| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Carl-Ludwig-Institute of Physiology, University of Leipzig, Leipzig, Germany
1Correspondence: Carl-Ludwig-Institute of Physiology, University of Leipzig, Liebigstrasse 27, D-04103 Leipzig, Germany. E-mail: wenr{at}medizin.uni-leipzig.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONHYPOXIA-INDUCIBLE TRANSCRIPTION...OXYGEN SENSING AND REGULATION...REGULATION OF HIF-1 ACTIVITYO2-REGULATED GENESCONCLUSIONSREFERENCES |
|---|
subunits recently led to the spectacular discovery of the molecular principles of oxygen sensing. This review aims to summarize our current knowledge of oxygen-regulated gene expression.—Wenger, R. H. Cellular adaptation to hypoxia: O2-sensing protein hydroxylases, hypoxia-inducible transcription factors, and O2-regulated gene expression.
Key Words: environmental adaptation • erythropoietin • oxygen • transcriptional regulation • vascular endothelial growth factor
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOXIA-INDUCIBLE TRANSCRIPTION...OXYGEN SENSING AND REGULATION...REGULATION OF HIF-1 ACTIVITYO2-REGULATED GENESCONCLUSIONSREFERENCES |
|---|
|
HYPOXIA-INDUCIBLE TRANSCRIPTION FACTORS |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOXIA-INDUCIBLE TRANSCRIPTION...OXYGEN SENSING AND REGULATION...REGULATION OF HIF-1 ACTIVITYO2-REGULATED GENESCONCLUSIONSREFERENCES |
|---|
, 2
). These studies revealed that several factors interact with the erythropoietin 3' HRE, one of which bound solely after hypoxic exposure. Semenza and Wang hence designated this DNA binding activity ‘hypoxia-inducible factor 1’ or ‘HIF-1’ (3)
. Apart from human hepatoma cell lines, HIF-1 DNA binding activity was detected in nuclear extracts derived from phenylhydrazine-treated mice, confirming that the in vitro cultured cell lines reflect the in vivo situation (4)
. HIF-1 activity was detected in various non-erythropoietin-producing cell lines cultured under hypoxic conditions (5
, 6)
, providing the first evidence that HIF-1 not only activates the erythropoietin gene, but is part of a widespread oxygen-sensing and signal transduction mechanism.
The cloning of the HIF-1- and ß subunits
Functional analyses revealed that HIF-1 is a phosphorylation-dependent and redox-sensitive protein, contacting DNA in the major groove (7)
. Biochemical purification of this protein yielded a heterodimer composed of a 120 kDa subunit and a 91–94 kDa ß subunit (8)
. These subunits were partially sequenced and the comparison with sequence databases allowed Wang and Semenza (9)
to identify HIF-1 as a novel protein and HIF-1ß as being identical to the previously identified heterodimerization partner of the dioxin receptor/aryl hydrocarbon receptor (AhR), called AhR nuclear translocator (ARNT) (10)
. The HIF-1 cDNA was cloned and fully sequenced, revealing that HIF-1 belongs to the same class of proteins as ARNT, containing a basic helix-loop-helix (bHLH) DNA binding domain and a common domain called PAS, an acronym for the first three members of this family, namely, Sim/ARNT/Per (for a review on the PAS protein family, see ref 11
). Later on, HIF-1 cDNA was independently cloned in a search for new heterodimerization partners of ARNT (12)
and by its capability to bind the transcriptional coactivator p300/CBP (13)
.
Additional HIFs
In addition to the ubiquitously expressed HIF-1, two other members of this family, HIF-2 (14
15
16
17)
and HIF-3 (18)
, were identified that show a more restricted tissue expression pattern. Functional comparison between HIF-1 and HIF-2 in vitro revealed many similarities concerning genomic organization, modular protein structure, hypoxic protein stabilization, heterodimerization with ARNT, DNA recognition, DNA binding, and trans-activation of reporter genes (19
20
21
22
23)
. The three HIF subunits show partially overlapping expression patterns in vitro and in vivo (20
, 24
, 25)
. Thus, HIF-2 and HIF-3 potentially might interact with the DNA binding site of HIF-1 target genes. However, further experiments are required to define the precise role of each member of the HIF family.
