Address correspondence and reprint requests to Prof. Dr. C. Richter-Landsberg at Department of Biology, Molecular Neurobiology, University of Oldenburg, POB 2503, D-26111 Oldenburg, Germany.
Abstract : H2O2-induced onset and execution of programmed cell death in mature rat brain oligodendrocytes in culture is accompanied by the induction and nuclear translocation of the transcription factors AP-1 and nuclear factor-κB (NF-κB), both of which have been discussed as regulators of cell death and survival. Supershift analysis of nuclear extracts indicated that the AP-1 complex consists of c-Jun, c-Fos, JunD, and possibly JunB proteins, and that the NF-κB complex contains p50, p65, and c-Rel proteins. The first signs of DNA fragmentation were seen already during the first hour after the treatment. DNA fragmentation could be prevented by the antioxidants pyrrolidine dithiocarbamate and vitamin E, by the nuclease inhibitor aurintricarboxylic acid, and by preincubation with the iron chelator deferoxamine (DFO). Additionally, DFO protected oligodendrocytes from H2O2-induced cytotoxic effects as assessed by the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay, and suppressed the formation of free radicals. DFO alone led to a slight increase and in combination with H2O2 synergistically induced DNA-binding activities of AP-1 and NF-κB in oligodendrocytes. Our data suggest that although low levels of H2O2 directly activate AP-1 and NF-κB and might contribute to signal transduction pathways promoting cell survival, the formation and action of hydroxyl radicals promote cell death mechanisms that can be attenuated by the iron chelator DFO.
Oxidative stress contributes to a number of acute and chronic neurodegenerative diseases (Halliwell, 1992). Cell degeneration induced by reactive oxygen species (ROS) occurs by either apoptosis or necrosis (Leist and Nicotera, 1998). Necrosis is characterized by a loss of plasma membrane integrity, the formation of large vacuoles, and cell swelling, whereas typical features of apoptotic cells undergoing programmed cell death (PCD) are nuclear changes that include chromatin margination and condensation, DNA fragmentation, membrane blebbing, and cell shrinkage (Leist and Nicotera, 1998). In the CNS, oligodendrocytes form the myelin sheath that comprises a lipid-rich multilamellar membrane, insulates axons, and increases nerve conductance velocity. Oligodendrocytes have the highest rate of oxidative metabolic activity of any cell in the brain and can myelinate as many as 50 internodes on multiple neuronal axons. Stress responses are triggered in oligodendrocytes, and toxic or metabolic disturbances might damage the myelin sheath or injure the entire oligodendroglial cell. Oligodendroglial cell death might contribute to the pathology of demyelinating disorders, such as multiple sclerosis (Boccaccio and Steinman, 1996 ; Luccinetti et al., 1996 ; Ludwin, 1997). The reaction of oligodendrocytes to free radical damage, as a consequence of ischemia, anoxia, and reoxygenation injury, is far from understood. Oligodendrocytes are vulnerable to hypoxia, which alters iron homeostasis and induces ferritin synthesis (Qi et al., 1995). It has been suggested that the susceptibility of oligodendrocytes to oxidative stress is related to their low glutathione content and high iron store (Thorburne and Juurlink, 1996 ; Juurlink et al., 1998). Oligodendrocytes are the predominant iron-containing cells in the brain (Connor and Menzies, 1996), and iron released from the intracellular stores can be pathogenic because it might mediate the formation of free radicals, in particular the damaging hydroxyl radical from peroxides via the Fenton reaction (Halliwell, 1992). We have demonstrated previously that mature oligodendrocytes in culture are highly susceptible to free radical damage exerted by hydrogen peroxide (H2O2) (Richter-Landsberg and Vollgraf, 1998). Brief exposure to H2O2 (30 min, 50-100 μM) within 24 h triggered the onset of PCD, which was accompanied by the appearance of fragmented and condensed 4,6-diamidino-2-phenylindole-stained nuclei and internucleosomal DNA cleavage. The onset of the apoptotic death program involved the transcriptional activation of the immediate early genes c-fos and c-jun.
