A ventral pallidal-thalamocortical circuit mediates the cognitive control of instrumental action.

Predictive learning can engage a selective form of cognitive control that biases choice between actions based on information about future outcomes that the learning provides. This influence has been hypothesized to depend on a feedback circuit in the brain through which the basal ganglia modulate activity in the prefrontal cortex; however, direct evidence for this functional circuit has proven elusive. Here, using an animal model of cognitive control, we found that the influence of predictive learning on decision making is mediated by an inhibitory feedback circuit linking the medial ventral pallidum and the mediodorsal thalamus, the activation of which causes disinhibition of the orbitofrontal cortex via reduced activation of inhibitory parvalbumin interneurons during choice. Thus, we found that, for this function, the mediodorsal thalamus serves as a pallidal-cortical relay through which predictive learning controls action selection, which has important implications for understanding cognitive control and its vicissitudes in various psychiatric disorders and addiction.

Immp2l knockdown in male mice increases stimulus-driven instrumental behaviour but does not alter goal-directed learning or neuron density in cortico-striatal circuits in a model of Tourette syndrome and autism spectrum disorder.

Cortico-striatal neurocircuits mediate goal-directed and habitual actions which are necessary for adaptive behaviour. It has recently been proposed that some of the core symptoms of autism spectrum disorder (ASD) and Gilles de la Tourette syndrome (GTS), such as tics and other repetitive behaviours, may emerge because of imbalances in these neurocircuits. We have recently developed a model of ASD and GTS by knocking down Immp2l, a mitochondrial gene frequently associated with these disorders. The current study sought to determine whether Immp2l knockdown (KD) in male mice alters flexible, goal- or cue- driven behaviour using procedures specifically designed to examine response-outcome and stimulus-response associations, which underlie goal-directed and habitual behaviour, respectively. Whether Immp2l KD alters neuron density in cortico-striatal neurocircuits known to regulate these behaviours was also examined. Immp2l KD mice and wild type-like mice (WT) were trained on Pavlovian and instrumental learning procedures where auditory cues predicted food delivery and lever-press responses earned a food outcome. It was demonstrated that goal-directed learning was not changed for Immp2l KD mice compared to WT mice, as lever-press responses were sensitive to changes in the value of the food outcome, and to contingency reversal and degradation. There was also no difference in the capacity of KD mice to form habitual behaviours compared to WT mice following extending training of the instrumental action. However, Immp2l KD mice were more responsive to auditory stimuli paired with food as indicated by a non-specific increase in lever response rates during Pavlovian-to-instrumental transfer. Finally, there were no alterations to neuron density in striatum or any prefrontal cortex or limbic brain structures examined. Thus, the current study suggests that Immp2l is not necessary for learned maladaptive goal or stimulus driven behaviours in ASD or GTS, but that it may contribute to increased capacity for external stimuli to drive behaviour. Alterations to stimulus-driven behaviour could potentially influence the expression of tics and repetitive behaviours, suggesting that genetic alterations to Immp2l may contribute to these core symptoms in ASD and GTS. Given that this is the first application of this battery of instrumental learning procedures to a mouse model of ASD or GTS, it is an important initial step in determining the contribution of known risk-genes to goal-directed versus habitual behaviours, which should be more broadly applied to other rodent models of ASD and GTS in the future.

High fat diet allows food-predictive stimuli to energize action performance in the absence of hunger, without distorting insulin signaling on accumbal cholinergic interneurons.

Obesity can disrupt how food-predictive stimuli control action performance and selection. These two forms of control recruit cholinergic interneurons (CIN) located in the nucleus accumbens core (NAcC) and shell (NAcS), respectively. Given that obesity is associated with insulin resistance in this region, we examined whether interfering with CIN insulin signaling disrupts how food-predictive stimuli control actions. To interfere with insulin signaling we used a high-fat diet (HFD) or genetic excision of the insulin receptor (InsR) from cholinergic cells. HFD left intact the capacity of food-predictive stimuli to energize performance of an action earning food when mice were tested hungry. However, it allowed this energizing effect to persist when the mice were tested sated. This persistence was linked to NAcC CIN activity but was not associated with distorted CIN insulin signaling. Accordingly, InsR excision had no effect on how food-predicting stimuli control action performance. Next, we found that neither HFD nor InsR excision altered the capacity of food-predictive stimuli to guide action selection. Yet, this capacity was associated with changes in NAcS CIN activity. These results indicate that insulin signaling on accumbal CINs does not modulate how food-predictive stimuli control action performance and selection. However, they show that HFD allows food-predictive stimuli to energize performance of an action earning food in the absence of hunger.

Cognitive effects of thalamostriatal degeneration are ameliorated by normalizing striatal cholinergic activity.

The loss of neurons in parafascicular thalamus (Pf) and their inputs to dorsomedial striatum (DMS) in Lewy body disease (LBD) and Parkinson's disease dementia (PDD) have been linked to the effects of neuroinflammation. We found that, in rats, these inputs were necessary for both the function of striatal cholinergic interneurons (CINs) and the flexible encoding of the action-outcome (AO) associations necessary for goal-directed action, producing a burst-pause pattern of CIN firing but only during the remapping elicited by a shift in AO contingency. Neuroinflammation in the Pf abolished these changes in CIN activity and goal-directed control after the shift in contingency. However, both effects were rescued by either the peripheral or the intra-DMS administration of selegiline, a monoamine oxidase B inhibitor that we found also enhances adenosine triphosphatase activity in CINs. These findings suggest a potential treatment for the cognitive deficits associated with neuroinflammation affecting the function of the Pf and related structures.

Basal forebrain cholinergic signaling in the basolateral amygdala promotes strength and durability of fear memories.

The basolateral amygdala (BLA) complex receives dense cholinergic projections from the nucleus basalis of Meynert (NBM) and the horizontal limb of the diagonal band of Broca (HDB). The present experiments examined whether these projections regulate the formation, extinction, and renewal of fear memories. This was achieved by employing a Pavlovian fear conditioning protocol and optogenetics in transgenic rats. Silencing NBM projections during fear conditioning weakened the fear memory produced by that conditioning and abolished its renewal after extinction. By contrast, silencing HDB projections during fear conditioning had no effect. Silencing NBM or HDB projections during extinction enhanced the loss of fear produced by extinction, but only HDB silencing prevented renewal. Next, we found that systemic blockade of nicotinic acetylcholine receptors during fear conditioning mimicked the effects produced by silencing NBM projections during fear conditioning. However, this blockade had no effect when given during extinction. These findings indicate that basal forebrain cholinergic signaling in the BLA plays a critical role in fear regulation by promoting strength and durability of fear memories. We concluded that cholinergic compounds may improve treatments for post-traumatic stress disorder by durably stripping fear memories from their fear-eliciting capacity.

Goal-directed actions transiently depend on dorsal hippocampus.

The role of the hippocampus in goal-directed action is currently unclear; studies investigating this issue have produced contradictory results. Here we reconcile these contradictions by demonstrating that, in rats, goal-directed action relies on the dorsal hippocampus, but only transiently, immediately after initial acquisition. Furthermore, we found that goal-directed action also depends transiently on physical context, suggesting a psychological basis for the hippocampal regulation of goal-directed action control.

Basolateral Amygdala Drives a GPCR-Mediated Striatal Memory Necessary for Predictive Learning to Influence Choice.

Predictive learning exerts a powerful influence over choice between instrumental actions. Nevertheless, how this learning is encoded in a sufficiently stable manner to influence choices that can occur much later in time is unclear. Here, we report that the basolateral amygdala (BLA) encodes predictive learning and establishes the memory necessary for future choices by driving the accumulation of delta-opioid receptors (DOPRs) on the somatic membrane of cholinergic interneurons in the nucleus accumbens shell (NAc-S). We found that the BLA controls DOPR accumulation via its influence on substance P release in the NAc-S, and that although DOPR accumulation is not necessary for predictive learning per se, it is necessary for the influence of this learning on later choice between actions. This study uncovers, therefore, a novel GPCR-based form of memory that is established by predictive learning and is necessary for such learning to guide the selection and execution of specific actions.

Ventral pallidal projections to mediodorsal thalamus and ventral tegmental area play distinct roles in outcome-specific Pavlovian-instrumental transfer.

Outcome-specific Pavlovian-instrumental transfer (PIT) demonstrates the way that reward-related cues influence choice between instrumental actions. The nucleus accumbens shell (NAc-S) contributes critically to this effect, particularly through its output to the rostral medial ventral pallidum (VP-m). Using rats, we investigated in two experiments the role in the PIT effect of the two major outputs of this VP-m region innervated by the NAc-S, the mediodorsal thalamus (MD) and the ventral tegmental area (VTA). First, two retrograde tracers were injected into the MD and VTA to compare the neuronal activity of the two populations of projection neurons in the VP-m during PIT relative to controls. Second, the functional role of the connection between the VP-m and the MD or VTA was assessed using asymmetrical pharmacological manipulations before a PIT test. It was found that, whereas neurons in the VP-m projecting to the MD showed significantly more neuronal activation during PIT than those projecting to the VTA, neuronal activation of these latter neurons correlated with the size of the PIT effect. Disconnection of the two pathways during PIT also revealed different deficits in performance: disrupting the VP-m to MD pathway removed the response biasing effects of reward-related cues, whereas disrupting the VP-m to VTA pathway preserved the response bias but altered the overall rate of responding. The current results therefore suggest that the VP-m exerts distinct effects on the VTA and MD and that these latter structures mediate the motivational and cognitive components of specific PIT, respectively.

Thalamocortical integration of instrumental learning and performance and their disintegration in addiction.

A recent focus of addiction research has been on the effect of drug exposure on the neural processes that mediate the acquisition and performance of goal-directed instrumental actions. Deficits in goal-directed control and a consequent dysregulation of habit learning processes have been described as resulting in compulsive drug seeking. Similarly, considerable research has focussed on the motivational and emotional changes that drugs produce and that result in changes in the incentive processes that modulate goal-directed performance. Although these areas have developed independently, we argue that the effects they described are likely not independent. Here we hypothesize that these changes result from a core deficit in the way the learning and performance factors that support goal-directed action are integrated at a neural level to maintain behavioural control. A dorsal basal ganglia stream mediating goal-directed learning and a ventral stream mediating various performance factors find several points of integration in the cortical basal ganglia system, most notably in the thalamocortical network linking basal ganglia output to a variety of cortical control centres. Recent research in humans and other animals is reviewed suggesting that learning and performance factors are integrated in a network centred on the mediodorsal thalamus and that disintegration in this network may provide the basis for a 'switch' from recreational to dysregulated drug seeking resulting in the well documented changes associated with addiction.

Dorsal and ventral streams: the distinct role of striatal subregions in the acquisition and performance of goal-directed actions.

Considerable evidence suggests that distinct neural processes mediate the acquisition and performance of goal-directed instrumental actions. Whereas a cortical-dorsomedial striatal circuit appears critical for the acquisition of goal-directed actions, a cortical-ventral striatal circuit appears to mediate instrumental performance, particularly the motivational control of performance. Here we review evidence that these distinct mechanisms of learning and performance constitute two distinct 'streams' controlling instrumental conditioning. From this perspective, the regulation of the interaction between these 'streams' becomes a matter of considerable importance. We describe evidence that the basolateral amygdala, which is heavily interconnected with both the dorsal and ventral subregions of the striatum, coordinates this interaction providing input to the final common path to action as a critical component of the limbic-motor interface.

