Foxa2 (HNF3β) Controls Multiple Genes Implicated in Metabolism-Secretion Coupling of Glucose-induced Insulin Release
2002; Elsevier BV; Volume: 277; Issue: 20 Linguagem: Inglês
10.1074/jbc.m111037200
ISSN1083-351X
AutoresHaiyan Wang, Benoit R. Gauthier, Kerstin A. Hagenfeldt-Johansson, Mariella Iezzi, Claes B. Wollheim,
Tópico(s)Diabetes and associated disorders
ResumoThe transcription factor Foxa2 is implicated in blood glucose homeostasis. Conditional expression of Foxa2 or its dominant-negative mutant DN-Foxa2 in INS-1 cells reveals that Foxa2 regulates the expression of genes important for glucose sensing in pancreatic β-cells. Overexpression of Foxa2 results in blunted glucose-stimulated insulin secretion, whereas induction of DN-Foxa2 causes a left shift of glucose-induced insulin release. The mRNA levels of GLUT2 and glucokinase are drastically decreased after induction of Foxa2. In contrast, loss of Foxa2 function leads to up-regulation of hexokinase (HK) I and II and glucokinase (HK-IV) mRNA expression. The glucokinase and the low Kmhexokinase activities as well as glycolysis are increased proportionally. In addition, induction of DN-Foxa2 also reduces the expression of β-cell KATP channel subunits Sur1 and Kir6.2 by 70%. Furthermore, in contrast to previous reports, induction of Foxa2 causes pronounced decreases in the HNF4α and HNF1α mRNA levels. Foxa2 fails to regulate the expression of Pdx1 transcripts. The expression of insulin and islet amyloid polypeptide is markedly suppressed after induction of Foxa2, while the glucagon mRNA levels are significantly increased. Conversely, Foxa2 is required for glucagon expression in these INS-1-derived cells. These results suggest that Foxa2 is a vital transcription factor evolved to control the expression of genes essential for maintaining β-cell glucose sensing and glucose homeostasis. The transcription factor Foxa2 is implicated in blood glucose homeostasis. Conditional expression of Foxa2 or its dominant-negative mutant DN-Foxa2 in INS-1 cells reveals that Foxa2 regulates the expression of genes important for glucose sensing in pancreatic β-cells. Overexpression of Foxa2 results in blunted glucose-stimulated insulin secretion, whereas induction of DN-Foxa2 causes a left shift of glucose-induced insulin release. The mRNA levels of GLUT2 and glucokinase are drastically decreased after induction of Foxa2. In contrast, loss of Foxa2 function leads to up-regulation of hexokinase (HK) I and II and glucokinase (HK-IV) mRNA expression. The glucokinase and the low Kmhexokinase activities as well as glycolysis are increased proportionally. In addition, induction of DN-Foxa2 also reduces the expression of β-cell KATP channel subunits Sur1 and Kir6.2 by 70%. Furthermore, in contrast to previous reports, induction of Foxa2 causes pronounced decreases in the HNF4α and HNF1α mRNA levels. Foxa2 fails to regulate the expression of Pdx1 transcripts. The expression of insulin and islet amyloid polypeptide is markedly suppressed after induction of Foxa2, while the glucagon mRNA levels are significantly increased. Conversely, Foxa2 is required for glucagon expression in these INS-1-derived cells. These results suggest that Foxa2 is a vital transcription factor evolved to control the expression of genes essential for maintaining β-cell glucose sensing and glucose homeostasis. The forkhead/winged-helix Foxa family of transcription factors, encoded by three genes Foxa1 (Hnf3α),Foxa2 (Hnf3β), and Foxa3(Hnf3γ), regulate hepatic and/or pancreatic gene expression (1Duncan S. Navas M. Dufort D. Rossant J. Stoffel M. Science. 