Orphan Nuclear Receptor Small Heterodimer Partner, a Novel ...

0888-8809/04/$15.00/0 Printed in U.S.A.

Molecular Endocrinology 18(4):776–790 Copyright © 2004 by The Endocrine Society doi: 10.1210/me.2003-0311

Orphan Nuclear Receptor Small Heterodimer Partner, a Novel Corepressor for a Basic HelixLoop-Helix Transcription Factor BETA2/NeuroD JOON-YOUNG KIM, KHOI CHU, HAN-JONG KIM, HYUN-A SEONG, KI-CHEOL PARK, SABYASACHI SANYAL, JUN TAKEDA, HYUNJUNG HA, MINHO SHONG, MING-JER TSAI, HUENG-SIK CHOI

AND

Hormone Research Center (J.-Y.K., H.-J.K., S.S., H.-S.C.), School of Biological Sciences and Technology, Chonnam National University, Kwangju 500-757, Republic of Korea; Department of Molecular and Cellular Biology (K.C., M.-J.T.), Baylor College of Medicine, Houston, Texas 77030; Department of Biochemistry (H.-A.S., H.H.), School of Life Sciences, Chungbuk National University, Cheongju 361-763, Republic of Korea; Laboratory of Endocrine Cell Biology (K.-C.P., M.S.), Department of Internal Medicine, Chungnam National University School of Medicine, Daejon 301721, Republic of Korea; and Department of Endocrinology (J.T.), Diabetes and Rheumatology, Division of Bioregulatory Medicine, Gifu University School of Medicine, Gifu-city, Gifu 500-8705, Japan Small heterodimer partner (SHP; NR0B2) is an atypical orphan nuclear receptor that lacks a conventional DNA binding domain (DBD) and represses the transcriptional activity of various nuclear receptors. In this study, we examined the novel cross talk between SHP and BETA2/NeuroD, a basic helix-loop-helix transcription factor. In vitro and in vivo protein interaction studies showed that SHP physically interacts with BETA2/NeuroD, but not its heterodimer partner E47. Moreover, confocal microscopic study and immunostaining results demonstrated that SHP colocalized with BETA2 in islets of mouse pancreas. SHP inhibited BETA2/ NeuroD-dependent transactivation of an E-box reporter, whereas SHP was unable to repress the E47-mediated transactivation and the E-box mutant reporter activity. In addition, SHP repressed the BETA2-dependent activity of glucokinase and cyclin-dependent kinase inhibitor p21 gene pro-

moters. Gel shift and in vitro protein competition assays indicated that SHP inhibits neither dimerization nor DNA binding of BETA2 and E47. Rather, SHP directly repressed BETA2 transcriptional activity and p300-enhanced BETA2/NeuroD transcriptional activity by inhibiting interaction between BETA2 and coactivator p300. We also showed that C-terminal repression domain within SHP is also required for BETA2 repression. However, inhibition of BETA2 activity was not observed by naturally occurring human SHP mutants that cannot interact with BETA2/NeuroD. Taken together, these results suggest that SHP acts as a novel corepressor for basic helix-loop-helix transcription factor BETA2/NeuroD by competing with coactivator p300 for binding to BETA2/NeuroD and by its direct transcriptional repression function. (Molecular Endocrinology 18: 776–790, 2004)

S

hormone receptor (TR), constitutive androstane receptor (CAR), retinoid X receptor (RXR) (1–5), hepatocyte nuclear factor 4␣ (HNF 4␣) (3, 6, 7), liver receptor homolog 1 (LRH-1) (8–10), androgen receptor (AR) (11), and liver X receptor (LXR) (12), estrogen-related receptor ␥ (ERR ␥) (13), glucocorticoid receptor (14), and pregnelone X receptor (15). Detailed structural analysis of SHP demonstrated that a receptor interacting domain and a repression domain were mapped

MALL HETERODIMER partner (SHP) is one of unique members of the orphan nuclear receptor superfamily that lacks a conventional DNA binding domain (DBD) (1). SHP regulates the transcriptional activity of various nuclear receptors via direct physical interaction, such as estrogen receptor (ER), thyroid Abbreviations: aa, Amino acid; AD, activation domain; AHR, aryl hydrocarbon receptor; ARNT, AHR nuclear translocator; bHLH, basic helix-loop-helix; CAR, constitutive androstane receptor; CBP, cAMP response element binding protein-binding protein; CMV, cytomegalovirus; DBD, DNA binding domain; EID1, E1A-like inhibitor of differentiation 1; ER, estrogen receptor; ERR ␥, estrogen-related receptor ␥; GST, glutathione-S-transferase; HA, hemagglutinin; HEB, HeLa E-box binding protein; HLH, helix-loop-helix; HNF 4␣, hepatocyte nuclear factor 4␣; hSHP, human SHP; IA, insulinoma-associated antigen-1; Id, inhibitor of differentiation; LBD, ligand-binding domain; LRH-1, liver receptor homolog 1; MOP4, member of PAS superfamily; mSHP, murine SHP;

ORF, open reading frame; PAS, Per-ARNT-Sim; RIPE3, rat insulin promoter element 3; RAR, retinoic acid receptor; RXR, retinoid X receptor; SDS, sodium dodecyl sulfate; SHP, small heterodimer partner; tk, thymidine kinase; TR, thyroid hormone receptor. Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community. 776

Kim et al. • Orphan Nuclear Receptor SHP

to the N terminus and the C terminus regions, respectively (5). Although the precise molecular mechanisms involved in the SHP-mediated repression of nuclear receptors remain unclear, the repression domain is indispensable for such repression (2, 3, 9). A number of reports have demonstrated that such repressive actions of SHP may involve 1) inhibition of DNA binding of the target protein (1, 15); 2) competition with coactivators for binding to the activation function 2 surfaces of nuclear receptors (15); 3) recruiting unknown corepressors through its intrinsic transrepression domain (2, 3, 9); 4) interacting with E1A-like inhibitor of differentiation 1 (EID-1) (16), which antagonize the p300 coactivation function; and 5) interacting with RNA polymerase II, an additional way how SHP represses both basal and induced transactivation (12). It is also reported that SHP acts as a negative regulator of basic helix-loop-helix (bHLH)-PAS (PerARNT-Sim) transcription factors, aryl hydrocarbon receptor (AHR)/AHR nuclear translocator (ARNT) by inhibiting DNA binding and transcriptional activation via direct interaction with ARNT (17). However, it remains to be determined whether SHP can act as a coregulator for other families of transcription factors. SHP is expressed in a wide variety of tissues, including heart, brain, liver, spleen, adrenal gland, small intestine, and pancreas (6, 7, 10, 13, 18). SHP gene consists of two exons and one intron (19), and SHP gene transcription is regulated by several members of the nuclear receptor superfamily, including bile acid receptor (8, 10), steroidogenic factor-1 (20), HNF 4␣ (7), LRH-1 (8–10), ERR ␥ (13), and ER␣ (21). Recently, we reported that human SHP gene promoter is synergistically activated by bHLH protein E2A and steroidogenic factor-1 (22). Previous reports have suggested that SHP plays a pivotal role in the regulation of cholesterol homeostasis via bile acid-activated regulatory cascade in the liver (8, 10, 23; reviewed in Ref. 24). The bile acid receptor-mediated SHP gene induction has been shown to inhibit the activity of orphan nuclear receptor LRH-1 that positively regulate the expression of cholesterol 7 ␣-hydroxylase (CYP7A1) gene, which catalyzes the rate-limiting step in bile acid biosynthesis. In addition, loss of function studies of SHP demonstrated that redundant pathways independent of SHP-mediated pathway are implicated in the negative feedback regulation of bile acid production (25, 26). The SHP-independent pathways have been suggested to include liver damage, changes in bile acid pool size, activation of the xenobiotic receptor, pregnelone X receptor, and activation of the JNK (25, 26). In addition, it has been reported that mutations in the SHP gene in human are associated with mild hyperinsulinemia and obesity via inhibition of HNF4␣, a positive regulator of insulin secretion (6). Genetic variation in the SHP locus has also been suggested to influence birth weight and have effects on body mass index (BMI), although mutations in SHP are not a common cause of severe human obesity (27, 28).

