Insulin Inhibits the Activation of Transcription by a C-terminal Fragment of the Forkhead Transcription Factor FKHR

The forkhead rhabdomyosarcoma transcription factor (FKHR) is a promising candidate to be the transcription factor that binds to the insulin response element of the insulin-like growth factor-binding protein-1 (IGFBP-1) promoter and mediates insulin inhibition of IGFBP-1 promoter activity. Cotransfection of mouse FKHR increased IGFBP-1 promoter activity 2–3-fold in H4IIE rat hepatoma cells; insulin inhibited FKHR-stimulated promoter activity ∼70%. A C-terminal fragment of mouse FKHR (residues 208–652) that contains the transcription activation domain fused to a Gal4 DNA binding domain potently stimulated Gal4 promoter activity. Insulin inhibited FKHR fragment-stimulated promoter activity by ∼70%. Inhibition was abolished by coincubation with the phosphatidylinositol-3 kinase inhibitor, LY294002. The FKHR 208–652 fragment contains two consensus sites for phosphorylation by protein kinase B (PKB)/Akt, Ser-253 and Ser-316. Neither site is required for insulin inhibition of promoter activity stimulated by the FKHR fragment, and overexpression of Akt does not inhibit FKHR fragment-stimulated Gal4 promoter activity. These results suggest that insulin- and phosphatidylinositol-3 kinase-dependent phosphorylation of another site in the fragment by a kinase different from PKB/Akt inhibits transcription activation by the fragment. Phosphorylation of this site also may be involved in insulin inhibition of transcription activation by full-length FKHR, but only after phosphorylation of Ser-253 by PKB/Akt.

The insulin-like growth factor-binding proteins (IGFBPs) 1 are a family of six proteins that bind IGF-I and IGF-II with high affinity and modulate their growth-promoting and other biological activities (1)(2)(3)(4). IGFBP-1 is distinctive among the IGFBPs in that its plasma levels are dynamically regulated in response to hormonal or metabolic changes, being increased in diabetes and fasting, and normalized by insulin treatment or refeeding, respectively (5). IGFBP-1 is thought to play a role in glucose homeostasis by inhibiting the insulin-like effects of IGF-I and IGF-II (5)(6)(7).
Insulin is the principal regulator of IGFBP-1 (5). It acts primarily at the level of transcription, rapidly inhibiting IGF-BP-1 transcription in diabetic rat liver (8) and H4IIE rat hepatoma cells (9,10). An insulin-response element (IRE), T(G/ A)TTTTG, was identified in the proximal promoter of the human and rat IGFBP-1 genes that is necessary for insulin inhibition of basal and dexamethasone-stimulated IGFBP-1 promoter activity (11)(12)(13)(14)(15). Two copies of the IRE are present as an inverted palindrome in IGFBP-1. Similar IREs have been identified in other genes whose expression is negatively regulated by insulin. Single copies of the IRE are present in the phosphoenolpyruvate carboxykinase, tyrosine aminotransferase, and apolipoprotein CIII genes (reviewed in Refs. 16 and 17); the glucose-6-phosphatase gene has three copies (18).
