HIF-1 Regulates Cytochrome Oxidase Subunits to Optimize Efficiency of Respiration in Hypoxic Cells

O2 is the ultimate electron acceptor for mitochondrial respiration, a process catalyzed by cytochrome c oxidase (COX). In yeast, COX subunit composition is regulated by COX5a and COX5b gene transcription in response to high and low O2, respectively. Here we demonstrate that in mammalian cells, expression of the COX4-1 and COX4-2 isoforms is O2 regulated. Under conditions of reduced O2 availability, hypoxia-inducible factor 1 (HIF-1) reciprocally regulates COX4 subunit expression by activating transcription of the genes encoding COX4-2 and LON, a mitochondrial protease that is required for COX4-1 degradation. The effects of manipulating COX4 subunit expression on COX activity, ATP production, O2 consumption, and reactive oxygen species generation indicate that the COX4 subunit switch is a homeostatic response that optimizes the efficiency of respiration at different O2 concentrations. Thus, mammalian cells respond to hypoxia by altering COX subunit composition, as previously observed in yeast, but by a completely different molecular mechanism. O2 is the ultimate electron acceptor for mitochondrial respiration, a process catalyzed by cytochrome c oxidase (COX). In yeast, COX subunit composition is regulated by COX5a and COX5b gene transcription in response to high and low O2, respectively. Here we demonstrate that in mammalian cells, expression of the COX4-1 and COX4-2 isoforms is O2 regulated. Under conditions of reduced O2 availability, hypoxia-inducible factor 1 (HIF-1) reciprocally regulates COX4 subunit expression by activating transcription of the genes encoding COX4-2 and LON, a mitochondrial protease that is required for COX4-1 degradation. The effects of manipulating COX4 subunit expression on COX activity, ATP production, O2 consumption, and reactive oxygen species generation indicate that the COX4 subunit switch is a homeostatic response that optimizes the efficiency of respiration at different O2 concentrations. Thus, mammalian cells respond to hypoxia by altering COX subunit composition, as previously observed in yeast, but by a completely different molecular mechanism. The goal of oxidative phosphorylation is the transfer of electrons through a series of acceptor cytochromes in order to generate a proton gradient within the inner mitochondrial membrane. The potential energy of this gradient is used to synthesize ATP. O2 is the ultimate electron acceptor, resulting in the production of H2O in a process that is catalyzed by cytochrome c oxidase (COX; complex IV). This process is not completely efficient; electron transfer to O2 may occur at complex I or III, resulting in generation of reactive oxygen species (ROS) that oxidize cellular proteins, lipids, and nucleic acids and, by doing so, cause cell dysfunction or death. COX, which is located in the inner mitochondrial membrane, is a dimer in which each monomer consists of 13 subunits (Tsukihara et al., 1996Tsukihara T. Aoyama H. Yamashita E. Tomizaki T. Yamaguchi H. Shinizawa-Itoh K. Nakashima R. Yaono R. Yoshikawa S. The whole structure of the 13-subunit oxidized cytochrome c oxidase at 2 A.Science. 1996; 272: 1136-1144Crossref PubMed Scopus (1844) Google Scholar). Subunits I, II, and III are encoded by the mitochondrial genome, constitute the catalytic core of the enzyme, and are highly conserved in eukaryotes. The crystal structure of bovine COX revealed that subunit IV (COX4) interacts with both COX1 and COX2 (Tsukihara et al., 1996Tsukihara T. Aoyama H. Yamashita E. Tomizaki T. Yamaguchi H. Shinizawa-Itoh K. Nakashima R. Yaono R. Yoshikawa S. The whole structure of the 13-subunit oxidized cytochrome c oxidase at 2 A.Science. 