NADPH-Cytochrome P-450 Reductase in the Plasma Membrane Modulates the Activation of Hypoxia-inducible Factor 1*

Hypoxia induces a group of physiologically important genes that include erythropoietin (EPO) and vascular endothelial growth factor (VEGF). Hypoxia-inducible factor 1 (HIF-1) was identified as a hypoxia-activated transcription factor; however, the molecular mechanisms that underlie hypoxia signal transduction in mammalian cells remain undefined. In this study, we found that a flavoprotein, NADPH-P450 reductase (NPR), could regulate the induction of EPO mRNA under hypoxic conditions. Hypoxic EPO mRNA induction in Hep3B cells was inhibited by diphenyleneiodonium chloride, which is an inhibitor of NADPH-dependent enzymes. NPR antisense cDNA was transfected into Hep3B cells, and NPR-deficient hepatocyte cells (NPR− cells) were established. NPR− cells lacked EPO induction under hypoxia, and HIF-1α in NPR− cells did not respond to either transcriptional activation or translocation to the nucleus based on electrophoretic mobility shift assays and reporter gene assay including hypoxia response element. In contrast, NPR overexpression in Hep3B cells enhanced the DNA binding activity of HIF-1α by luciferase reporter gene assay. A study with HeLa S3 cells produced the same results. Furthermore, anti-NPR IgG inhibited EPO induction. EPO induction inhibited by diphenyleneiodonium chloride was recovered by bovine serum albumin-NADPH (a covalent binding complex of bovine serum albumin and NADPH) as well as NADPH. These results suggested that NPR located at the plasma membrane regulates EPO expression in hypoxia, including HIF-1 activation and translocation. We further studied the expression of NPR and VEGF mRNAs in human tumor tissues and found that the NPR mRNA levels were correlated with the VEGF mRNA levels, suggesting that NPR might be an important factor in the hypoxic induction of genes such as VEGF in vivo.

Hypoxia induces a group of physiologically important genes that include erythropoietin (EPO) and vascular endothelial growth factor (VEGF). Hypoxia-inducible factor 1 (HIF-1) was identified as a hypoxia-activated transcription factor; however, the molecular mechanisms that underlie hypoxia signal transduction in mammalian cells remain undefined. In this study, we found that a flavoprotein, NADPH-P450 reductase (NPR), could regulate the induction of EPO mRNA under hypoxic conditions. Hypoxic EPO mRNA induction in Hep3B cells was inhibited by diphenyleneiodonium chloride, which is an inhibitor of NADPH-dependent enzymes. NPR antisense cDNA was transfected into Hep3B cells, and NPR-deficient hepatocyte cells (NPR ؊ cells) were established. NPR ؊ cells lacked EPO induction under hypoxia, and HIF-1␣ in NPR ؊ cells did not respond to either transcriptional activation or translocation to the nucleus based on electrophoretic mobility shift assays and reporter gene assay including hypoxia response element. In contrast, NPR overexpression in Hep3B cells enhanced the DNA binding activity of HIF-1␣ by luciferase reporter gene assay. A study with HeLa S3 cells produced the same results. Furthermore, anti-NPR IgG inhibited EPO induction. EPO induction inhibited by diphenyleneiodonium chloride was recovered by bovine serum albumin-NADPH (a covalent binding complex of bovine serum albumin and NADPH) as well as NADPH. These results suggested that NPR located at the plasma membrane regulates EPO expression in hypoxia, including HIF-1 activation and translocation. We further studied the expression of NPR and VEGF mRNAs in human tumor tissues and found that the NPR mRNA levels were correlated with the VEGF mRNA levels, suggesting that NPR might be an important factor in the hypoxic induction of genes such as VEGF in vivo.
