UC Irvine UC Irvine Previously Published Works Title Endogenous cannabinoid release within prefrontal-limbic pathways affects memory consolidation of emotional training. Permalink https://escholarshii juc/item/7p3407r3 Journal Proceedings of the National Academy of Sciences of the United States of America, 111(51) ISSN 0027-8424 Authors Morena, Maria Roozendaal, Benno Trezza, Viviana etal. Publication Date 2014-12-08 Dol 10.1073/pnas.1420285111 License https://creativecommons.or Peer reviewed WinesPNAS ® (CrossMatk Endogenous cannabinoid release within prefrontal- limbic pathways affects memory consolidation of emotional training Maria Morena‘, Benno Roozenda: Luigia Trabace’, Vincenzo Cuomo’ Depa James L. McGaugh*", Gustav Schelling®, and Patrizi *,Piray Atsak™s, ‘campolongo™* nt of Physiology and Pharmacology, Sapienza Univers of Rome, OO18S Rome, tay: “Deparment of Cogitve Newostence, Radboud University Medial Cerve, 6501 HE Nijmegen, The Netherlands "Donde's stitute fr Brain, Cognition and Benoa Radboud Unversity Nimegen, 6525 C2 Nimegen, Munich, Germany Department of Experimental and Clinical Meine, Unvesty of Foggia, 77122 Foggia, aly, and "Cente forthe Newroiology of Leaming and Memory, Department of Neurobiology and Sehavier, Univesiy ef Califor, Wine, CA 92687-3800, Contributed by James L MeGaugh, November 6, 2014 (sent for review September 16, 2014; reviewed by Bruce 5. MeEien, Mark Packard and Rafael Roesler) Previous studies have provided extensive evidence that adminis tration of cannabinoid drugs after training modulates the con- solidation of memory for an aversive experience, The present ‘experiments investigated whether the memory consolidation is regulated by endogenously released cannabinoids. The exper ‘ments first examined whether the endocannabinoids anandamide (AEA) and 2-arachidonoy! glycerol (2-AG) are released by aversive training. Inhibitory avoidance training with higher footshock intensity produced increased levels of AEA in the amygdala, hippocampus, and medial prefrontal cortex (mPFC) shortly after training in comparison with levels assessed in rats trained with lower footshock intensity or unshecked controls exposed only to the training apparatus. In contrast, 2-AG levels were not signif cantly elevated, The additional finding that posttaining infusions cof the fatty acd amide hydrolase (FAAH) inhibitor URBS97, which selectively increases AEA lavels at active synapses, administered into ‘the basolateral complex ofthe amygdala (BLA), hippocampus, or mPFC enhanced memory strongly suggests that the endogenously released ‘AEA modulates memory consolidation. Moreover, in support of the view that this emotional training-associated increase in endocannabe roid neurotransmission, and Is effects on memory enhancement, depends on the integrity af functional interactions between these different brain regions, we found that disruption of BLA activity blocked the training induced increases in AEA levels as well as the memory enhancement produced by URBSS7 administered into the hippocampus or mPFC. Thus, the findings provide evidence that ‘emotionally arousing training increases AEA levels within prefrontal limbic circuits and strongly suggest that this cannabinoid activation regulates emotional arousal effects on memory consoldation. snandamide | endocannabinoids | cannabinold receptors | inhibitry avoidance | emotional arousal tis well-established that stressful or emotionally srousing Jexperiences are well-remembered (1). In addition, there is extensive evidence that the basolateral complex of the amygdala (BLA), hippocampus, and medial preérontal cortex (mPFC) are all crucially involved in mediating emotional arousal effects on ‘memory consolidation (2, 3). The consolidation of memories of arousing experiences requires an orchestration of neural activity in these brain systems, and the cannabinoid system has emerged as ‘a key modulator of such function (1, 4-8). Endogenous eannabi- noid ligands (termed endocannabinoids, mainly N-arachidonoyl ethanolamine (anandamide; AEA) and 2-arachidonoyl glycerol (2-AG)] are released from postsynaptic membranes and feed- ‘back in a retrograde manner onto either excitatory or inhibitory presynaptic terminals, thus suppressing both excitatory and in- hibitory signaling within specific neuronal circuits (7) Extensive evidence indicates that exogenous cannabinoids ad- ‘ministered into this neural circuitry modulate memory processing prs rego. 1072pns 420285112 of emotionally arousing training (8-12). We previously showed that the synthetic cannabinoid agonist WINSS 212-2, administered into the BLA immediately after inhibitory avoidance training enhances the consolidation of long-term memory (8), Conversely, inhibition of endogenous cannabinoid signaling within