| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
T. Kietzmann is at the Institut für Biochemie und Molekulare Zellbiologie, Humboldtallee 23, 37073 Göttingen, Germany. J. Fandrey is at the Institut für Physiologie, IG1, Hufelandstrasse 55, 45147 Essen, Germany. H. Acker is at theMax-Planck-Institut für molekulare Physiologie, Otto-Hahn-Strasse 11, 44227 Dortmund, Germany.
 |
Abstract |
|---|
|
Introduction |
|---|
 Top Introduction O2 sensing and O2...Heme-containing oxidases as O2...H2O2 and derived O2...H2O2 and derived O2...Intracellular localization of...H2O2 and derived O2...Referemces |
|---|
ROS are produced in mitochondria during respiration when O2 is not completely reduced to water. The resulting oxygen intermediates are the superoxide anion radical (O2–•), hydrogen peroxide (H2O2), and the hydroxy radical (OH•). All of these compounds are able to damage membranes, oxidize proteins, and mutate DNA. These hazardous reactions are often referred to as oxidative stress, resulting in cell damage. However, in a limited fashion ROS might have regulatory functions other than to activate the defense mechanisms against ROS-induced cell damage. Since the production of ROS increases proportionally with the O2 tension, they are ideal candidates to act as messengers or mediators involved in the modification of transcription factors modulating gene activity in response to the ambient PO2.
This article will focus mainly on the O2-sensing process that leads to subsequent modulation of gene activity in primary rat hepatocyte cultures and human hepatoma (Hep G2) cells via generation of O2 radicals from H2O2 in an iron-dependent Fenton reaction.
|
O2 sensing and O2 radicals: insights from Escherichia coli |
|---|
|
TopIntroductionO2 sensing and O2...Heme-containing oxidases as O2...H2O2 and derived O2...H2O2 and derived O2...Intracellular localization of...H2O2 and derived O2...Referemces |
|---|
The enteric bacteria E. coli can live under aerobic and anaerobic conditions. Therefore, they must be able to adapt their metabolism to either aerobiosis or anaerobiosis, and they have developed mechanisms that allow the expression of the respective metabolic pathways only if required. In E. coli, the O2-dependent expression of metabolic enzymes is mainly regulated by the transcriptional regulators aerobic respiration control (Arc) A and fumarate nitrate reduction (FNR). The O2 sensory kinase ArcB of the two-component ArcA/ArcB system is located in the cytoplasmic membrane and catalyzes both its own autophosphorylation and the phosphorylation of ArcA. ArcA is a cytoplasmic protein that binds to DNA regulatory elements only when phosphorylated. Interestingly, O2 itself does not trigger the autophosphorylation of ArcB but changes in the redox state of the cytosol via a so-far unknown mechanism (2). Thus PO2 changes are somehow reflected by alterations in the cytosolic redox state.
The transcription factor FNR, which shares similarities with the cAMP receptor protein (CRP) of E. coli, upregulates the expression of up to 30 genes required for anaerobic metabolism in response to low PO2. The wild-type FNR protein was found to contain a redox-active Fe-S cluster that is extremely O2 labile. When isolated, the majority of the anoxically purified protein was a dimer compared with the aerobically purified FNR monomer. Modification of the Fe-S cluster by exposure to O2 led to dissociation of the dimer and conversion to the monomeric form, which displayed markedly decreased DNA-binding activity. Furthermore, neither aerobic respiration nor the respiratory chain was necessary for the inactivation of FNR by O2. Thus O2 itself or its derivatives in the cytoplasm directly react with FNR, and obviously no further proteins are involved in the signal transduction from O2 to FNR (8).
