Effects of Targeted Neuronal Nitric Oxide Synthase Gene Disruption and Nitro sup G -L-Arginine Methylester on the Threshold for Isoflurane Anesthesia

These investigations were approved by the Subcommittee for Research Animal Care of the Massachusetts General Hospital (Boston, MA).

Fifty knockout mice and 80 wild-type mice (SV129, Taconic, MA) with approximately the same body weight (18-26 g) of both sexes were used in this study. Knockout mice were generated by homologous recombination, as described previously. [17]These homozygous knockout mice showed no evidence of nNOS gene expression when DNA, mRNA, and proteins from brain extracts were analyzed. [17]Residual brain NOS catalytic activity was less than 5% over background activity as detected by the enzymatic conversion of radiolabeled arginine to citrulline in brain homogenates. [17]This residual activity is likely to be attributable to endothelial NOS. [17-19].

Determination of Baseline Values of Minimum Alveolar Concentration and the RRED⁵⁰in Wild-type and Knockout Mice. Baseline values of minimum alveolar concentration and RRED⁵⁰of isoflurane were established according to the methods described by Eger et al. [20-22]Mice were placed in individual Plexiglas chambers (25 cm long and 5.0 cm in diameter) for the determination of minimum alveolar concentration values. Each chamber was fitted with a rubber stopper at one end through which the mouse's tail and a rectal temperature probe protruded. Groups of eight mice were given isoflurane in oxygen (4 l/min total gas flow). A gas sample was continuously drawn from the expiratory limb of the circuit, and the anesthetic concentration was measured with an infrared analyzer (Datex Ultima, Helsinki, Finland). A rectal temperature probe was inserted under light general anesthesia, and temperature was monitored in each mouse. The rectal temperature was kept between 36.5 degrees Celsius and 38.0 degrees Celsius with heat lamps. Mice initially breathed approximately 1.5 vol% isoflurane for 60 min. A clamp (alligator clip) was applied to the tail for 1 min, and the mice were observed for movement in response to the stimulation. In every case, the tail was stimulated proximal to the previous test site. Only the middle third of the tail was used for tail-clamping. The concentration of the anesthetic agent at which the mouse exhibited motor activity (gross movements of the head, extremities, and/or body) was considered one that permitted a positive motor response. The anesthetic concentration was increased (or decreased) in steps of 0.12-0.15% until the positive response disappeared (or vice versa), with 20 min for equilibration allowed after each change of anesthetic concentration. Minimum alveolar concentration is defined as the concentration midway between the highest concentration that permitted movement in response to the stimulus and the lowest concentration that prevented movement. A typical minimum alveolar concentration study of a group of eight mice took approximately 4-5 h, including the initial equilibration period.

For the determination of baseline RRED⁵⁰, unrestrained mice were placed in individual wire-mesh cages rotated at 4 rpm in a 20-1 plastic chamber. [20-22]In one additional restrained mouse in the same chamber, rectal temperature was monitored and maintained as above. Controlled amounts of volatile anesthetic were delivered by temperature-compensated vaporizers in a flow of oxygen (4 l/min) and passed continuously through the chamber during the entire experimental period. After 60 min of equilibration, the mice were subjected to five complete turns of the rotator. Animals that rolled over two or more times were considered anesthetized. The concentration of the anesthetic agent was increased or decreased in 0.12-0.15% steps from the initial concentration that we tested, and the mice were reexamined for their ability to right themselves after a 30-min equilibration period. These changes and subsequent equilibration periods were continued until a sufficiently high anesthetic concentration was provided in which all animals failed the righting-reflex test and then until a sufficiently low concentration was provided in which all the animals righted themselves. RRED⁵⁰was calculated for each mouse as the mean value of the anesthetic concentrations that just permitted and just prevented the righting-reflex. A typical righting-reflex study of a group of seven mice took approximately 3-4 h, including the initial equilibration period.