1998; 281: 692-695Crossref PubMed Scopus (294) Google Scholar, 2Kaestner K.H. Hiemisch H. Schutz G. Mol. Cell. Biol. 1998; 18: 4245-4251Crossref PubMed Scopus (127) Google Scholar, 3Kaestner K. Katz J. Liu Y. Drucker D. Schutz G. Genes Dev. 1999; 13: 495-504Crossref PubMed Scopus (211) Google Scholar, 4Rausa F.M. Tan Y. Zhou H. Yoo K.W. Stolz D.B. Watkins S.C. Franks R.R. Unterman T.G. Costa R.H. Mol. Cell. Biol. 2000; 20: 8264-8282Crossref PubMed Scopus (98) Google Scholar, 5Shih D. Navas M. Kuwajima S. Duncan S. Stoffel M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 10152-10157Crossref PubMed Scopus (119) Google Scholar, 6Sund N. Ang S. Sackett S. Shen W. Daigle N. Magnuson M. Kaestner K. Mol. Cell. Biol. 2000; 20: 5175-5518Crossref PubMed Scopus (121) Google Scholar, 7Vallet V. Antoine B. Chafey P. Vandewalle A. Kahn A. Mol. Cell. Biol. 1995; 15: 5453-5460Crossref PubMed Scopus (44) Google Scholar, 8Wang J.C. Stafford J.M. Scott D.K. Sutherland C. Granner D.K. J. Biol. Chem. 2000; 275: 14717-14721Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar, 9Wu K.L. Gannon M. Peshavaria M. Offield M.F. Henderson E. Ray M. Marks A. Gamer L.W. Wright C.V. Stein R. Mol. Cell. Biol. 1997; 17: 6002-6013Crossref PubMed Google Scholar). Foxa1 and Foxa3 are required for maintaining glucose homeostasis by activation, respectively, of pancreatic glucagon and hepatic gluconeogenic enzymes (2Kaestner K.H. Hiemisch H. Schutz G. Mol. Cell. Biol. 1998; 18: 4245-4251Crossref PubMed Scopus (127) Google Scholar, 3Kaestner K. Katz J. Liu Y. Drucker D. Schutz G. Genes Dev. 1999; 13: 495-504Crossref PubMed Scopus (211) Google Scholar, 5Shih D. Navas M. Kuwajima S. Duncan S. Stoffel M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 10152-10157Crossref PubMed Scopus (119) Google Scholar). Targeted disruption of Foxa2 resulted in embryonic lethality with defective development of the foregut endoderm, from which the liver and pancreas arise (10Weinstein D.C. Ruiz i Altaba A. Chen W.S. Hoodless P. Prezioso V.R. Jessell T.M. Darnell J.E., Jr. Cell. 1994; 78: 575-588Abstract Full Text PDF PubMed Scopus (701) Google Scholar). Foxa2, which is expressed in islets, has been suggested as the upstream transactivator of Hnf4α, Hnf1α, Pdx1, and Hnf1β in the transcriptional hierarchy (1Duncan S. Navas M. Dufort D. Rossant J. Stoffel M. Science. 1998; 281: 692-695Crossref PubMed Scopus (294) Google Scholar, 9Wu K.L. Gannon M. Peshavaria M. Offield M.F. Henderson E. Ray M. Marks A. Gamer L.W. Wright C.V. Stein R. Mol. Cell. Biol. 1997; 17: 6002-6013Crossref PubMed Google Scholar, 11Kaestner K. Trends Endocrinol. Metab. 2000; 11: 281-285Abstract Full Text Full Text PDF PubMed Scopus (149) Google Scholar). Mutations in the genes encoding these pancreatic transcription factors are linked to four monogenic forms of MODY 1The abbreviations used are: MODYmaturity-onset diabetes of the youngDNdominant-negativePBSphosphate-buffered salineBSAbovine serum albuminIAPPislet amyloid polypeptideGDHglutamate dehydrogenaseANTadenine nucleotide translocatorGLP-1Rglucagon-like peptide-1 receptorUCPuncoupling protein (maturity-onset diabetes of the young): MODY1/HNF4α, MODY3/HNF1α, MODY4/IPF1(PDX1), and MODY5/HNF1β (12Hattersley A.T. Diabetes Med. 1998; 15: 15-24Crossref PubMed Scopus (266) Google Scholar, 13Ryffel G.U. J. Mol. Endocrinol. 2001; 27: 11-29Crossref PubMed Scopus (236) Google Scholar). However, the search for the association of FOXA2 mutations with MODY patients has not been successful (14Abderrahmani A. Chevre J.C. Otabe S. Chikri M. Hani E.H. Vaxillaire M. Hinokio Y. Horikawa Y. Bell G.I. Froguel P. Diabetes. 2000; 49: 306-308Crossref PubMed Scopus (10) Google Scholar, 15Hinokio Y. Horikawa Y. Furuta H. Cox N.J. Iwasaki N. Honda M. Ogata M. Iwamoto Y. Bell G.I. Diabetes. 