The SoxR protein, the transcriptional activator of the soxS gene that in turn activates the soxRS (superoxide response) regulon of E. coli, is also a sensor for ROS and a transcriptional regulator. SoxR is activated under oxidizing conditions, in contrast to FNR, which is active under anaerobic conditions. The purified SoxR protein is a homodimer containing two oxidized (2Fe-2S) clusters per peptide chain. In contrast to FNR, the DNA-binding affinity of SoxR is not strongly affected by its metal centers or their oxidation state. Instead, activation of the SoxR transcription complex is an allosteric event, which alters the structure of the DNA binding site (5). One-electron oxidation of the SoxR (2Fe-2S) clusters is the mechanism for the sensing of ROS and for activating the SoxR protein as a transcription factor. In this process, as well as in the deactivation of FNR, O2– may act as an oxidant generated in conjunction with the E. coli flavohemoglobin oxidase hemoglobin-like protein, which was proposed to function as an O2 sensor (10). This protein might be the prototype of the so-called heme-based O2 sensors. Similar proteins possessing heme and oxidase activity are found also in eukaryotic cells; one is the mammalian NADPH oxidase.
|
Heme-containing oxidases as O2 sensors in mammalian cells |
|---|
|
TopIntroductionO2 sensing and O2...Heme-containing oxidases as O2...H2O2 and derived O2...H2O2 and derived O2...Intracellular localization of...H2O2 and derived O2...Referemces |
|---|
In both the carotid body preparations and in Hep G2 spheroids, b-type cytochrome absorbance maxima were detected that were O2 sensitive and cyanide insensitive and thus typical of a nonrespiratory chain heme protein. Immunohistochemical staining with antibodies directed against components of the NADPH oxidase complex revealed the existence of proteins with great homology. The detection of H2O2 by rhodamine fluorescence in Hep G2 cells strongly suggested a heme-containing oxidase producing considerable amounts of ROS. Finally, H2O2 production was PO2 dependent, being highest at normoxia and lowest under hypoxia, and strongly supported the notion of the involvement of H2O2 and its derivatives in mediating cellular O2 sensing. Additional studies with pulmonary neuroepithelial bodies, rat carotid body preparations, bovine pulmonary artery cells, and primary hepatocytes confirmed that H2O2 production was O2 dependent within the relevant physiological range (cited in Refs. 7 and 11). The potential effect of PO2-dependent H2O2 production with respect to O2-dependent gene expression was demonstrated for erythropoietin (EPO) production in Hep G2 cells (for review, see Ref. 3). H2O2 production rate was positively correlated with the pericellular PO2. Under normoxia, EPO production is low because of high H2O2 levels. Under hypoxia, however, when H2O2 production decreases, full expression of the gene is permitted.
Considering that an oxidase is involved in the O2-sensing process, one should be able to inhibit oxidase activity and thus disturb O2-regulated gene expression. With the use of the flavin oxidase inhibitor diphenylene iodonium, the role for redox modulations and the involvement of an oxidase in the process of O2 sensing was corroborated: Diphenylene iodonium inhibited the hypoxia-dependent expression of the vascular-endothelial growth factor (VEGF), lactate dehydrogenase A (LDHA), glucose transporter 1 (GLUT1), and the normoxia-dependent placental growth factor (PLGF) gene in several cell lines (4). Still, an oxidase—mitochondrial or cytosolic—that is critical for O2-dependent gene expression via the PO2-dependent production of ROS has not yet been identified. An immediate role for O2 sensing of the phagocyte heme-containing high-output NADPH oxidase, which is composed of gp91, p47phox, p67phox, and p22phox subunits, is unlikely because cell lines deficient in gp91 and p22phox displayed an unimpaired O2-dependent modulation of gene expression (14). However, it does not exclude a NADPH oxidase isoenzyme, a "low-output oxidase," as the O2 sensor that is closely related but clearly distinct from the high-output oxidase in phagocytes (1).