the BLA with posttraining infusions of the cannabinoid type 1 (CB1) re- ceplor antagonist AMDS1 impairs inhibitory avoidance memory (8). We recently reported thatthe level of emotional arousal at the ‘ume of training isan important factor in determining cannabinoid effects on memory (13): WINSS,212-2 administration enhanced long-term object recognition memory when rats were trained u- dra high arousal concition but was ineffective with low-arousing training (13). Although the findings of our prior studies, as well as those of other investigators (6, 8, 11-13), indicate that adminis- tration of cannabinoids can modulate memory consolidation, studies have not yel investigated whether AEA or 2-AG are re- leased physiologically in response to emotionally arousing taining and normally play a role in creating strong memories for these experiences. The present experiments investigated ths issue, Rats wore trained on an inhibitory avoidance task under different arousal conditions and were kiled after training for determination of AEA and 2-AG levels in the amygdala, hippocampus, and mPEC: To investigate whether training-induced endocannabinoid release contributes to memory consolidation, we pharmacologically An involvement of exogenous cannabinoids in the regulation of smory for emotional events has emerged over the past de- cade. The present findings demonstrate that after an aversive training experience the endogenous cannabinoid anandamide is released into the amygdala, hippocampus, and medial pre- frontal cortex (mPFC) and normally plays arle in the formation of a strong memory trace, Furthermore, an intact basolateral amygdala is required to coordinate both hippocampal and ‘mPFC endocannabinoid activity during aversive memory con- solidation. Thus, our findings provide the frst evidence, to our knowledge, that a dynamic network involving anandamide signaling within this prefrontablimbic cuit modulates the consolidation of memory for emotionally arousing ta ‘thr corrosion MM, By Vi My Gi an design ema Ay LT, G5, ana PC. anayoed data ane MAF. BR. ILM, ane PC. owote the papers That conan sparing infomation anne at wan.aserlbotesinsplao rapa anes" sapsPNAS blocked AEA degradation in the BLA, hippocampus, or mPFC by locally infusing the faty acid anaide hydrolase (EAH) inhibitor ‘URBS97 immediately after inhibitory avoidance training. In contrast to direct cannabinoid receptor agonists, URBS9 sclec- tively augments AEA signaling at active synapses, thus mimick- ing the physiological condition normally occurring after an emo- tionally arousing training experience. ‘The experiments also investigated whether training-associated endocannabinoid newrotransmission within this newrocircuityy depends on functional interactions between the BLA. hippo- campus, and mPFC, Extensive evidence indicates that the BLA contributes to enhancement of memory for emotional events primarily by integrating neuromodulatory influences and mod- Ulating neural activity and synaptic plasticity in other brain regions (1). The BLA is known to project both directly and in- ditecily to the hippocampus (14, 15) and mPFC (16, 17), and it has been shown that a disruption of BLA activity blocks the memory-enhancing effects of drugs administered directly into either the hippocampus or mPFC (18, 19). Therefore, inthis last series of experiments, we investigated whether an intact BLA, tivity requized to allow for a normal endocannabinoid response in the hippocampus and mPFC after an aversive training experience. Results “Aversve Inhibitory Avoidance Training Increases AEA Levels Within the Amysdala, Hippocampus, and mPFC. This experiment inves- tigated whether inhibitory avoidance training elevates AEA snd 2-AG levels within the amygdala, hippocampus, and mPEC, and ‘whether the magnitude of the effects depends on the aversive- hese of the training procedure. Rals were trained on an Iltony avoidance task with either a lower (0.35-mA; lower F or higher footshock intensity (045-mA, higher FS) or were ex posed to the experimental apparatus without any footshock (30 FS). Fig, 1 shows that these two FS intensities resulted in di ferent a8-h retention latencies (F337 = 2041, P <-0.0001), For endocannabinoid measurements, rats were killed 10, 30, or rin aller taining. Our findings indicate thal the more aversive training condition elevated AEA, bul not 2-AG, levels in all thee brain zegions. As shown in Fig. 24, wo-way ANOVA for amygdala AEA levels revealed significant FS condition effect (Faso = 1692; P < 0.0001) but no time effect (59 = 0.003; P (98) or interaction between both factors (P50 =0.40;P = 081). Poste test indicated that the higher FS sigoiticantly increased TrovigB Test 800 J ONo rs 8007 QNoFS we owes ggg |Btomres 500 | Mtigher FS = 500 4 Higher FS 400 5 a00 200 Boj 200 a 200 100 © +00 ° — > ance ‘memory. (A) Step-through latences during inhibitory avoidance {eaning of rat trained under thre ilferentcondivons only exposed tothe Cortex without receiving any FS, no FS with lewer FS tens lover with a higher F intens, higher F5).