Trace and contextual fear conditioning require neural activity and NMDA receptor-dependent transmission in the medial prefrontal cortex

The contribution of the medial prefrontal cortex (mPFC) to the formation of memory is a subject of considerable recent interest. Notably, the mechanisms supporting memory acquisition in this structure are poorly understood. The mPFC has been implicated in the acquisition of trace fear conditioning, a task that requires the association of a conditional stimulus (CS) and an aversive unconditional stimulus (UCS) across a temporal gap. In both rat and human subjects, frontal regions show increased activity during the trace interval separating the CS and UCS. We investigated the contribution of prefrontal neural activity in the rat to the acquisition of trace fear conditioning using microinfusions of the γ-aminobutyric acid type A (GABA_A) receptor agonist muscimol. We also investigated the role of prefrontal N-methyl-d-aspartate (NMDA) receptor-mediated signaling in trace fear conditioning using the NMDA receptor antagonist 2-amino-5-phosphonovaleric acid (APV). Temporary inactivation of prefrontal activity with muscimol or blockade of NMDA receptor-dependent transmission in mPFC impaired the acquisition of trace, but not delay, conditional fear responses. Simultaneously acquired contextual fear responses were also impaired in drug-treated rats exposed to trace or delay, but not unpaired, training protocols. Our results support the idea that synaptic plasticity within the mPFC is critical for the long-term storage of memory in trace fear conditioning.The prefrontal cortex participates in a wide range of complex cognitive functions including working memory, attention, and behavioral inhibition (Fuster 2001). In recent years, the known functions of the prefrontal cortex have been extended to include a role in long-term memory encoding and retrieval (Blumenfeld and Ranganath 2006; Jung et al. 2008). The prefrontal cortex may be involved in the acquisition, expression, extinction, and systems consolidation of memory (Frankland et al. 2004; Santini et al. 2004; Takehara-Nishiuchi et al. 2005; Corcoran and Quirk 2007; Jung et al. 2008). Of these processes, the mechanisms supporting the acquisition of memory may be the least understood. Recently, the medial prefrontal cortex (mPFC) has been shown to be important for trace fear conditioning (Runyan et al. 2004; Gilmartin and McEchron 2005), which provides a powerful model system for studying the neurobiological basis of prefrontal contributions to memory. Trace fear conditioning is a variant of standard “delay” fear conditioning in which a neutral conditional stimulus (CS) is paired with an aversive unconditional stimulus (UCS). Trace conditioning differs from delay conditioning by the addition of a stimulus-free “trace” interval of several seconds separating the CS and UCS. Learning the CS–UCS association across this interval requires forebrain structures such as the hippocampus and mPFC. Importantly, the mPFC and hippocampus are only necessary for learning when a trace interval separates the stimuli (Solomon et al. 1986; Kronforst-Collins and Disterhoft 1998; McEchron et al. 1998; Takehara-Nishiuchi et al. 2005). This forebrain dependence has led to the hypothesis that neural activity in these structures is necessary to bridge the CS–UCS temporal gap. In support of this hypothesis, single neurons recorded from the prelimbic area of the rat mPFC exhibit sustained increases in firing during the CS and trace interval in trace fear conditioning (Baeg et al. 2001; Gilmartin and McEchron 2005). Similar sustained responses are not observed following the CS in delay conditioned animals or unpaired control animals. This pattern of activity is consistent with a working memory or “bridging” role for mPFC in trace fear conditioning, but it is not clear whether this activity is actually necessary for learning. We address this issue here using the γ-aminobutyric acid type A (GABA_A) receptor agonist muscimol to temporarily inactivate cellular activity in the prelimbic mPFC during the acquisition of trace fear conditioning.The contribution of mPFC to the long-term storage (i.e., 24 h or more) of trace fear conditioning, as opposed to a strictly working memory role (i.e., seconds to minutes), is a matter of some debate. Recent reports suggest that intact prefrontal activity at the time of testing is required for the recall of trace fear conditioning 2 d after training (Blum et al. 2006a), while mPFC lesions performed 1 d after training fail to disrupt the memory (Quinn et al. 2008). The findings from the former study may reflect a role for prelimbic mPFC in the expression of conditional fear rather than memory storage per se (Corcoran and Quirk 2007). However, blockade of the intracellular mitogen-activated protein kinase (MAPK) cascade during training impairs the subsequent retention of trace fear conditioning 48 h later (Runyan et al. 2004). Activation of the MAPK signaling cascade can result in the synthesis of proteins necessary for synaptic strengthening, providing a potential mechanism by which mPFC may participate in memory storage. To better understand the nature of the prefrontal contribution to long-term memory, more information is needed about fundamental plasticity mechanisms in this structure. Dependence on N-methyl-d-aspartate receptors (NMDAR) is a key feature of many forms of long-term memory, both in vitro and in vivo. The induction of long-term potentiation (LTP) in the hippocampus, a cellular model of long-term plasticity and information storage, requires NMDAR activation (Reymann et al. 1989). Genetic knockdown or pharmacological blockade of NMDAR-mediated neurotransmission in the hippocampus impairs several forms of hippocampus-dependent memory, including trace fear conditioning (Tonegawa et al. 1996; Huerta et al. 2000; Quinn et al. 2005), but it is unknown if activation of these receptors is necessary in the mPFC for the acquisition of trace fear conditioning. Data from in vivo electrophysiology studies have shown that stimulation of ventral hippocampal inputs to prelimbic neurons in mPFC produces LTP, and the induction of prefrontal LTP depends upon functional NMDARs (Laroche et al. 1990; Jay et al. 1995). If the role of mPFC in trace fear conditioning goes beyond simply maintaining CS information in working memory, then activation of NMDAR may be critical to memory formation. We test this hypothesis by reversibly blocking NMDAR neurotransmission with 2-amino-5-phosphonovaleric acid (APV) during training to examine the role of prefrontal NMDAR to the acquisition of trace fear conditioning.Another important question is whether mPFC contributes to the formation of contextual fear memories. Fear to the training context is acquired simultaneously with fear to the auditory CS in both trace and delay fear conditioning. Conflicting reports in the literature suggest the role of mPFC in contextual fear conditioning is unclear. Damage to ventral areas of mPFC prior to delay fear conditioning has failed to impair context fear acquisition (Morgan et al. 1993). Prefrontal lesions incorporating dorsal mPFC have in some cases been reported to augment fear responses to the context (Morgan and LeDoux 1995), while blockade of NMDAR transmission has impaired contextual fear conditioning (Zhao et al. 2005). Post-training lesions of mPFC impair context fear retention (Quinn et al. 2008) in trace and delay conditioning. Contextual fear responses were assessed in this study to determine the contribution of neuronal activity and NMDAR-mediated signaling in mPFC to the acquisition of contextual fear conditioning.     

The role of amygdala nuclei in the expression of auditory signaled two-way active avoidance in rats

Using a two-way signaled active avoidance (2-AA) learning procedure, where rats were trained in a shuttle box to avoid a footshock signaled by an auditory stimulus, we tested the contributions of the lateral (LA), basal (B), and central (CE) nuclei of the amygdala to the expression of instrumental active avoidance conditioned responses (CRs). Discrete or combined lesions of the LA and B, performed after the rats had reached an asymptotic level of avoidance performance, produced deficits in the CR, whereas CE lesions had minimal effect. Fiber-sparing excitotoxic lesions of the LA/B produced by infusions of N-methyl-d-aspartate (NMDA) also impaired avoidance performance, confirming that neurons in the LA/B are involved in mediating avoidance CRs. In a final series of experiments, bilateral electrolytic lesions of the CE were performed on a subgroup of animals that failed to acquire the avoidance CR after 3 d of training. CE lesions led to an immediate rescue of avoidance learning, suggesting that activity in CE was inhibiting the instrumental CR. Taken together, these results indicate that the LA and B are essential for the performance of a 2-AA response. The CE is not required, and may in fact constrain the instrumental avoidance response by mediating the generation of competing Pavlovian responses, such as freezing.Early studies of the neural basis of fear often employed avoidance conditioning procedures where fear was assessed by measuring instrumental responses that reduced exposure to aversive stimuli (e.g., Weiskrantz 1956; Goddard 1964; Sarter and Markowitsch 1985; Gabriel and Sparenborg 1986). Despite much research, studies of avoidance failed to yield a coherent view of the brain mechanisms of fear. In some studies, a region such as the amygdala would be found to be essential and in other studies would not. In contrast, rapid progress in understanding the neural basis of fear and fear learning was made when researchers turned to the use of Pavlovian fear conditioning (Kapp et al. 1984, 1992; LeDoux et al. 1984; Davis 1992; LeDoux 1992; Cain and Ledoux 2008a). It is now well established from such studies that specific nuclei and subnuclei of the amygdala are essential for the acquisition and storage of Pavlovian associative memories about threatening situations (LeDoux 2000; Fanselow and Gale 2003; Maren 2003; Maren and Quirk 2004; Schafe et al. 2005; Davis 2006).Several factors probably contributed to the fact that Pavlovian conditioning succeeded where avoidance conditioning struggled. First, avoidance conditioning has long been viewed as a two-stage learning process (Mowrer and Lamoreaux 1946; Miller 1948b; McAllister and McAllister 1971; Levis 1989; Cain and LeDoux 2008b). In avoidance learning, the subject initially undergoes Pavlovian conditioning and forms an association between the shock and cues in the apparatus. The shock is an unconditioned stimulus (US) and the cues are conditioned stimuli (CS). Subsequently, the subject learns the instrumental response to avoid the shock. Further, the “fear” aroused by the presence of the CS motivates learning of the instrumental response. Fear reduction associated with successful avoidance has even been proposed to be the event that reinforces avoidance learning (e.g., Miller 1948b; McAllister and McAllister 1971; Cain and LeDoux 2007). Given that Pavlovian conditioning is the initial stage of avoidance conditioning, as well as the source of the “fear” in this paradigm, it would be more constructive to study the brain mechanisms of fear through studies of Pavlovian conditioning rather than through paradigms where Pavlovian and instrumental conditioning are intermixed. Second, avoidance conditioning was studied in a variety of ways, but it was not as well appreciated at the time as it is today; that subtle differences in the way tasks are structured can have dramatic effects on the brain mechanisms required to perform the task. There was also less of an appreciation for the detailed organization of circuits in areas such as the amygdala. Thus, some avoidance studies examined the effects of removal of the entire amygdala or multiple subdivisions (for review, see Sarter and Markowitsch 1985). Finally, fear conditioning studies typically involved a discrete CS, usually a tone, which could be tracked from sensory processing areas of the auditory system to specific amygdala nuclei that process the CS, form the CS–US association, and control the expression of defense responses mediated by specific motor outputs. In contrast, studies of avoidance conditioning often involved diffuse cues, and the instrumental responses used to indirectly measure fear were complex and not easily mapped onto neural circuits.Despite the lack of progress in understanding the neural basis of avoidance responses, this behavioral paradigm has clinical relevance. For example, avoidance behaviors provide an effective means of dealing with fear in anticipation of a harmful event. When information is successfully used to avoid harm, not only is the harmful event prevented, but also the fear arousal, anxiety, and stress associated with such events; (Solomon and Wynne 1954; Kamin et al. 1963). Because avoidance is such a successful strategy to cope with danger, it is used extensively by patients with fear-related disorders to reduce their exposure to fear- or anxiety-provoking situations. Pathological avoidance is, in fact, a hallmark of anxiety disorders: In avoiding fear and anxiety, patients often fail to perform normal daily activities (Mineka and Zinbarg 2006).We are revisiting the circuits of avoidance conditioning from the perspective of having detailed knowledge of the circuit of the first stage of avoidance, Pavlovian conditioning. To most effectively take advantage of Pavlovian conditioning findings, we have designed an avoidance task that uses a tone and a shock. Rats were trained to shuttle back and forth in a runway in order to avoid shock under the direction of a tone. That is, the subjects could avoid a shock if they performed a shuttle response when the tone was on, but received a shock if they stayed in the same place (two-way signaled active avoidance, 2-AA). While the amygdala has been implicated in 2-AA (for review, see Sarter and Markowitsch 1985), the exact amygdala nuclei and their interrelation in a circuit are poorly understood.We focused on the role of amygdala areas that have been studied extensively in fear conditioning: the lateral (LA), basal (B), and central (CE) nuclei. The LA is widely thought to be the locus of plasticity and storage of the CS–US association, and is an essential part of the fear conditioning circuitry. The basal amygdala, which receives inputs from the LA (Pitkänen 2000), is not normally required for the acquisition and expression of fear conditioning (Amorapanth et al. 2000; Nader et al. 2001), although it may contribute under some circumstances (Goosens and Maren 2001; Anglada-Figueroa and Quirk 2005). The B is also required for the use of the CS in the motivation and reinforcement of responses in other aversive instrumental tasks (Killcross et al. 1997; Amorapanth et al. 2000). The CE, through connections to hypothalamic and brainstem areas (Pitkänen 2000), is required for the expression of Pavlovian fear responses (Kapp et al. 1979, 1992; LeDoux et al. 1988; Hitchcock and Davis 1991) but not for the motivation or reinforcement of aversive instrumental responses (Amorapanth et al. 2000; LeDoux et al. 2009). We thus hypothesized that damage to the LA or B, but not to the CE, would interfere with the performance of signaled active avoidance.     

