Loss-of-function and Dominant-negative Mechanisms Associated with Hepatocyte Nuclear Factor-1β Mutations in Familial Type 2 Diabetes Mellitus
1999; Elsevier BV; Volume: 274; Issue: 19 Linguagem: Inglês
10.1074/jbc.274.19.12975
ISSN1083-351X
AutoresHideaki Tomura, Hidekazu Nishigori, Kimie Sho, Kazuya Yamagata, Ituro Inoue, Jun Takeda,
Tópico(s)Epigenetics and DNA Methylation
ResumoHepatocyte nuclear factor (HNF)-1β, a homeodomain-containing transcription factor, regulates gene expression in a dimerized form in pancreas, liver, and some other tissues. Recent genetic studies have identified two HNF-1β mutations, R177X and A263fsinsGG, in subjects with a monogenic form of type 2 diabetes. Despite the defects being in the same gene, diverse severities of disease are observed in the affected subjects. To investigate the molecular mechanism by which mutations might cause various phenotypic features, wild type and mutant proteins were transiently expressed in insulin-producing (MIN6) and hepatic (HepG2) cells. Luciferase reporter assay showed that both mutations resulted in a marked reduction of transactivation activity. Because their dimerization activity was found to be intact by the yeast two-hybrid system, it was possible that they were dominant-negative to wild type activity. When co-expressed with wild type, both of the mutants significantly decreased wild type activity in HepG2 cells. In contrast, although A263fsinsGG functioned similarly in MIN6 cells, R177X failed to affect wild type activity in this cell line. Immunohistochemical analysis of the mutants suggests that this functional divergence might be generated by the modification of nuclear localization. These results suggest that HNF-1β mutations may impair pancreatic β-cell function by loss-of-function and dominant-negative mechanisms. Hepatocyte nuclear factor (HNF)-1β, a homeodomain-containing transcription factor, regulates gene expression in a dimerized form in pancreas, liver, and some other tissues. Recent genetic studies have identified two HNF-1β mutations, R177X and A263fsinsGG, in subjects with a monogenic form of type 2 diabetes. Despite the defects being in the same gene, diverse severities of disease are observed in the affected subjects. To investigate the molecular mechanism by which mutations might cause various phenotypic features, wild type and mutant proteins were transiently expressed in insulin-producing (MIN6) and hepatic (HepG2) cells. Luciferase reporter assay showed that both mutations resulted in a marked reduction of transactivation activity. Because their dimerization activity was found to be intact by the yeast two-hybrid system, it was possible that they were dominant-negative to wild type activity. When co-expressed with wild type, both of the mutants significantly decreased wild type activity in HepG2 cells. In contrast, although A263fsinsGG functioned similarly in MIN6 cells, R177X failed to affect wild type activity in this cell line. Immunohistochemical analysis of the mutants suggests that this functional divergence might be generated by the modification of nuclear localization. These results suggest that HNF-1β mutations may impair pancreatic β-cell function by loss-of-function and dominant-negative mechanisms. Noninsulin-dependent (type 2) diabetes mellitus is a genetic disorder characterized by elevated plasma glucose levels due to an absolute or relative deficiency of insulin. Recent progress in modern genetics research makes precise localization of disease genes within the genome possible, and this is especially relevant in the study of monogenic diseases. Maturity onset diabetes of the young (MODY) 1The abbreviation used is: MODY, maturity onset diabetes of the young; GLUT2, glucose transporter type 2; HNF, hepatocyte nuclear factor; HA, hemagglutinin; DOTAP, N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulfate. is a monogenic form of type 2 diabetes characterized by onset usually under 25 years of age and autosomal dominant inheritance (1Fajans S.S. Diabetes Care. 1990; 13: 49-64Crossref PubMed Scopus (165) Google Scholar). Genetic linkage studies first localized three genes responsible for the development of MODY on chromosomes 20, 7, and 12 (2Bell G.I. Xiang K.-S. Newman M.V. Wu S.-H. Wright L.G. Fajans S.S. Spielman R.S. Cox N.J. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 1484-1488Crossref PubMed Scopus (309) Google Scholar, 3Froguel P. Vaxillaire M. Sun F. Velho G. Zouali H. Butel M.O. Lesage S. Vionnet N. Clement K. Fougerousse F. Tanizawa Y. Weissenbach J. Beckmann J.S. Lathrop G.M. Passa P. Permutt M.A. Cohen D. Nature. 1992; 356: 162-164Crossref PubMed Scopus (571) Google Scholar, 4Vaxillaire M. Boccio V. Philippi A. Vigouroux C. Terwilliger J. Passa P. Beckmann J.S. Velho G. Lathrop G.M. Froguel P. Nat. Genet. 1995; 9: 418-423Crossref PubMed Scopus (219) Google Scholar). The third form of MODY was found to result from mutations in the gene encoding hepatocyte nuclear factor (HNF)-1α (5Yamagata K. Oda N. Kaisaki P.J. Menzel S. Furuta H. Vaxillaire M. Cox R.D. Lathrop G.M. Boriraj V.V. Chen X. Cox N.J. Oda Y. Yano H. LeBeau M.M. Yamada S. Nishigori H. Takeda J. Chevre J.-C. Fajans S.S. Hattersley A.T. Iwasaki N. Pedersen O. Polonsky K.S. Turner R.C. Froguel P. Bell G.I. Nature. 1996; 384: 455-458Crossref PubMed Scopus (1055) Google Scholar), a homeodomain-containing transcription factor (6Frain M. Swart G. Monaci P. Nicosia A. Stampfli S. Frank R. Cortese R. Cell. 1989; 59: 145-157Abstract Full Text PDF PubMed Scopus (338) Google Scholar, 7Tronche F. Yaniv M. BioEssays. 1992; 14: 579-589Crossref PubMed Scopus (194) Google Scholar, 8Mendel D.B. Crabtree G.R. J. Biol. Chem. 1991; 266: 677-680Abstract Full Text PDF PubMed Google Scholar). HNF-1α forms a homodimer or heterodimer with structurally related HNF-1β (9Mendel D.B. Hansen L.P. Graves M.K. Conley P.B. Crabtree G.R. Genes Dev. 1991; 5: 1042-1056Crossref PubMed Scopus (251) Google Scholar, 10Rey-Campos J. Chouard T. Yaniv M. EMBO J. 1991; 10: 1445-1457Crossref PubMed Scopus (241) Google Scholar, 11De Simone De Magistris L. Lazzaro D. Gerstner J. Monaci P. Nicosia A. Cortese R. EMBO J. 1991; 10: 1435-1443Crossref PubMed Scopus (152) Google Scholar), and they function together to regulate gene expression in liver, pancreas, and some other tissues. In this context, the HNF-1β gene was also screened for mutations in subjects with early onset type 2 diabetes/MODY, and two mutations of R177X and A263fsinsGG were found in two Japanese families (12Horikawa Y. Iwasaki N. Hara M. Furuta H. Hinokio Y. Cockburn B.N. Linder T. Yamagata K. Ogata M. Tomonaga O. Kuroki H. Kasahara T. Iwamoto Y. Bell G.I. Nat. Genet. 1997; 17: 384-385Crossref PubMed Scopus (757) Google Scholar, 13Nishigori H. Yamada S. Kohama T. Tomura H. Sho K. Horikawa Y. Bell G.I. Takeuchi T. Takeda J. Diabetes. 1998; 47: 1354-1355Crossref PubMed Scopus (133) Google Scholar). HNF-1β is a protein of 557 amino acids comprising three functional domains: a dimerization domain (residues 1–32), a DNA-binding domain with a POU subdomain and a homeosubdomain (residues 88–178 and 229–299, respectively), and a transactivation domain (residues 314–557) (9Mendel D.B. Hansen L.P. Graves M.K. Conley P.B. Crabtree G.R. Genes Dev. 1991; 5: 1042-1056Crossref PubMed Scopus (251) Google Scholar, 10Rey-Campos J. Chouard T. Yaniv M. EMBO J. 1991; 10: 1445-1457Crossref PubMed Scopus (241) Google Scholar, 11De Simone De Magistris L. Lazzaro D. Gerstner J. Monaci P. Nicosia A. Cortese R. EMBO J. 1991; 10: 1435-1443Crossref PubMed Scopus (152) Google Scholar). The R177X nonsense mutation generates a truncated protein of 176 amino acids with the N-terminal dimerization and POU domains. The A263fsinsGG frameshift mutation due to insertion of a GG dinucleotide also generates a truncated mutation of 264 amino acids that lacks a third helix structure at the C terminus of the homeodomain and the entire transactivation domain. Interestingly, although the same gene is mutated, differing clinical features including diverse severity and inconsistency of onset age of diabetes are observed in these families (12Horikawa Y. Iwasaki N. Hara M. Furuta H. Hinokio Y. Cockburn B.N. Linder T. Yamagata K. Ogata M. Tomonaga O. Kuroki H. Kasahara T. Iwamoto Y. Bell G.I. Nat. Genet. 1997; 17: 384-385Crossref PubMed Scopus (757) Google Scholar, 13Nishigori H. Yamada S. Kohama T. Tomura H. Sho K. Horikawa Y. Bell G.I. Takeuchi T. Takeda J. Diabetes. 