Mol Endocrinol, April 2004, 18(4):776–790

777

bHLH transcription factors comprise a large family of transcription factors, which bind to the specific DNA sequence CANNTG, known as E-box (reviewed in Ref. 29). The bHLH proteins play important roles in development, the cell-specific expression of diverse genes and the maintenance of many differentiated cell types (30–35). The E-proteins, including E2-2, HEB, E12 and E47, are ubiquitously expressed class A bHLH transcription factors and homodimerize or heterodimerize with class B bHLH proteins through the helix-loophelix (HLH) domains (35–37). Class B bHLH proteins including MyoD (33, 34), myogenin (35), BETA2/ NeuroD (32, 38–46) are tissue-specifically expressed and these proteins heterodimerize with class A bHLH proteins. These heterodimers activate cell type-specific gene transcription by binding to E-boxes (CANNTG) through the basic domains. In addition, the formation of active class A and class B heterodimer is modulated by the Id (inhibitor of differentiation) family members, another distinct class HLH proteins (reviewed in Refs. 29 and 47). The four Id proteins, Id-1, Id-2, Id-3, and Id-4, identified thus far have only HLH domain and lack the basic DNA binding region. Dimerization of Id proteins with bHLH proteins results in the inhibition of DNA binding of bHLH proteins. BETA2/NeuroD, a member of tissue-specific class B bHLH proteins was independently cloned and characterized as a positive regulator of insulin gene expression (48) and a factor required for neuronal differentiation (38). BETA2/NeuroD has been shown to play important roles in the development of the nervous system and the maintenance and formation of pancreatic and enteroendocrine cells (30–32, 49, 50). BETA2/ NeuroD and E47/E12 heterodimer binds to with high affinity and transactivates the E-box elements in their target gene promoters, including the insulin (42, 48, 51–55), glucagon (51) glucokinase (GK) (44), secretin (39, 40), cyclin-dependent kinase (cdk) inhibitor p21 (39), proopiomelanocortin (41), sulfonylurea receptor I (43) and IA-1genes (45) by binding to E-box (CANNTG) elements. In addition, activation of E-box-dependent transcription by BETA2/NeuroD is potentiated by the presence of the coactivator, p300/cAMP response element binding protein-binding protein (CBP). p300/ CBP interacts with BETA2/NeuroD through its Cterminal region containing cysteine/histidine (C/H3)and glutamine-rich (Q-rich) domains (39, 42, 52, 53). In this study, we demonstrate that SHP acts as a novel coregulator of the bHLH protein BETA2/NeuroD via physical interaction. We also elucidated the molecular mechanism underlying the repression of BETA2/NeuroD activity by SHP. We showed that SHP represses the BETA2/NeuroD-mediated reporter activity and the p300-enhanced transcriptional activity of BETA2/NeuroD by interfering with coactivator p300/ CBP binding to BETA2/NeuroD. In addition, the transcriptional repression domain of SHP is required for the inhibition of BETA2/NeuroD transactivity. These results suggest that SHP acts as a novel corepressor for a bHLH transcription factor BETA2/NeuroD and

778 Mol Endocrinol, April 2004, 18(4):776–790


expand the role of SHP as a coregulator for other family of transcription factors, as well as nuclear receptors by dual mechanism; by competing with coactivators and by its direct transcriptional repression function.

RESULTS SHP Physically Interacts with a bHLH Transcription Factor, BETA2/NeuroD

Fig. 1. Identification of BETA2/NeuroD as a Novel SHPInteracting Protein A, SHP interacts with BETA2/NeuroD in GST pull-down assay. BETA2/NeuroD, E47, Id-2, and mPer1 were labeled with [35S]methionine by in vitro translation and incubated with glutathione-sepharose beads containing bacterially expressed GST alone, GST-SHP fusion proteins. The input lane represents 10% of the total volume used in the binding assay. B, Yeast two-hybrid assay. The plasmids encoding LexA DBD or LexA-SHP were cotransformed with plasmids encoding B42AD fusions of BETA2, E47, or Id-2 into yeast strain EGY48. C, BETA2/NeuroD specifically interacts with SHP but not with other nuclear receptors analyzed. The plasmids en-

To identify novel SHP-interacting proteins, we performed a yeast two-hybrid screening using the fulllength human SHP as a bait. Screening of a human spleen cDNA library led to the identification of the helix loop helix protein, Id-2 as a novel SHP-interacting protein, suggesting that SHP can interact with distinct family of transcription factors such as bHLH proteins. Previously, it is reported that SHP acts as a negative regulator of bHLH-PAS transcription factors, AHR/ ARNT via direct interaction with ARNT (17). Based on this report and our observation, we investigated whether SHP could interact with bHLH proteins in glutathione-S-transferase (GST) pull-down assay and identified BETA2 as an SHP-interacting protein. As shown in Fig. 1A, SHP interacted with BETA2 as well as Id-2, whereas SHP did not interact with E47 and bHLH-PAS protein mPer1 (56). We could not observe any interaction of SHP with other bHLH proteins, MyoD and Stra13 (data not shown). To confirm the interaction between SHP and BETA2, a yeast twohybrid assay was performed by cotransforming the plasmids encoding LexA-SHP together with the plasmids for B42-BETA2, B42-E47, or control B42AD into yeast cells. Consistent with the results from GST pulldown assay, a strong interaction was observed between SHP and BETA2 (⬃17-fold), whereas no significant interaction of SHP with E47 (⬃1.8-fold) or control B42AD was observed, although E47 is a heterodimer partner of BETA2/NeuroD (Fig. 1B). To confirm the specificity of the interaction between SHP and BETA2, we examined the interaction between BETA2 and other nuclear receptors, including RXR, RAR, TR, and glucocorticoid receptor in a yeast two-hybrid assay in the absence or presence of cognate ligands. No significant interaction was detected between BETA2 and the nuclear receptors analyzed (Fig. 1C). Taken together, these results indicate that SHP specifically interacts with a tissue- specific bHLH transcription factor BETA2/neuroD.

coding LexA fusions of RXR, RAR, TR, or SHP were transformed with the plasmids for B42AD or B42-BETA2 into yeast cells. The transformants were grown in the selective medium in the absent or presence of cognate ligands and assayed for ␤-galactosidase activity. RXR; 1 ␮M 9-cis-retinoic acid, RAR; 1 ␮M all-trans retinoic acid, TR; 1 ␮M T3, respectively. The data are representative of at least three independent experiments.



779

Fig. 2. Interaction and Colocalization between SHP and BETA2/NeuroD in Vivo A and B, In vivo interaction assay. 293T cells were cotransfected with pcDNA3/HA-BETA2 (A) or pcDNA3/HA-E47 (B) and pEBG-SHP (GST-SHP) or GST alone (pEBG), as a control. After 48 h, cells were extracted and GST fusion proteins were purified with glutathione-sepharose beads (GST puri.) and analyzed on a SDS-polyacrylamide gel. The complex formation (upper left, GST puri.) and the amount of HA-BETA2 or HA-E47 used for the in vivo binding assay (upper right, Lysate) were determined by anti-HA antibody immunoblot. The same blot was stripped and reprobed with an anti-GST antibody (lower panel) to confirm the expression levels of the GST fusion protein (GST-hSHP) and the GST control (GST). C, Colocalization between SHP and BETA2. COS-7 cells were transiently transfected with pEGFP-SHP and pCDNA3/HA-BETA2 (see Materials and Methods). At 24 h after transfection, cells were fixed in 3.7% formaldehyde for 40 min. Fixed cells were mounted onto glass slides with PBS and observed with a laser-scanning confocal microscope. GFP-fused SHP was detected by autofluorescence, and HA-BETA2 was detected by staining with primary anti-HA antibody and rhodamine-conjugated secondary antibody. The yellow stain in the merged image depicts colocalization of SHP and BETA2. Shown are representative cells from one of three independent experiments. D, Immunostaining of SHP in mouse pancreas. SHP (green) was detected in the endocrine islet cells as indicated. Glucagonproducing ␣-cells (red) were detected in cells surrounding the core of the islet. Merged image shows the presence of SHP in the central core of the islet and in cells expressing glucagon. Nuclei were stained with 4⬘,6-diamidino-2-phenylindole (DAPI) (blue) in the merged image. No staining was observed when no primary antibody was added (control).

780


SHP Interacts and Colocalizes with BETA2/NeuroD in Vivo To verify the in vivo interaction between SHP and BETA2/NeuroD, we cotransfected mammalian expression vectors encoding either GST alone or GST-SHP together with hemagglutinin (HA)-BETA2 or HA-E47 into 293T cells. As shown in Fig. 2A, upper left, BETA2 was detected in the coprecipitate only when coexpressed with the GST-SHP but not with the control GST alone. In contrast, the interaction between SHP and E47 (Fig. 2B, upper left) was not observed, although the expression level of E47 was similar to that of BETA2. Expression level of GST- and HA-tagged proteins was also confirmed by Western blot analysis of crude cell lysates with the antibodies against HA and GST (Fig. 2, A and B, upper right and lower, respectively). These results indicated that SHP interacts with BETA2/NeuroD in vivo and were consistent with the results of the yeast two-hybrid and in vitro GST pull-down assay. To determine the subcellular localization of SHP and BETA2, immunofluorescence confocal microscopy was performed in COS-7 cells cotransfected with pEGFPSHP and pCDNA3/HA-BETA2. GFP-SHP protein was mainly distributed in the nucleus but also weakly detected in the cytoplasm (Fig. 2C). HA-BETA2 proteins are localized predominantly in the nucleus. SHP colocalized with BETA2 in the nucleus, as depicted in the merged image (Fig. 2C). However, SHP and ␤ were not colocalized in the cytoplasm. The result indicates that SHP and BETA2 colocalize mainly in the nucleus, rather than in the cytoplasm. Based on our previous report that BETA2/ NeuroD is expressed in islet cells (32), indirect immunostaining of SHP in mouse pancreas was performed to verify the physiological relevance of the interaction of SHP and BETA2. Specific immunostaining of SHP was observed in islets of Langerhans, where BETA2 is known to be expressed (Fig. 2D). SHP is localized in glucagon producing ␣-cells as depicted in the merged image (Fig. 2D, top middle). In the absence of primary antibody, no staining was observed (Fig. 2D, bottom). This result demonstrated that SHP is coexpressed with BETA2 in islets of pancreas. Taken together, these results suggest that SHP interacts and colocalizes with BETA2 in vivo. SHP Represses the BETA2/NeuroD-Mediated Transactivation It has been demonstrated that the heterodimer of BETA2/NeuroD and E47 transactivates their target genes by binding to E-box (CANNTG) elements (39– 41, 44–46, 48, 51–55). To investigate the functional significance of the interaction between SHP and BETA2/ NeuroD, we performed transient transfections assays in CV-1 cells using the reporter plasmids containing wild-type E-box [rat insulin promoter element 3 (RIPE3)], or mutant E-boxes (RIPE3 Em). In agreement with previous reports (42, 48, 52), BETA2 and E47 activated the RIPE3 reporter genes, but not the E-box


