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

J. Biol. Chem., Vol. 277, Issue 11, 9262-9267, March 15, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/11/9262    most recent M110377200v1 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 Katschinski, D. M. Articles by Wenger, R. H. Search for Related Content PubMed PubMed Citation Articles by Katschinski, D. M. Articles by Wenger, R. H. Social Bookmarking                      What's this?

Heat Induction of the Unphosphorylated Form of Hypoxia-inducible Factor-1 Is Dependent on Heat Shock Protein-90 Activity^*

Dörthe M. Katschinski^§¶, Lu Le^¶, Daniel Heinrich, Klaus F. Wagner^**, Thomas Hofer, Susann G. Schindler, and Roland H. Wenger^§§§¶¶

From the  Institute of Physiology and ^** Department of Anaesthesiology, Medical University of Lübeck, Ratzeburger Allee 160, D-23538 Lübeck, Germany and the  Institute of Physiology, University of Zürich, Winterthurerstrasse. 190, Zürich CH-8057, Switzerland

Received for publication, October 29, 2001, and in revised form, December 21, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Hypoxia-inducible factor (HIF)-1 is the oxygen-sensitive subunit of HIF-1, a transcriptional master regulator of oxygen homeostasis. Oxygen-dependent prolyl hydroxylation targets HIF-1 for ubiquitinylation and proteasomal degradation. Unexpectedly, we found that exposing mice to elevated temperatures resulted in a strong HIF-1 induction in kidney, liver, and spleen. To elucidate the molecular mechanisms responsible for this effect, HepG2 hepatoma cells were exposed to different temperatures (34-42 °C) under normoxic (20% O₂) or hypoxic (3% O₂) conditions. Heat was sufficient to stabilize mainly a phosphatase-resistant, low molecular weight form of HIF-1 (termed HIF-1_a). Heat-induced HIF-1_a accumulated in the nucleus but neither bound to DNA nor trans-activated reporter or target gene expression, demonstrating the need for post-translational modifications for these functions. The protein banding pattern of heat-induced HIF-1 in immunoblot analyses was clearly distinct from the HIF-1 pattern after prolyl hydroxylase inhibition (by hypoxia or iron chelation/replacement) or following proteasome inhibition, suggesting that heat stabilizes HIF-1 by a novel mechanism. Inhibition of the ATP-dependent chaperone activity of HSP90 by novobiocin or geldanamycin prevented heat-induced as well as hypoxia-induced HIF-1 accumulation, indicating a common role of the HSP90 chaperone activity in HIF-1 stabilization by these two environmental parameters.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The hypoxia-inducible factor-1 (HIF-1)¹ is a heterodimeric transcription factor composed of the two subunits HIF-1 and HIF-1, both belonging to the basic helix-loop-helix (bHLH)-Per/arylhydrocarbon receptor (AhR) nuclear translocator (ARNT)/Sim (PAS) protein superfamily (1). HIF-1 is identical to the previously described ARNT protein. HIF-1 is a critical regulator of the physiological adaptive response to hypoxia, since it activates genes regulating, among other processes, angiogenesis, erythropoiesis, and glucose metabolism. HIF-1 has also been described to be involved in tumor angiogenesis and ischemic diseases such as myocardial ischemia or stroke (reviewed in Refs. 2 and 3). Whereas ARNT is constitutively expressed, HIF-1 expression is induced in hypoxic cells with an exponential increase in expression as cells are exposed to decreased oxygen partial pressures. As determined in HeLa cells, the highest HIF-1 protein levels are reached at 0.5% oxygen (4, 5). Recent studies have shown that HIF-1 is modified by oxygen-dependent prolyl hydroxylation (6-9), allowing the binding of the von Hippel-Lindau protein (pVHL), which targets HIF-1 for ubiquitinylation and proteasomal degradation (10-17). Further activation of HIF-1 involves nuclear translocation, dimerization with ARNT, DNA binding, and recruitment of transcriptional co-activators. These processes are regulated at least in part by posttranslational modifications. It has been demonstrated that phosphorylation of HIF-1 via the Ras/Raf-MEK-p42/44 and the phosphatidylinositol 3-kinase-PTEN-Akt-GSK3 kinase pathways eventually results in elevated expression and/or transcriptional activity of HIF-1 (18-21).