2000; 49: 302-305Crossref PubMed Scopus (14) Google Scholar). Most recently, Sund et al. (16Sund N.J. Vatamaniuk M.Z. Casey M. Ang S.L. Magnuson M.A. Stoffers D.A. Matschinsky F.M. Kaestner K.H. Genes Dev. 2001; 15: 1706-1715Crossref PubMed Scopus (159) Google Scholar) have suggested that FOXA2 rather might be a candidate gene for familial hyperinsulinism. Pancreatic β-cell-specific deletion of Foxa2 resulted in postnatal death due to severe hyperinsulinemic hypoglycemia, and the down-regulation of ATP-sensitive K+ (KATP) channel subunits Sur1 and Kir6.2 has been demonstrated in these mutant mice (16Sund N.J. Vatamaniuk M.Z. Casey M. Ang S.L. Magnuson M.A. Stoffers D.A. Matschinsky F.M. Kaestner K.H. Genes Dev. 2001; 15: 1706-1715Crossref PubMed Scopus (159) Google Scholar). maturity-onset diabetes of the young dominant-negative phosphate-buffered saline bovine serum albumin islet amyloid polypeptide glutamate dehydrogenase adenine nucleotide translocator glucagon-like peptide-1 receptor uncoupling protein To assess whether Foxa2 indeed controls the expression of the transcription factors associated with MODY, we have established INS-1-derived stable cell lines, which allow conditional expression of the wild type Foxa2 or its dominant-negative mutant DN-Foxa2 under tight control of the reverse tetracycline-dependent transactivator (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar). DN-Foxa2 is a Myc-tagged truncated Foxa2 mutant protein that possesses the intact DNA-binding domain but lacks the transactivation domain (7Vallet V. Antoine B. Chafey P. Vandewalle A. Kahn A. Mol. Cell. Biol. 1995; 15: 5453-5460Crossref PubMed Scopus (44) Google Scholar). DN-Foxa2 exerts its dominant-negative function by competing with the endogenous Foxa2 for cognate DNA binding (7Vallet V. Antoine B. Chafey P. Vandewalle A. Kahn A. Mol. Cell. Biol. 1995; 15: 5453-5460Crossref PubMed Scopus (44) Google Scholar). The impact of altered Foxa2 function on glucose metabolism and insulin secretion was assessed in these stable clones. The gene expression profile before and after induction of the Foxa2 or DN-Foxa2 was quantified. Rat insulinoma INS-1 cell line-derived stable clones were cultured in RPMI 1640 in 11.2 mm glucose (18Wang H. Maechler P. Ritz-Laser B. Hagenfeldt K.A. Ishihara H. Philippe J. Wollheim C.B. J. Biol. Chem. 2001; 276: 25279-25286Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar), unless otherwise indicated. The first step stable clone INSrαβ, which expresses the reverse tetracycline-dependent transactivator, was described previously (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar, 18Wang H. Maechler P. Ritz-Laser B. Hagenfeldt K.A. Ishihara H. Philippe J. Wollheim C.B. J. Biol. Chem. 2001; 276: 25279-25286Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar). Plasmids used in the secondary stable transfection were constructed by subcloning the cDNAs encoding the mouse Foxa2 (kindly supplied by Prof. G. Schütz) and its dominant-negative mutant (DN-Foxa2) into the expression vector PUHD10–3 (a kind gift from Prof. H. Bujard). DN-Foxa2 (truncated mutation lacking the transactivation domain but containing the intact DNA-binding domain) (7Vallet V. Antoine B. Chafey P. Vandewalle A. Kahn A. Mol. Cell. Biol. 1995; 15: 5453-5460Crossref PubMed Scopus (44) Google Scholar) was PCR-amplified from Foxa2 cDNA using the following primers: 5′-gcaggatccgtaatggtgctcgggcttcaggtg-3′ and 5′-gcaggatccggcgccatggcgggcatgagcggctca3-′. The PCR fragment was subcloned into modified pcDNA3.1myc (Invitrogen, Groningen, The Netherlands) and