Halothane but Not Isoflurane Attenuates Interleukin 1β– induced Nitric Oxide Synthase in Vascular Smooth Muscle

With the approval of the Animal Use Committee of Wakayama Medical College, 103 male Wistar rats weighing 300–400 g were anesthetized with intraperitoneally injected pentobarbital (50 mg/kg). Midline abdominal incisions were followed by exsanguinations via  the abdominal aorta. Each descending thoracic aorta was then removed, cleaned of adherent connective tissue, and cut into rings approximately 3 mm long, which were used for the isometric force measurements, and four longitudinal strips, which were used for the determination of the cGMP and cAMP contents. The endothelium of each ring was removed by gently rubbing the intimal surface with a swab to exclude the possibility of NOS induction by cytokines within the endothelium. 14The aortic rings were bathed in 10-ml organ baths containing Krebs bicarbonate solution (composition: 119 mm NaCl, 4.7 mm KCl, 1.17 mm MgSO⁴, 25 mm NaHCO³, 1.18 mm KH²PO⁴, 2.5 mm CaCl², and 11 mm glucose), which was gassed with 5% CO²in 95% O²and maintained at 37°C. The Krebs solution contained indomethacin (10⁻⁵m) throughout the experiments to exclude the influences of prostaglandins.

Halothane and isoflurane were introduced into the gas mixture through agent-specific vaporizers (Fluotec 3, Cyprane, Keighley, United Kingdom, for halothane; Fortec, Cyprane, for isoflurane). Their concentrations in the resulting gas mixtures were monitored and adjusted by using an anesthetic agent monitor (model 303, Atom, Tokyo Japan). The concentrations of anesthetics in the bathing solution were measured by gas chromatography as described previously. 15The concentrations of halothane at 1, 2, and 3% in the bathing solution were 2.93 × 10⁻⁴, 5.87 × 10⁻⁴, and 8.80 × 10⁻⁴m, respectively, and that of isoflurane at 3% was 6.52 × 10⁻⁴m (n = 3).

To keep the experimental system free from lipopolysaccharide, the organ bath and surgical tools were heated at 250°C for 2 h. Heat-labile parts were immersed in 99.9% (vol/vol) ethanol for 12 h. The distilled water used for these experiments was suitable for injection in humans (Otsuka Pharmaceuticals, Tokyo, Japan).

Aortic rings were vertically fixed between two parallel tungsten wire hooks. The lower hook was attached to a support leg and the upper one to a force transducer (Nihondenki-sanei Co., Tokyo, Japan). The aortic rings were placed under a resting force of 3.0 g, which had been determined to be optimal in length–tension relation experiments. Isometric contractions and relaxations were recorded and displayed on a polygraph recorder using a force transducer and a carrier amplifier (Type N6682, Nihondenki-sanei Co., Tokyo, Japan). The rings were allowed to equilibrate in the control medium for 90 min, during which the bathing fluid was exchanged every 15 min. After equilibration, rings with and without an intact endothelium were tested for the presence and absence, respectively, of relaxant response to acetylcholine (10⁻⁶m) after precontraction with phenylephrine (3 × 10⁻⁷m). Contractile responses to phenylephrine and KCl (30 mm) were obtained twice, before and after the exposure to test agents. Contractile responses of second trial were expressed as a percentage relative to the maximal responses (phenylephrine, 10⁻⁵m; KCl, 30 mm) of the first trial. Phenylephrine and KCl were used as constrictors for the purpose of demonstrating that the effects of agents used in the study are not specific to one contractile mechanism.

In the preliminary study, IL-1β exposure (20 ng/ml) for 1, 3, 5, and 7 h inhibited the maximal contraction caused by phenylephrine (10⁻⁵m) in a time-dependent manner: the mean percentage contractions relative to that before exposure to IL-1β were 106.5 ± 5.8%, 81.0 ± 8.1%, 66.3 ± 7.1%, and 63.9 ± 9.2%, respectively (n = 5 each). The contractions were significantly reduced with time, but there was no significant difference between the contractions observed after 5- and 7-h exposures. Thus, in the subsequent experiment, rat aortic rings were exposed to IL-1β for 5 h.

