• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
Mol Cell Biol. Jul 2008; 28(14): 4588–4597.
Published online May 12, 2008. doi:  10.1128/MCB.01191-07
PMCID: PMC2447131

The MODY1 Gene for Hepatocyte Nuclear Factor 4α and a Feedback Loop Control COUP-TFII Expression in Pancreatic Beta Cells[down-pointing small open triangle]


Pancreatic islet beta cell differentiation and function are dependent upon a group of transcription factors that maintain the expression of key genes and suppress others. Knockout mice with the heterozygous deletion of the gene for chicken ovalbumin upstream promoter-transcription factor II (COUP-TFII) or the complete disruption of the gene for hepatocyte nuclear factor 4α (HNF4α) in pancreatic beta cells have similar insulin secretion defects, leading us to hypothesize that there is transcriptional cross talk between these two nuclear receptors. Here, we demonstrate specific HNF4α activation of a reporter plasmid containing the COUP-TFII gene promoter region in transfected pancreatic beta cells. The stable association of the endogenous HNF4α with a region of the COUP-TFII gene promoter that contains a direct repeat 1 (DR-1) binding site was revealed by chromatin immunoprecipitation. Mutation experiments showed that this DR-1 site is essential for HNF4α transactivation of COUP-TFII. The dominant negative suppression of HNF4α function decreased endogenous COUP-TFII expression, and the specific inactivation of COUP-TFII by small interfering RNA caused HNF4α mRNA levels in 832/13 INS-1 cells to decrease. This positive regulation of HNF4α by COUP-TFII was confirmed by the adenovirus-mediated overexpression of human COUP-TFII (hCOUP-TFII), which increased HNF4α mRNA levels in 832/13 INS-1 cells and in mouse pancreatic islets. Finally, hCOUP-TFII overexpression showed that there is direct COUP-TFII autorepression, as COUP-TFII occupies the proximal DR-1 binding site of its own gene in vivo. Therefore, COUP-TFII may contribute to the control of insulin secretion through the complex HNF4α/maturity-onset diabetes of the young 1 (MODY1) transcription factor network operating in beta cells.

Chicken ovalbumin upstream promoter-transcription factor II (COUP-TFII, also called NR2F2) is an orphan member of the steroid/thyroid hormone receptor superfamily classed in the same subfamily as hepatocyte nuclear factor 4α (HNF4α)/maturity-onset diabetes of the young 1 (MODY1) and retinoid X receptor (RXR) (4, 11). Several molecular mechanisms by which COUP-TFII controls gene expression in pancreatic islet beta cell differentiation and function have been shown previously. COUP-TFII binds DNA by a Zn finger DNA binding domain in a variety of hormone response elements (HRE) that contain imperfect AGGTCA direct or inverted repeats with various spacing patterns (3, 14). It can form heterodimeric complexes with RXR, the universal partner of many nuclear receptors, and as such acts as a repressor (15). We previously showed that COUP-TFII acts as an inhibitor of the glucose activation of the liver pyruvate kinase gene by binding to the glucose-responsive element (9). On most promoters, HNF4α response elements are also bound by COUP-TFII, which often behaves as a transcriptional repressor antagonizing the enhancement of transcription by HNF4α (8, 12, 24). In a functional study, the impaired synergy between COUP-TFII and the E276Q mutant form of human HNF4α on the HNF1 promoter was found to be due to their altered interactions (21).

Recently, we showed that heterozygous COUP-TFII gene deletion in mouse pancreatic beta cells leads to impaired glucose sensitivity and abnormal insulin secretion (1). These mutant mice presented hyperinsulinemia in fasting and fed states and impaired glucose tolerance. Interestingly, mice with the complete disruption of the HNF4α gene in beta cells have a similar phenotype, i.e., hyperinsulinemia in fasting and fed states and impaired glucose tolerance (6, 16) and glucose-stimulated insulin secretion defects (13). These observations raised the question of the possible interdependency of COUP-TFII and HNF4α. To address this issue, we investigated the capacity of HNF4α to regulate COUP-TFII expression and the possible cross-regulation of HNF4α and COUP-TFII. We also tested the idea that the expression of COUP-TFII, like that of some other transcription factors, is autoregulated.


Cell culture.

The rat insulinoma 832/13 INS-1 cell line (7), generously provided by C. Newgard, was used between passages 19 and 29. Cells were cultured in 5% CO2-95% O2 at 37°C in INS-1 medium (RPMI 1640 medium containing 11 mM d-glucose supplemented with 10% [vol/vol] heat-inactivated fetal bovine serum, 100 U of penicillin/ml, 100 U of streptomycin/ml, 10 mM HEPES, 1 mM sodium pyruvate [Invitrogen], and 50 μM beta-mercaptoethanol [Sigma]). Cells of the DN-HNF4α-26 line, a derivative of the INS-1 cell line that contains a plasmid encoding a doxycycline-inducible dominant negative form of HNF4α (DN-HNF4α), were maintained as described before (25). To induce DN-HNF4α, cells were incubated for 24 h with 500 ng of doxycycline/ml, and total RNA was extracted 24 h later.

Pancreatic islet isolation and culture.

