Isoflurane Causes Anterograde but Not Retrograde Amnesia for Pavlovian Fear Conditioning

The Committee on Animal Research of the University of California–San Francisco approved our study of 264 male specific-pathogen–free Sprague-Dawley rats (weight, 275–325 g; Charles River Laboratories, Hollister, CA). Animals were housed in our animal care facility for 1 week before study under 12-h cycles of light and dark, two animals per cage, and had continuous access to standard rat chow and tap water before the study. The dose–response effect of isoflurane on classic fear conditioning was measured in 16 groups trained to tone (n = 8 per group) and 15 groups trained to context (n = 8 per group, except n = 16 per group for the 0.12–minimum alveolar concentration [MAC] one-trial context groups; the number was increased to 16 after finding equivocal results with n = 8, see Discussion).

Anesthetic was delivered to each animal in its home cage. We replaced the normal cage cover with a cover containing ports for inflow, outflow, and sampling of isoflurane. This cover also supported an acrylic cylinder (12-cm diameter, 20-cm-high chimney) through which rats could be easily removed from the cage without altering the isoflurane concentration inside the cage.

Training was accomplished in four identical rectangular box-shaped chambers (25 × 20 × 17 cm) constructed of clear acrylic and located in a well-lit room. The top of each training chamber contained an 8-cm diameter port sealed with a rubber cork. Inlet and outlet ports allowed continuous ventilation through the chambers. A circular flow through the four chambers was maintained by a fan producing a background noise of 70 dB (A-scale; Sound Level Meter; Radio Shack, Ft. Worth, TX). Fresh gas inflow was provided by a 5-l/min oxygen flow through a calibrated isoflurane vaporizer. Carbon dioxide was removed with a soda lime canister, and gas concentrations were sampled from a port in the circle system. For concentrations of isoflurane less than 0.5 MAC, the floor of each training chamber consisted of 14 stainless steel rods (6-mm diameter), spaced 1.8 cm center to center wired to a shock scrambler (Gemini Avoidance System; San Diego Instruments, San Diego, CA). During 0.50 MAC isoflurane administration, the floor consisted of 18 more narrowly spaced bars to prevent the feet of obtunded animals from falling beneath the bars and thereby avoiding the foot shocks. 24A speaker was mounted on the rear wall of each training chamber. Training chambers were cleaned with 2% ammonium hydroxide before and after each animal occupied it.

Tone testing took place in chambers providing a different environment from that provided by the training chambers. The clear acrylic testing chambers had an A-frame roof. They measured 25 × 28 cm base and 21 × 28 cm sides, had a smooth floor, and were in a different room from the training chambers. The test room was lit with a 25-W red light bulb, whereas the training room was lit by conventional white-light fluorescent ceiling lamps. The testing cages were cleaned with a pine-scented solution, and there was no background noise. A speaker was mounted on the rear wall of each testing chamber.

On the training day, animals were brought to the training room and marked on the tail with a permanent felt-tipped pen for identification. After at least a 1-h period of rehabituation to their home cages, isoflurane was delivered for 30 min to these cages at the target anesthetic concentration. Each animal was then rapidly (< 10 s) transferred via  the 8-cm port to a training chamber charged at the target anesthetic concentration. Animals were allowed to explore the training chamber for 3 min before training began. Isoflurane concentrations were continuously measured by a respiratory gas analyzer (Capnomac II; Datex, Helsinki, Finland), which was calibrated with a standard commercially available gas mixture (Quick cal; Datex). Calibration was confirmed by gas chromatography.

For three-trial tone conditioning, animals received three tone-shock pairs consisting of a 30-s tone (90 dB, A-scale, 2,000 Hz) coterminating with a 2-s electric shock (11 Hz bipolar square wave, 2 mA for oxygen [0.00 MAC] groups and 3 mA for groups receiving isoflurane); shock pairs were 90 s apart. Within 30 s after the third shock, the posttraining isoflurane concentration was (1) increased by delivering 3.0 MAC isoflurane in 20 l/min oxygen until the concentration in the chamber reached 1.0 MAC, after which isoflurane was delivered at 1.0 MAC for 1 h; (2) maintained at the target concentration for 1 h; or (3) decreased to zero by returning the animal to its home cage (purged of isoflurane). For one-trial tone conditioning, animals received a single tone-shock pair as described above for the three-trial study. Within 5 s after the shock, the posttraining isoflurane concentration was (1) rapidly increased by injecting 10 ml of liquid isoflurane into the inflow port and delivering 3.0 MAC isoflurane in 20 l/min oxygen until the equilibrated chamber concentration reached 2.0 MAC, after which the chamber concentration was brought to and held at 1.0 MAC for 1 h; or (2) decreased to zero by returning the animal to its home cage (purged of isoflurane). For context conditioning, the identical procedures were used except that no tone was administered. An additional one-trial tone conditioned group was trained during 0.00 MAC, and the posttraining isoflurane concentration rapidly increased (as described for three-trial training) and maintained at 1.0 MAC for 5 h. Figure 1diagrams the sequence of tone-shock pairs for the three- and one-trial training procedures.

