| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Endocrinology, Department of Medicine, University of Colorado Health Sciences Center at Fitzsimons, Aurora, Colorado 80045
Address all correspondence and requests for reprints to: David F. Gordon, Ph.D., Endocrinology, University of Colorado Health Sciences Center-Fitzsimons, Mail Stop 8106, P.O. Box 6511, Aurora, Colorado 80049. E-mail: david.gordon{at}uchsc.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
- and ß-subunits combined in a heterodimeric structure (1). The -subunit is identical in all of the pituitary and placental glycoprotein hormones: TSH, LH, FSH, and chorionic gonadotropin. The TSHß subunit is structurally unique and confers biological specificity to intact TSH. We have shown that the highly restricted nature of TSHß gene expression is likely the result of the concerted action of two transcription factors, Pit-1 and GATA-2, acting on the TSHß gene (2, 3). Within the pituitary Pit-1, a POU (Pit-1, Unc86, Oct-1)-homeodomain factor, is expressed in somatotropes, lactotropes, and thyrotropes, whereas expression of GATA-2, a dual zinc finger protein (4), is restricted to thyrotropes and gonadotropes (5). Thus, this unique combination of Pit-1 and GATA-2, which only exists in thyrotropes, plays a critical role in cell type specification and maintenance of the differentiated state (5). Our previous studies have shown that Pit-1 and GATA-2 bind to adjacent sequences located within a composite cis-acting region on the proximal TSHß promoter (–135 to –88) (2, 6). This composite DNA element has a 5' Pit-1 site and a 3' GATA-2 site. Between these two sites are 16 bp that include overlapping putative Pit-1 and GATA-2 sites. Mutation of either flanking Pit-1 or GATA-2 site completely abolished protein binding and promoter activity. However, the central overlapping Pit-1/GATA-2 sites are also important because mutations within this region reduced occupancy by both factors and dramatically lowered transcription in primary TtT-97 thyrotropes (6). Moreover, gel mobility shift assays revealed the simultaneous occupancy by both Pit-1 and GATA-2 to the composite DNA element as a ternary complex (3). In addition, both factors can form stable protein-protein interactions (2). We mapped the interacting area of Pit-1 to the homeodomain, which also contributes to sequence-specific DNA binding (7). In cotransfection experiments, both factors are required to activate high levels of TSHß promoter activity. Because other transcription factors have been shown to interact with Pit-1 (8, 9) or GATA factors (10, 11), it is likely that additional coactivating factors, which do not bind DNA directly are recruited to the proximal promoter by virtue of their interaction with either Pit-1 or GATA-2 providing additional activation of TSHß gene transcription.
A recent study by Crawford et al. demonstrated that a large thyroid receptor-associated protein (TRAP220, MED220, PBP, DRIP205, ARC), originally defined as part of a large transcriptional mediator complex (SMCC, Mediator) that interacts with thyroid/steroid hormone receptors in a ligand-dependent manner (12, 13, 14, 15), could activate other groups of transcription factors including GATA family members (16). This group demonstrated that MED220/TRAP220, which itself does not bind DNA, could form complexes with at least five GATA family members including GATA-2 and that it could mediate GATA-3 transactivation of a GATA-responsive reporter in 3T3 fibroblasts. Interestingly, gene knockout studies in transgenic mice have implicated MED220 as a coactivator that may participate in the regulation of the TSHß gene in vivo because heterozygote null mice exhibited slightly impaired growth and had markedly reduced levels of TSHß transcripts in their pituitary glands resulting in hypothyroidism (17, 18). Embryonic fibroblasts from the MED220 null mice showed a significant dose-dependent 7-fold reduction in ligand and thyroid receptor-dependent transcription that could be rescued by exogenous TRAP220 (17). This confirmed that MED220/TRAP220 plays an important physiological role as a coactivator with ligand-occupied thyroid hormone receptor. These data suggested that MED220 may have more global functions in pituitary physiology and may physically interact and functionally cooperate with other transcription factors such as GATA-2 and/or Pit-1 that are present in thyrotrope cells to regulate expression of the TSHß gene. Moreover, these studies indicate that, in vivo, a threshold level of MED220 is required to maintain normal levels of the ß-subunit of TSH.
The MED220/TRAP220 complex has been shown to be part of a class of transcriptional cofactors devoid of histone acetyltransferase activity, suggesting that it is formed after other coactivator complexes have modified nucleosomal structure to activate thyroid hormone-activated genes (19, 20). This is distinct from other types of coactivators such as cAMP response element binding protein (CBP)/p300 (21, 22), steroid receptor coactivator-1 (23), and peroxisome proliferator-activated receptor 
coactivator-1 (24), which can be recruited to promoter regions of actively transcribed genes and allow an intrinsic histone acetyltransferase activity in part, to remodel chromatin and to relieve chromatin-mediated repression (25). MED220 also plays a critical role in normal mammalian development because its disruption by homologous recombination of homozygous null alleles in transgenic mice results in lethality at embryonic d 11 (17, 26). These embryos had multiple tissue defects affecting heart, liver, lung, eye, and brain (16, 17, 26, 27). Similar developmental defects had previously been described when GATA family member genes were disrupted in transgenic mice. The fact that this coactivator has been implicated in both activation of the TSHß gene, which ultimately controls levels of thyroid hormone in the body and with T3 receptors, which participate in the down-regulation of this gene, led us to examine expression of the MED220 gene in thyrotrope cells and further to see whether it could activate TSHß promoter activity by functionally and physically interacting with Pit-1 and GATA-2.