Pancreatic islets from 6- to 8-week-old male C57BL/6J mice were isolated by an adapted collagenase digestion method (1). Briefly, mice were anesthetized with a mixture of 3.5 × 105 Pa of isoflurane and 0.5 × 105 Pa of oxygen (Minerve), and type V collagenase P (Roche) was injected into the common bile ducts. Infused and distended pancreases were then removed and left to digest for 4 min at 37°C with gentle mixing. Islets were washed and handpicked in HEPES balanced salt solution (124 mM NaCl, 5 mM KCl, 0.8 mM MgSO4·7H2O, 1 mM NaH2PO4, 10 mM HEPES, 1.8 mM CaCl2, 14 mM NaHCO3, and 0.5% bovine serum albumin [BSA; Sigma], pH 7.4) containing 3 mM glucose under an inverted light microscope and were then separated into study groups. Islets were cultured overnight in RPMI 1640 medium (Invitrogen) containing 11 mM glucose supplemented with 10% heat-inactivated fetal bovine serum (Invitrogen), 1 mM sodium pyruvate (Invitrogen), 100 U of penicillin/ml, and 100 U of streptomycin (Invitrogen)/ml.


Pancreases were removed from 4-month-old male C57/BL6N mice, fixed overnight in 4% paraformaldehyde, and embedded in paraffin. Blocks were serially sectioned (5-μm thickness). Sections were immunostained for insulin by using a guinea pig anti-insulin antibody (1:2,000 dilution of A0564 from Dakocytomation), after which they were incubated with biotin-labeled goat anti-guinea pig immunoglobulin G (IgG) and then with peroxidase-labeled streptavidin, and results were developed with DAB (3, 3′-diaminobenzidine-tetrahydrochloride; Vectastain from Vector Laboratories). For COUP-TF staining, sections were incubated for 2 h with primary anti-human COUP-TFI/NR2F1 (PP-H8132-00 at a 1:100 dilution) or COUP-TFII/NR2F2 (PP-H7147-00 at a 1:50 dilution) mouse monoclonal antibody (Perseus Proteomics) in a solution of 3% BSA, 0.05% Tween 20, and phosphate-buffered saline (PBS). An indirect peroxidase labeling technique coupled with development by DAB from the EnVision+ System-HRP (Dakocytomation) was then used. For immunofluorescence staining, 832/13 INS-1 cells were grown on glass coverslips treated with poly-l-lysine (Sigma). Cells were washed and fixed in 4% paraformaldehyde for 10 min and blocked and permeabilized with a solution of 2% BSA and 0.05% Triton X-100 in PBS for 10 min. Then cells were incubated with the first COUP-TFII antibody (1:100), followed by a fluorescein isothiocyanate-conjugated goat anti-rabbit secondary antibody (1:400; Jackson ImmunoResearch). Cells were then mounted onto slides using Vectashield mounting medium with DAPI (4′-6-diamidino-2-phenylindole; Vector Laboratories).

Preparation of a recombinant virus expressing COUP-TFII and adenovirus infection.

The full-length human COUP-TFII cDNA was inserted into the KpnI/XhoI sites of the pAdTrack-CMV shuttle vector, which also contains a cytomegalovirus promoter directing the green fluorescent protein (GFP) reporter gene for monitoring the efficiency of adenovirus infection. The resulting recombinant adenovirus plasmid was designated Ad-hCOUP-TFII. Ad-hCOUP-TFII and the control plasmid (pAdTrack with no exogenous gene) were produced by INSERM, U649, Nantes, France. 832/13 INS-1 cells were seeded into 12-well tissue culture plates at a density of 0.9 × 106 cells/well in INS-1 medium and, 24 h later, were exposed overnight to adenovirus at 2 to 5 PFU per cell. The virus-containing medium was removed the next day and replaced by fresh medium for 36 h before the extraction of RNA (described below). After isolation, pancreatic islets were cultured for 2 h in 11 mM glucose before being exposed to adenovirus at 200 PFU/cell for 1 h and 30 min. After infection, islets were cultured in fresh 11 mM glucose medium for 3 days before RNA or protein extraction.

Nuclear extract preparation and EMSA.

Nuclear extracts from 832/13 INS-1 cells and mouse liver cells were prepared as described in references 1 and 9. The double-stranded oligonucleotide 5′-TGC AGC AGT CGT GTC AAA GTT CAC TAT ATA GAG-3′ was used as the COUP-TFII direct repeat 1 (DR-1) probe. The double-stranded M oligonucleotide 5′-TGC AGC AGT CGT GAT GCA TTT CAC TAT ATA GAG-3′ was used as a COUP-TFII DR-1 mutant form. Electrophoresis mobility shift assays (EMSA), probe labeling, binding reactions, and competitor and antibody supershift analyses were performed as reported previously (9). For the supershift analyses with the HNF4α antibody (kindly provided by M. Pontoglio and F. Ringeisen) and the COUP-TFII/NR2F2 antibody (PP-H7147-00 [1:500; Perseus Proteomics]), 1 μl of antiserum was included in the binding reaction mixtures.

Plasmids and site-directed mutagenesis.

The reporter plasmid −3000/luc (corresponding to positions −3047 to +873 relative to the defined COUP-TFII gene transcription initiation site [20]) from the fragment described previously (27) was subcloned into the pGL3-Basic vector (Promega) by using BamHI/BglII restriction sites. Then fragments from −688 (EcoRI), −328 (ApaI), −48 (SacI), +202 (FspI), +418 (ApaI), and +639 (SacI) to +873 relative to the transcription initiation site were subcloned into the pGL3-Basic plasmid. Site-directed mutagenesis of the reporter plasmid containing the fragment from −328 to +873 and a luciferase reporter gene (plasmid −328/luc) was performed by GenScript, NJ, to introduce five point mutations at the DR-1 site, as in the sequence of the double-stranded M oligonucleotide, yielding the plasmid −328 M/luc. Isolated clones were totally sequenced. The Renilla luciferase coding sequence from the pRL null vector (Promega) is controlled by the Rous sarcoma virus promoter (9). pcDNA3, encoding HNF4α, and pcDNA, encoding DN-HNF4α, have been described previously (22).

