Abstract
Volatile anesthetics reduce excitatory synaptic transmission by both presynaptic and postsynaptic mechanisms which include inhibition of depolarization-evoked increases in presynaptic Ca2+ concentration and blockade of postsynaptic excitatory glutamate receptors. The presynaptic sites of action leading to reduced electrically evoked increases in presynaptic Ca2+ concentration and Ca2+-dependent exocytosis are unknown. Endoplasmic reticulum (ER) of Ca2+ release via ryanodine receptor 1 (RyR1) and uptake by SERCA are essential for regulation intracellular Ca2+ and are potential targets for anesthetic action. Mutations in sarcoplasmic reticulum (SR) release channels mediate volatile anesthetic-induced malignant hyperthermia (MH), a potentially fatal pharmacogenetic condition characterized by unregulated Ca2+ release and muscle hypermetabolism. However, the impact of MH mutations on neuronal function are unknown. We used primary cultures of postnatal hippocampal neurons to analyze volatile anesthetic-induced changes in ER Ca2+ dynamics using a genetically encoded ER-targeted fluorescent Ca2+ sensor in both rat and mouse wild-type (WT) neurons and in mouse mutant neurons harboring the RYR1 T4826I MH-susceptibility mutation. The volatile anesthetic isoflurane reduced both baseline and electrical stimulation-evoked increases in ER Ca2+ concentration in neurons independent of its depression of presynaptic cytoplasmic Ca2+ concentrations. Isoflurane and sevoflurane, but not propofol, depressed depolarization-evoked increases in ER Ca2+ concentration significantly more in mouse RYR1 T4826I mutant neurons than in wild-type neurons. The RYR1 T4826I mutant neurons also showed markedly greater isoflurane-induced reductions in presynaptic cytosolic Ca2+ concentration and synaptic vesicle (SV) exocytosis. These findings implicate RyR1 as a molecular target for the effects of isoflurane on presynaptic Ca2+ handling.
- anesthesia
- calcium
- endoplasmic reticulum
- exocytosis
- isoflurane
- malignant hyperthermia
- presynaptic
- propofol
- sevoflurane
- synaptic vesicle
Significance Statement
Despite their essential clinical roles, the molecular and cellular mechanisms of action of general anesthetics are not fully understood. Malignant hyperthermia (MH) is a potentially fatal pharmacogenetic disorder that leads to dysregulation of intracellular Ca2+ handling in response to triggering by volatile anesthetics. While research on malignant hyperthermia has focused on skeletal muscle effects, much less is known about its neuronal effects. We identify neuronal endoplasmic reticulum (ER) Ca2+ regulation as a novel target for volatile anesthetic action and as a potential target in malignant hyperthermia. While depression of CNS electrical activity in vivo by anesthesia has been observed in another model of malignant hyperthermia, our study reveals fundamental presynaptic mechanisms of volatile anesthetics with implications for the development of more selective anesthetics and for prevention and treatment of malignant hyperthermia.
Introduction
Although volatile anesthetics are essential to modern medicine, a detailed understanding of their cellular and molecular mechanisms of action is incomplete despite 175 years of clinical use. In addition to their major effect of producing an unconsciousness that allows for painful procedures, volatile anesthetics also produce serious cardiovascular and respiratory side effects, and can trigger the rare but potentially fatal pharmacogenetic reaction malignant hyperthermia (MH). Malignant hyperthermia is a hypermetabolic syndrome characterized by excessive intracellular Ca2+ release from the sarcoplasmic reticulum (SR) in skeletal muscle leading to hyperthermia, tachycardia, and muscle rigidity (Halliday, 2003; H. Rosenberg et al., 2015). While the mechanisms underlying the skeletal muscle manifestations of MH are understood in reasonable detail, there are few reports of the effects of MH mutations on neuronal function. In the absence of hyperthermia and muscle rigidity, depression of CNS electrical activity has been reported in R163C-RYR1 MH-susceptible mice after exposure to the volatile anesthetic halothane (Aleman et al., 2020). However, this has not been investigated at the cellular level, in other MH models including the RYR1 T4826I mutation, or with modern anesthetic ethers such as isoflurane.
Volatile anesthetics modulate neurotransmission and communication between neuronal networks (Hemmings et al., 2005, 2019; Franks, 2006), including depression of synaptic transmission through both presynaptic and postsynaptic mechanisms (Hemmings et al., 2005). Volatile anesthetics inhibit activity-dependent Ca2+ influx into presynaptic terminals and Ca2+-dependent synaptic vesicle (SV) exocytosis by reducing neuronal excitability and Ca2+ entry. However, potential sites of action upstream of reduced Ca2+ entry and SV exocytosis are not fully understood.
Presynaptic endoplasmic reticulum (ER) Ca2+ concentration modulates Ca2+ entry involving ER Ca2+ sensing proteins. For example, reduced ER Ca2+ concentration is linked to reduced presynaptic Ca2+ influx (de Juan-Sanz et al., 2017). Ryanodine receptors (RyRs), the principal ER Ca2+ efflux channels, are essential for ER Ca2+ regulation and provide plausible targets for anesthetic action. For example, mutations in RyR1 increase Ca2+ efflux from skeletal muscle sarcoplasmic reticulum (SR) in response to MH-triggering agents including volatile anesthetics to initiate the pathologic features of MH including muscle rigidity and hypermetabolism. We used mice with the well characterized T4826I-RYR1 MH-susceptibility mutation (Yuen et al., 2012) to investigate the role of ER Ca2+ regulation in the presynaptic mechanisms of volatile anesthetic action and MH pathogenesis.
