The role of spatial representations in MTL regions in generalizing fear responses

Keywords
No Thumbnail Available
Issue Date
2016-08-03
Language
en
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Not only exaggerated reactions to threats, but more importantly deficits in suppressing fear in safe contexts can lead to fear-related disorders. It is therefore important to study the link between threat and context encoding. Animal studies demonstrated the importance of hippocampal place cells in representing contextual information in fear learning. Human fMRI virtual reality (VR) studies conducted so far showed that spatial locations within VR environments could be decoded from the (para)hippocampus and that this region also showed different activity patterns for threat and safe contexts, suggesting a role of medial temporal lobe (MTL) regions in contextual encoding. The relation between contextual representations and fear generalization has not been investigated. We hypothesised that inadequate context representations in MTL regions play a major role in overgeneralization of fear. To assess how the strength of contextual representations influences fear generalization, we designed a differential contextual fear conditioning VR paradigm for fMRI and tested it in 24 healthy individuals. We investigated whether we could decode the contexts within the VR from patterns in four MTL regions. Fear generalization was calculated from skin conductance responses (SCR) and compared the differential conditioning response between the safe and threat context. Multi voxel pattern analysis (MVPA) was used to decode activity patterns from MTL regions in the two contexts, where accurate decoding meant that the area contained information to dissociate the contexts. Our results demonstrate that spatial location could be decoded from patterns in the right entorhinal cortex, but not in other MTL regions. No correlation was found between individual SCR generalization scores and decoding accuracies in any MTL region. In conclusion, contextual information seems to be encoded in MTL regions but it remains unclear how this relates to fear generalization.
Description
Citation
Faculty
Faculteit der Sociale Wetenschappen