This page has been archived and is no longer updated
Oxygen sensing by HIF hydroxylases.
Author: C. J. Schofield
Keywords
Keywords for this Article
Add keywords to your Content
Save
|
Cancel
Share
|
Cancel
Revoke
|
Cancel
Rate & Certify
Rate Me...
Rate Me
!
Comment
Save
|
Cancel
Flag Inappropriate
The Content is
Objectionable
Explicit
Offensive
Inaccurate
Comment
Flag Content
|
Cancel
Delete Content
Reason
Delete
|
Cancel
Close
Full Screen
"REVIEWS The possibility that cells have specific interfaces with molecular oxygen that have a prime function in bio- logical control has long interested biologists. Specific ?oxygen-sensing? mechanisms have been defined in bacteria and yeast, but, until recently, have remained elusive in higher organisms. Studies have now identi- fied an unexpectedly direct link between the avail- ability of oxygen and an important transcriptional cascade that regulates many responses to hypoxia in higher organisms, including humans. Molecular dis- section of one of the most striking homeostatic responses to hypoxia ? the induction of the haematopoietic growth factor erythropoietin ? led to the discovery of the transcription factor hypoxia- inducible factor (HIF) 1 .Unexpectedly, it was found that this system operates in essentially all mammalian cells irrespective of their relevance to erythropoietin production 2 ,and that it directs many other responses to hypoxia. The same hypoxic responses can be induced by iron chelators or cobaltous ions, distinc- tive properties that led to the proposal of a ferropro- tein oxygen sensor, which was originally thought to be a haem protein. Work on the HIF signalling path- way has now shown that its transcriptional activity is regulated by the post-translational hydroxylation of specific residues. Hydroxylation is catalysed by a set of oxygen-dependent enzymes that belong to the 2-oxoglutarate-dependent-oxygenase superfamily, which are, in fact, non-haem, Fe 2+ -dependent enzymes (BOX 1).Here,we review recent findings and their implications for our understanding of cellular responses to hypoxia, and discuss the possibility of a wider role for post-translational hydroxylation in signalling. Hypoxia-inducible factor HIF is a heterodimeric transcription factor that is com- posed of two basic helix?loop?helix proteins ? HIF? and HIF? ?ofthe PAS FAMILY (PER, AHR, ARNT and SIM family) 3 .The HIF?/? dimer binds to a core DNA motif (G/ACGTG) in hypoxia-response elements (HREs) that are associated with a broad range of tran- scriptional targets (for reviews, see REFS 4?8). These target genes are centrally involved in both systemic responses to hypoxia, such as ANGIOGENESIS and ERYTHROPOIESIS, and in cellular responses, such as alterations in glucose/energy metabolism (see BOX 2 and FIG. 1;for a fully referenced version of this figure, see online supplementary infor- mation S1 (figure)). The HIF? subunit, which is identi- cal to ARNT, is a constitutive nuclear protein that has further roles in transcription. By contrast, the levels of HIF? subunits are highly inducible by hypoxia. There are three closely related forms of HIF?,each ofwhich is encoded by a distinct gene locus. HIF1? and HIF2? have a similar domain architecture and are regulated in a similar manner. HIF3? is less closely related and its reg- ulation is less well understood. HIF1? and HIF2? undergo proteolytic regulation that is dependent on the presence of two indepen- dently functioning oxygen-dependent degradation OXYGEN SENSING BY HIF HYDROXYLASES Christopher J. Schofield* and Peter J. Ratcliffe ? The transcription factor HIF (hypoxia-inducible factor) has a central role in oxygen homeostasis in animals ranging from nematode worms to man. Recent studies have shown that this factor is regulated by an unprecedented signalling mechanism that involves post-translational hydroxylation. This hydroxylation is catalysed by a set of non-haem, Fe 2+ -dependent enzymes that belong to the 2-oxoglutarate-dependent-oxygenase superfamily. The absolute requirement of these enzymes for molecular oxygen has provided new insights into the way cells sense oxygen. NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 5 | MAY 2004 | 343 *The Oxford Centre for Molecular Sciences and The Dyson Perrins Laboratory, Department of Chemistry, South Parks Road, Oxford OX1 3QY, UK. ? The Henry Wellcome Building of Genomic Medicine, Roosevelt Drive, Oxford OX3 7BN, UK. Correspondence to P.J.R. e-mail: pjr@well.ox.ac.uk doi:10.1038/nrm1366 PAS FAMILY (period circadian protein (PER), aryl-hydrocarbon receptor (AHR), aryl-hydrocarbon- receptor nuclear translocator (ARNT) and single-minded protein (SIM)). A group of interacting and structurally related basic helix?loop?helix transcription factors. ANGIOGENESIS The growth and proliferation of new blood vessels from existing vasculature. ERYTHROPOIESIS The formation of erythrocytes (red blood cells). � 2004 Nature Publishing Group 344 | MAY 2004 | VOLUME 5 www.nature.com/reviews/molcellbio REVIEWS a similar manner by hypoxia, cobaltous ions and iron chelators. These findings indicated that each sequence must have a discrete, but mechanistically related, interaction with the regulatory signal pathway(s), and focused analysis on these sites. HIF regulation by protein hydoxylation Detailed biochemical and genetic analyses have now shown that oxygen responsiveness is indeed conveyed by separate interactions with each of the three HIF? regions that transfer responsiveness to hypoxia. And, in each case, the process is similar and involves the oxygen- dependent enzymatic hydroxylation of specific residues (FIG. 2).Although other post-translational modifications of HIF? subunits (including phosphorylation and acetylation) have been defined, their interface with oxy- gen-sensitive pathways remains unclear. This review therefore focuses on the newly defined hydroxylation pathways. Prolyl hydroxylation. Hydroxylation of specific prolyl residues at conserved sites in NODDD and CODDD (FIG. 2) regulates independent interactions with the von Hippel?Lindau tumour suppressor (pVHL) 12?15 . pVHL is the recognition component of a multi-component ubiquitin ligase (pVHL?elonginB?elonginC?Cul2?Rbx) that targets HIF? subunits for proteasomal proteolysis by the UBIQUITIN?PROTEASOME PATHWAY 16,17 .Hydroxylation increases the affinity of HIF? peptides for the pVHL?elonginB?elonginC (VBC) complex by at least domains (NODDD and CODDD), which are located in the central region of the molecule (FIG. 2). HIF1? and HIF2? also possess two transactivation domains ? an internal activation domain (NAD) that overlaps with the CODDD and a carboxy-terminal activation domain (CAD; FIG. 2). CAD is not directly involved in proteolytic regulation, but has an enabling role in HIF-mediated transcription through its oxygen-regu- lated association with the CH-1 (cysteine/histidine rich) domain of the transcriptional co-activator p300. In addition, HIF1? is reported to be subject to a fur- ther control that involves its nuclear/cytoplasmic localization, which is probably mediated by its active exclusion from the nucleus in the presence of oxygen. However, the precise mechanisms and regulatory sequences that are involved are not yet known (for reviews, see REFS 5,7,9). For HIF3?, at least one of the oxygen-regulated processes ? that is, proteolytic regulation that involves a central domain that closely resembles the HIF1/2? CODDD ? seems to be conserved 10 . HIF3? is also regulated by several types of alternative splic- ing. Most striking, is the production of a shortened form that is composed of the amino-terminal basic helix?loop?helix and PAS domains, and omits the CODDD and other carboxy-terminal sequences. This protein ? known as inhibitory PAS protein (IPAS) ? forms transcriptionally inactive heterodimers with HIF1? (REF. 11).Interestingly, the process of alternative splicing that forms the IPAS transcript is induced by hypoxia in certain tissues, although it is not yet clear whether this represents a separate interface with oxy- gen-sensitive pathways or a counter-regulatory response to the activation of HIF transcriptional activity. An important aspect of the analysis of the domain architecture of HIF? subunits was the identification of specific regions that, when fused to heterologous proteins, could independently transfer responsiveness to hypoxia. So far, three regions that correspond to the NODDD, CODDD and CAD have been shown to operate in this way (FIG. 2).It was found that the responses produced by these domains were affected in Box 1 | 2-Oxoglutarate-dependent enzymes and related enzymes 2-Oxoglutarate-dependent oxygenases and related enzymes are involved in a wide range of metabolic and signalling processes. In plants, they catalyse steps in the biosynthesis of signalling molecules such as ethylene and the gibberellins, as well as several steps in flavonoid biosynthesis. In both bacteria and fungi, they are involved in the biosynthesis of various medicinally important antibiotics, including those of the ?-lactam family such as penicillins 36 ,clavulanic acid 48,96 and carbapenems 97,98 . The oxidative reactions that are catalysed in the biosynthesis of these antibiotics are unusual and have little precedent in synthetic chemistry. In mammals, in addition to the hydroxylation of hypoxia-inducible factor, 2-oxoglutarate-dependent oxygenases catalyse procollagen-extracellular-matrix formation 99 , as well as important metabolic reactions such as the biosynthesis of carnitine 100 and the metabolism of dietary phytanic acid (a product of chlorophyll metabolism). In phytanic-acid metabolism, genetic lesions in phytanoyl CoA hydroxylase ? a 2-oxoglutarate-dependent oxygenase ? lead to Refsum?s disease 101 .Finally, these oxygenases are involved in DNA repair in organisms ranging from bacteria to humans 74,75,102,103 . Box 2 | The HIF transcriptional cascade So far, more than 40 hypoxia-inducible-factor (HIF) target genes have been characterized by the functional delineation of HIF binding to hypoxia-response elements in transcriptional control sequences. In keeping with the complexity of oxygen homeostasis in large animals, these HIF transcriptional targets are involved in diverse aspects of cellular and integrative physiology, including, for example, energy metabolism and cell growth and migration (see FIG. 1;for a fully referenced version of this figure, see online supplementary information S1 (figure)). In addition, large-scale gene-expression arrays have been used to define global responses to hypoxia, and they indicate that, in any given cell, several hundred genes are positively or negatively regulated by hypoxia 104?106 .Furthermore, the complement of genes that is regulated by hypoxia differs greatly between cell types, so that, overall, on the order of thousands of genes might be regulated in this way. Genetic studies in HIF?