Several genes have been shown to be involved in both the positive and negative regulation of apoptotic cell death (Dragunow and Preston, 1995), yet the question whether changes in gene expression are causally related to the subsequent cell death is still unsolved. Prominent among these genes that are redox-regulated and activated during cellular insults are the AP-1 transcription factor and nuclear factor-κB (NF-κB). AP-1 is a protein complex containing Jun and Fos proteins or Jun dimers (Gass and Herdegen, 1995). Fos and Jun proteins belong to a large superfamily of proteins that possess a leucine zipper and a DNA-binding domain and include c-Fos, FosB, Fra-1, and Fra-2, and c-Jun, JunB, and JunD, respectively. The nuclear factor NF-κB consists of homo- or heterodimers of NF-κB-1 (p50), NF-κB-2 (p52), RelA (p65), RelB, or Rel (c-Rel) and exerts its transcriptional activation after phosphorylation, ubiquitination, and subsequent degradation of its inhibitory factor-κB (IκB) component (comprising IκBα, -β, -γ, and Bcl-3) (for review, see Flohé et al., 1997). NF-κB plays an important role in processes of the immune response and inflammation. Recent evidence suggests that it is also a crucial transcription factor for glial and neuronal cell function (O'Neill and Kaltschmidt, 1997). It might act as an important signal in neurodegenerative diseases (Lezoualc'h and Behl, 1998) and is involved in apoptotic events (Clements et al., 1997). The signal transduction pathways that are elicited by the diverse stimuli and stress situations are cell type-specific and depend on the age and physiological state of the cell, and hence NF-κB activation might exert either a protective or harmful function (Lipton, 1997 ; Mattson et al., 1997).
The role of AP-1 and NF-κB in oxidative stress-induced oligodendroglial cell death so far has not been investigated. Here we have examined the transcriptional modulation and DNA-binding activity of AP-1 and NF-κB transcription factors during apoptotic events in cultured rat brain oligodendrocytes that were triggered by treatment with low concentrations of H2O2. H2O2 easily penetrates cell membranes, and via the iron-mediated Fenton reaction hydroxyl radicals are formed (Halliwell, 1992). Additionally, the effects of antioxidants and the iron chelator deferoxamine (DFO) on the nature of cell death and the potential to rescue oligodendrocytes were tested. DFO is a well known therapeutic substance used in the treatment of chronic iron overload and might be effective in protecting against oxidative stress (for review, see Galey, 1997). DFO has been shown to suppress experimental allergic encephalomyelitis (Pedchenko and LeVine, 1998), an animal model of multiple sclerosis, and thus might be useful in the treatment of demyelinating disorders.
MATERIALS AND METHODS
Cell culture media were from GibcoBRL (Grand Island, NY, U.S.A.). Poly-l-lysine (PLL) was from Boehringer Mannheim (Germany). 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was from Pharmacia. 12-O-Tetradecanoylphorbol 13-acetate (TPA) and staurosporine were from Sigma. 2′, 7′ -Dichlorofluorescin diacetate (DCFH-DA) was from Molecular Probes.
Oligodendroglial cultures were prepared from the brains of 1-day-old Wistar rats as described before (Richter-Landsberg and Vollgraf, 1998). In brief, precursor cells, growing on the astrocytic cell layer, were separated by vigorous shaking (for 16 h at 220 rpm) and removed (McCarthy and de Vellis, 1989). To obtain mature oligodendrocytes, precursors were replated on PLL-coated culture dishes (3 × 106 cells per 10-cm dish) and kept for 7 days in serum-free Dulbecco's modified Eagle medium to which insulin (5 μg/ml), transferrin (5 μg/ml), and sodium selenite (25 ng/ml) supplement (Boehringer, Mannheim, Germany) was added. All cells were kept at 37°C in a 10% CO2 atmosphere. Growth medium was changed twice a week.
Cells were washed with phosphate-buffered saline (PBS) and scraped off the culture dishes with a rubber policeman. After centrifugation, the cell pellet was fixed in 6% glutaraldehyde in PBS for 2 h and washed several times in 0.1 M phosphate buffer (0.1 M Na2HPO4· H2O, 0.1 M NaH2PO4· 2H2O, pH 7.3). After osmication in 2% osmium tetroxide for 2 h and dehydration in a graded series of ethanol steps, samples were embedded in Spurr solution and polymerized for 8 h at 70°C. After contrasting with uranyl acetate, ultrathin sections (90 nm) were cut with an Ultramicrotome Ultracut E (Reichert, Wien, Austria), and observed and photographed with a Zeiss EM 109 electron microscope.