The protein chaperone complex heat shock protein 90 (HSP90)-p23 (reviewed in Refs. 22 and 23) is involved in the regulation of a variety of bHLH transcription factors including MyoD and the bHLH-PAS proteins Sim and AhR (24-26). HSP90 interacts with two different sites of AhR, one in the bHLH domain and the other in the PAS B domain (26, 27). The interaction of HSP90 with the ligand binding domain of AhR is important for maintaining the receptor in a high affinity ligand binding and repressed conformation (28-30). Furthermore, the HSP90 chaperone complex seems to regulate the intracellular localization of the AhR (31). As shown by in vitro pull-down assays, HSP90 can also bind HIF-1 (32). This observation has been confirmed by co-immunoprecipitation studies using an artificial enhanced green fluorescent protein-HIF-1 fusion protein ectopically overexpressed in heterologous COS-7 cells (33). However, because HSP90 is capable of binding many proteins, the functional significance of these observations remained unknown. In particular, the effect of physiological activation of HSPs by heat on HIF-1 activity has not been reported so far.

We therefore investigated the impact of heat shock (as evidenced by HSP90 induction and activation) in vivo and in vitro on the regulation of cellular HIF-1 expression, protein modifications, and subcellular localization as well as HIF-1 DNA binding and trans-activation activity. Our data indicate a novel role of HSP90 in the stabilization of HIF-1, which is not related to proteasome function or posttranslational modification of HIF-1 by phosphorylation.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Antibodies and Chemicals-- Antibodies were purchased from the suppliers indicated: mouse monoclonal anti-HSP90 (StressGen), mouse anti-HIF-1 (Novus), mouse anti-pVHL (Pharmingen), mouse anti-HIF-1 (Transduction Laboratories), mouse anti--actin (Sigma). Appropriate horseradish peroxidase-labeled secondary antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) and Promega. PD98059 and  protein phosphatase were supplied by Calbiochem and New England Biolabs, respectively. All other chemicals were obtained from Sigma. Petriperm dishes were purchased from Satorius.

Cell Culture and Reporter Gene Assays-- The human hepatoma cell lines HepG2 and Hep3B were obtained from American Type Culture Collection. Cells were grown in high glucose Dulbecco's modified Eagle's medium containing 10% fetal calf serum (Invitrogen) in a humidified 5% CO₂, 95% air atmosphere at 37 °C. For hypoxic exposure, the O₂ concentration in the cell culture incubator (Heraeus) was lowered to 3%. The B1 cell line (kind gift of J. Caro, Philadelphia, PA) is derived from Hep3B cells stably transfected with a reporter gene construct containing the 5'- and 3'-flanking regions of the erythropoietin gene linked to the luciferase reporter gene (10).

Animal Experiments-- The investigations were performed in strict accordance with National Institutes of Health guidelines (48) and approved by the local Governmental Commissions for the Care of Animals. For heat treatment, B6C3F1 mice were placed into a 37 °C incubator for different time periods. The rectal temperature was continuously recorded with a thermocouple. Recooling of the animals was accomplished by exposing the animals after heat treatment to 20 °C ambient temperature. Typically, rectal temperature increased up to 41 °C during heat treatment, reaching the maximum temperature ~2 h after exposing the mice to 37 °C. Recooling resulted in a drop of rectal temperature back to 37 °C within 20 min.

Immunoblot Analyses-- Following stimulation, cells were collected, washed with ice-cold PBS, and extracted with 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, 400 mM NaCl, 0.1% Nonidet P-40, 1 mM dithiothreitol, 1 µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 1 mM phenylmethylsulfonyl fluoride. Protein concentrations were determined using the Bradford method (Sigma). Proteins were separated by SDS-PAGE and transferred onto nitrocellulose membranes by semidry electroblotting (Bio-Rad). Membranes were stained with Ponceau S to verify equal protein loading per lane. After overnight blocking (5% nonfat milk powder in PBS), the blots were incubated for 1-2 h with primary antibodies against HIF-1, pVHL, HSP90, or -actin and detected with secondary horseradish peroxidase-conjugated antibodies. Chemiluminescence detection was performed by incubating the membranes with 100 mM Tris-HCl (pH 8.5), 2.65 mM H₂O₂, 0.45 mM luminol, and 0.625 mM coumaric acid for 1 min followed by exposure to x-ray films.