mutant reporter, RIPE3 Em. Interestingly, SHP markedly decreased the BETA2-E47-dependent transactivation of the RIPE3 reporter gene in a dose-dependent manner (Fig. 3A). However, SHP did not have any significant effect on that of RIPE3 Em or the basal activity of the reporter genes (Fig. 3B), suggesting that the BETA2-E47 repression by SHP is mediated through E-box. To test the effect of SHP on the E47 transactivity, we used the (523)4 luc reporter plasmid that contains multiple E-boxes specific for active homodimers of E47 (57–60). As expected, SHP was unable to repress the E47-dependent transactivation of the (523)4 luc reporter (Fig. 3C). This result indicates that SHP specifically inhibits the BETA2/NeuroD transactivity and supports our results showing that SHP specifically interacts with BETA2, but not E47. To examine whether the BETA2 repression by SHP is also occurring on the natural BETA2 target gene promoters, transient transfection was performed using the reporters containing ␤GK and p21 gene promoters. As previously reported (39, 44), BETA2 transactivates the GK and p21 luciferase reporter genes in CV-1 cells (Fig. 3, D and E). SHP also repressed the BETA2dependent transactivation of the GK and p21 genes, suggesting that SHP can regulate BETA2 transactivity on the natural promoter contexts. Taken together, these results demonstrated that SHP specifically inhibits the transcriptional activity of BETA2/NeuroD. Determination of Interaction Domain within SHP and BETA2/NeuroD To map the domains within SHP required for BETA2/ NeuroD interaction, we employed a number of previously described (5, 13) deletion constructs of SHP in yeast two-hybrid interaction assay (Fig. 4A). B42BETA2 interacted more weakly than wild-type SHP with LexA-W160X (1–159 aa), which contains the entire N terminus and the nuclear receptor-interacting domain. However, the ⌬N148 (1–3, 72–148 aa) construct which contains the interaction domain for the nuclear receptors including RXR (5) and ERR ␥ (13) did not show any significant interaction with BETA2. In addition, other deletion constructs ⌬E1X (1–71, 91–93 aa), ⌬120 (120–260 aa) and ⌬210 (210–260 aa), which lack a part or the complete interaction domain showed little or no interaction with BETA2 (Fig. 4B), indicating that in addition to the nuclear receptor interaction domain, the extra N terminus region of SHP is also required for the BETA2/NeuroD interaction. Taken together, these results suggest that SHP needs the entire N-terminal domain (1–159 aa) for the BETA2/ NeuroD interaction. Tissue-specific class B bHLH proteins heterodimerize with ubiquitous E proteins E47/E12 and bind to E-box elements through its bHLH domain (35), and it has been reported that BETA2/NeuroD interacts with a coactivator p300/CBP through its bHLH domain and activation domain (AD) (39, 42, 52). To identify the SHP interaction surface in BETA2/neuroD, we performed



781

Fig. 3. SHP Represses the Transcriptional Activity of BETA2/NeuroD. A, CV-1 cells were cotransfected with 0.2 ␮g of the reporter genes, RIPE3 (A), RIPE3 Em (B), (523)4 (C), ␤GK/luc (⫺1003/⫹196) (D) or p21/luc (⫺143/⫹8) (E), together with 0.1 ␮g of expression vectors for pcDNA3/HA-BETA2, pcDNA3/HA-E47 and indicated amounts of pcDNA3/HA-SHP. The empty vector pcDNA3 was used to adjust the total DNA amounts and 0.1 ␮g of pCMV-␤galactosidase expression plasmids were used as an internal control. The relative luciferase activity was normalized against ␤-galactosidase activity. The data are representative of at least three independent experiments.

yeast two hybrid interaction assays using B42AD fused BETA2 deletion constructs (Fig. 4C). As shown in Fig. 4D, LexA-SHP interacted with both B42BETA2N (1–165 aa) and B42-BETA2C (166–355 aa). However, the interactions of BETA2 deletion mutants with SHP were somewhat weaker than that of BETA2 wild type. To further confirm yeast two-hybrid interaction results, we performed GST pull-down assay using GST-SHP or GST-E47 together with 35S-labeled BETA2N or BETA2C. Consistent with the yeast interaction results, SHP interacted with both BETA2N and BETA2C with weaker affinity than wild type BETA2 (Fig. 4E). In a good agreement with previous reports that bHLH transcription factors dimerize through their bHLH domain (35–37), BETA2N containing bHLH domain interacted with GST-E47 but not GST alone.

BETA2C-containing activation domain did not interact with GST-E47. These results indicate that both N terminus and C terminus of BETA2/NeuroD are involved in the SHP interaction. SHP Inhibits Neither Heterodimerization nor E-Box Binding by E47 and BETA2/NeuroD To examine whether the repression of BETA2/NeuroDdependent transcription by SHP results from interfering with the dimerization or binding of BETA2/NeuroD and E47 to an E-box element, we performed an in vitro competition assay using GST-E47 and 35S-labeled BETA2 in the absence or presence of excessive amounts of purified His-tagged SHP or Id-2 proteins. As shown in Fig. 5A, the presence of SHP did not show any significant

782



Fig. 4. Detailed Mapping of the Interaction Surfaces within SHP and BETA2/NeuroD A, The schematic structures of mSHP wild-type and deletion constructs described previously (5). INT and REP indicate nuclear receptor interaction domain and repression domain of SHP, respectively. The numbers in parentheses indicate amino acid residues. B, N-terminal region (1–159 aa) of mSHP is required for the interaction with BETA2/NeuroD. Plasmids encoding LexA-mSHP wild type (WT) or LexA-mSHP deletion mutants were cotransformed with plasmids for B42AD and B42-BETA2 into yeast cells. C, BETA2/NeuroD wild type (WT) and BETA2/NeuroD deletion constructs, BETA2N (1–165 aa) and BETA2C (166–355 aa) are shown schematically. bHLH and AD indicate bHLH domain and transactivation domain, respectively. The numbers in the figure represent amino acids residues. D and E, Both bHLH domain and transactivation domain of ␤/NeuroD are involved in the interaction with SHP in yeast two-hybrid assay (D) and GST pull-down assay (E). D, The plasmids for LexADBD or LexA-SHP were transformed with the plasmids encoding B42 fusions of B42AD, BETA2/NeuroD wild type, BETA2N or BETA2C. Liquid ␤-galactosidase assay was performed described in Materials and Methods. E, BETA2 WT, BETA2N and BETA2C were labeled with [35S]methionine by in vitro translation, incubated with glutathione-sepharose beads containing bacterially expressed GST alone, GST-SHP, and GST-E47 fusion proteins and analyzed on a SDS-polyacrylamide gel.

effect on the interaction between E47 and BETA2 (Fig. 5A, lanes 3-, 5), whereas the interaction was significantly inhibited by Id-2 (Fig. 5A, lanes 7–9). Next, to test whether SHP inhibits binding of BETA2-E47 heterodimer to E-box element by allosteric modulation of the heterodimeric complex, we performed gel shift assay with purified GST-E47, GST-BETA2, His-SHP and His-Id-2 proteins. Distinctive E47 homodimer and E47-BETA2 heterodimer bands were observed on the E-box element, as previously reported (48) (Fig. 5B, lanes 4 and 6, respectively). Unexpectedly, the DNA binding of the E47

homodimer and E47-BETA2 heterodimer was not significantly affected by the presence of SHP (Fig. 5B, lanes 7–9), whereas Id-2 inhibited DNA binding of both E47 homodimer and E47-BETA2 heterodimer (Fig. 5B, lanes 10– 12). The amounts of purified His-tagged SHP and Id-2 proteins used in each reaction were confirmed by the Coomassie blue staining (Fig. 5C). Taken together, these results demonstrate that SHP does not inhibit heterodimerization or E-box binding of BETA2/NeuroD and E47. Because SHP did not inhibit the heterodimerization and DNA binding of the BETA2/NeuroD and E47, we



