Abstract
Volatile (inhaled) anesthetics cause amnesia at concentrations well below those that cause loss of consciousness and immobility; however, the underlying neuronal mechanisms are unknown. Although many anesthetics increase inhibitory GABAergic synaptic transmission, this effect occurs only at high concentrations (>100 μm). Molecular targets for low concentrations of inhaled anesthetics have not been identified. Here, we report that a tonic inhibitory conductance in hippocampal pyramidal neurons generated by α5 subunit-containing GABAA receptors is highly sensitive to low concentrations of the volatile anesthetic isoflurane (ISO) (25 and 83.3 μm). The α5 subunit is necessary for enhancement of the tonic current by these low concentrations of isoflurane because potentiation is absent in neurons from α5-/- mice. Furthermore, ISO (25 μm) potentiated recombinant human α5β3γ2L GABAA receptors, whereas this effect was not seen with α1β3γ2L GABAA receptors. These studies suggest that an increased tonic inhibition in the hippocampus may contribute to amnestic properties of volatile anesthetics.
Introduction
The general anesthetic state is characterized by multiple desirable endpoints, including amnesia, analgesia, sedation, and immobility in response to pain (Campagna et al., 2003). Since the classical description of the “Stages of Ether Anesthesia,” it has been recognized that different concentrations of anesthetics produce distinct behavioral effects (Gillespie, 1943). Memory is particularly sensitive to general anesthetics because amnesia occurs at concentrations well below those that cause sedation and immobility (Cook et al., 1978; Veselis et al., 2001; Pain et al., 2002). Commonly used inhaled anesthetics cause amnesia at concentrations fourfold lower than those required for immobility (Dwyer et al., 1992; Kandel et al., 1996; Dutton et al., 2001). The GABA subtype A receptor (GABAAR) is thought to be a primary target for most anesthetics, and the concentrations of halogenated anesthetics that potentiate GABAARs correlate with their abilities to induce general anesthesia (Mihic et al., 1994; Campagna et al., 2003). However, GABAARs that are sensitive to low concentrations of volatile anesthetics that cause amnesia have yet to be identified.
GABAARs are pentameric anion-selective ion channels that assemble from different classes of subunits (α1-6, β1-3, γ1-3, δ, π, θ, ϵ) (McKernan and Whiting, 1996). The subunit composition of GABAARs determines their pharmacological and biophysical properties as well as subcellular distribution patterns (Belelli et al., 1999). GABAARs regulate the transfer of information at inhibitory synapses by generating transient IPSCs. Recently, GABAARs have also been shown to generate a tonic form of inhibition through activation by low ambient concentrations of GABA (Bai et al., 2001; Semyanov et al., 2004). In CA1 pyramidal neurons, the tonic inhibitory conductance is generated by α5 subunit-containing GABAARs (α5GABAARs) (Caraiscos et al., 2004). Because the sensitivity of GABAARs to volatile anesthetics is influenced by the α isoform (Jenkins et al., 2001; Korpi et al., 2002), we tested the hypothesis that α5GABAARs that generate a tonic current in hippocampal pyramidal neurons are highly sensitive to low concentrations of the prototypic volatile anesthetic isoflurane (ISO). The results provide the first evidence of native GABAARs in hippocampal neurons that are sensitive to subanesthetic concentrations of ISO (25 μm). We also show that the α5 subunit is necessary for conferring sensitivity of pyramidal neurons to ISO (25 μm), using neurons from α5-/- mice.
Materials and Methods
Cell cultures and whole-cell recording. Experiments were approved by the Animal Care Committee of the University of Toronto. Primary cultures of hippocampal and cortical neurons were prepared as described previously (MacDonald et al., 1989) from α5-/- mice and wild-type (WT) α5+/+ littermates on postnatal day 1. The α5-/- mice were generated in a mixed 50% C57BL/6 and 50% 129SvEv genetic background, as described previously (Collinson et al., 2002). Given the limited availability of α5+/- breeding pairs, neuronal cultures were also prepared from Swiss White mice and were used to determine the threshold concentration of isoflurane that modulated the tonic and synaptic currents. All other experiments that involved neurons were performed using WT and α5-/- mice. The tonic current is readily detected in cultured hippocampal neurons not treated with pharmacological agents intended to increase the ambient concentration of GABA (Bai et al., 2001; Yeung et al., 2003). However, to increase the amplitude of the tonic current and facilitate its characterization, cultures were treated with the GABA transaminase inhibitor vigabatrin (100 μm) for 24 hr before recording (Wu et al., 2003).