Post-training reversible inactivation of the hippocampus enhances novel object recognition memory

Research on the role of the hippocampus in object recognition memory has produced conflicting results. Previous studies have used permanent hippocampal lesions to assess the requirement for the hippocampus in the object recognition task. However, permanent hippocampal lesions may impact performance through effects on processes besides memory consolidation including acquisition, retrieval, and performance. To overcome this limitation, we used an intrahippocampal injection of the GABA agonist muscimol to reversibly inactivate the hippocampus immediately after training mice in two versions of an object recognition task. We found that the inactivation of the dorsal hippocampus after training impairs object-place recognition memory but enhances novel object recognition (NOR) memory. However, inactivation of the dorsal hippocampus after repeated exposure to the training context did not affect object recognition memory. Our findings suggest that object recognition memory formation does not require the hippocampus and, moreover, that activity in the hippocampus can interfere with the consolidation of object recognition memory when object information encoding occurs in an unfamiliar environment.The medial temporal lobe plays an important role in recognition memory formation, as damage to this brain structure in humans, monkeys, and rodents impairs performance in recognition memory tasks (for review, see Squire et al. 2007). Within the medial temporal lobe, studies have consistently demonstrated that the perirhinal cortex is involved in this form of memory (Brown and Aggleton 2001; Winters and Bussey 2005; Winters et al. 2007, 2008; Balderas et al. 2008). In contrast, the role of the hippocampus in object recognition memory remains a source of debate. Some studies have reported novel object recognition (NOR) impairments in animals with hippocampal lesions (Clark et al. 2000; Broadbent et al. 2004, 2010), yet others have reported no impairments (Winters et al. 2004; Good et al. 2007). Differences in hippocampal lesion size and behavioral procedures among the different studies have been implicated as the source of discrepancy in these findings (Ainge et al. 2006), but previous studies have not examined the consequences of environment familiarity on the hippocampus dependence of object recognition memory.Previous studies addressing the role of the hippocampus in recognition memory relied on permanent, pre-training lesions (Clark et al. 2000; Broadbent et al. 2004; Winters et al. 2004; Good et al. 2007). Permanent lesions inactivate the hippocampus not only during the consolidation phase, but also during habituation, acquisition, and memory retrieval, potentially confounding interpretation of the results. Furthermore, permanent lesion studies require long surgery recovery times during which extrahippocampal changes may emerge to mask or compensate for the loss of hippocampal function. To overcome these problems, we reversibly inactivated the dorsal hippocampus after training mice in two versions of the object recognition task. We infused muscimol, a γ-aminobutyric acid (GABA) receptor type A agonist, into the dorsal hippocampus immediately after training in an object-place recognition task or immediately following training in a NOR task. Consistent with previous studies (Save et al. 1992; Galani et al. 1998; Mumby et al. 2002; Stupien et al. 2003; Aggleton and Brown 2005), we observed that hippocampal inactivation impairs object-place recognition memory. Interestingly, we observed that the degree of contextual familiarity can influence NOR memory formation. We found that when shorter periods of habituation to the experimental environment were used, hippocampal inactivation enhances long-term NOR memory. In contrast, after extended periods of contextual habituation, long-term recognition memory was unaltered by hippocampal inactivation. Together these results suggest that if familiarization with objects occurs at a stage in which the contextual environment is relatively novel, the hippocampus plays an inhibitory role on the consolidation of object recognition memory. Supporting this view, we observed that object recognition memory is unaffected by hippocampal inactivation when initial exploration of the objects occurred in a familiar environment.     

A molecular dissociation between cued and contextual appetitive learning

In appetitive Pavlovian learning, animals learn to associate discrete cues or environmental contexts with rewarding outcomes, and these cues and/or contexts can potentiate an ongoing instrumental response for reward. Although anatomical substrates underlying cued and contextual learning have been proposed, it remains unknown whether specific molecular signaling pathways within the striatum underlie one form of learning or the other. Here, we show that while the striatum-enriched isoform of adenylyl cyclase (AC5) is required for cued appetitive Pavlovian learning, it is not required for contextual appetitive learning. Mice lacking AC5 (AC5KO) could not learn an appetitive Pavlovian learning task in which a discrete signal light predicted reward delivery, yet they could form associations between context and either natural or drug reward, which could in turn elicit Pavlovian approach behavior. However, unlike wild-type (WT) mice, AC5KO mice could not use these Pavlovian conditioned stimuli to potentiate ongoing instrumental behavior in a Pavlovian-to-instrumental transfer paradigm. These data suggest that AC5 is specifically required for learning associations between discrete cues and outcomes in which the temporal relationship between conditioned stimulus (CS) and unconditioned stimulus (US) is essential, while alternative signaling mechanisms may underlie the formation of associations between context and reward. In addition, loss of AC5 compromises the ability of both contextual and discrete cues to modulate instrumental behavior.In Pavlovian learning, animals form associations between discrete or contextual stimuli in their environment to shape their behavior and make appropriate responses. In discrete cue appetitive Pavlovian conditioning, a single cue with a defined onset and offset that typically activates one sensory modality is provided, immediately followed by reward delivery (Hall 2002; Domjan 2006; Ito et al. 2006). Alternatively, behavior can be driven by context, an assortment of stimuli activating a number of sensory modalities that contribute to the representation of environmental space (Balsam 1985; Rudy and Sutherland 1995; Smith and Mizumori 2006). Collectively, these stimuli make up a context that is paired with reward delivery in contextual appetitive learning. One important distinction between these two forms of learning is that in cued conditioning, there is a discrete temporal relationship between conditioned stimulus (CS) and unconditioned stimulus (US). Thus, an animal can effectively anticipate timing of reward delivery from onset and offset of CS. In vivo studies of dopamine (DA) neuron activity have suggested this discrete temporal relationship can be encoded by DA neurons (Schultz et al. 1997; Schultz 1998a). In contrast, in many contextual Pavlovian conditioning tasks, US delivery is not predicted, it is delivered as the animal explores the environment; thus, the temporal relationship between contextual stimuli and reinforcement is not an essential component of the learned associations (Fanselow 2000). These two types of environmental stimuli may be encoded differently and mediated by different neural substrates.Lesion studies have elucidated the anatomical dissociations between cued and contextual appetitive learning. Using a modified Y-maze procedure, it has been suggested that contextual appetitive learning is hippocampus- and nucleus-accumbens (NAc) dependent, while cued learning is dependent on the basolateral nucleus of the amygdala (BLA) and the NAc (Ito et al. 2005, 2006). In addition, as the NAc processes glutamatergic inputs from the amygdala and the hippocampus (Groenewegen et al. 1999; Goto and Grace 2008), recent studies have indicated that disconnecting the hippocampus from the NAc shell can disrupt contextual appetitive conditioning (Ito et al. 2008). In addition to glutamatergic inputs, the NAc, as part of the ventral striatum, receives dense dopaminergic input from midbrain nuclei (Groenewegen et al. 1999). Temporal shifts in phasic DA release in striatal regions has been correlated with appetitive Pavlovian learning (Day et al. 2007), and models of striatal function suggest that DA-dependent modification of glutamatergic transmission in the striatum may underlie reinforcement learning (Reynolds et al. 2001; Reynolds and Wickens 2002).The cAMP pathway has been implicated in plasticity and learning in a number of neuronal structures (Abel et al. 1997; Ferguson and Storm 2004; Pittenger et al. 2006). Adenylyl cyclase (AC), the enzyme that makes cAMP, has nine membrane-bound isoforms, each with different expression patterns and regulatory properties (Hanoune and Defer 2001). AC5 is highly enriched in the striatum, with very low levels of expression in other regions of the brain (Mons et al. 1998; Iwamoto et al. 2003; Kheirbek et al. 2008, 2009), and genetic deletion of AC5 (AC5KO) severely compromises DA''s ability to modulate cAMP levels in the striatum (Iwamoto et al. 2003). Previous studies have shown that AC5KO mice were severely impaired in acquisition of a cued appetitive Pavlovian learning task, while formation of action–outcome contingencies in instrumental learning was intact (Kheirbek et al. 2008). Yet, it remains unknown whether the cAMP pathway in the striatum underlies all forms of appetitive Pavlovian learning, or how it contributes to the ability of Pavlovian cues to modulate instrumental behavior.In this study, we asked if genetic deletion of AC5 selectively impairs cued or contextual appetitive learning. In addition, we tested whether loss of AC5 affects the ability of conditioned cues or contexts to modulate instrumental behavior. Our data indicate that although loss of AC5 abolishes cued appetitive learning, contextual learning is spared. Although contextual stimuli could elicit approach behavior in AC5KO mice, they could not potentiate an ongoing instrumental response, highlighting the importance of this isoform of AC in Pavlovian–instrumental interactions.     

