- Email: [email protected]
High-frequency stimulation of the hippocampus blocks fear learning sensitization and return of extinguished fear
Neuroscience 286 (2015) 423–429
HIGH-FREQUENCY STIMULATION OF THE HIPPOCAMPUS BLOCKS FEAR LEARNING SENSITIZATION AND RETURN OF EXTINGUISHED FEAR O. DESCHAUX, a* O.-C. KOUMAR, b F. CANINI, c J.-L. MOREAU d AND R. GARCIA e
occurs following a traumatic event and is characterized by intrusive recollections of the event, avoidance of reminders of the event and hyperarousal. Moreover, PTSD is often co-morbid with phobias (Engdahl et al., 1998). Fear learning sensitization (i.e. potentiation of fear acquisition by initial fear experience) and fear return (i.e. reactivation of extinguished fear) have been suggested to contribute to the development (Dykman et al., 1997; Rau et al., 2005) and persistency (Milad et al., 2009) of phobias in PTSD, respectively. Because the hippocampus (HPC), which is dysfunctional in PTSD (Bremner et al., 2003), is involved in conditioned fear-related phenomena (Bast et al., 2001), fear learning sensitization and fear return may be associated with HPC dysfunction. Also, fear return in PTSD is associated with HPC deactivation (Milad et al., 2009), whereas activation of this area in healthy individuals is associated with less fear return (Milad et al., 2007). It is unknown whether a treatment, known to activate the HPC, can also reduce fear learning sensitization. Because this idea cannot be directly tested in humans, rodent studies can help to elucidate this issue. Fear learning sensitization and fear return have both been widely reported in rats. First, initial fear conditioning (FC1) (context-shock association) can sensitize subsequent fear acquisition. In these studies (e.g., Rau et al., 2005, 2009; Rau and Fanselow, 2009; Ponomarev et al., 2010; Long and Fanselow, 2012), rats are initially exposed to 15 shock administrations in context A; 24 h later, they are placed in context B and exposed to a single shock administration. When re-exposed to context A, animals typically show pronounced freezing response proportional to the 15 shocks, whereas when re-exposed to context B, they do not display freezing proportional to the single shock, but higher levels of freezing, indicating fear learning sensitization. Moreover, in a study, in which we used a similar paradigm, we found that the initial conditioning (with 15 shock administration) induces long-term reduction of activity in the dorsal HPC (Spennato et al., 2008). This HPC deactivation may facilitate fear leaning sensitization. Second, extinguished fear responses can return either spontaneously (spontaneous recovery) or following provocation. For instance, conditioned fear can be fully recovered in rats 10 days after its extinction (Quirk, 2002). Diverse behavioral paradigms have also been developed to provoke fear return. For example, fear returns in rats exposed to unsignaled shock of initial intensity (Rescorla and Heth, 1975; Bouton and Bolles, 1979) or lower intensity (Spennato et al., 2008; Deschaux
a
Institut de Biologie de Valrose, UMR CNRS 7277 – INSERM 1091, Universite´ Nice Sophia Antipolis, 06108 Nice, France b De´partement des Sciences de la Vie, Universite´ Nice Sophia Antipolis, 06108 Nice, France c
Institut de Recherches Biome´dicales des Arme´es, De´partement des Environnements Ope´rationnels, 38702 La Tronche, France d
Pharma Division, Hoffmann-La Roche, Basel, Switzerland
e
Institut de Neurosciences de la Timone, UMR7289, Aix Marseille Universite´ & CNRS, 13385 Marseille, France
Abstract—Patients with post-traumatic stress disorder (PTSD) present hippocampal (HPC) dysfunction, which may facilitate fear-related phenomena such as fear learning sensitization (i.e. potentiation of fear acquisition by initial fear conditioning (FC1)) and fear return (i.e. reactivation of extinguished fear). Fear return is sensitive to HPC high-frequency stimulation (HFS) in rats. The goal of the present study was to examine whether fear learning sensitization is also sensitive to HPC HFS in rats. We found in control conditions that, after FC1 (with 15 shock administrations) and extinction, conditioning in a different context with one shock administration was potentiated (proactive effect) and associated with fear return in the initial context (retroactive effect). Both phenomena were prevented by HPC HFS applied before the second conditioning. We also found that the effect of HPC HFS on fear learning sensitization required initial extinction. These findings suggest a pivotal role of the HPC in preventing proactive and retroactive effects of successive fear conditionings. These data also support the concept that HPC deactivation may be involved in fear learning sensitization and fear return in PTSD patients. Ó 2014 Published by Elsevier Ltd. on behalf of IBRO.
Key words: conditioned freezing, fear extinction, fear return, hippocampus, rats.
INTRODUCTION Acquired fear is at the root of several anxiety disorders, including post-traumatic stress disorder (PTSD), which *Corresponding author. Tel: +33-4-92-07-61-47; fax: +33-4-92-0768-85. E-mail address: [email protected] (O. Deschaux). Abbreviations: FC1, initial fear conditioning; FC2, second fear conditioning; HFS, high-frequency stimulation; HPC, hippocampus; PTSD, post-traumatic stress disorder; SCP, sub-conditioning procedure; VLFS, very low-frequency stimulation. http://dx.doi.org/10.1016/j.neuroscience.2014.12.001 0306-4522/Ó 2014 Published by Elsevier Ltd. on behalf of IBRO. 423
424
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
Table 1. Timing (from Day 1 to Day 27) of the treatments applied in the three contexts (A, B, C) for each group of rats [No-FC1: rats submitted to the second fear conditioning (FC2), but not the initial; FC1: rats submitted to the initial fear conditioning (FC1), but not the second; FC1 + FC2: rats submitted to both sessions of fear conditioning; VLFS: rats submitted to FC1, extinction training, and very low-frequency stimulation (VLFS) of the hippocampus, prior to FC2; HFS: rats submitted to FC1, extinction training (six sessions: two sessions/day 3 days), and high-frequency stimulation (HFS) of the hippocampus, prior to FC2; No-Ext: rats submitted to FC1 and HFS of the hippocampus prior to FC2, but not to extinction training] Group
Day 1: Days 2–4: FC1 (0.7 mA) Extinction (2 sessions/d)
Day 25a: Day 25b: Day 26: Brain stimulation (BS) FC2 (0.7 mA) Sensitization test
Day 27a: Generalization test
Day 27b: Extinction reversal test
No-FC1
Context A (no shock)
Context B (No BS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
FC1
Context A (15 shocks)
Context B (No BS)
Context C (0 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
Context B (No BS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
Context B (BS: HPC VLFS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
Context B (BS: HPC HFS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
Context B (BS: HPC HFS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
FC1 + FC2 Context A (15 shocks) VLFS
Context A (15 shocks)
HFS
Context A (15 shocks)
No-Ext
Context A (15 shocks)
Context A (30 min/ session) Context A (30 min/ session) Context A (30 min/ session) Context A (30 min/ session) Context A (30 min/ session)
et al., 2011, 2013; Zheng et al., 2013). Interestingly, this fear return is prevented by high-frequency stimulation (HFS) of the HPC (Deschaux et al., 2011) or a pharmacological treatment fluoxetine (Deschaux et al., 2013), which has been shown to suppress HPC deactivation (Spennato et al., 2008). From these observations, we hypothesized that if fear learning sensitization and fear return are both facilitated by HPC deactivation, HPC HFS, which is known to induce intra-HPC LTP in behaving rodents (Kamal et al., 2014; Wall et al., 2014), would impair fear learning sensitization as it blocks fear return. We used a behavioral paradigm provoking fear learning sensitization and fear return in which rats were submitted to initial conditioning with 15 shocks in context A, followed (or not) by an extinction training session. Rats received HPC HFS (or not) in context B, before being exposed to a single shock in context C. Fear learning sensitization and fear return were evaluated in contexts C and A, respectively. We also tested possible fear generalization in context B, wherein animals had no shock experience.