The extracellular solution contained (in mm): 140 NaCl, 1.3 CaCl2, 5.4 KCl, 25 HEPES, and 28 glucose, pH 7.4. Tetrodotoxin (300 nm) was added to block Na+ channels, and 6-cyano-7-nitro-quinoxaline-2,3-dione (10 μm) and 2-amino-4-phosphonovaleric acid (40 μm) were added to inhibit ionotropic glutamate receptors. Neurons were perfused using a computer-controlled multibarreled perfusion system with an exchange time of ∼3 msec (SF-77B; Warner Instruments, Hamden, CT). Whole-cell currents were recorded under voltage-clamp (-60 mV) conditions using an Axopatch 200 amplifier (Axon Instruments, Union City, CA) that was interfaced to a Digidata 1200 (InstruTech, Elmont, NY). To monitor series resistance, a hyperpolarizing voltage step of 10 mV was applied. Only cells that demonstrated a stable access resistance were used for data analysis. Patch electrodes were filled with a solution containing (in mm): 140 CsCl, 10 HEPES, 11 EGTA, 2 MgCl2, 1 CaCl2, 4 MgATP, and 2 tetraethylammonium, pH 7.3. Experiments were performed at room temperature.
ISO solutions were prepared and applied as described previously (Joo et al., 2001). The concentrations of ISO in samples obtained at the end of the perfusion barrels were measured previously using gas chromatography (Mihic et al., 1994; Joo et al., 2001). The highest concentration of ISO in the undiluted aqueous solution was 2500 μm. The potency of volatile anesthetics is most often described as the minimum alveolar concentration (MAC) that prevents movement in response to a noxious stimulus in 50% of subjects (Eger et al., 1965). The MAC equivalent for ISO in mice was calculated using a formula that adjusted the in vivo MAC value to room temperature (Joo et al., 2001). The aqueous-phase concentration of ISO equivalent to MAC for in vitro studies at room temperature is ∼253-280 μm (Franks and Lieb, 1993, 1998).
Heterologous expression. Human embryonic kidney (HEK) 293 cells were transiently transfected with cDNAs encoding human GABAAR α5 or α1, β3, and γ2L subunit isoforms (1:1:1 ratio) using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). All cDNAs were subcloned into the mammalian expression vectors pCDM8 or pcDNA3.1(+) for heterologous expression. HEK 293 cells were incubated for 24-48 hr before investigation. GABA-evoked currents were activated by applying GABA for 2 sec. ISO was preapplied for at least 30 sec before coapplication with GABA.
Data analysis and statistics. The amplitude of the tonic current and total charge transfer (Q) were measured as described previously (Wall and Usowicz, 1997; Bai et al., 2001). Drug efficacy was reported as percentage of control, where control equaled the amplitude of the tonic current revealed by bicuculline (BIC). Minimal tonic current was detected in α5-/- neurons (Caraiscos et al., 2004), so the ISO effects were reported as the change in Ihold. All results are reported as the mean ± SEM. Statistical significance was determined using a Kruskal-Wallis or one-way ANOVA followed by an appropriate post hoc test.
Results
Isoflurane selectively enhances tonic conductance
Hippocampal pyramidal neurons generated a persistent tonic inhibitory conductance and transient postsynaptic currents, as shown in Figure 1A. The GABAAR antagonist BIC (100 μm) inhibited miniature IPSCs (mIPSCs) and reduced the tonic current by 32.9 ± 5.6 pA (n = 19) (Fig. 1A). ISO caused a concentration-dependent increase in the tonic current, as evidenced by an inward shift in the baseline. Low (sub-MAC) concentrations of ISO (25 and 83.3 μm) selectively enhanced the tonic current to 149 and 197% of control, respectively, but failed to alter mIPSCs (p < 0.05) (Fig. 1, Table 1). ISO (250 μm) also increased the tonic current to 314% of control and caused a threefold increase in the duration of mIPSCs but reduced the frequency by 65% (Fig. 1B,C; Table 1). The differential ISO sensitivity of the tonic and synaptic conductances to ISO was further illustrated by comparing net charge transfer. ISO (25 μm) caused a 37-fold greater increase in charge transfer for the tonic current compared with time-averaged synaptic currents (Qtonic 33.7 ± 10.3 pC, n = 7vs Qsynaptic 0.4 ± 0.9, n = 6; p < 0.05). The relative increase in charge transfer was even greater for ISO (250 μm). Despite prolonging the time course of synaptic currents, ISO (250 μm) failed to increase synaptic charge transfer because the frequency of mIPSCs was reduced (Qtonic 117.1 ± 31.3 pC, n = 7vs Qsynaptic -1.2 ± 0.8 pC, n = 5; p < 0.05).
