Role of three FKHR phosphorylation sites in insulin inhibition of FKHR action in hepatocytes.

Insulin inhibits insulin-like growth factor binding protein-1 (IGFBP-1) transcription by preventing FKHR protein family members from binding a specific insulin response element in the IGFBP-1 promoter. In most cells, three serine/threonine moieties in FKHR family members are phosphorylated after insulin treatment or protein kinase B/Akt (PKB) transfection, and each of the three phosphorylated PKB sites contributes to insulin- or PKB-mediated inhibition of both the action and the nuclear localization of FKHR family members. In hepatocytes, however, the middle PKB site (PKB2) of FKHR was required for insulin to phosphorylate FKHR and was the only PKB site that participated in insulin inhibition of FKHR action, indicating that insulin utilizes a unique pathway to regulate FKHR action in hepatocytes. In studies presented here, plasmids expressing native or mutant FKHR forms, either with or without N-terminal fusion to green fluorescent protein (GFP), were transiently transfected into HEP G2 cells. All FKHR forms stimulated IGFBP-1 promoter activity, and mutating any of the three FKHR PKB sites impaired the ability of insulin both to inhibit FKHR-stimulated IGFBP-1 promoter activity and to induce FKHR accumulation in cytoplasm. Thus, in hepatocytes as in other cell lines, all three FKHR PKB sites participate in insulin-mediated inhibition of FKHR action and in insulin-mediated accumulation of FKHR in cytoplasm.

Conservation of an insulin response unit between mouse and human glucose-6-phosphatase catalytic subunit gene promoters: transcription factor FKHR binds the insulin response sequence.

Because overexpression of the glucose-6-phosphatase catalytic subunit (G-6-Pase) in both type 1 and type 2 diabetes may contribute to the characteristic increased rate of hepatic glucose production, we have investigated whether the insulin response unit (IRU) identified in the mouse G-6-Pase promoter is conserved in the human promoter. A series of human G-6-Pase-chloramphenicol acetyltransferase (CAT) fusion genes was transiently transfected into human HepG2 hepatoma cells, and the effect of insulin on basal CAT expression was analyzed. The results suggest that the IRU identified in the mouse promoter is conserved in the human promoter, but that an upstream multimerized insulin response sequence (IRS) motif that is only found in the human promoter appears to be functionally inactive. The G-6-Pase IRU comprises two distinct promoter regions, designated A and B. Region B contains an IRS, whereas region A acts as an accessory element to enhance the effect of insulin, mediated through region B, on basal G-6-Pase gene transcription. We have previously shown that the accessory factor binding region A is hepatocyte nuclear factor-1, and we show here that the forkhead protein FKHR is a candidate for the insulin-responsive transcription factor binding region B.

Activation of the insulin-like growth factor binding protein-1 promoter by progesterone receptor in decidualized human endometrial stromal cells.

Insulin-like growth factor binding protein-1 (IGFBP-1) is induced extensively when human endometrial stromal cells are decidualized by progestin and relaxin in a long-term primary culture system. The purpose of this study is to investigate whether progesterone receptor (PR) directly activates the IGFBP-1 gene promoter. In decidualized stromal cells, activity of the IGFBP-1 promoter (from -1.2 kb to +68 bp) containing putative progesterone-response elements (PREs) was increased 80-fold. Mutation of either 5' or 3' half-site of the putative PRE1 site (from -193 to -179 bp) reduced the promoter activity. Mutations that converted PRE1 closer to consensus PRE increased the promoter activity. In undifferentiated stromal cells, mutations of PRE sites had no effect on the promoter activity. When a PR expression vector (hPR1) was cotransfected, progestin increased promoter activity derived from p275CAT but not from p1.2CAT, suggesting that the function of PRE1 was repressed by the region from -1.2 kb to -275 bp in the promoter. Progestin did not increase promoter activity derived from p275CAT without cotransfection of hPR1, suggesting that endogenous PR alone is insufficient to activate PRE1. In summary, results indicate that the PRE1 site of the IGFBP-1 promoter mediates a direct activation of PR on transcription specifically in decidualized stromal cells.

FKHR binds the insulin response element in the insulin-like growth factor binding protein-1 promoter.