-subunit- or von- Hippel?Lindau-tumour-suppressor (pVHL)-defective cells, together with studies of the effects of nonspecific hydroxylase inhibitors such as cobaltous ions and iron chelators, have indicated that most of the responses to hypoxia are dependent on a functional HIF/pVHL/hydroxylase system. However, the extent to which the responses are the direct result of HIF-target- gene transcription, as opposed to indirect effects, is unclear. � 2004 Nature Publishing Group NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 5 | MAY 2004 | 345 REVIEWS Further analysis showed that the hydroxylation occurs on the ?-carbon 23 . NMR studies of a non-hydroxy- lated CAD polypeptide bound to CH-1 showed that Asn803 of HIF1? is part of an ?-helix that is deeply buried in the molecular interface 24,25 .These findings indicate how ?-hydroxylation of Asn803 prevents HIF1? binding to CH-1: it might disrupt both the hydrophobic interactions between the molecules and the ?-helix that is formed by CAD in this complex. Like prolyl hydroxylation, HIF asparaginyl hydroxyla- tion is also catalysed by a member of the 2-oxoglu- tarate-dependent-oxygenase superfamily, but one that is significantly different in sequence to the PHDs 26,27 . It was originally identified as a protein that binds HIF and was named FIH (factor inhibiting HIF) 28 . Asparaginyl hydroxylation therefore provides a sec- ond oxygen-regulated mechanism by which HIF? molecules that escape the prolyl-hydroxylation/degra- dation pathway are prevented from activating tran- scription through the p300 co-activator (FIG. 3). Genetic studies have clearly shown the non-redun- dant operation of each step in the physiological regula- tion of HIF. In Caenorhabditis elegans and Drosophila melanogaster, there is a single HIF prolyl hydroxylase, three orders of magnitude. X-ray crystallography of a hydroxylated HIF1? CODDD peptide bound to VBC showed that the exquisite discrimination between hydroxylated and non-hydroxylated sequences is medi- ated by two optimized hydrogen bonds, which are formed between the alcohol of the hydroxylated proline and two residues of pVHL (Ser111 and His115) 18,19 .In hypoxia, prolyl hydroxylation is suppressed, which allows the HIF? subunit to escape pVHL-mediated destruction and therefore to accumulate to high levels. Three closely related 2-oxoglutarate-dependent oxyge- nases with the capacity to catalyse HIF prolyl hydroxyla- tion ? known as prolyl hydroxylase domain (PHD)1, PHD2 and PHD3 ? have been identified 20,21 .They have an absolute requirement for molecular oxygen as a co-substrate and therefore provide a direct link between oxygen availability and the regulation of HIF (FIG. 3). Asparaginyl hydroxylation. Analysis of the CADs of human HIF1? and HIF2? showed that the oxygen- dependent association of this domain with the CH-1 domain of the p300 co-activator is blocked by the hydroxylation of a specific asparaginyl residue (Asn803 in human HIF1?; Asn851 in human HIF2?) 22 . UBIQUITIN?PROTEASOME PATHWAY A system of selective, energy (ATP)-consuming protein degradation that involves the linking of ubiquitins to specific proteins and the subsequent targeting of these polyubiquitylated proteins to the 26S proteasome (a multi- catalytic protease). Growth and apoptosis Insulin-like growth-factor-binding protein-1 Nip3 Endoglin Wilms? tumour suppressor ?-Fetoprotein (negative regulation) Calcitonin-receptor-like receptor Cell migration Chemokine receptor CXCR4 c-Met Transcriptional regulation DEC1 and DEC2 ETS1 p35srj Virus related Retrotransposon VL30 Transport Transferrin Transferrin receptor Ceruloplasmin Multidrug-resistance P-glycoprotein Matrix and barrier functions Procollagen prolyl hydroxylase-? 1 Intestinal trefoil factor Ecto-5?-nucleotidase Vasomotor regulation Endothelin-1 Adrenomedullin Tyrosine hydroxylase ? 1B -adrenergic receptor Inducible nitric-oxide synthase Endothelial nitric-oxide synthase Haem oxygenase Atrial natriuretic peptide Angiogenic signalling Vascular endothelial growth factor A Endothelial-gland-derived vascular endothelial growth factor Vascular-endothelial-growth-factor receptor-1 (Flt-1) Plasminogen-activator inhibitor-1 Energy metabolism Glucose transporter-1 Hexokinase-2 6-Phosphofructo-1-kinase L Glyceraldehyde-3-phosphate dehydrogenase Aldolase A Enolase-1 Phosphoglycerate kinase-1 Lactate dehydroxygenase A 6-Phosphofructo-2-kinase Carbonic anhydrase-9 Hormonal regulation Erythropoietin Leptin HIF Figure 1 | Direct transcriptional targets of HIF. The figure shows the protein products of the genes for which there is evidence of direct transcriptional activation by hypoxia-inducible factor (HIF). For a fully referenced version of this figure, see online supplementary information S1 (figure). Differentiated embryo chrondrocyte (DEC)1 and DEC2 are transcription factors of the basic helix?loop?helix family. ETS is a DNA-binding domain that defines a family of transcription factors. p35srj is a 35-kDa protein that contains a serine/glycine-rich junction. Nip3 is a member of the Bcl2 family of cell-death factors. CXCR is a receptor for the CXC family of chemokines. c-Met is the tyrosine-kinase product of the met proto-oncogene. � 2004 Nature Publishing Group 346 | MAY 2004 | VOLUME 5 www.nature.com/reviews/molcellbio REVIEWS a co-substrate (BOX 1;for reviews, see REFS 31,32). During catalysis, the splitting of molecular oxygen is coupled to both the hydroxylation of the prime substrate (for example, HIF) and the oxidative decarboxylation of 2-oxoglutarate to give succinate and CO 2 . One oxygen atom of molecular oxygen is incorporated into the alco- hol that is a result of oxidation of the prime substrate and the other is incorporated into succinate. The 2-oxoglutarate-dependent-oxygenase family is the largest of several families of non-haem, Fe 2+ -depen- dent enzymes that use a conserved two-histidine, one- carboxylate motif to coordinate Fe 2+ at the catalytic site. These endogenous protein ligands form a ?facial triad? that occupies three of the six possible coordination sites in an octahedral coordination geometry (FIG. 4;for a review, see REF. 32). In the enzyme?Fe 2+ complex, the remaining three coordination sites are occupied by two to three labile water molecules that are readily displaced by substrates and co-substrates. In the enzyme?Fe 2+ ? 2-oxoglutarate and enzyme?Fe 2+ ?2-oxoglutarate?O 2 complexes, 2-oxoglutarate occupies two of these coordi- nation sites and is ligated to Fe 2+ through its 1-carboxy- late and 2-oxo groups. After the loss of a water ligand, which is induced by substrate binding, dioxygen binds to the remaining site. The coordination chemistry of the 2-oxoglutarate- dependent oxygenases contrasts with that of haem- dependant oxygenases and oxidases, in which the por- phyrin ring and the distal ligating amino acid account for five of the six coordination positions. This leaves only one vacant site for binding molecular oxygen and prohibits the binding of a further (co-)substrate to the metal. In these enzymes, the role of the 2-oxoglutarate in the two-electron reduction of molecular oxygen is typi- cally taken by accompanying reductase/iron-sulphur proteins, which couple the reduction of O 2 to the oxida- tion of NAD(P)H. The flexibility in the coordination chemistry of the 2-oxoglutarate-dependent oxygenases, which is perhaps coupled to conformational changes during catalysis (see below), might facilitate the diverse oxidative reactions that are catalysed by these enzymes, including hydroxylations, oxidative ring closures and desaturations. The evolutionary origin of the oxygen-sensing func- tion of 2-oxoglutarate-dependent oxygenases in higher animals is unclear, although the superfamily is widely represented across both prokaryotes and eukaryotes. Members of the 2-oxoglutarate-dependent-oxygenase family and related enzymes oxidize both small- and large-molecule substrates and are involved in diverse biological functions. However, none of these processes point clearly to an ancestral oxygen-sensing mechanism, and in lower organisms other types of enzyme have been implicated in this role. The striking conservation of HIF prolyl hydroxylases across nematode worms, insects and vertebrates is apparently confined to higher eukaryotes. Neither the PHDs nor the HIF? orthologues have been identified so far in bacteria, yeast or plants, although a striking exception is the presence of PHD-related pro- teins in Pseudomonas aeruginosa and Vibrio cholera 33 . The function of these putative 2-oxoglutarate-dependent and genetic inactivation of this hydroxylase results in constitutive stabilization of the HIF? orthologues and activation of the transcriptional response 20,21,29 .In mammalian cells, a small interfering RNA approach has recently been used to specifically inactivate each of the three PHD enzymes in various cell types. In a range of cell lines, these studies have shown that inac- tivation of PHD2 alone is sufficient to upregulate HIF1? protein levels and HIF-target-gene expression in oxygenated cells, and it has been proposed that PHD2 is the most important enzyme in regulating oxygenated levels of HIF1? under these conditions (REF. 30).However, the extent to which these findings reflect a special activity of PHD2 on HIF1? or the rel- ative abundance of the PHD enzymes in oxygenated cell lines is not yet known. Transfection studies indi- cate that all three PHD enzymes have the potential to regulate HIF in cultured cells 21 .In the case of the HIF asparaginyl hydroxylase FIH, only one isoform has been identified so far, and genetic manipulation of this isoform has highlighted its non-redundant effect on HIF-dependent transcription under a range of conditions 28 . 