Mutagenesis of the rat insulin II 5'-flanking region defines sequences important for expression in HIT cells

In vivo evaluation of GG2-GG1/A2 element activity in the insulin promoter region using the CRISPR-Cas9 system

The insulin promoter is regulated by ubiquitous as well as pancreatic β-cell-specific transcription factors. In the insulin promoter, GG2-GG1/A2-C1 (bases - 149 to - 116 in the human insulin promoter) play important roles in regulating β-cell-specific expression of the insulin gene. However, these events were identified through in vitro studies, and we are unaware of comparable in vivo studies. In this study, we evaluated the activity of GG2-GG1/A2 elements in the insulin promoter region in vivo. We generated homozygous mice with mutations in the GG2-GG1/A2 elements in each of the Ins1 and Ins2 promoters by CRISPR-Cas9 technology. The mice with homozygous mutations in the GG2-GG1/A2 elements in both Ins1 and Ins2 were diabetic. These data suggest that the GG2-GG1/A2 element in mice is important for Ins transcription in vivo.

Repression of insulin gene transcription by indirect genomic signaling via the estrogen receptor in pancreatic beta cells

The mechanism whereby 17β-estradiol (E2) mediates insulin gene transcription has not been fully elucidated. In this study, exposure of hamster insulinoma (HIT-T15) cells to 5 × 10-9 to 1 × 10-7 M E2 led to a concentration-dependent decrease of insulin mRNA levels. Transient expression of the estrogen receptor (ER) in HIT-T15 cells revealed that estrogen receptor α (ERα) repressed transcription of the rat insulin II promoter in both ligand-dependent and ligand-independent manners. The N-terminal A/B domain of ERα was not required for either activity. However, the repression was absent with mutated ER lacking the DNA-binding domain. Moreover, introducing mutations in the D-box and P-box of the zinc finger of ER (C227S, C202L) also abolished the repression. Deletion of the insulin promoter region revealed that nucleotide positions - 238 to - 144 (relative to the transcriptional start site) were needed for ER repression of the rat insulin II gene. PDX1- and BETA2-binding sites were required for the repression, but an estrogen response element-like sequence or an AP1 site in the promoter was not involved. In conclusion, we found that estrogen repressed insulin mRNA expression in a beta cell line. In addition, the ER suppressed insulin gene transcription in a ligand-independent matter. These observations suggest ER may regulate insulin transcription by indirect genomic signaling.

A novel function of Onecut1 protein as a negative regulator of MafA gene expression

The transcription factor MafA is a key regulator of insulin gene expression and maturation of islet β cells. Despite its importance, the regulatory mechanism of MafA gene expression is still unclear. To identify the transcriptional regulators of MafA, we examined various transcription factors, which are potentially involved in β cell differentiation. An adenovirus-mediated overexpression study clearly demonstrated that Onecut1 suppresses the promoter activity of MafA through the Foxa2-binding cis-element on the MafA enhancer region (named area A). However, ChIP analysis showed that Foxa2 but not Onecut1 could directly bind to area A. Furthermore, overexpression of Onecut1 inhibited the binding of Foxa2 onto area A upon ChIP analysis. Importantly, insertion of a mutation in the Foxa2-binding site of area A significantly decreased the promoter activity of MafA. These findings suggest that Onecut1 suppresses MafA gene expression through the Foxa2-binding site. In the mouse pancreas, MafA expression was first detected at the latest stage of β cell differentiation and was scarcely observed in Onecut1-positive cells during pancreas development. In addition, Onecut1 expression was significantly increased in the islets of diabetic db/db mice, whereas MafA expression was markedly decreased. The improved glucose levels of db/db mice with insulin injections significantly reduced Onecut1 expression and rescued the reduction of MafA expression. These in vivo experiments also suggest that Onecut1 is a negative regulator of MafA gene expression. This study implicates the novel role of Onecut1 in the control of normal β cell differentiation and its involvement in β cell dysfunction under diabetic conditions by suppressing MafA gene expression.

Binding of activating transcription factor 6 to the A5/Core of the rat insulin II gene promoter does not mediate its transcriptional repression

Pancreatic β-cells have a well-developed endoplasmic reticulum due to their highly specialized secretory function to produce insulin in response to glucose and nutrients. It has been previously reported that overexpression of activating transcription factor 6 (ATF6) reduces insulin gene expression in part via upregulation of small heterodimer partner. In this study, we investigated whether ATF6 directly binds to the insulin gene promoter, and whether its direct binding represses insulin gene promoter activity. A bioinformatics analysis identified a putative ATF6 binding site in the A5/Core region of the rat insulin II gene promoter. Direct binding of ATF6 was confirmed using several approaches. Electrophoretic mobility shift assays in nuclear extracts from MCF7 cells, isolated rat islets and insulin-secreting HIT-T15 cells showed ATF6 binding to the native A5/Core of the rat insulin II gene promoter. Antibody-mediated supershift analyses revealed the presence of both ATF6 isoforms, ATF6α and ATF6β, in the complex. Chromatin immunoprecipitation assays confirmed the binding of ATF6α and ATF6β to a region encompassing the A5/Core of the rat insulin II gene promoter in isolated rat islets. Overexpression of the active (cleaved) fragment of ATF6α, but not ATF6β, inhibited the activity of an insulin promoter-reporter by 50%. However, the inhibitory effect of ATF6α was insensitive to mutational inactivation or deletion of the A5/Core. Therefore, although ATF6 binds directly to the A5/Core of the rat insulin II gene promoter, this direct binding does not appear to contribute to its repressive activity.

Pancreatic beta cells require NeuroD to achieve and maintain functional maturity

NeuroD, a transactivator of the insulin gene, is critical for development of the endocrine pancreas, and NeuroD mutations cause MODY6 in humans. To investigate the role of NeuroD in differentiated beta cells, we generated mice in which neuroD is deleted in insulin-expressing cells. These mice exhibit severe glucose intolerance. Islets lacking NeuroD respond poorly to glucose and display a glucose metabolic profile similar to immature beta cells, featuring increased expression of glycolytic genes and LDHA, elevated basal insulin secretion and O2 consumption, and overexpression of NPY. Moreover, the mutant islets appear to have defective K(ATP) channel-mediated insulin secretion. Unexpectedly, virtually all insulin in the mutant mice is derived from ins2, whereas ins1 expression is almost extinguished. Overall, these results indicate that NeuroD is required for beta cell maturation and demonstrate the importance of NeuroD in the acquisition and maintenance of fully functional glucose-responsive beta cells.

Prolactin regulatory element binding protein as a potential transcriptional factor for the insulin gene in response to glucose stimulation

AIMS/HYPOTHESIS

Prolactin regulatory element binding (PREB) protein has been identified as a factor that regulates prolactin promoter activity in rat anterior pituitary. PREB is located not only in the anterior pituitary but also in pancreas; however its role in the pancreas is not known. We therefore examined the role of PREB in insulin gene expression.

MATERIALS AND METHODS

To analyse the effects of PREB on insulin gene transcription, we employed the luciferase reporter gene assay and electrophoretic mobility shift assay (EMSA). In cells expressing or knocked down for PREB, insulin expression and secretion were determined.

RESULTS

PREB was located mainly in nuclei of rat pancreatic beta cells and its cell line, INS-1. A nuclear extract of INS-1 cells contained material that was recognised by PREB antiserum. This nuclear extract also showed insulin promoter binding activity that was super-shifted by PREB antiserum in EMSA studies. In the INS-1 cells, co-expression of PREB and the insulin promoter induced activity of the latter. The addition of glucose to the cells increased PREB expression. Deletional analysis of the insulin promoter showed that A3, a glucose-responsive cis-element in the insulin promoter, mediated the transcriptional effect of PREB. In addition, synthesised PREB bound the A3 element by EMSA, while a mutant of this motif in the insulin promoter abrogated the effect of PREB. Cells expressing or knocked down for PREB exhibited increased or decreased insulin expression, respectively.

CONCLUSIONS/INTERPRETATION

These results demonstrate that PREB may contribute to the regulation of insulin gene transcription and insulin secretion in response to glucose stimulation.

A switch from MafB to MafA expression accompanies differentiation to pancreatic beta-cells

Major insulin gene transcription factors, such as PDX-1 or NeuroD1, have equally important roles in pancreatic development and the differentiation of pancreatic endocrine cells. Previously, we identified and cloned another critical insulin gene transcription factor MafA (RIPE3b1) and reported that other Maf factors were expressed in pancreatic endocrine cells. Maf factors are important regulators of cellular differentiation; to understand their role in differentiation of pancreatic endocrine cells, we analyzed the expression pattern of large-Maf factors in the pancreas of embryonic and adult mice. Ectopically expressed large-Maf factors, MafA, MafB, or cMaf, induced expression from insulin and glucagon reporter constructs, demonstrating a redundancy in their function. Yet in adult pancreas, cMaf was expressed in both alpha- and beta-cells, and MafA and MafB showed selective expression in the beta- and alpha-cells, respectively. Interestingly, during embryonic development, a significant proportion of MafB-expressing cells also expressed insulin. In embryos, MafB is expressed before MafA, and our results suggest that the differentiation of beta-cells proceeds through a MafB+ MafA- Ins+ intermediate cell to MafB- MafA+ Ins+ cells. Furthermore, the MafB to MafA transition follows induction of PDX-1 expression (Pdx-1(high)) in MafB+ Ins+ cells. We suggest that MafB may have a dual role in regulating embryonic differentiation of both beta- and alpha-cells while MafA may regulate replication/survival and function of beta-cells after birth. Thus, this redundancy in the function and expression of the large-Maf factors may explain the normal islet morphology observed in the MafA knockout mice at birth.

