Until recently, 3,5-diiodothyronine (3,5-T2) has been considered an inactive by-product of triiodothyronine (T3) deiodination. However, studies from several laboratories have shown that 3,5-T2 has specific, nongenomic effects on mitochondrial oxidative capacity and respiration rate that are distinct from those due to T3. Nevertheless, little is known about the putative genomic effects of 3,5-T2. We have previously shown that hyperthyroidism induced by supraphysiological doses of 3,5-T2 inhibits hepatic iodothyronine deiodinase type 2 (D2) activity and lowers mRNA levels in the killifish in the same manner as T3 and T4, suggesting a pretranslational effect of 3,5-T2 (Garcia-G C, Jeziorski MC, Valverde-R C, Orozco A. Gen Comp Endocrinol 135: 201–209, 2004). The question remains as to whether 3,5-T2 would have effects under conditions similar to those that are physiological for T3. To this end, intact killifish were rendered hypothyroid by administering methimazole. Groups of hypothyroid animals simultaneously received 30 nM of either T3, reverse T3, or 3,5-T2. Under these conditions, we expected that, if it were bioactive, 3,5-T2 would mimic T3 and thus reverse the compensatory upregulation of D2 and tyroid receptor β1 and downregulation of growth hormone that characterize hypothyroidism. Our results demonstrate that 3,5-T2 is indeed bioactive, reversing both hepatic D2 and growth hormone responses during a hypothyroidal state. Furthermore, we observed that 3,5-T2 and T3 recruit two distinct populations of transcription factors to typical palindromic and DR4 thyroid hormone response elements. Taken together, these results add further evidence to support the notion that 3,5-T2 is a bioactive iodothyronine.
- deiodinase type 2
- thyroid hormone receptor β1
- thyroid hormone response element
iodothyronines or thyroid hormones (TH) are essential in regulating energy expenditure and development. Triiodothyronine (T3) is the bioactive TH, which modulates gene expression in virtually every vertebrate tissue through ligand-dependent transcription factors, the TH receptors (TR). Sequential deiodination of thyroxine (T4) generates T3 as well as other iodothyronines that have been considered inactive by-products, but, recently, interest has grown in identifying bioactive iodothyronines in addition to T4 and T3. Studies from several laboratories have suggested that 3,5-diiodothyronine (3,5-T2), a putative product of the deiodination pathway involved in T3 metabolism, could be a peripheral mediator of some effects of TH on mitochondrial oxidative capacity and respiration rate. To date, results in mammals suggest that 3,5-T2 has specific actions on oxygen consumption that are distinct from those of T3: they are not attenuated by inhibition of protein synthesis and are more rapid than those due to T3 (for review, see Ref. 12). Genomic effects of 3,5-T2 have been analyzed in only a few classic iodothyronine-dependent genes, such as thyroid stimulating hormone (TSH), thyroid receptor β2 (TRβ2), iodothyronine deiodinase type 1 (D1), and growth hormone (GH) (2, 3, 24). While 3,5-T2 inhibits TSH and TRβ2 expression in vivo and in vitro, respectively (2, 24), it stimulates expression of D1 and GH under both experimental conditions (2, 3, 24). Interestingly, the potency of 3,5-T2 with respect to these target genes seems to depend on the status of the deiodination pathways, as well as on the experimental model. Thus 3,5-T2 is equipotent to T3 in hypothyroid animals in which the deiodinating enzymes are inhibited. On the other hand, in euthyroid animals or in cell cultures, 3,5-T2 is less potent than T3 (see Ref. 12 for review).
One prevailing question is how 3,5-T2 is formed in the target cell. Although conversion of T3 to 3,5-T2 has not been demonstrated in mammals (21) or in teleosts in vitro (A. Orozco, unpublished observations; Ref. 31), indirect but robust evidence indicates that 3,5-T2 is indeed formed from T3 in vivo through deiodination (23). It is not yet known which deiodinase catalyzes the conversion of T3 to 3,5-T2, either in vivo or in vitro, but a likely candidate is iodothyronine deiodinase type 2 (D2). D2 catalyzes the removal of outer-ring iodine from T4 to generate the bioactive hormone T3 and thus determines T3 availability at the cellular level. Conversion of T3 to 3,5-T2 would require similar outer-ring deiodination.
