Modulating the shift from goal-directed to habitual behaviors by manipulation of dopaminergic striatonigral loops – DOPALOOPS

Using our high throughput operant conditioning chambers, we will first measure the dynamics of goal-directed acquisition and stimulus-response based habit formation in a cue-reward contingency task in wildtype mice. We will then use targeted virus-based vectors to optogenetically modulate dopaminergic neurons projecting from the ventral tegmental area (VTA) to the dorsal striatum. The behaviorally assessed mice will then be used to identify the associated molecular changes and cellular identities by analyzing hundreds of samples from relevant brain regions using highly cost-efficient bulk RNA-seq library preparations. Further characterization will be done using single-cell RNA-seq profiles and bulk ATAC-seq profiles from the relevant regions.
Finally, we propose to extend these experimental procedures to genetically modified mice that carry humanized or non-functional alleles of Foxp2, a transcription factor associated with speech development and speech evolution. As mice humanized for Foxp2 have been shown to be affected in CBG-dependent learning and automatization, our study will shed light on the function of dopaminergic loops in CBGs as well as their potential role in the evolution and development of human speech.

During the first period of the DOPALOOPS project, for aim 1, we have accomplished the development of software and hardware for the behavioural task. Improvements of previously existing setups include a new generation of photobeam gates (hardware) as well as a new version of the extinction task (software; both hard- & software will be made available open source). Data of pilot animals have been acquired and analysed (n = 5). These data indicate that we have identified a parameter to measure habitual behaviour in mice, which is a decrease in the proportion of trials during overtraining, during which the subject refrains from collecting the reward associated with a correct response. This indicates that an animal with extensive conditioning or overtraining increasingly automatizes its actions in a way that this action is performed even in the absence of reward, a behaviour characterized indeed as “habitual”. We have furthermore piloted the injection sites on the medial VTA to specifically label dopaminergic projections from the VTA to the dorsomedial striatum. In the range of these pilot injections, we determined stereotaxic injection coordinates, excluded lentiviral constructs as insufficient to label axonal projections, established a tissue clearing protocol (iDISCO), whole-tissue staining protocol (using anti-GFP labelling to visualize viral infection and tyrosine hydroxylse labelling dopaminergic neurons) and lightsheet imaging protocol (collaboration with Dr. Nicolas Renier, ICM).
For aim 2, we have developed a tissue punch tool (hardware) and validated a protocol for tissue collection and RNA extraction, yielding excellent RNA quality such as needed for the planned RNA sequencing tasks.
For aim 3, we have established an agreement for mouse logistics between LMU Munich and the ICM in Paris to enable shipment of Foxp2 mice which are currently being tested for performing tasks under aim 3.

Our perspective follow our original plan according to the positive results we obtained from our pilots experiments, therefore the next main steps are
-Molecular characterization of habit formation in WT and FoxP2 mice.
-Optogenetic manipulation of the VTA-DMS pathway to causally demonstrate the relevance of this circuit in habit formation.
These perspectives include experimental procedure and data collection and analyses which are detailled in our main reporting document.

N/A

Cortico-basal ganglia circuits (CBGs) are essential for the learning and automatization of behaviours. CBGs are topographically organized, dynamically interact, and dopamine is a key neurotransmitter in this context. A canonical hypothesis is that reward and behaviours are bridged by dopaminergic neurons that receive input from the reward-mediating ventral striatum and project to the behavior-mediating dorsal striatum. While this hypothesis of spiraling striatal-nigro-striatal loops is plausible and important and backed up by neuroanatomical and behavioral data, it has not been addressed at a functional or molecular level. Here, we propose to combine optogenetics, automated operant learning paradigms and single-cell RNA-sequencing to explore this hypothesis in a functional context. We will first measure and modulate the dynamics of CBG learning in wildtype mice and characterize the molecular networks and properties of the involved striatal and dopaminergic neurons. We will then extend these experimental procedures to genetically modified mice that carry humanized or non-functional alleles of Foxp2, a transcription factor associated with speech development and speech evolution. As mice humanized for Foxp2 have been shown to be affected in CBG-dependent learning and automatization, our study will shed light on the function of dopaminergic loops in CBGs as well as their potential role in the evolution and development of human speech.