28 –90% acetonitrile in 0.02 M ammonium acetate (pH 4) at a flow rate of 1.2 ml/min. Radioactivity in the eluate was monitored online using a Radiomatic A-500 flow scintillation detector (Packard, Meriden, CT).

Results Structural characteristics of bfDy Figure 1A shows the nucleotide sequence and corresponding amino acid sequence for bfDy. The cDNA contains 1328 nucleotides and the coding sequence 801 nucleotides, coding for a protein of 266 amino acids. Blasting of amphioxus genomic sequences in GenBank with the nucleotide sequence demonstrates the bfDy gene has only one exon. Alignment of the bfDy sequence with deiodinase sequences from representative species (human, mouse, chicken, frog, and fish) indicated 27–33% amino acid identity with vertebrate D1 sequences, 28 –32% identity with vertebrate D2 sequences, and 28 –33% identity with vertebrate D3 sequences. Similar amino acid identities were observed of bfDx and bfDt with the D1, D2, and D3 sequences. The phylogenetic relationship between the various amino acid sequences is illustrated in Fig. 1B. These results indicate that bfDy, bfDx, and bfDt cannot be categorized as a type 1, 2, or 3 deiodinase but form a separate branch on the phylogenetic tree. Although the average amino acid sequence identity of bfDy with vertebrate D1, D2, or D3 sequences amounts to roughly 30%, some parts of the amino acid sequences show a much higher conservation. This is particularly the case for the “core” sequence comprising amino acid residues 119 –167 of bfDy (Fig. 1C). This core bfDy sequence shows 48 –55% amino acid identity with vertebrate D1 sequences, 61– 65% identity with vertebrate D2 sequences, and 59 – 63% identity with vertebrate D3 sequences. Despite the high amino acid identity, there are also some notable differences, the most obvious being the presence of a Cys residue in bfDy at the position where Sec is present in bfDt, bfDx, and the vertebrate deiodinases. The latter is encoded by the opal UGA stop codon, whereas at the corresponding position, UGC codes for Cys in bfDy. In the 3⬘-untranslated region of all vertebrate deiodinase mRNA sequences, Sec insertion sequence elements have been identified that are required for read-


through at the UGA codon and incorporation of Sec (3). No such Sec insertion sequence element could be identified in the 3⬘-untranslated region of the bfDy mRNA sequence. Interestingly, UGA appears to function as a true stop codon at the end of the coding sequence in the bfDy mRNA (Fig. 1A). The region around the Sec/Cys residue is particularly well conserved in all known deiodinases as well as in bfDy (RPLVVCFGSYTCPPF). The second position upstream of this Cys/Sec residue is occupied by Cys in all D1 and D3 sequences and by Ala in all D2 sequences, whereas a Tyr is present in bfDy. The second amino acid downstream of the Sec/Cys residue is Pro in bfDy and either a Pro or Ser in other deiodinases. The nature of this residue has been associated with PTU sensitivity and with particular enzyme kinetics (17–19); all PTU-sensitive mammalian D1 proteins have a Ser at this position and all PTU-insensitive deiodinases (nonmammalian D1 and all known D2 and D3) feature a Pro residue. A second highly conserved sequence in all deiodinases is present at amino acids 152–167 in bfDy (NFLLVYIEEAHPSDGW). This sequence has been shown to be important for homodimerization of deiodinase proteins (20). It contains one of the two His residues that are highly conserved in all deiodinases. The other conserved His residue is also present in bfDy at position 178. Alignment with the vertebrate deiodinase sequences indicates that bfDy lacks the instability region present in hsD2 (amino acids 92–109), which is important for the ubiquitination of this enzyme (21). Examination of the bfDy amino acid sequence by different prediction programs identifies a putative transmembrane domain between residues 13 and 35. However, the exact location of this transmembrane domain and the orientation of the protein predicted by these programs vary. A single transmembrane domain has also been recognized at the N terminus of all known deiodinases. Enzyme characteristics of bfDy Deiodinase activity of lysates of COS1 cells transiently transfected with bfDy.pcDNA3.2 was undetectable using 1 nM [125I]T4 or [125I]T3 as substrate and 10 mM DTT as the cofactor (Fig. 2). Similar negative findings were obtained if 125I-labeled rT3, 3-iodotyrosine or 3,5-diiodotyrosine were tested as substrate or if reduced glutathione or reduced nicotinamide adenine dinucleotide phosphate were tested as possible cofactor (data not shown). When the report of Paris et al. (13) was published, indicating that not T3 but TA3 may be the active TH in amphioxus, we decided to test [125I]TA4 and [125I]TA3 as possible substrates for bfDy. Under the same conditions where bfDy failed to deiodinate T4 and T3, we observed rapid deiodi-