Immunofluorescence Analyses-- For immunofluorescence analyses, cells were fixed with freshly prepared 3.7% formaldehyde in PBS (pH 7.4) for 10 min, washed with PBS, permeabilized with 0.5% Triton X-100 for 5 min, and rinsed again with PBS. After blocking nonspecific binding sites with 3% bovine serum albumin in PBS for 30 min, the cells were incubated for 1 h with the anti-HIF-1 (Transduction Laboratories) or the anti-HSP90 antibody diluted 1:100 in 3% bovine serum albumin in PBS, followed by a fluorescein isothiocyanate-coupled secondary anti-mouse (Dako) antibody diluted 1:100 with 3% bovine serum albumin in PBS. After extensive washings with PBS, the slides were mounted and analyzed by fluorescence microscopy (Axioplan 2000; Zeiss).

Immunoprecipitation Analyses-- Following stimulation, cells were collected, washed with ice-cold PBS, and extracted with 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, 400 mM NaCl, 0.1% Nonidet P-40, 1 mM dithiothreitol, 1 µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 1 mM phenylmethylsulfonyl fluoride. 1 mg of each protein sample was incubated with 26 µg of anti-HIF-1 (Novus) antibodies and 100 µl of protein G/A-agarose (Oncogene) overnight. Agarose beads were washed three times with lysis buffer containing 2 mM sodium molybdate. Immunoprecipitated proteins were recovered by adding 0.5 M Tris-HCl (pH 6.8), 0.5% SDS, 0.5% bromphenol blue, 20% glycerol. Subsequently, samples were freed of agarose beads by filtration through Micropure 0.22 separators (Amicon), boiled, and separated by SDS-PAGE. HIF-1 and HSP90 were detected by immunoblot analysis as described above.

RNA Extraction and RNA Blot Analysis-- Total cellular RNA preparation and RNA blot analysis was carried out as described previously (34). Hybridization probes for glucose transporter-1, aldolase, and -actin were obtained by gel purification of the inserts and random-primed labeling as described previously (34). Radioactive signals were recorded by exposure to phosphorimaging screens (Fuji).

Electrophoretic Mobility Shift Assay-- Electrophoretic mobility shift assay was performed as described before (35). Oligonucleotides containing an HIF-1 DNA-binding site derived from the erythropoietin gene were gel-purified on 10% polyacrylamide gels prior to 5'-end-labeling of the sense strand with [-³²P]ATP (Amersham Biosciences, Inc.). Unincorporated nucleotides were removed by gel filtration over Bio-Gel P60 columns (Bio-Rad). Labeled sense strands were annealed to a 2-fold molar excess of unlabeled antisense strands. DNA-protein binding reactions were carried out for 4 h at 4 °C in a total volume of 20 µl of 10 mM Tris-HCl (pH 7.5), 50 mM KCl, 50 mM NaCl, 1 mM MgCl₂, 1 mM EDTA, 5 mM dithiothreitol, 5% glycerol, containing 5 µg of nuclear extract, 0.1 µg of sonicated calf thymus DNA (Sigma), and 2 × 10⁴ cpm of oligonucleotide probe. Samples were run on 4% nondenaturing polyacrylamide gels at 200 V in TBE buffer (89 mM Tris, 89 mM boric acid, 5 mM EDTA) at 4 °C. The gels were dried, and radioactive signals were recorded as described above.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Heat Induces HIF-1 Expression in Vivo-- Similar to other bHLH transcription factors, HIF-1 can form a stable association with HSP90 in vitro (33, 36). To obtain evidence as to whether HIF-1 stabilization is functionally linked to activation of the heat shock response in vivo, we exposed mice to an ambient temperature of 37 °C for up to 4 h. During this treatment, the rectal temperature increased up to 41 °C as soon as 2 h of exposure. Three animals were recooled following 3 h of 37 °C by exposing them back to room temperature. Subsequently, animals were sacrificed, and HSP90 as well as HIF-1 protein expression was determined in kidney, liver, spleen, lung, and testis (Fig. 1). In kidney and liver, a strong induction of HIF-1 expression was observed as soon as 1 h after exposure of the mice to an ambient temperature of 37 °C. This induction was detectable up to 3 h after recooling of the animals, although the rectal temperature had decreased back to normal values. In spleen, only a slight induction of HIF-1 protein expression with elevated core body temperature was observed, whereas in the lung no signal could be detected. In testis, a strong constitutive HIF-1 protein expression was observed, which remained unaffected by elevated core body temperature. Interestingly, heat-induced as well as constitutive expression of HIF-1 protein correlated with the expression of HSP90 in the different organs (Fig. 1), suggesting that HSP90 induction might be functionally linked to HIF-1 stabilization.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Heat induces HIF-1 expression in vivo. Mice were exposed to an ambient temperature of 37 °C for different time intervals and were then either sacrificed directly after the heat treatment or recooled by reexposure to 20 °C. Subsequently, HIF-1, HSP90, and -actin protein expression was determined in organ protein extracts. A representative immunoblot analysis of cellular protein probed with anti-HIF-1, anti-HSP90, or anti--actin antibodies is shown.

