HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 24, 18358-18365, June 16, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/24/18358    most recent M909536199v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Cha, J.-Y. Articles by Ahn, Y.-h. Search for Related Content PubMed PubMed Citation Articles by Cha, J.-Y. Articles by Ahn, Y.-h. Social Bookmarking                      What's this?

Identification of Transacting Factors Responsible for the Tissue-specific Expression of Human Glucose Transporter Type 2 Isoform Gene

COOPERATIVE ROLE OF HEPATOCYTE NUCLEAR FACTORS 1 AND 3^*

Ji-Young Cha, Ha-il Kim, Kyung-Sup Kim, Man-Wook Hur, and Yong-ho Ahn^§

From the Department of Biochemistry and Molecular Biology and the Institute of Genetic Science, Yonsei University College of Medicine, 134 Shinchon-dong, Seodaemun-gu, Seoul 120-752, South Korea

Received for publication, December 2, 1999, and in revised form, March 9, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We investigated transacting factors binding to the cis-element important in tissue-specific expression of the human glucose transporter type 2 isoform (GLUT2) gene. By transient transfection assay, we determined that the 227-base pair fragment upstream of the ATG start site contained promoter activity and that the region from +87 to +132 (site C) was responsible for tissue-specific expression. DNase I footprinting and electrophoretic mobility shift assay indicated that site C contained one binding site for hepatocyte nuclear factor 1 (HNF1) and two binding sites for HNF3. The mutations at positions +101 and +103, which are considered to be critical in binding HNF1 and HNF3, resulted in a 53% decrease in promoter activity, whereas the mutation of the proximal HNF3 binding site (+115 and +117) reduced promoter activity by 28%. The mutations of these four sites resulted in marked decrease (70%) in promoter activity as well as diminished bindings of HNF1 and HNF3. A to G mutation, which causes conversion of the HNF1 and HNF3 binding sequence to the NF-Y binding site, resulted in a 22% decrease in promoter activity. We identified that both HNF1 and HNF3 function as transcriptional activators in GLUT2 gene expression. Coexpression of the pGL+74 (+74 to +301) construct with the HNF1 and HNF3 expression vectors in NIH 3T3 cells showed the synergistic effect on GLUT2 promoter activity compared with the expression of HNF1, HNF3, or a combination of HNF1 and HNF3. These data suggest that HNF1 and HNF3 may be the most important players in the tissue-specific expression of the human GLUT2 gene.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The glucose transporter type 2 isoform (GLUT2)¹ is a transmembrane protein involved in maintaining glucose homeostasis by facilitating the entry of glucose through the plasma membrane. GLUT2 is predominantly expressed in hepatocytes, -cells of the pancreas, and to a lesser extent in absorptive epithelia of the intestine and kidney (12). GLUT2 plays an important role in absorbing glucose after meals in hepatocyte (3) as well as in insulin secretion in pancreatic -cells (4). The GLUT2 mRNA level is influenced by the blood glucose level and insulin (5, 6). The GLUT2 expression between hepatocytes and pancreatic -cells is contradictory in diabetic animal model; GLUT2 expression is increased in hepatocytes, whereas it is decreased in -cells (7-9).

Although the transcriptional regulation of GLUT2 expression under various metabolic states has been studied, the transacting factors regulating the tissue-specific expression of the GLUT2 gene are relatively unknown. So far, Pdx-1, a pancreatic homeodomain transcription factor, has been proposed to regulate the expression of the GLUT2 gene (10). However, this factor is not expressed in hepatocytes. Thus, liver-specific transcription factors, such as hepatocyte nuclear factor 1 (HNF1) (11, 12), CCAAT/enhancer-binding protein (13), HNF3 (14) and HNF4 (15) have been expected to regulate GLUT2 gene expression. None of these transcription factors are found exclusively in hepatocytes; rather, they are all expressed, although to a lesser degree, in other tissues. Thus, the tissue specificity and levels of gene expression depend on the concerted actions of several factors working together.

HNF1 is a homeodomain protein that plays a key role in the liver-specific expression of many genes during differentiation and development (16). HNF1 is required for the expression of GLUT2 and other liver-specific genes such as albumin, 1-antitrypsin, and fibrinogen (17-19). HNF1 associates with the highly related HNF1 (vHNF1) to form homo- or heterodimers (20). The expression of the HNF1 dominant negative form resulted in reduction of GLUT2 gene expression (21). Mutation of HNF1 resulted in decreased transcription of GLUT2 and was ultimately related to familial type 2 diabetes mellitus (22). The human GLUT2 promoter has a binding site for HNF1 at the 1030 bp region. However, deletion of the HNF1 binding site did not abolish GLUT2 promoter activity mediated by HNF1 (22). Thus, it should be determined whether other cis-elements for HNF1 exist in the human GLUT2 promoter or whether HNF1 indirectly regulates GLUT2 gene expression by other transcription factors.

HNF3, a member of the forkhead winged helix family, has been known to have a function in liver and gut development (23). HNF3, -3, and -3 bind to the same DNA sequence with different affinities. It was shown that HNF3 could activate the albumin gene expression by repositioning the nucleosomes in the albumin enhancer (24, 25).

