Journal Logo

Clinical Neuroscience

Anesthesia enhances spontaneous low-frequency oscillations in the brain

Zhang, Zhuo; Li, Fuquan; Li, Ming; Hu, Dewen*,

Author Information
doi: 10.1097/WNR.0000000000001418
  • Open

Abstract

Introduction

Spontaneous oscillations play an important role in the brain and body. Some researchers have linked spontaneous low-frequency oscillation (LFO) signals to basic functions of the brain [1]. Spontaneous LFO signals are associated with a plethora of cortical functions because they synchronize diverse operations across neuronal networks and coordinate gross excitability [2]. Characterizing these oscillations may also be helpful for revealing some other underlying mechanisms, such as the autoregulatory mechanism of the brain vasculature [3,4]. Consequently, spontaneous LFO signals have attracted widespread interest. Fox and Raichle [5] reviewed recent studies examining spontaneous fluctuations as being potentially important and revealing manifestations of spontaneous neuronal activity.

In such studies, the animals needed to be under anesthesia. However, because anesthesia can alter neural activity, vascular tone, and neurovascular coupling [6,7], it is believed that LFOs can be affected by the application of anesthetic and the depth of anesthesia. Therefore, it is necessary to characterize the influence of anesthetic on spontaneous oscillation signals.

In fact, many studies have focused on this issue. Mitra et al. [8] explored the spatiotemporal structure of infraslow (0.02–0.1 Hz) and higher frequency (1–4 Hz) spontaneous neural activity in awake and anesthetized mice and found that the spread direction of the oscillations was state-dependent (awake versus anesthesia). Pal et al. [9] studied the effect of three clinically used anesthetics (propofol, sevoflurane, and ketamine) on phase-amplitude coupling in the frontal cortex of rats and found that the anesthesia substantially impacted the coupling of metabolism and neural activity and also altered the temporal and spatial characteristics of spontaneous fluctuations. Astashev et al. [10] compared the anesthesia effects of two anesthetics (zoletil-xylazine and zoletil-nitrous oxide mixtures) on LFOs and found that myogenic, neurogenic, and endothelial oscillations can be changed diversely by different anesthetics. Some results have shown that the brain not only exhibits intrinsic-organized fluctuations in neuronal activity, but that these fluctuations impact brain function and behavior in interesting and important ways [11]. Klein et al. [12,13] observed differential spatial distribution of these LFO amplitudes within the human resting brain and provide a foundation for continued examination of LFO amplitude in populations. The results of the study by Fultz et al. [14] concluded that human sleep is associated with large coupled LFO in neuronal activity, blood oxygenation, and cerebrospinal fluid flow. And they also thought that the LFO is linked to the waste clearance in the human brain.

These studies clearly showed that anesthetic can affect LFOs [15,16]. However, the effect of anesthesia on oscillation frequency and amplitude across various anesthesia depths remains unclear. And the depth of anesthesia will likely to cause change in nervous system activity or hemodynamic parameters, which includes cerebral blood volume (CBV) and deoxygenated hemoglobin (Hb). In this study, these features and parameters of LFO signals are explored by gas-mouse model.

Methods

This study was performed in strict accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and was approved by the Medical Ethics Committee of 921 Hospital, Chinese Academy of Sciences.

Animal preparation and anesthesia concentration settings

Before the experiment, 54 male adult Kun Ming mice (40 ± 3 g) were fed a standard diet and water for 12 h in a room kept at 22 ± 2°C. The mice were randomly divided into three groups and were anesthetized at three concentrations (1.0, 1.2, and 1.5% isoflurane). During surgery, the isoflurane concentrations are set to be 1.0%. After removing the parenchyma and muscles under the scalp, the skull above the somatosensory cortex was exposed and thinned to build a window (size: 8 × 8 mm) in which the blood vessels are clearly visible. Then, the mice were positioned in a standard stereotaxic frame, and the thinned region was kept horizontal. A homeothermic blanket was used to maintain the rectal temperature at 37 ± 0.5°C, and the heartbeat and respiration was monitored as data recording. Details on the process of surgery and data observation are given in Refs. 17 and 18.