783

Fig. 5. Effects of SHP on the Dimerization and E-Box Binding of BETA2/NeuroD and E47 A, SHP does not interfere with the dimerization between BETA2/NeuroD and E47. BETA2 labeled with [35S] methionine were incubated with bead-bound GST-E47 fusion protein (lanes 2–9) in the absence or presence of His-SHP (lanes 3–5) or His Id-2 (lanes 7–9) protein and then analyzed on a SDS-polyacrylamide gel. Lane 1 as input represents 10% of the total volume of 35 S-labeled BETA2 used in the assay. B, SHP dose not block E-box binding of E47 and BETA2/NeuroD. [␥-32P] Labeled E-box probes were incubated with 1 ␮g of GST purified proteins (BETA2 and E47) and increasing amounts of His-SHP (lanes 7–9) or His Id-2 (lanes 10–12) protein in each reaction. DNA-protein complexes were analyzed on 4% polyacrylamide gel and analyzed by autoradiography. Lane 1, Free probe; lane 2, His-SHP; lane 3, His-Id-2. C, Coomassie blue staining of increasing amounts of purified His-tagged proteins, His-SHP (lanes 2–4) and His-Id-2 (lanes 5–7) used in each assay A and B. Lane 1 shows the protein marker and the numbers indicates protein molecular mass. D, CV-1 cells were cotransfected with 0.2 ␮g of Gal4-tk-luc reporter plasmid, 0.1 ␮g of pCMXGal4-BETA2 or pCMXGal4-E47 and indicated amounts of pcDNA3/HA-SHP. The relative luciferase activity was normalized against ␤-galactosidase activity and the normalized luciferase activity stimulated by Gal4-BETA2 or Gal4-E47 alone was set as 100%.

determined whether the inhibitory effect of SHP on the BETA2/NeuroD-mediated transcription is due to a direct repression of the BETA2/NeuroD activity. To exclude the effect of E47 on the dimerization and DNA binding with BETA2/NeuroD, we used Gal4DBD fused BETA2. In CV-1 cells, Gal4-BETA2 increased approximately 5-fold transcription of the reporter gene driven by Gal4 binding sites. Interestingly, coexpression of SHP significantly decreased Gal4-BETA2-dependent transactivation. In contrast, SHP was unable to repress the transcriptional activity of Gal4-E47 (Fig. 5D), suggesting that SHP directly inhibited the transcriptional activity of BETA2/NeuroD. Taken together, these results demonstrated that SHP directly inhibits transcriptional activity of BETA2/NeuroD, but not that of E47.

SHP Inhibits p300-Enhanced Transactivity of BETA2/NeuroD and the Binding of p300 to BETA2/NeuroD It has been demonstrated that p300/CBP enhances the transcriptional activity of BETA2/NeuroD by binding to BETA2/NeuroD via its C-terminal region containing C/H3 and glutamine rich (Q-rich) domain (39, 42, 52). To examine whether SHP might affect the p300-enhanced transcriptional activity of BETA2, we cotransfected the expression plasmids encoding Gal4DBD fused BETA2, p300 and SHP into CV-1 cells. Overexpression of p300 enhanced the transcriptional activity of Gal4-BETA2. SHP significantly decreased the p300-enhanced transcriptional activity of Gal4-BETA2, suggesting that SHP



Fig. 6. SHP Represses the p300-Enhanced Transactivity of BETA2/NeuroD by Interfereing with p300 Binding to BETA2/NeuroD A, CV-1 Cells were cotransfected with 0.2 ␮g of Gal4-tk-luc reporter plasmid, 0.1 ␮g of pCMXGal4N (Gal4DBD) or pCMXGal4BETA2, 0.5 ␮g of pcDNA3-p300 and indicated amounts of pcDNA3/HA-SHP and 0.1 ␮g of pCMV-␤-galactosidase expression plasmids were used as an internal control. The relative luciferase activity was normalized against ␤-galactosidase activity and the normalized luciferase activity stimulated by Gal4DBD was set as 1. B and C, SHP interferes with the interaction between BETA2/NeuroD and p300. p300C indicates the C-terminal region of p300 (1254–2141 aa) including C/H3 and Q-rich domains. p300C labeled with [35S]methionine was incubated with bead-bound GST alone (B and C, lane 2) or GST-BETA2 (B, lanes 3–9) or GST-E47 fusion protein (C, lanes 3–9). Increasing amounts of purified His-SHP (lanes 4–6) or His Id-2 (lanes 7–9) proteins were added to this binding reaction. D, 35S-labeled p300C was incubated with bead-bound GST alone or GST-SHP and then analyzed on a SDS-polyacrylamide gel. Lane 1 as input represents 10% of the total volume of 35S-labeled p300C used in the assay.

may interfere with the coactivation function of p300 for BETA2/NeuroD (Fig. 6A). To ensure that SHP directly interferes with p300 for binding to BETA2/NeuroD, we performed in vitro competition assay using 35S-labeled C-terminal 1254–2141 amino acid residues of p300 (p300C), which includes C/H3 and Q-rich regions required for BETA2/NeuroD and E47 interaction. In a good agreement with previous reports (2, 3, 9), 35Slabeled p300C strongly interacted with GST-BETA2 and GST-E47 (Fig. 6, B and C). Intriguingly, the interaction between p300C and GST-BETA2 was inhibited by the presence of SHP, whereas Id-2 did not show any significant inhibition of the p300C-BETA2 interaction. In contrast, neither SHP nor Id-2 inhibited the interaction between GSTE47 and p300C, and interaction between SHP and p300C was not observed (Fig. 6D). Taken together, these results suggest that SHP specifically inhibits the transcriptional activity of BETA2/NeuroD by competing with p300 for the BETA2/NeuroD interaction. Transrepression Domain of SHP Is Required for the BETA2 Repression It has been shown that C-terminal transrepression domain of SHP is necessary for the full repression of

nuclear receptors, although the detailed repression mechanism through the domain still remains unclear (2, 3, 9). Because W160X includes entire N-terminal BETA2/NeuroD interaction domain but not transrepression domain of SHP, W160X is useful to determine whether the transrepression domain of SHP is also required for the repression of BETA2/NeuroD. To address this hypothesis, we transfected mammalian expression vectors encoding wild-type or mutant of human SHP (HA-W157X), corresponding to W160X of mouse SHP together with Gal4-BETA2 into CV-1 cells. As shown in Fig. 7A, Gal4-BETA2 activated the reporter gene activity and SHP wild type significantly repressed the transcriptional activity of Gal4-BETA2. Interestingly, W157X also repressed the transcriptional activity of Gal4-BETA2, but at a relatively reduced rate. To eliminate the possibility that the reduced ability of W157X to repress BETA2 transactivity is not due to lower protein expression of mutant SHP, we performed Western blot analysis using anti-HA antibody. As shown in Fig. 7B, protein expression level of HA-W157X was similar to that of wild-type SHP, indicating that reduced transrepression function of W157X is caused at least in part by the loss of tran-



785

domain to entire C terminus was tested for the BETA2/ NeuroD-dependent transactivation of the E-box reporter gene. As expected, the truncated SHP mutant L98 did not show any significant inhibition on the BETA2/NeuroD-dependent transactivation of the Ebox reporter gene, whereas both wild-type SHP and the polymorphic form R216H substantially repressed BETA2 activity (Fig. 8A). To identify whether the lack of the inhibitory effect of L98 on the BETA2/NeuroD activity was due to a defect in the interaction with BETA2/NeuroD, we performed yeast two-hybrid assay. As shown in Fig. 8B, LexA-L98 showed no significant interaction with B42-BETA2, whereas LexASHP wild type and the polymorphic form R216H showed strong interaction with B42-BETA2. These results suggest that N terminus of SHP including the receptor interaction domain is essential for its interaction with BETA2/NeuroD and for the repression of BETA2/NeuroD activity by SHP. DISCUSSION

Fig. 7. The Transrepression Domain of SHP Is Required for the Repression of BETA2/NeuroD Activity A, CV-1 cells were cotransfected with 0.2 ␮g of Gal4-tk-luc reporter plasmid, 0.1 ␮g of pCMXGal4N (Gal4DBD) or pCMXGal4BETA2, and indicated amounts of pcDNA3/HA-SHP wild-type or pcDNA3/HA-W157X, a deletion construct of human SHP. The relative luciferase activity was normalized against ␤-galactosidase activity and the normalized luciferase activity and the normalized activity of the reporter gene stimulated by Gal4DBD alone was set as 1. The data are representative of at least three independent experiments. B, Western blot analysis of SHP wild type and W157X. COS-7 cells were cotransfected with 4 ␮g of pcDNA3 empty vector (lane 1) and 2–4 ␮g of pcDNA3/HA-SHP wild type (lanes 2 and 3), and pcDNA3/HA-W157X (lanes 4 and 5). Western blot analysis using anti-HA antibody was performed as described in Materials and Methods.

srepression domain in SHP. Taken together, these results suggest that transrepression domain of SHP is also required for the full repression of BETA2/NeuroD activity. Naturally Occurring Mutant of Human SHP Failed to Repress the BETA2 Transactivity It has been reported that naturally occurring mutants of human SHP gene neither interact with the orphan nuclear receptor ERR ␥ nor repress the transcriptional activity of ERR ␥ (13). To examine whether the human SHP mutant has any effects on the transcriptional activity of BETA2/NeuroD, SHP truncated mutant L98 (6, 13) that is deleted from middle receptor interaction

