Thiopental Inhibits Tumor Necrosis Factor α–induced Activation of Nuclear Factor κB through Suppression of IκB Kinase Activity

The following anesthetics and substances were used: GABA (Sigma, Deisenhofen, Germany), bicuculline-methochloride (Tocris Cookson Inc., Ellisville, MO), dichlorophenyl-methyl-amino-propyl-diethoxymethyl-phosphinic acid (CGP 52432; Tocris Cookson Inc.), pentobarbital, thiamylal (Surital; Pharmacia, Erlangen, Germany), secobarbital (Sigma), and thiopental (Byk-Gulden, Konstanz, Germany). Recombinant human TNF-α was purchased from Sigma. All other reagents were purchased from Sigma unless specified otherwise.

Peripheral blood mononuclear cells were isolated from buffy coats donated from healthy donors by density centrifugation on Ficoll-Hypaque (Amersham-Pharmacia, Freiburg, Germany) according to the manufacturer's recommendations. The cells were microscopically analyzed and counted in a Neubauer chamber. For the isolation of CD3⁺T lymphocytes, the peripheral blood mononuclear cells (3–4 × 10⁸) were incubated for 15 min on ice with anti CD3-antibodies conjugated to magnetic beats (Miltenyi Biotech, Bergisch-Gladbach, Germany). Separation of CD3⁺cells was performed using an L/S column (Miltenyi Biotech) and confirmed by fluorescence-associated cell sorting (> 85% purity). For electrophoretic mobility shift assays (EMSAs) > 5 × 10⁶, T lymphocytes were analyzed per sample.

Jurkat T cells (ACC 282; DSMZ, Braunschweig, Germany) and primary human T lymphocytes, which had been isolated as described previously, were maintained in Roswell Park Memorial Institute 1640 medium supplemented with 10% fetal calf serum, 1% glutamine, and 50 mg/ml penicillin and streptomycin (all from Gibco-BRL, Karlsruhe, Germany) and were grown in a humidified atmosphere containing 5% CO²at 37°C.

Total cell extracts were prepared using a high-salt detergent buffer, Totex (20 mm N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic) acid [HEPES], pH 7.9, 350 mm NaCl, 20% (v/v) glycerol, 1% (w/v) NP-40, 1 mm MgCl², 0.5 mm ethylenediaminetetraacetic acid [EDTA], 0.1 mm ethylene glycol-bis(β-aminoethyl ether)-N,N-tetraacetic acid [EGTA], 0.5 mm dithiothreitol, 0.1% phenylmethylsulfonyl fluoride [PMSF], 1% aprotinin). Cells were harvested by centrifugation, washed once in ice-cold PBS, and resuspended in four cell volumes of the detergent buffer. The cell lysate was incubated for 30 min on ice and then centrifuged for 5 min at 13,000 g  at 4°C. EMSAs were performed using a ³²P-labeled NF-κB oligonucleotide as previously described. 27,28The reaction mixture consisted of 37 μl purified water, 1 μl NF-κB oligonucleotides (25 ng/μl; Promega, Madison, WI), 5 μl kinase buffer, 5 μl γ-³²P-dATP (Amersham International, Braunschweig, Germany), and 1.5 μl T4 kinase (PNK buffer and PNK T4 kinase; New England Biolabs, Schwalbach, Germany) and was incubated for 30 min at 37°C. The protein content of the cell lysates was determined using a Bradford-Assay system (Bio-Rad Laboratories, München, Germany), and equal amounts of protein (30 μg) were added to a 20-μl reaction mixture containing 20 μg bovine serum albumin, 2 μg polydesoxyiosine-desoxycytosine (dI-dC; Roche, Mannheim, Germany), 2 μl buffer D+ (20 mm HEPES, pH 7.9, 20% glycerol, 100 mm KCl, 0.5 mm EDTA, 0.25% Nonidet P-40, 2 mm dithiothreitol, 0.1% PMSF), 4 μl 5× Ficoll buffer (20% Ficoll 400, 100 mm HEPES, 300 mm KCl, 10 mm dithiothreitol, 0.1% PMSF), 4 μl double-distilled water, and 1 μl NF-κB ³²P-labeled oligonucleotide. These samples were incubated at room temperature for 30 min and then loaded on an acrylamide gel containing 60 ml double-distilled water, 10 ml 30% acrylamide, 3.8 ml 10× Tris-borate-EDTA buffer (TBE: 900 mm TRIS-HCl, 900 mm boric acid, 20 mm EDTA [pH 8.0]), 400 μl ammonium persulfate, and 40 μl tetramethylethylenediamine. After running the gel in 0.5× TBE running buffer, gels were vacuum dried (Gel dryer 543; Biorad, Hercules, CA) for 30 min on a 3-MM chromatography filter (Whatman, Maidstone, United Kingdom) and exposed to x-ray film (Kodak, Stuttgart, Germany).

