Nitric Oxide Disrupts H2O2-dependent Activation of Nuclear Factor κB

Tumor necrosis factor α (TNF-α) exerts its effect by two distinct signaling pathways. It can trigger cytotoxicity in sensitive target cells. TNF-α can also promote nuclear factor κB (NF-κB) activity and regulate the expression of genes that interfere with apoptosis and thus conferring resistance to several apoptotic stimuli. We have observed that interferon-γ (IFN-γ) sensitizes human ovarian carcinoma cell lines to TNF-α-mediated apoptosis and further, IFN-γ induces the expression of the inducible nitric-oxide synthase (iNOS) and the generation of nitric oxide (NO). This study examines the role of NO in the sensitization of the ovarian carcinoma cell line AD10 to TNF-α-mediated cytotoxicity. Treatment of AD10 cells with the NOS inhibitor l-NMA blocked the IFN-γ-dependent sensitization whereas NO donors (S-nitroso-N-acetylpenicillamine) sensitized these cells to TNF-α cytotoxicity. Analysis of the activation status of NF-κB upon treatment with NO donors confirmed the inhibitory role of NO on both the NF-κB DNA-binding property and its activation. Moreover, the inhibition of NF-κB nuclear translocation by NO donors directly correlated with the intracellular concentration of H2O2 and was reversed by the addition of exogenous H2O2. These findings show that NO might interfere with TNF-α-dependent NF-κB activation by interacting with O⨪2 and reducing the generation of H2O2, a potent NF-κB activator. Therefore, NO-mediated disruption of NF-κB activation results in the removal of anti-apoptotic/resistance signals and sensitizes tumor cells to cytotoxic cytokines like TNF-α.

Tumor necrosis factor ␣ (TNF-␣) exerts its effect by two distinct signaling pathways. It can trigger cytotoxicity in sensitive target cells. TNF-␣ can also promote nuclear factor B (NF-B) activity and regulate the expression of genes that interfere with apoptosis and thus conferring resistance to several apoptotic stimuli. The development of resistance to either the immune system or chemo-immunotherapeutic strategies remains a disadvantage in the therapy of cancer, particularly in cases where recurrences and/or relapses occurred. Apoptosis has been accepted as a distinct pathological mechanism in tumors responding to anticancer therapies. Further, resistance to apoptosis in tumor cells has been recognized as a common pathway to multiple drug resistance (1,2). Multiple lines of evidence have implicated the activation of the transcription factor NF-B 1 as one of the primary signals in the onset of resistance to many apoptotic stimuli, particularly TNF-␣ (3)(4)(5).

We have observed that interferon-␥ (IFN-␥) sensitizes human ovarian carcinoma cell lines to TNF-␣-mediated apoptosis and further, IFN-␥ induces the expression of the inducible nitric-oxide synthase (iNOS) and the generation of nitric oxide (NO). This study examines the role of NO in the sensitization of the ovarian carcinoma cell line AD10 to TNF-␣-mediated cytotoxicity. Treatment of AD10 cells with the NOS inhibitor L-NMA blocked the IFN-␥-dependent sensitization whereas NO donors (S-nitroso-N-acetylpenicillamine) sensitized these cells to TNF-␣ cytotoxicity. Analysis of the activation status of NF-B upon treatment with NO donors
TNF-␣ is a proinflammatory cytokine that exerts a broad spectrum of biological effects by its interaction with two distinct cell surface receptors, TNFR1 and TNFR2 (6). Most cytotoxic effects of TNF-␣ are mediated by the TNFR1. It has been demonstrated that, upon interaction with TNF-␣, trimerization of TNFR1 takes place and results in cellular signaling leading to the recruitment of the TNFR1-associated death domain protein and the receptor-interacting protein to the receptor complex (7). The TNFR1-associated death domain protein interacts with the Fas-associated death domain to initiate the death pathway and engages several proteins such as the TNFR-associated factor-1, the TNFR-associated factor-2, and receptorinteracting protein to initiate the TNF signaling pathways such as the activation of NF-B (8).