Data acquisition and analysis

Changes in neural signals were detected by observing changes in hemodynamic parameters via the Optical Imaging (OI) system (Imager 3001; Optical Imaging Inc., Germantown, New York, USA), which was used to capture the primary somatosensory cortex under red (605 ± 10 nm) and green (546 ± 10 nm) illumination. The OI system consists of a high-precision Charge Coupled Device (CCD) (12-bit, 1024 × 1024 pixels, 60 dB), a cold light source and an optical path. Two front-to-front camera lenses (35 mm) were connected to the CCD to obtain a very shallow depth of field (~50 μm) and a broad field of view. In three given regions, data were acquired at 5 Hz with an observation window of 400 s for each recording. We focused on the somatosensory cortex surface by setting a proper object distance to visualize the cortical vessels most clearly [18].

In the optical imaging of intrinsic signals, the different metabolism signals can be monitored at different illumination. In our work, the wavelength of 546 ± 10 nm (green) and 605 ± 10 nm (red) are used to observed the oscillations of CBV and Hb concentration, respectively [18,19]. The mean time series was defined as the change in the average grey level within the region of interest (ROI). The most primary vein and its adjacent artery are, respectively, selected as the ROIs of veins and arteries, and the cortex between the artery ROI and vein ROI was selected as the ROI of cortex. The spectral properties of the processed time series were obtained using the traditional fast Fourier transform (FFT) method and the multitaper analysis method (MTM), and then, a statistical F-test was performed to verify the significance of the power density at each frequency point. Only the frequency points that met the following three conditions were chosen as valid LFO frequency points: (1) the FFT amplitude-frequency spectrum reaches its peak; (2) the MTM power spectrum reaches its peak; and (3) the F-value is larger than the threshold corresponding to the confidence level of 95% (F(2,12)). See the example in Fig. 1c. The frequency and amplitude difference between the three anesthesia depths was examined using t-test which was implemented by MATLAB (the MathWorks, Natick, MA, USA) software. The P values were corrected by the Benjamini–Hochberg algorithm for multiple comparisons, and the significant level was set to be 0.05. The strength of effect is evaluated by the Cohen’s d effect size (>0.8 means large effect).

Fig. 1
Fig. 1:
The frequency and amplitude characteristics of spontaneous LFO signals under different anesthesia depths. (a) The data were recorded from the somatosensory cortex (ROI1: artery, ROI2: vein, and ROI3: cortex); the three ROIs are indicated by forward slashes, backward slashes, and crossed lines. The relative position of the illumination source and the ROI is showed at the up-right corner. (b) An example of the mean time series of the observed intrinsic signal. (c) A demonstration of the determination of LFO frequencies (indicated by the dashed vertical line). An LFO frequency point must meet the following three conditions: (1) the FFT amplitude-frequency spectrum reaches its peak; (2) the MTM power spectrum reaches its peak; and (3) the F-value is larger than the threshold corresponding to the 95% confidence level (red horizontal line). (d–e) Frequency comparisons on the LFO from the whole region at different anesthesia depths. Panel (d): under green illumination; panel (e): under red illumination. There is no statistically significant difference in the frequency between the three groups. (f–g): Amplitude comparisons on the LFO from the whole region at different anesthesia depths. Panel (f): under green illumination, the amplitude increased significantly with the anesthesia depth (P 1.0 vs. 1.2% < 0.01, P 1.0 vs. 1.5% < 0.001, P 1.2 vs 1.5% < 0.001; ES1.0 vs 1.2% = 0.92, ES1.0 vs. 1.5% = 0.95, ES1.2 vs. 1.5% = 0.93). Panel (g): under red illumination, there was no significant difference in the amplitudes. Panels (h–m): Amplitude comparisons on the LFO from the different ROIs at different anesthetic depths (the three ROIs are indicated by forward slashes, backward slashes, and crossed lines). The results are consistent with the findings in (f–g). Under green illumination (panels h–j), the amplitude increase is significant (artery: P 1.0 vs. 1.2 = 0.048, P 1.0% vs. 1.5% < 0.001, P 1.2% vs. 1.5% < 0.001; ES1.0% vs. 1.2% = 0.85, ES1.0% vs. 1.5% = 0.99, ES1.2% vs. 1.5% = 0.98. vein: P 1.0% vs. 1.2% = 0.67, P 1.0% vs. 1.5% < 0.01, P 1.2% vs. 1.5% < 0.01; ES1.0% vs. 1.2% = 0.29, ES1.0% vs. 1.5% = 0.93, ES1.2% vs. 1.5% = 0.92. cortex: P 1.0% vs. 1.2% = 0.035, P 1.0% vs. 1.5% < 0.001, P 1.2% vs. 1.5% < 0.001; ES1.0% vs. 1.2% = 0.86, ES1.0% vs. 1.5% = 0.96, ES1.2% vs. 1.5% = 0.93), and under red illumination, no significant result is found (Panel k–m). Data are shown as the means, and the sample size was 54 (see the Methods section). Error bars indicate standard error. *Indicates that the difference is significant at the 0.05 level. **Indicates significance at the 0.01 level. ES, effect size; LFO, low-frequency oscillations; MTM, multitaper analysis method; n.s., no significant difference; ROI, region of interest.

