A novel function of eIF2alpha kinases as inducers of the phosphoinositide-3 kinase signaling pathway

Abstract

Phosphoinositide-3 kinase (PI3K) plays an important role in signal transduction in response to a wide range of cellular stimuli involved in cellular processes that promote cell proliferation and survival. Phosphorylation of the alpha subunit of the eukaryotic translation initiation factor eIF2 at Ser51 takes place in response to various types of environmental stress and is essential for regulation of translation initiation. Herein, we show that a conditionally active form of the eIF2alpha kinase PKR acts upstream of PI3K and turns on the Akt/PKB-FRAP/mTOR pathway leading to S6 and 4E-BP1 phosphorylation. Also, induction of PI3K signaling antagonizes the apoptotic and protein synthesis inhibitory effects of the conditionally active PKR. Furthermore, induction of the PI3K pathway is impaired in PKR(-/-) or PERK(-/-) mouse embryonic fibroblasts (MEFs) in response to various stimuli that activate each eIF2alpha kinase. Mechanistically, PI3K signaling activation is indirect and requires the inhibition of protein synthesis by eIF2alpha phosphorylation as demonstrated by the inactivation of endogenous eIF2alpha by small interfering RNA or utilization of MEFs bearing the eIF2alpha Ser51Ala mutation. Our data reveal a novel property of eIF2alpha kinases as activators of PI3K signaling and cell survival.

Figure 1.

Conditional activation of GyrB.PKR induces the phosphorylation and activation of Akt/PKB. (A–D) Protein extracts (50 μg) from serum-deprived HT1080 cells expressing either GyrB.PKR or GyrB.PKRK296H, untreated or treated with coumermycin (Coum; 100 ng/ml) for up to 12 or 24 h, were subjected to immunoblotting with antibodies against the indicated proteins. The ratio of phosphorylated to total protein is indicated. The ratio was set to 1 for each cell line in the absence of coumermycin treatment. The data represent one of four reproducible experiments.

Figure 2.

GyrB.PKR acts upstream of PI3K. (A–C) Serum-starved HT1080 cells expressing GyrB.PKR were left untreated or treated with coumermycin (100 ng/ml) in the absence or presence of (A) LY294002 (LY; 20 μM), (B) rapamycin (Rapa; 20 nM), or (C) wortmannin (Wort; 100 nM) for the indicated times. Protein extracts (50 μg) were subjected to immunoblotting with antibodies against the indicated proteins. The data represent one of three reproducible experiments.

Figure 3.

Activation of PI3K by GyrB.PKR. (A) Serum-starved GyrB.PKR or GyrB.PKRK296H-expressing cells were left untreated or treated with coumermycin (100 ng/ml) for 6 h. Protein extracts (500 μg) were subjected to immunoprecipitation with an anti-PI3K p85 antibody followed by an in vitro lipid kinase assay in the presence of [³²P-γ]ATP and phosphatidylinositol (PI) as a substrate. Radioactive PIP was visualized by TLC and autoradiography (a). P85 levels in the immunoprecipitates were detected by immunoblotting (b). The data represent one of two reproducible experiments. (B) GyrB.PKR-expressing cells were transiently transfected with scrambled control siRNA (SCR) or siRNA targeting the p85 subunit of PI3K for 72 h. (C) GyrB.PKR cells were transiently transfected with pCMV plasmid lacking or containing FLAG-tagged PTEN for 24 h. (D) GyrB.PKR cells were infected with adenovirus expressing a dominant negative mutant of the p85 subunit (dnp85) of PI3K or a control adenovirus for 24 h. (B–D) Transfected or infected cells were left untreated or treated with coumermycin (100 ng/ml) for 6 h, and protein extracts (30 μg) were subjected to immunoblotting for the indicated proteins. The data represent one of three reproducible experiments.

Figure 4.

Regulation of 4E-BP1 phosphorylation and its implication in cap-dependent translation in GyrB.PKR cells. (A and B) Serum-deprived GyrB.PKR cells were left untreated or treated with coumermycin (100 ng/ml) for 6 h, in the absence or presence of rapamycin (20 nM) and/or LY294002 (20 μM), as indicated. (A) Protein extracts (50 μg) were subjected to immunoblotting with antibodies against the indicated proteins. The data represent one of three reproducible experiments. (B) Cells expressing either GyrB.PKR (□) or GyrB.PKRK296H (■) were incubated in media lacking methionine and supplemented with 10% dialyzed fetal bovine serum for 1 h. Cells were then treated with LY294002 or rapamycin for 1 h followed by the addition of coumermycin for 4 h. Subsequently, [³⁵S]methionine was added to cells for a further 2 h followed by the quantification of radioactive TCA precipitates. (C, coumermycin; R, rapamycin; LY, LY294002). One hundred percent (%) protein synthesis represents the average value of ³⁵S-labeled proteins in untreated GyrB.PKR or GyrB.PKRK296H-expressing cells. Values represent the average of three separate experiments performed in triplicate.

Figure 5.

Control of PKR-mediated apoptosis by PI3K pathway. HT1080 cells expressing GyrB.PKR were left untreated or treated with coumermycin (100 ng/ml) in the absence or presence of either LY294002 (20 μM) or rapamycin (20 nM) for 24 h. Cells were harvested, fixed in ethanol, stained with PI, and subjected to flow cytometry analysis. The percentage (%) of apoptotic cells or cells in various phases of the cell cycle is indicated. Data represent one of four reproducible experiments.

Figure 6.

PKR or PERK mediates the induction of PI3K pathway in response to IFN, dsRNA or ER stress respectively. (A) PKR^+/+ and PKR^−/− MEFs were treated with mouse IFN-γ (100 IU/ml) for the indicated time points. Protein extracts (50 μg) were subjected to immunoblotting against the indicated proteins. (B and C) PKR^+/+ and PKR^−/− MEFs were transfected with dsRNA (10 μg/ml) for the indicated time points. Protein extracts (50 μg) were subjected to immunoblotting against the indicated proteins. (D) PERK^+/+ and PERK^−/− MEFs were treated with thapsigargin (TG; 1 μM) in the presence of serum for the indicated time points. Protein extracts (50 μg) were subjected to immunoblotting with phosphospecific antibodies against the indicated proteins. (A–D) Data represent one of three reproducible experiments.

Figure 7.

Induction of PI3K signaling by eIF2α kinases requires eIF2α phosphorylation. (A) GyrB.PKR cells were left untransfected or transfected with siRNA for the luciferase reporter gene (Luc; negative control) or siRNA for eIF2α for 72 h followed by treatment with 100 ng/ml coumermycin for 6 h. (B) GyrB.PKR and GyrB.PKRT487D cells were left untreated or treated with coumermycin for the indicated times. (C) HT1080 cells were left untreated or pretreated with 20 μM LY294002 for 1 h before treatment with 75 μM Sal003 for the indicated times. (D) eIF2α S/S and eIF2α A/A MEFs were left untreated or treated with thapsigargin (TG; 1 μM) for indicated times. (A–D) Protein extracts (50 μg) were subjected to immunoblotting against the indicated proteins. The data represents one out of three reproducible experiments.

Figure 8.

Model of PI3K activation by eIF2α kinases. Activation of the eIF2α kinases (namely PERK or PKR) leads to the translational inhibition of a protein (X) that negatively regulates PI3K activity. This leads to the activation of the downstream components of PI3K signaling as documented in this article. As explained in the Discussion, the negative regulator may act either directly on PI3K or indirectly by suppressing the activity of upstream activators of PI3K.