Materials and Methods
Animals
All animal procedures were performed in accordance with the Weill Cornell Medical College Institutional Animal Care and Use Committee regulations and conform to National Institutes of Health (Bethesda, MD) Guidelines for the Care and Use of Animals as well as ARRIVE guidelines, where appropriate. We used both wild-type (WT) Sprague Dawley rats (Charles River Strain code 400; Charles River Laboratories) and wild-type BALB/c mice (Charles River Strain code 028; Charles River Laboratories). Mice homozygous for the MH mutation T4826I-RYR1 in a BALB/c background were purchased from the University of California, Davis (Davis, CA; stock #042036-UCD) and bred for use. Homozygous T4826I-RYR1 mice were used, since 100% of homozygous T4826I-RYR1 mice develop fulminant MH. In contrast, only 17% of male heterozygous T4826I-RYR1 mice develop fulminant MH (Yuen et al., 2012).
Primary neuron culture
Bilateral hippocampi were dissected from postnatal rats or mice (P0–P1; 0–1 d old, both sexes) and plated on poly-ornithine-coated coverslips. Neurons were maintained in culture medium containing of MEM (Thermofisher Scientific, S1200038), 30 mm glucose, 0.1 g/l bovine transferrin (Millipore, 616420), 0.25 g/l insulin, 0.3 g/l glutamine, 5–10% fetal bovine serum (Atlanta Biologicals, S11510), 2% B-27 (Thermofisher Scientific, 17504-044). Cultures were incubated at 37°C with 95% air/5% CO2 in a humidified incubator before imaging. Transfection was performed on day 6 or 7 in vitro (DIV6, DIV7) using Ca2+ phosphate-mediated gene transfer to transfect a low percentage of cells. Live-cell imaging was performed on DIV14–DIV19. For each experiment, neurons were derived from at least three separate culture preparations to minimize artifacts from small variations in culture conditions. The N in each figure corresponds to the number of cells recorded per treatment group.
Plasmids
ER-GCaMP6-150 (Addgene, plasmid #86918), VAMP-mCherry, and synaptophysin-GCaMP6f (syn-GCaMP6) were gifts from Timothy Ryan (Weill Cornell Medicine). Synaptophysin-pHluorin (syn-pH) was a gift from Stephen Heinemann and Yongling Zhu (Salk Institute; pcDNA3-SypHluorin 2×, Addgene plasmid #37004).
Live-cell imaging
Experiments were performed using a Zeiss Axio Observer Z1 widefield fluorescence microscope with filter cubes and LEDs for eGFP and RFP illumination (Zeiss) and an Andor iXon1 EMCCD camera sampling at 10 Hz. Coverslips with attached cells were mounted in a custom closed-bath field stimulation perfusion chamber (total volume 263 μl), with solutions and chamber maintained at 37.0 ± 0.2°C by an in-line solution heater and imaging chamber heater (Warner Instruments). Perfusion was at 1 ml/min using a custom system with a multibarrel closed syringe manifold (Warner Instruments). Only one imaging experiment was acquired per coverslip.
Standard perfusion buffer for syn-GCaMP6 and ER-GCaMP6-160 imaging was Tyrode’s solution (119 mM NaCl, 2.5 mM KCl, 1.2 mM CaCl2, 2.8 mM MgCl2, 25 mM HEPES buffered to pH 7.4, 30 mM glucose). The standard buffer for syn-pH imaging was Tyrode’s solution with 2 mM CaCl2 and 2 mM MgCl2. In experiments where external Ca2+ was decreased to study effects of reduced Ca2+, the standard buffer for imaging was Tyrode’s solution with 1 mM CaCl2 and 3 mM MgCl2. All perfusate solutions contained 10 μm 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and 50 μm D,L-2-amino-5-phosphonovaleric acid (AP5; Tocris) to block recurrent excitation because of glutamatergic excitation. Neurons expressing syn-pH and syn-GCaMP6f were identified by their resting green fluorescence. ER-GCaMP6-150-expressing neurons were co-transfected with the presynaptic marker VAMP-mCherry to identify presynaptic boutons for selective quantification of presynaptic changes in fluorescence.
Cells transfected with syn-GCaMP6 and ER-GCaMP6-150 were electrically stimulated at 20 Hz for 1 s to mimic action potential (AP) trains of 20 AP. Cells transfected with syn-pH were stimulated with AP trains of 100 AP at 10 Hz for 10 s. Electrical stimulation was generated by field stimulation with a pulse generator (Master-9, A.M.P.I.), stimulus isolator (Model A385, World Precision Instruments) and platinum/iridium bath electrodes built into the imaging chamber to produce an electrical field of 10 V cm−1.
Anesthetic solutions
Volatile anesthetic solutions were prepared daily from saturated stock solutions in Tyrode’s buffer. A 12 mM saturated stock isoflurane solution was diluted into an experimental solution equivalent to the clinically relevant dose of ∼1 minimum alveolar concentration (MAC; 0.32 mM; Taheri et al., 1991). A 6 mM saturated stock sevoflurane solution was diluted into an experimental solution equivalent to ∼1 MAC (0.48 mM). Both anesthetic solutions were perfused using gas-tight glass syringes and tubing into the imaging chamber for 5 min before imaging to allow equilibration. Perfusate samples were taken from the chamber for determination of delivered anesthetic concentrations by gas chromatography (Shimadzu GC-2010 Plus) with external standard calibration (Ratnakumari and Hemmings, 1998). The reported mean values of 0.32 mM (range 0.21–0.56 mM) isoflurane and 0.46 mM (range 0.35–0.51 mM) sevoflurane reflect mean measurements from the bath samples collected. Propofol was diluted from a 50 mM stock solution in dimethylsulfoxide (DMSO) into Tyrode’s buffer to a final concentration of 1 μm (0.002% DMSO).