To assess the cytotoxic potential of the compounds, the MTT (tetrazolium) assay was carried out (Richter-Landsberg and Vollgraf, 1998). In brief, oligodendrocyte precursor cells were prepared as described above and plated on PLL-coated 96-microwell cell culture plates (2 × 104 cells per well) and grown for 7 days. Thereafter, the growth medium was removed, fresh medium (100 μl/well) was added either with or without H2O2 or the other compounds at the indicated concentrations, and cells were incubated for the indicated times. MTT solution (10 μl ; 5 mg/ml in PBS) was added to the wells, containing 100 μl of medium, and the plates were incubated for 4 h. Thereafter, 100 μl of a solubilization solution (10% sodium dodecyl sulfate in 0.01 mol/L HCl) was added and incubated overnight to dissolve the water-insoluble formazan salt. Quantification was then carried out with an ELISA reader at 595 nm using a 655-nm filter as a reference. Data are expressed as a percentage of the untreated controls, and values represent the means ± SD of eight microwells each of three independent experiments (n = 24).
2′,7′-Dichlorofluorescein (CDF) assay
Cultures were washed twice with Ringer solution (120 mM NaCl, 0.6 mM MgCl2, 1.9 mM CaCl2, 7.5 mM KCl, 24 mM NaHCO3, 2.4 mM HEPES, 11 mM D-glucose). DCFH-DA (50 μM) was added, and cells were incubated for 30 min at room temperature. Cultures were washed again twice with Ringer solution, and H2O2 was added. After 5 min, cells were scraped off the culture dishes and lysed in Tris buffer (50 mM Tris, pH 7.5, 1% sodium dodecyl sulfate). Cell lysates were vortexmixed and centrifuged, and fluorescence was measured in a fluorescence photometer SFM25 (Kontron Analytic) at 490 nm excitation and 515 nm emission wavelengths. Data are expressed as a percentage of the untreated control.
RNA extraction and reverse transcription
RNA was isolated by a modification of the method of Chomczynski and Sacchi (1987) as described before (Richter-Landsberg and Vollgraf, 1998). One microgram of RNA was used for reverse transcription in a final volume of 20 μl. First-strand synthesis conditions were as follows : 50 mM Tris-HCl, pH 8.3, 50 mM KCl, 10 mM MgCl2, 10 mM dithiothreitol, 0.5 mM spermidine, 0.25 mM of each dNTP, 12.5 pmol each of oligo(dT)16 and random hexamer primers, and 5-8 U of AMV reverse transcriptase (Promega, Madison, WI, U.S.A.). After a 1-h incubation at 42°C, the reaction mixture was diluted to 100 μl with water and stored at -80°C. Subsequently, 1-2 μl was used for PCR analysis.
PCR and primers
Primers were synthesized by Pharmacia (Freiburg, Germany). Primers were designed with the Primer3 software (White-head Institute for Biomedical Research) or taken from the literature. The following primers for AP-1 subunits were used (number of amplification cycles and annealing temperatures are indicated in parentheses) : c-fos, 5′-CTG TCC GTC TCT AGT GCC AAC TT-3′ and 5′-ATC TGT CTC CGC TTG GAG CGT AT-3′ (Gillardon et al., 1996) (34, 60°C) ; c-jun, 5′-GCT TCT CTA GTG CTC CGT AA-3′ and 5′-TCT AGG AGT CGT CAG AAT CC-3′ (Schäfer et al., 1996) (33, 52°C) ; junB, 5′-AGG GGG CGT CTA TGC TGG TC-3′ and 5′-GGC TGG GGG TGA CCG TAT GG-3′ (36, 62°C) ; junD, 5′-CCC TCA AAA GCC AGA ACA CC-3′ and 5′-CAC ACT CAA CAC GCA ACC AA-3′ (37, 60°C). Primers for NF-κB subunits were as follows : p50, 5′-CAC TGT GAG GAC GGG GTA TGC-3′ and 5′-AGC TGG CTT TGT AAT GTT GAC-3′ (Tan et al., 1994) (37, 65°C) ; p65, 5′-AAG ATC AAT GGC TAC ACA GG-3′ and 5′-CCT CAA TGT CTT CTT TCT GC-3′ (based on MEDLINE derived sequence, accession no. M61909) (32, 55°C) ; c-rel, 5′-TCG GTG TGT AAA GAA AAA GG-3′ and 5′-CAG TCA TTC AAC ACA AAA CG-3′ (MEDLINE accession no. X70690) (38, 47°C).