Influence of Acute L-NAME on Minimum Alveolar Concentration and RRED⁵⁰in Wild-type and Knockout Mice. We administered L-NAME at 10, 25, 50, 75, or 100 mg/kg as an intraperitoneal injection to wild-type and knockout mice to study the influence of L-NAME on the minimum alveolar concentration and RRED⁵⁰of isoflurane. Seven to nine mice were studied at each L-NAME dose. We chose these doses and the number of mice based upon prior studies. [23-25]L-NAME was dissolved in 0.2 ml of normal saline. L-NAME or placebo (0.2 ml of normal saline) was injected 15 min before commencing each study, and minimum alveolar concentration and RRED⁵⁰were determined in each group of mice as described above. As a separate set of experiments, we examined stereospecificity of the effects of L-NAME by administering L-arginine or D-arginine. After minimum alveolar concentration and RRED⁵⁰were determined in a group of mice pretreated with the most potent dose of L-NAME (50 mg/kg), either L-arginine (600 mg/kg) or D-arginine (600 mg/kg) were dissolved in 0.2 ml of normal saline and injected intraperitoneally. Minimum alveolar concentration and RRED⁵⁰were redetermined starting 15 min later. As an additional control experiment, the effects of ketamine on isoflurane minimum alveolar concentration in knockout mice were examined. Ketamine (50 mg/kg) was given as an intraperitoneal injection to eight knockout mice, and isoflurane minimum alveolar concentration was determined as above.

Influence of Week-long L-NAME Administration on Minimum Alveolar Concentration and Righting Reflex in Wild-type Mice. L-NAME (50 mg/kg in 0.2 ml of normal saline) was given to 20 wild-type mice by oral gavage every 12 h for 7 days followed on the 8th day by intraperitoneal L-NAME (50 mg/kg). [26]We determined RRED⁵⁰and minimum alveolar concentration before and after giving intraperitoneal L-NAME on the 8th day. We gave the same volume of normal saline (placebo) by gavage to another group of 20 wild-type mice, and their minimum alveolar concentration and RRED⁵⁰were determined. We chose the gavage route of administration because it has been used previously to deliver L-NAME on a chronic basis. [26]This portion of the study was done without the investigators knowing the treatment status of the animals.

Blood Gas Analysis. To rule out the presence of hypoxia, hypercapnia, and acidosis during these experiments, we sampled arterial blood from four mice in each group (wild-type mice, wild-type mice treated with L-NAME (50 mg/kg), and nNOS knockout mice) by percutaneous left ventricular puncture at the end of the minimum alveolar concentration experiments.

All data are presented as mean plus/minus SE. The data were analyzed using an analysis of variance followed by Dunnett's multiple comparison tests or an unpaired t-test. P < 0.05 was the criterion for statistical significance.

The baseline values of isoflurane minimum alveolar concentration and RRED⁵⁰of nNOS knockout mice did not differ from those of wild-type mice (minimum alveolar concentration (vol %): knockout 1.24 plus/minus 0.05, wild-type 1.24 plus/minus 0.05; RRED⁵⁰(vol %): knockout 0.56 plus/minus 0.01, wild-type 0.58 plus/minus 0.01; Table 1). The administration of 10, 25, 50, 75, and 100 mg/kg intraperitoneal L-NAME caused a decrease from isoflurane baseline minimum alveolar concentration of wild-type mice of 5.81 plus/minus 5.3%, 28.3 plus/minus 3.1% (P < 0.01), 24.0 plus/minus 4.3% (P < 0.01), 19.4 plus/minus 3.8% (P < 0.05), and 17.7 plus/minus 6.0% (P < 0.05), respectively (Table 2). The isoflurane RRED⁵⁰of wild-type mice also was reduced by 10, 25, 50, 75, and 100 mg/kg intraperitoneal L-NAME from its baseline value by -3.82 plus/minus 5.2% (increased), 19.6 plus/minus 7.1% (P < 0.01), 37.3 plus/minus 5.4% (P < 0.01), 32.3 plus/minus 5.1% (P < 0.01), and 29.2 plus/minus 6.1% (P < 0.01), respectively (Table 2). However, the responses of knockout mice to both tests were not altered by L-NAME injections at any of the L-NAME doses we examined (Table 2). In contrast, the administration of intraperitoneal ketamine (50 mg/kg) caused a significant reduction of the isoflurane minimum alveolar concentration value of knockout mice from its baseline value of 1.24 plus/minus 0.05 to 0.96 plus/minus 0.04 (vol%). No apparent increase of morbidity or mortality was noted in mice treated with intraperitoneal L-NAME at any dose. The reduction of minimum alveolar concentration and RRED⁵⁰in wild-type mice treated with 50 mg/kg L-NAME was completely reversed by an injection of 600 mg/kg intraperitoneal L-arginine, but not by administering D-arginine (Table 3).

