Sevoflurane Effects on Neuronal Energy Metabolism Correlate with Activity States While Mitochondrial Function Remains Intact

Abstract

During general anesthesia, alterations in neuronal metabolism may induce neurotoxicity and/or neuroprotection depending on the dose and type of the applied anesthetic. In this study, we investigate the effects of clinically relevant concentrations of sevoflurane (2% and 4%, i.e., 1 and 2 MAC) on different activity states in hippocampal slices of young Wistar rats. We combine electrophysiological recordings, partial tissue oxygen (p_tiO₂) measurements, and flavin adenine dinucleotide (FAD) imaging with computational modeling. Sevoflurane minimally decreased the cerebral metabolic rate of oxygen (CMRO₂) while decreasing synaptic transmission in naive slices. During pharmacologically induced gamma oscillations, sevoflurane impaired network activity, thereby decreasing CMRO₂. During stimulus-induced neuronal activation, sevoflurane decreased CMRO₂ and excitability while basal metabolism remained constant. In this line, stimulus-induced FAD transients decreased without changes in basal mitochondrial redox state. Integration of experimental data and computer modeling revealed no evidence for a direct effect of sevoflurane on key enzymes of the citric acid cycle or oxidative phosphorylation. Clinically relevant concentrations of sevoflurane generated a decent decrease in energy metabolism, which was proportional to the present neuronal activity. Mitochondrial function remained intact under sevoflurane, suggesting a better metabolic profile than isoflurane or propofol.

Keywords: metabolism; neuron; sevoflurane.

Figure 1

Sevoflurane-induced changes on cerebral metabolic rate of oxygen (CMRO₂) and synaptic activity in naive slices. (a) Representation of recording settings for simultaneous measurements of partial tissue oxygen pressure (p_tiO₂) and stimulus-induced population spikes (PSs). Left: Field potential (f.p.) electrode and Clark-style oxygen electrode (p_tiO₂) were positioned in the stratum pyramidale of area CA1 while a stimulation electrode (black dots) was placed in the stratum radiatum in area CA2. Right: Exemplary recording of p_tiO₂ depth profile in slices gassed with carbogen in an interface condition. Under these conditions, the p_tiO₂ of approximately 682 mmHg at the surface of the slices decayed until the lowest oxygen values at the core of the slice. (b) Exemplary p_tiO₂ depth profiles and CMRO₂ values as calculated using a reaction–diffusion model in the different conditions: control (CTL, black), under 2% sevoflurane (green), under 4% sevoflurane (red), and after washout (WO, grey). (c) Plots of recorded absolute CMRO₂ values (left) and normalized CMRO₂ changes (middle) showing a small but significant decrease under 2% sevoflurane, no further significant changes under 4% sevoflurane, and reversibility. Based on the experimental data, the computed relative adenosine triphosphate (ATP) consumption rate slightly decreased in the presence of sevoflurane (right). (d) Effects of sevoflurane on stimulus-induced PSs in area CA1. Left: Example traces of PSs in the control condition (dark grey background), under 2% and 4% sevoflurane (green and red backgrounds, respectively) and after washout (grey background). Right: Plot of absolute PS amplitude showing a concentration-dependent decrease under 2% and 4% sevoflurane. After washout, PSs increased, suggesting good reversibility and synaptic facilitation. All point line charts display the median in addition to the absolute values. The bar charts display mean + standard deviation (SD). Statistical comparison: repeated measures ANOVA and Bonferroni post hoc test. * = p < 0.05, ** = p < 0.01, *** = p < 0.001, n.s = p > 0.05.

