Cortico-striatal representations of multisensory decision-making

Feinstein Institute For Medical Research, Manhasset NY

Investigators

Linked publications & trials

Abstract

Perceptual decision-making is a complex cognitive function critical for health, reproductive success, and survival. In this process, an informed choice based on sensory evidence is made through engagement of circuits across the brain. Recent studies have found the striatum to causally affect sensory-guided decisions, adding to the well-established evidence that multiple regions of the cortex also reflect decision-making. However, these cortical decision structures have not been identified in large-scale pan-neuronal studies of cortical activity, where neural responses are closely related to movement, rather than to decisions. Furthermore, during decision learning, functional reorganization, not present in other cortical layers, could be distinctly identified in layer 5 neurons, which contains the majority of cortex-wide projections that converges onto the striatum. Despite these implications, the role of the cortico-striatal pathway in decision-making behavior is poorly understood. In this proposal, I seek to use wide-field calcium imaging to examine cortex-wide activity patterns of cortico-striatal neurons in mice performing perceptual decision tasks. To specifically express genetically- encodable calcium indicators (GECIs) in cortico-striatal neurons throughout of the dorsal cortex, I propose to develop a two-component intersectional viral technique where (1) a Cre-expressing monosynaptic retrograde virus, delivered downstream at the axon terminals, labels projection pathways by (2) driving upstream Cre-dependent gene expression transduced brain-wide by the a systemically delivered adeno-associated virus variant (AAV-PHP.eB) that crosses the blood-brain barrier. After performing an intact skull-clearing procedure, I will use wide-field calcium imaging to record neural activity across the dorsal cortex in mice trained on complex auditory-, somatosensory-, or visually-guided two-alternative forced choice decision-making tasks. Neural activity will be tracked throughout learning over the course of 2-4 weeks with simultaneous video recording of animal movements. I will then use a linear encoding model to relate behavioral variables (e.g. whisking, pupil size, choice, past reward, etc) to neural activity. Subsequently, by creating cortical maps quantifying each variable?s unique contribution in predicting neural activity, I will be able to identify regions activate during the decision process. Additionally, a comparison of cortico-striatal activity between novice and expert mice can show decision-specific cortex-wide changes. Using this approach, I will be able to reveal how cortico-striatal activity is shaped by learning, identify cortical regions encoding decision-related variables, and separate core versus modality-specific decision areas.