| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
 
|
|
|
||
|
|
Anesth Analg 2007; 105:1675-1680 © 2007 International Anesthesia Research Society doi: 10.1213/01.ane.0000287818.46240.83
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
BACKGROUND: Inhaled anesthetics may affect proteins at the interface between membrane lipids and the surrounding aqueous phase. The underlying solution chemistry is not known. Because the hydrophobicity of nonpolar protein components importantly influences their conformation, we tested the hypothesis that isoflurane affects the solubility of two nonpolar compounds, methane and toluene, in saline.
METHODS: Using a serial dilution technique, we determined the saline:gas partition coefficients (PCs) of methane and toluene at 37°C in the absence of isoflurane and in the presence of approximately 1%, 5%, and 15% isoflurane. We also measured the effect on the vapor pressure of benzene produced by saturating benzene with either cyclopropane or chloroethane, anesthetics used in a previous study to demonstrate that their equilibration with benzene decreased the solubility of benzene in water.
RESULTS: Clinically relevant concentrations of isoflurane (1% and 5%) did not affect the saline:gas PC of methane and toluene, but 15%–20% isoflurane increased the PC of methane (P < 0.05) but not toluene. Saturating benzene with cyclopropane or chloroethane, decreased the vapor pressure of benzene in proportion to the amount of anesthetic dissolved in the benzene.
CONCLUSION: Isoflurane has a weak antihydrophobic effect at concentrations far above the clinically relevant range, and this effect is unlikely to explain how anesthetics act. A previous study, which found that cyclopropane and chloroethane decreased the solubility of benzene in water, probably erred in its conclusion that these anesthetics interfered with the interaction of benzene and water. Instead, the anesthetics simply decreased the vapor pressure of benzene, doing so in accordance with Raoult's Law.
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Despite a century of inquiry, where and how inhaled anesthetics act at a molecular level remain uncertain. On a gross anatomic level, the spinal cord, rather than the brain, probably mediates the capacity of inhaled anesthetic to produce immobility, but no specific effect on receptors or ion channels presently explains immobility (1). Anesthetics inhibit light production by the water-soluble protein firefly luciferase, indicating a potential effect on protein conformation (2,3). The enormous number of small molecules with anesthetic properties, and the variety of nonanesthetic biologic effects they produce, suggest nonspecific effects rather than binding to a receptor as the basis for anesthesia. The Meyer-Overton correlation predicts a site of action in the lipid portion of membranes, and much effort linked lipid solubility with various actions that might account for anesthesia (4,5). A second physical theory postulated that anesthetics serve as a nidus leading to increased organization of water (clathrate formation) in the aqueous phase surrounding proteins and membranes (6). None of these theories have significant present support.
All inhaled anesthetics are small polar or polarizable molecules that concentrate at water/nonpolar interfaces (7–9). This finding points to the interface between nonpolar parts of membranes and the surrounding aqueous solution as the site of action. The interface includes "membrane bound" enzymes and other proteins whose nonpolar components reside in the membrane lipids and whose polar and ionic components contact water (10).
Nuclear magnetic resonance studies combined with large scale molecular simulations demonstrate that halothane alters the dynamics of the well-defined ion channel gramicidin A in a model membrane (11–13). Halothane molecules are located near tryptophan residues at the channel/lipid/water interface responsible for anchoring the channel in the membrane. Halothane may disrupt the interaction between tryptophan and surrounding lipids by forming indole-fluorine hydrogen bonds. Such an effect can be interpreted as an increase in the aqueous solubility of tryptophan.
Whether inhaled anesthetics interact with other membrane components at the interface to alter protein conformation or dynamics remains unknown. The polar nature (or potential for polarization) of inhaled anesthetics is important because it allows the anesthetic to straddle the interface between water and lipid phases. This suggests the possibility that anesthetics may increase the solubility of nonpolar compounds in water by a "salting-in" effect. This sort of phenomenon has been demonstrated for urea which can denature globular proteins by reducing hydrophobic forces responsible for tertiary protein structure (14), but only at concentrations far exceeding those needed to produce anesthesia.
We hypothesized that anesthetics alter the hydrophobicity of crucial membrane proteins. Small changes in solubility of nonpolar components can produce large changes in the thermodynamics controlling protein conformation (10,15–17). One of the authors previously tested this hypothesis in a simple model, finding that nitrous oxide, cyclopropane, and ethyl chloride decreased the solubility of benzene in water in proportion to their potency (18). We retested the hypothesis using a modern anesthetic (isoflurane), different test compounds (methane and toluene), and relatively low concentrations of the test compounds. We used a dilute solution to avoid the dimer formation that can occur with concentrated solutions of benzene (19). Finally, we tested the possibility that our previous findings resulted from a dilution of the benzene by dissolved anesthetics. Such a dilution would decrease the vapor pressure of the benzene and thereby decrease the amount of benzene that would be found in water equilibrated with such benzene.
