| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Laboratoire de Physiologie Générale et Comparée (H.G., S.M.D., N.B., I.S., B.A.D.), Unité Mixte de Recherche 8572, Centre Nationale de la Recherche Scientifique, Muséum National d’Histoire Naturelle, 75231 Paris, cedex 5, France; and Institut de Biologie Animale (E.J., B.D.), Universite de Lausanne, Batiment de Biologie, CH-1015 Lausanne, Switzerland
Address all correspondence and requests for reprints to: Dr. Barbara Demeneix, Laboratoire de Physiologie Generale et Comparee Museum National d’Histoire Naturelle, Unité Mixte de Recherche 8572, Centre Nationale de la Recherche Scientifique, 75231, Paris cedex 5, France. E-mail: demeneix{at}mnhn.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1, plays a specific role in mediating T3-dependent repression of hypothalamic TRH transcription. To investigate the structural basis of isoform specificity, we compared the transcriptional regulation and DNA binding obtained with chimeric and N-terminally deleted TRs. Using in vivo transfection assays to follow hypothalamic TRH transcription in the mouse brain, we found that TRß1 and chimeras with the TRß1 N terminus did not affect either transcriptional activation or repression from the rat TRH promoter, whereas N-terminally deleted TRß1 impaired T3-dependent repression. TR1 or chimeras with the TR1 N terminus reduced T3-independent transcriptional activation and blocked T3-dependent repression of transcription. Full deletion of the TR1 N terminus restored ligand-independent activation of transcription. No TR isoform specificity was seen after transcription from a positive thyroid hormone response element. Gel mobility assays showed that all TRs tested bound specifically to the main negative thyroid hormone response element in the TRH promoter (site 4). Addition of neither steroid receptor coactivator 1 nor nuclear extracts from the hypothalamic paraventricular nuclei revealed any TR isoform specificity in binding to site 4. Thus N-terminal sequences specify TR T3-dependent repression of TRH transcription but not DNA recognition, emphasizing as yet unknown neuron-specific contributions to protein-promoter interactions in vivo. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and c-erbA-ß (1, 2). TRs and ß are similar in overall structure, being most related in the Cys-rich DNA-binding (DBD) and C-terminal ligand-binding (LBD) domains (3). Further diversity occurs by alternative splicing of the TR primary transcripts generating C-terminal TR1 and TR2 variants (4, 5, 6). TR2 fails to bind T3 (7). N-terminal variants of TRß have been described previously (8, 9, 10).
During vertebrate brain development, TR and -ß genes show distinct spatio-temporal distribution patterns, with early and ubiquitous expression of TR mRNAs, but later, more restricted, expression of TRß mRNAs (3). In most species, timing of TRß expression correlates closely with known TH-dependent developmental changes (3, 11). These spatially and temporally defined patterns of TR and -ß expression are correlated with specific functions of these receptors in brain development (12, 13, 14). Moreover, TRß isoforms show brain region-specific profiles; although both TRß1 and TRß2 are found in the hypothalamus and pituitary, the TRß1 isoform has a generally wider pattern of expression (15, 16).
TRs act on target genes by binding to specific regulatory DNA sequences, called TH response elements (TREs) (for review, see Ref. 17). On positively regulated genes, TRs function as repressors of basal promoter activity in the absence of T3, and repression is relieved by T3. However, TRs also mediate ligand-dependent repression of certain target genes. In particular, the negative feedback effects of T3 on the hypothalamic/pituitary axis act through repression of genes encoding TRH (12, 18, 19) and TSH (20).
TRß1 and TRß2 have specific roles in mediating T3-dependent transcriptional repression of hypothalamic TRH (12, 19, 21), these effects being correlated with high levels of these TRß isoforms in the hypothalamus (15, 22). Increasing expression of TR, in either chick hypothalamic neurons in primary culture (12) or in mice hypothalami in vivo (21), blocked the negative feedback effects of T3 on TRH transcription. Similar results have been reported using cell lines (23, 24). These observations raise the key question as to what is the structural basis of the isoform specificity (TR vs. TRß) in mediating this physiological feedback of T3 on TRH. Indeed, although much current work addresses how TRs interact on negative TREs with nuclear corepressor and histone deacetylase (HDAC) partners (25, 26), little data are available on the structural basis for TRß vs. TR specificity in this regulation.
