Isoflurane Modulates Hippocampal Cornu Ammonis Pyramidal Neuron Excitability by Inhibition of Both Transient and Persistent Sodium Currents in Mice

Although general anesthetics have been used clinically for more than 170 yr, their molecular and cellular mechanisms of action are still not clear. Compared with intravenous general anesthetics like propofol, the molecular mechanisms of volatile anesthetics such as isoflurane and sevoflurane are more complex because they interact with multiple molecular targets.^1,2  Volatile anesthetics produce the desirable pharmacologic endpoints of amnesia, unconsciousness, and immobility, but also unwanted side effects including respiratory and cardiovascular depression and developmental neurotoxicity.^3–5  Therefore, understanding the various molecular targets that contribute to specific desired and undesired endpoints is important for a complete pharmacologic understanding and for development of novel anesthetics with improved safety profiles.

Volatile anesthetics depress neurotransmitter release with greater potency at excitatory than inhibitory synapses, which contributes to volatile anesthetic depression of central nervous system function.⁶  Volatile anesthetics at clinically relevant concentrations inhibit sodium currents in transfected cells,^7,8  nerve terminals,^8,9  and dorsal root ganglion neurons.¹⁰  Voltage-gated sodium channels are important targets for the presynaptic effects of volatile anesthetics on excitatory neurotransmitter release.^11,12  However, whether volatile anesthetics can directly depress activity of excitatory neurons by inhibiting sodium currents is unclear.

Initiation and propagation of action potentials are important for synaptic transmission and neuronal plasticity.¹³  Even a small change in action potential properties can lead to significant modulation of synaptic transmission.¹⁴  Transient sodium currents are crucial for membrane excitability, including initiation and propagation of action potentials.¹³  Voltage-gated sodium channels can also produce persistent and resurgent sodium currents to modulate excitability of neuronal networks.^15,16  Slowly inactivating or noninactivating sodium currents (persistent sodium currents) are activated at sub-threshold voltages and enhance repetitive firing.¹⁷  Resurgent sodium currents occur with repolarization after a previous period of depolarization; neurons with significant resurgent sodium currents share the capacity for rapid spontaneous firing or burst firing.¹⁶  Changes in persistent or resurgent sodium currents are implicated in several diseases including epilepsy, paramyotonia congenita, and extreme pain syndromes.^17,18  However, whether volatile anesthetics directly modulate persistent sodium currents and/or resurgent sodium currents is unknown.

The hippocampus is critical to many brain functions including learning and memory,¹⁹  which are sensitive to the actions of general anesthetics.^20–22  Pyramidal neurons are the principal excitatory neurons in the hippocampus and express multiple subtypes of voltage-gated sodium channels.²³  We designed the present study to test the hypothesis that the volatile anesthetic isoflurane directly modulates excitability of hippocampal pyramidal neurons by inhibiting both transient and persistent sodium currents.

Working solutions of isoflurane were prepared from saturated solutions (12 to 15 mM isoflurane, as measured by gas chromatography²⁴ ) in extracellular solution consisting of (in mM) 125 NaCl, 25 NaHCO₃, 1.25 NaH₂PO₄, 2.5 KCl, 1 MgCl₂, 2 CaCl₂, and 15 glucose, pH = 7.3 ± 0.5. The saturated solution was prepared by rotation in a gas-tight vial for at least 24 h. The saturated solution was diluted to experimental concentrations immediately before perfusion, and final concentrations of isoflurane were confirmed by gas chromatography.²⁴  Isoflurane 0.27 mM was used as the predicted minimum alveolar concentration (MAC, equivalent to the EC₅₀ for immobilization) for mouse adjusted to room temperature (~25°C).^25–27  Isoflurane was purchased from Abbott Pharmaceutical Co. Ltd. (China), tetrodotoxin was purchased from Alomone Labs (Israel), and other compounds were obtained from Sigma-Aldrich (China).