Decreased cellular oxygen tension, that is hypoxia, occurs under physiological conditions such as high altitudes and physiological exercise and under pathological conditions including ischemia, inflammation, and neoplasm. The cells adapt to hypoxia mainly via propagating energy metabolism, ventilation of blood flow, and erythropoiesis. Hypoxia stimulates a group of physiologically important genes such as erythropoietin (EPO) 1 and vascular endothelial growth factor (VEGF). Transcriptional activation of the EPO gene is controlled via an enhancer element located in the 3Ј-flanking region of the gene, hypoxia response element (HRE), and requires binding of a specific transcription factor termed hypoxia-inducible factor 1 (HIF-1) (1). Identification and cloning of HIF-1 revealed a heterodimeric protein consisting of two subunits: HIF-1␣ and HIF-1␤ (2). HIF-1␣ protein undergoes rapid degradation by proteasomes under normoxic conditions (3,4). Targeting for the protein degradation is determined by binding to the von Hippel-Lindau protein (VHL) to form a ubiquitin-ligating complex (5). In hypoxia, degradation of HIF-1␣ is blocked, and then HIF-1␣ is transferred into the nucleus to bind to HRE.
The mechanisms of sensing low oxygen and transduction of this signal to HIF-1 are not well understood. Recently, researchers have hypothesized that the mechanisms involve the generation of reactive oxygen species by NADPH-oxidase. Cytochrome b 558 is the redox core of the NADPH-oxidase complex in phagocytes (6) and B lymphocytes (7,8) and is a membranebound protein consisting of two subunits, p22 phox and gp91 phox (9). Although the presence of cytochrome b 558 was previously believed to be restricted to these cell types, its presence has also been demonstrated in other cells, such as rat carotid body cells (10) and HepG2 cells (11,12). In the carotid body, reactive oxygen species produced by NADPH-oxidase opens a K ϩ channel whose signal is transferred to HIF-1. In support of this model, exogenous H 2 O 2 and diphenyleneiodonium chloride (DPIC), an inhibitor of NADPH-dependent enzymes, are found to inhibit hypoxic stabilization of HIF-1␣ (13,14). However, Archer et al. (15) showed that NADPH-oxidase is not an oxygen sensor by using knockout mice lacking the gp91 phox gene. They found that the K ϩ channel of the knockout mouse was opened by hypoxia, although DPIC inhibited it efficiently.
A different model based on the role of the mitochondrial electron transport chain has been suggested as a mechanism of hypoxic response. Inhibition of the mitochondrial respiratory chain blocks HIF-1 DNA binding activity by electrophoretic mobility shift assays (EMSA) and also blocks the induction of EPO, VEGF, aldolase, and phosphoglycerate kinase mRNA, which are HIF-1 target genes in Hep3B cells under hypoxic conditions (16). Whereas, Srinivas et al. (17) found that mitochondrial DNA-less ( 0 ) cells that lack a respiratory chain protein have a normal response to hypoxia.
Cytochrome P-450 (P450) monooxygenase has been proposed as a microvascular oxygen-sensing protein. P450 enzymes produce a series of vasoactive metabolites from arachidonic acid * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ To whom correspondence should be addressed. Present address: School of Science and Technology, Kwansei Gakuin University, 2-1 Gakuen, Sanda 669-1337, Japan. Tel.: 81-795-65-7673; Fax: 81-795-65-7673; E-mail: imaoka@ksc.kwansei.ac.jp. 1 The abbreviations used are: EPO, erythropoietin; HIF, hypoxiainducible factor; VEGF, vascular endothelial growth factor; NPR, NADPH-P450 reductase; DPIC, diphenyleneiodonium chloride; NPR Ϫ cell, NPR-deficient hepatocyte cell; EMSA, electrophoretic mobility shift assay; HRE, hypoxia response element; P450, cytochrome P-450; pcDNA-aNPR, anti-sense NPR mRNA expression plasmid; pcDNA-sNPR, NPR expression plasmid; RT, reverse transcriptase; EDTA, ethylenediamine-N,N,NЈ,NЈ-tetraacetic acid; BSA, bovine serum albumin. (18,19). Harder et al. (20) demonstrated that the production of 20-hydroxyeicosatetraenoic acid by P450 in rat vascular tissue is directly dependent on the concentration of oxygen. P450 enzymes require molecular oxygen for their activity, and the majority of them require only very low PO 2 levels for normal activity. The unique characteristic of the extrahepatic P450 enzymes responsible for 20-hydroxyeicosatetraenoic acid formation is that a high level of PO 2 is required for the catalytic activity of these enzymes. Furthermore, CO, which efficiently binds to heme, inhibits chloramphenicol acetyltransferase activity including HRE element in rat aortic smooth muscle cells (21) or HIF-1 binding activity by EMSA into Hep3B cells (22). There is no definitive evidence that the P450 system contributes to gene regulation under hypoxic conditions. P450s require NADPH-P450 reductase (NPR) for electron transfer as well as molecular oxygen. NPR is known as a flavoprotein and requires NADPH. In this study, we observed that DPIC inhibited hypoxic EPO induction in Hep3B cells that did not express gp91 phox mRNA. This result suggested that DPIC did not inhibit gp91 phox (a component of NADPH-oxidase) in the cells, suggesting the presence of another flavoprotein that regulates hypoxic response of the cells. We found that NPR plays an important role in the regulation of HIF-1 activation, stabilization, and HRE binding under hypoxic conditions. We also propose that NPR is present in the plasma membrane, although it was previously believed to exist only in the endoplasmic reticulum. We further studied the correlation of NPR expression with VEGF and HIF-1 expression in human tumor tissues.