EXPERIMENTAL PROCEDURES Animals Forty-two adult male Wistar rats (De´pre´, Saint Doulchard, France), weighing between 280 and 350 g, were used. They were housed in individual home cages in a wellventilated and temperature-regulated (22 ± 2 °C) room with food and water ad libitum. A 12–12-h light/dark cycle (lights on from 7 a.m.) was imposed. Experiments were performed in accordance with the European Community Guidelines on the Care and Use of Laboratory Animals (86/609/EEC).
Electrode implantation For brain stimulation, rats were implanted prior to contextual fear conditioning. For this purpose, they were anesthetized with pentobarbital (65 mg/kg, i.p.) and were fixed in a stereotaxic frame. Stimulating electrodes made of twisted silver wires (90-lm diameter) were bilaterally lowered into the CA1 region of the dorsal HPC (3.2 mm posterior to bregma, 1.6 mm lateral to midline, 2.5 mm from dura). This area was chosen based on our previous study showing that its stimulation modulates (facilitates or inhibits) fear return as a function of the frequency used (Farinelli et al., 2006). The entire miniature system was fixed onto the skull by three screws and dental cement. Rats were then left undisturbed in their home cage for 5–7 days to recover before the experiments were started.
Apparatus All behavioral devices were purchased from Imetronic (Pessac, France). Three contexts (A, B and C) were used. Context A consisted of a gray plastic cuboid cage (each side, 40-cm length and width and 45-cm height), with a grid floor composed of 0.5-cm-diameter stainless steel rods placed 1.5 cm apart. Context B consisted of a clear plastic cylindrical cage (25-cm diameter, 25 cm height) with a plastic floor covered with animal bedding. Context C consisted of a clear plastic cage with a hexagonal prism shape (40 cm 30 cm 30 cm 20 cm 12 cm 12 cm breadth, 45-cm height), with a grid floor as in context A, and an additional background noise that was continuously provided (60 dB). White ceiling lights were on for context A, and a red light was on (with white ceiling lights off) in the room for contexts
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
425
Fig. 1. Histological representation of a coronal section showing placements of stimulating electrodes in the dorsal hippocampus.
B and C. During each behavioral test, each cage was placed inside a clear plastic chamber that was soundattenuating and temperature-regulated. Behavioral and brain stimulation procedures For behavioral tests, rats were submitted to a 27-day experimental procedure (Table 1). On Day 1, nonimplanted rats were randomly divided into three groups of seven rats each, comprising rats subjected to the second fear conditioning (FC2), but not to the first (No-FC1 group), rats subjected only to the FC1 (FC1 group), and rats subjected to both sessions of conditioning (FC1 + FC2 group). On the same day (Day 1), implanted rats were also randomly divided into three groups (seven rats each): all groups were subjected to FC1 and FC2, but two groups received either HPC very low-frequency stimulation (VLFS group) or HPC HFS (HPC group) after extinction training. The third group did not undergo
extinction training but received HPC HFS before FC2 (No-Ext group). All rats (non-implanted and implanted) were placed in context A (the cuboid cage, which was cleaned with a 50% ethanol and lemon scent solution before each rat was introduced). Three minutes later, all rats, except those in the No-FC1 group, were subjected to 15 electric foot-shocks (0.7 mA, 1 s; intertrial interval: 240–420 s). Sixty seconds after the last foot-shock application, they were removed from context A. Rats in the No-FS1 group were left undisturbed in context A for the same time without any shock presentation. On Day 2, all rats, except those in the No-Ext group, were re-exposed to context A twice (with 3–5 h interval) for 30 min each time (exposures A1 and A2). This schedule of context re-exposure was repeated on Days 3 and 4 (exposures A3 and A4 and exposures A5 and A6, respectively) to induce maximized extinction of conditioned freezing to context A.
Fig. 2. Percentage of freezing behavior (means ± SE) during the first 10 min of the six sessions (A1 to A6) of extinction training (after FC1) for NoFC1, FC1, FC1 + FC2, VLFS, and HFS groups. ⁄p < 0.05: statistically significant difference from the No-FC1 group.
426
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
Fig. 3. Percentage of freezing behavior (mean ± SE) during the first 10 min of tests in context C (test for fear generalization), in context B (test for fear generalization), and in context A (test for fear return) for all groups of rats (No-FC1, FC1, FC1 + FC2, VLFS, HFS, and No-Ext). ⁄p < 0.05 and ⁄⁄⁄ p < 0.0001: statistically significant difference from the No-FC1 group; #p < 0.05 and ###p < 0.0001: statistically significant difference from the FC1 group; ¤p < 0.05 and ¤¤¤p < 0.0001: statistically significant difference from the HFS group; £p < 0.05: statistically significant difference from the No-Ext group.
On Day 25 (Day 25a), all rats were placed in a new context (context B, the cylindrical cage that was cleaned with 70% ethanol and had animal bedding on its floor that was changed before each rat was introduced). During the 60 min of exposure to context B, implanted rats were connected to stimulating cables, which were relayed at the top of the experimental box by a multichannel rotating connector allowing free movement. After the first 30 min of acclimation to context B and to the connected cables, VLFS rats received VLFS (100 pulses of 500 lA delivered at 0.2 Hz), whereas rats in the HFS and No-Ext groups received HFS (100 pulses of 500 lA delivered at 100 Hz). These parameters of stimulation were chosen according to our previous study, showing that VLFS does not affect HPC-mPFC synaptic efficacy, whereas HFS enhances this synaptic efficacy in a long-term manner (Farinelli et al., 2006). When animals were removed from context B, they were placed in their home cages in the animal room where they were left undisturbed for 6 h. After this delay (Day 25b), they were placed in a new context (context C, the prism-shaped cage, which was cleaned with a 50%-ethanol and lemon scent solution before each animal was introduced). Three minutes later, all groups of rats, except the FC1 group, received a single electric foot-shock (0.7 mA, 1 s). Sixty seconds later, they were removed from context C. On Day 26, rats were tested for fear conditioning in context C for 30 min. On Day 27 (Day 27a), they were tested for 30 min for fear generalization in context B. Six hours later (Day 27b), they
were re-exposed to context A for 30 min to test fear return. Behavioral scoring, statistical analyses and histology The behavior of each animal was continuously recorded. Conditioned fear was assessed by measuring freezing behavior (Fanselow, 1994), which was scored using a time-sampling procedure. The time spent freezing was measured during the first 10 min of each context re-exposure (contexts A, B and C) because all rats were fully awake during this period. Data were analyzed by two-way ANOVAs for repeated measurements, followed by Fisher’s PLSD post hoc tests, as indicated in the text. Data are presented as mean ± SEM. Values were considered significant if p < 0.05. Upon completion of the experiments, histological analysis was performed as detailed elsewhere (Farinelli et al., 2006).
