Extinction learning: neurological features, therapeutic applications and the effect of aging

Increased hippocampal volume and gene expression following cognitive behavioral therapy in PTSD

The learning of fear extinction

NMDA Receptors Control Cue-Outcome Selectivity and Plasticity of Orbitofrontal Firing Patterns during Associative Stimulus-Reward Learning

Modulation of the extinction of fear learning.

The retention of information represents a defining trait of cognitive systems. The primary function of an organism’s memory is to retain engrams of unique experiences for extended periods, without interrupting memory trace during new learning processes. One of the cardinal unsolved issues in neuroscience lies in unveiling the mechanisms responsible for long-term memory. Despite this, a trustworthy experimental animal model for single-trial long-term memory remains undeveloped. To address this challenge, we aimed to create an animal model to study lifelong memory formation with a single trial. We used a mouse model of post-traumatic stress disorder (PTSD) for this purpose [1]. In this model, mice exposed to a powerful electrical foot shock develop highly persistent traumatic memories, which results in long-term behavioral changes [1]. Our hypothesis was that this model could uncover the mechanisms related to lifelong memory. The primary rationale for adopting the PTSD model as a model of lifelong memory pertains to the heightened durability of traumatic memories in comparison to conventional aversive memories [2]. Protein synthesis inhibitors, which are needed for long-term memory consolidation, can be used as amnesic agents to evaluate memory stability [3]. In a PTSD model using predator scent, the administration of a protein synthesis blocker prior to a traumatic experience disrupts the development of PTSD in mice [4]. However, it remains unclear how traumatic memory impairment affects PTSD development in the footshock model and whether the impairment persists long-term. Based on the evidence that the formation of PTSD necessitates the consolidation of associative memory, which is reliant on protein synthesis and coincides with changes in the stress response system, Siegmund and Watzhek [1] present a two-part proposal concerning the onset of PTSD. The two-part hypothesis regarding PTSD posits that PTSD formation involves sensory conditioning and sensitization processes that mutually reinforce one another. In line with this theorem, we considered the effects of studying context and exposure timing on the development of post-traumatic stress disorder, as it could potentially interfere with sensory conditioning formation [5]. The aim of this study is to create an experimental approach for developing enduring long-term memory through a single trial event on adult mice. We methodically analyzed behavioral expressions and the endurance of normal and traumatic fear memory, as well as their sensitivity to protein synthesis inhibition. Additionally, we investigated the effects of separating the timing of the associative and aversive elements of traumatic memory on PTSD development in a mouse model. The experiment involved male C57Bl/6 mice, aged between 3–4 months and 15–18 months (for an investigation into aged mice), which were placed in an electrified chamber. After 170 seconds, the mice experienced either one footshock (1.5 mA, 2 s) to elicit fear memory or three footshocks (1.5 mA, 10 s) for PTSD induction. After receiving the footshock, the mouse was held in the chamber for 60 seconds before being returned to its home cage. A memory test was conducted seven days later by placing the mice back in the same context. To assess the existence of standard PTSD symptoms, like sensitization and generalization, the creatures were exposed to an unfamiliar context and an unexpected auditory stimulus, respectively. In the experiment, animals were placed in unfamiliar or familiar safe contexts that differed from their previous ones. The duration of freezing was measured to assess fear and evaluate memory retention, sensitization, and generalization. Anxiety levels were evaluated using the elevated plus maze test. In the process of constructing a highly stable long-term memory model, we evaluated the behavioral performance of PTSD-induced (PTSD), fear-conditioned (FC), and active control (AC) groups of animals at 7 days, 1 month, and 3 months after exposure. As a result of PTSD induction, mice displayed increased fear levels for up to 3 months in the training context, along with heightened fear sensitization and generalization at 7 days following exposure, relative to the FC and AC groups. The PTSD group exhibited heightened freezing behavior within a month and a decreased number of entries to the closed arms during the initial three months following exposure when contrasted with the FC and AC groups. We additionally noted that the FC group displayed raised fear levels in contrast to the control animals up to 3 months post-exposure, albeit lower in magnitude than that of the PTSD group. The FC group, similar to the PTSD group, showed increased sensitization and generalization compared to control animals at 1 week post-exposure, but to a lesser extent than we observed in the PTSD group. The findings suggest that traumatic memory remains present for a minimum of three months, whereas fear memory in the fear conditioning paradigm diminishes in intensity during this time. Hence, PTSD induction effectively functions as an animal model of profoundly stable long-term memory, in opposition to the fear conditioning paradigm. As the high stable long-term memory model is designed for testing in senior animals, the impact of aging on traumatic memory formation and storage becomes an important inquiry at six months and one year following exposure. We analyzed the formation of both aversive and traumatic memory in mice aged 15–18 months, one week after footshock exposure. In the study involving older mice, the PTSD group exhibited increased levels of fear memory and sensitization when compared with both the FC and AC groups, in addition to displaying higher levels of generalization and anxiety when compared with the AC group. Furthermore, mice that underwent rear-conditioning showed heightened levels of fear during learning, fear sensitization, and generalization, although not anxiety. These findings demonstrate the formation of traumatic and aversive memory in aged mice, indicating that conducting memory tests six months and a year after exposure is an appropriate approach. To evaluate the durability of traumatic and aversive memory, we administered a protein synthesis inhibitor called cycloheximide (Chm) to interfere with memory formation. Mice were given a cycloheximide solution (95 mg/kg) via intraperitoneal injection 30 minutes before footshock (Chm-PTSD and Chm-FC groups) while control animals received a saline injection (Sal-PTSD, Sal-FC). After training, the Chm-FC group displayed a decrease in freezing rate when compared to the saline-injected group at the 7- and 30-day marks. The Chm-PTSD group exhibited fear levels comparable to those of the Sal-FC group in the training context, and did not display sensitization or generalization of fear. These findings suggest that inhibiting protein synthesis during memory formation resulted in complete amnesia for standard aversive memory and merely weakened traumatic memory. As a result, only the associative component of traumatic memory remained intact, while nonspecific symptoms of PTSD were absent. The strong resilience of memory in the PTSD model to severe disruptions, such as protein synthesis blockade, also confirms its stability, which facilitates exploration of lifelong memory mechanisms in the PTSD model. To examine the timing effect of the associative and aversive component of PTSD development, we evaluated traumatic memory formation when contextual memory formation was absent, and contextual exploration occurred three days before shock cessation. A week after exposure to immediate and intense footshock, mice in the experimental group exhibited lower levels of fear, reduced fear sensitization, and less anxiety compared to those in the PTSD-induced group. If an intense foot shock was preceded by a contextual study three days earlier, the animals exhibited lower levels of fear in an unfamiliar, safe context, and reduced anxiety compared to the mice induced with PTSD. In both cases, the fear level after the shock cessation was identical to that of the mice who experienced an immediate, moderate shock. Based on our findings, we concluded that aversive memory forms in the absence or prior formation of contextual memory in relation to traumatic exposure. However, PTSD induction does not occur under such circumstances. This indicates that contextual memory needs to be formed simultaneously with the traumatic event for the development of traumatic memory in the mouse model of PTSD. In summary, our findings indicate that aversive memory tends to fade over time, while traumatic memory remains stable for a minimum of 3 months following its induction. Therefore, PTSD proves to be a fitting candidate laboratory model for high, stable, and long-term memory. Aged mice aged between 15–18 months display the formation of traumatic memory and experience PTSD symptoms in a manner similar to adult animals aged between 3–4 months in the PTSD model. For the induction of PTSD in mice, contextual memory formation about the traumatic environment and the traumatic event occurrence may coincide, which is crucial.