Our laboratory has previously studied whether 3,5-T2 has a regulatory effect upon D2 in Fundulus heteroclitus (killifish). In this species, hepatic D2 expression is very abundant, among the highest in vertebrates so far examined (26, 28). The D2 gene is tightly downregulated by bioactive TH (for review: Refs. 4, 5). We have found that hyperthyroidism induced by a short-term immersion (up to 24 h) of killifish in seawater (SW) supplemented with 3,5-T2 significantly decreased both mRNA expression and hepatic activity of D2, mimicking the effects of T4 or T3 (10). Because these effects were obtained with supraphysiological doses of the three iodothyronines, in this work, we asked whether 3,5-T2, when administered in doses similar to those known to be physiological for T3, could reestablish the euthyroidal activity and hepatic expression of D2 in hypothyroid killifish. We also examined the effects of 3,5-T2 upon the hepatic expression of GH, which is upregulated in the pituitary by hyperthyroid status (1, 11, 32). Finally, as a first approach to study the action mechanism of 3,5-T2 in the killifish, we analyzed the hepatic expression of the TRβ1 and the TH response element (TRE)-transcription factor complexes formed in response to treatment with 3,5-T2 or T3.
MATERIALS AND METHODS
SW-adapted male Fundulus heteroclitus, ranging from 4 to 6 g, were collected from the estuarine creeks of the Matanzas River (St. Augustine, FL). After capture, fish were kept in tanks with running SW piped directly from the ocean at a temperature of ∼28°C. Animals were deparasitized after capture, fed ad libitum 24 h later (Silver Cup, Nelson and Sons), and maintained on a light-dark cycle of 14:10 h. Collection of the tissues for the different experiments was performed as follows: fish were decapitated, and the liver was rapidly removed. Each liver was divided in half to determine both deiodinase activity and mRNA levels. Segments used to measure deiodinase activity were homogenized in 10 volumes of ice-cold homogenization buffer [10 mM HEPES (Sigma), 0.25 M sucrose (Sigma), 10 mM EDTA (Sigma), pH 7]. Aliquots were quick-frozen in liquid nitrogen and stored at −70°C until assayed. To quantify mRNA, segments were pooled (n = 2–3/pool) for RNA extraction. All animal experimentation was conducted in accord with accepted standards of humane animal care, and procedures regarding handling and euthanasia of animals were reviewed and approved by the Animal Welfare Committee of our Institute.
Our laboratory (10) and others (9, 25) have established that the administration of hydrophobic drugs by immersion provides an efficient, noninvasive, and minimally stressful means of chronically administering these compounds in aquatic vertebrates. We used this method to deliver all treatments in the present work. The dose-response curves for T3 and 3,5-T2 were performed by supplementing the environmental SW with 10, 30, 100, and 300 nM of either iodothyronine (Sigma) for 24 h (vehicle: 0.01 N NaOH) or vehicle only. For time-response curves, a final concentration of 100 nM of either T3 or 3,5-T2 was administered by immersion for 3, 6, 12, 18, and 24 h (n = 6). Fish were killed after the treatment, the liver was dissected, and D2 activity was determined.
Since the aim of the present study was to analyze the effects of 3,5-T2 and to compare them to those exerted by physiological doses of T3, we treated the different experimental groups with methimazole (MMI; Sigma), an inhibitor of TH synthesis, at a final concentration of 4.5 mM, with or without simultaneous addition of T3, reverse T3 (rT3), or 3,5-T2 (Sigma) at a final concentration of 30 nM (vehicle: 0.01 N NaOH). Based on the dose-response curve results from euthyroid animals (see above), we simultaneously added different doses (10, 30, and 100 nM) of either T3 or 3,5-T2 to the MMI-treated fish. We found that 30 nM of exogenous T3 merely restored intrahepatic concentrations of this hormone, whereas higher doses elevated intrahepatic T3 above control levels (data not shown). Based on these results, we defined 30 nM as the effective immersion dose at which intrahepatic euthyroidism is attained. For each experiment, fish (n = 18/experimental group) were placed into tanks with running SW and constant aeration. After 12 h of acclimation, the water flow was shut off, and the water volume of each tank was adjusted to 6 liters. Control groups were handled in the same manner as the experimental groups. Fish were killed after 5 days of treatment. Six livers from control, MMI, MMI + T3, and MMI + 3,5-T2 groups were individually prepared for iodothyronine extraction and measurement of T3 (RIA). All of the experimental treatments were performed at least twice in independent groups.
