HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 16, 13508-13517, April 19, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/16/13508    most recent M200017200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Alvarez-Tejado, M. Articles by del Peso, L. Search for Related Content PubMed PubMed Citation Articles by Alvarez-Tejado, M. Articles by del Peso, L. Social Bookmarking                      What's this?

Lack of Evidence for the Involvement of the Phosphoinositide 3-Kinase/Akt Pathway in the Activation of Hypoxia-inducible Factors by Low Oxygen Tension^*

Miguel Alvarez-Tejado^§, Arántzazu Alfranca^¶, Julian Aragonés^§, Alicia Vara, Manuel O. Landázuri, and Luis del Peso^**

From the Servicio de Inmunología, Hospital de la Princesa, Universidad Autónoma de Madrid, Diego de León 62, 28006 Madrid, Spain

Received for publication, January 2, 2002, and in revised form, January 24, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Hypoxia-inducible factors (HIF) belong to an evolutionary conserved family of transcription factors, the activity of which is tightly regulated by oxygen levels. We have recently demonstrated that hypoxia activates the phosphoinositide 3-kinase (PI3K)/Akt pathway in some cell types, and other works have suggested that this pathway is involved in the activation of HIF. In the present work we studied the role of this pathway in the induction of HIF by hypoxia. Under hypoxic conditions the PI3K/Akt pathway was activated in some (PC12 and HeLa) but not all cell types (HepG2) tested, whereas the HIF protein was induced by hypoxia in all cases. Kinetics analysis showed that, when observed, the activation of PI3K/Akt occurred after HIF induction. In addition, the chemical inhibition of PI3K had no significant effect on the induction of the HIF protein or its transcriptional activity but prevented Akt activation. Accordingly, transient overexpression of a dominant negative form of the regulatory subunit of PI3K in HEK293T cells did not interfere with the induction of the HIF- protein by hypoxia or affect HIF-mediated transcription in any of the cell types tested. Moreover, forced activation of the PI3K/Akt pathway did not affect the transcriptional activity of HIF under normoxic or hypoxic conditions. Thus, our data suggest that the activation of PI3K/Akt by hypoxia is cell type-specific and, when observed, lies downstream of HIF activation or in a parallel pathway. Furthermore, the activity of the PI3K/Akt is not sufficient for the activation of HIF nor is it essential for its induction by hypoxia.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The reduction in oxygen levels in aerobic organisms triggers specific cellular and systemic adaptive responses. Among the molecular responses to hypoxia found in multicellular organisms, the activation of HIF¹ transcription factors is the best characterized and the one responsible for the activation of genes involved in energy metabolism, angiogenesis, and apoptosis. HIF proteins belong to the basic helix-loop-helix-Per/ARNT/Sim (bHLH-PAS) family of transcription factors. The functional HIF unit is a heterodimer composed of an HIF- and a HIF- (ARNT) subunit. The  subunit is common to all HIF complexes, whereas there are three different HIF- subunits that can participate in the complex: HIF-1, HIF-2 (EPAS), and HIF-3. Both  and  mRNAs as well as the HIF- protein are constitutively expressed; however, HIF- protein levels are tightly regulated by the level of oxygen. Under normoxic conditions HIF proteins show a remarkably short half-life (<5 min), whereas a decline in oxygen tension results on its accumulation (1). Further analysis demonstrated that a specific region of HIF, the oxygen-dependent degradation domain, mediates its targeting for ubiquitination and proteasomal degradation under normoxic conditions (1, 2). Importantly, the von Hippel-Lindau tumor suppressor protein (pVHL) binds to HIF and functions as part of an E3 ubiquitin ligase complex that mediates its degradation (3). Recently, several groups (4-7) have demonstrated that the interaction between pVHL and HIF depends on the hydroxylation of specific proline residues located within the oxygen-dependent degradation domain. Because proline hydroxylases require molecular oxygen and iron for their enzymatic activity, the reduction of oxygen levels or treatment with deferoxamine, an iron chelator, results in HIF- stabilization (4-7). Although control of HIF protein stability is the first step in regulating the activity of these transcription factors, hypoxia also affects the translocation of HIF to the nucleus (8) and its transactivation activity (1) by unknown mechanisms.

In addition to proline hydroxylation of HIF-, regulatory pathways have been reported to be important for the control of HIF- levels by hypoxia including diacylglycerol kinase (9), reactive oxygen species (10), MAP kinase cascades (11-13), the GTPase Rac1 (14), and the PI3K/Akt pathway (14-18). However, the mechanisms by which these molecules affect HIF- stability or function have not been clarified. Among these routes, the implication of the PI3K/Akt pathway has received much attention lately.

The PI3K belongs to a large family of lipid kinases that phosphorylate phosphatidylinositol phospholipids at position D3 of the inositol ring. Members of class I of the PI3K family are involved in the transduction of extracellular signals, and the lipid products of the reaction that they catalyze are potent second messengers (19). Among the targets of these second messengers is the protooncogenic serine/threonine kinase Akt. Upon activation of PI3K by extracellular stimuli, Akt translocates from the cytosol to the plasma membrane where it is phosphorylated at two specific residues (threonine 308 and serine 473), resulting in its full activation (19). Because of the central role of PI3K/Akt in the control of cell growth and survival (19-21), this pathway is frequently activated in human tumors by oncogenes such as ras that act upstream of this route or by the loss of the tumor suppressor gene PTEN/MMAC1 (22), which encodes for a phosphatase that dephosphorylates position D3 of inositol phospholipids (23).