To examine MED220 expression, we used the well-characterized TtT-97 thyrotropic tumor that is propagated in hypothyroid LAF1 mice and represents a source of pituitary-derived thyrotrope cells that express both subunits of TSH. Our current studies demonstrate that there are at least three different sized MED220 transcripts in TtT-97 thyrotrope cells, whereas Western blots, using two different polyclonal antibodies, detected a single nuclear protein of about 165 kDa, consistent with the full-length product of 1560 amino acids (aa). Furthermore, we show that MED220 functions as a transcriptional coactivator in combination with Pit-1 and GATA-2 to stimulate mouse TSHß promoter activity in cotransfection assays in CV-1 cells. We also demonstrate in vitro protein-protein interactions involving all three transcription factors by glutathione-S-transferase (GST) pull-down assays and have mapped the key interaction domains to the homeodomain of Pit-1, the dual zinc fingers of GATA-2, and the amino-terminal half of MED220. Furthermore, we show that the amino terminal half of MED220 (aa 1–626) is critically important for full functional cooperativity with Pit-1 and GATA2 on the TSHß promoter. Coimmunoprecipitation (CoIP) assays using nuclear extracts from TSHß expressing TtT-97 thyrotropes demonstrated in vivo protein-protein interactions involving all three factors. Finally, chromatin immunoprecipitation (ChIP) assays show that Pit-1, GATA2, and MED220 can occupy the proximal TSHß promoter in thyrotrope cells in vivo.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-subunit genes resulting in exuberant TSH production (29). These transplantable hyperplastic tumors were originally created by sc transplantation of enlarged pituitaries from hypothyroid LAF1 mice into hypothyroid LAF1 hosts. These diploid cells behave in many ways like hyperplastic pituitary thyrotrope cells in a hypothyroid in vivo environment and exhibit normal regulation by thyroid hormone (30). Because different sizes of MED220 transcripts have been reported by Northern blot analyses in several mouse and human tissues that were examined (12, 13), we performed Northern blot analyses with TtT-97 polyadenylated [poly(A)+] RNA with a series of human MED220 cDNA probes spanning the entire coding region. There is a high degree of sequence similarity at the nucleotide and aa levels (97.2%) between the reported human and mouse MED220 cDNA coding regions. A schematic diagram of the 5008-bp human MED220 cDNA insert (gift of Dr. Robert Roeder, Rockefeller University, New York, NY) is shown in Fig. 1A
with the sizes and relative codon locations of the six probes used in our analysis. Each probe was 32P-radiolabeled and hybridized to 10 µg poly(A)+ RNA from TtT-97 cells (Fig. 1B). Probe 1 encoding the N-terminal residues 1–327 detected three major bands of 8.0, 2.7, and 1.8 kb with similar autoradiographic intensity and two additional minor bands of approximately 4.3 kb that may represent unspliced intermediates or transcripts of related genes. The 8-kb band is identical with the size reported by Zhu et al. (12) from a variety of mouse tissues, whereas they only detected the 2.7-kb band in testes.
|
). This suggested that the 1.8-kb band found with probe 1 did not contain the central portion of the coding region and that additional truncated forms of MED220 may be present in thyrotropes. Probe 5 only detected the largest hybridizing band, although this difference may be biased due to its lower specific activity. Finally, probe 6 detected the 8- and 2.7-kb bands, and a novel 1.5-kb transcript possibly representing a C-terminal truncated form of the protein.
Additionally, we used the N-terminal probe 1 on a multiple tissue blot (CLONTECH, Mountain View, CA) containing 2 µg poly(A)+ RNA from a variety of mouse tissues. We detected the largest 8-kb transcript in all tissues. Heart and spleen contained the same three transcripts as was found with TtT-97 RNA, whereas liver and testes contained only the 8- and 2.7-kb bands, the latter being highly expressed in testes only (Fig. 1C). MED220 protein from thyrotrope cells was then detected by Western blot analysis. As shown in Fig. 1D, using nuclear extracts from two independent TtT-97 thyrotropic tumors, we detected a single MED220 protein of about 165 kDa, consistent with the full-sized mouse protein of 1560 aa (www.ensembl.org; ENSMUSP00000018304). Thus, the smaller transcripts detected on Northern blots may not be efficiently translated in thyrotropes.
MED220 Further Activates Mouse TSHß Promoter Activity in Combination with Pit-1 and GATA-2
Transcriptional activation of cell-specific genes often involves the functional cooperativity of combinations of transcription factors that bind to adjacent or overlapping composite cis-acting DNA elements. Recent studies have also demonstrated that a number of coactivators or corepressors can be recruited to actively transcribed genes via protein-protein interactions (31). In pituitary thyrotropes, two transcription factors, Pit-1 and GATA-2 are important for mediating thyrotrope-specific expression of the TSHß gene where they bind to adjacent sites on the proximal promoter, physically interact, and lead to synergistic increases in gene transcription (2, 3). Because MED220/TRAP220 +/– mice had low levels of TSHß in vivo we tested whether MED220 might act as a coactivator along with Pit-1 and GATA-2 by cotransfecting it alone or in combination with the other two thyrotrope factors. We performed these transient cotransfection experiments in CV1 monkey kidney cells because they do not contain endogenous levels of Pit-1 or GATA-2, although they are likely to contain MED220 because we have shown that normal kidney expresses the full-length 8-kb transcript (Fig. 1C). Figure 2A shows promoter activity of a –392/+40 mouse (m) TSHß 5' region linked to firefly luciferase in the absence or presence of Pit-1, GATA-2, and MED220 alone or in combination. Cotransfection of MED220 at doses up to 3 µg failed to stimulate TSHß promoter activity over an empty vector control (0.9- to 1.3-fold). Similar to our previous findings using optimized amounts of cotransfected plasmids (3), addition of 0.75 µg Pit-1 alone resulted in only a modest increase of 3.4-fold, whereas 3 µg GATA-2 failed to stimulate by itself. As before, cotransfection of Pit-1 and GATA-2 together led to a significant 12-fold increase demonstrating transcriptional cooperativity. Using these same amounts of Pit-1 and GATA-2, we then titrated in the MED220 expression vector in doses from 1–5 µg. There was a dose-dependent increase in TSHß promoter activity that reached a maximum of 24.5-fold relative to the vector control with 3 µg of plasmid. Further increases of 4 or 5 µg led to slightly lower promoter activity, suggesting a possible squelching mechanism due to competition for other factors. Figure 2A also shows that MED220 did not increase promoter activity when cotransfected with only GATA-2 (1.3- to 2.0-fold, respectively) whereas addition of MED220 at higher levels did increase the ability of Pit-1 to stimulate the TSHß promoter (6.0-fold over control). These studies demonstrated that MED220 can act as a coactivator along with both Pit-1 and GATA-2 on the TSHß promoter and may explain the lower level of TSHß gene expression in vivo in heterozygous MED220 null mice with one half of the normal gene levels of this factor. The effect was promoter specific because no increase in promoter activity was seen when similar combinations of Pit-1, GATA-2, and MED220 were tested on a cytomegalovirus (CMV)-luciferase construct (Fig. 2B).