Isolation of cDNA and Construction of Plasmid-The entire coding region of human NPR was isolated from a human liver cDNA library (CLONTECH, Palo Alto, CA) by PCR. PCR was performed with Pyrobest DNA polymerase as follows: 35 cycles of 94°C for 30 s, 55°C for 30 s, 72°C for 1.5 min. The NPR coding regions of 1.2 kb upstream and 1.3 kb downstream were amplified using each primer set. The primers were designed from the total nucleotide sequence (which lost 5Ј-terminal ATG) reported by Shephard et al. (23). The upstream primers were 5Ј-CCCCTCGAGATGGGAGACTCCCACGTGGACACCA (a single line indicates the XhoI site, and a double line indicates the initiation codon, nucleotides 1-25) and 5Ј-CGCCAGCTCGTACAGCACGT (antisense, nucleotides 1151-1170). The downstream primers were 5Ј-ATGCACCT-GGAATTGGACAT (sense, nucleotides 901-920) and 5Ј-GGGGAATTC-CTAGCTCCACACGTCCAGGGAGTAG (a line indicates the EcoRI site, nucleotides 2010 -2034). The upstream fragment was digested with XhoI and PvuII, and the downstream fragment was digested with PvuII and EcoRI. These two fragments were ligated with pBluescript cut with XhoI and EcoRI. The full-length NPR cDNA was cut out with XhoI and EcoRI and subcloned into pcDNA, an expression vector for mammalian cells. The resulting plasmid contained antisense NPR cDNA, called pcDNA-aNPR. The plasmid containing sense NPR cDNA, called pcDNA-sNPR, was inserted into the KpnI-EcoRI sites of pcDNA. The entire coding region of human HIF-1␣ was isolated from Hep3B cDNA by PCR. The PCR conditions were the same as those used for human NPR. The HIF-1␣ coding region was divided into three fragments and amplified by PCR. The primers for PCR were designed with reference to the nucleotide sequence of HIF-1␣ reported by Hogenesch (24). The upstream region (0.9 kb), the middle region (1.5 kb), and the downstream region (0.7 kb) were amplified using the following sets of primers: 5Ј-GGGGTACCACCGATTCACCATGGAGGGC (a single line indicates the KpnI site, and a double line indicates the initiation codon, nucleotides Ϫ11 to 9), 5Ј-ACTGTCCTGTGGTGACTTGT (antisense, nucleotides 894 -903), 5Ј-TATGACCTGCTTGGTGCTGA (sense, nucleotides 648 -667), 5Ј-TTCCTCAGGAACTGTAGTTC (antisense, nucleotides 2093-2112), 5Ј-CAGTTACAGTATTCCAGCAG (sense, nucleotides 1793-1812), and 5Ј-TTCGCGGCCGCCGTTAACTTGATCCAAAG-CTCT (a line indicates the NotI site, nucleotides 2453-2478), respectively. The three fragments were successively subcloned into pBluescript digested with KpnI and SalI, SalI and SpeI, and SpeI and NotI, respectively. The constructed plasmid including full-length HIF-1␣ was cut with KpnI and PstI. The fragment was subcloned into pQE82L vector, which contains a His tag sequence and is used for the expression of His tag protein in Escherichia coli (Qiagen).
