Differential regulation of gene expression by insulin and IGF-1 receptors correlates with phosphorylation of a single amino acid residue in the forkhead transcription factor FKHR

Abstract

The transcription factor FKHR is inhibited by phosphorylation in response to insulin and IGF-1 through Akt kinase. Here we show that FKHR phosphorylation in hepatocytes conforms to a hierarchical pattern in which phosphorylation of the Akt site at S(253), in the forkhead DNA binding domain, is a prerequisite for the phosphorylation of two additional potential Akt sites, T(24) and S(316). Using insulin receptor-deficient hepatocytes, we show that T(24) fails to be phosphorylated by IGF-1 receptors, suggesting that this residue is targeted by a kinase specifically activated by insulin receptors. Lack of T(24) phosphorylation is associated with the failure of IGF-1 to induce nuclear export of FKHR, and to inhibit expression of a reporter gene under the transcriptional control of the IGF binding protein-1 insulin response element. We propose that site-specific phosphorylation of FKHR is one of the mechanisms by which insulin and IGF-1 receptors exert different effects on gene expression.

Fig. 1. Insulin-dependent phosphorylation of wild-type and mutant FKHR constructs in SV40-transformed murine hepatocytes. (A) Deletion constructs utilized in the present study. The consensus Akt sites are indicated in the single-letter amino acid code. The phosphorylated or mutant residue is underlined. The forkhead domain is represented by the shaded box. (B) Replacement of S253D partially restores insulin-dependent phosphorylation of FKHR. Normal mouse hepatocytes were transiently transfected with wild-type and mutant FKHR constructs, and insulin-induced phosphorylation was studied using ³²P labeling. In some experiments, cells were treated with wortmannin prior to the addition of insulin (right hand panel). In the upper panel, following immunoprecipitation with anti-c–myc antibody and SDS–PAGE, proteins were transferred to nylon membranes for autoradiography. The filters were then re-probed with an anti-FKHR antiserum to normalize the protein content of each lane (lower panel). (C) T²⁴ and S³¹⁶ are phosphorylated in S253D mutant FKHR. Experiments were performed as indicated above, except that double and triple mutants were used, as shown at the top of each panel.

Fig. 1. Insulin-dependent phosphorylation of wild-type and mutant FKHR constructs in SV40-transformed murine hepatocytes. (A) Deletion constructs utilized in the present study. The consensus Akt sites are indicated in the single-letter amino acid code. The phosphorylated or mutant residue is underlined. The forkhead domain is represented by the shaded box. (B) Replacement of S253D partially restores insulin-dependent phosphorylation of FKHR. Normal mouse hepatocytes were transiently transfected with wild-type and mutant FKHR constructs, and insulin-induced phosphorylation was studied using ³²P labeling. In some experiments, cells were treated with wortmannin prior to the addition of insulin (right hand panel). In the upper panel, following immunoprecipitation with anti-c–myc antibody and SDS–PAGE, proteins were transferred to nylon membranes for autoradiography. The filters were then re-probed with an anti-FKHR antiserum to normalize the protein content of each lane (lower panel). (C) T²⁴ and S³¹⁶ are phosphorylated in S253D mutant FKHR. Experiments were performed as indicated above, except that double and triple mutants were used, as shown at the top of each panel.

Fig. 1. Insulin-dependent phosphorylation of wild-type and mutant FKHR constructs in SV40-transformed murine hepatocytes. (A) Deletion constructs utilized in the present study. The consensus Akt sites are indicated in the single-letter amino acid code. The phosphorylated or mutant residue is underlined. The forkhead domain is represented by the shaded box. (B) Replacement of S253D partially restores insulin-dependent phosphorylation of FKHR. Normal mouse hepatocytes were transiently transfected with wild-type and mutant FKHR constructs, and insulin-induced phosphorylation was studied using ³²P labeling. In some experiments, cells were treated with wortmannin prior to the addition of insulin (right hand panel). In the upper panel, following immunoprecipitation with anti-c–myc antibody and SDS–PAGE, proteins were transferred to nylon membranes for autoradiography. The filters were then re-probed with an anti-FKHR antiserum to normalize the protein content of each lane (lower panel). (C) T²⁴ and S³¹⁶ are phosphorylated in S253D mutant FKHR. Experiments were performed as indicated above, except that double and triple mutants were used, as shown at the top of each panel.

Fig. 2. The C-terminal half of FKHR contains negative regulatory elements for FKHR phosphorylation by insulin. Two truncated mutants were employed for these experiments, one encompassing amino acids 1–255 (Δ256) and one encompassing amino acids 1–317 (Δ318). Additional mutations were introduced as indicated at the top of each lane (see also Figure 1A). Phosphorylation experiments and quantification of the data were carried out as indicated in the legend to Figure 1.