Temperature-dependent HIF-1 Protein Accumulation-- To gain insight into the molecular basis of the heat-induced HIF-1 expression, human hepatoma HepG2 cells were exposed for 4 h to temperatures ranging from 34 to 42 °C under normoxic (20% O₂) or hypoxic (3% O₂) conditions. HIF-1 but not pVHL or -actin protein levels were induced by elevated incubation temperatures under both normoxic and hypoxic conditions (Fig. 2A). A similar effect was found when desferrioxamine or cobalt chloride was combined with heat (Fig. 2B). Heat-induced expression of HIF-1 started following exposure for 2 h (data not shown). A heat-induced increase in HIF-1 expression was also detectable in Petriperm cell culture dishes, in which the cells were grown attached to an oxygen-permeable silicone membrane. This prevents hypoxic effects that potentially might occur due to increased oxygen consumption following heat exposure (data not shown).

View larger version (49K): [in this window] [in a new window]   Fig. 2.   Heat induces HIF-1 expression in vitro. A, induction of HIF-1 protein after exposure of HepG2 cells to 34-42 °C under normoxic (20% O2) or hypoxic (3% O2) conditions for 4 h. A representative immunoblot analysis of cellular protein probed with anti-HIF-1, anti-pVHL, and anti--actin antibodies is shown. B, induction of HIF-1 protein after exposure of HepG2 cells to 125 µM desferrioxamine (DFX) or 50 µM CoCl2 at 37 °C or 42 °C for 4 h. Representative immunoblot analysis is shown of protein probed with anti-HIF-1.

Three major HIF-1 bands could be resolved by Western blot analysis (termed HIF-1_a, HIF-1_b, and HIF-1_c) (Figs. 2 and 3). Interestingly, heat induced primarily the low molecular weight form HIF-1_a in vitro and to some extent HIF-1_b, whereas hypoxia or the hypoxia-mimicking agents cobalt chloride and desferrioxamine induced primarily the higher molecular weight species HIF-1_b and HIF-1_c as well as intervening forms (Fig. 2, A and B). To determine the biochemical basis of the three HIF-1 forms, protein extracts of HepG2 cells exposed to 37 °C under hypoxic conditions or exposed to 42 °C under normoxic conditions were treated with -phosphatase. The HIF-1_c band disappeared after -phosphatase (400 units) treatment, whereas the HIF-1_a and HIF-1_b bands were not affected (Fig. 3A), even after increasing the amount of -phosphatase up to 2000 units (data not shown). These results suggest that HIF-1_c represents a phosphorylated form of HIF-1. A high resolution immunoblot with the organ protein extracts, derived from mice exposed for 4 h to elevated temperatures, together with a protein extract of hypoxic HepG2 cells, clearly indicated that the HIF-1_a form is also induced by heat in vivo (Fig. 3B).

View larger version (34K): [in this window] [in a new window]   Fig. 3.   Heat induces nonphosphorylated HIF-1. A, HepG2 cells were exposed to 37 °C under hypoxic conditions or 42 °C under normoxic conditions. Cell lysates were then treated with or without 400 units of -phosphatase. Representative immunoblot analysis is shown; proteins were probed with anti-HIF-1. B, organ protein extracts of mice exposed for 4 h to an ambient temperature of 37 °C were analyzed together with a protein extract of HepG2 cells exposed for 4 h to 3% O2 by high resolution HIF-1 immunoblotting.