The cross talk between different families of transcription factors plays important roles in the regulation of the eukaryotic gene expression. Nuclear receptors have been demonstrated to cross talk with other families of transcription factors including TGF-␤ signaling transducer Smads (61), nuclear factor (NF)-␬B (62) and activator protein (AP)-1 (63), through direct proteinprotein interaction and indirect mechanisms. Previously, it is reported that RXR and RAR directly interact with bHLH-PAS circadian transcription factors, CLOCK and MOP4 and inhibit their transcriptional activity by blocking their DNA binding (64). It is also reported that SHP acts as a negative regulator of bHLH-PAS transcription factors, AHR/ARNT by inhibiting DNA binding and transcriptional activation via direct interaction with ARNT (17). However, little is known about the interaction or cross talk between SHP and other bHLH transcription factors. Here, we demonstrated the novel cross talk between SHP and a bHLH transcription factor BETA2/NeuroD. We found that SHP physically interacts with BETA2/NeuroD in vivo and in vitro. We also demonstrated that SHP inhibits the transcriptional activity of BETA2/NeuroD by inhibiting its interaction with transcriptional coactivator, p300/CBP. BETA2 Inhibition mechanism by SHP is different from those of bHLH-PAS transcription factor AHR/ARNT by SHP and CLOCK by RXR in that SHP neither inhibits dimerization nor the DNA binding of BETA2/NeuroD and E47. This difference might be at least partly due to the difference in SHP interaction domains within BETA2 and ARNT. SHP is structurally and functionally distinct from other nuclear receptors and acts as a corepressor for its nuclear receptor partners. Therefore, the molecular mechanism underlying repression of BETA2/NeuroD by SHP can be different from that of bHLH-PAS transcription factor CLOCK by RXR.



Fig. 8. Naturally Occurring Mutant of Human SHP Fails to Repress the BETA2/NeuroD Transactivity A, CV-1 cells were cotransfected with 0.2 ␮g of E-box reporter plasmid and 0.4 ␮g of pcDNA3/HA-SHP wild-type (WT), polymorphic form R216H or truncated mutant type L98. B, The plasmids encoding LexA-SHP WT, LexA-R216H or LexA-L98 were cotransformed with B42-BETA2 into EGY48 yeast cells. L98; deletion of 9 bases and insertion of dinucleotide AC at codon 98 for Leucine resulting in deleting the receptor interaction domain and C-terminal transrepression domain. R216H; replacement of Arginine codon 216 by Histidine (6, 11). Liquid ␤-galactosidase assay was performed as described in Materials and Methods.

In an effort to identify SHP-interacting proteins, we isolated HLH protein Id-2 as a novel SHP-interacting protein, which interacts with bHLH transcription factors and represses their transcriptional activities. This finding prompted us to investigate whether SHP can interact with other bHLH proteins and we identified BETA2/NeuroD as a novel and specific in vitro- and in vivo-interacting protein of SHP. However, we could not observe any significant interactions between SHP and other bHLH proteins we analyzed including a ubiquitous bHLH protein E47, a tissue-specific bHLH protein MyoD and bHLH PAS protein Per. These distinct interaction results can be supported by the previous reports demonstrating that bHLH proteins show diverse interaction features and affinities with the different classes of bHLH proteins, although those interactions are mediated by helix loop helix region (65; reviewed in Ref. 29). In addition, SHP has been shown to interact with bHLH PAS protein ARNT, but not with AHR, although bHLH PAS protein AHR shares high homology with its heterodimer partner ARNT. We also demonstrated that overall structure of BETA2/NeuroD is implicated in SHP interaction. Therefore, our findings suggest that SHP has different interaction characteristics for the different members of bHLH proteins. Further studies should verify the potential interaction of SHP with other members of the bHLH protein family. It has been demonstrated that p300/CBP directly interacts with bHLH domain (34) and activation domain of BETA2/NeuroD (42, 52), as well as E47, and functions as a coactivator for BETA2/NeuroD and E47. In addition, SHP has been shown to compete with

p160 coactivators including steroid receptor coactivator-3 and transcription intermediary factor 2 for the binding to activation function 2 surfaces of nuclear receptors such as ER (2), HNF 4␣ (3), or LRH-1 (9) through receptor-interaction domain. Based on the interaction between SHP and HLH protein Id-2, we speculated that SHP might interact with bHLH domain of BETA2/NeuroD. However, unexpectedly, activation domain of BETA2/NeuroD as well as bHLH domain is involved in SHP interaction. We could not observe the blocking BETA2-E47 DNA binding by SHP. Rather, SHP inhibited the interaction between BETA2/NeuroD and coactivator p300 and repressed the p300-enhanced transcriptional activity of Gal4-BETA2. Our results suggest that SHP represses the transcriptional activity of bHLH protein BETA2/NeuroD, by interfering with p300 binding to the transcription factor BETA2/ NeuroD. In addition, the transrepression domain within SHP is also necessary for the full repression of BETA2/ NeuroD activity. Our findings suggest that the dual repression mechanism of nuclear receptors by SHP could be applicable to the regulation of the bHLH protein BETA2/NeuroD activity. However, the repression mechanism by SHP through its intrinsic repression domain is still elusive. It has been hypothesized that SHP may recruit unidentified corepressor complexes because SHP does not seem to interact with the nuclear receptor corepressor, or to recruit histone deacetylases (3, 5). However, a recent report has shown that SHP can recruit EID1 via the interaction surface within SHP encompassing putative helices 3 and 12, suggesting that inhibition of transcription by


SHP involves EID1 antagonism of CBP/p300-dependent coactivator functions (16). BETA2/NeuroD has been reported to play important roles in the development of the nervous system and the maintenance and formation of pancreatic and enteroendocrine cells (30–32, 49, 50). BETA2/NeuroD transactivates its target gene promoters by binding to E-box (CANNTG) elements as a heterodimer with E47/ E12. BETA2 has been shown to regulate various genes including insulin (42, 48, 51–53), glucagon (51) GK (44), secretin (39, 40), p21 (39), proopiomelanocortin (41), sulfonylurea receptor I (43) and IA-1 (45) genes. We showed that SHP inhibited the BETA2-dependent GK and p21 gene promoter activity, as well as the multiple E-box reporter activity, but not E47-mediated activation of E-box reporter. These results suggest that SHP might regulate BETA2 target genes through repression of BETA2 transactivity. This has functional significance in tissues including islets of pancreas that coexpress both BETA2 and SHP, although it remains to be determined whether SHP can act as a corepressor for BETA2 in the physiological condition. To gain better insights into the BETA2/NeuroD-mediated gene expression, it should be considered that the cellular protein ratio of a corepressor SHP and coactivators including p300 might play a key role in the BETA2/ NeuroD-mediated gene regulation.

MATERIALS AND METHODS Plasmids and DNA Construction CMV-E47 and the previously described reporter plasmids, RIPE3 (⫺126 to ⫺86 bp) and RIPE3 Em (mutated E-box) constructs were kind gifts from Dr. Roland Stein (52). The GK/Luc (⫺1003/⫹196) and p21/Luc (⫺143/⫹8) were generously provided by Dr. Yong-ho Ahn (66) and Dr. Aristidis Moustakas (67). The previously described (523)4 luciferase reporter was kindly donated by John Kim Choi (60). pCDNA3mPer1 plasmid was a gift from Dr. Hitoshi Okamura (56). The plasmids encoding LexA-RXR, RAR, TR, murine SHP (mSHP) wild-type and mSHP deletion constructs (1, 5), Gal4-thymidine kinase (tk)-luciferase reporter gene and pCR3.1-BETA2 are described previously (9, 48). B42-Id-2 was constructed by inserting PCR fragment into pB42AD/pJG4–5 vector (CLONTECH, Palo Alto, CA). PCR was performed using the human spleen cDNA library (Invitrogen, Carlsbad, CA) as a template and the primers designed based on the sequences of Id-2 from the GenBank (accession no. M97796). pcDNA3/ HA-Id-2 and His-Id-2 were constructed by insertion of open reading frame (ORF) fragment from B42 construct into a HA epitope-tagged pcDNA3 vector (pcDNA3/HA) and pET28b (Novagen, San Diego, CA) vector. B42AD fused BETA2 and E47 were constructed by inserting PCR fragments encoding the complete ORF of BETA2 and E47 into pB42AD/pJG4–5 vector. The ORF fragments of BETA2 and E47 from B42 constructs were subcloned into pcDNA3/HA, pGEX 4T-1 (Amersham Biosciences, Piscataway, NJ), pET28a, and pCMXGal4N vectors by appropriate restriction enzyme digestion. B42-BETA2N (1–165 aa) and B42-BETA2C (166–305 aa) were made by insertion of PCR fragments into pB42AD/ pJG4–5 vector and the ORF fragment of BETA2N and BETA2C were subcloned into pcDNA3/HA vectors. pcDNA3/ HA-human SHP (hSHP) wild type and hSHP L98, hSHP