In this study, we showed that human GLUT2 expression was regulated directly by HNF1 and HNF3. HNF1 and HNF3 synergistically activated transcription in NIH 3T3 cells. Thus, we propose that the combined action of the two liver-enriched factors, HNF1 and HNF3, plays a critical role in the tissue-specific expression of the GLUT2 gene. Furthermore, we showed that the promoter activity of the mutG construct containing the +103AG mutation, which was found in a Korean non-insulin-dependent diabetes mellitus patient, was in part compensated by the binding of NF-Y instead of HNF1 or HNF3.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Construction of Plasmids-- The promoter region (294 to +301 bp) of the human GLUT2 gene was cloned into the SmaI site of pGL3-basic reporter vector (Promega, Madison, WI) to generate pGL-294. Serial deletions from the 5'-end of pGL3-294 were created by polymerase chain reaction. The mutations were generated by polymerase chain reaction-based site-directed mutagenesis (Stratagene, La Jolla, CA; for sequences, see below). For pGL+74m3, the mutant pGL+74m1 was mutated again with the oligonucleotide m2. The sequences of constructs were confirmed by DNA sequencing. pRSV-HNF1 and pRSV-vHNF1 were gifts from Dr. M. Yaniv. pGem-HNF3 and pGem-HNF3 were provided by Dr. R. H. Costa.

Preparation of Nuclear Extracts-- Nuclear extracts from liver of male Sprague-Dawley rats or cell lines were prepared as described by Gorski et al. (26) or Dignam et al. (27). Protein concentration was determined according to Bradford (28). The extracts were frozen in aliquots and stored at 70 °C.

DNase I Footprinting Assay-- DNA fragments (294 to +16 bp or +46 to +301 bp) were labeled in each strand and purified as follows. The promoter region was amplified by polymerase chain reaction to obtain 294 to +16 and +48 to +301 promoter fragments, and the resulting fragments were subcloned into the EcoRV site of pT7Blue (R) vector. The promoter fragments were isolated by double digestion with EcoRI (BamHI) and PstI (HindIII) to obtain 5'-overhanging ends and 3'-overhanging ends. The fragments were labeled using Klenow fragment and [-³²P]dATP and then purified from agarose gel electrophoresis by a gel extraction kit (Qiagen, Valencia, CA). DNA-protein binding reactions were performed using 50,000 cpm (approximately 1 ng) of probe per reaction in a solution containing 10 mM HEPES, pH 7.9, 60 mM KCl, 7% glycerol, 1 mM EDTA, 1 mM dithiothreitol, 2 µg of poly(dI-dC), and the indicated amount of nuclear extracts. After a 20-min incubation on ice, 5 µl of DNase I freshly diluted in a solution containing 10 mM HEPES, pH 7.9, 60 mM KCl, 25 mM MgCl₂, 5 mM CaCl₂, and 7% glycerol was added to the reaction and kept at room temperature for 2 min. Dilution folds of DNase I were ranged from 1:200 to 1:2000 of stock (10 units/µl) depending on the amount of protein present in the reaction. Digestion reactions were stopped by adding 80 µl of a stop solution containing 20 mM Tris-Cl, pH 8.0, 20 mM EDTA, 250 mM NaCl, 0.5% SDS, 4 µg of yeast tRNA, and 10 µg of proteinase K. The samples were incubated for 30 min at 45 °C, extracted with phenol/chloroform, ethanol-precipitated, and resuspended in formamide dye. The samples were resolved in 6% polyacrylamide, 7 M urea sequencing gel. The protected regions were mapped with reference to the migration of Maxam-Gilbert A + G sequencing products (29).

Electrophoretic Mobility Shift Assay (EMSA) and Supershift Assay-- Probes for gel shift assays were labeled with ³²P in the presence of [-³²P]ATP and T4 polynucleotide kinase. Labeled double-stranded oligonucleotides were prepared by mixing a 5-fold molar excess of the complementary single-stranded DNAs in 50 mM NaCl, heating to 90 °C for 5 min, and then cooling to room temperature. The oligonucleotides used in these assays were as follows (the mutated nucleotides are underlined): C-1, 5'-GGCAAAGCACTTATTGATTAGATTCCCATC-3'; C-2, 5'-GATTCCCATCAATATTCAGCTGCCGC-3'; mutC, 5'- GGCAAAGCACTTATTGCTTAGATTCCCATC-3'; mutG, 5'- GGCAAAGCACTTATTGTTAGATTCCCATC-3'; m1, 5'- GGCAAAGCACTTATCGCTTAGATTCCCATC-3'; m2, 5'-GATTCCCAGCCATATTCAGCTGCCGC-3'.

The oligonucleotides for HNF1 (30), HNF3 (31), HNF4 (32), and NF-Y (33) were synthesized, and Oct-1 (5'-TGTCGAATGCAAATCACTAGAA-3') and Sp1 consensus oligonucleotides (5'-ATTCGATCGGGGCGGGGCGAGC-3') were purchased from Promega. The labeled probe (50,000 cpm) was combined with nuclear proteins in 25 mM Tris/HCl, pH 7.4, 80 mM KCl, 0.1 mM EDTA, 1 mM dithiothreitol, and 10% (v/v) glycerol. The nonspecific competitor, 1.5 µg of poly(dI-dC), was added to each binding reaction. In the supershift assays, anti-HNF1 and/or anti-HNF1 antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and the antibody against the B subunit of NF-Y was provided by Dr. R. Mantovani. Antibodies were added to the binding reaction 10 min before the addition of the labeled probe. Binding reaction mixtures were incubated for 20 min on ice and resolved on a nondenatured (5% w/v) acrylamide gel (29:1 (w/w) acrylamide/bisacrylamide) in 0.5× TBE at 4 °C. For competition assays, a 50-200-fold molar excess of various unlabeled competitor DNAs was added to the reaction mixture prior to the addition of the labeled probe. The dried gels were exposed to x-ray film at 70 °C with an intensifying screen.

