Research ArticleBiochemistry

Akt Determines Cell Fate Through Inhibition of the PERK-eIF2α Phosphorylation Pathway

See allHide authors and affiliations

Science Signaling  27 Sep 2011:
Vol. 4, Issue 192, pp. ra62
DOI: 10.1126/scisignal.2001630

Abstract

Metazoans respond to various forms of environmental stress by inducing the phosphorylation of the α subunit of eukaryotic translation initiation factor 2 (eIF2α) at serine-51, a modification that leads to global inhibition of mRNA translation. We demonstrate induction of the phosphorylation of eIF2α in mammalian cells after either pharmacological inhibition of the phosphoinositide 3-kinase (PI3K)–Akt pathway or genetic or small interfering RNA–mediated ablation of Akt. This increase in the extent of eIF2α phosphorylation also occurred in Drosophila cells and depended on the endoplasmic reticulum (ER)–resident protein kinase PERK, which was inhibited by Akt-dependent phosphorylation at threonine-799. The activity of PERK and the abundance of phosphorylated eIF2α (eIF2αP) were reduced in mouse mammary gland tumors that contained activated Akt, as well as in cells exposed to ER stress or oxidative stress. In unstressed cells, the PERK-eIF2αP pathway mediated survival and facilitated adaptation to the deleterious effects of the inactivation of PI3K or Akt. Inactivation of the PERK-eIF2αP pathway increased the susceptibility of tumor cells to death by pharmacological inhibitors of PI3K or Akt. Thus, we suggest that the PERK-eIF2αP pathway provides a link between Akt signaling and translational control, which has implications for tumor formation and treatment.

Introduction

Control of mRNA translation is a crucial process in regulating the expression of genes involved in cell growth, proliferation, and differentiation (1). Translational control is intimately involved in cancer development through the selective synthesis of proteins that influence tumor initiation, progression, and metastasis (2). Most of the regulation of mRNA translation is thought to be exerted at the level of initiation through the coordinated actions of several eukaryotic initiation factors (eIFs) that facilitate the recruitment of the ribosome to an mRNA and its positioning at the initiation codon (3). Metazoans respond to various forms of environmental stress by blocking the initiation process through inducing the phosphorylation of the α subunit of eIF2 at Ser51 (S51) to generate phosphorylated eIF2α (eIF2αP) (4). The phosphorylation of eIF2α is mediated by members of a family of kinases, each of which responds to distinct stimuli (4). The family includes the heme-regulated inhibitor (HRI), whose activity is induced by heme deficiency and which plays a role in the regulation of globin synthesis; general control nonrepressed 2 (GCN2), which is activated by uncharged transfer RNA (tRNA) in response to amino acid deficiency; the endoplasmic reticulum (ER)–resident protein kinase PERK (also known as PEK), whose activity is induced by the accumulation of unfolded proteins in the ER and which represents an essential arm of the unfolded protein response (UPR); and the RNA-activated protein kinase (PKR), an interferon (IFN)–inducible protein activated by double-stranded RNA (dsRNA) (4). These enzymes exhibit substantial amino acid sequence similarities, particularly in the protein kinase domain (KD), which explains their specificity toward eIF2α (4). eIF2αP is implicated in tumorigenesis through its ability to act as either a promoter of cell survival or an inducer of cell death in response to various types of stress, including DNA damage, oncogenic stress, ER stress, or stress in the tumor microenvironment (5, 6).

Genetic alterations that lead to an increase in the extent of phosphoinositide 3-kinase (PI3K) signaling are commonly observed in human cancers (7). Induction of PI3K activity after the stimulation of cells by hormones, mitogens, or growth factors results in the activation of the serine and threonine kinase Akt [also known as protein kinase B (PKB)], which phosphorylates several proteins involved in the regulation of cell survival and proliferation (8). Among these, the mammalian target of rapamycin (mTOR) is a protein kinase that stimulates protein synthesis by mediating, directly or indirectly, the phosphorylation of proteins implicated in cap-dependent mRNA translation (9).

We recently demonstrated that the phosphorylation of eIF2α by PKR mediates the proapoptotic properties of the tumor suppressor phosphatase and tensin homolog deleted from chromosome 10 (PTEN) independently of its inhibition of PI3K-Akt signaling (10); however, it is possible that inhibition of the PI3K-Akt pathway stimulates the phosphorylation of eIF2α through a kinase other than PKR. Here, we demonstrate that PERK acts downstream of Akt and promotes an adaptation process in response to inhibition of the PI3K-Akt pathway. We showed that PERK was a substrate of Akt, which has implications for the regulation of eIF2α phosphorylation and Akt signaling in response to stress. We further showed that inactivation of PERK and eIF2α phosphorylation had effects on promoting tumor death in response to pharmacological inhibition of the PI3K-Akt pathway. Hence, we suggest that in addition to the well-established role of Akt in the regulation of protein synthesis through mTOR, Akt can control protein synthesis through the PERK-eIF2αP pathway. Regulation of the PERK-eIF2αP arm by Akt affects the function of Akt in response to stress as well as the efficacy of tumor cell treatment with chemotherapeutic drugs that target the PI3K-Akt pathway.

