Inhalational Anesthetic Effects on Rat Cerebellar Nitric Oxide and Cyclic Guanosine Monophosphate Production

Superoxide dismutase and catalase were obtained from Sigma Chemical Company (St. Louis, MO). L-[sup 3 H]arginine (55 Ci/mmol) was purchased from Amersham (Arlington Heights, IL). Dowex AG50W-X8 cation exchange resin and Bio-Rad protein assay reagent were obtained from Bio-Rad Laboratories (Richmond, CA).

Sprague-Dawley rats were killed by cervical disarticulation and their cerebella were rapidly removed. The methods and procedures of this study were approved by the animal research committee of the University of Virginia. Nitric oxide synthase activity and cGMP accumulation in rat cerebellum slices were determined as described previously. [25]Briefly, thin slices (0.4 mm) were cut from the cerebella of 10-day-old rats in both sagittal and coronal planes using a McIlwain tissue chopper. The slices were incubated for 1 h in Krebs-Henseleit buffer containing 118 mM NaCl, 4.7 mM KCl, 2 mM CaCl², 1.2 mM MgSO⁴, 1.2 mM KH²PO⁴, 25 mM NaHCO³, and 11 mM glucose (Buffer A) continuously supplied with 95% oxygen/5% carbon dioxide at 37 degrees Celsius. Nitric oxide synthase activity was determined by measuring the conversion of L-[sup 3 H]cbarginine to L-[sup 3 H]citrulline in a 4-ml gas-tight vial capped with a Teflon septum. Supplying the vial with a specific gas mixture was achieved by temporarily inserting a 22-gauge needle for inlet and a similar needle for outlet at will. Buffer A (530 micro liter) equilibrated with 95% oxygen/5% carbon dioxide was injected into the vial and supplied with 95% oxygen/5% carbon dioxide for 2 min. A small volume (40 micro liter) of the gravity-packed cerebellum slices was injected into the vial, and a humidified 95% oxygen/5% carbon dioxide mixture with or without the anesthetic agent was passed through the vial for 4 min. After incubating the reaction mixture with or without the anesthetic agent for 15 min at 37 degrees Celsius, a 30-micro liter solution containing L-[sup 3 H]arginine (3 micro Ci) with buffer A or NMDA was injected. The amount of cerebellum slices and L-[sup 3 H]arginine used in these experiments ranged from 30–40 micro liter and 1.5–3 micro Ci, respectively. The reaction was stopped after 15 min by adding 400 micro liter ice-cold buffer A containing 10 mM arginine and 10 mM ethylene diaminetetraacetic acid. The mixture was transferred to an Eppendorf tube and centrifuged at 10,000g at 4 degrees Celsius. The supernatant was removed and 1 ml of 1 M trichloroacetic acid was added to the pellet. The sample was sonicated for 3 x 15 s and 50 micro liter aliquots were saved for protein determination. The sample was centrifuged and 0.85 ml supernatant was transferred to a 5-ml glass vial. Trichloroacetic acid in the supernatant was extracted three times with 2 ml diethyl ether. The sample was freeze-dried and then reconstituted with 1 ml of 20 mM HEPES, pH 5.5, containing 2 mM ethylene diaminetetraacetic acid. The recovery of L-[sup 3 H]-citrulline in this above procedure was 89%. The sample was applied to a 1-ml Dowex AG50WX-8 (Na sup + form) column and L-[sup 3 H]-citrulline eluted with 1.5 ml water. The radioactivity of the eluate was determined by liquid scintillation counting at 49% efficiency.

To measure cGMP, cerebellum slices equilibrated with a 95% oxygen/5% carbon dioxide mixture were supplied with gas with or without the anesthetic agent for 4 min. After 15 min, cerebellum slices were stimulated with buffer or NMDA at different concentrations. After 3 min, the reaction was stopped by adding 0.4 ml of 0.25 M HCl. The sample was sonicated for 3 x 15 s and 100 micro liter samples were separated in aliquots for protein determination. The rest of the volume in each sample was centrifuged and cGMP in the supernatant was analyzed by radioimmunoassay. [26] 

Whether anesthetics alter NO synthase activity by interfering with L-[sup 3 H]arginine uptake was determined because L-[sup 3 H]arginine uptake was the first step in measuring NO synthase activity. L-[sup 3 H]arginine uptake was performed as described for NO synthase activity measurement. The uptake of L-[sup 3 H]arginine by cerebellum slices was stopped by adding 0.4 ml ice-cold buffer A containing 10 mM arginine and 10 mM ethylene diaminetetraacetic acid. The sample was centrifuged and the supernatant was discarded. To the pellet, 1 ml distilled water was added and sonicated for 3 x 15 s. Samples were analyzed for radioactivity and protein content.