Fig. 3. (A) Different patterns of FKHR phosphorylation are mediated by insulin and IGF–1. Wild-type and mutant FKHR as indicated in Figure 1A were transfected into SV40-transformed hepatocytes derived from either normal mice (empty bars) or IR-deficient mice (full bars). Each experiment was repeated a minimum of three times. The data are expressed as the mean ± SD of the fold increase over basal phosphorylation levels in the absence of insulin (control hepatocytes) or IGF–1 (IR-deficient hepatocytes), and were quantitated using scanning densitometry and the NIH Image 1.60 software. The difference between the effect of insulin to phosphorylate the S253D/S316A mutant in WT and IR-deficient hepatocytes is statistically significant (P <0.05 analyzed by one-factor ANOVA). (B) T²⁴ is not phosphorylated in response to IGF–1 treatment in IR-deficient hepatocytes. The mutants employed for these experiments are indicated at the top of each lane. A representative experiment is shown.

Fig. 3. (A) Different patterns of FKHR phosphorylation are mediated by insulin and IGF–1. Wild-type and mutant FKHR as indicated in Figure 1A were transfected into SV40-transformed hepatocytes derived from either normal mice (empty bars) or IR-deficient mice (full bars). Each experiment was repeated a minimum of three times. The data are expressed as the mean ± SD of the fold increase over basal phosphorylation levels in the absence of insulin (control hepatocytes) or IGF–1 (IR-deficient hepatocytes), and were quantitated using scanning densitometry and the NIH Image 1.60 software. The difference between the effect of insulin to phosphorylate the S253D/S316A mutant in WT and IR-deficient hepatocytes is statistically significant (P <0.05 analyzed by one-factor ANOVA). (B) T²⁴ is not phosphorylated in response to IGF–1 treatment in IR-deficient hepatocytes. The mutants employed for these experiments are indicated at the top of each lane. A representative experiment is shown.

Fig. 4. Subcellular localization of phosphorylation site mutant FKHRs in insulin-treated SV40-transformed hepatocytes. Following transient transfection, hepatocytes were seeded into 4-well slide culture chambers and incubated in serum-free medium for 4 h prior to the addition of insulin. Myc epitope-tagged FKHR was visualized with anti-myc monoclonal antibody and FITC-conjugated anti-mouse IgG. (A and B) Wild-type FKHR; (C and D) S253A mutant; (E and F) S253D mutant; (G and H) S316A mutant; (I and J) T24A mutant; (K and L) T24D mutant.

Fig. 5. Role of T²⁴ in subcellular localization of FKHR. Experi- ments were performed as indicated in the legend to Figure 4. (A and B) T24A/S253D double mutant; (C and D) T24A/S253D/S316D triple mutant; (E and F) T24D/S253D/S316D triple mutant; (G and H) T24A/S253D/S316A triple mutant.

Fig. 6. Subcellular localization of an S253D/S316A double mutant FKHR in control and IR-deficient hepatocytes. Experiments were performed as indicated in the legend to Figure 4, except that IR-deficient hepatocytes were treated with IGF–1 to induce nuclear export of FKHR. (A and B) Nuclear exclusion of wild-type FKHR in insulin-treated SV40-transformed hepatocytes from normal mice; (C and D) nuclear exclusion of wild-type FKHR in IGF–1-treated IR-deficient hepatocytes; (E and F) nuclear exclusion of S253D/S316A FKHR in insulin-treated control hepatocytes; (G and H) nuclear exclusion of S253D/S316A FKHR in IGF–1-treated IR-deficient hepatocytes.

Fig. 7. Expression of an IGFBP–1/luciferase reporter gene in hepatocytes co-transfected with wild-type and S253D/S316A mutant FKHR. Control and IR-deficient hepatocytes were transfected with wild-type or mutant FKHR in addition to the IGFBP–1/luciferase reporter gene (p925GL3) (Ooi et al., 1992). Plasmid pRSV-β–galactosidase was used as an internal control of transfection efficiency (Ooi et al., 1992). After transfection, β–galactosidase and luciferase activity were measured in the same sample as described in Materials and methods. The data represent the mean ± SD from five independent experiments.

Fig. 8. Proposed model for multisite regulation of FKHR function. Under conditions in which insulin signaling is inactive, FKHR localizes to the nucleus. Following insulin stimulation, Akt translocates to the nucleus and phosphorylates S²⁵³ in the forkhead domain of FKHR. This primes FKHR for further phosphorylation of T²⁴ and S³¹⁶. T²⁴ phosphorylation is effected by an IR-specific kinase and is required for FKHR nuclear export. S³¹⁶ phosphorylation is effected by both IR and IGF–1R-stimulated kinases, and is not a requisite for FKHR nuclear export. Once FKHR is phosphorylated on T²⁴ and exported to the cytoplasm, pT²⁴ has the potential to bind 14-3-3 and cause cytoplasmic retention of FKHR (Brunet et al., 1999). The T²⁴ and S³¹⁶ kinases are also PIP₃ dependent, like Akt. Nevertheless, it is unlikely that the T²⁴ kinase is Akt, because Akt is activated to the same extent in control and IR-deficient hepatocytes (B.C.Park et al., 1999), and T²⁴ fails to be phosphorylated in IR-deficient hepatocytes.