Results

Inhibition of PI3K induces the phosphorylation of eIF2α and activates PERK

When human glioblastoma U87 cells, human breast cancer SkBr3 cells, or spontaneously immortalized mouse embryonic fibroblasts (MEFs) were treated with the PI3K inhibitor LY294002, we observed an increase in the extent of phosphorylation of eIF2α at Ser51 in a time-dependent manner (Fig. 1A and fig. S1, A and B). The abundance of eIF2αP was also increased after treatment of U87 cells with GDC-0941, which is a potent and specific inhibitor of PI3K (Fig. 1B) (11). Treatment of U87 cells with the mTOR inhibitor KU0063794 (12) or the mitogen-activated or extracellular signal–regulated protein kinase kinase 1 (MEK1) inhibitor PD98059 (13) did not affect the abundance of eIF2αP (fig. S1, C and D), suggesting that an increase in eIF2αP abundance was specific to the inhibition of PI3K. An increase in eIF2αP abundance in response to inhibition of PI3K is an evolutionary conserved process, because we also observed it in Drosophila melanogaster embryo Kc167 cells after treatment with LY294002 (fig. S2A). The efficacy of LY294002 and GDC-0941 in all cells was confirmed by monitoring the reduction in the extent of Akt phosphorylation at Ser473 as well as the inhibition of glycogen synthase kinase 3β (GSK-3β) phosphorylation at Ser9 or phosphorylation of ribosomal S6 protein at Ser235/236 (Fig. 1, A and B, and fig. S1, A and B).

Fig. 1

Induction of eIF2α phosphorylation upon inhibition of PI3K requires PERK and GCN2. (A and B) Human glioblastoma U87 cells were left untreated (A and B, lane 1) or were treated with LY294002 (20 μM) (A, lanes 2 and 3) or with GDC-0941 (5 or 10 μM) (B, lanes 2 and 3) for the indicated times. (C) Isogenic wild-type (WT) or PERK, GCN2 DKO MEFs were left untreated (lanes 1 and 3) or were treated with LY294002 (20 μM, lanes 2 and 4) for 6 hours. (D) Immortalized WT MEFs were left untreated (lane 1) or were treated with LY294002 (20 μM) for the indicated times. (A to D) Protein extracts (50 μg) were analyzed by Western blotting for the indicated proteins. The ratio of the amount of eIF2αP to that of total eIF2α or actin and of the amount of phosphorylated PERK to that of total PERK for each lane from three experiments is indicated in the accompanying bar graphs. Error bars indicate the SD. In this and all other figures, symbols that define statistical significance are as follows: #P < 0.002; ##P < 0.005; ###P < 0.01; ####P < 0.05; *P < 0.0001; **P < 0.0002; ***P < 0.0005; ****P < 0.001. A list of the statistical comparisons in each figure can be found in table S1.

To identify the kinase responsible for the phosphorylation of eIF2α, we performed experiments in Kc167 cells in which we used small interfering RNA (siRNA) to knock down either Drosophila PERK (dPERK) or Drosophila GCN2 (dGCN2), the two kinases that phosphorylate eIF2α in Drosophila. We found that the targeting of either dPERK (fig. S2B) or dGCN2 (fig. S2C) with siRNA prevented induction of the phosphorylation of Drosophila eIF2α (deIF2α) by LY294002. Because of the lack of availability of antibodies against dPERK or dGCN2, we verified siRNA-mediated silencing of targets by showing the lack of an induction of deIF2α phosphorylation in Kc167 cells after treatment with thapsigargin (fig. S2B) and ultraviolet C (UV-C) light (fig. S2C), stimuli that activate PERK (14) and GCN2 (15), respectively. These data implicated both dPERK and dGCN2 in the phosphorylation of eIF2α in response to PI3K inhibition.

To substantiate these observations in mammalian cells, we examined the phosphorylation of eIF2α in MEFs lacking both PERK and GCN2 [in so-called double knockout (DKO) cells]. We observed that, unlike in wild-type MEFs, induction of eIF2α phosphorylation was blocked in DKO MEFs after treatment with LY294002 (Fig. 1C). Additional experiments with MEFs lacking either PKR (16) or HRI (17) indicated that neither kinase was involved in the phosphorylation of eIF2α in response to PI3K inhibition (fig. S3). This is further supported by our work demonstrating that PKR mediates the phosphorylation of eIF2α downstream of PTEN and independently of the ability of PTEN to inhibit PI3K signaling (10). We next examined the phosphorylation of PERK at Thr980, an autophosphorylation site in the activation loop of the kinase that is essential for the phosphorylation of eIF2α (14). We found that treatment of wild-type MEFs with LY294002 led to a substantial induction of PERK phosphorylation at Thr980, which was accompanied by an increase in the extent of phosphorylation of eIF2α (Fig. 1D). The activation of PERK was not as a result of the induction of ER stress, because LY294002 did not affect the splicing of X-box binding protein 1 (XBP-1) mRNA (fig. S4), which serves as a reliable indicator of the UPR (18).

Inactivation of Akt leads to the phosphorylation of eIF2α

To determine the mechanism of PERK activation, we knocked down Drosophila Akt (dAkt) in Drosophila Kc167 cells with siRNA and found that this increased the basal amount of deIF2αP; however, this was not further increased after treatment with LY294002 (fig. S5A). When we performed experiments with wild-type MEFs and MEFs lacking both Akt1 and Akt2 (Akt DKO MEFs) (19), we observed that the deficiency in Akt1 and Akt2 increased the basal amount of eIF2αP compared to that in wild-type MEFs (Fig. 2A, lanes 1 and 4), which was further increased after knockdown of the remaining Akt isoform, Akt3, by siRNA (Fig. 2A, lane 7). We also observed that LY294002 induced eIF2α phosphorylation to a greater extent in wild-type MEFs than in Akt DKO MEFs treated with either a scrambled, control siRNA, or an siRNA against Akt3 (Fig. 2A). These data indicated that knockdown of Akt increased the amount of eIF2αP, which could not be further increased by PI3K inhibition.