1996; 272: 1136-1144Crossref PubMed Scopus (1844) Google Scholar). In mammalian cells, the first step of COX assembly is the association of COX1 with COX4 (Nijtmans et al., 1998Nijtmans L.G. Taanman J.W. Muijsers A.O. Speijer D. Van den Bogert C. Assembly of cytochrome-c oxidase in cultured human cells.Eur. J. Biochem. 1998; 254: 389-394Crossref PubMed Scopus (199) Google Scholar). Within the complex, COX4 binds ATP, leading to allosteric inhibition of COX activity at high ATP/ADP ratios and demonstrating a regulatory role for COX4 (Napiwotzki and Kadenbach, 1998Napiwotzki J. Kadenbach B. Extramitochondrial ATP/ADP ratios regulate cytochrome c oxidase activity via binding to the cytosolic domain of subunit IV.Biol. Chem. 1998; 379: 335-339Crossref Scopus (78) Google Scholar). Molecular modeling indicates that the spatial relationships of yeast homologs of COX1, COX2, and COX4 (which in yeast is designated COX5) are similar to the mammalian complex (Burke and Poyton, 1998Burke P.V. Poyton R.O. Structure/function of oxygen-regulated isoforms in cytochrome c oxidase.J. Exp. Biol. 1998; 201: 1163-1175Crossref Google Scholar). The yeast genome contains two genes encoding COX5 proteins, which show reciprocal patterns of expression: at high O2 concentrations, COX5a transcription is activated and COX5b transcription is repressed, whereas at low O2 concentrations, COX5a transcription is inactivated and COX5b transcription is derepressed (Lowry and Zitomer, 1984Lowry C.V. Zitomer R.S. Oxygen regulation of anaerobic and aerobic genes mediated by a common factor in yeast.Proc. Natl. Acad. Sci. USA. 1984; 81: 6129-6133Crossref Scopus (72) Google Scholar, Forsburg and Guarente, 1989Forsburg S.L. Guarente L. Communication between mitochondria and the nucleus in regulation of cytochrome genes in the yeast Saccharomyces cerevisiae.Annu. Rev. Cell Biol. 1989; 5: 153-180Crossref Scopus (128) Google Scholar, Kwast et al., 1998Kwast K.E. Burke P.V. Poyton R.O. Oxygen sensing and the transcriptional regulation of oxygen-responsive genes in yeast.J. Exp. Biol. 1998; 201: 1177-1195Crossref Google Scholar). Substitution of COX5b for COX5a increases the turnover number (TNmax) of COX by increasing the rate of electron transfer from heme a to the binuclear reaction center within COX1 (Waterland et al., 1991Waterland R.A. Basu A. Chance B. Poyton R.O. The isoforms of yeast cytochrome c oxidase alter the in vivo kinetic properties of the holoenzyme.J. Biol. Chem. 1991; 266: 4180-4186Abstract Full Text PDF Google Scholar, Allen et al., 1995Allen L.A. Zhao X.-J. Caughey W. Poyton R.O. Isoforms of yeast cytochrome c oxidase subunit V affect the binuclear reaction center and alter the kinetics of interaction with the isoforms of yeast cytochrome c.J. Biol. Chem. 1995; 270: 110-118Crossref Scopus (80) Google Scholar). Mammals possess paralagous genes encoding alternative isoforms of COX6, COX7a, COX8, and COX4 (Hutteman et al., 2001Hutteman M. Kadenbach B. Grossman L.I. Mammalian subunit IV isoforms of cytochrome c oxidase.Gene. 2001; 267: 111-123Crossref Scopus (133) Google Scholar). Although tissue-specific expression of these genes has been demonstrated, neither molecular nor physiological regulatory mechanisms have been elucidated. Hypoxia-inducible factor 1 (HIF-1) is a transcriptional activator that functions as a master regulator of oxygen homeostasis in all metazoan species. Previous studies have demonstrated that HIF-1 transactivates genes encoding glucose transporters and glycolytic enzymes in response to reduced O2 availability (Semenza et al., 1996Semenza G.L. Jiang B.-H. Leung S.W. Passantino R. Concordet J.