Preparation of Antibodies-Human NPR expressed in E. coli was a generous gift of Dr. S. Asahi (Department of Biology, Graduate School of Science, Osaka University). Human NPR was purified as described previously (25). Human HIF-1␣, which has a His tag, was expressed in E. coli as DH5␣ cells according to the manufacturer's instructions (Qiagen). The DH5␣ cells were cultured in a Jar Fermenter (K-0816, Able, Tokyo, Japan) and were collected by centrifugation. The expressed HIF-1␣ was solubilized in 6 M guanidine HCl. HIF-1␣ containing His tag in its N terminus was purified by affinity chromatography with a nickel-nitrilotriacetic acid-Sepharose gel (Qiagen) according to the manufacturer's instructions. The absorbed protein was eluted with 0.25 M imidazole. Purified HIF-1␣ solution was dialyzed against phosphate-buffered saline and used for the preparation of antibodies. The preparation of antibodies against NPR and HIF-1␣ is described elsewhere (26). Human NPR (50 -100 g) or HIF-1␣ (50 -100 g) was given to rabbits. An antibody against gp91 phox was a generous gift from Dr. E. F. Sato (Department of Biochemistry and Molecular Pathology, Osaka City University).
Cell Culture and Transfection-The human hepatoma cell line Hep3B was obtained from the Cell Resource Center for Biomedical Research at the Institute of Development, Aging and Cancer of Tohoku University (Sendai, Japan). The human cervix cell line HeLa S3 (JCRB0713) was provided by the Health Science Research Resources Bank (Osaka, Japan). Hep3B cells and HeLa S3 cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal calf serum. Stable transfectants of NPR knockdown lines were generated from Hep3B cells by transfection of antisense NPR expression plasmids (pcDNA-aNPR) (5 g). After the transfection, the cells were cultured in serum-free medium for 4 h. Then the cells received fresh medium containing 10% fetal calf serum. The clones were picked up as individual colonies and maintained in medium containing G418 (600 g/ml). The expression of NPR in Hep3B cells transfected with pcDNA-aNPR was checked by immunoblotting with an anti-NPR antibody. At least two independent clones were isolated for each expression plasmid. DPIC was dissolved in water and added to the medium (final concentration, 100 M). For hypoxic stimulation, the cells were incubated in 1% O 2 , 5% CO 2 and nitrogen-balanced with a modulator incubator chamber (Napco 7101, Winchester, VA).
Electrophoretic Mobility Shift Assay-The Hep3B cells were washed with phosphate-buffered saline and collected by centrifugation in icecold phosphate-buffered saline. The nuclear extracts for the cells were prepared by the method of Wang et al. (2). The oligonucleotides for  ). B, HIF-1 binding activity to HRE in the presence of DPIC. Hep3B cells were exposed to normoxia or hypoxia (1%) in the presence of DPIC. The nuclear extracts were prepared from Hep3B cells and analyzed by electrophoretic mobility shift assay.
Preparation of Neutrophils-Human neutrophils were isolated from the peripheral blood of healthy adult volunteers by sedimentation through mono-poly-resolving medium (Dainippon Pharmaceutical Company, Osaka, Japan).
Reporter Gene Assay with HRE-Complementary oligonucleotides (50 bp containing HRE of EPO) were synthesized (22) and subcloned into pGL3-SV40 digested with BamHI. The clone containing two copies of HRE was picked up and named pGL3-SV40HRE. Both pGL3-SV40HRE and pRL-TK were lipofected into the cultured cells (5 ϫ 10 5 Hep3B cells) in a 30-mm dish with GenePORTER TM 2 transfection reagent (Gene Therapy Systems, San Diego, CA). The luciferase activity was measured 48 h later with a luminometer (Lumat LB9507, Berthold, Tokyo Japan) using the dual luciferase reporter system (Toyo-Inki, Tokyo, Japan).
Tissue Characteristics-Microsomes of human bladder and Hep3B cells were prepared as described previously (31). Human bladder tissue samples were obtained by total cystectomy or transurethral resection at the Osaka City General Hospital and Osaka City University Hospital. None of these patients had received any type of therapy for bladder tumors prior to surgery. From the resected bladders, small pieces of tissue, including the bladder mucosa, were rapidly extracted and frozen in liquid nitrogen. All of the tissues were taken after informed consent had been obtained from the patients and their families.

Involvement of NADPH-dependent Enzymes in EPO Induction-We examined the effect of DPIC upon the induction of the EPO gene by hypoxia in human hepatoma cells (Hep3B cells).
