Donor treatment with a PHD-inhibitor activatingHIFs prevents graft injury and prolongs survivalin an allogenic kidney transplant modelW. M. Bernhardta,1,2, U. Gottmannb,1, F. Doyonb, B. Buchholza, V. Campeanc, J. Schodela, A. Reisenbuechlerb, S. Klausd,M. Arendd, L. Flippind, C. Willama, M. S. Wiesenera, B. Yardb, C. Warneckea, and K.-U. Eckardta

aDepartment of Nephrology and Hypertension, Friedrich–Alexander University, Erlangen–Nuremberg, Germany; bV Department of Medicine(Nephrology/Endocrinology/Rheumatology), University Hospital Mannheim, Mannheim, Germany; cInstitute for Pathology, Friedrich–Alexander University,Erlangen–Nuremberg, Germany; and dFibroGen, Inc., South San Francisco, CA

Edited by Gregg L. Semenza, Johns Hopkins University School of Medicine, Baltimore, MD, and approved October 12, 2009 (received for review May 5, 2009)

Long-term survival of renal allografts depends on the chronicimmune response and is probably influenced by the initial injurycaused by ischemia and reperfusion. Hypoxia-inducible transcrip-tion factors (HIFs) are essential for adaptation to low oxygen.Normoxic inactivation of HIFs is regulated by oxygen-dependenthydroxylation of specific prolyl-residues by prolyl-hydroxylases(PHDs). Pharmacological inhibition of PHDs results in HIF accumu-lation with subsequent activation of tissue-protective genes. Weexamined the effect of donor treatment with a specific PHDinhibitor (FG-4497) on graft function in the Fisher–Lewis rat modelof allogenic kidney transplantation (KTx). Orthotopic transplanta-tion of the left donor kidney was performed after 24 h of coldstorage. The right kidney was removed at the time of KTx (acutemodel) or at day 10 (chronic model). Donor animals received asingle dose of FG-4497 (40 mg/kg i.v.) or vehicle 6 h before donornephrectomy. Recipients were followed up for 10 days (acutemodel) or 24 weeks (chronic model). Donor preconditioning withFG-4497 resulted in HIF accumulation and induction of HIF targetgenes, which persisted beyond cold storage. It reduced acute renalinjury (serum creatinine at day 10: 0.66 � 0.20 vs. 1.49 � 1.36mg/dL; P < 0.05) and early mortality in the acute model andimproved long-term survival of recipient animals in the chronicmodel (mortality at 24 weeks: 3 of 16 vs. 7 of 13 vehicle-treatedanimals; P < 0.05). In conclusion, pretreatment of organ donorswith FG-4497 improves short- and long-term outcomes after allo-genic KTx. Inhibition of PHDs appears to be an attractive strategyfor organ preservation that deserves clinical evaluation.

hypoxia-inducible transcription factors � kidney injury � protection �transplantation

In patients suffering from end-stage renal disease, kidneytransplantation (KTx) offers the best form of renal replace-

ment therapy, leading to improved quality of life and survival incomparison to dialysis (1). However, organ shortage significantlylimits the opportunities for transplantation. Moreover, despiteimprovements in immunosuppressive therapy leading to a de-creased incidence of acute rejection episodes, there has beenlittle improvement in long-term graft survival. The mean half-life of renal allografts from cadaveric donors is �8 years (2).Accordingly, chronic allograft nephropathy contributes signifi-cantly to the need for initiation of regular dialysis (3). Reducingthe risk for late graft loss therefore remains a major challenge inKTx.

The mechanisms of chronic allograft failure reflect a complexinterplay between immune and nonimmune mechanisms (4).Acute kidney injury immediately after transplantation, which isinduced by the sequence of cold ischemia, warm ischemia, andreperfusion, in conjunction with an early immune response, notonly impairs early graft function but is associated with reducedlong-term graft survival (5, 6). Although causality of this link hasnot been proven, it is tempting to assume that minimizing early

graft injury improves long-term outcome. At the same time, theneed to extend donor criteria in light of organ shortage furtherincreases the risk for delayed graft function. Several approacheshave already been taken to reduce early graft dysfunction,including optimization of the perfusion fluid (7, 8) and the useof pharmaceuticals that interact with single steps in the patho-genesis of ischemia reperfusion injury (9). We reasoned that aneffective way to improve early and possibly late graft functioncould be the induction of an array of endogenous protectivemechanisms before the initiation of the acute injury associatedwith transplantation. Moreover, we hypothesized that the induc-tion of genes that have become inducible by hypoxia duringevolution could confer such protection. Testing this hypothesishas become possible because of recent progress in understandingthe mechanisms by which hypoxia influences gene expressionthrough activation of hypoxia-inducible transcription factors(HIFs). HIFs are heterodimers of a constitutive �-subunit(HIF-�) and one of two alternative oxygen-regulated �-subunits,HIF-1� or HIF-2�. Under normoxia, HIF� is rapidly degradedvia the ubiquitin–proteasome pathway. To target HIF-� fordegradation, two specific prolyl residues are hydroxylated byHIF-prolyl hydroxylases [prolyl hydroxylation domain (PHD)proteins] that require dioxygen and 2-oxoglutarate as cosub-strates (10, 11). Oxoglutarate analogues therefore act as com-petitive inhibitors of PHDs (PHD-I) and HIF stabilizers. Todate, far more than 100 HIF target genes have been identified,including erythropoietin (EPO), VEGF, glucose transporters,and heme oxygenase-1 (HO-1), which have the potential toconfer nephroprotection under different conditions (12). Weand others have previously shown that the application of PHD-Inhibitor stabilizes HIF in the kidney (13, 14) and confersprotection against prolonged warm renal ischemia (14, 15) andtoxic renal injury (16). Here, we have used a small moleculePHD-I (FG-4497) to test the hypothesis that stabilization of HIFin organ donors prevents early graft dysfunction in an allogenicrat kidney transplant model and that this effect translates into along-term improvement of allograft function.

ResultsFG-4497 Stabilizes HIF-� and Induces HIF Target Genes in HumanProximal Tubular Cells. FG-4497 (FibroGen, Inc.) was identifiedfrom a PHD-I library for its capacity to activate the HIF pathway

Author contributions: W.M.B. and K.-U.E. designed research; W.M.B., U.G., F.D., B.B., J.S.,A.R., B.Y., and C. Warnecke performed research; S.K., M.A., and L.F. contributed newreagents; W.M.B., U.G., F.D., B.B., V.C., J.S., C. Willam, and C. Warnecke analyzed data; andW.M.B., M.S.W., and K.-U.E. wrote the paper.

Conflict of interest statement: S.K., M.A., and L.F. are employees of FibroGen.

This article is a PNAS Direct Submission.

1W.M.B. and U.G. contributed equally to this work.

2To whom correspondence should be addressed. E-mail: wanja.bernhardt@uk-erlangen.de.

This article contains supporting information online at www.pnas.org/cgi/content/full/0903978106/DCSupplemental.

21276–21281 � PNAS � December 15, 2009 � vol. 106 � no. 50 www.pnas.org�cgi�doi�10.1073�pnas.0903978106

Dow

nloa

ded

by g

uest

on

Nov

embe

r 5,

202

0

in vitro (17). It has previously been demonstrated that 4–6 hafter treatment of rats with FG-4497, EPO production is induced(18) and protection occurs against ex vivo ischemic injury inisolated perfused kidneys (19). Moreover, FG-4497 was found toprotect mucosal cells in a murine colitis model (18) and neuronalcells (20). To test its ability to inhibit HIF degradation in kidneycells, increasing concentrations of FG-4497 were applied in vitro.Immunoblotting of protein extracts of HKC-8 cells (immortal-ized human tubular cells) showed stabilization of both HIF-�isoforms (HIF-1� and HIF-2�) after treatment with dipyridyl(positive control) or FG-4497 (0.5–100 �M) for 6 h (Fig. 1A).Analysis of the time course of HIF induction by FG-4497

revealed very rapid stabilization of HIF-� (15 min) and stableHIF protein levels for up to 6 h (Fig. 1B). To test the time courseof HIF target gene expression, mRNA levels of angiopoietin-like4 (AngPTL4), glucose transporter-1 (Glut-1), and VEGF weredetermined for up to 6 h. Expression levels of the genes wereinitially elevated after 2 h and increased further up to 6 h.

FG-4497 Stabilizes HIF-� in Rat Kidneys, and HIF Remains Stable AfterCold Ischemia. Next, we injected FG-4497 into rats to test the timecourse of HIF stabilization in the kidney in vivo. Both HIF-�isoforms were detectable in rat kidneys 1, 4, and 6 h after i.v.injection of FG-4497 at a dose of 40 mg/kg, which was found tostimulate HIF in prior dose-finding experiments (Methods; Fig.2A). There was no apparent difference between the extent ofHIF accumulation between 1 and 6 h. Based on these results, thetime course of target gene expression (Fig. 1C), and publisheddata (14, 18–21), we chose a 6-h interval between injection andkidney explantation for cold preservation. To determinewhether HIF-� is still detectable after cold ischemia, kidneyswere removed, perfused with University of Wisconsin solution,and maintained in this solution for 24 h at 4 °C following priortreatment with FG-4497. Immunohistochemistry revealed strongintranuclear accumulation of HIF-� after FG-4497 treatment,whereas cold ischemia did not induce HIF-� in vehicle (Veh)-treated rats (Fig. 2B). The expression pattern of the two HIF-�isoforms stabilized by FG 4497 resembles physiological activa-tion under systemic hypoxia (22); HIF-1� was expressed intubular epithelial cells [Fig. 2 A (Upper) and B (Left)], andHIF-2� was expressed in interstitial, glomerular, and endothelialcells [Fig. 2 A (Lower) and B (Right)].

FG-4497 Transactivates HIF Target Genes in Vivo. To examine theactivation of HIF target genes 6 h after i.v. injection of FG-4497(40 mg/kg) with and without subsequent cold storage for 24 h,we performed real-time PCR for selected known HIF targetgenes, which may be of particular relevance for a potentialnephroprotective effect. FG-4497 was found to increase HO-1and EPO 4–5-fold and more than 200-fold, respectively, inde-pendent of cold storage (Fig. 3). In addition, HIF target genesinvolved in glucose metabolism, such as Glut-1 or hexokinase,were increased (Fig. S1A). Given their potential pathophysio-logical relevance, we also tested the expression of inflammatorycytokines known to be inducible by HIF and found that IL-1�was induced approximately 2-fold by FG-4497 [0 h, of coldischemia (c.i.): 2.27 � 0.19; 24 h of c.i.: 2.23 � 0.49]. To identifyadditional genes that may be involved in mediating the protectiveeffect, we performed an oligonucleotide microarray analysis ofwhole kidney extracts. Table S1 shows additional genes that weremore than 2.0-fold up-regulated in whole kidney extracts 6 hafter application of FG-4497. The change of mRNA in renalcortical extract may underestimate the change in expression levelof the respective gene in specific cell populations that contributeonly a proportion to total kidney mRNA. The microarrayrevealed that FG-4497 up-regulated known HIF target genes andadditional genes not previously identified as HIF target genes.The activation of all these genes remained stable over 24 h ofcold storage. Change in expression of the 2 most strongly up- anddown-regulated genes (Rarres 2, Lox, and IGFBP3, Id2, respec-tively) was confirmed by Real Time-PCR (Fig. S1 B and C).Single genes, such as the sodium channel Scnn1a, were inducedby cold ischemia only but not by FG-4497. Taken together,FG-4497 stabilizes HIF and transactivates HIF target genes inrat kidneys, and this effect remains stable during cold ischemia,which is a prerequisite for a potentially protective effect duringreperfusion. We then tested the effect of donor pretreatmentwith FG-4497 in a rat model of allograft KTx.

Fig. 1. Induction of HIF-� and HIF target genes by FG-4497 in humanproximal tubular cells. (A) HKC-8 cells were exposed to dipyridyl (DP, positivecontrol) or FG-4497 (0.5, 1, 5, 10, 25, 50, and 100 �M) for 6 h. Immunoblottingfor HIF-� protein shows dose-dependent accumulation of HIF-1� and HIF-2�

by FG-4497, with a maximal level of HIF-� stabilization at concentrationsabove 10 �M. (B) Fifteen minutes after FG-4497 treatment (150 �M), HIF-1�

and HIF-2� were stabilized and remained stable over the observation periodof 6 h. (C) With a delay of approximately 2 h, the mRNA (real-time PCR) of theHIF target genes AngPTL4, Glut-1, and VEGF was significantly up-regulatedcompared with baseline (set as ‘‘1’’) and further increased up to 6 h aftertreatment with FG-4497. Normoxic untreated cells (N) served as a negativecontrol.

Bernhardt et al. PNAS � December 15, 2009 � vol. 106 � no. 50 � 21277

MED

ICA

LSC

IEN

CES

Dow

nloa

ded

by g

uest

on

Nov

embe

r 5,

202

0

Donor Treatment with FG-4497 Ameliorates Renal Function in theAcute Phase of Allograft KTx. To test the effect of FG-4497 on earlygraft function, the left kidney from a donor animal (Fisherstrain) treated with FG-4497 or Veh was transplanted ortho-topically into a recipient animal (Lewis strain) following 24 h ofcold storage, with a warm ischemia period of �30 min. Imme-diately after transplantation, the right kidney of the recipient wasremoved so that survival became graft-dependent and theoccurrence of delayed graft function predictably resulted in the

death of the recipient animal after 2–5 days. Animals were nottreated with immunosuppressants so as not to blunt the devel-opment of allograft injury. In control experiments, the sameprocedure was performed in isogenic animals (Lewis–Lewisstrain). In the allogenic constellation, kidney injury was severe,resulting in survival of only 6 (23.1%) of 26 animals in theVeh-treated group. FG-4497 pretreatment significantly reducedmortality, with 8 (53.3%) of 15 animals surviving (P � 0.019; Fig.4A). In addition, among animals surviving for 10 days, thosereceiving a kidney from a donor pretreated with FG-4497 (n �8) had significantly lower serum creatinine levels as comparedwith animals receiving a transplant from a Veh-treated donor(n � 6) (Fig. 4B). This functional improvement was paralleled byreduced structural damage, as reflected by an acute tubularnecrosis score (Fig. 4C). In contrast, animals subjected toisogenic transplantation all survived (Fig. 4A) and showed onlya very moderate increase in serum creatinine, with no differencebetween animals receiving kidneys from donors pretreated withFG-4497 or Veh (Fig. 4B). However, renal morphology wassignificantly better preserved in kidneys from FG-4497-pretreated donors (Fig. 4C), indicating an effect of FG-4497 onischemic injury that was independent of immunological mis-match. The kidney leukocyte infiltrate as assessed according tothe Banff 97 classification (23) (Table S2) and the number of Tlymphocytes (CD3� cells) (Fig. S2) did not reveal differencesbetween Veh- and FG-4497-treated groups.

Donor Treatment with FG-4497 Significantly Improves Long-TermGraft Survival. To investigate the long-term consequences ofprotection against early graft dysfunction induced by donorpretreatment with FG-4497, an additional group of animals wasstudied in which nephrectomy of the right kidney of the recipientanimal was delayed until day 10 after transplantation. Thisallowed animals to survive periods of early severe graft dysfunc-tion and assessment of the effect of the intervention on chronicgraft failure by studying survival rates. As in the acute setting, wechose not to treat rats with immunosuppressants to acceleratechronic allograft nephropathy. Fig. 5 illustrates that donor

Fig. 2. Kinetics of HIF accumulation after FG-4497 treatment in the kidneys of Fisher (F-344) rats before and after cold ischemia. b.w., body weight. (A) At 1,4, and 6 h after i.v. injection of FG-4497, HIF-1� was immunohistochemically detectable in tubular epithelial cells of rat kidneys and HIF-2� was detectable ininterstitial, endothelial, and glomerular cells. (B) After FG-4497 and cold ischemia of 24 h, HIF-� was still detectable to a comparable degree as before coldischemia. Veh treatment with or without cold ischemia did not lead to detectable HIF-� accumulation.

Fig. 3. Effect of FG-4497 on the expression of HIF target genes. Singlepotentially protective HIF target genes were quantified by real-time PCR 6 hafter FG-4497 treatment without cold ischemia (c.i.) or after an additional 24-hperiod of cold ischemia. The nephroprotective genes HO-1 and EPO werestrongly up-regulated after FG-4497 treatment irrespective of cold ischemia(*, P � 0.05).

21278 � www.pnas.org�cgi�doi�10.1073�pnas.0903978106 Bernhardt et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 5,

202

0

treatment with FG-4497 markedly prolonged graft-dependentsurvival in recipient animals by more than 50%. Two weeks aftertransplantation, when all animals in both groups were still alive,recipients of FG-4497-treated donors already showed a tendencytoward lower serum creatinine concentrations (1.45 � 0.66mg/dL vs. 2.75 � 1.55 mg/dL; P � 0.07). Isogenically trans-planted control animals showed no mortality within the obser-vation period.

FG-4497 Treatment Protects Human Proximal Tubular Cells fromApoptosis. To test whether HIF accumulation induced by FG-4497 protects cells under in vitro conditions mimicking ischemiareperfusion injury, we used an in vitro model of cell injuryinduced by oxygen–glucose deprivation and subsequent reoxy-genation. After 24 h of exposure to 1 Vol% O2 in a glucose-freemedium, cells were reoxygenated (21 Vol% O2) in glucose-containing medium for another 24 h. At the end of the exper-iment, the apoptosis rate was determined. Pretreatment for 6 hwith FG-4497 significantly reduced the rate of apoptosis byabout 30% in the HKC-8 cell line and in primary human tubularcells (Fig. S3A). In vivo, however, there was no detectabledifference in the number of apoptotic cells at day 10 after KTx(Fig. S3B) between surviving recipients of Veh- and FG-4497-treated donors as assessed by Tunel staining.

FG-4497 Has No Influence on NF-�B Signaling. Hydroxylation ofaspargyl-residues of ankyrin repeats by factor-inhibiting HIF(FIH) has been shown to influence signaling pathways in addi-tion to HIF signaling, such as Notch (24, 25) or NF-�B signaling(26). Therefore, we examined NF-�B expression after FG-4497treatment. Using a luciferase assay and electrophoretic mobilityshift assay we found no influence of FG-4497 on NF-�B induc-tion (Fig. S4 A and C) or on protein levels of its regulatorysubunit I�B (Fig. S4B).

DiscussionThis study translates recent knowledge on the molecular mech-anisms of hypoxia-inducible gene expression into a widelyaccepted animal model of allograft nephropathy. Donor pre-treatment using a single dose of a small molecule inhibitor ofHIF PHDs, FG-4497, which stabilizes HIF and activates HIFtarget genes in the kidney, significantly reduced the frequency ofdelayed graft function and markedly improved long-term out-come, suggesting opportunities for improvement of kidney func-tion after transplantation.

Previous experimental data suggest that short periods ofischemia/reperfusion can protect against subsequent ischemicinjury (27–29). However, the clinical application of this so-called‘‘ischemic preconditioning’’ is limited by (i) obvious difficultiesto induce controlled ischemia in vivo, (ii) sensitivity of theprotective effect to changes in the duration and sequence ofischemia and reperfusion episodes, and (iii) the failure toidentify unequivocally the molecular mediators. In contrast, theapproach used in this study relies on mimicking the response tohypoxia through inhibition of the enzymes that degrade thetranscription factor HIF. The analysis of mRNA expression

Fig. 4. Effect of FG-4497 on early graft function and renal morphology. Inthe acute setting, FG-4497 significantly improved survival within the initialdays after KTx (A), indicating a lower incidence of delayed graft function. Inaddition, in surviving animals, FG-4497 donor pretreatment resulted in anearlier and more pronounced decrease in serum creatinine from days 3–10after allogenic transplantation (Allo) when compared with donor pretreat-ment with Veh only (B). In isogenic KTx, there was no measurable differencein serum creatinine and survival between FG-4497 and Veh. (C) Blinded scoringfor acute tubular necrosis revealed better preservation of renal morphologyafter FG-4497 compared with Veh independent of the immunological con-stellation (*, P � 0.05).

Fig. 5. Effect of FG-4497 on long-term graft survival. Kaplan-Meier curvesafter allograft KTx in animals with and without pretreatment of the donorwith FG-4497. Animals that received a renal transplant from an FG-4497-treated donor had significantly better survival rates (black line) than animalstransplanted with a kidney from a Veh-treated donor (dotted gray line). Noneof the isogenic control animals died (dashed gray line). (*, P � 0.05).

Bernhardt et al. PNAS � December 15, 2009 � vol. 106 � no. 50 � 21279

MED

ICA

LSC

IEN

CES

Dow

nloa

ded

by g

uest

on

Nov

embe

r 5,

202

0

confirmed that a number of genes previously identified asHIF-dependent are inducible in the kidney, including EPO,which has previously been demonstrated to be nephroprotectivein renal ischemia-reperfusion models (30–33), and HO-1, whichhas also been shown to improve short- and long-term graftfunction in the same rat model (34, 35). In addition, HIFactivation facilitates cellular anaerobic metabolism by inducingGlut-1 or enzymes involved in glycolysis. Although it is possiblethat among the large number of genes induced by HIF, singlegenes like EPO and HO-1 are of particular importance, identi-fication of their relative contribution is of theoretical rather thanpractical relevance, because the concerted activation is likely tobe superior to single gene induction. In terms of target genespecificity, the parallel activation of HIF-1� and HIF-2� ispossibly of particular relevance. Although HIF-2� transactivatesEPO (36, 37), HIF-1� is the subunit stabilized in tubular cells,and target genes involved in glucose metabolism are predomi-nantly HIF-1�-dependent (38).

The design of this study sheds further light on the dependenceof long-term graft function on early graft injury. Although anassociation between the occurrence of early graft dysfunctionand long-term graft failure has been clearly demonstrated(39–41), causality of this link has not been proven to date. It isthus noteworthy that donor pretreatment limiting early graftdysfunction had a significant impact on long-term graft survival.The HIF system responds primarily to reduced availability ofmolecular oxygen, but changes in cellular function induced byHIF apparently also confer protection against more complextypes of cellular injury, including those induced by the sequenceof ischemia and reperfusion in vivo (14, 42, 43) and partiallymimicked by oxygen–glucose deprivation in vitro. HIF activationby FG-4497 reduced apoptosis in vitro and tubular necrosis invivo. Using Tunel staining, we did not observe reduced apoptosisrates in vivo after 10 days, but this may be attributable to survivalbias, because only animals not developing early graft failuresurvived until this time point. We also observed no effects of HIFstabilization on lymphocyte counts in allogenic animals at day 10after KTx, which might suggest that the observed protection ismost likely attributable to reduction of the ischemic injury ratherthan to a modulation of immunological factors. Nevertheless, thefar better outcome of transplantation between isogenic animalsindicates that the interplay between ischemia reperfusion injuryand immunological factors is of crucial importance both for earlyand late graft dysfunction. HIF has recently been shown toinfluence several aspects of the innate and adaptive immuneresponse (44–46), but donor treatment limits such effects to

immune cells residing in the graft. Although HIF may stimulatedendritic cells (47), any potential adverse effects of such stim-ulation are obviously offset by inhibition of ischemia reperfusioninjury. FIH has also been shown to hydroxylate defined residuesof ankyrin repeats, thereby possibly affecting the regulation ofother signaling pathways, including Notch (24, 25) and NF-�B(26). Although we found no evidence that FG-4497 influencesNF-�B or I�B-signaling, non-HIF-dependent effects of hydrox-ylase inhibition cannot be definitively ruled out.

Although current management of organ donors is largelyfocused on maintenance of physical homeostasis, this studysuggests that there is significant potential to improve the resultsof organ transplantation by inducing protective biological mech-anisms before organ explantation. A recent analysis of humankidney graft biopsies obtained immediately after engraftmentrevealed significant variability in the extent of baseline HIFexpression (48), consistent with animal experiments showingthat ischemia reperfusion injury, per se, is only a weak stimulusfor HIF activation (22). Interestingly, however, renal allograftswith higher baseline HIF expression tended to have betterfunctional outcomes (48), suggesting that the current experi-mental strategy of pharmacological HIF induction is potentiallysuitable for translation into clinical practice. The 6-h timeinterval between drug dosing and nephrectomies used in thisstudy would be achievable in human organ donors, and futurestudies will have to address whether shorter pretreatment peri-ods are sufficient to induce protection, even though HIF-dependent gene expression is not yet maximal. The potentialapplicability of our approach to humans is underscored by thefact that a small molecule inhibitor of PHDs was used, and othermembers of this class of drugs are already being developed forthe treatment of anemia through stimulation of endogenousEPO production (49).

MethodsDetails for cell culture, HIF, and I�B� protein extraction and immunoblotting,animal experiments/transplantation protocol, functional studies, tissue prep-aration, immunohistochemistry, morphological analysis, real-time PCR, oxy-gen–glucose deprivation, caspase 3/7 assay, oligonucleotide microarray anal-ysis, luciferase reporter assays, electrophoretic mobility shift assay, andstatistics are given in SI Materials and Methods.

ACKNOWLEDGMENTS. We thank B. Teschemacher, H. Fees, and A. Kosel forexcellent technical assistance. The work was supported by the CollaborativeResearch Center of the Deutsche Forschungsgemeinschaft (Grant SFB 423)‘‘Kidney Injury: Pathogenesis and Regenerative Mechanisms’’ (TP A14). B.B.was the recipient of a grant from the Interdisciplinary Centre for ClinicalResearch at the University of Erlangen–Nuremberg.

1. Wolfe RA, et al. (1999) Comparison of mortality in all patients on dialysis, patients ondialysis awaiting transplantation, and recipients of a first cadaveric transplant. N EnglJ Med 341:1725–1730.

2. Meier-Kriesche HU, Schold JD, Kaplan B (2004) Long-term renal allograft survival: Havewe made significant progress or is it time to rethink our analytic and therapeuticstrategies? Am J Transplant 4:1289–1295.

3. Vadivel N, Tullius SG, Chandraker A (2007) Chronic allograft nephropathy. SeminNephrol 27(4):414–429.

4. Kaplan B (2006) Overcoming barriers to long-term graft survival. Am J Kidney Dis 47(4suppl. 2):S52–S64.

5. Baid-Agrawal S, Frei UA (2007) Living donor renal transplantation: Recent develop-ments and perspectives. Nat Clin Pract Nephrol 3:31–41.

6. CeckaJM(2000)TheUNOSScientificRenalTransplantRegistry—2000.ClinTranspl2000:1–18.

7. Faenza A, et al. (2001) Kidney preservation with University of Wisconsin and Celsiorsolution: A prospective multicenter randomized study. Transplantation 72:1274–1277.

8. Hauet T, et al. (2003) A modified University of Wisconsin preservation solution withhigh-NA� low-K� content reduces reperfusion injury of the pig kidney graft. Trans-plantation 76:18–27.

9. Cugini D, et al. (2005) Inhibition of the chemokine receptor CXCR2 prevents kidneygraft function deterioration due to ischemia/reperfusion. Kidney Int 67:1753–1761.

10. Semenza GL (2004) Hydroxylation of HIF-1: Oxygen sensing at the molecular level.Physiology (Bethesda) 19:176–182.

11. Kaelin WG, Jr, Ratcliffe PJ (2008) Oxygen sensing by metazoans: The central role of theHIF hydroxylase pathway. Mol Cell 30:393–402.

12. Bernhardt WM, et al. (2007) Organ protection by hypoxia and hypoxia-induciblefactors. Methods Enzymol 435:221–245.

13. Warnecke C, et al. (2003) Activation of the hypoxia-inducible factor-pathway andstimulation of angiogenesis by application of prolyl hydroxylase inhibitors. FASEB J17:1186–1188.

14. Bernhardt WM, et al. (2006) Preconditional activation of hypoxia-inducible factorsameliorates ischemic acute renal failure. J Am Soc Nephrol 17:1970–1978.

15. Hill P, et al. (2008) Inhibition of hypoxia inducible factor hydroxylases protects againstrenal ischemia-reperfusion injury. J Am Soc Nephrol 19:39–46.

16. Weidemann A, et al. (2008) HIF activation protects from acute kidney injury. J Am SocNephrol 19:486–494.

17. Norman LO (1983) US Patent 4,379,752.18. Robinson A, et al. (2008) Mucosal protection by hypoxia-inducible factor prolyl hy-

droxylase inhibition. Gastroenterology 134:145–155.19. Rosenberger C, et al. (2008) Activation of hypoxia-inducible factors ameliorates hy-

poxic distal tubular injury in the isolated perfused rat kidney. Nephrol Dial Transplant23:3472–3478.

20. Milosevic J, et al. (2009) Non-hypoxic stabilization of hypoxia-inducible factor alpha(HIF-�): Relevance in neural progenitor/stem cells. Neurotox Res 15:367–380.

21. Schneider C, et al. (2009) Short-term effects of pharmacologic HIF stabilization onvasoactive and cytotrophic factors in developing mouse brain. Brain Res 1280:43–51.

22. Rosenberger C, et al. (2002) Expression of hypoxia-inducible factor-1� and -2� inhypoxic and ischemic rat kidneys. J Am Soc Nephrol 13:1721–1732.

23. Racusen LC, et al. (1999) The Banff 97 working classification of renal allograft pathol-ogy. Kidney Int 55(2):713–723.

21280 � www.pnas.org�cgi�doi�10.1073�pnas.0903978106 Bernhardt et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 5,

202

0

24. Coleman ML, et al. (2007) Asparaginyl hydroxylation of the Notch ankyrin repeatdomain by factor inhibiting hypoxia-inducible factor. J Biol Chem 282(33):24027–24038.

25. Zheng X, et al. (2008) Interaction with factor inhibiting HIF-1 defines an additionalmode of cross-coupling between the Notch and hypoxia signaling pathways. Proc NatlAcad Sci USA 105:3368–3373.

26. Cockman ME, et al. (2006) Posttranslational hydroxylation of ankyrin repeats inIkappaB proteins by the hypoxia-inducible factor (HIF) asparaginyl hydroxylase, factorinhibiting HIF (FIH). Proc Natl Acad Sci USA 103:14767–71472.

27. Bonventre JV (2002) Kidney ischemic preconditioning. Curr Opin Nephrol Hypertens11:43–48.

28. Murry CE, Jennings RB, Reimer KA (1986) Preconditioning with ischemia: A delay oflethal cell injury in ischemic myocardium. Circulation 74:1124–1136.

29. Schlaifer JD, Kerensky RA (1997) Ischemic preconditioning: Clinical relevance andinvestigative studies. Clin Cardiol 20:602–606.

30. Patel NS, et al. (2004) Pretreatment with EPO reduces the injury and dysfunctioncaused by ischemia/reperfusion in the mouse kidney in vivo. Kidney Int 66:983–989.

31. Sharples EJ, et al. (2004) Erythropoietin protects the kidney against the injuryand dysfunction caused by ischemia-reperfusion. J Am Soc Nephrol 15:2115–2124.

32. Yang CW, et al. (2003) Preconditioning with erythropoietin protects against subse-quent ischemia-reperfusion injury in rat kidney. FASEB J 17:1754–1755.

33. Vesey DA, et al. (2004) Erythropoietin protects against ischaemic acute renal injury.Nephrol Dial Transplant 19:348–355.

34. Martins PN, et al. (2005) Induction of heme oxygenase-1 in the donor reduces graftimmunogenicity. Transplant Proc 37:384–386.

35. Kotsch K, et al. (2006) Improved long-term graft survival after HO-1 induction inbrain-dead donors. Am J Transplant 6:477–486.

36. Rankin EB, et al. (2007) Hypoxia-inducible factor-2 (HIF-2) regulates hepatic erythro-poietin in vivo. J Clin Invest 117:1068–1077.

37. Warnecke C, et al. (2004) Differentiating the functional role of hypoxia-induciblefactor (HIF)-1alpha and HIF-2alpha (EPAS-1) by the use of RNA interference: Eryth-ropoietin is a HIF-2alpha target gene in Hep3B and Kelly cells. FASEB J 18:1462–1464.

38. Iyer NV, et al. (1998) Cellular and developmental control of O2 homeostasis byhypoxia-inducible factor 1 alpha. Genes Dev 12:149–162.

39. Shoskes DA, Cecka JM (1997) Effect of delayed graft function on short- and long-termkidney graft survival. Clin Transpl 297–303.

40. Peeters P, Terryn W, Vanholder R, Lameire N (2004) Delayed graft function in renaltransplantation. Curr Opin Crit Care 10:489–498.

41. Koning OH, et al. (1997) Risk factors for delayed graft function in cadaveric kidneytransplantation: A prospective study of renal function and graft survival after preser-vation with University of Wisconsin solution in multi-organ donors. European Multi-center Study Group. Transplantation 63:1620–1628.

42. Matsumoto M, et al. (2003) Induction of renoprotective gene expression by cobaltameliorates ischemic injury of the kidney in rats. J Am Soc Nephrol 14:1825–1832.

43. Cai Z, et al. (2008) Complete loss of ischaemic preconditioning-induced cardioprotec-tion in mice with partial deficiency of HIF-1 alpha. Cardiovasc Res 77:463–470.

44. Peyssonnaux C, et al. (2005) HIF-1alpha expression regulates the bactericidal capacityof phagocytes. J Clin Invest 115:1806–1815.

45. Zinkernagel AS, Johnson RS, Nizet V (2007) Hypoxia inducible factor (HIF) function ininnate immunity and infection. J Mol Med 85:1339–1346.

46. Cramer T, et al. (2003) HIF-1alpha is essential for myeloid cell-mediated inflammation.Cell 112:645–657.

47. Jantsch J, et al. (2008) Hypoxia and hypoxia-inducible factor-1 alpha modulate lipopo-lysaccharide-induced dendritic cell activation and function. J Immunol 180:4697–4705.

48. Rosenberger C, et al. (2007) Immunohistochemical detection of hypoxia-induciblefactor-1alpha in human renal allograft biopsies. J Am Soc Nephrol 18:343–351.

49. Hsieh MM, et al. (2007) HIF prolyl hydroxylase inhibition results in endogenous EPOinduction, erythrocytosis, and modest fetal hemoglobin expression in rhesus ma-caques. Blood 110:2140–2147.

Bernhardt et al. PNAS � December 15, 2009 � vol. 106 � no. 50 � 21281

MED

ICA

LSC

IEN

CES

Dow

nloa

ded

by g

uest

on

Nov

embe

r 5,

202

0