Cell Culture and Transient Transfection-- Cells used in this experiment were maintained as monolayer cultures and grown in appropriate media. Plasmid DNAs were purified on Qiagen Midiprep kit columns (Qiagen) at least twice independently. Cells were plated in six-well tissue culture plates at a density of 1 × 10⁶ cells/well in 2 ml of medium. After a 20-h attachment period, transfections were performed with LipofectAMINE PLUS reagent (Life Technologies, Inc.), following the manufacturer's protocol. Briefly, 0.5 µg of each construct of GLUT2 promoter; 0.1 µg of pCMV--galactosidase, with or without 20 ng of HNF1, HNF1, HNF3, and/or HNF3; and 4 µl of PLUS reagent and 2 µl of LipofectAMINE in 200 µl of OPTI-MEM I (Life Technologies, Inc.) medium lacking serum were mixed and added to cells. In order to transfect constant amounts of DNA, sample DNAs were supplemented with an appropriate control vector. After 3 h, the medium containing the LipofectAMINE-DNA complex was removed and replaced by appropriate medium (containing serum and antibiotics). Cells were then cultured further for 48 h and harvested in reporter lysis buffer (Promega). The lysed cells were centrifuged to remove cell debris, and the supernatant was collected. Luciferase assays were conducted with 10 µl of cell extracts and 50 µl of luciferase assay reagent (Promega). -Galactosidase activity was determined with 10 µl of cell extract and 190 µl of assay reagent containing O-nitrophenol--D-galactopyranoside in a colorimetric assay. Luciferase data were expressed as luciferase activity corrected by -galactosidase activity in the cell lysate. Each transfection was performed in triplicate and repeated 3-5 times.

Statistical Analysis-- All transfection studies were performed in 3-5 separate experiments, where triplicate dishes were transfected. The data were represented as mean ± S.D. Statistical analysis was carried out using Microsoft Excel^® (Microsoft).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Promoter Activities of Human GLUT2 Gene-- To identify the important region in GLUT2 transcription, a series of sequential 5'-deletions of the GLUT2 294-bp 5'-flanking region contiguous to the promoter and 301-bp 5'-untranslated region were generated and fused to a luciferase reporter gene (Fig. 1A). These constructs were transiently transfected in HIT-T15 cells, which express GLUT2, or NIH 3T3 cells, which do not express GLUT2, and luciferase activity was determined. Luciferase activity for all constructs was very high in HIT-T15 but low in NIH 3T3 cells (Fig. 1A). These differences in the activities indicated that the 5'-flanking region of the GLUT2 promoter might be responsible for tissue-specific expression. In HIT-T15 cells, luciferase activities for the pGL+34 construct remained high (Fig. 1A). However, deletion down to +73 resulted in a marked increase, whereas further deletion to +188 caused a drastic decrease in luciferase activity. These results indicated the presence of negative and positive regulatory elements in the 5'-untranslated region. To generalize that the 5'-flanking region of human GLUT2 promoter was responsible for tissue-specific expression, we have transfected the reporter constructs to MIN6, HepG2 (GLUT2-expressing) cells and HeLa (GLUT2-nonexpressing) cells (Fig. 1B). As expected, the promoter activity was high in MIN6 and HepG2 cells, whereas HeLa cells showed very low promoter activity.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Transcriptional activities of human GLUT2 promoter. A, the human GLUT2 promoter regions shown at the left were inserted into the pGL3-basic. The reporter constructs were transfected to HIT-T15 (open bar) and NIH 3T3 cells (hatched bar). B, transfection of the reporter constructs to HepG2 (filled bar), MIN6 (open bar), and HeLa (hatched bar) was performed. Luciferase activities were normalized on the basis of -galactosidase activity encoded by the co-transfected control plasmid, pCMV--galactosidase. Results are the mean ± S.D. of three independent experiments in triplicate.

Identification of Protein Binding Sites in the Human GLUT2 Promoter-- Since the promoter activities remained high down to the transcription start site (Fig. 1A), two promoter fragments (from 294 to +16 or from +46 to +301) were analyzed. Eight sites were protected with rat liver nuclear extract. These sites were designated as site H (294/281), site G (274/247), site F (230/196), site E (180/161), site D (154/135), site C (+87/+132), site B (+190/+216), and site A (+229/+257) (Fig. 2). Computer-based analysis of the sequence revealed several motifs resembling consensus sequences of many known nuclear factors (Fig. 3A). Site C contained three putative transacting factor-binding sites for HNF1 and two HNF3s in reverse orientation. A sequence corresponding to the consensus Oct-1-binding site was also present in site C. Site C and site A contained consensus sequences for HNFs that were mainly expressed in the liver, pancreas, and kidney. The deletion of the promoter region from +74 to +188 bp greatly decreased the promoter activity, suggesting that the factors binding to the footprinting region of site C may act as positive regulators.