When the present studies were begun, the transcription factor that binds to the IRE and mediates insulin inhibition of IGFBP-1 promoter activity had not been identified. One candidate, the forkhead transcription factor hepatocyte nuclear factor-3␤, bound to the IRE and stimulated IGFBP-1 promoter activity in NIH-3T3 cells (19), but binding of hepatocyte nuclear factor-3␤ to mutated IREs did not correlate with the effect of these mutations on insulin inhibition of promoter activity (17,20). An important clue to its identity came from recent studies in Caenorhabditis elegans. The inhibition of IGFBP-1 gene expression by insulin in rat hepatocytes and HepG2 human hepatocarcinoma cells appears to be mediated by a signal transduction pathway that involves the lipid kinase, phosphatidylinositol-3 kinase (PI-3 kinase) (21), and the serinethreonine protein kinase, protein kinase B (PKB)/Akt (22). In C. elegans, the transcription factor DAF-16 is negatively regulated by a signaling pathway that includes an insulin-like receptor (DAF-2), the p110 catalytic subunit of PI-3 kinase (AGE-1), and PKB/Akt (AKT1/AKT2) (23)(24)(25)(26). Mutations in DAF-2, AGE-1, or AKT-1/AKT-2 interrupt the normal developmental program, causing growth arrest, a metabolic shift to fat storage, increased longevity, and suppressed reproduction. Mutations in DAF-16, which by themselves have no effect, can reverse the effects of mutations in these genes, restoring normal growth, metabolism, reproduction, and lifespan. Daf-16 is a member of the forkhead rhabdomyosarcoma (FKHR) subfam-ily of the hepatocyte nuclear factor-3/forkhead ("winged helix") family of transcription factors (27,28). It is homologous to three human transcription factors: FKHR (29), FKHR-L1/AF6q21 (30,31), and AFX (32,33).
Regulation of DAF-16 by a PI-3 kinase-PKB/Akt signaling pathway in C. elegans raised the possibility that an FKHR-like protein also might mediate insulin inhibition of IGFBP-1 promoter activity in mammalian liver. Using a mouse FKHR cDNA clone that is orthologous to human FKHR (34), we show that mFKHR stimulates IGFBP-1 promoter activity in H4IIE rat hepatoma cells and that insulin inhibits this stimulation. A C-terminal mFKHR fragment (amino acids 208 -652), which contains the transactivation domain fused to a Gal4-DNA binding domain, stimulates Gal4 promoter activity; insulin also inhibits this stimulation. Insulin inhibition of Gal4 promoter activity stimulated by the mFKHR fragment is dependent on PI-3 kinase but does not require the two consensus PKB/Akt phosphorylation sites in the fragment and is not mimicked by constitutively active PKB/Akt.

EXPERIMENTAL PROCEDURES
Materials-LY294002, a selective inhibitor of PI-3 kinase (35), and PD98059, a selective inhibitor of mitogen-activated protein kinase kinase (36), were obtained from Calbiochem. All restriction enzymes were purchased from Life Technologies, Inc. Taq polymerase was obtained from PE Applied Biosystems (Foster City, CA). Bovine pancreatic insulin was purchased from Sigma; human recombinant insulin (Humulin U-100 regular) was obtained from Eli Lilly & Co. (Indianapolis, IN).
Cell Cultivation-H4IIE rat hepatoma cells (10) were grown as monolayer cultures in low glucose Dulbecco's minimal essential medium (Life Technologies Inc.) supplemented with 10% fetal bovine serum (Hyclone Laboratories, Logan, UT) and incubated in a humidified 95% air, 5% CO 2 atmosphere at 37°C. Cultures were passaged twice weekly (when they reached confluence) at a ratio of 1:10 using trypsin-EDTA. Fresh stocks were thawed after 8 -10 passages.