Fig. 2

Inactivation of Akt induces eIF2α phosphorylation. (A) WT cells were left untreated (lane 1) or were treated with LY294002 (20 μM) (lanes 2 and 3) for the indicated times. Akt1Akt2 DKO MEFs were treated with scrambled control siRNA (SCR, lanes 4 to 6) or with siRNA against Akt3 (lanes 7 to 9) in the absence (lanes 4 and 7) or presence of LY294002 (20 μM) (lanes 5, 6, 8, and 9) for the indicated times. (B to D) Immortalized WT MEFs were treated with the indicated concentrations of (C) Akt inhibitor VIII, (D) Akt inhibitor IX, or (E) Akt inhibitor XI for the indicated times. (A to D) Protein extracts (50 μg) were analyzed by Western blotting for the indicated proteins. The ratios of the amounts of eIF2αP to those of either total eIF2α or actin for each lane from three experiments are shown in the graphs. Error bars indicate the SD.

To further support these data, we performed experiments with pharmacological inhibitors of Akt, such as inhibitor VIII and inhibitor XI, both of which target the pleckstrin homology (PH) domain of Akt (20), and inhibitor IX, which directly inhibits Akt activity (20). We observed that all of the inhibitors caused a substantial induction in the phosphorylation of eIF2α in MEFs (Fig. 2, B to D) and human tumor cells (fig. S5, B to D). We verified the efficiencies of the inhibitors by monitoring the impaired phosphorylation of Akt at Ser473 and that of ribosomal S6 protein at Ser235 and Ser236 (Fig. 2, B to D, and fig. S5, B to D). Collectively, these data support the notion that Akt has an inhibitory effect on the phosphorylation of eIF2α.

Akt inactivates PERK by phosphorylation at Thr799

Mouse PERK contains seven serine and threonine residues, all of which conform to a canonical RxRxxS/T phosphorylation consensus site (in which x represents any amino acid residue) for Akt (Fig. 3A). To test whether PERK was a substrate of Akt, we performed in vitro kinase assays with a catalytically inactive glutathione S-transferase (GST) fusion of PERK (GST-PERK) and catalytically active Akt1 or Akt2. As a control, we included in the assays GST–GSK-3β, which is a well-characterized substrate of Akt. We found that active Akt induced the phosphorylation of GST-PERK, as indicated in assays with radioactive [γ-32P]adenosine 5′-triphosphate (ATP) (fig. S6, A and B) or with nonradioactive ATP after Western blotting analysis with antibodies specific for phosphorylated substrates of Akt (fig. S6C). Phosphorylated GST-PERK was subsequently subjected to mass spectrometry (MS), which identified that Akt-mediated phosphorylation occurred at Thr799 only (Fig. 3B). Thr799 of mouse PERK and the surrounding consensus Akt substrate sequence are conserved in human and rat PERK (Fig. 3B). Assays in transiently transfected COS-1 cells demonstrated the phosphorylation of Myc-tagged wild-type PERK but not of a Myc-tagged Thr799→Ala mutant of PERK (PERK T799A) after immunoprecipitation with antibody against Myc and Western blotting analysis with antibodies specific for phosphorylated substrates of Akt (Fig. 3C). Autophosphorylation of Thr980 of PERK, which is a marker of its activation, was more highly induced in Myc-PERK T799A than in wild-type Myc-PERK in transfected COS-1 cells that also expressed Myc-Akt1 (Fig. 3D). Furthermore, COS-1 cells expressing Myc-PERK T799A contained greater amounts of endogenous eIF2αP than did cells expressing wild-type Myc-PERK, consistent with the increased autophosphorylation capacity of Myc-PERK T799A at Thr980 (Fig. 3D). These data demonstrated that the phosphorylation of Thr799 plays an inhibitory role in the activation of PERK.

Fig. 3

Akt inactivates PERK by phosphorylation at Thr799. (A) Sequence alignment of putative Akt phosphorylation sites in mouse PERK. (B) Tandem MS (MS/MS) spectrum of the phosphopeptide SREGpTSSSIVFEDSGCGNASSK of mouse PERK phosphorylated by Akt1 and subjected to in-gel digestion with trypsin. The spectrum shows a peptide that contains the pThr799 residue identified by the analysis of the y and b fragments of the peptide. An intense neutral loss of phosphoric acid (−98) was observed from the precursor ion [peaks at a mass/charge ratio (m/z) of 1122.63]. Alignment of the amino acid sequence surrounding the Thr799 phosphorylation site indicates its conservation in mouse, human, and rat PERK orthologs. (C) COS-1 cells (5 × 105) were transfected with 5 μg of pcDNA empty vector (lane 1) or 5 μg of pcDNA vector encoding either Myc-tagged, WT mouse PERK (WT) (lane 2) or Myc-tagged mouse PERK T799A (lane 3). Forty-eight hours after transfection, protein extracts (500 μg) were subjected to immunoprecipitation with antibody against Myc followed by Western blotting analysis with antibody against phosphorylated substrates of Akt (top panel) or antibody against Myc (bottom panel). (D) COS-1 cells (5 × 105) were transfected with 2.5 μg of pcDNA vector encoding Myc-Akt1 (lanes 1 to 3) and 2.5 μg of pcDNA vector encoding Myc-tagged WT PERK (lane 1), Myc-PERK T799A (lane 2), or pcDNA vector alone (lane 3). Forty-eight hours after transfection, protein extracts (50 μg) were analyzed by Western blotting for the indicated proteins. The ratios of the amount of phosphorylated Myc-PERK to that of total Myc-PERK as well as the ratio of the amount of eIF2αP to that of total eIF2α for each lane from three experiments are shown in the graphs. Error bars indicate the SD. Abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; and Y, Tyr.