2-Oxoglutarate-dependent oxygenases The involvement of two discrete types of 2-oxoglu- tarate-dependent oxygenase in directing transcriptional responses to hypoxia raises an interesting question ? is this class of enzyme particularly suited to an oxygen- sensing function? Enzymes that metabolize molecular oxygen (dioxygen or O 2 ) are defined as: dioxygenases when both atoms of oxygen are incorporated into their products; mono-oxygenases when one atom is incorpo- rated into a substrate and the other is reduced to water; and oxidases when both atoms are reduced to water. The HIF hydroxylases are therefore dioxygenases and belong to a large group of enzymes that use 2-oxoglu- tarate (a citric-acid-cycle intermediate and oxoacid) as AHIF1? HIF1? B AB bHLH PAS NAD CODDDNODDD CAD bHLH PAS Hydroxylation Pro Pro Asn Figure 2 | The domain structure of HIF1? and HIF1?/ARNT. The figure highlights: basic helix?loop?helix (bHLH) domains; PAS (PER (period circadian protein), AHR (aryl-hydrocarbon receptor), ARNT (aryl-hydrocarbon-receptor nuclear translocator), SIM (single-minded protein)) domains, which consist of directly repeating A and B sequences that are contained within a region of sequence similarity; the amino-terminal oxygen-dependent degradation domain (NODDD); the carboxy-terminal oxygen-dependent degradation domain (CODDD); the amino- terminal activation domain (NAD); and the carboxy-terminal activation domain (CAD). Note the overlap of CODDD and NAD in hypoxia-inducible factor (HIF)?. The oxygen responsiveness of HIF? is conveyed by separate protein interactions with three regions of HIF? that transfer responsiveness to hypoxia (that is, NODDD, CODDD and CAD), and these interactions involve the oxygen-dependent enzymatic hydroxylation of specific residues in these regions. � 2004 Nature Publishing Group NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 5 | MAY 2004 | 347 REVIEWS and the asparaginyl hydroxylase FIH as HIF hydroxy- lases was achieved following similar structural predic- tions that were coupled to biological knowledge. The jelly-roll conformation is also present in numerous other enzymes and proteins (including the CUPIN SUPER- FAMILY), many of which are not 2-oxoglutarate-depen- dent oxygenases 39 . Mutational studies of both PHDs and FIH have con- firmed the assignment and crucial importance of the predicted Fe 2+ -coordinating residues 21,26 .For FIH, thepredictions have also been verified by X-ray crystal- lography 40?42 (FIG. 5c).These studies showed that FIH, like other 2-oxoglutarate-dependent oxygenases, con- tains a single Fe 2+ that is coordinated almost octahe- drally by a two-histidine, one-carboxylate motif (that is, by the side chains of His199, Asp201 and His279; FIGS 4a,5d). The 2-oxoglutarate co-substrate binds the Fe 2+ in the typical bidentate manner through its 1-car- boxylate and 2-oxo groups. However, the FIH residues that interact with the 5-carboxylate of 2-oxoglutarate are unusual. By contrast with certain other 2-oxoglu- tarate-dependent oxygenases, in which an Arg residue (on or close to the eighth strand of the jelly-roll motif) and a Ser/Thr residue 34,35,43 interact with this carboxylate, in FIH the interacting residues are Lys214 (on the fourth strand of the jelly-roll motif), Thr196 and Tyr145. This and other features identify FIH as a member of a distinct subfamily of 2-oxoglutarate-dependent oxygenases. The presence of a jelly-roll motif and sequence similarities also define FIH as one of the jumonji tran- scription factors, members of which have been impli- cated in cell growth and heart development (for a review, see REF. 44). Owing to the sequence similarities between these factors and zinc-dependent enzymes that contain a jelly-roll motif (for example, phospho- mannose isomerase), some of them have been pro- posed to be zinc-dependent transcription factors, but might in fact be iron- and 2-oxoglutarate-dependent oxygenases. Many, but not all, of the jumonji proteins contain an HXD/EX n H motif. Interestingly, several are predicted to contain a tyrosine in the place of D/E, which might indicate that the consensus sequence should be changed to HXD/E/YX n H or that these pro- teins have alternative roles. Others, possibly including jumonji itself, do not seem to contain the conserved metal-binding ligands, which indicates that they might produce their biological effect through different, non- oxidative mechanisms. X-ray crystallography of FIH in complex with the CAD of HIF1? or HIF2? shows that the enzyme?sub- strate interaction involves at least two distinct sites 41 .At the hydroxylation site, the CAD forms an extensive set of hydrogen bonds with FIH and forms a tight turn that presents the side chain of Asn803 towards the catalytic Fe 2+ of the active site. Consistent with the importance of this structure, the identity of the adja- cent residue (Val802), which is conserved throughout HIF? sequences, is also crucial for substrate activity 45 . Unusually, the active-site complex involves a hydrogen bond between the iron-coordinating carboxylate residue in FIH (Asp201) and the target asparagine in oxygenases is unknown and their isolated appearance indicates that they might have arisen in these species by horizontal gene transfer. Structures of 2-oxoglutarate-dependent oxygenases. X-ray crystallographic analyses of the 2-oxoglutarate- dependent oxygenases have highlighted a core of eight ?-strands that are folded into a ?jelly-roll? motif (also known as a double-stranded ?-helix) 34?36 (FIG. 5a,b).The three iron-coordinating residues of the facial triad ? HXD/EX n H (where X is any amino acid) ? are found on or close to the second and seventh ?-strands. Such an arrangement has been observed in all structural studies of 2-oxoglutarate-dependent oxygenases so far and is predicted, by sequence conservation and mutational analysis, to be present in many other members of this class 31,33,37 .Other classes of non-haem, Fe 2+ -dependent oxygenases (for example, the cyclo-oxygenases, para- hydroxyphenylpyruvate dioxygenase and catechol dioxygenases) 38 use different structural platforms to present the two-histidine, one-carboxylate motif. In addition, at the start of the catalytic cycle some use Fe 3+ rather than the Fe 2+ that is used by 2-oxoglutarate- dependent oxygenases. Importantly, inserts might occur between the ?-strands of the jelly-roll motif ? in par- ticular, between the fourth and fifth strands ? that potentially confuse simple sequence analyses and struc- tural predictions 31 .When such possibilities are allowed for, analysis of genomic data indicates that many other predicted proteins might be 2-oxoglutarate-dependent oxygenases with as-yet-unknown functions 27,33 .Indeed, the functional identification of both the PHD enzymes CUPIN SUPERFAMILY The cupins are a diverse family of plant proteins, all of which contain at least one double- stranded ?-helix or jelly-roll strucural motif. This motif is also present in all structurally characterized 2-oxoglutarate- dependent oxygenases including the HIF hydroxylases, and is characteristic of the jumonji transcription factors. O 2 high O 2 low pVHL-mediated proteolysis Blocked p300 co-activator recruitment Inactivation of HIF transcriptional activity HO OH PHDs FIH Pro Asn Active HIF hydroxylases Activation of HIF transcriptional activity Inactive HIF hydroxylases HIF? p300 Stable HIF? Unstable HIF? Co-activator recruitment Pro Asn Pro Asn Figure 3 | Dual regulation of HIF? subunits by prolyl and asparaginyl hydroxylation. In the presence of oxygen, active hypoxia-inducible factor (HIF) hydroxylases ? that is, prolyl hydroxylase domains (PHDs) and factor inhibiting HIF (FIH ) ? downregulate and inactivate HIF? subunits. PHDs hydroxylate a prolyl residue in the amino- and the carboxy-terminal oxygen- dependent degradation domains (NODDD and CODDD, respectively; see FIG. 2), which promotes von-Hippel?Lindau-tumour-suppressor (pVHL)-dependent proteolysis and results in the destruction of HIF? subunits. FIH, on the other hand, hydroxylates an asparaginyl residue in the carboxy-terminal activation domain (CAD; see FIG. 2), which blocks p300 co-activator recruitment and results in the inactivation of HIF?