DPIC is an inhibitor of NADPH-dependent enzymes such as NADPH-oxidase and NPR. The Hep3B cells were incubated under hypoxia (1% oxygen) for 3 h in the presence of DPIC (100 M). Exposure to DPIC markedly reduced the hypoxic induction of EPO (Fig. 1). The presence of NADPH (1 mM) together with DPIC recovered the induction of EPO mRNA in hypoxia, but the presence of NADH did not. These findings suggest that the NADPH-dependent enzyme NADPH-oxidase, a candidate oxygen sensor, plays a role in the hypoxic EPO mRNA induction. For that reason, we investigated the expression of FIG. 6. Dual luciferase reporter assay in HeLa S3 cells. A, RT-PCR using RNA from HeLa S3 cells with specific primers for gp91 phox . B, effect of NPR antisense cDNA and NPR sense cDNA on the luciferase activity. HeLa S3 cells were co-transfected with a reporter plasmid (pGL3-SV40HRE vector) containing the HIF-1 binding domain of the human EPO gene (HRE) and a control plasmid (pRL-TK vector), containing a basal T4 promoter. Either pcDNA-aNPR or pcDNA-sNPR was transfected together with pGL3-SV40HRE and pRL-TK vector, and the cells were incubated for 2 h under hypoxia or normoxia. The amount of luciferase activity was quantitated with the cell lysates. The ratio (reporter/control luciferase activity) obtained from the cell lysates under hypoxia was set at 100%. Luciferase activity is given as the mean Ϯ S.D. of six separate experiments. NADPH-oxidase component proteins, p22 phox and gp91 phox , using reverse transcriptase (RT)-PCR ( Fig. 2A). However, the two proteins were not detected in Hep3B cells. On the contrary, NPR, which is also an NADPH-dependent enzyme, was detected by RT-PCR. NPR protein was abundantly present in Hep3B cells (Fig. 2B).
Study of NPR Function Using NPR Ϫ Cells-We examined whether NPR regulates EPO mRNA induction by hypoxia. NPR Ϫ cells were prepared by transfection of pcDNA-aNPR plasmid. The cell lysates were subjected to immunoblotting with anti-human NPR antibody (Fig. 3A) or to cytochrome c enzyme assay. The NPR in NPR Ϫ cells was much less than that in control cells transfected with vector only. The NPR activity assayed by cytochrome c reduction in NPR Ϫ cells was also extensively reduced (data not shown). Hypoxic EPO mRNA induction using NPR Ϫ cells was investigated (Fig. 3B). The expression of EPO mRNA was suppressed by NPR knockdown, whereas it was induced by hypoxia in control cells. NPR Ϫ cells were next treated with CoCl 2 , which mimics hypoxia (32). The induction of EPO mRNA by CoCl 2 in NPR Ϫ cells occurred to the same degree as in control cells, indicating that these cells maintain the capacity of EPO induction. This study suggests that NPR is an important factor in hypoxic induction of EPO mRNA.
Regulation of HIF-1␣ Stabilization and HRE Binding by NPR-Next, we investigated whether NPR regulates the EPO expression at the transcriptional level in Hep3B cells. Hypoxia stabilized HIF-1␣ and increased the HIF-1␣ protein level, but the level was dramatically decreased in the presence of DPIC (Fig. 4A), indicating that DPIC inhibited HIF-1␣ stabilization (the HIF-1␣ mRNA level was not changed). We examined whether NPR influences the activation or stabilization of HIF-1 and the interaction of HIF-1 with HRE of the EPO gene (Fig.  4B). Hep3B cells were incubated under normoxia or hypoxia for 3 h, in the presence or absence of DPIC. The nuclear protein extracts were prepared from the cells, and the HIF-1 binding activity was examined with EMSA. The HRE oligonucleotide probe produced one major DNA-protein complex under hypoxia, and the protein complex disappeared when DPIC was added to the cells. We also investigated the binding of HIF-1 with HRE with a reporter gene assay including the plasmid constituted with HRE and the luciferase gene (Fig. 5). Under hypoxic conditions, the cells lipofected by the reporter gene were observed to induce the luciferase activity by 12.5-fold compared with that occurring under normoxic conditions. When