787

R216H and hSHP W157X (1–157 aa) were constructed by insertion of PCR fragments into EcoRI-XhoI digested pcDNA3/HA vector. LexA-hSHP, L98 and R216H constructs were subcloned by insertion of ORF fragments into pLexA/ pEG202 vector. The ORF fragment of hSHP wild type was subcloned into pET28a, pGEX 4T-1, and pEBG vectors (68). pcDNA3-p300C (1254–2141 aa) was constructed by inserting BamHI-XhoI digested DNA fragment from pcDNA3-p300 fulllength into BamHI-XhoI digested pcDNA3 vector. All plasmids were confirmed by automatic sequencing analysis. Yeast Two-Hybrid Assay and Liquid ␤-Gal Assay Yeast two hybrid interaction assays were performed as previously described (13). Briefly, the plasmids encoding LexA fusions of RXR, RAR, TR, or SHP were transformed with the plasmids for B42AD or B42-BETA2 into yeast cells. The transformants were grown in the selective medium in the absence or presence of cognate ligands and assayed for ␤-galactosidase activity. Yeast EGY48 strain was cotransformed with the plasmids encoding LexADBD fused fulllength human or murine SHP, or deletions together with the plasmids for B42AD fused BETA2, BETA2 N (1–165 aa), BETA2 C (166–305 aa), E47 or Id-2. The transformants were selected on plates (Ura-, His-, and Trp-) with appropriate selection markers and assayed for ␤-galactosidase activity. GST Pull-Down Assay and in Vitro Competition Assay A GST pull-down assay was performed according to the method described previously (13, 55). Briefly, BETA2, BETA2N (1–165 aa), BETA2C (166–355 aa), E47, Id-2, Per and p300C (1254–2141 aa) were labeled with [35S]methionine using TNT (transcription and translation)-coupled reticulocyte lysate system (Promega Corp., Madison, WI) according to the manufacturer’s instructions. GST fused and His-tagged proteins were expressed in Escherichia coli, BL21 (DE3) pLys bacterial culture, induced by 0.2 mM isopropyl-␤-D-thiogalactopyranoside and cells were extracted. His-tagged proteins were purified using Ni-NTA agarose beads (QIAGEN Inc., Valencia, CA) according to the manufacturer’s instructions, and GST fusion proteins were prebound with glutathionesepharose beads and then incubated with in vitro-translated [35S]methionine-labeled BETA2, E47 or SHP proteins in binding buffer containing 25 mM HEPES (pH 7.6), 120 mM NaCl, 0.2 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 20% glycerol at 4 C for 2–3 h. For in vitro competition assay, purified His-SHP or His-Id-2 was added to the binding reaction and beads were washed three times with the binding buffer, resuspended in 2⫻ sodium dodecyl sulfate (SDS) loading buffer, analyzed by SDS-PAGE gel and visualized by a phosphorimage analyzer (BAS-1500, Fuji, Tokyo, Japan). GST fused or His-tagged proteins used in each reaction were analyzed by SDS-PAGE gel and quantified by Coomassie blue staining. Ten percent of in vitro-translated BETA2, E47, and SHP used in each reaction were used as input. In Vivo Interaction Assay 293T cells grown in DMEM supplemented with 10% fetal bovine serum were plated in six-well flat-bottomed microplates (Nunc, Roskilde, Denmark) at a concentration of 2 ⫻ 105 cells per well the day before transfection as described previously (68). Briefly, 1 ␮g of each plasmid DNA was transfected into 293T cells with a calcium phosphate precipitation method. Forty-eight hours after transfection, cells were solubilized with 100 ␮l of lysis buffer [20 mM HEPES (pH 7.9), 10 mM EDTA, 0.1 M KCl, and 0.3 M NaCl] containing 0.1% Nonidet P-40, 10 ␮g/ml aprotinin, 10 ␮g/ml leupeptin, 10 mM sodium fluoride, 2 ␮g/ml ␣-1-antitrypsin, 2 mM sodium pyro-


phosphate, 25 mM sodium ␤-glycerophosphate, 1 mM sodium orthovanadate, and 1 mM PMSF. Eighty microliters of the cleared lysates were mixed with 15 ␮l of glutathioneSepharose beads (Amersham Biosciences) and rotated for 2 h at 4 C. The bound proteins were eluted by boiling in SDS sample buffer, subjected to SDS-polyacrylamide gel electrophoresis, and then transferred to polyvinylidene difluoride (PVDF) membranes (Millipore Corp., Billerica, MA). The membranes were probed with an anti-HA monoclonal antibody (12CA5) (Roche Molecular Biochemicals, Indianapolis, IN) and then developed using the ECL kit (Amersham Biosciences) according to the manufacturer’s instruction. Cell Culture and Transient Transfection Assay CV-1, 293T, and COS-7 cells were maintained with DMEM (Gibco BRL), supplemented with 10% fetal bovine serum (BioWhittaker, Walkersville, MD) and antibiotics (Life Technologies, Gaithersburg, MD). Cells were split in 24-well plates at densities of 2–20 ⫻ 104 cells/well the day before transfection. Transient transfections were performed using the SuperFect transfection reagent (QIAGEN Inc.) according to the manufacturer’s instruction. Cells were cotransfected with 0.2 ␮g of RIPE3, RIPE3 Em, (523)4, GK/Luc (⫺1003/⫹196), or p21/Luc (⫺2300/⫹8) reporter plasmids together with 0.1 ␮g of pcDNA3/HA expression vectors for BETA2 and E47 and indicated amounts of pcDNA3/HA-hSHP wild type, and mutants, L98 or R216H. Cells were also cotransfected with 0.2 ␮g of Gal4-tk-luciferase reporter plasmids together with 0.1 ␮g of pCMXGal4N, pCMXGal4-BETA2 or pCMXGal4-E47, and indicated amounts of pcDNA3/HA-SHP, pcDNA3/HAW157X or pcDNA3-p300. Total DNA used in each transfection was adjusted to 1 ␮g/well by adding appropriate amount of pcDNA3 empty vector and 0.1 ␮g of CMV (cytomegalovirus)-␤-galactosidase plasmids were cotransfected as an internal control. Cells were harvested approximately 40–48 h after the transfection for luciferase and ␤-galactosidase assays. The luciferase activity was normalized by ␤-galactosidase activity. Immunostaining Six-week-old mouse pancreas was isolated after perfusion of a solution consisting of 4% paraformaldehyde in PBS. The pancreas was dissected and further fixed in 4% paraformaldehyde in PBS for 16 h at 4 C. After paraffin embedding, sections were performed at 7 ␮M of thickness. Sections were dewaxed and rehydrated into PBS. Microwave antigen retrieval was performed with a citric acid buffer (Biogenex, San Ramon, CA). Sections were incubated with a goat anti-SHP (Santa Cruz Biotechnology, Santa Cruz, CA) at 1:4000 dilution and a rabbit antiglucagon (Linco Research, Inc., St. Louis, MO; at 1:1000). Secondary antibodies used were fluorescein isothiocyanate (FITC)-labeled-donkey antigoat and Cy3-labeled donkey antirabbit (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA). Slides were mounted with Slowfade mounting medium with 4⬘,6-diamidino-2phenylindole (DAPI) (Molecular Probes, Eugene, OR). Confocal Microscopy COS-7 cells were grown on uncoated glass coverslips (Bellco Glass Inc.) and transfected with pEGFP-SHP and pCDNA3/ HA-BETA2 by the LipofectAmine method (Life Technologies Inc.). At 24 h after transfection, cells were washed three times with cold PBS and fixed with 3.7% formaldehyde for 40 min. Fixed cells were mounted on glass slides with PBS and observed with a laser-scanning confocal microscope (Olympus Corp., Lake Success, NY). For detection of pCDNA3/HABETA2, cells mounted on glass slides 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 1⫻ 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 1⫻ PBS, and incubated for 1 h with rhodamine-conjugated antirabbit secondary antibody (Jackson ImmunoResearch Laboratories Inc.) at 37 C. Western Blot Analysis Western blot analysis was performed as previously described (13). Briefly, COS-7 cells were transfected with 2–4 ␮g of pcDNA3/HA-SHP and pcDNA3/HA-W157X or pcDNA3 empty vectors. Forty-eight hours after transfection, cell lysates were prepared and separated on 15% SDS-polyacrylamide gel, and proteins were transferred to nitrocellulose membrane (Amersham Biosciences). The membrane was blocked with 5% nonfat dry milk in 1⫻ TBST buffer and incubated with anti-HA monoclonal antibody (12CA5) (Roche). Membrane was washed three times with 1⫻ TBST and incubated with horseradish peroxidase-conjugated secondary antibody (Amersham Pharmacia Biotech). Bound antibodies were visualized using ECL Kit (Amersham Pharmacia Biotech) according to the manufacturer’s protocol. Gel Mobility Shift Assays Oligonucleotides corresponding to the E-box motif in the rat II insulin gene promoter (48) (5⬘-GAGCCCCTCTGGCCATCTGCTGATCC-3⬘) were end-labeled with [␥-32P]ATP using T4 polynucleotide kinase. Bacterially expressed GST-BETA2 and GST-E47 proteins were purified using GST-sepharose beads (Amersham Pharmacia Biotech) and His-tagged proteins (His-SHP and His-Id-2) were purified as described previously (55). GST fused or His-tagged proteins used in each reaction were analyzed by SDS-PAGE gel and quantified by Coomassie blue staining. Gel shift assays contained 10 mM Tris (pH 8.0), 40 mM KCl, 0.05% Nonidet P-40, 6% glycerol, 1 mM dithiothreitol, and 1 ␮g of poly(dI-dC). One microgram of GST-purified proteins (BETA2 and E47) and increasing amounts of His-SHP or His-Id-2 were mixed with 10,000 cpm of end-labeled oligonucleotide probes in 20 ␮l of each reaction. After 20 min incubation, DNA protein complexes were analyzed on 4% polyacrylamide gel in 1⫻ TBE (90 mM Tris, 90 mM boric acid, 2 mM EDTA). Gels were dried and analyzed by autoradiography.