-P. Maire P. Giallongo A. Hypoxia response elements in the aldolase A, enolase 1, and lactate dehydrogenase A gene promoters contain essential binding sites for hypoxia-inducible factor 1.J. Biol. Chem. 1996; 271: 32529-32537Crossref PubMed Scopus (1250) Google Scholar, Iyer et al., 1998Iyer N.V. Kotch L.E. Agani F. Leung S.W. Laughner E. Wenger R.H. Gassmann M. Gearhart J.D. Lawler A.M. Yu A.Y. Semenza G.L. Cellular and developmental control of O2 homeostasis by hypoxia-inducible factor 1α.Genes Dev. 1998; 12: 149-162Crossref PubMed Scopus (1933) Google Scholar, Seagroves et al., 2001Seagroves T.N. Ryan H.E. Lu H. Wouters B.G. Knapp M. Thibault P. Laderoute K. Johnson R.S. Transcription factor HIF-1 is a necessary mediator of the Pasteur effect in mammalian cells.Mol. Cell. Biol. 2001; 21: 3436-3444Crossref PubMed Scopus (442) Google Scholar). HIF-1 is a heterodimeric protein composed of a constitutively expressed HIF-1β subunit and an O2-regulated HIF-1α or HIF-2α subunit (Wang et al., 1995Wang G.L. Jiang B.-H. Rue E.A. Semenza G.L. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension.Proc. Natl. Acad. Sci. USA. 1995; 92: 5510-5514Crossref PubMed Scopus (4701) Google Scholar, Jiang et al., 1996Jiang B.-H. Rue E. Wang G.L. Roe R. Semenza G.L. Dimerization, DNA binding, and transactivation properties of hypoxia-inducible factor 1.J. Biol. Chem. 1996; 271: 17771-17778Crossref PubMed Scopus (855) Google Scholar, Tian et al., 1997Tian H. McKnight S.L. Russell D.W. Endothelial PAS domain protein 1 (EPAS1), a transcription factor selectively expressed in endothelial cells.Genes Dev. 1997; 11: 72-82Crossref PubMed Scopus (1010) Google Scholar, Wiesener et al., 1998Wiesener M.S. Turley H. Allen W.E. Willam C. Eckardt K.U. Talks K.L. Wood S.M. Gatter K.C. Harris A.L. Pugh C.W. et al.Induction of endothelial PAS domain protein-1 by hypoxia: characterization and comparison with hypoxia-inducible factor-1α.Blood. 1998; 92: 2260-2268Crossref PubMed Google Scholar). Under aerobic conditions, HIF-1α and HIF-2α are subjected to prolyl hydroxylation by enzymes that utilize O2 as a substrate (Ivan et al., 2001Ivan M. Kondo K. Yang H. Kim W. Valiando J. Ohh M. Salic A. Asara J.M. Lane W.S. Kaelin Jr., W.G. HIF-α targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing.Science. 2001; 292: 464-468Crossref PubMed Scopus (3600) Google Scholar, Jaakkola et al., 2001Jaakkola P. Mole D.R. Tian Y.M. Wilson M.I. Gielbert J. Gaskell S.J. Kriegsheim A.V. Hebestreit H.F. Mukherji M. Schofield C.J. Maxwell P.H. Pugh C.W. Ratcliffe P.J. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation.Science. 2001; 292: 468-472Crossref PubMed Scopus (4128) Google Scholar, Yu et al., 2001Yu F. White S.B. Zhao Q. Lee F.S. HIF-1α binding to VHL is regulated by stimulus-sensitive proline hydroxylation.Proc. Natl. Acad. Sci. USA. 2001; 98: 9630-9635Crossref PubMed Scopus (616) Google Scholar). Hydroxylation is required for binding of the von Hippel-Lindau (VHL) protein (Maxwell et al., 1999Maxwell P.H. Wiesener M.S. Chang G.W. Clifford S.C. Vaux E.C. Cockman M.E. Wykoff C.C. Pugh C.W. Maher E.R. Ratcliffe P.J. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis.Nature. 1999; 399: 271-275Crossref PubMed Scopus (3863) Google Scholar), which targets HIF-1α and HIF-2α for ubiquitination and proteasomal degradation (Salceda and Caro, 1997Salceda S. Caro J. Hypoxia-inducible factor 1α (HIF-1α) protein is rapidly degraded by the ubiquitin-proteasome system under normoxic conditions. Its stabilization by hypoxia depends on redox-induced changes.J. Biol. Chem. 1997; 272: 22642-22647Crossref PubMed Scopus (1334) Google Scholar). The interaction of HIF-1α with the coactivators CBP and p300 is blocked by O2-dependent asparaginyl