Several works (14-18) indicate that the PI3K/Akt pathway might be involved in the induction of HIF in transformed cells. It was first reported that the induction of the angiogenic cytokine vascular endothelial growth factor (VEGF) by hypoxia was enhanced in ras-transformed fibroblast and that this effect was mediated by PI3K/Akt (16). Later, the same group demonstrated that in PTEN-deficient glioblastoma cells the induction of VEGF by hypoxia was dependent on PI3K/Akt (18). Moreover, the stabilization of the HIF- protein by hypoxia was prevented by the overexpression of PTEN, whereas the activation of Akt was sufficient to promote HIF- stabilization under normoxia (18). Other groups have also demonstrated that the inhibition of PI3K with drugs or by overexpression of dominant negative forms prevented hypoxic induction of HIF- protein as well as its transcriptional activity (14, 15, 17, 24).

Recent reports show the induction of HIF under normoxic conditions by certain stimuli such as cytokines (25, 26), hormones (27, 28), growth factors (17, 29-31), and viral-encoded receptors (32). The activation of HIF is dependent on the PI3K/Akt pathway in some cases, such as EGF (17, 31) and insulin-like growth factor 1 (IGF-1) (30) stimulation, but not others. The mechanism by which PI3K/Akt mediates the induction of HIF by hypoxia or signaling agonists is not yet clear, but it could involve the mammalian target of rapamycin (mTOR) kinase (17, 31), which acts downstream of the PI3K/Akt route to control the translation of specific mRNAs. However, in PC12 cells, a cell type that activates the PI3K/Akt route in response to hypoxia (33, 34), chemical inhibition of PI3K does not affect the induction of HIF- proteins by hypoxia but prevents the activation of Akt (34). In addition, Sandau et al. (25) found that chemical or genetic interference with the PI3K/Akt pathway completely prevented the HIF- accumulation stimulated by the tumor necrosis factor (TNF) but had only a minor effect on the hypoxic induction. Furthermore, the enhancement of VEGF induction by hypoxia found in cells with a constitutively active PI3K/Akt route is not mediated by HIF (35). Finally, it has been reported (24) that inhibition of PI3K affects the hypoxic induction of HIF-1 but not HIF-2. Thus, the involvement of the PI3K/Akt route in the regulation of HIF activity by hypoxia remains controversial.

The aim of this work was to elucidate the possible role of the PI3K/Akt pathway on the induction of HIF by hypoxia. We found that the activation of the PI3K/Akt pathway by hypoxia is cell type-specific and not a general effect of hypoxia. In addition, we observed that chemical or genetic inhibition of PI3K had, at best, a modest effect on HIF- protein or activity, although these treatments severely affected signaling downstream of PI3K. In addition, activation of the PI3K/Akt route with growth factors or by the overexpression of active mutants of PI3K or Akt did not result in the induction of HIF- protein or activity. Thus, we propose that PI3K/Akt activity is not sufficient for HIF induction nor is it essential for its regulation by hypoxia.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture and Reagents-- The human hepatoma (HepG2) and the human embryo kidney (HEK293T) cell lines were cultured in Dulbecco's modified Eagle's medium (Invitrogen) whereas the cervix carcinoma cell line HeLa and the rat pheochromocytoma PC12 cells were maintained in RPMI 1640 medium with GLUTAMAX-I (Invitrogen). The culture media were supplemented with 10% fetal calf serum in the case of HepG2, HeLa, and HEK293T cells and with 10% horse serum (Invitrogen) and 5% fetal bovine serum for PC12 cells. In all cases the culture media were supplemented with 100 units/ml penicillin and 100 µg/ml streptomycin, and the cells were grown in a humidified atmosphere containing 5% CO₂ at 37 °C. Hypoxia (1% O₂) was induced by a culture of cells inside an air-tight chamber with inflow and outflow valves that was infused with a mixture of 1% O₂, 5% CO₂, and 94% N₂ (S.E. Carburos Metalicos S.A., Madrid, Spain). In those experiments in which drugs were used, the compounds were added 30-60 min prior to transfer to hypoxia/normoxia conditions. Wortmannin, LY294002, PD98059, rapamycin, and NGF were from Calbiochem. EGF was from Promega (Madison, WI), and IGF and HGF were from Sigma. Anti-Akt (catalog no. 9272, anti-phospho Akt serine 473 (catalog no. 9271), and anti-phospho S6K (catalog no. 9205) were from Cell Signaling (Beverly MA); anti-phospho Erk V8031 was from Promega, anti-HIF-1 from BD Transduction Laboratories, and anti-HIF-2/EPAS from Novus Biologicals (Littleton, CO).