Data analysis and statistics
Live-cell imaging recordings were analyzed using ImageJ (https://imagej.nih.gov/ij/) with the TimeSeries Analyzer plugin (rsb.info.nih.gov/ij/plugins/time-series.html) to measure fluorescence over time using the Background Correction plugin (https://imagej.nih.gov/ij/plugins/download/Background_Correction_.java) to compensate for between experiment variations in background fluorescence. Transfected boutons (∼20–50 boutons per neuron) were selected from control images before drug application and analyzed using 2-μm diameter regions of interest (ROIs). ROIs were selected based on their response to control stimulation in Tyrode’s buffer. Background corrected fluorescence changes were normalized to baseline fluorescence as ΔF/F0 = (F – F0)/F0. Baseline fluorescence (F0) was defined as the mean of 10 frames before stimulus onset, and peak fluorescence (F) was defined as the mean of the five consecutive frames with the highest values immediately following electrical stimulation. Boutons with a signal-to-noise ratio more than or equal to four were used in the analysis.
Statistical significance was tested by paired or unpaired Student’s t tests and by paired one-way or two-way ANOVA with Tukey’s post hoc test, with p < 0.05 considered statistically significant (Table 1). Datasets were assayed for normality with the Shapiro–Wilk test. Sample size was determined by a power analysis of 0.8 with a 5% error, yielding an effect size of n = 6. Statistical analysis and graph preparation used GraphPad Prism v9.3 (GraphPad) and Adobe Illustrator.
Statistical table
Results
Isoflurane reduces presynaptic Ca2+ concentration and synaptic vesicle exocytosis in wild-type and malignant hyperthermia mutant hippocampal neurons
To examine the neuronal consequences of a known human malignant hyperthermia (MH) mutation, we used homozygous knock-in mice with the homologous T4826I-RYR1 mutation (Yuen et al., 2012). This mutation increases Ca2+ efflux through ryanodine receptor 1 (RyR1), an endoplasmic reticulum (ER) Ca2+ efflux channel. Similar to human carriers, mice with the T4826I-RYR1 mutation are phenotypically normal unless challenged with an MH-triggering agent, which can produce skeletal muscle SR Ca2+ release and fulminant MH (Yuen et al., 2012). We examined the effects of this MH-susceptibility mutation on presynaptic function in the presence of isoflurane, by measuring presynaptic Ca2+ concentrations and SV exocytosis (Yuen et al., 2012).
Primary postnatal (DIV6–DIV7) mouse hippocampal neuron cultures were transfected with either syn-GCaMP6f or syn-pH for live-cell imaging. First, we measured presynaptic cytosolic Ca2+ (syn-GCaMP6f) and synaptic vesicle exocytosis (syn-pH) in both wild-type and T4826I-RYR1 mutant mouse neurons (Fig. 1). In control solutions, traces remain stable over two stimulations for both sensors in both wild-type and T4826I-RYR1 mutant mouse neurons (Fig. 1b–g).
The T4826I-RYR1 malignant hyperthermia mutation does not affect presynaptic cytosolic Ca2+ influx or synaptic vesicle exocytosis compared with wild-type mouse hippocampal neurons. a, Schematic diagram of the imaging protocol. Neurons were stimulated electrically with 20 action potentials (APs) at 20 Hz (for Ca2+ measurements using syn-GCaMP6f) or 100 APs at 10 Hz [for synaptic vesicle (SV) exocytosis measurements using syn-pH]. Representative average traces of (b) syn-GCaMP6f and (c) syn-pH responses to electrical stimulation in a wild-type (WT) mouse neuron in control (black) and sham (gray) conditions (50 boutons, DIV 16). Representative average traces of (d) syn-GCaMP6f and (e) syn-pH responses to electrical stimulation in a T4826I-RYR1 malignant hyperthermia susceptible mouse (MH) neuron in control (black) and sham (gray) conditions (50 boutons, DIV 16). Effects of time control stimulation on peak (f) syn-GCaMP6f and (g) syn-pH measurements in T4826I-RYR1 compared with wild-type mouse neurons normalized to their respective controls (d: p = 0.5305, unpaired t test, n = 7 WT, 8 MH; e: p = 0.8734, unpaired t test, n = 7 WT, 6 MH).