Reactive oxygen species (ROS) have also been implicated in the signaling pathways initiated by TNF-␣. Stimulation of mammalian cells with TNF-␣ triggers the generation of various ROS (9,10). Hence, the use of antioxidants results in the inhibition of various TNF-␣-related effects such as the activation of transcription factors, gene expression, and cytotoxicity. In addition, the use of exogenous ROS mimics the biological activity of TNF-␣ (11). These data support the hypothesis that ROS function as second messengers for TNF-␣-mediated signaling. In biological systems the most important ROS generated upon TNF-␣ stimulation are the result of enzymatic partial reduction of oxygen yielding superoxide (O 2 . ), which is either immediately reduced by superoxide dismutase to hydrogen peroxide (H 2 O 2 ) or alternatively reacts rapidly with nitric oxide (NO) to generate ONOO Ϫ (12)(13)(14). However, the regulatory role of NO in TNF-␣ signaling via the disruption of ROSdependent activation of NF-B has not been established. Several lines of evidence showed that resistant tumors could be sensitized to TNF-␣-mediated cytotoxicity by various cytokines or pharmacological treatments (15)(16)(17)(18)(19)(20). Recently, we have reported that IFN-␥ induced the sensitization of the human ovarian carcinoma AD10 cell line to Fas-mediated apo ptosis and the sensitization was due in part to the generation of nitric oxide by the induction of iNOS in these cells (21). NO has been identified as a potential second messenger based on its ability to chemically interact with a broad range of regulatory proteins. Furthermore, NO can interact with metal cluster-and thiol-containing proteins (for review, see Ref. 22) resulting in the modification of both the structures and functions of these proteins. Although NO has been shown to react very rapidly with O 2 . , the only biological effect to this chemical reaction has been assigned to the generation of ONOO Ϫ , a proposed cytotoxic derivative (23,24). Herein, we hypothesize that NO is interfering with the TNF-␣-mediated signaling by chemically reacting with O 2 . . Since can serve as a precursor to H 2 O 2 , which is a proposed activator of the anti-apoptotic transcription factor NF-B, the reaction of O 2 . with NO will interfere with the activation of NF-B and will result in the removal of anti-apoptotic signals and sensitization of the tumor cells to TNF-␣ cytotoxicity. This study has been designed to test this hypothesis, and the following have been examined: (a) the molecular mechanism by which IFN-␥ sensitizes the human ovarian carcinoma cell line to TNF-␣-induced cytotoxicity, (b) the specific role of NO in the disruption of TNF-␣-mediated generation of H 2 O 2 , and, subsequently, (c) the mechanism by which NO can disrupt the TNF-␣-dependent NF-B activation.

EXPERIMENTAL PROCEDURES
Cell Culture and Reagents-The AD10 cell line is an adriamycinresistant, MDR phenotype-expressing subline derived from the human ovarian carcinoma cell line A2780 and was obtained from Dr. Ozols (Fox Chase Cancer Center, Philadelphia, PA). The PC-3 cell line is a metastatic bone-derived human prostatic adenocarcinoma, CRL-1435, obtained from ATCC (American Type Culture Collection, Manassas, VA). Cell cultures were maintained as monolayers on plastic dishes in RPMI 1640 medium (MediaTech, Inc., Herndon, VA), supplemented with 10% heat-inactivated FBS (Gemini Bio-Products, Inc., Calabasas, CA), 1% L-glutamine (Life Technologies, Inc.), 1% pyruvate (Life Technologies, Inc.), 1% nonessential amino acids (Life Technologies, Inc.), and incubated at 37°C and 5% CO 2 . For every experimental condition, the cells were cultured in 1% FBS 24 h prior to treatments. In cases where SNAP (kindly provided and synthesized by Dr. Jon Fukuto, UCLA, Los Angeles, CA) was used, 500 M photo-activated SNAP was added to the cultured cells 2 h prior to stimulation with cytokines unless otherwise indicated in the text. For iNOS induction, cultured cells were stimulated 18 h with 100 units of human recombinant IFN-␥ (PeproTech, Inc., Rocky Hill, NJ). For guanylate cyclase-related effects, cells were incubated in the presence of the cGMP analogue 8-bromo-cGMP instead of SNAP or blocked using 300 M 1H-(1,2,4)oxadiazolo-[4,3-a]quinoxalin-1-one (Alexis Corp., San Diego, CA).
