Histamine in the basolateral amygdala promotes inhibitory avoidance learning independently of hippocampus

Significance

Integrity of the brain histaminergic system is necessary for long-term memory (LTM) but not short-term memory of step-down inhibitory avoidance (IA). Histamine depletion in hippocampus or basolateral amygdala (BLA) impairs LTM of that task. Histamine infusion into either structure restores LTM in histamine-depleted rats. The restoring effect in BLA occurs even when hippocampal activity was impaired. Cyclic adenosine monophosphate (cAMP) responsive-element-binding protein phosphorylation correlates anatomically and temporally with histamine-induced memory recall. Thus, histaminergic neurotransmission appears critical to provide the brain with the plasticity necessary for IA through recruitment of alternative circuits. Our findings indicate that the histaminergic system comprises parallel, coordinated pathways that provide compensatory plasticity when one brain structure is compromised.

Abstract

Recent discoveries demonstrated that recruitment of alternative brain circuits permits compensation of memory impairments following damage to brain regions specialized in integrating and/or storing specific memories, including both dorsal hippocampus and basolateral amygdala (BLA). Here, we first report that the integrity of the brain histaminergic system is necessary for long-term, but not for short-term memory of step-down inhibitory avoidance (IA). Second, we found that phosphorylation of cyclic adenosine monophosphate (cAMP) responsive-element-binding protein, a crucial mediator in long-term memory formation, correlated anatomically and temporally with histamine-induced memory retrieval, showing the active involvement of histamine function in CA1 and BLA in different phases of memory consolidation. Third, we found that exogenous application of histamine in either hippocampal CA1 or BLA of brain histamine-depleted rats, hence amnesic, restored long-term memory; however, the time frame of memory rescue was different for the two brain structures, short lived (immediately posttraining) for BLA, long lasting (up to 6 h) for the CA1. Moreover, long-term memory was formed immediately after training restoring of histamine transmission only in the BLA. These findings reveal the essential role of histaminergic neurotransmission to provide the brain with the plasticity necessary to ensure memorization of emotionally salient events, through recruitment of alternative circuits. Hence, our findings indicate that the histaminergic system comprises parallel, coordinated pathways that provide compensatory plasticity when one brain structure is compromised.

Emotionally arousing experiences create long-term memories that are initially labile, but over time become insensitive to disruption through a process known as consolidation (1). One-trial fear-motivated learning tasks, such as step-down inhibitory avoidance (IA), a hippocampal-dependent associative learning (2), have largely contributed to the knowledge of consolidation process, and convincing evidence indicates that the CA1 region of the hippocampus, the basolateral amygdala (BLA) and the medial prefrontal cortex, are crucially involved in this process (3, 4). However, the actual contribution of each region remains poorly characterized. For instance, it is suggested that they are part of mostly independent circuitries specialized in encoding specific aspects of information, e.g., the emotional component in the BLA and the cognitive aspect in the hippocampus (2, 5). If this relation between structures holds true for an intact brain, however, examples of learning recovery of individuals bearing large brain lesions suggest that the brain can adapt dynamically, and interconnected systems can provide compensation for selective damage (2). Indeed, hippocampal damage resulted in retrograde amnesia for acquired fear memories (6), but did not prevent new learning in rodents (7) as well as in humans (8). Therefore, hippocampal damage before new learning can be overcome, possibly through recruitment of an alternative circuit. The identity of the compensatory structures is still unknown, but the BLA could potentially be one, because it operates to a certain extent in parallel with the CA1 region in memory processing (9).

