Nonhalogenated Alkanes Cyclopropane and Butane Affect Neurotransmitter-gated Ion Channel and G-protein–coupled Receptors: Differential Actions on GABAAand Glycine Receptors

These studies were approved by the Animal Care and Use Committee of the University of Texas (Austin, Texas).

The cDNAs encoding the following receptor subunits were used for nuclear injections: human wild-type α¹or α¹(S267C) glycine (subcloned in pCIS2 vector) 10; wild-type human α¹(subcloned in pBK-CMV N/B-200 vector); α², β¹, γ^2S, and α²mutants (S270X) (subcloned in pCIS2 vector) and β²(subcloned in pCDM8 vector) GABA^A11; human NR1 and NR2A NMDA (subcloned in pcDNA Amp vector) 12; and human 5-HT^3A(subcloned in pBK-CMV N/B-200 vector). 13The cDNAs of rat GluR1 and GluR2 AMPA receptor subunits (subcloned in pBluescript SK⁻vector) 14; the cDNAs of rat α⁴and β²nACh receptor subunits (subcloned in pSP64 and pSP65 vectors, respectively) 15; the cDNAs of rat GIRK1 16and mouse GIRK2 17(subcloned in pBluescript II KS⁻vector); and the cDNA of rat M¹receptor (subcloned in pGEM vector) 18were used for cRNA synthesis. In vivo  transcripts were prepared using the mCAP Capping Kit (Stratagene, La Jolla, CA).

Isolation of Xenopus laevis  oocytes was conducted as described previously. 19Isolated oocytes were placed in modified Barth's saline (MBS) containing 88 mm NaCl, 1 mm KCl, 2.4 mm NaHCO³, 0.82 mm MgSO⁴, 0.91 mm CaCl², 0.33 mm Ca(NO³)², and 10 mm N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES) adjusted to pH 7.5. Single cDNAs or combinations of the cDNAs encoding the following receptor subunits were injected into the animal poles of oocytes by a blind method 20: the α¹or α¹(S267C) glycine receptor subunit cDNA (1 ng/ 30 nl); the α¹, β², and γ^2SGABA^Areceptor subunits cDNAs (2 ng/30 nl in a 1:1:2 molar ratio); the α²(wild-type or mutants) and β¹GABA^Areceptor subunits cDNAs (1.5 ng/30 nl in a 1:1 molar ratio); the NR1 and NR2A NMDA receptor subunits cDNAs (1.5 ng/30 nl in a 1:1 molar ratio); the 5-HT^3Areceptor subunit cDNA (1.5 ng/30 nl). The GluR1 and GluR2 AMPA receptor subunits cRNAs, the α⁴and β²nACh receptor subunits cRNAs, the GIRK1 and GIRK2 receptor subunits cRNAs (30 ng/30 nl in a 1:1 molar ratio), or M¹receptor subunit cRNA (30 ng/30 nl) were injected into cytoplasm of oocytes. The injected oocytes were singly placed in Corning cell wells (Corning Glass Works, Corning, NY) containing incubation medium (sterile MBS supplemented with 10 mg/l streptomycin, 100,000 U/l penicillin, 50 mg/l gentamycin, 90 mg/l theophylline, and 220 mg/l pyruvate) and incubated at 15–19°C. One to 4 days after injection, the oocytes were used in electrophysiological recording. 19 

