The author(s) have made the following declarations about their contributions: Conceived and designed the experiments: KJ AAF. Performed the experiments: KJ BBL EK KMM. Analyzed the data: KJ BBL EK AAF. Contributed reagents/materials/analysis tools: AAF. Wrote the paper: KJ EK AAF. Supervised and consulted on experiments: BSM AAF.
Current address: Center for Neural Science, New York University, New York, New York, United States of America
The authors have declared that no competing interests exist.
An intensely stressful experience can itself activate memories that are unrelated to the stressful experience. This previously unknown property of stress could help explain how traumatic memories become pathological.
Inappropriate recollections and responses in stressful conditions are hallmarks of post-traumatic stress disorder and other anxiety and mood disorders, but how stress contributes to the disorders is unclear. Here we show that stress itself reactivates memories even if the memory is unrelated to the stressful experience. Forced-swim stress one day after learning enhanced memory recall. One-day post-learning amnestic treatments were ineffective unless administered soon after the swim, indicating that a stressful experience itself can reactivate unrelated consolidated memories. The swim also triggered inter-hemispheric transfer of a lateralized memory, confirming stress reactivates stable memories. These novel effects of stress on memory required the hippocampus although the memories themselves did not, indicating hippocampus-dependent modulation of extrahippocampal memories. These findings that a stressful experience itself can activate memory suggest the novel hypothesis that traumatic stress reactivates pre-trauma memories, linking them to memory for the trauma and pathological facilitation of post-traumatic recall.
This work identifies a powerful effect of stressful experiences on memories. We report that a single intensely stressful experience can activate memories in a situation that has essentially no physical or motivational relationship to the stressful experience. Using a forced-swim test as a stressor in rats, we find that this treatment was able to activate unrelated memories formed 24 hours earlier. We also find that the hippocampus of the brain is required for this effect of stress but that recall of the memories themselves does not. The ability of stress to activate memories that are unrelated to the stressful event may help to explain how memories can sometimes become pathological and uncontrollable following traumatic events, as in post-traumatic stress disorder. Our findings suggest the novel hypothesis that the stress of the traumatic event activates neutral, unrelated memories, which then become associated with the traumatic event. Subsequent normal recall of the neutral memories can more easily trigger inappropriate recall of the traumatic event, initiating another bout of stress and inappropriate associations of neutral and traumatic memories.
Inappropriate negative responses in emotionally neutral circumstances and the development of negative associations to harmless stimuli are core, debilitating features of post-traumatic stress disorder (PTSD), depression, and a host of anxiety and mood disorders. The possibility that stress itself might promote inappropriate associations between unrelated memories and events has not been explored, although a central role for stress and memory in these disorders is established
On Day 1, food-deprived rats were trained to perform left/right discriminations for food reward on a T-maze. On Day 2, one group of rats was forced to swim in a covered bucket for 20 min. The other group of rats was placed in the same bucket but with only 1 cm deep water so the animal was not forced to swim. On Day 3 all rats completed three non-reinforced T-maze trials to test retention of Day 1 memory. The groups did not differ on Day 1, (
(A) Experiment 1a—Rats trained on Day 1 in the appetitive left/right discrimination were either forced to swim (group Sw1, black,
We investigated the possibility that the learning task itself is stressful, and therefore the stressful forced-swim experience may be serving as a memory activation cue. We repeated Experiment 1a, and at different stages of the protocol, we sacrificed animals to measure their corticosterone levels as an estimate of physiological stress. Relative to cage control levels, serum corticosterone increased over 300% immediately after the swim, which was the only significant increase of all the time points we assayed (
These results indicate the stressful forced swim enhanced the expression of the 24-h-old memory.