Regulation of insulin gene expression by overlapping DNA-binding elements

The transcription factor MafA/RIPE3b1 is an important regulator of insulin gene expression. MafA binds to the insulin enhancer element RIPE3b (C1-A2), now designated as insulin MARE (Maf response element). The insulin MARE element shares an overlapping DNA-binding region with another insulin enhancer element A2. A2.2, a beta-cell-specific activator, like the MARE-binding factor MafA, binds to the overlapping A2 element. Our previous results demonstrated that two nucleotides in the overlapping region are required for the binding of both factors. Surprisingly, instead of interfering with each other's binding activity, the MafA and the A2-binding factors co-operatively activated insulin gene expression. To understand the molecular mechanisms responsible for this functional co-operation, we have determined the nucleotides essential for the binding of the A2.2 factor. Using this information, we have constructed non-overlapping DNA-binding elements and their derivatives, and subsequently analysed the effect of these modifications on insulin gene expression. Our results demonstrate that the overlapping binding site is essential for maximal insulin gene expression. Furthermore, the overlapping organization is critical for MafA-mediated transcriptional activation, but has a minor effect on the activity of A2-binding factors. Interestingly, the binding affinities of both MafA and A2.2 to the overlapping or non-overlapping binding sites were not significantly different, implying that the overlapping binding organization may increase the activation potential of MafA by physical/functional interactions with A2-binding factors. Thus our results demonstrate a novel mechanism for the regulation of MafA activity, and in turn beta-cell function, by altering expression and/or binding of the A2.2 factor. Our results further suggest that the major downstream targets of MafA will in addition to the MARE element have a binding site for the A2.2 factor.

Cell-specific cytotoxicity of human pancreatic adenocarcinoma cells using rat insulin promoter thymidine kinase-directed gene therapy

The formation of a normal pancreas and the activation of insulin production are, in part, dependent on the expression and activation of the pancreatic duodenal homeobox gene 1 (PDX-1). The expression of PDX-1 also has been detected in various human pancreatic ductal adenocarcinoma (PDA) cell lines. This has made it possible to generate a cancer cell-specific gene expression system to treat human pancreatic cancer. In this study, we have developed a cell-specific cytotoxic model of PDA cells using the expression of herpes simplex virus thymidine kinase (TK) under the control of the rat insulin promoter (RIP-TK). We have shown that the cell-specific cytotoxicity in human PDA cells depends on the presence of PDX-1. Our results also demonstrate that in vivo PDA-specific cytotoxicity can be achieved with RIP-TK using an intraperitoneal liposomal gene delivery method followed by a short period of ganciclovir treatment in severe combined immunodeficient (SCID) mice. Furthermore, PDX-1 protein was found in all six freshly isolated human pancreas cancer specimens and two liver metastasis samples that were group-tested, suggesting the feasibility of using RIP-TK gene therapy in humans. This study may provide an alternative strategy for the future treatment of pancreatic cancer.

Specific gene expression and therapy for pancreatic cancer using the cytosine deaminase gene directed by the rat insulin promoter

Suicide gene therapy has been shown to be an effective means of destroying pancreatic cancer cells, but cell-specific delivery of the gene is required to limit host toxicity. The objective of this study is to determine whether the rat insulin promoter (RIP) will permit cell-specific gene delivery and subsequent cell death in human pancreatic cancer cells. The RIP DNA was amplified using polymerase chain reaction (PCR), and the purified fragment was inserted into pCR-Blunt II-TOPO plasmid at the SpeI site, which contains the coding sequence of yeast cytosine deaminase (CD). Transfection assays were carried out using both RIP-lacZ and RIP-CD DNA constructs in two human pancreatic cancer cell lines, PANC-1 and MIA PaCa-2. Reporter assays using X-gal staining were performed, and the in vitro cytotoxicity was examined in RIP-CD-transfected cells treated with 5-flucytosine for 5 days. The expression levels of CD protein in the transfected cells were determined 2 days after transfection by Western blot analysis. The expression levels of insulin promoter factor (IPF-1/PDX-1) in these human pancreatic cell lines, as well as in freshly isolated human pancreatic cancer specimens, were determined using in situ immunohistochemistry analysis. After transfection with RIP-lacZ, only PANC-1 cells, but not MIA PaCa-2 cells, were positive for RIP-lacZ expression, indicating that RIP-directed reporter gene expression occurred only in PANC-1 cells. After transfection with RIP-CD and treatment with 5-flucytosine, PANC-1 cells had a significantly increased cell death rate compared with that of MIA PaCa-2 cells, suggesting that RIP-directed suicide gene expression occurred only in PANC-1 cells. Western blot analysis demonstrated that only PANC-1 cells were able to express the CD protein and that significantly increased levels of PDX-1 were found in PANC-1 but not in Mia PaCa-2 cells. In situ immunohistochemical analysis of both cell lines showed that PDX-1 was only expressed in the nuclei of PANC-1 cells and not in MIA PaCa-2 cells. Furthermore, two freshly isolated human pancreatic cancer specimens had significantly increased levels of PDX-1. The RIP is activated in PANC-1 cells, but not in Mia PaCa-2 cells, and the mechanism of activation is via PDX-1. Pancreatic cancer-specific cytotoxicity can be achieved with the use of RIP-CD and 5-flucytosine treatment in vitro. Significantly increased levels of PDX-1 have been found in human pancreatic cancer specimens. These results suggest that RIP could be used for cell-specific suicide gene therapy to target human pancreatic tumors.

Identification of beta-cell-specific insulin gene transcription factor RIPE3b1 as mammalian MafA

Of the three critical enhancer elements that mediate beta-cell-specific and glucose-responsive expression of the insulin gene, only the identity of the transcription factor binding to the RIPE3b element (RIPE3b1) has remained elusive. Using a biochemical purification approach, we have identified the RIPE3b1 factor as a mammalian homologue of avian MafA/L-Maf (mMafA). The avian MafA is a cell-type determination factor that expressed ectopically can trigger lens differentiation program, but no mammalian homologue of avian MafA has previously been identified. Here, we report cloning of the human mafA (hMafA) and demonstrate that it can specifically bind the insulin enhancer element RIPE3b and activate insulin-gene expression. In addition, mMafA has a very restrictive cellular distribution and is selectively expressed in pancreatic beta but not in alpha cells. We suggest that mMafA has an essential role in the function and differentiation of beta-cells and thus may be associated with the pathophysiological origins of diabetes.

Insulin gene transcription is mediated by interactions between the p300 coactivator and PDX-1, BETA2, and E47

Pancreatic beta-cell-type-specific expression of the insulin gene requires both ubiquitous and cell-enriched activators, which are organized within the enhancer region into a network of protein-protein and protein-DNA interactions to promote transcriptional synergy. Protein-protein-mediated communication between DNA-bound activators and the RNA polymerase II transcriptional machinery is inhibited by the adenovirus E1A protein as a result of E1A's binding to the p300 coactivator. E1A disrupts signaling between the non-DNA-binding p300 protein and the basic helix-loop-helix DNA-binding factors of insulin's E-element activator (i.e., the islet-enriched BETA2 and generally distributed E47 proteins), as well as a distinct but unidentified enhancer factor. In the present report, we show that E1A binding to p300 prevents activation by insulin's beta-cell-enriched PDX-1 activator. p300 interacts directly with the N-terminal region of the PDX-1 homeodomain protein, which contains conserved amino acid sequences essential for activation. The unique combination of PDX-1, BETA2, E47, and p300 was shown to promote synergistic activation from a transfected insulin enhancer-driven reporter construct in non-beta cells, a process inhibited by E1A. In addition, E1A inhibited the level of PDX-1 and BETA2 complex formation in beta cells. These results indicate that E1A inhibits insulin gene transcription by preventing communication between the p300 coactivator and key DNA-bound activators, like PDX-1 and BETA2:E47.

The NeuroD1/BETA2 sequences essential for insulin gene transcription colocalize with those necessary for neurogenesis and p300/CREB binding protein binding

NeuroD1/BETA2 is a key regulator of pancreatic islet morphogenesis and insulin hormone gene transcription in islet beta cells. This factor also appears to be involved in neurogenic differentiation, because NeuroD1/BETA2 is able to induce premature differentiation of neuronal precursors and convert ectoderm into fully differentiated neurons upon ectopic expression in Xenopus embryos. We have identified amino acid sequences in mammalian and Xenopus NeuroD1/BETA2 that are necessary for insulin gene expression and ectopic neurogenesis. Our results indicate that evolutionarily conserved sequences spanning the basic helix-loop-helix (amino acids [aa] 100 to 155) and C-terminal (aa 156 to 355) regions are important for both of these processes. The transactivation domains (AD1, aa 189 to 299; AD2, aa 300 to 355) were within the carboxy-terminal region, as analyzed by using GAL4:NeuroD1/BETA2 chimeras. Selective activation of mammalian insulin gene enhancer-driven expression and ectopic neurogenesis in Xenopus embryos was regulated by two independent and separable domains of NeuroD1/BETA2, located between aa 156 to 251 and aa 252 to 355. GAL4:NeuroD1/BETA2 constructs spanning these sequences demonstrated that only aa 252 to 355 contained activation domain function, although both aa 156 to 251 and 300 to 355 were found to interact with the p300/CREB binding protein (CBP) coactivator. These results implicate p300/CBP in NeuroD1/BETA2 function and further suggest that comparable mechanisms are utilized to direct target gene transcription during differentiation and in adult islet beta cells.