Akt antagonizes the PERK-eIF2αP pathway in stressed and tumor cells

The PERK-eIF2αP pathway is an important element of the switch from pro-survival to pro-death signaling during chronic or severe ER stress that leads to induction of the activating transcription factor 4 (ATF4) and CCAAT/enhancer binding protein (C/EBP) homologous protein (CHOP) (21, 22). On the other hand, activation of Akt in ER-stressed cells is largely associated with cell survival (23, 24). Thus, we wished to examine the possible connection between PERK and Akt in response to ER stress. When wild-type MEFs and isogenic Akt DKO MEFs (19) were treated with thapsigargin, we observed that Akt promoted cell survival in response to ER stress (Fig. 4A); however, phosphorylation of PERK at Thr980 and of eIF2α was more highly induced in DKO MEFs than in wild-type MEFs (Fig. 4B). In addition, we observed the ability of Akt to impair the phosphorylation of eIF2α in response to ER stress in experiments with human HT1080 cells in which all three Akt isoforms were knocked down by siRNA (fig. S7). Moreover, Akt-mediated phosphorylation of PERK was impaired in Akt DKO MEFs compared to that in wild-type MEFs after treatment with thapsigargin (Fig. 4C). Together, these data suggested that Akt inhibited PERK by phosphorylating PERK at Thr799 in response to ER stress.

Fig. 4

PERK counteracts the pro-survival properties of Akt in ER-stressed cells. (A) Isogenic WT MEFs and Akt DKO MEFs were treated with thapsigargin (1 μM, TG) for the indicated times. Cell death (% of cells in sub-G1) was determined by propidium iodide staining and flow cytometric analysis. Histograms show the quantification of results from three independent experiments; error bars indicate the SD. (B and C) WT and Akt DKO MEFs were treated with thapsigargin (1 μM, TG) for the indicated times. (B) Protein extracts (50 μg) were analyzed by Western blotting for the indicated proteins. The ratios of the amount of eIF2αP to that of total eIF2α, the amount of Thr980-phosphorylated PERK to that of total PERK, and the amounts of ATF4 or CHOP to that of actin are indicated for each lane. (C) Protein extracts (500 μg) were subjected to immunoprecipitation with antibody against PERK followed by Western blotting analysis with antibody against phosphorylated Akt substrates (top panel) or antibody against PERK (bottom panel). The ratios of the amount of phosphorylated PERK or eIF2αP to that of total PERK or eIF2α, respectively, as well as the ratio of ATF4 and CHOP to actin for each lane from three separate experiments are shown in the graphs. Error bars indicate the SD. (D) Akt DKO MEFs were transfected with 10 μg of pcDNA vector alone or 10 μg of pcDNA vector encoding Myc-PERK K618A in the presence of 5 μg of plasmid encoding GFP. Cells were treated with thapsigargin for 18 or 24 hours and stained with Hoechst 33342. Cell death (% sub-G1) in cells expressing GFP was assessed by cell sorting and flow cytometric analysis. Histograms show the quantification of results from three independent experiments; error bars indicate the SD.

We further observed that the increased extents of PERK activity and eIF2α phosphorylation in Akt DKO MEFs compared to those in wild-type MEFs were associated with the increased abundances of ATF4 and CHOP, which indicated that the higher susceptibility of the Akt DKO MEFs to cell death in response to ER stress might be mediated, in part, by the activation of the ATF4-CHOP pathway (Fig. 4B). We confirmed the proapoptotic role of PERK in experiments that showed that the blockade of endogenous PERK by a Myc-tagged dominant-negative and catalytically inactive form of PERK (Myc-PERK K618A) caused a 40% decrease in the susceptibility of Akt DKO MEFs to cell death in response to ER stress (Fig. 4D). Collectively, these data demonstrated that Akt-mediated phosphorylation of PERK prevented the induction of apoptosis under prolonged ER stress.

Contrary to its function in ER stress, Akt plays a proapoptotic role in cells subjected to oxidative stress (25). Consistent with previous studies (25), we observed that wild-type MEFs were more susceptible than were Akt DKO MEFs to death after treatment with hydrogen peroxide (H2O2) (Fig. 5A). We also saw that the extent of H2O2-induced phosphorylation of PERK at Thr980 and of eIF2αP was greater in Akt DKO MEFs than in wild-type MEFs (Fig. 5B). Similarly, H2O2-dependent phosphorylation of eIF2α was more induced in human HT1080 cells in which all three Akt isoforms were targeted by siRNA than in cells with intact Akt (fig. S7). We further noticed that Akt-mediated phosphorylation of PERK was impaired in H2O2-treated DKO MEFs compared to that in H2O2-treated wild-type MEFs (Fig. 5C). These data demonstrated that Akt activation resulted in the inhibition of PERK-eIF2αP arm in cells under oxidative stress. To address the biological role of Akt-mediated phosphorylation of PERK, we reconstituted PERK KO MEFs with either Myc-tagged wild-type PERK (Myc-PERK) or Myc-PERK T799A followed by treatment with H2O2. We observed that the proapoptotic effects of H2O2 were reduced by 50% in PERK KO cells reconstituted with PERK T799A compared to that in mock-transfected cells or cells reconstituted with wild-type PERK (Fig. 5D). These data confirmed that the inactivation of PERK by phosphorylation at Thr799 promoted the proapoptotic effects of Akt in cells subjected to oxidative stress. Thus, PERK antagonizes Akt-mediated cell death in response to oxidative stress.