Cytotoxicity Assay-TNF-␣-mediated cytotoxicity was assessed using recombinant TNF-␣ at the concentrations of 0.01, 0.1, and 1 ng/ml in a 24-h incubation assay. The lactate dehydrogenase (LDH)-based Cyto-Tox 96™ assay (Promega, Madison, WI) was used to determine cytotoxicity (25). Briefly, 1 ϫ 10 4 cells/sample, in quadruplicate, were distributed into a 96-well flat-bottom microtiter plate (Costar, Cambridge, MA) and cultured at a low serum concentration (0.1% FBS) 18 h prior to each treatment. After incubation for each different experimental condition, released LDH into the culture supernatants was measured with a 30-min coupled enzymatic assay, which results in the conversion of a tetrazolium salt (INT) into a red formazan product that is read at 490 nm in an automated plate reader (Emax, Molecular Devices, Sunnyvale, CA). Percentage cytotoxicity was calculated using the spontaneous release-corrected OD as follows: % cytotoxicity ϭ (OD of experimental well/OD of maximum release control well) ϫ 100.
Reverse Transcription-Polymerase Chain Reaction (RT-PCR)-Total RNA was extracted and purified from ϳ1 ϫ 10 6 cells for each experimental condition by a single-step monophasic solution of phenol and guanidine isothiocyanate-chloroform using Trizol ® reagent (Life Technologies, Inc.). 1 g of total RNA was reverse-transcribed to first strand cDNA for 1 h at 42°C with 200 units of SuperScript™ II reverse transcriptase and 20 M random hexamer primers (Life Technologies, Inc.). Amplification of 1/10 of these cDNA by PCR was performed using the following gene-specific primers: TNF-␣ (forward) (5Ј-AAG CCT GTA GCC CAT GTT GTA GC-3Ј) and TNF-␣ (reverse) (5Ј-GAA GAC CCC TCC CAG ATA GAT G-3Ј) (342-base pair expected product). Internal control for equal cDNA loading in each reaction was assessed using the following gene-specific glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primers: GAPDH sense (5Ј-GAA CAT CAT CCC TGC CTC TAC TG-3Ј), GAPDH antisense (5Ј-GTT GCT GTA GCC AAA TTC GTT G-3Ј) (355-base pair expected product). PCR amplifications of each specific DNA sequence were carried out using the "Hot Start" method using Platinum Taq™ polymerase (Life Technologies, Inc.) followed by a two-step thermal cycling incubations (95°C/15 s; 60°C/30 s for 30 cycles and a final extension at 72°C/10 min). The numbers of PCR cycles were established based on preliminary titration of the relative amount of amplified product for each gene representing the linear phase of the amplification process. The amplified products were resolved on 1.5% agarose gel electrophoresis, and their relative concentrations were assessed by densitometric analysis of the digitized ethidium bromide (EtBr)-stained image, performed on a Macintosh computer (Apple Computer Inc., Cupertino, CA.) using the public domain NIH Image program (developed at the United States National Institutes of Health and available on the Internet).