A novel inhibitory PAS protein (IPAS), a splice variant of HIF-3, has recently been discovered by Poellinger and co-workers (26)
. IPAS lacks a trans-activation domain, thus serving as a natural HIF antagonist. IPAS is strongly expressed in the corneal epithelium of the eye, an organ in which HIF-dependent angiogenesis is suppressed despite profound tissue hypoxia. Three other naturally occurring potential HIF-1 antagonists were reported: aHIF, an antisense RNA complementary to the 3' untranslated region of HIF-1 (27)
; HIF-1Z, a zinc-inducible isoform lacking exon 12 (28)
; and a dominant-negative HIF-1 isoform lacking exons 11 and 12 (29)
. The physiological relevance of these transcripts still needs to be defined.
Expression and physiological functions of HIFs in vivo
The existence of several HIFs, partially overlapping in expression and function, suggested that they might be functionally redundant. However, embryonic lethal phenotypes were found in mouse gene targeting experiments at both the HIF-1 and HIF-2 loci. These experiments demonstrated that HIF-1 is required for mesenchymal cell survival during embryonic development. HIF-1 knockout embryos die around midgestation, showing cardiovascular malformations and open neural tube defects (30
31
32)
. Mice containing only one mutant HIF-1 allele develop normally but show impaired physiological responses to chronic hypoxia such as reduced polycythemia, right ventricular hypertrophy, pulmonary hypertension, pulmonary vascular remodeling, and electrophysiological responses (33
, 34)
. Carotid bodies isolated from these mice show a defective neural activity after exposure to hypoxia (35)
. Finally, although the ventilatory response to acute hypoxia was normal, ventilatory adaptation to chronic hypoxia was impaired in partially HIF-1-deficient mice (35)
. These results clearly demonstrate the central role of HIF-1 in the adaptation to hypoxia not only at the cellular but at the systemic level.
Diverse phenotypes related to the cardiovascular system were reported for HIF-2 knockout embryos. On one genetic background, the organ of Zuckerkandl (the major catecholamine source during development) does not produce sufficient amounts of catecholamines, leading to embryonic lethality, presumably because of bradycardia (36)
. On a different genetic background, HIF-2 was found to play an important role for remodeling of the primary vascular network (37)
. The expression pattern and functional role of HIF-3 still have to be elucidated.
Several tissues are naturally hypoxic despite normoxic-inspired oxygen fractions. This tissue hypoxia often serves as an important, HIF-mediated developmental signal. For example, conditional knockout mice lacking HIF-1 expression specifically in the cartilaginous growth plate of the developing bone show gross skeletal malformations and die perinatally, probably due to tracheal abnormalities (38)
. In these mice, cells in the hypoxic interior of the growth plate die as a result of defective HIF-1-mediated growth arrest. During placenta development, hypoxia maintains trophoblast proliferation and prevents differentiation toward an invasive phenotype (39)
. Indeed, ARNT knockout mice revealed that HIF-1 is essential for mammalian placentation (40)
. In addition, least part of this developmental process is regulated by HIF-1-mediated TGF-ß3 expression (41)
. HIF-1-regulated TGF-ß3 expression is involved in the scarless wound healing in hypoxic embryonic skin (42)
.
Expression studies in healthy mice revealed that the function of HIF-1 might reach far beyond hypoxia adaptation: HIF-1 was found to be expressed in diverse normoxic organs within distinct cell types (43)
, suggesting that additional factors such as cellular differentiation and response to growth factors regulate HIF-1 expression. Expression of HIF-1 in normoxic mouse brain, for example, appears to be involved in the regulation of the circadian rhythm, as HIF-1 was shown to interact with the clock component Per1 (44)
. A novel isoform of HIF-1 whose expression is regulated by an alternative promoter/first exon combination is expressed in postmeiotic spermatids of the mouse testis; the corresponding HIF-1 protein is located in the midpiece of the spermatozoal tail (45)
. The same isoform was reported to be up-regulated in activated T cells (46)
. However, further studies will have to reveal the function of this alternative HIF-1 isoform at these specific locations.