Plasmid Constructions-- The FKHR reporter plasmid pPr2xFKHR-Luc was constructed by cloning the insulin response sequence (5'-CAAAACAAACTTATTTTG-3') from the IGF-binding protein 1 gene promoter (36) into a prolactin-luciferase vector (37). The HIF reporter plasmid, p9xHIF1-Luc, has been described (9). To generate p9xHIF-EGFP, the HindIII/HpaI fragment from p9xHIF-Luc containing the minimal promoter fused to nine copies of HIF binding sites was cloned into the promoterless pEGFP C1 vector (CLONTECH). The expression plasmids producing N-terminal epitope-tagged FKHR and different mutants of PI3K and Akt have been described elsewhere (21, 38).

Western Blot-- Immediately after treatments the cells were washed with ice-cold phosphate-buffered saline and harvested in 70-200 µl of a lysis buffer containing 2% SDS, 10% glycerol, 10 mM dithiothreitol, 62 mM Tris, pH 6.8, and 0.004% bromphenol blue. Lysates were sonicated for 4 s, centrifuged at 4 °C for 2 min at 14,000 × g, and resolved on 8-10% SDS-polyacrylamide gels. Proteins were then transferred to nitrocellulose membranes (Bio-Rad) blocked previously with 5% nonfat dry milk in Tris-buffered saline-T (50 mM Tris, pH 7.6, 150 mM NaCl, and 0.1% Tween-20) and incubated overnight at 4 °C with the indicated antibodies. Immunolabeling was detected by enhanced chemiluminescence (ECL, Amersham Biosciences) and visualized with a digital luminescent image analyzer (Fujifilm LAS-1000 CH). The intensity of each band was quantified with the Image Reader v1.8 software (Science Lab software, Fuji Photo Film); normalized values are shown on Figs. 1-3 and 5.

Reporter Assays-- Cells were transfected by Lipofectin (Invitrogen), HepG2, and PC12 or calcium phosphate, HeLa, and HEK293T with the indicated reporter plasmid(s) together with Renilla luciferase plasmid (pRLTK, Promega). To analyze FKHR-mediated transcription, cells were transfected with a reporter plasmid containing FKHR binding sites (pPr2xFKHR-Luc) together with a FKHR (pCDNA3-FLAG-FKHR)-encoding plasmid (21). In those experiments where the effect of different inhibitors was tested, cells were first transfected and 12-24 h after transfection were split into the appropriate number of wells (24-well plates) for treatments. Drugs or vehicles were added to duplicate wells, and cells were incubated for 30-60 min and then transferred to hypoxic conditions or left at normoxia for 6 additional hours. In those experiments where the effect of different PI3K/Akt constructs were tested, cells were transfected with the indicated plasmids and 24 h later one-half of the transfected wells were transferred to hypoxic conditions for 6 h, and the remaining wells were left at control (normoxic) conditions. Finally, cells were harvested, and firefly and Renilla luciferase activities were determined using a dual luciferase system (Promega). Firefly luciferase activity was normalized based on the Renilla luciferase activity. The average of duplicate samples for each condition is represented in Table II and Figs. 4 and 6.

Flow Cytometry-- HEK293T cells were transfected with pEGFP, p85-EGFP, or p9xHIF-EGFP, and 24-36 h later the cells were incubated at hypoxic or normoxic conditions in the presence of complete or serum-free media for the indicated periods of time. The expression of the enhanced green fluorescent protein (EGFP) or of EGFP fusion proteins was analyzed by flow cytometry in a FACSCalibur apparatus (BD PharMingen). Data are presented as green fluorescence intensity versus cell number or side scatter.

Statistical Analysis-- Where indicated, experimental data were analyzed using the Prism^TM GraphPad (version 2.0) software. Significant differences were determined by the analysis of variance (ANOVA) test followed by the Newman-Keuls test; the p values <0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Akt Activation by Hypoxia Is Cell Type-specific and Occurs after HIF Induction-- We have recently reported that the route of PI3K/Akt is activated by hypoxia in PC12 cells (33), and other groups have found similar results in this and other cell types (34, 39). In addition, it has been suggested that this pathway is critical for the induction of HIF proteins by hypoxia (14, 17, 18). Thus, we decided to study whether activation of this signaling pathway is a general effect of hypoxia as well as to assess its role on the induction of HIF. To this end we chose three different cell lines commonly used in studies of the biological effects of hypoxia: PC12, HepG2, and HeLa. HIF-1 is the principal member of the HIF family expressed in HeLa and HepG2 cells, whereas HIF-2/EPAS is expressed in PC12 cells (40).

These cell lines were incubated under hypoxia (1% O₂) or treated with deferoxamine, a chemical agent that mimics some of the molecular effects of hypoxia (41), for different periods of time, and the activation of PI3K/Akt was monitored by the phosphorylation of Akt (33). As a control for the activation of the PI3K/Akt pathway, cells were stimulated under normoxic conditions with growth factors for different periods of time. Hypoxia or deferoxamine treatments resulted in the induction of HIF- proteins in all the cases, although Akt activation was observed in PC12 and HeLa cells but not in HepG2 cells (Fig. 1). In contrast, the PI3K/Akt pathway was efficiently activated by growth factors in all cell types (Fig. 1). Moreover, under hypoxia the accumulation of HIF- protein seemed to precede the activation of the PI3K/Akt pathway in HeLa and PC12 cells (Fig. 1, B and C). Interestingly, none of the growth factors tested was able to induce any significant accumulation of HIF protein (Fig. 1, A-C), although the treatments resulted in a potent and in some cases sustained activation of the PI3K/Akt pathway.