BackgroundNew evidence suggests that the centromedial amygdala (CMA) and the basolateral amygdala (BLA) play different roles in threat processing. Our study aimed to investigate the effects of trauma and post‐traumatic stress disorder (PTSD) on the functional connectivity (FC) of the amygdala and its subregions.MethodsTwenty‐seven patients with typhoon‐related PTSD, 33 trauma‐exposed controls (TEC), and 30 healthy controls (HC) were scanned with a 3‐Tesla magnetic resonance imaging scanner. The FCs of the BLA, the CMA, and the amygdala as a whole were examined using a seed‐based approach, and then, the analysis of variance was used to compare the groups.ResultsWe demonstrated that the BLA had a stronger connectivity with the prefrontal cortices (PFCs) and angular gyrus in the PTSD group than in the TEC group. Additionally, compared with the PTSD and the HC groups, the TEC group exhibited decreased and increased BLA FC with the ventromedial PFC and postcentral gyrus (PoCG), respectively. Furthermore, the PTSD group showed abnormal FC between the salience network and default‐mode network, as well as the executive control network. Compared with the HC group, the TEC group and the PTSD group both showed decreased BLA FC with the superior temporal gyrus (STG). Finally, the FCs between the bilateral amygdala (as a whole) and the vmPFC, and between the BLA and the vmPFC have a negative correlation with the severity of PTSD.ConclusionsDecreased BLA‐vmPFC FC and increased BLA‐PoCG FC may reflect PTSD resilience factors. Trauma leads to decreased connectivity between the BLA and the STG, which could be further aggravated by PTSD.

