HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 38, 34773-34784, September 20, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/38/34773    most recent M203902200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dirmeier, R. Articles by Poyton, R. O. Search for Related Content PubMed PubMed Citation Articles by Dirmeier, R. Articles by Poyton, R. O. Social Bookmarking                      What's this?

Exposure of Yeast Cells to Anoxia Induces Transient Oxidative Stress

IMPLICATIONS FOR THE INDUCTION OF HYPOXIC GENES^*

Reinhard Dirmeier, Kristin M. O'Brien, Marcella Engle, Athena Dodd, Erick Spears, and Robert O. Poyton^§

From the Department of Molecular, Cellular, and Developmental Biology, University of Colorado, Boulder, Colorado 80309-0347 and  AMC Cancer Research Center, Denver, Colorado 80214

Received for publication, April 22, 2002, and in revised form, June 23, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The mitochondrial respiratory chain is required for the induction of some yeast hypoxic nuclear genes. Because the respiratory chain produces reactive oxygen species (ROS), which can mediate intracellular signal cascades, we addressed the possibility that ROS are involved in hypoxic gene induction. Recent studies with mammalian cells have produced conflicting results concerning this question. These studies have relied almost exclusively on fluorescent dyes to measure ROS levels. Insofar as ROS are very reactive and inherently unstable, a more reliable method for measuring changes in their intracellular levels is to measure their damage (e.g. the accumulation of 8-hydroxy-2'-deoxyguanosine (8-OH-dG) in DNA, and oxidative protein carbonylation) or to measure the expression of an oxidative stress-induced gene, e.g. SOD1. Here we used these approaches as well as a fluorescent dye, carboxy-H₂-dichloro-dihydrofluorescein diacetate (carboxy-H₂-DCFDA), to determine whether ROS levels change in yeast cells exposed to anoxia. These studies reveal that the level of mitochondrial and cytosolic protein carbonylation, the level of 8-OH-dG in mitochondrial and nuclear DNA, and the expression of SOD1 all increase transiently during a shift to anoxia. These studies also reveal that carboxy-H₂-DCFDA is an unreliable reporter of ROS levels in yeast cells shifted to anoxia. By using two-dimensional electrophoresis and mass spectrometry (matrix-assisted laser desorption ionization time-of-flight), we have found that specific proteins become carbonylated during a shift to anoxia and that some of these proteins are the same proteins that become carbonylated during peroxidative stress. The mitochondrial respiratory chain is responsible for much of this carbonylation. Together, these findings indicate that yeast cells exposed to anoxia experience transient oxidative stress and raise the possibility that this initiates the induction of hypoxic genes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In many organisms, adaptation to changing oxygen concentrations is achieved both by the short term effects of oxygen on energy metabolism and the long term effects of oxygen on the expression of oxygen-responsive genes in the nucleus (1, 2). These oxygen-regulated nuclear genes can be placed into one of the two following groups: aerobic genes, which are transcribed optimally in the presence of air; and hypoxic genes, which are transcribed optimally under anoxic or micro-aerophilic conditions. In some case, proteins have both aerobic and hypoxic isoforms that are interchangeable but functionally different (3, 4). The oxygen-responsive transcription factors that regulate these genes have been identified in a number of different organisms (5-8), but despite a great deal of progress in understanding how these transcription factors function, the nature of the more proximal events involved in oxygen sensing have remained elusive. Two fundamentally different types of model have been proposed. In the first model, oxygen has a direct effect on transcription either by acting on a transcription factor itself or by affecting a component that interacts with a transcription factor (8-11). In a second model, oxygen is sensed more distally by a proximal oxygen sensor, a signal is produced, and the signal initiates a signal transduction cascade that affects distal transcription factors (for review see Ref. 12).

Two different hemoproteins have been implicated as possible oxygen sensors for the induction of hypoxic genes in eucaryotic cells. One is a multisubunit cytochrome b-like NAD(P)H oxidase present in the plasma membrane of mammalian cells. It has been proposed that in the presence of oxygen this protein generates superoxide, which is converted to hydrogen peroxide and other reactive oxygen species (ROS)¹ that serve to facilitate degradation of Hif-1, a subunit of the global hypoxic transcription factor Hif-1 (13). This leads to the suppression of expression of Hif-1-regulated hypoxic genes under normoxic conditions. According to this model, ROS levels drop in cells exposed to reduced atmospheric oxygen; this leads to the stabilization of Hif-1 and the induction of Hif-1-regulated hypoxic genes. The second hemoprotein that has been implicated in the induction of hypoxic genes is cytochrome c oxidase. Studies with yeast using cytochrome c oxidase-deficient yeast nuclear mutants, a respiratory-deficient ^o mutant, or respiratory inhibitors have shown that cytochrome c oxidase and the mitochondrial respiratory chain are required for the induction of nuclear hypoxic genes when cells are exposed to anoxia (14, 15). Similar studies with mammalian tissue culture cells, using respiratory-deficient ^o cells, respiratory inhibitors, and xenomitochondrial cybrids, have reported that the respiratory chain is required for the induction of hypoxic genes in Hep3B hepatoma cells (16) and osteosarcoma cells (17). Studies with Hep3B cells have also reported that the respiratory chain produces reactive oxygen species when cells are shifted to hypoxic conditions, that the concentrations of these ROS increase throughout the cell, and that this increase in ROS concentration serves to stabilize Hif-1, which then leads to the induction of Hif-1-dependent nuclear hypoxic genes (16, 18). It is not clear if the respiratory chain is required for hypoxic gene induction in all mammalian cell types (e.g. see Refs. 19 and 20). Also unclear is whether ROS influence the activity of the prolyl or asparagine hydroxylases that have recently been implicated in the stabilization and function of Hif-1 (8, 21).

The models that have been proposed for how cytochrome b-like NAD(P)H oxidase and cytochrome c oxidase function as oxygen sensors for Hif-1-dependent induction of hypoxic nuclear genes in mammalian cells have both focused on ROS as intermediaries that connect a proximal oxygen sensor with the distal transcription factor. However, they are diametrically opposed with regards to the effects of hypoxia on intracellular ROS levels. Whereas the first model requires that ROS levels decrease in cells exposed to hypoxia, the second model requires that they increase. Experiments designed to determine how hypoxia affects ROS levels in mammalian cells have produced conflicting results, often with the same cell type. For example, one group has reported that ROS levels in Hep3B hepatoma cells increase upon exposure to hypoxia (18), whereas another group has reported that ROS levels decrease in HepG2 (22) or Hep3B hepatoma cells to hypoxia (23). These studies used fluorescent dyes (either 2',7'- dichlorofluorescein diacetate or dihydrorhodamine 123) to measure ROS levels in cells exposed to hypoxia. Although these dyes are widely used to assess intracellular H₂O₂ levels, the mechanism by which they become oxidized is not entirely clear (24). Because these dyes produced conflicting results in mammalian cells shifted from normoxic to hypoxic conditions, it has been suggested that they lack the necessary precision to monitor ROS levels in cells experiencing changes in oxygen concentration (25). This would not be surprising because ROS are transient, highly reactive, and probably oxidize macromolecules (proteins, lipids, and nucleic acids) in their immediate vicinity.

