Anesthetics Inhibit Acetylcholine-promoted Guanine Nucleotide Exchange of Heterotrimeric G Proteins of Airway Smooth Muscle

After the study was approved by the Institutional Animal Care and Use Committee (Mayo Foundation, Rochester, Minnesota), porcine tracheas were obtained from an abattoir or by euthanasia of research animals. In our experience, we have found no physiologic difference in tracheal smooth muscle obtained from these two sources (unpublished observations). For tissue obtained from research animals, euthanasia was accomplished by intramuscular injection of telazol (Ft. Dodge Animal Health, Inc., Ft. Dodge, IA) (10 ml/kg) and xylazine (6 mg/kg) and intravenous injection of 400–600 mg phenobarbital sodium, followed by exsanguination by bilateral transection of the carotid arteries. For studies using tissue obtained form both sources, the extrathoracic tracheas were excised and immersed in chilled physiologic salt solution with a composition of 110.5 mm NaCl, 25.7 mm NaHCO³, 5.6 mm dextrose, 3.4 mm KCl, 2.4 mm CaCl², 1.2 mm KH²PO⁴, and 0.8 mm Mg²SO⁴. After removal of fat, connective tissue, and epithelium, tracheal smooth muscle was cut into strips (1.5 cm long × 0.25 cm wide), frozen in liquid nitrogen and stored at −70°C until its was used to prepare crude membranes for investigation.

A crude membrane fraction of ASM homogenate was prepared according to a modification of previously described methods.22,23Approximately 350 mg frozen tissue, the amount obtained from a single animal, was ground to a fine powder in liquid nitrogen using a mortar and pestle. The dry powder was suspended for 15 min in ice-cold lysis buffer composed of 20 mm HEPES (pH 8.0), 1 mm EDTA, 0.1 mm phenylmethysulfonyl fluoride, 10 μg/ml leupeptin, and 2 μg/ml aprotinin and then gently homogenized on ice with a Dounce tissue grinder (approximately 10–12 strokes). The homogenate was filtered through a 250-μm nylon filter (Small Parts, Inc., Miami Lakes, FL) and centrifuged at 87,000g  for 30 min (4°C). The pellet was washed with lysis buffer and then resuspended by gentle vortex in assay buffer composed of 50 mm Tris-HCl (pH 7.4), 2 mm EDTA, 100 mm NaCl, 4.8 mm MgCl², and 1 μm guanosine-5′-diphosphate (GDP), creating a crude membrane emulsion that was again filtered as described above. A portion of the crude membrane emulsion was solubilized in 6 ml NaOH, 0.1 N, and heated (30 min) for protein concentration determination by Lowry assay.24The homogenate was then diluted with assay buffer to a protein concentration of 2.5 mg/ml.

This assay, graphically depicted in figure 1, is based on the fundamental principle that the exchange of GTP for GDP at the Gα subunit of heterotrimeric G proteins causes the dissociation of this subunit from the membrane-delimited receptor and the Gβγ dimer into solution. In a system with preserved receptor–heterotrimeric G-protein coupling, the extent of this dissociation is promoted by the binding of an agonist to the receptor in the presence of GTP. In the current study, exogenous acetylcholine was used to stimulate muscarinic receptors, and a nonhydrolyzable, radioactive form of GTP, [³⁵S]GTPγS, was used as the reporter for the Gα subunit dissociation. The native, soluble Gα^q/11isotype was then immunoprecipitated from solution using rabbit polyclonal antiserum generated against native, recombinant rat brain Gα^q.

To perform the assay, Gα^q/11nucleotide exchange was initiated by the addition of 29 nm (final concentration) [³⁵S]GTPγS (specific activity 1.25 μCi/pmol) to the crude membrane emulsion (containing 125 μg protein) at 30°C. Reactions were terminated with 600 μl ice-cold immunoprecipitation buffer of the following composition: 50 mm Tris-HCl (pH 7.5), 20 mm MgCl², 150 mm NaCl, 2 μg/ml aprotinin, 0.5% (volume/volume) IGEPAL CA-630, 100 μm GDP, and 100 μm GTP. All the reaction tubes were then briefly vortex mixed and gently rotated for 30 min (4°C). Finally, the samples were centrifuged at 12,500g  for 15 min (4°C), and the supernatant containing soluble, native proteins was used for immunoprecipitation of [³⁵S]GTPγS-bound Gα^q/11(fig. 1).

