General Anesthetic Isoflurane Modulates Inositol 1,4,5-Trisphosphate Receptor Calcium Channel Opening

THE inositol 1,4,5-trisphosphate receptor (InsP₃R) is an intracellular Ca²⁺-release channel found mostly on the membrane of the endoplasmic reticulum (ER). Activation of InsP₃R by inositol 1,4,5-trisphosphate (InsP₃) causes Ca²⁺ release from the ER lumen into the cytoplasm, where it acts as a second messenger to regulate many physiological processes such as cell survival and neurogenesis.^1–4  InsP₃R overactivation also regulates some pathological processes,^1,5  especially apoptosis and neurodegeneration.^6–8  InsP₃R channel activity can be regulated by a diverse array of interacting proteins,⁴  including neurodegenerative-associated proteins such as huntingtin-associated protein 1 and Alzheimer’s disease-mutant presenilin 1 and presenilin 2.^9,10 

More than 260 million patients worldwide receive surgeries with general anesthesia each year. The diversity of the molecular mechanisms of general anesthetics is still not clear. Activation of InsP₃Rs may play important roles in anesthetic-mediated regulation of intracellular Ca²⁺ homeostasis and some physiological and pathological processes.^11–14  General anesthetics, especially isoflurane, may precondition cells and provide cytoprotection by moderately activating InsP₃Rs and Ca²⁺ release from the ER.¹²  Isoflurane may also cause apoptosis and neurodegeneration by causing abnormal Ca²⁺ release from the ER via overactivation of InsP₃R.^{3,11,13,15–17}  Sensitized InsP₃R activity in Alzheimer’s and Huntington diseases seems to render neurons vulnerable to isoflurane-induced Ca²⁺ release from the ER and subsequent apoptosis.^15,17,18  Isoflurane increases levels of β-site amyloid β-precursor protein-cleaving enzyme¹⁶  and aggregation of mutated huntingtin proteins¹⁷ via activation of InsP₃R. General anesthetics may cause neuronal apoptosis by disruption of intracellular Ca²⁺ homeostasis.^18–21  Despite these strong associations of anesthetic exposure and InsP₃R activation in normal and pathological conditions, there is a lack of evidence to support direct activation of the InsP₃R by isoflurane.

In this study, we report for the first time that clinically relevant concentrations of isoflurane directly modulate the activity of the InsP₃R channel and sensitize the channel to basal levels of InsP₃, resulting in InsP₃R-mediated Ca²⁺ release from the ER and amplification of InsP₃-induced [Ca²⁺]_i signals and induction of cell apoptosis. These results suggest that the InsP₃R may be one of the molecular mechanisms of anesthetic-mediated pharmacological and toxic effects in neurodegeneration.

DT40 cells lacking the genes for all three isoforms of InsP₃Rs (DT40-KO, RIKEN Cell Bank No. RCB 1467, Ibaraki, Japan) and DT40-KO cells stably overexpressing the rat type 3 InsP₃R (DT40-R3)²²  were used in this study. Although the work was conducted in a chicken lymphocyte cell line, the channel studied is the rat recombinant InsP₃R channel. We elected to use InsP₃R-3 instead of other InsP₃R subtypes because InsP₃R-3 has been shown to have relatively higher opening probability activated by InsP₃R agonists and is a useful in vitro model to examine the InsP₃R activation by its agonists.²²  Given the high level of primary sequence homology among mammalian InsP₃Rs, and the substantial similarities in the major features of the regulation of endogenous and recombinant InsP₃R channels by cytoplasmic InsP₃ and Ca²⁺ (biphasic activation and inhibition by Ca²⁺; and monotonic, saturable activation by InsP₃),⁴  we expect that our findings that isoflurane affects Ca²⁺ and InsP₃ regulation of InsP₃R channels have relevance for similar regulations of neuronal InsP₃R channels, and thereby have relevance for Ca²⁺ homeostasis in neurons. Cells were maintained in suspension culture in Roswell Park Memorial Institute 1640 medium containing 10% fetal calf serum, 1% chicken serum, penicillin (100 units/ml), streptomycin (100 μg/ml), and glutamine (2 mM) at 37°C in a 95% air and 5% CO₂ humidified incubator.