We next tested whether the swim would enhance memory for a negatively conditioned response using aversively reinforced left/right discrimination on a Y-maze. A “short” training protocol was used as follows: On Day 1, naïve rats were conditioned in four or five reinforced trials to make a left or right turn to avoid foot-shock (
We repeated Experiment 2a, this time prolonging the interval between the forced swim and the retention test to 6 d in an effort to evaluate whether some lingering condition, like enhanced arousal at the time of the retention test, might account for the swim-induced enhancement of memory. Day 1 learning did not differ between the groups (
Once again, the forced swim enhanced the expression of the Day 1 memory. The enhancement did not depend on whether learning was appetitively (Expt. 1) or aversively (Expt. 2) reinforced. The memory enhancement was long lasting, at least for 6 d.
Experiment 3a tested whether the swim would enhance an already strong memory. To form a strong aversively reinforced left/right discrimination memory on Day 1, naïve rats received the “intensive” training protocol on the Y-maze as follows: They received training trials until they reached a criterion of 9 correct choices out of 10 consecutive trials and then they were given 30 additional trials. On Day 2 the rats were given either the forced swim (Sw4) or they were handled (NoSw4). On Day 3, the safe and punished arms were switched for reversal training, so the rat had to escape to the opposite arm than on Day 1. The reversal test was used to assess memory because after intensive training, control rats would perform perfectly on an extinction test, making it problematic to observe enhanced expression of memory. The groups did not differ on Day 1. The number of errors increased significantly from Day 1 to Day 3 in both groups (Sw4: t12 = 3.5,
In Experiment 3a, either the stressful swim enhanced the expression of left/right discrimination memory as it did in Experiments 1 and 2, or the swim impaired left/right discrimination learning on the reversal test. Experiment 3b was performed to distinguish between these possibilities. Rats were forced to swim and then 24 h later given intensive left/right discrimination training (SwD0). Learning in this group was not different from the initial learning of the groups in Experiment 3a (Sw4 and NoSw4;
Experiments 1, 2, and 3 all demonstrate the forced swim enhanced the expression of memory. This phenomenon was robust; the memory enhancement persisted at least 6 d. It was observed for both appetitive and aversive conditioning, for weak and strong memories, and whether memory was assessed by extinction or reversal tests.
Stress modulates memories that are undergoing cellular consolidation
On Day 1, the intensive training protocol was used to condition an aversively reinforced left/right discrimination in naïve rats. On Day 2, the rats were divided into five groups. To replicate the enhancing effect of swim (Experiment 3a), one group of rats was only handled (NoSw-NoECS) and a second group was forced to swim (Sw-NoECS). To test if the memory was consolidating 24 h after training, rats in a third group were given ECS and not forced to swim (NoSw-ECS). To examine the effect of ECS after the swim, rats in the remaining two groups were given ECS immediately (Sw-ECS) or 5 h after swim (Sw-delECS).
On Day 1, the groups did not differ. Day 3 testing replicated the significant group effect of Experiment 3a (F4,58 = 4.5,
All rats were trained in the intensive training protocol on Day 1. On Day 2, ECS was delivered immediately after the swim in the Sw-ECS group (black,
The ECS in the NoSw-ECS group did not alter the expression of memory because their performance was not different from the NoSw-NoECS group's performance (post hoc,
Unexpectedly, the Sw-ECS group differed from the Sw-NoECS and Sw-delECS groups (post hoc,
The results together suggest that the swim activated a stable memory, making it transiently sensitive to amnestic treatment. However, it is also possible that the rapid reversal learning caused by forced swim followed immediately by ECS is a result of a change such as increased arousal that persisted at the time of the retention test.
Experiment 5 was designed to test the hypothesis suggested by the results of Experiment 4, that the swim makes an already consolidated memory sensitive to amnestic treatment. Propranolol reliably causes amnesia for a recently activated one-trial inhibitory avoidance memory
Inhibitory avoidance was tested using a box with separate bright and dark compartments. On Day 1, naive rats were conditioned by foot-shock to inhibit the preference for entering the dark compartment. On Day 2, the rats were divided into five groups. To test whether the memory already consolidated within 24 h, the first group was injected (1 ml/kg i.p.) with saline (NoSw-Sal) and the second group was injected with Propranolol (NoSw-Pro). To test whether the swim made the memory sensitive to Propranolol, the third group was forced to swim and then injected with saline (Sw-Sal), and the fourth and fifth groups were forced to swim and then injected with Propranolol either immediately (Sw-Pro) or 5 h (Sw-delPro) after swim. The Propranolol dose administered was 10 mg/ml/kg i.p.