For protein determination, the sonicate was hydrolyzed in 0.5 M NaOH at 60 degrees Celsius for 30 min and then reacted with the Bio-Rad protein reagent. Bovine serum albumin was used as the standard. Anesthetic concentration in the reaction sample buffer was determined as described previously. [16] 

Data were expressed as the means +/- SEM. Paired t tests were used to compare control values with anesthetic at each concentration of NMDA. Comparisons between groups were made using one-way analysis of variance with means testing by Bonferroni t test when appropriate. Differences were considered significant at P < 0.05.

The effect of halothane and isoflurane on NO synthase activity was assessed by measuring the conversion of L-[sup 3 H]arginine to L-[sup 3 H]citrulline. L-[sup 3 H]citrulline production of the control was 23,994 +/- 4,312 cpm/mg protein (100%). Halothane significantly increased the control NO synthase activity at 1.5% but not at 3% concentration (Figure 1(A)). Nitric oxide synthase activity stimulated by NMDA (0.5 M) was 62 +/- 6% greater than the control, which was not significantly altered by halothane at 1.5% and 3%(Figure 1(A)). Isoflurane at 1.5% and 3% significantly enhanced NMDA-stimulated NO synthase activity, with no effect on the control (Figure 1(B)).

Whether anesthetics alter NO synthase activity by interfering with L-[sup 3 H]arginine uptake by cerebellum slices was determined. L-[sup 3 H]arginine uptake by control and NMDA-stimulated conditions was not significantly altered by either halothane or isoflurane at 1.5% and 3% concentrations (Table 1).

The effect of different concentrations of halothane and isoflurane on cGMP accumulation in cerebellum slices was determined in the absence or presence of various concentrations of NMDA. Basal cGMP concentration was not significantly altered by either halothane or isoflurane at 0.75%, 1.5%, or 3% concentration (data not shown). However, cGMP production stimulated by NMDA (50 micro Meter, 100 micro Meter, and 500 micro Meter) was significantly attenuated by halothane at 1.5% and 3%(Figure 2(B and C)) but not at 0.75%(Figure 2(A)). Isoflurane at 0.75% did not produce any significant effect on cGMP production at different NMDA concentrations (Figure 3(A)). A significant inhibition of cGMP production was observed at 1.5% isoflurane when stimulated with 50 micro Meter but not with 100 micro Meter or 500 micro Meter NMDA (Figure 3(B)). Isoflurane at 3% produced a significant attenuation of cGMP accumulation at all concentrations of NMDA tested (Figure 3(C)).

To determine whether superoxide anion is involved in the observed anesthetic responses, the effect of a mixture of superoxide dismutase and catalase on cGMP accumulation was examined. A mixture of superoxide dismutase and catalase (100 units each per milliliter) caused no significant effect on anesthetic inhibition of NMDA-stimulated cGMP concentration (Table 2). Studies by Martin et al. [27]have suggested that glycine, a coagonist of the NMDA receptor, attenuated the inhibitory effect of anesthetics on NMDA-induced ion channel opening. Glycine (100 micro Meter), when tested on the anesthetic inhibition of NMDA-induced cGMP accumulation but did not produce any significant effect, although cGMP accumulation by NMDA stimulation was significantly increased (Table 3).

In this study, we investigated the effects of inhalational anesthetics on NMDA-stimulated NO and cGMP production in rat cerebellum slices. Halothane and isoflurane inhibited NMDA-induced cGMP content of cerebellum slices. Unlike previous studies showing a lack of effect of halothane and isoflurane on neuronal NO synthase, the current studies investigated the action using isolated brain slices with intact cellular milieu. Despite their inhibitory action on NMDA-stimulated cGMP accumulation in cerebellum slices, isoflurane significantly enhanced NMDA-stimulated NO synthase activity, whereas halothane produced no significant effect. This study presents new insights into the mechanism of inhalational anesthetic action on NMDA-stimulated NO synthase activity in the rat cerebellum.