To help resolve the controversy concerning how ROS levels change in eucaryotic cells exposed to low levels of oxygen and, at the same time, to begin to address whether ROS are involved in hypoxic gene induction in yeast, we have shifted cells from normoxic to anoxic conditions and have monitored oxidative stress in four different ways: 1) by using carboxy-H₂-DCFDA, which becomes oxidized upon exposure to hydrogen peroxide and ROS; 2) by assessing levels of oxidative protein and 3) DNA damage; and 4) by examining the expression of SOD1, an oxidative stress-induced gene.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Yeast Strains and Media-- The following Saccharomyces cerevisiae strains were used: JM43 (Mat his4-580 leu2-3, trp1-289, 112 ura3-52 [⁺]) (26); JM43 ⁰ (Mat his4-580 trp1-289 leu2-3, 112 ura3-52 [⁰] (27)); and JM43 GD100 (28). They were grown in SSG-TEA, supplemented with Tween 80, ergosterol, silicon antifoam, and amino acids and uracil, as needed (15).

Growth Conditions-- Aerobic cultures and pre-cultures were grown in a shaker (200 rpm) at 28 °C and harvested in logarithmic growth phase. For shift experiments between aerobic and anaerobic conditions, mid-logarithmic phase aerobic pre-cultures were inoculated into a New Brunswick Scientific BioFloIIc fermentor, equipped with a gas-flow train that allows cells to be cultured to steady state at defined oxygen concentrations (29). These fermentor cultures were grown in air to a cell density of ~2 × 10⁷ cells per ml; then the gas entering the fermentor vessel was changed to O₂-free N₂ containing 2.5% CO₂. Anaerobic cultures were grown in the fermentor in the dark at 28 °C and 200 rpm and sparged with a gas mixture containing 97.5% N₂ and 2.5% CO₂. The gas mixture was passed through an Oxyclear O₂ absorber (Lab-Clear, Oakland, CA) to prevent trace O₂ from entering the fermentor. Temperature, pH, sparge rate, and dissolved oxygen concentration in the fermentor were monitored and maintained as described (15, 29). Both anoxic and shift cultures were grown in the dark to prevent photoinhibition (30). Cell growth was followed by measuring turbidity with a Klett-Summerson colorimeter (No. 54 green filter), and cells were harvested in mid-logarithmic phase. During harvest, cultures were chilled to 4 °C, washed twice with ice-cold distilled water, and either processed immediately or frozen in liquid nitrogen.

Preparation of Mitochondrial and Cytosolic Cell Fractions-- Cells were spheroplasted as described (3), except that bovine serum albumin was omitted from the post-spheroplast and lysis buffers for experiments in which the carbonyl content in cytosolic or mitochondrial proteins was quantitated. All steps after spheroplasting were performed at 4 °C. The spheroplasts were harvested by centrifugation (5 min at 3000 × g), washed gently in post-spheroplast buffer (1.5 M sorbitol, 1 mM Na₂EDTA, 0.1% bovine serum albumin, pH 7.0), and sedimented at 3000 × g. The pelleted cells were resuspended in lysis buffer (0.6 M mannitol, 2 mM Na₂EDTA, 0.1% bovine serum albumin, pH 7.4) and lysed in a Sorvall Omnimixer (3 s at low speed and 25 s at full speed). The cell lysate was then centrifuged for 5 min at 1,900 × g to pellet unbroken cells, nuclei, and debris. The supernatant contained the mitochondria as well as cytosolic proteins. It was decanted and centrifuged for 10 min at 12,100 × g to pellet the mitochondria. The resulting supernatant was saved as the cytosolic fraction and frozen in liquid nitrogen. The mitochondrial pellet was washed in mitochondrial lysis buffer minus bovine serum albumin, pH 7.0, homogenized with a glass-Teflon homogenizer, and centrifuged at 1651 × g for 5 min to pellet cell debris that was trapped in the pellet. The supernatant, containing the mitochondria, was decanted and centrifuged at 23,000 × g for 10 min. The mitochondrial pellet was resuspended in Milli-Q H₂O and immediately frozen in liquid nitrogen.

Isolation of Nuclear and Mitochondrial DNA-- Mitochondrial DNA was obtained from mitochondria, isolated as above, using the method of Querol and Barrio (31). Nuclear DNA was isolated using the Y-DER^TM YEAST DNA Extraction Kit (Pierce).

Assay of Reactive Oxygen Species with Carboxy-H₂-DCFDA-- Carboxy-H₂-DCFDA was used to try to assess hydrogen peroxide levels in aerobic JM43 cultures, anaerobic JM43 cultures, aerobic JM43^o cultures, and in aerobic JM43 cultures treated with peroxide or peroxide plus the spin-trap N-tert-butyl--phenylnitrone. Carboxy-H₂-DCFDA, dissolved in Me₂SO, was added to cultures at a final concentration of 10 µM 1 h before measurements were taken. For measurements taken during a shift from normoxic to anoxic conditions, carboxy-H₂-DCFDA was added to cultures at a final concentration of 11.8 µM 105 min prior to the shift, and samples were taken every 20 min for 5 h. For these shift studies, total intracellular levels of carboxy-H₂-DCFDA were estimated by adding 1% hydrogen peroxide to aliquots of cells sampled from the fermentor every 20 min after the shift and measuring fluorescence as described below. Fluorescence was measured in three 100-µl aliquots of either steady state or shifted cells that were diluted 10-fold in 50 mM NaPO₄, pH 7.0, and sonicated briefly to disperse cells. Fluorescence (wavelength, 515-545 nm) of 5,000 cells was measured in each aliquot (15,000 total) using a BD PharMingen fluorescent-activated cell sorter equipped with a 15-milliwatt argon laser.