RIA of iodothyronines.
To determine intrahepatic T3 levels, individual livers were homogenized, and iodothyronines were extracted as previously described (10). Livers (average wet weight 100 ± 7 mg) were homogenized in 10 volumes of methanol-ammonium hydroxide (99:1) solution. The homogenates were centrifuged for 10 min at 700 g. The supernatants were collected, evaporated in a speed vacuum, and suspended 1:5 (initial weight/volume) in assay buffer Tris·HCl (0.05 M; pH 8.6). Hepatic content of T3 was measured by RIA as previously described (27). The inter- and intra-assay coefficients of variation were 9.5 and 6.6%, respectively. The incubation mixture contained assay buffer and a working dilution (1:8,000) of anti-T3 serum (Sigma), standard (standard curve, 7.8–1,000 pg/dl), the radioactive solution (10 pg/100 μl of the labeled T3 plus 10 mg/10 ml of 8-anilino-1-naphthalene sulfonic acid, Sigma), and 50 μl of the experimental sample. Free and antibody-bound radioactive T3 were separated using 0.5% activated charcoal/dextran suspension (Sigma).
Enzyme activities were measured in duplicate as previously described (26). The total volume of the reaction mixture was 100 μl and contained 1 nM 125I-T4 (sp. act. 1,200 μCi/μg; New England Nuclear) plus 25 mM DTT (Calbiochem) and liver homogenate at a protein concentration of ∼100 μg. Assays were incubated for 1 h at 37°C. The released, acid-soluble 125I was isolated by chromatography on Dowex 50W-X2 columns (BioRad). Specific activity was calculated as previously described (29) and expressed as femtomoles of 125iodide per hour per milligram. Protein content was measured by the Bradford method (BioRad).
Measurement of D2 mRNA.
RNA was isolated from treated or control livers using TRIzol (Invitrogen), and cDNA was reverse transcribed (Superscript, Invitrogen) from 10 μg RNA using a gene-specific primer (5′ TTC AGA GCT CAT CTA CTA TCG T 3′). The concentration of D2 mRNA was measured in duplicate by competitive PCR, as previously described (10). The standard curve ranged from 104 to 109 molecules/μl. Oligonucleotides used (sense: 5′ CAA ACA GGT GAA ACT TGG CT 3′ and antisense: 5′ TCG TCG ATG TAG ACC AGC 3′) amplified a product of 270 bp (40 s at 94°, 40 s at 65°, 30 s at 72° for 35 cycles). Identical PCRs from the RNA samples before the reverse transcription reaction yielded no detectable products, indicating that the RNA was not contaminated with genomic DNA. Results are expressed as molecules per microgram of total mRNA used in the reverse transcription.
Measurement of GH mRNA.
Total RNA was extracted from male killifish brains (including the pituitary) with TRIzol and used as a template for reverse transcription of cDNA using an oligo-dT-polylinker primer. A degenerate primer (5′ TGY TTY AAR AAR GAY ATG CAY AAR 3′) that contained a sequence conserved in 35 teleost GH cDNAs reported in GeneBank (CFKKDMHKV) was used with an oligo-dT-polylinker primer to generate a first fragment, which included the stop codon of the killifish GH cDNA. The 5′ end of the cDNA was amplified using rapid amplification of cDNA ends. A cDNA was reversed transcribed using a specific antisense primer. This cDNA was purified and tailed with dCTP using terminal transferase, and an oligo-dG-polylinker primer was used in combination with exact primers in a series of nested amplifications to generate the 5′ end. The open reading frame of the GH cDNA clone obtained had 64% identity to the other 35 teleost GH genes reported in GeneBank. This sequence was inserted into a vector (pGEM-T, Promega) and used to construct a standard curve. Primers specific for this sequence were designed for real-time PCR (see below).