Figure 2a shows the protocol used to image neurons transfected with syn-GCaMP6f or syn-pH and treated with a clinically relevant concentration of isoflurane. In wild-type mouse neurons, isoflurane depressed stimulation-evoked increases in presynaptic cytosolic Ca2+ and synaptic vesicle exocytosis (Fig. 1b,c) to a degree comparable to that described in rat neurons (Baumgart et al., 2015). The T4826I-RYR1 mutation markedly enhanced isoflurane inhibition of both depolarization evoked increases in presynaptic cytosolic Ca2+ concentration (Fig. 2b) and SV exocytosis (Fig. 2c) compared with wild-type neurons, with greater inhibition by isoflurane of presynaptic cytosolic Ca2+ (p = 0.0015; Fig. 2d) and exocytosis (p = 0.0148; Fig. 2e) in mutant neurons. Wild-type mouse neurons exhibited 35% depression of stimulation-evoked increases in presynaptic cytosolic Ca2+ concentration and 41% depression of SV exocytosis (Fig. 2d,e), while T4826I-RYR1 neurons treated with isoflurane exhibited 75% depression of presynaptic cytosolic Ca2+ concentration (WT isoflurane vs MH isoflurane: p = 0.0013; Fig. 2d) and 63% depression of SV exocytosis (WT isoflurane vs MH isoflurane: p = 0.0250; Fig. 2e). These results suggest that the T4826I-RYR1 MH mutation enhance the actions of isoflurane and contributes to neuronal dysfunction in MH-susceptible neurons in response to isoflurane through alterations in presynaptic ER Ca2+ regulation. These results also indicate that RyR1 makes a functionally important contribution to synaptic function in hippocampal neurons.
Isoflurane effects on presynaptic cytosolic Ca2+ influx and synaptic vesicle exocytosis are reduced in T4826I-RYR1 compared with wild-type mouse hippocampal neurons. a, Schematic diagram of the imaging protocol. Neurons were stimulated electrically with 20 action potentials (APs) at 20 Hz (for Ca2+ measurements using syn-GCaMP6f) or 100 APs at 10 Hz [for synaptic vesicle (SV) exocytosis measurements using syn-pH]. Representative average traces of (b) syn-GCaMP6f and (c) syn-pH responses to electrical stimulation in a T4826I-RYR1 neuron in control (black) and isoflurane (blue) conditions (50 boutons, DIV 16). Effects of isoflurane on peak (d) syn-GCaMP6f and (e) syn-pH measurements in T4826I-RYR1 malignant hyperthermia susceptible (MH) compared with wild-type (WT) mouse neurons normalized to their respective controls (d: p = 0.0013, unpaired t test, n = 6 WT, 10 MH; e: p = 0.0250, unpaired t test, n = 7 WT, 6 MH).
Isoflurane reduces endoplasmic reticulum Ca2+ concentration independent of inhibition of Ca2+ influx
These findings led us to investigate the role of ER Ca2+ as a target for the presynaptic effects of isoflurane, using rat neuron cultures to optimize the protocols. Using ER-GCaMP6-150, a genetically encoded, ER-targeted Ca2+ sensor, we found that isoflurane significantly depressed baseline ER Ca2+ concentration in wild-type rat hippocampal neurons by 5% (p = 0.0098; Fig. 3). To restrict analysis to presynaptic boutons, we used co-transfection with the presynaptic marker VAMP-mCherry to analyze boutons co-expressing ER-GCaMP6-150 and VAMP-mCherry (Fig. 4a–c). Using the protocol shown in Figure 4d, isoflurane reversibly inhibited stimulation-evoked increases in ER Ca2+ concentration by ∼57% (p = 0.0091; Fig. 4e,f).
Isoflurane reduces resting endoplasmic reticulum Ca2+ concentration in rat hippocampal neurons. Baseline ER-GCaMP6-150 fluorescence was measured in wild-type rat neurons treated with isoflurane compared with a separate time control neuron [p = 0.0098, unpaired t test, n = 9 (left), n = 7 (right)].
Isoflurane reduces stimulation-evoked increases in endoplasmic reticulum Ca2+ concentration. a–c, Fluorescence images of a rat hippocampal neuron co-transfected with ER-GCaMP6-150 (green) and VAMP mCherry (red). Live-cell imaging measuring [Ca2+]ER (a) before and (b) during electrical stimulation. c, Snapshot of the presynaptic marker VAMP-mCherry. d, Schematic diagram of the protocol used in b, c. e, Representative average traces of ER-GCaMP6-150 fluorescence changes with 20 action potentials (APs) at 20 Hz stimulations for control, isoflurane, and washout conditions (n = 1, 50 boutons, DIV 16). f, Peaks of ER-GCaMP6-150 fluorescence over three stimulations of 20 AP each at 20 Hz for control (white circle) and isoflurane (blue circle) conditions. Control versus isoflurane, p = 0.0091; isoflurane versus washout, p = 0.0047; control versus washout p = 0.9967; one-way ANOVA, n = 8.
Since isoflurane reduced both baseline and stimulation-evoked increases in ER Ca2+ concentration, we reanalyzed data from Figure 4b,c using arbitrary fluorescence units (AFUs) rather than normalized F to mitigate a potential effect of the isoflurane-induced change in baseline fluorescence on ΔF/F0. The degree of inhibition of unnormalized F was comparable to that of normalized ΔF/F0 (p = 0.0002; Fig. 5b,c). There was no significant difference between results obtained using ΔF/F0 values normalized to control values or raw AFU values normalized to control values in the same set of isoflurane-treated neurons (Fig. 5d). We therefore used ΔF/F0 for subsequent experiments.
Isoflurane depression of baseline ER Ca2+ concentration was not sufficient to disrupt normalization using ΔF/F0. a, Schematic diagram of the protocol used in b, c. b, Representative average traces in arbitrary fluorescence units (AFU) of ER-GCaMP6-150 fluorescence changes with 20 action potentials (APs) at 20 Hz stimulations for control, isoflurane, and washout conditions (n = 1, 50 boutons, DIV 16). c, Arbitrary fluorescence unit (AFU) peaks of ER-GCaMP6-150 fluorescence over three stimulations of 20 AP each at 20 Hz for control (white circle) and isoflurane (blue circle) conditions. (control vs isoflurane p = 0.0002, isoflurane vs washout p = 0.0002, control vs washout p = 0.0425, one-way ANOVA, n = 8). d, Box and whisker plot comparing the same data set measured two ways: ΔF/F0 or total AFU of isoflurane-treated normalized to its respective control condition (p = 0.3457, paired t test, n = 8).