The next day, we assessed freezing to tone and context. For tone testing, each rat trained to tone was placed in the triangular (A-frame) testing chamber in the different room. After 3 min of exploration, a tone (90 dB, A-scale, 2,000 Hz) was continuously sounded for 8 min. Shocks were not administered. For context testing, each rat trained to context was returned to the chamber in which it was trained for 8 min. Neither tone nor shock was administered. Four animals were tested simultaneously, one in each of the four chambers. The test sessions were observed via  a videocamera and videotaped for later scoring by a blinded observer. Each animal was scored for 2 s once every 8 s. Behavior was judged as freezing if there was no visible movement except for breathing. 25,26 

For each group, the isoflurane concentration was calculated as the mean of the concentrations measured in the home cages, in the training cages before, and in the training cages after training of that group. The MAC fractions were calculated by dividing by 1.49 vol%, the value we took for MAC. 27 

A freezing score was calculated for each animal as the number of observations judged to be freezing divided by the total number of observations. 25These freezing scores are probability estimates amenable to analysis with parametric statistics. 22–24,28For each group score, the mean and standard error of the mean were calculated. After an arcsine transformation, 29two-factor analysis of variance (ANOVA) StatView (Abacus Concepts, Inc., Berkeley, CA) was used for overall comparison of the posttraining isoflurane treatments. A two-tailed t  test was used to calculate the statistical significance of the difference between group freezing scores for selected training conditions. A P  value < 0.05 was regarded as significant for all comparisons.

In addition, the freezing score of each animal was dichotomously classified as either freezing or nonfreezing, using the previously determined classification thresholds of 19 and 5% for tone and context training, respectively. 24Logistic regression analysis was then applied to provide the EC⁵⁰and standard error values for suppression of fear conditioning to tone and to context for each of the posttraining isoflurane treatments (BMDP Statistical Software; University of California Press, Berkeley, CA). (The 0.00 MAC groups were arbitrarily assigned a concentration value of 0.01 MAC so that logarithms of the concentrations could be calculated for these groups.) A pooled EC⁵⁰was calculated for each training procedure (three-trial tone, three-trial context, one-trial tone, and one-trial context) after determining that within each training procedure there was no significant statistical difference between the EC⁵⁰values of the increased, maintained, or decreased posttraining treatments. A pooled EC⁵⁰was calculated by pooling the freezing scores from all posttraining treatments in a training paradigm. A two-tailed t  test was used to calculate the statistical significance of the difference between the EC⁵⁰values.

For the target groups in the three-trial tone paradigm, termed 0.25, 0.38, and 0.50 MAC, the measured isoflurane concentrations were within ± 0.02 MAC of the target concentrations (fig. 2). For the target groups in the three-trial context paradigm, termed 0.12, 0.25, and 0.38 MAC, the measured isoflurane concentrations were within ± 0.01 MAC of the target (fig. 3). For the target groups in the one-trial tone paradigm, termed 0.00, 0.25, and 0.38 MAC, the measured isoflurane concentrations were within ± 0.01 MAC of the target concentration (fig. 2). For the target groups in the one-trial context paradigm, termed 0.00, 0.12, and 0.25 MAC, the measured isoflurane concentrations were within ± 0.01 MAC of the target (fig. 3).

For the groups that received increased isoflurane after training, the concentration in the training chambers reached 1.0 MAC at 2.2 ± 0.3 min (mean ± SD) after the third training trial of the three-trial training procedure, and reached 1.0 MAC at 0.6 ± 0.2 min and 2.0 MAC at 1.2 ± 0.4 min, respectively, after the single trial of the one-trial training procedure.

Rats trained with shocks signaled by a tone proceeding the shock showed minimal freezing during the 1 min before the onset of the test tone; the highest baseline group score before the tone was 3 ± 2%. Figure 2shows the freezing scores during the test tone for the three-trial tone-shocked trained animals (i.e. , for freezing during testing 24 h after training). The abscissa shows the steady state isoflurane concentration during tone conditioning, and the three curves show the effect of increasing, maintaining unchanged, or discontinuing isoflurane after training. These posttraining treatments resulted in similar freezing scores measured 24 h later. Two-factor ANOVA yielded significant main effects of training concentration (F^2,63= 22.9, P < 0.001), but no significant main effects of posttraining changes of isoflurane (F^2,63= 0.72, P = 0.49), and the interaction between the two was not significant (F^4,63= 0.8, P = 0.53). In addition, table 1shows the EC⁵⁰values of isoflurane during training that provided suppression of tone-shock associations measured at 24 h for each of the posttraining isoflurane treatments. There were no statistically significant differences between the EC⁵⁰values for these treatments.