Acknowledgments We thank Drs. Roland Stein, Jae Woon Lee, Aristidis Moustakas, Hitoshi Okamura, and John Kim Choi for kind gifts of the plasmids. We also thank Drs. Yong-Ho Ahn, Yoon-Kwang Lee, David D. Moore, and In-Kyu Lee for helpful advice and discussion and Dr. Ji-Young Cha, Dr. Wang Li, and Ji-Ho Soe for technical assistance.

Received August 18, 2003. Accepted January 23, 2004. Address all correspondence and requests for reprints to: Hueung-Sik Choi, Ph.D., Hormone Research Center, Chonnam National University, Kwangju 500-757, Republic of Korea. E-mail: [email protected]. This work was supported by Korea Research Foundation Grant KRF-2000-015-DS0037 and in part by Hormone Research Center Grant HRC199920001. M.J.T. is supported by National Institutes of Health Grant HD17379. Present address for S.S.: Department of Biosciences at Novum, Karolinska Institute, S-14157 Huddinge, Sweden.



REFERENCES 1. Seol W, Choi HS, Moore DD 1996 An orphan nuclear hormone receptor that lacks a DNA binding domain and heterodimerizes with other receptors. Science 272: 1336–1339 2. Johansson L, Bavner A, Thomsen JS, Farnegardh M, Gustafsson JA, Treuter E 2000 The orphan nuclear receptor SHP utilizes conserved LXXLL-related motifs for interactions with ligand-activated estrogen receptors. Mol Cell Biol 20:1124–1133 3. Lee YK, Dell H, Dowhan DH, Hadzopoulou-Cladaras M, Moore DD 2000 The orphan nuclear receptor SHP inhibits hepatocyte nuclear factor 4 and retinoid X receptor transactivation: two mechanisms for repression. Mol Cell Biol 20:187–195 4. Seol W, Hanstein B, Brown M, Moore DD 1998 Inhibition of estrogen receptor action by the orphan receptor SHP (short heterodimer partner). Mol Endocrinol 12:1551–1557 5. Seol W, Chung M, Moore DD 1997 Novel receptor interaction and repression domains in the orphan receptor SHP. Mol Cell Biol 17:7126–7131 6. Nishigori H, Tomura H, Tonooka N, Kanamori M, Yamada S, Sho K, Inoue I, Kikuchi N, Onigata K, Kojima I, Kohama T, Yamagata K, Yang Q, Matsuzawa Y, Miki T, Seino S, Kim MY, Choi HS, Lee YK, Moore DD, Takeda J 2001 Mutations in the small heterodimer partner gene are associated with mild obesity in Japanese subjects. Proc Natl Acad Sci USA 98:575–580 7. Shih DQ, Screenan S, Munoz KN, Philipson L, Pontoglio M, Yaniv M, Polonsky KS, Stoffel M 2001 Loss of HNF-1␣ function in mice leads to abnormal expression of genes involved in pancreatic islet development and metabolism. Diabetes 50:2472–2480 8. Goodwin B, Jones SA, Price RR, Watson MA, McKee DD, Moore LB, Galardi C, Wilson JG, Lewis MC, Roth ME, Maloney PR, Willson TM, Kliewer SA 2000 A regulatory cascade of the nuclear receptors FXR, SHP-1, and LRH-1 represses bile acid biosynthesis. Mol Cell 6:517–526 9. Lee YK, Moore DD 2002 Dual mechanisms for repression of the monomeric orphan receptor liver receptor homologous protein-1 by the orphan small heterodimer partner. J Biol Chem 277:2463–2467 10. Sinal CJ, Tohkin M, Miyata M, Ward JM, Lambert G, Gonzalez FJ 2000 Targeted disruption of the nuclear receptor FXR/BAR impairs bile acid and lipid homeostasis. Cell 102: 731–744 11. Gobinet J, Auzou G, Nicolas JC, Sultan C, Jalaguier S 2001 Characterization of the interaction between androgen receptor and a new transcriptional inhibitor, SHP. Biochemistry 40:15369–15377 12. Brendel C, Schoonjans K, Botrugno OA, Treuter E, Auwerx J 2002 The small heterodimer partner interacts with the liver X receptor ␣ and represses its transcriptional activity. Mol Endocrinol 16:2065–2076 13. Sanyal S, Kim JY, Kim HJ, Takeda J, Lee YK, Moore DD, Choi HS 2002 Differential regulation of the orphan nuclear receptor small heterodimer partner (SHP) gene promoter by orphan nuclear receptor ERR isoforms. J Biol Chem 277: 1739–1748 14. Borgius LJ, Steffensen KR, Gustafsson JA, Treuter E 2002 Glucocorticoid signaling is perturbed by the atypical orphan receptor and corepressor SHP. J Biol Chem 277: 49761–49766 15. Ourlin JC, Lasserre F, Pineau T, Fabre JM, Sa-Cunha A, Maurel P, Vilarem MJ, Pascussi JM 2003 The small heterodimer partner interacts with the pregnane X receptor and represses its transcriptional activity. Mol Endocrinol 17: 1693–1703 16. Bavner A, Johansson L, Toresson G, Gustafsson JA, Treuter E 2002 A transcriptional inhibitor targeted by the

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

789

atypical orphan nuclear receptor SHP. EMBO Rep 3:478–484 Klinge CM, Jernigan SC, Risinger KE, Lee JE, Tyulmenkov VV, Falkner KC, Prough RA 2001 Short heterodimer partner (SHP) orphan nuclear receptor inhibits the transcriptional activity of aryl hydrocarbon receptor (AHR)/AHR nuclear translocator (ARNT). Arch Biochem Biophys 390:64–70 Johansson L, Thomsen JS, Damdimopoulos AE, Spyrou G, Gustafsson JA, Treuter E 1999 The orphan nuclear receptor SHP inhibits agonist-dependent transcriptional activity of estrogen receptors ER␣ and ER␤. J Biol Chem 274: 345–353 Lee HK, Lee YK, Park SH, Kim YS, Lee JW, Kwon HB, Soh JM, Moore DD, Choi HS 1998 Structure and expression of the orphan nuclear receptor SHP gene. J Biol Chem 273: 14398–14402 Lee YK, Parker KL, Choi HS, Moore DD 1999 Activation of the promoter of the orphan receptor SHP by orphan receptors that bind DNA as monomers. J Biol Chem 274: 20869–20873 Lai K, Harnish DC, Evans MJ 2003 Estrogen receptor ␣ regulates expression of the orphan receptor SHP. J Biol Chem 278:36418–36429 Kim HJ, Kim JY, Park YY, Choi HS 2003 Synergistic activation of the human orphan nuclear receptor SHP gene promoter by basic helix-loop-helix protein E2A and orphan nuclear receptor SF-1. Nucleic Acids Res 31:6860–6872 Denson LA, Sturm E, Echevarria W, Zimmermann TL, Makishima M, Mangelsdorf DJ, Karpen SJ 2001 The nuclear orphan receptor, shp, mediates bile acid-induced feedback inhibition of the bile acid transporter, ntcp. Gastroenterology 121:140–147 Chiang JY 2002 Bile acid regulation of gene expression: roles of nuclear hormone receptors. Endocr Rev 23: 443–463 Kerr TA, Saeki S, Schneider M, Schaefer K, Berdy S, Redder T, Shan B, Russell DW, Schwarz M 2002 Loss of nuclear receptor SHP impairs but does not eliminate negative feedback regulation of bile acid synthesis. Dev Cell 2:713–720 Wang L, Lee YK, Bundman D, Han Y, Thevananther S, Kim CS, Chua SS, Wei P, Heyman RA, Karin M, Moore DD 2002 Redundant pathways for negative feedback regulation of bile acid production. Dev Cell 2:721–731 Mitchell SM, Weedon MN, Owen KR, Shields B, WilkinsWall B, Walker M, McCarthy MI, Frayling TM, Hattersley AT 2003 Genetic variation in the small heterodimer partner gene and young-onset type 2 diabetes, obesity, and birth weight in U. K. subjects. Diabetes 52:1276–1279 Hung CC, Farooqi IS, Ong K, Luan J, Keogh JM, Pembrey M, Yeo GS, Dunger D, Wareham NJ, O’ Rahilly S 2003 Contribution of variants in the small heterodimer partner gene to birthweight, adiposity, and insulin levels: mutational analysis and association studies in multiple populations. Diabetes 52:1288–1291 Massari ME, Murre C 2000 Helix-loop-helix proteins: regulators of transcription in eucaryotic organisms. Mol Cell Biol 20:429–440 Liu M, Pereira FA, Price SD, Chu MJ, Shope C, Himes D, Eatock RA, Brownell WE, Lysakowski A, Tsai MJ 2000 Essential role of BETA2/NeuroD1 in development of the vestibular and auditory systems. Genes Dev 14:2839–2854 Liu M, Pleasure SJ, Collins AE, Noebels JL, Naya FJ, Tsai MJ, Lowenstein DH 2000 Loss of BETA2/NeuroD leads to malformation of the dentate gyrus and epilepsy. Proc Natl Acad Sci USA 97:865–870 Naya FJ, Huang HP, Qiu Y, Mutoh H, DeMayo FJ, Leiter AB, Tsai MJ 1997 Diabetes, defective pancreatic morphogenesis, and abnormal enteroendocrine differentiation in BETA2/neuroD-deficient mice. Genes Dev 11:2323–2334 Olson E 1990 MyoD family: a paradigm for development? Genes Dev 4:1454–1461


