Understanding the mechanism of how fear memory can be extinguished could provide potential therapeutic strategies for the treatment of posttraumatic stress disorders. Here we show that infusion of CB1 receptor antagonist into the infralimbic (IL) subregion of the medial prefrontal cortex (mPFC) retarded cue-alone–induced reduction of fear-potentiated startle. Conversely, cannabinoid agonist WIN55212-2 (WIN) facilitated the extinction. Unexpectedly, administration of WIN without cue-alone trials reduced startle potentiation in a dose-dependent manner. The effect of cannabinoid agonists was mimicked by endocannabinoid uptake or fatty acid amide hydrolase inhibitors. Rats were trained with 10 conditioned stimulus (CS+) (yellow light)–shock pairings. Extinction training with CS+ (yellow light)-alone but not CS− (blue light)-alone trials decreased fear-potentiated startle. Intra-IL infusion of WIN before CS−-alone trials decreased startle potentiation, suggesting that the cannabinoid agonist decreased conditioned fear irrespective of whether the rats underwent CS+- or CS−-alone trials. Cannabinoid agonists activated extracellular signal-regulated kinases (ERKs) in mPFC slices, and ERK inhibitor blocked the effect of cannabinoid agonists on fear-potentiated startle. These results suggest that CB1 receptors acting through the phosphorylation of ERK are involved not only in the extinction of conditioned fear but also in the adaptation to aversive situations in general.
Pavlovian fear conditioning is a behavioral procedure in which a cue (conditioned stimulus [CS]) induces a fear response when it is repeatedly paired with a noxious stimulus, often a footshock (unconditioned stimulus [US]) (Pavlov 1927; Davis 2000; LeDoux 2000). However, a conditioned response gradually disappears after animals are repeatedly exposed alone to the cue without the footshock, a process termed extinction (Pavlov 1927; Falls et al. 1992; Rescorla 2001; Myers and Davis 2002). Psychological analysis and animal studies indicate that memory extinction is an active learning process that inhibits expression of the original association rather than erasing it (Myers and Davis 2002; Sotres-Bayon et al. 2004). The infralimbic (IL) subregion of the medial prefrontal cortex (mPFC) plays an important role in the extinction of fear memory (Maren and Quirk 2004; Sotres-Bayon et al. 2004; Milad et al. 2006; Quirk et al. 2006). The IL neurons exert an inhibitory tone over the main output regions of the amygdala (Quirk et al. 2003; Likhtik et al. 2005). Lesions or the inhibition of protein synthesis in the IL impair recall of extinction (Morgan et al. 1993; Quirk et al. 2000; Santini et al. 2004), whereas IL stimulation that mimics extinction-inducing tone presentations reduces fear responses (Milad and Quirk 2002; Milad et al. 2004). Furthermore, recall of extinction is correlated with plasticity in IL single units (Milad and Quirk 2002), IL-evoked potentials (Herry and Garcia 2002), and IL glucose utilization (Barrett et al. 2003).
Mutant mice lacking CB1 receptors were impaired specifically in the extinction of conditioned fear, which could be mimicked by systemic administration of the CB1 receptor antagonist. These results suggest that endocannabinoids (eCBs) are critically involved in the extinction of aversive memory (Marsicano et al. 2002; Barad et al. 2006). Indeed, activation of CB1 receptors has been shown to facilitate extinction (Chhatwal et al. 2005) and block reconsolidation (Lin et al. 2006) and mediate extinction through a habituation-like process (Kamprath et al. 2006). Activation of CB1 receptors induces phosphorylation of extracellular signal-regulated kinases (ERKs) (Bouaboula et al. 1995; Derkinderen et al. 2003). In the ventromedial prefrontal cortex, extinction training led to higher levels of phosphorylated ERKs in wild-type mice as compared with those of CB1-deficient mice (Cannich et al. 2004). In addition, a postextinction infusion of ERK inhibitor into the mPFC blocked recall of extinction (Hugues et al. 2004, 2006). These results raise the possibility that CB1 receptors within the mPFC play a critical role in the modulation of fear memory.
In the present study, we aim to accomplish 3 goals. First, we hope to demonstrate that the extinction of fear memory requires activation of CB1 receptors in the mPFC. Second, we aim to determine the specificity of the conditioning stimulus in the action of the cannabinoid agonist. Finally, we hope to delineate the signal pathway behind the action of the cannabinoid agonist.
