| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Second Division, Department of Internal Medicine, Hamamatsu University School of Medicine, Hamamatsu, Shizuoka 431-3192, Japan
Address all correspondence and requests for reprints to: Shigekazu Sasaki, Second Division, Department of Internal Medicine, Hamamatsu University School of Medicine, 1–20-1 Handayama, Hamamatsu, Shizuoka 431-3192, Japan. E-mail: sasakis{at}hama-med.ac.jp.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
T1, revealed that T3 treatment recruited histone deacetylase 3, reduced the acetylation of histone H4, and caused the dissociation of TRAP220 within 15–30 min. The reduction of histone H4 acetylation was transient, whereas the dissociation of TRAP220 persisted for a longer period. In the negative regulation of the TSHß gene by T3-TR we report that 1) GATA2 is the major transcriptional activator of the TSHß gene, 2) the putative NRE previously reported is not required, 3) TR-DNA-binding domain directly interacts with the Zn finger region of GATA2, and 4) histone deacetylation and TRAP220 dissociation are important. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
- and ß-chains. The ß-chain (TSHß) is specific to TSH, whereas the -chain (TSH) is common to all glycoprotein hormones. Transcription for both TSH and ß genes is known to be repressed by thyroid hormone (T3) in thyrotrophs (1). The effect of T3 is mainly mediated through thyroid hormone receptors (TRs), which are encoded by TR and -ß genes. The TR locus generates mainly TRl and -2 through alternative splicing, whereas different promoters in the TRß locus generate TRßl and ß2. In patients with resistance to thyroid hormone who exhibit the syndrome of the inappropriate secretion of TSH (2), abnormalities have been identified exclusively in the TRß, not the TR, gene. In agreement, mice lacking the TRß gene display inappropriate secretion of TSH (3), whereas no apparent alteration of TSHß expression is detected in TR null mice (4), and mice having neither TR nor TRß show dramatic overexpression of the TSHß gene (5). We recently reported that TRß2 is the major TR isoform expressed in TT1, a thyrotroph cell line (6, 7). This suggests that TRß2 plays a central role in the T3-dependent negative regulation of TSH genes. Moreover, TRß2 null mice were reported to exhibit central resistance to thyroid hormone not only in the pituitary but also in the hypothalamus (8, 9). For transcriptional activation by T3 in positive regulation, TR heterodimerizes with the retinoid X receptor (RXR) on the positive T3-responsive element (positive TRE) (10). Unliganded TR recruits corepressors including the nuclear receptor corepressor (NCoR) and silencing mediator of retinoic acid and thyroid hormone receptors (SMRT), resulting in the association of histone deacetylases (HDACs) to repress the transcription (silencing). In contrast, T3-bound TR (T3-TR) interacts with coactivators including p160 family proteins and p300/cAMP response element-binding protein (CREB)-binding protein (CBP) through their LXXLL motifs (L is leucine and X is any amino acid). For this interaction, the activation function-2 (AF-2) domain in the C-terminal region of TR plays a pivotal role. These coactivators stimulate transcription through their histone acetyltransferase (HAT) activity (11). T3-TR also recruits TR-associated protein 220 (TRAP220), a component of TRAP/SRB/MED-containing cofactor (TRAP/SMCC) complex that modulates the function of the carboxyl-terminal domain (CTD) in RNA polymerase II (RNA pol-II) (12). Chromatin immunoprecipitation (ChIP) assays of positively regulated genes indicated that TRAP220 may be recruited after the transient association of p160 family proteins and p300 (13). In estrogen (E2)-induced transactivation, the recruitment of TRAP220 and steroid receptor coactivator 1 (SRC-1), a p160 family coactivator, is reciprocal (14), and TRAP220 may play a role in the cyclic entry of the preinitiation complex (15, 16).
The molecular mechanism underlying T3-dependent negative regulation has been unclear. The short DNA sequence immediately downstream of the transcription start site of the TSHß gene, which contains a single half-site-like sequence (GGGTCA), has been postulated to mediate inhibition by T3-TR (1, 17, 18, 19) and is referred to as negative TRE (nTRE) (1, 20) or a negative regulatory element (NRE) (21). Using in vitro experiments including gel shift assays, several studies have demonstrated the interaction between TR and putative NRE (20, 21, 22, 23, 24, 25). Nevertheless, if TR-RXR heterodimer binds with NRE, it is difficult to explain why T3-TR on DNA does not recruit coactivators to stimulate transcription, and why unliganded TR does not interact with corepressors to cause silencing. To address such a discrepancy, we postulated the involvement of a thyrotroph-specific cofactor that might switch T3-TR from a transcriptional activator to repressor and proposed a model in which the TR monomer-HDAC2 complex may be recruited to NRE in a T3-dependent fashion (21). In vivo footprinting, however, failed to demonstrate direct contact between TR and the X region (GGGTCA) (21).
The DNA sequence between –271 and –80 bp in the mouse TSHß promoter (corresponding to the sequence between –269 and –78 bp in the human TSHß gene) was reported to be sufficient for maximal promoter activity in thyrotroph (26). Gordon et al. (27) reported that the expression of the TSHß gene requires not only a pituitary-specific transcription factor, Pit1, but also the hematopoietic transcription factor, GATA2. GATA2 exhibits drastic cooperative function with Pit1 even in kidney-derived CV1 cells through the two GATA-responsive elements (GATA-REs) close to the Pit1-binding site (28). Using in vivo methods, Dasen et al. (29) demonstrated that Pit1 and GATA2 direct the differentiation from pituitary precursor cells to the thyrotroph lineage. In our previous study, we reported that negative regulation of the TSHß promoter is easily detected even in CV1 cells when TR is coexpressed with Pit1 and GATA2 (6). This observation suggested that thyrotroph-specific factors other than Pit1 and GATA2 may not be essential to mediate T3-TR-dependent negative regulation of the TSHß gene. Receptor-ligand specificity was observed as in the positive regulation of T3-target genes (6). Deletion analyses of TRß in CV1 cells revealed that the DNA-binding domain (DBD) of TR is indispensable for T3-dependent negative regulation, consistent with the findings of in vivo analysis (30) and also the study of the TSH gene (31). On the other hand, mutant TRß2, E457A, which has a glutamic acid to alanine substitution in AF-2, fails to interact with coactivators. Recently, Ortiga-Carvalho et al. (32) reported that TSH secretion was elevated in homozygotic mice with E457A mutation. Thus, T3-dependent inhibition may require an intact DBD and AF-2 domain.
