Isoflurane activates the type 1 ryanodine receptor to induce anesthesia in mice

Abstract

Inhaled anesthetics were first introduced into clinical use in the 1840s. Molecular and transgenic animal studies indicate that inhaled anesthetics act through several ion channels, including γ-aminobutyric acid type A receptors (GABAARs) and two-pore domain K+ (K2P) channels, but other targets may mediate anesthetic effects. Mutations in the type 1 ryanodine receptor (RyR1), which is a calcium release channel on the endoplasmic reticulum membrane, are relevant to malignant hyperthermia, a condition that can be induced by inhaled anesthetics. However, it was previously uncertain whether inhaled anesthetics directly interact with RyR1. In our study, we demonstrated that isoflurane and other inhaled anesthetics activate wild-type RyR1. By employing systematic mutagenesis, we discovered that altering just one amino acid residue negates the response to isoflurane, thus helping us to pinpoint the potential binding site. Knock-in mice engineered to express a mutant form of RyR1 that is insensitive to isoflurane exhibited resistance to the loss of righting reflex (LORR) when exposed to isoflurane anesthesia. This observation suggests a connection between RyR1 activation and the anesthetic effects in vivo. Moreover, it was shown that RyR1 is involved in the neuronal response to isoflurane. Additionally, administering new RyR1 agonists, which share the same binding site as isoflurane, resulted in a sedation-like state in mice. We propose that isoflurane directly activates RyR1, and this activation is pertinent to its anesthetic/sedative effects.

Fig 1. Inhaled anesthetics activate RyR1.

(A) RyR1 is a calcium release channel located on the membrane of the endoplasmic reticulum (ER). (B) Experimental method for monitoring intracellular calcium ([Ca²⁺]i). (C) Quantification of the intracellular calcium signal. F₀ represents the averaged intensity of baseline fluorescence, whereas F_max represents the peak intensity following the drug administration. (D–F) Caffeine responses for RyR1, RyR2, and RyR3. The time-course reaction (F/F₀) and the peak (F_max/F₀) are plotted for conditions with and without expression. Caffeine was administered after measuring the baseline fluorescence for 30 s, as shown by the arrows. N = 4. (G–I) Response of RyR1, RyR2, and RyR3 to isoflurane. N = 4. (J–L) Response of RyR1, RyR2, and RyR3 to sevoflurane. N = 4. (M–O) Response of RyR1, RyR2, and RyR3 to halothane. N = 4. (P–R) Response of RyR1, RyR2, and RyR3 to chloroform. N = 6. (S) Normalized peak intensity for each anesthetic (1.25 mM). Data are normalized so that the peaks of 2.50 mM caffeine and the control (without pharmacological agents) are 1.0 and 0, respectively, for each isoform (equation 1 in Materials and methods). N = 4–6. (T) RyR1-selective index. I_x represents the normalized peak intensity at 1.25 mM anesthetics for each isoform. N = 4–6. Data are represented as Mean ± SD. For the insets of panels D–R, the data were fitted with logistic functions (equation 4 in Materials and methods). RyR1, the type 1 ryanodine receptor; ER, endoplasmic reticulum; RyR2, the type 2 ryanodine receptor; RyR3, the type 3 ryanodine receptor; Caf., Caffeine; Iso., Isoflurane; Sevo., Sevoflurane; Halo., Halothane.

Fig 2. Identification of RyR1 residues responsible for the sensitivity to isoflurane.

