- Original Articles
- Open Access
Tumor Necrosis Factor and Reactive Oxygen Species Cooperative Cytotoxicity Is Mediated via Inhibition of NF-κB
Molecular Medicinevolume 6, pages1028–1041 (2000)
Tumor necrosis factor alpha (TNFα) plays a key role in pathogenesis of brain injury. However, TNFα exhibits no cytotoxicity in primary cultures of brain cells. This discrepancy suggests that other pathogenic stimuli that exist in the setting of brain injury precipitate TNFα cytotoxicity. The hypothesis was tested that reactive oxygen species (ROS), that are released early after brain injury, act synergistically with TNFα in causing cell death.
Materials and Methods
Cultured human and rat brain capillary endothelial cells (RBEC), and cortical astrocytes were treated with TNFα alone or together with different doses of H2O2, and apoptotic cell death and DNA fragmentation were measured by means of 3′-OH-terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) and Hoechst fluorescence assay, respectively. The effect of H2O2 on TNFα-induced activation of nuclear factor kappa B (NF-κB) was measured by Western blots of cytoplasmic and nuclear extracts of RBEC using anti-inhibitor of NF-κB (IκB) and anti-p65 subunit of NF-κB antibodies. Nuclear translocation of NF-κB was investigated by immunofluorescent staining of RBEC with anti-p65 antibodies.
TNFα alone had no cytotoxic effect in brain endothelial cells and astrocytes at concentrations up to100 ng/ml. Co-treatment with 5–10 µM of H2O2 caused a two-fold increase in the number of apoptotic cells 24 hr later. Similar doses (1–3 µM) of H2O2 initiated early DNA fragmentation. H2O2 inhibited TNFα-induced accumulation of p65 in the nucleus, although it had no effect on degradation of the IκB in cytoplasm. Immunostaining confirmed that H2O2 inhibited p65 transport to the nucleus.
Reactive oxygen species could act synergistically with TNFα in causing cytotoxicity via inhibition of a cytoprotective branch of TNFα signaling pathways, which starts with NF-κB activation.
The pleiotropic cytokine tumor necrosis factor alpha (TNFα) exerts biological activity in CNS (1–4). TNFα effects in brain parenchyma are shown to play a key role in brain injury (5–8). High TNFα levels have been detected in brain trauma (9–11) and ischemia (12–15). Neutralization of TNFα by TNFα-binding protein had a protective effect against focal ischemia (16,17) and trauma (18) and an inhibitor of TNFα synthesis, dexanabinol has a protective effect in closed head injury (CHI) and MCAO (19,20).
However, in vitro studies demonstrate that TNFα is not cytotoxic in brain cells. It even causes protection of cultured neurons (21–23). Cultured cortical astrocytes and brain endothelial cells treated with TNFα for 48 hr exhibit no signs of apoptosis (24). The discrepancy between observations of a TNFα pathogenic function in animal models of brain injury and its lack of cytotoxic effect on brain cells in vitro suggests that other pathogenic stimuli contribute to TNFα cytotoxicity in the setting of brain injury.
Reactive oxygen species are among the most toxic mediators released early after brain injury. The brain is extremely vulnerable to oxidative damage (25). We have shown that the synthetic spin-trap antioxidant from the nitroxide family, Tempol, improved recovery and protected the blood-brain barrier in a rat model of CHI (26). Similar protection was found after CHI in heat-acclimated rats, in which the endogenous antioxidants have been shown to be elevated (27). On the other hand, TNFα levels and activity were not affected in Tempol-treated or heat-acclimated animals (28), suggesting that ROS could alter TNFα signaling rather than TNFα synthesis and thus precipitate TNFα cytotoxicity. Similarly, the same spin-trap molecule, was used in studies of bacterial and cultured mammalian cells and was shown to provide cytoprotection from the toxicity induced by TNFα (29). Transcription of many pro-inflammatory, immune, and apoptotic genes, which are induced by TNFα, is dependent on activation of nuclear factor kappa B (NF-κB). Each step of NF-κB activation and DNA binding is redox sensitive (30). Taken together, these observations suggest that the point of intersection of TNFα and ROS, which both accompany brain insults, could be NF-κB. The present study was designed to test this hypothesis. We demonstrate here that sublethal doses of H2O2 abrogate natural resistance of different types of brain cells to TNFα by inhibiting TNFα-induced activation of NF-κB.
Materials and Methods
Human brain capillary endothelial cells (HBEC) cultures have been previously described (31). Rat brain capillary endothelial cells (RBEC) were prepared from adult WKY rat brains as for human cultures except that fetal bovine serum was substituted for human serum and 90 µg/ml heparin was added to the medium. The purity of the both HBEC and RBEC was > 95% as determined by positive immunostaining for von Willebrand factor (Factor VIII), and angiotensin-converting enzyme, incorporation of acetylated low-density lipoprotein, and by negative staining for glial cells (GFAP, galactocerebroside, ED-2), muscle cells (α-actin) and pericytes (tropomyosin). Human brain capillary endothelial cells were at passage 4. Rat brain capillary endothelial cells were at passages 7–12 and were seeded from four different batches of brain tissue. Cortical astrocyte cultures were established from 3-day-old Sprague-Dawley rats as described (24).
Visualization and quantitation of apoptotic cells was performed by means of 3′-OH-terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNNEL) using “In Situ Cell Death Detection Kit (POD)” (Boehringer Mannheim, Germany). Human brain capillary endothelial cells and RBEC were treated with 15 ng/ml human TNFα (Endogen, Woburn, MA, USA) and 20 ng/ml rat TNFα (Chemicon International Inc., Temecula, CA, USA), respectively, or/and with different doses of H2O2 for 24 hr, fixed with 4% paraformaldehyde for 30 min and stained according to the manufacture’s protocol. The samples were analyzed with a Zeiss Axiovert 100 light microscope (20× objective). Digitized images of 15 microscopic fields per each experimental condition were generated using a digital CCD Camera C4742-95-12 (Hamamatsu) and Zeiss Axio-Version 2 software. The same microscope and camera settings were used for all samples. The number of apoptotic cells within each image was determined by means of Scion Image (NIH Image for PC) computer program. Briefly, background was subtracted from each image and each image was transformed into a binary image, which permitted measurement of the area occupied by all of the cells in the image (area T). All the images were then reversed to multi-gray mode, and the average optical density of nonstained cells was measured for all images acquired. The value of the mean density of nonstained cells was subtracted from all the images and the remaining areas of higher density (area A) (these were positively stained apoptotic nuclei) were again thresholded to binary images and measured.