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The experiment determined how graded concentrations of isoflurane affect the solubility of methane and toluene in normal saline. To minimize contaminants, glassware was washed with Contrad70 (Decon Labs, King of Prussia, PA) 2.5% by volume, rinsed with hot water several times, and then rinsed with organic-free distilled water and dried.
Two empty size E gas cylinders were purged repeatedly with air to remove all trace of gaseous contaminants. Then the tanks were evacuated with a vacuum pump. Sufficient pure methane (99.0%+, Sigma-Aldrich, St. Louis, MO) and nitrogen were added to one tank to give a calculated methane concentration of 9.90%. Exactly 400 µL of toluene (Sigma Chemicals Company, St. Louis, MO) was added to the second tank, which was then pressurized to 281 psig with pure nitrogen, yielding a 0.10% toluene concentration.
Glass syringes (50 mL) were loaded with slightly more than 12 or 20 mL of saline (0.9% NaCl; Baxter Healthcare Corporation, Deerfield, IL). The lower volume was used in the toluene experiments because of toluene's relatively high saline:gas partition coefficient (PC) (approximately 2.5). Gas from the appropriate cylinder was then bled into the syringe to give a total volume of 40 mL. The syringe gas-liquid contents were mixed by rolling, and the gas was ejected. Fresh gas from the E cylinder was introduced and the process repeated three times. The syringes then were placed in a chamber heated to 37°C for at least 1 h. The chamber was equipped with a device that rolled each syringe around its long axis every 15 s. This "rolling tonometer" provided equilibration without introducing gas bubbles into the liquid phase.
Methane (or toluene) concentrations in the gas phase of each syringe were determined with a gas chromatograph (GC) (GOW-MAC Series 580, Bethlehem, PA) using a 10-ft long stainless steel column with SF-96 on Chromasorb WHP and a flame ionization detector (FID) and a gas sampling loop to provide consistent injection volumes. Serial dilutions of both methane and toluene produced linear plots in the concentration range used in the study. A calibration curve revealed some saturation of the FID at isoflurane concentrations exceeding 10%. A polynomial fit to the data provided an estimate of isoflurane concentration from FID output. The GC output was stable over time with a coefficient of variation for determinations of a given concentration of <0.5%.
After the initial measurement, all gas in each syringe was expelled along with sufficient saline to bring the remaining saline volume to exactly 12 mL (toluene) or 20 mL (methane). We then added exactly 28 or 20 mL of room air. The concentration of the test compound was determined in the gas phase after a second hour in the tonometer. The process was repeated a third time.
This approach gives the initial gas-phase concentration (Co) and two subsequent values (C1 and C2) that reflect the amount of test compound that was dissolved in the saline phase and subsequently diffused back into the gas phase during the second and third equilibrations. Concentration declines in a first-order fashion, so a plot of the natural log of concentration versus "step" is linear, and this plot was used to precisely estimate Co and C1. The saline:gas PC was calculated as:
|
|
where VG equals the volume of room air, and VS equals the volume of saline.
The saline:gas PC was determined four times (i.e., in four syringes) in the absence of anesthetic and (separately) in the presence of approximately 1%, 5%, and 15% isoflurane. Gas from the standard methane and toluene tanks was directed through a standard isoflurane vaporizer (1% and 5% isoflurane concentrations) or through a glass bubbler (15%) to get appropriate concentrations of both isoflurane and the test compound to insert as the initial gas phase of each syringe. For the second and third steps, isoflurane at the test concentration was prepared by directing air through the vaporizer. The resulting gas containing air and isoflurane was used to replace the gas expelled during the second and third steps.
We next tested the effect of cyclopropane or ethyl chloride on the vapor pressure of benzene at room temperature (23°C). We added 8–10 mL of benzene to twelve 50-mL glass syringes. To four syringes we added 30 mL of air (benzene syringes). To four we added 100% cyclopropane (cyclopropane syringes). Initially, the benzene absorbed the bubbles of cyclopropane, but after a few minutes, bubbling produced an increasing gas volume within the syringe despite continued shaking of the mixture. When the gas volume of cyclopropane reached 35 mL, we stopped the inflow, ejected the gas, and replaced it with 35 mL of fresh cyclopropane. To the remaining four syringes we added ethyl chloride (ethyl chloride syringes). As with cyclopropane, the benzene absorbed the inflowing bubbles of ethyl chloride, but unlike cyclopropane, the absorption continued for several minutes. The volume of fluid increased from the initial 8–10 mL to as much as 14–15 mL, and the mixture became warm to touch. Eventually, a gas volume of 35 mL was obtained, ejected, and replaced.