Pre-exposure to context affects learning strategy selection in mice

The multiple memory systems hypothesis proposes that different types of learning strategies are mediated by distinct neural systems in the brain. Male and female mice were tested on a water plus-maze task that could be solved by either a place or response strategy. One group of mice was pre-exposed to the same context as training and testing (PTC) and the other group was pre-exposed to a different context (PDC). Our results show that the PTC condition biased mice to place strategy use in males, but this bias was dependent on the presence of ovarian hormones in females.The participation of different brain areas in place and response learning strategies has been studied extensively (White and McDonald 2002; Gold 2004; Mizumori et al. 2004). Place strategy is an allocentric navigation strategy that depends on extramaze cues. Response strategy is an egocentric navigation strategy based on proprioceptive cues. Inactivation of the hippocampus biased animals to response strategy use, and inactivation of the striatum biased animals to place strategy use (Packard and McGaugh 1996; Lee et al. 2008). Furthermore, glutamate infusion into the hippocampus strengthened place strategy use and, conversely, glutamate infusion into the striatum enhanced response strategy use (Packard 1999). These studies suggest that the hippocampus system mediates place strategy, while the striatum system mediates response strategy.Various factors can modulate learning strategy use, including training intensity (Packard and McGaugh 1996; Martel et al. 2007). A recent study investigated the influence of training on strategy use on a probe trial conducted 1 h after training (Martel et al. 2007). Male mice displayed enhanced place strategy use when trained on 12 or 22 trials compared with four trials, suggesting an effect of training intensity on strategy choice (Martel et al. 2007). This study further investigated the effect of pre-exposure to the training and testing context (PTC). Pre-exposure enhanced place strategy use in male mice after only four trials relative to animals pre-exposed to a different context (PDC). These results suggest that a sufficient exposure to the training and testing context promotes place strategy use in mice.The type of strategy used by rats is affected by both biological sex and gonadal steroids. Male rats typically employ a place strategy, especially during the early phase of training, on both land and water T-mazes (Packard and McGaugh 1992, 1996; Packard and Teather 1997). However, strategy use by female rats depends on hormonal conditions (Dohanich 2002; Dohanich et al. 2009). Place strategy is preferred by intact female rats on the day of proestrus when estradiol levels are elevated, and by ovariectomized rats treated with estradiol (Korol and Kolo 2002; Korol 2004; Korol et al. 2004). In contrast, response strategy is more often displayed by intact females on diestrus, and by ovariectomized females that did not receive estradiol replacement (Korol and Kolo 2002; Korol et al. 2004). To date, the effects of biological sex and gonadal steroids on learning strategy have not been studied in mice.In this study, we developed a modified version of the dual-solution water plus-maze task to further investigate the role of PTC compared with PDC in male and female mice. We hypothesized that strategy choice in both sexes would be dependent on context pre-exposure, and ovarian hormones would influence strategy choice in females. Our results show that PTC significantly enhanced place strategy use in male mice. Although there was no significant difference between PTC and PDC female mice, ovariectomy significantly reduced place strategy use in the PTC females, suggesting that ovarian hormones play a significant role in strategy use in female mice.Sixteen male and 39 female 129/Sve strain mice were obtained at 2–3 mo of age from Charles River Laboratories (Boston, MA). Mice were housed in groups of four on a 12/12 light/dark cycle (lights on at 07:00 h) with free access to food and water. All protocols followed the guidelines from a protocol approved by the Animal Care and Use Committee of Tulane University in accordance with National Institutes of Health Guide for the Care and Use of Laboratory Animals.Mice were pre-exposed for 5 min to the dry plus-maze either in the context of the subsequent training and testing (PTC), or in a different context in a different room (PDC), 30 min prior to the first training trial. The maze consisted of four clear Plexiglas arms (40 cm in length, 10 cm in width, and 40 cm in height). During the pre-exposure, mice were able to visit three arms of the maze. The rooms had different visual cues surrounding the maze. No extramaze cues were placed directly at the end of any arm. After the pre-exposure, the animal was placed in its home cage. The maze was wiped clean with 70% ethanol between trials.After pre-exposures, the maze was filled to 1.5 cm above the Plexiglas escape platform (15 cm in height) with room-temperature water colored opaque with white nontoxic tempera paint. Mice were trained in the water plus-maze task (Fig. 1A). The training was ended when the animals made six correct choices or reached nine trials. The animals that made fewer than four correct choices during training were not included in the study. Trials were continued until the mouse reached the platform or a maximum of 1 min. Each trial was separated by an intertrial interval of 4 min. Throughout the training trials, one arm (north) was blocked off by a white Plexiglas shield, creating a T-shaped maze. Mice were placed in the start arm of the maze (south) and were allowed to swim to the escape platform, which was consistently located in one arm of the maze for each animal and alternated between animals (east or west). Entry of the entire animal into the maze arm that contained the escape platform was scored as a correct response during the training trials, and entry of the entire animal into the maze arm that did not contain the escape platform was scored as an incorrect response. Mice were allowed to remain on the escape platform for 15 sec before being returned to their cages. Mice that failed to find the escape platform within 60 sec were manually guided to the platform. The water was distributed across all arms of the maze and the maze walls were wiped down to reduce intramaze cues between training and probe trials. One hour after training, mice were tested on a probe trial (Fig. 1B) in order to determine their relative use of “place” and “response” strategy. On the probe trial, mice were placed into the start arm 180° opposite the start arm used during training (i.e., end of the north arm) and were allowed to make an entry into either the east or west maze arm. The white Plexiglas shield blocked the south arm during the probe trial. Mice were designated as using place or response strategy based on the probe trial. Place strategy was designated as entry of the entire animal into the arm with the platform, and response strategy was designated as entry of the entire animal into the opposite arm.Open in a separate windowFigure 1.The effects of pre-exposure to the training and testing context (PTC) or to a different context (PDC) on strategy selection of male mice. (A) Schematic diagram of the water plus-maze. Mice were released from the south arm during training trials and from the north arm during the probe test. (B) More male mice used place strategy than response strategy when pre-exposed to the same context prior to training and testing (PTC, n = 5) compared with male mice pre-exposed to a different context (PDC, n = 7, P < 0.05). (C) Latency curves show the actual latency to escape to the platform. Two-way ANOVA (non-repeated measures) revealed no significant difference across training trials in escape latencies between PDC and PTC mice (P > 0.5), although a significant effect of trial indicated that mice reduced their escape latencies across trials (P < 0.001). Values represent mean ± S.E.M.Sixteen male mice were randomly divided into two groups based on pre-exposure context, PTC or PDC. Four of the 16 males were not included in the study for failure to reach criterion (four correct out of nine trials) or failure to escape to the platform due to floating, which is a behavior commonly seen in this strain (Wolfer et al. 1997). On the probe trial PTC males used the place strategy significantly more often than PDC males (P < 0.05, χ² = 5.182, Fig. 1B). Four of five PTC males used place strategy, whereas only one of seven PDC males used place strategy. Pre-exposure of animals to the same or different context prior to training did not affect the latency to escape the platform during training. Latency to find the platform during training trials revealed a significant effect of trial (F_(8,89) = 3.830, P = 0.0007, non-repeated measures two-way ANOVA) but not pre-exposure condition (F_(1,89) = 0.103, P = 0.75, non-repeated measures two-way ANOVA; Fig. 1C). Moreover, the average swim speed of PDC male mice (6 ± 1.6 cm/sec, n = 7) was not significantly different than the average swim speed of PTC male mice (6 ± 2.5 cm/sec, n = 5; P = 0.34, t = 0.9 [t-test]). Together, these data suggest that the pre-exposure condition did not influence learning during the training period, but PTC did enhance place strategy use in the probe trial in male mice.Female mice at 3 mo of age were randomly divided into two groups: mice that would receive ovariectomy (Ovx), and a sham surgery group (Sh). Mice were anesthetized with a ketamine (80 mg/kg) and xylazine (8 mg/kg) mixture. The first group of mice (n = 20) received ovariectomy using a dorsolateral approach. The other group (n = 19) of female mice received sham surgery, which consisted of ovary exposure only. Animals were injected with the pain reliever, buprenorphine (5 mg/kg), immediately after the surgery. One week after the surgery, vaginal smears were collected from all females, including Ovx as handled controls, at the same time each morning by lavage to track their estrus cycles (Marcondes et al. 2002). After two regular cycles, Sh animals were trained and tested on the day of proestrus (high estradiol).Ovariectomy has been reported to affect anxiety levels (Walf et al. 2006), and anxiety levels may alter performance on water maze tasks. To assay possible anxiety differences between Sh and Ovx, female mice were tested on open field and elevated plus-maze (EPM) 2 wk after the surgery in a room different from the rooms used in water maze tasks. A single mouse was placed in the center of a white, Plexiglas chamber measuring 43 cm in length × 43 cm in width × 18 cm in height. The animal explored the novel environment for 15 min, and movements were monitored by a camera interfaced with a tracking system (US HVS Image). The area was divided into 16 virtual squares (10.75 × 10.75 cm) by the program, and the middle four squares were defined as the center area. The Plexiglas chamber was wiped clean with 70% ethanol between trials. The EPM consisted of four arms (5 cm in width × 30 cm in length) arranged perpendicularly in a plus shape and elevated 38 cm above the floor. Two arms were enclosed by 15.5-cm dark Plexiglas walls and two arms were open. Each animal was placed in the center of the EPM facing a closed arm and allowed to move freely for 5 min. Behavior was monitored by a camera interfaced with the tracking system.Animals with high anxiety levels tend to spend less time in the open arms of the EPM and in the center of the open field. The percent time spent in the open arms of the EPM by Ovx mice (37.9% ± 7.5%, n = 14) was not significantly different than the percent of time spent in the open arms by Sh mice (27.8% ± 6.2%, n = 15; P = 0.30, t = 1.1). The percent time spent in the center of the open field by Ovx mice (35.1% ± 7.1%, n = 14) was not significantly different from Sh mice (29.0% ± 7.1%, n = 15; P = 0.55, t = 0.61). These results indicate that ovarian hormones did not have a significant effect on the anxiety levels of the female mice tested in this study.Two weeks after the anxiety tests, the Ovx and Sh groups were divided randomly into two groups based on the pre-exposure context: Ovx PTC, Ovx PDC, Sh PTC, Sh PDC. Sh females with regular estrus cycles were trained and tested on the day of proestrus. Five Ovx and seven Sh animals were not included in the study because of floating, failing to reach criterion (four correct out of nine trials), or exhibiting irregular estrus cycles. Five of eight Sh PTC and only one of six Sh PDC females used place strategy; however, this difference was not significant (P > 0.05, χ² = 2.94, Fig. 2A). Therefore, the pre-exposure condition did not significantly affect strategy use in females at proestrus.Open in a separate windowFigure 2.The effects of ovarian hormone status and pre-exposure to the training and testing (PTC) or to a different context (PDC) on strategy selection of female mice. (A) When pre-exposed to the same context prior to training and testing (PTC), more gonadally intact female mice at proestrus (Sh, n = 8) used place strategy than response strategy compared with ovariectomized female mice (Ovx, n = 8, P < 0.05). When pre-exposed to a context different than the training and testing context (PDC), gonadally intact female mice at proestrus (Sh, n = 6) and ovariectomized mice (Ovx, n = 6) used response strategy rather than place strategy. (B) Latency curves show the actual latency to escape to the platform. Two-way ANOVA (non-repeated measures) revealed no significant differences across training trials in escape latencies between sham and ovariectomized PTC and PDC mice (P > 0.5), although a significant effect of trial indicated that mice reduced their escape latencies across trials (P < 0.0001). Values represent mean ± S.E.M.Interestingly, ovariectomy did significantly affect strategy use in PTC females. Five of eight Sh PTC and only one of eight Ovx PTC females used place strategy (P < 0.05, χ² = 4.267, Fig. 2A). One of six Sh PDC females and zero of the six Ovx PDC animals used place strategy (Fig. 2A). Therefore, both Sh and Ovx PDC females used response strategy, and ovarian hormones did not enhance place strategy use in PDC females (P > 0.05, χ² = 1.09, Fig. 2A). Ovarian hormones did enhance place strategy use in PTC females. Furthermore, PTC did not enhance place strategy use in Ovx animals. Similar to males, there was a significant effect of training trial on latency to find the platform in female animals (F_(8,189) = 10.32, P < 0.0001, Fig. 2B). Ovarian hormones or pre-exposure to either context also did not affect escape latency during training in PTC or PDC females (F_(3,189) = 0.33, P = 0.80, Fig. 2B). In addition, there was no significant difference in the average swim speed between groups (F_(3,14) = 0.15, P = 0.93, one-way ANOVA). The average swim speed for each group was as follows: Ovx PTC (5 ± 1.5 cm/sec, n = 5), Ovx PDC (5 ± 1.0 cm/sec, n = 4), Sh PTC (5 ± 1.8 cm/sec, n = 5), Sh PDC (6 ± 1.5 cm/sec, n = 4). The numbers of animals are lower because in some cases, speed was not measured. Together, these data suggest that ovarian hormones and pre-exposure condition did not influence learning during the training period, but ovarian hormones did enhance place strategy use in the probe trial in only PTC mice.Consistent with previous literature (Martel et al. 2007), we found that ∼80% of PTC males favored the use of place strategy. In addition, 63% of PTC females on proestrus also used place strategy. Ovx female mice used response strategy regardless of the pre-exposure condition. These results confirm that pre-exposure to the training and testing context significantly increased the use of place strategy or reduced response strategy in male mice, while female mice on proestrus were not significantly different than chance. Ovariectomy diminished the use of place strategy and enhanced response strategy use in our study, implicating ovarian hormones in strategy choice.Male rats rely initially on a hippocampus-dependent place strategy, and then switch to a striatum-based response strategy over training (Packard and McGaugh 1996; Packard 1999). This suggests that response strategy is incrementally learned with repeated exposure to the same task. However, a sufficient amount of time to explore the extramaze cues during or before training increased place strategy use in male mice (Martel et al. 2007). In addition, it has been proposed that the presence of an increased number of salient extramaze cues favors place strategy use in rats (Restle 1957). Therefore, it is possible that pre-exposing mice to the learning environment allowed them to build a cognitive map that facilitated the use of a spatial place strategy. Another possible advantage of pre-exposure for place strategy use is that it may reduce the impact of non-mnemonic factors, such as anxiety, on performance (Cain 1998). Indeed, it was shown that peripheral injection and infusion of anxiogenic drugs into the basolateral amygdala biased rats toward the use of response strategy (Packard 1999; Wingard and Packard 2008; Packard and Gabriele 2009).While PTC female mice were not significantly different than PDC female mice, ovariectomy did reduce place strategy choice in the PTC mice. An emerging theory proposes that estradiol modulates cognitive performance via shifts in learning strategy (Korol and Kolo 2002; Daniel and Lee 2004; Korol 2004; McElroy and Korol 2005; Zurkovsky et al. 2007). Shifts in strategy use occurred across the estrus cycle in rats such that the hippocampus-dependent strategy was favored when estradiol levels were high (Korol et al. 2004). Similarly, estradiol treatment in ovariectomized rats increased hippocampus-dependent place strategy and impaired response strategy use compared with nontreated ovariectomized females (Korol and Kolo 2002). Our results showing that the lack of ovarian hormones reduced place strategy and increased response strategy use in PTC mice are consistent with these studies.In summary, we present a new design to a traditional dual-solution land plus-maze. One issue with the land maze version of the task is that it requires food deprivation. The possible increase in the appetite as a result of ovariectomy (Wade 1975) or disruption in the estrus cycle in response to food deprivation (Daniel et al. 1999) could confound the results in females in tasks that present food reward. In order to avoid these confounds, we used a modified version of a water-escape plus-maze (Packard and Wingard 2004). In this design, compared with the water-escape plus-maze, the clear Plexiglas maze itself is filled with water, instead of placing the plus-maze into a water maze, allowing a better view of extramaze visual cues. However, unlike rats, mice tend to be prey animals when in the water; therefore they are highly motivated to escape the water (Francis et al. 1995; Van Dam et al. 2006). Consequently, the stressful nature of the task prevents mice from utilizing the spatial cues as efficiently (Frick et al. 2000). Therefore, we pre-exposed the mice to the maze while it was dry, allowing them to build a cognitive map before they were released in water. The water plus-maze is important not only for the design of future studies, but also for the evaluation of previous studies that investigated learning strategies using tasks dependent on food deprivation.     

Entorhinal cortical Island cells regulate temporal association learning with long trace period

Temporal association learning (TAL) allows for the linkage of distinct, nonsynchronous events across a period of time. This function is driven by neural interactions in the entorhinal cortical–hippocampal network, especially the neural input from the pyramidal cells in layer III of medial entorhinal cortex (MECIII) to hippocampal CA1 is crucial for TAL. Successful TAL depends on the strength of event stimuli and the duration of the temporal gap between events. Whereas it has been demonstrated that the neural input from pyramidal cells in layer II of MEC, referred to as Island cells, to inhibitory neurons in dorsal hippocampal CA1 controls TAL when the strength of event stimuli is weak, it remains unknown whether Island cells regulate TAL with long trace periods as well. To understand the role of Island cells in regulating the duration of the learnable trace period in TAL, we used Pavlovian trace fear conditioning (TFC) with a 60-sec long trace period (long trace fear conditioning [L-TFC]) coupled with optogenetic and chemogenetic neural activity manipulations as well as cell type-specific neural ablation. We found that ablation of Island cells in MECII partially increases L-TFC performance. Chemogenetic manipulation of Island cells causes differential effectiveness in Island cell activity and leads to a circuit imbalance that disrupts L-TFC. However, optogenetic terminal inhibition of Island cell input to dorsal hippocampal CA1 during the temporal association period allows for long trace intervals to be learned in TFC. These results demonstrate that Island cells have a critical role in regulating the duration of time bridgeable between associated events in TAL.