A novel glucose-responsive element in the human insulin gene functions uniquely in primary cultured islets

Insulin gene transcription is limited to the beta cells within the mammalian pancreas and, like insulin secretion, is regulated by glucose. Our previous studies in primary cultured beta cells suggested the presence of a strong glucose-responsive enhancer element between base pairs -341 and -260 of the human insulin promoter, the same region in which a transcriptional repressor had been identified in beta-cell tumor lines. In an attempt to map these promoter activities and resolve these conflicting data, we designed minienhancer constructs spanning this region, and tested them in primary cultured and immortalized cells. One sequence, the Z element (base pairs -292 to -243), functions as both a potent glucose-responsive transcriptional enhancer in primary cultured islet cells and as a transcriptional repressor in immortalized beta and nonbeta cells and in primary fibroblasts. In addition, the Z element binds a novel glucose-responsive protein complex that is found in the nuclei of primary cultured islet cells, but not in the nuclei of tumor cells or primary cultured fibroblasts. These data demonstrate a critical role for the Z element in human insulin gene transcription and its regulation by glucose.

p300 mediates transcriptional stimulation by the basic helix-loop-helix activators of the insulin gene

Pancreatic beta-cell-type-specific and glucose-inducible transcription of the insulin gene is mediated by the basic helix-loop-helix factors that bind to and activate expression from an E-box element within its enhancer. The E-box activator is a heteromeric complex composed of a beta-cell-enriched factor, BETA2/NeuroD, and ubiquitously distributed proteins encoded by the E2A and HEB genes. Previously, we demonstrated that the adenovirus type 5 E1A proteins repressed stimulation by the E-box activator in beta cells. In this study, our objective was to determine how E1A repressed activator function. The results indicate that E1A reduces activation by binding to and sequestering the p300 cellular coactivator protein. Thus, we show that expression of p300 in beta cells can relieve inhibition by E1A, as well as potentiate activation by the endogenous insulin E-box transcription factors. p300 stimulated activation from GAL4 (amino acids 1 to 147) fusion constructs of either BETA2/NeuroD or the E2A-encoded E47 protein. The sequences spanning the activation domains of BETA2/NeuroD (amino acids 156 to 355) and E47 (amino acids 1 to 99 and 325 to 432) were required for this response. The same region of BETA2/NeuroD was shown to be important for binding to p300 in vitro. The sequences of p300 involved in E47 and BETA2/NeuroD association resided between amino acids 1 and 1257 and 1945 and 2377, respectively. A mutation in p300 that abolished binding to BETA2/NeuroD also destroyed the ability of p300 to activate insulin E-box-directed transcription in beta cells. Our results indicate that physical and functional interactions between p300 and the E-box activator factors play an important role in insulin gene transcription.

Functional characterization of the transactivation properties of the PDX-1 homeodomain protein

Pancreas formation is prevented in mice carrying a null mutation in the PDX-1 homeoprotein, demonstrating a key role for this factor in development. PDX-1 can also bind to and activate transcription from cis-acting regulatory sequences in the insulin and somatostatin genes, which are expressed in pancreatic islet beta and delta cells, respectively. In this study, we compared the functional properties of PDX-1 with those of the closely related Xenopus homeoprotein XIHbox8. Analysis of chimeras between PDX-1, XIHbox8, and the DNA-binding domain of the Saccharomyces cerevisiae transcription factor GAL4 revealed that their transactivation domain was contained within the N-terminal region (amino acids 1 to 79). Detailed mutagenesis of this region indicated that transactivation is mediated by three highly conserved sequences, spanning amino acids 13 to 22 (subdomain A), 32 to 38 (subdomain B), and 60 to 73 (subdomain C). These sequences were also required by PDX-1 to synergistically activate insulin enhancer-mediated transcription with another key insulin gene activator, the E2A-encoded basic helix-loop-helix E2-5 and E47 proteins. These results indicated that N-terminal sequences conserved between the mammalian PDX-1 and Xenopus XIHbox8 proteins are important in transcriptional activation. Stable expression of the PDX-1 deltaABC mutant in the insulin- and PDX-1-expressing betaTC3 cell line resulted in a threefold reduction in the rate of endogenous insulin gene transcription. Strikingly, the level of the endogenous PDX-1 protein was reduced to very low levels in these cells. These results suggest that PDX-1 is not absolutely essential for insulin gene expression in betaTC3 cells. We discuss the possible significance of these findings for insulin gene transcription in islet beta cells.

Induction of insulin and islet amyloid polypeptide production in pancreatic islet glucagonoma cells by insulin promoter factor 1

Insulin promoter factor 1 (IPF1), a member of the homeodomain protein family, serves an early role in pancreas formation, as evidenced by the lack of pancreas formation in mice carrying a targeted disruption of the IPF1 gene [Jonsson, J., Carlsson, L., Edlund, T. & Edlund, H. (1994) Nature (London) 371, 606-609]. In adults, IPF1 expression is restricted to the beta-cells in the islets of Langerhans. We report here that IPF1 induces expression of a subset of beta-cell-specific genes (insulin and islet amyloid polypeptide) when ectopically expressed in clones of transformed pancreatic islet alpha-cells. In contrast, expression of IPF1 in rat embryo fibroblasts factor failed to induce insulin and islet amyloid polypeptide expression. This is most likely due to the lack of at least one other essential insulin gene transcription factor, the basic helix-loop-helix protein Beta 2/NeuroD, which is expressed in both alpha- and beta-cells. We conclude that IPF1 is a potent transcriptional activator of endogenous insulin genes in non-beta islet cells, which suggests an important role of IPF1 in beta-cell maturation.

BETA3, a novel helix-loop-helix protein, can act as a negative regulator of BETA2 and MyoD-responsive genes

Using degenerate PCR cloning we have identified a novel basic helix-loop-helix (bHLH) transcription factor, BETA3, from a hamster insulin tumor (HIT) cell cDNA library. Sequence analysis revealed that this factor belongs to the class B bHLH family and has the highest degree of homology with another bHLH transcription factor recently isolated in our laboratory, BETA2 (neuroD) (J. E. Lee, S. M. Hollenberg, L. Snider, D. L. Turner, N. Lipnick, and H. Weintraub, Science 268:836-844, 1995; F. J. Naya, C. M. M. Stellrecht, and M.-J. Tsai, Genes Dev. 8:1009-1019, 1995). BETA2 is a brain- and pancreatic-islet-specific bHLH transcription factor and is largely responsible for the tissue-specific expression of the insulin gene. BETA3 was found to be tissue restricted, with the highest levels of expression in HIT, lung, kidney, and brain cells. Surprisingly, despite the homology between BETA2 and BETA3 and its intact basic region, BETA3 is unable to bind the insulin E box in bandshift analysis as a homodimer or as a heterodimer with the class A bHLH factors E12, E47, or BETA1. Instead, BETA3 inhibited both the E47 homodimer and the E47/BETA2 heterodimer binding to the insulin E box. In addition, BETA3 greatly repressed the BETA2/E47 induction of the insulin enhancer in HIT cells as well as the MyoD/E47 induction of a muscle-specific E box in the myoblast cell line C2C12. In contrast, expression of BETA3 had no significant effect on the GAL4-VP16 transcriptional activity. Immunoprecipitation analysis demonstrates that the mechanism of repression is via direct protein-protein interaction, presumably by heterodimerization between BETA3 and class A bHLH factors.

c-jun inhibits insulin control element-mediated transcription by affecting the transactivation potential of the E2A gene products

Pancreatic beta-cell-type-specific transcription of the insulin gene is principally controlled by trans-acting factors which influence insulin control element (ICE)-mediated expression. The ICE activator is composed, in part, of the basic helix-loop-helix proteins E12, E47, and E2-5 encoded by the E2A gene. Previous experiments showed that ICE activation in beta cells was repressed in vivo by the c-jun proto-oncogene (E. Henderson and R. Stein, Mol. Cell. Biol. 14:655-662, 1994). Here we focus on the mechanism by which c-Jun inhibits ICE-mediated activation. c-Jun was shown to specifically repress the transactivation potential of the E2A proteins. Thus, we found that the activity of GAL4:E2A fusion constructs was inhibited by c-Jun. The transrepression capabilities of c-Jun were detected only in pancreatic islet cell lines that contained a functional ICE activator. Repression of GAL4:E2A was mediated by the basic leucine zipper regions of c-Jun, which are also the essential regions of this protein necessary for controlling ICE activator-stimulated expression in vivo. The specific target of c-Jun repression was the transactivation domain (located between amino acids 345 and 408 in E12 and E47) conserved in E12, E47, and E2-5. In contrast, the activation domain unique to the E12 and E47 proteins (located between amino acids 1 and 99) was unresponsive to c-Jun. Our results indicate that c-Jun inhibits insulin gene transcription in beta cells by reducing the transactivation potential of the E2A proteins present in the ICE activator complex.