Transfection and reporter gene assay.

Transient transfections were carried out with Lipofectamine 2000 reagent (Invitrogen). Transfected cells were cultured in INS-1 medium and harvested 24 h after transfection. Cell extracts were assayed for reporter enzyme activities by using the dual luciferase kit (Promega) as described previously (9).

siRNA-mediated silencing and transfection.

A 21-nucleotide RNA molecule was designed and synthesized by Qiagen SA (2-For-Silencing small interfering RNA [siRNA]). This siRNA sequence targeted mouse and rat COUP-TFII RNA molecules (GenBank accession numbers NM_009697 and NM_080778, respectively) but not the human COUP-TFII RNA molecule (GenBank accession number NM_021005) or the mouse COUP-TFI RNA molecule (GenBank accession number NM_010151) and corresponded to position 1023 relative to the mouse gene start codon. The COUP-TFII siRNA sequences were AGUGUGCUUUGGAAGAGUAdTdT (sense) and UACUCUUCCAAAGCACACUdGdG (antisense); the nonspecific control siRNA used was from Qiagen SA. 832/13 INS-1 cells were grown to 75 to 80% confluence in 10-cm-diameter dishes. Then cells were trypsinized and transiently transfected by electroporation with the Amaxa Nucleofector II device with solution T and program T20 (Amaxa Biosystems) by using 68 pmol of siRNA duplex per 1.2 × 106 cells. Nucleofection was done with cells in INS-1 medium in 12-well plates, the culture medium was changed 24 h after electroporation, and RNA and protein extractions were done 24 h later.

Isolation of mRNA from 832/13 INS-1 cells and mouse pancreatic islets and detection by real-time quantitative PCR (RT-QPCR).

Total RNA was extracted from cultured cells and purified using the RNA-PLUS reagent according to the instructions of the manufacturer (Q-BIOgene). Reverse transcription was performed with 2 μg of total RNA by using SuperScript II reverse transcriptase (Invitrogen) or 1 μg of total RNA by using the iScript cDNA synthesis kit (Bio-Rad), depending on the gene and according to the manufacturers’ protocols.

Total RNA was extracted from handpicked islets by using the Absolutely RNA microprep kit according to the instructions of the manufacturer (Stratagene). Each RNA or nuclear extract sample was prepared from a total of 200 to 400 islets from about three mice. RNA (0.5 μg) was reverse transcribed for RT-QPCR.

RT-QPCR was performed with 6.25 ng of reverse-transcribed total RNA and a mixture containing a 10 μM concentration of each primer (Eurogentec), 2 mM MgCl2, and 1× LightCycler DNA master Sybr green I mix in a LightCycler apparatus (Roche). The relative amounts of the different mRNAs were quantified using the threshold cycle methodology. All samples were normalized to the threshold cycle value of the cyclophilin reference mRNA. Forward and reverse primers used for the specific amplification of cDNA fragments, designed to hybridize to rat transcripts, were as follows: 5′-CAG AGC CAG CAG CAC ATC GAG-3′ and 5′-TTA AGT TCC TGC GGA CGC TCC T-3′ for the COUP-TFI gene (121 bp), 5′-CGC TCC TTG CCG CTG CT-3′ and 5′-AAG AGC TTT CCG AAC CGT GTT-3′ for the COUP-TFII gene (289 bp), 5′-TTG CCA TTC CTG GAC CCA AA-3′ and 5′-ATG GCA CTG GTG GCA AGT CC-3′ for the cyclophilin gene (325 bp), 5′-AAA TGT GCA GGT GTT GAC CA-3′ and 5′-CAC GCT CCT CCT GAA GAA TC-3′ for the HNF4α gene (178 bp), 5′-AGC AGT GCT GGC TAC CTT CAA-3′ and 5′-AAT ATG TAG CCA CCC CCT TGG-3′ for the peroxisome proliferator-activated receptor α (PPARα) gene (97 bp), 5′-GCC CAG CTT AAT GCC ATC TTT-3′ and 5′-CAA AAG GGC TGC CTT CTG TAA-3′ for the BETA2/Neuro-D1 gene (113 bp), and 5′-GCG CTG AGA GTC CGT GAG-3′ and 5′-CCG GGG TAG GGA GCT ACA-3′ for the Pdx1 gene (60 bp).

We checked that these COUP-TFII primers specifically amplified the rat (and mouse) COUP-TFII gene; they did not amplify the COUP-TFII gene from cDNA prepared from a human cell line expressing COUP-TFII.

Protein analysis by Western blotting and enhanced chemiluminescence detection.