|
. A previously identified transactivation domain located within the N terminus (TAD) as well as the POU-specific and POU-homeodomains responsible for sequence-specific DNA binding to cognate recognition sites are shown by the black or shaded boxes with the aa number shown at the top. The relative locations of GST fusions for the N terminus (residues 2–124), the POU-specific domain (residues 109–200), the POU-homeodomain (residues 198–291), and the DNA binding domain (residues 109–291) are shown compared with the full-length protein. Fusion proteins were purified on glutathione-Sepharose beads and size-separated by electrophoresis on sodium dodecyl sulfate (SDS)-protein gels. Figure 3B shows the Coomassie blue staining of normalized amounts of each Pit-1 fusion as well as GST alone. We synthesized full-length human MED220 and an N-terminal fragment containing residues from 1–327 in a coupled 35S-labeled transcription/translation reaction using T7 RNA polymerase and rabbit reticulocyte lysates and used them to test interactions with various GST fusions of Pit-1. Figure 3C shows stable interactions between full-length MED220 and full-length Pit-1, the homeodomain alone or the POU+ homeodomain accounting for about 20% of the input radiolabeled MED220. The POU-specific domain exhibited a modest interaction, whereas no interaction was observed with the N-terminal 124 residues or with the GST-negative control. We also found the same pattern of Pit-1 GST fusion interaction with the N-terminal fragment of MED220 (residues 1–327). This mapped the interaction domain of MED220 to the homeodomain of Pit-1, a region of the protein that we have previously shown to interact with GATA-2. These data also identified the N terminus of MED220 as the interaction domain with the homeodomain of Pit-1, a region of MED220 that is distinct from the two nuclear receptor binding domains (NR boxes, LXXLL) that participate in protein interactions with helix 12 of liganded steroid/thyroid receptors.
|
, including the relative locations of the dual zinc finger domains Znf1 (N) and Znf2 (C) that delineate the DNA binding domain. Shown below are the truncated domains that include the N terminus, C terminus, each separate zinc finger, as well as the dual finger domain. After purification, we showed that each construct produced the expected size of the GST fusion as shown on a representative Coomassie blue-stained protein gel (Fig. 4B) in parallel with known amounts of BSA as a standard.
|
). At the top of Fig. 4C are shown the results of the GATA-2 fusions with the Pit-1 HD. The input Pit-1 HD shows at least three closely spaced bands of the appropriate size indicating some premature termination in the coupled translation and/or posttranslational modifications. Although no signal was detected with the GST control, we demonstrated an equivalent interaction of the Pit-1 homeodomain with full-length GATA-2, the dual zinc fingers (Znf1+2), or with the most distal zinc finger, Znf2. The proximal zinc finger (Znf1) did form a weak interaction that appeared to only involve the smaller of the three closely spaced bands shown with the 35S-labeled input signal. We did not detect any interaction with the N or C terminus that allowed us to map the interaction domain, suggesting that the Pit-1 HD interaction was predominantly determined by the distal zinc finger (Znf2) of GATA-2 with a minor contribution from the proximal zinc finger.
Using a full-length MED220 radiolabeled protein, we repeated the interaction analysis with these same GST-GATA-2 fusions. Figure 4C shows that MED220 did not interact with the GST control or with the N terminus but, in contrast to the Pit-1 HD, appears to interact equally with each zinc finger of GATA-2. Additionally, we demonstrated an additional interaction with the C terminus of GATA-2 not seen with Pit-1. These data demonstrate that both Pit-1 and MED220 each can form stable protein-protein interactions with the dual zinc fingers of GATA-2, whereas MED220 in contrast to Pit-1 can also interact with the GATA-2 C terminus.
The LXXLL Motif of MED220 Is Not Required for the Interaction with Pit-1 or GATA-2
Because different domains of MED220 were found to be involved in the interaction with a subset of different GATA family members (GATA-1, GATA-4, or GATA-6) (16), we mapped which areas of this important coactivator interacted with both Pit-1 and GATA-2, which transcriptionally cooperate with MED220 on the TSHß gene. A schematic diagram of the full-length MED220 coding region is shown at the top of Fig. 5A with the two closely spaced nuclear receptor (NR) boxes with the sequence LXXLL indicated, along with a basic region, serine-rich domain, and a charged aa domain located within the C terminus. A number of other cofactors contain the LXXLL motif that is both necessary and sufficient in mediating their binding to nuclear steroid/thyroid receptors in the presence of ligand (31). Because we previously showed that full-length MED220 could interact with the homeodomain of Pit-1 (Fig. 3C) and the dual zinc fingers of GATA-2 (Fig. 4C), we tested whether the NR boxes of MED220 are required for these interactions. We expressed a series of separate truncated MED220 fragments (Fig. 5A) as 35S-lableled proteins synthesized in a coupled transcription/translation system and incubated each with GST alone, or with either the Pit-1 HD or the GATA-2 dual zinc fingers, the regions that interacted with the full-length protein. The results are shown in Fig. 5B. For both Pit-1 and GATA-2, we show relatively strong interactions, when compared with the input signal, with MED220 residues 1–327 (N terminus) and with a fragment containing residues 327–626, the latter including the first NR box. The MED220 fragment of residues 440–740 including both NR boxes showed a weaker interaction than with the N-terminal fragment. However, we could detect no interaction with two additional C-terminal fragments, including residues 740-1130 or 1130–1581. None of the labeled MED220 fragments interacted with the GST control. These results show that the same regions of MED220 can interact similarly with both Pit-1 and GATA-2 and that each maps to a broad region at the N-terminal half of the protein but not to the C-terminal half. Because the N-terminal 327-aa fragment lacks either LXXLL motif, this demonstrates that the NR boxes are not absolutely required for protein-protein interactions of MED220 to either Pit-1 and GATA-2. Because the two fragments containing either the first NR box (residues 326–626) or both LXXLL motifs (residues 440–740) still formed a physical interaction, we cannot rule out that this motif does not participate in a more global interaction involving the entire N-terminal half of MED220. MED220’s interactions with GATA-2 are similar to the interactions formed with GATA-1, and are distinct from those with GATA 4 and GATA-6 (16). Thus, distinct domains of MED220 are used by this cofactor when it interacts with different partner transcription factors, perhaps allowing it a broader and more global role in the regulation of eukaryotic transcription.
|
. Similar to our previous result (Fig. 2) in the absence of MED220, the combination of Pit-1 and GATA2 increased TSHß promoter activity 10.2 ± 0.49-fold (n = 6) when compared with an empty vector control. In Fig. 6, we set this activity at 1 (dotted line). The full-length MED220 construct (aa 1–1581) stimulated the promoter 25.2 ± 1.1-fold (n = 6) over the empty vector control or 2.47-fold higher than the activity of Pit-1 plus GATA2 alone. Next, we tested the function of two N-terminal fragments that retained strong protein-protein interactions with Pit-1 and GATA2. The shorter mutant (aa 1–327) showed diminished cooperativity of promoter activity of 1.71-fold over that of Pit-1 plus GATA2 alone, whereas the MED220 mutant containing aa 1–626 had 2.23-fold activity, which was not significantly different when compared with the full-length protein. In contrast, the C-terminal remaining fragment containing aa 627-1581, which interacted weakly with Pit-1 and GATA2, exhibited a much lower activity of 1.31-fold. Finally, we showed that the most C-terminal mutant (aa 1130–1581), which did not interact with Pit-1 or GATA2, did not stimulate the TSHß promoter (0.97-fold) in the presence of both cotransfected partner factors. Thus, there is a strong correlation between regions of MED220 important for protein-protein interactions with Pit-1 plus GATA2 and maximal transcriptional activity of the TSHß promoter.
