1203
Journal of Cell Science 112, 1203-1212 (1999)
Printed in Great Britain © The Company of Biologists Limited 1999
JCS0249
Induction and nuclear translocation of hypoxia-inducible factor-1 (HIF-1):
heterodimerization with ARNT is not necessary for nuclear accumulation of
HIF-1α
Dmitri Chilov1, Gieri Camenisch1, Ivica Kvietikova1,*, Urs Ziegler2, Max Gassmann1 and Roland H. Wenger1,‡
1Institute
of Physiology and 2Institute of Anatomy, University of Zürich-Irchel, CH-8057 Zürich, Switzerland
*Present address: Institut de Biochimie, Université de Lausanne, Ch. des Boveresses 155, CH-1066 Epalinges, Switzerland
‡Author for correspondence (e-mail: [email protected])
Accepted 10 February; published on WWW 23 March 1999
SUMMARY
Hypoxia-inducible factor-1 (HIF-1) is a master regulator of
mammalian oxygen homeostasis. HIF-1 consists of two
subunits, HIF-1α and the aryl hydrocarbon receptor
nuclear translocator (ARNT). Whereas hypoxia prevents
proteasomal degradation of HIF-1α, ARNT expression is
thought to be oxygen-independent. We and others
previously showed that ARNT is indispensable for HIF-1
DNA-binding and transactivation function. Here, we have
used ARNT-mutant mouse hepatoma and embryonic stem
cells to examine the requirement of ARNT for
accumulation and nuclear translocation of HIF-1α in
hypoxia. As shown by immunofluorescence, HIF-1α
accumulation in the nucleus of hypoxic cells was
independent of the presence of ARNT, suggesting that
nuclear translocation is intrinsic to HIF-1α. Coimmunoprecipitation of HIF-1α together with ARNT could
be performed in nuclear extracts but not in cytosolic
INTRODUCTION
In response to reduced oxygenation, activation of the hypoxiainducible factor-1 (HIF-1) regulates transcription of several
genes involved in oxygen homeostasis (reviewed in Bunn and
Poyton, 1996; Semenza, 1998; Wenger and Gassmann, 1997).
The transcription factor HIF-1 was originally discovered as a
critical factor binding to the hypoxia-inducible 3′ enhancer of
the erythropoietin (Epo) gene. Subsequently, HIF-1 was shown
to be involved in oxygen-dependent expression of many other
genes, including vascular endothelial growth factor (VEGF),
glycolytic enzymes, glucose transporter-1 (Glut-1), transferrin,
inducible nitric oxide synthase and heme oxygenase-1 (Wenger
and Gassmann, 1997). HIF-1 is a heterodimeric complex
composed of the two basic-helix-loop-helix (bHLH) PerARNT-Sim (PAS) subunits HIF-1α and HIF-1β (Wang et al.,
1995a). As determined in HeLa cells, highest HIF-1α protein
levels are reached at 0.5% oxygen (Jiang et al., 1996b) by a
process that involves redox-dependent proteolytic stabilization
to prevent HIF-1α ubiquitinylation and rapid degradation in
fractions, implying that formation of the HIF-1 complex
occurs in the nucleus. A proteasome inhibitor and a thiolreducing agent could mimic hypoxia by inducing HIF-1α
in the nucleus, indicating that escape from proteolytic
degradation is sufficient for accumulation and nuclear
translocation of HIF-1α. During biochemical separation,
both HIF-1α and ARNT tend to leak from the nuclei in the
absence
of
either
subunit,
suggesting
that
heterodimerization is required for stable association within
the nuclear compartment. Nuclear stabilization of the
heterodimer might also explain the hypoxically increased
total cellular ARNT levels observed in some of the cell lines
examined.
Key words: Dioxin receptor, Erythropoietin, Gene expression,
Oxygen, Protein stability
proteasomes (Huang et al., 1998, 1996; Salceda and Caro,
1997).
HIF-1β is identical to the aryl hydrocarbon receptor nuclear
translocator (ARNT), which was first cloned as a
heterodimerization partner of the aryl hydrocarbon receptor
(AhR), also known as the dioxin receptor (reviewed in
Gassmann and Wenger, 1997; Hankinson, 1995; Schmidt and
Bradfield, 1996). By using an ARNT-mutant cell line
(Hepa1C4) derived from Hepa1 mouse hepatoma cells, we
(Gassmann et al., 1997; Gradin et al., 1996) and others
(Forsythe et al., 1996; Salceda et al., 1996; Wood et al., 1996)
have shown that ARNT is indispensable for HIF-1 DNA
binding and transactivation.
Mice homozygous for a targeted deletion in the gene
encoding HIF-1α are not viable and die around midgestation,
mainly due to defective vascularization, heart malformations
and failure in neuronal tube closure (Iyer et al., 1998). In
addition, embryonic stem (ES) cell-derived solid tumour
formation is also affected in the absence of HIF-1α (Carmeliet
et al., 1998; Ryan et al., 1998). Similar to HIF-1α, ARNT
1204 D. Chilov and others
deficiency is also embryonic lethal (Kozak et al., 1997;
Maltepe et al., 1997), indicating that the heterodimeric HIF-1
complex is a non-redundant master regulator of oxygen
homeostasis. Interestingly, targeted deletion of HIF-2α also
results in embryonic lethality, because the impaired function of
the organ of Zuckerkandl results in reduced catecholamine
synthesis (Tian et al., 1998).
HIF-1α protein stabilization in hypoxia is generally
considered to be the rate-limiting step in HIF-1 activation.
However, a (weaker) concomitant increase in ARNT protein
levels has also been observed by immunoblot analyses of
nuclear extracts derived from many different cell types (Iyer et
al., 1998; Jiang et al., 1997a,b, 1996a,b; Lee et al., 1997; Liu
et al., 1998; Martin et al., 1998; Wang et al., 1995a,b,c). The
significance of these observations has not been further
investigated, but hypoxically increased ARNT levels in nuclear
extracts might be related to HIF-1α nuclear translocation. To
clarify the role of ARNT in HIF-1α regulation, we analyzed
HIF-1α and ARNT levels and subcellular localization
following exposure to hypoxia in cell lines that are either wild
type or deficient for ARNT.
MATERIALS AND METHODS
Cell culture and hypoxic induction
The human HeLa epitheloid carcinoma and Hep3B hepatoma cell
lines were obtained from American Type Culture Collection (ATCC
numbers CCL-2 and HB-8064, respectively). The human LN229
glioblastoma cell line (Wenger et al., 1997) was a kind gift of E. G.
Van Meir (Lausanne, Switzerland). The mouse L929 fibroblast (ATCC
CCL-1), and the Hepa1 (also termed Hepa1c1c7), Hepa1C4 and
VT{2} hepatoma cell lines, originally developed by O. Hankinson and
co-workers (Hoffman et al., 1991), were kindly provided by V.
O’Donnall (Bern, Switzerland), L. Poellinger (Stockholm, Sweden)
and J. Caro (Philadelphia, PA), respectively. Wild-type and ARNTdeficient (Maltepe et al., 1997) R1 ES cells were kind gifts of C.
Simon (Chicago, IL). All cells were cultured in DMEM medium (high
glucose, Gibco-BRL) supplemented with 10% heat-inactivated fetal
calf serum (FCS, Boehringer-Mannheim), 100 i.u./ml penicillin, 100
µg/ml streptomycin, 1× MEM non-essential amino acids, 2 mM Lglutamine and 1 mM sodium pyruvate (all Gibco-BRL) in a
humidified atmosphere containing 5% CO2 at 37°C. Oxygen tensions
in the incubator (Forma Scientific, model 3319) were either 140 mm
Hg (20% O2 v/v, normoxia) or 7 mm Hg (1% O2 v/v, hypoxia). The
cells were exposed to hypoxia for 4 hours. Where indicated,
lactacystin (10 mM in DMSO, Calbiochem) or N-(2mercaptopropionyl)-glycine (NMPG, 1 M in H2O, Fluka) was added
to normoxic cells.
Protein extractions
Nuclear extracts and cytoplasmic fractions were prepared as described
previously (Kvietikova et al., 1995). Briefly, 1×108 cells were washed
twice, collected in ice-cold PBS and pelleted. After incubation in
buffer A (10 mM Tris-HCl, pH 7.8, 1.5 mM MgCl2, 10 mM KCl) on
ice for 10 minutes, the cells were lysed by 10 strokes of a Dounce
homogenizer and centrifuged for 10 minutes at 1500 g. Cytoplasmic
fractions were obtained by re-centrifugation of the supernatant for 10
minutes at 20,000 g. The pelleted nuclei were extracted with buffer C
(420 mM KCl, 20 mM Tris-HCl, pH 7.8, 1.5 mM MgCl2, 20%
glycerol) at 4°C for 30 minutes with gentle agitation. Immediately
before use, buffers A and C were supplemented with 0.5 mM
dithiothreitol (DTT), 1 mM Na3VO4, and a protease inhibitor cocktail
consisting of 0.4 mM phenylmethylsulfonyl fluoride and 2 µg/ml each
of leupeptin, pepstatin and aprotinin (all obtained from Sigma). The
nuclear extract was centrifuged and the supernatant was dialyzed
twice against buffer D (20 mM Tris-HCl, pH 7.8, 100 mM KCl, 0.2
mM EDTA, 20% glycerol). For total cellular lysates, the culture
medium was removed, the cells were washed twice with ice-cold PBS
and incubated in lysis buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA,
150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1%
SDS plus protease inhibitor cocktail) for 10 minutes on ice. The
lysates were sonicated and centrifuged at 700 g for 20 minutes. The
supernatant was collected and stored frozen at −80°C. Protein
concentrations were determined by the Bradford protein assay (BioRad) or the BCA assay (Pierce) using bovine serum albumin as a
standard.
Immunoblot analysis
Protein extracts were electrophoresed through SDS-polyacrylamide
gels and transferred to nitrocellulose membranes (Schleicher and
Schuell) using standard procedures. Equal protein loading and
transfer was verified by Ponceau staining (Sigma). The membrane
was blocked with 4% (w/v) instant non-fat milk powder and
incubated for 2 hours with the indicated antibodies diluted in PBS
containing 4% milk powder. An anti-HIF-1α chicken polyclonal IgY
and a mouse monoclonal IgG1 antibody (designated mgc3) were
raised against a bacterially expressed GST-HIF-1α fusion protein
containing amino acids 530-825 of human HIF-1α. The generation
and purification of these antibodies are described in more detail
elsewhere (Camenisch et al., 1999). Rabbit anti-ARNT polyclonal
IgG antibodies and the mouse anti-ARNT monoclonal antibody
2B10 (Hord and Perdew, 1994) were kindly provided by L.
Poellinger and G. H. Perdew (Pennsylvania, PA) or purchased from
Affinity BioReagents. Rabbit-anti Sp1 polyclonal IgG antibodies
were purchased from SantaCruz Biotechnology. The respective
horseradish peroxidase-coupled secondary antibodies were: rabbit
anti-chicken (Promega), goat anti-mouse (Pierce) or goat anti-rabbit
(Sigma), and Super Signal Chemiluminescent Substrate (Pierce) was
used for their detection.
Immunoprecipitation
Nuclear extracts (200 µg) and cytoplasmic fractions (2 mg) from
mouse hepatoma cell lines were incubated with the affinity-purified
anti-HIF-1α IgY antibody (30 ng) overnight on ice followed by
incubation with a rabbit anti-chicken antibody (Promega), or with a
rabbit anti-ARNT antibody. Protein A-Sepharose (Pharmacia) was
added and incubated with rotation for 30 minutes at 4°C. Following
centrifugation of the precipitate at 15,000 g, the pellet was washed
three times with either buffer A or buffer C (see above) for
cytoplasmic and nuclear fractions, respectively, and twice with 10 mM
Tris-HCl, pH 7.5. Finally, the precipitates were analyzed by
immunoblotting using either the anti-HIF-1α IgY or the anti-ARNT
2B10 antibody.
Electrophoretic mobility shift assay (EMSA)
An oligonucleotide probe derived from the Epo 3′ enhancer was
purified and labelled as described previously (Kvietikova et al., 1995).
DNA-protein binding reactions were carried out for 20 minutes at 4°C
in a total volume of 20 µl containing 4-5 µg of nuclear extract, 0.10.4 µg of sonicated, denatured calf thymus DNA (Sigma) and 1×104
cpm of oligonucleotide probe in 10 mM Tris-HCl, pH 7.5, 50 mM
KCl, 50 mM NaCl, 1 mM MgCl2, 1 mM EDTA, 5 mM DTT and 5%
glycerol and run on 4% non-denaturating polyacrylamide gels.
Electrophoresis was performed at 200 V in TBE buffer (89 mM Tris,
89 mM boric acid, 5 mM EDTA) at 4°C and dried gels were
autoradiographed. In vitro transcription and translation of ARNT were
performed using the plasmid pBSArntKS+ (Hoffman et al., 1991) and
the TNT® coupled transcription-translation rabbit reticulocyte lysate
system (Promega) under conditions recommended by the
manufacturers. For supershift analysis, 1 µl of anti-HIF-1α mgc3
Subcellular localization of HIF-1α and ARNT 1205
hybridoma supernatant was added to the completed EMSA reaction
mixture and incubated for 16 hours at 4°C prior to loading.
Indirect immunofluorescence microscopy
Adherent cells were fixed with 4% formaldehyde in PBS (pH 8.0) for
10 minutes, washed three times with PBS, permeabilized with 0.5%
Triton X-100 in PBS for 5 minutes and rinsed again three times with
PBS (all at room temperature). After blocking nonspecific binding by
adding 10% FCS in PBS for 30 minutes, the cells were incubated with
either the affinity-purified anti-HIF-1α IgY (1-2 ng/µl in PBS) or the
monoclonal anti-ARNT 2B10 antibody at 37°C for 1 hour. Primary
antibodies were detected by incubation with either an FITCconjugated rabbit anti-chicken antibody (Promega) or Cy3-conjugated
donkey anti-mouse antibody (Jackson Immunoresearch Laboratory) at
room temperature for 30 minutes. Following extensive washing in
PBS and mounting in Dabco solution (Sigma), cells were analyzed by
confocal laser scanning microscopy (CLSM-310, Zeiss). For antibody
depletion experiments, the IgY antibody was incubated with a 100fold excess of GST-HIF-1α fusion protein for 30 minutes on ice. The
solution was centrifuged at 10,000 g for 30 minutes and the
supernatant was used to stain the cells as described above.
RESULTS
ARNT is induced in nuclear extracts but not in total
cellular lysates from cells exposed to hypoxia
HIF-1α and ARNT protein levels were examined in
cytoplasmic fractions and nuclear extracts derived from HeLa,
Hep3B, LN229, L929 and Hepa1 cells that were cultured at
either normoxia or hypoxia. The samples were analyzed by
immunoblotting using an affinity-purified chicken anti-HIF-1α
IgY or a rabbit anti-ARNT antibody. As shown in Fig. 1A,
Fig. 1. Immunoblot analysis of HIF-1α and ARNT expression in
subcellular fractions of various human and mouse cell lines. (A) The
cell lines were cultured under normoxic (20% O2) or hypoxic (1%
O2) conditions for 4 hours, and 25 µg of protein from nuclear
extracts (n) or cytoplasmic fractions (c) were separated by SDSPAGE followed by immunoblotting and detection with the polyclonal
anti-HIF-1α IgY antibody (upper panel) or the rabbit anti-ARNT
antibody (lower panel). (B) ARNT expression analyzed in total
cellular lysates (50 µg). Ponceau staining of the membrane prior
incubation with antibodies confirmed equal loading and blotting
efficiency.
HIF-1α protein was detected exclusively in nuclear extracts of
hypoxically induced cells. No additional bands were observed,
confirming the specificity of the IgY antibody (not shown).
While ARNT was not detectable in normoxic nuclear extracts
of HeLa, Hep3B and Hepa1 cells, it was found in LN229 and
L929 cells. Unexpectedly, following exposure to hypoxia,
increased ARNT protein levels were found in nuclear extracts
of all cell lines. No HIF-1α and ARNT could be detected in
the same amount of cytoplasmic protein derived from the same
cell lines (Fig. 1A). We next tested whether nuclear ARNT
accumulation in hypoxia is due to increased cellular ARNT
protein concentrations. As shown in Fig. 1B, ARNT levels in
total cellular lysates remained unchanged after hypoxic
induction of the human cell lines HeLa, Hep3B and LN229,
whereas ARNT was repeatedly induced in the mouse L929 and
Hepa1 cell lines. HIF-1α was not detectable in normoxic total
cellular lysates (not shown).
Taken together, these results might imply that hypoxic nuclear
ARNT accumulation was a result of translocation from the
cytoplasm to the nucleus. At least in some cell lines, increased
ARNT production and/or reduced degradation also contributes
to the elevated ARNT levels in nuclear extracts. However, the
fact that no cytoplasmic ARNT could be detected in normoxia
Fig. 2. Immunoblot analysis of HIF-1α and ARNT expression in
wild-type and ARNT-mutant mouse hepatoma cells. (A) Wild-type
Hepa-1 and ARNT-mutant Hepa1C4 cells were exposed for 4 hours
to hypoxic conditions, and 50 µg each of nuclear extracts (n) and
cytoplasmic fractions (c) were analyzed by immunoblotting. Note
that the amount of loaded protein was increased twofold compared to
Fig. 1, allowing the detection of cytoplasmic HIF-1α. Representative
results from a single immunoblot are shown. (B) Hepa1 and
Hepa1C4 cells were exposed to hypoxia, and the nuclei extracted
with increasing NaCl concentrations. Following incubation with the
anti-HIF-1α antibody, the membrane was incubated with an anti-Sp1
control antibody. Ponceau staining of the membranes prior to
incubation with antibodies confirmed equal loading and blotting
efficiency.
1206 D. Chilov and others
Fig. 3. Electrophoretic mobility shift assay (EMSA) of nuclear
extracts derived from wild-type and ARNT-mutant mouse hepatoma
cells. Nuclear fractions were prepared from normoxic or hypoxic
wild-type Hepa1 and ARNT-mutant Hepa1C4 cells. Nuclear protein
extracts (5 µg) were incubated with a 32P-labelled oligonucleotide
probe carrying the HIF-1 binding site derived from the Epo 3′
enhancer. The HIF-1 complex is supershifted by the monoclonal antiHIF-1α antiboy mgc3. HIF-1 DNA-binding activity in hypoxic
ARNT-mutant Hepa1C4 nuclear extracts is reconstituted following
addition of in vitro translated wild-type ARNT protein. Migration
positions of hypoxia-inducible (HIF-1), constitutive (ATF-1/CREB-1
family members) and nonspecific factors are indicated.
provides evidence against the ‘translocation hypothesis’. Thus,
we assume that nuclear ARNT was lost during isolation (but
before extraction) of the nuclei rather than being induced by
hypoxia. This would suggest a higher affinity of ARNT for the
nuclear compartment in hypoxia (i.e. in the presence of HIF-1α)
than in normoxia (i.e. in the absence of HIF-1α).
Nuclear HIF-1α accumulation in an ARNT-mutant
hepatoma cell line
To analyze whether ARNT is necessary for hypoxic induction
and nuclear accumulation of HIF-1α, we used an ARNTmutant Hepa1-derived subline (termed Hepa1C4) that was
originally selected for its resistance against 3,4-benzopyrene
treatment (Hoffman et al., 1991). We and others previously
showed that in Hepa1C4 cells (1) ARNT mRNA levels were
decreased, (2) hypoxic response of endogenous as well as
reporter gene transcription was impaired and (3) no functional
HIF-1 DNA-binding activity was detected (Forsythe et al.,
1996; Gassmann et al., 1997; Gradin et al., 1996; Salceda et
al., 1996; Wenger et al., 1998; Wood et al., 1996). As shown
in Fig. 2A, ARNT protein was also hardly detectable in
nuclear extracts from normoxic and hypoxic Hepa1C4 cells.
Interestingly, HIF-1α was still induced in nuclear extracts
from Hepa1C4 cells, although at lower levels compared to
wild-type Hepa1 cells. Reduced HIF-1α levels in nuclear
extracts from Hepa1C4 cells were paralleled by increased
HIF-1α levels in the cytoplasmic fraction (Fig. 2A). As for
ARNT, these findings suggest that the affinity of HIF-1α for
the nuclear compartment is higher in the presence of its
heterodimerization partner. To directly assess this hypothesis,
a titration experiment was performed, using increasing
concentrations of NaCl to extract the nuclei. As shown in Fig.
2B, at least 300 mM NaCl was necessary to extract HIF-1α
from wild-type cells, whereas in ARNT-mutant cells, HIF-1α
had already leaked out of the nuclei in the absence of NaCl.
In addition, HIF-1α was more efficiently extracted in
Hepa1C4 than in Hepa1 cells with all the NaCl concentrations
used. Correspondingly, the HIF-1α content was much lower
in cytoplasmic fractions from Hepa1 than from Hepa1C4
cells. In contrast, the unrelated transcription factor Sp-1 was
efficiently extracted at 300 mM NaCl in both cell lines. Taken
together, these results suggest that ARNT is required to retain
HIF-1α in the nucleus but not for HIF-1α nuclear
translocation.
Fig. 4. Co-immunoprecipitation of HIF1α:ARNT complexes from nuclear
extracts of hypoxic hepatoma cells.
Nuclear extracts (n, 200 µg) and
cytoplasmic fractions (c, 2 mg) were
prepared from normoxic and hypoxic
wild-type Hepa1 and ARNT-mutant
Hepa1C4 cells. Following incubation
with either the anti-HIF-1α IgY
antibody, followed by a rabbit anti-IgY
antibody or the rabbit anti-ARNT
antibody, the immune complexes were
precipitated by addition of protein A sepharose. The precipitates were analyzed by immunoblotting using the monoclonal anti-ARNT
2B10 antibody or the polyclonal anti-HIF-1α IgY antibody.
Subcellular localization of HIF-1α and ARNT 1207
Reconstitution of HIF-1 DNA-binding activity in
nuclear extracts from Hepa1C4 cells
To prove that nuclear HIF-1α from Hepa1C4 cells is capable
of forming a functional complex with ARNT, EMSAs were
performed using an Epo 3′ enhancer-derived HIF-1 DNAbinding oligonucleotide (Kvietikova et al., 1995). Hepa1C4
nuclear extracts were complemented with exogenous ARNT
protein that was translated in vitro using reticulocyte lysates.
As shown in Fig. 3, synthetic ARNT could reconstitute HIF-1
DNA-binding despite a general inhibitory effect of the
reticulocyte lysate on DNA-binding activity of HIF-1 and ATF1/CREB-1 family members, which assemble the constitutive
factor (Kvietikova et al., 1995). Supershift experiments using
the monoclonal anti-HIF-1α antibody mgc3 confirmed the
identity of the HIF-1 band (Fig. 3). Unprogrammed
reticulocyte lysate lacking HIF-1α and ARNT cDNAs did not
yield any detectable bands (not shown).
Formation of the HIF-1 heterodimer in the nucleus
To assess the site of HIF-1 heterodimer formation, coimmunoprecipitation experiments were performed using the
affinity-purified chicken anti-HIF-1α IgY or the rabbit antiARNT antibody. Since chicken IgY antibodies have a very low
affinity for protein A or G, a rabbit anti-IgY antibody was used
as an intermediary for the precipitation with protein Asepharose beads. The precipitates were analyzed by
immunoblotting using either the chicken anti-HIF-1α IgY or the
mouse anti-ARNT 2B10 antibody. As shown in Fig. 4 (left),
ARNT was co-immunoprecipitated together with HIF-1α from
200 µg hypoxic nuclear extract but not from 2 mg cytoplasmic
fraction of Hepa1 cells. On the other hand, HIF-1α was coimmunoprecipitated together with ARNT from 200 µg hypoxic
nuclear extract but not from 2 mg cytoplasmic fraction of Hepa1
cells (Fig. 4, right). Since this represents 40-fold more protein
than minimally required to detect cytoplasmic HIF-1α in an
Fig. 5. Indirect
immunofluorescence analysis of
HIF-1α expression in normoxic and
hypoxic human and mouse cell
lines. The various cell lines were
cultured under normoxic or hypoxic
conditions for 4 hours, and prepared
for immunofluorescence as
described in Materials and methods.
The cells were then incubated with
the affinity-purified anti-HIF-1α
IgY antibody followed by a FITCconjugated rabbit anti-chicken
antibody and analyzed by CLSM
(HeLa: b,c; Hep3B: f,g; Hepa1: j,k;
Hepa1C4: n,o; VT{2}: r,s). The
corresponding transmission images
are shown in the outer columns.
Normoxic and hypoxic Hepa1 cells
were stained with either the antiHIF-1α IgY (j,k), or with the IgY
antibody that had been depleted of
specific anti-HIF-1α antibodies by
pre-incubation with GST-HIF-1α
fusion protein (v,w).
1208 D. Chilov and others
Fig. 6. Indirect immunofluorescence analysis of ARNT expression in normoxic and hypoxic mouse and human cell lines. Wild-type Hepa1
(b,c), ARNT-mutant Hepa1C4 (f,g) and HeLa cells (j,k) were incubated with the monoclonal anti-ARNT antibody 2B10 followed by a Cy3conjugated donkey anti-mouse antibody. The corresponding transmission images are shown in the outer columns.
immunoblot (see Fig. 2A), these results suggest that HIF1α:ARNT heterodimer formation occurs mainly in the nucleus.
No co-immunoprecipitations were obtained from normoxic
Hepa1 or normoxic and hypoxic Hepa1C4 extracts.
Subcellular localization of HIF-1α in wild-type and
ARNT-mutant hepatoma cells
Since HIF-1α was prone to leak from the nuclei during
biochemical separation in the absence of ARNT (see above),
we examined the subcellular localization of HIF-1α by
indirect immunofluorescence and confocal laser scanning
microscopy (CLSM). Normoxic and hypoxic HeLa, Hep3B,
wild-type Hepa1, ARNT-mutant Hepa1C4 as well as ARNTreconstituted VT{2} (Hoffman et al., 1991) cells were
analyzed using the chicken anti-HIF-1α IgY antibody. In
normoxically cultured cells, very weak immunofluorescence
was detected in the cytoplasm and nuclei of HeLa and
Hep3B cells (Fig. 5b,f), but not in the other cell lines (Fig.
5j,n,r). Upon hypoxic exposure, all cell lines, including
ARNT-mutant Hepa1C4 cells, responded with a drastic
increase in HIF-1α levels in the nucleus but not in the
nucleoli (Fig. 5c,g,k,o,s), demonstrating that ARNT is not
required for hypoxic accumulation and nuclear translocation
of HIF-1α. In contrast to the immunoblot results (Fig. 2A),
HIF-1α was not detectable in the cytoplasm of hypoxic
Hepa1C4 cells (Fig. 5o), supporting our notion that the
presence of HIF-1α in cytoplasmic fractions is mainly due
to leakage from the nucleus. In human HeLa and Hep3B
cells, the cytoplasmic fluorescence was brighter than that of
mouse Hepa1 cells. To ensure that cytoplasmic staining in
HeLa and Hep3B cells was not an experimental artifact,
HeLa cells were stained using the monoclonal anti-HIF-1α
antibody mgc3, and a similar cytoplasmic fluorescence was
observed (data not shown). To confirm the specificity of the
chicken anti-HIF-1α IgY antibody, it was pre-absorbed with
recombinant GST-HIF-1α fusion protein and removed by
centrifugation. As demonstrated in Fig. 5w, this treatment
completely abolished binding of the antibody to hypoxic
Hepa1 cells.
Induction and subcellular localization of ARNT in
hypoxia
It has previously been reported that, unlike the dioxin receptor,
ARNT levels remain spatially and temporarily constant in the
nucleus following treatment with dioxin-analogs (Pollenz,
1996). In contrast, in our immunoblotting experiments (see
Figs 1A, 2A), as well as in those of others (see Introduction),
ARNT seemed to be regulated by hypoxia. Because these data
might be compromised by artifacts occurring during
biochemical
separation,
we
performed
indirect
immunofluorescence experiments using the anti-ARNT
monoclonal antibody 2B10. As shown in Fig. 6b,c, the ARNT
level in the nucleus of Hepa1 cells increased following
exposure to hypoxia. A nonspecific cross-reactivity of the
2B10 antibody gave rise to background signals in Hepa1 and
Hepa1C4 cells, which overwhelmed normoxic ARNT signals
in Hepa1 cells. Opposite to Hepa1 cells, no hypoxic ARNT
induction was observed in HeLa cells (Fig. 6j,k). These results
confirm our immunoblot data using total cellular extracts (Fig.
1B), and show that (1) ARNT is a nuclear protein and its
disappearance from normoxic nuclear extracts is due to leakage
during preparation of the nuclei, and (2) certain cell lines are
Subcellular localization of HIF-1α and ARNT 1209
capable of inducing ARNT protein levels in a hypoxiadependent manner.
Subcellular localization of HIF-1α in wild-type and
ARNT-deficient embryonic stem cells
An ARNT point mutation (Gly326→Asp) in Hepa1C4 cells
has recently been reported, which leads to increased
proteolytic susceptibility and a higher turnover rate
(Numayama-Tsuruta et al., 1997). Despite reduced ARNT
mRNA and protein levels, as well as lack of HIF-1 DNAbinding and HIF-1-mediated gene activation (Gassmann et
al., 1997; Gradin et al., 1996; Wenger et al., 1998), this
ARNT mutation did not abrogate its ability to translocate to
the nucleus and to form heterodimers with AhR (NumayamaTsuruta et al., 1997). Thus, residual amounts of ARNT might
have been responsible for the nuclear localization of HIF-1α
observed in Hepa1C4 cells (Fig. 5). To rule out this
possibility, ES cells in which the ARNT gene has been
targeted by homologous recombination (Maltepe et al., 1997)
were examined. As shown in Fig. 7A, HIF-1α was detected
by immunoblotting in nuclear extracts of hypoxic ARNT+/+
as well as ARNT−/− ES cells, albeit at a lower level in the
latter cell line. Nuclear ARNT levels were increased in
hypoxic ARNT+/+ ES cells, but, as expected, could not be
detected in ARNT−/− ES cells. Staining of ARNT+/+ and
ARNT−/− ES cells with the anti-HIF-1α IgY antibody and
analysis by CLSM revealed a weak homogenous
immunofluorescence in normoxic cells (Fig. 7Bb,f). In
hypoxic wild-type ES cells, HIF-1α accumulated in the nonnucleolar part of the nuclei (Fig. 7Bc,d). Nuclear
accumulation of HIF-1α was also clearly observed in hypoxic
ARNT−/− ES cells. As can be seen in Fig. 7Bg, HIF-1α
signals were variable from cell to cell. However, using
CLSM, scanning through multiple horizontal sections of the
cell aggregates revealed that the HIF-1α was present in the
nuclei of a majority of the cells (data not shown). Thus, these
data confirmed that whereas hypoxic HIF-1α accumulation
and nuclear translocation is ARNT-independent, HIF-1α
stability in the nucleus during isolation of the nuclei is
lowered in the absence of ARNT.
Subcellular localization of HIF-1α following
proteasome inhibition and thiol reduction
Salceda and Caro (1997) reported that the proteasome inhibitor
lactacystin
and
the
thiol-reducing
agent
N-(2mercaptopropionyl)-glycine (NMPG) induced HIF-1α protein
in normoxic cells. It is conceivable that additional hypoxiainducible post-stabilization steps (e.g. phosphorylation) might
Fig. 7. Hypoxic HIF-1α nuclear accumulation in ARNT+/+ and ARNT−/− embryonic stem (ES) cells. (A) Nuclear extracts (n) and cytoplasmic
fractions (c) from normoxic and hypoxic ARNT+/+ and ARNT−/− ES cells were analyzed by immunoblotting using the anti-HIF-1α IgY or the
rabbit anti-ARNT antibody. (B) Indirect immunofluorescence analysis of normoxic and hypoxic ARNT+/+ (b,c) and ARNT−/− (f,g) ES cells
using the anti-HIF-1α IgY antibody. The corresponding phase-contrast images are shown in the outer columns.
1210 D. Chilov and others
Fig. 8. Hypoxic nuclear accumulation of HIF-1α can
be mimicked by lactacystin and NMPG. (A) HeLa
cells were exposed for 6 hours to either 1% O2, 10,
20 or 80 µM lactacystin, or to 1 mM NMPG, and
analyzed by immunoblotting using the anti-HIF-1α
IgY antibody. (B) Indirect immunofluorescence
analysis of HeLa cells exposed for 6 hours to either
1% O2 (hypoxia), 20 µM lactacystin or 1 mM
NMPG. Despite the stronger induction of lactacystin
at 80 µM (Fig. 8A), this concentration was not used
because it was toxic and interfered with the structural
integrity of the cells (not shown). The cells were
prepared and stained with the anti-HIF-1α IgY
antibody as described in Materials and methods.
be required for nuclear translocation. Therefore, we first
assessed HIF-1α protein induction in total cellular lysates
derived from HeLa cells. As shown in Fig. 8A, lactacystin and
NMPG induced a series of HIF-1α isoforms with different
mobilities compared to hypoxically induced HIF-1α,
indicating alternative post-translational modifications.
However, nuclear accumulation of HIF-1α was observed by
immunofluorescence as a result of the treatment with all of
these stimuli (Fig. 8B). In addition, we found that the iron
chelator desferrioxamine induced protein stability and nuclear
accumulation of HIF-1α (data not shown). Thus, hypoxia as
well as hypoxia-mimicking agents that interfere with the
process of protein modification/degradation lead to an increase
in protein levels that is sufficient for nuclear translocation of
HIF-1α.
DISCUSSION
In the present study, we used ARNT-mutant Hepa1C4 cells
(Numayama-Tsuruta et al., 1997) as well as ARNT-deficient
ES cells (Maltepe et al., 1997) to demonstrate that HIF-1α
protein accumulation and nuclear translocation can occur
independently of ARNT. This finding is not compatible with
the term ‘nuclear translocator’ in ‘ARNT’. There is a similar
discrepancy between biochemical fractionation and
immunofluorescence data, as we show in this work for the HIF1α:ARNT complex, which has also been reported for the
AhR:ARNT complex (reviewed in Hankinson, 1995; Schmidt
and Bradfield, 1996). In Hepa1, but not in Hepa1C4 cells, the
liganded AhR can be purified from the nucleus. The
complementing ARNT was hence believed to be responsible
for nuclear translocation of AhR. That this model is wrong has
been demonstrated by immunofluorescence experiments
showing that AhR and ARNT are translocated to the nucleus
independently (Hord and Perdew, 1994; Pollenz et al., 1994),
and that following ligand binding the AhR translocates to the
nucleus where it is rapidly depleted, whereas ARNT remains
at constant levels in the nucleus (Pollenz, 1996). Thus, upon
homogenization almost all of the ARNT present in a cell is
found outside of the nuclei, apparently because it leaches out
of the nuclei (Hord and Perdew, 1994).
A plausible explanation for the failure of subcellular
fractionation methods to accurately preserve the distribution of
AhR and ARNT observed in vivo might be that heterodimer
formation in the nucleus renders the complex more stably
associated with the nuclear compartment during isolation of the
nuclei. Hence monomeric subunits are more prone to leakage.
In this work, we found partial loss of both HIF-1 subunits
during isolation of the nuclei in the absence of one of the
respective heterodimerization partners in Hepa1C4 and ES
cells (lacking ARNT) or under normoxic conditions (lacking
HIF-1α). Whether this is due to interaction of the heterodimer
with the nuclear scaffold and/or specific DNA binding is a
matter of further investigations. In this context, it is noteworthy
that Poellinger and coworkers reported a conformational
change of HIF-1α upon interaction with ARNT, as shown by
altered resistance to proteolytic digestion in vitro (Kallio et al.,
1997). Even if this effect might be due to steric hindrance, it
provides evidence that increased proteolytic stability in vivo
during isolation of the nuclei could also contribute to the
difference observed in monomeric versus heterodimeric
protein levels of the HIF-1 subunits.
Our findings on ARNT-independent HIF-1α induction in
Hepa1C4 and ARNT−/− ES cells are in striking contrast to a
recent publication where the same cell lines were used as HIF1α-deficient cell culture models (An et al., 1998). Therein, the
authors claimed that HIF-1α hypoxically stabilizes wild-type
Subcellular localization of HIF-1α and ARNT 1211
p53 protein, as substantiated by the lack of hypoxic p53
induction in the ARNT-deficient cell lines. However, since
hypoxia normally induces HIF-1α in these cells, the simplified
model of increased p53 due to interaction with HIF-1α needs
refinement (Wenger et al., 1998).
Recently, nuclear localization signals (NLS) were identified
in the amino termini of AhR, ARNT and HIF-1α as well as a
second NLS in the carboxy-terminal part of HIF-1α (Eguchi
et al., 1997; Ikuta et al., 1998; Kallio et al., 1998), consistent
with their ability to translocate to the nucleus independently.
However, these data were based on transient transfection
experiments with fusion protein vectors, conditions that lead
to unphysiological accumulation of products already in
normoxia, probably because of the high copy number of
transfected plasmids or because the fusion partners may
stabilize the HIF-1α protein (Kallio et al., 1998). In contrast,
we showed that endogenous HIF-1α can accumulate and
translocate to the nucleus in hypoxic conditions only.
Interestingly, a nuclear export signal was also identified in the
conserved bHLH region of AhR (Ikuta et al., 1998). Given the
rapid degradation of HIF-1α upon reoxygenation (Huang et al.,
1996), it will be important to determine whether HIF-1α
degradation occurs inside or outside the nucleus, probably
involving a nuclear export signal.
We are grateful to E. G. Van Meir, O. Hankinson, V. O’Donnall, L.
Poellinger, J. Caro, C. Simon and G. H. Perdew for the generous gift
of cell lines, plasmids and antibodies; to P. Spielmann and F. Parpan
for excellent technical assistance; to S. Kunz for advice; to D. Stroka
and I. Desbaillets for critically reading the manuscript; to C. Gasser
for the artwork; and to C. Bauer for support. This work was supported
by the Swiss National Science Foundation (31-47111.96) and the
Wolfermann-Nägeli-Stiftung. R.H.W. is a recipient of the
‘Sondermassnahmen des Bundes zur Förderung des akademischen
Nachwuchses’.
REFERENCES
An, W. G., Kanekal, M., Simon, M. C., Maltepe, E., Blagosklonny, M. V.
and Neckers, L. M. (1998). Stabilization of wild-type p53 by hypoxiainducible factor 1α. Nature 392, 405-408.
Bunn, H. F. and Poyton, R. O. (1996). Oxygen sensing and molecular
adaptation to hypoxia. Physiol. Rev. 76, 839-885.
Camenisch, G., Tini, M., Chilov, D., Kvietikova, I., Srinivas, V., Caro, J.,
Spielmann, P., Wenger, R. H. and Gassmann, M. (1999). General
applicability of chicken egg yolk antibodies: the performance of IgY
immunoglobulins raised against the hypoxia-inducible factor 1α. FASEB J.
13, 81-88.
Carmeliet, P., Dor, Y., Herbert, J. M., Fukumura, D., Brusselmans, K.,
Dewerchin, M., Neeman, M., Bono, F., Abramovitch, R., Maxwell, P. et
al. (1998). Role of HIF-1α in hypoxia-mediated apoptosis, cell proliferation
and tumour angiogenesis. Nature 394, 485-490.
Eguchi, H., Ikuta, T., Tachibana, T., Yoneda, Y. and Kawajiri, K. (1997).
A nuclear localization signal of human aryl hydrocarbon receptor nuclear
translocator/hypoxia-inducible factor 1β is a novel bipartite type recognized
by the two components of nuclear pore-targeting complex. J. Biol. Chem.
272, 17640-17647.
Forsythe, J. A., Jiang, B. H., Iyer, N. V., Agani, F., Leung, S. W., Koos, R.
D. and Semenza, G. L. (1996). Activation of vascular endothelial growth
factor gene transcription by hypoxia-inducible factor 1. Mol. Cell. Biol. 16,
4604-4613.
Gassmann, M., Kvietikova, I., Rolfs, A. and Wenger, R. H. (1997). Oxygenand dioxin-regulated gene expression in mouse hepatoma cells. Kidney Int.
51, 567-574.
Gassmann, M. and Wenger, R. H. (1997). HIF-1, a mediator of the molecular
response to hypoxia. News Physiol. Sci. 12, 214-218.
Gradin, K., McGuire, J., Wenger, R. H., Kvietikova, I., Whitelaw, M. L.,
Toftgård, R., Tora, L., Gassmann, M. and Poellinger, L. (1996).
Functional interference between hypoxia and dioxin signal transduction
pathways: competition for recruitment of the Arnt transcription factor. Mol.
Cell. Biol. 16, 5221-5231.
Hankinson, O. (1995). The aryl hydrocarbon receptor complex. Annu. Rev.
Pharmacol. Toxicol. 35, 307-340.
Hoffman, E. C., Reyes, H., Chu, F. F., Sander, F., Conley, L. H., Brooks,
B. A. and Hankinson, O. (1991). Cloning of a factor required for activity
of the Ah (dioxin) receptor. Science 252, 954-958.
Hord, N. G. and Perdew, G. H. (1994). Physicochemical and
immunocytochemical analysis of the aryl hydrocarbon receptor nuclear
translocator: characterization of two monoclonal antibodies to the aryl
hydrocarbon receptor nuclear translocator. Mol. Pharmacol. 46, 618-626.
Huang, E. L., Gu, J., Schau, M. and Bunn, H. F. (1998). Regulation of
hypoxia-inducible factor 1α is mediated by an oxygen-dependent domain
via the ubiquitin-proteasome pathway. Proc. Natl. Acad. Sci. USA 95, 79877992.
Huang, L. E., Arany, Z., Livingston, D. M. and Bunn, H. F. (1996).
Activation of hypoxia-inducible transcription factor depends primarily upon
redox-sensitive stabilization of its α subunit. J. Biol. Chem. 271, 3225332259.
Ikuta, T., Eguchi, H., Tachibana, T., Yoneda, Y. and Kawajiri, K. (1998).
Nuclear localization and export signals of the human aryl hydrocarbon
receptor. J. Biol. Chem. 273, 2895-2904.
Iyer, N. V., Kotch, L. E., Agani, F., Leung, S. W., Laughner, E., Wenger,
R. H., Gassmann, M., Gearhart, J. D., Lawler, A. M., Yu, A. Y. et al.
(1998). Cellular and developmental control of O2 homeostasis by hypoxiainducible factor 1α. Genes Dev. 12, 149-162.
Jiang, B. H., Agani, F., Passaniti, A. and Semenza, G. L. (1997a). V-SRC
induces expression of hypoxia-inducible factor 1 (HIF-1) and transcription
of genes encoding vascular endothelial growth factor and enolase 1:
involvement of HIF-1 in tumor progression. Cancer Res. 57, 5328-5335.
Jiang, B. H., Rue, E., Wang, G. L., Roe, R. and Semenza, G. L. (1996a).
Dimerization, DNA binding, and transactivation properties of hypoxiainducible factor 1. J. Biol. Chem. 271, 17771-17778.
Jiang, B. H., Semenza, G. L., Bauer, C. and Marti, H. H. (1996b). Hypoxiainducible factor 1 levels vary exponentially over a physiologically relevant
range of O2 tension. Am. J. Physiol. 271, C1172-1180.
Jiang, B. H., Zheng, J. Z., Leung, S. W., Roe, R. and Semenza, G. L.
(1997b). Transactivation and inhibitory domains of hypoxia-inducible factor
1α. Modulation of transcriptional activity by oxygen tension. J. Biol. Chem.
272, 19253-19260.
Kallio, P. J., Okamoto, K., O’Brien, S., Carrero, P., Makino, Y., Tanaka,
H. and Poellinger, L. (1998). Signal transduction in hypoxic cells: inducible
nuclear translocation and recruitment of the CBP/p300 coactivator by the
hypoxia-inducible factor-1α. EMBO J. 17, 6573-6586.
Kallio, P. J., Pongratz, I., Gradin, K., McGuire, J. and Poellinger, L.
(1997). Activation of hypoxia-inducible factor 1α: posttranscriptional
regulation and conformational change by recruitment of the Arnt
transcription factor. Proc. Natl. Acad. Sci. USA 94, 5667-5672.
Kozak, K. R., Abbott, B. and Hankinson, O. (1997). Arnt-deficient mice and
placental differentiation. Dev. Biol. 191, 297-305.
Kvietikova, I., Wenger, R. H., Marti, H. H. and Gassmann, M. (1995). The
transcription factors ATF-1 and CREB-1 bind constitutively to the hypoxiainducible factor-1 (HIF-1) DNA recognition site. Nucl. Acids Res. 23, 45424550.
Lee, P. J., Jiang, B. H., Chin, B. Y., Iyer, N. V., Alam, J., Semenza, G. L.
and Choi, A. M. (1997). Hypoxia-inducible factor-1 mediates
transcriptional activation of the heme oxygenase-1 gene in response to
hypoxia. J. Biol. Chem. 272, 5375-5381.
Liu, Y. X., Christou, H., Morita, T., Laughner, E., Semenza, G. L. and
Kourembanas, S. (1998). Carbon monoxide and nitric oxide suppress the
hypoxic induction of vascular endothelial growth factor gene via the 5′
enhancer. J. Biol. Chem. 273, 15257-15262.
Maltepe, E., Schmidt, J. V., Baunoch, D., Bradfield, C. A. and Simon, M.
C. (1997). Abnormal angiogenesis and responses to glucose and oxygen
deprivation in mice lacking the protein ARNT. Nature 386, 403-407.
Martin, C., Yu, A. Y., Jiang, B. H., Davis, L., Kimberly, D., Hohimer, A.
R. and Semenza, G. L. (1998). Cardiac hypertrophy in chronically anemic
fetal sheep – increased vascularization is associated with increased
myocardial expression of vascular endothelial growth factor and hypoxiainducible factor 1. Am. J. Obst. Gynecol. 178, 527-534.
Numayama-Tsuruta, K., Kobayashi, A., Sogawa, K. and Fujii-Kuriyama,
1212 D. Chilov and others
Y. (1997). A point mutation responsible for defective function of the arylhydrocarbon-receptor nuclear translocator in mutant Hepa-1c1c7 cells. Eur.
J. Biochem. 246, 486-495.
Pollenz, R. S. (1996). The aryl-hydrocarbon receptor, but not the arylhydrocarbon receptor nuclear translocator protein, is rapidly depleted in
hepatic and nonhepatic culture cells exposed to 2,3,7,8-tetrachlorodibenzop-dioxin. Mol. Pharmacol. 49, 391-398.
Pollenz, R. S., Sattler, C. A. and Poland, A. (1994). The aryl hydrocarbon
receptor and aryl hydrocarbon receptor nuclear translocator protein show
distinct subcellular localizations in Hepa 1c1c7 cells by
immunofluorescence microscopy. Mol. Pharmacol. 45, 428-438.
Ryan, H. E., Lo, J. and Johnson, R. S. (1998). HIF-1α is required for
solid tumor formation and embryonic vascularization. EMBO J. 17, 30053015.
Salceda, S., Beck, I. and Caro, J. (1996). Absolute requirement of aryl
hydrocarbon receptor nuclear translocator protein for gene activation by
hypoxia. Arch. Biochem. Biophys. 334, 389-394.
Salceda, S. and Caro, J. (1997). Hypoxia-inducible factor 1α (HIF-1α)
protein is rapidly degraded by the ubiquitin-proteasome system under
normoxic conditions. Its stabilization by hypoxia depends on redox-induced
changes. J. Biol. Chem. 272, 22642-22647.
Schmidt, J. V. and Bradfield, C. A. (1996). Ah receptor signaling pathways.
Ann. Rev. Cell. Dev. Biol. 12, 55-89.
Semenza, G. L. (1998). Hypoxia-inducible factor 1: master regulator of O2
homeostasis. Curr. Opin. Genet. Dev. 8, 588-594.
Tian, H., Hammer, R. E., Matsumoto, A. M., Russell, D. W. and McKnight,
S. L. (1998). The hypoxia-responsive transcription factor EPAS1 is essential
for catecholamine homeostasis and protection against heart failure during
embryonic development. Genes Dev. 12, 3320-3324.
Wang, G. L., Jiang, B. H., Rue, E. A. and Semenza, G. L. (1995a). Hypoxiainducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by
cellular O2 tension. Proc. Natl. Acad. Sci. USA 92, 5510-5514.
Wang, G. L., Jiang, B. H. and Semenza, G. L. (1995b). Effect of altered
redox states on expression and DNA-binding activity of hypoxia-inducible
factor 1. Biochem. Biophys. Res. Commun. 212, 550-556.
Wang, G. L., Jiang, B. H. and Semenza, G. L. (1995c). Effect of protein
kinase and phosphatase inhibitors on expression of hypoxia-inducible factor
1. Biochem. Biophys. Res. Commun. 216, 669-675.
Wenger, R. H., Camenisch, G., Desbaillets, I., Chilov, D. and Gassmann,
M. (1998). Up-regulation of hypoxia-inducible factor-1α is not sufficient
for hypoxic/anoxic p53 induction. Cancer Res. 58, 5678-5680.
Wenger, R. H. and Gassmann, M. (1997). Oxygen(es) and the hypoxiainducible factor-1. Biol. Chem. 378, 609-616.
Wenger, R. H., Kvietikova, I., Rolfs, A., Gassmann, M. and Marti, H. H.
(1997). Hypoxia-inducible factor-1α is regulated at the post-mRNA level.
Kidney Int. 51, 560-563.
Wood, S. M., Gleadle, J. M., Pugh, C. W., Hankinson, O. and Ratcliffe, P.
J. (1996). The role of the aryl hydrocarbon receptor nuclear translocator
(ARNT) in hypoxic induction of gene expression. Studies in ARNTdeficient cells. J. Biol. Chem. 271, 15117-15123.