Materials and Methods
All procedures were approved by the Institutional Animal Care and Use Committee of the College of Medicine, National Cheng-Kung University. Male Sprague-Dawley rats (175–200 g), anesthetized with sodium pentobarbital (50 mg/kg, intraperitoneally [i.p.]), were mounted on stereotaxic apparatus, and a cannula made of 22 gauge stainless steel tubing was implanted into the IL cortex at the following coordinates: anteroposterior +3.2 mm, mediolateral ±0.5 mm, and dorsoventral –5.2 mm according to Paxinos and Watson (2004). The rats were monitored and handled daily and were given 7 days to recover. WIN55212-2 (WIN), HU210, AM404, and AM251 were obtained from Tocris Cooks Ltd. (Bristol, UK); URB597 was obtained from Cayman Chemical Co. (Ann Arbor, MI). The drugs were dissolved in dimethyl sulfoxide (DMSO) and further diluted with saline before infusion. Drug and vehicle solutions had a final concentration of 20–50% DMSO in saline; drugs were administered at a volume of 0.5 μL and a rate of 0.1 μL/min. The rats were administered the drugs in a different room from the testing room.
Behavioral Apparatus and Procedures
Rats were trained and tested in a stabilimeter device. Behavioral experiments of fear conditioning and extinction training were performed in standard operant chambers (San Diego Instruments, San Diego, CA). The acoustic startle stimulus used was 50-ms white noise at an intensity of 95 dB; the visual CS was a 3.7-s light produced by an 8-W fluorescent bulb attached to the back of the stabilimeter; and the US was a 0.6-mA footshock of a duration of 0.5 s. Rats were placed in the startle test boxes for 10 min and returned to their home cages on 3 consecutive days to habituate them to the test chamber and to minimize the effect of contextual conditioning.
Experiment 1: Effects of CB1 Antagonist on Extinction
On day 1 of the experiment, rats were placed in startle boxes and received 10 light–footshock pairings with an intertrial interval (ITI) of 2 min. On day 2, the rats were returned to the startle boxes and received 3 sessions of 10 presentations of the 3.7-s light in the absence of either the shock or the startle-elicited noise burst (light-alone trials). The sessions were conducted 10 min apart, with an ITI of 1 min. The CB1 receptor antagonist AM251 was infused into the IL of the rats 30 min before the CS-alone trials. On day 3, the rats were tested for fear-potentiated startle. The test involved 10 startle-eliciting noise bursts presented alone (noise-alone trial) and 10 noise bursts presented 3.2 s after the onset of the 3.7-s light (light–noise trials). The 2 trial types were presented in a balanced mixed order (ITI, 30 s). The percentage of fear-potentiated startles was calculated as follows: ([startle amplitude on CS–noise minus noise-alone trials]/[noise-alone trials]) × 100.
Experiment 2: Effects of Cannabinoid Agonist on the Facilitation of Extinction
Rats were conditioned on day 1 of the experiment, and startle potentiation was assessed on the following day (day 2, Test 1). On day 3, WIN was infused into the IL of the rats 30 min before the conduction of CS-alone trials (1 block of 10 presentations of light-alone trials). We used a 10-trial protocol instead of 30 trials because it is then easier to detect the effects of drugs on extinction. A separate group of rats that received WIN infusion but no extinction training served as No-CS controls. The rats were tested for fear-potentiated startle 24 h after extinction training (day 4, Test 2). In order to examine whether WIN-treated rats exhibited reinstatement of fear, the rats were administered US-alone trials (10 unpaired footshocks) on day 5 and retention of memory was tested 24 h later (day 6, Test 3).
Experiment 3: Effects of Cannabinoid Agonists, eCB Transporter Inhibitor or Amide Hydrolase Inhibitor on Fear-Potentiated Startle
The experimental procedure was similar to that of experiment 2, the only difference being that the rats were administered a CB1 agonist but did not receive extinction training on day 3.
Experiment 4: Effects of Cannabinoid Agonist on Shock Reactivity and Baseline Startle
WIN was injected into the ILs of conditioned rats, which were then placed in the training box and administered 3 unpaired footshocks and 42 startle stimuli (0.6-mA, 0.5-s shocks; 95-dB noise-burst startle). The same group of rats was returned to the same startle box 3 days later, injected with vehicle, and administered identical footshocks and startle stimuli.