RESULTS Histological controls revealed that the tips of electrodes were located in the CA1 region of the dorsal HPC in all implanted rats (Fig. 1). Extinction training Fig. 2 (extinction training in context A) shows that during the first 10 min of the first session of context re-
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
exposure (session A1), levels of freezing were high in rats submitted to the 15-shock conditioning (FC1, FC1 + FC2, VLFS and HFS groups), but there was no freezing in rats that did not receive shock (No-FC1 group). Extinction training (sessions A1-A6) of the conditioned rats resulted in a progressive suppression of freezing behavior. A two-way ANOVA performed on these data (five groups, six sessions) revealed main effects of group (F4,30 = 7.13; p < 0.001) and session (F5,150 = 74.41; p < 0.0001). A significant Group x Session interaction was also found (F15,150 = 5.46; p < 0.0001). Post hoc Fisher’s PLSD tests indicated that each conditioned group differed from the No-FC1 group during the first extinction session (A1: all ps < 0.001). Another two-way ANOVA on data from the four conditioned groups (FC1, FC1 + FC2, VLFS, and HFS, six sessions) showed that there was no effect of group or interaction between group and session, indicating that the four groups similarly extinguished their conditioned fear. Fear sensitization, generalization and return tests Fig. 3 shows that, during re-exposure to context C, only rats that were submitted to FC2 (No-FC1, FC1 + FC2, VLFS, HFS, and No-Ext) displayed conditioned freezing, which was potentiated, except in rats that underwent extinction training and received HPC HFS (HFS group). Rats that extinguished the initial fear and were not submitted to FC2 continued displaying low levels of freezing during their re-exposure to context C (FC1). During re-exposure to context B, none of the six groups showed significant conditioned freezing. Finally, during re-exposure to context A, all groups that were conditioned twice, except those submitted to extinction training and stimulated with HFS (HFS group), presented high levels of freezing behavior. Rats that were conditioned once, either during the FC1 only (FC1 group) or the FC2 only (No-FC1 group), did not exhibit freezing in context A. A two-way ANOVA performed on these data (six groups three contexts: C, B and A) revealed main effects of group (F5,36 = 24.63, p < 0.0001) and context (F2,72 = 60.57, p < 0.0001). Group x Context interaction was also significant (F10,72 = 10.18, p < 0.0001). Post hoc Fisher’s PLSD tests showed that, during the sensitization test, rats that were not submitted to FC1 (No-FC1 group) and animals that underwent extinction training and received HPC HFS (HFS group) did not differ from each other, but each differed from each of the other groups (FC1: both ps < 0.05; FC1 + FC2, VLFS and No-Ext: all ps < 0.05). In addition, during the same test, the FC1 group differed from the FC1 + FC2, VLFS, and NoExt groups (all ps < 0.0001), but the latter three groups did not differ from each other. During the generalization test, no group differed from the others. Finally, during the fear return test, the three groups with very low levels of freezing (No-FC1, FC1 and HFS) did not differ from each other, but each differed from each of the 3 other groups (FC1 + FC2, VLFS
427
and No-Ext: all ps < 0.0001). However, rats that did not undergo fear extinction (No-Ext) differed from both the FC1 + FC2 and VLFS groups (both ps < 0.05).
DISCUSSION In the present study, initial conditioning (with 15 shocks in context A) induced sensitization of subsequent conditioning (with a single shock in context C), independently of extinction training (in context A). We demonstrated for the first time that HPC HFS, applied several hours before the second conditioning, prevented this fear learning sensitization. We also observed that the above paradigm produced fear return when rats were re-exposed to context A. This context-specific fear return (no fear generalization was found) was also prevented by HPC HFS. Prior stress can potentiate subsequent contextual and auditory fear conditioning (e.g., Baratta et al., 2007) or contextual fear conditioning, but not auditory (e.g., Cordero et al., 2003). In these studies, the stressful situation was different from the fear conditioning paradigm itself (uncontrollable tail-shock or restrain stress versus foot-shock). However, we based the present study on the enhancement of contextual fear learning caused by prior exposure to several shocks. As reported by other authors (e.g., Rau et al., 2005, 2009; Rau and Fanselow, 2009; Ponomarev et al., 2010; Long and Fanselow, 2012), the advantage of such a procedure is that it works for both contextual and auditory fear conditioning. In addition, although the mechanisms underlying fear learning potentiation remain unknown, it resists to fear extinction and to amnestic (N-Methyl-D-Aspartate antagonist) treatment (Rau et al., 2005). However, here we observed that this phenomenon did not resist to HPC HFS. Interestingly, HPC HFS specifically impaired fear sensitization, without affecting fear learning per se. In context C, rats that received HFS displayed levels of freezing behavior similar to those of the ones submitted only to a single shock. In other words, rats that received HPC HFS acquired the second conditioning, without any sensitization (no proactive effect). Surprisingly, fear sensitization was still present in rats that received HFS, but were not trained for extinction, revealing the importance of extinction to obtain the effect of HPC stimulation. We had shown that after conditioning with several shocks and fear extinction, rats can express fear return when retrained with a shock intensity that is too weak to induce, by itself, significant fear conditioning (Spennato et al., 2008). This protocol, which we named sub-conditioning procedure (SCP), led to similar findings when we used an auditory-cue fear conditioning task (Deschaux et al., 2011, 2013; Zheng et al., 2013). We also had found that HPC HFS, applied in the case of auditory fear conditioning, suppressed SCP-induced fear return (Deschaux et al., 2011). The present study, using a different paradigm, extends this observation. The fear response during the fear return test corresponded neither to fear generalization, since there was no freezing response during the generalization test, nor to spontaneous recovery, since there was no freezing behavior in rats that were submitted
428
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
only to initial conditioning, despite the 3-week delay between extinction and retention testing. The latter observation disagrees with data showing full fear return 2 weeks after fear extinction (Quirk, 2002). One of the possible neurophysiological actions of the HPC HFS is activation of the HPC-cortical input. First, HFS of the dorsal HPC induces long-term potentiation in the medial prefrontal cortex (mPFC; Romcy-Pereira and Pavlides, 2004), probably through an indirect connection from the dorsal HPC to the ventral HPC and from the ventral HPC to the mPFC (Rocher et al., 2004). Second, HFS of the ventral HPC abolishes fear return, but this effect is not observed when the mPFC is inactivated (Deschaux et al., 2011). Third, HFS of the mPFC also abolishes fear return (Deschaux et al., 2011; Zheng et al., 2013; Nachon et al., 2014). More important, the effect of mPFC HFS on conditioned fear depends on initial extinction training (Nachon et al., 2014). Hence, one may consider that activation of the HPC-mPFC input inhibits fear return through interactions with the initially formed extinction circuits in the amygdala, a key structure for fear conditioning and extinction (Herry et al., 2010). Regarding fear learning sensitization, the HPC-cortical-amygdala circuit may also be involved. Indeed, fear learning sensitization is associated with specific molecular and cellular changes in the amygdala (Ponomarev et al., 2010). A pharmacological treatment (isoflurane) blocking amygdala changes was found to suppress fear learning sensitization, independently of fear extinction (Ponomarev et al., 2010). One can suggest that HPC HFS could have inhibited these amygdala changes through the HPC-cortical-amygdala circuit, which, on the contrary, required fear extinction. We previously found a 15-shock exposure decreased synaptic efficacy in the dorsal HPC in a long-term manner (Spennato et al., 2008). In addition, in this condition, HFS of a HPC input failed to induce HPC long-term potentiation (Spennato et al., 2008). These changes in HPC physiology may also explain why HPC HFS failed to suppress fear sensitization in the present study. However, following extinction, it is possible that restoration of normal HPC functioning, could have facilitated the effect of HPC HFS, which further blocked both fear learning sensitization and fear return. This idea is, in part, also supported by our previous study, showing that a pharmacological treatment (fluoxetine) that reversed the electrophysiological effects of the 15-shock exposure, abolished SCPinduced fear return (Spennato et al., 2008). Taken together, these data suggest a pivotal role of the HPC or HPC-mPFC input in controlling both proactive (fear sensitization) and retroactive (fear return) effects of successive fear conditionings. These data also support the idea that fear learning sensitization and fear return in PTSD patients (Dykman et al., 1997; Milad et al., 2009, respectively) may be in part related to HPC dysfunction, and mainly deactivation.