On Day 1, all animals rapidly moved to the dark compartment and step-through latencies were not different between the groups (all averages <10 s; see
(A) Experiment 5—Propranolol caused amnesia of inhibitory avoidance memory only if it was administered after the forced swim. Rats were trained in the inhibitory avoidance paradigm on Day 1. On Day 2, they were either forced to swim (Sw) or just handled (NoSw), and immediately afterwards injected with 10 mg/ml/kg Propranolol or saline (NoSw-Sal
Propranolol attenuated inhibitory avoidance to a level that was below the level of the no swim control rats, suggesting the original memory was disrupted. In contrast, ECS administration in Experiment 4 attenuated the enhanced left/right discrimination to the level of the no swim controls, suggesting ECS disrupted the updating but not the original avoidance memory. These differences may be attributed to the treatment doses, the task differences in left/right discrimination, and inhibitory avoidance and/or differences in the brain regions that are critical for these behaviors. Left/right discrimination, for example, is not sensitive to hippocampal dysfunction (see Experiment 8), while inhibitory avoidance is
Propranolol blocks the adrenergic component of stress, so the effect of the drug on memory in Experiment 5 also suggests that stress can alter a stable, consolidated memory. We used Dexamethasone, a potent suppressant of the hypothalamic-pituitary-adrenal (HPA) axis to investigate further whether the stress of the forced swim triggers the memory alteration (
Using different memory paradigms, Experiments 4–6 revealed that the swim made conditioned avoidance susceptible to amnestic treatment, and activation of both the adrenergic and HPA components of stress are crucial for the phenomenon. The failure of ECS and Propranolol to affect the day-old memory in Experiments 4 and 5 caused us to reject the hypothesis that memory was undergoing cellular consolidation at the time of the forced swim. Nonetheless, the forced swim improved expression of left/right discrimination memory and made the expression of conditioned avoidance memories sensitive to ECS and Propranolol, phenomena that are normally triggered by memory activation
Together, Experiments 1–6 demonstrate that a stressful swim reactivates consolidated memories causing them to be strengthened or, alternatively, causing them to weaken when an amnestic treatment followed the swim. While we are not certain of the mechanism, these results can be explained if the swim-elicited stress response coupled with the activated memory to enhance reconsolidation
The hypothesis that discrimination memory was activated by the forced swim was tested using the phenomenon of IHT of lateralized memory. Learning under functional hemidecortication when one (“non-trained”) hemicortex is inactivated causes a “lateralized memory state,” in which subsequent expression of the memory relies on the (“trained”) hemicortex that was active during learning
The hypothesis that the swim activates memory makes two strong predictions. First, if the forced swim activates memory, then it should induce IHT. The second prediction is that IHT should be blocked by inactivating the trained hemisphere to prevent memory activation during the swim.
On Day 1, left/right discrimination memory was lateralized using the protocol of Goldowitz et al.
The groups did not differ on Day 1 (
The intensive training protocol was administered under unilateral CSD (shading) on Day 1, which led to the formation of a lateralized left/right discrimination memory (arrow). The next day the forced swim was administered with an intact brain in group Lat-Sw (black,
The swim modified discrimination memory by enhancing its expression, by switching it from a consolidated to a labile state, and by modifying what part of the brain could retrieve it, a process thought to require synapse-specific plasticity. We conclude that the stressful swim activated memory.
