| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
State Key Laboratory of Molecular Biology (P.-L.X., S.-F.S., Y.-Y.K., Q.Z., M.L., Y.-H.X., Y.W.) and Key Laboratory of Proteomics (Y.-Q.L., J.-P.D.), Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China
Address all correspondence and requests for reprints to: Yuan Wang or You-Hua Xie or Jian-Ping Ding, Institute of Biochemistry and Cell Biology, 320 Yueyang Road, Shanghai 200031, China. E-mail:wangy{at}sibs.ac.cn (Y.W.), yhxie{at}sibs.ac.cn (Y.-H.X.), jpding{at}sibs.ac.cn (J.-P.D.).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-hydroxylase promoter and increases the transcription of the endogenous cholesterol 7--hydroxylase in Huh7 cells. The interaction between SRC-1 and hLRH-1 assumes a unique pattern that involves primarily a region containing the glutamine-rich domain of SRC-1, and helix 1 and activation function-2 of hLRH-1 LBD. Mutagenesis and molecular modeling studies indicate that, similar to mouse LRH-1, the coactivator-binding cleft of hLRH-1 LBD is not optimized. An interaction between helix 1 of hLRH-1 LBD and a region containing the glutamine-rich domain of SRC-1 can provide an additional stabilizing force and enhances the recruitment of SRC-1. Similar interaction is observed between hLRH-1 and SRC-2/transcriptional intermediary factor 2 or SRC-3/acetyltransferase. Moreover, transcriptional intermediary factor 2 and acetyltransferase also potentiate the transcriptional activity of hLRH-1, suggesting a functional redundancy among SRC family members. These findings collectively demonstrate an important functional role of helix 1 in cofactor recruitment and reveal a novel molecular mechanism of transcriptional regulation and cofactor recruitment mediated by hLRH-1. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
) (2). LRH-1 and SF-1 exhibit distinct yet overlapping expression patterns. SF-1 exists mainly in steroidogenic tissues as a key regulator for the development and differentiated function of the adrenal gland and gonads (3, 4). LRH-1 has been demonstrated to express in liver, pancreas, intestine, ovary, adrenal, and preadipocytes (5, 6, 7, 8), and its biological functions are just being unveiled. LRH-1 acts as an important tissue-specific transcriptional activator in bile acid and cholesterol homeostasis by regulating several key enzymes and transporters (9, 10, 11, 12, 13). Among them, the cholesterol 7--hydroxylase (CYP7A1) gene encodes the first and rate-limiting enzyme in bile acid synthesis, and LRH-1 has been identified as an essential regulator for its transcription. Besides, LRH-1 can modulate the expression of aromatase (8) and several liver-enriched transcriptional factors such as hepatocyte nuclear factor (HNF)1, HNF3ß, and HNF4 (14). In addition to these cellular genes, our previous works have also shown that human (h) LRH-1 can activate hepatitis B virus enhancer II, which in turn regulates hepatic viral gene expression and replication (15, 16).
|
The p160 steroid receptor coactivator (SRC) family contains three homologous members, SRC-1 (nuclear receptor coativator-1), SRC-2 [transcriptional intermediary factor 2 (TIF2), glucocorticoid receptor-interacting protein-1, nuclear receptor coactivator-2], and SRC-3 [acetyltransferase (ACTR), p300/cBP-interacting protein, receptor-associated coactivator-3, amplified in breast cancer-1, thyroid hormone receptor activator module-1] (24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34). These coactivators interact with NRs in a ligand-dependent manner to recruit histone acetyltransferases, histone methyltransferases, coactivator-associated arginine methyltransferase 1, and protein methyltransferase-1 (35, 36, 37, 38) to specific enhancer/promoter regions, which facilitates chromatin remodeling, assembly of general transcription factors, and transcriptional regulation of target genes. SRC-1 plays an important role in SF-1 mediated transcription. A mammalian two-hybrid assay suggests that SF-1 may interact with SRC-1 via AF-2 and a proximal element [amino acids (aa) 187–245] in the hinge region, although a direct interaction between SF-1 and SRC-1 has not been demonstrated (39, 40). These data suggest that the hinge region and other regions of the LBD in addition to AF-2 may possess important regulatory functions for the transcriptional activity of FTZ-F1-related receptors. This notion is further supported by analyzing the activities of two other hLRH-1 variants resulted from alternative splicing (11, 15). The longer variant (hB1F2) that contains 46 additional amino acids in the middle of the A/B domain is also a transcriptional activator. However, the shorter variant (CPFv2) that lacks most of the hinge region and the N-terminal part of the LBD (aa 153–324 of hLRH-1) is inert in transcriptional activation, suggesting that the hinge region and/or the proximal LBD are important for the activity of hLRH-1.
In this paper, we use mutagenesis and reporter gene assay methods to characterize the functional roles of the hinge region and the LBD of hLRH-1. We identified several structural segments in these regions that can greatly affect the transcriptional activity of hLRH-1. In particular, helix 1 of the LBD is shown to be essential for the transcriptional activity of hLRH-1. We also discovered that SRC-1 could potentiate the transcriptional activity of hLRH-1 through a direct interaction with its LBD. This interaction involves primarily a region containing the glutamine-rich domain of SRC-1, and helix 1 and AF-2 of hLRH-1. We demonstrate that overexpressed SRC-1 can potentiate hLRH-1-mediated activation of the liver-specific CYP7A1 promoter and increase the endogenous CYP7A1 transcription in Huh7 cells. Similarly, TIF2 and ACTR also interact directly with hLRH-1 and potentiate the transcriptional activity of hLRH-1. With the availability of the recently reported crystal structure of the LBD of mouse LRH-1 (mLRH-1) (41), we have built a three-dimensional homology model of hLRH-1 LBD and discuss the biochemical and mutagenesis data and the potential functional roles of these structural segments in the context of the model. These findings illustrate a unique property of LRH-1 in cofactor recruitment and demonstrate a novel interaction mode between members of the SRC family and an NR.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). The transcriptional activities of these fusion proteins were examined in Huh7 cells with reporter assays. As shown in Fig. 1B, GAL4-hLRH-1141–495 containing the complete hinge region and the LBD was fully active. Truncation to residue 219 partially impaired the activity. However, truncation to residue 228 caused a dramatic increase in the transcriptional activity, suggesting a putative strong repression element between aa 219 and 227. Further truncations to residues 244, 256, and 259 resulted in moderate decrease of activities. Because residue 256 is the putative boundary between the hinge region and the LBD, these results indicate that hLRH-1 LBD alone is active. However, truncation to residue 262 not only completely abolished the transcriptional activity but also showed a repression. Further truncation to residue 325, which is the C terminus of the internal deletion in the CPFv2 variant, displayed a similar repression. Moreover, GAL4-hLRH-1141–481, in which activation function-2 (AF-2) was partially deleted, also lost the transcriptional activity and behaved as a repressor.