For total-protein extraction from 832/13 INS-1 or INS-1 DN-HNF4α-26 cells, 0.9 × 106 cells were washed in cold PBS. Cell pellets were lysed in 100 μl of lysis buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 5 mM EDTA, 30 mM Na4P2O7, 50 mM NaF with 1% Triton, 10 mg of leupeptin/ml, 10 mg of pepstatin/ml, 10 mg of aprotinin/ml, and 1 mM phenylmethylsulfonyl fluoride). Nuclear extracts from pancreatic islets were obtained using NE-PER nuclear and cytoplasmic extraction reagents (Pierce). Immunoblotting procedures were as described previously (1). Blots were developed with ECL SuperSignal West Pico chemiluminescent reagents (Pierce) and visualized using the high-end LAS-3000 imaging system (Fujifilm). Bands were quantified by densitometry using the Multigauge 3.0 image processor program (Fujifilm), and band intensities were normalized to those obtained for the appropriate loading controls. Dilutions and sources of antibodies were as follows: anti-Myc tag, 1:1,000 (catalog no. sc-40; Santa Cruz); anti-COUP-TFII/NR2F2, 1:500 (PP-H7147-00; Perseus Proteomics); anticyclophilin, 1:3,000 (catalog no. 07-313; Upstate); and anti-GAPDH (glyceraldehyde-3-phosphate dehydrogenase), 1:200 (catalog no. sc-25778; Santa Cruz).

ChIP assays.

832/13 INS-1 cells were cultured in INS-1 medium, and chromatin immunoprecipitation (ChIP) was done as described previously (17). Supernatants were incubated with polyclonal antibodies directed against HNF4α (5 μl/mg of protein [catalog no. sc-8987; Santa Cruz Biotechnologies, Santa Cruz, CA]), with normal rabbit IgG (catalog no. sc-2027; Santa Cruz) as the control, or a monoclonal antibody directed against COUP-TFII/NR2F2 (4 μg/mg of protein [PP-H7147-00; Perseus Proteomics]), with normal mouse IgG (catalog no. 12-371; Upstate) as the control. The primer sequences used to amplify the COUP-TFII promoter region by PCR were 5′-GCT AGG ACC GGG CTG TTC-3′ and 5′-TGA ACT TTG ACA CGA CTG CTG-3′. The PCR primers for the COUP-TFII gene coding region were 5′-CAG CAG CAG CAC ATC GAG-3′ and 5′-GGC ACT ACT GGC ACT GGT TG-3′.

Statistical analysis.

Quantitative results are expressed as the mean with the standard deviation. The Mann-Whitney test, a nonparametric statistical program accepted to be appropriate when the number of experiments is less than 10, was used for statistical analyses, and null hypotheses were rejected at P values of ≥0.05. All experiments were performed at least three times.


COUP-TFII is expressed in adult mouse pancreatic islet beta cells and in the 832/13 INS-1 beta cell line.

We found previously using rabbit polyclonal COUP-TF antibodies that COUP-TFII and COUP-TFI are expressed in mouse pancreatic islets (27). Here, using mouse monoclonal antibodies specific to COUP-TFII and COUP-TFI, we observed that COUP-TFII protein expression was restricted to mouse (Fig. (Fig.1A)1A) and human (data not shown) pancreatic beta cells. We detected COUP-TFI protein in mouse islet nonbeta cells (Fig. (Fig.1C).1C). These results suggest that these two proteins have specific functions in mouse pancreatic islets. We then analyzed COUP-TFII protein expression in a beta cell line, 832/13 INS-1 (7), and found that COUP-TFII protein was expressed in the nuclei of these cells (Fig. 1E and F). This cell line was used as a model to study whether HNF4α can regulate COUP-TFII expression at the molecular level.

FIG. 1.
COUP-TFII expression in adult mouse pancreas and in the 832/13 INS-1 cell line. (A to F) Immunostaining of pancreatic sections from adult mice by using mouse monoclonal antibodies against COUP-TFII (A), insulin (B and D), and COUP-TFI (C) and a secondary ...

HNF4α binds and transactivates the COUP-TFII gene promoter.

We reported previously that the first 3,000 nucleotides upstream of the transcription initiation site are sufficient for the transcription of the mouse COUP-TFII gene in insulin-positive cells from transgenic mice (27). To identify key promoter elements within this region, 832/13 INS-1 cells were transiently transfected with reporter plasmids carrying the luciferase gene controlled by various portions of the mouse COUP-TFII gene 5′ regulatory region (Fig. (Fig.2A).2A). This deletion analysis identified two major regions responsible for different promoter activities: a negative element between nucleotides −688 and −328 relative to the transcription start site, the deletion of which led to increased activity, and a strong positive element between nucleotides −328 and −48.

FIG. 2.
Mapping the COUP-TFII promoter and its transactivation by HNF4α. (A) 832/13 INS-1 cells were transiently cotransfected with a luciferase reporter (luc) gene driven by various lengths of the mouse COUP-TFII promoter, designated by their 5′-end ...