For immunoprecipitation, rabbit anti-Gα^q/11antiserum was bound to protein A-agarose beads by incubating the beads in immunoprecipitation buffer (4°C) containing 1:200 (volume/volume) anti-Gα^q/11antiserum for at least 2 h before the assay (antibody-coated beads). Unbound antibody was removed by washing the beads twice with immunoprecipitation buffer; the antibody-coated beads were stored at 4°C until used for experiments. The supernatant of the assay was first precleared with normal rabbit serum (1:100 dilution) and 30 μl antibody-free protein A-agarose beads for 60 min (4°C) to reduce subsequent nonspecific protein binding to immune complexes or the agarose matrix. The beads were then pelleted by centrifugation at 3,260g  (10 min at 4°C), and the precleared samples were transferred into fresh tubes and incubated for 2 h (4°C) with 40 μl antibody-coated beads. The beads were then washed four times by repeated pelleting and centrifugation at 3,260g  (10 min at 4°C) followed by resuspension in immunoprecipitation buffer (1 ml). Finally, the washed beads were placed in 4 ml Ultimate Gold scintillation cocktail (Packard Bioscience, Meriden, CT), and radioactivity was quantified using a Beckman model LS6000IC liquid scintillation counter (Beckman, Palo Alto, CA). Background radioactivity measurements were determined by performing tandem experiments without protein and were less than 10% of the radioactivity of the basal Gα^q/11nucleotide exchange measurements. The amount of radioactivity above background measurements was taken to indicate the amount of Gα^q/11dissociated from the membrane into the soluble fraction due to the exchange of [³⁵S]GTPγS for GDP. Values were normalized to the amount of protein present in the membrane emulsion before preclearance.

Stock solutions of assay buffer with saturating concentrations of halothane were prepared by mixing halothane in the assay buffer by stirring overnight in a ground-glass flask.16These stocks were diluted with fresh assay buffer to achieve the desired concentration of halothane. Assay tubes were capped with Teflon-coated rubber stoppers (Alltech Associates, Inc., Deerfield, IL) immediately after the addition of halothane-containing solutions. Actual halothane concentrations in solution under assay conditions were measured by gas chromatography according to the method of Van Dyke and Wood.25Hexanol was added as appropriate directly to the assay buffer. We have verified in previous work using gas chromatography that this procedure provides concentrations of hexanol in aqueous solution expected on the basis of its density and molecular weight.7 

To determine the time course for the extent of exchange of [³⁵S]GTPγS for GDP at Gα^q/11(defined as Gα^q/11[³⁵S]GTPγS/GDP exchange), crude membrane samples were incubated without (basal exchange) or with (acetylcholine-promoted exchange) 100 μm acetylcholine. The reactions were terminated at 1, 3, 5, 10, and 20 min after initiating the assay reactions with [³⁵S]GTPγS. To determine the effect of acetylcholine concentration on Gα^q/11[³⁵S]GTPγS/GDP exchange, measurements were obtained in separate experiments using crude membrane samples incubated without (basal exchange) or with 10, 30, 60, and 100 μm acetylcholine, and the reactions were terminated at 10 min after initiating the assay reactions.

Assays were performed in the presence or absence of either halothane or 10 mm hexanol. The effects of these agents on Gα^q/11[³⁵S]GTPγS/GDP exchange were determined in separate experiments using samples incubated without (to assess effects on basal nucleotide exchange) or with 30 or 100 μm acetylcholine (to assess anesthetic effects on acetylcholine-promoted nucleotide exchange) present in the assay buffer. All reactions were terminated at 10 min after initiating the assay reactions. Hexanol, 10 mm, produces maximal functional effects on ASM and was chosen so that the current results could be compared to the results of our previous work.9The aqueous halothane concentration was 0.38 ± 0.02 mm, which did not vary significantly over the duration of an experiment (preliminary data not shown) and is a concentration within the range previously shown to inhibit ASM contraction.12,26,27Each condition was assayed in triplicate.

[³⁵S]GTPγS (specific activity 1 μCi/μl) was purchased from Amersham Biosciences (Piscataway, NJ). Rabbit polyclonal antiserum generated against recombinant rat brain Gα^qprotein and rabbit normal serum were purchased from Calbiochem (EMD Biosciences, Inc. Affiliate, San Diego, CA). Protein A-agarose beads were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The Lowry protein assay kits were purchased from Bio-Rad (Hercules, CA). All other chemicals were purchased from Sigma Chemical Company (St. Louis, MO).

