university of zurich - uzh...jens ko¨ditz,1 jutta nesper,2 marieke wottawa,1 daniel p. stiehl,3...

University of ZurichZurich Open Repository and Archive

Winterthurerstr. 190

CH-8057 Zurich

http://www.zora.uzh.ch

Year: 2007

Oxygen-dependent ATF-4 stability is mediated by the PHD3oxygen sensor

Köditz, J; Nesper, J; Wottawa, M; Stiehl, D P; Camenisch, G; Franke, C; Myllyharju,J; Wenger, R H; Katschinski, D M

Köditz, J; Nesper, J; Wottawa, M; Stiehl, D P; Camenisch, G; Franke, C; Myllyharju, J; Wenger, R H; Katschinski,D M (2007). Oxygen-dependent ATF-4 stability is mediated by the PHD3 oxygen sensor. Blood,110(10):3610-3617.Postprint available at:http://www.zora.uzh.ch

Posted at the Zurich Open Repository and Archive, University of Zurich.http://www.zora.uzh.ch

Originally published at:Blood 2007, 110(10):3610-3617.

Köditz, J; Nesper, J; Wottawa, M; Stiehl, D P; Camenisch, G; Franke, C; Myllyharju, J; Wenger, R H; Katschinski,D M (2007). Oxygen-dependent ATF-4 stability is mediated by the PHD3 oxygen sensor. Blood,110(10):3610-3617.Postprint available at:http://www.zora.uzh.ch

Posted at the Zurich Open Repository and Archive, University of Zurich.http://www.zora.uzh.ch

Originally published at:Blood 2007, 110(10):3610-3617.

Oxygen-dependent ATF-4 stability is mediated by the PHD3oxygen sensor

Abstract

The activating transcription factor-4 (ATF-4) is translationally induced under anoxic conditions,mediates part of the unfolded protein response following endoplasmic reticulum (ER) stress, and is acritical regulator of cell fate. Here, we identified the zipper II domain of ATF-4 to interact with theoxygen sensor prolyl-4-hydroxylase domain 3 (PHD3). The PHD inhibitors dimethyloxalylglycine(DMOG) and hypoxia, or proteasomal inhibition, all induced ATF-4 protein levels. Hypoxic inductionof ATF-4 was due to increased protein stability, but was independent of the ubiquitin ligase vonHippel-Lindau protein (pVHL). A novel oxygen-dependent degradation (ODD) domain was identifiedadjacent to the zipper II domain. Mutations of 5 prolyl residues within this ODD domain orsiRNA-mediated down-regulation of PHD3, but not of PHD2, was sufficient to stabilize ATF-4 undernormoxic conditions. These data demonstrate that PHD-dependent oxygen-sensing recruits both thehypoxia-inducible factor (HIF) and ATF-4 systems, and hence not only confers adaptive responses butalso cell fate decisions.

HEMATOPOIESIS

Oxygen-dependent ATF-4 stability is mediated by the PHD3 oxygen sensorJens Koditz,1 Jutta Nesper,2 Marieke Wottawa,1 Daniel P. Stiehl,3 Gieri Camenisch,3 Corinna Franke,2 Johanna Myllyharju,4

Roland H. Wenger,3 and Dorthe M. Katschinski1

1Department of Heart and Circulatory Physiology, Center of Physiology and Pathophysiology, Georg-August University Gottingen, Gottingen, Germany; 2CellPhysiology Group, Medical Faculty, Martin-Luther University Halle, Halle, Germany; 3Institute of Physiology and Zurich Center for Integrative Human Physiology,University of Zurich, Zurich, Switzerland; and 4Collagen Research Unit, Biocenter Oulu and Department of Medical Biochemistry and Molecular Biology,University of Oulu, Oulu, Finland

The activating transcription factor-4(ATF-4) is translationally induced underanoxic conditions, mediates part of theunfolded protein response following en-doplasmic reticulum (ER) stress, and isa critical regulator of cell fate. Here, weidentified the zipper II domain of ATF-4to interact with the oxygen sensor prolyl-4-hydroxylase domain 3 (PHD3). ThePHD inhibitors dimethyloxalylglycine(DMOG) and hypoxia, or proteasomal

inhibition, all induced ATF-4 protein lev-els. Hypoxic induction of ATF-4 was dueto increased protein stability, but wasindependent of the ubiquitin ligase vonHippel–Lindau protein (pVHL). A noveloxygen-dependent degradation (ODD)domain was identified adjacent to thezipper II domain. Mutations of 5 prolylresidues within this ODD domain orsiRNA-mediated down-regulation ofPHD3, but not of PHD2, was sufficient to

stabilize ATF-4 under normoxic condi-tions. These data demonstrate that PHD-dependent oxygen-sensing recruits boththe hypoxia-inducible factor (HIF) andATF-4 systems, and hence not only con-fers adaptive responses but also cellfate decisions. (Blood. 2007;110:3610-3617)

© 2007 by The American Society of Hematology

Introduction

When O2 delivery is impaired, the resulting hypoxia activateshomeostatic mechanisms at the systemic and cellular level.1 Thisresponse involves changes in gene expression mediated by hypoxia-inducible factor-1 (HIF-1), the master transcription factor ofoxygen-regulated genes. HIF-1 is a heterodimeric protein compris-ing the oxygen-sensitive �-subunit (HIF-1�, or the more cell-type–specifically expressed HIF-2�) and the oxygen-insensitive �-sub-unit.2 Oxygen-regulated gene expression involves binding of HIFto cis-regulatory hypoxia response elements (HREs) of HIF targetgenes such as erythropoietin or vascular endothelial growth factor.3

The molecular basis for the hypoxia-induced stability and activityof HIF-1� and HIF-2� is the O2-dependent hydroxylation ofdistinct prolyl residues.4-6 A family of oxygen-, iron- and 2-oxoglu-tarate-dependent prolyl-4-hydroxylases has been described re-cently to hydroxylate the oxygen-labile � subunits of HIF-1 andHIF-2.5,7,8 This family consists of 3 members called prolyl-4-hydroxylase domain (PHD) 1, PHD2, PHD3, or HIF prolylhydroxylase (HPH) 3, HPH2, and HPH1, respectively.4,5 Followingprolyl-4-hydroxylation of the critical prolyl residues under nor-moxic conditions, the ubiquitin ligase von Hippel–Lindau tumorsuppressor protein (pVHL) recognizes HIF� subunits and targetsthem for rapid ubiquitination and proteasomal degradation.9 Bind-ing of pVHL strictly requires prior modification of human HIF-1�and HIF-2� by prolyl-4-hydroxylation at prolines 402 and 564 orprolines 405 or 531, respectively.10,11 Limited oxygen supplyprevents HIF� hydroxylation and degradation.12 In addition toprotein stability, oxygen-dependent C-terminal asparagine hydroxy-lation of HIF� by factor-inhibiting HIF (FIH) prevents transcrip-

tional cofactor recruitment, thereby fine-tuning HIF-1 activity aftera further decrease in oxygen availability.13 Most interestingly, inaddition to HIF�, ankyrin repeats present in I�B and NF-�B familymembers have recently been described to be hydroxylated by FIH,demonstrating that hydroxylation is not restricted to the HIFsignaling pathway.14

Besides similarities in the hydroxylation reaction in vitro, the3 PHDs differ in their ability to hydroxylate HIF-1� in vivo and intheir organ-specific expression pattern.15-17 Moreover, the pheno-types of PHD knock-out mice demonstrate divergent roles of the3 PHDs during embryonic development.18 These data indicate thatunder physiologic conditions, PHD1, PHD2, and PHD3 mediatedifferent, probably even HIF-independent, oxygen-regulated signaltransduction pathways.

By searching for novel targets of PHD3 using yeast 2-hybridtechnology, we identified activating transcription factor-4 (ATF-4)as a novel interaction partner, and we found that PHD3 confersoxygen-dependent ATF-4 protein stability in a pVHL-independentmanner. ATF-4–deficient mice are severely anemic during fetaldevelopment, apparently because of an impairment in definitivehematopoiesis.19 In addition, overexpression of ATF-4 near ne-crotic areas in tumor tissues has been described, suggesting a roleof ATF-4 in cancer.20 Since ATF-4 influences DNA repair as well ascell fate decisions via its target genes like GADD153, the presenteddata may help to understand the physiologic and pathophysiologicrole of ATF-4 for hematopoiesis, cancer physiology, and tumortherapy resistance.20

Submitted June 7, 2007; accepted August 6, 2007. Prepublished onlineas Blood First Edition paper, August 7, 2007; DOI 10.1182/blood-2007-06-094441.

An Inside Blood analysis of this article appears at the front of this issue.

The publication costs of this article were defrayed in part by page chargepayment. Therefore, and solely to indicate this fact, this article is herebymarked ‘‘advertisement’’ in accordance with 18 USC section 1734.

© 2007 by The American Society of Hematology

3610 BLOOD, 15 NOVEMBER 2007 � VOLUME 110, NUMBER 10

For personal use only. at UNIVERSITAETSSPITAL on July 15, 2008. www.bloodjournal.orgFrom

http://bloodjournal.hematologylibrary.org

http://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtl

Materials and methods

Plasmids

All cloning work was carried out using Gateway technology (Invitrogen,Carlsbad, CA). Full-length ATF-4 was amplified by polymerase chainreaction (PCR) from a HeLa cDNA pool. Full-length PHD1–3 wasamplified by PCR from plasmids pcDNA3.1PHD1 and pcDNA3.1PHD3(kindly provided by F. Oehme and I. Flamme, Wuppertal, Germany), orpcDNA3.1HA-PHD2 (kind gift from W. G. Kaelin Jr, Boston, MA).Site-directed mutagenesis or deletion mutants were performed with theATF-4 entry vector using the QuickChange kit (Stratagene, La Jolla, CA).To generate plasmids expressing fusion proteins, DNA inserts weretransferred from entry clones to destination vectors using LR Clonaserecombination enzyme mix (Invitrogen). To generate plasmids expressingthe Gal4-BD and the Gal4-AD fused to an oxygen-dependent degradation(ODD) domain of interest, PCR-amplified DNA fragments were cloned intothe EcoRI-digested pM3VP16 (BD Clontech, Palo Alto, CA). The Gal4containing luciferase reporter gene vector pGREx5E1bLuc was kindlyprovided by D. Peet (University of Adelaide, Australia).

Yeast 2-hybrid analyis

Yeast 2-hybrid analyses were performed using the ProQuest 2-hybridsystem by following the manufacturer’s recommendations (Invitrogen).Full-length PHD3 fused to the Gal4-BD was expressed from pDEST32-PHD3 in Saccharomyces cerevisiae strain MaV203 (Invitrogen). To test forself-activity, cotransformants of pDEST32-PHD3 and pExpAD502, encod-ing the Gal4-AD, were tested on selection plates. A ProQuest human braincDNA library was transformed into MaV203 pDEST32-PHD3, and transfor-mants were screened on synthetic drop-out medium lacking tryptophan,leucine, and histidine, as well as 10 mM 3-amino-1,2,4-triazole (3-AT;Sigma, St Louis, MO). Positive clones were further assayed for growth onsynthetic dropout medium lacking tryptophan, leucine, and uracil, and for�-galactosidase activity, to characterize the strength of the interaction. ThepExpAD502 plasmids encoding putative PHD3-interacting proteins wereisolated, retested for interaction, and checked for self-activity. Inserts ofconfirmed interactors were sequenced to ensure in-frame coding sequencewith the activation domain of Gal4.

Protein analysis

Proteins were separated by SDS-PAGE and detected by immunoblotting.Antibodies used were mouse anti–HIF-1� (Transduction Laboratories,Franklin Lakes, NJ), rabbit anti–maltose-binding protein (MBP) (NewEngland Biolabs, Ipswich, MA), mouse anti–�-actin (Sigma), mouseanti-V5 (Invitrogen), rabbit anti–ATF-4 (Santa Cruz Biotechnology, SantaCruz, CA), rabbit anti-PHD2 (Abcam, Cambridge, United Kingdom), andrabbit anti-PHD3 (Abcam). Immunocomplexes were detected by usinghorseradish-conjugated goat anti-rabbit or goat anti-mouse antibodies(Santa Cruz Biotechnology) and incubation of the membranes with 100 mMTris-HCl (pH 8.5), 2.65 mM H2O2, 0.45 mM lumin, and 0.625 mMcoumaric acid for 1 minute, followed by exposure to X-ray films (Amer-sham, Arlington Heights, IL) or detection of the signal with the LAS3000camera system (Fiji Film Europe, Dusseldorf, Germany).

Protein expression and purification

MBP fusion proteins were expressed in the Escherichia coli strain TB1(New England Biolabs) transformed with pMal-c2x or derivatives andpurified using the MBP purification system according to the manufacturer’sinstructions (New England Biolabs). His-PHD3 was purified from insectcells as described previously.21

In vitro pull-downs

Purified MBP–ATF-4 fusion protein (10 �g) was bound in buffer 1 (20 mMTris-HCl [pH 7.5], and 200 mM NaCl) to bovine serum albumin (BSA)–blocked amylose-resin (New England Biolabs) for 1 hour at 4°C. Bound

proteins were incubated with His-PHD3 in 500 �L 10 mM Na-phosphate,(pH 7.0), 5 mM MgCl2, 100 mM NaCl, 100 �M FeSO4, and 0.1 mg/mLBSA for 1 hour at 30°C. The resins were washed 5 times, and boundproteins were eluted with SDS-PAGE loading buffer and subjected toimmunoblot analysis.

Cell culture

HeLa cells were cultured in high-glucose Dulbecco modified Eagle mediumcontaining 10% fetal calf serum, 50 U/mL penicillin G, and 50 �g/mLstreptomycin (Invitrogen) in a humidified 5% CO2, 95% air atmosphere at37°C. For hypoxic conditions, O2 levels were decreased to 0.2% or 1% O2

with N2 in an oxygen-controlled incubator (Binder, Tuttlingen, Germany).TS-20 cells are derivatives of 3T3 fibroblasts, which harbor a temperature-sensitive defect in the E1 ubiquitin–activating enzyme.22,23 H38–5 cells arereconstituted with a wild-type allele of the E1 enzyme. Both cell lines werekindly provided by C. Borner (Center for Biochemistry and Molecular CellResearch, Freiburg, Germany). In some experiments, cells were treatedwith 10 �M MG132 (Alexis, Grunberg, Germany), 1 mM DMOG (FrontierScientific, Carnforth, United Kingdom), 20 �g/mL cycloheximide (Sigma),or 300 nM thapsigargin (AppliChem, Darmstadt, Germany) for theindicated time periods.

Transient transfection and luciferase assay

A total of 1 �g pcDNA3.1Dest-V5-ATF-4 or various ATF-4 mutants weretransfected together with 0.2 �g pEGFPC1 by calcium phosphate coprecipi-tation. For luciferase assays, HeLa cells were transfected with 250 ng fireflyluciferase reporter vector pATFx2-Luc, containing 2 cAMP responseelement (CRE) sites (kind gift of T. Hai, Ohio State University, Columbus,OH) or pH3SVL containing 3 HRE sites and 2.5 ng of pRLSV40, encodingrenilla luciferase, for internal standardization (Promega, Madison, WI). Fortransient transfection of the Gal-4BD/Gal-4AD fusion proteins, 62 ngpM3XVP16 was transfected along with 250 ng luciferase reporter vectorpGREx5E1bLuc. The total amount of plasmid DNA was adjusted to 1 �gby addition of empty pcDNA3.1 vector (Invitrogen).

Gene silencing by RNAi

The following stealth RNAi (Invitrogen) sequences were used: control siRNA,forward 5�-gcuccggagaacuaccagaguauua-3�; control siRNA, reverse 5�-uaauacu-cugguaguucuccggagc-3�; PHD2 siRNA, forward 5�-ggacgaaagccaugguugcu-uguua-3�; PHD2 siRNA, reverse 5�-uaacaagcaaccauggcuuucgucc-3�; PHD3siRNA, forward 5�-gcuauccgggaaaugga-acagguua-3�; and PHD3 siRNA, reverse5�-uaaccuguuccauuucccggauagc-3�. HeLa cells were transfected with 80 nMsiRNAs using Lipofectamin 2000 (Invitrogen).

RNA extraction and real-time RT-PCR

After total RNA extraction, reverse transcription was performed with 1 �g ofRNA (Fermentas) in a total volume of 20 �L. Subsequently, mRNA expressionlevels were quantified with 1 �L of cDNA reaction by real-time PCR using aSybrGreen Q-PCR reagent kit (Stratagene) in combination with the MX3000Plight cycler (Stratagene). Initial template concentrations of each sample werecalculated by comparison with serial dilutions of a calibrated standard. Primerswere as follows: PHD2 forward, 5�-gaaagccatggttgcttgtt-3�; PHD2 reverse,5�-ttgccttctggaaaaattcg-3�; PHD3 forward, 5�-atcgacaggctggtcctcta-3�; PHD3reverse, 5�-cttggcatcccaattcttgt-3�; ATF-4 forward, 5�-tcaaacctcatgggttctcc-3�;ATF-4 reverse, 5�-gtgtcatccaacgtggtcag-3�; hL28 forward, 5�-gcatctgcaatggatggt-3�; and hL28 reverse, 5�-tgttcttgcggatcatgtgt-3�.

Results

Identification of ATF-4 as a novel PHD-binding protein

To identify novel interaction partners of PHD3 by yeast 2-hybridanalysis, we screened a human brain cDNA library fused to Gal4activation domain (AD) with full-length PHD3 fused to Gal4

ATF-4 STABILITY IS REGULATED BY PHD3 3611BLOOD, 15 NOVEMBER 2007 � VOLUME 110, NUMBER 10




binding domain (BD) as bait. As shown in Figure 1A, the PHD3bait alone was unable to confer yeast growth. A total of 4 out of69 prey clones were found to be independent AD fusions to thecDNA of HIF-2�, corresponding to amino acid (aa) 73-870 (Figure1A), aa 379-870, aa 398-870, and aa 451-870 (data not shown). All4 HIF-2� clones contain at least the C-terminal part of the ODDdomain (aa 404-569) which includes proline 531. This site isknown to be modified by PHD3, hence confirming the reliability ofour screen.

An additional 4 prey clones derived from this screen corre-sponded to ATF-4 fusion proteins encompassing ATF-4 (NationalCenter for Biotechnology information data base24 accession num-ber NM 001675) aa 27-351 (Figure 1A), aa 42-351, aa 51-351,and aa 83-351 (data not shown). No other proteins of theATF/CREB family were identified in this screen. No protein-protein interaction was found in further yeast 2-hybrid analysesusing PHD1, PHD2, or FIH fused to Gal4-BD together withfull-length ATF-4 fused to Gal4-AD (data not shown). ATF-4comprises 2 zipper domains: a C-terminal basic leucine zipper aswell as a zipper II. Through its C-terminal basic leucine zipperdomain, it dimerizes with members of either the ATF/CREB familyor other basic leucine zipper transcription factor families like c-Fosor c-Jun.25 The zipper II as the main PHD3-interaction domain wasverified by lack of interaction of the ATF-4 zipper II deletionmutants (ATF-4� aa 89-125 and ATF-4� aa 89-181) with PHD3(Figure 2).

The binding of full-length ATF-4 to PHD3 was shown in vitroby MBP pull-down assays (Figure 1B). Likewise, ATF-4 taggedwith the V5-epitope synthesized in rabbit reticulocyte lysatesbound to a bacterially expressed MBP-PHD3 fusion protein (datanot shown). In contrast to the recently described WD-repeat proteinmitogen-activated protein kinase (MAPK) organizer (Morg)–1,which affects HIF-1 most likely via modulation of PHD3 activity,26

ATF-4 did not influence PHD activity and thereby HIF stability/activity, as determined by HIF-dependent reporter gene assays aftercoexpression of HIF-2� and PHD3 (Figure 1C). We additionallyexcluded an effect of ATF-4 on the HRE-driven luciferase activity

Figure 1. Interaction of ATF-4 with PHD3. (A) Theyeast reporter strain MaV203 expressing Gal4-BD-PHD3 and Gal4-AD-ATF-4 aa 27–351 or Gal4-AD-HIF2� aa 73–870 was assayed for histidine auxotrophy.(B) His-PHD3 expressed in insect cells is captured byimmobilized MBP–ATF-4. Antibodies against MBP orPHD3 were used for immunoblot detection as indicated.(C) HeLa cells were cotransfected with a firefly lucif-erase reporter gene plasmid driven by HREs togetherwith a constitutively expressed renilla luciferase controlvector as depicted and the expression plasmidspcDNA3.1HIF-2�, pcDNA3.1PHD3, or pcDNA3.1V5-ATF-4. After 24 hours, HeLa cells were lysed andanalyzed for firefly (FL) and renilla luciferase (RL)activities. (D) HeLa cells were cotransfected with anHRE-driven firefly luciferase reporter gene plasmid anda constitutively expressed renilla luciferase control vec-tor as in panel C and the empty control vector pcDNA3.1or the expression plasmid pcDNA3.1V5ATF-4. Subse-quently, cells were incubated at 20% O2 or 1% O2 for24 hours, lysed, and analyzed for firefly and renillaluciferase activities (C,D). Shown are mean values ofthe FL/RL ratios (� SD) of 3 independent experiments.

Figure 2. Mapping the PHD3 binding domain of ATF-4. (A) Schematic drawing ofthe protein domains of ATF-4. Arrows (27, 42, 51, and 83) indicate the insertion of the4 independent ATF-4 prey-clones identified by yeast 2-hybrid screening of the humanbrain cDNA library with PHD3 as bait. (B) Yeast reporter strain MaV203 expressingGal4-BD-PHD3 and various domains of ATF-4 as Gal4-AD-fusions was assayed forhistidine and uracil auxotrophy. (C) Yeast reporter strain MaV203 expressingGal4-BD-PHD3 and various deletion variants of ATF-4 as Gal4-AD fusions wasassayed for histidine and uracil auxotrophy.

3612 KODITZ et al BLOOD, 15 NOVEMBER 2007 � VOLUME 110, NUMBER 10




by coexpressing ATF-4 together with the reporter gene plasmid at20% or 1% oxygen. (Figure 1D).

Hypoxia and PHD inhibitors stabilize ATF-4

Previous reports have established a regulatory mechanism fortranslational control of ATF-4 in response to endoplasmic reticu-lum (ER) stress, amino acid starvation, and anoxia.27 ExposingHeLa cells to the ER stress–inducing agent thapsigargin as well asto the proteasome inhibitor MG132 both resulted in enhancedATF-4 levels (Figure 3A). To further confirm the involvement ofthe ubiquitin-proteasome system in the degradation of ATF-4, thetemperature-sensitive E1 mutant cell line TS20 was cultured undernormoxic conditions at 34°C and 39°C. As demonstrated in Figure3C, TS20 cells do not express ATF-4 at 34°C, whereas growing thecells at 39°C resulted in accumulation of ATF-4. Simultaneousexposure to 39°C of the E1-corrected cells H38–5 did not result inincreased ATF-4, suggesting that in addition to translation, ATF-4might also be regulated on the level of protein degradation.

Interestingly, DMOG, which is able to induce HIF-1� viainhibition of PHDs, also induced ATF-4 reporter gene activity,further supporting the existence of an alternative ATF-4 inductionpathway (Figure 3B). If ATF-4 was regulated by PHD3, it shouldbe stabilized not only by DMOG but also by hypoxia. Thus, wenext determined ATF-dependent reporter gene activity using aluciferase expression construct under the control of 2 CRE sites.A strong increase in reporter gene activity was observed afterexposing transfected HeLa cells to 1% O2 (Figure 3B). ExposingHeLa cells for 4 hours to 1% or 0.2% O2 enhanced the levels ofATF-4 similar to HIF-1� (Figure 4A; compare lanes 3 and 5 withlane 1). Inhibition of endogenous PHDs with DMOG was followedby increased ATF-4 protein in normoxia (Figure 4A; compare lanes1 and 2) and enhanced the hypoxic response (Figure 4A; lanes 4and 6), indicating that hypoxia-induced ATF-4 was mediated byPHDs. After 24 hours, however, when the hypoxic stabilization ofHIF-1� was still clearly visible, hypoxic ATF-4 induction was no

longer detectable. This transient hypoxia-induced expression ofATF-4 inversely correlated with the strong hypoxic induction ofPHD3 after 24 hours. In this regard it is interesting to note that werecently demonstrated that the hypoxic induction of PHD3 corre-lates with increased hydroxylation activity, which compensates fordecreased oxygen levels.28,29

If PHDs mediated the degradation of ATF-4 in normoxia, thehypoxia-induced ATF-4 protein should become unstable followingreoxygenation. Therefore, HeLa cells were exposed for 4 hours to0.2% O2, followed by reoxygenation in the presence of cyclohexi-mide to inhibit de novo translation. ATF-4 was rapidly degradedwith a half-life (t1/2) of 9 minutes (Figure 4B). However, addingDMOG at the time of hypoxia resulted in a prolonged ATF-4protein half-life of 21 minutes.

Because the commercially available anti–ATF-4 antibodies areof minor quality, the treatment by hypoxia and DMOG wasrepeated using HeLa cells transiently transfected with V5-taggedATF-4. As shown in Figure 5A, V5–ATF-4 but not V5–ATF-3 orV5-ARNT could only be detected under hypoxic conditions or aftertreatment with DMOG. A cotransfected green fluorescent protein(GFP) expression vector ensured equal transfection and immuno-blotting efficiencies.

Identification of an ODD domain in ATF-4

To explore whether the interaction of ATF-4 with PHD3 mediateshypoxia-inducible stability of ATF-4, we transiently overexpressedV5-tagged wild-type ATF-4 and various ATF-4 deletion mutantsunder normoxic or hypoxic conditions. In contrast to the wild-typeATF-4, ATF-4� aa 89-181, an ATF-4 variant missing the zipper IIwhich is responsible for the interaction with PHD3 as described,plus an adjacent 48-aa domain, was stable in normoxia. Interest-ingly, deletion of just the 48-aa stretch (ATF-4� aa 133-181)

Figure 3. Hypoxia as well as proteasome and PHD inhibition stabilize ATF-4.(A) HeLa cells were treated for 8 hours with 300 nM thapsigargin or 10 �M MG132.Subsequently, cells were lysed and analyzed by immunoblotting. (B) HeLa cells weretransiently transfected with an ATF-dependent firefly (FL) luciferase reporter geneplasmid (pATFx2-Luc) together with the renilla luciferase (RL) control plasmidpRLSV40. Following exposure to 20% or 1% O2, with or without the addition of 1 mMDMOG, cells were lysed and luciferase activities were determined. (C) TS-20 cells(with a temperature-sensitive E1 ubiquitin–activating enzyme defect) and E1-reconstituted H38–5 cells were exposed to 34°C or 39°C. Subsequently, cells werelysed and analyzed by immunoblotting. Shown are mean values of the FL/RL ratios(� SD) of 3 independent experiments.

Figure 4. Kinetics of hypoxic stabilization and reoxygenation-induced destabi-lization of ATF-4. (A) HeLa cells were incubated at 20%, 1%, or 0.2% O2, with orwithout the addition of 1 mM DMOG. After 4 or 24 hours, the cells were lysed andanalyzed by immunoblotting. As a control for the induction of ATF-4 by ER stress, onedish of HeLa cells was treated with 300 nM thapsigargin for 4 hours. (B) HeLa cellswere incubated at 0.2% O2 with or without the addition of 1 mM DMOG. At 4 hourslater, 20 �g/mL cycloheximide was added, and cells were reoxygenated at 20% O2.After the indicated time points, cells were lysed and analyzed by immunoblotting.





adjacent to the zipper II or an even smaller deletion (ATF-4� aa154-181) was sufficient to generate increased ATF-4 levels innormoxia (Figure 5B). Determining the half-life of ATF-4 andATF-4� aa 154-181 in reoxygenation, a higher stability of theATF-4� aa 154-181 mutant (t1/2, 24 minutes) compared with thewild-type ATF-4 (t1/2, 13 minutes) was detected (Figure 5C,D).

Since the ATF-4 wt protein seems to be highly instable undernormoxic conditions, we further characterized the identified ODDin a more sensitive reporter gene assay. Therefore, we determined ifthe identified ODD domain of ATF-4 was also sufficient to promotethe degradation of a heterologous protein under normoxia. To thisend, HeLa cells were transiently cotransfected with a Gal-4 drivenfirefly luciferase reporter gene plasmid together with expressionvectors containing either Gal-4AD/BD in frame with the HIF-2�ODD aa 404-569 and HIF-2� ODD aa 404-569 with pointmutations of the 2 critical prolines (P405A/P531A), or ATF-4aa 133-183, and exposed for 24 hours at 20% or 1% O2.

A constitutively active renilla luciferase reporter gene plasmid wascotransfected for standardization purposes. HeLa cells expressingthe HIF-2� ODD aa 404-569 showed a significant increase inluciferase activity under hypoxic conditions compared with nor-moxic conditions, and this effect was abolished by mutation of theprolines 405 and 531 to alanine (Figure 6A). Similar to the HIF-2�ODD, ATF-4aa 133-183 conferred oxygen-dependent stability tothe heterologous Gal4DB/Gal4AD fusion protein. The criticalATF-4 domain contains 5 prolyl residues (P156, P162, P164, P167,and P174). Mutation of all 5 prolines to alanine (V5–ATF-45 � P A) led to increased ATF-4 levels in normoxia (Figure 6B)and abolished the destabilizing effect in the Gal4AD/BD ATF-4 aa133-183 fusion protein (Figure 6A), suggesting that these prolylresidues define a novel ODD domain within ATF-4. Point mutationof each proline alone was not sufficient to generate a similar effectlike V5–ATF-4 5 � P A (data not shown), indicating a redun-dant role of each prolyl residue.

Figure 5. ATF-4 aa 154–181 harbors a novel ODDdomain. (A) HeLa cells were transiently transfected withV5–ATF-4, V5–ATF-3, or V5–ARNT expression vectors.To control for equal transfection efficiencies, cells werecotransfected with pEGFPC1. At 6 hours after transfec-tion, cells were exposed to 1% O2 or treated with 1 mMDMOG for 24 hours. Subsequently, cells were lysed andanalyzed by immunoblotting. (B) HeLa cells were tran-siently transfected with different ATF-4 variants andtreated as in panel A. (C) HeLa cells were transientlytransfected with V5–ATF-4 or V5–ATF-4� aa154-181.To control for equal transfection efficiencies, cells werecotransfected with pEGFPC1. At 6 hours after transfec-tion, cells were exposed to 1% O2. After exposure to 1%O2 for 24 hours, the cells were treated with 20 �g/mLcycloheximide and were reoxygenated with 20% O2.Cells were lysed at different reoxygenation time points,and total protein was analyzed by immunoblotting. Bandintensities of 3 independent experiments were deter-mined and half-life of the V5–ATF-4 and V5–ATF-4�aa154-181 proteins were calculated. (D) Decline ofV5–ATF-4 and V5–ATF-4� aa154-181 after reoxygen-ation calculated from 3 independently performed experi-ments as described in panel C. Shown are mean values(� SD) of 3 independent experiments.

Figure 6. The oxygen-dependent stability of ATF-4 isproline dependent. (A) HeLa cells were transiently trans-fected with a constitutively expressed renilla luciferase (RL)control vector, an expression plasmid containing the Gal4AD/BD (pM3VP16) fused to ATF-4 aa 133–183; ATF-4 aa133-183, in which prolines P156, P162, P164, P167, andP174 have been mutated to alanine; HIF-2� aa 404-568; orHIF-2� aa 404–568 P405/531A together with the Gal4-driven firefly luciferase (FL) plasmid pGRExE1bLuc. Aftertransfection cells, were exposed to 20% or 1% O2 for24 hours, lysed, and analyzed for FL and RL activities.Shown are mean values plus or minus SD of the FL/RLratiosof 3 independent experiments. (B) HeLa cells were tran-siently transfected with V5–ATF-4, V5–ATF-4� aa 154-81, orV5–ATF-4 5 � P A, in which prolines P156, P162, P164,P167, and P174 have been mutated to alanine. To control forequal transfection efficiencies cells were cotransfected withpEGFPC1. At 6 hours after transfection, cells were exposedto 20% or 1% O2 for 24 hours. Subsequently, the cells werelysed and analyzed by immunoblotting. (C) Multiple se-quence alignment of the proposedATF-4 ODD domain.





Notably, the normoxic induction of V5–ATF-4 5 � P Awas less pronounced than V5–ATF-4� aa 154-181 (Figure 6B).Despite multiple proline mutations, binding to PHD3 as well asactivation of an ATF-4–dependent reporter gene (pCHOP-AARE-TK-Luc) was not affected as tested by yeast 2-hybrid analysisand reporter gene assay (data not shown). Aligning the ATF-4sequences of mammalian and nonmammalian vertebrate speciesindicates that P164, P167, and P174 are highly conserved,possibly indicating their specific role in regulation of ATF-4stability (Figure 6C).

PHD3 regulates ATF-4 protein stability

To determine the role of the PHDs in the regulation of ATF-4protein stability, HeLa cells were transfected with siRNA targetingPHD2 or PHD3 and exposed to either 20% O2 or 1% O2 for24 hours. PHD1 was not further investigated, since it was notdetectable at the protein level in HeLa cells (data not shown).Moreover, in contrast to PHD2 and PHD3, PHD1 expression is notinduced by hypoxia.8 In cells transfected with PHD2 and PHD3siRNA, an additive up-regulation of HIF-1� in normoxia wasfound (Figure 7A). In contrast to HIF-1�, ATF-4 protein levelswere up-regulated exclusively by PHD3 silencing but not by PHD2silencing. As demonstrated here, the hypoxic induction of ATF-4was transient and not detectable after incubation of HeLa cells for24 hours at 1% O2. However, after treatment of HeLa cells withPHD3 siRNA, ATF-4 was up-regulated even after exposing thecells for 24 hours to hypoxia. PHD3 down-regulation exerted itseffect via increased ATF-4 protein stability as demonstrated inreoxygenation experiments (Figure 7B). No significant differencesin ATF-4 mRNA levels were found after treatment of HeLa cellswith PHD3 siRNA compared with mock-treated control cells orafter treatment with PHD2 siRNA (Figure 7C). The siRNAexperiments were repeated with independent siRNAs as well asshort hairpin constructs, confirming the previous results (data notshown). In addition, FIH neither interacted with ATF-4 nor had FIHsiRNA any effect on ATF-4 protein levels (data not shown).

Hydroxylated HIF-1� is strongly bound by pVHL. Thus, wenext determined whether PHD-dependent ATF-4 hydroxylationleads to subsequent pVHL interaction. Therefore, MBP–ATF-4or MBP–HIF-2� ODD were treated with HA-PHD2 or His-PHD3 in the presence of oxygen, 2-oxoglutarate, ferrous iron,

and ascorbic acid. The interaction with His6-Trx–tagged pVHL–elongin B–elongin C (VBC) complex was examined in a MBPpull-down assay. However, MBP-HIF-2� ODD but not MBP–ATF-4 interacted with pVHL following incubation with PHD2or PHD3 (data not shown).

Discussion

ATF-4, also known as CREB2, TAXREB67, or C/ATF, is a memberof the ATF/CREB family of basic-leucine zipper transcriptionfactors.30 ATF-4–dependent gene expression has been described forseveral genes, including CHOP/GADD153, which is involved instress regulatory pathways.31-33 ATF-4 mRNA is present in alltissues examined thus far, whereas ATF-4 protein levels varydependent on a variety of extracellular signals and stress condi-tions.25 This regulation has been widely attributed to preferentialtranslation of the ATF-4 mRNA.27 In this work, up-regulation ofATF-4 protein was found after treatment with hypoxia or the PHDinhibitor DMOG, demonstrating that mechanisms other thantranslational control might be additionally involved in the regula-tion of ATF-4 protein stability.

Using PHD3 as bait, we identified ATF-4 as a new PHD-interacting protein. The interaction site was mapped to a leucinezipper motif of ATF-4, which is also called zipper II. In thepresent work, we demonstrate that deletion of aa 154-181, aregion close to the zipper II, results in oxygen-independentexpression of ATF-4. Oxygen-regulated ATF-4 expression wasat least in part dependent on 5 conserved prolyl residues withinthis region. Therefore, aa 154-181 of ATF-4 can be consideredas a novel ODD domain.

siRNA experiments strongly suggested a specific role for PHD3in ATF-4 protein stability. Despite several attempts, we did notsucceed to directly demonstrate ATF-4 hydroxylation by massspectrometry or 2-oxoglutarate turnover. Apart from technicaldifficulties, the conditions for efficient ATF-4 hydroxylation mightbe different from those for HIF-1� hydroxylation. Indeed, majordifferences become obvious when the 2 proteins are compared:first, whereas HIF-1� levels can be regulated by all 3 PHDs, ATF-4levels are modulated specifically by PHD3; second, HIF-1� is

Figure 7. ATF-4 expression is controlled by PHD3.(A) HeLa cells were transiently transfected with siRNAtargeting PHD2 or PHD3 and exposed to 20% or 1% O2.After 24 hours, the cells were lysed and analyzed byimmunoblotting. One dish of HeLa cells was treated for8 hours with 300 nM thapsigargin (t). Subsequently, cellswere lysed and analyzed together with the protein ex-tracts of the siRNA-transfected cells. (B) Hela cells weretransiently transfected with siRNA targeting PHD3 or acontrol siRNA and exposed to 1% O2 for 4 hours withsubsequent reoxygenation.After the indicated time points,cells were lysed and analyzed by immunoblotting.(C) HeLa cells were transiently transfected with siRNAtargeting PHD2 or PHD3 and exposed to 20% or 1% O2.After 24 hours, the cells were lysed and total RNA wasextracted and analyzed for ATF-4, PHD2, and PHD3mRNA levels by real-time reverse transcriptase (RT)–PCR. Shown are mean values plus or minus SD of3 independent experiments of ratios to the ribosomalprotein L28 mRNAlevels, which were used as housekeep-ing controls.





stabilized even under prolonged hypoxia, whereas hypoxic stabili-zation of ATF-4 is transient and inversely correlates with hypoxicup-regulation of PHD3; third, as a result of PHD-dependentmodification, HIF-1� but not ATF-4 can be bound to pVHL. Thislast finding is consistent with a previous report demonstrating thatoxygen-dependent ATF-4 destruction is independent of pVHL.20

Rather than pVHL, proteasome-dependent degradation of ATF-4has been demonstrated to rely on the SCF�TRCP ubiquitin ligase.34

Further studies will be needed to investigate whether SCF func-tions as an ATF-4 ubiquitin E3 ligase in an oxygen- and PHD3-dependent manner. It is, however, still possible that despiteprotein-protein interaction, proline residue-dependency, and DMOGinhibition, PHD3 regulates ATF-4 without direct proline hydroxyla-tion. Future experiments will show whether PHD3 confers ATF-4protein stability by recruiting a yet unidentified protein in atrimolecular complex, which might then be regulated byhydroxylation.

Of note, PHD3 is the best hypoxia-inducible protein of allPHDs. We recently observed that PHD3 is still functional evenunder 0.2% O2 conditions, albeit at a reduced rate.28 As aconsequence, a protein specifically regulated by PHD3 should onlybe transiently stabilized, like we found for ATF-4. Our results thusprovide a possible explanation for the existence of more than onePHD with similar oxygen affinities: compared with HIF�, targetsof only one PHD family member will have different stabilizationkinetics, thereby orchestrating the hypoxic response.

Besides HIF-mediated signal transduction pathways, HIF-independent pathways have been described to be involved in thecellular response to low oxygen concentrations using genearrays. The molecular mechanisms leading to HIF-independenthypoxia-inducible mRNA induction remains largely unknown.Among the HIF-independent oxygen-regulated genes are ATF-4target genes like CHOP10/GADD153 and ATF-3.35-38 Hypoxia-inducible regulation of GADD153 has been implicated to play arole in cell stress response, DNA-repair, and anoxia-induced celldeath.39 Our data provide evidence that in addition to thetranslational control of ATF-4 under anoxic conditions, PHD3-dependent hydroxylation may play a major role in controlling

ATF-4 target gene expression in normoxia and hypoxia. Tightoxygen-dependent regulation of ATF-4 by PHD3 may be themolecular basis for the cell fate after exposure to severehypoxia. Our data also show that PHD-mediated regulation ofprotein levels is not unique for the HIF signal transductionpathway, suggesting that posttranslational, PHD-mediated modi-fications are used in additional pathways that need tight andrapid oxygen-dependent regulation.

Acknowledgments

This work was supported by grants from the Bundesministeriumfur Bildungund Forschung (BMBF) (NBL3 to D.M.K.), DeutscheForschungsgemeischoft (DFG) (Ka1269/8–1 and DFG Ka1269/9–1 to D.M.K.), Schweicerischer National Fund (SNF) (3100A0–116047/1 to R.H.W.), the 6th Framework Program of the EuropeanCommission/SBF (Euroxy LSHC-CT-2003–502932/SBF no.03.0647–2 to R.H.W.), the Acadamy of Finland (200471, 202469,and 44843 Center of Excellence Program 200-2005 to J.M.), andthe Sigrid Juselius Foundation (to J.M.).

Authorship

Contribution: G.C., R.H.W., and D.M.K. performed experiments,analyzed data, and wrote the paper. J.K., J.N., D.P.S., and J.M.performed experiments and analyzed data. M.W. and C.F. per-formed experiments. J.K. and J.N. contributed equally to this work.

Conflict-of-interest disclosure : The authors declare no compet-ing financial interests.

Correspondence: Dorthe M. Katschinski, Department of Heartand Circulatory Physiology, Center of Physiology and Patho-physiology, Georg-August University Gottingen, Humboldtallee23, D-37073, Gottingen, Germany; e-mail: [email protected].

References

1. Schofield CJ, Ratcliffe PJ. Signalling hypoxia byHIF hydroxylases. Biochem Biophys Res Com-mun. 2005;30:30.

2. Wang GL, Semenza GL. Purification and charac-terization of hypoxia-inducible factor 1. J BiolChem. 1995;270:1230-1237.

3. Wenger RH, Stiehl DP, Camenisch G. Integrationof oxygen signaling at the consensus HRE. SciSTKE. 2005;2005:re12.

4. Jaakkola P, Mole DR, Tian YM, et al. Targeting ofHIF-� to the von Hippel-Lindau ubiquitylationcomplex by O2-regulated prolyl hydroxylation.Science. 2001;292:468-472.

5. Ivan M, Kondo K, Yang H, et al. HIF� targeted forVHL-mediated destruction by proline hydroxyla-tion: implications for O2 sensing. Science. 2001;292:464-468.

6. Masson N, Ratcliffe PJ. HIF prolyl and asparagi-nyl hydroxylases in the biological response to in-tracellular O2 levels. J Cell Sci. 2003;116:3041-3049.

7. Bruick RK, McKnight SL. A conserved family ofprolyl-4-hydroxylases that modify HIF. Science.2001;294:1337-1340.

8. Epstein AC, Gleadle JM, McNeill LA, et al. C. el-egans EGL-9 and mammalian homologs define afamily of dioxygenases that regulate HIF by prolylhydroxylation. Cell. 2001;107:43-54.

9. Maxwell PH, Wiesener MS, Chang GW, et al. Thetumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteoly-sis. Nature. 1999;399:271-275.

10. Min JH, Yang H, Ivan M, Gertler F, Kaelin WG Jr,Pavletich NP. Structure of an HIF-1�-pVHL com-plex: hydroxyproline recognition in signaling. Sci-ence. 2002;296:1886-1889.

11. Hon WC, Wilson MI, Harlos K, et al. Structuralbasis for the recognition of hydroxyproline inHIF-1� by pVHL. Nature. 2002;417:975-978.

12. Hirsila M, Koivunen P, Gunzler V, Kivirikko KI,Myllyharju J. Characterization of the human prolyl4-hydroxylases that modify the hypoxia-induciblefactor. J Biol Chem. 2003;278:30772-30780.

13. Lando D, Peet DJ, Whelan DA, Gorman JJ,Whitelaw ML. Asparagine hydroxylation of theHIF transactivation domain a hypoxic switch. Sci-ence. 2002;295:858-861.

14. Cockman ME, Lancaster DE, Stolze IP, et al.Posttranslational hydroxylation of ankyrin repeatsin I�B proteins by the hypoxia-inducible factor(HIF) asparaginyl hydroxylase, factor inhibitingHIF (FIH). Proc Natl Acad Sci U S A. 2006;103:14767-14772.

15. Lieb ME, Menzies K, Moschella MC, Ni R, Taub-man MB. Mammalian EGLN genes have distinct

patterns of mRNA expression and regulation. Bio-chem Cell Biol. 2002;80:421-426.

16. Rohrbach S, Simm A, Pregla R, Franke C,Katschinski DM. Age-dependent increase ofprolyl-4-hydroxylase domain (PHD) 3 expressionin human and mouse heart. Biogerontology.2005;6:165-171.

17. Menzies K, Liu B, Kim WJ, Moschella MC, Taub-man MB. Regulation of the SM-20 prolyl hydroxy-lase gene in smooth muscle cells. Biochem Bio-phys Res Commun. 2004;317:801-810.

18. Takeda K, Ho VC, Takeda H, Duan LJ, Nagy A,Fong GH. Placental but not heart defects are as-sociated with elevated hypoxia-inducible factor �levels in mice lacking prolyl hydroxylase domainprotein 2. Mol Cell Biol. 2006;26:8336-8346.

19. Masuoka HC, Townes TM. Targeted disruption ofthe activating transcription factor 4 gene results insevere fetal anemia in mice. Blood. 2002;99:736-745.

20. Ameri K, Lewis CE, Raida M, Sowter H, Hai T,Harris AL. Anoxic induction of ATF-4 through HIF-1–independent pathways of protein stabilizationin human cancer cells. Blood. 2004;103:1876-1882.

21. Oehme F, Jonghaus W, Narouz-Ott L, Huetter J,Flamme I. A nonradioactive 96-well plate assayfor the detection of hypoxia-inducible factor prolyl





hydroxylase activity. Anal Biochem. 2004;330:74-80.

22. Monney L, Otter I, Olivier R, et al. Defects in theubiquitin pathway induce caspase-independentapoptosis blocked by Bcl-2. J Biol Chem. 1998;273:6121-6131.

23. Salceda S, Caro J. Hypoxia-inducible factor 1�(HIF-1�) protein is rapidly degraded by the ubiq-uitin-proteasome system under normoxic condi-tions: its stabilization by hypoxia depends on re-dox-induced changes. J Biol Chem. 1997;272:22642-22647.

24. National Center for Biotechnology Informationdatabase, Genbank. http://www.ncbi.nlm.nih.gov/Genbank/index.html. Accessed August 11,2003.

25. Rutkowski DT, Kaufman RJ. All roads lead toATF4. Dev Cell. 2003;4:442-444.

26. Hopfer U, Hopfer H, Jablonski K, Stahl RA, WolfG. The novel WD-repeat protein Morg1 acts as amolecular scaffold for hypoxia-inducible factorprolyl hydroxylase 3 (PHD3). J Biol Chem. 2006;281:8645-8655.

27. Blais JD, Filipenko V, Bi M, et al. Activating tran-scription factor 4 is translationally regulated byhypoxic stress. Mol Cell Biol. 2004;24:7469-7482.

28. Stiehl DP, Wirthner R, Koditz J, Spielmann P, Ca-menisch G, Wenger RH. Increased prolyl 4-hy-droxylase domain proteins compensate for de-creased oxygen levels: evidence for an

autoregulatory oxygen-sensing system. J BiolChem. 2006;281:23482-23491.

29. Wirthner R, Kuppusamy B, Stiehl DP, et al. Deter-mination and modulation of prolyl-4-hydroxylasedomain (PHD) oxyen sensor activity. MethodsEnzymol. 2007;in press.

30. Hai T, Hartman MG. The molecular biology andnomenclature of the activating transcription fac-tor/cAMP responsive element binding family oftranscription factors: activating transcription fac-tor proteins and homeostasis. Gene. 2001;273:1-11.

31. Fawcett TW, Martindale JL, Guyton KZ, Hai T,Holbrook NJ. Complexes containing activatingtranscription factor (ATF)/cAMP-responsive-ele-ment-binding protein (CREB) interact with theCCAAT/enhancer-binding protein (C/EBP)-ATFcomposite site to regulate Gadd153 expressionduring the stress response. Biochem J. 1999;339:135-141.

32. Chen C, Dudenhausen E, Chen H, Pan YX,Gjymishka A, Kilberg MS. Amino-acid limitationinduces transcription from the human C/EBP�gene via an enhancer activity located down-stream of the protein coding sequence. BiochemJ. 2005;391:649-658.

33. Ohoka N, Yoshii S, Hattori T, Onozaki K, HayashiH. TRB3, a novel ER stress-inducible gene, isinduced via ATF4-CHOP pathway and is involvedin cell death. EMBO J. 2005;24:1243-1255.

34. Lassot I, Segeral E, Berlioz-Torrent C, et al. ATF4

degradation relies on a phosphorylation-depen-dent interaction with the SCF(�TrCP) ubiquitinligase. Mol Cell Biol. 2001;21:2192-2202.

35. Ameri K, Hammond EM, Culmsee C, et al. Induc-tion of activating transcription factor 3 by anoxiais independent of p53 and the hypoxic HIF signal-ling pathway. Oncogene. 2007;26:284-289.

36. Carmeliet P, Dor Y, Herbert JM, et al. Role ofHIF-1� in hypoxia-mediated apoptosis, cell prolif-eration and tumour angiogenesis. Nature. 1998;394:485-490.

73. Carriere A, Carmona MC, Fernandez Y, et al. Mi-tochondrial reactive oxygen species control thetranscription factor CHOP-10/GADD153 and adi-pocyte differentiation: a mechanism for hypoxia-dependent effect. J Biol Chem. 2004;279:40462-40469.

38. Elvidge GP, Glenny L, Appelhoff RJ, Ratcliffe PJ,Ragoussis J, Gleadle JM. Concordant regulationof gene expression by hypoxia and 2-oxogluta-rate-dependent dioxygenase inhibition: the role ofHIF-1�, HIF-2�, and other pathways. J BiolChem. 2006;281:15215-15226.

39. Han XJ, Chae JK, Lee MJ, You KR, Lee BH, KimDG. Involvement of GADD153 and cardiacankyrin repeat protein in hypoxia-induced apopto-sis of H9c2 cells. J Biol Chem. 2005;280:23122-23129.





university of zurich - uzh...jens ko¨ditz,1 jutta nesper,2 marieke wottawa,1 daniel p. stiehl,3...

Documents