Stress impairs extinction learning, and these deficits depend, in part, on stress-induced norepinephrine (NE) release in the basolateral amygdala (BLA). For example, systemic or intra-BLA administration of propranolol reduces the immediate extinction deficit (IED), an impairment in extinction learning that occurs when extinction trials are administered soon after fear conditioning. Here, we explored whether locus coeruleus (LC)-NE regulates stress-induced changes in spike firing in the BLA and consequent extinction learning impairments. Rats were implanted with recording arrays in the BLA and, after recovery from surgery, underwent a standard auditory fear conditioning procedure. Fear conditioning produced an immediate and dramatic increase in the spontaneous firing of BLA neurons that persisted (and in some units, increased further) up to an hour after conditioning. This stress-induced increase in BLA firing was prevented by systemic administration of propranolol. Conditioning with a weaker footshock caused smaller increases in BLA firing rate, but this could be augmented by chemogenetic activation of the LC. Conditioned freezing in response to a tone paired with a weak footshock was immune to the IED, but chemogenetic activation of the LC before the weak conditioning protocol increased conditioned freezing behavior and induced an IED; this effect was blocked with intra-BLA infusions of propranolol. These data suggest that stress-induced activation of the LC increases BLA spike firing and causes impairments in extinction learning. Stress-induced increases in BLA activity mediated by LC-NE may be a viable therapeutic target for individuals with stress- and trauma-related disorders.SIGNIFICANCE STATEMENT Patients with post-traumatic stress disorder (PTSD) show heightened amygdala activity; elevated levels of stress hormones, including norepinephrine; and are resistant to the extinction of fear memories. Here, we show that stress increases basolateral amygdala (BLA) spike firing. This could be attenuated by systemic propranolol and mimicked by chemogenetic activation of the locus coeruleus (LC), the source of forebrain norepinephrine (NE). Finally, we show that LC-NE activation is sufficient to produce extinction deficits, and this is blocked by intra-BLA propranolol. Stress-induced increases in BLA activity mediated by LC-NE may be a viable therapeutic target for individuals with PTSD and related disorders.

Artificial association of pre-stored information to generate a qualitatively new memory.

Cholinergic neurotransmission in the basolateral amygdala during cued fear extinction

NMDA receptor blockade in the basolateral amygdala disrupts consolidation of stimulus-reward memory and extinction learning during reinstatement of cocaine-seeking in an animal model of relapse

Pharmacologic augmentation of extinction learning during exposure therapy for PTSD.

Extinction of threat memory is a measure of behavioral flexibility. In the absence of additional reinforcement, the extinction of learned behaviors allows animals and humans to adapt to their changing environment. Extinction mechanisms and their therapeutic implications for maladaptive learning have been extensively studied. However, how aging affects extinction learning is much less understood. Using a rat model of olfactory threat extinction, we show that the extinction of olfactory threat memory is impaired in aged Sprague-Darley rats. Following extinction training, long-term depression (LTD) in the piriform cortex (PC) was inducible ex vivo in aged rats and was NMDA receptor (NMDAR)-independent. On the other hand, adult rats acquired successful olfactory threat extinction, and LTD was not inducible following extinction training. Neuronal cFos activation in the posterior PC correlated with learning and extinction performance in rats. NMDAR blockade either systemically or locally in the PC during extinction training prevented successful extinction in adult rats, following which NMDAR-dependent LTD became inducible ex vivo. This suggests that extinction learning employs NMDAR-dependent LTD mechanisms in the PC of adult rats, thus occluding further LTD induction ex vivo. The rescue of olfactory threat extinction in aged rats by D-cycloserine, a partial NMDAR agonist, suggests that the impairment in olfactory threat extinction of aged animals may relate to NMDAR hypofunctioning and a lack of NMDAR-dependent LTD. These findings are consistent with an age-related switch from NMDAR-dependent to NMDAR-independent LTD in the PC. Optimizing NMDAR function in sensory cortices may improve learning and flexible behavior in the aged population.

