Brain-state-independent neural representation of peripheral stimulation in rat olfactory bulb

Abstract

It is critical for normal brains to perceive the external world precisely and accurately under ever-changing operational conditions, yet the mechanisms underlying this fundamental brain function in the sensory systems are poorly understood. To address this issue in the olfactory system, we investigated the responses of olfactory bulbs to odor stimulations under different brain states manipulated by anesthesia levels. Our results revealed that in two brain states, where the spontaneous baseline activities differed about twofold based on the local field potential (LFP) signals, the levels of neural activities reached after the same odor stimulation had no significant difference. This phenomenon was independent of anesthetics (pentobarbital or chloral hydrate), stimulating odorants (ethyl propionate, ethyl butyrate, ethyl valerate, amyl acetate, n-heptanal, or 2-heptanone), odor concentrations, and recording sites (the mitral or granular cell layers) for LFPs in three frequency bands (12-32 Hz, 33-64 Hz, and 65-90 Hz) and for multiunit activities. Furthermore, the activity patterns of the same stimulation under these two brain states were highly similar at both LFP and multiunit levels. These converging results argue the existence of mechanisms in the olfactory bulbs that ensure the delivery of peripheral olfactory information to higher olfactory centers with high fidelity under different brain states.

Fig. 1.

LFP signals in the GCL under different brain states generated by different anesthetics. (A) LFP signals in three frequency bands were separated from the raw data in LBS and HBS (Upper and Lower). (B and D) The averaged time courses of LFP signals in the GCL for the three bands with chloral hydrate (n = 12) and pentobarbital (n = 13) as anesthetics, respectively. (C and E) Statistical analyses of the data in B and D, respectively. *HBS vs. LBS at rest; #rest vs. stimulated in HBS; Δrest vs. stimulated in LBS; P < 0.001 for all comparisons; no significant differences between the peak activities in LBS and HBS for any bands any anesthetic.

Fig. 2.

The LFP signals of different odorants at different recording sites in LBS and HBS. (A) Raw LFP signals at recording site three elicited by five odorants with 2-s stimulation. (B) The LFP signals of the five odorants corresponding to A at nine recording sites. None of the pattern pairs for a given odorant in LBS and HBS is significantly different (P < 0.02) (Table S2).

Fig. 3.

The effects of odor concentration on the LFP signals in LBS and HBS. (A) Raw LFP signals at three concentrations with 2-s stimulation in LBS (Upper) and HBS (Lower). (B) Statistical analyses of the LFP signals in three frequency bands; the LFP signals are significantly correlated with concentration. *, #, and Δ, are the same as in Fig. 1; P < 0.02 for all comparisons; no significant difference between the peak activities in LBS and HBS for all three bands at all three concentrations (n = 14) (Table S5).

Fig. 4.

Electrical recording in the MCL under LBS and HBS. (A and B) LFP signals (Top), multiunit spiking (Middle), and histogram of spiking (Bottom) at two recording sites with 2-s stimulation in LBS and HBS. (C) The averaged time courses of LFP signals in three bands (29 recording sites from 19 rats). (D) The statistical comparisons of data in C. (E) Multiunit signals and the statistical analyses. (F) Correlation between LFP and multiunit signals. (G) The distribution patterns of firing rate at rest and after stimulation in LBS and HBS, respectively; the last three bars were the number of neurons with firing frequency from 11 to 15, and > 15, respectively. *, #, and Δ are the same as in Fig. 1; P < 0.001 for all situations; no significant difference between the peak activities in LBS and HBS for all three LFP bands and multiunit signals; circles and filled circles in C and E are LBS and HBS, respectively.