The reporter assay system using CV1 cells provides an ideal experimental platform with which to study the difference between positive and negative regulation by T3, because the CV1 cell line is one of the cultured cells most frequently used for the study of positive regulation. Using CV1 cells cotransfected with GATA2 and Pit1, we have reevaluated the function of putative NRE in the TSHß promoter and found that negative regulation by T3-TR was preserved after complete destruction of NRE. The TSHß-D4-chloramphenicol acetyltransferase (CAT) construct containing only the binding sites for Pit1 and GATA2 was stimulated by GATA2 alone without Pit1, and this transactivation was specifically inhibited by T3-TR. We found that GATA2 associates with TR in vitro and in vivo through direct interaction between the Zn finger domain of GATA2 (GATA2-Zf) and TR-DBD. Finally, histone deacetylation and dissociation of a TRAP/SMCC complex were found to be important in inhibition by T3-TR.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). In the reporter assay using CV1 cells, negative regulation by T3-TR was observed using these constructs (Fig. 1B). Because the X region (GGGTCA) contains the reported transcription start site of the human TSHß gene (33), NRE deletion might affect the stability of mRNA. To exclude this possibility, we also introduced various mutations in NRE to create mX, mY, mZ, mXZ, mYZ, and mXYZ (Fig. 1A). Again, we found the T3-dependent inhibition of these mutant reporters (Fig. 1, C and D), suggesting that NRE is not essential for the inhibition of the TSHß promoter by T3-TR. To identify the region responsible for inhibition, we made a series of deletion mutants, including TSHß-D3-CAT, D4-CAT, and D5-CAT (Fig. 1E). As shown in Fig. 1F, both TSHß-D3-CAT and D4-CAT were stimulated by Pit1 and GATA2, and suppressed by T3-TR. The basal transactivation level of TSHß-D4-CAT in the presence of Pit1 and GATA2 was much higher than that of the wild-type promoter. This indicates that the sequence between –81 bp and –29 bp may have some inhibitory function. Notably, TSHß-D4-CAT possesses only the Pit1-binding site and GATA-REs, but no other sequence similar to the reported TRE-half-site (AGGTCA). Thus, we speculated that the interaction of TR with Pit1 or GATA2 might play a key role in T3-dependent inhibition. Transactivation was abolished for TSHß-D5-CAT, suggesting that the downstream GATA-RE is necessary. In addition, this excludes the possibility of a cryptic GATA-RE in the plasmid backbone.
|
, lanes 5–10). Interestingly, however, we found that GATA2 alone without Pit1 could transactivate TSHß-D4-CAT (Fig. 2B, lanes 5–7). The magnitude adjusted by control cytomegalovirus (CMV)-CAT was approximately 4-fold greater than that of the wild-type promoter activated by Pit1 and GATA2 (compare lanes 8–10 in Fig. 2A with lanes 5–7 in Fig. 2B). Transactivation of TSHß-D4-CAT by GATA2 was specifically suppressed by T3-TRß2 (Fig. 2C) and TRß1 (data not shown), but not by vitamin D/vitamin D receptor (VDR) or retinoic acid /retinoic acid receptor (RAR) (Fig. 2, C and D).
|
, E and F, these reporter genes were activated by the coexpression of GATA2 and repressed by T3. These results suggest the essential role of GATA2 to transactivate the TSHß promoter and to mediate negative regulation by T3-TR.
GATA2-Zf Plays an Important Role in Negative Regulation of the TSHß Gene by T3-TR
To test whether T3-TR suppresses the activation of other GATA-RE, we generated (GATA-RE)2-tk-CAT containing two copies of GATA-RE derived from the CD34 gene (36) fused to the thymidine kinase (tk) promoter. As expected, (GATA-RE)2-tk-CAT was stimulated by GATA2 and negatively regulated by T3-TR (Fig. 3A). Among the six members of the GATA family, GATA1, -2, and -3 play pivotal roles in the transcription of hematopoiesis-related genes (37). The central Zn finger domain is known to recognize GATA-RE and is highly conserved among GATA family members, whereas the homology in N- and C-terminal regions is low (Fig. 3B). We found that the transcription of (GATA-RE)2-tk-CAT was also activated by mouse GATA1 and -3, and this activation was suppressed by T3-TR (Fig. 3A). Similarly, transactivation of TSHß-D4-CAT was also stimulated by GATA1 and -3, and repressed by T3-TRß2 (Fig. 3C). These findings suggest that the function of the Zn-finger domain conserved among GATA1, -2, and -3 may be important not only for the recognition of GATA-REs but also for negative regulation. To confirm this, we generated three expression plasmids for GATA2 mutants, i.e. GATA2-NZ, -Zf and -ZC, in which the N- and/or C-terminal region was truncated (Fig. 3D, middle panel). TSHß-D4-CAT was stimulated by both GATA2-NZ and -ZC (Fig. 3D, left panel) and suppressed by T3-TR (Fig. 3D, right panel). Although the fold repression appeared smaller in NZ and ZC than wild-type GATA2, these differences were not statistically significant. These data suggested the importance of the Zf region common to GATA2-NZ and -ZC. It was difficult to evaluate the effect of GATA2-Zf itself because of its low transactivation activity.
|
, left panel). It should be noted that T3-TRß2-dependent inhibition of the transactivation by C295A was impaired and that the reduction in the presence of T3 was not statistically significant (Fig. 3D, right panel). These results suggested that T3-TR targets GATA2-Zf.
