| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Hormone Research Center (J.-Y.K., H.-J.K., Y.-Y.P., H.-S.C.), School of Biological Sciences and Technology, Chonnam National University, Gwangju 500-757; Department of Biochemistry and Molecular Biology (K.T.K., S.C.P.), The Aging and Apoptosis Research Center, Seoul National University College of Medicine, Seoul 110-799; Department of Biochemistry (H.-A.S., H.H.), School of Life Sciences, Research Center for Bioresource and Health, Chungbuk National University, Cheongju 361-763; Laboratory of Endocrine Cell Biology (K.C.P., M.S.), National Research Laboratory Program, Chungnam National University College of Medicine, Daejeon 301-721; and Department of Internal Medicine (I.-K.L.), Keimyung University School of Medicine, Daegu 700-712, Korea
Address all correspondence and requests for reprints to: Hueng-Sik Choi, Ph.D., Hormone Research Center, Chonnam National University, Gwangju 500-757, Korea. E-mail: hsc{at}chonnam.chonnam.ac.kr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, ß, and 
. In vivo and in vitro protein interaction studies showed that SHP physically interacted with HNF3. Adenovirus-mediated overexpression of SHP significantly decreased the mRNA levels of glucose-6-phosphase (G6Pase), cholesterol 7--hydroxylase (CYP7A1), and phosphoenolpyruvate carboxykinase (PEPCK) in HepG2 cells and rat primary hepatocytes. Moreover, the mRNA level of G6Pase was notably increased by down-regulation of SHP with small interfering RNA. Interestingly, HNF3 transactivity was still repressed by SHP
128–139 that fails to repress nuclear receptors. Mapping of interaction domain revealed that SHP interacted with forkhead DNA binding domain of HNF3. Gel mobility shift and chromatin immunoprecipitation assays demonstrated that SHP inhibits DNA binding of HNF3. These results suggest that SHP is involved in the regulation of G6Pase, CYP7A1, and PEPCK gene expression via novel mechanism of inhibition of HNF3 activity and expand the role of SHP as a coregulator of other family of transcription factors in addition to nuclear receptors. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, androgen receptor, ER-related receptor (ERR), liver receptor homolog-1 (LRH-1), liver X receptor (LXR), glucocorticoid receptor (GR), and pregnane X receptor (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12). Although the precise repression mechanism of SHP still remains unclear, in addition to inhibition of DNA binding (1, 12), dual repression mechanism has been suggested: coactivator competition and recruitment of unknown corepressors through its transrepression domain (5, 6, 9). In addition, a previous report suggested the involvement of a corepressor, E1A-like inhibitor of differentiation 1 (EID-1) in the repression of nuclear receptors by SHP (13). Recently, we demonstrated that EID-1 recruitment to SHP via the loop region between helices H6 and H7 within SHP is essential for the nuclear receptor repression by SHP (14). In addition, SHP has been reported to repress basic helix-loop-helix (bHLH)-PAS transcription factors, aryl hydrocarbon receptor/aryl hydrocarbon receptor nuclear translocator, and a bHLH protein BETA2 by inhibiting DNA binding and by displacing coactivator p300, respectively (15, 16). Therefore, it is hypothesized that the repressive mechanism of SHP can be diverse and applicable to other family of transcription factors as well as nuclear receptors.
SHP is expressed in various tissues, including heart, liver, spleen, adrenal gland, small intestine, and pancreas (4, 8, 16, 17). It has been demonstrated that SHP gene promoter is up-regulated by several nuclear receptors such as FXR (18, 19, 20), ERR (8), LRH-1 (21), SF-1 (21, 22), HNF4 (23), LXR (24), and ER (25). Recently, we reported that a bHLH protein E47 and SREBP-1 are also involved in the activation of human SHP promoter (22, 26). SHP has been suggested to play a pivotal role in the regulation of cholesterol homeostasis via bile acid-activated regulatory cascade in the liver (18, 19, 20, 27, 28). Bile acids repress the expression of cholesterol 7--hydroxylase (CYP7A1) gene, which catalyzes the rate-limiting step in bile acid biosynthesis via feedback inhibition mechanism (18, 19, 20). CYP7A1 gene expression is regulated by a number of factors including LRH-1 and HNF4 whose activities are repressed by SHP (29). Therefore, FXR-mediated induction SHP by bile acids results in the repression of CYP7A1 gene at least partly via repression of LRH-1 and HNF4. Recently, SHP has been shown to repress the phosphoenolpyruvate carboxykinase (PEPCK) gene expression by antagonizing HNF4 and GR activity (10). However, the role of SHP in the hepatic gene regulation by controlling other family of transcription factors still remains to be elucidated.