After 7 days of either L-NAME or placebo (normal saline) gavage-feeding, minimum alveolar concentration and RRED⁵⁰of wild-type mice did not differ from their baseline values (Table 4). When placebo-gavaged mice were given L-NAME (50 mg/kg) intraperitoneal injections on the 8th day, there was a significant reduction of minimum alveolar concentration and RRED⁵⁰by 24.0 plus/minus 3.2% (P < 0.001) and 26.2 plus/minus 7.4% (P < 0.01), respectively, but not in the week-long L-NAME-gavaged mice (Table 4). During week-long L-NAME gavage-feeding, 4 of 20 mice died in the L-NAME-gavaged group of undetermined causes. No untoward effects were observed in the placebo-gavaged group.

The results of blood gas analyses are shown in Table 5. No significant hypoxia, hypercapnia, or acidosis was present in any of the mice. No significant differences were noted among groups of these measured indices.

Our study revealed that targeted disruption and inactivation of the nNOS gene in mice did not alter their sensitivity to isoflurane anesthesia. Mice that were congenitally deficient in nNOS had an isoflurane minimum alveolar concentration and RRED⁵⁰identical to that of wild-type mice. Our study confirms the results of Johns et al. [12]and demonstrates a reduction of the minimum alveolar concentration for isoflurane by acute L-NAME treatment in wild-type mice. In addition, we demonstrated that the RRED⁵⁰for isoflurane was decreased by acute L-NAME administration. The L-NAME-induced reduction of minimum alveolar concentration and RRED⁵⁰in wild-type mice was completely reversible by subsequent L-arginine administration but not by D-arginine, suggesting the specific site of action of L-NAME is NOS.

The precise mechanisms of general anesthesia remain largely unknown. A number of studies suggest that the anesthetic actions of general anesthetics are, at least in part, due to inhibition of brain NMDA receptors. [27,28],* The NMDA receptor is the best characterized of the various excitatory glutamate-activated receptors in mammalian brain. Both competitive and noncompetitive antagonism of the NMDA receptor have been shown to reduce the minimum alveolar concentration of volatile anesthetics. [27],* Ketamine is believed to produce anesthesia by noncompetitive antagonism of the NMDA receptor, [28]and we demonstrated that ketamine reduced the minimum alveolar concentration of isoflurane in nNOS knockout mice. Thus it is likely that the NMDA receptor mechanism is intact in nNOS knockout mice. Activation of the NMDA receptor has been shown to result in a Calcium²+ -dependent increase in cGMP via the production of nitric oxide in the CNS. [29]Current evidence also indicates that NMDA receptor activation and the subsequent release of nitric oxide play a pivotal role in spinal nociceptive processing in chronic pain models [2,30]and long-term potentiation [19]as a persistent increase of synaptic strength implicated in certain forms of learning and memory. Moore et al. reported antinociceptive action of NOS inhibitors given by various routes (intraperitoneal, intracerebroventricular, and oral) evaluated by three nociception paradigms in mice. [23-25]Although most of the evidence is accumulated from studies of chronic pain models and long-term potentiation, the weight of prior observations suggests that nitric oxide plays a major role as a neurotransmitter in the central nervous system.