The entorhinal cortex (EC) is a unique and fascinating structure, constituting a highly parallel interface between phylogenetically old cortex (hippocampus) and higher neocortical areas. The literature related to the EC has increased dramatically in the past 30 years. A simple PubMed search for EC lists around 20–40 articles per year in the early 1980s. This more than doubled by the beginning of the 1990s and doubled again by the start of the new century. Throughout the current decade, some 220–250 papers per year relating to structure, function, and pathology of the EC have been published, and the numbers are still rising. It has become increasingly apparent that the EC can no longer be regarded as a simple relay between the hippocampus and neocortex, but is a dynamic processor of both efferent and afferent information. In recent years, there has been burgeoning interest in synaptic plasticity at entorhinal synapses particularly in relationship with learning and memory functions of limbic structures. Neuroplasticity in the EC clearly takes many forms. In addition to the “input” and “output” roles that have long been ascribed to the superficial and deep layers, functions of the entorhinal area have been studied with respect to differences in the medial and lateral divisions, prominent state-dependent theta- and gamma-frequency population activities, membrane conductances that shape cellular and network activities, the neurophysiology of synaptic inputs from other regions and interlaminar connections, and the powerful roles of modulatory transmitter systems and local inhibition. The discovery of grid cells in the EC has led to strong interest in the role of the area in spatial processing. On the other side of the coin, long-term adaptive changes in EC function also appear to play pivotal roles in neuropathological states, particularly epilepsy, schizophrenia, and Alzheimer's disease. Contributions to this special issue of Neural Plasticity provide an overview of some of the theoretical and methodological approaches that are being applied to understand the functions and mechanisms of neuroplasticity within the EC. The research contained in this special issue deals mainly with findings derived from animal research. It is clear, however, that research using a variety of techniques is providing insight into entorhinal function in human and clinical populations. The significance of plasticity in the EC is best understood within the context provided by anatomical knowledge of the extrinsic and intrinsic connectivity of this area. Canto et al. have provided an excellent overview of the anatomical organization of the EC. Their analysis of intrinsic and extrinsic connections, in relationship to laminar organization and neuronal subtypes, heavily underscores our increasing perception of the EC as a dynamic interactive processor essential to hippocampal function, rather than a passive route of information entry and exit. The EC is tightly intertwined with circuitries of the hippocampal formation and other cortical areas, and its particular contributions to sensory processing, learning, memory, and motor behavior have been difficult to define. Lipton and Eichenbaum present evidence dealing with the complementary roles of the EC and hippocampus in episodic memory. They have found that medial EC neurons show stronger trajectory-dependent firing whereas hippocampal place cells show greater spatial specificity. Based on these observations, they propose roles of the hippocampus and medial EC in encoding sequences of events and disambiguating overlapping experiences, respectively. A contribution by Bevilaqua et al. also highlights the role of the EC in extinction learning and how this is affected by ageing. They raise the possibility that ageing-associated impairment of extinction may occur in part because of degenerative changes in the EC. Several articles in the special issue have investigated mechanisms of long-term, activity-dependent synaptic plasticity. Ma et al. observed differences in the induction of long-term synaptic potentiation (LTP) in the superficial neurones of the EC; LTP in horizontal (potentially extrinsic) inputs required NMDA receptor activation whereas LTP in interlaminar (intrinsic) inputs did not. This underlines the potential complexity of information processing in neurones that provides the bulk of hippocampal afferent inputs. Hernandez et al. have found that prenatal malnutrition has lasting effects on the capacity of the EC to express LTP, and it is possible that this reduced plasticity may contribute to deficits in learning and memory in the adult. In addition to LTP, there is a growing body of literature dealing with characteristics of long-term depression (LTD) in the EC. Kourrich et al. have investigated the postsynaptic signaling mechanisms that mediate entorhinal LTD, and provide further evidence for the roles of calcium-dependent signaling and protein phosphatases in the expression of LTD in the superficial layers. Much of the interest in long-term changes in synaptic strength has traditionally been driven by interest in mnemonic processes and neurological disorders. Short-term activity-dependent changes in synaptic strength also have powerful influences on ongoing synaptic transmission in the EC. In this issue, Chamberlain et al. have continued their work on the role of presynaptic NMDA autoreceptors in excitatory transmission in layer V of the EC. They demonstrate that presynaptic NR2B subunits are critical in enhancement of both spontaneous glutamate release and frequency-dependent facilitation of action potential-driven release. The kinetics of these receptors provide an optimal facilitation frequency of 3–6 Hz, and the authors speculate on the possible involvement of the autoreceptors in generation of delta/theta oscillations in mnemonic processing and epileptogenesis. The EC is subject to the powerful influence of several neuromodulatory transmitters including acetylcholine, serotonin, and dopamine. One of the articles in the special issue, from Caruana and Chapman, describes the dose-dependent, bidirectional effect of dopamine on the amplitude of evoked synaptic responses in layer II of the lateral EC. The significance of this bidirectional effect for entorhinal function is as yet unclear, but there are interesting parallels with bidirectional effects of dopamine in prefrontal cortex. Oscillatory neuronal population activities are known to contribute to the synchronization of neuronal activity in cortical areas throughout the brain, and the mechanisms that generate theta- and gamma-frequency activities in the EC are being examined. Kainic acid induces powerful gamma oscillations in layer III of the EC, and Stanger et al. demonstrate here that the GLUK5 subunit is required for kainate-induced gamma activity in the EC. Morgan et al. also examined kainate-induce oscillations in the beta/gamma range in EC with respect to the action of cannabinoid receptors. Their experiments suggest that CB1 receptors are constitutively active in the EC and that antagonists or inverse agonists enhanced beta/gamma oscillations in layer II but suppressed oscillations in layer V, again pointing to differential control of synchrony in neurons providing afferent input to the hippocampus and those receiving output from it. Hasselmo and Brandon have examined how oscillations in membrane potential in entorhinal neurons, combined with the unique persistent firing activity in these cells, may contribute to several phenomena. This provides an excellent example of how quantitative analysis of cellular and network properties of the EC can lead to a greater understanding of mechanisms of cognition and behavior. The role of oscillatory activity in regulating large-scale interactions between the hippocampal formation and the neocortex has come under increased study in the past several years as several labs have begun to examine concurrent changes in oscillatory EEG activity in hippocampus and neocortex associated with sleep and waking states. In the current issue, Axmacher et al. describe results obtained from scalp and intracranial EEG recordings from epilepsy patients that show increased correlations between oscillations in the neocortex and medial temporal lobe including the rhinal cortex during non-REM sleep; the increased correlation may support mechanisms related to memory consolidation. Our hope is that this special issue of Neural Plasticity will serve to emphasize the diversity of phenomena related to neural plasticity in the EC, and to highlight the importance of these effects, particularly in relationship to the increasing perception of the participation of the EC in mnemonic function of temporal lobe memory circuits and its related roles in integration of spatial information. C. Andrew Chapman Roland S. G. Jones Min Jung