3263

FIG. 2. HPLC analysis of the deiodination of [125I]TA3 (A), [125I]TA4 (B), and [125I]T3 (C) by lysates of bfDy-expressing COS1 cells and summary of results with all substrates (D). Reaction conditions: 1 nM substrate, 0.5 mg lysate protein/ml, 10 mM DTT; incubation for 60 min at 25 C. Reaction mixtures were processed and analyzed by HPLC as described in Materials and Methods.

nation of TA4 and TA3 (Fig. 2). The major product generated from [125I]TA4 eluted at the position of [125I]rTA3 from the HPLC column, and the major product from [125I]TA3 coeluted with synthetic 3,3⬘-TA2. These results indicated that TA4 and TA3 only underwent IRD and that further deiodination of rTA3 and 3,3⬘-TA2 was negligible. Because TA3 was deiodinated faster than TA4, we set out to characterize the deiodinase activity of bfDy using TA3 as the substrate. In all experiments, lysates of COS1 cells transfected with empty vector were used as controls, which were found to be completely devoid of deiodinase activity. Because B. floridae is a poikilotherm living in subtropical coastal waters, we realized that the optimal reaction temperature for bfDy may be well less than 37 C, which is routinely used for enzyme analysis. Therefore, we incubated COS1 cell lysates for 60 min with 1 nM [125I]TA3 and 10 mM DTT at different temperatures, i.e. 10, 20, 25, 30, and 37 C. The results indicated an optimum at 25 C with a strong decrease in activity if the temperature was increased to 37 C (Fig. 3A). Therefore, all subsequent experiments were conducted at 25 C. Next, the effect of DTT concentration was investigated. This involved incubation of bfDy-expressing cell lysates for 60 min at 25 C with labeled TA3 in the absence or presence of 0.1, 1, 10, or 100 mM DTT. The results indicated that deiodinase activity strictly required the addition of DTT with an optimum concentration of 10 mM (Fig. 3B). Hence, all further studies were carried out with 10 mM DTT as the cofactor.

FIG. 3. Determination of optimal conditions for analysis of bfDy activity. Unless indicated otherwise, reaction mixtures contained 1 nM [125I]TA3, 10 mM DTT, and 25–50 ␮g/ml bfDy-expressing COS1 cell protein and were incubated for 60 min at 25 C. A, Effects of incubation temperature, with an optimum at 25 C. B, Effects of DTT concentration, with an optimum at 10 mM. C, Effects of incubation time and lysate protein, with fairly linear responses up to 60 min and 30 ␮g/ml, respectively. D, Loss of deiodinase activity by incubation of bfDy-expressing COS1 cell lysates for 2– 60 min at 25 or 37 C. Deiodinase activity was measured by subsequent addition of 1 nM [125I]TA3 and incubation for 60 min at 25 C.