The DBD of steroidogenic factor 1/adrenal 4 binding protein (SF-1/Ad4BP) has been reported to interact with GATA4 (38). A mutant SF-1/Ad4BP with an amino acid substitution from glycine to glutamic acid at codon 35 (G35E) in the P-box fails to synergize with GATA4 although this mutant preserves the direct physical binding to GATA4 (39). The corresponding mutation from glycine to glutamic acid at codon 182 (G182E) in the P-box of TRß2 failed to repress the transactivation of TSHß-D4-CAT by GATA2, whereas another mutation (glycine to alanine, G182A) did not affect the inhibitory effect (Fig. 3E). Identical results were obtained when corresponding mutations in TRß1s, G129E and G129A, were used (data not shown). Collectively, the Zn finger domain of GATA2 and DBD of TRßs is important to mediate T3-dependent repression.
GATA2-Zf Interacts with TR-DBD in Vitro and in Vivo
As illustrated in Fig. 4, A and B, full-length GATA2 (FL) and truncated mutant GATA2s (N and Zf) were produced in Escherichia coli as glutathione-S-transferase (GST) fusion protein (Fig. 4B). As shown in Fig. 4C, GST-FL, -N, and -Zf, but not GST alone, interacted with TRß2 in a T3-independent fashion (Fig. 4C). These results indicated that GATA2-Zf interacts with TRß2. Next, to confirm the in vivo association between TR and GATA2, the expression plasmids for myc-tagged GATA2 and FLAG-tagged TRß2 were transfected into 293T cells. Whole-cell extracts were immunoprecipitated using anti-FLAG antibody and analyzed by Western blotting with anti-myc antibody. As shown in Fig. 4D, myc-tagged GATA2 was coimmunoprecipitated with anti-FLAG antibody in a T3-independent fashion. Similar results were obtained in reciprocal experiments, where FLAG-tagged GATA2 and wild-type TRß2 were coexpressed in 293T cells and the whole-cell extract was precipitated with anti-FLAG antibody and analyzed by Western blotting with anti-TRß2 antibody (Fig. 4E). To map the interacting region of TR, we generated TR deletion mutants including C1, C2, L178X, G251X, and V336X (Fig. 5A). GST-Zf interacted with every TR mutant if it contained DBD, but not with C2 in which DBD was deleted (Fig. 5B). These experiments demonstrated that TR-DBD functions as the interface to bind GATA2-Zf. Although mutant SF-1/Ad4BP, G35E, reportedly failed to synergize with GATA4, it can physically interact with GATA4 (39). Similarly, corresponding TRß2 mutants, G182E, bound GST-Zf with comparable affinity for wild-type TRß2 and G182A (Fig. 5C). Identical results were also observed when we employed corresponding mutant TRß1s, G129E and G129A (data not shown). Thus, the property of the interaction between TRß2 and GATA2 was similar to that between SF-1/Ad4BP and GATA4.
|
|
, A and D), we postulated that T3 binding to TR might interfere with some cofactor(s) for GATA2. GATA2 interacts with CBP/p300 (40, 41) and HDAC3 (42), whereas TR constitutively binds with HDAC2 (21), and NCoR/SMRT has been known to form a complex with HDAC3 (43). We examined the effects of the overexpression of various cofactors including CBP, p300, HDAC2 and -3, NCoR, SMRT, ligand-dependent corepressor, and receptor-interacting protein 140 on suppression by T3-TR; however, no significant alteration was observed in our reporter assay (data not shown).
TRAP220 (Fig. 6A) has been reported to function as a coactivator for liganded nuclear receptors, including TR (44, 45), VDR (46), estrogen receptor (ER) (15), glucocorticoid receptor (GR) (47), and peroxisomal proliferator-activated receptor (48). Interestingly, TRAP220 also interacts with GATA2 as well as Pit1 (49), resulting in stimulation of the expression of the TSHß gene (49, 50). Whereas TRAP220 does not have histone acetyltransferase (HAT) activity, it is a constituent of the TRAP/SMCC complex that regulates the function of CTD in RNA pol-II (12). To study the effect of TRAP220, we transiently transfected full-length TRAP220 together with TSHß-D4-CAT and GATA2 into CV1 cells. In accordance with the previous report with GATA2 (49) and GATA3 (51), wild-type TRAP220 significantly stimulated GATA2-mediated transactivation (Fig. 6B). The enhancement was modest probably due to the presence of endogenous TRAP220, which is known to be ubiquitously expressed at various levels (45). The GST pull-down assay showed direct interaction of GATA2-Zf with the N-terminal region of TRAP220 (Fig. 6C), consistent with previous reports (49, 51). Truncated TRAP220 with both N- and C-terminal deletions [dominant-negative TRAP220 (dnTRAP220)] that retains two LXXLL motifs has been known to exhibit a dominant-negative effect on positive regulation by T3-TR (45). As shown in Fig. 6C (right panel), dnTRAP220 did not bind with GATA2-Zf. This result is in agreement with the recent report from Gordon et al. (49) and suggests that the interaction surfaces of TRAP220 for GATA2 and TRß2 are different. As shown in Fig. 6D, overexpression of dnTRAP220 did not impair the transactivation by GATA2 but relieved the inhibition by T3-TR. In contrast, the truncated SRC-1 (amino acids 1–771) or p300 (amino acids 1–670), both of which possess a nuclear localization signal and all LXXLL motifs but no HAT domain, did not exhibit a significant effect on T3-dependent inhibition (data not shown). These results suggest that TRAP220 plays an important role in the T3-TR-dependent negative regulation of the TSHß gene.
|
T1 cells (7) were cross-linked and immunoprecipitated with the specific antibody against GATA2, followed by PCR amplification with the specific primers that encompass a 150-bp sequence in the promoter region (Fig. 7A). As shown in Fig. 7B (upper panel) and Fig. 7C, we observed the enrichment of the TSHß promoter with ChIP assays using anti-GATA2 antibody over the signal derived from control IgG. There was no enrichment of the same DNA on the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) gene (Fig. 7B, lower panel, and Fig. 7D). These results were in agreement with a previous report (49). We failed to detect endogenous TRß2 in TT1 cells (7) using ChIP assay probably due to the problem with the antibody against TRß2. To confirm the association of TRß2, we used a DNA pull-down assay (21). As illustrated in Fig. 7E, nuclear extract from 293T cells transfected with Pit1, GATA2, and TRß2 was incubated with three copies of Pit1- and GATA2-binding sites conjugated to magnetic beads. The DNA-bound complex was washed and eluted from beads. The proteins in the complex were analyzed using Western blotting using specific antibodies. As expected, not only Pit1 and GATA2 but also TRß2 were detected in a T3-independent manner (Fig. 7F). These bindings were specific because they were abolished by a 50-fold excess amount of free DNA that harbored three copies of Pit1- and GATA2-binding sites.