The role of the H2O2 as the signaling molecule in the O2 response has been substantiated in studies investigating the modulation by O2 of aldolase A (ALDA), phosphoenolpyruvate carboxykinase (PCK), glucokinase (GK), and tyrosine hydroxylase (TH) gene expression (Ref. 4; see below).
|
H2O2 and derived O2 radicals in O2 sensing in primary hepatocytes |
|---|
|
TopIntroductionO2 sensing and O2...Heme-containing oxidases as O2...H2O2 and derived O2...H2O2 and derived O2...Intracellular localization of...H2O2 and derived O2...Referemces |
|---|
Rat hepatocytes can be isolated and brought into culture in which they serve as an excellent model for investigation of PO2-dependent gene expression. According to their zonated representation in the liver, in primary rat hepatocyte cultures the glucagon-dependent activation of the PCK gene was maximal under a PO2 that mimicked the high or periportal O2 tension and only half maximal under a PO2 that mimicked the low or perivenous O2 tensions. In contrast, insulin activated the gene for GK maximally under low or perivenous PO2 and only half maximally under a high PO2 such as in the periportal area (7). As outlined above, H2O2 has been proposed as a putative intracellular messenger between the O2 sensor and the transcriptional machinery controlling the EPO gene in human hepatoma cells (3). Likewise, primary rat hepatocytes released H2O2 depending on the O2 tension. The highest H2O2 concentration (~1.2 nM) was measured under high or periportal PO2 (~70 mmHg), whereas under low or perivenous PO2 (35 mmHg) only concentrations of ~0.8 nM were recorded. If H2O2 is the intracellular messenger from the O2 sensor, the external application of H2O2 to cells cultured under low perivenous PO2 should mimic a high PO2, resulting in an induction of normoxia-activated genes and in a suppression of hypoxia-activated genes. This was indeed found: in primary rat hepatocyte cultures, H2O2 mimicked the (higher) periportal PO2 with respect to the modulation of the glucagon-dependent activation of the PCK gene and the insulin-dependent activation of the GK gene. Maximal PCK and lowest GK mRNA levels were observed when the H2O2 concentration was highest and vice versa when H2O2 was scavenged by simultaneous administration of catalase, as shown in Fig. 1
(7).
|
Fe3+ + OH– + OH•). This reaction is potentially of great importance in the O2 signaling cascade since the OH• are highly reactive, with a diffusion area of only a few nanometers. OH• could be generated in the ultimate vicinity of Fe-containing residues of protein or even transcription factors and directly react at its site of generation with Fe-S clusters or cysteine residues of regulatory proteins (3, 11). Studies in primary hepatocytes provide an example of the involvement of the Fenton reaction. It was found that in hepatocytes the Fenton reaction-mediated formation of OH• was maximal under arterial PO2 and reduced to about half maximal under venous PO2. Impairment of the Fenton reaction by short-term treatment of hepatocytes cultured under arterial PO2 with the iron chelator desferrioxamine (DSF) and the OH• scavenger dimethylthiourea (DMTU) resulted in a similar decrease of OH• formation as usually measured by reducing the PO2 to venous PO2 values and respective reciprocal O2-dependent modulation of the glucagon-dependent induction of the PCK mRNA and the insulin-dependent induction of the GK mRNA. Despite an arterial PO2, the addition of DSF and DMTU lowered the glucagon-induced PCK mRNA levels to a degree otherwise seen under low venous PO2 conditions. Reciprocally, insulin-elevated GK mRNA levels were increased to values usually observed under low-PO2 conditions (7). Thus it seems likely that a local Fenton reaction was involved in the reciprocal modulation of the glucagon-activated PCK gene and the insulin-activated GK gene by O2.
|
H2O2 and derived O2 radicals in O2 sensing in Hep G2 cells |
|---|
|
TopIntroductionO2 sensing and O2...Heme-containing oxidases as O2...H2O2 and derived O2...H2O2 and derived O2...Intracellular localization of...H2O2 and derived O2...Referemces |
|---|
To further support this hypothesis, the H2O2-sensitive dye dihydrorhodamine 123 was applied as an OH• scavenger to Hep G2 cells. Dihydrorhodamine 123 is a nonfluorescent agent that scavenges the OH• generated from H2O2 in an iron-dependent Fenton reaction and is thereby converted into the fluorescent rhodamine 123 (11). Thus dihydrorhodamine 123 should be able to counteract the H2O2-mediated effects. It was found that in Hep G2 cells exposed to moderate hypoxia (PO2 ~ 35 mmHg) the addition of dihydrorhodamine 123 completely antagonized the inhibitory action of exogenous H2O2 on EPO production (Fig. 2). Moreover, under normoxia EPO production could be increased on application of dihydrorhodamine 123 alone, confirming the role of H2O2 and OH• in repressing EPO production (Fig. 2). Thus scavenging of OH• mimics hypoxic conditions and releases the inhibition of EPO gene expression under a high PO2.
|
|
Intracellular localization of OH• formed by Fenton reaction in primary rat hepatocytes and Hep G2 cells |
|---|
|
TopIntroductionO2 sensing and O2...Heme-containing oxidases as O2...H2O2 and derived O2...H2O2 and derived O2...Intracellular localization of...H2O2 and derived O2...Referemces |
|---|
; Refs. 7 and 11). This very localized perinuclear fluorescence in both cell types has not been demonstrated so far and can therefore be considered the first three-dimensional visualization of an intracellular Fenton reaction.
|
) located in the cytoplasm of primary hepatocytes and Hep G2 cells near the nucleus were undoubtedly identified as containing iron by corresponding X-ray spectra (Fig. 3, E and F). The X-ray microanalysis in these "cytoplasmic dense particles" did not provide information about the ionic state of the iron or whether the iron is bound to proteins. However, it was possible to colocalize iron and dihydrorhodamine 123 conversion by H2O2 in an iron-mediated Fenton reaction to the close vicinity of the nucleus. |
H2O2 and derived O2 radicals in the regulation of transcription factor activity |
|---|
|
TopIntroductionO2 sensing and O2...Heme-containing oxidases as O2...H2O2 and derived O2...H2O2 and derived O2...Intracellular localization of...H2O2 and derived O2...Referemces |
|---|
). Candidate targets of H2O2-derived O2 radicals are transcription factors like hypoxia-inducible factor 1 (HIF-1), nuclear factor-
B (NF-B) or activator protein 1 (AP-1), which might be modified by OH• since all of these protein complexes are known to be redox sensitive (for review, see Ref. 2).
|
and a HIF-1ß-aryl hydrocarbon receptor nuclear translocator (ARNT) subunit, both of which are members of the basic helix-loop-helix-Per, ARNT, Sim (PAS) family of transcription factors. This family of transcription factors has been growing because of a set of recent cloning experiments. One of the cloned factors, which was first termed endothelial Per, ARNT, Sim (PAS) protein-1, (EPAS-1) shared a 48% homology to HIF-1. This molecule was cloned independently by several groups and also was called HIF-related factor (HRF), HIF-like factor (HLF), and member of the PAS superfamily-2 (MOP-2). According to its high homology to HIF-1, its ability to dimerize with HIF-1ß-(ARNT), and its ability to induce transcription at the same DNA-response element as HIF-1, it was suggested that this protein be named HIF-2 (13). HIF-1 was found to be redox sensitive. In HeLa and Hep 3B cells, H2O2 destabilized the HIF-1 protein, which lead to decreased binding of the HIF-1 complex to the EPO enhancer DNA element (6). Recently, H2O2 was found not only to reduce the hypoxia-induced HIF-1 protein accumulation but also the induction by hypoxia of the HIF-2 (EPAS-1) protein in HeLa cells (15). According to the notion that an oxidase could be involved in generating ROS, the authors concluded that the abundance of both HIF-1 and HIF-2 appears to be regulated by strikingly similar O2-sensing and signal transduction pathways (15). The regulation of HIF-1 DNA-binding activity by ROS is apparently evolutionarily conserved. In nuclear extracts from the Drosophila melanogaster SL2 cell line cultured for 4-16 h under hypoxia, a hypoxia-inducible complex binding to the mouse EPO-HIF element and the PGK-HIF element was present. This Drosophila complex, called HIF-D, showed sensitivity to redox modifications. Omission of the reducing agent dithiothreitol from the buffers during nuclear extract preparation and in the binding buffer abolished HIF-D binding to the mouse EPO enhancer. Addition of the oxidative agent diamide pronounced this effect. HIF-D-binding activity could subsequently be restored after dithiothreitol addition (12).
Thus mimicry of high PO2 by H2O2 leads to a destabilization of HIF-1 and HIF-2 and of (probably redox-modulated) proteasomal degradation, resulting in a suppression of hypoxia-dependent gene activation. The nature of the ROS interacting with HIF-1 or HIF-2 and the reaction mechanism is yet unknown. It is tempting to speculate that under an arterial PO2 with high cellular H2O2 levels, an iron-dependent Fenton reaction could take place near the HIF-1 or HIF-2 proteins. Exploiting the extremely short diffusion distances of the generated OH•, a very localized oxidation would lead to an oxidation and thus to the destabilization of the HIF-1 protein and HIF-2 protein.
The transcription factor AP-1, consisting of dimers formed by members of the Fos, Jun, or ATF protein families, was also shown to be involved in the modulation by O2 of gene expression. It was found that AP-1 was hypoxia inducible in its DNA-binding activity and in its transactivation capability in HeLa and Hep 3B cells. In transient transfection studies with PC-12 cell gene constructs containing the TH promoter from nucleotides (nt) –284 to + 27 fused to the chloramphenicol acetyltransferase (CAT) gene, the region from nt –284 to –150 was found to confer O2 regulation. This region contains an AP-1, AP-2, and HIF-1 elements. Mutation of the AP-1 sequence prevented stimulation of transcription of the TH-CAT gene by hypoxia. Gel shift assays revealed enhanced protein interactions at the AP-1 and HIF-1 elements of the native gene. Supershift assays showed that c-Fos and JunB bind to the AP-1 element during hypoxia and that these protein levels were stimulated by hypoxia. The role of the AP-1 proteins in the regulation of gene activity by hypoxia was substantiated in a study in which human VEGF promoter luciferase gene constructs were investigated in C6 glioma cells. It was found that an AP-1 site (–1129/–1123) was involved in potentiating the induction by hypoxia mediated by HIF-1. In the case of the AP-1 complex, cysteine residue 154 in c-Fos and cysteine residue 272 in c-Jun are responsible for redox sensitivity of the two proteins. Both dimerization partners can be converted to an inactive state by chemical oxidation of these cysteine residues. Thus, like HIF-1, AP-1 is a good candidate where the above-described Fenton reaction could be important (for review, see Ref. 2). In contrast, the dimerization and enhanced DNA-binding activity of NF-B under high levels of ROS (9) should favor gene expression under high PO2 levels. However, the role of H2O2 and derived OH• in the activation of NF-B subunits has not yet been completely elucidated, and the mechanism of NF-B activation in response to O2 is poorly understood.
Derived from the above-described findings, one can draw the following model for the signaling cascade regulating O2-dependent gene activity (Fig. 4). The PO2 is measured by the cells via a heme protein. This sensor produces H2O2 in a proportion directly dependent on the PO2. The produced H2O2 reaches concentrations that are well below the levels exerting oxidative stress but can freely diffuse inside the cell. To mediate the O2 response, H2O2 is degraded in an iron-dependent Fenton reaction, yielding OH– and the highly reactive OH•. Under a high PO2, the produced OH• can oxidize sulfhydroxyl groups in certain candidate transcription factors, thus shifting the balance between reduced and oxidized forms to the oxidized state. The transcription factor may then bind to the normoxia-response elements in genes that are preferentially expressed at a high PO2. Low O2 tensions reduce the OH• levels, and binding of the reduced transcription factor to the hypoxia-responsive elements in several genes again initiates the O2 modulation of either basal expression, e.g., the EPO gene, or, as in the case with GK, insulin (hormone)-dependent expression (Fig. 4). The role of different types of ROS as intracellular messenger molecules in control of gene transcription, cytokine, growth factor, and hormone production, and action, ion transport, neuromodulation, and apoptosis is now widely recognized in the literature. In addition, ROS generated in a reaction dependent on PO2 seem to play a significant physiological role in the O2-sensing process that allows fine tuning of cellular function according to tissue PO2 distribution.
|
Referemces |
|---|
|
TopIntroductionO2 sensing and O2...Heme-containing oxidases as O2...H2O2 and derived O2...H2O2 and derived O2...Intracellular localization of...H2O2 and derived O2...Referemces |
|---|
. Blood 92: 2260–2268, 1998.This article has been cited by other articles:
|
|
 |
|
  M. F. Essop Cardiac metabolic adaptations in response to chronic hypoxia J. Physiol., November 1, 2007; 584(3): 715 - 726. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Freyssenet Energy sensing and regulation of gene expression in skeletal muscle J Appl Physiol, February 1, 2007; 102(2): 529 - 540. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Hoffman, J. D. Salter, and P. S. Brookes Response of mitochondrial reactive oxygen species generation to steady-state oxygen tension: implications for hypoxic cell signaling Am J Physiol Heart Circ Physiol, January 1, 2007; 292(1): H101 - H108. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Fandrey, T. A. Gorr, and M. Gassmann Regulating cellular oxygen sensing by hydroxylation Cardiovasc Res, September 1, 2006; 71(4): 642 - 651. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-G. Scharf, T. G. Unterman, and T. Kietzmann Oxygen-Dependent Modulation of Insulin-Like Growth Factor Binding Protein Biosynthesis in Primary Cultures of Rat Hepatocytes Endocrinology, December 1, 2005; 146(12): 5433 - 5443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Droge Oxidative Aging and Insulin Receptor Signaling J. Gerontol. A Biol. Sci. Med. Sci., November 1, 2005; 60(11): 1378 - 1385. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Fandrey Oxygen-dependent and tissue-specific regulation of erythropoietin gene expression Am J Physiol Regulatory Integrative Comp Physiol, June 1, 2004; 286(6): R977 - R988. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Liu, U. Berchner-Pfannschmidt, U. Moller, M. Brecht, C. Wotzlaw, H. Acker, K. Jungermann, and T. Kietzmann A Fenton reaction at the endoplasmic reticulum is involved in the redox control of hypoxia-inducible gene expression PNAS, March 23, 2004; 101(12): 4302 - 4307. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Dahan and L. J. Teppema Influence of anaesthesia and analgesia on the control of breathing Br. J. Anaesth., July 1, 2003; 91(1): 40 - 49. [Full Text] [PDF]
|
|
|
|
 |
|
 Z.-Z. Yang, A. Y. Zhang, F.-X. Yi, P.-L. Li, and A.-P. Zou Redox regulation of HIF-1alpha levels and HO-1 expression in renal medullary interstitial cells Am J Physiol Renal Physiol, June 1, 2003; 284(6): F1207 - F1215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. J Teppema, D. Nieuwenhuijs, E. Sarton, R. Romberg, C. N Olievier, D. S Ward, and A. Dahan Antioxidants prevent depression of the acute hypoxic ventilatory response by subanaesthetic halothane in men J. Physiol., November 1, 2002; 544(3): 931 - 938. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Raj and L. Shimoda Oxygen-dependent signaling in pulmonary vascular smooth muscle Am J Physiol Lung Cell Mol Physiol, October 1, 2002; 283(4): L671 - L677. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