Extensive evidence indicates that emotionally significant experiences activate many hormones and neurotransmitters, including histamine, that regulate the consolidation of newly acquired memories (10, 11). Histamine is synthesized from histidine by histidine-decarboxylase (HDC) (12) and released in the brain from varicosities of axons that ramify extensively throughout the central nervous system. The only source of histaminergic fibers is the hypothalamic tuberomamillary nucleus (TMN) (13). The histaminergic system is crucial in the sleep–wake cycle and is implicated in various brain functions, including the modulation of hippocampal synaptic plasticity (12, 14). Interestingly, when infused into the CA1 region immediately after training of an IA task, histamine induced a dose-dependent promnesic effect through activation of H₂ receptors without altering locomotor activity, exploratory behavior, anxiety state, or retrieval of the avoidance response (15). Consistently, posttraining injections into the dorsal hippocampus of histamine H₂ or H₃ receptor agonists improved memory consolidation after contextual fear conditioning through a mechanism involving extracellular signal-regulated kinase (ERK)₂ phosphorylation (16). Several neurotransmitters, such as dopamine, glutamate, and norepinephrine, activate the ERK cascade in the hippocampus (17), and histamine may interact with these neurotransmitters to orchestrate ERK₂ phosphorylation that appears to play a critical role in consolidating emotional memories (17). Histamine modulates memory of emotionally arousing experiences also in the BLA. Administration of H₃ receptor antagonists into the BLA impaired consolidation of fear memories (18), whereas H₃ receptor agonists ameliorated the expression of adverse memories (19). This effect involved H₂ receptors and was accompanied by a bimodal modulation of the local cholinergic tone (18, 19).

Although these findings indicate that administration of histaminergic ligands modulates memory consolidation, studies have not yet investigated the role of endogenous histamine in creating memories for emotionally arousing training. To answer this question, we performed a first set of the experiments to examine how brain histamine depletion obtained by using intralateral ventricle (LV) administration of a-fluoromethylhistidine (a-FMHis), a suicide inhibitor of HDC (20), affected IA memory processes. With the second set of experiments, we investigated IA training-induced CREB phosphorylation in the brain of normal and histamine-depleted rats. The last set of experiments learned whether the local infusion of histamine in the BLA or the CA1 region, respectively, overcame a-FMHis–induced amnesia for IA training.

Results

Depletion of Histamine Impairs Long-Term but Not Short-Term Memory.

Administration of a-FMHis (5 µg/µL) quickly suppressed baseline and histamine H₃ receptor antagonist-evoked release of histamine from the TMN of freely moving rats, as 180 min after injection, histamine release values decreased below the sensitivity of the method (Fig. S1). Histamine release was restored to control levels after approximately 3 d.

To investigate the role of histamine in short- and long-term memory retention, we examined the performance of rats treated with a-FMHis 24 h before training in the one-trial step-down IA. The control group received intra-LV infusion of equivalent volume of saline. Retention test was carried out in different groups of animals at 2 h, 24 h, or 7 d after training. There were no differences in training performances in any group examined (Fig. 1). Latency of the a-FMHis group did not significantly differ from that of controls at the 2-h retention test (Fig. 1A), but it was significantly shorter at the 24-h retention test (unpaired t test, P < 0.0001 vs. controls; Fig. 1B). However, the latency of the a-FMHis group and controls did not differ at the 7-d retention test (Fig. 1C), indicating that all animals formed a memory trace of the training experience. Restoration of histamine release to control levels about 3 d after treatment with a-FMHis may explain this finding. To test this hypothesis, a-FMHis was administered repeatedly on the second and sixth day after training. Under this condition, memory retention tested 7 d after training was impaired (unpaired t test, P < 0.0001 vs. controls; Fig. 1D). Thus, histamine depletion blocked long-term memory while leaving short-term memory intact. The finding that a-FMHis had no effect on short-term memory rules out the possibility of its influence on acquisition or retrieval mechanisms. Moreover, no changes in exploratory locomotor activity in the open field test were observed between controls and a-FMHis–treated rats.

Fig. 1.

Effect of histamine acute depletion through a-FMHis administration on IA task. The schematic drawings above A and D show the sequence of procedures and treatment administrations. Rats implanted with an infusion cannula in the LV received a-FMHis or saline (SAL) 24 h before training. Retention test was performed 2 h (A), 24 h (B), and 7 d (C) after training. Latencies of a-FMHis groups did not significantly differ from respective controls on the 2-h and the 7-d retention test, but they were significantly shorter on the 24-h retention test. Data are expressed as means ± SEM of 8–15 animals for each group; unpaired t test, ****P < 0.0001 vs. controls. (D) a-FMHis or SAL was infused into the LV 24 h before and 2 and 6 d after training. Latencies of a-FMHis group were significantly shorter on the 7-d retention test. Data are expressed as means ± SEM of 12–15 animals for each group (unpaired t test, ****P < 0.0001 vs. controls).