The insulin response element (IRE) in the IGFBP-1 promoter, and in other gene promoters, contains a T(A/G)TTT motif essential for insulin inhibition of transcription. Studies presented here test whether FKHR may be the transcription factor that confers insulin inhibition through this IRE motif. Immunoblots using antiserum to the synthetic peptide FKHR413-430, RNase protection, and Northerns blots show that FKHR is expressed in HEP G2 human hepatoma cells. Southwestern blots, electromobility shift assays, and DNase I protection assays show that Escherichia coli-expressed GST-FKHR binds specifically to IREs from the IGFBP-1, PEPCK and TAT genes; however, unlike HNF3beta, another protein proposed to be the insulin regulated factor, GST-FKHR does not bind the insulin unresponsive G/C-A/C mutation of the IGFBP-1 IRE. When HEP G2 cells were cotransfected with FKHR expression vectors and with IGFBP-1 promoter plasmids containing either native or mutant IREs, FKHR expression induced a 5-fold increase in activity of the native IGFBP-1 promoter but no increase in activity of promoter constructs containing insulin unresponsive IRE mutants. These data suggest that FKHR, and/or a related family member, is the important T(G/A)TTT binding protein that confers the inhibitory effect of insulin on gene transcription.

Hepatic nuclear factor 3 and high mobility group I/Y proteins bind the insulin response element of the insulin-like growth factor-binding protein-1 promoter.

The insulin response element (IRE) of the human insulin-like growth factor-binding protein-1 (IGFBP-1) promoter contains a palindrome of the T(A/G)TTT sequence crucial to hormonal regulation of many genes. In initial studies of how this IRE participates in hormonal regulation, the electromobility shift assay was used under a variety of conditions to identify IRE-binding proteins. An exhaustive search identified five proteins that specifically bind this IRE; purified proteins were used to show that all five are related to either the high mobility group I/Y (HMGI/Y) or hepatic nuclear factor 3 (HNF3) protein families. Further studies used purified HNF3 and HMGI proteins to show: 1) eah protects the IGFBP-1 IRE from deoxyribonuclease I (DNaseI) digestion; and 2) HNF3 but not HMGI/Y binds to the related phosphoenolpyruvate carboxykinase and Apo CIII IREs. A series of IRE mutants with variable responsiveness to insulin were used to show that the presence of a TGTTT sequence in the mutants did parallel, but HMGI/Y and HNF3 binding to the mutants did not parallel, the ability of the mutants to confer the inhibitory effect of insulin. In contrast, HNF3 binding to these IRE mutants roughly correlates with response of the mutants to glucocorticoids. The way by which HNF3 and/or other as yet unidentified IRE-binding proteins confer insulin inhibition to IGFBP-1 transcription and the role of HMGI/Y in IRE function have yet to be established.

Hepatic nuclear factor 3- and hormone-regulated expression of the phosphoenolpyruvate carboxykinase and insulin-like growth factor-binding protein 1 genes.

The rate of transcription of the hepatic phosphoenolpyruvate carboxykinase (PEPCK) and insulin-like growth factor-binding protein 1 (IGFBP-1) genes is stimulated by glucocorticoids and inhibited by insulin. In both cases, the effect of insulin is dominant, since it suppresses both basal and glucocorticoid-stimulated PEPCK or IGFBP-1 gene transcription. Analyses of both promoters by transfection of PEPCK or IGFBP-1-chloramphenicol acetyltransferase fusion genes into rat hepatoma cells has led to the identification of insulin response sequences (IRSs) in both genes. The core IRS, T(G/A)TTTTG, is the same in both genes, but the PEPCK promoter has a single copy of this element whereas the IGFBP-1 promoter has two copies arranged as an inverted palindrome. The IGFBP-1 IRS and PEPCK IRS both bind the alpha and beta forms of hepatic nuclear factor 3 (HNF-3), although the latter does so with a sixfold-lower relative affinity. Both the PEPCK and the IGFBP-1 IRSs also function as accessory factor binding sites required for the full induction of gene transcription by glucocorticoids. A combination of transient transfection and DNA binding studies suggests that HNF-3 is the accessory factor that supports glucocorticoid-induced gene transcription. In both genes, the HNF-3 binding site overlaps the IRS core motif(s). A model in which insulin is postulated to mediate its negative effect on glucocorticoid-induced PEPCK and IGFBP-1 gene transcription indirectly by inhibiting HNF-3 action is proposed.

Glucocorticoids and insulin regulate expression of the human gene for insulin-like growth factor-binding protein-1 through proximal promoter elements.

