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.     

Extinction circuits for fear and addiction overlap in prefrontal cortex

Extinction is a form of inhibitory learning that suppresses a previously conditioned response. Both fear and drug seeking are conditioned responses that can lead to maladaptive behavior when expressed inappropriately, manifesting as anxiety disorders and addiction, respectively. Recent evidence indicates that the medial prefrontal cortex (mPFC) is critical for the extinction of both fear and drug-seeking behaviors. Moreover, a dorsal-ventral distinction is apparent within the mPFC, such that the prelimbic (PL-mPFC) cortex drives the expression of fear and drug seeking, whereas the infralimbic (IL-mPFC) cortex suppresses these behaviors after extinction. For conditioned fear, the dorsal-ventral dichotomy is accomplished via divergent projections to different subregions of the amygdala, whereas for drug seeking, it is accomplished via divergent projections to the subregions of the nucleus accumbens. Given that the mPFC represents a common node in the extinction circuit for these behaviors, treatments that target this region may help alleviate symptoms of both anxiety and addictive disorders by enhancing extinction memory.Emotional memories, both in the aversive and appetitive domains, are important for guiding behavior. Regulating the expression of these memories is critical for mental health. Extinction of classical conditioning is one form of emotion regulation that is easily modeled in animals. In the aversive domain, a conditioned stimulus (CS) is typically paired with a shock, while in the appetitive domain, a CS is paired with the availability of food or drug reward. Repeated presentation of the CS in the absence of the reinforcer leads to extinction of conditioned fear or drug-seeking behaviors. In recent years, there have been great advances in our understanding of the neural circuitry responsible for this form of inhibitory learning (for reviews, see Cammarota et al. 2005; Maren 2005; Myers and Davis 2007; Quirk and Mueller 2008). The prefrontal cortex has been strongly implicated in fear expression (Powell et al. 2001; Vidal-Gonzalez et al. 2006; Corcoran and Quirk 2007) and fear extinction (Herry and Garcia 2002; Milad and Quirk 2002; Gonzalez-Lima and Bruchey 2004; Hugues et al. 2004; Burgos-Robles et al. 2007; Hikind and Maroun 2008; Lin et al. 2008; Mueller et al. 2008; Sotres-Bayon et al. 2008), and more recently, in expression of drug seeking after extinction (Peters et al. 2008a,b). These findings are consistent with a well-documented role of the prefrontal cortex in executive function and emotional regulation (Miller 2000; Fuster 2002; Quirk and Beer 2006; Sotres-Bayon et al. 2006).In this review, we propose that the medial prefrontal cortex (mPFC) regulates the expression of both fear and drug memories after extinction, through divergent projections to the amygdala and nucleus accumbens, respectively. Extinction failure in the aversive domain can lead to anxiety disorders (Delgado et al. 2006; Milad et al. 2006), while extinction failure in the appetitive domain can lead to relapse in addicted subjects (Kalivas et al. 2005; Garavan and Hester 2007). A common neural circuit for extinction of fear and drug memories would suggest shared mechanisms and treatment strategies across both domains.     

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.     

Reciprocal patterns of c-Fos expression in the medial prefrontal cortex and amygdala after extinction and renewal of conditioned fear

After extinction of conditioned fear, memory for the conditioning and extinction experiences becomes context dependent. Fear is suppressed in the extinction context, but renews in other contexts. This study characterizes the neural circuitry underlying the context-dependent retrieval of extinguished fear memories using c-Fos immunohistochemistry. After fear conditioning and extinction to an auditory conditioned stimulus (CS), rats were presented with the extinguished CS in either the extinction context or a second context, and then sacrificed. Presentation of the CS in the extinction context yielded low levels of conditioned freezing and induced c-Fos expression in the infralimbic division of the medial prefrontal cortex, the intercalated nuclei of the amygdala, and the dentate gyrus (DG). In contrast, presentation of the CS outside of the extinction context yielded high levels of conditioned freezing and induced c-Fos expression in the prelimbic division of the medial prefrontal cortex, the lateral and basolateral nuclei of the amygdala, and the medial division of the central nucleus of the amygdala. Hippocampal areas CA1 and CA3 exhibited c-Fos expression when the CS was presented in either context. These data suggest that the context specificity of extinction is mediated by prefrontal modulation of amygdala activity, and that the hippocampus has a fundamental role in contextual memory retrieval.Considerable interest has emerged in recent years in the neural mechanisms underlying the associative extinction of learned fear (Maren and Quirk 2004; Myers et al. 2006; Quirk and Mueller 2008). Notably, extinction is a useful model for important aspects of exposure-based therapies for the treatment of human anxiety disorders such as panic disorder and post-traumatic stress disorder (PTSD) (Bouton et al. 2001, 2006). During extinction, a conditioned stimulus (CS) is repeatedly presented in the absence of the unconditioned stimulus (US), a procedure that greatly reduces the magnitude and probability of the conditioned response (CR). After the extinction of fear, there is substantial evidence that extinction does not erase the original fear memory, but results in a transient inhibition of fear. For example, extinguished fear responses return after the mere passage of time (i.e., spontaneous recovery) or after a change in context (i.e., renewal) (Bouton et al. 2006; Ji and Maren 2007). In other words, extinguished fear is context specific. The return of fear after extinction is a considerable challenge for maintaining long-lasting fear suppression after exposure-based therapies (Rodriguez et al. 1999; Hermans et al. 2006; Effting and Kindt 2007; Quirk and Mueller 2008).In the last several years, considerable progress has been made in understanding the neural mechanisms underlying the context specificity of fear extinction. For example, lesions or inactivation of the hippocampus prevent the renewal of fear when an extinguished CS is presented outside of the extinction context (Corcoran and Maren 2001, 2004; Corcoran et al. 2005; Ji and Maren 2005, 2008; Hobin et al. 2006). In addition, neurons in the basolateral complex of the amygdala exhibit context-specific spike firing to extinguished CSs (Hobin et al. 2003; Herry et al. 2008), and this requires hippocampal input (Maren and Hobin 2007). Indeed, amygdala neurons that fire more to extinguished CSs outside of the extinction context are monosynaptically excited by hippocampal stimulation (Herry et al. 2008). In contrast, neurons that responded preferentially to extinguished CSs in the extinction context receive synaptic input from the medial prefrontal cortex (Herry et al. 2008).The prevalent theory of the interactions between the prefrontal cortex, hippocampus, and amygdala that lead to regulation of fear by context assumes that when animals experience an extinguished CS in the extinction context, the hippocampus drives prefrontal cortex inhibition of the amygdala to suppress fear (Hobin et al. 2003; Maren and Quirk 2004; Maren 2005). When animals encounter an extinguished CS outside of the extinction context, the hippocampus is posited to inhibit the prefrontal cortex and thereby promote amygdala activity required to renew fear. The hippocampus may also drive fear renewal through its direct projections to the basolateral amygdala (Herry et al. 2008). Although this model accounts for much of the extant literature on the context specificity of extinction, it is not known whether the nodes of this hypothesized neural network are coactive during the retrieval of fear and extinction memories. As a first step in addressing this issue, we used ex vivo c-Fos immunohistochemistry (e.g., Knapska et al. 2007) to generate a functional map of the neural circuits involved in the contextual retrieval of fear memory after extinction. Our results reveal reciprocal activity in prefrontal-amygdala circuits involved in extinction and renewal and implicate the hippocampus in hierarchical control of contextual memory retrieval within these circuits.     

The amygdala is not necessary for unconditioned stimulus inflation after Pavlovian fear conditioning in rats

The basolateral complex (BLA) and central nucleus (CEA) of the amygdala play critical roles in associative learning, including Pavlovian conditioning. However, the precise role for these structures in Pavlovian conditioning is not clear. Recent work in appetitive conditioning paradigms suggests that the amygdala, particularly the BLA, has an important role in representing the value of the unconditioned stimulus (US). It is not known whether the amygdala performs such a function in aversive paradigms, such as Pavlovian fear conditioning in rats. To address this issue, Experiments 1 and 2 used temporary pharmacological inactivation of the amygdala prior to a US inflation procedure to assess its role in revaluing shock USs after either overtraining (Experiment 1) or limited training (Experiment 2), respectively. Inactivation of the BLA or CEA during the inflation session did not affect subsequent increases in conditioned freezing observed to either the tone conditioned stimulus (CS) or the conditioning context in either experiment. In Experiment 3, NBQX infusions into the BLA impaired the acquisition of auditory fear conditioning with an inflation-magnitude US, indicating that the amygdala is required for associative learning with intense USs. Together, these results suggest that the amygdala is not required for revaluing an aversive US despite being required for the acquisition of fear to that US.Pavlovian fear conditioning in rats is a behavioral model used to investigate the neurobiology underlying the development and maintenance of fear learning and memory (Grillon et al. 1996; LeDoux 1998, 2000; Bouton et al. 2001; Maren 2001b, 2005; Kim and Jung 2006). In this model, an innocuous conditioned stimulus (CS), such as a tone, is paired with an aversive unconditioned stimulus (US), such as a footshock. After one or more pairings, the rat learns that the CS predicts the US. As a consequence, CS presentations alone elicit a conditioned fear response (CR), which includes increases in heart rate, arterial blood pressure, hypoalgesia, potentiated acoustic startle, stress hormone release, and freezing (somatomotor immobility).The amygdala has been identified as one of the major regions in which fear memories are encoded and stored. Within the amygdala, the basolateral complex of the amygdala (BLA; consisting of the lateral, basolateral, and basomedial nuclei) and the central nucleus of the amygdala (CEA) receive convergent CS and US information and are involved in the acquisition of fear memories (LeDoux 1998, 2000; Fendt and Fanselow 1999; Davis and Whalen 2001; Maren 2001b; Schafe et al. 2001; Fanselow and Gale 2003; Wilensky et al. 2006; Zimmerman et al. 2007). In addition, the CEA has an important role in the expression of fear CRs (Fendt and Fanselow 1999; LeDoux 2000; Davis and Whalen 2001; Maren 2001b; Fanselow and Gale 2003). In support of this, many studies have shown that either permanent or temporary lesions of the BLA or CEA prevent the acquisition and/or expression of fear memories (Helmstetter 1992; Helmstetter and Bellgowan 1994; Campeau and Davis 1995; Maren et al. 1996a,b; Killcross et al. 1997; Muller et al. 1997; Walker and Davis 1997; Cousens and Otto 1998; Maren 1998, 1999, 2001a,b; Wilensky et al. 1999, 2000, 2006; Goosens and Maren 2001, 2003; Nader et al. 2001; Fanselow and Gale 2003; Gale et al. 2004; Koo et al. 2004; Zimmerman et al. 2007).In addition to its role in encoding CS–US associations during conditioning, recent work suggests that the amygdala is also involved in representing properties of the US itself. For example, temporary or permanent lesions of the BLA reduce both decrements in conditioned responding after devaluation of a food US (Hatfield et al. 1996; Killcross et al. 1997; Blundell et al. 2001; Balleine et al. 2003; Everitt et al. 2003; Pickens et al. 2003; Holland 2004) and increments in conditional responding after inflation of a shock US (Fanselow and Gale 2003). Moreover, recent electrophysiological studies in primates indicate that amygdala neurons represent the value of both aversive and appetitive outcomes (Paton et al. 2006; Belova et al. 2007, 2008; Salzman et al. 2007). These studies suggest that one function of the BLA is to represent specific properties of biologically significant events, such as the food or shock USs that are typically used in Pavlovian conditioning paradigms. By this view, the BLA may represent specific sensory properties of USs that shape the nature of learned behavioral responses to the US (Balleine and Killcross 2006) and allow CSs to gain access to the incentive value of the US (Everitt et al. 2003).In contrast to this view, we recently reported that rats with neurotoxic BLA lesions exhibit normal US revaluation after Pavlovian fear conditioning (Rabinak and Maren 2008). In this study, auditory fear conditioning (75 CS–US trials) with a moderate footshock (1 mA) was followed by several exposures (five US-alone trials) to an intense footshock (3 mA) during an inflation session. Both intact rats and rats with BLA lesions exhibit a robust increase in conditional freezing to the auditory CS during a subsequent retention test (Rabinak and Maren 2008). Control experiments suggested that this was due to a revaluation of the US with which the CS was associated, rather than nonassociative sensitization of fear engendered by exposure to intense shock. These data reveal that the BLA may not be necessary for representing properties of shock USs during Pavlovian fear conditioning. To address these issues further, we have examined the consequence of reversible pharmacological manipulations of the amygdala during US inflation on conditional fear responses established with either extensive or limited training.     

Hippocampal and extrahippocampal systems compete for control of contextual fear: Role of ventral subiculum and amygdala

Two neural systems, a hippocampal system and an extrahippocampal system compete for control over contextual fear, and the hippocampal system normally dominates. Our experiments reveal that output provided by the ventral subiculum is critical for the hippocampal system to win this competition. Bilateral electrolytic lesions of the ventral subiculum after conditioning, but not before conditioning, impaired contextual fear conditioning. Reversibly inactivating this region by bilateral injections of muscimol produced the same results—no impairment when the injection occurred prior to conditioning but a significant impairment when this region was inactivated after conditioning. Thus, the extrahippocampal system can support contextual fear conditioning if the ventral subiculum is disabled before conditioning but not if it is disabled after conditioning. Our experiments also reveal that the basolateral region of the amygdala (BLA) is where the two systems compete for associative control of the fear system. To test this hypothesis we reasoned that the extrahippocampal system would also acquire associative control over the fear system, even if the hippocampal system were functional, if the basal level of plasticity potential in the BLA could be increased. We did this by injecting the D1 dopamine agonist, {"type":"entrez-protein","attrs":{"text":"SKF82958","term_id":"1156217255","term_text":"SKF82958"}}SKF82958, into the BLA just prior to conditioning. This treatment resulted in a significant increase in freezing when the ventral subiculum was disabled prior to the test. These results are discussed in relationship to the idea that D1 agonists increase plasticity potential by increasing the pool of available extrasynaptic GluR1 receptors in the population of neurons supporting acquired fear.It is generally believed that contextual fear conditioning depends on the hippocampus. However, it is now clear that an extrahippocampal system exists that can also support contextual fear conditioning. The last statement is based on the fact that damage to the hippocampus prior to conditioning has a minor impact on contextual fear (Maren et al. 1997; Frankland et al. 1998; Cho et al. 1999; Wiltgen et al. 2006). In fact, the conclusion that the hippocampus is normally involved in contextual fear conditioning is now based primarily on the finding that damage to the hippocampus after conditioning greatly impairs fear to the context in which conditioning occurs (Maren et al. 1997; Frankland et al. 1998; Anagnostaras et al. 1999; Bannerman et al. 1999; Richmond et al. 1999; Fanselow 2000; Rudy et al. 2004; Wiltgen et al. 2006). Maren et al. (1997) were the first to appreciate the implications of this set of findings. Specifically, they proposed that (1) in the normal animal there is competition between the hippocampal system and an extrahippocampal system for support of contextual fear conditioning, and (2) the hippocampal system normally dominates the extrahippocampal system—preventing it from acquiring the information needed to support fear to the context. This competition hypothesis provides a reasonable account of the lesion data. If the hippocampus is damaged prior to conditioning, then the extrahippocampal system will be able to acquire control over the fear system and generate a fear response at the time of testing. However, if the hippocampal system is functional during conditioning, it will (1) acquire control of the fear system, and (2) prevent the acquisition of control by the extrahippocampal system. Thus, if the hippocampal system is damaged after conditioning, the expression of contextual fear will be impaired, because the information that was acquired by the hippocampal system will not be available and the extrahippocampal system never acquired the relevant information.Maren et al.''s competition hypothesis is accepted by a number of other researchers (Wiltgen and Fanselow 2003; Rudy et al. 2004; Driscoll et al. 2005; Wiltgen et al. 2006). Nevertheless, very little is known about the mediators of this competition. The experiments in this study were aimed at filling some of the gaps in our knowledge of the neural basis of this competition. They are organized around two hypotheses:

• The ability of the hippocampal system to dominate the extrahippocampal system depends on information it provides through the ventral subiculum, a major output region of the hippocampus.
• The locus of the competition is the basolateral region of the amygdala (BLA), which is thought to be critical to the acquisition of conditioned fear and is where information from the hippocampal and extrahippocampal system can converge.