We focused on the N-terminal region of TR and TRß because this region shows the largest differences between isoforms. Chimeric TRs were created by interchanging the main domains of rat (r) TR1 and rTRß1, producing ßß, ß, ß, and ßß. A number of N-terminally deleted TRs were also constructed. We compared the transcriptional activities and TRE binding capacities of these wild-type and modified receptors. Their transcriptional effects in the T3-dependent repression of TRH expression were followed in vivo directly in the hypothalamus of newborn mice. In this experimental paradigm we are exploiting the unique physiology of the hypothalamic nuclei that express endogenous TRH and we can thus analyze the transcriptional effects of the different TRs in an integrated context (21). We also addressed whether TR isoforms showed specific binding on the main negative TRE in the rat TRH promoter. The results show that the N-terminal sequence is sufficient to confer TR isoform specificity for ligand-dependent transcriptional repression of TRH, but that a negative TRE within this promoter does not by itself recognize this specificity.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 (vs. TRß1) responsible for abrogating T3-dependent inhibition of transcription from the TRH promoter, four chimeric receptors were created (Fig. 1A
), and a series of N-terminally deleted TRs were subcloned into the same expression vector (Fig. 1B). Using an in vivo transfection assay with a TRH-luciferase (TRH-luc) construct, we examined the effects of these chimeric and deleted receptors on the TRH promoter activity, in situ, in the hypothalami of hypothyroid newborn mice. The TRH-luc plasmid used contains the regulatory regions that are necessary and sufficient to observe a physiologically relevant T3-dependent transcriptional regulation in the hypothalamus of newborn mice (21).
|
, A and B, and Fig. 3, A and B). This repression was always significant with a P value consistently less than 0.05. To verify the specificity of this response we mutated the predominant negative TRE in the rat promoter (between bp -59 and -54). This region corresponds to site 4 (TGACCT) identified in the human and mouse TRH promoters (32, 34). Figure 2A shows that there was no T3-dependent repression from the mutated promoter (Fig. 2A, right columns). In fact, mutation of this site significantly depressed the level of TRH expression in the absence of hormone (Fig. 2A, P < 0.05).
|
|
, second pair of columns from the left). Using an N-terminally deleted TRß1 (
Nß1) significantly modified this result (Fig. 2B, third pair of columns from the left), abrogating repression in the presence of ligand while not affecting activation of TRH transcription in the absence of T3. Using chimeras with the TRß1 N terminus, i.e. TRßß or TRß, permitted T3-dependent inhibition of TRH transcription (Fig. 2B, far right columns). Indeed, chimeric TRßß and TRß are compatible with statistically significant decreases in TRH transcription (40% and 50%, respectively) in T3-injected animals, the decreases being equivalent to those seen in controls (P < 0.001).
To further examine TR N terminus specificity, we used a similar approach with TR1. Cotransfecting rTR1 with TRH-luc reduces ligand-independent (-T3) TRH-luc transcription by 42%, and addition of T3 does not further reduce TRH-luc transcription (Fig. 3A, second pair of columns from the left). We next tested the effect of expressing chimeras bearing either the TR1 N terminus or the TR1 N terminus with the TR1 DBD (TRßß, ß). These chimeras provoke a further, significant, reduction in ligand-independent TRH-luc transcription and, again, no ligand-dependent regulation is induced (Fig. 3A, far right columns). Thus, the presence of the TR1 amino terminus is sufficient to cause a reduction in basal transcription and the loss of T3-dependent negative feedback on TRH transcription.
To determine which part of the N terminus accounts for the specificity of TR1, we used N-terminally deleted constructions. As shown in Fig. 3B, the levels of TRH transcription with the N19 mutant was not different from that with the wild-type TR1, whether with or without T3. Further deletion of the N terminus (N30 or N1) restored ligand-dependent inhibition (P < 0.05). As regards activation of transcription without T3, only removal of the entire N terminus (N1) restored basal levels of transcriptional activation giving profiles of transcription similar to controls.
Taken together, these results show that the TRß1 amino-terminal sequence is necessary and sufficient to confer T3-dependent transcriptional repression of TRH. In contrast, the TR1 amino-terminal sequence blocks T3-independent activation of TRH transcription.