The HNF3 (also known as Foxa) is a member of the forkhead gene family, which is characterized by the presence of a highly conserved 100-amino acid (aa), monomeric DNA-binding or forkhead domain (reviewed in Ref.30). HNF3 proteins were initially discovered as liver-enriched transcription factors, which bind to and regulate the promoters of transthyretin and 1-antitrypsin genes (31). HNF3 proteins in mammals are encoded by three unlinked genes, designated as HNF3 (Foxa1), HNF3ß (Foxa2), and HNF3 (Foxa3) (32). Their winged helix domains share more than 90% homology in their aa sequence and therefore bind to similar DNA target sequences within hepatocyte-specific regulatory regions and exhibit functional redundancy in hepatocytes. HNF3 proteins play a pivotal role in the regulation of metabolism and the development of foregut endoderm (reviewed in Ref.33, 34, 35). HNF3 binds to cis-regulatory elements in its target genes including gluconeogenic enzymes such as PEPCK and glucose-6-phosphatase (G6pase) in the liver (36, 37, 38). HNF3 is also involved in the regulation of CYP7A1 expression (39, 40, 41). Genetic analysis in mice has shown that HNF3 is required for the full activation of glucagon gene in the pancreas (42, 43). HNF3ß is required for the development of the node and notochord and for visceral endoderm formation (44, 45). HNF3 plays a pivotal role in the activation of gluconeogenic enzymes to prevent hypoglycemia during fasting in the liver (46).
In this study, we identified a forkhead transcription factor HNF3 as a novel target of SHP repression and elucidated the molecular mechanism by which HNF3 transactivation is inhibited by SHP. We showed that SHP repressed the HNF3-mediated transactivation of G6Pase and CYP7A1 genes. SHP inhibits DNA binding of HNF3 via interaction with forkhead domain. Our findings suggest that SHP is implicated in the transcriptional regulation of G6Pase and CYP7A1 genes via inhibition of the HNF3 function.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Transactivation, LXR, LRH-1, GR, and pregnane X receptor, which are also expressed in the liver and regulate transcription of various target genes (6, 9, 10, 11, 12). To identify the hepatic transcriptional factors whose activity is regulated by SHP, we examined the effect of SHP on the transcriptional activity of several hepatic transcription factors by transient transfection assay in HepG2 cells. Of those transcription factors, we found that HNF3 transactivation is regulated by SHP. As shown in Fig. 1A
, SHP strongly repressed the HNF3-dependent activity of the 6X Foxa TATA-luc reporter gene in a dose-dependent manner, which contains six copies of HNF3 binding site from mouse cdx-2 promoter (47). In contrast, SHP had little effect on the HNF1-mediated transactivation of organic anion transporting polypeptide (OATP)-C128/luc, although HNF1 markedly increased the activity of OATP-C128/luc reporter gene as previously reported (48) (Fig. 1B). To examine whether the repression of HNF3 activity is partly due to the reduced protein expression of HNF3 by SHP, we performed Western blot analysis using anti-hemagluttin (HA) antibody. As shown in Fig. 1C, overexpression of HA-SHP had little effect on the HA-HNF3 protein expression. Taken together, these results suggest that SHP repressed the transcriptional activity of forkhead transcription factor HNF3.
|
(Foxa1), HNF3ß (Foxa2), and HNF3 (Foxa3), which share high homology in their forkhead domain. Recently, HNF6 has been shown to coactivate HNF3ß, but not HNF3 or HNF3 transcriptional activity although HNF6 interacts with forkhead domain (47). Therefore, we investigated the effect of SHP on the transcriptional activity of HNF3ß and HNF3. As shown in Fig. 2, A and B, SHP repressed the transcriptional activity of HNF3ß as well as HNF3 in HepG2 cells. Because orphan nuclear receptor DAX-1 is structurally similar to SHP and represses nuclear receptors via different mechanism from that of SHP (14), we tested the effect of DAX-1 on the HNF3 transactivity. Interestingly, DAX-1 had no significant effect on either HNF3 or HNF3ß activity, suggesting that repression target of SHP and DAX-1 may be different, as shown in the repression of nuclear receptor mCAR and SF-1.
|
as previously reported (47) and G6Pase promoter is activated by HNF3 (38), we used G6Pase promoter to test the effect of SHP on the HNF3 transactivation. The G6Pase wild-type (WT) reporter plasmid covers the region from –1227 to +57 bp of human G6Pase promoter and G6Pase insulin response unit (IRU) mutant reporter contains three mutations (49). G6Pase promoter was significantly activated by HNF3, whereas HNF3 had little effect on the activity of IRU mutant reporter, implying that HNF3 transactivates G6Pase promoter through three IRU sequences. As shown in Fig. 2C
, SHP hugely repressed both basal and HNF3-mediated transactivation of G6Pase WT promoter, whereas that of IRU mutant reporter was not significantly affected by SHP. DAX-1 significantly repressed the basal transactivity of IRU mutant reporter, as well as G6Pase WT reporter (data not shown). This observation suggests that DAX-1 regulates other transcription factors involved in transactivation of G6Pase promoter, although because DAX-1 is not expressed in liver, the repression of G6Pase gene promoter by DAX-1 may not be physiologically relevant. This observation indicates that SHP also represses the HNF3 transactivation. Taken together, these results suggest that SHP represses the transcriptional activity of all three isoforms of HNF3.