-subunit transcriptional activity. In hypoxia, HIF hydroxylases are inactive and these processes are suppressed, which allows the formation of a transcriptionally active complex. � 2004 Nature Publishing Group 348 | MAY 2004 | VOLUME 5 www.nature.com/reviews/molcellbio REVIEWS different 2-oxoglutarate-dependent oxygenases have indicated a common mechanism in which an enzyme?Fe 2+ complex first binds 2-oxoglutarate, then its prime substrate and, finally, molecular oxygen (FIG. 6). The binding of 2-oxoglutarate and the prime substrate is reversible. The binding of the prime substrate dis- places a water molecule that is coordinated to the Fe 2+ , and this allows Fe 2+ to react with TRIPLET-STATE MOLECULAR OXYGEN.Structures of the 2-oxoglutarate-dependent-oxy- genase clavaminate synthase bound to Fe 2+ , 2-oxoglu- tarate and substrate were compared in the presence and absence of nitric oxide (NO; which was used as a substi- tute for molecular oxygen to produce a stable complex; FIG. 4b). Various other Fe 2+ and 2-oxoglutarate complexes of these enzymes were also studied and the results indi- cate that the binding of molecular oxygen might involve the rearrangement of the 1-carboxylate of 2-oxoglu- tarate, and the approach of molecular oxygen through the jelly-roll motif 48 (FIGS 5c,6).Interestingly, if such a process occurs in HIF hydroxylases, it could provide amechanism for oxygen sensing by regulating the access of oxygen to the active site. Binding of oxygen is followed by the oxidative decar- boxylation of 2-oxoglutarate to give succinate, CO 2 and a ferryl species (Fe IV =O) at the iron centre (FIG. 6).This highly reactive intermediate can then oxidize an unacti- vated C?H bond in the prime substrate. Evidence for intermediates comes from substrate-analogue studies 49 , model compounds 50 and spectroscopic analyses 51?53 .The prime-substrate product is released before succinate. The point at which CO 2 is released is uncertain. The sequential binding of co-substrate and prime substrate, which is necessary to trigger oxygen binding, is probably important to limit the generation of reactive oxidizing species in the absence of prime substrate. The genera- tion of such species in a prime-substrate-uncoupled manner can inactivate 2-oxoglutarate and the related oxygenases through self-oxidation, which sometimes leads to fragmentation 54,55 .Typically, the uncoupled turnover of 2-oxoglutarate occurs at ~5% of the rate of its coupled turnover in the presence of saturating concentrations of prime substrate. Several 2-oxoglutarate-dependent oxygenases, including procollagen prolyl hydroxylase and antho- cyanidin synthase, also have a requirement for ascorbate for full catalytic activity 56,57 .Although ascorbate might stimulate activity by reducing Fe 3+ to Fe 2+ (either free in solution or at the active site), the stimulation of oxygenase activity by ascorbate might occur by other mechanisms, for instance, by promoting completion of uncoupled cycles. For uncoupled reaction cycles that are catalysed by procollagen prolyl hydroxylase in the absence of prime substrate, the oxidation of 2-oxoglutarate to succinate has been shown to be stoichiometrically coupled to ascorbate 58 .It is believed that one role of ascorbate is to function as a surrogate reducing substrate to ?rescue? the enzyme in the event of the uncoupled production of a ferryl (Fe IV =O) intermedi- ate. Ongoing structural studies on anthocyanidin syn- thase indicate that ascorbate might also have a role in the binding of prime substrate to the enzyme 57 . HIF1? (Asn803; FIGS 5d,6). This might require disrup- tion before oxygen can bind at the active site and might therefore affect the oxygen sensitivity of the hydroxyla- tion reaction. Distal to this site, a portion of CAD that is necessary for efficient hydroxylation adopts an ?-heli- cal conformation and interacts with a hydrophobic region on the surface of FIH. Comparisons of FIH?CAD crystal structures with NMR structures of CAD bound to the CH-1 domain of p300 show that the conformation of the CAD region surrounding Asn803 changes from an ?-helical conformation when bound to p300 (REFS 24,25) to an extended loop when bound to FIH 41 .The CAD that is present in HIF? subunits has been classified as an ?intrinsically unstructured protein?, such that in the absence of binding partners the CAD region is disordered. The apparently disordered state for the isolated CAD might reflect the requirement to adopt different conformational states when bound to alternative partners. Although structural data for PHD?HIF? com- plexes is not yet available, mutational and deletional studies of HIF?-subunit NODDD and CODDD pro- lyl hydroxylation sites indicate that recognition by PHD enzymes is probably also complex 15,46,47 and involves further interactions that are discrete from the hydroxylation site. Enzyme mechanism. Although there are exceptions, spectroscopic, kinetic and structural studies on several TRIPLET-STATE MOLECULAR OXYGEN Most ?natural? molecules exist in the singlet state ? that is, they contain paired electrons. However, the most stable form of molecular oxygen (O 2 ) is the triplet state, in which there are two unpaired electrons. O OO Fe 2+ O His199 His279 OAsp201 O O ? Fe 2+ O His199 O O His279 OAsp201 O O O ? L L Arg293 Thr172 Arg297 His279 Nitric oxide 1-carboxylate His144 Tyr299 ab Glu146 Figure 4 | The coordination chemistry of 2-oxoglutarate-dependent oxygenases. a | The facial triad of metal-ion-binding residues (highlighted by blue text and the pink shaded region) in the octahderal coordination of Fe 2+ at the active site of 2-oxoglutarate-dependent oxygenases 32 . The factor inhibiting hypoxia-inducible factor (FIH) numbering system is used. Structural evidence shows that the 1-carboxylate of 2-oxoglutarate (green) can adopt both of the coordination positions that are shown. L represents ligand, which can be either water or dioxygen 48,96 . b | The crystal structure of the active site of the 2-oxoglutarate-dependent- oxygenase clavaminate synthase. The enzyme is complexed with Fe 2+ (purple sphere), 2-oxoglutarate (upper yellow structure), the substrate N-acetyl arginine (lower yellow structure), and nitric oxide, which was used as a substitute for molecular oxygen to produce a stable complex 48 . � 2004 Nature Publishing Group NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 5 | MAY 2004 | 349 REVIEWS Oxygen. Assays of HIF hydroxylation by cell extracts and recombinant PHD enzymes have been conducted in graded-hypoxia conditions using a controlled- atmosphere work station. A reduction in hydroxylation activity was detected when the oxygen concentration was reduced from 20% to 10%, and progressive inhibi- tion was observed at lower oxygen tensions 20 .This indi- cates that the effective K m of PHD enzymes for oxygen might be high in relation to the physiological range of oxygen tensions in tissues (lower K m values can reflect tighter substrate binding), which supports a sensing function. Although results that are obtained under in vitro conditions using non-physiological substrate concentrations are difficult to interpret, this result is consistent with recent studies of 2-oxoglutarate turnover by recombinant PHD enzymes, in which the pO 2 (partial pressure of oxygen) was measured directly in the reaction mix. Under these conditions, the apparent K m for oxygen for each of the PHD enzymes was in the range of 230?250 �M ? a result that contrasts with measurements on procollagen prolyl hydroxylase, which had a K m of 40 �M (REF. 46).The data for the PHD enzymes are broadly in line with studies of the induc- tion of both HIF1? protein expression and HIF DNA- binding activity by graded hypoxia in tissue-culture cells 59 , and indicate that PHD enzymes do manifest the appropriate sensitivity to hypoxia. For the asparaginyl hydroxylase FIH, similar studies indicate that its apparent K m for oxygen is lower (being ~90 �M) than that for the PHD enzymes and that its catalytic properties towards other co-substrates are also distinct (the K m for 2-oxoglutarate is 25 �M for FIH and 55?60 �M for the PHDs, and the K m for ascorbate is 260 �M for FIH and 140?170 �M for the PHDs) 60 . However, how these parameters translate into kinetic behaviour in vivo is still unclear. Transfection experi- ments have indicated that, at least when overexpressed, FIH can reduce HIF transcriptional activity even at low oxygen tensions 26,28 .This result apparently contrasts with data for transfected PHD3. In this case, suppres- sion of HIF transcriptional activity was observed in oxy- genated, but not hypoxic, cells 21 .However, the studies (by different groups) are difficult to compare, particu- larly as the cells were not pre-equilibrated in hypoxic conditions prior to transfection. Furthermore, as FIH could, in theory, simply compete with CH-1 to bind CAD, it is not yet clear whether all of the suppressive function of FIH on HIF transcriptional activity is enzymatic. Nitric oxide. NO is an in vitro inhibitor of the 2-oxog- lutarate-dependent oxygenases. Several studies of the HIF/hydroxylase pathway have shown both apparent inhibition of hydroxylation and activation of HIF tran- scriptional activity by different NO donors 61?63 .A mechanism for this can be readily envisaged, because NO is known to compete with O 2 for binding to the iron at the active site of 2-oxoglutarate-dependent oxy- genases 48 .Nevertheless, other studies have shown that NO donors might reduce HIF transcriptional activity in hypoxia 64,65 , and in one study the promotion of Studies with several enzymes have shown that certain substrate analogues and mutants can also stimulate uncoupled 2-oxoglutarate turnover. This leads to the idea that uncoupled turnover might reflect a proof- reading process, in which incorrect substrates or incorrectly bound substrates can be rejected following the activation of molecular oxygen 35 . Control of HIF-hydroxylase activity At present, it is not yet clear whether the HIF hydroxy- lases have evolved unique catalytic features or are rela- tively ?ordinary? 2-oxoglutarate-dependent oxygenases that simply use their absolute requirement for molecu- lar oxygen in a signalling role. It is also unclear how such a simple and direct interface with oxygen could be used in an intact organism to direct diverse responses in cells that operate at substantially different oxygen tensions (BOX 3).Nevertheless, emerging data indicate that the enzymes are sensitive to moderate hypoxia 20,46 , and that their activity might be controlled at several other levels, which potentially provide flexibility for directing physi- ological responses to hypoxia. K m The Michaelis constant. A kinetic parameter for a specific substrate in an enzyme-catalysed reaction. Providing certain conditions are met, the K m for a substrate can equate to its binding constant, and the lower the value of K m ,the tighter the substrate binds. pO 2 The partial pressure that is exerted by molecular oxygen in a mixture of gases. It is also used to define the concentration of molecular oxygen in a solution or biological tissue that is at equilibrium with such a gas mixture. C N Fe 2+ ab cd 64 4 5 2 7 6 3 8 1 82 53 7 1 Insert His279 His199 Fe 2+ Asp201 Asn803 C? Insert Figure 5 | Structural insights into 2-oxoglutarate-dependent oxygenases. a | A schematic representation of the jelly-roll (double-stranded ?