Heat-induced HIF-1_a Translocates into the Nucleus under Normoxic Conditions-- Immunofluorescence studies were performed to analyze the subcellular localization of heat-induced HIF-1_a. As shown in Fig. 4, nuclear accumulation of HIF-1_a was detected following exposure of HepG2 cells to hypoxic conditions at 37 °C as well as at 42 °C. A similar effect could be observed following exposure of the cells to 42 °C at normoxic conditions but not to 37 °C. In the same cell line, a strong induction of HSP90 was observed at 42 °C in the cytoplasm as well as in the nucleus. Because the same antibody was used for the immunoblot and immunofluorescence studies, it has to be assumed (although this cannot be formally proven) that all three forms of HIF-1 (i.e. HIF-1_a-c) are recognized in the immunofluorescence experiments as well. Strikingly, nuclear HSP90 was detected under hypoxic conditions at 37 °C. Neither these immunofluorescence nor Western blot analyses (data not shown) provided evidence for an increase of total HSP90 cellular protein under hypoxic conditions, indicating that hypoxia induces nuclear translocation of HSP90.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Subcellular localization of HIF-1. HepG2 cells were exposed to 37 °C or 42 °C under normoxic (20% O2) or hypoxic (3% O2) conditions for 4 h. The cells were prepared for immunofluorescence analysis and incubated with anti-HIF-1 or anti-HSP90 antibodies followed by fluorescein isothiocyanate-conjugated secondary antibodies as described under "Experimental Procedures."

Heat Induces HIF-1_a but Not HIF-1 trans-Activation Activity-- Next, we analyzed the functional activities of heat-induced compared with hypoxia-induced HIF-1. Therefore, DNA binding and reporter gene studies were performed. Despite the increase in protein concentration, exposing cells to 42 °C did not result in increased DNA binding activity (Fig. 5A). Consequently, the transcriptional activity of heat-induced HIF-1 remained unchanged in B1 cells exposed to 42 °C (Fig. 5B). Since the luciferase protein was found to be thermoinstable (data not shown and Ref. 37), reporter gene expression was determined at the mRNA level. In contrast to heat, exposing B1 cells to hypoxia resulted in a strong induction of luciferase mRNA expression, which was not further increased by heat.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Heat-induction of HIF-1 is not sufficient for induction of HIF-1 trans-activation activity. A, lack of heat-induced HIF-1 DNA-binding. HepG2 cells were incubated at 37 and 42 °C under normoxic (20% O2) or hypoxic (3% O2) conditions for 4 h. Nuclear protein was extracted and incubated with 32P-labeled oligonucleotides containing the wild type or a mutant HIF-1 binding site derived from the erythropoietin 3' hypoxia response element. Antibodies were added to the samples where indicated to confirm the specificity of the supershifted (ss) protein-DNA complexes. B, luciferase mRNA Northern blotting of B1 cells (Hep3B cells stably transfected with the regulatory elements of the erythropoietin gene linked to the luciferase reporter gene) exposed to 37 and 42 °C under normoxic (20% O2) or hypoxic (3% O2) conditions for 4 h. Shown are mean values ± S.D. of three independent experiments. C, Northern blotting of HepG2 cells exposed to 37 and 42 °C under normoxic (20% O2) or hypoxic (3% O2) conditions for 4 h using radioactively labeled cDNA probes for glucose transporter-1 (Glut-1), aldolase, HIF-1, and -actin mRNA.

Finally, HIF-1 target gene expression was evaluated following exposure to heat and/or hypoxia. In agreement with the lack of DNA-binding and heat-induced reporter gene activity, hypoxia but not heat induced the HIF-1 target genes glucose transporter-1 (Glut-1) and aldolase in HepG2 cells (Fig. 5C). In the same samples, no change in HIF-1 mRNA expression was detectable, demonstrating that neither hypoxia nor heat-induced HIF-1 protein expression is caused by increased mRNA expression.

Heat Induction of HIF-1_a,b Is Not Caused by Inhibition of Prolyl Hydroxylation or Proteasomal Activity-- There are several possible regulatory steps during which heat might block the HIF-1 degradation process, including prolyl hydroxylation, ubiquitinylation, and proteasomal degradation. By high resolution Western blotting, HIF-1 was detected as multiple bands, designated HIF-1_a (likely to represent the native HIF-1 protein), HIF-1_b (representing a phosphatase-resistant form), and HIF-1_c (representing the phosphorylated form). Interestingly, the protein banding pattern of HIF-1 induced by heat corresponded to the pattern obtained with recombinant HIF-1 in vitro transcribed and translated (IVTT) in rabbit reticulocyte lysates (Fig. 6). Because the HIF-1_a and HIF-1_b bands were still present in IVTTs performed in the presence of the iron chelator desferrioxamine (DFX; Fig. 6), an inhibitor of HIF-1 prolyl hydroxylation (6, 7), it is unlikely that HIF-1_b is the hydroxylated form of HIF-1. These results were confirmed by IVTTs performed under hypoxic conditions (data not shown).

