Propofol Enhances Memory Formation via  an Interaction with the Endocannabinoid System

• Propofol is associated with a higher incidence of traumatic memories. Despite barbiturates and benzodiazepines, propofol increases endocannabinoid concentrations in the rat brain

• This unique effect of propofol suggests that inhibition of cannabinoid signaling might provide a therapeutic strategy for blocking posttraumatic memory formation

PROPOFOL is a commonly used agent for general anesthesia and for sedation in patients undergoing intensive care treatment (ICU). It is known to reduce postoperative nausea and vomiting1and is associated with postoperative mood alterations and a higher incidence of dreaming compared with other general anesthetics. However, the use of propofol for general anesthesia or for sedation of critically ill patients in the ICU is not universally successful with respect to preventing traumatic memories from perioperative awareness and ICU treatment.2There is extensive evidence that the occurrence of traumatic experiences associated with perioperative awareness or ICU treatment could result in stress-related disorders such as posttraumatic stress disorder and impaired long-term health-related quality of life outcomes.3,4One clinical study, investigating propofol's effects on memory, reported that propofol inhibits conscious memory processing in human subjects soon after memory encoding and that it impairs the encoding of material into long-term memory.5In another study, propofol administration to rats induced amnesia of training on an inhibitory avoidance task.6However, in both studies propofol was administered before learning, thus revealing propofol's effect on the encoding of new information. No studies are available regarding propofol's effects on the consolidation of traumatic memories. However, because patients often have experienced stressful events, such as preoperative fear and anxiety, car accidents, myocardial infarctions, or acute respiratory distress shortly before induction of general anesthesia or sedation with propofol, it is crucial to investigate the effects of propofol administered shortly after the acquisition of new information, a time window when the memory trace is consolidated into stable long-term memory.

Propofol inhibits the enzyme fatty acid amide hydrolase, which is known to degrade endocannabinoids, especially anandamide.7Like propofol, the endocannabinoid system recently has been shown to be crucially involved in mood control in animals8,9and the regulation of nausea and vomiting in humans during stress.10Thus, some of the mentioned propofol effects could be attributable to an activation of the endocannabinoid system.11Propofol administration to mice has been shown to increase endocannabinoid content within the brain, an effect that could not be detected with other sedative agents, such as midazolam or thiopental.7In addition, endocannabinoid plasma concentrations increased moderately in patients undergoing propofol anesthesia but decreased in patients undergoing general anesthesia with a volatile agent such as sevoflurane12or isoflurane.13The endocannabinoid system consists of endocannabinoid ligands, the endogenous cannabinoid receptors 1 and 2 (CB1 and CB2), and enzymes involved in the synthesis and metabolism of endocannabinoids.14Endocannabinoids (i.e. , anandamide and 2-arachidonoylglycerol) are synthesized on demand through cleavage of membrane precursors and serve as retrograde messengers at central synapses.15They bind to CB1 receptors on axon terminals to regulate ion channel activity and neurotransmitter release16and are degraded intracellularly by specific enzymes: anandamide is mainly degraded by fatty acid amide hydrolase and 2-arachidonoylglycerol by monoacylglycerol lipase.17CB1 receptors are highly expressed in several brain regions and in lower densities outside the brain.18,19In contrast, CB2 receptors have a more restricted distribution and are found mainly on immune cells and in low numbers in the brainstem20and some other brain regions.21Both CB1 and CB2 receptors primarily signal through inhibitory G proteins.22 