TR1 and TRß1 Exert Equivalent Transcriptional Effects on a Positive TRE
To determine whether the differential effects of TR1 and TRß1 were specific to the TRH promoter, we tested their effects on a positively regulated promoter, that of the malic enzyme (ME) gene, with a ME-chloramphenicol acetyltransferase (CAT) construct used in an identical in vivo cotransfection paradigm as described above. As shown in Fig. 4, we find that, as for the controls (P < 0.05), both TRß1 and TR1 do not give any significantly different effects on transcription from the ME promoter in the presence of T3 (P < 0.05 and P < 0.01, respectively). This emphasizes that isoform-specific effects on transcription measured in the hypothalamus are restricted to the negatively regulated promoter, TRH, and obviates the need to test the chimeras on this TRE construct.
|
The TRE sequences within the rat TRH promoter used for gel-shift analysis were chosen in accordance with the predominant negative TRE (site 4: TGACCT) identified in the human and mouse TRH promoters (32, 34), the physiological significance of which was demonstrated in the in vivo transfection assay (Fig. 2A). In the rat, the sequence (including site 4) encompasses bases -74 to -34. In addition, sequences corresponding in position (but not identity) with two weak monomeric binding sites for TRs in the human promoter were also used (bases +9 to +47), hereafter named potential sites 5 and 6.
As shown in Fig. 5A, site 4 binds rTR1 and rTRß1 as monomers (lanes 1 and 3) and heterodimers with RXRß (lanes 2 and 4), no distinct homodimers being seen. Homodimers for rTRß1 are evident, however, on the positive TRE from the myelin basic protein (MBP) gene, which was used as a control for homodimer binding (Fig. 5D) (31). Mutating site 4 (Fig. 5B) leads to a loss of all forms of binding. The interaction with potential sites 5 and 6 was also examined, but no significant TR binding was seen (Fig. 5C). To assess the specificity of TR/RXRß binding to site 4, we titrated binding of labeled site 4 against increasing amounts of unlabeled probe. As seen in Fig. 5E, both rTR1/RXRß and rTRß1/RXRß heterodimers strongly bind site 4, and this binding is antagonized with increasing amounts of nonradiolabeled site 4 probe. No competition is seen in the presence of a nonspecific (mutated site 4) probe. These data demonstrate that the sequence from -74 to -34 (site 4) of the rat TRH promoter contains a specific binding site for TR and TR/RXRß.
|
). This is consistent with the binding activity of wild-type TRs (Fig. 5A) on this negative TRE. In some lanes a faint band is seen at the level of homodimers, but this was not reproducible from gel to gel. The effect of T3 on TR/RXRß complexes binding on site 4, with wild-type as well as chimeric TRs, was examined by carrying out identical assays as described above (Fig. 6B). Densitometric analyses performed with similar gels revealed no significant differences in binding of TR/RXR isoforms in the presence or absence of T3.
|
). Addition of ligand produces a shift for TRs with SRC-1, both with and without RXR (compare lanes 2, 5, 8, and 11 with lanes 3, 6, 9, and 12, respectively).
|
, SRC-1 abrogated ligand-independent activation in a dose-dependent manner from the TRH promoter. T3-dependent transcriptional regulation was thus lost (Fig. 7B, far right columns). Combining SRC-1 and TR expression restored T3-dependent transcriptional regulation (Fig. 7C, two last pairs of columns), producing exactly the same result with TR and TRß.
Nuclear Extracts from the Hypothalamic Paraventricular Nucleus (PVN) Do Not Reveal Any TR/RXR Isoform Specificity
To determine whether there was a tissue-specific component that could affect TR isoform binding to site 4, we used nuclear extracts from the PVN of the hypothalamus and from the cerebellar cortex of adult mice. As shown in Fig. 8, addition of PVN extracts increased heterodimeric binding to site 4 (compare lanes 2, 3, 4, and 5 with 8, 9, 11, and 12, respectively). Densitometric analysis showed the increase to be 2.15 times the level in the absence of extract (n = 8, P = 0.0024). In contrast, addition of nuclear extracts from the cerebellum enhanced heterodimeric binding to a lesser degree (1.38-fold control values, n = 8, P = 0.0024). Similarly, extracts of PVN without added TRs showed a signal on site 4 at the level of the TR/RXR shift (lane 7) that was absent both for the mutated site 4 (lanes 10 and 13) and for extracts of cerebellum (lane 14).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and TRß genes throughout evolution (39) and correlates with tissue-specific and developmentally distinct patterns of TR vs. TRß expression (15). To address the question of the structural basis for this specificity, we used chimeric and deleted mutant receptors using transcription and gel-shift assays. The transcriptional studies were based on a physiological relevant in vivo assay. Indeed, it has been shown previously that hypothyroidism and T3, respectively, up- and down-regulate transcription from the TRH-luc construct when it is introduced into the hypothalamus of newborn mice (21). The T3-dependent repression observed can be attributed to endogenous TRs, known to be expressed in the hypothalamus (15, 22). Furthermore, it has been shown in this experimental paradigm, that using the same expression vectors for each TR isoform cotransfected with TRH-luc, increases T3 binding in the hypothalamus to equivalent amounts (
2-fold) (21), while having differential effects on TRH-luc transcription.