For Gal4 promoter assays (38), the reporter plasmid pFR-Luc (Stratagene) was used. It contains a CMV promoter, five copies of the Gal4 binding element, and a luciferase reporter. The activation domains to be tested were inserted into pFA-CMV-EcoRV, modified from pFA-CMV (Stratagene) by insertion of an EcoRV linker between the BamHI and XbaI sites. C-terminal fragments (208 -652) of mouse FKHR (wild-type or containing S253A or S316A mutations) were obtained by PCR amplification of the corresponding pCMV5-FKHR plasmids (34) using the sense primer, 5Ј-CGTAGATATCAATTCAATTCGCCACAATCTG-3Ј; and the antisense primer, 5Ј-GCTCTAGATTAGCCTGACACCCAGCT-GTG-3Ј. Shorter C-terminal FKHR fragments (256 -652 and 317-652) were generated using the sense primers 5Ј-CGTAGATATCAACAACA-GTAAATTTGCTAAGAGC-3Ј and 5Ј-CGTAGATATCAATGCTAGTAC-CATCAGTGGG-3Ј, respectively, and the same antisense primer. The PCR fragments were digested with EcoRV and XbaI and sequentially ligated into the EcoRV and XbaI sites of pFA-CMV-EcoRV to generate pFA-FKHR 208 -652, pFA-FKHR 256 -652, and pFA-FKHR 317-652 constructs. In some experiments, plasmids pM3 and pM3VP16 (CLON-TECH, Palo Alto, CA), containing the Gal4 DNA binding domain alone or fused to amino acids 411-455 of the herpes simplex viral protein, were transfected as controls. The sequence of all constructs was confirmed using a Rhodamine fluorescent terminator sequencing kit (PE Applied Biosystems). DNA sequence analysis was kindly performed by George Poy (NIDDK).
Transfection-H4IIE cells were transfected as described previously (39). The day before transfection, cells (1.75-8 million/dish) were plated onto 6-cm tissue culture dishes in 3 ml of Dulbecco's minimal essential medium containing 10% serum and were typically 70 -100% confluent at the time of transfection. DEAE dextran (Amersham Pharmacia Biotech) stock solution (2 mg/ml in 0.15 M NaCl) was diluted with an equal volume of Tris-buffered saline (25 mM Tris, pH 7.5, 137 mM NaCl, 5 mM KCl, 0.7 mM CaCl 2 , 0.5 mM MgCl 2 , 0.6 mM Na 2 HPO 4 ). Plasmid DNA (100 l) was mixed with 100 l of DEAE dextran-Tris-buffered saline and incubated at room temperature. After 15 min, 190 l of the mixture was added to each dish and, 15 min later, 3 ml of Dulbecco's minimal essential medium containing 10% serum was added and the incubation continued overnight. After medium change and a second overnight incubation, the medium was replaced with serum-free Dulbecco's minimal essential medium (containing 0.1% bovine serum albumin) with or without insulin (bovine insulin (1 g/ml) or recombinant human insulin (0.25 or 1 g/ml) as indicated). The cells were harvested for luciferase assay after a 24-h incubation.
Luciferase and ␤-Galactosidase Assays-Cells were washed twice with ice-cold phosphate-buffered saline. After the addition of 360 l of lysis buffer (25 mM glycylglycine, 15 mM MgSO 4 , 4 mM EGTA, 1 mM dithiothreitol, 1% Triton X-100), cells were scraped and centrifuged (4°C, 14,000 rpm, 15 min). The supernatant (100 l) was added to assay buffer (25 mM glycylglycine, 15 mM MgSO 4 , 4 mM EGTA, 15 mM potassium phosphate, pH 7.8, 1 mM dithiothreitol, 2 mM ATP) and assayed immediately after the addition of D-Luciferine (Biosynth, Naperville, IL) using a Lumat LB 9507 luminometer (EG&G Berthold, Germany). ␤-Galactosidase in cell lysates was assayed according to the manufacturer's instructions (Tropix, Bedford, MA). In later experiments, luciferase and ␤-galactosidase activity were measured in the same tube using the Dual Light chemiluminescent reporter gene assay kit (Tropix, PE Applied Biosystems) as specified by the manufacturer, allowing 1 h at room temperature for luciferin fluorescence to decay before measuring ␤-galactosidase activity. Assays were performed in duplicate.
Northern Blot Hybridization-Total RNA was prepared from H4IIE cells using the acid guanidine thiocyanate-phenol-chloroform method (40). The RNA was fractionated by electrophoresis on 1.5% agarose gels and blotted to nitrocellulose membranes (Amersham Pharmacia Biotech) as described previously (10). The blots were hybridized at 42°C (41) with a full-length rat IGFBP-1 cDNA (GenBank TM accession number M89791) probe prepared by PCR amplification (10). The probe was labeled by random priming using a Prim-It kit (Stratagene, La Jolla, CA). Autoradiographs and photographs of gels stained with ethidium bromide were scanned, and the images were analyzed using the software program NIH Image 1.61/ppc (NIH, Bethesda, MD).