The comparison of transcription factor DNA binding motifs in the latter region in different species revealed a conserved HRE, a DR-1 with a single-base-pair spacer, that is known to bind nuclear hormone receptors, including HNF4α (Fig. (Fig.2B)2B) (14). Thus, we tested the ability of HNF4α to bind the DR-1 binding site of the COUP-TFII gene promoter in vitro and in vivo. In antibody-mediated supershift assays using a probe containing the DR-1 element (Fig. (Fig.3A),3A), the incubation of 832/13 INS-1 cell nuclear extracts with the radiolabeled COUP-TFII DR-1 probe resulted in the formation of several complexes. The specificities of these complexes were analyzed by competition with unrelated oligonucleotides and displacement or supershifting by specific antibodies. All the complexes were specific, as they could be displaced by a 50-fold excess of the unlabeled COUP-TFII DR-1 oligonucleotide (Fig. (Fig.3A,3A, lane 5) but not by an unrelated oligonucleotide (Fig. (Fig.3A,3A, lane 6). The faster-migrating band corresponded to HNF4α binding activity, as it was displaced by anti-HNF4α antibodies (Fig. (Fig.3A,3A, lane 2) whereas the addition of anti-upstream stimulatory factor (anti-USF) antibodies did not affect this binding activity (Fig. (Fig.3A,3A, lane 3). In addition, none of the binding activities were altered by the inclusion of a 50-fold excess of an unlabeled mutant version of the COUP-TFII DR-1 oligonucleotide (Fig. (Fig.3A,3A, lane 4). These results suggest that all the complexes could contain members of the nuclear receptor superfamily, including HNF4α. Liver cell nuclear extracts known to contain abundant HNF4α proteins were tested by EMSA. In this case, stronger complexes that could be displaced with the anti-HNF4α serum were observed. These results suggest that this DR-1 element is an HNF4α DNA binding site. The in vivo relevance of the observed HNF4α binding activity was analyzed in the context of chromatin in intact 832/13 INS-1 cells by ChIP of endogenous HNF4α (Fig. (Fig.3B).3B). HNF4α antibodies efficiently and specifically immunoprecipitated the COUP-TFII promoter DNA, indicating that there is a stable association between the endogenous factor and the COUP-TFII promoter in vivo.

FIG. 3.
HNF4α binds to the COUP-TFII DR-1 site in vitro and in vivo. (A) Results of EMSA using the COUP-TFII DR-1 probe are shown. The probe was incubated with 832/13 INS-1 cell nuclear extracts (lanes 1 to 6) and mouse liver cell nuclear extracts (lanes ...

To test the involvement of HNF4α in the regulation of the bp −328 region of the COUP-TFII gene, we first transfected 832/13 INS-1 cells with the −328/luc reporter construct in the absence or presence of a wild-type HNF4α expression plasmid. Figure Figure2C2C shows that HNF4α expression generated 25-fold activation of the promoter but did not activate the bp −48 COUP-TFII gene construct lacking the DR-1 DNA binding site. To further characterize the specific HNF4α-dependent activation, we used the selective DN-HNF4α mutant protein, which specifically forms defective heterodimers with wild-type HNF4α, thereby preventing DNA binding and hence transcriptional activation by HNF4α (5). DN-HNF4α had no effect on the binding of PPARγ-RXRα heterodimers to a PPAR response element (22) or on the binding of COUP-TFII to a DR-1 site in these cells (data not shown). The overexpression of DN-HNF4α results in 40% inhibition of the basal activity of the COUP-TFII gene promoter bp −328 region evoked by endogenous HNF4α expression (Fig. (Fig.2C,2C, compare bars 1 and 3) and suppresses transactivation by HNF4α (Fig. (Fig.2C,2C, compare bar 2 and bars 4 and 5). To definitively demonstrate the implication of HNF4α in the control of the COUP-TFII gene promoter bp −328 region, critical bases in the DR-1 site were mutated. The mutant sequence failed to bind HNF4α, as shown by EMSA (Fig. (Fig.3A,3A, lane 4). The activity of the mutant reporter plasmid (−328 M/luc) in 832/13 INS-1 cells was then tested. The mutant showed a lower basal level of activity, 50% of the wild-type activity (Fig. (Fig.2C,2C, compare bars 6 and 1). In addition, the mutations in the DR-1 site led to a complete loss of HNF4α transactivation (Fig. (Fig.2C,2C, compare bars 7 and 8 and bar 6). Thus, we conclude that the COUP-TFII promoter is specifically activated by HNF4α.

Mechanism of interdependency of the COUP-TFII and HNF4α genes.

Given the strength of the HNF4α transactivation effect on the −328 COUP-TFII promoter construct, we wanted to assess the influence of transactivation on endogenous COUP-TFII expression in beta cells. We used DN-HNF4α-26 cells, which are derived from the INS-1 cell line and contain a plasmid encoding doxycycline-inducible DN-HNF4α (25). As shown in Fig. Fig.4A,4A, in the presence of doxycycline, the c-Myc-tagged DN-HNF4α protein was induced. As a consequence of the dominant negative suppression of HNF4α function, there was a significant decrease in endogenous COUP-TFII gene expression (Fig. (Fig.4B4B).

FIG. 4.
COUP-TFII mRNA expression in INS-1 DN-HNF4α-expressing cells. Cells were cultured in the presence (+) or absence (−) of 500 ng of doxycycline/ml for 24 h. (A) The upper panel shows DN-HNF4α expression as demonstrated by ...

A transcriptional network with the reciprocal activation of the HNF1α and HNF4α genes has been described previously (10). These observations led us to test whether such reciprocal cross-regulation between COUP-TFII and HNF4α occurs in pancreatic beta cells in two ways. Firstly, endogenous COUP-TFII expression was down-regulated, almost completely, by introducing a specific COUP-TFII siRNA into 832/13 INS-1 cells by electroporation (Fig. 5A and B). Under these conditions, HNF4α gene transcript levels were decreased by 40% (Fig. (Fig.5C).5C). Secondly, we overexpressed COUP-TFII by using a recombinant adenovirus encoding a human COUP-TFII protein (Ad-hCOUP-TFII) in 832/13 INS-1 cells and in mouse pancreatic islets. Based on GFP reporter expression, 80% of 832/13 INS-1 cells and 50% of islet beta cells were infected (data not shown). EMSA using the COUP-TFII binding probe showed that COUP-TFII was strongly expressed in 832/13 INS-1 cells (Fig. (Fig.6A).6A). Similarly, after the infection of pancreatic islets with Ad-hCOUP-TFII, COUP-TFII was readily detected by Western blotting (Fig. (Fig.6A).6A). As shown in Fig. Fig.6B,6B, adenovirus-mediated induction of COUP-TFII resulted in a 300% increase in HNF4α mRNA expression in 832/13 INS-1 cells and a 100% increase in HNF4α mRNA expression in pancreatic islets. A known target of HNF4α, the expression of the nuclear receptor PPARα, was also induced under these conditions (Fig. (Fig.6C).6C). Together, these results indicate that there is cross talk between HNF4α and COUP-TFII.