Procedures were approved by the Animal Ethics Committee of Sichuan University (Chengdu, China). Randomization and blinding methods were not used in these electrophysiologic recordings. C57BL/6 mice (28 male and 25 female, 53 in total) at 7 to 10 postnatal days were anesthetized with ketamine/xylazine (60/10 mg/kg) and decapitated. The brain was rapidly removed and put into ice-cold oxygenated (95% O₂/5% CO₂) sucrose-substituted dissecting solution containing (in mM): 87 NaCl, 25 NaHCO₃, 1.25 NaH₂PO₄, 2.5 KCl, 7.5 MgCl₂, 75 sucrose, and 25 glucose. Horizontal hippocampal slices (thickness of 270 μm) were cut using a vibratome (Leica VT1000 A, USA), then incubated for 30 min at 37°C and then at room temperature (23° to 25°C) in extracellular solution aerated with 95% O₂/5% CO₂. After incubation, hippocampal slices were mounted in the recording chamber for electrophysiologic recordings at room temperature.⁸ 

Hippocampal slices were placed in a recording chamber and continuously perfused with extracellular solution at 2 ml/min. Pyramidal neurons in the hippocampal cornu ammonis region were directly visualized and identified by their shape and size with infrared differential interference contrast imaging microscopy. Whole-cell voltage-clamp recordings were applied to record sodium currents. Electrophysiologic recordings were conducted using an Axopatch 200B amplifier, 1,440 Digidata and coupled with pClamp 10.2 software (Molecular Devices, USA). Currents were sampled at 20 kHz and filtered at 5 kHz. The external solution was the same as the incubation solution but supplemented with 25 mM TEA-Cl to block potassium currents. The resistance of glass pipettes was 3 to 4 MΩ and the internal pipette solution contained (in mM): 110 CsF, 9 NaCl, 1.8 MgCl₂, 4 Mg-ATP, 0.3 Na-GTP, 0.09 EGTA, 0.018 CaCl₂, 9 HEPES, 10 TEA-Cl. CsOH was used to adjust to pH = 7.38 (290 to 310 mOsm). Tetrodotoxin-sensitive currents were confirmed by subtraction after application of 500 nM tetrodotoxin. Series resistance was compensated by approximately 70 to 75%, and cells were rejected when series resistance exceeded 15 MΩ. Action potentials were recorded under current-clamp mode with an extracellular solution containing (in mM): 125 NaCl, 25 NaHCO₃, 1.25 NaH₂PO₄, 2.5 KCl, 1 MgCl₂, 2 CaCl₂, 15 glucose, and 10 μM bicuculline or 100 μM picrotoxin to block possible effects from γ-aminobutyric acid–mediated (GABAergic) inputs. The internal pipette solution for current-clamp mode contained (in mM): 122 K-methanesulfonate, 9 NaCl, 1.8 MgCl₂, 4 Mg-ATP, 0.3 Na-GTP, 0.09 EGTA, 0.018 CaCl₂, and 9 HEPES (pH adjusted to 7.35 with KOH, 290 to 310 mOsm).

No power calculations were conducted before the study. Sample sizes were based on our experience with similar experimental designs. No data were lost, but data were not collected when cell series resistance exceeded 15 MΩ. Electrophysiologic data were analyzed using Clampfit 10.0 software (Molecular Devices), Graph-Pad Prism 6 (Graph-Pad Software Inc., USA), and Origin 10.0 (OriginLab, USA). Half-maximal inhibitory concentration (IC₅₀) values were obtained by least squares fitting to the Hill equation: Y = 1/(1 + 10((logIC₅₀ − X) · h), where Y is the effect, X is the measured concentration of isoflurane, and h is the Hill slope. For transient sodium currents, voltage-dependence of half-maximal activation and half-maximal inactivation were fitted to the Boltzmann equation: G/G_max = 1/[1 + exp (V_½ − V/k)], where G = conductance, G/G_max = normalized conductance; V_½ = voltage of half-maximal activation or inactivation, and k = slope factor. Sodium current recovery traces and time constants (τ) were determined by fitting to the mono-exponential function: Y = exp (−τ · n) + A_P, where τ is the time constant, A_P is the plateau, and n is stimulus number based on complete recovery time. Inhibition of peak transient sodium current was normalized as %inhibition versus control and compared by chi-square test. For persistent sodium currents, currents were measured as the preserved currents from 450 to 475 ms after stimulus. Resurgent sodium current refers to the peak current during the first 50 ms following step back. Both persistent and resurgent sodium currents were corrected for each individual trace by subtracting current recorded in the presence of tetrodotoxin. Normal distribution of data was tested by the Shapiro–Wilk test. All data are presented as mean ± SD. Repeated measures data were analyzed by two-way ANOVA with Bonferroni post hoc testing. Two-tailed independent-samples or paired Student’s t test were used for comparison between control and isoflurane conditions. Statistical significance was set as P < 0.05.