A number of key findings arise from the data presented here using this approach. First, the N-terminal domain of TR1 reduces ligand-independent transcriptional activation of TRH and abrogates ligand-dependent transcriptional repression. Second, these effects are not shared by the TRß1 N terminus, and deletion of the TRß1 N terminus actually removes ligand-dependent repression from the TRH promoter. Third, the structural differences in the TRß1 and TR1 isoforms that confer each of these specific transcriptional effects do not significantly affect their binding to the negative TRE in the TRH promoter. These results emphasize that N-terminal sequences can confer TR isoform specificity for ligand-dependent transcriptional repression of TRH but not DNA recognition. This highlights the importance of neuron-specific promoter context to specify protein-DNA interactions and transcriptional effects. Such cell-specific protein-DNA interactions will be particularly crucial for such key transcriptional regulations underlying negative feedback loops in the hypothalamo/hypophyseal system.
The TRH and TSHß genes are both negatively regulated by T3. Transcriptional repression of these genes involves TRß isoforms. In the pituitary, the TRß2 isoform has been shown to be the most potent regulator of the TSHß gene, although TRß1, and to a lesser extent, TR isoforms are capable of mediating repression of the TSH gene (14, 40). In the hypothalamus, TRH gene expression has been shown to be enhanced in mice devoid of the TR1, -ß1, and -ß2 isoforms (41). In TRß-/- mice (40), the regulation of the TRH gene has not yet been studied. More recently, a predominant role for the TRß2 isoform in regulating the TRH gene has been revealed by using TRß2-/- mice (19). In these mice there is a loss of TRH up- or down-regulation, in hypothyroid and hyperthyroid animals, respectively.
The data reported here provide evidence that overexpression of a truncated TRß1 isoform in the hypothalamus can impair TRH regulation in vivo, whereas a full-length TRß1 isoform is compatible with TRH regulation in vivo. A possible criticism of our present studies could be that we use the rTRß1 and chimeric proteins derived from it rather than rTRß2. Using the in vivo, hypothalamic transcription assay, we found TRß2 and TRß1 equally compatible in ligand-dependent repression of TRH (21). In TRß2-/- mice, where TRH regulation is impaired (19), it is possible that the expression of the other TR isoforms is also affected in the hypothalamus. Indeed, in these TRß2-/- mice, TRß1 expression was examined only qualitatively in the whole brain (14), and very little data are available on the effects of expression of one TR isoform on another, particularly in the hypothalamus, where isoform expression is a key consideration. Thus, we cannot exclude a role for TRß1 in the T3-dependent repression of the TRH gene. It is possible that both TRß1 and TRß2 isoforms contribute to the regulation of the TRH gene, perhaps through different mechanisms involving specific comodulators.
That the greatest differences between TRß and TR lie in their N termini suggests a significance for this domain in T3-dependent transcriptional repression. Our data on the transcriptional effects of the chimeric proteins show unambiguously that the N terminus does determine T3-dependent repression of TRH transcription. The chimeras bearing the TRß N terminus (ß and ßß) permitted ligand-dependent repression, whereas the two bearing the TR N terminus (ßß and ß) impaired ligand-dependent repression. Moreover, deletion of the full TR N terminus restored both activation and ligand-dependent repression. In contrast, deletion of the TRß N terminus removed the characteristic ligand-dependent repression seen with the wild-type receptor and the two chimeras, ß and ßß. Thus, the full TR N terminus contains a sequence that annuls activation in the absence of ligand, whereas the full N terminus of TRß confers repression in the presence of ligand. Amino acid differences (16%) in the other domains could account for the different transcriptional effects of TRNß1 vs. TRN1.
