Die Rolle von Dopamin in cortico-striatalen Schaltkreisen bei der Steuerung zielgerichteten Verhaltens

Abstract

Instrumentelles Verhalten eines Organismus kann entweder zielgerichtet sein oder automatisiert als Gewohnheit ablaufen. Zielgerichtetes Verhalten wird durch den kausalen Zusammenhang zwischen Handlung und Ergebnis, sowie dem Wert des Ergebnisses bestimmt und passt sich somit flexibel an sich ändernde Umweltbedingungen an. Dahingegen läuft das Gewohnheitsverhalten ohne bewusste Kontrolle des deklarativen Lern – und Gedächtnissystems ab. Durch die Bildung einer Gewohnheit können so kognitive Ressourcen gespart und für gleichzeitig ablaufende Prozesse, die aufmerksamkeitsfordernd sind, genutzt werden. Zellkörperläsionsstudien zeigen nun auf, dass sowohl das zielgerichtete, als auch das automatisierte Gewohnheitsverhalten von zwei unterschiedlichen cortico-striatalen Schaltkreisen, dem assoziativen und dem sensomotorischen Schaltkreis, gesteuert werden. Ebenso haben Studien ergeben, dass beide Netzwerke zahlreiche DA-Afferenzen aus dem Mittelhirn erhalten. Zudem ist bekannt, dass DA an vielen kognitiven und motivationalen Prozessen maßgeblich beteilig ist. So gibt es mehrere Hypothesen wonach DA den Anreiz-Wert einer Belohnung kodiert oder als wichtiges Lernsignal fungiert. Jedoch ist über die Rolle von DA beim zielgerichteten Verhalten innerhalb des assoziativen Schaltkreises noch wenig bekannt Die Ergebnisse der vorliegenden Arbeiten zeigen auf, dass DA nicht an der Bildung von R-O-Assoziationen beteiligt ist, d.h. nicht alle Lernvorgänge unterstützt, die für zielgerichtetes Verhalten notwendig sind. Innerhalb des assoziativen cortico-striatalen Schaltkreises scheint DA zudem keine einheitliche Rolle einzunehmen. Vielmehr moduliert DA in den einzelnen Arealen dieses Schaltkreises unterschiedliche Prozesse. Im PL kodiert DA den Anreiz-Wert einer Belohnung bzw. die durchschnittlich zu erwartende Belohnungsmenge. Im pDMS moduliert DA eingehende kontextuelle Informationen vom EC und trägt auf diese Weise zur Detektierung des kausalen Zusammenhangs zwischen Handlung und Ergebnis bei. Im NAc core steuert DA bei einer instrumentellen Konditionierung motivationale Leistungen, bei einer klassischen Konditionierung die Bildung von Assoziationen des CS+ mit den spezifisch sensorischen Eigenschaften der Belohnung bzw. das Abrufen des Wertes einer Belohnung. Aufgrund dieser Erkenntnisse ist es fraglich, ob weiterhin an der Idee festgehalten werden kann, dass die Funktion von DA mittels einer einzigen Hypothese erklärbar ist. Vielmehr scheint es so zu sein, dass DA unterschiedliche Informationen kodiert und übermittelt, wobei in verschiedenen Hirnarealen jeweils andere Teile dieser Information genutzt werden, um unterschiedliche kognitive Prozesse zu steuern.

The prelimbic region of the prefrontal cortex (PL), the posterior subregion of the dorsomedial striatum (pDMS), the entorhinal cortex (EC) and the nucleus accumbens core (NAc core) are components of the associative cortico-striatal circuit mediating goal-directed behavior. Goal-directed behavior requires the learning of action-outcome contingencies and of outcome values, processes that could depend on dopamine (DA) signals in the PL, pDMS and NAC core. As DA is supposed to play a critical role in associative learning, we examined in 3 experiments whether dopamine signals conveyed to these brain regions are critical for goal-directed behavior. In experiment 1, rats with 6-hydroxydopamine (6-OHDA) or vehicle infusion into the PL and pDMS were trained to press two levers, one delivering food pellets, the other a sucrose solution. Thereafter, we tested whether animals were sensitive (i) to a selective degradation of one of two outcomes using a specific satiety procedure, and, (ii) to a selective degradation of one of two contingencies controlling instrumental choice behavior. Rats with PL DA depletion displayed a reduced rate of lever pressing but were sensitive to outcome devaluation and contingency degradation. Thus, PL DA seems to be critical for instrumental performance, but not for instrumental conditioning. Rats with pDMS DA depletion had intact overall response rates and were sensitive to a selective outcome devaluation; however, they failed to show a contingency degradation effect. Therefore, pDMS DA signaling seems not to be involved in instrumental performance, but in instrumental conditioning by supporting the detection of changes in the causal relationships between an action and its consequences. The capacity to detect changes in the causal efficacy of actions is mediated not only by the pDMS but also by the EC. In experiment 2, we examined whether interactions between the EC and the pDMS are required to detect changes of the instrumental contingency. Rats that received EC-pDMS disconnection lesions, i.e. unilateral cell body lesions of the EC and contralateral dopamine depletions of the pDMS, were trained to press two levers, one delivering food pellets, the other a sucrose solution. Thereafter, we tested whether rats were sensitive (i) to a selective devaluation of the value of one of two outcomes using a specific satiety procedure, and (ii) to a selective degradation of one of two contingencies controlling instrumental choice behavior. Our results reveal that rats with EC-pDMS disconnection lesions were sensitive to outcome devaluation. However, unlike rats with sham lesions or unilateral EC and pDMS lesions, rats with EC-pDMS disconnection lesions showed a reduced sensitivity to contingency degradation. These findings suggest that the EC and the pDMS are part of a neural system that supports the detection of changes in the causal relationship between an action and its consequences. Considerable evidence suggests that DA in the core subregion of the nucleus accumbens is not only involved in Pavlovian conditioning but also supports instrumental performance. However, it is largely unknown whether NAc DA is required for outcome encoding, which plays an important role both in Pavlovian stimulus-outcome learning and instrumental action-outcome learning. Therefore, in experiment 3, we tested rats with 6-OHDA induced DA depletion of the NAc core for their sensitivity to outcome devaluation in a Pavlovian and an instrumental task. Results indicate that 6-OHDA-lesioned animals were sensitive to outcome devaluation in an instrumental task. This finding provides support to the notion that NAc core DA may not be crucial in encoding action-outcome associations. However, during instrumental conditioning lever pressing rates in 6-OHDA-lesioned animals were markedly lower which could reflect an impaired behavioral activation. By contrast, after outcome-specific devaluation in a Pavlovian task, performance in 6-OHDA-lesioned animals was impaired, i.e. their magazine-directed responding was non-selectively reduced. One possibility to explain non-selective responding is that NAc core DA depletion impaired the ability of conditioned stimuli to activate the memory of the current value of the reinforcer. Taken together, the results from all three experiments reveal that DA signals conveyed to the PL, pDMS and NAC core are not critical for the learning of action-outcome associations. In fact, the experiments show that the role of DA in the associative cortico-striatal network is not uniformly, but rather complex, in that DA mediates different aspects of behavior in each target area; i.e. instrumental performance in the PL, processing of contextual information in the pDMS, and the ability of conditioned stimuli to activate reinforcer value information in the NAc core.This leads to the suggestion, that there is no single hypothesis, which is able to account for all the different functions of DA in the associative cortico-striatal network.