|
, G and H). TRAP220 was found to dissociate during the same time period (Fig. 7I). Whereas the reduction of histone H4 acetylation was substantially attenuated at 2 h, the dissociation of TRAP220 persisted for 16 h. We detected no significant alteration of HDAC2 association (data not shown). These findings indicate that histone H4 deacetylation and dissociation of TRAP220 may be important in the negative regulation of the TSHß promoter in vivo. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
plus ß genes (5). A similar finding was obtained in an resistance to thyroid hormone subject with a large-scale deletion in the TRß gene (52). We have developed a highly sensitive reporter assay system using nonpituitary CV1 cells coexpressed with Pit1 and GATA2, which is suitable to study the negative regulation of the TSHß gene (6). When Pit1, GATA2, and TR were coexpressed in CV1 cells, the basal level of the TSHß gene expression was remarkably elevated and was suppressed by T3. Without Pit1 and GATA2, unliganded TR alone did not transactivate the TSHß promoter at all (6). These findings provide compelling evidence that unliganded TR per se is not a transcriptional activator of the TSHß gene.
Mutation in the Pit1 gene in human subjects causes combined pituitary hormone deficiency, including TSH deficiency (53), suggesting the important contribution of Pit1 to TSH production in vivo; however, the mechanism has not been clarified as to how Pit1 works for the TSHß gene activation (27, 29). We found that the TSHß-D4-CAT construct, which possesses the Pit1-binding site and GATA-REs only, was stimulated by GATA2 alone without Pit1 (Fig. 2B), although GATA2 requires the coexistence of Pit1 for transactivation of the wild-type promoter. Although the deleted sequence (nucleotide –81/–29) contains the potential Pit1 binding site reported previously (54), overexpression of Pit1 did not affect the activity of the heterologous tk-promoter fused to this region (data not shown). Recently, we obtained the finding that there is an as yet unidentified inhibitor against GATA2-driven transactivation that acts on the sequence (nucleotide –81/–54) (Fig. 1C). We have found that the suppressive function of this sequence is not dependent on the orientation of the sequence or the distance from a TATA box, and that Pit1 is considered to protect GATA2 from this inhibitor (Kashiwabara, Y., S. Sasaki, A. Matsushita, K. Nagayama, K. Ohba, H. Iwaki, H. Matsunaga, H. Misawa, K. Ishizuka, and H. Nakamura, manuscript submitted).
The Putative NRE Is Not Required for Negative Regulation of the TSHß Gene, and TR-DBD Interacts with GATA2-Zf in a T3-Independent Manner
There have been many studies that characterized the negative regulation of the TSHß gene, but most did not consider the endogenous activator of this gene (18, 19, 55, 56, 57). NRE in the TSHß gene was originally defined by the observation that its deletion abolished the basal transcriptional activity of the TSHß promoter (19); however, the interpretation of this study was based on the hypothesis that unliganded TR was a transcriptional activator of the TSHß gene, which is not the case. In addition, this study was performed using human kidney-derived 293 cells that lack Pit1 (58) and GATA2 (59). Although the somatotroph cell line GH3, which expresses TRH receptor, TRß1, TRß2, and Pit1, has often been used (17, 18, 20, 21), this cell line does not express endogenous GATA2 (27). Indeed, cotransfection of GATA2 activated the TSHß promoter and enabled the detection of the inhibition by T3 (Fig. 3, E and F). Other studies used the JEG3 choriocarcinoma cell line (20, 24, 31, 60), but no attention was paid to the interaction between TR and endogenous GATA2 and 3 in this cell line (61).
Another experimental problem in previous studies concerns the reporter constructs used for assays. To gain high sensitivity, firefly luciferase-based reporter plasmids have been employed (20, 56, 57, 62). The possibility has been pointed out that firefly luciferase cDNA itself is subject to T3-dependent inhibition (34, 35). We have recently found that conventional firefly-luciferase cDNA, but not modified Renilla-luciferase cDNA (hRluc, Promega), harbors a sequence that is stimulated by phorbol 12-0-tetradecanoate-13-acetate (TPA) and repressed by T3-TR (Misawa, H., S. Sasaki, A. Matsushita, K. Nagayama, K. Ohba, H. Iwaki, H. Matsunaga, S. Suzuki, K. Ishizuka, H. Nakamura, manuscript in preparation). Some studies used parental reporter plasmids containing a pUC-derived AP-1 site (18, 19), although Jun/Fos-related transcription factors are known to mediate T3-dependent inhibition (63, 64). In our TSHß-CAT, the artificial AP-1 site in the plasmid backbone was deleted (21), and we confirmed that overexpression of Jun and Fos did not affect the transcriptional activity of TSHß-CAT (data not shown). As mentioned above, the definition of nTRE/NRE should be reevaluated, and we have demonstrated that, at least in CV1 cells, the reported nTRE is not essential for the T3-dependent inhibition. However, in vivo function of nTRE remains to be determined. Using degenerated PCR primers and cDNA from TT1 cells, we recently identified a high-mobility group protein, Sox 11, as a binding factor to the Z region in NRE (Fig. 1A). Unexpectedly, Sox11 relieved T3-TR dependent inhibition of the TSHß promoter in CV1 cells (data not shown); therefore, a transgenic mouse in which nTRE is disrupted should be examined in the future.