In principle, rodent memory activation could be triggered by internal variables like the level of a circulating hormone
Temporary inactivation of hippocampus with long-acting (6–10 h) tetrodotoxin (TTX)
The intensive training protocol was used to first determine whether bilateral hippocampal inactivation by TTX impairs Day 1 learning in naïve rats. Injecting TTX (D1-TTX,
(A) Experiment 8a—Bilateral TTX inactivation of dorsal hippocampus in the D1-TTX (black,
The intensive training protocol and reversal test were next used to determine whether bilateral hippocampal inactivation by TTX impairs Day 3 retrieval. Eighteen naïve rats were given intensive left/right discrimination training on Day 1. On Day 3, an hour before reversal training, rats were injected in both hippocampi either with TTX (D3-TTX) or saline (D3-Sal). The groups were not different on Day 1. The TTX injection did not impair Day 3 retrieval compared to the saline-injected controls (t16 = 1.8,
Experiments 8a and 8b demonstrated that the left/right discrimination memory could be acquired and recalled independently of the hippocampus. This put us in a position to ask whether hippocampus is necessary for OCAM itself.
If the hippocampus is important for OCAM, then bilateral inactivation of hippocampus during the stressful swim should block the swim-induced enhancement of memory. Naïve rats received intensive left/right discrimination training on Day 1. On Day 2, they received bilateral injections of either TTX (Sw-TTX) or saline (Sw-Sal) in the dorsal hippocampi. One hour later, all rats were forced to swim for 20 min. On Day 1, the groups did not differ. On the Day 3 reversal test (
These data suggest the hippocampus was necessary for the swim-induced memory enhancement.
If the hippocampus is necessary for OCAM, then bilateral inactivation of hippocampus during the swim should also block swim-induced IHT of a lateralized memory. On Day 1, naïve rats received intensive left/right discrimination training under CSD in one hemicortex to cause lateralized memory formation. On Day 2, 21 rats received bilateral intrahippocampal injections of lidocaine (Lat-Sw-Lid) and 11 rats were injected with saline (Lat-Sw-Sal). Lidocaine was used instead of TTX because lidocaine only blocks neural transmission for ∼30 min. Immediately after the injection, the rats were forced to swim. Two hours later CSD was elicited in the opposite side to the Day 1 training and memory was assessed by the reversal test. The groups did not differ on Day 1; the groups did differ on Day 2 (t30 = 2.3,
Left/right discrimination was conditioned on Day 1 using the intensive training protocol with one hemicortex inactivated by cortical spreading depression (CSD). On Day 2 rats received bilateral hippocampal injections of saline (Lat-Sw-Sal, white,
Because the lidocaine-injected rats expressed no Day 1 memory, this result indicates that OCAM required a functional hippocampus during the swim.
The results of this set of experiments suggest that stress can activate memory, even if the memory is unrelated to the stressful experience. We use the term “memory activation” in the established sense that the term is used in the consolidation and reconsolidation literatures, to mean that memory is in a labile (“active”) state rather than an inert (“inactive”) state
The activation by forced-swim stress was independent of the conditioned and external contextual stimuli that were present during learning, leading us to call the phenomenon OCAM (alternative interpretations are considered and rejected in the section that follows). OCAM seems to be a general phenomenon that does not depend on whether the conditioned response is rapidly extinguished (Experiment 1) or persistent (Experiments 2–10) or whether the activated memory is acquired during single (Experiment 5) or multiple (all other experiments) appetitively (Experiment 1) or aversively (all other experiments) conditioned trials that reinforce an inhibitory (Experiment 5) or an active (all other experiments) conditioned response. OCAM also seems to occur whether or not memory expression depends on the hippocampus (inhibitory avoidance) or the neocortex (left/right discrimination). Both beta-adrenergic activation (
Forced swimming is probably arousing, in which case, can the results be explained by a post-learning facilitation of memory caused by enhanced arousal during the retrieval test 1 to 6 d after the swim? We think this explanation is unlikely for several reasons. To our knowledge, the longest reported post-training interval for an effective memory facilitating treatment is 6 h, and no treatments have been effective 9 h after learning
The corticosterone data as well as the ineffectiveness of Propranolol administration 5 h after the swim (Experiment 5) make it unlikely that persistently altered stress hormone changes alone account for the swim-induced memory activation and enhancement. Although unlikely, it is nonetheless still possible that other endogenous hormonal changes persisted after the swim and account for the enhanced memory on the retention tests 24 h or even 6 d later (Experiment 2b), for example by inducing preservative behavior. All the results, however, cannot be readily explained by some unidentified change in hormonal expression at the time of the retention tests. It is particularly difficult for the possibility of a persistent hormonal change to explain the IHT results of Experiments 7 and 10 because the swim triggered IHT, which is not merely an enhancement of memory or retrieval but a change in the information content that can be localized to a brain region.