Glucocorticoids stimulate, while insulin inhibits, the hepatic transcription of insulin-like growth factor-binding protein-1 (IGFBP-1). In the present studies, human HEP G2 hepatoma cells were transiently transfected with human (h)IGFBP-1 promoter constructs. Activity of a construct containing the first 1205 base pairs (bp) of the hIGFBP-1 promoter was stimulated 6-9.5-fold by dexamethasone, and this increase was inhibited approximately 76% by insulin. Deletion and site-directed mutations of the hIGFBP-1 promoter (a) identified two glucocorticoid response elements, located within the first 200 bp of the promoter, which are essential for dexamethasone-stimulated promoter activity and which specifically bind human glucocorticoid receptor; (b) showed that a recently characterized insulin-responsive element, located approximately 110 bp 5' to the transcription start site (Suwanichkul, A., Morris, S.L., and Powell, D. R. (1993) J. Biol. Chem. 268, 17063-17068), confers the entire inhibitory effect of insulin not only on basal but also on glucocorticoid-stimulated promoter activity; and (c) showed that this insulin-responsive element is essential for maximal glucocorticoid-stimulated activity. These studies suggest that the interaction of proteins that bind to a cluster of cis elements located in the first 200 bp of the hIGFBP-1 promoter are of major importance in modulating the opposing effects of glucocorticoids and insulin on hepatic hIGFBP-1 expression.

Characterization of the chromosomal gene and promoter for human insulin-like growth factor binding protein-5.

To better understand the regulation of insulin-like growth factor binding protein-5 (IGFBP-5) expression, we cloned the IGFBP-5 gene from human genomic libraries and identified a region in the 5' flanking sequence which functions as a promoter. The human IGFBP-5 gene is divided into four exons which, primarily due to a first intron of approximately 25 kilobases, span approximately 33 kilobases of DNA. Southern analysis identified a single copy of the IGFBP-5 gene in the haploid human genome, and several independent mapping strategies found this gene tightly linked with, and in opposite transcriptional orientation to, the IGFBP-2 gene at chromosomal region 2q33-34. Primer extension studies identified the IGFBP-5 mRNA cap site 772 base pairs (bp) 5' to the first nucleotide of the translation start codon. Analysis of the 5'-flanking sequence identified a potential TATA element beginning 33 bp 5' to the mRNA cap site. When a DNA fragment containing this cap site and 461 bp of upstream sequence was placed 5' to the chloramphenicol acetyltransferase reporter gene and transfected into MDA-MB-468 human breast cancer cells, it directed chloramphenicol acetyltransferase expression in an orientation-specific manner, suggesting that this region contains elements essential for IGFBP-5 promoter activity.

Insulin-like growth factor (IGF) suppression of IGFBP-1 production: evidence for mediation by the type I IGF receptor.

The regulation of insulin-like growth factor binding protein-1 (IGFBP-1) by its ligands, IGF-I and IGF-II, was studied in continuous cultures of HepG2 human hepatoma cells. Both IGF-I and IGF-II in concentrations as low as 1-10 nmol/l caused significant suppression of IGFBP-I protein levels. This suppression was accompanied by decreased IGFBP-1 mRNA levels occurring within 2-4 h of exposure to IGF-I or IGF-II, and by a significant decrease in IGFBP-1 promoter activity. IGF-I and IGF-II were equipotent in suppressing basal levels of IGFBP-1 protein, mRNA and promoter activity. IGF-I, IGF-II, and IGF-analogs with low IGFBP-1 affinity, (des 1-3)IGF-I and long R3IGF-I, all potently suppressed the previously characterized increase in IGFBP-1 protein levels and promoter activity induced by cAMP and theophylline. In contrast, [Leu-27]IGF-II, which interacts with the type II but not type I IGF receptor, had no effect on IGFBP-1 protein levels or promoter activity. Our data indicate that IGFBP-1 production is inhibited by its ligands, IGF-I and IGF-II, and that this effect is probably mediated at the transcriptional level. The effects of IGF-I and IGF-II apparently occur as a result of binding to the type I IGF receptor, and are similar to the previously characterized suppressive effects of insulin on IGFBP-1 transcription mediated through the insulin receptor. When considered with previous data regarding expression of IGFBP-1 and the type I IGF receptor, our results suggest that IGF regulation of IGFBP-1 may play an as yet undefined role in fetal development and postnatal hepatic regeneration.

HNF1 activates transcription of the human gene for insulin-like growth factor binding protein-1.