FIG. 5.
HNF4α mRNA expression in 832/13 INS-1 cells transfected with COUP-TFII siRNA. (A and C) 832/13 INS-1 cells were transfected with unrelated (control) or COUP-TFII-specific siRNA by electroporation and cultured for 48 h in INS-1 medium. Total RNA ...
FIG. 6.
Effects of adenovirus-mediated overexpression of COUP-TFII in 832/13 INS-1 and mouse pancreatic beta cells. 832/13 INS-1 cells and islets were cultured in INS-1 medium and infected (at 2 or 5 PFU/cell and 200 PFU/cell, respectively) with adenovirus expressing ...

Autoregulation of the COUP-TFII promoter.

The observation of a COUP-TFII-DR-1 complex, as shown in Fig. Fig.6A6A (left panel), led us to hypothesize that COUP-TFII might control its own expression. We tested whether endogenous rat COUP-TFII mRNA expression was modified when 832/13 INS-1 cells were infected with the human COUP-TFII-expressing adenovirus. As shown in Fig. Fig.6D,6D, Ad-hCOUP-TFII decreased the endogenous rat COUP-TFII mRNA expression by 30%. Next, we assessed in vivo promoter occupancy by ChIP from 832/13 INS-1 cells. Endogenous COUP-TFII occupied the region of the rat COUP-TFII gene promoter that contains the DR-1 DNA binding site, between nucleotides −57 and −69 (Fig. (Fig.6E).6E). These data suggest that COUP-TFII has the capacity to directly repress its expression.


Until now, no ligand has been found for COUP-TFII, which exhibits constitutive transcriptional activity as an activator or a suppressor in different cell types. It is therefore of interest to identify the transcription factors that modulate COUP-TFII gene expression. The central issue addressed here was the transcriptional regulation of the murine COUP-TFII gene by the nuclear receptor HNF4α/MODY1. Unexpectedly, HNF4α expression was found to be controlled by COUP-TFII expression in pancreatic beta cells.

We have shown here that HNF4α binds to a DR-1 DNA binding site in the proximal bp −328 region of the COUP-TFII promoter in vitro and in vivo and activates both reporter gene expression and endogenous COUP-TFII gene expression in beta cells. When cells were transiently transfected with a series of promoter-reporter constructs, we observed that the deletion of an upstream region (bp −688 to −328) significantly increased the activity of the bp −328 region of the COUP-TFII promoter. The bp −688 construct still mediated HNF4α activation, but this response was less than that with the bp −328 construct (data not shown). We speculate that the repressive element may modulate the endogenous activation of COUP-TFII by HNF4α and may explain the lesser effect of HNF4α on endogenous COUP-TFII mRNA expression.

In addition, we have demonstrated that COUP-TFII positively regulates HNF4α expression. To assess the importance of HNF4α gene modulation, we measured the expression of the PPARα target gene. The strong suppression of COUP-TFII in 832/13 INS-1 cells decreased HNF4α gene transcript levels by 40%. There were no changes in PPARα target gene expression, however, probably because of the remaining HNF4α mRNA. Nevertheless, when COUP-TFII was overexpressed, it led to a marked increase in HNF4α, with a statistically significant elevation in PPARα gene expression.

In addition, the 6.8-kbp P1 and 4.1-kbp P2 HNF4α promoter regions (2) are transactivated by COUP-TFII in COS cells, suggesting possible transcriptional control of HNF4α gene expression by COUP-TFII (our unpublished data).

The reciprocal cross-regulation between COUP-TFII and HNF4α raises some important questions with respect to the transcriptional activities of these factors. Other regulatory events in the same circuit may prevent undesirably high intracellular expression levels. In a similar vein, we have shown that COUP-TFII negatively regulates its own expression. In addition, COUP-TFII binds to the same DNA binding site as HNF4α, suggesting possible competition for occupancy of this site (schematically depicted in Fig. Fig.77).

FIG. 7.
Schematic representation of the positive (arrows) and negative (flat-headed symbol) effects of COUP-TFII integrated in a complex regulatory network. Adapted from Médecine Sciences (23) with permission of the publisher.

The physiological relevance of the cross-regulation between COUP-TFII and HNF4α, which remains to be established in vivo, may be in an amplification loop controlling an acute response to a specific signal, such as those received in the fasting state. The level of HNF4α expression is high during fasting, decreases in the liver (26) and in the pancreas (our unpublished data) upon refeeding, and correlates well with the variations in COUP-TFII transcript and protein levels (A. Perilhou, unpublished data).