Experiment 5: Effects of Cannabinoid Agonist on Conditioned Fear Independent of Stimulus
In this experiment, the conditioning stimulus (CS) was yellow light or blue light; only the yellow light (CS+) was paired with US. Rats were trained with 10 CS+–US pairings; 24 h later, they were tested for memory retention (Test 1, day 2). On day 3, the rats were assigned to CS+ or CS− groups: those in the CS+ group received 3 blocks of 10 presentations of yellow-light–alone trials, and those in the CS− group were administered the same number of blue-light–alone trials. Memory retention was assessed 24 h later (Test 2, day 4). In the next experiment, rats were conditioned and tested as described for the above experiments. On the following day, the rats were exposed to CS−-alone trials; then, on day 3, WIN (1 μg, n = 6) or vehicle (n = 6) was infused into the IL of the rats 30 min before the CS−-alone trials. Memory retention was assessed 24 h after drug application (Test 2).
Experiment 6: Effects of Cannabinoid Agonist on ERK Phosphorylation and Western Blot Analysis
In the first experiment, Male Sprague-Dawley rats (175–200 g) were decapitated and their brains rapidly removed and placed in cold oxygenated artificial cerebrospinal fluid (ACSF) solution. Subsequently, the brain was hemisected, transverse slices of a 450-μm thickness were cut, and the mPFC slices were placed in a beaker of oxygenated ACSF at room temperature for at least 1 h. The ACSF solution had the following composition (in mM): NaCl 117, KCl 4.7, CaCl2 2.5, MgCl2 1.2, NaHCO3 25, NaH2PO4 1.2, and glucose 11. The ACSF was bubbled continuously with 95% O2–5% CO2 and had a pH of 7.4. Slices were incubated with WIN (2 μM) for 15 min and then washed to remove the WIN. One hour later, the IL subregion was dissected out under microscope and was frozen and stored at −80 °C until processing.
In the second experiment, rats were conditioned and 24 h later were given an intra-IL infusion of WIN (1 μg), vehicle, U0126 (2 μg) plus WIN, or U0126 only. One hour after drug application, the tissues of the IL subregion were dissected out.
The tissues were sonicated briefly in ice-cold buffer (50 mM Tris–HCl, pH 7.5, 0.3 M sucrose, 5 mM ethylenediaminetetraacetic acid, 2 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 20 μg/mL leupeptin, 4 μg/mL aprotinin, and 1% Triton X-100). Following sonication, the samples were centrifuged at 14 000 rpm for 30 min and the supernatant was obtained. The protein concentration in the soluble fraction was then measured using a Bradford assay, with bovine serum albumin as the standard. Equivalent amounts of protein for each sample were resolved in 8.5% sodium dodecyl sulfate (SDS)–polyacrylamide gels, blotted electrophoretically to Immobilon, and blocked overnight in Tris-buffered saline (TBS) buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl) containing 2% bovine serum albumin. For the detection of the phosphorylated forms of ERK, blots were incubated with anti-phospho-ERK antibody (1:1000, Cell Signaling Technology Inc., Beverly, MA). To control the content of the specific protein per lane, membranes were stripped with 100 mM β-mercaptoethanol and 2% SDS in 62.5 mM Tris–HCl, pH 6.8, for 30 min at 70 °C and reprobed with rabbit anti-pan-ERK antibody (Cell Signaling Technology Inc.). The membranes were then washed with TBS-T buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 0.05% Tween-20) 4 times and incubated with a chemiluminescent substrate, ECL plus (Amersham Biosciences, Hong Kong), for 3 min. Chemiluminescent signals were visualized by exposing the membrane to KODAK BioMax light films, and the entire film was digitized using a Gel Doc video camera. Band intensities (optical density units, OD/mm2) were quantified densitometrically using Quantity One software (Bio-Rad, Hercules, CA). Western blots were developed in the linear range used for densitometry. The levels of pERK in the WIN-treated rats were expressed as a percentage of the level in the vehicle controls.
At the end of the experiments, animals received an overdose of pentobarbital (100 mg/kg) and the brains were removed from the skull and fixed in buffered 4% paraformaldehyde (pH 7.4) for 48 h. Brains were sectioned using a sliding MicroSlicer (DTK-1000, Ted Pella Inc., Redding, CA), and sections (40-μm thick) were stained for Nissl bodies.
Single-factor analysis of variance (ANOVA) and Newman–Keuls post hoc comparisons were used to analyze the difference in startle potentiation between the AM251-treated and vehicle control groups in experiment 1 and the dose-dependent effect of WIN on fear-potentiated startle in experiment 3. The paired t-test was used to analyze the differences in startle amplitude before and after a reminder shock in the drug-treated rats in experiment 2, the shock sensitivity and baseline startle amplitude in experiment 4, and the difference before and after the administration of CS‑-alone trials in experiment 5. All values are presented as mean ± standard error of the mean; P values <0.05 were considered to be statistically significant.