Pancreatic beta cells express a diverse set of homeobox genes

Homeobox genes, which are found in all eukaryotic organisms, encode transcriptional regulators involved in cell-type differentiation and development. Several homeobox genes encoding homeodomain proteins that bind and activate the insulin gene promoter have been described. In an attempt to identify additional beta-cell homeodomain proteins, we designed primers based on the sequences of beta-cell homeobox genes cdx3 and lmx1 and the Drosophila homeodomain protein Antennapedia and used these primers to amplify inserts by PCR from an insulinoma cDNA library. The resulting amplification products include sequences encoding 10 distinct homeodomain proteins; 3 of these proteins have not been described previously. In addition, an insert was obtained encoding a splice variant of engrailed-2, a homeodomain protein previously identified in the central nervous system. Northern analysis revealed a distinct pattern of expression for each homeobox gene. Interestingly, the PCR-derived clones do not represent a complete sampling of the beta-cell library because no inserts encoding cdx3 or lmx1 protein were obtained. Beta cells probably express additional homeobox genes. The abundance and diversity of homeodomain proteins found in beta cells illustrate the remarkable complexity and redundancy of the machinery controlling beta-cell development and differentiation.

Isolation and characterization of a novel transcription factor that binds to and activates insulin control element-mediated expression

Pancreatic beta-cell-type-specific transcription of the insulin gene is principally regulated by a single cis-acting DNA sequence element, termed the insulin control element (ICE), which is found within the 5'-flanking region of the gene. The ICE activator is a heteromeric complex composed of an islet alpha/beta-cell-specific factor associated with the ubiquitously distributed E2A-encoded proteins (E12, E47, and E2-5). We describe the isolation and characterization of a cDNA for a protein present in alpha and beta cells, termed INSAF for insulin activator factor, which binds to and activates ICE-mediated expression. INSAF was isolated from a human insulinoma cDNA library. Transfection experiments demonstrated that INSAF activates ICE expression in insulin-expressing cells but not in non-insulin-expressing cells. Cotransfection experiments showed that activation by INSAF was inhibited by Id, a negative regulator of basic helix-loop-helix (bHLH) protein function. INSAF was also shown to associate in vitro with the bHLH protein E12. In addition, affinity-purified INSAF antiserum abolished the formation of the activator-specific ICE-binding complex. Immunohistochemical studies indicate that INSAF is restricted in terms of its expression pattern, in that INSAF appears to be detected only within the nuclei of islet pancreatic alpha and beta cells. All of these data are consistent with the proposal that INSAF is either part of the ICE activator or is antigenically related to the specific activator required for insulin gene transcription.

Isolation and characterization of a novel transcription factor that binds to and activates insulin control element-mediated expression

Pancreatic beta-cell-type-specific transcription of the insulin gene is principally regulated by a single cis-acting DNA sequence element, termed the insulin control element (ICE), which is found within the 5'-flanking region of the gene. The ICE activator is a heteromeric complex composed of an islet alpha/beta-cell-specific factor associated with the ubiquitously distributed E2A-encoded proteins (E12, E47, and E2-5). We describe the isolation and characterization of a cDNA for a protein present in alpha and beta cells, termed INSAF for insulin activator factor, which binds to and activates ICE-mediated expression. INSAF was isolated from a human insulinoma cDNA library. Transfection experiments demonstrated that INSAF activates ICE expression in insulin-expressing cells but not in non-insulin-expressing cells. Cotransfection experiments showed that activation by INSAF was inhibited by Id, a negative regulator of basic helix-loop-helix (bHLH) protein function. INSAF was also shown to associate in vitro with the bHLH protein E12. In addition, affinity-purified INSAF antiserum abolished the formation of the activator-specific ICE-binding complex. Immunohistochemical studies indicate that INSAF is restricted in terms of its expression pattern, in that INSAF appears to be detected only within the nuclei of islet pancreatic alpha and beta cells. All of these data are consistent with the proposal that INSAF is either part of the ICE activator or is antigenically related to the specific activator required for insulin gene transcription.

Transcription of the insulin gene: towards defining the glucose-sensitive cis-element and trans-acting factors

Previous work has shown that the sequence -196 to -247 of the rat insulin I gene mediates the stimulatory effect of glucose in fetal islets. We have used adult rat and human islets to delineate the glucose-sensitive cis-element to the sequence -193 to -227. In electrophoretic mobility shift assays, a 22 bp nucleotide corresponding to the sequence -206 to -227 bound all the nuclear proteins that could be bound by the entire minienhancer sequence -196 to -247. The rat insulin I sequence -206 to -227 formed three major complexes; in contrast, the corresponding human insulin sequence formed one single band with human and rat islet nuclear extracts, corresponding to the complex C1 of the rat insulin gene. Incubation of islets with varying glucose levels resulted in a dose-dependent increase in the intensity of the C1 band, while the other nuclear complexes formed with the insulin sequence, or the AP1 and SP1 binding activities used as control, were glucose insensitive. This is thus the first demonstration of a physiologic glucose-sensitive trans-acting factor for the insulin gene, whose further study may markedly enhance our understanding of the regulation of insulin biosynthesis in normal and diabetic beta cells. Furthermore, once cloned, the introduction of this glucose sensitive factor may enable the construction of truly physiologic artificial beta cells.

The insulin gene contains multiple transcriptional elements that respond to glucose

The beta cells in the pancreatic islets of Langerhans increase insulin gene transcription in response to increased glucose concentration. We have mapped sequences within the rat insulin I gene 5'-flanking DNA (rInsI promoter) that direct this transcriptional response to glucose. When linked to chloramphenicol acetyltransferase and expressed in cultured beta cells, no single mutation of the rInsI promoter removes its ability to respond to glucose, although several mutations cause marked reductions in basal chloramphenicol acetyltransferase expression. A 50-bp sequence isolated from the rInsI promoter, the Far-FLAT minienhancer, can confer glucose responsiveness to nonresponsive promoters. Fine mapping of this minienhancer further localizes a glucose response to the sequence GGCCATCTGGCC, or the Far element. Nuclear extracts from islets grown in various glucose concentrations demonstrate a glucose-stimulated increase in a protein complex that binds the Far element and contains the transcription factors Pan-1 and Pan-2. Overexpression of intact or partially deleted Pan-1 ablates the Far-directed transcriptional response to glucose. We conclude that the full glucose response of the insulin promoter involves the interaction of multiple sequence elements. Part of this response, however, results from activation of a complex binding at the Far element.

The insulin gene contains multiple transcriptional elements that respond to glucose

The beta cells in the pancreatic islets of Langerhans increase insulin gene transcription in response to increased glucose concentration. We have mapped sequences within the rat insulin I gene 5'-flanking DNA (rInsI promoter) that direct this transcriptional response to glucose. When linked to chloramphenicol acetyltransferase and expressed in cultured beta cells, no single mutation of the rInsI promoter removes its ability to respond to glucose, although several mutations cause marked reductions in basal chloramphenicol acetyltransferase expression. A 50-bp sequence isolated from the rInsI promoter, the Far-FLAT minienhancer, can confer glucose responsiveness to nonresponsive promoters. Fine mapping of this minienhancer further localizes a glucose response to the sequence GGCCATCTGGCC, or the Far element. Nuclear extracts from islets grown in various glucose concentrations demonstrate a glucose-stimulated increase in a protein complex that binds the Far element and contains the transcription factors Pan-1 and Pan-2. Overexpression of intact or partially deleted Pan-1 ablates the Far-directed transcriptional response to glucose. We conclude that the full glucose response of the insulin promoter involves the interaction of multiple sequence elements. Part of this response, however, results from activation of a complex binding at the Far element.

Glucose-induced transcription of the insulin gene is mediated by factors required for beta-cell-type-specific expression

The insulin gene is expressed exclusively in pancreatic islet beta cells. The principal regulator of insulin gene transcription in the islet is the concentration of circulating glucose. Previous studies have demonstrated that transcription is regulated by the binding of trans-acting factors to specific cis-acting sequences within the 5'-flanking region of the insulin gene. To identify the cis-acting control elements within the rat insulin II gene that are responsible for regulating glucose-stimulated expression in the beta cell, we analyzed the effect of glucose on the in vivo expression of a series of transfected 5'-flanking deletion mutant constructs. We demonstrate that glucose-induced transcription of the rat insulin II gene is mediated by sequences located between -126 and -91 bp relative to the transcription start site. This region contains two cis-acting elements that are essential for directing pancreatic beta-cell-type-specific expression of the rat insulin II gene, the insulin control element (ICE; -100 to -91 bp) and RIPE3b1 (-115 to -107 bp). The gel mobility shift assay was used to determine whether the formation of the ICE- and RIPE3b1-specific factor-DNA element complexes were affected in glucose-treated beta-cell extracts. We found that RIPE3b1 binding activity was selectively induced by about eightfold. In contrast, binding to other insulin cis-acting element sequences like the ICE and RIPE3a2 (-108 to -99 bp) were unaffected by these conditions. The RIPE3b1 binding complex was shown to be distinct from the glucose-inducible factor that binds to an element located between -227 to -206 bp of the human and rat insulin I genes (D. Melloul, Y. Ben-Neriah, and E. Cerasi, Proc. Natl. Acad. Sci. USA 90:3865-3869, 1993). We have also shown that mannose, a sugar that can be metabolized by the beta cell, mimics the effects of glucose in the in vivo transfection assays and the in vitro RIPE3b1 binding assays. These results suggested that the RIPE3b1 transcription factor is a primary regulator of glucose-mediated transcription of the insulin gene. However, we found that mutations in either the ICE or the RIPE3b1 element reduced glucose-responsive expression from transfected 5'-flanking rat insulin II gene constructs. We therefore conclude that glucose-regulated transcription of the insulin gene is mediated by cis-acting elements required for beta-cell-type-specific expression.

