James M. Sonner

In this study, we measured the minimum alveolar anesthetic concentration (MAC) in several mouse strains, including strains used in the construction of genetically engineered mice. This is important because defined genetic modifications are used increasingly to test mechanisms of inhaled anesthetic action, and background variability in MAC can potentially influence the interpretation of these studies. We investigated the effect of strain on MAC for desflurane, isoflurane, halothane, ethanol, the experimental anesthetic 1-chloro-1,2,2-trifluorocyclobutane, and convulsive 50% effective dose (the dose required to produce convulsions in 50% of animals) of the nonimmobilizer 1,2-dichlorohexafluorocyclobutane. These drugs were studied in eight inbred strains, including both laboratory and wild mouse strains (129/J, 129/SvJ, 129/Ola Hsd, C57BL/6NHsd, C57BL/6J, DBA/2J, Spret/Ei, and Cast/Ei), one hybrid strain (B6129F2/J, derived from the C57BL/6J and 129/J strains), and one outbred strain (CD-1). To test our ability to detect effects in a genetically modified mouse, we compared these data with those for a mouse lacking the gamma (neuronal) isoform of the protein kinase C gene (PKCgamma). We also assessed whether amputating the tail tip of mice (a standard method of obtaining tissue for genetic analysis) increased MAC (e.g., by sensitization of the spinal cord). MAC and convulsant 50% effective dose values differed modestly among strains, with a range of 17% to 39% from the lowest to highest values for MAC using conventional anesthetics, and up to 48% using the experimental anesthetic 1-chloro-1,2,2-trifluorocyclobutane. Convulsivity to the nonimmobilizer varied by 47%. Amputating the tail tip did not affect MAC. PKCgamma knockout mice had significantly higher MAC values than control animals for isoflurane, but not for halothane or desflurane, which implies that protein phosphorylation by PKCgamma can alter sensitivity to isoflurane. Anesthetic potency differs by modest amounts among inbred, outbred, wild, and laboratory mouse strains. Absence of the neural form of protein kinase C increases minimum alveolar anesthetic concentration for isoflurane, indicating that protein phosphorylation by the gamma-isoform of protein kinase C (PKCgamma) can influence the potency of this anesthetic.

We previously demonstrated that desflurane and two nonimmobilizers dose-dependently decrease learning and memory in rats. This suggests that although they do not suppress movement in response to noxious stimuli, nonimmobilizers act like inhaled anesthetics in their effects on learning and memory. Like most conventional anesthetics, nonimmobilizers have a greater affinity for lipid than for aqueous phases. In the present study, we examined the effect of ethanol on learning and memory to test the hypothesis that a large part of the capacity of anesthetics to affect learning depends on an action on a lipid (nonpolar) phase. Unlike volatile anesthetics and nonimmobilizers, ethanol has a greater affinity for water than for lipids. Thus, if our hypothesis is correct, ethanol should be relatively less potent in its suppression of memory. Rats receiving various doses of ethanol were conditioned to fear a light followed by a footshock. Fear conditioning to the light was subsequently assessed by measurement of potentiation of the acoustic startle reflex in the presence, compared with the absence, of light. Ethanol up to 0.54 minimum alveolar anesthetic concentration (MAC) did not abolish fear, but 0.82 MAC ethanol did abolish learning. Expressed as a fraction of MAC or predicted MAC, ethanol is less potent than desflurane or the nonimmobilizer 1,2-dichlorohexafluorocyclobutane in suppressing learning. This finding is consistent with the hypothesis that the capacity of anesthetics and nonimmobilizers to impair learning and memory depends mostly on an action at a nonpolar site. Abolition of learning and memory is an important property of inhaled anesthetics. This effect primarily results from an action at a lipid (nonpolar) site, rather than a polar site or a water-lipid interface.

We previously demonstrated that anesthetics and non-immobilizers suppress learning and memory in rats. In the training portion of the test, rats received a light plus a footshock and learned to associate the two, as evidenced by subsequent potentiation of the response (jumping) to light plus a noise

We previously demonstrated that anesthetics and non-immobilizers suppress learning and memory in rats. In the training portion of the test, rats received a light plus a footshock and learned to associate the two, as evidenced by subsequent potentiation of the response (jumping) to light plus a noise (fear-potentiated startle). However, anesthetics and nonimmobilizers also decreased the response of animals receiving footshocks during training, which suggests that the reduction in fear-potentiated startle might reflect analgesia, rather than an impairment of learning and memory. Furthermore, although we previously demonstrated that the nonimmobilizer 2,3-dichlorohexafluorocyclobutane (2N) could completely abolish learning, we did not demonstrate the minimal dose required. In the present study, we eliminated analgesia as a confounding factor by training rats breathing desflurane and 2N with footshock intensities that produced responses at least equal to those produced in control animals. Both desflurane and 2N suppressed learning at 0.2 times the minimum alveolar anesthetic concentration (MAC) or the MAC predicted from lipid solubility, despite the increased footshock intensity. This partial pressure of desflurane equals that previously shown to suppress learning at lower footshock intensities. We conclude that suppression of learning and memory by desflurane and 2N does not result from decreased sensitivity to the unconditioned stimulus (the footshock) and that the potency of 2N is consistent with its lipophilicity. General anesthesia eliminates recall of intraoperative events, including pain. Using an animal model, we refuted the hypothesis that lack of recall results from the analgesia (i.e., the reduced response to painful stimuli produced by inhaled drugs) rather than from a direct effect on learning.

Nonimmobilizers (previously called nonanesthetics) do not prevent movement in response to a noxious stimulus, even at doses predicted to produce anesthesia.We hypothesized they would also lack antinociceptive effects. We tested this prediction using the tail-flick latency (TFL) test. As predicted, the two nonimmobilizers tested (1,2-dichlorohexafluorocyclobutane and perfluoropentane) did not alter TFL, whereas desflurane and nitrous oxide both lengthened TFL (nitrous oxide at a lower minimum alveolar anesthetic concentration [MAC]-multiple than desflurane). In addition, we found that 0.1 MAC desflurane had a hyperalgesic effect (shortened TFL). Implications: We studied the response of animals inhaling anesthetics or nonimmobilizers (compounds predicted to be anesthetics from the Meyer-Overton relation) to painful stimuli. Nonimmobilizers had no effect on these responses; at a low partial pressure, desflurane was hyperalgesic; nitrous oxide and, at higher partial pressures, desflurane were antinociceptive. (Anesth Analg 1998;86:629-34)