Western blot assays with the whole cell lysates were performed to examine the expression of these fusion proteins. Detected with an anti-GAL4 monoclonal antibody, the expression levels of all fusion proteins are comparable (Fig. 1C). Therefore, the differences in the transcriptional activities of the fusion proteins are not likely due to the slightly varying protein expression levels or protein stability; instead, reflect an intrinsic property of hLRH-1. Although bands of lower molecular weight impurities were observed in some lanes, they should not alter the result interpretation. Collectively, the results suggest that hLRH-1 contains a structural element between aa 219 and 227 of the hinge region that can repress the transcriptional activity of hLRH-1 and a crucial structural element at the N terminus of the LBD that plays an important role in the transcriptional activity of hLRH-1.
Helix 1 of the LBD Is Essential for the Activity of hLRH-1
The reporter assays reveal that the structural segment containing aa 259–261 at the N terminus of the LBD is crucial for the transcriptional activity of hLRH-1. Secondary structure prediction and homology modeling studies show that this segment belongs to an -helix (aa 257–264) at the N terminus of the LBD (Figs. 2A and 3). This helix (helix 1) forms part of the conserved sandwich framework structure found in other nuclear receptors (42). The hydrophobic side of helix 1 packs with helices 8 and 9 via extensive hydrophobic interactions and its hydrophilic side is exposed to the solvent.
|
|
). As shown in Fig. 2C, the TSS251AAA mutant, in which three amino acids (T249S250S251) located at the linker from the hinge region to helix 1 of the LBD were altered to alanines, retained a transcriptional activity similar to wild-type GAL4-hLRH-1141–495. In contrast, mutations within helix 1, specifically E261G, IL260AA, and LL263AA, caused moderate to dramatic decrease of activities. Similar results were obtained from the transfection experiments using a luciferase reporter driven by a natural hLRH-1-responsive promoter, the enhancer II and core promoter of hepatitis B virus (HBV ENII/Cp). In Huh7 cells, the exogenous expression of wild-type hLRH-1 (0.5 µg) activated the reporter by about 7.5-fold compared with the empty vector. The TSS251AAA mutant exhibited a transcriptional activity similar to wild-type hLRH-1. A moderate activation was observed with the E261G mutant (4.6-fold). However, the IL260AA mutant only activated the reporter by 1.8-fold and the LL263AA mutant barely showed any transcriptional activity. In both reporter gene assays, the AF-2 partial deletion mutant of hLRH-1 (AF-2del) was inert in transactivation. Taken together, the reporter assays clearly indicate that helix 1 (aa 257–264) is essential for the transcriptional activity of hLRH-1.
The functional roles of these residues and the effects of these mutations on the transcriptional activity of hLRH-1 were examined in the context of the three-dimensional model of hLRH-1 LBD (Fig. 3). Modeling studies show that I259 and L262 point their side chains inward toward the hydrophobic core and form extensive hydrophobic interactions with residues of helices 8 and 9. E261 points its side chain outward on the surface and forms a salt bridge with the positively charged side chain of K264. These two charged residues form an electrostatic patch on the LBD surface and could take part in interactions with a potential ligand or coactivator. Mutation of E261G would disrupt its interaction with K264 and is likely to affect its interaction with a ligand or coactivator. The side chains of L260 and L263 are partially exposed on the surface and form a hydrophobic patch on the protein surface. Specifically, the side chain of L260 has hydrophobic contacts with the side chain of I255 at the N terminus of LBD; the side chain of L263 has contacts with the aliphatic side chains of G399 and R400 of helix 8. Molecular dynamics simulations indicate that all of these mutations do not affect the main-chain conformation of helix 1 and the overall sandwich framework of NR structure. However, the mutations IL260AA or LL263AA could affect the hydrophobic interactions with residues of helices 8 and 9 and consequently the precise position of helix 1. Thus, these mutations would affect the interactions of hLRH-1 LBD with a ligand or coactivator and the transcriptional activity of hLRH-1.
Coactivator SRC-1 Potentiates the Transcriptional Activity of hLRH-1
It is well documented that several transcriptional cofactors, including CBP, p300, SRC-1, and silencing mediator of retinoic acid and thyroid hormone receptors (SMRT), are often involved in the coregulation of NRs (17). We carried out cotransfection and reporter assays in Huh7 cells to examine whether any of these cofactors can affect the transcriptional activity of hLRH-1. The results show that SRC-1 can potentiate the activity of GAL4-hLRH-1141–495 by about 2.7-fold; SMRT can repress it by about 3-fold, whereas CBP and p300 exhibit barely any effects (Fig. 4A). Similar results were obtained with reporter assays using pENII/CpLuc. SRC-1 can stimulate the reporter by 2.7-fold compared with hLRH-1 alone, whereas SMRT can repress it by 3.5-fold. CBP and p300 do not show any obvious effects (Fig. 4A).
|
). A Western blot was performed to rule out the possibility that the coexpression of SRC-1 might increase the expression of GAL4-hLRH-1141–495, which could also enhance the reporter gene transcription. The results demonstrate that SRC-1 does not stimulate the expression of GAL4-hLRH-1141–495, whereas the transfection with an 8-fold GAL4-hLRH-1141–495 expression plasmid does induce an increased expression (Fig. 4C). Therefore, the coactivation of hLRH-1 by SRC-1 is due to a functional correlation between these two factors rather than through other secondary effects.
