Differential effects of isoflurane on high-frequency and low-frequency γ oscillations in the cerebral cortex and hippocampus in freely moving rats

Abstract

Background: Cortical γ oscillations are thought to play a role in conscious cognitive functions. Suppression of 40-Hz γ activity was implicated in the loss of consciousness during general anesthesia. However, several experimental studies found that γ oscillations were preserved in anesthesia. The authors investigated the concentration-dependent effect of isoflurane on spontaneous γ oscillations in two frequency bands and three distinct brain regions in the rat.

Methods: Adult Sprague-Dawley rats were chronically implanted with epidural and coaxial depth electrodes to record cortical field potentials in frontal cortex, visual cortex, and hippocampus in waking and at steady-state isoflurane concentrations of 0.4, 0.8, and 1.2%. The γ power was calculated for the frequency bands 30-50 and 70-140 Hz. Temporal variation and interregional synchrony of γ activity were analyzed using wavelet transform. Loss of consciousness was indexed by the loss of righting reflex.

Results: Rats lost their righting reflex at 0.8 ± 0.1% isoflurane. High-frequency γ power was decreased by isoflurane in a concentration-dependent manner (P < 0.001, 50% decrease at 0.8% isoflurane) in all brain regions. Low-frequency γ power was unaffected by isoflurane. The duration and interregional synchrony of high-frequency γ bursts was also reduced (P l < 0.001, 40% decrease at 0.8% isoflurane).

Conclusions: Distinction between high- and low-frequency γ bands is important when evaluating the effect of general anesthetics on brain electrical activity. Spontaneous 40-Hz γ power does not indicate the state of consciousness. The attenuation and interregional desynchronization of high-frequency γ oscillations appear to correlate with the loss of consciousness.

Figure 1

Examples of γ oscillations in the frontal cortex of one rat at four levels of isoflurane. All traces are original recordings, sampled at 2.5 kHz after 1-500 Hz analog band-pass filtering, and subsequently down-sampled to 500 Hz. All traces are 3s long. The γ bursts are marked with asterisks.

Figure 2

Time-frequency power as calculated by the complex Morlet wavelet transform in three brain regions of a waking rat. Note the high temporal dynamics of γ power in all cases.

Figure 3

Power spectra in the three different brain regions and at four levels of consciousness in one rat. Power spectra were obtained with Welch's spectral estimation method using 250 points at 500 Hz sampling frequency and 90% window overlap. Note the progressive decrease in high-frequency (70-140 Hz) γ power with an increasing concentration of isoflurane.

Figure 4

Concentration-dependent effect of isoflurane on cortical and hippocampal γ power from all experiments. In each of the three brain regions, isoflurane suppressed high-frequency (70-140 Hz) γ power but it did not significantly alter low-frequency (30-50Hz) γ power.

Figure 5

The effect of isoflurane (Iso) on the properties of γ bursts. A: Example of thresholding of high-frequency γ bursts at median value after 1 Hz low-pass filtering; data are from frontal cortex. SD=standard deviation. B: Summary of data from all experiments for γ burst duration, intraburst power, and interregional burst synchrony in four states of consciousness in the low (30-50 Hz) and high (70-140 Hz) γ band. Isoflurane produced significant decreases in high-frequency γ burst duration and synchrony (linear trend, p<0.001).