Odors in space

Abstract

As an evolutionarily ancient sense, olfaction is key to learning where to find food, shelter, mates, and important landmarks in an animal's environment. Brain circuitry linking odor and navigation appears to be a well conserved multi-region system among mammals; the anterior olfactory nucleus, piriform cortex, entorhinal cortex, and hippocampus each represent different aspects of olfactory and spatial information. We review recent advances in our understanding of the neural circuits underlying odor-place associations, highlighting key choices of behavioral task design and neural circuit manipulations for investigating learning and memory.

Keywords: cognitive map; entorhinal cortex; hippocampus; learning and memory; olfaction; piriform cortex (PC); virtual reality.

Figure 1

Brain circuitry for integrating olfactory-spatial information. (A) Schematic of brain regions involved in odor-place coding, with arrows indicating connections between brain regions. The lateral olfactory tract (LOT) traditionally defines the boundary between anterior (a) and posterior (p) PCx. OB, olfactory bulb; M, mitral cells; T, tufted cells; AON, anterior olfactory nucleus; PCx, piriform cortex; LEC, lateral entorhinal cortex; MEC, medial entorhinal cortex; HPC, hippocampus. (B) Circuitry between PCx, LEC, and HPC with main excitatory projection cell types. Information from olfactory areas OB and AON is relayed to PCx via layer I. PCx layer IIa semilunar (SL) cells project to LEC layer I, and LEC projects back to PCx via layer IIb CB+ pyramidal cells (Pyr). LEC projects to dentate gyrus granule cells (GC) from layer IIa RLN+ fan cells (Fan) via the perforant pathway. Information received from LEC is routed through HPC via the trisynaptic circuit, from granule cells to pyramidal cells in CA3 and CA1, and back to LEC to deep layer Vb pyramidal cells. (C) Flow chart indicating the flow of odor and spatial information from the main brain regions implicated in odor-place coding: AON, PCx, LEC, and HPC. Odor information from PCx and spatial information from HPC are directed to LEC. The LEC then relays processed odor information to HPC and processed spatial information back to PCx. The AON receives direct input from HPC and has been suggested as an alternate integrator of odor and spatial information.

Figure 2

Odor-spatial paradigms for characterizing neural activity across brain areas. Left-top: A typical VR linear track set-up including a rodent head-fixed under a recording device (e.g., two-photon microscope) walking on a sphere or wheel while odor is delivered through a nose cone and visuals are shown on screens. Left-bottom: Simplified depictions of select VR linear tracks featured in this review. Odors are represented by translucent red and purple blocks on the track. Water drops represent reward. Specific visual contexts shown in green and gray. Beneath each track are simplified depictions of featured cell activities from each study. Bump represents an uptick in activity. Cell descriptions are labeled in the first panel for each experiment. Experiments labeled by citation above track depictions. Right-top: Schematic of a freely-moving rodent, allocentric spatial perspective represented in green, egocentric represented in red. Right-bottom: Freely-moving task from Igarashi et al. (50), depicting the correct paths to reward depending on odor presentation, odor A signals a reward at the left, odor B signals reward at the right. Bottom-right: Task from Poo et al. (27) with four odor trials from the same port depicted.

Figure 3

Cell activity changes with olfactory-spatial learning across the PCx-HPC circuit. Summary model of the flow of information along the HPC-LEC-PCx pathway. Schematics represent circuitry in a novel environment (left) and after an association is learned (right). Synchrony between areas is represented by a blue wavy line. Arrows between areas represent projections from cell populations with indicated tuning. Spatial (scarlet), odor (chartreuse), and conjunctive cell (orange) activity are represented by colored outlines. Learning is represented in blue between circuitry models, with synchrony and synaptic plasticity icons to symbolize the changes made between the two timepoints. Gray arrows to and from LEC represent connections that exist but are not thought to be synchronized before learning. Dashed boundaries represent cell populations that contain more than one firing type; for example, the DG exhibits odor cue cells and place cells even in novel environments, while the CA1 in the learned condition contains place cells and odor-place cells.