DPIC was added, the induction of luciferase activity by hypoxia was inhibited, and it almost matched the normoxia level (Fig. 5A). We next lipofected both the reporter gene and either pcDNA-aNPR, pcDNA-sNPR, or empty vector to Hep3B cells (Fig. 5B). The induction of luciferase activity was also FIG. 7. Effect of anti-NPR IgG on HIF-1 activation and translocation to the nucleus. A, HIF-1 binding activity to HRE in the presence of anti-NPR IgG. Hep3B cells were exposed to normoxia or hypoxia (1%) in the presence of anti-NPR IgG or control IgG. The nuclear extracts were prepared from Hep3B cells and analyzed by electrophoretic mobility shift assay. B, induction of luciferase by hypoxia in the presence of anti-NPR IgG. Hep3B cells were co-transfected with a reporter plasmid (pGL3-SV40HRE vector) containing the HIF-1-binding domain of the human EPO gene (HRE) and a control plasmid (pRL-TK vector) containing a basal T4 promoter. At 48 h post-transfection, the cells were incubated for 2 h under hypoxia or normoxia. An anti-NPR IgG was added to the transfected Hep3B cells. The amount of luciferase activity was quantitated with the cell lysates. The ratio (reporter/control luciferase activity) obtained from the cell lysates under hypoxia was set at 100%. Luciferase activity is given as the mean Ϯ S.D. of six separate experiments. Detection was done with horseradish peroxidase-conjugated goat anti-rabbit IgG and ECL. The gp91 phox in the neutrophils appeared as a thick, smeared band around 75 kDa by anti-gp91 phox antibody, but anti-NPR antibody did not produce such a band in the neutrophils, indicating that the antibody did not cross-react with gp91 phox . inhibited by the reduction in NPR. In contrast, NPR overexpression increased the luciferase activity under hypoxic conditions, although there was no effect under normoxic conditions. These results indicate that the expression levels of NPR in Hep3B cells are important for hypoxic response. We further studied the hypoxic response in another cell line, HeLa S3, which expresses NADPH-oxidase (Fig. 6A), although Hep3B did not express NADPH-oxidase. Either pcDNA-aNPR or pcDNA-sNPR was transfected into HeLa S3 cells, and the hypoxic response was investigated (Fig. 6B). As in Hep3B cells, the induction of luciferase activity in hypoxia was inhibited by the reduction in NPR and was increased by the overexpression of NPR. These results indicate that NPR is important for the activation of HIF-1 and the interaction of HIF-1 with HRE.
Elucidation of NPR Localization-To elucidate the function of NPR, NPR IgG was added to Hep3B cells under hypoxic conditions (Fig. 7A). DNA-protein complex under hypoxia by EMSA was reduced by anti-NPR IgG, although control IgG had no effect. We next expressed transient fusion luciferase proteins containing the HRE and added anti-NPR IgG to the cells (Fig. 7B). The addition of control IgG did not affect the luciferase activity induced by hypoxia, whereas anti-NRP IgG reduced it by 50% in comparison. Because the IgG protein cannot pass through the plasma membrane, NPR IgG may bind with NPR located on the surface of the plasma membrane. This antibody against NPR did not cross-react with gp91 phox , which was expressed abundantly in human neutrophils, indicating that the specificity of an anti-NPR antibody is high (Fig. 7, C  and D). We have supplied evidence that BSA-NADPH (a covalent complex of NADPH and BSA) recovered the EPO induction inhibited by DPIC as well as NADPH (Fig. 8). We concluded that NADPH acted on the plasma membrane, because neither BSA-NADPH nor NPR IgG can pass through the membrane. DISCUSSION NPR is a major NADPH-dependent enzyme transferring electrons to microsomal P450 and heme oxygenase (33). NPR is distinct from many other flavoproteins in that it contains one molecule each of FAD and FMN per polypeptide chain. NPR reduces NADPH in FMN. We have shown here a new function of NPR, as a regulator of HIF-1 activation and gene induction by hypoxia.
