Volatile Anesthetics Differentially Affect Immunostimulated Expression of Inducible Nitric Oxide Synthase: Role of Intracellular Calcium

Halothane was obtained from Hoechst Marion Roussel (Bad Soden, Germany). Isoflurane and enflurane were purchased from Abbott (Wiesbaden, Germany). Desflurane was provided by Pharmacia & Upjohn (Erlangen, Germany). γIFN and polyclonal IL-1β–neutralizing antibodies were obtained from R&D Systems (Wiesbaden, Germany). All other reagents, media and materials were obtained from Sigma Aldrich Chemical (Deisenhofen, Germany) unless otherwise indicated.

J774A.1 cells, a murine macrophage-like cell line, were obtained from American Type Culture Collection (Rockville, MD), and cultured in bottles (Becton Dickinson, Heidelberg, Germany) containing RPMI 1640 medium, 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 μg/ml). For stimulation experiments, confluent cells were harvested and then resuspended and plated in two different 24-well culture plates (Costar, Cambridge, MA) at a concentration of 1 × 10⁶/ml. Two hours after incubation at 37°C in 5% CO², supernatant was aspired to remove nonadherent cells and replaced by fresh culture medium. One cell culture plate was then incubated at 37°C in 5% CO²for 18 h in a chamber preequilibrated with 0.25 minimum alveolar concentration (MAC), 0.5 MAC, 1.0 MAC, and 2.0 MAC volatile anesthetic. The MAC values used were isoflurane 1.15%, enflurane 1.75%, halothane 0.75%, and desflurane 7.2%. The other cell culture plate (anesthetic control) was treated in parallel and incubated at 37°C in 5% CO²without volatile anesthetic. Equilibration with the volatile anesthetic was achieved and maintained by introducing an anesthetic/air mixture and controlling the concentration of the anesthetic and the CO²by using a gas analyzer (PM 8050, Draeger Medizintechnik, Luebeck, Germany) that continuously monitored the gas concentration in the incubator in a closed-loop system. After 1 h of incubation in the presence or absence of volatile anesthetic, macrophages were stimulated with lipopolysaccharide (1 μg/ml), γIFN (200 U/ml), or lipopolysaccharide (1 μg/ml) plus γIFN (200 U/ml) and then further incubated at 37°C in 5% CO²with and without volatile anesthetic. Nitrite, a stable metabolite and measure for NO production, was determined in the supernatant 1 h before and 18 h after stimulation. In addition, the expression of iNOS mRNA and iNOS protein was analyzed 4 h and 18 h after stimulation, respectively. To assess the time course of the effect of volatile anesthetics on immunostimulated NO production, cells were exposed to volatile anesthetics 1 h before and 2 h, 4 h, 6 h, and 12 h after stimulation. Nitrite concentration in the culture supernatant was measured 18 h after stimulation and compared with the nitrite level of immunostimulated cells not exposed to volatile anesthetics that were run in parallel. In additional experiments, calcium-ionophore ionomycin (1 μM), diacylglycerol (Calbiochem, Bad Soden, Germany; 3 μg/ml), phorbol myristate acetate (1 μg/ml), C2-ceramide (12 μM), or polyclonal interleukin-1β (IL-1β)–neutralizing antibodies (1 ng/ml) were added 30 min before immunostimulation in the presence or absence of 1.0 and 2.0 MAC of inhalational anesthetics, and nitrite concentrations in the culture supernatant and iNOS expression were determined.

Nitrite concentration in the culture supernatant medium was assessed as previously described. 23In brief, Griess color reagent was freshly prepared by combining (1:1) reagent A (0.1% N-[1-naphthyl]etyhlendiamine) and reagent B (1%, w/v, sulfanilamide in 5% H³PO⁴). Samples (100 μl) were transferred to a 96-well plate (Costar). Griess color reagent (100 μl) was added, and 10 min after incubation at 21°C, absorbance was read at 560 nm using a Multiskan MCC/340 plate reader (Titertek Instruments Inc., Huntsville, AL). Standard curves were constructed by using sodium nitrite (0–120 μM) dissolved in supplemented culture medium.