(8) Step-through ltences on = 46- retention test. The increase of FS intent during the traning tal enhanced mmemary consoldation ofthe inhibitory avoidance task? «O05, **P 00% vir the no FS group, “P< 0.05 ve the lower FS group, Rerulte represent mean = SEM (n= 6-7 per group. S34 |v pnas orgies 10.1073 1420285111 amygdala AEA levels at 10, 30, and 60 min after the training tial compared with rats given the lower FS (30 min: P < 0.05; 10 and 60 min: P < (101) oF those unshocked (10 and 30 min: P < 0.05; 60 min: P < 0.04). As shown in Fig. 28, two-way ANOVA for AEA levels in the hippocampus revealed a significant FS con- dition effect (Fay = 14.25; P < 0.0001), no time effect (Fae 1.27; P= 0.29), and a significant interaction between both factors (Faq = 2.73; P = 0.08), Post hoc tests indicated that the higher FS significantly increased hippocampal AEA levels 10 min alter the Uaining trial compared with the unshocked controls (P< 0.01). Two-way ANOVA for AEA levels in the mPFC revealed a significant FS condition effect (Fay P< 01007) but no time effect (F4» = 2.94; P = 0.06) or interaction between both factors (F,.9 = 1.54: P = 0.20; Fig. 20). Post hoc tests indicated that the higher FS significantly increased AEA levels in the mPFC 60 min after the training trial compared with unshocked rats (P < 0.05; Fig. 2C). Rats exposed to the lower FS did not differ in AEA levels compared with unshocked rats ia the three ‘brain areas at any time point. Inhibitory avoidance training did not affect 2-AG levels in any of the brain regions at any time point (S/ Results and Table Si). In plasma, AEA levels were Significantly elevated 60 min afler inhibitory avoidance training with the higher, but not lower, FS (57 Resuls and Table 82). Pesttraining Infusions of the FAAH Inhibitor URBS97 ints the BLA, ‘mPFC Enhance Inhibitory Avoidance Memory via n of CBI Receptors. To determine whether the elevated AEA levels induced by the higher FS contribute to the enhancement of memory of this more aversive training experi- ence, we next investigated whether exogenous augmentation of training-induced AEA levels with posttraining infusions of the FAAH inhibitor URBS97 into the BLA, hippocampus, or mPFC modulates memory consolidation, Our findings indicate that [URBS97 administered into all three brain regions enhances mem- cory consolidation ‘As shown in Fig. 44, 48-h retention latencies of rats admin- istered vehicle into the BLA immediately after training were significantly longer than their entrance latencies during the waining trial ((4 = -3.63; P = 0.004), indicating that the rats retained memory of the shock experience. Posttraining URBS97 into the BLA enhanced 48-b retention latencies in an inverted Usshape relationship (Fag = 3.63; P = 0.02). Post hoc tests in- dicated that the 10-ng dose of URBS97 produced retention la- tencies that were significantly longer than those of rats given vehicle (P < 0.05). Lower or higher doses (3 or 30 ng) were in- elfective. Fig. 3B shows that URBS97 infused posttraining into the hippocampus also enhanced 48-h retention latencies (Fy. = 3.30; P = 0,03). Post hoc tests indicated that the 10-1g dose of ‘URBS97 significantly enhanced retention performance (P< 0.05), with lower or higher doses (3 or 30 ng) being ineffective. Fig. 3C shows that URB597 infusions into the mPFC also en- hanced 48-h retention latencies (Fy «5 = 3.97; P= 0.01), Post hoc comparisons indicated that the 30-ng dose of URBS9 signifi cantly enhanced retention performance (P < 0.05). Lower doses G or 10 ng) were inelfectiv. ‘Next, we examined whether the elevated AEA levels, induced ‘by URBS97 administration, enhance memory consolidation via an activation of CBI receptors, Rats were administered the effective dose of URBS07 together with a nonimpairing dose of the CEL receptor antagonist AM2S1 immediately after the training tal (Our findings indicate that the URBS97 effect om memory con- solidation depends on CBI receptor activity in all three brain regions investigated. Fig. 3D shows 48-h retention test latencies of rals administered URBS97 together with AMDS1 into the BLA. Two-way ANOVA for zetention latencies revealed a significant URBS97 effect (F, = 4.27; P = 0.04), a significant AM25I effect (Fine = 547; P= 003), and a significant interaction between both faciors (Fis, = 545; P = 0.03). Post hoe tests indicated thatA rasta B Hippocampus ; 7 dd oly Effect ofthe level of emotonal arousal ofa inhibitory avoldance tain on AEA levels inthe amgdal, hippocampus, and