We hypothesized that depression of presynaptic Ca2+ influx by isoflurane triggers a reduction in the amount of cytosolic Ca2+ available for sequestration by the ER, thereby reducing ER Ca2+ uptake and reducing the increase in intraluminal ER Ca2+ concentration in response to electrical stimulation. To test this mechanism, we lowered extracellular Ca2+ from 1.2 mM to 1 mM to reduce the stimulation-evoked increase in presynaptic cytosolic Ca2+ to a similar degree as observed with isoflurane exposure. This reduction in extracellular Ca2+ concentration led to a 40% reduction in presynaptic Ca2+ influx in wild-type rat hippocampal neurons in the absence of isoflurane, which is comparable to the reduction produced by 0.30 mM isoflurane (Fig. 6b). However, compared with the reduction in ER Ca2+ observed with reduced extracellular Ca2+, the degree of isoflurane inhibition of the stimulation-evoked increase in ER Ca2+ concentration was greater (p = 0.0380; Fig. 6c). Thus, the reduction in evoked ER Ca2+ concentration by isoflurane is not fully attributable to its inhibition of presynaptic cytosolic Ca2+ influx, supporting an additional mechanism(s) of action.
Reduction of stimulation-evoked increase in endoplasmic reticulum Ca2+ concentration by isoflurane is not dependent on reduced Ca2+ influx. a, Schematic diagram of the protocol used. Box and whisker plot comparing (b) syn-GCaMP6f or (c) ER-GCaMP6-150-transfected cells stimulated with 20 action potentials (APs) at 20 Hz with 0.30 (±0.11) mM isoflurane normalized to control with 1.2 mM Ca2+ Tyrode’s solution, or control with 1 mM Ca2+ Tyrode’s solution normalized to 1.2 mM Ca2+ Tyrode’s solution [b: p = 0.7963, unpaired t test, n = 8 (left), 5 (right); c: p = 0.0380, unpaired t test, n = 7 (left), 6 (right)].
The T4826I-RYR1 malignant hyperthermia mutation potentiates volatile anesthetic depression of stimulation-evoked increases in endoplasmic reticulum Ca2+
We compared the effects of isoflurane on stimulation-evoked increases in ER Ca2+ in wild-type and T4826I-RYR1 mutant mouse hippocampal neurons to explore the impact of this MH-susceptibility mutation on anesthetic effects. Isoflurane produced a large reduction in the stimulation-evoked increase in ER Ca2+ concentration in T4826I-RYR1 mouse hippocampal neurons compared with wild-type neurons (p = 0.0002; Fig. 7b,e). Sevoflurane, another volatile anesthetic trigger of MH, had a similar effect (p = 0.0295; Fig. 7c,e). Both volatile anesthetics reduced stimulation evoked increases in ER Ca2+ concentration to a greater extent in T4826I-RYR1 neurons (MH isoflurane vs MH control 80% inhibition: p < 0.0001; MH sevoflurane vs MH control 63% inhibition: p < 0.0001) compared with in wild-type neurons (WT isoflurane vs WT control 31% inhibition: p = 0.0009; WT sevoflurane vs WT control 23% inhibition: p = 0.0367). In contrast, propofol, an intravenous anesthetic that is not a trigger of MH, reduced evoked ER Ca2+ concentration in wild-type neurons but did not potentiate the reduction in T4826I-RYR1 mutant neurons (Fig. 7d,e; not significant; WT propofol vs WT control 27% inhibition: p = 0.0052; MH propofol vs MH control 39% inhibition: p = 0.0011). The effects of isoflurane or sevoflurane on evoked increases in ER Ca2+ in T4826I-RYR1 neurons were significant compared with their effects on ER Ca2+ in wild-type neurons (Fig. 7e), while the effect of propofol on ER Ca2+ was not different between T4826I-RYR1 and wild-type neurons (p = 0.9258;Fig. 7e). These results suggest that this MH-susceptibility mutation has marked effects on presynaptic Ca2+ handing in the presence of triggering volatile anesthetics, but not nontriggering intravenous anesthetics.
Volatile anesthetics potentiate stimulation-evoked reduction of endoplasmic reticulum Ca2+ concentration in T4826I-RYR1 mouse neurons. a, Schematic diagram of the protocol. Neurons were stimulated electrically with 20 action potentials (APs) at 20 Hz. Representative average traces of T4826I-RYR1 malignant hyperthermia (MH) mouse neurons transfected with ER-GCaMP6-150 perfused with (b) 0.34 mM isoflurane, (c) 0.46 mM sevoflurane, or (d) 1 μm propofol. Peak ER-GCaMP6-150 effect of isoflurane, sevoflurane, or propofol on T4826I-RYR1 compared with wild-type (WT) mouse neurons normalized to their respective controls (e: p = WT isoflurane vs MH isoflurane: 0.0002, WT sevoflurane vs MH sevoflurane: 0.0295, WT propofol vs MH propofol: 0.9258, two-way ANOVA, n = 8, 7, 4, 6, 7, 8, respectively).