-helix) core motif that is found in 2-oxoglutarate- dependent oxygenases. Inserts might occur between the ?-strands of the jelly-roll motif ? in particular, between the fourth and fifth strands (see label). The three iron-coordinating residues of the facial triad are found on or close to the second and seventh ?-strands (orange circles) 31,36 . b | The jelly-roll core motif of factor inhibiting hypoxia-inducible factor (FIH). c | The X-ray crystal structure of FIH showing the Fe 2+ (orange sphere) and the substrate (yellow). The jelly-roll core motif of FIH is shown in red, whereas other strands and helices are shown in green. Note the apparent ?tunnel? through the double-stranded ?-helix 40?42 . The insert between the fourth and fifth ?-strands of the jelly-roll core motif is highlighted. d | Part of the active site of FIH showing the FIH residues that bind Fe 2+ (grey), the ligating N-oxalylglycine (a 2-oxoglutarate analogue; green), and the side chain of Asn803 of hypoxia-inducible factor-1? (yellow), the methylene (C?) of which becomes hydroxylated 41 . � 2004 Nature Publishing Group 350 | MAY 2004 | VOLUME 5 www.nature.com/reviews/molcellbio REVIEWS active-site Fe 2+ can be substituted by Co 2+ ,Cu 2+ ,Zn 2+ and Mn 2+ .So, the classic hypoxia-mimetic function of cobal- tous ions can probably be explained by their direct inhi- bition of the HIF hydroxylases, although the direct binding of cobalt to the HIF? CODDD has been reported, which might contribute to HIF? stability 68 . From the biological perspective, these findings raise an important question: to what extent is HIF-hydroxy- lase activity regulated by physiological or pathological changes in cellular Fe 2+ availability? Experiments in tissue-culture cells indicate that such effects could be important, particularly with respect to cancer. The acti- vation of oncogenic pathways is often associated with the upregulation of HIF? protein levels even in oxygenated cells (for reviews, see REFS 69,70). Using hydroxylation- specific antibodies, it has been shown that the HIF1? CODDD is not always fully hydroxylated in such cells 71 . As oxygen availability should not be limiting in these cul- tures, this indicates that other mechanisms are limiting hydroxylase activity. In a separate line of investigation, it has been shown that supplementation with iron or ascorbate dramatically reduces HIF1? levels in a manner that is dependent on a functional hydroxylase system 72 . hydroxylase activity by NO donors has been reported 66 . Overall, this indicates that either NO ? or another component of the chemical probes that were used in these studies ? has more than one action on the HIF system. Interestingly, a recent study has indicated that NO might reduce the activation of HIF in hypoxia, because it might effectively relieve cellular hypoxia by acting as an endogenous inhibitor of cytochrome-c oxi- dase (complex IV of the mitochrondrial electron-trans- port chain) 67 . Iron/ascorbate. Iron binding by the two-histidine, one- carboxylate motif is relatively labile, particularly for Fe 2+ enzymes such as the 2-oxoglutarate-dependent oxyge- nases. In keeping with this, many 2-oxoglutarate-depen- dent oxygenases are readily inhibited by iron chelators: inhibition of both the PHDs and FIH by this mechanism readily explains the stabilization of HIF? and the activa- tion of HIF transcriptional activity that is observed when cells are exposed to the iron-chelator desferrioxamine. Active-site Fe 2+ might also be substituted by other metals that are unable to support the catalytic cycle. Indeed, spectroscopic and structural studies have shown that the Fe 2+ O His OH 2 OH 2 H 2 O His OAsp Fe 2+ O His O OH 2 O His OAsp O O O ? Fe 2+ O His O O His OAsp O O O ? N H O H NH 2 O N H O H NH 2 O O + OO Fe 3+ O His His OAsp ?O O ? N H O H NH 2 O N H O O NH 2 OH Fe IV O His O O His OAsp O O O ? O O ? Fe 2+ O His O O His OAsp O O ? Binding of molecular oxygen (O 2 ) to Fe 2+ Release of succinate Binding of H 2 O Release of hydroxylated HIF??CAD Binding of HIF??CAD Release of H 2 O Binding of 2-oxoglutarate Hydroxylation step Oxidative decarboxylation step Release of CO 2 Ferryl[Fe IV =O] intermediate Hydrogen bond Asn803 of HIF??CAD 2-Oxoglutarate FIH?Fe 2+ .nH 2 O Figure 6 | Mechanism of the 2-oxoglutarate-dependent-oxygenase FIH. The figure shows an outline of the mechanism of the 2-oxoglutarate-dependent-oxygenase FIH (factor inhibiting HIF (hypoxia-inducible factor)), which is a HIF asparaginyl hydroxylase. Fe 2+ is bound to the enzyme through the ?facial triad? of two histidinyl residues and one aspartyl residue 40?42 , and the remaining three coordination sites are occupied by two to three labile water molecules. Studies with FIH and other 2-oxoglutarate-dependent oxygenases indicate that, in most cases, the sequential binding of 2-oxoglutarate and then the protein substrate (HIF? subunit) to the active site occurs 31,32,41,52,53 . Binding of the latter displaces a water molecule from Fe 2+ , which leaves a vacant coordination site. Together, these processes allow the productive binding of triplet-state molecular oxygen. Subsequent oxidative decarboxylation of 2-oxoglutarate generates carbon dioxide, succinate and a ferryl (Fe IV =O) species. The latter is responsible for hydroxylating the ?-position of Asn803 in the CAD (carboxy-terminal activation domain) of HIF? subunits. Release of the hydroxylated product probably precedes that of succinate. The point at which carbon dioxide leaves the active site is unclear. Coloured text is used in the figure to highlight the changes that occur at each step. � 2004 Nature Publishing Group NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 5 | MAY 2004 | 351 REVIEWS increased susceptibility to particular types of tumour that have indirect connections with the hypoxia response. For example, defects in the genes that encode the succinate dehydrogenase subunits B, C and D 76,77 have been associ- ated with familial paraglioma, a type of tumour that most commonly affects the oxygen-sensitive carotid body and the incidence of which is greatly increased in populations living at a high altitude 77 .In this case, impaired hydroxylase activity might be caused by both reduced 2-oxoglutarate concentrations and elevated suc- cinate concentrations, because the latter is both a product and a weak inhibitor of 2-oxoglutarate-dependent oxyge- nases. Defective fumarate hydratase has been associated with familial susceptibility to papillary renal carcinoma, leiomyomata of the skin, and uterine fibroids 78 .Uterine fibroids are a type of tumour that is (albeit rarely) associ- ated with excessive erythropoeitin production. However, the dysregulation of HIF has not yet been clearly demon- strated in these tumours and it is possible that different mechanisms underlie the tumour predisposition. Enzyme synthesis/abundance. As the 2-oxoglutarate- dependent oxygenases are not equilibrium enzymes ? that is, they do not catalyse the reversal of hydroxylation ? the abundance of the HIF hydroxylases will dictate the rate of HIF hydroxylation at any given concentration of oxygen. Studies of their tissue distribution show large dif- ferences in expression, at least at the messenger RNA level. For example, PHD1 is particularly abundant in the testis, whereas PHD3 is strongly expressed in the heart 79 . However, whether these differences affect the oxygen dependence of HIF induction in vivo remains to be deter- mined. Interestingly, the PHD enzymes can be induced to a high level in response to various stimuli. So far, this has been most completely studied in the context of the response to hypoxia. Both PHD2 and PHD3 (but not PHD1) are strongly induced by hypoxia 20,30 .In keeping with this, the prior incubation of cells in hypoxic condi- tions results in an enhanced total HIF-hydroxylase activ- ity 20 and a more rapid destruction of HIF1? when cell cultures are re-oxygenated 80 .Induction of the PHDs by hypoxia is at least partly dependent on the transcriptional activity of HIF, so this mechanism effectively provides a feedback control on HIF signalling 30,81 . The expression of certain PHD enzymes is also highly inducible by other stimuli. PHD3 has previously been identified in different cell types as a gene product that is induced by p53, by stimuli that induce smooth- muscle differentiation and by nerve-growth-factor withdrawal 82?84 .PHD1 mRNA has been reported to be an oestrogen-inducible transcript in breast cancer cell lines 85 .However, it is unclear at present how these char- acteristics impinge on the regulation of the HIF system. Interestingly, both PHD1 (REF. 86) and PHD3 (REF. 83) have been reported to have growth-suppressive effects, although, again, it is not known whether these arise from effects on the HIF system or effects on other pathways. Other controls on the rate of HIF hydroxylation could be generated by: the production of alternative hydroxylase transcripts 46 that encode less-active or It therefore seems probable that under normal condi- tions of tissue culture, HIF-hydroxylase activity can be limited by the availability of iron and/or ascorbate. The mode of action of ascorbate in this situation is unclear. It is possible that ascorbate functions, as dis- cussed above, to reduce the catalytic iron centre in the event of uncoupled activation. However, as similar effects are observed with iron and ascorbate, it is possible that ascorbate functions in a more general way to pro- mote the availability of Fe 2+ .Several lines of evidence indicate that rapidly growing cancers are associated with a cellular iron deficiency in the tumour mass 73 . Understanding the extent to which this contributes to HIF activation, and potentially to tumour angiogenesis, should be of interest in future studies. The precise defini- tion of the in vivo roles of ascorbate is of increasing inter- est, not only because of its potential involvement in the hypoxic response, but also because of its possible role in DNA repair through stimulation of the 2-oxoglutarate- dependent-oxygenase AlkB 74,75 and related enzymes. 