|
shows that a CoIP using an antibody against Pit-1 detects a 50 kDa GATA2 protein (lane 1) as well as GATA2 in TtT-97 nuclear extracts (lane 2), whereas GATA2 was not detected with a CoIP using the same amount of control rabbit IgG (lane 3). The reciprocal CoIP demonstrates that both the 33- and 31-kDa forms of Pit-1 are present in the GATA2-antibody immunoprecipitated (IP) complex (Fig. 7B). Finally, using a combination of two goat polyclonal antibodies directed against MED220 in CoIP experiments, we detected both Pit-1 (Fig. 7C) and GATA2 (Fig. 7D) in the IP complexes and in nonprecipitated TtT-97 nuclear extracts. These studies demonstrate that stable protein-protein interactions between these three transcription factors can occur in vivo in TSHß-expressing thyrotrope cells.
|
, we observed enrichment of the TSHß promoter with ChIP assays using Pit-1, GATA-2, and MED220 over the signal derived from control IgG. There was no enrichment of the same DNA on the glyceraldehyde 3-phosphate dehydrogenase gene. These data indicate that these three transcription factors occupy the proximal TSHß promoter in vivo under hypothyroid conditions where the endogenous gene is maximally expressed.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
) using two different polyclonal antibodies, suggesting that the smaller transcripts are not translated efficiently or are not stable in pituitary derived cells.
Secondly, we provided evidence by cotransfection studies that MED220 acts as a transcriptional coactivator on the proximal TSHß promoter but not with a CMV promoter (Fig. 2). We observed the highest promoter activity with a combination of Pit-1, GATA-2, and an intermediate dose of MED220 (3 µg). Doses less than 3 µg of MED220 did not show increased promoter activity over the synergistic activation observed with a combination of Pit-1 plus GATA-2 in the absence of MED220. This dosage effect is consistent with the in vivo finding that showed lower promoter activity with one half the dose of this cofactor in vivo found in the haploinsufficient MED220 mice. Our data showing that MED220 addition along with cotransfected GATA-2 did not transcriptionally activate the TSHß promoter differs from a previous study reported by Crawford et al. (16). In that study using NIH 3T3 fibroblasts, a trimerized CßE GATA-3 binding site fused to a minimal rabbit ß-globin promoter/human GH reporter was minimally activated by either MED220 or GATA-3 but was activated 5-fold with a combination of both factors. Thus transcriptional cooperativity of MED220 with GATA family members may be promoter specific or may be dependent on a particular GATA isoform. In support of the latter possibility, it has been demonstrated that individual GATA family members, via the conserved dual zinc finger domains, form distinct interactions with different domains of MED220 (16). For example, mouse GATA-1 interacted with residues 1–380 of MED220 and not with residues from 440-1560, whereas GATA-4 interacted with MED220 at residues from 740-1130 and GATA-6 with residues from 230–626. In the present study, we observed identical interactions with both the dual zinc fingers of GATA-2 and the homeodomain of Pit-1 with MED220 residues 1–327, which does not include the dual LXXLL motifs, as well as with MED220 fragments containing residues 327–626 and residues 440–740, which contain one or both LXXLL motifs, respectively. Thus, both GATA-1 and GATA-2 can interact with the N terminus of MED220, whereas GATA-4 and GATA-6 do not. Because all GATA family members contain a highly conserved dual zinc finger, it appears that sequences in either the N or C terminus, which differ between GATA family members are likely to dictate which domains of MED220 can form protein-protein interactions in the absence of DNA. The fact that the dual zinc fingers of GATA-2 and the homeodomain of Pit-1, which interact with each other (Fig. 4), and in addition interact with the same region of MED220 (Fig. 5B), suggests that all three factors may form a unique three-protein complex that is responsible for high levels of TSHß promoter activity in pituitary thyrotropes.
Our studies have also shown that the homeodomain of Pit-1 can physically interact with primarily the second zinc finger of GATA-2 (residues 340–412) with a minor contribution from the first zinc finger (residues 280–340, Fig. 4C). No interactions were observed with either the amino or carboxy-termini of GATA-2. These results are consistent with a previous report that mapped the interaction of the homeodomain of Pit-1 to the second zinc finger and an adjacent cluster of basic residues of GATA-2 (5). Similar interactions have also been described for the homeodomain transcription factor Nkx2.5 with GATA-4. These factors can physically and functionally interact on the atrial naturietic factor promoter where the interaction domains also map to the homeodomain of Nkx2.5 and the second zinc finger domain of GATA-4 (32). Additionally, the zinc fingers of GATA-6 directly interact with the homeodomain of thyroid transcription factor-1 to regulate the surfactant protein-C gene (33). Our findings that MED220 can interact with and enhance transcription of the TSHß promoter in concert with Pit-1 and GATA-2 may be a more general mechanism used by combinations of homeodomain proteins with GATA family members on other gene promoters.
Also, we demonstrated by CoIP studies that the interactions between Pit-1, GATA-2, and MED220 found in vitro also occur in vivo in TSHß expressing thyrotrope cells and that all three factors can occupy sites on the endogenous TSHß promoter in thyrotropes by ChIP assays. These data strongly argue that these factors functionally collaborate to enable high levels of transcription by a mechanism involving both protein-DNA and protein-protein interactions on the proximal TSHß promoter.
Both Pit-1 and GATA-2 can bind to and simultaneously occupy a complex composite element within the proximal mouse TSHß promoter (2). Because they can form protein-protein interactions in the absence of DNA, they may first bind separately to DNA and then interact with each other. Alternatively, they may first interact with each other in the absence of DNA and then be recruited to the TSHß promoter. After DNA binding, there may be a novel interface that becomes exposed which allows the recruitment of additional factors such as MED220 and/or CBP/p300, which has been shown to be a coactivator on the rat prolactin promoter in concert with Pit-1 after protein kinase A stimulation (34). Moreover, it was shown that two widely separated cysteine/histidine-rich domains of CBP located within residues 118–737 and 1677–2441, formed a protein-protein interaction with Pit-1. Similarly, it has been shown that GATA family members can bind to CBP and result in transcriptional synergy (11, 35, 36) where the GATA factors, as well as histones, may be directly acetylated. Histone acetylation has been correlated with chromatin remodeling and relief of histone repression. There may then be a sequential and temporal exchange (31) of bound CBP for other coactivators such as MED220, which we have shown can interact with both Pit-1 and GATA-2 (Figs. 3 and 5). Such an ordered recruitment of CBP/p300 and MED220 has recently been demonstrated in vivo on the thyroid hormone-stimulated human iodothyronine deiodinase type I gene, after ligand administration using ChIP assays (20). This would likely result in the recruitment of additional proteins within a large transcriptional mediator complex onto the proximal TSHß promoter.