Recent evidence indicates an important role for endocannabinoids and CB1 receptor activation in enhancing the memory consolidation of emotionally arousing experiences.23,24Moreover, it recently has been shown that the fatty acid amide hydrolase inhibitor URB597 enhances memory acquisition and consolidation in rats.25These findings suggest that propofol might modulate memory consolidation of emotionally arousing experiences via  an interaction with the endocannabinoid system. To investigate this issue, in a first experiment, anesthetic doses of propofol were administered to rats by intraperitoneal injection, immediately and 30, 90, and 180 min after aversively motivated inhibitory avoidance training, a widely used animal model to assess drug effects on emotional memory consolidation. In a second experiment, we evaluated whether the propofol effect on the consolidation of inhibitory avoidance memory is specific for this anesthetic by administering anesthetic doses of the benzodiazepine midazolam or the barbiturate pentobarbital immediately after inhibitory avoidance training. In the last experiment, we investigated whether the memory-enhancing effect of propofol depends on concurrent CB1 activity by administering a nonimpairing dose of the CB1 receptor antagonist rimonabant 30 min before propofol injection; we also studied whether propofol administration modulates endocannabinoid release in rats.

Male adult Sprague-Dawley rats (350–450 g at the time of training; Charles River Laboratories, Calco, Italy) were housed individually and maintained in a temperature-controlled environment (20°± 1°C) under a 24-h light-dark cycle (7:00 am to 7:00 pm lights on) with unlimited access to food and water. All procedures involving animal care or treatments were approved by the Italian Ministry of Health (Rome, Italy) and performed in compliance with the guidelines of the US National Institutes of Health and the Italian Ministry of Health (D.L. 116/92), the Declaration of Helsinki, and the Guide for the Care and Use of Mammals in Neuroscience and Behavioral Research (National Research Council 2004).

2,6-Diisopropyl phenol (propofol, 250, 300, or 350 mg/kg), purchased from Sigma-Aldrich (Milan, Italy), was dissolved in a vehicle containing 100% sesame oil. Midazolam (30, 50, or 70 mg/kg; Ratiopharm, Ulm, Germany) was dissolved in saline, and pentobarbital (60, 70, or 80 mg/kg; Sigma-Aldrich, St. Louis, MO) was dissolved in a vehicle containing 40% propylene glycol (1,2-propanediol), 10% ethanol, and 50% distilled water. Drug solutions were freshly prepared before each experiment and administered by intraperitoneal injection in a volume of 1 ml/kg immediately after the training trial. To control for time specificity, propofol was administered to different groups of rats either 30, 90, or 180 min after the training trial. To assess whether CB1 receptors are involved in mediating the propofol effect on memory consolidation, the CB1 receptor antagonist rimonabant (1 mg/kg; donated by the National Institute of Mental Health, Chemical Synthesis and Drug Supply Program, Bethesda, MD) was dissolved in a vehicle containing 5% polyethylene glycol, 5% TWEEN 80, and 90% saline9and administered immediately after training, whereas propofol was given 30 min later.

Rats were trained and tested in an inhibitory avoidance apparatus consisting of a trough-shaped alley (91 cm long, 15 cm deep, 20 cm wide at the top, and 6.4 cm wide at the bottom) divided into two compartments, separated by a sliding door that opened by retracting into the floor. The starting compartment (31 cm long) was made of opaque white plastic and illuminated by a lamp; the shock compartment (60 cm long) was made of two dark, electrifiable metal plates and was not illuminated.26Training and testing were performed during the light phase, between 10:00 am and 2:00 pm, and were conducted in dim light conditions in a sound-attenuated room. Animals were handled 1 min each for 2 days before the training day.

For training, the rats were placed into the starting compartment of the apparatus, facing away from the door, and were permitted to explore the apparatus. After the rats stepped completely into the dark compartment, the sliding door was closed and a single, inescapable foot shock (0.35 mA) was delivered for 1 s. The animals were removed from the shock compartment 15 s after termination of the foot shock. Retention was tested 48 h later. On the retention test trial, the rats were placed into the starting compartment, and the latency to reenter the shock compartment with all four paws (maximum latency of 600 s) was recorded and used as a measure of retention. Longer latencies were interpreted as indicating better retention.27Immediately after the training and testing of each animal, the apparatus was cleaned with a 70% ethanol solution.

To be included in the test phase, rats they had to reach a minimum criterion on the training test (before treatment), which is 60 s maximum to step in the dark compartment of the maze.

All the analyses were performed by the same observer, who was unaware of animal treatment.