Effects of ISO on tonic and synaptic conductances in hippocampal neurons from Swiss White mice. A, Current recorded before and during the application of BIC (100 μm). The dashed line depicts the baseline current before drug application. BIC revealed a tonic current as evidenced by an outward shift in the baseline current and blocked the transient inward mIPSCs. The bottom traces are temporal expansions of segments showing abolishment of mIPSCs by BIC. B, ISO potentiated the tonic current in a concentration-dependent manner. C, The bar graph illustrates the percentage change in peak tonic current by ISO. The dashed line indicates 100% (*p < 0.05). D, Averaged mIPSCs from a single neuron show that 250 μm but not 25 or 83.3 μm ISO (data not shown) prolonged the decay. E, PEN inhibited mIPSCs but not ISO (250 μm)-enhanced tonic current. BIC (100 μm) inhibited the ISO-enhanced current.
Effects of ISO on mIPSCs from hippocampal Swiss White, WT, and α5−/− mice
To ensure that ISO (250 μm) generated an inward current by targeting tonic but not synaptic receptors, penicillin (PEN) (5 mm) and ISO were coapplied because PEN selectively inhibits mIPSCs (Yeung et al., 2003). PEN inhibited mIPSCs but failed to reduce the ISO-enhanced tonic current, whereas BIC reversibly inhibited the ISO-enhanced tonic current (Fig. 1E).
A 10-fold higher (clinically irrelevant) concentration of ISO (2500 μm) activated a large inward current and also completely inhibited the synaptic currents. The mIPSCs were no longer superimposed on the ISO-activated inward current, as shown in Figure 1B. These actions are consistent with reports that show that high concentrations of ISO reduce the vesicular release of neurotransmitter and have multiple actions on GABAARs, including direct gating and channel blockade (Banks and Pearce, 1999).
The α5 subunit is necessary for ISO (25 μm) modulation
To determine whether α5GABAARs are required for the exquisite sensitivity of the tonic conductance to ISO (25 μm), currents were recorded from α5-/- and WT hippocampal neurons (Fig. 1B,C). We also examined cortical neurons as a negative control because α5GABAARs are expressed in low levels in the cortex.
ISO (25 and 83.3 μm) activated an inward current in WT but not α5-/- neurons (25 μm: WT 13.6 ± 3.8 pA, n = 14 vs α5-/- 0.4 ± 2.0 pA, n = 25; 83.3 μm: WT 29.2 ± 5.5 pA, n = 3vs α5-/- 0 ± 0.3 pA, n = 3; p < 0.01) (Fig. 2A,B). Notably, the increase in tonic current was similar in neurons from WT and Swiss White mice. ISO (250 μm) activated an inward current that was multiple-fold greater in WT compared with α5-/- neurons (WT 42.7 ± 8.8 pA, n = 14 vs α5-/- 13.4 ± 4.4 pA, n = 25; p < 0.01) (Fig. 2A,B). ISO (250 μm) slowed the rise time but reduced the frequency of mIPSCs in WT and α5-/- neurons, as shown in neurons from Swiss White mice (Table 1, Fig. 2C). No differences were detected in the amplitude, time course, or frequency of mIPSCs for WT and α5-/- neurons in the absence or presence of ISO (Table 1). ISO (2500 μm) abolished mIPSCs and generated a large inward current in both α5-/- and WT neurons (224.8 ± 46.0 pA, n = 10 vs 210.0 ± 27.1 pA, n = 19, respectively). These results suggest that the relative contribution of α5GABAARs to the macroscopic current activated by ISO (2500 μm) was obscured by activation of other GABAAR subtypes.
Effects of ISO on tonic and synaptic conductances in WT and α5-/- neurons. A, The dashed gray line depicts the baseline current before drug application. The BIC-sensitive tonic current was also reduced in α5-/- compared with WT neurons (5.9 ± 0.7 pA, n = 26 vs 24.6 ± 7.7 pA, n = 14; p < 0.01). B, The bar graph illustrates the reduced effect of ISO (25 and 83.3 μm) on the holding current in hippocampal neurons from WT (black) and α5-/- (gray) mice (*p < 0.01). In cortical neurons that express low levels of the α5 subunit, ISO caused similar changes in the holding current. Error bars represent SEM. C, ISO, 250 μm but not 25 μm, prolongs mIPSCs in WT and α5-/- neurons.