The pathophysiology of maturity-onset diabetes of the young is caused by mutations in five separate genes encoding transcription factors, including the HNF4α, or MODY1, gene (Fig. (Fig.7).7). HNF4α/MODY1 is involved in transcriptional regulatory networks in both the liver and the pancreas (19, 25). It has been hypothesized previously that maturity-onset diabetes of the young may result from a collapse of cell-specific transcription networks due to the haploinsufficiency of key network genes (19). In this context, our genetic analysis of COUP-TFII function in pancreatic beta cells is relevant in that haploinsufficient mice have an impaired insulin secretion phenotype (1) and we have previously observed the loss of glucose-stimulated insulin secretion in 832/13 INS-1 cells treated with COUP-TFII siRNA (Perilhou, unpublished).

In conclusion, the results shown here are of particular interest in that the COUP-TFII gene can be considered as a candidate MODYX gene. The COUP-TFII promoter contains a functional DR-1 binding site for the HNF4α protein, and the molecular basis of many MODY1 phenotypes remains to be defined (18). Future studies should therefore include the search for mutations in the human COUP-TFII gene in patients with MODYX.


This work was supported by grants from Nestlé France 2007; from the Programme National de Recherches sur le Diabète-INSERM/ARD (grant no. PNRD-A06064KS); from the Agence Nationale pour la Recherche (ANR 2005 Cardiovasculaire, Obésité et Diabète grant no. ANR-05-PCOD-088-01); from Bonus Qualité Recherche, Université Paris 5; from the Swiss National Science Foundation (grant no. 32-66907.01 to C.B.W.); and from ADA (grant no. 7-04-RA-106 to D.K.S.). A.P. is the recipient of a doctoral fellowship from the Ministère de l'Enseignement Supérieur et de la Recherche and from Fondation pour la Recherche Médicale.

We are grateful to C. Newgard for the 832/13 INS-1 cell line and to F. Petit for the gift of some of the COUP-TFII promoter constructs. We thank all the INSERM, U649, team for their expertise in adenovirus preparation. We thank P. Bossard for critical reading of the manuscript, and T. Becker for helpful discussion of the siRNA experiments. We thank K. Kaestner and J. Ferrer for helpful discussions.


[down-pointing small open triangle]Published ahead of print on 12 May 2008.