     

Early extinction after fear conditioning yields a context-independent and short-term suppression of conditional freezing in rats

Extinction of Pavlovian fear conditioning in rats is a useful model for therapeutic interventions in humans with anxiety disorders. Recently, we found that delivering extinction trials soon (15 min) after fear conditioning yields a short-term suppression of fear, but little long-term extinction. Here, we explored the possible mechanisms underlying this deficit by assessing the suppression of fear to a CS immediately after extinction training (Experiment 1) and the context specificity of fear after both immediate and delayed extinction training (Experiment 2). We also examined the time course of the immediate extinction deficit (Experiment 3). Our results indicate that immediate extinction produces a short-lived and context-independent suppression of conditional freezing. Deficits in long-term extinction were apparent even when the extinction trials were given up to 6 h after conditioning. Moreover, this deficit was not due to different retention intervals that might have influenced the degree of spontaneous recovery after immediate and delayed extinction (Experiment 4). These results suggest that fear suppression under immediate extinction may be due to a short-term, context-independent habituation process, rather than extinction per se. Long-term extinction memory only develops when extinction training occurs at least six hours after conditioning.Pavlovian fear conditioning and extinction are important behavioral models for studying the brain mechanisms underlying the acquisition, storage, retrieval, and suppression of traumatic fear (LeDoux 2000; Maren 2001, 2005; Kim and Jung 2005). In this procedure, an emotionally neutral stimulus, such as a tone, is paired with an aversive stimulus (US), such as an electric foot shock. After a few tone–foot shock pairings, the previous neutral tone becomes a potent conditioned stimulus (CS) and acquires the ability to elicit fear responses, such as freezing (CR). However, with repeated presentations of the CS-alone, the previously acquired CR gradually subsides, a process called extinction (Davis et al. 2003; Maren and Quirk 2004; Kim and Jung 2005; Myers and Davis 2007). The behavioral processes and the underlying neural mechanisms of extinction have attracted extensive attention in contemporary research of learning and memory (Bouton et al. 2006). Indeed, it has been suggested that failure to extinguish fear may contribute to post-traumatic stress disorder (PTSD) (Bouton et al. 2001; Rothbaum and Davis 2003). To avoid the possible long-term consequences and costs of PTSD or other anxiety disorders, clinical interventions are essential. While early interventions may manage the stress response to trauma, their efficacy has been challenged, because the acute intense stress of the traumatic experience might actually exacerbate relapse of fear (McNally 2003; Rothbaum and Davis 2003; Gray and Litz 2005). Thus, it is essential to learn when these interventions generate the best long-term extinction of fear responses.In a recent study, we demonstrated that delivering extinction trials shortly after fear conditioning yields poor long-term fear reduction (Maren and Chang 2006; but, see Myers et al. 2006). We observed that conditional freezing decreased during extinction training, but recovered completely 24 h later. This was true even when we gave 225 massed extinction trials 15 min after fear conditioning. However, in these experiments the within-session decrease in fear in rats that underwent extinction was similar to that in rats that were not exposed to extinction trials. Thus, it is unclear to what extent the short-term fear suppression we observed was due to a loss of fear to the context, the auditory CS, or both. It is also not clear whether fear suppression was due to extinction or, alternatively, another learning process such as habituation.To examine these issues further, in the present study we first assessed fear suppression to the auditory CS after immediate extinction by probing CS fear 15 min after extinction training. In a second experiment, we examined whether short-term fear suppression to the CS is renewed outside of the extinction context, as context specificity is one of the hallmarks of extinction (Bouton 2002; Ji and Maren 2007). In the third and fourth experiments, we examined the temporal delay necessary between conditioning and extinction to yield long-term suppression of fear. In our previous work (Maren and Chang 2006), all phases of training were conducted in the same context. Therefore, fear to the context decreased conditional freezing to the tone, particularly when extinction occurred shortly after conditioning, a time at which sensitized context fear was high. In an effort to isolate fear to the tone CS during extinction, we conducted extinction and test sessions in a context that was different from the conditioning context (i.e., an ABB procedure, where each letter denotes the context used for conditioning, extinction, and test, respectively). Our results reveal that delivering CS-alone trials shortly after fear conditioning produces a short-lived and context-independent suppression of freezing. This fear suppression may be due to a short-term, context-independent habituation process, rather than extinction. Furthermore, poor long-term extinction occurs even when the extinction trials were administered up to 6 h after conditioning.     

Impact of predatory threat on fear extinction in Lewis rats

Humans with post-traumatic stress disorder (PTSD) are deficient at extinguishing conditioned fear responses. A study of identical twins concluded that this extinction deficit does not predate trauma but develops as a result of trauma. The present study tested whether the Lewis rat model of PTSD reproduces these features of the human syndrome. Lewis rats were subjected to classical auditory fear conditioning before or after exposure to a predatory threat that mimics a type of traumatic stress that leads to PTSD in humans. Exploratory behavior on the elevated plus maze 1 wk after predatory threat exposure was used to distinguish resilient vs. PTSD-like rats. Properties of extinction varied depending on whether fear conditioning and extinction occurred before or after predatory threat. When fear conditioning was carried out after predatory threat, PTSD-like rats showed a marked extinction deficit compared with resilient rats. In contrast, no differences were seen between resilient and PTSD-like rats when fear conditioning and extinction occurred prior to predatory threat. These findings in Lewis rats closely match the results seen in humans with PTSD, thereby suggesting that studies comparing neuronal interactions in resilient vs. at-risk Lewis rats might shed light on the causes and pathophysiology of human PTSD.Following a severe traumatic event, some individuals manifest a syndrome, known as post-traumatic stress disorder (PTSD), characterized by repeated painful recollection of the trauma, avoidance of trauma reminders, intrusive thoughts, startle, hyperarousal, and disturbed sleep. Lifetime prevalence of PTSD ranges from 1.4% to 11.2% in representative samples (Afifi et al. 2010). Review of heritability studies indicate that there is a significant genetic component to PTSD (Nugent et al. 2008) as shared genes explain approximately 25%–38% of variability in PTSD symptom clusters and total symptoms (Afifi et al. 2010). Moreover, PTSD heritability coincides with that of other psychiatric conditions such as generalized anxiety, panic disorder, and depression (Chantarujikapong et al. 2001; Fu et al. 2007), suggesting that these disorders gain expression through common biological pathways.Although our understanding of PTSD has improved recently, we still have a limited grasp of the factors that predispose some to be at risk for PTSD, as well as those contributing to PTSD expression following trauma. In part, this situation results from the ethical limitations associated with human studies. For example, humans cannot be randomly assigned to trauma, and, importantly, the invasive techniques required to study the pathophysiology of PTSD can be used only in animals. Thus, a promising approach toward understanding the underlying pathophysiology of PTSD would be to study the disease in a valid animal model of the human syndrome.Fortunately, much work has already been performed to define an animal model of PTSD that reproduces the salient features of the human syndrome (see Adamec et al. 2006; Cohen et al. 2006a; Siegmund and Wotjak 2006). The most promising research has focused on the impact of exposing rodents to species-relevant threatening stimuli that mimic the kind of life-and-death circumstances that precipitates PTSD in humans. Indeed, rodents exposed to predators or their odor develop long-lasting (3 wk or more) manifestations of anxiety as seen in a variety of behavioral assays including the elevated plus maze (EPM), social interaction test, and acoustic startle (Adamec and Shallow 1993; Blanchard et al. 2003; Adamec et al. 2006). The inherent strength of this species-relevant stimulus was demonstrated in studies where predator odor served as an unconditioned stimulus to support cued or contextual fear conditioning (Blanchard et al. 2001; McGregor et al. 2002). As is the case with human PTSD, differential vulnerability to predatory threat was also observed in rodents. In one study, for instance, the propensity of different strains of rats to develop extreme behavioral manifestations of anxiety (EBMAs) as a result of predatory threat has been characterized, revealing that a much higher proportion (50%) of Lewis rats (an inbred strain) develops EBMAs as a result of an intense predatory threat compared with 10% of Fisher rats and 20% of Sprague–Dawley rats (Cohen et al. 2006b).Although these results are promising, it remains unclear whether Lewis rats also exhibit traits that parallel the pathophysiology of human PTSD. One such factor, thought to play a particularly critical role in the persistence of PTSD, is a compromised ability to extinguish fear memories (for review, see Quirk and Mueller 2008). Two main lines of evidence support this notion. First, in functional imaging studies, the brain structures that normally support fear expression and extinction (for review, see Pape and Pare 2010) show abnormal activity patterns in PTSD (Rauch et al. 2006; Shin et al. 2006; Bremner et al. 2008; Milad et al. 2009). Second, several studies have reported that individuals with PTSD are deficient at extinguishing classically conditioned fear responses (Orr et al. 2000; Peri et al. 2000; Blechert et al. 2007; Milad et al. 2008, 2009). Of particular interest, a study of identical twins discordant for trauma exposure has revealed that this extinction deficit was not a pre-existing condition but developed as a result of trauma (Milad et al. 2008). Given the possibility that an inability to extinguish fear might contribute to the maintenance of PTSD, we therefore tested whether Lewis rats reproduced the properties of extinction seen in human PTSD.     

Impairments in fear conditioning in mice lacking the nNOS gene

The fear conditioning paradigm is used to investigate the roles of various genes, neurotransmitters, and substrates in the formation of fear learning related to contextual and auditory cues. In the brain, nitric oxide (NO) produced by neuronal nitric oxide synthase (nNOS) functions as a retrograde neuronal messenger that facilitates synaptic plasticity, including the late phase of long-term potentiation (LTP) and formation of long-term memory (LTM). Evidence has implicated NO signaling in synaptic plasticity and LTM formation following fear conditioning, yet little is known about the role of the nNOS gene in fear learning. Using knockout (KO) mice with targeted mutation of the nNOS gene and their wild-type (WT) counterparts, the role of NO signaling in fear conditioning was investigated. Plasma levels of the stress hormone corticosterone were measured to determine the relationship between physiological and behavioral response to fear conditioning. Contextual fear learning was severely impaired in male and female nNOS KO mice compared with WT counterparts; cued fear learning was slightly impaired in nNOS KO mice. Sex-dependent differences in both contextual and cued fear learning were not observed in either genotype. Deficits in contextual fear learning in nNOS KO mice were partially overcome by multiple trainings. A relationship between increase in plasma corticosterone levels following footshock administration and the magnitude of contextual, but not cued freezing was also observed. Results suggest that the nNOS gene contributes more to optimal contextual fear learning than to cued fear learning, and therefore, inhibition of the nNOS enzyme may ameliorate context-dependent fear response.Anxiety disorders, such as post-traumatic stress disorder (PTSD), constitute the most prevalent mental illnesses in the United States, costing nearly one-third of the country''s total health bill (Greenberg et al. 1999). The treatment of these disorders requires overcoming complications such as reluctance to seek mental health treatment and an extremely high comorbidity rate with other affective disorders, reaching 80% (Brady 1997; Solomon and Davidson 1997). Emerging evidence suggests that dysfunctions underlying acquired anxiety and PTSD include an abnormal reaction to stress, which is mediated by specific neurochemical and neuroanatomical substrates (Yehuda and McFarlane 1995; Adamec 1997). Pharmacotherapies that target neuronal signaling molecules, such as nitric oxide (NO), may play a role in the treatment of these disorders.In the brain, N-methyl-d-aspartate receptor (NMDAR) activation and calcium influx into the cell activates the neuronal nitric oxide synthase (nNOS) enzyme to produce NO, which has the role of retrograde messenger (Snyder 1992). NO is involved in memory formation and synaptic plastic events such as late-phase long-term potentiation (LTP) (Lu et al. 1999; Arancio et al. 2001; Puzzo et al. 2006). Behavioral evidence in invertebrates (Lewin and Walters 1999; Muller 2000; Kemenes et al. 2002; Matsumoto et al. 2006) and vertebrates (Medina and Izquierdo 1995; Rickard et al. 1998; Ota et al. 2008) suggest that NO has a major role in consolidation of long-term memory (LTM). Recently, studies have shown that site-specific pharmacological blockade of NO signaling in rats impairs contextual (Resstel et al. 2008) and cued (Schafe et al. 2005) fear learning. However, the role of the nNOS gene in fear conditioning has not been investigated.In the present study, fear conditioning was investigated in homozygous nNOS knockout (KO) and wild-type (WT) mice. In the fear-conditioning paradigm, the association of a footshock (unconditioned stimulus; US), with a specific context and a neutral stimulus (auditory cue) results in learned fear. Re-exposure to the conditioning context and to the previously neutral auditory cue (conditioned stimulus; CS) elicits a freezing response in the absence of the aversive US. Thus, the fear-conditioning paradigm includes both contextual and cued fear learning components, which can be measured in separate tests. Fear conditioning recruits both the amygdala (emotional cue learning) and the hippocampus (spatial/contextual learning) (Phillips and LeDoux 1992; Goosens and Maren 2004; Mei et al. 2005). The involvement of these brain regions in fear learning and anxiety has been confirmed by animal and human imaging studies (LeDoux 1998; Rauch et al. 2006).We report that nNOS KO mice showed a severe deficiency in contextual fear learning and a less marked deficit in cued fear learning compared with WT mice after a single fear-conditioning session. This deficiency was partially improved by multiple (four) fear-conditioning sessions. In addition, we observed that plasma levels of corticosterone, the primary stress hormone in rodents, are related to contextual fear learning ability.     