mPFC, measured at sltferent time points ater the training The higher FS contin induced an increaein AEA levels inthe amygdala (A 1,30, and 6O min after the taining, inthe hippocampus 8) a 10 min ale the Vaining, an inthe mPFC (C) a 60 min afte the Vaining. “P< 0.05, *P-< O01 vs. the eorespanaing no FS group, *P <0.05, "P< 001 vs the corresponding lower F group. Revs reprevent mean retention latencies of rats given URBS97 (10 ng) were significant longer than those of rats given vehicle (P < (005). Retention | tencies of rats given a nonimpairing dose of AM251 (0.14 ng) together with URES? were significantly shorter than those of rats treated with URBS97 alone (P < 0.05), Fig. 3E shows 48-h re- tention latencies of rats given URBS97 and AM251 into the hippocampus, Two-way ANOVA for retention latencies revealed a significant URBSO7 effect (Fy «y = 842; P = 0.006), a significant AMBSI effect (Fey = 11.13; P = 0,002), and a significant in- teraction between both factors (Fy = 10.74; P = 0.002), Post hhoe comparisons indicated that retention latencies of rats given URE597 (10 ng) were significantly longer than those of vehicle- treated rats (P< 0.01), Retention latencies of rats given a non- impairing dose of AM251 (0.28 ng) together with URBS7 were significantly shorter than those of rats treated with URBS97 alone (P< 001). Fig. 3F shows 48-b retention latencies of rals adm istered URBS97 and AM251 into the mPFC. Two-way ANOVA revealed a significant URES97 effect (Fy s» = 5.58; P = 0.03), no AM2S1 effect (Fy, 35 =3.87; P= 0.06), and a significant interaction between both faciors (Fy, = 4.41; P = 0.04), Post hoc compar- isons indicated that retention latencies of rats given URE597 (G0 ng) were significantly longer than those of rls given vehicle (P <0). Retention latencies of rats given a nonimpairing dose fof AM25I (0.28 ng) together with URBS97 were significantly shorter than those of rats treated with URBS97 alone (P < 0.03). ‘An ntact BLA ls Required for Enabling the Traning-Induced increase of ‘AEA Leves Within the Hippocampus and mPFC. It is well-stablished that prefrontablimbic circuits regulate emotional arousal effects fon memory consolidation. In this experiment, we investigated ‘whether functional interactions between the BLA, hippocampus, and mPFC are necessary for modulating endocannabinoid activity in response to an aversive training experience, We made bilateral excitotoxic lesions of the BLA and subsequently measured APA and 2-AG levels in the hippocampus and mPFC after inhibitory fvoidance training. Rats were trained withthe higher FS and killed 10 or 60 min later for hippocampus and mPFC dissection (in the previous experiments, we found that AEA levels in the hippo- ‘campus and mPFC were clevated 10 and 60 min after training, respectively; Fig. 2 B and C). As shown in Fig 4 and B, bilateral [BLA lesions abolished the traiing-induced increases in AEA levels in both the hippocampus and mPFC (hippocampus: ty = 7-11; P < 0.0001; mPFC: 1,, = 2.79; P = 0,02). Bilateral BLA lesions did not affect 2-AG levels in either the hippocampus or mPFC after inhibitory avoidance training (by = 0.41; P = 0.69: fy, 0.12; P = 0.91, respectively; Table 83). These findings in- dicate that an intact BLA is necessary for inducing a training- associated increase of AEA levels within the hippocampus and mPFC, and provide important support for the view that = SEM (n= £8 per grouph functional interactions between these brain regions are re- quired to coordinate endocannabinoid responses after emo- tionally arousing training. ‘An ntact BLA is Required for Enabling the Memory-Enhancing Effect ‘of Pesttraining URESS7 Infusions into the Hippecampus and mPFC. To further investigate the significance of functional interactions between these brain regions in regulating endocannabinoid effects on memory consolidation, we examined whether an intact BLA is also required for enabling the memory-enhancing effect Aan Biman ‘cae lll ee Ae lee oe on FAH inhibitor URB587, either alone or together with the CBI receptor antagonist AMZSI, on 48h retention. Immediate postvaining bilteal Infons of URES8? (URS: 3, 10,030 nh inte the BLA (2), hppocarmpu (8), ‘or mPFC (C) enhanced merry consolidation, Concurrent adminisvation of ‘onimparring dose of AMDS! (010-028 ng pe side) blacked the memory fennancing eects of URBS87 (10 or 301mg) inthe BLA (0, hippocampus (2), fF mPFC {F. Representalive. photomicrographs (Nikon 80" microscope ‘orginal magnifiaton 2») iusvating the placement of the cannula and reese ipintne BLA (6), hippocampus i, and mPFC °P «005, "*P-<001 vs the coresponding vehi (Veh) group: *P < 005, "P < 001 vs. the cr ‘responding AMZ gfoup Result represent mean = SEM (n= 13 per grou. eumoscaenceA siopscanus Bere G 30 7Shencee 29-7DShamLes 60 , [meats makes ove a roo wo Role ofthe BLA in modulating the endocannabinei response to emo tral permanent lesions of the BLA blocked the Urainingsnduced increase of AEA levels in the hippocampus (A) and mPFC {8}, 10 and 60 min afer the raining tial, respectively. "P< O08, =P < 0.0 vs, the coresponaing sham lesion group (Sham Ls). Resus represent mean + SEM {n D mere Dvn [BLAS 20-0059 ” nally arousing inhibitory avoidsnce training inthe hippocsmpus and PEC. 7 per group). Bilateral lesions of the BLA blacked the memory-enhancing effects induced by URBSOY infused into the hippocampus (0 ng C) oF mPFC (0 ng). P05, "Pc Oot vs. the corresponding vehicle group; “P < 0.01 vs. the corresponding BLA lesion group. Results represent mean 1 SEM (n= 10-18 per group) ‘of URBS97 when infused into either the hippocampus or mPEC, Concerning the hippocampus, two-way ANOVA. for training latencies revealed no BLA lesion effect (F;,ae = 1.46; P = 0.23) and no difference between later drug treatment groups (F, 0.28; P = 0.60) of interaction between both factors (Fag = 0.58; P = 041). Two-way ANOVA for 48h retention latencies revealed a significant BLA lesion effect (Frac = 1210; P (0.001), a significant URBS97 etfect (Fxg = 4.13; P = 0.048), and a significant interaction between these two factors (P,.q = 484; P= 0.03), As shown in Fig. 4C, URBS97 (10 ng) infused into the hippocampus of sham-esioned rats enhanced 48-h retention relative to their corresponding vehicle group (P < 005). Further- ‘more, and importantly, BLA lesions blocked the memory-enhancing effect of URBS07 infused into the hippocampus (P < 0.01). Next, we examined URBS97 effects in the mPFC. Two-way ANOVA for taining latencies revealed no BLA lesion effect ia = 199; P = 0.17) and no difference between later drug treatment groups (Fy «1 = 1.04; P = 0,31) oF interaction between both factors (F,4: = 0.46; P = 0.50), Two-way ANOVA for 48-h retention test latencies revealed a significant BLA lesion effect (Fa = 11.80; P = 0.001), a significant URBS97 effect (Fy, 6.26; P = 0.02), and a significant interaction hetween these two factors (Fy; = 4.91; P = 0.03). As shown in Fig. 4D, post hoc tests indicated that URB597 (30 ng) infused into the mPFC of sham-lesioned rats enhanced 48-h retention compared with the vehicle group (P < 0.05). BLA lesions blocked the memory- enhancing effect of URBS97 infused into the mPEC (P < 0.0). “Thus, out findings indicate that the BLA is critically involved in regulating the memory-cnhancing effects of endocannabinoids in the hippocampus and mPRC, Discussion The present findings indicate that a dynamic network involving AEA signaling within the BLA, hippocampus, and mPFC mod- tlates the consolidation of memory for emotionally arousing training, Moreover, our results suggest that the BLA. plays a crucial role in coordinating hippocampal and mPFC endocanna- Dinoid activity within this nevrocircutry during the consolidation cof memories for aversive experiences, ‘Rats trained under a higher arousal condition retained better memory of the aversive event than did rats trained under a lower arousal consition, and this effect was paralleled by an increase in AEA levels within the amygdala, hippocampus, and mPFC. In contrast, 2-AG levels were unaffected. The increase in AEA content showed a temporal-dependent response during the early phases of memory consolidation, with a sustained increase of AEA levels in the amygdala (from 10 to 60 min alter training) fand an increase only at 10 and 60 min afler training in the S36 |v pas orgies 10.1073 1420285111 hippocampus and mPFC, respectively. The increase of AEA signaling induced by postraining administration of the FAAHL inhibitor URBS97 enhanced 48-h inhibitory avoidance reten- tion performance through indirect activation of CBI receptors. AGditionally, we found that this training-induced increase in endocannabinoid neurotransmission depends on the integrity of functional interactions between these brain regions, because disruption of BLA activity resulted in a loss of the training- induced AEA response as well as the URBS07 effect in the hip- pocampus and mPFC Tis well-known that CB1 receptors, highly expressed throughout the limbic stem (20), modulate neuronal signaling and synaptic plasticity (21, 22), thereby regulating emotional behavior and memory processes activated by emotionally arousing events (8, 9 23-26). However, studies have reported conflicting findings con. ceming cannabinoid effects on memory consolidation (fora review, see ref. 6), Cannabinoid effects on memory are highly dependent on the level of emotional arousal induced by the stessflness of the experimental condition used inthe different studies (6, 13, 27). (Our findings are consistent with prior evidence in indicating that inhibitory avoidance