T3-TR inhibited the GATA2-dependent transactivation of TSHß-D4-CAT, which possesses only the Pit1-binding site and GATA-REs, but no sequence similar to the TRE-half-site. CD34-derived GATA-RE activated by GATA2 was also suppressed by T3-TR (Fig. 3A). A functional GATA-RE exists in the endothelin-1 promoter (65, 66), the transcription of which was enhanced by GATA2 and repressed by T3-TR (data not shown). The TSH promoter, which also contains functional GATA-RE (61), is stimulated by transcription factors including Ptx1, Lhx3a, CREB, and GATA2, but only the transactivation driven by GATA2 is inhibited by T3-TR (Nakano, K., S. Sasaki, A. Matsushita, K. Nagayama, K. Ohba, H. Iwaki, H. Matsunaga, S. Suzuki, H. Misawa, K. Ishizuka, and H. Nakamura, manuscript in preparation). Therefore, transactivation by GATA2 is indispensable for negative regulation by T3-TR. The finding that the transactivation by a mutant GATA2, C295A, is resistant to T3-TR-mediated suppression (Fig. 3D) supports the notion that cross-talk between TR and GATA2 is a prerequisite for T3-dependent inhibition. Recently, pituitary-specific knockout of the GATA2 gene was reported (67). Although GATA2 is important for the enhanced TSHß expression in hypothyroidism, it appears to be dispensable for thyrotroph cell fate. This phenomenon may be explained, in part, by the finding that these mice were reported to exhibit an elevated level of GATA3 (67), the function of which can be inhibited by T3-TRß2 (Fig. 3C).
The DBD of TR is known to play essential roles in the negative regulation of TSH (31) and -ß genes (6, 30). Although TR-DBD directly recognizes positive TRE, the function of the DBD of nuclear receptors is not limited to DNA binding. For example, GR interacts with nuclear factor (NF)-
B through its DBD and induces the glucocorticoid-dependent inhibition of NF-B-mediated transactivation (68, 69). DBDs of GR and RAR are necessary for interaction with AP-1 and the ligand-dependent inhibition of AP-1-induced transactivation (70, 71). Transcription of the proopiomelanocortin gene is stimulated by an orphan receptor, Nur77, and this transactivation is repressed by glucocorticoid-bound GR through interaction between GR-DBD and Nur77-DBD (72).
Cross-talk between nuclear hormone receptors and GATA family transcription factors has been reported. Ligand-bound GR is known to repress the transactivation by GATA1 (73). ER also interacts with GATA1 (74), and we recently found that E2 bound ER modestly but significantly repressed GATA1-dependent transactivation in CV1 cells (data not shown). Tremblay and Viger (38) reported the interaction between SF-1/Ad4BP and GATA4. They demonstrated that a mutation in the P-box of SF-1/Ad4BP (G35E) abolished the synergism with GATA4 although physical interaction is maintained (39). Similarly, the corresponding mutation in TRß2 (G182E) lost T3-dependent inhibition (Fig. 3, E and F) whereas it could interact with GATA2 in vitro (Fig. 5C). In the future, it should be examined whether the P-box in TR-DBD may be one of the determinants for receptor specificity as in the case with positive TRE.
The Mechanism by Which TR-GATA2 Complex Mediates T3 Signaling in the Negative Regulation of the TSHß Gene
Several mechanisms are considered to explain how the TR-GATA2 complex sensitizes T3 signaling. The first is that T3-TR may recruit an HDAC-related factor to the TR-GATA2 complex. Previously, it was reported that NCoR and SMRT enhanced the basal transcriptional activity of TSH and -ß promoters in a firefly luciferase-based reporter assay (56). We reported the association of HDAC2 with TR in vitro (21). In our present study, overexpression of NCoR, SMRT, or HDAC2 had no effect on the suppression of the TSHß gene by T3-TR (data not shown). Ligand-dependent corepressors such as ligand-dependent corepressor (75) and receptor-interacting protein 140 (76) also exhibited no effects (data not shown). HDAC3 interacts with GATA2 (42) and forms a complex with NCoR/SMRT (43), but its overexpression did not affect the suppression of TSHß by T3-TR (data not shown). FOG1 and 2 interacted with N-terminal Zn fingers to repress GATA-dependent transcription (77). FOG2, which plays a role in the function of the GATA family in nonerythroid tissues (78), interacts with nuclear hormone receptors including RXR (79) and chicken ovalbumin upstream promoter transcription factor (80). Unfortunately, we could not examine the effect of FOG2 on the inhibition by T3-TR, because the basal activation of TSH ß-D4-CAT by GATA2 was completely suppressed by FOG2 (data not shown). The N-terminal Zn finger mutant of GATA2, C295A, is known to resist the FOG1 function (81). Because C295A is also resistant to the suppressive function by T3-TR (Fig. 3D), a common mechanism may exist in suppression by T3-TR and FOGs.
The second possibility is that T3-TR may interfere with the HAT activity required for transactivation by GATA2. GATA2 has been reported to associate with CBP (40) and p300 (41). p300 was reported to acetylate GATA1 protein, resulting in enhanced DNA binding affinity and transactivating function (82). Other investigators, however, suggested that histone acetylation by p300 is not the main determinant of GATA2-dependent transactivation (41). In our reporter assay, overexpression of CBP or p300 did not exert significant effects on the negative regulation of the TSHß gene (data not shown). SRC-1, the disruption of which causes the elevation of TSHß expression in vivo (83), is an important coactivator of TR in the positive regulation of the T3-target genes. We observed, however, that SRC-1 overexpression did not affect the negative regulation of the TSHß promoter under our experimental conditions (data not shown).
Finally, TRAP220 may mediate T3 signaling in negative regulation. TRAP220 is a non-HAT coactivator that physically interacts with T3-TR through its LXXLL motif (45). Notably, it functions as a coactivator for GATA family transcription factors (51) as well as Pit1 (49), and the phenotype of a TRAP220-null mouse is similar to that of a GATA family member-null mouse (51). Interestingly and importantly, the heterozygote of the TRAP220-null mouse exhibits reduced expression of the TSHß gene (50). This is consistent with the recent observation reported by Gordon et al. (49) and our findings that overexpression of wild-type TRAP220 modestly but significantly enhanced the transactivation of TSHß-D4-CAT by GATA2 (Fig. 6B). Our GST pull-down assay showed the interaction of GATA2-Zf with the N-terminal region of TRAP220, which is different from LXXLL motifs (Fig. 6C). On the other hand, dominant-negative type TRAP220 (dnTRAP220) containing two LXXLL motifs, which causes inhibition of positive regulation by T3-TR (45), abolished the negative regulation of TSHß-D4-CAT (Fig. 6D). It is unlikely that dnTRAP220 competes with other coactivator proteins, because the truncated SRC-1 or p300, which contains LXXLL motifs but lacks the HAT domain, did not exhibit such an effect (data not shown). It has been reported that E2-bound ER specifically distinguishes the LXXLL motif of TRAP220 from that of p160 (15). These findings suggested an important role of TRAP220 in T3-TR induced suppression.