Effect of Histamine Depletion on IA-Induced Increase of Cyclic Adenosine Monophosphate Responsive-Element-Binding Protein Phosphorylation in Rat BLA and Hippocampus.

Cyclic adenosine monophosphate (cAMP) responsive-element-binding protein (CREB) is a crucial mediator in the formation of long-term memory (21), and an increase in CREB phosphorylation at Ser-133 in the hippocampus is specifically associated with IA memory formation (22, 23). In particular, a significant increase of pCREB in the hippocampus occurs immediately after training, followed by a delayed increase of Ser-133 pCREB 3–6 h later (22). It is conceivable that for a short period after training, the hippocampus acts in concert with the amygdala that contributes emotional values (2).

To examine the influences of histamine depletion on CREB phosphorylation following IA training, rats were euthanized 10 min or 5 h after training, and pCREB levels were assessed in the amygdala and the CA1 of rats given saline or a-FMHis infusions in the lateral ventricle (LV) 24 h before training. Controls received saline into the LV and no foot shock (untrained). pCREB levels were measured also in rats that received saline into the LV and were placed directly over the electrified grid (saline/foot-shocked) to control for potential changes in CREB phosphorylation due to the foot shock itself. Representative immunoblots and the densitometric analysis are shown in Fig. 2 A (10 min) and B (5 h). Ten min after training, pCREB levels measured in controls and in saline/foot-shocked animals were not significantly different (Fig. 2A). Conversely, both amygdala and CA1 of rats given saline or a-FMHis displayed a significant increase of pCREB density [one-way ANOVA and Bonferroni’s multiple comparison test (MCT), amygdala: F_3,15 = 11.64; P < 0.0007; CA1: F_3,15 = 7.3; P < 0.004] compared with controls (Fig. 2A). Five hours after training, no difference of pCREB density was found among the amygdala of controls and trained rats treated with saline or a-FMHis (Fig. 2B, Left). In the CA1, however, only trained rats treated with saline, but not those with a-FMHis, showed a significant increase of pCREB density compared with controls (one-way ANOVA and Bonferroni’s MCT, F_2,19 = 7.498; P < 0.004; Fig. 2B). Therefore, a-FMHis administration caused amnesia of IA training on the 24-h retention test (Fig. 1B) and blocked the increase of hippocampal pCREB associated with IA training (Fig. 2B).

Fig. 2.

Effect of histamine (HA) depletion on the IA training-induced increase of CREB phosphorylation in rat amygdala and CA1. The schematic drawing on top of the figure displays the sequence of procedures and treatment administrations. Rats were implanted with an infusion cannula in the LV. (A) Representative immunoblots and densitometric quantification show that at 10 min after training pCREB/CREB ratio was increased in the amygdala and the CA1 of all rats undergone IA training, independently of having received saline or a-FMHis (n = 3–5 animals for each group). (B) At 5 h after training, pCREB/CREB ratio increased only in the CA1 of rats infused with SAL. Data are expressed as means ± SEM of 3–7 animals for each group; *P < 0.05, **P < 0.01 vs. controls, one-way ANOVA and Bonferroni’s MCT.

Effects of Histamine or 8-Bromoadenosine-3′,5′-cAMP Infusion into the BLA on a-FMHis–Induced Amnesia.