Multiple transcriptional factors can interact with the ENII/Cp region (43). To exclude the possibility that SRC-1 might activate the reporter pENII/CpLuc through other cellular factors instead of hLRH-1, a mutant reporter pENIIm/CpLuc was made, in which the critical hLRH-1 binding site in the ENII B1 element was mutated so that ENII is inert to the activation by hLRH-1 (15, 16). As expected, an overexpression of hLRH-1 could hardly activate this reporter. Coexpression of SRC-1 could not further stimulate the reporter (Fig. 4D), suggesting that the coactivation by SRC-1 is primarily via hLRH-1. Cotransfection assay results performed in HeLa cells that lack endogenous hLRH-1 support the above suggestion. Despite a weak activation by hLRH-1 on the ENII/Cp in HeLa cells, coexpression of SRC-1 stimulates the reporter activity only in the presence of hLRH-1 (data not shown).
SRC-1 Physically Interacts with hLRH-1
The functional correlation between hLRH-1 and SRC-1 implicates that there might be a physical interaction between the two factors. To explore this possibility, a mammalian cell two-hybrid assay was firstly performed in Huh7 cells. A considerable activation of the reporter was observed by the coexpression (Fig. 5A, left panel) of GAL4-hLRH-1141–495 and VP16-SRC-1, in which the full-length SRC-1 was fused with the activation domain of VP16 (VP16-AD). A similar activation was observed in a reciprocal two-hybrid assay, in which the full-length SRC-1 was fused with the GAL4-DBD and the full-length hLRH-1 fused with the VP16-AD (Fig. 5A, right panel). Similar results were obtained from experiments in HeLa cells (data not shown). These data suggest an interaction between SRC-1 and hLRH-1 in mammalian cells.
|
), we used the GST-hLRH-1186–495 fusion protein in the pull-down assays. As shown in Fig. 5B, the immobilized GST-hLRH-1186–495 could interact specifically with a 35[S]-labeled in vitro translated full-length SRC-1. As a positive control, the GST-retinoic acid receptor (RAR) fusion protein can interact strongly with the labeled SRC-1 only in the presence of its specific ligand all-trans retinoic acid (ATRA). No interaction was observed between GST alone and SRC-1. These results clearly suggest that SRC-1 interacts directly with hLRH-1 and that the interaction might occur in the absence of a ligand or any posttranslational modifications of hLRH-1.
The Region Containing the Q-Rich Domain of SRC-1 Is Primarily Involved in Its Interaction with hLRH-1
To characterize the region of SRC-1 involved in its interaction with hLRH-1, several truncated SRC-1 were fused to GST. The truncated SRC-1 were designed to contain specific regions that have been reported to be important in protein-protein interaction and transcriptional regulation, including the basic-helix-loop-helix/Per/ARNT/Sim (bHLH-PAS) domain, nuclear receptor-interacting domain (NID), CBP/P300-interacting domain (AD1), Q-rich domain, and activation domain 2 (AD2) (44) (Fig. 6A). Pull-down assays were performed with the 35[S]-labeled in vitro translated full-length hLRH-1. As shown in Fig. 6B, strong interactions exist between hLRH-1 and GST-SRC-11029–1255 or GST-SRC-11255–1441. The interactions between hLRH-1 and GST-SRC-1568–780 or GST-SRC-11321–1441 are much weaker. Other GST-SRC-1 fusion proteins, as well as GST alone, do not display any interactions with hLRH-1.
|
). Marked activations of the reporters were observed with the full-length, aa 1029–1255, and aa 1255–1441 of SRC-1, whereas slight activations were detected with aa 568–780 and aa 1321–1441. No activation was observed with other truncated SRC-1 proteins. These results are consistent with the data from the pull-down experiments.
The NID of SRC-1 is implicated in mediating the interaction between SRC-1 and nuclear receptors through its internal LXXLL motifs (45). However, our data show that there is only a weak interaction between the NID of SRC-1 (aa 568–780) and hLRH-1. To further assess the role of the NID in the interaction, a full-length SRC-1 containing a mutant NID (NIDm) was made, in which all three LXXLL motifs in the NID were replaced by LXXAA (Fig. 6D). As shown in Fig. 6E, the interaction between the NIDm and hLRH-1 remains strong. In contrast, the interaction between the NIDm and RAR is significantly impaired, consistent with the biochemical data that the NID is required for the interaction with RAR (35). These results demonstrate that the NID of SRC-1 plays a minor role for its interaction with hLRH-1.
Cotransfections with the expression plasmids of GAL4-hLRH-1141–495 and mutant SRC-1 support the pull-down assay results. As shown in Fig. 6F, SRC-1 containing the mutant NID can still coactivate hLRH-1, though to a somewhat lesser extent compared with the wild-type SRC-1. Both SRC-11–1321 and SRC-11–1255 exhibit similar coactivation capabilities as the wild-type SRC-1, whereas SRC-11–1029 and SRC-1d1029–1255 do not stimulate hLRH-1. Similar results were obtained from the cotransfection assays with the pENII/CpLuc reporter. Moreover, GST pull-down assay using SRC-1d1029–1255 indicates that the aa 1029–1255 deletion abrogates the binding of SRC-1 to GST-hLRH-1186–495, whereas it does not affect the binding of SRC-1 to GST-RAR (Fig. 6E). Collectively, SRC-1 interacts directly with hLRH-1 primarily via the C-terminal region aa 1029–1255 that contains the Q-rich domain.
