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.     

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.     

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.     

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.     

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.     

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.     

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.     

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).     

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.     

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.

     

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.     

Odor-specific habituation arises from interaction of afferent synaptic adaptation and intrinsic synaptic potentiation in olfactory cortex

Segmentation of target odorants from background odorants is a fundamental computational requirement for the olfactory system and is thought to be behaviorally mediated by olfactory habituation memory. Data from our laboratory have shown that odor-specific adaptation in piriform neurons, mediated at least partially by synaptic adaptation between the olfactory bulb outputs and piriform cortex pyramidal cells, is highly odor specific, while that observed at the synaptic level is specific only to certain odor features. Behavioral data show that odor habituation memory at short time constants corresponding to synaptic adaptation is also highly odor specific and is blocked by the same pharmacological agents as synaptic adaptation. Using previously developed computational models of the olfactory system we show here how synaptic adaptation and potentiation interact to create the observed specificity of response adaptation. The model analyzes the mechanisms underlying the odor specificity of habituation, the dependence on functioning cholinergic modulation, and makes predictions about connectivity to and within the piriform neural network. Predictions made by the model for the role of cholinergic modulation are supported by behavioral results.Filtering sensory input is critical for information processing tasks such as background segmentation, and shifting processing power away from redundant, stable, or repetitive stimuli toward dynamic, novel stimuli. A critical aspect of this filtering however, is stimulus specificity. Under most circumstances it may be most beneficial to selectively filter the redundant stimulus, while maintaining responsiveness to different, though perhaps highly similar stimuli.In the olfactory system, short-term habituation to stable or repeated odorants involves a metabotropic glutamate receptor (mGluR)-dependent depression of afferent synapses to the piriform cortex (Best and Wilson 2004). Blockade of group III mGluR receptors prevents cortical adaptation odors (Best and Wilson 2004), and reduces short-term habituation of odor-evoked reflexes (Best et al. 2005) and odor investigation (Yadon and Wilson 2005; Bell et al. 2008; McNamara et al. 2008). This short-term habituation is highly odor specific, with minimal cross-adaptation of piriform cortical single-unit responses or cross-habituation of behavioral responses to similar odors, including between mixtures and their components (Wilson 2000; Cleland et al. 2002). Interestingly, there is an experience-dependent component to short-term habituation odor specificity. The odor specificity is most pronounced for familiar odors, with very brief (<20 sec) exposure to odors producing more generalization, and longer exposures (>50 sec) sufficient to permit strong odor specificity in cortex adaptation (Wilson 2003).The homosynaptic nature of afferent synaptic depression underlying cortical adaptation (Wilson 1998; Best and Wilson 2004) may contribute to this odor specificity. However, the experience dependence suggests that there may be an additional process involved. In fact, theoretical views of piriform cortical function suggest that the cortex learns previous patterns of input via potentiation of intracortical association fiber synapses (Hasselmo et al. 1990; Barkai et al. 1994; Haberly 2001; Linster et al. 2003). This autoassociative process essentially creates a template of previous network activity, against which new input patterns can be compared, allowing enhanced discrimination between similar patterns, as well as completion of degraded patterns (Barkai et al. 1994; Barnes et al. 2008). In support of this hypothesis, previous work has demonstrated that disruption of normal synaptic potentiation in association fiber synapses through blockade of cholinergic muscarinic receptors (Patil et al. 1998; Linster et al. 2003), reduces odor specificity of cortical adaptation (Wilson 2001b), prevents the effects of odor experience on subsequent behavioral cross-habituation (Fletcher and Wilson 2002), and disrupts odor discrimination (Linster et al. 2001).The present series of studies further explored the role of combined afferent synaptic depression and intracortical association fiber synaptic potentiation on the specificity of cortical adaptation and odor habituation. Using a computational model of the olfactory system (Linster et al. 2007), the results suggest that activity-dependent association fiber plasticity is necessary to account for the specificity of odor habituation. Furthermore, in behavioral experiments blockade of cholinergic muscarinic receptors during habituation enhances generalization of odor habituation, consistent with the modeling and with previous electrophysiological results.     