The application of BIC to cortical neurons from WT and α5-/- mice failed to reveal a tonic conductance (3.6 ± 0.8 pA, n = 7 vs 4.6 ± 1.0 pA, n = 14, respectively), as reported previously (Caraiscos et al., 2004). ISO (25 and 250 μm) activated a small inward current that was similar in amplitude in WT and α5-/- neurons (Fig. 2B). The current activated by ISO (2500 μm) was not different in α5-/- and WT cortical neurons (69.0 ± 12.1 pA, n = 7 vs 115.9 ± 20.1 pA, n = 12).
The α5 subunit confers increased ISO efficacy compared with the α1 subunit
To examine the specific role of α5 subunits in conferring sensitivity to low concentrations of ISO, currents were recorded from human α5β3γ2L GABAARs. We also examined α1β3γ2L GABAARs because the α1 subunit is widely expressed in the brain (Fritschy and Mohler, 1995). In contrast, the α5 subunit is localized primarily to extrasynaptic regions of hippocampal neurons (Brunig et al., 2002). Currents were evoked by concentrations of GABA intended to mimic the saturating (600 μm) and subsaturating (3 μm) concentrations of neurotransmitter present in the synaptic cleft and extracellular space, respectively (Lerma et al., 1986). The same GABA concentrations were used for α5β3γ2L and α1β3γ2L GABAARs because the potency of GABA is similar for both receptors (Caraiscos et al., 2004).
ISO (25 and 83.3 μm) potentiated the current evoked by GABA (3 μm) from α5β3γ2L GABAARs to 131 and 162% of control, respectively (p < 0.05) (Fig. 3A, B). Low concentrations of ISO caused a similar potentiation of the tonic conductance in neurons from Swiss White mice and GABA-evoked current from α5β3γ2L GABAARs. In contrast, the enhancement by ISO (25 μm) was not seen for α1β3γ2L GABAARs, as shown by the concentration-response plots (Fig. 3A,B). The maximal enhancement of both α5β3γ2L and α1β3γ2L GABAARs occurred at ISO (250 μm), although the increase was 1.4-fold greater for α5β3γ2L GABAARs. High concentrations of ISO directly gate GABAARs when applied in the absence of GABA. To determine whether the maximum current evoked by ISO (250 μm) was attributed, in part, to direct gating of GABAARs, we measured currents evoked by ISO (250 μm) alone. The amplitude of the ISO-evoked current was considerably less than the GABA-evoked (3 μm) current from α5β3γ2L (17.3 ± 5.8%; n = 5) and α1β3γ2L GABAARs (12.8 ± 2.1%; n = 4). This result indicated that direct gating failed to account for the large inward current activated by the coapplication of ISO (≥250 μm) and GABA.
ISO modulates recombinantα5β3γ2L and α1β3γ2L GABAARs. A, Dual effects of ISO (25 and 250 μm) on GABA (3 μm)-evoked currents recorded from α5β3γ2L and α1β3γ2L GABAARs. B, Concentration-response relationships for ISO enhancement of GABA-evoked current. Data points represent the mean ± SEM. ISO caused a greater increase in current recorded from α5GABAARs compared with α1GABAARs (p < 0.05). C, ISO (250 μm) had a minimal effect on GABA (600 μm)-evoked current, whereas (2500 μm) completely inhibited current from α1β2γ2L but not α5β2γ2L GABAARs (data not shown).
To mimic the concentration of GABA in the synaptic cleft, GABA (600 μm) was applied to α5β3γ2L and α1β3γ2L GABAARs. ISO (250 μm) modestly reduced the current from α5β3γ2L and α1β3γ2L to 84.0 ± 3.5% (n = 6) and 80.4 ± 5.4% (n = 7) of control, respectively (Fig. 3C). Consistent with complete inhibition of the mIPSCs, ISO (2500 μm) blocked the GABA-evoked current from α1β3γ2L GABAARs (1.0 ± 1.2% of control; n = 5; p < 0.05). In contrast, the current from α5β3γ2L GABAARs was reduced to 38.5 ± 11.5% of control (n = 5; p < 0.05) (Fig. 3C).