Viability of T4826I-RYR1 mouse neurons
Given the finding that the T4826I-RYR1 MH-susceptibility mutation greatly enhances isoflurane depression of electrical stimulation-evoked increases in presynaptic Ca2+, SV exocytosis, and ER Ca2+ concentration, it was important to show that T4826I-RYR1 neurons remain viable after isoflurane treatment, i.e., that the effects are not because of enhanced neurotoxicity. Most neurons exhibited a return in responsiveness to stimulation after washout of isoflurane (Fig. 8). There was a complete return of responsiveness for SV exocytosis (syn-pH) and presynaptic Ca2+ concentration (syn-GCaMP6) in T4826I-RYR1 neurons (Fig. 8b,c). However, there was only a partial return of responsiveness in ER Ca2+ (ER-GCaMP6-150; Fig. 8d). It is likely that the 5-min washout period was insufficient for recovery of ER Ca2+ in T4826I-RYR1 neurons. Together these findings show that isoflurane depression of stimulation-evoked increases in ER Ca2+ concentration, presynaptic Ca2+ concentration, and SV exocytosis in T4826I-RYR1 neurons is not the result of irreversible cell death. Moreover, the mutant neurons were not more sensitive to electrical stimulation than wild-type neurons, and the mutant neurons remained stable and responsive over multiple electrical stimulations in the absence of isoflurane (Fig. 9).
T4826I-RYR1 malignant hyperthermia susceptible mouse neurons remain stable and responsive over repeated stimulations. a, Schematic diagram of the protocol. Neurons were stimulated electrically with 20 action potentials (APs) at 20 Hz. Representative average traces of T4826I-RYR1 mouse neurons perfused with control solution transfected with (b) syn-pH, (c) syn-GCaMP6f, or (d) ER-GCaMP6-150.
Enhanced isoflurane-induced inhibition of stimulation-evoked presynaptic synaptic vesicle exocytosis, cytosolic Ca2+ concentration, and ER Ca2+ concentration in T4826I-RYR1 malignant hyperthermia susceptible mouse neurons is reversible. a, Schematic diagram of the protocol. Neurons were stimulated electrically with 20 action potentials (APs) at 20 Hz. Representative average traces of T4826I-RYR1 mouse neurons perfused with isoflurane transfected with (b) syn-pH, (c) syn-GCaMP6f, or (d) ER-GCaMP6-150.
Discussion
General anesthetics induce a complex drug-induced coma that has intrigued neuropharmacologists since its initial demonstration in 1846 because of its reversible effects on memory, consciousness, and movement in response to pain (Hemmings et al., 2019). General anesthetics have marked effects on synaptic transmission, but our understanding of the mechanisms involved in their presynaptic effects remain incomplete. Moreover, the role anesthetic effects on presynaptic neuronal intracellular Ca2+ store regulation has not been investigated previously. Here, we show that volatile anesthetics, but not propofol, disrupt neuronal ER Ca2+ handling, and that these effects are enhanced in a mouse model of malignant hyperthermia susceptibility. We identified inhibitory effects of isoflurane on stimulation-evoked increases in neuronal ER Ca2+ concentration in wild-type neurons, demonstrating presynaptic ER Ca2+ handing as a neuronal target for anesthetic effects. Moreover, we showed that the inhibitory effects of isoflurane on stimulation-evoked increases in presynaptic ER Ca2+, cytosolic Ca2+, and SV exocytosis are enhanced in the well characterized T4826I-RYR1 mouse model of human malignant hyperthermia.
Most previous studies of anesthetic effects on the ER have focused on its role in modulating cell viability (Q.J. Wang et al., 2008; Zhai et al., 2015; Liu et al., 2016). Our studies are the first to analyze the functional impact of anesthetic effects on ER Ca2+ on neuronal function in intact hippocampal neurons. Using fluorescent biosensors to measure ER Ca2+ in live neurons, we found that isoflurane reduced both resting baseline and electrical activity-evoked increases in presynaptic ER Ca2+ concentration. The mechanisms of isoflurane effects on ER Ca2+ regulation are better understood in other cell types, particularly in skeletal muscle and cardiomyocytes (Davies et al., 2000; Pabelick et al., 2004; Dworschak et al., 2006; Klingler et al., 2014), but have not been characterized in neurons. Anesthetic depression of stimulation-evoked ER Ca2+ could be attributed to reductions in presynaptic cytosolic Ca2+ influx, but by reducing external Ca2+ concentration to mimic the isoflurane-induced reduction in presynaptic cytosolic Ca2+ we were able to distinguish between the known inhibition of presynaptic cytosolic Ca2+ by isoflurane and a distinct effect to reduce ER Ca2+ concentration. These data indicate a distinct mechanism underlying modulation of ER Ca2+ by volatile anesthetics mediated by RyR1.