2-oxoglutarate and succinate. Use of the citric-acid-cycle intermediates 2-oxoglutarate, as a co-substrate, and suc- cinate, as a product, provides a potential interface between the HIF hydroxylases and energy metabo- lism. Although the possibility of dual control by oxy- gen and energy metabolism is appealing, it is not clear whether cellular 2-oxoglutarate levels are ever limit- ing.Intriguingly, however, genetic defects in several citric-acid-cycle enzymes have been associated with an SYSTEMIC HYPOXIA A reduction in the partial pressure of oxygen (pO 2 ) throughout the organism. TISSUE ISCHAEMIA Inadequate blood supply to a tissue, which causes other metabolic abnormalities in addition to hypoxia (for example, the defective delivery of substrates and removal of waste products). Box 3 | The oxygen dependence of HIF activation Studies of cultured cells showed an exponential induction of hypoxia-inducible factor (HIF)1? protein and the DNA-binding activity of HIF as oxygen concentration is reduced over the range 10.0?0.6% (REF. 59). This pattern correlates well with erythropoietin production (one of the most striking homeostatic responses to hypoxia) by the hepatoma cell line Hep3b (REF. 107).Although there are only a few such detailed studies of graded hypoxia, analyses of HIF-target-gene expression in a range of tissue- culture cells, which are derived from widely differing sources, have indicated surprisingly similar characteristics, with maximal induction being observed at ~1% oxygen. This response range overlaps that of oxygen tensions in vivo in normal tissues, although a tissue pO 2 (partial pressure of oxygen) equivalent to 1% oxygen would rarely be found in non-pathological states in the intact organism. In keeping with this, immunohistochemical staining of human and rodent tissues for induced HIF? proteins showed that HIF?-subunit levels are generally low in the basal state, but are strikingly increased by SYSTEMIC HYPOXIA or TISSUE ISCHAEMIA. Interestingly, by contrast with the tissue-culture observations, marked cell-type- specific differences in oxygen thresholds for HIF activation seem to exist in vivo.This is highlighted by rodent kidneys, in which a marked heterogeneity in pO 2 is generated by the countercurrent exchange of oxygen between blood vessels that loop from the cortex into the medulla and papilla. Cells near the papillary tip are at a pO 2 of a few mmHg physiologically (equivalent to <1% oxygen), but do not express HIF1?. Nevertheless, papillary HIF1? is strongly induced by systemic hypoxia 108 .In the renal cortex, the pO 2 is much higher. Here, HIF1? is strongly induced in proximal nephron segments by systemic hypoxia, but not in the adjacent distal nephron segments that presumably experience a similar pO 2 (REF. 108).Nevertheless, HIF1? can be induced in the distal nephron by cobaltous ions, which indicates that the regulatory hydroxylase pathway is functioning. The HIF system therefore seems to be set differently in different cells, in a manner that is appropriate for the physiological control of oxygen homeostasis. � 2004 Nature Publishing Group 1. Semenza, G. L. & Wang, G. L. A nuclear factor induced by hypoxia via de novo protein synthesis binds to the human erythropoietin gene enhancer at a site required for transcriptional activation. Mol. Cell. Biol. 12, 5447?5454 (1992). 2. Maxwell, P. H., Pugh, C. W. & Ratcliffe, P. J. Inducible operation of the erythropoietin 3? enhancer in multiple cell lines: evidence for a widespread oxygen sensing mechanism. Proc. Natl Acad. Sci. USA 90, 2423?2427 (1993). 3. Wang, 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 O 2 tension. Proc. Natl Acad. Sci. USA 92, 5510?5514 (1995). 4. Semenza, G. L. Hypoxia-inducible factor 1: oxygen homeostasis and disease pathophysiology. Trends Mol. Med. 7, 345?350 (2001). 5. Wenger, R. H. Cellular adaptation to hypoxia: O 2 -sensing protein hydroxylases, hypoxia-inducible transcription factors, and O 2 -regulated gene expression. FASEB J. 16, 1151?1162 (2002). 6. Pugh, C. W. & Ratcliffe, P. J. Regulation of angiogenesis by hypoxia: role of the HIF system. Nature Med. 9, 677?684 (2003). 7. Huang, L. E. & Bunn, H. F. Hypoxia-inducible factor and its biomedical relevance. J. Biol. Chem. 278, 19575?19578 (2003). 8. Semenza, G. L. Targeting HIF-1 for Cancer Therapy. Nature Rev. Cancer 3, 721?732 (2003). 9. Masson, N. & Ratcliffe, P. J. HIF prolyl and asparaginyl hydroxylases in the biological response to intracellular O 2 levels. J. Cell Sci. 116, 3041?3049 (2003). 10. Maynard, M. A. et al. Multiple splice variants of the human HIF-3? locus are targets of the von Hippel?Lindau E3 ubiquitin ligase complex. J. Biol. Chem. 278, 11032?11040 (2003). 11. Makino, Y. et al. Inhibitory PAS domain protein is a negative regulator of hypoxia-inducible gene expression. Nature 414, 550?554 (2001). 12. Ivan, M. et al. HIF? targeted for VHL-mediated destruction by proline hydroxylation: implications for O 2 sensing. Science 292, 464?468 (2001). 13. Jaakkola, P. et al. Targeting of HIF-? to the von Hippel?Lindau ubiquitylation complex by O 2 -regulated prolyl hydroxylation. Science 292, 468?472 (2001). 14. Yu, F., White, S. B., Zhao, Q. & Lee, F. S. HIF-1? binding to VHL is regulated by stimulus-sensitive proline hydroxylation. Proc. Natl Acad. Sci. USA 98, 9630?9635 (2001). References 12, 13 and 14 define prolyl hydroxylation as the key oxygen-dependent modification that promotes the interaction of HIF? subunits with pVHL. 15. Masson, N., Willam, C., Maxwell, P. H., Pugh, C. W. & Ratcliffe, P. J. Independent function of two destruction domains in hypoxia-inducible factor-? chains activated by prolyl hydroxylation. EMBO J. 20, 5197?5206 (2001). This paper shows that HIF? subunits contain two prolyl hydroxylation sites, each of which can interact independently with pVHL. 16. Maxwell, P. H. et al. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 399, 271?275 (1999). 17. Ohh, M. et al. Ubiquitination of hypoxia-inducible factor requires direct binding to the ?-domain of the von Hippel? Lindau protein. Nature Cell Biol. 2, 423?427 (2000). 18. Hon, W. C. et al. Structural basis for the recognition of hydroxyproline in HIF-1? by pVHL. Nature 417, 975?978 (2002). 19. Min, J.-H. et al. Structure of an HIF-1??pVHL complex: hydroxyproline recognition in signaling. Science 296, 1886?1889 (2002). References 18 and 19 describe X-ray crystal structures of hydroxylated HIF1? peptides bound to a VBC complex and show how hydrogen bonding to the hydroxylated peptide can discriminate between hydroxylated and non-hydroxylated HIF? subunits. 20. Epstein, A. C. R. et al. C. elegans EGL-9 and mammalian homologues define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell 107, 43?54 (2001). 21. Bruick, R. K. & McKnight, S. L. A conserved family of prolyl- 4-hydroxylases that modify HIF. Science 294, 1337?1340 (2001). References 20 and 21 define the HIF? prolyl hydroxylase (PHD) enzymes and demonstrate their conservation in C. elegans and D. melanogaster. 22. Lando, D., Peet, D. J., Whelan, D. A., Gorman, J. J. & Whitelaw, M. L. Asparagine hydroxylation of the HIF transactivation domain: a hypoxic switch. Science 295, 858?861 (2002). This paper defines asparaginyl hydroxylation as the key oxygen-dependent modification that regulates HIF? CAD activity by preventing its interaction with the CH-1 domain of p300. 23. McNeill, L. A. et al. Hypoxia-inducible factor asparaginyl hydroxylase (FIH-1) catalyses hydroxylation at the ?-carbon of asparagine-803. J. Biochem. 367, 571?575 (2002). 352 | MAY 2004 | VOLUME 5 www.nature.com/reviews/molcellbio REVIEWS and 3-methylcytosine by oxidation of the methyl group 74,75 raises the possibility that signalling and reg- ulation involving oxygen might also occur through the direct modification of (methylated) nucleic acids. This might extend to 5-methylcytosine, which is a common site of methylation in higher eukaryotes that is important in the regulation of gene expression. To gether with database predictions of additional 2- oxoglutarate-dependent enzymes that are related to the HIF hydroxylases 41 , including the jumonji pro- teins, these observations indicate that 2-oxoglutarate- dependent oxygenases might be widely involved in cell signalling. Indeed, it might be that many of the jumonji proteins would be more accurately described as 2-oxoglutarate-dependent oxygenases that are involved in transcriptional regulation. In considering potential targets, it is important to recognize that, as far as we know, the reactions catal- ysed by these enzymes are irreversible. In the case of HIF? subunits, their rapid re-synthesis enables the inactivation to be reversed. We would therefore pre- dict that other protein substrates that are involved in irreversible signalling pathways should also be rapidly turned over, or be targeted to other enzymes that could reduce, or otherwise modify, the hydroxylation site to reverse the direction of signalling (for example, by the reduction or phosphorylation of hydroxyl groups). Overall, the new insights into the regulation of HIF by protein hydroxylation offer many exciting possibili- ties for future research. As well as furthering our understanding of the physiological response to hypoxia, there is the possibility ? which is attracting widespread interest ? that pharmacological manipu- lation of these signalling pathways by hydroxylase inhibitors 60,92?95 might offer a new route to the therapy of ischaemic/hypoxic disease. interfering products; the post-translational (allosteric) modification of the enzymes; or the post-translational modification of HIF?