Role of amygdala and hippocampus in the neural circuit subserving conditioned defeat in Syrian hamsters

We examined the roles of the amygdala and hippocampus in the formation of emotionally relevant memories using an ethological model of conditioned fear termed conditioned defeat (CD). Temporary inactivation of the ventral, but not dorsal hippocampus (VH, DH, respectively) using muscimol disrupted the acquisition of CD, whereas pretraining VH infusions of anisomycin, a protein synthesis inhibitor, failed to block CD. To test for a functional connection between the VH and basolateral amygdala (BLA), we used a classic functional connectivity design wherein injections are made unilaterally in brain areas either on the same or opposite sides of the brain. A functional connection between the BLA and VH necessary for the acquisition of CD could not be found because unilateral inactivation of either BLA alone (but not either VH alone) was sufficient to disrupt CD. This finding suggested instead that there may be a critical functional connection between the left and right BLA. In our final experiment, we infused muscimol unilaterally in the BLA and assessed Fos immunoreactivity on the contralateral side following exposure to social defeat. Inactivation of either BLA significantly reduced defeat-induced Fos immunoreactivity in the contralateral BLA. These experiments demonstrate for the first time that whereas the VH is necessary for the acquisition of CD, it does not appear to mediate the plastic changes underlying CD. There also appears to be a critical interaction between the two BLAs such that bilateral activation of this brain area must occur in order to support fear learning in this model, a finding that is unprecedented to date.Our laboratory has taken a novel approach to examine the behavioral and physiological changes that accompany social experience by studying a striking behavioral response that is exhibited following social defeat. When a Syrian hamster is paired with a larger, more aggressive opponent and is defeated, it subsequently becomes highly submissive and fails to defend its own home cage even against a smaller, nonaggressive intruder. We call this change in the behavior of the defeated hamster conditioned defeat (CD) (Portegal et al. 1993) and believe that it is a valuable model with which to study neural and behavioral plasticity following exposure to a biologically relevant stressor.One of the critical structures subserving CD is the amygdala; temporary inactivation of its major subnuclei, including the basolateral amygdala (BLA), blocks the acquisition of CD (Jasnow and Huhman 2001). Together with the findings that protein synthesis inhibition in the BLA effectively disrupts CD (Markham and Huhman 2008) and that overexpression of cAMP response element binding protein (CREB) in the BLA enhances CD (Jasnow et al. 2005), the data support the hypothesis that the BLA is a critical site for plasticity related to CD.One brain region that we have largely overlooked, but which receives considerable attention for its role in learning and memory, is the hippocampus. Several groups have now gathered anatomical and behavioral data demonstrating functionally specific dissociation between the dorsal (DH) and ventral (VH) regions of the hippocampus (Risold and Swanson 1996; Moser and Moser 1998; Bannerman et al. 2004; McEown and Treit 2009). While the DH is critical for spatial relationships (O''Keefe and Nadel 1978; Moser et al. 1993; Eichenbaum 1996) and has been shown to play an important role in social recognition in hamsters (Lai et al. 2005), the VH appears to be involved in the production of behaviors produced in response to aversive stimuli (Trivedi and Coover 2004; Pentkowski et al. 2006).Considering how critically important the hippocampus and amygdala are in relation to fear and memory, some studies are beginning to suggest that these areas may functionally interact to modulate memory function (Akirav and Richter-Levin 2002; McGaugh et al. 2002; McGaugh 2004; Vouimba et al. 2007). The BLA projects to the hippocampus (Amaral and Insausti 1992), and high-frequency stimulation of the BLA combined with tetanic stimulation of the perforant pathway facilitates hippocampal long-term potentiation (LTP) (Ikegaya et al. 1996). Additionally, electrolytic lesions of the VH produce a deficit in the acquisition of fear to a contextual conditioned stimulus, and NMDA lesions of the BLA cause a nonselective deficit in the acquisition of fear to both contextual and acoustic conditioned stimuli (Maren and Fanselow 1995). Although our laboratory has previously demonstrated that the BLA is critically involved in the acquisition of CD (Jasnow and Huhman 2001), the role of the hippocampus has yet to be investigated. The aim of the present study was to examine whether the VH and DH are involved in mediating CD and also to determine whether there is a functional interaction between the hippocampus and the amygdala in the acquisition of CD.     