Experiment 1: Intra-IL Infusion of CB1 Antagonist Blocks CS-Alone–Induced Reduction of Fear Memory
In the first experiment, we aimed to determine whether intra-IL infusion of a CB1 antagonist affected extinction. Rats were trained with 10 pairings of light and footshock followed by extinction training, as described in the Materials and Methods. The CB1 receptor antagonist AM251 (2 μg) or vehicle was infused into the ILs of the rats 30 min before CS-alone trials were conducted. ANOVA showed a significant effect in the experimental group (F2,15 = 21.6, P < 0.001), and Newman–Keuls post hoc tests revealed that the level of startle potentiation in the AM251 rats was significantly higher than that of the vehicle rats (P < 0.01) and was equivalent to the conditioned, no extinction rats (P = 0.68) (Fig. 1A). These results suggest that the infusion of AM251 into the IL before CS-alone trials blocked reduction of fear memory. The locations of microinjection tips for the rats in these experiments are shown in Figure 1B. Figure 1C shows representative photomicrographs illustrating the placement of the infusion needle tip in the IL in the vehicle- or AM251-treated rats.
Experiment 2: Intra-IL Administration of WIN Facilitates Extinction
In the next experiment, we examined whether a cannabinoid agonist facilitated extinction: WIN, at a dose of 0.05 μg, was chosen because it did not affect the fear-potentiated startle (see below). Rats were conditioned, and WIN (0.05 μg) or vehicle was infused into the IL 30 min before CS-alone trials were conducted (1 block of 10 presentations of light-only trials). A separate group of rats that received WIN infusion but no extinction training served as a No-CS control. As shown in Figure 2A, the fear-potentiated startle in Test 2 in the CS+ WIN-treated group was significantly lower than in Test 1 (n = 6, P < 0.05). In contrast, the startle amplitudes in Test 2 in the No-CS and vehicle control groups were equivalent to those in Test 1 (P > 0.5). Thus, WIN at a dose that does not affect startle potentiation facilitates extinction of fear memory.
In order to examine whether WIN affects the reinstatement of fear, the rats were given 10 unpaired footshocks. Figure 2A reveals that the level of startle potentiation in the CS+ WIN-treated rats in Test 3 did not differ from that in Test 1 (n = 6, P > 0.1) after US-alone trials. Thus, WIN-treated rats exhibited significant reinstatement after US trials. The infusion cannula tip locations are shown in Figure 2B.
We examined whether memory would recover spontaneously in the WIN-treated rats. The animals were trained according to the above paradigm and then tested for memory recovery 5 days after extinction training (day 9). Consistent with the results shown in Figure 2A, the fear-potentiated startle in Test 2 in the CS+ WIN-treated group was significantly less than that in Test 1 (n = 6, P < 0.05) (Fig. 2C); on the other hand, the startle amplitudes in Test 2 in the No-CS and vehicle control groups were equivalent to those in Test 1 (P > 0.5). Furthermore, the level of startle potentiation in the CS+ WIN–treated rats in Test 3 when the retention of memory was assessed 5 days after CS-alone trials did not differ from that in Test 1 (n = 6, P > 0.1) (Fig. 2C). The locations of the microinjection tips for the rats in these experiments are shown in Figure 2D.
Experiment 3: Intra-IL Infusion of Cannabinoid Agonists, eCB Transporter Inhibitor or Amide Hydrolase Inhibitor Decreases Fear-Potentiated Startle
If reduction of startle potentiation induced by CS-alone trials required activation of the CB1 receptors, then intra-IL infusion of cannabinoids may affect conditioned fear. Therefore, conditioned rats were infused with different doses of the CB1 receptor agonist WIN or vehicle, and memory retention was then assessed 24 h after drug application (Test 2). Local injection of WIN into the IL was found to significantly reduce fear-potentiated startle. ANOVA for startle scores showed a significant effect in the experimental groups (F3,25 = 23.69, P < 0.001), and Newman–Keuls post hoc tests showed that the 3 WIN groups differed from the vehicle group (P < 0.001). Furthermore, less startle occurred in the high-dose groups (1 and 10 μg) than in the low-dose group (0.1 μg, P < 0.05), indicating a dose-dependent effect (Fig. 3A). The locations of microinjection tips for rats in these experiments are shown in Figure 3B. Two rats had cannulae placed in the anterior cingulate cortex and 2 rats in the dorsal peduncular nucleus. Infusion of WIN to these areas did not reduce fear-potentiated startle, and the data were pooled. The startle potentiations were 188.3 ± 10.1% (n = 4) before and 180.1 ± 19.3% after treatment with WIN (1 μg, n = 4), suggesting that the effect was specific to the IL.