Whereas liganded RAR and VDR interact with TRAP220 (12), only the T3-TR-dependent pathway was disturbed in TRAP220-deficient fibroblasts (84). This may explain why TR, but not RAR or VDR, mediates ligand-dependent repression (Fig. 2D), although GATA2-Zf interacts with these two receptors as well as TR (Ref. 85 and data not shown). In the inhibition of the TSHß promoter by T3, the interaction of TR with both GATA2 and TRAP220 may direct ligand/receptor specificity. It is unknown why RA-bound RAR enhances GATA2-induced transactivation (Fig. 2D), although this is in agreement with the previous report (85).
Both TRAP220 and the Alteration of Histone Acetylation Play Essential Roles in the Negative Regulation of the TSHß Gene
In the study of T3-dependent transactivation, Sharma and Fondell (13) reported that the recruitment of p160 and p300 occurred transiently for 30–60 min, whereas acetylation of histone H4 was maintained for 4 h, and that TRAP220 was subsequently recruited to positive TRE after the dissociation of p160 and p300. In our ChIP assay of the TSHß gene, the association of HDAC3, reduction of histone H4 acetylation, and dissociation of TRAP220 were detected within 15–30 min (Fig. 7, G–I). The duration of histone H4 deacetylation was transient (2–4 h), whereas that of TRAP220 dissociation was maintained for 16 h. The short duration of histone H4 deacetylation may explain why we failed to detect the effect of HDAC3 overexpression on T3-dependent inhibition (data not shown).
The mechanism of T3-TR antagonism against GATA2 appears to be similar to that of the antagonism of liganded GR against NF-B. First, cross-talk occurs through the protein-protein interaction between DNA-binding transcription factors and nuclear hormone receptors. Second, the reduction of histone H4 acetylation was also observed when liganded GR inhibited the NF-B-induced transactivation of granulocyte-macrophage colony-stimulating factor promoter (86). Third, transactivation by NF-B is mediated by a protein complex almost identical to TRAP/SMCC (87). Fourth, TRAP220 interacts with liganded TR as well as GR (47). Finally, Nissen and Yamamoto (88) reported that, when the promoters of IL-8 and intercellular adhesion molecule-1 were transactivated by NF-B, liganded GR interfered with the phosphorylation of CTD of RNA pol-II, the function of which is controlled by TRAP/SMCC complex (12). This observation supports the possibility that liganded GR targets not only NF-B but also TRAP/SMCC complex. As the TRAP/SMCC-related complex lacking TRAP220 is known to inhibit transcription (89), we speculate that liganded TR and GR may reduce the stability of the interaction of TRAP/SMCC complex with GATA2 or NF-B, respectively, resulting in the dissociation of TRAP220 from the complex. The ChIP assay in positive regulation showed that histone acetylation is required for association of the TRAP/SMCC complex (13). In T3-dependent inhibition, the transient reduction of histone acetylation (Fig. 7) may be required for the dissociation of TRAP220. Currently, we speculate that, on the TSHß promoter, Pit1-GATA2 complex associates with TRAP/SMCC complex as well as TRß2 (Fig. 8). Upon T3 binding, TRß2 modifies the function of TRAP220, resulting in destabilization of the interaction between TRAP220 and Pit1-GATA2-complex. Simultaneously, T3 binding to TRß2 temporally induces the recruitment of HDAC3 or dissociation of HAT-related molecules.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). SwaI and BglII were introduced in the N- and C-terminal regions of the CAT coding region in TSH ß-CAT and TSH ß-D4-CAT using site-directed mutagenesis. Then, this region was replaced with SwaI and BglII fragments of Renilla luciferase cDNA (hRluc) from phRluc-null vector (Promega Corp., Madison, WI), generating TSHß-hRluc and TSHß-D4-hRluc. We confirmed that Renilla luciferase cDNA (hRluc) in phRluc-null vector does not mediate artificial T3-depend repression (data not shown). The two copies of the back-to-back double GATA-REs derived from the mouse CD34 promoter fused to the thymidine kinase promoter in (GATA-RE)2-tk-Luc (a gift from Dr. Towatari, Nagoya University, Nagoya, Japan) were amplified by PCR. The PCR product was subcloned into the NheI site in the pCAT3 basic vector (Promega Corp.) to create (GATA-RE) 2-tk-CAT. The expression vectors that contain cDNA for wild-type human TRß1 (pCMX-hTRß1), rat TRß2 (pCMX-rTRß2), and human TRß deletion mutants (pCI-C1 and pCI-C2) were described previously (6, 21). Using site-directed mutagenesis, glycine at codon 129 in pCMX-hTRß1 was substituted to glutamic acid (G182E) or alanine (G182A), and stop codons were introduced at codon 178 (leucine), 251 (glycine), or 336 (valine) in pCMX-hTRß1 to generate pCMX-hTRß1-L178X, G251X, and V336X, respectively. The expression plasmids for human Pit1 (pCB-hPit1) and mouse GATA2 (pcDNA3-mGATA2) have been described elsewhere (6). Using pcDNA3-mGATA2 as a template, PCR was performed with primer mGATA2U; tatctcgaggtaccatggaggtggcgcctgagcagcctcgctgga and primer mGATA2D2; gcccctcgagtctagactacttccgattccgggtctggat for GATA2-NZ, with primer mGATA2U2; acccgaattcaagcttggtaccatggctcgctcctgctcagaaggccg and primer mGATA2D; tatctcgagctagcccatggcagtcaccat for GATA2-ZC, and with primer mGATA2U2; and primer mGATA2D2 for GATA2-Zf. The amplified fragments were digested with KpnI and XhoI, followed by subcloning into pcDNA3 to create pcDNA3-mGATA2-NZ, -ZC, and -Zf. The Zn finger domain of mouse GATA2 (codon 280–400) was amplified by PCR with primer mGATA2U2 and mGATA2D2. The amplified fragment was subcloned into the EcoRI/XhoI site in the pGEX-4T-1 vector (Amersham Biosciences, Piscataway, NJ) to generate pGEX-4T-1-GATA2-Zf (plasmid for GST-Zf). In pcDNA3-mGATA2, cysteine at codon 295 and codon 349 was substituted to alanine to generate C295A and C349A, respectively. All constructs were confirmed by DNA sequencing.