Sleeping parameters were determined in different groups of rats. To determine sleeping onset and recovery, immediately after anesthetic administration each rat was placed on its back once every 30 s until it was unable to right itself within 30 s. Sleeping onset was defined as the interval between anesthetic injection and the time the rat was unable to turn itself upright at least twice within 1 min. Then each rat was left undisturbed on its back until it spontaneously regained its righting reflexes, defined as having at least three paws under its body. Complete recovery of the righting reflex was defined as the rat being able to turn itself upright. The time between loss and recovery of righting reflex for each rat was defined as sleeping time (cutoff = 180 min).28All of the analyses were performed by the same observer, who was unaware of animal treatment.

In accordance with Patel's protocol in mice,7rats were treated with propofol (300 mg/kg, intraperitoneally) or with its vehicle and killed 8 or 40 min after administration.

Brain and plasma samples were subjected to a lipid extraction process, and the endocannabinoid content of the lipid extracts was determined using isotope-dilution liquid chromatography-mass spectrometry as described previously.12The brain tissue was collected and stored at −80°C. Before the extraction process, tissues were weighted and homogenized in polypropylene tubes (Sarstedt, Numbrecht, Germany) and kept in ice water. Five hundred μl of the described homogenized tissue solution was transferred to a 2-ml Eppendorf tube, and 20 μl of internal standard and 1 ml methyl tertiary butyl ether (Sigma-Aldrich, Italy) were added to extract the endocannabinoids. The mixture was vortexed for 1 min and centrifuged at 12,000g  for 6 min. The clear supernatant was transferred into a clean 5-ml polypropylene tube (Sarstedt) and evaporated under vacuum at 37°C. The residue of all evaporated samples was reconstituted in 100 μl acetonitrile, vortexed for 30 s, and sonicated in 4°C water for 15 min. A 20-μl aliquot of the clear solution was used for liquid chromatography-tandem mass spectrometry analysis. All samples were injected in duplicates.

The training and retention latencies of rats were analyzed with one-way ANOVA. Time-dependent effects of propofol, the interactions between propofol and rimonabant, and propofol effects on endocannabinoid concentrations were analyzed with two-way ANOVAs. The source of the detected significances was determined by Tukey–Kramer post hoc  tests. To determine whether learning had occurred, paired t  tests were used to compare the training and retention latencies of the vehicle groups. Sleeping parameters were analyzed with Kruskal-Wallis one-way ANOVA on ranks or Mann–Whitney U test because of their nonnormal distribution. StatView software (SAS Institute, Cary, NC) was used to conduct statistical analyses. Normal data are expressed as mean ± SD; nonparametric data are expressed as median and percentiles. Two-tailed testing was used for all the analyses. P  values of < 0.05 were considered statistically significant. The number of rats per group is indicated in the figures and tables.

This experiment examined whether immediate posttraining administration of propofol would enhance 48-h retention performance of inhibitory avoidance training. Average step-through latencies for all groups during training (i.e. , before foot shock and drug treatment) were 17.6 ± 13.7 s (mean ± SD). One-way ANOVA for training latencies revealed no significant differences between groups (F^3,46= 0.93, P = 0.43). The 48-h retention latencies of rats given vehicle immediately after training were significantly longer than their entrance latencies during the training trial (t =−5.59, P = 0.0002), indicating that the rats retained memory of the shock experience. As shown in figure 1, propofol induced dose-dependent retention enhancement. One-way ANOVA for 48-h retention latencies revealed a significant treatment effect (F^3,43= 7.82, P = 0.0003). Post hoc  analysis indicated that rats treated with the higher doses of propofol (300 or 350 mg/kg) had significantly longer retention than did those treated with vehicle or with 250 mg/kg propofol (P < 0.01 and P < 0.05 for 300 and 350 mg/kg, respectively). The lower dose of propofol (250 mg/kg), which did not induce anesthesia, did not induce retention enhancement. Three of 12 rats given 350 mg/kg propofol died of respiratory depression.