Determination of 8-Hydroxy-2'-deoxyguanosine and 2-Deoxyguanosine Levels in DNA Samples-- Isolated mitochondrial or nuclear DNA was hydrolyzed as described (32), using nuclease P1 and alkaline phosphatase. 8-Hydroxy-2'-deoxyguanosine (8-OH-dG) levels were quantified using the isocratic high pressure liquid chromatography (HPLC) method described by McCabe et al. (33). All data were analyzed using ESA (Chelmsford, MA) CoulArray for Windows software and expressed in terms of the ratio 8-OH-dG:10⁵ dG. Reagents were HPLC grade, and Milli-Q water was further purified using a Waters Sep-Pak solid phase extraction column.

Quantitation of Protein-Carbonyl Content-- Carbonyl content of mitochondrial and cytosolic protein fractions was determined after derivatization with 2,4-dinitrophenyl hydrazine (DNPH) (34). Nucleic acids in cytosolic fractions were removed by treatment with streptomycin sulfate, prior to the addition DNPH (35). Derivatized protein samples were fractionated by HPLC on a Waters model 626 fitted with a Waters 600S controller and a Waters 996 Photodiode Array Detector, using Zorbax GF 450 and Zorbax GF 250 gel filtration columns run in series. The columns were calibrated and eluted (flow rate 1 ml/min) with 6 M guanidinium hydrochloride, 0.5 M KPO₄, pH 2.5. Eluants were monitored for absorbance at 366 and 280 nm (34).

SDS-PAGE of DNP-derivatized Proteins-- Mitochondrial and cytosolic protein samples for SDS-PAGE were prepared as follows. One volume of sample containing 5 or 10 µg of protein was added to an equal volume of 12% SDS. One additional volume of 20 mM DNPH in 10% trifluoroacetic acid was added, and the mixture was incubated for 35 min at room temperature. The derivatization reaction was stopped by adding 1.05 volumes of 2 M Tris base, 30% glycerol (final concentration of 0.52 M). Derivatized protein samples were separated on 10% polyacrylamide gels (resolving gel: 10% (w/v) 32:1 acrylamide:bisacrylamide, 0.1% (w/v) SDS, 0.4 M Tris, pH 8.8; stacking gel: 3.5% (w/v) 32:1 acrylamide:bisacrylamide 0.1% (w/v) SDS, 0.125 M Tris, pH 6.8) Gels were run at 110 V until the yellow front of underivatized DNPH reached the bottom of the gel.

Two-dimensional Gel Electrophoresis of Carbonylated Proteins-- Aliquots of mitochondria (125 µg of protein) or cytosol (70-90 µg of protein) were diluted 5-fold in lysis buffer containing 9 M urea, 2.5 M thiourea, 5% (w/v) CHAPS, 12.5 mM DTT, and 1% (v/v) carrier ampholytes, pH 3-10, and incubated for 1 h at room temperature. Rehydration buffer (7 M urea, 2 M thiourea, 4% (w/v) CHAPS, 10 mM DTT, 0.5% (v/v) Triton X-100, 1% (v/v) carrier ampholytes, pH 3-10, and a trace of bromphenol blue) was then added to bring the total volume up to 250 µl. The samples were incubated at room temperature for 20 min and then used to rehydrate 13-cm Immobiline Drystrips (Amersham Biosciences) with a pH 3-10 linear gradient. The strips were rehydrated overnight in an Immobiline Drystrip Reswelling Tray (Amersham Biosciences). First dimension isoelectric focusing was carried out at 20 °C using a Multiphor II flatbed system (Amersham Biosciences) at 300 V for 1 min, 300 V for 4 h, 3000 V for 5 h, and 3000 V for 10 h with a EPS 3501 xl (Amersham Biosciences) electrophoresis power supply. After isoelectric focusing, the Immobiline strips were derivatized with DNPH using a modification of the method of Reinheckel et al. (36). They were incubated in a solution of 10 mM DNPH in 10% trifluoroacetic acid for 20 min with agitation and then incubated with agitation for 10 min at room temperature in each of the following solutions: 1) 150 mM Tris-HCl, pH 6.8, 8 M urea, 20% (v/v) glycerol, and 2% (w/v) SDS; 2) 150 mM Tris-HCl, pH 6.8, 8 M urea, 20% (v/v) glycerol, 2% (w/v) SDS, and 1% (w/v) DTT; and 3) 150 mM Tris-HCl, pH 6.8, 8 M urea, 20% (v/v) glycerol, 2% (w/v) SDS, 4% (w/v) iodoacetamide and a trace amount of bromphenol blue. The Immobiline strips were loaded onto a 20 × 20-cm polyacrylamide gel composed of 8-18% (w/v) 37.5:1 acrylamide:bisacrylamide, 0.37 M Tris, pH 8.8. The 20 × 20-cm gradient gels were poured using a Gradient Mixer GM-1 (Amersham Biosciences) and a Protean II xi Multi-Gel Casting Chamber (Bio-Rad). The gels were run at 10 °C on Protean II xl m Multicell Unit (Bio-Rad) at a constant 15 mA per gel for ~17 h in a running buffer containing 40 mM glycine, 200 mM Tris, and 0.1% SDS. After electrophoresis, proteins were visualized by silver staining (37).

Western Immunoblotting-- Proteins were transferred onto Hybond ECL (Amersham Biosciences) nitrocellulose membranes using a Semi-Phor (Hoefer) Blotting apparatus. Oxidatively modified proteins were detected with anti-DNP antibodies (Dako) (at a dilution of 1:4000) and secondary horseradish peroxidase-linked antibodies (at a dilution of 1:25000), followed by a chemiluminescence reaction using a chemiluminescence detection kit (PerkinElmer Life Sciences). Gels were scanned, and the resulting images were evaluated using Melanie 3 (Geneva Bioinformatics, Geneva, Switzerland) software.