Oocytes expressing GABA^A, glycine, AMPA, or M¹receptors were placed in a rectangular chamber (∼100-μl volume) and perfused (2 ml/min) with MBS. Oocytes expressing NMDA receptors were perfused with Ba²⁺Ringer's solution (115 mm NaCl, 2.5 mm KCl, 1.8 mm BaCl², and 10 mm HEPES adjusted to pH7.4) to minimize the effects of secondarily activated Ca²⁺-dependent Cl⁻currents, and oocytes expressing nACh receptors were perfused with Ba²⁺Ringer's solution containing 1 μm atropine sulfate. For the 5-HT^3Areceptors, oocytes were perfused with low-Ca²⁺Ringer's solution (115 mm NaCl, 2.5 mm KCl, 0.18 mm CaCl², and 10 mm HEPES adjusted to pH7.4) to decrease inhibitory effect of calcium ion. The animal poles of oocytes were impaled with two glass electrodes (0.5–10 Ω) filled with 3 m KCl, and the oocytes were voltage clamped at −70 mV by using a Warner Instruments model OC-752B oocyte clamp (Hamden, CT). Glycine, GABA, kainic acid (for AMPA receptors), or ACh (for M¹receptor) dissolved in MBS was applied to the oocytes for 20 or 30 s to reach a maximal response. Likewise, L-glutamate plus 10 μm glycine (for NMDA receptors) or ACh (for nACh receptor) dissolved in Ba²⁺Ringer's solution was applied to the oocytes for 20 s, and 5-HT in low-Ca²⁺Ringer's solution was applied for 30 s. To study the effects on both wild-type and mutant GABA^Aor glycine receptors, the experiments were performed at EC⁵of agonist (concentrations that produced 5% of the maximal currents produced by 1 mm of glycine or GABA). To study the effects on the NMDA, AMPA, nACh, M¹, or 5-HT^3Areceptors, the experiments were performed at the half-maximal effective concentration (EC⁵⁰) of agonists. Based on the concentration–response relation from our previous work, 4,21,22which was performed under the same conditions as the current study (receptor: EC⁵⁰, Hill coefficient; α¹glycine: 206 ± 18 μm, 2.1 ± 0.1; α¹β²γ^2SGABA^A: 94 ± 3 μm, 1.2 ± 0.04; NR1/NR2A: 2.1 ± 0.3 μm, 1.3 ± 0.1; GluR1/GluR2: 91 ± 4 μm, 1.4 ± 0.1; 5-HT^3A: 1.8 ± 0.2 μm, 2.1 ± 0.1; α⁴β²nACh: 60 ± 3 μm, 0.9 ± 0.1; M¹ACh: 0.9 μm, 1.1), we used agonist concentrations to obtain EC⁵or EC⁵⁰for each receptor as follows: 30–60 μm glycine for wild-type α¹and α¹(S267C) glycine receptors, 10–30 μm GABA for α¹β²γ^2SGABA^A, and 1–2 μm GABA for α²β¹GABA^Areceptors; 60 μm ACh for nACh receptors; 1.5–2 μm 5-HT for 5-HT^3Areceptors; 2–3 μm L-glutamate plus 10 μm glycine for NMDA receptors; 100 μm kainic acid for AMPA receptors; and 1 μm ACh for M¹receptors. To obtain a control response, the agonists were repeatedly applied until a consistent response was observed. A 5- to 20-min washout period was allowed between drug applications.

Each solution (20 ml) in a closed vial was bubbled with 100% cyclopropane or butane for at least 10 min at room temperature to obtain a saturated solution. The saturated solution was pumped into the rectangular chamber via  a roller pump (ColeParmer Instrument Co., Chicago, IL) through 18-gauge polyethylene tubing. The concentrations of cyclopropane and butane in the chamber were quantified by gas chromatography, and the values were 0.46 ± 0.03 atm (n = 3) and 0.29 ± 0.01 atm (n = 3), respectively. These concentrations correspond to 5 minimum alveolar concentration (MAC) of cyclopropane 23and 1 MAC of butane. 24To test the effects of 1 and 2 MAC of cyclopropane, the saturated solutions were diluted with cyclopropane-free solution immediately before application. Cyclopropane or butane was preapplied for 1 min before being coapplied with agonists, unless specified. Effects of anesthetics were expressed as the fraction of control responses, which were calculated by averaging the control responses before and after application of anesthetics. To rule out the involvement of hypoxia, the effect of a perfusion solution bubbled with 100% nitrogen was evaluated in a previous study; replacement of oxygen by nitrogen did not alter the function of several ligand-gated ion channels. 21 

To address the mechanism of cyclopropane action on the α¹glycine receptor, we examined the glycine concentration–response relation in the presence or absence of cyclopropane, butane, or isoflurane. For other receptors, such as nACh, NMDA, and 5-HT^3A, we examined the effects of 5 MAC cyclopropane on the maximal response to agonists. According to the concentration–response relations obtained from our previous work, 21,221 mm ACh, 100 μm L-glutamate plus 10 μm glycine, or 30 μm 5-HT were applied to obtain maximal responses.