Transient sodium channel current was activated by depolarization to 0 mV from holding potentials of −70 or −120 mV. At a clinically relevant concentration (0.55 mM, ~2 MAC), isoflurane inhibited transient sodium channel current in a voltage-dependent manner (fig. 1A). Inhibition was significantly greater at the physiologic holding potential of −70 mV (37.8% ± 4.6% inhibition, n = 9) than at a hyperpolarized holding potential of −120 mV (3.1% ± 7.6% inhibition, n = 9, P = 0.012 vs. −70 mV; fig. 1B). The median inhibitory concentration of isoflurane for inhibition of transient sodium channel current was 1.0 ± 0.3 mM (~3.7 MAC) from a holding potential of −70 mV, whereas the currents from a holding potential of −120 mV were minimally inhibited (median inhibitory concentration = 3.6 ± 0.7 mM, ~13.3 MAC; fig. 1C). With repeated 5-ms pulses depolarizations at 50 Hz (fig. 1D), normalizing sodium channel current of each pulse to that of the first pulse (Pulse_n/Pulse₁) removed the effect of resting block by isoflurane.²⁴  Thus the reduced sodium channel current at the 10th pulse reflected activity-dependent inhibition as a result of repeated membrane depolarizations (fig. 1E). From a holding potential of −70 mV, isoflurane reduced Pulse₁₀/Pulse₁ ratio from 0.49 ± 0.02 to 0.38 ± 0.03 (fig. 1F, n = 7, P = 0.010). This suggests that isoflurane leads to progressive inhibition of sodium channel current during trains of action potentials as a result of delayed recovery from inactivation produced by isoflurane.

To test the effects of isoflurane on voltage-gated sodium channel activation, sodium current was activated by a series of voltage steps from −70 to +70 mV preceded by a −110 mV prepulse to relieve channel inactivation (fig. 2A). Isoflurane at 0.42 mM (~1.6 MAC) did not shift the current–voltage relationship, and maximum sodium channel current for both control and isoflurane conditions occurred at −20 mV (fig. 2B). Voltage-dependent activation was unchanged by isoflurane: voltage-dependence of half-maximal activation was −23.9 ± 0.2 mV for control compared to −24.2 ± 0.2 mV with isoflurane (fig. 2C, n = 8, P = 0.144). The effect of isoflurane on steady-state inactivation was determined by eliciting currents at 0 mV after a 20-ms prepulse to voltages from −120 to +10 mV (fig. 2D). Normalized sodium channel current/maximum sodium channel current values reflected the fraction of channels inactivated during the prepulse.⁹  Isoflurane (0.42 mM, ~1.6 MAC) shifted the steady-state inactivation curve in the hyperpolarizing direction; with a voltage of half-maximal inactivation of −46.7 ± 0.9 mV for control compared with −53.2 ± 0.8 mV with isoflurane (fig. 2E, n = 11, P < 0.0001). These data suggest that isoflurane inhibits peak sodium channel current by increasing the fraction of inactivated channels.

As neuronal firing frequency depends in part on how fast voltage-gated sodium channels can cycle through its various activation states, we measured the time-course of recovery from inactivation. Peak sodium channel current was recorded in response to two 10-ms pulses to 0 mV from a holding potential of −70 mV, where the duration between the two pulses was varied from 1 to 400 ms (fig. 2F). Recovery time-courses were fit to a mono-exponential function for both control and isoflurane conditions, indicating that channels predominantly entered a single fast-inactivated state.²⁴  Isoflurane increased the time required for channel recovery, with the recovery time constant increasing from 7.5 ± 0.6 to 12.7 ± 1.3 ms (fig. 2G, n = 13, P < 0.0001).