Protein Identification by MALDI Mass Spectrometry-- Silver-stained spots were cut out of the gels for in-gel digestion and destained with 1 ml of 50 mM sodium thiosulfate, 15 mM potassium ferricyanide followed by 4 washes in 1 ml of Milli-Q H₂O. The spots were then equilibrated for 20 min in 500 µl of 100 mM ammonium bicarbonate and then incubated for 20 min in 500 µl of 50% acetonitrile, 50 mM ammonium bicarbonate. The spots were dried, rehydrated for digestion with 5 µg/ml porcine trypsin (Promega, Madison, WI) in 25 mM ammonium bicarbonate, and incubated at 37 °C overnight. The reaction was stopped by adding 1 µl of 88% formic acid. The peptides were extracted from the gel matrix by sonication for 20 min and then concentrated using Zip Tips (Millipore Corp., Bedford, MA). Peptide mass figure printing was performed using a PerkinElmer Life Sciences Voyager-DE STR, operating in delayed reflector mode at an accelerating voltage of 20 kV. The peptide samples were co-crystallized with matrix on a gold-coated sample plate using 0.6 µl of matrix (-cyano-4-hydroxytranscinnamic acid) and 0.6 µl of sample. After external calibration with protein standards, the spectra were internally calibrated using trypsin autolysis products. The monoisotope peptide masses were assigned and then used in data base searches with PeptIdent (ca.expansy.org/tools/peptident.html). Cysteines were treated with iodoacetamide to form carboxyamidomethyl cysteine, and methionine was considered to be oxidized. One missed cleavage was allowed and a minimum of 4 matching peptides was required for a match. A pI range of ± 2.00 and a molecular weight range of ±20% were used for the data base search. Criteria used to identify each protein included the extent of sequence coverage, the number of peptides matched, and the match with the theoretical pI and M_r as determined from the gel.

RNA Isolation and Northern Blotting-- Total RNA, isolated as described previously (38), was denatured and separated on 1.5% agarose gels containing MOPS/formaldehyde buffer (20 mM MOPS, 40 mM sodium acetate, 8 mM EDTA, and 220 mM formaldehyde (39)). The RNA was transferred to Nytran Plus membranes (Schleicher & Schuell) and hybridized as described (40) using DNA probes to ACT1 and SOD1. The probes (a 520-bp StyI fragment of ACT1, and a 1041-bp fragment made to correspond to the SOD1-coding sequence and 41 nucleotides of its 5'-flanking sequence) were prepared by random-primer labeling of double-stranded DNA fragments using [-³²P]dCTP (PerkinElmer Life Sciences) (cf. Ref. 41). Stringency washes were performed as described previously (42) and according to the manufacturer's recommendations.

Miscellaneous-- Protein concentration was determined according to Lowry et al. (43). All gases were Matheson certified standards and were obtained from Airgas (Los Angeles, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Recent studies with mammalian cells have produced conflicting results concerning whether exposure to hypoxia is a form of oxidative stress (18, 22, 23). These studies have relied on fluorescent dyes to measure H₂O₂ and other ROS. It is not clear yet if the discrepant results come from the use of different protocols with these dyes or from inherent problems (e.g. cellular permeability, dye leakage, or inaccessibility to all cellular compartments) associated with the dyes themselves. Because ROS are very reactive and inherently unstable, it has been suggested that a more reliable method for measuring changes in their intracellular levels is to measure their damage, e.g. by measuring the accumulation of oxidized nucleosides in DNA or oxidized amino acid side chains on proteins (44). Here we used both approaches in order to determine whether, and how, ROS levels change in yeast cells exposed to anoxia.

Measurement of ROS Levels in Whole Yeast Cells with Carboxy-H₂-DCFDA-- Previous studies with yeast have used carboxy-H₂-DCFDA in conjunction with a fluorescent-activated cell sorter to assess H₂O₂ levels in cells during apoptosis (45), during ischemia reperfusion (46), and in farnesol-treated cells (47). This carboxylated analog of 2',7'-dichlorofluorescein has been reported to be better retained by yeast cells than its parent compound. Although 2',7'-dichlorofluorescein is commonly used in fluorometric assays to determine hydrogen peroxide levels, it may measure levels of other reactive oxygen species as well (24, 48). Before using carboxy-H₂-DCFDA to measure changes in H₂O₂ levels in yeast cells shifted from normoxia to anoxia, we first assessed its ability to detect changes in ROS levels in whole cells under conditions that would be expected to alter ROS levels. Fluorescence was measured in steady state cultures of strain JM43 grown in the presence and absence of air and in the respiration-deficient ^o strain JM43^o. As expected, the average fluorescence is reduced in JM43 cells grown in the absence of air and in JM43 ^o (Table I). Somewhat surprising, however, is the finding that although respiration is completely inhibited in JM43 cells grown in the absence of air or in JM43 ^o cells, the levels of fluorescence do not decrease by more than 30%. To test further the reactivity of carboxy-H₂-DCFDA to H₂O₂ in yeast, we measured ROS levels in whole cells treated with H₂O₂ and H₂O₂ plus the spin trap N-tert-butyl--phenylnitrone. Hydrogen peroxide (1 mM) increased the average fluorescence levels to 122.69 ± 1.33% of the control (Table I); the addition of N-tert-butyl--phenylnitrone abolished this increase in fluorescence and reduced it to levels below that of the control, untreated culture. Because all of the fluorescent measurements were taken from a cell sorter that registered fluorescence from within viable cells, and not their surrounding medium, these results suggest that carboxy-H₂-DCFDA is responsive to intracellular H₂O₂ levels in yeast and that respiration affects the production of this H₂O₂.

View this table: [in this window] [in a new window]   Table I Effects of respiration and hydrogen peroxide on carboxy-H2-DCFDA fluorescence in whole yeast cells Mid-logarithmic phase cultures of JM43 (grown aerobically in the presence or absence of 1 mM H2O2 or H2O2 plus 5 mM of the spin trap N-tert-butyl--phenyl nitrone (BPN)) or JM43° were incubated with 10 µM carboxy-H2-DCFDA for 1 h. Three 100-µl aliquots were withdrawn, diluted 10-fold with 50 mM NaPO4, pH 7.0, sonicated briefly to disperse clumps, and then analyzed with a BD PharMingen Fluorescent-activated Cell Sorter. Measurements were taken from 5,000 cells in each of three aliquots (15,000 total). The values represent the mean values ± S.E. and are expressed as a percentage of the values obtained from JM43 cells grown aerobically with no additions.