Discussion
Many neurodepressive drugs are thought to reduce neuronal excitability by increasing inhibitory synaptic transmission. However, a picture is emerging of a family of tonic GABAARs that “fine tune” neuronal excitability by regulating a conductance leak (Semyanov et al., 2004). Here, we report that low concentrations of the volatile anesthetic ISO (25 μm) selectively enhance the tonic but not synaptic current in hippocampal pyramidal neurons. The α5 subunit is necessary for enhancement by ISO (25 and 83.3 μm) because potentiation is absent in α5-/- neurons. Furthermore, studies using recombinant human α5β3γ2L and α1β3γ2L GABAARs show that the α5 subunit confers sensitivity to these low concentrations of ISO.
Our results are consistent with previous reports that show relatively high concentrations of volatile anesthetics are required to enhance inhibitory synaptic transmission in hippocampal neurons. Other investigators report that the lowest concentration of ISO that prolongs inhibitory synaptic currents in cultured rat hippocampal neurons and CA1 pyramidal neurons in hippocampal slices is ∼100 μm with a half-maximal effective concentration of 320 μm (Jones and Harrison, 1993; Banks and Pearce, 1999). The 2.5-fold enhancement of spontaneous IPSCs in CA1 pyramidal neurons by ISO (300 μm) (Banks and Pearce, 1999) is similar to the prolongation of mIPSCS by ISO (250 μm) we observed. These results indicate that a considerably greater ISO concentration is required to potentiate the synaptic compared with tonic current.
The α subunit isoform influences the sensitivity of GABAARs to halogenated ethers, including ISO (Jenkins et al., 2001). Positive modulation of GABAAR activity has been attributed to interactions of anesthetics with specific amino acid residues in transmembrane domains 2 and 3 of the α subunit (Mihic et al., 1997; Krasowski et al., 1998; Nishikawa et al., 2002). Mutations of Ser-270 or Ala-291 in the α subunit result in recombinant GABAARs that are insensitive to clinically relevant concentrations of ISO (Krasowski et al., 1998). Amino acid residues that contribute to the putative anesthetic binding cavity are conserved across all α isoforms, suggesting that the differential effects of ISO on α5GABAARs compared with α1GABAARs result from interactions with yet unidentified components of the binding cavity or transduction pathways. The structural basis for the enhanced sensitivity of α5GABAARs to ISO warrants additional investigation.
Our results support a growing body of literature that shows structurally diverse GABAergic drugs preferentially enhance a tonic conductance. We previously showed that the intravenous anesthetic propofol caused a greater increase in the tonic compared with synaptic conductance (Bai et al., 2001). This increase in tonic conductance by propofol reduces neuronal excitability (Bieda and MacIver, 2004). Until we determined that α5GABAARs generated the tonic current (Caraiscos et al., 2004), it was impossible to discern whether the pharmacology of the tonic conductance resulted from subunit composition and/or conditions of receptor activation. The efficacy of ISO is greater for α5GABAARs compared with the α1GABAARs, and the increase in GABAAR activity by anesthetics is greatest when receptors are activated by low GABA concentrations (Harris et al., 1995). Thus, both subunit composition and low GABA occupancy contribute to the preferential enhancement of the tonic conductance.
Low concentrations of ISO suppress learning in humans and animals. The steady-state concentration of ISO that impairs memory in rats is estimated to be ∼25% of MAC (Dutton et al., 2001). The neural correlates underlying complex behavioral phenomena including learning and memory remain unknown. α5GABAARs may play an important role in cognitive function because mice lacking the α5 subunit display enhanced learning and memory for hippocampal-dependent behaviors (Collinson et al., 2002; Crestani et al., 2002). In light of our findings that α5-/- neurons appear to be insensitive to low concentrations of ISO (83.3 μm; 33% MAC), we speculate that modulation of α5GABAARs in hippocampal neurons contributes to the amnestic properties of volatile anesthetics.
Footnotes
This research was supported by grants from the Canadian Institutes of Health Research (V.B.C., J.G.N., J.F.M., and B.A.O), a Career Scientist Award (B.A.O.), and the Canadian Anesthesiologists' Society (K.E.Y.-T.). We thank Lidia Brandes and Elzbieta Czerwinska for technical support.
Correspondence should be addressed to Dr. Beverley A. Orser, Department of Physiology, Room 3318, Medical Sciences Building, 1 King's College Circle, Toronto, Ontario M5S 1A8, Canada. Email: beverley.orser{at}utoronto.ca.
Copyright © 2004 Society for Neuroscience 0270-6474/04/248454-05$15.00/0