- Email: [email protected]
Effects of d -cycloserine on extinction of learned fear to an olfactory cue
Neurobiology of Learning and Memory 87 (2007) 476–482 www.elsevier.com/locate/ynlme
EVects of D-cycloserine on extinction of learned fear to an olfactory cue Marianne Weber, Joe Hart, Rick Richardson ¤ School of Psychology, University of New South Wales, Sydney 2052, Australia Received 11 July 2006; revised 28 November 2006; accepted 21 December 2006 Available online 1 February 2007
Abstract D-cycloserine (DCS), a partial NMDA receptor agonist, facilitates extinction of learned fear in rats and has been used to treat anxiety disorders in clinical populations. However, research into the eVects of DCS on extinction is still in its infancy, with visual cues being the primary fear-eliciting stimuli under investigation. In both human and animal subjects odors have been found to associate strongly with aversive events. Therefore, this study examined the generality of the eVects of DCS on extinction by testing odor cues. Sprague–Dawley rats were conditioned and extinguished to an odor using varying parameters, injected with either saline or DCS (15 mg/kg) following extinction, and then tested for a freezing response 24 h later. Experiment 1 demonstrated that after 3 odor-shock pairings, rats did not display short-term extinction and DCS had no eVect on long-term extinction. Experiment 2 demonstrated that after 3 odor-noise pairings, rats displayed signiWcant short-term extinction and DCS signiWcantly facilitated long-term extinction. Following 2 odor-shock pairings in Experiment 3, half the rats displayed short-term extinction (“extinguishers”) and half did not (“non-extinguishers”). DCS facilitated longterm extinction in the “extinguishers” condition but not in the “non-extinguishers” condition. In Experiment 4, following 2 odor-shock pairings and an extra extinction session, DCS had a signiWcant facilitatory eVect on long-term extinction. Thus, extinction of freezing to an odor cue was facilitated by systemic injections of DCS, but only when some amount of within-session extinction occurred prior to injection. © 2006 Elsevier Inc. All rights reserved.
Keywords: Rats; D-Cycloserine; Fear extinction; Odors; Freezing
1. Introduction Anxiety disorders such as Phobia, post-traumatic stress disorder (PTSD), and Obsessive–Compulsive Disorder are often treated with exposure-based cognitive behavior therapy (Andrews, Grino, Hunt, & Page, 1994; Foa & Kozak, 1986; Thyer, Baum, & Reid, 1988; Zarate & Agras, 1994). This type of therapy involves exposing patients to a fearful, or anxiety-provoking, stimulus within a calm and safe environment. Gradually, over successive presentations of the stimulus within this environment, patients learn to respond without fear or anxiety. This type of treatment is procedurally very similar to ‘extinction’ training in animal models of emotional learning. For example, learned fear responses to
*
Corresponding author. Fax: +61 2 9385 3641. E-mail address: [email protected] (R. Richardson).
1074-7427/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.nlm.2006.12.010
a conditioned stimulus (CS; e.g., a light that had previously been paired with shock) can be reduced by repeatedly presenting the CS in the absence of shock. Eventually the animal learns that the CS no longer predicts the occurrence of shock and fear reactions decrease in amplitude and/or frequency. Although very simple procedurally, extinction has proven to be an extremely diYcult process to understand. Nonetheless, considerable eVort has been directed at trying to identify techniques that render extinction—and potentially exposure therapy—more eVective. One signiWcant advance in this area has been the demonstration that D-cycloserine (DCS) can facilitate fear extinction. DCS was used because of its indirect agonist properties at N-methylD-aspartate receptors, which are critical for the neuronal plasticity required for emotional memory (see, Davis, Ressler, Rothbaum, & Richardson, 2006). Pre-clinical studies have demonstrated that DCS is a potent facilitator of fear
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
extinction in rats (Ledgerwood, Richardson, & Cranney, 2003, 2004, 2005; Parnas, Weber, & Richardson, 2005; Walker, Ressler, Lu, & Davis, 2002; Yang & Lu, 2005). In addition, DCS has been found to enhance the eVectiveness of exposure-based therapy in people treated for acrophobia (Ressler et al., 2004) and Social Anxiety Disorder (Hofmann et al., 2006). However, research on the eVects of DCS on extinction is still in its infancy, and it is still not known whether DCS is eVective at facilitating extinction to any, or all, types of feared stimuli, or whether its eVectiveness is limited in some way. The aim of the current study was to test the ‘generality’ of the eVects of DCS on extinction. Both pre-clinically and clinically, DCS has primarily been used in conjunction with exposure to visual cues. However, fear-eliciting cues are not always visually-based. Odors, in particular, have been found to be especially eVective in cuing emotional memory in nonclinical populations (Herz, 1998; Herz, Eliassen, Beland, & Souza, 2004), and can evoke strong trauma-related memories in people with PTSD (Vermetten & Bremner, 2003). Olfactory stimuli are also extremely salient and powerful cues for memory retrieval in rats. For example, conditioned fear associations using olfactory cues in rats are encoded more rapidly (Paschall & Davis, 2002), and require more trials to extinguish than do visual cues (Richardson, Tronson, Bailey, & Parnas, 2002). These Wndings suggest that odor cues may be processed by a diVerent memory system. Indeed, in rats the neuroanatomical pathways subserving olfactory information are slightly diVerent than for stimuli of other sensory modalities. For example, visual and auditory sensations are processed within the thalamus before being sent to the amygdala, which is the brain region crucial for recalling emotional associations. However, olfactory information is transmitted directly to the amygdala (Price, 1973), suggesting that odors have ‘privileged’ access to the amygdala and may prompt recall for emotional associations more quickly and easily. Therefore, because (1) odor-evoked memories can be powerful triggers for the emotional component of memory, and (2) the neuroanatomical substrates of odor-evoked emotional memories may diVer to that for visually-evoked memories, the present study was designed to determine whether DCS facilitates extinction of conditioned fear to an odor CS in rats. 2. Methods and materials 2.1. Subjects Experimentally-naive, male Sprague–Dawley rats (450–700 g) obtained from the breeding colony maintained by the School of Psychology at the University of New South Wales were used. The rats were housed in groups of eight in plastic boxes (67 cm long £ 40 cm wide £ 22 cm high) in a colony room with a natural light cycle. Food and water were continuously available. All experimental procedures were approved by the Animal Care and Ethics Committee of the University of New South Wales and adhered to the ethical guidelines described in The Australian Code of Practice for the Care and Use of Animals for ScientiWc Purposes (2004). Prior to experimentation, all rats were handled for approximately 5 min on 3 consecutive days.
477
2.2. Apparatus In all experiments conditioning occurred in one context and extinction training and testing occurred in a diVerent context. All experimental cages were located within a sound- and light-attenuating wood cabinet where a ventilation fan provided a 60-dB ambient noise level in each chamber. Conditioning occurred in one of 4 identical cages (20 cm long £ 12 cm wide £ 12 cm high). The front wall, rear wall, and ceiling of each cage were made of Plexiglas, and the side walls and Xoor consisted of stainless steel bars, with 1.3 cm between each bar. Each cage was suspended 13 cm above the Xoor of the wood cabinet and a tray of bedding was placed under each cage, which was changed between each rat. During conditioning, the doors to the wood cabinets remained open and the room was illuminated by a red light. The context used for extinction training and testing in Experiments 1 and 2 diVered slightly from that used in Experiments 3 and 4. In Experiments 1 and 2, extinction training and testing occurred in one of 2 identical cages (30 cm long £ 22.5 cm wide £ 30 cm high) with a clear Perspex front, black and white striped side and rear walls, and a wood hinged lid. The Xoor consisted of stainless steel bars, with 1.0 cm between each bar, and was suspended 8.5 cm above the Xoor of the wood cabinet. In Experiments 3 and 4, extinction training and testing occurred in one of 2 identical clear Perspex cages (29 cm long £ 28 cm wide £ 29 cm high). The Xoor consisted of stainless steel bars, with 1.5 cm between each bar, and was suspended 10 cm above the Xoor of the wood cabinets. A white light bulb in each cabinet provided illumination at all times during extinction and testing. Between sessions, these chambers were wiped with 0.5% eucalyptus solution and the bedding below the grid Xoor was changed. Extinction and test sessions were recorded for later assessment.