Reverse transcription (Superscript) from 10 μg of total hepatic RNA was performed using an oligo(dT) primer (final volume of 20 μl). Real-time PCR was carried out in duplicate using β-actin as internal standard in reactions that contained 1 μl aliquots of the reverse transcription reaction described above, corresponding to 0.5 μg total RNA, 7.5 μl of TAQurate GREEN Real-Time PCR MasterMix (EPICENTRE), and 500 nM forward and reverse primers in a final volume of 15 μl. A 233-bp fragment from killifish β-actin (accession number: CN985078) was cloned with oligonucleotides 5′ GCG ACA TCA AGG AGA AGC T 3′ (sense) and 5′ CGA CGT CGC ACT TCA TGAT 3′ (antisense) and used to construct a standard curve (103–108 molecules/μl). This same combination of primers was used to measure β-actin in the experimental samples (3 s at 94°C, 8 s at 61°C, 9 s at 72°C for 45 cycles). For GH cDNA, the standard curve ranged from 101 to 106 molecules/μl. The oligonucleotides used were 5′ GAT CTC CCC CAA ACT GTC A 3′ (sense) and 5′ GAC TCA TCA GCT TCC AGA CT 3′ (antisense). These primers generated a product of 146 bp (3 s at 94°C, 5 s at 55°C, 7 s at 72°C for 45 cycles). Detection and data analyses were carried out on a Light Cycler instrument, according to the manufacturer's instructions (Roche Molecular Biochemicals, Mannheim, Germany). Fluorescence analysis was performed after each cycle, and the PCR products generated were analyzed with a melting curve to verify the specificity of the amplified products. The mRNA concentration (molecules/μl), obtained by comparison with the standard curve, was normalized to the concentration of β-actin in each experimental sample. As indicated for D2 expression, identical PCRs from the RNA samples before the reverse transcription reaction yielded no detectable products. Results are expressed as molecules/μg of total mRNA used in the reverse transcription.
Measurement of Na-K-ATPase α1-mRNA.
Reverse transcription and real-time PCRs were conducted as described above. Based on the reported sequence for the killifish α1-subunit of the Na-K-ATPase cDNA (accession number AY057072; Ref. 30), a fragment containing the open-reading frame was amplified (sense: 5′ GGG AAG TTT TGA AAA AGA AAA TTG 3′ and antisense: 5′ GCC TTG CAA CAC GAT GGT G 3′ primers), subcloned (pGEM-T, Promega), and used to construct a standard curve (103–108 molecules/μl). A product of 256 bp was amplified with sense (5′ GGA ACT GCC AGA GGA ATT G 3′) and antisense (5′ GGA GAC CTT CTG GCA CAT TA 3′) primers (3 s at 94°C, 6 s at 64°C, 10 s at 72°C for 45 cycles). Identical PCRs from the RNA samples before the reverse transcription reaction yielded no detectable products.
Measurement of TRβ1 mRNA.
Total RNA was extracted from male killifish livers (TRIzol, Invitrogen) and used as a template for reverse transcription of cDNA using an oligo-dT-polylinker primer. A pair of degenerate primers (sense: 5′ TGA GTG CAG IGG GGG TGA AG 3′ and antisense: 5′ GAC ATG ATC TCC ATR CAR CA 3′) that contained highly conserved TRβ1 cDNA sequences previously reported in several teleosts (22) was used to clone a 230-bp fragment that had 84% identity to the other teleost TRβ1 genes reported. This fragment was inserted into a vector (pGEM-T, Promega) and used to construct a standard curve (103–108 molecules/μl). Primers specific for this sequence were designed for real-time PCR.