Examination of and ß N termini shows virtually no intraspecific homology between rTR1, rTRß1, and rTRß2 (8). Thus, we looked for interspecific homologies within the N terminus of each isoform. Comparison of aa 19–30 of chicken TR, rTR, and human TR shows the conservation of 10 aa, with a 5-aa core (KRKRK), within otherwise dissimilar N termini. Results from other groups (29) also indicate that this region (aa 19–30) of rTR1 carries transactivation properties on positive TREs. Our results corroborate the idea that these aa contribute to isoform transactivation specificity, as their deletion restores ligand-dependent repression. Interestingly, Hadzic et al. (42), using a positive TRE, showed that full-length chicken TR in HeLa cells enhances T3-dependent transcription more efficiently than an N terminus-shortened form, and this preferential activation is due to the N-terminal activation function-1 domain (aa 21–30). The opposite effect seen here emphasizes the importance of this sequence and suggests that interaction of this sequence with other proteins will be promoter dependent. The role of TR in blocking regulation of TRH transcription may also have physiological relevance. In the hypothalamic/pituitary axis, TR0/0 mutants are hypersensitive to TH (43), suggesting that they are involved at some level of the negative feedback controls.
Using putative negative TREs within the rat TRH promoter and first exon in an EMSA, we found specific binding only on the TRE between bp -74 and -43 (site 4). This site 4 is well conserved across species (27, 32, 44), and our transcription experiment (Fig. 2A) with the mutated site 4 demonstrates its essential role in T3-dependent regulation of the rat TRH gene. As to the weaker, potential negative TRE [sites 5 and 6 of the human TRH promoter, (32)] we found no binding of any TR to this site. The specific binding of TRs to site 4 fits with the data of Satoh et al. (34), who showed that this site in the mouse TRH promoter binds TRß1.
This brings us back to the central problem of how the N terminus of TR1 could hinder, in the absence of T3, TRH activation through site 4. There is, in fact, a near-canonical CRE (cAMP-responsive element) site, juxtaposed to site 4 (45). This CRE is conserved among species and has been shown to be responsible for transcriptional activation of TRH induced through -melanocyte-stimulating hormone signaling (46). A plausible hypothesis would be that the N terminus of TR1 (most likely aa 30–50) specifically interferes with this pathway in the absence of T3. Alternatively, the N terminus of TR1 could hinder T3-independant activation possibly mediated by TRß isoforms, by competition for binding on site 4.
Neither gel shift analyses nor transcription studies with TRH or ME promoters indicated any influence of the DBD in TR specificity. This suggests that the DBD is of lesser importance for conferring ligand-dependent repression than the N terminus. However, in an in vitro assay, the DBD was shown to be vital for interaction of TRß1 with a histone deacetylase (HDAC2) and for transcriptional repression of TSHß (26). In the presence of ligand, both TRß1 and HDAC2 interacted with a negative TRE in the TSHß promoter (26). However, TRß1 and TR1 bind equally well to HDAC2 in vitro (26). This interaction may involve the N terminus as N-terminally deleted TRß1 showed diminished interactions with HDAC2. This deletion analysis was limited to TRß1; therefore, whether such interactions involve TR isoform specificities and whether they can be extended to the TRH promoter remains to be investigated.
Coactivators and corepressors modulate nuclear receptor activity. Although little data are available for negative TREs, on positive TREs it is known that unliganded TRs interact with corepressor proteins such as nuclear corepressor (NCoR) and silencing mediator for retinoid and thyroid hormone receptors (SMRT) mediating ligand-independent repression (47, 48, 49, 50). Ligand binding releases corepressor proteins from TRs (43, 45) and recruits coactivator proteins such as the steroid receptor coactivator-1 (SRC-1) (for review, see Ref. 35), switching transcription of positively regulated genes from repressive states to active states. As we have recently shown low levels of expression of NCoR and SMRT mRNA in the PVN (37), we concentrated on potential differential interactions of TR1 and TRß1 on site 4 with a typical coactivator (SRC-1), which is highly expressed in the PVN (36). Our data show both TRs had similar binding profiles, interacting with SRC-1 in the presence, but not in the absence, of T3. Other authors (44, 51) observed a shift on the TRH promoter in the presence of TRß1 with NCoR and SMRT on site 4. These data indicate that TR-comodulator interactions are similarly modulated by T3, whether on a positive or on a negative TRE.