To examine whether propofol influences the consolidation phase of memory processing, rats were treated with propofol (300 mg/kg) immediately or 30, 90, or 180 min after training. Average step-through latencies for all groups during training, before foot shock and drug treatment, were 16.6 ± 13.0 s (mean ± SD). Two-way ANOVA for training latencies revealed no significant differences between groups (main effect of treatment F^1,78= 0.77, P = 0.38; main effect of time of administration F^3,78= 2.0, P = 0.12; interaction F^3,78= 1.54, P = 0.21). Two-way ANOVA for 48-h retention latencies revealed a significant main effect of propofol (F^1,78= 17.64, P < 0.0001) as well as a significant main effect of time of administration (F^3,78= 3.76, P = 0.014). Moreover, there was a statistically significant interaction effect between treatment and time of administration (F^3,78= 4.76, P = 0.0042). As shown in figure 2, post hoc  analysis indicated that rats treated with propofol either immediately or 30 min after training had significantly longer retention latencies than did those given vehicle (P < 0.01). Retention latencies of rats injected with propofol immediately or 30 min posttraining were significantly longer than were those of rats given propofol 180 min after the training (P < 0.01).

To determine whether the propofol effect on inhibitory avoidance memory enhancement is specific for this anesthetic, rats were treated with anesthetic doses of midazolam (30, 50, or 70 mg/kg, intraperitoneally) or pentobarbital (60, 70, or 80 mg/kg, intraperitoneally) immediately after inhibitory avoidance training. For midazolam, average step-through latencies for all groups during training, before foot shock and drug treatment, were 17.7 ± 13.9 s (mean ± SD). One-way ANOVA for training latencies revealed no significant differences between groups (F^3,34= 0.17, P = 0.92). As shown in figure 3A, one-way ANOVA for 48-h retention latencies indicated that midazolam did not significantly enhance retention latencies (F^3,34= 0.09, P = 0.97).

For pentobarbital, average step-through latencies for both groups during training, before foot shock and drug treatment, were 17.2 ± 14.2 s (mean ± SD). One-way ANOVA for training latencies revealed no significant differences between groups (F^3,34= 0.34, P = 0.79). As shown in figure 3B, one-way ANOVA for 48-h retention latencies indicated that pentobarbital did not significantly enhance retention latencies (F^3,34= 0.21, P = 0.89).

This experiment examined whether the memory-enhancing effect of propofol depends on a concurrent activation of CB1 receptors. To address this issue, we investigated whether the CB1 receptor antagonist rimonabant (1 mg/kg) administered intraperitoneally immediately after inhibitory avoidance training would block the retention enhancement induced by propofol given 30 min later. Average step-through latencies for all groups during training, before foot shock and drug treatment, were 15.2 ± 11.8 s. The 48-h retention latencies of rats given vehicle after training were significantly longer than their entrance latencies during the training trial (P = 0.0001). As shown in figure 4, posttraining administration of rimonabant blocked the retention enhancement induced by propofol (300 mg/kg). Two-way ANOVA for 48-h retention latencies revealed a significant rimonabant plus propofol interaction effect (F^1,27= 11.70, P = 0.002). Post hoc  comparison revealed that retention latencies of rats given propofol alone were significantly longer than were those of vehicle-treated rats (P < 0.01). Most importantly, retention latencies of rats given an otherwise nonimpairing dose of rimonabant together with propofol were significantly shorter than those of rats treated with propofol alone (P < 0.01).

Table 1shows the effects of propofol, midazolam, and pentobarbital on sleeping parameters.

Kruskal-Wallis ANOVA revealed no statistically significant effect on sleeping onset (H⁶= 10.27, P = 0.11). However, Kruskal-Wallis ANOVA revealed a statistically significant effect for sleeping time (H⁶= 19.64, P = 0.002). Post hoc  comparisons (Mann–Whitney U test with Bonferroni correction) revealed that rats given 50 mg/kg midazolam slept for a shorter amount of time than did rats given 70 or 80 mg/kg pentobarbital or those given 350 mg/kg propofol. None of the rats treated with the lower doses of midazolam (30 mg/kg) or propofol (250 mg/kg) lost righting reflex.