Nuclear Extract Preparation-1 ϫ 10 6 cultured cells treated under different experimental conditions were washed twice with ice-cold Dulbecco's phosphate-buffered saline (MediaTech, Inc., Herndon, VA). P-40 lysis buffer (10 mM Tris-HCl, pH 7.4, 10 mM NaCl, 3 mM MgCl 2 , 0.5% Nonidet P-40, 0.1 mM EDTA) was added to the top of the washed cells and incubated on ice for 5 min. Lysed cells were collected by gentle pipetting three to four times and transfered to a microcentrifuge tube. Nuclear pellets for each experimental condition were generated by two consecutive centrifugation and washing steps at 1200 rpm. Nuclear pellets were lysed in buffer C (20 mM HEPES, pH 7.9, 25% glycerol, 0.42 M NaCl, 1.5 mM MgCl 2 , 0.1 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, and 0.5 mM dithiothreitol). Total nuclear protein concentrations were determined using the method of Bradford (26).
Electrophoretic Mobility Shift Assay (EMSA)-Nuclear protein extracts (2 g) were assayed for DNA interaction by EMSA as described previously with modifications (27). The double-stranded NF-B consensus binding sequence (5Ј-AGT TGA GGG GAC TTT CCC AGG C-3Ј) oligonucleotide was radiolabeled with [␥-32 P]ATP (ICN Pharmaceuticals, Inc. Costa Mesa, CA.) by incubation with 10 units of T4 polynucleotide kinase (New England Biolabs, Beverly, MA.) and further purified using QIAquick nucleotide removal kit (Qiagen, Valencia, CA.). After the DNA-binding reaction, the samples were resolved on 4 -15% Tris-HCl-polyacrylamide minigels (Bio-Rad) and the gels were dried and autoradiographed. Specificity of the DNA-binding reaction was determined by competition assays performed with 100-fold excess of unlabeled NF-B or unrelated oligonucleotide (i.e. AP-1: 5Ј-GAT CGA ACT GAC CGC CCG CGG CCC GT-3Ј). The relative concentrations of specific NF-B shifted bands were assessed by densitometric analysis of the digitized autoradiographic images using the NIH Image program described above.
Determination of Intracellular H 2 O 2 Generation-1 ϫ 10 6 cells were cultured in a six-well plate for 18 h in culture medium supplemented with 1% FBS. In some instances, the minimal serum-cultured cells were treated with 500 M photo-activated SNAP 2 h prior to stimulation with 10 or 100 units/ml TNF-␣. Intracellular H 2 O 2 levels were evaluated using the fluorescent cell permeable probe, 2Ј,7Ј-dichlorofluorescein diacetate (H 2 DCFDA) (Molecular Probes, Inc., Eugene, OR). Then, the culture medium was replaced with Dulbecco's phosphate-buffered saline, pH 7.4, containing 5 M H 2 DCFDA. Fluorescence intensity was analyzed on an EPICS ® XL-MCL flow cytometer (Beckman Coulter Inc., Fullerton, CA).
Transfections and Reporter Gene System-The intracellular activation of NF-B was determined by transient transfection of AD10 cells with the pNF-B-d2EGFP reporter vector (CLONTECH, Palo Alto, CA). 7 ϫ 10 6 cultured cells were transfected with 10 g of DNA using 60 l of Lipofectamine reagent (Life Technologies, Inc.) according to the manufacturer's recommendations. Transfected cells were then distributed onto a six-well culture plate and incubated under different experimental conditions. The relative fluorescence intensity was analyzed on an EPICS ® XL-MCL flow cytometer.
Statistical Analysis-The experimental values were expressed as the means Ϯ standard error of the mean (S.E.) for the number of separate experiments indicated in each case. One-way analysis of variance was used to compare variances within groups and among them. Bartlett's tests were used to establish the homogeneity of variance on the basis of the differences among standard deviations (S.D.). Whenever needed, post hoc unpaired multiple comparison tests (Bonferroni's test) and Student's t test were used for comparison between two groups. Significant differences were considered for those probabilities Ͻ 5% (p Ͻ 0.05).