training induces the release of the endo- cannabinoid ABA in the amygdala, hippocampus, and mPFC only under highly aversive conditions. This endocannabinoid response was not detectable when animals received a lower FS. “To investigate whether the endogenous release of AEA after Inhibitory avoidance training is linked to the stabilization of the ‘memory trace, we tested whether pharmacological amplification of AEA signaling would affect memory consolidation for in- hibitory avoidance training. Our findings indicate that the FAH, inhibitor URBS97, which increases endogenous ABA levels in the synaptic cleft (28), enhanced 48-h retention when infused into the BLA, hippocampus, or mPFC after training. URBS97 also affects, to a certain extent, the hydrolysis of other fatty-acid ethanolamices that do not bind to CBI receptors. Our result that the URB597 effect was blocked by concurrent administration of the CBI receptor antagonist AM251 indicates that the URB597- induced memory enhancement involves an increase in ABA signaling and activation of CBI receptors. Ou findings are, thus, consistent with previous evidence indicating that a blockade of endocannabinoid! neurotransmission in the hippocampus (29), amygdala (30), and mPFC (12) impairs the consolidation of aver- sively motivated taining experiences. ‘We previously reported that endocannabinoid effects on ‘memory consolidation depend on the level of emotional arousal at the time of encoding (6, 13, 27) Such findings suggest that an interaction with stress hormones is important in determining the modulatory effects of cannabinoids on memory processes. The BLA is known to influence the association of environmentalinformation with emotional significance by engaging stress- related hormones and neurotransmatters to modulate memory procestes (31, 32). We have previously shown that intra-BLA ‘administration of the CBI receptor antagonist AM251 blocks the ability of systemically administered corticosterone to facilitate the consolidation of memory of inhibitory avoidance training (8) Such findings indicate that glucocorticoids recruit endocanna- binoid signaling in the BLA to modulate memory consolidation (G3, 34), Our current results add to this evidence the findings that, after a stressful event, the endogenous release of AEA in prefrontalsimbic areas enhances memory consolidation for aver- sive experiences via an activation of CBI receptors. Findings inthe literature indicate a direct correlation between FS intensity and norepinephrine release in the amygdala (35). Therefore, iis ike that a certain degree of emotional arousal, induced by the higher FS intensity leads to an optimal activation of stress mediators that is necessary to recruit the endocannabinoid system and, ultimately, ‘mediate the formation of a strong memory trace Considerable evidence indicates that the BLA modulates memory consolidation by coordinating the activation of other brain regions and by regulating memory processes that take place elsewhere in the brain (I, 31, 36-38). This modulation might involve either direct or indirect neural connections to various limbic structures, including the hippocampus and mPFC (14, 15, 39, 40), Our findings provide additional evi- lence supporting this implication in indicating that a functional BLA is required to enable the modulatory effect of ABA within the hippocampus and mPFC on memory consolidation. These findings are also consistent with evidence that interactions of the BLA with the hippocampus (18, 41, 42) and mPFC (19) regulate emotional arousal effects on memory consolidation ‘Our findings fucther indicate that, depending on the brain re- gion, AEA increases with distinct temporal windows during tearly phases of memory consolidation, We show that activation of the BLA regulates the endocannabinoid response in the hippocampus and mPFC, as permanent lesions of the BLA blocked the waining-induced increase of AEA levels and the smemory-enhancing effects of URBS97 in both the hippocampus and mPFC, Previous findings suggest that the initiation of a strong emo- tional experience activates memory-related neuroplasticity in the amygdala and hippocampus whereas suppressing PFC function- ing (43). Such a model suggests a rapid activation of the BLA and hippocampus after a stressful experience, followed by an inhibitory phase. Moreover, there is evidence that corticosterone rapidly but transiently enhances the frequency of miniature postsynaptic currents (mEPSCs) in the hippocampal CAL area (Gi), whereas it also rapidly increases mEPSCs in the BLA, but lover a longer time window (45). Accordingly, we found a rapid dand transient inerease of AEA levels in the hippocampus and 4 rapid but enduring increase within the amygdala. It is well- accepted that CBL receptor