It is unlikely that stress itself could have served as a reminder cue for activating the conditioned response. At least in Experiment 1, endogenous corticosterone levels were distinct between the learning and swim experiences. Levels were elevated by 300% by the forced swim, but compared to the levels of naïve animals taken from the cage and sacrificed, corticosterone levels were unaltered by either appetitive learning or retrieval regardless of the intervening swim experience. While it is still possible that an unidentified stress-triggered hormonal change acted as a reminder for the memory activation, this hypothesis is not falsifiable. We formulated the alternative, falsifiable hypothesis that memory was activated out-of-context. Even within the literature on endogenous state-dependent learning and recall, we are unaware of any other reports of memory activation in the absence of the external stimuli that were present during learning. If the OCAM hypothesis is falsified, it will be valuable to learn what stimulus was the reminder that triggered memory activation during the swim. At the very least, this may provide an experimental model of the mental process that in people appears to be context-free recall.
Although some features of the forced-swim procedure were common to the training and retention procedures such as handling by a human, the results of the control animals contradict the possibility that uncontrolled external features cued the swim-induced memory activation. The behavioral testing and swim environments were different rooms to minimize contextual similarities. Although in the first experiments, all the animals were transported from the vivarium for both the training and the swim procedures, the rats that received the control swim procedures did not show evidence of memory activation. In subsequent experiments, we excluded the possibility that transport from the vivarium to the laboratory was the trigger for the memory activation by putting the rats to swim in a bucket that was placed just below their home cage in the vivarium. The swim still elicited a robust enhancement of memory in all these experiments. Furthermore, evidence of memory activation was not observed in any of the many control groups, even though the control animals received the same environmental exposures and handling as the rats that were forced to swim, with the exception of swimming. These procedures included transport the rare times it was done, being manipulated by hand, and actions to dry the fur with paper towels. It is nonetheless possible that some feature of being in the bucket with deep water was not reproduced by the control swim procedure of being in a similar bucket with shallow water, and that feature was common to the L/R discrimination or inhibitory avoidance learning and sufficient to activate the relevant memory.
We see no alternative to concluding that the forced swim activated a consolidated memory in the absence of external conditioned and contextual stimuli and in the absence of the conditioned response. Neither external conditioned nor contextual stimuli were present during the swim to elicit the conditioned responses, and the rat could not express these responses during the swim. While the rule of parsimony requires concluding that OCAM occurs in rats, this conclusion may be unintuitive. It seems, however, that stress-triggered OCAM could provide the basis for an obviously adaptive biological advantage. People commonly review recent past experience in response to current adversity in the effort to identify the cause of the adversity and increase the possibility of avoiding it in the future. This cognitive ability would also confer an adaptive advantage to lower animals. This ability would benefit from causal reasoning, and there is evidence, albeit inconclusive, that causal reasoning occurs in rats
The forced swim modified a consolidated memory and no anterograde learning effects were detected (Experiment 3b). The results of the first three experiments that assayed enhanced memory expression did not distinguish between whether the forced swim had its effect on memory storage or the process of memory retrieval, including for example by inducing perseveration during the reversal tests for memory strength. However, the results of Experiments 4 and 5 with amnestic agents indicate the swim-induced modifications only occurred soon after the swim but not after a 5-h delay. This strongly suggests the effect of swim was not on retrieval itself, which occurred a day later. The results of Experiment 2b are also consistent with an effect on storage rather than retrieval, because 6 d after the swim, we also observed the enhanced expression of intensively conditioned left/right discrimination memory (