SHP Physically Interacts with HNF3 in Vivo and in Vitro
To examine whether the HNF3 repression by SHP is mediated through direct protein-protein interaction, we performed in vitro glutathione-S-transferase (GST) pull-down assay. GST alone and GST-SHP proteins were bacterially expressed and 35S-labeled HNF3, HNF3ß, and HNF3 were synthesized by in vitro translation. As shown in Fig. 3A, GST-SHP interacted with all three isoforms of HNF3 family, whereas no significant interactions were observed between HNF3 family and GST alone, demonstrating that SHP directly interacts with HNF3 in vitro. To confirm the in vivo interaction between SHP and HNF3 family, we cotransfected mammalian expression vectors encoding either GST alone or GST-SHP together with expression vectors for HA tagged HNF3, HNF3ß, and HNF3 into 293T cells. As shown in Fig. 3B, top, all three HNF3 isoforms were detected in the coprecipitate only when coexpressed with the GST-SHP but not with the control GST alone. Expression levels of GST-SHP and HA-tagged HNF3 proteins was confirmed by Western blot analysis of crude cell lysates with the antibodies against GST and HA (Fig. 3B, middle and bottom, respectively). This result demonstrates that SHP interacts with HNF3 family in vivo.
|
, HA-HNF3, -HNF3ß, and -HNF3 proteins were localized predominantly in the nucleus even in the presence of SHP. This result suggests that the HNF3 repression by SHP may not be due to the nuclear exclusion of HNF3. Taken together, these results demonstrate that SHP physically interacts with HNF3 in vivo but has no significant effect on the subcellular localization of HNF3.
Effect of SHP on the Expression of HNF3 Target Genes
It has been reported that HNF3 directly binds to and up-regulates G6Pase and CYP7A1 gene promoters (38, 39, 40, 41). To investigate whether SHP can repress the natural target gene promoters of HNF3, transient transfection assays were performed using the reporters containing G6Pase and CYP7A1 gene promoters. As shown in Fig. 4A and B, HNF3, HNF3ß, and HNF3 significantly transactivated both G6Pase and CYP7A1 luciferase reporter genes in HepG2 cells. HNF3 and HNF3ß showed the strongest transactivation of CYP7A1 and G6Pase reporters, respectively. SHP significantly repressed the both basal and HNF3-dependent transactivation of the G6Pase and CYP7A1 genes in a dose-dependent manner. This result demonstrates that SHP represses HNF3-mediated transactivation of G6Pase and CYP7A1 genes.
|
, the mRNA levels of both G6Pase and CYP7A1 were significantly decreased by overexpression of SHP, but not by null adenovirus in HepG2 cells. In addition, SHP obviously reduced the mRNA levels of G6Pase and PEPCK genes in both low- and high-glucose conditions in rat hepatocytes (Fig. 4D). These observations demonstrated that SHP repressed the G6Pase, CYP7A1, and PEPCK gene expression in cultured cells and rat primary hepatocytes.
Because SHP expressed in HepG2 cells, we further confirmed whether G6Pase gene expression is up-regulated by down-regulation of endogenous SHP using small interfering RNA (si-SHP/I or II) into HepG2 cells. si-SHP/II, but not si-SHP/I significantly reduced endogenous SHP expression (Fig. 4E). G6Pase mRNA level was increased by down-regulation of SHP by si-SHP/II, but not by si-SHP/I, suggesting that endogenous SHP can repress the G6Pase expression. Taken together, these results suggest that SHP can inhibit the gene expression of G6Pase, CYP7A1, and PEPCK via repression of HNF3-mediated activation.
H6-H7 Loop Region of SHP Is Not Involved in the Repression of HNF3 by SHP
Recently, we reported that a loop region between helices H6 and H7 within SHP plays an important role in recruitment of EID-1 for the repression of nuclear receptors (14). To elucidate the mechanism underlying the inhibition of HNF3 transactivity by SHP, we examined the effect of SHP128–139, a SHP mutant lacking the loop region, on the HNF3 transactivity. Interestingly, SHP128–139 significantly repressed the HNF3 transactivity to the extent of SHP WT (Fig. 5A). As shown Fig. 5B, however, the deletion mutant failed to repress the transactivity of mCAR, as previously reported (14). This observation suggests that the repression mechanism of HNF3 by SHP can be different from that of nuclear receptors. To test whether HNF3 can interact with SHP WT and SHP128–139, we performed in vitro GST pull-down assay. HNF3 interacted with both WT and SHP128–139 with similar affinity (Fig. 5C). These results suggest that, in contrast to the repression of nuclear receptors by SHP, the loop region between H6-H7 of SHP is not essential for the HNF3 repression and that the recruitment of EID-1 to SHP may not be involved in the HNF3 repression.
|
transactivity. Unexpectedly, W157X, as well as REP failed to repress the HNF3 transactivity (Fig. 5D). This result prompted us to further investigate whether overall structure or the region containing the divided region of SHP is required for HNF3 repression. To address this, we generated N, a deletion construct of SHP, which deletes N-terminal 97 aa and contains nuclear receptor interaction domain (INT) and REP domain (Fig. 5E). As shown in Fig. 5D, N significantly repressed the HNF3 transactivity, suggesting that at least the region containing both INT and REP domains, which also contains the deleted region, but not the overall structure is sufficient for HNF3 repression.
To map the region required for interaction with HNF3, GST pull-down assay was performed using GST-W157X, GST-REP, or GST-HNF3. As shown in Fig. 5F, neither W157X nor REP interacted with 35S-HNF3, whereas 35S-N, as well as 35S-SHP WT, strongly interacted with GST-HNF3, supporting the result from Fig. 5D. These observations suggest that the region containing both INT and REP domains of SHP is necessary and sufficient for interaction with and repression of HNF3. Taken together, these results indicate that the repression mechanism of HNF3 by SHP can be different from that of nuclear receptors and BETA2.