To further test the idea that IA memory consolidation may depend on histaminergic signaling, we investigated whether administration of histamine (1 µg/µL) into the BLA counteracts the amnesic effect of a-FMHis. We tested also the membrane-permeable analog of cAMP, 8-bromoadenosine 3′,5′-cAMP (8Br-cAMP), considering that IA memory formation involved cAMP/cAMP-dependent protein kinase signaling pathway (22), and required CREB phosphorylation (24, 25). Rats received a-FMHis or saline into LV 24 h before training, and saline, histamine, or 8Br-cAMP (1.25 µg/µL) into the BLA bilaterally immediately or 110 min after training. Retention test was carried out 24 h after training. Latency for all groups during training did not differ (Fig. 3 A and B). Fig. 3A shows the latency of rats given infusions of histamine or 8Br-cAMP into the BLA immediately after training. One-way ANOVA performed on the retention test revealed a significant difference across groups (F_5,74 = 20.56; P < 0.0001). Further analysis with Bonferroni’s MCT (Fig. 3A) showed that latencies of rats treated with a-FMHis (LV) and saline (BLA) were significantly shorter than those of all of the other groups, including animals treated with a-FMHis intra-LV and histamine or 8Br-cAMP intra-BLA (Fig. 3A). Thus, both histamine and 8Br-cAMP fully antagonized the amnesic effect of a-FMHis. Interestingly, infusion of either histamine or 8Br-cAMP per se enhanced step-down latency of rats treated with saline or with a-FMHis, compared with control group (Fig. 3A), thus suggesting a memory improving effect. Conversely, intra-BLA infusions of either histamine or 8Br-cAMP 110 min after training did not reverse a-FMHis–elicited amnesia (Fig. 3B). Indeed, although one-way ANOVA performed on the retention latency displayed a significant difference across groups (F_5,58 = 5.748; P < 0.0003), Bonferroni’s MCT analysis revealed that latencies of all rat groups infused with a-FMHis (LV) did not differ significantly, and were significantly shorter compared with controls (Fig. 3B).

Fig. 3.

Effect of HA or 8Br-cAMP infusion into the BLA on the amnesia of IA training induced by a-FMHis. The schematic drawing on top of the figure displays the sequence of procedures and treatment administrations. Rats were implanted with infusion cannulae in the LV and in the BLA bilaterally. Rats receiving saline into both the LV and BLA served as controls. Retention test was performed 24 h after training. (A) HA, 8Br-cAMP, or SAL was bilaterally infused into the BLA immediately after training. Data are expressed as means ± SEM of 9–15 animals for each group; *P < 0.05, ***P < 0.001 vs. controls, ^####P < 0.0001 vs. “a-FMHis (LV) + HA (BLA)” and “a-FMHis (LV) + 8Br-cAMP (BLA),” one-way ANOVA, and Bonferroni’s MCT. (B) HA, 8Br-cAMP, or SAL was bilaterally infused into the BLA 110 min after training. Data are expressed as means ± SEM of 8–24 animals for each group; **P < 0.01 vs. controls, one-way ANOVA and Bonferroni’s MCT.

Effects of Histamine or 8Br-cAMP Infusion into Hippocampal CA1 Region on a-FMHis–Induced Amnesia.

Rats received a-FMHis or saline into the LV 24 h before training, and saline, histamine, (1 µg/µl) or 8Br-cAMP (1.25 µg/µL) bilaterally into the CA1 at different posttraining times (Fig. 4, Top). Retention test was carried out 24 h after training. Latencies for all groups during training did not differ (Fig. 4 A–C). Histamine or 8Br-cAMP given immediately after training fully antagonized the effect of a-FMHis (one-way ANOVA and Bonferroni’s MCT, F_5.63 = 31.12, P < 0.0001; Fig. 4A). Histamine or 8Br-cAMP fully antagonized the amnesic effect of a-FMHis also when administered into CA1 110 min (F_5,62 = 6.762, P < 0.0001; Fig. 4B) or 6 h (F_5,80 = 26.17, P < 0.0001; Fig. 4C) after training. Intra-CA1 infusion of histamine 12 h after training failed to antagonize a-FMHis–elicited amnesia, whereas 8Br-cAMP retained its effect (F_5,74 = 23.5, P < 0.0001; Fig. 4D).

Fig. 4.

Effect of HA or 8Br-cAMP infusion into the CA1 region of hippocampus on the amnesia of IA training induced by a-FMHis. The schematic drawing on top of the figure displays the sequence of procedures and treatment administrations. Rats were implanted with infusion cannulae in the LV and in the CA1 bilaterally. Rats receiving saline into both the LV and CA1 served as controls. Retention test was carried out 24 h after training. HA, 8Br-cAMP, or SAL were bilaterally infused into the CA1. (A) Immediately after training. Data are expressed as means ± SEM of 10–12 animals for each group; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 vs. controls, ^####P < 0.0001 vs. “a-FMHis (LV) + HA (CA1),” ^$$$$P < 0.0001 vs. “a-FMHis (LV) + 8Br-cAMP (CA1).” (B) One hundred ten minutes after training. Data are expressed as means ± SEM of 7–19 animals for each group; ***P < 0.001 vs. controls, ^###P < 0.001 vs. “a-FMHis (LV) + HA (CA1),” ^$P < 0.05 vs. “a-FMHis (LV) + 8Br-cAMP (CA1).” (C) Six hours after training. Data are expressed as means ± SEM of 11–17 animals for each group; ****P < 0.0001 vs. controls, ^####P < 0.0001 vs. “a-FMHis (LV) + HA (CA1),” ^$$$$P < 0.0001 vs. “a-FMHis (LV) + 8Br-cAMP (CA1).” (D) Twelve hours after training. Data are expressed as means ± SEM of 11–17 animals for each group; **P < 0.01, ****P < 0.0001 vs. controls, ^$$$$P < 0.0001 vs. “a-FMHis (LV) + 8Br-cAMP (CA1)” (one-way ANOVA and Bonferroni’s MCT).