The finding that carboxy-H₂-DCFDA fluorescence levels are reduced in anoxic cells made it possible to examine how carboxy-H₂-DCFDA fluorescence changes as cells are shifted from normoxic to anoxic conditions. To do this we made use of a fermentor system capable of growing cells at any desired oxygen concentration and for regulated shifts between oxygen concentrations (29). When the sparge gas in this system is shifted from oxygen (normoxia) to 97.5% N₂, 2.5% CO₂ (anoxia), the oxygen concentration in the fermentor falls rapidly and decreases to less than 0.5% of its normoxic level within 5 min (Fig. 1, inset). As shown in Fig. 1 there is an immediate decline in carboxy-H₂-DCFDA fluorescence during the shift. After 180 min the level of fluorescence reaches a plateau, which approximates that measured in anoxic JM43 cells. The decline in carboxy-H₂-DCFDA fluorescence after a shift to anoxia can be interpreted in at least two ways. First, it may reflect a decline in H₂O₂ levels. Because the oxygen concentration decreases during a shift it would not be surprising if H₂O₂ levels would decrease as well. Alternatively, the decline in fluorescence may result from a decrease in carboxy-H₂-DCFDA stability or level inside the cells. In order to decide between these possibilities, we did an experiment designed to measure the total amount of dye within cells after a shift. Cells in the presence of carboxy-H₂-DCFDA were shifted from normoxic to anoxic conditions, as above. Aliquots of cells were collected every 20 min after the shift; the dye was oxidized by adding 1% hydrogen peroxide, and fluorescence was measured. From Fig. 1 it can be seen that the decrease of fluorescence after a shift to hypoxia correlates with the decrease of intracellular levels of carboxy-H₂-DCFDA. This finding supports the conclusion that the decline in fluorescence observed in the first experiment is a result of a decrease in the amount of intracellular dye. This decrease may reflect breakdown of the dye in the media, a decrease in the rate of uptake of the dye, or an increase in the efflux of the dye from the cell. Regardless of the reason for this decrease in intracellular dye concentration during the shift, the decrease itself makes it clear that carboxy-H₂-DCFDA is unsuitable for assessing oxidative stress in yeast cells shifted from normoxia to anoxia and emphasizes the need for additional ways of determining whether cells shifted from normoxia to anoxia experience oxidative stress.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   ROS levels in whole cells after a shift from air to anoxia. Two different shift experiments were performed to address the effects of a shift from normoxia to anoxia on carboxy-H2-DCFDA fluorescence. In the first experiment (closed symbols) 11.8 µM carboxy-H2-DCFDA was added to an aerobic culture of JM43, and after ~90 min, cells were shifted from air to N2/CO2 (97.5/2.5%), and samples were taken at the times indicated. In the second experiment (open symbols) total intracellular levels of carboxy-H2-DCFDA were estimated by adding an excess of H2O2 to aliquots of cells taken from the fermentor at different times after a shift. Cells harvested from the fermentor at each time point were diluted 10-fold in 50 mM NaPO4, pH 7.0, and sonicated briefly to disperse cells. H2O2 was added, and fluorescence was measured after 10 min. For both experiments fluorescence was measured in three separate aliquots of 5,000 cells each (total of 15,000 cells at each time point) with a BD PharMingen fluorescent-activated cell sorter. The inset shows the dissolved oxygen concentration in the fermentor during the shift. The dissolved oxygen concentration in fully aerated growth media prior to the shift was used as 100%.

The Levels of 8-OH-dG in Mitochondrial and Nuclear DNA Increase Transiently during a Shift from Normoxia to Anoxia-- One common method for determining cellular exposure to oxidative stress is to measure oxidative DNA damage (49). This is most easily done by assaying levels of 8-OH-dG that accumulate in DNA by using HPLC in conjunction with electrochemical detection. The accumulation of this oxidatively modified deoxynucleoside is an easily assayed product of oxidative damage to DNA. In normoxic yeast cells harvested prior to a shift to anoxia, the average 8-OH-dG levels in mtDNA are much higher, 6-7-fold, than levels in nDNA (Table II). This has been reported for cells from other organisms as well (50), and most likely results from the proximity of mtDNA to mitochondrial respiration, the major source of ROS in most cells. During a shift from normoxia to anoxia, the levels of 8-OH-dG in both mtDNA and nDNA show transient increases. The average increase we observe is about 3.5-fold for mtDNA and 2.3-fold for nDNA (Table II). Data from three independent shift experiments indicate that 8-OH-dG levels in mtDNA increase to a maximum between 90 and 180 min after a shift and then decline. Similarly, the levels of 8-OH-dG levels in nDNA start to increase between 30 and 60 min and then decline. These findings indicate that both nuclear and mitochondrial DNA are exposed to transient oxidative stress when yeast cells are shifted from normoxia to anoxia.

Effects of a Shift to Anoxia on Protein Carbonylation-- Another commonly used measure of oxidative stress in cells is protein carbonylation (34). Carbonyl groups (i.e. aldehyde or ketone groups) result from the oxidation of some amino acids (51) and serve as useful markers for metal-catalyzed protein oxidation that occurs under conditions of oxidative stress. Protein carbonyls are easily quantitated after derivatization with 2,4-dinitrophenyl hydrazine. The 2,4-dinitrophenyl hydrazine is converted to 2,4-dinitrophenyl hydrazone by interaction with carbonyl groups, and the DNP-protein conjugates are subjected to analysis by HPLC (52, 53). To assess the reliability of this method for assaying oxidative stress in yeast cells, we first measured carbonylation of mitochondrial and cytosol fractions from cells with different levels of respiration. For this study, we used three related strains: JM43 and two strains, JM43^o and JM43GD100, derived from it. JM43 has a fully functional mitochondrial respiratory chain and actively respires under normoxic conditions. In contrast, JM43^o, a mitochondrial mutant that completely lacks mitochondrial DNA (27), and JM43GD100, a nuclear pet mutant deleted for a cytochrome c oxidase assembly factor (28), both lack functional respiratory chains and are respiration-deficient. In aerobically grown JM43 cells the levels of mitochondrial protein carbonylation are 6-7 times higher than levels of cytosolic protein carbonylation (Fig. 2). Anaerobically grown JM43 cells do not respire and have reduced carbonylation of both mitochondrial and cytosolic proteins. Similarly, the carbonylation levels in mitochondrial and cytosolic proteins from JM43^o and JM43GD100 are reduced in the absence of respiration. These findings clearly indicate that protein carbonylation is affected by respiration. They also suggest that most of the protein carbonylation of both mitochondrial and cytosolic proteins results from ROS released by mitochondrial respiration and that some carbonylation of mitochondrial and cytosolic proteins results from ROS produced by other processes (e.g. metal-based Fenton chemistry, microsomal electron transport, or cytosolic oxidation reactions).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effect of mitochondrial respiration on carbonyl content of yeast proteins. Mitochondrial and cytosolic fractions were prepared from JM43 cells grown in semisynthetic galactose medium under aerobic (+O2) or anaerobic (O2) conditions and JM43o and JM43DG100 grown under aerobic conditions. The carbonyl contents of mitochondrial and cytosolic proteins were determined after derivatization with DNPH. DNP-derivatized proteins (~100 µg of protein) were separated by HPLC on Zorbax GF 450 and 250 gel filtration columns connected in series and monitored with a photodiode detector tuned to 280 nm (for protein) and 366 nm (for DNP). The ratio of absorbance at 366 and 280 nm was used to normalize the column for loading and for calculating mmol of carbonyl per mol of protein (34). Bars represent mean values ± S.D. (n = 3).