View larger version (40K): [in this window] [in a new window]   Fig. 6.   Distinct banding patterns of HIF-1 are induced by heat, hypoxia, and proteasomal inhibition. HepG2 cells were exposed to the indicated temperatures or oxygen concentrations with or without the addition of lactacystin for 4 h. Representative immunoblot analysis is shown of cellular protein or recombinant HIF-1, IVTT with or without the presence of 125 µM desferrioxamine (DFX), probed with anti-HIF-1 antibodies.

Compared with heat and IVTT, hypoxia resulted in a distinct HIF-1 banding pattern. Hypoxia induced the phosphatase treatment-resistant HIF-1_b and the phosphorylated HIF-1_c species, the latter of which is not found following heat or IVTT. This observation suggests that heat induced HIF-1 at least partially by a pathway distinct from hypoxia. We further excluded heat-mediated proteasome inhibition as a potential mechanism of HIF-1 stabilization, as evidenced by the distinct pattern of HIF-1 following treatment of the cells with the proteasome inhibitor lactacystin (Fig. 6). Unlike heat, lactacystin induced a series of additional (most probably polyubiquitinylated) HIF-1 bands with lower electrophoretic mobilities than HIF-1_a, HIF-1_b, and HIF-1_c (Fig. 6).

HSP90 Activity Is Required for HIF-1 Induction by Heat as Well as by Hypoxia-- The HSP90-inhibiting drugs novobiocin and geldanamycin were used to gain insight into a possible contribution of HSP90 activity to the heat-induced expression of HIF-1. Therefore, HepG2 cells were subjected to 42 °C with or without the addition of increasing concentrations of novobiocin or geldanamycin (Fig. 7A). It is important to note that the doses of geldanamycin and novobiocin employed were not cytotoxic as determined by a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide survival assay (data not shown). At lower doses, both HSP90 inhibitors diminished the prominent heat-induced HIF-1_a form but led to an increase of the HIF-1_b form. Higher doses of both HSP90 inhibitors even led to the disappearance of heat-induced HIF-1_a and HIF-1_b. The addition of the MEK1 inhibitor PD98059 did not affect heat induction of HIF-1_a and HIF-1_b, confirming that the MEK-ERK kinase pathway is not involved in heat regulation of HIF-1 (Fig. 7A).

View larger version (47K): [in this window] [in a new window]   Fig. 7.   HSP90 activity is required for both heat and hypoxic induction of HIF-1. A, HepG2 cells were incubated at 37 or 42 °C under normoxic (20% O2) or hypoxic (3% O2) conditions for 4 h with or without the addition of the HSP90 inhibitor novobiocin and geldanamycin (GM) or the MEK1 inhibitor PD98059. B, HepG2 cells were incubated at 37 or 42 °C for 4 h under normoxic conditions. Subsequently, protein extracts were prepared, and the HIF-1/HSP90 interaction was detected by co-immunoprecipitation assays using anti-HIF-1 antibodies. C, HepG2 cells were incubated at 37 °C under hypoxic (3% O2) conditions for 4 h with or without the addition of geldanamycin. Representative immunoblot analysis of cellular protein probed with anti-HIF-1 and anti--actin antibodies.

The direct interaction of HIF-1 and HSP90 was determined by co-immunoprecipitation assays in protein extracts of HepG2 cells exposed to 37 °C compared with 42 °C. In good agreement with the hypothesis that HSP90 is mediating heat-induced stabilization of HIF-1, more HSP90 was found to be complexed with HIF-1 after exposure to 42 °C compared with 37 °C (Fig. 7B).