2.3. Procedure In all experiments the CS was a grape odor (0.1 ml in a plastic jar; #182380019 from Wild Flavours, Heidelberg) and the method of presentation involved the odor jar being placed under the experimental cage for the required time period and then being removed and capped with an air-tight lid. The unconditioned stimulus (US) varied for each experiment. The procedures for Experiments 1 and 2 were identical except that a diVerent US was used. In Experiment 1 the US was a 1 s, 0.6 mA footshock and in Experiment 2 the US was a 100 ms, 120 dB white noise stimulus (rise-fall time <5 ms). In each experiment a 2 £ 2 factorial design was employed, where the Wrst factor was drug condition (DCS or saline) and the second factor was condition (extinction or no extinction). This produced 4 groups: extinction + DCS (E-DCS), extinction + saline (E-Sal), no-extinction + DCS, and no-extinction + saline; the latter two groups were subsequently pooled into a single group (NE-Pooled). On Day 1, all rats were pre-exposed to the conditioning context for 20 min. Three hours later they were returned to the conditioning cages and received 3 pairings of the CS and US. That is, 120 s after being placed in the conditioning cages, the odor-jar was placed beneath the cage. Ten seconds later, the US was presented and then the odor-jar was removed. The 3 CS–US pairings occurred 120 s apart and rats were removed from the conditioning cages 120 s after the last pairing. On Day 2, rats allocated to the extinction groups were placed in the extinction cages. After 2 min in the cage, rats were presented with the jar containing the odor 6 times (each exposure was 2-min long, with a 2-min ITI). The rats were removed from the extinction cages 120 s after the last extinction trial, and half were injected with saline (E-Sal) and half were injected with DCS (E-DCS). Rats that did not receive extinction training were removed from their home cage and injected with either saline or DCS (NE-pooled). After injection all rats were returned to their home cage. On Day 3, all rats were tested for odor-elicited freezing. That is, 2 min after being placed in the cage, the odor was presented for 3 min. The procedure for Experiments 3 and 4 was identical to Experiments 1 and 2 except that rats received 2 CS–US pairings and the US was a 0.5 s, 0.6 mA footshock. After extinction training in Experiment 3 rats were
478
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
divided into 2 groups where half received an injection of DCS and half received saline before being returned to the home cage. All rats were tested for odor-elicited freezing the following day. For the purpose of analysis, these groups were further divided equally into ‘extinguishers’ and ‘nonextinguishers’. Those rats that had the lowest levels of freezing at the end of extinction were put into the “extinguishers” condition while those with the highest levels of freezing at the end of extinction were placed in the “non-extinguishers” condition. In Experiment 4, rats received a second extinction training session that occurred 45–90 min after the Wrst. After this second extinction session, half the rats received DCS and half received saline. All rats were tested for odor-elicited freezing the following day.
2.4. Data analysis Each rat was scored for freezing during the 6 extinction trials (or the 12 trials in Experiment 4) and during the test on Day 3. Freezing was characterized by an absence of all bodily movements except those required for breathing (Fanselow, 1980), and was measured using a time sampling method. An observation was made every 5 s and a positive or negative count was made. A percentage score was then calculated for the proportion of the total observation period spent freezing. For the test session on Day 3, freezing during the 120 s prior to the Wrst odor presentation (pre-CS) was scored as an indication of baseline levels of fear. Analysis of variance (ANOVA) was the primary statistical approach, and post hoc comparisons were made with Tukey’s honestly signiWcant diVerence (HSD) test.
2.5. Drugs D-cycloserine (Sigma–Aldrich, Castle Hill, New South Wales) was freshly dissolved in sterile saline at a concentration of 15 mg/ml and injected subcutaneously in a volume of 1 ml/kg.
100
3.1. Experiment 1 Rats that received extinction training displayed high levels of conditioned freezing to each of the 6 odor presentations during the extinction session (Fig. 1a). Overall, levels of freezing did not signiWcantly decline across the extinction session, as conWrmed by a mixed-design ANOVA with 6 time points, F < 1. In addition, rats allocated to the saline condition (n D 8) displayed similar levels of freezing to rats that were to receive DCS (n D 8). At test, rats that received DCS but no extinction training did not display a diVerent level of freezing to rats that received saline and no extinction [t (6) <1]. Therefore, these groups were pooled (NE-Pooled). Overall, there were no group diVerences in the amount of pre-CS freezing [F (2, 23) D 2.49, p D .11]. Additionally, there were no group diVerences in the amount of freezing to the odor CS [F (2, 23) D 1.85, p D .18] (Fig. 1b). Thus, rats that received DCS immediately after extinction training to an odor CS previously paired with a shock US did not freeze less at test than rats that received saline, indicating that DCS did not facilitate extinction. However, no decline in the fear response was observed in either extinction group across the six 2-minute extinction trials. It may be the case that DCS does not facilitate extinction retention if no short-term extinction learning
b
75
% Freezing
% Freezing
a
3. Results
50 25
DCS SAL
100 75 50 25 0
0 0
1
2
3
4
5
6
E-DCS
Trial
d
100 75
% Freezing
% Freezing
c
50 25
E-SAL
NE-Pooled
Condition
DCS SAL
100 75 50 25 0
0 0
1
2
3
Trial
4
5
6
E-DCS
E-SAL
NE-Pooled
Condition
Fig. 1. (a) Mean freezing during extinction trials in Experiment 1 where rats had received odor–shock pairings. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (b) Mean freezing during test in Experiment 1. Rats had either been extinguished (E) or not (NE), and injected with saline (SAL) or DCS. Half of the rats in the NE group were injected with saline while the others were injected with DCS; these rats were collapsed into a single control group (pooled). (c) Mean freezing during extinction trials in Experiment 2 where rats had received odor–noise pairings. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (d) Mean freezing during test in Experiment 2. Rats had been extinguished (E) or not (NE), and injected with saline (SAL) or DCS. Half of the rats in the NE group were injected with saline while the others were injected with DCS; these rats were collapsed into a single control group (pooled).
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
is observed. Alternatively, it may be the case that DCS does not aVect extinction of conditioned fear when odor cues are used. The aim of Experiment 2 was to condition rats to an odor cue in a manner in which short-term extinction was observed. That is, rats were conditioned with a US that is less aversive than a shock, namely a loud noise. 3.2. Experiment 2 Two rats did not display a robust conditioned response to the odor (mean levels of freezing were less than 10% across the 6 extinction trials) and were excluded from all statistical analysis. Rats that received extinction training displayed high levels of conditioned freezing to the odor during the Wrst extinction trial and displayed signiWcantly lower levels of freezing across trials [mixed-design ANOVA with 6 time points, F (5, 120) D 39.0, p D .001] (Fig. 1c). Rats that were to receive saline (n D 12) did not diVer on the amount of freezing during each extinction trial to rats that were to receive DCS (n D 14) [non-signiWcant interaction, F < 1 and a non-signiWcant between groups eVect, F (1, 24) D 1.17, p D .29]. At test, rats that received no extinction training and DCS did not display a diVerent level of freezing to rats that received saline and no extinction [t (10) <1]. Therefore, these groups were pooled (NE-Pooled, n D 12). PreCS freezing was signiWcantly higher for rats that received extinction training and saline injections (E-SAL) than rats in the E-DCS and NE-Pooled conditions [F (2, 37) D 4.01, p D .03, with post hoc comparisons signiWcant at p < .05]. Thus, freezing to the CS was analyzed with an ANCOVA, with pre-CS freezing as a covariate. During the 3 min odor CS presentation, rats that received DCS following extinction (E-DCS) froze signiWcantly less than rats that received saline following extinction (E-Sal), and signiWcantly less than rats that were not given extinction training (NE-Pooled; F (2, 37) D 5.32, p D .01, with post hoc comparisons signiWcant at p < .05). Further, rats that received extinction training and saline injections (E-SAL) did not display signiWcantly diVerent levels of freezing to rats that did not receive extinction training (NE-Pooled) (Fig. 1d). Experiment 2 demonstrated facilitated extinction to an odor CS when DCS was injected after an extinction training session in which short-term extinction was observed. This experiment demonstrates that long-term extinction of odor-elicited fear can be facilitated by DCS, and suggests that the failure to observe facilitation of extinction retention in Experiment 1 was due to the high level of freezing maintained during extinction training. However, the diVerence in US modality between the two experiments could also account for the diVerence in the eVectiveness of DCS. Therefore, the aim of Experiment 3 was to demonstrate within-session extinction after odor–shock pairings by giving fewer odor–shock pairings than in Experiment 1, and also by reducing the duration of the US.