Because many of the effects of NMDA receptor activation appear to be mediated ultimately via production of nitric oxide, [1,2]our findings, along with the report by Johns et al., [12]of an L-NAME-induced reduction of halothane minimum alveolar concentration appear consistent with the above hypothesis. In contrast, our discovery that nNOS knockout mice had the same minimal anesthetic requirements as wild-type mice was surprising. Studies of this nNOS knockout mouse demonstrated that long-term potentiation and formalin-induced pain behavior (a test of spinally mediated central sensitization) were intact in these animals. [19,31]Because these knockout mice are completely congenitally deficient in the nNOS enzyme, [17]we considered the possibilities that other isoforms of NOS may serve to generate nitric oxide in the CNS or the existence of other unknown compensatory mechanisms.

As a control experiment, we examined the effects of L-NAME in the knockout mice. Interestingly, the isoflurane minimum alveolar concentration or RRED⁵⁰of knockout mice were not decreased by bolus intraperitoneal L-NAME injection at doses between 10 and 100 mg/kg. Because non-nNOS enzymes, such as eNOS, would be inhibited by L-NAME administration, our observations strongly suggest that, in knockout mice, mechanisms independent of nitric oxide compensate for the chronic congenital deficiency of nitric oxide and preserve the pathways mediating the responses to tail-clamping and the righting reflex against anesthetic depression. In contrast, long-term potentiation was blocked by NOS inhibitors in the mutant mice just as it was in wild-type mice, suggesting the importance of alternative nitric oxide isoforms in this particular form of synaptic plasticity. [19]Immunostaining demonstrated the endothelial form of NOS (eNOS) in the CA1 region of hippocampus from which long-term potentiation was induced. [18,19].

We examined the effects of ketamine on isoflurane minimum alveolar concentration to learn whether the inability of L-NAME to reduce minimum alveolar concentration in these knockout mice was a generalized phenomenon. An intraperitoneal injection of 50 mg/kg ketamine decreased the isoflurane minimum alveolar concentration of knockout mice by more than 20%. This suggests that the failure of L-NAME to reduce anesthetic requirements in knockout mice is quite specific and not a generalized phenomenon in these mutant knockout animals.

Our observation that nNOS knockout mice have an anesthetic requirement identical to that of wild-type mice provides evidence that nitric oxide is not the sole messenger of pain perception or maintaining consciousness. Because of the vital importance of these processes for protecting all life forms, it is likely that multiple mechanisms coexist to mediate these responses. The dose-response curves of L-NAME effects on isoflurane minimum alveolar concentration and RRED⁵⁰of our current study, as well as the study of Johns et al., [12]provide evidence that there may be a plateau of antinociceptive effect at high L-NAME doses. In addition, the largest reduction of anesthetic requirements by L-NAME was approximately 30-50% in both studies. Our observations suggest the coexistence of multiple pathways for maintaining pain perception and consciousness. Another possible explanation for the existence of the dose plateau effect is the incompleteness of inhibition of nNOS by L-NAME. In the rat, bolus intravenous administration of 10-40 mg/kg L-NAME attenuated nNOS catalytic activity by approximately 50%. [32]On the other hand, nNOS activity is inhibited by 90% after a single intravenous injection of L-NAME (10-40 mg/kg) in cats. [33]Therefore, the degree of nNOS inhibition after a single systemic dose of L-NAME varies widely among species. No data are available for mice. It is conceivable that there was a plateau effect of L-NAME on minimum alveolar concentration reduction in our study because nNOS was only partially inhibited by the route and dose of L-NAME that we selected. We do not know why our study showed less dose-dependency of L-NAME on minimum alveolar concentration than that of Johns et al. [12]This difference may be related to differences of route of administration and species.