1998; 47: 1354-1355Crossref PubMed Scopus (133) Google Scholar). In this study, the molecular mechanisms by which mutations in the HNF-1β gene might cause diabetes and generate the diverse phenotypic features were addressed by functional analysis of the mutant proteins. The mutations were generated by polymerase chain reaction-based site-directed mutagenesis and cloned in pSP72 (Promega, Madison, WI) to generate the pSP-R177X and pSP-A263fsinsGG and also in the expression vector pCMV-6b to generate the pCMV-R177X and pCMV-A263fsinsGG. Wild type HNF-1β was also cloned in pSP72 and pCMV6b to generate pSP-WT and pCMV-WT, respectively. For luciferase reporter assay, the promoter region (nucleotides −1296/+312) of the human gene for GLUT2 (14Takeda J. Kayano T. Fukumoto H. Bell G.I. Diabetes. 1993; 42: 773-777Crossref PubMed Scopus (59) Google Scholar), a liver and pancreatic β-cell glucose transporter, was cloned in the pGL3-Basic Reporter vector (Promega) to generate the pGL3-GT2. The consensuscis-element for HNF-1α/-1β binding locates at nucleotide −1030 of the promoter region of the GLUT2 gene. For immunohistochemical analysis, the nucleotide sequence encoding the HA epitope of YPYDVPDYA was introduced in frame at the 3′ end of the sequence for R177X, A263fsinsGG, and wild type in the expression vector to generate the pCMV-R177X-HA, pCMV-A263fsinsGG-HA, and pCMV-WT-HA, respectively. For verification of dimerization activity, yeast two-hybrid analysis (15Field S. Song O. Nature. 1989; 340: 245-247Crossref PubMed Scopus (4875) Google Scholar) was performed to monitor interaction of the mutant protein with wild type. The entire coding region of the wild type and mutant proteins were fused to the yeast GAL4 DNA-binding and transcriptional activation domains in pAS2–1 and pACT2 (CLONTECH), respectively. HepG2 and MIN6 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10 and 15% fetal calf serum, respectively. Cells were transfected with the liposomal DOTAP/nucleic acid mixture including 5 μg of DOTAP (Roche Molecular Biochemicals), 333 ng of pGL3-GT2, 0–96 ng of pCMV-WT, pCMV-R177X, and pCMV-A263fsinsGG, and 17 ng of pRL (Renillaluciferase)-SV40. Luciferase reporter assay was performed using a Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions. Renilla luciferase activity was used to normalize transfection efficiencies among experiments. Wild type and mutant proteins were prepared by in vitro transcription and translation with pSP-WT, pSP-R177X, and pSP-A263fsinsGG using a TNT Coupled Reticulocyte Lysate System (Promega) according to the manufacturer's instructions. For electrophoretic mobility shift assay, in vitrotranslated products of the mutant and wild type proteins were incubated in a 20-μl reaction containing 10 mm Hepes, pH 7.4, 10% glycerol, 25 mm KCl, 5 mm MgCl2, 2 mm spermidine, 0.1 mm EDTA, 0.5 mmdithiothreitol, 2 μg of poly(dI-dC), and 0.3 pmol of32P-labeled double-stranded oligonucleotides (10,000 cpm) of 5′-ACCTCAGTAAAGATTAACCATCA-3′. DNA-protein complexes were resolved on a 5% nondenaturing polyacrylamide gel in 0.25 × TBE (45 mm Tris borate, 45 mm boric acid, and 2 mm EDTA). The gel was then dried and exposed to x-ray film for autoradiography. Verification of two-hybrid interactions of wild type and mutant proteins was performed according to the manufacturer's instructions. Briefly, the yeast strain YRG-2 was transformed by electroporation with the above bait and prey constructs in pAS2–1 and pACT2, respectively. The yeast competent cells used were purchased from Stratagene (La Jolla, CA). Cotransformation mixtures were plated on synthetic dropout medium (SD/-Trp/-Leu/-His) and incubated at 30 °C for several days to select for colonies expressing interacting hybrid proteins. If the target protein interacts with a test protein, a functional GAL4 activator is reconstituted, and the expression of theHIS3 reporter gene is activated. For qualitative blue/white screening of the His+-cotransformants, colony lift β-galactosidase filter assay was performed. MIN6 and HepG2 cells were transfected with the pCMV-R177X-HA, pCMV-A263fsinsGG-HA, and pCMV-WT-HA and grown on the slide containing appropriate medium. After 48 h of culture, cells were fixed with 3% formaldehyde, washed with phosphate-buffered saline, and incubated in the primary antibody at 37 °C for 30 min. After washing with phosphate-buffered saline, the cells were incubated in the fluorescence-labeled secondary antibody at 37 °C for 30 min. Localization of fusion proteins within the cells was analyzed using fluorescence