The 12 syringes then were rotated to equilibrate the gas phase with the benzene and the benzene-anesthetic mixtures. The gas volumes remained constant with the benzene syringes but decreased slightly with the cyclopropane syringes. Cyclopropane was added to these syringes to restore the gas volume to 35 mL. The gas volumes decreased considerably in the ethyl chloride syringes as the temperature of the solution decreased and solubility of ethyl chloride in benzene increased. The gas volume in these syringes was repeatedly increased to 35 mL by addition of ethyl chloride. The rotation of the syringes proceeded for at least an hour, longer for the ethyl chloride syringes.
After completing equilibration of the gas and liquid phases, a 20 mL gas sample was taken from each syringe into a 100 mL syringe and the gas phase immediately enlarged to 100 mL by drawing in 80 mL. The resulting gas mixture was diluted two more times so that the final gas mixture was diluted by a factor of 20 (i.e., to 5% of the initial concentration). This gas was analyzed by gas chromatography (see above). The GC was calibrated with primary standards that produced peak heights in the range found in the gas samples.
Having analyzed the gas samples in the cyclopropane and ethyl chloride syringes, we removed 0.24 mL of liquid from each of the residual liquid phases and transferred this 0.24 mL to evacuated 600-mL flasks. After allowing an hour to complete vaporization of the liquid, the resulting gas concentration was measured using gas chromatography. From the relative concentrations, we calculated the molar concentrations initially found in the liquid (e.g., if the benzene concentration equaled the ethyl chloride concentration, then the molar concentration of benzene constituted 50% of the total; if it were a third of the ethyl chloride concentration, then the molar concentration would constitute 25% of the total).
Output of the FID was attenuated to give peaks on a strip-chart recorder in the 40–70-division range under most circumstances. The baseline was subtracted from peak height and multiplied by attenuation to give FID output, which was converted to concentration using the calibration curves discussed above. Saline:gas PCs were calculated and plotted against isoflurane concentration. Linear regression analysis (Graphpad V 4.01, San Diego, CA) was used to determine if the slope of the relationship differed from zero. One-way ANOVA was also used to determine differences between saline:gas PCs at different isoflurane concentrations. When the overall F test was significant, individual comparisons were performed with Newman-Keuls posttest.
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
ANOVA revealed no effect of isoflurane on toluene PC (Table 1). In contrast, the methane PC determined at the highest isoflurane concentration exceeded that determined in the absence of isoflurane (P < 0.05).
|
Plots of PCs versus measured isoflurane concentrations illustrate the data for methane (Fig. 1) and toluene (Fig. 2). In the plots (but not for the analysis), control values with zero anesthetic concentration are assigned a value of 0.1% to allow the use of a logarithmic plot. Linear regression for methane PC versus isoflurane revealed a positive slope that significantly exceeded zero (P < 0.001). The linear best-fit line in Figure 1 presents a curve because of the logarithmic scale.
|
|
In the benzene syringes, the concentration of benzene was 12.50% ± 0.14% (mean ± sd). The addition of cyclopropane significantly (P < 0.02) decreased the benzene concentration to 11.30% ± 0.78%. Addition of ethyl chloride decreased the benzene concentration to 6.22% ± 0.45% (P < 0.001). The respective molar percentages of benzene were benzene 100.0% ± 1.1%; cyclopropane (one flask lost) 88.1% ± 0.2%; and ethyl chloride 43.2% ± 0.8%. The decrease in benzene concentration was proportional to the decrease in the molar percentage of benzene (Fig. 3).
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In contrast to our previous results with benzene, we find that clinically relevant concentrations of isoflurane do not affect the saline:gas PC of either methane or toluene. Fifteen percent isoflurane increased methane's PC significantly, but the absolute magnitude of the change was modest.
The methods used in this study follow closely those used by Coburn and Eger (20), and produced saline: gas PCs similar to those previously reported. We added methane and toluene to air to produce standard concentrations (9.9% and 0.1%, respectively) and assume that these dilute mixtures behaved as ideal gases. We did not include a bulk nonpolar phase such as olive oil because of the great solubility of toluene in olive oil; inadvertent inclusion of a tiny droplet of oil in the saline used for analysis would substantially (and falsely) increase the PC (21).