sequenced. The stable transfection and the clone selection and screening procedures were described previously (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar). Immunoblotting procedures were performed as described previously using enhanced chemiluminescence (Pierce) for detection (18Wang H. Maechler P. Ritz-Laser B. Hagenfeldt K.A. Ishihara H. Philippe J. Wollheim C.B. J. Biol. Chem. 2001; 276: 25279-25286Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar). The dilutions for antibodies against Foxa2 C terminus (Santa Cruz Biotechnology, Heidelberg, Germany) and Myc-tag (19Wang H. Maechler P. Antinozzi P.A. Hagenfeldt K.A. Wollheim C.B. J. Biol. Chem. 2000; 275: 35953-35959Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar) were 1:5,000 and 1:10, respectively. Nuclear extracts were isolated from the cells cultured with or without 500 ng/ml doxycycline for 24 h. For immunofluorescence, cells grown on polyornithine-treated glass coverslips were cultured for 24 h with or without 500 ng/ml doxycycline. The cells were then washed, fixed in 4% paraformaldehyde, and permeabilized with 0.1% Triton X-100 in phosphate-buffered saline containing 1% BSA (PBS-BSA). The preparation was then blocked with PBS-BSA before incubating with the first antibodies, anti-Foxa2 (1:500 dilution) and mouse monoclonal anti-Myc-tag, (1:2 dilution), followed by the second antibody labeling. Nuclear extracts from INS-1 cells grown in culture medium with or without 500 ng/ml doxycycline for 24 h were prepared according to Schreiber et al.(20Schreiber E. Matthias P. Muller M. Schaffner W. EMBO J. 1988; 7: 4221-4229Crossref PubMed Scopus (193) Google Scholar). Cells in 12-well plates were cultured in 11.2 mm glucose medium with or without 500 ng/ml doxycycline for 19 h, followed by an additional 5 h equilibration in 2.5 mm glucose medium. Insulin secretion was measured over a period of 30 min, in Krebs-Ringer-Bicarbonate-HEPES buffer (KRBH, 140 mm NaCl, 3.6 mm KCl, 0.5 mmNaH2PO4, 0.5 mm MgSO4, 1.5 mm CaCl2, 2 mmNaHCO3, 10 mm HEPES, 0.1% BSA) containing the indicated concentrations of glucose. Insulin content was determined after extraction with acid ethanol following the procedures of Wang et al. (21Wang H. Maechler P. Hagenfeldt K.A. Wollheim C.B. EMBO J. 1998; 17: 6701-6713Crossref PubMed Google Scholar). Insulin was detected by radioimmunoassay using rat insulin as standard (21Wang H. Maechler P. Hagenfeldt K.A. Wollheim C.B. EMBO J. 1998; 17: 6701-6713Crossref PubMed Google Scholar). Cytosolic proteins were extracted, according to Wang and Iynedjian (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar), from cells cultured in 11.2 mm glucose medium in the presence or absence of 500 ng/ml doxycycline for 24 h. Total hexokinase activity was measured at 30 °C by a glucose-6-phosphate dehydrogenase-coupled assay in a fluorometer (Lambda Bio20, PerkinElmer Life Sciences) estimation of NADH production (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar). Glucokinase activity and high affinity hexokinase activity were calculated, respectively, as the differences in NADH produced at 100, 0.5, and 0 mm glucose and expressed in nmol/min (=milliunits) per mg of protein. Cells in 24-well dishes were cultured in 2.5 mm glucose medium with or without 500 ng/ml doxycycline for 24 h. The rate of glycolysis was estimated from the production of [3H]water fromd-[5-3H]glucose according to Wang and Iynedjian (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar). Cells in 10-cm diameter dishes were cultured in 2.5 mm glucose medium with or without 500 ng/ml doxycycline for 16 h, followed by an additional 8 h in culture medium with 2.5, 