The Forkhead DNA Binding Domain of HNF3 Is Required for the SHP Interaction
A recent report showed that SHP interacts with C-terminal transactivation domain of Foxo1 and represses Foxo1-mediated transactivation via competition with coactivator cAMP response element binding protein-binding protein (CBP) (51). In contrast, no common coactivators for HNF3 family have been characterized yet, except for HNF6, which coactivates HNF3ß but not HNF3 and . To investigate whether SHP can repress the transactivation mediated by C-terminal transactivation domain of HNF3, we employed a mammalian one-hybrid assay using Gal4-HNF3 C2 containing C-terminal transactivation domain of HNF3 (270–466 aa). SHP and p300 alone or both had little effect on the Gal4-HNF3 C2 transactivity (Fig. 6A), whereas SHP and p300 repressed and enhanced Gal4-BETA2 transactivity, respectively (Fig. 6B). SHP also repressed p300-dependent activation of Gal4-BETA2, as previously reported (16). This result demonstrates that C-terminal transactivation domain of HNF3 is not involved in the repression of HNF3 transactivity by SHP.
|
deletion constructs (Fig. 6C) and performed GST pull-down assay using 35S-labeled HNF3 constructs. GST-SHP interacted with HNF3 N1, which encompasses N-terminal transactivation domain and forkhead domain, although the interaction affinity of HNF3 N1 to SHP was slightly reduced compared with that of WT HNF3 (Fig. 6D, upper). However, we could not observe any significant interaction between GST-SHP and HNF3 N2, which possesses N-terminal transactivation domain. This observation indicates that forkhead domain may be essential for SHP interaction. To rule out the possibility that C-terminal domain may also be involved in SHP interaction, we tested the interaction between SHP and C-terminal domain of HNF3. As shown in Fig. 6D, lower, GST-SHP interacted with HNF3 C1, which contains forkhead domain and C-terminal transactivation domain, whereas no interaction was observed between SHP and HNF3 C2 containing C-terminal transactivation domain. In a good agreement with the results from Fig. 6A, this result demonstrates that C-terminal domain is not essential for SHP interaction. Taken together, these results suggest that the forkhead domain of HNF3 is essential for the SHP interaction.
SHP Inhibits DNA Binding of HNF3
Because SHP interacted with forkhead domain of HNF3, we hypothesized that the repression of HNF3 transactivity by SHP might be due to the inhibition of DNA binding of HNF3. To test this hypothesis, we performed gel shift assay using HNF3 and SHP proteins synthesized by in vitro translation and a consensus HNF3 oligonucleotides as a probe. As shown in Fig. 7A, all three isoforms of HNF3 proteins formed distinct sizes of DNA-protein complexes with the consensus HNF3 oligonucleotides. Interestingly, DNA-protein complexes of HNF3 and HNF3 were obviously decreased by addition of SHP, but not by the unprogrammed lysates. SHP also reduced DNA binding of HNF3ß, although the inhibition of DNA binding of HNF3ß by SHP was slightly weak, compared with those of HNF3 and HNF3 by SHP. These results suggest that SHP can inhibit DNA binding of HNF3 in vitro.
|
were transfected into 293 cells together with HA alone or HA-SHP expression vectors. As shown in Fig. 7B, the bands were observed only in FLAG-HNF3 transfected cells but not in FLAG alone transfected cells. Consistent with the result from the gel shift assay, HA-SHP markedly reduced the HNF3 binding to G6Pase promoter.
To further confirm the inhibition of DNA binding of endogenous HNF3 proteins by SHP, we performed ChIP assay after transfection of HA-SHP into HepG2 cells using anti-HNF3 antibody for immunoprecipitation. As shown in Fig. 7C, endogenous HNF3 binds to proximal (–307/+13 bp, lanes 3, 4) but not distal (–1299/–1010 bp, lanes 11 and 12) region of G6Pase promoter and ectopically expressed SHP significantly reduced endogenous HNF3 binding to DNA (lanes 5 and 6), demonstrating that SHP inhibits the DNA binding of HNF3 in vivo. However, neither endogenous nor exogenous SHP bound to G6Pase promoter, as well as control IgG (lanes 7 and 8). Taken together, these results suggest that SHP can repress the HNF3 transactivation via inhibition of DNA binding of HNF3.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, LRH-1, and GR (6, 9, 10). Recently, we reported that SHP represses a bHLH transcription factor BETA2 by competing with a coactivator, P300/CBP (16). However, it has not been fully investigated whether SHP can regulate other family of transcription factors including forkhead transcription factors. In the present study, we identified a forkhead transcription factor HNF3 as a novel target of SHP repression. We demonstrated that SHP repressed HNF3 mediated transactivation of G6Pase and CYP7A1 genes via inhibition of DNA binding of HNF3. Recently, it has been reported that SHP represses the transcriptional activity of Foxo1 (FKHR) (51). The report showed that SHP displaces a coactivator, CBP for the interaction with Foxo1 by interacting with C-terminal domain of Foxo1, which is also responsible for CBP interaction. However, the mechanism underlying the repression of HNF3 seems to be different from that of Foxo1 by SHP. Our results demonstrated that SHP inhibits DNA binding of HNF3. This observation is supported by the domain mapping study showing that SHP directly interacts with a forkhead DNA binding domain of HNF3 but not with C-terminal transactivation domain. Therefore, we suggest that the difference in the repression mechanism of HNF3 and Foxo1 by SHP is at least partly due to the difference in the interaction characteristic.