View larger version (77K): [in this window] [in a new window]   Fig. 2.   DNase I footprinting analysis of human GLUT2 promoter. The DNA fragments 294 to +16, coding strand A and +46 to +301, coding strand B were labeled at their 3'-ends and mixed with rat liver nuclear extract at protein concentrations (in µg) noted above the respective lanes. Lane A+G shows Maxam-Gilbert sequencing reaction products of the probe DNA. The regions protected from DNase I digestion are indicated by boxes with their names (sites A to H).

View larger version (46K): [in this window] [in a new window]   Fig. 3.   Summary of the consensus sequences in the human GLUT2 promoter. A, the transcription initiation site is designated +1 (35). The boxed areas represent the protected regions of human GLUT2 promoter identified by the DNase I footprinting experiment. The known consensus sequences are underlined. B, comparison of potential HNF1 and HNF3 binding sites in human, mouse, and rat GLUT2 promoter sequences.

As shown in Fig. 3B, site C is well conserved between species. The +87/+132 region of the human GLUT2 promoter matched well with the 131/86 region of mouse (34) and the +492/+541 region of rat (13). A computer search suggested that these regions are potential binding sites for HNFs in these species. Also, it is interesting to note that site B of human GLUT2 promoter region was also conserved both in mouse (34) and rat (13), although the transcription factors binding to the site B are unknown.

HNF1 and HNF3 Binding to Site C Is Critical in GLUT2 Promoter Activity-- On the basis of DNase I footprinting and computer-based analysis, we attempted to identify the nuclear factor(s) binding to site C by EMSA. Site C was divided into two overlapping halves, site C-1 and site C-2 (Fig. 4A). Site C-1 (+87 to +116) included putative HNF1 and HNF3 binding sites. The HNF1 binding site covers the HNF3 binding site. Site C-2 (+107 to +132) included putative HNF3 and Oct-1 consensus sites. In site C, two HNF3 binding sites were inversely oriented. For site C-1, multiple bands were observed by the addition of rat liver nuclear extracts (Fig. 4B, lanes 1 and 2). These bands were specific, since they competed by an excess of unlabeled oligonucleotide site C-1 (lane 3). Band b was selectively abolished by an oligonucleotide HNF1 site of the L-pyruvate kinase promoter (lane 4). Band e decreased in the presence of an excess of the unlabeled oligonucleotide corresponding to HNF3 sequences in the transthyretin promoter (lane 5). The addition of the HNF1 and HNF3 consensus oligonucleotides abolished bands b and e (lane 6). The unrelated HNF4 oligonucleotide did not affect protein binding (lane 7). Band c represented the complex containing Oct-1 that was competed by Oct-1 consensus oligonucleotide (data not shown). Supershift experiments using anti-HNF1 and/or  antibodies revealed that HNF1 and HNF1 were binding (lanes 8 and 9), although HNF1 is the predominant member of the HNF1 family in hepatocytes (35). When both antibodies against HNF1 and HNF1 were used together, band b was completely supershifted (lane 10). Gel retardation using oligonucleotide HNF1 as a probe gave band b, and it was shifted by anti-HNF1 antibody (lanes 12 and 13). The identity of band f, the fastest migrating band, was not determined in this study. It's interesting to note that an A to G mutation at +103 in site C-1, which was found in the course of screening GLUT2 promoter mutations of non-insulin-dependent diabetes mellitus patients, resulted in the appearance of band d (lane 11 and Fig. 4C). To identify the transcription factor binding to this mutated region, we compared this sequence to a number of transcription factor consensus sequences and found that the sequence was identical to the NF-Y consensus sequences. This was confirmed by a cold competitor and NF-Y consensus oligonucleotide (Fig. 4C, lane 3 and lanes 5 and 6, respectively). Site C-1 abolished bands e and f, but not band d (lane 4). The identity of band d was further confirmed by an anti-NF-Y B subunit antibody (lane 7). HNF3 cold oligonucleotide abolished band e (lane 8).

View larger version (32K): [in this window] [in a new window]   Fig. 4.   EMSAs of site C. A, the sequences of the site C fragment (+87 to +132) of the human GLUT2 gene and the positions of the probes used in EMSAs are shown. The positions of the probes are indicated by the thick lines. The sequences similar to consensus motifs for HNFs are underlined and boxed. B, 32P-labeled site C-1 (lanes 1-10), mutG (lane 11), or HNF1 (lanes 12 and 13) consensus oligonucleotides were incubated with rat liver nuclear extracts (5 µg) in the absence (lane 2) or presence of a 100-fold molar excess of the indicated cold competitor: site C-1 (lane 3), HNF (lane 4), HNF3 (lane 5), both HNF1 and HNF3 (lane 6), or HNF4 (lane 7) consensus oligonucleotide. Supershift assay using antibodies against HNF1 and HNF1 confirmed the HNF1 binding to this probe (lanes 8-10 and 13). C, 32P-labeled mutG was incubated with rat liver nuclear extracts (5 µg) in the absence (lane 2) or presence of a 100-fold molar excess of the indicated cold competitors: mutG (lane 3), site C-1 (lane 4), NF-Y (lane 5), HNF3 (lane 8). Unlabeled NF-Y consensus oligonucleotide was added in a 200-fold molar excess (lane 6). The mutG probe was incubated with antibody directed against NF-Y B subunit (lane 7). D, 32P-labeled site C-2 was incubated with rat liver nuclear extracts (5 µg) in the absence (lane 2) or presence of a 100-fold molar excess of the indicated cold competitor: site C-2 (lane 3), HNF1 (lane 4), HNF3 (lane 5), Oct-1 (lane 6), or Sp-1 (lane 7) consensus oligonucleotide. The bands representing specific DNA-protein complexes (bands b, c, d, e, and f) and the supershifted complex (bands a and s) were indicated.