6, 12, and 24 mm glucose. Total RNA was extracted and blotted to nylon membranes as described previously (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar). The membrane was prehybridized and then hybridized to 32P-labeled random primer cDNA probes according to Wang and Iynedijian (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar). To ensure equal RNA loading and even transfer, all membranes were stripped and re-hybridized with a "housekeeping gene" probe cyclophilin. cDNA fragments used as probes for Foxa2, Hnf1α, Hnf4α, glucokinase, hexokinase I, Glut2, l-pyruvate kinase, insulin, Sur1, Kir6.2, and Pdx1 mRNA detection were digested from the corresponding plasmids. cDNA probes for rat islet amyloid polypeptide (IAPP), glucagon, Nkx6.1, Nkx2.2, Isl-1, β2/NeuroD, aldolase B, adenine nucleotide translocators 1 and 2 (ANT1, ANT2), mitochondrial uncoupling protein 2 (UCP2), mitochondrial glutamate dehydrogenase (GDH), citrate synthase, glyceraldehydes-3 phosphate dehydrogenase (GAPDH), hexokinase II and glucagon-like peptide-1 receptor (GLP-1R) were prepared by RT-PCR and confirmed by sequencing. Results are expressed as mean ± S.E., and statistical analyses were performed by Student's t test for unpaired data. We have established over 10 clones positively expressing Foxa2 and DN-Foxa2, respectively, using the parental INS-rαβ (INS-r3) cells (17Wang H. Iynedjian P.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4372-4377Crossref PubMed Scopus (117) Google Scholar, 18Wang H. Maechler P. Ritz-Laser B. Hagenfeldt K.A. Ishihara H. Philippe J. Wollheim C.B. J. Biol. Chem. 2001; 276: 25279-25286Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar). The clones designated as Foxa2#51 and DN-Foxa2#45, which displayed the highest inducible protein levels without leakage under noninduced state, were selected for the present study. As illustrated in Fig. 1,A and C, the INS-1-derived cells express endogenous Foxa2 in the nucleus. Foxa2 protein was overexpressed in all of the cells treated with 500 ng/ml doxycycline for 24 h. As predicted, the antibody against the carboxyl terminus of Foxa2 did not detect DN-Foxa2 with the COOH-terminal deletion (7Vallet V. Antoine B. Chafey P. Vandewalle A. Kahn A. Mol. Cell. Biol. 1995; 15: 5453-5460Crossref PubMed Scopus (44) Google Scholar) (Fig. 1B). As shown in the Western blotting (Fig. 1B) and immunostaining (Fig. 1D) with a monoclonal anti-Myc antibody, this Myc-tagged DN-Foxa2 protein was induced in a doxycycline-dependent and an all-or-none manner. Induction of DN-Foxa2 did not interrupt the endogenous Foxa2 expression (Fig. 1B), and the induced DN-Foxa2 protein was localized in the nucleus of DN-Foxa2#45 cells (Fig. 1, B and D). We also performed an electrophoretic mobility shift assay (data not shown) using the Foxa2-binding site containing glucagon G2 element as a probe (22Philippe J. Mol. Endocrinol. 1995; 9: 368-374PubMed Google Scholar). Induction of Foxa2 led to a 10-fold increase in the signal density of Foxa2 binding, whereas induction of DN-Foxa2 almost completely abolished the binding activity of endogenous Foxa2 (data not shown). As demonstrated in Fig. 2A, overexpression of Foxa2 almost completely blunted glucose-stimulated insulin release. The cellular insulin content was reduced by 46.3 ± 5.1% (p < 0.001) after induction of Foxa2 for 24 h (see also Fig. 6 for the decrease in insulin mRNA levels). Secretion data were therefore normalized for cellular insulin content. In contrast, induction of DN-Foxa2 resulted in a left shift of the dose-response curve of glucose-stimulated insulin release (Fig. 2B) without altering insulin content. To verify the