There is no way to directly observe single-channel activity of the InsP₃R in the native ER membrane because performing patch clamp electrophysiology on the ER directly has not been achieved. Because the outer nuclear membrane is continuous with the ER membrane topologically, and because there is no evidence to suggest that the biochemical properties of the outer nuclear membrane are distinct from those of the rest of the ER, nuclear patch clamping of the ER membrane is a tool to record ER-localized ion channel activities, including the InsP₃R.⁴  We have examined the effects of isoflurane on InsP₃Rs located on the outer nuclear membrane. Preparation of isolated nuclei from DT40 cells was performed as described.^10,22–26  Cells were washed twice with phosphate buffered saline and suspended in a nuclear isolation solution containing 150 mM KCl, 250 mM sucrose, 1.5 mM 2-mercaptoethanol, 10 mM Tris·HCl, 0.05 mM phenylmethylsulfonyl fluoride, and protease inhibitor mixture (Complete; Roche Molecular Biochemicals, Indianapolis, IN), adjusted to pH 7.3 with KOH. Isolated nuclei were placed in a recording chamber containing a standard bath solution: 140 mM KCl, 10 mM HEPES, and 0.5 mM 1,2-bis(o-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid, ≈200 nM free Ca²⁺, adjusted to pH 7.3 with KOH. During the nuclear isolation protocol and the introduction of nuclei onto the stage of the patch clamp microscope, any endogenous InsP₃ is washed away. Selected intact nuclei were patch clamped and single-channel activities were recorded in the on-nucleus configuration. Recording pipettes with resistances of 8 to 10 MΩ were used. The pipette solution contained 140 mM KCl, 0.5 mM adenosine triphosphate, and 10 mM HEPES (pH 7.3), with varying concentrations of isoflurane, InsP₃, and Ca²⁺. The free Ca²⁺ concentration was varied by addition of an appropriate Ca²⁺ chelator, as previously described.²⁷  Single-channel currents were amplified using an Axopatch-200B amplifier (Molecular Devices, Downingtown, PA), filtered at 1 kHz, and digitized at 5 kHz with an ITC-16 interface (Instrutech, Port Washington, NY) and Pulse+ Pulse Fit software (HEKA Electronik, Bellmore, NY). All recordings were performed at room temperature with the pipette electrode at −40 mV relative to the reference bath electrode. Single-channel analyses were performed on recordings exhibiting only a single channel using QuB (University of Buffalo, Buffalo, New York) and Igor Pro (WaveMetrics, Lake Oswego, OR) softwares. Figures were generated using Igor Pro software and Adobe Illustrator (Adobe Systems Incorporated, San Jose, CA).

DT40-KO or DT40-R3 cells were seeded onto glass coverslips coated with 0.01% poly-l-ornithine for 1 h before measurements. Cells were loaded with 2.5 μM acetoxymethyl ester form of calcium sensitive dye Fura-2 (Molecular Probes, Grand Island, NY) for 30 min at room temperature in Hanks’ balanced salt solution (Sigma, Pittsburgh, PA) containing 1.8 mM CaCl₂ and 0.8 mM MgCl₂, pH 7.4. Coverslips were then placed in a sealed perfusion chamber (Warner Instruments, Hamden, CT) and continuously perfused at room temperature with Hanks’ balanced salt solution. Fura-2 was alternately excited at 340 and 380 nm, and emitted fluorescence (510 nm) was collected and recorded using a charged-coupled device-based imaging system running IPLab v3.7 software (Biovision Technologies, Exton, PA). The data are presented as the ratio of fluorescence intensities recorded at 340 and 380 nm excitations (F340/F380) during baseline and isoflurane application. F340/F380 ratios were recorded from more than 30 cells in at least three separate experiments. For analyses of [Ca²⁺]_i oscillations, ratio images were collected from more than 60 cells during 20 min in at least four separate experiments. The percentage of cells with obvious [Ca²⁺]_i responses (a single transient increase or sustained oscillations), peak [Ca²⁺]_i, area under the [Ca²⁺]_iversus time curve, and oscillation frequency were determined and analyzed as described previously.²⁸  Hanks’ balanced salt solution samples were collected from inflow and outflow tubes of the recording chamber to determine the anesthetic concentration to which the cells were exposed. High-performance liquid chromatography measurements showed that isoflurane was consistently maintained at 0.4 mM (not shown). This concentration corresponds to a minimum alveolar concentration of approximately 1.