34. Weintraub H, Dwarki V, Verma I, Davis R, Hollenberg S, Snider L, Lassar A, Tapscott S 1991 Muscle-specific transcriptional activation by. MyoD. Genes Dev 5:1377–1386 35. Lassar AB, Davis RL, Wright WE, Kadesh T, Murre C, Voronova A, Baltimore D, Weintraub H 1991 Functional activity of myogenic HLH proteins requires hetero-oligomerization with E12/E47-like proteins in vivo. Cell 66: 305–315 36. Murre C, Mccaw PS, Vaessin H, Caudy M, Jan LY, Jan YN, Cabrera CV, Buskin JN, Hauschka SD, Lassar AB, Weintraub H, Baltimore D 1989 Interactions between heterologous helix-loop-helix proteins generate complexes that bind specifically to a common DNA sequence. Cell 58: 537–544 37. Zhuang Y, Cheng P, Weintraub H 1996 B-lymphocyte development is regulated by the combined dosage of three basic helix-loop-helix genes, E2A, E2–2, and HEB. Mol Cell Biol 16:2898–2905 38. Lee JE, Hollenberg SM, Snider L, Turner DL, Lipnick N, Weintraub H 1995 Conversion of Xenopus ectoderm into neurons by NeuroD, a basic helix-loop-helix protein. Science 268:836–844 39. Mutoh H, Naya FJ, Tsai MJ, Leiter AB 1998 The basic helix-loop-helix transcription factor BETA2 interacts with p300 to coordinate differentiation of secretin-expressing enteroendocrine cells. Genes Dev 12:820–830 40. Mutoh H, Fung BP, Naya FJ, Tsai MJ, Nishitani J, Leiter AB 1997 The basic helix-loop-helix transcription factor BETA2/ NeuroD is expressed in mammalian enteroendocrine cells and activates secretin gene expression. Proc Natl Acad Sci USA 94:3560–3564 41. Poulin G, Turgeon B, Drouin J 1997 NeuroD1/BETA2 contributes to cell-specific transcription of the proopiomelanocortin gene. Mol Cell Biol 17:6673–6682 42. Sharma A, Moore M, Marcora E, Lee JE, Qiu Y, Samaras S, Stein R 1999 The NeuroD1/BETA2 sequences essential for insulin gene transcription colocalize with those necessary for neurogenesis and p300/CREB binding protein binding. Mol Cell Biol 19:704–713 43. Malecki, MT, Jhala US, Antonellis A, Fields L, Doria A, Orban T, Saad M, Warram JH, Montminy M, Krolewski AS 1999 Mutations in NEUROD1 are associated with the development of type 2 diabetes mellitus. Nat Genet 23: 323–328 44. Kim JW, Seghers V, Cho JH, Kang Y, Kim S, Ryu Y, Baek K, Aguilar-Bryan L, Lee YD, Bryan J, Suh-Kim H 2002 Transactivation of the mouse sulfonylurea receptor I gene by BETA2/NeuroD. Mol Endocrinol 16:1097–1107 45. Moates JM, Nanda S, Cissell MA, Tsai MJ, Stein R 2003 BETA2 Activates transcription from the upstream glucokinase gene promoter in islet ␤-cells and gut endocrine cells. Diabetes 52:403–408 46. Breslin MB, Zhu M, Lan MS 2003 NeuroD1/E47 regulates the E-box element of a novel zinc-finger transcription factor, IA-1, in developing nervous system. J Biol Chem 278: 38991–38997 47. Norton JD 2000 ID helix-loop-helix proteins in cell growth, differentiation and tumorigenesis. J Cell Sci 113:3897–3905 48. Naya FJ, Stellrecht CMM, Tsai MJ 1995 Tissue-specific regulation of the insulin gene by a novel basic helix-loophelix transcription factor. Genes Dev 9:1009–1019 49. Miyata T, Maeda T, Lee JE 1999 NeuroD is required for differentiation of the granule cells in the cerebellum and hippocampus. Genes Dev 13:1647–1652 50. Kim WY, Fritzsch B, Serls A, Bakel LA, Huang EJ, Reichardt LF, Barth DS, Lee JE 2001 NeuroD-null mice are deaf due to a severe loss of the inner ear sensory neurons during development. Development 128:417–426


51. Dumonteil E, Laser B, Constant I, Philippe J 1998 Differential regulation of the glucagon and insulin I gene promoters by the basic helix-loop-helix transcription factors E47 and BETA2. J Biol Chem 273:19945–19954 52. Qiu Y, Sharma A, Stein R 1998 p300 Mediates transcriptional stimulation by the basic helix-loop-helix activators of the insulin gene. Mol Cell Biol 18:2957–2964 53. Qiu Y, Guo M, Huang S, Stein R 2002 Insulin gene transcription is mediated by interactions between the p300 coactivator and PDX-1, BETA2, and E47. Mol Cell Biol 22:412–420 54. Glick E, Leshkowitz D, Walker MD 2000 Transcription factor BETA2 acts cooperatively with E2A and PDX1 to activate the insulin gene promoter. J Biol Chem 275:2199–2204 55. Ohneda K, Mirmira RG, Wang JH, Johnson JD, German MS 2000 The homeodomain of PDX-1 mediates multiple protein-protein interactions in the formation of a transcriptional activation complex on the insulin promoter. Mol Cell Biol 20:900–911 56. Yagita K, Yamaguchi S, Tamanini F, van Der Horst GT, Hoeijmakers JH, Yasui A, Loros JJ, Dunlap JC, Okamura H 2000 Dimerization and nuclear entry of mPER proteins in mammalian cells. Genes Dev 14:1353–1363 57. Ruezinsky D, Beckmann H, Kadesch T 1991 Modulation of the IgH enhancer’s cell type specificity through a genetic switch. Genes Dev 5:29–37 58. Shen CP, T. Kadesch T 1995 B-cell-specific DNA binding by an E47 homodimer. Mol Cell Biol 15:4518–4524 59. Carter RS, Ordentlich P, Kadesch T 1997 Selective utilization of basic helix-loop-helix-leucine zipper proteins at the immunoglobulin heavy-chain enhancer. Mol Cell Biol 17: 18–23 60. Zhao F, Vilardi A, Neely RJ, Choi JK 2001 Promotion of cell cycle progression by basic helix-loop-helix E2A. Mol Cell Biol 21:6346–6357 61. Chipuk JE, Cornelius SC, Pultz NJ, Jorgensen JS, Bonham MJ, Kim SJ, Danielpour D 2002 The androgen receptor represses transforming growth factor-␤ signaling through interaction with Smad3. J Biol Chem 277:1240–1248 62. Ray A, Prefontaine KE 1994 Physical association and functional antagonism between the p65 subunit of transcription factor NF-␬B and the glucocorticoid receptor. Proc Natl Acad Sci USA 91:752–756 63. Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B, Lin SC, Heyman RA, Rose DW, Glass CK, Rosenfeld MG 1996 A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 85:403–414 64. McNamara P, Seo SB, Rudic RD, Sehgal A, Chakravarti D, FitzGerald GA 2001 Regulation of CLOCK and MOP4 by nuclear hormone receptors in the vasculature: a humoral mechanism to reset a peripheral clock. Cell 105:877–889 65. Langlands K, Yin X, Anand G, Prochownik EV 1997 Differential interactions of Id proteins with basic-helix-loop-helix transcription factors. J Biol Chem 272:19785–19793 66. Cha JY, Kim HI, Im SS, Li TZ, Ahn YH 2001 HNF1 and/or HNF3 may contribute to the tissue specific expression of glucokinase gene. Exp Mol Med 33:59–63 67. Pardali K, Kurisaki A, Moren A, ten Dijke P, Kardassis D, Moustakas A 2000 Role of Smad proteins and transcription factor Sp1 in p21(Waf1/Cip1) regulation by transforming growth factor-␤. J Biol Chem 275:29244–29256 68. Jung H, Kim T, Chae HZ, Kim KT, Ha H 2001 Regulation of macrophage migration inhibitory factor and thiol-specific antioxidant protein PAG by direct interaction. J Biol Chem 276:15504–15510

Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.