Discussion

Aversive memories can follow different processing routes, engaging multiple independent circuits. Emotionally relevant experiences activate the histaminergic system (26, 27), and there is an abundant literature demonstrating that activation of histamine receptors (mostly H₂) in the BLA (18, 19), the dorsal hippocampus (15, 16) or the nucleus basalis magnocellularis (28) modulates the consolidation of memory associated to aversive events.

Here we examined IA memory formation in rats temporarily depleted of histamine by LV injections of a-FMHis, an irreversible histidine decarboxylase inhibitor, that completely suppressed spontaneous and evoked histamine release from the tuberomamillary nucleus. Our results provide strong evidence that intact histamine neurotransmission is required specifically for the establishment of long-term aversive memory, whereas short-term memory formation is independent of histamine neurotransmission. In fact, fear memories associated to IA were quickly forgotten when the brain histaminergic system was silenced. Furthermore, we found that memory can be restored in brain histamine-depleted rats, by local injections of histamine in the CA1 region of the dorsal hippocampus or the BLA, with a completely different time course: Whereas in the BLA only immediate posttraining infusion of histamine allows long-term memory formation, in the CA1, reinstatement of histamine consolidates aversive memory even 6 h after training, confirming that the hippocampus is engaged in IA memory processing for a period longer than the amygdala (2, 29).

The unique finding in this study is that the histaminergic transmission in the BLA is crucial for the early phase of IA memory consolidation that occurred despite the blockade of histaminergic neurotransmission in the hippocampus. This is surprising, given the important contribution of hippocampal histamine receptors to IA consolidation (15, 16).

The different effects of histamine depletion on short- and long-term memories are in line with reports of several drug treatments impairing long-term memory, while leaving short-term memory intact (30–32). These findings suggest that short-term and long-term memories are separate processes, a view convincingly supported by pharmacological manipulations blocking short-term memory while keeping long-term memory intact, such as intraentorhinal cortex administration of AMPA receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), GABA_A receptor agonist muscimol, or 5-HT1A receptor antagonist 1-(2-methoxyphenyl)-4-(4-(2-phthalimido) butylpiperazine (NAN) (32). However, when administered into the CA1 region, CNQX or muscimol impair both short-term and long-term memory (32), thus suggesting selective involvement of different receptors in different brain regions.

Memory consolidation lasts several hours or days and relies on molecular events occurring with different timings in several brain regions, including the hippocampus and the amygdala (4). CREB phosphorylation represents a crucial step for the establishment of long-term memories (21, 23, 25, 33), but is not a requirement for short-term memory formation in, e.g., conditioned taste aversion (34). The present study reports that rats given either saline or a-FMHis and submitted to the IA learning paradigm showed an increase of hippocampal and amygdalar CREB phosphorylation 10 min after training, which is agreement with previous reports (22, 33). The lack of difference between histamine-depleted and not-depleted animals indicates that histamine modulates long-term memory formation, influencing mechanisms other than early CREB activation in these two brain regions. Our results suggest instead that histamine activates the CREB pathway to exert its mnemonic effects later during the temporal progression of memory consolidation. Indeed, pCREB levels were augmented 5 h after training only in the CA1 of saline-treated rats that well remembered the IA training. This event did not occur in the CA1 of a-FMHis–treated rats, which displayed long-term memory impairments. This observation fits well with previous reports that IA training increases hippocampal pCREB levels 3–6 h after training (22), suggesting a causal correlation between pCREB levels in the hippocampus and long-term memory formation (23, 33).