Percentage of apoptotic cells was calculated as follows:
where: A is the sum of the TUNEL-positive areas in the image, and T is the sum of the areas occupied by all cells in the image.
Quantitation of DNA fragmentation was performed by means of a fluorescent, cell-permeable, DNA-binding dye, Hoechst 33342 as previously described (24). Hoechst fluorescence upon binding to DNA is inversely proportional to the degree of DNA fragmentation in the cells undergoing apoptosis (32). Briefly, TNFα/H2O2-treated astrocytes or RBEC, plated in 96 well microtiter plates were incubated with 25 µM Hoechst 33342 (Molecular Probes, Eugene, OR, USA) in PBS added at 100 µl/well for 45 min at 37°C. Cell fluorescence was measured using a CytoFluor 4000 fluorescent plate reader (PerSeptive Biosystems, Framingham, MA, USA) at excitation/emission wavelengths of 360/460 nm. Background fluorescence was measured on each plate and subtracted. Each data point was a mean of fluorescence readings of eight wells (variability was less than 30%). The percentage of apoptotic cells was calculated from Hoechst fluorescence by means of the following formula:
where Fmax is fluorescence of untreated healthy control cultures, Fmin is fluorescence of cells treated with the cytotoxic alkylating agent methyl iodide, and F is fluorescence of unknown sample.
Immunofluorescent Staining for p65 Subunit of NF-κB
Endothelial cells were treated with 20 ng/ml TNFα and with or without various doses of H2O2 for 25 min. Cells were fixed with ethanol for 2 min and then with 3.7% formaldehyde for 5 min and immunostained with rabbit polyclonal antibody directed against the p65 subunit of NF-κB (Santa Cruz, cat. #sc109) or with mouse monoclonal antibody against the activated form of p65 (Boehringer-Manheim Cat. #1697838) both antibodies were at 1:50 dilution according to Kaltschmidt et al. (69). Detection was performed with anti-rabbit and anti-mouse corresponding biotinylated secondary antibody, followed by addition of streptavidin-Cy3. Digitized images of the fluorescent cells were generated using the same microscope and camera as for TUNEL experiments (40× objective).
Preparation of Cytosolic and Nuclear Extracts
Rat brain capillary endothelial cells were grown to confluency in 60-mm dishes. Rat TNFα (Chemicon International, Temecula, CA, USA) was added to the cells at 20 ng/ml with or without 2 µM H2O2. At the indicated times, cells were placed on ice, washed twice with PBS, and then scraped off into 800-µl PBS containing protease inhibitor cocktail (Boehringer Mannheim), phosphatase inhibitors (10 mM NaF, 1 mM Na3VO4), and 1 mM dithiothreitol (DTT). Cells were pelleted in a microcentrifuge for 1 min at 2,500 rpm, resuspended in 5 volumes (∼100 µl/dish) of a low salt buffer A (10 mM HEPES, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, pH = 7.9) and incubated for 15 min on ice. At the end of incubation NP-40 was added to the lysates at final concentration 0.1%. Samples were vigorously vortexed for 20 sec and centrifuged at 11,000 rpm for 1 min. Cytoplasmic fraction was transferred to a new Eppendorf tube, and frozen at −70°C. The pellet was resuspended in 30–50 µl/dish high salt buffer B (20 mM HEPES, 400 mM NaCl, 50 mM KCl, 1 mM EDTA, 1 mM EGTA, 10% (w/v) glycerol, pH = 7.9). The samples were shaken at a high speed for 30 min and then microcentrifuged at 14,000 rpm for 5 min. The supernatant was frozen at −70°C. Prior to freezing, a 2-µl aliquot from each cytosolic and nuclear extract was taken for protein determination (Bio-Rad Laboratories, Hercules, CA, USA). All reagents were ice-cold, and all procedures were performed on ice.
Western Blots for NF-κB Studies
All the buffers, 4–12% Tris-glycine gradient mini-gels, nitrocellulose membranes, and electrophoresis equipment were from Novex (San Diego, CA, USA). Cytosolic or nuclear extracts were boiled in equal volumes of loading buffer/1 mM DTT for 3 min and loaded on a gel at 10 µg protein/lane for determination of p65 concentrations in nuclear extracts, and at 15 µg protein/lane for IκB determination in cytosolic fractions. Separated proteins were transferred to nitrocellulose membrane. Immunoblotting was performed as previously described (33). Anti-IκBα and anti-p65 rabbit polyclonal antibodies were from Santa Cruz Biotechnology (SC-372 and SC-203, respectively).
Electrophoretic mobility shift assays (EMSA) were performed by using the Gel Shift Assay System (Promega) according to the manufacturer’s instruction. Nuclear extracts were prepared from untreated control cells, and cells treated with TNFα alone or together with 5 µM H2O2. The NF-κB consensus oligonucleotide (AGTTGAGGGGACTTTCCCAGGC) was endlabeled using [γ-32P]ATP. The reaction mixture contained 4 µl 5× gel shift binding buffer, 1 µl of 32P labeled NF-κB (15,000 cpm) consensus oligonucleotide probe, 10 µg of nuclear extract in a total volume of 20 µl. The reaction was incubated at room temperature for 20 min. After incubation, the samples were loaded on 4% nondenaturing polyacrylamide gel and were electrophoresed at 150 volts for 3 hr. The gel was dried and autoradiograph was developed. After the autoradiography, the bands were cut out, counted, and the specific radioactivity associated with each band was calculated.