Both Helix 1 and AF-2 of hLRH-1 Are Important for the Recruitment of SRC-1
Both AF-2 and helix 1 of hLRH-1 are essential for its transcriptional activity. Whether these structural elements are also required for the recruitment of SRC-1 was investigated by mammalian two-hybrid assays. Several GAL4-hLRH-1 fusion proteins (as in Fig. 1B) were examined for their potential interactions with VP16-SRC-1. Obvious interactions were observed with aa 141–495, aa 219–495, and aa 256–495 of hLRH-1. In contrast, no interaction was observed with those truncated hLRH-1s in which either helix 1 (aa 262–495) or AF-2 (aa 141–481) was disrupted or partially deleted (Fig. 7A).
|
). Compared with wild-type hLRH-1, the E261G mutant displays a weaker interaction with GST-SRC-11029–1255, whereas other mutants, namely IL260AA, LL263AA, and AF-2del, show much more severe impairment in the interaction (Fig. 7B). The data are well consistent with the reporter assay results shown in Fig. 2C, indicating that the more severe the activity is affected by the mutations, the weaker the interaction is between hLRH-1 and SRC-1. Moreover, based on the weak interactions between SRC-1 and the hLRH-1 mutants containing point mutations in helix 1, it can be deduced that the interaction between AF-2 and SRC-1 is weak. Collectively, these data indicate that both helix 1 and AF-2 of hLRH-1 play important roles in the recruitment of SRC-1.
Interestingly, overexpression of SRC-1 could dramatically enhance the transcriptional activities of those hLRH-1 mutants that contain point mutations of the hydrophobic residues in helix 1 (Fig. 7C). The GAL4-hLRH-1141–495 containing IL260AA or LL263AA mutations have very weak transcriptional activities in Huh7 cells. However, if SRC-1 was coexpressed, the activities of these mutants could increase dramatically by about 12- and 16-fold, respectively. Although the coexpression of SRC-1 did not rescue the transcriptional activity of the AF-2 partial deletion mutant, it seems that the overexpression of SRC-1 can partially compensate the impaired interactions between SRC-1 and hLRH-1 mutants that contain point mutations of the hydrophobic residues in helix 1.
SRC-1 Potentiates hLRH-1-Mediated Activation of the CYP7A1 Promoter
To further demonstrate the physiological relevance of SRC-1 for hLRH-1 signaling, we investigate whether SRC-1 can potentiate hLRH-1 mediated activation of the liver-specific CYP7A1 promoter and the transcription of the endogenous CYP7A1 gene in hepatic cells. A consensus hLRH-1 binding site is located in the core promoter of hCYP7A1 gene (11). A reporter assay was first performed as previously done with the ENII/Cp promoter. As shown in Fig. 8A, hLRH-1 and SRC-1 activated a CYP7A1 promoter reporter about 2.9- and 1.6-fold, respectively. However, when coexpressed with hLRH-1, SRC-1 enhanced the reporter activity up to 8-fold, indicating that SRC-1 can potentiate hLRH-1-dependent activation of the CYP7A1 promoter. To investigate whether overexpression of SRC-1 has an impact on the transcription of the endogenous CYP7A1 gene, semiquantitative RT-PCR assays were performed. As shown in Fig. 8B, overexpression of hLRH-1 or SRC-1 increased the transcription of the endogenous CYP7A1 slightly in Huh7 cells. When coexpressed with hLRH-1, SRC-1 could increase the transcription of the endogenous CYP7A1 up to 60%. Taken together, SRC-1 can potentiate hLRH-1-mediated activation of the transcription of the CYP7A1 gene.
|
). It is thus interesting to investigate whether TIF2 and ACTR can also interact with hLRH-1 and potentiate its transcriptional activity in a similar manner as SRC-1. Cotransfection analysis was performed in Huh7 cells. The activity of GAL4-hLRH-1141–495 was enhanced by all three SRCs with TIF2 showing a dramatic potentiation (up to 13.5-fold) (Fig. 9B). Similar results were obtained using the ENII/Cp reporter (data not shown). Potentiation of hLRH-1 activity by all three SRCs was also observed in cotransfection assays with the CYP7A1 promoter reporter (Fig. 9C). However, with the CYP7A1 promoter, TIF2 did not show a dramatic potentiation, which was different from the results performed with the GAL4 based reporter and the pENII/CpLuc reporter. Thus, the relative strength of potentiation by different SRCs appears to depend on the promoter used.
|
), suggesting that CBP might be recruited by SRC-1 or ACTR. Interestingly, CBP failed to enhance TIF2 mediated coactivation of hLRH-1.
To investigate a possible physical interaction between hLRH-1 and TIF2 or ACTR, GST pull-down assays were performed using GST-hLRH-1186–495 and in vitro-translated TIF2 or ACTR. Similar to SRC-1, both TIF2 and ACTR interact with hLRH-1 directly (Fig. 9D). Subsequently, the interaction regions between hLRH-1 and TIF2 or ACTR were investigated with mammalian two-hybrid and GST pull-down assays. Because C-terminal domains (AD1, Q-rich, and AD2) are less conserved in sequence among SRCs (46), we tethered the part beginning from the Q-rich region to the C-terminal of TIF2 or ACTR with GAL4-DBD. As shown in Fig. 9E, the C-terminal part of TIF2 or ACTR interacts with hLRH-1 in Huh7 cells. The interaction is abrogated when helix 1 or AF-2 of hLRH-1 LBD was deleted. GST pull-down assays also revealed that NID of TIF2 or ACTR interacts with hLRH-1 weakly, although it interacts with ligand-bound RAR strongly (Fig. 9F). Therefore, the interaction between hLRH-1 and TIF2 or ACTR is similar to that between hLRH-1 and SRC-1. Take together, these results suggest a functional redundancy among members of SRC family in the potentiation of the transcriptional activity of hLRH-1.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Many coregulators have been documented to potentiate or repress the activities of NRs through direct or indirect interactions (17). Several factors including Dax-1, CBP/p300, TReP-132, DP103 (19, 20, 21, 22), and SRC-1 (39, 40) have been implicated in modulating the transcriptional activity of SF-1. Small heterodimer partner has been found to interact with AF-2 of mLRH-1 and represses its transcriptional activity in bile acid synthesis (53). Here we demonstrate that SRC-1 as well as its homologous coactivators TIF2 and ACTR interact directly with hLRH-1 and potentiate the transcriptional activity of hLRH-1 on GAL4 responsive reporter, HBV ENII/Cp, and the CYP7A1 promoter. Moreover, overexpressed SRC-1 enhances the endogenous transcription of the CYP7A1 gene in Huh7 cells, although not as significantly as it does on the CYP7A1 promoter reporter, which probably due to a much tighter regulation of the endogenous CYP7A1 expression in cells. Because the purified E. coli-expressed hLRH-1 is capable of interacting with the in vitro translated SRCs, it seems that the interaction can occur in the absence of a potential ligand. SRC-1 has also been shown to coactivate SF-1. Nevertheless, a direct interaction between SF-1 and SRC-1 has not been clearly demonstrated yet (39, 40). For SF-1, the phosphorylation of residue S203 in the hinge region by the MAPK pathway maximizes the SF-1 mediated transcription and its interaction with cofactors (23). Because the hinge region is the least conserved region (<25% sequence identity) between SF-1 and hLRH-1, whether a similar phosphorylation mechanism exists for hLRH-1 awaits future investigation.