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Hypoxia does not activate the PI3K/Akt pathway in all cell types. Cells were cultured on 24-well (HepG2, HeLa, and HEK293T) or 6-well (PC12) plates and grown for 48-72 h. Then cells were challenged with deferoxamine (DFX, 380 µM), HGF (15 ng/ml), IGF-1 (30 nM), EGF (10 ng/ml), or NGF (100 ng/ml) at standard atmospheric O2 (normoxia) or transferred to hypoxia chambers (Hx, 1% O2) for the indicated periods of time. A, HepG2 cells. B, HeLa cells. C, PC12 cells. The results shown are representative of several independent experiments (HepG2, n = 3; HeLa, n = 2; PC12, n>3). The intensity of each band, relative to that in basal (Nx) conditions, is shown. Nx, untreated cells kept under normoxic conditions; WB, Western blot; AU, arbitrary unit.

Pharmacological Inhibition of the PI3K/Akt Pathway Has No Significant Effect on Hypoxic Induction of the HIF Protein or Its Transcriptional Activity-- The data presented above indicated that the activation of the PI3K/Akt pathway is not a general consequence of hypoxia. We then studied the requirement of PI3K on HIF activation using two structurally non-related inhibitors of PI3K, wortmannin and LY294002. In these studies we included the MEK1/2 inhibitor (MAP kinases cascade) PD98059, because it attenuates the hypoxia-induced transcriptional activity of HIF without affecting HIF protein levels (11, 42). We also included rapamycin, which inhibits mTOR/FRAP (a protein downstream of the PI3K/Akt pathway) that has been suggested to be involved in the regulation of the translation of HIF- mRNA (17, 31). Prior to testing the effect of these compounds on the induction of HIF protein, we determined the optimal dose of each compound required to inhibit its target enzymes on each cell type (Fig. 2). To this end we pretreated cell lines with the indicated concentrations of each compound (see Fig. 2 legend) for 30-60 min and then challenged the lines for 10 min with a growth factor able to induce the pathways of interest on each cell type. PI3K activity was assessed by the level of Akt phosphorylation, MEK activity by the level of the phosphorylation of Erk1/2, and the effect of rapamycin on FRAP was monitored by the level of phosphorylation of S6K at residue Thr-389, which is a target for mTOR/FRAP (Fig. 2). As expected, the sensitivity of each cell line to these compounds was slightly different (see Fig. 2 and Table I). Thus, for the next experiments we chose the dose of each inhibitor required for complete (>95%) inhibition of the growth factor-induced phosphorylation of the relevant target (Table I). As has been reported (43), in the case of PD98059 only a partial inhibition of Erk1/2 was observed in HeLa and PC12 cells.

View larger version (44K): [in this window] [in a new window]   Fig. 2.   Determination of the minimal dose of each inhibitor required for inhibition of its target pathway. Cells were incubated with the different doses of each inhibitor for 30-60 min and then stimulated in the presence of drugs with the indicated growth factor for 10 min. A, HepG2 cells; HGF, 15 ng/ml HGF. B, HeLa cells; EGF, 10 ng/ml EGF. C, PC12 cells; NGF, 100 ng/ml NGF. The doses of inhibitors used were 33, 100, and 300 nM wortmannin; 3.3, 10, and 30 µM LY294002; 16.6, 50, and 150 µM PD98059; 8, 25, and 75 nM rapamycin. The intensity of each band, relative to that in basal (control) conditions, is shown. One representative experiment is shown (HepG2, n = 2; HeLa, n = 2; PC12, n = 2). WB, Western blot; AU, arbitrary unit.

Next we tested the effect of these inhibitors on HIF protein induction by hypoxia and deferoxamine. Cells were pretreated for 30-60 min with the selected dose of each inhibitor (Table I, underlined numbers) and then incubated under hypoxia or treated with deferoxamine for an additional 6 h, always in the presence of the drugs. Fig. 3 shows that except for cycloheximide, which was included as a positive control (33), the rest of compounds had, at best, a partial effect on the induction of the HIF protein by hypoxia or deferoxamine treatment in all three cell lines. We intentionally show experiments where wortmannin and LY294002 had some effect on HIF levels, but this effect was not reproducible.² Importantly, treatment with LY294002 totally prevented the activation of Akt induced by hypoxia in PC12 cells without affecting HIF- induction (Fig. 3C). The induction of phospho-Akt was also strongly reduced (87% inhibition of hypoxia-induced Akt phosphorylation) in the presence of wortmannin, whereas the induction of EPAS was unaffected (Fig. 3C). The lack of complete inhibition of Akt activity in the presence of wortmannin could be due to the instability of this compound on aqueous solutions (44).

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Effect of wortmannin, LY294002, PD98059, and rapamycin on HIF- induction by hypoxia and deferoxamine. Cells were incubated with the indicated inhibitors for 30-60 min and then, in the presence of drugs, transferred to hypoxia, treated with deferoxamine, or left at normoxia for 6 h. A, HepG2 cells; Wt, 300 nM wortmannin; LY, 10 µM LY294002; PD, 50 µM PD98059; Rp, 75 nM rapamycin; Chx, 10 µM cycloheximide. B, HeLa cells; Wt, 33 nM wortmannin; LY, 3.3 µM LY294002; PD, 50 µM PD98059; Rp, 50 nM rapamycin; Chx, 10 µM cycloheximide. C, PC12 cells; Wt, 100 nM wortmannin; LY, 10 µM LY294002; PD, 50 µM PD98059; Rp, 75 nM rapamycin; Chx, 10 µM cycloheximide. Similar results were obtained in several independent experiments (HepG2, n = 3; HeLa, n = 2; PC12, n = 4). WB, Western blot; AU, arbitrary unit.