By analyzing mitochondrial and cytosolic fractions taken from yeast cells after a shift to anoxia (Fig. 3), we have found that the levels of mitochondrial protein carbonylation decline from 40 mmol of carbonyl/mol of mitochondrial protein to ~10 mmol of carbonyl/mol of mitochondrial protein during the first 100 min, and then increase to a maximum of nearly 75 mmol per mol of mitochondrial protein at 180-200 min. Mitochondrial protein carbonylation then declines, reaching about twice its anoxic value of 8 mmol per mol of mitochondrial protein by 300 min. Cytosolic protein carbonylation also increases after a shift, reaching a maximum at ~150 min after a shift (Fig. 3) and starts to decline again after 180 min. An analysis of the carbonylated proteins by immunoblotting with anti-DNP antibodies (Fig. 4) confirms the drop in protein carbonylation shown in Fig. 3 and reveals that some proteins become carbonylated beginning between 90 and 120 min after the shift. Two aspects of these findings are interesting. First, the initial drop in mitochondrial protein carbonylation suggests that the carbonylated mitochondrial proteins that were present in aerobically grown cells at the time of the shift are either degraded or exported into the cytosol. The slight increase in cytosolic protein carbonylation that is concomitant with the decrease in mitochondrial carbonylation may reflect the export of some carbonylated proteins. Second, the transient increase in mitochondrial and cytosolic protein carbonylation after a shift correlates well with the transient increase of 8-OH-dG levels in mtDNA and nDNA, respectively.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Protein carbonylation after a shift to anoxia. JM43 cells were maintained in steady state normoxic growth in the fermentor for 6 generations. Then the process gas was shifted from air to 97.5% N2, 2.5% CO2. Cells were harvested at the times indicated, and mitochondrial and cytosolic fractions were prepared, derivatized with DNPH, and analyzed by HPLC, as described under "Experimental Procedures." Approximately 100 µg of mitochondrial or cytosolic protein were run for each analysis. Data shown are representative of 5 independent experiments.

View larger version (90K): [in this window] [in a new window]   Fig. 4.   Pattern of carbonylated mitochondrial and cytosolic proteins after a shift from aerobic to anaerobic conditions. JM43 cells were maintained in steady state aerobic growth in the fermentor for 6 generations, and then the process gas was shifted from air to 97.5% N2, 2.5% CO2. Cells were harvested at the times indicated, and mitochondrial and cytosolic fractions were prepared. They were derivatized with DNPH. The derivatized proteins were separated by SDS-PAGE, blotted to nitrocellulose filters, and detected with anti-DNP antibodies. 10 µg of cytosolic (top panel) and 5 µg of mitochondrial protein (bottom panel) were loaded onto each lane. Lane 1, initial time point just prior to shift. Lanes 2-9 are 30, 90, 120, 150, 180, 210, 240, and 270 min, respectively, after the shift.

The findings that two independent indicators of oxidative stress, 8-OH-dG and protein carbonylation levels, increase when yeast cells are shifted from normoxia to anoxia clearly indicate that there is a rise in ROS levels after a shift to hypoxia and suggest that cells exposed to anoxia experience oxidative stress. Importantly, this increase in oxidative stress was not detected with carboxy-H₂-DCFDA, probably because the dye leaked out of the cells during exposure to anoxia (Fig. 1).

Carbonylation in Response to Oxidative Stress Affects Specific Proteins-- To determine whether the increase in protein carbonylation affects specific proteins or has a more general widespread effect, we subjected mitochondrial and cytosolic fractions to two-dimensional electrophoresis followed by immunoblot analysis with anti-DNP antibodies. Fractions from normoxic cells were compared with mitochondrial and cytosolic fractions taken at those times after a shift when protein carbonylation is maximal. A silver stain of two-dimensional electrophoresis gels of the mitochondrial fraction taken before a shift and at 210 min after a shift reveals ~400 protein spots (Fig. 5, A and B); this represents a significant portion of the mitochondrial proteome in yeast (~450 proteins, www.incyte.com/sequence/proteome/databases/ypd.shtml). By comparing the silver stain patterns of mitochondria taken before and after the shift, it appears that there is no obvious change in pattern (Fig. 5, A versus B). This indicates that there is no gross degradation or synthesis of mitochondrial proteins during this period. By comparing the anti-DNP staining pattern it is clear that specific mitochondrial proteins are carbonylated and that their level of carbonylation increases after the shift (Fig. 5, C and D). These proteins range in size from apparent molecular masses of 23-87 kDa and have apparent isoelectric points between 5.6 and 7.7 (Table III). The level of carbonylation on nine of these proteins, indicated in Fig. 5, show an average increase of more than 2-fold with maximum increases up to 17-fold (Table III).

View larger version (75K): [in this window] [in a new window]   Fig. 5.   Two-dimensional electrophoresis of mitochondrial proteins. Samples were taken before (A and C) and 180 min after (B and D) a shift from air to 97.5% N2, 2.5% CO2. A and B are silver stains of the entire gel, and C and D are immunoblots with anti-DNP serum of A and B, respectively. Those proteins whose level of carbonylation on average increases by more than 2-fold are indicated on both the silver stains and immunoblots.

The silver-stained two-dimensional gels of the post-mitochondrial cytosolic fractions also resolve ~400 protein spots (Fig. 6, A and B). As was the case for the mitochondrial fraction, the silver stain patterns of cytosolic fractions taken before the shift were very similar to those taken after the shift, again indicating that there is little, if any, protein degradation after the shift (Fig. 6, A and B). Anti-DNP staining makes it clear that specific cytosolic proteins are carbonylated and that their level of carbonylation increases after the shift (Fig. 6, C and D). They have apparent molecular masses between 33 and 77 kDa and apparent isoelectric points between 5.3 and 7.5. The level of carbonylation of 14 of these proteins, indicated in Fig. 6, showed average increases by more than 2-fold and some increased by as much as 14-fold (Table III).

View larger version (80K): [in this window] [in a new window]   Fig. 6.   Two-dimensional electrophoresis of cytosolic proteins. Samples were taken before (A and C) and 180 min after (B and D) a shift from air to 97.5% N2, 2.5% CO2. A and B are silver stains of the entire gel, and C and D are immunoblots with anti-DNP serum of A and B, respectively. Those proteins whose level of carbonylation on average increases by more than 2-fold are indicated on both the silver stains and immunoblots.