The linkage of temporally discontiguous events, called temporal association learning (TAL), is an essential function for episodic memory formation; for animals, when an event took place, and in what order a series of events occurred is directly linked to adaptation to continuous changes in the environment (Eichenbaum 2000; Tulving 2002a,b; Kitamura et al. 2015a; Kitamura 2017; Pilkiw and Takehara-Nishiuchi 2018). The entorhinal cortical–hippocampal (EC-HPC) network in particular is currently considered to bridge the temporal discontinuity between events (Solomon et al. 1986; Moyer et al. 1990; Wallenstein et al. 1998; McEchron et al. 1999; Eichenbaum 2000; Huerta et al. 2000; Ryou et al. 2001; Takehara et al. 2003; Chowdhury et al. 2005; Esclassan et al. 2009; Morrissey et al. 2012; Suter et al. 2013; Sellami et al. 2017; Wilmot et al. 2019).Two major excitatory inputs to HPC arise from the superficial layers of the EC (Fig. 1A), forming the direct (monosynaptic), and indirect (trisynaptic) pathways (Amaral and Witter 1989; Amaral and Lavenex 2007; Kitamura 2017; Kitamura et al. 2017). While pyramidal cells in EC layer III (ECIII cells) project directly to CA1 (Kohara et al. 2014; Kitamura et al. 2015b), the trisynaptic pathway originates from excitatory Reelin⁺ stellate cells in EC layer II (ECII) projecting directly to DG, CA3, and CA2 (Fig. 1B; Tamamaki and Nojyo 1993; Varga et al. 2010). CalbindinD-28K⁺/Wolfram syndrome 1 (Wfs1)⁺ pyramidal cells, another excitatory neural population in EC layer II called “Island cells,” form cell clusters along the ECII/ECI border (Alonso and Klink 1993; Fujimaru and Kosaka 1996; Klink and Alonso 1997; Kawano et al. 2009; Varga et al. 2010; Kitamura et al. 2014; Ray et al. 2014) and directly project to the GABAergic interneurons of stratum lacunosum (SL-INs) in HPC CA1 and drive feedforward inhibition to HPC CA1 pyramidal cells (Fig. 1B; Kitamura et al. 2014; Surmeli et al. 2016; Kitamura 2017; Ohara et al. 2018; Yang et al. 2018; Zutshi et al. 2018).Open in a separate windowFigure 1.Circuit schematic diagram of the medial entorhinal cortex (MEC)–hippocampal (HPC) circuit. (A) Major projections in the entorhinal cortical (EC)-HPC network. ECIII neurons (green) project directly to CA1. ECII Ocean cells (ECIIo, purple) project to the dentate gyrus (DG) (light blue)/CA3 (pink) initiating the trisynaptic pathway. ECII Island cells (ECIIi, blue) project directly into CA1. (B) ECIII projections (green) excite the distal portions of CA1 pyramidal cell (yellow) dendrites in the stratum moleculare. Island cells (ECIIi, blue) excite the interneurons of stratum lacunosum (SL-INs, red), which in turn inhibit the distal dendrites of CA1 pyramidal cells in SL.Trace fear conditioning (TFC) has been established as one suitable animal model for TAL (Fendt and Fanselow 1999; Maren 2001; Kim and Jung 2006) that can be also used as a translational bridge between animal and human learning (Clark and Squire 1998; Buchel and Dolan 2000; Delgado et al. 2006). Lesion, pharmacological, molecular, and optogenetic manipulation, as well as disease models in medial entorhinal cortex (MEC), demonstrate that MEC is crucial for TFC and temporal learning (Ryou et al. 2001; Woodruff-Pak 2001; Runyan et al. 2004; Esclassan et al. 2009; Gilmartin and Helmstetter 2010; Suh et al. 2011; Morrissey et al. 2012; Shu et al. 2016; Hales et al. 2018; Yang et al. 2018; Heys et al. 2020). Specifically, MECIII inputs into the HPC CA1 pyramidal cells are essential for the formation of TFC (Yoshida et al. 2008; Suh et al. 2011; Kitamura et al. 2014; Kitamura 2017). However, the temporal association function driven by MECIII neurons must be regulated for optimal adaptive memory formation, as too strong an association of a particular pair of events may interfere with associations of other useful pairs, whereas too weak an association for a given pair of events, in terms of weaker impact of events or longer duration of temporal gap between events, would not result in an effective memory (Kitamura et al. 2015a; Marks et al. 2020). In a naturalistic context, this would mean that more distant/quieter sounds, less intense somatic sensations (e.g., pain), or increased temporal distance between any two events would signal that the events are less likely to be causally associated, therefore less relevant, and less likely to be stored and recalled. In fact, successful TFC depends on the strength of event stimuli and duration of temporal gap between events (Stiedl and Spiess 1997; Misane et al. 2005; Kitamura et al. 2014; Kitamura 2017). However, the underlying regulatory mechanism for TAL remains hidden. Previously we demonstrated that feedforward inhibition by Island cells acts as a gating controller for the MECIII inputs to the distal dendrites of HPC CA1 pyramidal cells in stratum moleculare (SM) (Kitamura et al. 2014) to control TFC when weaker (in this case diminished footshock intensity) unconditioned stimuli were delivered for TFC, indicating that Island cell activity controls the temporal association when the strength of two discontinuous events are relatively weaker. However, the way in which the EC-HPC network regulates TFC with a longer trace period still remains unknown. Because the activation of Island cells would result in a net inhibitory effect on the local network in CA1, imposing a tight and specific regulation on associations of events across the temporal gap in TAL (Crestani et al. 2002; Moore et al. 2010; Kitamura et al. 2014, 2015b), we hypothesized that the length of the temporal gap between events would also be modulated by this mechanism. In this study, we examined the role of the regulatory input to this circuit arising specifically from the Island cells in the MECII using apoptotic elimination of Island cells, chemogenetic neural inhibition, and optogenetic terminal inhibition methods within an L-TFC protocol to give a thorough and complete assessment of the circuit involvement while considering each technique''s unique features.     

A brief nap is beneficial for human route-learning: The role of navigation experience and EEG spectral power

Here, we examined the effect of a daytime nap on changes in virtual maze performance across a single day. Participants either took a short nap or remained awake following training on a virtual maze task. Post-training sleep provided a clear performance benefit at later retest, but only for those participants with prior experience navigating in a three-dimensional (3D) environment. Performance improvements in experienced players were correlated with delta-rich stage 2 sleep. Complementing observations that learning-related brain activity is reiterated during post-navigation NREM sleep in rodents, the present data demonstrate that NREM sleep confers a performance advantage for spatial memory in humans.A growing body of animal and human literature suggests that the consolidation of memories occurs optimally during periods of post-learning sleep. Nonrapid eye movement sleep (NREM), in particular, may be beneficial for the offline consolidation of hippocampus-dependent learning. The neurophysiological basis for this hypothesis is derived largely from electrophysiological studies in rodents, demonstrating that patterns of hippocampal place cell activity first seen during waking exploration are later reexpressed during post-learning sleep (Wilson and McNaughton 1994; Kudrimoti et al. 1999; Nadasdy et al. 1999; Ji and Wilson 2007). Behavioral studies in humans meanwhile demonstrate that NREM sleep is beneficial for declarative memory performance, relative to equivalent periods of wakefulness (Plihal and Born 1997; Tucker et al. 2006). However, the memory tasks typically employed in human research are quite different from those used in rodents, with human studies most often focusing on the memorization of verbal or visual stimuli (Plihal and Born 1997; Schabus et al. 2004; Clemens et al. 2005; Ellenbogen et al. 2006; Tucker et al. 2006; Daurat et al. 2008). Thus far, sleep-dependent memory reactivation has not been established to be directly beneficial for memory performance in an animal model, as the protocols employed in this research typically involve well-learned simple tasks which do not easily lend themselves to measurement of learning across time (Wilson and McNaughton 1994; Kudrimoti et al. 1999). Although the hippocampal memory reactivation described in rodents is a possible explanation for the effect of NREM sleep on human declarative memory, widely divergent methodologies employed across species prohibit confidence in this conclusion.Bridging this conceptual gap, a small handful of studies have begun to explore the relationship between spatial navigation and NREM sleep in humans. Notably, a PET study by Peigneux et al. (2004) demonstrated that learning-related hippocampal activity seen while training on a virtual maze task is again expressed during post-learning human sleep. Furthermore, this hippocampal reactivation strongly predicted overnight improvement on the task (Peigneux et al. 2004). Additional studies have suggested a link between sleep and other types of spatial-related learning, including mental rotation performance (Plihal and Born 1999), the ability to reproduce a complex figure (Clemens et al. 2006; Tucker and Fishbein 2008), performance on a computerized version of Milner''s (1965) “bolt head” maze (Tucker and Fishbein 2008), and memory for the location of verbal information on a screen (Daurat et al. 2008).Yet it remains unclear whether sleep, relative to wakefulness, provides a performance benefit for human route-learning in the context of a realistic spatial environment. Navigation through virtual environments is a strongly hippocampus-dependent task (Peigneux et al. 2004; Astur et al. 2005) and provides an experimental model closely paralleling the spatial exploration tasks employed in the rodent literature. However, the few studies reporting effects of sleep on human navigation performance have been contradictory. Using a navigation task similar to that of Peingeux et al. (2004), Orban et al. (2006) failed to detect any effect of post-learning sleep deprivation on maze performance but did find evidence of altered task-related brain activity, concluding that sleep supports “covert” memory reorganization (Orban et al. 2006). In direct contrast, Ferrara et al. found that spatial memory is improved when a retention interval falls across a night of sleep, relative to when route memory must be retained during daytime wakefulness, or across a night of sleep deprivation (Ferrara et al. 2006, 2008).The present study clarifies these issues by examining the effect of a post-learning nap on complex route-learning in a three-dimensional (3D) virtual environment. When controls are tested at a different time of day than sleep participants, circadian confounds may present a substantial problem. Alternatively, overnight protocols employing sleep-deprived subjects necessarily suffer from confounds related to this sleep deprivation during the retention interval. The use of a daytime nap as a sleep intervention avoids these pitfalls by allowing all subjects to be trained and tested at the same circadian time, and in the absence of sleep deprivation. A series of recent studies confirm that a daytime nap is sufficient to induce performance improvements on declarative and procedural memory tasks, relative to wake subjects (Mednick et al. 2003; Backhaus and Junghanns 2006; Nishida and Walker 2006; Tucker et al. 2006; Lahl et al. 2008; Tucker and Fishbein 2008).Participants (n = 53, 34 female) were trained on a virtual maze-learning task at 12:30 pm. Following training, nap participants lay down for a 1.5-h sleep opportunity. These subjects were allowed to obtain as much NREM sleep as possible but were awoken at the first signs of REM (see Table 

Time-dependent transformations of memory representations differ along the long axis of the hippocampus

Research has shown that sleep is beneficial for the long-term retention of memories. According to theories of memory consolidation, memories are gradually reorganized, becoming supported by widespread, distributed cortical networks, particularly during postencoding periods of sleep. However, the effects of sleep on the organization of memories in the hippocampus itself remains less clear. In a 3-d study, participants encoded separate lists of word–image pairs differing in their opportunity for sleep-dependent consolidation. Pairs were initially studied either before or after an overnight sleep period, and were then restudied in a functional magnetic resonance imaging (fMRI) scan session. We used multivariate pattern similarity analyses to examine fine-grained effects of consolidation on memory representations in the hippocampus. We provide evidence for a dissociation along the long axis of the hippocampus that emerges with consolidation, such that representational patterns for object–word memories initially formed prior to sleep become differentiated in anterior hippocampus and more similar, or overlapping, in posterior hippocampus. Differentiation in anterior hippocampal representations correlated with subsequent behavioral performance. Furthermore, representational overlap in posterior hippocampus correlated with the duration of intervening slow wave sleep. Together, these results demonstrate that sleep-dependent consolidation promotes the reorganization of memory traces along the long axis of the hippocampus.