However, we did not observe an increase of pCREB in the amygdala 5 h after training, consistent with an active involvement of this structure only in the early phases of IA consolidation (9, 35). A brief temporal window is supported also by experiments showing that reinstatement of histaminergic transmission in the BLA of histamine-depleted rats with exogenous histamine not only restored, but even improved the memory for IA, when local infusions of histamine were performed immediately but not 110 min after training.

The IA training activates molecular changes with different temporal progression in multiple brain areas, such as amygdala, hippocampus, entorhinal cortex, and parietal cortex (36, 37). However, experimental evidence indicates that CA1 and BLA operate also in parallel (9). Therefore, it is plausible that IA memory is spared after inactivation of the histaminergic system in the hippocampus because BLA histaminergic system takes over the process of consolidation, for example, enabling the storage of such information in other brain systems such as the entorhinal cortex or parietal cortex. Many reports have shown that IA training with high-intensity foot shocks reduced the memory impairment caused by hippocampal inactivation (38–40), thus suggesting that consolidation of intense emotional experiences do not require a functioning hippocampus. However, this fear memory is acquired slowly and forgotten when remote memory is tested (41). We suggest that histamine in the BLA promoted inhibitory avoidance learning independently of hippocampus by increasing the emotional value of IA training. This effect is probably achieved through activation of histamine H₂ receptors, whose responses occur through adenylyl cyclase stimulation (42). The activation of cAMP/PKA pathway, which targets CREB signaling, is crucial in IA memory processing (43). Moreover, many reports indicate that activation of H₂ receptors potentiates the memory in aversive tasks, including IA (15, 16, 28). Consistently with this hypothesis, we showed that the nonhydrolysable cAMP analog, 8-Br-cAMP, produces effects similar to those of histamine. When given into the BLA immediately after training, both compounds rescued memory in histamine-depleted animals, but failed to produce this effect when they were administered 110 min after training. Thus, the involvement of BLA sufficient to exert IA long-term memory formation is required only for a brief temporal window. Also, when given in the CA1 of histamine-depleted rats, histamine and 8Br-cAMP had a similar effect in rescuing IA memory. However, histamine was effective up to 6 h after training, whereas 8Br-cAMP rescued the memory of IA training in histamine-depleted rats also when given 12 h after training. Molecular changes in the hippocampus of rats trained with IA begin immediately after training and progress for approximately 20 h (36, 44). The effect of histamine lasts up to 6 h; however, the involvement of other hippocampal modulatory neurotransmission that might contribute to a later phase of memory consolidation cannot be ruled out (22).

Taken together, the present findings suggest that (i) the integrity of the histaminergic system in the brain is crucial for IA long-term, but not short-term, memory formation; (ii) histamine-depleted rats did not express long-term aversive memory and displayed a lack of hippocampal CREB phosphorylation; (iii) long-term memory was formed immediately following posttraining restoration of histamine transmission in the BLA independently of hippocampus. Therefore, despite current views conferring to the BLA essentially a modulatory function in IA memory formation, this and other evidence (9) indicates that it plays a role also in consolidation of this task. A prudent interpretation of these results is that following the local activation of the histaminergic system, the BLA takes over the functions of the hippocampus in the consolidation process and renders hippocampus no longer crucial for long-term memory formation.

Materials and Methods

Animals.

Male Wistar rats (3 mo old, 300–330 g) purchased from Centro de Reprodução e Experimentação de Animais de Laboratorio of the Universidade Federal do Rio Grande do Sul (our regular provider) were used. They were housed four to a cage with water and food ad libitum, under a 12-h light/dark cycle (lights on at 7:00 AM). The temperature of the animal room was maintained at 22–24 °C. All procedures were in accordance with the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals (45) and were approved by Pontifical Catholic University of Rio Grande do Sul.

Surgery.

