Neurobiology of Learning and Memory最新文献

Cue-potentiated feeding (CPF) describes instances where food intake is increased by exposure to conditioned cues associated with food, often in the absence of hunger. CPF effects have been reported in a range of experimental protocols developed by researchers working across diverse fields spanning behavioural neuroscience, social psychology and ecology. Here we review the evolution of research on cue-potentiated feeding in animal models to identify important behavioural parameters and key neural circuits and pharmacological systems underlying the effect. Overall, evidence indicates that social, discrete and contextual stimuli can be used to elicit CPF effects across multiple species, though effects are often subtle and sensitive to procedural variables. While regular exposure to food cues is thought to be a key risk factor for overeating in so-called ‘obesogenic’ environments, further work is needed to identify whether CPF promotes positive energy balance and weight gain over the longer term. We suggest several methodological and conceptual areas for inquiry to elucidate the contribution of CPF to the regulation of food choice and energy intake.

Although early studies were able to demonstrate a negative impact of stress on working memory performance, present research findings are heterogeneous. Numerous further studies found no effects or even improved performance, with the direction of these stress effects likely depending on the underlying biological mechanisms. The aim of this study was to investigate receptor-specific effects, as part of the stress-induced cortisol response, on working memory performance. Healthy, male participants (N=318, mean age 25.4 ± 5.1y) were exposed to the Trier Social Stress Test (TSST), a social-evaluative stress manipulation, or a non-stress control condition after they had received either spironolactone (blockade of the mineralocorticoid receptor, MR) or mifepristone (blockade of the glucocorticoid receptor, GR) or a placebo. Both substances are potent antagonists with high affinity for the respective receptors. To assess working memory, we implemented the n-back task subsequent to stress exposure, number of correct responses and reaction times served as outcome measures. We did not find effects of stress on working memory for any outcome measure, i.e. correct responses and reaction times. Yet, post hoc tests revealed that the group that received mifepristone exhibited longer reaction times under medium load conditions when compared to the placebo group, which might be an indication of the GR’s involvement in task performance. We conclude that working memory performance is not affected by acute stress, at least under these prevalent conditions.

Humans and animals can quickly learn a new strategy when a previously-rewarding strategy is punished. It is difficult to model this with reinforcement learning methods, because they tend to perseverate on previously-learned strategies − a hallmark of impaired response to punishment. Past work has addressed this by augmenting conventional reinforcement learning equations with ad hoc parameters or parallel learning systems. This produces reinforcement learning models that account for reversal learning, but are more abstract, complex, and somewhat detached from neural substrates. Here we use a different approach: we generalize a recently-discovered neuron-level learning rule, on the assumption that it captures a basic principle of learning that may occur at the whole-brain-level. Surprisingly, this gives a new reinforcement learning rule that accounts for adaptation and lose-shift behavior, and uses only the same parameters as conventional reinforcement learning equations. In the new rule, the normal reward prediction errors that drive reinforcement learning are scaled by the likelihood the agent assigns to the action that triggered a reward or punishment. The new rule demonstrates quick adaptation in card sorting and variable Iowa gambling tasks, and also exhibits a human-like paradox-of-choice effect. It will be useful for experimental researchers modeling learning and behavior.

Exercise provides a range of cognitive benefits, including improved memory performance. Previously, we demonstrated that 14 days of continuous voluntary wheel-running exercise enables learning in a hippocampus-dependent Object Location Memory (OLM) task under insufficient, subthreshold training conditions in adult mice. Whether similar exercise benefits can be obtained from consistent intermittent exercise as continuous exercise is unknown. Here, we examine whether intermittent exercise (the weekend warrior effect: 2 days of exercise a week for 7 weeks) displays similar or distinct cognitive benefits as previously examined with 14 days of continuous exercise. We find that both continuous and intermittent exercise parameters similarly enable hippocampus-dependent OLM compared to the 2-day exercise control group. Mice receiving intermittent exercise maintained cognitive benefits following a 7-day sedentary delay, whereas mice that underwent 14 continuous days of exercise showed diminished cognitive benefits as previously reported. Further, compared to continuous exercise, intermittent exercise mice exhibited persistently elevated levels of the genes Acvr1c and Bdnf which we know to be critically involved in hippocampus-dependent long-term memory in the dorsal hippocampus. Together findings suggest that consistent intermittent exercise persistently enables hippocampal-dependent long-term memory. Understanding the optimal parameters for persistent cognitive function and the mechanisms mediating persistent effects will aid in therapeutic pursuits investigating the mitigation of cognitive ailments.

Habitual instrumental behaviour is believed to rely on stimulus–response (S–R) associations. However, the method most commonly used to identify habitual behaviour, outcome devaluation, provides only indirect evidence of S-R control. Therefore, it is important to have a better understanding of the S-R association believed to underlie habitual responding. Under free-operant conditions, the context itself likely serves as at least part of the relevant stimuli in the association, and so modifications to the predictive power of the context should alter the expression of habits. The following experiments investigated how changes to the relationship between the training context and performance of the response, either by changing the context during testing or by exposing animals to the context alone, without the response lever present, impacted behavioural control during a devaluation test. We found evidence that the training context is important for the expression of habits; testing animals in a different context than where they were trained resulted in increased goal-directed control (Experiment 1). Furthermore, context alone exposure also increased goal-directed control with animals that received context alone exposure showing stronger devaluation effects, whether the context alone exposure happened on the last day of training (Experiment 2) or throughout training (Experiment 3). These findings are consistent with prior reports that the training context is important for the expression of habits and extends these findings by using sensory-specific satiety as a means for devaluation and by using context alone exposure to alter behavioural control.