Insulin Inhibits Transactivation of a Gal4 Promoter by a C-terminal Fragment of Mouse FKHR-(208 -652)-To determine the mechanism by which insulin inhibits FKHR-stimulated IGFBP-1 promoter activity, we used a yeast Gal4 promoter system (38,42) in which DNA binding is determined by recognition of a Gal4 binding element so that differences in Gal4 promoter activity are determined by the transcription activation domain that is fused to the Gal4 DNA binding domain. H4IIE cells were cotransfected with a reporter plasmid containing a luciferase gene coupled to five copies of the Gal4 binding element and activation plasmids in which the Gal4 DNA binding domain was fused to a C-terminal fragment of mouse FKHR (amino acids 208 -652) that contains the activation domain. This fragment corresponds to the fragment of the human FKHR gene (residues 211-655) that was identified in a human alveolar rhabdomyosarcoma fused to the DNA binding domain of Pax3 following chromosomal translocation (29). Expression of proteins containing the human FKHR fragment fused to the Pax3 (or Gal4) DNA binding domain stimulated promoter activity in different cell lines (43)(44)(45). Similarly, proteins in which the mouse FKHR C-terminal fragment was fused to the Gal4 DNA binding domain potently activated the Gal4 promoter in H4IIE cells (Fig. 2). Insulin inhibited promoter activity by 69 Ϯ 4% (S.D., n ϭ 4) in cells that were transfected with pFA-FKHR 208 -652, indicating that this fragment contains the necessary sites for insulin inhibition of FKHR-stimulated promoter activity.
LY294002 also abolished insulin inhibition of Gal4 promoter activity stimulated by the C-terminal FKHR fragment (Fig.  3C). In the absence of inhibitor, luciferase activity in H4IIE cells transfected with pFA-FKHR 208 -652 was 36 Ϯ 11% (S.D., n ϭ 6) of the activity in the absence of insulin. By contrast, no significant decrease in luciferase activity was observed following insulin treatment in cells that had been preincubated with LY294002 (103 Ϯ 16%; S.D., n ϭ 6). These results indicate that PI-3 kinase also is involved in insulin inhibition of Gal4 promoter activity stimulated by the C-terminal FKHR fragment.
Constitutively Active Akt Does Not Inhibit the Stimulation of Gal4 Promoter Activity by FKHR 208 -652-As PKB/Akt acts downstream of PI-3 kinase to mediate insulin inhibition of IGFBP-1 promoter activity (22), we examined whether cotransfection of constitutively active myr Akt would mimic insulin and inhibit the stimulation of Gal4 promoter activity by the Luciferase activity was determined in cell lysates and is expressed relative to the activity seen in cells transfected with pM3VP16 (and not incubated with insulin) in the same experiment (taken as 100%). The relative activity in the presence of insulin was calculated and plotted as described in Fig. 1. The mean Ϯ S.D. of four experiments is shown. C-terminal FKHR fragment (Fig. 4). Cotransfection of pCEFL-HA-myr-Akt with pFA-FKHR 208 -652 and the Gal4-luciferase reporter plasmid did not significantly decrease Gal4 promoter activity in the absence of insulin, compared with cotransfection with pCEFL empty vector or pCEFL-HA-Akt (wild-type), and promoter activity was inhibited to the same extent by insulin. By contrast, in parallel experiments, cotransfection of pCEFL-HA-myr-Akt markedly inhibited the stimulation of IGFBP-1 promoter activity by full-length FKHR. 3 Thus, overexpression of Akt inhibits transcription activation by full-length FKHR but not by the C-terminal FKHR fragment.