hydroxylation mediated by FIH-1 (Lando et al., 2002Lando D. Peet D.J. Gorman J.J. Whelan D.A. Whitelaw M.L. Bruick R.K. FIH-1 is an asparaginyl hydroxylase enzyme that regulates the transcriptional activity of hypoxia-inducible factor.Genes Dev. 2002; 16: 1466-1471Crossref PubMed Scopus (1115) Google Scholar). Under hypoxic conditions, the rate of hydroxylation declines, leading to HIF-1 activation and thus providing a mechanism by which changes in oxygenation are transduced to the nucleus as changes in gene expression. In this study, we tested the hypothesis that in mammalian cells the expression of COX4-1 and COX4-2 is O2 regulated through the activity of HIF-1. To investigate whether COX4-2 mRNA expression is O2 regulated in human cells, Hep3B (liver), HeLa (uterus), Hct116 (colon), and A594 (lung) cells were incubated at 20% or 1% O2 for 24 hr. COX4-2 mRNA expression was induced under hypoxic conditions in all four cell lines as determined by reverse transcriptase (RT)-PCR (Figure 1A). Hypoxia-induced expression of COX4-2 mRNA was also observed in primary cultures of mouse pulmonary artery smooth muscle cells (mPASMC; Figure 1B). COX4-2 protein expression was induced in hypoxic HeLa cells as determined by immunoblot assay (Figure 1C) using an antibody that specifically recognizes COX4-2 (Figure S1). COX4-1 mRNA expression was not O2 regulated in mPASMC, HeLa, or Hep3B cells as determined by both conventional RT-PCR (Figures 1B and S2) and quantitative real-time RT-PCR (qrtPCR; Figure 1D). COX4-2, but not COX4-1, mRNA expression was induced in HeLa and Hep3B cells exposed to CoCl2, desferrioxamine (DFX), or dimethyloxalylglycine (DMOG), which are inducers of HIF-1α expression that inhibit hydroxylase activity (Epstein et al., 2001Epstein A.C. Gleadle J.M. McNeill L.A. Hewitson K.S. O'Rourke J. Mole D.R. Mukherji M. Metzen E. Wilson M.I. Dhanda A. et al.C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation.Cell. 2001; 107: 43-54Abstract Full Text Full Text PDF PubMed Scopus (2441) Google Scholar, Jaakkola et al., 2001Jaakkola P. Mole D.R. Tian Y.M. Wilson M.I. Gielbert J. Gaskell S.J. Kriegsheim A.V. Hebestreit H.F. Mukherji M. Schofield C.J. Maxwell P.H. Pugh C.W. Ratcliffe P.J. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation.Science. 2001; 292: 468-472Crossref PubMed Scopus (4128) Google Scholar) and thereby increase the expression of mRNAs encoded by HIF-1 target genes such as GLUT1 (Figures 1D and S2). COX4-2 mRNA expression was increased in liver and lung tissue from mice exposed to 10% O2 for 3 weeks relative to control mice exposed to room air (21% O2), indicating that COX4-2 induction is a physiological response to hypoxia (Figure 1E). Taken together the data shown in Figure 1 demonstrate that expression of the COX4I2 gene, which encodes COX4-2, is induced in response to hypoxia and suggest that this response may be regulated by HIF-1. We searched the human COX4I2 gene for matches to a consensus sequence, which was generated from hypoxia-response elements (HREs) in the EPO, ALDA, Vegf, CP, ID2, BNIP3, and NOXA genes (Semenza et al., 1996Semenza G.L. Jiang B.-H. Leung S.W. Passantino R. Concordet J.