The NMDA antagonist MK-801 disrupts reconsolidation of a cocaine-associated memory for conditioned place preference but not for self-administration in rats

Recent research suggests that drug-related memories are reactivated after exposure to environmental cues and may undergo reconsolidation, a process that can strengthen memories. Conversely, reconsolidation may be disrupted by certain pharmacological agents such that the drug-associated memory is weakened. Several studies have demonstrated disruption of memory reconsolidation using a drug-induced conditioned place preference (CPP) task, but no studies have explored whether cocaine-associated memories can be similarly disrupted in cocaine self-administering animals after a cocaine priming injection, which powerfully reinstates drug-seeking behavior. Here we used cocaine-induced CPP and cocaine self-administration to investigate whether the N-methyl-D-aspartate receptor antagonist (+)-5methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine maleate (MK-801) given just prior to reactivation sessions would suppress subsequent cocaine-primed reinstatement (disruption of reconsolidation). Systemic injection of MK-801 (0.05 or 0.20 mg/kg administered intraperitoneally) in rats just prior to reactivation of the cocaine-associated memory in the CPP context attenuated subsequent cocaine-primed reinstatement, while no disruption occurred in rats that did not receive reactivation in the CPP context. However, in rats trained to self-administer cocaine, systemic administration of MK-801 just prior to either of two different types of reactivation sessions had no effect on subsequent cocaine-primed reinstatement of lever-pressing behavior. Thus, systemic administration of MK-801 disrupted the reconsolidation of a cocaine-associated memory for CPP but not for self-administration. These findings suggest that cocaine-CPP and self-administration do not use similar neurochemical processes to disrupt reconsolidation or that cocaine-associated memories in self-administering rats do not undergo reconsolidation, as assessed by lever-pressing behavior under cocaine reinstatement conditions.The ability to disrupt previously consolidated memories in a reactivation-dependent manner is thought to be due to the disruption of a memory reconsolidation process. This disruption of reconsolidation has been observed in a wide variety of tasks and species (Nader et al. 2000b; Sara 2000; Alberini 2005; Riccio et al. 2006). Early reconsolidation experiments primarily focused on aversive learning paradigms, with an emphasis on disruption of reconsolidation as a potential treatment for posttraumatic stress disorder (Misanin et al. 1968; Nader et al. 2000a; Debiec and Ledoux 2004; Brunet et al. 2008). Only more recently have investigators demonstrated that appetitive memories also undergo reconsolidation; most, but not all (Yim et al. 2006), studies found a disruption of expression for the drug-associated memory, suggesting the potential to target the reconsolidation process as a treatment for drug addiction (Lee et al. 2005; Miller and Marshall 2005; Milekic et al. 2006; Valjent et al. 2006; Brown et al. 2007; Kelley et al. 2007; Sadler et al. 2007; Fricks-Gleason and Marshall 2008; Milton et al. 2008a, b).Miller and Marshall (2005) showed that reconsolidation of cocaine conditioned place preference (CPP) in the rat could be disrupted by either pre- or post-treatment of a phosphorylation inhibitor of extracellular signal-regulated kinase (1/2) (ERK) in a reactivation-dependent manner. Other studies have shown that protein synthesis inhibitors (Milekic et al. 2006), a matrix metalloproteinase (MMP) inhibitor (Brown et al. 2007), a β-noradrenergic receptor antagonist (Bernardi et al. 2006; Robinson and Franklin 2007a; Fricks-Gleason and Marshall 2008), and an N-methyl-D-aspartate (NMDA) receptor antagonist (Kelley et al. 2007; Sadler