Gel retardation assay of a site C-2 probe revealed two shifted bands (Fig. 4D, lane 2). These bands were competed out by an excess of unlabeled probe (lane 3). The lower one had the same mobility as band e in the site C-1 probe and was competed by oligonucleotide HNF3 (lane 5). The slow mobility band (band c) represented a complex containing Oct-1, because it was completely competed out by the consensus sequence for Oct-1 (lane 6). The consensus oligonucleotide for Oct-1 also abolished band d. The unrelated Sp1 oligonucleotide did not compete out any protein binding (lane 7).

Next, we determined the effect of mutations on the binding affinity of HNF1, HNF3, and Oct-1 with the promoter. The HNF-binding sites in site C were disrupted by site-directed mutagenesis, and the resulting changes in binding activity were determined (Fig. 5). Using oligonucleotide C-1 as a probe, the retarded bands disappeared with a 50-fold molar excess of cold competitors (Fig. 5A, lanes 3 and 4). The bands b and e were slightly competed out by the addition of a 200-fold molar excess of the unlabeled mutC (A to C mutation at +103) or mutG (A to G mutation at +103) (lanes 5-8). However, the mutant m1 (T to C mutation at +101 and A to G mutation at +103) showed decreased competition when compared with a self-competitor, mutC or mutG, suggesting that the base changes at +101 and +103 in site C-1 may decrease HNF1 and HNF3 binding (lanes 9 and 10).

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Competition assays of site C-1 and site C-2. Competition assay of EMSA using probe C-1 (A) or C-2 (B) was performed in the absence or presence of competing oligonucleotides (a 50- or 100-fold molar excess); mutC (+103AC), mutG (+103AG), m1 (two-base mutation at +101TC and +103AC), or m2 (two-base mutation at +115TG and +117AC). Five µg of rat liver nuclear extract was used in each lane.

In the case of the site C-2 probe, Oct-1 and HNF3 bindings were efficiently competed out by C-2 and partially competed out by C-1, which contained HNF3 and Oct-1 binding sites too (Fig. 5B, lanes 2-4 and lane 7). Mutant oligonucleotide m2 (T to G mutation at +115 and A to C mutation at +117) could not compete out protein binding even in the presence of a 200-fold molar excess of C-2 probe (lanes 5 and 6).

To correlate the decreased HNF1 and HNF3 binding with GLUT2 promoter activity, we introduced the mutC, mutG, m1, m2, and m3 substitutions at their respective positions in the pGL+74 (+74 to +301) construct. The effects of the mutations within the promoter were tested by transient transfection assay in HIT-T15 cells (Fig. 6). The C or G mutation of the +103 base reduced GLUT2 promoter activity to 70 and 78% of wild type pGL+74 expression, respectively. Two-base substitutions of site C-1 (pGL+74m1) resulted in a 53% (p < 0.05) decrease in activity, whereas two-base mutations of site C-2 (pGL+74m2) resulted in a 28% decrease in GLUT2 promoter activity. The pGL+74m3 mutant, in which both site C-1 and C-2 were mutated, reduced transcriptional activity by 70% (p < 0.05). These data showed that the HNF1 and HNF3 binding sites in site C act as positive cis-elements in HIT-T15 cells and that site C-1 is more important than site C-2 in the expression of the GLUT2 gene.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Mutational effects on human GLUT2 promoter. HIT-T15 cells were transiently transfected with 0.5 µg of pGL+74, pGL+74mutC, pGL+74mutG, pGL+74m1, pGL+74m2, or pGL+74m3 and 0.1 µg of pCMV--galactosidase. The boldface base(s) shown at the left are the sequences of introduced mutations at the specific position(s) (see "Results" for details). Cells were harvested 48 h after transfection, and the luciferase activity was measured. Luciferase activities have been normalized on the basis of -galactosidase activity. Results are the mean ± S.D. of three independent experiments in triplicate.

Cooperative Activation of Human GLUT2 Promoter by HNF1 and HNF3 in NIH 3T3 Cells-- Finally, we examined whether these transcription factors were able to transactivate the human GLUT2 gene in a non-HNF-expressing system. As shown in Fig. 7, the HNF1 and HNF3 were not expressed in NIH 3T3 cells, whereas they were expressed in HIT-T15 cells as well as in hepatocytes. Therefore, we overexpressed the subtypes of HNF1 and HNF3 alone or in various combinations in NIH 3T3 cells to investigate their effects on GLUT2 gene expression. Except for HNF1, the subtypes of HNFs showed positive transacting effects on the GLUT2 promoter. HNF1, HNF3, and HNF3 resulted in 1.9-, 2.0-, and 3.6-fold increases in promoter activity, respectively. Coexpression of HNF1 and HNF3 proteins showed the highest activation (8.7-fold) of promoter activity, whereas other coexpressions of HNF1 and HNF3 or of HNF1 and HNF3 showed activation effects by 3.1- and 3.4-fold, respectively (Fig. 8).

View larger version (29K): [in this window] [in a new window]   Fig. 7.   Expression of HNFs in various cell types. 32P-Labeled site C-1, mutG, or HNF1 oligonucleotides were incubated with several nuclear extracts: NIH 3T3 cells (lanes 1-3), rat liver (lanes 4-6), or HIT-T15 cells (lanes 7-9). Five µg of nuclear extracts was used in each reaction.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Cooperative activation of HNF1 and HNF3 on human GLUT2 promoter activity. NIH 3T3 cells were transiently transfected with 0.5 µg of pGL+74 construct and 20 ng of HNF1 and/or HNF3 expression vectors, as indicated. In all cases, equal amounts of DNA (0.64 µg) were transfected into cells by adding control expression vectors devoid of insert DNA. Cells were harvested 48 h after transfection, and the luciferase activity was measured. Luciferase activities were normalized on the basis of -galactosidase activity produced by the co-transfected control plasmid, pCMV--galactosidase. Results are the mean ± S.D. of three independent experiments in triplicate.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Although the GLUT2 gene is mainly expressed in liver and pancreatic -cells, the transcription factors and cis-elements responsible for the tissue-specific expression are largely unknown. PDX-1, a pancreatic homeodomain transcription factor, has been showed to activate the expression of the mouse GLUT2 gene (10). However, its expression is not correlated with reduced GLUT2 gene expression in diabetes, which suggests that transacting factors other than PDX-1 may play some roles in GLUT2 gene expression. In this study, we showed that the 5'-untranslated region of the human GLUT2 gene plays a critical role in transcriptional activation, and specifically, site C (+87 to +132) is responsible for tissue-specific expression of the human GLUT2 gene. Also, a serial deletion study revealed that deletion down to the +189 region abolished GLUT2 promoter activity, which suggested that multiple transcription initiation sites exist in this 5'-untranslated region (36).

In EMSA, site C-1 (+87 to +116), which contained overlapping HNF3 and HNF1 binding sites, provided the HNF1/probe and HNF3/probe complexes (Fig. 4B). Competition experiments indicated that HNF1 and HNF3 bound to site C-1 in a mutually exclusive manner, although HNF1 binding seemed to be preferential. In supershift experiments, most of the HNF1 subtypes present in hepatocytes or HIT-T15 cells may be the HNF1 form rather than HNF1 (Fig. 4B). This result suggests that most of HNF1 acts as HNF1 homodimeric forms in the cells where GLUT2 is mainly expressed. In contrast to hepatocytes or HIT-T15 cells, HNF1 is the predominant form in the kidney (37), where GLUT2 is expressed at a lower level. Thus, it could be assumed that the differential expression of these HNF1 subtypes may be one of the factors regulating tissue-specific GLUT2 expression.

We have compared the sequences of site C in human promoter with that of mouse and rat. As shown in Fig. 3B, these sequences were well conserved in mouse (GTII region; Ref. 34) and rat (13) promoters and may be real HNF1 and HNF3 binding sites. Bonny et al. (34) suggested that the GTII region might be a cis-element having a stimulatory role on GLUT2. Thus far, transacting factors binding to this region in mice have not been known. Based on our experiment and the homology relationship between human and mouse sequences of the site C and GTII region, it is assumed that HNFs might bind to the GTII region and be working as positive regulators.

Transfection studies using HIT-T15, MIN6, HepG2, NIH 3T3, and HeLa cells revealed the differences in promoter activities (Fig. 1, A and B). These activities were well correlated with the status of HNF1 and HNF3 mRNA expression (data not shown). From these observations, it is suggested that GLUT2 promoter could be activated by the expression of HNF1 and HNF3 regardless of the origins or types of cells.

It is possible that the subtypes of HNFs may directly or indirectly interact with each other in controlling GLUT2 gene expression, because site C contains both HNF1 and HNF3 binding sites that are located closely together. Previous studies have suggested that an important role of HNF3 could be to cooperate with other factors bound to closely oriented cis-elements (38-41). In the case of the aldolase B gene, HNF3 competitively antagonized HNF1-dependent transactivation (42). Here, we report another type of HNF1 and HNF3 relationship in activating the GLUT2 promoter, in which we assume there is cooperative activation. Recent papers have shown that HNF3 may be involved in the transition of chromatin from an inactive to an active conformation (24, 25). In the GLUT2 promoter, binding HNF3 to site C-2 may contribute to a higher affinity for HNF1, perhaps by opening the chromatin.

Interestingly, in the course of screening non-insulin-dependent diabetes mellitus patients who may have possible mutations at the upstream region of the GLUT2 gene, we found A to G mutation in the +103 region, which belongs to site C-1. By this mutation, the consensus sequence for HNF1 and HNF3 is changed to that of NF-Y. Thus, it was expected that this mutation definitely caused a decrease in promoter activity. However, functional assay of this mutant showed only a 22% decrease in the promoter activity, whereas the m1 construct, which has lost its ability to bind HNF1 and HNF3, showed a 50% decrease in the promoter activity (Fig. 6). These results can be explained by assuming that NF-Y binding instead of HNFs binding to site C-1 might compensate the promoter activity, because NF-Y is known to act as a positive transacting factor in many other genes (43).