Figure 3shows the percent freezing at each concentration step for animals trained to context with unsignaled shocks (i.e. , without tones) using the three-trial procedure. Memory for context at each concentration step was not significantly altered by posttraining changes of isoflurane concentrations. Two-factor ANOVA yielded significant main effects of training concentration (F^2,63= 34.1, P < 0.001), but no significant main effects of changes in posttraining concentration (F^2,63= 0.09, P = 0.91), and the interaction between the two was not significant (F^4,63= 0.7, P = 0.99). The EC⁵⁰values resulting from the posttraining treatments (table 1) showed no statistically significant difference.

Figure 2also shows the percent freezing at each concentration step for animals trained to tone using the one-trial procedure. Again, posttraining changes in isoflurane concentration resulted in minimal differences in freezing at 24 h. Two-factor ANOVA yielded significant main effects of training concentration (F^2,42= 17.9, P < 0.001) but no significant main effects of changes in posttraining concentration (F^1,42= 1.0, P = 0.32), and the interaction between the two was not significant (F^2,42= 0.4, P = 0.74). The EC⁵⁰values resulting from the treatments (table 1) showed no statistically significant difference.

In addition, extending the posttraining treatment of isoflurane to 5 h did not significantly alter freezing at 24 h for groups receiving one-trial training to tone at 0.00 MAC: freezing for the 5-h 1.0 MAC treatment was 59 ± 11% (mean ± SE) compared with 67 ± 10% for the 1-h 1.0 MAC treatment and 69 ± 7% for the 0.00 MAC treatment.

Figure 3also shows the percent freezing at each concentration step for animals trained to context using the one-trial procedure. Again, posttraining changes in isoflurane concentration resulted in minimal differences in freezing at 24 h. Two-factor ANOVA yielded significant main effects of training concentration (F^2,58= 15.9, P < 0.001) but no significant main effects of changes in posttraining concentration (F^1,58= 0.2, P = 0.65), and the interaction between the two was not significant (F^2,58= 2.8, P = 0.07). The EC⁵⁰values resulting from the treatments (table 1) showed no statistically significant difference.

We found that posttraining administration of isoflurane did not provide retrograde amnesia for Pavlovian fear conditioning. We looked for retrograde amnesia after training to tone-shocks and context-shocks because different neural substrates underlie these paradigms. Processing by the amygdala is required for both training to tone and training to context, whereas training to context in addition requires processing by the hippocampus. 22,23Thus, our findings indicate that rapidly increasing posttraining isoflurane to 1.0 MAC did not disrupt either amygdalar or hippocampal processing.

We studied three-trial training because this procedure provided robust fear conditioning. However, three-trial training created a 180-s interval between the end of the first trial and the end of the third trial, an interval during which memory processing might occur. Therefore, after finding that isoflurane did not cause retrograde amnesia for the three-trial training, we studied one-trial training to minimize the delay between completion of the training and the posttraining change of isoflurane concentration. For one-trial training, we accelerated the increase in posttraining concentrations by simultaneously injecting liquid isoflurane into the inflow port and administering 3.0 MAC isoflurane, achieving training chamber isoflurane concentrations of 1.0 MAC within 0.6 min and 2.0 MAC within 1.2 min. Even so, isoflurane did not cause retrograde amnesia. Therefore, our findings suggest that isoflurane disrupted memory processes at the time of, or within the first minutes after, the conditioning procedure and did not significantly affect memory processes that occurred later.

We looked for retrograde amnesia in animals that were already partially anesthetized because this condition resembles the clinical situation where an agent could be called on to provide retrograde amnesia. In such situations, the partial suppression of memory processing by the existing anesthesia might enhance the ability of additional isoflurane to create retrograde amnesia. We did not find a significant difference (at P = 0.05) between the increased versus  decreased posttraining groups of any training condition (i.e. , the same training protocol and isoflurane concentration). Of the six training conditions during which administration of isoflurane during training suppressed but did not abolish fear conditioning of posttraining discontinued groups, posttraining increases in isoflurane appeared to enhance suppression (not significantly) for four conditions (0.38 and 0.50 MAC three-trial tone, 0.38 MAC one-trial tone, and 0.12 MAC one-trial context) but not for two (0.25 MAC one-trial tone and 0.38 MAC three-trial context). Therefore, according to a sign test, even this post hoc  approach failed to show a significant difference. Initial study of the 0.12 MAC one-trial context condition (n = 8 per group) yielded P = 0.06 for the difference between posttraining treatments. Because a significantly different result for this training condition could suggest a clinical role for applying isoflurane immediately after an untoward event, we studied an additional eight animals per group. The difference between the enlarged groups (n = 16 per group) yielded P = 0.09. Thus, we did not find any evidence of significant retrograde amnesia.