Similar results were obtained with another cannabinoid agonist, HU210 (Fig. 3C). In order to determine whether AM251 reversed the effects of WIN and HU210, AM251 (2 μg) and WIN (1 μg) (n = 5) or AM251 and HU210 (1 μg) (n = 5) were sequentially infused into the IL with an interval of 20–25 min. As shown in Figure 3C, AM251 blocked the effects of WIN and HU210 such that there was no difference in the startle amplitude between the vehicle, WIN/AM251, and HU210/AM251 groups (F2,12 = 0.39, P> 0.5). AM251 alone had no effect on fear-potentiated startle (n = 6, P = 0.58 vs. vehicle). The locations of the microinjection tips for rats in these experiments are shown in Figure 3D.
However, the reduction of fear-potentiated startle by WIN could be due to its continued presence and activation of CB1 receptors. To test this possibility, a parallel set of rats was conditioned and infused with WIN (1 μg) as described in the above studies. AM251 or vehicle was infused 30 min before testing. If the reduction of fear-potentiated startle by WIN was due to its continued presence, then pretest administration of AM251 should block the effect of WIN. Figure 4A shows that pretest administration of AM251 did not affect the WIN-induced reduction of fear-potentiated startle, which was measured to be 79.8 ± 18.9% (n = 6) in the AM251 group and 86.0 ± 24.8% (n = 6, P > 0.5) in the vehicle control group. Neither differed from the WIN-treated, no-pretest infusion group of rats (64.7 ± 10.6%, n = 6, P> 0.5) ruling out any lingering effects of WIN. The locations of microinjection tips for rats in these experiments are shown in Figure 4B.
Initially, we planned to test whether the systemic administration of a cannabinoid agonist facilitated extinction as observed when one was infused into the IL. However, it is known that systemic cannabinoid agonists produce profound locomotor and analgesic effects (Tsou et al. 1996; Herzberg et al. 1997) that complicate the interpretation of the results. We therefore performed experiments in which the rats were injected i.p. with AM404 (10 mg/kg) or vehicle 30 min before extinction training, and fear-potentiated startle was measured 24 h after injection. Consistent with a previous report (Chhatwal et al. 2005), AM404 was found to facilitate extinction. The startle potentiation was significantly lower in the AM404-treated rats (86.5 ± 12.7%, n = 6) as compared with the vehicle-treated rats (186.6 ± 16.3%, n = 6, P < 0.001).
The eCB diffusion is limited by high-affinity cellular reuptake (Beltramo et al. 1997; Piomelli et al. 1999). The inhibition of reuptake prolongs the action of released eCB and hence increases CB1 receptor activation. To determine the involvement of eCB in the reduction of fear memory, we examined the effect of AM404 on the startle potentiation. AM404 (0.2 μg) or vehicle was infused into the IL 24 h after training, and the results showed that fear-potentiated startle was significantly less in the AM404-treated rats (83.2 ± 20.6%, n = 6) as compared with that of the vehicle-treated rats (178.5 ± 9.1%, n = 6, P < 0.01). In addition to blocking eCB reuptake, AM404 also produces several effects such as action on the vanilloid receptor (Smart and Jerman 2000) and inhibition of fatty acid amide hydrolase (FAAH) (Jarrahian et al. 2000). We therefore determined whether the effect of AM404 was mediated by the CB1 receptor by infusing AM251 into the IL before AM404. Figure 5A shows that rats treated with AM251 and AM404 exhibited significantly higher startle levels than those of AM404-only–treated rats (P < 0.01) and had equivalent startle levels to the vehicle controls (P = 0.82). This result indicates that eCB uptake inhibitor mimics the effect of CB1 agonists on the reduction of startle potentiation.
Anandamide (AEA) hydrolysis is catalyzed by the enzyme FAAH (Bracey et al. 2002), and a blockade of AEA hydrolysis with FAAH inhibitor URB597 produced an anxiolytic effect (Kathuria et al. 2003). We also tested the effect of URB597 on fear-potentiated startle and found that infusion of URB597 (0.3 μg) into the IL reduced fear-potentiated startle; this effect could be blocked by AM251 (F2,15 = 9.03, P < 0.01) (Fig. 5A). The locations of microinjection tips for rats in these experiments are shown in Figure 5B.
Experiment 4: Effects on Shock Reactivity and Baseline Startle Activity
We assessed whether a cannabinoid agonist and antagonist produced an analgesic effect and affected baseline anxiety by measuring the shock reactivity and baseline startle, respectively, according to the methods described by Chhatwal et al. (2005). The results revealed that there was no difference in shock sensitivity (n = 6 rats, P = 0.77) (Fig. 5C) or baseline startle amplitude (n = 6 rats, P = 0.46) in rats administered WIN or vehicle (Fig. 5D); similarly, there was no difference in shock sensitivity (n = 6 rats, P = 0.41) or baseline startle amplitude (n = 6 rats, P = 0.75) in rats given AM251 or vehicle. Thus, intra-IL administration of WIN or AM251 has no effect on the shock sensitivity or baseline startle amplitude.
WIN Reduces Fear Independent of Conditioning Stimulus
The infusion of a cannabinoid agonist into the IL reduced fear-potentiated startle in the absence of cue-alone trials, suggesting that the effect was not stimulus specific. We therefore tested the possibility that CB1 receptors in the IL elicit a general reduction of fear. To this end, rats were trained using a differential conditioning protocol: yellow light and blue light. Only the yellow light (CS+) was paired with US; blue light (CS−) was not. Rats were administered 10 CS+–US pairings and were assigned to a CS+ or CS− group. Rats in the CS+ group received 3 blocks of 10 presentations of yellow-light–alone trials; rats in the CS− group were given the same number of blue-light–alone trials. Figure 6A shows that fear-potentiated startle was 182.6 ± 27.2% before and 75.2 ± 20.6% (n = 6, P < 0.001) after the CS+ trials; on the other hand, fear-potentiated startle was 198.1 ± 14.3% before and 149.3 ± 24.1% (n = 8, P = 0.13) after the CS− trials, indicating that the rats were able to discriminate between CS+ and the CS−. The rats that received CS−–no shock conditioning (Test 1: 36.4 ± 7.6%) and CS− extinction training (Test 2: 41.7 ± 13.2%, n = 6) served as controls.
Next, a separate group of rats was conditioned and tested as in the above experiment; WIN (1 μg, n = 6) or vehicle (n = 6) was then infused into the IL 30 min before CS−-only trials. Figure 6B shows that fear-potentiated startle was significantly reduced in WIN-treated rats. In contrast, there was no difference between the results of Test 1 and Test 2 in the vehicle-treated rats. These results suggest that intra-IL administration of WIN reduces conditioned fear independent of conditioning stimulus.
Involvement of ERKs in the Action of CB1 Agonists
We explored the possible signal pathway underlying the cannabinoid agonist-mediated reduction of startle potentiation. Rats were decapitated and mPFC slices taken; the slices were then incubated with WIN (2 μM) for 15 min and washed to remove WIN. ERK phosphorylation was analyzed with an antibody directly against the active form of ERKs. As shown in Figure 7A, WIN treatment significantly increased the phosphorylated levels of ERK44 (156.8 ± 16.4%, n = 6, P < 0.01) and ERK42 (131.5 ± 11.6%, n = 6, P < 0.05) relative to the vehicle controls. No change was observed when the cellular extract was blotted with an antibody that recognizes ERKs independent of their phosphorylation state, suggesting that the observed pERKs increments were not due to an increase in the total amount of ERKs. Furthermore, when mitogen-activated protein kinase kinase (MEK) inhibitor U0126 (10 μM) was present, ERK phosphorylation induced by WIN was abolished (ERK44: 99.2 ± 15.3%; ERK42: 94.0 ± 10.2%, n = 6) (Fig. 7A). As a control, U0126 was found to have no effect on ERK phosphorylation in the absence of a cannabinoid agonist.
The next experiment examined whether administration of a cannabinoid agonist in vivo induced ERK phosphorylation. Rats were conditioned and 24 h later were administered an intra-IL infusion of WIN (1 μg), vehicle, U0126 (2 μg) plus WIN, or U0126 only (n = 6 in each group). One hour after drug application, the tissues of the IL subregion were dissected out and ERK phosphorylation was measured. Figure 7B shows that WIN treatment significantly increased the phosphorylated levels of ERK44 (132.7 ± 4.4%, P < 0.01) and ERK42 (133.6 ± 5.9%, P < 0.01) relative to the vehicle controls. U0126 completely abrogated the effect of WIN (ERK44: 99.0 ± 1.1%; ERK42: 101.1 ± 1.1%). As a control, U0126 had no effect on ERK phosphorylation in the absence of a cannabinoid agonist (ERK44: 100.8 ± 1.7%; ERK42: 100.5 ± 1.6%, n = 6).