Acknowledgments—This study was supported by F. HoffmannLa Roche (J.L.M.), Universite´ de Nice-Sophia Antipolis (O.D.), Institut de Recherches Biome´dicales des Arme´es (F.C.) and Institut de Neurosciences de la Timone (R.G.).
REFERENCES Baratta MV, Christianson JP, Gomez DM, Zarza CM, Amat J, Masini CV, Watkins LR, Maier SF (2007) Controllable versus uncontrollable stressors bi-directionally modulate conditioned but not innate fear. Neuroscience 146:1495–1503. Bast T, Zhang WN, Feldon J (2001) Hippocampus and classical fear conditioning. Hippocampus 11:828–831. Bouton ME, Bolles RC (1979) Role of conditioned contextual stimuli in reinstatement of extinguished fear. J Exp Psychol Anim Behav Process 5:368–378. Bremner JD, Vythilingam M, Vermetten E, Southwick SM, McGlashan T, Nazeer A, Khan S, Vaccarino LV, Soufer R, Garg PK, Ng CK, Staib LH, Duncan JS, Charney DS (2003) MRI and PET study of deficits in hippocampal structure and function in women with childhood sexual abuse and posttraumatic stress disorder. Am J Psychiatry 160:924–932. Cordero MI, Venero C, Kruyt ND, Sandi C (2003) Prior exposure to a single stress session facilitates subsequent contextual fear conditioning in rats. Evidence for a role of corticosterone. Horm Behav 44:338–345. Deschaux O, Motanis H, Spennato G, Moreau JL, Garcia R (2011) Re-emergence of extinguished auditory-cued conditioned fear following a sub-conditioning procedure: effects of hippocampal and prefrontal tetanic stimulations. Neurobiol Learn Mem 95:510–518. Deschaux O, Zheng X, Lavigne J, Nachon O, Cleren C, Moreau JL, Garcia R (2013) Post-extinction fluoxetine treatment prevents stress-induced reemergence of extinguished fear. Psychopharmacology 225:209–216. Dykman RA, Ackerman PT, Newton JEO (1997) Posttraumatic stress disorder: a sensitization reaction. Integr Physiol Behav Sci 32:9–18. Engdahl B, Dikel TN, Eberly R, Blank Jr A (1998) Comorbidity and course of psychiatric disorders in a community sample of former prisoners of war. Am J Psychiatry 155:1740–1745. Fanselow MS (1994) Neural organization of the defensive behavior system responsible for fear. Psychon Bull Rev 1:429–438. Farinelli M, Deschaux O, Hugues S, Thevenet A, Garcia R (2006) Hippocampal train stimulation modulates recall of fear extinction independently of prefrontal cortex synaptic plasticity and lesions. Learn Mem 13(329):334. Herry C, Ferraguti F, Singewald N, Letzkus JJ, Ehrlich I, Lu¨thi A (2010) Neuronal circuits of fear extinction. Eur J Neurosci 31:599–612. Kamal A, Ramakers GM, Altinbilek B, Kas MJ (2014) Social isolation stress reduces hippocampal long-term potentiation: effect of animal strain and involvement of glucocorticoid receptors. Neuroscience 256:262–270. Long VA, Fanselow MS (2012) Stress-enhanced fear learning in rats is resistant to the effects of immediate massed extinction. Stress 15:627–636. Milad MR, Pitman RK, Ellis CB, Gold AL, Shin LM, Lasko NB, Zeidan MA, Handwerger K, Orr SP, Rauch SL (2009) Neurobiological basis of failure to recall extinction memory in posttraumatic stress disorder. Biol Psychiatry 66:1075–1082. Milad MR, Wright CI, Orr SP, Pitman RK, Quirk GJ, Rauch SL (2007) Recall of fear extinction in humans activates the ventromedial prefrontal cortex and hippocampus in concert. Biol Psychiatry 62:446–454. Nachon O, Cleren C, Husson S, Huguet C, Auclair J, Faure S, Akirav I, Moreau JL, Garcia R (2014) Prefrontal tetanic stimulation, following fear reconditioning, facilitates expression of previously acquired extinction. Neurobiol Learn Mem 113:62–68. Ponomarev I, Rau V, Eger EI, Harris RA, Fanselow MS (2010) Amygdala transcriptome and cellular mechanisms underlying stress-enhanced fear learning in a rat model of posttraumatic stress disorder. Neuropsychopharmacology 35:1402–1411. Quirk GJ (2002) Memory for extinction of conditioned fear is longlasting and persists following spontaneous recovery. Learn Mem 9:402–407.
O. Deschaux et al. / Neuroscience 286 (2015) 423–429 Rau V, DeCola JP, Fanselow MS (2005) Stress-induced enhancement of fear learning: an animal model of posttraumatic stress disorder. Neurosci Biobehav Rev 29:1207–1223. Rau V, Fanselow MS (2009) Exposure to a stressor produces a long lasting enhancement of fear learning in rats. Stress 12:125–133. Rau V, Oh I, Laster M, Eger 2nd EI, Fanselow MS (2009) Isoflurane suppresses stress-enhanced fear learning in a rodent model of post-traumatic stress disorder. Anesthesiology 110:487–495. Rescorla RA, Heth CD (1975) Reinstatement of fear to an extinguished conditioned stimulus. J Exp Psychol Anim Behav Process 1:88–96. Rocher C, Spedding M, Munoz C, Jay TM (2004) Acute stressinduced changes in hippocampal/prefrontal circuits in rats: effects of antidepressants. Cereb Cortex 14:224–229.
429
Romcy-Pereira R, Pavlides C (2004) Distinct modulatory effects of sleep on the maintenance of hippocampal and medial prefrontal cortex LTP. Eur J Neurosci 20:3453–3462. Spennato G, Zerbib C, Mondadori C, Garcia R (2008) Fluoxetine protects hippocampal plasticity during conditioned fear stress and prevents fear learning potentiation. Psychopharmacology 196:583–589. Wall AM, Corcoran AE, O’Halloran KD, O’Connor JJ (2014) Effects of prolyl-hydroxylase inhibition and chronic intermittent hypoxia on synaptic transmission and plasticity in the rat CA1 and dentate gyrus. Neurobiol Dis 62:8–17. Zheng X, Deschaux O, Lavigne J, Nachon O, Cleren C, Moreau JL, Garcia R (2013) Prefrontal high-frequency stimulation prevents sub-conditioning procedure-provoked, but not acute stressprovoked, reemergence of extinguished fear. Neurobiol Learn Mem 101:33–38.
(Accepted 2 December 2014) (Available online 15 December 2014)
HIGH-FREQUENCY STIMULATION OF THE HIPPOCAMPUS BLOCKS FEAR LEARNING SENSITIZATION AND RETURN OF EXTINGUISHED FEAR O. DESCHAUX, a* O.-C. KOUMAR, b F. CANINI, c J.-L. MOREAU d AND R. GARCIA e
occurs following a traumatic event and is characterized by intrusive recollections of the event, avoidance of reminders of the event and hyperarousal. Moreover, PTSD is often co-morbid with phobias (Engdahl et al., 1998). Fear learning sensitization (i.e. potentiation of fear acquisition by initial fear experience) and fear return (i.e. reactivation of extinguished fear) have been suggested to contribute to the development (Dykman et al., 1997; Rau et al., 2005) and persistency (Milad et al., 2009) of phobias in PTSD, respectively. Because the hippocampus (HPC), which is dysfunctional in PTSD (Bremner et al., 2003), is involved in conditioned fear-related phenomena (Bast et al., 2001), fear learning sensitization and fear return may be associated with HPC dysfunction. Also, fear return in PTSD is associated with HPC deactivation (Milad et al., 2009), whereas activation of this area in healthy individuals is associated with less fear return (Milad et al., 2007). It is unknown whether a treatment, known to activate the HPC, can also reduce fear learning sensitization. Because this idea cannot be directly tested in humans, rodent studies can help to elucidate this issue. Fear learning sensitization and fear return have both been widely reported in rats. First, initial fear conditioning (FC1) (context-shock association) can sensitize subsequent fear acquisition. In these studies (e.g., Rau et al., 2005, 2009; Rau and Fanselow, 2009; Ponomarev et al., 2010; Long and Fanselow, 2012), rats are initially exposed to 15 shock administrations in context A; 24 h later, they are placed in context B and exposed to a single shock administration. When re-exposed to context A, animals typically show pronounced freezing response proportional to the 15 shocks, whereas when re-exposed to context B, they do not display freezing proportional to the single shock, but higher levels of freezing, indicating fear learning sensitization. Moreover, in a study, in which we used a similar paradigm, we found that the initial conditioning (with 15 shock administration) induces long-term reduction of activity in the dorsal HPC (Spennato et al., 2008). This HPC deactivation may facilitate fear leaning sensitization. Second, extinguished fear responses can return either spontaneously (spontaneous recovery) or following provocation. For instance, conditioned fear can be fully recovered in rats 10 days after its extinction (Quirk, 2002). Diverse behavioral paradigms have also been developed to provoke fear return. For example, fear returns in rats exposed to unsignaled shock of initial intensity (Rescorla and Heth, 1975; Bouton and Bolles, 1979) or lower intensity (Spennato et al., 2008; Deschaux
a
Institut de Biologie de Valrose, UMR CNRS 7277 – INSERM 1091, Universite´ Nice Sophia Antipolis, 06108 Nice, France b De´partement des Sciences de la Vie, Universite´ Nice Sophia Antipolis, 06108 Nice, France c
Institut de Recherches Biome´dicales des Arme´es, De´partement des Environnements Ope´rationnels, 38702 La Tronche, France d
Pharma Division, Hoffmann-La Roche, Basel, Switzerland
e
Institut de Neurosciences de la Timone, UMR7289, Aix Marseille Universite´ & CNRS, 13385 Marseille, France
Abstract—Patients with post-traumatic stress disorder (PTSD) present hippocampal (HPC) dysfunction, which may facilitate fear-related phenomena such as fear learning sensitization (i.e. potentiation of fear acquisition by initial fear conditioning (FC1)) and fear return (i.e. reactivation of extinguished fear). Fear return is sensitive to HPC high-frequency stimulation (HFS) in rats. The goal of the present study was to examine whether fear learning sensitization is also sensitive to HPC HFS in rats. We found in control conditions that, after FC1 (with 15 shock administrations) and extinction, conditioning in a different context with one shock administration was potentiated (proactive effect) and associated with fear return in the initial context (retroactive effect). Both phenomena were prevented by HPC HFS applied before the second conditioning. We also found that the effect of HPC HFS on fear learning sensitization required initial extinction. These findings suggest a pivotal role of the HPC in preventing proactive and retroactive effects of successive fear conditionings. These data also support the concept that HPC deactivation may be involved in fear learning sensitization and fear return in PTSD patients. Ó 2014 Published by Elsevier Ltd. on behalf of IBRO.
Key words: conditioned freezing, fear extinction, fear return, hippocampus, rats.
INTRODUCTION Acquired fear is at the root of several anxiety disorders, including post-traumatic stress disorder (PTSD), which *Corresponding author. Tel: +33-4-92-07-61-47; fax: +33-4-92-0768-85. E-mail address: [email protected] (O. Deschaux). Abbreviations: FC1, initial fear conditioning; FC2, second fear conditioning; HFS, high-frequency stimulation; HPC, hippocampus; PTSD, post-traumatic stress disorder; SCP, sub-conditioning procedure; VLFS, very low-frequency stimulation. http://dx.doi.org/10.1016/j.neuroscience.2014.12.001 0306-4522/Ó 2014 Published by Elsevier Ltd. on behalf of IBRO. 423
424
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
Table 1. Timing (from Day 1 to Day 27) of the treatments applied in the three contexts (A, B, C) for each group of rats [No-FC1: rats submitted to the second fear conditioning (FC2), but not the initial; FC1: rats submitted to the initial fear conditioning (FC1), but not the second; FC1 + FC2: rats submitted to both sessions of fear conditioning; VLFS: rats submitted to FC1, extinction training, and very low-frequency stimulation (VLFS) of the hippocampus, prior to FC2; HFS: rats submitted to FC1, extinction training (six sessions: two sessions/day 3 days), and high-frequency stimulation (HFS) of the hippocampus, prior to FC2; No-Ext: rats submitted to FC1 and HFS of the hippocampus prior to FC2, but not to extinction training] Group
Day 1: Days 2–4: FC1 (0.7 mA) Extinction (2 sessions/d)
Day 25a: Day 25b: Day 26: Brain stimulation (BS) FC2 (0.7 mA) Sensitization test
Day 27a: Generalization test
Day 27b: Extinction reversal test
No-FC1
Context A (no shock)
Context B (No BS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
FC1
Context A (15 shocks)
Context B (No BS)
Context C (0 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
Context B (No BS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
Context B (BS: HPC VLFS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
Context B (BS: HPC HFS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
Context B (BS: HPC HFS)
Context C (1 shock)
Context C (30 min)
Context B (30 min)
Context A (30 min)
FC1 + FC2 Context A (15 shocks) VLFS
Context A (15 shocks)
HFS
Context A (15 shocks)
No-Ext
Context A (15 shocks)
Context A (30 min/ session) Context A (30 min/ session) Context A (30 min/ session) Context A (30 min/ session) Context A (30 min/ session)
et al., 2011, 2013; Zheng et al., 2013). Interestingly, this fear return is prevented by high-frequency stimulation (HFS) of the HPC (Deschaux et al., 2011) or a pharmacological treatment fluoxetine (Deschaux et al., 2013), which has been shown to suppress HPC deactivation (Spennato et al., 2008). From these observations, we hypothesized that if fear learning sensitization and fear return are both facilitated by HPC deactivation, HPC HFS, which is known to induce intra-HPC LTP in behaving rodents (Kamal et al., 2014; Wall et al., 2014), would impair fear learning sensitization as it blocks fear return. We used a behavioral paradigm provoking fear learning sensitization and fear return in which rats were submitted to initial conditioning with 15 shocks in context A, followed (or not) by an extinction training session. Rats received HPC HFS (or not) in context B, before being exposed to a single shock in context C. Fear learning sensitization and fear return were evaluated in contexts C and A, respectively. We also tested possible fear generalization in context B, wherein animals had no shock experience.
EXPERIMENTAL PROCEDURES Animals Forty-two adult male Wistar rats (De´pre´, Saint Doulchard, France), weighing between 280 and 350 g, were used. They were housed in individual home cages in a wellventilated and temperature-regulated (22 ± 2 °C) room with food and water ad libitum. A 12–12-h light/dark cycle (lights on from 7 a.m.) was imposed. Experiments were performed in accordance with the European Community Guidelines on the Care and Use of Laboratory Animals (86/609/EEC).
Electrode implantation For brain stimulation, rats were implanted prior to contextual fear conditioning. For this purpose, they were anesthetized with pentobarbital (65 mg/kg, i.p.) and were fixed in a stereotaxic frame. Stimulating electrodes made of twisted silver wires (90-lm diameter) were bilaterally lowered into the CA1 region of the dorsal HPC (3.2 mm posterior to bregma, 1.6 mm lateral to midline, 2.5 mm from dura). This area was chosen based on our previous study showing that its stimulation modulates (facilitates or inhibits) fear return as a function of the frequency used (Farinelli et al., 2006). The entire miniature system was fixed onto the skull by three screws and dental cement. Rats were then left undisturbed in their home cage for 5–7 days to recover before the experiments were started.
Apparatus All behavioral devices were purchased from Imetronic (Pessac, France). Three contexts (A, B and C) were used. Context A consisted of a gray plastic cuboid cage (each side, 40-cm length and width and 45-cm height), with a grid floor composed of 0.5-cm-diameter stainless steel rods placed 1.5 cm apart. Context B consisted of a clear plastic cylindrical cage (25-cm diameter, 25 cm height) with a plastic floor covered with animal bedding. Context C consisted of a clear plastic cage with a hexagonal prism shape (40 cm 30 cm 30 cm 20 cm 12 cm 12 cm breadth, 45-cm height), with a grid floor as in context A, and an additional background noise that was continuously provided (60 dB). White ceiling lights were on for context A, and a red light was on (with white ceiling lights off) in the room for contexts
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
425
Fig. 1. Histological representation of a coronal section showing placements of stimulating electrodes in the dorsal hippocampus.
B and C. During each behavioral test, each cage was placed inside a clear plastic chamber that was soundattenuating and temperature-regulated. Behavioral and brain stimulation procedures For behavioral tests, rats were submitted to a 27-day experimental procedure (Table 1). On Day 1, nonimplanted rats were randomly divided into three groups of seven rats each, comprising rats subjected to the second fear conditioning (FC2), but not to the first (No-FC1 group), rats subjected only to the FC1 (FC1 group), and rats subjected to both sessions of conditioning (FC1 + FC2 group). On the same day (Day 1), implanted rats were also randomly divided into three groups (seven rats each): all groups were subjected to FC1 and FC2, but two groups received either HPC very low-frequency stimulation (VLFS group) or HPC HFS (HPC group) after extinction training. The third group did not undergo
extinction training but received HPC HFS before FC2 (No-Ext group). All rats (non-implanted and implanted) were placed in context A (the cuboid cage, which was cleaned with a 50% ethanol and lemon scent solution before each rat was introduced). Three minutes later, all rats, except those in the No-FC1 group, were subjected to 15 electric foot-shocks (0.7 mA, 1 s; intertrial interval: 240–420 s). Sixty seconds after the last foot-shock application, they were removed from context A. Rats in the No-FS1 group were left undisturbed in context A for the same time without any shock presentation. On Day 2, all rats, except those in the No-Ext group, were re-exposed to context A twice (with 3–5 h interval) for 30 min each time (exposures A1 and A2). This schedule of context re-exposure was repeated on Days 3 and 4 (exposures A3 and A4 and exposures A5 and A6, respectively) to induce maximized extinction of conditioned freezing to context A.
Fig. 2. Percentage of freezing behavior (means ± SE) during the first 10 min of the six sessions (A1 to A6) of extinction training (after FC1) for NoFC1, FC1, FC1 + FC2, VLFS, and HFS groups. ⁄p < 0.05: statistically significant difference from the No-FC1 group.
426
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
Fig. 3. Percentage of freezing behavior (mean ± SE) during the first 10 min of tests in context C (test for fear generalization), in context B (test for fear generalization), and in context A (test for fear return) for all groups of rats (No-FC1, FC1, FC1 + FC2, VLFS, HFS, and No-Ext). ⁄p < 0.05 and ⁄⁄⁄ p < 0.0001: statistically significant difference from the No-FC1 group; #p < 0.05 and ###p < 0.0001: statistically significant difference from the FC1 group; ¤p < 0.05 and ¤¤¤p < 0.0001: statistically significant difference from the HFS group; £p < 0.05: statistically significant difference from the No-Ext group.
On Day 25 (Day 25a), all rats were placed in a new context (context B, the cylindrical cage that was cleaned with 70% ethanol and had animal bedding on its floor that was changed before each rat was introduced). During the 60 min of exposure to context B, implanted rats were connected to stimulating cables, which were relayed at the top of the experimental box by a multichannel rotating connector allowing free movement. After the first 30 min of acclimation to context B and to the connected cables, VLFS rats received VLFS (100 pulses of 500 lA delivered at 0.2 Hz), whereas rats in the HFS and No-Ext groups received HFS (100 pulses of 500 lA delivered at 100 Hz). These parameters of stimulation were chosen according to our previous study, showing that VLFS does not affect HPC-mPFC synaptic efficacy, whereas HFS enhances this synaptic efficacy in a long-term manner (Farinelli et al., 2006). When animals were removed from context B, they were placed in their home cages in the animal room where they were left undisturbed for 6 h. After this delay (Day 25b), they were placed in a new context (context C, the prism-shaped cage, which was cleaned with a 50%-ethanol and lemon scent solution before each animal was introduced). Three minutes later, all groups of rats, except the FC1 group, received a single electric foot-shock (0.7 mA, 1 s). Sixty seconds later, they were removed from context C. On Day 26, rats were tested for fear conditioning in context C for 30 min. On Day 27 (Day 27a), they were tested for 30 min for fear generalization in context B. Six hours later (Day 27b), they
were re-exposed to context A for 30 min to test fear return. Behavioral scoring, statistical analyses and histology The behavior of each animal was continuously recorded. Conditioned fear was assessed by measuring freezing behavior (Fanselow, 1994), which was scored using a time-sampling procedure. The time spent freezing was measured during the first 10 min of each context re-exposure (contexts A, B and C) because all rats were fully awake during this period. Data were analyzed by two-way ANOVAs for repeated measurements, followed by Fisher’s PLSD post hoc tests, as indicated in the text. Data are presented as mean ± SEM. Values were considered significant if p < 0.05. Upon completion of the experiments, histological analysis was performed as detailed elsewhere (Farinelli et al., 2006).
RESULTS Histological controls revealed that the tips of electrodes were located in the CA1 region of the dorsal HPC in all implanted rats (Fig. 1). Extinction training Fig. 2 (extinction training in context A) shows that during the first 10 min of the first session of context re-
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
exposure (session A1), levels of freezing were high in rats submitted to the 15-shock conditioning (FC1, FC1 + FC2, VLFS and HFS groups), but there was no freezing in rats that did not receive shock (No-FC1 group). Extinction training (sessions A1-A6) of the conditioned rats resulted in a progressive suppression of freezing behavior. A two-way ANOVA performed on these data (five groups, six sessions) revealed main effects of group (F4,30 = 7.13; p < 0.001) and session (F5,150 = 74.41; p < 0.0001). A significant Group x Session interaction was also found (F15,150 = 5.46; p < 0.0001). Post hoc Fisher’s PLSD tests indicated that each conditioned group differed from the No-FC1 group during the first extinction session (A1: all ps < 0.001). Another two-way ANOVA on data from the four conditioned groups (FC1, FC1 + FC2, VLFS, and HFS, six sessions) showed that there was no effect of group or interaction between group and session, indicating that the four groups similarly extinguished their conditioned fear. Fear sensitization, generalization and return tests Fig. 3 shows that, during re-exposure to context C, only rats that were submitted to FC2 (No-FC1, FC1 + FC2, VLFS, HFS, and No-Ext) displayed conditioned freezing, which was potentiated, except in rats that underwent extinction training and received HPC HFS (HFS group). Rats that extinguished the initial fear and were not submitted to FC2 continued displaying low levels of freezing during their re-exposure to context C (FC1). During re-exposure to context B, none of the six groups showed significant conditioned freezing. Finally, during re-exposure to context A, all groups that were conditioned twice, except those submitted to extinction training and stimulated with HFS (HFS group), presented high levels of freezing behavior. Rats that were conditioned once, either during the FC1 only (FC1 group) or the FC2 only (No-FC1 group), did not exhibit freezing in context A. A two-way ANOVA performed on these data (six groups three contexts: C, B and A) revealed main effects of group (F5,36 = 24.63, p < 0.0001) and context (F2,72 = 60.57, p < 0.0001). Group x Context interaction was also significant (F10,72 = 10.18, p < 0.0001). Post hoc Fisher’s PLSD tests showed that, during the sensitization test, rats that were not submitted to FC1 (No-FC1 group) and animals that underwent extinction training and received HPC HFS (HFS group) did not differ from each other, but each differed from each of the other groups (FC1: both ps < 0.05; FC1 + FC2, VLFS and No-Ext: all ps < 0.05). In addition, during the same test, the FC1 group differed from the FC1 + FC2, VLFS, and NoExt groups (all ps < 0.0001), but the latter three groups did not differ from each other. During the generalization test, no group differed from the others. Finally, during the fear return test, the three groups with very low levels of freezing (No-FC1, FC1 and HFS) did not differ from each other, but each differed from each of the 3 other groups (FC1 + FC2, VLFS
427
and No-Ext: all ps < 0.0001). However, rats that did not undergo fear extinction (No-Ext) differed from both the FC1 + FC2 and VLFS groups (both ps < 0.05).
DISCUSSION In the present study, initial conditioning (with 15 shocks in context A) induced sensitization of subsequent conditioning (with a single shock in context C), independently of extinction training (in context A). We demonstrated for the first time that HPC HFS, applied several hours before the second conditioning, prevented this fear learning sensitization. We also observed that the above paradigm produced fear return when rats were re-exposed to context A. This context-specific fear return (no fear generalization was found) was also prevented by HPC HFS. Prior stress can potentiate subsequent contextual and auditory fear conditioning (e.g., Baratta et al., 2007) or contextual fear conditioning, but not auditory (e.g., Cordero et al., 2003). In these studies, the stressful situation was different from the fear conditioning paradigm itself (uncontrollable tail-shock or restrain stress versus foot-shock). However, we based the present study on the enhancement of contextual fear learning caused by prior exposure to several shocks. As reported by other authors (e.g., Rau et al., 2005, 2009; Rau and Fanselow, 2009; Ponomarev et al., 2010; Long and Fanselow, 2012), the advantage of such a procedure is that it works for both contextual and auditory fear conditioning. In addition, although the mechanisms underlying fear learning potentiation remain unknown, it resists to fear extinction and to amnestic (N-Methyl-D-Aspartate antagonist) treatment (Rau et al., 2005). However, here we observed that this phenomenon did not resist to HPC HFS. Interestingly, HPC HFS specifically impaired fear sensitization, without affecting fear learning per se. In context C, rats that received HFS displayed levels of freezing behavior similar to those of the ones submitted only to a single shock. In other words, rats that received HPC HFS acquired the second conditioning, without any sensitization (no proactive effect). Surprisingly, fear sensitization was still present in rats that received HFS, but were not trained for extinction, revealing the importance of extinction to obtain the effect of HPC stimulation. We had shown that after conditioning with several shocks and fear extinction, rats can express fear return when retrained with a shock intensity that is too weak to induce, by itself, significant fear conditioning (Spennato et al., 2008). This protocol, which we named sub-conditioning procedure (SCP), led to similar findings when we used an auditory-cue fear conditioning task (Deschaux et al., 2011, 2013; Zheng et al., 2013). We also had found that HPC HFS, applied in the case of auditory fear conditioning, suppressed SCP-induced fear return (Deschaux et al., 2011). The present study, using a different paradigm, extends this observation. The fear response during the fear return test corresponded neither to fear generalization, since there was no freezing response during the generalization test, nor to spontaneous recovery, since there was no freezing behavior in rats that were submitted
428
O. Deschaux et al. / Neuroscience 286 (2015) 423–429
only to initial conditioning, despite the 3-week delay between extinction and retention testing. The latter observation disagrees with data showing full fear return 2 weeks after fear extinction (Quirk, 2002). One of the possible neurophysiological actions of the HPC HFS is activation of the HPC-cortical input. First, HFS of the dorsal HPC induces long-term potentiation in the medial prefrontal cortex (mPFC; Romcy-Pereira and Pavlides, 2004), probably through an indirect connection from the dorsal HPC to the ventral HPC and from the ventral HPC to the mPFC (Rocher et al., 2004). Second, HFS of the ventral HPC abolishes fear return, but this effect is not observed when the mPFC is inactivated (Deschaux et al., 2011). Third, HFS of the mPFC also abolishes fear return (Deschaux et al., 2011; Zheng et al., 2013; Nachon et al., 2014). More important, the effect of mPFC HFS on conditioned fear depends on initial extinction training (Nachon et al., 2014). Hence, one may consider that activation of the HPC-mPFC input inhibits fear return through interactions with the initially formed extinction circuits in the amygdala, a key structure for fear conditioning and extinction (Herry et al., 2010). Regarding fear learning sensitization, the HPC-cortical-amygdala circuit may also be involved. Indeed, fear learning sensitization is associated with specific molecular and cellular changes in the amygdala (Ponomarev et al., 2010). A pharmacological treatment (isoflurane) blocking amygdala changes was found to suppress fear learning sensitization, independently of fear extinction (Ponomarev et al., 2010). One can suggest that HPC HFS could have inhibited these amygdala changes through the HPC-cortical-amygdala circuit, which, on the contrary, required fear extinction. We previously found a 15-shock exposure decreased synaptic efficacy in the dorsal HPC in a long-term manner (Spennato et al., 2008). In addition, in this condition, HFS of a HPC input failed to induce HPC long-term potentiation (Spennato et al., 2008). These changes in HPC physiology may also explain why HPC HFS failed to suppress fear sensitization in the present study. However, following extinction, it is possible that restoration of normal HPC functioning, could have facilitated the effect of HPC HFS, which further blocked both fear learning sensitization and fear return. This idea is, in part, also supported by our previous study, showing that a pharmacological treatment (fluoxetine) that reversed the electrophysiological effects of the 15-shock exposure, abolished SCPinduced fear return (Spennato et al., 2008). Taken together, these data suggest a pivotal role of the HPC or HPC-mPFC input in controlling both proactive (fear sensitization) and retroactive (fear return) effects of successive fear conditionings. These data also support the idea that fear learning sensitization and fear return in PTSD patients (Dykman et al., 1997; Milad et al., 2009, respectively) may be in part related to HPC dysfunction, and mainly deactivation.
Acknowledgments—This study was supported by F. HoffmannLa Roche (J.L.M.), Universite´ de Nice-Sophia Antipolis (O.D.), Institut de Recherches Biome´dicales des Arme´es (F.C.) and Institut de Neurosciences de la Timone (R.G.).
REFERENCES Baratta MV, Christianson JP, Gomez DM, Zarza CM, Amat J, Masini CV, Watkins LR, Maier SF (2007) Controllable versus uncontrollable stressors bi-directionally modulate conditioned but not innate fear. Neuroscience 146:1495–1503. Bast T, Zhang WN, Feldon J (2001) Hippocampus and classical fear conditioning. Hippocampus 11:828–831. Bouton ME, Bolles RC (1979) Role of conditioned contextual stimuli in reinstatement of extinguished fear. J Exp Psychol Anim Behav Process 5:368–378. Bremner JD, Vythilingam M, Vermetten E, Southwick SM, McGlashan T, Nazeer A, Khan S, Vaccarino LV, Soufer R, Garg PK, Ng CK, Staib LH, Duncan JS, Charney DS (2003) MRI and PET study of deficits in hippocampal structure and function in women with childhood sexual abuse and posttraumatic stress disorder. Am J Psychiatry 160:924–932. Cordero MI, Venero C, Kruyt ND, Sandi C (2003) Prior exposure to a single stress session facilitates subsequent contextual fear conditioning in rats. Evidence for a role of corticosterone. Horm Behav 44:338–345. Deschaux O, Motanis H, Spennato G, Moreau JL, Garcia R (2011) Re-emergence of extinguished auditory-cued conditioned fear following a sub-conditioning procedure: effects of hippocampal and prefrontal tetanic stimulations. Neurobiol Learn Mem 95:510–518. Deschaux O, Zheng X, Lavigne J, Nachon O, Cleren C, Moreau JL, Garcia R (2013) Post-extinction fluoxetine treatment prevents stress-induced reemergence of extinguished fear. Psychopharmacology 225:209–216. Dykman RA, Ackerman PT, Newton JEO (1997) Posttraumatic stress disorder: a sensitization reaction. Integr Physiol Behav Sci 32:9–18. Engdahl B, Dikel TN, Eberly R, Blank Jr A (1998) Comorbidity and course of psychiatric disorders in a community sample of former prisoners of war. Am J Psychiatry 155:1740–1745. Fanselow MS (1994) Neural organization of the defensive behavior system responsible for fear. Psychon Bull Rev 1:429–438. Farinelli M, Deschaux O, Hugues S, Thevenet A, Garcia R (2006) Hippocampal train stimulation modulates recall of fear extinction independently of prefrontal cortex synaptic plasticity and lesions. Learn Mem 13(329):334. Herry C, Ferraguti F, Singewald N, Letzkus JJ, Ehrlich I, Lu¨thi A (2010) Neuronal circuits of fear extinction. Eur J Neurosci 31:599–612. Kamal A, Ramakers GM, Altinbilek B, Kas MJ (2014) Social isolation stress reduces hippocampal long-term potentiation: effect of animal strain and involvement of glucocorticoid receptors. Neuroscience 256:262–270. Long VA, Fanselow MS (2012) Stress-enhanced fear learning in rats is resistant to the effects of immediate massed extinction. Stress 15:627–636. Milad MR, Pitman RK, Ellis CB, Gold AL, Shin LM, Lasko NB, Zeidan MA, Handwerger K, Orr SP, Rauch SL (2009) Neurobiological basis of failure to recall extinction memory in posttraumatic stress disorder. Biol Psychiatry 66:1075–1082. Milad MR, Wright CI, Orr SP, Pitman RK, Quirk GJ, Rauch SL (2007) Recall of fear extinction in humans activates the ventromedial prefrontal cortex and hippocampus in concert. Biol Psychiatry 62:446–454. Nachon O, Cleren C, Husson S, Huguet C, Auclair J, Faure S, Akirav I, Moreau JL, Garcia R (2014) Prefrontal tetanic stimulation, following fear reconditioning, facilitates expression of previously acquired extinction. Neurobiol Learn Mem 113:62–68. Ponomarev I, Rau V, Eger EI, Harris RA, Fanselow MS (2010) Amygdala transcriptome and cellular mechanisms underlying stress-enhanced fear learning in a rat model of posttraumatic stress disorder. Neuropsychopharmacology 35:1402–1411. Quirk GJ (2002) Memory for extinction of conditioned fear is longlasting and persists following spontaneous recovery. Learn Mem 9:402–407.
O. Deschaux et al. / Neuroscience 286 (2015) 423–429 Rau V, DeCola JP, Fanselow MS (2005) Stress-induced enhancement of fear learning: an animal model of posttraumatic stress disorder. Neurosci Biobehav Rev 29:1207–1223. Rau V, Fanselow MS (2009) Exposure to a stressor produces a long lasting enhancement of fear learning in rats. Stress 12:125–133. Rau V, Oh I, Laster M, Eger 2nd EI, Fanselow MS (2009) Isoflurane suppresses stress-enhanced fear learning in a rodent model of post-traumatic stress disorder. Anesthesiology 110:487–495. Rescorla RA, Heth CD (1975) Reinstatement of fear to an extinguished conditioned stimulus. J Exp Psychol Anim Behav Process 1:88–96. Rocher C, Spedding M, Munoz C, Jay TM (2004) Acute stressinduced changes in hippocampal/prefrontal circuits in rats: effects of antidepressants. Cereb Cortex 14:224–229.
429
Romcy-Pereira R, Pavlides C (2004) Distinct modulatory effects of sleep on the maintenance of hippocampal and medial prefrontal cortex LTP. Eur J Neurosci 20:3453–3462. Spennato G, Zerbib C, Mondadori C, Garcia R (2008) Fluoxetine protects hippocampal plasticity during conditioned fear stress and prevents fear learning potentiation. Psychopharmacology 196:583–589. Wall AM, Corcoran AE, O’Halloran KD, O’Connor JJ (2014) Effects of prolyl-hydroxylase inhibition and chronic intermittent hypoxia on synaptic transmission and plasticity in the rat CA1 and dentate gyrus. Neurobiol Dis 62:8–17. Zheng X, Deschaux O, Lavigne J, Nachon O, Cleren C, Moreau JL, Garcia R (2013) Prefrontal high-frequency stimulation prevents sub-conditioning procedure-provoked, but not acute stressprovoked, reemergence of extinguished fear. Neurobiol Learn Mem 101:33–38.
(Accepted 2 December 2014) (Available online 15 December 2014)