We subjected the 9 mitochondrial and 14 cytosolic protein spots whose level of carbonylation on average increased by at least 2-fold after a shift to in-gel digestion with trypsin, followed by peptide mass fingerprinting and MALDI analysis of the eluted peptides. These proteins were identified based on their apparent molecular weights, isoelectric points, intracellular location, and MALDI patterns (Table IV). Interestingly, some of the proteins that experience enhanced carbonylation (glyceraldehyde-3-phosphate dehydrogenase, pyruvate decarboxylase, enolase, and aconitase) in cells shifted to hypoxia are identical to proteins that experience enhanced carbonylation in cells exposed to peroxidative stress (54). This provides further support for the conclusion that cells shifted to anoxia experience oxidative stress.

With the exception of phosphoglycerate kinase, which was identified in both mitochondrial and cytosolic fractions, no carbonylated mitochondrial protein was found in the cytosol. This makes it unlikely that the decline in carbonylation observed soon after the shift (Fig. 3) is due to the release of carbonylated proteins from mitochondria. Phosphoglycerate kinase is an abundant cytosolic protein so the finding that it is present in both fractions may represent cross-contamination of the mitochondrial fraction by the cytosol. Alternatively, it is possible that this enzyme is one of many that is encoded by a single nuclear gene but that is shared between mitochondrial and cytosolic compartments (e.g. see Refs. 55 and 56).

Induction of SOD1 after a Shift to Anoxia-- Yeast genes that encode stress proteins are aerobic genes whose expression is maximal in air and reduced under anoxic conditions (cf. Refs. 14 and 57). They are also induced by oxidative stress brought about by exposure to oxidants such as H₂O₂ and paraquat (57-59). In view of our finding above that yeast cells experience transient oxidative stress upon being shifted to anoxia, it was of interest to ask if a stress-response gene is also induced transiently during a shift to anoxia. To address this we investigated the expression of SOD1, the gene for Cu,Zn-superoxide dismutase, after a shift to anoxia. During a shift to anoxia, mRNA levels from this gene decline to about 30% of their aerobic level within 4 h after the shift and then increase to levels that exceed their normoxic levels (Fig. 7). The finding that SOD1 mRNA levels initially decline is expected for an aerobic gene because cells experience reduced oxygen concentrations during a shift from normoxia to anoxia. Moreover, the increase in SOD1 expression after 4 h of exposure to anoxia is consistent with the conclusion reached above that cells exposed to anoxia experience oxidative stress. The time of induction of SOD1 comes at a time that is after proteins and DNA show maximal oxidative damage, again indicating that yeast cells shifted to anoxia experience oxidative stress and that this stress is not immediate but delayed by 2-3 h.

View larger version (8K): [in this window] [in a new window]   Fig. 7.   Transcription of the SOD1 after shift from aerobic to anaerobic conditions. JM43 cells were maintained in steady state normoxic growth in the fermentor for at least 6 generations; the sparge gas was then switched from air to 97.5% N2, 2.5% CO2. Cells were harvested, and total RNA was isolated and subjected to Northern blot analyses as described under "Experimental Procedures." The relative signal intensity was measured using an Amersham Biosciences Storm 860 PhosphorImager. Transcript levels were normalized to the level of ACT1 mRNA and are presented as a decimal percent of their steady state levels, under aerobic conditions.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The results of this study demonstrate that yeast cells exposed to anoxia experience transient oxidative stress, that specific proteins become selectively carbonylated during exposure to anoxia, that some of these proteins are also selectively carbonylated during peroxidative stress, and that most of the protein carbonylation in yeast cells comes from ROS generated by mitochondrial respiration. These findings, together with those of previous studies with yeast (15) and mammalian cells (16, 18), raise the question of whether mitochondrially generated ROS function as "signals" in a signal transduction pathway for the induction of yeast hypoxic nuclear genes.

Exposure to Anoxia Is a Type of Oxidative Stress-- The conclusion that yeast cells experience transient oxidative stress when exposed to anoxia is based on three different observations. First, mtDNA, and to a lesser extent, nDNA accumulate elevated levels of 8-OH-dG. After a shift to anoxia 8-OH-dG levels in mtDNA increase an average of nearly 4-fold, whereas 8-OH-dG levels in nDNA increase by about 2.5-fold. The peak increase in 8-OH-dG levels comes earlier after the shift for nDNA than for mtDNA. After they peak, the levels of 8-OH-dG in both mtDNA and nDNA decline to their normoxic levels. This decline probably represents the results of base excision and nucleotide excision repair (60, 61).

The second observation that supports the conclusion that exposure to anoxia induces oxidative stress comes from measuring levels of protein carbonylation at different times after a shift from normoxia to anoxia. The level of mitochondrial protein carbonylation in normoxic yeast cells harvested before such a shift is nearly 10 times higher than the level of cytosolic protein carbonylation. Because levels of both mitochondrial and cytosolic protein carbonylation are higher in respiratory-proficient yeast strains than in respiratory-deficient strains, it is clear that mitochondrial respiration is responsible for the bulk of this carbonylation. After a shift from normoxia to anoxia mitochondrial protein carbonylation initially drops, then increases dramatically between 120 and 200 min after the shift, and finally drops back to its anoxic level. The level of carbonylation of cytosolic proteins increases slightly between 50 and 200 min after a shift and then declines. It is interesting that the oxidation of nDNA as well as cytosolic protein precedes the oxidation of mtDNA and mitochondrial protein. This suggests that the nucleo-cytosol experiences oxidative stress earlier than the mitochondrion.

The overall level of protein carbonylation during a shift reflects the balance between protein oxidation and the degradation of oxidized protein. Although it is generally recognized that the rate of protein oxidation is determined by the rate of ROS production by the electron transport chain and by the subsequent alterations in steady state NADH and NAD(P)H levels, it is not at all clear what determines the rate of oxidized protein degradation. Indeed, although many oxidized cytosolic proteins are more susceptible to degradation by the proteosome (e.g. see Ref. 62), some oxidized proteins are resistant to degradation (63). Moreover, it is not yet known how mitochondria degrade oxidized proteins. Nonetheless, our data support the overall conclusion that an increase in ROS leads to enhanced oxidation of protein and DNA after a shift to anoxia and that the oxidative damage that is caused is repaired.