Results

Low-frequency oscillation frequencies under different anesthesia depths

To answer whether LFO frequencies drift with anesthesia depth, we examined LFO frequencies at different anesthesia depths.

LFO frequencies were determined from the averaged time series of all pixels using MTM and FFT methods (see Data analysis). Figure 1d and e compares the oscillation frequencies under different anesthetic depths (panel d: green illumination and panel e: red illumination). No significant difference was found in LFO frequencies between the three anesthetic concentrations under green or red illumination, which implies that anesthetic depth has no effect on the frequency of spontaneous LFO signals.

Low-frequency oscillation amplitudes under different anesthesia depths

The amplitude values of the LFO signals were extracted by calculating the FFT amplitude of the averaged time series of the whole region (all pixels) at the oscillation frequency. Under green illumination, an increasing trend in LFO amplitudes was found in most animals as the depth of anesthesia increased. Figure 1f shows the mean values of the amplitudes of the oscillations under green illumination. The mean amplitude of the spontaneous LFO signals was substantially affected by the concentration of the anesthetic. The mean amplitude increased as anesthesia deepened. However, under red illumination, the mean amplitude did not significantly change with the depth of anesthesia (Fig. 1g).

LFO amplitudes in different ROIs (artery, vein, and cortex) were also studied separately, and the results are consistent with the findings for the whole region (Fig. 1f and g). With the increasing depth of anesthesia, LFO amplitudes increased in all the ROIs under green illumination (Fig. 1hj), while under red illumination, the amplitude difference is insignificant in all ROIs (panels h–j). Thus, we suggest that anesthesia can affect the LFO amplitude under green illumination but not under red illumination.

Amplitudes at other frequencies under different anesthesia depths

To study the anesthesia effect at other frequencies, we also examined the amplitude-frequency curves within 0–1 Hz under different anesthesia depths (Fig. 2). Figure 2a shows the amplitude-frequency characteristics under green illumination. The amplitude significantly increased in the ultralow band (0–0.2 Hz) and low band (0.2–1 Hz) with the deepening of the anesthesia. Furthermore, the anesthesia effect showed differences in different subbands. It seems that the amplitudes affected by the anesthesia were more sensitive within ultralow band than within the low band. To evaluate the different anesthesia effects in different sub-bandwidths, we calculated the total energy within different subbands at different anesthesia depths (see the bar diagrams). In the bar diagrams embedded in Fig. 2a, the energy within ultralow band increased by 1242.8% from 1 to 1.5% isoflurane, while within the low band, the energy increased by only 465.6%.

Fig. 2
Fig. 2:
The amplitude-frequency spectrum under different anesthesia depths. (a) Under green illumination; (b) under red illumination. Blue: curves under 1.0% isoflurane; yellow: 1.2% isoflurane; and purple: 1.5% isoflurane. The colored shadow indicates the stand error of each curve. The bar diagrams compare the energy within the ultralow band (0–0.2 Hz, dark rectangular area) and low band (0.2–1 Hz, light rectangular area). The energy within the ultralow band is more sensitive to anesthesia growth under green illumination: the ultralow band energy increased 1242.8%, which is a much larger increase than the increase of 465.6% seen within the low band (ultralow band: P 1.0 vs. 1.2% < 0.01, P 1.0 vs. 1.5% < 0.001, P 1.2 vs. 1.5% < 0.001; ES1.0 vs. 1.2% = 0.88, ES1.0 vs. 1.5% = 0.95, ES1.2 vs. 1.5% = 0.93. Low band: P 1.0 vs. 1.2% < 0.01, P 1.0 vs. 1.5% < 0.001, P 1.2 vs. 1.5% < 0.01; ES1.0 vs. 1.2% = 0.90, ES1.0 vs. 1.5% = 0.96, ES1.2 vs. 1.5% = 0.89). Under red illumination, the energy is insensitive to anesthesia depth in any frequency band. Error bars indicate standard error. *Indicates that the difference is significant at the 0.05 level. **Indicates that the difference is significant at the 0.01 level. ES, effect size; n.s., no significant difference.

Figure 2b shows the amplitude-frequency curves within the ultralow band and low band under red illumination. The three amplitude-frequency curves under different anesthesia depths are almost the same. In the bar diagrams, the energy within different sub-bandwidths shows little significant difference when the depth of anesthesia changes.

Next, the results in Fig. 2 imply that the anesthetic could enhance the intrinsic signals under green illumination, especially within the lower band, and had no effect on the signal under red illumination. This finding is consistent with the anesthesia effect on LFO amplitudes in Fig. 1fm.

Discussion

LFOs may play an important role in brain regulation mechanisms and functional connections [5]. Some work has focused on amplitudes, frequencies, correlations, etc., and have found that these parameters may be affected by the stimulation or the type of anesthetic [9,10]. However, the effect of anesthesia depth on amplitude-frequency has not been reported. Our findings showed that the frequency of LFOs is stable at different anesthesia depths, while the amplitude can be enhanced by anesthesia in the artery, vein, and cortex. It has been reported that isoflurane causes vasodilation [20], which can explain the amplitude enhancement after anesthesia application.

It should be noted that the fluctuation enhancement was observed only under green illumination and not under red illumination. And because CBV is the major component under the green illumination and Hb concentration is the major component under red, the finding shows that only the CBV was enhanced by anesthesia, and the Hb concentration was not. Because the Hb concentration is the ratio of the Hb amount to the CBV [18], a stable Hb concentration implied that Hb amount increased as the CBV increased. Therefore, it can be reasoned that anesthesia can promote nervous system activity and blood supplement. Thrane et al. [21] reported that the Electrocorticography (ECoG) power in the low-frequency band increased after isoflurane was applied. This could explain the increase in the CBV induced by anesthesia.

In summary, we provided evidence of the effects of anesthesia on blood flow and nervous system signals and confirmed that CBV oscillations can be enhanced by isoflurane. We also observed that the oscillation frequency and the Hb concentration are insensitive to the anesthesia depth. Our findings may provide some reference for anesthesia in animal experiments. Many researchers have studied the role of spontaneous oscillations in humans [22,23]. For example, Spironelli and Angrilli [24] compared spontaneous activations partly differ and overlap by Electroencephalography (EEG), functional magnetic resonance imaging (FMRI), and Near-infrared spectroscopy (NIRS) when the humans were sitting or horizontal bed rest. And Astashev et al. [10] proved that low-oscillation processes have universal nature and are independent from the species or regulatory mechanisms. So, we believe that the mice model is appropriate for studying the so-called resting state in humans.

Acknowledgements

Conflicts of interest

There are no conflicts of interest.