Hypoxia not only affects HIF- protein levels but also its subcellular localization (8) and transactivation function (45, 46). Because it could be possible that the inhibition of PI3K had some effect on the transcriptional activity of HIF without affecting HIF- protein levels as described for MEK (11, 42), we investigated the effect of chemical inhibitors on the transcription of a firefly gene under the control of hypoxia response elements (HRE) (9). Table II shows the results of several independent experiments; we included experiments with different levels of induction of HIF activity to illustrate that the effect of different inhibitors was independent of the range of induction. As published previously (11, 42), PD98059 treatment resulted in a significant inhibition of the transcriptional response induced by hypoxia (Table II) or deferoxamine treatment² in all the tested cell lines. In contrast, treatment with wortmannin, LY294002, or rapamycin had no reproducible effect. In some experiments these drugs had some inhibitory effect whereas in others they apparently enhanced HIF activity. The effects of wortmannin and LY294002 were particularly inconsistent, and in some experiments these compounds had opposite effects on HIF activity. These data were in sharp contrast with the highly reproducible effects of these drugs on Akt activity and other PI3K-mediated phenomena.³ The average effect of the PI3K inhibitors and rapamycin on HIF activity was small (<30%), not significant (p > 0.05), and inconsistent among the different cell types tested, in contrast to the PD98059 effect (Table II).

Several works have described the inhibition of HIF-mediated transcription by wortmannin or LY294002 (14, 17, 18). However, we found that at doses that effectively inhibit PI3K (Fig. 2 and Table I), neither wortmannin nor LY294002 had a significant effect on any of the three cell types tested (Table II). Thus, we decided to study the effect of higher doses of these inhibitors on HIF-mediated transcription. In HepG2 and PC12 cells (Fig. 4, A and C), HRE-driven transcription was not affected even at the highest doses of wortmannin (300 nM). However concentrations of LY294002 higher than those required for the inhibition of PI3K (Fig. 2) resulted in a clear inhibition of HIF-mediated transcription in HepG2 and HeLa cells (Fig. 4, A and B). HIF activity was also sensitive to increasing doses of wortmannin in HeLa cells, but not in PC12 cells although this effect was not as important as that of LY294002 (Fig. 4B). As expected, PD98059 had an important inhibitory effect at all doses tested in all cell types (Fig. 4). Finally, rapamycin did not have any effect at all doses tested in HepG2 and PC12 cells and only affected HeLa cells at the highest dose (Fig. 4).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Effect of wortmannin, LY294002, PD98059, and rapamycin on the induction of HIF-mediated transcription by hypoxia. Cells were transfected with an HRE-driven firefly luciferase reporter plasmid together with a Renilla luciferase expression plasmid. 24 h after transfection cells were split into several wells and, when attached, pretreated with increasing doses of inhibitors for 30-60 min. Then, in the presence of drugs, cells were transferred to hypoxia (Hx, 1% O2) or left at normoxia (N) for 6 h. The firefly luciferase activity normalized to the Renilla luciferase activity is represented as -fold over control samples. A, HepG2 cells; Wt, 100 and 300 nM wortmannin; LY294002, 10 and 30 µM LY294002; PD98059, 16.6 and 50 µM PD98059; Rp, 50 and 150 nM rapamycin. B, HeLa cells; Wt, 33, 100, and 300 nM wortmannin; LY294002, 3.3, 10, and 30 µM LY294002; PD98059, 16.6, 50, and 150 µM PD98059; Rp, 16.6, 50, and 150 nM rapamycin. C, PC12 cells; Wt, 100 and 300 nM wortmannin; LY294002, 10 and 30 µM LY294002; Rp, 50 and 150 nM rapamycin.

Genetic Interference with the PI3K/Akt Pathway Has No Effect on HIF- Protein or Its Transcriptional Activity-- In addition to the chemical inhibition of PI3K, we decided to study the effect of the expression of a dominant negative form (p85) of this molecule. To study the effect of this construct on the induction of endogenous HIF- protein, we chose the HEK293T cells because they can be transiently transfected with high efficiency. As a control for the inhibitory effect of p85, we monitored the transcriptional activity of FKHR, a transcription factor of the forkhead family, the activity of which is regulated by the PI3K/Akt pathway. The activity of FKHR is inhibited through direct phosphorylation by Akt at three independent residues (21, 47-51). After phosphorylation by this route, FKHR is sequestered in the cytoplasm by binding to the 14-3-3 proteins, preventing it from entering the nucleus (21, 48, 51). Thus, inhibition of the PI3K/Akt pathway results in the transcriptional activation of FKHR (47-49, 51, 52).