Neither of the Two Potential PKB/Akt Phosphorylation Sites in the C-terminal FKHR Fragment Are Required for the Inhibition of FKHR 208 -652 Stimulation of Gal4
Promoter Activity by Insulin-Two of the three potential PKB/Akt phosphorylation sites (46) in FKHR are present in the FKHR 208 -652 fragment, Ser-253 (RRRAAS*) and Ser-316 (RPRTSS*) (Fig. 5). Phosphorylation of both sites by a PI-3 kinase/PKB-Akt pathway has been demonstrated in vitro and in vivo (34,47). To determine whether these sites are required for insulin inhibi-3 pCEFL-myr-Akt inhibited IGFBP-1 promoter activity to a greater extent than maximally effective concentrations of insulin. This suggests that overexpression of Akt may act more efficiently or by additional mechanisms than insulin. Despite the marked inhibition of promoter activity by pCEFL-myr-Akt, insulin treatment resulted in further inhibition. This may reflect the fact that although myristylation localizes Akt to the plasma membrane where it can be phosphorylated by PDK-1, activation of PI-3 kinase by insulin enhances the activation of PDK-1, resulting in increased phosphorylation and activation of Akt (56, 57).  -length mFKHR, 3 g, lanes 1-3), or pFR-Luc (2 g) and pFA-FKHR 208 -652 (C-terminal fragment, 3 g, lanes 4 -6); pRSV-␤-galactosidase (40 ng); and pCEFL, pCEFL-HA-Akt, or pCEFL-HA-myr-Akt (3 g) as indicated. Cells were treated (ϩ, hatched bars) or not treated (Ϫ, solid bars) with human insulin (0.25 g/ml). Luciferase activity was determined and is normalized to the activity with pCEFL and the same reporter gene in each experiment (taken as 100%). The relative activity in the presence and absence of insulin ((ϩinsulin/Ϫinsulin) ϫ 100) was determined for each experiment, and activity in the presence of insulin was plotted as described in Fig. 1. The mean Ϯ S.D. of three to five experiments is shown.
alone (Control), or supplemented with LY294002 (50 M; LY). Thirty min later, human insulin (0.25 or 1 g/ml) was added to half of the dishes. Luciferase activity in cell lysates was determined. Relative luciferase activity ((ϩinsulin/Ϫinsulin) ϫ 100) is plotted for Control and LY-treated samples. Activity in the absence of insulin under the same experimental conditions is taken as 100%. The mean Ϯ S.D. for six experiments is plotted. In the absence of insulin, luciferase activity in cells treated with LY294002 was 167 Ϯ 34% of that in cells not treated with the inhibitor. tion of Gal4 promoter activity, we mutated Ser-253 or Ser-316 to Ala in the FKHR 208 -652 fragment. Neither mutation significantly reduced insulin inhibition compared with wild-type FKHR 208 -652. Insulin inhibition also was not affected when Ser-253 was deleted in the shorter C-terminal FKHR fragment (256 -652). These results indicate that Ser-253 and Ser-316 are not required for insulin inhibition of Gal4 promoter activity stimulated by C-terminal FKHR fragments. To exclude the possibility that one of the two sites might be sufficient for insulin inhibition with the second site compensating for the mutated or deleted site, we also examined two constructs in which both sites were deleted or mutated, FKHR 256 -652/ S316A and FKHR 317-652 (Fig. 5). Insulin inhibition still was observed. We conclude that neither PKB/Akt site is required for insulin inhibition of promoter activity in the C-terminal FKHR fragment. DISCUSSION FKHR or a closely related protein is a promising candidate to be the transcription factor in H4IIE cells that binds to the IRE and is responsible for the inhibition of IGFBP-1 promoter activity by insulin. Mouse FKHR stimulates IGFBP-1 promoter activity in H4IIE cells 2-3-fold. FKHR binds to the IGFBP-1 IRE, and an intact IRE is required for FKHR stimulation of IGFBP-1 promoter activity (48). FKHR-like mRNA was detected in H4IIE cells by reverse transcription PCR. 4 Insulin inhibited mFKHR-stimulated IGFBP-1 promoter activity in H4IIE cells by ϳ70%. During review and revision of this manuscript, several other groups reported similar stimulation of IGFBP-1 promoter activity by full-length FKHR or the related protein, AFX, and inhibition of FKHR-stimulated promoter activity by insulin or constitutively active PKB/Akt (48 -52).
To analyze the mechanism for insulin inhibition of FKHRstimulated IGFBP-1 promoter activity, we have studied the activation of a Gal4-luciferase reporter plasmid by a C-termi-nal fragment of mFKHR (amino acids 208 -652) fused to a Gal4 DNA binding domain. DNA binding is based on recognition of the Gal4 binding element. The C-terminal mFKHR fragment contains the transcription activation domain and part of the forkhead domain (residues 155-255). 5 As previously reported for the corresponding fragment of human FKHR (residues 211-655) (43)(44)(45), mFKHR 208 -652 potently stimulated Gal4 promoter activity.
Insulin inhibited the stimulation of Gal4 promoter activity by the mFKHR 208 -652 fragment by ϳ70%. This demonstrates that the C-terminal FKHR fragment contains the information necessary for insulin inhibition of FKHR-stimulated transcription. By contrast, insulin did not inhibit the stimulation of Gal4 promoter activity by a C-terminal fragment (comparable to mFKHR 256 -652) of another member of the FKHR subfamily, AFX, in NIH3T3 cells that overexpress insulin receptors (49). This difference may reflect the low conservation of nucleotide sequences in the transactivation domains of FKHR and AFX or the different cell contexts.
Inhibition of FKHR-stimulated gene transcription in H4IIE cells by insulin is initiated by binding to the insulin receptor; H4IIE cells lack IGF-I receptors with which insulin might cross-react (9,53). Insulin inhibition of IGFBP-1 transcription does not require new protein synthesis (10), indicating that inhibition results from post-translational events such as altering the phosphorylation state of FKHR. PI-3 kinase appears to be involved in transmission of the signal from the insulin receptor to IGFBP-1. The PI-3 kinase inhibitor, LY294002, greatly reduced insulin inhibition of IGFBP-1 mRNA abundance, as previously seen in rat hepatocytes (21) and HepG2 cells (22). LY294002 also abolished insulin inhibition of mFKHR fragment-stimulated Gal4 promoter activity, indicating that insulin regulation of promoter activity also was PI-3 kinase-dependent. Mitogen-activated protein kinase activation was not involved.
PKB/Akt is the best characterized downstream effector of 4 Reverse transcription PCR of H4IIE mRNA using primers from the 3Ј-end of mouse FKHR amplified a fragment of the predicted size (M. Tomizawa, unpublished results). This region is more highly conserved in human FKHR (29) than in the other two human members of the FKHR subfamily, FKHR-L1 (30) and AFX (32). 5 The complete forkhead domain must be present, however, to form the winged-helix conformation that is necessary for DNA binding (61,62). PI-3 kinase (54 -56). Overexpression of PKB/Akt inhibited IGF-BP-1 promoter activity in HepG2 cells, and a kinase-dead mutant of PKB/Akt blocked insulin inhibition (22), suggesting that insulin inhibition of IGFBP-1 expression was mediated by PI-3 kinase activation of PKB/Akt. PKB/Akt is translocated to the nucleus of cells that had been incubated with IGF-I (57), raising the possibility that the kinase might directly phosphorylate FKHR. The FKHR-related proteins have three conserved PKB/ Akt consensus phosphorylation sites (RXRXXS/T) (46) located at sites equivalent to Thr-24, Ser-253, and Ser-316 of mFKHR. These sites become phosphorylated after incubation with insulin or IGF-I, or overexpression of PKB/Akt (34,47,58). Mutation of all three PKB/Akt phosphorylation sites in FKHR abolished the inhibition of FKHR-stimulated activity of promoters containing one or more copies of the IRE by insulin or myr Akt (50,51). These results suggest that phosphorylation of one or more of the PKB/Akt sites in full-length FKHR, most likely by PKB/ Akt, decreases the stimulation of promoter activity by FKHR.
Our results with the C-terminal mFKHR 208 -652 fragment are quite different. Neither of the two consensus PKB/Akt phosphorylation sites in the fragment, Ser-253 or Ser-316, is required for insulin inhibition of promoter activity. Mutation or deletion of one or both of these sites did not affect insulin inhibition, indicating that insulin inhibits the stimulation of Gal4 promoter activity by C-terminal mFKHR fragments by modifying sites other than the two PKB/Akt sites in the fragment. Moreover, overexpression of PKB/Akt by transfection of myr Akt did not significantly decrease Gal4 promoter activity by mFKHR 208 -652 in the absence of insulin compared with transfection with wild-type Akt or empty vector. By contrast, in the same experiments, cotransfection with myr Akt profoundly decreased the stimulation of IGFBP-1 promoter activity by full-length mFKHR. These results suggest that, in contrast to full-length mFKHR, insulin inhibition of transactivation by the C-terminal FKHR fragment does not involve the two PKB/Akt consensus phosphorylation sites or phosphorylation by PKB/Akt.
It is tempting to speculate that the mechanism responsible for insulin inhibition of transcription by the C-terminal FKHR fragment also may be involved in the inhibition of transcription by full-length FKHR. We propose that the presumed phosphorylation site (not Ser-253 or Ser-316) that is responsible for insulin inhibition of transcription activation by FKHR 208 -652 is accessible to an insulin-responsive kinase, whereas this site only becomes accessible to the kinase in full-length FKHR after a cascade of phosphorylations (probably initiated by phosphorylation of Ser-253 by PKB/Akt (34,47)) and associated conformational changes. The insulin-responsive kinase is PI-3 kinase-dependent but distinct from PKB/Akt. Consistent with this proposal, Nakae et al. 6 have shown that Ser-253 is the gatekeeper for phosphorylation of Thr-24 and Ser-316 in fulllength mFKHR and that Thr-24 is phosphorylated by a different kinase than Ser-253. Insulin inhibition of phosphoenolpyruvate carboxykinase gene expression also requires a downstream effector of PI-3 kinase that is distinct from PKB/ Akt, the atypical protein kinase C, and the small GTPase Rac (59). Studies are in progress to identify the site in the mFKHR 208 -652 fragment whose phosphorylation in response to insulin decreases transcriptional activity and the kinase that is responsible for this phosphorylation.
The molecular mechanism for insulin inhibition of FKHRstimulated promoter activity remains unclear. FKHR-L1 (58) and mFKHR (52) are exported from the nucleus to the cytoplasm following incubation with IGF-I or overexpression of PKB/Akt. Insulin causes a similar redistribution of mFKHR in hepatocytes. 6 Export of FKHR to the cytoplasm could account for the decrease in FKHR-stimulated promoter activity after treatment with insulin or IGF-I. We do not know whether the C-terminal mFKHR fragments used in the present study undergo a similar redistribution after incubation with insulin. Even if FKHR fragments are exported to the cytoplasm following insulin treatment as is full-length FKHR, insulin still may induce a parallel direct inhibition of FKHR-stimulated transcription. Transcriptional activity of Pho4, a transcription factor in budding yeast that activates the expression of genes in response to phosphate starvation, is determined by four distinct phosphorylation sites that regulate nuclear import, nuclear export, and transcription activation (60). To completely shut off Pho4-mediated transcription, phosphorylation of both the sites that regulate nuclear localization and the site that regulates transactivation is required, indicating that several independent mechanisms play essential roles in regulating the transcriptional activity of Pho4.