In contrast to the absence of retrograde amnesia for Pavlovian fear conditioning in the current study, previous investigators found retrograde amnesia for instrumental learning by applying posttraining treatments with electroconvulsive shock, hypothermia, and anesthetics such as diethyl ether, halothane, and propofol. 7–16,30Characteristically, those treatments were applied within 3–5 min or more of training. The retrograde amnestic effect by those anesthetics was dose-dependant and required a depth of anesthesia to at least the loss of the righting reflex. 7–12,31Therefore, the procedures we used to apply posttraining isoflurane treatments were within the time frame and dose that previous investigators used to produce retrograde amnesia for instrumental learning.

Perhaps the difference between our results and those of previous investigators stems from differences in the processes underlying the two forms of learning. Such differences are suggested by the effects of posttraining infusion of receptor agonists and antagonists involved in synaptic transmission. For example, posttraining infusion of scopolamine, a muscarinic acetylcholine antagonist, provided retrograde amnesia for inhibitory avoidance but not Pavlovian fear conditioning. 32,33Post-training intraamygdalar infusion of the γ-aminobutyric acid–mediated agonist, muscimol, suppressed inhibitory avoidance learning but not Pavlovian fear conditioning. 13,14,34Similarly, posttraining infusion of the N -methyl-d-aspartate receptor antagonist AP5 (DL-2-amino-5-phosphonovalerate) interfered with inhibitory avoidance but not Pavlovian fear conditioning. 28,35Furthermore, lesions of the amygdala have less disruptive effect on inhibitory avoidance than Pavlovian fear conditioning. 36–39 

Therefore, isoflurane, scopolamine, and muscimol (and perhaps therefore, γ-aminobutyric acid–mediated agents such as propofol and midazolam) do not provide retrograde amnesia for Pavlovian fear conditioning. 33,34In contrast, Vazdarjanova and McGaugh 40found that posttraining intraamygdalar infusion of lidocaine did create retrograde amnesia for Pavlovian fear conditioning to context. As Wilensky et al.  pointed out, lidocaine affects not only Na⁺channels but also cyclic adenosine monophosphate–dependent intracellular signaling agents such as cyclic adenosine monophosphate–dependent protein kinase A. 34In addition, posttraining disruption of protein kinase A activity can impair memory consolidation for Pavlovian fear conditioning. 41,42Taken together, these studies suggest that, for Pavlovian fear conditioning, synaptic transmission processes involved in memory formation are relatively invulnerable to posttraining treatments, whereas the subsequent intracellular events that stabilize memory may be vulnerable to such treatments.

Perhaps the differences between our results and those of previous investigators who found retrograde amnesia stem from other differences in methodology. Ours is the first to study isoflurane, which may have different amnestic effects than other agents. The studies with halothane evaluated appetitive (reward), not fear conditioning. 10The pungency of diethyl ether and its rapid posttraining administration might have produced additional learning that interfered with target memories associated with the instrumental training and thus caused retrograde amnesia by a different process. Posttraining injections may have provided their own learning experiences. We sought to overcome some of these limitations by administering posttraining isoflurane to animals that had already received nearly amnestic concentrations of anesthetic as well as to unanesthetized animals.

Although isoflurane did not provide retrograde amnesia, it created robust dose-dependent anterograde amnesia as shown by the significant ANOVA main effects of training concentrations. The pooled EC⁵⁰values for training to tone versus  context for the three-trial (P < 0.05) and one-trial procedures (P < 0.05) confirm our previous reports, 21,24that training to context was more vulnerable to the effect of isoflurane than training to tone. The pooled EC⁵⁰values for three-trial versus  one-trial training to tone (P < 0.05) and context (P < 0.05) show that the one-trial training procedure was more vulnerable to the effect of isoflurane than the three-trial procedure.

In summary, we found that isoflurane provided intense anterograde but not retrograde amnesia for Pavlovian (classic) fear conditioning. Fear conditioning to context was more vulnerable to the effect of isoflurane than fear conditioning to tone. The one-trial training procedure was more vulnerable to the effect of isoflurane than the three-trial procedure. In contrast to previous reports that anesthetics can cause retrograde amnesia for instrumental (operant) forms of learning, we found that isoflurane did not cause retrograde amnesia for Pavlovian fear conditioning.