The hippocampus has long been considered critical for encoding new memories; however, the effects of consolidation on hippocampal memory traces has remained an area of active research. Memories are thought to be stabilized for the long term as they become distributed across neocortical networks (Buzsáki 1989; Alvarez and Squire 1994; McClelland et al. 1995), a process supported by mechanisms during sleep (Diekelmann and Born 2010; Rasch and Born 2013). Whereas much research has been devoted to understanding the hippocampal contributions to the long-term retention of memories, open questions remain in considering how sleep-dependent consolidation affects the organization of hippocampal traces.The hippocampus has previously been shown to be critical for binding disparate elements of an experience together (Cohen and Eichenbaum 1993; Davachi 2006). Theories suggest that the hippocampus quickly encodes new experiences, while the cortex, with a slower learning rate, gradually comes to represent the central features from this hippocampal trace, resulting in abstracted memories that can be integrated into long-term cortical stores (McClelland et al. 1995). Prior research has demonstrated evidence for a coordinated hippocampal–cortical dialogue during sleep (Andrade et al. 2011; Bergmann et al. 2012; Ngo et al. 2020) as well as enhanced hippocampal–cortical functional connectivity after learning, facilitating the retention of memories (Tambini et al. 2010; Tompary et al. 2015; Murty et al. 2017; Cowan et al. 2021). Reports suggest consolidation results in more integrated cortical memory traces in the cortex (Richards et al. 2014; Tompary and Davachi 2017; Cowan et al. 2020); however, it remains an open question whether the active consolidation processes that support memory reorganization across hippocampal–cortical networks also transform hippocampal memory traces.Research on the fate of the hippocampal trace with consolidation has often focused on questions about the permanence of memories in the hippocampus. Theories of systems consolidation have classically debated whether the hippocampal trace is time-limited (Alvarez and Squire 1994), or, rather, whether the hippocampus continues to represent memories in perpetuity (Nadel and Moscovitch 1997; Winocur and Moscovitch 2011; Moscovitch et al. 2016; Sekeres et al. 2018a). Another theory posits that while the original hippocampal trace is transient, during retrieval the hippocampus reconstructs details of an experience from cortical traces (Barry and Maguire 2019). Much research in this vein has focused on investigating changes in hippocampal blood-oxygenation level-dependent (BOLD) univariate activation with time (Bosshardt et al. 2005a,b; Takashima et al. 2006, 2009; Gais et al. 2007; Sterpenich et al. 2007, 2009; Yamashita et al. 2009; Milton et al. 2011; Watanabe et al. 2012; Ritchey et al. 2015; Baran et al. 2016; Dandolo and Schwabe 2018) and the effects of hippocampal lesions in animals and humans (Winocur et al. 2001; Frankland and Bontempi 2005; Winocur and Moscovitch 2011; Moscovitch et al. 2016) with mixed results. Interestingly, pinpointing these effects along the long axis of the hippocampus has also proven unclear. Some reports have found that only the anterior hippocampus exhibits time-dependent changes in retrieval-related univariate activation, with evidence of decreases with delay (Takashima et al. 2006; Milton et al. 2011; Dandolo and Schwabe 2018), but also evidence of greater activation for more remote, compared with recent, memories (Bosshardt et al. 2005a,b). At the same time, other studies have found decreases in univariate activation only in the posterior hippocampus (Bosshardt et al. 2005b; Takashima et al. 2009; Yamashita et al. 2009; Milton et al. 2011; Watanabe et al. 2012; Ritchey et al. 2015; Sekeres et al. 2018b).Because of these conflicting findings, instead of asking just about dependence or overall changes in activation in the hippocampus, theories and empirical research have instead increasingly considered the organization of memory representations in the hippocampus (Robin and Moscovitch 2017; Sekeres et al. 2018a). Broadly, using representational similarity analyses, several studies have shown that hippocampal memory representations tend to become differentiated over learning, particularly for memories with overlapping content (LaRocque et al. 2013; Schlichting et al. 2015; Chanales et al. 2017; Brunec et al. 2020). Furthermore, it has been suggested that information is represented at different scales or “granularity” along the long axis of the hippocampus, in line with place field size differences (Kjelstrup et al. 2008; Komorowski et al. 2013), with anterior hippocampus representing more similar, coarse-grained, or gist-like information, while the posterior hippocampus represents fine-grained, detail-oriented representations (Evensmoen et al. 2013; Poppenk et al. 2013; Robin and Moscovitch 2017; Brunec et al. 2018, 2020). However, limited work has investigated whether this representational organization is altered with consolidation. Reports have shown that memory representations sharing overlapping content become more similar over a delay (Tompary and Davachi 2017; Audrain and McAndrews 2020), yet other work has found that hippocampal representations were not modulated by time (Ritchey et al. 2015; Ezzyat et al. 2018). Intriguingly, reports indicating greater differentiation in memories in anterior compared with posterior hippocampus with consolidation (Tompary and Davachi 2017; Dandolo and Schwabe 2018; Ezzyat et al. 2018) raise the possibility that the representational granularity along the anteroposterior axis may be transformed with consolidation. Thus, more work is needed to understand how consolidation influences the representational structure of memories in the hippocampus. In particular, despite much research connecting sleep to consolidation (Diekelmann and Born 2010; Rasch and Born 2013), it remains unknown whether sleep-dependent processes facilitate such delay-dependent transformations to the hippocampus.Active processes in the sleeping brain seem to be optimized for systems consolidation. Currently, the best mechanistic evidence for sleep-dependent consolidation comes from studies on hippocampal replay showing the repeated reactivation of encoding-related patterns of hippocampal activity (Buzsáki 1989; Wilson and McNaughton 1994; Girardeau and Zugaro 2011), which seems to be coordinated with replay in areas of the cortex (Ji and Wilson 2007; Peyrache et al. 2009; Wierzynski et al. 2009). It is thought that the coupling between oscillations during non-REM sleep stages (particularly slow wave sleep [SWS])—including sharp wave ripples that support replay, thalamocortical spindles, and slow oscillations—facilitates the hippocampal–cortical dialogue and information transfer to the cortex (Buzsáki 1996; Sirota et al. 2003; Steriade 2006; Clemens et al. 2011; Mölle and Born 2011; Staresina et al. 2015). Indeed, our previously published work from the present study provided supporting evidence that the density of thalamocortical sleep spindles (11–16 Hz) during overnight sleep is related to enhanced hippocampal–cortical functional connectivity measures, and increased similarity, or greater representational overlap, among memories in the ventromedial prefrontal cortex (vmPFC) (Cowan et al. 2020). Yet, while some prior work has shown that features of sleep, including spindle density and the duration of non-REM SWS, are related to decreased retrieval-related hippocampal activation for memoranda learned prior to sleep (Takashima et al. 2006; Baran et al. 2016; Hennies et al. 2016), it remains unclear how the reactivation of hippocampal traces during replay may impact the way memories are organized along the long axis of the hippocampus.To examine the effects of sleep-dependent consolidation on the neural representation of memories in the hippocampus, we designed a within-participant 3-d study using overnight polysomnography (PSG), functional magnetic resonance imaging (fMRI), and behavioral measures of memory (Fig. 1). In this study, aspects of which have been previously published (Cowan et al. 2020), participants first studied a list of word–image pairs before sleeping overnight (Sleep List), during which PSG was recorded. Upon waking in the morning, participants studied a new list of pairs (Morning List). The word–image pairs from these two lists were then restudied while undergoing an fMRI scan, intermixed with a third, novel list of pairs (Single Study List). Associative memory was tested immediately after the scan and again 24 h later. We compared measures of multivariate pattern similarity and univariate BOLD signal for the lists learned prior to, or after, sleep to probe how modulating the opportunity for sleep-dependent consolidation impacts the way memories are organized across the long axis of the hippocampus. Furthermore, our design allowed us to examine how features of overnight sleep are related to the representational organization of memories learned prior to the sleep period, as well as the behavioral benefit of changes to the organization of these memories. Thus, our study provides a novel examination of the effects of sleep-dependent consolidation on the representation of memories along the long axis of the hippocampus.Open in a separate windowFigure 1.Study design. For all encoding and restudy sessions, participants were asked to form an association between a word and an image. Participants first encoded the Sleep List (blue) before sleeping overnight while polysomnography was recorded. The next morning (day 2), participants encoded a second set of novel word–image pairs (Morning List). After a short delay (∼2 h), participants restudied these two sets of pairs, intermixed with novel pairs (Single Study List) in the functional magnetic resonance imaging (fMRI) scanner. Source memory was tested immediately after the scan and after a 24-h delay (day 3).     

The effect of zolpidem on targeted memory reactivation during sleep

According to the active system consolidation theory, memory consolidation during sleep relies on the reactivation of newly encoded memory representations. This reactivation is orchestrated by the interplay of sleep slow oscillations, spindles, and theta, which are in turn modulated by certain neurotransmitters like GABA to enable long-lasting plastic changes in the memory store. Here we asked whether the GABAergic system and associated changes in sleep oscillations are functionally related to memory reactivation during sleep. We administered the GABA_A agonist zolpidem (10 mg) in a double-blind placebo-controlled study. To specifically focus on the effects on memory reactivation during sleep, we experimentally induced such reactivations by targeted memory reactivation (TMR) with learning-associated reminder cues presented during post-learning slow-wave sleep (SWS). Zolpidem significantly enhanced memory performance with TMR during sleep compared with placebo. Zolpidem also increased the coupling of fast spindles and theta to slow oscillations, although overall the power of slow spindles and theta was reduced compared with placebo. In an uncorrected exploratory analysis, memory performance was associated with slow spindle responses to TMR in the zolpidem condition, whereas it was associated with fast spindle responses in placebo. These findings provide tentative first evidence that GABAergic activity may be functionally implicated in memory reactivation processes during sleep, possibly via its effects on slow oscillations, spindles and theta as well as their interplay.

Sleep supports the consolidation of newly acquired memories (Mednick et al. 2011; Klinzing et al. 2019). According to the active system consolidation theory, new memories and their associated neuronal activation patterns become spontaneously reactivated (replayed) following learning in the sleeping brain (Wilson and McNaughton 1994; Diekelmann and Born 2010). These reactivations allow for the redistribution and integration of the memory representations from hippocampal to neocortical sites for long-term storage (Rasch and Born 2007; Klinzing et al. 2019). Memory reactivation during sleep has been proposed to rely on the synchronized interplay of electrophysiological oscillations characteristic of non–rapid eye movement (NREM) sleep, mainly neocortical slow oscillations (SOs, <1 Hz), thalamocortical spindles (9–15 Hz), and hippocampal ripples (80–200 Hz) (Mölle et al. 2009; Staresina et al. 2015; Helfrich et al. 2018; Ngo et al. 2020). Particularly, sleep spindles and their intricate phase coupling to SO have been suggested to be mechanistically involved in memory consolidation processes during sleep (Ulrich 2016; Antony et al. 2019). It has been proposed that memories become reinstated by spindle events, specifically during the up-state of slow oscillations, allowing for the flow of information between different brain sites as well as the induction of lasting structural and functional plastic changes in the learning-associated neuronal networks (Rosanova and Ulrich 2005; Peyrache and Seibt 2020). In addition to sleep spindles, neocortical and hippocampal theta activity (4–8 Hz) is also phase-locked to SO during NREM sleep (Gonzalez et al. 2018; Cox et al. 2019; Krugliakova et al. 2020), and this coupling has been related to memory consolidation during sleep (Schreiner et al. 2018).A number of neuromodulators seem to be involved in the generation of sleep spindles, SO and associated memory processing, most notably GABA (γ-aminobutyric acid), which is the major inhibitory neurotransmitter (Lancel 1999; Ulrich et al. 2018). Sleep spindles and sleep-dependent memory processing can be boosted by targeting the GABAergic system pharmacologically (Mednick et al. 2013). Zolpidem is one of the most frequently used drugs in this regard, binding to GABA_A receptors at the same location as benzodiazepines, thereby acting as a GABA_A receptor agonist (Lemmer 2007). Zolpidem increases the time spent in slow-wave sleep (SWS) and reduces the amount of rapid eye movement (REM) sleep (Kanno et al. 2000; Uchimura et al. 2006; Zhang et al. 2020). Zolpidem also increases the density and power of sleep spindles (Dijk et al. 2010; Lundahl et al. 2012; Mednick et al. 2013; Niknazar et al. 2015; Zhang et al. 2020) as well as the coupling of spindles to SO (Niknazar et al. 2015; Zhang et al. 2020), and it was further found to enhance declarative memory consolidation during sleep, with postsleep performance improvements being associated with higher spindle density and spindle power as well as with SO–spindle coupling (Kaestner et al. 2013; Mednick et al. 2013; Zhang et al. 2020).However, it remains unclear whether the changes in sleep stages, sleep spindles, and SO–spindle coupling after pharmacological manipulation with zolpidem are functionally related to the mechanisms underlying sleep-dependent memory consolidation such as memory reactivation. Over the last few years, targeted memory reactivation (TMR) has been increasingly applied to manipulate memory reactivation during sleep experimentally by presenting learning-associated reminder cues like odors or sounds (Oudiette and Paller 2013; Hu et al. 2020; Klinzing and Diekelmann 2020). TMR biases sleep-related neuronal replay events toward the reactivated memory contents (Lewis and Bendor 2019) and enhances subsequent recall performance (Rudoy et al. 2009; Diekelmann et al. 2011; Schreiner et al. 2015; Cairney et al. 2018). Although a few studies observed modulations of SOs (Rihm et al. 2014), sleep spindles (Cox et al. 2014), and SO–spindle coupling (Bar et al. 2020) with TMR during sleep, studies on the role of specific neurotransmitters and particularly on the role of GABAergic neurotransmission and associated changes in sleep oscillations for targeted memory reactivation are entirely lacking. One previous study tested the effect of pharmacologically increased GABAergic activity by administering the benzodiazepine clonazepam after cued reactivation of a declarative memory during wakefulness (Rodríguez et al. 2013). Clonazepam increased memory performance when it was administered after reactivation with an incomplete reminder cue, suggesting that increasing GABAergic neurotransmission may enhance the restabilization of reactivated declarative memories in humans during wakefulness.In the present study, we tested the effect of modulating GABAergic activity with zolpidem on targeted memory reactivation during sleep and associated changes in sleep spindles as well as SO–spindle and SO–theta coupling. We hypothesized that zolpidem enhances the beneficial effects of targeted memory reactivation on memory performance and that this enhancement is associated with increases in spindle density, spindle power, SO–spindle coupling, and possibly SO–theta coupling, and the amount of SWS. Participants were trained on a memory task including 30 sound–word associations in the evening (Forcato et al. 2020) and received an oral dose of 10 mg zolpidem (n = 11) or placebo (n = 11) after training before a full night of sleep in the sleep lab (Fig. 1). During the night, incomplete reminder cues (sounds + first syllable of the associated words) were played again via in-ear headphones during SWS. The next morning, participants were trained on an interference memory task to probe the stability of the original memory, which was tested 30 min later.Open in a separate windowFigure 1.Experimental design and memory task. (A) All subjects took part in a training session at ∼22.30, were administered with placebo (n = 11) or 10 mg of zolpidem (n = 11) before going to bed at 23:00, and received targeted memory reactivation during the first SWS period. After ∼8 h of sleep, in the morning, subjects learned an interference task and were tested on the original memory task in a testing session 30 min after the interference task. (B) Training: First, subjects were presented with 30 sound–word associations for learning. For each association, the sound was presented first for 2900 msec. The sound then continued accompanied by the word written on the screen and spoken aloud for 1500 msec. After a 4000-msec break, the next association was presented in the same way. After all associations were presented once, participants completed an immediate cued recall test. For each association, the sound was presented for 2900 msec. The sound then continued accompanied by the first syllable of the associated word for 1500 msec. Participants were then given 5000 msec to say the complete word aloud (sound continued during the entire period). Independently of their response, the correct answer was then presented on the screen and via headphones for 1500 msec. Reactivation: Each sound was first presented alone for an average of 2900 msec; the sound then continued accompanied by the first syllable of each word for another 1500 msec. After a 7000-msec break, the next sound–syllable pair was presented until all 30 pairs had been presented once. Testing: Each sound was presented for 500 msec and then the sound continued and subjects had 5000 msec to say the associated word aloud. After a break of 4000 msec, the procedure continued for the rest of the 30 associations. Adapted from Forcato et al. (2020).     

LTP in hippocampal area CA1 is induced by burst stimulation over a broad frequency range centered around delta

Long-term potentiation (LTP) is typically studied using either continuous high-frequency stimulation or theta burst stimulation. Previous studies emphasized the physiological relevance of theta frequency; however, synchronized hippocampal activity occurs over a broader frequency range. We therefore tested burst stimulation at intervals from 100 msec to 20 sec (10 Hz to 0.05 Hz). LTP at Schaffer collateral–CA1 synapses was obtained at intervals from 100 msec to 5 sec, with maximal LTP at 350–500 msec (2–3 Hz, delta frequency). In addition, a short-duration potentiation was present over the entire range of burst intervals. We found that N-methyl-d-aspartic acid (NMDA) receptors were more important for LTP induction by burst stimulation, but L-type calcium channels were more important for LTP induction by continuous high-frequency stimulation. NMDA receptors were even more critical for short-duration potentiation than they were for LTP. We also compared repeated burst stimulation with a single primed burst. In contrast to results from repeated burst stimulation, primed burst potentiation was greater when a 200-msec interval (theta frequency) was used, and a 500-msec interval was ineffective. Whole-cell recordings of postsynaptic membrane potential during burst stimulation revealed two factors that may determine the interval dependence of LTP. First, excitatory postsynaptic potentials facilitated across bursts at 500-msec intervals but not 200-msec or 1-sec intervals. Second, synaptic inhibition was suppressed by burst stimulation at intervals between 200 msec and 1 sec. Our data show that CA1 synapses are more broadly tuned for potentiation than previously appreciated.Long-term potentiation (LTP) is used as a model for studying synaptic events during learning and memory (Bliss and Collingridge 1993; Morris 2003; Lynch 2004). At most synapses, LTP is triggered by postsynaptic Ca²⁺ influx through N-methyl-d-aspartic acid (NMDA) glutamate receptors (Collingridge et al. 1983; Harris et al. 1984; Herron et al. 1986) and, under some conditions, through L-type voltage-gated Ca²⁺ channels (Grover and Teyler 1990, 1994; Morgan and Teyler 1999). LTP was discovered in the dentate gyrus (Bliss and Lomo 1973) following several seconds of 10–100 Hz stimulation of the perforant path. Since then, many LTP studies have used similar long, high-frequency stimulation (HFS) protocols, most typically 100 Hz, 1 sec (Bliss and Collingridge 1993). Although effective, HFS does not resemble physiological patterns of activity (Albensi et al. 2007). Patterned stimulation resembling physiological activity, in particular theta burst stimulation, is also effective for LTP induction (Larson et al. 1986; Staubli and Lynch 1987; Capocchi et al. 1992; Nguyen and Kandel 1997). Theta burst stimulation consists of short bursts (4–5 stimuli at 100 Hz) repeated at 5 Hz, which lies within the hippocampal theta frequency range (4–12 Hz) (Bland 1986; Buzsáki 2002). Primed burst stimulation, another form of patterned stimulation, involves delivery of a priming stimulus followed by a single short burst (Larson and Lynch 1986; Rose and Dunwiddie 1986). The temporal requirements for primed burst LTP are quite precise (Diamond et al. 1988; Greenstein et al. 1988; Larson and Lynch 1989): Intervals less than 140 msec or greater than 200 msec are ineffective.The mechanisms underlying theta frequency-dependent LTP have been studied primarily using the primed burst protocol (Larson and Lynch 1986, 1988, 1989; Pacelli et al. 1989; Davies and Collingridge 1996). Activation of GABA_B autoreceptors during the priming stimulus suppresses GABA release during the following burst (Davies et al. 1990; Lambert and Wilson 1994; Olpe et al. 1994), allowing greater postsynaptic depolarization (Larson and Lynch 1986; Pacelli et al. 1989) and more effective NMDA receptor activation (Davies and Collingridge 1996). Consequently, temporal requirements for primed burst potentiation match the time course of GABA_B autoreceptor-mediated suppression of GABA release (Davies et al. 1990; Davies and Collingridge 1993; Mott et al. 1993).Besides theta, hippocampal activity is observed at other frequencies, notably sharp waves (0.01–5 Hz) (Buzsáki 1986, 1989; Suzuki and Smith 1987) and low-frequency oscillations (≤1 Hz) (Wolansky et al. 2006; Moroni et al. 2007). These lower frequencies dominate during slow wave sleep (Buzsáki 1986; Suzuki and Smith 1987; Wolansky et al. 2006; Moroni et al. 2007), and contribute to hippocampal memory processing (Buzsáki 1989; Pennartz et al. 2002). While synchronized population activity over frequencies from <1 Hz to 12 Hz is associated with hippocampal memory function, previous LTP studies have focused on theta. We therefore investigated burst stimulation at frequencies from 0.05 Hz to 10 Hz. We found that CA1 synapses potentiate to some degree over this entire range and that maximal potentiation occurs around delta frequency rather than theta.     

Sleep spectral power correlates of prospective memory maintenance

Prospective memory involves setting an intention to act that is maintained over time and executed when appropriate. Slow wave sleep (SWS) has been implicated in maintaining prospective memories, although which SWS oscillations most benefit this memory type remains unclear. Here, we investigated SWS spectral power correlates of prospective memory. Healthy young adult participants completed three ongoing tasks in the morning or evening. They were then given the prospective memory instruction to remember to press “Q” when viewing the words “horse” or “table” when repeating the ongoing task after a 12-h delay including overnight, polysomnographically recorded sleep or continued daytime wakefulness. Spectral power analysis was performed on recorded sleep EEG. Two additional groups were tested in the morning or evening only, serving as time-of-day controls. Participants who slept demonstrated superior prospective memory compared with those who remained awake, an effect not attributable to time-of-day of testing. Contrary to prior work, prospective memory was negatively associated with SWS. Furthermore, significant increases in spectral power in the delta-theta frequency range (1.56 Hz–6.84 Hz) during SWS was observed in participants who failed to execute the prospective memory instructions. Although sleep benefits prospective memory maintenance, this benefit may be compromised if SWS is enriched with delta–theta activity.

Prospective memory refers to the maintenance, retrieval, and execution of a previously formed intention (Einstein and McDaniel 1990). Successful prospective memory is essential for a large number of tasks in daily life, such as remembering to attend a doctor''s appointment, to pick up a prescribed medication after that appointment, and to also pick up other needed items (e.g., groceries) while at the drugstore. The above described hypothetical sequence of events integrates previously studied prospective memory variants including time-based (i.e., maintaining a memory to complete an intention at a prespecified time; e.g., Esposito et al. 2015; Occhionero et al. 2017), activity-based (i.e., maintaining a memory to perform an intention before or after a particular activity; e.g., Occhionero et al. 2020), and cue-based (i.e., relying on external cues to prompt a maintained memory for a set intention; e.g., Scullin and McDaniel 2010; Leong et al. 2019b; Scullin et al. 2019).When it is required that memories be maintained across longer periods of time, prospective memory may become less reliable unless sleep occurs (Scullin and McDaniel 2010; Diekelmann et al. 2013a,b; Grundgeiger et al. 2014; Leong et al. 2019a,b; Scullin et al. 2019). Sleep appears to most strongly aid spontaneous retrieval of cue-based prospective memories (Leong et al. 2019a). Several reports have found that slow wave sleep (SWS) supports spontaneous retrieval of cue-based prospective memory intentions (e.g., Diekelmann et al. 2013a; Leong et al. 2019b), although at least one study found an association with rapid eye movement (REM) sleep instead (Scullin et al. 2019). Cue-based prospective memory is hypothesized to be a type of associative memory that binds prospective components (the prospective memory cue) and retrospective components (maintenance of the memory for the prospective memory intention when presented with the cue; Diekelmann et al. 2013b; Leong et al. 2019a).Rodent and human literature, implementing a variety of invasive and noninvasive brain imaging techniques, show that cortical slow oscillations (SOs; <1 Hz) and fast thalamocortical sleep spindles during SWS facilitate associative memory retention (Niknazar et al. 2015; Latchoumane et al. 2017; Helfrich et al. 2018; Mikutta et al. 2019; Muehlroth et al. 2019), whereas faster oscillations, such as those in the theta frequency band (∼4–7 Hz), may inhibit declarative associative memory (Marshall et al. 2011). We therefore hypothesize that prospective memory performance, like other studied associative memory variants, should benefit from oscillations during SWS (Klinzing et al. 2019). However, it remains unknown which SWS microarchitectural features may facilitate or inhibit prospective memory performance.Here, we aimed to first replicate prior findings that prospective memories are better maintained across a 12-h interval including sleep compared with an equivalent interval of wakefulness (e.g., Scullin and McDaniel 2010). We next explored whether sleep-associated memory maintenance might be linked to SWS microarchitectural features. To our knowledge, this is the first experiment to examine whether SWS oscillations differentiate successful from unsuccessful prospective memory performance. Given the role of hippocampal engagement in both associative memory binding (e.g., Yonelinas et al. 2019) and oscillatory coupling during SWS that supports associative memory (Niknazar et al. 2015; Latchoumane et al. 2017; Helfrich et al. 2018; Mikutta et al. 2019; Muehlroth et al. 2019), we hypothesized that prospective memory performance would be supported by SWS and specifically SOs and sleep spindle activity.     

Greater dependence on working memory and restricted familiarity in orangutans compared with rhesus monkeys

The prefrontal cortex is larger than would be predicted by body size or visual cortex volume in great apes compared with monkeys. Because prefrontal cortex is critical for working memory, we hypothesized that recognition memory tests would engage working memory in orangutans more robustly than in rhesus monkeys. In contrast to working memory, the familiarity response that results from repetition of an image is less cognitively taxing and has been associated with nonfrontal brain regions. Across three experiments, we observed a striking species difference in the control of behavior by these two types of memory. First, we found that recognition memory performance in orangutans was controlled by working memory under conditions in which this memory system plays little role in rhesus monkeys. Second, we found that unlike the case in monkeys, familiarity was not involved in recognition memory performance in orangutans, shown by differences with monkeys across three different measures. Memory in orangutans was not improved by use of novel images, was always impaired by a concurrent cognitive load, and orangutans did not accurately identify images seen minutes ago. These results are surprising and puzzling, but do support the view that prefrontal expansion in great apes favored working memory. At least in orangutans, increased dependence on working memory may come at a cost in terms of the availability of familiarity.

The prefrontal cortex is critical for a suite of cognitive control processes that are conspicuous in human cognition (Miller 2000; Rougier et al. 2005; Braver et al. 2009). One such process is working memory, which actively maintains representations in a state of heightened access for further processing (Baddeley and Hitch 1974; Engle 2010). Working memory and cognitive control are positively correlated with measures of general intelligence in humans, implicating the prefrontal cortex as a key neural structure in the evolution of human cognition (Gray et al. 2003; Unsworth and Engle 2007; Cole et al. 2012). Some neuroanatomical studies have found that the prefrontal cortex is larger than would be predicted based on body size and visual cortex volume in apes compared with monkeys (Rilling 2006; Passingham and Smaers 2014). These findings suggest that the prefrontal cortex expanded disproportionately in great apes during primate evolution. Disproportionate expansion of the prefrontal cortex in great apes may have resulted in greater capacity for cognitive control functions, such as working memory, compared with monkeys. Thus, studies directly comparing working memory in monkeys and apes are critical to understanding the evolution of intelligence in primates.The role of cognitive control, and thus the prefrontal cortex, varies among memory systems. For instance, working memory relies heavily on cognitive control, consuming substantial cognitive resources, and is known to depend on frontal brain areas (Goldman-Rakic 1995; Fuster 1997). In contrast, familiarity, which is a strength-based memory signal that codes for whether or not a percept has previously been experienced (Kelley and Jacoby 1998; Yonelinas 2002), relies substantially less on cognitive control, consumes fewer cognitive resources, and has been mostly associated with nonfrontal areas of the brain such as the perirhinal cortex (Bachevalier and Mishkin 1986; Brown and Aggleton 2001; Haskins et al. 2008; Tu et al. 2011; O''Neil et al. 2012). Thus, working memory and familiarity vary in the degree to which they rely on cognitive control, and the degree to which they rely on prefrontal areas of the brain.If a relatively large prefrontal cortex enhances cognitive control and working memory, then we should expect recognition memory tests to engage working memory relatively more in apes than in monkeys. We evaluated this hypothesis by comparing the contributions of working memory and familiarity with recognition memory performance in orangutans and rhesus monkeys. Because the orangutans and monkeys here compared have different experience with cognitive testing, we aimed to compare the relative contributions of working memory and familiarity in each species, rather than the absolute accuracy of the two species in a particular memory test. This follows a logic similar to that used in many comparative anatomical studies; for example, those cited here that found the prefrontal cortex is larger in apes relative to body weight or visual cortex volume, rather than simply absolute volume.The relative contributions of working memory and familiarity to behavior can be measured in visual recognition memory tests. In these tests subjects study a sample image at the beginning of each trial and after a delay they are presented with a test consisting of the recently seen sample image among distractors (Fig. 1). The images used in these tests can either be repeated, such that the subject''s job is to determine which image in a set of familiar images was seen most recently, or the images can be trial unique, such that at test subjects need to discriminate a previously seen image from novel distractors. Working memory is critical for solving tests with repeating images, but much less so for tests using trial unique images, where familiarity plays a much greater role (Brady and Hampton 2018a). Monkeys (Jitsumori et al. 1988; Basile and Hampton 2013a) and apes (Harlow 1944; Hayes and Thompson 1953) are more accurate and better tolerate long delay intervals in tests with trial-unique stimuli, when familiarity can support performance. Experimentally naïve monkeys require comparatively little training to demonstrate proficient use of familiarity as a mnemonic cue, compared with the training required to become proficient in using working memory (Mishkin and Delacour 1975). Active working memory and passive familiarity are independent mnemonic processes that can be doubly dissociated. Working memory is impaired by a concurrent cognitive load imposed during the memory interval, while familiarity is not affected (Logie 1986; Jacoby et al. 1989; Basile and Hampton 2013a; Brady and Hampton 2018a). Completing the double dissociation, equating the familiarity of the sample and distractor images during study impairs choice based on familiarity, but not working memory (Brady and Hampton 2018a). Thus, recognition memory tests may allow us to compare the relative contributions of these two memory processes with recognition performance across species.Open in a separate windowFigure 1.Recognition memory tests with repeating and trial-unique images. (A) In tests with trial-unique images, each image was only used once as a sample or a distractor within a session. (B) When tested with repeating images, the images were the same on each trial. The sample image was pseudorandomly selected each trial such that each image appeared equally often as the sample or as a distractor.One might expect orangutans to show greater dependence on working memory compared with rhesus monkeys for at least two reasons. First, working memory is highly refined in humans and orangutans are more closely related to humans phylogenetically, sharing a common ancestor 13 million to 14 million years ago (Stewart and Disotell 1998), whereas rhesus monkeys and humans shared a common ancestor ∼32 million years ago (Roos and Zinner 2015). Second, orangutans have a relatively larger prefrontal cortex compared with monkeys (Rilling 2006; Passingham and Smaers 2014). We compared the ability of rhesus monkeys and orangutans to maintain images from different sets in working memory. We also determined the extent to which familiarity contributed to recognition memory performance. Across three experiments, we observed striking species differences. We found that in orangutans, recognition memory performance for both repeating and trial-unique images was controlled by working memory. In contrast, monkeys relied on working memory for repeating images, and on familiarity for trial-unique images. Furthermore, monkeys dramatically outperformed orangutans in tests that exceeded the capacity and duration of working memory, and thus depended on familiarity.     