479
3.3. Experiment 3 As noted in Section 2, rats in this experiment were divided into two groups: “non-extinguishers” and “extinguishers”. Overall, levels of freezing did not signiWcantly decline across the extinction session for rats in the “nonextinguishers” condition, as conWrmed by a mixed-design ANOVA with 6 time points, F < 1 (Fig. 2a). In addition, rats allocated to the saline condition (n D 6) displayed similar levels of freezing to rats that were to receive DCS (n D 6) [non-signiWcant interaction F < 1, and non-signiWcant between groups eVect, F (1, 10) D 1.11, p D 0.32]. At test, there were no group diVerences in the amount of pre-CS freezing [t (10) <1], and no group diVerences in the amount of freezing to the odor CS [t (10), <1] (Fig. 2b). Conversely, rats in the “extinguishers” condition displayed high levels of conditioned freezing to the odor during the Wrst extinction trial and signiWcantly lower levels of freezing across trials [mixed-design ANOVA with 6 time points, F (5,50) D 2.65, p D 0.03] (Fig. 2c). Rats that were to receive saline (n D 6) did not diVer on the amount of freezing during each extinction trial to rats that were to receive DCS (n D 6) [non-signiWcant interaction and non-signiWcant between groups eVect, both Fs < 1]. At test, there were no group diVerences in the amount of pre-CS freezing [t (10) D 2.09, p D 0.08], however, during the 3 min odor CS presentation, rats that received DCS froze signiWcantly less than rats that received saline [t (10) D 2.31, p D 0.04] (Fig. 2d). In this experiment DCS facilitated extinction in rats that displayed within-session extinction (“extinguishers”) but not in rats that maintained high levels of freezing before the DCS injection (“non-extinguishers”). Experiment 4 aimed to further explore the idea that within-session extinction is necessary in order for DCS to facilitate long-term extinction. 3.4. Experiment 4 One rat in the DCS group displayed a high level of freezing during the pre-CS period at test (>50%) and was excluded from all statistical analysis. By the end of the second extinction training session, and before the drug injection, all rats displayed signiWcant levels of extinction [mixed-design ANOVA with 12 time points, F (11, 143) D 13.62, p D .005] (Fig. 3a). Rats that were to receive saline (n D 8) did not diVer on the amount of freezing during each extinction trial to rats that were to receive DCS (n D 7) [non-signiWcant interaction F (11, 143) D 1.03, p D .42, and non-signiWcant between groups eVect, F < 1]. At test, the two groups did not diVer in the level of pre-CS freezing [t (13) <1], however during the 3 min odor presentation, rats that received DCS froze signiWcantly less than rats that received saline [t (13) D 2.43, p D .03] (Fig. 3b). 4. Discussion This study demonstrated that post-extinction, systemic injections of DCS facilitated extinction retention with an
480
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
100
b
75
100 75
% Freezing
% Freezing
a
50 25
50 25
DCS SAL
0
0 0
1
2
3
4
5
6
DCS
Trial
d
100 75
100 75
% Freezing
% Freezing
c
50 25
SAL
Condition
50 25
DCS SAL
0
0 0
1
2
3
4
5
6
DCS
Trial
SAL
Condition
Fig. 2. (a) Mean freezing during extinction trials in Experiment 3 for rats in the “non-extinguishers” condition. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (b) Mean freezing of rats in the “non-extinguishers” condition during test in Experiment 3. (c) Mean freezing during extinction trials in Experiment 3 for rats in the “extinguishers” condition. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (d) Mean freezing of rats in the “extinguishers” condition during test in Experiment 3.
a
Session 1
b
Session 2
% Freezing
100
100
75
75
50
50
25
25
DCS SAL
0
1
2
3
4
5
6
1
2
3
Trial
4
5
6
0 DCS
SAL
Condition
Fig. 3. (a) Mean freezing during extinction trials in Experiment 4 across 12 trials that were separated into two groups of 6 by 45–90 min. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (b) Mean freezing at test in Experiment 4.
aversive odor CS only if short-term, or within-session, extinction of conditioned freezing had been observed prior to injection. This occurred whether short-term extinction was observed following conditioning with a noise US (Experiment 2), or fewer odor–shock pairings (Experiment 3—“extinguishers”), or when additional extinction training was given after odor–shock pairings (Experiment 4). When high levels of freezing were maintained throughout the extinction session, DCS-treated rats displayed the same level of extinction retention as did saline-treated rats (as shown in Experiment 1 and the “non-extinguishers” in Experiment 3). Together, these experiments demonstrate that (1) DCS can be used eVectively to facilitate extinction with odor cues, and (2) DCS is ineVective at facilitating extinction retention if no short-term extinction learning occurs prior to injection. The dissociation between the eVects of DCS on extinction learning in the experiments in this study and the
hypothesized diVerences between the amounts of short- and long-term learning across extinction training provide an interesting perspective from which to speculate on the mechanisms underlying the facilitatory eVects of DCS on extinction retention. DCS exerts its eVects in the brain by binding to the glycine regulatory site of the NMDA complex. When the glycine site on the NMDA complex is activated, cell depolarization is ampliWed by an increased frequency of channel openings, and therefore excitatory neurotransmission through the NMDA receptor is facilitated (Kleckner & Dingledine, 1988). Normal activation of NMDA receptors has been shown to be critical for the neural plasticity underlying learning (e.g., long-term potentiation; Bear & Malenka, 1994) and is, more speciWcally, one mechanism by which experiences are translated from shortto long-term memory (Sweatt, 1999). Given that DCS augments neurotransmission through NMDA receptors, and working under the assumption that long-term memories
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
cannot be formed without short-term learning, it is possible that DCS facilitates extinction retention by increasing the amount of short-term extinction learning that is translated into long-term memory. For example, in Experiments 2, 3 (“extinguishers”), and 4, there was a signiWcant amount of short-term extinction exhibited across the extinction trials and the translation of that learning into long-term memory was increased in rats that received DCS. The saline-treated rats that received extinction training in these experiments displayed very little extinction retention at test (indeed in Experiment 1, these rats responded similarly to the noextinction control groups) and therefore either did not form a long-term extinction memory, or the memory was not strong enough to reach threshold for behavioral expression. However, in the case where no short-term extinction learning occurs, regardless of whether NMDA receptors are activated or not, no translation into long-term memory can occur. For example, there was no apparent short-term extinction learning in Experiments 1 and 3 (“non-extinguishers”) because rats displayed a consistently high level of conditioned freezing across all six extinction trials. In this case, the absence of short-term learning meant there could be no translation from short- to long-term memory and therefore DCS could not facilitate this transfer. Hence, at test there were no group diVerences. However, another interpretation of our results can be derived from the notion that rather than having a “quantitative” eVect on memory (i.e., the amount transferred), DCS has a “qualitative” eVect on memory. Rescorla (2004) noted that what is learnt during an extinction session is less stable with time than what is learnt during initial acquisition. The Wnding that extinguished responses often ‘spontaneously recover’ after a passage of time following non-reinforcement is a clear demonstration of this idea and can be used to explain the current data. In other words, rather than facilitating the neural process underlying the translation of extinction learning from short- to long-term memory, such that more is learnt, the results of the current study could suggest that DCS fortiWes any learning that does occur during extinction training and renders it more stable over time. In this way, the eVects of DCS on extinction could be interpreted as facilitating extinction retention by impairing spontaneous recovery. SpeciWcally, in Experiments 1 and 3 (“non-extinguishers”) rats had not shown a signiWcant decrease in freezing by the end of the extinction session. Thus, it was impossible to observe any spontaneous recovery of the original learning because no inhibition of the original response was observed. It follows then that it was impossible to observe any impairment of spontaneous recovery by DCS. Therefore, the following day, rats that had received an injection of DCS displayed a similar level of freezing to the level they exhibited on the Wnal trial of extinction training, and similar levels of freezing to the saline group. However, in the experiments where rats exhibited a signiWcant decrease in freezing by the Wnal extinction trial, the originally learned responses to the CS had been inhibited (to diVerent degrees, depending on the condition-