Control experiments were carried out with the following primers for glyceraldehyde-3-phosphate dehydrogenase (gapdh), 5′-CCC ACG GCA AGT TCA ACG GCA-3′ and 5′-TGG CAG GTT TCT CCA GGC GGC-3′ (Fort et al., 1985) (25, 60°C), and β-actin, 5′-GGC ATG TGC AAG GCC GGC TTC-3′ and 5′-GGA TGG CAT GAG GGA GCG CGT-3′ (Nudel et al., 1983) (28, 72°C).
PCR was carried out in a volume of 23 μl containing 10 mM Tris-HCl, pH 9.0, 50 mM KCl, 2 mM MgCl2, 0.2 mM of each dNTP, and 0.8 μM of each primer. After an initial denaturation for 2 min at 95°C, 2 μl of Taq polymerase (0.5 U/μl) was added, and 20-40 cycles were performed in a Biometra Thermocycler, each consisting of 30 s of denaturation at 95°C, 30 s of primer annealing at 52-65°C, depending on the primer pair used, and 1 min of elongation at 72°C. After a terminal extension for 5 min at 72°C, amplification products were stored at 4°C, and 5-10 μl was used for analysis by gel electrophoresis on 1.5% agarose gel. Gels were evaluated quantitatively with a gel documentation system and matching software EasyImage Plus (Herolab GmbH, Wiesloch, Germany). Values were normalized to the corresponding GAPDH expression levels.
Assay for DNA fragmentation
Total genomic DNA was isolated from cell pellets of oligodendrocytes (3-6 × 106 cells for each experimental condition). DNA samples (2-10 μg) and a 1-kb DNA ladder standard (GIBCO) were separated on 1.5% agarose gels, visualized by ethidium bromide staining, and photographed.
Electrophoretic mobility shift assay (EMSA)
Nuclear extracts were prepared from oligodendrocytes at different time points after treatment as described (Schreiber et al., 1997). In brief, 3 × 106 cells were washed twice with ice-cold PBS, harvested in 400 μl of buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 2 mM dithiothreitol, 10 μg/ml each of aprotinin/leupeptin), and incubated on ice for 5 min. After addition of 20 μl of 20% Nonidet P-40, extracts were vortex-mixed and centrifuged for 30 s. Pellets were resuspended in 150 μl of buffer B (buffer A plus 400 mM NaCl, 1% Nonidet P-40) and gently mixed (15 min, 4°C). After centrifugation (5 min, 4°C), 15 μl of glycerol was added to supernatants, which could be stored at —80°C.
For mobility shifts, 1 μg of extract was incubated with 0.5 ng of 32P-labeled probe (~25,000 cpm) for 20 min on ice in a 20-μl reaction mixture [10 mM HEPES, pH 8.0, 5% glycerol, 50 mM NaCl, 5 mM MgCl2, 2 mM dithiothreitol, 0.1 mM EDTA, and 1 μg of poly (dI-dC) as unspecific competitor]. For supershifts, 0.5 μl of antibody solution (Santa Cruz) was added to the reaction, and incubation was continued for an additional 10 min before loading onto 4% polyacrylamide gels. Electrophoresis was performed in 0.5 × TBE (45 mM Tris, 45 mM boric acid, 1 mM EDTA, pH 8.3). Gels were dried and exposed for autoradiography. Experiments were carried out twice with similar results. The following oligonucleotides were used as probes (factor binding sites are underlined) : NF-κB-1, 5′-GATCTGAGGGGACTTTCCCAGG-3′, containing one NF-κB binding motif ; TRE (phorbol ester response element) from the plasmid 5×TRE CAT, containing five AP-1 binding motifs (5′-AAGCTTGATGAGTCAGCCGGATCC-3′) (Angel et al., 1987). Double-stranded oligonucleotides were labeled with Klenow enzyme (GibcoBRL) and [α-32P]dCTP.