Data are reported as mean ± sd. For time-course studies, a two-way repeated-measures analysis of variance was used to test the effect of time and acetylcholine on Gα^q/11nucleotide exchange. For the determination of the effects of acetylcholine, halothane, and hexanol on Gα^q/11nucleotide exchange, one- or two-way repeated-measures analysis of variance was used as appropriate. Post hoc  testing was performed using the Student-Newman-Keuls test. A value of P < 0.05 indicated statistical significance.

In the absence of acetylcholine (basal nucleotide exchange), there was a time-dependent increase in Gα^q/11[³⁵S]GTPγS/GDP exchange (fig. 2). At all time points, there was significantly more Gα^q/11[³⁵S]GTPγS/GDP exchange in samples treated with 100 μm acetylcholine than in those without acetylcholine, with a maximal difference reached at 20 min (35.9 ± 2.9 vs.  9.8 ± 1.2 fmol/mg protein, respectively). The increase in Gα^q/11[³⁵S]GTPγS/GDP exchange was concentration dependent compared with basal values (fig. 3).

Halothane significantly inhibited the increase in Gα^q/11[³⁵S]GTPγS/GDP exchange (assayed at 10 min) induced by 30 and 100 μm acetylcholine (14.2 ± 1.7 vs.  22.0 ± 2.4 and 22.1 ± 2.5 vs.  30.8 ± 4.2 fmol/mg protein, respectively) (fig. 4), i.e. , by 59% and 39%, respectively, when compared to basal values (fig. 4). Likewise, hexanol significantly inhibited the increase in Gα^q/11[³⁵S]GTPγS/GDP exchange induced by 30 μm acetylcholine from 27.3 ± 2.1 fmol/mg protein to 19.8 ± 1.9 fmol/mg protein, a 68% inhibition when compared to basal values (fig. 5). A similar tendency was observed during stimulation with 100 μm acetylcholine but did not reach statistical significance. Neither halothane nor hexanol had an effect on basal Gα^q/11[³⁵S]GTPγS/GDP exchange (i.e. , in the absence of acetylcholine) (figs. 4 and 5).

The major finding of this study is that both halothane and hexanol, anesthetics that inhibit ASM contraction,7,9,12,28inhibit acetylcholine-promoted Gα^q/11[³⁵S]GTPγS/GDP exchange without affecting basal, intrinsic nucleotide exchange in the absence of receptor stimulation. These data demonstrate in tissue and through direct biochemical assessment that some anesthetics inhibit receptor–heterotrimeric G-protein coupling. This mechanism could be responsible at least in part for the observed effects of anesthetics on numerous receptor–heterotrimeric G-protein-mediated cellular processes, including ASM relaxation.

Muscarinic receptors activate one or more subfamilies of heterotrimeric G proteins, which are comprised of α, β, and γ subunits. The Gα subunit contains a binding site for GTP and GDP and possesses GTPase (catalytic) activity. The Gβγ subunits are tightly associated and anchor the Gα subunit to the cytoplasmic surface of the cell membrane. In its resting state, the G protein exists as an inactive Gαβγ trimer, with GDP occupying the nucleotide binding site of the Gα subunit. The binding of an agonist to the receptor promotes GDP release and subsequent GTP (or [³⁵S]GTPγS) binding to the α subunit (i.e. , nucleotide exchange). This exchange of nucleotides triggers dissociation of the Gαβγ complex from the receptor and separation of the Gα subunit from the Gβγ dimer. The hydrolysis of bound GTP by the intrinsic GTPase activity of the Gα subunit permits reassociation of the subunits into a heterotrimer and terminates the activation of the effector. In ASM, this process increases both intracellular calcium concentration and the sensitivity of the contractile machinery to a given concentration of intracellular calcium (i.e. , calcium sensitivity), both of which induce smooth muscle contraction.29 

The heterotrimeric G proteins are classified according to the identity of the Gα subunit into four major subfamilies,30and within each subfamily, there are several isotypes. For example, the Gα^qsubfamily consists of at least five distinct proteins, including Gα^qitself, Gα¹¹, Gα¹⁴, and others. The role of different heterotrimer G-protein subfamilies in mediating ASM contraction is not completely delineated, but Gα^qsubfamily proteins, such as G^qand G¹¹, transduce receptor-mediated increases in both intracellular calcium concentration and calcium sensitivity in ASM.20 