Insulin-like growth factor binding protein-1 (IGFBP-1) is expressed primarily in the liver, kidney, and uterus. Basal IGFBP-1 promoter activity in human HEP G2 hepatoma cells is dependent upon a proximal promoter element that binds hepatic nuclear factor 1 (HNF1), a protein that is likely to be an important factor regulating the expression of many genes in liver and kidney. To test whether HNF1 activates IGFBP-1 transcription, HEP G2 cells and HeLa cells were cotransfected transiently with HNF1 expression vectors and with IGFBP-1 promoter/chloramphenicol acetyltransferase reporter gene constructs. HNF1 increased IGFBP-1 promoter activity in both HEP G2 and HeLa cells. Gel mobility-shift assays and additional transfections in HeLa cells showed that expressed full-length and carboxy-terminal truncated forms of HNF1 could each bind the HNF1 cis element of the IGFBP-1 promoter; however, significant trans-activation only occurred in the presence of the full-length HNF1 protein, similar to past experience with these two HNF1 forms and the albumin promoter. Further studies showed that IGFBP-1 promoter constructs containing mutations with high or low affinity for HNF1 responded to HNF1 expression with increased or decreased activity, respectively, relative to the native promoter. These studies suggest that HNF1 and/or related proteins play a role in hepatic, and perhaps also renal, expression of IGFBP-1.

Identification of a promoter element which participates in cAMP-stimulated expression of human insulin-like growth factor-binding protein-1.

In hepatocytes, insulin-like growth factor-binding protein-1 (IGFBP-1) levels are increased by glucocorticoids and by agents that raise intracellular cAMP levels such as glucagon, theophylline, forskolin, and cAMP analogues. In contrast, insulin lowers IGFBP-1 levels, an effect dominant over the glucocorticoid and cAMP effects. Previous studies showed that dibutyryl cAMP (Bt2cAMP) and theophylline increase IGFBP-1 promoter activity in HEP G2 human hepatoma cells and that insulin abolishes this increase. In studies reported here, HEP G2 cells were used to further evaluate the role of cAMP in stimulating IGFBP-1 expression. Initial studies found that either 0.5 or 5.0 mM Bt2cAMP alone, or the combination of 0.5 mM Bt2cAMP and 2 mM theophylline, increased IGFBP-1 protein levels, mRNA levels, and promoter activity, but that the addition of theophylline to Bt2cAMP was required to give a approximately 5-fold increase in promoter activity. Deletion mutations of the IGFBP-1 promoter were used to show that much of the effect of Bt2cAMP and theophylline was conferred by the region between 269 and 246 base pairs (bp) 5' of the IGFBP-1 mRNA cap site. DNase I protection assays showed that HEP G2 nuclear extract footprinted the region from 273 to 249 bp 5' of the cap site; this region, designated P2, has a central CGTCA motif common to cAMP-responsive elements (CREs). Mutating the CGTCA motif in the 1205-bp IGFBP-1 promoter construct to TAGCA led to a 51% decrease in the ability of Bt2cAMP and theophylline to stimulate IGFBP-1 promoter activity above control levels. In addition, cotransfection of the catalytic subunit of cAMP-dependent protein kinase A (PKA) with the native 1205-bp IGFBP-1 promoter construct stimulated IGFBP-1 promoter activity 3.9-fold, but the TAGCA mutation decreased by 73% the ability of PKA to stimulate IGFBP-1 promoter activity above control levels. Mutating the CGTCA motif to TAGCA also blocked the ability of both crude HEP G2 nuclear extract and recombinant CRE-binding protein to bind to the P2 element. These data suggest that the P2 element is a CRE that confers at least part of the stimulatory effect of cAMP on the human IGFBP-1 promoter.

Dexamethasone stimulates expression of insulin-like growth factor binding protein-1 in HEP G2 human hepatoma cells.