Reverse transcription and real-time PCRs were conducted as described above. A product of 180 bp was amplified with sense (5′ TGA GTG CAG GGG GGG GTG AAG 3′) and antisense (5′ GCA GCT CAC AGA ACA TGG GC 3′) primers (3 s at 94°C, 5 s at 65°C, 8 s at 72°C for 45 cycles). Identical PCRs from the RNA samples before the reverse transcription reaction yielded no detectable products.
Nuclear proteins were obtained as previously described (20). In brief, untreated fish (control) or fish treated with MMI only or MMI plus either T3 or 3,5-T2 (30 nM) for 5 days were killed, and the livers were resuspended 1:10 (wt/vol) in a hypotonic buffer (10 mM HEPES, pH 7.9, 10 mM KCl, 1 mM EDTA, 5 mM DTT). After mechanical disruption, the tissues were incubated at 4°C for 30 min, and the supernatants were collected. The quality of the nuclei was evaluated with a Trypan blue stain (1:1). Nuclei were centrifuged at 800 g for 10 min. The pellet was resuspended in hypertonic buffer (50 mM Tris·HCl, pH 7.5, 400 mM KCl, 400 mM NaCl, 10% glycerol, 1 mM EDTA, 5 mM DTT, 0.5 mM PMSF) and agitated at 4°C for 30 min. After centrifugation at 16,000 g for 25 min, the supernatant was diluted in an equal volume of dilution buffer (20 mM HEPES, pH 7.9, 50 mM KCl, 20% glycerol, 0.2 mM EDTA, 0.5 mM PMSF, 1 mM DTT) for protein quantification by the Bradford method.
Electrophoretic mobility shift assay.
Nuclear protein extracts (10 μg) of liver were incubated with 32P-labeled TRE oligonucleotides for either DR4 (5′ AGC TTC AGG_TCA_CAG GAG_GTC_AGA GAG 3′) or palindromic sites (5′ AAG ATT CAG_GTC_ATG_ACC_TGA GGA GA 3′) (Santa Cruz Biotechnology, Santa Cruz, CA). The binding reactions were incubated on ice for 40 min in a buffer containing 20 mM HEPES, 50 mM KCl, 20% glycerol, 0.2 mM EDTA, 0.5 mM PMSF, 1 mM DTT, 1 μg/μl BSA, and 1 μg/μl poly(dI-dC) (Pharmacia). The reaction mixture was loaded onto a 5% non-denaturating polyacrylamide gel and resolved at 120 V over the course of 4 h. The gel was dried, and the DNA-protein complexes were visualized by exposure in a Storage Phosphor Screen (Molecular Dynamics, Sunnyvale, CA). The screens were read in a Storm Phosphorimager and analyzed with the ImageQuant software (Molecular Dynamics).
Results obtained in all experiments were analyzed using a one-way analysis of variance coupled with Bonferroni's multiple-comparison test (control vs. treatments). Differences were considered statistically significant at P values of ≤0.01.
RESULTS AND DISCUSSION
In the present study, we investigated the genomic or extra-mitochondrial thyromimetic effects of 3,5-T2 at concentrations similar to those physiological for T3. Previous data from our laboratory had shown that, as in other vertebrates (4, 5, 16), hepatic D2 activity in the killifish is highly sensitive to thyroidal status (10). Thus, and due to the difficulties of measuring 3,5-T2, we initially established the concentration range of its regulatory action on D2 by comparing the effects of 3,5-T2 to those of T3 in intact (euthyroid) killifish. Dose- and time-dependent activities for both hormones are shown in Fig. 1. The well-known downregulatory effect of T3 was indistinguishable from that exerted by 3,5-T2, both showing a dose-dependent inhibition, which attained a plateau at an immersion dose of 100 nM and as early as 3 h postimmersion. These findings suggested that both hormones had comparable potencies at equivalent doses. To analyze 3,5-T2 effects in more detail, we used a model of hypothyroidism where the administration of active TH was anticipated to maintain euthyroidal expression of the various thyroid-dependent genes. As shown in Fig. 2, intrahepatic T3 levels were significantly decreased by the administration of MMI alone. As expected, this decrease was reversed by coadministration of T3 but not 3,5-T2. In fact, levels of endogenous intrahepatic T3 in fish treated with 3,5-T2 combined with MMI were no higher than in fish treated with MMI alone. This finding is very important, because it shows that the livers of animals treated with MMI coadministered with 3,5-T2 are indeed T3 deficient.