Of note, the HSP90 inhibitor geldanamycin was able to diminish also the hypoxia-induced expression of HIF-1 (Fig. 7C). At lower doses, the native HIF-1_a form was more susceptible to inhibition of HSP90 activity than the phosphorylated HIF-1_c form. These data indicate that protein stabilization of HIF-1 by heat as well as by hypoxia involves ATP-dependent HSP90 activity.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Early work on hypoxia-inducible gene expression suspected mechanisms comparable with those elicited by the general stress response, most prominently represented by the heat-shock response. However, with the discovery of HIF-1 it became evident that the specific hypoxic response is distinct from the heat-shock response (38). Thus, we were surprised to find heat induction of HIF-1 in mice exposed to elevated temperatures. It seems unlikely that heat-induced HIF-1 expression is simply mediated by hypoxia due to an increased oxygen consumption, since the temperature effects were similar in standard polystyrol cell culture dishes and in Petriperm dishes, in which the cells grow attached to oxygen-permeable silicone membranes. As previously shown in our laboratory, this prevents hypoxic effects caused by high cell densities or increased metabolic rates (39). More important, the protein pattern of HIF-1 induced by heat was clearly distinct from the pattern induced by hypoxia; while heat induced mainly the native form of HIF-1 (HIF-1_a), hypoxia induced a series of additional bands with lower electrophoretic mobilities (HIF-1_b and HIF-1_c), which were prone (HIF-1_c) or resistant (HIF-1_b) to phosphatase treatment. Although we can not fully exclude the possibility that secondary hypoxia plays an additional role for heat-induced HIF-1 induction in vivo, our data indicate that the mechanism of heat-induced HIF-1 expression in vivo appears to be similar to the mechanism found in vitro, since also the HIF-1_a form is induced by heat in vivo. Thus, heat-induced protein expression is mainly due to protein stabilization and to a lesser extent dependent on protein modifications. In line with this notion, Jewell et al. (5) reported that immediately after the onset of hypoxia, the high mobility (native) form of HIF-1 became first detectable on Western blots, whereas following prolonged exposure additional (modified) forms with lower electrophoretic mobility became apparent.

The treatment with the HSP90 inhibitors novobiocin or geldanamycin prevented heat-induced as well as hypoxia-induced HIF-1 expression. Geldanamycin and novobiocin specifically bind to ATP-binding pockets of HSP90, blocking the ATP-dependent maturation process of HSP90-dependent complexes in an intermediate state by interfering with the association of HSP90 with Hsc70 and finally the recruitment of the co-chaperone p23 that has been reported to stabilize the interaction between HSP90 and HSP90-regulated proteins (40, 41). Occupancy of the ATP binding pocket by ansamycin drugs prevents refolding of mature proteins and causes the degradation of HSP90-regulated proteins (42, 43). Thus, the chaperone activity of HSP90 is involved in HIF-1 stabilization, and an increase in HSP90 protein by heat might stabilize HIF-1 even under normoxic conditions.

HSP90 is a highly conserved and abundant protein in the cytosol of both eukaryotic and prokaryotic cells (22, 23). It represents the most abundant heat shock protein under normal conditions and is up-regulated by various stress conditions including heat. HSP90 functions as a general molecular chaperone in vitro and in vivo, acts in concert with many other heat shock and nonheat shock proteins, and shows a broad substrate specificity. The chaperone complex HSP90-p23 has impact on the regulation of a variety of bHLH transcription factors including AhR and Sim as well as various protein kinases such as pp60v-src, steroid hormone receptors, and the tumor suppressor p53 (24-26, 44). Many of these proteins are activated by binding of ligands. HSP90 is thought to stabilize these proteins and maintain them in an activable form. Especially for bHLH transcription factors like AhR and Sim, the release of HSP90 appears to be required for activation, since only the repressed, non-DNA-binding forms of the receptors show interaction with HSP90. Although a protein interaction of HSP90 with HIF-1 has been described previously (33, 36), the functional meaning of this interaction was not clear so far. We show for the first time that the heat-induced stabilization of HIF-1 is linked to the function of HSP90. The correlation of HSP90 and heat-induced as well as constitutive HIF-1 expression in the animal experiments suggests a general role of HSP90-mediated stabilization of HIF-1 also in vivo. In accordance with the depressing function of HSP90 for other bHLH transcription factors, the heat-induced HIF-1 was functionally inactive as demonstrated by DNA-binding and reporter gene assays as well as by the lack of target gene expression. This suggests that additional posttranslational modifications of HIF-1 are required to activate these functions.