Persistent currents after fast-inactivated transient currents were recorded with 500-ms voltage steps from −70 to +20 mV (fig. 3A). The mean residual currents during 450 to 475 ms were calculated as persistent sodium currents.¹⁸  The largest persistent currents were found at a voltage step to −40 mV (fig. 3B). Isoflurane inhibited persistent sodium currents at nearly all voltage steps from −60 to 0 mV (fig. 3C). The median inhibitory concentration of isoflurane for persistent sodium currents was 0.4 ± 0.1 mM (~1.5 MAC) from a holding potential of −70 mV with depolarization to −40 mV (the voltage for maximal persistent currents; fig. 3D). Persistent sodium currents were corrected for each individual trace by subtracting current recorded in the presence of tetrodotoxin (500 nM). When persistent sodium current was evoked by a ramp depolarization stimulus (from −80 to 0 mV at 30 mV/s),²⁸  isoflurane decreased persistent sodium current density (fig. 3E and 3F, n = 11). Persistent sodium current was tetrodotoxin-sensitive as it was inhibited completely by 500 nM tetrodotoxin (fig. 3E). Resurgent sodium current is the residual fraction of the slow sodium current that can be activated by the next depolarization.¹⁶  Isoflurane inhibited the resurgent current evoked by voltage steps from −70 to 0 mV with a 20-ms prepulse to +30 mV (fig. 3G-H, n = 7, all P < 0.05 for voltages from −70 to −10 mV).

Sodium currents at sub-threshold voltages determine the threshold of neuronal activation and are critical to synaptic transmission.²⁹  To test the effects of isoflurane on sub-threshold sodium currents, we applied successive 5-mV step depolarizations at the same rate as the ramp depolarization²⁹  (fig. 4A). Both transient and steady-state sodium channel current were evoked at these potentials. The largest transient and steady-state sodium channel current densities (current/conductance) were found at a voltage step to −60 mV, with current densities of −24.0 ± 7.9 and −2.0 ± 0.3 sodium channel current densities (current/conductance), respectively (fig. 4B). Isoflurane (0.42 mM, ~1.6 MAC) inhibited the transient sodium current at voltage steps to −50 and −45 mV, and steady-state sodium currents at voltage steps to −60 mV (fig. 4C and 4D, n = 9).

Single action potential were evoked by injection of 60 pA current for 100 ms in current-clamp mode (fig. 5A). Isoflurane reduced action potential amplitude from 104.5 ± 5.1 to 90.0 ± 3.0 mV (fig. 5B left, n = 13, P = 0.001) and increased action potential width from 2.3 ± 0.2 to 2.6 ± 0.2 ms (fig. 5B right, n = 13, P = 0.001). By phase-plane plot voltage-time derivative analysis, isoflurane inhibited the whole time course of somatic spikes recorded from hippocampal cornu ammonis pyramidal neurons (fig. 5C). Firing of action potentials was activated by 1-s series of current injection from 120 to 60 pA (fig. 5D). Isoflurane reduced firing frequency from 10.2 ± 1.6 to 3.3 ± 1.0 Hz for 30 pA injections (fig. 5E, n = 20, P = 0.0002), from 13.2 ± 1.5 to 7.2 ± 1.0 Hz for 60 pA injections (fig. 5E, n = 20, P < 0.0001), and from 12.5 ± 1.5 to 7.6 ± 1.0 Hz for 90 pA injections (fig. 5E, n = 20, P < 0.0001). Isoflurane hyperpolarized the resting membrane potential from −54.6 ± 2.3 to −58.7 ± 2.1 mV (fig. 5F, n = 16, P = 0.001). Rheobase was increased by isoflurane from 21.4 ± 5.7 to 42.9 ± 11.5 pA (fig. 5G, n = 14, P = 0.019), and input resistance was increased from 220 ± 17 to 241 ± 17 MΩ (fig. 5H, n = 12, P < 0.0001). The effects of isoflurane on action potentials firing frequency, rheobase, neuronal input resistance, and resting membrane potential represent the combined results with recordings under bicuculline or picrotoxin.

Specific molecular targets mediate the neurophysiologic actions of general anesthetics on critical properties and processes such as neuronal excitability, axonal conduction, and synaptic transmission.^30,31  Voltage-gated sodium channels are essential ion channels in mediating the rising phase of action potentials^13,32  and are implicated as presynaptic targets for general anesthetics.³  Although early studies on the squid giant axon found that action potentials were relatively insensitive to clinical concentrations of general anesthetics,³³  more recent studies report significant sensitivity of sodium channels to volatile anesthetics at clinically relevant concentrations.^9,34  Supporting in vivo animal studies show that intravenous administration of the voltage-gated sodium channels blocker lidocaine reduces MAC for several volatile anesthetics both in animals and humans.^35–37  Moreover, intrathecal injection of tetrodotoxin reduces MAC for isoflurane in rats,³⁵  whereas intrathecal veratridine, a toxin that promotes the voltage-gated sodium channels open state, increases MAC and antagonizes the effect of intrathecal tetrodotoxin.³⁸ 

Volatile anesthetics including isoflurane and sevoflurane inhibit heterologously expressed recombinant brain voltage-gated sodium channels at clinical concentrations by selectively interacting with the inactivated state of the channel.³¹  Our findings support and extend these findings by demonstrating effects of isoflurane on sodium channel current and action potentials in native hippocampal neurons in situ. Voltage- and activity-dependent inhibition of voltage-gated sodium channels by volatile anesthetics can be explained by preferential interaction with the inactivated state to impede transition from the inactivated state back to the resting state.^24,31  The median inhibitory concentration of isoflurane for inhibition of transient sodium channel currents from a holding potential of −70 mV is relevant to the clinically used concentrations, whereas isoflurane did not inhibit transient sodium channel currents at clinically relevant concentrations from a holding potential of −120 mV. Even perfusion with saturated isoflurane stock solution (12 to 14 mM, ~48 MAC) only depressed transient sodium channel current by ~60% from a holding potential of −120 mV. The partial inactivation state sodium channels existing at a holding potential of −70 mV can explain this result. Volatile anesthetics including isoflurane also display activity-dependent inhibition of neuronal sodium channels in isolated rat neurophysical nerve terminals,⁹  neuroblastoma cells,²⁴  and dorsal root ganglion neurons,¹⁰  and in the bacterial voltage-gated sodium channels homologue voltage-gated sodium channel of Bacillus halodurans,^39,40  but there was no pervious evidence that volatile anesthetics directly depress activity of excitatory neurons by inhibiting neuronal sodium channels.

The major anion in our pipette solution was fluorion, which has been reported to affect inactivation of voltage-gated sodium channels in squid giant axons.⁴¹  We used fluorion instead of chlorine in the pipette solution for voltage-gated sodium channels recordings^7,42  to avoid changes in intracellular chlorine that might affect neuronal excitability, signaling, and plasticity of hippocampal pyramidal neurons.^43,44  Since isoflurane may also potentiate chlorine permeable γ-aminobutyric acid type A receptors, the level of intracellular chlorine has important consequences for its actions in a manner similar to that recently shown for propofol.^45,46 

We provide evidence that isoflurane at clinically relevant concentrations inhibits transient sodium channel currents, persistent sodium currents, and resurgent sodium currents at physiologic holding potentials in mouse hippocampal cornu ammonis pyramidal neurons. For transient sodium channel current, isoflurane inhibits peak sodium current and modulates channel gating by shifting the voltage-dependence of voltage-gated sodium channels inactivation, and by delaying recovery from inactivation. Isoflurane also inhibits persistent sodium currents and resurgent sodium currents at clinically relevant concentrations.^6,47  From a holding potential of −70 mV, the median inhibitory concentration of isoflurane for persistent sodium current inhibition was lower than for transient sodium channel current (0.4 ± 0.1 vs. 1.0 ± 0.3 mM), indicating that isoflurane selectively inhibits the persistent currents of sodium channels. This inhibition by isoflurane of persistent sodium current at clinically relevant concentrations might reduce neuronal excitability such as lower action potentials firing frequency and increased rheobase.

At least three activation states have been identified for neuronal sodium channels depending on membrane potential: resting (closed), activated (open), and inactivated.¹⁶  transient sodium channel currents are evoked by step depolarization and rapidly activate, whereas the subsequent residual sodium currents that last throughout the step are referred to as persistent sodium currents.¹⁷  Upon repolarization, closed voltage-gated sodium channels reopen to produce resurgent currents, known as resurgent sodium currents.¹⁶  Although these three components of sodium currents all flow through the same channels, they have different kinetics and physiologic roles. Transient sodium channel currents are pivotal for initiation and propagation of action potentials,³²  whereas persistent sodium currents and resurgent sodium currents amplify responses of neurons to synaptic input and enhance repetitive firing.¹⁸  The rapid kinetics and all-or-none properties of transient sodium channel currents mediate the upstroke of fast neuronal action potentials.¹³  Persistent and resurgent sodium currents are activated by small synaptic depolarization and are slower than transient sodium channel currents. Despite their small amplitudes compared with transient sodium channel currents, persistent and resurgent sodium currents can profoundly alter neuronal firing behavior, especially at subthreshold voltages.^16,17  In dendrites, persistent sodium currents can boost distal synaptic potentials to propagate to the neuronal soma.¹⁵  In proximal axons and peripheral axons, persistent sodium currents can affect initiation of action potentials.^48,49  Both persistent and resurgent sodium currents also play critical roles in the regulation of neuronal firing behavior because of their high densities near the axon initial segment.⁵⁰  Resurgent sodium current is enhanced by specific pathogenic mutations in sodium channels, thereby increasing action potential firing rate and duration, causing neuronal hyperexcitability.⁵¹  Inhibition by isoflurane of these neuronal sodium channel currents provides neurophysiologic mechanisms to support previous reports that isoflurane reduces excitability of cornu ammonis pyramidal neurons by depressing action potentials.

Neuronal information coding occurs by varying frequencies and patterns of action potentials between somata and axons, where they can rapidly propagate information over distance. action potentials are fundamentally important for synaptic transmission and neuroplasticity.^52,53  Neurotransmitter release is largely determined by the shape, frequency, and pattern of presynaptic action potentials during the process of information transfer.¹⁴  Release of many neurotransmitters is inhibited by volatile anesthetics in a sodium channel–dependent manner.^54,55  Thus voltage-gated sodium channels is a plausible presynaptic molecular target for volatile anesthetics.

Isoflurane decreased excitability of mouse cornu ammonis pyramidal neurons by shaping the action potentials and hyperpolarizing the resting membrane potential. The generation of action potentials depends on multiple ion channels including voltage-gated sodium channels for rapid depolarization, voltage-gated calcium channels for slow depolarization, and potassium channels for repolarization.⁵⁶  Analysis of action potential kinetics (voltage-time derivative) showed that isoflurane slowed action potential kinetics throughout, consistent with net inhibition. Isoflurane decreased action potential amplitude mainly from inhibition of transient voltage-gated sodium channel currents, as seen in rat neurohypophysial terminals.⁹  Prolonged action potential duration impairs neuronal responses to high frequency stimuli, enhancing the depression of synaptic transmission. The increased action potential half-width by isoflurane is likely attributable to inhibition of potassium channels.⁵⁷  Isoflurane at clinical concentrations can enhance and stabilize the inactivation state of voltage-gated sodium channels, resulting in activity-dependent reduction of sodium channel current with high-frequency stimuli, which could partly contribute to slow the rapid depolarization of action potential by isoflurane. Inhibition by isoflurane of persistent sodium channel current may also contribute to its effect on action potential amplitude and duration. Isoflurane slightly hyperpolarized resting membrane potential possibly as a result of enhancement of volatile anesthetic-sensitive background K⁺ channels.⁵⁸  The effects of isoflurane on resting membrane potential have been reported to be varied^47,59  and voltage-dependent;⁶⁰  expression of different K⁺ channel subtypes may contribute to these differential effects.⁵⁸ 

In conclusion, isoflurane can inhibit three components of sodium channel current, which contribute to its depression of neuronal excitability and action potentials in hippocampal cornu ammonis pyramidal neurons. Voltage-dependent and activity-dependent inhibition by isoflurane, and possibly other volatile anesthetics, depresses neuronal excitability and may contribute to anesthetic effects on neurotransmitter release and synaptic plasticity.

Supported by grant No. 81771486 (to Dr. Zhou) and No. 81600918 (to Dr. Liang) from the National Natural Science Foundation of China (Beijing, China), and grant No. R01 GM-58055 from the National Institutes of Health (Bethesda, Maryland; to Dr. Hemmings).

The authors declare no competing interests.