All experiments were carried out with 7-day-old cultured oligodendrocytes derived from rat brains. Oligodendrocyte progenitor cells show a characteristic bipolar morphology, and after 1 week in culture cells have matured and oligodendrocytes have extended numerous membranous processes that stain positively for galactocerebroside, myelin basic protein, proteolipid protein, and the myelin-associated glycoproteins MAG and MOG, and thus are considered mature oligodendrocytes (Richter-Landsberg and Heinrich, 1995 ; Richter-Landsberg and Vollgraf, 1998).
Accumulation of reactive oxygen species
To test if exposure of oligodenbrocytes to H2O2 at concentrations that exert half-maximal cytotoxicity (100 μM) and lead to the onset of the apoptotic program (Richter-Landsberg and Vollgraf, 1998) result in an accumulation of peroxides, oligodendrocytes were incubated with DCFH-DA. DCFH-DA permeats cellular membranes and is intracellularly cleaved into 2′,7′ -dichlorofluorescin (DCFH), which upon interaction with peroxides is oxidized to the fluorescent DCF (Page et al., 1993). Cells were preloaded with DCFH-DA and exposed to H2O2 for 5 min. Figure 1 indicates that low concentrations of H2O2 exerted a small but distinct increase in DCF fluorescence, which was increased further at a higher concentration (1 mM). This increase in DCF fluorescence could be attenuated when cells were preincubated with the antioxidants vitamin E or pyrrolidine dithiocarbamate (PDTC), and also in the presence of the iron chelator DFO (Fig. 1).
Early response gene activation during stress situations has been described to occur on the level of transcription. H2O2 induces the onset of the apoptotic death program in oligodendrocytes within 24 h (see Fig. 8 below). To investigate whether the expression levels of the AP-1 family members c-fos, c-jun, junB, and junD and the NF-κB family members p50, p65, and c-rel are individually affected by oxidative stress, total RNA was isolated from control and H2O2-treated cells at different time points after treatment. RNA was reverse transcribed and the resulting cDNA was subjected to PCR (RT-PCR). Reaction conditions were chosen so that amplification was in the linear range, and amplification of gapdh and β-actin was carried out from each cDNA as a control for equal loading. Figure 2a shows that 100 μM H2O2 exerts relatively small increases in the mRNA levels of all members of the AP-1 family tested. It is interesting that the highest increase is reached in the AP-1 family member junD (Fig. 2a). Figure 2b demonstrates that all three members of the NF-κB family that were tested here were transcriptionally activated by the treatment with 100 μM H2O2, and a maximal increase of ~250% was reached after 6 h. In contrast to c-jun and c-fos, the incubation with higher concentrations of H2O2 (up to 1 mM) did not further stimulate the mRNA levels of the NF-κB family members (data not shown). The level of actin mRNA expression was stimulated only slightly during the first 2 h and then was back to control levels (Fig. 2b).
Effects of H2O2 on AP-1 and NF-κB DNA binding activity
To examine whether the transcriptional activation of the early response genes is accompanied by activation on the protein level, EMSAs with nuclear extracts of oligodendrocytes dendrocytes and 32P-labeled oligonucleotides containing specific AP-1 or NF-κB binding sites, respectively, were carried out. Oligodendrocytes were treated with H2O2 (100 μM) for 30 min. Nuclear extracts of the cells were prepared from untreated cells, immediately after the treatment, or at the indicated times after a recovery period of up to 24 h. Figure 3 demonstrates that oxidative stress induced the DNA-binding activity of AP-1 and that DNA-binding activity was stimulated after a 30-min treatment with 100 μM H2O2 immediately after the treatment. DNA-binding activity was further enhanced during the recovery period and reached a maximum 6 h after the treatment with H2O2 (Fig. 3). Preincubation of oligodendrocytes with DFO (1 mM) alone led to a slight induction in AP-1 DNA-binding activity. When cells were preincubated with DFO (1 mM) for 6 h, treated with H2O2 (100 μM, 30 min), and kept for 2 h of recovery time in the presence of DFO, the DNA-binding activity of AP-1 was significantly induced and stronger than after the treatment with H2O2 only (Fig. 3). Thus, the iron chelator modulated DNA-binding activity of the AP-1 protein complex. AP-1 DNA-binding activity could also be induced by the treatment with the phorbol ester TPA (100 nM), which is known to activate AP-1 DNA-binding activity and served as a positive control. In contrast, treatment of the cells with staurosporine (200 nM) for 24 h, which effectively leads to the onset of apoptosis in oligodendrocytes (Richter-Landsberg and Vollgraf, 1998), caused only a slight induction of AP-1 DNA binding. To assess the specificity of the AP-1 complex and identify the contributing family members, supershift experiments were carried out by adding specific antibodies against c-Fos, c-Jun, JunB, and JunD, respectively, to the nuclear extracts before EMSA (Fig. 4). Nuclear extracts were prepared from cells treated with 100 μM H2O2 for 30 min after a 2-h recovery period. Protein-antibody recognition can be visualized by a decrease in the mobility of the DNA-protein complex and/or a diminution of the AP-1 complex. Figure 4 indicates that the AP-1 complex consists mainly of c-Fos and to a lesser extent c-Jun. Also, small amounts of JunB and JunD could be detected in the autoradiogram depicted in Fig. 4.