Experiments with the GIRK1/2 channels were performed with a high potassium (hK) solution containing 2 mm NaCl, 96 mm KCl, 1 mm MgCl², 1.5 mm CaCl², and 5 mm HEPES adjusted to pH 7.5. Initially, oocytes expressing GIRK1/2 were bathed in MBS and changed in hK solution. Because basal current obtained in physiologic buffer (MBS) is small, hK solution was used to reverse the driving force of the channel and provide a large inward current, as reported previously. 5After stable responses were established, the anesthetics were applied in the hK solution for 2 min before returning to the anesthetic-free hK solution. Effect of the anesthetics was expressed as the fraction of hK responses. All experiments were performed at room temperature (23°C).

Our previous studies showed that serine at position 267 located between transmembrane domains 2 and 3 of the glycine α¹subunit and the corresponding amino acids of GABA^Aα and β subunits are critical for the positive modulation of the receptor functions by several anesthetics and ethanol. 25–27Accordingly, we tested the effect of cyclopropane on the homomeric α¹(S267C) mutant glycine receptor in which serine at position 267 is substituted with cysteine. Also, we studied effects of cyclopropane on α²mutant GABA^Areceptors in which the corresponding sS270 was substituted with amino acids with larger volume of the side chain, i.e. , isoleucine (S270I), asparagine (S270N), threonine (S270T), tyrosine (S270Y), and tryptophan (S270W). The mutated glycine α¹(S267C) subunit and GABA^Aα²(S270X) subunits were constructed by the method of site-directed mutagenesis as described previously. 28The glycine α¹(S267C) subunit and GABA^Aα²(S270X) subunit were confirmed by DNA sequencing.

Xenopus laevis  female frogs were purchased from Xenopus Express (Homosassa, FL). Cyclopropane, butane, glycine, L-glutamate, kainic acid, acetylcholine chloride, and 5-hydroxytryptamine hydrochloride were obtained from Sigma Aldrich (St. Louis, MO). GABA was obtained from Research Biochemical International (Natick, MA).

Data are represented as mean ± SEM. Statistical analysis was performed by one-way analysis of variance for multiple comparisons and unpaired Student t  test for comparisons between two groups. Differences were considered as statistically significant at a P  value less than 0.05. The values of the EC⁵⁰and the Hill coefficient were calculated by nonlinear regression using GraphPad Prism software version 3.0 (GraphPad Inc., San Diego, CA). Concentration–response data for glycine were fitted to the equation)

where I represents the current, I^maxrepresents the maximal current, A represents the agonist (glycine) concentration, and n represents the Hill coefficient. Molecular volumes of anesthetics and side chain of amino acid were calculated with Spartan 5.0 software (Wavefunction Inc., San Diego, CA).

In oocytes expressing GABA^Aor glycine receptors, inward chloride currents were observed in response to the applications of agonists, and effects of anesthetics were studied by coapplication with agonists (fig. 1). For α¹β²γ^2SGABA^Areceptors, cyclopropane showed little effect at 1 or 2 MAC (fig. 2A). Even at 5 MAC, it only modestly potentiated the receptor (18 ± 2%). To test subunit specificity, we also used the α¹β²GABA^Areceptor and cyclopropane potentiated the α²β¹GABA^Areceptor to a similar extent as the α¹β²γ^2SGABA^Areceptor (16%± 2%). Thus, it appears that the γ subunit does not influence cyclopropane action on GABA^Areceptors. On the other hand, cyclopropane at 1, 2, and 5 MAC significantly potentiated the current responses of the α¹glycine receptors by 39, 62, and 161%, respectively. Butane was tested at 1 MAC, which is the highest concentration possible in our system. Butane significantly potentiated the α¹β²GABA^Areceptors by 13% and α¹glycine receptors by 64% (fig. 2B). We compared cyclopropane with isoflurane and halothane actions on glycine receptors. Isoflurane at 1, 2, and 5 MAC potentiated the current responses of the α¹glycine receptors by 101, 209, and 484%, respectively. Similarly, halothane potentiated by 139, 253, and 534% (1, 2, and 5 MAC, respectively). Isoflurane and halothane seem to be roughly three times more effective than cyclopropane (fig. 3).