The role of the amygdala-striatal pathway in the acquisition and performance of goal-directed instrumental actions.

The posterior dorsomedial striatum (pDMS) is essential for the acquisition and expression of the specific response-outcome (R-O) associations that underlie goal-directed action. Here we examined the role of a pathway linking the basolateral amygdala (BLA) and pDMS in such goal-directed learning. In Experiment 1, rats received unilateral lesions of the BLA and were implanted with cannula targeting the pDMS in either the ipsilateral (control) or contralateral (disconnection) hemisphere. After initial training, rats received infusions of muscimol to inactivate the pDMS immediately before sessions in which novel R-O associations were introduced. Sensitivity to devaluation by specific satiety was then assessed. Whereas rats in the ipsilateral group used the recently acquired associations to direct performance following devaluation, those in the contralateral group could not, indicating that BLA-pDMS disconnection prevented the acquisition of the new R-O associations. Indeed, evidence suggested that these rats relied instead on learning acquired during prior training to direct performance following devaluation. In Experiment 2, rats underwent similar surgery and training except they received muscimol infusions immediately before devaluation testing. Those in the ipsilateral group showed a selective devaluation effect, again based on the most recently introduced R-O associations. In contrast, rats in the contralateral group showed nonselective performance after devaluation indicating that the BLA-DMS pathway is also required for the expression of selective R-O associations. Together these results suggest that input from the BLA is essential for specific R-O learning by the pDMS.

The ventral striato-pallidal pathway mediates the effect of predictive learning on choice between goal-directed actions.

The nucleus accumbens shell (NAc-S) plays an important role in the way stimuli that predict reward affect the performance of, and choice between, goal-directed actions in tests of outcome-specific Pavlovian-instrumental transfer (PIT). The neural processes involved in PIT downstream of the ventral striatum are, however, unknown. The NAc-S projects prominently to the ventral pallidum (VP), and in the current experiments, we assessed the involvement of the NAc-S to VP projection in specific PIT in rats. We first compared expression of the immediate-early gene c-Fos in the medial (VP-m) and lateral (VP-l) regions of the VP and in addition, used the retrograde tracer Fluoro-gold combined with c-Fos to assess the involvement of these pathways during PIT. Although there was no evidence of differential activation in neurons in the VP-l, the VP-m showed a selective increase in activity in rats tested for PIT compared with appropriate controls, as did NAc-S neurons projecting to the VP-m. To confirm that VP-m activity is important for PIT, we inactivated this region before test and found this inactivation blocked the influence of predictive learning on choice. Finally, to confirm the functional importance of the NAc-S to VP-m pathway we used a disconnection procedure, using asymmetrical inactivation of the NAc-S and either the ipsilateral or contralateral VP-m. Specific PIT was blocked but only by inactivation of the NAc-S and VP-m in contralateral hemispheres. These results suggest that the NAc-S and VP-m form part of a circuit mediating the effects of predictive learning on choice.

The orbitofrontal cortex, predicted value, and choice.

Considerable evidence suggests that choice between goal-directed actions depends on two incentive processes encoding the reward value of the goal or outcome and the predicted value of an action based on outcome-related stimuli. Although incentive theories generally assume that these processes are mediated by a common associative mechanism, a number of recent findings suggest that they are dissociable; the reward value of an action is derived from consummatory experience with the outcome itself, whereas the predicted value of an action is based on the presence of outcome-associated stimuli from which estimates of the likelihood of an outcome are derived. Importantly, the orbitofrontal cortex (OFC) in rodents appears to mediate the effect of outcome-related stimuli on choice; OFC lesions disrupt the influence of Pavlovian stimuli on choice in tests of outcome-specific Pavlovian-instrumental transfer. However, the influence of outcome-related stimuli on choice involves a larger circuit including the OFC, the ventral striatum, and the amygdala. How these structures interact, however, is not yet fully understood and is an important question for future research.