Odor fear conditioning modifies piriform cortex local field potentials both during conditioning and during post-conditioning sleep

Abstract

Background: Sleep plays an active role in memory consolidation. Sleep structure (REM/Slow wave activity [SWS]) can be modified after learning, and in some cortical circuits, sleep is associated with replay of the learned experience. While the majority of this work has focused on neocortical and hippocampal circuits, the olfactory system may offer unique advantages as a model system for exploring sleep and memory, given the short, non-thalamic pathway from nose to primary olfactory (piriform cortex), and rapid cortex-dependent odor learning.

Methodology/principal findings: We examined piriform cortical odor responses using local field potentials (LFPs) from freely behaving Long-Evans hooded rats over the sleep-wake cycle, and the neuronal modifications that occurred within the piriform cortex both during and after odor-fear conditioning. We also recorded LFPs from naïve animals to characterize sleep activity in the piriform cortex and to analyze transient odor-evoked cortical responses during different sleep stages. Naïve rats in their home cages spent 40% of their time in SWS, during which the piriform cortex was significantly hypo-responsive to odor stimulation compared to awake and REM sleep states. Rats trained in the paired odor-shock conditioning paradigm developed enhanced conditioned odor evoked gamma frequency activity in the piriform cortex over the course of training compared to pseudo-conditioned rats. Furthermore, conditioned rats spent significantly more time in SWS immediately post-training both compared to pre-training days and compared to pseudo-conditioned rats. The increase in SWS immediately after training significantly correlated with the duration of odor-evoked freezing the following day.

Conclusions/significance: The rat piriform cortex is hypo-responsive to odors during SWS which accounts for nearly 40% of each 24 hour period. The duration of slow-wave activity in the piriform cortex is enhanced immediately post-conditioning, and this increase is significantly correlated with subsequent memory performance. Together, these results suggest the piriform cortex may go offline during SWS to facilitate consolidation of learned odors with reduced external interference.

Figure 1. Representative data from one animal showing transition between behavioral states.

(A) Left, a representative example showing a transition from SWS into REM sleep and returning to SWS as recorded in the anterior piriform cortex (PCX). Each point on the line graph represents a fourteen second time window. The waveforms below are raw LFP and EMG data showing the same window of time as the line graphs above. Note the change in both LFP and EMG frequencies when the animal enters REM sleep. Right, an example of the same animal transitioning from SWS to REM to Awake then returning to SWS. SWS is characterized by high delta power activity and relatively low EMG activity. REM is typified by low delta LFP activity and very low EMG activity. Awake state is distinguished by high frequency activity (lower delta) in the LFP and high frequency EMG waveforms. (B) A mean hypnogram recorded in the anterior piriform cortex of 4 naïve rats placed individually in the recording chamber at 3 p.m. for 24 h.

Figure 2. Odor evoked activity in the piriform cortex changes across behavioral states.

(A) Representative odor evoked anterior piriform cortex activity during awake, REM, and SWS from one animal. During both awake and REM odor stimulation increased activity in the gamma (35–85 Hz) frequency band. There was no obvious odor-evoked activity in the piriform cortex during SWS. (B) During Awake and REM, there was significantly greater odor evoked activity in all frequency bands (mean odor-evoked activity ± SEM) compared to SWS.

Figure 3. Odor-fear conditioning enhances odor-evoked freezing and odor-evoked gamma oscillations.

(A) Paired (Test Cue+Context and Test Cue Only) odor-shock animals significantly increased odor-evoked freezing over the course of odor/shock conditioning trials. Furthermore, during odor only tests the following day, (B) the Paired animals maintained their odor-evoked freezing response when tested either in the context of the conditioning chamber (n = 6) or in a novel context (n = 4). Unpaired animals (n = 4), however, showed significantly less odor-evoked freezing. During the post-conditioning test, the Unpaired animals show no freezing response to the odor. The legend in Figure B applies also to Figure A. (C) Mean level of freezing behavior during the testing session 24 hr post-conditioning across the three groups. (D) Power spectrum analysis showed Paired animals had a significant increase in odor-evoked gamma frequency activity over the course of conditioning trials. Odor-evoked gamma (35–85 Hz) activity was significantly higher on average in the second half of trials compared to the first half in Paired animals. There was no change in Unpaired animals. Asterisks signify significant difference between groups.

Figure 4. Paired rats increased time spent in post-conditioning SWS.

(A) Following odor aversion conditioning, paired animals spent more time in SWS recorded in the piriform cortex than on baseline days (mean SWS duration ± SEM). This increase was seen only in odor/shock Paired animals. Immediately following conditioning (marked by Day Conditioning), Paired animals spent significantly more time in SWS than on baseline days and more time than Unpaired animals. This increase was maintained on the day of testing. There was no significant change in REM or total sleep after conditioning in either group (not shown). Asterisks signify significant difference between groups. (B) There was a significant correlation between the amount of time spent in SWS immediately after training and the duration of odor-evoked freezing (strength of memory) observed the next day.

Figure 5. Recording electrode tip locations.

Coronal sections of rat's brain with distances from Bregma indicated. The black dots represent the recording site of each piriform cortex LFP. Atlas plates from Paxinos and Watson .