-subunit target regions, which would affect enzyme?substrate interactions. For exam- ple, evidence has been provided for the functionally important phosphorylation of a conserved Thr residue in CAD (Thr796 in the human HIF1? CAD) 87 that is close to the asparaginyl hydroxylation site and might therefore affect the action of FIH. Protein hydroxylation in non-HIF signalling? Overall, accruing data on the HIF hydroxylases indicates the existence of numerous controls that could provide the flexibility that is required to govern the physiological activation of HIF by hypoxia. Furthermore, the discovery of protein hydroxylation as a new regulatory pathway that controls HIF raises the possibility that similar processes could be involved in signalling other biological responses to hypoxia. Recently, it has been proposed that the large subunit of RNA polymerase II is also targeted for pVHL-dependent ubiquitylation by prolyl hydroxyla- tion 88 .Interestingly, on the basis of the responses of iron regulatory protein-2 to the inhibitory 2-oxoglutarate analogue dimethyl-oxalylglycine, it has been proposed that 2-oxoglutarate-dependent oxygenases are also involved in cellular iron sensing 89 .Similar arguments have indicated a role for 2-oxoglutarate-dependent dioxygenases in the HIF-independent, but oxygen-sensi- tive, generation of the signalling molecule phosphatidic acid 90 .A 2-oxoglutarate-dependent oxygenase has also been described that hydroxylates the Asp and Asn residues in human epidermal-growth-factor domains 91 . And, although the functional significance of this hydroxylation is unknown, it seems reasonable to pro- pose that it is involved in signalling. The discovery that a 2-oxoglutarate-dependent-oxygenase AlkB can repair DNA that is methylated at 1-methyladenine � 2004 Nature Publishing Group NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 5 | MAY 2004 | 353 REVIEWS 24. Freedman, S. J. et al. Structural basis for recruitment of CBP/p300 by hypoxia-inducible factor-1?. Proc. Natl Acad. Sci. USA 99, 5367?5372 (2002). 25. Dames, S. A., Martinez-Yamout, M., Guzman, R. N. D., Dyson, H. J. & Wright, P. E. Structural basis for Hif-1?/CBP recognition in the cellular hypoxic response. Proc. Natl Acad. Sci. USA 99, 5271?5276 (2002). References 24 and 25 describe the NMR solution structure of HIF1? CAD complexed to the CH-1 domain of p300 and predict how hydroxylation of Asn803 in the HIF1? CAD might disrupt the interaction. 26. Lando, D. et al. FIH-1 is an asparaginyl hydroxylase enzyme that regulates the transcriptional activity of hypoxia-inducible factor. Genes Dev. 16, 1466?1471 (2002). 27. Hewitson, K. S. et al. Hypoxia inducible factor (HIF) asparagine hydroxylase is identical to Factor Inhibiting HIF (FIH) and is related to the cupin structural family. J. Biol. Chem. 277, 26351?26355 (2002). References 26 and 27 define FIH as a HIF? asparaginyl hydroxylase. 28. Mahon, P. C., Hirota, K. & Semenza, G. L. FIH-1: a novel protein that interacts with HIF-1? and VHL to mediate repression of HIF-1 transcriptional activity. Genes Dev. 15, 2675?2686 (2001). 29. Lavista-Llanos, S. et al. Control of the hypoxic response in Drosophila melanogaster by the basic helix-loop-helix PAS protein Similar. Mol. Cell. Biol. 22, 6842?6853 (2002). 30. Berra, E. et al. HIF prolyl-hydroxylase 2 is the key oxygen sensor setting low steady-state levels of HIF-1? in normoxia. EMBO J. 22, 4082?4090 (2003). This paper provides genetic evidence for the crucial importance of PHD2 in setting the oxygenated levels of HIF1?. 31. Schofield, C. J. & Zhang, Z. Structural and mechanistic studies on 2-oxoglutarate-dependent oxygenases and related enzymes. Curr. Opin. Struct. Biol. 9, 722?731 (1999). 32. Hegg, E. L. & Que, L. The 2-His-1-carboxylate facial triad ? an emerging structural motif in mononuclear non-heme iron(II) enzymes. Eur. J. Biochem. 250, 625?629 (1997). 33. Aravind, L. & Koonin, E. V. The DNA-repair protein AlkB, EGL-9, and leprecan define new families of 2-oxoglutarate- and iron-dependent dioxygenases. Genome Biol. 2, RESEARCH0007 (2001). 34. Valegard, K. et al. Structure of a cephalosporin synthase. Nature 394, 805?809 (1998). 35. Lee, H.-J. et al. Kinetic and crystallographic studies on deacetoxycephalosporin C synthase (DAOCS). J. Mol. Biol. 308, 937?948 (2001). 36. Roach, P. L. et al. Crystal-structure of isopenicillin N-synthase is the first from a new structural family of enzymes. Nature 375, 700?704 (1995). 37. Myllyharju, J. & Kivirikko, K. I. Characterization of the iron- and 2-oxoglutarate-binding sites of human prolyl 4-hydroxylase. EMBO J. 16, 1173?1180 (1997). 38. Bugg, T. D. H. Oxygenases: mechanisms and structural motifs for O 2 activation. Curr. Opin. Chem. Biol. 5, 550?555 (2001). 39. Dunwell, J. M., Culham, A., Carter, C. E., Sosa-Aguirre, C. R. & Goodenough, P. W. Evolution of functional diversity in the cupin superfamily. Trends Biochem. Sci. 26, 740?746 (2001). 40. Dann, C. E. I., Bruick, R. K. & Deisenhofer, J. Structure of factor-inhibiting hypoxia-inducible factor 1: an asparaginyl hydroxylase involved in the hypoxic response pathway. Proc. Natl Acad. Sci. USA 99, 15351?15356 (2002). 41. Elkins, J. M. et al. Structure of factor-inhibiting hypoxia- inducible factor (HIF) reveals mechanism of oxidative modification of HIF-1?. J. Biol. Chem. 278, 1802?1806 (2003). 42. Lee, C., Kim, S. J., Jeong, D. G., Lee, S. I. & Ryu, S. E. Structure of human FIH-1 reveals a unique active site pocket and interaction sites for HIF-1 and von Hippel?Lindau. J. Biol. Chem. 278, 7558?7563 (2003). References 40, 41 and 42 describe X-ray crystal structures of the HIF? asparaginyl hydroxylase FIH. Reference 41 also describes the structure of the FIH?Fe 2+ ?2-oxoglutarate?HIF?-CAD complex, which highlights an unusual two-site interaction. 43. Lee, H.-J. et al. Alteration of the co-substrate selectivity of deacetoxycephalosporin C synthase. J. Biol. Chem. 276, 18290?18295 (2001). 44. Clissold, P. M. & Ponting, C. P. JmjC: cupin metalloenzyme- like domains in jumonji, hairless and phospholipase A 2 ?. Trends Biochem. Sci. 26, 7?9 (2001). 45. Linke, S. et al. Substrate requirements of the oxygen- sensing asparaginyl hydroxylase factor inhibiting HIF. J. Biol. Chem. 279, 14391?14397 (2004). 46. Hirsila, M., Koivunen, P., Gunzler, V., Kivirikko, K. I. & Myllyharju, J. Characterization of the human prolyl 4-hydroxylases that modify the hypoxia-inducible factor HIF. J. Biol. Chem. 278, 30772?30780 (2003). 47. Huang, L. E., Pete, E. A., Schau, M., Milligan, J. & Gu, J. Leu-574 of HIF-1? is essential for the von Hippel?Lindau (VHL)-mediated degradation pathway. J. Biol. Chem. 277, 41750?41755 (2002). 48. Zhang, Z. et al. Crystal structure of a clavaminate synthase?Fe(II)?2-oxoglutarate?substrate?NO complex: evidence for metal centred rearrangements. FEBS Lett. 517, 7?12 (2002). 49. Wu, M., Moon, H. S., Begley, T. P., Myllyharju, J. & Kivirikko, K. I. Mechanism-based inactivation of the human prolyl-4-hydroxylase by 5-oxaproline-containing peptides: evidence for a prolyl radical intermediate. J. Am. Chem. Soc. 121, 587?588 (1999). 50. Rohde, J.-U. et al. Crystallographic and spectroscopic characterisation of a nonheme Fe(IV)=O complex. Science 299, 1037?1039 (2003). 51. Price, J. C., Barr, E. W., Tirupati, B., Bollinger, J. M. & Krebs, C. The first direct characterization of a high-valent iron intermediate in the reaction of an ?-ketoglutarate- dependent dioxygenase: a high-spin Fe(IV) complex in taurine/?-ketoglutarate dioxygenase (TauD) from Escherichia coli. Biochemistry 42, 7497?7508 (2003). 52. Pavel, E. G. et al. Circular dichroism and magnetic circular dichroism spectroscopic studies of the non-heme ferrous active site in clavaminate synthase and its interaction with ?-ketoglutarate cosubstrate. J. Am. Chem. Soc. 120, 743?753 (1998). 53. Zhou, J. et al. Spectroscopic studies of substrate interactions with clavaminate synthase 2, a multifunctional ?-KG-dependent non-heme iron enzyme: correlation with mechanisms and reactivities. J. Am. Chem. Soc. 123, 7388?7398 (2001). 54. Ryle, M. J. et al. O 2 and ?-ketoglutarate-dependent tyrosyl formation in TauD, an ?-keto acid-dependent non-heme iron dioxygenase. Biochemistry 42, 1854?1862 (2003). 55. Barlow, J. N., Zhang, Z. H., John, P., Baldwin, J. E. & Schofield, C. J. Inactivation of 1-aminocyclopropane- 1-carboxylate oxidase involves oxidative modifications. Biochemistry 36, 3563?3569 (1997). 56. Kivirikko, K. I. & Myllyharju, J. Prolyl 4-hydroxylases and their protein disulfide isomerase subunit. Matrix Biol. 16, 357?368 (1998). 57. Wilmouth, R. C. et al. Structure and mechanism of anthocyanidin synthase from Arabidopsis thaliana. Structure 10, 93?103 (2002). 58. Myllyla, R., Majamaa, K., Gunzler, V., Hanauske-Abel, H. M. & Kivirikko, K. I. Ascorbate is consumed stoichiometrically in the uncoupled reactions catalyzed by prolyl 4-hydroxylase and lysyl hydroxylase. J. Biol. Chem. 259, 5403?5405 (1984). 59. Jiang, B.-H., Semenza, G. L., Bauer, C. & Marti, H. H. Hypoxia-inducible factor 1 levels vary exponentially over a physiologically relevant range of O 2 tension. Am. J. Physiol. 271, C1172?C1180 (1996). 60. Koivunen, E., Hirsila, M., Gunzler, V., Kivirikko, K. I. & Myllyharju, J. Catalytic properties of the asparaginyl hydroxylase (FIH) in the oxygen sensing pathway are distinct from those of its prolyl-4-hydroxylases. J. Biol. Chem. 279, 9899?9904 (2004). Reference 60 analyses the kinetic properties of the HIF asparaginyl hydroxylase FIH and contrasts them with the properties of the HIF prolyl hydroxylases. 61. Kimura, H. et al. Hypoxia response element of the human vascular endothelial growth factor gene mediates transcriptional regulation by nitric oxide: control of hypoxia- inducible factor-1 activity by nitric oxide. Blood 95, 189?197 (2000). 