Fig. 5

PERK antagonizes the proapoptotic function of Akt in response to oxidative stress. (A) Isogenic WT and Akt DKO MEFs were treated with H2O2 (0.75 or 1 mM) for 1 hour. Cell death (% sub-G1) was determined by propidium iodide staining and flow cytometric analysis. Histograms show the quantification of results from three independent experiments; error bars indicate the SD. (B and C) WT and Akt DKO MEFs were treated with increasing concentrations of H2O2 for 1 hour. (B) Protein extracts (50 μg) were analyzed by Western blotting for the indicated proteins. The ratios of the amount of eIF2αP to that of total eIF2α and of the amount of PERK phosphorylated at Thr980 to that of total PERK for each lane are indicated. (C) Protein extracts (500 μg) were subjected to immunoprecipitation with antibody against PERK followed by Western blotting with antibody against phosphorylated Akt substrates (top panel) or antibody against PERK (bottom panel). The ratios of the amount of phosphorylated PERK at either T799 or T980 to that of total PERK as well as the ratio of the amount of eIF2αP to that of total eIF2α from three separate experiments are shown in the graphs. Error bars indicate the SD. (D) PERK KO MEFs were transfected with 10 μg of pcDNA vector alone or 10 μg of pcDNA vector encoding either Myc-tagged WT PERK or Myc-PERK T799A in the presence of 5 μg of vector encoding GFP. Cells were treated with H2O2 (0.5 mM) for 1 hour and stained with Hoechst 33342. Cell death (% sub-G1) was assessed in cells expressing GFP by cell sorting and flow cytometric analysis. Histograms show the quantification of results from three independent experiments; error bars indicate the SD.

We further examined whether inactivation of the PERK-eIF2αP pathway by Akt could be observed in vivo. To this end, we used transgenic mice expressing a constitutively active form of Akt1 bearing the Thr308→Asp and Ser473→Asp substitutions (Akt1-DD) together with an oncogenic version of ErbB2/Neu (NDL), which is a member of the epidermal growth factor receptor (EGFR) family, under the control of the mouse mammary tumor virus (MMTV) promoter (26). Expression of Akt1-DD in NDL mice substantially accelerates the progression of mammary gland tumors compared to that in NDL mice alone (26). When the NDL transgenic mice were used, we observed that mammary tumors expressing the NDL transgene exhibited increased PERK activity, which was detected by immunohistochemical (IHC) analysis of PERK phosphorylated at Thr980, compared to that in mammary tumors from mice coexpressing NDL and Akt1-DD, in which we could not detect PERK activity by IHC (Fig. 6). We also examined PERK activity and eIF2α phosphorylation in tumor samples by Western blotting analysis, and we observed that the amounts of PERK phosphorylated at Thr980 and of eIF2αP in NDL tumors were increased compared to those in NDL tumors expressing Akt1-DD (Fig. 6). The activity of Akt1-DD in tumors was determined by measuring the increased amounts of Akt phosphorylated at Ser473 by IHC and Western blotting (Fig. 6). These data demonstrated the inhibition of the PERK-eIF2αP pathway by activated Akt in vivo.

Fig. 6

Akt inhibits PERK activity and eIF2α phosphorylation in mouse mammary gland tumors in vivo. (A) Mammary tumors from transgenic mice expressing either NDL 2-5 alone (NDL-CON) or NDL 2-5 together with the constitutively active Akt1-DD (NDL-Akt1DD) were subjected to IHC analysis for PERK phosphorylated at T980 (PERK-pT980), Akt phosphorylated at Ser473 (Akt-pS473), and hematoxylin and eosin (H&E) staining (left panel). The black arrows indicate staining of PERK phosphorylated at T980. (B) Lysates of mammary tumors from two different NDL-CON or NDL-Akt1DD mice were analyzed by Western blotting for the indicated proteins. The ratio of the amount of phosphorylated PERK to that of total PERK, the amount of phosphorylated Akt to that of total Akt, and the amount of eIF2αP to that of total eIF2α from three separate experiments is shown in the graphs (right panel).

Phosphorylation of eIF2α is an adaptive response to the inhibition of PI3K and Akt that has implications in cancer treatment

Next, we investigated the biological role of eIF2αP in response to inhibition of the PI3K-Akt pathway. To this end, we performed experiments with MEFs bearing a S51A knock-in homozygous mutation of eIF2α (that is, eIF2α A/A MEFs) together with their isogenic wild-type counterparts (that is, eIF2α S/S MEFs) (27). We observed that treatment of both types of MEF with LY294002 increased Go-G1 arrest and induced apoptosis, effects that were more prominent in eIF2α A/A MEFs than in eIF2α S/S MEFs (Fig. 7A and fig. S8A). Consistent with these findings, treatment of the MEFs with Akt inhibitor VIII, IX, or XI resulted in an enhanced induction of death in eIF2α A/A MEFs than in eIF2α S/S MEFs, as documented by flow cytometric analysis (Fig. 7, B and C, and fig. S8B). Collectively, these data demonstrate that eIF2αP promoted cell survival in response to disruption of the PI3K-Akt pathway. The data also implied that elimination of the cytoprotective effects of eIF2αP might render tumor cells more susceptible to death after treatment with pharmacological inhibitors of the PI3K-Akt pathway. To address this possibility, we knocked down PERK in HT1080 fibrosarcoma cells with siRNA, as demonstrated by the decreased abundance of PERK and the impaired induction of eIF2α phosphorylation after treatment with inhibitors of PI3K and Akt (fig. S9). We observed that treatment with either the PI3K inhibitor GDC-0941 or Akt inhibitor XI induced cell death more efficiently in cells with inactivated PERK than in cells with intact PERK (Fig. 7, D and E). These findings demonstrated that elimination of the cytoprotective effects of the PERK-eIF2αP pathway rendered tumor cells more susceptible to death after treatment with inhibitors of PI3K-Akt signaling.

Fig. 7

The PERK-eIF2αP pathway promotes cell survival and reduces the efficacy of antitumor treatment by inhibitors of PI3K and Akt. (A to C) Isogenic eIF2α S/S and eIF2α A/A MEFs were treated with (A) LY294002 (20 μM) for 24 or 48 hours, (B) the Akt inhibitor VIII (50 μM) for 12 hours, or (C) Akt inhibitor XI (50 μM) for 48 hours. (D and E) HT1080 cells were treated with either scrambled control siRNA or siRNA against PERK in the absence or presence of either (D) GDC-0941 (10 μM) or (E) Akt inhibitor XI (50 μM) for the indicated times. (A to E) Cell death (% of cells in sub-G1) was assessed by propidium iodide staining and flow cytometric analysis. Histograms in (A) show the quantification of results from seven independent experiments; error bars indicate the SD. Histograms in (B) to (E) show the quantification of results from three independent experiments; error bars indicate the SD.

Discussion

Our study reveals an essential role for Akt signaling in the inhibition of eIF2α phosphorylation. Although genetic analysis implicates both PERK and GCN2 in this process, it is not presently known how GCN2 activity is controlled by Akt. A previous study in budding yeast indicated an indirect role of TOR in the regulation of GCN2, the only kinase that phosphorylates eIF2α in this organism. That is, inhibition of TOR by rapamycin induces the dephosphorylation of GCN2 at Ser577, a modification that leads to GCN2 activation and induction of eIF2α phosphorylation (28). However, Ser577 is not conserved in other GCN2 orthologs (28), suggesting that the inhibition of GCN2 by mTOR is not a universal mechanism. Consistent with this notion, pharmacological inhibition of mTOR did not affect eIF2α phosphorylation in U87 cells (fig. S1C), indicating that mammalian GCN2 is controlled by the PI3K-Akt pathway at a level different from that of mTOR. Nevertheless, our biochemical and biological data demonstrate a role for PERK in the induction of eIF2α phosphorylation in response to the inhibition of PI3K, Akt, or both. Our study further demonstrates that the ability of Akt to function as a “checkpoint” of survival or death is affected by the inactivation of PERK by phosphorylation at Thr799 (fig. S10).

The biological implications of our findings are highlighted by the ability of PERK and Akt to respond to ER stress and oxidative stress, two forms of stress that are intimately linked to tumorigenesis. Specifically, ER stress is induced in solid tumors that are deprived of nutrients, glucose, and oxygen as a result of limited tumor vascularization (29, 30). Under these conditions, tumor cells have developed mechanisms to adapt to ER stress to maintain their survival and growth. Mild forms of ER stress in tumors lead to the activation of PERK, which promotes adaptation through translational and transcriptional mechanisms (29); however, tumor adaptation becomes defective under conditions of severe or chronic forms of ER stress (21, 22), which lead to the induction of a proapoptotic program that is orchestrated not only by the sustained activation of PERK (Fig. 4) but also by the inactivation of Akt (31). Cell adaptation through PERK and eIF2αP also involves inhibition of the production of reactive oxygen species (ROS) and reduction of damage caused by oxidative stress as part of an antioxidant mechanism that involves the induction of ATF4 (32, 33) and expression of NF-E2 (nuclear factor erythroid-derived 2)–related factor 2 (Nrf2) (34). The antioxidant function of PERK has profound effects on tumor promotion (35, 36) as well as tumor resistance to chemotherapeutic drugs (36). Contrary to the effects of PERK and eIF2αP, Akt increases ROS synthesis through inhibition of the transcriptional function of FoxO proteins (25). Akt may further increase ROS synthesis by inhibiting the PERK-eIF2αP pathway, which in turn facilitates cell death from oxidative stress (Fig. 5) (25). The ability of Akt to respond to oxidative stress has also been linked to the induction of senescence in mouse and human fibroblasts (25, 37). The pro-senescent function of Akt may have implications not only in the regulation of tumorigenesis but also in the development of metabolic diseases, such as insulin resistance and diabetes (38). Unlike that of Akt, the role of PERK in the regulation of senescence is not known; however, the abundance of eIF2αP is increased in senescent cells as part of the UPR (39, 40). Several observations support the notion that the UPR is compromised by aging (41), which indicates that PERK and eIF2αP may play roles in senescence initiated by activated Akt.

Our study provides evidence of roles for PERK and eIF2αP in anticancer therapies that target the PI3K pathway. Specifically, we showed that the ability of activated Akt to impair the PERK-eIF2αP pathway was a property of tumor cells grown in vitro and in vivo (Fig. 6). We also showed that inactivation of the cytoprotective effects of the PERK-eIF2αP pathway increased the susceptibility of human tumor cells to death by inhibitors of PI3K or Akt (Fig. 7). This is consistent with many studies that showed that pharmacological inhibitors of PI3K and Akt signaling are inducers of apoptosis in several types of cancer; however, their best use in the clinic is in combination with other chemotherapies (42). Our data showed that inactivation of PERK and eIF2αP may provide a suitable means to improve the efficacy of current chemotherapies that target PI3K-Akt signaling. Our data suggest that the phosphorylation of eIF2α may represent an important mechanism of addiction of tumor cells to chemotherapeutic drugs; hence, it is a potential target for anticancer therapy.

Materials and Methods

Plasmids

Plasmids encoding Myc-tagged wild-type PERK, Myc-PERK K618A, and GST-PERK K618A were described previously (14).

Cell culture and treatments

PKR KO, PERK KO, and HRI KO MEFs lacking PKR (encoded by EIF2AK2), PERK (EIF2AK3), and HRI (EIF2AK1), respectively, and their isogenic wild-type counterparts were maintained in culture as described previously (43). eIF2α S/S and eIF2α A/A MEFs were cultured as described elsewhere (27). The culture of Akt DKO MEFs and their isogenic wild-type MEFs was also described previously (25). HT1080 and U87 cells were cultured as previously described (10, 44). Drosophila Kc167 cells were obtained from the Drosophila Genomics Resource Center and were grown in Shields and Sang insect medium containing yeast extract (1 g/liter), bactopeptone (2.5 g/liter), and 5% heat-inactivated fetal calf serum. LY294002 (Biomol), H2O2 (Bio-Rad), thapsigargin (Sigma), the Akt inhibitors VIII, IX, and XI (EMD Chemicals), GDC-0941 (Selleck Chemicals), PD98059 (Selleck Chemicals), and KU0063794 (Bethyl Laboratories) were used at the concentrations described in the figure legends.

Protein extraction, immunoprecipitation, and Western blotting analysis

Extraction of proteins from mouse and human cells was performed as described previously (10). Protein extraction from Drosophila Kc167 cells was also performed as described elsewhere (45). Western blotting analysis of protein extracts was performed as described previously (10). Primary antibodies used in this study were as follows: rabbit polyclonal antibodies against dAkt pSer505, Akt/PKB pSer473, total Akt/PKB, and S6-pSer235/236 (Cell Signaling); rabbit serum against eIF2α pSer51 (Invitrogen); rabbit polyclonal antibody against eIF2α (Santa Cruz); antibody against Myc (Santa Cruz Biotechnology); mouse monoclonal antibody against actin (ICN); rabbit monoclonal antibody against PERK pThr980 (Cell Signaling); in-house–generated mouse monoclonal antibody against PERK; antibody against phosphorylated Akt substrates (Cell Signaling); antibody against ATF4 (Proteintech Group); and antibody against CHOP (Santa Cruz Biotechnology). The secondary antibodies used were horseradish peroxidase (HRP)–conjugated antibody against mouse immunoglobulin G (IgG) and HRP-conjugated antibody against rabbit IgG (Amersham Pharmacia Biotech). Myc-tagged proteins were immunoprecipitated from total protein extracts (500 μg) with antibody against the Myc tag (2 μg) conjugated to anti–mouse IgG Sepharose beads (Sigma). Endogenous PERK was immunoprecipitated from total protein extracts (500 μg) with monoclonal antibody against PERK (5 μg) coupled to anti–mouse IgG Sepharose beads (Sigma). Quantification of the bands in the linear range of exposure was performed by densitometry with National Institutes of Health Image 1.54 software.

Transfections

Transient transfections of COS-1 cells were performed with the Lipofectamine Plus Reagent (Invitrogen) according to the manufacturer’s protocol. Cells (5 × 105) were seeded in 60-mm plates and 5 μg of plasmid DNA was used for transfection. Cells were then incubated at 37°C for 48 hours, which included any treatments.

RNA interference

MEFs were transfected with siRNA with the Amaxa Nucleofector system (MEF kit 1) according to the manufacturer’s specifications. HT1080 cells were transfected with siRNA with Lipofectamine 2000 according to the manufacturer’s instructions. Treatment of cells with siRNA against mouse Akt3 (SMARTpool, Dharmacon), human Akt1, Akt2, and Akt3 (SMARTpool, Dharmacon), human PERK (SMARTpool, Dharmacon), or scrambled control siRNA (Dharmacon) was performed for 48 hours.

Drosophila RNA interference and analysis

dsRNAs to target Drosophila Akt/PKB, S6K, PERK, and GCN2 were synthesized by in vitro transcription (IVT) in 20-μl reactions with a T7 MEGAscript RNAi (RNA interference) kit (Ambion). DNA templates for IVT were generated by reverse transcription–polymerase chain reaction (RT-PCR) assay from total Drosophila cellular RNA with the TRIzol reagent as specified by the manufacturer (Invitrogen). As a silencing control, dsRNA against the human transferrin receptor (TfR) was generated by RT-PCR assay from total RNA obtained from human HT1080 cells. Primers (which incorporated a 5′ and 3′ T7 promoter) for the synthesis of dsRNAs against dAkt/dPKB, dS6K, dPERK, and dGCN2 are listed in the Supplementary Materials. Drosophila Kc167 cells (1 × 106) were diluted in serum-free medium (1 ml) and incubated with dsRNA (20 μg) in a six-well cell culture dish for 30 min at room temperature, followed by the addition of medium (2 ml) containing 10% fetal bovine serum. The cells were incubated for an additional 3 days to enable turnover of the target protein.