et al. 2007) can also disrupt the reconsolidation of drug-associated CPP memories. Studies by Lee and colleagues have shown that Zif268 antisense oligodeoxynucleotide infused into the basolateral amygdala prior to reactivation of memory for a cocaine-associated cue (the conditioned stimulus or CS) disrupts the ability of cocaine-associated cues to establish subsequent acquisition of a new instrumental response (Lee et al. 2005), and the ability of a drug-associated cue to induce relapse under a second-order schedule (Lee et al. 2006a). Thus, cocaine-associated memories appear to undergo reconsolidation in both Pavlovian and operant conditioning paradigms.Relapse to drug-seeking or drug-taking behavior can occur after re-exposure to three types of stimuli: the drug itself, drug-associated contextual and discrete cues, and stress; and all of these may promote relapse in humans (for review, see Epstein et al. 2006). Only a few CPP studies (Valjent et al. 2006; Brown et al. 2007) and no self-administration studies to our knowledge have tested whether the drug-associated memory can be rendered susceptible to disruption by pharmacological agents such that subsequent cocaine-primed reinstatement is suppressed. This drug-primed effect is observed in humans, producing relapse (Ludwig et al. 1974; Jaffe et al. 1989), and in rats, producing robust reinstatement of drug-seeking behavior in both CPP and self-administration tasks (McFarland and Ettenberg 1997; McFarland and Kalivas 2001; Sanchez and Sorg 2001; Kalivas and McFarland 2003). The development of a treatment strategy that makes use of the reconsolidation process will ultimately need to be powerful enough to diminish drug-seeking behavior in the presence of sizable doses of the drug itself. Therefore, the primary goal of this study was to determine whether drug-primed reinstatement could be suppressed in rats that have the memory reactivated in the presence of a pharmacological agent in cocaine self-administering rats. Since we previously have demonstrated the ability to disrupt cocaine-primed reinstatement only in animals in which the memory was reactivated using cocaine-induced CPP, we also tested the extent to which the same parameters used to disrupt reconsolidation in a cocaine-induced CPP task would disrupt reconsolidation in a cocaine self-administration task under conditions of drug-induced reinstatement.To examine this question, we chose the noncompetitive NMDA receptor antagonist (+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine maleate (MK-801). MK-801 has been shown to disrupt reconsolidation of spatial tasks (Przybyslawski and Sara 1997), fear tasks (Lee et al. 2006b), amphetamine-induced CPP (Sadler et al. 2007), cocaine-induced CPP (Kelley et al. 2007), and sucrose self-administration (Lee and Everitt 2008). Importantly, the two studies examining CPP using MK-801 did not explore whether MK-801 suppressed drug-seeking behavior in a manner that was dependent on whether the memory was reactivated, leaving open the possibility that it was not a reconsolidation process that was disrupted by MK-801.Here we demonstrate that MK-801 injected prior to cocaine-primed reinstatement of CPP disrupted subsequent cocaine-primed reinstatement of CPP, and this disruption was dependent on CPP contextual reactivation since injection of MK-801 and cocaine in the home cage did not disrupt subsequent cocaine-primed reinstatement of CPP. However, drug-seeking behavior in animals trained for cocaine self-administration was not disrupted when rats were reactivated under the same parameters that disrupted cocaine-induced CPP or when rats were given a reactivation session identical to their self-administration sessions. We thus demonstrate for the first time that memories associated with cocaine-induced CPP and cocaine self-administration are not similarly susceptible to disruption by MK-801.     