Because no coactivators for HNF3 family have been characterized yet, except for HNF6 which coactivates HNF3ß but not HNF3 and , we tested the effects of several coactivators including p300/CBP, steroid receptor coactivator-1, and peroxisome proliferator-activated receptor- coactivator (PGC)-1 on the HNF3 mediated transactivity. In contrast to Foxo1, p300/CBP and steroid receptor coactivator-1 had little effect on HNF3-mediated transactivation. Interestingly, PGC-1 coactivated HNF3 transactivity but not those of HNF3 and ß (data not shown). These results suggest that the each member of HNF3 family can be differentially regulated by distinct coactivators. In addition, SHP could not repress the Gal4-HNF3 transactivation mediated by C-terminal transactivation domain, whereas Gal4-Foxo1 C-terminal domain whose activity is reported to be activated and repressed by CBP and SHP, respectively. Therefore, further study is needed to elucidate the preferences between the members of HNF3 to coactivators, and it still remains to be determined whether SHP represses other family of forkhead transcription factors via different mechanism of interaction and repression.
Recently, it has been reported that insulin signaling represses HNF3ß-mediated transcription via HNF3ß nuclear exclusion by phosphatidylinositol 3-kinase and protein kinase B/Akt cascade signaling pathway (50). However, SHP had no significant effect on the nuclear localization of HNF3, suggesting that nuclear exclusion of HNF3 may not be involved in the HNF3 repression by SHP. Previous reports showed that HNF3 promotes gene activation by altering chromatin structure by binding to a linker histone site on the nucleosome core (53). HNF3 also directly binds to core histones H3 and H4 and opens chromatin via C-terminal transactivation domain, which previously was implicated in transcriptional activation (54). However, the C-terminal domain must be targeted to specific genomic sites by the DNA binding domain. Our domain mapping study demonstrated SHP interacts with forkhead DNA binding domain but not with C-terminal transactivation domain of HNF3. We also demonstrated that SHP reduced HNF3 DNA binding in vitro and inhibited HNF3 binding to chromatin structure of G6Pase promoter in vivo. Therefore, we suggest that the inhibition of DNA binding by SHP is an alternative mode of repression of HNF3 transactivation. We also found that adenovirus-mediated overexpression of SHP markedly decreased mRNA level of HNF3ß (data not shown), suggesting that SHP can repress the HNF3-mediated transactivation via repression of HNF3 gene expression as well as repression of its transactivity. We are currently under investigation which factors are involved in the SHP-mediated repression of HNF3 gene expression.
In addition to competition with coactivators for the binding to AF-2 region of nuclear receptors, EID-1 antagonism of CBP/p300-dependent coactivator functions is implicated the inhibition of nuclear receptors by SHP (13). More recently, we reported that the extra amino acids within helices H6 and H7 of SHP form a bulged loop extended from the surface of the protein and the loop region is required for both recruitment of EID-1 and the repressive function of SHP (14). SHP128–139, which has the similar ability to interact with nuclear receptors failed to repress the transcriptional activity of nuclear receptors, HNF4, mCAR, and ERR due to the significantly diminished interaction with EID-1. In contrast, we observed that SHP128–139 still possesses the ability to interact with and repress HNF3 activity. This result suggests that EID-1 recruitment through this region of SHP may not be implicated in HNF3 repression. Moreover, in contrast to nuclear receptors and BETA2, W157X failed to interact with and repress the HNF3 activity, as well as REP domain alone (Fig. 5D). Rather, the region containing both INT and REP, which also contains the divided region, is necessary and sufficient for HNF3 interaction and repression. Therefore, our observations suggest that the repression mechanism of HNF3 by SHP may be different from the repression mechanism of nuclear receptors or BETA2.
The transcriptional regulation of gluconeogenic genes encoding G6Pase and PEPCK, and CYP7A1 involved in bile acid synthesis is controlled by complex interactions of a variety of transcription factors and coregulatory proteins in the liver. Of those transcription factors, nuclear receptors including GR and HNF4 play important roles in controlling G6Pase and PEPCK gene expression (10, 36, 51, 55, 56). CYP7A1 gene transcription is also regulated by GR, HNF4, and LRH-1 whose activities are repressed by SHP (18, 19, 20, 39, 40, 41). In addition to nuclear receptors, forkhead transcription factors including HNF3 and FOXO are also involved in the transcriptional regulation of these genes (36, 37, 38, 39, 40, 41, 49, 51, 57). The coactivator protein PGC-1 has been identified as an important coactivator for nuclear receptors HNF4 and GR, as well as FOXO inducing the expression of PEPCK, G6Pase, and CYP7A1 genes (10, 56, 57, 58, 59). In contrast, SHP seems to counteract with PGC-1 in that SHP represses the CYP7A1, PEPCK, and G6Pase promoters by antagonizes the functions of LRH-1, HNF4 GR, and Foxo1. Here, we demonstrated that both basal and HNF3-mediated transactivation of G6Pase and CYP7A1 promoters are repressed by SHP. We also showed that adenovirus-mediated overexpression decreased the mRNA levels of G6Pase, CYP7A1, and PEPCK in cultured cells and rat primary hepatocytes. Therefore, we suggest that repression of HNF3 by SHP is another mechanism of transcriptional regulation of these genes, in addition to the repression of nuclear receptor signaling. However, to gain better insights into the regulation of G6Pase, PEPCK, and CYP7A1 gene transcription by SHP, a detailed understanding is required about how much portion of transcription of these genes is mediated by HNF3 and how SHP is involved in the synergism between HNF3 and nuclear receptors involved in activation of these genes in vivo.