In conclusion, we identified that the liver-enriched transcription factors HNF1 and HNF3 are two of the main stimulatory trans-acting factors, and they may be responsible for tissue-specific expression of the GLUT2 gene. Furthermore, our results emphasize a new type of functional interplay between HNF1 and HNF3 to drive human GLUT2 gene expression.

	ACKNOWLEDGEMENTS

We thank Dr. M. Yaniv for providing pRSV-HNF1 and pRSV-vHNF1, Dr. R. H. Costa for providing pGem-HNF3 and pGem-HNF3, Dr. R. Mantovani for providing the antibody against B subunit of NF-Y, and Dr. Jun-ichi Miyazaki for providing MIN6 beta cells.

	FOOTNOTES

^* This work was supported by Korea Science and Engineering Foundation Grant 96-0403-14-01-3.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Recipient of a scholarship from the Brain Korea 21 Project For Medical Science, Ministry of Education, South Korea.

^§ To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, Yonsei University College of Medicine, 134 Shinchon-dong, Seodaemun-gu, Seoul 120-752, Korea. Tel.: 82-2-361-5187; Fax: 82-2-312-5041; E-mail: yha111@yumc.yonsei.ac.kr.

Published, JBC Papers in Press, March 31, 2000, DOI 10.1074/jbc.M909536199

	ABBREVIATIONS

The abbreviations used are: GLUT2, glucose transporter type 2 isoform; HNF, hepatocyte nuclear factor; NF-Y, nuclear factor-Y; EMSA, electrophoretic mobility shift assay; bp, base pair(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				C. Dos Santos, P. Bougneres, and D. Fradin A Single-Nucleotide Polymorphism in a Methylatable Foxa2 Binding Site of the G6PC2 Promoter Is Associated With Insulin Secretion In Vivo and Increased Promoter Activity In Vitro Diabetes, February 1, 2009; 58(2): 489 - 492. [Abstract] [Full Text] [PDF]

				E. D. Wederell, M. Bilenky, R. Cullum, N. Thiessen, M. Dagpinar, A. Delaney, R. Varhol, Y. Zhao, T. Zeng, B. Bernier, et al. Global analysis of in vivo Foxa2-binding sites in mouse adult liver using massively parallel sequencing Nucleic Acids Res., August 1, 2008; 36(14): 4549 - 4564. [Abstract] [Full Text] [PDF]

				H. S. Freitas, G. F. Anhe, K. F. S. Melo, M. M. Okamoto, M. Oliveira-Souza, S. Bordin, and U. F. Machado Na+-Glucose Transporter-2 Messenger Ribonucleic Acid Expression in Kidney of Diabetic Rats Correlates with Glycemic Levels: Involvement of Hepatocyte Nuclear Factor-1{alpha} Expression and Activity Endocrinology, February 1, 2008; 149(2): 717 - 724. [Abstract] [Full Text] [PDF]

				M. E Cerf Transcription factors regulating {beta}-cell function. Eur. J. Endocrinol., November 1, 2006; 155(5): 671 - 679. [Abstract] [Full Text] [PDF]

				C. Haumaitre, M. Fabre, S. Cormier, C. Baumann, A.-L. Delezoide, and S. Cereghini Severe pancreas hypoplasia and multicystic renal dysplasia in two human fetuses carrying novel HNF1{beta}/MODY5 mutations Hum. Mol. Genet., August 1, 2006; 15(15): 2363 - 2375. [Abstract] [Full Text] [PDF]

				S.-S. Im, S.-Y. Kang, S.-Y. Kim, H.-i. Kim, J.-W. Kim, K.-S. Kim, and Y.-H. Ahn Glucose-Stimulated Upregulation of GLUT2 Gene Is Mediated by Sterol Response Element-Binding Protein-1c in the Hepatocytes Diabetes, June 1, 2005; 54(6): 1684 - 1691. [Abstract] [Full Text] [PDF]

				M. S. Winzell, G. Pacini, C. B. Wollheim, and B. Ahren Beta-Cell-Targeted Expression of a Dominant-Negative Mutant of Hepatocyte Nuclear Factor-1{alpha} in Mice: Diabetes Model with {beta}-Cell Dysfunction Partially Rescued by Nonglucose Secretagogues Diabetes, December 1, 2004; 53(suppl_3): S92 - S96. [Abstract] [Full Text] [PDF]

				S.-Y. Kim, H.-i. Kim, T.-H. Kim, S.-S. Im, S.-K. Park, I.-K. Lee, K.-S. Kim, and Y.-H. Ahn SREBP-1c Mediates the Insulin-dependent Hepatic Glucokinase Expression J. Biol. Chem., July 16, 2004; 279(29): 30823 - 30829. [Abstract] [Full Text] [PDF]