Methane and toluene were chosen as test nonpolar molecules because they form the "R" groups of alanine and phenylalanine, two hydrophobic amino acids. Methane is the smallest alkane and is sparingly soluble in water. Toluene is larger and has a planar structure with a cloud of 
electrons above and below the plane. Because of the interaction of these electrons with water, toluene is about 75 times more soluble in saline than is methane.
Our results do not support the notion that inhaled anesthetics act by changing the solubility of nonpolar parts of proteins and other biologic macromolecules in saline. The only effect we saw was a 15%–20% increase in the PC for methane at 15% isoflurane, a concentration far exceeding that producing clinical anesthesia.
On the other hand, anesthetic molecules concentrate at membrane/water interfaces, and some evidence points to the membrane/water interface as the site at which inhaled anesthetics act (9). The interface concentration could exceed that in the near-by bulk nonpolar phase by 5–10-fold. The concentration of anesthetic molecules occurs because of their weakly polar, amphiphilic nature prompts them to straddle the interface between water and nonpolar regions. Thus, we speculate that the actual anesthetic concentration at the site of action in a biologic system may lie in the range where we noticed an effect of isoflurane on methane solubility. Although a 15%–20% increase in methane solubility seems relatively small, such a change could profoundly affect the conformation of a protein with the usual quantity of nonpolar groups because hydrophobicity of these groups is such a strong thermodynamic component determining protein structure (17).
The enhanced solubility seen with methane is consistent with several facets of anesthetic effects on protein chemistry. The destabilization of folded proteins with decreasing temperature provides additional evidence of the importance of hydrophobic interactions in determining tertiary structure (22) and is consistent with the decrease in minimum alveolar anesthetic concentration as temperature decreases (23). Urea, a small polar molecule similar to many anesthetics, is used to "denature" (unfold) proteins and acts by increasing the solubility of nonpolar protein components (14).
The current results differ from those obtained by one of us (C.B.) in a previous study of the effect of nitrous oxide, cyclopropane and ethyl chloride on the solubility of benzene in water (18). These anesthetics decreased the solubility of benzene in water in proportion to their potency. Our results showing that solution of cyclopropane and ethyl chloride decreases the vapor pressure of benzene explains the previous findings. Benzene was present as a bulk liquid phase in the previous study and the vapor pressure of benzene was assumed to be constant. However, the present results show that this assumption was incorrect because the anesthetics would dissolve in the benzene, thereby diluting the benzene and decreasing its vapor pressure, and thus decreasing the amount of benzene dissolved in water. Indeed, the decreases in vapor pressure found in the present study approximate the decreases in solubility found in the previous study. And the decreases in vapor pressure in the present study were shown to correlate with the dilution of the benzene by the dissolved cyclopropane and ethyl chloride (Fig. 3). These findings, of course, are simply a result of Rauolt's Law. Finally, benzene may have formed dimers because of the concentrated solutions used in the previous experiment, an effect that would complicate both the chemistry and analysis (19). We avoided that possibility in the present experiment by the use of dilute solutions of methane and toluene.
Although the solubility of toluene in saline was unchanged statistically by isoflurane, there is a suggestion of a decline in PC at the highest isoflurane concentration. It is curious that isoflurane should have increased the solubility of methane and, if anything, decreased the solubility of toluene. The diametrically opposite effect of isoflurane on these two organic compounds suggests that the changes may have resulted from random effects and further strengthens our conclusion of "no effect." However, a suggestion that opposite effects should not occur also assumes a unitary theory of narcosis, which may be incorrect.
The small number of compounds studied in the present report limits the strength of the negative result. Different results might have been obtained with different anesthetic or test compounds. A change in the saline:gas PC for toluene might have been seen at still higher isoflurane concentrations, but our highest concentration, 15%, already vastly exceeds any clinically relevant concentration. Finally, measurements were made at a single temperature, 37°C. The lack of a range of temperatures precludes further thermodynamic analysis of the methane data.
In conclusion, isoflurane produces a weak antihydrophobic effect in aqueous solutions of methane, but not toluene, at concentrations far above the clinical level. These results argue against alteration in the aqueous solubility of nonpolar groups on macromolecules as a primary component of the mechanism of action of inhaled anesthetics.
|
Footnotes |
|---|
Accepted for publication August 10, 2007.
Supported in part by NIH grant 1PO1GM47818.