activation mediates a decrease in GABA release in the BLA (46). Thus, the rapid increase in amygdala AEA content observed in our study may represent one of the most rapid actions of BLA activation after experiencing an aversive event. In particular, the AEA.mediated activation of BI receptors in the BLA may decrease feedforward inhibition ‘vig inhibitory intemeurons, thercby increasing the activity of BLA projection neurons (47), On the other hand, basal mPFC activity is known to provide an inhibitory influence on BLA ac- tivity (16, 48), whereas stress and glucocorticoids are known to suppress mPFC functions (49). It has previously been reported that glucocorticoid infusions into the mPFC enhance memory consolidation of an inhibitory avoidance training via functional and bidirectional interactions with the BLA (19). Thus, the present findings provide additional evidence that increased BLA Activation during emotionally arousing conditions (38, 49, 50) enhances the consolidation of different kinds of information via its projections to other brain regions Considered together, our findings indicate that, asa response to emotional events, endocannabinoids are released withia the prefrontal-limbie neutocircuitry, where they enhance the con- solidation of memory for emotionally arousing events. These findings are of crucial importance, helping to understand the neural underpinnings of the temporal interactions between limbic regions after experiencing a stzssful event and shedding light on the neural mectianism involved in the process of aversive memory formation, Materials and Methods Animals. Male Sprague-Davley rts (320-370 g atthe ime of behavioral experiments; Charles iver were housed indvialy in atemperaturecontvoled (29-11 "0 vivarum room and maintained under 3 12424 hana cle (720 AM to 700 PI Iights on. Food and water were avaable ad itu All procedures were in complace with alan law (DL 26/4, a European Union ‘Seecive QOTOSHE), the Decavaon of Hebinkl, and Guldeline forthe Care and Use of Mammals in Neurosinee and Behavioral Research) Surgery, Rats were anesthetized with sodium pentobarbital {0 mag, Lp). siven atropine sulfate (04 mghc, ip) to maintain respiration, and sub Fequertyinjectea with 3m ofsline fc) to facitate clearance of rugs and prevent dehydration. The ras were then placed ina sereotaie rare (Kopt Instruments, and two sailestee! guide cannulae (23-gauge) were implanted bistealy, with the cannula tps 2 mm above the BLA 1S mm ‘coordinates anteroposterior (AP), 2.8 men fom Bregma; mediolateral (MU, {35.0 mm from micline; dorsoventral (OV), 63 mom from sul surface] of 115 mm above the Cl region of the dovsal hippocampus (11 mm coor inate AP, “34 mm: ML, +1.8 mm; DY, -2.7 mm] othe prelimbic region ef TRPFC [17 mm: coordinates: AP, 43.7 mm: ML, 20.7 mm; DV, -24 mm] (19, 52, 53). Siylets (15. or Tl-mmlong 00 insect election pis) were Inverted into each cannula to maintain patency The NMIOA Solution Sigma-Aldrich; 25 ug in 02 pL of phosphate buen) ‘wae baceiled into 10s Hamikon miconringe gauge) driven bya te reotanc miniump (Stocking) and inured into the BLA over 8304 period (4 “he needle wat retuned i place for an adtlonl 3 mint optimize fon, Rats were allowed to recover fom suger for 10d before traning and handled see times for 1 min before taining Inhibitory Aveldance Apparatus and Experimental Procedures. Rats were ined nan inhibitory avoldance apparatus for decals, sees Materials and [Methoor and ref. 8). Nonoperated rats ware place in the starting Con partment and alowed to treely explore the apparatus. After the at stepped Ineo the dark Shoe) compartment the sling dor was dosed anda single inescapable FS with an intensity of either 035 mA Jower #3) oF 0.45 mA (fugher #5) was delvered for 1+ Other rats were unshocked {no FS} upon entering the dark compartment, Cannulated rats received higher FS in tenses (045-085 mA) to ensure memory in all experimental foups. Rats vere removed ftom he shock compartment 15 slater and returned to thet home eager. Some groups of rats wer killed at 10, 3, or 60 min ater the {raining tral for endocannabinoid detection. Others were lef undsturbed pd tested fr retention 48 later For retention testing the rat was placed ineo the starting compartvent and the latency to reenter the sock com partment with al four paws var meatured (maximum latency of 600) Drug restment. URB597 (Tocris 3, 10, oF 307g) warinfured postraining into ‘he BLA, hippocampus, of mPFC To examine whether the URBSST effec wat mediates by acivation of CBI receptors, other gtoups of rat received the tecve dose of URESS7 (10 ng BLA and hippocampus, 30 ng mPFC) concur ‘enti with a nerimpairng dove ofthe CB1 receptor antagonist AMIS} [Toc (0. ng for BLA (i, 0.28 ng fr hippocampus ad mPFC]. Doves were based on pilot