Cell Culture and Transient Transfection
Monkey kidney cell line CV1 was grown in monolayer culture at 37 C under 5% CO2/95% air in DMEM containing 10% fetal calf serum, penicillin G (100 units/ml), and streptomycin (100 µg/ml). CV1 cells were trypsinized and plated in 60-mm dishes for 24 h before transient transfection using the calcium-phosphate technique (6). Cells at a density of 106 cells per plate were transfected with the TSHß-CAT reporter gene (3.6 µg) along with pCMX-rTRß2 (0.4 µg), ß-galactosidase expression vector pCH111 (1.8 µg), pCB-hPit1 (0.2 µg), pcDNA3-mGATA2 (0.4 µg), and empty vector (pCMX) as carrier DNA to adjust the total DNA to 7.2 µg/dish (6). After cells were exposed to calcium phosphate/DNA precipitates for 20 h, the medium was replaced with fresh DMEM containing 5% fetal calf serum depleted of thyroid hormone (91) or the same medium supplemented with 1 µM T3. After incubation for an additional 24 h, cells were harvested and CAT activity was measured and normalized by ß-galactosidase activity as described previously (6). The rat somatotroph cell line, GH3 cell, was kept in DMEM supplemented with 10% calf bovine serum, and transfection was performed as described previously (21). The thyrotroph cell line TT1 was a kind gift from Dr. P. L. Mellon (University of California, San Diego, CA) (7) and described elsewhere (6).
GST Pull-Down Assay
E. coli (DH5) that had been transformed with pGEX-4T-1-GATA2-Zf were induced with 0.1 mM isopropyl-1-thio-ß-galactopyranoside for 4 h. The E. coli pellet was sonicated and the fusion proteins were mixed with glutathione-Sepharose beads (Amersham Pharmacia Biotech, Uppsala, Sweden) for purification. Receptor proteins were translated in vitro using rabbit reticulocyte lysate (Promega Corp.) in the presence of [35S]methionine. Radiolabeled receptors were incubated with GST fusion proteins in binding buffer [150 mM NaCl, 20 mM Tris HCl (pH 7.5), 0.3% Nonidet P-40, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 2 µg/ml leupeptin, 2 µg/ml aprotinin] for 3 h at 4 C, and washed three times with the binding buffer. Bound protein was analyzed by 10–14% SDS-PAGE and visualized using the BAS-1000 autoradiography system (Fuji Film, Tokyo, Japan).
Immunoprecipitation
Expression vectors for FLAG-tagged TRß2 (pCMX-FLAG-ratTRß2) and myc-tagged GATA2 (pCDNA3.1-myc-GATA2) were cotransfected into 293T cells by the calcium-phosphate method. After 24 h incubation in the presence or absence of 1 µM T3, cells were harvested and washed twice with ice-cold PBS. Cell pellets were lysed in hypotonic buffer [20 mM HEPES (pH 7.9), 10 mM KCl, 10% glycerol, 1 µM EDTA, 0.2% Nonidet-P40, 3 µg/ml aprotinin, and leupeptin] and incubated on ice for 15 min. After centrifugation at 14,000 rpm for 5 min, the pellet was resuspended in high-salt buffer [20 mM HEPES (pH 7.9), 420 mM NaCl, 20% glycerol, 1 µM EDTA, 0.2% Nonidet P-40, 3 µg/ml aprotinin, and leupeptin] and gently agitated at 4 C with for 30 min. The supernatants were collected after centrifugation at 14,000 rpm for 10 min and incubated with anti-FLAG M2 affinity gel (Sigma, St. Louis, MO) in binding buffer [150 mM NaCl, 20 mM Tris HCl (pH 7.5), 0.3% Nonidet P-40, 3 µg/ml aprotinin, and leupeptin at 4 C) overnight. After four washes with hypotonic buffer, immunocomplexes were resolved by SDS-PAGE and Western blotted using anti-c-Myc antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA) and analyzed by anti-FLAG M2 antibody (Sigma).