Insulin-like growth factor binding protein-1 (IGFBP-1) serum levels are increased by glucocorticoids and glucagon, and are decreased by insulin; these effects seem to reflect changes in hepatic IGFBP-1 expression. Human HEP G2 hepatoma cells respond to cAMP or insulin with stimulation or inhibition of IGFBP-1 expression, respectively; however, dexamethasone alone does not stimulate IGFBP-1 expression. Studies presented here show that in the presence of cAMP and theophylline, nontransfected HEP G2 cells respond to further addition of dexamethasone with a approximately 70% rise in IGFBP-1 mRNA levels, and HEP G2 cells transfected with IGFBP-1 promoter constructs respond to further addition of dexamethasone with a approximately 70% rise in IGFBP-1 protein levels and a approximately 9-fold rise in IGFBP-1 promoter activity. The stimulatory effect of dexamethasone and cAMP, conferred to the IGFBP-1 promoter between 357 and 103 base pairs 5' to the mRNA cap site, is inhibited by insulin. Thus, the cis elements necessary to confer multihormonal regulation of IGFBP-1 expression appear to reside in a short stretch of DNA just upstream from the IGFBP-1 transcription start site.

Insulin inhibits transcription of the human gene for insulin-like growth factor-binding protein-1.

Insulin rapidly lowers serum insulin-like growth factor-binding protein-1 (IGFBP-1) levels in vivo. In studies reported here, HEP G2 cells were used as a model system to investigate how insulin achieves this effect. When HEP G2 cells were incubated with 100 nM insulin for 6, 14, or 24 h, IGFBP-1 protein levels in conditioned medium fell to approximately 50% of control values. This apparently was due to a fall in the rate of IGFBP-1 protein synthesis, since HEP G2 cells incorporated 46% less [35S]methionine into IGFBP-1 during a 4-h incubation with 100 nM insulin. IGFBP-1 mRNA levels were similarly affected by 100 nM insulin, falling to 45% of control values after 2 h, and to 9% of control values after 4 h of incubation with this hormone. The fall in IGFBP-1 mRNA level is consistent with data from nuclear transcription assays. HEP G2 nuclei isolated from cells that were incubated with 100 nM insulin for 2 h synthesized only approximately 1/3 the number of IGFBP-1 transcripts as did control nuclei. Further evidence that insulin decreases IGFBP-1 gene transcription comes from transient transfections using chimeric IGFBP-1 promoter-chloramphenicol acetyltransferase reporter gene constructs. IGFBP-1 promoter activity fell to approximately 50% of control values when HEP G2 cells transfected with a construct containing the first 1205 base pairs of the IGFBP-1 promoter were incubated with 100 nM insulin for 6, 14, or 24 h. Insulin lowered both IGFBP-1 protein levels and promoter activity in a dose-dependent manner. A half-maximal effect was found at approximately 1 nM insulin and a maximal effect was found at approximately 10 nM insulin in each instance. Transfections with constructs containing smaller IGFBP-1 promoter fragments showed that the region spanning from 103 to 529 base pairs 5' to the IGFBP-1 mRNA cap site was necessary to demonstrate the inhibitory effect of insulin. These studies indicate that insulin lowers IGFBP-1 protein levels, at least in part, by rapidly decreasing the rate of IGFBP-1 gene transcription, and suggest that this insulin-mediated fall in transcription is conferred through a specific region of the IGFBP-1 promoter.

The promoter of the human gene for insulin-like growth factor binding protein-1. Basal promoter activity in HEP G2 cells depends upon liver factor B1.

Characterization of the human insulin-like growth factor binding protein-1 (IGFBP-1) promoter was initiated to facilitate study of developmental and hormonal factors regulating IGFBP-1 production. The region immediately 5' to the IGFBP-1 mRNA capsite is typical of a eukaryotic promoter, with a TATA sequence beginning 28 base pairs (bp) and a CCAAT promoter element beginning 72 bp upstream from this capsite. A 1.3-kilobase insert containing the IGFBP-1 capsite and 1205 bp of this putative IGFBP-1 promoter region directs expression of the reporter gene chloramphenicol acetyltransferase (CAT) in an orientation-specific manner in transfected HEP G2 cells, and the capsite identified for the CAT mRNA is identical to that identified for native IGFBP-1 mRNA. These observations suggest that the 1.3-kilobase insert contains the IGFBP-1 promoter. This promoter was further characterized by deletion analysis, site-directed mutagenesis, gel mobility shift assays, and DNaseI protection assays. These studies identify the CCAAT box region as the major cis element involved in basal IGFBP-1 promoter activity in HEP G2 cells, demonstrate that increased basal promoter activity is associated with the binding of at least one HEP G2 nuclear factor to the CCAAT box region, and indicate that the DNA binding factor(s) responsible for increased basal promoter activity is related to liver factor B1. These observations suggest that liver B1 is the major trans-acting factor stimulating basal IGFBP-1 promoter activity in HEP G2 cells.