Statistical analysis was carried out by oneway ANOVA followed by Dunnett’s test and by one-way ANOVA for repeated measures followed by Turkey test using SigmaStat Software (p values < .05 were considered statistically significant). All graph data are presented as Mean ± SD.
TNFα and H2O2 Have a Synergistic Effect on Induction of Apoptosis
It has previously been shown that human TNFα was not cytotoxic for cultured HBEC at doses of 15 ng/ml (250 EU/ml), although this dose caused cell activation (34). Similarly in RBEC, while causing activation of ICAM-1 adhesion ligand, TNFα exhibited no cytotoxicity at the dose of 20 ng/ml (24). Thus, these doses of TNFα were adopted for current investigation. To test the hypothesis that the combination of TNFα and ROS will have a synergistic effect on cell viability, HBEC were treated with a combination of TNFα (15 ng/ml) and H2O2 (100 µM) for 4 hr. Cell viability was monitored under phase contrast field. When added separately, TNFα and H2O2 caused no morphological changes (Fig. 1B and C) when compared to control cultures (Fig. 1A), addition of both agonists resulted in cell death (Fig. 1D).
To further investigate the effect of H2O2 cells were treated with lower doses of H2O2 (5 and 10 µM for HBEC and RBEC, respectively) in the presence of TNFα for 24 hr and the number of apoptotic cells was quantitated with TUNEL assay. The results are presented in Figure 2A (HBEC) and Figure 2B (RBEC). The top panel of each figure presents photomicrographs of endothelial cultures after treatment. The bottom pannel presents binary images of corresponding photomicrographs with highlighted TUNEL-positive areas. TNFα alone for 24 hr remained healthy, with few TUNEL-positive cells (photomicrographs a and binary images d in both Fig. 2A and B). Similarly, H2O2 alone had no effect on cell viability (photomicrograph b and binary image e in both Fig. 2A and B). Addition of H2O2 together with TNFα resulted in the appearance of many TUNEL-positive cells exhibiting apoptotic morphology (shrinking of the cytoplasm, chromatin condensation, pyknosis) (photomicrograph c and binary image f in both Fig. 2A and B). Results of quantitation of percentage of TUNEL-positive nuclei with ScionImage program are presented in Table 1. It demonstrates that endothelial cell cultures treated with both TNFα and H2O2 exhibited two times more apoptotic cells than control cultures or cultures treated with TNFα or H2O2 alone (p < .05).
To study the synergistic effect of TNFα and H2O2 in more detail, we have used the Hoechst assay, which is apparently more sensitive indicator of DNA fragmentation (32). Besides endothelial cells, cortical astrocytes also have been studied. Cells were treated with TNFα in combination with various doses (0–10 µM) of H2O2. Because DNA fragmentation occurs early in the course of apoptosis, the cultures were evaluated 15–16 hr after addition of TNFα and of H2O2. Lower doses of H2O2 (1–3 µM) were required to initiate DNA fragmentation measured in this assay compared to TUNEL assay as cells need a longer time to undergo fullblown apoptosis. Rat brain capillary endothelial cells were more sensitive to TNFα/H2O2 treatment than astrocytes. Addition of as little as 0.3–1 µM of H2O2 together with 10 ng/ml TNFα resulted in apoptosis of the majority of the RBEC (Fig. 3A), whereas 1–3 µM H2O2 caused the same effect in astrocytes (Fig. 3B).
H2O2 Inhibits TNFα-induced Activation of NF-κB
Transcription of many TNFα-activated genes depends on the transcription factor, NF-κB. However, NF-κB activation could be altered by ROS and antioxidants. Thus, we hypothesized that H2O2 might interfere with TNFα-induced NF-κB activation precipitating cell death. To test this hypothesis, we have studied NF-κB activation in RBEC. Cells were treated with 20 ng/ml TNFα for various times, and then IκB and p65 levels were determined in cytoplasmic and nuclear extracts, respectively, by means of Western blotting and ScionImage analysis as described in Materials and Methods. The results of these experiments are presented in Figure 4A–D. IκB degradation in cytoplasm began as early as 10 min after TNFα addition; at 20 min IκB levels in cytoplasm dropped to 22.5 ± 6.2% (mean ± SD; n = 4) of baseline levels and remained that low for another 10 min. At 1 hr after TNFα addition, IκB levels returned to about 70% of control values (Fig. 4A and B). Addition of 2 µM H2O2 had no significant effect on IκB degradation. IκB degradation had almost identical kinetics in the absence or presence of H2O2. Addition of 2 µM H2O2 alone without TNFα caused no degradation of IκB. At 20 min, IκB levels were 109.8 ± 0.05% control (mean ± SD; n = 2). There was a constitutive p65 presence in the nucleus of brain cells. In all experiments (n = 5), antibody identified two bands. We suggest that one of the bands is a phosphorylated form of p65. Accumulation of p65 subunit in the nucleus paralleled degradation of IκB in the cytoplasm and peaked at 20 min after TNFα addition (153.0 ± 25.8% baseline; mean ± SD; n = 5) (Fig. 4C and D). Addition of 2 µM H2O2 to the cells simultaneously with TNFα completely inhibited p65 translocation to the nucleus: measured p65 levels at 20 min were 109.5 ± 21.1% baseline (p < 0.05 vs. TNFα). Addition of 2 µM H2O2 alone without TNFα resulted in no significant increase of p65 levels in the nucleus. At 20 min, p65 levels were 110.6 ± 5.7% control (mean ± SD; n = 3).
The results of Western blots were confirmed with electrophoretic mobility shift assay. Nuclear extracts of untreated HBEC cells had little DNA-binding activity. Drastic up-regulation of DNA-binding activity was caused by TNFα alone. However, addition of 5 µM H2O2 together with TNFα inhibited DNA binding by 40% (Fig. 4E and F).