clonal variability, we randomly chose another clone DN-Foxa2# 2 and studied the effects of DN-Foxa2 induction on glucose-stimulated insulin secretion. As shown in Fig. 2C, induction of DN-Foxa2 in this clone also led to a typical left-shift of glucose-dependent insulin release, suggesting a common phenomenon rather than a clonal peculiarity. Next, we examined the gene expression patterns in these cell lines to elucidate the mechanisms underlying the changes in insulin secretion.Figure 6Foxa2 suppresses β-cell gene expression and promotes glucagon levels. Cells were cultured in 2.5 mm glucose medium with or without 500 ng/ml doxycycline for 16 h and were then further incubated for 8 h at the indicated glucose concentrations. The gene expression profile in Foxa2#51 (A) and DN-Foxa2#45 (B) cells was quantified by Northern blotting. 20 μg of total RNA samples were analyzed by hybridizing with indicated cDNA probes.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Northern blot analysis of the gene expression pattern in Foxa2#51 and DN-Foxa2#45 cells cultured in indicated concentrations of glucose and treated with or without 500 ng/ml doxycycline for 24 h is described in the legend to Fig. 3. Consistent with the immunoblotting (Fig. 1B), DN-Foxa2 mRNA was induced in an all-or-none manner, and such induction did not alter endogenous Foxa2 mRNA expression (Fig. 3B). The mRNA levels of the KATP channel subunits Sur1 and Kir6.2 were reduced by 60 and 70%, respectively, after dominant-negative suppression of Foxa2 function (Fig. 3B). However, overexpression of Foxa2 alone was not sufficient to promote the expression of Sur1 and Kir6.2 (Fig. 3A). The mRNA levels of mitochondrial GDH, citrate synthase, and adenine nucleotide translocators 1 and 2 (ANT1 and ANT2) were not modulated by Foxa2 (Fig. 3, A and B). On the other hand, overexpression of Foxa2 caused up-regulation of UCP2 (Fig. 3A), whereas induction of DN-Foxa2 did not affect the expression of UCP2 mRNA (Fig. 3B). Furthermore, overexpression of Foxa2 resulted in down-regulation of glucagon-like peptide-1 receptor (GLP-1R) (Fig. 3A). Persistent hyperinsulinemic hypoglycemia of infancy has been linked to mutations in the genes encoding Sur1, Kir6.2, glucokinase, and GDH (23Glaser B. Kesavan P. Heyman M. Davis E. Cuesta A. Buchs A. Stanley C.A. Thornton P.S. Permutt M.A. Matschinsky F.M. Herold K.C. N. Engl. J. Med. 1998; 338: 226-230Crossref PubMed Scopus (519) Google Scholar, 24Meissner T. Beinbrech B. Mayatepek E. Hum. Mutat. 1999; 13: 351-361Crossref PubMed Scopus (40) Google Scholar, 25Nestorowicz A. Wilson B.A. Schoor K.P. Inoue H. Glaser B. Landau H. Stanley C.A. Thornton P.S. Clement J.P.t. Bryan J. Aguilar-Bryan L. Permutt M.A. Hum. Mol. Genet. 1996; 5: 1813-1822Crossref PubMed Scopus (241) Google Scholar, 26Stanley C.A. Fang J. Kutyna K. Hsu B.Y. Ming J.E. Glaser B. Poncz M. Diabetes. 2000; 49: 667-673Crossref PubMed Scopus (159) Google Scholar, 27Thomas P.M. Cote G.J. Wohllk N. Haddad B. Mathew P.M. Rabl W. Aguilar-Bryan L. Gagel R.F. Bryan J. Science. 1995; 268: 426-429Crossref PubMed Scopus (747) Google Scholar, 28Kane C. Shepherd R.M. Squires P.E. Johnson P.R. James R.F. Milla P.J. Aynsley-Green A. Lindley K.J. Dunne M.J. Nat. Med. 1996; 2: 1344-1347Crossref PubMed Scopus (225) Google Scholar). Increased glucose-dependent insulin release was also observed in the UCP2-deleted mouse (29Zhang C.Y. Baffy G. Perret P. Krauss S. Peroni O. Grujic D. Hagen T. Vidal-Puig A.J. Boss O. Kim Y.B. Zheng X.X. Wheeler M.B. Shulman G.I. Chan C.B. Lowell B.B. Cell. 2001; 105: 745-755Abstract Full Text Full Text PDF PubMed Scopus (825) Google Scholar), whereas decreased insulin secretion was reported after overexpression of UCP2 in islets (30Chan C.B., De Leo D. Joseph J.W. McQuaid T.S. Ha X.F. Xu F. Tsushima R.G. Pennefather P.S. Salapatek A.M. Wheeler M.B. Diabetes. 