Glucose-induced transcription of the insulin gene is mediated by factors required for beta-cell-type-specific expression

The insulin gene is expressed exclusively in pancreatic islet beta cells. The principal regulator of insulin gene transcription in the islet is the concentration of circulating glucose. Previous studies have demonstrated that transcription is regulated by the binding of trans-acting factors to specific cis-acting sequences within the 5'-flanking region of the insulin gene. To identify the cis-acting control elements within the rat insulin II gene that are responsible for regulating glucose-stimulated expression in the beta cell, we analyzed the effect of glucose on the in vivo expression of a series of transfected 5'-flanking deletion mutant constructs. We demonstrate that glucose-induced transcription of the rat insulin II gene is mediated by sequences located between -126 and -91 bp relative to the transcription start site. This region contains two cis-acting elements that are essential for directing pancreatic beta-cell-type-specific expression of the rat insulin II gene, the insulin control element (ICE; -100 to -91 bp) and RIPE3b1 (-115 to -107 bp). The gel mobility shift assay was used to determine whether the formation of the ICE- and RIPE3b1-specific factor-DNA element complexes were affected in glucose-treated beta-cell extracts. We found that RIPE3b1 binding activity was selectively induced by about eightfold. In contrast, binding to other insulin cis-acting element sequences like the ICE and RIPE3a2 (-108 to -99 bp) were unaffected by these conditions. The RIPE3b1 binding complex was shown to be distinct from the glucose-inducible factor that binds to an element located between -227 to -206 bp of the human and rat insulin I genes (D. Melloul, Y. Ben-Neriah, and E. Cerasi, Proc. Natl. Acad. Sci. USA 90:3865-3869, 1993). We have also shown that mannose, a sugar that can be metabolized by the beta cell, mimics the effects of glucose in the in vivo transfection assays and the in vitro RIPE3b1 binding assays. These results suggested that the RIPE3b1 transcription factor is a primary regulator of glucose-mediated transcription of the insulin gene. However, we found that mutations in either the ICE or the RIPE3b1 element reduced glucose-responsive expression from transfected 5'-flanking rat insulin II gene constructs. We therefore conclude that glucose-regulated transcription of the insulin gene is mediated by cis-acting elements required for beta-cell-type-specific expression.

c-jun inhibits transcriptional activation by the insulin enhancer, and the insulin control element is the target of control

Selective transcription of the insulin gene in pancreatic beta cells is regulated by its enhancer, located between nucleotides -340 and -91 relative to the transcription start site. One of the principal control elements within the enhancer is found between nucleotides -100 and -91 (GCCATCTGCT, referred to as the insulin control element [ICE]) and is regulated by both positive- and negative-acting transcription factors in the helix-loop-helix (HLH) family. It was previously shown that the c-jun proto-oncogene can repress insulin gene transcription. We have found that c-jun inhibits ICE-stimulated transcription. Inhibition of ICE-directed transcription is mediated by sequences within the carboxy-terminal region of the protein. These c-jun sequences span an activation domain and the basic leucine zipper DNA binding-dimerization region of the protein. Both regions of c-jun are conserved within the other members of the jun family: junB and junD. These proteins also suppress ICE-mediated transcription. The jun proteins do not appear to inhibit insulin gene transcription by binding directly to the ICE. c-jun and junB also block the trans-activation potential of two skeletal muscle-specific HLH proteins, MyoD and myogenin. These results suggests that the jun proteins may be common transcription control factors used in skeletal muscle and pancreatic beta cells to regulate HLH-mediated activity. We discuss the possible significance of these observations to insulin gene transcription in pancreatic beta cells.

c-jun inhibits transcriptional activation by the insulin enhancer, and the insulin control element is the target of control

Selective transcription of the insulin gene in pancreatic beta cells is regulated by its enhancer, located between nucleotides -340 and -91 relative to the transcription start site. One of the principal control elements within the enhancer is found between nucleotides -100 and -91 (GCCATCTGCT, referred to as the insulin control element [ICE]) and is regulated by both positive- and negative-acting transcription factors in the helix-loop-helix (HLH) family. It was previously shown that the c-jun proto-oncogene can repress insulin gene transcription. We have found that c-jun inhibits ICE-stimulated transcription. Inhibition of ICE-directed transcription is mediated by sequences within the carboxy-terminal region of the protein. These c-jun sequences span an activation domain and the basic leucine zipper DNA binding-dimerization region of the protein. Both regions of c-jun are conserved within the other members of the jun family: junB and junD. These proteins also suppress ICE-mediated transcription. The jun proteins do not appear to inhibit insulin gene transcription by binding directly to the ICE. c-jun and junB also block the trans-activation potential of two skeletal muscle-specific HLH proteins, MyoD and myogenin. These results suggests that the jun proteins may be common transcription control factors used in skeletal muscle and pancreatic beta cells to regulate HLH-mediated activity. We discuss the possible significance of these observations to insulin gene transcription in pancreatic beta cells.

E1A-mediated inhibition of myogenesis correlates with a direct physical interaction of E1A12S and basic helix-loop-helix proteins

The observation that adenovirus E1A gene products can inhibit differentiation of skeletal myocytes suggested that E1A may interfere with the activity of myogenic basic helix-loop-helix (bHLH) transcription factors. We have examined the ability of E1A to mediate repression of the muscle-specific creatine kinase (MCK) gene. Both the E1A12S and E1A13S products repressed MCK transcription in a concentration-dependent fashion. In contrast, amino-terminal deletion mutants (d2-36 and d15-35) of E1A12S were defective for repression. E1A12S also repressed expression of a promoter containing a multimer of the MCK high-affinity E box (the consensus site for myogenic bHLH protein binding) that was dependent, in C3H10T1/2 cells, on coexpression of a myogenin bHLH-VP16 fusion protein. A series of coprecipitation experiments with glutathione S-transferase fusion and in vitro-translated proteins demonstrated that E1A12S, but not amino-terminal E1A deletion mutants, could bind to full-length myogenin and E12 and to deletion mutants of myogenin and E12 that spare the bHLH domains. Thus, the bHLH domains of myogenin and E12, and the high-affinity E box, are targets for E1A-mediated repression of the MCK enhancer, and domains of E1A required for repression of muscle-specific gene transcription also mediate binding to bHLH proteins. We conclude that E1A mediates repression of muscle-specific gene transcription through its amino-terminal domain and propose that this may involve a direct physical interaction between E1A and the bHLH region of myogenic determination proteins.

E1A-mediated inhibition of myogenesis correlates with a direct physical interaction of E1A12S and basic helix-loop-helix proteins

The observation that adenovirus E1A gene products can inhibit differentiation of skeletal myocytes suggested that E1A may interfere with the activity of myogenic basic helix-loop-helix (bHLH) transcription factors. We have examined the ability of E1A to mediate repression of the muscle-specific creatine kinase (MCK) gene. Both the E1A12S and E1A13S products repressed MCK transcription in a concentration-dependent fashion. In contrast, amino-terminal deletion mutants (d2-36 and d15-35) of E1A12S were defective for repression. E1A12S also repressed expression of a promoter containing a multimer of the MCK high-affinity E box (the consensus site for myogenic bHLH protein binding) that was dependent, in C3H10T1/2 cells, on coexpression of a myogenin bHLH-VP16 fusion protein. A series of coprecipitation experiments with glutathione S-transferase fusion and in vitro-translated proteins demonstrated that E1A12S, but not amino-terminal E1A deletion mutants, could bind to full-length myogenin and E12 and to deletion mutants of myogenin and E12 that spare the bHLH domains. Thus, the bHLH domains of myogenin and E12, and the high-affinity E box, are targets for E1A-mediated repression of the MCK enhancer, and domains of E1A required for repression of muscle-specific gene transcription also mediate binding to bHLH proteins. We conclude that E1A mediates repression of muscle-specific gene transcription through its amino-terminal domain and propose that this may involve a direct physical interaction between E1A and the bHLH region of myogenic determination proteins.