H2O2 Inhibits Nuclear Transport of p65 Subunit of NF-κB
Two subunits comprise NF-κB, which belong to Rel family of proteins. The most frequently found NF-κB heterodimer is p65/p50. Interaction of the subunits with the inhibitor of NF-κB (IκB) masks the nuclear translocation sequence of NF-κB and retains the heterodimer in cytoplasm. Upon activation with TNFα and other agonists, IκB gets phosphorylated by a specific kinase, ubiquitinated, and subsequently degraded by proteasome peptidases (35). Removal of IκB-alpha uncovers the nuclear localization signals of subunits of NF-κB, allowing the complex to enter the nucleus, bind to DNA, and affect gene expression. The results of Western blots presented in Figure 4 demonstrate that, although TNFα-induced IκB degradation was not affected by H2O2 (Fig. 4A and B), and should result in release of a free p65/p55 heterodimer in cytoplasm, the p65 subunit did not appear in the nucleus (Fig. 4C and D), suggesting that transport of p65 from cytoplasm to the nucleus was compromised. To confirm this assumption, RBEC were treated with 20 ng/ml TNF-α or with 5 µM H2O2 or with both agonists for 30 min and then fixed and immunostained with anti-p65 antibodies. In control, untreated cells (Fig. 5A) and in the cells treated with H2O2 alone (Fig. 5B), immunofluorescence was spread over cytoplasm and was not found in the nucleus. Treatment with TNFα caused redistribution of fluorescence with maximal signal coming from the nucleus (arrows) (Fig. 5C). Although there was some background fluorescence in the nuclei of control and H2O2-treated cells (Fig. 5A and B), it had an appearance of small granules, whereas in TNF-treated cells (Fig. 5C), the whole nucleus was highlighted. Addition of TNFα together with H2O2 resulted in no accumulation of fluorescence in the nucleus, and most of the bright spots are localized in perinuclear forming a dense circle (Fig. 5D). This pattern was somewhat different from that of control cells, and was consistently observed in all captured images. Similar results were obtained in HBEC using an antibody directed against an activated form of p65 (data not shown).
Inhibitor of NF-κB Causes Apoptosis
To investigate the role NF-κB in survival of endothelial cells, HBEC were incubated with different doses (1-10 µM) of the NF-κB inhibitor BAY 11-7082 (E-3-[4-methylphenylsulfonyl]-2-propenenitrile) (Biomol Research Laboratories, Inc. Plymouth Meeting, PA, USA). Cell morphology was analyzed 4 hr later with a Zeiss Axiovert 100 light microscope (20× objective) and the images were captured as described in Materials and Methods. At a dose as low as 1 µM, the inhibitor caused apoptosis (Fig. 1E).
Ischemic and traumatic brain injury are accompanied by oxidative stress (36–38) and release of proinflammatory cytokines, TNFα and IL-1β (8,39). Although these pathogenic reactions have been extensively studied, there is no consensus of opinion on mechanisms of their action and interaction. Oxidants have been shown to stimulate signaling pathways usually triggered by growth factors (activation of protein tyrosine kinases and phosphatases, PKC and mitogen-activated kinases, phospholipases Cγ and A2 and Ca2+ (40). Similarly, it has been thought that TNFα also induces release of ROS (mainly H2O2) in mitochondria (41) and through NADPH oxidase (42) and then ROS act as messengers in TNFα signaling pathways leading to NF-κB-dependent transcription of pro-inflammatory genes (43–45) and to cell death (46). However, although H2O2 activates NF-κB in some cell types, it fails to do so in human endothelial cells (47,48), in lymphoblastoid (49), and in monocytic cell lines (50). Moreover, since the first observations of the anti-apoptotic effects of NF-κB have been published (51–53), the role of NF-κB in cell death has been revised. Much evidence has emerged that demonstrates that NF-κB activates transcription of protective genes in different types of cells (54), including brain cells (55,56) rather than causing cell death. Furthermore, inhibition of NF-κB sensitizes neurons to cytotoxic effects of amyloid beta (57), and activation of NF-κB promotes neuronal survival (58,59). These observations suggest an alternative mechanism for interaction between TNFα and ROS in induction of cell death.
In this work, we present evidence supporting the hypothesis that ROS cooperate with TNFα and induce cell death via inhibition of NF-κB. For the first time, we demonstrate that low doses of H2O2, which are not capable of causing cell death on their own, synergize with TNFα and unmask TNFα cytotoxicity in cultured brain endothelial cells and astrocytes. This conclusion is based on the results of morphological studies and TUNEL staining for apoptotic cells, as well as on quantitation of DNA fragmentation with Hoechst 33342, which is a cell-permeable DNA-binding fluorescent dye. Hoechst fluorescence is inversely proportional to the degree of DNA fragmentation (60). A synergistic effect of TNFα and H2O2 was demonstrated in both assays. When treated with TNFα alone for 24 hr, HBEC, RBEC, and astrocytes showed a low rate of apoptosis, most probably caused by culture conditions. However, addition of low doses of H2O2, together with TNFα, results in early DNA fragmentation followed by appearance of TUNEL-positive apoptotic cells. Interestingly, in our preliminary experiments, we used a cell-impermeable fluorescent dye, ethidium homodimer, to assess cell viability. Healthy cells exclude ethidium, but those with a damaged plasma membrane, generally necrotic cells, accumulate the dye and become highly fluorescent. In these experiments, doses of H2O2 required to produce membrane permeability (ethidium assay) were higher (10–20 µM depending on the culture; data not shown) than those needed to cause DNA fragmentation (15 µM) in the presence of the same dose of TNFα. This dichotomy fits the definition of necrosis and apoptosis well. Necrosis is an uncontrolled degenerative phenomenon invariably caused by noxious stimuli and is the result of irreversible failure of membrane function. In contrast, apoptosis is a death process that involves a series of well-organized events that require active cell participation, and is primarily caused by physiological stimuli. Previous observations showing that low doses of ROS induce apoptosis, whereas necrosis occurs in cells exposed to higher doses of ROS (61–63) are in accordance with our findings.