SRC-1 contains several functional domains, including a basic bHLH-PAS domain, a central nuclear receptor interacting domain (NID), a loosely defined Q-rich domain, and two activation domains (44). The NID features three LXXLL motifs (NR box) that have been demonstrated to be necessary and sufficient for mediating the direct interaction between SRC-1 and several ligand-bound NRs, such as RAR and estrogen receptor (ER) (35, 45). This appears not the case for hLRH-1. Our data show that strong physical and functional interactions are mainly observed between hLRH-1 and the aa 1029–1255 region of SRC-1 that contains the Q-rich domain. Deletion of this region results in complete loss of the interaction and thus the coactivation of hLRH-1. In contrast, the NID of SRC-1 (aa 568–780) is shown to have only a weak interaction with hLRH-1 in cells and in vitro. Mutations of the LXXLL motifs in the NID do not obviously alter the interaction between SRC-1 and hLRH-1 and have only a minor effect on SRC-1 coactivation. Therefore, the NID of SRC-1 plays a minor role in its interaction and coactivation of hLRH-1. The Q-rich domain of SRC-1 can mediate the ligand-independent activity of the androgen receptor (54, 55). However, in that case, it is the N-terminal AF-1 domain of the androgen receptor that is involved in interaction and potentiation. Therefore, this is the first report that SRC-1 can interact with the LBD of an NR by means of a region containing the Q-rich domain. On the other hand, aa 1–1255 of SRC-1 that does not contain AD2 is fully active in potentiating hLRH-1, suggesting that AD2 is not necessary for the coactivation of hLRH-1. However, it is noteworthy that AD2 (as in aa 1255–1441) of SRC-1 can interact with hLRH-1 under our experimental conditions. The same region was previously shown to have interactions with SF-1 in a mammalian two-hybrid assay (39, 40). Therefore, we cannot exclude the possibility that AD2 is also required in specific cell types or for other hLRH-1-dependent promoters because it has been reported that the function of AD2 may depend on the context of cell types and promoters (38, 44).
Structural and biochemical studies of mLRH-1 LBD suggest that the cofactor recruitment by the conserved core structure is not optimized and the binding of coregulators to mLRH-1 is relatively weaker than to other NRs (41). To enhance cofactor recruitment, other structural modules may be required to provide additional stabilizing forces. Indeed, our data show that in addition to AF-2 that is usually involved in coactivator recruitment, helix 1 of hLRH-1 LBD is also essential for the transcriptional activity of hLRH-1 and plays an important role in the recruitment of SRC-1. Point mutations in helix 1 would impair its interaction with SRC-1 and reduce the transcriptional activity of hLRH-1, suggesting that AF-2 of hLRH-1 has a relatively weak interaction with SRC-1. Structural comparisons of several NRs [RAR-related orphan receptor, retinoid X receptor (RXR), peroxisome proliferator-activated receptor (PPAR), ER, and RAR] and the structural model of hLRH-1 indicate that the sandwich framework of NR structures is well conserved. Helix 1 is located at the N terminus of the LBD, is exposed on the outer surface, and forms part of the sandwich framework. The only structural element that has a great flexibility is the region that comprises helix 2. Docking experiments based on the crystal structure of several NR LBDs in complexes with SRC-1 or SMRT peptides containing the conserved LXXLL motif indicate that both SRC-1 and SMRT peptides can bind to hLRH-1 LBD in the surface cleft formed by helices 3 and 4 and AF-2 (Figs. 3 and 10). However, as in mLRH-1, the residues that compose the coactivator-binding cleft in hLRH-1 comprise four residues (R315, M329, Q398, and N484) different from those in other receptors. R315 forms part of the so-called "electrostatic clamp" that stabilizes the receptor-coactivator complex (56). The corresponding residue in other receptors is lysine. In the structural model of hLRH-1 LBD with docked SRC-1 peptide, the side chain of R315 has steric conflict with one of the conserved leucine residues of the SRC-1 peptide. M329 and N484 have relatively large side chains compared with the usually smaller residues (A/S/T/V/P) at these positions in other receptors and appear to also have close contacts with the docked SRC-1 peptide. Mutations of these two residues in mLRH-1 to the smaller residues in RXR (M394V/N549T) increase significantly the activation of mutant mLRH-1 by other NR coregulators (41). Therefore, the amino acid differences at these positions appear to be primarily responsible for the weak interactions between SRC-1 and AF-2 of mouse and hLRH-1. On the other hand, the modeling studies show that the C terminus of SMRT fragment runs across helix 3 and extends further to reach helix 1 and has the potential to interact with the ILELL segment of helix 1. It is conceivable that SRC-1 may also extend along the same path and has interactions with helix 1 of LBD during coactivation.