Finally, we determined the proper conditions for analysis of bfDy activity at 25 C and 10 mM DTT with respect to incubation time and lysate concentration. The results of these experiments indicated that the conversion of TA3 to TA2 was fairly linear with time up to 60 min of incubation and with lysate protein up to 30 ␮g/ml (Fig. 3C). Therefore, all further studies were done using 25 ␮g lysate protein/ml and an incubation time of 60 min. The substrate specificity of bfDy was studied by analysis of the effects of increasing concentrations (1 nM to 1 ␮M) of unlabeled TA3, TA4, T3, and T4 on the deiodination of 1 nM [125I]TA3. In all cases, a concentration-dependent decrease in [125I]TA3 deiodination was observed with IC50 values of 9 nM for TA3, 15 nM for TA4, 60 nM for T4, and 90 nM for T3 (Fig. 4A). The kinetics of TA3 and TA4 deiodination by bfDy were studied by Michaelis-Menten analysis of the deiodination rate as a function of increasing substrate concentrations (Fig. 4, B and D). This provided apparent Michaelis-Menten constant (Km) values of 6.8 and 68 nM and maximum velocity (Vmax) values of 1.4 and 5.4 pmol/min䡠mg lysate protein for TA3 and TA4, respectively.

3264

Klootwijk et al.


FIG. 4. Characteristics of bfDy deiodinase activity. Unless indicated otherwise, reaction mixtures contained 1 nM [125I]TA3, 10 mM DTT, and 25–50 ␮g/ml bfDy-expressing COS1 cell protein and were incubated for 60 min at 25 C. A, Effects of increasing concentrations (1 nM to 1 ␮M) of unlabeled TA3, TA4, T3, and T4 on deiodination of [125I]TA3. B, Michaelis-Menten analysis of the rate of deiodination of TA3 by bfDy at increasing TA3 concentrations. Apparent Km and Vmax values are indicated. C, Effects of increasing concentrations (1 ␮M to 1 mM) of the inhibitors PTU, IAc, GTG, and IOP on bfDy deiodinase activity. D, Michaelis-Menten analysis of the rate of deiodination of TA4 by bfDy at increasing TA4 concentrations. Apparent Km and Vmax values are indicated.

A number of Sec-targeted deiodinase inhibitors have been identified. Among them, PTU selectively inhibits mammalian D1, probably by reaction with an enzymeselenenyl iodide intermediate generated during catalysis, resulting in a stable, inactive enzyme-PTU adduct. IAc and GTG are two other deiodinase inhibitors that react avidly with the native selenolate (E-Se⫺) moiety of all selenodeiodinases (3, 22). We have tested the effects of increasing concentrations (1 ␮M to 1 mM) of these inhibitors on the deiodinase activity of bfDy. The results are presented in Fig. 4C. No inhibition whatsoever was observed with PTU and IAc. A dose-dependent inhibition was found for GTG, with an IC50 value of approximately 100 ␮M. This is two to three orders of magnitude higher than the GTG concentration required to inhibit selenodeiodinases. We also tested the possible effects of IOP, an x-ray contrast agent that potently and competitively inhibits D1, D2, and D3 (3). We observed dose-dependent inhibition of bfDy by IOP with an IC50 value of approximately 100 ␮M, much higher than the IOP concentrations inhibiting D1, D2, or D3. The temperature sensitivity of bfDy was further investigated by incubating bfDy-expressing cell lysates for dif-


ferent time periods at 25 or 37 C in the presence of 10 mM DTT but in the absence of substrate. Subsequently, bfDy activity was measured by addition of 1 nM [125I]TA3 and continued incubation for 60 min at 25 C. The results show that bfDy was rapidly inactivated at 37 C with a half-life of less than 2 min, whereas more than 60% of enzyme activity remained after a 60-min incubation at 25 C (Fig. 3D). The rapid decrease in bfDy activity at 37 C was not prevented in the presence of multiple protease inhibitors (data not shown). The failure of bfDy to deiodinate T3 and T4 cannot be explained by their lack of binding to the enzyme, as indicated by their potent inhibition of TA3 deiodination. Apparently, T3 and T4 bind with relatively high affinity to bfDy, but this does not result in their deiodination. The possibility was considered that the presence of Cys instead of Sec allowed deiodination of TA3 and TA4 by bfDy but prevented the deiodination of T3 and T4. This was tested by analysis of the deiodination of TA3 and T3 by wild-type hsD3 and its Sec⬎Cys mutant. Lysates of cells expressing these proteins showed similar deiodinase activities toward both T3 and TA3. However, although cell lysates containing hsD3Cys were as effective as lysates with wild-type hsD3 in catalyzing deiodination of TA3 and T3, the products 3,3⬘-TA2 and 3,3⬘-T2 were rapidly further deiodinated in the inner ring by wild-type hsD3 but not by hsD3Cys (Fig. 5).