The expression of p85 inhibited endogenous PI3K activity as demonstrated by the potent induction of FKHR activity found even with the minimal amount of p85 (Fig. 5C). Increasing the amount of p85 did not result in higher FKHR activity, suggesting that the maximal inhibition of PI3K had been achieved. However, p85 had no effect on the induction of HIF- protein by hypoxia (Fig. 5A) despite a high proportion of cells expressing the p85 construct (Fig. 5B).

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Effect of PI3K on the induction of the HIF- protein by hypoxia on HEK293T cells. HEK293T cells were grown on 60-mm dishes and transiently transfected with pcDNA3 vector (negative control for flow cytometry) or plasmids coding for EGFP (pEGFP) alone (2 µg) or fused in-frame to p85 (1 or 2 µg) together with the PPr2xFKHR-Luc reporter plasmid (200 ng), the FKHR coding plasmid (200 ng), and Renilla expression plasmids (200 ng). In all cases the total amount of DNA was completed to 7 µg/plate with pCDNA3 vector. 12 h after transfection plates were split in two, cells were allowed to attach for 12 h and then incubated under normoxic (N) or hypoxic (HP) conditions for 6 additional hours. After treatment, cells were harvested, divided into aliquots, and processed for Western blot, flow cytometry, or luciferase and Renilla activities. A, the effect of p85 on HIF- induction. B, transfection efficiency (percent of EGFP-positive cells) and mean fluorescence intensity (MFI). The data presented correspond to transfected cells exposed to hypoxia; the same results were obtained with normoxic samples. C, the effect of p85 on FKHR activity. The experiment was repeated twice with the same result. AU, arbitrary unit.

Next we investigated the effect of p85 as well as dominant positive forms of PI3K (p110-CAAX) and Akt (Myr-Akt) on the transcriptional activity of HIF. Fig. 6 shows that the overexpression of p85 had no effect on the hypoxia-induced transcriptional activity of HIF in HepG2 and HeLa cells. However, p85 expression had a clear-cut positive effect on FKHR-mediated transcription (Fig. 6), which was more pronounced in the presence of serum because of the higher endogenous PI3K activity found in this condition. The expression of p85 had no effect on HIF activity in PC12 cells.² In addition, we found that forced activation of the PI3K/Akt pathway induced by the expression of p110-CAAX or Myr-Akt had no effect on the basal (normoxic) or hypoxia-induced transcription activity of the HRE-driven promoter tested. In contrast, these constitutive active forms of PI3K and, in particular, Akt resulted in the inhibition of FKHR-mediated transcription (Fig. 6); in this case the effect was more patent in the absence of serum because under these conditions FKHR is more active because of decreased activity of the endogenous PI3K/Akt pathway.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Effect of PI3K/Akt pathway on the induction of HIF-mediated transcription. HepG2 (A) or HeLa (B) cells were transfected with an HIF- or FKHR-responsive firefly luciferase reporter plasmid together with a Renilla luciferase expression plasmid and different combinations of constructs coding for active PI3K/Akt mutants or dominant negative PI3K. 24 h after transfection cells transfected with the HRE reporter plasmid were transferred to hypoxia or left at normoxic conditions for 6 h, whereas cells transfected with the FKHR-responsive reporter were serum-starved or left in complete media for the same period of time at normoxic conditions. The effect of PI3K and Akt constructs on FKHR-mediated transcription is shown (upper panels). We assigned a value of 100 (horizontal line) to the normalized luciferase activity of the FKHR-responsive reporter in both serum-free and complete media when no PI3K or Akt construct was cotransfected, and the rest of the samples are represented as a percent of this value (average value is shown). The effect of the PI3K and Akt constructs on the HIF-mediated transcription in both normoxic and hypoxic conditions is shown (lower panels). We assigned a value of 1 (horizontal line) to the firefly luciferase activity of the HIF-responsive reporter normalized by Renilla activity when no PI3K or Akt construct was cotransfected; the rest of the samples were expressed as a fold of this value. Each symbol represents the result of an independent experiment, each one done in duplicate. The horizontal bar symbol represents the mean of all the experiments shown. The mean value is shown.

Taking advantage of the high transfection efficiency of the HEK293T cell line, we additionally studied the effect of the dominant negative forms of PI3K and active Akt on HIF activity on this cell line by a different approach. We studied HIF transcriptional activity with the aid of a reporter construct that expresses the EGFP under the control of an HRE-driven promoter. Together with this construct, we cotransfected the cells with the FKHR reporter plasmid and measured FKHR and HIF activities in the same transfected cells instead of in parallel experiments. In these experiments HIF activity was measured by flow cytometry analysis of green fluorescence (mean fluorescence intensity), whereas FKHR activity was monitored by luciferase activity in lysates of the same transfected cells. As shown in Fig. 7, the EGFP protein was induced by hypoxia in HEK293T cells regardless of the presence or absence of p85. In contrast, p85 induced strong activation of FKHR-mediated transcription in a dose-dependent manner. In addition, the transfection of the active forms of PI3K² or Akt did not induce HRE-mediated transcription under normoxia or cooperate with hypoxia, although their inhibitory effect on FKHR was evident at all amounts of plasmid transfected (Fig. 7).

