The brains of humans and animals have the amazing capability of extracting abstract relationships between external stimuli efficiently. Knowing such regularities helps us compute and react to novel information flexibly without prior experience. The olfactory system is no exception. Animals need to infer commonalities across different odors sharing similar meaning, while preserving individual distinctions across these odors. They must also locate and predict sources of odors to optimize their proximity to the odor-emitting objects. The neural mechanisms underlying these critical olfactory cognitions are still full of questions. In two independent fMRI experiments, we investigated how the human brain represents 1) categorically organized odor objects, and 2) two-dimensional odor space constructed by two-odorant mixtures. In the first experiment, we took advantage of the anatomical organization of the olfactory system and delivered the GABA(B) receptor agonist baclofen to suppress associative input to human piriform cortex. Multi-voxel pattern analyses revealed that baclofen disrupted pattern separation of within-category odors in the piriform cortex, and disrupted pattern separation of odor categories in the orbitofrontal cortex (OFC) and the hippocampus. In the second experiment, we uncovered evidence of grid-like coding with a two-dimensional virtual landscape constructed only of odors. We found that humans can learn to navigate through an odor space, and that the internal maps of the space take the form of periodic, hexagonally symmetric patterns in the entorhinal cortex, the ventromedial prefrontal cortex, and the anterior piriform cortex, consistent with a grid-like arrangement. Results from these experiments should advance our understanding of coding mechanisms by which the olfactory brain represents odors with relational organizations in different applications. They complement and extend earlier literature on olfactory cognition, and advance olfaction as a model system for further studies of relational memory.

Last modified

• 11/04/2019

Creator

• Xiaojun Bao

DOI

• https://doi.org/10.21985/n2-5esr-3930

Subject

• Interdepartmental Neuroscience Program (NUIN)

Keyword

• Neurosciences

Date created

• 2018-01-01

Resource type

• Dissertation

Rights statement

• In Copyright