The activation and translocation of NF-κB to the nucleus is preceded by the phosphorylation and proteolytic degradation of the inhibitory IκB-α proteins. This process is readily detectable in Western blots. 18To determine whether thiopental may interfere with the degradation of IκB-α, Jurkat T cells were pretreated with different doses of thiopental (200, 400, or 1,000 μg/ml) for 105 min and subsequently stimulated with TNF-α (20 U/ml) for 15 min. These time points were chosen on the basis of the previously published time course of TNF-α–mediated degradation of IκB-α. 29Total cell extracts of Jurkat T cells (30 μg) were boiled in Laemmli sample buffer and subjected to 10% SDS-PAGE. Prior to transfer, gels were equilibrated for 15 min in cathode buffer (25 mm Tris, 40 mm glycin, 10% methanol). Proteins were transferred at 0.8 mA/cm²for 1 h onto Immobilon P membranes (Millipore Corp., Eschborn, Germany) preequilibrated in methanol (15 s), double-distilled water (2 min each side) and anode buffer II (25 mm Tris–10% methanol), using a semidry blotting apparatus (Bio-Rad Laboratories). Equal loading and transfer were monitored by Ponceau S staining of the membranes. Nonspecific binding sites were blocked by immersing the membrane into blocking solution (TBST [10 mm Tris-HCl, pH 8.0, 150 mm NaCl, 0,1% Tween-20 (v/v)] containing 2% bovine serum albumin) overnight at 4°C. Membranes were washed in TBST and incubated in a 1:1,000 dilution of anti-IκBα antibody or antiphosphorylated-IκBα antibody (IκBα^P) (cat. No. 9241; Cell Signaling Technology Inc., Beverly, MA) in blocking solution for 1 h at room temperature, followed by extensive washing with TBST. Bound antibody was decorated with goat antirabbit–horseradish peroxidase conjugate (Amersham-Pharmacia) and diluted 1:5,000 in blocking solution for 30 min at room temperature. After washing four times (5 min each), the immunocomplexes were detected using ECL Western blotting reagents (Amersham-Pharmacia) according to the manufacturer's instructions. Exposure to Kodak XAR-5 films was performed for 15 s to 1 min.

Lymphocytes were treated with vehicle, with TNF-α alone, or with TNF-α and thiopental at specified concentrations and harvested by centrifugation in ice-cold phosphate-buffered saline containing phosphatase inhibitor cocktail set II (Calbiochem, La Jolla, CA). Equal amounts (500 μg) of whole cell protein extracts were obtained in immunoprecipitation buffer (50 mm HEPES [pH 7.6], 250 mm NaCl, 10% glycerol, 1 mm EDTA) containing 0.1% NP40 with protease and phosphatase inhibitors. The cell lysate was cleared and incubated for 2 h at 4°C with an antibody against IKKα (cat. No. sc-7606 AC; Santa Cruz Biotechnology Inc., Santa Cruz, CA). The immunoprecipitates were washed extensively with immunoprecipitation buffer. One portion of the immunoprecipitated IKKα complexes was run on a separate 10% SDS-PAGE and was Western blotted with the IKKα antibody to check for equal loading. The remaining portion was used to perform the in vitro  kinase assay by incubating the immunoprecipitates with 4 μg IκBα (cat. No. sc-4094; Santa Cruz Biotechnology Inc.) in 20 μl kinase buffer containing 20 mm Tris-HCl (pH 7.6), 10 mm MgCl², 0.5 mm dithiothreitol, 100 μm ATP, and 5 μCi [γ-³²P]ATP for 30 min. The immunoprecipitates were subjected to SDS-PAGE, dried, and visualized by autoradiography.