62. Sandau, K. B., Zhou, J., Kietzmann, T. & Brune, B. Regulation of the hypoxia-inducible factor 1? by the inflammatory mediators nitric oxide and tumor necrosis factor-? in contrast to desferroxamine and phenylarsine oxide. J. Biol. Chem. 276, 39805?39811 (2001). 63. Metzen, E., Zhou, J., Jelkmann, W., Fandrey, J. & Brune, B. Nitric oxide impairs normoxic degradation of HIF-1? by inhibition of prolyl hydroxylases. Mol. Biol. Cell 14, 3470?3481 (2003). 64. Sogawa, K. et al. Inhibition of hypoxia-inducible factor 1 activity by nitric oxide donors in hypoxia. Proc. Natl Acad. Sci. USA 95, 7368?7373 (1998). 65. Huang, L. E., Willmore, W. G., Gu, J., Goldberg, M. A. & Bunn, H. F. Inhibition of hypoxia-inducible factor 1 activation by carbon monoxide and nitric oxide. J. Biol. Chem. 274, 9038?9044 (1999). 66. Wang, F. et al. HIF-1?-prolyl hydroxylase: molecular target of nitric oxide in the hypoxic signal transduction pathway. Biochem. Biophys. Res. Commun. 295, 657?662 (2002). 67. Hagen, T., Taylor, C. T., Lam, F. & Moncada, S. Redistribution of intracellular oxygen in hypoxia by nitric oxide: effect on HIF-1?. Science 302, 1975?1978 (2003). 68. Yuan, Y., Hilliard, G., Ferguson, T. & Millhorn, D. E. Cobalt inhibits the interaction between hypoxia inducible factor-? and von Hippel?Lindau protein by direct binding to hypoxia inducible factor-?. J. Biol. Chem. 278, 15911?15916 (2003). 69. Maxwell, P. H., Pugh, C. W. & Ratcliffe, P. J. Activation of the HIF pathway in cancer. Curr. Opin. Genet. Dev. 11, 293?299 (2001). 70. Semenza, G. L. Involvement of hypoxia-inducible factor 1 in human cancer. Intern. Med. 41, 79?83 (2002). 71. Chan, D. A., Sutphin, P. D., Denko, N. C. & Giaccia, A. J. Role of prolyl hydroxylation in oncogenically stabilized hypoxia-inducible factor-1?. J. Biol. Chem. 277, 40112?40117 (2002). 72. Knowles, H. J., Raval, R. R., Harris, A. L. & Ratcliffe, P. J. Effect of ascorbate on the activity of hypoxia inducible factor (HIF) in cancer cells. Cancer Res. 63, 1764?1768 (2003). 73. Le, N. T. V. & Richardson, D. R. The role of iron in cell cycle progression and the proliferation of neoplastic cells. Biochim. Biophys. Acta 1603, 31?46 (2002). 74. Falnes, P. O., Johansen, R. F. & Seeberg, E. AlkB-mediated oxidative demethylation reverses DNA damage in Escherichia coli. Nature 419, 178?182 (2002). 75. Trewick, S. C., Henshaw, T. F., Hausinger, R. P., Lindahl, T. & Sedgwick, B. Oxidative demethylation by Escherichia coli AlkB directly reverts DNA base damage. Nature 419, 174?178 (2002). 76. Baysal, B. E. et al. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science 287, 848?851 (2000). 77. Baysal, B. E. Hereditary paraganglioma targets diverse paraganglia. J. Med. Genet. 39, 617?622 (2002). 78. Tomlinson, I. P. et al. Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nature Genet. 30, 406?410 (2002). 79. Lieb, M. E., Menzies, K., Moschella, M. C., Ni, R. & Taubman, M. B. Mammalian EGLN genes have distinct patterns of mRNA expression and regulation. Biochem. Cell Biol. 80, 421?426 (2002). 80. Berra, E., Richard, D. E., Gothie, E. & Pouyssegur, J. HIF-1- dependent transcriptional activity is required for oxygen- mediated HIF-1? degradation. FEBS Lett. 491, 85?90 (2001). 81. del Peso, L. et al. The von Hippel?Lindau/hypoxia-inducible factor (HIF) pathway regulates the transcription of the HIF- proline hydroxylase genes in response to low oxygen. J. Biol. Chem. 278, 48690?48695 (2003). 82. Wax, S. D., Rosenfield, C. L. & Taubman, M. B. Identification of a novel growth factor-responsive gene in vascular smooth muscle cells. J. Biol. Chem. 269, 13041?13047 (1994). 83. Madden, S. L., Galella, E. A., Riley, D., Bertelsen, A. H. & Beaudry, G. A. Induction of cell growth regulatory genes by p53. Cancer Res. 56, 5384?5390 (1996). 84. Lipscomb, E. A., Sarmiere, P. D. & Freeman, R. S. SM-20 is a novel mitochondrial protein that causes caspase- dependent cell death in nerve growth factor-dependent neurons. J. Biol. Chem. 276, 5085?5092 (2001). 85. Seth, P. I. K., Porter, D. & Polyak, K. Novel estrogen and tamoxifen induced genes identified by SAGE (serial analysis of gene expression). Oncogene 21, 836?843 (2002). 86. Erez, N. et al. Expression of prolyl-hydroxylase-1 (PHD1/EGLN2) suppresses hypoxia inducible factor-1? activation and inhibits tumor growth. Cancer Res. 63, 8777?8783 (2003). 87. Gradin, K., Takasaki, C., Fujii-Kuriyama, Y. & Sogawa, K. The transcriptional activation function of the HIF-like factor requires phosphorylation at a conserved threonine. J. Biol. Chem. 277, 23508?23514 (2002). 88. Kuznetsova, A. V. et al. von Hippel?Lindau protein binds hyperphosphorylated large subunit of RNA polymerase II through a proline hydroxylation motif and targets it for ubiquitination. Proc. Natl Acad. Sci. USA 100, 2706?2711 (2003). 89. Hanson, E. S., Rawlins, M. L. & Leibold, E. A. Oxygen and iron regulation of iron regulatory protein 2. J. Biol. Chem. 278, 40337?40342 (2003). 90. Martin-Puig, S. et al. Role of iron (II)-2-oxoglutarate- dependent dioxygenases in the generation of hypoxia- induced phosphatidic acid through HIF1/2 and VHL- independent mechanisms. J. Biol. Chem. 279, 9504?9511 (2004). 91. Lavaissiere, L. et al. Overexpression of human aspartyl(asparaginyl) ?-hydroxylase in hepatocellular carcinoma and cholangiocarcinoma. J. Clin. Invest. 98, 1313?1323 (1996). � 2004 Nature Publishing Group 354 | MAY 2004 | VOLUME 5 www.nature.com/reviews/molcellbio REVIEWS 92. Ivan, M. et al. Biochemical purification and pharmacological inhibition of a mammalian prolyl hydroxylase acting on hypoxia-inducible factor. Proc. Natl Acad. Sci. USA 99, 13459?13464 (2002). 93. Schlemminger, I. et al. Analogues of dealanylalahopcin are inhibitors of human HIF prolyl hydroxylases. Bioorg. Med. Chem. Lett. 13, 1451?1454 (2003). 94. Mole, D. R. et al. 2-oxoglutarate analogue inhibitors of HIF prolyl hydroxylase. Bioorg. Med. Chem. Lett. 13, 2677?2680 (2003). 95. Warnecke, C. et al. Activation of the hypoxia-inducible factor-pathway and stimulation of angiogenesis by application of prolyl hydroxylase inhibitors. FASEB J. 17, 1186?1188 (2003). 96. Zhang, Z. H. et al. Structural origins of the selectivity of the trifunctional oxygenase clavaminic acid synthase. Nature Struct. Biol. 7, 127?133 (2000). 97. Stapon, A., Li, R. F. & Townsend, C. A. Carbapenem biosynthesis: confirmation of stereochemical assignments and the role of CarC in the ring stereoinversion process from L-proline. J. Am. Chem. Soc. 125, 8486?8493 (2003). 98. Clifton, I. J. et al. Crystal structure of carbapenem synthase (CarC). J. Biol. Chem. 278, 20843?20850 (2003). 99. Kivirikko, K. I., Myllyla, R. & Pihlajaniemi, T. Protein hydroxylation: prolyl 4-hydroxylase, an enzyme with four cosubstrates and a multifunctional subunit. FASEB J. 3, 1609?1617 (1989). 100. Vaz, F. M. & Wanders, R. J. A. Carnitine biosynthesis in mammals. Biochem. J. 361, 417?429 (2002). 101. Jansen, G. A. et al. Human phytanoyl CoA hydroxylase: resolution of the gene structure and the molecular basis of Refsum?s disease. Hum. Mol. Genet. 9, 1195?1200 (2000). 102. Aas, P. A. et al. Human and bacterial oxidative demethylases repair alkylation damage in both RNA and DNA. Nature 421, 859?863 (2003). 103. Welford, R. W. D., Schlemminger, I., McNeill, L. A., Hewitson, K. S. & Schofield, C. J. The selectivity and inhibition of AlkB. J. Biol. Chem. 278, 10157?10161 (2003). 104. Koong, A. C. et al. Candidate genes for the hypoxic tumor phenotype. Cancer Res. 60, 883?887 (2000). 105. Wykoff, C. C., Pugh, C. W., Maxwell, P. H., Harris, A. L. & Ratcliffe, P. J. Identification of novel hypoxia-dependent and independent target genes of the von Hippel?Lindau (VHL) tumor suppressor by mRNA differential expression profiling. Oncogene 19, 6297?6305 (2000). 106. Jiang, Y. et al. Gene expression profiling in a renal cell carcinoma cell line: dissecting VHL and hypoxia- dependent pathways. Mol. Cancer Res. 1, 453?462 (2003). 107. Fandrey, J. & Bunn, H. F. In vivo and in vitro regulation of the erythropoietin mRNA: measurement by competitive polymerase chain reaction. Blood 81, 617?623 (1993). 108. Rosenberger, C. et al. Expression of hypoxia-inducible factor-1? and -2? in hypoxic and ischemic rat kidneys. J. Am. Soc. Nephrol. 13, 1721?1732 (2002). Acknowledgements We thank all our colleagues for their dedicated contributions to the Oxford work in the area of hypoxic signalling, T. Murray-Rust for assis- tance with the figures, and the Wellcome Trust, the Biotechnology and Biological Sciences Research Council, the Medical Research Council and the European Union for funding our research. Competing interests statement The authors declare competing financial interests: see Web version for details. Online links DATABASES The following terms in this article are linked online to: Interpro: http://www.ebi.ac.uk/interpro/ PAS domain Swiss-Prot: http://us.expasy.org/sprot/ Cul2 | elonginB | elonginC | FIH | HIF1? | HIF2? | HIF3? | pVHL | Rbx SUPPLEMENTARY INFORMATION See online article: S1 (figure) Access to this links box is available online. � 2004 Nature Publishing Group "
Add Content to Group
|
Bookmark
|
Keywords
|
Flag Inappropriate
share
Close
Digg
Facebook
MySpace
Google+
Comments
Close
Please Post Your Comment
*
The Comment you have entered exceeds the maximum length.
Submit
|
Cancel
*
Required
Comments
Please Post Your Comment
No comments yet.
Save Note
Note
View
Public
Private
Friends & Groups
Friends
Groups
Save
|
Cancel
|
Delete
Please provide your notes.
Next
|
Prev
|
Close
|
Edit
|
Delete
Genetics
Gene Inheritance and Transmission
Gene Expression and Regulation
Nucleic Acid Structure and Function
Chromosomes and Cytogenetics
Evolutionary Genetics
Population and Quantitative Genetics
Genomics
Genes and Disease
Genetics and Society
Cell Biology
Cell Origins and Metabolism
Proteins and Gene Expression
Subcellular Compartments
Cell Communication
Cell Cycle and Cell Division
Scientific Communication
Career Planning
Loading ...
Scitable Chat
Register
|
Sign In
Visual Browse
Close
Comments
CloseComments
Please Post Your Comment