Chemical reactivity and redox potentials of ROS range from reducing to oxidizing: superoxide anion radical can be reduced to H2O2 by superoxide dismutase, the latter can be further reduced by the Fenton reaction with iron to the hydroxyl radical, which is capable of oxidizing nucleic acids, lipids, and proteins. Employment of specific scavengers for different types of ROS has demonstrated that some but not all of these species activate NF-κB (64). It has been demonstrated that all the steps of NF-κB activation (IκB phosphorylation and degradation, p65/p55 nuclear translocation, and DNA binding) are redox-sensitive (65–68). Our data suggests that low doses of ROS have no effect on the steps involved in degradation of IκB but alter TNFα-induced nuclear translocation of at least the p65 subunit of NF-κB. Quantitation of IκB levels in TNFα-treated RBEC demonstrated almost complete disappearance of the inhibitor from the cytoplasm at 20 min; IκB gene itself has an NF-κB binding site, so it is quickly resynthesized (69). By 60 min ∼60% of IκB protein could be detected in TNFα-stimulated RBEC. These kinetics of IκB degradation and resynthesis did not change in the presence of H2O2, which means that free NF-κB heterodimer is released in the cytoplasm and should be transported to the nucleus. However, we were unable to detect increased levels of p65 in the nucleus in the cells treated with TNFα and H2O2, although TNFα alone triggered a 1.5-fold increase of p65 levels in the nucleus. This result was confirmed by gel-shift assay and immunostaining experiments, which demonstrated that TNFα failed to induce translocation of the p65 subunit to the nucleus in the presence of H2O2 and instead caused accumulation of the p65 in the perinuclear area. Similarly, high doses of amyloid beta peptide, known to elicit production of ROS, have been shown to inhibit nuclear transport of p65 (57,70).
Our data strongly suggest that H2O2 precipitates TNFα cytotoxicity by inhibiting transcription of the NF-κB-dependent protective genes. However, other mechanisms are not excluded by these studies. Thus treatment of the cells with low concentrations of H2O2 induces activation of caspases, cysteine proteases that constitute part of apoptotic machinery (63). In addition, TNFα-induced activation of sphingomyelinase and consequent release of ceramide, a phospholipid messenger implicated in apoptosis, could be prevented by antioxidants and stimulated by H2O2 in astrocytes (71). In our studies, we have shown that physiological doses of C-2 ceramide failed to induce apoptosis in cultured astrocytes and RBEC but higher doses were apoptotic (33). Another possibility is that ROS by interfering with ceramide metabolism allow higher intracellular levels of ceramide and cause cell death.
There is evidence that NF-κB is constitutively expressed in neuronal cells and mediates their resistance to different types of stress (55,72,73), and that neurons from mice lacking the p50 subunit of NF-κB are more vulnerable to excitotoxic stress (74). Our Western blot data demonstrate constitutive presence of p65 in the nucleus of brain microvascular cells and inhibition of this NF-κB activity in HBEC cultures by low doses of NF-κB inhibitor BAY11-7082 resulted in rapid apoptosis. Our results are consistent with findings of Taglialatela et al. (76). However, observations in animals are not consistent. Thus suppression of NF-κB activity in brain by a specific inhibitor resulted in DNA fragmentation (76), but TNFα receptor knockout mice, which exhibited delayed up-regulation of NF-κB after traumatic brain injury, had a larger average lesion volume and blood brain barrier breach than wild-type animals (77). Also, p50 knockout mice tolerated ischemic injury better than wild-type animals (74). NF-κB activity was shown to increase after brain trauma (79) and in the model of transient focal ischemia 72 hr after reperfusion. (78). In the latter model and also in the model of intracerebral hemorrhage, activated NF-κB colocalized with apoptotic cells (78,80). However, in a rat model of permanent MCAO, activated NF-κB immunore-activity decreased from basal levels already at 2 hr after onset of ischemia and remained undetectable up to 5 days (75). Interestingly, antioxidant-dependent protection against transient focal ischemia was associated with inhibition of NF-κB (81) but in global ischemia antioxidants inhibited only persistent NF-κB activity in hyppocampal CA1 neurons, whereas transient activation of NF-κB seemed to be protective (82). Overexpression of Mn-SOD in human breast cancer MCF-7 cells completely abolished TNF-mediated NF-κB activation, and caused apoptosis (83). These contradictions call for more detailed studies of NF-B activation in vivo. Time, localization, and variations in NF-κB heterodimer composition should be taken into account.
In conclusion, the data presented here suggests a new pharmacological approach to the treatment of brain injury. Instead of targeting TNFα and ROS separately, one might want to interfere with the cross-talk mechanisms of these two pathogenic pathways that modulate NF-κB dependent anti-apoptotic signaling.
Tracey KJ, Cerami A. (1993) Tumor necrosis factor, other cytokines and disease. Ann. Rev. Cell Biol. 9: 317–342.
Rothwell NJ, Hopkins SJ. (1995) Cytokines and the nervous system II: action and mechanisms of action. Trends Neurosci. 18: 130–136.
Merril JE, Benveniste EN. (1996) Cytokines in inflammatory brain lesions: helpful and harmful. Trends Neurosci. 19: 331–338.
Barger S. (1998) Tumor necrosis factor, the good, the bad and the umbra. In: Mattson MP (ed.) Neuroprotective Signal Transduction. Humana Press, Totowa, NJ, pp. 163–183.
Hallenbeck JM. (1997) Cytokines, macrophages, and leukocytes in brain ischemia. Neurology 49(suppl 4): S5–9.
Mattson MP, Barger SW, Furukawa K, Bruce AJ, Wyss-Coray T, Mark RJ, Mucke L. (1997) Cellular signaling role of TGF beta, TNF alpha and beta APP in brain injury responses and Alzheimer’s disease. Brain Res. Rev. 23: 47–61.