Contextual fear conditioning is a protocol used to assess associative learning across species, including fish. Here, our goal was to expand the analysis of behavioral parameters that may reflect aversive behaviors in a contextual fear conditioning protocol using adult zebrafish (Danio rerio) and to verify how such parameters can be modulated. First, we analyzed the influence of an aversive stimulus (3 mild electric shocks for 5 s each at frequencies of 10, 100 or 1000 Hz) on fish behavior, and their ability to elicit fear responses in the absence of shock during a test session. To confirm whether the aversive responses are context-dependent, behaviors were also measured in a different experimental environment in a test session. Furthermore, we investigated the effects of dizocilpine (MK-801, 2 mg/kg, i.p.) on fear-related responses. Zebrafish showed significant changes in baseline activity immediately after shock exposure in the training session, in which 100 Hz induced robust contextual fear responses during the test session. Importantly, when introduced to a different environment, animals exposed to the aversive stimulus did not show any differences in locomotion and immobility-related parameters. MK-801 administered after the training session reduced fear responses during the test, indicating that glutamate NMDA-receptors play a key role in the consolidation of contextual fear-related memory in zebrafish. In conclusion, by further exploring fear-related behaviors in a contextual fear conditioning task, we show the effects of different shock frequencies and confirm the importance of context on aversive responses for associative learning in zebrafish. Additionally, our data support the use of zebrafish in contextual fear conditioning tasks, as well as for advancing pharmacological studies related to associative learning in translational neurobehavioral research.

In an animal model of compulsive drug use, a subset of rats continues to self-administer cocaine despite footshock consequences and is considered punishment resistant. We recently found that punishment resistance is associated with habits that persist under conditions that typically encourage a transition to goal-directed control. Given that random ratio (RR) and random interval (RI) schedules of reinforcement influence whether responding is goal-directed or habitual, we investigated the influence of these schedules on punishment resistance for cocaine or food. Male and female Sprague Dawley rats were trained to self-administer either intravenous cocaine or food pellets on a seeking-taking chained schedule of reinforcement, with the seeking lever requiring completion of either an RR20 or RI60 schedule. Rats were then given four days of punishment testing with footshock administered at the completion of seeking on a random one-third of trials. For cocaine-trained rats, the RI60 schedule led to greater punishment resistance (i.e., more trials completed) than the RR20 schedule in males and females. For food-trained rats, the RI60 schedule led to greater punishment resistance (i.e., higher reward rates) than the RR20 schedule in female rats, although male rats showed punishment resistance on both RR20 and RI60 schedules. For both cocaine and food, we found that seeking responses were suppressed to a greater degree than reward rate with the RI60 schedule, whereas response rate and reward rate were equally suppressed with the RR20 schedule. This dissociation between punishment effects on reward rate and response rate with the RI60 schedule can be explained by the nonlinear relation between these variables on RI schedules, but it does not account for the enhanced resistance to punishment. Overall, the results show greater punishment resistance with the RI60 schedule as compared to the RR20 schedule, indicating that schedules of reinforcement are an influencing factor on resistance to negative consequences.

Labilization-reconsolidation, which relies on retrieval, has been considered an opportunity to attenuate the negative aspects of traumatic memories. A therapeutic strategy based on reconsolidation blockade is deemed more effective than current therapies relying on memory extinction. Nevertheless, extremely stressful memories frequently prove resistant to this process. Here, after inducing robust fear memory in mice through strong fear conditioning, we examined the possibility of rendering it susceptible to pharmacological modulation based on the degree of generalized fear (GF). To achieve this, we established an ordered gradient of GF, determined by the perceptual similarity between the associated context (CA) and non-associated contexts (CB, CC, CD, and CE) to the aversive event. We observed that as the exposure context became less similar to CA, the defensive pattern shifted from passive to active behaviors in both male and female mice. Subsequently, in conditioned animals, we administered propranolol after exposure to the different contexts (CA, CB, CC, CD or CE). In males, propranolol treatment resulted in reduced freezing time and enhanced risk assessment behaviors when administered following exposure to CA or CB, but not after CC, CD, or CE, compared to the control group. In females, a similar change in behavioral pattern was observed with propranolol administered after exposure to CC, but not after the other contexts. These results highlight the possibility of indirectly manipulating a robust contextual fear memory by controlling the level of generalization during recall. Additionally, it was demonstrated that the effect of propranolol on reconsolidation would not lead to a reduction in fear memory per se, but rather to its reorganization resulting in greater behavioral flexibility (from passive to active behaviors). Finally, from a clinical viewpoint, this would be of considerable relevance since following this strategy could make the treatment of psychiatric disorders associated with traumatic memory formation more effective and less stressful.