DNA Pull-Down Assay
Using the calcium-phosphate technique, 293T cells cultured in 20 x 10 cm dishes were cotransfected with the expression plasmid for GATA2 (pcDNA3-mGATA2, 5 µg), Pit1 (pCB6+-hPit1, 3 µg), and TRß2 (pcDNA3-rTRß2, 6 µg). Cells were rinsed with PBS and harvested with 5 ml of PBS 24 h after the transfection. After centrifugation at 2000 rpm for 10 min at 4 C, the cell pellets were suspended with one packed cell volume of buffer A [10 mM HEPES (pH 7.9), 1.5 mM MgCl2, 10 mM KCl]. Cells were incubated on ice for 15 min and shared with a 26 G needle five times. The homogenate was centrifuged at 14,000 rpm for 1 min, and nuclei pellets were suspended with one packed cell volume of buffer C [20 mM HEPES (pH 7.9), 25% glycerol, 420 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA]. Elution was performed for 30 min at 4 C and centrifuged 14,000 rpm at 4 C. The samples were dialyzed against buffer D [20 mM HEPES (pH 7.9), 20% glycerol, 0.2 EDTA 0.5 mM dithiothreitol, 42 mM (NH4)2SO4] at 4 C for 2 h. After centrifugation at 14,000 rpm at 4 C, the supernatant was stored in liquid nitrogen. The basic procedure of the DNA pull-down assay is described elsewhere (21). The pUC19-based vector containing three copies of Pit1 and GATA2 binding sites of human TSHß promoter was subjected to PCR with the biotinylated M13 primer and 3'-specific primer. The PCR products were purified with a PCR purification column (QIAGEN, Chatsworth, CA). Biotinylated PCR fragment (40 µg) was conjugated with 10 mg of M280 magnetic beads (Dynal) in 4 ml of TEN buffer [10 mM Tris HCl (pH 7.5), 1 mM EDTA, 100 mM NaCl] for 1 h at room temperature. The DNA-conjugated beads were precipitated with a magnet and resuspended in 4 ml of TEN buffer and then blocked with 0.5% nonfat milk in binding buffer [20 mM Tris-HCl (pH 7.5), 10% glycerol, 2 mM EDTA, 0.02 Triton X-100, 100 mM KCl] for 1 h at room temperature. The nuclear extract (180 µl) from transfected 293T cells was mixed with the same volume of binding buffer with or without T3 and ultracentrifuged at 40,000 rpm for 30 min. After blocking, the beads were precipitated and resuspended with 100 µl of binding buffer with or without T3. The bead suspensions were mixed with 300 µl of supernatant from the ultracentrifuged nuclear extract and incubated at 4 C for 4 h with rotation. The suspensions were precipitated with a magnet, washed twice with binding buffer, and precipitated. Proteins on beads were eluted with 7 µl of BC500 buffer on ice for 5 min. The elution was analyzed by Western blotting.
ChIP Assay and Real-Time PCR
Approximately 106 TT1 cells were grown in 60-mm dishes. After the addition of T3, the cells were cross-linked by formaldehyde (1% final concentration) for 10 min at room temperature. After cross-linking was terminated by the addition of glycine (0.125 M final concentration), cells were washed twice with ice-cold PBS, and collected by centrifugation. The cell pellet was resuspended in 200 µl of SDS lysis buffer (50 mM Tris-HCl, 10 mM EDTA, 1% SDS, 0.5 mM phenylmethylsulfonyl fluoride, 2 µg/ml leupeptin, 2 µg/ml aprotinin), and incubated for 15 min on ice. Samples were sonicated for 10 sec three times and centrifuged at 14,000 rpm at 4 C. The supernatants were diluted 10-fold with ChIP dilution buffer [50 mM Tris-HCl, 167 mM NaCl, 1.1% Triton X-100, 0.11% sodium deoxycholate (DOC)] supplemented with protease inhibitors. Chromatin solution (2 µl) were precleared with 60 µl of 50% protein G-Sepharose/salmon sperm DNA slurry (Upstate Biotechnology, Lake Placid, NY), and incubated with 4 µl of antiserum against GATA2 (Santa Cruz Biotechnology), acetylated histone H4 (Upstate Biotechnology), HDAC3 (Santa Cruz Biotechnology), and TRAP220 (Santa Cruz Biotechnology) overnight at 4 C. Immunoprecipitated proteins were recovered with 20 µl of 50% protein G-Sepharose/salmon sperm DNA for 2 h and washed with low-salt buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.1% DOC). Pellets were washed with high-salt buffer (50 mM Tris-HCl, 500 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.1% DOC), followed by one wash with LiCl wash solution (10 mM Tris-HCl, 250 mM LiCl, 1 mM EDTA, 0.5% Nonidet P-40, 0.5% DOC), and two washes with 1x Tris-EDTA. Protein-DNA complexes were eluted with the elution buffer (10 mM Tris-HCl, 300 mM NaCl, 5 mM EDTA, 0.5% SDS), and cross-linking was reversed by heating at 65 C for 4 h. DNA was extracted with phenol-chloroform-isoamyl-alcohol (25:24:1) and precipitated with 20 µg of glycogen as a carrier. Samples were dissolved in 20 µl of TE. Using the SYBR Green I kit and a Light Cycler (Roche Diagnostics, Mannheim, Germany), the precipitated DNA was quantified by real-time PCR with primers designed to encompass the 150-bp sequence between the Pit1-binding site and the downstream GATA-RE in the TSHß promoter (forward primer, 5'-taatgtcagagtttccaggg-3'; reverse primer, 5'-ccctctgatcttcttgtt-3') (Fig. 7A). The thermal cycling conditions were 10 min at 95 C, followed by 40 cycles of 10 sec at 95 C for denaturing, 10 sec at 62 C for annealing, and 7 sec at 72 C for extension. PCR signals were analyzed using Light Cycler software version 3.5 (Roche Diagnostics).
Statistical Analysis
The statistical significance of values obtained for CAT assays and ChIP assays was determined by ANOVA and Fisher’s Protected Least Significant Difference test using StatView J-4.5 software (Abacus Concepts, Berkeley, CA).
|
ACKNOWLEDGMENTS |
|---|
T1 cells. |
FOOTNOTES |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online January 23, 2007
Abbreviations: Ad4BP, Adrenal 4-binding protein; AF-2, activation function-2; CBP, CREB-binding protein; ChIP, chromatin immunoprecipitation; CTD, carboxyl-terminal domain; DBD, DNA-binding domain; dnTRAP220, dominant-negative TRAP220; DOC, deoxycholate; E2, estrogen; ER, estrogen receptor; FOG, Friend of GATA; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GATA-RE, GATA-responsive element; GATA2-Zf, Zn-finger domain of GATA2; GR, glucocorticoid receptor; GST, glutathione S-transferase; GST-Zf, Zn finger domain of GST; HAT, histone acetyltransferase; HDAC, histone deacetylase; NCoR, nuclear receptor corepressor; NF-B, nuclear factor-B; NRE, negative regulatory element; nTRE, negative TRE; RAR, retinoic acid receptor; SF-1, steroidogenic factor 1; SMCC, SRB/MED-containing cofactor; SMRT, silencing mediator of retinoic acid and thyroid hormone receptor; SRC-1, steroid receptor coactivator 1; TR, thyroid hormone receptor; TRAP, TR-associated protein; TRE, T3-responsive element; VDR, vitamin D receptor.
Received for publication May 16, 2006. Accepted for publication January 16, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSION |
|---|