The ventral medial prefrontal cortex (vMPFC) facilitates the cardiac baroreflex response through N-methyl-D-aspartate (NMDA) receptor activation and nitric oxide (NO) formation by neuronal NO synthase (nNOS) and soluble guanylate cyclase (sGC) triggering. Glutamatergic transmission is modulated by the cannabinoid receptor type 1 (CB1) and transient receptor potential vanilloid type 1 (TRPV1) receptors, which may inhibit or stimulate glutamate release in the brain, respectively. Interestingly, vMPFC CB1 receptors decrease cardiac baroreflex responses, while TRPV1 channels facilitate them. Therefore, the hypothesis of the present study is that the vMPFC NMDA/NO pathway is regulated by both CB1 and TRPV1 receptors in the modulation of cardiac baroreflex activity. In order to test this assumption, we used male Wistar rats that had stainless steel guide cannulae bilaterally implanted in the vMPFC. Subsequently, a catheter was inserted into the femoral artery, for cardiovascular recordings, and into the femoral vein for assessing baroreflex activation. The increase in tachycardic and bradycardic responses observed after the microinjection of a CB1 receptors antagonist into the vMPFC was prevented by an NMDA antagonist as well as by the nNOS and sGC inhibition. NO extracellular scavenging also abolished these responses. These same pharmacological manipulations inhibited cardiac reflex enhancement induced by TRPV1 agonist injection into the area. Based on these results, we conclude that vMPFC CB1 and TRPV1 receptors inhibit or facilitate the cardiac baroreflex activity by stimulating or blocking the NMDA activation and NO synthesis.

PTSD and Combat-Related Injuries: Functional Neuroanatomy