(A) Transient transfection of RyR1, RyR2, and the chimeric receptors (ChR.1-3, ChR.1′-3′) into HEK 293T cells. The time-course reaction (F/F₀) and the peak of the isoflurane response normalized by the caffeine response are shown. N = 4. (B) The RyR1-RyR2 chimeric receptor 10 (ChR.10) and ChR.10′. N = 4. (C) The chimeric receptors, ChR.10, ChR.13, ChR.13-6, and the amino acid substitution mutants. N = 8. (D) Normalized isoflurane responses of M4000 mutants with single amino acid substitutions. The amino acid substitution was color-coded as follows: cyan for basic amino acids; red for acidic amino acids; orange for amino acids with polar uncharged side chains; green for amino acids with alkyl chains; and purple for amino acids with aromatic rings. N = 4–12. (E) Time-course reaction and normalized isoflurane response of RyR1(M4000F). N = 4. (F) Stable cell lines that express RyR1(WT) and RyR1(M4000F). Basal buffer without pharmacological agents, 1.25 mM caffeine, or 0.63 mM anesthetics (isoflurane, sevoflurane, and halothane) were given. N = 4. Data are represented as Mean ± SD. ** Adjusted P-values (Adj. P) < 0.01 by the Steel’s test (multiple comparisons between RyR1 and other groups). For the calculation of the Normalized Iso. Response (equation 1 in Materials and methods), the peaks at 5.0 mM caffeine and the control (without pharmacological agents) were normalized to 1.0 and 0, respectively, in the transient expression experiment (A–E), while the peaks at 1.25 mM caffeine and the control were 1.0 and 0, respectively, in the experiments with stable cell lines (F). ChR., Chimeric receptor; RyR1, the type 1 ryanodine receptor; RyR2, the type 2 ryanodine receptor; Caf., Caffeine; Iso., Isoflurane; Sevo., Sevoflurane; Halo., Halothane.

Fig 3. Putative isoflurane-binding site on RyR1.

(A, B) Putative binding formation with 4-chloro-m-cresol (4-CmC) (A) and isoflurane (B). (C) Time-course reaction to 4-CmC, isoflurane, and caffeine of each amino acid substitution mutant. N = 4. (D) Normalized peak intensity of responses to 4-CmC and isoflurane for each amino acid substitution mutant. N = 4. The data are normalized so that the peaks of 5.0 mM caffeine and the control (without pharmacological agents) are 1.0 and 0, respectively (equation 1 in Materials and methods). Data are represented as Mean ± SD. 4-CmC, 4-chloro-m-cresol; RyR1, the type 1 ryanodine receptor; WT, wild-type; Caf., Caffeine; Iso., Isoflurane.

Fig 4. Resistance to isoflurane in RyR1(M4003F) KI mice.

(A) Amino acid sequence around M4000 (rabbit RyR1) and the generation of RyR1(M4003F) KI mice. (B) The genomic sequences of the KI site for each genotype. (C) Dose–response curves for isoflurane-induced loss of righting reflex (LORR). Data were fitted with logistic functions (equation 5 in Materials and methods). N = 11 (WT), 14 (heterozygous [Hetero] KI), and 13 (homozygous [Homo] KI). The shaded regions represent 95% CIs. (D) Comparison of the concentrations at which the mice reached LORR (Mean ± SD with individual data points). N = 11 (WT), 14 (Hetero KI), and 13 (Homo KI). The adjusted P-values (Adj. P) by the Dunnett’s test are shown. (E) Representative transition of EEG signals, the spectrograms, and EMG signals during isoflurane treatment. At the baseline, mice were kept awake by gentle handling. (F) Normalized delta power (0.4–4 Hz) on EEG during isoflurane treatment for each animal. The mean values and individual data are shown. N = 8 (WT), 10 (Hetero KI), and 9 (Homo KI). (G) Changes in delta power over the baseline (0.0% isoflurane). Mean ± SD is shown. N = 8 (WT), 10 (Hetero KI), and 9 (Homo KI). Data were fitted using logistic functions (equation 6 in Materials and methods). RyR1, the type 1 ryanodine receptor; ssODN, single-stranded oligodeoxynucleotide; WT, wild-type; Hetero, heterozygous; Homo, homozygous; PAM, protospacer-adjacent motif; Iso., Isoflurane; LORR, loss of righting reflex; EEG, electroencephalogram; EMG, electromyogram; The mice illustration was modified from Openclipart (https://openclipart.org/).

Fig 5. Inhibition of neuronal RyR1 alters the responses to isoflurane.