Heat induction of HIF-1 was sufficient to induce nuclear translocation of the HIF-1_a form. Concomitantly, a tremendous induction and nuclear accumulation of HSP90 was observed. Thus, it is tempting to speculate that HSP90 itself inhibited the DNA binding activity of the native HIF-1-ARNT complex. Following hypoxic induction, however, HIF-1 becomes activated by phosphorylation (and/or other modifications), which might prevent the interaction with HSP90. Accordingly, heat did not inhibit DNA-binding or reporter gene activation under hypoxic conditions. Similar data have been reported for the function of HSP90 in glucocorticoid receptor regulation. Applying heat stress to both cell cultures or whole organisms led to considerable loss of cytosolic glucocorticoid binding capacity that coincided with a decreased amount of the glucocorticoid receptor protein present in the cytosolic fraction and its increase in the nuclei (45-47). Since we were able to demonstrate a slight nuclear accumulation of HSP90 after hypoxic versus normoxic exposure at normothermic temperature, one could postulate a physiologic role of the HSP90-HIF-1 interaction for the hypoxic activation of HIF-1. In analogy to AhR and Sim, HSP90 could support the conformational maturation of HIF-1 and protect it from proteasomal degradation. This hypothesis is further substantiated by the study of Minet et al. (33), who demonstrated that inhibition of the function of HSP90 with geldanamycin impaired CoCl₂-induced HIF-1 activation as determined by reporter gene assays in transiently transfected cells. In agreement with this model, we found that the HSP90 inhibitor geldanamycin was able to inhibit hypoxic stabilization of HIF-1.

In summary, our studies imply that HSP90 plays a functional role in HIF-1 regulation by physiologic variations of both environmental parameters, oxygen partial pressure and temperature.

	ACKNOWLEDGEMENTS

We thank J. Caro for the gift of the B1 cell line, G. Fletschinger for the artwork, and J. Fandrey and W. Jelkmann for support.

	FOOTNOTES

^* 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.

^§ Supported by grants from the Forschungsschwerpunkt Onkologie of the Medical University of Lübeck.

^¶ Supported by Deutsche Forschungsgemeinschaft Grant GRK288.

To whom correspondence should be addressed. Tel.: 49-451-500-4152; Fax: 49-451-500-4151; E-mail: katschinski@physio.mu-luebeck.de.

^§§ Supported by Deutsche Forschungsgemeinschaft Grant We2672/1-1.

^¶¶ Present address: Carl-Ludwig-Institute of Physiology, University of Leipzig, Liebigstrasse 27, D-04103 Leipzig, Germany.

Published, JBC Papers in Press, January 4, 2002, DOI 10.1074/jbc.M110377200

	ABBREVIATIONS

The abbreviations used are: HIF, hypoxia-inducible factor; AhR, arylhydrocarbon receptor; ARNT, AhR nuclear translocator; bHLH, basic-helix-loop-helix; Glut, glucose transporter; HSP, heat shock protein; IVTT, in vitro transcribed and translated; PAS, Per-Arnt-Sim; pVHL, von Hippel-Lindau protein; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; PBS, phosphate-buffered saline.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				M. Nikinmaa, L. Leveelahti, E. Dahl, E. Rissanen, K. T. Rytkonen, and A. Laurila Population origin, development and temperature of development affect the amounts of HSP70, HSP90 and the putative hypoxia-inducible factor in the tadpoles of the common frog Rana temporaria J. Exp. Biol., June 15, 2008; 211(12): 1999 - 2004. [Abstract] [Full Text] [PDF]

				B. Andriopoulos, S. Hegedusch, J. Mangin, H.-D. Riedel, U. Hebling, J. Wang, K. Pantopoulos, and S. Mueller Sustained Hydrogen Peroxide Induces Iron Uptake by Transferrin Receptor-1 Independent of the Iron Regulatory Protein/Iron-responsive Element Network J. Biol. Chem., July 13, 2007; 282(28): 20301 - 20308. [Abstract] [Full Text] [PDF]

				J. Fang, Q. Zhou, L.-Z. Liu, C. Xia, X. Hu, X. Shi, and B.-H. Jiang Apigenin inhibits tumor angiogenesis through decreasing HIF-1{alpha} and VEGF expression Carcinogenesis, April 1, 2007; 28(4): 858 - 864. [Abstract] [Full Text] [PDF]

				B. Fu, J. Xue, Z. Li, X. Shi, B.-H. Jiang, and J. Fang Chrysin inhibits expression of hypoxia-inducible factor-1{alpha} through reducing hypoxia-inducible factor-1{alpha} stability and inhibiting its protein synthesis Mol. Cancer Ther., January 1, 2007; 6(1): 220 - 226. [Abstract] [Full Text] [PDF]