A recent report demonstrated HIF-1 induction by exposing mice to heat, leading to an increase in core body temperature as found in fever. As shown by the different modification pattern of HIF-1, this effect is not due to tissue hypoxia secondary to heat but occurs rather by direct protein stabilization involving HSP90 (47)
. Myocardial mechanical stress was reported to induce HIF-1 via stress-activated channels (48)
. This effect seems to be mediated via common kinase signaling pathways (see below).
HIFs are expressed in a multitude of diseases associated with tissue hypoxia, including ischemia and tumorigenesis. However, this review focuses on the physiological functions of HIFs; pathophysiological functions were reviewed previously (49)
.
|
OXYGEN SENSING AND REGULATION OF HIF SUBUNITS
|
|---|
|
TOPABSTRACTINTRODUCTIONHYPOXIA-INDUCIBLE TRANSCRIPTION...OXYGEN SENSING AND REGULATION...REGULATION OF HIF-1 ACTIVITYO2-REGULATED GENESCONCLUSIONSREFERENCES |
|---|
. The breakthrough experiments were performed by two research groups, headed by Ratcliffe (51)
and Kaelin (52)
, who discovered that an oxygen-dependent enzymatic activity covalently modifies a domain of HIF-1 known as the oxygen-dependent degradation (ODD) domain (53)
. To the surprise of all researchers involved in the hunt for the oxygen sensor, this modification turned out to be the hydroxylation of two proline residues (Pro402 and Pro564) by a prolyl-4-hydroxylase, similar to the proline modification known from the collagens (51
, 52
, 54
, 55)
.
Molecular cloning of the HIF-prolyl hydroxylases was greatly supported by C. elegans genetics. The egl-9 gene encoded an orphan prolyl hydroxylase and, indeed, egl-9-mutant C. elegans constitutively expressed HIF-1 (56)
. Using the egl-9 sequence as an electronic bait in database searches, Ratcliffe and co-workers identified three mammalian homologs designated ‘prolyl hydroxylase domain containing’ (PHD) 1, 2, and 3 (56)
. The same three HIF-prolyl hydroxylases (termed HPH 3, 2, and 1, respectively) were cloned by Bruick and McKnight based on sequence similarity to collagen-modifying prolyl hydroxylases (57)
.
The PHDs/HPHs are dioxygenases requiring oxygen and 2-oxoglutarate as cosubstrates. PHDs/HPHs contain iron liganded by two histidine and one aspartic acid residues. Oxygen binding to this central moiety requires the vitamin C-dependent maintenance of iron in its ferrous state. The PHDs/HPHs transfer one oxygen atom onto the proline residues of the HIF subunits; the second oxygen atom reacts with 2-oxoglutarate, yielding succinate and carbon dioxide as products. Under oxygen-deprived conditions or when the PHDs/HPHs are inactivated by competitive substrate analogs, the HIF prolines remain unmodified. The same effects were obtained by iron chelation, oxidation, or replacement by transition metal ions such as cobalt, treatments known to mimic hypoxia-inducible expression. The most convincing proof for the role of PHDs/HPHs as oxygen sensors was the demonstration that HIF subunit modification was reduced under hypoxic conditions similar to those known to induce HIF-1 (56
, 58)
.
Whitelaw and co-workers recently discovered that the carboxyl-terminal trans-activation region of HIF subunits is hydroxylated in an oxygen-dependent manner (59)
. In contrast to proline modification, regulating stability, the carboxyl-terminal hydroxylation modifies asparagine residues (Asn803 and Asn851 in HIF-1 and HIF-2, respectively), regulating activity rather than stability by interfering with the recruitment of the transcriptional coactivator CBP/p300 (59)
. Although the corresponding asparaginyl hydroxylase has not been identified, these results are consistent with the observation that stabilization alone is not sufficient to fully activate HIFs under normoxic conditions (60)
.