Despite various studies, thus far the molecular mechanisms responsible for oxygen sensing and downstream pathways uti-lized by the hypoxic signal are still poorly understood. Recent studies of oxygen sensing have explored the role of oxygen as an electron acceptor in a variety of nonmitochondrial redox systems involving electron transport and have focused on an NADPH-oxidase from phagocytes that catalyzes the production of hydrogen peroxide (12). We utilized hypoxic EPO induction to test the hypothesis that NADPH-oxidase is an oxygen sensor. Although we observed that DPIC as an NADPH-oxidase inhibitor inhibited EPO induction in Hep3B cells by hypoxia, NADPH-oxidase mRNA was not detected in Hep3B cells by RT-PCR. These results indicate that DPIC is probably not an inhibitor of NADPH-oxidase in Hep3B cells. Our data provide evidence that NPR is the target of DPIC.
We cloned cells that had low NPR expression (NPR Ϫ cells) to elucidate whether the target of DPIC is NPR. The EPO gene was not induced in the NPR Ϫ cells under hypoxic conditions, whereas the hypoxic EPO induction in the NPR Ϫ cells was normally induced by cobalt, which is known to induce EPO mRNA, as happens with hypoxia (32). These findings were consistent with our results, in which DPIC had little or no effect on EPO induction by cobalt. Therefore, we studied whether NPR influences either transportation of HIF-1 to the nucleus or its HRE binding. As expected, binding of HIF-1 to HRE of the EPO gene examined by EMSA and reporter gene assay was extensively reduced in Hep3B cells lipofected with pcDNA-aNPR. The same experiments with HeLa cells that expressed NADPH-oxidase produced the same results. These results suggest that NPR but not NADPH-oxidase is important for the activation of HIF-1 under hypoxic conditions in general.
Moreover, we found novel evidence that NPR is located on the plasma membrane. This was demonstrated by both the anti-NPR antibody, which did not cross-react with NADPHoxidase, and BSA-NADPH. Thus far it has been understood that NPR and P450 are located in the endoplasmic reticulum; however, in this study, anti-NPR IgG also inhibited the binding of HIF-1 to HRE of the EPO gene. An important function of NPR in the plasma membrane is the regulation of EPO induction and HIF-1 activation.
The activation of HIF-1 by hypoxia appears to be complex and to involve changes in protein stability, nuclear localization, DNA binding capability, and transcriptional activation function (34,35). In the presence of oxygen, HIF-1 is targeted for destruction by an ubiquitin-protein isopeptide ligase containing the von Hippel-Lindau tumor suppressor protein (pVHL) (36 -39). Cells lacking functional pVHL cannot degrade HIF-1 and thus overproduce HIF-1 protein encoded by HIF-1 target genes (5). Recent studies have indicated that proline hydroxylation governs HIF-1 turnover in the presence of oxygen (40 -42). Hydroxylation of proline is necessary for pVHL to bind to HIF-1␣ for degradation. Although a conserved HIF-VHL-prolyl hydroxylase pathway in Caenorhabditis elegans has been defined (43), human HIF prolyl hydroxylase has not been identified. We hypothesize that NPR might modify HIF prolyl hydroxylase activity through the reduction of an unknown factor, because we have observed that both HIF-1 degradation and DNA binding have diminished under hypoxic conditions in cells lacking NPR. If HIF prolyl hydroxylase directly senses oxygen and promotes pVHL-dependent ubiquitylation, hypoxia could induce the EPO gene in the NPR Ϫ cells. In addition, the inhibition of EPO induction by CO cannot be explained by this system, in which some hemeprotein works (21,22).
Recent studies have suggested that an increase in the VEGF levels was recognized in hypoxic disease. Tumors become hypoxic, and then hypoxia causes the expression of VEGF, which is regulated by HIF-1. More recently, Li et al. (44) found that NPR was induced in human bladder tumors. In this study, we found that the expression of NPR was well correlated with that of VEGF (Fig. 9). The HIF-1␣ protein level was also high in tissues that had a high expression of NPR. These findings are consistent with in vitro studies where NPR overexpression in Hep3B cells and HeLa S3 cells increased HRE binding activity by luciferase assay. These findings suggest that NPR is important for HIF-1␣ stabilization and VEGF expression in vivo.
In conclusion, NPR regulates gene expression under hypoxic conditions by modulating the HIF-1␣ activation and its HRE binding. We also found that NPR was present in the plasma membrane and had the function of regulating gene expression. We have presented evidence for the relation of NPR to HIF-1␣ expression in human bladder tissues and have suggested that NPR is an important factor in hypoxic expression of genes such as VEGF in vivo.