FIG. 5. Conversion of T3 to 3,3⬘-T2 and 3⬘-T1 and of TA3 to 3,3⬘-TA2 and 3⬘-TA1 by COS1 cell lysates expressing wild-type hsD3 (A) or hsD3Cys mutant (B). Incubations mixtures contained 1 nM [125I]T3 or [125I]TA3, 0.004 – 0.4 mg lysate protein/ml, and 10 mM DTT and were incubated for 60 min at 37 C.


Discussion This is the first report of a naturally occurring nonselenodeiodinase catalyzing the deiodination of iodothyronine derivatives. The amino acid sequence of this amphioxus enzyme shows similar amino acid identities with vertebrate D1, D2, or D3 sequences. However, because it only catalyzes IRD, its activity is more akin to D3. It is clear that bfDy contains a Cys residue, where all known deiodinases feature a Sec residue. Another remarkable property of bfDy is that it catalyzes specifically the deiodination of TA3 and TA4, whereas it shows no activity toward T3 and T4. It has a high affinity for its substrates and is characterized by Km values of 6.8 nM for TA3 and 68 nM for TA4. The presence of Cys instead of Sec in the active center in bfDy explains its low susceptibility to inhibition by IAc and GTG, while it is also not affected by PTU. Like most deiodinases, it requires a thiol cofactor and shows optimal activity in the presence of 10 mM DTT. bfDy is remarkably heat labile; it is rapidly denaturated at 37 C but is quite stable at 25 C. In the selenodeiodinases, the Sec residue appears to participate directly in the catalytic process (3). Replacement of the Sec residue by inert amino acids, such as Ala or Leu, results in the complete inactivation of the enzymes (5–7). Substitution of Sec by Cys results in a large drop in deiodinase activity, although this is offset by a higher efficiency of the synthesis of the Cys mutant than of the wild-type Sec protein. Kinetic analysis of the Cys mutants vs. wild-type deiodinases shows primarily a marked reduction in Vmax for D1 (5), a large increase in apparent Km value for D2 (6), and changes in both Km and Vmax for D3, affecting T4 deiodination more than T3 (7). The Km value for TA3 deiodination by bfDy (6.8 nM) is similar to the Km value of T3 for wild-type hsD3 (6.6 nM) but substantially lower than the Km value of T3 for hsD3Cys (33 nM) (7). No Km values have been reported for TA3 deiodination by hsD3 and hsD3Cys. Although T3 and T4 are not deiodinated by bfDy, they potently inhibit the deiodination of TA3 by this enzyme with IC50 values less than 100 nM. This suggests that T3 and T4 bind to the active center of bfDy without being deiodinated. It could be argued that the catalytic Cys residue in bfDy is capable of deiodinating the inner ring of TA3 and TA4 but not that of T3 and T4. This could be due to the much lower reactivity of Cys than Sec in combination with differences in electron density of the inner ring of T3 and T4 vs. TA3 and TA4. However, replacement of Sec by Cys in hsD3 does not specifically diminish deiodination of T3 and T4 compared with TA3 and TA4. We noticed, however, that wild-type hsD3 actively further deiodinates the TA3 product 3,3⬘-TA2 to 3⬘-TA1 (and the T3 product