View larger version (54K): [in this window] [in a new window]   Fig. 7.   Effect of PI3K/Akt pathway on HIF transcriptional activity. HEK293T cells were grown on 60-mm dishes and transiently transfected with pcDNA3 vector or plasmids coding for p85 (0.2-3 µg) or Myr-Akt (0.2-3 µg) together with p9xHIF-EGFP (200 ng) and PPr2xFKHR-Luc (200 ng) reporter plasmids, FKHR coding plasmid (200 ng), and Renilla expression plasmids (200 ng). In all cases the total amount of DNA was 7 µg/plate with a pCDNA3 vector. 12 h after transfection, plates were split in two, and the cells were allowed to attach for 24 h and then were incubated under normoxic or hypoxic conditions for 15 additional hours. R59949 (20 µM), a compound that prevents HIF activation by hypoxia (9), was included as a control of inhibition of HIF activity. After treatment cells were harvested, part of the cells were analyzed by flow cytometry, the rest of the cells were lysed, and luciferase and Renilla activities in lysates were measured. The experiment was repeated twice with the same result. M.F.I., mean fluorescence intensity.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The activation of HIF transcription factors by hypoxia is critical for the induction of genes such as VEGF, pyruvate kinase, glucose transporter-1, tyroxine hydroxylase, and EPO that mediate the adaptive responses to low oxygen levels such as metabolic changes, angiogenesis, and erythropoiesis. A vast effort has been made toward the understanding of the regulation of HIF by oxygen. These studies have shown that hypoxia induces the stabilization of HIF- proteins (1, 2), the transactivation activity mediated by the  subunits (1), and possibly their translocation to the nucleus (8). Among these effects, only the mechanism of stabilization of the HIF- protein upon hypoxic stimulation begins to be understood.

In this work we have studied the role of the PI3K/Akt on the hypoxic induction of the HIF- protein and HIF transcriptional activity. Several lines of evidence have led us to suggest that the activation of this route is not sufficient for the induction of the HIF- protein. First, stimulation with several growth factors that activate the PI3K/Akt route did not result in the induction of HIF- in any of the cell types tested (Fig. 1). Second, the expression of constitutively active forms of PI3K or Akt did not induce HIF-mediated transcription under normoxic conditions (Figs. 6 and 7), suggesting that they were unable to induce the HIF- protein and/or its transcriptional activity. Moreover, the transient expression of active Akt in HEK293T did not induce HIF-, and PC12 cell lines stably expressing p110-CAAX or Myr-Akt showed levels of EPAS protein identical to controls under normoxic and hypoxic conditions.²

On the other hand, the data presented here suggest that the PI3K/Akt route is not essential for the induction of HIF by hypoxia. We and others have found that PI3K/Akt is activated by hypoxia in some cell types such as PC12 cells (Fig. 1 and Refs. 33 and 34), HeLa (Fig. 1), PTEN-deficient glioblastoma cell lines (18), HT1080 fibrosarcoma (39) cells, and canine kidney epithelial cells (39). In contrast, this effect is not observed in other cell types including HepG2 (Fig. 1), HEK293T,² MDA-MB-468, and T47D breast cancer cell lines (24), PC-3 prostate cancer cells (17), and 3T3 cells (31). However, the HIF- protein and HRE-mediated transcription is potently induced by hypoxia in all cases. Moreover, the activation of HIF- and PI3K/Akt by hypoxia, when shown, are dissociated in time and mechanistically; Akt phosphorylation is detected after HIF- induction (Fig. 1), and PI3K inhibitors prevent activation of Akt but not HIF induction (Fig. 3). Although the PI3K/Akt pathway is not generally activated by hypoxia, it could be proposed that the basal activity of this pathway is required for HIF induction. However, our results argue against this possibility. The HIF- protein and its activity are basically not affected by chemical or genetic inhibition of this pathway, whereas under this experimental setting the activity of the PI3K route is strongly diminished as demonstrated by the activity of downstream molecules such as Akt and FHKR. Thus, either the PI3K/Akt activity is not essential for HIF activation or the residual activity remaining after chemical or genetic interference is sufficient to allow HIF activation.

In contrast to our data (Fig. 3 and Table II) and the data of others (25, 34, 35), several works demonstrate an inhibitory effect of wortmannin and LY294002 on the HIF- protein and its activity induced by hypoxia (14, 17, 18). However, in addition to differences in the cell types studied or experimental conditions such as the use of anoxia instead of hypoxia (18), it is evident that in these works the optimal doses of PI3K inhibitors were not determined, and it is possible that their conclusions are based on the use of doses of these inhibitors that were too high. Accordingly, we found that HIF activity was, in some cell types, inhibited when cells were treated with doses of inhibitors above those determined to effectively inhibit PI3K (Fig. 4). In this regard, some works describe the inhibition of HIF at concentrations of 100 µM LY294002 (14, 17), a dose that we found to be toxic for HEK293T cells.² Furthermore, these works show that whereas LY294002 (50-100 µM) inhibits the induction of HIF-, the treatment with wortmannin had virtually no effect (14, 17). We have also found that the activation of HIF is more sensitive in most cell types to increasing doses of LY294002 than to wortmannin. Because both compounds are PI3K inhibitors, these data suggest that LY294002 could have nonspecific effects. In this regard, it has been recently reported (53) that LY294002 but not wortmannin inhibits casein kinase II at doses similar to those required for PI3K inhibition (IC₅₀ of 6.9 µM for CKII and 10 µM for PI3K). The lack of a consistent effect of PI3K inhibitors on HIF- protein and activity was in agreement with the lack of effect of p85, which exerted an effective inhibition of PI3K as assessed by the activation of FKHR-mediated transcription.