Flow cytometric analysis

Cells (2 × 105 cells) were seeded in 100-mm plates and treated with LY294002 (refreshed every 12 hours), thapsigargin, H2O2, or the Akt inhibitors VIII, IX, or XI at the concentrations and times indicated in the text. Cells were detached with phosphate-buffered saline (PBS) containing 1 mM EDTA and centrifuged at 900g for 5 min. Cells were fixed by gently adding ice-cold 70% ethanol (4 ml) to the pellet, and samples were then stored at −20°C overnight. For staining, the ethanol was removed and cells were resuspended in 0.5 ml of PBS containing propidium iodide (50 μg/ml; P4170, Sigma) and ribonuclease (RNase) (20 μg/ml; Sigma). Cells were incubated at room temperature in the dark for 30 min and subjected to flow cytometric analysis with a FACScan flow cytometer. PERK KO MEFs and Akt1 and Akt2 DKO MEFs were transfected with a pcDNA plasmid encoding green fluorescent protein (GFP) alone or together with plasmids encoding Myc-PERK, Myc-PERK T799A, or Myc-PERK K618A. Forty-eight hours after transfection, cells were treated with H2O2 (0.5 mM) for 2 and 4 hours or with thapsigargin (1 μM) for 18 and 24 hours. The GFP-positive cells were gated and the population of cells in sub-G1 was quantified by flow cytometry after staining with Hoechst 33342 (10 μg/ml).

In vitro phosphorylation assays

In vitro kinase assays with Akt and GST-PERK K618A or GST–GSK-3β K85R/K86R as substrates were performed with [γ-32P]ATP, 25 mM magnesium acetate, and 0.25 mM cold ATP. For Akt1 (Upstate), the kinase reaction buffer contained 40 mM Mops/NaOH and 1 mM EDTA; for Akt1 (Biomol), 25 mM Mops, 12.5 mM β-glycerophosphate, 5 mM EGTA, 2 mM EDTA, 25 mM MgCl2, and 0.25 mM dithiothreitol (DTT); and for GST-Akt2 (Biomol), 50 mM tris, 0.5 mM DTT, 1 mM EGTA, and 0.2 mM sodium orthovanadate. Kinase reactions were incubated for 30 min at 30°C. Samples were subjected to SDS–polyacrylamide gel electrophoresis (SDS-PAGE), stained with Coomassie blue, and subjected to autoradiography. Nonradioactive in vitro kinase assays with Akt were performed as described earlier in the presence of 3 mM ATP. The MS procedure used is described in the Supplementary Materials.

Statistical analysis

All quantitative variables are presented as means ± SD. We compared the differences of three or more groups with one-way analysis of variance (ANOVA) and post-test analyses with GraphPad Prism 5, and the differences of two groups with two-tailed Student’s t test (GraphPad Prism 5). P < 0.05 was considered statistically significant. In the figure legends, symbols that define statistical significance are as follows: #P < 0.002; ##P < 0.005; ###P < 0.01; ####P < 0.05; *P < 0.0001; **P < 0.0002; ***P < 0.0005; ****P < 0.001. A list of the statistical comparisons in each figure can be found in table S1.

Supplementary Materials

www.sciencesignaling.org/cgi/content/full/4/192/ra62/DC1

Materials and Methods

Fig. S1. Specificity of induction of eIF2αP by PI3K inhibitors.

Fig. S2. Induction of eIF2α phosphorylation by PI3K inhibition in Drosophila cells.

Fig. S3. PKR and HRI are not involved in the induction of eIF2α phosphorylation by PI3K inhibition.

Fig. S4. PI3K inhibition does not result in induction of ER stress.

Fig. S5. Akt inactivation results in the induction of eIF2α phosphorylation.

Fig. S6. PERK is phosphorylated by Akt in vitro.

Fig. S7. Akt diminishes the induction of eIF2α phosphorylation in response to oxidative or ER stress in human cells.

Fig. S8. Induction of PERK activation and eIF2α phosphorylation by PI3K or Akt inactivation promotes cell survival.

Fig. S9. Inactivation of PERK increases the efficacy of tumor treatment with PI3K and Akt inhibitors.

Fig. S10. Model of the inhibition of the PERK-eIF2αP arm by Akt.

Table S1. Results of statistical analysis.

References

References and Notes

  1. Acknowledgments: We thank D. Ron for Myc- and GST-tagged PERK constructs and for PERK KO and PERK, GCN2 DKO MEFs; R. J. Kaufman for HRI KO eIF2α S/S and eIF2α A/A MEFs; N. Hay for wild-type and Akt1 and Akt2 DKO MEFs; P. N. Tsichlis for the Myc-Akt1 plasmid; N. Benlimame for IHC analysis of mouse tissues; H. Muaddi for help in some experiments; and members of the Koromilas lab for helpful discussions. Mass spectrometric analysis was performed by the proteomics platform of the Eastern Quebec Genomics Center, Quebec, Canada. IHC analysis was performed at the Georges and Olga Minarik Research Pathology Facility at the Segal Cancer Centre, Jewish General Hospital, Montreal, Canada. Funding: The work was supported by MOP-38160 grant from the Canadian Institutes of Health Research (CIHR) and a grant from the Quebec Breast Cancer Foundation (QBCF) to A.E.K., a joint grant from the Cancer Research Society (CRS) and QBCF to W.J.M., MOP-12182 from CIHR to B.P., and grants DK60596 and DK53307 from the NIH to M.H. Z.M. is the recipient of a Canada Graduate Studentship Doctoral Award from the CIHR. Author contributions: Z.M. and A.E.K. designed the study, interpreted the data, and wrote the paper; Z.M., J.L.K., and S.W. performed the experiments; Z.M. performed the statistical analysis of the data; B.P. contributed to MS analysis; S.C. and W.J.M. provided mammary tumors from Akt-DD and NDL transgenic mice; and M.H. contributed to the design of experiments and data analysis. Competing interests: The authors declare that they have no competing interests.
View Abstract

Navigate This Article