The forced swim activated consolidated memories that were 24 h old, to the best of our knowledge mimicking the basic phenomenon of reconsolidation, possibly with an important distinction. Reconsolidation is said to occur when a consolidated memory is retrieved and the activation converts the memory from a biochemically stable state to a labile state
We investigated the OCAM effect in several memories, but we only assayed each memory in isolation, so whether stress activates all or a subset of the rat's memories remains an open question. We suspect that the answer will be complex because whether and how a memory is modified after retrieval depends on the strength and age of the memory
Blocking hippocampal activity during the swim prevented both the swim-induced memory enhancement and the swim-induced IHT of lateralized memory for left/right discrimination, the learning or expression of which is insensitive to hippocampal inactivation. This suggests that hippocampal activity during the swim was necessary for the out-of-context activation of an extra-hippocampal memory. The results do not indicate whether the role of hippocampus was only to mediate the response to stress or whether hippocampal memories were specifically activated. The data demonstrate the hippocampus plays a role in memory beyond its role in associative memory storage
The results of Experiments 8, 9, and 10 suggest that OCAM is a hippocampus-dependent process that appears to alter memory in extrahippocampal sites. OCAM is a common feature of human conscious recollection, but despite a recent suggestion that hippocampus is important for recollection in rats
The findings presented here indicate that under acute stress, the hippocampus is involved in activating a set of arbitrary memories that can be stored at both hippocampal
While further investigations of the effects of stress on multiple, concurrent memories are warranted, our observations indicate that stressful experience alters diverse associative memories. We only found evidence of memory enhancement, for both weak and strong associations; it however remains possible that other forms of memory that we did not test were weakened by the stress. Nonetheless, at this point, our observations suggest that in stress-induced OCAM, stress acts to generally strengthen memory rather than acting to strengthen some and weaken others. If confirmed, this may help understand the memory dysfunction in PTSD and other stress-related mood disorders. We hypothesize that stress-triggered memory activation creates a condition where multiple memories coactivate, and through mechanisms of synaptic plasticity
The experiments were conducted in accordance with Institutional (SUNY, IACUC 07-197-05) and NIH guidelines, and the directive of the European Communities Council (6/609/EEC).
Male rats of the Long-Evans strain weighing 350–450 g were used. The experiments were performed during the light period (07:00 to 19:00) of a 12 h:12 h cycle. Rats were habituated to handling by the experimenter for 3–5 d prior to behavioral testing.
Trunk blood was collected under Halothane anesthesia. After overnight storage, the blood was centrifuged at 4,000 rpm for 10 min; the supernatant was withdrawn and then stored frozen until assayed by radioimmunoassay.
More than 10 experiments were performed requiring the use of a large number of behavioral and experimental manipulations. Here in the
Rats were forced to swim individually for 20 min in a covered bucket (diameter 30 cm) filled to 30 cm with 27°C water. The bucket lid had six small holes to allow air in and the experimenter to observe the rat. Afterwards the rats were dried with paper towels and returned to the home cage. All animals survived the experience and none required additional follow-up care. Unless stated otherwise, the control rats spent the same time in the experimental room and were treated like the experimental animals with the exception that they were not put into the bucket and forced to swim.
The rats were food-deprived to 85% of their weight. During 5–6 d they were habituated to the T-maze (45×12×15 cm (l× w × h) arms) and to eat 3 cocoa puffs (General Mills, Minneapolis, MN) during 2 min at the choice point. All rats then received five acquisition trials. A trial began by placing the rat in the start arm and ended when the rat entered a choice arm by half a body length or 120 s elapsed. If the rat entered the goal arm, it was given 3 cocoa puffs. If the rat did not enter a choice arm within 120 s, it was placed in the goal arm, given 3 cocoa puffs, and an error was scored. On Day 3, the rats were allowed 120 s to make a choice on each of three unreinforced retention trials; all rats responded within 120 s.