Glucose modulates the binding of an islet-specific factor to a conserved sequence within the rat I and the human insulin promoters

In cultured rat and human pancreatic islets, glucose stimulated transcription of the rat insulin I gene through the mini-enhancer (FF) located between residues -196 and -247. The glucose-sensitive element was delineated to the region -193 to -227. The mini-enhancer bound islet nuclear proteins to form three major complexes (C1-C3). A 22-bp subfragment, spanning the sequence -206 to -227, was sufficient to retain all binding activities of the entire FF. The homologous sequence of the human insulin promoter interacted with rat islet nuclear extracts to form a single complex, corresponding to the C1 complex of the rat insulin I sequence. C1 was present only in insulin-producing cells; it was the major complex detected in isolated human islets with both rat and human insulin sequences. Furthermore, the DNA binding activity of the C1 factor(s) was selectively modulated by extracellular glucose in a dose-dependent manner; a 4.5-fold increase in binding intensity was detected when rat islets were incubated for 1-3 h in the presence of 20 vs. 1-2 mM glucose. We therefore suggest that the factor(s) involved in the C1 complex corresponds to the glucose-sensitive factor and, consequently, may play a determining role in glucose-regulated expression of the insulin gene.

Cell-specific helix-loop-helix factor required for pituitary expression of the pro-opiomelanocortin gene

Pro-opiomelanocortin (POMC)-expressing cells appear to be the first pituitary cells committed to hormone production. In this work, we have identified an element of the POMC promoter which confers cell-specific activity. This element did not exhibit any activity on its own and required at least one other element of the promoter to manifest its cell-specific activity. Fine mutagenesis of this element indicated that a CANNTG motif is responsible for activity. This E-box motif is typical of binding sites for helix-loop-helix (HLH) transcription factors; however, the POMC cell-specific E box cannot be replaced by other E boxes like the kappa E2 site of the immunoglobulin gene or a muscle-specific E box. Similar E boxes which are present in the insulin gene promoter were shown to contribute to the pancreatic specificity of the insulin promoter. However, E-box-binding proteins found in nuclear extracts from POMC-expressing AtT-20 cells and from insulin-expressing cells have different electrophoretic mobilities. The AtT-20 proteins were named CUTE (for corticotroph upstream transcription element-binding) proteins, and they were not found in any other cells. CUTE proteins have DNA-binding properties characteristic of HLH transcription factors. Overexpression of the dominant negative HLH protein Id or of the ubiquitous positive HLH factor rat Pan-2 decreased or augmented POMC promoter activity, respectively. These observations are consistent with the hypothesis that CUTE factors might be heterodimers. This hypothesis was further supported by antibody shift experiments and by abrogation of DNA binding in the presence of bacterially expressed Id protein. Thus, the cell-specific CUTE proteins and their binding site in the POMC promoter appear to be important determinants for cell specificity of this promoter. The requirement for HLH factors in POMC transcription also presents the possibility that these factors are involved in differentiation of pituitary cells, in analogy with the role of HLH factors in muscle development.

An insulinoma nuclear factor binding to GGGCCC motifs in human insulin gene

Cell specific expression of the insulin gene is achieved through transcriptional mechanisms operating on 5' flanking DNA elements. In the enhancer of rat I insulin gene, two elements, the Nir and Far boxes, located at positions -104 and -233 respectively and containing the same octameric motif are essential for B cell specific transcription activity. Homologous sequences are present in the human insulin gene. While studying the binding of nuclear proteins from insulinoma cells to the -258/+241 region of the human insulin gene, we observed a previously undetected protein binding site in the intron I region between nucleotides +160 and +175. The binding activity was present in insulin producing cells such as RIN and HIT insulinoma cells but not in fibroblasts or insulin negative fibroblast x RIN hybrid cells. DNAse I footprinting and gel retardation/methylation interference experiments allowed us to define the core binding site of the intron binding factor as a GGGCCC hexamer. This factor is also capable to bind to a related sequence, contiguous to the Far-like element in rat and human insulin genes. The binding of the GGGCCC binding factor in this critical region of the insulin gene enhancer may participate in the regulation of insulin gene expression.

Cell-specific helix-loop-helix factor required for pituitary expression of the pro-opiomelanocortin gene

Pro-opiomelanocortin (POMC)-expressing cells appear to be the first pituitary cells committed to hormone production. In this work, we have identified an element of the POMC promoter which confers cell-specific activity. This element did not exhibit any activity on its own and required at least one other element of the promoter to manifest its cell-specific activity. Fine mutagenesis of this element indicated that a CANNTG motif is responsible for activity. This E-box motif is typical of binding sites for helix-loop-helix (HLH) transcription factors; however, the POMC cell-specific E box cannot be replaced by other E boxes like the kappa E2 site of the immunoglobulin gene or a muscle-specific E box. Similar E boxes which are present in the insulin gene promoter were shown to contribute to the pancreatic specificity of the insulin promoter. However, E-box-binding proteins found in nuclear extracts from POMC-expressing AtT-20 cells and from insulin-expressing cells have different electrophoretic mobilities. The AtT-20 proteins were named CUTE (for corticotroph upstream transcription element-binding) proteins, and they were not found in any other cells. CUTE proteins have DNA-binding properties characteristic of HLH transcription factors. Overexpression of the dominant negative HLH protein Id or of the ubiquitous positive HLH factor rat Pan-2 decreased or augmented POMC promoter activity, respectively. These observations are consistent with the hypothesis that CUTE factors might be heterodimers. This hypothesis was further supported by antibody shift experiments and by abrogation of DNA binding in the presence of bacterially expressed Id protein. Thus, the cell-specific CUTE proteins and their binding site in the POMC promoter appear to be important determinants for cell specificity of this promoter. The requirement for HLH factors in POMC transcription also presents the possibility that these factors are involved in differentiation of pituitary cells, in analogy with the role of HLH factors in muscle development.

Differential expression of the two nonallelic proinsulin genes in the developing mouse embryo

In the mouse, insulin is produced from two similar but nonallelic genes that encode proinsulins I and II. We have investigated expression of these two genes during mouse embryonic development, using a PCR to detect the two gene transcripts and immunocytochemistry to visualize the two corresponding proteins. At appearance of the dorsal pancreatic anlage at day 9.5 of gestation, both mRNAs could be detected in the embryos, and both proteins were present together in the same cells of the developing pancreas. At days 9.5 and 10.5, when the ventral anlage appears, there were fewer proinsulin II mRNAs than proinsulin I mRNAs. At day 12.5 this ratio was reversed. Proinsulin II mRNA, but not proinsulin I mRNA, could be detected at day 8.5 in the prepancreatic embryo. Proinsulin II mRNA, but not proinsulin I mRNA, was also found in the heads of embryos at day 9.5 and at all later stages studied. These results indicate that the two proinsulin genes are regulated independently, at least in part. They also suggest that insulin might play a role as a growth factor in the developing mouse brain.

Multiple elements in the upstream glucokinase promoter contribute to transcription in insulinoma cells

beta-cell type-specific expression of the upstream glucokinase promoter was studied by transfection of fusion genes and analysis of DNA-protein interactions. A construct containing 1,000 bp of 5'-flanking DNA was efficiently expressed in HIT M2.2.2 cells, a beta-cell-derived line that makes both insulin and glucokinase, but not in NIH 3T3 cells, a heterologous cell line. In a series of 5' deletion mutations between bases -1000 and -100 (relative to a base previously designated +1), efficient expression in HIT cells was maintained until -280 bp, after which transcription decreased in a stepwise manner. The sequences between -180 and -1 bp contributing to transcriptional activity in HIT cells were identified by studying 28 block transversion mutants that spanned this region in 10-bp steps. Two mutations reduced transcription 10-fold or more, while six reduced transcription between 3- and 10-fold. Three mutationally sensitive regions of this promoter were found to bind to a factor that was expressed preferentially in pancreatic islet beta cells. The binding sites, designated upstream promoter elements (UPEs), shared a consensus sequence of CAT(T/C)A(C/G). Methylation of adenine and guanine residues within this sequence prevented binding of the beta-cell factor, as did mutations at positions 2, 3, and 5. Analysis of nuclear extracts from different cell lines identified UPE-binding activity in HIT M2.2.2 and beta-TC-3 cells but not in AtT-20, NIH 3T3, or HeLa cells; the possibility of a greatly reduced amount in alpha-TC-6 cells could not be excluded. UV laser cross-linking experiments supported the beta-cell type expression of this factor and showed it to be approximately 50 kDa in size. Gel mobility shift competition experiments showed that this beta-cell factor is the same that binds to similar elements, termed CT boxes, in the insulin promoter. Thus, a role for these elements (UPEs or CT boxes), and the beta-cell factor that binds to them, in determining the expression of genes in the beta cells of pancreatic islets is suggested.