-P. Maire P. Giallongo A. Hypoxia response elements in the aldolase A, enolase 1, and lactate dehydrogenase A gene promoters contain essential binding sites for hypoxia-inducible factor 1.J. Biol. Chem. 1996; 271: 32529-32537Crossref PubMed Scopus (1250) Google Scholar, Shima et al., 1996Shima D.T. Kuroki M. Deutsch U. Ng Y.S. Adamis A.P. D'Amore P. The mouse gene for vascular endothelial growth factor. Genomic structure, definition of the transcriptional unit, and characterization of transcriptional and post-transcriptional regulatory sequences.J. Biol. Chem. 1996; 271: 3877-3883Abstract Full Text Full Text PDF PubMed Scopus (307) Google Scholar, Bruick, 2000Bruick R.K. Expression of the gene encoding the proapoptotic Nip3 protein is induced by hypoxia.Proc. Natl. Acad. Sci. USA. 2000; 97: 9082-9087Crossref PubMed Scopus (634) Google Scholar, Mukhopadhyay et al., 2000Mukhopadhyay C.K. Mazumder B. Fox P.L. Role of hypoxia-inducible factor-1 in transcriptional activation of ceruloplasmin by iron deficiency.J. Biol. Chem. 2000; 275: 21048-21054Crossref PubMed Scopus (210) Google Scholar, Kim et al., 2004Kim J.Y. Ahn H.J. Ryu J.H. Suk K. Park J.H. BH3-only protein Noxa is a mediator of hypoxic cell death induced by hypoxia-inducible factor 1α.J. Exp. Med. 2004; 199: 113-124Crossref Scopus (227) Google Scholar, Lofstedt et al., 2004Lofstedt T. Jogi A. Sigvardsson M. Gradin K. Poellinger L. Pahlman S. Axelson H. Induction of ID2 expression by hypoxia-inducible factor-1: a role in dedifferentiation of hypoxic neuroblastoma cells.J. Biol. Chem. 2004; 279: 39223-39231Crossref Scopus (117) Google Scholar) and consisted of the core HIF-1-binding site 5′-RCGTG-3′ followed, after 1–8 nucleotides, by a CAC sequence (consensus, 5′-CACAG-3′). Candidate HREs were identified in the 5′-flanking region (HRE1) and intron 1 (HRE2) of COX4I2 (Figure 2A). A 2.1 kb DNA fragment spanning these HREs was inserted into the promoterless luciferase reporter plasmid pGL2-Basic. A second plasmid containing 2.7 kb spanning the analogous region of the human COX4I1 gene, which encodes COX4-1, was also constructed. The reporter plasmids were transfected into human embryonic kidney 293 cells, which were incubated under four conditions: control nonhypoxic conditions (20% O2); hypoxia (1% O2); and nonhypoxic conditions in the presence of 100 μM CoCl2 or DFX. Analysis of cell lysates 24 hr later revealed that COX4I2 promoter activity was induced by hypoxia, CoCl2, or DFX, whereas COX4I1 promoter activity was unaffected (Figure 2B). These results are consistent with the analyses of COX4-1 and COX4-2 mRNA expression (Figure 1) and indicate that hypoxia selectively induces COX4I2 gene transcription. Immunoblot assays revealed that HIF-1α protein expression was induced by hypoxia in HeLa and 293 cells, whereas HIF-2α expression was poorly induced in hypoxic 293 cells (Figure 2C). COX4I2 promoter activity in HeLa cells was dramatically increased by hypoxia or cotransfection of an expression vector encoding HIF-2α, whereas cotransfection of an expression vector encoding HIF-1α had a more modest effect (Figures 2D and S4) despite high levels of HIF-1α protein in transfected cells (Figure S3). In 293 cells, COX4I2 promoter activity was greatly increased by cotransfection of a HIF-2α expression vector, whereas hypoxia or cotransfection of HIF-1α expression vector had a more modest effect (Figures 2E and S4). These data suggest that HIF-2α is a more potent transactivator of the COX4I2 promoter than HIF-1α, and the immunoblot data (Figure 2C) suggest that the more modest effect of hypoxia on COX4I2 promoter activity in 293 cells is due to the limited induction of HIF-2α expression. Analysis of deletion constructs containing only HRE1 or HRE2 suggested that both sequences are required for maximal induction of COX4I2 promoter activity in response to hypoxia or overexpression of HIF-1α or HIF-2α (Figures 2D, 2E, and S4). The data presented in Figure 2 suggest that both HIF-1α and HIF-2α can contribute to increased COX4I2 promoter activity in hypoxic cells. To further test this hypothesis, HeLa and 293 cells were cotransfected with pCEP4-HIF-1αDN, which encodes a dominant-negative form of HIF-1α that competes with endogenous HIF-1α and HIF-2α for dimerization with HIF-1β but results in transcriptionally inactive heterodimers (Jiang et al., 1996Jiang B.