Cyclopropane and butane shifted the glycine concentration–response curve leftward as was seen for isoflurane (fig. 4). In the experiments with cyclopropane, nonlinear regression analysis of the concentration response curves yielded the glycine EC⁵⁰value of 118 ± 6 μm for control and 69 ± 5 μm for cyclopropane (P < 0.05), and the Hill coefficients for control and cyclopropane were 2.0 and 2.3, respectively. In the experiments with butane, the glycine EC⁵⁰value of 142 ± 6 μm for control and 109 ± 4 μm for butane (P < 0.05) and the Hill coefficients both for control and for cyclopropane were 1.6. In the experiments with isoflurane, we obtained a glycine EC⁵⁰value of 126 ± 6 μm for control and 31 ± 8 μm for isoflurane (P < 0.05), and the Hill coefficients for control and isoflurane were 1.8 and 1.6, respectively. These results suggest that both cyclopropane and butane increase the apparent affinity for glycine as do other volatile anesthetics. The enhancing effect of cyclopropane was completely abolished in the glycine α¹(S267C) receptor (−8%± 4%;fig. 2A), indicating that cyclopropane interacts with the putative binding pocket composed of serine at position 267 as do ethanol and anesthetics including isoflurane and enflurane. 25–27 

For the experiments testing the nACh, glutamate, or 5-HT^3Areceptors, oocytes expressing the α⁴β²nACh, NR1/NR2A NMDA, GluR1/GluR2 AMPA, or 5-HT^3Areceptors represented inward cation currents in response to agonists (fig. 1). Cyclopropane inhibited the nACh receptor in a concentration-dependent manner (fig. 5A). One MAC cyclopropane markedly suppressed the nACh receptor function by 70 ± 4%. Similarly, cyclopropane concentration-dependently inhibited the NMDA and AMPA glutamate receptors. These receptors were significantly inhibited by 1 MAC cyclopropane (NMDA, −29 ± 5%; AMPA, −51 ± 5%). The 5-HT^3Areceptors were less sensitive than the nACh and glutamate receptors and slightly suppressed by 2 MAC. At concentrations of agonists that produce maximal responses (EC¹⁰⁰), the inhibitory effect of 5 MAC cyclopropane on the 5-HT^3Areceptors was not changed (EC⁵⁰, 23 ± 4%; EC¹⁰⁰, 17 ± 4%). On the other hand, the inhibitory effects of cyclopropane on the NMDA and the nACh receptors were decreased at agonist EC¹⁰⁰(NMDA, 5 MAC: EC⁵⁰, −66 ± 4%; EC¹⁰⁰, −33 ± 3%; nACh, 1 MAC: EC⁵⁰, −70 ± 1%; EC¹⁰⁰, −54 ± 3%). Cyclopropane enhanced the function of the GIRK1/2 channels (fig. 5A). One MAC cyclopropane slightly but significantly increased potassium current by 9 ± 1%. Xenopus  oocyte expression system is also well characterized for the study of G-protein–coupled receptors. Activation of M¹receptors expressed in oocytes results in activation of phospholipase C, mobilization of calcium stores, and activation of an endogenous Ca²⁺-dependent Cl⁻current. Thus, inward currents are observed in response to ACh. 4Cyclopropane had little effect on the M¹receptors even at 5 MAC (−5%± 4%;fig. 5A).

Of interest, 1 MAC butane provided similar effects to those of 1 MAC cyclopropane on the most of the receptors studied (nACh receptors, −72%± 4%; 5-HT^3Areceptors, −3%± 2%; NMDA receptors, −30%± 6%; GIRK1/2 channels, 7%± 1%; M¹receptors, 2%± 3%;fig. 5B). In contrast, butane had little effect on the AMPA receptors (−2%± 3%).