The contractile response to KCl (30 mm) and the dose–contractile response relation for phenylephrine were obtained for each aortic ring before exposure to test agents. To determine the effects of halothane or isoflurane on IL-1β–induced inhibition of the contractile response, three to five rings obtained from the same rat were randomly tested during each condition: (1) control (without any test agent); (2) IL-1β (20 ng/ml) plus halothane (1, 2, or 3); (3) IL-1β (20 ng/ml) plus isoflurane (3%); (4) IL-1β (20 ng/ml) alone; (5) halothane (3%) alone; (6) isoflurane (3%) alone; and (7) IL-1β (20 ng/ml) plus a recombinant human IL-1 receptor antagonist (10⁻³m). The rings of the control group were left unexposed to any test agent for 5 h. After exposure to IL-1β or an anesthetic, the aortic rings were repeatedly washed with fresh bathing fluid for at least 30 min before the response to KCl and phenylephrine was obtained again. After the second response to phenylephrine had leveled off, N  ^G-l-nitro-arginine (final concentration, 3 × 10⁻⁵m) was added to the bath.

Four longitudinal aortic strips from each rat were suspended in organ baths, not under tension, and allowed to equilibrate for 90 min. One of the four strips was not exposed to any of the test agents (control), and the remaining three strips were exposed to IL-1β alone, halothane alone, or IL-1β plus halothane for 5 h.

The strips were then repeatedly washed with fresh fluid for 30 min, quick-frozen in liquid nitrogen, homogenized in 6% volume-to-volume ratio trichloroacetic acid, and subjected to ether extraction and succinylation. The cGMP and cAMP in each sample were measured using a commercial radioimmunoassay kit (¹²⁵I kit;Yamasa, Chiba, Japan). The strips used for the time-course measurement were frozen quickly in liquid nitrogen after exposure to agents for 0, 1, 3, or 5 h, then the cGMP and cAMP contents were measured in a same method as described previously. All rings were treated with indomethacin (10⁻⁵m) throughout the experiments.

To prepare the cDNA probe, the total RNA (2 μg) from the aorta isolated from a rat treated with lipopolysaccharide (0555:B5, Sigma Chemical, St. Louis, MO; 20 mg/kg) for 4 h was reverse-transcribed by incubation with 1 μm oligo(dT)15 primer, 3 mm MgCl, 400 U murine Moloney leukemia virus reverse transcriptase, 500 μm dNTP, 0.01 mm dithiothreitol, 75 mm KCl, and 50 mm Tris-HCl, pH 8.3, in a final volume of 20 μl at 37°C for 1 h.

Polymerase chain reaction primers for rat iNOS were synthesized according to the oligonucleotide sequences corresponding to amino acid residues 944–951 and 1130–1136: GATCAATAACCTGAAGCCCG and GCCCTTTTTTGCTCCATAGG. 16The cDNA was amplified by 40 cycles of polymerase chain reaction in a 40-μl reaction volume containing reverse-transcribed cDNA generated from 100 ng of total RNA, 1.25 U of Taq DNA polymerase, and each primer at 0.5 μm denaturing at 94°C for 1 min, annealing at 60°C for 1.5 min, and extension at 72°C for 2 min. The polymerase chain reaction product of the expected size (577 bp) was used as a cDNA probe for iNOS.