NF-κB DNA binding activity is shown in Fig. 5. The nuclear translocation of NF-κB occurred within the first 2 h after H2O2 treatment (30 min, 100 μM) and decreased again during the next 24 h. Supershift analysis with specific antibodies against the p50, p65, and c-Rel subunits revealed that all three members of the NF-κB family contributed to the complex (Fig. 5). In contrast to Jurkat cells, where we observed a strong response after TPA (100 nM, 6 h) treatment (data not shown), in oligodendrocytes TPA exerted only a slight increase in NF-κB DNA-binding activity (Fig. 5, right). The incubation of the cells with staurosporine (200 nM, 24 h) did not induce NF-κB binding activity. The DNA-binding activity of NF-κB was affected by the iron chelator DFO. Whereas DFO (1 mM, 8 h) alone led to a weak activation, DFO in combination with H2O2 exerted a strong induction as analyzed after 2 h of recovery, which significantly exceeded the effect of H2O2 alone (Fig. 5, right).
To investigate whether DFO protects oligodendrocytes against H2O2-mediated cytotoxicity, cells were preincubated with DFO (1 mM, 6 h) and then subjected to H2O2 (50-250 μM, 30 min). After a recovery period in fresh medium supplemented with DFO, cell viability was determined by measuring the mitochondrial activity using the MTT assay. Figure 6 demonstrates that 1 mM DFO was effective in protecting oligodendrocytes from the cytotoxic effects of H2O2 up to a concentration of 250 μM.
Ultrastructural examination by electron microscopy confirmed that H2O2 (100 μM, 30 min, 24-h recovery) caused morphological changes in oligodendrocytes. The chromatin was more condensed and marginated, and nuclei appeared strongly distorted (Fig. 7). Apoptotic cell death is often accompanied by DNA fragmentation into multiples of nucleosome length of ~180 bp. We have shown before that H2O2 (10-500 μM, 30 min) induced internucleosomal DNA fragmentation in oligodendrocytes after a recovery period of 16-24 h (Richter-Landsberg and Vollgraf, 1998). Here we have isolated DNA from control and H2O2-treated oligodendrocytes during various times of recovery. Time course analysis indicated that the first signs of DNA fragmentation after treatment with H2O2 (100 μM, 30 min) were observable as early as 1 h, and DNA fragmentation was already most prominent after 2-3 h of recovery (Fig. 8). H2O2-induced DNA fragmentation could be attenuated by preincubation with antioxidants PDTC or vitamin E, and the nuclease inhibitor aurintricarboxylic acid (ATA) (Fig. 9a). When cells were preincubated with DFO for 6 h and then subjected to oxidative stress (100 μM H2O2, 30 min), DNA fragmentation could be suppressed (Fig. 9b). This effect was concentration-dependent and most effective at a DFO concentration of 1 mM. DFO alone did not induce the appearance of a DNA ladder.