We next investigated the effect of ERK inhibition on cannabinoid agonist-mediated fear reduction. Rats were given WIN (1 μg), vehicle, or HU210 plus U0126 (2 μg), and fear-potentiated startle was assessed 24 h after drug application. The startle potentiation in the WIN-treated rats (52.8 ± 15.8%, n = 5) was significantly less than that of the vehicle group (188.2 ± 20.9%, n = 5, P < 0.01). The effect of WIN on the startle potentiation was blocked by U0126 treatment (194.0 ± 21.3%, n = 5) (Fig. 8A). The locations of microinjection tips for rats in these experiments are shown in Figure 8B. A similar result was observed with HU210: the startle potentiation in HU210 (1 μg)-treated rats (60.4 ± 15.5%, n = 5) was significantly less than that in the vehicle group (180.8 ± 20.4%, n = 5, P < 0.01). The effect of HU210 on the startle potentiation was blocked by U0126 treatment (171.4 ± 26.0%, n = 6) (Fig. 8C). The locations of microinjection tips for rats in these experiments are shown in Figure 8D. Thus, U0126 blocked the effect of CB1 agonists on fear-potentiated startle at the same dose that inhibited CB1 agonist-induced ERK phosphorylation. A control experiment revealed that U0126 on its own did not affect startle potentiation.
Cannabinoid Receptors in the IL Play a Critical Role in the Extinction of Fear Memory
Studies of extinction in animals and humans indicate that the mPFC is critically involved in this behavioral process (Myers and Davis 2002; Maren and Quirk 2004; Phelps et al. 2004; Sotres-Bayon et al. 2004); however, the cellular mechanism of extinction within the mPFC remains to be defined. In the present study, we have demonstrated that 1) preextinction infusion of the CB1 receptor antagonist into the IL blocked extinction; 2) infusion of a cannabinoid agonist into the IL 30 min before extinction training facilitated extinction of conditioned fear; 3) infusion of a cannabinoid agonist into the IL 24 h after training reduced fear-potentiated startle in the absence of extinction training; and 4) the effect of cannabinoid agonists was mimicked by an eCB uptake inhibitor and FAAH inhibitor. These results indicate that cannabinoid receptors in the IL are not only critically involved in the experimental extinction but also in animals’ ability to adapt to aversive situations in general.
Measurement of the levels of 2 major eCBs, AEA and 2-arachidonoylglycerol, in brain punches of the mPFC and the basolateral amygdala (BLA) complex showed significant increases in the levels in the BLA complex but not in the mPFC at the end of the cue presentation (Marsicano et al. 2002), implying that the BLA is a predominant site for extinction. The discrepancies between our data and previous findings might be due to the different areas from which tissue was taken for sampling. Recently, the same laboratory performed measurements specifically on the IL and found that the phosphorylated level of ERK44 was increased during extinction training (Cannich et al. 2004). Alternatively, the discrepancy might be due to contextual variables (conditioning and extinction in different contexts vs. conditioning and extinction in the same context in the present study). When conditioning and extinction took place in the same context, CS induced release of eCBs in the IL, resulting in extinction that displayed characteristics of spontaneous recovery and reinstatement. Thus, it is suggested that CS preferentially activates the CS–US memory if conditioning and extinction are in the same context (Garcia 2002). Conversely, CS induced release of eCBs in the amygdala if conditioning and extinction were in different contexts, resulting in extinction as well as a blockade of reconsolidation (Lin et al. 2006). Consequently, animals exhibited less reinstatement and spontaneous recovery. Taken together, these results suggest that activation of CB1 receptors in the IL is critically involved in the extinction of fear memory.
Administration of Cannabinoid Agonists into the IL Elicits a General Decrease in Conditioned Fear Independent of Conditioning Stimulus
Infusion of cannabinoid agonists into the IL 24 h after training reduced fear-potentiated startle in the absence of extinction training. The effect of cannabinoid agonists was mimicked by an eCB uptake inhibitor and FAAH inhibitor. Taking into account that administration of the CB1 antagonist to the IL retarded extinction whereas infusion of the CB1 agonist facilitated extinction, it is likely that the cannabinoid receptors in the IL are involved not only in extinction but also in adaptation to aversive situations in general. Most mammals, including rats and mice, are dichromats, having only 2 classes of cone photopigment: short-wavelength–sensitive and medium-wavelength–sensitive cone cells (Bowmaker 1998; Jacobs et al. 2001). As a result, they can distinguish only a fraction of the visible light range seen by humans. Short-wavelength–sensitive cone cells are most sensitive to blue light; therefore, rats are able to discriminate between blue and yellow light. This hypothesis is confirmed at the behavioral level in the present study. In response to repeated CS+ exposure, the rats exhibited an extinction phenomenon such that previously acquired fear-potentiated startle decreased gradually. In contrast, extinction was not observed when the rats had undergone CS−-alone trials. More importantly, intra-IL infusion of WIN decreased the startle potentiation irrespective of whether the rats received CS+- or CS−-alone trials. These results suggest that WIN produces a general reduction in fear independent of conditioning stimulus and are consistent with the idea of cannabinoid-mediated long-term adaptation to aversive situations (Kamprath et al. 2006).