The third observation that supports the conclusion that yeast cells exposed to anoxia experience oxidative stress comes from following the expression of SOD1 after a shift to anoxia. SOD1 is an aerobic gene whose expression is induced above its normoxic levels in hyperbolic cells and in cells exposed to oxidants (57, 59); it is repressed in cells grown to steady state under anoxic conditions (57). We have found that during a shift to anoxia SOD1 mRNA levels initially decline and then increase. The increase in SOD1 expression occurs after the peaks in protein carbonylation and 8-OH-dG levels, suggesting that the transient increase in ROS that preceded these peaks in DNA and protein oxidation also stimulates the expression of SOD1.

Finally, our findings with carboxy-H₂-DCFDA raise serious questions concerning the reliability of using fluorescein-based analogs for assaying ROS levels in cells shifted from normoxic to anoxia. Although these compounds have been widely used, they are problematic for a variety of reasons. First, it is not clear what ROS they detect. Second, it is not known whether they are capable of sampling all cellular compartments. Third, their oxidized forms are unstable. Fourth, they leak out of cells. The latter is a particular problem for the type of experiments performed here where cells are loaded with the dye, shifted from normoxia to anoxia, and followed for 6 h.

Specific Proteins Become Selectively Carbonylated upon Exposure to Anoxia-- Interestingly, not all proteins acquire enhanced levels of carbonylation after a shift to anoxia. This suggests that only certain proteins are the targets of oxidants generated during the shift. Some of these proteins (glyceraldehyde-3-phosphate dehydrogenase, pyruvate decarboxylase, enolase, and aconitase) are identical to proteins that are modified during exposure of yeast cells to hydrogen peroxide (53). The finding that specific proteins become carbonylated after a shift to anoxia is also interesting because it clearly demonstrates that protein carbonylation (i.e. oxidation) is selective. However, it is not known what determines this selectivity. It is clear that protein carbonylation is not determined by protein size, pI, abundance, or the presence of metal-binding sites. Also unclear is whether selective protein oxidation has any physiological meaning for cellular adaptation to anoxia. In this context, it is important to note that carbonylation is merely a convenient measure for protein oxidation and that there are many other types of amino acid side chain oxidation (e.g. formation of o-tyrosine, n-tyrosine, or dityrosine). It is possible that any one of these post-translational protein modifications may have an important role in physiological adaptation to growth in different oxygen environments.

Do Mitochondrially Generated ROS Function as Signals in a Signaling Pathway?-- The findings reported here are interesting in the context of studies on hypoxic gene induction in mammalian cells, where it has been proposed that mitochondrially generated ROS participate as signals in a signaling pathway that mediates hypoxic stabilization of the  subunit of the HIF-1 transcription factor (18). It is not known whether mammalian mitochondria experience transient changes in protein and mtDNA oxidation, as shown here for yeast. However, our findings that mitochondrial respiration is required for the induction of hypoxic genes in yeast (15) together with these findings from mammalian cells suggest that mitochondrially generated ROS may play a pivotal role in the induction of some hypoxic genes in eucaryotes. From the finding that protein carbonylation and 8-OH-dG levels in mitochondrial DNA increase when yeast cells are shifted to hypoxia, it is possible to envision three mechanisms by which mitochondrially produced ROS could participate in hypoxic gene induction (Fig. 8). In the first pathway, ROS oxidize a mitochondrial protein, which initiates a signaling pathway to the nucleus. In the second pathway, free ROS are released from mitochondria and initiate a signaling pathway to the nucleus. In the third pathway, ROS modify mitochondrial gene expression via oxidative damage to mtDNA, which initiates a signaling pathway to the nucleus. In the first and third pathways, the ROS signal originates in the mitochondrion and starts a signaling pathway while still in the mitochondrion, but in the second pathway, mitochondrially produced ROS leave the mitochondrion and start a signaling pathway in the cytosol or nucleus. At present, it is not known which, if any, of these pathway(s) is (are) operative. However, microarray analysis of gene expression data suggests that some yeast genes involved in the oxidative stress response are induced by hypoxia (64) but that hypoxic yeast genes are not induced by exogenously added oxidants, like hydrogen peroxide (59, 65). These microarray data are consistent with the hypothesis that the stress response that yeast cells experience during a shift to anoxia comes from within the cell and cannot be mimicked by exogenous oxidants.

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Three ways that ROS generated by the mitochondrial respiratory chain may be involved in hypoxic gene induction in yeast. In the first, ROS oxidize a mitochondrial protein, which initiates a signaling pathway to the nucleus. In the second, free ROS are released from mitochondria and initiate a signaling pathway to the nucleus. In the third, ROS modify mitochondrial gene expression via oxidative damage to mtDNA, which initiates a signaling pathway to the nucleus.

Conclusions-- The findings reported here clearly demonstrate that yeast cells experience oxidative stress when exposed to anoxia. This is surprising. Indeed, it raises many questions. First, does the respiratory chain produce more ROS as the oxygen concentration is decreasing, or does the oxidative stress experienced by cells exposed to anoxia come from an increase in the level of reduced pyridine nucleotides and other redox active compounds that accumulate when respiration slows? Second, do the enhanced levels of 8-OH-dG and protein carbonylation observed during a shift to anoxia result from increased levels of ROS and, if so, where does the oxygen that is present in these ROS come from as cells experience reduced oxygen concentrations? Third, are specific genes affected by the increase in 8OH-dG levels in mtDNA or nDNA or is the distribution of this oxidized base random? If specific genes are oxidized does this affect their function or expression and, if so, is this important for hypoxic gene induction? Fourth, do mitochondrial or cytosolic proteins that are specifically oxidized in cells exposed to anoxia play a role in signaling pathways from the mitochondrion to the nucleus that function to induce hypoxic genes? These questions are currently under study.

	ACKNOWLEDGEMENTS

We gratefully acknowledge the assistance of Dr. N. Ahn and co-workers in performing MALDI analysis.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant GM 30228 (to R. O. P.) and a National Institutes of Health postdoctoral fellowship (to K. O.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ To whom correspondence should be addressed. Tel.: 303-493-3823; Fax: 303-492-3883; E-mail: Poyton@spot.Colorado.edu.

Published, JBC Papers in Press, June 27, 2002, DOI 10.1074/jbc.M203902200

	ABBREVIATIONS

The abbreviations used are: ROS, reactive oxygen species; 8-OH-dG, 8-hydroxy-2'-deoxyguanosine; carboxy-H₂-DCFDA, carboxy-H₂-dichloro-dihydrofluorescein diacetate; DNPH, 2,4-dinitrophenyl hydrazine; DNP, 2,4-dinitrophenol; MALDI, matrix-assisted laser desorption ionization; HPLC, high pressure liquid chromatography; DTT, dithiothreitol; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid; MOPS, 4-morpholinepropanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES