Effect of D-cycloserine in conjunction with fear extinction training on extracellular signal-regulated kinase activation in medial prefrontal cortex and amygdala in rat

Abstract

D-cycloserine (DCS) is currently under clinical trials for a number of neuropsychiatric conditions and has been found to augment fear extinction in rodents and exposure therapy in humans. However, the molecular mechanism of DCS action in these multiple modalities remains unclear. Here, we describe the effect of DCS administration, alone or in conjunction with extinction training, on neuronal activity (c-fos) and neuronal plasticity (phospho-extracellular signal-regulated kinase, pERK) markers using immunohistochemistry. We found that intraperitoneal administration of DCS in untrained young rats (24–28 days old) increased c-fos and pERK-stained neurons in both the prelimbic (PL) and infralimbic (IL) division of the medial prefrontal cortex (mPFC) and reduced pERK levels in the lateral nucleus (CeL) of the central amygdala (CeA). Moreover, DCS administration significantly increased GluA1, GluN1, GluN2A, and GluN2B expression in mPFC. In a separate set of animals, we found that DCS facilitated fear extinction and increased pERK levels in IL, PL, intercalated cells and CeL, compared to saline control. In synaptoneurosomal preparation, we found that extinction training increased iGluR protein expression in the mPFC, compared to context animals. No significant difference in protein expression was observed between extinction-saline and extinction-DCS groups in the mPFC. In contrast, in the amygdala DCS in conjunction with extinction training led to an increase in iGluR subunit expression, compared to extinction-saline group. Our data suggest that the efficacy of DCS in neuropsychiatric disorders may be partly due to its ability to affect neuronal activity and signaling in the mPFC and amygdala subnuclei.

Introduction

There has been considerable interest in the neural mechanisms of extinction learning, given the use of extinction-based therapies for treatment of anxiety disorders (Davis, 2002; Ressler et al., 2004). A role of glutamate receptors in fear processes is evident from studies showing that pharmacological blockade of NMDA receptors in the basolateral amygdala (BLA) blocks both acquisition and extinction of fear (Miserendino et al., 1990; Cox & Westbrook, 1994; Baker & Azorlosa, 1996). D-cycloserine (DCS), an agonist at the glycine binding-site of the NMDA receptor, has been shown to augment the extinction of conditioned fear (Walker et al., 2002; Ledgerwood et al., 2003). In humans, DCS is effective in posttraumatic stress disorder, specific phobias, social anxiety, obsessive compulsive disorders, as well as in extinction of drugs of abuse (reviewed in Hofmann et al., 2006; Ganasen et al., 2010; Myer and Carlezon, 2012). DCS has also been shown to influence paradigms that do not involve behavioral training (Papp and Moryl, 1996; Jacome et al., 2011). Thus, DCS alone and in conjunction with behavioral therapy may have specific effects that remain to be fully tested.

Studies in animals (Paré et al., 2004; Maren and Quirk, 2004) and humans (Phelps and LeDoux, 2005) indicate that interactions between the medial prefrontal cortex (mPFC) and the amygdala are critically involved in extinction learning (LeDoux, 2000; Morgan et al., 1993; Quirk et al., 2003; Rosenkranz et al., 2003). Specifically, recent studies have established a role for the infralimbic (IL)-PFC in consolidation of fear extinction (Morgan et al., 2003; Sierra-Mercado et al., 2006; Milard and Quirk, 2002). Previous studies also implicate MAPK/ERK signaling pathways in consolidation of auditory and contextual fear conditioning (Schafe et al., 2000; Trifilieff et al., 2006), as well as extinction of conditioned fear in the mPFC, BLA, and hippocampus (Hugues et al., 2004, 2006; Herry et al., 2006; Fisher et al., 2007; Tronson et al., 2009). C-fos is a neuronal activity marker, and increased c-fos staining, suggestive of higher neuronal activity, has been observed in PL and IL after fear extinction (Herry and Mons, 2004; Knapska and Maren, 2009; Park and Choi, 2009; Kim et al., 2010). Evidence also indicates that amygdala GABAergic intercalated cells (ITC) are prime candidates for mediating mPFC influences during extinction (Pare et al., 2004). Additionally, local infusion of DCS in the lateral amygdala has been shown to facilitate fear extinction, suggesting DCS enhancement of fear extinction is mediated via amygdala action (Walker et al., 2002; Ledgerwood et al., 2003). DCS has also been shown to facilitate re-extinction via local infusion in the IL (Chang and Maren, 2011), and has been proposed to erase conditioned fear (Richardson et al., 2004; Mao et al., 2006; Lin et a l., 2010). While many studies point to DCS influencing extinction circuitry, a thorough investigation into various sites of DCS action is lacking. To elucidate the mechanism by which DCS facilitates fear extinction, or DCS alone may modulate signaling, we tested the effect of DCS on pERK and iGluRs in the amygdala and mPFC.

Materials and methods

Subjects

Male Sprague-Dawley rats, 24–28 days old, (Harlan, Denver, CO, USA) were used in the study. Animals were group housed on a 12:12 light-dark cycle with ad libitum access to food and water. Prior to behavioral procedures, animals were handled at the approximate time of day in which the procedures were to be carried out. All procedures took place in the light phase of the light-dark cycle. All procedures were approved by the Creighton University Institutional Animal Care and Use Committee and conformed to the NIH Guide for the Care and Use of Laboratory Animals.

Time-dependent effect of DCS administration on p-ERK, c-fos, and iGluRs expression in the mPFC and amygdala

D-cycloserine (C6880, Sigma-Aldrich, St. Louis, MO, USA) was freshly dissolved in sterile isotonic saline and injected intraperitoneally (IP) at a volume of 1 ml/kg. Control rats were injected with saline at a volume of 1 ml/kg. Drug doses were based on previous studies in rats (Walker et al., 2002; Ledgerwood et al., 2003). For studying the effect of DCS on pERK and c-fos expression in the mPFC and amygdala, in time dependent manner, animals were divided into two groups: saline (n=5) and DCS treated (n=5) (30 min, 1h, 6h, and 24h). After time-dependent administration of DCS or saline, rats were anesthetized by isoflurane (14043-225-06, Webster Veterinary, Devens, MA, USA) and then intracardially perfused with 0.9% saline followed by 4% paraformaldehyde (PFA) in 0.1M phosphate buffer (PB). Brains were removed, postfixed in 4% PFA overnight, cryoprotected in 10%, 20%, and 30% sucrose solution, followed by freezing in n-pentane on dry ice and stored at -80°C. For studying the effect of DCS on iGluR expression, animals were divided into two groups; saline (n=4) and DCS (n=4). Saline or DCS (30 mg/kg) was administered (IP) and 6h after treatment, animals were anesthetized and their mPFC and amygdala collected in synaptoneurosome buffer for synaptoneurosome preparation.

DCS effect on fear expression and extinction recall

Fear Conditioning/Extinction Training/Extinction Recall Testing

Prior to behavioral procedures, animals were handled for three days at the approximate time of the day the fear conditioning/extinction procedures were to occur. All behavioral procedures took place during the light phase of the light/dark cycle. For fear conditioning, rats were placed in a Plexiglas rodent chamber (chamber A; model 2325-0241 San Diego Instruments, San Diego, CA) with a metal grid floor that was enclosed in a sound-attenuating chamber illuminated with white light. Chamber A was cleaned with a 19.5% ethanol/1% vanilla solution to give the chamber a distinct scent. For extinction training and CS testing, rats were placed in a novel Plexiglas chamber (chamber B, model 2325-0241 San Diego Instruments, San Diego, CA) with different visual cues and a solid Plexiglas floor to minimize generalization to the conditioning chamber. Chamber B was cleaned with a 70% ethanol solution, scented with linen-scented air freshener, and illuminated with white light. White noise was provided in each isolation chamber with a fan. A web camera (Logitech QuickCam) was mounted at the top of each chamber to record all sessions. Prior to conditioning (day 0), animals were acclimated to chamber A for 30 min. On the day of fear conditioning (day 1), animals were placed in chamber A for 3 min followed by five presentations of a conditioned stimulus (CS). The CS was an 85 dB, 3 kHz tone delivered for 10 sec with a 1 min inter-trial interval (ITI). Depending on the training group, CS was presented alone (context), co-terminated with a noxious unconditioned stimulus (US) (conditioned), or presented with random presentations of the US (unpaired). The US was a 0.8 mA scrambled foot-shock delivered for 2 sec. Rats were removed from chamber A 1 min after the final CS-US pairing (conditioned), CS presentation alone (context), or US presentation (unpaired).

Prior to activity on day 2, conditioned animals were split into four groups; those that were injected with DCS or saline 24 h after conditioning (cond-DCS or cond-sal, respectively) or those that were to undergo extinction training followed by injection with DCS or saline (ext-DCS and ext-sal, respectively). On day 2, animals that were to undergo extinction training were placed into chamber B and, after a 2 min acclimation, they were given 6 presentations of the CS for 2 min with a 1 min ITI. Immediately following extinction training, animals were injected with either DCS (30 mg/kg, ext-DCS) or saline (ext-sal) and placed back into the home cage. Conditioned animals that did not undergo extinction training but were injected with DCS or saline (cond-DCS or cond-sal, respectively) were injected at the same time point as the respective extinction animals. Cond-DCS and cond-sal animals were injected with either DCS (30 mg/kg, cond-DCS) or saline (cond-sal) and placed back into their home cages. Context animals were placed into chamber B on day 2 without any presentation of the CS.

Twenty-four hours later (day 3), context, unpaired, cond-DCS, cond-sal, ext-DCS, and ext-sal animals were placed into chamber B to analyze the expression of fear or fear extinction recall in all groups. Animals were placed into chamber B and, after a 2 min acclimation, presented with 1 CS presentation for 2 min and removed from the chamber 2 min after the CS presentation. Behavioral freezing was measured at all stages of behavioral testing during conditioning (day 1), extinction training (day 2), and testing (day 3) visually as the absence of all non-respiratory movement every 5 sec. Scores of 0 for immobility and 1 for mobility were averaged and divided by the total number of readings during the CS presentation to derive a percent freezing. Behavioral freezing was also monitored with the Freeze Monitor System (San Diego Instruments, San Diego, CA) software to verify visual scores.

DCS effect on ERK and iGluR after fear conditioning or extinction

Separate animals were used to observe the behavioral, immunohistochemical, and synaptoneurosomal iGluR changes due to DCS administration. Behavioral analysis occurred as described above. To determine ERK activation and iGluR expression, animals were not subjected to the behavioral procedures on day 3. Rather, 6 hours after injection with DCS or saline on day 2, animals were sacrificed and their brains were isolated for immunohistochemical analysis (n=4) or synaptoneurosomal preparation (n=4) for iGluR expression. This time point was determined based on results from the previous section looking at the time-dependent effect of DCS on pERK expression and influence on iGluR expression. Six hours after injection, animals were anesthetized with isoflurane, decapitated, and brain samples were collected and processed for immunohistochemical or synaptoneurosomal analysis as described below.

Immunohistochemistry

Diaminobenzidine (DAB)-Immunostaining for ERK/pERK and c-fos in rat brain was performed using a previously described DAB staining procedure (Xu and Pandey, 2000). Coronal brain sections (20μm) were obtained using a cryostat (Leica CM 1900, Heidelberger, Germany). Sections were washed with 0.01 M phosphate buffer saline (PBS) (2 ×10 min), treated with 0.3% hydrogen peroxide (H324-500, Fisher Scientific) diluted in PBS containing 0.25% Triton X-100 (PBST) for 30 min followed by washing with 0.01M PBS (6 × 5 min), and incubated with 10% normal goat serum (NGS) (005-000-121, Jackson Immuno Research Laboratories, West Grove, PA, USA) for 1h at room temperature. Sections were incubated overnight at room temperature in p44/42 MAPK (ERK1/2) (rabbit polyclonal, 9102, 1:2000 dilution), phospho-p44/42 MAPK (pERK1/2) (Thr202/Tyr204) (rabbit monoclonal, 4370S, 1:2000 dilution) (Cell Signaling Technology, Danvers, MA, USA), or c-fos (rabbit polyclonal, sc-52, 1:5000 dilution) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) diluted in PBST (containing 0.01% sodium azide) overnight at room temperature. Sections were washed with 0.01M PBS (3 × 15min) and incubated with 1:2000 diluted secondary antibody [(goat anti-rabbit biotinylated BA01000, 1:2000 (Vector Laboratories, Burlingame, CA))] in PBST for 2h at room temperature. Afterward, sections were washed with PBS (3 × 10 min) and incubated with avidin-biotinylated-peroxidase complex (PK-7100, Vectastain Elite ABC Kit; Vector Laboratories). Following 3 × 15 min washes with PBS, sections were incubated with ImmPACT^™ DAB (SK-4105, Vector Laboratories) according to the instructions provided with the kit. After washing with PBS (3 × 15 min) and then water, sections were mounted on slides and air-dried. The dried sections were dehydrated with 70%, 90%, and 100% alcohol and xylene (5 min in each), and cover slips were mounted with Permount® (SP15-100, Fisher Scientific, New Jersey, USA). For negative brain sections, an identical protocol to the one mentioned above was used, except that 3% NGS in PBST containing 0.01% sodium azide was substituted for the primary antibody. Sections were visualized using Zeiss Axioscope Z 1051-078 microscope (Zeiss, Chester, VA, USA) using spot basic image acquisition software (SPOT^™ Imaging Solutions, Sterling Heights, MI, USA). The top border of PL was identified by aligning the grid with the top of the corpus callosum and the PL/IL border was identified using the Rat Brain Atlas (Paxinos and Watson, 2005). Moreover, in order to avoid improper discrimination between PL and IL we chose not to perform cell counting very close to the border of PL/IL. Images of five random regions of each section of mPFC (PL, IL) and amygdala (CeL, CeM, BA, and ITC) were taken at 63X and pERK- or c-fos-positive neurons were counted manually and averaged. Slides were coded randomly to keep the individual counting the neurons blind to the experimental group. Five such sections were analyzed from each animal.

Synaptoneurosome preparation

Synaptoneurosomes were prepared from amygdala and mPFC as described in Villasana et al. (2006) with modifications. After isoflurane anesthesia, rats were decapitated and their brains were rapidly removed and placed on an ice-cold platform for dissection of the mPFC and amygdala from the brain sample. Isolated mPFC and amygdala samples were then placed in synaptoneurosome buffer containing (in mM) 10 HEPES, 1 EDTA, 2 EGTA, 0.5 DTT, and 0.5 PMSF, as well as (in μg/mL) 10 Leupeptin, 50 Soybean Trypsin Inhibitor, 5 Pepstatin, and 50 Aprotinin (pH=7.0) and homogenized. From this step forward, the homogenate was kept ice-cold at all times to minimize proteolysis throughout the isolation procedure. The homogenate was diluted further with the same volume of synaptoneurosome buffer and briefly and gently sonicated by delivering 3 pulses using an output power of 1 using a Sonic dismembrator Model 100 (Fischer Scientific, NJ, USA). The sample was loaded into a 1.0 ml Luer-lock syringe (BD syringes) and filtered twice through three layers of a pre-wetted 100 μm pore nylon filter (CMN-0105-D, Small Parts, Inc. Logansport, IN, USA) held in a 13 mm diameter filter holder (XX3001200, Millipore, MA, USA). The resulting filtrate was loaded into a 1 ml Luer-lock syringe and filtered through a pre-wetted 5 μm pore hydrophilic filter (CMN-0005-D, Small Parts, Inc. Logansport, IN, USA) held in a 13 mm diameter filter holder. The resulting filtrate was centrifuged at 1000 X g for 10 min. The pellet obtained corresponded to the synaptoneurosome fraction. Isolated synaptoneurosomes were resuspended in 75 μl of buffer solution containing 0.32 M sucrose and 1 mM NaHCO₃ (pH 7.0).

Western blotting

An equal amount (20μg/well) of synaptoneurosomes were loaded on 10% Sodium Dodecyl Sulfate (SDS) gel run at 100V for 1.5h (Mini PROTEAN^® Tetra Cell, BIO-RAD, Hercules, CA, USA); transfer was performed at 100V for 1h to nitrocellulose membrane using wet transfer (Mini-PROTEAN^® II Cell, BIO-RAD, Hercules, CA, USA). Nitrocellulose membrane was blocked in 5% milk in TBST (Tris-Buffered Saline Tween-20 (TBST; 140 mM NaCl, 20 mM Tris, and 0.1% Tween-20)) for 1h at room temperature, then incubated in GluA1 (rabbit monoclonal, 05-855R, 1:2000 dilution), GluA2 (rabbit polyclonal, AB 1768, 1:2000 dilution), GluN1 (mouse monoclonal, 05-432, 1:1000 dilution), GluN2A (rabbit polyclonal, AB1555P, 1:1000 dilution), or GluN2B (rabbit polyclonal, 06-600, 1:1000 dilution) (Millipore, Temecula, CA, USA) antibody overnight at 4°C, followed by washing with TBST (6 × 5 min) and incubated with appropriate horseradish peroxidase (HRP)-conjugated secondary antibody for 1h at room temperature, followed by washing with TBST. All secondary antibodies were prepared in 5% milk-TBST and used at 1:5000 dilution [(goat anti-rabbit HRP, 7074, goat anti-mouse HRP, 7076 (Cell Signaling Technology)]. Blots were developed using the ECL Plus Western Blotting Detection System kit (RPN2132; GE Healthcare, Piscataway, NJ, USA). Actin (rabbit polyclonal, A-5060, 1:1000 dilution) (Sigma, USA) was used as a loading control for all western blotting experiments. Images of blots were taken with an MTI CCD 72S camera (Nikon, Japan), and optical densities of the bands were analyzed using MCID Basic software version 7.0 software (Imaging Research Inc., Cambridge, UK).

Statistics

All values given as mean ± SEM. We compared the differences of means using unpaired t-test, one-way ANOVA, or two-way ANOVA with Bonferonni post-hoc test where appropriate. Data were analyzed with GraphPad Prism 4 software; values p≤0.05 were considered significantly different.

Results

Numerous studies over the last two decades have studied the neural circuits of auditory fear conditioning and fear extinction in rodents and have identified the mPFC and amygdala to be crucial in these processes. Studies have also demonstrated a crucial role of the NMDA receptor activation and mitogen-activated protein kinase (MAPK) signaling pathway in fear learning (Lu et al., 2001; Lin et al., 2003; Chen et al., 2005; Herry et al., 2006). In the first set of experiments, we studied the time-dependent effect of intraperitoneal DCS administration on c-fos and pERK and iGluRs in mPFC and amygdala subnuclei. The time-dependent experiments were performed to delineate potential short-term and long-term effects of DCS that may be responsible for acute and sustained effects of DCS in various paradigms. The second set of experiments explored the specific role of DCS administration 24h after fear conditioning and immediately after extinction on ERK activation and AMPA and NMDA receptor subunit expression in the mPFC and amygdala.

Effects of DCS on pERK and c-fos in the mPFC and amygdala

We performed c-fos and pERK immunostaining, which are markers for neuronal activity and synaptic plasticity, respectively (Bullitt, 1990; Kandel, 2001). DCS significantly increased pERK levels in both the PL (P<0.0001) and IL (P=0.0002) regions of the mPFC 30 min after administration versus saline control (Figure 1A). Interestingly, pERK staining in both the IL and PL was dramatically increased, with the highest increase in the PL region at 30 min (Figure 1B, P<0.0001). Moreover, DCS administration resulted in a sustained increase in pERK-stained neurons at all other time points after injection both in the PL (P<0.0001 at 1h, 6h, and 24h, respectively) and IL (P<0.0001 at 1h and 24h, P=0.0002 at 6h) compared to saline control. Interestingly, the regional pERK staining appeared to be time-dependent. At 30 min after injection, the pERK staining was highest in the PL, which reduced in subsequent time points. In contrast, pERK staining in the IL reached a stable plateau at 30 min, 1h, and 6h and declined only at 24h after administration of DCS. The pERK staining was most robust in layer V and VI after 30 minutes. Other layers of the mPFC, including layers II/III and V, which connect to the BLA (Gabbott et al., 2005), showed higher number of pERK-stained neurons at 1h, 6h, and 24h after DCS treatment. Layer VI neurons connect with the lateral hypothalamus (Gabbott et al., 2005), which is crucial for autonomic responses. Thus, DCS may potentially affect behaviors such as anxiety-like behavior by modulating autonomic output (Ho et al., 2005; Wu et al., 2008). DCS also significantly increased c-fos staining at 30 min and 1h after injection relative to saline control (Figure 2). Moreover, c-fos staining was significantly increased both in PL and IL after DCS treatment at 30 min (PL: P=0.0011 and IL: P=0.0003) and 1h (PL: P=0.0429 and IL: P=0.0021). Interestingly, elevation of c-fos expression persisted 24h after administration of DCS treatment in the PL (P=0.0493, Figure 2). Several studies have reported induction of c-fos expression after acute (Robertson and Fibiger, 1992; Kinon and Lieberman, 1996) and chronic (Kontkanen et al., 2002) treatment with antipsychotic drugs in mPFC and may stimulate the endogenous fear reduction system.

FIG.1.

D-cycloserine increases pERK expression in the mPFC in a time-dependent manner. (A) Representative immunohistochemical staining (DAB) of coronal sections of rat mPFC 30 min, 1h, 6h, and 24h after DCS injection versus saline (Sal) control. (B) DCS increased the pERK expression in a time-dependent manner in the PL and IL at all time points versus saline control. Data are presented as mean ± SEM. *** represents P<0.001 (n=5).

FIG. 2.

DCS increases c-fos expression in the mPFC in a time-dependent manner. (A) DAB staining of coronal sections after injection of saline or DCS shows increased c-fos staining in the PL and IL regions. (B) DCS increased c-fos expression in a time-dependent manner in the PL and IL versus saline (Sal) control (n=5). 6h after DCS administration, the number of c-fos-stained neurons in the PL significantly decreased versus saline control. Data are presented as mean ± SEM, *, **, and *** represent P<0.05, P<0.01, and P<0.001, respectively. # represents decrease in c-fos expression in DCS-treated group compared to saline group, P<0.05).

We also assessed the effect of DCS on pERK expression in the amygdala and observed a sustained and significant reduction of pERK staining localized to the medial subdivision of the central nucleus of the amygdala (CeM) at all time points (P<0.0001 and P=0.0001 at 30 min to 6h and 24h, respectively) after injection of DCS. We also observed a decrease in pERK expression in the lateral subdivision of the central nucleus of the amygdala (CeL) after DCS administration (P=0.0076 and P=0.0192 at 1h and 6h, respectively). In DCS-treated groups, pERK expression was higher in CeL compared to CeM at each time point, but reached significance at 30 min and 6h (P=0.0009 and P=0.0133 at 30 min and 6h, respectively). Interestingly, we did not see a reduction in pERK expression in other nuclei of the amygdala (Figure 3). These results demonstrate that systemic administration of DCS affects mPFC and specific nuclei in the amygdala.

FIG. 3.

Systemic DCS administration decreases pERK staining in the medial and lateral subdivision of the central amygdala (CeA). (A) Representative DAB staining of rat coronal sections of the amygdala after saline or DCS injection. (B) DCS significantly decreased pERK expression in the CeM at all time points tested and 1h and 6h in CeL compared to saline (n=4) (unpaired t-test). Data are presented as mean ± SEM. # represents P<0.05, ±±represents P<0.01, ***and ### represent P<0.001.

Effect of DCS on iGluRs expression in mPFC and amygdala

To study the effect of DCS on iGluR expression in the mPFC and amygdala, we prepared the synaptoneurosomes of both brain regions from saline- or DCS-treated animals 6h after administration. DCS treatment significantly increased the expression of GluA1 (P=0.0404), GluN1 (P=0.0153), GluN2A (P=0.0217), and GluN2B (P=0.0120) in the mPFC but not in the amygdala (Figure 4A & B). One potential mechanism of DCS-induced increase in iGluRs in the mPFC may be via modulation of synaptic plasticity; previous studies have shown that DCS administration recovers behavioral deficits by changing the expression level of AMPA and GluNs (Yadav et al., 2012).

FIG. 4.

Systemic administration of DCS for 6h significantly increases the expression level of iGluR in the mPFC. (A) Normalized expression level of iGluR proteins shows that DCS significantly increased GluA1, GluN1, GluN2A, and GluN2B expression in the mPFC but not in the amygdala. (B) Representative protein expression from synaptoneurosome preparation collected 6h after systemic injection of saline or DCS. Data are presented as mean ± SEM. * represents P<0.05.

Effects of DCS on fear conditioning and fear extinction recall

Previous studies have identified that both systemic and intra-amygdala infusion of DCS facilitates fear extinction, as assessed by fear-potentiated startle and freezing behavior (Walker et al., 2002; Ledgerwood et al., 2003). Similar to previous work, we found that systemic injection of DCS at 30 mg/kg significantly reduced freezing during extinction recall testing (Figure 5C). During fear conditioning training, animals presented with paired CS-US showed significantly higher freezing response compared to the context group (P<0.01 and P<0.001 for cond-saline and cond-DCS respectively). We also studied the effect of shock by presenting unpaired CS and US, and observed freezing response equal to the context group (Figure 5A). We also studied the effect of DCS on fear conditioning. Animals were administered with saline or DCS 24h after fear conditioning. The freezing response was similar in the two groups, suggesting that DCS does not reduce fear in the absence of extinction training (Figure 5C). This finding is in line with an earlier study by Walker et al. (2002). For the extinction training, animals were predetermined for the treatment of DCS or saline after fear conditioning. A mixed factor ANOVA shows no difference between groups predetermined for ext-sal [F(1,16)=0.8, P=0.4] or ext-DCS [F(1,15)=0.2, P=0.6] during extinction training (Figure 5B). However, 24h after extinction training, there is a significant effect of the drug on freezing behavior during exposure to the CS, with the group that received DCS immediately after extinction training showing significantly less freezing than the saline group [unpaired t-test; P=0.003, F=3.8] (Figure 5C).

FIG. 5.

DCS facilitates fear extinction. Percent freezing during (A) fear conditioning, (B) extinction training, and (C) extinction testing. DCS administration after fear extinction training (arrow) significantly facilitates fear extinction 24h after extinction training (unpaired t-test; P=0.003, F=3.8). Data are presented as mean ± SEM. ** represents P<0.01.

Effects of DCS following fear conditioning and extinction training on ERK activation and expression of iGluRs in the mPFC and amygdala

We next explored the effect of DCS administration 24h after conditioning and immediately after extinction training on pERK activation in the mPFC and amygdala. We found that the extinction trained-saline group had a significant increase in pERK-positive neurons in the IL compared to no extinction groups (P< 0.001 for context, unpaired, cond-sal, and cond-DCS in one-way ANOVA) (Figure 6A–B). Our findings are in agreement with Kim et al. (2009). In addition to increased pERK-positive neurons in the IL, we also found significantly higher pERK-positive neurons in the PL after extinction training (P<0.001 for context, unpaired, cond-sal, and cond-DCS in one-way ANOVA) (Figure 6A–B). Additionally, we found that DCS treatment (30 mg/kg) immediately after extinction training facilitated extinction-induced pERK expression in the IL (P<0.01) and PL (P<.001) compared to saline control (Figure 6A–B). These data suggest that both the PL and IL are active during extinction. Increase in pERK-positive neurons in the PL is in contrast to the study by Kim et al. (2009). However, contrasting evidence is also reported supporting our finding (Herry and Mons, 2004; Santini et al. 2004; Park and Choi, 2009; Kim et al., 2010; Barrett et al., 2003).

FIG. 6.

DCS administration after extinction training increases the number of pERK-positive neurons, and extinction training modulates iGluR protein expression in the mPFC. (A) Extinction training significantly increased the number of pERK-stained neurons in both the PL and IL compared to context, unpaired, cond-sal, and cond-DCS. Ext-DCS further increased the number of pERK-stained neurons in both the PL and IL compared to ext-sal (one-way ANOVA). Data are presented as mean ± SEM (n=6). (B) DAB staining of coronal sections of context, unpaired, condition-saline (cond-sal), condition-DCS (cond-DCS), extinction-saline (ext-sal), and extinction-DCS (ext-DCS) shows an increase in pERK staining in the PL and IL in the ext-DCS group compared to the ext-sal and other groups. (C) Normalized expression level of iGluR proteins shows that extinction training increases GluA1, GluN1, and GluN2A expression in mPFC compared to context, unpaired, cond-sal, and cond-DCS (one-way ANOVA). Administration of DCS after extinction training (ext-DCS) versus ext-sal showed similar protein expression levels (n=6). (D) Representative protein expression from synaptoneurosome preparation collected from all groups. Data are presented as mean ± SEM. * (# or $), ** (## or $$), and *** (### or $$$) represent P<0.05, P<0.01, and P<0.001 versus context and unpaired groups, conditioned control, or extinction-saline control, respectively.

Increase in pERK in the IL by DCS suggests enhanced activity in the IL and may potentially facilitate extinction. Our study is supported by various past findings (Farinelli et al., 2006; Herry and Garcia, 2002; Hugues and Garcia, 2007; Milad and Guirk, 2002). We observed no difference in the expression of pERK in the cond-sal group compared to the context and unpaired groups. Moreover, DCS administration in conjunction with fear conditioning showed no difference compared to the cond-sal group, suggesting that the effect of DCS alone on pERK may be neutralized by fear conditioning.

Neuronal plasticity involves changes at the level of AMPA and NMDA receptor expression, which leads to strengthening or reversal of strengthening of synapses. In order to gain insight into the molecular changes at the synaptic level that may underlie the facilitatory effect of DCS on fear extinction, we isolated synaptoneurosomes that are composed of glutamatergic synapses at the dendritic spine, containing both presynaptic and postsynaptic components (Hollingsworth et al., 1985; Hampson et al., 1992; Heynen et al., 2000). We observed a significant increase in expression of GluA1 (P<0.001 for context and unpaired and P<0.01 for cond-sal and cond-DCS), GluN1 (P<0.01 for context and unpaired, P<0.001 for cond-sal and P<0.05 for cond-DCS), and GluN2A (P<0.001 for context, unpaired, cond-sal and cond-DCS) in synaptoneurosomal preparation in the mPFC in the extinction-saline group in comparison to no extinction groups (one way ANOVA). This result correlates with an increase in the synaptic plasticity marker pERK in mPFC by extinction training (Kim et al., 2010; our results). Additionally, GluA1 (P<0.001 for context, unpaired, cond-sal and cond-DCS) and GluN2A (P<0.001 for context, unpaired and cond-sal and P<0.01 for cond-DCS) was significantly higher in ext-DCS group compared to no extinction groups (one-way ANOVA) (Figure 6C and D). However, no difference was observed between extinction groups treated with either saline or DCS (one-way ANOVA, P>0.05). This is in contrast to the enhanced pERK by DCS after extinction training the in mPFC, and may indicate that, although DCS increases signaling, it does not have a significant effect on protein expression at that time point. Moreover, we observed no difference in the expression of iGluRs in the cond-sal group compared to the context and unpaired groups; DCS administration in conjunction with fear conditioning did not show any difference compared to the cond-sal group.

In addition to the prefrontal cortex, DCS treatment immediately after extinction training significantly affected pERK activation in the amygdala (Figure 7). The modulation of pERK staining in the amygdala appeared to be region-specific (Figure 7A). In our study, we did not observe fear conditioning-induced pERK activation in the LA compared to the context and unpaired groups (data not shown). Additionally, DCS treatment in conjunction with fear conditioning training did not modulate pERK expression. A previous study by Schafe et al. (2000) demonstrated a transient increase in pERK expression (60 min) after fear conditioning in the LA. The possible explanation may be the difference in time of analysis; in our experiment, we did pERK staining 24h after fear conditioning, suggesting that the increase in pERK expression may not be persistent. However, we did observe an increase in pERK expression in the basal amygdala (BA) (P<0.01) (Figure 7A). This result is in agreement with previous findings (Schafe et al., 2000 and Brambilla et al., 1997), demonstrating a crucial role of ERK/MAPK activation in the BA in memory consolidation. Collectively, these finding suggest that ERK/MAPK activation is an important event, whereby long-term memories are established. We observed an increase in pERK expression in the BA in the ext-sal group compared to the context and unpaired groups, in agreement with Herry et al. (2006). DCS in conjunction with extinction training significantly increased pERK staining in both the CeL (P<0.001) and the BSTIA regions that includes ITC (P<0.001) compared to extinction-saline rats (Figure 7B). Thus, our results demonstrate that DCS facilitates the cortico-amygdala circuitry involved in fear extinction, which may underlie its ability to facilitate fear extinction. In expression experiments, we found that there was a trend that expression of iGluR was lower in the ext-sal group compared to fear conditioning groups, but this was not significant (one-way ANOVA, Figure 7D and E). Interestingly, ext-DCS appeared to reverse the effect observed in the ext-sal group. This significant increase from the ext-saline to the ext-DCS group was observed in GluA1 and GluN2B expression (one-way ANOVA, P<0.05) (Figure 7D and E). This trend of an increase in expression of glutamate receptors by DCS in the amygdala compared to saline is in contrast to trends observed in the mPFC, where protein expression appears unchanged between these two groups and may underlie the effect of DCS on facilitating fear extinction.

FIG. 7.

DCS administration after fear extinction training increases pERK staining in selective regions in the amygdala and also increases GluA1 and GluN2B protein level. (A and B) Histological counts of pERK-stained neurons in the BA, ITC, and CeM and CeL regions of the amygdala. DCS administration in conjunction with extinction decreased the number of pERK-stained neurons in the BA and increased pERK staining in the ITC and CeL compared to context, unpaired, cond-sal, cond-DCS, and ext-saline (one-way ANOVA). (C) Representative DAB staining of coronal sections collected from context, unpaired, cond-sal, cond-DCS, ext-sal, and ext-DCS groups. (D) Administration of DCS following extinction training increased GluA1 and GluN2B expression (one-way ANOVA) (n=6). (E) Representative protein expression from synaptoneurosome preparation collected from all groups. Data are presented as mean ± SEM. * (# or $), ** (## or $$), and *** (### or $$$) represent P<0.05, P<0.01, and P<0.001 versus context and unpaired groups, conditioned control, or extinction-saline control, respectively.

Discussion

One of the most important findings of this study is that DCS alone or in conjunction with extinction training led to a strong facilitation of pERK in the mPFC. Additionally, DCS alone led to an increase in c-fos in both the IL and PL regions, reduced pERK levels in the CeM, and increased iGluR expression in the mPFC. The increase in pERK and iGluRs suggests that DCS may augment local synaptic plasticity in the mPFC implicated in extinction and re-extinction (Hugues et al., 2004; Santini et al., 2008; Peters et al., 2009), as well as in various models of extinction of addiction. This is a novel finding that may partly explain why the effects of DCS are not limited to the fear extinction paradigm, but also influences substance abuse and social behaviors that have been shown to involve mPFC activity (Aupperle et al., 2009; Bertotto et al., 2006; Yadav et al., 2012).

Accumulating evidence points to an extinction circuitry that involves the IL, ITC, and CeA (Sierra-Mercado et al., 2011; Thompson et al., 2010; Chang et al., 2010; Sotres-Bayon and Quirk, 2010; Peters et al., 2009; Laurent and Westbrook, 2009; Likhtik et al., 2008 and Amano et al., 2010). The mPFC sends excitatory input to the GABAergic interneurons of both ITC and CeL (McDonald et al., 1996; Sesack et al., 1989; Quirk et al., 2003), which in turn inhibit CeM output neurons, resulting in a decrease in fear response (Quirk et al., 2006). In addition studies have also reported that the BA has two populations of neurons designated as “fear neurons” and “extinction neurons” that show CS responsiveness upon fear conditioning and fear extinction, respectively. Interestingly, the extinction neurons have dense reciprocal connections with the mPFC (McDonald, 1998). Extinction neurons in the BA may exert an inhibitory effect on the CeM neurons, leading to the suppression of conditioned fear (Likhtik et al., 2008). Our results suggest that DCS immediately after extinction modulates pERK in all of these mPFC and amygdala nuclei compared to saline.

Based on the profound effect of DCS alone on pERK in mPFC, we predict that the DCS effect on the various amygdala subnuclei, when given in conjunction with extinction training, may be downstream of its effect on the mPFC. Our current finding also correlates with work of Santini et al. (2008), showing that the intrinsic excitability of IL pyramidal neurons increased after extinction training, ultimately leading to an increase in synaptic plasticity, marked by increased pERK expression. We observed no significant difference between the ext-saline and ext-DCS groups in mPFC iGluR expression, nor did there appear to be an obvious trend associated with DCS administration. However, compared to groups that did not undergo fear extinction training (context, unpaired, cond-sal, and cond-DCS), there was a significant increase in GluA1 and GluN2A expression associated with extinction training (both ext-saline and ext-DCS) in the mPFC, as well as an increase in GluN1 in mPFC of ext-sal animals. Increased GluA1 expression in extinction-trained rats is consistent with the hypothesis that extinction induces LTP in the mPFC (Rosenkranz and Grace, 2001; Rosenkranz et al., 2003; Royer and Pare, 2002; Quirk et al., 2003). However, lack of change in the mPFC due to DCS administration after extinction training suggests that, while ext-DCS increases pERK activity in the mPFC, it does not significantly alter protein expression in the mPFC at the same time point. Conversely, it is possible that extinction training alone increases levels of GluA1 and GluN2A to saturation, such that DCS administration cannot significantly change the level of iGluR expression in the mPFC. Another possibility is that the 6h time point is not sufficient to induce the changes in protein synthesis in the mPFC after DCS administration. Meanwhile, the dramatic effect of DCS administration on iGluR expression in the amygdala is in agreement with intra-amygdala infusion of DCS potentiating fear extinction (Walker et al., 2002), while no observable difference in synaptoneurosomal protein expression in the mPFC is consistent with intra-mPFC infusion of DCS not facilitating fear extinction but rather influencing re-extinction (Chang and Maren, 2011). Thus it appears that DCS may have a dual effect on cortico-amygdala circuitry with different temporal patterns. Specifically, DCS may have a more short-term effect in the amygdala and a longer-lasting effect in the mPFC. Further studies will be required to determine the precise temporal pattern of DCS action.

We found that DCS may facilitate fear extinction consolidation by augmenting GluN2B expression in the amygdala, thereby strengthening amygdalar synaptic transmission. Activation of NMDA receptors in the mPFC and/or amygdala is required for consolidation of extinction memory (Santini et al., 2001; Herry and Garcia, 2002; Milad and Quirk, 2002). However, previous studies have shown that a pre-training blockade of GluN2B in the LA leads to an impairment of acquisition of fear extinction but not the consolidation of fear extinction (Sotres-Bayon et al., 2007; Laurent et al., 2008; Dalton et al., 2012; Gourley et al. 2009; Bauer et al. 2002). Although these findings suggest that NMDA receptors, specifically GluN2B-containing NMDA receptors, in the amygdala are not necessary for consolidation of fear extinction, other studies have shown a crucial role of the BLA in the consolidation of fear extinction (Akirav et al. 2006; Berlau and McGaugh, 2006). It will be necessary in future studies to determine the exact role of DCS administration on GluN2B-containing NMDA receptor function after fear extinction training.

Accumulating evidence suggests that MAPK/ERK-dependent activation of transcription is important for consolidation of long-term extinction memory (Herry et al., 2006; Herry and Mons, 2004). Immediate post-extinction infusion of the MAPK inhibitor PD098059 in the mPFC impairs the extinction of both auditory and contextual fear memory, while showing no effect if administered 4h after extinction training (Hugues et al., 2004 and 2006). Studies have also shown a requirement for immediate protein synthesis in consolidation of extinction memory (Santini et al., 2004), since infusion of anisomycin, a protein synthesis inhibitor, before extinction training impairs extinction recall. Interestingly, infusion of anisomycin 4h after extinction training has no effect on fear extinction recall, suggesting a short window during which extinction-related protein synthesis occurs. Consistent with a time-dependent role of molecular processes in the consolidation of extinction, we chose a 6h time point after extinction training for our study to allow the molecular processes required for extinction recall in this time window. Similar to the mPFC, in the hippocampus (HPC), the MAPK/ERK pathway controls the rapid local translation during stabilization of newly formed memories (Kelleher et al. 2004; Thomas and Huganir, 2004). Additional studies have demonstrated the expression of pERK in the HPC during extinction training (Fisher et al., 2007; Tronson et al., 2009), similar to the mPFC. Interestingly, HPC is known to be involved in extinction (Cannich et al., 2004; Corcoran et al., 2005), and projections from the HPC to the mPFC have been shown to display LTP (Laroche et al., 2000; Romcy-Pereira and Pavlides, 2004). Studies have also shown MAPK/ERK-dependent mRNA translation in the BLA to be involved in acquisition of extinction (Lin et al., 2003). Moreover, DCS facilitates extinction by activating pERK expression in the BLA, which is impaired by infusion of MAPK inhibitors, as well as transcription (actinomycin D) and translation (Anisomycin) inhibitors (Yang and Lu, 2005), suggesting that DCS facilitates extinction by activating the MAPK-dependent signaling cascade and new protein synthesis in the amygdala. In our study, we observed an increase in pERK and iGluRs (GluA1 and GluN2B) due to DCS administration after extinction training in the amygdala, supportive of MAPK/ERK being an upstream regulator of protein synthesis (Kelleher et al., 2004; Thomas & Huganir, 2004). Our study is in line with these converging findings demonstrating a crucial role of MAPK in fear extinction.

Increase in pERK expression in both the PL and IL during extinction training suggest activation of both of these regions of the mPFC. The activation of the PL during fear extinction may be accounted for by two reasons. First, an extinction session involves retrieval of fear, at least in the beginning of the session, which involves activation of the PL. In contrast, the IL will activate during a later phase of extinction training. Thus, the similar pattern of PL and IL activation of pERK by DCS after 6h of extinction training probably represents two temporally distinct events. A second explanation of pERK expression in the PL is unrelated to its role in fear conditioning. The PL and IL are involved synergistically to bring appropriate response in a situation in which discontinuity of context must be detected (Gisquet-Verrier and Delatour, 2006). This explanation suggests that the role of the PL might be interchangeable rather than contradictory to that of the IL in some circumstances.

Extinction is thought to involve inhibitory learning such that the original fear memory persists and a new memory imparts safety to the previously aversive CS (Bouton & King, 1983). The behavioral features of spontaneous recovery, renewal, and reinstatement of fear suggest that the original memory remains intact after extinction training. At a molecular level, the precise effect of extinction on fear conditioning-induced enhancement of thalamic-lateral amygdala synapses are less clear. There is evidence for both reversal of fear conditioning-induced enhancement and inhibitory learning due to extinction training (Mao et al., 2006; Kim et al., 2007). Which process dominates likely depends on the strength of conditioning and the extinction training protocol. It has been previously proposed that DCS leads to erasure of fear memory (Mao et al., 2006). This hypothesis is supported by impaired spontaneous recovery and reinstatement by DCS in conjunction with extinction training (Ledgerwood et al., 2004). Moreover, at a molecular level, extinction training with DCS, unlike extinction training alone, reverses fear conditioning-induced surface GluA1 increase in the amygdala, raising the possibility of a memory erasure hypothesis rather than inhibitory learning (Mao et al., 2006). Our findings that DCS enhances pERK activation in the mPFC, ITC, and CeL, which form the fear extinction circuitry, suggests that DCS may further facilitate the already engaged extinction circuitry. Additionally, we observed an increase in iGluR expression in the amygdala 6h after extinction training in the DCS groups compared to saline, suggesting that protein synthesis or trafficking of iGluRs is enhanced due to DCS administration and may represent molecular correlates of synaptic strengthening. These results suggest that DCS may enhance new learning. Overall, our results suggest that DCS action on fear extinction may not be limited to the LA and may involve other brain regions, including the mPFC, ITC, and CeL, which may emanate from specific GluN2 subunits expressed in these regions. Thus future studies will be required to address the effect of DCS on synaptic strengthening not limited to the LA.

Abbreviations

DCS

D-cycloserine

pERK

phospho-extracellular signal-related kinase

mPFC

medial prefrontal cortex

CeA

central amygdala

CeM

medial subdivision of the central amygdala

CeL

lateral subdivision of the central amygdala

ITC

intercalated cells of the amygdala

IL

infralimbic prefrontal cortex

PL

prelimbic prefrontal cortex

LA

lateral amygdala

BLA

basolateral amygdala

HPC

hippocampus