Multiple elements in the upstream glucokinase promoter contribute to transcription in insulinoma cells

beta-cell type-specific expression of the upstream glucokinase promoter was studied by transfection of fusion genes and analysis of DNA-protein interactions. A construct containing 1,000 bp of 5'-flanking DNA was efficiently expressed in HIT M2.2.2 cells, a beta-cell-derived line that makes both insulin and glucokinase, but not in NIH 3T3 cells, a heterologous cell line. In a series of 5' deletion mutations between bases -1000 and -100 (relative to a base previously designated +1), efficient expression in HIT cells was maintained until -280 bp, after which transcription decreased in a stepwise manner. The sequences between -180 and -1 bp contributing to transcriptional activity in HIT cells were identified by studying 28 block transversion mutants that spanned this region in 10-bp steps. Two mutations reduced transcription 10-fold or more, while six reduced transcription between 3- and 10-fold. Three mutationally sensitive regions of this promoter were found to bind to a factor that was expressed preferentially in pancreatic islet beta cells. The binding sites, designated upstream promoter elements (UPEs), shared a consensus sequence of CAT(T/C)A(C/G). Methylation of adenine and guanine residues within this sequence prevented binding of the beta-cell factor, as did mutations at positions 2, 3, and 5. Analysis of nuclear extracts from different cell lines identified UPE-binding activity in HIT M2.2.2 and beta-TC-3 cells but not in AtT-20, NIH 3T3, or HeLa cells; the possibility of a greatly reduced amount in alpha-TC-6 cells could not be excluded. UV laser cross-linking experiments supported the beta-cell type expression of this factor and showed it to be approximately 50 kDa in size. Gel mobility shift competition experiments showed that this beta-cell factor is the same that binds to similar elements, termed CT boxes, in the insulin promoter. Thus, a role for these elements (UPEs or CT boxes), and the beta-cell factor that binds to them, in determining the expression of genes in the beta cells of pancreatic islets is suggested.

Hepatocyte nuclear factor 1 alpha is expressed in a hamster insulinoma line and transactivates the rat insulin I gene

Systematic mutational analysis previously identified two primary regulatory elements within a minienhancer (-247 to -198) of the rat insulin I promoter that are critical for transcriptional activity. The Far box (-241 to -232) and the FLAT element (-222 to -208) synergistically upregulate transcription and, together, are sufficient to confer tissue-specific and glucose-responsive transcriptional activity on a heterologous promoter. Detailed analysis of the FLAT element further revealed that, in addition to the positive regulatory activity it mediates in tandem with the Far box, it is a site for negative regulatory control. A portion of the FLAT element bears considerable sequence similarity to the consensus binding site for hepatocyte nuclear factor 1 alpha (HNF1 alpha; LF-B1) a liver-enriched homeodomain-containing transcription factor. Here we show that the HNF1-like site within the FLAT element exhibited positive transcriptional activity in both HepG2 and HIT cells and bound similar, but distinguishable, nuclear protein complexes in the respective nuclear extracts. Screening of a hamster insulinoma cDNA library with a PCR-derived probe encompassing the DNA-binding domain of rat HNF1 alpha resulted in isolation of a hamster HNF1 alpha (hHNF1 alpha) cDNA homolog. Specific antiserum identified the HNF1 alpha protein as one component of a specific FLAT-binding complex in HIT nuclear extracts. Expression of the hHNF1 alpha cDNA in COS cells resulted in transactivation of reporter constructs containing multimerized segments of the rat insulin I minienhancer. Thus, HNF1 alpha, one component of a DNA-binding complex involved in transcriptional regulation of the rat insulin I gene, may play a significant role in nonhepatic as well as hepatic gene transcription.

Control of insulin gene expression by glucose

Northern-blot analysis was used to demonstrate that an increase in extracellular glucose concentration increased the content of preproinsulin mRNA 2.3-fold in the beta-cell line HIT T15. A probe for the constitutively expressed glyceraldehyde-3-phosphate dehydrogenase was used as a control. Mannoheptulose blocked this effect of glucose. A stimulatory effect on preproinsulin mRNA levels was also observed in response to mannose and to 4-methyl-2-oxopentanoate. However, galactose and arginine were ineffective. Glucagon, forskolin and dibutyryl cyclic AMP also elicited an increase in HIT-cell preproinsulin mRNA. The ability of the 5' upstream region of the preproinsulin gene to mediate the effect of glucose and other metabolites on transcription was studied by using a bacterial reporter gene technique. HIT cells were transfected with a plasmid, pOK1, containing the upstream region of the rat insulin-1 gene (-345 to +1) linked to chloramphenicol acetyltransferase (CAT). Co-transfection with a plasmid pRSV beta-gal containing beta-galactosidase driven by the Rous sarcoma virus promoter was used as a control for the efficiency of transfection; expression of CAT activity in transfected HIT cells was normalized by reference to expression of beta-galactosidase. Glucose caused a dose-dependent increase in expression of CAT activity, with a half-maximal effect at 5.5 mM and a maximum response of 4-fold. Mannoheptulose blocked this effect of glucose. Other metabolites (mannose, 4-methyl-2-oxopentanoate and leucine plus glutamine) were also able to increase insulin promoter-driven CAT expression, but galactose and arginine were ineffective. The stimulatory effect of glucose on CAT expression was not blocked by verapamil and was inhibited by increasing extracellular Ca2+ from 0.4 to 5 mM. Both dibutyryl cyclic AMP and forskolin caused an increase in insulin promoter-driven gene expression in the presence of 1 mM-glucose, but neither agent further increased the level of expression occurring in the presence of a maximally stimulating glucose concentration. The phorbol ester phorbol 12-myristate 13-acetate (PMA) also increased insulin promoter-driven CAT expression in the presence of 1 mM-, but not 11 mM-glucose. Staurosporine blocked the stimulatory effect not only of PMA but also of glucose and of dibutyryl cyclic AMP. We conclude that the 5' upstream region of the insulin gene contains sequences responsible for mediating the stimulatory effect of glucose on insulin-gene transcription.(ABSTRACT TRUNCATED AT 400 WORDS)

The insulin and islet amyloid polypeptide genes contain similar cell-specific promoter elements that bind identical beta-cell nuclear complexes

The pancreatic beta cell makes several unique gene products, including insulin, islet amyloid polypeptide (IAPP), and beta-cell-specific glucokinase (beta GK). The functions of isolated portions of the insulin, IAPP, and beta GK promoters were studied by using transient expression and DNA binding assays. A short portion (-247 to -197 bp) of the rat insulin I gene, the FF minienhancer, contains three interacting transcriptional regulatory elements. The FF minienhancer binds at least two nuclear complexes with limited tissue distribution. Sequences similar to that of the FF minienhancer are present in the 5' flanking DNA of the human IAPP and rat beta GK genes and also the rat insulin II and mouse insulin I and II genes. Similar minienhancer constructs from the insulin and IAPP genes function as cell-specific transcriptional regulatory elements and compete for binding of the same nuclear factors, while the beta GK construct competes for protein binding but functions poorly as a minienhancer. These observations suggest that the patterns of expression of the beta-cell-specific genes result in part from sharing the same transcriptional regulators.

The insulin and islet amyloid polypeptide genes contain similar cell-specific promoter elements that bind identical beta-cell nuclear complexes

The pancreatic beta cell makes several unique gene products, including insulin, islet amyloid polypeptide (IAPP), and beta-cell-specific glucokinase (beta GK). The functions of isolated portions of the insulin, IAPP, and beta GK promoters were studied by using transient expression and DNA binding assays. A short portion (-247 to -197 bp) of the rat insulin I gene, the FF minienhancer, contains three interacting transcriptional regulatory elements. The FF minienhancer binds at least two nuclear complexes with limited tissue distribution. Sequences similar to that of the FF minienhancer are present in the 5' flanking DNA of the human IAPP and rat beta GK genes and also the rat insulin II and mouse insulin I and II genes. Similar minienhancer constructs from the insulin and IAPP genes function as cell-specific transcriptional regulatory elements and compete for binding of the same nuclear factors, while the beta GK construct competes for protein binding but functions poorly as a minienhancer. These observations suggest that the patterns of expression of the beta-cell-specific genes result in part from sharing the same transcriptional regulators.

Regulatory regions of rat insulin I gene necessary for expression in transgenic mice

Ten transgenic mouse lines harboring the -346/-103 fragment of the rat insulin I enhancer linked to a heterologous promoter and a reporter gene (Eins-Ptk-CAT construct) were produced. Expression of the hybrid transgene was essentially observed in pancreas and to a lesser extent in brain. These results indicate that the rat insulin I promoter is dispensable for pancreatic expression. This insulin gene sequence is the shortest fragment described as conferring tissue-specific expression in transgenic mice. Two short homologous sequences in the rat insulin I enhancer fragment used, IEB2 and IEB1, have been described as playing a dominant role in the regulation of HIT hamster insulinoma cell-specific transcription of the insulin gene (1). We investigated whether the combination of IEB2 and IEB1 sequences is sufficient to confer specific expression in transgenic mice to a IEB2-IEB1-Ptk-CAT gene construct. No CAT activity was observed neither in pancreas nor in any other organ examined in 19 different transgenic mice. Moreover in transient expression experiments in RIN2A rat insulinoma cells, the IEB sequences had a very weak or no enhancer activity. These observations contribute to the conclusion that DNA regulatory elements other than the IEB sequences are necessary for gene expression in vivo.