IFN-␥-mediated Sensitization of the Human Ovarian
Carcinoma AD10 to TNF-␣-induced Cytotoxicity Is Due, in Part, to the Generation of Nitric Oxide-To investigate the role of nitric oxide on the sensitization of the human ovarian carcinoma AD10 cell line to TNF-␣-mediated cytotoxicity, we first stimulated quiescent AD10 cells with IFN-␥ in the presence or absence of 1 mM potent NOS inhibitor L-NMA. The sensitivity of AD10 cells to the cytotoxic effect of increasing concentrations of TNF-␣ (0.01, 0.1, and 1 ng/ml) was evaluated by the release of LDH into the culture medium after 24 h of incubation. Exposure of AD10 cells to IFN-␥ (100 units/ml) for 18 h sensitized the tumor cells to TNF-␣-mediated cytotoxicity and the degree of sensitization increased with increasing concentrations of TNF-␣. Sensitization by IFN-␥ was significantly decreased in the presence of 1 mM NOS inhibitor L-NMA (Fig. 1A).
To confirm the specific role of nitric oxide in the sensitization of AD10 cells, we assessed the cytotoxic effect of TNF-␣ in the presence of 10 and 100 M nitric oxide donor SNAP. After incubation for 18 h with different concentrations of SNAP, we observed a significant increase in the sensitivity of AD10 cells to TNF-␣-mediated cytotoxicity in a 24-h assay that directly correlated with the concentrations of SNAP (Fig. 1B).

Nitric Oxide and Pyrrolidine Dithiocarbamate (PDTC) Inhibit TNF-␣-induced Expression of Endogenous TNF-␣ in AD10
Cells-The transcription factor NF-B has been demonstrated to tightly regulate the gene expression of TNF-␣, establishing a self-regulatory loop in tumor cells that secrete TNF-␣ that in turn activates NF-B (28). Furthermore, PDTC has been shown to inhibit TNF-␣-mediated activation of NF-B in several cell types and in macrophages (29). To demonstrate the specific effect of nitric oxide on the NF-B-mediated expression of TNF-␣, we incubated AD10 cells with 1, 10, 100, and 500 M SNAP for 18 h and then stimulated the cells with 100 units/ml TNF-␣ for 4 h. The relative levels of endogenously generated TNF-␣ were assessed by amplification of the specific TNF-␣ cDNA using RT-PCR. The constitutive expression of TNF-␣ by AD10 cells was demonstrated and a significant increased level was observed upon treatment with exogenous TNF-␣. Moreover, this increased level of TNF-␣ was blocked following treatment of the cells with SNAP (500 M nitric oxide donor) up to the complete disappearance of the amplified TNF-␣ mRNA ( Fig. 2A). These findings suggest that NO inhibits NF-B and consequently down-regulates TNF-␣ mRNA expression. Similar results to those observed with AD10 cells were obtained with the human prostatic adenocarcinoma cell line PC-3. The expression of TNF-␣ messenger RNA in PC-3 was decreased approximately 4 -5-fold upon treatment with 500 M SNAP, suggesting the role of nitric oxide in the NF-B-dependent expression of TNF-␣.
To confirm the control of NF-B on TNF-␣ expression, we examined the relative levels of expression of endogenous TNF-␣ mRNA after treatment of AD10 cells with 1, 10, 100, and 500 M PDTC for 18 h followed with TNF-␣ (100 units/ml) stimulation for 4 h. Endogenous TNF-␣ gene expression of TNF-␣-stimulated cells decreased in the presence of PDTC but was never completely blocked as was observed above following treatment with SNAP (Fig. 2B). These results confirm the role of ROS in the activation of the transcription factor NF-B and the subsequent expression of TNF-␣.