Recently, Arenos et al. (2006) have shown that freezing to a tone was increased in the AM251-treated rats, an observation consistent with a previous report (Marsicano et al. 2002) and with the present results, which show that blockage of CB1 receptors increases fear-potentiated startle during extinction training. Interestingly, freezing behavior during the context test decreased when the rats were treated with AM251 before the training or before the testing, indicating that expression of contextual fear is impaired by AM251 (Arenos et al. 2006). This observation raises the possibility that activation of CB1 receptors enhances the expression of inhibitory (extinction) memory in which uncontrolled associative cues such as olfactory, visual, auditory, or tactile cues were associated with no shock, resulting in a general reduction of fear independent of conditioning stimulus. Alternatively, cannabinoid agonists may have disrupted consolidation, although this is less likely as the CB1-deficient mice and the rats administered with AM251 after training did not exhibit impairment of memory consolidation (Marsicano et al. 2002; Arenos et al. 2006).
Involvement of ERKs in the Action of Cannabinoid Agonists
In the present study, we have demonstrated that activation of cannabinoid receptors in the IL induced ERKs phosphorylation. MEK inhibitor blocked the effect of cannabinoid agonists on the fear-potentiated startle at the same dose that inhibited cannabinoid agonist-induced ERK phosphorylation. Activation of cannabinoid receptors may activate ERKs through the stimulation of a pertussis toxin-sensitive G protein to inhibit adenylyl cyclase and cAMP-dependent protein kinase, leading to a decrease in constitutive phosphorylation of c-Raf at the inhibitory site (Ser259) and an increase in ERK phosphorylation (Davis et al. 2003). ERK is known to be required for several forms of synaptic plasticity and long-term memory (English and Sweatt 1996, 1997; Impey et al. 1998; Schafe et al. 2000). Thus, activation of ERK by cannabinoid agonists induced neuronal plasticity in the mPFC, which subsequently caused sufficient excitation of intercalated neurons to produce relatively more inhibition on the central nucleus of the amygdala, leading to the suppression of conditioned fear (Royer and Pare 2002; Quirk et al. 2003; Likhtik et al. 2005). This is evidenced by a recent report showing that bursting firing of ventromedial PFC neurons is necessary for the consolidation of extinction memory (Burgos-Robles et al. 2007). Alternatively, ventromedial PFC projection neurons activated BLA inhibitory interneurons, leading to feedforward inhibition of BLA principal neurons (Rosenkranz and Grace 2001; Rosenkranz et al. 2003). It has been shown that cannabinoid agonists enhance whereas the CB1 antagonist blocks the response of mPFC neurons to olfactory cues previously paired with a footshock, indicating that cannabinoid receptors within the mPFC could modulate emotional memory through functional input from the amygdala (Laviolette and Grace 2006). Taken together, these results suggest that cannabinoid receptors within the IL and the integrity of the mPFC–amygdala–mPFC circuitry are important in the regulation of emotionally learned fear.
In summary, there are 3 important findings of this study. First, the induction of extinction of fear memory by CS-alone trials is retarded by the CB1 receptor antagonist; on the other hand, activation of CB1 receptors in the IL facilitates extinction. Second, cannabinoid agonists reduce fear-potentiated startle independent of conditioning stimulus. Third, activation of CB1 receptors in the IL induces ERKs phosphorylation, and MEK inhibitor blocks the effect of the cannabinoid agonist on fear-potentiated startle at the same dose that inhibits cannabinoid agonist–induced ERK phosphorylation. We suggest that cannabinoid receptors in the IL are involved not only in extinction but also in adaptation to aversive situations in general, independent of emotional value.
National Science Council (NSC94-2752-B-006-001-PAE); National Health Research Institute (N08I97N); Landmark Project (A0031) of the National Cheng-Kung University of Taiwan.
We thank Dr Min-Der Lai for critical comments on the manuscript.
Conflict of Interest: None declared.