There is considerable evidence that some anesthetics, including halothane and hexanol, can inhibit cellular processes activated by heterotrimeric G-protein–coupled receptors in ASM, producing the functional effect of ASM relaxation.8,9,12,26,27,31–34In a series of experiments in differentiated ASM tissue using pharmacologic probes, we suggested that this effect could be localized, at least in part, to the GPRC.9,27However, these measurements of cellular function (e.g. , intracellular calcium concentration and isometric force) did not permit direct assessment of anesthetic effects on the GPRC, because effects on more distal elements of the signaling pathways could not be excluded (e.g. , effects on calcium channels). A similar limitation applies to previous work inferring anesthetic effects on the GPRC from measurements of downstream effector function in other intact tissues or expression systems, such as Xenopus  oocytes.5,13–15,35Finally, studies of ligand binding to muscarinic receptors derived from various tissues provide indirect evidence of anesthetic effects on the GPRC, because receptor–heterotrimeric G-protein interactions determine ligand–receptor affinity.1,2However, this technique is indirect, which can lead to misinterpretation and erroneous conclusions.

Assessment of nucleotide exchange at the Gα subunit in cellular membrane preparations from specific tissues provides a direct measure of receptor–heterotrimeric G-protein coupling. Using the technique described in the current study, the exchange of GDP for GTP at a particular Gα protein can be measured, with subfamily specificity determined by the epitope to which the antibody is raised for the immunoprecipitation step. Most previous studies using these techniques have been conducted using experimental systems in which receptors and heterotrimeric G proteins are overexpressed in mammalian or insect cells22,23,36; one report using freshly dissociated gut smooth muscle cells has been published.37The current study is the first to demonstrate the use of this technique to measure agonist-promoted nucleotide exchange using crude membrane prepared from smooth muscle tissue.

In addition to the specificity of the antibody used for immunoprecipitation, the ability of the experimental techniques used in the current study to detect agonist-promoted heterotrimeric G-protein nucleotide exchange in a crude membrane preparation is limited by several factors. These include the magnitude of the intrinsic, basal nucleotide exchange, the amount of the endogenous Gα subunit isotype of interest expressed in the tissue, and the extent to which the Gα subunit isotype of interest is coupled to and dissociates from the membrane receptor and Gβγ dimer with agonist binding. In the case of Gα^q/11, the antibody used is highly specific for both the Gα^qand Gα¹¹subunits,36a finding confirmed by our preliminary data showing no cross-reactivity with Gαⁱor Gα^ssubfamily proteins (data not shown). The kinetics of basal, intrinsic nucleotide exchange for the Gα^q/11using either recombinant, pure protein38,39or crude membrane prepared from mammalian cells in which the receptor and the heterotrimer G-protein subunits have been enriched36is low. Although this was also true in the current study, the basal Gα^q/11nucleotide exchange was still sufficient to conduct a reliable assessment of a possible anesthetic effect, because the background radioactivity was only 10% or less of the radioactivity of this measurement. The low level of Gα^q/11nucleotide exchange observed in the absence of exogenous acetylcholine also optimized our ability to examine anesthetic effects on receptor–G-protein coupling, because the relative difference in the magnitude of the acetylcholine-promoted versus  basal Gα^q/11nucleotide exchange was approximately fourfold to fivefold (fig. 3). The magnitude of this difference is similar to that reported in previous work.36,38,39 

The observation that acetylcholine significantly increased Gα^q/11nucleotide exchange demonstrated functional coupling between the muscarinic receptor and Gα^q/11in porcine ASM. The time course for agonist-promoted nucleotide exchange measured in the current study was similar to that reported by others using a similar crude membrane preparation.37However, the rate of acetylcholine-promoted nucleotide exchange was slower than that anticipated based on kinetic measurements of other heterotrimeric G-protein–mediated signals obtained in intact, undisrupted cells or tissue, such as free calcium concentration. The explanation for this difference is most likely the loss of soluble, heterotrimeric G-protein regulatory proteins, known as GTPase-activating proteins, caused by cell membrane disruption in our preparation. These proteins, which are present in intact smooth muscle cells,40accelerate G-protein nucleotide exchange by approximately 200-fold and thus confer the speed needed for reliable intracellular signaling.41–43Nevertheless, the crude membrane preparation is a well-established model that provides direct, unambiguous measurement of membrane receptor coupling to heterotrimeric G proteins and thus is appropriate for examining the hypothesis proposed in the current study.

