Odor-evoked activity in the mouse lateral entorhinal cortex

Abstract

The entorhinal cortex is a brain area with multiple reciprocal connections to the hippocampus, amygdala, perirhinal cortex, olfactory bulb and piriform cortex. As such, it is thought to play a large role in the olfactory memory process. The present study is the first to compare lateral entorhinal and anterior piriform cortex odor-evoked single-unit and local field potential activity in mouse. Recordings were made in urethane-anesthetized mice that were administered a range of three pure odors and three overlapping odor mixtures. Results show that spontaneous as well as odor-evoked unit activity was lower in lateral entorhinal versus piriform cortex. In addition, units in lateral entorhinal cortex were responsive to a more restricted set of odors compared to piriform. Conversely, odor-evoked power change in local field potential activity was greater in the lateral entorhinal cortex in the theta band than in piriform. The highly odor-specific and restricted firing in lateral entorhinal cortex suggests that it may play a role in modulating odor-specific, experience- and state-dependent olfactory coding.

Figure 1

Electrode placement in the a) LEC of 13 animals and b) aPCX of 9 animals. Coronal stereotaxic images showing the approximate placement of the final tip location of recording electrode tracks (black dots). Images adapted from (Franklin and Paxinos, 2008).

Figure 2

Representative odorant responsivity to 4 sample odors in a single unit in the LEC from a single mouse. The top half shows waveform activity in response to a single presentation of the named odorant. The bottom half shows raster plots (dots) of responsivity to each presentation of an odor. Rasterplots and peristimulus time histograms (PSTH) show reliability of responses. Data for rasters and PSTHs obtained by extraction of single-unit activity from the raw recordings (see Methods). Overlapping single-unit waveforms shown on top. Odorant presentation time is enclosed in the shaded area starting from 0 seconds to 2 seconds.

Figure 3

Spontaneous activity (a) and the maximal odor-evoked response to the best odor (b) in aPCX versus LEC. Both spontaneous activity (p <.01) and aPCX maximal odor-evoked responses (p < .001) were significantly greater than those in LEC. Error bars represent ± 1 SEM.

Figure 4

a) Single-unit odor receptive fields in the aPCX are significantly more broad than in LEC. Responses are normalized to best odor within each cell. Therefore, a measure of 1 is the response to the best odor with subsequent responses being expressed as a percentage of that response strength. The strength of LEC responses to the fourth, fifth, and worst odors was significantly lower than matched responses in the aPCX (p<.05). Error bars represent ± 1 SEM. b) Receptive field slopes in aPCX and LEC plotted for each unit recorded. Here, aPCX cells demonstrate a reduced slope compared LEC units. Response magnitude of each unit to each odor in c) aPCX and d) LEC. aPCX units are much more broadly tuned compared to LEC.

Figure 5

Respiration entrainment in LEC vs. aPCX units. a) Averaged respiration waveform, aPCX single-unit activity displayed as rasters (middle) and as a phase plot (bottom) relative to the respiratory cycle. This unit showed highly significant entrainment to the respiratory cycle as assessed with Rayleigh statistics. b) The proportion of units firing in phase with respiration in the LEC and aPCX).Significantly more aPCX units fired in phase with respiration compared to LEC.

Figure 6

Representative field potential recordings from a) aPCX and b) LEC showing LFP as a sonogram and waveform as well as individually digitally filtered theta, beta, and gamma components. Odorant stimulation duration (2 sec.) is depicted by the shaded area. Odors evoked enhanced oscillatory activity in LEC compared to the aPCX. c) Spontaneous LFP activity in aPCX and LEC. While theta band power was significantly higher in aPCX compared to LEC, beta and gamma band spontaneous activity were higher in LEC than in aPCX. d) Odor-evoked oscillations in the aPCX and LEC across all odors. LEC evoked oscillations were significantly greater than aPCX specifically within the theta band. Error bars represent ± 1 SEM.

Figure 7

Power change in odor-evoked theta oscillations recorded in LEC and aPCX in response to pure, monomolecular odors and odor mixtures. Pure odors elicited a significantly larger power change than odor mixtures in aPCX (p<.01) though this difference was absent in LEC. In addition, pure odors elicited a larger power change than mixtures in both LEC and aPCX (F(1, 116) = 4.09, p <.05). Furthermore, LEC theta power change in response to both types of odors was greater than in aPCX (F(1, 116) = 12.14, p <.001).

Figure 8

a) Raster plots of spontaneous and odor evoked beta waves in aPCX and LEC of a representative dual recording animal are shown in the top panels with respiration phase shown in the bottom panel. Note the high correlation of beta activity to respiration in aPCX and the weaker correlation in LEC. The transition between inhalation and exhalation occurs at time 0. b) Sonogram of oscillatory activity in LEC and aPCX relative to respiration (bottom panel) through a 1 second time span. Beta activity occurred consistently in phase with respiration in aPCX while not being phase locked to any phase of respiration in LEC. c) Mean computed vector length and angle across all 4 animals (transition between inhalation and exhalation occurs at angle 0). Note circle radius is 0.5.