microscopy. The mouse anti-HA primary antibody and secondary antibody of fluorescein isothiocyanate-conjugated anti-mouse IgG were purchased from Roche Molecular Biochemicals. Nuclear DNA was stained with 4′,6-diamidino-2-phenylindole. To investigate the molecular mechanism by which HNF-1β mutations cause impaired glucose tolerance, transactivation activities of R177X and A263fsinsGG were analyzed in liver cell and pancreatic β-cell lines by luciferase reporter assay using the promoter of the human gene for GLUT2, which mediates facilitative glucose transport in these tissues. Wild type HNF-1β bound to the cis-element and efficiently increased reporter gene activity directed by transcription from the GLUT2 gene promoter in HepG2 and MIN6 cells (Figs. 1 and 2A). When the region from −1056 to −1026 containing the cis-element for binding was deleted from the promoter segment, the reporter activity was significantly reduced (85 ± 4%, p < 0.001). However, the deleted reporter construct still represented a significant increase of activity mediated by wild type HNF-1β, suggesting that GLUT2 gene expression could also be indirectly regulated by other transcription factors whose expression is up-regulated by HNF-1β. When the HNF-1β mutations of R177X and A263fsinsGG were expressed, a marked reduction of transactivation activity of the mutant proteins was observed both in MIN6 and HepG2 cells.Figure 2A, transactivation activity of wild type (WT) and mutant proteins. pGL3-GT2 and each test plasmid were cotransfected into MIN6 and HepG2 cells. The relative luciferase activity (Firefly/Renilla luciferase) of each construct at 0, 6, 12, and 24 ng for MIN6 cells and at 0, 24, 48, and 96 ng for HepG2 cells was measured by four independent experiments. Mean ± S.D. is shown. **, p < 0.01. B, co-expression of wild type and each mutant protein. pGL3-GT2, wild type, and each mutant plasmid were co-transfected into MIN6 and HepG2 cells. The effect of increasing amounts (12 and 24 ng for MIN6 cells and 24, 48, and 96 ng for HepG2 cells) of each test plasmid on the wild type activity was examined. The open bar on the left side indicates the endogenous transcription activity. The other bars indicate the activities of the WT (wild type) alone (open bar), WT+R177X (hatched bar), and WT+A263fsinsGG (filled bar). **, p < 0.01.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Because A263fsinsGG retains the greater part of the DNA-binding domain, it is possible that this mutation may interfere with DNA binding of wild type in a dominant-negative manner. To determine whether A263fsinsGG may compete with wild type to bind the target DNA sequence to diminish normal activity, the electrophoretic mobility shift assay was performed, resulting in no DNA binding activity in A263fsinsGG (Fig. 1A). Because the C terminus of the helix-turn-helix motif of Drosophila homeodomain proteins has been shown to be necessary to confer the DNA binding specificity (16Treisman J. Gonczy P. Vashishtha M. Harris E. Desplan C. Cell. 1989; 59: 553-562Abstract Full Text PDF PubMed Scopus (398) Google Scholar), the third helix structure of the DNA-binding domain of human HNF-1β, in which 21 amino acids are looped out of the turn between helix 2 and helix 3 unlike other homeodomain proteins (8Mendel D.B. Crabtree G.R. J. Biol. Chem. 1991; 266: 677-680Abstract Full Text PDF PubMed Google Scholar), also appears to be crucial for DNA binding. Because the dimerization domain at the N terminus is retained in both mutations, the nonfunctional truncated proteins may form a heterodimer with wild type HNF-1β and so interfere with the normal function. To test this possibility, dimerization activity of the mutants was examined using the yeast two-hybrid system (15Field S. Song O. Nature. 