The N terminus of MED220 has recently been shown to be necessary and sufficient to form stable interactions with additional proteins making up the intact TRAP/Mediator complex (37). Initial mapping experiments have localized the key region involved in this interaction to aa 108–390, which excludes the two LXXLL motifs important for interaction with ligand occupied steroid/thyroid receptors. Our data also shows that the N terminus (aa 1–327) of MED220 can also functionally and physically interact with Pit-1 and GATA-2 (Figs. 3B and 5B) on the TSHß promoter, although other interactions with more C-terminal regions were required for maximal function. We demonstrated such maximal functional TSHß promoter activity in the presence of Pit-1 and GATA-2 using an N-terminal deletion mutant of MED220 (aa 1–626), which was not significantly different when compared with the wild-type protein (Fig. 6). This region, which contained only the first LXXLL motif, also interacted strongly with Pit-1 and GATA2. Using a similar fragment (aa 1–670), Malik et al. (37) mapped full functional activity on a thyroid-responsive template in transient transfection assays or in reconstituted in vitro transcription assays. A C-terminal MED220 deletion (aa 1130–1581) neither physically interacted with Pit-1 and GATA2 nor had any additional function on TSHß promoter activity (Fig. 6). Thus, the N terminus of MED220 may act as a tether between the two thyrotrope-restricted transcription factors and recruit the other proteins of the Mediator/TRAP complex. When it is not present in sufficient levels, much of the Mediator complex would not be expected to be recruited to the TSHß promoter, resulting in lower rates of transcription, and is consistent with the hypothyroid phenotype of the MED220 null heterozygote mouse.
Because liganded thyroid hormone receptor has been shown to functionally and physically interact with MED220, it may also play an important role in modulating levels of this factor that are available to be recruited to the TSHß promoter. In the absence of thyroid hormone, thyroid receptor ß1/2 would not be able to bind to the two LXXLL motifs of MED220 (38, 39), thus freeing it to be recruited to the proximal TSHß promoter in combination with Pit-1 and GATA-2. However, in the presence of thyroid hormone, TRß1/2 undergoes a conformational change that would allow it to interact with MED220 and perhaps sequester it away from the Pit-1/GATA-2 complex on the TSHß promoter. This would result in a reduced transcriptional rate for this gene. Alternatively, if MED220 is competed from the Pit-1/GATA-2 complex, these factors might more easily dissociate from the promoter and allow liganded TRß to bind to a negative thyroid response element in the proximal promoter, and thus recruit histone deacetylases (40, 41) that modify histones and result in a fully repressed state. The latter mechanism is favored, supported by several recent studies demonstrating the requirement of an intact DNA binding domain of TRß in the negative regulation of the TSHß gene in vitro (42) and in vivo (43). Nakano et al. (42) used a combination of Pit-1 and GATA-2 to stimulate a TSHß (–128/+37) chloramphenicol acetyltransferase (CAT) construct in CV-1 cells and cotransfected wild-type or mutant TRß1 constructs in the absence or presence of T3. They found that unliganded TRß1 did not stimulate promoter activity, whereas a mutation lacking the N terminus and DNA binding domain of TRß1 lost the ability of T3-treated cells to negatively regulate TSHß-CAT activity. Moreover, using a gene targeting approach in transgenic mice, Shibusawa et al. (43) replaced the wild-type TRß gene with a P-box (GS125) mutant that abolished DNA binding in vitro without altering ligand and cofactor interactions. Homozygous mutant mice demonstrated central thyroid hormone resistance with 20-fold higher serum TSH in the face of 2- to 3-fold higher T3 and T4 levels that were similar to those of TRß homozygous null mice.
In summary, we have shown that MED220 can act as a coactivator in concert with Pit-1 and GATA-2 on the TSHß promoter where it uses similar broad domains in the amino terminal half of the protein to interact perhaps simultaneously with both Pit-1 and GATA-2 as well as to other protein components within a large transcriptional complex. Thus, the TSHß gene is activated by a unique combination of transcription factors present in pituitary thyrotropes, including those that act via binding to the proximal promoter as well as others that are recruited to the promoter via protein-protein interactions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
was transformed, clones checked for the correct orientation, and the sequences verified by DNA sequencing. Construction of full-length GST human GATA-2 was performed by PCR with a HindIII site (underlined) at the amino terminus and a BglII site at the carboxy terminus using a sense primer, 5'-GCAAGCTTCCATGGAGGTGGCGCCGGAG-3' and an antisense primer 5'-GCAGATCTCTAGCCCATGGCGGTCACCATGCT-3' with an annealing temperature of 59 C as described (2). Construction of GST fusions with GATA-2 subfragments used a similar strategy using the following primer sets: (residues 1–280), sense 5'-GCAAGCTTCCATGGAGGTGGCGCCGGAG and antisense 5'-GCAGATCTCTAAGGGGTGAAGCTGGAGGCCGG-3'; (residues 280–340), sense 5'-GCAAGCTTCCCCTAAGCAGCGCAGCAAGGCT and antisense 5'-GCAGATCTCTACGACAGTCTTCGCTTGGGCTT-3'; (residues 340–412), sense 5'-GCAAGCTTCCTCGGCCGCCAGAAGAGCCGGC-3' and antisense 5'-GCAGATCTCTACTCCGCCCCTTTCTTGCTCTT-3', (residues 280–412), sense 5'-GCAAGCTTCCCCTAAGCAGCGCAGCAAGGCT-3' and antisense 5'-GCAGATCTCTACTCCGCCCCTTTCTTGCTCTT-3'; and (residues 396–480), sense 5'-GCAAGCTTCCCGGAACCGGAAGATGTCCAAC-3' and antisense 5'-GCAGATCTCTAGCCCATGGCGGTCACCATGCT-3'. All fragments were directionally cloned into the HindIII to BglII sites of pGEX-DFGK and verified by sequencing.
Generation of Plasmid Constructs for in Vitro Transcription
The homeodomain of rPit-1 (residues 198–291) was amplified by PCR using the following primer set that incorporates a NotI site at each end, sense 5'-GAGCGGCCGCATGCAGGTCGGAGCTTTGTAC-3' and antisense 5'-GAGCGGCCGCTTATCTGCACACAAGATG-3' with an annealing temperature of 60.2 C. Fragments were cut with NotI and cloned into NotI linearized pTNT (Promega, Madison, WI) and verified by DNA sequencing.
A fragment containing the first 327 residues of hMED220 was amplified by PCR using a similar strategy with an EcoRI site at the amino terminus and a NotI site at the carboxy terminus using a sense strand primer 5'-GCGAATTCCCATGAAAGCTCAGGGGGAAACC-3'and an antisense primer 5'-GAGCGGCCGCTAAATTCCTGTGCAGTTCTG-3' with an annealing temperature of 59.3 C. After digestion with EcoRI and NotI, the fragment was directionally cloned into the same sites of pTNT. Full-length hMED220 in pGEM5Zf+ was provided by Dr. R. Roeder and was used as the template for the following PCRs using primer sets with an EcoRI and NotI sites into pTNT: (residues 327–626), sense 5'-GCGAATTCCCATGATTCCATTGTTTGAAACT-3' and antisense 5'-GAGCGGCCGCTAAATTCCTGTGCAGTTCTG-3'; (residues 441–740), sense 5'-GCGAATTCCCATGGAAGTGTGTCCTCTCTCAG-3' and antisense 5'-GAGCGGCCGCTATGGAGTGATGTGTGGCGT-3' (residues 740-1130), sense 5'-GCGAATTCCCATGCCAGCTCCAAGCCAGTGT-3' and antisense 5'-GAGCGGCCGCTAACTGCTACTTAACTTGGA-3' and; (residues 1130–1581), sense 5'-GCGAATTCCCATGAGTATGTATTCTAGCCAG-3' and antisense GAGCGGCCGCTAATTCCCAATCAGGGCCAC-3'.
Northern Analysis of MED220 Transcripts in TtT-97 Thyrotropic Tumors
RNA was isolated from TtT-97 thyrotropic tumors from a hypothyroid mouse using the guanidinium isothiocyanate-CsCl method (44). Poly(A)+ RNA was isolated by affinity chromatography over an oligo(deoxythymidine) cellulose column (type 7; Pharmacia Biotech, Inc., Piscataway, NJ). The RNA was size-fractionated through a 1% agarose gel containing 6% formaldehyde as described (45) and transferred to a nylon membrane (Nytran, 0.2 mm; Schleicher and Schuell, Keene, NH) and fragments were fixed by ultraviolet light cross-linking (model FB-UVXL-1000; Fisher, Pittsburgh, PA) and hybridized to a series of different coding region cDNA probes. Each probe was 32P-radiolabeled by random priming (9 nucleotide oligomers) with an exonuclease-free Klenow DNA polymerase using a Prime-It II kit (Stratagene, La Jolla, CA). Radiolabeled standards of 
-phage DNA cut with HindIII were prepared by filling in-3' ends with 32P-deoxy-CTP and incubation with 10 U reverse transcriptase (Promega) at 40 C for 40 min and purified by Sephadex G50 column chromatography were fractionated in parallel. Prehybridization, hybridization, and wash conditions have been described previously (45).
Generation of MED220 deletion mutant constructs for cotransfection studies
Two N-terminal and two C-terminal deletion mutants of MED220 were constructed using a PCR approach. An N-terminal fragment containing aa 2–327 or 2–626 was amplified from 1 ng of full-length MED220 as a template using a common sense strand oligonucleotide 5'-GCGCGGCCGCAAAGCTCAGGGGGAAAC-3' (NotI site underlined) containing aa 2–7 and a unique antisense strand oligonucleotide 5'-GAGCGGCCGCTAAATTCCTGTGCAGTTCTG-3' (aa 327–322) or 5'-GAGCGGCCGCTAATGAGGAGGGGTCGGACT (aa 627–621). A construct containing aa 627-1581 was amplified using a sense strand oligonucleotide (aa 627–633) 5'-GTGCGGCCGCCACACGCCGCCACCTGTCTCTT-3' and an antisense strand primer (aa 1581–1576) 5'-GAGCGGCCGCTAATTCCCAATCAGGGCCAC-3'. Finally, a shorter C-terminal deletion containing aa 1130–1581 was amplified using the same antisense primer as above with a unique sense strand primer (aa 1130–1137) 5'-GCGCGGCCGCAGTATGTATTCTAGCCAGGGGTC-3'. After amplification of the correctly sized product, they were gel purified, cloned into pCR2.1, and the nucleotide sequence verified. Recombinant plasmids were excised by digestion with NotI and cloned into NotI cut pCGN2. Inserts with the correct orientation were selected by restriction enzyme mapping and verified by DNA sequencing using an antisense primer downstream of the NotI site.
Cotransfection of MED220 on TSHß Promoter Activity
Construction of the –392 to +40 mTSHß firefly luciferase vector in pA3luc has been described previously (46) as has the construction of plasmids containing hGATA-2 and rPit-1 with a CMV promoter and containing a hemagglutinin tag at their amino terminus (pCGN2) (2). Full-length human MED220 with a CMV promoter in pCIN4 was a kind gift from Dr. Robert Roeder. CV-1 cells were transiently transfected using a previously described calcium phosphate method (47) as described (2). Approximately 750,000 cells were cotransfected with 10 µg of a –392/+40 mTSHß firefly luciferase construct alone or in combination with either 3 µg of pCGN-2-hGATA-2 and/or 0.75 µg of pCGN-2-rPit-1, and/or 0.5 to 5 µg pCin4/MED220 as described in Results. In addition, as a negative control, the mTSHß firefly luciferase construct was replaced by a CMV firefly luciferase construct and cotransfected into CV-1 cells alone and with the same combinations of pCGN-2-hGATA-2, pCGN-2-rPit-1 and 3 µg pCin4/TRAP-220 as described above. Each transfection also contained 25 ng Renilla luciferase plasmid (Promega) as an internal transfection control. The total amount of plasmid used in a transfection assay was adjusted to a total of 16.75 µg with a nonspecific control plasmid. A Rous sarcoma virus-luciferase plasmid and an empty pA3 luciferase plasmid were transfected in parallel as positive and negative controls, respectively. Cells were harvested after 48 h of incubation at 37 C, lysed with Passive Lysis Buffer (Promega), subjected to freeze thaw extraction, and assayed for dual firefly and Renilla luciferase activity. Luciferase activity was measured in a Monolight 3010 luminometer using a Dual Luciferase Reporter Assay System (Promega). Firefly luciferase light units were normalized to Renilla luciferase activity. All transfections were performed in duplicate at least four times. Statistically significant differences were tested by two-way ANOVA (48).
In Vitro Protein-Protein Binding Assays
Bacterial extracts containing the recombinant GST Pit-1 or GATA-2 fusions or GST alone were prepared essentially as described previously (49) with some modifications (3). The bacterial cells were harvested, and the cell pellet was resuspended in 20 ml of fusion protein buffer (150 mM NaCl, 16 mM NaH2PO4, 4 mM Na2HPO4, 1% Triton X-100, 2.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, supplemented with a protease inhibitor mixture) (Complete; Roche, Indianapolis, IN). Bacteria were lysed by sonication, and supernatant was obtained as described (49). Supernatant (50 ml) containing each GST fusion were mixed with 0.5 ml of a 50% slurry of glutathione-Sepharose 4B (Pharmacia) for 1 h at 4 C followed by washing three times with 1 ml buffer. Concentration of bound protein was determined using a Bio-Rad (Richmond, CA) assay and by SDS-PAGE and Coomassie blue staining with varying amounts of BSA electrophoresed in parallel lanes.
Coding sequences for full-length MED220 and rPit-1 or truncated forms of each protein were cloned into the vector pTNT, and proteins synthesized and radiolabeled with [35S]methionine (NEN Life Science Products, Boston, MA) using a reticulocyte lysate-coupled transcription-translation system (TNT, Promega) under conditions suggested by the supplier. Approximately 2 µg of each GST fusion or GST alone were immobilized onto Sepharose beads (25% slurry) and mixed with 5 µl of 35S-labeled hGATA2, in a total volume of 500 µl of binding buffer [40 mM HEPES (pH 7.5), 100 mM NaCl, 0.5 mM EDTA, 5 mM MgCl2, 1 mM dithiothreitol, 0.05% Nonidet P-40, 0.5 mM phenylmethylsulfonyl fluoride] and supplemented with a protease inhibitor mixture (Complete, Roche). The suspension was incubated on a rotator at 4 C for 1 h, and beads were allowed to settle by gravity for 10 min and washed as described (2). The beads were resuspended in 65 µl of 2x treatment buffer [0.125 M Tris-Cl (pH 6.8), 4% SDS, 20% glycerol, 10% 2-mercaptoethanol], boiled for 90 sec, and analyzed by SDS-PAGE and autoradiography. Exposure times were 16–40 h.
CoIP Assays in TtT-97 Thyrotropes
Immunoprecipitations were performed by magnetic bead separation (MACS separation; Miltenyi Biotec, Gladbach, Germany) as recommended by the supplier with minor modifications. All antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). Nuclear extracts of TtT-97 thyrotropic tumors were prepared as described (50) with the addition of 1 mM sodium vanadate (pH 10) to inhibit phosphatases and 5 µM zinc sulfate was included for zinc finger proteins. Approximately 300 µg (3 µg/µl) in 750 µl of 50 mM Tris-Cl (pH 8), 1% Nonidet P-40 (NP40), 25 mM NaCl, 1 mM sodium vanadate, 1 mM EDTA, 10 µM zinc sulfate, 1 mM dithiothreitol, 2% glycerol, and a cocktail of protease inhibitors was added to a microfuge tube at 4 C and 50 µl of protein G microbeads included for 30 min with constant rotation to reduce background binding, followed by application to a microcolumn in the magnetic field of a micro-MACS separator. The void volume was collected and incubated for 3 h with 2 µg of rabbit polyclonal antibodies against GATA-2 (SC9008), Pit-1 (SC402), or with a combination of two goat polyclonal MED220 antibodies (SC5334, SC5335), or with control IgG (SC2025). Protein G microbeads were added to each tube (50 µl) and allowed to bind at 4 C for 1 h before being applied to a prewashed microcolumn in binding buffer in the field of a magnetic separator. The column was washed four times with 300 µl binding buffer, proteins eluted with two successive 35-µl aliquots of 2x Laemmli sample buffer, electorphoresed on 10% polyacrylamide protein gels in parallel with nonprecipitated nuclear extracts and electroblotted onto polyvinylidene fluoride membranes. The proteins were analyzed by Western blotting using the indicated primary antibodies in Fig. 6 and detected with horseradish peroxidase conjugated secondary antibodies and a chemiluminescent substrate (SuperSignal West Pico; Pierce Biotechnology, Rockford, IL).
ChIP Assays in TtT-97 Thyrotropes
The ChIP method was a modification of the method by Boyd and Farnham (51). Three grams of TtT-97 thyrotropic tumor tissue were excised from a hypothyroid LAF1 mouse, finely minced in 1 ml PBS with a razor blade at room temperature for 5 min, transferred to a 50-ml conical tube in 15 ml PBS and the solution adjusted to 1% formaldehyde. The tube was rotated at 25 C in a hybridization oven (Fisher) for 15 min and the solution adjusted to 0.125 M glycine, followed by rotation for 5 min. The tube was placed on ice and centrifuged at 4 C at 1500 rpm for 5 min to pellet cells, supernatant was aspirated and cells washed with 10 ml PBS, recentrifuged and washed with an additional 50 ml ice-cold PBS. Cells were transferred to a 15-ml Wheaton Dounce homogenizer (Kontes Glass Co., Vineland, NJ) in 10 ml cell lysis buffer [5 mM piperazine-1,4-bis(2-ethane sulfonic acid) (pH 8.0), 85 mM KCl, 0.5% NP40, 0.5 mM phenylmethylsulfonyl fluoride, and one Complete mini tablet (protease inhibitor cocktail, Roche)] and dounced 15 times on ice with the B (tight) pestle to release the nuclei, followed by incubation on ice for 15 min and douncing for five more times. Contents were transferred to a 15-ml conical polypropylene tube and centrifuged for 3500 rpm for 5 min at 4 C. The nuclear pellet was resuspended in 10 ml nuclear lysis buffer [50 mM Tris-Cl (pH 8), 10 mM EDTA, and 1% SDS, 0.5 mM phenylmethylsulfonyl fluoride and one Complete mini tablet (Roche)], gently resuspended with a pasteur pipet and 1.5-ml aliquots were frozen at –80 C in 2-ml microfuge tubes. Tubes were thawed on ice, 200 mg acid washed glass beads added (Sigma, St. Louis, MO; G1277), and chromatin was sonicated 25–30 times for 15-sec pulses with a microtip using a Fisher sonicator at 4 C to an average size of 300-1000 bp. Tubes were centrifuged to pellet debris and supernatants. Approximately 70 µg of chromatin (1–2 µg/µl) was diluted 10-fold with IP dilution buffer [0.01% SDS, 1% NP40, 1.2 mM EDTA, 16.7 mM Tris-Cl (pH 8.0), and 167 mM NaCl] and precleared twice with 50 µl protein A+G beads (Santa Cruz) containing 3.3 µg salmon sperm DNA for 1 h at 4 C. Supernatants were removed to a new tube and 2.5 µg of polyclonal antibody for Pit-1, GATA-2, or MED220 (same as for CoIP studies), or rabbit IgG was added and tubes rotated overnight at 4 C. To the tube was added 25 µl protein G microbeads (MACS separation, Miltenyi Biotec, Gladbach, Germany) for 1 h with rotation and complexes captured on microcolumns with a magnetic separator. IP chromatin was washed with five 1-ml aliquots of ChIP wash buffer [100 mM Tris-Cl (pH 8), 500 mM LiCl, 1% NP40, 1% deoxycholic acid] and eluted with three 100-µl aliquots of 50 mM NaHCO3 and 1% SDS. Cross-links were reversed by adding 10 µg ribonuclease A and adjusting to 0.3 M NaCl before a 4-h incubation at 65 C, followed by addition of 6 µl 0.5 M EDTA, 6 µl Tris-Cl (pH 6.5), and 6 µl 20 mM proteinase K (Sigma) and incubation at 42 C for 90 min. DNA was purified on QIAGEN (Valencia, CA) minipurification columns and eluted in 60 µl 10 mM Tris-Cl (pH 8), 1 mM EDTA as recommended by the commercial supplier. PCR was performed on 4–8 µl of each sample for 25–31 cycles using Taq Gold polymerase (Applied Biosystems) using oligonucleotide primers for TSHß (–219/–135) sense 5'-AGAAGAGAGGAAGATGCATGCTATAAT-3', antisense sense 5'-TCATACTGAACCCCAAATAAAACTTG-3' with an annealing temperature of 55 C or specific for the coding region of glyceraldehyde 3-phosphate dehydrogenase sense 5'-ATGGTGAAGGTCGGTGTGAACG-3', antisense sense 5'-CCTTCTCCATGGTGGTGAAGAC-3' with an annealing temperature of 53 C.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
First Published Online January 5, 2006
Abbreviations: aa, Amino acid; CAT, chloramphenicol acetyltransferase; CBP, cAMP response element binding protein; ChIP, chromatin immunoprecipitation; CMV, cytomegalovirus; CoIP, coimmunoprecipitation; GST, glutathione-S-transferase; HD, homeodomain; IP, immunoprecipitated; MACS, magnetic bead separation; m, mouse; MED, mediator; NP40, Nonidet P-40; poly(A), polyadenylated; POU, Pit-1, Unc86, Oct-1; SDS, sodium dodecyl sulfate; TRAP, thyroid receptor-associated protein.
Received for publication March 7, 2005. Accepted for publication December 29, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
. Mol Cell 12:1137–1149[CrossRef][Medline]-subunits after thyroxine administration: comparison of mouse hypothyroid pituitary and thyrotropic tumors. Endocrinology 112:2050–2053 and thyrotropin ß-subunit messenger ribonucleic acids by thyroid hormones. Endocrinology 116:873–878 gene by the thyroid hormone receptor. J Biol Chem 274:22345–22353-subunit producing tumor function normally. Mol Cell Endocrinol 71:93–103[CrossRef][Medline]NURSA Molecule Pages Link:
This article has been cited by other articles:
 |
|
 |
|
  S. Ray, D. Dutta, M. A. K. Rumi, L. N. Kent, M. J. Soares, and S. Paul Context-dependent Function of Regulatory Elements and a Switch in Chromatin Occupancy between GATA3 and GATA2 Regulate Gata2 Transcription during Trophoblast Differentiation J. Biol. Chem., February 20, 2009; 284(8): 4978 - 4988. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Magklara and C. L. Smith A Composite Intronic Element Directs Dynamic Binding of the Progesterone Receptor and GATA-2 Mol. Endocrinol., January 1, 2009; 23(1): 61 - 73. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Nagayama, S. Sasaki, A. Matsushita, K. Ohba, H. Iwaki, H. Matsunaga, S. Suzuki, H. Misawa, K. Ishizuka, Y. Oki, et al. Inhibition of GATA2-dependent transactivation of the TSH{beta} gene by ligand-bound estrogen receptor {alpha} J. Endocrinol., October 1, 2008; 199(1): 113 - 125. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Belakavadi, P. K. Pandey, R. Vijayvargia, and J. D. Fondell MED1 Phosphorylation Promotes Its Association with Mediator: Implications for Nuclear Receptor Signaling Mol. Cell. Biol., June 15, 2008; 28(12): 3932 - 3942. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Li, P. Gade, S. C. Nallar, A. Raha, S. K. Roy, S. Karra, J. K. Reddy, S. P. Reddy, and D. V. Kalvakolanu The Med1 Subunit of Transcriptional Mediator Plays a Central Role in Regulating CCAAT/Enhancer-binding Protein-{beta}-driven Transcription in Response to Interferon-{gamma} J. Biol. Chem., May 9, 2008; 283(19): 13077 - 13086. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Viger, S. M. Guittot, M. Anttonen, D. B. Wilson, and M. Heikinheimo Role of the GATA Family of Transcription Factors in Endocrine Development, Function, and Disease Mol. Endocrinol., April 1, 2008; 22(4): 781 - 798. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. K. Kim, S. B. Song, K. I. Kang, M. Rhee, and K. E. Kim Activation of the Thyroid-Stimulating Hormone {beta}-Subunit Gene by LIM Homeodomain Transcription Factor Lhx2 Endocrinology, July 1, 2007; 148(7): 3468 - 3476. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Vijayvargia, M. S. May, and J. D. Fondell A Coregulatory Role for the Mediator Complex in Prostate Cancer Cell Proliferation and Gene Expression Cancer Res., May 1, 2007; 67(9): 4034 - 4041. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Matsushita, S. Sasaki, Y. Kashiwabara, K. Nagayama, K. Ohba, H. Iwaki, H. Misawa, K. Ishizuka, and H. Nakamura Essential Role of GATA2 in the Negative Regulation of Thyrotropin {beta} Gene by Thyroid Hormone and Its Receptors Mol. Endocrinol., April 1, 2007; 21(4): 865 - 884. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Duval, M. D. Jonsen, S. E. Diamond, P. Murapa, A. Jean, and A. Gutierrez-Hartmann Differential Utilization of Transcription Activation Subdomains by Distinct Coactivators Regulates Pit-1 Basal and Ras Responsiveness Mol. Endocrinol., January 1, 2007; 21(1): 172 - 185. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Stumpf, C. Waskow, M. Krotschel, D. van Essen, P. Rodriguez, X. Zhang, B. Guyot, R. G. Roeder, and T. Borggrefe The mediator complex functions as a coactivator for GATA-1 in erythropoiesis via subunit Med1/TRAP220 PNAS, December 5, 2006; 103(49): 18504 - 18509. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
