LIN: Forschungsabteilungen > Akkustik, Lernen, Sprache > Unterpunkt Ebene 3 > Unterpunkt Ebene 4
Titel: LIN Layout
Adaptive behavior requires constant semantic evaluation of stimuli and the consequences of a subject's own actions. This is made possible by the interaction between sensory processing systems of the brain and so-called reinforcement-evaluating brain systems (Scheich and Ohl 2010, Scheich et al. 2011). Several projects in the department address the neuronal mechanisms mediating this fundamental interaction.
One of the projects focuses on the interaction between auditory cortex and corpus striatum. It has been shown that the ventral striatum is involved in behavioral strategy forming based on reward prediction of sensory cues. However, how auditory information is relayed to and processed within the ventral striatum is largely unknown. Our work has revealed several of the basic properties of auditory stimulus representation in the striatum. Additionally we have analyzed how these mechanisms are modulated once sound stimuli have been associated with a motivational value. For example, analysis of cortico-striatal spectral coherence during learning has revealed a learning-induced enhancement of cortico-striatal interplay during the early stages of learning (figure below).
A second project focuses on the fact that learning can be motivated by both reward and punishment. We study the mechanisms that lead to learning in both cases and also in the case when reward and punishment are combined in one training scenario (Ilango et al. 2010). Analysis of this newly developed learning paradigm has revealed detailed interactions between reinforcement-evaluating brain systems, e.g. the opponent action of the VTA and the lateral habenula (Shumake et al. 2010, Ilango et al. 2011). Currently we have transferred a modified version of this training protocol to the species mouse which allows us to study the potentially differential influence of reward- and punishment-motivated learning on the synaptic proteome. We also have begun to use single-photon emission computed tomography (SPECT) to study dissociable effects of different reinforcement scenarios that occur in learning.
Learning-induced increase of task-dependent coherence between ventral striatum and auditory cortex. A. Average corticostriatal coherence without sensory stimulation revealed intermittent coherence of the ongoing activity in the 8-16 Hz frequency range. B. Auditory stimulation revealed an alternating pattern of corticostriatal coherence in the beta and lower gamma range. Coherence values proved to be significantly lower (white encircled areas) and significantly higher than baseline levels (black encircled areas). C. The same stimulus as in B. was used as a conditioning cue in a Go/NoGo task. A significant rise of corticostriatal coherence occurs only during initial learning but not in later training stages.