At the end of stimulation (18 h), J774 mouse macrophages (10⁶/well) were washed twice with phosphate-buffered saline (PBS) and lysed by treatment with 50 μl boiling lysis solution (1% sodium dodecyl sulfate, 1.0 mM sodium ortho-vanadate, 10 mM Tris, pH 7.4). Lysed samples were transferred to a microcentrifuge tube and boiled for additional 5 min. To reduce viscosity, the samples were passed several times through a 27-gauge needle to shear DNA. Protein concentration of each sample was measured in a 10-fold–diluted aliquot using the bicinchoninic acid assay from Pierce (Rockford, IL). To all samples an equal volume of two times the concentrated electrophoresis sample buffer (125 mM Tris-HCl, pH 6.8, 4% sodium dodecyl sulfate, 10% glycerol, 0.006% bromophenol blue, and 2%β-mercaptoethanol) was added, and the samples were boiled for 5 min. Per lane, 15 μg total cell extract protein, which was approximately 15 μl, was used for iNOS Western blot analysis.

Cell extract proteins were separated by a standard sodium dodecyl sulfate–polyacrylimide gel electrophoresis (SDS-PAGE; 7.5 ExelGel SDS Homogenous, Pharmacia & Upjohn, Erlangen, Germany) and electrophoretically transferred to a nitrocellulose membrane (0.45 μm, Schleicher & Schuell, Keene, NH) using a NovaBlot Electrophoretic Transfer Kit (Pharmacia & Upjohn). The membrane was then immersed for 1 h at 25°C in PBS containing 5% nonfat dry milk, washed twice with Tween-PBS (0.05% Tween 20), and incubated overnight at 4°C with anti-iNOS monoclonal antibody (Transduction Laboratories, Lexington, KY) diluted in PBS containing 1% nonfat dry milk. The membrane was then washed three times with Tween-PBS and incubated for 1 h at 25°C with sheep antimouse immunoglobulin G biotin conjugate (Boehringer Mannheim, Mannheim, Germany). After further washing steps, the membrane was incubated for 45 min at 25°C with avidin-peroxidase. For visualization of the protein, the membrane was washed again, incubated with ECL detection reagents (Amersham International, Braunschweig, Germany) and exposed to Kodak XR film (Kodak, Stuttgart, Germany). Lysate from immunostimulated RAW 264.7 cells were used as a positive control for iNOS protein (Transduction Laboratories).

Four hours after immunostimulation, total RNA was extracted from 15 × 10⁶J774 cells using peqGOLD TriFas FL (peqlab Biotechnologie, Erlangen, Germany) and the method of Chomczynski. 24Ten micrograms of total RNA per sample were denatured, run on formaldehyde/agarose gels, and transferred to Biodyne B nylon membranes (Pall Filtron, Dreieich, Germany). These were then ultraviolet cross-linked, prehybridized, hybridized with a α-[³²P]dCTP random primer–labeled cDNA plasmid probe specific for iNOS, washed, and exposed to Kodak XAR films as previously described. 25The iNOS cDNA clone was obtained from Dr. C. Nathan (Beatrice and Samuel A. Seaver Laboratory, Department of Medicine, Cornell University Medical College, New York, NY). The blots were also hybridized with a γ-[³²P]ATP end-labeled oligoprobe specific for mouse 16S mitochondrial RNA to normalize the iNOS mRNA data for RNA loading and transfer. For quantitative comparison of mRNA levels, bands were analyzed using a BAS 3000 phosphor imager (Fuji Medical Systems, Stamford, CT), and the intensities were expressed as phosphor imager units. After correcting for background activity, the ratios of iNOS mRNA levels to mitochondrial RNA levels were calculated. These values were then expressed as percent of mRNA levels of cells immunostimulated in the absence of volatile anesthetics.