Table 2shows the effects of rimonabant on propofol in inducing anesthesia. Mann–Whitney U test showed no difference between rats pretreated with rimonabant compared with rats pretreated with vehicle on sleeping onset or time induced by propofol (U = 5.0, P = 0.11; U = 11.000, P = 0.75, respectively), indicating that the anesthetic effect of propofol is independent from the indirect activation of the endocannabinoid system.

Two-way ANOVA for propofol effects on endocannabinoid content revealed a statistically significant interaction between treatment and time of administration (F^1,19= 7.1, P = 0.015). Post hoc  comparisons revealed that propofol increases anandamide concentrations in rat brains 8 min after administration (P < 0.05, table 3).

The current findings indicate that propofol, at anesthetic doses, enhances memory consolidation of inhibitory avoidance training in rats when administered immediately after the training experience. This memory enhancement is blocked by coadministration of the CB1 cannabinoid receptor antagonist rimonabant, suggesting that the enhancing effect of propofol on memory consolidation depends on an indirect activation of CB1 receptors. In contrast, midazolam and pentobarbital, two anesthetics that do not increase endocannabinoid signaling,7did not enhance the consolidation of memory of inhibitory avoidance training.

The current findings may appear at odds with preclinical and clinical findings indicating that propofol induces amnesia. For example, Veselis et al.  5reported that propofol inhibits conscious memory processes in human subjects soon after memory encoding and that it impairs the acquisition or encoding of material into long-term memory. In addition, propofol has been reported to induce amnesia of training in rats on the same inhibitory avoidance task used in the current study.6However, a critical difference between these investigations and the current study is that in the human studies, memory function was assessed shortly after drug administration, whereas in the preclinical study, rats were given the drug before training. Therefore, acute pharmacologic effects could have influenced directly both the acquisition and retention of the training. In contrast, in our study the drug was administered after  the training and was not present during the acquisition phase. Thus, the enhancing effects of propofol on retention performance in our study are likely mediated by specific influences on the consolidation of memory of the training experience.29The use of posttraining drug manipulation is a widely accepted method for effectively dissociating memory processes from secondary behavioral effects of nonassociative nature, such as those related to sensory sensitivity.30Because retention testing took place 48 h after training and drug treatment, these findings further exclude residual pharmacologic effects as having a direct influence on behavior during retention testing. Moreover, the effect of posttraining propofol administration on retention enhancement was time dependent: propofol administration immediately or 30 min after inhibitory avoidance training resulted in memory enhancement, whereas administration of propofol 90 or 180 min after training was ineffective. Together these findings provide evidence that propofol enhances time-dependent processes underlying the consolidation of memory for emotionally arousing experiences. The posttraining drug administration protocol used in the current article has a translational value to humans. Acute sedation or even the induction of anesthesia immediately after  a traumatic experience (e.g. , in the consolidation phase of a traumatic memory) is a common clinical scenario in emergency medicine and in the ICU.

Our findings demonstrate that propofol is able to enhance memory consolidation when administered immediately after the exposure to a traumatic event and that this effect on memory depends on an indirect activation of the endocannabinoid system. In accordance with the behavioral data, we also found that propofol administration increases anandamide concentrations in the rat brain 8 min after injection, whereas anandamide plasma concentrations remain unaffected. Our data are in accordance with preclinical and clinical evidence. Patel et al.  7demonstrated increased concentrations of anandamide in the mouse brain after systemic administration of propofol in contrast to the administration of benzodiazepines, barbiturates, or volatile anesthetics; the effect of propofol on anandamide concentrations is mediated by an inhibition of fatty acid amide hydrolase, the major degradation enzyme of anandamide.7In humans undergoing general anesthesia, plasma concentrations of the endocannabinoid anandamide remained unchanged during propofol anesthesia but were significantly reduced during anesthesia with volatile agents.12,13 