(A) AAV-mediated expression in the brain. (B) Dose–response curves for loss of righting reflex (LORR) of mice expressing mCherry or RyR1-BsSV. Data were fitted with logistic functions (equation 5 in Materials and methods). N = 18 for each group. The shaded regions represent 95% CIs. (C) Comparison of the concentrations at which mice reach LORR (Mean ± SD with individual data points). N = 18 for each group. The P-value by the Student t test is shown. (D) Cultures of cortical neurons on HD-MEA. (E, F) Representative raster plots and the total spike counts in baseline (E) and the isoflurane treatment (2.1% isoflurane) (F). (G) Raster plots across the sequential exposure to isoflurane and the recovery for mCherry and RyR1-BsSV expressions. The above black bars represent the detected prolonged burst periods. Four independent wells are indicated, respectively. The extracellular calcium concentration was adjusted to 2.0 mM before the experiment. (H) Duration of the prolonged burst periods. Data are presented as Mean ± SD with individual data points. N = 4 for both mCherry (gray) and RyR1-BsSV (red) expressions. *P < 0.05 by the two-sample Wilcoxon test. Camk2a, Ca²⁺/calmodulin-dependent protein kinase II α; AAV, adeno-associated virus; RyR1-BsSV, brain-specific splicing variant of the type 1 ryanodine receptor; Iso., Isoflurane; LORR, loss of righting reflex; HD-MEA, high-density multielectrode array; Recov., Recovery. The mice illustration was modified from Openclipart (https://openclipart.org/).

Fig 6. Identification of RyR1 agonists that share the binding site with isoflurane.

(A) Scheme of the chemical screening. Following in silico screening, the agonistic actions of some of the hit compounds were tested with tetracycline-induced expressions of RyR1(WT) and RyR1(M4000F). (B) Results of the in vitro screening. As controls, stimulations by caffeine, isoflurane, and basal buffer without chemicals were included. 4-bromophenol (4-BP) is identified as a hit compound. (C) Putative binding formation of 4-BP. (D) Dose-dependent effects of 4-BP and the isomers 3-BP and 2-BP on RyR1(WT), RyR1(M4000F), RyR2, and RyR3. Data were fitted using logistic functions (equation 4 in Materials and methods). N = 4. (E, F) Effects of methyl phenols (MPs) (E) and ethyl phenols (EPs) (F) on RyR1. N = 5. (G) Effects of a series of 4-alkylphenols on RyR1, RyR2, and RyR3. N = 5. (H) Effects of 4-MP, 4-EP, and 4-PP on RyR1(WT) and RyR1(M4000F). The time-course reactions and the peak intensities are shown, respectively. N = 4. Data are presented as Mean ± SD. RyR1 and the M4000F mutant were expressed through tetracycline induction in the stable cell lines except for the data shown in panel H. RyR1, the type 1 ryanodine receptor; F_max/F₀, the maximum fluorescence intensity; 4-BP, 4-bromophenol; RyR2, the type 2 ryanodine receptor; RyR3, the type 3 ryanodine receptor; 4-BP, 4-bromophenol; 3-BP, 3-bromophenol; 2-BP, 2-bromophenol; 4-MP, 4-methylphenol; 3-MP, 3-methylphenol; 2-MP, 2-methylphenol; 4-EP, 4-ethylphenol; 3-EP, 3-ethylphenol; 2-EP, 2-ethylphenol; 4-PP, 4-propylphenol; Caf., Caffeine.

Fig 7. Isoflurane-mimicking RyR1 agonists induce a sedation-like state.

(A, B) Representative transitions of EEG and EMG signals (A) and EEG power spectra (B) following the injection of 3-BP or saline as the control. N = 3 for both groups. In panel B, the line represents the mean value, and the shaded regions indicate the SEM. (C, D) Representative transitions of EEG and EMG signals (C) and EEG power spectra (Mean ± SEM) (D), following the injection of 4-MP or saline as the control. N = 3 for both groups. In panel D, the line represents the mean value, and the shaded regions show the SEM. (E) Dose–response curves of the loss of righting reflex (LORR) by isoflurane after saline or 3-BP. N = 9 for both groups. The shaded regions show 95% CIs. (F) Comparison of the concentrations at which mice reach LORR (Mean ± SD with individual data points). N = 9 for both groups. (G) LORR dose–response curves with isoflurane after saline or 4-MP injection. N = 9 for both groups. The shaded regions represent 95% CIs. (H) Comparing the concentrations at which mice reached LORR (Mean ± SD with individual data points). N = 9 for both groups. The data in panels E and G were fitted with logistic functions (equation 5 in Materials and methods). The P-value by the Student t test is shown in panels F and H. 3-BP, 3-bromophenol; 4-MP, 4-methylphenol; EEG, electroencephalogram; EMG, electromyogram; Iso., Isoflurane; LORR, loss of righting reflex.