To test the idea that the small size of cyclopropane precludes its binding to serine 270 in GABA α²subunit, we tested five α²mutant GABA^Areceptors coexpressed with β¹subunit with increasing volumes of amino acids at position 270. EC⁵values were determined for each of the GABA^Amutant receptors, which were 0.2–0.5 μm for α²(S270T)β¹, 0.4–0.8 μm for α²(S270N)β¹, 0.5–0.8 μm for α²(S270I)β¹, 0.3–0.7 μm for α²(S270Y)β¹, and 0.02–0.075 μm for α²(S270W)β¹receptor. Cyclopropane (5 MAC) enhanced four of five mutants to a similar extent as α^2β1wild type, and did not affect α²(S270W)β¹receptors (fig. 6).

Neither cyclopropane (1–5 MAC) nor butane (1 MAC) produced current in any receptor studied (in the absence of neurotransmitter). Isoflurane as well as halothane (up to 5 MAC) had no effect on basal current of the glycine receptors.

In this study, we found marked difference in the effects of nonhalogenated alkane anesthetics between GABA^Aand glycine receptors, i.e. , they enhanced glycine receptors but not GABA^Areceptors at clinical concentrations. The lack of the effect on GABA^Areceptors is consistent with a previous report. 9It has been shown that most volatile anesthetics markedly enhance both GABA^Aand glycine receptors, 29–31although some anesthetics, such as pentobarbital, propofol, and etomidate, are more effective on GABA^Areceptors than glycine receptors. 29,32,33Neither GABA^Areceptors nor glycine receptors are affected by anesthetics such as ketamine. 34In this context, nonhalogenated alkane anesthetics cyclopropane and butane belong to a distinct category. This result leads to the question of the mechanism underlying the differential effects of these anesthetics on GABA^Aand glycine receptors. Our results with glycine α¹(S267C) receptors suggest that cyclopropane interacts with the putative anesthetic and ethanol binding pocket composed of serine 267. 25–27A small change of volume from serine (volume of the residue: 53 ų) to cysteine (65 ų) abolished the potentiation by cyclopropane. The molecular volume of cyclopropane and butane are 70 and 99 ų, respectively, and are relatively small compounds compared to other representative anesthetics, such as isoflurane (144 ų) and halothane (110 ų). We hypothesized that the volume of the putative binding pocket of GABA^Areceptor may be larger than that of glycine receptor and that because molecular volumes of cyclopropane and butane are small, they might not be able to interact with GABA^Areceptors as tightly as glycine receptors and may not exert an effect strong enough to cause conformational change on the GABA^Areceptors. This “critical volume” hypothesis for anesthetic actions on GABA^Areceptors is based on the work of Jenkins et al. , 35who proposed that the volume of the binding cavity is 250–370 ų. Therefore, we attempted to augment the enhancing effect of cyclopropane on the GABA^Areceptor by a substitution of serine at position 270 in the α²subunit with larger amino acids, including threonine (70 ų), asparagine (71 ų), isoleucine (100 ų), tyrosine (116 ų), and tryptophan (137 ų). Contrary to our hypothesis, increasing the volume of the side chain did not produce significant potentiation. Conversely, the potentiation was switched to inhibition in the α²(S270W)β¹receptors. In a previous study, 35a small anesthetic, chloroform (90 ų), provided much greater enhancement when the volume of amino acid at position 270 was increased by approximately 20–30 ų. Because cyclopropane is 20 ųsmaller than chloroform, we predicted that an increase of 40–50 ųwould augment the action of cyclopropane. However, we increased the volume of the side chains at position 270 by 17, 18, 47, 63, and 84 ų, with no evidence of enhancement of action. As a result, the small volume of nonhalogenated alkane anesthetics does not appear to be the only determinant of the insensitivity of the GABA^Areceptor. Because the putative binding pockets on the glycine and GABA^Areceptors are supposed to be lined with many different amino acids, the binding environment for cyclopropane may be different between the glycine and GABA^Areceptors. It also should be noted that neither cyclopropane nor butane has a dipole moment. Perhaps polarity is more important for the binding of anesthetics to GABA^Areceptors than glycine receptors.