First, the mechanism of inhalational anesthetic action on NO synthase activity in different tissues is controversial. [15,16,19,23,24]Contrary to our expectation, in this study, isoflurane enhanced the activity of NMDA-stimulated NO synthase in cerebellum slices, whereas halothane produced no significant effect. The effect of isoflurane on NO synthase activity might be due to a direct action on NO synthase, as reported previously. [15]However, recent studies show that isoflurane or other inhalational anesthetics do not alter isolated NO synthase activity, excluding a direct interaction of anesthetics with NO synthase. [16–18]Then how can isoflurane enhance the activity of NO synthase in cerebellum slices? Isoflurane may increase the cytosolic free calcium ([Ca sup 2+]i) and enhance NMDA-stimulated NO synthase activity in the intact cerebellum. Anesthetic agents have been shown to increase the [Ca²+]i in unstimulated cells [28,29]and to inhibit receptor activated increase in [Ca²+]i. [30,31]Isoflurane has been shown to reduce glutamate and NMDA-mediated calcium fluxes in brain slices by approximately 60% and thereby delay cellular injury induced by simulated ischemia. [32]Recently Miao et al. [33]observed that both halothane and isoflurane decreased the intrasynaptosomal calcium level, presumably by inhibiting calcium entry through specific neuronal calcium channels. Isoflurane may interact with other cofactors of NO synthase and enhance NMDA-stimulated NO synthase activity. The isoflurane enhancement of NMDA-stimulated NO synthase activity is supported by a recent report in which isoflurane increased the cGMP content of primary cultures of cerebral neurons stimulated with NMDA, presumably through increased production of NO. [34] 

Another possibility is that isoflurane may interfere with L-[sup 3 H]arginine uptake by cerebellum slices and alter NO synthase activity, as assessed by the conversion of L-[sup 3 H]arginine to L-[sup 3 H]citrulline. In our study, isoflurane showed no significant effect on L-[sup 3 H]arginine uptake by cerebellum slices under control and NMDA-stimulated conditions while increasing the NMDA stimulated L-[sup 3 H]citrulline production. This suggests that the isoflurane enhancement of NO synthase activity is not due to an increase in L-[sup 3 H]arginine uptake and that L-[sup 3 H]citrulline produced inside the tissue is not extruded. One other possibility is that isoflurane may increase the concentration of L-[sup 3 H]citrulline without activating NO synthase. In NO-producing cellular systems, including the brain and endothelium, L-citrulline formed by the action of NO synthase activity is recycled to L-arginine by argininosuccinate synthetase and argininosuccinate lyase. [35]If isoflurane inhibits the L-citrulline recycling pathway, then it is likely to increase the concentration of L-citrulline in the cell with no effect on NO synthase activity. It remains to be determined whether isoflurane interferes with the L-citrulline recycling pathway in the brain.

Second, both halothane and isoflurane attenuated NMDA-stimulated cGMP accumulation in cerebellum slices. These data correspond with our recent report showing anesthetic inhibition of NMDA-stimulated GMP production in the rat brain by an autoradiographic method [36]and with several other studies that investigated the anesthetic effect on cGMP production in the brain and vascular smooth muscle. [7,8,20,24]It is possible that anesthetics may decrease cGMP concentration in brain slices by activating phosphodiesterases. Although there is no literature reporting an effect of anesthetics on specific phosphodiesterases that hydrolyze cGMP, several studies show that anesthetics inhibit or produce no significant effect on phosphodiesterase activity of the brain and other sources. [21,37,38]The inhibitory effect of anesthetics on phosphodiesterases is further supported by studies showing that anesthetics act as antagonists of calmodulin, [39,40]which is an activator of cGMP phosphodiesterases. Anesthetic inhibition of a receptor-mediated increase in [Ca²+]i appears to be an important mechanism for the decrease in cGMP production. Puil et al. [30]have reported that isoflurane and halothane attenuated a glutamate-stimulated increase in [Ca²+]i in cultured hippocampal neurons. Similarly, volatile anesthetics have been shown to depress evoked transmitter release by reducing calcium entry. [31]A recent report shows that isoflurane increases the cGMP content in primary cultures of cerebral neurons in response to NMDA stimulation. [34]This discrepancy may be due to the difference between cerebellum slices and cultured cerebral neurons used in these studies, although we observe a decrease in cGMP concentration in cultured hippocampal neurons (Z Zuo and RA Johns, unpublished data).

