Different vulnerability of fast and slow cortical oscillations to suppressive effect of spreading depolarization: state-dependent features potentially relevant to pathogenesis of migraine aura

Abstract

Background: Spreading depolarization (SD), underlying mechanism of migraine aura and potential activator of pain pathways, is known to elicit transient local silencing cortical activity. Sweeping across the cortex, the electrocorticographic depression is supposed to underlie spreading negative symptoms of migraine aura. Main information about the suppressive effect of SD on cortical oscillations was obtained in anesthetized animals while ictal recordings in conscious patients failed to detect EEG depression during migraine aura. Here, we investigate the suppressive effect of SD on spontaneous cortical activity in awake animals and examine whether the anesthesia modifies the SD effect.

Methods: Spectral and spatiotemporal characteristics of spontaneous cortical activity following a single unilateral SD elicited by amygdala pinprick were analyzed in awake freely behaving rats and after induction of urethane anesthesia.

Results: In wakefulness, SD transiently suppressed cortical oscillations in all frequency bands except delta. Slow delta activity did not decline its power during SD and even increased it afterwards; high-frequency gamma oscillations showed the strongest and longest depression under awake conditions. Unexpectedly, gamma power reduced not only during SD invasion the recording cortical sites but also when SD occupied distant subcortical/cortical areas. Contralateral cortex not invaded by SD also showed transient depression of gamma activity in awake animals. Introduction of general anesthesia modified the pattern of SD-induced depression: SD evoked the strongest cessation of slow delta activity, milder suppression of fast oscillations and no distant changes in gamma activity.

Conclusion: Slow and fast cortical oscillations differ in their vulnerability to SD influence, especially in wakefulness. In the conscious brain, SD produces stronger and spatially broader depression of fast cortical oscillations than slow ones. The frequency-specific effects of SD on cortical activity of awake brain may underlie some previously unexplained clinical features of migraine aura.

Keywords: Animal models; Aura; Cortical spreading depression; Migraine; Spreading depolarization.

Fig. 1

Pathways of SD propagation from the amygdala to the cortex. SD was triggered in the amygdala (Am) by its micro-injury via preliminary implanted guide cannula. Cortical activity was recorded in the frontal (Cx fr) and occipital (Cx occ) regions of the cortex using implanted recording electrodes. SD propagated from the injury site to the cortical regions by invading temporal cortex (A) and the striatum/frontal pole (B)

Fig. 2

Depression of spontaneous cortical activity induced by unilateral SD in awake rats. Typical recordings of dc potential (A), filtered ECoG (B) and spectrogram (C) of the 800-s fragment obtained in homotopic sites of the right (Cx, R) and left (Cx, L) occipital cortex of the two hemispheres immediately after a focal microinjury of the right amygdala (marked by red dashed line at the onset of recordings). Calibration bars – 2 mV (A) and 0.2 mV (B). The time scale is the same in A, B, C and shown below the spectrogram. A single SD event (dc shift) appeared in the right occipital cortex in 150 s after its initiation in the amygdala (A) and induced mild suppression of ipsilateral ECoG amplitude (B, C)

Fig. 3

Depression of spontaneous cortical activity induced by unilateral SD in urethane-anesthetized rats. Typical recordings of dc potential (A), filtered ECoG (B) and spectrogram (C) of the 800-s fragment obtained in homotopic sites of the right (Cx, R) and left (Cx, L) occipital cortex of the two hemispheres immediately after a focal microinjury of the right amygdala (marked by red dashed line at the onset of recordings) in the rat shown in Fig. 2 under awake conditions. Calibration bars – 2 mV (A) and 0.2 mV (B). The time scale is the same in A, B, C and shown below the spectrogram. A single SD event (dc shift) appeared in the right occipital cortex in 220 s after its initiation in the amygdala (A) and induced pronounced ECoG flattering in the ipsilateral cortex (B, C)

Fig. 4

Effect of unilateral SD on ECoG power in different frequency bands in awake rats. Graphs show mean power of delta (1–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), beta (12–25 Hz) and gamma (25–50 Hz) oscillations (marked on the right Y-axis) in the frontal (left fragments) and occipital (right fragments) cortices ipsilateral (upper fragments) and contralateral (lower fragments) to SD (n = 7). Within each band, lines with shadows mark baseline activity power, circles mark power for 10-s intervals following SD initiation, dark circles indicate intervals significantly differed from the baseline level (p < 0.05, one-way ANOVA). Gray vertical areas in the fragments show periods of DC potential shift (depolarization phase of SD) in respective cortical regions

Fig. 5

Effect of unilateral SD on ECoG power in different frequency bands in anesthetized rats. Graphs show mean power of delta (1–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), beta (12–25 Hz) and gamma (25–50 Hz) oscillations (marked on the right Y-axis) in the frontal (left fragments) and occipital (right fragments) cortices ipsilateral (upper fragments) and contralateral (lower fragments) to SD (n = 6). Within each band, lines with shadows mark baseline activity power, circles mark power for 10-s intervals following SD initiation, dark circles indicate intervals significantly differed from the baseline level (p < 0.05, one-way ANOVA). Gray vertical areas in the fragments show periods of DC potential shift (depolarization phase of SD) in respective cortical regions

Fig. 6

Magnitudes of SD-induced ECoG depression in different frequency bands in awake rats. The bars represent percentages of average power of cortical oscillations for each frequency range in the ipsilateral (dark bars) and contralateral (light bars) regions during depolarization phase of SD (dc-shift) relative to the baseline. *—p < 0.05 – significant difference from the respective baseline level within each frequency band

Fig. 7

Magnitudes of SD-induced ECoG depression in different frequency bands in anesthetized rats. The bars represent percentages of average power of cortical oscillations for each frequency range in the ipsilateral (dark bars) and contralateral (light bars) regions during depolarization phase of SD (dc-shift) relative to the baseline. *—p < 0.05 – significant difference from respective baseline levels within each frequency band

Fig. 8

Frontal gamma oscillations are depressed only after injury triggering SD. Mean power of gamma activity recorded in the ipsilateral frontal cortex of awake rats following amygdala stimulation induced (dark circles, n = 7) and not induced (white circles, sham injury, n = 6) SD. The baseline level is marked by line with shadow. Only stimulation triggering SD elicited transient reduction of frontal gamma power