In the present communication, we show that the onset and execution of the oxidative stress-induced apoptotic death program in oligodendrocytes is accompanied by the induction and nuclear translocation of the transcription factors AP-1 and NF-κB. Both transcription factors have been discussed as regulators of cell death, survival, and regeneration. Whereas nuclear translocation of NF-κB occurred within the first 2 h after the stimulus and declined again within 24 h, the increase in DNA-binding activity of AP-1 was observed immediately after the treatment and remained elevated for a longer time. Supershift analysis of the nuclear extracts indicated that the AP-1 complex in oligodendrocytes consists of c-Jun, c-Fos, JunD, and possibly JunB (Fig. 4). RT-PCR analysis further indicated that c-jun, c-fos, and junD mRNA expression levels were stimulated after oxidative stress. To exert its function as a transcription factor, c-Fos has to dimerize with proteins of the Jun or other families of transcription factors, whereas c-Jun in the AP-1 complex forms either homodimers or heterodimers with Fos or other transcription factors (Gass and Herdegen, 1995). The individual family members were shown to influence DNA binding and the transactivation potential of AP-1 in different ways, and distinct expression patterns following brain ischemia and injury were observed (Raghupathi and McIntosh, 1996 ; for review, see Tong et al., 1998). Their role in the regulation of apoptotic cell death has been investigated mostly in neuronal cells (Dragunow and Preston, 1995). In sympathetic neurons, an increase of c-Jun was associated with apoptosis due to nerve growth factor deprivation (Eilers et al., 1998), and c-Jun was expressed in response to axotomy in a variety of systems (Herdegen et al., 1997). JunD levels were elevated after focal brain injury (Pennypacker, 1997) and transient fore-brain ischemia (Kamme and Wieloch, 1996). The activation of c-Jun and JunD in response to stress situations and during neurodegenerative diseases is involved in the regulation of death and survival, which depend on the diverse signal transduction pathways and the specificity of the c-Jun complexes (Herdegen et al., 1997). It has been suggested that an early and persistent c-Jun expression is correlated with the potential to regenerate. It is interesting that c-Jun immunoreactivity was found in neurons in the vicinity of subacute plaques and might be a consistent reaction not only to axonal damage, but also to demyelinating processes (Martin et al., 1996). c-Jun N-terminal kinases, also known as stress-activated protein kinases, regulate the activity of c-Jun. c-Jun N-terminal kinases are activated in primary glial cultures by tumor necrosis factor-α, ceramide, UV irradiation, and several other stresses (Zhang et al., 1996) and by H2O2 in the CG4 oligodendroglia cell line (Bhat and Zhang, 1999). Its involvement in the regulation of oligodendroglia cell death and the onset of H2O2-induced PCD remains to be established.
The activation of NF-κB by ROS has been reported in many studies (for reviews, see Flohé et al., 1997 ; Piette et al., 1997). Prominent stimuli are tumor necrosis factor-α, TPA, or interleukin-1, all of which seem to exert their activation via ROS formation, because they can be blocked by antioxidants (Schreck et al., 1992). Hydroperoxides, including products of the multiple lipoxygenases and cyclooxygenases, in general might act as activators of NF-κB (Flohé et al., 1997 ; Piette et al., 1997). Here we report for the first time that oxidative stress exerted by low concentrations of H2O2 leads to the transcriptional activation, nuclear translocation, and an increase in DNA-binding activity of NF-κB in primary cultures of mature oligodendrocytes. The transcriptional activation and nuclear translocation are transient and decrease after 6 h. Supershift analysis demonstrates that p50, p65, and c-Rel proteins contribute to the NF-κB complex (Fig. 5). The p65 protein of the complex plays a key role in NF-κB function and is required for transactivation (Flohé et al., 1997). The activation of NF-κB involves cascades of phosphorylation and dephosphorylation. It remains to be elucidated in which way H2O2 or peroxides activate the kinases that subsequently phosphorylate IκB and lead to its ubiquitination and degradation (Palombella et al., 1994 ; Flohé et al., 1997), finally permitting the NF-κB complex to enter the nucleus. Some evidence has accumulated that the activation of NF-κB might rescue cells from oxidative stress-induced apoptosis (Mattson et al., 1997 ; Tagliatela et al., 1997), and increased NF-κB seems to be an important signal in a number of neurodegenerative diseases. Thus, NF-κB activation may be an important signal to prevent cellular degeneration (O'Neill and Kaltschmidt, 1997). Yet whether NF-κB is protective or not might depend on the activation and contribution of the individual subunits and again on the specificity of the individual cells.