As TR and TRß did not interact differentially with SRC-1 on site 4, we further examined potential TR-SRC-1 interactions in vivo. We observed no synergy between SRC-1 and TR1. Moreover, the fact that both combinations of SRC-1/TR1 and SRC-1/TRß1 did not affect TRH-luc regulation in vivo indicates that the differential effects of TR1 are probably due to another, isoform-specific, partner. Our gelshift experiments performed with nuclear extracts from PVN or cerebellar cortex did not reveal any significant complexes with any TR on site 4, thus emphasizing the need for more sensitive identification strategies (such as the yeast double-hybrid system). The fact that PVN nuclear extracts enhanced TR/RXR binding, more than cerebellar controls, could be due to lower TRß mRNA expression in the cerebellum as compared with PVN (15).
In conclusion, the N-terminal sequence of TR and TRß is sufficient to confer isoform specificity for ligand-dependent transcriptional repression of TRH, but not for recognition of the negative TRE within this promoter. A hypothesis to explain this isoform specificity could be that TR and TRß adopt distinct conformations on negative TREs but not so distinctly on positive TREs. The amino termini of each isoform would be the main factor determining these conformational differences on negative TREs. Indeed, it is possible that N- to C-terminal interactions of certain nuclear receptors could define receptor interactions with comodulator proteins, and that such receptor-comodulator interfaces will be modified as a function of the response element involved.
Recently, the unique N terminus of the TRß2 isoform has been shown, in the absence of T3, to recruit coactivators (52, 53) and to interfere with corepressor function (54). These interactions provide an explanation for the T3-independent activation described for TRß2 on the TRH promoter in CV-1 cells (24). However, another report shows no difference between TRß2 or TRß1 on the T3-dependent repression of the TRH promoter (34). It is important to recall that comodulators are not ubiquitously expressed in the brain (36, 37, 55). Thus it is critical to analyze comodulator expression in the hypothalamus and particularly in the PVN, where TRH is expressed and regulated.
Determination of the profiles of comodulator proteins expressed in the PVN will be an essential prelude to dissecting how isoform N terminus specificity relates to physiological feedback on TRH transcription in the hypothalamus. Once such profiles are obtained, it will become relevant to carry out analyses of TR interactions with either comodulator proteins or other chromatin-modulating proteins that are specific to TRH neurons. Such studies will contribute to understanding how ligand-dependent transcriptional repression is obtained in the physiological context of hypothalamic feedback.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The TRH-luc construct contains a rat TRH gene 5'-fragment extending from -547 to +84 bp cloned upstream of the firefly luciferase-coding region (27).
The TRH4-luc construct was obtained by deleting bp -59 to -54 (site 4) of the initial TRH-luc construct, using the QuikChange XL Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA) according to the manufacturer’s conditions. The primers used for this deletion are described below:
5'-CGC CCC CTC CCC GCA CAG GCG CCG CG-3'
5'-CGC GGC GCC TGT GCG GGG AGG GGG CG-3'
A chloramphenicol acetyltransferase reporter gene construct, ME(-315/+3)-CAT, has been previously described (28).
rTR1 and rTRß1 cDNAs were subcloned into pSG5 plasmid (Stratagene). All the following mutants were also created in pSG5.
Four rTR chimeras were created and correspond to the exchange of domains between TR1 and TRß1 (Fig. 1). The chimera ß (412 aa) contains the TR1 aa Met1-Ser177 and Glu222-Asp456 of TRß1. ßß (470 aa) contains the TRß1 aa Met1-Ala229 and Val170-Val410 of TR1. ßß (407 aa) contains the TR1 aa Met1-Tyr46 and Leu96-Asp456 of TRß1. The chimera ß (459aa) contains the TRß1 aa Met1-Leu101 and Cys53-Val410 of TR1.
An N-terminal deletion mutant of rTRß1 was prepared by replacing the HindIII–XbaI fragment of pSG5-rTRß1 with a double-stranded oligonucleotide containing a HindIII restriction site, the Kozak sequence, and the initiation codon, a nucleotide sequence encoding 6 aa starting from position 90 of rTRß1, and an XbaI site.