The von Hippel-Lindau tumor suppressor protein (pVHL) mediates HIF-1 protein degradation under normoxic conditions
Hydroxylation of the HIF-1 ODD domain confers rapid degradation to the constitutively translated protein (58
, 61)
. In fact, the HIF-1 protein usually remains undetectable under normoxic conditions. Proteasomal inhibitors or mutation of the ubiquitin-activating enzyme E1 stabilize HIF-1, demonstrating that under normoxic conditions HIF-1 is degraded by ubiquitinylation and proteasomal degradation (53
, 62
63
64
65
66)
. As discovered by Maxwell and co-workers (23)
, this degradation is mediated by pVHL, binding to the hydroxylated but not to the nonhydroxylated HIF-1 ODD domain (51
, 52)
. pVHL is part of an ubiquitin ligase linking the HIF subunits via elongin C and a complex composed of several not completely characterized proteins to the ubiquitinylation machinery. Cells deficient in functional pVHL show constitutively high HIF subunit levels and expression of many oxygen-regulated genes, leading to highly vascularized hemangioblastoma tumor formation, the most frequent manifestation of hereditary VHL disease. Tumor mutations of pVHL affect either binding to HIF-1 or binding to elongins C and B, consistent with a role of HIFs in the tumorigenesis of hemangioblastoma and renal cell carcinoma (67
68
69
70
71
72
73
74)
.
Alternative mechanisms stabilizing HIF-1 under normoxic conditions
Apart from the relatively well-characterized mechanisms of hypoxic HIF subunit stabilization, many growth factors and cytokines are known to stabilize HIF-1 under normoxic conditions, including insulin, insulin-like growth factors 1 and 2, epidermal growth factor, fibroblast growth factor 2, interleukin 1ß, tumor necrosis factor , angiotensin II, thrombin, transforming growth factor ß1, platelet-derived growth factor, and hepatocyte growth factor (75
76
77
78
79
80
81
82
83
84)
. Despite this great diversity, most of these growth factors might stabilize HIF-1 via common cellular kinase pathways (see below) activated by cell type-specific receptors. However, so far it is not known how HIF-1 is finally stabilized by these signaling pathways.
Nitric oxide (NO) has been reported to normoxically stabilize HIF-1 (85
86
87
88
89)
. Contradictory reports have been published showing that treatment with NO (as well as carbon monoxide) or ectopic expression of inducible NO synthase (which is itself a HIF-1 target) abrogated HIF-1 activity (90
91
92
93)
. This discrepancy might be explained by the finding that NO effects seem to be transient because initial NO treatment increased, but prolonged NO treatment decreased, HIF-1 protein and hypoxic erythropoietin expression (94)
. As with the growth factors and cytokines, these in vitro data might be restricted to the cell culture models used and do not necessarily reflect the physiological situation in vivo.
|
REGULATION OF HIF-1 ACTIVITY |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOXIA-INDUCIBLE TRANSCRIPTION...OXYGEN SENSING AND REGULATION...REGULATION OF HIF-1 ACTIVITYO2-REGULATED GENESCONCLUSIONSREFERENCES |
|---|
protein stabilization alone is not sufficient to activate HIF-1 under normoxic conditions. Solely blocking HIF-1 ubiquitinylation and proteasomal degradation does not yield transcriptionally active HIF-1 (62
, 65)
. Additional, partially hypoxia-inducible mechanisms seem to be involved in HIF activation. Part of this hypoxic activation might be explained by the lack of asparaginyl hydroxylation (see above), but the full HIF-1 activation cascade includes post-translational protein phosphorylation, nuclear translocation, ARNT heterodimerization, DNA binding, recruitment of general and tissue-specific transcriptional cofactors, and target gene trans-activation. Each step may be under the control of the oxygen concentration and other environmental parameters, growth factors, hormones, nutrients, etc. Cross-talk with other signaling pathways allows for the cell type-specific and even target gene-specific fine-tuning of the general hypoxic response. These functions are based on HIF-1 protein–protein interactions (Table 1
).
|
Phosphorylation of HIF-1
Early reports showed that phosphorylation plays an important role in HIF-1 activation (95
, 96)
, but replacement of potential phosphoacceptor sites in the critical region of the ODD domain does not affect its role in hypoxic protein stabilization (63
, 64)
. Thus, protein phosphorylation is secondary to hypoxic stabilization and seems to be required to regulate the transcriptional activity of HIF-1. Several reports showed that the p42/p44 (Erk2/Erk1) mitogen-activated protein kinases (MAPKs) can phosphorylate HIF subunits leading to similar electrophoretic migration properties as the hypoxically modified protein (97
98
99
100
101
102)
. However, the corresponding phosphoamino acid residues were not identified. MAPK-dependent phosphorylation enhances the transcriptional activity of HIFs, but in nonexcitable cells the p42/p44 MAPKs are not activated by hypoxia (97)
. In excitable cells, hypoxia might indirectly activate MAPKs, but HIF -protein stabilization is independent of MAPK-dependent phosphorylation (98)
. Because many of these growth factor pathways recruit the p42/p44 MAPKs via the receptor tyrosine kinase 
Ras/Raf MEK pathway, the results may explain how growth factors activate HIF-1 function. However, they do not explain how growth factor stimulation stabilizes the HIF-1 protein.