Barone FC, Feuerstein GZ. (1999) Inflammatory mediators and stroke: new opportunities for novel therapeutics. J. Cereb. Blood Flow Metab. 19: 819–834.
Shohami E, Ginis I, Hallenbeck JM. (1999) Dual role of tumor necrosis factor alpha in brain injury. Cytokine Growth Factor Rev. 10: 119–130.
Shohami E, Novikov M, Bass R, Yamin A, Gallily R. (1994) Closed head injury triggers early production of TNF and IL-6 by brain tissue. J. Cereb. Blood Flow Metab. 14: 615–619.
Taupin V, Toulmond S, Serrano A, Benavides J, Zavala F. (1993) Increase in IL-6, IL-1 and TNF levels in rat brain following traumatic lesion. Influence of pre- and post-traumatic treatment with Ro5 4864, a peripheral-type (p site) benzo-diazepine ligand. J. Neuroimmunol. 42: 177–185.
Fan L, Young PR, Barone FC, Feuerstein GZ, Smith DH, McIntosh TK. (1996) Experimental brain injury induces differential expression of tumor necrosis factor-α mRNA in the CNS. Molec. Brain Res. 36: 287–291.
Liu T, Clark RK, McDonnel PC, Toung PR, White RF, Barone FC, Feuerstein FZ. (1994) Tumor necrosis factor-alpha expression in ischemic neurons. Stroke 25: 1481–1488.
Meirstrell ME III, Botchkina GI, Wang H, et al. (1997) Tumor necrosis factor is a brain damaging cytokine in cerebral ischemia. Shock 8: 341–348.
Yamasaki T, Kikuchi H, Moritake K, et al. (1992) A morphological and ultrastructural investigation of normal mouse brain tissue after intracerebral injection of tumor necrosis factor. J. Neurosurg. 77: 279–287.
Uno H, Matsuyama T, Akita H, Nishimura H, Sugita M. (1997) Induction of tumor necrosis factor-alpha in the mouse hippocampus following transient forebrain ischemia. J. Cereb. Blood Flow Metab. 17: 491–499.
Dawson D, Martin D, Hallenbeck JM. (1996) Inhibition of tumor necrosis factor-alpha reduces focal cerebral ischemic injury in the spontaneously hypertensive rat. Neurosci. Lett. 218: 41–44.
Nawashiro H, Martin D, Hallenbeck JM. (1997) Inhibition of tumor necrosis factor ameliorated brain infarction in mice. J. Cereb. Blood Flow Metab. 17: 229–232.
Shohami E, Bass R, Wallach D, Yamin A, Gallily R. (1996) Inhibition of tumor necrosis factor (TNF-alpha) activity in rat brain is associated with cerebroprotecion after closed head injury. J. Cereb. Blood Flow Metab. 16: 378–384.
Shohami E, Gallily R, Mechoulam R, Bass R, Ben-Hur T. (1997) Cytokine production in the brain following closed head injury: dexanabinol (HU-211) is a novel TNFα inhibitor and an effective neuroprotectant. J. Neuroimmunol. 72: 169–177.
Leker R, Shohami E, Abramsky O, Ovadia H. (1999) Dexanabinol: a novel neuroprotective drug in experimental focal cerebral ischemia. J. Neurol. Sci. 162: 114–115.
Ito A, Horigome K. (1995) Ceramide prevents neuronal programmed cell death induced by nerve growth factor deprivation. J. Neurochem. 65: 463–466.
Goodman Y, Mattson MP. (1996) Ceramide protects hippocampal neurons against excitotoxic and oxidative insults, and amyloid beta-peptide toxicity. J. Neurochem. 66: 869–872.
Liu J, Ginis I, Spatz M, Hallenbeck JM. (2000) Hypoxic preconditioning protects cultured neurons against hypoxic stress via TNF-α and ceramide. Am. J. Physiol. (Cell Physiol.) 278: C144.
Ginis I, Schweizer U, Brenner M, Liu J, Azzam N, Spatz M, Hallenbeck JM. (1999) TNF-α pre-treatment prevents subsequent activation of cultured brain cells with TNF-α and hypoxia via ceramide. Am. J. Physiol. (Cell Physiol.) 276: C1171–1183.
Halliwell B. (1992) Reactive oxygen species and the central nervous system. J. Neurochem. 59: 1609–1623.
Shohami E, Beit-Yannai E, Horowitz M, Kohen R. (1997) Oxidative stress in closed-head injury: brain antioxidant capacity as an indicator of functional outcome. J. Cereb. Blood Flow Metab. 17: 1007–1019.
Shohami E, Novikov M, Horowitz M. (1994) Long term exposure to heat reduces edema formation after closed head injury in the rat. Acta Neurochir. Suppl. (Wien) 60: 443–445.
Trembovler V, Beit-Yannai E, Younis F, Gallily R, Horowitz M, Shohami E. (1999) Antioxidants attenuate acute toxicity of tumor necrosis factor-alpha induced by brain injury in rat. J. Interferon Cytokine Res. 19: 791–795.
Pogrebniak H, Matthews W, Mitchell J, Russo A, Samuni A, Pass H. (1991) Spin trap protection from tumor necrosis factor cytotoxicity. J. Surg. Res. 50: 469–474.
Kaltschmidt B, Sparna T, Kaltschmidt C. (1999) Activation of NF-κB by reactive oxygen intermediates in nervous system. Antioxidants Redox Signaling 1: 129–144.
Spatz M, Kawai N, Merkel N, Bembry J, McCarron RM. (1997) Functional properties of cultured endothelial cells derived from large microvessels of human brain. Am. J. Physiol. 272: C231–C239.
Ginis I, Faller DV. (1997) Protection from apoptosis in human neutrophils is determined by the surface of adhesion. Am. J. Physiol. 272: C295–C309.
Ginis I, Mentzer SJ, Faller DV. (1995) Characterization of a hypoxia-responsive adhesion molecule for leukocytes on human endothelial cells. J. Immunol. 155: 802–810.