Memory for individual items is related to nonreinforced preference change

It is commonly assumed that memories contribute to value-based decisions. Nevertheless, most theories of value-based decision-making do not account for memory influences on choice. Recently, new interest has emerged in the interactions between these two fundamental processes, mainly using reinforcement-based paradigms. Here, we aimed to study the role memory processes play in preference change following the nonreinforced cue-approach training (CAT) paradigm. In CAT, the mere association of cued items with a speeded motor response influences choices. Previous studies with this paradigm showed that a single training session induces a long-lasting effect of enhanced preferences for high-value trained stimuli, that is maintained for several months. We hypothesized that CAT increases memory of trained items, leading to enhanced accessibility of their positive associative memories and in turn to preference changes. In two preregistered experiments, we found evidence that memory is enhanced for trained items and that better memory is correlated with enhanced preferences at the individual item level, both immediately and 1 mo following CAT. Our findings suggest that memory plays a central role in value-based decision-making following CAT, even in the absence of external reinforcements. These findings contribute to new theories relating memory and value-based decision-making and set the groundwork for the implementation of novel nonreinforced behavioral interventions that lead to long-lasting behavioral change.

Value-based decision-making and memory are both extensively studied processes in cognitive psychology and cognitive neuroscience (Fellows 2017). Most theories of value-based decision-making have focused on processes related to the incremental learning of value following external reinforcement, but have not explicitly addressed the role of memory per se. Thus, fundamental questions remain regarding interactions between memory and value-based decisions, which have been gaining attention in recent years.Several recent empirical studies have demonstrated interactions between episodic memory and value-based decision-making. For example, memory for past events has been shown to bias value-based decisions (Duncan and Shohamy 2016), differently for choices of novel versus choices of familiar options (Duncan et al. 2019), and choice behavior and fMRI signals during value-based decision-making were better explained by episodic memory for individual past choices than by a standard reinforcement learning model (Bornstein et al. 2017). Another study has found that during sampling of episodic memories of previous choices, the retrieved context influenced present choices, deviating from the predictions of standard reinforcement learning models (Bornstein and Norman 2017). Other studies have demonstrated that the long time known effect of choices on future preferences is related to memory processes (Chammat et al. 2017; DuBrow et al. 2019; Luettgau et al. 2020). At the neural level, the ventromedial prefrontal cortex (vmPFC) and the hippocampus both have been shown to play a role in memory processes and value-based decisions (Weilbächer and Gluth 2017) and recent studies have been further emphasizing that the hippocampus bridges between past experience and future decisions (Bakkour et al. 2019; Biderman et al. 2020).All these studies, and many others, highlighted the interaction between memory and value-based decision-making involving external reinforcements. However, everyday life involves decisions and associations that are not directly reinforced. Thus, it remains unclear whether memory plays a general role in value-based decision-making even without external reinforcements.To better understand the role of memory processes in shaping preferences independently of external reinforcements, we used a novel behavioral change paradigm, named cue-approach training (CAT). In this paradigm, associating images of items with a neutral cue and a speeded motor response results in a consistent preference enhancement without external reinforcement, which is maintained for months (Schonberg et al. 2014; Bakkour et al. 2018; Salomon et al. 2018, 2019; Botvinik-Nezer et al. 2020). During CAT, images of items are consistently paired with a neutral cue and a speeded motor response (“Go items”), while other items are presented without the cue or the response (“NoGo items”). One training session with several presentations of all items leads to long-lasting preference changes, measured as the likelihood of choosing Go over NoGo items that had similar initial subjective values (Schonberg et al. 2014). Results from over 30 samples with this paradigm have demonstrated a replicable effect on various types of stimuli, including snack food items, fruits and vegetables, unfamiliar faces, fractal art images, and positive affective images (Bakkour et al. 2016, 2017; Veling et al. 2017; Zoltak et al. 2017; Bakkour et al. 2018; Salomon et al. 2018, 2019; Botvinik-Nezer et al. 2020), revealing the potential of the CAT paradigm as an experimental platform for value-based decision-making without external reinforcements (Schonberg and Katz 2020).The underlying mechanisms of the change of preferences following CAT are not yet fully understood (Schonberg et al. 2014; Bakkour et al. 2017; Salomon et al. 2019; Botvinik-Nezer et al. 2020; Schonberg and Katz 2020). The long-lasting nature of the effect, which has been shown to last for up to 6 mo following a single training session (Schonberg et al. 2014; Salomon et al. 2018, 2019; Botvinik-Nezer et al. 2020), raises the hypothesis that memory processes are involved in its maintenance. Furthermore, previous studies have found enhanced memory for Go compared with NoGo items with other types of Go–NoGo tasks (Chiu and Egner 2015a,b; Yebra et al. 2019) and for items for which participants have a sense of agency (Murty et al. 2015). One recent study provided preliminary evidence suggesting that memory is involved in preference change following a similar nonreinforced Go/NoGo training task (Chen et al. 2021).We hypothesized that CAT enhances memory of Go items, which in turn leads to preferring these items over NoGo items. Previous neuroimaging findings with CAT that suggested possible interactions between hippocampal fMRI activity and subsequent preferences 1 mo following CAT, provide additional evidence in support of this hypothesis (Botvinik-Nezer et al. 2020). Therefore, here we set out to test the role memory processes play in the behavioral change of preferences following CAT, in the short and in the long term.We propose an underlying mechanism for the CAT effect, in which preference change following CAT results from a boost in memory encoding of positive Go items, which in itself is a consequence of enhanced perceptual processing of Go items (Schonberg et al. 2014; Botvinik-Nezer et al. 2020). We hypothesize that the enhanced encoding of Go items, as well as the greater perceptual activation in response to them, increases accessibility of attributes and associations of these specific Go items (Anderson 1983; Bhatia 2013). Furthermore, we hypothesized that preference changes, reflected in the binary choice phase, are due to the enhanced accessibility of memory associations of the Go items, which tips the scales in favor of the Go items when the associations are positive.In order to test memory for individual items, in the current work we introduced a memory recognition task following CAT. In two independent preregistered experiments and one pilot experiment, memory was evaluated following a long (16 repetitions) or short (a single exposure) CAT training session, before the probe phase that evaluated post-training preferences. We then tested our predictions that (1) memory will be stronger for Go compared with NoGo items following CAT (more accurate and faster responses in the recognition task) and (2) that memory will be related to choices (better remembered Go items will be chosen over worse remembered NoGo items). Since the link between better memory and enhanced choices is hypothesized to be related to positive associated memories, we tested the relationship between memory and choices separately for choices between low-value and choices between high-value items. These hypotheses were tested both in the short term (immediately or a few days after CAT) and in a 1-mo follow-up.     

The Black Box effect: sensory stimulation after learning interferes with the retention of long-term object location memory in rats

Reducing sensory experiences during the period that immediately follows learning improves long-term memory retention in healthy humans, and even preserves memory in patients with amnesia. To date, it is entirely unclear why this is the case, and identifying the neurobiological mechanisms underpinning this effect requires suitable animal models, which are currently lacking. Here, we describe a straightforward experimental procedure in rats that future studies can use to directly address this issue. Using this method, we replicated the central findings on quiet wakefulness obtained in humans: We show that rats that spent 1 h alone in a familiar dark and quiet chamber (the Black Box) after exploring two objects in an open field expressed long-term memory for the object locations 6 h later, while rats that instead directly went back into their home cage with their cage mates did not. We discovered that both visual stimulation and being together with conspecifics contributed to the memory loss in the home cage, as exposing rats either to light or to a cage mate in the Black Box was sufficient to disrupt memory for object locations. Our results suggest that in both rats and humans, everyday sensory experiences that normally follow learning in natural settings can interfere with processes that promote long-term memory retention, thereby causing forgetting in form of retroactive interference. The processes involved in this effect are not sleep-dependent because we prevented sleep in periods of reduced sensory experience. Our findings, which also have implications for research practices, describe a potentially useful method to study the neurobiological mechanisms that might explain why normal sensory processing after learning impairs memory both in healthy humans and in patients suffering from amnesia.

One of the most puzzling phenomena of memory is that we forget, and since its beginning as a scientific discipline, psychology has been trying to find out why and how this happens (Ribot 1882; Ebbinghaus 1885; Müller and Pilzecker 1900; Burnham 1903)? Addressing this question, Jenkins and Dallenbach (1924) published a remarkable study in 1924 suggesting that much forgetting arises from continued mental activity caused by ongoing everyday experiencing that normally follows learning in natural settings. Their intriguing findings were not systematically pursued during the next decades, as the focus shifted to exploring the role of prior or subsequent learning on forgetting; that is, effects of proactive or retroactive interference of highly similar material on memory retention. This research program eventually led into a dead end (Tulving and Madigan 1970; Wixted 2004), and interference research in humans slowed down in the 1970s. In recent years, however, interest about the neurobiological bases of interference began to emerge again (Appleby and Wiskott 2009; Bartko et al. 2010; Blake et al. 2010; Butterly et al. 2012; Luu et al. 2012; Martínez et al. 2012; Winocur et al. 2012; Peters et al. 2013; Alber et al. 2014; Censor et al. 2014; Martínez et al. 2014; McDevitt et al. 2014; Albasser et al. 2015; Eugenia et al. 2016; Koen and Rugg 2016; Ge et al. 2019; Peters and Smith 2020).In their original experiment, Jenkins and Dallenbach (1924) used sleep to reduce the amount of interference after learning. They found that when their participants went about their normal (university campus) day after learning a list of nonsense syllables, their ability to recall the lists 1, 2, 4, or 8 h later was always poorer than when instead they slept during the time between learning and test. Jenkins and Dallenbach (1924) concluded that their results “indicate that forgetting is not so much a matter of the decay of old impressions and associations than a matter of the interference, inhibition, or obliteration of the old by the new.” Their findings were replicated by others, confirming that being asleep, compared with being awake and active, indeed improves memory retention (Van Ormer 1932; Ekstrand 1967). However, it remained an open question whether it is the reduction of sensory stimulation and new learning, which would usually occur during wakefulness, that prevents retroactive interference, or whether a specific, possibly sleep-dependent, memory facilitation process was at play (Ekstrand 1967, 1972).Noting that participants in the sleep condition did not immediately fall asleep in the original experiment, but that they experienced increased quiescence shortly after learning, Minami and Dallenbach (1946) tested the retroactive interference explanation of forgetting more directly, by controlling the amount of stimulation after learning in awake animals. This remarkable experiment used Periplaneta americana (American cockroach) and a little treadmill. After learning to suppress their natural tendency to run into a dark shelter box in a bright open alley (encouraged by an electrical shock received in the dark shelter), the cockroaches were either placed on a running treadmill in a transparent box, or in a normally lit circular transparent resting chamber, where they were not able to fall asleep but experienced notably less activity than the cockroaches on the treadmill. The outcome was that cockroaches who were forced to move presented with more forgetting than those who were not, suggesting that sleep—notwithstanding its possible beneficial effect on memory—may not be necessary to promote memory retention; rather, reducing the amount of stimulation and activity after learning may be critical for attenuating retroactive interference and thus forgetting.Some six decades later, a series of experiments picked up this original line of inquiry. Exploring in humans whether memory for short prose, word lists, or spatial knowledge benefits from reduced stimulation after learning, these studies have invariably replicated the main finding that spending a 10-min retention interval in quiet wakefulness in a dimly lit room after learning leads to better memory for the learned material than participating in unrelated cognitive tasks during the retention interval (Dewar et al. 2007, 2010). Increased memory for the acquired material following quiet wakefulness is long-lasting and can be detected up to 7 d after learning (Dewar et al. 2012; Alber et al. 2014). Even in amnesic patients 10 min of reduced sensory stimulation, compared with participating in cognitive tasks, enhances memory retention for verbal material (Cowan et al. 2004; Dewar et al. 2009, 2010). This lends strong support to the suggestion that the memory loss in amnesia arises from an increased vulnerability to interference shortly after encoding (Warrington and Weiskrantz 1974; Hardt et al. 2013)Similar results have been obtained in rodents in studies exploring the role of perirhinal cortex in object recognition memory. Rats with lesions to the perirhinal cortex typically show robust impairments in object recognition tasks (Brown and Aggleton 2001; Mumby et al. 2002, 2007; Norman and Eacott 2005; Albasser et al. 2015). However, if rats are placed into a dark box during the retention interval between the encoding phase and the test phase of an object recognition task, rats with lesions to perirhinal cortex no longer show a memory deficit and perform as well as intact animals (McTighe et al. 2010). Thus, reduction of sensory stimulation between encoding and test appears to enhance memory for objects even in rats with perirhinal cortex lesions. This finding recapitulates the outcomes of the studies with human patients suffering from amnesia after hippocampal damage.The aim of the current experiments was to determine whether reducing sensory stimulation after encoding would also enhance hippocampus-dependent memory in rats. To do this, we used a spontaneous object exploration task that assesses memory for object locations (Ennaceur and Delacour 1988; Hardt et al. 2010; Migues et al. 2016, 2019). Using this approach, we replicated in rats the basic effect that quiet wakefulness promotes memory retention as previously observed in humans. Specifically, here we show that following learning, everyday activity in the home cage with cage mates impairs object location memory in rats, while reducing sensory stimulation in a dark chamber, without sleep, promotes it.     