Our results showed that cyclopropane and butane affect other receptors. These anesthetics inhibited the nACh and NMDA receptors, slightly potentiated the GIRK1/2 channels, and did not change the 5-HT^3Aand M¹receptors. For the AMPA receptors, only cyclopropane produced a marked inhibition. GABA^Areceptors are thought to be a primary target of anesthetics because most volatile and nonvolatile anesthetics augment the channel activity at clinical concentrations. In terms of the nonhalogenated alkanes cyclopropane and butane, however, this is not the case. Based on sensitivity to clinical concentrations of cyclopropane and butane, glycine, nACh, and NMDA receptors are likely candidates. The nACh receptors are the most sensitive among the receptors tested in this study. Recently, nACh receptors were proposed as targets for anesthetics because volatile anesthetics and some intravenous anesthetics, such as thiopental, inhibit the function of nACh receptors. 36,37However, nACh receptors appear not to mediate immobility in vivo , based on a study using a nACh receptor antagonist. 38Furthermore, F6 (1,2-dichlorohexafluorocyclobutane), a nonimmobilizer, also inhibits nACh receptors expressed in Xenopus  oocytes. 39Thus, the inhibition of nACh receptors is not likely involved in immobility by anesthetics, although their role in other aspects of anesthesia remains to be elucidated.

Glycine receptors are the main inhibitory receptors in the spinal cord and brainstem, and volatile anesthetics enhance the function of these receptors. In this study, the glycine receptors were moderately potentiated by nonhalogenated alkane anesthetics. While the magnitude of the enhancing effect is less than for isoflurane or halothane, glycine receptors are probably target molecules. Indeed, this assumption is supported by in vivo  study using rats, which demonstrated that MAC of cyclopropane is increased by an intrathecal administration of strychnine, a glycine receptor antagonist. 40 

Glutamate plays a major role in synaptic excitation in the CNS and is critical for information storage in memory and learning. 41The NMDA receptors are known to mediate nociceptive neurotransmission in the CNS, and both the NMDA and AMPA receptors are important for memory. Inhibition of NMDA receptors is thought to be responsible for anesthetic actions of ketamine. 42Cyclopropane and butane moderately inhibited the NMDA receptors. The AMPA receptors were more strongly suppressed by cyclopropane rather than the NMDA receptors. In contrast, butane had little or no effect on the AMPA receptors, suggesting that the suppression of the AMPA receptors is not essential for the anesthesia by nonhalogenated alkane anesthetics. Given a similarity in effects of cyclopropane and butane on most of receptors, except the AMPA receptors, a comparison of behaviors produced in vivo  between these two anesthetics may elucidate which clinical manifestations can be explained by the AMPA receptors.

It is interesting that nonhalogenated alkane anesthetics potentiate GIRK channels, because volatile anesthetics inhibit the GIRK channels expressed in oocytes. 43Recently, the GIRK channels were shown to be potentiated by ethanol and to be important for ethanol actions, including analgesia. 5–7The potentiation by cyclopropane and butane may in part attribute to their analgesic actions. It is, however, unclear whether the GIRK channels are involved in immobility so far.

Previous studies showed that inhalational anesthetics have very similar effects on GABA^Aand glycine receptors and suggested a common site and mechanism of action for these two neurotransmitter-gated ion channels. 3,23The current work with cyclopropane and butane provides the first evidence that GABA^Aand glycine receptors can be differentially sensitive to some inhalational agents. Our mutation studies suggest a common site between transmembrane domains 2 and 3 that accommodates volatile anesthetics, including cyclopropane, but this site must differ subtly between GABA^Aand glycine receptors.

In summary, our findings suggest that the glycine and NMDA receptors but not GABA^Areceptors may contribute anesthesia (immobility) produced by nonhalogenated alkanes cyclopropane and butane.

The authors thank the laboratories quoted herein for kindly providing cDNAs. The authors also thank James R. Trudell, Ph.D. (Professor, Department of Anesthesia, Stanford University, Palo Alto, California), for calculating molecular volumes.