Data are presented as mean ± SD unless otherwise specified. For each volatile anesthetic studied, data for nitrite production at different anesthetic concentrations, different stimulatory conditions, and different time points after stimulation, respectively, were compared with control values using analysis of variance (ANOVA) followed by unpaired, two-tailed t  tests with Bonferroni correction of the t  statistic for multiple comparisons. Individual comparisons were carried out using the Mann–Whitney U test. Significance was accepted at P < 0.05.

We first studied the effect of volatile anesthetics on immunostimulated NO production. Table 1shows that halothane and enflurane, and isoflurane and desflurane, significantly inhibit the nitrite release into the culture supernatant in response to lipopolysaccharide and γIFN. In contrast, all volatile anesthetics studied increased the nitrite release after costimulation with lipopolysaccharide plus γIFN. The inhibitory effect of volatile anesthetics on the induced nitrite release in the culture supernatant after a single stimulation with lipopolysaccharide or γIFN occurred dose-dependently with an 50% inhibitory concentration (IC⁵⁰) of about 0.5 MAC, as shown in figures 1A and 1Bfor halothane and isoflurane, respectively. Similar data were obtained for enflurane and desflurane (not shown). At an anesthetic concentration of 0.25 MAC, we already found a significant inhibition of the induced NO production that continuously increased to almost complete inhibition at a concentration of 2 MAC. With costimulation with lipopolysaccharide and γIFN, however, NO production was significantly increased (between 10% and 20%). With all anesthetics, maximum increase of costimulated nitrite release occurred at 0.5 MAC, showing a lower but still elevated nitrite level compared with controls, with anesthetic concentration of 1 MAC and 2 MAC.

To further investigate the mechanism of the modulatory effect of volatile anesthetics on nitrite production and whether there is a sensitive time period after lipopolysaccharide stimulation, the onset time of anesthetic exposure from 1 h before stimulation to 12 h after stimulation was varied. As shown in figure 2, a distinct time-dependence of the modulating effects on nitrite production could be detected (fig. 2). Regarding single stimuli, incubation with anesthetics 1 h before stimulation had the most profound inhibitory effect on nitrite release. The inhibitory effect of volatile anesthetics on lipopolysaccharide- and γIFN-stimulated nitrite accumulation was significantly less (about 50%) when anesthetics were applied 6 h after stimulation and absent when applied 12 h after stimulation. In contrast, nitrite production after costimulation with lipopolysaccharide plus γIFN was maximally potentiated by halothane, and by other anesthetics studied (data not shown), when cells were exposed to the anesthetic 4 h after costimulation. Inhibitory and stimulatory effects of all volatile anesthetics studied were limited to the first 12 h after stimulation, indicating that they interfere with the induction of NOS expression and not with NOS activity. In fact, induced NOS activity in the homogenate of immunostimulated cells could not be inhibited by 2 MAC of any volatile anesthetic (data not shown).

To determine whether volatile anesthetics modulate the iNOS protein expression of immunostimulated J774 macrophages, we analyzed total cell lysates by Western blot using murine iNOS-specific antibodies. As shown in figure 3A, iNOS protein, which was detected as a 130-kDa protein band only in stimulated cells, almost completely disappeared when cells were exposed to 2 MAC isoflurane during stimulation (lanes 3 and 6). The inhibitory effect of volatile anesthetics on nitrite production after single stimulation was caused by a specific reduction of the expression of inducible NOS and not by a reduction in the viability of the cells, which were studied by Trypan blue exclusion. In line with our findings on costimulated NOS activity, iNOS protein expression in costimulated cells was significantly increased by incubation with a volatile anesthetic (fig. 3B). This up-regulation of iNOS expression did not occur in unstimulated cells with volatile anesthetics alone (fig. 3B, lane 4).

Northern blot analysis revealed that the modulatory effect of inhalational anesthetics on induction of iNOS protein and enzyme activity was due to an interference with iNOS gene transcription (figs. 4A and 4B). Whereas single immunostimulation with lipopolysaccharide and γIFN resulted in a distinct induction of iNOS mRNA in the absence of volatile anesthetic (fig. 4, lanes 6 and 7), iNOS transcription was almost completely suppressed after 4 h of incubation with 2 MAC isoflurane (fig. 4, lanes 2 and 3). The ratios of the quantified iNOS mRNA and mRNA levels that were determined as a control for appropriate loading and transfer confirmed a specific modulatory effect of volatile anesthetics on iNOS gene transcription. In contrast to single immunostimulation, costimulation of the cells with lipopolysaccharide plus γIFN induced a substantially higher iNOS mRNA level which was no longer inhibited by volatile anesthetics (fig. 4A, lane 1 vs.  5).

In three separate experiments, iNOS transcription induced by a single immunostimulation with lipopolysaccharide or γIFN was significantly suppressed in the presence of 2 MAC isoflurane (by more than 80%) compared with iNOS mRNA levels in the absence of volatile anesthetics (fig. 4C). In contrast to single stimulation, mean mRNA level in costimulated cells was increased by 2 MAC isoflurane. Although not reaching statistical significance (P = 0.03), the potentiating effect on costimulated iNOS transcription correlated well with the up-regulation in iNOS protein expression and induced enzyme activity that we observed under these conditions.

To investigate the underlying mechanism by which volatile anesthetics may modulate iNOS expression in immunostimulated macrophages, we assessed the role of intracellular calcium mobilization in iNOS expression using the calcium ionophore ionomycin. As shown in figures 3A and 4B, addition of ionomycin 1 μM 30 min before single stimulation of the cells with lipopolysaccharide or γIFN completely restored the ability of lipopolysaccharide and γIFN to induce iNOS protein and mRNA expression in cells exposed to a volatile anesthetic that, in the absence of ionomycin, were strongly inhibited to express iNOS protein and mRNA, respectively. The antagonistic effect of ionomycin on the isoflurane-induced inhibition of iNOS protein expression was also reflected by induced NOS activity determined by the nitrite accumulation in the culture supernatant (Tab. 2). Whereas ionomycin per se  was unable to induce nitrite expression in unstimulated J774 cells, it significantly augmented the nitrite production of macrophages not exposed to anesthetics in response to a single stimulation with lipopolysaccharide or γIFN. The presence of ionomycin significantly restored nitrite production that was profoundly reduced after exposure to 2 MAC isoflurane to levels measured in singly stimulated cells in the presence of ionomycin and absence of volatile anesthetic.

Costimulated cells, however, showed a much higher levels of nitrite production and iNOS mRNA, respectively, which was even further increased by exposure to isoflurane but was not significantly affected by ionomycin, regardless of an incubation with the volatile anesthetic (tab. 2;figs. 4B and 4C).

Because halothane was described to increase IL-1β release by alveolar macrophages 26and IL-1β is known to induce iNOS in macrophages, 27,28we tested the hypothesis that the stimulatory effect of anesthetics on costimulated J774 cells is mediated by IL-1β. However, the addition of polyclonal IL-1β–neutralizing antibodies before costimulation with lipopolysaccharide plus γIFN could not prevent the stimulatory effect of anesthetics on iNOS expression (data not shown). Similarly, the presence of cell-permeable diacylglycerol and phorbol myristate acetate (both activators of protein kinase C) during stimulation of the cells had no significant effect on the modulation of nitrite production seen under exposure to 1 MAC volatile anesthetic, as shown in figure 5for enflurane and desflurane. Similar results were obtained when cells were stimulated in the presence of cell-permeable C2-ceramide, which is known as an activator of ceramide-activated protein kinase (data not shown). Thus, with exception of the calcium ionophore ionomycin, none of the protein kinase activators studied was able to prevent the changes in iNOS expression in immunostimulated macrophages mediated by volatile anesthetics.

NO generated by inducible NOS has been established as an important factor in the pathogenesis of sepsis. 29–31Volatile anesthetics such as halothane and isoflurane have been reported to interfere with the activity of constitutive endothelial and brain NOS. 8–10,15 

In this study, we have shown that volatile anesthetics can have both a suppressive and a potentiating effect on immunostimulated NO production and iNOS gene expression in murine macrophages, depending on the stimulus used. Our findings suggest that volatile anesthetics do not directly interfere with already expressed iNOS activity, as assessed by the lack of effect on nitrite production in lysates of stimulated cells and activated cells 12 h after stimulation. Moreover, Northern blot analysis and immunoblotting of the cell extracts revealed that the modulatory effect of volatile anesthetics on stimulated NO production affected both the transcriptional and the translational level of iNOS gene expression within the first hours after immunostimulation.

During costimulation with lipopolysaccharide plus γIFN, which resulted in a strong induction of NOS activity, iNOS protein, and mRNA, the presence of volatile anesthetics further increased NO production and iNOS expression in J774 macrophages. Time– and dose–response analysis of the stimulatory effect of volatile anesthetics revealed a nonlinear relation with a maximum at a concentration of 0.5 MAC after 4 h of costimulation. With increasing the concentrations above 0.5 MAC, the up-regulatory effect on costimulated nitrite production decreased. Similarly, exposure of the cells to inhalational anesthetics 1 h before or 2 h and 6 h after costimulation of the cells resulted in significantly less stimulatory effect, indicating that an anesthetic-sensitive step is located around the hour 4 of stimulation.

We therefore assumed that volatile anesthetics could potentiate costimulated iNOS gene expression by stimulating a relatively late event, such as the release of a proinflammatory cytokine implicated in iNOS up-regulation. Because volatile anesthetics have been shown to increase IL-1β release in macrophages 26and interleukin-1β is known to induce iNOS in macrophages, 27,28we evaluated whether an increase in IL-1β by volatile anesthetics might account for the augmented iNOS expression in costimulated J774 cells. However, we found no evidence for this hypothesis because neutralizing IL-1β antibodies failed to prevent the up-regulation of iNOS expression induced by volatile anesthetics. Alternatively, additional activation of protein kinase C, which has been shown to occur by volatile anesthetics 32and to play a role in lipopolysaccharide-stimulated iNOS induction, 33must be considered. However, our data do not support this notion because protein kinase C activators had no effect on nitrite production by costimulated nor by singly stimulated cells. Another possible explanation for the finding of an up-regulated iNOS expression by volatile anesthetics could be that anesthetics increase the stability of iNOS mRNA transcripts by inhibition of ribonucleolytic action similar to a mechanism that has been reported for IL-10. 34 

In contrast to the potentiating effect on costimulated iNOS expression, after single stimulation with lipopolysaccharide or γIFN, which provides only a weak iNOS induction, volatile anesthetics were found to strongly inhibit nitrite production and iNOS gene expression at the transcriptional level. The assumption of a direct and additional inhibitory effect on the translational level (i.e. , iNOS protein synthesis) is not supported by our findings because in this case iNOS protein and nitrite production should also be reduced after costimulation with lipopolysaccharide plus γIFN. Likewise, the differential effect on iNOS expression according to the stimulatory condition does not advocate for the possibility of a generally reduced iNOS mRNA half-life, but instead indicates an inhibitory effect of the volatile anesthetics on the signal transduction upstream of iNOS gene transcription.

Furthermore, we demonstrate for the first time that the inhibitory effect of inhalational anesthetics can be completely antagonized by the calcium ionophore ionomycin. In addition, we found that the presence of ionomycin during single stimulation with lipopolysaccharide or γIFN enhanced the expression but not the activity of iNOS, whereas NO production of costimulated and unstimulated cells, respectively, was unaffected by ionomycin. In previous experiments, we also observed an inhibitory effect of verapamil (150 μM) and EGTA (2 mM) on immunostimulated nitrite production by J774 macrophages (unpublished observations). These findings indicate that cytosolic-free Ca²⁺potentiates those immune stimuli that moderately induce iNOS expression in J774 macrophages.

In several studies, volatile anesthetics have been shown to reversibly inhibit voltage-dependent calcium channels and affect intracellular calcium mobilization, resulting in a decreased concentration of intracellular free Ca²⁺. 12–15,35Our data therefore support the conclusion that volatile anesthetics mediate the inhibitory effect on iNOS expression in singly stimulated cells by inhibiting mobilization of cytosolic free Ca²⁺.

In line with our findings, there are several reports demonstrating a significant role of Ca²⁺as a priming signal in the transcriptional regulation of iNOS in murine peritoneal macrophages and rat Kupffer cells after a single stimulation with lipopolysaccharide or γIFN. 16–19In those studies, an increase in cytosolic free Ca²⁺enhanced iNOS expression, whereas blocking of cytosolic free Ca²⁺significantly decreased iNOS expression and NO release. In contrast, in immunostimulated RAW 264.7 cells, an increase in intracellular free Ca²⁺has been shown to inhibit iNOS transcription and NO production. 19Moreover, RAW 264.7 cells respond to a single stimulation with lipopolysaccharide with a considerably higher production of iNOS activity than J774 cells (unpublished observations). This could explain the discrepancy between the effect of volatile anesthetics observed in RAW 264.7 cells 22and J774 macrophages after single immunostimulation. Whereas Zuo and Johns 22reported that halothane and isoflurane up-regulate constitutive and inducible NOS expression in RAW 264.7 cells after a single stimulation with lipopolysaccharide, we found that inhalational anesthetics potentiate iNOS expression in J774 macrophages only during costimulation with lipopolysaccharide plus γIFN. Thus, the effect of volatile anesthetics and intracellular free Ca²⁺on the transcriptional regulation of iNOS appears to partly depend on the stimulus used and the cell type studied.

There are several lines of evidence regarding the mechanism linking impaired up-regulation of cytosolic free calcium by volatile anesthetics to inhibition of further downstream activated transcription factors regulating iNOS gene expression. First, immediate early genes jun -B and c-fos  have been shown to be up-regulated by intracellular calcium 36and inhibited by halothane. 20,21Second, fos and jun are known to form the AP-1 transcription factor that could in addition stimulate the promoter activity of the murine iNOS gene, together with other transcription factors such as nuclear factor NF-κB. 37,38In the light of these findings, we would like to speculate that volatile anesthetics might inhibit iNOS expression by interfering with calcium mobilization and subsequent activator protein-1 activation when stimuli are used that induce only submaximal promoter activity of iNOS gene. In contrast, when cells were costimulated with lipopolysaccharide plus γIFN and the iNOS promoter was already maximally activated by nuclear factor NF-κB 39and IFN regulatory factor-1, 40additional transcriptional activators such as AP-1 are redundant. Thus, costimulation of J774 macrophages with lipopolysaccharide plus γIFN results in a strong iNOS induction, and volatile anesthetics can no longer impair iNOS expression by their interference with intracellular Ca²⁺mobilization.

Furthermore, we addressed the question whether volatile anesthetics could exert their modulatory effect on iNOS expression by interference with the activation of protein kinase C and ceramide-activated protein kinase that has been shown to be required for iNOS expression. 33,41,42However, neither activation of protein kinase C by phorbol myristate acetate and diacylglycerol nor the activation of ceramide-activated protein kinase by cell-permeable C2-ceramide could overcome the modulatory effect of volatile anesthetics on iNOS expression.

We have shown that volatile anesthetics time- and dose-dependently inhibit nitrite production, iNOS protein, and mRNA expression at clinically relevant concentrations in J774 murine macrophages within the first hours after single stimulation with lipopolysaccharide or γIFN. This inhibitory effect could be completely antagonized by ionomycin, suggesting an impaired up-regulation of intracellular free calcium as the pathophysiologic mechanism. By contrast, nitrite production and iNOS expression after costimulation with lipopolysaccharide plus γIFN are moderately enhanced by volatile anesthetics, and this occurrence is not affected by ionomycin. The mechanisms for this potentiating effect are presently not known but may be related to an increased stability of iNOS mRNA. The results of the current study add further evidence that volatile anesthetics can have a specific and differential effect on signal transduction and gene expression at clinically relevant concentrations. Thus, our findings may have important implications for experimental studies investigating iNOS induction and also for clinical medicine in regard to anesthesia administration during major surgery and analgosedation in critically ill patients.