Proteolytic activation of caspase-3 was measured in lysates from DT40 wild-type (DT40-wt), DT40-KO, and DT40-R3 cells after exposure to isoflurane using previously described protocols.¹³  Caspase-3 activity was measured using Ac-DEVD-AFC hydrolysis kit (Caspase-3; Calbiochem, Billerica, MA). In brief, DT40-KO or DT40-R3 cells grown in six well plates were exposed to isoflurane (2.4%) for 24 h which resulted in 0.8 mM isoflurane in the culture medium²⁹  and reliably produced isoflurane toxicity in this anesthesia exposure cell model,¹³  harvested via trypsinization, and washed with phosphate buffered saline. The cell pellet was gently resuspended in CelLytic M lysis buffer with protease inhibitor cocktail (Sigma). Lysate was centrifuged and the resultant supernatant was used for the assay. Ac-DEVD-AFC, the caspase substrate, was added at a final concentration of 50 μM and the samples were incubated for 45 min at 37°C. Ac-DEVD-AFC hydrolysis was monitored by fluorescence emission of the released AFC (excitation, 400 nm; emission, 500 nm) using a multiwavelength excitation dual-wavelength-emission fluorometer (Delta RAM; Photon Technology International, Edison, NJ). Under these excitation and emission conditions, Ac-DEVD-AFC hydrolysis produced a yellow-green fluorescence as suggested by the manufacturer (Calbiochem).

Statistical analyses of data were done with IGOR Pro software (Wavemetrics). All data are presented as mean ± SEM. Sample size for the electrophysiological and Ca²⁺ imaging experiments were used based on previous experiences.^27,30  Experimenters were not blinded to conditions. To avoid error introduced by subjectivity, all electrophysiological experiments in which nuclear membrane patches were successfully isolated were recorded and analyzed.

Statistical comparisons between data points were done using two-tailed tests. All statistically significant differences between data points stated were established using unpaired t tests (with Bonferroni correction for multiple comparisons within a family of inferences). All lacks of statistically significant differences between data points stated were established by ANOVA. Unless a more stringent P value is stated, P value less than 0.05 was used for rejecting the null hypothesis.

Previous studies indicated that InsP₃Rs are involved in isoflurane-mediated increases in [Ca²⁺]_i in PC12 cells, DT40 chicken B lymphocytes, Huntington’s striatal neuronal cell lines, and primary cortical neurons.^11,13,16,31  To test whether isoflurane might directly modulate the activity of the InsP₃R, we measured single InsP₃R channel currents in native ER membranes by nuclear patch clamp electrophysiology.^{4,10,23,24,26,28}  Homo-tetrameric InsP₃R-3 channel activities were recorded by patch clamping outer membranes of nuclei isolated from DT40-KO cells stably expressing the rat type 3 InsP₃R.^10,22,32  This isoform has ligand regulation and permeation properties similar to other InsP₃R isoforms, but with robust gating that provides sensitive detection of modulation of channel activity in patch clamp electrophysiology. InsP₃R-3 channel activity was observed with 2 µM Ca²⁺ and either 1 or 10 µM InsP₃ in the pipette solution (fig. 1, A–C) in 90% or more of patches (table 1; 18 of 20 and 31 of 34 for 1 and 10 µM InsP₃, respectively). Surprisingly, with 400 µM isoflurane in the pipette solution, InsP₃R channel activity could be elicited in the absence of InsP₃ (fig. 1D), although with low open probability (P_o) and in less than 20% of patches (table 1; 15 of 85). The channels activated by isoflurane were identified as InsP₃R by their sensitivity to the InsP₃R antagonists heparin (100 µg/ml) or xestospongin C (1 µM) (fig. 1, E and F; table 1). The mean open probability, open duration (t_o), and closed time (t_c) of the InsP₃R-3 channels activated by 400 µM isoflurane were similar to those of channels activated by subsaturating 1 µM InsP₃ (fig. 1, G–I), suggesting that isoflurane activates InsP₃R-3 channels with similar kinetics to the endogenous agonist InsP₃ but with lower efficacy. Interestingly, the concentration dependence of the modification of InsP₃R-3 channel activity by isoflurane was biphasic: at low isoflurane concentration (<400 µM), increases in isoflurane concentration increased channel P_o, but channel P_o decreased as isoflurane concentration was increased beyond 400 µM. (fig. 2). Thus, InsP₃R-3 channels are activated within a narrow range of isoflurane concentration. This is different from InsP₃ activation of the channel, in which channel P_o increases with [InsP₃] until the saturating [InsP₃] is reached. Beyond that, further increase in [InsP₃] does not enhance channel activity any more.³² 

Gating of InsP₃Rs is under complex allosteric regulation by InsP₃ and [Ca²⁺]_i. In general, InsP₃R channel activity is biphasically regulated by [Ca²⁺]_i with maximum channel P_o observed over a broad range of [Ca²⁺]_i in the presence of saturating (10 µM) InsP₃. In the presence of subsaturating [InsP₃], sensitivity of the channel to inhibition by high [Ca²⁺] is enhanced, resulting in a narrower P_oversus [Ca²⁺]_i dependence.^4,32  Thus, we investigated the [InsP₃] and [Ca²⁺]_i dependencies of isoflurane-activated InsP₃R channel activity. At low (0.4 or 0.9 µM) or high (6 and 10 µM) [Ca²⁺]_i, 400 µM of isoflurane-activated channel P_o was substantially lower (P_o < 0.006) than that observed in 2 µM [Ca²⁺]_i, (P_o = 0.2; fig. 3), resulting in a narrow biphasic [Ca²⁺]_i dependence.

Although isoflurane activated InsP₃R-3 channel activity in the absence of InsP₃ in the pipette solution, the sensitivity of isoflurane activation of the channel to the competitive inhibitor heparin suggests that isoflurane may activate the InsP₃R-3 by sensitizing the channel to low [InsP₃] generated locally in the patched nuclear membrane. To test this, we recorded InsP₃R channel activity in DT40-R3 cells with 400 µM isoflurane, 2 µM Ca²⁺, and a range of [InsP₃] in the pipette solution. As expected, InsP₃ enhanced channel P_o in a dose-dependent manner (fig. 4, A–D). At very low (0.1 and 0.5 µM) [InsP₃], 400 µM isoflurane potentiated channel activity to a level higher than by either agonists alone (fig. 4E). In contrast, at higher (1 µM) [InsP₃], 400 µM isoflurane no longer potentiated channel activity significantly (fig. 4E). In saturating (10 µM) [InsP₃], 400 µM isoflurane did not enhance channel activity measurably (fig. 4E). These results suggest that isoflurane increases the functional sensitivity of the InsP₃R-3 to InsP₃ only at low, subsaturating [InsP₃] (<0.1 µM).

Our single-channel recordings indicate that isoflurane at clinically relevant concentrations potentiates the activity of the InsP₃R-3 at low levels of [InsP₃] that can exist in unstimulated cells in basal conditions. To determine whether this effect influences intracellular Ca²⁺ signaling, we measured [Ca²⁺]_i in individual DT40-KO or DT40-R3 cells kept in complete growth medium. Application of 400 µM isoflurane resulted in a significant transient increases in [Ca²⁺]_i in the DT40-R3 cells that were absent in DT40-KO cells lacking InsP₃R expression (fig. 5, A–C).

To further study the potentiating effects of isoflurane on intracellular Ca²⁺ signaling in weakly stimulated cells, the cells were perfused with complete growth medium containing a low concentration (50 ng/ml) of anti-immunoglobulin M antibody to weakly stimulate the B-cell receptor to generate low, but higher than basal levels of InsP₃.³³  Eighty-two ± 4% of DT40-R3 cells responded to anti-immunoglobulin M antibody with either a single transient increase in [Ca²⁺]_i or with [Ca²⁺]_i oscillations (fig. 5, D and E). Whereas the addition of 400 µM isoflurane did not change the percentage of cells responding to anti-immunoglobulin M antibody (fig. 5, D and E), it enhanced the percentage of cells that responded with a single prolonged [Ca²⁺]_i peak (fig. 5E), consistent with a stronger InsP₃R response. Because of the variable responses of cells (oscillations vs. single-peak responses), we quantified the [Ca²⁺]_i responses in all cells by the areas under the [Ca²⁺]_i-versus-time curves. The presence of isoflurane significantly increased the area under curve for cells responding with a single-peak transient (fig. 5F; P < 0.001) or with [Ca²⁺]_i oscillations (fig. 5G; P < 0.01). These results suggest that isoflurane, at a clinically used concentration, potentiates low-level InsP₃-mediated [Ca²⁺]_i signaling.

Prolonged exposure to isoflurane is associated with widespread cell death in diverse in vivo and in vitro systems.^13,16  Because activation of InsP₃R-mediated Ca²⁺ signaling has been linked to apoptosis,^34,35  and considering the effects of isoflurane on InsP₃R gating observed here, we assessed the role of this effect in isoflurane-mediated apoptosis. Caspases are a family of cysteine proteases that play crucial roles in apoptosis. Activation of caspase-3 is a central event in the progression of programmed cell death. We therefore monitored caspase-3 cleavage as a measure of apoptosis. In agreement with our previous observations of cells expressing endogenous InsP₃R,^13,16  isoflurane triggered apoptosis in the DT40-R3 cells, but not in the InsP₃R-deficient cells (fig. 5H). These observations support a role for exaggerated activation of InsP₃R in isoflurane-induced apoptosis, consistent with previous observations.^11,13,16 

We have, for the first time, demonstrated that isoflurane at clinically relevant concentrations modulates the activity of the InsP₃R Ca²⁺-release channel at the single-channel level. The modulating effects of isoflurane on the InsP₃R regulate Ca²⁺ release from the ER that results in exaggerated [Ca²⁺]_i signaling. We have also demonstrated, for the first time at the molecular level, that isoflurane causes Ca²⁺ release from the ER via this activation of InsP₃R and can therefore affect intracellular Ca²⁺ homeostasis, regulation of cytosolic Ca²⁺ oscillations, and cell survival. These results provide novel insights into possible molecular mechanisms of anesthesia-mediated effects on neurodegeneration and cognitive function.

It was previously demonstrated that both isoflurane and halothane may increase [Ca²⁺]_i primarily by inducing Ca²⁺ release from the intracellular Ca²⁺ stores in neurons.³⁶  Our results provide the first evidence that an inhalational anesthetic can modulate activation of InsP₃R channels. Interestingly, isoflurane activates InsP₃R channels with a biphasic dose–response with optimal concentrations approximately 0.4 to 1 mM, close to clinically used concentrations for general anesthesia. This activation of InsP₃R channels by isoflurane showed a strong Ca²⁺ concentration dependence, with the optimum [Ca²⁺]_i of 2 µM, qualitatively similar to the [Ca²⁺] dependence of InsP₃-activated channel activity. The InsP₃ competitive antagonist, heparin, blocked the ability of isoflurane to activate InsP₃Rs. Whether isoflurane requires InsP₃ for it to activate the channel, or isoflurane activation is not only direct but also sensitive to heparin inhibition, is unclear. The former may be likely because isoflurane (0.4 mM) potentiated activation of the InsP₃R by low concentrations of InsP₃, but failed to further enhance InsP₃R-3 activity at saturating [InsP₃]. Nevertheless, further studies are needed to investigate possible biochemical interactions between isoflurane and the InsP₃R. Together, these results establish the InsP₃R as a novel target of isoflurane and perhaps other inhalational or intravenous general anesthetics.

The biphasic effects of [Ca²⁺]_i on InsP₃R activation play important roles on intracellular Ca²⁺ oscillations, waves, and spreading of global Ca²⁺ signals.³⁷  Our finding that isoflurane activates InsP₃R-3 channels raised the possibility that it could enhance [Ca²⁺]_i signals. Application of 400 µM isoflurane resulted in a transient increase in [Ca²⁺]_i in DT40 cells expressing InsP₃R-3 but not in DT40 cells lacking the channels. In addition, [Ca²⁺]_i signals generated by immunoglobulin M stimulation of DT40-R3 cells, both single-peaks and oscillations, were more prominent in the presence of isoflurane (fig. 5, F and G).

We previously showed in in vitro and in vivo models that exposure to isoflurane for prolonged durations significantly induced apoptosis that required activation of InsP₃Rs.^11,13  Our results here suggest that this is medicated at least in part by isoflurane activation of InsP₃R channel gating. The effects of isoflurane on changes in [Ca²⁺]_i are in good agreement with the conclusions reached from the single-channel recordings of the effects of isoflurane on InsP₃R activity observed in the current study. It has long been known that anesthetics including halothane activate the other major ER Ca²⁺-release channels, the RyR channel complex,³⁸  and this is thought to be the basis for malignant hyperthermia.³⁹  Similar to InsP₃Rs, RyRs are expressed throughout the nervous system and play important roles in both normal cell functions⁴⁰  and in various neurodegenerative diseases.^2,41–43  Our previous study indicated a role of RyR activation in isoflurane-induced apoptosis in neuronal tissue cultures.²⁹  Both InsP₃R and RyR contribute to regulation of intracellular Ca²⁺ homeostasis and may interact with each other through Ca²⁺-induced Ca²⁺ release in a common pathway in normal neuronal function and neurodegeneration. Excessive Ca²⁺ release from the ER via these release channels could cause Ca²⁺ overload in mitochondria and depletion of ER Ca²⁺, both of which can contribute to apoptosis.^6,34,44  In addition, mitochondrial Ca²⁺ overload causes cytochrome C release, activating caspase-3, which can cleave the InsP₃R, resulting in a constituent Ca²⁺ leak from the ER.⁴⁵  Thus, excessive or prolonged activation of InsP₃R by isoflurane may set in motion a cascade of events resulting in apoptosis in different tissue culture cells including neurons^{3,11,13,15–17}  and in developing brains.¹⁶  Our data suggest that the InsP₃R may represent an ideal target for prevention of the harmful side effects of inhalational anesthetics. Inhibition of excessive activation of the InsP₃R may ameliorate or prevent anesthesia-mediated neurodegeneration as demonstrated in animal models.¹⁶  This is especially relevant in pediatric^46,47  and aged patients,⁴⁸  who seem to be the most vulnerable to the harmful side effects of anesthetics. Unfortunately, there is no good pharmacology for the InsP₃R. Although heparin is an InsP₃R antagonist, its poor penetration across cell membranes limits its use for protection against anesthetic neurotoxicity in animals or patients. It should be noted that anesthetics have also been shown to be protective against various stresses, also by activation of InsP₃Rs in different tissue culture models.^{3,12,14,49,50}  As demonstrated here (fig. 6), mild or moderate activation of InsP₃R and moderate Ca²⁺ release from the ER by isoflurane provides cytoprotection, possibly via physiological Ca²⁺ uptake into mitochondria and stimulation of adenosine triphosphate production,²⁵  or by activation of AKT and microtubule-associated protein kinase/extracellular-signal-regulated kinases cytoprotective pathways.^12,14,50  Nevertheless, it is prudent to minimize the use of general anesthetics as much as possible, so that their beneficial effects can be used and the harmful effects be minimized.

In summary, our results indicate that the commonly used inhalational anesthetic isoflurane modulates gating of the InsP₃R Ca²⁺-release channel, enhancing ER Ca²⁺ release and contributing to isoflurane-mediated apoptosis. These results suggest that InsP₃Rs are molecular targets of general anesthetics and that these receptors may provide the basis for some pharmacologic effects of general anesthetics and therapeutic interventions for anesthesia-mediated cell death by apoptosis.

The authors thank Qingcheng Meng, Ph.D. (Department of Anesthesiology, University of Pennsylvania, Philadelphia, Pennsylvania), for assistance in measuring isoflurane concentrations used in the electrophysiology and Ca²⁺ measurement experiments. The authors appreciate the valuable discussions and support from Roderic Eckenhoff, M.D., Maryellen Eckenhoff, Ph.D., and Lee Fleisher, M.D. (Department of Anesthesiology and Critical Care, University of Pennsylvania).

This work was supported by the National Institutes of Health (Baltimore, Maryland) awards (GM084979 and GM084979-02S1 to Dr. Wei, GM056328 and MH059937 to Dr. Foskett, and K08-GM073224 to Dr. Wei); March of Dimes Birth Defects Foundation Research (White Plains, New York; Grant 12-FY08-167 to Dr. Wei); and the Research Fund at the Department of Anesthesiology at the University of Pennsylvania (Philadelphia, Pennsylvania).

The authors declare no competing interests.