Nitric Oxide Disrupts the H 2 O 2 -dependent Activation of NF-B in AD10 Cells-To determine whether nitric oxide could interfere with the TNF-␣-mediated activation of NF-B, we examined the NF-B DNA-binding activity by EMSA. As shown in Fig. 3, nuclear extracts from TNF-␣-stimulated AD10 cells exhibited an increased binding activity specific for the NF-B heterodimer p65-p50. H 2 O 2 also induced specific NF-B binding activity in AD10 cells after 30 min of incubation. Further, NF-B binding activity was significantly inhibited by the incubation of AD10 cells with 500 M SNAP for 2 h prior to stimulation with TNF-␣ for 30 min. The impaired NF-B binding activity by SNAP was restored by the addition of H 2 O 2 to similar levels as those detected in the H 2 O 2 -stimulated AD10 cells. Thus, these results suggest that the step at which nitric oxide interferes preceded the step at which H 2 O 2 is generated after stimulation of AD10 cells with TNF-␣.
Noteworthy, AD10 cells exhibit a constitutive level of NF-B binding activity that is not affected by nitric oxide (Fig. 3, lanes  1 and 10), whereas in TNF-␣-stimulated cells the NF-B binding activity decreased below the basal levels in the presence of nitric oxide (Fig. 3, lane 6).
Nitric Oxide Decreases TNF-␣-dependent Generation of H 2 O 2 -To examine whether nitric oxide affects the generation of H 2 O 2 in AD10 cells stimulated with TNF-␣, we determined the intracellular generation of H 2 O 2 using the fluorescent cellpermeable probe, H 2 DCFDA. AD10 cells were incubated in the presence or absence of 500 M SNAP and then stimulated with 10 and 100 units/ml TNF-␣, respectively, for 15 min. Fluorescence cytometric analysis of these experimental groups revealed a significant increase in H 2 O 2 levels generated by the TNF-␣ treatment. Incubation of the TNF-␣-stimulated cells in the presence of SNAP significantly reduced the relative amount of H 2 O 2 generated by these cells (Fig. 4). These data suggest that nitric oxide is affecting the intracellular biogen-eration of H 2 O 2 by superoxide dismutase via its chemical interaction with TNF-␣-induced O 2 . .

Exogenous H 2 O 2 Restored the Nitric Oxide-mediated Blocking of the TNF-␣-dependent Activation of NF-B-Nitric oxide
has been shown to directly affect the structure of NF-B and decrease its DNA-binding ability due to thiol modification of critical amino acid residues (30). To determine the direct effect of nitric oxide on the activation of NF-kB, we used an enhanced green fluorescent protein-based reporter system driven by four tandem-repeated B responsive elements linked to the thymidine kinase minimal promoter (pNF-B-d2EGFP). We transiently transfected AD10 cells with the pNF-B-d2EGFP reporter vector and then stimulated the cells in the presence or absence of 500 M SNAP. Cytofluorometric analysis of these cells revealed a significant activation of the reporter gene by TNF-␣ and H 2 O 2 , and the extent of activation was a function of the concentrations used. The TNF-␣-induced activation of the NF-B-dependent reporter gene was significantly decreased in the presence of 500 M SNAP (Fig. 5), corroborating the findings obtained in the NF-B binding assay. The inhibitory activity of SNAP on the TNF-␣-induced activation of the NF-Bdependent reporter gene was significantly rescued by stimulation with 200 M exogenous H 2 O 2 (Fig. 5). These data confirm the inhibitory effect of nitric oxide on the H 2 O 2 -dependent activation of NF-B in TNF-␣-treated AD10 cells. We also noticed that untreated AD10 cells were able to maintain basal levels of NF-B activation that were not inhibited by treatment with nitric oxide, corroborating the findings observed in the binding assay in Fig. 3.

