Habit system (S-R learning): dorsolateral striatum
Striatum-dependent procedural memory
Hypothesis
Intra-BLA administration of an anxiogenic biases rats towards the use of habit memory
Intra-BLA infusion of RS-79948 is anxiogenic
Peripheral administration of an anxiogenic drug enhances and impairs response and place learning, respectively
The functional integrity of the BLA may be critical for these effects
Acute stress favours habits over goal-directed actions when it is administered before the extinction test (after learning)
Stress may modulate the engagement of hippocampus-based declarative and striatum-based procedural memory systems in classification learning
This may be observable in functional activity patterns assessed with fMRI
Sample
Rodents male Charles River Long-Evans rats
Rodents male Charles River Long-Evans rats
Humans students (34 men, 34 women; 18–32 years) Exclusion of 17 subjects
Humans students (30 men, 30 women; 18–30 years) Exclusion of 1 subject
Methods
Training (5 consecutive days, six trials/day) in single-solution water plus maze task (hippocampus-dependent place learning versus dorsal striatum-dependent response learning)
Intra-BLA infusions of RS-79948 (place task , response task ) or saline (place task , response task ) immediately following training on days 1–3
Additional groups (place task , response task ) received intra-BLA infusions of RS-79948 2 h after training
Anxiogenic potential of RS-79948: standard anxiety test with an automated elevated plus maze apparatus after intra-BLA infusions of RS 79948 () or saline ()
Training (6 consecutive days, six trials/day) in single-solution water plus maze task (hippocampus-dependent place learning versus dorsal striatum-dependent response learning)
Infusions of either peripheral and intra-BLA saline (place task , response task ), peripheral RS-79948 and intra-BLA saline (place task , response task ), peripheral saline and intra-BLA bupivacaine (place task , response task ), or peripheral RS 79948 and intra-BLA bupivacaine (place task , response task ) immediately following training on days 1–3
Training in two instrumental actions (high versus low probability) leading to a food outcome: three randomized trial types with different outcomes: (i) Chocolate (75 trials) (ii) Orange (75 trials) (iii) Neutral (75 trials)
Selective devaluation of one outcome (chocolate or orange) through eating to satiety
Socially evaluated cold pressor test (SECPT) versus control condition
Subjective stress ratings immediately after SECPT or control condition
Extinction test with 75 trials of the three trial types (25 min after SECPT/control condition and approximately 40 min after training)
Assessment of explicit task knowledge after extinction test (free recall and multiple-choice questionnaire)
Hunger and pleasantness ratings before learning, before and after devaluation, and before the extinction test
SECPT/control condition before learning
Subjective stress ratings immediately after SECPT or control condition
Weather-prediction task (PCL task) and visuomotor control task (25 min after SECPT/control condition) in the scanner: 200 randomized trials (100 trials per task; 1 out of 14 different cue patterns is presented per trial)
Assessment of explicit task knowledge (10 items questionnaire) outside the scanner
Learning strategy analysis with a mathematical model (comparison of a participant’s actual responses and the expected responses using a declarative or procedural strategy)
Induction of stress/arousal
BLA: injection of beta-adrenergic antagonist (RS 79948)
Peripheral injection of beta-adrenergic antagonist (RS 79948)
Socially evaluated cold pressure test (SECPT)
Socially evaluated cold pressure test (SECPT)
Physiological stress parameters
None measured
None measured
Blood pressure before, during, and after the SECPT or control condition
Cortisol: saliva samples after arrival at the laboratory, just before, just after, and 20 and 50 min after the SECPT or control condition
Blood pressure before, during, and after the SECPT or control condition
Cortisol: saliva samples before and immediately after the SECPT or control condition as well as before and after the learning task (25 and 90 min after the SECPT/control condition, resp.)
Affected memory phase
Consolidation
Consolidation
Retrieval
Encoding/consolidation
Behavioural results
Posttraining immediate intra-BLA infusions of RS-79948 (relative to delayed infusion or saline) impaired acquisition of the place task and enhanced acquisition of the response task
Anxiogenic potential of RS 79948: intra-BLA infusions of RS 79948 (relative to saline) lead to more and less time spent in the closed and open arms of the maze, respectively
BLA inactivation blocks the enhancement of response learning
Posttraining peripheral RS 79948 injections impair place learning
BLA inactivation blocks the impairment of place learning
SECPT increased subjective stress ratings
Habitual behaviour after SECPT: participants indicated that they do not want the devalued outcome any more but still chose the referring action
Goal-directed behaviour after control condition: control participants did not prefer the devalued food anymore; thus they did not prefer the associated action anymore
No stress effects on explicit task knowledge
SECPT increased subjective stress ratings
Comparable learning curves in the PCL task between SECPT/control condition: gradual improvement of classification performance across training
Stress effects on learning strategy during PCL: decreased use of single-cue-based strategies (hippocampus-dependent) and increased use of multicue-based strategies (striatum-dependent)
Stress effects on explicit task knowledge: participants in the SECPT condition remembered fewer details of the PCL task
Physiological results (stress parameters)
None measured
None measured
Increase in blood pressure during SECPT
Increased salivary cortisol before extinction (20 min after SECPT)
The SECPT-induced increase in cortisol levels (baseline-peak) is associated with habit performance
Increase in blood pressure during SECPT
Systolic blood pressure during the SECPT/control condition correlates with the use of multicue-based strategies
Increased salivary cortisol immediately before the PCL task in the scanner (25 min after SECPT)
Salivary cortisol levels (across both groups and for all time points) correlate with PCL performance
The SECPT-induced increase in cortisol levels (baseline-peak) is not associated with PCL performance
Salivary cortisol levels 25 and 90 min after the SECPT correlate with the use of multicue-based strategies
Neuroimaging results
None
None
None
Activated brain areas during the PCL task: caudate nucleus, putamen, hippocampus, parahippocampal cortex, orbitofrontal cortex, cingulate cortex, and inferior frontal cortex
Activity of the hippocampus correlated with the use of single-cue strategies, and activity of the putamen and the caudate nucleus correlated with multicue strategies during the PCL task
Activation of the striatum during PCL in both groups, but no significant activation of medial temporal lobe structures in the stress group
Caudate nucleus activity correlates with salivary cortisol levels (across both groups and for all time points), but not with the increase in cortisol levels (baseline-peak) after SECPT
PCL performance in the stress group is positively correlated with activity in the right caudate nucleus and the left putamen, but negatively correlated with activity in the left hippocampus
PCL performance in the control group is correlated with activity in the left hippocampus, but not with striatal activity
Conclusions
Intra-BLA infusions of RS 79948 can bias rats towards using habit memory by impairing cognitive memory
Intra-BLA infusions of RS 79948 exert an anxiogenic effect at the same dose that impairs and enhances cognitive and habit memory, respectively
Emotional state can modulate the degree of interference between cognitive and habit memory systems (release habit memory from competing/inhibitory influences of cognitive memory)
The functional integrity of the BLA is not necessary for the acquisition of place and response learning
The functional integrity of the BLA is critical in order for peripheral injections of RS 79948 to impair hippocampus-dependent cognitive memory and enhance dorsal striatum-dependent habit memory, respectively
Acute stress before extinction testing can abolish sensitivity of performance to outcome value
Acute stress can make behaviour habitual without affecting processes involved in learning (encoding, consolidation)
Stress does not affect the acquisition of the PCL task, but it changes the nature of classification learning from flexible, declarative learning to inflexible, procedural learning
Stress impairs the hippocampus-dependent system and allows the striatum to control behaviour, which rescues task performance
Attempts to engage the declarative system in PCL after stress disrupt performance