In the absence of receptor stimulation, neither halothane nor hexanol had an effect on basal, intrinsic Gα^q/11[³⁵S]GTPγS/GDP exchange (figs. 4 and 5). By contrast, Pentyala et al.  16found that halothane and other volatile anesthetics modulated the binding of guanine nucleotides to recombinant Gα subunits in aqueous solution, thereby inhibiting the exchange of GDP for GTPγS. They did not study Gα^qsubfamily proteins because nucleotide exchange is not detectable in these purified subunits, unlike in membrane preparations as demonstrated by the current and previous studies.44,45However, for reasons that we have not been able to elucidate, we have not been able to duplicate their findings on intrinsic, basal nucleotide exchange using either purified, recombinant Gαⁱ-1 protein or endogenous Gαⁱin a porcine ASM membrane preparation.17 

In contrast to basal measurements, both halothane and hexanol significantly inhibited acetylcholine-promoted [³⁵S]GTPγS/GDP exchange in concentrations that produce anesthesia in vivo  and ASM relaxation in vitro .7,32,46This effect is consistent with functional studies of heterotrimeric G-protein–mediated processes in cells3and expression systems.1,5,47In studies of ASM, we found in previous work that hexanol and halothane inhibited activation of downstream effectors induced by stimulation of muscarinic receptors with acetylcholine and by direct stimulation of heterotrimeric G proteins using tetrafluoroaluminate.9,27Tetrafluoroaluminate directly activates heterotrimeric G proteins by binding to the α subunit next to the β phosphate of GDP and mimicking the γ phosphate of GTP.48Because the magnitude of these effects was similar for both methods of activation, we interpreted these findings as demonstrating a direct interaction of anesthetics with the heterotrimer, acting to stabilize the heterotrimer and inhibit its dissociation. However, these interpretations must now be revised based on the current finding that the anesthetics did not affect intrinsic, basal Gα^q/11nucleotide exchange. We now propose that some anesthetics stabilize the entire GPRC, rather than just the heterotrimeric G protein, thereby attenuating muscarinic receptor coupling to the heterotrimeric G protein G^q/11.

The experimental techniques used in the current study can provide only a functional assessment of the interaction between the membrane delimited muscarinic receptor and the associated heterotrimeric G protein G^q/11. They cannot ascertain with which of the possible protein targets, either the muscarinic receptor or the G^q/11heterotrimer subunits, the anesthetic molecules interacted to produce the observed effects. For example, it is possible that Gα^q/11possesses an anesthetic binding region that is at its receptor binding domain, which could interfere with receptor coupling but have no effect on basal Gα^q/11nucleotide exchange. Another plausible interpretation of our data is that the anesthetic molecules interacted directly with the muscarinic receptor only, as previously demonstrated for the rhodopsin receptor,18,19thereby only interfering with the ability of acetylcholine to activate Gα^q/11nucleotide exchange. The techniques used in this study also cannot distinguish between binding of anesthetic molecules directly to a protein or at a hydrophobic binding site created at the interface between one of the possible protein targets and the lipid membrane. Determining the target sites (e.g. , protein, protein–protein interface, or lipid–protein interface) would involve the use of biophysical methods that directly assess binding interactions between the anesthetic and these possible targets. Nevertheless, regardless of the biophysical mechanisms, the current work provides compelling evidence for an anesthetic effect on muscarinic receptor–heterotrimeric G-protein coupling. The experiments in the current study in isolation do not exclude actions on other components of the heterotrimeric G-protein–mediated signaling pathway in situ , but our previous work suggests that such effects are not present, at least in ASM.31 

In summary, to our knowledge, these are the first data to demonstrate, using a direct measurement technique, that anesthetics inhibit receptor-promoted nucleotide exchange at the Gα subunit of heterotrimeric G proteins. The preponderance of previous work indicates that anesthetics bind to membrane-delimited receptors but not G proteins and have no effect on basal, intrinsic nucleotide exchange at Gα in the absence of receptor activation. Taken together with these previous observations, our findings suggest that halothane and hexanol interact either with the receptor itself or at the interface between receptor and G protein. Accordingly, these interactions prevent the ability of acetylcholine to promote Gα^q/11nucleotide exchange and its subsequent dissociation from the muscarinic receptor and the Gβγ dimer. These data are consistent with the ability of anesthetics to interfere with cellular processes mediated by heterotrimeric G proteins,1–5including our previous observations of anesthetic effects on muscarinic receptor–G-protein regulation of ASM contraction. These findings also have implications in other organ systems, including the central nervous system, given the ubiquitous nature of heterotrimeric G proteins in mediating signaling pathways and cell function.