To determine whether the inhibitory effect of thiopental on NF-κB activation is mediated through a GABA receptor–dependent agonistic mechanism, we performed EMSAs using total cell extracts from Jurkat T cells after a 2-h incubation with thiopental (400 or 1,000 μg/ml) or GABA (3 and 10 mm). To evaluate whether GABA^Aantagonism with bicuculline (10 and 100 μm) or GABA^Bantagonism with CGP 52432 (100 and 500 μm) would be able to prevent the thiopental-mediated inhibitory effect, cells were pretreated with these antagonists for 1 h before thiopental incubation. One hour before harvesting, the cells were stimulated with TNF-α (1 ng/ml) for 1 h, after which total cell protein extracts were prepared and analyzed for the DNA binding activity of NF-κB by EMSA.

To determine whether the inhibitory effect of thiopental could be explained by direct targeting of the NF-κB, cell extracts from TNF-stimulated Jurkat T cells were pooled to achieve equality of activation and were than subsequently fractionated. Extracts were incubated with thiopental (100, 200, 400, or 1,000 μg/ml) for 1 or 2 h in a humidified atmosphere containing 5% CO²at 37°C in a cell-free system. The cell extracts were analyzed for the DNA binding activity of NF-κB by EMSA.

Because the phosphorylated form of IκBα (IκBα^P) is highly transient and is therefore difficult to capture, the proteasome inhibitor MG 132 (Calbiochem Corp.) was used to inhibit the IκBα degradation and to visualize the phosphorylated form of IκBα. Consequently, cells were treated with 10 or 20 μm MG 132 for 1 h before the addition of thiopental or vehicle, and after TNF-α stimulation the whole cell lysates were then collected as described previously.

To determine whether thiopental inhibits signaling pathways leading to IκB phosphorylation, we measured IKK activity in TNF-stimulated Jurkat T cells and CD3⁺T lymphocytes in vitro . To detect IKK activity, Jurkat T cells and CD3⁺T lymphocytes were incubated with TNF-α (1 ng/ml) for 15 min (described by Rossi et al.  30; data not shown) after incubation with various concentrations of thiopental (200, 400, 1,000 μg/ml) for 2 h. Equal loading of IKK was determined by Western blotting for IKKα.

To investigate whether the inhibitory effect of thiopental is confined to the thio-group at the C2 position of the pyrimidine ring, we analyzed the effect of different pairs of structural barbiturate analogs (thiopental–pentobarbital and thiamylal–secobarbital) differing exclusively in this substituent on TNF-induced NF-κB activation. Therefore, Jurkat T cells were incubated with equimolar concentrations of thiopental or pentobarbital as well as thiamylal or secobarbital for 2 h. One hour before harvesting, the cells were stimulated with TNF-α (1 ng/ml) for 1 h, after which total cell extracts were prepared. The cell extracts were analyzed for the DNA binding activity of NF-κB by EMSA.

To test whether NF-κB inhibition is due to the thio-group of thiopental, we precoincubated thiopental with dithiothreitol for 1 h at room temperature in double-distilled water, since thiopental would react with the large quantities of free sulfhydryls in dithiothreitol to chemically modify the sulfur atom by developing disulfide bonds. Jurkat T cells were incubated with dithiothreitol alone (0.01, 0.1, 1, or 5 mm), with thiopental alone (1,000 μg/ml), or with dithiothreitol (0.1, 1, or 5 mm) together with thiopental (1,000 μg/ml) for 2 h. One hour before harvesting, the cells were stimulated with TNF-α (5 U/ml) for 1 h, after which total cell extracts were prepared and analyzed for NF-κB DNA binding activity by EMSA.

Autoradiographs of the kinase assay experiments were evaluated by volume quantification and local median background correction using two-dimensional scanning (Personal Densitometer; Amersham-Pharmacia). Differences in measured variables between the experimental conditions were assessed using one-way analysis of variance on ranks followed by a nonparametric Student-Newman-Keuls test for multiple comparisons. Results were considered statistically significant at P < 0.05. The tests were performed using the SigmaStat software package (Jandel Scientific, San Rafael, CA).

Treatment of Jurkat T cells with TNF-α induced NF-κB DNA binding activity (fig. 1, lanes 1  and 2 ), which was inhibited by pretreatment of cells with 1,000 μg/ml thiopental as previously reported (fig. 1, lane  4). 19In contrast, incubation of Jurkat cells with GABA alone (3 and 10 mm) had no effect on the activation of NF-κB (fig. 1, lanes 5  and 6 ). The addition of TNF-α to GABA-treated cells resulted in NF-κB activation similar to that observed in control cells (fig. 1, lanes 7  and 8 vs. lane 2 ). Pretreatment of Jurkat T cells with either the GABA^A-antagonist bicuculline (10 and 100 μm) or the GABA^B-antagonist CGP 52432 (100 and 500 μm) did not prevent the inhibition of NF-κB by thiopental (fig. 1, lanes 9–12 ).

In vitro  incubation of cell extracts containing activated NF-κB with different concentrations of thiopental (100, 200, 400, 1,000 μg/ml as indicated) for 1 h (fig. 2, lanes 2–6 ) and 2 h (fig. 2, lanes 8–12 ) had no effect on NF-κB DNA binding activity as detected by EMSA.

We have previously demonstrated that thiopental (1,000 μg/ml) prevents the degradation of IκBα. 19As shown in the Western blot depicted in figure 3, Jurkat T cells that were treated with the proteasome inhibitor MG 132 (10 or 20 μm) and subsequently stimulated with TNF contained IκBα^Pin much larger amounts (fig. 3, lanes 3  and 4, top ), had a higher IκBα content (fig. 3, lanes 3  and 4, middle ), and showed a reduced activation of NF-κB (fig. 3, EMSA, lanes 3  and 4, bottom ) as compared to TNF alone (fig. 3, lane 2 ). This pattern remained unchanged if Jurkat T cells were pretreated with 10 μm MG 132, followed by incubation with thiopental at concentrations of 200 or 400 μg/ml (fig. 3, lanes 5  and 7, top, middle , and bottom ), i.e. , IκBα^Paccumulated, IκBα was degraded, and consequently, no NF-κB was activated. In contrast, after preincubation with 10 μm MG 132, followed by incubation with 1,000 μg/ml thiopental and subsequent TNF stimulation, no IκBα^Pcould be detected (fig. 3, lane 9, top ), much more IκBα was present (fig. 3, lane 9, middle ), and consequently, much less NF-κB was activated (fig. 3, lane 9, bottom ).

Basal IKK activity was low in unstimulated CD3⁺T cells (fig. 4, lane 1, top ), whereas the subsequent incubation with TNF-α caused an increase in IKK activity (fig. 4, lane 2, top ). Lower concentrations of thiopental (200 and 400 μg/ml) had no major effect on the TNF-induced increase in IKK activity (fig. 4, lanes 3  and 4, top ). In contrast, IKK activity remained at the level of unstimulated controls if the cells were treated with TNF in the presence of high concentrations of thiopental (1,000 μg/ml;fig. 4, lane 5, top ). Under these conditions, densitometric analysis revealed a 60% decrease of IKK activity (fig. 4, bottom ). IKK recovery was comparable under all conditions as determined by immunoblotting for IKKα (fig. 4, top ). Similar results could also be obtained using Jurkat T cells instead of CD3⁺T cells (data not shown).

Pretreatment of Jurkat T cells with 4 mm thiopental inhibited the activation of NF-κB, whereas an equimolar concentration of pentobarbital did not (fig. 5, lanes 3  and 4 ). In addition, 4 mm thiamylal suppressed the DNA binding of NF-κB to its DNA probe, whereas the equimolar concentration of secobarbital had no detectable effect (fig. 5, lanes 9  and 10 ). Lower equimolar concentrations of thiopental–pentobarbital and thiamylal–secobarbital did not affect DNA binding of NF-κB to its DNA probe in a detectable manner (fig. 5, lanes 5–8  and 11–14 ).

Incubation of Jurkat T cells with dithiothreitol followed by subsequent stimulation with TNF-α showed no detectable inhibition at 0.01 mm and 0.1 mm dithiothreitol (fig. 6, lanes 3  and 4 ). In contrast, NF-κB activation was suppressed at 1 and 5 mm dithiothreitol alone (fig. 6, lanes 5  and 6 ) and, as previously shown, by thiopental alone (1,000 μg/ml;fig. 6, lane 7 ). Precoincubation of thiopental (1,000 μg/ml) with dithiothreitol (0.1 mm) attenuated the inhibitory effect of thiopental on the DNA binding activity of NF-κB (fig. 6, lane 8 ). Because of the intrinsic inhibitory properties of higher concentrations of dithiothreitol on NF-κB activation, an attenuation of the thiopental-mediated suppressing effect could not be demonstrated if dithiothreitol was present at concentrations of 1 or 5 mm (fig. 6, lanes 9  and 10 ).

Barbiturates may be beneficial in patients with severe head injury and refractory intracranial hypertension. This conclusion is based on a series of clinical studies showing that administration of barbiturates can reduce intracranial pressure and increase cerebral perfusion pressure following brain trauma, as long as systemic hemodynamic stability is maintained. 31,32Despite these favorable effects, accumulating evidence suggests that the long-term administration of high doses of thiopental is associated with a suppression of immune functions. For example, it has been shown that thiopental treatment causes a profound increase in the incidence of nosocomial infections, which may in turn contribute to the high mortality rate of these patients. 2,33Evidence from our previous work suggested that inhibition of the activation of NF-κB by thiopental may be involved in mediating the immunosuppressive effects of this agent. 19However, the molecular mechanism by which thiopental exerts these effects remains unknown. Thus, it was the aim of the current study to identify the molecular mechanism by which thiopental inhibits the activation of NF-κB.

Barbiturates are potent agonists of the receptors for the ubiquitous inhibitory neurotransmitter GABA. 34Interestingly, immunomodulatory actions through the GABA^Areceptor complex have been reported. 26,35,36Furthermore, Tian et al.  25recently demonstrated the presence of functional GABA^Areceptors on T cells, which mediate the lymphoproliferating effects of GABA and can be pharmacologically manipulated in a manner similar to neuronal GABA^Areceptors. Thus, it would be tempting to speculate that the inhibitory effect of thiopental on NF-κB is mediated through GABA receptor stimulation. However, suppression of the NF-κB DNA binding was not mimicked by incubation of the cells with GABA, nor could pretreatment of T cells with GABA^A(bicuculline) or GABA^B(CGP) antagonists attenuate the thiopental-mediated inhibition of NF-κB. These findings strongly argue against a key role of the GABA receptor complex in mediating the inhibition of NF-κB by thiopental.

In a next step, we systematically evaluated if and how thiopental may interfere with different components of the signal transduction pathway leading to the activation of NF-κB. Both the physicochemical properties of thiobarbiturates and the fact that other agents have been shown to directly modulate the DNA binding activity of NF-κB prompted us to examine whether thiopental may act in a similar fashion. 37However, our experiments using a cell-free system strongly suggest that the inhibitory effect of this agent is not due to a direct molecular targeting of the p50/p60 heterodimer protein and must therefore occur further upstream. The translocation of free, active NF-κB into the nucleus is preceded by the ubiquitination and degradation of its inhibitor IκB. 18This requires the phosphorylation of IκB. 13Therefore, our observation that less phosphorylated IκB is detectable after TNF stimulation in the presence of thiopental provided a first hint toward a potential target of its inhibitory action. Theoretically, this finding could be due to an increased degradation of phosphorylated IκB. However, this can be excluded because the experiments were performed in the presence of the proteasome inhibitor MG 132 at concentrations proven to effectively prevent its breakdown, and the amount of unphosphorylated IκB remained unchanged as compared to TNF stimulation alone. Alternatively, suppression of the accumulation of phosphorylated IκB on TNF stimulation in the presence of thiopental could be the result of an interference with its formation. Therefore, we evaluated whether thiopental may abrogate the activity of IKK, the enzyme complex responsible for the phosphorylation of IκB. The observation that thiopental suppresses the increase in IKK activity, which can otherwise be observed on TNF stimulation, identifies this enzyme complex as a target of thiobarbiturates.

This raises the question of how thiopental alters IKK activity. Previous studies have demonstrated that thio-group reactive agents may block IKK activity and prevent the subsequent activation of NF-κB. 38Structural analyses of IKK subunits suggest that cysteine residues are present in the activation loop and within the kinase domain of IKKα and IKKβ at sites critical for enzymatic activity 16,39,40and could therefore serve as molecular targets for these agents. Thus, we hypothesized that thiopental acts in a similar fashion, i.e. , the inhibitory action depends on its thio-group. To test this hypothesis, we compared the effects of the thiobarbiturates thiopental and thiamylal on the activation of NF-κB with those of their structural oxyanalogs pentobarbital and secobarbital. Interestingly, in contrast to the former agents, the latter two oxybarbiturates failed to inhibit NF-κB, if present in equimolar amounts. These results have two important implications. First, they support the hypothesis that the thio-group is of functional importance for the inhibitory action of barbiturates. Second, the difference between thiobarbiturates and oxybarbiturates in their ability to inhibit the activation of NF-κB described herein would provide an explanation for the results of previous reports showing that thiobarbiturates are more potent suppressors of immune responses that depend on the appropriate activation of NF-κB as compared to their oxyanalogs. 7,41 

In another attempt to define the functional role of the thio-group within the barbiturate molecule for the suppression of NF-κB, we performed a series of experiments with dithiothreitol, which contains large quantities of free sulfhydryls. As could be expected from this molecular structure and in agreement with the results of previous studies, 27dithiothreitol inhibited the activation of NF-κB at concentrations of 1 mm or higher. However, precoincubation of thiopental with dithiothreitol at lower concentrations that did not exert any intrinsic inhibitory effect, attenuated the inhibition of NF-κB by thiopental. These findings strongly suggest that interaction of the thiol-group containing thiopental with the sulfhydryl groups within the dithiothreitol molecule may have limited the availability of this functional group for interactions with the IKK complex, and in turn attenuated the inhibitory action of thiopental. Therefore, these results suggest that the sulfur atom at the C2 position within the thiobarbiturates is a structural requirement for the inhibitory action of these agents on the activation of NF-κB.

Identification of the mechanism and the molecular structure responsible for thiopental-mediated immunosuppression could form a basis for the development of new strategies for the therapy of intracranial hypertension. For example, separation of the neuroprotective from the immunomodulating effects may be beneficial in patients whose survival depends on lowering the intracranial pressure but who are simultaneously at a high risk for the development of nosocomial infections. Thus, the results of the current study provide a molecular rationale for future investigations that systematically examine the comparative clinical efficacy of oxybarbiturates versus  thiobarbiturates in improving the outcome after severe head injury.

In conclusion, inhibition of the NF-κB DNA binding activity by thiopental is not due to GABA receptor stimulation and does not involve direct targeting of activated NF-κB. Our results rather indicate that thiopental suppresses the NF-κB activating signaling cascade by altering IKK activity and that the thio-group at the C2 position within the barbiturate molecule plays a key role in mediating this effect.