Three successive N-terminal deletion mutants of rTR1, including the first methionine and containing amino acids starting from position 19, 30, or 51 of the initial rTR1 (29), were subcloned into pSG5.
All constructs were sequenced through 400 nucleotides starting from the pSG5 HindIII site and translated in vitro, using the TNT Coupled Reticulocyte Lysate System (Promega Corp., Madison, WI).
Human SRC-1a in pSG5 (30) was kindly provided by Dr. Chatterjee (Cambridge, UK) and Dr. Collingwood (Richmond, CA).
Treatment of Animals, in Vivo Transfection, and Luciferase Assay
All animal studies were conducted in accordance with the highest standards of human care and according to the principles and procedures described in Guidelines for Care and Use of Experimental Animals.
Female OF1 mice (Janvier, Le Genest St. Isle, France) were mated. To induce fetal and neonatal hypothyroidism, dams were given an iodine-deficient food containing 0.15% 6-n-propyl-2-thiouracil (PTU) at d 14 of pregnancy (Harlan, Gannat, France). The 6-n-propyl-2-thiouracil diet was continued throughout the lactation period. For evaluating T3 effects on reporter gene expression, hypothyroid pups were injected sc with 250 µg of T3/100 g of body weight (b.w.) (in 9% saline). Controls received saline (9%) injections. DNA/polyethylenimine complexations, in vivo transfection, and luciferase assay were carried out as described previously (21).
Dissection of Hypothalami
Given the highly tissue-specific nature of TRH transcription, one of the most important steps in ensuring reproducibility is careful and consistent microdissection of the hypothalamic areas transfected. Brains were rapidly removed and placed on a petri dish in contact with ice. A precise 2-mm3 block of hypothalamic tissue enveloping the paraventricular nucleus was dissected out and transferred to ice-cold lysis buffer for luciferase or CAT assay.
CAT Assay
Hypothalami were homogenized in 150 µl 250 mM Tris-HCl (pH 7.4). The homogenates were centrifuged for 10 min at 4 C (11,000 x g). The supernatants were removed, and aliquots (50 µl) were transferred to Eppendorf tubes (Eppendorf North America, Inc., Madison, WI) containing 40 µl of 250 mM Tris-HCl buffer (pH 7.4) and heated at 65 C (10 min). The reaction was started by adding 10 µl of butyryl-coenzyme A (0.53 mM) and [14C]chloramphenicol (0.01 mM, 1.85 kBq per tube). The mixture was incubated at 37 C (1 h), and butyrylated forms of [14C]chloramphenicol were extracted after centrifugation (4 C, 11,000 x g) by addition of 2 volumes of 2,6,10,14-tetramethylpentadecane/xylene, 2:1. Supernatants were removed, and the products were quantified in a scintillation counter (LKB, Rockville, MD).
EMSAs
TRs for EMSA were obtained by in vitro transcription and translation using the TNT Coupled Reticulocyte Lysate System (Promega Corp.). To quantify protein production, [35S]methionine incorporation and direct visualization on SDS-PAGE (10%) were used. As a source for RXRß, we used nuclear extract of Sf9 cells infected with a recombinant baculovirus overexpressing mouse RXRß as described previously (31). As a source for comodulators, the bacterial pGEX system was used (Amersham Pharmacia Biotech, Arlington Heights, IL) to obtain glutathione-S-transferase-SRC-1-ID (31), which includes aa 186–401 of hSRC-1, according to GenBank accession no. U40396.
Double-stranded oligonucleotides used as probes were radiolabeled with [-32P]dCTP by a fill-in reaction using a Klenow fragment of DNA polymerase. Unincorporated [-32P] dCTP was removed by G-50 Sephadex chromatography. Sequences of the upper strand of each oligonucleotides are as follows:
-Site 4 from -74 to -34 bp of the rat TRH promoter [encompassing the negative TRE homologous to site 4 in the human TRH promoter (32)], and its mutant, MUT 4 [mutations were chosen in accordance with those characterized by Hollenberg et al. (32)]:
Site 4: 5'-GCGCCCCCTCCCCGCTGACCTCACAGGCGCCGCGTCTCCA-3'
MUT 4: 5'-GCGCCCCCTCCCCGCTAAAATCACAGGCGAAAAAAAACCA-3'
Potential sites 5 and 6 from +9 to +47 of the rat TRH promoter [equivalent in position to those defined in the human TRH promoter (32)] are as follows:
Potential sites 5 and 6: 5'-GACCCTGGATTCGGGAGTATTGCAAACTCTACCCAGCCAG-3'
Sequence from the MBP (myelin basic protein) promoter (31):
MBPTRE: 5'-GAT CAG AAC AAT GGG AGC TCG GCT GAG GAC ACG GC-3'
For the binding reaction, the proteins (5 µl of reticulocyte lysate programmed for each TR protein per reaction and, when necessary, 1 µg of SRC-1 or 0.5 µl of a Sf9-RXRß extract), were preincubated in the presence of 3 x 104 cpm of labeled probe for 20 min at room temperature, in a buffer with 25 mM HEPES (pH 7.5), 5 mM MgCl2, 1 mM EDTA, 10% glycerol, 40 mM KCl, 1 mM dithiothreitol, 2 µg of poly (dI-dC), and 150 nmol of T3 (when necessary) in a total volume of 20 µl.
For brain nuclear extracts, PVN and cerebellar cortex were dissected out from 2.5-month-old male mice and processed according to Beckmann et al. (33). Nuclear extracts (10 µg) were added to the binding reaction, to a final volume of 30 µl.
The complexes were separated on a 5% native polyacrylamide gel. For DNA binding competition experiments, a 5- to 200-fold molar excess (as indicated) of the unlabeled double-stranded competitor oligonucleotide was added for an additional 15 min after the incubation reaction. Gels were visualized by autoradiography. Gelshift quantifications were performed using Molecular Dynamics, Inc. 445 SI PhosphorImager (Amersham Pharmacia Biotech) and ImageQuant 1.1 software (Molecular Dynamics, Inc., Sunnyvale, CA; Amersham Pharmacia Biotech).
Statistical Analysis of Results
In vivo gene transfer results are expressed as means ± SEM per group. After ANOVA analysis where appropriate, the t test was used to analyze differences between groups. Differences were considered significant at P < 0.05. In all cases, typical experiments are shown. Each experiment was carried out with n 
10, repeated at least three times and provided the same results.
Gelshift quantifications were analyzed using Friedman Nonparametric Repeated Measures Test.
|
ACKNOWLEDGMENTS |
|---|
1 constructs, and Dr. Chatterjee (Cambridge, UK) and Dr. Collingwood (Richmond, CA) for hSRC-1 in pSG5. |
FOOTNOTES |
|---|
1 H. G. and S. D. contributed equally to this work. 
Abbreviations: aa, Amino acids; b.w., body weight; CAT, chloramphenicol acetyltransferase; DBD, DNA-binding domain; HDAC, histone deacetylase; hSRC, human SRC; LBD, ligand-binding domain; MBP, myelin basic protein; ME, malic enzyme; NCoR, nuclear receptor corepressor; rTR, rat thyroid hormone receptor; RXR, retinoic X receptor; SRC-1, steroid receptor coactivator 1; TH, thyroid hormone; TR, thyroid hormone receptor; TRE, thyroid response element; TRH-luc, TRH-luciferase.
Received for publication July 11, 2001. Accepted for publication March 12, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and ß thyroid hormone receptor genes. EMBO J 9:1519–1528[Medline]
isoforms generated by alternative splicing differentially activate myosin HC gene transcription. Nature 334:539–542[CrossRef][Medline]
- and ß-forms for the acetic acid analog and failure of the human testis and kidney -2 products to bind T3. Mol Endocrinol 4:227–234[Abstract]
thyroid hormone receptor oligonucleotides. J Biol Chem 270:12100–12108- and ß-thyroid hormone receptor mRNAs, including the ß 2-subtype, in the developing mammalian nervous system. J Neurosci 12:2288–2302[Abstract]
and ß c-erbA genes in developing rat brain. Mol Endocrinol 5:1339–1350[Abstract]
1, ß1, but not ß2 contains a strong thyroid hormone-dependent transactivation function. Proc Natl Acad Sci USA 92:5600–56041 and ß1. J Biol Chem 273:24239–24248 and TRß in the control of thyroid hormone production and post-natal development. EMBO J 18:623–631[CrossRef][Medline]
is essential for transcriptional activation and interaction with the general transcription factor TFIIB. Mol Cell Biol 15:4507–4517[Abstract]
. Proc Natl Acad Sci USA 98:349–354