Memory deficits are associated with impaired ability to modulate neuronal excitability in middle-aged mice

Normal aging disrupts hippocampal neuroplasticity and learning and memory. Aging deficits were exposed in a subset (30%) of middle-aged mice that performed below criterion on a hippocampal-dependent contextual fear conditioning task. Basal neuronal excitability was comparable in middle-aged and young mice, but learning-related modulation of the post-burst afterhyperpolarization (AHP)—a general mechanism engaged during learning—was impaired in CA1 neurons from middle-aged weak learners. Thus, modulation of neuronal excitability is critical for retention of context fear in middle-aged mice. Disruption of AHP plasticity may contribute to contextual fear deficits in middle-aged mice—a model of age-associated cognitive decline (AACD).Plasticity of intrinsic neuronal excitability increases the overall storage capacity of neurons and therefore likely plays a critical role in learning and memory (Zhang and Linden 2003). Increased neuronal excitability via reductions of the post-burst afterhyperpolarization (AHP) is hypothesized as a general mechanism underlying learning and memory tasks (Disterhoft et al. 1986; Disterhoft and Oh 2006). The AHP serves to limit subsequent firing in response to excitation (Madison and Nicoll 1984; Lancaster and Adams 1986; Storm 1990; Sah and Bekkers 1996). Generally speaking, the size of the AHP is inversely related to neuronal excitability, and the measurement of the AHP is routinely used as an index of neuronal excitability.Our laboratory and others have shown that AHP reductions are observed in hippocampal neurons from animals that learn hippocampal-dependent tasks including trace eyeblink conditioning in rabbit and rat (de Jonge et al. 1990; Moyer Jr et al. 1996, 2000; Kuo 2004) and spatial water maze in rat and mouse (Oh et al. 2003; Tombaugh et al. 2005; Ohno et al. 2006b). Learning-related reductions in the AHP have also been observed in cortical neurons following odor discrimination (Saar et al. 1998) and extinction learning (Santini et al. 2008). In vitro, activity-dependent plasticity of the AHP is induced using physiologically relevant stimuli (Kaczorowski et al. 2007). Because the AHP serves to limit subsequent firing, learning-related reductions in the AHP are poised to facilitate mechanisms crucial for information storage, such as long-term potentiation (LTP), synaptic integration (Sah and Bekkers 1996), metaplasticity (Le Ray et al. 2004), and spike-timing dependent plasticity (STDP) (Le Ray et al. 2004).Hippocampal neurons from naïve aged rodents and rabbits show a decrement in basal excitability evidenced by a robust enhancement of the AHP (Landfield and Pitler 1984; Moyer Jr et al. 1992, 2000; Oh et al. 1999; Kumar and Foster 2002, 2004; Power et al. 2002; Hemond and Jaffe 2005; Murphy et al. 2006b; Gant and Thibault 2008). Enhancement of the AHP in hippocampal neurons in aged animals correlates with impaired performance on learning paradigms that depend on a functional hippocampus, such as trace eyeblink and spatial water maze (Moyer Jr et al. 2000; Tombaugh et al. 2005; Murphy et al. 2006a). Pharmaceuticals aimed at reducing the AHP and increasing basal excitability (Moyer Jr et al. 1992; Moyer Jr and Disterhoft 1994) have been successful at restoring performance of aged rats on trace eyeblink conditioning (Deyo et al. 1989; Straube et al. 1990; Kowalska and Disterhoft 1994). Interestingly, AHPs from neurons recorded from aged learners are indistinguishable from young learners; both are reduced compared to that of aged weak-learners (Moyer Jr et al. 2000; Tombaugh et al. 2005). These data suggest that mechanisms that permit learning-related modulation of the AHP are also critical determinants of learning abilities in an aged population. To date, age-related impairments in hippocampal-dependent tasks and biophysical alterations in hippocampal neurons have largely focused on studies that compare animals at extreme ends of the aging spectrum.In an effort to better understand physiological changes that underlie the onset of early cognitive decline, the development of rodent models of “normal” age-associated cognitive decline (AACD), as well as mild cognitive impairment (MCI), is critical (Pepeu 2004). Therefore, we set out to characterize the development of age-related deficits indicative of hippocampal dysfunction in middle-aged C57Bl6/SJL mice and to examine the biophysical changes in hippocampal neurons that accompany such deficits.Recently, age-related deficits in contextual fear memory following trace fear conditioning were reported in a subset of middle-aged rats (Moyer Jr and Brown 2006). Because the dorsal hippocampus is critical for trace and contextual fear conditioning in mice and rats (McEchron et al. 1998; Chowdhury et al. 2005; Misane et al. 2005), trace fear conditioning is an ideal paradigm for exploring cellular mechanisms that underlie early-age-related cognitive decline.Here we investigate the effects of “early” aging on trace fear conditioning by comparing performance outcomes of young (2 mo, n = 7; and 4 mo, n = 8) and middle-aged (8 mo, n = 22) male C57/SJL F1 hybrid mice. Mice were trained and tested singly, and the experimenter was blind to the training and retention status of the mice. All animal procedures were approved by the Northwestern University Animal Care and Use Committee. Preliminary data were reported previously (Kaczorowski 2006).To assess hippocampal function with aging, young and middle-aged mice were trained on a trace fear conditioning task followed by retention tests of the auditory conditioned stimulus (CS) and contextual CS memory. The basic protocol for trace fear conditioning has been described previously (Ohno et al. 2006a). Mice were trained in a Plexiglas conditioning chamber with a stainless-steel floor grid used for shock delivery. After the baseline period (150 sec), mice received four pairings of the CS (tone; 15 sec, 3 kHz, 75 dB) and US (shock; 1 sec, 0.7 mA). The CS and unconditioned stimulus (US) were separated by a 30-sec empty trace interval. The intertrial interval was set at 210 ± 10 sec. The training chamber was wiped with 95% ethyl alcohol, illuminated with a 10-W bulb in an otherwise dark room, and provided with 65-dB white noise to make it distinct. During training on trace fear conditioning, no effect of age was observed on measures of baseline freezing (F_(2,34) = 2.0, P = 0.15), the expression of freezing during tone (F_(2,34) = 0.6, P = 0.6), or post-shock freezing (F_(2,34) = 0.2, P = 0.8), suggesting that middle-aged and young mice do not differ in measures of anxiolysis or expression of behavioral freezing (measured index of fear) (Fig. 1A).Open in a separate windowFigure 1.Onset of early aging deficits in 8-mo-old middle-aged mice. (A) Baseline (BL) freezing and auditory CS freezing during trace fear conditioning was similar between young (2 mo and 4 mo) and middle-aged (8 mo) mice. (B1) Mean baseline freezing and retention of the auditory CS memory (tones 1–4) were comparable in young (2 mo and 4 mo) and middle-aged (8 mo) mice. (B2) Middle-aged mice showed a significant decrease in freezing compared to young (2 mo and 4 mo) mice when exposed to the original context chamber where they had been trained 1 d earlier; (*) P < 0.05.Retention of the auditory CS:US memory was tested 24 h later in a novel context that differed in its location, size, scent, lighting, background noise, and flooring (bedding) compared to the training chamber. Data from three mice (one young, two middle-aged) were excluded because of video malfunction. Following a 150-sec baseline, mice received four presentations of the tone CS in the absence of footshock. Neither baseline freezing (F_(2,31) = 2.6, P = 0.1) nor conditional freezing in response to the tone CS (F_(2,31) = 1.6, P = 0.2) differed between young and middle-aged mice (Fig. 1B1). Thus, retention of the auditory CS following trace fear conditioning was intact in middle-aged compared to young mice. Although deficits in retention of auditory trace fear have been reported in aged mice and rats (Blank et al. 2003; McEchron et al. 2004; Villarreal et al. 2004), the results herein agree with report of intact trace fear memory in middle-aged rats (Moyer Jr and Brown 2006).One hour after this testing, retention of the contextual fear memory was assessed by placing mice in the original context (in the absence of the tone and footshock) and measuring freezing for 10 min. A subtle but significant difference in freezing was observed as a function of age (F_(2,31) = 4.3, P = 0.02; Fig. 1B ​2).2). A student''s post-hoc t-test revealed that mean freezing (collapsed across 10 min) of middle-aged mice was reduced compared to 2-mo (P < 0.05) and 4-mo (P < 0.05) young mice. Contextual fear memory deficits have been similarly reported in aged mice (Fukushima et al. 2008). Studies that failed to observe contextual fear deficits in aged (>18 mo) mice may result from a floor effect because young mice showed weak conditioning to the context (∼30% freezing) (Feiro and Gould 2005; Gould and Feiro 2005) or employment of delay (Corcoran et al. 2002; Feiro and Gould 2005; Gould and Feiro 2005) compared to trace procedures (Moyer Jr and Brown 2006). Although differences in experimental parameters are plausible, heterogeneity in the performance of aged mice may make detection of age-related impairments difficult owing to increased variability.Open in a separate windowFigure 2.Selective deficits on retention of contextual fear in middle-aged weak-learner mice. (A,B) Summary plot and histogram show young mice (100%, n = 6) at 2 mo of age showed robust recall of contextual fear memory (range 75%–99%) with mean and standard deviation (SD) of 91% ± 10%, whereas retention of middle-aged mice (n = 21) varied to a much greater extent (range 21%–95%; M, SD = 74% ± 19%). Distribution of middle-aged mice relative to their mean percent freezing shows two distinct populations. Middle-aged mice with freezing levels less than 3 SD from the mean freezing in young wild-type (WT) mice (61%, dashed line) were characterized as having weak contextual fear memory (weak learners) and those with freezing levels ≥62% as having strong contextual fear memory (learners). (C) Baseline (BL) freezing, expression of post-shock freezing and freezing during retention tests for auditory CS and trace CS memories, were comparable in both weak-learners and learners. (D) Selective deficits in retention of contextual fear memories were observed in middle-aged weak learners as compared to middle-aged learners; (*) P < 0.05.Previous studies in the rat report heterogeneity in spatial water maze and contextual fear conditioning in middle-aged and/or aged rats compared to young animals (Fischer et al. 1992; Wyss et al. 2000; Moyer Jr and Brown 2006). Therefore, we determined if middle-aged impairments of context fear (Fig. 2A) were driven by a subset of impaired mice. The degree of age-related impairment in each middle-aged mouse was determined by comparison to a reference group of young mice tested concurrently (shown in Fig. 1). The behavioral criterion for retention of contextual fear in middle-aged mice was set at 61%, which was 3 standard deviations (SD) below the mean freezing in young mice (mean and SD, 91% ± 10%; Fig. 1). A bimodal distribution of freezing of middle-aged mice was observed (Fig. 2B), where 70% of middle-aged mice performed above criterion and were labeled learners (n = 14), and 30% of middle-aged mice performed below criterion and were labeled as weak learners (n = 6). Comparison on measures of baseline freezing (F_(1,18) = 1.8, P = 0.2) and expression of post-shock freezing (F_(1,18) = 2.1, P = 0.2) revealed no differences between the groups during auditory trace fear training (Fig. 2C). Similarly, no differences in baseline freezing (F_(1,18) = 0.03, P = 0.9) or acquisition/recall of conditioned auditory trace fear (tone, F_(1,18) = 0.08, P = 0.8; trace, F_(1,18) = 2.4, P = 0.1) were observed 24 h later during retention tests. Thus, deficits ascribed to middle-aged weak learners were limited to contextual processing/retention, where middle-aged weak learners responded to the contextual CS with significantly lower levels of freezing compared to middle-aged learners (F_(1,18) = 47, P = 0.001; Fig. 2D). To summarize, we found that onset of cognitive decline in the C56Bl6/SJL mice was first apparent in a subset of middle-aged mice. Middle-aged weak learners showed a mild but specific deficit in hippocampal-dependent contextual learning/memory (spatial learning) but not hippocampal-dependent auditory trace learning/memory (temporal learning), assessed following trace fear conditioning.Given that contextual fear deficits occurred in a subset of middle-age mice, we were able to directly assess age-related alterations in excitability and AHP plasticity in CA1 neurons as they relate to learning abilities (learners vs. weak learners). Within 1 h of cessation of behavioral tests, middle-aged learners and weak learners were decapitated under deep halothane anesthesia and their brains quickly removed and placed into ice-cold artificial cerebral spinal fluid (aCSF): 125 mM NaCl, 25 mM glucose, 25 mM NaHCO₃, 2.5 mM KCl, 1.25 mM NaH₂PO₄, 2 mM CaCl₂, 1 MgCl₂ (pH 7.5, bubbled with 95%O₂/5%CO₂). Naïve mice were removed from their home cage and underwent identical decapitation procedures. Slices (300 μm) of the dorsal hippocampus and adjacent cortex were made using a Leica vibratome. The slices were first incubated for 30 min at 34°C in bubbled aCSF, and held at room temperature in bubbled aCSF for 1–4 h before use. Recording electrodes prepared from thin-walled capillary glass were filled with potassium methylsulfate-based internal solution and had a resistance of 5–6 MΩ.Whole-cell current-clamp recordings were performed on CA1 hippocampal pyramidal neurons of middle-aged learners (n = 36, 14 mice) and weak-learners (n = 15, 6 mice), as well as middle-aged naïve mice (n = 35 cells, 18 mice). Neuronal excitability was compared by measuring the post-burst AHP generated by 25 action potentials at 50 Hz (Fig. 3A), a stimulus shown to reliably evoke an AHP of sizable—but not maximal—amplitude from hippocampal neurons of mice (Ohno et al. 2006b). A significant difference in the peak amplitude of the AHP from learners, weak learners, and naïve mice was observed (F_(2,83) = 5, P < 0.01). Because the peak AHP and sAHP amplitudes did not differ between neurons from weak-learners and naïve mice (Fig. 3B). No differences in membrane resistance (F_(2,83) = 1.6, P = 0.2) or action potential properties, elicited using a brief (2 msec) near threshold current step (pA), were observed (Open in a separate window^aP < 0.05 compared to Weak L.^bP < 0.05 compared to Naïve.^cP < 0.05 compared to Pooled.Open in a separate windowFigure 3.Learning-related AHP plasticity is impaired in middle-aged weak-learner mice. (A) Representative traces showing the sAHP is reduced in neurons from (black) middle-aged learners compared to (blue) weak-learner mice and (gray) naïve mice. (Inset) The medium AHP (mAHP) of neurons from (black) learner mice was decreased compared to (blue) weak-learner mice and (gray) naïve mice. (B) No differences in the AHP from naïve and weak learners were observed; therefore, their data were pooled, and mean AHP was plotted by time on a log scale. (Inset) The mean amplitude of the peak AHP (1 msec) and sAHP (600 msec) was significantly reduced in neurons from learners compared to AHPs from weak learners and naïve labeled control; (*)P < 0.05.The results presented here are important in two respects. First, we demonstrate that the successful acquisition and recall of trace fear conditioning results in a significant reduction in the AHP in CA1 hippocampal neurons from the mouse. Our data are similar to previous reports showing learning-related reductions of the AHP in hippocampal neurons following training on hippocampal-dependent tasks (Disterhoft and Oh 2007) and thus strengthen the case for neuronal excitability change as a general mechanism underlying hippocampal-dependent learning. Second, we demonstrate that the onset of age-related cognitive decline in the C56Bl6/SJL mouse (termed “weak learners”) first manifests as a specific deficit in spatial associative learning in a subset of middle-age mice. These data, combined with a previous report from middle-aged rats (Moyer Jr and Brown 2006), suggest that initiation of age-related hippocampal dysfunction results in specific spatial—as opposed to temporal—deficits in associative learning and memory during middle age. By combining trace fear conditioning with whole-cell patch-clamp recordings in middle-aged mice, we revealed that “early” age-related impairments in spatial associative learning—like those in the aged hippocampus (Tombaugh et al. 2005)—result in part from an impairment of AHP plasticity of hippocampal neurons. Because AHP reductions are poised to facilitate mechanisms crucial for information storage, it is interesting that trace fear conditioning facilitates the long-term potentiation (LTP) of field excitatory postsynaptic potentials in the CA1 region of the rat hippocampus (Song et al. 2008).Generally speaking, both LTP and activation of AHP currents (I_AHP and sI_AHP) are sensitive to changes in intracellular Ca²⁺ (Storm 1990; Sah 1996; Malenka and Nicoll 1999). Thus, dysregulation of Ca²⁺ homeostasis in the hippocampus of middle-aged rats via enhancement of Ca²⁺-induced Ca²⁺ release (CICR) is an important finding (Gant et al. 2006). Age-related enhancement of Ca²⁺-dependent AHPs has been shown to raise the threshold for induction of LTP (Kumar and Foster 2004). These data support our hypothesis that impairments in contextual fear reported herein, as well as deficits in spatial water maze reported in middle-aged rats (Frick et al. 1995; Markowska 1999; Kadish et al. 2009), result from dysfunction of AHP plasticity.Studies in middle-aged mice have important implications for the treatment of “normal” age-associated cognitive decline (AACD), as well as mild cognitive impairment (MCI) (Pepeu 2004). Further studies aim to examine alterations in cholinergic function in our middle-aged mouse model, as the cholinergic agonist carbachol suppressed the AHP in neurons from naïve middle-aged mice (Supplemental Fig. 1). Activation of cholinergic receptors shape neuronal excitability and synaptic throughput (Tai et al. 2006) through multiple Ca²⁺-dependent processes (Gahwiler and Brown 1987; Tai et al. 2006). Restoration of cholinergic function has been shown to rescue deficits on hippocampal-dependent tasks in aged rodent and mouse models of Alzheimer''s disease (AD) (Disterhoft and Oh 2006), as well as in human AD patients (Cummings et al. 1998; Morris et al. 1998; Pettigrew et al. 1998), and therefore is a potential target aimed at the rescue of early age-related cognitive decline.     