In summary, the data presented herein indicate that PI3K/Akt pathway activity is neither sufficient for HIF- induction nor does it play an essential role in its induction by hypoxia. Still, it is possible that this pathway may play an indirect or modulatory role not related to the machinery activated by hypoxia on the induction of the HIF- protein or its activity. In this regard, it has been recently reported (31) that HER2 (neu) signaling promotes the translation of HIF-1 mRNA, through the activation of the PI3K/Akt/FRAP pathway, resulting in the accumulation of HIF- protein under normoxic conditions. Accordingly, we found that the inhibition of translation by cycloheximide treatment completely prevented HIF- induction by hypoxia, but the lack of effect of rapamycin (Fig. 3, Table II, and Ref. 16) on HIF induction by hypoxia argues against a role for mTOR/FRAP, at least under our experimental conditions. However, the PI3K/Akt pathway also activates rate-limiting steps of the general protein translation machinery by rapamycin-insensitive pathways such as the regulation of eIF2B activity (54). Thus it is possible that the PI3K/Akt pathway might modulate the induction of HIF- by hypoxia through its effects on the translation machinery. The differences observed for the involvement of the PI3K/Akt pathway on the hypoxic induction of HIF among different works could be explained by the differential dependence on this pathway for protein translation in distinct cell types and/or experimental conditions. Finally, the involvement of the PI3K/Akt route in the general aspects of cell biology that might modulate the induction of HIF by hypoxia, such as its effect on the translation machinery, could also explain the variability of the effect of the PI3K/Akt inhibitors on the HIF-meditated transcription observed by us.

Recent works indicate that the activation of the HIF protein or at least the stabilization of HIF- is controlled by oxygen levels through proline hydroxylases (55). In light of these works, the implication of several molecules and pathways on the induction of HIF should be re-evaluated. That is the case of the reactive oxygen species, the implication of which on HIF induction is now seriously questioned (56, 57). The data presented above argue against a central role for the PI3K/Akt pathway in the induction of the HIF protein or its activity by hypoxia. In agreement with our data, a recent article (55) shows that in the nematode Caenorhabditis elegans the stabilization of HIF- induced by low oxygen tension is not altered in animals harboring a loss of function mutation in the PI3K gene age-1. Because the machinery responsible for O₂ sensing and induction of HIF- is conserved from nematodes to mammals, the normal function of this machinery in age-1 worms indicates either that the PI3K is not required for HIF induction by hypoxia or that the mammalian and nematode systems differ at some essential points.

	ACKNOWLEDGEMENTS

We thank E. Temes for her interest in this project and suggestions, R. González-Amaro for critically reviewing the manuscript, and F. Sánchez-Madrid for providing some of the plasmids used in this work.

	Note Added in Proof

While this manuscript was in press, Simon and co-workers reported similar findings (Arsham, A. M., Plas, D. R., Thompson, C. B., and Simon, M. C. (February 21, 2002) J. Biol. Chem. 10.1074/jbc.M111162200).

	FOOTNOTES

^* This work was supported in part by Fondo de Investigaciones Sanitarias Grant FIS01/0264, Ministerio de Educación y Cultura Grant FEDER 2FD/997-1870, and Ministerio de Ciencia y Tecnología Grant SAF 2001/0215.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

These authors contributed equally to this work.

^§ Supported by a postdoctoral fellowship from the Comunidad Autónoma de Madrid.

^¶ Supported by graduate fellowships from the Ministerio de Educación y Cultura.

These senior authors contributed equally to this work.

^** Recipient of a Contrato de Investigación from the Fondo de Investigaciones Sanitarias. To whom correspondence should be addressed. Tel.: 34-91-5202371; Fax: 34-91-5202374; E-mail: lpeso@hlpr.insalud.es.

Published, JBC Papers in Press, January 28, 2002, DOI 10.1074/jbc.M200017200

² M. Álvarez-Tejado, A. Alfranca, J. Aragones, M. O. Landázuri, and L. del Peso, unpublished observations.

³ Specifically, with respect to the effect of wortmannin and LY294002 on the viability of PC12 cells (33) and on the induction of BAD phosphorylation by cytokines (20).

	ABBREVIATIONS

The abbreviations used are: HIF, hypoxia-inducible factor(s); ARNT, aryl receptor nuclear translocator; EPAS, endothelial Per/ARNT/Sim; PI3K, phosphoinositide 3-kinase; VEGF, vascular endothelial growth factor; PTEN, phosphatase and tension homolog deleted from chromosome 10; EGF, epidermal growth factor; IGF, insulin-like growth factor; mTOR, mammalian target of rapamycin; NGF, nerve growth factor; HGF, hepatocyte growth factor; FKHR, transcription factor of the forkhead family; EGFP, enhanced green fluorescent protein; MAP, mitogen-activated protein; MEK, MAP kinase/extracellular signal-related kinase kinase; FRAP, FKBP (FK506-binding protein)-rapamycin-associated protein; HRE, hypoxia response element.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES