Extracellular Superoxide Dismutase Functions as a Major Repressor of Hypoxia-Induced Erythropoietin Gene Expression
http://www.100md.com
内分泌学杂志 2005年第1期
Departments of Medicine (I.N.Z., R.J.F.) and Cell Biology (R.J.F.), Duke University Medical Center, Durham, North Carolina 27710
Address all correspondence and requests for reprints to: Rodney J. Folz, M.D., Ph.D., Division of Pulmonary, Allergy, and Critical Care, Department of Medicine, Duke University Medical Center, Box 2620/Room 331 MSRB, Durham, North Carolina 27710. E-mail: rodney.folz@duke.edu.
Abstract
Hypoxia and biological responses to hypoxia are commonly encountered in both normal and pathologic cellular processes. Here we report that extracellular superoxide dismutase (EC-SOD) plays a major role in regulating the magnitude of hypoxia-induced erythropoietin (Epo) gene expression, thus implicating superoxide as an intermediary signal transduction molecule critical to this process. We found that mice which have the EC-SOD gene inactivated show a marked more than 100-fold elevation in hypoxia-induced Epo gene expression, compared with wild-type controls, which was both dose and time dependent. These mice also showed a significant increase in serum Epo levels after 1 d hypoxia. Interestingly, despite elevated Epo levels, reciprocal changes in hematocrit and reticulocyte counts were not found, suggesting that this newly synthesized Epo lacks functional hematopoietic effects. When EC-SOD was overexpressed in Hep3B cells, we found a significant reduction in Epo gene induction by both CoCl2 (50 μM) and hypoxia (1% O2). Similar findings were noted with another hypoxia-inducible gene, carbonic anhydrase IX. We conclude that EC-SOD functions as a major repressor of hypoxia-induced Epo gene expression, which implicates superoxide as a signaling intermediate whose downstream effects, at least in part, may be mediated by HIF-1.
Introduction
ADAPTATION OF VERTEBRATES to the oxygen-rich environment require regulatory mechanisms that allow them to respond to a changing oxygen concentration. When mammalian organisms are exposed to a hypoxic environment, the red blood cell count increases significantly to boost oxygen delivery to the hypoxic tissues. Erythropoietin (Epo) is the major hormonal signal involved in this hypoxia-induced response (1). Epo is synthesized primarily in kidney, and secondarily in liver, where it is released into the circulation and binds to its major target, the Epo receptor. The interaction between ligand and receptor stimulates production of red blood cells in the bone marrow. The regulation of Epo expression by hypoxia is quite complex and involves first stabilization/activation of hypoxia-inducible factor 1 (HIF-1), which then translocates into the nucleus, where it dimerizes with the arylhydrocarbon receptor nuclear translocator (HIF-1?) and binds the hypoxia-responsive enhancer located in the 3'-region of the Epo gene (for review see Refs. 2 and 3). The molecular mechanisms involved in oxygen-dependent stabilization of HIF-1 has recently been elucidated in great detail. Briefly, under normoxic conditions, HIF-1 protein is hydroxylated at its prolyl residues located in the oxygen-dependent domain (ODD) by a family of prolyl hydroxylase enzymes (4, 5). Hydroxylation of HIF-1 initiates binding of the Von Hipple-Lindau protein that acts as an E3 ubiquitin ligase that then promotes the degradation of HIF-1 by proteosomes. The activity of these prolyl hydroxylases inherently depends on oxygen but also require 2-oxoglutarate, ascorbic acid, and iron. On the other hand, prolyl hydroxylase activities are inhibited by cobaltous ions and by iron chelation, explaining why HIF-1 is activated by transition metals and by the iron-chelating agent desferrioxamine (6).
Although this model of HIF-1 activation explains induction of Epo by hypoxia and cobalt chloride, it fails to explain the role reactive oxygen species (ROS) play in this process. The role of superoxide radicals in hypoxia-mediated transduction pathways is not clearly defined and sometimes controversial. For example, it has been proposed that levels of ROS increase under hypoxic conditions due to malfunctioning of the mitochondrial electron transport chain. The role of mitochondria-generated ROS in the regulation of HIF-1 activity is not clear, and several studies have reported conflicting and confusing results. For example, it has been shown that mitochondria generated superoxide radicals, and their reaction product hydrogen peroxide, are required for induction of HIF-1 activity and other subsequent downstream target genes in hypoxic tissues (7, 8). On the other hand, using similar pharmacological and genetic approaches, others have provided compelling evidence that the mitochondrial respiratory chain is not essential for the regulation of HIF-1 activity by oxygen (9, 10). Furthermore, an alternative model for HIF-1 regulation has been proposed. In this model, hypoxia leads to a decreased production of ROS due to inhibition of nicotinamide adenine dinucleotide phosphate oxidase activity. The fast diffusing ROS oxidize and destabilize HIF-1 protein resulting in a decreased expression of hypoxia-inducible genes, including Epo. The decrease in superoxide production after hypoxia leads to HIF-1 stabilization and activation of gene expression (11, 12).
A major class of antioxidant enzymes involved in maintaining homeostatic levels of ROS, such as superoxide, are represented by the superoxide dismutases (SODs). SODs are a family of metalloenzymes that catalyze the dismutation of O2– to H2O2. Mammalian cells possess three distinct forms of SODs; extracellular SOD (EC-SOD) present in the extracellular spaces, manganese SOD, found exclusively in the mitochondria, and copper-zinc SOD located in the cytoplasm (13, 14). EC-SOD is the principal enzymatic antioxidant in extracellular spaces (15) and plays an important role in the protection of mammalian organisms against superoxide. Genetic disruption of EC-SOD decreases survival time of mice exposed to high oxygen tension (16). Overexpression of EC-SOD in transgenic mice preserves lung function and reduces inflammation caused by influenza pneumonia (17) and hyperoxia (18). The human and mouse EC-SOD is a homotetrameric protein with molecular mass of 135 kDa, whereas the rat enzyme is dimeric (19, 20). Expression levels of EC-SOD differs dramatically among tissues, but the highest expression is found in kidney, lung, and heart (21, 22). EC-SOD is glycosylated and exhibits affinity for heparin or heparan sulfate. Thus, although detectable in blood plasma, most of the enzyme exists bound to the extracellular matrix milieu (19, 23).
We have begun to explore the role EC-SOD plays in regulating cellular signaling processes mediated by ROS produced in response to hypoxia. Currently, the identity, source, cellular localization, target, and factors that regulate levels of ROS important for hypoxia-induced gene transcription, are not yet known. In the present study, we employ EC-SOD knockout mice and Hep3B cells overexpressing EC-SOD to demonstrate a novel role for EC-SOD in the modulation of hypoxia-induced Epo gene regulation.
Materials and Methods
Reagents
Oligonucleotides were obtained from Invitrogen Life Technologies (Carlsbad, CA). 32P-deoxy-ATP (3000 Ci/mmol) was purchased from Amersham (Arlington Heights, IL). Anti-HIF-1 (sc-13515X) and anti-HNF-4 (sc-6556X) IgG were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). All other chemicals and enzymes were from Roche Molecular Biochemicals (Indianapolis, IN), Sigma Chemical Co. (St. Louis, MO), or Invitrogen Life Technologies.
Animals
All experiments involving animals were approved by the Duke University Medical Center Institutional Animal Care and Use Committee. Homozygous EC-SOD KO mice (C57BL/6 x 129/Sv) were generated by homologous recombination (16). Control wild-type (Wt) mice (C57BL/6 x 129/Sv) were obtained from Taconic (Germantown, NY). All of the experiments in this study were performed using 6- to 8-wk-old male mice.
Immunocytochemistry
Wt and EC-SOD KO mice were euthanized with Nembutal (65 mg/kg ip) and were vascularly perfused with Hanks’ balanced solution via the left ventricle until the kidneys were thoroughly blanched. The kidneys were then perfusion fixed with 4% paraformaldehyde in PBS, excised, and then embedded in paraffin. Four to 6-μm-thick sections of kidney were treated with 0.3% hydrogen peroxide and 10% normal goat serum for 30 min to minimize endogenous peroxidase activity and nonspecific staining. The sections were then incubated at room temperature with 0.1% trypsin for 30 min and washed afterward with PBS (pH 7.6). The sections were then incubated with rabbit antimouse EC-SOD antibodies (18) at a concentration of 0.5 μg/ml in PBS-T for 1 h at room temperature. Biotinylated secondary antibody and the avidin-biotin-peroxidase complex (Vectastain kit, Burlingame, CA) were subsequently applied to the sections; diaminobenzidine tetrahydrochloride was used as peroxidase substrate.
Measurement of EC-SOD, Epo, and CA IX mRNA using RT-PCR or TaqMan
Total RNA was prepared from tissue and cultured cells using Trizol reagent (Invitrogen Life Technologies). The synthesis of single-stranded DNA from RNA was performed using SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen Life Technologies), according to the protocol provided by the manufacturer.
The PCR product of EC-SOD mRNA-derived cDNA was quantified using a dye-labeled fluorogenic oligonucleotide probe with 6-carboxy-fluoresceine (FAM) as a label and 6-carboxy-tetramethylrhodamine (TAMRA) as a quencher at the 5'- and 3'-ends, respectively. Mouse EC-SOD-specific primers used for PCR were the forward 5'-AGG TGG ATG CTG CCG AGA T-3' and reverse 5'-TCC AGA CTG AAA TAG GCC TCA AG-3' primers. The nucleotide sequence for the mouse EC-SOD detection probe was 5'-(FAM)-CAT CAG CCA CGC TGC CAC CG-(TAMRA)-3'. The mouse Epo gene was amplified using the forward primer 5'-GAG GCA GAA AAT GTC ACG ATG-3' and reverse primer 5'-CTT CCA CCT CCA TTC TTT TCC-3'. The human Epo gene was amplified using the forward primer 5'-ACC AAC ATT GCT TGT GCC AC-3' and reverse primer 5'-TCT GAA TGC TTC CTG CTC TGG-3'. The level of mRNA expression was detected using the labeled probe 5'-(FAM)-TGC AGA AGG TCC CAG ACT GAG TGA AAA TA-3'-(TAMRA) for mouse Epo and 5'-(FAM)-TCC AGT GCC AGC AAT GAC ATC TCA GG-3'-(TAMRA) for human Epo. The mRNA levels of Epo and EC-SOD were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression. Mouse GAPDH was detected using TaqMan Rodent GAPDH Control Reagent (VIC Probe) (Applied Biosystems, Foster City, CA). Human GAPDH was amplified using the forward 5'-CCA TGT TCG TCA TGG GTG TGA-3' and reverse 5'-CAT GGA CTG TGG TCA TGA GT-3' primers and detected using the labeled probe 5'-(VIC)-TCC TGC ACC ACC AAC TGC TTA GCA-3'(TAMRA).
The reaction mixture for PCR contains 5 μl of single-stranded DNA diluted 1:5 in water, 400 nM gene-specific forward primer, 400 nM gene-specific reverse primer, 200 nM gene-specific labeled probe, 40 nM GAPDH forward primer, 40 nM GAPDH reverse primer, 200 nM probe for GAPDH and 25 μl of TaqMan Universal PCR Master Mix (PE Applied Biosystems) in a total volume of 50 μl. All procedures, including data analysis, were performed on the ABI PRISM 7700 Sequence Detection System using the software provided with the instrument. The final determination of EC-SOD and Epo mRNA levels was calculated after normalization to GAPDH mRNA levels.
The human EC-SOD gene was amplified using the forward primer 5'-TCT GCC TTT GAG CTT CTC CTC TGC TC-3' and reverse primer 5'-AAG GGG GAA GAT CGT CAG GTC AAA G-3'. The human CA IX gene was amplified using forward primer 5'-TAT CTG CAC TCC TGC CCT CT-3' and reverse primer 5'-GCT GGC TTC TCA CAT TCT CC-3'. PCRs were performed using SureStart Taq Polymerase (Stratagene, La Jolla, CA) at 95 C for 5 min, and 34 cycles at 94 C for 1 min, 60 C for 1 min, 72 C for 1 min. The amplified DNA was separated on 1.5% agarose gel and visualized by staining with ethidium bromide. Human GAPDH mRNA was amplified as an internal control. Control RT-PCR experiments, in which the reverse transcriptase was omitted, showed no amplification of specific products.
Mouse hypoxic exposure
EC-SOD KO and Wt mice were exposed to a continuous flow of normobaric hypoxia (7% oxygen) or room air (normoxia) for 4 h in an air-tight cabinet. In a second set of hypoxic conditions, mice were exposed to 10% oxygen for 4 h or 1, 4, and 7 d. Immediately after exposures, the mice were euthanized, and kidneys were excised and used for the RNA isolation.
ELISA for Epo
To measure mouse Epo levels, we followed the method for measuring human Epo using an ELISA from Biomerica (Newport, CA). The mouse Epo that was used as a standard was purchased from R&D Systems (Minneapolis, MN) and reconstituted in PBS containing 0.1% BSA. Due to a lack of manufacturer provided information on the activity for mouse Epo, we assumed that 1 ng of mouse Epo corresponded to approximately 10 mU of activity. This assumption was made based on a comparison of ED50 for human and mouse Epo in their ability to stimulate proliferation of TF-1 cell line. Dilutions of mouse Epo were made in sterile PBS containing 40 mg/ml BSA. Serum samples were collected by cardiac puncture from Wt and EC-SOD KO mice at the exposure and times indicated. To obtain 400 μl of serum, which is required for duplicate assays, we pooled serum from at least three mice in each experimental group.
Nuclear extract preparation and EMSA
Nuclear extracts from Hep3B cells were prepared using a mini-extraction procedure (12). Protein concentrations were determined by a modified Bradford method (Bio-Rad, Hercules, CA) using BSA as standard. The HIF-1 and HNF-4 oligonucleotide probes used in the gel shift assays consisted of the sequences 5'-GAT CTC TGT ACG TGA CCA CAC TCA CCT C-3' and 5'-GAT CCT CAG CTT GTA CTT TGG TAC AAC TA-3', respectively. EMSA was performed in 10 μl of 20 mM HEPES (pH 7.9), 1 mM dithiothreitol, 5% glycerol, 0.2 mM EDTA, 0.05% Nonidet P-40, 32P-end-labeled oligonucleotide probe (40,000 cpm) and 100 ng poly(deoxyinosine-deoxycytosine). The reaction was started by adding 2 μg of nuclear extract. In supershift experiments, the nuclear extracts were preincubated with corresponding IgG for 30 min on ice. The free and protein-bound probes were separated on a 5% polyacrylamide gel in 0.25x TBE buffer, then dried and exposed to x-ray film.
Epo enhancer construct
An Epo enhancer-reporter construct was made by amplifying the 160-bp thymidine kinase (tk) promoter from pBLCAT2 and cloning it into the BglII and HindIII sites of pGL3-Basic vector containing the firefly luciferase reporter gene (Promega, Madison, WI), resulting in pGL3-tk plasmid. All sequence numbering for the human Epo gene are based on the nucleotide sequence derived from GenBank (accession no. M11319). The human Epo enhancer construct was created by amplification of the nucleotides 3324–3579 from human genomic DNA using corresponding primers containing a KpnI restriction site. The amplified sequence was ligated into the KpnI site immediately upstream of the thymidine kinase promoter in pGL3-tk. The forward orientation of the Epo enhancer, the number of repeats, and the sequence integrity were verified by sequencing.
Cell culture hypoxic exposures and transfections
The human liver cell line, Hep3B, was obtained from ATCC (Manassas, VA). Hep3B cells were cultured in MEM supplemented with nonessential amino acids and 10% heat-inactivated fetal bovine serum. Cells were maintained at 37 C and 5% CO2 in a humidified incubator. Wt Hep3B cells were stably transfected with the pcDNA3-EC-SOD plasmid that expresses human EC-SOD using the cytomegalovirus promoter. Permanently transfected clones expressing human EC-SOD were selected using geneticine (40 μg/ml). The isolated clones were analyzed for EC-SOD expression by RT-PCR and Western blot.
For hypoxic exposures, the cells were seeded at 90% confluence on 35-mm dishes and incubated in medium containing 50 μM CoCl2 or exposed to 1% oxygen with 5% CO2. The duration of exposure to hypoxia or cobalt chloride was 4 h. Total cellular total RNA was obtained using Trizol (Invitrogen Life Technologies).
Transfection assays were carried out using LipofectAmine reagent (Invitrogen Life Technologies). Briefly, Hep3B cells were seeded at 90% confluence on 24-well plates, incubated for 24 h, and then treated with LipofectAmine for 5 h according to the manufacturer’s protocol. After adding the complete medium, the cells were incubated overnight and then CoCl2 was added to a final concentration of 50 μM. Four hours later, cell extracts were assayed for luciferase activity using a Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions. All experiments were performed at least in triplicate, and the data were normalized to Renilla luciferase (control reporter) to control for differences in transfection efficiency.
EC-SOD expression vector
The human EC-SOD cDNA was amplified using specific primers with EcoRI restriction sites from the pBluescript-7B14 plasmid carrying human EC-SOD genomic clone (21). The amplified band was extracted from the gel, treated with EcoRI, and cloned into corresponding pcDNA3 restriction site. The orientation and integrity of EC-SOD cDNA was confirmed by sequencing.
Western blot assay
For Western blot analysis, 20 μg of total cellular or nuclear extract were separated on a 12% sodium dodecyl sulfate/polyacrylamide gel and transferred to an Immobilon-P membrane (Millipore Corp., Bedford, MA). The membrane was subsequently incubated for 1 h at room temperature with rabbit antihuman EC-SOD, diluted 1:100 in 5% nonfat milk, 0.04% Tween 20 in PBS. Horseradish peroxidase-conjugated goat-antirabbit IgG was used as a secondary antibody. The specific complexes were visualized using enhanced chemiluminescence plus kit (Amersham Biosciences, Piscataway, NJ) according to manufacturer’s instructions.
Hematocrit and reticulocytes count
After exposure to room air or hypoxia, mice were euthanized with Nembutal, and blood was collected via cardiac puncture. The blood was analyzed for hematocrit and reticulocyte count. To measure hematocrit, the blood was placed into heparinized capillary tubes (VWR International, West Chester, PA) and centrifuged 1000 x g for 30 min at 4 C. The percentage of reticulocytes was determined using FACStar Plus fluorescence-activated cell sorter and Thiazole Orange-based kit (Retic-Count) from Becton Dickinson (Franklin Lakes, NJ).
Statistical analysis
Statistical analysis was performed using t test. Values of P < 0.05 were considered significant.
Results
Immunolocalization of EC-SOD in mouse kidney
Because the kidney is the major tissue which expresses EC-SOD, we first wanted to understand its localization (24). Using a previously characterized rabbit antimouse EC-SOD peptide purified polyclonal antibody, we found positive immunostaining for EC-SOD in both the renal cortex and the outermost medulla of mouse kidney (Fig. 1A). On the basis of cell morphology, the immunostained elements were identified as proximal convoluted tubules. Most of the staining occurs in the juxtamedullary cortex and outermost medulla where more than half of the tubules revealed strong staining and the remaining subpopulation showed only light immunoreactivity or no staining. A small amount of staining was observed in glomeruli. Heavy EC-SOD staining was detected adjacent to vascular smooth muscle cells and endothelium of large renal vessels (Fig. 1C). The immunoreactivity described above was completely absent in identically prepared sections obtained from EC-SOD KO mice (Fig. 1, B and D). Immunoreactivity in Wt kidneys was also absent using antibodies preabsorbed to EC-SOD peptide (data not shown).
FIG. 1. Immunohistochemical localization of EC-SOD in mouse kidney. Kidney sections from Wt and EC-SOD KO mice were incubated with mouse EC-SOD affinity purified polyclonal IgG. Positive brown labeling is detected in the proximal convoluted tubules in the cortex and outer medulla of Wt mice (A) with no labeling observed in EC-SOD knockout mice (B). Positive brown labeling for EC-SOD is present within the cytoplasm and nucleus of proximal tubule cells. Minimal labeling was observed in glomeruli (g). Extensive labeling was observed around smooth muscle cells of renal blood vessels (bv) (C) with no labeling detected in vessels from knockout mice (D). Magnification, x400.
EC-SOD functions as a major supressor of hypoxia-induced Epo mRNA induction
Because EC-SOD is expressed at high levels in the renal cortex, also the site of hypoxia-inducible Epo expression, we decided to test the hypothesis that EC-SOD could be involved in the regulation of Epo gene expression. To determine the effect of EC-SOD on expression of Epo, we exposed Wt and EC-SOD KO mice to 4 h of either room air (normoxia) or 7% oxygen (hypoxia). Using real-time RT-PCR, we found that, after hypoxia, Wt animals increased kidney Epo mRNA levels 16-fold (Fig. 2A), which is similar to previously published studies that demonstrate an approximate 20-fold increase of Epo mRNA levels after 4 h hypoxia (25). However, EC-SOD KO mice exposed to identical hypoxic conditions, showed a striking 246-fold increase in Epo mRNA levels when compared with normoxia-treated EC-SOD KO animals. Thus, EC-SOD KO mice had a 15 times greater Epo mRNA induction after hypoxia then did Wt mice. The disruption of the EC-SOD gene did not significantly change Epo mRNA levels in the kidney under normoxic conditions. These results demonstrate that the absence of EC-SOD dramatically augments Epo induction by hypoxia. There were no changes in renal EC-SOD mRNA levels for Wt mice after exposure to hypoxia (data not shown).
FIG. 2. Effects of EC-SOD on hypoxia-induced Epo mRNA and protein levels. A, Real-time RT-PCR was used to quantitate the levels of Epo mRNA in whole kidney after 4 h of exposure to either room air or 7% oxygen in EC-SOD KO and Wt mice. The data are representative of the mean ± SE from four mice. *, P < 0.01 vs. normoxia mice; **, P < 0.01 between hypoxic Wt and EC-SOD KO mice. B, Epo levels in serum determined by ELISA. Standard curves with recombinant mouse Epo (mEpo, gray circles) or human Epo (hEpo, black squares) were obtained using the protocol described in Materials and Methods section. Data were analyzed by nonlinear regression analysis. Dashed line, Detection limit for this ELISA. Dotted line, Absorbance and relative Epo concentration in serum of EC-SOD KO mice exposed to 10% hypoxia for 1 d. The table insert shows mouse serum Epo levels in either Wt or EC-SOD KO mice exposed to 10% hypoxia for 1 d.
To determine the time course effect of hypoxia on Epo induction, we exposed Wt and EC-SOD KO mice to 10% oxygen for 4 h and 1, 4, and 7 d. The induction of Epo mRNA in Wt mice reached a maximum of 9-fold after 1 d, decreased to 3.9-fold at d 4, and then returned back to the baseline at d 7 (Table 1). Strikingly, the level of Epo mRNA in EC-SOD null mice increased 86-fold at 4 h, peaked at 1733-fold after 1 d hypoxic exposure, decreased to 7.2-fold at d 4, and then returned to normal after 7 d. The magnitude of Epo mRNA increase was dependent on the oxygen concentration. For example, 7% hypoxia exposure for 4 h caused a 16-fold increase in Epo mRNA in Wt mice and a 246-fold increase in EC-SOD KO mice, whereas a 4-h exposure to 10% oxygen resulted in an induction of only 2- and 68-fold, respectively (Fig. 2A and Table 1). The level of EC-SOD mRNA in the kidney of Wt mice remained unchanged during the course of hypoxic exposure except for a slight decrease in mRNA levels after 7 d of continuous hypoxia (Table 2), whereas no EC-SOD-specific mRNA was detected in EC-SOD KO mice, as expected.
TABLE 1. Induction of Epo mRNA in mouse kidney after whole-animal exposure to 10% oxygen for the times indicated
TABLE 2. Changes in EC-SOD mRNA levels in mouse kidney after exposure to 10% oxygen for the times indicated
Because there is an approximately 80% similarity in amino acid sequences between mouse and human Epo, and due to the absence of any commercially available ELISA for mouse Epo, we used a human Epo ELISA kit to measure mouse Epo levels. As shown in Fig. 2B, the human Epo ELISA kit has a similar detection limit and calibration curve for both mouse and human Epo. This ELISA was not sensitive enough to detect Epo protein levels in the blood of Wt or KO mice exposed to normoxia or Wt mice exposed to hypoxia. However, EC-SOD KO mice exposed to 10% hypoxia for 1 d, showed a significant increase in Epo blood levels to 52 mU/ml (Fig. 2B). Because blood levels of Epo were below the detection limits under normoxic conditions, we could not calculate the fold increase for mouse plasma Epo protein levels.
Marked increase of Epo levels in EC-SOD KO mice did not change the rate of erythropoiesis in response to hypoxia
Our data demonstrate that exposure of EC-SOD KO mice to hypoxia for 1 d significantly increases both Epo mRNA in the kidney and Epo plasma protein levels. To analyze the physiological effect of Epo induction, we measured both hematocrit and reticulocyte counts of mice exposed to hypoxia. After exposure of Wt mice to 10% oxygen for 7 d, the hematocrit increased from 52.2 ± 0.7 to 62.4 ± 2.5%. EC-SOD KO mice, however, showed a significantly lower hematocrit after 1 and 4 d of hypoxia compared with Wt mice, but by 7 d there was no significant difference (Fig. 3A). The reticulocyte counts were also elevated after hypoxia reaching a maximum at 4 d hypoxia exposure and returning back to normal by 7 d (Fig. 3B), similar to previously published data (26). There was a significant difference in reticulocyte counts between Wt and EC-SOD KO mice observed after 4 h and 1 d of hypoxia. After 4 h of hypoxia, EC-SOD KO mice showed higher reticulocyte counts compare to Wt mice, whereas 1 d of hypoxia resulted in lower reticulocyte counts. These data suggest that, although EC-SOD KO mice showed increased Epo expression in response to hypoxia, it does not necessary lead to a corresponding increase in reticulocytosis and erythropoiesis.
FIG. 3. Effect of hypoxia on hematocrit and reticulocyte count in Wt and EC-SOD KO mice. The mice were exposed to 10% oxygen for the time indicated. At the end of the exposure, the blood was collected by heart puncture and hematocrit (A) and reticulocyte count (B) determined as described in Materials and Methods. The data are representative of the mean ± SD from three mice. *, P < 0.05 vs. normoxia mice; **, P < 0.05 between Wt and EC-SOD KO mice. NS, Not significant.
EC-SOD attenuates hypoxia inducible Epo expression and HIF-1 DNA binding
To determine the mechanism by which EC-SOD markedly represses hypoxia-inducible Epo expression, we generated Hep3B clones that overexpress human EC-SOD. The extent of EC-SOD overexpression was determined by RT-PCR (data not shown) and Western blot (Fig. 4A). Two clones were selected based on the magnitude of EC-SOD expression, one clone with a relatively low expression (SOD3 low) and a second clone with high EC-SOD expression (SOD3 high). EC-SOD protein levels were measured in total cellular extract (Fig. 4A), in cell culture medium, and in nuclear extract (data not shown), all by Western blot. We found that EC-SOD protein levels were increased in all three compartments (the nucleus, cytoplasm, and extracellular medium). These results suggest that EC-SOD overexpressed in Hep3B cells is not only secreted into the extracellular space but is also localized to the nuclear fraction.
FIG. 4. Overexpression of EC-SOD attenuates induction of hypoxia-regulated genes in Hep3B cells. A, Western blot of total cellular extract from cells described above exposed for 4 h to PBS or CoCl2 (50 μM). B and C, Induction of Epo (B) and CA IX (C) mRNA levels in cells exposed to CoCl2 (50 μM) or hypoxia (1% O2, 5% CO2, 94% N2) for 4 h. The amplification and quantitation of Epo mRNA levels was performed using TaqMan and primers/probe specific for human Epo. Levels of CA IX mRNA in Hep3B cells were detected by RT-PCR. The expression of Epo and CA IX was normalized to GAPDH. The data are representative of the mean ± SD from four measurements. *, P < 0.05 vs. wt cells.
The exposure of Wt Hep3B cells to CoCl2 (50 μM) or 1% oxygen for 4 h, caused increases in Epo mRNA levels of up to 29- and 50-fold, respectively (Fig. 4B). The EC-SOD overexpressing clones (SOD3 low and SOD3 high) markedly reduced this Epo induction in a dose-dependent manner. For instance, in cells exposed to 50 μM CoCl2, low levels of EC-SOD overexpression reduced Epo induction from 28.5- to 14.6-fold and in clones expressing high levels of EC-SOD to 6.8-fold. Similarly, overexpression of EC-SOD attenuates Epo mRNA induction after exposure to 1% oxygen for 4 h from 49.8-fold in Wt Hep3B cells to 26.3 and 13.1-fold in cells expressing low and high levels EC-SOD, respectively.
To analyze the effect of EC-SOD overexpression on induction of other hypoxia-inducible genes, we measured mRNA levels of carbonic anhydrase (CA) IX. CA IX is strongly up-regulated under hypoxia conditions in HepG2 cells (27) and other tumor cells (28). We found that elevated levels of EC-SOD significantly attenuated induction of CA IX by hypoxia in a dose-dependent manner (Fig. 4C).
To better elucidate the molecular mechanism of EC-SOD’s repression of hypoxia-induced Epo expression, we constructed a hypoxia-sensitive reporter plasmid by ligating two hypoxia response elements from the Epo enhancer derived from human genomic DNA, and placed them upstream of the luciferase reporter gene expressed under the control of the tk promoter (Fig. 5A). The exposure of Wt Hep3B cells transfected with pGL3-tk-Epo(2x) to 50 μM CoCl2 for 4 h led to a 5.5-fold induction in reporter activity (Fig. 5B). As a negative control, transfection of Hep3B cells with the enhancerless pGL3-tk luciferase vector resulted in no increase in luciferase activity after hypoxic exposure. These data indicate that activation of the reporter by CoCl2 is dependent on the presence of the Epo enhancer element containing a functional binding site for HIF-1. In Hep3B cells overexpressing EC-SOD, there was an approximate 30% decrease in CoCl2-induced reporter gene expression compared with Wt cells. When Wt Hep3B cells were transfected with pGL3-tk-Epo(2x) and exposed to 1% oxygen, reporter gene activity increased approximately 8-fold relative to identical cells exposed to 21% oxygen. However, when SOD3 high Hep3B cells were used, reporter gene induction by hypoxia was significantly attenuated by 35% to 5.2-fold (Fig. 5C). These data indicate that EC-SOD attenuates hypoxia-induced Epo gene expression, at least in part, through reducing HIF-1 activity in response to hypoxia.
FIG. 5. Role of HIF-1 in EC-SOD mediated Epo induction by hypoxia. A, Schematic representation of the Epo enhancer/reporter construct. The enhancer region of human Epo gene was concatenated twice and cloned in the forward orientation upstream of tk minimal promoter. The putative HIF-1 binding sites are depicted as the black rectangles. The original reporter plasmid used for this construct was pGL3-Basic with the firefly luciferase gene as a reporter. B, Induction of the reporter by CoCl2 (50 μM) using transient transfection of Wt Hep3B cells or cells that stably overexpressing EC-SOD at high (SOD3 high) levels. C, Same as B but cells were exposed to 21% oxygen or 1% oxygen for 4 h at 37 C. After exposure the firefly luciferase activity was measured and normalized to Renilla luciferase activity produced by the cotransfected control plasmid pRL-CMV. Results shown are the mean ± SD from at least two independent transfection experiments, each performed in quadruplicate. *, P < 0.05. D, Wt Hep3B cells and clone overexpressing EC-SOD at high levels (SOD3 high) were exposed to PBS or 50 μM CoCl2 for 4 h at 37 C. Nuclear extracts were incubated with HIF-1-specific oligonucleotide and separated on a polyacrylamide gel as described in Materials and Methods. For supershift experiments nonimmune (n.i.) or HIF-1 (anti-HIF-1)-specific antibodies were added to the incubation mixtures. Arrows on the right side show HIF-1 and supershifted HIF-1-specific bands. All binding reactions were separated on 5% nondenaturing polyacrylamide gel, dried, and exposed to the x-ray film.
Because hypoxia-mediated induction of Epo is mostly regulated by the HIF-1 transcription factor, we decided to analyze whether EC-SOD overexpression affects HIF-1 DNA binding. Using EMSA and a HIF-1-specific oligonucleotide, we detected a significant increase in the intensity of the HIF-1-specific band in Wt Hep3B cells after CoCl2 exposure (Fig. 5D, compare lane 1 and 2). A similar exposure using SOD3 high Hep3B cells showed that CoCl2-treated cells also increased HIF-1 levels, but to a lesser extent compared with Wt cells. (Fig. 5D, lanes 3 and 4). The specificity of this assay was demonstrated by supershifting the DNA-protein complex with anti-HIF-1 antibodies (Fig. 5D, lanes 5–8). Although nonimmune IgG did not change the electrophoretic mobility of shifted bands (Fig. 5D, lanes 9 and 10), incubation with anti-HIF-1 IgG resulted in the appearance of new slower-migrating band (see arrow on right side of Fig. 5D, lanes 6 and 8). The intensity of this HIF-1 supershifted band was slightly decreased in cells overexpressing EC-SOD (compare lanes 6 and 8 in Fig. 5D). These data support our transient transfection reporter gene experiments and indicate that EC-SOD overexpression, at least in part, modulates hypoxia-mediated changes in HIF-1 levels. Additionally, we analyzed HNF-4 binding activity using EMSA and supershift assay but did not detect any differences between either PBS vs. CoCl2-treated cells or Wt vs. EC-SOD overexpressing cells (data not shown).
Discussion
The presence of high levels of EC-SOD mRNA and protein in the kidney, (24) and EC-SOD’s distinctive localization to specific proximal tubular epithelial cells, suggest a potentially significant role of this antioxidant enzyme in the normal and, perhaps, pathological kidney. Interestingly, EC-SOD appears to have a similar pattern of expression as does that of Epo. For example, in the kidney, Epo is primarily synthesized in the renal cortex with no detectable expression in the medulla. There are contradictory data about which renal cells produce this hormone. Peritubular interstitial cells have been identified as the major source of Epo production (29, 30). However, in contrast to these studies, other groups have detected Epo production primarily by proximal tubular epithelial cells using in situ hybridization (31), immunochemistry (32), and transgenic mice (33). The apparent close proximity of sites of EC-SOD and Epo expression suggest the possible importance of EC-SOD in the regulation of Epo expression. Indeed, we show that mice lacking the EC-SOD gene have a 193-fold higher Epo induction, in response to hypoxia, compared with Wt mice.
The molecular mechanisms mediating this profound effect of EC-SOD on hypoxia-induced Epo expression is not known, but several hypothesis can be considered. Because EC-SOD is the major extracellular antioxidant enzyme that reduces extracellular levels of superoxide, we infer that ROS serve as signaling molecule that regulate hypoxia-induced gene expression in renal proximal tubules. If ROS concentrations increase under hypoxic conditions due to malfunctioning of mitochondrial electron transport chain, EC-SOD KO mice would have significantly higher levels of ROS after hypoxic exposure. Because hypoxia generated superoxide radicals are required for induction of HIF-1 activity and other downstream target genes (7, 8), we can predict that disruption of EC-SOD function will further increase kidney ROS levels that, in turn, will increase HIF-1 levels to a greater extent compared with Wt mice. However, the role ROS play in mediating HIF-1 stabilization remains controversial because recent studies show that the absence of a functional mitochondrial electron transport chain does not interfere with the regulation of HIF-1 and HIF-1 target genes (9).
The role of ROS in hypoxia-mediated gene induction should be considered in relation with another important messenger nitric oxide (NO). It has been shown that NO inhibits hypoxia-inducible transcription of Epo gene through suppression of HIF-1 expression, DNA binding activity, and transcriptional activity (34, 35, 36). The mechanism by which NO exerts its effects on HIF-1 stability are not clearly understood but seem to involve the intracellular S-nitrosylation of cysteine residues in the ODD domain or in critical thiols of ubiquitin-activating enzymes (37). Because superoxide rapidly reacts with NO to form peroxynitrite, elevated levels of superoxide would be predicted to decrease physiologic NO concentrations (38). Furthermore, EC-SOD KO mice have an impaired ability to dismutate superoxide into hydrogen peroxide that could lead to decreases in H2O2 levels. Because NO and H2O2 are known to block the induction of Epo and accumulation of HIF-1 (39), EC-SOD KO mice would be predicted to enhance hypoxia-induced Epo gene expression (Fig. 6).
FIG. 6. Schematic illustration depicting the possible mechanims by which EC-SOD regulates hypoxia-inducible Epo gene transcription. Hypoxia-mediated Epo gene expression is mediated primarily by HIF-1. HIF-1 is a heterodimeric protein composed of two discreet subunits, HIF-1 and ARNT. Under normoxia conditions, HIF-1 is hydroxylated by a prolyl hydroxylases and subsequently ubiquitinated and completely degraded. After hypoxia, an oxygen sensor protein, presumed to be a nicotinamide adenine dinucleotide phosphate oxidase and illustrated here as Renox, signals, by an as yet unknown pathway, leading to the stabilization and accumulation of HIF-1. HIF-1/ARNT heteroduplex then translocates into the nucleus enhancing gene transcription via the hypoxia response element (HRE). Under these conditions, both NO and H2O2 have both been shown to completely inhibit and attenuate this hypoxia-induced gene transcription response. We hypothesized that in tissues lacking EC-SOD expression, levels of superoxide will be markedly increased, which would lead to decreased levels of NO, which would further augment Epo gene expression. Moreover, without EC-SOD, the production of hydrogen peroxide in the extracellular space would be predicted to be diminished, which would additionally remove its repressor activity on hypoxia-inducible Epo gene expression. For a more detailed description, see Discussion.
In addition to the above possible mechanisms, there is small chance that the significant differences in hypoxia-mediated Epo induction seen in Wt and EC-SOD KO mice can be attributed to subtle genetic differences between the two strains of animals used. In our experiments, we have used C57BL/6 x 129/Sv mice purchased from Taconic (as controls) and EC-SOD KO mice that were maintained in the original strain, also of C57BL/6 x 129/Sv, as described previously (16). Although the animals are all of the same background strain, there is the potential due to breeding effects for the animals to have divergent response. However, the experiments performed with Hep3B cells overexpressing EC-SOD provided additional support that EC-SOD does play a critical role in the regulation of Epo expression and that the differences observed in our in vivo model are unlikely to be due to that of genetic variations.
The physiological relevance of such a remarkable up-regulation of Epo expression in EC-SOD KO mice is not clear at this time. Our data suggest that increasing Epo concentrations in the blood of EC-SOD KO mice in response to hypoxia does not cause an expected significant increase in erythropoiesis compared with Wt mice. These results suggest that the nascent Epo being produced in the EC-SOD KO mice may show reduced or lack hematopoietic activities. Recent studies have demonstrated that carbamylated Epo and certain Epo mutants do not bind the classical Epo receptor, and hence did not show any hematopoietic activity (40). Nevertheless, these Epo variants were able to confer cytoprotective benefits despite nonhematopoietic effects. Several other studies indicate that Epo’s physiologic effects are not restricted to just the stimulation of red blood cells production, but seem to be quite pleitropic. For instance, it has been shown that increasing Epo levels protect neurons from ischemic stroke (41, 42) as well as prevent light-induced retinal degradation (43). Preconditioning cardiomyocytes with Epo in vivo as well as in vitro protect them from ischemic injury. Moreover, a single dose of Epo delivered at the time of cardiac reperfusion had a profound beneficial therapeutic effect on the ischemic heart but did not change hematocrit (44). Thus, we speculate that the increased Epo being produced in EC-SOD KO mice to have poor activity with the classical Epo receptor, perhaps due to oxidative modifications that are a direct result of EC-SOD’s absence in the kidney. Whether this enhanced Epo production retains cytoprotoective properties remains to be studied.
In summary, we show that EC-SOD is primarily immunolocalized in proximal tubule cells in the kidney and dramatically represses the hypoxia-inducible Epo gene induction. In cell culture studies, overexpression of EC-SOD decreased hypoxia-induced Epo gene expression in vivo and in vitro in a concentration-dependent fashion. The mechanism for EC-SOD’s profound effect on hypoxia-induced Epo gene expression is dependent, at least in part, on stabilization of HIF-1. Epo produced under hypoxic conditions in EC-SOD KO mice appears to have lost hematopoietic activity. Overall, our results implicate superoxide as an important component of the signal transduction pathway that leads to amplification of hypoxia-inducible Epo gene expression.
Acknowledgments
We thank Dr. Murat Arcasoy for critical review of this manuscript. We thank members of the Duke University Cell Culture Facility for help in obtaining and culturing of mammalian cells.
References
Zhu H, Jackson T, Bunn HF 2002 Detecting and responding to hypoxia. Nephrol Dial Transplant 17(Suppl 1):3–7
Michiels C, Minet E, Mottet D, Raes M 2002 Regulation of gene expression by oxygen: NF-B and HIF-1, two extremes. Free Radic Biol Med 33:1231–1242
Huang LE, Bunn HF 2003 Hypoxia-inducible factor and its biomedical relevance. J Biol Chem 278:19575–19578
Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell SJ, Kriegsheim AV, Hebestreit HF, Mukherji M, Schofield CJ, Maxwell PH, Pugh CW, Ratcliffe PJ 2001 Targeting of HIF- to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 292:468–472
Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, Salic A, Asara JM, Lane WS, Kaelin Jr WG 2001 HIF targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 292:464–468
Epstein AC, Gleadle JM, McNeill LA, Hewitson KS, O’Rourke J, Mole DR, Mukherji M, Metzen E, Wilson MI, Dhanda A, Tian YM, Masson N, Hamilton DL, Jaakkola P, Barstead R, Hodgkin J, Maxwell PH, Pugh CW, Schofield CJ, Ratcliffe PJ 2001 C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell 107:43–54
Chandel NS, Maltepe E, Goldwasser E, Mathieu CE, Simon MC, Schumacker PT 1998 Mitochondrial reactive oxygen species trigger hypoxia-induced transcription. Proc Natl Acad Sci USA 95:11715–11720
Agani FH, Pichiule P, Chavez JC, LaManna JC 2000 The role of mitochondria in the regulation of hypoxia-inducible factor 1 expression during hypoxia. J Biol Chem 275:35863–35867
Vaux EC, Metzen E, Yeates KM, Ratcliffe PJ 2001 Regulation of hypoxia-inducible factor is preserved in the absence of a functioning mitochondrial respiratory chain. Blood 98:296–302
Srinivas V, Leshchinsky I, Sang N, King MP, Minchenko A, Caro J 2001 Oxygen sensing and HIF-1 activation does not require an active mitochondrial respiratory chain electron-transfer pathway. J Biol Chem 276:21995–21998
Ehleben W, Porwol T, Fandrey J, Kummer W, Acker H 1997 Cobalt and desferrioxamine reveal crucial members of the oxygen sensing pathway in HepG2 cells. Kidney Int 51:483–491
Fandrey J, Frede S, Ehleben W, Porwol T, Acker H, Jelkmann W 1997 Cobalt chloride and desferrioxamine antagonize the inhibition of erythropoietin production by reactive oxygen species. Kidney Int 51:492–496
Fridovich I 1997 Superoxide anion radical (O2–.), superoxide dismutases, and related matters. J Biol Chem 272:18515–18517
Zelko IN, Mariani TJ, Folz RJ 2002 Superoxide dismutase multigene family: a comparison of the CuZn-SOD (SOD1), Mn-SOD (SOD2), and EC-SOD (SOD3) gene structures, evolution, and expression. Free Radic Biol Med 33:337–349
Marklund SL 1984 Extracellular superoxide dismutase and other superoxide dismutase isoenzymes in tissues from nine mammalian species. Biochem J 222:649–655
Carlsson LM, Jonsson J, Edlund T, Marklund SL 1995 Mice lacking extracellular superoxide dismutase are more sensitive to hyperoxia. Proc Natl Acad Sci USA 92:6264–6268
Suliman HB, Ryan LK, Bishop L, Folz RJ 2001 Prevention of influenza-induced lung injury in mice overexpressing extracellular superoxide dismutase. Am J Physiol Lung Cell Mol Physiol 280:L69–L78
Folz RJ, Abushamaa AM, Suliman HB 1999 Extracellular superoxide dismutase in the airways of transgenic mice reduces inflammation and attenuates lung toxicity following hyperoxia. J Clin Invest 103:1055–1066
Marklund SL 1982 Human copper-containing superoxide dismutase of high molecular weight. Proc Natl Acad Sci USA 79:7634–7638
Willems J, Zwijsen A, Slegers H, Nicolai S, Bettadapura J, Raymackers J, Scarcez T 1993 Purification and sequence of rat extracellular superoxide dismutase B secreted by C6 glioma. J Biol Chem 268:24614–24621
Folz RJ, Crapo JD 1994 Extracellular superoxide dismutase (SOD3): tissue-specific expression, genomic characterization, and computer-assisted sequence analysis of the human EC SOD gene. Genomics 22:162–171
Ookawara T, Imazeki N, Matsubara O, Kizaki T, Oh-Ishi S, Nakao C, Sato Y, Ohno H 1998 Tissue distribution of immunoreactive mouse extracellular superoxide dismutase. Am J Physiol 275:C840–C847
Oury TD, Day BJ, Crapo JD 1996 Extracellular superoxide dismutase in vessels and airways of humans and baboons. Free Radic Biol Med 20:957–965
Folz RJ, Guan J, Seldin MF, Oury TD, Enghild JJ, Crapo JD 1997 Mouse extracellular superoxide dismutase: primary structure, tissue-specific gene expression, chromosomal localization, and lung in situ hybridization. Am J Respir Cell Mol Biol 17:393–403
Chikuma M, Masuda S, Kobayashi T, Nagao M, Sasaki R 2000 Tissue-specific regulation of erythropoietin production in the murine kidney, brain, and uterus. Am J Physiol Endocrinol Metab 279:E1242–E1248
Mylrea KC, Abbrecht PH 1970 Hematologic responses of mice subjected to continuous hypoxia. Am J Physiol 218:1145–1149
Sonna LA, Cullivan ML, Sheldon HK, Pratt RE, Lilly CM 2003 Effect of hypoxia on gene expression by human hepatocytes (HepG2). Physiol Genom 12:195–207
Wykoff CC, Beasley NJ, Watson PH, Turner KJ, Pastorek J, Sibtain A, Wilson GD, Turley H, Talks KL, Maxwell PH, Pugh CW, Ratcliffe PJ, Harris AL 2000 Hypoxia-inducible expression of tumor-associated carbonic anhydrases. Cancer Res 60:7075–7083
Koury ST, Bondurant MC, Koury MJ 1988 Localization of erythropoietin synthesizing cells in murine kidneys by in situ hybridization. Blood 71:524–527
Lacombe C, Da Silva JL, Bruneval P, Fournier JG, Wendling F, Casadevall N, Camilleri JP, Bariety J, Varet B, Tambourin P 1988 Peritubular cells are the site of erythropoietin synthesis in the murine hypoxic kidney. J Clin Invest 81:620–623
Shanks JH, Hill CM, Lappin TR, Maxwell AP 1996 Localization of erythropoietin gene expression in proximal renal tubular cells detected by digoxigenin-labelled oligonucleotide probes. J Pathol 179:283–287
Maxwell AP, Lappin TR, Johnston CF, Bridges JM, McGeown MG 1990 Erythropoietin production in kidney tubular cells. Br J Haematol 74:535–539
Loya F, Yang Y, Lin H, Goldwasser E, Albitar M 1994 Transgenic mice carrying the erythropoietin gene promoter linked to lacZ express the reporter in proximal convoluted tubule cells after hypoxia. Blood 84:1831–1836
Liu Y, Christou H, Morita T, Laughner E, Semenza GL, 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
Huang LE, Willmore WG, Gu J, Goldberg MA, Bunn HF 1999 Inhibition of hypoxia-inducible factor 1 activation by carbon monoxide and nitric oxide. Implications for oxygen sensing and signaling. J Biol Chem 274:9038–9044
Sogawa K, Numayama-Tsuruta K, Ema M, Abe M, Abe H, Fujii-Kuriyama Y 1998 Inhibition of hypoxia-inducible factor 1 activity by nitric oxide donors in hypoxia. Proc Natl Acad Sci USA 95:7368–7373
Palmer LA, Gaston B, Johns RA 2000 Normoxic stabilization of hypoxia-inducible factor-1 expression and activity: redox-dependent effect of nitrogen oxides. Mol Pharmacol 58:1197–1203
Ortega Mateo A, Amaya Aleixandre de Artinano 2000 Nitric oxide reactivity and mechanisms involved in its biological effects. Pharmacol Res 42:421–427
Huang LE, Arany Z, Livingston DM, Bunn HF 1996 Activation of hypoxia-inducible transcription factor depends primarily upon redox-sensitive stabilization of its subunit. J Biol Chem 271:32253–32259
Leist M, Ghezzi P, Grasso G, Bianchi R, Villa P, Fratelli M, Savino C, Bianchi M, Nielsen J, Gerwien J, Kallunki P, Larsen AK, Helboe L, Christensen S, Pedersen LO, Nielsen M, Torup L, Sager T, Sfacteria A, Erbayraktar S, Erbayraktar Z, Gokmen N, Yilmaz O, Cerami-Hand C, Xie QW, Coleman T, Cerami A, Brines M 2004 Derivatives of erythropoietin that are tissue protective but not erythropoietic. Science 305:239–242
Brines ML, Ghezzi P, Keenan S, Agnello D, de Lanerolle NC, Cerami C, Itri LM, Cerami A 2000 Erythropoietin crosses the blood-brain barrier to protect against experimental brain injury. Proc Natl Acad Sci USA 97:10526–10531
Siren AL, Fratelli M, Brines M, Goemans C, Casagrande S, Lewczuk P, Keenan S, Gleiter C, Pasquali C, Capobianco A, Mennini T, Heumann R, Cerami A, Ehrenreich H, Ghezzi P 2001 Erythropoietin prevents neuronal apoptosis after cerebral ischemia and metabolic stress. Proc Natl Acad Sci USA 98:4044–4049
Grimm C, Wenzel A, Groszer M, Mayser H, Seeliger M, Samardzija M, Bauer C, Gassmann M, Reme CE 2002 HIF-1-induced erythropoietin in the hypoxic retina protects against light-induced retinal degeneration. Nat Med 8:718–724
Parsa CJ, Matsumoto A, Kim J, Riel RU, Pascal LS, Walton GB, Thompson RB, Petrofski JA, Annex BH, Stamler JS, Koch WJ 2003 A novel protective effect of erythropoietin in the infarcted heart. J Clin Invest 112:999–1007(Igor N. Zelko and Rodney )
Address all correspondence and requests for reprints to: Rodney J. Folz, M.D., Ph.D., Division of Pulmonary, Allergy, and Critical Care, Department of Medicine, Duke University Medical Center, Box 2620/Room 331 MSRB, Durham, North Carolina 27710. E-mail: rodney.folz@duke.edu.
Abstract
Hypoxia and biological responses to hypoxia are commonly encountered in both normal and pathologic cellular processes. Here we report that extracellular superoxide dismutase (EC-SOD) plays a major role in regulating the magnitude of hypoxia-induced erythropoietin (Epo) gene expression, thus implicating superoxide as an intermediary signal transduction molecule critical to this process. We found that mice which have the EC-SOD gene inactivated show a marked more than 100-fold elevation in hypoxia-induced Epo gene expression, compared with wild-type controls, which was both dose and time dependent. These mice also showed a significant increase in serum Epo levels after 1 d hypoxia. Interestingly, despite elevated Epo levels, reciprocal changes in hematocrit and reticulocyte counts were not found, suggesting that this newly synthesized Epo lacks functional hematopoietic effects. When EC-SOD was overexpressed in Hep3B cells, we found a significant reduction in Epo gene induction by both CoCl2 (50 μM) and hypoxia (1% O2). Similar findings were noted with another hypoxia-inducible gene, carbonic anhydrase IX. We conclude that EC-SOD functions as a major repressor of hypoxia-induced Epo gene expression, which implicates superoxide as a signaling intermediate whose downstream effects, at least in part, may be mediated by HIF-1.
Introduction
ADAPTATION OF VERTEBRATES to the oxygen-rich environment require regulatory mechanisms that allow them to respond to a changing oxygen concentration. When mammalian organisms are exposed to a hypoxic environment, the red blood cell count increases significantly to boost oxygen delivery to the hypoxic tissues. Erythropoietin (Epo) is the major hormonal signal involved in this hypoxia-induced response (1). Epo is synthesized primarily in kidney, and secondarily in liver, where it is released into the circulation and binds to its major target, the Epo receptor. The interaction between ligand and receptor stimulates production of red blood cells in the bone marrow. The regulation of Epo expression by hypoxia is quite complex and involves first stabilization/activation of hypoxia-inducible factor 1 (HIF-1), which then translocates into the nucleus, where it dimerizes with the arylhydrocarbon receptor nuclear translocator (HIF-1?) and binds the hypoxia-responsive enhancer located in the 3'-region of the Epo gene (for review see Refs. 2 and 3). The molecular mechanisms involved in oxygen-dependent stabilization of HIF-1 has recently been elucidated in great detail. Briefly, under normoxic conditions, HIF-1 protein is hydroxylated at its prolyl residues located in the oxygen-dependent domain (ODD) by a family of prolyl hydroxylase enzymes (4, 5). Hydroxylation of HIF-1 initiates binding of the Von Hipple-Lindau protein that acts as an E3 ubiquitin ligase that then promotes the degradation of HIF-1 by proteosomes. The activity of these prolyl hydroxylases inherently depends on oxygen but also require 2-oxoglutarate, ascorbic acid, and iron. On the other hand, prolyl hydroxylase activities are inhibited by cobaltous ions and by iron chelation, explaining why HIF-1 is activated by transition metals and by the iron-chelating agent desferrioxamine (6).
Although this model of HIF-1 activation explains induction of Epo by hypoxia and cobalt chloride, it fails to explain the role reactive oxygen species (ROS) play in this process. The role of superoxide radicals in hypoxia-mediated transduction pathways is not clearly defined and sometimes controversial. For example, it has been proposed that levels of ROS increase under hypoxic conditions due to malfunctioning of the mitochondrial electron transport chain. The role of mitochondria-generated ROS in the regulation of HIF-1 activity is not clear, and several studies have reported conflicting and confusing results. For example, it has been shown that mitochondria generated superoxide radicals, and their reaction product hydrogen peroxide, are required for induction of HIF-1 activity and other subsequent downstream target genes in hypoxic tissues (7, 8). On the other hand, using similar pharmacological and genetic approaches, others have provided compelling evidence that the mitochondrial respiratory chain is not essential for the regulation of HIF-1 activity by oxygen (9, 10). Furthermore, an alternative model for HIF-1 regulation has been proposed. In this model, hypoxia leads to a decreased production of ROS due to inhibition of nicotinamide adenine dinucleotide phosphate oxidase activity. The fast diffusing ROS oxidize and destabilize HIF-1 protein resulting in a decreased expression of hypoxia-inducible genes, including Epo. The decrease in superoxide production after hypoxia leads to HIF-1 stabilization and activation of gene expression (11, 12).
A major class of antioxidant enzymes involved in maintaining homeostatic levels of ROS, such as superoxide, are represented by the superoxide dismutases (SODs). SODs are a family of metalloenzymes that catalyze the dismutation of O2– to H2O2. Mammalian cells possess three distinct forms of SODs; extracellular SOD (EC-SOD) present in the extracellular spaces, manganese SOD, found exclusively in the mitochondria, and copper-zinc SOD located in the cytoplasm (13, 14). EC-SOD is the principal enzymatic antioxidant in extracellular spaces (15) and plays an important role in the protection of mammalian organisms against superoxide. Genetic disruption of EC-SOD decreases survival time of mice exposed to high oxygen tension (16). Overexpression of EC-SOD in transgenic mice preserves lung function and reduces inflammation caused by influenza pneumonia (17) and hyperoxia (18). The human and mouse EC-SOD is a homotetrameric protein with molecular mass of 135 kDa, whereas the rat enzyme is dimeric (19, 20). Expression levels of EC-SOD differs dramatically among tissues, but the highest expression is found in kidney, lung, and heart (21, 22). EC-SOD is glycosylated and exhibits affinity for heparin or heparan sulfate. Thus, although detectable in blood plasma, most of the enzyme exists bound to the extracellular matrix milieu (19, 23).
We have begun to explore the role EC-SOD plays in regulating cellular signaling processes mediated by ROS produced in response to hypoxia. Currently, the identity, source, cellular localization, target, and factors that regulate levels of ROS important for hypoxia-induced gene transcription, are not yet known. In the present study, we employ EC-SOD knockout mice and Hep3B cells overexpressing EC-SOD to demonstrate a novel role for EC-SOD in the modulation of hypoxia-induced Epo gene regulation.
Materials and Methods
Reagents
Oligonucleotides were obtained from Invitrogen Life Technologies (Carlsbad, CA). 32P-deoxy-ATP (3000 Ci/mmol) was purchased from Amersham (Arlington Heights, IL). Anti-HIF-1 (sc-13515X) and anti-HNF-4 (sc-6556X) IgG were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). All other chemicals and enzymes were from Roche Molecular Biochemicals (Indianapolis, IN), Sigma Chemical Co. (St. Louis, MO), or Invitrogen Life Technologies.
Animals
All experiments involving animals were approved by the Duke University Medical Center Institutional Animal Care and Use Committee. Homozygous EC-SOD KO mice (C57BL/6 x 129/Sv) were generated by homologous recombination (16). Control wild-type (Wt) mice (C57BL/6 x 129/Sv) were obtained from Taconic (Germantown, NY). All of the experiments in this study were performed using 6- to 8-wk-old male mice.
Immunocytochemistry
Wt and EC-SOD KO mice were euthanized with Nembutal (65 mg/kg ip) and were vascularly perfused with Hanks’ balanced solution via the left ventricle until the kidneys were thoroughly blanched. The kidneys were then perfusion fixed with 4% paraformaldehyde in PBS, excised, and then embedded in paraffin. Four to 6-μm-thick sections of kidney were treated with 0.3% hydrogen peroxide and 10% normal goat serum for 30 min to minimize endogenous peroxidase activity and nonspecific staining. The sections were then incubated at room temperature with 0.1% trypsin for 30 min and washed afterward with PBS (pH 7.6). The sections were then incubated with rabbit antimouse EC-SOD antibodies (18) at a concentration of 0.5 μg/ml in PBS-T for 1 h at room temperature. Biotinylated secondary antibody and the avidin-biotin-peroxidase complex (Vectastain kit, Burlingame, CA) were subsequently applied to the sections; diaminobenzidine tetrahydrochloride was used as peroxidase substrate.
Measurement of EC-SOD, Epo, and CA IX mRNA using RT-PCR or TaqMan
Total RNA was prepared from tissue and cultured cells using Trizol reagent (Invitrogen Life Technologies). The synthesis of single-stranded DNA from RNA was performed using SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen Life Technologies), according to the protocol provided by the manufacturer.
The PCR product of EC-SOD mRNA-derived cDNA was quantified using a dye-labeled fluorogenic oligonucleotide probe with 6-carboxy-fluoresceine (FAM) as a label and 6-carboxy-tetramethylrhodamine (TAMRA) as a quencher at the 5'- and 3'-ends, respectively. Mouse EC-SOD-specific primers used for PCR were the forward 5'-AGG TGG ATG CTG CCG AGA T-3' and reverse 5'-TCC AGA CTG AAA TAG GCC TCA AG-3' primers. The nucleotide sequence for the mouse EC-SOD detection probe was 5'-(FAM)-CAT CAG CCA CGC TGC CAC CG-(TAMRA)-3'. The mouse Epo gene was amplified using the forward primer 5'-GAG GCA GAA AAT GTC ACG ATG-3' and reverse primer 5'-CTT CCA CCT CCA TTC TTT TCC-3'. The human Epo gene was amplified using the forward primer 5'-ACC AAC ATT GCT TGT GCC AC-3' and reverse primer 5'-TCT GAA TGC TTC CTG CTC TGG-3'. The level of mRNA expression was detected using the labeled probe 5'-(FAM)-TGC AGA AGG TCC CAG ACT GAG TGA AAA TA-3'-(TAMRA) for mouse Epo and 5'-(FAM)-TCC AGT GCC AGC AAT GAC ATC TCA GG-3'-(TAMRA) for human Epo. The mRNA levels of Epo and EC-SOD were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression. Mouse GAPDH was detected using TaqMan Rodent GAPDH Control Reagent (VIC Probe) (Applied Biosystems, Foster City, CA). Human GAPDH was amplified using the forward 5'-CCA TGT TCG TCA TGG GTG TGA-3' and reverse 5'-CAT GGA CTG TGG TCA TGA GT-3' primers and detected using the labeled probe 5'-(VIC)-TCC TGC ACC ACC AAC TGC TTA GCA-3'(TAMRA).
The reaction mixture for PCR contains 5 μl of single-stranded DNA diluted 1:5 in water, 400 nM gene-specific forward primer, 400 nM gene-specific reverse primer, 200 nM gene-specific labeled probe, 40 nM GAPDH forward primer, 40 nM GAPDH reverse primer, 200 nM probe for GAPDH and 25 μl of TaqMan Universal PCR Master Mix (PE Applied Biosystems) in a total volume of 50 μl. All procedures, including data analysis, were performed on the ABI PRISM 7700 Sequence Detection System using the software provided with the instrument. The final determination of EC-SOD and Epo mRNA levels was calculated after normalization to GAPDH mRNA levels.
The human EC-SOD gene was amplified using the forward primer 5'-TCT GCC TTT GAG CTT CTC CTC TGC TC-3' and reverse primer 5'-AAG GGG GAA GAT CGT CAG GTC AAA G-3'. The human CA IX gene was amplified using forward primer 5'-TAT CTG CAC TCC TGC CCT CT-3' and reverse primer 5'-GCT GGC TTC TCA CAT TCT CC-3'. PCRs were performed using SureStart Taq Polymerase (Stratagene, La Jolla, CA) at 95 C for 5 min, and 34 cycles at 94 C for 1 min, 60 C for 1 min, 72 C for 1 min. The amplified DNA was separated on 1.5% agarose gel and visualized by staining with ethidium bromide. Human GAPDH mRNA was amplified as an internal control. Control RT-PCR experiments, in which the reverse transcriptase was omitted, showed no amplification of specific products.
Mouse hypoxic exposure
EC-SOD KO and Wt mice were exposed to a continuous flow of normobaric hypoxia (7% oxygen) or room air (normoxia) for 4 h in an air-tight cabinet. In a second set of hypoxic conditions, mice were exposed to 10% oxygen for 4 h or 1, 4, and 7 d. Immediately after exposures, the mice were euthanized, and kidneys were excised and used for the RNA isolation.
ELISA for Epo
To measure mouse Epo levels, we followed the method for measuring human Epo using an ELISA from Biomerica (Newport, CA). The mouse Epo that was used as a standard was purchased from R&D Systems (Minneapolis, MN) and reconstituted in PBS containing 0.1% BSA. Due to a lack of manufacturer provided information on the activity for mouse Epo, we assumed that 1 ng of mouse Epo corresponded to approximately 10 mU of activity. This assumption was made based on a comparison of ED50 for human and mouse Epo in their ability to stimulate proliferation of TF-1 cell line. Dilutions of mouse Epo were made in sterile PBS containing 40 mg/ml BSA. Serum samples were collected by cardiac puncture from Wt and EC-SOD KO mice at the exposure and times indicated. To obtain 400 μl of serum, which is required for duplicate assays, we pooled serum from at least three mice in each experimental group.
Nuclear extract preparation and EMSA
Nuclear extracts from Hep3B cells were prepared using a mini-extraction procedure (12). Protein concentrations were determined by a modified Bradford method (Bio-Rad, Hercules, CA) using BSA as standard. The HIF-1 and HNF-4 oligonucleotide probes used in the gel shift assays consisted of the sequences 5'-GAT CTC TGT ACG TGA CCA CAC TCA CCT C-3' and 5'-GAT CCT CAG CTT GTA CTT TGG TAC AAC TA-3', respectively. EMSA was performed in 10 μl of 20 mM HEPES (pH 7.9), 1 mM dithiothreitol, 5% glycerol, 0.2 mM EDTA, 0.05% Nonidet P-40, 32P-end-labeled oligonucleotide probe (40,000 cpm) and 100 ng poly(deoxyinosine-deoxycytosine). The reaction was started by adding 2 μg of nuclear extract. In supershift experiments, the nuclear extracts were preincubated with corresponding IgG for 30 min on ice. The free and protein-bound probes were separated on a 5% polyacrylamide gel in 0.25x TBE buffer, then dried and exposed to x-ray film.
Epo enhancer construct
An Epo enhancer-reporter construct was made by amplifying the 160-bp thymidine kinase (tk) promoter from pBLCAT2 and cloning it into the BglII and HindIII sites of pGL3-Basic vector containing the firefly luciferase reporter gene (Promega, Madison, WI), resulting in pGL3-tk plasmid. All sequence numbering for the human Epo gene are based on the nucleotide sequence derived from GenBank (accession no. M11319). The human Epo enhancer construct was created by amplification of the nucleotides 3324–3579 from human genomic DNA using corresponding primers containing a KpnI restriction site. The amplified sequence was ligated into the KpnI site immediately upstream of the thymidine kinase promoter in pGL3-tk. The forward orientation of the Epo enhancer, the number of repeats, and the sequence integrity were verified by sequencing.
Cell culture hypoxic exposures and transfections
The human liver cell line, Hep3B, was obtained from ATCC (Manassas, VA). Hep3B cells were cultured in MEM supplemented with nonessential amino acids and 10% heat-inactivated fetal bovine serum. Cells were maintained at 37 C and 5% CO2 in a humidified incubator. Wt Hep3B cells were stably transfected with the pcDNA3-EC-SOD plasmid that expresses human EC-SOD using the cytomegalovirus promoter. Permanently transfected clones expressing human EC-SOD were selected using geneticine (40 μg/ml). The isolated clones were analyzed for EC-SOD expression by RT-PCR and Western blot.
For hypoxic exposures, the cells were seeded at 90% confluence on 35-mm dishes and incubated in medium containing 50 μM CoCl2 or exposed to 1% oxygen with 5% CO2. The duration of exposure to hypoxia or cobalt chloride was 4 h. Total cellular total RNA was obtained using Trizol (Invitrogen Life Technologies).
Transfection assays were carried out using LipofectAmine reagent (Invitrogen Life Technologies). Briefly, Hep3B cells were seeded at 90% confluence on 24-well plates, incubated for 24 h, and then treated with LipofectAmine for 5 h according to the manufacturer’s protocol. After adding the complete medium, the cells were incubated overnight and then CoCl2 was added to a final concentration of 50 μM. Four hours later, cell extracts were assayed for luciferase activity using a Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions. All experiments were performed at least in triplicate, and the data were normalized to Renilla luciferase (control reporter) to control for differences in transfection efficiency.
EC-SOD expression vector
The human EC-SOD cDNA was amplified using specific primers with EcoRI restriction sites from the pBluescript-7B14 plasmid carrying human EC-SOD genomic clone (21). The amplified band was extracted from the gel, treated with EcoRI, and cloned into corresponding pcDNA3 restriction site. The orientation and integrity of EC-SOD cDNA was confirmed by sequencing.
Western blot assay
For Western blot analysis, 20 μg of total cellular or nuclear extract were separated on a 12% sodium dodecyl sulfate/polyacrylamide gel and transferred to an Immobilon-P membrane (Millipore Corp., Bedford, MA). The membrane was subsequently incubated for 1 h at room temperature with rabbit antihuman EC-SOD, diluted 1:100 in 5% nonfat milk, 0.04% Tween 20 in PBS. Horseradish peroxidase-conjugated goat-antirabbit IgG was used as a secondary antibody. The specific complexes were visualized using enhanced chemiluminescence plus kit (Amersham Biosciences, Piscataway, NJ) according to manufacturer’s instructions.
Hematocrit and reticulocytes count
After exposure to room air or hypoxia, mice were euthanized with Nembutal, and blood was collected via cardiac puncture. The blood was analyzed for hematocrit and reticulocyte count. To measure hematocrit, the blood was placed into heparinized capillary tubes (VWR International, West Chester, PA) and centrifuged 1000 x g for 30 min at 4 C. The percentage of reticulocytes was determined using FACStar Plus fluorescence-activated cell sorter and Thiazole Orange-based kit (Retic-Count) from Becton Dickinson (Franklin Lakes, NJ).
Statistical analysis
Statistical analysis was performed using t test. Values of P < 0.05 were considered significant.
Results
Immunolocalization of EC-SOD in mouse kidney
Because the kidney is the major tissue which expresses EC-SOD, we first wanted to understand its localization (24). Using a previously characterized rabbit antimouse EC-SOD peptide purified polyclonal antibody, we found positive immunostaining for EC-SOD in both the renal cortex and the outermost medulla of mouse kidney (Fig. 1A). On the basis of cell morphology, the immunostained elements were identified as proximal convoluted tubules. Most of the staining occurs in the juxtamedullary cortex and outermost medulla where more than half of the tubules revealed strong staining and the remaining subpopulation showed only light immunoreactivity or no staining. A small amount of staining was observed in glomeruli. Heavy EC-SOD staining was detected adjacent to vascular smooth muscle cells and endothelium of large renal vessels (Fig. 1C). The immunoreactivity described above was completely absent in identically prepared sections obtained from EC-SOD KO mice (Fig. 1, B and D). Immunoreactivity in Wt kidneys was also absent using antibodies preabsorbed to EC-SOD peptide (data not shown).
FIG. 1. Immunohistochemical localization of EC-SOD in mouse kidney. Kidney sections from Wt and EC-SOD KO mice were incubated with mouse EC-SOD affinity purified polyclonal IgG. Positive brown labeling is detected in the proximal convoluted tubules in the cortex and outer medulla of Wt mice (A) with no labeling observed in EC-SOD knockout mice (B). Positive brown labeling for EC-SOD is present within the cytoplasm and nucleus of proximal tubule cells. Minimal labeling was observed in glomeruli (g). Extensive labeling was observed around smooth muscle cells of renal blood vessels (bv) (C) with no labeling detected in vessels from knockout mice (D). Magnification, x400.
EC-SOD functions as a major supressor of hypoxia-induced Epo mRNA induction
Because EC-SOD is expressed at high levels in the renal cortex, also the site of hypoxia-inducible Epo expression, we decided to test the hypothesis that EC-SOD could be involved in the regulation of Epo gene expression. To determine the effect of EC-SOD on expression of Epo, we exposed Wt and EC-SOD KO mice to 4 h of either room air (normoxia) or 7% oxygen (hypoxia). Using real-time RT-PCR, we found that, after hypoxia, Wt animals increased kidney Epo mRNA levels 16-fold (Fig. 2A), which is similar to previously published studies that demonstrate an approximate 20-fold increase of Epo mRNA levels after 4 h hypoxia (25). However, EC-SOD KO mice exposed to identical hypoxic conditions, showed a striking 246-fold increase in Epo mRNA levels when compared with normoxia-treated EC-SOD KO animals. Thus, EC-SOD KO mice had a 15 times greater Epo mRNA induction after hypoxia then did Wt mice. The disruption of the EC-SOD gene did not significantly change Epo mRNA levels in the kidney under normoxic conditions. These results demonstrate that the absence of EC-SOD dramatically augments Epo induction by hypoxia. There were no changes in renal EC-SOD mRNA levels for Wt mice after exposure to hypoxia (data not shown).
FIG. 2. Effects of EC-SOD on hypoxia-induced Epo mRNA and protein levels. A, Real-time RT-PCR was used to quantitate the levels of Epo mRNA in whole kidney after 4 h of exposure to either room air or 7% oxygen in EC-SOD KO and Wt mice. The data are representative of the mean ± SE from four mice. *, P < 0.01 vs. normoxia mice; **, P < 0.01 between hypoxic Wt and EC-SOD KO mice. B, Epo levels in serum determined by ELISA. Standard curves with recombinant mouse Epo (mEpo, gray circles) or human Epo (hEpo, black squares) were obtained using the protocol described in Materials and Methods section. Data were analyzed by nonlinear regression analysis. Dashed line, Detection limit for this ELISA. Dotted line, Absorbance and relative Epo concentration in serum of EC-SOD KO mice exposed to 10% hypoxia for 1 d. The table insert shows mouse serum Epo levels in either Wt or EC-SOD KO mice exposed to 10% hypoxia for 1 d.
To determine the time course effect of hypoxia on Epo induction, we exposed Wt and EC-SOD KO mice to 10% oxygen for 4 h and 1, 4, and 7 d. The induction of Epo mRNA in Wt mice reached a maximum of 9-fold after 1 d, decreased to 3.9-fold at d 4, and then returned back to the baseline at d 7 (Table 1). Strikingly, the level of Epo mRNA in EC-SOD null mice increased 86-fold at 4 h, peaked at 1733-fold after 1 d hypoxic exposure, decreased to 7.2-fold at d 4, and then returned to normal after 7 d. The magnitude of Epo mRNA increase was dependent on the oxygen concentration. For example, 7% hypoxia exposure for 4 h caused a 16-fold increase in Epo mRNA in Wt mice and a 246-fold increase in EC-SOD KO mice, whereas a 4-h exposure to 10% oxygen resulted in an induction of only 2- and 68-fold, respectively (Fig. 2A and Table 1). The level of EC-SOD mRNA in the kidney of Wt mice remained unchanged during the course of hypoxic exposure except for a slight decrease in mRNA levels after 7 d of continuous hypoxia (Table 2), whereas no EC-SOD-specific mRNA was detected in EC-SOD KO mice, as expected.
TABLE 1. Induction of Epo mRNA in mouse kidney after whole-animal exposure to 10% oxygen for the times indicated
TABLE 2. Changes in EC-SOD mRNA levels in mouse kidney after exposure to 10% oxygen for the times indicated
Because there is an approximately 80% similarity in amino acid sequences between mouse and human Epo, and due to the absence of any commercially available ELISA for mouse Epo, we used a human Epo ELISA kit to measure mouse Epo levels. As shown in Fig. 2B, the human Epo ELISA kit has a similar detection limit and calibration curve for both mouse and human Epo. This ELISA was not sensitive enough to detect Epo protein levels in the blood of Wt or KO mice exposed to normoxia or Wt mice exposed to hypoxia. However, EC-SOD KO mice exposed to 10% hypoxia for 1 d, showed a significant increase in Epo blood levels to 52 mU/ml (Fig. 2B). Because blood levels of Epo were below the detection limits under normoxic conditions, we could not calculate the fold increase for mouse plasma Epo protein levels.
Marked increase of Epo levels in EC-SOD KO mice did not change the rate of erythropoiesis in response to hypoxia
Our data demonstrate that exposure of EC-SOD KO mice to hypoxia for 1 d significantly increases both Epo mRNA in the kidney and Epo plasma protein levels. To analyze the physiological effect of Epo induction, we measured both hematocrit and reticulocyte counts of mice exposed to hypoxia. After exposure of Wt mice to 10% oxygen for 7 d, the hematocrit increased from 52.2 ± 0.7 to 62.4 ± 2.5%. EC-SOD KO mice, however, showed a significantly lower hematocrit after 1 and 4 d of hypoxia compared with Wt mice, but by 7 d there was no significant difference (Fig. 3A). The reticulocyte counts were also elevated after hypoxia reaching a maximum at 4 d hypoxia exposure and returning back to normal by 7 d (Fig. 3B), similar to previously published data (26). There was a significant difference in reticulocyte counts between Wt and EC-SOD KO mice observed after 4 h and 1 d of hypoxia. After 4 h of hypoxia, EC-SOD KO mice showed higher reticulocyte counts compare to Wt mice, whereas 1 d of hypoxia resulted in lower reticulocyte counts. These data suggest that, although EC-SOD KO mice showed increased Epo expression in response to hypoxia, it does not necessary lead to a corresponding increase in reticulocytosis and erythropoiesis.
FIG. 3. Effect of hypoxia on hematocrit and reticulocyte count in Wt and EC-SOD KO mice. The mice were exposed to 10% oxygen for the time indicated. At the end of the exposure, the blood was collected by heart puncture and hematocrit (A) and reticulocyte count (B) determined as described in Materials and Methods. The data are representative of the mean ± SD from three mice. *, P < 0.05 vs. normoxia mice; **, P < 0.05 between Wt and EC-SOD KO mice. NS, Not significant.
EC-SOD attenuates hypoxia inducible Epo expression and HIF-1 DNA binding
To determine the mechanism by which EC-SOD markedly represses hypoxia-inducible Epo expression, we generated Hep3B clones that overexpress human EC-SOD. The extent of EC-SOD overexpression was determined by RT-PCR (data not shown) and Western blot (Fig. 4A). Two clones were selected based on the magnitude of EC-SOD expression, one clone with a relatively low expression (SOD3 low) and a second clone with high EC-SOD expression (SOD3 high). EC-SOD protein levels were measured in total cellular extract (Fig. 4A), in cell culture medium, and in nuclear extract (data not shown), all by Western blot. We found that EC-SOD protein levels were increased in all three compartments (the nucleus, cytoplasm, and extracellular medium). These results suggest that EC-SOD overexpressed in Hep3B cells is not only secreted into the extracellular space but is also localized to the nuclear fraction.
FIG. 4. Overexpression of EC-SOD attenuates induction of hypoxia-regulated genes in Hep3B cells. A, Western blot of total cellular extract from cells described above exposed for 4 h to PBS or CoCl2 (50 μM). B and C, Induction of Epo (B) and CA IX (C) mRNA levels in cells exposed to CoCl2 (50 μM) or hypoxia (1% O2, 5% CO2, 94% N2) for 4 h. The amplification and quantitation of Epo mRNA levels was performed using TaqMan and primers/probe specific for human Epo. Levels of CA IX mRNA in Hep3B cells were detected by RT-PCR. The expression of Epo and CA IX was normalized to GAPDH. The data are representative of the mean ± SD from four measurements. *, P < 0.05 vs. wt cells.
The exposure of Wt Hep3B cells to CoCl2 (50 μM) or 1% oxygen for 4 h, caused increases in Epo mRNA levels of up to 29- and 50-fold, respectively (Fig. 4B). The EC-SOD overexpressing clones (SOD3 low and SOD3 high) markedly reduced this Epo induction in a dose-dependent manner. For instance, in cells exposed to 50 μM CoCl2, low levels of EC-SOD overexpression reduced Epo induction from 28.5- to 14.6-fold and in clones expressing high levels of EC-SOD to 6.8-fold. Similarly, overexpression of EC-SOD attenuates Epo mRNA induction after exposure to 1% oxygen for 4 h from 49.8-fold in Wt Hep3B cells to 26.3 and 13.1-fold in cells expressing low and high levels EC-SOD, respectively.
To analyze the effect of EC-SOD overexpression on induction of other hypoxia-inducible genes, we measured mRNA levels of carbonic anhydrase (CA) IX. CA IX is strongly up-regulated under hypoxia conditions in HepG2 cells (27) and other tumor cells (28). We found that elevated levels of EC-SOD significantly attenuated induction of CA IX by hypoxia in a dose-dependent manner (Fig. 4C).
To better elucidate the molecular mechanism of EC-SOD’s repression of hypoxia-induced Epo expression, we constructed a hypoxia-sensitive reporter plasmid by ligating two hypoxia response elements from the Epo enhancer derived from human genomic DNA, and placed them upstream of the luciferase reporter gene expressed under the control of the tk promoter (Fig. 5A). The exposure of Wt Hep3B cells transfected with pGL3-tk-Epo(2x) to 50 μM CoCl2 for 4 h led to a 5.5-fold induction in reporter activity (Fig. 5B). As a negative control, transfection of Hep3B cells with the enhancerless pGL3-tk luciferase vector resulted in no increase in luciferase activity after hypoxic exposure. These data indicate that activation of the reporter by CoCl2 is dependent on the presence of the Epo enhancer element containing a functional binding site for HIF-1. In Hep3B cells overexpressing EC-SOD, there was an approximate 30% decrease in CoCl2-induced reporter gene expression compared with Wt cells. When Wt Hep3B cells were transfected with pGL3-tk-Epo(2x) and exposed to 1% oxygen, reporter gene activity increased approximately 8-fold relative to identical cells exposed to 21% oxygen. However, when SOD3 high Hep3B cells were used, reporter gene induction by hypoxia was significantly attenuated by 35% to 5.2-fold (Fig. 5C). These data indicate that EC-SOD attenuates hypoxia-induced Epo gene expression, at least in part, through reducing HIF-1 activity in response to hypoxia.
FIG. 5. Role of HIF-1 in EC-SOD mediated Epo induction by hypoxia. A, Schematic representation of the Epo enhancer/reporter construct. The enhancer region of human Epo gene was concatenated twice and cloned in the forward orientation upstream of tk minimal promoter. The putative HIF-1 binding sites are depicted as the black rectangles. The original reporter plasmid used for this construct was pGL3-Basic with the firefly luciferase gene as a reporter. B, Induction of the reporter by CoCl2 (50 μM) using transient transfection of Wt Hep3B cells or cells that stably overexpressing EC-SOD at high (SOD3 high) levels. C, Same as B but cells were exposed to 21% oxygen or 1% oxygen for 4 h at 37 C. After exposure the firefly luciferase activity was measured and normalized to Renilla luciferase activity produced by the cotransfected control plasmid pRL-CMV. Results shown are the mean ± SD from at least two independent transfection experiments, each performed in quadruplicate. *, P < 0.05. D, Wt Hep3B cells and clone overexpressing EC-SOD at high levels (SOD3 high) were exposed to PBS or 50 μM CoCl2 for 4 h at 37 C. Nuclear extracts were incubated with HIF-1-specific oligonucleotide and separated on a polyacrylamide gel as described in Materials and Methods. For supershift experiments nonimmune (n.i.) or HIF-1 (anti-HIF-1)-specific antibodies were added to the incubation mixtures. Arrows on the right side show HIF-1 and supershifted HIF-1-specific bands. All binding reactions were separated on 5% nondenaturing polyacrylamide gel, dried, and exposed to the x-ray film.
Because hypoxia-mediated induction of Epo is mostly regulated by the HIF-1 transcription factor, we decided to analyze whether EC-SOD overexpression affects HIF-1 DNA binding. Using EMSA and a HIF-1-specific oligonucleotide, we detected a significant increase in the intensity of the HIF-1-specific band in Wt Hep3B cells after CoCl2 exposure (Fig. 5D, compare lane 1 and 2). A similar exposure using SOD3 high Hep3B cells showed that CoCl2-treated cells also increased HIF-1 levels, but to a lesser extent compared with Wt cells. (Fig. 5D, lanes 3 and 4). The specificity of this assay was demonstrated by supershifting the DNA-protein complex with anti-HIF-1 antibodies (Fig. 5D, lanes 5–8). Although nonimmune IgG did not change the electrophoretic mobility of shifted bands (Fig. 5D, lanes 9 and 10), incubation with anti-HIF-1 IgG resulted in the appearance of new slower-migrating band (see arrow on right side of Fig. 5D, lanes 6 and 8). The intensity of this HIF-1 supershifted band was slightly decreased in cells overexpressing EC-SOD (compare lanes 6 and 8 in Fig. 5D). These data support our transient transfection reporter gene experiments and indicate that EC-SOD overexpression, at least in part, modulates hypoxia-mediated changes in HIF-1 levels. Additionally, we analyzed HNF-4 binding activity using EMSA and supershift assay but did not detect any differences between either PBS vs. CoCl2-treated cells or Wt vs. EC-SOD overexpressing cells (data not shown).
Discussion
The presence of high levels of EC-SOD mRNA and protein in the kidney, (24) and EC-SOD’s distinctive localization to specific proximal tubular epithelial cells, suggest a potentially significant role of this antioxidant enzyme in the normal and, perhaps, pathological kidney. Interestingly, EC-SOD appears to have a similar pattern of expression as does that of Epo. For example, in the kidney, Epo is primarily synthesized in the renal cortex with no detectable expression in the medulla. There are contradictory data about which renal cells produce this hormone. Peritubular interstitial cells have been identified as the major source of Epo production (29, 30). However, in contrast to these studies, other groups have detected Epo production primarily by proximal tubular epithelial cells using in situ hybridization (31), immunochemistry (32), and transgenic mice (33). The apparent close proximity of sites of EC-SOD and Epo expression suggest the possible importance of EC-SOD in the regulation of Epo expression. Indeed, we show that mice lacking the EC-SOD gene have a 193-fold higher Epo induction, in response to hypoxia, compared with Wt mice.
The molecular mechanisms mediating this profound effect of EC-SOD on hypoxia-induced Epo expression is not known, but several hypothesis can be considered. Because EC-SOD is the major extracellular antioxidant enzyme that reduces extracellular levels of superoxide, we infer that ROS serve as signaling molecule that regulate hypoxia-induced gene expression in renal proximal tubules. If ROS concentrations increase under hypoxic conditions due to malfunctioning of mitochondrial electron transport chain, EC-SOD KO mice would have significantly higher levels of ROS after hypoxic exposure. Because hypoxia generated superoxide radicals are required for induction of HIF-1 activity and other downstream target genes (7, 8), we can predict that disruption of EC-SOD function will further increase kidney ROS levels that, in turn, will increase HIF-1 levels to a greater extent compared with Wt mice. However, the role ROS play in mediating HIF-1 stabilization remains controversial because recent studies show that the absence of a functional mitochondrial electron transport chain does not interfere with the regulation of HIF-1 and HIF-1 target genes (9).
The role of ROS in hypoxia-mediated gene induction should be considered in relation with another important messenger nitric oxide (NO). It has been shown that NO inhibits hypoxia-inducible transcription of Epo gene through suppression of HIF-1 expression, DNA binding activity, and transcriptional activity (34, 35, 36). The mechanism by which NO exerts its effects on HIF-1 stability are not clearly understood but seem to involve the intracellular S-nitrosylation of cysteine residues in the ODD domain or in critical thiols of ubiquitin-activating enzymes (37). Because superoxide rapidly reacts with NO to form peroxynitrite, elevated levels of superoxide would be predicted to decrease physiologic NO concentrations (38). Furthermore, EC-SOD KO mice have an impaired ability to dismutate superoxide into hydrogen peroxide that could lead to decreases in H2O2 levels. Because NO and H2O2 are known to block the induction of Epo and accumulation of HIF-1 (39), EC-SOD KO mice would be predicted to enhance hypoxia-induced Epo gene expression (Fig. 6).
FIG. 6. Schematic illustration depicting the possible mechanims by which EC-SOD regulates hypoxia-inducible Epo gene transcription. Hypoxia-mediated Epo gene expression is mediated primarily by HIF-1. HIF-1 is a heterodimeric protein composed of two discreet subunits, HIF-1 and ARNT. Under normoxia conditions, HIF-1 is hydroxylated by a prolyl hydroxylases and subsequently ubiquitinated and completely degraded. After hypoxia, an oxygen sensor protein, presumed to be a nicotinamide adenine dinucleotide phosphate oxidase and illustrated here as Renox, signals, by an as yet unknown pathway, leading to the stabilization and accumulation of HIF-1. HIF-1/ARNT heteroduplex then translocates into the nucleus enhancing gene transcription via the hypoxia response element (HRE). Under these conditions, both NO and H2O2 have both been shown to completely inhibit and attenuate this hypoxia-induced gene transcription response. We hypothesized that in tissues lacking EC-SOD expression, levels of superoxide will be markedly increased, which would lead to decreased levels of NO, which would further augment Epo gene expression. Moreover, without EC-SOD, the production of hydrogen peroxide in the extracellular space would be predicted to be diminished, which would additionally remove its repressor activity on hypoxia-inducible Epo gene expression. For a more detailed description, see Discussion.
In addition to the above possible mechanisms, there is small chance that the significant differences in hypoxia-mediated Epo induction seen in Wt and EC-SOD KO mice can be attributed to subtle genetic differences between the two strains of animals used. In our experiments, we have used C57BL/6 x 129/Sv mice purchased from Taconic (as controls) and EC-SOD KO mice that were maintained in the original strain, also of C57BL/6 x 129/Sv, as described previously (16). Although the animals are all of the same background strain, there is the potential due to breeding effects for the animals to have divergent response. However, the experiments performed with Hep3B cells overexpressing EC-SOD provided additional support that EC-SOD does play a critical role in the regulation of Epo expression and that the differences observed in our in vivo model are unlikely to be due to that of genetic variations.
The physiological relevance of such a remarkable up-regulation of Epo expression in EC-SOD KO mice is not clear at this time. Our data suggest that increasing Epo concentrations in the blood of EC-SOD KO mice in response to hypoxia does not cause an expected significant increase in erythropoiesis compared with Wt mice. These results suggest that the nascent Epo being produced in the EC-SOD KO mice may show reduced or lack hematopoietic activities. Recent studies have demonstrated that carbamylated Epo and certain Epo mutants do not bind the classical Epo receptor, and hence did not show any hematopoietic activity (40). Nevertheless, these Epo variants were able to confer cytoprotective benefits despite nonhematopoietic effects. Several other studies indicate that Epo’s physiologic effects are not restricted to just the stimulation of red blood cells production, but seem to be quite pleitropic. For instance, it has been shown that increasing Epo levels protect neurons from ischemic stroke (41, 42) as well as prevent light-induced retinal degradation (43). Preconditioning cardiomyocytes with Epo in vivo as well as in vitro protect them from ischemic injury. Moreover, a single dose of Epo delivered at the time of cardiac reperfusion had a profound beneficial therapeutic effect on the ischemic heart but did not change hematocrit (44). Thus, we speculate that the increased Epo being produced in EC-SOD KO mice to have poor activity with the classical Epo receptor, perhaps due to oxidative modifications that are a direct result of EC-SOD’s absence in the kidney. Whether this enhanced Epo production retains cytoprotoective properties remains to be studied.
In summary, we show that EC-SOD is primarily immunolocalized in proximal tubule cells in the kidney and dramatically represses the hypoxia-inducible Epo gene induction. In cell culture studies, overexpression of EC-SOD decreased hypoxia-induced Epo gene expression in vivo and in vitro in a concentration-dependent fashion. The mechanism for EC-SOD’s profound effect on hypoxia-induced Epo gene expression is dependent, at least in part, on stabilization of HIF-1. Epo produced under hypoxic conditions in EC-SOD KO mice appears to have lost hematopoietic activity. Overall, our results implicate superoxide as an important component of the signal transduction pathway that leads to amplification of hypoxia-inducible Epo gene expression.
Acknowledgments
We thank Dr. Murat Arcasoy for critical review of this manuscript. We thank members of the Duke University Cell Culture Facility for help in obtaining and culturing of mammalian cells.
References
Zhu H, Jackson T, Bunn HF 2002 Detecting and responding to hypoxia. Nephrol Dial Transplant 17(Suppl 1):3–7
Michiels C, Minet E, Mottet D, Raes M 2002 Regulation of gene expression by oxygen: NF-B and HIF-1, two extremes. Free Radic Biol Med 33:1231–1242
Huang LE, Bunn HF 2003 Hypoxia-inducible factor and its biomedical relevance. J Biol Chem 278:19575–19578
Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell SJ, Kriegsheim AV, Hebestreit HF, Mukherji M, Schofield CJ, Maxwell PH, Pugh CW, Ratcliffe PJ 2001 Targeting of HIF- to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 292:468–472
Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, Salic A, Asara JM, Lane WS, Kaelin Jr WG 2001 HIF targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 292:464–468
Epstein AC, Gleadle JM, McNeill LA, Hewitson KS, O’Rourke J, Mole DR, Mukherji M, Metzen E, Wilson MI, Dhanda A, Tian YM, Masson N, Hamilton DL, Jaakkola P, Barstead R, Hodgkin J, Maxwell PH, Pugh CW, Schofield CJ, Ratcliffe PJ 2001 C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell 107:43–54
Chandel NS, Maltepe E, Goldwasser E, Mathieu CE, Simon MC, Schumacker PT 1998 Mitochondrial reactive oxygen species trigger hypoxia-induced transcription. Proc Natl Acad Sci USA 95:11715–11720
Agani FH, Pichiule P, Chavez JC, LaManna JC 2000 The role of mitochondria in the regulation of hypoxia-inducible factor 1 expression during hypoxia. J Biol Chem 275:35863–35867
Vaux EC, Metzen E, Yeates KM, Ratcliffe PJ 2001 Regulation of hypoxia-inducible factor is preserved in the absence of a functioning mitochondrial respiratory chain. Blood 98:296–302
Srinivas V, Leshchinsky I, Sang N, King MP, Minchenko A, Caro J 2001 Oxygen sensing and HIF-1 activation does not require an active mitochondrial respiratory chain electron-transfer pathway. J Biol Chem 276:21995–21998
Ehleben W, Porwol T, Fandrey J, Kummer W, Acker H 1997 Cobalt and desferrioxamine reveal crucial members of the oxygen sensing pathway in HepG2 cells. Kidney Int 51:483–491
Fandrey J, Frede S, Ehleben W, Porwol T, Acker H, Jelkmann W 1997 Cobalt chloride and desferrioxamine antagonize the inhibition of erythropoietin production by reactive oxygen species. Kidney Int 51:492–496
Fridovich I 1997 Superoxide anion radical (O2–.), superoxide dismutases, and related matters. J Biol Chem 272:18515–18517
Zelko IN, Mariani TJ, Folz RJ 2002 Superoxide dismutase multigene family: a comparison of the CuZn-SOD (SOD1), Mn-SOD (SOD2), and EC-SOD (SOD3) gene structures, evolution, and expression. Free Radic Biol Med 33:337–349
Marklund SL 1984 Extracellular superoxide dismutase and other superoxide dismutase isoenzymes in tissues from nine mammalian species. Biochem J 222:649–655
Carlsson LM, Jonsson J, Edlund T, Marklund SL 1995 Mice lacking extracellular superoxide dismutase are more sensitive to hyperoxia. Proc Natl Acad Sci USA 92:6264–6268
Suliman HB, Ryan LK, Bishop L, Folz RJ 2001 Prevention of influenza-induced lung injury in mice overexpressing extracellular superoxide dismutase. Am J Physiol Lung Cell Mol Physiol 280:L69–L78
Folz RJ, Abushamaa AM, Suliman HB 1999 Extracellular superoxide dismutase in the airways of transgenic mice reduces inflammation and attenuates lung toxicity following hyperoxia. J Clin Invest 103:1055–1066
Marklund SL 1982 Human copper-containing superoxide dismutase of high molecular weight. Proc Natl Acad Sci USA 79:7634–7638
Willems J, Zwijsen A, Slegers H, Nicolai S, Bettadapura J, Raymackers J, Scarcez T 1993 Purification and sequence of rat extracellular superoxide dismutase B secreted by C6 glioma. J Biol Chem 268:24614–24621
Folz RJ, Crapo JD 1994 Extracellular superoxide dismutase (SOD3): tissue-specific expression, genomic characterization, and computer-assisted sequence analysis of the human EC SOD gene. Genomics 22:162–171
Ookawara T, Imazeki N, Matsubara O, Kizaki T, Oh-Ishi S, Nakao C, Sato Y, Ohno H 1998 Tissue distribution of immunoreactive mouse extracellular superoxide dismutase. Am J Physiol 275:C840–C847
Oury TD, Day BJ, Crapo JD 1996 Extracellular superoxide dismutase in vessels and airways of humans and baboons. Free Radic Biol Med 20:957–965
Folz RJ, Guan J, Seldin MF, Oury TD, Enghild JJ, Crapo JD 1997 Mouse extracellular superoxide dismutase: primary structure, tissue-specific gene expression, chromosomal localization, and lung in situ hybridization. Am J Respir Cell Mol Biol 17:393–403
Chikuma M, Masuda S, Kobayashi T, Nagao M, Sasaki R 2000 Tissue-specific regulation of erythropoietin production in the murine kidney, brain, and uterus. Am J Physiol Endocrinol Metab 279:E1242–E1248
Mylrea KC, Abbrecht PH 1970 Hematologic responses of mice subjected to continuous hypoxia. Am J Physiol 218:1145–1149
Sonna LA, Cullivan ML, Sheldon HK, Pratt RE, Lilly CM 2003 Effect of hypoxia on gene expression by human hepatocytes (HepG2). Physiol Genom 12:195–207
Wykoff CC, Beasley NJ, Watson PH, Turner KJ, Pastorek J, Sibtain A, Wilson GD, Turley H, Talks KL, Maxwell PH, Pugh CW, Ratcliffe PJ, Harris AL 2000 Hypoxia-inducible expression of tumor-associated carbonic anhydrases. Cancer Res 60:7075–7083
Koury ST, Bondurant MC, Koury MJ 1988 Localization of erythropoietin synthesizing cells in murine kidneys by in situ hybridization. Blood 71:524–527
Lacombe C, Da Silva JL, Bruneval P, Fournier JG, Wendling F, Casadevall N, Camilleri JP, Bariety J, Varet B, Tambourin P 1988 Peritubular cells are the site of erythropoietin synthesis in the murine hypoxic kidney. J Clin Invest 81:620–623
Shanks JH, Hill CM, Lappin TR, Maxwell AP 1996 Localization of erythropoietin gene expression in proximal renal tubular cells detected by digoxigenin-labelled oligonucleotide probes. J Pathol 179:283–287
Maxwell AP, Lappin TR, Johnston CF, Bridges JM, McGeown MG 1990 Erythropoietin production in kidney tubular cells. Br J Haematol 74:535–539
Loya F, Yang Y, Lin H, Goldwasser E, Albitar M 1994 Transgenic mice carrying the erythropoietin gene promoter linked to lacZ express the reporter in proximal convoluted tubule cells after hypoxia. Blood 84:1831–1836
Liu Y, Christou H, Morita T, Laughner E, Semenza GL, 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
Huang LE, Willmore WG, Gu J, Goldberg MA, Bunn HF 1999 Inhibition of hypoxia-inducible factor 1 activation by carbon monoxide and nitric oxide. Implications for oxygen sensing and signaling. J Biol Chem 274:9038–9044
Sogawa K, Numayama-Tsuruta K, Ema M, Abe M, Abe H, Fujii-Kuriyama Y 1998 Inhibition of hypoxia-inducible factor 1 activity by nitric oxide donors in hypoxia. Proc Natl Acad Sci USA 95:7368–7373
Palmer LA, Gaston B, Johns RA 2000 Normoxic stabilization of hypoxia-inducible factor-1 expression and activity: redox-dependent effect of nitrogen oxides. Mol Pharmacol 58:1197–1203
Ortega Mateo A, Amaya Aleixandre de Artinano 2000 Nitric oxide reactivity and mechanisms involved in its biological effects. Pharmacol Res 42:421–427
Huang LE, Arany Z, Livingston DM, Bunn HF 1996 Activation of hypoxia-inducible transcription factor depends primarily upon redox-sensitive stabilization of its subunit. J Biol Chem 271:32253–32259
Leist M, Ghezzi P, Grasso G, Bianchi R, Villa P, Fratelli M, Savino C, Bianchi M, Nielsen J, Gerwien J, Kallunki P, Larsen AK, Helboe L, Christensen S, Pedersen LO, Nielsen M, Torup L, Sager T, Sfacteria A, Erbayraktar S, Erbayraktar Z, Gokmen N, Yilmaz O, Cerami-Hand C, Xie QW, Coleman T, Cerami A, Brines M 2004 Derivatives of erythropoietin that are tissue protective but not erythropoietic. Science 305:239–242
Brines ML, Ghezzi P, Keenan S, Agnello D, de Lanerolle NC, Cerami C, Itri LM, Cerami A 2000 Erythropoietin crosses the blood-brain barrier to protect against experimental brain injury. Proc Natl Acad Sci USA 97:10526–10531
Siren AL, Fratelli M, Brines M, Goemans C, Casagrande S, Lewczuk P, Keenan S, Gleiter C, Pasquali C, Capobianco A, Mennini T, Heumann R, Cerami A, Ehrenreich H, Ghezzi P 2001 Erythropoietin prevents neuronal apoptosis after cerebral ischemia and metabolic stress. Proc Natl Acad Sci USA 98:4044–4049
Grimm C, Wenzel A, Groszer M, Mayser H, Seeliger M, Samardzija M, Bauer C, Gassmann M, Reme CE 2002 HIF-1-induced erythropoietin in the hypoxic retina protects against light-induced retinal degeneration. Nat Med 8:718–724
Parsa CJ, Matsumoto A, Kim J, Riel RU, Pascal LS, Walton GB, Thompson RB, Petrofski JA, Annex BH, Stamler JS, Koch WJ 2003 A novel protective effect of erythropoietin in the infarcted heart. J Clin Invest 112:999–1007(Igor N. Zelko and Rodney )