Total RNA for Northern blot hybridization was prepared from aortic strips that were exposed to (1) IL-1β (20 ng/ml), (2) IL-1β (20 ng/ml) plus halothane (3%), or (3) no test agent (control) for 5 h by the acid phenol method. 17Samples of 20 μg total RNA were fractionated on a 1.2% agarose gel and transfered to a nylon membrane (Hybond N, Amersham Co., Tokyo, Japan). The cDNA probes for rat iNOS and human 18S ribosomal RNA were labelled with dCTP by nick translation or random-primer labeling, respectively. Hybridization was conducted for approximately 48 h at 42°C. The membrane was washed under high stringency conditions (2 × 15 min with 4 × sodium citrate–sodium chloride buffer [SSC] at room temperature and 1 × 60 min with 0.3 × SSC, 0.1% sodium dodecyl sulphate at 52°C), the radioactivity was measured by an imaging analyzer (BAS2000; Fuji Photo Film Co. Ltd. Tokyo, Japan), and the relative content of iNOS mRNA to ribosomal RNA was calculated.

To determine whether the effect of halothane on iNOS mRNA expression is specific, inducible-type cyclooxygenase (cyclooxygenase 2 [COX-2]) mRNA was investigated by means of Northern blot hybridization with regard to the effect of halothane on COX-2 mRNA expression. Probes used for Northern analysis were a 1 kb 3′ end rat COX-2 cDNA. Northern blot analysis was conducted as previously described, and the radioactivity was then measured by an imaging analyzer, and the relative content of COX-2 mRNA to β actin was calculated.

The protein fractions of homogenates from aortic tissues that were exposed to (1) IL-1β (20 ng/ml), (2) IL-1β (20 ng/ml) plus halothane (3%), or (3) no test agent (control) for 5 h were separated on denatuturing 7.5% polyacrylamide gels, followed by blotting onto polyvinilidene difluoride membranes. The subsequent steps were performed using a commercial kit according to the manufacturer’s instructions (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD). The antibody used in this experiment was anti-macNOS pAb (Transduction Laboratories, Lexington, KY). Signals were visualized by the horseradish peroxidase–tetramethylbenzidine reaction. These signals were measured by imaging analyzer. Then the density was expressed as a percentage relative to that of the IL-1β group.

The magnitude of the contractions observed after exposure to IL-1β or the test agents were expressed as a percentage relative to the maximal contraction (phenylephrine, 10⁻⁵m; KCl, 30 mm) that was obtained before the exposure. The data are expressed as mean ± SEM. Data from the study were evaluated by analysis of variance for repeated measurements follower by the Scheffé F test. Responses that differed from control values at P < 0.05 were considered statistically significant. All statistical analyses were performed using Statview II (Abacus Concepts, Inc., Berkeley, CA) on an Apple Macintosh computer (Apple Computer, Inc., Cupertino, CA). In this study, n represents the number of animals, not rings.

All drugs used in this study were obtained from Sigma Chemical Co. (St. Louis, MO) unless specified as follows. Human recombinant IL-1β was gift from Otsuka Pharmaceuticals (Tokushima, Japan). Human IL-1 receptor antagonist, halothane, and isoflurane were obtained from R & D Systems Inc. (Minneapolis, MN), Hoechst Japan Limited (Osaka, Japan), and Dainabot (Osaka, Japan), respectively. Indomethacin was dissolved in 99% vol/vol ethanol, and all other chemicals were dissolved in distilled water.

Exposure of aortic rings without endothelium to IL-1β for 5 h significantly attenuated the contractile responses to phenylephrine and KCl, and the dose–contractile response curve for phenylephrine was shifted to the right and downward (fig. 1). The maximal contraction caused by phenylephrine (final concentration, 10⁻⁵m) was significantly reduced from 2,823 ± 182 to 1,784 ± 367 mg (fig. 1A; n = 7;P < 0.01). EC⁵⁰in the IL-1β groups (2.7 ± 1.2 × 10⁻⁷m) was higher (P < 0.01) than that in control (1.7 ± 1.1 × 10⁻⁸m). The addition of N  ^G-l-nitro-arginine (final concentration, 3 × 10⁻⁵m) completely restored the contractions in response to phenylephrine that had been inhibited by previous IL-1β exposure (fig. 1A).

Contractions caused by 30 mm KCl were also significantly reduced from 112.7 ± 5.4 to 38.8 ± 6.1% (fig. 2; n = 7;P < 0.01). IL-1β exposure attenuated the contractile responses evoked by both phenylephrine and KCl, suggesting that the effects of IL-β on contractility were independent on contractile mechanism.

Exposure to IL-1β and a recombinant human IL-1 receptor antagonist (10 μg/ml) completely abolished the IL-1β–induced inhibition of contractile responses to phenylephrine (n = 3).

The contractions caused by phenylephrine and KCl after the rings had been exposed to halothane (3%) or isoflurane (3%) alone for 5 h were comparable to those of the control rings, which had not been exposed to any test agents. Simultaneous exposure to halothane (1.0–3.0%) plus IL-1β (20 ng/ml) significantly reduced the IL-1β–induced inhibition of the contractile response to phenylephrine in a concentration-dependent manner (fig. 1). The addition of N  ^G-l-nitro-arginine (3 × 10⁻⁵m) did not affect the contractions of the control rings or of rings that had been exposed to halothane or isoflurane but not IL-1β. The IL-1β–induced inhibition of contraction caused by 30 mm KCl was also significantly reduced by simultaneous exposure to halothane (1.0–3.0%;fig. 2;P < 0.05; n = 7).

Simultaneous exposure to isoflurane (3.0%) and IL-1β (20 ng/ml) had no effect on the IL-1β–induced inhibition of the contractile response to phenylephrine (fig. 3; n = 7).

Exposure of the endothelium-denuded rat aorta to IL-1β (20 ng/ml) for 5 h caused an increase in the tissue cGMP contents in an exposure time–dependent manner. The concentrations of cGMP measured in aortic strips that were exposed to halothane plus IL-1β were significantly lower than those of strips exposed to IL-1β alone (fig. 4A). Exposure of endothelium-denuded rat aorta to halothane alone had no effects on the cGMP contents during the exposing period (fig. 4B).

The cAMP contents of the IL-1β (20 ng/ml) group were slightly higher than those of control. The cAMP contents of the IL-1β (20 ng/ml) plus halothane (3%) group were almost same as those of the IL-1β group. Exposure to halothane (3%) did not alter the cAMP accumulation during the exposure period (1, 3, and 5 h;fig. 4C).

Northern blot hybridization using the synthesized cDNA probes for rat iNOS showed a single band corresponding to the size of iNOS mRNA (approximately 4.4 kb, fig. 5). No hybridization signal for iNOS mRNA was observed in the control aortae by Northern blot hybridization. Exposure to IL-1β for 5 h induced the expression of iNOS mRNA. Exposure to IL-1β in combination with halothane attenuated the expression of iNOS mRNA.

Northern blot hybridization using the synthesized cDNA probes for rat COX-2 showed a single band corresponding to the size of COX-2 mRNA (approximately 1 kb;fig. 6). Halothane inhibited the COX-2 mRNA expression that was evoked by IL-1β.

The antiserum reacted exclusively with a protein band corresponding to approximately 130 kd in homogenates of rat aortic tissue treated with IL-1β alone. The signal produced after stimulation of the rat aorta with IL-1β plus halothane was greatly attenuated compared with that produced when stimulated with IL-1β alone. No band was detected in homogenates from nonstimulated tissues (fig. 7).

The current study has demonstrated that the exposing of rat aortic rings without endothelium to IL-1β for 5 h results in hyporesponsiveness to vasoactive agents (phenylephrine and KCl) in association with an increase in the tissue cGMP content and increased the expression of iNOS mRNA and iNOS protein, in agreement with previous reports. 4,18These results suggest that hyporesponsiveness, independent on the difference in the contraction mechanism, was a consequence of IL-1β–induced iNOS expression after the accumulation of cGMP content. The current study has also clearly demonstrated that halothane, but not isoflurane, significantly inhibited the vascular response induced by IL-1β. Halothane also suppressed the accumulation of cGMP content and the expression of iNOS mRNA and iNOS protein induced by IL-1β in rat vascular smooth muscle. In addition, IL-1β also caused the increase of the expression of COX-2 mRNA in rat VSM, and halothane suppressed this expression, suggesting that the effect of halothane on mRNA expression is not specific for iNOS.

The effects of volatile anesthetics, including halothane and isoflurane, on vasorelaxation mediated via  the NO–cGMP pathway have been the subjects of several investigations recently. Several mechanisms by which halothane may inhibit acetylcholine- or bradykinin-induced vasorelaxation mediated via  endothelium-derived NO have been proposed. 19In the current study, these sites of action are not involved in halothane-induced inhibition, because the endothelium was removed from the aortic rings. On the other hand, halothane reportedly induces vasorelaxation, which is mediated via  an increase in the tissue cGMP content. Halothane at concentrations more than 2.0% has been shown to increase the cGMP content of isolated rat aorta 20and of dog middle cerebral artery. 21However, this effect of halothane was not observed in the current study, in which halothane antagonized the IL-1β–induced increase in the cGMP content.

In addition to acting on the endothelium, several reports have suggested that halothane affects the soluble guanylyl cyclase within the VSM. Hart et al.  22demonstrated that halothane inhibits the soluble guanylyl cyclase that is present in VSM and catalyzes the formation of cGMP from guanosine triphosphate. Although the inhibitory effect of halothane on IL-1β–induced hyporesponsiveness might be a result of this effect of halothane on soluble guanyly cyclase, in our study the contractile responses to phenylephrine or KCl of aortic rings exposed to halothane alone were comparable to that of the control rings, and exposure to halothane alone had no effect on cGMP content. Thus, it is unlikely that the inhibitory effect of halothane on the IL-1β–induced hyporesponsiveness of the VSM and the increase in the cGMP content may be a result of its inhibitory action on the soluble guanylyl cyclase activity of VSM. It is suggested that the inhibitory effect of halothane on the expression of iNOS mRNA and iNOS protein may cause the following inhibition of IL-1β–induced hyporesponsiveness.

It has been reported that cAMP plays a major role as a second messenger in the IL-1β–evoked induction of iNOS mRNA in cultured VSM cells. 11Activation of IL-1 receptor leads to stimulation of AC, with subsequent activation of cAMP–protein kinase A pathway. The stimulative effect of cAMP on iNOS induction in VSM cells could occur at the level of nuclear factor κB and guanosine triphosphate:cyclohydrolase activity. IL-1β can also cause COX-2 induction and after the increase of cAMP content. Because our cyclic nucleotides measurements were performed in the presence of indomethacin, cAMP contents measured in this study were not consequences of induction of COX-2. Our cyclic nucleotides measurements revealed that IL-1β increased cAMP contents in VSM of rat aortic tissues. Furthermore, simultaneous exposure to IL-1β plus forskolin (an AC stimulator) enhanced the IL-1β–induced inhibition of the contractile response of the aortic rings to phenylephrine, and exposure to IL-1β plus 2′,3′-dideoxyadenosine (an AC inhibitor) attenuated the IL-1β–induced inhibition (data not shown). From these findings, it is suggested that cAMP might contribute to iNOS induction evoked by IL-1β even in the VSM of rat aortic tissues.

There has been no report that halothane affects the accumulation of cAMP contents induced by IL-1β in VSM cells of rat aortic tissues. In the current study, cAMP contents of aortic tissues exposed to halothane alone were similar to those of control group during the exposure period. Furthermore, cAMP contents of the halothane plus IL-1β group were also similar to those of the IL-1β group. These results suggest that halothane might not have any effects on the cAMP accumulation in the VSM cells of rat aortic tissues. Thus, it is unlikely that halothane interferes with the iNOS induction by inhibiting the pathway through which AC–cAMP was associated.

Although the mechanism by which halothane down-regulates iNOS expression remains unknown, it may be mediated through the alteration in the rate of transcription or stability of iNOS mRNA. In the promoter region of iNOS, PKC responsive cis  elements, such as binding site of nuclear factor κB, which is activated either by direct phosphorylation or by the phosphorylation of associated regulatory proteins, are present. Activation of PKC has been reported to be of no consequence, 18to up-regulate, 23or to down-regulate 24the cytokine-induced increase in iNOS gene expression in VSM cells. Halothane has been shown to activate PKC 25; therefore, if PKC down-regulates the cytokine-induced increase in iNOS gene expression, this may be one candidate for the mechanism by which halothane down-regulates iNOS gene expression.

Zuo and Johns 26have reported that halothane and isoflurane, at final concentrations of 2%, up-regulate the iNOS expression stimulated by lipopolysaccharide in cultured macrophages. Otherwise, our Northern blot hybridization experiment revealed that halothane significantly inhibited the iNOS mRNA expression induced by IL-1β. The reasons for these differences between our results and those of Zuo and Johns are not apparent; however, differences concerning iNOS induction between cultured cells and VSM tissues may explain this discrepancy. It has been reported that cycloheximide, a protein synthesis inhibitor, up-regulates iNOS mRNA expression evoked by lipopolysaccharide plus interferon γ in cultured VSM cells, whereas in VSM tissues cycloheximide attenuates the induction. 27The cycloheximide-dependent induction has also been demonstrated in mouse peritoneal macrophages. 28This indicates that the induction of iNOS mRNA in VSM tissues requires de novo  protein synthesis, proteins or transcription factors that seems to be constitutively expressed in VSM cells in culture. Furthermore, in this study, halothane suppressed COX-2 mRNA expression, which was similar to the effect of halothane on iNOS mRNA expression, suggesting that the effects of halothane on mRNA may not be specific to iNOS. From our results, together with those of Zuo and Johns 26and concomitant reports concerning cycloheximide, the effect of halothane on mRNA expression appears to be similar to those of cycloheximide as a protein synthesis inhibitor. In addition, Ghantous et al.  8demonstrated that halothane inhibits protein synthesis in guinea pig liver slices. This protein synthesis inhibitor-like effect of halothane might be one of the possible mechanisms for these mRNA suppressions. Further studies will determine the exact site of action or mechanism of halothane in the mRNA expression pathway.

In the current study, isoflurane failed to affect the IL-1β–induced hyporesponsiveness to contractile agents, which was contrary to the effects of halothane. The vascular effects of volatile anesthetics have been demonstrated to vary depending on the anesthetics used. Halothane in vivo  does not change systemic vascular resistance, whereas isoflurane and sevoflurane decrease it. 29In addition, the in vitro  effects of these anesthetics on isolated vascular preparations differ from each other. Halothane relaxes isolated porcine coronary artery to a greater degree than isoflurane. 30It has been demonstrated that isoflurane is also capable of inhibiting protein synthesis in guinea pig liver slices, but to a lesser extent than halothane. 9Assuming that the effects of halothane on iNOS might be caused by protein synthesis inhibition, it is not surprising that isoflurane failed to affect the IL-1β–induced hyporesponsiveness, since even halothane, which seems to be more potent in the protein synthesis inhibition than isoflurane, partially reversed the hyporesponsiveness.

In summary, we demonstrated that halothane suppressed the expression of iNOS mRNA and iNOS protein, then inhibited the hyporesponsiveness to vasoactive agents in association with an increase in the tissue cGMP content evoked by IL-1β exposure in VSM tissues. These results suggest that halothane affects the physiologic reaction mediated by the NO-signaling pathway. Determining the mechanism underlying halothane’s inhibition of iNOS pathway would provide the contribution for understanding the mechanisms of inhalational anesthetics, which remain unclear even in the clinical situation.