Phosphorylation and inactivation of glycogen synthase kinase 3 by protein kinase A

Abstract

Glycogen synthase kinase 3 (GSK-3) is implicated in multiple biological processes including metabolism, gene expression, cell fate determination, proliferation, and survival. GSK-3 activity is inhibited through phosphorylation of serine 21 in GSK-3 alpha and serine 9 in GSK-3 beta. These serine residues of GSK-3 have been previously identified as targets of protein kinase B (PKB/Akt), a serine/threonine kinase located downstream of phosphatidylinositol 3-kinase. Here, we show that serine 21 in GSK-3 alpha and serine 9 in GSK-3 beta are also physiological substrates of cAMP-dependent protein kinase A. Protein kinase A physically associates with, phosphorylates, and inactivates both isoforms of GSK-3. The results indicate that depending on the stimulatory context, the activity of GSK-3 can be modulated either by growth factors that work through the phosphatidylinositol 3-kinase-protein kinase B cascade or by hormonal stimulation of G protein-coupled receptors that link to changes in intracellular cAMP levels.

Figure 1

PKA-dependent phosphorylation and inactivation of GSK-3 induced by cAMP. Subconfluent Rat1, NIH 3T3, and HEK293 cells were starved in a serum-free medium for 12–24 h and stimulated with 8-Br-cGMP or 8-Br-cAMP (2 mM, 30 min), forskolin (15 μM, 30 min), isoproterenol (10 μM, 30 min), or IGF-1 (75 ng/ml, 10 min). Cells were stimulated with forskolin and isoproterenol in the presence of 0.1 mM of 3-isobutyl-1-methylxanthine to inhibit cellular phosphodiesterase activity. The cell-permeable PKA inhibitor H89 (10 μM) was added to culture where indicated 1 h before stimulation. Cell lysates were prepared and analyzed for GSK-3 phosphorylation at serine 21 and 9 by using GSK-3α and -β phospho-specific antibodies and for GSK-3 kinase assays as described in Materials and Methods. (A) 8-Br-cAMP, forskolin, and isoproterenol induce phosphorylation and inhibition of GSK-3 in a PKA-dependent manner in Rat1 cells. GSK-3 kinase activities are presented as percentage of the activity in unstimulated cells. The results are means ± SD of three independent experiments. (B) 8-Br-cAMP and forskolin stimulate GSK-3 phosphorylation in a PKA-dependent manner in NIH 3T3 and HEK293 cells.

Figure 2

Dissociation of PKA-mediated phosphorylation and inactivation of GSK-3 from a functional PI3K–PKB signaling pathway. (A) PKA-mediated phosphorylation of GSK-3 is independent of PI3K activity. After 12–24 h of incubation in serum-free medium, Rat1 and NIH 3T3 cells were stimulated with 8-Br-cAMP (2 mM, 30 min), IGF-1 (75 ng/ml, 10 min, Rat1), or insulin (0.1 μM, 10 min, NIH 3T3) in the absence or presence of the PI3K inhibitors, wortmannin (125 nM) or LY294002 (12.5 μM). The cells were preincubated with wortmannin or LY294002 for at least 1 h before addition of 8-Br-cAMP, IGF-1, or insulin. Phosphorylation of GSK-3α at serine 21 and GSK-3β at serine 9 was analyzed by immunoblotting with GSK-3 phospho-specific antibodies as in Fig. 1. (B) Elevation of cAMP levels does not increase phosphorylation of PKB at serine 473. Rat1 and NIH 3T3 cells were starved and stimulated with 8-Br-cGMP, 8-Br-cAMP, forskolin, isoproterenol, IGF, or insulin, as detailed in Fig. 1. PKB phosphorylation levels were analyzed by immunoblotting with a PKB phospho-specific antibody that recognizes PKB phosphorylated at serine 473 (New England Biolabs). Reprobing with a PKB antibody reactive with total PKB (New England Biolabs) shows similar loading among samples. (C) 8-Br-cAMP or elevation of endogenous cAMP levels induces no or limited increases in PKB activity. Rat1 and NIH 3T3 cells were stimulated as in B. PKB activity toward the crosstide–paramyosin substrate was determined as detailed in Materials and Methods. The crosstide–paramyosin substrate phosphorylated at a serine site corresponding to serine 21/9 of GSK-3 is indicated with crosstide-p. The intensities of the crosstide-p bands were quantified by densitometry. The values beneath each lane represent relative intensities (%) with bands induced by IGF-1 in Rat1 cells and by insulin in NIH 3T3 cells defined as 100%. For comparison with PKB activity, a fraction of each lysate was analyzed for levels of GSK-3 phosphorylation at serine 21 and 9 by immunoblotting with GSK-3α and -β phospho-specific antibodies (Bottom). Data shown are representative of three independent experiments.

Figure 3

Physical association between PKA and GSK-3. (A) HEK293, Rat1, and NIH 3T3 were transfected with an empty vector or GSK-3 vectors that express rat GSK-3α (pMT2-GSK3α) or HA epitope-tagged human GSK-3β (pcDNA3- HA-GSK3β). Two days after transfection, cells were starved in serum-free medium for 12 h before lysing. GSK-3 was immunoprecipitated (IP) by using isoform-selective GSK-3α or -β antibodies as described in Materials and Methods. The immunocomplex was analyzed by Western blotting (WB) with an anti-PKAc antibody. The membranes were reprobed with an anti-GSK-3 antibody (reactive with both isoforms) to confirm the efficiency and specificity of immunoprecipitation. The locations of PKAc, GSK-3α, GSK-3β, HA-GSK-3β, and the Ig heavy chain (IgH) are indicated. (B) Lysates from vector or HA-GSK-3β-transfected HEK293 cells were immunoprecipitated with anti-HA monoclonal antibody. The immunocomplex was subjected to Western blotting with an anti-PKAc antibody and then with an anti-GSK-3 antibody as in A. In the reciprocal experiment (Right), PKAc was immunoprecipitated from the same lysates. The immunocomplex was analyzed by Western blotting for GSK-3 and then for PKAc. (C) Untransfected HEK293 and NIH 3T3 cells were starved in serum-free medium for 12–24 h and stimulated with forskolin or vehicle (control) as described in Fig. 1. GSK-3β or PKA RII was immunoprecipitated and the immunocomplex subjected to Western blotting with the indicated antibodies.

Figure 4

In vivo and in vitro phosphorylation of GSK-3 by PKA. (A) Transfection of PKAc induces GSK-3 phosphorylation at serine 21 and 9 in intact cells. HEK293 and NIH 3T3 cells were cotransfected with pcDNA3-GSK-3β (labeled as GSK-3β) or pMT2-GSK-3α (GSK-3α) along with an empty vector (Vector) or pcDNA3-PKAc (PKAc). Two days after transfection, cells were starved for 12 h in serum-free medium and lysates prepared. Total cell lysates were analyzed for phosphorylation of transfected and endogenous GSK-3 by immunoblotting with GSK-3α and -β phospho-specific antibodies. Expression of cellular and transfected GSK-3α or -β was determined by immunoblotting with an anti-GSK-3 antibody. The relative locations of cellular and transfected GSK-3α or -β are indicated (Right). (B) PKAc phosphorylates GSK-3 at serine 21 and 9 in vitro. Immunoprecipitated HA-GSK-3β (Top Left), immunoprecipitated GSK-3α (Top Right), recombinant GSK-3β (Bottom Left), and purified GSK-3α (Bottom Right) were used as substrates for PKAc in a kinase reaction. The immunoprecipitates or recombinant/purified GSK-3α or -β were incubated without PKA (No PKA), with PKA (PKA), or with PKA in the presence of the PKA inhibitor PKI (PKA + PKI). Phosphorylation of the GSK-3α or -β substrates by PKA was determined by immunoblotting with GSK-3α and -β phospho-specific antibodies. Levels of GSK-3α or -β substrates were assessed by immunoblotting with an anti-GSK-3 antibody. Bands corresponding to phosphorylated GSK-3α or -β, total GSK-3α or -β, and the Ig heavy chain (IgH) are indicated (Right).