DISCUSSION
The activation of the transcription factor NF-B by TNF-␣ and many other stimuli has been implicated in the development of resistance of tumor cells to a variety of cytotoxic molecules including TNF-␣ (3,5). NF-B is an oxidative stressresponsive transcription factor that has been shown to respond to small concentrations of exogenous H 2 O 2 or to reactive oxygen species endogenously generated as part of the signaling cascade triggered by many molecules such as TNF-␣ (31-33). We have reported that the IFN-␥-induced sensitization of the human ovarian carcinoma AD10 cell line to Fas-mediated apoptosis is due in part to the generation of nitric oxide, or its reaction products, by iNOS in these cells (21). In the present study, evidence is presented for the first time that demonstrates that NO also sensitizes tumor cells to TNF-␣-mediated cytotoxicity. Further, we describe a novel molecular mechanism by which nitric oxide disrupts the H 2 O 2 -dependent activation of NF-B resulting in sensitization of the AD10 cells to TNF-␣ cytotoxicity.
The specific role of nitric oxide in tumor biology is not estab- lished. A broad spectrum of activities has been assigned to either the physiology or the pathophysiology of nitric oxide in tumor cells (for a review, see Ref. 34). Low output of nitric oxide has been correlated with increased blood flow and new blood vessels feeding the tumor area (35). In addition, the generation of nitric oxide by tumor cells may inhibit the activation and proliferation or may increase the apoptosis of surrounding lymphocytes that can account for the immune suppression accompanying tumor growth (data not shown). Furthermore, high intratumoral output of nitric oxide could inhibit the activation of caspases and therefore antagonizes the pro-apoptotic signals (36,37). However, the opposite effect has also been observed in many other systems whereby the generation of high output of nitric oxide, either by iNOS induction or by the use of NO donors, inhibits tumor growth and metastasis (38). Therefore, the final outcome of NO-mediated effects would be determined by many factors including the local concentration and sources of nitric oxide and the presence of reactive molecules that might redirect the redox status in the tumor cell.
In the human ovarian carcinoma AD10 cell line stimulated with the pro-inflammatory cytokine IFN-␥, we observed a markedly increased sensitivity of these tumor cells to the cytotoxic effect of TNF-␣. IFN-␥ also induces iNOS expression in these cells (21). Sensitization to TNF-␣ was antagonized by the use of the specific NOS inhibitor L-NMA and was mimicked by the use of the NO donor SNAP, confirming the role of nitric oxide in the sensitization process (Fig. 1). Frequently, IFN-␥ treatment alone might not be sufficient to induce iNOS expression in cultured cells. The participation of IFN-␥ in the induction of iNOS is generally directed to the potentiation of the activity of pro-inflammatory cytokines like TNF-␣, interleukin-1, or bacterial lipopolysaccharide. These cytokines and/or the bacterial lipopolysaccharide have been shown to activate the transcription factor NF-B, setting the basal threshold for the induction of the expression of iNOS that might be enhanced by the action of IFN-␥ (39). We observed that untreated AD10 cells (which constitutively secrete TNF-␣) display a constitutive level of activation of NF-B (Fig. 3 and 5). Therefore, the basal activation of NF-B in AD10 cells could explain why the treatment with IFN-␥ alone was sufficient to induce iNOS and subsequently generate nitric oxide.