3265

3,3⬘-T2 to 3⬘-T1), whereas hsD3Cys and bfDy are much less active in this respect. PTU is a selective and potent inhibitor of D1 (IC50, ⬃5 ␮M), probably by targeting the enzyme-selenenyl iodide intermediate of the catalytic process, in a reaction that requires the presence of a Ser residue two positions downstream of Sec (17–19). As D2 and D3 contain Pro instead of Ser at this position, they are much less sensitive to PTU inhibition (IC50, ⬎1 mM). D1 is also potently inhibited by IAc (IC50, ⬃2 ␮M) and GTG (IC50, ⬃50 nM) that target the native E-SeH form of the enzyme, whereas 50% inhibition of D2 and D3 require approximately 1 mM IAc and 1–2 ␮M GTG (22). The present study indicates that bfDy is even much less sensitive to all these inhibitors than D2 and D3, which probably reflects the presence of Cys instead of Sec and of Pro instead of Ser in its active center. A remarkable difference is also the presence of Tyr in bfDy two positions upstream of Sec, where D1 and D3 feature a Cys residue and D2 an Ala residue. Because Tyr is much larger than Cys and Ala, the substrate-binding site of bfDy may be smaller that of D1, D2, and D3. It is tempting to speculate that this is related to the inability of bfDy to deiodinate T3 and T4 in contrast to the rapid deiodination of TA3 and TA4. The optimum temperature for bfDy activity is approximately 25 C, close to the ambient temperature of this poikilotherm’s habitat. Above 30 C, the deiodinase activity of bfDy shows a steep decrease, resulting from the denaturation of the protein. A similar temperature optimum was found previously for a selenodeiodinase (hrDx) expressed in the sea squirt Halocynthia roretzi that catalyzes the ORD of T4 and rT3 (23). This low temperature optimum of bfDy is therefore not related to the absence of the Sec residue but is probably associated with a high flexibility of the protein structure that also makes it susceptible to denaturation at higher temperatures. It is unknown if this denaturation only occurs in cell lysates or if it also occurs in situ in transfected cells. In the latter case, the yield of active bfDy may be further increased by culturing the cells at temperatures less than 37 C. The identification of a nonselenodeiodinase in amphioxus that specifically catalyzes the IRD of TA3 and TA4 is interesting in view of recent findings regarding the role of TH in the metamorphosis of this organism (13). In vivo, this metamorphosis is accelerated by administration of T4, T3, TA4, or TA3, and it is inhibited by administration of thyroid-blocking agents. The effects of T4 and TA4 but not of T3 and TA3 are inhibited by administration of IOP (13). Amphioxus does not have a thyroid gland but an endostyle that contains dispersed thyroid follicle-like structures (12). The amphioxus genome contains genes that are highly homologous to vertebrate genes involved in TH

3266

Klootwijk et al.


synthesis, such as the Na/I symporter and thyroid peroxidase, although a thyroglobulin-like gene is not readily identified in amphioxus (12). Amphioxus also expresses a gene that is the homolog of the vertebrate TR. Interestingly, this TR is activated by TA3 but not by T3 (13). It appears, therefore, that the TH-dependent stimulation of amphioxus metamorphosis is not mediated by T3 but by TA3 binding to TR, and this may well hold true for other TH actions in this organism. If TA3 is the major active TH in amphioxus, multiple enzymes may be involved in the regulation of its intracellular concentration and, thus, its biological activity in a tissue-specific manner. That T4 stimulates amphioxus metamorphosis in vivo, although only TA3 is capable of stimulating the TR in vitro, indicates the presence of enzymes catalyzing ORD and side chain modification. Recent in vivo studies support this notion by showing that TA3 and TA4 are major TH metabolites in amphioxus, but serum levels have not been determined (24). Conversion of the Ala side chain of T4 and T3 to the acetic acid side chain of TA4 and TA3 is probably a two-step process. In the first step, T4 or T3 would be converted to 3,3⬘,5,5⬘-tetraiodothyronamine or 3,3⬘,5-triiodothyronamine by an aromatic amino acid decarboxylase or similar enzyme. In the second step, the ethylamine intermediates would undergo oxidative deamination to TA4 and TA3 by monoamine oxidase or similar enzyme (25). Generation of TA3 from T4 may take place in the following pathways: T4 3 T3 3 TA3 or T4 3 TA4 3 TA3. Mammalian deiodinases are capable of deiodinating both iodothyronine and iodothyroacetic acid derivatives (26, 27) (this article and our unpublished observations), but this may not be the case in amphioxus. In fact, we found that bfDy effectively deiodinates the inner ring of TA4 and TA3 but not T4 and T3. Supposedly, this enzyme is important for the regulation of TH bioactivity by catalyzing the degradation of TA3 and TA4. If there is also an enzyme in amphioxus that specifically catalyzes the deiodination of T4 and T3 but not of TA4 and TA3 remains to be investigated. Molecular characterization of all the different deiodinases in amphioxus is required to fully understand TH regulation in this species. The present findings in combination with recent reports by Paris et al. (11–13, 24) raise the question if TA3 is an ancient bioactive hormone and that T3 has taken over this function later in evolution. More research in invertebrates is required to answer this question. We have characterized a selenodeiodinase in the sea squirt H. roretzi, which catalyzes the ORD of T4 and T3 but is less active in the deiodination of TA4 (23). Other ascidians, Ciona intestinalis and Ciona savignyi, have two selenodeiodinase genes, but the encoded enzyme activities have not been studied. In-


terestingly, the genome of the sea anemone Nematostella vectensis contains four genes apparently coding for nonselenodeiodinase, and we are currently investigating their enzyme activities. It is remarkable that in humans and other mammals, TA3 is still an active hormone showing high affinity for different TH-related proteins, including the T3 receptors TR␣1 and TR␤1/2 (28), the deiodinases D1, D2, and D3 (26, 27), and the plasma binding protein transthyretin (29). It has been shown before that Sec is not an essential amino acid in mammalian deiodinases. Although the selenodeiodinases show much greater catalytic efficiency than their Cys mutants, this is offset by a much lower efficiency for the synthesis of selenoproteins at least in transfected cells (5–7). The present study indicates that a natural nonselenodeiodinase very effectively catalyzes the deiodination of its substrates. It is possible that in vitro analysis of deiodinases, in the presence of high concentrations of an artificial cofactor, leads to overestimation of the activity of Cys vs. Sec deiodinases. It is also possible that Cys deiodinases are not capable of catalyzing all possible deiodinations. A glimpse of that may be seen in the poor IRD of 3,3⬘-T2 and 3,3⬘-TA2 by hsD3Cys vs. hsD3Sec. This requires further investigation.

Acknowledgments Address all correspondence and requests for reprints to: Theo J. Visser, Ph.D., Department of Internal Medicine, Erasmus University Medical Center, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands. E-mail: [email protected]. This work was supported by was supported by The Netherlands Organization for Scientific Research Grant 9120.6093 (to E.C.H.F.). Disclosure Summary: The authors have nothing to disclose.

References 1. Oetting A, Yen PM 2007 New insights into thyroid hormone action. Best Pract Res Clin Endocrinol Metab 21:193–208 2. Yen PM 2001 Physiological and molecular basis of thyroid hormone action. Physiol Rev 81:1097–1142 3. Gereben B, Zavacki AM, Ribich S, Kim BW, Huang SA, Simonides WS, Zeo¨ld A, Bianco AC 2008 Cellular and molecular basis of deiodinase-regulated thyroid hormone signaling. Endocr Rev 29:898 – 938 4. Valverde C, Orozco A, Becerra A, Jeziorski MC, Villalobos P, Solís JC 2004 Halometabolites and cellular dehalogenase systems: an evolutionary perspective. Int Rev Cytol 234:143–199 5. Berry MJ, Maia AL, Kieffer JD, Harney JW, Larsen PR 1992 Substitution of cysteine for selenocysteine in type I iodothyronine deiodinase reduces the catalytic efficiency of the protein but enhances its translation. Endocrinology 131:1848 –1852 6. Buettner C, Harney JW, Larsen PR 2000 The role of selenocysteine


7.

8.

9. 10.

11. 12.

13.

14.

15.

16.

133 in catalysis by the human type 2 iodothyronine deiodinase. Endocrinology 141:4606 – 4612 Kuiper GG, Klootwijk W, Visser TJ 2003 Substitution of cysteine for selenocysteine in the catalytic center of type III iodothyronine deiodinase reduces catalytic efficiency and alters substrate preference. Endocrinology 144:2505–2513 Salvatore D, Harney JW, Larsen PR 1999 Mutation of the Secys residue 266 in human type 2 selenodeiodinase alters 75Se incorporation without affecting its biochemical properties. Biochimie 81: 535–538 Wu SY, Green WL, Huang WS, Hays MT, Chopra IJ 2005 Alternate pathways of thyroid hormone metabolism. Thyroid 15:943–958 Scanlan TS, Suchland KL, Hart ME, Chiellini G, Huang Y, Kruzich PJ, Frascarelli S, Crossley DA, Bunzow JR, Ronca-Testoni S, Lin ET, Hatton D, Zucchi R, Grandy DK 2004 3-Iodothyronamine is an endogenous and rapid-acting derivative of thyroid hormone. Nat Med 10:638 – 642 Paris M, Laudet V 2008 The history of a developmental stage: metamorphosis in chordates. Genesis 46:657– 672 Paris M, Brunet F, Markov GV, Schubert M, Laudet V 2008 The amphioxus genome enlightens the evolution of the thyroid hormone signaling pathway. Dev Genes Evol 218:667– 680 Paris M, Escriva H, Schubert M, Brunet F, Brtko J, Ciesielski F, Roecklin D, Vivat-Hannah V, Jamin EL, Cravedi JP, Scanlan TS, Renaud JP, Holland ND, Laudet V 2008 Amphioxus postembryonic development reveals the homology of chordate metamorphosis. Curr Biol 18:825– 830 Holland LZ, Albalat R, Azumi K, Benito-Gutie´rrez E, Blow MJ, Bronner-Fraser M, Brunet F, Butts T, Candiani S, Dishaw LJ, Ferrier DE, Garcia-Ferna`ndez J, Gibson-Brown JJ, Gissi C, Godzik A, Hallbo¨o¨k F, Hirose D, Hosomichi K, Ikuta T, Inoko H, Kasahara M, Kasamatsu J, Kawashima T, Kimura A, Kobayashi M, Kozmik Z, Kubokawa K, Laudet V, Litman GW, McHardy AC, Meulemans D, Nonaka M, Olinski RP, Pancer Z, Pennacchio LA, Pestarino M, Rast JP, Rigoutsos I, Robinson-Rechavi M, Roch G, Saiga H, Sasakura Y, Satake M, Satou Y, Schubert M, Sherwood N, Shiina T, Takatori N, Tello J, Vopalensky P, Wada S, Xu A, Ye Y, Yoshida K, Yoshizaki F, Yu JK, Zhang Q, Zmasek CM, de Jong PJ, Osoegawa K, Putnam NH, Rokhsar DS, Satoh N, Holland PW 2008 The amphioxus genome illuminates vertebrate origins and cephalochordate biology. Genome Res 18:1100 –1111 Putnam NH, Butts T, Ferrier DE, Furlong RF, Hellsten U, Kawashima T, Robinson-Rechavi M, Shoguchi E, Terry A, Yu JK, Benito-Gutie´rrez EL, Dubchak I, Garcia-Ferna`ndez J, GibsonBrown JJ, Grigoriev IV, Horton AC, de Jong PJ, Jurka J, Kapitonov VV, Kohara Y, Kuroki Y, Lindquist E, Lucas S, Osoegawa K, Pennacchio LA, Salamov AA, Satou Y, Sauka-Spengler T, Schmutz J, Shin-I T, Toyoda A, Bronner-Fraser M, Fujiyama A, Holland LZ, Holland PW, Satoh N, Rokhsar DS 2008 The amphioxus genome and the evolution of the chordate karyotype. Nature 453:1064 – 1071 Tamura K, Dudley J, Nei M, Kumar S 2007 MEGA4: molecular


17.