The Y-maze had opaque walls (40×10×30 cm; 120° between arms) and an electrifiable floor made of parallel rods. Each rat was habituated to the maze for 5 min before being trained to escape from a fixed start arm to one of the two choice arms. On each trial, the rat was placed in the start arm and 5 s later foot-shocks (50 Hz, 0.5 mA, 0.5 s) were delivered every 3 s until the rat escaped to the goal arm or 60 s elapsed. The response was correct if the rat escaped directly to the goal, and an error was scored if the rat entered the incorrect arm by at least half of its body. If 60 s elapsed, the rat was put into the goal arm and an error was scored. Each rat was allowed to spend approximately 30 s in the goal arm before it was placed in the home cage (“short” training) or in the start arm (“intensive” training).
In the short training protocol (Experiment 2) the first choice was always considered an error and the other arm was designated the goal. Training continued until either three correct choices or three errors. Rats that made three errors were excluded from the study (
In the intensive training protocol (Experiments 3, 4, and 6–10) the first choice was always considered an error and the other arm was designated the goal. After 9 of 10 consecutive responses were correct, an additional 30 trials were given. Retention was tested by reversal training in which the rats had to escape to the opposite arm than on Day 1 (the Day 1 error arm). The number of errors to the criterion of four consecutive correct responses was used to compare acquisition and retention. More errors during the Day 3 reversal test indicated better retention of Day 1 memory.
The apparatus consisted of two plastic boxes connected with a guillotine door. The brightly lit white start box (30×20×13 cm) had white plastic walls, a metal parallel rod floor, and a Plexiglas ceiling. The dark shock box (25×15×13 cm) had a dark gray plastic ceiling and walls and an electrifiable floor. Rats received two baseline trials and one acquisition trial at 30 min intervals. The rat was placed in the start compartment with its back towards the door, and approximately 5 s later, the guillotine door was raised when the rat was not facing it. After entering the shock compartment, the door was closed and on the baseline trials, 15 s later the rat was returned to its home cage. On the third trial (acquisition), once the rat entered the black compartment, it received two 0.6-mA, 2 s foot-shocks (50 Hz) separated by 1 s, and immediately afterwards the rat was returned to the home cage. Retention was measured without reinforcement by the latency to enter the black compartment. If 300 s elapsed, the rat was removed and the step-through latency was set to 300 s.
Rats were implanted under Nembutal anesthesia (50 mg/kg) with a pair of stainless steel injection guide cannulae aimed 1.5 mm above the injection targets in the dorsal hippocampus as described in detail
The rats were pre-treated with atropine (1 mg/kg) and 5 min later anesthetized by a mixture of ketamine hydrochloride (90 mg/kg) and xylazine hydrochloride (14 mg/kg). Two trephine holes (3 mm diameter) were made over both fronto-parietal cortices without damaging the dura, and each was fitted with an aluminum well (4 mm inner, 6 mm outer diameter, 5 mm high) that allowed free access to the dura in the course of the experiment. This assembly was fixed to the skull with dental acrylic and two anchoring screws. The exposed dura was protected from desiccation by saline-soaked cotton, and both wells were covered with a metal cap. One day was allowed for recovery. Unilateral repeating waves of CSD were elicited using the method of Burešova
Rats were placed in a plastic holding cage next to the forced swim bucket. A pair of electrodes was clipped to the ears and an ECS (50 mA, 50 Hz, 1 s) was delivered. After the treatment, the rats were returned to their home cage to recover.
Average measures ± SEM are reported. Significant differences confirmed by ANOVA were followed by Newman-Keuls post hoc tests. The results of these pair-wise comparisons are reported in the main text, and the statistical details are given in the corresponding figure legends. Chi-square and
We thank Adam Cassino for experimental assistance and Robert Sutherland for valuable comments.
cortical spreading depression
electro-convulsive shock
hypothalamic-pituitary-adrenal
inter-hemispheric transfer
Out-of-context activation of memory
post-traumatic stress disorder
tetrodotoxin