Ohara Y, McCarron RM, Hoffman TT, Sugano H, Bembrey J, Lenz FA, Spatz M. (in press) Adrenergic mediation of TNFa-stimulated ICAM-1 expression on human brain microvascular endothelial cells. Acta Neurochir.
Baeuerle PA. (1998) IkappaB-NF-kappaB structures: at the interface of inflammation control. Cell 95: 729–731.
Beit-Yannai E, Zhang R, Trembovler V, Samuni A, Shohami E. (1996) Cerebroprotective effect of stable nitroxide radicals in closed head injury in the rat. Brain Res. 717: 22–28.
Love S. (1998) Oxidative stress in brain ischemia. Brain Pathol. 9: 119–131.
Beit-Yannai E, Kohen R, Horowitz M, Trembovler V, Shohami E. (1997) Changes in biological reducing activity in rat brain following closed head injury: a cyclic voltammetry study in normal and acclimated rats. J. Cereb. Blood Flow Metabol. 17: 273–279.
Feuerstein GZ, Wang X, Barone FC (1998). The role of cytokines in the neuropathology of stroke and neurotrauma. Neuroimmunomodulation 5: 143–159.
Kamata H, Hirata H. (1999) Redox regulation of cellular signalling. Cell Signal 11: 1–14.
Schulze-Osthoff K, Bakker AC, Vanhaesebroeck B, Beyaert R, Jacob WA, Fiers W. (1992) Cytotoxic activity of tumor necrosis factor is mediated by early damage of mitochondrial functions. Evidence for the involvement of mitochondrial radical generation. J. Biol. Chem. 267: 5317–5323.
Lo YY, Cruz TF. (1995) Involvement of reactive oxygen species in cytokine and growth factor induction of c-fos expression in chondrocytes. J. Biol. Chem. 270: 11727–11730.
Schreck R, Rieber P, Baeuerle PA. (1991) Reactive oxygen intermediates as apparently widely used messengers in the activation of the NF-kappa B transcription factor and HIV-1. EMBO J. 10: 2247–2258.
Rahman A, Kefer J, Bando M, Niles WD, Malik AB. (1998) E-selectin expression in human endothelial cells by TNF-alpha-induced oxidant generation and NF-kappaB activation. Am. J. Physiol. 275: L533–544.
Lo YYC, Wong JMS, Cruz TF. (1996) Reactive oxygen species mediate cytokine activation of c-Jun NH2-terminal kinases. J. Biol. Chem. 271: 15703–15707.
Tenneti L, D’Emilia DM, Troy CM, Lipton SA. (1998) Role of caspases in N-methyl-D-aspartate-induced apoptosis in cerebrocortical neurons. J. Neurochem. 71: 946–959.
Jornot L, Petersen H, Junod AF. (1997) Modulation of the DNA binding activity of transcription factors CREP, NFkappaB and HSF by H2O2 and TNF alpha. Differences between in vivo and in vitro effects. FEBS Lett. 416: 381–386.
Bowie AG, Moynagh PN, O’Neill LAJ. (1997) Lipid peroxidation is involved in the activation of NF-kappaB by tumor necrosis factor but not interleukin-1 in the human endothelial cell line ECV304. Lack of involvement of H2O2 in NF-kappaB activation by either cytokine in both primary and transformed endothelial cells. J. Biol. Chem. 272: 25941–25950.
Brennan P, O’Neill LA. (1995) Effects of oxidants and antioxidants on nuclear factor kappa B activation in three different cell lines: evidence against a universal hypothesis involving oxygen radicals. Biochim. Biophys. Acta 1260: 167–175.
Israel N, Gougerot-Pocidalo MA, Aillet F, Virelizier JL. (1992) Redox status of cells influences constitutive or induced NF-kappa B translocation and HIV long terminal repeat activity in human T and monocytic cell lines. J. Immunol 149: 3386–3393.
Beg AA, Baltimore D. (1996) An essential role for NF-B in preventing TNF-induced cell death. Science 274: 782–784.
Wang C-Y, Mayo MW, Baldwin AS, Jr. (1996) TNF- and cancer therapy-induced apoptosis: potentiation by inhibition of NF-B. Science 274: 784–787.
Van Antwerp DJ, Seamus J, Martin SJ, Kafri T, Green DR, Verma IM. (1996) Suppression of TNF-induced apoptosis by NF-B. Science 274: 787–789.
Baichwal VR, Baeuerle PA. (1997) Activate NF-kappa B or die? Curr. Biol. 7: R94–96.
Lezoualc’h F, Sagara Y, Holsboer F, Behl C. (1998) High constitutive NF-kappaB activity mediates resistance to oxidative stress in neuronal cells. J. Neurosci. 18: 3224–3232.
Lezoualc’h F, Engert S, Berning B, Behl C. (2000) Corticotropin-releasing hormone-mediated neuroprotection against oxidative stress is associated with the increased release of non-amyloidogenic amyloid beta precursor protein and with the suppression of nuclear factor-kappaB. Mol. Endocrinol. 14: 147–159.
Kaltschmidt B, Uherek M, Wellmann H, Volk B, Kaltschmidt C. (1999) Inhibition of NF-kappaB potentiates amyloid beta-mediated neuronal apoptosis. Proc. Natl. Acad. Sci. USA 96: 9409–9414.
Heck S, Lezoualc’h F, Engert S, Behl C. (1999) Insulin-like growth factor-1-mediated neuroprotection against oxidative stress is associated with activation of nuclear factor kappaB. J. Biol. Chem. 274: 9828–9835.
Barger SW, Horster D, Furukawa K, Goodman Y, Krieglstein J, Mattson MP. (1995) Tumor necrosis factors alpha and beta protect neurons against amyloid beta-peptide toxicity: evidence for involvement of a kappa B-binding factor and attenuation of peroxide and Ca2+ accumulation. Proc. Natl. Acad. Sci. USA 92: 9328–9332.