Figure 2

Sevoflurane effects on gamma oscillations and changes in CMRO₂. (a) Exemplary recording of gamma oscillation induction with acetylcholine and physostigmine with subsequent treatment with 2% and 4% sevoflurane and washout (top: online spectrogram, bottom: corresponding f.p. trace). (b) Details of recorded network activity and simultaneous changes in p_tiO₂ of (a), corresponding power spectrum analysis, and calculated CMRO₂ for each experimental condition of (a) (control: black; induced gammas: grey; gamma oscillations under 2% isoflurane: green; gammas under 4%: red; sevoflurane washout: grey). (c) Calculated absolute and normalized CMRO₂ and modeled relative ATP consumption rate showing high energy demand during gamma oscillations. CMRO₂ significantly decreased under 2% and 4% sevoflurane. After washout, CMRO₂ was higher than before the induction of gamma oscillations. Concerning gamma oscillations: under 2% sevoflurane, the frequency decreased insignificantly while the power remained unchanged or even slightly increased. Under 4% sevoflurane, gammas were abolished in almost all experiments. After washout effects were reversible. All point line charts display the median in addition to the absolute values. The bar charts display mean +SD. Statistical comparison: repeated measures ANOVA for the CMRO₂, Friedman test for the gamma oscillations (power and frequency). After Bonferroni correction significance given with * = p < 0.05, ** = p < 0.01, *** = p < 0.001, n.s = p > 0.05.

Figure 3

Sevoflurane-induced changes in CMRO₂ and synaptic activity during electrical stimulation. (a) Exemplary recording of simultaneous stimulus-induced (20 Hz, 2 s) p_tiO₂ and extracellular potassium ([K⁺]_o) increases during control conditions (grey) and under sevoflurane treatment (2%: green, 4%: red). Plot of absolute change in stimulus-induced [K_o]⁺ increases and representation of recording settings. (b) Plots of the basal and stimulus-induced changes in CMRO₂ under control (grey-black), 2% sevoflurane (green), and 4% sevoflurane (red) and normalized changes. The normalized CMRO₂ data show no significant changes in the basal oxygen consumption under sevoflurane treatment and a significant decrease in the stimulus-induced oxygen consumption under 4% sevoflurane. (c) Absolute basal and stimulus-induced CMRO₂ changes and relative ATP consumption rate. As the basal CMRO₂ and ATP consumption remained similar, the stimulus-induced changes in metabolism decreased significantly under 4% sevoflurane. All point line charts display the median in addition to the absolute values. The bar charts display mean +SD. Statistical comparison: repeated measures ANOVA for CMRO₂ and Friedman test for the [K⁺]_o. After Bonferroni correction significance given with ** = p < 0.01, *** = p < 0.001, n.s = p > 0.05.

Figure 4

Effects of sevoflurane on flavin adenine dinucleotides (FAD). (a) Simultaneous recordings of stimulus-induced changes in FAD autofluorescence and [K⁺]_o or [Ca²⁺]_o were performed in area CA1 as represented along with example traces of FAD/[K⁺]_o in control (black) and under 4% sevoflurane (red). Note that the decrease in FAD peak and undershoot correlated with decreased [K⁺]_o rises as tissue excitability decreases as well. (b) Left: Averaged stimulus-induced FAD transients in control (black trace) and under 4% sevoflurane (red trace). Stimulus-induced FAD signals typically have two components: a first oxidative peak immediately after stimulation followed by a reductive undershoot. In the presence of 4% sevoflurane, both components diminished as synaptic activity decreased. (c) Ca²⁺ input onto neurons decreased under 4% sevoflurane, generating the boundary conditions for an activity-dependent reduction in oxidative metabolism. Statistical comparison: repeated measures ANOVA. After Bonferroni correction significance given with * = p < 0.05, ** = p < 0.01. (d) Modeling of relative changes in FAD signaling during electrical stimulation without sevoflurane (black) and with 4% isoflurane (red) for the pyruvate dehydrogenase (pdhc), α-ketoglutarate dehydrogenase (kgdhc), succinate dehydrogenase (succdh), and mitochondrial glycerol-3-phosphate dehydrogenase (g3pdh). The solid line and shaded area depict the mean and SD of simulated FAD signals for the individual slices.