To better understand the marked difference in the minimum alveolar concentration and RRED⁵⁰response to an acute L-NAME injection of wild-type and knockout mice, we examined the effects of week-long L-NAME treatment in wild-type mice. A previous investigation demonstrated that L-NAME gavage at doses larger than 10 mg/kg resulted in a tenfold reduction of the arterial wall cGMP content, suggesting inactivation of the nitric oxide-cGMP pathway. [26]It has been reported that another potent NOS inhibitor, N^G-nitro-L-arginine, when given at 50 mg/kg intraperitoneally twice a day for 4 days, inhibits nNOS activity by 95% in rats, and NOS activity did not return to normal for at least 5 days. [34]We noted that wild-type mice given an L-NAME gavage every 12 h for a week demonstrated a minimum alveolar concentration and RRED⁵⁰identical to that of untreated wild-type mice as well as saline-gavaged mice. Further challenges with intraperitoneal L-NAME on the 8th day decreased the isoflurane minimum alveolar concentration and RRED sub 50 of saline-gavaged mice but not of L-NAME-gavaged mice. These results demonstrate neural tolerance to L-NAME administration and suggest the existence of compensatory alternative pathways when the brain's nitric oxide-cGMP pathways are inactivated by either congenital gene disruption or pharmacologic inhibition.

Although the isoflurane minimum alveolar concentration and RRED sub 50 values were measured as inspired gas concentrations in the current study, our baseline values are similar to previously reported values (minimum alveolar concentration 1.33 - 1.41 vol%, RRED⁵⁰0.57 - 0.66 vol%). [21,35]Ideally, volatile anesthetic gas concentrations used for the determination of anesthetic potency should be measured as alveolar concentrations because of the moderate solubility of these gases. However, it has been suggested that the inspired-to-alveolar difference of isoflurane concentration might be negligible in the mouse (or other small animals), in which the ventilation per gram of body tissue is high and blood uptake will have less influence on the rate of alveolar concentration increase. [36]There was no apparent hypoxia, hypercapnia, or acidosis in the mice immediately after minimum alveolar concentration studies, ruling out the possibility that hypoxia or hypercapnia impaired minimum alveolar concentration reduction (Table 5). There were no differences among wild-type mice, wild-type mice treated with bolus L-NAME, or nNOS knockout mice of these respiratory indexes, suggesting the difference in anesthetic requirements is not due to differing respiratory status.

We did not measure hemodynamic parameters because of the logistics of the apparatus and the size of the mouse. Although arterial hypertension due to L-NAME administration per se should not affect minimum alveolar concentration, [37]we cannot rule out the possibility that any differences between knockout and wild-type mice are due to differing hemodynamic responses to isoflurane and/or L-NAME. We have not observed significant differences in mean systemic arterial blood pressure values or the ability to autoregulate cerebral blood flow over a wide range of systemic arterial blood pressure between wild-type and nNOS knockout mice. [38].

In summary, mice that lack nNOS have a minimum alveolar concentration and RRED⁵⁰identical to that of wild-type mice. Treatment with NOS inhibitors reduces the minimum alveolar concentration and RRED⁵⁰of wild-type mice, but not knockout mice, suggesting that other non-nitric oxide nociceptive mechanisms exist in the knockout mice to compensate for this mutation. Our studies confirm that acute NOS inhibition has analgesic or anesthetic effects in the wild-type mouse, whereas week-long NOS inhibition by L-NAME induces compensatory adaptations of the nervous system, making wild-type mice once again sensitive to normal levels of inhalational anesthesia. Week-long administration of L-NAME appeared to mimic the normal anesthetic sensitivity of the nNOS knockout mouse, and similar adaptive nervous processes may have occurred. We speculate that the nitric oxide-cGMP pathway plays an important role in pain perception and maintaining consciousness, but nNOS is not an essential nociceptive pathway and can be compensated for by alternative pathways when congenitally inactivated by mutation or chemically inactivated during adulthood for a week. Further studies using isoform-specific NOS inhibitors or other congenital NOS isoform knockouts (e.g., the endothelial NOS knockout mouse) will provide additional information about the role of the nitric oxide-cGMP pathway in modulating pain perception and consciousness, as well as possible alternative nociceptive pathways in the central nervous system.

* Perkins W, Morrow D: A dose dependent reduction in halothane M.A.C. in rats with a competitive N-methyl-D-aspartate (NMDA) receptor antagonist (abstract). Anesth Analg 74:S234, 1992.