481
ing and extinction parameters). The following day, salinetreated rats displayed spontaneous recovery of conditioned freezing and DCS-treated rats displayed an impairment in spontaneous recovery. Indeed, in Experiments 3 (“extinguishers”) and 4, the level of freezing exhibited at test by DCS-treated rats matched their level of freezing during the last extinction trial (see Figs. 2c and d, and 3a and b). From this perspective, DCS may facilitate extinction retention by strengthening the long-term memory such that it is not susceptible to degradation during the retention interval. It will be diYcult to design experiments to distinguish between these two interpretations. Whichever account is eventually supported, however, the present results suggest that in order for DCS to have a facilitatory eVect on extinction there needs to be at least some within-session extinction. The current study demonstrated that in an animal preparation, fear responses elicited by an odor can be extinguished, and further, that DCS facilitated the long-term extinction of those responses. Most clinical and preclinical studies to date have investigated the eVects of DCS on extinction using visual cues. However, in both humans and animals stimuli from other sensory senses can be equally, if not more, eVective at retrieving traumatic memories. Odors, in particular, can be powerful for triggering the emotional component of memory. For example, Herz (1998) demonstrated that odors were equally eVective at cuing recall for emotionally arousing pictures as tactile, visual, or auditory cues, however the emotional ‘potency’ of the memories was signiWcantly greater for odor-evoked memories. In a clinical setting, the emotional quality of memories is clearly an important factor underlying the maintenance of anxiety disorders. Our Wndings suggest that DCS can facilitate extinction of learned fear to odor cues if there is some within-session extinction to those cues during the extinction training session. It would be of interest whether these Wndings are limited to odor cues, and indeed, to animal preparations. Previous experiments investigating the eVects of DCS on fear extinction in human studies have not reported any data on within-session loss of fear, but from the perspective of the present results, this may be a critical factor in determining whether DCS is eVective at enhancing the long-term loss of fear in clinical samples. Acknowledgment This research was supported by a grant from the National Health and Medical Research Council (#300538). References Andrews, G., Grino, R., Hunt, C., & Page, A. (1994). The Treatment of Anxiety Disorders: Clinican’s Guide and Patients Manual. Cambridge University Press. The Australian Code of Practice for the Care and Use of Animals for ScientiWc Purposes. (7th ed.) 2004. Canberra: Australian Government Publishing Service. Bear, M. F., & Malenka, R. C. (1994). Synaptic plasticity: LTP and LTD. Current Opinion in Neurobiology, 4(3), 389–399.
482
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
Davis, M., Ressler, K. J., Rothbaum, B. O., & Richardson, R. (2006). EVects of D-cycloserine on extinction: translation from preclinical to clinical work. Biological Psychiatry, 60, 369–375. Fanselow, M. S. (1980). Conditioned and unconditional components of post-shock freezing. Pavlovian Journal of Biological Science, 15(4), 177–182. Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: exposure to corrective information. Psychological Bulletin, 99(1), 20–35. Herz, R. S. (1998). Are odors the best cues to memory? A cross-modal comparison of associative memory stimuli. Annals of the New York Academy of Sciences, 855, 670–674. Herz, R. S., Eliassen, J., Beland, S., & Souza, T. (2004). Neuroimaging evidence for the emotional potency of odor-evoked memory. Neuropsychologia, 42(3), 371–378. Hofmann, S. G., Meuret, A. E., Smits, J. A., Simon, N. M., Pollack, M. H., Eisenmenger, K., et al. (2006). Augmentation of exposure therapy with D-cycloserine for social anxiety disorder. Archives of General Psychiatry, 63(3), 298–304. Kleckner, N. W., & Dingledine, R. (1988). Requirement for glycine in activation of NMDA-receptors expressed in Xenopus oocytes. Science, 241(4867), 835–837. Ledgerwood, L., Richardson, R., & Cranney, J. (2003). EVects of D-cycloserine on extinction of conditioned freezing. Behavioral Neuroscience, 117(2), 341–349. Ledgerwood, L., Richardson, R., & Cranney, J. (2004). D-cycloserine and the facilitation of extinction of conditioned fear: consequences for reinstatement. Behavioral Neuroscience, 118(3), 505–513. Ledgerwood, L., Richardson, R., & Cranney, J. (2005). D-cycloserine facilitates extinction of learned fear: eVects on reacquisition and generalized extinction. Biological Psychiatry, 57(8), 841–847. Parnas, A. S., Weber, M., & Richardson, R. (2005). EVects of multiple exposures to D-cycloserine on extinction of conditioned fear in rats. Neurobiology of Learning and Memory, 83(3), 224–231.
Paschall, G. Y., & Davis, M. (2002). Olfactory-mediated fear-potentiated startle. Behavioral Neuroscience, 116(1), 4–12. Price, J. L. (1973). An autoradiographic study of complementary laminar patterns of termination of aVerent Wbers to the olfactory cortex. Journal of Comparative Neurology, 150(1), 87–108. Rescorla, R. A. (2004). Spontaneous recovery. Learning and Memory, 11(5), 501–509. Ressler, K. J., Rothbaum, B. O., Tannenbaum, L., Anderson, P., Graap, K., Zimand, E., et al. (2004). Cognitive enhancers as adjuncts to psychotherapy: use of D-cycloserine in phobic individuals to facilitate extinction of fear. Archives of General Psychiatry, 61(11), 1136–1144. Richardson, R., Tronson, N., Bailey, G. K., & Parnas, A. S. (2002). Extinction of conditioned odor potentiation of startle. Neurobiology of Learning and Memory, 78, 426–440. Sweatt, J. D. (1999). Toward a molecular explanation for long-term potentiation. Learning and Memory, 6(5), 399–416. Thyer, B. A., Baum, M., & Reid, L. D. (1988). Exposure techniques in the reduction of fear: a comparative review of the procedure in animals and humans. Advances in Behavior Research and Therapy, 10, 105–127. Vermetten, E., & Bremner, J. D. (2003). Olfaction as a traumatic reminder in posttraumatic stress disorder: case reports and review. Journal of Clinical Psychiatry, 64(2), 202–207. Walker, D. L., Ressler, K. J., Lu, K. T., & Davis, M. (2002). Facilitation of conditioned fear extinction by systemic administration or intra-amygdala infusions of D-cycloserine as assessed with fear-potentiated startle in rats. Journal of Neuroscience, 22(6), 2343–2351. Yang, Y. L., & Lu, K. T. (2005). Facilitation of conditioned fear extinction by D-cycloserine is mediated by mitogen-activated protein kinase and phosphatidylinositol 3-kinase cascades and requires de novo protein synthesis in basolateral nucleus of amygdala. Neuroscience, 134(1), 247–260. Zarate, R., & Agras, W. S. (1994). Psychosocial treatment of phobia and panic disorders. Psychiatry, 57(2), 133–141.
EVects of D-cycloserine on extinction of learned fear to an olfactory cue Marianne Weber, Joe Hart, Rick Richardson ¤ School of Psychology, University of New South Wales, Sydney 2052, Australia Received 11 July 2006; revised 28 November 2006; accepted 21 December 2006 Available online 1 February 2007
Abstract D-cycloserine (DCS), a partial NMDA receptor agonist, facilitates extinction of learned fear in rats and has been used to treat anxiety disorders in clinical populations. However, research into the eVects of DCS on extinction is still in its infancy, with visual cues being the primary fear-eliciting stimuli under investigation. In both human and animal subjects odors have been found to associate strongly with aversive events. Therefore, this study examined the generality of the eVects of DCS on extinction by testing odor cues. Sprague–Dawley rats were conditioned and extinguished to an odor using varying parameters, injected with either saline or DCS (15 mg/kg) following extinction, and then tested for a freezing response 24 h later. Experiment 1 demonstrated that after 3 odor-shock pairings, rats did not display short-term extinction and DCS had no eVect on long-term extinction. Experiment 2 demonstrated that after 3 odor-noise pairings, rats displayed signiWcant short-term extinction and DCS signiWcantly facilitated long-term extinction. Following 2 odor-shock pairings in Experiment 3, half the rats displayed short-term extinction (“extinguishers”) and half did not (“non-extinguishers”). DCS facilitated longterm extinction in the “extinguishers” condition but not in the “non-extinguishers” condition. In Experiment 4, following 2 odor-shock pairings and an extra extinction session, DCS had a signiWcant facilitatory eVect on long-term extinction. Thus, extinction of freezing to an odor cue was facilitated by systemic injections of DCS, but only when some amount of within-session extinction occurred prior to injection. © 2006 Elsevier Inc. All rights reserved.
Keywords: Rats; D-Cycloserine; Fear extinction; Odors; Freezing
1. Introduction Anxiety disorders such as Phobia, post-traumatic stress disorder (PTSD), and Obsessive–Compulsive Disorder are often treated with exposure-based cognitive behavior therapy (Andrews, Grino, Hunt, & Page, 1994; Foa & Kozak, 1986; Thyer, Baum, & Reid, 1988; Zarate & Agras, 1994). This type of therapy involves exposing patients to a fearful, or anxiety-provoking, stimulus within a calm and safe environment. Gradually, over successive presentations of the stimulus within this environment, patients learn to respond without fear or anxiety. This type of treatment is procedurally very similar to ‘extinction’ training in animal models of emotional learning. For example, learned fear responses to
*
Corresponding author. Fax: +61 2 9385 3641. E-mail address: [email protected] (R. Richardson).
1074-7427/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.nlm.2006.12.010
a conditioned stimulus (CS; e.g., a light that had previously been paired with shock) can be reduced by repeatedly presenting the CS in the absence of shock. Eventually the animal learns that the CS no longer predicts the occurrence of shock and fear reactions decrease in amplitude and/or frequency. Although very simple procedurally, extinction has proven to be an extremely diYcult process to understand. Nonetheless, considerable eVort has been directed at trying to identify techniques that render extinction—and potentially exposure therapy—more eVective. One signiWcant advance in this area has been the demonstration that D-cycloserine (DCS) can facilitate fear extinction. DCS was used because of its indirect agonist properties at N-methylD-aspartate receptors, which are critical for the neuronal plasticity required for emotional memory (see, Davis, Ressler, Rothbaum, & Richardson, 2006). Pre-clinical studies have demonstrated that DCS is a potent facilitator of fear
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
extinction in rats (Ledgerwood, Richardson, & Cranney, 2003, 2004, 2005; Parnas, Weber, & Richardson, 2005; Walker, Ressler, Lu, & Davis, 2002; Yang & Lu, 2005). In addition, DCS has been found to enhance the eVectiveness of exposure-based therapy in people treated for acrophobia (Ressler et al., 2004) and Social Anxiety Disorder (Hofmann et al., 2006). However, research on the eVects of DCS on extinction is still in its infancy, and it is still not known whether DCS is eVective at facilitating extinction to any, or all, types of feared stimuli, or whether its eVectiveness is limited in some way. The aim of the current study was to test the ‘generality’ of the eVects of DCS on extinction. Both pre-clinically and clinically, DCS has primarily been used in conjunction with exposure to visual cues. However, fear-eliciting cues are not always visually-based. Odors, in particular, have been found to be especially eVective in cuing emotional memory in nonclinical populations (Herz, 1998; Herz, Eliassen, Beland, & Souza, 2004), and can evoke strong trauma-related memories in people with PTSD (Vermetten & Bremner, 2003). Olfactory stimuli are also extremely salient and powerful cues for memory retrieval in rats. For example, conditioned fear associations using olfactory cues in rats are encoded more rapidly (Paschall & Davis, 2002), and require more trials to extinguish than do visual cues (Richardson, Tronson, Bailey, & Parnas, 2002). These Wndings suggest that odor cues may be processed by a diVerent memory system. Indeed, in rats the neuroanatomical pathways subserving olfactory information are slightly diVerent than for stimuli of other sensory modalities. For example, visual and auditory sensations are processed within the thalamus before being sent to the amygdala, which is the brain region crucial for recalling emotional associations. However, olfactory information is transmitted directly to the amygdala (Price, 1973), suggesting that odors have ‘privileged’ access to the amygdala and may prompt recall for emotional associations more quickly and easily. Therefore, because (1) odor-evoked memories can be powerful triggers for the emotional component of memory, and (2) the neuroanatomical substrates of odor-evoked emotional memories may diVer to that for visually-evoked memories, the present study was designed to determine whether DCS facilitates extinction of conditioned fear to an odor CS in rats. 2. Methods and materials 2.1. Subjects Experimentally-naive, male Sprague–Dawley rats (450–700 g) obtained from the breeding colony maintained by the School of Psychology at the University of New South Wales were used. The rats were housed in groups of eight in plastic boxes (67 cm long £ 40 cm wide £ 22 cm high) in a colony room with a natural light cycle. Food and water were continuously available. All experimental procedures were approved by the Animal Care and Ethics Committee of the University of New South Wales and adhered to the ethical guidelines described in The Australian Code of Practice for the Care and Use of Animals for ScientiWc Purposes (2004). Prior to experimentation, all rats were handled for approximately 5 min on 3 consecutive days.
477
2.2. Apparatus In all experiments conditioning occurred in one context and extinction training and testing occurred in a diVerent context. All experimental cages were located within a sound- and light-attenuating wood cabinet where a ventilation fan provided a 60-dB ambient noise level in each chamber. Conditioning occurred in one of 4 identical cages (20 cm long £ 12 cm wide £ 12 cm high). The front wall, rear wall, and ceiling of each cage were made of Plexiglas, and the side walls and Xoor consisted of stainless steel bars, with 1.3 cm between each bar. Each cage was suspended 13 cm above the Xoor of the wood cabinet and a tray of bedding was placed under each cage, which was changed between each rat. During conditioning, the doors to the wood cabinets remained open and the room was illuminated by a red light. The context used for extinction training and testing in Experiments 1 and 2 diVered slightly from that used in Experiments 3 and 4. In Experiments 1 and 2, extinction training and testing occurred in one of 2 identical cages (30 cm long £ 22.5 cm wide £ 30 cm high) with a clear Perspex front, black and white striped side and rear walls, and a wood hinged lid. The Xoor consisted of stainless steel bars, with 1.0 cm between each bar, and was suspended 8.5 cm above the Xoor of the wood cabinet. In Experiments 3 and 4, extinction training and testing occurred in one of 2 identical clear Perspex cages (29 cm long £ 28 cm wide £ 29 cm high). The Xoor consisted of stainless steel bars, with 1.5 cm between each bar, and was suspended 10 cm above the Xoor of the wood cabinets. A white light bulb in each cabinet provided illumination at all times during extinction and testing. Between sessions, these chambers were wiped with 0.5% eucalyptus solution and the bedding below the grid Xoor was changed. Extinction and test sessions were recorded for later assessment.
2.3. Procedure In all experiments the CS was a grape odor (0.1 ml in a plastic jar; #182380019 from Wild Flavours, Heidelberg) and the method of presentation involved the odor jar being placed under the experimental cage for the required time period and then being removed and capped with an air-tight lid. The unconditioned stimulus (US) varied for each experiment. The procedures for Experiments 1 and 2 were identical except that a diVerent US was used. In Experiment 1 the US was a 1 s, 0.6 mA footshock and in Experiment 2 the US was a 100 ms, 120 dB white noise stimulus (rise-fall time <5 ms). In each experiment a 2 £ 2 factorial design was employed, where the Wrst factor was drug condition (DCS or saline) and the second factor was condition (extinction or no extinction). This produced 4 groups: extinction + DCS (E-DCS), extinction + saline (E-Sal), no-extinction + DCS, and no-extinction + saline; the latter two groups were subsequently pooled into a single group (NE-Pooled). On Day 1, all rats were pre-exposed to the conditioning context for 20 min. Three hours later they were returned to the conditioning cages and received 3 pairings of the CS and US. That is, 120 s after being placed in the conditioning cages, the odor-jar was placed beneath the cage. Ten seconds later, the US was presented and then the odor-jar was removed. The 3 CS–US pairings occurred 120 s apart and rats were removed from the conditioning cages 120 s after the last pairing. On Day 2, rats allocated to the extinction groups were placed in the extinction cages. After 2 min in the cage, rats were presented with the jar containing the odor 6 times (each exposure was 2-min long, with a 2-min ITI). The rats were removed from the extinction cages 120 s after the last extinction trial, and half were injected with saline (E-Sal) and half were injected with DCS (E-DCS). Rats that did not receive extinction training were removed from their home cage and injected with either saline or DCS (NE-pooled). After injection all rats were returned to their home cage. On Day 3, all rats were tested for odor-elicited freezing. That is, 2 min after being placed in the cage, the odor was presented for 3 min. The procedure for Experiments 3 and 4 was identical to Experiments 1 and 2 except that rats received 2 CS–US pairings and the US was a 0.5 s, 0.6 mA footshock. After extinction training in Experiment 3 rats were
478
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
divided into 2 groups where half received an injection of DCS and half received saline before being returned to the home cage. All rats were tested for odor-elicited freezing the following day. For the purpose of analysis, these groups were further divided equally into ‘extinguishers’ and ‘nonextinguishers’. Those rats that had the lowest levels of freezing at the end of extinction were put into the “extinguishers” condition while those with the highest levels of freezing at the end of extinction were placed in the “non-extinguishers” condition. In Experiment 4, rats received a second extinction training session that occurred 45–90 min after the Wrst. After this second extinction session, half the rats received DCS and half received saline. All rats were tested for odor-elicited freezing the following day.
2.4. Data analysis Each rat was scored for freezing during the 6 extinction trials (or the 12 trials in Experiment 4) and during the test on Day 3. Freezing was characterized by an absence of all bodily movements except those required for breathing (Fanselow, 1980), and was measured using a time sampling method. An observation was made every 5 s and a positive or negative count was made. A percentage score was then calculated for the proportion of the total observation period spent freezing. For the test session on Day 3, freezing during the 120 s prior to the Wrst odor presentation (pre-CS) was scored as an indication of baseline levels of fear. Analysis of variance (ANOVA) was the primary statistical approach, and post hoc comparisons were made with Tukey’s honestly signiWcant diVerence (HSD) test.
2.5. Drugs D-cycloserine (Sigma–Aldrich, Castle Hill, New South Wales) was freshly dissolved in sterile saline at a concentration of 15 mg/ml and injected subcutaneously in a volume of 1 ml/kg.
100
3.1. Experiment 1 Rats that received extinction training displayed high levels of conditioned freezing to each of the 6 odor presentations during the extinction session (Fig. 1a). Overall, levels of freezing did not signiWcantly decline across the extinction session, as conWrmed by a mixed-design ANOVA with 6 time points, F < 1. In addition, rats allocated to the saline condition (n D 8) displayed similar levels of freezing to rats that were to receive DCS (n D 8). At test, rats that received DCS but no extinction training did not display a diVerent level of freezing to rats that received saline and no extinction [t (6) <1]. Therefore, these groups were pooled (NE-Pooled). Overall, there were no group diVerences in the amount of pre-CS freezing [F (2, 23) D 2.49, p D .11]. Additionally, there were no group diVerences in the amount of freezing to the odor CS [F (2, 23) D 1.85, p D .18] (Fig. 1b). Thus, rats that received DCS immediately after extinction training to an odor CS previously paired with a shock US did not freeze less at test than rats that received saline, indicating that DCS did not facilitate extinction. However, no decline in the fear response was observed in either extinction group across the six 2-minute extinction trials. It may be the case that DCS does not facilitate extinction retention if no short-term extinction learning
b
75
% Freezing
% Freezing
a
3. Results
50 25
DCS SAL
100 75 50 25 0
0 0
1
2
3
4
5
6
E-DCS
Trial
d
100 75
% Freezing
% Freezing
c
50 25
E-SAL
NE-Pooled
Condition
DCS SAL
100 75 50 25 0
0 0
1
2
3
Trial
4
5
6
E-DCS
E-SAL
NE-Pooled
Condition
Fig. 1. (a) Mean freezing during extinction trials in Experiment 1 where rats had received odor–shock pairings. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (b) Mean freezing during test in Experiment 1. Rats had either been extinguished (E) or not (NE), and injected with saline (SAL) or DCS. Half of the rats in the NE group were injected with saline while the others were injected with DCS; these rats were collapsed into a single control group (pooled). (c) Mean freezing during extinction trials in Experiment 2 where rats had received odor–noise pairings. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (d) Mean freezing during test in Experiment 2. Rats had been extinguished (E) or not (NE), and injected with saline (SAL) or DCS. Half of the rats in the NE group were injected with saline while the others were injected with DCS; these rats were collapsed into a single control group (pooled).
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
is observed. Alternatively, it may be the case that DCS does not aVect extinction of conditioned fear when odor cues are used. The aim of Experiment 2 was to condition rats to an odor cue in a manner in which short-term extinction was observed. That is, rats were conditioned with a US that is less aversive than a shock, namely a loud noise. 3.2. Experiment 2 Two rats did not display a robust conditioned response to the odor (mean levels of freezing were less than 10% across the 6 extinction trials) and were excluded from all statistical analysis. Rats that received extinction training displayed high levels of conditioned freezing to the odor during the Wrst extinction trial and displayed signiWcantly lower levels of freezing across trials [mixed-design ANOVA with 6 time points, F (5, 120) D 39.0, p D .001] (Fig. 1c). Rats that were to receive saline (n D 12) did not diVer on the amount of freezing during each extinction trial to rats that were to receive DCS (n D 14) [non-signiWcant interaction, F < 1 and a non-signiWcant between groups eVect, F (1, 24) D 1.17, p D .29]. At test, rats that received no extinction training and DCS did not display a diVerent level of freezing to rats that received saline and no extinction [t (10) <1]. Therefore, these groups were pooled (NE-Pooled, n D 12). PreCS freezing was signiWcantly higher for rats that received extinction training and saline injections (E-SAL) than rats in the E-DCS and NE-Pooled conditions [F (2, 37) D 4.01, p D .03, with post hoc comparisons signiWcant at p < .05]. Thus, freezing to the CS was analyzed with an ANCOVA, with pre-CS freezing as a covariate. During the 3 min odor CS presentation, rats that received DCS following extinction (E-DCS) froze signiWcantly less than rats that received saline following extinction (E-Sal), and signiWcantly less than rats that were not given extinction training (NE-Pooled; F (2, 37) D 5.32, p D .01, with post hoc comparisons signiWcant at p < .05). Further, rats that received extinction training and saline injections (E-SAL) did not display signiWcantly diVerent levels of freezing to rats that did not receive extinction training (NE-Pooled) (Fig. 1d). Experiment 2 demonstrated facilitated extinction to an odor CS when DCS was injected after an extinction training session in which short-term extinction was observed. This experiment demonstrates that long-term extinction of odor-elicited fear can be facilitated by DCS, and suggests that the failure to observe facilitation of extinction retention in Experiment 1 was due to the high level of freezing maintained during extinction training. However, the diVerence in US modality between the two experiments could also account for the diVerence in the eVectiveness of DCS. Therefore, the aim of Experiment 3 was to demonstrate within-session extinction after odor–shock pairings by giving fewer odor–shock pairings than in Experiment 1, and also by reducing the duration of the US.
479
3.3. Experiment 3 As noted in Section 2, rats in this experiment were divided into two groups: “non-extinguishers” and “extinguishers”. Overall, levels of freezing did not signiWcantly decline across the extinction session for rats in the “nonextinguishers” condition, as conWrmed by a mixed-design ANOVA with 6 time points, F < 1 (Fig. 2a). In addition, rats allocated to the saline condition (n D 6) displayed similar levels of freezing to rats that were to receive DCS (n D 6) [non-signiWcant interaction F < 1, and non-signiWcant between groups eVect, F (1, 10) D 1.11, p D 0.32]. At test, there were no group diVerences in the amount of pre-CS freezing [t (10) <1], and no group diVerences in the amount of freezing to the odor CS [t (10), <1] (Fig. 2b). Conversely, rats in the “extinguishers” condition displayed high levels of conditioned freezing to the odor during the Wrst extinction trial and signiWcantly lower levels of freezing across trials [mixed-design ANOVA with 6 time points, F (5,50) D 2.65, p D 0.03] (Fig. 2c). Rats that were to receive saline (n D 6) did not diVer on the amount of freezing during each extinction trial to rats that were to receive DCS (n D 6) [non-signiWcant interaction and non-signiWcant between groups eVect, both Fs < 1]. At test, there were no group diVerences in the amount of pre-CS freezing [t (10) D 2.09, p D 0.08], however, during the 3 min odor CS presentation, rats that received DCS froze signiWcantly less than rats that received saline [t (10) D 2.31, p D 0.04] (Fig. 2d). In this experiment DCS facilitated extinction in rats that displayed within-session extinction (“extinguishers”) but not in rats that maintained high levels of freezing before the DCS injection (“non-extinguishers”). Experiment 4 aimed to further explore the idea that within-session extinction is necessary in order for DCS to facilitate long-term extinction. 3.4. Experiment 4 One rat in the DCS group displayed a high level of freezing during the pre-CS period at test (>50%) and was excluded from all statistical analysis. By the end of the second extinction training session, and before the drug injection, all rats displayed signiWcant levels of extinction [mixed-design ANOVA with 12 time points, F (11, 143) D 13.62, p D .005] (Fig. 3a). Rats that were to receive saline (n D 8) did not diVer on the amount of freezing during each extinction trial to rats that were to receive DCS (n D 7) [non-signiWcant interaction F (11, 143) D 1.03, p D .42, and non-signiWcant between groups eVect, F < 1]. At test, the two groups did not diVer in the level of pre-CS freezing [t (13) <1], however during the 3 min odor presentation, rats that received DCS froze signiWcantly less than rats that received saline [t (13) D 2.43, p D .03] (Fig. 3b). 4. Discussion This study demonstrated that post-extinction, systemic injections of DCS facilitated extinction retention with an
480
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
100
b
75
100 75
% Freezing
% Freezing
a
50 25
50 25
DCS SAL
0
0 0
1
2
3
4
5
6
DCS
Trial
d
100 75
100 75
% Freezing
% Freezing
c
50 25
SAL
Condition
50 25
DCS SAL
0
0 0
1
2
3
4
5
6
DCS
Trial
SAL
Condition
Fig. 2. (a) Mean freezing during extinction trials in Experiment 3 for rats in the “non-extinguishers” condition. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (b) Mean freezing of rats in the “non-extinguishers” condition during test in Experiment 3. (c) Mean freezing during extinction trials in Experiment 3 for rats in the “extinguishers” condition. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (d) Mean freezing of rats in the “extinguishers” condition during test in Experiment 3.
a
Session 1
b
Session 2
% Freezing
100
100
75
75
50
50
25
25
DCS SAL
0
1
2
3
4
5
6
1
2
3
Trial
4
5
6
0 DCS
SAL
Condition
Fig. 3. (a) Mean freezing during extinction trials in Experiment 4 across 12 trials that were separated into two groups of 6 by 45–90 min. Either saline or DCS (15 mg/kg) was injected immediately after extinction. (b) Mean freezing at test in Experiment 4.
aversive odor CS only if short-term, or within-session, extinction of conditioned freezing had been observed prior to injection. This occurred whether short-term extinction was observed following conditioning with a noise US (Experiment 2), or fewer odor–shock pairings (Experiment 3—“extinguishers”), or when additional extinction training was given after odor–shock pairings (Experiment 4). When high levels of freezing were maintained throughout the extinction session, DCS-treated rats displayed the same level of extinction retention as did saline-treated rats (as shown in Experiment 1 and the “non-extinguishers” in Experiment 3). Together, these experiments demonstrate that (1) DCS can be used eVectively to facilitate extinction with odor cues, and (2) DCS is ineVective at facilitating extinction retention if no short-term extinction learning occurs prior to injection. The dissociation between the eVects of DCS on extinction learning in the experiments in this study and the
hypothesized diVerences between the amounts of short- and long-term learning across extinction training provide an interesting perspective from which to speculate on the mechanisms underlying the facilitatory eVects of DCS on extinction retention. DCS exerts its eVects in the brain by binding to the glycine regulatory site of the NMDA complex. When the glycine site on the NMDA complex is activated, cell depolarization is ampliWed by an increased frequency of channel openings, and therefore excitatory neurotransmission through the NMDA receptor is facilitated (Kleckner & Dingledine, 1988). Normal activation of NMDA receptors has been shown to be critical for the neural plasticity underlying learning (e.g., long-term potentiation; Bear & Malenka, 1994) and is, more speciWcally, one mechanism by which experiences are translated from shortto long-term memory (Sweatt, 1999). Given that DCS augments neurotransmission through NMDA receptors, and working under the assumption that long-term memories
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
cannot be formed without short-term learning, it is possible that DCS facilitates extinction retention by increasing the amount of short-term extinction learning that is translated into long-term memory. For example, in Experiments 2, 3 (“extinguishers”), and 4, there was a signiWcant amount of short-term extinction exhibited across the extinction trials and the translation of that learning into long-term memory was increased in rats that received DCS. The saline-treated rats that received extinction training in these experiments displayed very little extinction retention at test (indeed in Experiment 1, these rats responded similarly to the noextinction control groups) and therefore either did not form a long-term extinction memory, or the memory was not strong enough to reach threshold for behavioral expression. However, in the case where no short-term extinction learning occurs, regardless of whether NMDA receptors are activated or not, no translation into long-term memory can occur. For example, there was no apparent short-term extinction learning in Experiments 1 and 3 (“non-extinguishers”) because rats displayed a consistently high level of conditioned freezing across all six extinction trials. In this case, the absence of short-term learning meant there could be no translation from short- to long-term memory and therefore DCS could not facilitate this transfer. Hence, at test there were no group diVerences. However, another interpretation of our results can be derived from the notion that rather than having a “quantitative” eVect on memory (i.e., the amount transferred), DCS has a “qualitative” eVect on memory. Rescorla (2004) noted that what is learnt during an extinction session is less stable with time than what is learnt during initial acquisition. The Wnding that extinguished responses often ‘spontaneously recover’ after a passage of time following non-reinforcement is a clear demonstration of this idea and can be used to explain the current data. In other words, rather than facilitating the neural process underlying the translation of extinction learning from short- to long-term memory, such that more is learnt, the results of the current study could suggest that DCS fortiWes any learning that does occur during extinction training and renders it more stable over time. In this way, the eVects of DCS on extinction could be interpreted as facilitating extinction retention by impairing spontaneous recovery. SpeciWcally, in Experiments 1 and 3 (“non-extinguishers”) rats had not shown a signiWcant decrease in freezing by the end of the extinction session. Thus, it was impossible to observe any spontaneous recovery of the original learning because no inhibition of the original response was observed. It follows then that it was impossible to observe any impairment of spontaneous recovery by DCS. Therefore, the following day, rats that had received an injection of DCS displayed a similar level of freezing to the level they exhibited on the Wnal trial of extinction training, and similar levels of freezing to the saline group. However, in the experiments where rats exhibited a signiWcant decrease in freezing by the Wnal extinction trial, the originally learned responses to the CS had been inhibited (to diVerent degrees, depending on the condition-