|
Three homologous members of the SRC family have been identified in human and rodents, which produce proteins about 160 kDa in size and have an overall sequence similarity of 50–55% among them. They share high homology in functional domain, including bHLH-PAS, NID, AD1, AD2, and Q-rich domain. All three SRCs have been detected by Northern blot in many tissues including liver and pancreas where hLRH-1 expression is enriched. Protein-protein interaction assays have indicated that SRCs share many common features that permit interaction with and coactivation of NRs (45, 56, 57). Nevertheless, functional specificity among SRCs is still apparent based on the studies with SRC knockout (KO) mice. For example, partial resistance to steroid and thyroid hormones was observed in SRC-1 KO mice, whereas development of reproductive functions seems impaired in TIF2 and ACTR KO mice (49, 58, 59). It is likely that selective recruitment of SRCs by different NRs may mediate specific transcription signals. On the other hand, the viable phenotypes of SRC KO mice suggest that SRC family members may be able to partially compensate each other’s function in vivo (49, 50, 60). The expression of known LRH-1 target genes has so far not been examined in SRC KO mice. In this study, all three SRCs share similar unique features in the interaction with hLRH-1, suggesting a functional redundancy among SRC family members in the potentiation of the transcriptional activity of hLRH-1. As a relatively new receptor, the number of known target genes of LRH-1 is still limited. Thus, it is also possible that some of the reported phenotypes with SRC KO mice may be directly or indirectly related to LRH-1 signaling. Identification of more LRH-1 target genes and careful examination on the expression of LRH-1 target genes in SRC KO mice may provide important information on the in vivo functional interaction between LRH-1 and SRCs.
Interestingly, in our cotransfection analyses, TIF2 shows a relatively greater potential in coactivating hLRH-1 than SRC-1 and ACTR on the pG5Luc and pENII/CpLuc reporters, but not on the CYP7A1 promoter reporter. Coexpression of CBP and SRC-1 or ACTR shows an elevated hLRH-1 activity, whereas this is not the case in the coexpression of CBP and TIF2, which is also observed in COS cells (data not shown). Because no obvious difference exists in the interaction with hLRH-1 among three SRCs in our studies, the different coactivation of hLRH-1 by SRCs on different promoters may be caused by selective recruitment of secondary coactivators by SRCs (61, 62). Therefore, the temporal and spatial expression of SRCs may partially determine the transcriptional activation and functions of hLRH-1 in diverse biological processes such as development and metabolism. It will be interesting to examine more SRC-related coactivators on different hLRH-1 regulated promoters.
In contrast to the well-defined molecular events that trigger the ligand-dependent activation of hormone NR, the understanding is still limited in how the transcriptional activities of the constitutively active orphan NRs are regulated. In this study, we demonstrate that SRCs interact directly with hLRH-1 in a novel pattern and potentiates its transcriptional activity. Our findings also illustrate the unique features of hLRH-1 responsible for cofactor recruitment.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The pENII/CpLuc reporter was made by cloning the enhancer II and the core promoter (ENII/Cp) region of HBV subtype adr1 (15) in the pGL2basic vector (Promega, Madison, WI). Point mutations that eliminate the hLRH-1 binding site in the ENII B1 element (5'-GATCAACtACaGAtCTcGAG-3', mutations in lowercase letters) were introduced by overlapping PCR as described previously (15). The resulting reporter plasmid was designated pENIIm/CpLuc. The CYP7A1 promoter reporter was made by inserting the –911 to 25 region of hCYP7A1 promoter (11) (the transcription start site is designated 1) amplified by PCR into the SmaI/XhoI sites of pGL2basic. The GAL4-dependent pG5Luc reporter was made by replacing the chloramphenicol acetyltransferase reporter gene in the pG5CAT (CLONTECH, Palo Alto, CA) with the luciferase gene.
The complete and partial AF2-deleted (aa 1–481) hLRH-1 cDNA (15) were cloned in the pcDNA3 vector (Invitrogen Life Technologies, Carlsbad, CA), respectively, resulting in pcDNA3-hLRH-1 and pAF2-del. After digestion with EcoRI and XbaI, the complete hLRH-1 cDNA was released from the pcDNA3-hLRH-1 and inserted in the pVP16 vector (CLONTECH) to get pVP16-hLRH-1.
5'-Deletions of the hinge region and the proximal LBD of hLRH-1 were amplified and cloned in the pM vector (CLONTECH), in frame to the GAL4-DBD. The resulting constructs represented fragments aa 141–495, aa 186–495, aa 219–495, aa 228–495, aa 236–495, aa 244–495, aa 256–495, aa 259–495, aa 262–495, aa 325–495, and aa 141–481 of hLRH-1, respectively.
Human SRC-1 cDNA cloned in the pCR3.1 vector (pCR3.1-hSRC-1) was kindly provided by Dr. M. J. Tsai (Dallas, TX). Original expression plasmids of hTIF2 (TIF2/pSG5) and human ACTR (pCMX-ACTR) were generous gifts from Dr. P. Chambon (Strasburg, France) and Dr. R. M. Evans (La Jolla, CA), respectively. The full-length SRC-1 cDNA was amplified by PCR. After digestion with XbaI, it was subcloned downstream from the VP16 activation domain in the blunted EcoRI and XbaI sites of the pVP16 vector to generate pVP16-SRC-1. Fragments of SRC-1 were also cloned in the pVP16 vector. The resulting constructs represented fragments aa 1–568, aa 568–780, aa 780-1029, aa 1029–1255, aa 1255–1441, aa 1321–1441, and aa 1029–1441 of SRC-1, respectively. C-terminal fragments of TIF2 and ACTR cDNAs were also inserted in the pVP16 vector. The resulting constructs represented aa 1165–1464 of TIF2 and aa 1142–1412 of ACTR. To make pGAL4-SRC-1, the full-length SRC-1 cDNA in the pVP16-SRC-1 was digested with SalI and XbaI and inserted in the same sites of the pM vector. Truncated SRC-1 fragments representing aa 1–1029, aa 1–1255, and aa 1–1321 were amplified by PCR. These fragments were digested with KpnI and XbaI and inserted in the same sites of the pCR3.1 vector.