				J. K. Divine, S. P. McCaul, and T. C. Simon HNF-1{alpha} and endodermal transcription factors cooperatively activate Fabpl: MODY3 mutations abrogate cooperativity Am J Physiol Gastrointest Liver Physiol, June 9, 2003; 285(1): G62 - G72. [Abstract] [Full Text] [PDF]

				D. Melloul, S. Marshak, and E. Cerasi Regulation of pdx-1 Gene Expression Diabetes, December 1, 2002; 51(90003): S320 - 325. [Abstract] [Full Text] [PDF]

				Y. Bai, M. Pontoglio, T. Hiesberger, A. M. Sinclair, and P. Igarashi Regulation of kidney-specific Ksp-cadherin gene promoter by hepatocyte nuclear factor-1beta Am J Physiol Renal Physiol, October 1, 2002; 283(4): F839 - F851. [Abstract] [Full Text] [PDF]

				D.-K. Lee, D. Suh, H. J. Edenberg, and M.-W. Hur POZ Domain Transcription Factor, FBI-1, Represses Transcription of ADH5/FDH by Interacting with the Zinc Finger and Interfering with DNA Binding Activity of Sp1 J. Biol. Chem., July 19, 2002; 277(30): 26761 - 26768. [Abstract] [Full Text] [PDF]

				N. Ban, Y. Yamada, Y. Someya, K. Miyawaki, Y. Ihara, M. Hosokawa, S. Toyokuni, K. Tsuda, and Y. Seino Hepatocyte Nuclear Factor-1{alpha} Recruits the Transcriptional Co-Activator p300 on the GLUT2 Gene Promoter Diabetes, May 1, 2002; 51(5): 1409 - 1418. [Abstract] [Full Text] [PDF]

				J.-y. Cha, H.-s. Kim, H.-i. Kim, S.-s. Im, S.-y. Kim, J.-w. Kim, B.-i. Yeh, and Y.-h. Ahn Analysis of Polymorphism of the GLUT2 Promoter in NIDDM Patients and its Functional Consequence to the Promoter Activity Ann. Clin. Lab. Sci., April 1, 2002; 32(2): 114 - 122. [Abstract] [Full Text] [PDF]

				K. A. Hagenfeldt-Johansson, P. L. Herrera, H. Wang, A. Gjinovci, H. Ishihara, and C. B. Wollheim {beta}-Cell-Targeted Expression of a Dominant-Negative Hepatocyte Nuclear Factor-1{alpha} Induces a Maturity-Onset Diabetes of the Young (MODY)3-Like Phenotype in Transgenic Mice Endocrinology, December 1, 2001; 142(12): 5311 - 5320. [Abstract] [Full Text] [PDF]

				W. Shen, L. M. Scearce, J. E. Brestelli, N. J. Sund, and K. H. Kaestner Foxa3 (Hepatocyte Nuclear Factor 3gamma ) Is Required for the Regulation of Hepatic GLUT2 Expression and the Maintenance of Glucose Homeostasis during a Prolonged Fast J. Biol. Chem., November 9, 2001; 276(46): 42812 - 42817. [Abstract] [Full Text] [PDF]

				M. Párrizas, M. A. Maestro, S. F. Boj, A. Paniagua, R. Casamitjana, R. Gomis, F. Rivera, and J. Ferrer Hepatic Nuclear Factor 1-{alpha} Directs Nucleosomal Hyperacetylation to Its Tissue-Specific Transcriptional Targets Mol. Cell. Biol., May 1, 2001; 21(9): 3234 - 3243. [Abstract] [Full Text]

				L. J. Bischof, C. C. Martin, C. A. Svitek, B. T. Stadelmaier, L. A. Hornbuckle, J. K. Goldman, J. K. Oeser, J. C. Hutton, and R. M. O’Brien Characterization of the Mouse Islet-Specific Glucose-6-Phosphatase Catalytic Subunit-Related Protein Gene Promoter by In Situ Footprinting: Correlation With Fusion Gene Expression in the Islet-Derived {beta}TC-3 and Hamster Insulinoma Tumor Cell Lines Diabetes, March 1, 2001; 50(3): 502 - 514. [Abstract] [Full Text]

				D. Schmoll, K. S. Walker, D. R. Alessi, R. Grempler, A. Burchell, S. Guo, R. Walther, and T. G. Unterman Regulation of Glucose-6-phosphatase Gene Expression by Protein Kinase Balpha and the Forkhead Transcription Factor FKHR. EVIDENCE FOR INSULIN RESPONSE UNIT-DEPENDENT AND -INDEPENDENT EFFECTS OF INSULIN ON PROMOTER ACTIVITY J. Biol. Chem., November 10, 2000; 275(46): 36324 - 36333. [Abstract] [Full Text] [PDF]

				E. Ben-Shushan, S. Marshak, M. Shoshkes, E. Cerasi, and D. Melloul A Pancreatic beta -Cell-specific Enhancer in the Human PDX-1 Gene Is Regulated by Hepatocyte Nuclear Factor 3beta (HNF-3beta ), HNF-1alpha , and SPs Transcription Factors J. Biol. Chem., May 11, 2001; 276(20): 17533 - 17540. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/24/18358    most recent M909536199v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Cha, J.-Y. Articles by Ahn, Y.-h. Search for Related Content PubMed PubMed Citation Articles by Cha, J.-Y. Articles by Ahn, Y.-h. Social Bookmarking                      What's this?