Nakamura et al. [41]have suggested that isoflurane attenuates endothelium-dependent relaxation by inhibiting the production of NO, whereas halothane inhibits the action of NO in vascular smooth muscle. [41]Further studies in rat cerebellar slices suggested that isoflurane may interact with receptor sites. [20]Terasako et al. [20]observed that both halothane and isoflurane at 2% inhibited NMDA-stimulated cGMP production in rat cerebellar slices to a similar degree, whereas halothane but not isoflurane significantly suppressed the D-aspartate-induced cGMP production. Because D-aspartate stimulation increases [Ca²+]i by a pathway different from NMDA receptor stimulation to activate NO synthase, they suggested that isoflurane may interact with the NMDA receptor, G protein, or coupled calcium channel. [20]In our studies, cGMP accumulation with 50 micro Meter NMDA but not with high concentrations was inhibited by 1.5% isoflurane, which was consistent with a competition between isoflurane and NMDA for the receptor. The anesthetic action at the NMDA receptor level corresponds with the work of Martin et al. [27]In their study, MK801, an NMDA receptor antagonist, binding to cerebral cortex membranes was measured as an index of the NMDA ionic channel opening. Anesthetics depressed glutamate stimulation of MK 801 binding, suggesting an interference with the NMDA receptor ion channel. Glycine, a coagonist of the NMDA receptor, partially reversed the effect of anesthetics, suggesting that anesthetics may interact with the glycine binding site. In our study, exogenous glycine did not alter anesthetic inhibition of NMDA-stimulated cGMP accumulation, suggesting that the glycine binding site is not involved in anesthetic inhibition of cGMP accumulation.

The paradox in this study is that isoflurane enhances NMDA-stimulated NO synthase activity with a simultaneous attenuation of cGMP content of cerebellum slices. An increase in NO synthase activity is expected to increase cGMP concentration in the tissue. On the contrary, there are conditions under which an increased production of NO may adversely affect the accumulation of cGMP content. In a previous study, we observed an increase in NO production as measured by the conversion of L-[sup 3 H]arginine to L-[sup 3 H]citrulline under hyperoxic conditions. [42]However, when the NO-producing enzyme system was coupled to a bioassay system consisting of vascular smooth muscle cells, the cGMP accumulation in vascular smooth muscle cells was drastically reduced under hyperoxic conditions. This suggested an involvement of superoxide in inactivating NO and subsequently decreasing the concentration of cGMP in the tissue. Whether superoxide is involved in the inhibition of cGMP accumulation in cerebellum slices is not known. Shayevitz et al. [43]reported that halothane and isoflurane increase pulmonary artery endothelial cell sensitivity to oxidant stress. They have suggested that anesthetics may produce oxidant load by enhancing the conversion of xanthine dehydrogenase to xanthine oxidase or by attenuating antioxidant systems, including glutathione reductase. This is supported by the studies of Yoshida and Okabe [44]that showed attenuation of acetylcholine-induced relaxation of canine mesenteric arteries by sevoflurane by a superoxide-mediated mechanism. In that study, the attenuation of relaxation was partially reversed by superoxide dismutase or deferoxamine, indicating that superoxide anion or hydroxyl anion was in part responsible for the effect of sevoflurane. This indicates that superoxide may be responsible for the anesthetic-induced decrease in cGMP content in brain slices. However, in our study, a mixture of superoxide dismutase and catalase (100 units each per milliliter) produced no significant effect on the anesthetic inhibition of NMDA-stimulated cGMP production, suggesting that superoxide generation may not be involved in anesthetic inhibition of NMDA-stimulated cGMP production.