To introduce an internal deletion of aa 1029–1255 into SRC-1, two SRC-1 fragments (aa 1–1028 and aa 1256–1441) were amplified by PCR. After digestion with KpnI and BglII, BglII and XbaI, respectively, these two fragments were ligated in the KpnI and XbaI sites of the pCR3.1 vector. Point mutations that change all three LXXLL motifs to LXXAA in the NID of SRC-1 were introduced by overlapping PCR-mediated mutagenesis. After digestion with KpnI and BamHI, the final fragment replaced the KpnI-BamHI fragment in pCR3.1-hSRC-1.
The pRC/RSV-mCBP-HA-RK and pCMVß-p300-CHA expression plasmids were kindly provided by Dr. R. H. Goodman (Portland, OR). The pCMV-mSMRT-FL expression plasmid was a generous gift from Dr. R. M. Evans (La Jolla, CA).
For expression of GST-hLRH-1186–495 fusion protein, the aa 186–495 fragment of hLRH-1 amplified by PCR was digested with EcoRI and inserted in the SmaI and EcoRI sites of the pGEX-3X vector (Amershan Pharmacia Biotech, Uppsala, Sweden). For expression of GST-RAR, the ScaI fragment of plasmid pSG5-hRAR (gift from Dr. Z. Chen, Shanghai, China) containing the hinge region and the LBD of hRAR (aa 143–462) was subcloned in the blunted EcoRI site of the pGEX-3X vector. Fragments of SRC-1 were also cloned in the pGEX-3X vector, resulting in constructs representing aa 1–568, aa 568–780, aa 780-1029, aa 1029–1255, aa 1255–1441, and aa 1321–1441 of SRC-1, respectively. Fragments of TIF2 (aa 577–796) and ACTR (aa 562–783), which contain the NID of TIF2 or ACTR, respectively, were also cloned in the pGEX-3X vector.
Oligonucleotide-Directed Mutagenesis
Using the pcDNA3-hLRH-1 as the template, point mutations were introduced into the hinge region and helix 1 with GeneEditor In Vitro Site-Directed Mutagenesis System (Promega) according to the manufacturer’s protocol. All mutations were confirmed by sequencing. The after oligonucleotides were used in the mutagenesis (mutations in lowercase letters). TSS251AAA, 5'-TGGATAGTTACCAGgCcgcagCTCCAGCAAGCATC-3'; IL260AA, 5'-CCACAT-CTGgcAgctAA CTTTTGAAG-3'; E261G, 5'-CTGATACTGGgACTTTTGAAG-3'; LL263AA, 5'-CATCTGATACTG GAAgcT-gcaAAGTGTGAGCCAG-3'; and H1_del, 5'-AGCTCTCCA-GCAAGCATCCCACATCTG AAGTGTGAGCCAGATGAGC-CTCAAGTCCAGG-3'.
Transfection and Luciferase Assays
Huh7, HeLa, and COS cells were grown in DMEM supplemented with 10% fetal calf serum at 37 C and 5% CO2. Cells were seeded in 35-m dishes at a density of 2.5 x 105 cells per dish and grown for 24 h. One to 4 h before transfection, medium was refreshed. Transient transfections were carried out with the calcium-phosphate precipitation method as previously described (15). Three micrograms of total DNA were used for each transfection, which included 0.5 µg of the pCMV-lacZ (Promega) as an internal control to monitor cell viability and to normalize transfection efficiencies among different transfections. When necessary, appropriate amount of the empty vectors (pM, pcDNA3, pCR3.1, and pVP16) were added to ensure equal quantity of DNA among different transfections. Forty-eight hours post transfection, cells were harvested and lysed in 1x Reporter Lysis Buffer (Promega). Luciferase activities were determined with Luciferase Assay System (Promega). ß-Galactosidase activities were measured according to a standard colorimetric method (63). Luciferase activities of different transfections were normalized by ß-galactosidase activities. Each transfection was performed in duplicate dishes and repeated at least three times.
Western Blotting
Huh7 cells were seeded in 35-mm dishes at a density of 2.5 x 105 cells per dish and were transfected with appropriate expression plasmids and 0.5 µg of the pCMV-lacZ. Forty-eight hours post transfection, cells were harvested. A small proportion of the cells were removed for the measurement of the ß-galactosidase activity, which was used to monitor cell viability and to normalize the transfection efficiencies among different transfections. The remaining cells were lysed. To ensure equal loading, the amount of cell lysates for each transfection that would be subjected to 10% Tris-glycine SDS-PAGE was adjusted based on the ß-galactosidase activity. After electrophoresis, proteins were transferred onto nitrocellulose membrane (PROTRAN; Schleicher & Shnell GmbH, Dassel, Germany). Immunoblotting was carried out with an anti-GAL4-DBD monoclonal antibody SC-510 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), followed by a rabbit antimouse Ig/horseradish peroxidase (Dako Corp., Carpinteria, CA) as the secondary antibody. Peroxidase activities were detected by the enhanced chemiluminescence reaction with the Western Blot Luminol Reagent (Santa Cruz Biotechnology, Inc.) according to the manufacturer’s instruction.
GST Pull-Down Assay
GST fusion proteins were expressed in E. coli. BL21(DE3) cultures by the induction with 0.2 mM isopropyl-ß-D-thiogalactoside for 2 h at 37 C. Fusion proteins were purified with glutathione-Sepharose 4B beads according to the manufacturer’s instruction (Amersham Pharmacia Biotech). Purified proteins were quantified by comparing the sample lanes with a BSA standard on coomassic-stained SDS-PAGE. SRC-1, TIF2, ACTR, hLRH-1, and hRAR proteins were synthesized in vitro with the TNT Quick Coupled Transcription/translation System (Promega) in the presence of 35[S] methionine (Amersham Pharmacia) according to the manufacturer’s protocol. Pull-down assays were performed using the purified GST fusion proteins and the appropriate labeled proteins. Briefly, 2 µg of a GST fusion protein bound to glutathione-Sepharose beads was incubated with 10 µl of a labeled protein in the binding buffer containing 200 mM KCl at 4 C for 1.5 h. After extensive washing, the mixture was boiled and resolved in 8% or 10% SDS-PAGE. The gel was fixed and dried. Signals were detected by autoradiography.