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

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

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.     

Prenatal choline availability alters the context sensitivity of Pavlovian conditioning in adult rats

The effects of prenatal choline availability on Pavlovian conditioning were assessed in adult male rats (3–4 mo). Neither supplementation nor deprivation of prenatal choline affected the acquisition and extinction of simple Pavlovian conditioned excitation, or the acquisition and retardation of conditioned inhibition. However, prenatal choline availability significantly altered the contextual control of these learned behaviors. Both control and choline-deprived rats exhibited context specificity of conditioned excitation as exhibited by a loss in responding when tested in an alternate context after conditioning; in contrast, choline-supplemented rats showed no such effect. When switched to a different context following extinction, however, both choline-supplemented and control rats showed substantial contextual control of responding, whereas choline-deficient rats did not. These data support the view that configural associations that rely on hippocampal function are selectively sensitive to prenatal manipulations of dietary choline during prenatal development.There is increasing evidence that variations in maternal dietary choline intake during the second half of pregnancy cause structural, biochemical, and physiological changes in basal forebrain neurons and their projections to the hippocampal complex as well as long-term cognitive changes in the offspring (e.g., Meck and Williams 2003; McCann et al. 2006; Meck et al. 2008). We know, for instance, that the adult offspring of pregnant rats supplemented with 4.5 times the amount of choline in the standard laboratory diet display improved memory capacity and precision on the radial-arm maze (e.g., Meck et al. 1988, 1989; Meck and Williams 1997b, 1999; Tees 1999a), Morris water maze (e.g., Tees 1999b; Tees and Mohammadi 1999; Yang et al. 2000; Brandner 2002), as well as facilitation of sustained attention and interval timing (e.g., Meck and Williams 1997a, c; Mohler et al. 2001; Cheng et al. 2006, 2008a, b; Cheng and Meck 2007) compared with offspring of dams fed a standard diet. Choline deficiency during the same developmental time frame, embryonic days (ED) 12–17, results in impaired performance on some, but not all, of these behavioral measures (e.g., Meck and Williams 1999, 2003). Furthermore, perinatal choline supplementation can alter behavior following a variety of developmental disorders, including the alleviation of abnormalities associated with fetal alcohol syndrome in rats (Thomas et al. 2000, 2004, 2007; Wagner and Hunt 2006), attenuation of some of the motor deficits observed in a Mecp21^lox mouse model of Rett syndrome (Nag and Berger-Sweeney 2007), and the improvement of sensory gating in a DBA/2 mouse model of schizophrenia that exhibits reduced numbers of hippocampal a7 nicotinic receptors (Stevens et al. 2008).These choline-induced alterations in cognitive function are accompanied by changes in the size and shape of basal forebrain cholinergic neurons (e.g., Williams et al. 1998; McKeon-O’Malley et al. 2003); modifications in acetylcholine turnover and choline transporter expression in the septum and hippocampus (Cermak et al. 1999; Mellott et al. 2007b); modulation of hippocampal neurogenesis, gene expression, phospholipase D activity, NGF levels, and MAPK and CREB activation (e.g., Holler et al. 1996; Sandstrom et al. 2002; Mellott et al. 2004, 2007a; Glenn et al. 2007); changes in dendritic fields and spine density in CA1 and dentate gyrus (DG) regions of the hippocampus (Meck et al. 2008); as well as modification of the neuropathological response to status epilepticus (e.g., Holmes et al. 2002; Wong-Goodrich et al. 2008a) and thresholds for eliciting long-term potentiation (LTP) in the hippocampus (Pyapali et al. 1998; Jones III et al. 1999). Together, these findings suggest that alterations in choline availability during early development may have specific impact on the ontogeny and later functioning of basal forebrain cholinergic neurons as well as efferent neurons involved in hippocampal LTP (Montoya et al. 2000). These findings also predict that behaviors that rely on the hippocampus are likely to be most affected by this dietary manipulation.Although choline is well known as the precursor for the neurotransmitter acetylcholine, it may be especially crucial to young or developing mammals for a number of other reasons (see Blusztajn and Wurtman 1983; Blusztajn 1998; Zeisel 2000, 2004, 2005). It is the precursor of certain phospholipids (e.g., phosphatidylcholine, sphingomyelin, and plasmenylcholine), which constitute the bulk of phospholipids in all biological membranes. Thus, there may be a particularly high demand for choline during prenatal and neonatal periods associated with rapid neurogenesis and synaptogenesis. Choline can also be enzymatically oxidized to betaine (mostly in peripheral tissues) and the methyl groups of betaine can then be used to resynthesize methionine from homocysteine. Changes in methionine availability alter the methylation of regulatory sequences of genes and of histones, leading to alterations in the patterns of gene expression (e.g., Waterland and Michels 2007; Nafee et al. 2008). Choline is also the precursor of two signaling molecules, platelet-activating factor, and sphingosylphosphorylcholine. Changes in choline availability may also alter membrane synthesis, methylation, and signaling broadly throughout the brain and periphery as well as more restricted effects on cholinergic neuronal pathways (e.g., Zeisel and Blusztajn 1994; Meck and Williams 2003).One common distinction in the Pavlovian-conditioning literature is between tasks that are sensitive to manipulation of the hippocampal formation from those that are not (e.g., Ross et al. 1984; Meck 1988; Schmajuk and Buhusi 1997; Holland et al. 1999). For example, simple excitatory Pavlovian conditioning is typically found to be unaffected by lesions of the hippocampus, while conditional discriminations in which animals must rely on combinations of predictive cues to respond correctly are disrupted by hippocampal damage (e.g., Jarrard and Davidson 1990, 1991). If prenatal choline availability is altering the development of cholinergic neurons in the basal forebrain that project to the hippocampus (see Meck and Williams 2003), our dietary manipulation might only be expected to affect conditioning tasks that require hippocampal involvement, not relatively simple tasks such as excitatory conditioning which do not rely on the hippocampus (e.g., Green and Woodruff-Pak 