Labarca C, Paigen K. (1980) A simple, rapid, and sensitive DNA assay procedure. Anal. Biochem. 102: 344–352.
Lennon SV, Martin SJ, Cotter TG. (1991) Dose-dependent induction of apoptosis in human tumour cell lines by widely diverging stimuli. Cell Prolif. 24: 203–214.
Dypbukt JM, Ankarcrona M, Burkitt M, Sjoholm A, Strom K, Orrenius S, Nicotera P. (1994) Different prooxidant levels stimulate growth, trigger apoptosis, or produce necrosis of insulin-secreting RINm5F cells. The role of intracellular polyamines. J. Biol. Chem. 269: 30553–30560.
Hampton MB, Orrenius S. (1997) Dual regulation of caspase activity by hydrogen peroxide: implications for apoptosis. FEBS Lett. 414: 552–556.
Shrivastava A, Aggarwal BB. (1999) Antioxidants differentially regulate activation of nuclear Factor-kB, activator protein-1, c-jun aminoterminal kinases, and apoptosis induced by tumor necrosis factor: evidence that JNK and NF-κB activation are not linked to apoptosis. Antioxidants Redox Signal. 1: 181–191.
Matthews JR, Hay RT. (1995) Regulation of the DNA binding activity of NF-kappa B. Int. J. Biochem. Cell Biol. 27: 865–879.
Sen CK, Packer L. (1996) Antioxidant and redox regulation of gene transcription. FASEB J. 10: 709–720.
Flohe L, Brigelius-Flohe R, Saliou C, Traber MG, Packer L. (1997) Redox regulation of NF-kappa B activation. Free Radic. Biol. Med. 22: 1115–1126.
Piette J, Piret B, Bonizzi G, Schoonbroodt S, Merville MP, Legrand-Poels S, Bours V. (1997) Multiple redox regulation in NF-kappaB transcription factor activation. Biol. Chem. 378: 1237–1245.
Sun SC, Ganchi PA, Ballard DW, Greene WC. (1993) NF-kappa B controls expression of inhibitor I kappa B alpha: evidence for an inducible autoregulatory pathway. Science 259: 1912–1915.
Kaltschmidt B, Uherek M, Volk B, Baeuerle PA, Kaltschmidt C. (1997) Transcription factor NF-kappaB is activated in primary neurons by amyloid beta peptides and in neurons surrounding early plaques from patients with Alzheimer disease. Proc. Natl. Acad. Sci. USA 94: 2642–2647.
Singh I, Pahan K, Khan M, Singh AK. (1998) Cytokine-mediated induction of ceramide production is redox-sensitive. Implications to pro-inflammatory cytokine-mediated apoptosis in demyelinating diseases. J. Biol. Chem. 273: 20354–20362.
Kaltschmidt C, Kaltschmidt B, Neumann H, Wekerle H, Baeuerle PA. (1994) Constitutive NF-kappa B activity in neurons. Mol. Cell Biol. 14: 3981–3992.
Furukawa K, Mattson MP. (1998) The transcription factor NF-kappaB mediates increases in calcium currents and decreases in NMDA- and AMPA/kainate-induced currents induced by tumor necrosis factor-alpha in hippocampal neurons. J. Neurochem. 70: 1876–1886.
Yu Z, Zhou D, Bruce-Keller AJ, Kindy MS, Mattson MP. (1999) Lack of the p50 subunit of nuclear factor-kappaB increases the vulnerability of hippocampal neurons to excitotoxic injury. J. Neurosci. 19: 8856–8865.
Botchkina GI, Geimonen E, Bilof ML, Villarreal O, Tracey KJ. (1999) Loss of NF-kappaB activity during cerebral ischemia and TNF cytotoxicity. Mol. Med. 5: 372–381.
Taglialatela G, Kaufmann JA, Trevino A, Perez-Polo JR. (1998) Central nervous system DNA fragmentation induced by the inhibition of nuclear factor kappa B. Neuroreport 9: 489–493.
Sullivan PG, Bruce-Keller AJ, Rabchevsky AG, Christakos S, Clair DK, Mattson MP, Scheff SW. (1999) Exacerbation of damage and altered NF-kappaB activation in mice lacking tumor necrosis factor receptors after traumatic brain injury. J. Neurosci. 19: 6248–6256.
Schneider A, Martin-Villalba A, Weih F, Vogel J, Wirth T, Schwaninger M. (1999) NF-kappaB is activated and promotes cell death in focal cerebral ischemia. Nat. Med. 5: 554–559.
Yang K, Mu XS, Hayes RL. (1995) Increased cortical nuclear factor-kappa B (NF-kappa B) DNA binding activity after traumatic brain injury in rats. Neurosci. Lett. 197: 101–104.
Hickenbottom SL, Grotta JC, Strong R, Denner LA, Aronowski J. (1999) Nuclear factor-kappaB and cell death after experimental intracerebral hemorrhage in rats. Stroke 30: 2472–2477 (discussion, 2477-8).
Carroll JE, Howard EF, Hess DC, Wakade CG, Chen Q, Cheng C. (1998) Nuclear factor-kappa B activation during cerebral reperfusion: effect of attenuation with N-acetylcysteine treatment. Brain Res. Mol. Brain Res. 56: 186–191.
Clemens JA, Stephenson DT, Yin T, Smalstig EB, Panetta JA, Little SP. (1998) Drug-induced neuroprotection from global ischemia is associated with prevention of persistent but not transient activation of nuclear factor-kappaB in rats. Stroke 29: 677–682.
Manna SK, Zhang HJ, Yan T, Oberley LW, Aggarwal BB. (1998) Overexpression of manganese superoxide dismutase suppresses tumor necrosis factor-induced apoptosis and activation of nuclear transcription factor-B and activated protein-1. J. Biol. Chem. 273: 13245–13254.
The authors thank Joliet Bembry for preparation of endothelial cells, and Dace Klimanis and Christl Reutzler for assistance with Western blots and TUNEL staining.