We measured the effects of exogenous T3, rT3, and 3,5-T2 on expression and activity of hepatic D2. Our hypothesis was that, if 3,5-T2 were bioactive, the fish would remain euthyroid in those groups treated with MMI plus T3 and 3,5-T2 but not with the inactive metabolite rT3. Blocking TH production with MMI significantly increased expression of D2 mRNA by 1.8-fold (Fig. 3A), which was paralleled by a 4.6-fold increase in hepatic D2 activity (Fig. 3B). Concomitant treatment with MMI and T3 prevented the MMI-induced increase in hepatic D2 mRNA and enzymatic activity (Fig. 3). These results are consistent with the tight regulation of D2 expression exerted by thyroidal status to maintain TH homeostasis (4, 5, 16). In addition, these data further demonstrate that 30 nM T3 can maintain normal physiological levels of D2 expression.
Despite the hypothyroid levels of endogenous intrahepatic T3 in fish treated with 3,5-T2, D2 mRNA expression and enzymatic activity were maintained at levels similar to those of untreated controls or fish treated with MMI+T3 (Fig. 3). These data strongly suggest that 3,5-T2 is sufficient to maintain D2 expression in liver at euthyroid levels.
Treatment with the inactive iodothyronine rT3 had no effect on MMI-induced increases in D2. This result differs from mammalian D2 regulation, where rT3 increases D2 degradation in the cerebral cortex (for review see Refs. 4, 5). This is the first study of the rT3 effect on the regulation of a teleostean D2. Although we cannot offer an explanation for the lack of effect of rT3 upon D2 regulation, this result adds to the unique aspects of iodothyronine deiodinases in teleosts.
As another test of our hypothesis, we analyzed the expression of GH, a gene that is classically upregulated by TH in the pituitary (11, 15). Here we demonstrated GH expression in the liver of killifish, as reported earlier in trout (33). This finding allowed the same end tissue to be analyzed throughout the study. As expected (14) and shown in Fig. 4, hepatic GH mRNA levels were reduced by threefold in hypothyroid animals. The effects of MMI on GH expression were reversed by 30 nM T3 but not by rT3, similar to the effects seen on D2 expression. When 3,5-T2 was coadministered with MMI, hepatic GH mRNA levels were similar to those in untreated or MMI+T3-treated fish. These data add further support for the hypothesis that 3,5-T2, like T3, can reverse the impaired expression of hepatic GH associated with the hypothyroidal state.
To determine whether the effects of 3,5-T2 on D2 and GH expression were specifically related to thyroidal status, we also measured the hepatic expression of the Na-K-ATPase α1-subunit gene. Na-K-ATPase α1 has been cloned from killifish, and its hepatic expression has been demonstrated (30). In contrast to the α2- and β-genes, the expression of the α1-isoform is unaffected by TH depletion (13), thus making it a suitable negative control for TH bioactivity. Hypothyroidism had no effect on hepatic Na-K-ATPase α1 expression (Fig. 5). Furthermore, treatment with replacement doses of T3, 3,5-T2, or rT3 induced no changes in Na-K-ATPase α1 expression. Although these results suggest that 3,5-T2 bioactivity could be limited to the expression of TH-dependent genes, additional work involving other nonthyroid-dependent genes is needed to confirm this possibility.