At least 1 wk after their arrival, the animals were anesthetized (75 mg/kg ketamine plus 10 mg/kg xylazine) and placed on a stereotaxic frame (Stellar; Stoeling). A stainless steel cannula (22 gauge) was implanted in the LV and fixed to the skull by using dental cement, according to the following coordinates: anterior, −0.9 mm; lateral, −1.5 mm; ventral, −3.6 mm (46), and used for a-FMHis administration. Rats were also implanted bilaterally with 22-gauge guide cannulae 1 mm above the CA1 area of the hippocampus or 1 mm above the BLA. The coordinates were anterior, −4.2 mm; lateral, ±3.0 mm; ventral, −1.8 mm for the CA1, and anterior, −2.4 mm; lateral, ±5.1 mm; ventral, −7.5 mm for the BLA (46). Cannulae placements were verified postmortem as described in detail in SI Materials and Methods. Animals were allowed 7 d to recover from surgery before behavioral procedures. Animals were handled once daily for three consecutive days, and all behavioral procedures was conducted between 8:00 and 11:00 AM.

Inhibitory Avoidance Task.

The apparatus consisted in a 50 × 25 × 25 cm Plexiglas box with a 5-cm-high, 8-cm-wide, 25-cm-long Formica platform on the left end of a grid of 1-mm caliber bronze bars spaced 0.8 mm apart. The rats were gently placed on the platform facing the left rear corner. When they stepped down, placing their four paws on the grid, they received a 2-s 0.5-mA scrambled foot shock and then were immediately withdrawn from the training box. Retention test was carried out 2 h, 24 h, or 7 d after the training session. The procedure was the same except that the foot shock was omitted. In the retention test, the step-down latency was 300 s. Latency to step down was measured with an automated stopwatch.

Drugs and Infusion Procedures.

At the time of drug microinfusions, the animals were gently restrained by hand, and the injection needle (30 gauge) was fitted tightly into the guides, extending 1 mm from the tip of the guide cannulae. The injection needle was connected to a 10-μL Hamilton microsyringe, and the infusions were performed at a rate of 0.5 μL/30 s. The infusion cannula was left in place for an additional 60 s to minimize backflow. It was then carefully withdrawn and placed on the other side, where the procedure was repeated. The entire bilateral infusion procedure took approximately 90 s. Infusions into the BLA were performed immediately or 110 min after training; into the CA1 immediately; or 110 min, 3 h or 6 h after training. The drugs used were histamine (1 μg/uL) and 8Br-cAMP (1.25 μg/uL) purchased from Sigma-Aldrich. a-FMHis (5 μg/uL) was synthesized at Abbott Laboratories. The doses were chosen among those found to be effective in previous papers from our group or others (6, 7). The drugs had no effects on locomotion, exploration, or performance in an elevated plus maze (6). The volume of drugs infused was 0.5 µL per side in the BLA and 1 µL per side into the CA1 area and LV. Control groups received equal volumes of sterile saline (0.9%).

Microdialysis Experiments.

The effects of a-FMHis on brain histamine release were evaluated in freely moving rats implanted with microdialysis probes (Fig. S1). For details on surgery, experimental protocols, and HPLC-fluorimetric assay to quantify histamine, see SI Materials and Methods.

Western Blotting Analysis.

Animals were killed 10 min or 5 h after training, the brain was dissected out on ice, and the amygdala and the CA1 were immediately isolated. The pooled structures (left and right) were individually homogenized in 200 µL of ice-cold lysis buffer containing protease and phosphatase inhibitors [50 mM Tris⋅HCl (pH 7.5), 50 mM NaCl, 10 mM EGTA, 5 mM EDTA, 2 mM sodium pyrophosphate, 4 mM p-nitrophenyl phosphate, 1 mM Na₃VO₄, 1.1 mM PMSF, 20 μg/μL leupeptin, 50 μg/μL aprotinin, 0.1% SDS] and centrifuged at 13.8 × g at 4 °C for 15 min. The following procedure is described in detail in SI Materials and Methods. For each sample, a ratio of p^Ser133-CREB/CREB densities was calculated, and then all of the individual rates were expressed as a percentage of the average of ratios obtained from control group.

Data and Statistical Analysis.

Statistical analysis was performed by using Prism Software (GraphPad). Data are expressed as means ± SEM. Inhibitory avoidance latencies as well as pCREB/CREB ratio were analyzed with unpaired t test or one-way ANOVA. The source of the detected significances was determined by Bonferroni’s multiple comparison post hoc test. P values less than 0.05 were considered statistically significant. The number of rats per group is indicated in the figure legends.