References

1. Buzsaki G. Rhythms of the Brain. 2006, Oxford: Oxford University Press
2. Mather M, Nga L. Age differences in thalamic low-frequency fluctuations. NeuroReport. 2013; 24:349–353
3. Diehl RR, Linden D, Lücke D, Berlit P. Phase relationship between cerebral blood flow velocity and blood pressure. A clinical test of autoregulation. Stroke. 1995; 26:1801–1804
4. Hu HH, Kuo TB, Wong WJ, Luk YO, Chern CM, Hsu LC, Sheng WY. Transfer function analysis of cerebral hemodynamics in patients with carotid stenosis. J Cereb Blood Flow Metab. 1999; 19:460–465
5. Fox MD, Raichle ME. Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging. Nat Rev Neurosci. 2007; 8:700–711
6. Keilholz SD, Pan WJ, Billings J, Nezafati M, Shakil S. Noise and non-neuronal contributions to the BOLD signal: applications to and insights from animal studies. Neuroimage. 2016; 154:267–281
7. Lalou DA, Czosnyka M, Donnelly J, Lavinio A, Pickard JD, Garnett M, Czosnyka Z. Influence of general anaesthesia on slow waves of intracranial pressure. Neurol Res. 2016; 38:587–592
8. Mitra A, Kraft A, Wright P, Acland B, Snyder AZ, Rosenthal Z, et al. Spontaneous infra-slow brain activity has unique spatiotemporal dynamics and laminar structure. Neuron. 2018; 98:297–305.e6
9. Pal D, Silverstein BH, Sharba L, Li D, Hambrecht-Wiedbusch VS, Hudetz AG, Mashour GA. Propofol, sevoflurane, and ketamine induce a reversible increase in delta-gamma and theta-gamma phase-amplitude coupling in frontal cortex of rat. Front Syst Neurosci. 2017; 11:41
10. Astashev ME, Serov DA, Tankanag AV. Anesthesia effects on the low frequency blood flow oscillations in mouse skin. Skin Res Technol. 2019; 25:40–46
11. Fox MD, Snyder AZ, Vincent JL, Raichle ME. Intrinsic fluctuations within cortical systems account for intertrial variability in human behavior. Neuron. 2007; 56:171–184
12. Zuo XN, Di Martino A, Kelly C, Shehzad ZE, Gee DG, Klein DF, et al. The oscillating brain: complex and reliable. Neuroimage. 2010; 49:1432–1445
13. Klein C, Liem F, Hänggi J, Elmer S, Jäncke L. The “silent” imprint of musical training. Hum Brain Mapp. 2016; 37:536–546
14. Fultz NE, Bonmassar G, Setsompop K, Stickgold RA, Rosen BR, Polimeni JR, Lewis LD. Coupled electrophysiological, hemodynamic, and cerebrospinal fluid oscillations in human sleep. Science. 2019; 366:628–631
15. Cho D, Shin TJ, Ham J, Choi DH, Kim S, Jeong S, et al. Differential modulation of thalamo-parietal interactions by varying depths of isoflurane anesthesia. PLoS One. 2017; 12:e0175191
16. Jiang S, Miao B, Chen Y. Prolonged duration of isoflurane anesthesia impairs spatial recognition memory through the activation of JNK1/2 in the hippocampus of mice. NeuroReport. 2017; 28:386–390
17. Huang L, Liu Y, Gui J, Li M, Hu D. Stimulus-dependent modulation of spontaneous low-frequency oscillations in the rat visual cortex. NeuroReport. 2014; 25:823–828
18. Li-ming LIU. Spatio-temporal analysis of stimuli-modulated spontaneous low frequency oscillations. Chinese Science Bulletin. 2007; 52:1475–1483
19. Pouratian N, Cannestra AF, Martin NA, Toga AW. Intraoperative optical intrinsic signal imaging: a clinical tool for functional brain mapping. Neurosurg Focus. 2002; 13:e1
20. Redfors B, Shao Y, Omerovic E. Influence of anesthetic agent, depth of anesthesia and body temperature on cardiovascular functional parameters in the rat. Lab Anim. 2014; 48:6–14
21. Thrane AS, Rangroo Thrane V, Zeppenfeld D, Lou N, Xu Q, Nagelhus EA, Nedergaard M. General anesthesia selectively disrupts astrocyte calcium signaling in the awake mouse cortex. Proc Natl Acad Sci USA. 2012; 109:18974–18979
22. Svein Aslak L, Per K, Alan B, Aneta S, Kirkeboen KA. The effects of general anesthesia on human skin microcirculation evaluated by wavelet transform. Anesth Analg. 2007; 105:1012–1019
23. Per K, Svein Aslak L, Alan B, Aneta S, Hebe Désirée K, Knut Arvid KE. Low-frequency oscillations of the laser Doppler perfusion signal in human skin. Microvasc Res. 2006; 72:120–127
24. Spironelli C, Angrilli A. Posture used in fMRI-PET elicits reduced cortical activity and altered hemispheric asymmetry with respect to sitting position: an EEG resting state study. Front Hum Neurosci. 2017; 11:621
Keywords:

anesthetic concentration; multitaper analysis method; optical imaging; spontaneous low-frequency oscillations

© 2020 Wolters Kluwer Health | Lippincott Williams & Wilkins