RT-PCR
Semiquantitative RT-PCR was performed to determine the relative levels of the CYP7A1 transcript among different transfection assays. Huh7 cells were grown in 60-mm dishes and transfected with expression plasmids using the FuGene 6 reagent according to the manufacturer’s instruction (Roche Molecular Biochemicals, Indianapolis, IN). Thirty-six hours post transfection, total RNA was isolated with the TRIzol reagent (Invitrogen Life Technologies). Two micrograms of RNA from each sample were used in the reverse transcription reaction with the Moloney murine leukemia virus reverse transcriptase (Promega). For the PCR, the following primer pair was used to amplify a 416-bp fragment of hCYP7A1 cDNA (forward: 5'-GCATCATAGCTCTTTACCCAC-3', reverse: 5'-GGTGTTCTGCAGTCCTGTAAT-3'). Meanwhile, a 145-bp fragment of the housekeeping GAPDH cDNA was also amplified (forward: 5'-CCATGACAACTTTGGTATCGTG-3', reverse: 5'-GCCAGTAG-AGGCAGGGATGA-3'). PCRs were carried out with the following condition: 94 C 1 min, 50 C 1 min, and 72 C 30 sec for 33 cycles for CYP7A1 and 19 cycles for GAPDH. The amplified products were separated on 2% agarose gels and visualized by ethidium-bromide staining. Band density was calculated using the Totallab gel scanning and quantification software. All RT-PCR assays were repeated at least three times.
Detection of SRCs and hLRH-1 in Huh7 cells was performed with standard RT-PCR using after primers to produce fragments located in the C-terminal region of SRCs and the hinge region of hLRH-1, respectively. (SRC-1 forward: 5'-CGAGGTGCTTTTTCACCTGGC-3', reverse: 5'-CAT GGAG-CTCCCAGGGCTTA-3'; TIF2 forward: 5'-CCAAATCAACTAAGACTTCAAC-3', reverse: 5'-GTTACTGTACATGCTGGTGT-TTG-3'; ACTR forward: 5'-CCTCCGCAACAGTTTCCATATC-3', reverse: 5'-AGACATGGGCATGGGGTTCAT-3'; hLRH-1 forward: 5'-TCCAAAGGCCTACCTC TGAAC-3', reverse: 5'-AACTATCCATATATGAATAGC-3'). PCRs were carried out with the following condition: 94 C 45 sec, 54 C 45 sec, and 72 C 50 sec for 30 cycles. The amplified products were separated on 1% agarose gel and visualized by ethidium-bromide staining.
Homology Modeling
The LBD of hLRH-1 shares 87% sequence identity and is homologous in function with mLRH-1 LBD. To understand the molecular basis of hLRH-1 transcriptional activity and the potential functional roles of the hinge region and the LBD in the mechanism of transcriptional activation, we built a three-dimensional model of hLRH-1 LBD based on the crystal structure of mLRH-1 LBD (Protein Data Bank entry 1PK5) (41). Sequence alignment between the LBDs from mouse and hLRH-1 was performed using the program CLUSTAL-W (64). The initial model was obtained using the homology-modeling program MODELLER (65). The less conserved regions and buried side chains were manually oriented to have favorable interactions with each other. Models of SRC-1 and SMRT structural segments were docked based on the structures of PPAR
, RXR, and RAR-related orphan receptor in their complexes with SRC-1 and SMRT peptides. The corresponding structures of these complexes were superimposed onto the structural model of hLRH-1 LBD based on the structural alignment of the sandwich framework of nuclear receptors. Molecular dynamics simulation of hLRH-1 LBD model was performed using the program GROMACS (66, 67) in water bath according to the standard procedures of molecular simulations.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
This work was supported by the National Natural Science Foundation of China (30100088, 30393112, 30125011, and 30170223), High Technology R&D Project (2001AA221261, 2002BA711A13) from Ministry of Science and Technology, and Innovation Program of Chinese Academy of Sciences (KSCX2-SW-208, KSCX1-SW-17). Y.-H.X. is supported by the Qi Ming Xin Program (01QA14046) from Shanghai Science and Technology Committee, and M.L. by a fellowship from CAS for outstanding Ph.D students.
Abbreviations: aa, Amino acids; ACTR, acetyltransferase; AD, activation domain; AF-2, activation function-2 domain; ATRA, all-trans retinoic acid; bHLH-PAS, basic-helix-loop-helix/Per/ARNT/Sim; CBP, cAMP response element binding protein (CREB)-binding protein; CYP7A1, cholesterol 7--hydroxylase; DBD, DNA binding domain; ENII/Cp, enhancer II and core promoter of HBV; FTZ-F1, Fushi tarazu factor 1; GST, glutathione-S-transferase; h, human; HBV, hepatitis B virus; HNF, hepatocyte nuclear factor; KO, knockout; LRH-1, liver receptor homolog-1; LBD, ligand binding domain; NID, nuclear receptor-interacting domain; NIDm, mutant NID; NR, nuclear receptor; Q-rich, glutamine-rich; PPAR, peroxisome proliferator-activated receptor; RAR, retinoic acid receptor; RXR, retinoid X receptor; SF-1, steroidogenic factor-1; SMRT, silencing mediator of retinoic acid and thyroid hormone receptors; SRC, steroid receptor coactivator; TIF2, transcriptional intermediary factor 2.
Received for publication September 1, 2003. Accepted for publication May 3, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