DNA cleavage into oligonucleosomal-sized fragments is one of the key features of apoptotic cell death (Leist and Nicotera, 1998). We have shown before that a typical DNA ladder is observed in oligodendrocytes after treatment with low concentrations of H2O2 (10-100 μM) during a 24-h recovery period (Richter-Landsberg and Vollgraf, 1998). The present study indicates that the first signs of DNA fragmentation are seen already during the first hour after the treatment (Fig. 8). Oxidative stress-induced DNA fragmentation in oligodendrocytes could be attenuated by the nuclease inhibitor ATA, and by the preincubation with the antioxidants vitamin E or PDTC in a concentration-dependent manner (Fig. 9a). PDTC represents a radical scavenger with chelating properties for heavy metals, but also has been shown to inhibit NF-κB activation by suppressing the release of the inhibitory subunit IκB, and does not interfere with the induction of DNA-binding activity of AP-1 (Schreck et al., 1992). DNA fragmentation could also be prevented by preincubation of oligodendrocytes with the iron chelator DFO (Fig. 9b). DFO protected oligodendrocytes from H2O2-induced cytotoxic effects (Fig. 6) and suppressed the formation of free radicals, as assessed by DCF measurements (Fig. 1). On the other hand, DFO alone led to a slight increase and in combination with H2O2 synergistically induced DNA-binding activities of AP-1 and NF-κB in oligodendrocytes. DFO chelates Fe3+ and has been shown to block the Fenton reaction, i.e., the formation of the hydroxyl radical from H2O2 mediated by Fe2+ (Galey, 1997). Oligodendrocytes require a high iron content for metabolic processes and lipid biosynthesis, are the major iron-containing cells in the brain, and contain high levels of transferrin and ferritin (Connor and Menzies, 1996). They express predominantly the H subunit of ferritin, which rapidly can sequester and release iron (Blissman et al., 1996). Oxidative stress might induce a local release of iron from the intracellular stores, and it has been suggested that because of the high iron content and requirement, oligodendrocytes are more susceptible to oxidative damage than other cells in the brain. The data presented here suggest that DFO rescues oligodendrocytes from H2O2-induced cytotoxic effects and the onset of apoptosis, as monitored by DNA fragmentation, by preventing the iron-catalyzed formation of the hydroxyl radical.
The question remains which form of oxidant species activates the DNA-binding activity of AP-1 and NF-κB in oligodendrocytes. Several lines of evidence support the hypothesis that low levels of H2O2 may act as a modulator of cell function (Remacle et al., 1995) and are directly involved in the activation of NF-κB (Sen and Packer, 1996). Our data point to the conclusion that indeed low levels of H2O2 are directly involved in the activation of AP-1 and NF-κB and might contribute to signal transduction pathways promoting cell survival. Yet the cellular defense program activated by the stress response may not be sufficient, if the devastating hydroxyl radicals are formed. DFO prevents the formation of hydroxyl radicals, and the incubation of oligodendrocytes with DFO in combination with H2O2 potentiated the effect on the DNA-binding activity of the transcription factors observed by H2O2 alone. Thus, it might be speculated that a pretreatment with DFO augments or stabilizes the local accumulation of H2O2, thereby leading to a further increase in DNA-binding activity of AP-1 and NF-κB. The DNA-binding activity of both transcription factors reaches a maximum after 2-6 h, whereas the first signs of the initiation of PCD are observable as early as 1 h after the treatment. Hence, without DFO pretreatment, the DNA binding and nuclear translocation of AP-1 and NF-κB might not be timely and sufficient to prevent the onset and subsequent execution of the apoptotic death program, and the formation and action of the hydroxyl radicals rather promote cell death mechanisms.
In conclusion, the data presented in this article suggest that activation of transcription factors AP-1 and NF-κB is associated with the onset of apoptosis and the regulation of cell survival in oligodendrocytes after exposure to H2O2, and thus might have an impact on the pathogenesis of demyelinating diseases of the CNS. In the presence of the iron chelator DFO, the oxidative damage is limited and the activation of the transcription factors seems to sustain cell viability.
We thank Angelika Spanjer for expert technicl assistance with the cell cultures and Udo Gansel for help with the DNA ladders. This work is part of the doctoral thesis of Ulrich Vollgraf.