-H. Rue E. Wang G.L. Roe R. Semenza G.L. Dimerization, DNA binding, and transactivation properties of hypoxia-inducible factor 1.J. Biol. Chem. 1996; 271: 17771-17778Crossref PubMed Scopus (855) Google Scholar). HIF-1αDN inhibited hypoxia-induced COX4I2 promoter activity in a dose-dependent manner in both 293 and HeLa cells (Figure 3A), which is similar to its effect on reporter plasmid p2.1 (Figure S5), which contains a HIF-1-dependent HRE from the ENO1 gene (Semenza et al., 1996Semenza G.L. Jiang B.-H. Leung S.W. Passantino R. Concordet J.-P. Maire P. Giallongo A. Hypoxia response elements in the aldolase A, enolase 1, and lactate dehydrogenase A gene promoters contain essential binding sites for hypoxia-inducible factor 1.J. Biol. Chem. 1996; 271: 32529-32537Crossref PubMed Scopus (1250) Google Scholar). To determine the effect of specifically interfering with HIF-1α or HIF-2α expression, HeLa and 293 cells were transfected with: empty vector (EV); expression vector encoding short hairpin RNA (shRNA) directed against HIF-1α (Shr-HIF-1α) or HIF-2α (Shr-HIF-2α); or vector encoding a scrambled negative control sequence (Scr). Immunoblot assays revealed that target protein expression was reduced in Shr-transfected cells (Figure S6). The cells were cotransfected with reporter plasmid p2.1 or the COX4I2 promoter reporter. Shr-HIF-1α significantly inhibited hypoxia-induced p2.1 in both HeLa and 293 cells, whereas Shr-HIF-2α had no significant effect on hypoxia-induced p2.1 activity in HeLa or 293 cells (Figure S7). In contrast, Shr-HIF-2α significantly inhibited hypoxia-induced COX4I2 promoter activity in HeLa cells but not in 293 cells (Figure 3B), which have low HIF-2α expression (Figure 2C). These data indicate that: different gene promoters (ENO1 versus COX4I2) are differentially regulated by HIF-1α versus HIF-2α and that regulation of the same gene (COX4I2) in different cell types (293 versus HeLa) may vary depending upon the extent to which HIF-2α is induced by hypoxia. To demonstrate that the HIF-1-binding sites in HRE1 and HRE2 were required for hypoxia-induced COX4I2 promoter activity, CG→AA mutations that destroy the HIF-1-binding sites were introduced into the HREs (Figure 2A). Mutation of HRE2 resulted in a partial loss of activity, whereas mutation of both HREs resulted in a complete loss of hypoxia-induced COX4I2 promoter activity in both HeLa and 293 cells (compare Figure 3C with Figures 2D and 2E). To demonstrate that HIF-1 binds to these sequences within living cells, chromatin immunoprecipitation (ChIP) assays were performed using DFX-treated HeLa and 293 cells. Binding of HIF-2α to HRE1 and HRE2 in HeLa cells and binding of HIF-1α to HRE2 in 293 cells was demonstrated when PCR was performed using a single set of flanking primers (Figures 3D and 3E). A more sensitive but nonquantitative nested PCR was required to demonstrate the binding of HIF-1α to HRE2 in HeLa cells. No PCR product was detected when nonspecific IgG was used instead of specific antibody against HIF-1α or HIF-2α for ChIP assay of DFX-treated cells or when chromatin from untreated cells was used for ChIP with antibodies against HIF-1α or HIF-2α (data not shown). The data presented in Figure 1, Figure 2, Figure 3 show that COX4I2 expression is induced by HIF-1 and that HIF-2α-containing heterodimers preferentially, but not exclusively, bind to two HREs within the 5′-flanking region and intron 1 to activate gene transcription. We next investigated whether HIF-1 regulates COX4-1 protein levels using a subunit-specific antibody (Figure S1). HeLa and Hct116 cells were cultured for 24 or 48 hr in the presence or absence of CoCl2. Exposure of the cells to CoCl2 induced increased HIF-1α expression as expected but also resulted in a dramatic reduction in COX4-1 protein levels, whereas β-actin levels were unaffected (Figure 4A). A time course analysis of CoCl2-treated HeLa cells revealed a gradual reduction in COX4-1 protein levels over 24 hr, whereas β-actin protein levels were unchanged (Figure 4B). In addition to CoCl2, the HIF-1 hydroxylase inhibitors DMOG and dihydroxybenzoate (DHB) increased HIF-1α and decreased COX4-1 protein levels (Figure 4C). To investigate whether the effect of hydroxylase inhibitors on COX4-1 was due to their induction of HIF-1 activity, we utilized two other HIF-1 gain-of-function (GOF) models. First, we analyzed RCC4 human renal carcinoma cells in which VHL loss-of-function (LOF) results in constitutive high-level expression of HIF-1α and HIF-2α (Hu et al., 2003Hu C.J. Wang L.Y. Chodosh L.A. Keith B. Simon M.C. Differential roles of hypoxia-inducible factor 1α (HIF-1α) and HIF-2α in hypoxic gene regulation.Mol. Cell. Biol. 2003; 23: 9361-9374Crossref PubMed Scopus (956) Google Scholar). In these cells, COX4-1 protein levels were low even in the absence of CoCl2 treatment, whereas in a subclone stably transfected with a VHL expression vector to inhibit HIF-1α and HIF-2α expression under nonhypoxic conditions, COX4-1 protein levels were increased in the absence but not in the presence of CoCl2 (Figure 4D). In a second HIF-1 GOF model, we analyzed COX4-1 levels in HeLa cells that were infected with an adenovirus encoding either β-galactosidase (AdLacZ) or a constitutively active form of HIF-1α (AdCA5) that activates target gene expression under nonhypoxic conditions (Kelly et al., 2003Kelly B.D. Hackett S.F. Hirota K. Oshima Y. Cai Z. Berg-Dixon S. Rowan A. Yan Z. Campochiaro P.A. Semenza G.L. Cell type-specific regulation of angiogenic growth factor gene expression and induction of angiogenesis in nonischemic tissue by a constitutively active form of hypoxia-inducible factor 1.Circ. Res. 2003; 93: 1074-1081Crossref PubMed Scopus (465) Google Scholar). COX4-1 levels were dramatically reduced by infection with AdCA5 as compared to AdLacZ (Figure 4E). Thus, HIF-1 GOF leads to loss of COX4-1 protein expression. We also investigated the effect of HIF-1 LOF using two independent experimental approaches. In wild-type mouse embryo fibroblasts, which express HIF-1α but not HIF-2α, COX4-1 protein levels were reduced in response to hypoxia, whereas in HIF-1α null cells, no induction of HIF-1α or loss of COX4-1 protein occurred (Figure 4F). In the second LOF approach, expression of Shr-HIF-1α in HeLa cells partially inhibited the induction of HIF-1α in response to CoCl2 and partially inhibited the loss of COX4-1 protein in response to CoCl2, whereas β-actin protein levels were unaffected by CoCl2 or Shr-HIF-1α (Figure 4G). To investigate the mechanism by which COX4-1 was degraded in response to hypoxia, cells were treated with the proteasome inhibitor MG132, which blocks HIF-1α degradation (Figure 4H). However, CoCl2-induced degradation of COX4-1 was not blocked by MG132. In pulse-chase assays, degradation of COX4-1 was dramatically increased by CoCl2 (Figure 4I). The results presented in Figure 4 demonstrate that COX4-1 degradation is induced in cells exposed to hypoxia or CoCl2 by a HIF-1-dependent, proteasome-independent mechanism. The activity of the mitochondrial protease LON is induced by hypoxia, ischemia, or endoplasmic reticulum stress (Hori et al., 2002Hori O. Ichinoda F. Tamatani T. Yamaguchi A. Sato N. Ozawa K. Kitao Y. Miyazaki M. Harding H.P. Ron D. et al.Transmission of cell stress from endoplasmic reticulum to mitochondria: enhanced expression of Lon protease.J. Cell Biol. 2002; 157: 1151-1160Crossref PubMed Scopus (161) Google Scholar). We hypothesized that LON is responsible for hypoxia-induced degradation of COX4-1. LON mRNA expression was induced in: HeLa cells exposed to 1% O2 for 24 or 48 hr (Figures 5A and S8); heart, lung, and liver of hypoxic mice (Figure 5B); and AdCA5-infected cells (Figure 5C). LON mRNA expression was induced by hypoxia in RCC4-VHL cells but was constitutively expressed in the VHL-deficient parental RCC4 cells (Figure 5D, left panels). RCC4 cells stably transfected with expression vector encoding Shr-HIF-1α expressed lower levels of LON mRNA than did RCC4 cells transfected with empty vector (Figure 5D, right panels). Compared to scrambled control shRNA, shRNA against LON mRNA blocked degradation of COX4-1 protein that was induced by hypoxia or CoCl2 (Figure 5E). Analysis of the human LON gene promoter revealed the presence of five potential HIF-1-binding sites within 0.6 kb 5′ to the transcription start site (Figure S9). Activity of the full-length promoter was markedly induced by hypoxia, and deletion analysis suggested that the middle and proximal, but not distal, candidate HIF-1 sites contributed to the response in HeLa (Figure 5F) and 293 (Figure S10) cells. These results were corroborated by ChIP assays, which demonstrated the binding of protein complexes containing HIF-1α and HIF-2α to the proximal and middle, but not distal, HIF-1 sites in DFX-treated, but not in untreated, HeLa cells (Figure 5G). The data presented in Figure 5 indicate that HIF-1 induces LON expression and that LON is required for the degradation of COX4-1 protein in hypoxic cells. Little is known regarding the functional properties of the human COX4-1 and COX4-2 proteins. We analyzed the effect of increasing (Figure S11) or decreasing (Figures S12 and S13) the levels of these proteins in 293T cells cultured under nonhypoxic and hypoxic conditions. 293T cells were transiently transfected with expression vectors encoding either FLAG- or V5-epitope-tagged COX4-1 or COX4-2 protein. Immunoblot assays revealed that all four proteins were expressed at high levels and localized to the mitochondrial fraction of cell lysates (Figure S14), and all subsequent experiments were performed with FLAG-tagged proteins. Compared to cells transfected with EV, nonhypoxic cells transfected with vector encoding COX4-1 manifested modest, but significant, increases in COX activity and O2 consumption (COX and JO2, respectively, in Figure 6A, left panels), which is consistent with the already high level of COX4-1 in these cells (Figure S12). COX4-2 GOF also increased COX and JO2 in nonhypoxic cells. COX4-2, but not COX4-1, increased COX and JO2 in hypoxic cells. Immunoblot assay revealed that overexpression of FLAG-COX4-1 successfully overwhelmed the proteolytic machinery in hypoxic cells (Figure S11), thus ruling out the possibility that the ineffectiveness of FLAG-COX4-1 to increase COX and JO2 in hypoxic cells was due to its degradation. ATP levels were increased in nonhypoxic cells with either COX4-1 or COX4-2 GOF, but only COX4-2 GOF increased ATP levels in hypoxic cells. COX4-1 LOF led to a reduction in COX, JO2, and ATP in nonhypoxic cells but had no effect on these indices in hypoxic cells (Figure 6A, right panels). COX4-2 LOF had no effect in nonhypoxic cells but led to reduced COX, JO2, and ATP in hypoxic cells. We hypothesized that COX4-1 and COX4-2 may be necessary to optimize electron transfer chain activity under nonhypoxic and hypoxic conditions, respectively. Nonoptimal efficiency may result in decreased ATP production and/or increased production of H2O2 due