2001; 50: 1302-1310Crossref PubMed Scopus (316) Google Scholar) and INS-1 cells (31Lameloise N. Muzzin P. Prentki M. Assimacopoulos-Jeannet F. Diabetes. 2001; 50: 803-809Crossref PubMed Scopus (214) Google Scholar). We could rule out the possible involvement of GDH and UCP2 in the enhanced glucose-stimulated insulin secretion observed in β-cells deficient in Foxa2 function, since their expression was not altered by induction of DN-Foxa2. The rodent pancreatic β-cell expresses high levels of the glucose transporter Glut2, which allows rapid equilibration of glucose across the plasma membrane (32Thorens B. Sarkar H.K. Kaback H.R. Lodish H.F. Cell. 1988; 55: 281-290Abstract Full Text PDF PubMed Scopus (661) Google Scholar, 33Schuit F.C. Huypens P. Heimberg H. Pipeleers D.G. Diabetes. 2001; 50: 1-11Crossref PubMed Scopus (316) Google Scholar). This is associated with extremely low levels of high affinity hexokinase isoforms (hexokinases I, II, and III) to optimize glucose sensing in the physiological blood glucose range. A β-cell-specific promoter in the glucokinase (hexokinase IV) gene maintains a precise expression level of this rate-limiting enzyme for glucose metabolism, which determines the glucose sensing in pancreatic β-cells (reviewed in Refs. 33Schuit F.C. Huypens P. Heimberg H. Pipeleers D.G. Diabetes. 2001; 50: 1-11Crossref PubMed Scopus (316) Google Scholar, 34Matschinsky F.M. Diabetes. 1996; 45: 223-241Crossref PubMed Scopus (0) Google Scholar, 35Matschinsky F.M. Glaser B. Magnuson M.A. Diabetes. 1998; 47: 307-315Crossref PubMed Scopus (291) Google Scholar). Alterations of glucokinase activity by gene manipulation or pharmacological inhibition, or by naturally occurring genetic mutations, have been demonstrated to change the physiological threshold of β-cell glucose sensing (reviewed in Refs. 33Schuit F.C. Huypens P. Heimberg H. Pipeleers D.G. Diabetes. 2001; 50: 1-11Crossref PubMed Scopus (316) Google Scholar, 34Matschinsky F.M. Diabetes. 1996; 45: 223-241Crossref PubMed Scopus (0) Google Scholar, 35Matschinsky F.M. Glaser B. Magnuson M.A. Diabetes. 1998; 47: 307-315Crossref PubMed Scopus (291) Google Scholar). As shown in Fig. 4, overexpression of Foxa2 in Foxa2#51 cells reduced the glucokinase mRNA level by 60%, whereas induction of DN-Foxa2 in DN-Foxa2#45 raised the glucokinase expression by 2-fold. The increased glucokinase expression after induction of DN-Foxa2 was also demonstrated in another clone, DN-Foxa2#2 (Fig. 7). The INS-1-derived clones expressed hexokinases I and II (but not III) mRNAs at barely detectable levels, and induction of Foxa2 and DN-Foxa2 resulted in, respectively, down- and up-regulation of these mRNA levels (Fig. 4). Overexpression of Foxa2 also caused a 90% reduction of Glut2 mRNA expression, while induction of DN-Foxa2 left-shifted the glucose dose-dependent increase in Glut2 transcript level (Fig. 4). The suppressive effects of Foxa2 on glucose sensing were also reflected by the blunted glucose responsiveness ofl-pyruvate kinase and aldolase B mRNA expression (Fig. 4).Figure 7Graded overexpression of Foxa2 suppresses HNF4α and HNF1α expression and induction of DN-Foxa2 increases glucokinse expression.A, Foxa2#39 cells were cultured for 24 h in normal (11.2 mm) glucose medium containing, respectively, 0, 75, 150, and 500 ng/ml doxycycline. Samples from two independent experiments were demonstrated in parallel.B, DN-Foxa2#2 cells were cultured in 2.5 mm glucose medium with or without 500 ng/ml doxycycline for 16 h and were then further incubated for 8 h at the indicated glucose concentrations. The gene expression was quantified by Northern blotting. 