1989; 340: 245-247Crossref PubMed Scopus (4875) Google Scholar). The bait and prey DNA constructs were introduced into the same host cells to verify mutual interaction. When either pAS2-WT (bait) or pACT2-WT (prey) for wild type was introduced into the cells with a counter vector, no reporter activity was observed, suggesting that HNF-1β alone does not have an autonomous function that activates GAL4 gene expression. When both pAS2–1-WT and pACT2-WT were introduced together, the efficient growth of transformants in the selection medium and blue color formation by β-galactosidase assay were observed, indicating that the specific HNF-1β dimerization should be readily detected by the two-hybrid system. Using this method, the dimerizing activity of A263fsinsGG and R177X was examined in relation to wild type, resulting in efficient growth of cotransformants in the selection medium and generation of β-galactosidase activity (TableI). The positive interaction of each mutation with wild type was confirmed by switching the test molecules in bait and prey constructs.Table IVerification of dimerization activity of wild type and mutant proteinsPreyBaitpAS2–1HNF-1βR177XA263fsinsGGpACT2−−−−HNF-1β−+++R177X−+±NDA263fsinsGG−+ND+Dimerization activity of the HNF-1β mutants was examined using yeast two-hybrid system. − indicates no growth of cotransformants in the selection media (SD/−Trp/−Leu/−His). + indicates the efficient growth of the cotransformants in the selection media and the generation of positive β-galactosidase activity. ± indicates the efficient growth of the cotransformants in the selection medium but extremely weak generation of β-galactosidase activity. pAS2–1 and pACT2 are bait and prey plasmid vectors, respectively. ND, not done. Open table in a new tab Dimerization activity of the HNF-1β mutants was examined using yeast two-hybrid system. − indicates no growth of cotransformants in the selection media (SD/−Trp/−Leu/−His). + indicates the efficient growth of the cotransformants in the selection media and the generation of positive β-galactosidase activity. ± indicates the efficient growth of the cotransformants in the selection medium but extremely weak generation of β-galactosidase activity. pAS2–1 and pACT2 are bait and prey plasmid vectors, respectively. ND, not done. To determine whether A263fsinsGG or R177X acts as a dominant-negative regulator, varied amounts (molar ratio, 1:1 to 1:2 in MIN6 cells and 1:0.5 to 1:2 in HepG2 cells) of the mutant constructs were expressed together with wild type, and the luciferase reporter activity driven by the GLUT2 gene promoter was measured (Fig. 2B). When equimolar amounts of DNA for the wild type and each mutant were transfected into HepG2 cells, both of the mutations significantly reduced wild type activity. With increasing amounts of mutant DNA, both of the mutations were also found to reduce normal activity, suggesting that A263fsinsGG and R177X both function as a dominant-negative regulator in hepatic cells. Interestingly, however, when the mutations were expressed in MIN6 cells, although A263fsinsGG also similarly reduced wild type activity in this cell line, R177X failed to affect wild type activity, suggesting that A263fsinsGG and R177X function differently in pancreatic β-cells. To determine whether a difference in intracellular localization of mutant proteins could affect transcription activity, localization of HA-tagged proteins within the HepG2 and MIN6 cells was examined by fluorescence microscopy (Fig. 3). Strong signals for A263fsinsGG and wild type were similarly observed only in the nucleus in both cells. In contrast, the distribution of R177X within the cells was found to be differently patterned and markedly modified among the cells. In MIN6 cells, 90% of the cells transfected with R177X showed strong signals only in the cytoplasm, whereas the other 6 and 4% of the cells showed relatively weak signals in both nucleus and cytoplasm or only in the nucleus, respectively; 50–100 cells in two experiments were counted to estimate the frequency of each pattern. On the other hand, 92% of the transfected HepG2 cells showed strong staining in the nucleus (nucleus only, 19%; nucleus and cytoplasm, 73%). The other 8% of the cells showed only cytoplasmic staining. Although the frequency of nuclear or cytoplasmic staining varied to some extent in two experiments, the general patterns of intracellular localization were similar. The modified localization of R177X within the cells might be due to lack of a short stretch of basic residues (residues 229–237) in the POU domain, which has been suggested as a nuclear localization signal (16Treisman J. Gonczy P. Vashishtha M. Harris E. Desplan C. Cell. 1989; 59: 553-562Abstract Full Text PDF PubMed Scopus (398) Google Scholar,17Dingwall C. Laskey R.A. Annu. Rev. Cell Biol. 1986; 2: 367-390Crossref PubMed Scopus (336) Google Scholar). Liver and pancreatic