Anesthetic interaction with soluble guanylyl cyclase may decrease cGMP accumulation. Studies by Terasako et al. [20]showed no significant effect of halothane and isoflurane on sodium nitroprusside-induced cGMP production in rat cerebellum. This suggested that halothane and isoflurane do not interact with soluble guanylyl cyclase. However, studies by Hart et al. [45]showed that halothane inhibited NO-induced cGMP production in denuded rat aortic rings. Thus it appears that inhalational anesthetics may interfere with NO activation of soluble guanylyl cyclase. Hart et al. [45]have suggested that inhalational anesthetics may compete with NO for binding to the heme group of soluble guanylyl cyclase. If this is true, anesthetics are expected to inhibit NO or sodium nitroprusside activation of isolated soluble guanylyl cyclase. However, we found that there was no significant effect of inhalational anesthetics on NO or sodium nitroprusside activation of isolated soluble guanylyl cyclase. [23]In addition, NO, sodium nitroprusside, and nitroglycerin-stimulated cGMP accumulation in cultured vascular smooth muscle cell or endothelial cell-vascular smooth muscle cell cocultures was not significantly altered by 0–5% halothane and isoflurane. [21]These studies suggest that the anesthetic inhibition of NO-induced cGMP accumulation may not be due to a direct competition between NO and inhalational anesthetics for binding to soluble guanylyl cyclase in the brain.

The mechanism of anesthetic inhibition of NMDA-stimulated cGMP accumulation may be due to the ability of anesthetics to interact with multiple neuronal pathways. For example anesthetics have been shown to increase the concentration of neurotransmitters, including gamma-aminobutyric acid [46]and serotonin, [47]which may feed back on the NO-cGMP pathway and alter NMDA-induced cGMP production in the cerebellum. Serotonin at low nanomolar concentrations has been shown to inhibit NMDA-stimulated cGMP production in rat cerebellum. [48]The inhibitory serotonin receptors appear to interact with NMDA receptors at both post-synaptic sites and presynaptic terminals. This suggests that inhalational anesthetics interact with other neurotransmitter pathways that modulate NMDA-induced cGMP accumulation in the cerebellum.

Although our interest in investigating the mechanism of anesthetic inhibition of the NO pathway was stimulated by our observations that inhibition of neuronal NO synthase decreases minimum alveolar concentration for anesthetics, [9,10]the possibility that inhibition of the NO pathway could account for other actions of anesthetics on neuronal physiology should be considered. Several studies suggest that NO plays a role in regulating cerebral blood flow (CBF), electroencephalogram (EEG) activity, and cerebral oxygen consumption. Administration of L-N^G-nitro-L-arginine methyl ester, an inhibitor of NO synthase, in different animal models has been shown to reduce CBF under various conditions, suggesting that NO contributes to CBF regulation. [49–51]3-morpholinosydnonimine, an NO donor, administered after 3 min of middle cerebral artery occlusion in rats, enhanced the recovery of CBF and EEG amplitude and reduced the size of the infarct. [52]Nitric oxide also appears to regulate EEG activity [52]and cerebral oxygen consumption. [53,54]If anesthetics inhibit the NO pathway, we would expect a decrease in CBF and EEG. However, inhalational anesthetics have been shown to increase CBF and decrease cerebral oxygen consumption and EEG activity. [55–57]Several studies have shown that anesthetic-induced increases in CBF are prevented by inhibitors of NO synthase, suggesting that anesthetics may actually stimulate the NO pathway. [51,58]Alternatively, these results could be explained by inhibition of basal NO unrelated to the anesthetic action. Indeed, Todd et al. [59]have shown that NO is not a primary mediator of anesthetic CBF effects. [59]Overall, these studies suggest that the anesthetic inhibition of the NO signaling pathway is not related to anesthetic effects on CBF, EEG activity, and cerebral oxygen consumption, which must be mediated by other mechanisms.

In conclusion, we have shown that isoflurane significantly enhanced NMDA-stimulated NO synthase activity, with a simultaneous decrease in cGMP accumulation in cerebellum slices. Halothane produced no significant effect on NMDA-stimulated NO synthase activity but attenuated NMDA-stimulated cGMP accumulation.

The authors thank Ninghong Da, Rama Goswami, and Kimberly Parekh for their technical assistance in measuring cGMP samples by radioimmunoassay. They also thank Dr. Zhiyi Zuo for his comments.