In summary, we demonstrated that SHP interacts with forkhead domain of HNF3 and inhibits its DNA binding, resulting in the repression of HNF3-dependent transactivation of G6Pase and CYP7A1 genes. Therefore, we suggest that SHP plays a pivotal role in the gluconeogenesis and bile acid metabolism via a novel mechanism of repression of transactivation of HNF3, in addition to the repression of nuclear receptors such as LRH-1, HNF4, and GR. Our findings also support the hypothesis that SHP can serve as a coregulator for wide range of transcription factor family including bHLH and forkhead transcription factors other than nuclear receptors.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
128–139, pGEX4T-1-SHP128–139, pcDNA3/HA-SHP, pcDNA3/HA-W157X, pEBG-SHP, pEGFP-SHP, pGEX4T-1-SHP, pcDNA3-p300, pCMXGal4N and Gal4-BETA2 were described previously (14, 16). HA- or FLAG-tagged rat HNF3 and HNF3ß and murine HNF3 were constructed by inserting PCR fragments of ORF into EcoRI and XhoI digested HA or FLAG epitope-tagged pcDNA3 vector (pcDNA3/HA or pcDNA3/FLAG). Deletion constructs of HNF3 (C1 and C2) and SHP (REP and N) were constructed by inserting PCR fragments into EcoRI/XhoI digested pcDNA3/HA vector. Gal4-HNF3 C2 was subcloned by insertion of EcoRI/XhoI digested fragment from HA-HNF3 C2 into EcoRI/SalI digested pCMXGAL4N vector. GST-W157X, GST-REP, and GST-HNF3 were subcloned by insertion of EcoRI/XhoI digested fragments from pcDNA3/HA-W157X and -REP into pGEX4T-1 vector. All plasmids were confirmed by automatic sequencing analysis.
Cell Culture and Transient Transfection Assay
HepG2, 293T, NIH3T3, and HeLa cells were maintained with DMEM (Invitrogen Life Technologies, Carlsbad, CA), supplemented with 10% fetal bovine serum (FBS; Cambrex Bioscience Walkersville, Inc., Walkersville, MD) and antibiotics (Invitrogen Life Technologies). Cells were split in 24-well plates at densities of 2–8 x 104 cells/well the day before transfection. Transient transfections were performed using the SuperFect transfection reagent (QIAGEN, Valencia, CA) according to the manufacturer’s instruction. Cells were cotransfected with various reporter plasmids together with expression vectors encoding various transcription factors and indicated amounts of pcDNA3/HA-hSHP WT, and mutant. Total DNA used in each transfection was adjusted to 1 µg/well by adding appropriate amount of pcDNA3 empty vector and CMV (cytomegalovirus)-ß-galactosidase plasmids were cotransfected as an internal control. Cells were harvested approximately 40–48 h after the transfection for luciferase and ß-galactosidase assays. The luciferase activity was normalized by ß-galactosidase activity.
Isolation and Culture of Hepatocytes
Hepatocytes were isolated from 4-month Fisher 344 rat livers by double collagenase perfusion, and plated 2 x 106 cells onto collagen precoated 60 mm culture dish with Williams E media Sigma (St. Louis, MO) supplemented with 10% FBS, fungizone (Invitrogen Life Technologies), kanamycin, 2 M dexamethasone as previously described (61). After 4 h incubation, medium were changed to fresh 5.5 mM glucose DMEM with 10% FBS for removing unattached hepatocytes. Primary hepatocytes were infected with adenovirus and shifted to serum-free 5.5 mM or 25 mM glucose media.
Western Blot Analysis
Western blot analysis was performed as previously described (16). Briefly, 293T cells were transfected with 1 µg of pcDNA/HA-HNF3 together with 1–2 µg of pcDNA3/HA-SHP or pcDNA3 empty vectors. Forty-eight hours after transfection, cell lysates were prepared and separated on 12% SDS-polyacrylamide gel, and proteins were transferred to nitrocellulose membrane (Amersham Biosciences, Piscataway, NJ). The membranes were probed with an anti-HA monoclonal antibody (12CA5) (Roche Molecular Biochemicals, Indianapolis, IN) and then developed using the ECL kit (Amersham Bioscience) according to the manufacturer’s instruction.
GST Pull-Down Assay
A GST pull-down assay was performed according to the method described previously (16). Briefly, HNF3, ß, , and HNF3 deletion mutant (HNF3 N1 and N2, HNF3 C1 and C2) were labeled with [35S]methionine using TNT-coupled reticulocyte lysate system (Promega Corp., Madison, WI) according to the manufacturer’s instructions. GST alone and GST fused SHP (GST-SHP) or SHP128–139 (GST-SHP128–139) proteins were prepared as previously described (14, 16). GST fusion proteins were prebound with glutathione-Sepharose beads (Amersham Biosciences) and then incubated with in vitro-translated [35S]methionine-labeled proteins in binding buffer containing 25 mM HEPES (pH 7.6), 150 mM NaCl, 0.2 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and 20% glycerol at 4 C for 2–3 h. The beads were washed three times with the binding buffer, analyzed by SDS-PAGE gel and visualized by a phosphorimage analyzer (BAS-1500, Fuji, Tokyo, Japan).
In Vivo Interaction Assay
293T Cells grown in DMEM supplemented with 10% FBS were plated in six-well flat-bottomed microplates at a concentration of 2 x 105 cells per well the day before transfection as described previously (16). Briefly, 1 µg of each plasmid was transfected into 293T cells with a calcium phosphate precipitation method. Forty-eight hours after transfection, cells were solubilized and the cleared lysates were mixed with 15 µl of glutathione-Sepharose beads and rotated for 2 h at 4 C. The bound proteins were eluted by boiling in SDS sample buffer, subjected to SDS-PAGE, and then transferred to polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA). The membranes were probed with an anti-HA monoclonal antibody (12CA5) (Roche) and then developed using the ECL kit.
Confocal Microscopy
Confocal microscopy was performed as previously described (16). Briefly, NIH3T3 cells were grown on uncoated glass coverslips and transfected with pEGFP-SHP and pCDNA3/HA-HNF3, ß and by the LipofectAmine method (Invitrogen Life Technologies). At 24 h after transfection, cells were fixed with 3.7% formaldehyde for 40 min, mounted on glass slides and observed with a laser-scanning confocal microscope (Olympus Corp., Lake Success, NY). For detection of pCDNA3/HA-HNF3, cells were permealized with 2 ml PBS containing 0.1% Triton X-100 and 0.1 M glycine at room temperature, incubated for 15 min, washed three times with 1x PBS, and blocked with 3% (wt/vol) BSA in PBS for 10 min at room temperature. Cells were incubated with primary anti-HA antibody for 1 h at 37 C, washed three times with 1x PBS, and incubated for 1 h with rhodamine-conjugated antirabbit secondary antibody (Jackson ImmunoResearch Laboratories Inc., West Grove, PA) at 37 C.
Preparation of Recombinant Adenovirus
For the ectopic expression of the human (Fig. 4C) or mouse SHP (Fig. 4D), adenoviral delivery systems were used as described previously (26, 62). Briefly, the cDNA encoding hSHP was cloned into pAd-YC2 shuttle vector. For homologous recombination, pAd-YC2 shuttle vector (5 µg) and a rescue vector, pJM17 (5 µg) were cotransfected into 293 cells. The mouse SHP (mSHP) cDNA was cloned by insertion of BamHI/XbaI digested fragment into BglII/XbaI site of pAdTrack-CMV vector. Recombination of the AdTrack-CMV-mSHP with adenoviral gene carrier vector was performed by transformation to pretransformed adEasy-BJ21 competent cells.
siRNA Expression
The siRNAs for SHP (si-SHP/I and /II) were chemically synthesized (Dharmacon Research, Lafayette, CO), deprotected, annealed, and transfected according to the manufacturer’s instructions. HepG2 cells were transfected with siRNA using Oligofectamine reagent (QIAGEN). Forty-eight hours after transfection, total RNA was isolated for RT-PCR for SHP (30 cycles) and for ß-actin (25 cycles) as a control. The sequences of siRNA are as follows: si-SHP/I, sense 5'-CCTGCCATCCTTCTGGCAGdTdT-3'; si-SHP/II sense 5'-GTCCA TTCCGACCAGCCTGdTdT-3'.
RT-PCR Analyses
HepG2 cells (Fig. 4C) or rat primary hepatocytes (Fig. 4D) were infected with adenovirus containing hSHP (100 pfu/cells) or mSHP (40 pfu/cells), respectively. After viral infection for 24 h, cells were harvested for total RNA isolation using the TRIzol reagent (Invitrogen Life Technologies) according to the manufacturer’s protocol. The mRNAs of SHP, CYP7A1, G6Pase, and PEPCK were analyzed by RT-PCR as described previously (26). Briefly, first-strand cDNA was synthesized from 1 µg of total RNA utilizing an anchored oligo(dT) primers and reverse transcriptase. The resulting first-strand cDNA was then amplified to measure mRNA levels of human/rat (h/r) SHP, h/r G6Pase, hCYP7A1, rPEPCK and h/r ß-actin by 25/27, 30/29, 35, 27 and 25/23 cycles of PCR, respectively, using specific primers. The mRNA levels of ß-actin served as an internal control for the RT-PCR analysis. The primer sequences used for PCR of human SHP, CYP7A1, and ß-actin are previously described (26). The primers used for PCR of h/r G6Pase, rSHP, rPEPCK, and rß-actin are as follows: human G6Pase, forward 5'-GTGAATTACCAAGACTCCCAG-3' and reverse 5'-GCCCATGGCATGGCCAGAGGG-3'; rat G6Pase, forward 5'-GTGGGTCCTGGACACTGACT-3' and reverse 5'-AAAGTGAGCCGCAAGGTA GA-3'; rat SHP, forward 5'-CCTTGGATGTCCTAGGCAAG-3' and reverse 5'-CACCACTGTTGGGTTCCTCT-3'; rat PEPCK, forward 5'-CTGGTTCCGGAAAGACAAAA and reverse 5'-GCTCGGAGCTCCCTCTCTAT; rat ß-actin, forward 5'-CGTGAAAAGATGACCCAGATCATGTT and reverse 5'-GCTCATTGCCGATAGTGATGACCTG-3'. The sizes of PCR products of h/r SHP, h/r G6Pase, hCYP7A1, rPEPCK, and h/r ß-actin were 390/213, 305/274, 350, 297 and 350 or 415 bp, respectively.