NF-B has been shown to be a key transcription factor con-trolling TNF-␣ gene expression in many cells, either as a major activator or synergistically in association with other transcription factors (28). Thus, the significant basal activation of the NF-B in AD10 cells might explain the presence of a constitutive expression of TNF-␣ by these cells (Fig. 2, A and B, last  lanes). Moreover, TNF-␣ has been implicated as a survival cytokine used by tumor cells either to control anti-apoptotic mechanisms or promoting cellular proliferation (40 -42). Therefore, the maintenance of a self-regulated loop in which the expression of TNF-␣ is perpetuated by the TNF-␣-mediated basal activation of NF-B could play a major role in the survival and/or proliferation of tumor cells. PDTC has been shown to be a potent and specific inhibitor of the NF-B-mediated expression of TNF-␣ (29,43). Untreated AD10 cells exhibited a basal expression of TNF-␣, which was enhanced by stimulation with exogenous TNF-␣ and subsequently inhibited by PDTC (Fig. 2B). Similarly, using the nitric oxide donor SNAP, we were able to completely abrogate the expression of endogenous TNF-␣ ( Fig. 2A). In contrast, nitric oxide was unable to block the basal expression of endogenous TNF-␣ in the absence of exogenous stimulation. These results strongly suggest the inhibitory role of nitric oxide on TNF-␣-induced activation of NF-B and consequently resulting in the disruption of TNF-␣ gene expression. TNF-␣ induces the generation of ROS that may serve as second messengers in the activation of divergent pathways related to the cell death processes (44 -46). Stimulation of many cell types with TNF-␣ results in the generation of intracellular superoxide (O 2 . ) (10 (48,49). Further, we have found that the addition of NO donors to TNF-␣-stimulated AD10 cells inhibited either the DNA binding activity of NF-B (Fig. 3) or its activation (Fig. 5). This inhibition was restored to the normal H 2 O 2 -stimulated level by treatment with exogenous H 2 O 2 . In contrast, nitric oxide did not affect the NF-B activation in untreated AD10 cells, confirming the previous observation with TNF-␣ gene expression. These results suggest the presence of at least two pathways in the activation of NF-B in AD10 cells that may differ in their sensitivity to H 2 O 2 and the selectivity of nitric oxide to affect just one of these two pathways. The inactivation of NF-B upon NO treatment was not mediated by guanylate cyclase activation since the cGMP analogue 8-bromo-cGMP had no effect on NF-B and we could not block the inhibitory effect of NO on NF-B activation by the use of the guanylate cyclase blocker 1H-(1,2,4)oxadiazolo-[4,3-a]quinoxalin-1-one (data not shown).
Previous reports have implicated the role of nitric oxide on the activation of NF-B. NO has been shown to increase the expression of the NF-B inhibitory subunit IB or affects its cellular stability by inhibiting protein degradation (50). Due to the rapid generation of H 2 O 2 upon TNF-␣ treatment and the immediate activation NF-B (less than 15 min) in AD10 cells, it is highly unlikely that secondary regulatory factors like the induction of IB interfered with the rapid NF-B activation. It is likely that, in the long run, a combination of both mechanisms may account for the total inhibitory role of nitric oxide on the TNF-␣-induced activation of NF-B. An alternative proposed mechanism implicated in the inhibition of the NF-B activity by NO is via the alteration of critical thiol groups, resulting in the disruption of the NF-B structure and subsequently affecting its DNA-binding ability (30). However, the in vivo situation may be much more complex due to the high concentrations of glutathione and other redoxactive proteins within the cell, which may prevent the modification of thiol groups.
In conclusion, our findings suggest that the mechanism by which NO sensitizes the human ovarian carcinoma cell line to TNF-␣-mediated apoptosis is due to the specific disruption of the TNF-␣-induced generation of H 2 O 2 and the subsequent inhibition of the NF-B-dependent expression of anti-apoptotic genes. These results can be extended to other solid tumor cells, as observed with the human prostatic adenocarcinoma cell line PC-3. As shown in Fig. 6, the survival autocrine-paracrine loop involving the NF-B-dependent expression of TNF-␣ could be interrupted by the inhibitory activity that nitric oxide exerts on the TNF-␣-induced activation of NF-B. Furthermore, in an in vivo situation, the exposure of tumor cells to pro-inflammatory cytokines such as IFN-␥ will promote the induction of iNOS by the tumor cells or neighboring lymphocytes and which in turn will result in the generation of nitric oxide. Hence, the endogenously generated or the exogenously provided NO would scavenge the TNF-␣-generated O 2 . and decrease the H 2 O 2 -dependent activation of NF-B. Based on these molecular events, a new mechanism of NO-mediated sensitization to apoptosis is revealed.