The basolateral complex of the amygdala (BLA) appears to be a critical site for mediating drug effects on memory performance, including those of propofol. One study reported that permanent neurotoxic lesions of the BLA produced with N -methyl-d-aspartate blocked the amnestic effect of pretraining propofol administration of rats trained on an inhibitory avoidance task.6We recently have shown that the endocannabinoid system in the BLA is involved in the enhanced consolidation of inhibitory avoidance memory and that CB1 activity within the BLA is essential for mediating glucocorticoid effects on long-term memory.23–31Based on these findings, a new model has emerged.32–33In this model, stress-induced glucocorticoids bind to membrane-bound receptors in the BLA that activate a G-protein signaling cascade that induces endocannabinoid synthesis. The ensuing release of endocannabinoid ligands could diffuse to local γ-aminobutyric acid–mediated (GABAergic) terminals and inhibit γ-aminobutyric acid release onto noradrenergic terminals in the BLA. The end result of this process is an increased norepinephrine release within the BLA and subsequently an enhancement of emotional memory consolidation. Many sedative and anamnestic effects of general anesthetics, including those of propofol, crucially depend on γ-aminobutyric acid release. The current findings demonstrate that the enhancing effects of propofol on memory consolidation depend on concomitant CB1 receptor activity, so we hypothesize that the anamnestic effects of propofol are mediated by an endocannabinoid-induced inhibition of γ-aminobutyric acid release, resulting in a more pronounced memory consolidation during stressful conditions when glucocorticoid signaling is high.34 

The pharmacokinetic properties of midazolam, pentobarbital, and propofol differ to a large extent, but all three drugs share the pharmacodynamic capability to potentiate γ-aminobutyric acid neurotransmission.35Our results showing that rats treated with midazolam (50 mg/kg) slept less than did rats treated with propofol (350 mg/kg) or pentobarbital (70 or 80 mg/kg) are in accordance with clinical evidence showing that midazolam has a shorter half-life than propofol and barbiturates.35However, neither rats treated with the higher dose of midazolam nor the ones treated with pentobarbital showed differences in the sleeping parameters compared with those treated with propofol. Although propofol enhances memory consolidation through an activation of the endocannabinoid system, the anesthetic effect of propofol does not depend on this activation. The CB1 receptor antagonist rimonabant blocks the propofol-enhancing effect on memory consolidation but does not influence propofol's effects on sleeping. On the whole, these data suggest that, unlike midazolam and pentobarbital, propofol induces selective effects on memory consolidation, which are linked to the activation of the endocannabinoid system and not related to the potentiation of GABAergic neurotransmission.

These findings, together with the results showing that midazolam and pentobarbital, at anesthetic doses, did not influence memory consolidation strongly corroborate the hypothesis that propofol's effects on memory consolidation are not attributable to a general nonspecific anesthetic effect.

In summary, our study demonstrates that propofol enhances memory consolidation via  an endocannabinoid-mediated mechanism. These effects are markedly different from those of other direct GABAergic agents such as midazolam or pentobarbital. These findings from animal experiments suggest that propofol should be used with caution in individuals during the aftermath of an acute traumatic event and may help to explain the increased incidence of aversive memories from intraoperative awareness seen in patients undergoing total intravenous anesthesia with propofol.36Likewise, the findings suggest that pharmacologic manipulation of endocannabinoid signaling could be a useful intervention aimed at blocking memory consolidation immediately after a traumatic event.

The authors thank Viviana Trezza, Ph.D., Pharm.D. (Assistant Professor, Department of Biology, University of RomaTre, Rome, Italy), for critical reading of the manuscript; Flavia Chiarotti, Ph.D., M.Sc. (Researcher, Istituto Superiore di sanità, Rome, Italy), for statistical advice; and Daniela Valeri (Technician, Department of Physiology and Pharmacology, Sapienza University of Rome, Rome, Italy), Antonia Manduca, Pharm.D. (Ph.D. Student, Department of Biology, University of RomaTre), and Cosima Giaffreda and Jessica Miele (Master Students, Department of Physiology and Pharmacology, Sapienza University of Rome) for technical help.