DHPG activation of group 1 mGluRs in BLA enhances fear conditioning

Group 1 metabotropic glutamate receptors are known to play an important role in both synaptic plasticity and memory. We show that activating these receptors prior to fear conditioning by infusing the group 1 mGluR agonist, (R.S.)-3,5-dihydroxyphenylglycine (DHPG), into the basolateral region of the amygdala (BLA) of adult Sprague–Dawley rats enhances freezing normally supported by a weak footshock. This effect of DHPG was blocked when it was co-infused with either the general group 1 mGluR1 antagonist, (R,S)-1-aminoindan-1,5 dicarboxylic acid (AIDA), or with the selective mGluR5 antagonist, 2-methyl-6-(phenylethynyl)-pyridine (MPEP). These results support previous findings by Rodrigues and colleagues that mGluR5s in the lateral region of the amygdala make an import contribution to fear conditioning. More importantly, they support the general ideas embedded in the concept of metaplasticity, as per Abraham, and the synaptic-tagging hypothesis per Frey and Morris—that the processes that specify the content of experience can be experimentally separated from those needed to acquire the memory.The last decade has seen an increased appreciation of the view that the plasticity state of neurons—their ability to be modified by experience—is not fixed. Instead, the effect of experience depends on the physiological or biochemical state of the neurons or synapses that receive and store information contained in the experience, and this state is variable. This perspective is captured by the concept called “metaplasticity” (Abraham and Bear 1996; Abraham and Tate 1997; Abraham 2008), which recognizes that prior events can change the general plasticity or modifiability of neurons and synapses that will potentially store information contained in a subsequent target experience. This idea is also embedded in the synaptic-tagging hypothesis (Frey and Morris 1997, 1998) that will be described in some detail in the Discussion section. This view is important because it recognizes that the processes that specify the content of experience can be experimentally separated from those that are needed to store the memory for the experience.As reviewed by Abraham (2008), the range of prior events that can potentially determine a neuron''s state of plasticity is quite large and can be mediated by their effects on many components of the machinery that supports changes in synaptic strength. They include modification in NMDA-receptor function, AMPA-receptor trafficking, neuronal excitability, and epigenetic modifications.One way that the potential for plasticity can be altered is by activation of group 1 metabotropic glutamate receptors (mGluR1s and mGlur5s) (Abraham 2008). A number of reports based on the in vitro long-term potentiation (LTP) methodology suggest that activating this class of receptors prior to the delivery of a relatively weak inducing stimulus can transform a normally short-lasting form of LTP into a more persistent form (Cohen and Abraham 1996; Cohen et al. 1998; Raymond et al. 2000). For example, an infusion of (R.S.)-3,5-dihydroxyphenylglycine (DHPG), a group 1 mGluR agonist, into the bathing medium 30 min prior to the delivery of a weak inducing stimulus can significantly enhance the persistence of the resulting LTP, while having no effect on the baseline response (Cohen et al. 1998; Raymond et al. 2000). This effect of DHPG, however, is blocked (Raymond et al. 2000) when its administration is accompanied by an infusion of the group 1 mGluR antagonist, (R,S)-1-aminoindan-1,5 dicarboxylic acid (AIDA).Group 1 mGluRs also have been implicated in fear conditioning. For example, Rodrigues et al. (2002) reported that one subtype, mGluR5, is localized in dendrites and spines in neurons in the lateral nucleus of the amygdala, and fear conditioning can be significantly impaired if the mGluR5 antagonist, 2-methyl-6-(phenylethynyl)-pyridine (MPEP), is infused into the lateral nucleus prior to conditioning.These findings, from the studies of synaptic plasticity and fear conditioning, provide the empirical basis for the hypothesis that motivates the experiments reported here—that the plasticity potential of neurons in the amygdala that support fear conditioning can be increased by activating group 1 mGluRs prior to the conditioning experience.To evaluate this idea, we followed a strategy similar to that used to study the role of mGluRs in LTP (Cohen and Abraham 1996; Cohen et al. 1998; Raymond et al. 2000). In these in vitro LTP experiments, a relatively weak inducing stimulus was used to generate a short-lasting LTP. Infusing the group 1 agonist DHPG prior to the delivery of the inducing stimulus transformed this short-lasting LTP into a more persistent form. Their results suggest that the activation of group 1 mGluR1 receptors independently initiates processes that make an important contribution to long-lasting LTP. To apply this strategy to fear conditioning, we used a very low level of shock—one that by itself produced an almost undetectable level of conditioned fear (as measured by freezing, the innate defensive response of rodents). We then determined if the activation of group 1 mGluRs prior to the conditioning experience would transform this outcome and produce a stronger level of freezing. DPHG was infused into the basolateral region of the amygdala (BLA) to activate the mGluRs.     

Nuclear disconnection within the amygdala reveals a direct pathway to fear

It is widely believed that a descending serial circuit consisting of neural projections from the basolateral complex (BLA) to the central nucleus (CEA) of the amygdala mediates fear expression. Here we directly test this hypothesis and show that disconnecting the BLA and CEA with asymmetric neurotoxic lesions after Pavlovian fear conditioning in rats completely abolishes the expression of conditional freezing. These results demonstrate that neural projections from the BLA to CEA are essential for the expression of learned fear responses.Long-standing anatomical models of the brain circuitry underlying learned fear posit serial flow of information through the amygdala to engage the expression of fear responses (LeDoux 2000; Maren 2001). Specifically, conditioning-induced plasticity in the lateral nucleus of the amygdala (LA) is thought to drive learned fear through excitatory axonal projections to the basal nuclei of the amygdala (BA; basolateral and basomedial nuclei), which in turn send unidirectional and excitatory synaptic projections to the medial division of the central nucleus of the amygdala (CEm). Neurons in the CEm project to brain structures involved in the production of a variety of fear responses (Pitkänen et al. 1997; Swanson and Petrovich 1998). Alternatively, neurons in LA might excite neurons in CEm by limiting inhibitory input from the intercalated cell masses (ITC) interposed between the basolateral complex (BLA; lateral and basal nuclei) and the central nucleus of the amygdala (CEA) (Paré et al. 2004). In either case, the BLA is well positioned to drive learned fear responses via anatomical connections with the CEA.Although an extensive literature demonstrates the importance of both the BLA and CEA in the acquisition and expression of fear (LeDoux et al. 1990; Lee et al. 1996; Amorapanth et al. 2000; Goosens and Maren 2001), it is not known whether a functional connection between the two structures is essential for the expression of learned fear. Indeed, the BLA and CEA make independent contributions to aversively motivated learning under some conditions (Killcross et al. 1997; Amorapanth et al. 2000). Moreover, recent work in appetitive conditioning paradigms challenges the necessity of serial circuits in the amygdala for associative learning processes (Holland and Gallagher 1999; Everitt et al. 2003; Balleine and Killcross 2006). It is therefore essential to determine whether serial connections between the BLA and CEA are involved in the expression of fear memories as widely assumed in the literature.To address this issue, we made asymmetric neurotoxic lesions of the BLA and CEA after Pavlovian fear conditioning in rats. That is, we placed BLA lesions in one hemisphere and CEA lesions in the contralateral hemisphere, thereby producing a functional disconnection of the two brain regions. Control animals received neurotoxic BLA and CEA lesions in the same hemisphere, thereby leaving both structures and their connections intact in one hemisphere. This disconnection strategy (e.g., Olton et al. 1982) capitalizes on the fact that projections from the BLA to CEA are both ipsilateral and unidirectional (Pitkänen et al. 1997). It has been used by several groups to assess the contribution of connections between brain areas to learning and memory processes, including conditioned stimulus processing (Han et al. 1999) and appetitive spatial learning (Ito et al. 2008), for example.Fear conditioning was conducted in standard observation chambers (see Supplemental Methods) and consisted of five pairings of an auditory conditional stimulus (CS) (2 kHz, 10 sec, 80 dB) with a footshock unconditioned stimulus (US) (2 sec, 1 mA); the intertrial interval (ITI) was 1 min. Freezing behavior served as the measure of conditional fear. Twenty-four hours after fear conditioning, the rats were deeply anesthetized, and amygdala lesions were made by infusing N-methyl-d-aspartate (NMDA; 20 mg/mL in 100 mM phosphate buffered saline with a pH 7.4; Sigma) into the CEA and BLA through a 28-g injection cannula attached to a Hamilton syringe via polyethylene tubing. Control rats received unilateral lesions of the BLA and CEA in the same hemisphere (IPSI, n = 13) or sham surgery (SHAM, n = 16), whereas experimental rats received the unilateral BLA and CEA lesions in opposite hemispheres (CONTRA, n = 11). Hence, functional connectivity between the BLA and CEA was left intact in one hemisphere among rats in the IPSI group, whereas the asymmetric lesions in rats in the CONTRA group eliminated this connection. Retention tests assessing fear to the conditioning context and tone CS were conducted in separate sessions one week after recovery from surgery. Histological examination of coronal brain sections obtained from the subjects after the experiment revealed selective CEA and BLA lesions in each group; representative lesions are illustrated in Figure 1, A and B. The lesions spared fibers of passage (Fig. 1C).Open in a separate windowFigure 1.Disconnection of the BLA and CEA impairs the expression of conditional fear. (A) Schematic representation of a typical BLA (shaded) and CEA (black) lesions among rats with unilateral lesions placed in the same hemisphere (IPSI) or in opposite hemispheres (CONTRA). (B) Photomicrograph of a thionin-stained coronal section from a representative rat in the CONTRA group. The image has been cropped to include only the left and right amygdale. Broken lines encircle CEA and BLA lesions in the left and right hemispheres, respectively. (C, left) Photomicrograph of a thionin-stained coronal section from a representative rat in the IPSI group. The image has been cropped to focus on the lesion in the right amygdala. Broken lines encircle the CEA and BLA lesions. (C, right) Photomicrograph of an AuCl-stained coronal section adjacent to that shown at left indicates that NMDA infusions into the amygdala did not damage myelinated axons in the vicinity of the lesion. (D) Mean percentage of freezing (± SEM) during the conditioning session (collapsed across CSs and ITIs) and retention tests in rats in each of the three groups. The expression of fear to the conditioning context and tone CS was severely impaired by disconnection of the BLA and CEA in the CONTRA group. *P < 0.05.As shown in Figure 1D (left panel), all rats acquired similar levels of conditional freezing (collapsed across the CS and ITI for each trial) by the end of the presurgical fear conditioning session (main effect of trial, F_(5,185) = 42.1, P < 0.0001); there was neither a main effect of group (F < 1) nor a group × trial interaction (F < 1.3). One week after recovery from surgery, conditional freezing to the conditioning context and the auditory CS was assessed in separate retention tests. Importantly, functional disconnection of the BLA and the CEA after fear conditioning severely impaired the expression of conditioned freezing to both the conditioning context and the auditory CS (Fig. 1D, right). Indeed, the expression of fear to the auditory CS was essentially abolished. Freezing among rats in the CONTRA group was significantly lower than that in both the SHAM and IPSI groups during both retention tests (main effect of group, F_(2,37) = 9.98, P < 0.001), and this difference did not interact with the nature of the retention test (F < 1). Hence, functional disconnection of the BLA and CEA produced an impairment in the expression of conditional fear that was much greater than that produced by comparable lesions that spared this connection.The logic of the disconnection procedure requires that focal CEA and BLA lesions in each hemisphere do not damage the adjacent BLA or CEA, respectively. For instance, if CEA lesions produce retrograde degeneration in the neighboring BLA, rats in the CONTRA group would effectively have bilateral lesions in the BLA. Such an outcome would be expected to yield the massive deficits in conditional freezing that we have observed. Inspection of thionin-stained coronal sections suggested that, as in previous studies (Goosens and Maren 2001), our lesions were indeed selective for the targeted areas (Fig. 1B). However, to increase our confidence that the BLA adjacent to a CEA lesion was in fact functional, we performed c-fos immunohistochemistry on brain sections obtained from a separate group of rats trained and tested as previously described (SHAM, CONTRA, and IPSI groups; n = 8 per group). In addition to these rats, we included a group of rats that were not conditioned (NO-SHOCK, n = 8) to quantify fear-induced increases in c-fos expression. All rats were sacrificed 90 min after the retention test to the auditory CS. Amygdaloid c-fos expression in the SHAM rats and the intact hemisphere of the IPSI rats did not differ, and these groups were therefore collapsed into a single SHAM group.As shown in Figure 2, SHAM rats exhibited a greater density of c-fos positive nuclei in both the CEA and BLA relative to nonshocked controls. Quantification of these data confirmed this observation and revealed significant differences in c-fos expression among the groups in both the CEA (Fig. 2, left; F_(2,40) = 5.59, P < 0.01) and the BLA (Fig. 2, right; F_(2,40) = 5.07, P < 0.01). Importantly, BLA c-fos expression adjacent to CEA lesions in the CONTRA group was similar to that in intact SHAM rats, and significantly greater than that in nonshocked controls (Fig. 2B, right). This suggests that the failure of CONTRA rats to express conditional fear responses was not due to a failure to engage the intact BLA but rather to the functional disconnection of BLA activity from CEA output. Indeed, the CEA adjacent to a BLA lesion exhibited a marked reduction in c-fos expression relative to SHAM controls (Fig. 2, left), indicating that BLA lesions failed to drive ipsilateral CEA neurons important for the expression of learned fear. It is possible that the absence of c-fos expression in the CEA, in this case, is due to encroachment of the adjacent BLA lesion. However, thionin-stained sections revealed that BLA lesions were selective. Moreover, we observed intact BLA c-fos expression in rats with adjacent CEA lesions suggesting that nearby lesions per se do not disrupt c-fos expression.Open in a separate windowFigure 2.Functional disconnection of the BLA and CEA revealed by c-fos expression. Mean density (± SEM) of c-fos positive nuclei in the CEA (left) and BLA (right) in rats from each of the three groups. Disconnection of the BLA and CEA eliminated fear-related increases in c-fos expression in the CEA (left), but not the BLA (right). *P < 0.05.These results reveal that a functional connection between the BLA and CEA is required for the expression of learned fear. Considering that BLA projections to the CEA are largely unidirectional, our data reveal that a serial circuit from the BLA to the CEA mediates the expression of conditional fear responses. Consistent with this view, there are numerous reports that permanent lesions or reversible inactivation of either the BLA or CEA prevent the expression of conditioned fear (Lee et al. 1996; Maren 1999; Zimmerman et al. 2007). It is now apparent that the necessity for both the BLA and CEA in fear expression arises from the functional connectivity between them. Anatomically, this connection might involve a direct excitatory projection from BA to CEm (Paré et al. 1999) or indirect projections from LA to ITC and the lateral division of the central nucleus (CEl), both of which project to CEm (Smith and Paré 1994; Paré et al. 2004). Recent data reveal, however, that selective immunotoxic lesions of the ITC do not impair the expression of conditioned freezing (Likhtik et al. 2008). Moreover, CEl projections to CEm are inhibitory, making it unlikely that that an LA-CEl projection drives learned fear responses via CEm (Paré et al. 2004). Hence, it appears that the most likely route by which fear CSs drive learned fear involves projections from BA to CEm. Consistent with this, selective BA lesions disrupt the expression of conditioned freezing when made either before (Goosens and Maren 2001) or after (Anglada-Figueroa and Quirk 2005) fear conditioning.The dependence of conditional fear on a serial circuit between the BLA and CEA stands in contrast to the independent roles these areas have been proposed to play in appetitive conditioning paradigms (Holland and Gallagher 1999; Everitt et al. 2003; Balleine and Killcross 2006). For example, CEA, but not BLA, lesions have been reported to produce deficits in the acquisition and expression of autoshaped conditioned responses (CRs), indicating that the CEA has an independent contribution to CR expression for food-motivated responses (Parkinson et al. 2000). Moreover, the BLA has a role in the attribution of incentive salience to rewarding stimuli independent of the generation of CRs to those stimuli (Hatfield et al. 1996). However, in aversive conditioning, it appears that the BLA may not encode the motivational properties of the shock US (Rabinak and Maren 2008), but rather CS–US associations that are essential for organizing conditional fear responses by the CEA. Indeed, there is an emerging body of data suggesting that these associations may be established not only in sensory afferents in the BLA but also in the CEA (Wilensky et al. 2006; Zimmerman et al. 2007). Synaptic plasticity in projections from BA to CEm may be the essential substrate underlying the functional connectivity between these structures that is essential for the expression of fear memory.     

Stimulation of the lateral geniculate,superior colliculus,or visual cortex is sufficient for eyeblink conditioning in rats

The role of the cerebellum in eyeblink conditioning is well established. Less work has been done to identify the necessary conditioned stimulus (CS) pathways that project sensory information to the cerebellum. A possible visual CS pathway has been hypothesized that consists of parallel inputs to the pontine nuclei from the lateral geniculate nucleus (LGN), superior colliculus (SC), pretectal nuclei, and visual cortex (VCTX) as reported by Koutalidis and colleagues in an earlier paper. The following experiments examined whether electrical stimulation of neural structures in the putative visual CS pathway can serve as a sufficient CS for eyeblink conditioning in rats. Unilateral stimulation of the ventral LGN (Experiment 1), SC (Experiment 2), or VCTX (Experiment 3) was used as a CS paired with a periorbital shock unconditioned stimulus. Stimulation was delivered to the hemisphere contralateral to the conditioned eye. Rats in all experiments were given five 100-trial sessions of paired or unpaired eyeblink conditioning with the stimulation CS followed by three paired sessions with a light CS. Stimulation of each visual area when paired with the unconditioned stimulus supported acquisition of eyeblink conditioned responses (CRs) and substantial savings when switched to a light CS. The results provide evidence for a unilateral parallel visual CS pathway for eyeblink conditioning that includes the LGN, SC, and VCTX inputs to the pontine nuclei.Pavlovian eyeblink (eyelid closure and nictitating membrane movement) conditioning is established by pairing a conditioned stimulus (CS), usually a tone or light, with an unconditioned stimulus (US) that elicits the eyeblink reflex. The eyeblink conditioned response (CR) emerges over the course of paired training, occurs during the CS, and precedes the US (Gormezano et al. 1962; Schneiderman et al. 1962). Neurobiological investigations of Pavlovian eyeblink conditioning have primarily focused on the cerebellum, which is the site of memory formation and storage (Thompson 2005). The anterior interpositus nucleus is necessary for acquisition and retention of the eyeblink CR (Lavond et al. 1985; Krupa and Thompson 1997; Freeman Jr. et al. 2005; Thompson 2005; Ohyama et al. 2006). Lobule HVI and the anterior lobe of the cerebellar cortex (lobules I–V) contribute to acquisition, retention, and timing of the CR (McCormick and Thompson 1984; Perrett et al. 1993; Perrett and Mauk 1995; Attwell et al. 1999, 2001; Medina et al. 2000; Nolan and Freeman Jr. 2005; Nolan and Freeman 2006). The brainstem nuclei that comprise the proximal ends of the CS and US input pathways to the cerebellum have also been identified.The pontine nuclei (PN) and inferior olive (IO) receive CS and US information, respectively, and are the primary sensory relays into the interpositus nucleus and cerebellar cortex (Thompson 2005). Conditioned stimulus information converges in the PN, which receives projections from lower brainstem, thalamus, and cerebral cortex (Glickstein et al. 1980; Brodal 1981; Schmahmann and Pandya 1989; Knowlton et al. 1993; Campolattaro et al. 2007). The lateral pontine nuclei (LPN) are the sources of auditory CS information projected into the cerebellum. Lesions of the LPN block CR retention to a tone CS, but have no effect on CRs to a light CS (Steinmetz et al. 1987). Thus, CS inputs from different sensory modalities may be segregated at the level of the PN. Neurons in the PN project CS information into the contralateral cerebellum via mossy fibers in the middle cerebellar peduncle that synapse primarily on granule cells in the cerebellar cortex and on neurons in the deep nuclei (Bloedel and Courville 1981; Brodal 1981; Steinmetz and Sengelaub 1992). Stimulation of the PN acts as a supernormal CS supporting faster CR acquisition than conditioning with peripheral stimuli (Steinmetz et al. 1986, 1989; Rosen et al. 1989; Steinmetz 1990; Tracy et al. 1998; Freeman Jr. and Rabinak 2004). The primary focus of these experiments was to investigate the most proximal components of the CS pathway in eyeblink conditioning. There has been less emphasis on identifying the critical CS pathways that project information to the PN.Recent studies using lesions, inactivation, stimulation, and neural tract tracing have provided evidence that the auditory CS pathway that is necessary for acquisition and retention of eyeblink conditioning is comprised of converging inputs to the medial auditory thalamic nuclei (MATN), and a direct ipsilateral projection from the MATN to the PN (Halverson and Freeman 2006; Campolattaro et al. 2007; Freeman et al. 2007; Halverson et al. 2008). Unilateral lesions of the MATN, contralateral to the conditioned eye, block acquisition of eyeblink CRs to a tone CS but have no effect on conditioning with a light CS (Halverson and Freeman 2006). Inactivation of the MATN with muscimol blocks acquisition and retention of CRs to an auditory CS, and decreases metabolic activity in the PN (Halverson et al. 2008). The MATN has a direct projection to the PN and stimulation of the MATN supports rapid CR acquisition (Campolattaro et al. 2007). Our current model of the auditory CS pathway consists of converging inputs to the MATN, and direct unilateral thalamic input to the PN (Halverson et al. 2008).Less work has been done to identify the visual CS pathway necessary for eyeblink conditioning. A possible parallel visual CS pathway has been hypothesized, which includes parallel inputs to different areas of the PN from the lateral geniculate nucleus (LGN), superior colliculus (SC), visual cortex (VCTX), and pretectal nuclei (Koutalidis et al. 1988). In the Koutalidis et al. study, lesions of the LGN, SC, VCTX, or pretectal nuclei alone had only a partial effect on CR acquisition with a light CS. Lesions of any two of these structures together produced a more severe impairment on acquisition and combined lesions of all of these areas completely blocked CR acquisition to a light CS (Koutalidis et al. 1988). Each visual area investigated in the Koutalidis et al. study has a direct projection to the PN that could be important for eyeblink conditioning. The ventral LGN projects to the medial, and to a lesser extent, the lateral PN (Graybiel 1974; Wells et al. 1989). The superficial, intermediate, and deep layers of SC send projections to both the dorsomedial and dorsolateral PN (Redgrave et al. 1987; Wells et al. 1989). The VCTX has a direct projection to the rostral and lateral portions of the PN (Glickstein et al. 1972; Baker et al. 1976; Mower et al. 1980; Wells et al. 1989). The pretectal nuclei also have a direct projection to both the medial and lateral PN (Weber and Harting 1980; Wells et al. 1989). However, stimulation of the anterior pretectal nucleus is not an effective CS for eyeblink conditioning (Campolattaro et al. 2007). The failure to establish conditioning with stimulation of the anterior pretectal nucleus as a CS suggests that there may be differences in the efficacy of the various visual inputs to the PN for cerebellar learning. The following experiments investigated the sufficiency of stimulation of the LGN, SC, or primary VCTX as a CS for eyeblink conditioning in rats.     

Extinction as new learning versus unlearning: considerations from a computer simulation of the cerebellum

Like many forms of Pavlovian conditioning, eyelid conditioning displays robust extinction. We used a computer simulation of the cerebellum as a tool to consider the widely accepted view that extinction involves new, inhibitory learning rather than unlearning of acquisition. Previously, this simulation suggested basic mechanistic features of extinction and savings in eyelid conditioning, with predictions born out by experiments. We review previous work showing that the simulation reproduces behavioral phenomena and lesion effects generally taken as evidence that extinction does not reverse acquisition, even though its plasticity is bidirectional with no site dedicated to inhibitory learning per se. In contrast, we show that even though the sites of plasticity are, in general, affected in opposite directions by acquisition and extinction training, most synapses do not return to their naive state after acquisition followed by extinction. These results suggest caution in interpreting a range of observations as necessarily supporting extinction as unlearning or extinction as new inhibitory learning. We argue that the question “is extinction reversal of acquisition or new inhibitory learning?” is therefore not well posed because the answer may depend on factors such as the brain system in question or the level of analysis considered.Pavlovian eyelid conditioning is robustly bidirectional. Conditioned responses that are acquired via training that pairs a conditioned stimulus (CS) with an unconditioned stimulus (US) can be rapidly extinguished with CS-alone training or unpaired CS-US training (Gormezano et al. 1983; Napier et al. 1992; Macrae and Kehoe 1999; Kehoe and Macrae 2002; Kehoe and White 2002; Weidemann and Kehoe 2003). Whether extinction involves unlearning or separate inhibitory learning that suppresses conditioned response expression remains an important issue for both behavioral theories and for investigations of underlying neural mechanisms (Pavlov 1927; Hull 1943; Konorski 1948, 1967; Rescorla and Wagner 1972; Mackintosh 1974; Rescorla 1979; Bouton 1993, 2002; Falls 1998; Myers and Davis 2002; Kehoe and White 2002). Here, we addressed this issue using a computer simulation of the cerebellum that is capable of emulating many aspects of eyelid conditioning. Although simulation results cannot resolve such issues, several aspects of the simulation''s behavior are instructive. Even though the sites of plasticity are, in general, affected in opposite directions by acquisition and extinction training, the simulation can emulate several behavioral phenomena that are generally taken as evidence that extinction does not involve unlearning. Moreover, we found that the strengths of most synapses are quite different from their naive state following acquisition and then extinction. Independent of the overall biological accuracy of this simulation, these results highlight a variety of implications for ongoing debates about the roles of unlearning versus new learning in extinction.A combination of factors makes it possible to analyze the neural basis of eyelid conditioning in detail, and to build and test computer simulations of its cerebellar mechanisms (Medina and Mauk 2000). Foremost among these is the close association between eyelid conditioning and the cerebellum (Thompson 1986; Raymond et al. 1996; Mauk and Donegan 1997). Previous studies from several labs have shown that (1) cerebellar output drives the motor pathways that produce the conditioned responses (McCormick and Thompson 1984), (2) presentation of a CS is conveyed to the cerebellum via activation of certain of its mossy fiber inputs (Steinmetz et al. 1986; Hesslow et al. 1999), and (3) presentation of the US is conveyed via activation of certain climbing fiber inputs to the cerebellum (Fig. 1A; Mauk et al. 1986). These factors are complemented by the extent to which the synaptic organization and physiology of the cerebellum are known (Eccles et al. 1967; Ito 1984), as are the behavioral properties of eyelid conditioning (Gormezano et al. 1983; Kehoe and Macrae 2002). These advantages combine with the speed of current computers to make possible the construction of biologically detailed and large-scale computer simulations of the cerebellum that can then be thoroughly tested using standard eyelid conditioning protocols (Medina et al. 2000, 2001, 2002; Medina and Mauk 1999, 2000).Open in a separate windowFigure 1Emulation of eyelid conditioning in a computer simulation of the cerebellum. (A) A schematic representation of the simulation and how it was trained using an eyelid conditioning-like protocol. The output of the simulation comes from the summed activity of the six cerebellar deep nucleus cells (blue box). The CS was conveyed to the simulation by phasic activation of 18 of the 600 mossy fibers and tonic activation of six mossy fibers (green box). The US was emulated by a brief excitatory conductance applied to the single climbing fiber. The remainder of the simulation consisted of 10,000 granule cells, 900 Golgi cells, 20 stellate/basket cells, and 20 Purkinje cells with essentials of the connectivity as shown. (B) Acquisition, extinction, and savings by the simulation. Each panel shows the equivalent of 10 d of acquisition (left panel), extinction (center), and reacquisition (right) training. Individual sweeps are averages of 10 trials, which are clustered together to approximate the equivalent of one daily session of eyelid conditioning. These sessions are numbered at the left, progressing from front to back. The blue portion of the sweeps denotes the presence of the CS. (C) The strength of the mossy fiber-to-nucleus synapses in the simulation over the three phases of training. The synapses that progressively increase in strength during acquisition and reacquisition and decrease during extinction are the six that are tonically activated by the CS. Note that extinction training only slowly and thus incompletely reverses the strengthened synapses. Savings during reacquisition in the simulation is largely attributable to this residual plasticity. The continued increase in the strength of these synapses does not produce a comparable increase in response amplitude, rather, it reflects the tendency for the network to transfer plasticity from cortex (pauses in Purkinje activity produced by LTD) to the nucleus (increased strength of mossy fiber-to-nucleus synapses). How long this process continues depends on a number of unknown factors.The present results are more easily appreciated with a brief review of previous studies (Medina et al. 2000, 2001, 2002) showing how the simulation emulates acquisition, extinction, and savings during reacquisition. These phenomena are shown for the simulation in the three panels of Figure 1B. The underlying essential elements can be summarized briefly. Presentation of a CS activates subsets of granule cells, and these subsets change somewhat over the duration of the CS. Paired training induces long-term depression (LTD) at CS-activated granule-to-Purkinje synapses that are activated when the US is presented. This leads to a learned and well timed decrease in the activity of Purkinje cells during the CS (Hesslow and Ivarsson 1994), which leads to the induction of long-term potentiation (LTP) at mossy fiber-to-nucleus synapses activated by the CS. As this plasticity develops, nucleus cells encounter during the CS strong excitation combined with release from inhibition and therefore discharge robustly, thereby driving the expression of conditioned responses (McCormick and Thompson 1984). These steps suggest that learning first occurs in the cerebellar cortex, before robust conditioned responses are seen. We have observed evidence for this latent learning in cerebellar cortex (Ohyama and Mauk 2001).During extinction, CS-activated granule-to-Purkinje synapses undergo LTP because their activation occurs in the absence of climbing fiber activity. The essential suppression of climbing fiber activity below the typical level of 1 Hz is produced by inhibition from cerebellar output (Sears and Steinmetz 1991; Hesslow and Ivarsson 1996; Kenyon et al. 1998a,b; Miall et al. 1998), which is robust during the expression of conditioned responses. This prediction of the simulation is supported by observations that blocking inhibition of climbing fibers prevents extinction (Medina et al. 2002).We have shown previously that savings during reacquisition results, at least in part, from plasticity in the cerebellar deep nucleus that is relatively resistant to extinction (Medina et al. 2001). The strengths of the CS-activated mossy fiber-to-nucleus synapses in the simulation are shown in Figure 1C for acquisition, extinction, and reacquisition. Because learned pauses in Purkinje cell activity are still present early in extinction training, the strengths of CS-activated mossy fiber-to-nucleus synapses continue to increase. Once conditioned responses are fully extinguished, due to the restoration of robust Purkinje cell activity during the CS via the induction of LTP at CS-activated granule-to-Purkinje synapses, then CS-activated mossy fiber-to-nucleus synapses begin to undergo LTD and decrease in strength. The rate at which these synapses decrease in strength with additional extinction training depends on unknown factors such as the level of Purkinje activity required for induction of LTD. These results show in principle, however, that plasticity in the cerebellar cortex is sufficient to extinguish conditioned responses, and that a network displaying fully extinguished conditioned responses can still contain strengthened mossy fiber-to-nucleus synapses. In the simulation, savings occur largely because this residual plasticity in the cerebellar nucleus enhances the conditioned responses produced by the relearning of decreased activity in the Purkinje cells. In support, we have shown in rabbits that plasticity in the cerebellar nucleus persists following extinction, and that a measure of the magnitude of this residual plasticity correlates with the rate of reacquisition (Medina et al. 2001).Here, we used the mechanisms of extinction in this simulation to stimulate discussion regarding the issue of extinction as unlearning versus extinction as new learning.     

Ventral lateral geniculate input to the medial pons is necessary for visual eyeblink conditioning in rats

The conditioned stimulus (CS) pathway that is necessary for visual delay eyeblink conditioning was investigated in the current study. Rats were initially given eyeblink conditioning with stimulation of the ventral nucleus of the lateral geniculate (LGNv) as the CS followed by conditioning with light and tone CSs in separate training phases. Muscimol was infused into the medial pontine nuclei (MPN) after each training phase to examine conditioned response (CR) retention to each CS. The spread of muscimol infusions targeting the MPN was examined with fluorescent muscimol. Muscimol infusions into the MPN resulted in a severe impairment in retention of CRs with the LGNv stimulation and light CSs. A less severe impairment was observed with the tone CS. The results suggest that CS information from the LGNv and light CSs is relayed to the cerebellum through the MPN. Retrograde tracing with fluoro-gold (FG) showed that the LGNv and nucleus of the optic tract have ipsilateral projections to the MPN. Unilateral inputs to the MPN from the LGNv and nucleus of the optic tract may be part of the visual CS pathway that is necessary for visual eyeblink conditioning.The neural substrates of associative motor learning have been studied extensively using eyeblink conditioning (Christian and Thompson 2003; Thompson 2005). Eyeblink conditioning is typically established by pairing a tone or light conditioned stimulus (CS) with an unconditioned stimulus (US) that elicits the eyeblink reflex. An eyeblink conditioned response (CR) emerges over the course of paired training, and the peak of eyelid closure occurs at the onset time of the US. Results from experiments using temporary lesions of the cerebellar deep nuclei or cerebellar cortex indicate that the anterior interpositus nucleus and cerebellar cortex are necessary for acquisition, expression, and extinction of eyeblink conditioning (Krupa et al. 1993; Hardiman et al. 1996; Krupa and Thompson 1997; Garcia and Mauk 1998; Medina et al. 2001; Bao et al. 2002; Freeman et al. 2005a). Blocking cerebellar output with inactivation of the superior cerebellar peduncle, red nucleus, or brainstem motor nuclei selectively blocks CR expression but not acquisition, providing further evidence that learning occurs in the cerebellum (Chapman et al. 1990; Krupa et al. 1993, 1996; Krupa and Thompson 1995).Sensory stimuli from every modality are sent to the pontine nuclei (PN), which receive projections from the lower brainstem, thalamus, and cerebral cortex (Glickstein et al. 1980; Brodal 1981; Mihailoff et al. 1989; Schmahmann and Pandya 1989; Wells et al. 1989; Knowlton et al. 1993; Campolattaro et al. 2007). Neurons in the PN project CS information to the cerebellum via mossy fibers in the middle cerebellar peduncle that synapse on granule cells in the cerebellar cortex and on neurons in the interpositus nucleus (Bloedel and Courville 1981; Brodal 1981; Steinmetz and Sengelaub 1992; Mihailoff 1993). Lesions of the middle cerebellar peduncle impair eyeblink conditioning with auditory, somatosensory, and visual CSs (Lewis et al. 1987). Bilateral electrolytic lesions of the dorsolateral and lateral pontine nuclei (LPN) block retention of CRs to an auditory CS but have no effect on light-elicited CRs (Steinmetz et al. 1987). Inactivation of the contralateral LPN blocks CRs to a tone CS but not to lateral reticular nucleus stimulation in rabbits (Bao et al. 2000). Moreover, stimulation of the LPN or middle cerebellar peduncle is a sufficient CS for eyeblink conditioning (Steinmetz et al. 1986, 1987; Tracy et al. 1998; Bao et al. 2000; Freeman and Rabinak 2004; Freeman et al. 2005b; Campolattaro and Freeman 2008). The findings from the lesion, inactivation, and stimulation studies provide evidence that the PN is the proximal part of the CS pathway for cerebellar learning. These studies also indicate that the LPN is the primary source of auditory CS input to the cerebellum.Only a few studies have examined the visual CS pathway necessary for eyeblink conditioning. The dorsal and ventral divisions of the lateral geniculate nucleus of the thalamus (LGNd, LGNv), pretectal nuclei, visual cortex (VCTX), and superior colliculus (SC) comprise a hypothesized parallel visual CS pathway for eyeblink conditioning (Koutalidis et al. 1988). Combined lesions of all of these visual areas completely block acquisition, lesions of two visual areas produce a partial impairment, and lesions in one visual area do not impair CR acquisition (Koutalidis et al. 1988). Stimulation of the VCTX, SC, and LGNv support eyeblink conditioning, and each of these structures has a direct unilateral projection to the PN that could be important for eyeblink conditioning (Halverson et al. 2009). The lesion and stimulation studies provide evidence that structures in the hypothesized visual CS pathway are independently capable of supporting conditioning. An important aspect of the visual CS pathway proposed in the Koutalidis et al. (1988) study is distributed projections of each visual area to different regions of the PN. The important projections were hypothesized to be from the VCTX to the rostral portion of the PN, from both the SC and pretectal nuclei to the dorsolateral PN, and the LGNv projection to the medial pontine nuclei (MPN) (Koutalidis et al. 1988). Lesions of the VCTX were substituted for LGN lesions in the Koutalidis et al. (1988) study due to technical problems with animal survival. The LGNv projection to the MPN was therefore not examined in their combined lesion group. Stimulation of the anterior pretectal nucleus is not a sufficient CS to support eyeblink conditioning (Campolattaro et al. 2007). The direct PN projection from the VCTX is not necessary for CR retention to a light CS, as lesions do not prevent eyeblink conditioning to a light CS in dogs or monkeys (Hilgard and Marquis 1935, 1936). Moreover, lesions of the entire cerebral cortex do not prevent acquisition or retention of delay eyeblink conditioning to a tone or light CS in rabbits (Oakley and Russell 1972, 1977). The LGNv and SC, therefore, are likely sources of visual input to the PN that is necessary for eyeblink conditioning.The current experiment investigated whether information from the LGNv and a visual CS (light) share similar inputs into the MPN and whether those inputs are different from an auditory CS. The visual projections to the MPN were also investigated with the retrograde tracer fluoro-gold (FG) to identify structures that may be involved with the relay of CS information during eyeblink conditioning. In the conditioning experiment, rats received three phases of training, with each phase consisting of three acquisition sessions followed by a muscimol infusion into the MPN, and then a saline recovery session. Each rat received unilateral stimulation of the LGNv (contralateral to the trained eye) during phase 1 of training followed by either a tone or light CS in phases 2 and 3 (order of stimulus presentation was counterbalanced). One group received LGNv stimulation in phase 1 followed by a light CS in phase 2, and a tone CS in phase 3 (SLT). The other group received the tone CS in phase 2, and light CS in phase 3 (STL).     

Interactions between prefrontal cortex and cerebellum revealed by trace eyelid conditioning

Eyelid conditioning has proven useful for analysis of learning and computation in the cerebellum. Two variants, delay and trace conditioning, differ only by the relative timing of the training stimuli. Despite the subtlety of this difference, trace eyelid conditioning is prevented by lesions of the cerebellum, hippocampus, or medial prefrontal cortex (mPFC), whereas delay eyelid conditioning is prevented by cerebellar lesions and is largely unaffected by forebrain lesions. Here we test whether these lesion results can be explained by two assertions: (1) Cerebellar learning requires temporal overlap between the mossy fiber inputs activated by the tone conditioned stimulus (CS) and the climbing fiber inputs activated by the reinforcing unconditioned stimulus (US), and therefore (2) trace conditioning requires activity that outlasts the presentation of the CS in a subset of mossy fibers separate from those activated directly by the CS. By use of electrical stimulation of mossy fibers as a CS, we show that cerebellar learning during trace eyelid conditioning requires an input that persists during the stimulus-free trace interval. By use of reversible inactivation experiments, we provide evidence that this input arises from the mPFC and arrives at the cerebellum via a previously unidentified site in the pontine nuclei. In light of previous PFC recordings in various species, we suggest that trace eyelid conditioning involves an interaction between the persistent activity of delay cells in mPFC-a putative mechanism of working memory-and motor learning in the cerebellum.Eyelid conditioning is a form of associative learning that has proven useful for mechanistic studies of learning (Thompson 1986). All variants of eyelid conditioning involve pairing a conditioned stimulus (CS, typically a tone) with a reinforcing unconditioned stimulus (US, mild electrical stimulation near the eye) to promote learned eyelid closure in response to the CS (also known as a conditioned response). Delay eyelid conditioning, where the CS and US overlap in time (Fig. 1A , left), is largely unaffected by forebrain lesions (Solomon et al. 1986; Mauk and Thompson 1987; Kronforst-Collins and Disterhoft 1998; Weible et al. 2000; Powell et al. 2001; McLaughlin et al. 2002) and engages the cerebellum relatively directly (but see Halverson and Freeman 2006). Presentation of the tone and the US are conveyed to the cerebellum via activation of mossy fibers and climbing fibers, respectively (Fig. 1B; Mauk et al. 1986; Steinmetz et al. 1987, 1989; Sears and Steinmetz 1991; Hesslow 1994; Hesslow et al. 1999). In addition, output via a cerebellar deep nucleus is required for the expression of conditioned responses (McCormick and Thompson 1984). This relatively direct mapping of stimuli onto inputs and of output onto behavior makes delay eyelid conditioning a powerful tool for the analysis of cerebellar learning and computation (Mauk and Donegan 1997; Medina and Mauk 2000; Medina et al. 2000, 2002; Hansel et al. 2001; Ohyama et al. 2003).Open in a separate windowFigure 1.The procedures, neural pathways, and putative signals involved in delay and trace eyelid conditioning. (A) Stimulus timing for delay (left) and trace (right) training trials. For delay conditioning, the US overlaps in time with the tone CS. In this and subsequent figures, green is used to indicate the presentation of the CS for delay conditioning. For trace conditioning, the US is presented after CS offset, and “trace interval” refers to the period between CS offset and US onset. For convenience, we used red and maroon regions to represent the CS and trace interval, respectively. Sample conditioned eyelid responses are shown below, for which an upward deflection indicates closure of the eyelid. (B) Schematic representation of the pathways engaged by delay conditioning. The CS and US, respectively, engage mossy fibers and climbing fibers relatively directly, and forebrain input is not required for normal learning. (C) The signals hypothesized to engage the cerebellum during trace conditioning. The activity of mossy fibers directly activated by the tone CS does not significantly outlast the stimulus. Thus, a forebrain structure is thought to provide an input that overlaps in time with the US and is necessary to produce cerebellar learning.Trace eyelid conditioning, where the US is presented after tone offset (Fig. 1A, right), has attracted interest for its potential to reveal the nature of interactions between the forebrain and cerebellum as well as the learning mechanisms within these systems. This potential stems from the sensitivity of trace conditioning not only to lesions of cerebellum but also to lesions of hippocampus, medial prefrontal cortex (mPFC), or mediodorsal thalamic nucleus (Woodruff-Pak et al. 1985; Moyer Jr. et al. 1990; Kronforst-Collins and Disterhoft 1998; Weible et al. 2000; Powell et al. 2001; McLaughlin et al. 2002; Powell and Churchwell 2002; Simon et al. 2005). Given the general inability of forebrain lesions to affect delay conditioning, these results have promoted the general interpretation that the forebrain and cerebellum interact to mediate trace conditioning (Weiss and Disterhoft 1996; Clark and Squire 1998; Clark et al. 2002).Here we test the specific hypotheses that (Fig. 1C) (1) cerebellar learning requires that mossy fiber and climbing fiber inputs overlap in time (or nearly so) and (2) that cerebellar learning in trace conditioning occurs in response to a forebrain-driven mossy fiber input that outlasts the CS to overlap with the US rather than the inputs activated by the tone CS (Clark et al. 2002). The data provide direct support for both assertions and, together with recent anatomical studies (Buchanan et al. 1994; Weible et al. 2007), reveal a pathway between the mPFC and cerebellum that is necessary for the expression of trace eyelid responses. When combined with previous recordings from PFC in primates and rodents (Funahashi et al. 1989; Bodner et al. 1996; Fuster et al. 2000; Narayanan and Laubach 2006), these data support the hypothesis that trace eyelid conditioning is mediated by interactions between working memory-related persistent activity in mPFC and motor learning mechanisms in the cerebellum.     

Genetic inactivation of D-amino acid oxidase enhances extinction and reversal learning in mice

Activation of the N-methyl-d-aspartate receptor (NMDAR) glycine site has been shown to accelerate adaptive forms of learning that may benefit psychopathologies involving cognitive and perseverative disturbances. In this study, the effects of increasing the brain levels of the endogenous NMDAR glycine site agonist D-serine, through the genetic inactivation of its catabolic enzyme D-amino acid oxidase (DAO), were examined in behavioral tests of learning and memory. In the Morris water maze task (MWM), mice carrying the hypofunctional Dao1^G181R mutation demonstrated normal acquisition of a single platform location but had substantially improved memory for a new target location in the subsequent reversal phase. Furthermore, Dao1^G181R mutant animals exhibited an increased rate of extinction in the MWM that was similarly observed following pharmacological administration of D-serine (600 mg/kg) in wild-type C57BL/6J mice. In contextual and cued fear conditioning, no alterations were found in initial associative memory recall; however, extinction of the contextual fear memory was facilitated in mutant animals. Thus, an augmented level of D-serine resulting from reduced DAO activity promotes adaptive learning in response to changing conditions. The NMDAR glycine site and DAO may be promising therapeutic targets to improve cognitive flexibility and inhibitory learning in psychiatric disorders such as schizophrenia and anxiety syndromes.The N-methyl-d-aspartate receptor (NMDAR) has an important role in excitatory neurotransmission and contributes to numerous brain processes, including synaptic plasticity, learning, and memory formation (Nicoll 2003). Activation of NMDARs requires membrane depolarization in addition to concurrent binding of glutamate to NMDAR2 (NR2) and glycine to the NMDAR1 (NR1) subunit (Johnson and Ascher 1987; Clements and Westbrook 1991). D-serine has also been shown to be an endogenous co-agonist for the NR1 glycine site, acting with high selectivity and a potency similar to or greater than that of glycine (Matsui et al. 1995). In the brain, the localization of D-serine closely resembles that of NMDARs (Schell et al. 1997), and D-serine has been reported to be the predominant physiologic co-agonist for the maintenance of NMDAR-mediated currents in the hippocampus, retina, and hypothalamus (Mothet et al. 2000; Yang et al. 2003). Moreover, in vivo studies have demonstrated that the NMDAR glycine site is not saturated at the synapses of several brain regions (Fuchs et al. 2005). Consequently, increasing D-serine levels may modulate neurotransmission and behavioral responses reliant on NMDAR activity.The NMDAR glycine site has been implicated in the pathophysiology and treatment of a number of psychiatric conditions (Coyle and Tsai 2004; Millan 2005). Blockade of the NMDAR with noncompetitive antagonists like phencyclidine results in the production and exacerbation of schizophrenic-like symptoms in humans and animals (Javitt and Zukin 1991; Krystal et al. 1994). Genetic studies have associated genes that mediate D-serine synthesis and degradation with a vulnerability to schizophrenia, and levels of D-serine are decreased in the CSF and serum of schizophrenic patients (Chumakov et al. 2002; Hashimoto et al. 2003, 2005; Schumacher et al. 2004; Morita et al. 2007). These observations prompted clinical trials with direct and indirect activators of the NMDAR glycine site, including D-serine, and improvements were revealed when these compounds were added to conventional antipsychotic regimes, particularly with the negative and cognitive symptoms of schizophrenia (Tsai et al. 1998; Coyle and Tsai 2004; Heresco-Levy et al. 2005). Furthermore, altered NMDAR activation has also been shown to affect extinction, a learning process that may be of benefit in anxiety illnesses, such as post-traumatic stress syndrome and obsessive-compulsive disorder (Davis et al. 2006). In rodents, extinction was shown to be impaired following inhibition of NMDARs in contextual fear conditioning, inhibitory avoidance, and eyeblink conditioning tasks (Kehoe et al. 1996; Lee and Kim 1998; Szapiro et al. 2003). In contrast, the partial NMDAR agonist D-cycloserine facilitated the extinction of fear memories in rodents and individuals with phobias and other anxiety disorders (Ressler et al. 2004; Ledgerwood et al. 2005; Norberg et al. 2008). Thus, the NMDAR glycine site and its related modulatory proteins may be important targets for the amelioration of psychopathologies involving cognitive dysfunction and maladaptive behaviors.Endogenous levels of D-serine in the brain are regulated by its catabolic enzyme, D-amino acid oxidase (DAO); by the D-serine synthesis enzyme, serine racemase (Srr); and by neuronal and glial transporters (Foltyn et al. 2005; Martineau et al. 2006). Agents targeting such proteins may prove to be an effective method of increasing cerebral D-serine and occupancy of the NMDAR glycine site, which could overcome the difficulties D-serine and similar compounds have with penetrating the blood-brain barrier (Coyle and Tsai 2004; Bauer et al. 2005). Inhibiting DAO function in the brain is of particular interest as it would circumvent any nephrotoxicity associated with high levels of systemic D-serine (Maekawa et al. 2005a). DAO is a peroxisomal flavoprotein that at physiological pH is highly selective for D-serine, and in the brain, DAO is located predominantly in astrocytes (Mothet et al. 2000). An inverse correlation between the brain distribution of DAO and D-serine evinces the efficacy of this enzyme, with the most abundant DAO expression located in the D-serine-sparse hindbrain and cerebellum (Schell et al. 1995; Moreno et al. 1999). To study the effects of limiting DAO function, we tested a line of mice carrying a single point mutation (G181R) that results in a complete lack of DAO activity and consequently augmented D-serine in serum and brain (Sasaki et al. 1992; Hashimoto et al. 1993). These mice have previously been shown to exhibit an in vitro increase in NMDAR-mediated excitatory postsynaptic currents in dorsal horn neurons of the spinal cord and an in vivo elevation of cGMP that is indicative of augmented NMDAR activity (Wake et al. 2001; Almond et al. 2006). This demonstrates that reduced DAO function is capable of augmenting NMDAR activation, and it may follow that cognitive and extinction processes influenced by NMDARs are enhanced in Dao1^G181R mutant mice. To investigate this possibility, we assessed the effects of the Dao1^G181R mutation on learning, memory, and extinction in Morris water maze (MWM) and in contextual and cued fear conditioning paradigms.     