Dr. Eger is a paid consultant to Baxter Healthcare Corp. Baxter Healthcare Corp. donated the isoflurane used in these studies.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
- Sonner JM, Antognini JF, Dutton RC, Flood P, Gray AT, Harris RA, Homanics GE, Kendig J, Orser B, Raines DE, Rampil IJ, Trudell J, Vissel B, Eger EI II. Inhaled anesthetics and immobility: mechanisms, mysteries, and minimum alveolar anesthetic concentration. Anesth Analg 2003;97:718–40
[Abstract/Free Full Text] - Franks NP, Jenkins A, Conti E, Lieb WR, Brick P. Structural basis for the inhibition of firefly luciferase by a general anesthetic. Biophys J 1998;75:2205–11[Web of Science][Medline]
- Takehara K, Kamaya H, Ueda I. Inhibition of firefly luciferase by alkane analogues. Biochim Biophys Acta 2005;1721:124–9[Medline]
- Matubayasi N, Ueda I. Is membrane expansion relevant to anesthesia? Mean excess volume. Anesthesiology 1983;59:541–6[Medline]
- van den Brink-van der Laan, Chupin V, Killian JA, de KB. Small alcohols destabilize the KcsA tetramer via their effect on the membrane lateral pressure. Biochemistry 2004;43:5937–42
- Pauling L. A molecular theory of general anesthesia. Science 1961;134:15–21
[Free Full Text] - Wilson MA, Pohorille A. Molecular dynamics of a water-lipid bilayer interface. J Am Chem Soc 1994;116:1490–501[Web of Science][Medline]
- Xu Y, Tang P. Amphiphilic sites for general anesthetic action? Evidence from 129Xe-[1H] intermolecular nuclear Overhauser effects. Biochim Biophys Acta 1997;1323:154–62[Medline]
- Pohorille A, Wilson MA, New MH, Chipot C. Concentrations of anesthetics across the water-membrane interface; the Meyer-Overton hypothesis revisited. Toxicol Lett 1998;100–101:421–30
- Pratt LR, Pohorille A. Hydrophobic effects and modeling of biophysical aqueous solution interfaces. Chem Rev 2002; 102:2671–92[Web of Science][Medline]
- Tang P, Eckenhoff RG, Xu Y. General anesthetic binding to gramicidin A: the structural requirements. Biophys J 2000;78: 1804–9[Web of Science][Medline]
- Tang P, Xu Y. Large-scale molecular dynamics simulations of general anesthetic effects on the ion channel in the fully hydrated membrane: the implication of molecular mechanisms of general anesthesia. Proc Natl Acad Sci U S A 2002;99:16035–40
[Abstract/Free Full Text] - Liu Z, Xu Y, Tang P. Molecular dynamics simulations of C2F6 effects on gramicidin A: implications of the mechanisms of general anesthesia. Biophys J 2005;88:3784–91[Web of Science][Medline]
- Bresloe R. Hydrophobic and antihydrophobic effects on organic reactions in aqueous solution. In: Cramer CJ, Truhlar DG, eds. Structure and reactivity in aqueous solution. Washington, DC: American Chemical Society, 1994:291–302
- Southall NT, Dill KA, Haymet ADJ. A view of the hydrophobic effect. J Phys Chem B 2002;106:521–33
- Chandler D. Interfaces and the driving force of hydrophobic assembly. Nature 2005;437:640–7[Medline]
- Alonso DOV, Dill KA. Solvent denaturation and stabilization of globular proteins. Biochemistry 1991;30:5974–85[Medline]
- Buffington C, Turndorf H. Anesthetics alter the solubility of nonpolar compounds in water. Bull N Y Acad Med 1976; 52:838–41[Web of Science][Medline]
- Tucker EE, Christian SD. A prototype hydrophobic interaction, the dimerization of benzene in water. J Phys Chem 1979; 83:426–7[Web of Science]
- Coburn CM, Eger EI II. The partial pressure of isoflurane or halothane does not affect their solubility in rabbit blood or brain or human brain: inhaled anesthetics obey Henry's Law. Anesth Analg 1986;65:960–2
[Abstract/Free Full Text] - Ionescu P, Eger EI II, Trudell J. Direct determination of oil/saline partition coefficients. Anesth Analg 1994;79:1056–8
[Abstract/Free Full Text] - Hummer G, Garde S, Garcia AE, Pohorille A, Pratt LR. An information theory model of hydrophobic interactions. Proc Natl Acad Sci U S A 1996;93:8951–5
[Abstract/Free Full Text] - Eger EI II, Saidman LJ, Brandstater B. Temperature dependence of halothane and cyclopropane anesthesia in dogs: correlation with some theories of anesthetic action. Anesthesiology 1965;26:764–70[Web of Science][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