20 μg of total RNA samples were analyzed by hybridizing with indicated cDNA probes.View Large Image Figure ViewerDownload Hi-res image Download (PPT) To confirm the Northern blot analysis, we also measured the activities of glucokinase and high affinity hexokinase. As seen in Table I, the glucokinase activity was reduced by 60% following overexpression of Foxa2 and was increased 2.5-fold by dominant-negative suppression of Foxa2 function. Similarly, the high affinity hexokinase activity was down-regulated by 50% and up-regulated by 3-fold, respectively, by induction of Foxa2 and DN-Foxa2. Thus, Foxa2 is essential for the transcriptional regulation of enzymes controlling the β-cell glucose phosphorylation. This conclusion was corroborated by the measurements of glycolytic flux, which was decreased by 60% after overexpression of Foxa2 and increased by 2-fold after induction of DN-Foxa2 (Fig. 5).Table IEffects of induction of Foxa2 and DN-Foxa2 on the activities of glucokinase and high affinity hexokinaseFoxa2#51DN-Foxa2#45−Dox+Dox−Dox+DoxGlucokinase10.18 ± 0.973.94 ± 1.241-ap < 0.0001.9.53 ± 1.7024.60 ± 3.781-ap < 0.0001.High affinity hexokinase1.30 ± 0.360.56 ± 0.351-bp < 0.005.1.32 ± 0.464.20 ± 1.531-ap < 0.0001.Enzyme activities were measured using cytosolic proteins isolated from cells cultured with or without 500 ng/ml doxycycline for 24 h and expressed as milliunits/mg of protein. Data represent means ± S.E. of seven to nine separate experiments.1-a p < 0.0001.1-b p < 0.005. Open table in a new tab Enzyme activities were measured using cytosolic proteins isolated from cells cultured with or without 500 ng/ml doxycycline for 24 h and expressed as milliunits/mg of protein. Data represent means ± S.E. of seven to nine separate experiments. Foxa2 has been previously suggested as a master transactivator of the pancreatic transcription factors, Hnf4α, Hnf1α, Hnf1β, and Pdx1, in the transcriptional hierarchy (1Duncan S. Navas M. Dufort D. Rossant J. Stoffel M. Science. 1998; 281: 692-695Crossref PubMed Scopus (294) Google Scholar, 9Wu K.L. Gannon M. Peshavaria M. Offield M.F. Henderson E. Ray M. Marks A. Gamer L.W. Wright C.V. Stein R. Mol. Cell. Biol. 1997; 17: 6002-6013Crossref PubMed Google Scholar,11Kaestner K. Trends Endocrinol. Metab. 2000; 11: 281-285Abstract Full Text Full Text PDF PubMed Scopus (149) Google Scholar). The results we obtained were unexpected and in disagreement with previous reports (1Duncan S. Navas M. Dufort D. Rossant J. Stoffel M. Science. 1998; 281: 692-695Crossref PubMed Scopus (294) Google Scholar, 9Wu K.L. Gannon M. Peshavaria M. Offield M.F. Henderson E. Ray M. Marks A. Gamer L.W. Wright C.V. Stein R. Mol. Cell. Biol. 1997; 17: 6002-6013Crossref PubMed Google Scholar). We found that Pdx1 expression was not significantly affected by induction of Foxa2 or DN-Foxa2 (Fig. 6). Isl-1 is the only pancreatic transcription factor, the expression of which requires Foxa2 function (Fig. 6B). Overexpression of Foxa2 suppressed rather than enhanced the expression of Hnf4α and Hnf1α mRNAs (Fig. 6A). To verify whether this is due to a clonal variability or a paradoxical effect of high level overexpression, we also studied the effect of graded overexpression of Foxa2 on mRNA levels of Hnf4α and Hnf1α in another randomly selected clone, Foxa2#39 (Fig. 7). Titrated overexpression of Foxa2 by 3.5-, 10-, and 20-fold at 75, 150, and 500 ng/ml of doxycycline all caused significant inhibition of Hnf4α and Hnf1α
Referência(s)