To establish the role of 3,5-T2 in modulating expression of TH-responsive genes, we examined the binding of hepatic nuclear protein extracts to oligonucleotides corresponding to known mammalian TREs. Canonical TREs identified in mammals include the AGGTCA half-site, repeated and separated by a four-base spacer (DR4), as well as a 12-base palindrome formed by inverted duplication of the half-site. Extracts from untreated fish formed two specific, protein-DNA complexes, with labeled oligonucleotides corresponding to either the DR4 or palindromic sequences (Fig. 6). Treatment with MMI virtually eliminated the paired protein-oligonucleotide complexes for both TREs. In animals treated with MMI+T3, the intensity of the lower molecular weight complex returned to normal levels, whereas the higher molecular weight band was unchanged. Conversely, cotreatment with 3,5-T2 restored formation of only the high molecular weight complex with both oligonucleotides. These data not only demonstrate that T3 and 3,5-T2 stimulate binding of proteins to TREs, but suggest that the two THs activate different DNA-binding proteins. In this context, the upper band suggests the presence of heavier or additional proteins in the 3,5-T2 recruited complex. Nevertheless, it is difficult to speculate about the composition of these complexes recruited by TRE, because variable and distinct recruitment patterns of coactivators occur on individual target genes regulated by TH (18). The exact mechanism of T3- and 3,5-T2-dependent recruitment of nuclear transcription factors or other DNA-binding proteins remains unknown.
Although other studies indicate that 3,5-T2 exhibits low affinity for the canonical TR in both mammalian (17) and teleostean (6, 7) species, our results strongly suggest that both T3 and 3,5-T2 act directly on TH-dependent genes in the liver by binding and activating TRs and interacting with TREs in the promoter regions of these genes. In this context, Baur et al. (3) also suggested an activation of nuclear T3/retinoid X-receptor heterodimers in rat pituitary in the presence of 3,5-T2. These authors observed that 3,5-T2 rapidly activates proteins that can specifically bind to the DR4 TRE that was identified in the promoter region of the human Dio1 gene and that this DNA-protein interaction was as potent as when stimulated by T3. In the present study, we evaluated the possible participation of TR in 3,5-T2 gene regulation. Indeed, as shown in Fig. 7, the fact that the upregulation of TRβ1 expression by hypothyroidism was reversed by T3 and 3,5-T2 indirectly suggests the participation of TRβ1 as part of the transcriptional machinery that mediates the action of both iodothyronines. This matter deserves further attention and is currently being studied in our laboratory.
Attempts to demonstrate 3,5-T2 bioactivity have been made since the late 1980s, and now it is well accepted that 3,5-T2 has clear effects on oxygen consumption that are distinct from those of T3 and are not dependent on protein synthesis (12, 19). In support of these nongenomic effects, a very low affinity of 3,5-T2 for TR α1, β1, and β2 (∼100× lower than T3) has been described in mammals (17) and teleosts (6, 7). Nevertheless, studies using rat models or in vitro systems to examine the effects of 3,5-T2 on malic enzyme, GH and TRβ2 (2), TSH (3), and IGF binding protein type 4 (8) have shown that 3,5-T2 can indeed influence mRNA expression or concentration. However, in most of these studies, 3,5-T2 at doses up to 100-fold greater than those of T3 were used to generate comparable effects. Our laboratory's previous finding that fish hepatic D1 and D2 are downregulated transcriptionally by supraphysiological doses of both T3 and 3,5-T2 (10) also provided evidence that 3,5-T2 has effects within the nucleus. The present work extends these findings by showing that, at equivalent doses, both 3,5-T2 and T3 maintain transcription of TH-responsive genes at euthyroidal levels. Furthermore, 3,5-T2 and T3 induce formation of two distinct TRE-protein complexes. Although not all components of these complexes have yet been identified, both appear to contain TRβ1. Whether 3,5-T2 has thyromimetic effects on all TH-target genes or an effect in all vertebrates is still an open question. Understanding TH metabolism and action across different vertebrates may require us to reconsider established schemes of thyroid physiology.
This work was partially supported by Programa de Apoyo a Proyectos de Investigación e Innovación Tecnológica (PAPIIT) Grants PAPIIT IN222405 and PAPIIT IN202807.
We gratefully acknowledge Patricia Villalobos and Anaid Antaramian for technical assistance. We would like to express our gratitude to Dr. Dorothy Pless for critically reviewing the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society