18.

19.

20.

21.

22. 23.

24.

25.

26.

27.

28.

29.

3267

evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24:1596 –1599 Callebaut I, Curcio-Morelli C, Mornon JP, Gereben B, Buettner C, Huang S, Castro B, Fonseca TL, Harney JW, Larsen PR, Bianco AC 2003 The iodothyronine selenodeiodinases are thioredoxin-fold family proteins containing a glycoside hydrolase clan GH-A-like structure. J Biol Chem 278:36887–36896 Goemann IM, Gereben B, Harney JW, Zhu B, Maia AL, Larsen PR 2010 Substitution of serine for proline in the active center of type 2 iodothyronine deiodinase substantially alters its in vitro biochemical properties with dithiothreitol but not its function in intact cells. Endocrinology 151:821– 829 Kuiper GG, Klootwijk W, Morvan Dubois G, Destree O, Darras VM, Van der Geyten S, Demeneix B, Visser TJ 2006 Characterization of recombinant Xenopus laevis type I iodothyronine deiodinase: substitution of a proline residue in the catalytic center by serine (Pro132Ser) restores sensitivity to 6-propyl-2-thiouracil. Endocrinology 147:3519 –3529 Leonard JL, Simpson G, Leonard DM 2005 Characterization of the protein dimerization domain responsible for assembly of functional selenodeiodinases. J Biol Chem 280:11093–11100 Dentice M, Bandyopadhyay A, Gereben B, Callebaut I, Christoffolete MA, Kim BW, Nissim S, Mornon JP, Zavacki AM, Zeo¨ld A, Capelo LP, Curcio-Morelli C, Ribeiro R, Harney JW, Tabin CJ, Bianco AC 2005 The Hedgehog-inducible ubiquitin ligase subunit WSB-1 modulates thyroid hormone activation and PTHrP secretion in the developing growth plate. Nat Cell Biol 7:698 –705 Visser TJ 1996 Pathways of thyroid hormone metabolism. Acta Med Austriaca 23:10 –16 Shepherdley CA, Klootwijk W, Makabe KW, Visser TJ, Kuiper GG 2004 An ascidian homolog of vertebrate iodothyronine deiodinases. Endocrinology 145:1255–1268 Paris M, Hillenweck A, Bertrand S, Delous G, Escriva H, Zalko D, Cravedi JP, Laudet V 2010 Active metabolism of thyroid hormone during metamorphosis of amphioxus. Integr Comp Biol 50:63–74 Wood WJ, Geraci T, Nilsen A, DeBarber AE, Scanlan TS 2009 Iodothyronamines are oxidatively deaminated to iodothyroacetic acids in vivo. Chembiochem 10:361–365 Koehrle J, Auf’mkolk M, Rokos H, Hesch RD, Cody V 1986 Rat liver iodothyronine monodeiodinase. Evaluation of the iodothyronine ligand-binding site. J Biol Chem 261:11613–11622 Rutgers M, Heusdens FA, Visser TJ 1989 Metabolism of triiodothyroacetic acid (TA3) in rat liver. I. Deiodination of TA3 and TA3 sulfate by microsomes. Endocrinology 125:424 – 432 Jeyakumar M, Webb P, Baxter JD, Scanlan TS, Katzenellenbogen JA 2008 Quantification of ligand-regulated nuclear receptor corepressor and coactivator binding, key interactions determining ligand potency and efficacy for the thyroid hormone receptor. Biochemistry 47:7465–7476 Rosen HN, Murrell JR, Liepnieks JJ, Benson MD, Cody V, Moses AC 1994 Threonine for alanine substitution at position 109 of transthyretin differentially alters human transthyretin’s affinity for iodothyronines. Endocrinology 134:27–34