Insulin gene expression in nonexpressing cells appears to be regulated by multiple distinct negative-acting control elements

Selective transcription of the insulin gene in pancreatic beta cells is regulated by its enhancer, located between nucleotides -340 and -91 relative to the transcription start site. Transcription from the enhancer is controlled by both positive- and negative-acting cellular factors. Cell-type-specific expression is mediated principally by a single cis-acting enhancer element located between -100 and -91 in the rat insulin II gene (referred to as the insulin control element [ICE]), which is acted upon by both of these cellular activities. Analysis of the effect of 5' deletions within the insulin enhancer has identified a region between nucleotides -217 and -197 that is also a site of negative control. Deletion of these sequences from the 5' end of the enhancer leads to transcription of the enhancer in non-insulin-producing cells, even though the ICE is intact. Derepression of this ICE-mediated effect was shown to be due to the binding of a ubiquitously distributed cellular factor to a sequence element which resides just upstream of the ICE (i.e., between nucleotides -110 and -100). We discuss the possible relationship of these results to cell-type-specific regulation of the insulin gene.

Insulin gene expression in nonexpressing cells appears to be regulated by multiple distinct negative-acting control elements

Selective transcription of the insulin gene in pancreatic beta cells is regulated by its enhancer, located between nucleotides -340 and -91 relative to the transcription start site. Transcription from the enhancer is controlled by both positive- and negative-acting cellular factors. Cell-type-specific expression is mediated principally by a single cis-acting enhancer element located between -100 and -91 in the rat insulin II gene (referred to as the insulin control element [ICE]), which is acted upon by both of these cellular activities. Analysis of the effect of 5' deletions within the insulin enhancer has identified a region between nucleotides -217 and -197 that is also a site of negative control. Deletion of these sequences from the 5' end of the enhancer leads to transcription of the enhancer in non-insulin-producing cells, even though the ICE is intact. Derepression of this ICE-mediated effect was shown to be due to the binding of a ubiquitously distributed cellular factor to a sequence element which resides just upstream of the ICE (i.e., between nucleotides -110 and -100). We discuss the possible relationship of these results to cell-type-specific regulation of the insulin gene.

Pancreatic beta-cell-type-specific transcription of the insulin gene is mediated by basic helix-loop-helix DNA-binding proteins

The pancreatic beta-cell-specific expression of the insulin gene is mediated, at least in part, by the interaction of unique trans-acting beta-cell factors with a cis-acting DNA element found within the insulin enhancer (5'-GC CATCTG-3'; referred to as the insulin control element [ICE]) present in the rat insulin II gene between positions -100 and -91. This sequence element contains the consensus binding site for a group of DNA-binding transcription factors called basic helix-loop-helix proteins (B-HLH). As a consequence of the similarity of the ICE with the DNA sequence motif associated with the cis-acting elements of the B-HLH class of binding proteins (CANNTG), the ability of this class of proteins to regulate cell-type-specific expression of the insulin gene was addressed. Cotransfection experiments indicated that overexpression of Id, a negative regulator of B-HLH protein function, inhibits ICE-mediated activity. Antibody to the E12/E47 B-HLH proteins attenuated the formation, in vitro, of a previously described (J. Whelan, S. R. Cordle, E. Henderson, P. A. Weil, and R. Stein, Mol. Cell. Biol. 10:1564-1572, 1990) beta-cell-specific activator factor(s)-ICE DNA complex. Both of these B-HLH proteins (E12 and E47) bound efficiently and specifically to the ICE sequences. The role of B-HLH proteins in mediating pancreatic beta-cell-specific transcription of the insulin gene is discussed.

Pancreatic beta-cell-type-specific transcription of the insulin gene is mediated by basic helix-loop-helix DNA-binding proteins

The pancreatic beta-cell-specific expression of the insulin gene is mediated, at least in part, by the interaction of unique trans-acting beta-cell factors with a cis-acting DNA element found within the insulin enhancer (5'-GC CATCTG-3'; referred to as the insulin control element [ICE]) present in the rat insulin II gene between positions -100 and -91. This sequence element contains the consensus binding site for a group of DNA-binding transcription factors called basic helix-loop-helix proteins (B-HLH). As a consequence of the similarity of the ICE with the DNA sequence motif associated with the cis-acting elements of the B-HLH class of binding proteins (CANNTG), the ability of this class of proteins to regulate cell-type-specific expression of the insulin gene was addressed. Cotransfection experiments indicated that overexpression of Id, a negative regulator of B-HLH protein function, inhibits ICE-mediated activity. Antibody to the E12/E47 B-HLH proteins attenuated the formation, in vitro, of a previously described (J. Whelan, S. R. Cordle, E. Henderson, P. A. Weil, and R. Stein, Mol. Cell. Biol. 10:1564-1572, 1990) beta-cell-specific activator factor(s)-ICE DNA complex. Both of these B-HLH proteins (E12 and E47) bound efficiently and specifically to the ICE sequences. The role of B-HLH proteins in mediating pancreatic beta-cell-specific transcription of the insulin gene is discussed.

Cooperativity of sequence elements mediates tissue specificity of the rat insulin II gene

The 5'-flanking region of the rat insulin II gene (-448 to +50) is sufficient for tissue-specific expression. To further determine the tissue-specific cis-acting element(s), important sequences defined by linker-scanning mutagenesis were placed upstream of a heterologous promoter and transfected into insulin-producing and -nonproducing cells. Rat insulin promoter element 3 (RIPE3), which spans from -125 to -86, was shown to confer beta-cell-specific expression in either orientation. However, two subregions of RIPE3, RIPE3a and RIPE3b (defined by linker-scanning mutations), displayed only marginal activities. These results suggest that the two subregions cooperate to confer tissue specificity, presumably via their cognate binding factors.

Identification of a pancreatic beta-cell insulin gene transcription factor that binds to and appears to activate cell-type-specific expression: its possible relationship to other cellular factors that bind to a common insulin gene sequence

The insulin gene is expressed almost exclusively in pancreatic beta-cells. Previous work in our laboratory has shown that pancreatic beta-cell-specific expression of the rat insulin II gene is controlled by a number of positive and negative cis-acting DNA elements within the enhancer. We have shown that one element within the enhancer, located between nucleotides -100 and -91 (GCCATCTGCT; referred to as the insulin control element [ICE]) relative to the transcription start site, is controlled by both positive- and negative-acting cellular transcription factors. The positive-acting factor appears to be uniquely active in beta-cells. To identify the nucleotides within the ICE that mediate positive cell-type-specific regulation, point mutations within this element were generated and assayed for their effects on expression. Base pairs -97, -94, -93, and -92 were found to be crucial for the activator function of this region, while mutations at base pairs -100, -96, and -91 had little or no effect on activity. The gel mobility shift assay was used to determine whether specific cellular factors associated directly with the ICE. Several specific protein-DNA complexes were detected in extracts prepared from insulin-producing and non-insulin-producing cells, including a complex unique to beta-cell extracts. The ability of unlabeled wild-type and point mutant versions of the ICE to compete for binding to these cellular factors demonstrated that the beta-cell-specific complex appears to contain the insulin gene activator protein(s). Interestingly, the adenovirus type 2 major late promoter upstream element (USE; GCCACGTGAC) also competed in the gel mobility shift assay for binding of cellular proteins to the ICE. These results suggested that the cellular factor that binds to the USE (i.e., USF) also interacts with the ICE. This was directly demonstrated by showing that ICE and USE sequences completed for the USF required for adenovirus type 2 major late promoter transcription in vitro and by showing that reticulocyte lysate-translated human USF products bound to the ICE. However, the USE sequences were unable to stimulate beta-cell-type-specific activity in vivo. We discuss the possible relationship of these observations to positive and negative control mediated by the ICE.

Identification of a pancreatic beta-cell insulin gene transcription factor that binds to and appears to activate cell-type-specific expression: its possible relationship to other cellular factors that bind to a common insulin gene sequence

The insulin gene is expressed almost exclusively in pancreatic beta-cells. Previous work in our laboratory has shown that pancreatic beta-cell-specific expression of the rat insulin II gene is controlled by a number of positive and negative cis-acting DNA elements within the enhancer. We have shown that one element within the enhancer, located between nucleotides -100 and -91 (GCCATCTGCT; referred to as the insulin control element [ICE]) relative to the transcription start site, is controlled by both positive- and negative-acting cellular transcription factors. The positive-acting factor appears to be uniquely active in beta-cells. To identify the nucleotides within the ICE that mediate positive cell-type-specific regulation, point mutations within this element were generated and assayed for their effects on expression. Base pairs -97, -94, -93, and -92 were found to be crucial for the activator function of this region, while mutations at base pairs -100, -96, and -91 had little or no effect on activity. The gel mobility shift assay was used to determine whether specific cellular factors associated directly with the ICE. Several specific protein-DNA complexes were detected in extracts prepared from insulin-producing and non-insulin-producing cells, including a complex unique to beta-cell extracts. The ability of unlabeled wild-type and point mutant versions of the ICE to compete for binding to these cellular factors demonstrated that the beta-cell-specific complex appears to contain the insulin gene activator protein(s). Interestingly, the adenovirus type 2 major late promoter upstream element (USE; GCCACGTGAC) also competed in the gel mobility shift assay for binding of cellular proteins to the ICE. These results suggested that the cellular factor that binds to the USE (i.e., USF) also interacts with the ICE. This was directly demonstrated by showing that ICE and USE sequences completed for the USF required for adenovirus type 2 major late promoter transcription in vitro and by showing that reticulocyte lysate-translated human USF products bound to the ICE. However, the USE sequences were unable to stimulate beta-cell-type-specific activity in vivo. We discuss the possible relationship of these observations to positive and negative control mediated by the ICE.