Making context memories independent of the hippocampus

We present evidence that certain learning parameters can make a memory, even a very recent one, become independent of the hippocampus. We confirm earlier findings that damage to the hippocampus causes severe retrograde amnesia for context memories, but we show that repeated learning sessions create a context memory that is not vulnerable to the damage. The findings demonstrate that memories normally dependent on the hippocampus are incrementally strengthened in other memory networks with additional learning. The latter provides a new account for patterns of hippocampal retrograde amnesia and how memories may become independent of the hippocampus.Contextual fear conditioning can be supported by two neural systems, one that contains the hippocampus (HPC), and one that does not. Evidence for this assertion comes from studies in which the HPC, in rats, is damaged either before or after the contextual fear conditioning. Extensive damage to the HPC before conditioning has little effect on contextual fear conditioning (Maren et al. 1997; Frankland et al. 1998; Wiltgen et al. 2006). This result can only mean that there is a non-HPC memory system that can support fear of context. In contrast, there is unequivocal evidence that moderate to extensive damage to the HPC soon after learning severely impairs the ability of the conditioning context to evoke fear, suggesting that the HPC normally makes a major contribution to this type of memory (Kim and Fanselow 1992; Maren et al. 1997; Frankland et al. 1998; Anagnostaras et al. 1999; Debiec et al. 2002; Lehmann et al. 2007b; Sutherland et al. 2008; Wang et al. 2009).The dissociable effects of pre- and post-training HPC damage on contextual fear conditioning have been interpreted as suggesting that: (1) When the HPC is intact during learning it interferes with other systems and prevents them from acquiring an independent contextual fear conditioning memory, and (2) when the HPC is absent, these other systems are released from this interference and are able to rapidly acquire an independent memory (Maren et al. 1997; Frankland et al. 1998; Fanselow and Poulos 2004; Driscoll et al. 2005; Lehmann et al. 2006; Sutherland et al. 2006). The latter interference from the HPC on the other memory systems has been termed overshadowing. Supplemental Figure S1 depicts data from our laboratory demonstrating the overshadowing phenomenon and the dissociable effects of HPC damage induced before and after contextual fear conditioning.Very little, however, is known about the parameters determining the extent to which the HPC system interferes with the non-HPC system for control over contextual fear. The purpose of the current study is to provide some insight into this issue. Typically, contextual fear conditioning in rats is conducted in a single conditioning session in which a configuration of static background cues is paired with several footshocks. When returned to the conditioning context, rats display several species-specific defensive responses including freezing (i.e., absence of movement except for breathing). Several theorists have proposed that non-HPC systems are more likely to be recruited when there are multiple experiences with similar events, which, in turn, would mitigate the necessity of the HPC for memory expression (O''Keefe and Nadel 1978; Sherry and Schacter 1987; McClelland et al. 1995; O''Reilly and Rudy 2001; White and McDonald 2002). Accordingly, we hypothesized that repeated contextual fear conditioning sessions separated by hours and days would overcome the HPC interference or overshadowing effect. In other words, with repeated learning sessions, enough information would be incrementally captured by the non-HPC system to support a contextual fear memory that would survive complete damage to the HPC.Adult male rats received 11 fear-conditioning sessions across 6 d. In each session, they were placed in a context and received mild footshocks (Shock Context). Concurrently, the rats were exposed 10 times to another context in which they never received shock (No-Shock Context). The No-Shock Context served as a control condition to measure whether the rats simply showed generalized fear or could show context-specific memory. Within 72 h following the last conditioning session, rats either received sham surgery or complete lesions of the HPC using the neurotoxin N-methyl-d-aspartic acid (NMDA) (Lehmann et al. 2007a). Rats were then tested for retention in both the Shock and No-Shock Contexts in a counterbalanced order. In addition, in a single learning episode, another group of rats received a matching number of shocks (i.e., 12 shocks) and context exposure (i.e., 17 min), and then received surgery 7–10 d after conditioning. The latter interval is identical to the interval between the initial conditioning session and surgery in the repeated learning condition. Figure 1 illustrates and describes the design of the experiments.Open in a separate windowFigure 1.Illustration of the experimental design used in (A) the single conditioning session and (B) repeated conditioning session experiments. In A the rats were initially placed in the conditioning chamber for 17 min and received the first of 12 footshocks (1 mA/2 sec) at the 300-sec mark, and then one every following 58 sec after shock offset. Seven to 10 d later, the rats either received sham or HPC damage (Sx). Approximately 10 d after, the rats were returned to the chamber to assess freezing over a 5-min retention test. In B the rats were placed initially in the conditioning chamber for 1 min and received a shock at the 45-sec mark (Shock Context). Approximately 45 min later, the rats were placed in a different chamber for 1 min and did not receive shock (No-Shock Context). The procedure was repeated twice daily for five consecutive days, and the Shock and No-Shock chamber order was counterbalanced according to the principles of a Latin Square design. The rats then received sham or HPC damage 1–3 d later. The rats'' retention was assessed in both contexts ∼10 d after surgery in both the Shock and No-Shock Context in a counterbalanced order with a 24-h span between tests. Importantly, the number of shocks, context exposure time, and interval between initial learning and surgery were matched between both experiments.When all shocks were delivered in a single session, HPC damage caused profound retrograde amnesia. As illustrated in Figure 2A, the HPC rats displayed significantly less freezing than control rats during the retention test (t₍₈₎ = 23.895, P < 0.001). This result replicates all previous studies in which the HPC was damaged days after a single contextual fear conditioning training session (Kim and Fanselow 1992; Maren et al. 1997; Frankland et al. 1998; Anagnostaras et al. 1999; Debiec et al. 2002; Lehmann et al. 2007b; Sutherland et al. 2008).Open in a separate windowFigure 2.Mean (± SEM) percent time freezing by Sham and HPC rats during the retention test of the (A) single conditioning (12 shocks) experiment and (B) repeated conditioning session experiment. In A the HPC rats showed significantly less freezing (P < 0.001) than the Sham rats, suggesting that the damage caused profound retrograde amnesia for contextual fear conditioning learned in a single session 7–10 d before surgery. In B the performance of the HPC rats did not significantly differ from the Sham rats, and they exhibited significantly more freezing in the Shock Context than the No-Shock Context (P < 0.001). Consequently, repeated conditioning sessions prevented the retrograde amnesic effects normally observed in contextual fear conditioning following HPC damage, suggesting that other neural networks were now able to support the memory.In striking contrast, memory for contextual fear conditioning was spared when the HPC was damaged after repeated conditioning sessions. Figure 2B shows the percent time spent freezing during the retention test in the Shock and No-Shock Contexts. An ANOVA with between-group factor (Lesion: Sham and HPC) and within-group factor (Context: Shock and No-Shock) revealed a significant main effect of Context (F_(1,14) = 84.731, P < 0.001), indicating that the rats displayed higher levels of freezing in the Shock than in the No-Shock Context. The effect of Lesion (F_(1,14) = 4.280, P = 0.058) was not significant, nor was the Lesion × Context interaction (F_(1,14) = 0.877, P = 0.369), suggesting that extensive HPC damage did not impair memory. The tendency for an effect of Lesion is due to the HPC rats freezing less than the Sham rats in the No-Shock Context (P = 0.06) rather than freezing less in the Shock Context (P = 0.457).The repeated conditioning sessions clearly enabled a contextual fear representation to be established in non-HPC memory systems. However, it is surprising that the HPC damage did not impair the ability to discriminate between the Shock and No-Shock Context, because evidence suggests that context discrimination is dependent on the HPC (see Moscovitch et al. 2006). Indeed, studies of rats with HPC damage induced before learning have shown that contextual fear conditioning is acquired quickly by non-HPC systems in a single session, but the ability to discriminate between the training context and a new context is lost (Frankland et al. 1998; Antoniadis and McDonald 2000; Winocur et al. 2007). Hence, it is significant in the present study that the HPC damage did not impair context discrimination abilities in the rats that received repeated learning episodes. The latter appear to have established a context representation, outside of the HPC, that was not bereft of details. Yet, one should consider that the rats in the repeated sessions experiment received experience in both the Shock and No-Shock Contexts prior to surgery, and this discrimination training procedure may have established two different non-HPC representations. It remains possible that HPC damage would impair the ability to discriminate the Shock Context from a new context, which is what is found in anterograde amnesia studies (Frankland et al. 1998; Antoniadis and McDonald 2000; Winocur et al. 2007). To address this possibility, a new experiment examined whether HPC-damaged rats could discriminate the Shock Context from a Novel Context. Rats were trained with the same repeated learning protocol as described earlier, with the exception that the rats were never placed in the No-Shock Context prior to surgery. One to 3 d following learning, the rats either received Sham or complete HPC damage. They were then tested for retention in the Shock and the Novel (i.e., No-Shock) Context in a counterbalanced order. Figure 3 shows the percent time spent freezing during the retention test in the Shock and Novel Contexts. An ANOVA with between-group factor (Lesion: Sham and HPC) and within-group factor (Context: Shock and Novel) revealed that the rats froze significantly more in the Shock than the Novel Context (F_(1,10) = 57.393, P < 0.001). However, no significant difference was found between the HPC and Sham groups (F_(1,10) = 0.597, P = 0.458) and the Lesion × Context interaction did not reach significance (F_(1,10) = 0.123, P = 0.733). Thus, as in the previous repeated sessions experiment, the HPC damage did not cause retrograde amnesia for contextual fear conditioning and, more importantly, the HPC damage did not impair the ability to discriminate between the original context and new context.Open in a separate windowFigure 3.Mean (± SEM) percent time freezing by Sham and HPC rats in the Shock and Novel Contexts during the retention tests of the discrimination experiment. The rats exhibited significantly more freezing in the Shock than the Novel Context (P < 0.001), and the HPC rats did not significantly differ from the Sham rats, suggesting that the HPC-damaged rats remembered the specific meaning of the Shock Context as well as control rats. Hence, repeated conditioning sessions established a context-rich representation in non-HPC systems, which supports successful context discriminations.The absence of amnesia for contextual fear conditioning in the current study is not due to insufficient damage to the HPC. We calculated (see Lehmann et al. 2007b) that an average of 83% of the HPC was damaged across rats (smallest: 64%; largest: 90%) in the repeated learning experiments (see Supplemental material for more histological details). The amount of HPC damage is substantially more than that found in most studies reporting impairments for contextual fear conditioning following HPC damage (Kim and Fanselow 1992; Maren et al. 1997; Frankland et al. 1998; Anagnostaras et al. 1999; Debiec et al. 2002) and more than for the single-session experiment (average 76%) in which we currently report amnesia. Therefore, the amount of HPC damage inflicted in the rats in this study is certainly sufficient to disrupt HPC-dependent memories.Like others (Kim and Fanselow 1992; Maren et al. 1997; Frankland et al. 1998; Anagnostaras et al. 1999; Debiec et al. 2002; Lehmann et al. 2007b; Sutherland et al. 2008), we found that damage to the HPC after a single contextual fear-conditioning session involving multiple shocks produces profound retrograde amnesia for contextual fear conditioning. However, in two separate experiments, distributing shock across multiple conditioning sessions prevented this amnesia. In one case, the rats experienced Context–Shock pairings in one context and no shock in another context. Following this training, rats with damage to the HPC did not differ from control rats in the absolute amount of freezing in the training context nor in their ability to discriminate between the two contexts. In the second case, rats only received the multiple Context–Shock sessions. Rats with damage to the HPC could not be distinguished from control rats during the test in the training context or in their responses to a novel context. These findings provide new support for the general idea that contextual fear conditioning can be supported by both HPC and non-HPC systems. This conclusion is supported by (1) the finding that damage to the HPC following a single conditioning session virtually eliminates freezing during the test, implying the importance of the HPC system, and (2) that following multiple conditioning sessions, damage to the HPC has no effect on either contextual fear displayed in the training context or their ability to discriminate the training context from other contexts, suggesting the existence of non-HPC systems that can support contextual fear. The findings also reveal that the overshadowing or interference by the HPC over the non-HPC memory systems for control over contextual fear is not absolute. Following a single conditioning session, removal of the HPC produced a devastating retrograde amnesia, illustrating substantial overshadowing. However, distributing conditioning across several sessions completely attenuated the effects of damage to the HPC, revealing that non-HPC systems can support contextual fear conditioning despite the HPC, and revealed the importance of multiple sessions for this to occur.The overshadowing by the HPC is based on the familiar idea in associative learning at the behavioral level, where through a competitive process some of the cues that redundantly predict a reinforcer acquire the ability to generate strong conditioned responding, while other equally predictive, but less salient cues do not (Stout et al. 2003). Conditioning to the less potent cues proceeds more effectively if the more potent competitors are absent. Following the same principle, if the HPC representation is active, then learning in the non-HPC systems suffers strong interference. In contrast, in the absence of the HPC representation, learning in non-HPC systems is released from this interfering effect of the HPC. Thus, the learning rate in non-HPC networks is potently lowered by the activity of the HPC. However, with repeated learning, other structures, which are overshadowed by the HPC, may cumulatively build a representation that achieves HPC independence. The current findings clearly support this hypothesis, whereby repeated learning episodes incrementally established a contextual fear-conditioning representation outside of the HPC that mitigated the usual retrograde amnesic effects of HPC damage.One important question is where does the HPC interference occur? Biedenkapp and Rudy (2009) recently reported that the HPC competes with the basolateral region of the amygdala during fear conditioning. Previously, Guarraci et al. (1999) found that the amount of conditioned fear produced by training could be increased if the dopamine D1 receptor agonist {"type":"entrez-protein","attrs":{"text":"SKF82958","term_id":"1156217255","term_text":"SKF82958"}}SKF82958 was injected into the basolateral region. Biedenkapp and Rudy (2009) reasoned that if this is the area where the HPC interferes with non-HPC systems for the association with shock, then a local infusion of {"type":"entrez-protein","attrs":{"text":"SKF82958","term_id":"1156217255","term_text":"SKF82958"}}SKF82958 before a single session of contextual fear conditioning should attenuate the interference and allow the non-HPC system to gain more control over contextual fear. Their data supported this hypothesis, which leads to the possibility that with multiple conditioning sessions, the non-HPC system gradually gains association with these fear-supporting neurons in this region of the brain.Patients with bilateral damage to the HPC often exhibit temporally graded retrograde amnesia, such that recently acquired memories are lost, whereas remote memories, especially those acquired years before the damage, are more likely to be spared (Scoville and Milner 1957; Rempel-Clower et al. 1996). This pattern of amnesia is taken as evidence for temporally based systems consolidation, whereby over time the essential support for memories is “switched” from dependence on the HPC to neocortical networks (McClelland et al. 1995; Squire and Alvarez 1995; Anagnostaras et al. 2001; Meeter and Murre 2004; Squire et al. 2004; Wiltgen et al. 2004; Frankland and Bontempi 2005). Our research, however, points to another process for becoming independent of the HPC, a change in the strength of the representation in non-HPC systems during learning rather than a consolidation process linked to the passage of time since the learning episode. A study of a former London taxi driver with bilateral HPC damage alludes to this possibility (Maguire et al. 2006). This amnesic patient showed greater retrograde amnesia for roads that he used less commonly than the major arteries that he used regularly. Hence, greater exposure to the major arteries established memories in non-HPC systems, whereas roads with less exposure remained dependent on the HPC regardless of the age of the memory. Our findings add support to this view, because studies examining the effects of complete HPC damage after a single conditioning episode suggest that the HPC is permanently involved in contextual fear conditioning (Lehmann et al. 2007b; Sutherland et al. 2008); yet, with repeated learning episodes we clearly demonstrated that the memory rapidly becomes independent of the HPC. The latter is important because the process for memories becoming independent of the HPC need not require systems consolidation.In conclusion, this is the first example of intact contextual fear memories following complete HPC damage induced soon after learning. Importantly, repetition of the learning episode underlies the change in memory from HPC dependent to HPC independent. We argue that each learning episode incrementally establishes a representation in non-HPC memory systems—a representation that ultimately becomes sufficiently strong to support memory expression without the HPC. The current findings also demonstrate the critical need to consider learning parameters when discussing patterns of retrograde amnesia and the role of the HPC in memory.     

Social modulation of associative fear learning by pheromone communication

Mice communicate through visual, vocal, and olfactory cues that influence innate, nonassociative behavior. We here report that exposure to a recently fear-conditioned familiar mouse impairs acquisition of conditioned fear and facilitates fear extinction, effects mimicked by both an olfactory chemosignal emitted by a recently fear-conditioned familiar mouse and by the putative stress-related anxiogenic pheromone β-phenylethylamine (β-PEA). Together, these findings suggest social modulation of higher-order cognitive processing through pheromone communication and support the concurrent excitor hypothesis of extinction learning.Social communication in mammals has evolved to facilitate reproductive behavior and for protection against environmental threat and predation. Mice communicate information about imminent danger through vocal (Seyfarth and Cheney 2003), visual (Kavaliers et al. 2001; Langford et al. 2006), and odor or pheromone cues (Rottman and Snowdon 1972), each with profound influences on defensive responding. There is also evidence of social empathy in mice (Langford et al. 2006). Mice will sensitize to pain-inducing stimuli simply by observing a conspecific that is currently experiencing pain. Importantly, sensitization occurs only when the conspecific is familiar with the observer (i.e., sibling or cage mate), a clear example of social modulation of an innate behavior. Müller-Velten (1966) provided the first evidence of a functional alarm chemosignal in mice by showing that animals would avoid a pathway in which the odor of a stressed mouse was present. Subsequent studies have shown effects of mammalian olfactory chemosignals on a variety of defensive behaviors such as analgesia, vigilance, and avoidance (Rottman and Snowdon 1972; Mackay-Sim and Laing 1981; Fanselow 1985; Zalaquett and Thiessen 1991). To date, research on social modulation of behavior has focused primarily on observational learning and innate or nonassociative processes. Two recent studies have demonstrated an influence of fear-related chemosignals on associative learning in humans (Chen et al. 2006; Prehn et al. 2006), evidence that supports the hypothesis that social modulation of behavior extends to higher-order cognitive processing.In the following experiments, we asked whether exposure to a familiar mouse recently fear conditioned or trained for fear extinction would influence associative fear learning in a conspecific. We find that exposure to a recently fear-conditioned mouse impairs acquisition of conditioned fear, while the same experience facilitates the extinction of conditioned fear; effects mimicked by exposure to an olfactory chemosignal emitted from fear-conditioned mice and by the putative anxiogenic pheromone, β-phenylethylamine (β-PEA). Interestingly, we find that exposure to a recently extinction-trained mouse results in an inhibition of fear extinction learning, an effect not related to an olfactory chemosignal emitted by a recently extinguished mouse or by exposure to β-PEA. These data suggest that mice communicate information about their experience, in part through pheromone communication, with different effects on associative learning depending on the valence of the task.