1. Bardoux, P., P. Zhang, D. Flamez, A. Perilhou, T. Lavin, J. Tanti, K. Hellemans, E. Gomas, C. Godard, F. Andreelli, M. Buccheri, A. Kahn, Y. Le Marchand-Brustel, R. Burcelin, F. Schuit, and M. Vasseur-Cognet. 2005. Essential role of chicken ovalbumin upstream promoter-transcription factor II in insulin secretion and insulin sensitivity revealed by conditional gene knockout. Diabetes 541357-1363. [PubMed]
2. Briancon, N., A. Bailly, F. Clotman, P. Jacquemin, F. P. Lemaigre, and M. C. Weiss. 2004. Expression of the alpha7 isoform of hepatocyte nuclear factor (HNF) 4 is activated by HNF6/OC-2 and HNF1 and repressed by HNF4alpha1 in the liver. J. Biol. Chem. 27933398-33408. [PubMed]
3. Cooney, A., S. Tsai, B. O'Malley, and M. Tsai. 1992. Chicken ovalbumin upstream promoter transcription factor (COUP-TF) dimers bind to different GGTCA response elements, allowing COUP-TF to repress hormonal induction of the vitamin D3, thyroid hormone, and retinoic acid receptors. Mol. Cell. Biol. 124153-4163. [PMC free article] [PubMed]
4. Escriva, H., R. Safi, C. Hanni, M. C. Langlois, P. Saumitou-Laprade, D. Stehelin, A. Capron, R. Pierce, and V. Laudet. 1997. Ligand binding was acquired during evolution of nuclear receptors. Proc. Natl. Acad. Sci. USA 946803-6808. [PMC free article] [PubMed]
5. Gourdon, L., D. Lou, M. Raymondjean, M. Vasseur-Cognet, and A. Kahn. 1999. Negative cyclic AMP response elements in the promoter of the L-type pyruvate kinase gene. FEBS Lett. 4599-14. [PubMed]
6. Gupta, R., M. Vatamaniuk, C. Lee, R. Flaschen, J. Fulmer, F. Matschinsky, S. A. Duncan, and K. Kaestner. 2005. The MODY1 gene HNF4α regulates selected genes involved in insulin secretion. J. Clin. Investig. 1151006-1015. [PMC free article] [PubMed]
7. Hohmeier, H., H. Mulder, G. Chen, R. Henkel-Rieger, M. Prentki, and C. Newgard. 2000. Isolation of INS-1-derived cell lines with robust ATP-sensitive K+ channel-dependent and -independent glucose-stimulated insulin secretion. Diabetes 49424-430. [PubMed]
8. Ladias, J., M. Hadzopoulou-Cladaras, D. Kardassis, P. Cardot, J. Cheng, V. Zannis, and C. Cladaras. 1992. Transcriptional regulation of human apolipoprotein genes ApoB, ApoCIII, and ApoAII by members of the steroid hormone receptor superfamily HNF-4, ARP-1, EAR-2, and EAR-3. J. Biol. Chem. 26715849-15860. [PubMed]
9. Lou, D. Q., M. Tannour, L. Selig, D. Thomas, A. Kahn, and M. Vasseur-Cognet. 1999. Chicken ovalbumin upstream promoter-transcription factor II, a new partner of the glucose response element of the L-type pyruvate kinase gene, acts as an inhibitor of the glucose response. J. Biol. Chem. 27428385-28394. [PubMed]
10. Maestro, M. A., C. Cardalda, S. F. Boj, R. F. Luco, J. M. Servitja, and J. Ferrer. 2007. Distinct roles of HNF1beta, HNF1alpha, and HNF4alpha in regulating pancreas development, beta-cell function and growth. Endocr. Dev. 1233-45. [PubMed]
11. Mangelsdorf, D., C. Thummel, M. Beato, P. Herrlich, G. Schutz, K. Umesono, B. Blumberg, P. Kastner, M. Mark, P. Chambon, and R. Evans. 1995. The nuclear receptor superfamily: the second decade. Cell 83835-839. [PubMed]
12. Mietus-Snyder, M., F. M. Sladek, G. S. Ginsburg, C. F. Kuo, J. A. Ladias, J. E. Darnell, Jr., and S. K. Karathanasis. 1992. Antagonism between apolipoprotein AI regulatory protein 1, Ear3/COUP-TF, and hepatocyte nuclear factor 4 modulates apolipoprotein CIII gene expression in liver and intestinal cells. Mol. Cell. Biol. 121708-1718. [PMC free article] [PubMed]
13. Miura, A., K. Yamagata, M. Kakei, H. Hatakeyama, N. Takahashi, K. Fukui, T. Nammo, K. Yoneda, Y. Inoue, F. M. Sladek, M. A. Magnuson, H. Kasai, J. Miyagawa, F. J. Gonzalez, and I. Shimomura. 2006. Hepatocyte nuclear factor-4alpha is essential for glucose-stimulated insulin secretion by pancreatic beta-cells. J. Biol. Chem. 2815246-5257. [PubMed]
14. Nakshatri, H., and P. Bhat-Nakshatri. 1998. Multiple parameters determine the specificity of transcriptional response by nuclear receptors HNF4, ARP-1, PPAR, RAR and RXR through common response elements. Nucleic Acids Res. 262491-2499. [PMC free article] [PubMed]
15. Park, J.-I., S. Tsai, and M. Tsai. 2003. Molecular mechanism of chicken ovalbumin upstream promoter-transcription factor (COUP-TF) actions. Keio J. Med. 52174-181. [PubMed]
16. Pearson, E. R., S. F. Boj, A. M. Steele, T. Barrett, K. Stals, J. P. Shield, S. Ellard, J. Ferrer, and A. T. Hattersley. 2007. Macrosomia and hyperinsulinaemic hypoglycaemia in patients with heterozygous mutations in the HNF4A gene. PLoS Med. 4e118. [PMC free article] [PubMed]
17. Pedersen, K., P. Zhang, C. Doumen, M. Charbonnet, D. Lu, C. Newgard, J. Haycock, A. Lange, and D. Scott. 2007. The promoter for the gene encoding the catalytic subunit of rat glucose-6-phosphatase contains two distinct glucose-responsive regions. Am. J. Physiol. Endocrinol. Metab. 292E788-E801. [PubMed]
18. Rowley, C., L. Staloch, J. Divine, S. McCaul, and T. Simon. 2006. Mechanisms of mutual functional interactions between HNF-4alpha and HNF-1alpha revealed by mutations that cause maturity onset diabetes of the young. Am. J. Physiol. Gastrointest. Liver Physiol. 290G466-G475. [PubMed]
19. Servitja, J., and J. Ferrer. 2004. Transcriptional networks controlling pancreatic development and beta cell function. Diabetologia 47597-613. [PubMed]
20. Soosaar, A., K. Neuman, H. Nornes, and T. Neuman. 1996. Cell type specific regulation of COUP-TF II promoter activity. FEBS Lett. 39195-100. [PubMed]
21. Suaud, L., Y. Hemimou, P. Formstecher, and B. Laine. 1999. Functional study of the E276Q mutant hepatocyte nuclear factor-4α found in type 1 maturity-onset diabetes of the young. Diabetes 481162-1167. [PubMed]
22. Taylor, D. G., S. Haubenwallner, and T. Leff. 1996. Characterization of a dominant negative mutant form of the HNF-4 orphan receptor. Nucleic Acids Res. 242930-2935. [PMC free article] [PubMed]
23. Velho, G., C. Bellanné-Chantelot, and J. Timsit. 2003. MODY, a model of genotype/phenotype interactions in type 2 diabetes. Med. Sci. 19854-859. (In French.) [PubMed]
24. Viollet, B., A. Kahn, and M. Raymondjean. 1997. Protein kinase A-dependent phosphorylation modulates DNA-binding activity of hepatocyte nuclear factor 4. Mol. Cell. Biol. 174208-4219. [PMC free article] [PubMed]
25. Wang, H., P. Maechler, P. Antinozzi, K. Hagenfeldt, and C. B. Wollheim. 2000. Hepatocyte nuclear factor 4α regulates the expression of pancreatic β-cell genes implicated in glucose metabolism and nutrient-induced insulin secretion. J. Biol. Chem. 27535953-35959. [PubMed]
26. Yoon, C. J., P. Puigserver, G. Chen, J. Donovan, Z. Wu, J. Rhee, G. Adelmant, J. Stafford, R. C. Kahn, D. K. Granner, C. B. Newgard, and B. M. Spiegelman. 2001. Control of hepatic gluconeogenesis through the transcriptional coactivator PGC-1. Nature 413131-138. [PubMed]
27. Zhang, P., M. Bennoun, C. Godard, P. Bossard, I. Leclerc, A. Kahn, and M. Vasseur-Cognet. 2002. Expression of COUP-TFII in metabolic tissues during development. Mech. Dev. 119109-114. [PubMed]

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...