experiment performed nou laboratory. ltetal shan oF Builesoned Fats were administered the effective dose of UR8S97 into ether the hippo «arpus or mPFC Drugs were sce in 5% (volvol PEG, 5% (volvl Twen 430, and 80% (volvo saline The infusion procedure i aetiled in S Materia tnd Methods barbital (100 mg/kg, ip) and perfused Wranscardilly wth 09% saline eumoscaencePNAS The brains were removed to verily cannula placement and the size and location of lesions (for detail see 5) Materials ana Method!) Endecannabineld Measurement A 10,30 oF 60 min aftr inhibitory avoidance ‘raining (oF, lower orhigherF,ratewere decapitated and the amygdala, hippocampus and mPFC were rapidly elsected for AEA and 2:AG measure rent (5) Statics Oata ar expressed az mean 1 SEM. Inhibitory avoidance atences tt well at endocannabinad levels were analyzed with one- or tworvey esau 80) Memon —A ena of onset, cence 27551) 288-25 Hesougn DOI} Making lng memares Revemberng esha Pec at dead Se Uk osu Bate as faotendsal # MeGaugh J GOI!) Meany moduli. Behav Neuere! 12 Sota PE 2008 Courinatin of multe memory sper. Neusbis eum Mem eppaseee, Proc aca USA Yeung 20978671 2. Gone sera Kana ath) Endcanesindsnedited ogre ago 7a. 09) Endocannainclr nthe 1 bata aygls It ewe OSA ou) a8, 4. Mariana Ge 2002 Ihe eraoperous anil em cont erie maar: Nature e037 500530, to, Aleut Dor) ee ef enobinos naling emainal and nen-eatinal 1. De Olea Aue eno Diehl. uf A 200 Diferent role ofthe ‘meena newb! ann Sash 1a. fubeart 5 Mayer © ooh M BoD) Iasurent ot cna receptor inthe 12, Campolnge ta Ot) Novelvndved emotional wou modus canatinld fees on recogition maar and sreeceea sey eurprehphameee ay ppotmpal domains ar hypetbalamie Behar stems Bi es Bin Rs Re Sezai 15, Paramen Ranks Seande neu Pkinen A199) Projection fame 16. Lig Pete Jar Pae R. Part D 2008) ronal contol ofthe ayia heures 248-137, 1. Peecaray 4 Ves (198) lctaphylagil ud ofthe respon of meal 1, oarensal hy Mcnugh J (29) Basar arya eos Expdieur 90h, 19, Raszena Bs] cuccorcod tec on marr onalton sepend Sy) arse! 2905) 12996508. 20, Tur, tov Safudefere MC Mae, Walker M (38) nmuraktenemil ‘Tevinsion of oranda centre the canal nets stems ure one esp)350- stone frend TF 92) Mull fro of endocansbin 24, Gaonr EAs | Goo} Cambio eepor sito ne Bult Solsnce) Neuve 2996) ore He88 Toni eel 201) Camabnadtraromion nthe buaater amygdla modules Sivas sot ‘tay moss recon mete at epenrgom enrol ware 2a, Rabun e306) Modubtin 9 sie tough leckade of sande ray nat ea 970-0 SSE |_ vpn ores 10.1073 1420285111 ANOVAS. The source of the detected significances was determined by ACKNOWLEDGMENTS. This study was supported by grants from the Human Frontier Science Program {RGYOO77), Mini'y of Education, Universities and Researen (tal) MIUR (RBFRTOXKHS; 20108N3MXN_003) (to PC}, and Vigon’ Project [The German Academic Cxchange Service (PAAD}] Wo GS. and PC). 29, de Ofvers Ahr La. 00) Sires reson res the hippocampal endo Bucher © bale & Martin Paani My, Bia #2000) Aoesive memory ‘slstn Lea Mer aeeaneae 2x, MeGaugh it Ronendalb 200 feo adel es hormones informing sing Imemoresin te brn Cut Opn Neurbi 2205-798, 22, Rasrenal Mtwen , hat 5 08) Sess meer nthe yg Nat ov howard thn 22, He MN Kaas ine) cen 8S Be) ap etn in nb o (Seaman crrt by ghewetndhemonet ine Pychonervondinon? Ses pny 2a5)791-73, 2. Quart Galen Reonendsl3, McGaugh (198 Noresneptine eles sowartenv, 26 Hermans Ue 5 2010 How te amygdsa affects emotional memery by senag 27. Pakard MG, rgd Je BOG Amypaae and "emotanal” medlaton of et 58, MeGaugh JL BOM) The emyaesle medals the conalinon of memati of pape He Pare OW) Parc orapt< nto he srygels forthe satan, onaone fear jel Re Sap 03 [Npvorrpus nce by cate fer aravanng, hearse 2) We 622, ‘erage interes eral: ean’ of meray coldston ded by hips. ‘evpal sheets ceptor avon Pros Hat Ae USA 6 EYE «2, Dieund OM, Cpe AM, Pk eleven Zales Pm 0M The tempera 1. Kart t 3 Bea) Mirtle receptor a diparate fornongenomc ‘meduaion of pecenpalghtamsie anniaionby cericrerone Peal eed Se Usa vous zoe 3207 45, Kant berger trarann 6 Seht Goth M QOT) Metals of amyl ‘pons tothe Se homore cocoons Pro Na Aad 5 USA TB} Cldeanoshnodedaed corel ef epic tnsmsan Pye! er 90) 308-98, scones 48, Qu Genie OR 2O0) nhtion ofthe rgd Key to pathologie Ate? 49, hat ets 200 esa prerontl caren determines hon seer onotity 50. Daan Rnd flety ar ecte ye: Rle of peel eax rar ‘hn el pct sth 6098, 5. NiionalReseren Count 203) Guts forthe Care and Use of Maras in 52, Moat eta GOT} Gucci trac th te ippocnpalensocstnainld tosaraso isos 52. Campolonge 200 Fatunduce ity fee toy tarot enhances emery coriation Proc Ma eed St USA 03}40D-800 sRasreasalh, tfan OX Buarany), De Quarnn 0), Mecaugh 280) The 15, He’, 2011) Propel enhances meron fematon va neraton with the endecannsinid em, Anerhaiagy NAG} TL-88