Another important modulator of HIF-1 activity is the receptor tyrosine kinase PI3-kinase PTEN Akt kinase pathway. Specific inhibition of the PI3-kinase was shown to inhibit many of the HIF-1-dependent cellular responses (103
104
105
106)
and loss of the tumor suppressor function of PTEN facilitated HIF-1-mediated gene expression (107)
. Akt-1 might inhibit glycogen synthase kinase-3-mediated inactivation of HIF-1 (104)
. Akt-1 was suggested to induce HIF-1 protein translation rather than stability (108)
, providing a putative mechanism of how this phosphorylation pathway can overcome the oxygen sensor-mediated HIF-1 degradation under normoxic conditions.
Besides phosphorylation, many studies suggested that reduction oxidation (redox) -dependent processes might trigger HIF-1 function (22
, 62
, 109
110
111
112
113
114)
, probably even by direct protein oxidation (111)
. However, the specificity of these effects for the physiological regulation of HIF-1 in vivo remains obscure.
Nuclear localization of HIF-1 and the role of ARNT
In contrast to HIF-1, ARNT is not regulated by hypoxia. Increases of ARNT sometimes observed in nuclear extracts of hypoxic cells (115)
are most likely due to artifacts occurring during extract preparation (116)
. In mammalian transcription factors, ARNT serves primarily as a heterodimerization partner of other PAS proteins such as AhR or HIF subunits. Formation of the HIF-1 heterocomplex is mandatory for DNA binding and trans-activation (117
118
119
120
121)
. However, despite the term ‘nuclear translocator’, ARNT is not required for HIF-1 (or AhR) translocation into the nucleus (116)
. Rather, HIF-1 autonomously translocates into the nucleus due to the presence of a nuclear localization signal (122)
. Hypoxia might activate nuclear translocation of HIF-1 by a so far unknown mechanism (122)
, but overexpressed HIF-1 constitutively localizes to the nucleus under normoxic conditions (60)
. Thus, nuclear translocation might be regulated by normoxic inhibition rather than hypoxic activation, and overexpression of HIF-1 might saturate this inhibitory mechanism. The p14ARF tumor suppressor protein was reported to sequester HIF-1 into the nucleolus, thereby inhibiting its trans-activation function (123)
.
Transcriptional cofactor recruitment by HIF-1
HIF-1 contains two trans-activation (TA) domains, one overlapping with the ODD domain and the second at the carboxyl terminus (12
, 53
, 63
, 124)
. Regulation of the transcriptional activity of HIF-1 occurs via post-translational hydroxylation, phosphorylation, and redox modifications of these two TA domains. The TA domains confer transcriptional activation of target genes mainly by the recruitment of general transcriptional coactivators including CBP/p300, SRC-1, or TIF2 (13
, 22
, 112
, 122
, 125
, 126)
. Modification of the carboxyl-terminal TA domain by asparaginyl hydroxylation inhibits interaction with the CH1 domain of CBP/p300 (59)
. The CPB/p300 cofactor is a histone acetyltransferase known to be recruited by many transcription factors. Histone acetylation modulates chromatin structure to ‘open’ the genomic locus for increased transcriptional activity. CBP/p300 is known to directly acetylate some transcription factors; it is tempting to speculate that HIF-1 might be an acetylation target, which could provide another level of HIF-1 regulation.
Livingston and co-workers reported that p35srj binds to the CH1 domain of CBP/p300 and blocks the interaction with HIF-1 resulting in a marked inhibition of trans-activation (127)
. Moreover, Semenza and co-workers (128)
recently cloned a factor inhibiting HIF-1 (FIH-1). FIH-1 is a transcriptional repressor that interacts with HIF-1 and pVHL. FIH-1 recruits histone deacetylases, directly or via pVHL, and hence counteracts the recruitment of the histone acetylases CBP/p300 (128)
. It will be interesting to determine whether asparaginyl hydroxylation of the carboxyl-terminal TA domain blocks the interaction with CBP/p300 by direct steric hindrance or by hydroxylation-dependent, oxygen-regulated recruitment of transcriptional corepressors.
A reductive environment provided by the redox factor Ref-1 enhances the recruitment of the transcriptional coactivators SRC-1, TIF2 and CBP/p300 (112)
. Redox-sensitive cysteine residues that affected trans-activation (in the carboxyl-terminal TA domain) or DNA binding (in the amino-terminal DNA binding domain) of HIF subunits were identified (22
, 111)
. A model was proposed in which reduced thioredoxin translocates into the nucleus of hypoxic cells and transmits the redox signal to Ref-1, which in turn interacts with the cysteines. Indeed, overexpression of thioredoxin/Ref-1 was shown previously to amplify the hypoxic signal (110)
. However, the functional importance of the cysteine residue could not be confirmed in other mutation experiments (19)
and a nuclear translocation of thioredoxin under hypoxia was not observed in other reports (58)
. Therefore, the specificity of thioredoxin/Ref-1 in maintaining HIF-1 in a reduced state as part of the hypoxia-signaling mechanism awaits further clarification.
Several negative feedback regulatory pathways apparently limit the hypoxic response. On the mRNA level, a down-regulation of HIF-1 mRNA concentrations by an unknown mechanism was found after prolonged exposure to hypoxia (121
, 129)
. On the protein level, a ‘HIF-1 proteasome targeting factor’ (HPTF) was proposed (130)
. According to this model, HPTF is transcriptionally activated by HIF-1 and in turn targets the HIF-1 protein for degradation. This factor might be identical to hypoxia-inducible HIF-prolyl hydroxylase family members (56)
. Whether these members are by themselves HIF-1 targets is unknown. Finally, on the trans-activation level, HIF-1-dependent induction of p35srj was reported, resulting in a marked inhibition of trans-activation (127)
.
HIF-1 DNA binding and transcriptional complex assembly
Once stabilized and activated by hypoxia, HIF-1 binds to the consensus HIF-1 DNA binding site (HBS) A/(G)CGTG present in the HREs of many oxygen-regulated genes (131)
. The conserved core HBS contains a CpG dinucleotide, lowering the number of actual HBS sequences found in the genome compared with the frequency of a random tetranucleotide sequence. CpG methylation of the erythropoietin 3' HBS abrogates both HIF-1 DNA binding and gene activation, and endogenous erythropoietin expression in hepatoma cells is inversely related to the degree of CpG methylation of the erythropoietin 3' HBS (132)
. In contrast to most HIF-1-dependent genes where the HBSs are located in methylation-free CpG islands, the erythropoietin 3' HBS is present in a locus with average G+C content and suppressed CpG dinucleotide frequency. Such loci are usually methylated and hence ‘silenced’ in mammalian genomes. Thus, there must be a selective pressure to keep this site methylation-free. This could be the constitutive binding of a transcription factor to the HBS. Since HIF-1 activity can be detected only under hypoxic conditions, the specific constitutive binding of ATF-1/CREB-1 (133
, 134)
to the HBS might prevent CpG methylation. In support of this model, in vivo footprinting showed that the HBS is occupied in normoxia in the absence of HIF-1 (135)
.
An HBS is necessary but not sufficient for efficient hypoxic gene activation. Whereas the HBS is the minimal DNA domain required for the interaction with HIF-1, a functional HRE usually contains neighboring DNA binding sites for additional transcription factors. Although these elements are not hypoxia inducible per se, they might amplify the hypoxic response or confer tissue-restricted activity to a HRE. Examples include HIF-1 cooperation with the ATF-1/CREB-1 factor in thez lactate dehydrogenase A gene (125
, 136)
, with AP-1 binding factors in the VEGF gene (137)
, and with the orphan receptor hepatic nuclear factor 4 (HNF-4) in the erythropoietin HRE (138)
. The molecular mechanism of the interplay between these distinct transcription factors is manifested by the cooperative binding of CBP/p300 because high-affinity binding of CBP/p300 requires more than one protein–protein interaction. Thus, whereas only the HBS confers hypoxic inducibility, other elements are required to form a fully functional transcriptional enhancer complex.
Alternatively, multimerization of HBSs can form a functional HRE. Two adjacent HBSs were detected in the transferrin HRE (139)
and two or three adjacent HBSs were found in the HREs of genes encoding several glycolytic enzymes, glucose transporter 1, etc. (50)
.
|
O2-REGULATED GENES |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOXIA-INDUCIBLE TRANSCRIPTION...OXYGEN SENSING AND REGULATION...REGULATION OF HIF-1 ACTIVITYO2-REGULATED GENESCONCLUSIONSREFERENCES |
|---|
two dozen HIF-1 target genes are known (Table 2
). However, microarray expression analyses revealed many more potential HIF-1 target genes in pVHL-deficient cells constitutively overexpressing HIF-1 (140)
. Future experiments will show whether these novel genes are direct or indirect HIF-1 targets.
|
HIF-1 target genes involved in iron metabolism
Iron is required for heme formation and is the most common limiting factor in erythropoiesis. Hypoxia was found to increase the expression of transferrin, probably to enhance the iron transport to erythroid tissues (139)
. The transferrin receptor is a hypoxia-inducible HIF-1 target gene, enabling cellular transferrin uptake (141
, 142)
. Finally, ceruloplasmin was shown to be a HIF-1 target gene (143)
. Ceruloplasmin, also known as a ferroxidase, is required to oxidize ferrous to ferric iron. Because only ferric iron can be bound by transferrin, hypoxic ceruloplasmin induction is likely to support iron supply to erythroid tissues.
HIF-1 target genes involved in vascular biology
The vascular system underlies tight oxygen-regulated control. VEGF is the most prominent HIF-1 target gene involved in vascular biology (144
145
146)
. HIF-1 seems to regulate flt-1, one of the receptors for VEGF (147)
, but this finding was questioned in another report (131)
. Recently, an endocrine-gland-derived VEGF (EG-VEGF) was reported that might be a HIF-1 target (148)
. The induction of angiogenesis leads to an increase in the vascular density and hence a decrease in the oxygen diffusion distance. However, local blood flow under pathophysiological conditions is controlled by modulation of the vascular tone through production of NO (inducible nitric oxide synthase), CO (heme oxygenase 1), endothelin 1, adrenomedullin, or activation of the 1B-adrenergic receptor, all of which involve HIF-1 target genes, too (149
150
151
152
153
154)
. Transgenic mice expressing HIF-1 under the control of a keratinocyte-specific promoter develop hypervascularity but, in contrast to VEGF transgenic mice, no vessel leakage or inflammation was observed (155)
. This indicates that HIF-1 mediates angiogenesis by mechanisms far more complex than simple VEGF induction, probably by recruiting additional target genes involved in vessel maturation.
HIF-1 target genes involved in glucose metabolism
Under conditions of limited oxygen supply, anaerobic glycolysis becomes the predominant form of cellular ATP generation (Pasteur effect). Many genes involved in glucose uptake and glycolysis were identified as HIF-1 target genes (50)
. Enhanced lactate production and hence a decrease in pH results from the increase in anaerobic glycolysis, potentially limiting this source of ATP despite sufficient glucose supply. Transmembrane carbonic anhydrases were reported to be HIF-1 target genes (156)
. Carbonic anhydrases regulate the pH by converting protons and bicarbonate to carbon dioxide, which can be taken up by erythrocytes for transportation to the lung.
The constitutive activation of anaerobic metabolism in tumor cells (Warburg effect) provides an explanation for the cell autonomous response to oxygen deficiency (157
, 158)
. Thus, HIF-1 function in hypoxic tumor adaptation involves not only VEGF-mediated angiogenesis, but also increased glycolysis, pH buffering, and probably other key steps in tumor progression.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOXIA-INDUCIBLE TRANSCRIPTION...OXYGEN SENSING AND REGULATION...REGULATION OF HIF-1 ACTIVITYO2-REGULATED GENESCONCLUSIONSREFERENCES |
|---|
schematically summarizes our current knowledge on the functional principles involved in the regulation of HIF-1. Many exciting questions still have to be answered. What are the spatial and temporal functional differences between the three HIF prolyl hydroxylases? What is their preferred oxygen partial pressure? How are they regulated? Will they show a redundant phenotype in knockout mice? What is the nature of the HIF asparaginyl hydroxylase? Where are the HIF phosphoacceptor sites located? How does phosphorylation regulate HIF stability and function? What is the function of HIF-3? And how many further HIF target genes will be uncovered by modern gene array technology? Regarding our understanding of the molecular-physiological hypoxia adaptation, the last few years brought us a big step forward. Regarding a possible application of this knowledge in the treatment of major diseases associated with tissue hypoxia, such as stroke, infarction or cancer, we just stand at the beginning.
|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOXIA-INDUCIBLE TRANSCRIPTION...OXYGEN SENSING AND REGULATION...REGULATION OF HIF-1 ACTIVITYO2-REGULATED GENESCONCLUSIONSREFERENCES |
|---|
. J. Biol. Chem. 271,21262-21267 regulates the VEGF expression and is potentially involved in lung and vascular development. Proc. Natl. Acad. Sci. USA 94,4273-4278 and developmentally expressed in blood vessels. Mech. Dev. 63,51-60[CrossRef][Medline]
-class hypoxia inducible factor subunit, HIF3. Gene Exp. 7,205-213
. J. Biol. Chem. 274,2060-2071. Blood 92,2260-2268 in response to hypoxia: their stabilization and redox signal-induced interaction with CBP/p300. EMBO J 18,1905-1914[CrossRef][Medline]
and HIF-2 in normal human tissues, cancers, and tumor-associated macrophages. Am. J. Pathol. 157,411-421-subunits, HIF1, HIF2 and HIF3, in rat liver. Biochem. J. 354,531-537[CrossRef][Medline]
gene. Biochem. J. 362,71-79[CrossRef][Medline]
. Genes Dev. 12,149-162 is required for solid tumor formation and embryonic vascularization. EMBO J. 17,3005-3015[CrossRef][Medline]
-null embryos is not associated with VEGF deficiency but with mesenchymal cell death. Dev. Biol. 209,254-267[CrossRef][Medline]
. J. Clin. Invest. 103,691-696[Medline]
deficiency impairs pulmonary arterial myocyte electrophysiological responses to hypoxia. Am. J. Physiol. 281,L202-L208. Proc. Natl. Acad. Sci. USA 99,821-826 plays an important role in vascular remodeling. Proc. Natl. Acad. Sci. USA 97,8386-8391 is essential for chondrocyte growth arrest and survival. Genes Dev. 15,2865-2876 during the late stages of mouse spermiogenesis. Mol. Endocrinol. 16,234-243 in activated T lymphocytes. J. Biol. Chem. 276,48754-48763 is dependent on heat shock protein-90 activity. J. Biol. Chem. 277,9262-9267 in response to mechanical stresses: regulation by stretch-activated channels and the phosphatidylinositol 3-kinase signaling pathway. Circ. Res. 90,25e-33e to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 292,468-472 targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 292,464-468 is mediated by an O2-dependent degradation domain via the ubiquitin-proteasome pathway. Proc. Natl. Acad. Sci. USA 95,7987-7992 chains activated by prolyl hydroxylation. EMBO J. 20,5197-5206[CrossRef][Medline]
binding to VHL is regulated by stimulus-sensitive proline hydroxylation. Proc. Natl. Acad. Sci. USA 98,9630-9635 in response to hypoxia is instantaneous. FASEB J. 15,1312-1314 activation cascade: implications for HIF-1 gene therapy. FASEB J. 15,2715-2717 under normoxic and hypoxic conditions. Biochim. Biophys. Acta. 1493,125-134[Medline]
(HIF-1) protein is rapidly degraded by the ubiquitin-proteasome system under normoxic conditions. Its stabilization by hypoxia depends on redox-induced changes. J. Biol. Chem. 272,22642-22647 subunit. J. Biol. Chem. 272,11205-11214 (HIF-) proteins. Biochem. Biophys. Res. Commun. 260,557-561[CrossRef][Medline]
by the ubiquitin-proteasome pathway. J. Biol. Chem. 274,6519-6525 protein expression is controlled by oxygen-regulated ubiquitination that is disrupted by deletions and missense mutations. Proc. Natl. Acad. Sci. USA 97,4748-4753