Possible mechanisms for the presynaptic effects of isoflurane on ER Ca2+ regulation include a direct effect on ER Ca2+ regulation that leads to depression of presynaptic cytosolic Ca2+ and SV exocytosis, or separate effects on ER Ca2+ and presynaptic cytosolic Ca2+ concentration regulation that combine to inhibit SV exocytosis. The first possibility suggests that isoflurane depression of increases in ER Ca2+ is upstream of its depression of presynaptic cytosolic Ca2+ influx. Pharmacological inhibition of Ca2+ uptake into the ER by the SERCA pump reduces stimulation-evoked increases in both presynaptic cytosolic Ca2+ and SV exocytosis involving a temperature-dependent positive feedback loop in which ER Ca2+ content controls presynaptic cytosolic Ca2+ influx and SV exocytosis (de Juan-Sanz et al., 2017). Although the mechanism is unclear, STIM1, an ER Ca2+ sensor, is essential for normal CNS function; a conditional STIM1 knock-out results in marked learning defects and impaired cerebellum-regulated motor activity (Hartmann et al., 2014; Garcia-Alvarez et al., 2015; de Juan-Sanz et al., 2017). Cognitive dysfunction and impaired motor activity are also features of isoflurane anesthesia, so depression of ER Ca2+ through enhanced efflux via RyRs (Glover et al., 2004) might activate the STIM1-mediated positive feedback loop contributing to depression of presynaptic cytosolic Ca2+ and SV vesicle exocytosis, similar to the effect of SERCA inhibition (de Juan-Sanz et al., 2017). The second possibility suggests that the isoflurane effect on ER Ca2+ is secondary to its effect on presynaptic Ca2+ concentration. These mechanisms might each contribute to isoflurane depression of SV exocytosis, ER Ca2+ concentration through the STIM1-mediated positive feedback loop, and presynaptic cytosolic Ca2+ because of reduced Ca2+ entry. Distinguishing between these mechanisms and their individual contributions will be technically challenging as they are inextricably linked, but our results indicate multiple mechanisms.
Isoflurane depression of ER Ca2+ increases involves a distinct mechanism that could also contribute to some of its undesirable neurologic side effects (e.g., neurotoxicity, cognitive dysfunction). Hereditary ER dysfunction has been associated with cognitive dysfunction (Liu et al., 2012); if isoflurane effects on ER Ca2+ contribute to cognitive dysfunction, blocking or reducing its effects on ER Ca2+ might ameliorate this undesirable aspect of anesthesia (Rundshagen, 2014).
The mechanisms by which isoflurane reduces ER Ca2+ concentration could involve reduced uptake and/or enhanced efflux (Glover et al., 2004). Isoflurane could activate ER Ca2+ uptake or efflux pathways, including smooth endoplasmic reticulum Ca2+ ATPase (SERCA), inositol triphosphate receptor (IP3R), or RYR mediated mechanisms (Lytton et al., 1992). It is unlikely SERCA is the major mechanism for the effects we observed, as SERCA is not known to be modulated by isoflurane or directly affected by the T4826I-RYR1 mutation. Because of the known effects of volatile anesthetics on RyR1 in skeletal muscle in MH, direct interaction with RyR1 is likely involved in the neuronal effects of volatile anesthetics on ER Ca2+. A recent pilot study in heterozygous R163C-RYR1 MH-susceptible mice showed enhanced suppression of CNS electrical activity by the volatile anesthetic halothane in vivo consistent with CNS effects of an MH-associated RYR1 mutation (Aleman et al., 2020). Caffeine, an RyR agonist, enhanced halothane suppression of EEG power in R163C-RYR1 mice, suggesting that a volatile anesthetic has functional CNS phenotype in MH-susceptible mice. However, RyR1 might not be the sole mechanism contributing to this phenotype, as volatile anesthetics are promiscuous drugs that suppress EEG power, and it is possible that their effects involve interactions with multiple Ca2+ influx and efflux pathways (Baumgart et al., 2015). Isoflurane might also directly or indirectly alter production or localization of endogenous modulators of presynaptic Ca2+ channels and transporters.
Specific mutations that alter SR Ca2+ handling in skeletal muscle confer MH susceptibility (Ali et al., 2003). We hypothesized that RyRs are involved in neuronal function though presynaptic ER Ca2+ regulation, and that MH-susceptibility mutations also alter the neuronal effects of volatile anesthetics mediated by ER Ca2+ regulatory pathways. Most MH-susceptibility mutations occur in RYR1, which encodes an SR/ER Ca2+ efflux channel (H. Rosenberg et al., 2015). RyR1 is the predominant isoform expressed in skeletal muscle, and appears to be expressed in brain as well (Hakamata et al., 1992; Furuichi et al., 1994), but the effects of volatile anesthetics on neurons expressing MH-susceptibility mutations had not been investigated previously.
RyR isoform localization and function in neurons have not been clearly resolved. All three isoforms (RyR1–RyR3) are expressed in brain, however their roles in neuronal function have not been characterized. RyR expression has been detected in several hippocampal neuron compartments including in presynaptic terminals, dendritic spines, and somata (Nakanishi et al., 1992; Seymour-Laurent and Barish, 1995; Hertle and Yeckel, 2007; Shimizu et al., 2008; Wayman et al., 2012). However, because of poor antibody specificity, subcellular localization of RyR1 and other RyR isoforms in hippocampal neurons is not well undefined (Hiess et al., 2022). Our findings suggest that RyR1 is functionally expressed in hippocampal neurons, a significant advance in understanding the neuronal roles of RyRs. Contradictory results have suggested either the presence or absence of RyR1 in the hippocampus (Mori et al., 2000; Galeotti et al., 2008). We found that an MH-susceptibility mutation in RyR1 led to enhanced anesthetic-induced depression of electrical stimulation-evoked increases in presynaptic ER Ca2+, cytosolic Ca2+, and SV exocytosis. We also provide novel evidence that RyR1 plays a critical role in synaptic transmission of hippocampal neurons.
T4826I-RYR1 mutant mice are a suitable model for studying MH susceptibility: they have no overt phenotype in the absence of volatile anesthetics, as is typical in humans with this and other MH-susceptibility mutations. We used homozygous T4826I-RYR1 mice since 100% of homozygous T4826I-RYR1 mice develop fulminant MH while only 17% of male and no female heterozygous T4826I-RYR1 mice develop fulminant MH in response to halothane as a trigger (Yuen et al., 2012). Such complete penetrance is essential to interpretation of our in vitro findings. Our results show that the T4826I-RYR1 mutation exacerbates isoflurane depression of stimulus-evoked increases in ER Ca2+, as well as of presynaptic cytosolic Ca2+ and SV exocytosis. This effect was specific to the volatile anesthetics sevoflurane and isoflurane, while propofol, a mechanistically distinct intravenous anesthetic, had no significant effect on ER Ca2+ in MH-susceptible neurons compared with wild-type neurons. This is consistent with the MH triggering activity of volatile anesthetics, which can be lethal in triggering MH, while propofol is safe for MH-susceptible patients (M.B. Rosenberg, 1991; Gupta et al., 2021). The marked effects of volatile anesthetics are not the result of cell death or loss of responsivity, as the effects were largely reversible.
Our results imply that an MH episode might involve direct neurologic effects, including depression of synaptic transmission with possible short-term and long-term effects not yet investigated clinically (Fig. 10). Heat-induced central nervous system (CNS) damage because of MH has been reported (Gore and Isaacson, 1949). However, the pattern of brain injury seen in MH differs from the pattern seen in hypoxic-ischemic brain injury (Forrest et al., 2015). This coupled with our results suggests that MH-induced CNS injury is not exclusively because of hyperthermia-induced cytotoxicity, but also involves direct effects on neuronal Ca2+ regulation.
Overview of possible presynaptic volatile anesthetic mechanisms. Schematic diagram of isoflurane depression of presynaptic ER Ca2+ concentration, cytosolic Ca2+ concentration, and synaptic vesicle exocytosis in wild-type (left half) and MH-susceptible (right half) mice (ER: endoplasmic reticulum; MH: malignant hyperthermia; syn-pH: synaptophysin-pHlourin; syn-GCaMP6: synaptophysin-GCaMP6f, SV: synaptic vesicle).
Treatment of MH involves early administration of dantrolene, an RyR antagonist. Although dantrolene is lipid soluble with a molecular mass of 314, properties that generally predict blood-brain barrier permeability, it exhibits limited CNS penetration (J. Wang et al., 2020), and there is conflicting evidence for its to cross the blood-brain barrier (Meyler et al., 1981; Wuis et al., 1989). Future studies should focus on elucidating the impact of MH on neuronal function and cytotoxicity, as well as mitigating possible dysfunction and neurotoxicity in MH, and possibly other neurologic diseases.
Limitations
Use of primary dissociated neurons as an experimental model does not fully recapitulate cellular function and interactions in vivo. Our focus on presynaptic function might have overlooked other relevant mechanisms observed in intact neuronal circuits, including postsynaptic and glial contributions. We focused our studies on the hippocampus as a model given the large body of data available on its fundamental neurophysiology, sensitivity to general anesthetics, and critical role in anesthetic actions (Rudolph and Antkowiak, 2004; D.S. Wang and Orser, 2011; Hemmings et al., 2019). The fundamental cellular mechanisms of Ca2+ regulation are shared between many neuron types, however observations in hippocampal neurons might not translate to other neuronal types because of cell type-specific differences. Use of a single RyR1 mutation model for MH susceptibility from the many known human mutations (Hopkins, 2011) is a potential limitation, however there are limited viable animal models for human MH susceptibility that are as well characterized as the T4826I-RYR1 mutation (Yuen et al., 2012).
Future directions
Understanding the mechanisms of anesthetic effects on ER Ca2+ has implications for the presynaptic mechanisms of volatile anesthetics. Future studies will examine the role of the STIM1 feedback loop in anesthetic mechanisms using knock-down models (de Juan-Sanz et al., 2017). It is also important to understand the effects of dantrolene on anesthetic effects in MH-susceptible neurons. Use of fluorescence imaging of neurons treated with dantrolene is not technically possible because of its intense absorption at critical wavelengths, so alternative methods will be required.
In conclusion, we used optogenetic tools to study volatile anesthetic effects on presynaptic Ca2+ regulation and synaptic vesicle exocytosis in rodent hippocampal neurons. We identified and characterized anesthetic effects on intracellular Ca2+ regulation, specifically on stimulation-evoked changes in ER Ca2+ concentration, in hippocampal neurons from both wild-type and malignant hyperthermia-susceptible mice. We identified reduced baseline and stimulation-evoked increases in neuronal ER Ca2+ concentration as a novel mechanism of volatile anesthetic action. We also revealed a novel neuronal phenotype of MH susceptibility in response to the volatile anesthetics isoflurane and sevoflurane, which trigger MH. Finally, we provide functional evidence for a presynaptic role of RyR1 in synaptic transmission in the hippocampus. Taken together, our findings implicate RyR1 in presynaptic function and pave the way understanding and improving general anesthesia, and in treating those with MH-susceptibility mutations. Our results provide several important advances in elucidating the role of presynaptic ER Ca2+ mechanisms and RyR1 function in neuropharmacology and the actions of volatile anesthetics.
Acknowledgments
Acknowledgments: We thank Timothy Ryan (Weill Cornell Medicine) and members of his laboratory for generously providing plasmids and technical advice.
Footnotes
H.C.H. is the Editor-in-Chief of the British Journal of Anaesthesia. All other authors declare no competing financial interests.
This work was supported by National Institutes of Health Grants GM58055 (to H.C.H.) and GM133115 (to V.O.).
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license, which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.