481
ing and extinction parameters). The following day, salinetreated rats displayed spontaneous recovery of conditioned freezing and DCS-treated rats displayed an impairment in spontaneous recovery. Indeed, in Experiments 3 (“extinguishers”) and 4, the level of freezing exhibited at test by DCS-treated rats matched their level of freezing during the last extinction trial (see Figs. 2c and d, and 3a and b). From this perspective, DCS may facilitate extinction retention by strengthening the long-term memory such that it is not susceptible to degradation during the retention interval. It will be diYcult to design experiments to distinguish between these two interpretations. Whichever account is eventually supported, however, the present results suggest that in order for DCS to have a facilitatory eVect on extinction there needs to be at least some within-session extinction. The current study demonstrated that in an animal preparation, fear responses elicited by an odor can be extinguished, and further, that DCS facilitated the long-term extinction of those responses. Most clinical and preclinical studies to date have investigated the eVects of DCS on extinction using visual cues. However, in both humans and animals stimuli from other sensory senses can be equally, if not more, eVective at retrieving traumatic memories. Odors, in particular, can be powerful for triggering the emotional component of memory. For example, Herz (1998) demonstrated that odors were equally eVective at cuing recall for emotionally arousing pictures as tactile, visual, or auditory cues, however the emotional ‘potency’ of the memories was signiWcantly greater for odor-evoked memories. In a clinical setting, the emotional quality of memories is clearly an important factor underlying the maintenance of anxiety disorders. Our Wndings suggest that DCS can facilitate extinction of learned fear to odor cues if there is some within-session extinction to those cues during the extinction training session. It would be of interest whether these Wndings are limited to odor cues, and indeed, to animal preparations. Previous experiments investigating the eVects of DCS on fear extinction in human studies have not reported any data on within-session loss of fear, but from the perspective of the present results, this may be a critical factor in determining whether DCS is eVective at enhancing the long-term loss of fear in clinical samples. Acknowledgment This research was supported by a grant from the National Health and Medical Research Council (#300538). References Andrews, G., Grino, R., Hunt, C., & Page, A. (1994). The Treatment of Anxiety Disorders: Clinican’s Guide and Patients Manual. Cambridge University Press. The Australian Code of Practice for the Care and Use of Animals for ScientiWc Purposes. (7th ed.) 2004. Canberra: Australian Government Publishing Service. Bear, M. F., & Malenka, R. C. (1994). Synaptic plasticity: LTP and LTD. Current Opinion in Neurobiology, 4(3), 389–399.
482
M. Weber et al. / Neurobiology of Learning and Memory 87 (2007) 476–482
Davis, M., Ressler, K. J., Rothbaum, B. O., & Richardson, R. (2006). EVects of D-cycloserine on extinction: translation from preclinical to clinical work. Biological Psychiatry, 60, 369–375. Fanselow, M. S. (1980). Conditioned and unconditional components of post-shock freezing. Pavlovian Journal of Biological Science, 15(4), 177–182. Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: exposure to corrective information. Psychological Bulletin, 99(1), 20–35. Herz, R. S. (1998). Are odors the best cues to memory? A cross-modal comparison of associative memory stimuli. Annals of the New York Academy of Sciences, 855, 670–674. Herz, R. S., Eliassen, J., Beland, S., & Souza, T. (2004). Neuroimaging evidence for the emotional potency of odor-evoked memory. Neuropsychologia, 42(3), 371–378. Hofmann, S. G., Meuret, A. E., Smits, J. A., Simon, N. M., Pollack, M. H., Eisenmenger, K., et al. (2006). Augmentation of exposure therapy with D-cycloserine for social anxiety disorder. Archives of General Psychiatry, 63(3), 298–304. Kleckner, N. W., & Dingledine, R. (1988). Requirement for glycine in activation of NMDA-receptors expressed in Xenopus oocytes. Science, 241(4867), 835–837. Ledgerwood, L., Richardson, R., & Cranney, J. (2003). EVects of D-cycloserine on extinction of conditioned freezing. Behavioral Neuroscience, 117(2), 341–349. Ledgerwood, L., Richardson, R., & Cranney, J. (2004). D-cycloserine and the facilitation of extinction of conditioned fear: consequences for reinstatement. Behavioral Neuroscience, 118(3), 505–513. Ledgerwood, L., Richardson, R., & Cranney, J. (2005). D-cycloserine facilitates extinction of learned fear: eVects on reacquisition and generalized extinction. Biological Psychiatry, 57(8), 841–847. Parnas, A. S., Weber, M., & Richardson, R. (2005). EVects of multiple exposures to D-cycloserine on extinction of conditioned fear in rats. Neurobiology of Learning and Memory, 83(3), 224–231.
Paschall, G. Y., & Davis, M. (2002). Olfactory-mediated fear-potentiated startle. Behavioral Neuroscience, 116(1), 4–12. Price, J. L. (1973). An autoradiographic study of complementary laminar patterns of termination of aVerent Wbers to the olfactory cortex. Journal of Comparative Neurology, 150(1), 87–108. Rescorla, R. A. (2004). Spontaneous recovery. Learning and Memory, 11(5), 501–509. Ressler, K. J., Rothbaum, B. O., Tannenbaum, L., Anderson, P., Graap, K., Zimand, E., et al. (2004). Cognitive enhancers as adjuncts to psychotherapy: use of D-cycloserine in phobic individuals to facilitate extinction of fear. Archives of General Psychiatry, 61(11), 1136–1144. Richardson, R., Tronson, N., Bailey, G. K., & Parnas, A. S. (2002). Extinction of conditioned odor potentiation of startle. Neurobiology of Learning and Memory, 78, 426–440. Sweatt, J. D. (1999). Toward a molecular explanation for long-term potentiation. Learning and Memory, 6(5), 399–416. Thyer, B. A., Baum, M., & Reid, L. D. (1988). Exposure techniques in the reduction of fear: a comparative review of the procedure in animals and humans. Advances in Behavior Research and Therapy, 10, 105–127. Vermetten, E., & Bremner, J. D. (2003). Olfaction as a traumatic reminder in posttraumatic stress disorder: case reports and review. Journal of Clinical Psychiatry, 64(2), 202–207. Walker, D. L., Ressler, K. J., Lu, K. T., & Davis, M. (2002). Facilitation of conditioned fear extinction by systemic administration or intra-amygdala infusions of D-cycloserine as assessed with fear-potentiated startle in rats. Journal of Neuroscience, 22(6), 2343–2351. Yang, Y. L., & Lu, K. T. (2005). Facilitation of conditioned fear extinction by D-cycloserine is mediated by mitogen-activated protein kinase and phosphatidylinositol 3-kinase cascades and requires de novo protein synthesis in basolateral nucleus of amygdala. Neuroscience, 134(1), 247–260. Zarate, R., & Agras, W. S. (1994). Psychosocial treatment of phobia and panic disorders. Psychiatry, 57(2), 133–141.