Lifelong reductions of PKMζ in ventral hippocampus of nonhuman primates exposed to early-life adversity due to unpredictable maternal care

Protein kinase Mζ (PKMζ) maintains long-term potentiation (LTP) and long-term memory through persistent increases in kinase expression. Early-life adversity is a precursor to adult mood and anxiety disorders, in part, through persistent disruption of emotional memory throughout life. Here we subjected 10- to 16-wk-old male bonnet macaques to adversity by a maternal variable-foraging demand paradigm. We then examined PKMζ expression in their ventral hippocampi as 7- to 12-yr-old adults. Quantitative immunohistochemistry reveals decreased PKMζ in dentate gyrus, CA1, and subiculum of subjects who had experienced early-life adversity due to the unpredictability of maternal care. Adult animals with persistent decrements of PKMζ in ventral hippocampus express timid rather than confrontational responses to a human intruder. Persistent down-regulation of PKMζ in the ventral hippocampus might reduce the capacity for emotional memory maintenance and contribute to the long-lasting emotional effects of early-life adversity.

Early-life adversity is associated with an increased vulnerability to stress-related disorders that is maintained into adulthood, suggesting a very long-lived effect on emotional memory by the early-life event (Coplan et al. 1996). Although several structural and neurochemical sequelae of early-life adversity have been reported (Teicher et al. 2003; Jackowski et al. 2011), the direct effects of early-life adversity on the molecular substrates maintaining long-term memory storage have not been explored.Accumulating evidence supports a crucial role for the autonomously active, atypical protein kinase C (PKC) isoform protein kinase Mζ (PKMζ) in maintaining synaptic long-term potentiation (LTP), a putative physical substrate for memory, and long-term memory storage (Ling et al. 2002; Pastalkova et al. 2006; Glanzman 2013; Sacktor and Fenton 2018). The autonomous activity of PKMζ is due to its unusual structure that differs from other PKC isoforms (Sacktor et al. 1993). Most PKCs consist of two domains: a catalytic domain and an autoinhibitory regulatory domain that suppresses the catalytic domain. Therefore, most PKCs are inactive until second messengers bind to the regulatory domain and induce a conformational change that releases the autoinhibition. Because second messengers that activate PKCs such as Ca²⁺ or diacylglycerol have short half-lives, most PKCs are only transiently activated.PKMζ, in contrast, consists of an independent PKCζ catalytic domain, and the absence of an autoinhibitory regulatory domain results in autonomous and thus persistent activity once the kinase is synthesized. PKMζ mRNA is transcribed from an internal promoter within the PKCζ/PKMζ gene that is active only in neural tissue (Hernandez et al. 2003). The mRNA is translationally repressed and transported to dendrites of neurons (Muslimov et al. 2004). High-frequency afferent synaptic activity during LTP induction or learning derepresses PKMζ mRNA translation, triggering new synthesis of PKMζ protein (Osten et al. 1996; Hernandez et al. 2003; Tsokas et al. 2016; Hsieh et al. 2017).Once increased, the steady-state amount of PKMζ remains elevated during LTP or long-term memory maintenance. Recent work with quantitative immunohistochemistry (IHC) shows that spatial conditioning induces persistent increases of PKMζ in somatic and selective dendritic compartments of dorsal hippocampal CA1 pyramidal cells that can last at least 1 mo (Hsieh et al. 2021). The persistent increases are preferentially expressed in CA1 pyramidal cells that were activated during the formation of the memory, specifically at the termination zone of the Schaffer collateral/commissural inputs from subfield CA3. In contrast, persistent PKMζ increases are not evident in stratum lacunosum-moleculare, the termination zone that originates in entorhinal cortex that nonetheless is capable of expressing PKMζ. Postsynaptic domain-specific PKMζ expression patterns hint at distinct circuit-specific modifications of cortical–hippocampal synaptic function by maturational and experiential factors.Persistent changes in PKMζ expression are also associated with changes in the capacity for learning and memory across the life span of animals. Decreased memory ability in aged rats is associated with decreased training-induced, persistent PKMζ expression in prelimbic cortex, and increases in PKMζ are crucial for the cognition-enhancing effects of environmental enrichment in the aged animals (Chen et al. 2016). Hara et al. extended the connection between PKMζ and cognitive function to nonhuman primates (NHPs), showing that levels of PKMζ expression in dentate gyrus (DG) axospinous synapses correlate with successful performance on cognitive tasks in young and aged monkeys (Hara et al. 2012). These studies suggest that persistent down-regulation of PKMζ may comprise an important pathophysiological mechanism for cognitive impairment.Here we used a validated NHP model of early-life adversity, maternal variable-foraging demand (VFD), to explore the links between adversity in infancy and PKMζ expression in adulthood (Coplan et al. 1996; Jackowski et al. 2011). Previous studies of the VFD paradigm have revealed that both infants and their mothers exposed to VFD show significant cerebrospinal fluid (CSF) elevations of the stress neuropeptide, corticotropin-releasing factor (CRF). Moreover, the magnitude of CRF change in mothers and infants are positively correlated, suggesting synchronization of maternal–infant stress responses to the VFD stressor (Coplan et al. 2005). From a behavioral standpoint, maternal social rank plays a negligible role in determining an aggregate score of maternal–infant proximity, suggesting preferential attention of mothers to their infants. During the VFD condition, maternal social rank predicts >80% of the variance of maternal–infant proximity, suggesting mothering patterns are interrupted by preferential orientations to social rank; the latter determines food accessibility (Coplan et al. 2015). Dominant females show relative increases in maternal–infant proximity, whereas subordinate females show relative reductions in maternal–infant proximity. Neither pattern of attachment ameliorates an abnormal association between CSF oxytocin concentrations and hypothalamic-pituitary-adrenal (HPA) axis activity (Coplan et al. 2015). Offspring exposed to VFD rearing assessed both as juveniles and as full adults demonstrate persistent increases in CSF CRF concentrations in comparison with controls reared under non-VFD conditions (Coplan et al. 1996, 2001).Our prior neurohistological studies pointed to the DG as a region particularly vulnerable to VFD exposure, as shown by reduced trophic signaling and neurogenesis (Jackowski et al. 2011; Perera et al. 2011; Schoenfeld et al. 2021). We therefore hypothesized that early-life adversity due to unpredictable maternal care (for brevity, subsequently referred to as “early-life adversity”) reduces the persistent expression of PKMζ within the DG of ventral intrahippocampal neurocircuitry that mediates affective memory processing (Fanselow and Dong 2010). We used PKMζ antisera validated by the lack of immunostaining in PKMζ-null mice (Hsieh et al. 2021) to examine PKMζ expression in ventral hippocampus (NHP anterior hippocampus) in both DG granule cell layer and the stratum moleculare of the suprapyramidal blade that receives direct input from entorhinal cortex, as well as other regions encompassing the hippocampal formation, including the hilus, CA3, CA1, and subiculum.To assess behavioral correlates of hippocampal PKMζ expression, we used a stress-inducing paradigm designed specifically for singly housed bonnet macaque male NHPs, which we refer to as the “human exposure response” (Jackowski et al. 2011; Hamel et al. 2017), which is a variation of the paradigm used in human exposure studies by Kalin et al. in rhesus macaques (Kalin and Shelton 1989). On exposure to a direct human presence, singly housed adult male bonnet macaques react with a dichotomy of responses—confrontational versus timid (see the Materials and Methods) (Jackowski et al. 2011). In our macaque colony, groups of fully adult males are necessarily housed individually to prevent injury sustained during male agonistic encounters, whereas adult females and/or juveniles are safely housed in social groups. Because group housing of nursing females and/or juveniles of both sexes elicits a range of behaviors intrinsic to the species’ social repertoire (Rosenblum et al. 2001; Coplan et al. 2015) that complicates behavioral analyses to human exposure, we restricted our current study to male macaques.     

The basolateral amygdala and nucleus accumbens core mediate dissociable aspects of drug memory reconsolidation

A distributed limbic-corticostriatal circuitry is implicated in cue-induced drug craving and relapse. Exposure to drug-paired cues not only precipitates relapse, but also triggers the reactivation and reconsolidation of the cue-drug memory. However, the limbic cortical-striatal circuitry underlying drug memory reconsolidation is unclear. The aim of this study was to investigate the involvement of the nucleus accumbens core and the basolateral amygdala in the reconsolidation of a cocaine-conditioned stimulus-evoked memory. Antisense oligodeoxynucleotides (ASO) were infused into each structure to knock down the expression of the immediate-early gene zif268, which is known to be required for memory reconsolidation. Control infusions used missense oligodeoxynucleotides (MSO). The effects of zif268 knockdown were measured in two complementary paradigms widely used to assess the impact of drug-paired CSs upon drug seeking: the acquisition of a new instrumental response with conditioned reinforcement and conditioned place preference. The results show that both intranucleus accumbens core and intrabasolateral amygdala zif268 ASO infusions at memory reactivation impaired the reconsolidation of the memory underlying a cocaine-conditioned place preference. However, knockdown of zif268 in the nucleus accumbens at memory reactivation had no effect on the memory underlying the conditioned reinforcing properties of the cocaine-paired CS measured subsequently, and this is in contrast to the marked impairment observed previously following intrabasolateral amygdala zif268 ASO infusions. These results suggest that both the basolateral amygdala and nucleus accumbens core are key structures within limbic cortical-striatal circuitry where reconsolidation of a cue-drug memory occurs. However reconsolidation of memory representations formed during Pavlovian conditioning are differentially localized in each site.Through Pavlovian association with the effects of addictive drugs, a conditioned stimulus (CS) acquires both general motivational and sensory-specific conditioned reinforcing properties (Everitt et al. 2000). These associations contribute to the high likelihood of relapse in addicted individuals, yet the extinction of drug CSs by nonreinforced exposure has proved to be of limited therapeutic utility (Conklin and Tiffany 2002). In abstinent humans, drug CSs evoke salient and persistent memories of drug-taking experiences, inducing craving and relapse (Childress et al. 1988; O''Brien et al. 1992), while in animals they also precipitate relapse to, or reinstatement of, drug-seeking behavior (de Wit and Stewart 1981; Meil and See 1996; Fuchs et al. 1998; Weiss 2000). Thus, disrupting drug-related memories might significantly diminish relapse propensity on subsequent exposure to drug-paired CSs, and thereby promote abstinence.Exposure to a drug-associated CS also triggers a process of memory reconsolidation, which restabilizes the reactivated and labile memory (Nader 2003). While reconsolidation may adaptively update memories (Dudai 2006; Hupbach et al. 2007; Rossato et al. 2007; Lee 2009), its disruption may reduce the impact of intrusive or aberrant memories on behavior subsequently (Lee et al. 2005, 2006; Brunet et al. 2008; Kindt et al. 2009; Taubenfeld et al. 2009). The reconsolidation of CS–cocaine memories has been shown to depend upon protein synthesis and expression of the plasticity-associated immediate-early gene, zif268, in the basolateral amygdala (BLA), since zif268 knockdown at memory reactivation disrupted the acquired conditioned reinforcing properties of the CS measured in drug-seeking tasks days or weeks later (Lee et al. 2005, 2006).Although the BLA has an established role in CS-drug memory reconsolidation, it remains unclear whether other sites within limbic cortical-ventral striatal circuitry participate in this process. The nucleus accumbens core (AcbC) is a primary candidate, as zif268 is up-regulated in the AcbC as well as in the BLA following exposure to cocaine CSs (Thomas et al. 2003). Furthermore, the AcbC, which is strongly implicated in Pavlovian influences on drug seeking and relapse (Cardinal et al. 2002; Kalivas and McFarland 2003), has been shown to be a site where the reconsolidation of a drug conditioned place preference (CPP) memory can be disrupted (Miller and Marshall 2005).Given the evidence of increased zif268 expression in the AcbC following CS-drug memory reactivation, we investigated its requirement in the reconsolidation of cocaine-associated memories. To address this issue, we employed two different but complementary paradigms widely used to measure the conditioned effects of CSs associated with drugs of abuse: the acquisition of a new instrumental response with conditioned reinforcement (ANR) and CPP. These procedures have been used successfully to investigate the mechanisms underlying the reconsolidation of appetitive Pavlovian memories, but it is likely that they depend upon different associative mechanisms (Everitt et al. 1991; White and McDonald 1993) that in turn depend upon different neural loci within limbic cortical-striatal circuitry (Cardinal et al. 2002). Therefore, to enable a full comparison with the functional involvement of the BLA, we investigated the necessity for BLA zif268 expression in drug memory reconsolidation as assessed in the CPP paradigm.