Gel Mobility Shift Assays
The double-stranded HNF3 consensus oligonucleotides (38) were labeled by filling-in with [-32P]dCTP using the Klenow fragment of DNA polymerase I. The oligonucleotide sequences are as follows: sense 5'-GGCCTGACCTGTTTGCTTTTCTAC-3' and antisense 5'-GGGTAGAAAAGCAAACAGGTCAGG-3'). HNF3, ß, , and SHP proteins were prepared by in vito translation using a coupled transcription and translation system (TNT-coupled reticulocyte lysate system, Promega). Binding reactions were carried out in binding buffer containing 10 mM Tris (pH 8.0), 40 mM KCl, 0.05% Nonidet P-40, 6% glycerol, 1 mM dithiothreitol, and 1 µg of poly (dI-dC). HNF3 proteins (1 µl) and indicated amounts of SHP or unprogrammed lysates were mixed with 10,000 cpm of labeled HNF3 consensus oligonucleotides in 20 µl of each reaction. After 15 min incubation, DNA protein complexes were analyzed on 5% polyacrylamide gel in 0.5x TBE (90 mM Tris, 90 mM boric acid, 2 mM EDTA). Gels were dried and analyzed by autoradiography.
ChIP Assay
The ChIP assay was performed as described previously (22, 26). Briefly, 293T cells (Fig. 7B) were transfected with 1 µg of pcDNA3/FLAG-HNF3 together with 3 µg of pcDNA3/HA-SHP or pcDNA3/HA empty vectors. At 30 h after transfection, the cells were fixed with 1% formaldehyde and harvested. For immunoprecipitation, anti-FLAG antibodies (M2) (Stratagene, La Jolla, CA) and protein A Sepharose beads CL-4B (Amersham Biosciences) were used. HepG2 cells (Fig. 7C) were transfected with 4 µg of pcDNA3/HA or pcDNA3/HA-SHP. Soluble chromatin was immunoprecipitated with antibodies against HNF3 (N-19) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and SHP (Q-14) (Santa Cruz Biotechnology, Inc.) or nonimmune mouse IgG. The final DNA extractions were amplified by 30 cycles of PCR using two pairs of primers encompassing proximal (–307/+13 bp) or distal (–1299/–1010 bp) region of human G6Pase promoter. The primers used for PCR are as follows; proximal, forward 5'-GATGTAGACTCTGTCCTG-3' and reverse 5'-GATTGCTCTGCTATGAGT-3'; distal, forward 5'-CACTTGGCCAGGCATGGTGG-3' and reverse 5'-TTTGAAACAGGGTCTCACTCT-3'.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: aa, Amino acid; bHLH, basic helix-loop-helix; CBP, cAMP response element binding protein-binding protein; ChIP, chromatin immunoprecipitation; CMV, cytomegalovirus; CYP7A1, cholesterol 7--hydroxylase; EID1, E1A-like inhibitor of differentiation 1; ER, estrogen receptor; ERR, ER-related receptor ; FBS, fetal bovine serum; FXR, farnesoid X receptor; G6Pase, glucose-6-phosphase; GR, glucocorticoid receptor; GST, glutathione-S-transferase; HA, hemagglutinin; HNF4, hepatocyte nuclear factor 4; h/r, human/rat; INT, nuclear receptor interaction domain; IRU, insulin response unit; LRH-1, liver receptor homolog-1; LXR, liver X receptor; mCAR, mouse constitutive androstane receptor; MSHP, mouse SHP; OATP, organic anion transporting polypeptide; PGC, peroxisome proliferator-activated receptor- coactivator; REP, C-terminal transrepression domain of SHP; SF-1, steroidogenic factor 1; SHP, small heterodimer partner; si, small interfering; tk, thymidine kinase; WT, wild-type.
Received for publication May 25, 2004. Accepted for publication August 31, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and ERß. J Biol Chem 274:345–353 and represses its transcriptional activity. Mol Endocrinol 16:2065–2076 function in mice leads to abnormal expression of genes involved in pancreatic islet development and metabolism. Diabetes 50:2472–2480. Mol Endocrinol 17:386–394 regulates expression of the orphan receptor SHP. J Biol Chem 278:36418–36429: a key mediator of the effect of bile acids on gene expression. Hepatology 37:622–631[CrossRef][Medline]
-hydroxylase gene (CYP7A1). J Lipid Res 41:1–111-antitrypsin genes. Mol Cell Biol 9:1415–1425