β-cells play a central role in regulation of glucose homeostasis by glucose uptake and disposal and insulin secretion, respectively, in response to the levels of plasma glucose. Because expression of the genes involved in such liver- and β-cell-specific function is regulated by a tissue-specific subset of transcription factors, molecular defects of these factors could lead to the development of impaired glucose tolerance. In this study, the HNF-1β mutations of R177X and A263fsinsGG were found to have markedly reduced activity of transactivation of the human GLUT2 gene in liver or pancreatic β-cell lines. These results are consistent with the contribution of the mutations in the development of diabetes in patients with the mutation. Because a functional loss of GLUT2 has been suggested to be responsible for hyperglycemia and relative hypoinsulinemia in the fed state in patients with Fanconi-Bickel syndrome and early onset diabetes in mice lacking GLUT2 (18Santer R. Schneppenheim R. Dombrowski A. Gotze H. Steinmann B. Schaub J. Nat. Genet. 1997; 17: 324-326Crossref PubMed Scopus (248) Google Scholar, 19Guillam M.-T. Hummler E. Schaerer E. Wu J.-Y. Birnbaum M.J. Beermann F. Schmidt A. Deriaz N. Thorens B. Nat. Genet. 1997; 17: 327-330Crossref PubMed Scopus (344) Google Scholar), the decreased expression of GLUT2, which could be generated by HNF-1β mutations, might be involved in the pathogenesis of hyperglycemia in patients in concert with other target gene defects. In this regard, further identification of the target genes of HNF-1β, which are expressed in pancreatic β-cells and liver, is important to better understand the pathogenesis of HNF-1β-deficient diabetes. The present study also shows that the two mutations function differently in pancreatic β-cells, possibly due to the difference of intracellular localization of the protein. Because the DNA binding activity of wild type is found to be reduced in the presence of A263fsinsGG (Fig. 1C), mutation in pancreatic β cells could interfere with the wild type regulation of target gene expression. Accordingly, the differing functional properties of the mutations in pancreatic β-cells may explain, at least in part, the differing severity of insulin secretion deficiency in the affected subjects. However, because only two mutations have so far been reported, further identification of families with the mutation will be necessary to understand to what extent these distinct molecular behaviors could account for differing clinical features in patients. Because the primary structures of HNF-1β and HNF-1α, which recognize the same DNA sequence, are highly related and the tissues that express these proteins are mostly overlapped except lung and ovary (6Frain M. Swart G. Monaci P. Nicosia A. Stampfli S. Frank R. Cortese R. Cell. 1989; 59: 145-157Abstract Full Text PDF PubMed Scopus (338) Google Scholar, 7Tronche F. Yaniv M. BioEssays. 1992; 14: 579-589Crossref PubMed Scopus (194) Google Scholar, 8Mendel D.B. Crabtree G.R. J. Biol. Chem. 1991; 266: 677-680Abstract Full Text PDF PubMed Google Scholar, 9Mendel D.B. Hansen L.P. Graves M.K. Conley P.B. Crabtree G.R. Genes Dev. 1991; 5: 1042-1056Crossref PubMed Scopus (251) Google Scholar, 10Rey-Campos J. Chouard T. Yaniv M. EMBO J. 1991; 10: 1445-1457Crossref PubMed Scopus (241) Google Scholar, 11De Simone De Magistris L. Lazzaro D. Gerstner J. Monaci P. Nicosia A. Cortese R. EMBO J. 1991; 10: 1435-1443Crossref PubMed Scopus (152) Google Scholar), the structure and function relationships in HNF-1β may also apply to HNF-1α. The possible nuclear localization sequence of HNF-1β, which is absent in R177X, is also found in HNF-1α (residues 197–205), so the modification of nuclear targeting could be observed in HNF-1α mutations involved in this region (20Yamada S. Tomura H. Nishigori H. Sho K. Mabe H. Iwatani N. Takumi T. Kito Y. Moriya N. Muroya K. Ogata T. Onigata K. Morikawa A. Inoue I. Takeda J. Diabetes. 1999; 48: 645-648Crossref PubMed Scopus (53) Google Scholar). Accordingly, functional analysis of the HNF-1β mutations should be helpful to clarify the molecular mechanism of impaired insulin secretion not only in HNF-1β-deficient diabetes but also in the related form of HNF-1α-deficient diabetes. We thank Dr. J. Miyazaki (Osaka University, Osaka, Japan) for providing MIN6 cells.
Referência(s)