2000).In the current series of experiments, we examined the effects of prenatal choline supplementation and deficiency using a series of appetitive Pavlovian-conditioning tasks, all of which require associative learning. Our rationale was to determine whether variations in choline availability during prenatal development altered the learning of a simple association between the conditioned (CS) and unconditioned (US) stimuli (e.g., noise → food sequence), or if the dietary manipulation primarily affected conditioning tasks that require more complex relational processing and intact septal-hippocampal function (e.g., context A = tone → food; context B = noise → no food).In order to assess the importance of prenatal choline availability on associative learning, we investigated basic aspects of appetitive Pavlovian conditioning, i.e., conditioned excitation and extinction (e.g., Pavlov 1927) in experiment 1. In this paradigm, rats first receive repeated trials in which the CS occurs just before the presentation of the US, i.e., in a noise → food sequence. During this initial phase of training, the rat develops an increasing tendency to perform the conditioned response (CR) in the presence of the CS indicating that it expects the occurrence of the US. Typically, the probability of the CR increases in a negatively accelerating fashion until it reaches an asymptotic level. If the CS is then repeatedly presented in the absence of the reinforcing US (i.e., noise → no food), then the CR gradually declines; this is referred to as extinction of the CR (Gallistel and Gibbon 2000).One behavioral phenomenon that has been shown to be sensitive to hippocampal manipulation is the discriminative use of contextual cues to control conditioned responding (e.g., Holland and Bouton 1999). Typically, when CS-US pairings occur in one training environment or context, there is a small loss of responding to the CS if it is subsequently presented to the animal in the presence of a different set of contextual cues (e.g., Lovibond et al. 1984; Hall and Honey 1990; Honey et al. 1990; Kaye and Mackintosh 1990). However, this typical decrement in responding with a context switch is not observed in rats with electrolytic or aspiration lesions of the hippocampus (e.g., Good et al. 1997).In order to assess the effects of prenatal choline availability on contextual control of conditioned responding, we employed a renewal design (e.g., Bouton and Bolles 1979) in experiment 2. In this design, rats receive conditioning in one physical context (context A) prior to extinction in either the same context or a context different from that in which they received the initial CS-US pairings (context B). Finally, all of the rats are retested in the original conditioning context (i.e., context A). Bouton and colleagues (e.g., Bouton and Bolles 1979; Frohardt et al. 2000) have found that when rats that are conditioned in context A followed by extinction training in context B are later returned to the original training context for the final testing phase, they show a substantial recovery of the initial CR. Presumably, stimuli contained within the original training context act as reminder cues in this ABA condition, retrieving the memory for the initial acquisition phase (A) of the experiment during the final test phase as opposed to the more recent extinction phase (B). Rats in the AAA condition have no effective cues to discriminate the different phases of the experiment and as a consequence cannot selectively retrieve a specific memory from the sequence. In contrast, test session responding for the ABA condition should be more similar to the low levels observed at the end of the initial extinction phase due to the availability of differential contextual cues. This renewal design is particularly useful in that it provides for the potential to observe treatment effects in both the extinction and the renewal test phases of the experiment. Specifically, either the loss of responding with a context switch during extinction or the response recovery in the renewal test (or both) may be affected by prenatal choline availability. More importantly, those two effects may be due to either the same mechanism (e.g., processing of contextual stimuli) or two different mechanisms (e.g., context conditioning and memory retrieval)—potentially resulting in nonlinear effects of prenatal choline availability across the two experimental phases.A second basic type of associative learning, conditioned inhibition, in which the animal learns to predict the absence of an important event, was described by Pavlov (1927). A typical conditioned-inhibition task consists of training with two types of intermixed trials: On reinforced trials, one CS is followed by reinforcement (e.g., noise → food). On other trials, the same CS is paired with a second stimulus in the absence of the reinforcement (i.e., light/noise → no food). It is presumed that under these training conditions animals learn that the noise predicts the occurrence of the food, while light, the “conditioned inhibitor,” comes to predict the absence of food. That is, light “inhibits” the learned response to noise alone.A relatively small number of studies have examined the neural substrates of inhibitory learning. Aspiration lesions of the hippocampus, for example, impaired a relatively complex phenomenon called “blocking” of excitatory conditioning, but not the learning of conditioned inhibition (e.g., Solomon 1977; Chan et al. 2001). These data suggest that the hippocampal complex is not required for learning conditioned inhibition. Thus, in order to further assess whether prenatal choline availability affects basic associative learning, experiment 3 was designed to evaluate conditioned inhibition in supplemented (SUP), deficient (DEF), and control (CON) rats. In this experiment, rats were given randomly mixed presentations of reinforced and nonreinforced trial types. As training proceeds, the rats should learn to respond more on reinforced trials than on nonreinforced trials. After acquisition of the discrimination, the rats were presented with a retardation test phase in which the inhibitory light CS was paired with food. Rescorla (1969) described this retardation test as one of the critical measures of conditioned inhibition. Presumably, if the CS is a true inhibitor and predicts the absence of reinforcement at the outset of the retardation test, then acquisition of conditioned responding to the cue should be relatively slow during this phase of testing. Consequently, tests of conditioned inhibition should distinguish among prenatal choline treatment groups if inhibitory mechanisms are strengthened or weakened by prenatal choline availability.     

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.